821 files changed, 94853 insertions, 33215 deletions

diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index f8642759116..f5120046ff8 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -20,6 +20,10 @@ menuconfig DRM
 	  details.  You should also select and configure AGP
 	  (/dev/agpgart) support if it is available for your platform.
 
+config DRM_MIPI_DSI
+	bool
+	depends on DRM
+
 config DRM_USB
 	tristate
 	depends on DRM
@@ -79,6 +83,8 @@ config DRM_KMS_CMA_HELPER
 
 source "drivers/gpu/drm/i2c/Kconfig"
 
+source "drivers/gpu/drm/bridge/Kconfig"
+
 config DRM_TDFX
 	tristate "3dfx Banshee/Voodoo3+"
 	depends on DRM && PCI
@@ -188,6 +194,10 @@ source "drivers/gpu/drm/tilcdc/Kconfig"
 
 source "drivers/gpu/drm/qxl/Kconfig"
 
+source "drivers/gpu/drm/bochs/Kconfig"
+
 source "drivers/gpu/drm/msm/Kconfig"
 
 source "drivers/gpu/drm/tegra/Kconfig"
+
+source "drivers/gpu/drm/panel/Kconfig"
diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index cc08b845f96..dd2ba426974 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -13,15 +13,18 @@ drm-y       :=	drm_auth.o drm_buffer.o drm_bufs.o drm_cache.o \
 		drm_crtc.o drm_modes.o drm_edid.o \
 		drm_info.o drm_debugfs.o drm_encoder_slave.o \
 		drm_trace_points.o drm_global.o drm_prime.o \
-		drm_rect.o drm_vma_manager.o drm_flip_work.o
+		drm_rect.o drm_vma_manager.o drm_flip_work.o \
+		drm_modeset_lock.o
 
 drm-$(CONFIG_COMPAT) += drm_ioc32.o
 drm-$(CONFIG_DRM_GEM_CMA_HELPER) += drm_gem_cma_helper.o
 drm-$(CONFIG_PCI) += ati_pcigart.o
+drm-$(CONFIG_DRM_PANEL) += drm_panel.o
 
 drm-usb-y   := drm_usb.o
 
-drm_kms_helper-y := drm_crtc_helper.o drm_dp_helper.o
+drm_kms_helper-y := drm_crtc_helper.o drm_dp_helper.o drm_probe_helper.o \
+		drm_plane_helper.o
 drm_kms_helper-$(CONFIG_DRM_LOAD_EDID_FIRMWARE) += drm_edid_load.o
 drm_kms_helper-$(CONFIG_DRM_KMS_FB_HELPER) += drm_fb_helper.o
 drm_kms_helper-$(CONFIG_DRM_KMS_CMA_HELPER) += drm_fb_cma_helper.o
@@ -31,6 +34,7 @@ obj-$(CONFIG_DRM_KMS_HELPER) += drm_kms_helper.o
 CFLAGS_drm_trace_points.o := -I$(src)
 
 obj-$(CONFIG_DRM)	+= drm.o
+obj-$(CONFIG_DRM_MIPI_DSI) += drm_mipi_dsi.o
 obj-$(CONFIG_DRM_USB)   += drm_usb.o
 obj-$(CONFIG_DRM_TTM)	+= ttm/
 obj-$(CONFIG_DRM_TDFX)	+= tdfx/
@@ -56,6 +60,9 @@ obj-$(CONFIG_DRM_SHMOBILE) +=shmobile/
 obj-$(CONFIG_DRM_OMAP)	+= omapdrm/
 obj-$(CONFIG_DRM_TILCDC)	+= tilcdc/
 obj-$(CONFIG_DRM_QXL) += qxl/
+obj-$(CONFIG_DRM_BOCHS) += bochs/
 obj-$(CONFIG_DRM_MSM) += msm/
 obj-$(CONFIG_DRM_TEGRA) += tegra/
 obj-y			+= i2c/
+obj-y			+= panel/
+obj-y			+= bridge/
diff --git a/drivers/gpu/drm/armada/Kconfig b/drivers/gpu/drm/armada/Kconfig
index 40d371521fe..50ae88ad4d7 100644
--- a/drivers/gpu/drm/armada/Kconfig
+++ b/drivers/gpu/drm/armada/Kconfig
@@ -5,6 +5,7 @@ config DRM_ARMADA
 	select FB_CFB_COPYAREA
 	select FB_CFB_IMAGEBLIT
 	select DRM_KMS_HELPER
+	select DRM_KMS_FB_HELPER
 	help
 	  Support the "LCD" controllers found on the Marvell Armada 510
 	  devices.  There are two controllers on the device, each controller
diff --git a/drivers/gpu/drm/armada/armada_crtc.c b/drivers/gpu/drm/armada/armada_crtc.c
index d8e398275ca..81c34f949df 100644
--- a/drivers/gpu/drm/armada/armada_crtc.c
+++ b/drivers/gpu/drm/armada/armada_crtc.c
@@ -478,11 +478,12 @@ static int armada_drm_crtc_mode_set(struct drm_crtc *crtc,
 	unsigned i;
 	bool interlaced;
 
-	drm_framebuffer_reference(crtc->fb);
+	drm_framebuffer_reference(crtc->primary->fb);
 
 	interlaced = !!(adj->flags & DRM_MODE_FLAG_INTERLACE);
 
-	i = armada_drm_crtc_calc_fb(dcrtc->crtc.fb, x, y, regs, interlaced);
+	i = armada_drm_crtc_calc_fb(dcrtc->crtc.primary->fb,
+				    x, y, regs, interlaced);
 
 	rm = adj->crtc_hsync_start - adj->crtc_hdisplay;
 	lm = adj->crtc_htotal - adj->crtc_hsync_end;
@@ -567,10 +568,10 @@ static int armada_drm_crtc_mode_set(struct drm_crtc *crtc,
 	}
 
 	val = CFG_GRA_ENA | CFG_GRA_HSMOOTH;
-	val |= CFG_GRA_FMT(drm_fb_to_armada_fb(dcrtc->crtc.fb)->fmt);
-	val |= CFG_GRA_MOD(drm_fb_to_armada_fb(dcrtc->crtc.fb)->mod);
+	val |= CFG_GRA_FMT(drm_fb_to_armada_fb(dcrtc->crtc.primary->fb)->fmt);
+	val |= CFG_GRA_MOD(drm_fb_to_armada_fb(dcrtc->crtc.primary->fb)->mod);
 
-	if (drm_fb_to_armada_fb(dcrtc->crtc.fb)->fmt > CFG_420)
+	if (drm_fb_to_armada_fb(dcrtc->crtc.primary->fb)->fmt > CFG_420)
 		val |= CFG_PALETTE_ENA;
 
 	if (interlaced)
@@ -608,7 +609,7 @@ static int armada_drm_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 	struct armada_regs regs[4];
 	unsigned i;
 
-	i = armada_drm_crtc_calc_fb(crtc->fb, crtc->x, crtc->y, regs,
+	i = armada_drm_crtc_calc_fb(crtc->primary->fb, crtc->x, crtc->y, regs,
 				    dcrtc->interlaced);
 	armada_reg_queue_end(regs, i);
 
@@ -616,7 +617,7 @@ static int armada_drm_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 	wait_event(dcrtc->frame_wait, !dcrtc->frame_work);
 
 	/* Take a reference to the new fb as we're using it */
-	drm_framebuffer_reference(crtc->fb);
+	drm_framebuffer_reference(crtc->primary->fb);
 
 	/* Update the base in the CRTC */
 	armada_drm_crtc_update_regs(dcrtc, regs);
@@ -637,7 +638,7 @@ static void armada_drm_crtc_disable(struct drm_crtc *crtc)
 	struct armada_crtc *dcrtc = drm_to_armada_crtc(crtc);
 
 	armada_drm_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
-	armada_drm_crtc_finish_fb(dcrtc, crtc->fb, true);
+	armada_drm_crtc_finish_fb(dcrtc, crtc->primary->fb, true);
 
 	/* Power down most RAMs and FIFOs */
 	writel_relaxed(CFG_PDWN256x32 | CFG_PDWN256x24 | CFG_PDWN256x8 |
@@ -678,6 +679,7 @@ static void armada_load_cursor_argb(void __iomem *base, uint32_t *pix,
 				       base + LCD_SPU_SRAM_WRDAT);
 			writel_relaxed(addr | SRAM_WRITE,
 				       base + LCD_SPU_SRAM_CTRL);
+			readl_relaxed(base + LCD_SPU_HWC_OVSA_HPXL_VLN);
 			addr += 1;
 			if ((addr & 0x00ff) == 0)
 				addr += 0xf00;
@@ -904,7 +906,7 @@ static int armada_drm_crtc_page_flip(struct drm_crtc *crtc,
 	int ret;
 
 	/* We don't support changing the pixel format */
-	if (fb->pixel_format != crtc->fb->pixel_format)
+	if (fb->pixel_format != crtc->primary->fb->pixel_format)
 		return -EINVAL;
 
 	work = kmalloc(sizeof(*work), GFP_KERNEL);
@@ -912,7 +914,7 @@ static int armada_drm_crtc_page_flip(struct drm_crtc *crtc,
 		return -ENOMEM;
 
 	work->event = event;
-	work->old_fb = dcrtc->crtc.fb;
+	work->old_fb = dcrtc->crtc.primary->fb;
 
 	i = armada_drm_crtc_calc_fb(fb, crtc->x, crtc->y, work->regs,
 				    dcrtc->interlaced);
@@ -941,7 +943,7 @@ static int armada_drm_crtc_page_flip(struct drm_crtc *crtc,
 	 * will _not_ drop that reference on successful return from this
 	 * function.  Simply mark this new framebuffer as the current one.
 	 */
-	dcrtc->crtc.fb = fb;
+	dcrtc->crtc.primary->fb = fb;
 
 	/*
 	 * Finally, if the display is blanked, we won't receive an
diff --git a/drivers/gpu/drm/armada/armada_drv.c b/drivers/gpu/drm/armada/armada_drv.c
index 62d0ff3efdd..8ab3cd1a8cd 100644
--- a/drivers/gpu/drm/armada/armada_drv.c
+++ b/drivers/gpu/drm/armada/armada_drv.c
@@ -68,15 +68,7 @@ void __armada_drm_queue_unref_work(struct drm_device *dev,
 {
 	struct armada_private *priv = dev->dev_private;
 
-	/*
-	 * Yes, we really must jump through these hoops just to store a
-	 * _pointer_ to something into the kfifo.  This is utterly insane
-	 * and idiotic, because it kfifo requires the _data_ pointed to by
-	 * the pointer const, not the pointer itself.  Not only that, but
-	 * you have to pass a pointer _to_ the pointer you want stored.
-	 */
-	const struct drm_framebuffer *silly_api_alert = fb;
-	WARN_ON(!kfifo_put(&priv->fb_unref, &silly_api_alert));
+	WARN_ON(!kfifo_put(&priv->fb_unref, fb));
 	schedule_work(&priv->fb_unref_work);
 }
 
@@ -128,6 +120,7 @@ static int armada_drm_load(struct drm_device *dev, unsigned long flags)
 		return -ENOMEM;
 	}
 
+	platform_set_drvdata(dev->platformdev, dev);
 	dev->dev_private = priv;
 
 	/* Get the implementation specific driver data. */
@@ -180,7 +173,7 @@ static int armada_drm_load(struct drm_device *dev, unsigned long flags)
 	if (ret)
 		goto err_kms;
 
-	ret = drm_irq_install(dev);
+	ret = drm_irq_install(dev, platform_get_irq(dev->platformdev, 0));
 	if (ret)
 		goto err_kms;
 
@@ -381,7 +374,7 @@ static int armada_drm_probe(struct platform_device *pdev)
 
 static int armada_drm_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&armada_drm_driver, pdev);
+	drm_put_dev(platform_get_drvdata(pdev));
 	return 0;
 }
 
@@ -409,7 +402,7 @@ static struct platform_driver armada_drm_platform_driver = {
 
 static int __init armada_drm_init(void)
 {
-	armada_drm_driver.num_ioctls = DRM_ARRAY_SIZE(armada_ioctls);
+	armada_drm_driver.num_ioctls = ARRAY_SIZE(armada_ioctls);
 	return platform_driver_register(&armada_drm_platform_driver);
 }
 module_init(armada_drm_init);
diff --git a/drivers/gpu/drm/armada/armada_fbdev.c b/drivers/gpu/drm/armada/armada_fbdev.c
index 948cb14c561..fd166f532ab 100644
--- a/drivers/gpu/drm/armada/armada_fbdev.c
+++ b/drivers/gpu/drm/armada/armada_fbdev.c
@@ -181,10 +181,8 @@ void armada_fbdev_lastclose(struct drm_device *dev)
 {
 	struct armada_private *priv = dev->dev_private;
 
-	drm_modeset_lock_all(dev);
 	if (priv->fbdev)
-		drm_fb_helper_restore_fbdev_mode(priv->fbdev);
-	drm_modeset_unlock_all(dev);
+		drm_fb_helper_restore_fbdev_mode_unlocked(priv->fbdev);
 }
 
 void armada_fbdev_fini(struct drm_device *dev)
diff --git a/drivers/gpu/drm/armada/armada_gem.c b/drivers/gpu/drm/armada/armada_gem.c
index 887816f4347..bb9b642d848 100644
--- a/drivers/gpu/drm/armada/armada_gem.c
+++ b/drivers/gpu/drm/armada/armada_gem.c
@@ -433,7 +433,6 @@ armada_gem_prime_map_dma_buf(struct dma_buf_attachment *attach,
 
 	if (dobj->obj.filp) {
 		struct address_space *mapping;
-		gfp_t gfp;
 		int count;
 
 		count = dobj->obj.size / PAGE_SIZE;
@@ -441,12 +440,11 @@ armada_gem_prime_map_dma_buf(struct dma_buf_attachment *attach,
 			goto free_sgt;
 
 		mapping = file_inode(dobj->obj.filp)->i_mapping;
-		gfp = mapping_gfp_mask(mapping);
 
 		for_each_sg(sgt->sgl, sg, count, i) {
 			struct page *page;
 
-			page = shmem_read_mapping_page_gfp(mapping, i, gfp);
+			page = shmem_read_mapping_page(mapping, i);
 			if (IS_ERR(page)) {
 				num = i;
 				goto release;
diff --git a/drivers/gpu/drm/ast/Makefile b/drivers/gpu/drm/ast/Makefile
index 8df4f284ee2..171aa0622b6 100644
--- a/drivers/gpu/drm/ast/Makefile
+++ b/drivers/gpu/drm/ast/Makefile
@@ -4,6 +4,6 @@
 
 ccflags-y := -Iinclude/drm
 
-ast-y := ast_drv.o ast_main.o ast_mode.o ast_fb.o ast_ttm.o ast_post.o
+ast-y := ast_drv.o ast_main.o ast_mode.o ast_fb.o ast_ttm.o ast_post.o ast_dp501.o
 
-obj-$(CONFIG_DRM_AST) := ast.o
\ No newline at end of file
+obj-$(CONFIG_DRM_AST) := ast.o
diff --git a/drivers/gpu/drm/ast/ast_dp501.c b/drivers/gpu/drm/ast/ast_dp501.c
new file mode 100644
index 00000000000..5da4b62285f
--- /dev/null
+++ b/drivers/gpu/drm/ast/ast_dp501.c
@@ -0,0 +1,410 @@
+
+#include <linux/firmware.h>
+#include <drm/drmP.h>
+#include "ast_drv.h"
+MODULE_FIRMWARE("ast_dp501_fw.bin");
+
+int ast_load_dp501_microcode(struct drm_device *dev)
+{
+	struct ast_private *ast = dev->dev_private;
+	static char *fw_name = "ast_dp501_fw.bin";
+	int err;
+	err = request_firmware(&ast->dp501_fw, fw_name, dev->dev);
+	if (err)
+		return err;
+
+	return 0;
+}
+
+static void send_ack(struct ast_private *ast)
+{
+	u8 sendack;
+	sendack = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x9b, 0xff);
+	sendack |= 0x80;
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x9b, 0x00, sendack);
+}
+
+static void send_nack(struct ast_private *ast)
+{
+	u8 sendack;
+	sendack = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x9b, 0xff);
+	sendack &= ~0x80;
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x9b, 0x00, sendack);
+}
+
+static bool wait_ack(struct ast_private *ast)
+{
+	u8 waitack;
+	u32 retry = 0;
+	do {
+		waitack = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd2, 0xff);
+		waitack &= 0x80;
+		udelay(100);
+	} while ((!waitack) && (retry++ < 1000));
+
+	if (retry < 1000)
+		return true;
+	else
+		return false;
+}
+
+static bool wait_nack(struct ast_private *ast)
+{
+	u8 waitack;
+	u32 retry = 0;
+	do {
+		waitack = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd2, 0xff);
+		waitack &= 0x80;
+		udelay(100);
+	} while ((waitack) && (retry++ < 1000));
+
+	if (retry < 1000)
+		return true;
+	else
+		return false;
+}
+
+static void set_cmd_trigger(struct ast_private *ast)
+{
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x9b, ~0x40, 0x40);
+}
+
+static void clear_cmd_trigger(struct ast_private *ast)
+{
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x9b, ~0x40, 0x00);
+}
+
+#if 0
+static bool wait_fw_ready(struct ast_private *ast)
+{
+	u8 waitready;
+	u32 retry = 0;
+	do {
+		waitready = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd2, 0xff);
+		waitready &= 0x40;
+		udelay(100);
+	} while ((!waitready) && (retry++ < 1000));
+
+	if (retry < 1000)
+		return true;
+	else
+		return false;
+}
+#endif
+
+static bool ast_write_cmd(struct drm_device *dev, u8 data)
+{
+	struct ast_private *ast = dev->dev_private;
+	int retry = 0;
+	if (wait_nack(ast)) {
+		send_nack(ast);
+		ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x9a, 0x00, data);
+		send_ack(ast);
+		set_cmd_trigger(ast);
+		do {
+			if (wait_ack(ast)) {
+				clear_cmd_trigger(ast);
+				send_nack(ast);
+				return true;
+			}
+		} while (retry++ < 100);
+	}
+	clear_cmd_trigger(ast);
+	send_nack(ast);
+	return false;
+}
+
+static bool ast_write_data(struct drm_device *dev, u8 data)
+{
+	struct ast_private *ast = dev->dev_private;
+
+	if (wait_nack(ast)) {
+		send_nack(ast);
+		ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x9a, 0x00, data);
+		send_ack(ast);
+		if (wait_ack(ast)) {
+			send_nack(ast);
+			return true;
+		}
+	}
+	send_nack(ast);
+	return false;
+}
+
+#if 0
+static bool ast_read_data(struct drm_device *dev, u8 *data)
+{
+	struct ast_private *ast = dev->dev_private;
+	u8 tmp;
+
+	*data = 0;
+
+	if (wait_ack(ast) == false)
+		return false;
+	tmp = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd3, 0xff);
+	*data = tmp;
+	if (wait_nack(ast) == false) {
+		send_nack(ast);
+		return false;
+	}
+	send_nack(ast);
+	return true;
+}
+
+static void clear_cmd(struct ast_private *ast)
+{
+	send_nack(ast);
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x9a, 0x00, 0x00);
+}
+#endif
+
+void ast_set_dp501_video_output(struct drm_device *dev, u8 mode)
+{
+	ast_write_cmd(dev, 0x40);
+	ast_write_data(dev, mode);
+
+	msleep(10);
+}
+
+static u32 get_fw_base(struct ast_private *ast)
+{
+	return ast_mindwm(ast, 0x1e6e2104) & 0x7fffffff;
+}
+
+bool ast_backup_fw(struct drm_device *dev, u8 *addr, u32 size)
+{
+	struct ast_private *ast = dev->dev_private;
+	u32 i, data;
+	u32 boot_address;
+
+	data = ast_mindwm(ast, 0x1e6e2100) & 0x01;
+	if (data) {
+		boot_address = get_fw_base(ast);
+		for (i = 0; i < size; i += 4)
+			*(u32 *)(addr + i) = ast_mindwm(ast, boot_address + i);
+		return true;
+	}
+	return false;
+}
+
+bool ast_launch_m68k(struct drm_device *dev)
+{
+	struct ast_private *ast = dev->dev_private;
+	u32 i, data, len = 0;
+	u32 boot_address;
+	u8 *fw_addr = NULL;
+	u8 jreg;
+
+	data = ast_mindwm(ast, 0x1e6e2100) & 0x01;
+	if (!data) {
+
+		if (ast->dp501_fw_addr) {
+			fw_addr = ast->dp501_fw_addr;
+			len = 32*1024;
+		} else if (ast->dp501_fw) {
+			fw_addr = (u8 *)ast->dp501_fw->data;
+			len = ast->dp501_fw->size;
+		}
+		/* Get BootAddress */
+		ast_moutdwm(ast, 0x1e6e2000, 0x1688a8a8);
+		data = ast_mindwm(ast, 0x1e6e0004);
+		switch (data & 0x03) {
+		case 0:
+			boot_address = 0x44000000;
+			break;
+		default:
+		case 1:
+			boot_address = 0x48000000;
+			break;
+		case 2:
+			boot_address = 0x50000000;
+			break;
+		case 3:
+			boot_address = 0x60000000;
+			break;
+		}
+		boot_address -= 0x200000; /* -2MB */
+
+		/* copy image to buffer */
+		for (i = 0; i < len; i += 4) {
+			data = *(u32 *)(fw_addr + i);
+			ast_moutdwm(ast, boot_address + i, data);
+		}
+
+		/* Init SCU */
+		ast_moutdwm(ast, 0x1e6e2000, 0x1688a8a8);
+
+		/* Launch FW */
+		ast_moutdwm(ast, 0x1e6e2104, 0x80000000 + boot_address);
+		ast_moutdwm(ast, 0x1e6e2100, 1);
+
+		/* Update Scratch */
+		data = ast_mindwm(ast, 0x1e6e2040) & 0xfffff1ff;		/* D[11:9] = 100b: UEFI handling */
+		data |= 0x800;
+		ast_moutdwm(ast, 0x1e6e2040, data);
+
+		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x99, 0xfc); /* D[1:0]: Reserved Video Buffer */
+		jreg |= 0x02;
+		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x99, jreg);
+	}
+	return true;
+}
+
+u8 ast_get_dp501_max_clk(struct drm_device *dev)
+{
+	struct ast_private *ast = dev->dev_private;
+	u32 boot_address, offset, data;
+	u8 linkcap[4], linkrate, linklanes, maxclk = 0xff;
+
+	boot_address = get_fw_base(ast);
+
+	/* validate FW version */
+	offset = 0xf000;
+	data = ast_mindwm(ast, boot_address + offset);
+	if ((data & 0xf0) != 0x10) /* version: 1x */
+		return maxclk;
+
+	/* Read Link Capability */
+	offset  = 0xf014;
+	*(u32 *)linkcap = ast_mindwm(ast, boot_address + offset);
+	if (linkcap[2] == 0) {
+		linkrate = linkcap[0];
+		linklanes = linkcap[1];
+		data = (linkrate == 0x0a) ? (90 * linklanes) : (54 * linklanes);
+		if (data > 0xff)
+			data = 0xff;
+		maxclk = (u8)data;
+	}
+	return maxclk;
+}
+
+bool ast_dp501_read_edid(struct drm_device *dev, u8 *ediddata)
+{
+	struct ast_private *ast = dev->dev_private;
+	u32 i, boot_address, offset, data;
+
+	boot_address = get_fw_base(ast);
+
+	/* validate FW version */
+	offset = 0xf000;
+	data = ast_mindwm(ast, boot_address + offset);
+	if ((data & 0xf0) != 0x10)
+		return false;
+
+	/* validate PnP Monitor */
+	offset = 0xf010;
+	data = ast_mindwm(ast, boot_address + offset);
+	if (!(data & 0x01))
+		return false;
+
+	/* Read EDID */
+	offset = 0xf020;
+	for (i = 0; i < 128; i += 4) {
+		data = ast_mindwm(ast, boot_address + offset + i);
+		*(u32 *)(ediddata + i) = data;
+	}
+
+	return true;
+}
+
+static bool ast_init_dvo(struct drm_device *dev)
+{
+	struct ast_private *ast = dev->dev_private;
+	u8 jreg;
+	u32 data;
+	ast_write32(ast, 0xf004, 0x1e6e0000);
+	ast_write32(ast, 0xf000, 0x1);
+	ast_write32(ast, 0x12000, 0x1688a8a8);
+
+	jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd0, 0xff);
+	if (!(jreg & 0x80)) {
+		/* Init SCU DVO Settings */
+		data = ast_read32(ast, 0x12008);
+		/* delay phase */
+		data &= 0xfffff8ff;
+		data |= 0x00000500;
+		ast_write32(ast, 0x12008, data);
+
+		if (ast->chip == AST2300) {
+			data = ast_read32(ast, 0x12084);
+			/* multi-pins for DVO single-edge */
+			data |= 0xfffe0000;
+			ast_write32(ast, 0x12084, data);
+
+			data = ast_read32(ast, 0x12088);
+			/* multi-pins for DVO single-edge */
+			data |= 0x000fffff;
+			ast_write32(ast, 0x12088, data);
+
+			data = ast_read32(ast, 0x12090);
+			/* multi-pins for DVO single-edge */
+			data &= 0xffffffcf;
+			data |= 0x00000020;
+			ast_write32(ast, 0x12090, data);
+		} else { /* AST2400 */
+			data = ast_read32(ast, 0x12088);
+			/* multi-pins for DVO single-edge */
+			data |= 0x30000000;
+			ast_write32(ast, 0x12088, data);
+
+			data = ast_read32(ast, 0x1208c);
+			/* multi-pins for DVO single-edge */
+			data |= 0x000000cf;
+			ast_write32(ast, 0x1208c, data);
+
+			data = ast_read32(ast, 0x120a4);
+			/* multi-pins for DVO single-edge */
+			data |= 0xffff0000;
+			ast_write32(ast, 0x120a4, data);
+
+			data = ast_read32(ast, 0x120a8);
+			/* multi-pins for DVO single-edge */
+			data |= 0x0000000f;
+			ast_write32(ast, 0x120a8, data);
+
+			data = ast_read32(ast, 0x12094);
+			/* multi-pins for DVO single-edge */
+			data |= 0x00000002;
+			ast_write32(ast, 0x12094, data);
+		}
+	}
+
+	/* Force to DVO */
+	data = ast_read32(ast, 0x1202c);
+	data &= 0xfffbffff;
+	ast_write32(ast, 0x1202c, data);
+
+	/* Init VGA DVO Settings */
+	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xcf, 0x80);
+	return true;
+}
+
+void ast_init_3rdtx(struct drm_device *dev)
+{
+	struct ast_private *ast = dev->dev_private;
+	u8 jreg;
+	u32 data;
+	if (ast->chip == AST2300 || ast->chip == AST2400) {
+		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd1, 0xff);
+		switch (jreg & 0x0e) {
+		case 0x04:
+			ast_init_dvo(dev);
+			break;
+		case 0x08:
+			ast_launch_m68k(dev);
+			break;
+		case 0x0c:
+			ast_init_dvo(dev);
+			break;
+		default:
+			if (ast->tx_chip_type == AST_TX_SIL164)
+				ast_init_dvo(dev);
+			else {
+				ast_write32(ast, 0x12000, 0x1688a8a8);
+				data = ast_read32(ast, 0x1202c);
+				data &= 0xfffcffff;
+				ast_write32(ast, 0, data);
+			}
+		}
+	}
+}
diff --git a/drivers/gpu/drm/ast/ast_drv.c b/drivers/gpu/drm/ast/ast_drv.c
index 5137f15dba1..44074fbcf7f 100644
--- a/drivers/gpu/drm/ast/ast_drv.c
+++ b/drivers/gpu/drm/ast/ast_drv.c
@@ -94,9 +94,7 @@ static int ast_drm_thaw(struct drm_device *dev)
 	ast_post_gpu(dev);
 
 	drm_mode_config_reset(dev);
-	drm_modeset_lock_all(dev);
 	drm_helper_resume_force_mode(dev);
-	drm_modeset_unlock_all(dev);
 
 	console_lock();
 	ast_fbdev_set_suspend(dev, 0);
@@ -198,7 +196,6 @@ static const struct file_operations ast_fops = {
 
 static struct drm_driver driver = {
 	.driver_features = DRIVER_MODESET | DRIVER_GEM,
-	.dev_priv_size = 0,
 
 	.load = ast_driver_load,
 	.unload = ast_driver_unload,
diff --git a/drivers/gpu/drm/ast/ast_drv.h b/drivers/gpu/drm/ast/ast_drv.h
index 9833a1b1acc..5d6a87573c3 100644
--- a/drivers/gpu/drm/ast/ast_drv.h
+++ b/drivers/gpu/drm/ast/ast_drv.h
@@ -61,9 +61,17 @@ enum ast_chip {
 	AST2200,
 	AST2150,
 	AST2300,
+	AST2400,
 	AST1180,
 };
 
+enum ast_tx_chip {
+	AST_TX_NONE,
+	AST_TX_SIL164,
+	AST_TX_ITE66121,
+	AST_TX_DP501,
+};
+
 #define AST_DRAM_512Mx16 0
 #define AST_DRAM_1Gx16   1
 #define AST_DRAM_512Mx32 2
@@ -102,6 +110,12 @@ struct ast_private {
 	 * we have. */
 	struct ttm_bo_kmap_obj cache_kmap;
 	int next_cursor;
+	bool support_wide_screen;
+
+	enum ast_tx_chip tx_chip_type;
+	u8 dp501_maxclk;
+	u8 *dp501_fw_addr;
+	const struct firmware *dp501_fw;	/* dp501 fw */
 };
 
 int ast_driver_load(struct drm_device *dev, unsigned long flags);
@@ -368,4 +382,14 @@ int ast_mmap(struct file *filp, struct vm_area_struct *vma);
 
 /* ast post */
 void ast_post_gpu(struct drm_device *dev);
+u32 ast_mindwm(struct ast_private *ast, u32 r);
+void ast_moutdwm(struct ast_private *ast, u32 r, u32 v);
+/* ast dp501 */
+int ast_load_dp501_microcode(struct drm_device *dev);
+void ast_set_dp501_video_output(struct drm_device *dev, u8 mode);
+bool ast_launch_m68k(struct drm_device *dev);
+bool ast_backup_fw(struct drm_device *dev, u8 *addr, u32 size);
+bool ast_dp501_read_edid(struct drm_device *dev, u8 *ediddata);
+u8 ast_get_dp501_max_clk(struct drm_device *dev);
+void ast_init_3rdtx(struct drm_device *dev);
 #endif
diff --git a/drivers/gpu/drm/ast/ast_fb.c b/drivers/gpu/drm/ast/ast_fb.c
index 7b33e14e44a..a28640f47c2 100644
--- a/drivers/gpu/drm/ast/ast_fb.c
+++ b/drivers/gpu/drm/ast/ast_fb.c
@@ -65,7 +65,7 @@ static void ast_dirty_update(struct ast_fbdev *afbdev,
 	 * then the BO is being moved and we should
 	 * store up the damage until later.
 	 */
-	if (!in_interrupt())
+	if (drm_can_sleep())
 		ret = ast_bo_reserve(bo, true);
 	if (ret) {
 		if (ret != -EBUSY)
diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c
index af0b868a9df..a2cc6be9798 100644
--- a/drivers/gpu/drm/ast/ast_main.c
+++ b/drivers/gpu/drm/ast/ast_main.c
@@ -66,12 +66,16 @@ uint8_t ast_get_index_reg_mask(struct ast_private *ast,
 static int ast_detect_chip(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
+	uint32_t data, jreg;
 
 	if (dev->pdev->device == PCI_CHIP_AST1180) {
 		ast->chip = AST1100;
 		DRM_INFO("AST 1180 detected\n");
 	} else {
-		if (dev->pdev->revision >= 0x20) {
+		if (dev->pdev->revision >= 0x30) {
+			ast->chip = AST2400;
+			DRM_INFO("AST 2400 detected\n");
+		} else if (dev->pdev->revision >= 0x20) {
 			ast->chip = AST2300;
 			DRM_INFO("AST 2300 detected\n");
 		} else if (dev->pdev->revision >= 0x10) {
@@ -104,6 +108,59 @@ static int ast_detect_chip(struct drm_device *dev)
 			DRM_INFO("AST 2000 detected\n");
 		}
 	}
+
+	switch (ast->chip) {
+	case AST1180:
+		ast->support_wide_screen = true;
+		break;
+	case AST2000:
+		ast->support_wide_screen = false;
+		break;
+	default:
+		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd0, 0xff);
+		if (!(jreg & 0x80))
+			ast->support_wide_screen = true;
+		else if (jreg & 0x01)
+			ast->support_wide_screen = true;
+		else {
+			ast->support_wide_screen = false;
+			ast_write32(ast, 0xf004, 0x1e6e0000);
+			ast_write32(ast, 0xf000, 0x1);
+			data = ast_read32(ast, 0x1207c);
+			data &= 0x300;
+			if (ast->chip == AST2300 && data == 0x0) /* ast1300 */
+				ast->support_wide_screen = true;
+			if (ast->chip == AST2400 && data == 0x100) /* ast1400 */
+				ast->support_wide_screen = true;
+		}
+		break;
+	}
+
+	ast->tx_chip_type = AST_TX_NONE;
+	jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa3, 0xff);
+	if (jreg & 0x80)
+		ast->tx_chip_type = AST_TX_SIL164;
+	if ((ast->chip == AST2300) || (ast->chip == AST2400)) {
+		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd1, 0xff);
+		switch (jreg) {
+		case 0x04:
+			ast->tx_chip_type = AST_TX_SIL164;
+			break;
+		case 0x08:
+			ast->dp501_fw_addr = kzalloc(32*1024, GFP_KERNEL);
+			if (ast->dp501_fw_addr) {
+				/* backup firmware */
+				if (ast_backup_fw(dev, ast->dp501_fw_addr, 32*1024)) {
+					kfree(ast->dp501_fw_addr);
+					ast->dp501_fw_addr = NULL;
+				}
+			}
+			/* fallthrough */
+		case 0x0c:
+			ast->tx_chip_type = AST_TX_DP501;
+		}
+	}
+
 	return 0;
 }
 
@@ -129,7 +186,7 @@ static int ast_get_dram_info(struct drm_device *dev)
 	else
 		ast->dram_bus_width = 32;
 
-	if (ast->chip == AST2300) {
+	if (ast->chip == AST2300 || ast->chip == AST2400) {
 		switch (data & 0x03) {
 		case 0:
 			ast->dram_type = AST_DRAM_512Mx16;
@@ -189,53 +246,6 @@ static int ast_get_dram_info(struct drm_device *dev)
 	return 0;
 }
 
-uint32_t ast_get_max_dclk(struct drm_device *dev, int bpp)
-{
-	struct ast_private *ast = dev->dev_private;
-	uint32_t dclk, jreg;
-	uint32_t dram_bus_width, mclk, dram_bandwidth, actual_dram_bandwidth, dram_efficency = 500;
-
-	dram_bus_width = ast->dram_bus_width;
-	mclk = ast->mclk;
-
-	if (ast->chip == AST2100 ||
-	    ast->chip == AST1100 ||
-	    ast->chip == AST2200 ||
-	    ast->chip == AST2150 ||
-	    ast->dram_bus_width == 16)
-		dram_efficency = 600;
-	else if (ast->chip == AST2300)
-		dram_efficency = 400;
-
-	dram_bandwidth = mclk * dram_bus_width * 2 / 8;
-	actual_dram_bandwidth = dram_bandwidth * dram_efficency / 1000;
-
-	if (ast->chip == AST1180)
-		dclk = actual_dram_bandwidth / ((bpp + 1) / 8);
-	else {
-		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd0, 0xff);
-		if ((jreg & 0x08) && (ast->chip == AST2000))
-			dclk = actual_dram_bandwidth / ((bpp + 1 + 16) / 8);
-		else if ((jreg & 0x08) && (bpp == 8))
-			dclk = actual_dram_bandwidth / ((bpp + 1 + 24) / 8);
-		else
-			dclk = actual_dram_bandwidth / ((bpp + 1) / 8);
-	}
-
-	if (ast->chip == AST2100 ||
-	    ast->chip == AST2200 ||
-	    ast->chip == AST2300 ||
-	    ast->chip == AST1180) {
-		if (dclk > 200)
-			dclk = 200;
-	} else {
-		if (dclk > 165)
-			dclk = 165;
-	}
-
-	return dclk;
-}
-
 static void ast_user_framebuffer_destroy(struct drm_framebuffer *fb)
 {
 	struct ast_framebuffer *ast_fb = to_ast_framebuffer(fb);
@@ -304,17 +314,32 @@ static u32 ast_get_vram_info(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 	u8 jreg;
-
+	u32 vram_size;
 	ast_open_key(ast);
 
+	vram_size = AST_VIDMEM_DEFAULT_SIZE;
 	jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xaa, 0xff);
 	switch (jreg & 3) {
-	case 0: return AST_VIDMEM_SIZE_8M;
-	case 1: return AST_VIDMEM_SIZE_16M;
-	case 2: return AST_VIDMEM_SIZE_32M;
-	case 3: return AST_VIDMEM_SIZE_64M;
+	case 0: vram_size = AST_VIDMEM_SIZE_8M; break;
+	case 1: vram_size = AST_VIDMEM_SIZE_16M; break;
+	case 2: vram_size = AST_VIDMEM_SIZE_32M; break;
+	case 3: vram_size = AST_VIDMEM_SIZE_64M; break;
 	}
-	return AST_VIDMEM_DEFAULT_SIZE;
+
+	jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0x99, 0xff);
+	switch (jreg & 0x03) {
+	case 1:
+		vram_size -= 0x100000;
+		break;
+	case 2:
+		vram_size -= 0x200000;
+		break;
+	case 3:
+		vram_size -= 0x400000;
+		break;
+	}
+
+	return vram_size;
 }
 
 int ast_driver_load(struct drm_device *dev, unsigned long flags)
@@ -363,6 +388,7 @@ int ast_driver_load(struct drm_device *dev, unsigned long flags)
 	if (ast->chip == AST2100 ||
 	    ast->chip == AST2200 ||
 	    ast->chip == AST2300 ||
+	    ast->chip == AST2400 ||
 	    ast->chip == AST1180) {
 		dev->mode_config.max_width = 1920;
 		dev->mode_config.max_height = 2048;
@@ -390,6 +416,7 @@ int ast_driver_unload(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 
+	kfree(ast->dp501_fw_addr);
 	ast_mode_fini(dev);
 	ast_fbdev_fini(dev);
 	drm_mode_config_cleanup(dev);
@@ -449,7 +476,7 @@ int ast_dumb_create(struct drm_file *file,
 	return 0;
 }
 
-void ast_bo_unref(struct ast_bo **bo)
+static void ast_bo_unref(struct ast_bo **bo)
 {
 	struct ttm_buffer_object *tbo;
 
@@ -458,16 +485,13 @@ void ast_bo_unref(struct ast_bo **bo)
 
 	tbo = &((*bo)->bo);
 	ttm_bo_unref(&tbo);
-	if (tbo == NULL)
-		*bo = NULL;
-
+	*bo = NULL;
 }
+
 void ast_gem_free_object(struct drm_gem_object *obj)
 {
 	struct ast_bo *ast_bo = gem_to_ast_bo(obj);
 
-	if (!ast_bo)
-		return;
 	ast_bo_unref(&ast_bo);
 }
 
diff --git a/drivers/gpu/drm/ast/ast_mode.c b/drivers/gpu/drm/ast/ast_mode.c
index 7fc9f7272b5..114aee941d4 100644
--- a/drivers/gpu/drm/ast/ast_mode.c
+++ b/drivers/gpu/drm/ast/ast_mode.c
@@ -81,7 +81,7 @@ static bool ast_get_vbios_mode_info(struct drm_crtc *crtc, struct drm_display_mo
 	u32 refresh_rate_index = 0, mode_id, color_index, refresh_rate;
 	u32 hborder, vborder;
 
-	switch (crtc->fb->bits_per_pixel) {
+	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
 		vbios_mode->std_table = &vbios_stdtable[VGAModeIndex];
 		color_index = VGAModeIndex - 1;
@@ -115,11 +115,17 @@ static bool ast_get_vbios_mode_info(struct drm_crtc *crtc, struct drm_display_mo
 		else
 			vbios_mode->enh_table = &res_1280x1024[refresh_rate_index];
 		break;
+	case 1360:
+		vbios_mode->enh_table = &res_1360x768[refresh_rate_index];
+		break;
 	case 1440:
 		vbios_mode->enh_table = &res_1440x900[refresh_rate_index];
 		break;
 	case 1600:
-		vbios_mode->enh_table = &res_1600x1200[refresh_rate_index];
+		if (crtc->mode.crtc_vdisplay == 900)
+			vbios_mode->enh_table = &res_1600x900[refresh_rate_index];
+		else
+			vbios_mode->enh_table = &res_1600x1200[refresh_rate_index];
 		break;
 	case 1680:
 		vbios_mode->enh_table = &res_1680x1050[refresh_rate_index];
@@ -175,14 +181,17 @@ static bool ast_get_vbios_mode_info(struct drm_crtc *crtc, struct drm_display_mo
 		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x8d, refresh_rate_index & 0xff);
 		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x8e, mode_id & 0xff);
 
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x91, 0xa8);
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x92, crtc->fb->bits_per_pixel);
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x93, adjusted_mode->clock / 1000);
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x94, adjusted_mode->crtc_hdisplay);
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x95, adjusted_mode->crtc_hdisplay >> 8);
+		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x91, 0x00);
+		if (vbios_mode->enh_table->flags & NewModeInfo) {
+			ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x91, 0xa8);
+			ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x92, crtc->primary->fb->bits_per_pixel);
+			ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x93, adjusted_mode->clock / 1000);
+			ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x94, adjusted_mode->crtc_hdisplay);
+			ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x95, adjusted_mode->crtc_hdisplay >> 8);
 
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x96, adjusted_mode->crtc_vdisplay);
-		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x97, adjusted_mode->crtc_vdisplay >> 8);
+			ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x96, adjusted_mode->crtc_vdisplay);
+			ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x97, adjusted_mode->crtc_vdisplay >> 8);
+		}
 	}
 
 	return true;
@@ -340,7 +349,7 @@ static void ast_set_offset_reg(struct drm_crtc *crtc)
 
 	u16 offset;
 
-	offset = crtc->fb->pitches[0] >> 3;
+	offset = crtc->primary->fb->pitches[0] >> 3;
 	ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0x13, (offset & 0xff));
 	ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xb0, (offset >> 8) & 0x3f);
 }
@@ -365,7 +374,7 @@ static void ast_set_ext_reg(struct drm_crtc *crtc, struct drm_display_mode *mode
 	struct ast_private *ast = crtc->dev->dev_private;
 	u8 jregA0 = 0, jregA3 = 0, jregA8 = 0;
 
-	switch (crtc->fb->bits_per_pixel) {
+	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
 		jregA0 = 0x70;
 		jregA3 = 0x01;
@@ -389,7 +398,7 @@ static void ast_set_ext_reg(struct drm_crtc *crtc, struct drm_display_mode *mode
 	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xa8, 0xfd, jregA8);
 
 	/* Set Threshold */
-	if (ast->chip == AST2300) {
+	if (ast->chip == AST2300 || ast->chip == AST2400) {
 		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xa7, 0x78);
 		ast_set_index_reg(ast, AST_IO_CRTC_PORT, 0xa6, 0x60);
 	} else if (ast->chip == AST2100 ||
@@ -404,7 +413,7 @@ static void ast_set_ext_reg(struct drm_crtc *crtc, struct drm_display_mode *mode
 	}
 }
 
-void ast_set_sync_reg(struct drm_device *dev, struct drm_display_mode *mode,
+static void ast_set_sync_reg(struct drm_device *dev, struct drm_display_mode *mode,
 		      struct ast_vbios_mode_info *vbios_mode)
 {
 	struct ast_private *ast = dev->dev_private;
@@ -415,10 +424,10 @@ void ast_set_sync_reg(struct drm_device *dev, struct drm_display_mode *mode,
 	ast_io_write8(ast, AST_IO_MISC_PORT_WRITE, jreg);
 }
 
-bool ast_set_dac_reg(struct drm_crtc *crtc, struct drm_display_mode *mode,
+static bool ast_set_dac_reg(struct drm_crtc *crtc, struct drm_display_mode *mode,
 		     struct ast_vbios_mode_info *vbios_mode)
 {
-	switch (crtc->fb->bits_per_pixel) {
+	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
 		break;
 	default:
@@ -427,7 +436,7 @@ bool ast_set_dac_reg(struct drm_crtc *crtc, struct drm_display_mode *mode,
 	return true;
 }
 
-void ast_set_start_address_crt1(struct drm_crtc *crtc, unsigned offset)
+static void ast_set_start_address_crt1(struct drm_crtc *crtc, unsigned offset)
 {
 	struct ast_private *ast = crtc->dev->dev_private;
 	u32 addr;
@@ -451,9 +460,13 @@ static void ast_crtc_dpms(struct drm_crtc *crtc, int mode)
 	case DRM_MODE_DPMS_STANDBY:
 	case DRM_MODE_DPMS_SUSPEND:
 		ast_set_index_reg_mask(ast, AST_IO_SEQ_PORT, 0x1, 0xdf, 0);
+		if (ast->tx_chip_type == AST_TX_DP501)
+			ast_set_dp501_video_output(crtc->dev, 1);
 		ast_crtc_load_lut(crtc);
 		break;
 	case DRM_MODE_DPMS_OFF:
+		if (ast->tx_chip_type == AST_TX_DP501)
+			ast_set_dp501_video_output(crtc->dev, 0);
 		ast_set_index_reg_mask(ast, AST_IO_SEQ_PORT, 0x1, 0xdf, 0x20);
 		break;
 	}
@@ -490,7 +503,7 @@ static int ast_crtc_do_set_base(struct drm_crtc *crtc,
 		ast_bo_unreserve(bo);
 	}
 
-	ast_fb = to_ast_framebuffer(crtc->fb);
+	ast_fb = to_ast_framebuffer(crtc->primary->fb);
 	obj = ast_fb->obj;
 	bo = gem_to_ast_bo(obj);
 
@@ -623,7 +636,7 @@ static const struct drm_crtc_funcs ast_crtc_funcs = {
 	.destroy = ast_crtc_destroy,
 };
 
-int ast_crtc_init(struct drm_device *dev)
+static int ast_crtc_init(struct drm_device *dev)
 {
 	struct ast_crtc *crtc;
 	int i;
@@ -710,7 +723,7 @@ static const struct drm_encoder_helper_funcs ast_enc_helper_funcs = {
 	.mode_set = ast_encoder_mode_set,
 };
 
-int ast_encoder_init(struct drm_device *dev)
+static int ast_encoder_init(struct drm_device *dev)
 {
 	struct ast_encoder *ast_encoder;
 
@@ -729,10 +742,24 @@ int ast_encoder_init(struct drm_device *dev)
 static int ast_get_modes(struct drm_connector *connector)
 {
 	struct ast_connector *ast_connector = to_ast_connector(connector);
+	struct ast_private *ast = connector->dev->dev_private;
 	struct edid *edid;
 	int ret;
-
-	edid = drm_get_edid(connector, &ast_connector->i2c->adapter);
+	bool flags = false;
+	if (ast->tx_chip_type == AST_TX_DP501) {
+		ast->dp501_maxclk = 0xff;
+		edid = kmalloc(128, GFP_KERNEL);
+		if (!edid)
+			return -ENOMEM;
+
+		flags = ast_dp501_read_edid(connector->dev, (u8 *)edid);
+		if (flags)
+			ast->dp501_maxclk = ast_get_dp501_max_clk(connector->dev);
+		else
+			kfree(edid);
+	}
+	if (!flags)
+		edid = drm_get_edid(connector, &ast_connector->i2c->adapter);
 	if (edid) {
 		drm_mode_connector_update_edid_property(&ast_connector->base, edid);
 		ret = drm_add_edid_modes(connector, edid);
@@ -746,7 +773,56 @@ static int ast_get_modes(struct drm_connector *connector)
 static int ast_mode_valid(struct drm_connector *connector,
 			  struct drm_display_mode *mode)
 {
-	return MODE_OK;
+	struct ast_private *ast = connector->dev->dev_private;
+	int flags = MODE_NOMODE;
+	uint32_t jtemp;
+
+	if (ast->support_wide_screen) {
+		if ((mode->hdisplay == 1680) && (mode->vdisplay == 1050))
+			return MODE_OK;
+		if ((mode->hdisplay == 1280) && (mode->vdisplay == 800))
+			return MODE_OK;
+		if ((mode->hdisplay == 1440) && (mode->vdisplay == 900))
+			return MODE_OK;
+		if ((mode->hdisplay == 1360) && (mode->vdisplay == 768))
+			return MODE_OK;
+		if ((mode->hdisplay == 1600) && (mode->vdisplay == 900))
+			return MODE_OK;
+
+		if ((ast->chip == AST2100) || (ast->chip == AST2200) || (ast->chip == AST2300) || (ast->chip == AST2400) || (ast->chip == AST1180)) {
+			if ((mode->hdisplay == 1920) && (mode->vdisplay == 1080))
+				return MODE_OK;
+
+			if ((mode->hdisplay == 1920) && (mode->vdisplay == 1200)) {
+				jtemp = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd1, 0xff);
+				if (jtemp & 0x01)
+					return MODE_NOMODE;
+				else
+					return MODE_OK;
+			}
+		}
+	}
+	switch (mode->hdisplay) {
+	case 640:
+		if (mode->vdisplay == 480) flags = MODE_OK;
+		break;
+	case 800:
+		if (mode->vdisplay == 600) flags = MODE_OK;
+		break;
+	case 1024:
+		if (mode->vdisplay == 768) flags = MODE_OK;
+		break;
+	case 1280:
+		if (mode->vdisplay == 1024) flags = MODE_OK;
+		break;
+	case 1600:
+		if (mode->vdisplay == 1200) flags = MODE_OK;
+		break;
+	default:
+		return flags;
+	}
+
+	return flags;
 }
 
 static void ast_connector_destroy(struct drm_connector *connector)
@@ -777,7 +853,7 @@ static const struct drm_connector_funcs ast_connector_funcs = {
 	.destroy = ast_connector_destroy,
 };
 
-int ast_connector_init(struct drm_device *dev)
+static int ast_connector_init(struct drm_device *dev)
 {
 	struct ast_connector *ast_connector;
 	struct drm_connector *connector;
@@ -810,7 +886,7 @@ int ast_connector_init(struct drm_device *dev)
 }
 
 /* allocate cursor cache and pin at start of VRAM */
-int ast_cursor_init(struct drm_device *dev)
+static int ast_cursor_init(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 	int size;
@@ -847,7 +923,7 @@ fail:
 	return ret;
 }
 
-void ast_cursor_fini(struct drm_device *dev)
+static void ast_cursor_fini(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 	ttm_bo_kunmap(&ast->cache_kmap);
@@ -965,7 +1041,7 @@ static void ast_i2c_destroy(struct ast_i2c_chan *i2c)
 	kfree(i2c);
 }
 
-void ast_show_cursor(struct drm_crtc *crtc)
+static void ast_show_cursor(struct drm_crtc *crtc)
 {
 	struct ast_private *ast = crtc->dev->dev_private;
 	u8 jreg;
@@ -976,7 +1052,7 @@ void ast_show_cursor(struct drm_crtc *crtc)
 	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xcb, 0xfc, jreg);
 }
 
-void ast_hide_cursor(struct drm_crtc *crtc)
+static void ast_hide_cursor(struct drm_crtc *crtc)
 {
 	struct ast_private *ast = crtc->dev->dev_private;
 	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xcb, 0xfc, 0x00);
diff --git a/drivers/gpu/drm/ast/ast_post.c b/drivers/gpu/drm/ast/ast_post.c
index 977cfb35837..38d437f3a26 100644
--- a/drivers/gpu/drm/ast/ast_post.c
+++ b/drivers/gpu/drm/ast/ast_post.c
@@ -78,7 +78,7 @@ ast_set_def_ext_reg(struct drm_device *dev)
 	for (i = 0x81; i <= 0x8f; i++)
 		ast_set_index_reg(ast, AST_IO_CRTC_PORT, i, 0x00);
 
-	if (ast->chip == AST2300) {
+	if (ast->chip == AST2300 || ast->chip == AST2400) {
 		if (dev->pdev->revision >= 0x20)
 			ext_reg_info = extreginfo_ast2300;
 		else
@@ -102,23 +102,32 @@ ast_set_def_ext_reg(struct drm_device *dev)
 
 	/* Enable RAMDAC for A1 */
 	reg = 0x04;
-	if (ast->chip == AST2300)
+	if (ast->chip == AST2300 || ast->chip == AST2400)
 		reg |= 0x20;
 	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xb6, 0xff, reg);
 }
 
-static inline u32 mindwm(struct ast_private *ast, u32 r)
+u32 ast_mindwm(struct ast_private *ast, u32 r)
 {
+	uint32_t data;
+
 	ast_write32(ast, 0xf004, r & 0xffff0000);
 	ast_write32(ast, 0xf000, 0x1);
 
+	do {
+		data = ast_read32(ast, 0xf004) & 0xffff0000;
+	} while (data != (r & 0xffff0000));
 	return ast_read32(ast, 0x10000 + (r & 0x0000ffff));
 }
 
-static inline void moutdwm(struct ast_private *ast, u32 r, u32 v)
+void ast_moutdwm(struct ast_private *ast, u32 r, u32 v)
 {
+	uint32_t data;
 	ast_write32(ast, 0xf004, r & 0xffff0000);
 	ast_write32(ast, 0xf000, 0x1);
+	do {
+		data = ast_read32(ast, 0xf004) & 0xffff0000;
+	} while (data != (r & 0xffff0000));
 	ast_write32(ast, 0x10000 + (r & 0x0000ffff), v);
 }
 
@@ -154,28 +163,28 @@ static u32 mmctestburst2_ast2150(struct ast_private *ast, u32 datagen)
 {
 	u32 data, timeout;
 
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
-	moutdwm(ast, 0x1e6e0070, 0x00000001 | (datagen << 3));
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000001 | (datagen << 3));
 	timeout = 0;
 	do {
-		data = mindwm(ast, 0x1e6e0070) & 0x40;
+		data = ast_mindwm(ast, 0x1e6e0070) & 0x40;
 		if (++timeout > TIMEOUT_AST2150) {
-			moutdwm(ast, 0x1e6e0070, 0x00000000);
+			ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
 			return 0xffffffff;
 		}
 	} while (!data);
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
-	moutdwm(ast, 0x1e6e0070, 0x00000003 | (datagen << 3));
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000003 | (datagen << 3));
 	timeout = 0;
 	do {
-		data = mindwm(ast, 0x1e6e0070) & 0x40;
+		data = ast_mindwm(ast, 0x1e6e0070) & 0x40;
 		if (++timeout > TIMEOUT_AST2150) {
-			moutdwm(ast, 0x1e6e0070, 0x00000000);
+			ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
 			return 0xffffffff;
 		}
 	} while (!data);
-	data = (mindwm(ast, 0x1e6e0070) & 0x80) >> 7;
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
+	data = (ast_mindwm(ast, 0x1e6e0070) & 0x80) >> 7;
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
 	return data;
 }
 
@@ -184,18 +193,18 @@ static u32 mmctestsingle2_ast2150(struct ast_private *ast, u32 datagen)
 {
 	u32 data, timeout;
 
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
-	moutdwm(ast, 0x1e6e0070, 0x00000005 | (datagen << 3));
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000005 | (datagen << 3));
 	timeout = 0;
 	do {
-		data = mindwm(ast, 0x1e6e0070) & 0x40;
+		data = ast_mindwm(ast, 0x1e6e0070) & 0x40;
 		if (++timeout > TIMEOUT_AST2150) {
-			moutdwm(ast, 0x1e6e0070, 0x00000000);
+			ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
 			return 0xffffffff;
 		}
 	} while (!data);
-	data = (mindwm(ast, 0x1e6e0070) & 0x80) >> 7;
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
+	data = (ast_mindwm(ast, 0x1e6e0070) & 0x80) >> 7;
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
 	return data;
 }
 #endif
@@ -215,7 +224,7 @@ static int cbrscan_ast2150(struct ast_private *ast, int busw)
 	u32 patcnt, loop;
 
 	for (patcnt = 0; patcnt < CBR_PATNUM_AST2150; patcnt++) {
-		moutdwm(ast, 0x1e6e007c, pattern_AST2150[patcnt]);
+		ast_moutdwm(ast, 0x1e6e007c, pattern_AST2150[patcnt]);
 		for (loop = 0; loop < CBR_PASSNUM_AST2150; loop++) {
 			if (cbrtest_ast2150(ast))
 				break;
@@ -237,7 +246,7 @@ cbr_start:
 	passcnt = 0;
 
 	for (dlli = 0; dlli < 100; dlli++) {
-		moutdwm(ast, 0x1e6e0068, dlli | (dlli << 8) | (dlli << 16) | (dlli << 24));
+		ast_moutdwm(ast, 0x1e6e0068, dlli | (dlli << 8) | (dlli << 16) | (dlli << 24));
 		data = cbrscan_ast2150(ast, busw);
 		if (data != 0) {
 			if (data & 0x1) {
@@ -254,7 +263,7 @@ cbr_start:
 		goto cbr_start;
 
 	dlli = dll_min[0] + (((dll_max[0] - dll_min[0]) * 7) >> 4);
-	moutdwm(ast, 0x1e6e0068, dlli | (dlli << 8) | (dlli << 16) | (dlli << 24));
+	ast_moutdwm(ast, 0x1e6e0068, dlli | (dlli << 8) | (dlli << 16) | (dlli << 24));
 }
 
 
@@ -365,10 +374,12 @@ void ast_post_gpu(struct drm_device *dev)
 	ast_open_key(ast);
 	ast_set_def_ext_reg(dev);
 
-	if (ast->chip == AST2300)
+	if (ast->chip == AST2300 || ast->chip == AST2400)
 		ast_init_dram_2300(dev);
 	else
 		ast_init_dram_reg(dev);
+
+	ast_init_3rdtx(dev);
 }
 
 /* AST 2300 DRAM settings */
@@ -403,6 +414,7 @@ struct ast2300_dram_param {
 /*
  * DQSI DLL CBR Setting
  */
+#define CBR_SIZE0            ((1  << 10) - 1)
 #define CBR_SIZE1            ((4  << 10) - 1)
 #define CBR_SIZE2            ((64 << 10) - 1)
 #define CBR_PASSNUM          5
@@ -423,88 +435,84 @@ static const u32 pattern[8] = {
 	0x7C61D253
 };
 
-#if 0 /* unused in DDX, included for completeness */
 static int mmc_test_burst(struct ast_private *ast, u32 datagen)
 {
 	u32 data, timeout;
 
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
-	moutdwm(ast, 0x1e6e0070, 0x000000c1 | (datagen << 3));
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
+	ast_moutdwm(ast, 0x1e6e0070, 0x000000c1 | (datagen << 3));
 	timeout = 0;
 	do {
-		data = mindwm(ast, 0x1e6e0070) & 0x3000;
+		data = ast_mindwm(ast, 0x1e6e0070) & 0x3000;
 		if (data & 0x2000) {
 			return 0;
 		}
 		if (++timeout > TIMEOUT) {
-			moutdwm(ast, 0x1e6e0070, 0x00000000);
+			ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
 			return 0;
 		}
 	} while (!data);
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
 	return 1;
 }
-#endif
 
 static int mmc_test_burst2(struct ast_private *ast, u32 datagen)
 {
 	u32 data, timeout;
 
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
-	moutdwm(ast, 0x1e6e0070, 0x00000041 | (datagen << 3));
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000041 | (datagen << 3));
 	timeout = 0;
 	do {
-		data = mindwm(ast, 0x1e6e0070) & 0x1000;
+		data = ast_mindwm(ast, 0x1e6e0070) & 0x1000;
 		if (++timeout > TIMEOUT) {
-			moutdwm(ast, 0x1e6e0070, 0x0);
+			ast_moutdwm(ast, 0x1e6e0070, 0x0);
 			return -1;
 		}
 	} while (!data);
-	data = mindwm(ast, 0x1e6e0078);
+	data = ast_mindwm(ast, 0x1e6e0078);
 	data = (data | (data >> 16)) & 0xffff;
-	moutdwm(ast, 0x1e6e0070, 0x0);
+	ast_moutdwm(ast, 0x1e6e0070, 0x0);
 	return data;
 }
 
-#if 0 /* Unused in DDX here for completeness */
 static int mmc_test_single(struct ast_private *ast, u32 datagen)
 {
 	u32 data, timeout;
 
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
-	moutdwm(ast, 0x1e6e0070, 0x000000c5 | (datagen << 3));
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
+	ast_moutdwm(ast, 0x1e6e0070, 0x000000c5 | (datagen << 3));
 	timeout = 0;
 	do {
-		data = mindwm(ast, 0x1e6e0070) & 0x3000;
+		data = ast_mindwm(ast, 0x1e6e0070) & 0x3000;
 		if (data & 0x2000)
 			return 0;
 		if (++timeout > TIMEOUT) {
-			moutdwm(ast, 0x1e6e0070, 0x0);
+			ast_moutdwm(ast, 0x1e6e0070, 0x0);
 			return 0;
 		}
 	} while (!data);
-	moutdwm(ast, 0x1e6e0070, 0x0);
+	ast_moutdwm(ast, 0x1e6e0070, 0x0);
 	return 1;
 }
-#endif
 
 static int mmc_test_single2(struct ast_private *ast, u32 datagen)
 {
 	u32 data, timeout;
 
-	moutdwm(ast, 0x1e6e0070, 0x00000000);
-	moutdwm(ast, 0x1e6e0070, 0x00000005 | (datagen << 3));
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000000);
+	ast_moutdwm(ast, 0x1e6e0070, 0x00000005 | (datagen << 3));
 	timeout = 0;
 	do {
-		data = mindwm(ast, 0x1e6e0070) & 0x1000;
+		data = ast_mindwm(ast, 0x1e6e0070) & 0x1000;
 		if (++timeout > TIMEOUT) {
-			moutdwm(ast, 0x1e6e0070, 0x0);
+			ast_moutdwm(ast, 0x1e6e0070, 0x0);
 			return -1;
 		}
 	} while (!data);
-	data = mindwm(ast, 0x1e6e0078);
+	data = ast_mindwm(ast, 0x1e6e0078);
 	data = (data | (data >> 16)) & 0xffff;
-	moutdwm(ast, 0x1e6e0070, 0x0);
+	ast_moutdwm(ast, 0x1e6e0070, 0x0);
 	return data;
 }
 
@@ -533,7 +541,7 @@ static int cbr_scan(struct ast_private *ast)
 
 	data2 = 3;
 	for (patcnt = 0; patcnt < CBR_PATNUM; patcnt++) {
-		moutdwm(ast, 0x1e6e007c, pattern[patcnt]);
+		ast_moutdwm(ast, 0x1e6e007c, pattern[patcnt]);
 		for (loop = 0; loop < CBR_PASSNUM2; loop++) {
 			if ((data = cbr_test(ast)) != 0) {
 				data2 &= data;
@@ -568,11 +576,11 @@ static u32 cbr_scan2(struct ast_private *ast)
 
 	data2 = 0xffff;
 	for (patcnt = 0; patcnt < CBR_PATNUM; patcnt++) {
-		moutdwm(ast, 0x1e6e007c, pattern[patcnt]);
+		ast_moutdwm(ast, 0x1e6e007c, pattern[patcnt]);
 		for (loop = 0; loop < CBR_PASSNUM2; loop++) {
 			if ((data = cbr_test2(ast)) != 0) {
 				data2 &= data;
-				if (!data)
+				if (!data2)
 					return 0;
 				break;
 			}
@@ -583,106 +591,35 @@ static u32 cbr_scan2(struct ast_private *ast)
 	return data2;
 }
 
-#if 0 /* unused in DDX - added for completeness */
-static void finetuneDQI(struct ast_private *ast, struct ast2300_dram_param *param)
+static u32 cbr_test3(struct ast_private *ast)
 {
-	u32 gold_sadj[2], dllmin[16], dllmax[16], dlli, data, cnt, mask, passcnt;
-
-	gold_sadj[0] = (mindwm(ast, 0x1E6E0024) >> 16) & 0xffff;
-	gold_sadj[1] = gold_sadj[0] >> 8;
-	gold_sadj[0] = gold_sadj[0] & 0xff;
-	gold_sadj[0] = (gold_sadj[0] + gold_sadj[1]) >> 1;
-	gold_sadj[1] = gold_sadj[0];
-
-	for (cnt = 0; cnt < 16; cnt++) {
-		dllmin[cnt] = 0xff;
-		dllmax[cnt] = 0x0;
-	}
-	passcnt = 0;
-	for (dlli = 0; dlli < 76; dlli++) {
-		moutdwm(ast, 0x1E6E0068, 0x00001400 | (dlli << 16) | (dlli << 24));
-		/* Wait DQSI latch phase calibration */
-		moutdwm(ast, 0x1E6E0074, 0x00000010);
-		moutdwm(ast, 0x1E6E0070, 0x00000003);
-		do {
-			data = mindwm(ast, 0x1E6E0070);
-		} while (!(data & 0x00001000));
-		moutdwm(ast, 0x1E6E0070, 0x00000000);
+	if (!mmc_test_burst(ast, 0))
+		return 0;
+	if (!mmc_test_single(ast, 0))
+		return 0;
+	return 1;
+}
 
-		moutdwm(ast, 0x1E6E0074, CBR_SIZE1);
-		data = cbr_scan2(ast);
-		if (data != 0) {
-			mask = 0x00010001;
-			for (cnt = 0; cnt < 16; cnt++) {
-				if (data & mask) {
-					if (dllmin[cnt] > dlli) {
-						dllmin[cnt] = dlli;
-					}
-					if (dllmax[cnt] < dlli) {
-						dllmax[cnt] = dlli;
-					}
-				}
-				mask <<= 1;
-			}
-			passcnt++;
-		} else if (passcnt >= CBR_THRESHOLD) {
-			break;
-		}
-	}
-	data = 0;
-	for (cnt = 0; cnt < 8; cnt++) {
-		data >>= 3;
-		if ((dllmax[cnt] > dllmin[cnt]) && ((dllmax[cnt] - dllmin[cnt]) >= CBR_THRESHOLD)) {
-			dlli = (dllmin[cnt] + dllmax[cnt]) >> 1;
-			if (gold_sadj[0] >= dlli) {
-				dlli = (gold_sadj[0] - dlli) >> 1;
-				if (dlli > 3) {
-					dlli = 3;
-				}
-			} else {
-				dlli = (dlli - gold_sadj[0]) >> 1;
-				if (dlli > 4) {
-					dlli = 4;
-				}
-				dlli = (8 - dlli) & 0x7;
-			}
-			data |= dlli << 21;
-		}
-	}
-	moutdwm(ast, 0x1E6E0080, data);
+static u32 cbr_scan3(struct ast_private *ast)
+{
+	u32 patcnt, loop;
 
-	data = 0;
-	for (cnt = 8; cnt < 16; cnt++) {
-		data >>= 3;
-		if ((dllmax[cnt] > dllmin[cnt]) && ((dllmax[cnt] - dllmin[cnt]) >= CBR_THRESHOLD)) {
-			dlli = (dllmin[cnt] + dllmax[cnt]) >> 1;
-			if (gold_sadj[1] >= dlli) {
-				dlli = (gold_sadj[1] - dlli) >> 1;
-				if (dlli > 3) {
-					dlli = 3;
-				} else {
-					dlli = (dlli - 1) & 0x7;
-				}
-			} else {
-				dlli = (dlli - gold_sadj[1]) >> 1;
-				dlli += 1;
-				if (dlli > 4) {
-					dlli = 4;
-				}
-				dlli = (8 - dlli) & 0x7;
-			}
-			data |= dlli << 21;
+	for (patcnt = 0; patcnt < CBR_PATNUM; patcnt++) {
+		ast_moutdwm(ast, 0x1e6e007c, pattern[patcnt]);
+		for (loop = 0; loop < 2; loop++) {
+			if (cbr_test3(ast))
+				break;
 		}
+		if (loop == 2)
+			return 0;
 	}
-	moutdwm(ast, 0x1E6E0084, data);
-
-} /* finetuneDQI */
-#endif
+	return 1;
+}
 
-static void finetuneDQI_L(struct ast_private *ast, struct ast2300_dram_param *param)
+static bool finetuneDQI_L(struct ast_private *ast, struct ast2300_dram_param *param)
 {
-	u32 gold_sadj[2], dllmin[16], dllmax[16], dlli, data, cnt, mask, passcnt;
-
+	u32 gold_sadj[2], dllmin[16], dllmax[16], dlli, data, cnt, mask, passcnt, retry = 0;
+	bool status = false;
 FINETUNE_START:
 	for (cnt = 0; cnt < 16; cnt++) {
 		dllmin[cnt] = 0xff;
@@ -690,16 +627,8 @@ FINETUNE_START:
 	}
 	passcnt = 0;
 	for (dlli = 0; dlli < 76; dlli++) {
-		moutdwm(ast, 0x1E6E0068, 0x00001400 | (dlli << 16) | (dlli << 24));
-		/* Wait DQSI latch phase calibration */
-		moutdwm(ast, 0x1E6E0074, 0x00000010);
-		moutdwm(ast, 0x1E6E0070, 0x00000003);
-		do {
-			data = mindwm(ast, 0x1E6E0070);
-		} while (!(data & 0x00001000));
-		moutdwm(ast, 0x1E6E0070, 0x00000000);
-
-		moutdwm(ast, 0x1E6E0074, CBR_SIZE1);
+		ast_moutdwm(ast, 0x1E6E0068, 0x00001400 | (dlli << 16) | (dlli << 24));
+		ast_moutdwm(ast, 0x1E6E0074, CBR_SIZE1);
 		data = cbr_scan2(ast);
 		if (data != 0) {
 			mask = 0x00010001;
@@ -727,9 +656,13 @@ FINETUNE_START:
 			passcnt++;
 		}
 	}
+	if (retry++ > 10)
+		goto FINETUNE_DONE;
 	if (passcnt != 16) {
 		goto FINETUNE_START;
 	}
+	status = true;
+FINETUNE_DONE:
 	gold_sadj[0] = gold_sadj[0] >> 4;
 	gold_sadj[1] = gold_sadj[0];
 
@@ -753,7 +686,7 @@ FINETUNE_START:
 			data |= dlli << 21;
 		}
 	}
-	moutdwm(ast, 0x1E6E0080, data);
+	ast_moutdwm(ast, 0x1E6E0080, data);
 
 	data = 0;
 	for (cnt = 8; cnt < 16; cnt++) {
@@ -778,162 +711,116 @@ FINETUNE_START:
 			data |= dlli << 21;
 		}
 	}
-	moutdwm(ast, 0x1E6E0084, data);
-
+	ast_moutdwm(ast, 0x1E6E0084, data);
+	return status;
 } /* finetuneDQI_L */
 
-static void finetuneDQI_L2(struct ast_private *ast, struct ast2300_dram_param *param)
+static void finetuneDQSI(struct ast_private *ast)
 {
-	u32 gold_sadj[2], dllmin[16], dllmax[16], dlli, data, cnt, mask, passcnt, data2;
+	u32 dlli, dqsip, dqidly;
+	u32 reg_mcr18, reg_mcr0c, passcnt[2], diff;
+	u32 g_dqidly, g_dqsip, g_margin, g_side;
+	u16 pass[32][2][2];
+	char tag[2][76];
+
+	/* Disable DQI CBR */
+	reg_mcr0c  = ast_mindwm(ast, 0x1E6E000C);
+	reg_mcr18  = ast_mindwm(ast, 0x1E6E0018);
+	reg_mcr18 &= 0x0000ffff;
+	ast_moutdwm(ast, 0x1E6E0018, reg_mcr18);
 
-	for (cnt = 0; cnt < 16; cnt++) {
-		dllmin[cnt] = 0xff;
-		dllmax[cnt] = 0x0;
-	}
-	passcnt = 0;
 	for (dlli = 0; dlli < 76; dlli++) {
-		moutdwm(ast, 0x1E6E0068, 0x00001400 | (dlli << 16) | (dlli << 24));
-		/* Wait DQSI latch phase calibration */
-		moutdwm(ast, 0x1E6E0074, 0x00000010);
-		moutdwm(ast, 0x1E6E0070, 0x00000003);
-		do {
-			data = mindwm(ast, 0x1E6E0070);
-		} while (!(data & 0x00001000));
-		moutdwm(ast, 0x1E6E0070, 0x00000000);
-
-		moutdwm(ast, 0x1E6E0074, CBR_SIZE2);
-		data = cbr_scan2(ast);
-		if (data != 0) {
-			mask = 0x00010001;
-			for (cnt = 0; cnt < 16; cnt++) {
-				if (data & mask) {
-					if (dllmin[cnt] > dlli) {
-						dllmin[cnt] = dlli;
-					}
-					if (dllmax[cnt] < dlli) {
-						dllmax[cnt] = dlli;
-					}
-				}
-				mask <<= 1;
-			}
-			passcnt++;
-		} else if (passcnt >= CBR_THRESHOLD2) {
-			break;
-		}
+		tag[0][dlli] = 0x0;
+		tag[1][dlli] = 0x0;
 	}
-	gold_sadj[0] = 0x0;
-	gold_sadj[1] = 0xFF;
-	for (cnt = 0; cnt < 8; cnt++) {
-		if ((dllmax[cnt] > dllmin[cnt]) && ((dllmax[cnt] - dllmin[cnt]) >= CBR_THRESHOLD2)) {
-			if (gold_sadj[0] < dllmin[cnt]) {
-				gold_sadj[0] = dllmin[cnt];
-			}
-			if (gold_sadj[1] > dllmax[cnt]) {
-				gold_sadj[1] = dllmax[cnt];
-			}
-		}
+	for (dqidly = 0; dqidly < 32; dqidly++) {
+		pass[dqidly][0][0] = 0xff;
+		pass[dqidly][0][1] = 0x0;
+		pass[dqidly][1][0] = 0xff;
+		pass[dqidly][1][1] = 0x0;
 	}
-	gold_sadj[0] = (gold_sadj[1] + gold_sadj[0]) >> 1;
-	gold_sadj[1] = mindwm(ast, 0x1E6E0080);
-
-	data = 0;
-	for (cnt = 0; cnt < 8; cnt++) {
-		data >>= 3;
-		data2 = gold_sadj[1] & 0x7;
-		gold_sadj[1] >>= 3;
-		if ((dllmax[cnt] > dllmin[cnt]) && ((dllmax[cnt] - dllmin[cnt]) >= CBR_THRESHOLD2)) {
-			dlli = (dllmin[cnt] + dllmax[cnt]) >> 1;
-			if (gold_sadj[0] >= dlli) {
-				dlli = (gold_sadj[0] - dlli) >> 1;
-				if (dlli > 0) {
-					dlli = 1;
-				}
-				if (data2 != 3) {
-					data2 = (data2 + dlli) & 0x7;
-				}
-			} else {
-				dlli = (dlli - gold_sadj[0]) >> 1;
-				if (dlli > 0) {
-					dlli = 1;
-				}
-				if (data2 != 4) {
-					data2 = (data2 - dlli) & 0x7;
+	for (dqidly = 0; dqidly < 32; dqidly++) {
+		passcnt[0] = passcnt[1] = 0;
+		for (dqsip = 0; dqsip < 2; dqsip++) {
+			ast_moutdwm(ast, 0x1E6E000C, 0);
+			ast_moutdwm(ast, 0x1E6E0018, reg_mcr18 | (dqidly << 16) | (dqsip << 23));
+			ast_moutdwm(ast, 0x1E6E000C, reg_mcr0c);
+			for (dlli = 0; dlli < 76; dlli++) {
+				ast_moutdwm(ast, 0x1E6E0068, 0x00001300 | (dlli << 16) | (dlli << 24));
+				ast_moutdwm(ast, 0x1E6E0070, 0);
+				ast_moutdwm(ast, 0x1E6E0074, CBR_SIZE0);
+				if (cbr_scan3(ast)) {
+					if (dlli == 0)
+						break;
+					passcnt[dqsip]++;
+					tag[dqsip][dlli] = 'P';
+					if (dlli < pass[dqidly][dqsip][0])
+						pass[dqidly][dqsip][0] = (u16) dlli;
+					if (dlli > pass[dqidly][dqsip][1])
+						pass[dqidly][dqsip][1] = (u16) dlli;
+				} else if (passcnt[dqsip] >= 5)
+					break;
+				else {
+					pass[dqidly][dqsip][0] = 0xff;
+					pass[dqidly][dqsip][1] = 0x0;
 				}
 			}
 		}
-		data |= data2 << 21;
-	}
-	moutdwm(ast, 0x1E6E0080, data);
-
-	gold_sadj[0] = 0x0;
-	gold_sadj[1] = 0xFF;
-	for (cnt = 8; cnt < 16; cnt++) {
-		if ((dllmax[cnt] > dllmin[cnt]) && ((dllmax[cnt] - dllmin[cnt]) >= CBR_THRESHOLD2)) {
-			if (gold_sadj[0] < dllmin[cnt]) {
-				gold_sadj[0] = dllmin[cnt];
-			}
-			if (gold_sadj[1] > dllmax[cnt]) {
-				gold_sadj[1] = dllmax[cnt];
-			}
-		}
+		if (passcnt[0] == 0 && passcnt[1] == 0)
+			dqidly++;
 	}
-	gold_sadj[0] = (gold_sadj[1] + gold_sadj[0]) >> 1;
-	gold_sadj[1] = mindwm(ast, 0x1E6E0084);
-
-	data = 0;
-	for (cnt = 8; cnt < 16; cnt++) {
-		data >>= 3;
-		data2 = gold_sadj[1] & 0x7;
-		gold_sadj[1] >>= 3;
-		if ((dllmax[cnt] > dllmin[cnt]) && ((dllmax[cnt] - dllmin[cnt]) >= CBR_THRESHOLD2)) {
-			dlli = (dllmin[cnt] + dllmax[cnt]) >> 1;
-			if (gold_sadj[0] >= dlli) {
-				dlli = (gold_sadj[0] - dlli) >> 1;
-				if (dlli > 0) {
-					dlli = 1;
-				}
-				if (data2 != 3) {
-					data2 = (data2 + dlli) & 0x7;
-				}
-			} else {
-				dlli = (dlli - gold_sadj[0]) >> 1;
-				if (dlli > 0) {
-					dlli = 1;
-				}
-				if (data2 != 4) {
-					data2 = (data2 - dlli) & 0x7;
-				}
+	/* Search margin */
+	g_dqidly = g_dqsip = g_margin = g_side = 0;
+
+	for (dqidly = 0; dqidly < 32; dqidly++) {
+		for (dqsip = 0; dqsip < 2; dqsip++) {
+			if (pass[dqidly][dqsip][0] > pass[dqidly][dqsip][1])
+				continue;
+			diff = pass[dqidly][dqsip][1] - pass[dqidly][dqsip][0];
+			if ((diff+2) < g_margin)
+				continue;
+			passcnt[0] = passcnt[1] = 0;
+			for (dlli = pass[dqidly][dqsip][0]; dlli > 0  && tag[dqsip][dlli] != 0; dlli--, passcnt[0]++);
+			for (dlli = pass[dqidly][dqsip][1]; dlli < 76 && tag[dqsip][dlli] != 0; dlli++, passcnt[1]++);
+			if (passcnt[0] > passcnt[1])
+				passcnt[0] = passcnt[1];
+			passcnt[1] = 0;
+			if (passcnt[0] > g_side)
+				passcnt[1] = passcnt[0] - g_side;
+			if (diff > (g_margin+1) && (passcnt[1] > 0 || passcnt[0] > 8)) {
+				g_margin = diff;
+				g_dqidly = dqidly;
+				g_dqsip  = dqsip;
+				g_side   = passcnt[0];
+			} else if (passcnt[1] > 1 && g_side < 8) {
+				if (diff > g_margin)
+					g_margin = diff;
+				g_dqidly = dqidly;
+				g_dqsip  = dqsip;
+				g_side   = passcnt[0];
 			}
 		}
-		data |= data2 << 21;
 	}
-	moutdwm(ast, 0x1E6E0084, data);
-
-} /* finetuneDQI_L2 */
+	reg_mcr18 = reg_mcr18 | (g_dqidly << 16) | (g_dqsip << 23);
+	ast_moutdwm(ast, 0x1E6E0018, reg_mcr18);
 
-static void cbr_dll2(struct ast_private *ast, struct ast2300_dram_param *param)
+}
+static bool cbr_dll2(struct ast_private *ast, struct ast2300_dram_param *param)
 {
-	u32 dllmin[2], dllmax[2], dlli, data, data2, passcnt;
+	u32 dllmin[2], dllmax[2], dlli, data, passcnt, retry = 0;
+	bool status = false;
 
-
-	finetuneDQI_L(ast, param);
-	finetuneDQI_L2(ast, param);
+	finetuneDQSI(ast);
+	if (finetuneDQI_L(ast, param) == false)
+		return status;
 
 CBR_START2:
 	dllmin[0] = dllmin[1] = 0xff;
 	dllmax[0] = dllmax[1] = 0x0;
 	passcnt = 0;
 	for (dlli = 0; dlli < 76; dlli++) {
-		moutdwm(ast, 0x1E6E0068, 0x00001300 | (dlli << 16) | (dlli << 24));
-		/* Wait DQSI latch phase calibration */
-		moutdwm(ast, 0x1E6E0074, 0x00000010);
-		moutdwm(ast, 0x1E6E0070, 0x00000003);
-		do {
-			data = mindwm(ast, 0x1E6E0070);
-		} while (!(data & 0x00001000));
-		moutdwm(ast, 0x1E6E0070, 0x00000000);
-
-		moutdwm(ast, 0x1E6E0074, CBR_SIZE2);
+		ast_moutdwm(ast, 0x1E6E0068, 0x00001300 | (dlli << 16) | (dlli << 24));
+		ast_moutdwm(ast, 0x1E6E0074, CBR_SIZE2);
 		data = cbr_scan(ast);
 		if (data != 0) {
 			if (data & 0x1) {
@@ -957,44 +844,31 @@ CBR_START2:
 			break;
 		}
 	}
+	if (retry++ > 10)
+		goto CBR_DONE2;
 	if (dllmax[0] == 0 || (dllmax[0]-dllmin[0]) < CBR_THRESHOLD) {
 		goto CBR_START2;
 	}
 	if (dllmax[1] == 0 || (dllmax[1]-dllmin[1]) < CBR_THRESHOLD) {
 		goto CBR_START2;
 	}
+	status = true;
+CBR_DONE2:
 	dlli  = (dllmin[1] + dllmax[1]) >> 1;
 	dlli <<= 8;
 	dlli += (dllmin[0] + dllmax[0]) >> 1;
-	moutdwm(ast, 0x1E6E0068, (mindwm(ast, 0x1E6E0068) & 0xFFFF) | (dlli << 16));
-
-	data  = (mindwm(ast, 0x1E6E0080) >> 24) & 0x1F;
-	data2 = (mindwm(ast, 0x1E6E0018) & 0xff80ffff) | (data << 16);
-	moutdwm(ast, 0x1E6E0018, data2);
-	moutdwm(ast, 0x1E6E0024, 0x8001 | (data << 1) | (param->dll2_finetune_step << 8));
-
-	/* Wait DQSI latch phase calibration */
-	moutdwm(ast, 0x1E6E0074, 0x00000010);
-	moutdwm(ast, 0x1E6E0070, 0x00000003);
-	do {
-		data = mindwm(ast, 0x1E6E0070);
-	} while (!(data & 0x00001000));
-	moutdwm(ast, 0x1E6E0070, 0x00000000);
-	moutdwm(ast, 0x1E6E0070, 0x00000003);
-	do {
-		data = mindwm(ast, 0x1E6E0070);
-	} while (!(data & 0x00001000));
-	moutdwm(ast, 0x1E6E0070, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0068, ast_mindwm(ast, 0x1E720058) | (dlli << 16));
+	return status;
 } /* CBRDLL2 */
 
 static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *param)
 {
 	u32 trap, trap_AC2, trap_MRS;
 
-	moutdwm(ast, 0x1E6E2000, 0x1688A8A8);
+	ast_moutdwm(ast, 0x1E6E2000, 0x1688A8A8);
 
 	/* Ger trap info */
-	trap = (mindwm(ast, 0x1E6E2070) >> 25) & 0x3;
+	trap = (ast_mindwm(ast, 0x1E6E2070) >> 25) & 0x3;
 	trap_AC2  = 0x00020000 + (trap << 16);
 	trap_AC2 |= 0x00300000 + ((trap & 0x2) << 19);
 	trap_MRS  = 0x00000010 + (trap << 4);
@@ -1008,22 +882,35 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 
 	switch (param->dram_freq) {
 	case 336:
-		moutdwm(ast, 0x1E6E2020, 0x0190);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0190);
 		param->wodt          = 0;
 		param->reg_AC1       = 0x22202725;
 		param->reg_AC2       = 0xAA007613 | trap_AC2;
 		param->reg_DQSIC     = 0x000000BA;
 		param->reg_MRS       = 0x04001400 | trap_MRS;
 		param->reg_EMRS      = 0x00000000;
-		param->reg_IOZ       = 0x00000034;
+		param->reg_IOZ       = 0x00000023;
 		param->reg_DQIDLY    = 0x00000074;
 		param->reg_FREQ      = 0x00004DC0;
 		param->madj_max      = 96;
 		param->dll2_finetune_step = 3;
+		switch (param->dram_chipid) {
+		default:
+		case AST_DRAM_512Mx16:
+		case AST_DRAM_1Gx16:
+			param->reg_AC2   = 0xAA007613 | trap_AC2;
+			break;
+		case AST_DRAM_2Gx16:
+			param->reg_AC2   = 0xAA00761C | trap_AC2;
+			break;
+		case AST_DRAM_4Gx16:
+			param->reg_AC2   = 0xAA007636 | trap_AC2;
+			break;
+		}
 		break;
 	default:
 	case 396:
-		moutdwm(ast, 0x1E6E2020, 0x03F1);
+		ast_moutdwm(ast, 0x1E6E2020, 0x03F1);
 		param->wodt          = 1;
 		param->reg_AC1       = 0x33302825;
 		param->reg_AC2       = 0xCC009617 | trap_AC2;
@@ -1033,7 +920,7 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->reg_IOZ       = 0x00000034;
 		param->reg_DRV       = 0x000000FA;
 		param->reg_DQIDLY    = 0x00000089;
-		param->reg_FREQ      = 0x000050C0;
+		param->reg_FREQ      = 0x00005040;
 		param->madj_max      = 96;
 		param->dll2_finetune_step = 4;
 
@@ -1053,14 +940,14 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		break;
 
 	case 408:
-		moutdwm(ast, 0x1E6E2020, 0x01F0);
+		ast_moutdwm(ast, 0x1E6E2020, 0x01F0);
 		param->wodt          = 1;
 		param->reg_AC1       = 0x33302825;
 		param->reg_AC2       = 0xCC009617 | trap_AC2;
 		param->reg_DQSIC     = 0x000000E2;
 		param->reg_MRS       = 0x04001600 | trap_MRS;
 		param->reg_EMRS      = 0x00000000;
-		param->reg_IOZ       = 0x00000034;
+		param->reg_IOZ       = 0x00000023;
 		param->reg_DRV       = 0x000000FA;
 		param->reg_DQIDLY    = 0x00000089;
 		param->reg_FREQ      = 0x000050C0;
@@ -1083,7 +970,7 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 
 		break;
 	case 456:
-		moutdwm(ast, 0x1E6E2020, 0x0230);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0230);
 		param->wodt          = 0;
 		param->reg_AC1       = 0x33302926;
 		param->reg_AC2       = 0xCD44961A;
@@ -1097,7 +984,7 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 4;
 		break;
 	case 504:
-		moutdwm(ast, 0x1E6E2020, 0x0270);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0270);
 		param->wodt          = 1;
 		param->reg_AC1       = 0x33302926;
 		param->reg_AC2       = 0xDE44A61D;
@@ -1111,7 +998,7 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 4;
 		break;
 	case 528:
-		moutdwm(ast, 0x1E6E2020, 0x0290);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0290);
 		param->wodt          = 1;
 		param->rodt          = 1;
 		param->reg_AC1       = 0x33302926;
@@ -1127,7 +1014,7 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 3;
 		break;
 	case 576:
-		moutdwm(ast, 0x1E6E2020, 0x0140);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0140);
 		param->reg_MADJ      = 0x00136868;
 		param->reg_SADJ      = 0x00004534;
 		param->wodt          = 1;
@@ -1145,7 +1032,7 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 3;
 		break;
 	case 600:
-		moutdwm(ast, 0x1E6E2020, 0x02E1);
+		ast_moutdwm(ast, 0x1E6E2020, 0x02E1);
 		param->reg_MADJ      = 0x00136868;
 		param->reg_SADJ      = 0x00004534;
 		param->wodt          = 1;
@@ -1163,7 +1050,7 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 3;
 		break;
 	case 624:
-		moutdwm(ast, 0x1E6E2020, 0x0160);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0160);
 		param->reg_MADJ      = 0x00136868;
 		param->reg_SADJ      = 0x00004534;
 		param->wodt          = 1;
@@ -1196,7 +1083,7 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 	case AST_DRAM_4Gx16:
 		param->dram_config = 0x133;
 		break;
-	}; /* switch size */
+	} /* switch size */
 
 	switch (param->vram_size) {
 	default:
@@ -1218,106 +1105,98 @@ static void get_ddr3_info(struct ast_private *ast, struct ast2300_dram_param *pa
 
 static void ddr3_init(struct ast_private *ast, struct ast2300_dram_param *param)
 {
-	u32 data, data2;
+	u32 data, data2, retry = 0;
 
-	moutdwm(ast, 0x1E6E0000, 0xFC600309);
-	moutdwm(ast, 0x1E6E0018, 0x00000100);
-	moutdwm(ast, 0x1E6E0024, 0x00000000);
-	moutdwm(ast, 0x1E6E0034, 0x00000000);
+ddr3_init_start:
+	ast_moutdwm(ast, 0x1E6E0000, 0xFC600309);
+	ast_moutdwm(ast, 0x1E6E0018, 0x00000100);
+	ast_moutdwm(ast, 0x1E6E0024, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0034, 0x00000000);
 	udelay(10);
-	moutdwm(ast, 0x1E6E0064, param->reg_MADJ);
-	moutdwm(ast, 0x1E6E0068, param->reg_SADJ);
+	ast_moutdwm(ast, 0x1E6E0064, param->reg_MADJ);
+	ast_moutdwm(ast, 0x1E6E0068, param->reg_SADJ);
 	udelay(10);
-	moutdwm(ast, 0x1E6E0064, param->reg_MADJ | 0xC0000);
+	ast_moutdwm(ast, 0x1E6E0064, param->reg_MADJ | 0xC0000);
 	udelay(10);
 
-	moutdwm(ast, 0x1E6E0004, param->dram_config);
-	moutdwm(ast, 0x1E6E0008, 0x90040f);
-	moutdwm(ast, 0x1E6E0010, param->reg_AC1);
-	moutdwm(ast, 0x1E6E0014, param->reg_AC2);
-	moutdwm(ast, 0x1E6E0020, param->reg_DQSIC);
-	moutdwm(ast, 0x1E6E0080, 0x00000000);
-	moutdwm(ast, 0x1E6E0084, 0x00000000);
-	moutdwm(ast, 0x1E6E0088, param->reg_DQIDLY);
-	moutdwm(ast, 0x1E6E0018, 0x4040A170);
-	moutdwm(ast, 0x1E6E0018, 0x20402370);
-	moutdwm(ast, 0x1E6E0038, 0x00000000);
-	moutdwm(ast, 0x1E6E0040, 0xFF444444);
-	moutdwm(ast, 0x1E6E0044, 0x22222222);
-	moutdwm(ast, 0x1E6E0048, 0x22222222);
-	moutdwm(ast, 0x1E6E004C, 0x00000002);
-	moutdwm(ast, 0x1E6E0050, 0x80000000);
-	moutdwm(ast, 0x1E6E0050, 0x00000000);
-	moutdwm(ast, 0x1E6E0054, 0);
-	moutdwm(ast, 0x1E6E0060, param->reg_DRV);
-	moutdwm(ast, 0x1E6E006C, param->reg_IOZ);
-	moutdwm(ast, 0x1E6E0070, 0x00000000);
-	moutdwm(ast, 0x1E6E0074, 0x00000000);
-	moutdwm(ast, 0x1E6E0078, 0x00000000);
-	moutdwm(ast, 0x1E6E007C, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0004, param->dram_config);
+	ast_moutdwm(ast, 0x1E6E0008, 0x90040f);
+	ast_moutdwm(ast, 0x1E6E0010, param->reg_AC1);
+	ast_moutdwm(ast, 0x1E6E0014, param->reg_AC2);
+	ast_moutdwm(ast, 0x1E6E0020, param->reg_DQSIC);
+	ast_moutdwm(ast, 0x1E6E0080, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0084, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0088, param->reg_DQIDLY);
+	ast_moutdwm(ast, 0x1E6E0018, 0x4000A170);
+	ast_moutdwm(ast, 0x1E6E0018, 0x00002370);
+	ast_moutdwm(ast, 0x1E6E0038, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0040, 0xFF444444);
+	ast_moutdwm(ast, 0x1E6E0044, 0x22222222);
+	ast_moutdwm(ast, 0x1E6E0048, 0x22222222);
+	ast_moutdwm(ast, 0x1E6E004C, 0x00000002);
+	ast_moutdwm(ast, 0x1E6E0050, 0x80000000);
+	ast_moutdwm(ast, 0x1E6E0050, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0054, 0);
+	ast_moutdwm(ast, 0x1E6E0060, param->reg_DRV);
+	ast_moutdwm(ast, 0x1E6E006C, param->reg_IOZ);
+	ast_moutdwm(ast, 0x1E6E0070, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0074, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0078, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E007C, 0x00000000);
 	/* Wait MCLK2X lock to MCLK */
 	do {
-		data = mindwm(ast, 0x1E6E001C);
+		data = ast_mindwm(ast, 0x1E6E001C);
 	} while (!(data & 0x08000000));
-	moutdwm(ast, 0x1E6E0034, 0x00000001);
-	moutdwm(ast, 0x1E6E000C, 0x00005C04);
-	udelay(10);
-	moutdwm(ast, 0x1E6E000C, 0x00000000);
-	moutdwm(ast, 0x1E6E0034, 0x00000000);
-	data = mindwm(ast, 0x1E6E001C);
+	data = ast_mindwm(ast, 0x1E6E001C);
 	data = (data >> 8) & 0xff;
 	while ((data & 0x08) || ((data & 0x7) < 2) || (data < 4)) {
-		data2 = (mindwm(ast, 0x1E6E0064) & 0xfff3ffff) + 4;
+		data2 = (ast_mindwm(ast, 0x1E6E0064) & 0xfff3ffff) + 4;
 		if ((data2 & 0xff) > param->madj_max) {
 			break;
 		}
-		moutdwm(ast, 0x1E6E0064, data2);
+		ast_moutdwm(ast, 0x1E6E0064, data2);
 		if (data2 & 0x00100000) {
 			data2 = ((data2 & 0xff) >> 3) + 3;
 		} else {
 			data2 = ((data2 & 0xff) >> 2) + 5;
 		}
-		data = mindwm(ast, 0x1E6E0068) & 0xffff00ff;
+		data = ast_mindwm(ast, 0x1E6E0068) & 0xffff00ff;
 		data2 += data & 0xff;
 		data = data | (data2 << 8);
-		moutdwm(ast, 0x1E6E0068, data);
+		ast_moutdwm(ast, 0x1E6E0068, data);
 		udelay(10);
-		moutdwm(ast, 0x1E6E0064, mindwm(ast, 0x1E6E0064) | 0xC0000);
+		ast_moutdwm(ast, 0x1E6E0064, ast_mindwm(ast, 0x1E6E0064) | 0xC0000);
 		udelay(10);
-		data = mindwm(ast, 0x1E6E0018) & 0xfffff1ff;
-		moutdwm(ast, 0x1E6E0018, data);
+		data = ast_mindwm(ast, 0x1E6E0018) & 0xfffff1ff;
+		ast_moutdwm(ast, 0x1E6E0018, data);
 		data = data | 0x200;
-		moutdwm(ast, 0x1E6E0018, data);
+		ast_moutdwm(ast, 0x1E6E0018, data);
 		do {
-			data = mindwm(ast, 0x1E6E001C);
+			data = ast_mindwm(ast, 0x1E6E001C);
 		} while (!(data & 0x08000000));
 
-		moutdwm(ast, 0x1E6E0034, 0x00000001);
-		moutdwm(ast, 0x1E6E000C, 0x00005C04);
-		udelay(10);
-		moutdwm(ast, 0x1E6E000C, 0x00000000);
-		moutdwm(ast, 0x1E6E0034, 0x00000000);
-		data = mindwm(ast, 0x1E6E001C);
+		data = ast_mindwm(ast, 0x1E6E001C);
 		data = (data >> 8) & 0xff;
 	}
-	data = mindwm(ast, 0x1E6E0018) | 0xC00;
-	moutdwm(ast, 0x1E6E0018, data);
+	ast_moutdwm(ast, 0x1E720058, ast_mindwm(ast, 0x1E6E0068) & 0xffff);
+	data = ast_mindwm(ast, 0x1E6E0018) | 0xC00;
+	ast_moutdwm(ast, 0x1E6E0018, data);
 
-	moutdwm(ast, 0x1E6E0034, 0x00000001);
-	moutdwm(ast, 0x1E6E000C, 0x00000040);
+	ast_moutdwm(ast, 0x1E6E0034, 0x00000001);
+	ast_moutdwm(ast, 0x1E6E000C, 0x00000040);
 	udelay(50);
 	/* Mode Register Setting */
-	moutdwm(ast, 0x1E6E002C, param->reg_MRS | 0x100);
-	moutdwm(ast, 0x1E6E0030, param->reg_EMRS);
-	moutdwm(ast, 0x1E6E0028, 0x00000005);
-	moutdwm(ast, 0x1E6E0028, 0x00000007);
-	moutdwm(ast, 0x1E6E0028, 0x00000003);
-	moutdwm(ast, 0x1E6E0028, 0x00000001);
-	moutdwm(ast, 0x1E6E002C, param->reg_MRS);
-	moutdwm(ast, 0x1E6E000C, 0x00005C08);
-	moutdwm(ast, 0x1E6E0028, 0x00000001);
-
-	moutdwm(ast, 0x1E6E000C, 0x7FFF5C01);
+	ast_moutdwm(ast, 0x1E6E002C, param->reg_MRS | 0x100);
+	ast_moutdwm(ast, 0x1E6E0030, param->reg_EMRS);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000005);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000007);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000003);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000001);
+	ast_moutdwm(ast, 0x1E6E002C, param->reg_MRS);
+	ast_moutdwm(ast, 0x1E6E000C, 0x00005C08);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000001);
+
+	ast_moutdwm(ast, 0x1E6E000C, 0x00005C01);
 	data = 0;
 	if (param->wodt) {
 		data = 0x300;
@@ -1325,30 +1204,23 @@ static void ddr3_init(struct ast_private *ast, struct ast2300_dram_param *param)
 	if (param->rodt) {
 		data = data | 0x3000 | ((param->reg_AC2 & 0x60000) >> 3);
 	}
-	moutdwm(ast, 0x1E6E0034, data | 0x3);
+	ast_moutdwm(ast, 0x1E6E0034, data | 0x3);
 
-	/* Wait DQI delay lock */
-	do {
-		data = mindwm(ast, 0x1E6E0080);
-	} while (!(data & 0x40000000));
-	/* Wait DQSI delay lock */
-	do {
-		data = mindwm(ast, 0x1E6E0020);
-	} while (!(data & 0x00000800));
 	/* Calibrate the DQSI delay */
-	cbr_dll2(ast, param);
+	if ((cbr_dll2(ast, param) == false) && (retry++ < 10))
+		goto ddr3_init_start;
 
-	moutdwm(ast, 0x1E6E0120, param->reg_FREQ);
+	ast_moutdwm(ast, 0x1E6E0120, param->reg_FREQ);
 	/* ECC Memory Initialization */
 #ifdef ECC
-	moutdwm(ast, 0x1E6E007C, 0x00000000);
-	moutdwm(ast, 0x1E6E0070, 0x221);
+	ast_moutdwm(ast, 0x1E6E007C, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0070, 0x221);
 	do {
-		data = mindwm(ast, 0x1E6E0070);
+		data = ast_mindwm(ast, 0x1E6E0070);
 	} while (!(data & 0x00001000));
-	moutdwm(ast, 0x1E6E0070, 0x00000000);
-	moutdwm(ast, 0x1E6E0050, 0x80000000);
-	moutdwm(ast, 0x1E6E0050, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0070, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0050, 0x80000000);
+	ast_moutdwm(ast, 0x1E6E0050, 0x00000000);
 #endif
 
 
@@ -1358,10 +1230,10 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 {
 	u32 trap, trap_AC2, trap_MRS;
 
-	moutdwm(ast, 0x1E6E2000, 0x1688A8A8);
+	ast_moutdwm(ast, 0x1E6E2000, 0x1688A8A8);
 
 	/* Ger trap info */
-	trap = (mindwm(ast, 0x1E6E2070) >> 25) & 0x3;
+	trap = (ast_mindwm(ast, 0x1E6E2070) >> 25) & 0x3;
 	trap_AC2  = (trap << 20) | (trap << 16);
 	trap_AC2 += 0x00110000;
 	trap_MRS  = 0x00000040 | (trap << 4);
@@ -1375,7 +1247,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 
 	switch (param->dram_freq) {
 	case 264:
-		moutdwm(ast, 0x1E6E2020, 0x0130);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0130);
 		param->wodt          = 0;
 		param->reg_AC1       = 0x11101513;
 		param->reg_AC2       = 0x78117011;
@@ -1390,7 +1262,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 3;
 		break;
 	case 336:
-		moutdwm(ast, 0x1E6E2020, 0x0190);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0190);
 		param->wodt          = 1;
 		param->reg_AC1       = 0x22202613;
 		param->reg_AC2       = 0xAA009016 | trap_AC2;
@@ -1403,10 +1275,25 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->reg_FREQ      = 0x00004DC0;
 		param->madj_max      = 96;
 		param->dll2_finetune_step = 3;
+		switch (param->dram_chipid) {
+		default:
+		case AST_DRAM_512Mx16:
+			param->reg_AC2   = 0xAA009012 | trap_AC2;
+			break;
+		case AST_DRAM_1Gx16:
+			param->reg_AC2   = 0xAA009016 | trap_AC2;
+			break;
+		case AST_DRAM_2Gx16:
+			param->reg_AC2   = 0xAA009023 | trap_AC2;
+			break;
+		case AST_DRAM_4Gx16:
+			param->reg_AC2   = 0xAA00903B | trap_AC2;
+			break;
+		}
 		break;
 	default:
 	case 396:
-		moutdwm(ast, 0x1E6E2020, 0x03F1);
+		ast_moutdwm(ast, 0x1E6E2020, 0x03F1);
 		param->wodt          = 1;
 		param->rodt          = 0;
 		param->reg_AC1       = 0x33302714;
@@ -1417,7 +1304,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->reg_DRV       = 0x000000FA;
 		param->reg_IOZ       = 0x00000034;
 		param->reg_DQIDLY    = 0x00000089;
-		param->reg_FREQ      = 0x000050C0;
+		param->reg_FREQ      = 0x00005040;
 		param->madj_max      = 96;
 		param->dll2_finetune_step = 4;
 
@@ -1440,7 +1327,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		break;
 
 	case 408:
-		moutdwm(ast, 0x1E6E2020, 0x01F0);
+		ast_moutdwm(ast, 0x1E6E2020, 0x01F0);
 		param->wodt          = 1;
 		param->rodt          = 0;
 		param->reg_AC1       = 0x33302714;
@@ -1473,7 +1360,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 
 		break;
 	case 456:
-		moutdwm(ast, 0x1E6E2020, 0x0230);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0230);
 		param->wodt          = 0;
 		param->reg_AC1       = 0x33302815;
 		param->reg_AC2       = 0xCD44B01E;
@@ -1488,7 +1375,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 3;
 		break;
 	case 504:
-		moutdwm(ast, 0x1E6E2020, 0x0261);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0261);
 		param->wodt          = 1;
 		param->rodt          = 1;
 		param->reg_AC1       = 0x33302815;
@@ -1504,7 +1391,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 3;
 		break;
 	case 528:
-		moutdwm(ast, 0x1E6E2020, 0x0120);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0120);
 		param->wodt          = 1;
 		param->rodt          = 1;
 		param->reg_AC1       = 0x33302815;
@@ -1520,7 +1407,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 3;
 		break;
 	case 552:
-		moutdwm(ast, 0x1E6E2020, 0x02A1);
+		ast_moutdwm(ast, 0x1E6E2020, 0x02A1);
 		param->wodt          = 1;
 		param->rodt          = 1;
 		param->reg_AC1       = 0x43402915;
@@ -1536,7 +1423,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 		param->dll2_finetune_step = 3;
 		break;
 	case 576:
-		moutdwm(ast, 0x1E6E2020, 0x0140);
+		ast_moutdwm(ast, 0x1E6E2020, 0x0140);
 		param->wodt          = 1;
 		param->rodt          = 1;
 		param->reg_AC1       = 0x43402915;
@@ -1567,7 +1454,7 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 	case AST_DRAM_4Gx16:
 		param->dram_config = 0x123;
 		break;
-	}; /* switch size */
+	} /* switch size */
 
 	switch (param->vram_size) {
 	default:
@@ -1588,110 +1475,102 @@ static void get_ddr2_info(struct ast_private *ast, struct ast2300_dram_param *pa
 
 static void ddr2_init(struct ast_private *ast, struct ast2300_dram_param *param)
 {
-	u32 data, data2;
-
-	moutdwm(ast, 0x1E6E0000, 0xFC600309);
-	moutdwm(ast, 0x1E6E0018, 0x00000100);
-	moutdwm(ast, 0x1E6E0024, 0x00000000);
-	moutdwm(ast, 0x1E6E0064, param->reg_MADJ);
-	moutdwm(ast, 0x1E6E0068, param->reg_SADJ);
+	u32 data, data2, retry = 0;
+
+ddr2_init_start:
+	ast_moutdwm(ast, 0x1E6E0000, 0xFC600309);
+	ast_moutdwm(ast, 0x1E6E0018, 0x00000100);
+	ast_moutdwm(ast, 0x1E6E0024, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0064, param->reg_MADJ);
+	ast_moutdwm(ast, 0x1E6E0068, param->reg_SADJ);
 	udelay(10);
-	moutdwm(ast, 0x1E6E0064, param->reg_MADJ | 0xC0000);
+	ast_moutdwm(ast, 0x1E6E0064, param->reg_MADJ | 0xC0000);
 	udelay(10);
 
-	moutdwm(ast, 0x1E6E0004, param->dram_config);
-	moutdwm(ast, 0x1E6E0008, 0x90040f);
-	moutdwm(ast, 0x1E6E0010, param->reg_AC1);
-	moutdwm(ast, 0x1E6E0014, param->reg_AC2);
-	moutdwm(ast, 0x1E6E0020, param->reg_DQSIC);
-	moutdwm(ast, 0x1E6E0080, 0x00000000);
-	moutdwm(ast, 0x1E6E0084, 0x00000000);
-	moutdwm(ast, 0x1E6E0088, param->reg_DQIDLY);
-	moutdwm(ast, 0x1E6E0018, 0x4040A130);
-	moutdwm(ast, 0x1E6E0018, 0x20402330);
-	moutdwm(ast, 0x1E6E0038, 0x00000000);
-	moutdwm(ast, 0x1E6E0040, 0xFF808000);
-	moutdwm(ast, 0x1E6E0044, 0x88848466);
-	moutdwm(ast, 0x1E6E0048, 0x44440008);
-	moutdwm(ast, 0x1E6E004C, 0x00000000);
-	moutdwm(ast, 0x1E6E0050, 0x80000000);
-	moutdwm(ast, 0x1E6E0050, 0x00000000);
-	moutdwm(ast, 0x1E6E0054, 0);
-	moutdwm(ast, 0x1E6E0060, param->reg_DRV);
-	moutdwm(ast, 0x1E6E006C, param->reg_IOZ);
-	moutdwm(ast, 0x1E6E0070, 0x00000000);
-	moutdwm(ast, 0x1E6E0074, 0x00000000);
-	moutdwm(ast, 0x1E6E0078, 0x00000000);
-	moutdwm(ast, 0x1E6E007C, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0004, param->dram_config);
+	ast_moutdwm(ast, 0x1E6E0008, 0x90040f);
+	ast_moutdwm(ast, 0x1E6E0010, param->reg_AC1);
+	ast_moutdwm(ast, 0x1E6E0014, param->reg_AC2);
+	ast_moutdwm(ast, 0x1E6E0020, param->reg_DQSIC);
+	ast_moutdwm(ast, 0x1E6E0080, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0084, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0088, param->reg_DQIDLY);
+	ast_moutdwm(ast, 0x1E6E0018, 0x4000A130);
+	ast_moutdwm(ast, 0x1E6E0018, 0x00002330);
+	ast_moutdwm(ast, 0x1E6E0038, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0040, 0xFF808000);
+	ast_moutdwm(ast, 0x1E6E0044, 0x88848466);
+	ast_moutdwm(ast, 0x1E6E0048, 0x44440008);
+	ast_moutdwm(ast, 0x1E6E004C, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0050, 0x80000000);
+	ast_moutdwm(ast, 0x1E6E0050, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0054, 0);
+	ast_moutdwm(ast, 0x1E6E0060, param->reg_DRV);
+	ast_moutdwm(ast, 0x1E6E006C, param->reg_IOZ);
+	ast_moutdwm(ast, 0x1E6E0070, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0074, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0078, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E007C, 0x00000000);
 
 	/* Wait MCLK2X lock to MCLK */
 	do {
-		data = mindwm(ast, 0x1E6E001C);
+		data = ast_mindwm(ast, 0x1E6E001C);
 	} while (!(data & 0x08000000));
-	moutdwm(ast, 0x1E6E0034, 0x00000001);
-	moutdwm(ast, 0x1E6E000C, 0x00005C04);
-	udelay(10);
-	moutdwm(ast, 0x1E6E000C, 0x00000000);
-	moutdwm(ast, 0x1E6E0034, 0x00000000);
-	data = mindwm(ast, 0x1E6E001C);
+	data = ast_mindwm(ast, 0x1E6E001C);
 	data = (data >> 8) & 0xff;
 	while ((data & 0x08) || ((data & 0x7) < 2) || (data < 4)) {
-		data2 = (mindwm(ast, 0x1E6E0064) & 0xfff3ffff) + 4;
+		data2 = (ast_mindwm(ast, 0x1E6E0064) & 0xfff3ffff) + 4;
 		if ((data2 & 0xff) > param->madj_max) {
 			break;
 		}
-		moutdwm(ast, 0x1E6E0064, data2);
+		ast_moutdwm(ast, 0x1E6E0064, data2);
 		if (data2 & 0x00100000) {
 			data2 = ((data2 & 0xff) >> 3) + 3;
 		} else {
 			data2 = ((data2 & 0xff) >> 2) + 5;
 		}
-		data = mindwm(ast, 0x1E6E0068) & 0xffff00ff;
+		data = ast_mindwm(ast, 0x1E6E0068) & 0xffff00ff;
 		data2 += data & 0xff;
 		data = data | (data2 << 8);
-		moutdwm(ast, 0x1E6E0068, data);
+		ast_moutdwm(ast, 0x1E6E0068, data);
 		udelay(10);
-		moutdwm(ast, 0x1E6E0064, mindwm(ast, 0x1E6E0064) | 0xC0000);
+		ast_moutdwm(ast, 0x1E6E0064, ast_mindwm(ast, 0x1E6E0064) | 0xC0000);
 		udelay(10);
-		data = mindwm(ast, 0x1E6E0018) & 0xfffff1ff;
-		moutdwm(ast, 0x1E6E0018, data);
+		data = ast_mindwm(ast, 0x1E6E0018) & 0xfffff1ff;
+		ast_moutdwm(ast, 0x1E6E0018, data);
 		data = data | 0x200;
-		moutdwm(ast, 0x1E6E0018, data);
+		ast_moutdwm(ast, 0x1E6E0018, data);
 		do {
-			data = mindwm(ast, 0x1E6E001C);
+			data = ast_mindwm(ast, 0x1E6E001C);
 		} while (!(data & 0x08000000));
 
-		moutdwm(ast, 0x1E6E0034, 0x00000001);
-		moutdwm(ast, 0x1E6E000C, 0x00005C04);
-		udelay(10);
-		moutdwm(ast, 0x1E6E000C, 0x00000000);
-		moutdwm(ast, 0x1E6E0034, 0x00000000);
-		data = mindwm(ast, 0x1E6E001C);
+		data = ast_mindwm(ast, 0x1E6E001C);
 		data = (data >> 8) & 0xff;
 	}
-	data = mindwm(ast, 0x1E6E0018) | 0xC00;
-	moutdwm(ast, 0x1E6E0018, data);
+	ast_moutdwm(ast, 0x1E720058, ast_mindwm(ast, 0x1E6E0008) & 0xffff);
+	data = ast_mindwm(ast, 0x1E6E0018) | 0xC00;
+	ast_moutdwm(ast, 0x1E6E0018, data);
 
-	moutdwm(ast, 0x1E6E0034, 0x00000001);
-	moutdwm(ast, 0x1E6E000C, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0034, 0x00000001);
+	ast_moutdwm(ast, 0x1E6E000C, 0x00000000);
 	udelay(50);
 	/* Mode Register Setting */
-	moutdwm(ast, 0x1E6E002C, param->reg_MRS | 0x100);
-	moutdwm(ast, 0x1E6E0030, param->reg_EMRS);
-	moutdwm(ast, 0x1E6E0028, 0x00000005);
-	moutdwm(ast, 0x1E6E0028, 0x00000007);
-	moutdwm(ast, 0x1E6E0028, 0x00000003);
-	moutdwm(ast, 0x1E6E0028, 0x00000001);
-
-	moutdwm(ast, 0x1E6E000C, 0x00005C08);
-	moutdwm(ast, 0x1E6E002C, param->reg_MRS);
-	moutdwm(ast, 0x1E6E0028, 0x00000001);
-	moutdwm(ast, 0x1E6E0030, param->reg_EMRS | 0x380);
-	moutdwm(ast, 0x1E6E0028, 0x00000003);
-	moutdwm(ast, 0x1E6E0030, param->reg_EMRS);
-	moutdwm(ast, 0x1E6E0028, 0x00000003);
-
-	moutdwm(ast, 0x1E6E000C, 0x7FFF5C01);
+	ast_moutdwm(ast, 0x1E6E002C, param->reg_MRS | 0x100);
+	ast_moutdwm(ast, 0x1E6E0030, param->reg_EMRS);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000005);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000007);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000003);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000001);
+
+	ast_moutdwm(ast, 0x1E6E000C, 0x00005C08);
+	ast_moutdwm(ast, 0x1E6E002C, param->reg_MRS);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000001);
+	ast_moutdwm(ast, 0x1E6E0030, param->reg_EMRS | 0x380);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000003);
+	ast_moutdwm(ast, 0x1E6E0030, param->reg_EMRS);
+	ast_moutdwm(ast, 0x1E6E0028, 0x00000003);
+
+	ast_moutdwm(ast, 0x1E6E000C, 0x7FFF5C01);
 	data = 0;
 	if (param->wodt) {
 		data = 0x500;
@@ -1699,30 +1578,23 @@ static void ddr2_init(struct ast_private *ast, struct ast2300_dram_param *param)
 	if (param->rodt) {
 		data = data | 0x3000 | ((param->reg_AC2 & 0x60000) >> 3);
 	}
-	moutdwm(ast, 0x1E6E0034, data | 0x3);
-	moutdwm(ast, 0x1E6E0120, param->reg_FREQ);
+	ast_moutdwm(ast, 0x1E6E0034, data | 0x3);
+	ast_moutdwm(ast, 0x1E6E0120, param->reg_FREQ);
 
-	/* Wait DQI delay lock */
-	do {
-		data = mindwm(ast, 0x1E6E0080);
-	} while (!(data & 0x40000000));
-	/* Wait DQSI delay lock */
-	do {
-		data = mindwm(ast, 0x1E6E0020);
-	} while (!(data & 0x00000800));
 	/* Calibrate the DQSI delay */
-	cbr_dll2(ast, param);
+	if ((cbr_dll2(ast, param) == false) && (retry++ < 10))
+		goto ddr2_init_start;
 
 	/* ECC Memory Initialization */
 #ifdef ECC
-	moutdwm(ast, 0x1E6E007C, 0x00000000);
-	moutdwm(ast, 0x1E6E0070, 0x221);
+	ast_moutdwm(ast, 0x1E6E007C, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0070, 0x221);
 	do {
-		data = mindwm(ast, 0x1E6E0070);
+		data = ast_mindwm(ast, 0x1E6E0070);
 	} while (!(data & 0x00001000));
-	moutdwm(ast, 0x1E6E0070, 0x00000000);
-	moutdwm(ast, 0x1E6E0050, 0x80000000);
-	moutdwm(ast, 0x1E6E0050, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0070, 0x00000000);
+	ast_moutdwm(ast, 0x1E6E0050, 0x80000000);
+	ast_moutdwm(ast, 0x1E6E0050, 0x00000000);
 #endif
 
 }
@@ -1768,8 +1640,8 @@ static void ast_init_dram_2300(struct drm_device *dev)
 			ddr2_init(ast, &param);
 		}
 
-		temp = mindwm(ast, 0x1e6e2040);
-		moutdwm(ast, 0x1e6e2040, temp | 0x40);
+		temp = ast_mindwm(ast, 0x1e6e2040);
+		ast_moutdwm(ast, 0x1e6e2040, temp | 0x40);
 	}
 
 	/* wait ready */
diff --git a/drivers/gpu/drm/ast/ast_tables.h b/drivers/gpu/drm/ast/ast_tables.h
index 95fa6aba26b..4c761dcea97 100644
--- a/drivers/gpu/drm/ast/ast_tables.h
+++ b/drivers/gpu/drm/ast/ast_tables.h
@@ -42,7 +42,7 @@
 #define HBorder                 0x00000020
 #define VBorder                 0x00000010
 #define WideScreenMode		0x00000100
-
+#define NewModeInfo		0x00000200
 
 /* DCLK Index */
 #define VCLK25_175     		0x00
@@ -67,6 +67,11 @@
 #define VCLK106_5   		0x12
 #define VCLK146_25  		0x13
 #define VCLK148_5   		0x14
+#define VCLK71      		0x15
+#define VCLK88_75   		0x16
+#define VCLK119     		0x17
+#define VCLK85_5     		0x18
+#define VCLK97_75     		0x19
 
 static struct ast_vbios_dclk_info dclk_table[] = {
 	{0x2C, 0xE7, 0x03},					/* 00: VCLK25_175	*/
@@ -90,6 +95,10 @@ static struct ast_vbios_dclk_info dclk_table[] = {
 	{0x28, 0x49, 0x80},					/* 12: VCLK106.5        */
 	{0x37, 0x49, 0x80},					/* 13: VCLK146.25       */
 	{0x1f, 0x45, 0x80},					/* 14: VCLK148.5        */
+	{0x47, 0x6c, 0x80},					/* 15: VCLK71       */
+	{0x25, 0x65, 0x80},					/* 16: VCLK88.75    */
+	{0x77, 0x58, 0x80},					/* 17: VCLK119      */
+	{0x32, 0x67, 0x80},				    /* 18: VCLK85_5     */
 };
 
 static struct ast_vbios_stdtable vbios_stdtable[] = {
@@ -225,41 +234,63 @@ static struct ast_vbios_enhtable res_1600x1200[] = {
 	 (SyncPP | Charx8Dot), 0xFF, 1, 0x33 },
 };
 
-static struct ast_vbios_enhtable res_1920x1200[] = {
-	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz */
-	 (SyncNP | Charx8Dot), 60, 1, 0x34 },
-	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz */
-	 (SyncNP | Charx8Dot), 0xFF, 1, 0x34 },
+/* 16:9 */
+static struct ast_vbios_enhtable res_1360x768[] = {
+	{1792, 1360, 64,112, 795,  768, 3, 6, VCLK85_5,	         /* 60Hz */
+	 (SyncPP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x39 },
+	{1792, 1360, 64,112, 795,  768, 3, 6, VCLK85_5,	         /* end */
+	 (SyncPP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x39 },
+};
+
+static struct ast_vbios_enhtable res_1600x900[] = {
+	{1760, 1600, 48, 32, 926,  900, 3, 5, VCLK97_75,	/* 60Hz CVT RB */
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x3A },
+	{1760, 1600, 48, 32, 926,  900, 3, 5, VCLK97_75,	/* end */
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x3A }
 };
 
+static struct ast_vbios_enhtable res_1920x1080[] = {
+	{2200, 1920, 88, 44, 1125, 1080, 4, 5, VCLK148_5,	/* 60Hz */
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x38 },
+	{2200, 1920, 88, 44, 1125, 1080, 4, 5, VCLK148_5,	/* 60Hz */
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x38 },
+};
+
+
 /* 16:10 */
 static struct ast_vbios_enhtable res_1280x800[] = {
+	{1440, 1280, 48, 32,  823,  800, 3, 6, VCLK71,	/* 60Hz RB */
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 35 },
 	{1680, 1280, 72,128,  831,  800, 3, 6, VCLK83_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode), 60, 1, 0x35 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x35 },
 	{1680, 1280, 72,128,  831,  800, 3, 6, VCLK83_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode), 0xFF, 1, 0x35 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x35 },
 
 };
 
 static struct ast_vbios_enhtable res_1440x900[] = {
+	{1600, 1440, 48, 32,  926,  900, 3, 6, VCLK88_75,	/* 60Hz RB */
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x36 },
 	{1904, 1440, 80,152,  934,  900, 3, 6, VCLK106_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode), 60, 1, 0x36 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x36 },
 	{1904, 1440, 80,152,  934,  900, 3, 6, VCLK106_5,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode), 0xFF, 1, 0x36 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x36 },
 };
 
 static struct ast_vbios_enhtable res_1680x1050[] = {
+	{1840, 1680, 48, 32, 1080, 1050, 3, 6, VCLK119,	/* 60Hz RB */
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x37 },
 	{2240, 1680,104,176, 1089, 1050, 3, 6, VCLK146_25,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode), 60, 1, 0x37 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x37 },
 	{2240, 1680,104,176, 1089, 1050, 3, 6, VCLK146_25,	/* 60Hz */
-	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode), 0xFF, 1, 0x37 },
+	 (SyncPN | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x37 },
 };
 
-/* HDTV */
-static struct ast_vbios_enhtable res_1920x1080[] = {
-	{2200, 1920, 88, 44, 1125, 1080, 4, 5, VCLK148_5,	/* 60Hz */
-	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode), 60, 1, 0x38 },
-	{2200, 1920, 88, 44, 1125, 1080, 4, 5, VCLK148_5,	/* 60Hz */
-	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode), 0xFF, 1, 0x38 },
+static struct ast_vbios_enhtable res_1920x1200[] = {
+	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz */
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 60, 1, 0x34 },
+	{2080, 1920, 48, 32, 1235, 1200, 3, 6, VCLK154,	/* 60Hz */
+	 (SyncNP | Charx8Dot | LineCompareOff | WideScreenMode | NewModeInfo), 0xFF, 1, 0x34 },
 };
+
 #endif
diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c
index 32aecb34dbc..b8246227bab 100644
--- a/drivers/gpu/drm/ast/ast_ttm.c
+++ b/drivers/gpu/drm/ast/ast_ttm.c
@@ -80,7 +80,7 @@ static int ast_ttm_global_init(struct ast_private *ast)
 	return 0;
 }
 
-void
+static void
 ast_ttm_global_release(struct ast_private *ast)
 {
 	if (ast->ttm.mem_global_ref.release == NULL)
@@ -102,7 +102,7 @@ static void ast_bo_ttm_destroy(struct ttm_buffer_object *tbo)
 	kfree(bo);
 }
 
-bool ast_ttm_bo_is_ast_bo(struct ttm_buffer_object *bo)
+static bool ast_ttm_bo_is_ast_bo(struct ttm_buffer_object *bo)
 {
 	if (bo->destroy == &ast_bo_ttm_destroy)
 		return true;
@@ -208,7 +208,7 @@ static struct ttm_backend_func ast_tt_backend_func = {
 };
 
 
-struct ttm_tt *ast_ttm_tt_create(struct ttm_bo_device *bdev,
+static struct ttm_tt *ast_ttm_tt_create(struct ttm_bo_device *bdev,
 				 unsigned long size, uint32_t page_flags,
 				 struct page *dummy_read_page)
 {
@@ -259,7 +259,9 @@ int ast_mm_init(struct ast_private *ast)
 
 	ret = ttm_bo_device_init(&ast->ttm.bdev,
 				 ast->ttm.bo_global_ref.ref.object,
-				 &ast_bo_driver, DRM_FILE_PAGE_OFFSET,
+				 &ast_bo_driver,
+				 dev->anon_inode->i_mapping,
+				 DRM_FILE_PAGE_OFFSET,
 				 true);
 	if (ret) {
 		DRM_ERROR("Error initialising bo driver; %d\n", ret);
@@ -324,7 +326,6 @@ int ast_bo_create(struct drm_device *dev, int size, int align,
 	}
 
 	astbo->bo.bdev = &ast->ttm.bdev;
-	astbo->bo.bdev->dev_mapping = dev->dev_mapping;
 
 	ast_ttm_placement(astbo, TTM_PL_FLAG_VRAM | TTM_PL_FLAG_SYSTEM);
 
diff --git a/drivers/gpu/drm/bochs/Kconfig b/drivers/gpu/drm/bochs/Kconfig
new file mode 100644
index 00000000000..5f8b0c2b9a4
--- /dev/null
+++ b/drivers/gpu/drm/bochs/Kconfig
@@ -0,0 +1,12 @@
+config DRM_BOCHS
+	tristate "DRM Support for bochs dispi vga interface (qemu stdvga)"
+	depends on DRM && PCI
+	select DRM_KMS_HELPER
+	select DRM_KMS_FB_HELPER
+	select FB_SYS_FILLRECT
+	select FB_SYS_COPYAREA
+	select FB_SYS_IMAGEBLIT
+	select DRM_TTM
+	help
+	  Choose this option for qemu.
+	  If M is selected the module will be called bochs-drm.
diff --git a/drivers/gpu/drm/bochs/Makefile b/drivers/gpu/drm/bochs/Makefile
new file mode 100644
index 00000000000..844a5561492
--- /dev/null
+++ b/drivers/gpu/drm/bochs/Makefile
@@ -0,0 +1,4 @@
+ccflags-y := -Iinclude/drm
+bochs-drm-y := bochs_drv.o bochs_mm.o bochs_kms.o bochs_fbdev.o bochs_hw.o
+
+obj-$(CONFIG_DRM_BOCHS)	+= bochs-drm.o
diff --git a/drivers/gpu/drm/bochs/bochs.h b/drivers/gpu/drm/bochs/bochs.h
new file mode 100644
index 00000000000..7eb52dd44b0
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs.h
@@ -0,0 +1,163 @@
+#include <linux/io.h>
+#include <linux/fb.h>
+#include <linux/console.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_helper.h>
+
+#include <ttm/ttm_bo_driver.h>
+#include <ttm/ttm_page_alloc.h>
+
+/* ---------------------------------------------------------------------- */
+
+#define VBE_DISPI_IOPORT_INDEX           0x01CE
+#define VBE_DISPI_IOPORT_DATA            0x01CF
+
+#define VBE_DISPI_INDEX_ID               0x0
+#define VBE_DISPI_INDEX_XRES             0x1
+#define VBE_DISPI_INDEX_YRES             0x2
+#define VBE_DISPI_INDEX_BPP              0x3
+#define VBE_DISPI_INDEX_ENABLE           0x4
+#define VBE_DISPI_INDEX_BANK             0x5
+#define VBE_DISPI_INDEX_VIRT_WIDTH       0x6
+#define VBE_DISPI_INDEX_VIRT_HEIGHT      0x7
+#define VBE_DISPI_INDEX_X_OFFSET         0x8
+#define VBE_DISPI_INDEX_Y_OFFSET         0x9
+#define VBE_DISPI_INDEX_VIDEO_MEMORY_64K 0xa
+
+#define VBE_DISPI_ID0                    0xB0C0
+#define VBE_DISPI_ID1                    0xB0C1
+#define VBE_DISPI_ID2                    0xB0C2
+#define VBE_DISPI_ID3                    0xB0C3
+#define VBE_DISPI_ID4                    0xB0C4
+#define VBE_DISPI_ID5                    0xB0C5
+
+#define VBE_DISPI_DISABLED               0x00
+#define VBE_DISPI_ENABLED                0x01
+#define VBE_DISPI_GETCAPS                0x02
+#define VBE_DISPI_8BIT_DAC               0x20
+#define VBE_DISPI_LFB_ENABLED            0x40
+#define VBE_DISPI_NOCLEARMEM             0x80
+
+/* ---------------------------------------------------------------------- */
+
+enum bochs_types {
+	BOCHS_QEMU_STDVGA,
+	BOCHS_UNKNOWN,
+};
+
+struct bochs_framebuffer {
+	struct drm_framebuffer base;
+	struct drm_gem_object *obj;
+};
+
+struct bochs_device {
+	/* hw */
+	void __iomem   *mmio;
+	int            ioports;
+	void __iomem   *fb_map;
+	unsigned long  fb_base;
+	unsigned long  fb_size;
+
+	/* mode */
+	u16 xres;
+	u16 yres;
+	u16 yres_virtual;
+	u32 stride;
+	u32 bpp;
+
+	/* drm */
+	struct drm_device  *dev;
+	struct drm_crtc crtc;
+	struct drm_encoder encoder;
+	struct drm_connector connector;
+	bool mode_config_initialized;
+
+	/* ttm */
+	struct {
+		struct drm_global_reference mem_global_ref;
+		struct ttm_bo_global_ref bo_global_ref;
+		struct ttm_bo_device bdev;
+		bool initialized;
+	} ttm;
+
+	/* fbdev */
+	struct {
+		struct bochs_framebuffer gfb;
+		struct drm_fb_helper helper;
+		int size;
+		bool initialized;
+	} fb;
+};
+
+#define to_bochs_framebuffer(x) container_of(x, struct bochs_framebuffer, base)
+
+struct bochs_bo {
+	struct ttm_buffer_object bo;
+	struct ttm_placement placement;
+	struct ttm_bo_kmap_obj kmap;
+	struct drm_gem_object gem;
+	u32 placements[3];
+	int pin_count;
+};
+
+static inline struct bochs_bo *bochs_bo(struct ttm_buffer_object *bo)
+{
+	return container_of(bo, struct bochs_bo, bo);
+}
+
+static inline struct bochs_bo *gem_to_bochs_bo(struct drm_gem_object *gem)
+{
+	return container_of(gem, struct bochs_bo, gem);
+}
+
+#define DRM_FILE_PAGE_OFFSET (0x100000000ULL >> PAGE_SHIFT)
+
+static inline u64 bochs_bo_mmap_offset(struct bochs_bo *bo)
+{
+	return drm_vma_node_offset_addr(&bo->bo.vma_node);
+}
+
+/* ---------------------------------------------------------------------- */
+
+/* bochs_hw.c */
+int bochs_hw_init(struct drm_device *dev, uint32_t flags);
+void bochs_hw_fini(struct drm_device *dev);
+
+void bochs_hw_setmode(struct bochs_device *bochs,
+		      struct drm_display_mode *mode);
+void bochs_hw_setbase(struct bochs_device *bochs,
+		      int x, int y, u64 addr);
+
+/* bochs_mm.c */
+int bochs_mm_init(struct bochs_device *bochs);
+void bochs_mm_fini(struct bochs_device *bochs);
+int bochs_mmap(struct file *filp, struct vm_area_struct *vma);
+
+int bochs_gem_create(struct drm_device *dev, u32 size, bool iskernel,
+		     struct drm_gem_object **obj);
+int bochs_gem_init_object(struct drm_gem_object *obj);
+void bochs_gem_free_object(struct drm_gem_object *obj);
+int bochs_dumb_create(struct drm_file *file, struct drm_device *dev,
+		      struct drm_mode_create_dumb *args);
+int bochs_dumb_mmap_offset(struct drm_file *file, struct drm_device *dev,
+			   uint32_t handle, uint64_t *offset);
+
+int bochs_framebuffer_init(struct drm_device *dev,
+			   struct bochs_framebuffer *gfb,
+			   struct drm_mode_fb_cmd2 *mode_cmd,
+			   struct drm_gem_object *obj);
+int bochs_bo_pin(struct bochs_bo *bo, u32 pl_flag, u64 *gpu_addr);
+int bochs_bo_unpin(struct bochs_bo *bo);
+
+extern const struct drm_mode_config_funcs bochs_mode_funcs;
+
+/* bochs_kms.c */
+int bochs_kms_init(struct bochs_device *bochs);
+void bochs_kms_fini(struct bochs_device *bochs);
+
+/* bochs_fbdev.c */
+int bochs_fbdev_init(struct bochs_device *bochs);
+void bochs_fbdev_fini(struct bochs_device *bochs);
diff --git a/drivers/gpu/drm/bochs/bochs_drv.c b/drivers/gpu/drm/bochs/bochs_drv.c
new file mode 100644
index 00000000000..9c13df29fd2
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_drv.c
@@ -0,0 +1,222 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+
+#include "bochs.h"
+
+static bool enable_fbdev = true;
+module_param_named(fbdev, enable_fbdev, bool, 0444);
+MODULE_PARM_DESC(fbdev, "register fbdev device");
+
+/* ---------------------------------------------------------------------- */
+/* drm interface                                                          */
+
+static int bochs_unload(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+
+	bochs_fbdev_fini(bochs);
+	bochs_kms_fini(bochs);
+	bochs_mm_fini(bochs);
+	bochs_hw_fini(dev);
+	kfree(bochs);
+	dev->dev_private = NULL;
+	return 0;
+}
+
+static int bochs_load(struct drm_device *dev, unsigned long flags)
+{
+	struct bochs_device *bochs;
+	int ret;
+
+	bochs = kzalloc(sizeof(*bochs), GFP_KERNEL);
+	if (bochs == NULL)
+		return -ENOMEM;
+	dev->dev_private = bochs;
+	bochs->dev = dev;
+
+	ret = bochs_hw_init(dev, flags);
+	if (ret)
+		goto err;
+
+	ret = bochs_mm_init(bochs);
+	if (ret)
+		goto err;
+
+	ret = bochs_kms_init(bochs);
+	if (ret)
+		goto err;
+
+	if (enable_fbdev)
+		bochs_fbdev_init(bochs);
+
+	return 0;
+
+err:
+	bochs_unload(dev);
+	return ret;
+}
+
+static const struct file_operations bochs_fops = {
+	.owner		= THIS_MODULE,
+	.open		= drm_open,
+	.release	= drm_release,
+	.unlocked_ioctl	= drm_ioctl,
+#ifdef CONFIG_COMPAT
+	.compat_ioctl	= drm_compat_ioctl,
+#endif
+	.poll		= drm_poll,
+	.read		= drm_read,
+	.llseek		= no_llseek,
+	.mmap           = bochs_mmap,
+};
+
+static struct drm_driver bochs_driver = {
+	.driver_features	= DRIVER_GEM | DRIVER_MODESET,
+	.load			= bochs_load,
+	.unload			= bochs_unload,
+	.fops			= &bochs_fops,
+	.name			= "bochs-drm",
+	.desc			= "bochs dispi vga interface (qemu stdvga)",
+	.date			= "20130925",
+	.major			= 1,
+	.minor			= 0,
+	.gem_free_object        = bochs_gem_free_object,
+	.dumb_create            = bochs_dumb_create,
+	.dumb_map_offset        = bochs_dumb_mmap_offset,
+	.dumb_destroy           = drm_gem_dumb_destroy,
+};
+
+/* ---------------------------------------------------------------------- */
+/* pm interface                                                           */
+
+static int bochs_pm_suspend(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	struct bochs_device *bochs = drm_dev->dev_private;
+
+	drm_kms_helper_poll_disable(drm_dev);
+
+	if (bochs->fb.initialized) {
+		console_lock();
+		fb_set_suspend(bochs->fb.helper.fbdev, 1);
+		console_unlock();
+	}
+
+	return 0;
+}
+
+static int bochs_pm_resume(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	struct bochs_device *bochs = drm_dev->dev_private;
+
+	drm_helper_resume_force_mode(drm_dev);
+
+	if (bochs->fb.initialized) {
+		console_lock();
+		fb_set_suspend(bochs->fb.helper.fbdev, 0);
+		console_unlock();
+	}
+
+	drm_kms_helper_poll_enable(drm_dev);
+	return 0;
+}
+
+static const struct dev_pm_ops bochs_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(bochs_pm_suspend,
+				bochs_pm_resume)
+};
+
+/* ---------------------------------------------------------------------- */
+/* pci interface                                                          */
+
+static int bochs_kick_out_firmware_fb(struct pci_dev *pdev)
+{
+	struct apertures_struct *ap;
+
+	ap = alloc_apertures(1);
+	if (!ap)
+		return -ENOMEM;
+
+	ap->ranges[0].base = pci_resource_start(pdev, 0);
+	ap->ranges[0].size = pci_resource_len(pdev, 0);
+	remove_conflicting_framebuffers(ap, "bochsdrmfb", false);
+	kfree(ap);
+
+	return 0;
+}
+
+static int bochs_pci_probe(struct pci_dev *pdev,
+			   const struct pci_device_id *ent)
+{
+	int ret;
+
+	ret = bochs_kick_out_firmware_fb(pdev);
+	if (ret)
+		return ret;
+
+	return drm_get_pci_dev(pdev, ent, &bochs_driver);
+}
+
+static void bochs_pci_remove(struct pci_dev *pdev)
+{
+	struct drm_device *dev = pci_get_drvdata(pdev);
+
+	drm_put_dev(dev);
+}
+
+static DEFINE_PCI_DEVICE_TABLE(bochs_pci_tbl) = {
+	{
+		.vendor      = 0x1234,
+		.device      = 0x1111,
+		.subvendor   = 0x1af4,
+		.subdevice   = 0x1100,
+		.driver_data = BOCHS_QEMU_STDVGA,
+	},
+	{
+		.vendor      = 0x1234,
+		.device      = 0x1111,
+		.subvendor   = PCI_ANY_ID,
+		.subdevice   = PCI_ANY_ID,
+		.driver_data = BOCHS_UNKNOWN,
+	},
+	{ /* end of list */ }
+};
+
+static struct pci_driver bochs_pci_driver = {
+	.name =		"bochs-drm",
+	.id_table =	bochs_pci_tbl,
+	.probe =	bochs_pci_probe,
+	.remove =	bochs_pci_remove,
+	.driver.pm =    &bochs_pm_ops,
+};
+
+/* ---------------------------------------------------------------------- */
+/* module init/exit                                                       */
+
+static int __init bochs_init(void)
+{
+	return drm_pci_init(&bochs_driver, &bochs_pci_driver);
+}
+
+static void __exit bochs_exit(void)
+{
+	drm_pci_exit(&bochs_driver, &bochs_pci_driver);
+}
+
+module_init(bochs_init);
+module_exit(bochs_exit);
+
+MODULE_DEVICE_TABLE(pci, bochs_pci_tbl);
+MODULE_AUTHOR("Gerd Hoffmann <kraxel@redhat.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/bochs/bochs_fbdev.c b/drivers/gpu/drm/bochs/bochs_fbdev.c
new file mode 100644
index 00000000000..561b8447412
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_fbdev.c
@@ -0,0 +1,214 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include "bochs.h"
+
+/* ---------------------------------------------------------------------- */
+
+static struct fb_ops bochsfb_ops = {
+	.owner = THIS_MODULE,
+	.fb_check_var = drm_fb_helper_check_var,
+	.fb_set_par = drm_fb_helper_set_par,
+	.fb_fillrect = sys_fillrect,
+	.fb_copyarea = sys_copyarea,
+	.fb_imageblit = sys_imageblit,
+	.fb_pan_display = drm_fb_helper_pan_display,
+	.fb_blank = drm_fb_helper_blank,
+	.fb_setcmap = drm_fb_helper_setcmap,
+};
+
+static int bochsfb_create_object(struct bochs_device *bochs,
+				 struct drm_mode_fb_cmd2 *mode_cmd,
+				 struct drm_gem_object **gobj_p)
+{
+	struct drm_device *dev = bochs->dev;
+	struct drm_gem_object *gobj;
+	u32 size;
+	int ret = 0;
+
+	size = mode_cmd->pitches[0] * mode_cmd->height;
+	ret = bochs_gem_create(dev, size, true, &gobj);
+	if (ret)
+		return ret;
+
+	*gobj_p = gobj;
+	return ret;
+}
+
+static int bochsfb_create(struct drm_fb_helper *helper,
+			  struct drm_fb_helper_surface_size *sizes)
+{
+	struct bochs_device *bochs =
+		container_of(helper, struct bochs_device, fb.helper);
+	struct drm_device *dev = bochs->dev;
+	struct fb_info *info;
+	struct drm_framebuffer *fb;
+	struct drm_mode_fb_cmd2 mode_cmd;
+	struct device *device = &dev->pdev->dev;
+	struct drm_gem_object *gobj = NULL;
+	struct bochs_bo *bo = NULL;
+	int size, ret;
+
+	if (sizes->surface_bpp != 32)
+		return -EINVAL;
+
+	mode_cmd.width = sizes->surface_width;
+	mode_cmd.height = sizes->surface_height;
+	mode_cmd.pitches[0] = mode_cmd.width * ((sizes->surface_bpp + 7) / 8);
+	mode_cmd.pixel_format = drm_mode_legacy_fb_format(sizes->surface_bpp,
+							  sizes->surface_depth);
+	size = mode_cmd.pitches[0] * mode_cmd.height;
+
+	/* alloc, pin & map bo */
+	ret = bochsfb_create_object(bochs, &mode_cmd, &gobj);
+	if (ret) {
+		DRM_ERROR("failed to create fbcon backing object %d\n", ret);
+		return ret;
+	}
+
+	bo = gem_to_bochs_bo(gobj);
+
+	ret = ttm_bo_reserve(&bo->bo, true, false, false, 0);
+	if (ret)
+		return ret;
+
+	ret = bochs_bo_pin(bo, TTM_PL_FLAG_VRAM, NULL);
+	if (ret) {
+		DRM_ERROR("failed to pin fbcon\n");
+		ttm_bo_unreserve(&bo->bo);
+		return ret;
+	}
+
+	ret = ttm_bo_kmap(&bo->bo, 0, bo->bo.num_pages,
+			  &bo->kmap);
+	if (ret) {
+		DRM_ERROR("failed to kmap fbcon\n");
+		ttm_bo_unreserve(&bo->bo);
+		return ret;
+	}
+
+	ttm_bo_unreserve(&bo->bo);
+
+	/* init fb device */
+	info = framebuffer_alloc(0, device);
+	if (info == NULL)
+		return -ENOMEM;
+
+	info->par = &bochs->fb.helper;
+
+	ret = bochs_framebuffer_init(bochs->dev, &bochs->fb.gfb, &mode_cmd, gobj);
+	if (ret)
+		return ret;
+
+	bochs->fb.size = size;
+
+	/* setup helper */
+	fb = &bochs->fb.gfb.base;
+	bochs->fb.helper.fb = fb;
+	bochs->fb.helper.fbdev = info;
+
+	strcpy(info->fix.id, "bochsdrmfb");
+
+	info->flags = FBINFO_DEFAULT;
+	info->fbops = &bochsfb_ops;
+
+	drm_fb_helper_fill_fix(info, fb->pitches[0], fb->depth);
+	drm_fb_helper_fill_var(info, &bochs->fb.helper, sizes->fb_width,
+			       sizes->fb_height);
+
+	info->screen_base = bo->kmap.virtual;
+	info->screen_size = size;
+
+#if 0
+	/* FIXME: get this right for mmap(/dev/fb0) */
+	info->fix.smem_start = bochs_bo_mmap_offset(bo);
+	info->fix.smem_len = size;
+#endif
+
+	ret = fb_alloc_cmap(&info->cmap, 256, 0);
+	if (ret) {
+		DRM_ERROR("%s: can't allocate color map\n", info->fix.id);
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static int bochs_fbdev_destroy(struct bochs_device *bochs)
+{
+	struct bochs_framebuffer *gfb = &bochs->fb.gfb;
+	struct fb_info *info;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	if (bochs->fb.helper.fbdev) {
+		info = bochs->fb.helper.fbdev;
+
+		unregister_framebuffer(info);
+		if (info->cmap.len)
+			fb_dealloc_cmap(&info->cmap);
+		framebuffer_release(info);
+	}
+
+	if (gfb->obj) {
+		drm_gem_object_unreference_unlocked(gfb->obj);
+		gfb->obj = NULL;
+	}
+
+	drm_fb_helper_fini(&bochs->fb.helper);
+	drm_framebuffer_unregister_private(&gfb->base);
+	drm_framebuffer_cleanup(&gfb->base);
+
+	return 0;
+}
+
+void bochs_fb_gamma_set(struct drm_crtc *crtc, u16 red, u16 green,
+			u16 blue, int regno)
+{
+}
+
+void bochs_fb_gamma_get(struct drm_crtc *crtc, u16 *red, u16 *green,
+			u16 *blue, int regno)
+{
+	*red   = regno;
+	*green = regno;
+	*blue  = regno;
+}
+
+static struct drm_fb_helper_funcs bochs_fb_helper_funcs = {
+	.gamma_set = bochs_fb_gamma_set,
+	.gamma_get = bochs_fb_gamma_get,
+	.fb_probe = bochsfb_create,
+};
+
+int bochs_fbdev_init(struct bochs_device *bochs)
+{
+	int ret;
+
+	bochs->fb.helper.funcs = &bochs_fb_helper_funcs;
+
+	ret = drm_fb_helper_init(bochs->dev, &bochs->fb.helper,
+				 1, 1);
+	if (ret)
+		return ret;
+
+	drm_fb_helper_single_add_all_connectors(&bochs->fb.helper);
+	drm_helper_disable_unused_functions(bochs->dev);
+	drm_fb_helper_initial_config(&bochs->fb.helper, 32);
+
+	bochs->fb.initialized = true;
+	return 0;
+}
+
+void bochs_fbdev_fini(struct bochs_device *bochs)
+{
+	if (!bochs->fb.initialized)
+		return;
+
+	bochs_fbdev_destroy(bochs);
+	bochs->fb.initialized = false;
+}
diff --git a/drivers/gpu/drm/bochs/bochs_hw.c b/drivers/gpu/drm/bochs/bochs_hw.c
new file mode 100644
index 00000000000..dbe619e6aab
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_hw.c
@@ -0,0 +1,177 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include "bochs.h"
+
+/* ---------------------------------------------------------------------- */
+
+static void bochs_vga_writeb(struct bochs_device *bochs, u16 ioport, u8 val)
+{
+	if (WARN_ON(ioport < 0x3c0 || ioport > 0x3df))
+		return;
+
+	if (bochs->mmio) {
+		int offset = ioport - 0x3c0 + 0x400;
+		writeb(val, bochs->mmio + offset);
+	} else {
+		outb(val, ioport);
+	}
+}
+
+static u16 bochs_dispi_read(struct bochs_device *bochs, u16 reg)
+{
+	u16 ret = 0;
+
+	if (bochs->mmio) {
+		int offset = 0x500 + (reg << 1);
+		ret = readw(bochs->mmio + offset);
+	} else {
+		outw(reg, VBE_DISPI_IOPORT_INDEX);
+		ret = inw(VBE_DISPI_IOPORT_DATA);
+	}
+	return ret;
+}
+
+static void bochs_dispi_write(struct bochs_device *bochs, u16 reg, u16 val)
+{
+	if (bochs->mmio) {
+		int offset = 0x500 + (reg << 1);
+		writew(val, bochs->mmio + offset);
+	} else {
+		outw(reg, VBE_DISPI_IOPORT_INDEX);
+		outw(val, VBE_DISPI_IOPORT_DATA);
+	}
+}
+
+int bochs_hw_init(struct drm_device *dev, uint32_t flags)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct pci_dev *pdev = dev->pdev;
+	unsigned long addr, size, mem, ioaddr, iosize;
+	u16 id;
+
+	if (/* (ent->driver_data == BOCHS_QEMU_STDVGA) && */
+	    (pdev->resource[2].flags & IORESOURCE_MEM)) {
+		/* mmio bar with vga and bochs registers present */
+		if (pci_request_region(pdev, 2, "bochs-drm") != 0) {
+			DRM_ERROR("Cannot request mmio region\n");
+			return -EBUSY;
+		}
+		ioaddr = pci_resource_start(pdev, 2);
+		iosize = pci_resource_len(pdev, 2);
+		bochs->mmio = ioremap(ioaddr, iosize);
+		if (bochs->mmio == NULL) {
+			DRM_ERROR("Cannot map mmio region\n");
+			return -ENOMEM;
+		}
+	} else {
+		ioaddr = VBE_DISPI_IOPORT_INDEX;
+		iosize = 2;
+		if (!request_region(ioaddr, iosize, "bochs-drm")) {
+			DRM_ERROR("Cannot request ioports\n");
+			return -EBUSY;
+		}
+		bochs->ioports = 1;
+	}
+
+	id = bochs_dispi_read(bochs, VBE_DISPI_INDEX_ID);
+	mem = bochs_dispi_read(bochs, VBE_DISPI_INDEX_VIDEO_MEMORY_64K)
+		* 64 * 1024;
+	if ((id & 0xfff0) != VBE_DISPI_ID0) {
+		DRM_ERROR("ID mismatch\n");
+		return -ENODEV;
+	}
+
+	if ((pdev->resource[0].flags & IORESOURCE_MEM) == 0)
+		return -ENODEV;
+	addr = pci_resource_start(pdev, 0);
+	size = pci_resource_len(pdev, 0);
+	if (addr == 0)
+		return -ENODEV;
+	if (size != mem) {
+		DRM_ERROR("Size mismatch: pci=%ld, bochs=%ld\n",
+			size, mem);
+		size = min(size, mem);
+	}
+
+	if (pci_request_region(pdev, 0, "bochs-drm") != 0) {
+		DRM_ERROR("Cannot request framebuffer\n");
+		return -EBUSY;
+	}
+
+	bochs->fb_map = ioremap(addr, size);
+	if (bochs->fb_map == NULL) {
+		DRM_ERROR("Cannot map framebuffer\n");
+		return -ENOMEM;
+	}
+	bochs->fb_base = addr;
+	bochs->fb_size = size;
+
+	DRM_INFO("Found bochs VGA, ID 0x%x.\n", id);
+	DRM_INFO("Framebuffer size %ld kB @ 0x%lx, %s @ 0x%lx.\n",
+		 size / 1024, addr,
+		 bochs->ioports ? "ioports" : "mmio",
+		 ioaddr);
+	return 0;
+}
+
+void bochs_hw_fini(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+
+	if (bochs->mmio)
+		iounmap(bochs->mmio);
+	if (bochs->ioports)
+		release_region(VBE_DISPI_IOPORT_INDEX, 2);
+	if (bochs->fb_map)
+		iounmap(bochs->fb_map);
+	pci_release_regions(dev->pdev);
+}
+
+void bochs_hw_setmode(struct bochs_device *bochs,
+		      struct drm_display_mode *mode)
+{
+	bochs->xres = mode->hdisplay;
+	bochs->yres = mode->vdisplay;
+	bochs->bpp = 32;
+	bochs->stride = mode->hdisplay * (bochs->bpp / 8);
+	bochs->yres_virtual = bochs->fb_size / bochs->stride;
+
+	DRM_DEBUG_DRIVER("%dx%d @ %d bpp, vy %d\n",
+			 bochs->xres, bochs->yres, bochs->bpp,
+			 bochs->yres_virtual);
+
+	bochs_vga_writeb(bochs, 0x3c0, 0x20); /* unblank */
+
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_BPP,         bochs->bpp);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_XRES,        bochs->xres);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_YRES,        bochs->yres);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_BANK,        0);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_VIRT_WIDTH,  bochs->xres);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_VIRT_HEIGHT,
+			  bochs->yres_virtual);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_X_OFFSET,    0);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_Y_OFFSET,    0);
+
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_ENABLE,
+			  VBE_DISPI_ENABLED | VBE_DISPI_LFB_ENABLED);
+}
+
+void bochs_hw_setbase(struct bochs_device *bochs,
+		      int x, int y, u64 addr)
+{
+	unsigned long offset = (unsigned long)addr +
+		y * bochs->stride +
+		x * (bochs->bpp / 8);
+	int vy = offset / bochs->stride;
+	int vx = (offset % bochs->stride) * 8 / bochs->bpp;
+
+	DRM_DEBUG_DRIVER("x %d, y %d, addr %llx -> offset %lx, vx %d, vy %d\n",
+			 x, y, addr, offset, vx, vy);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_X_OFFSET, vx);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_Y_OFFSET, vy);
+}
diff --git a/drivers/gpu/drm/bochs/bochs_kms.c b/drivers/gpu/drm/bochs/bochs_kms.c
new file mode 100644
index 00000000000..dcf2e55f4ae
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_kms.c
@@ -0,0 +1,294 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include "bochs.h"
+
+static int defx = 1024;
+static int defy = 768;
+
+module_param(defx, int, 0444);
+module_param(defy, int, 0444);
+MODULE_PARM_DESC(defx, "default x resolution");
+MODULE_PARM_DESC(defy, "default y resolution");
+
+/* ---------------------------------------------------------------------- */
+
+static void bochs_crtc_load_lut(struct drm_crtc *crtc)
+{
+}
+
+static void bochs_crtc_dpms(struct drm_crtc *crtc, int mode)
+{
+	switch (mode) {
+	case DRM_MODE_DPMS_ON:
+	case DRM_MODE_DPMS_STANDBY:
+	case DRM_MODE_DPMS_SUSPEND:
+	case DRM_MODE_DPMS_OFF:
+	default:
+		return;
+	}
+}
+
+static bool bochs_crtc_mode_fixup(struct drm_crtc *crtc,
+				  const struct drm_display_mode *mode,
+				  struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static int bochs_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
+				    struct drm_framebuffer *old_fb)
+{
+	struct bochs_device *bochs =
+		container_of(crtc, struct bochs_device, crtc);
+	struct bochs_framebuffer *bochs_fb;
+	struct bochs_bo *bo;
+	u64 gpu_addr = 0;
+	int ret;
+
+	if (old_fb) {
+		bochs_fb = to_bochs_framebuffer(old_fb);
+		bo = gem_to_bochs_bo(bochs_fb->obj);
+		ret = ttm_bo_reserve(&bo->bo, true, false, false, 0);
+		if (ret) {
+			DRM_ERROR("failed to reserve old_fb bo\n");
+		} else {
+			bochs_bo_unpin(bo);
+			ttm_bo_unreserve(&bo->bo);
+		}
+	}
+
+	if (WARN_ON(crtc->primary->fb == NULL))
+		return -EINVAL;
+
+	bochs_fb = to_bochs_framebuffer(crtc->primary->fb);
+	bo = gem_to_bochs_bo(bochs_fb->obj);
+	ret = ttm_bo_reserve(&bo->bo, true, false, false, 0);
+	if (ret)
+		return ret;
+
+	ret = bochs_bo_pin(bo, TTM_PL_FLAG_VRAM, &gpu_addr);
+	if (ret) {
+		ttm_bo_unreserve(&bo->bo);
+		return ret;
+	}
+
+	ttm_bo_unreserve(&bo->bo);
+	bochs_hw_setbase(bochs, x, y, gpu_addr);
+	return 0;
+}
+
+static int bochs_crtc_mode_set(struct drm_crtc *crtc,
+			       struct drm_display_mode *mode,
+			       struct drm_display_mode *adjusted_mode,
+			       int x, int y, struct drm_framebuffer *old_fb)
+{
+	struct bochs_device *bochs =
+		container_of(crtc, struct bochs_device, crtc);
+
+	bochs_hw_setmode(bochs, mode);
+	bochs_crtc_mode_set_base(crtc, x, y, old_fb);
+	return 0;
+}
+
+static void bochs_crtc_prepare(struct drm_crtc *crtc)
+{
+}
+
+static void bochs_crtc_commit(struct drm_crtc *crtc)
+{
+}
+
+static void bochs_crtc_gamma_set(struct drm_crtc *crtc, u16 *red, u16 *green,
+				 u16 *blue, uint32_t start, uint32_t size)
+{
+}
+
+/* These provide the minimum set of functions required to handle a CRTC */
+static const struct drm_crtc_funcs bochs_crtc_funcs = {
+	.gamma_set = bochs_crtc_gamma_set,
+	.set_config = drm_crtc_helper_set_config,
+	.destroy = drm_crtc_cleanup,
+};
+
+static const struct drm_crtc_helper_funcs bochs_helper_funcs = {
+	.dpms = bochs_crtc_dpms,
+	.mode_fixup = bochs_crtc_mode_fixup,
+	.mode_set = bochs_crtc_mode_set,
+	.mode_set_base = bochs_crtc_mode_set_base,
+	.prepare = bochs_crtc_prepare,
+	.commit = bochs_crtc_commit,
+	.load_lut = bochs_crtc_load_lut,
+};
+
+static void bochs_crtc_init(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct drm_crtc *crtc = &bochs->crtc;
+
+	drm_crtc_init(dev, crtc, &bochs_crtc_funcs);
+	drm_mode_crtc_set_gamma_size(crtc, 256);
+	drm_crtc_helper_add(crtc, &bochs_helper_funcs);
+}
+
+static bool bochs_encoder_mode_fixup(struct drm_encoder *encoder,
+				     const struct drm_display_mode *mode,
+				     struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static void bochs_encoder_mode_set(struct drm_encoder *encoder,
+				   struct drm_display_mode *mode,
+				   struct drm_display_mode *adjusted_mode)
+{
+}
+
+static void bochs_encoder_dpms(struct drm_encoder *encoder, int state)
+{
+}
+
+static void bochs_encoder_prepare(struct drm_encoder *encoder)
+{
+}
+
+static void bochs_encoder_commit(struct drm_encoder *encoder)
+{
+}
+
+static const struct drm_encoder_helper_funcs bochs_encoder_helper_funcs = {
+	.dpms = bochs_encoder_dpms,
+	.mode_fixup = bochs_encoder_mode_fixup,
+	.mode_set = bochs_encoder_mode_set,
+	.prepare = bochs_encoder_prepare,
+	.commit = bochs_encoder_commit,
+};
+
+static const struct drm_encoder_funcs bochs_encoder_encoder_funcs = {
+	.destroy = drm_encoder_cleanup,
+};
+
+static void bochs_encoder_init(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct drm_encoder *encoder = &bochs->encoder;
+
+	encoder->possible_crtcs = 0x1;
+	drm_encoder_init(dev, encoder, &bochs_encoder_encoder_funcs,
+			 DRM_MODE_ENCODER_DAC);
+	drm_encoder_helper_add(encoder, &bochs_encoder_helper_funcs);
+}
+
+
+int bochs_connector_get_modes(struct drm_connector *connector)
+{
+	int count;
+
+	count = drm_add_modes_noedid(connector, 8192, 8192);
+	drm_set_preferred_mode(connector, defx, defy);
+	return count;
+}
+
+static int bochs_connector_mode_valid(struct drm_connector *connector,
+				      struct drm_display_mode *mode)
+{
+	struct bochs_device *bochs =
+		container_of(connector, struct bochs_device, connector);
+	unsigned long size = mode->hdisplay * mode->vdisplay * 4;
+
+	/*
+	 * Make sure we can fit two framebuffers into video memory.
+	 * This allows up to 1600x1200 with 16 MB (default size).
+	 * If you want more try this:
+	 *     'qemu -vga std -global VGA.vgamem_mb=32 $otherargs'
+	 */
+	if (size * 2 > bochs->fb_size)
+		return MODE_BAD;
+
+	return MODE_OK;
+}
+
+static struct drm_encoder *
+bochs_connector_best_encoder(struct drm_connector *connector)
+{
+	int enc_id = connector->encoder_ids[0];
+	struct drm_mode_object *obj;
+	struct drm_encoder *encoder;
+
+	/* pick the encoder ids */
+	if (enc_id) {
+		obj = drm_mode_object_find(connector->dev, enc_id,
+					   DRM_MODE_OBJECT_ENCODER);
+		if (!obj)
+			return NULL;
+		encoder = obj_to_encoder(obj);
+		return encoder;
+	}
+	return NULL;
+}
+
+static enum drm_connector_status bochs_connector_detect(struct drm_connector
+							*connector, bool force)
+{
+	return connector_status_connected;
+}
+
+struct drm_connector_helper_funcs bochs_connector_connector_helper_funcs = {
+	.get_modes = bochs_connector_get_modes,
+	.mode_valid = bochs_connector_mode_valid,
+	.best_encoder = bochs_connector_best_encoder,
+};
+
+struct drm_connector_funcs bochs_connector_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.detect = bochs_connector_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = drm_connector_cleanup,
+};
+
+static void bochs_connector_init(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct drm_connector *connector = &bochs->connector;
+
+	drm_connector_init(dev, connector, &bochs_connector_connector_funcs,
+			   DRM_MODE_CONNECTOR_VIRTUAL);
+	drm_connector_helper_add(connector,
+				 &bochs_connector_connector_helper_funcs);
+}
+
+
+int bochs_kms_init(struct bochs_device *bochs)
+{
+	drm_mode_config_init(bochs->dev);
+	bochs->mode_config_initialized = true;
+
+	bochs->dev->mode_config.max_width = 8192;
+	bochs->dev->mode_config.max_height = 8192;
+
+	bochs->dev->mode_config.fb_base = bochs->fb_base;
+	bochs->dev->mode_config.preferred_depth = 24;
+	bochs->dev->mode_config.prefer_shadow = 0;
+
+	bochs->dev->mode_config.funcs = (void *)&bochs_mode_funcs;
+
+	bochs_crtc_init(bochs->dev);
+	bochs_encoder_init(bochs->dev);
+	bochs_connector_init(bochs->dev);
+	drm_mode_connector_attach_encoder(&bochs->connector,
+					  &bochs->encoder);
+
+	return 0;
+}
+
+void bochs_kms_fini(struct bochs_device *bochs)
+{
+	if (bochs->mode_config_initialized) {
+		drm_mode_config_cleanup(bochs->dev);
+		bochs->mode_config_initialized = false;
+	}
+}
diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c
new file mode 100644
index 00000000000..b9a695d9279
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_mm.c
@@ -0,0 +1,544 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include "bochs.h"
+
+static void bochs_ttm_placement(struct bochs_bo *bo, int domain);
+
+/* ---------------------------------------------------------------------- */
+
+static inline struct bochs_device *bochs_bdev(struct ttm_bo_device *bd)
+{
+	return container_of(bd, struct bochs_device, ttm.bdev);
+}
+
+static int bochs_ttm_mem_global_init(struct drm_global_reference *ref)
+{
+	return ttm_mem_global_init(ref->object);
+}
+
+static void bochs_ttm_mem_global_release(struct drm_global_reference *ref)
+{
+	ttm_mem_global_release(ref->object);
+}
+
+static int bochs_ttm_global_init(struct bochs_device *bochs)
+{
+	struct drm_global_reference *global_ref;
+	int r;
+
+	global_ref = &bochs->ttm.mem_global_ref;
+	global_ref->global_type = DRM_GLOBAL_TTM_MEM;
+	global_ref->size = sizeof(struct ttm_mem_global);
+	global_ref->init = &bochs_ttm_mem_global_init;
+	global_ref->release = &bochs_ttm_mem_global_release;
+	r = drm_global_item_ref(global_ref);
+	if (r != 0) {
+		DRM_ERROR("Failed setting up TTM memory accounting "
+			  "subsystem.\n");
+		return r;
+	}
+
+	bochs->ttm.bo_global_ref.mem_glob =
+		bochs->ttm.mem_global_ref.object;
+	global_ref = &bochs->ttm.bo_global_ref.ref;
+	global_ref->global_type = DRM_GLOBAL_TTM_BO;
+	global_ref->size = sizeof(struct ttm_bo_global);
+	global_ref->init = &ttm_bo_global_init;
+	global_ref->release = &ttm_bo_global_release;
+	r = drm_global_item_ref(global_ref);
+	if (r != 0) {
+		DRM_ERROR("Failed setting up TTM BO subsystem.\n");
+		drm_global_item_unref(&bochs->ttm.mem_global_ref);
+		return r;
+	}
+
+	return 0;
+}
+
+static void bochs_ttm_global_release(struct bochs_device *bochs)
+{
+	if (bochs->ttm.mem_global_ref.release == NULL)
+		return;
+
+	drm_global_item_unref(&bochs->ttm.bo_global_ref.ref);
+	drm_global_item_unref(&bochs->ttm.mem_global_ref);
+	bochs->ttm.mem_global_ref.release = NULL;
+}
+
+
+static void bochs_bo_ttm_destroy(struct ttm_buffer_object *tbo)
+{
+	struct bochs_bo *bo;
+
+	bo = container_of(tbo, struct bochs_bo, bo);
+	drm_gem_object_release(&bo->gem);
+	kfree(bo);
+}
+
+static bool bochs_ttm_bo_is_bochs_bo(struct ttm_buffer_object *bo)
+{
+	if (bo->destroy == &bochs_bo_ttm_destroy)
+		return true;
+	return false;
+}
+
+static int bochs_bo_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
+				  struct ttm_mem_type_manager *man)
+{
+	switch (type) {
+	case TTM_PL_SYSTEM:
+		man->flags = TTM_MEMTYPE_FLAG_MAPPABLE;
+		man->available_caching = TTM_PL_MASK_CACHING;
+		man->default_caching = TTM_PL_FLAG_CACHED;
+		break;
+	case TTM_PL_VRAM:
+		man->func = &ttm_bo_manager_func;
+		man->flags = TTM_MEMTYPE_FLAG_FIXED |
+			TTM_MEMTYPE_FLAG_MAPPABLE;
+		man->available_caching = TTM_PL_FLAG_UNCACHED |
+			TTM_PL_FLAG_WC;
+		man->default_caching = TTM_PL_FLAG_WC;
+		break;
+	default:
+		DRM_ERROR("Unsupported memory type %u\n", (unsigned)type);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static void
+bochs_bo_evict_flags(struct ttm_buffer_object *bo, struct ttm_placement *pl)
+{
+	struct bochs_bo *bochsbo = bochs_bo(bo);
+
+	if (!bochs_ttm_bo_is_bochs_bo(bo))
+		return;
+
+	bochs_ttm_placement(bochsbo, TTM_PL_FLAG_SYSTEM);
+	*pl = bochsbo->placement;
+}
+
+static int bochs_bo_verify_access(struct ttm_buffer_object *bo,
+				  struct file *filp)
+{
+	struct bochs_bo *bochsbo = bochs_bo(bo);
+
+	return drm_vma_node_verify_access(&bochsbo->gem.vma_node, filp);
+}
+
+static int bochs_ttm_io_mem_reserve(struct ttm_bo_device *bdev,
+				    struct ttm_mem_reg *mem)
+{
+	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
+	struct bochs_device *bochs = bochs_bdev(bdev);
+
+	mem->bus.addr = NULL;
+	mem->bus.offset = 0;
+	mem->bus.size = mem->num_pages << PAGE_SHIFT;
+	mem->bus.base = 0;
+	mem->bus.is_iomem = false;
+	if (!(man->flags & TTM_MEMTYPE_FLAG_MAPPABLE))
+		return -EINVAL;
+	switch (mem->mem_type) {
+	case TTM_PL_SYSTEM:
+		/* system memory */
+		return 0;
+	case TTM_PL_VRAM:
+		mem->bus.offset = mem->start << PAGE_SHIFT;
+		mem->bus.base = bochs->fb_base;
+		mem->bus.is_iomem = true;
+		break;
+	default:
+		return -EINVAL;
+		break;
+	}
+	return 0;
+}
+
+static void bochs_ttm_io_mem_free(struct ttm_bo_device *bdev,
+				  struct ttm_mem_reg *mem)
+{
+}
+
+static int bochs_bo_move(struct ttm_buffer_object *bo,
+			 bool evict, bool interruptible,
+			 bool no_wait_gpu,
+			 struct ttm_mem_reg *new_mem)
+{
+	return ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
+}
+
+
+static void bochs_ttm_backend_destroy(struct ttm_tt *tt)
+{
+	ttm_tt_fini(tt);
+	kfree(tt);
+}
+
+static struct ttm_backend_func bochs_tt_backend_func = {
+	.destroy = &bochs_ttm_backend_destroy,
+};
+
+static struct ttm_tt *bochs_ttm_tt_create(struct ttm_bo_device *bdev,
+					  unsigned long size,
+					  uint32_t page_flags,
+					  struct page *dummy_read_page)
+{
+	struct ttm_tt *tt;
+
+	tt = kzalloc(sizeof(struct ttm_tt), GFP_KERNEL);
+	if (tt == NULL)
+		return NULL;
+	tt->func = &bochs_tt_backend_func;
+	if (ttm_tt_init(tt, bdev, size, page_flags, dummy_read_page)) {
+		kfree(tt);
+		return NULL;
+	}
+	return tt;
+}
+
+struct ttm_bo_driver bochs_bo_driver = {
+	.ttm_tt_create = bochs_ttm_tt_create,
+	.ttm_tt_populate = ttm_pool_populate,
+	.ttm_tt_unpopulate = ttm_pool_unpopulate,
+	.init_mem_type = bochs_bo_init_mem_type,
+	.evict_flags = bochs_bo_evict_flags,
+	.move = bochs_bo_move,
+	.verify_access = bochs_bo_verify_access,
+	.io_mem_reserve = &bochs_ttm_io_mem_reserve,
+	.io_mem_free = &bochs_ttm_io_mem_free,
+};
+
+int bochs_mm_init(struct bochs_device *bochs)
+{
+	struct ttm_bo_device *bdev = &bochs->ttm.bdev;
+	int ret;
+
+	ret = bochs_ttm_global_init(bochs);
+	if (ret)
+		return ret;
+
+	ret = ttm_bo_device_init(&bochs->ttm.bdev,
+				 bochs->ttm.bo_global_ref.ref.object,
+				 &bochs_bo_driver,
+				 bochs->dev->anon_inode->i_mapping,
+				 DRM_FILE_PAGE_OFFSET,
+				 true);
+	if (ret) {
+		DRM_ERROR("Error initialising bo driver; %d\n", ret);
+		return ret;
+	}
+
+	ret = ttm_bo_init_mm(bdev, TTM_PL_VRAM,
+			     bochs->fb_size >> PAGE_SHIFT);
+	if (ret) {
+		DRM_ERROR("Failed ttm VRAM init: %d\n", ret);
+		return ret;
+	}
+
+	bochs->ttm.initialized = true;
+	return 0;
+}
+
+void bochs_mm_fini(struct bochs_device *bochs)
+{
+	if (!bochs->ttm.initialized)
+		return;
+
+	ttm_bo_device_release(&bochs->ttm.bdev);
+	bochs_ttm_global_release(bochs);
+	bochs->ttm.initialized = false;
+}
+
+static void bochs_ttm_placement(struct bochs_bo *bo, int domain)
+{
+	u32 c = 0;
+	bo->placement.fpfn = 0;
+	bo->placement.lpfn = 0;
+	bo->placement.placement = bo->placements;
+	bo->placement.busy_placement = bo->placements;
+	if (domain & TTM_PL_FLAG_VRAM) {
+		bo->placements[c++] = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED
+			| TTM_PL_FLAG_VRAM;
+	}
+	if (domain & TTM_PL_FLAG_SYSTEM) {
+		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+	}
+	if (!c) {
+		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+	}
+	bo->placement.num_placement = c;
+	bo->placement.num_busy_placement = c;
+}
+
+static inline u64 bochs_bo_gpu_offset(struct bochs_bo *bo)
+{
+	return bo->bo.offset;
+}
+
+int bochs_bo_pin(struct bochs_bo *bo, u32 pl_flag, u64 *gpu_addr)
+{
+	int i, ret;
+
+	if (bo->pin_count) {
+		bo->pin_count++;
+		if (gpu_addr)
+			*gpu_addr = bochs_bo_gpu_offset(bo);
+		return 0;
+	}
+
+	bochs_ttm_placement(bo, pl_flag);
+	for (i = 0; i < bo->placement.num_placement; i++)
+		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
+	if (ret)
+		return ret;
+
+	bo->pin_count = 1;
+	if (gpu_addr)
+		*gpu_addr = bochs_bo_gpu_offset(bo);
+	return 0;
+}
+
+int bochs_bo_unpin(struct bochs_bo *bo)
+{
+	int i, ret;
+
+	if (!bo->pin_count) {
+		DRM_ERROR("unpin bad %p\n", bo);
+		return 0;
+	}
+	bo->pin_count--;
+
+	if (bo->pin_count)
+		return 0;
+
+	for (i = 0; i < bo->placement.num_placement; i++)
+		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
+	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+int bochs_mmap(struct file *filp, struct vm_area_struct *vma)
+{
+	struct drm_file *file_priv;
+	struct bochs_device *bochs;
+
+	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
+		return drm_mmap(filp, vma);
+
+	file_priv = filp->private_data;
+	bochs = file_priv->minor->dev->dev_private;
+	return ttm_bo_mmap(filp, vma, &bochs->ttm.bdev);
+}
+
+/* ---------------------------------------------------------------------- */
+
+static int bochs_bo_create(struct drm_device *dev, int size, int align,
+			   uint32_t flags, struct bochs_bo **pbochsbo)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct bochs_bo *bochsbo;
+	size_t acc_size;
+	int ret;
+
+	bochsbo = kzalloc(sizeof(struct bochs_bo), GFP_KERNEL);
+	if (!bochsbo)
+		return -ENOMEM;
+
+	ret = drm_gem_object_init(dev, &bochsbo->gem, size);
+	if (ret) {
+		kfree(bochsbo);
+		return ret;
+	}
+
+	bochsbo->bo.bdev = &bochs->ttm.bdev;
+	bochsbo->bo.bdev->dev_mapping = dev->anon_inode->i_mapping;
+
+	bochs_ttm_placement(bochsbo, TTM_PL_FLAG_VRAM | TTM_PL_FLAG_SYSTEM);
+
+	acc_size = ttm_bo_dma_acc_size(&bochs->ttm.bdev, size,
+				       sizeof(struct bochs_bo));
+
+	ret = ttm_bo_init(&bochs->ttm.bdev, &bochsbo->bo, size,
+			  ttm_bo_type_device, &bochsbo->placement,
+			  align >> PAGE_SHIFT, false, NULL, acc_size,
+			  NULL, bochs_bo_ttm_destroy);
+	if (ret)
+		return ret;
+
+	*pbochsbo = bochsbo;
+	return 0;
+}
+
+int bochs_gem_create(struct drm_device *dev, u32 size, bool iskernel,
+		     struct drm_gem_object **obj)
+{
+	struct bochs_bo *bochsbo;
+	int ret;
+
+	*obj = NULL;
+
+	size = ALIGN(size, PAGE_SIZE);
+	if (size == 0)
+		return -EINVAL;
+
+	ret = bochs_bo_create(dev, size, 0, 0, &bochsbo);
+	if (ret) {
+		if (ret != -ERESTARTSYS)
+			DRM_ERROR("failed to allocate GEM object\n");
+		return ret;
+	}
+	*obj = &bochsbo->gem;
+	return 0;
+}
+
+int bochs_dumb_create(struct drm_file *file, struct drm_device *dev,
+		      struct drm_mode_create_dumb *args)
+{
+	struct drm_gem_object *gobj;
+	u32 handle;
+	int ret;
+
+	args->pitch = args->width * ((args->bpp + 7) / 8);
+	args->size = args->pitch * args->height;
+
+	ret = bochs_gem_create(dev, args->size, false,
+			       &gobj);
+	if (ret)
+		return ret;
+
+	ret = drm_gem_handle_create(file, gobj, &handle);
+	drm_gem_object_unreference_unlocked(gobj);
+	if (ret)
+		return ret;
+
+	args->handle = handle;
+	return 0;
+}
+
+static void bochs_bo_unref(struct bochs_bo **bo)
+{
+	struct ttm_buffer_object *tbo;
+
+	if ((*bo) == NULL)
+		return;
+
+	tbo = &((*bo)->bo);
+	ttm_bo_unref(&tbo);
+	*bo = NULL;
+}
+
+void bochs_gem_free_object(struct drm_gem_object *obj)
+{
+	struct bochs_bo *bochs_bo = gem_to_bochs_bo(obj);
+
+	bochs_bo_unref(&bochs_bo);
+}
+
+int bochs_dumb_mmap_offset(struct drm_file *file, struct drm_device *dev,
+			   uint32_t handle, uint64_t *offset)
+{
+	struct drm_gem_object *obj;
+	int ret;
+	struct bochs_bo *bo;
+
+	mutex_lock(&dev->struct_mutex);
+	obj = drm_gem_object_lookup(dev, file, handle);
+	if (obj == NULL) {
+		ret = -ENOENT;
+		goto out_unlock;
+	}
+
+	bo = gem_to_bochs_bo(obj);
+	*offset = bochs_bo_mmap_offset(bo);
+
+	drm_gem_object_unreference(obj);
+	ret = 0;
+out_unlock:
+	mutex_unlock(&dev->struct_mutex);
+	return ret;
+
+}
+
+/* ---------------------------------------------------------------------- */
+
+static void bochs_user_framebuffer_destroy(struct drm_framebuffer *fb)
+{
+	struct bochs_framebuffer *bochs_fb = to_bochs_framebuffer(fb);
+	if (bochs_fb->obj)
+		drm_gem_object_unreference_unlocked(bochs_fb->obj);
+	drm_framebuffer_cleanup(fb);
+	kfree(fb);
+}
+
+static const struct drm_framebuffer_funcs bochs_fb_funcs = {
+	.destroy = bochs_user_framebuffer_destroy,
+};
+
+int bochs_framebuffer_init(struct drm_device *dev,
+			   struct bochs_framebuffer *gfb,
+			   struct drm_mode_fb_cmd2 *mode_cmd,
+			   struct drm_gem_object *obj)
+{
+	int ret;
+
+	drm_helper_mode_fill_fb_struct(&gfb->base, mode_cmd);
+	gfb->obj = obj;
+	ret = drm_framebuffer_init(dev, &gfb->base, &bochs_fb_funcs);
+	if (ret) {
+		DRM_ERROR("drm_framebuffer_init failed: %d\n", ret);
+		return ret;
+	}
+	return 0;
+}
+
+static struct drm_framebuffer *
+bochs_user_framebuffer_create(struct drm_device *dev,
+			      struct drm_file *filp,
+			      struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	struct drm_gem_object *obj;
+	struct bochs_framebuffer *bochs_fb;
+	int ret;
+
+	DRM_DEBUG_DRIVER("%dx%d, format %c%c%c%c\n",
+	       mode_cmd->width, mode_cmd->height,
+	       (mode_cmd->pixel_format)       & 0xff,
+	       (mode_cmd->pixel_format >> 8)  & 0xff,
+	       (mode_cmd->pixel_format >> 16) & 0xff,
+	       (mode_cmd->pixel_format >> 24) & 0xff);
+
+	if (mode_cmd->pixel_format != DRM_FORMAT_XRGB8888)
+		return ERR_PTR(-ENOENT);
+
+	obj = drm_gem_object_lookup(dev, filp, mode_cmd->handles[0]);
+	if (obj == NULL)
+		return ERR_PTR(-ENOENT);
+
+	bochs_fb = kzalloc(sizeof(*bochs_fb), GFP_KERNEL);
+	if (!bochs_fb) {
+		drm_gem_object_unreference_unlocked(obj);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	ret = bochs_framebuffer_init(dev, bochs_fb, mode_cmd, obj);
+	if (ret) {
+		drm_gem_object_unreference_unlocked(obj);
+		kfree(bochs_fb);
+		return ERR_PTR(ret);
+	}
+	return &bochs_fb->base;
+}
+
+const struct drm_mode_config_funcs bochs_mode_funcs = {
+	.fb_create = bochs_user_framebuffer_create,
+};
diff --git a/drivers/gpu/drm/bridge/Kconfig b/drivers/gpu/drm/bridge/Kconfig
new file mode 100644
index 00000000000..884923f982d
--- /dev/null
+++ b/drivers/gpu/drm/bridge/Kconfig
@@ -0,0 +1,5 @@
+config DRM_PTN3460
+	tristate "PTN3460 DP/LVDS bridge"
+	depends on DRM
+	select DRM_KMS_HELPER
+	---help---
diff --git a/drivers/gpu/drm/bridge/Makefile b/drivers/gpu/drm/bridge/Makefile
new file mode 100644
index 00000000000..b4733e1fbd2
--- /dev/null
+++ b/drivers/gpu/drm/bridge/Makefile
@@ -0,0 +1,3 @@
+ccflags-y := -Iinclude/drm
+
+obj-$(CONFIG_DRM_PTN3460) += ptn3460.o
diff --git a/drivers/gpu/drm/bridge/ptn3460.c b/drivers/gpu/drm/bridge/ptn3460.c
new file mode 100644
index 00000000000..98fd17ae491
--- /dev/null
+++ b/drivers/gpu/drm/bridge/ptn3460.c
@@ -0,0 +1,343 @@
+/*
+ * NXP PTN3460 DP/LVDS bridge driver
+ *
+ * Copyright (C) 2013 Google, Inc.
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/i2c.h>
+#include <linux/gpio.h>
+#include <linux/delay.h>
+
+#include "drmP.h"
+#include "drm_edid.h"
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+
+#include "bridge/ptn3460.h"
+
+#define PTN3460_EDID_ADDR			0x0
+#define PTN3460_EDID_EMULATION_ADDR		0x84
+#define PTN3460_EDID_ENABLE_EMULATION		0
+#define PTN3460_EDID_EMULATION_SELECTION	1
+#define PTN3460_EDID_SRAM_LOAD_ADDR		0x85
+
+struct ptn3460_bridge {
+	struct drm_connector connector;
+	struct i2c_client *client;
+	struct drm_encoder *encoder;
+	struct drm_bridge *bridge;
+	struct edid *edid;
+	int gpio_pd_n;
+	int gpio_rst_n;
+	u32 edid_emulation;
+	bool enabled;
+};
+
+static int ptn3460_read_bytes(struct ptn3460_bridge *ptn_bridge, char addr,
+		u8 *buf, int len)
+{
+	int ret;
+
+	ret = i2c_master_send(ptn_bridge->client, &addr, 1);
+	if (ret <= 0) {
+		DRM_ERROR("Failed to send i2c command, ret=%d\n", ret);
+		return ret;
+	}
+
+	ret = i2c_master_recv(ptn_bridge->client, buf, len);
+	if (ret <= 0) {
+		DRM_ERROR("Failed to recv i2c data, ret=%d\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int ptn3460_write_byte(struct ptn3460_bridge *ptn_bridge, char addr,
+		char val)
+{
+	int ret;
+	char buf[2];
+
+	buf[0] = addr;
+	buf[1] = val;
+
+	ret = i2c_master_send(ptn_bridge->client, buf, ARRAY_SIZE(buf));
+	if (ret <= 0) {
+		DRM_ERROR("Failed to send i2c command, ret=%d\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int ptn3460_select_edid(struct ptn3460_bridge *ptn_bridge)
+{
+	int ret;
+	char val;
+
+	/* Load the selected edid into SRAM (accessed at PTN3460_EDID_ADDR) */
+	ret = ptn3460_write_byte(ptn_bridge, PTN3460_EDID_SRAM_LOAD_ADDR,
+			ptn_bridge->edid_emulation);
+	if (ret) {
+		DRM_ERROR("Failed to transfer edid to sram, ret=%d\n", ret);
+		return ret;
+	}
+
+	/* Enable EDID emulation and select the desired EDID */
+	val = 1 << PTN3460_EDID_ENABLE_EMULATION |
+		ptn_bridge->edid_emulation << PTN3460_EDID_EMULATION_SELECTION;
+
+	ret = ptn3460_write_byte(ptn_bridge, PTN3460_EDID_EMULATION_ADDR, val);
+	if (ret) {
+		DRM_ERROR("Failed to write edid value, ret=%d\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static void ptn3460_pre_enable(struct drm_bridge *bridge)
+{
+	struct ptn3460_bridge *ptn_bridge = bridge->driver_private;
+	int ret;
+
+	if (ptn_bridge->enabled)
+		return;
+
+	if (gpio_is_valid(ptn_bridge->gpio_pd_n))
+		gpio_set_value(ptn_bridge->gpio_pd_n, 1);
+
+	if (gpio_is_valid(ptn_bridge->gpio_rst_n)) {
+		gpio_set_value(ptn_bridge->gpio_rst_n, 0);
+		udelay(10);
+		gpio_set_value(ptn_bridge->gpio_rst_n, 1);
+	}
+
+	/*
+	 * There's a bug in the PTN chip where it falsely asserts hotplug before
+	 * it is fully functional. We're forced to wait for the maximum start up
+	 * time specified in the chip's datasheet to make sure we're really up.
+	 */
+	msleep(90);
+
+	ret = ptn3460_select_edid(ptn_bridge);
+	if (ret)
+		DRM_ERROR("Select edid failed ret=%d\n", ret);
+
+	ptn_bridge->enabled = true;
+}
+
+static void ptn3460_enable(struct drm_bridge *bridge)
+{
+}
+
+static void ptn3460_disable(struct drm_bridge *bridge)
+{
+	struct ptn3460_bridge *ptn_bridge = bridge->driver_private;
+
+	if (!ptn_bridge->enabled)
+		return;
+
+	ptn_bridge->enabled = false;
+
+	if (gpio_is_valid(ptn_bridge->gpio_rst_n))
+		gpio_set_value(ptn_bridge->gpio_rst_n, 1);
+
+	if (gpio_is_valid(ptn_bridge->gpio_pd_n))
+		gpio_set_value(ptn_bridge->gpio_pd_n, 0);
+}
+
+static void ptn3460_post_disable(struct drm_bridge *bridge)
+{
+}
+
+void ptn3460_bridge_destroy(struct drm_bridge *bridge)
+{
+	struct ptn3460_bridge *ptn_bridge = bridge->driver_private;
+
+	drm_bridge_cleanup(bridge);
+	if (gpio_is_valid(ptn_bridge->gpio_pd_n))
+		gpio_free(ptn_bridge->gpio_pd_n);
+	if (gpio_is_valid(ptn_bridge->gpio_rst_n))
+		gpio_free(ptn_bridge->gpio_rst_n);
+	/* Nothing else to free, we've got devm allocated memory */
+}
+
+struct drm_bridge_funcs ptn3460_bridge_funcs = {
+	.pre_enable = ptn3460_pre_enable,
+	.enable = ptn3460_enable,
+	.disable = ptn3460_disable,
+	.post_disable = ptn3460_post_disable,
+	.destroy = ptn3460_bridge_destroy,
+};
+
+int ptn3460_get_modes(struct drm_connector *connector)
+{
+	struct ptn3460_bridge *ptn_bridge;
+	u8 *edid;
+	int ret, num_modes;
+	bool power_off;
+
+	ptn_bridge = container_of(connector, struct ptn3460_bridge, connector);
+
+	if (ptn_bridge->edid)
+		return drm_add_edid_modes(connector, ptn_bridge->edid);
+
+	power_off = !ptn_bridge->enabled;
+	ptn3460_pre_enable(ptn_bridge->bridge);
+
+	edid = kmalloc(EDID_LENGTH, GFP_KERNEL);
+	if (!edid) {
+		DRM_ERROR("Failed to allocate edid\n");
+		return 0;
+	}
+
+	ret = ptn3460_read_bytes(ptn_bridge, PTN3460_EDID_ADDR, edid,
+			EDID_LENGTH);
+	if (ret) {
+		kfree(edid);
+		num_modes = 0;
+		goto out;
+	}
+
+	ptn_bridge->edid = (struct edid *)edid;
+	drm_mode_connector_update_edid_property(connector, ptn_bridge->edid);
+
+	num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
+
+out:
+	if (power_off)
+		ptn3460_disable(ptn_bridge->bridge);
+
+	return num_modes;
+}
+
+struct drm_encoder *ptn3460_best_encoder(struct drm_connector *connector)
+{
+	struct ptn3460_bridge *ptn_bridge;
+
+	ptn_bridge = container_of(connector, struct ptn3460_bridge, connector);
+
+	return ptn_bridge->encoder;
+}
+
+struct drm_connector_helper_funcs ptn3460_connector_helper_funcs = {
+	.get_modes = ptn3460_get_modes,
+	.best_encoder = ptn3460_best_encoder,
+};
+
+enum drm_connector_status ptn3460_detect(struct drm_connector *connector,
+		bool force)
+{
+	return connector_status_connected;
+}
+
+void ptn3460_connector_destroy(struct drm_connector *connector)
+{
+	drm_connector_cleanup(connector);
+}
+
+struct drm_connector_funcs ptn3460_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ptn3460_detect,
+	.destroy = ptn3460_connector_destroy,
+};
+
+int ptn3460_init(struct drm_device *dev, struct drm_encoder *encoder,
+		struct i2c_client *client, struct device_node *node)
+{
+	int ret;
+	struct drm_bridge *bridge;
+	struct ptn3460_bridge *ptn_bridge;
+
+	bridge = devm_kzalloc(dev->dev, sizeof(*bridge), GFP_KERNEL);
+	if (!bridge) {
+		DRM_ERROR("Failed to allocate drm bridge\n");
+		return -ENOMEM;
+	}
+
+	ptn_bridge = devm_kzalloc(dev->dev, sizeof(*ptn_bridge), GFP_KERNEL);
+	if (!ptn_bridge) {
+		DRM_ERROR("Failed to allocate ptn bridge\n");
+		return -ENOMEM;
+	}
+
+	ptn_bridge->client = client;
+	ptn_bridge->encoder = encoder;
+	ptn_bridge->bridge = bridge;
+	ptn_bridge->gpio_pd_n = of_get_named_gpio(node, "powerdown-gpio", 0);
+	if (gpio_is_valid(ptn_bridge->gpio_pd_n)) {
+		ret = gpio_request_one(ptn_bridge->gpio_pd_n,
+				GPIOF_OUT_INIT_HIGH, "PTN3460_PD_N");
+		if (ret) {
+			DRM_ERROR("Request powerdown-gpio failed (%d)\n", ret);
+			return ret;
+		}
+	}
+
+	ptn_bridge->gpio_rst_n = of_get_named_gpio(node, "reset-gpio", 0);
+	if (gpio_is_valid(ptn_bridge->gpio_rst_n)) {
+		/*
+		 * Request the reset pin low to avoid the bridge being
+		 * initialized prematurely
+		 */
+		ret = gpio_request_one(ptn_bridge->gpio_rst_n,
+				GPIOF_OUT_INIT_LOW, "PTN3460_RST_N");
+		if (ret) {
+			DRM_ERROR("Request reset-gpio failed (%d)\n", ret);
+			gpio_free(ptn_bridge->gpio_pd_n);
+			return ret;
+		}
+	}
+
+	ret = of_property_read_u32(node, "edid-emulation",
+			&ptn_bridge->edid_emulation);
+	if (ret) {
+		DRM_ERROR("Can't read edid emulation value\n");
+		goto err;
+	}
+
+	ret = drm_bridge_init(dev, bridge, &ptn3460_bridge_funcs);
+	if (ret) {
+		DRM_ERROR("Failed to initialize bridge with drm\n");
+		goto err;
+	}
+
+	bridge->driver_private = ptn_bridge;
+	encoder->bridge = bridge;
+
+	ret = drm_connector_init(dev, &ptn_bridge->connector,
+			&ptn3460_connector_funcs, DRM_MODE_CONNECTOR_LVDS);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		goto err;
+	}
+	drm_connector_helper_add(&ptn_bridge->connector,
+			&ptn3460_connector_helper_funcs);
+	drm_sysfs_connector_add(&ptn_bridge->connector);
+	drm_mode_connector_attach_encoder(&ptn_bridge->connector, encoder);
+
+	return 0;
+
+err:
+	if (gpio_is_valid(ptn_bridge->gpio_pd_n))
+		gpio_free(ptn_bridge->gpio_pd_n);
+	if (gpio_is_valid(ptn_bridge->gpio_rst_n))
+		gpio_free(ptn_bridge->gpio_rst_n);
+	return ret;
+}
+EXPORT_SYMBOL(ptn3460_init);
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.c b/drivers/gpu/drm/cirrus/cirrus_drv.c
index 953fc8aea69..08ce520f61a 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.c
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.c
@@ -11,6 +11,7 @@
 #include <linux/module.h>
 #include <linux/console.h>
 #include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
 
 #include "cirrus_drv.h"
 
@@ -75,6 +76,41 @@ static void cirrus_pci_remove(struct pci_dev *pdev)
 	drm_put_dev(dev);
 }
 
+static int cirrus_pm_suspend(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	struct cirrus_device *cdev = drm_dev->dev_private;
+
+	drm_kms_helper_poll_disable(drm_dev);
+
+	if (cdev->mode_info.gfbdev) {
+		console_lock();
+		fb_set_suspend(cdev->mode_info.gfbdev->helper.fbdev, 1);
+		console_unlock();
+	}
+
+	return 0;
+}
+
+static int cirrus_pm_resume(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+	struct cirrus_device *cdev = drm_dev->dev_private;
+
+	drm_helper_resume_force_mode(drm_dev);
+
+	if (cdev->mode_info.gfbdev) {
+		console_lock();
+		fb_set_suspend(cdev->mode_info.gfbdev->helper.fbdev, 0);
+		console_unlock();
+	}
+
+	drm_kms_helper_poll_enable(drm_dev);
+	return 0;
+}
+
 static const struct file_operations cirrus_driver_fops = {
 	.owner = THIS_MODULE,
 	.open = drm_open,
@@ -103,11 +139,17 @@ static struct drm_driver driver = {
 	.dumb_destroy = drm_gem_dumb_destroy,
 };
 
+static const struct dev_pm_ops cirrus_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(cirrus_pm_suspend,
+				cirrus_pm_resume)
+};
+
 static struct pci_driver cirrus_pci_driver = {
 	.name = DRIVER_NAME,
 	.id_table = pciidlist,
 	.probe = cirrus_pci_probe,
 	.remove = cirrus_pci_remove,
+	.driver.pm = &cirrus_pm_ops,
 };
 
 static int __init cirrus_init(void)
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.h b/drivers/gpu/drm/cirrus/cirrus_drv.h
index b6aded73838..117d3eca5e3 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.h
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.h
@@ -222,7 +222,7 @@ void cirrus_fbdev_fini(struct cirrus_device *cdev);
 void cirrus_driver_irq_preinstall(struct drm_device *dev);
 int cirrus_driver_irq_postinstall(struct drm_device *dev);
 void cirrus_driver_irq_uninstall(struct drm_device *dev);
-irqreturn_t cirrus_driver_irq_handler(DRM_IRQ_ARGS);
+irqreturn_t cirrus_driver_irq_handler(int irq, void *arg);
 
 				/* cirrus_kms.c */
 int cirrus_driver_load(struct drm_device *dev, unsigned long flags);
diff --git a/drivers/gpu/drm/cirrus/cirrus_fbdev.c b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
index b27e95666fa..32bbba0a787 100644
--- a/drivers/gpu/drm/cirrus/cirrus_fbdev.c
+++ b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
@@ -39,7 +39,7 @@ static void cirrus_dirty_update(struct cirrus_fbdev *afbdev,
 	 * then the BO is being moved and we should
 	 * store up the damage until later.
 	 */
-	if (!in_interrupt())
+	if (drm_can_sleep())
 		ret = cirrus_bo_reserve(bo, true);
 	if (ret) {
 		if (ret != -EBUSY)
@@ -233,6 +233,9 @@ static int cirrusfb_create(struct drm_fb_helper *helper,
 	info->apertures->ranges[0].base = cdev->dev->mode_config.fb_base;
 	info->apertures->ranges[0].size = cdev->mc.vram_size;
 
+	info->fix.smem_start = cdev->dev->mode_config.fb_base;
+	info->fix.smem_len = cdev->mc.vram_size;
+
 	info->screen_base = sysram;
 	info->screen_size = size;
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_main.c b/drivers/gpu/drm/cirrus/cirrus_main.c
index 78e76f24343..99c1983f99d 100644
--- a/drivers/gpu/drm/cirrus/cirrus_main.c
+++ b/drivers/gpu/drm/cirrus/cirrus_main.c
@@ -255,7 +255,7 @@ int cirrus_dumb_create(struct drm_file *file,
 	return 0;
 }
 
-void cirrus_bo_unref(struct cirrus_bo **bo)
+static void cirrus_bo_unref(struct cirrus_bo **bo)
 {
 	struct ttm_buffer_object *tbo;
 
@@ -264,17 +264,13 @@ void cirrus_bo_unref(struct cirrus_bo **bo)
 
 	tbo = &((*bo)->bo);
 	ttm_bo_unref(&tbo);
-	if (tbo == NULL)
-		*bo = NULL;
-
+	*bo = NULL;
 }
 
 void cirrus_gem_free_object(struct drm_gem_object *obj)
 {
 	struct cirrus_bo *cirrus_bo = gem_to_cirrus_bo(obj);
 
-	if (!cirrus_bo)
-		return;
 	cirrus_bo_unref(&cirrus_bo);
 }
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_mode.c b/drivers/gpu/drm/cirrus/cirrus_mode.c
index adabc3daaa5..49332c5fe35 100644
--- a/drivers/gpu/drm/cirrus/cirrus_mode.c
+++ b/drivers/gpu/drm/cirrus/cirrus_mode.c
@@ -102,7 +102,7 @@ static bool cirrus_crtc_mode_fixup(struct drm_crtc *crtc,
 	return true;
 }
 
-void cirrus_set_start_address(struct drm_crtc *crtc, unsigned offset)
+static void cirrus_set_start_address(struct drm_crtc *crtc, unsigned offset)
 {
 	struct cirrus_device *cdev = crtc->dev->dev_private;
 	u32 addr;
@@ -149,7 +149,7 @@ static int cirrus_crtc_do_set_base(struct drm_crtc *crtc,
 		cirrus_bo_unreserve(bo);
 	}
 
-	cirrus_fb = to_cirrus_framebuffer(crtc->fb);
+	cirrus_fb = to_cirrus_framebuffer(crtc->primary->fb);
 	obj = cirrus_fb->obj;
 	bo = gem_to_cirrus_bo(obj);
 
@@ -268,13 +268,13 @@ static int cirrus_crtc_mode_set(struct drm_crtc *crtc,
 	sr07 = RREG8(SEQ_DATA);
 	sr07 &= 0xe0;
 	hdr = 0;
-	switch (crtc->fb->bits_per_pixel) {
+	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
 		sr07 |= 0x11;
 		break;
 	case 16:
-		sr07 |= 0xc1;
-		hdr = 0xc0;
+		sr07 |= 0x17;
+		hdr = 0xc1;
 		break;
 	case 24:
 		sr07 |= 0x15;
@@ -291,13 +291,13 @@ static int cirrus_crtc_mode_set(struct drm_crtc *crtc,
 	WREG_SEQ(0x7, sr07);
 
 	/* Program the pitch */
-	tmp = crtc->fb->pitches[0] / 8;
+	tmp = crtc->primary->fb->pitches[0] / 8;
 	WREG_CRT(VGA_CRTC_OFFSET, tmp);
 
 	/* Enable extended blanking and pitch bits, and enable full memory */
 	tmp = 0x22;
-	tmp |= (crtc->fb->pitches[0] >> 7) & 0x10;
-	tmp |= (crtc->fb->pitches[0] >> 6) & 0x40;
+	tmp |= (crtc->primary->fb->pitches[0] >> 7) & 0x10;
+	tmp |= (crtc->primary->fb->pitches[0] >> 6) & 0x40;
 	WREG_CRT(0x1b, tmp);
 
 	/* Enable high-colour modes */
@@ -308,6 +308,9 @@ static int cirrus_crtc_mode_set(struct drm_crtc *crtc,
 
 	WREG_HDR(hdr);
 	cirrus_crtc_do_set_base(crtc, old_fb, x, y, 0);
+
+	/* Unblank (needed on S3 resume, vgabios doesn't do it then) */
+	outb(0x20, 0x3c0);
 	return 0;
 }
 
@@ -453,7 +456,7 @@ static void cirrus_encoder_commit(struct drm_encoder *encoder)
 {
 }
 
-void cirrus_encoder_destroy(struct drm_encoder *encoder)
+static void cirrus_encoder_destroy(struct drm_encoder *encoder)
 {
 	struct cirrus_encoder *cirrus_encoder = to_cirrus_encoder(encoder);
 	drm_encoder_cleanup(encoder);
@@ -492,7 +495,7 @@ static struct drm_encoder *cirrus_encoder_init(struct drm_device *dev)
 }
 
 
-int cirrus_vga_get_modes(struct drm_connector *connector)
+static int cirrus_vga_get_modes(struct drm_connector *connector)
 {
 	int count;
 
@@ -502,14 +505,7 @@ int cirrus_vga_get_modes(struct drm_connector *connector)
 	return count;
 }
 
-static int cirrus_vga_mode_valid(struct drm_connector *connector,
-				 struct drm_display_mode *mode)
-{
-	/* Any mode we've added is valid */
-	return MODE_OK;
-}
-
-struct drm_encoder *cirrus_connector_best_encoder(struct drm_connector
+static struct drm_encoder *cirrus_connector_best_encoder(struct drm_connector
 						  *connector)
 {
 	int enc_id = connector->encoder_ids[0];
@@ -543,7 +539,6 @@ static void cirrus_connector_destroy(struct drm_connector *connector)
 
 struct drm_connector_helper_funcs cirrus_vga_connector_helper_funcs = {
 	.get_modes = cirrus_vga_get_modes,
-	.mode_valid = cirrus_vga_mode_valid,
 	.best_encoder = cirrus_connector_best_encoder,
 };
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c
index 75becdeac07..92e6b778609 100644
--- a/drivers/gpu/drm/cirrus/cirrus_ttm.c
+++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c
@@ -80,7 +80,7 @@ static int cirrus_ttm_global_init(struct cirrus_device *cirrus)
 	return 0;
 }
 
-void
+static void
 cirrus_ttm_global_release(struct cirrus_device *cirrus)
 {
 	if (cirrus->ttm.mem_global_ref.release == NULL)
@@ -102,7 +102,7 @@ static void cirrus_bo_ttm_destroy(struct ttm_buffer_object *tbo)
 	kfree(bo);
 }
 
-bool cirrus_ttm_bo_is_cirrus_bo(struct ttm_buffer_object *bo)
+static bool cirrus_ttm_bo_is_cirrus_bo(struct ttm_buffer_object *bo)
 {
 	if (bo->destroy == &cirrus_bo_ttm_destroy)
 		return true;
@@ -208,7 +208,7 @@ static struct ttm_backend_func cirrus_tt_backend_func = {
 };
 
 
-struct ttm_tt *cirrus_ttm_tt_create(struct ttm_bo_device *bdev,
+static struct ttm_tt *cirrus_ttm_tt_create(struct ttm_bo_device *bdev,
 				 unsigned long size, uint32_t page_flags,
 				 struct page *dummy_read_page)
 {
@@ -259,7 +259,9 @@ int cirrus_mm_init(struct cirrus_device *cirrus)
 
 	ret = ttm_bo_device_init(&cirrus->ttm.bdev,
 				 cirrus->ttm.bo_global_ref.ref.object,
-				 &cirrus_bo_driver, DRM_FILE_PAGE_OFFSET,
+				 &cirrus_bo_driver,
+				 dev->anon_inode->i_mapping,
+				 DRM_FILE_PAGE_OFFSET,
 				 true);
 	if (ret) {
 		DRM_ERROR("Error initialising bo driver; %d\n", ret);
@@ -329,7 +331,6 @@ int cirrus_bo_create(struct drm_device *dev, int size, int align,
 	}
 
 	cirrusbo->bo.bdev = &cirrus->ttm.bdev;
-	cirrusbo->bo.bdev->dev_mapping = dev->dev_mapping;
 
 	cirrus_ttm_placement(cirrusbo, TTM_PL_FLAG_VRAM | TTM_PL_FLAG_SYSTEM);
 
@@ -375,26 +376,6 @@ int cirrus_bo_pin(struct cirrus_bo *bo, u32 pl_flag, u64 *gpu_addr)
 	return 0;
 }
 
-int cirrus_bo_unpin(struct cirrus_bo *bo)
-{
-	int i, ret;
-	if (!bo->pin_count) {
-		DRM_ERROR("unpin bad %p\n", bo);
-		return 0;
-	}
-	bo->pin_count--;
-	if (bo->pin_count)
-		return 0;
-
-	for (i = 0; i < bo->placement.num_placement ; i++)
-		bo->placements[i] &= ~TTM_PL_FLAG_NO_EVICT;
-	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
-	if (ret)
-		return ret;
-
-	return 0;
-}
-
 int cirrus_bo_push_sysram(struct cirrus_bo *bo)
 {
 	int i, ret;
diff --git a/drivers/gpu/drm/drm_agpsupport.c b/drivers/gpu/drm/drm_agpsupport.c
index e301d653d97..dde205cef38 100644
--- a/drivers/gpu/drm/drm_agpsupport.c
+++ b/drivers/gpu/drm/drm_agpsupport.c
@@ -53,7 +53,7 @@
  */
 int drm_agp_info(struct drm_device *dev, struct drm_agp_info *info)
 {
-	DRM_AGP_KERN *kern;
+	struct agp_kern_info *kern;
 
 	if (!dev->agp || !dev->agp->acquired)
 		return -EINVAL;
@@ -198,17 +198,15 @@ int drm_agp_enable_ioctl(struct drm_device *dev, void *data,
 int drm_agp_alloc(struct drm_device *dev, struct drm_agp_buffer *request)
 {
 	struct drm_agp_mem *entry;
-	DRM_AGP_MEM *memory;
+	struct agp_memory *memory;
 	unsigned long pages;
 	u32 type;
 
 	if (!dev->agp || !dev->agp->acquired)
 		return -EINVAL;
-	if (!(entry = kmalloc(sizeof(*entry), GFP_KERNEL)))
+	if (!(entry = kzalloc(sizeof(*entry), GFP_KERNEL)))
 		return -ENOMEM;
 
-	memset(entry, 0, sizeof(*entry));
-
 	pages = (request->size + PAGE_SIZE - 1) / PAGE_SIZE;
 	type = (u32) request->type;
 	if (!(memory = agp_allocate_memory(dev->agp->bridge, pages, type))) {
@@ -393,14 +391,16 @@ int drm_agp_free_ioctl(struct drm_device *dev, void *data,
  * Gets the drm_agp_t structure which is made available by the agpgart module
  * via the inter_module_* functions. Creates and initializes a drm_agp_head
  * structure.
+ *
+ * Note that final cleanup of the kmalloced structure is directly done in
+ * drm_pci_agp_destroy.
  */
 struct drm_agp_head *drm_agp_init(struct drm_device *dev)
 {
 	struct drm_agp_head *head = NULL;
 
-	if (!(head = kmalloc(sizeof(*head), GFP_KERNEL)))
+	if (!(head = kzalloc(sizeof(*head), GFP_KERNEL)))
 		return NULL;
-	memset((void *)head, 0, sizeof(*head));
 	head->bridge = agp_find_bridge(dev->pdev);
 	if (!head->bridge) {
 		if (!(head->bridge = agp_backend_acquire(dev->pdev))) {
@@ -439,7 +439,7 @@ void drm_agp_clear(struct drm_device *dev)
 {
 	struct drm_agp_mem *entry, *tempe;
 
-	if (!drm_core_has_AGP(dev) || !dev->agp)
+	if (!dev->agp)
 		return;
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return;
@@ -460,35 +460,20 @@ void drm_agp_clear(struct drm_device *dev)
 }
 
 /**
- * drm_agp_destroy - Destroy AGP head
- * @dev: DRM device
- *
- * Destroy resources that were previously allocated via drm_agp_initp. Caller
- * must ensure to clean up all AGP resources before calling this. See
- * drm_agp_clear().
- *
- * Call this to destroy AGP heads allocated via drm_agp_init().
- */
-void drm_agp_destroy(struct drm_agp_head *agp)
-{
-	kfree(agp);
-}
-
-/**
  * Binds a collection of pages into AGP memory at the given offset, returning
  * the AGP memory structure containing them.
  *
  * No reference is held on the pages during this time -- it is up to the
  * caller to handle that.
  */
-DRM_AGP_MEM *
+struct agp_memory *
 drm_agp_bind_pages(struct drm_device *dev,
 		   struct page **pages,
 		   unsigned long num_pages,
 		   uint32_t gtt_offset,
 		   u32 type)
 {
-	DRM_AGP_MEM *mem;
+	struct agp_memory *mem;
 	int ret, i;
 
 	DRM_DEBUG("\n");
diff --git a/drivers/gpu/drm/drm_buffer.c b/drivers/gpu/drm/drm_buffer.c
index 39a71834031..0406110f83e 100644
--- a/drivers/gpu/drm/drm_buffer.c
+++ b/drivers/gpu/drm/drm_buffer.c
@@ -114,7 +114,7 @@ int drm_buffer_copy_from_user(struct drm_buffer *buf,
 
 	for (idx = 0; idx < nr_pages; ++idx) {
 
-		if (DRM_COPY_FROM_USER(buf->data[idx],
+		if (copy_from_user(buf->data[idx],
 			user_data + idx * PAGE_SIZE,
 			min(PAGE_SIZE, size - idx * PAGE_SIZE))) {
 			DRM_ERROR("Failed to copy user data (%p) to drm buffer"
diff --git a/drivers/gpu/drm/drm_bufs.c b/drivers/gpu/drm/drm_bufs.c
index 471e051d295..68175b54504 100644
--- a/drivers/gpu/drm/drm_bufs.c
+++ b/drivers/gpu/drm/drm_bufs.c
@@ -261,7 +261,7 @@ static int drm_addmap_core(struct drm_device * dev, resource_size_t offset,
 		struct drm_agp_mem *entry;
 		int valid = 0;
 
-		if (!drm_core_has_AGP(dev)) {
+		if (!dev->agp) {
 			kfree(map);
 			return -EINVAL;
 		}
@@ -303,9 +303,6 @@ static int drm_addmap_core(struct drm_device * dev, resource_size_t offset,
 
 		break;
 	}
-	case _DRM_GEM:
-		DRM_ERROR("tried to addmap GEM object\n");
-		break;
 	case _DRM_SCATTER_GATHER:
 		if (!dev->sg) {
 			kfree(map);
@@ -366,7 +363,7 @@ static int drm_addmap_core(struct drm_device * dev, resource_size_t offset,
 		list->master = dev->primary->master;
 	*maplist = list;
 	return 0;
-	}
+}
 
 int drm_addmap(struct drm_device * dev, resource_size_t offset,
 	       unsigned int size, enum drm_map_type type,
@@ -483,9 +480,6 @@ int drm_rmmap_locked(struct drm_device *dev, struct drm_local_map *map)
 		dmah.size = map->size;
 		__drm_pci_free(dev, &dmah);
 		break;
-	case _DRM_GEM:
-		DRM_ERROR("tried to rmmap GEM object\n");
-		break;
 	}
 	kfree(map);
 
@@ -662,13 +656,13 @@ int drm_addbufs_agp(struct drm_device * dev, struct drm_buf_desc * request)
 		DRM_DEBUG("zone invalid\n");
 		return -EINVAL;
 	}
-	spin_lock(&dev->count_lock);
+	spin_lock(&dev->buf_lock);
 	if (dev->buf_use) {
-		spin_unlock(&dev->count_lock);
+		spin_unlock(&dev->buf_lock);
 		return -EBUSY;
 	}
 	atomic_inc(&dev->buf_alloc);
-	spin_unlock(&dev->count_lock);
+	spin_unlock(&dev->buf_lock);
 
 	mutex_lock(&dev->struct_mutex);
 	entry = &dma->bufs[order];
@@ -811,13 +805,13 @@ int drm_addbufs_pci(struct drm_device * dev, struct drm_buf_desc * request)
 	page_order = order - PAGE_SHIFT > 0 ? order - PAGE_SHIFT : 0;
 	total = PAGE_SIZE << page_order;
 
-	spin_lock(&dev->count_lock);
+	spin_lock(&dev->buf_lock);
 	if (dev->buf_use) {
-		spin_unlock(&dev->count_lock);
+		spin_unlock(&dev->buf_lock);
 		return -EBUSY;
 	}
 	atomic_inc(&dev->buf_alloc);
-	spin_unlock(&dev->count_lock);
+	spin_unlock(&dev->buf_lock);
 
 	mutex_lock(&dev->struct_mutex);
 	entry = &dma->bufs[order];
@@ -1021,13 +1015,13 @@ static int drm_addbufs_sg(struct drm_device * dev, struct drm_buf_desc * request
 	if (order < DRM_MIN_ORDER || order > DRM_MAX_ORDER)
 		return -EINVAL;
 
-	spin_lock(&dev->count_lock);
+	spin_lock(&dev->buf_lock);
 	if (dev->buf_use) {
-		spin_unlock(&dev->count_lock);
+		spin_unlock(&dev->buf_lock);
 		return -EBUSY;
 	}
 	atomic_inc(&dev->buf_alloc);
-	spin_unlock(&dev->count_lock);
+	spin_unlock(&dev->buf_lock);
 
 	mutex_lock(&dev->struct_mutex);
 	entry = &dma->bufs[order];
@@ -1181,7 +1175,7 @@ int drm_addbufs(struct drm_device *dev, void *data,
  * \param arg pointer to a drm_buf_info structure.
  * \return zero on success or a negative number on failure.
  *
- * Increments drm_device::buf_use while holding the drm_device::count_lock
+ * Increments drm_device::buf_use while holding the drm_device::buf_lock
  * lock, preventing of allocating more buffers after this call. Information
  * about each requested buffer is then copied into user space.
  */
@@ -1202,13 +1196,13 @@ int drm_infobufs(struct drm_device *dev, void *data,
 	if (!dma)
 		return -EINVAL;
 
-	spin_lock(&dev->count_lock);
+	spin_lock(&dev->buf_lock);
 	if (atomic_read(&dev->buf_alloc)) {
-		spin_unlock(&dev->count_lock);
+		spin_unlock(&dev->buf_lock);
 		return -EBUSY;
 	}
 	++dev->buf_use;		/* Can't allocate more after this call */
-	spin_unlock(&dev->count_lock);
+	spin_unlock(&dev->buf_lock);
 
 	for (i = 0, count = 0; i < DRM_MAX_ORDER + 1; i++) {
 		if (dma->bufs[i].buf_count)
@@ -1387,16 +1381,16 @@ int drm_mapbufs(struct drm_device *dev, void *data,
 	if (!dma)
 		return -EINVAL;
 
-	spin_lock(&dev->count_lock);
+	spin_lock(&dev->buf_lock);
 	if (atomic_read(&dev->buf_alloc)) {
-		spin_unlock(&dev->count_lock);
+		spin_unlock(&dev->buf_lock);
 		return -EBUSY;
 	}
 	dev->buf_use++;		/* Can't allocate more after this call */
-	spin_unlock(&dev->count_lock);
+	spin_unlock(&dev->buf_lock);
 
 	if (request->count >= dma->buf_count) {
-		if ((drm_core_has_AGP(dev) && (dma->flags & _DRM_DMA_USE_AGP))
+		if ((dev->agp && (dma->flags & _DRM_DMA_USE_AGP))
 		    || (drm_core_check_feature(dev, DRIVER_SG)
 			&& (dma->flags & _DRM_DMA_USE_SG))) {
 			struct drm_local_map *map = dev->agp_buffer_map;
diff --git a/drivers/gpu/drm/drm_cache.c b/drivers/gpu/drm/drm_cache.c
index bb8f5801218..a6b690626a6 100644
--- a/drivers/gpu/drm/drm_cache.c
+++ b/drivers/gpu/drm/drm_cache.c
@@ -32,6 +32,12 @@
 #include <drm/drmP.h>
 
 #if defined(CONFIG_X86)
+
+/*
+ * clflushopt is an unordered instruction which needs fencing with mfence or
+ * sfence to avoid ordering issues.  For drm_clflush_page this fencing happens
+ * in the caller.
+ */
 static void
 drm_clflush_page(struct page *page)
 {
@@ -44,7 +50,7 @@ drm_clflush_page(struct page *page)
 
 	page_virtual = kmap_atomic(page);
 	for (i = 0; i < PAGE_SIZE; i += size)
-		clflush(page_virtual + i);
+		clflushopt(page_virtual + i);
 	kunmap_atomic(page_virtual);
 }
 
@@ -125,15 +131,15 @@ drm_clflush_sg(struct sg_table *st)
 EXPORT_SYMBOL(drm_clflush_sg);
 
 void
-drm_clflush_virt_range(char *addr, unsigned long length)
+drm_clflush_virt_range(void *addr, unsigned long length)
 {
 #if defined(CONFIG_X86)
 	if (cpu_has_clflush) {
-		char *end = addr + length;
+		void *end = addr + length;
 		mb();
 		for (; addr < end; addr += boot_cpu_data.x86_clflush_size)
-			clflush(addr);
-		clflush(end - 1);
+			clflushopt(addr);
+		clflushopt(end - 1);
 		mb();
 		return;
 	}
diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c
index d6cf77c472e..fe94cc10cd3 100644
--- a/drivers/gpu/drm/drm_crtc.c
+++ b/drivers/gpu/drm/drm_crtc.c
@@ -37,35 +37,78 @@
 #include <drm/drm_crtc.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_fourcc.h>
+#include <drm/drm_modeset_lock.h>
+
+#include "drm_crtc_internal.h"
 
 /**
  * drm_modeset_lock_all - take all modeset locks
  * @dev: drm device
  *
  * This function takes all modeset locks, suitable where a more fine-grained
- * scheme isn't (yet) implemented.
+ * scheme isn't (yet) implemented. Locks must be dropped with
+ * drm_modeset_unlock_all.
  */
 void drm_modeset_lock_all(struct drm_device *dev)
 {
-	struct drm_crtc *crtc;
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_modeset_acquire_ctx *ctx;
+	int ret;
 
-	mutex_lock(&dev->mode_config.mutex);
+	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
+	if (WARN_ON(!ctx))
+		return;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
-		mutex_lock_nest_lock(&crtc->mutex, &dev->mode_config.mutex);
+	mutex_lock(&config->mutex);
+
+	drm_modeset_acquire_init(ctx, 0);
+
+retry:
+	ret = drm_modeset_lock(&config->connection_mutex, ctx);
+	if (ret)
+		goto fail;
+	ret = drm_modeset_lock_all_crtcs(dev, ctx);
+	if (ret)
+		goto fail;
+
+	WARN_ON(config->acquire_ctx);
+
+	/* now we hold the locks, so now that it is safe, stash the
+	 * ctx for drm_modeset_unlock_all():
+	 */
+	config->acquire_ctx = ctx;
+
+	drm_warn_on_modeset_not_all_locked(dev);
+
+	return;
+
+fail:
+	if (ret == -EDEADLK) {
+		drm_modeset_backoff(ctx);
+		goto retry;
+	}
 }
 EXPORT_SYMBOL(drm_modeset_lock_all);
 
 /**
  * drm_modeset_unlock_all - drop all modeset locks
  * @dev: device
+ *
+ * This function drop all modeset locks taken by drm_modeset_lock_all.
  */
 void drm_modeset_unlock_all(struct drm_device *dev)
 {
-	struct drm_crtc *crtc;
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_modeset_acquire_ctx *ctx = config->acquire_ctx;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
-		mutex_unlock(&crtc->mutex);
+	if (WARN_ON(!ctx))
+		return;
+
+	config->acquire_ctx = NULL;
+	drm_modeset_drop_locks(ctx);
+	drm_modeset_acquire_fini(ctx);
+
+	kfree(ctx);
 
 	mutex_unlock(&dev->mode_config.mutex);
 }
@@ -74,6 +117,8 @@ EXPORT_SYMBOL(drm_modeset_unlock_all);
 /**
  * drm_warn_on_modeset_not_all_locked - check that all modeset locks are locked
  * @dev: device
+ *
+ * Useful as a debug assert.
  */
 void drm_warn_on_modeset_not_all_locked(struct drm_device *dev)
 {
@@ -84,8 +129,9 @@ void drm_warn_on_modeset_not_all_locked(struct drm_device *dev)
 		return;
 
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
-		WARN_ON(!mutex_is_locked(&crtc->mutex));
+		WARN_ON(!drm_modeset_is_locked(&crtc->mutex));
 
+	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
 	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
 }
 EXPORT_SYMBOL(drm_warn_on_modeset_not_all_locked);
@@ -114,6 +160,13 @@ static const struct drm_prop_enum_list drm_dpms_enum_list[] =
 
 DRM_ENUM_NAME_FN(drm_get_dpms_name, drm_dpms_enum_list)
 
+static const struct drm_prop_enum_list drm_plane_type_enum_list[] =
+{
+	{ DRM_PLANE_TYPE_OVERLAY, "Overlay" },
+	{ DRM_PLANE_TYPE_PRIMARY, "Primary" },
+	{ DRM_PLANE_TYPE_CURSOR, "Cursor" },
+};
+
 /*
  * Optional properties
  */
@@ -213,6 +266,17 @@ static const struct drm_prop_enum_list drm_encoder_enum_list[] =
 	{ DRM_MODE_ENCODER_TVDAC, "TV" },
 	{ DRM_MODE_ENCODER_VIRTUAL, "Virtual" },
 	{ DRM_MODE_ENCODER_DSI, "DSI" },
+	{ DRM_MODE_ENCODER_DPMST, "DP MST" },
+};
+
+static const struct drm_prop_enum_list drm_subpixel_enum_list[] =
+{
+	{ SubPixelUnknown, "Unknown" },
+	{ SubPixelHorizontalRGB, "Horizontal RGB" },
+	{ SubPixelHorizontalBGR, "Horizontal BGR" },
+	{ SubPixelVerticalRGB, "Vertical RGB" },
+	{ SubPixelVerticalBGR, "Vertical BGR" },
+	{ SubPixelNone, "None" },
 };
 
 void drm_connector_ida_init(void)
@@ -231,28 +295,13 @@ void drm_connector_ida_destroy(void)
 		ida_destroy(&drm_connector_enum_list[i].ida);
 }
 
-const char *drm_get_encoder_name(const struct drm_encoder *encoder)
-{
-	static char buf[32];
-
-	snprintf(buf, 32, "%s-%d",
-		 drm_encoder_enum_list[encoder->encoder_type].name,
-		 encoder->base.id);
-	return buf;
-}
-EXPORT_SYMBOL(drm_get_encoder_name);
-
-const char *drm_get_connector_name(const struct drm_connector *connector)
-{
-	static char buf[32];
-
-	snprintf(buf, 32, "%s-%d",
-		 drm_connector_enum_list[connector->connector_type].name,
-		 connector->connector_type_id);
-	return buf;
-}
-EXPORT_SYMBOL(drm_get_connector_name);
-
+/**
+ * drm_get_connector_status_name - return a string for connector status
+ * @status: connector status to compute name of
+ *
+ * In contrast to the other drm_get_*_name functions this one here returns a
+ * const pointer and hence is threadsafe.
+ */
 const char *drm_get_connector_status_name(enum drm_connector_status status)
 {
 	if (status == connector_status_connected)
@@ -264,11 +313,33 @@ const char *drm_get_connector_status_name(enum drm_connector_status status)
 }
 EXPORT_SYMBOL(drm_get_connector_status_name);
 
+/**
+ * drm_get_subpixel_order_name - return a string for a given subpixel enum
+ * @order: enum of subpixel_order
+ *
+ * Note you could abuse this and return something out of bounds, but that
+ * would be a caller error.  No unscrubbed user data should make it here.
+ */
+const char *drm_get_subpixel_order_name(enum subpixel_order order)
+{
+	return drm_subpixel_enum_list[order].name;
+}
+EXPORT_SYMBOL(drm_get_subpixel_order_name);
+
 static char printable_char(int c)
 {
 	return isascii(c) && isprint(c) ? c : '?';
 }
 
+/**
+ * drm_get_format_name - return a string for drm fourcc format
+ * @format: format to compute name of
+ *
+ * Note that the buffer used by this function is globally shared and owned by
+ * the function itself.
+ *
+ * FIXME: This isn't really multithreading safe.
+ */
 const char *drm_get_format_name(uint32_t format)
 {
 	static char buf[32];
@@ -293,14 +364,16 @@ EXPORT_SYMBOL(drm_get_format_name);
  * @obj_type: object type
  *
  * Create a unique identifier based on @ptr in @dev's identifier space.  Used
- * for tracking modes, CRTCs and connectors.
+ * for tracking modes, CRTCs and connectors. Note that despite the _get postfix
+ * modeset identifiers are _not_ reference counted. Hence don't use this for
+ * reference counted modeset objects like framebuffers.
  *
- * RETURNS:
+ * Returns:
  * New unique (relative to other objects in @dev) integer identifier for the
  * object.
  */
-static int drm_mode_object_get(struct drm_device *dev,
-			       struct drm_mode_object *obj, uint32_t obj_type)
+int drm_mode_object_get(struct drm_device *dev,
+			struct drm_mode_object *obj, uint32_t obj_type)
 {
 	int ret;
 
@@ -324,16 +397,33 @@ static int drm_mode_object_get(struct drm_device *dev,
  * @dev: DRM device
  * @object: object to free
  *
- * Free @id from @dev's unique identifier pool.
+ * Free @id from @dev's unique identifier pool. Note that despite the _get
+ * postfix modeset identifiers are _not_ reference counted. Hence don't use this
+ * for reference counted modeset objects like framebuffers.
  */
-static void drm_mode_object_put(struct drm_device *dev,
-				struct drm_mode_object *object)
+void drm_mode_object_put(struct drm_device *dev,
+			 struct drm_mode_object *object)
 {
 	mutex_lock(&dev->mode_config.idr_mutex);
 	idr_remove(&dev->mode_config.crtc_idr, object->id);
 	mutex_unlock(&dev->mode_config.idr_mutex);
 }
 
+static struct drm_mode_object *_object_find(struct drm_device *dev,
+		uint32_t id, uint32_t type)
+{
+	struct drm_mode_object *obj = NULL;
+
+	mutex_lock(&dev->mode_config.idr_mutex);
+	obj = idr_find(&dev->mode_config.crtc_idr, id);
+	if (!obj || (type != DRM_MODE_OBJECT_ANY && obj->type != type) ||
+	    (obj->id != id))
+		obj = NULL;
+	mutex_unlock(&dev->mode_config.idr_mutex);
+
+	return obj;
+}
+
 /**
  * drm_mode_object_find - look up a drm object with static lifetime
  * @dev: drm device
@@ -341,7 +431,9 @@ static void drm_mode_object_put(struct drm_device *dev,
  * @type: type of the mode object
  *
  * Note that framebuffers cannot be looked up with this functions - since those
- * are reference counted, they need special treatment.
+ * are reference counted, they need special treatment.  Even with
+ * DRM_MODE_OBJECT_ANY (although that will simply return NULL
+ * rather than WARN_ON()).
  */
 struct drm_mode_object *drm_mode_object_find(struct drm_device *dev,
 		uint32_t id, uint32_t type)
@@ -351,13 +443,10 @@ struct drm_mode_object *drm_mode_object_find(struct drm_device *dev,
 	/* Framebuffers are reference counted and need their own lookup
 	 * function.*/
 	WARN_ON(type == DRM_MODE_OBJECT_FB);
-
-	mutex_lock(&dev->mode_config.idr_mutex);
-	obj = idr_find(&dev->mode_config.crtc_idr, id);
-	if (!obj || (obj->type != type) || (obj->id != id))
+	obj = _object_find(dev, id, type);
+	/* don't leak out unref'd fb's */
+	if (obj && (obj->type == DRM_MODE_OBJECT_FB))
 		obj = NULL;
-	mutex_unlock(&dev->mode_config.idr_mutex);
-
 	return obj;
 }
 EXPORT_SYMBOL(drm_mode_object_find);
@@ -377,7 +466,7 @@ EXPORT_SYMBOL(drm_mode_object_find);
  * since all the fb attributes are invariant over its lifetime, no further
  * locking but only correct reference counting is required.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, error code on failure.
  */
 int drm_framebuffer_init(struct drm_device *dev, struct drm_framebuffer *fb,
@@ -438,7 +527,7 @@ static struct drm_framebuffer *__drm_framebuffer_lookup(struct drm_device *dev,
  *
  * If successful, this grabs an additional reference to the framebuffer -
  * callers need to make sure to eventually unreference the returned framebuffer
- * again.
+ * again, using @drm_framebuffer_unreference.
  */
 struct drm_framebuffer *drm_framebuffer_lookup(struct drm_device *dev,
 					       uint32_t id)
@@ -463,7 +552,7 @@ EXPORT_SYMBOL(drm_framebuffer_lookup);
  */
 void drm_framebuffer_unreference(struct drm_framebuffer *fb)
 {
-	DRM_DEBUG("FB ID: %d\n", fb->base.id);
+	DRM_DEBUG("%p: FB ID: %d (%d)\n", fb, fb->base.id, atomic_read(&fb->refcount.refcount));
 	kref_put(&fb->refcount, drm_framebuffer_free);
 }
 EXPORT_SYMBOL(drm_framebuffer_unreference);
@@ -471,10 +560,12 @@ EXPORT_SYMBOL(drm_framebuffer_unreference);
 /**
  * drm_framebuffer_reference - incr the fb refcnt
  * @fb: framebuffer
+ *
+ * This functions increments the fb's refcount.
  */
 void drm_framebuffer_reference(struct drm_framebuffer *fb)
 {
-	DRM_DEBUG("FB ID: %d\n", fb->base.id);
+	DRM_DEBUG("%p: FB ID: %d (%d)\n", fb, fb->base.id, atomic_read(&fb->refcount.refcount));
 	kref_get(&fb->refcount);
 }
 EXPORT_SYMBOL(drm_framebuffer_reference);
@@ -486,7 +577,7 @@ static void drm_framebuffer_free_bug(struct kref *kref)
 
 static void __drm_framebuffer_unreference(struct drm_framebuffer *fb)
 {
-	DRM_DEBUG("FB ID: %d\n", fb->base.id);
+	DRM_DEBUG("%p: FB ID: %d (%d)\n", fb, fb->base.id, atomic_read(&fb->refcount.refcount));
 	kref_put(&fb->refcount, drm_framebuffer_free_bug);
 }
 
@@ -527,8 +618,9 @@ EXPORT_SYMBOL(drm_framebuffer_unregister_private);
  * drm_framebuffer_cleanup - remove a framebuffer object
  * @fb: framebuffer to remove
  *
- * Cleanup references to a user-created framebuffer. This function is intended
- * to be used from the drivers ->destroy callback.
+ * Cleanup framebuffer. This function is intended to be used from the drivers
+ * ->destroy callback. It can also be used to clean up driver private
+ *  framebuffers embedded into a larger structure.
  *
  * Note that this function does not remove the fb from active usuage - if it is
  * still used anywhere, hilarity can ensue since userspace could call getfb on
@@ -591,7 +683,7 @@ void drm_framebuffer_remove(struct drm_framebuffer *fb)
 		drm_modeset_lock_all(dev);
 		/* remove from any CRTC */
 		list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
-			if (crtc->fb == fb) {
+			if (crtc->primary->fb == fb) {
 				/* should turn off the crtc */
 				memset(&set, 0, sizeof(struct drm_mode_set));
 				set.crtc = crtc;
@@ -613,20 +705,28 @@ void drm_framebuffer_remove(struct drm_framebuffer *fb)
 }
 EXPORT_SYMBOL(drm_framebuffer_remove);
 
+DEFINE_WW_CLASS(crtc_ww_class);
+
 /**
- * drm_crtc_init - Initialise a new CRTC object
+ * drm_crtc_init_with_planes - Initialise a new CRTC object with
+ *    specified primary and cursor planes.
  * @dev: DRM device
  * @crtc: CRTC object to init
+ * @primary: Primary plane for CRTC
+ * @cursor: Cursor plane for CRTC
  * @funcs: callbacks for the new CRTC
  *
  * Inits a new object created as base part of a driver crtc object.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, error code on failure.
  */
-int drm_crtc_init(struct drm_device *dev, struct drm_crtc *crtc,
-		   const struct drm_crtc_funcs *funcs)
+int drm_crtc_init_with_planes(struct drm_device *dev, struct drm_crtc *crtc,
+			      struct drm_plane *primary,
+			      void *cursor,
+			      const struct drm_crtc_funcs *funcs)
 {
+	struct drm_mode_config *config = &dev->mode_config;
 	int ret;
 
 	crtc->dev = dev;
@@ -634,8 +734,9 @@ int drm_crtc_init(struct drm_device *dev, struct drm_crtc *crtc,
 	crtc->invert_dimensions = false;
 
 	drm_modeset_lock_all(dev);
-	mutex_init(&crtc->mutex);
-	mutex_lock_nest_lock(&crtc->mutex, &dev->mode_config.mutex);
+	drm_modeset_lock_init(&crtc->mutex);
+	/* dropped by _unlock_all(): */
+	drm_modeset_lock(&crtc->mutex, config->acquire_ctx);
 
 	ret = drm_mode_object_get(dev, &crtc->base, DRM_MODE_OBJECT_CRTC);
 	if (ret)
@@ -643,15 +744,19 @@ int drm_crtc_init(struct drm_device *dev, struct drm_crtc *crtc,
 
 	crtc->base.properties = &crtc->properties;
 
-	list_add_tail(&crtc->head, &dev->mode_config.crtc_list);
-	dev->mode_config.num_crtc++;
+	list_add_tail(&crtc->head, &config->crtc_list);
+	config->num_crtc++;
+
+	crtc->primary = primary;
+	if (primary)
+		primary->possible_crtcs = 1 << drm_crtc_index(crtc);
 
  out:
 	drm_modeset_unlock_all(dev);
 
 	return ret;
 }
-EXPORT_SYMBOL(drm_crtc_init);
+EXPORT_SYMBOL(drm_crtc_init_with_planes);
 
 /**
  * drm_crtc_cleanup - Clean up the core crtc usage
@@ -668,6 +773,8 @@ void drm_crtc_cleanup(struct drm_crtc *crtc)
 	kfree(crtc->gamma_store);
 	crtc->gamma_store = NULL;
 
+	drm_modeset_lock_fini(&crtc->mutex);
+
 	drm_mode_object_put(dev, &crtc->base);
 	list_del(&crtc->head);
 	dev->mode_config.num_crtc--;
@@ -675,18 +782,27 @@ void drm_crtc_cleanup(struct drm_crtc *crtc)
 EXPORT_SYMBOL(drm_crtc_cleanup);
 
 /**
- * drm_mode_probed_add - add a mode to a connector's probed mode list
- * @connector: connector the new mode
- * @mode: mode data
+ * drm_crtc_index - find the index of a registered CRTC
+ * @crtc: CRTC to find index for
  *
- * Add @mode to @connector's mode list for later use.
+ * Given a registered CRTC, return the index of that CRTC within a DRM
+ * device's list of CRTCs.
  */
-void drm_mode_probed_add(struct drm_connector *connector,
-			 struct drm_display_mode *mode)
+unsigned int drm_crtc_index(struct drm_crtc *crtc)
 {
-	list_add_tail(&mode->head, &connector->probed_modes);
+	unsigned int index = 0;
+	struct drm_crtc *tmp;
+
+	list_for_each_entry(tmp, &crtc->dev->mode_config.crtc_list, head) {
+		if (tmp == crtc)
+			return index;
+
+		index++;
+	}
+
+	BUG();
 }
-EXPORT_SYMBOL(drm_mode_probed_add);
+EXPORT_SYMBOL(drm_crtc_index);
 
 /*
  * drm_mode_remove - remove and free a mode
@@ -712,7 +828,7 @@ static void drm_mode_remove(struct drm_connector *connector,
  * Initialises a preallocated connector. Connectors should be
  * subclassed as part of driver connector objects.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, error code on failure.
  */
 int drm_connector_init(struct drm_device *dev,
@@ -728,7 +844,7 @@ int drm_connector_init(struct drm_device *dev,
 
 	ret = drm_mode_object_get(dev, &connector->base, DRM_MODE_OBJECT_CONNECTOR);
 	if (ret)
-		goto out;
+		goto out_unlock;
 
 	connector->base.properties = &connector->properties;
 	connector->dev = dev;
@@ -738,9 +854,17 @@ int drm_connector_init(struct drm_device *dev,
 		ida_simple_get(connector_ida, 1, 0, GFP_KERNEL);
 	if (connector->connector_type_id < 0) {
 		ret = connector->connector_type_id;
-		drm_mode_object_put(dev, &connector->base);
-		goto out;
+		goto out_put;
 	}
+	connector->name =
+		kasprintf(GFP_KERNEL, "%s-%d",
+			  drm_connector_enum_list[connector_type].name,
+			  connector->connector_type_id);
+	if (!connector->name) {
+		ret = -ENOMEM;
+		goto out_put;
+	}
+
 	INIT_LIST_HEAD(&connector->probed_modes);
 	INIT_LIST_HEAD(&connector->modes);
 	connector->edid_blob_ptr = NULL;
@@ -757,7 +881,11 @@ int drm_connector_init(struct drm_device *dev,
 	drm_object_attach_property(&connector->base,
 				      dev->mode_config.dpms_property, 0);
 
- out:
+out_put:
+	if (ret)
+		drm_mode_object_put(dev, &connector->base);
+
+out_unlock:
 	drm_modeset_unlock_all(dev);
 
 	return ret;
@@ -785,11 +913,21 @@ void drm_connector_cleanup(struct drm_connector *connector)
 		   connector->connector_type_id);
 
 	drm_mode_object_put(dev, &connector->base);
+	kfree(connector->name);
+	connector->name = NULL;
 	list_del(&connector->head);
 	dev->mode_config.num_connector--;
 }
 EXPORT_SYMBOL(drm_connector_cleanup);
 
+/**
+ * drm_connector_unplug_all - unregister connector userspace interfaces
+ * @dev: drm device
+ *
+ * This function unregisters all connector userspace interfaces in sysfs. Should
+ * be call when the device is disconnected, e.g. from an usb driver's
+ * ->disconnect callback.
+ */
 void drm_connector_unplug_all(struct drm_device *dev)
 {
 	struct drm_connector *connector;
@@ -801,6 +939,18 @@ void drm_connector_unplug_all(struct drm_device *dev)
 }
 EXPORT_SYMBOL(drm_connector_unplug_all);
 
+/**
+ * drm_bridge_init - initialize a drm transcoder/bridge
+ * @dev: drm device
+ * @bridge: transcoder/bridge to set up
+ * @funcs: bridge function table
+ *
+ * Initialises a preallocated bridge. Bridges should be
+ * subclassed as part of driver connector objects.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
 int drm_bridge_init(struct drm_device *dev, struct drm_bridge *bridge,
 		const struct drm_bridge_funcs *funcs)
 {
@@ -824,6 +974,12 @@ int drm_bridge_init(struct drm_device *dev, struct drm_bridge *bridge,
 }
 EXPORT_SYMBOL(drm_bridge_init);
 
+/**
+ * drm_bridge_cleanup - cleans up an initialised bridge
+ * @bridge: bridge to cleanup
+ *
+ * Cleans up the bridge but doesn't free the object.
+ */
 void drm_bridge_cleanup(struct drm_bridge *bridge)
 {
 	struct drm_device *dev = bridge->dev;
@@ -836,6 +992,19 @@ void drm_bridge_cleanup(struct drm_bridge *bridge)
 }
 EXPORT_SYMBOL(drm_bridge_cleanup);
 
+/**
+ * drm_encoder_init - Init a preallocated encoder
+ * @dev: drm device
+ * @encoder: the encoder to init
+ * @funcs: callbacks for this encoder
+ * @encoder_type: user visible type of the encoder
+ *
+ * Initialises a preallocated encoder. Encoder should be
+ * subclassed as part of driver encoder objects.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
 int drm_encoder_init(struct drm_device *dev,
 		      struct drm_encoder *encoder,
 		      const struct drm_encoder_funcs *funcs,
@@ -847,27 +1016,46 @@ int drm_encoder_init(struct drm_device *dev,
 
 	ret = drm_mode_object_get(dev, &encoder->base, DRM_MODE_OBJECT_ENCODER);
 	if (ret)
-		goto out;
+		goto out_unlock;
 
 	encoder->dev = dev;
 	encoder->encoder_type = encoder_type;
 	encoder->funcs = funcs;
+	encoder->name = kasprintf(GFP_KERNEL, "%s-%d",
+				  drm_encoder_enum_list[encoder_type].name,
+				  encoder->base.id);
+	if (!encoder->name) {
+		ret = -ENOMEM;
+		goto out_put;
+	}
 
 	list_add_tail(&encoder->head, &dev->mode_config.encoder_list);
 	dev->mode_config.num_encoder++;
 
- out:
+out_put:
+	if (ret)
+		drm_mode_object_put(dev, &encoder->base);
+
+out_unlock:
 	drm_modeset_unlock_all(dev);
 
 	return ret;
 }
 EXPORT_SYMBOL(drm_encoder_init);
 
+/**
+ * drm_encoder_cleanup - cleans up an initialised encoder
+ * @encoder: encoder to cleanup
+ *
+ * Cleans up the encoder but doesn't free the object.
+ */
 void drm_encoder_cleanup(struct drm_encoder *encoder)
 {
 	struct drm_device *dev = encoder->dev;
 	drm_modeset_lock_all(dev);
 	drm_mode_object_put(dev, &encoder->base);
+	kfree(encoder->name);
+	encoder->name = NULL;
 	list_del(&encoder->head);
 	dev->mode_config.num_encoder--;
 	drm_modeset_unlock_all(dev);
@@ -875,25 +1063,25 @@ void drm_encoder_cleanup(struct drm_encoder *encoder)
 EXPORT_SYMBOL(drm_encoder_cleanup);
 
 /**
- * drm_plane_init - Initialise a new plane object
+ * drm_universal_plane_init - Initialize a new universal plane object
  * @dev: DRM device
  * @plane: plane object to init
  * @possible_crtcs: bitmask of possible CRTCs
  * @funcs: callbacks for the new plane
  * @formats: array of supported formats (%DRM_FORMAT_*)
  * @format_count: number of elements in @formats
- * @priv: plane is private (hidden from userspace)?
+ * @type: type of plane (overlay, primary, cursor)
  *
- * Inits a new object created as base part of a driver plane object.
+ * Initializes a plane object of type @type.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, error code on failure.
  */
-int drm_plane_init(struct drm_device *dev, struct drm_plane *plane,
-		   unsigned long possible_crtcs,
-		   const struct drm_plane_funcs *funcs,
-		   const uint32_t *formats, uint32_t format_count,
-		   bool priv)
+int drm_universal_plane_init(struct drm_device *dev, struct drm_plane *plane,
+			     unsigned long possible_crtcs,
+			     const struct drm_plane_funcs *funcs,
+			     const uint32_t *formats, uint32_t format_count,
+			     enum drm_plane_type type)
 {
 	int ret;
 
@@ -918,23 +1106,53 @@ int drm_plane_init(struct drm_device *dev, struct drm_plane *plane,
 	memcpy(plane->format_types, formats, format_count * sizeof(uint32_t));
 	plane->format_count = format_count;
 	plane->possible_crtcs = possible_crtcs;
+	plane->type = type;
 
-	/* private planes are not exposed to userspace, but depending on
-	 * display hardware, might be convenient to allow sharing programming
-	 * for the scanout engine with the crtc implementation.
-	 */
-	if (!priv) {
-		list_add_tail(&plane->head, &dev->mode_config.plane_list);
-		dev->mode_config.num_plane++;
-	} else {
-		INIT_LIST_HEAD(&plane->head);
-	}
+	list_add_tail(&plane->head, &dev->mode_config.plane_list);
+	dev->mode_config.num_total_plane++;
+	if (plane->type == DRM_PLANE_TYPE_OVERLAY)
+		dev->mode_config.num_overlay_plane++;
+
+	drm_object_attach_property(&plane->base,
+				   dev->mode_config.plane_type_property,
+				   plane->type);
 
  out:
 	drm_modeset_unlock_all(dev);
 
 	return ret;
 }
+EXPORT_SYMBOL(drm_universal_plane_init);
+
+/**
+ * drm_plane_init - Initialize a legacy plane
+ * @dev: DRM device
+ * @plane: plane object to init
+ * @possible_crtcs: bitmask of possible CRTCs
+ * @funcs: callbacks for the new plane
+ * @formats: array of supported formats (%DRM_FORMAT_*)
+ * @format_count: number of elements in @formats
+ * @is_primary: plane type (primary vs overlay)
+ *
+ * Legacy API to initialize a DRM plane.
+ *
+ * New drivers should call drm_universal_plane_init() instead.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int drm_plane_init(struct drm_device *dev, struct drm_plane *plane,
+		   unsigned long possible_crtcs,
+		   const struct drm_plane_funcs *funcs,
+		   const uint32_t *formats, uint32_t format_count,
+		   bool is_primary)
+{
+	enum drm_plane_type type;
+
+	type = is_primary ? DRM_PLANE_TYPE_PRIMARY : DRM_PLANE_TYPE_OVERLAY;
+	return drm_universal_plane_init(dev, plane, possible_crtcs, funcs,
+					formats, format_count, type);
+}
 EXPORT_SYMBOL(drm_plane_init);
 
 /**
@@ -952,11 +1170,13 @@ void drm_plane_cleanup(struct drm_plane *plane)
 	drm_modeset_lock_all(dev);
 	kfree(plane->format_types);
 	drm_mode_object_put(dev, &plane->base);
-	/* if not added to a list, it must be a private plane */
-	if (!list_empty(&plane->head)) {
-		list_del(&plane->head);
-		dev->mode_config.num_plane--;
-	}
+
+	BUG_ON(list_empty(&plane->head));
+
+	list_del(&plane->head);
+	dev->mode_config.num_total_plane--;
+	if (plane->type == DRM_PLANE_TYPE_OVERLAY)
+		dev->mode_config.num_overlay_plane--;
 	drm_modeset_unlock_all(dev);
 }
 EXPORT_SYMBOL(drm_plane_cleanup);
@@ -972,65 +1192,24 @@ EXPORT_SYMBOL(drm_plane_cleanup);
  */
 void drm_plane_force_disable(struct drm_plane *plane)
 {
+	struct drm_framebuffer *old_fb = plane->fb;
 	int ret;
 
-	if (!plane->fb)
+	if (!old_fb)
 		return;
 
 	ret = plane->funcs->disable_plane(plane);
-	if (ret)
+	if (ret) {
 		DRM_ERROR("failed to disable plane with busy fb\n");
+		return;
+	}
 	/* disconnect the plane from the fb and crtc: */
-	__drm_framebuffer_unreference(plane->fb);
+	__drm_framebuffer_unreference(old_fb);
 	plane->fb = NULL;
 	plane->crtc = NULL;
 }
 EXPORT_SYMBOL(drm_plane_force_disable);
 
-/**
- * drm_mode_create - create a new display mode
- * @dev: DRM device
- *
- * Create a new drm_display_mode, give it an ID, and return it.
- *
- * RETURNS:
- * Pointer to new mode on success, NULL on error.
- */
-struct drm_display_mode *drm_mode_create(struct drm_device *dev)
-{
-	struct drm_display_mode *nmode;
-
-	nmode = kzalloc(sizeof(struct drm_display_mode), GFP_KERNEL);
-	if (!nmode)
-		return NULL;
-
-	if (drm_mode_object_get(dev, &nmode->base, DRM_MODE_OBJECT_MODE)) {
-		kfree(nmode);
-		return NULL;
-	}
-
-	return nmode;
-}
-EXPORT_SYMBOL(drm_mode_create);
-
-/**
- * drm_mode_destroy - remove a mode
- * @dev: DRM device
- * @mode: mode to remove
- *
- * Free @mode's unique identifier, then free it.
- */
-void drm_mode_destroy(struct drm_device *dev, struct drm_display_mode *mode)
-{
-	if (!mode)
-		return;
-
-	drm_mode_object_put(dev, &mode->base);
-
-	kfree(mode);
-}
-EXPORT_SYMBOL(drm_mode_destroy);
-
 static int drm_mode_create_standard_connector_properties(struct drm_device *dev)
 {
 	struct drm_property *edid;
@@ -1052,6 +1231,21 @@ static int drm_mode_create_standard_connector_properties(struct drm_device *dev)
 	return 0;
 }
 
+static int drm_mode_create_standard_plane_properties(struct drm_device *dev)
+{
+	struct drm_property *type;
+
+	/*
+	 * Standard properties (apply to all planes)
+	 */
+	type = drm_property_create_enum(dev, DRM_MODE_PROP_IMMUTABLE,
+					"type", drm_plane_type_enum_list,
+					ARRAY_SIZE(drm_plane_type_enum_list));
+	dev->mode_config.plane_type_property = type;
+
+	return 0;
+}
+
 /**
  * drm_mode_create_dvi_i_properties - create DVI-I specific connector properties
  * @dev: DRM device
@@ -1234,6 +1428,16 @@ static int drm_mode_group_init(struct drm_device *dev, struct drm_mode_group *gr
 	return 0;
 }
 
+void drm_mode_group_destroy(struct drm_mode_group *group)
+{
+	kfree(group->id_list);
+	group->id_list = NULL;
+}
+
+/*
+ * NOTE: Driver's shouldn't ever call drm_mode_group_init_legacy_group - it is
+ * the drm core's responsibility to set up mode control groups.
+ */
 int drm_mode_group_init_legacy_group(struct drm_device *dev,
 				     struct drm_mode_group *group)
 {
@@ -1310,7 +1514,7 @@ static void drm_crtc_convert_to_umode(struct drm_mode_modeinfo *out,
  * Convert a drm_mode_modeinfo into a drm_display_mode structure to return to
  * the caller.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 static int drm_crtc_convert_umode(struct drm_display_mode *out,
@@ -1353,7 +1557,7 @@ static int drm_crtc_convert_umode(struct drm_display_mode *out,
  *
  * Called by the user via ioctl.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 int drm_mode_getresources(struct drm_device *dev, void *data,
@@ -1406,9 +1610,9 @@ int drm_mode_getresources(struct drm_device *dev, void *data,
 	mutex_unlock(&file_priv->fbs_lock);
 
 	drm_modeset_lock_all(dev);
-	mode_group = &file_priv->master->minor->mode_group;
-	if (file_priv->master->minor->type == DRM_MINOR_CONTROL) {
+	if (!drm_is_primary_client(file_priv)) {
 
+		mode_group = NULL;
 		list_for_each(lh, &dev->mode_config.crtc_list)
 			crtc_count++;
 
@@ -1419,6 +1623,7 @@ int drm_mode_getresources(struct drm_device *dev, void *data,
 			encoder_count++;
 	} else {
 
+		mode_group = &file_priv->master->minor->mode_group;
 		crtc_count = mode_group->num_crtcs;
 		connector_count = mode_group->num_connectors;
 		encoder_count = mode_group->num_encoders;
@@ -1433,7 +1638,7 @@ int drm_mode_getresources(struct drm_device *dev, void *data,
 	if (card_res->count_crtcs >= crtc_count) {
 		copied = 0;
 		crtc_id = (uint32_t __user *)(unsigned long)card_res->crtc_id_ptr;
-		if (file_priv->master->minor->type == DRM_MINOR_CONTROL) {
+		if (!mode_group) {
 			list_for_each_entry(crtc, &dev->mode_config.crtc_list,
 					    head) {
 				DRM_DEBUG_KMS("[CRTC:%d]\n", crtc->base.id);
@@ -1460,12 +1665,12 @@ int drm_mode_getresources(struct drm_device *dev, void *data,
 	if (card_res->count_encoders >= encoder_count) {
 		copied = 0;
 		encoder_id = (uint32_t __user *)(unsigned long)card_res->encoder_id_ptr;
-		if (file_priv->master->minor->type == DRM_MINOR_CONTROL) {
+		if (!mode_group) {
 			list_for_each_entry(encoder,
 					    &dev->mode_config.encoder_list,
 					    head) {
 				DRM_DEBUG_KMS("[ENCODER:%d:%s]\n", encoder->base.id,
-						drm_get_encoder_name(encoder));
+						encoder->name);
 				if (put_user(encoder->base.id, encoder_id +
 					     copied)) {
 					ret = -EFAULT;
@@ -1491,13 +1696,13 @@ int drm_mode_getresources(struct drm_device *dev, void *data,
 	if (card_res->count_connectors >= connector_count) {
 		copied = 0;
 		connector_id = (uint32_t __user *)(unsigned long)card_res->connector_id_ptr;
-		if (file_priv->master->minor->type == DRM_MINOR_CONTROL) {
+		if (!mode_group) {
 			list_for_each_entry(connector,
 					    &dev->mode_config.connector_list,
 					    head) {
 				DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
 					connector->base.id,
-					drm_get_connector_name(connector));
+					connector->name);
 				if (put_user(connector->base.id,
 					     connector_id + copied)) {
 					ret = -EFAULT;
@@ -1538,7 +1743,7 @@ out:
  *
  * Called by the user via ioctl.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 int drm_mode_getcrtc(struct drm_device *dev,
@@ -1546,7 +1751,6 @@ int drm_mode_getcrtc(struct drm_device *dev,
 {
 	struct drm_mode_crtc *crtc_resp = data;
 	struct drm_crtc *crtc;
-	struct drm_mode_object *obj;
 	int ret = 0;
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
@@ -1554,19 +1758,17 @@ int drm_mode_getcrtc(struct drm_device *dev,
 
 	drm_modeset_lock_all(dev);
 
-	obj = drm_mode_object_find(dev, crtc_resp->crtc_id,
-				   DRM_MODE_OBJECT_CRTC);
-	if (!obj) {
+	crtc = drm_crtc_find(dev, crtc_resp->crtc_id);
+	if (!crtc) {
 		ret = -ENOENT;
 		goto out;
 	}
-	crtc = obj_to_crtc(obj);
 
 	crtc_resp->x = crtc->x;
 	crtc_resp->y = crtc->y;
 	crtc_resp->gamma_size = crtc->gamma_size;
-	if (crtc->fb)
-		crtc_resp->fb_id = crtc->fb->base.id;
+	if (crtc->primary->fb)
+		crtc_resp->fb_id = crtc->primary->fb->base.id;
 	else
 		crtc_resp->fb_id = 0;
 
@@ -1607,14 +1809,13 @@ static bool drm_mode_expose_to_userspace(const struct drm_display_mode *mode,
  *
  * Called by the user via ioctl.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 int drm_mode_getconnector(struct drm_device *dev, void *data,
 			  struct drm_file *file_priv)
 {
 	struct drm_mode_get_connector *out_resp = data;
-	struct drm_mode_object *obj;
 	struct drm_connector *connector;
 	struct drm_display_mode *mode;
 	int mode_count = 0;
@@ -1638,13 +1839,11 @@ int drm_mode_getconnector(struct drm_device *dev, void *data,
 
 	mutex_lock(&dev->mode_config.mutex);
 
-	obj = drm_mode_object_find(dev, out_resp->connector_id,
-				   DRM_MODE_OBJECT_CONNECTOR);
-	if (!obj) {
+	connector = drm_connector_find(dev, out_resp->connector_id);
+	if (!connector) {
 		ret = -ENOENT;
 		goto out;
 	}
-	connector = obj_to_connector(obj);
 
 	props_count = connector->properties.count;
 
@@ -1672,10 +1871,12 @@ int drm_mode_getconnector(struct drm_device *dev, void *data,
 	out_resp->mm_height = connector->display_info.height_mm;
 	out_resp->subpixel = connector->display_info.subpixel_order;
 	out_resp->connection = connector->status;
+	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
 	if (connector->encoder)
 		out_resp->encoder_id = connector->encoder->base.id;
 	else
 		out_resp->encoder_id = 0;
+	drm_modeset_unlock(&dev->mode_config.connection_mutex);
 
 	/*
 	 * This ioctl is called twice, once to determine how much space is
@@ -1742,11 +1943,23 @@ out:
 	return ret;
 }
 
+/**
+ * drm_mode_getencoder - get encoder configuration
+ * @dev: drm device for the ioctl
+ * @data: data pointer for the ioctl
+ * @file_priv: drm file for the ioctl call
+ *
+ * Construct a encoder configuration structure to return to the user.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_getencoder(struct drm_device *dev, void *data,
 			struct drm_file *file_priv)
 {
 	struct drm_mode_get_encoder *enc_resp = data;
-	struct drm_mode_object *obj;
 	struct drm_encoder *encoder;
 	int ret = 0;
 
@@ -1754,13 +1967,11 @@ int drm_mode_getencoder(struct drm_device *dev, void *data,
 		return -EINVAL;
 
 	drm_modeset_lock_all(dev);
-	obj = drm_mode_object_find(dev, enc_resp->encoder_id,
-				   DRM_MODE_OBJECT_ENCODER);
-	if (!obj) {
+	encoder = drm_encoder_find(dev, enc_resp->encoder_id);
+	if (!encoder) {
 		ret = -ENOENT;
 		goto out;
 	}
-	encoder = obj_to_encoder(obj);
 
 	if (encoder->crtc)
 		enc_resp->crtc_id = encoder->crtc->base.id;
@@ -1777,21 +1988,27 @@ out:
 }
 
 /**
- * drm_mode_getplane_res - get plane info
+ * drm_mode_getplane_res - enumerate all plane resources
  * @dev: DRM device
  * @data: ioctl data
  * @file_priv: DRM file info
  *
- * Return an plane count and set of IDs.
+ * Construct a list of plane ids to return to the user.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
  */
 int drm_mode_getplane_res(struct drm_device *dev, void *data,
-			    struct drm_file *file_priv)
+			  struct drm_file *file_priv)
 {
 	struct drm_mode_get_plane_res *plane_resp = data;
 	struct drm_mode_config *config;
 	struct drm_plane *plane;
 	uint32_t __user *plane_ptr;
 	int copied = 0, ret = 0;
+	unsigned num_planes;
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
@@ -1799,15 +2016,28 @@ int drm_mode_getplane_res(struct drm_device *dev, void *data,
 	drm_modeset_lock_all(dev);
 	config = &dev->mode_config;
 
+	if (file_priv->universal_planes)
+		num_planes = config->num_total_plane;
+	else
+		num_planes = config->num_overlay_plane;
+
 	/*
 	 * This ioctl is called twice, once to determine how much space is
 	 * needed, and the 2nd time to fill it.
 	 */
-	if (config->num_plane &&
-	    (plane_resp->count_planes >= config->num_plane)) {
+	if (num_planes &&
+	    (plane_resp->count_planes >= num_planes)) {
 		plane_ptr = (uint32_t __user *)(unsigned long)plane_resp->plane_id_ptr;
 
 		list_for_each_entry(plane, &config->plane_list, head) {
+			/*
+			 * Unless userspace set the 'universal planes'
+			 * capability bit, only advertise overlays.
+			 */
+			if (plane->type != DRM_PLANE_TYPE_OVERLAY &&
+			    !file_priv->universal_planes)
+				continue;
+
 			if (put_user(plane->base.id, plane_ptr + copied)) {
 				ret = -EFAULT;
 				goto out;
@@ -1815,7 +2045,7 @@ int drm_mode_getplane_res(struct drm_device *dev, void *data,
 			copied++;
 		}
 	}
-	plane_resp->count_planes = config->num_plane;
+	plane_resp->count_planes = num_planes;
 
 out:
 	drm_modeset_unlock_all(dev);
@@ -1823,19 +2053,22 @@ out:
 }
 
 /**
- * drm_mode_getplane - get plane info
+ * drm_mode_getplane - get plane configuration
  * @dev: DRM device
  * @data: ioctl data
  * @file_priv: DRM file info
  *
- * Return plane info, including formats supported, gamma size, any
- * current fb, etc.
+ * Construct a plane configuration structure to return to the user.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
  */
 int drm_mode_getplane(struct drm_device *dev, void *data,
-			struct drm_file *file_priv)
+		      struct drm_file *file_priv)
 {
 	struct drm_mode_get_plane *plane_resp = data;
-	struct drm_mode_object *obj;
 	struct drm_plane *plane;
 	uint32_t __user *format_ptr;
 	int ret = 0;
@@ -1844,13 +2077,11 @@ int drm_mode_getplane(struct drm_device *dev, void *data,
 		return -EINVAL;
 
 	drm_modeset_lock_all(dev);
-	obj = drm_mode_object_find(dev, plane_resp->plane_id,
-				   DRM_MODE_OBJECT_PLANE);
-	if (!obj) {
+	plane = drm_plane_find(dev, plane_resp->plane_id);
+	if (!plane) {
 		ret = -ENOENT;
 		goto out;
 	}
-	plane = obj_to_plane(obj);
 
 	if (plane->crtc)
 		plane_resp->crtc_id = plane->crtc->base.id;
@@ -1888,19 +2119,21 @@ out:
 }
 
 /**
- * drm_mode_setplane - set up or tear down an plane
+ * drm_mode_setplane - configure a plane's configuration
  * @dev: DRM device
  * @data: ioctl data*
  * @file_priv: DRM file info
  *
- * Set plane info, including placement, fb, scaling, and other factors.
+ * Set plane configuration, including placement, fb, scaling, and other factors.
  * Or pass a NULL fb to disable.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
  */
 int drm_mode_setplane(struct drm_device *dev, void *data,
-			struct drm_file *file_priv)
+		      struct drm_file *file_priv)
 {
 	struct drm_mode_set_plane *plane_req = data;
-	struct drm_mode_object *obj;
 	struct drm_plane *plane;
 	struct drm_crtc *crtc;
 	struct drm_framebuffer *fb = NULL, *old_fb = NULL;
@@ -1915,35 +2148,42 @@ int drm_mode_setplane(struct drm_device *dev, void *data,
 	 * First, find the plane, crtc, and fb objects.  If not available,
 	 * we don't bother to call the driver.
 	 */
-	obj = drm_mode_object_find(dev, plane_req->plane_id,
-				   DRM_MODE_OBJECT_PLANE);
-	if (!obj) {
+	plane = drm_plane_find(dev, plane_req->plane_id);
+	if (!plane) {
 		DRM_DEBUG_KMS("Unknown plane ID %d\n",
 			      plane_req->plane_id);
 		return -ENOENT;
 	}
-	plane = obj_to_plane(obj);
 
 	/* No fb means shut it down */
 	if (!plane_req->fb_id) {
 		drm_modeset_lock_all(dev);
 		old_fb = plane->fb;
-		plane->funcs->disable_plane(plane);
-		plane->crtc = NULL;
-		plane->fb = NULL;
+		ret = plane->funcs->disable_plane(plane);
+		if (!ret) {
+			plane->crtc = NULL;
+			plane->fb = NULL;
+		} else {
+			old_fb = NULL;
+		}
 		drm_modeset_unlock_all(dev);
 		goto out;
 	}
 
-	obj = drm_mode_object_find(dev, plane_req->crtc_id,
-				   DRM_MODE_OBJECT_CRTC);
-	if (!obj) {
+	crtc = drm_crtc_find(dev, plane_req->crtc_id);
+	if (!crtc) {
 		DRM_DEBUG_KMS("Unknown crtc ID %d\n",
 			      plane_req->crtc_id);
 		ret = -ENOENT;
 		goto out;
 	}
-	crtc = obj_to_crtc(obj);
+
+	/* Check whether this plane is usable on this CRTC */
+	if (!(plane->possible_crtcs & drm_crtc_mask(crtc))) {
+		DRM_DEBUG_KMS("Invalid crtc for plane\n");
+		ret = -EINVAL;
+		goto out;
+	}
 
 	fb = drm_framebuffer_lookup(dev, plane_req->fb_id);
 	if (!fb) {
@@ -1999,16 +2239,18 @@ int drm_mode_setplane(struct drm_device *dev, void *data,
 	}
 
 	drm_modeset_lock_all(dev);
+	old_fb = plane->fb;
 	ret = plane->funcs->update_plane(plane, crtc, fb,
 					 plane_req->crtc_x, plane_req->crtc_y,
 					 plane_req->crtc_w, plane_req->crtc_h,
 					 plane_req->src_x, plane_req->src_y,
 					 plane_req->src_w, plane_req->src_h);
 	if (!ret) {
-		old_fb = plane->fb;
 		plane->crtc = crtc;
 		plane->fb = fb;
 		fb = NULL;
+	} else {
+		old_fb = NULL;
 	}
 	drm_modeset_unlock_all(dev);
 
@@ -2027,6 +2269,9 @@ out:
  *
  * This is a little helper to wrap internal calls to the ->set_config driver
  * interface. The only thing it adds is correct refcounting dance.
+ * 
+ * Returns:
+ * Zero on success, errno on failure.
  */
 int drm_mode_set_config_internal(struct drm_mode_set *set)
 {
@@ -2041,19 +2286,19 @@ int drm_mode_set_config_internal(struct drm_mode_set *set)
 	 * crtcs. Atomic modeset will have saner semantics ...
 	 */
 	list_for_each_entry(tmp, &crtc->dev->mode_config.crtc_list, head)
-		tmp->old_fb = tmp->fb;
+		tmp->old_fb = tmp->primary->fb;
 
 	fb = set->fb;
 
 	ret = crtc->funcs->set_config(set);
 	if (ret == 0) {
-		/* crtc->fb must be updated by ->set_config, enforces this. */
-		WARN_ON(fb != crtc->fb);
+		crtc->primary->crtc = crtc;
+		crtc->primary->fb = fb;
 	}
 
 	list_for_each_entry(tmp, &crtc->dev->mode_config.crtc_list, head) {
-		if (tmp->fb)
-			drm_framebuffer_reference(tmp->fb);
+		if (tmp->primary->fb)
+			drm_framebuffer_reference(tmp->primary->fb);
 		if (tmp->old_fb)
 			drm_framebuffer_unreference(tmp->old_fb);
 	}
@@ -2062,14 +2307,19 @@ int drm_mode_set_config_internal(struct drm_mode_set *set)
 }
 EXPORT_SYMBOL(drm_mode_set_config_internal);
 
-/*
- * Checks that the framebuffer is big enough for the CRTC viewport
- * (x, y, hdisplay, vdisplay)
+/**
+ * drm_crtc_check_viewport - Checks that a framebuffer is big enough for the
+ *     CRTC viewport
+ * @crtc: CRTC that framebuffer will be displayed on
+ * @x: x panning
+ * @y: y panning
+ * @mode: mode that framebuffer will be displayed under
+ * @fb: framebuffer to check size of
  */
-static int drm_crtc_check_viewport(const struct drm_crtc *crtc,
-				   int x, int y,
-				   const struct drm_display_mode *mode,
-				   const struct drm_framebuffer *fb)
+int drm_crtc_check_viewport(const struct drm_crtc *crtc,
+			    int x, int y,
+			    const struct drm_display_mode *mode,
+			    const struct drm_framebuffer *fb)
 
 {
 	int hdisplay, vdisplay;
@@ -2100,6 +2350,7 @@ static int drm_crtc_check_viewport(const struct drm_crtc *crtc,
 
 	return 0;
 }
+EXPORT_SYMBOL(drm_crtc_check_viewport);
 
 /**
  * drm_mode_setcrtc - set CRTC configuration
@@ -2111,7 +2362,7 @@ static int drm_crtc_check_viewport(const struct drm_crtc *crtc,
  *
  * Called by the user via ioctl.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 int drm_mode_setcrtc(struct drm_device *dev, void *data,
@@ -2119,7 +2370,6 @@ int drm_mode_setcrtc(struct drm_device *dev, void *data,
 {
 	struct drm_mode_config *config = &dev->mode_config;
 	struct drm_mode_crtc *crtc_req = data;
-	struct drm_mode_object *obj;
 	struct drm_crtc *crtc;
 	struct drm_connector **connector_set = NULL, *connector;
 	struct drm_framebuffer *fb = NULL;
@@ -2137,26 +2387,24 @@ int drm_mode_setcrtc(struct drm_device *dev, void *data,
 		return -ERANGE;
 
 	drm_modeset_lock_all(dev);
-	obj = drm_mode_object_find(dev, crtc_req->crtc_id,
-				   DRM_MODE_OBJECT_CRTC);
-	if (!obj) {
+	crtc = drm_crtc_find(dev, crtc_req->crtc_id);
+	if (!crtc) {
 		DRM_DEBUG_KMS("Unknown CRTC ID %d\n", crtc_req->crtc_id);
 		ret = -ENOENT;
 		goto out;
 	}
-	crtc = obj_to_crtc(obj);
 	DRM_DEBUG_KMS("[CRTC:%d]\n", crtc->base.id);
 
 	if (crtc_req->mode_valid) {
 		/* If we have a mode we need a framebuffer. */
 		/* If we pass -1, set the mode with the currently bound fb */
 		if (crtc_req->fb_id == -1) {
-			if (!crtc->fb) {
+			if (!crtc->primary->fb) {
 				DRM_DEBUG_KMS("CRTC doesn't have current FB\n");
 				ret = -EINVAL;
 				goto out;
 			}
-			fb = crtc->fb;
+			fb = crtc->primary->fb;
 			/* Make refcounting symmetric with the lookup path. */
 			drm_framebuffer_reference(fb);
 		} else {
@@ -2227,18 +2475,16 @@ int drm_mode_setcrtc(struct drm_device *dev, void *data,
 				goto out;
 			}
 
-			obj = drm_mode_object_find(dev, out_id,
-						   DRM_MODE_OBJECT_CONNECTOR);
-			if (!obj) {
+			connector = drm_connector_find(dev, out_id);
+			if (!connector) {
 				DRM_DEBUG_KMS("Connector id %d unknown\n",
 						out_id);
 				ret = -ENOENT;
 				goto out;
 			}
-			connector = obj_to_connector(obj);
 			DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
 					connector->base.id,
-					drm_get_connector_name(connector));
+					connector->name);
 
 			connector_set[i] = connector;
 		}
@@ -2267,7 +2513,6 @@ static int drm_mode_cursor_common(struct drm_device *dev,
 				  struct drm_mode_cursor2 *req,
 				  struct drm_file *file_priv)
 {
-	struct drm_mode_object *obj;
 	struct drm_crtc *crtc;
 	int ret = 0;
 
@@ -2277,14 +2522,13 @@ static int drm_mode_cursor_common(struct drm_device *dev,
 	if (!req->flags || (~DRM_MODE_CURSOR_FLAGS & req->flags))
 		return -EINVAL;
 
-	obj = drm_mode_object_find(dev, req->crtc_id, DRM_MODE_OBJECT_CRTC);
-	if (!obj) {
+	crtc = drm_crtc_find(dev, req->crtc_id);
+	if (!crtc) {
 		DRM_DEBUG_KMS("Unknown CRTC ID %d\n", req->crtc_id);
 		return -ENOENT;
 	}
-	crtc = obj_to_crtc(obj);
 
-	mutex_lock(&crtc->mutex);
+	drm_modeset_lock(&crtc->mutex, NULL);
 	if (req->flags & DRM_MODE_CURSOR_BO) {
 		if (!crtc->funcs->cursor_set && !crtc->funcs->cursor_set2) {
 			ret = -ENXIO;
@@ -2308,13 +2552,28 @@ static int drm_mode_cursor_common(struct drm_device *dev,
 		}
 	}
 out:
-	mutex_unlock(&crtc->mutex);
+	drm_modeset_unlock(&crtc->mutex);
 
 	return ret;
 
 }
+
+
+/**
+ * drm_mode_cursor_ioctl - set CRTC's cursor configuration
+ * @dev: drm device for the ioctl
+ * @data: data pointer for the ioctl
+ * @file_priv: drm file for the ioctl call
+ *
+ * Set the cursor configuration based on user request.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_cursor_ioctl(struct drm_device *dev,
-			void *data, struct drm_file *file_priv)
+			  void *data, struct drm_file *file_priv)
 {
 	struct drm_mode_cursor *req = data;
 	struct drm_mode_cursor2 new_req;
@@ -2325,6 +2584,21 @@ int drm_mode_cursor_ioctl(struct drm_device *dev,
 	return drm_mode_cursor_common(dev, &new_req, file_priv);
 }
 
+/**
+ * drm_mode_cursor2_ioctl - set CRTC's cursor configuration
+ * @dev: drm device for the ioctl
+ * @data: data pointer for the ioctl
+ * @file_priv: drm file for the ioctl call
+ *
+ * Set the cursor configuration based on user request. This implements the 2nd
+ * version of the cursor ioctl, which allows userspace to additionally specify
+ * the hotspot of the pointer.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_cursor2_ioctl(struct drm_device *dev,
 			   void *data, struct drm_file *file_priv)
 {
@@ -2332,7 +2606,14 @@ int drm_mode_cursor2_ioctl(struct drm_device *dev,
 	return drm_mode_cursor_common(dev, req, file_priv);
 }
 
-/* Original addfb only supported RGB formats, so figure out which one */
+/**
+ * drm_mode_legacy_fb_format - compute drm fourcc code from legacy description
+ * @bpp: bits per pixels
+ * @depth: bit depth per pixel
+ *
+ * Computes a drm fourcc pixel format code for the given @bpp/@depth values.
+ * Useful in fbdev emulation code, since that deals in those values.
+ */
 uint32_t drm_mode_legacy_fb_format(uint32_t bpp, uint32_t depth)
 {
 	uint32_t fmt;
@@ -2374,11 +2655,12 @@ EXPORT_SYMBOL(drm_mode_legacy_fb_format);
  * @data: data pointer for the ioctl
  * @file_priv: drm file for the ioctl call
  *
- * Add a new FB to the specified CRTC, given a user request.
+ * Add a new FB to the specified CRTC, given a user request. This is the
+ * original addfb ioclt which only supported RGB formats.
  *
  * Called by the user via ioctl.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 int drm_mode_addfb(struct drm_device *dev,
@@ -2551,11 +2833,13 @@ static int framebuffer_check(const struct drm_mode_fb_cmd2 *r)
  * @data: data pointer for the ioctl
  * @file_priv: drm file for the ioctl call
  *
- * Add a new FB to the specified CRTC, given a user request with format.
+ * Add a new FB to the specified CRTC, given a user request with format. This is
+ * the 2nd version of the addfb ioctl, which supports multi-planar framebuffers
+ * and uses fourcc codes as pixel format specifiers.
  *
  * Called by the user via ioctl.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 int drm_mode_addfb2(struct drm_device *dev,
@@ -2615,7 +2899,7 @@ int drm_mode_addfb2(struct drm_device *dev,
  *
  * Called by the user via ioctl.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 int drm_mode_rmfb(struct drm_device *dev,
@@ -2669,7 +2953,7 @@ fail_lookup:
  *
  * Called by the user via ioctl.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 int drm_mode_getfb(struct drm_device *dev,
@@ -2692,7 +2976,8 @@ int drm_mode_getfb(struct drm_device *dev,
 	r->bpp = fb->bits_per_pixel;
 	r->pitch = fb->pitches[0];
 	if (fb->funcs->create_handle) {
-		if (file_priv->is_master || capable(CAP_SYS_ADMIN)) {
+		if (file_priv->is_master || capable(CAP_SYS_ADMIN) ||
+		    drm_is_control_client(file_priv)) {
 			ret = fb->funcs->create_handle(fb, file_priv,
 						       &r->handle);
 		} else {
@@ -2713,6 +2998,25 @@ int drm_mode_getfb(struct drm_device *dev,
 	return ret;
 }
 
+/**
+ * drm_mode_dirtyfb_ioctl - flush frontbuffer rendering on an FB
+ * @dev: drm device for the ioctl
+ * @data: data pointer for the ioctl
+ * @file_priv: drm file for the ioctl call
+ *
+ * Lookup the FB and flush out the damaged area supplied by userspace as a clip
+ * rectangle list. Generic userspace which does frontbuffer rendering must call
+ * this ioctl to flush out the changes on manual-update display outputs, e.g.
+ * usb display-link, mipi manual update panels or edp panel self refresh modes.
+ *
+ * Modesetting drivers which always update the frontbuffer do not need to
+ * implement the corresponding ->dirty framebuffer callback.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_dirtyfb_ioctl(struct drm_device *dev,
 			   void *data, struct drm_file *file_priv)
 {
@@ -2767,10 +3071,8 @@ int drm_mode_dirtyfb_ioctl(struct drm_device *dev,
 	}
 
 	if (fb->funcs->dirty) {
-		drm_modeset_lock_all(dev);
 		ret = fb->funcs->dirty(fb, file_priv, flags, r->color,
 				       clips, num_clips);
-		drm_modeset_unlock_all(dev);
 	} else {
 		ret = -ENOSYS;
 	}
@@ -2792,7 +3094,7 @@ out_err1:
  *
  * Called by the user via ioctl.
  *
- * RETURNS:
+ * Returns:
  * Zero on success, errno on failure.
  */
 void drm_fb_release(struct drm_file *priv)
@@ -2816,6 +3118,20 @@ void drm_fb_release(struct drm_file *priv)
 	mutex_unlock(&priv->fbs_lock);
 }
 
+/**
+ * drm_property_create - create a new property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @num_values: number of pre-defined values
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy.
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
 struct drm_property *drm_property_create(struct drm_device *dev, int flags,
 					 const char *name, int num_values)
 {
@@ -2826,6 +3142,8 @@ struct drm_property *drm_property_create(struct drm_device *dev, int flags,
 	if (!property)
 		return NULL;
 
+	property->dev = dev;
+
 	if (num_values) {
 		property->values = kzalloc(sizeof(uint64_t)*num_values, GFP_KERNEL);
 		if (!property->values)
@@ -2846,6 +3164,9 @@ struct drm_property *drm_property_create(struct drm_device *dev, int flags,
 	}
 
 	list_add_tail(&property->head, &dev->mode_config.property_list);
+
+	WARN_ON(!drm_property_type_valid(property));
+
 	return property;
 fail:
 	kfree(property->values);
@@ -2854,6 +3175,24 @@ fail:
 }
 EXPORT_SYMBOL(drm_property_create);
 
+/**
+ * drm_property_create - create a new enumeration property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @props: enumeration lists with property values
+ * @num_values: number of pre-defined values
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy.
+ *
+ * Userspace is only allowed to set one of the predefined values for enumeration
+ * properties.
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
 struct drm_property *drm_property_create_enum(struct drm_device *dev, int flags,
 					 const char *name,
 					 const struct drm_prop_enum_list *props,
@@ -2882,6 +3221,24 @@ struct drm_property *drm_property_create_enum(struct drm_device *dev, int flags,
 }
 EXPORT_SYMBOL(drm_property_create_enum);
 
+/**
+ * drm_property_create - create a new bitmask property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @props: enumeration lists with property bitflags
+ * @num_values: number of pre-defined values
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy.
+ *
+ * Compared to plain enumeration properties userspace is allowed to set any
+ * or'ed together combination of the predefined property bitflag values
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
 struct drm_property *drm_property_create_bitmask(struct drm_device *dev,
 					 int flags, const char *name,
 					 const struct drm_prop_enum_list *props,
@@ -2910,14 +3267,12 @@ struct drm_property *drm_property_create_bitmask(struct drm_device *dev,
 }
 EXPORT_SYMBOL(drm_property_create_bitmask);
 
-struct drm_property *drm_property_create_range(struct drm_device *dev, int flags,
-					 const char *name,
+static struct drm_property *property_create_range(struct drm_device *dev,
+					 int flags, const char *name,
 					 uint64_t min, uint64_t max)
 {
 	struct drm_property *property;
 
-	flags |= DRM_MODE_PROP_RANGE;
-
 	property = drm_property_create(dev, flags, name, 2);
 	if (!property)
 		return NULL;
@@ -2927,21 +3282,90 @@ struct drm_property *drm_property_create_range(struct drm_device *dev, int flags
 
 	return property;
 }
+
+/**
+ * drm_property_create - create a new ranged property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @min: minimum value of the property
+ * @max: maximum value of the property
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy.
+ *
+ * Userspace is allowed to set any interger value in the (min, max) range
+ * inclusive.
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
+struct drm_property *drm_property_create_range(struct drm_device *dev, int flags,
+					 const char *name,
+					 uint64_t min, uint64_t max)
+{
+	return property_create_range(dev, DRM_MODE_PROP_RANGE | flags,
+			name, min, max);
+}
 EXPORT_SYMBOL(drm_property_create_range);
 
+struct drm_property *drm_property_create_signed_range(struct drm_device *dev,
+					 int flags, const char *name,
+					 int64_t min, int64_t max)
+{
+	return property_create_range(dev, DRM_MODE_PROP_SIGNED_RANGE | flags,
+			name, I642U64(min), I642U64(max));
+}
+EXPORT_SYMBOL(drm_property_create_signed_range);
+
+struct drm_property *drm_property_create_object(struct drm_device *dev,
+					 int flags, const char *name, uint32_t type)
+{
+	struct drm_property *property;
+
+	flags |= DRM_MODE_PROP_OBJECT;
+
+	property = drm_property_create(dev, flags, name, 1);
+	if (!property)
+		return NULL;
+
+	property->values[0] = type;
+
+	return property;
+}
+EXPORT_SYMBOL(drm_property_create_object);
+
+/**
+ * drm_property_add_enum - add a possible value to an enumeration property
+ * @property: enumeration property to change
+ * @index: index of the new enumeration
+ * @value: value of the new enumeration
+ * @name: symbolic name of the new enumeration
+ *
+ * This functions adds enumerations to a property.
+ *
+ * It's use is deprecated, drivers should use one of the more specific helpers
+ * to directly create the property with all enumerations already attached.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
 int drm_property_add_enum(struct drm_property *property, int index,
 			  uint64_t value, const char *name)
 {
 	struct drm_property_enum *prop_enum;
 
-	if (!(property->flags & (DRM_MODE_PROP_ENUM | DRM_MODE_PROP_BITMASK)))
+	if (!(drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
+			drm_property_type_is(property, DRM_MODE_PROP_BITMASK)))
 		return -EINVAL;
 
 	/*
 	 * Bitmask enum properties have the additional constraint of values
 	 * from 0 to 63
 	 */
-	if ((property->flags & DRM_MODE_PROP_BITMASK) && (value > 63))
+	if (drm_property_type_is(property, DRM_MODE_PROP_BITMASK) &&
+			(value > 63))
 		return -EINVAL;
 
 	if (!list_empty(&property->enum_blob_list)) {
@@ -2968,6 +3392,14 @@ int drm_property_add_enum(struct drm_property *property, int index,
 }
 EXPORT_SYMBOL(drm_property_add_enum);
 
+/**
+ * drm_property_destroy - destroy a drm property
+ * @dev: drm device
+ * @property: property to destry
+ *
+ * This function frees a property including any attached resources like
+ * enumeration values.
+ */
 void drm_property_destroy(struct drm_device *dev, struct drm_property *property)
 {
 	struct drm_property_enum *prop_enum, *pt;
@@ -2985,6 +3417,16 @@ void drm_property_destroy(struct drm_device *dev, struct drm_property *property)
 }
 EXPORT_SYMBOL(drm_property_destroy);
 
+/**
+ * drm_object_attach_property - attach a property to a modeset object
+ * @obj: drm modeset object
+ * @property: property to attach
+ * @init_val: initial value of the property
+ *
+ * This attaches the given property to the modeset object with the given initial
+ * value. Currently this function cannot fail since the properties are stored in
+ * a statically sized array.
+ */
 void drm_object_attach_property(struct drm_mode_object *obj,
 				struct drm_property *property,
 				uint64_t init_val)
@@ -3005,6 +3447,19 @@ void drm_object_attach_property(struct drm_mode_object *obj,
 }
 EXPORT_SYMBOL(drm_object_attach_property);
 
+/**
+ * drm_object_property_set_value - set the value of a property
+ * @obj: drm mode object to set property value for
+ * @property: property to set
+ * @val: value the property should be set to
+ *
+ * This functions sets a given property on a given object. This function only
+ * changes the software state of the property, it does not call into the
+ * driver's ->set_property callback.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
 int drm_object_property_set_value(struct drm_mode_object *obj,
 				  struct drm_property *property, uint64_t val)
 {
@@ -3021,6 +3476,20 @@ int drm_object_property_set_value(struct drm_mode_object *obj,
 }
 EXPORT_SYMBOL(drm_object_property_set_value);
 
+/**
+ * drm_object_property_get_value - retrieve the value of a property
+ * @obj: drm mode object to get property value from
+ * @property: property to retrieve
+ * @val: storage for the property value
+ *
+ * This function retrieves the softare state of the given property for the given
+ * property. Since there is no driver callback to retrieve the current property
+ * value this might be out of sync with the hardware, depending upon the driver
+ * and property.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
 int drm_object_property_get_value(struct drm_mode_object *obj,
 				  struct drm_property *property, uint64_t *val)
 {
@@ -3037,10 +3506,22 @@ int drm_object_property_get_value(struct drm_mode_object *obj,
 }
 EXPORT_SYMBOL(drm_object_property_get_value);
 
+/**
+ * drm_mode_getproperty_ioctl - get the current value of a connector's property
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * This function retrieves the current value for an connectors's property.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_getproperty_ioctl(struct drm_device *dev,
 			       void *data, struct drm_file *file_priv)
 {
-	struct drm_mode_object *obj;
 	struct drm_mode_get_property *out_resp = data;
 	struct drm_property *property;
 	int enum_count = 0;
@@ -3059,17 +3540,17 @@ int drm_mode_getproperty_ioctl(struct drm_device *dev,
 		return -EINVAL;
 
 	drm_modeset_lock_all(dev);
-	obj = drm_mode_object_find(dev, out_resp->prop_id, DRM_MODE_OBJECT_PROPERTY);
-	if (!obj) {
+	property = drm_property_find(dev, out_resp->prop_id);
+	if (!property) {
 		ret = -ENOENT;
 		goto done;
 	}
-	property = obj_to_property(obj);
 
-	if (property->flags & (DRM_MODE_PROP_ENUM | DRM_MODE_PROP_BITMASK)) {
+	if (drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
+			drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
 		list_for_each_entry(prop_enum, &property->enum_blob_list, head)
 			enum_count++;
-	} else if (property->flags & DRM_MODE_PROP_BLOB) {
+	} else if (drm_property_type_is(property, DRM_MODE_PROP_BLOB)) {
 		list_for_each_entry(prop_blob, &property->enum_blob_list, head)
 			blob_count++;
 	}
@@ -3091,7 +3572,8 @@ int drm_mode_getproperty_ioctl(struct drm_device *dev,
 	}
 	out_resp->count_values = value_count;
 
-	if (property->flags & (DRM_MODE_PROP_ENUM | DRM_MODE_PROP_BITMASK)) {
+	if (drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
+			drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
 		if ((out_resp->count_enum_blobs >= enum_count) && enum_count) {
 			copied = 0;
 			enum_ptr = (struct drm_mode_property_enum __user *)(unsigned long)out_resp->enum_blob_ptr;
@@ -3113,7 +3595,7 @@ int drm_mode_getproperty_ioctl(struct drm_device *dev,
 		out_resp->count_enum_blobs = enum_count;
 	}
 
-	if (property->flags & DRM_MODE_PROP_BLOB) {
+	if (drm_property_type_is(property, DRM_MODE_PROP_BLOB)) {
 		if ((out_resp->count_enum_blobs >= blob_count) && blob_count) {
 			copied = 0;
 			blob_id_ptr = (uint32_t __user *)(unsigned long)out_resp->enum_blob_ptr;
@@ -3175,10 +3657,23 @@ static void drm_property_destroy_blob(struct drm_device *dev,
 	kfree(blob);
 }
 
+/**
+ * drm_mode_getblob_ioctl - get the contents of a blob property value
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * This function retrieves the contents of a blob property. The value stored in
+ * an object's blob property is just a normal modeset object id.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_getblob_ioctl(struct drm_device *dev,
 			   void *data, struct drm_file *file_priv)
 {
-	struct drm_mode_object *obj;
 	struct drm_mode_get_blob *out_resp = data;
 	struct drm_property_blob *blob;
 	int ret = 0;
@@ -3188,12 +3683,11 @@ int drm_mode_getblob_ioctl(struct drm_device *dev,
 		return -EINVAL;
 
 	drm_modeset_lock_all(dev);
-	obj = drm_mode_object_find(dev, out_resp->blob_id, DRM_MODE_OBJECT_BLOB);
-	if (!obj) {
+	blob = drm_property_blob_find(dev, out_resp->blob_id);
+	if (!blob) {
 		ret = -ENOENT;
 		goto done;
 	}
-	blob = obj_to_blob(obj);
 
 	if (out_resp->length == blob->length) {
 		blob_ptr = (void __user *)(unsigned long)out_resp->data;
@@ -3209,6 +3703,17 @@ done:
 	return ret;
 }
 
+/**
+ * drm_mode_connector_update_edid_property - update the edid property of a connector
+ * @connector: drm connector
+ * @edid: new value of the edid property
+ *
+ * This function creates a new blob modeset object and assigns its id to the
+ * connector's edid property.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_connector_update_edid_property(struct drm_connector *connector,
 					    struct edid *edid)
 {
@@ -3244,19 +3749,40 @@ static bool drm_property_change_is_valid(struct drm_property *property,
 {
 	if (property->flags & DRM_MODE_PROP_IMMUTABLE)
 		return false;
-	if (property->flags & DRM_MODE_PROP_RANGE) {
+
+	if (drm_property_type_is(property, DRM_MODE_PROP_RANGE)) {
 		if (value < property->values[0] || value > property->values[1])
 			return false;
 		return true;
-	} else if (property->flags & DRM_MODE_PROP_BITMASK) {
+	} else if (drm_property_type_is(property, DRM_MODE_PROP_SIGNED_RANGE)) {
+		int64_t svalue = U642I64(value);
+		if (svalue < U642I64(property->values[0]) ||
+				svalue > U642I64(property->values[1]))
+			return false;
+		return true;
+	} else if (drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
 		int i;
 		uint64_t valid_mask = 0;
 		for (i = 0; i < property->num_values; i++)
 			valid_mask |= (1ULL << property->values[i]);
 		return !(value & ~valid_mask);
-	} else if (property->flags & DRM_MODE_PROP_BLOB) {
+	} else if (drm_property_type_is(property, DRM_MODE_PROP_BLOB)) {
 		/* Only the driver knows */
 		return true;
+	} else if (drm_property_type_is(property, DRM_MODE_PROP_OBJECT)) {
+		struct drm_mode_object *obj;
+		/* a zero value for an object property translates to null: */
+		if (value == 0)
+			return true;
+		/*
+		 * NOTE: use _object_find() directly to bypass restriction on
+		 * looking up refcnt'd objects (ie. fb's).  For a refcnt'd
+		 * object this could race against object finalization, so it
+		 * simply tells us that the object *was* valid.  Which is good
+		 * enough.
+		 */
+		obj = _object_find(property->dev, value, property->values[0]);
+		return obj != NULL;
 	} else {
 		int i;
 		for (i = 0; i < property->num_values; i++)
@@ -3266,6 +3792,20 @@ static bool drm_property_change_is_valid(struct drm_property *property,
 	}
 }
 
+/**
+ * drm_mode_connector_property_set_ioctl - set the current value of a connector property
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * This function sets the current value for a connectors's property. It also
+ * calls into a driver's ->set_property callback to update the hardware state
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_connector_property_set_ioctl(struct drm_device *dev,
 				       void *data, struct drm_file *file_priv)
 {
@@ -3332,6 +3872,21 @@ static int drm_mode_plane_set_obj_prop(struct drm_mode_object *obj,
 	return ret;
 }
 
+/**
+ * drm_mode_getproperty_ioctl - get the current value of a object's property
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * This function retrieves the current value for an object's property. Compared
+ * to the connector specific ioctl this one is extended to also work on crtc and
+ * plane objects.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_obj_get_properties_ioctl(struct drm_device *dev, void *data,
 				      struct drm_file *file_priv)
 {
@@ -3388,6 +3943,22 @@ out:
 	return ret;
 }
 
+/**
+ * drm_mode_obj_set_property_ioctl - set the current value of an object's property
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * This function sets the current value for an object's property. It also calls
+ * into a driver's ->set_property callback to update the hardware state.
+ * Compared to the connector specific ioctl this one is extended to also work on
+ * crtc and plane objects.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_obj_set_property_ioctl(struct drm_device *dev, void *data,
 				    struct drm_file *file_priv)
 {
@@ -3447,6 +4018,18 @@ out:
 	return ret;
 }
 
+/**
+ * drm_mode_connector_attach_encoder - attach a connector to an encoder
+ * @connector: connector to attach
+ * @encoder: encoder to attach @connector to
+ *
+ * This function links up a connector to an encoder. Note that the routing
+ * restrictions between encoders and crtcs are exposed to userspace through the
+ * possible_clones and possible_crtcs bitmasks.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_connector_attach_encoder(struct drm_connector *connector,
 				      struct drm_encoder *encoder)
 {
@@ -3462,23 +4045,20 @@ int drm_mode_connector_attach_encoder(struct drm_connector *connector,
 }
 EXPORT_SYMBOL(drm_mode_connector_attach_encoder);
 
-void drm_mode_connector_detach_encoder(struct drm_connector *connector,
-				    struct drm_encoder *encoder)
-{
-	int i;
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == encoder->base.id) {
-			connector->encoder_ids[i] = 0;
-			if (connector->encoder == encoder)
-				connector->encoder = NULL;
-			break;
-		}
-	}
-}
-EXPORT_SYMBOL(drm_mode_connector_detach_encoder);
-
+/**
+ * drm_mode_crtc_set_gamma_size - set the gamma table size
+ * @crtc: CRTC to set the gamma table size for
+ * @gamma_size: size of the gamma table
+ *
+ * Drivers which support gamma tables should set this to the supported gamma
+ * table size when initializing the CRTC. Currently the drm core only supports a
+ * fixed gamma table size.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_crtc_set_gamma_size(struct drm_crtc *crtc,
-				  int gamma_size)
+				 int gamma_size)
 {
 	crtc->gamma_size = gamma_size;
 
@@ -3492,11 +4072,24 @@ int drm_mode_crtc_set_gamma_size(struct drm_crtc *crtc,
 }
 EXPORT_SYMBOL(drm_mode_crtc_set_gamma_size);
 
+/**
+ * drm_mode_gamma_set_ioctl - set the gamma table
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Set the gamma table of a CRTC to the one passed in by the user. Userspace can
+ * inquire the required gamma table size through drm_mode_gamma_get_ioctl.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_gamma_set_ioctl(struct drm_device *dev,
 			     void *data, struct drm_file *file_priv)
 {
 	struct drm_mode_crtc_lut *crtc_lut = data;
-	struct drm_mode_object *obj;
 	struct drm_crtc *crtc;
 	void *r_base, *g_base, *b_base;
 	int size;
@@ -3506,12 +4099,11 @@ int drm_mode_gamma_set_ioctl(struct drm_device *dev,
 		return -EINVAL;
 
 	drm_modeset_lock_all(dev);
-	obj = drm_mode_object_find(dev, crtc_lut->crtc_id, DRM_MODE_OBJECT_CRTC);
-	if (!obj) {
+	crtc = drm_crtc_find(dev, crtc_lut->crtc_id);
+	if (!crtc) {
 		ret = -ENOENT;
 		goto out;
 	}
-	crtc = obj_to_crtc(obj);
 
 	if (crtc->funcs->gamma_set == NULL) {
 		ret = -ENOSYS;
@@ -3551,11 +4143,25 @@ out:
 
 }
 
+/**
+ * drm_mode_gamma_get_ioctl - get the gamma table
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Copy the current gamma table into the storage provided. This also provides
+ * the gamma table size the driver expects, which can be used to size the
+ * allocated storage.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_gamma_get_ioctl(struct drm_device *dev,
 			     void *data, struct drm_file *file_priv)
 {
 	struct drm_mode_crtc_lut *crtc_lut = data;
-	struct drm_mode_object *obj;
 	struct drm_crtc *crtc;
 	void *r_base, *g_base, *b_base;
 	int size;
@@ -3565,12 +4171,11 @@ int drm_mode_gamma_get_ioctl(struct drm_device *dev,
 		return -EINVAL;
 
 	drm_modeset_lock_all(dev);
-	obj = drm_mode_object_find(dev, crtc_lut->crtc_id, DRM_MODE_OBJECT_CRTC);
-	if (!obj) {
+	crtc = drm_crtc_find(dev, crtc_lut->crtc_id);
+	if (!crtc) {
 		ret = -ENOENT;
 		goto out;
 	}
-	crtc = obj_to_crtc(obj);
 
 	/* memcpy into gamma store */
 	if (crtc_lut->gamma_size != crtc->gamma_size) {
@@ -3601,11 +4206,28 @@ out:
 	return ret;
 }
 
+/**
+ * drm_mode_page_flip_ioctl - schedule an asynchronous fb update
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * This schedules an asynchronous update on a given CRTC, called page flip.
+ * Optionally a drm event is generated to signal the completion of the event.
+ * Generic drivers cannot assume that a pageflip with changed framebuffer
+ * properties (including driver specific metadata like tiling layout) will work,
+ * but some drivers support e.g. pixel format changes through the pageflip
+ * ioctl.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_page_flip_ioctl(struct drm_device *dev,
 			     void *data, struct drm_file *file_priv)
 {
 	struct drm_mode_crtc_page_flip *page_flip = data;
-	struct drm_mode_object *obj;
 	struct drm_crtc *crtc;
 	struct drm_framebuffer *fb = NULL, *old_fb = NULL;
 	struct drm_pending_vblank_event *e = NULL;
@@ -3619,13 +4241,12 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 	if ((page_flip->flags & DRM_MODE_PAGE_FLIP_ASYNC) && !dev->mode_config.async_page_flip)
 		return -EINVAL;
 
-	obj = drm_mode_object_find(dev, page_flip->crtc_id, DRM_MODE_OBJECT_CRTC);
-	if (!obj)
+	crtc = drm_crtc_find(dev, page_flip->crtc_id);
+	if (!crtc)
 		return -ENOENT;
-	crtc = obj_to_crtc(obj);
 
-	mutex_lock(&crtc->mutex);
-	if (crtc->fb == NULL) {
+	drm_modeset_lock(&crtc->mutex, NULL);
+	if (crtc->primary->fb == NULL) {
 		/* The framebuffer is currently unbound, presumably
 		 * due to a hotplug event, that userspace has not
 		 * yet discovered.
@@ -3647,7 +4268,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 	if (ret)
 		goto out;
 
-	if (crtc->fb->pixel_format != fb->pixel_format) {
+	if (crtc->primary->fb->pixel_format != fb->pixel_format) {
 		DRM_DEBUG_KMS("Page flip is not allowed to change frame buffer format.\n");
 		ret = -EINVAL;
 		goto out;
@@ -3680,7 +4301,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 			(void (*) (struct drm_pending_event *)) kfree;
 	}
 
-	old_fb = crtc->fb;
+	old_fb = crtc->primary->fb;
 	ret = crtc->funcs->page_flip(crtc, fb, e, page_flip->flags);
 	if (ret) {
 		if (page_flip->flags & DRM_MODE_PAGE_FLIP_EVENT) {
@@ -3698,7 +4319,7 @@ int drm_mode_page_flip_ioctl(struct drm_device *dev,
 		 * Failing to do so will screw with the reference counting
 		 * on framebuffers.
 		 */
-		WARN_ON(crtc->fb != fb);
+		WARN_ON(crtc->primary->fb != fb);
 		/* Unref only the old framebuffer. */
 		fb = NULL;
 	}
@@ -3708,11 +4329,19 @@ out:
 		drm_framebuffer_unreference(fb);
 	if (old_fb)
 		drm_framebuffer_unreference(old_fb);
-	mutex_unlock(&crtc->mutex);
+	drm_modeset_unlock(&crtc->mutex);
 
 	return ret;
 }
 
+/**
+ * drm_mode_config_reset - call ->reset callbacks
+ * @dev: drm device
+ *
+ * This functions calls all the crtc's, encoder's and connector's ->reset
+ * callback. Drivers can use this in e.g. their driver load or resume code to
+ * reset hardware and software state.
+ */
 void drm_mode_config_reset(struct drm_device *dev)
 {
 	struct drm_crtc *crtc;
@@ -3736,16 +4365,66 @@ void drm_mode_config_reset(struct drm_device *dev)
 }
 EXPORT_SYMBOL(drm_mode_config_reset);
 
+/**
+ * drm_mode_create_dumb_ioctl - create a dumb backing storage buffer
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * This creates a new dumb buffer in the driver's backing storage manager (GEM,
+ * TTM or something else entirely) and returns the resulting buffer handle. This
+ * handle can then be wrapped up into a framebuffer modeset object.
+ *
+ * Note that userspace is not allowed to use such objects for render
+ * acceleration - drivers must create their own private ioctls for such a use
+ * case.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_create_dumb_ioctl(struct drm_device *dev,
 			       void *data, struct drm_file *file_priv)
 {
 	struct drm_mode_create_dumb *args = data;
+	u32 cpp, stride, size;
 
 	if (!dev->driver->dumb_create)
 		return -ENOSYS;
+	if (!args->width || !args->height || !args->bpp)
+		return -EINVAL;
+
+	/* overflow checks for 32bit size calculations */
+	cpp = DIV_ROUND_UP(args->bpp, 8);
+	if (cpp > 0xffffffffU / args->width)
+		return -EINVAL;
+	stride = cpp * args->width;
+	if (args->height > 0xffffffffU / stride)
+		return -EINVAL;
+
+	/* test for wrap-around */
+	size = args->height * stride;
+	if (PAGE_ALIGN(size) == 0)
+		return -EINVAL;
+
 	return dev->driver->dumb_create(file_priv, dev, args);
 }
 
+/**
+ * drm_mode_mmap_dumb_ioctl - create an mmap offset for a dumb backing storage buffer
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Allocate an offset in the drm device node's address space to be able to
+ * memory map a dumb buffer.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_mmap_dumb_ioctl(struct drm_device *dev,
 			     void *data, struct drm_file *file_priv)
 {
@@ -3758,6 +4437,21 @@ int drm_mode_mmap_dumb_ioctl(struct drm_device *dev,
 	return dev->driver->dumb_map_offset(file_priv, dev, args->handle, &args->offset);
 }
 
+/**
+ * drm_mode_destroy_dumb_ioctl - destroy a dumb backing strage buffer
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * This destroys the userspace handle for the given dumb backing storage buffer.
+ * Since buffer objects must be reference counted in the kernel a buffer object
+ * won't be immediately freed if a framebuffer modeset object still uses it.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, errno on failure.
+ */
 int drm_mode_destroy_dumb_ioctl(struct drm_device *dev,
 				void *data, struct drm_file *file_priv)
 {
@@ -3769,9 +4463,14 @@ int drm_mode_destroy_dumb_ioctl(struct drm_device *dev,
 	return dev->driver->dumb_destroy(file_priv, dev, args->handle);
 }
 
-/*
- * Just need to support RGB formats here for compat with code that doesn't
- * use pixel formats directly yet.
+/**
+ * drm_fb_get_bpp_depth - get the bpp/depth values for format
+ * @format: pixel format (DRM_FORMAT_*)
+ * @depth: storage for the depth value
+ * @bpp: storage for the bpp value
+ *
+ * This only supports RGB formats here for compat with code that doesn't use
+ * pixel formats directly yet.
  */
 void drm_fb_get_bpp_depth(uint32_t format, unsigned int *depth,
 			  int *bpp)
@@ -3843,7 +4542,7 @@ EXPORT_SYMBOL(drm_fb_get_bpp_depth);
  * drm_format_num_planes - get the number of planes for format
  * @format: pixel format (DRM_FORMAT_*)
  *
- * RETURNS:
+ * Returns:
  * The number of planes used by the specified pixel format.
  */
 int drm_format_num_planes(uint32_t format)
@@ -3878,7 +4577,7 @@ EXPORT_SYMBOL(drm_format_num_planes);
  * @format: pixel format (DRM_FORMAT_*)
  * @plane: plane index
  *
- * RETURNS:
+ * Returns:
  * The bytes per pixel value for the specified plane.
  */
 int drm_format_plane_cpp(uint32_t format, int plane)
@@ -3924,7 +4623,7 @@ EXPORT_SYMBOL(drm_format_plane_cpp);
  * drm_format_horz_chroma_subsampling - get the horizontal chroma subsampling factor
  * @format: pixel format (DRM_FORMAT_*)
  *
- * RETURNS:
+ * Returns:
  * The horizontal chroma subsampling factor for the
  * specified pixel format.
  */
@@ -3959,7 +4658,7 @@ EXPORT_SYMBOL(drm_format_horz_chroma_subsampling);
  * drm_format_vert_chroma_subsampling - get the vertical chroma subsampling factor
  * @format: pixel format (DRM_FORMAT_*)
  *
- * RETURNS:
+ * Returns:
  * The vertical chroma subsampling factor for the
  * specified pixel format.
  */
@@ -3995,6 +4694,7 @@ EXPORT_SYMBOL(drm_format_vert_chroma_subsampling);
 void drm_mode_config_init(struct drm_device *dev)
 {
 	mutex_init(&dev->mode_config.mutex);
+	drm_modeset_lock_init(&dev->mode_config.connection_mutex);
 	mutex_init(&dev->mode_config.idr_mutex);
 	mutex_init(&dev->mode_config.fb_lock);
 	INIT_LIST_HEAD(&dev->mode_config.fb_list);
@@ -4009,6 +4709,7 @@ void drm_mode_config_init(struct drm_device *dev)
 
 	drm_modeset_lock_all(dev);
 	drm_mode_create_standard_connector_properties(dev);
+	drm_mode_create_standard_plane_properties(dev);
 	drm_modeset_unlock_all(dev);
 
 	/* Just to be sure */
@@ -4016,6 +4717,8 @@ void drm_mode_config_init(struct drm_device *dev)
 	dev->mode_config.num_connector = 0;
 	dev->mode_config.num_crtc = 0;
 	dev->mode_config.num_encoder = 0;
+	dev->mode_config.num_overlay_plane = 0;
+	dev->mode_config.num_total_plane = 0;
 }
 EXPORT_SYMBOL(drm_mode_config_init);
 
@@ -4091,5 +4794,6 @@ void drm_mode_config_cleanup(struct drm_device *dev)
 	}
 
 	idr_destroy(&dev->mode_config.crtc_idr);
+	drm_modeset_lock_fini(&dev->mode_config.connection_mutex);
 }
 EXPORT_SYMBOL(drm_mode_config_cleanup);
diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c
index 01361aba033..78b37f3febd 100644
--- a/drivers/gpu/drm/drm_crtc_helper.c
+++ b/drivers/gpu/drm/drm_crtc_helper.c
@@ -29,6 +29,7 @@
  *      Jesse Barnes <jesse.barnes@intel.com>
  */
 
+#include <linux/kernel.h>
 #include <linux/export.h>
 #include <linux/moduleparam.h>
 
@@ -72,165 +73,31 @@ void drm_helper_move_panel_connectors_to_head(struct drm_device *dev)
 }
 EXPORT_SYMBOL(drm_helper_move_panel_connectors_to_head);
 
-static bool drm_kms_helper_poll = true;
-module_param_named(poll, drm_kms_helper_poll, bool, 0600);
-
-static void drm_mode_validate_flag(struct drm_connector *connector,
-				   int flags)
-{
-	struct drm_display_mode *mode;
-
-	if (flags == (DRM_MODE_FLAG_DBLSCAN | DRM_MODE_FLAG_INTERLACE |
-		      DRM_MODE_FLAG_3D_MASK))
-		return;
-
-	list_for_each_entry(mode, &connector->modes, head) {
-		if ((mode->flags & DRM_MODE_FLAG_INTERLACE) &&
-				!(flags & DRM_MODE_FLAG_INTERLACE))
-			mode->status = MODE_NO_INTERLACE;
-		if ((mode->flags & DRM_MODE_FLAG_DBLSCAN) &&
-				!(flags & DRM_MODE_FLAG_DBLSCAN))
-			mode->status = MODE_NO_DBLESCAN;
-		if ((mode->flags & DRM_MODE_FLAG_3D_MASK) &&
-				!(flags & DRM_MODE_FLAG_3D_MASK))
-			mode->status = MODE_NO_STEREO;
-	}
-
-	return;
-}
-
-/**
- * drm_helper_probe_single_connector_modes - get complete set of display modes
- * @connector: connector to probe
- * @maxX: max width for modes
- * @maxY: max height for modes
- *
- * LOCKING:
- * Caller must hold mode config lock.
- *
- * Based on the helper callbacks implemented by @connector try to detect all
- * valid modes.  Modes will first be added to the connector's probed_modes list,
- * then culled (based on validity and the @maxX, @maxY parameters) and put into
- * the normal modes list.
- *
- * Intended to be use as a generic implementation of the ->fill_modes()
- * @connector vfunc for drivers that use the crtc helpers for output mode
- * filtering and detection.
- *
- * RETURNS:
- * Number of modes found on @connector.
- */
-int drm_helper_probe_single_connector_modes(struct drm_connector *connector,
-					    uint32_t maxX, uint32_t maxY)
-{
-	struct drm_device *dev = connector->dev;
-	struct drm_display_mode *mode;
-	struct drm_connector_helper_funcs *connector_funcs =
-		connector->helper_private;
-	int count = 0;
-	int mode_flags = 0;
-	bool verbose_prune = true;
-
-	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n", connector->base.id,
-			drm_get_connector_name(connector));
-	/* set all modes to the unverified state */
-	list_for_each_entry(mode, &connector->modes, head)
-		mode->status = MODE_UNVERIFIED;
-
-	if (connector->force) {
-		if (connector->force == DRM_FORCE_ON)
-			connector->status = connector_status_connected;
-		else
-			connector->status = connector_status_disconnected;
-		if (connector->funcs->force)
-			connector->funcs->force(connector);
-	} else {
-		connector->status = connector->funcs->detect(connector, true);
-	}
-
-	/* Re-enable polling in case the global poll config changed. */
-	if (drm_kms_helper_poll != dev->mode_config.poll_running)
-		drm_kms_helper_poll_enable(dev);
-
-	dev->mode_config.poll_running = drm_kms_helper_poll;
-
-	if (connector->status == connector_status_disconnected) {
-		DRM_DEBUG_KMS("[CONNECTOR:%d:%s] disconnected\n",
-			connector->base.id, drm_get_connector_name(connector));
-		drm_mode_connector_update_edid_property(connector, NULL);
-		verbose_prune = false;
-		goto prune;
-	}
-
-#ifdef CONFIG_DRM_LOAD_EDID_FIRMWARE
-	count = drm_load_edid_firmware(connector);
-	if (count == 0)
-#endif
-		count = (*connector_funcs->get_modes)(connector);
-
-	if (count == 0 && connector->status == connector_status_connected)
-		count = drm_add_modes_noedid(connector, 1024, 768);
-	if (count == 0)
-		goto prune;
-
-	drm_mode_connector_list_update(connector);
-
-	if (maxX && maxY)
-		drm_mode_validate_size(dev, &connector->modes, maxX,
-				       maxY, 0);
-
-	if (connector->interlace_allowed)
-		mode_flags |= DRM_MODE_FLAG_INTERLACE;
-	if (connector->doublescan_allowed)
-		mode_flags |= DRM_MODE_FLAG_DBLSCAN;
-	if (connector->stereo_allowed)
-		mode_flags |= DRM_MODE_FLAG_3D_MASK;
-	drm_mode_validate_flag(connector, mode_flags);
-
-	list_for_each_entry(mode, &connector->modes, head) {
-		if (mode->status == MODE_OK)
-			mode->status = connector_funcs->mode_valid(connector,
-								   mode);
-	}
-
-prune:
-	drm_mode_prune_invalid(dev, &connector->modes, verbose_prune);
-
-	if (list_empty(&connector->modes))
-		return 0;
-
-	list_for_each_entry(mode, &connector->modes, head)
-		mode->vrefresh = drm_mode_vrefresh(mode);
-
-	drm_mode_sort(&connector->modes);
-
-	DRM_DEBUG_KMS("[CONNECTOR:%d:%s] probed modes :\n", connector->base.id,
-			drm_get_connector_name(connector));
-	list_for_each_entry(mode, &connector->modes, head) {
-		drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V);
-		drm_mode_debug_printmodeline(mode);
-	}
-
-	return count;
-}
-EXPORT_SYMBOL(drm_helper_probe_single_connector_modes);
-
 /**
  * drm_helper_encoder_in_use - check if a given encoder is in use
  * @encoder: encoder to check
  *
- * LOCKING:
- * Caller must hold mode config lock.
- *
- * Walk @encoders's DRM device's mode_config and see if it's in use.
+ * Checks whether @encoder is with the current mode setting output configuration
+ * in use by any connector. This doesn't mean that it is actually enabled since
+ * the DPMS state is tracked separately.
  *
- * RETURNS:
- * True if @encoder is part of the mode_config, false otherwise.
+ * Returns:
+ * True if @encoder is used, false otherwise.
  */
 bool drm_helper_encoder_in_use(struct drm_encoder *encoder)
 {
 	struct drm_connector *connector;
 	struct drm_device *dev = encoder->dev;
+
+	/*
+	 * We can expect this mutex to be locked if we are not panicking.
+	 * Locking is currently fubar in the panic handler.
+	 */
+	if (!oops_in_progress) {
+		WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+		WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+	}
+
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
 		if (connector->encoder == encoder)
 			return true;
@@ -242,19 +109,25 @@ EXPORT_SYMBOL(drm_helper_encoder_in_use);
  * drm_helper_crtc_in_use - check if a given CRTC is in a mode_config
  * @crtc: CRTC to check
  *
- * LOCKING:
- * Caller must hold mode config lock.
- *
- * Walk @crtc's DRM device's mode_config and see if it's in use.
+ * Checks whether @crtc is with the current mode setting output configuration
+ * in use by any connector. This doesn't mean that it is actually enabled since
+ * the DPMS state is tracked separately.
  *
- * RETURNS:
- * True if @crtc is part of the mode_config, false otherwise.
+ * Returns:
+ * True if @crtc is used, false otherwise.
  */
 bool drm_helper_crtc_in_use(struct drm_crtc *crtc)
 {
 	struct drm_encoder *encoder;
 	struct drm_device *dev = crtc->dev;
-	/* FIXME: Locking around list access? */
+
+	/*
+	 * We can expect this mutex to be locked if we are not panicking.
+	 * Locking is currently fubar in the panic handler.
+	 */
+	if (!oops_in_progress)
+		WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head)
 		if (encoder->crtc == crtc && drm_helper_encoder_in_use(encoder))
 			return true;
@@ -279,33 +152,17 @@ drm_encoder_disable(struct drm_encoder *encoder)
 		encoder->bridge->funcs->post_disable(encoder->bridge);
 }
 
-/**
- * drm_helper_disable_unused_functions - disable unused objects
- * @dev: DRM device
- *
- * LOCKING:
- * Caller must hold mode config lock.
- *
- * If an connector or CRTC isn't part of @dev's mode_config, it can be disabled
- * by calling its dpms function, which should power it off.
- */
-void drm_helper_disable_unused_functions(struct drm_device *dev)
+static void __drm_helper_disable_unused_functions(struct drm_device *dev)
 {
 	struct drm_encoder *encoder;
-	struct drm_connector *connector;
 	struct drm_crtc *crtc;
 
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		if (!connector->encoder)
-			continue;
-		if (connector->status == connector_status_disconnected)
-			connector->encoder = NULL;
-	}
+	drm_warn_on_modeset_not_all_locked(dev);
 
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
 		if (!drm_helper_encoder_in_use(encoder)) {
 			drm_encoder_disable(encoder);
-			/* disconnector encoder from any connector */
+			/* disconnect encoder from any connector */
 			encoder->crtc = NULL;
 		}
 	}
@@ -318,40 +175,28 @@ void drm_helper_disable_unused_functions(struct drm_device *dev)
 				(*crtc_funcs->disable)(crtc);
 			else
 				(*crtc_funcs->dpms)(crtc, DRM_MODE_DPMS_OFF);
-			crtc->fb = NULL;
+			crtc->primary->fb = NULL;
 		}
 	}
 }
-EXPORT_SYMBOL(drm_helper_disable_unused_functions);
 
 /**
- * drm_encoder_crtc_ok - can a given crtc drive a given encoder?
- * @encoder: encoder to test
- * @crtc: crtc to test
+ * drm_helper_disable_unused_functions - disable unused objects
+ * @dev: DRM device
  *
- * Return false if @encoder can't be driven by @crtc, true otherwise.
+ * This function walks through the entire mode setting configuration of @dev. It
+ * will remove any crtc links of unused encoders and encoder links of
+ * disconnected connectors. Then it will disable all unused encoders and crtcs
+ * either by calling their disable callback if available or by calling their
+ * dpms callback with DRM_MODE_DPMS_OFF.
  */
-static bool drm_encoder_crtc_ok(struct drm_encoder *encoder,
-				struct drm_crtc *crtc)
+void drm_helper_disable_unused_functions(struct drm_device *dev)
 {
-	struct drm_device *dev;
-	struct drm_crtc *tmp;
-	int crtc_mask = 1;
-
-	WARN(!crtc, "checking null crtc?\n");
-
-	dev = crtc->dev;
-
-	list_for_each_entry(tmp, &dev->mode_config.crtc_list, head) {
-		if (tmp == crtc)
-			break;
-		crtc_mask <<= 1;
-	}
-
-	if (encoder->possible_crtcs & crtc_mask)
-		return true;
-	return false;
+	drm_modeset_lock_all(dev);
+	__drm_helper_disable_unused_functions(dev);
+	drm_modeset_unlock_all(dev);
 }
+EXPORT_SYMBOL(drm_helper_disable_unused_functions);
 
 /*
  * Check the CRTC we're going to map each output to vs. its current
@@ -384,9 +229,6 @@ drm_crtc_prepare_encoders(struct drm_device *dev)
  * @y: vertical offset into the surface
  * @old_fb: old framebuffer, for cleanup
  *
- * LOCKING:
- * Caller must hold mode config lock.
- *
  * Try to set @mode on @crtc.  Give @crtc and its associated connectors a chance
  * to fixup or reject the mode prior to trying to set it. This is an internal
  * helper that drivers could e.g. use to update properties that require the
@@ -396,8 +238,8 @@ drm_crtc_prepare_encoders(struct drm_device *dev)
  * drm_crtc_helper_set_config() helper function to drive the mode setting
  * sequence.
  *
- * RETURNS:
- * True if the mode was set successfully, or false otherwise.
+ * Returns:
+ * True if the mode was set successfully, false otherwise.
  */
 bool drm_crtc_helper_set_mode(struct drm_crtc *crtc,
 			      struct drm_display_mode *mode,
@@ -413,6 +255,8 @@ bool drm_crtc_helper_set_mode(struct drm_crtc *crtc,
 	struct drm_encoder *encoder;
 	bool ret = true;
 
+	drm_warn_on_modeset_not_all_locked(dev);
+
 	saved_enabled = crtc->enabled;
 	crtc->enabled = drm_helper_crtc_in_use(crtc);
 	if (!crtc->enabled)
@@ -501,7 +345,7 @@ bool drm_crtc_helper_set_mode(struct drm_crtc *crtc,
 			continue;
 
 		DRM_DEBUG_KMS("[ENCODER:%d:%s] set [MODE:%d:%s]\n",
-			encoder->base.id, drm_get_encoder_name(encoder),
+			encoder->base.id, encoder->name,
 			mode->base.id, mode->name);
 		encoder_funcs = encoder->helper_private;
 		encoder_funcs->mode_set(encoder, mode, adjusted_mode);
@@ -536,7 +380,7 @@ bool drm_crtc_helper_set_mode(struct drm_crtc *crtc,
 	 * are later needed by vblank and swap-completion
 	 * timestamping. They are derived from true hwmode.
 	 */
-	drm_calc_timestamping_constants(crtc);
+	drm_calc_timestamping_constants(crtc, &crtc->hwmode);
 
 	/* FIXME: add subpixel order */
 done:
@@ -552,8 +396,7 @@ done:
 }
 EXPORT_SYMBOL(drm_crtc_helper_set_mode);
 
-
-static int
+static void
 drm_crtc_helper_disable(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -581,25 +424,21 @@ drm_crtc_helper_disable(struct drm_crtc *crtc)
 		}
 	}
 
-	drm_helper_disable_unused_functions(dev);
-	return 0;
+	__drm_helper_disable_unused_functions(dev);
 }
 
 /**
  * drm_crtc_helper_set_config - set a new config from userspace
  * @set: mode set configuration
  *
- * LOCKING:
- * Caller must hold mode config lock.
- *
  * Setup a new configuration, provided by the upper layers (either an ioctl call
- * from userspace or internally e.g. from the fbdev suppport code) in @set, and
+ * from userspace or internally e.g. from the fbdev support code) in @set, and
  * enable it. This is the main helper functions for drivers that implement
  * kernel mode setting with the crtc helper functions and the assorted
  * ->prepare(), ->modeset() and ->commit() helper callbacks.
  *
- * RETURNS:
- * Returns 0 on success, -ERRNO on failure.
+ * Returns:
+ * Returns 0 on success, negative errno numbers on failure.
  */
 int drm_crtc_helper_set_config(struct drm_mode_set *set)
 {
@@ -636,11 +475,14 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set)
 				(int)set->num_connectors, set->x, set->y);
 	} else {
 		DRM_DEBUG_KMS("[CRTC:%d] [NOFB]\n", set->crtc->base.id);
-		return drm_crtc_helper_disable(set->crtc);
+		drm_crtc_helper_disable(set->crtc);
+		return 0;
 	}
 
 	dev = set->crtc->dev;
 
+	drm_warn_on_modeset_not_all_locked(dev);
+
 	/*
 	 * Allocate space for the backup of all (non-pointer) encoder and
 	 * connector data.
@@ -676,19 +518,19 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set)
 	save_set.mode = &set->crtc->mode;
 	save_set.x = set->crtc->x;
 	save_set.y = set->crtc->y;
-	save_set.fb = set->crtc->fb;
+	save_set.fb = set->crtc->primary->fb;
 
 	/* We should be able to check here if the fb has the same properties
 	 * and then just flip_or_move it */
-	if (set->crtc->fb != set->fb) {
+	if (set->crtc->primary->fb != set->fb) {
 		/* If we have no fb then treat it as a full mode set */
-		if (set->crtc->fb == NULL) {
+		if (set->crtc->primary->fb == NULL) {
 			DRM_DEBUG_KMS("crtc has no fb, full mode set\n");
 			mode_changed = true;
 		} else if (set->fb == NULL) {
 			mode_changed = true;
 		} else if (set->fb->pixel_format !=
-			   set->crtc->fb->pixel_format) {
+			   set->crtc->primary->fb->pixel_format) {
 			mode_changed = true;
 		} else
 			fb_changed = true;
@@ -718,12 +560,13 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set)
 				if (new_encoder == NULL)
 					/* don't break so fail path works correct */
 					fail = 1;
-				break;
 
 				if (connector->dpms != DRM_MODE_DPMS_ON) {
 					DRM_DEBUG_KMS("connector dpms not on, full mode switch\n");
 					mode_changed = true;
 				}
+
+				break;
 			}
 		}
 
@@ -772,11 +615,11 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set)
 		}
 		if (new_crtc) {
 			DRM_DEBUG_KMS("[CONNECTOR:%d:%s] to [CRTC:%d]\n",
-				connector->base.id, drm_get_connector_name(connector),
+				connector->base.id, connector->name,
 				new_crtc->base.id);
 		} else {
 			DRM_DEBUG_KMS("[CONNECTOR:%d:%s] to [NOCRTC]\n",
-				connector->base.id, drm_get_connector_name(connector));
+				connector->base.id, connector->name);
 		}
 	}
 
@@ -789,34 +632,34 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set)
 			DRM_DEBUG_KMS("attempting to set mode from"
 					" userspace\n");
 			drm_mode_debug_printmodeline(set->mode);
-			set->crtc->fb = set->fb;
+			set->crtc->primary->fb = set->fb;
 			if (!drm_crtc_helper_set_mode(set->crtc, set->mode,
 						      set->x, set->y,
 						      save_set.fb)) {
 				DRM_ERROR("failed to set mode on [CRTC:%d]\n",
 					  set->crtc->base.id);
-				set->crtc->fb = save_set.fb;
+				set->crtc->primary->fb = save_set.fb;
 				ret = -EINVAL;
 				goto fail;
 			}
 			DRM_DEBUG_KMS("Setting connector DPMS state to on\n");
 			for (i = 0; i < set->num_connectors; i++) {
 				DRM_DEBUG_KMS("\t[CONNECTOR:%d:%s] set DPMS on\n", set->connectors[i]->base.id,
-					      drm_get_connector_name(set->connectors[i]));
+					      set->connectors[i]->name);
 				set->connectors[i]->funcs->dpms(set->connectors[i], DRM_MODE_DPMS_ON);
 			}
 		}
-		drm_helper_disable_unused_functions(dev);
+		__drm_helper_disable_unused_functions(dev);
 	} else if (fb_changed) {
 		set->crtc->x = set->x;
 		set->crtc->y = set->y;
-		set->crtc->fb = set->fb;
+		set->crtc->primary->fb = set->fb;
 		ret = crtc_funcs->mode_set_base(set->crtc,
 						set->x, set->y, save_set.fb);
 		if (ret != 0) {
 			set->crtc->x = save_set.x;
 			set->crtc->y = save_set.y;
-			set->crtc->fb = save_set.fb;
+			set->crtc->primary->fb = save_set.fb;
 			goto fail;
 		}
 	}
@@ -953,8 +796,16 @@ void drm_helper_connector_dpms(struct drm_connector *connector, int mode)
 }
 EXPORT_SYMBOL(drm_helper_connector_dpms);
 
-int drm_helper_mode_fill_fb_struct(struct drm_framebuffer *fb,
-				   struct drm_mode_fb_cmd2 *mode_cmd)
+/**
+ * drm_helper_mode_fill_fb_struct - fill out framebuffer metadata
+ * @fb: drm_framebuffer object to fill out
+ * @mode_cmd: metadata from the userspace fb creation request
+ *
+ * This helper can be used in a drivers fb_create callback to pre-fill the fb's
+ * metadata fields.
+ */
+void drm_helper_mode_fill_fb_struct(struct drm_framebuffer *fb,
+				    struct drm_mode_fb_cmd2 *mode_cmd)
 {
 	int i;
 
@@ -967,26 +818,47 @@ int drm_helper_mode_fill_fb_struct(struct drm_framebuffer *fb,
 	drm_fb_get_bpp_depth(mode_cmd->pixel_format, &fb->depth,
 				    &fb->bits_per_pixel);
 	fb->pixel_format = mode_cmd->pixel_format;
-
-	return 0;
 }
 EXPORT_SYMBOL(drm_helper_mode_fill_fb_struct);
 
-int drm_helper_resume_force_mode(struct drm_device *dev)
+/**
+ * drm_helper_resume_force_mode - force-restore mode setting configuration
+ * @dev: drm_device which should be restored
+ *
+ * Drivers which use the mode setting helpers can use this function to
+ * force-restore the mode setting configuration e.g. on resume or when something
+ * else might have trampled over the hw state (like some overzealous old BIOSen
+ * tended to do).
+ *
+ * This helper doesn't provide a error return value since restoring the old
+ * config should never fail due to resource allocation issues since the driver
+ * has successfully set the restored configuration already. Hence this should
+ * boil down to the equivalent of a few dpms on calls, which also don't provide
+ * an error code.
+ *
+ * Drivers where simply restoring an old configuration again might fail (e.g.
+ * due to slight differences in allocating shared resources when the
+ * configuration is restored in a different order than when userspace set it up)
+ * need to use their own restore logic.
+ */
+void drm_helper_resume_force_mode(struct drm_device *dev)
 {
 	struct drm_crtc *crtc;
 	struct drm_encoder *encoder;
 	struct drm_crtc_helper_funcs *crtc_funcs;
-	int ret, encoder_dpms;
+	int encoder_dpms;
+	bool ret;
 
+	drm_modeset_lock_all(dev);
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 
 		if (!crtc->enabled)
 			continue;
 
 		ret = drm_crtc_helper_set_mode(crtc, &crtc->mode,
-					       crtc->x, crtc->y, crtc->fb);
+					       crtc->x, crtc->y, crtc->primary->fb);
 
+		/* Restoring the old config should never fail! */
 		if (ret == false)
 			DRM_ERROR("failed to set mode on crtc %p\n", crtc);
 
@@ -1009,155 +881,9 @@ int drm_helper_resume_force_mode(struct drm_device *dev)
 						     drm_helper_choose_crtc_dpms(crtc));
 		}
 	}
+
 	/* disable the unused connectors while restoring the modesetting */
-	drm_helper_disable_unused_functions(dev);
-	return 0;
+	__drm_helper_disable_unused_functions(dev);
+	drm_modeset_unlock_all(dev);
 }
 EXPORT_SYMBOL(drm_helper_resume_force_mode);
-
-void drm_kms_helper_hotplug_event(struct drm_device *dev)
-{
-	/* send a uevent + call fbdev */
-	drm_sysfs_hotplug_event(dev);
-	if (dev->mode_config.funcs->output_poll_changed)
-		dev->mode_config.funcs->output_poll_changed(dev);
-}
-EXPORT_SYMBOL(drm_kms_helper_hotplug_event);
-
-#define DRM_OUTPUT_POLL_PERIOD (10*HZ)
-static void output_poll_execute(struct work_struct *work)
-{
-	struct delayed_work *delayed_work = to_delayed_work(work);
-	struct drm_device *dev = container_of(delayed_work, struct drm_device, mode_config.output_poll_work);
-	struct drm_connector *connector;
-	enum drm_connector_status old_status;
-	bool repoll = false, changed = false;
-
-	if (!drm_kms_helper_poll)
-		return;
-
-	mutex_lock(&dev->mode_config.mutex);
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-
-		/* Ignore forced connectors. */
-		if (connector->force)
-			continue;
-
-		/* Ignore HPD capable connectors and connectors where we don't
-		 * want any hotplug detection at all for polling. */
-		if (!connector->polled || connector->polled == DRM_CONNECTOR_POLL_HPD)
-			continue;
-
-		repoll = true;
-
-		old_status = connector->status;
-		/* if we are connected and don't want to poll for disconnect
-		   skip it */
-		if (old_status == connector_status_connected &&
-		    !(connector->polled & DRM_CONNECTOR_POLL_DISCONNECT))
-			continue;
-
-		connector->status = connector->funcs->detect(connector, false);
-		if (old_status != connector->status) {
-			const char *old, *new;
-
-			old = drm_get_connector_status_name(old_status);
-			new = drm_get_connector_status_name(connector->status);
-
-			DRM_DEBUG_KMS("[CONNECTOR:%d:%s] "
-				      "status updated from %s to %s\n",
-				      connector->base.id,
-				      drm_get_connector_name(connector),
-				      old, new);
-
-			changed = true;
-		}
-	}
-
-	mutex_unlock(&dev->mode_config.mutex);
-
-	if (changed)
-		drm_kms_helper_hotplug_event(dev);
-
-	if (repoll)
-		schedule_delayed_work(delayed_work, DRM_OUTPUT_POLL_PERIOD);
-}
-
-void drm_kms_helper_poll_disable(struct drm_device *dev)
-{
-	if (!dev->mode_config.poll_enabled)
-		return;
-	cancel_delayed_work_sync(&dev->mode_config.output_poll_work);
-}
-EXPORT_SYMBOL(drm_kms_helper_poll_disable);
-
-void drm_kms_helper_poll_enable(struct drm_device *dev)
-{
-	bool poll = false;
-	struct drm_connector *connector;
-
-	if (!dev->mode_config.poll_enabled || !drm_kms_helper_poll)
-		return;
-
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		if (connector->polled & (DRM_CONNECTOR_POLL_CONNECT |
-					 DRM_CONNECTOR_POLL_DISCONNECT))
-			poll = true;
-	}
-
-	if (poll)
-		schedule_delayed_work(&dev->mode_config.output_poll_work, DRM_OUTPUT_POLL_PERIOD);
-}
-EXPORT_SYMBOL(drm_kms_helper_poll_enable);
-
-void drm_kms_helper_poll_init(struct drm_device *dev)
-{
-	INIT_DELAYED_WORK(&dev->mode_config.output_poll_work, output_poll_execute);
-	dev->mode_config.poll_enabled = true;
-
-	drm_kms_helper_poll_enable(dev);
-}
-EXPORT_SYMBOL(drm_kms_helper_poll_init);
-
-void drm_kms_helper_poll_fini(struct drm_device *dev)
-{
-	drm_kms_helper_poll_disable(dev);
-}
-EXPORT_SYMBOL(drm_kms_helper_poll_fini);
-
-bool drm_helper_hpd_irq_event(struct drm_device *dev)
-{
-	struct drm_connector *connector;
-	enum drm_connector_status old_status;
-	bool changed = false;
-
-	if (!dev->mode_config.poll_enabled)
-		return false;
-
-	mutex_lock(&dev->mode_config.mutex);
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-
-		/* Only handle HPD capable connectors. */
-		if (!(connector->polled & DRM_CONNECTOR_POLL_HPD))
-			continue;
-
-		old_status = connector->status;
-
-		connector->status = connector->funcs->detect(connector, false);
-		DRM_DEBUG_KMS("[CONNECTOR:%d:%s] status updated from %s to %s\n",
-			      connector->base.id,
-			      drm_get_connector_name(connector),
-			      drm_get_connector_status_name(old_status),
-			      drm_get_connector_status_name(connector->status));
-		if (old_status != connector->status)
-			changed = true;
-	}
-
-	mutex_unlock(&dev->mode_config.mutex);
-
-	if (changed)
-		drm_kms_helper_hotplug_event(dev);
-
-	return changed;
-}
-EXPORT_SYMBOL(drm_helper_hpd_irq_event);
diff --git a/drivers/gpu/drm/drm_crtc_internal.h b/drivers/gpu/drm/drm_crtc_internal.h
new file mode 100644
index 00000000000..a2945ee6d67
--- /dev/null
+++ b/drivers/gpu/drm/drm_crtc_internal.h
@@ -0,0 +1,38 @@
+/*
+ * Copyright © 2006 Keith Packard
+ * Copyright © 2007-2008 Dave Airlie
+ * Copyright © 2007-2008 Intel Corporation
+ *   Jesse Barnes <jesse.barnes@intel.com>
+ * Copyright © 2014 Intel Corporation
+ *   Daniel Vetter <daniel.vetter@ffwll.ch>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+/*
+ * This header file contains mode setting related functions and definitions
+ * which are only used within the drm module as internal implementation details
+ * and are not exported to drivers.
+ */
+
+int drm_mode_object_get(struct drm_device *dev,
+			struct drm_mode_object *obj, uint32_t obj_type);
+void drm_mode_object_put(struct drm_device *dev,
+			 struct drm_mode_object *object);
+
diff --git a/drivers/gpu/drm/drm_dp_helper.c b/drivers/gpu/drm/drm_dp_helper.c
index 9e978aae897..08e33b8b13a 100644
--- a/drivers/gpu/drm/drm_dp_helper.c
+++ b/drivers/gpu/drm/drm_dp_helper.c
@@ -206,13 +206,17 @@ i2c_dp_aux_prepare_bus(struct i2c_adapter *adapter)
  * i2c_dp_aux_add_bus() - register an i2c adapter using the aux ch helper
  * @adapter: i2c adapter to register
  *
- * This registers an i2c adapater that uses dp aux channel as it's underlaying
+ * This registers an i2c adapter that uses dp aux channel as it's underlaying
  * transport. The driver needs to fill out the &i2c_algo_dp_aux_data structure
  * and store it in the algo_data member of the @adapter argument. This will be
  * used by the i2c over dp aux algorithm to drive the hardware.
  *
  * RETURNS:
  * 0 on success, -ERRNO on failure.
+ *
+ * IMPORTANT:
+ * This interface is deprecated, please switch to the new dp aux helpers and
+ * drm_dp_aux_register().
  */
 int
 i2c_dp_aux_add_bus(struct i2c_adapter *adapter)
@@ -346,3 +350,421 @@ int drm_dp_bw_code_to_link_rate(u8 link_bw)
 	}
 }
 EXPORT_SYMBOL(drm_dp_bw_code_to_link_rate);
+
+/**
+ * DOC: dp helpers
+ *
+ * The DisplayPort AUX channel is an abstraction to allow generic, driver-
+ * independent access to AUX functionality. Drivers can take advantage of
+ * this by filling in the fields of the drm_dp_aux structure.
+ *
+ * Transactions are described using a hardware-independent drm_dp_aux_msg
+ * structure, which is passed into a driver's .transfer() implementation.
+ * Both native and I2C-over-AUX transactions are supported.
+ */
+
+static int drm_dp_dpcd_access(struct drm_dp_aux *aux, u8 request,
+			      unsigned int offset, void *buffer, size_t size)
+{
+	struct drm_dp_aux_msg msg;
+	unsigned int retry;
+	int err;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.address = offset;
+	msg.request = request;
+	msg.buffer = buffer;
+	msg.size = size;
+
+	/*
+	 * The specification doesn't give any recommendation on how often to
+	 * retry native transactions, so retry 7 times like for I2C-over-AUX
+	 * transactions.
+	 */
+	for (retry = 0; retry < 7; retry++) {
+
+		mutex_lock(&aux->hw_mutex);
+		err = aux->transfer(aux, &msg);
+		mutex_unlock(&aux->hw_mutex);
+		if (err < 0) {
+			if (err == -EBUSY)
+				continue;
+
+			return err;
+		}
+
+
+		switch (msg.reply & DP_AUX_NATIVE_REPLY_MASK) {
+		case DP_AUX_NATIVE_REPLY_ACK:
+			if (err < size)
+				return -EPROTO;
+			return err;
+
+		case DP_AUX_NATIVE_REPLY_NACK:
+			return -EIO;
+
+		case DP_AUX_NATIVE_REPLY_DEFER:
+			usleep_range(400, 500);
+			break;
+		}
+	}
+
+	DRM_DEBUG_KMS("too many retries, giving up\n");
+	return -EIO;
+}
+
+/**
+ * drm_dp_dpcd_read() - read a series of bytes from the DPCD
+ * @aux: DisplayPort AUX channel
+ * @offset: address of the (first) register to read
+ * @buffer: buffer to store the register values
+ * @size: number of bytes in @buffer
+ *
+ * Returns the number of bytes transferred on success, or a negative error
+ * code on failure. -EIO is returned if the request was NAKed by the sink or
+ * if the retry count was exceeded. If not all bytes were transferred, this
+ * function returns -EPROTO. Errors from the underlying AUX channel transfer
+ * function, with the exception of -EBUSY (which causes the transaction to
+ * be retried), are propagated to the caller.
+ */
+ssize_t drm_dp_dpcd_read(struct drm_dp_aux *aux, unsigned int offset,
+			 void *buffer, size_t size)
+{
+	return drm_dp_dpcd_access(aux, DP_AUX_NATIVE_READ, offset, buffer,
+				  size);
+}
+EXPORT_SYMBOL(drm_dp_dpcd_read);
+
+/**
+ * drm_dp_dpcd_write() - write a series of bytes to the DPCD
+ * @aux: DisplayPort AUX channel
+ * @offset: address of the (first) register to write
+ * @buffer: buffer containing the values to write
+ * @size: number of bytes in @buffer
+ *
+ * Returns the number of bytes transferred on success, or a negative error
+ * code on failure. -EIO is returned if the request was NAKed by the sink or
+ * if the retry count was exceeded. If not all bytes were transferred, this
+ * function returns -EPROTO. Errors from the underlying AUX channel transfer
+ * function, with the exception of -EBUSY (which causes the transaction to
+ * be retried), are propagated to the caller.
+ */
+ssize_t drm_dp_dpcd_write(struct drm_dp_aux *aux, unsigned int offset,
+			  void *buffer, size_t size)
+{
+	return drm_dp_dpcd_access(aux, DP_AUX_NATIVE_WRITE, offset, buffer,
+				  size);
+}
+EXPORT_SYMBOL(drm_dp_dpcd_write);
+
+/**
+ * drm_dp_dpcd_read_link_status() - read DPCD link status (bytes 0x202-0x207)
+ * @aux: DisplayPort AUX channel
+ * @status: buffer to store the link status in (must be at least 6 bytes)
+ *
+ * Returns the number of bytes transferred on success or a negative error
+ * code on failure.
+ */
+int drm_dp_dpcd_read_link_status(struct drm_dp_aux *aux,
+				 u8 status[DP_LINK_STATUS_SIZE])
+{
+	return drm_dp_dpcd_read(aux, DP_LANE0_1_STATUS, status,
+				DP_LINK_STATUS_SIZE);
+}
+EXPORT_SYMBOL(drm_dp_dpcd_read_link_status);
+
+/**
+ * drm_dp_link_probe() - probe a DisplayPort link for capabilities
+ * @aux: DisplayPort AUX channel
+ * @link: pointer to structure in which to return link capabilities
+ *
+ * The structure filled in by this function can usually be passed directly
+ * into drm_dp_link_power_up() and drm_dp_link_configure() to power up and
+ * configure the link based on the link's capabilities.
+ *
+ * Returns 0 on success or a negative error code on failure.
+ */
+int drm_dp_link_probe(struct drm_dp_aux *aux, struct drm_dp_link *link)
+{
+	u8 values[3];
+	int err;
+
+	memset(link, 0, sizeof(*link));
+
+	err = drm_dp_dpcd_read(aux, DP_DPCD_REV, values, sizeof(values));
+	if (err < 0)
+		return err;
+
+	link->revision = values[0];
+	link->rate = drm_dp_bw_code_to_link_rate(values[1]);
+	link->num_lanes = values[2] & DP_MAX_LANE_COUNT_MASK;
+
+	if (values[2] & DP_ENHANCED_FRAME_CAP)
+		link->capabilities |= DP_LINK_CAP_ENHANCED_FRAMING;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_dp_link_probe);
+
+/**
+ * drm_dp_link_power_up() - power up a DisplayPort link
+ * @aux: DisplayPort AUX channel
+ * @link: pointer to a structure containing the link configuration
+ *
+ * Returns 0 on success or a negative error code on failure.
+ */
+int drm_dp_link_power_up(struct drm_dp_aux *aux, struct drm_dp_link *link)
+{
+	u8 value;
+	int err;
+
+	/* DP_SET_POWER register is only available on DPCD v1.1 and later */
+	if (link->revision < 0x11)
+		return 0;
+
+	err = drm_dp_dpcd_readb(aux, DP_SET_POWER, &value);
+	if (err < 0)
+		return err;
+
+	value &= ~DP_SET_POWER_MASK;
+	value |= DP_SET_POWER_D0;
+
+	err = drm_dp_dpcd_writeb(aux, DP_SET_POWER, value);
+	if (err < 0)
+		return err;
+
+	/*
+	 * According to the DP 1.1 specification, a "Sink Device must exit the
+	 * power saving state within 1 ms" (Section 2.5.3.1, Table 5-52, "Sink
+	 * Control Field" (register 0x600).
+	 */
+	usleep_range(1000, 2000);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_dp_link_power_up);
+
+/**
+ * drm_dp_link_configure() - configure a DisplayPort link
+ * @aux: DisplayPort AUX channel
+ * @link: pointer to a structure containing the link configuration
+ *
+ * Returns 0 on success or a negative error code on failure.
+ */
+int drm_dp_link_configure(struct drm_dp_aux *aux, struct drm_dp_link *link)
+{
+	u8 values[2];
+	int err;
+
+	values[0] = drm_dp_link_rate_to_bw_code(link->rate);
+	values[1] = link->num_lanes;
+
+	if (link->capabilities & DP_LINK_CAP_ENHANCED_FRAMING)
+		values[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
+
+	err = drm_dp_dpcd_write(aux, DP_LINK_BW_SET, values, sizeof(values));
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_dp_link_configure);
+
+/*
+ * I2C-over-AUX implementation
+ */
+
+static u32 drm_dp_i2c_functionality(struct i2c_adapter *adapter)
+{
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL |
+	       I2C_FUNC_SMBUS_READ_BLOCK_DATA |
+	       I2C_FUNC_SMBUS_BLOCK_PROC_CALL |
+	       I2C_FUNC_10BIT_ADDR;
+}
+
+/*
+ * Transfer a single I2C-over-AUX message and handle various error conditions,
+ * retrying the transaction as appropriate.  It is assumed that the
+ * aux->transfer function does not modify anything in the msg other than the
+ * reply field.
+ */
+static int drm_dp_i2c_do_msg(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg)
+{
+	unsigned int retry;
+	int err;
+
+	/*
+	 * DP1.2 sections 2.7.7.1.5.6.1 and 2.7.7.1.6.6.1: A DP Source device
+	 * is required to retry at least seven times upon receiving AUX_DEFER
+	 * before giving up the AUX transaction.
+	 */
+	for (retry = 0; retry < 7; retry++) {
+		mutex_lock(&aux->hw_mutex);
+		err = aux->transfer(aux, msg);
+		mutex_unlock(&aux->hw_mutex);
+		if (err < 0) {
+			if (err == -EBUSY)
+				continue;
+
+			DRM_DEBUG_KMS("transaction failed: %d\n", err);
+			return err;
+		}
+
+
+		switch (msg->reply & DP_AUX_NATIVE_REPLY_MASK) {
+		case DP_AUX_NATIVE_REPLY_ACK:
+			/*
+			 * For I2C-over-AUX transactions this isn't enough, we
+			 * need to check for the I2C ACK reply.
+			 */
+			break;
+
+		case DP_AUX_NATIVE_REPLY_NACK:
+			DRM_DEBUG_KMS("native nack\n");
+			return -EREMOTEIO;
+
+		case DP_AUX_NATIVE_REPLY_DEFER:
+			DRM_DEBUG_KMS("native defer");
+			/*
+			 * We could check for I2C bit rate capabilities and if
+			 * available adjust this interval. We could also be
+			 * more careful with DP-to-legacy adapters where a
+			 * long legacy cable may force very low I2C bit rates.
+			 *
+			 * For now just defer for long enough to hopefully be
+			 * safe for all use-cases.
+			 */
+			usleep_range(500, 600);
+			continue;
+
+		default:
+			DRM_ERROR("invalid native reply %#04x\n", msg->reply);
+			return -EREMOTEIO;
+		}
+
+		switch (msg->reply & DP_AUX_I2C_REPLY_MASK) {
+		case DP_AUX_I2C_REPLY_ACK:
+			/*
+			 * Both native ACK and I2C ACK replies received. We
+			 * can assume the transfer was successful.
+			 */
+			if (err < msg->size)
+				return -EPROTO;
+			return 0;
+
+		case DP_AUX_I2C_REPLY_NACK:
+			DRM_DEBUG_KMS("I2C nack\n");
+			return -EREMOTEIO;
+
+		case DP_AUX_I2C_REPLY_DEFER:
+			DRM_DEBUG_KMS("I2C defer\n");
+			usleep_range(400, 500);
+			continue;
+
+		default:
+			DRM_ERROR("invalid I2C reply %#04x\n", msg->reply);
+			return -EREMOTEIO;
+		}
+	}
+
+	DRM_DEBUG_KMS("too many retries, giving up\n");
+	return -EREMOTEIO;
+}
+
+static int drm_dp_i2c_xfer(struct i2c_adapter *adapter, struct i2c_msg *msgs,
+			   int num)
+{
+	struct drm_dp_aux *aux = adapter->algo_data;
+	unsigned int i, j;
+	struct drm_dp_aux_msg msg;
+	int err = 0;
+
+	memset(&msg, 0, sizeof(msg));
+
+	for (i = 0; i < num; i++) {
+		msg.address = msgs[i].addr;
+		msg.request = (msgs[i].flags & I2C_M_RD) ?
+			DP_AUX_I2C_READ :
+			DP_AUX_I2C_WRITE;
+		msg.request |= DP_AUX_I2C_MOT;
+		/* Send a bare address packet to start the transaction.
+		 * Zero sized messages specify an address only (bare
+		 * address) transaction.
+		 */
+		msg.buffer = NULL;
+		msg.size = 0;
+		err = drm_dp_i2c_do_msg(aux, &msg);
+		if (err < 0)
+			break;
+		/*
+		 * Many hardware implementations support FIFOs larger than a
+		 * single byte, but it has been empirically determined that
+		 * transferring data in larger chunks can actually lead to
+		 * decreased performance. Therefore each message is simply
+		 * transferred byte-by-byte.
+		 */
+		for (j = 0; j < msgs[i].len; j++) {
+			msg.buffer = msgs[i].buf + j;
+			msg.size = 1;
+
+			err = drm_dp_i2c_do_msg(aux, &msg);
+			if (err < 0)
+				break;
+		}
+		if (err < 0)
+			break;
+	}
+	if (err >= 0)
+		err = num;
+	/* Send a bare address packet to close out the transaction.
+	 * Zero sized messages specify an address only (bare
+	 * address) transaction.
+	 */
+	msg.request &= ~DP_AUX_I2C_MOT;
+	msg.buffer = NULL;
+	msg.size = 0;
+	(void)drm_dp_i2c_do_msg(aux, &msg);
+
+	return err;
+}
+
+static const struct i2c_algorithm drm_dp_i2c_algo = {
+	.functionality = drm_dp_i2c_functionality,
+	.master_xfer = drm_dp_i2c_xfer,
+};
+
+/**
+ * drm_dp_aux_register() - initialise and register aux channel
+ * @aux: DisplayPort AUX channel
+ *
+ * Returns 0 on success or a negative error code on failure.
+ */
+int drm_dp_aux_register(struct drm_dp_aux *aux)
+{
+	mutex_init(&aux->hw_mutex);
+
+	aux->ddc.algo = &drm_dp_i2c_algo;
+	aux->ddc.algo_data = aux;
+	aux->ddc.retries = 3;
+
+	aux->ddc.class = I2C_CLASS_DDC;
+	aux->ddc.owner = THIS_MODULE;
+	aux->ddc.dev.parent = aux->dev;
+	aux->ddc.dev.of_node = aux->dev->of_node;
+
+	strlcpy(aux->ddc.name, aux->name ? aux->name : dev_name(aux->dev),
+		sizeof(aux->ddc.name));
+
+	return i2c_add_adapter(&aux->ddc);
+}
+EXPORT_SYMBOL(drm_dp_aux_register);
+
+/**
+ * drm_dp_aux_unregister() - unregister an AUX adapter
+ * @aux: DisplayPort AUX channel
+ */
+void drm_dp_aux_unregister(struct drm_dp_aux *aux)
+{
+	i2c_del_adapter(&aux->ddc);
+}
+EXPORT_SYMBOL(drm_dp_aux_unregister);
diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
index d9137e49c4e..8218078b613 100644
--- a/drivers/gpu/drm/drm_drv.c
+++ b/drivers/gpu/drm/drm_drv.c
@@ -286,6 +286,45 @@ static int drm_version(struct drm_device *dev, void *data,
 }
 
 /**
+ * drm_ioctl_permit - Check ioctl permissions against caller
+ *
+ * @flags: ioctl permission flags.
+ * @file_priv: Pointer to struct drm_file identifying the caller.
+ *
+ * Checks whether the caller is allowed to run an ioctl with the
+ * indicated permissions. If so, returns zero. Otherwise returns an
+ * error code suitable for ioctl return.
+ */
+static int drm_ioctl_permit(u32 flags, struct drm_file *file_priv)
+{
+	/* ROOT_ONLY is only for CAP_SYS_ADMIN */
+	if (unlikely((flags & DRM_ROOT_ONLY) && !capable(CAP_SYS_ADMIN)))
+		return -EACCES;
+
+	/* AUTH is only for authenticated or render client */
+	if (unlikely((flags & DRM_AUTH) && !drm_is_render_client(file_priv) &&
+		     !file_priv->authenticated))
+		return -EACCES;
+
+	/* MASTER is only for master or control clients */
+	if (unlikely((flags & DRM_MASTER) && !file_priv->is_master &&
+		     !drm_is_control_client(file_priv)))
+		return -EACCES;
+
+	/* Control clients must be explicitly allowed */
+	if (unlikely(!(flags & DRM_CONTROL_ALLOW) &&
+		     drm_is_control_client(file_priv)))
+		return -EACCES;
+
+	/* Render clients must be explicitly allowed */
+	if (unlikely(!(flags & DRM_RENDER_ALLOW) &&
+		     drm_is_render_client(file_priv)))
+		return -EACCES;
+
+	return 0;
+}
+
+/**
  * Called whenever a process performs an ioctl on /dev/drm.
  *
  * \param inode device inode.
@@ -315,9 +354,6 @@ long drm_ioctl(struct file *filp,
 	if (drm_device_is_unplugged(dev))
 		return -ENODEV;
 
-	atomic_inc(&dev->ioctl_count);
-	++file_priv->ioctl_count;
-
 	if ((nr >= DRM_CORE_IOCTL_COUNT) &&
 	    ((nr < DRM_COMMAND_BASE) || (nr >= DRM_COMMAND_END)))
 		goto err_i1;
@@ -347,72 +383,89 @@ long drm_ioctl(struct file *filp,
 
 	DRM_DEBUG("pid=%d, dev=0x%lx, auth=%d, %s\n",
 		  task_pid_nr(current),
-		  (long)old_encode_dev(file_priv->minor->device),
+		  (long)old_encode_dev(file_priv->minor->kdev->devt),
 		  file_priv->authenticated, ioctl->name);
 
 	/* Do not trust userspace, use our own definition */
 	func = ioctl->func;
 
-	if (!func) {
+	if (unlikely(!func)) {
 		DRM_DEBUG("no function\n");
 		retcode = -EINVAL;
-	} else if (((ioctl->flags & DRM_ROOT_ONLY) && !capable(CAP_SYS_ADMIN)) ||
-		   ((ioctl->flags & DRM_AUTH) && !drm_is_render_client(file_priv) && !file_priv->authenticated) ||
-		   ((ioctl->flags & DRM_MASTER) && !file_priv->is_master) ||
-		   (!(ioctl->flags & DRM_CONTROL_ALLOW) && (file_priv->minor->type == DRM_MINOR_CONTROL)) ||
-		   (!(ioctl->flags & DRM_RENDER_ALLOW) && drm_is_render_client(file_priv))) {
-		retcode = -EACCES;
-	} else {
-		if (cmd & (IOC_IN | IOC_OUT)) {
-			if (asize <= sizeof(stack_kdata)) {
-				kdata = stack_kdata;
-			} else {
-				kdata = kmalloc(asize, GFP_KERNEL);
-				if (!kdata) {
-					retcode = -ENOMEM;
-					goto err_i1;
-				}
-			}
-			if (asize > usize)
-				memset(kdata + usize, 0, asize - usize);
-		}
+		goto err_i1;
+	}
 
-		if (cmd & IOC_IN) {
-			if (copy_from_user(kdata, (void __user *)arg,
-					   usize) != 0) {
-				retcode = -EFAULT;
+	retcode = drm_ioctl_permit(ioctl->flags, file_priv);
+	if (unlikely(retcode))
+		goto err_i1;
+
+	if (cmd & (IOC_IN | IOC_OUT)) {
+		if (asize <= sizeof(stack_kdata)) {
+			kdata = stack_kdata;
+		} else {
+			kdata = kmalloc(asize, GFP_KERNEL);
+			if (!kdata) {
+				retcode = -ENOMEM;
 				goto err_i1;
 			}
-		} else
-			memset(kdata, 0, usize);
-
-		if (ioctl->flags & DRM_UNLOCKED)
-			retcode = func(dev, kdata, file_priv);
-		else {
-			mutex_lock(&drm_global_mutex);
-			retcode = func(dev, kdata, file_priv);
-			mutex_unlock(&drm_global_mutex);
 		}
+		if (asize > usize)
+			memset(kdata + usize, 0, asize - usize);
+	}
 
-		if (cmd & IOC_OUT) {
-			if (copy_to_user((void __user *)arg, kdata,
-					 usize) != 0)
-				retcode = -EFAULT;
+	if (cmd & IOC_IN) {
+		if (copy_from_user(kdata, (void __user *)arg,
+				   usize) != 0) {
+			retcode = -EFAULT;
+			goto err_i1;
 		}
+	} else if (cmd & IOC_OUT) {
+		memset(kdata, 0, usize);
+	}
+
+	if (ioctl->flags & DRM_UNLOCKED)
+		retcode = func(dev, kdata, file_priv);
+	else {
+		mutex_lock(&drm_global_mutex);
+		retcode = func(dev, kdata, file_priv);
+		mutex_unlock(&drm_global_mutex);
+	}
+
+	if (cmd & IOC_OUT) {
+		if (copy_to_user((void __user *)arg, kdata,
+				 usize) != 0)
+			retcode = -EFAULT;
 	}
 
       err_i1:
 	if (!ioctl)
 		DRM_DEBUG("invalid ioctl: pid=%d, dev=0x%lx, auth=%d, cmd=0x%02x, nr=0x%02x\n",
 			  task_pid_nr(current),
-			  (long)old_encode_dev(file_priv->minor->device),
+			  (long)old_encode_dev(file_priv->minor->kdev->devt),
 			  file_priv->authenticated, cmd, nr);
 
 	if (kdata != stack_kdata)
 		kfree(kdata);
-	atomic_dec(&dev->ioctl_count);
 	if (retcode)
 		DRM_DEBUG("ret = %d\n", retcode);
 	return retcode;
 }
 EXPORT_SYMBOL(drm_ioctl);
+
+/**
+ * drm_ioctl_flags - Check for core ioctl and return ioctl permission flags
+ *
+ * @nr: Ioctl number.
+ * @flags: Where to return the ioctl permission flags
+ */
+bool drm_ioctl_flags(unsigned int nr, unsigned int *flags)
+{
+	if ((nr >= DRM_COMMAND_END && nr < DRM_CORE_IOCTL_COUNT) ||
+	    (nr < DRM_COMMAND_BASE)) {
+		*flags = drm_ioctls[nr].flags;
+		return true;
+	}
+
+	return false;
+}
+EXPORT_SYMBOL(drm_ioctl_flags);
diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 8835dcddfac..dfa9769b26b 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -70,6 +70,8 @@
 #define EDID_QUIRK_FORCE_REDUCED_BLANKING	(1 << 7)
 /* Force 8bpc */
 #define EDID_QUIRK_FORCE_8BPC			(1 << 8)
+/* Force 12bpc */
+#define EDID_QUIRK_FORCE_12BPC			(1 << 9)
 
 struct detailed_mode_closure {
 	struct drm_connector *connector;
@@ -125,6 +127,9 @@ static struct edid_quirk {
 	{ "SAM", 596, EDID_QUIRK_PREFER_LARGE_60 },
 	{ "SAM", 638, EDID_QUIRK_PREFER_LARGE_60 },
 
+	/* Sony PVM-2541A does up to 12 bpc, but only reports max 8 bpc */
+	{ "SNY", 0x2541, EDID_QUIRK_FORCE_12BPC },
+
 	/* ViewSonic VA2026w */
 	{ "VSC", 5020, EDID_QUIRK_FORCE_REDUCED_BLANKING },
 
@@ -605,347 +610,347 @@ static const struct drm_display_mode edid_cea_modes[] = {
 	{ DRM_MODE("640x480", DRM_MODE_TYPE_DRIVER, 25175, 640, 656,
 		   752, 800, 0, 480, 490, 492, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 2 - 720x480@60Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 27000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 3 - 720x480@60Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 27000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 4 - 1280x720@60Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 1390,
 		   1430, 1650, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 5 - 1920x1080i@60Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 6 - 1440x480i@60Hz */
 	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
 		   1602, 1716, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 7 - 1440x480i@60Hz */
 	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
 		   1602, 1716, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 8 - 1440x240@60Hz */
 	{ DRM_MODE("1440x240", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
 		   1602, 1716, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 9 - 1440x240@60Hz */
 	{ DRM_MODE("1440x240", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1478,
 		   1602, 1716, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 10 - 2880x480i@60Hz */
 	{ DRM_MODE("2880x480i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 11 - 2880x480i@60Hz */
 	{ DRM_MODE("2880x480i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 12 - 2880x240@60Hz */
 	{ DRM_MODE("2880x240", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 13 - 2880x240@60Hz */
 	{ DRM_MODE("2880x240", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 14 - 1440x480@60Hz */
 	{ DRM_MODE("1440x480", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1472,
 		   1596, 1716, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 15 - 1440x480@60Hz */
 	{ DRM_MODE("1440x480", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1472,
 		   1596, 1716, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 16 - 1920x1080@60Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 17 - 720x576@50Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 18 - 720x576@50Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 19 - 1280x720@50Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 1720,
 		   1760, 1980, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 20 - 1920x1080i@50Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 21 - 1440x576i@50Hz */
 	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
 		   1590, 1728, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 22 - 1440x576i@50Hz */
 	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
 		   1590, 1728, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 23 - 1440x288@50Hz */
 	{ DRM_MODE("1440x288", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
 		   1590, 1728, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 24 - 1440x288@50Hz */
 	{ DRM_MODE("1440x288", DRM_MODE_TYPE_DRIVER, 27000, 1440, 1464,
 		   1590, 1728, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 25 - 2880x576i@50Hz */
 	{ DRM_MODE("2880x576i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 26 - 2880x576i@50Hz */
 	{ DRM_MODE("2880x576i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 27 - 2880x288@50Hz */
 	{ DRM_MODE("2880x288", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 28 - 2880x288@50Hz */
 	{ DRM_MODE("2880x288", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 29 - 1440x576@50Hz */
 	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
 		   1592, 1728, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 30 - 1440x576@50Hz */
 	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
 		   1592, 1728, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 31 - 1920x1080@50Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 32 - 1920x1080@24Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2558,
 		   2602, 2750, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 24, },
+	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 33 - 1920x1080@25Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 25, },
+	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 34 - 1920x1080@30Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 30, },
+	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 35 - 2880x480@60Hz */
 	{ DRM_MODE("2880x480", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2944,
 		   3192, 3432, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 36 - 2880x480@60Hz */
 	{ DRM_MODE("2880x480", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2944,
 		   3192, 3432, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 37 - 2880x576@50Hz */
 	{ DRM_MODE("2880x576", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2928,
 		   3184, 3456, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 38 - 2880x576@50Hz */
 	{ DRM_MODE("2880x576", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2928,
 		   3184, 3456, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 39 - 1920x1080i@50Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 72000, 1920, 1952,
 		   2120, 2304, 0, 1080, 1126, 1136, 1250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 40 - 1920x1080i@100Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 41 - 1280x720@100Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 148500, 1280, 1720,
 		   1760, 1980, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 42 - 720x576@100Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 43 - 720x576@100Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 44 - 1440x576i@100Hz */
 	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
 		   1590, 1728, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 45 - 1440x576i@100Hz */
 	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
 		   1590, 1728, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 46 - 1920x1080i@120Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 47 - 1280x720@120Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 148500, 1280, 1390,
 		   1430, 1650, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 48 - 720x480@120Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 54000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 49 - 720x480@120Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 54000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 50 - 1440x480i@120Hz */
 	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1478,
 		   1602, 1716, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 51 - 1440x480i@120Hz */
 	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1478,
 		   1602, 1716, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 52 - 720x576@200Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 108000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 200, },
+	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 53 - 720x576@200Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 108000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 200, },
+	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 54 - 1440x576i@200Hz */
 	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1464,
 		   1590, 1728, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 200, },
+	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 55 - 1440x576i@200Hz */
 	{ DRM_MODE("1440x576i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1464,
 		   1590, 1728, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 200, },
+	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 56 - 720x480@240Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 108000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 240, },
+	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 57 - 720x480@240Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 108000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 240, },
+	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 58 - 1440x480i@240 */
 	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1478,
 		   1602, 1716, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 240, },
+	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 59 - 1440x480i@240 */
 	{ DRM_MODE("1440x480i", DRM_MODE_TYPE_DRIVER, 108000, 1440, 1478,
 		   1602, 1716, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 240, },
+	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 60 - 1280x720@24Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 59400, 1280, 3040,
 		   3080, 3300, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 24, },
+	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 61 - 1280x720@25Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 3700,
 		   3740, 3960, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 25, },
+	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 62 - 1280x720@30Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 3040,
 		   3080, 3300, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 30, },
+	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 63 - 1920x1080@120Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 297000, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	 .vrefresh = 120, },
+	 .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 64 - 1920x1080@100Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 297000, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	 .vrefresh = 100, },
+	 .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 };
 
 /*
@@ -984,9 +989,13 @@ static const u8 edid_header[] = {
 	0x00, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00
 };
 
- /*
- * Sanity check the header of the base EDID block.  Return 8 if the header
- * is perfect, down to 0 if it's totally wrong.
+/**
+ * drm_edid_header_is_valid - sanity check the header of the base EDID block
+ * @raw_edid: pointer to raw base EDID block
+ *
+ * Sanity check the header of the base EDID block.
+ *
+ * Return: 8 if the header is perfect, down to 0 if it's totally wrong.
  */
 int drm_edid_header_is_valid(const u8 *raw_edid)
 {
@@ -1005,9 +1014,16 @@ module_param_named(edid_fixup, edid_fixup, int, 0400);
 MODULE_PARM_DESC(edid_fixup,
 		 "Minimum number of valid EDID header bytes (0-8, default 6)");
 
-/*
- * Sanity check the EDID block (base or extension).  Return 0 if the block
- * doesn't check out, or 1 if it's valid.
+/**
+ * drm_edid_block_valid - Sanity check the EDID block (base or extension)
+ * @raw_edid: pointer to raw EDID block
+ * @block: type of block to validate (0 for base, extension otherwise)
+ * @print_bad_edid: if true, dump bad EDID blocks to the console
+ *
+ * Validate a base or extension EDID block and optionally dump bad blocks to
+ * the console.
+ *
+ * Return: True if the block is valid, false otherwise.
  */
 bool drm_edid_block_valid(u8 *raw_edid, int block, bool print_bad_edid)
 {
@@ -1077,6 +1093,8 @@ EXPORT_SYMBOL(drm_edid_block_valid);
  * @edid: EDID data
  *
  * Sanity-check an entire EDID record (including extensions)
+ *
+ * Return: True if the EDID data is valid, false otherwise.
  */
 bool drm_edid_is_valid(struct edid *edid)
 {
@@ -1096,14 +1114,15 @@ EXPORT_SYMBOL(drm_edid_is_valid);
 
 #define DDC_SEGMENT_ADDR 0x30
 /**
- * Get EDID information via I2C.
+ * drm_do_probe_ddc_edid() - get EDID information via I2C
+ * @adapter: I2C device adaptor
+ * @buf: EDID data buffer to be filled
+ * @block: 128 byte EDID block to start fetching from
+ * @len: EDID data buffer length to fetch
  *
- * \param adapter : i2c device adaptor
- * \param buf     : EDID data buffer to be filled
- * \param len     : EDID data buffer length
- * \return 0 on success or -1 on failure.
+ * Try to fetch EDID information by calling I2C driver functions.
  *
- * Try to fetch EDID information by calling i2c driver function.
+ * Return: 0 on success or -1 on failure.
  */
 static int
 drm_do_probe_ddc_edid(struct i2c_adapter *adapter, unsigned char *buf,
@@ -1114,7 +1133,8 @@ drm_do_probe_ddc_edid(struct i2c_adapter *adapter, unsigned char *buf,
 	unsigned char xfers = segment ? 3 : 2;
 	int ret, retries = 5;
 
-	/* The core i2c driver will automatically retry the transfer if the
+	/*
+	 * The core I2C driver will automatically retry the transfer if the
 	 * adapter reports EAGAIN. However, we find that bit-banging transfers
 	 * are susceptible to errors under a heavily loaded machine and
 	 * generate spurious NAKs and timeouts. Retrying the transfer
@@ -1140,10 +1160,10 @@ drm_do_probe_ddc_edid(struct i2c_adapter *adapter, unsigned char *buf,
 			}
 		};
 
-	/*
-	 * Avoid sending the segment addr to not upset non-compliant ddc
-	 * monitors.
-	 */
+		/*
+		 * Avoid sending the segment addr to not upset non-compliant
+		 * DDC monitors.
+		 */
 		ret = i2c_transfer(adapter, &msgs[3 - xfers], xfers);
 
 		if (ret == -ENXIO) {
@@ -1212,7 +1232,7 @@ drm_do_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
 		if (i == 4 && print_bad_edid) {
 			dev_warn(connector->dev->dev,
 			 "%s: Ignoring invalid EDID block %d.\n",
-			 drm_get_connector_name(connector), j);
+			 connector->name, j);
 
 			connector->bad_edid_counter++;
 		}
@@ -1232,7 +1252,7 @@ drm_do_get_edid(struct drm_connector *connector, struct i2c_adapter *adapter)
 carp:
 	if (print_bad_edid) {
 		dev_warn(connector->dev->dev, "%s: EDID block %d invalid.\n",
-			 drm_get_connector_name(connector), j);
+			 connector->name, j);
 	}
 	connector->bad_edid_counter++;
 
@@ -1242,10 +1262,10 @@ out:
 }
 
 /**
- * Probe DDC presence.
+ * drm_probe_ddc() - probe DDC presence
+ * @adapter: I2C adapter to probe
  *
- * \param adapter : i2c device adaptor
- * \return 1 on success
+ * Return: True on success, false on failure.
  */
 bool
 drm_probe_ddc(struct i2c_adapter *adapter)
@@ -1259,12 +1279,12 @@ EXPORT_SYMBOL(drm_probe_ddc);
 /**
  * drm_get_edid - get EDID data, if available
  * @connector: connector we're probing
- * @adapter: i2c adapter to use for DDC
+ * @adapter: I2C adapter to use for DDC
  *
- * Poke the given i2c channel to grab EDID data if possible.  If found,
+ * Poke the given I2C channel to grab EDID data if possible.  If found,
  * attach it to the connector.
  *
- * Return edid data or NULL if we couldn't find any.
+ * Return: Pointer to valid EDID or NULL if we couldn't find any.
  */
 struct edid *drm_get_edid(struct drm_connector *connector,
 			  struct i2c_adapter *adapter)
@@ -1282,7 +1302,7 @@ EXPORT_SYMBOL(drm_get_edid);
  * drm_edid_duplicate - duplicate an EDID and the extensions
  * @edid: EDID to duplicate
  *
- * Return duplicate edid or NULL on allocation failure.
+ * Return: Pointer to duplicated EDID or NULL on allocation failure.
  */
 struct edid *drm_edid_duplicate(const struct edid *edid)
 {
@@ -1405,7 +1425,8 @@ mode_is_rb(const struct drm_display_mode *mode)
  * @rb: Mode reduced-blanking-ness
  *
  * Walk the DMT mode list looking for a match for the given parameters.
- * Return a newly allocated copy of the mode, or NULL if not found.
+ *
+ * Return: A newly allocated copy of the mode, or NULL if not found.
  */
 struct drm_display_mode *drm_mode_find_dmt(struct drm_device *dev,
 					   int hsize, int vsize, int fresh,
@@ -1586,15 +1607,16 @@ bad_std_timing(u8 a, u8 b)
 
 /**
  * drm_mode_std - convert standard mode info (width, height, refresh) into mode
+ * @connector: connector of for the EDID block
+ * @edid: EDID block to scan
  * @t: standard timing params
- * @timing_level: standard timing level
  *
  * Take the standard timing params (in this case width, aspect, and refresh)
  * and convert them into a real mode using CVT/GTF/DMT.
  */
 static struct drm_display_mode *
 drm_mode_std(struct drm_connector *connector, struct edid *edid,
-	     struct std_timing *t, int revision)
+	     struct std_timing *t)
 {
 	struct drm_device *dev = connector->dev;
 	struct drm_display_mode *m, *mode = NULL;
@@ -1615,7 +1637,7 @@ drm_mode_std(struct drm_connector *connector, struct edid *edid,
 	vrefresh_rate = vfreq + 60;
 	/* the vdisplay is calculated based on the aspect ratio */
 	if (aspect_ratio == 0) {
-		if (revision < 3)
+		if (edid->revision < 3)
 			vsize = hsize;
 		else
 			vsize = (hsize * 10) / 16;
@@ -2132,6 +2154,7 @@ do_established_modes(struct detailed_timing *timing, void *c)
 
 /**
  * add_established_modes - get est. modes from EDID and add them
+ * @connector: connector to add mode(s) to
  * @edid: EDID block to scan
  *
  * Each EDID block contains a bitmap of the supported "established modes" list
@@ -2182,8 +2205,7 @@ do_standard_modes(struct detailed_timing *timing, void *c)
 			struct drm_display_mode *newmode;
 
 			std = &data->data.timings[i];
-			newmode = drm_mode_std(connector, edid, std,
-					       edid->revision);
+			newmode = drm_mode_std(connector, edid, std);
 			if (newmode) {
 				drm_mode_probed_add(connector, newmode);
 				closure->modes++;
@@ -2194,6 +2216,7 @@ do_standard_modes(struct detailed_timing *timing, void *c)
 
 /**
  * add_standard_modes - get std. modes from EDID and add them
+ * @connector: connector to add mode(s) to
  * @edid: EDID block to scan
  *
  * Standard modes can be calculated using the appropriate standard (DMT,
@@ -2211,8 +2234,7 @@ add_standard_modes(struct drm_connector *connector, struct edid *edid)
 		struct drm_display_mode *newmode;
 
 		newmode = drm_mode_std(connector, edid,
-				       &edid->standard_timings[i],
-				       edid->revision);
+				       &edid->standard_timings[i]);
 		if (newmode) {
 			drm_mode_probed_add(connector, newmode);
 			modes++;
@@ -2415,7 +2437,7 @@ cea_mode_alternate_clock(const struct drm_display_mode *cea_mode)
  * drm_match_cea_mode - look for a CEA mode matching given mode
  * @to_match: display mode
  *
- * Returns the CEA Video ID (VIC) of the mode or 0 if it isn't a CEA-861
+ * Return: The CEA Video ID (VIC) of the mode or 0 if it isn't a CEA-861
  * mode.
  */
 u8 drm_match_cea_mode(const struct drm_display_mode *to_match)
@@ -2442,6 +2464,22 @@ u8 drm_match_cea_mode(const struct drm_display_mode *to_match)
 }
 EXPORT_SYMBOL(drm_match_cea_mode);
 
+/**
+ * drm_get_cea_aspect_ratio - get the picture aspect ratio corresponding to
+ * the input VIC from the CEA mode list
+ * @video_code: ID given to each of the CEA modes
+ *
+ * Returns picture aspect ratio
+ */
+enum hdmi_picture_aspect drm_get_cea_aspect_ratio(const u8 video_code)
+{
+	/* return picture aspect ratio for video_code - 1 to access the
+	 * right array element
+	*/
+	return edid_cea_modes[video_code-1].picture_aspect_ratio;
+}
+EXPORT_SYMBOL(drm_get_cea_aspect_ratio);
+
 /*
  * Calculate the alternate clock for HDMI modes (those from the HDMI vendor
  * specific block).
@@ -2562,25 +2600,43 @@ add_alternate_cea_modes(struct drm_connector *connector, struct edid *edid)
 	return modes;
 }
 
-static int
-do_cea_modes(struct drm_connector *connector, const u8 *db, u8 len)
+static struct drm_display_mode *
+drm_display_mode_from_vic_index(struct drm_connector *connector,
+				const u8 *video_db, u8 video_len,
+				u8 video_index)
 {
 	struct drm_device *dev = connector->dev;
-	const u8 *mode;
+	struct drm_display_mode *newmode;
 	u8 cea_mode;
-	int modes = 0;
 
-	for (mode = db; mode < db + len; mode++) {
-		cea_mode = (*mode & 127) - 1; /* CEA modes are numbered 1..127 */
-		if (cea_mode < ARRAY_SIZE(edid_cea_modes)) {
-			struct drm_display_mode *newmode;
-			newmode = drm_mode_duplicate(dev,
-						     &edid_cea_modes[cea_mode]);
-			if (newmode) {
-				newmode->vrefresh = 0;
-				drm_mode_probed_add(connector, newmode);
-				modes++;
-			}
+	if (video_db == NULL || video_index >= video_len)
+		return NULL;
+
+	/* CEA modes are numbered 1..127 */
+	cea_mode = (video_db[video_index] & 127) - 1;
+	if (cea_mode >= ARRAY_SIZE(edid_cea_modes))
+		return NULL;
+
+	newmode = drm_mode_duplicate(dev, &edid_cea_modes[cea_mode]);
+	if (!newmode)
+		return NULL;
+
+	newmode->vrefresh = 0;
+
+	return newmode;
+}
+
+static int
+do_cea_modes(struct drm_connector *connector, const u8 *db, u8 len)
+{
+	int i, modes = 0;
+
+	for (i = 0; i < len; i++) {
+		struct drm_display_mode *mode;
+		mode = drm_display_mode_from_vic_index(connector, db, len, i);
+		if (mode) {
+			drm_mode_probed_add(connector, mode);
+			modes++;
 		}
 	}
 
@@ -2674,21 +2730,13 @@ static int add_hdmi_mode(struct drm_connector *connector, u8 vic)
 static int add_3d_struct_modes(struct drm_connector *connector, u16 structure,
 			       const u8 *video_db, u8 video_len, u8 video_index)
 {
-	struct drm_device *dev = connector->dev;
 	struct drm_display_mode *newmode;
 	int modes = 0;
-	u8 cea_mode;
-
-	if (video_db == NULL || video_index >= video_len)
-		return 0;
-
-	/* CEA modes are numbered 1..127 */
-	cea_mode = (video_db[video_index] & 127) - 1;
-	if (cea_mode >= ARRAY_SIZE(edid_cea_modes))
-		return 0;
 
 	if (structure & (1 << 0)) {
-		newmode = drm_mode_duplicate(dev, &edid_cea_modes[cea_mode]);
+		newmode = drm_display_mode_from_vic_index(connector, video_db,
+							  video_len,
+							  video_index);
 		if (newmode) {
 			newmode->flags |= DRM_MODE_FLAG_3D_FRAME_PACKING;
 			drm_mode_probed_add(connector, newmode);
@@ -2696,7 +2744,9 @@ static int add_3d_struct_modes(struct drm_connector *connector, u16 structure,
 		}
 	}
 	if (structure & (1 << 6)) {
-		newmode = drm_mode_duplicate(dev, &edid_cea_modes[cea_mode]);
+		newmode = drm_display_mode_from_vic_index(connector, video_db,
+							  video_len,
+							  video_index);
 		if (newmode) {
 			newmode->flags |= DRM_MODE_FLAG_3D_TOP_AND_BOTTOM;
 			drm_mode_probed_add(connector, newmode);
@@ -2704,7 +2754,9 @@ static int add_3d_struct_modes(struct drm_connector *connector, u16 structure,
 		}
 	}
 	if (structure & (1 << 8)) {
-		newmode = drm_mode_duplicate(dev, &edid_cea_modes[cea_mode]);
+		newmode = drm_display_mode_from_vic_index(connector, video_db,
+							  video_len,
+							  video_index);
 		if (newmode) {
 			newmode->flags |= DRM_MODE_FLAG_3D_SIDE_BY_SIDE_HALF;
 			drm_mode_probed_add(connector, newmode);
@@ -2728,7 +2780,7 @@ static int
 do_hdmi_vsdb_modes(struct drm_connector *connector, const u8 *db, u8 len,
 		   const u8 *video_db, u8 video_len)
 {
-	int modes = 0, offset = 0, i, multi_present = 0;
+	int modes = 0, offset = 0, i, multi_present = 0, multi_len;
 	u8 vic_len, hdmi_3d_len = 0;
 	u16 mask;
 	u16 structure_all;
@@ -2774,32 +2826,84 @@ do_hdmi_vsdb_modes(struct drm_connector *connector, const u8 *db, u8 len,
 	}
 	offset += 1 + vic_len;
 
-	if (!(multi_present == 1 || multi_present == 2))
-		goto out;
+	if (multi_present == 1)
+		multi_len = 2;
+	else if (multi_present == 2)
+		multi_len = 4;
+	else
+		multi_len = 0;
 
-	if ((multi_present == 1 && len < (9 + offset)) ||
-	    (multi_present == 2 && len < (11 + offset)))
+	if (len < (8 + offset + hdmi_3d_len - 1))
 		goto out;
 
-	if ((multi_present == 1 && hdmi_3d_len < 2) ||
-	    (multi_present == 2 && hdmi_3d_len < 4))
+	if (hdmi_3d_len < multi_len)
 		goto out;
 
-	/* 3D_Structure_ALL */
-	structure_all = (db[8 + offset] << 8) | db[9 + offset];
+	if (multi_present == 1 || multi_present == 2) {
+		/* 3D_Structure_ALL */
+		structure_all = (db[8 + offset] << 8) | db[9 + offset];
 
-	/* check if 3D_MASK is present */
-	if (multi_present == 2)
-		mask = (db[10 + offset] << 8) | db[11 + offset];
-	else
-		mask = 0xffff;
+		/* check if 3D_MASK is present */
+		if (multi_present == 2)
+			mask = (db[10 + offset] << 8) | db[11 + offset];
+		else
+			mask = 0xffff;
+
+		for (i = 0; i < 16; i++) {
+			if (mask & (1 << i))
+				modes += add_3d_struct_modes(connector,
+						structure_all,
+						video_db,
+						video_len, i);
+		}
+	}
 
-	for (i = 0; i < 16; i++) {
-		if (mask & (1 << i))
-			modes += add_3d_struct_modes(connector,
-						     structure_all,
-						     video_db,
-						     video_len, i);
+	offset += multi_len;
+
+	for (i = 0; i < (hdmi_3d_len - multi_len); i++) {
+		int vic_index;
+		struct drm_display_mode *newmode = NULL;
+		unsigned int newflag = 0;
+		bool detail_present;
+
+		detail_present = ((db[8 + offset + i] & 0x0f) > 7);
+
+		if (detail_present && (i + 1 == hdmi_3d_len - multi_len))
+			break;
+
+		/* 2D_VIC_order_X */
+		vic_index = db[8 + offset + i] >> 4;
+
+		/* 3D_Structure_X */
+		switch (db[8 + offset + i] & 0x0f) {
+		case 0:
+			newflag = DRM_MODE_FLAG_3D_FRAME_PACKING;
+			break;
+		case 6:
+			newflag = DRM_MODE_FLAG_3D_TOP_AND_BOTTOM;
+			break;
+		case 8:
+			/* 3D_Detail_X */
+			if ((db[9 + offset + i] >> 4) == 1)
+				newflag = DRM_MODE_FLAG_3D_SIDE_BY_SIDE_HALF;
+			break;
+		}
+
+		if (newflag != 0) {
+			newmode = drm_display_mode_from_vic_index(connector,
+								  video_db,
+								  video_len,
+								  vic_index);
+
+			if (newmode) {
+				newmode->flags |= newflag;
+				drm_mode_probed_add(connector, newmode);
+				modes++;
+			}
+		}
+
+		if (detail_present)
+			i++;
 	}
 
 out:
@@ -2947,11 +3051,9 @@ monitor_name(struct detailed_timing *t, void *data)
  * @connector: connector corresponding to the HDMI/DP sink
  * @edid: EDID to parse
  *
- * Fill the ELD (EDID-Like Data) buffer for passing to the audio driver.
- * Some ELD fields are left to the graphics driver caller:
- * - Conn_Type
- * - HDCP
- * - Port_ID
+ * Fill the ELD (EDID-Like Data) buffer for passing to the audio driver. The
+ * Conn_Type, HDCP and Port_ID ELD fields are left for the graphics driver to
+ * fill in.
  */
 void drm_edid_to_eld(struct drm_connector *connector, struct edid *edid)
 {
@@ -3035,9 +3137,10 @@ EXPORT_SYMBOL(drm_edid_to_eld);
  * @sads: pointer that will be set to the extracted SADs
  *
  * Looks for CEA EDID block and extracts SADs (Short Audio Descriptors) from it.
- * Note: returned pointer needs to be kfreed
  *
- * Return number of found SADs or negative number on error.
+ * Note: The returned pointer needs to be freed using kfree().
+ *
+ * Return: The number of found SADs or negative number on error.
  */
 int drm_edid_to_sad(struct edid *edid, struct cea_sad **sads)
 {
@@ -3094,9 +3197,11 @@ EXPORT_SYMBOL(drm_edid_to_sad);
  * @sadb: pointer to the speaker block
  *
  * Looks for CEA EDID block and extracts the Speaker Allocation Data Block from it.
- * Note: returned pointer needs to be kfreed
  *
- * Return number of found Speaker Allocation Blocks or negative number on error.
+ * Note: The returned pointer needs to be freed using kfree().
+ *
+ * Return: The number of found Speaker Allocation Blocks or negative number on
+ * error.
  */
 int drm_edid_to_speaker_allocation(struct edid *edid, u8 **sadb)
 {
@@ -3128,10 +3233,9 @@ int drm_edid_to_speaker_allocation(struct edid *edid, u8 **sadb)
 
 			/* Speaker Allocation Data Block */
 			if (dbl == 3) {
-				*sadb = kmalloc(dbl, GFP_KERNEL);
+				*sadb = kmemdup(&db[1], dbl, GFP_KERNEL);
 				if (!*sadb)
 					return -ENOMEM;
-				memcpy(*sadb, &db[1], dbl);
 				count = dbl;
 				break;
 			}
@@ -3143,9 +3247,12 @@ int drm_edid_to_speaker_allocation(struct edid *edid, u8 **sadb)
 EXPORT_SYMBOL(drm_edid_to_speaker_allocation);
 
 /**
- * drm_av_sync_delay - HDMI/DP sink audio-video sync delay in millisecond
+ * drm_av_sync_delay - compute the HDMI/DP sink audio-video sync delay
  * @connector: connector associated with the HDMI/DP sink
  * @mode: the display mode
+ *
+ * Return: The HDMI/DP sink's audio-video sync delay in milliseconds or 0 if
+ * the sink doesn't support audio or video.
  */
 int drm_av_sync_delay(struct drm_connector *connector,
 		      struct drm_display_mode *mode)
@@ -3187,6 +3294,9 @@ EXPORT_SYMBOL(drm_av_sync_delay);
  *
  * It's possible for one encoder to be associated with multiple HDMI/DP sinks.
  * The policy is now hard coded to simply use the first HDMI/DP sink's ELD.
+ *
+ * Return: The connector associated with the first HDMI/DP sink that has ELD
+ * attached to it.
  */
 struct drm_connector *drm_select_eld(struct drm_encoder *encoder,
 				     struct drm_display_mode *mode)
@@ -3194,6 +3304,8 @@ struct drm_connector *drm_select_eld(struct drm_encoder *encoder,
 	struct drm_connector *connector;
 	struct drm_device *dev = encoder->dev;
 
+	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
 		if (connector->encoder == encoder && connector->eld[0])
 			return connector;
@@ -3203,11 +3315,12 @@ struct drm_connector *drm_select_eld(struct drm_encoder *encoder,
 EXPORT_SYMBOL(drm_select_eld);
 
 /**
- * drm_detect_hdmi_monitor - detect whether monitor is hdmi.
+ * drm_detect_hdmi_monitor - detect whether monitor is HDMI
  * @edid: monitor EDID information
  *
  * Parse the CEA extension according to CEA-861-B.
- * Return true if HDMI, false if not or unknown.
+ *
+ * Return: True if the monitor is HDMI, false if not or unknown.
  */
 bool drm_detect_hdmi_monitor(struct edid *edid)
 {
@@ -3237,6 +3350,7 @@ EXPORT_SYMBOL(drm_detect_hdmi_monitor);
 
 /**
  * drm_detect_monitor_audio - check monitor audio capability
+ * @edid: EDID block to scan
  *
  * Monitor should have CEA extension block.
  * If monitor has 'basic audio', but no CEA audio blocks, it's 'basic
@@ -3244,6 +3358,7 @@ EXPORT_SYMBOL(drm_detect_hdmi_monitor);
  * audio format, assume at least 'basic audio' support, even if 'basic
  * audio' is not defined in EDID.
  *
+ * Return: True if the monitor supports audio, false otherwise.
  */
 bool drm_detect_monitor_audio(struct edid *edid)
 {
@@ -3282,10 +3397,13 @@ EXPORT_SYMBOL(drm_detect_monitor_audio);
 
 /**
  * drm_rgb_quant_range_selectable - is RGB quantization range selectable?
+ * @edid: EDID block to scan
  *
  * Check whether the monitor reports the RGB quantization range selection
  * as supported. The AVI infoframe can then be used to inform the monitor
  * which quantization range (full or limited) is used.
+ *
+ * Return: True if the RGB quantization range is selectable, false otherwise.
  */
 bool drm_rgb_quant_range_selectable(struct edid *edid)
 {
@@ -3312,16 +3430,119 @@ bool drm_rgb_quant_range_selectable(struct edid *edid)
 EXPORT_SYMBOL(drm_rgb_quant_range_selectable);
 
 /**
+ * drm_assign_hdmi_deep_color_info - detect whether monitor supports
+ * hdmi deep color modes and update drm_display_info if so.
+ *
+ * @edid: monitor EDID information
+ * @info: Updated with maximum supported deep color bpc and color format
+ *        if deep color supported.
+ *
+ * Parse the CEA extension according to CEA-861-B.
+ * Return true if HDMI deep color supported, false if not or unknown.
+ */
+static bool drm_assign_hdmi_deep_color_info(struct edid *edid,
+                                            struct drm_display_info *info,
+                                            struct drm_connector *connector)
+{
+	u8 *edid_ext, *hdmi;
+	int i;
+	int start_offset, end_offset;
+	unsigned int dc_bpc = 0;
+
+	edid_ext = drm_find_cea_extension(edid);
+	if (!edid_ext)
+		return false;
+
+	if (cea_db_offsets(edid_ext, &start_offset, &end_offset))
+		return false;
+
+	/*
+	 * Because HDMI identifier is in Vendor Specific Block,
+	 * search it from all data blocks of CEA extension.
+	 */
+	for_each_cea_db(edid_ext, i, start_offset, end_offset) {
+		if (cea_db_is_hdmi_vsdb(&edid_ext[i])) {
+			/* HDMI supports at least 8 bpc */
+			info->bpc = 8;
+
+			hdmi = &edid_ext[i];
+			if (cea_db_payload_len(hdmi) < 6)
+				return false;
+
+			if (hdmi[6] & DRM_EDID_HDMI_DC_30) {
+				dc_bpc = 10;
+				info->edid_hdmi_dc_modes |= DRM_EDID_HDMI_DC_30;
+				DRM_DEBUG("%s: HDMI sink does deep color 30.\n",
+						  connector->name);
+			}
+
+			if (hdmi[6] & DRM_EDID_HDMI_DC_36) {
+				dc_bpc = 12;
+				info->edid_hdmi_dc_modes |= DRM_EDID_HDMI_DC_36;
+				DRM_DEBUG("%s: HDMI sink does deep color 36.\n",
+						  connector->name);
+			}
+
+			if (hdmi[6] & DRM_EDID_HDMI_DC_48) {
+				dc_bpc = 16;
+				info->edid_hdmi_dc_modes |= DRM_EDID_HDMI_DC_48;
+				DRM_DEBUG("%s: HDMI sink does deep color 48.\n",
+						  connector->name);
+			}
+
+			if (dc_bpc > 0) {
+				DRM_DEBUG("%s: Assigning HDMI sink color depth as %d bpc.\n",
+						  connector->name, dc_bpc);
+				info->bpc = dc_bpc;
+
+				/*
+				 * Deep color support mandates RGB444 support for all video
+				 * modes and forbids YCRCB422 support for all video modes per
+				 * HDMI 1.3 spec.
+				 */
+				info->color_formats = DRM_COLOR_FORMAT_RGB444;
+
+				/* YCRCB444 is optional according to spec. */
+				if (hdmi[6] & DRM_EDID_HDMI_DC_Y444) {
+					info->color_formats |= DRM_COLOR_FORMAT_YCRCB444;
+					DRM_DEBUG("%s: HDMI sink does YCRCB444 in deep color.\n",
+							  connector->name);
+				}
+
+				/*
+				 * Spec says that if any deep color mode is supported at all,
+				 * then deep color 36 bit must be supported.
+				 */
+				if (!(hdmi[6] & DRM_EDID_HDMI_DC_36)) {
+					DRM_DEBUG("%s: HDMI sink should do DC_36, but does not!\n",
+							  connector->name);
+				}
+
+				return true;
+			}
+			else {
+				DRM_DEBUG("%s: No deep color support on this HDMI sink.\n",
+						  connector->name);
+			}
+		}
+	}
+
+	return false;
+}
+
+/**
  * drm_add_display_info - pull display info out if present
  * @edid: EDID data
  * @info: display info (attached to connector)
+ * @connector: connector whose edid is used to build display info
  *
  * Grab any available display info and stuff it into the drm_display_info
  * structure that's part of the connector.  Useful for tracking bpp and
  * color spaces.
  */
 static void drm_add_display_info(struct edid *edid,
-				 struct drm_display_info *info)
+                                 struct drm_display_info *info,
+                                 struct drm_connector *connector)
 {
 	u8 *edid_ext;
 
@@ -3351,6 +3572,9 @@ static void drm_add_display_info(struct edid *edid,
 			info->color_formats |= DRM_COLOR_FORMAT_YCRCB422;
 	}
 
+	/* HDMI deep color modes supported? Assign to info, if so */
+	drm_assign_hdmi_deep_color_info(edid, info, connector);
+
 	/* Only defined for 1.4 with digital displays */
 	if (edid->revision < 4)
 		return;
@@ -3380,6 +3604,9 @@ static void drm_add_display_info(struct edid *edid,
 		break;
 	}
 
+	DRM_DEBUG("%s: Assigning EDID-1.4 digital sink color depth as %d bpc.\n",
+			  connector->name, info->bpc);
+
 	info->color_formats |= DRM_COLOR_FORMAT_RGB444;
 	if (edid->features & DRM_EDID_FEATURE_RGB_YCRCB444)
 		info->color_formats |= DRM_COLOR_FORMAT_YCRCB444;
@@ -3390,11 +3617,11 @@ static void drm_add_display_info(struct edid *edid,
 /**
  * drm_add_edid_modes - add modes from EDID data, if available
  * @connector: connector we're probing
- * @edid: edid data
+ * @edid: EDID data
  *
  * Add the specified modes to the connector's mode list.
  *
- * Return number of modes added or 0 if we couldn't find any.
+ * Return: The number of modes added or 0 if we couldn't find any.
  */
 int drm_add_edid_modes(struct drm_connector *connector, struct edid *edid)
 {
@@ -3406,7 +3633,7 @@ int drm_add_edid_modes(struct drm_connector *connector, struct edid *edid)
 	}
 	if (!drm_edid_is_valid(edid)) {
 		dev_warn(connector->dev->dev, "%s: EDID invalid.\n",
-			 drm_get_connector_name(connector));
+			 connector->name);
 		return 0;
 	}
 
@@ -3438,11 +3665,14 @@ int drm_add_edid_modes(struct drm_connector *connector, struct edid *edid)
 	if (quirks & (EDID_QUIRK_PREFER_LARGE_60 | EDID_QUIRK_PREFER_LARGE_75))
 		edid_fixup_preferred(connector, quirks);
 
-	drm_add_display_info(edid, &connector->display_info);
+	drm_add_display_info(edid, &connector->display_info, connector);
 
 	if (quirks & EDID_QUIRK_FORCE_8BPC)
 		connector->display_info.bpc = 8;
 
+	if (quirks & EDID_QUIRK_FORCE_12BPC)
+		connector->display_info.bpc = 12;
+
 	return num_modes;
 }
 EXPORT_SYMBOL(drm_add_edid_modes);
@@ -3456,7 +3686,7 @@ EXPORT_SYMBOL(drm_add_edid_modes);
  * Add the specified modes to the connector's mode list. Only when the
  * hdisplay/vdisplay is not beyond the given limit, it will be added.
  *
- * Return number of modes added or 0 if we couldn't find any.
+ * Return: The number of modes added or 0 if we couldn't find any.
  */
 int drm_add_modes_noedid(struct drm_connector *connector,
 			int hdisplay, int vdisplay)
@@ -3495,14 +3725,23 @@ int drm_add_modes_noedid(struct drm_connector *connector,
 }
 EXPORT_SYMBOL(drm_add_modes_noedid);
 
+/**
+ * drm_set_preferred_mode - Sets the preferred mode of a connector
+ * @connector: connector whose mode list should be processed
+ * @hpref: horizontal resolution of preferred mode
+ * @vpref: vertical resolution of preferred mode
+ *
+ * Marks a mode as preferred if it matches the resolution specified by @hpref
+ * and @vpref.
+ */
 void drm_set_preferred_mode(struct drm_connector *connector,
 			   int hpref, int vpref)
 {
 	struct drm_display_mode *mode;
 
 	list_for_each_entry(mode, &connector->probed_modes, head) {
-		if (drm_mode_width(mode)  == hpref &&
-		    drm_mode_height(mode) == vpref)
+		if (mode->hdisplay == hpref &&
+		    mode->vdisplay == vpref)
 			mode->type |= DRM_MODE_TYPE_PREFERRED;
 	}
 }
@@ -3514,7 +3753,7 @@ EXPORT_SYMBOL(drm_set_preferred_mode);
  * @frame: HDMI AVI infoframe
  * @mode: DRM display mode
  *
- * Returns 0 on success or a negative error code on failure.
+ * Return: 0 on success or a negative error code on failure.
  */
 int
 drm_hdmi_avi_infoframe_from_display_mode(struct hdmi_avi_infoframe *frame,
@@ -3535,7 +3774,14 @@ drm_hdmi_avi_infoframe_from_display_mode(struct hdmi_avi_infoframe *frame,
 	frame->video_code = drm_match_cea_mode(mode);
 
 	frame->picture_aspect = HDMI_PICTURE_ASPECT_NONE;
+
+	/* Populate picture aspect ratio from CEA mode list */
+	if (frame->video_code > 0)
+		frame->picture_aspect = drm_get_cea_aspect_ratio(
+						frame->video_code);
+
 	frame->active_aspect = HDMI_ACTIVE_ASPECT_PICTURE;
+	frame->scan_mode = HDMI_SCAN_MODE_UNDERSCAN;
 
 	return 0;
 }
@@ -3578,7 +3824,7 @@ s3d_structure_from_display_mode(const struct drm_display_mode *mode)
  * 4k or stereoscopic 3D mode. So when giving any other mode as input this
  * function will return -EINVAL, error that can be safely ignored.
  *
- * Returns 0 on success or a negative error code on failure.
+ * Return: 0 on success or a negative error code on failure.
  */
 int
 drm_hdmi_vendor_infoframe_from_display_mode(struct hdmi_vendor_infoframe *frame,
diff --git a/drivers/gpu/drm/drm_edid_load.c b/drivers/gpu/drm/drm_edid_load.c
index 9081172ef05..0a235fe61c9 100644
--- a/drivers/gpu/drm/drm_edid_load.c
+++ b/drivers/gpu/drm/drm_edid_load.c
@@ -31,8 +31,9 @@ module_param_string(edid_firmware, edid_firmware, sizeof(edid_firmware), 0644);
 MODULE_PARM_DESC(edid_firmware, "Do not probe monitor, use specified EDID blob "
 	"from built-in data or /lib/firmware instead. ");
 
-#define GENERIC_EDIDS 5
+#define GENERIC_EDIDS 6
 static const char *generic_edid_name[GENERIC_EDIDS] = {
+	"edid/800x600.bin",
 	"edid/1024x768.bin",
 	"edid/1280x1024.bin",
 	"edid/1600x1200.bin",
@@ -44,6 +45,24 @@ static const u8 generic_edid[GENERIC_EDIDS][128] = {
 	{
 	0x00, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00,
 	0x31, 0xd8, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+	0x05, 0x16, 0x01, 0x03, 0x6d, 0x1b, 0x14, 0x78,
+	0xea, 0x5e, 0xc0, 0xa4, 0x59, 0x4a, 0x98, 0x25,
+	0x20, 0x50, 0x54, 0x01, 0x00, 0x00, 0x45, 0x40,
+	0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01,
+	0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0xa0, 0x0f,
+	0x20, 0x00, 0x31, 0x58, 0x1c, 0x20, 0x28, 0x80,
+	0x14, 0x00, 0x15, 0xd0, 0x10, 0x00, 0x00, 0x1e,
+	0x00, 0x00, 0x00, 0xff, 0x00, 0x4c, 0x69, 0x6e,
+	0x75, 0x78, 0x20, 0x23, 0x30, 0x0a, 0x20, 0x20,
+	0x20, 0x20, 0x00, 0x00, 0x00, 0xfd, 0x00, 0x3b,
+	0x3d, 0x24, 0x26, 0x05, 0x00, 0x0a, 0x20, 0x20,
+	0x20, 0x20, 0x20, 0x20, 0x00, 0x00, 0x00, 0xfc,
+	0x00, 0x4c, 0x69, 0x6e, 0x75, 0x78, 0x20, 0x53,
+	0x56, 0x47, 0x41, 0x0a, 0x20, 0x20, 0x00, 0xc2,
+	},
+	{
+	0x00, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00,
+	0x31, 0xd8, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
 	0x05, 0x16, 0x01, 0x03, 0x6d, 0x23, 0x1a, 0x78,
 	0xea, 0x5e, 0xc0, 0xa4, 0x59, 0x4a, 0x98, 0x25,
 	0x20, 0x50, 0x54, 0x00, 0x08, 0x00, 0x61, 0x40,
@@ -141,7 +160,7 @@ static int edid_size(const u8 *edid, int data_size)
 	return (edid[0x7e] + 1) * EDID_LENGTH;
 }
 
-static u8 *edid_load(struct drm_connector *connector, const char *name,
+static void *edid_load(struct drm_connector *connector, const char *name,
 			const char *connector_name)
 {
 	const struct firmware *fw = NULL;
@@ -242,7 +261,7 @@ out:
 
 int drm_load_edid_firmware(struct drm_connector *connector)
 {
-	const char *connector_name = drm_get_connector_name(connector);
+	const char *connector_name = connector->name;
 	char *edidname = edid_firmware, *last, *colon;
 	int ret;
 	struct edid *edid;
@@ -263,7 +282,7 @@ int drm_load_edid_firmware(struct drm_connector *connector)
 	if (*last == '\n')
 		*last = '\0';
 
-	edid = (struct edid *) edid_load(connector, edidname, connector_name);
+	edid = edid_load(connector, edidname, connector_name);
 	if (IS_ERR_OR_NULL(edid))
 		return 0;
 
diff --git a/drivers/gpu/drm/drm_fb_cma_helper.c b/drivers/gpu/drm/drm_fb_cma_helper.c
index 61b5a47ad23..f27c883be39 100644
--- a/drivers/gpu/drm/drm_fb_cma_helper.c
+++ b/drivers/gpu/drm/drm_fb_cma_helper.c
@@ -429,13 +429,8 @@ EXPORT_SYMBOL_GPL(drm_fbdev_cma_fini);
  */
 void drm_fbdev_cma_restore_mode(struct drm_fbdev_cma *fbdev_cma)
 {
-	if (fbdev_cma) {
-		struct drm_device *dev = fbdev_cma->fb_helper.dev;
-
-		drm_modeset_lock_all(dev);
-		drm_fb_helper_restore_fbdev_mode(&fbdev_cma->fb_helper);
-		drm_modeset_unlock_all(dev);
-	}
+	if (fbdev_cma)
+		drm_fb_helper_restore_fbdev_mode_unlocked(&fbdev_cma->fb_helper);
 }
 EXPORT_SYMBOL_GPL(drm_fbdev_cma_restore_mode);
 
diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index 0a19401aff8..d5d8cea1a67 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -45,13 +45,13 @@ static LIST_HEAD(kernel_fb_helper_list);
  * DOC: fbdev helpers
  *
  * The fb helper functions are useful to provide an fbdev on top of a drm kernel
- * mode setting driver. They can be used mostly independantely from the crtc
+ * mode setting driver. They can be used mostly independently from the crtc
  * helper functions used by many drivers to implement the kernel mode setting
  * interfaces.
  *
  * Initialization is done as a three-step process with drm_fb_helper_init(),
  * drm_fb_helper_single_add_all_connectors() and drm_fb_helper_initial_config().
- * Drivers with fancier requirements than the default beheviour can override the
+ * Drivers with fancier requirements than the default behaviour can override the
  * second step with their own code.  Teardown is done with drm_fb_helper_fini().
  *
  * At runtime drivers should restore the fbdev console by calling
@@ -59,7 +59,7 @@ static LIST_HEAD(kernel_fb_helper_list);
  * should also notify the fb helper code from updates to the output
  * configuration by calling drm_fb_helper_hotplug_event(). For easier
  * integration with the output polling code in drm_crtc_helper.c the modeset
- * code proves a ->output_poll_changed callback.
+ * code provides a ->output_poll_changed callback.
  *
  * All other functions exported by the fb helper library can be used to
  * implement the fbdev driver interface by the driver.
@@ -120,7 +120,7 @@ static int drm_fb_helper_parse_command_line(struct drm_fb_helper *fb_helper)
 		mode = &fb_helper_conn->cmdline_mode;
 
 		/* do something on return - turn off connector maybe */
-		if (fb_get_options(drm_get_connector_name(connector), &option))
+		if (fb_get_options(connector->name, &option))
 			continue;
 
 		if (drm_mode_parse_command_line_for_connector(option,
@@ -142,12 +142,12 @@ static int drm_fb_helper_parse_command_line(struct drm_fb_helper *fb_helper)
 				}
 
 				DRM_INFO("forcing %s connector %s\n",
-					 drm_get_connector_name(connector), s);
+					 connector->name, s);
 				connector->force = mode->force;
 			}
 
 			DRM_DEBUG_KMS("cmdline mode for connector %s %dx%d@%dHz%s%s%s\n",
-				      drm_get_connector_name(connector),
+				      connector->name,
 				      mode->xres, mode->yres,
 				      mode->refresh_specified ? mode->refresh : 60,
 				      mode->rb ? " reduced blanking" : "",
@@ -232,7 +232,7 @@ static struct drm_framebuffer *drm_mode_config_fb(struct drm_crtc *crtc)
 
 	list_for_each_entry(c, &dev->mode_config.crtc_list, head) {
 		if (crtc->base.id == c->base.id)
-			return c->fb;
+			return c->primary->fb;
 	}
 
 	return NULL;
@@ -273,15 +273,7 @@ int drm_fb_helper_debug_leave(struct fb_info *info)
 }
 EXPORT_SYMBOL(drm_fb_helper_debug_leave);
 
-/**
- * drm_fb_helper_restore_fbdev_mode - restore fbdev configuration
- * @fb_helper: fbcon to restore
- *
- * This should be called from driver's drm ->lastclose callback
- * when implementing an fbcon on top of kms using this helper. This ensures that
- * the user isn't greeted with a black screen when e.g. X dies.
- */
-bool drm_fb_helper_restore_fbdev_mode(struct drm_fb_helper *fb_helper)
+static bool restore_fbdev_mode(struct drm_fb_helper *fb_helper)
 {
 	struct drm_device *dev = fb_helper->dev;
 	struct drm_plane *plane;
@@ -291,7 +283,8 @@ bool drm_fb_helper_restore_fbdev_mode(struct drm_fb_helper *fb_helper)
 	drm_warn_on_modeset_not_all_locked(dev);
 
 	list_for_each_entry(plane, &dev->mode_config.plane_list, head)
-		drm_plane_force_disable(plane);
+		if (plane->type != DRM_PLANE_TYPE_PRIMARY)
+			drm_plane_force_disable(plane);
 
 	for (i = 0; i < fb_helper->crtc_count; i++) {
 		struct drm_mode_set *mode_set = &fb_helper->crtc_info[i].mode_set;
@@ -310,7 +303,40 @@ bool drm_fb_helper_restore_fbdev_mode(struct drm_fb_helper *fb_helper)
 	}
 	return error;
 }
-EXPORT_SYMBOL(drm_fb_helper_restore_fbdev_mode);
+/**
+ * drm_fb_helper_restore_fbdev_mode - restore fbdev configuration
+ * @fb_helper: fbcon to restore
+ *
+ * This should be called from driver's drm ->lastclose callback
+ * when implementing an fbcon on top of kms using this helper. This ensures that
+ * the user isn't greeted with a black screen when e.g. X dies.
+ *
+ * Use this variant if you need to bypass locking (panic), or already
+ * hold all modeset locks.  Otherwise use drm_fb_helper_restore_fbdev_mode_unlocked()
+ */
+static bool drm_fb_helper_restore_fbdev_mode(struct drm_fb_helper *fb_helper)
+{
+	return restore_fbdev_mode(fb_helper);
+}
+
+/**
+ * drm_fb_helper_restore_fbdev_mode_unlocked - restore fbdev configuration
+ * @fb_helper: fbcon to restore
+ *
+ * This should be called from driver's drm ->lastclose callback
+ * when implementing an fbcon on top of kms using this helper. This ensures that
+ * the user isn't greeted with a black screen when e.g. X dies.
+ */
+bool drm_fb_helper_restore_fbdev_mode_unlocked(struct drm_fb_helper *fb_helper)
+{
+	struct drm_device *dev = fb_helper->dev;
+	bool ret;
+	drm_modeset_lock_all(dev);
+	ret = restore_fbdev_mode(fb_helper);
+	drm_modeset_unlock_all(dev);
+	return ret;
+}
+EXPORT_SYMBOL(drm_fb_helper_restore_fbdev_mode_unlocked);
 
 /*
  * restore fbcon display for all kms driver's using this helper, used for sysrq
@@ -325,12 +351,25 @@ static bool drm_fb_helper_force_kernel_mode(void)
 		return false;
 
 	list_for_each_entry(helper, &kernel_fb_helper_list, kernel_fb_list) {
-		if (helper->dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+		struct drm_device *dev = helper->dev;
+
+		if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 			continue;
 
+		/* NOTE: we use lockless flag below to avoid grabbing other
+		 * modeset locks.  So just trylock the underlying mutex
+		 * directly:
+		 */
+		if (!mutex_trylock(&dev->mode_config.mutex)) {
+			error = true;
+			continue;
+		}
+
 		ret = drm_fb_helper_restore_fbdev_mode(helper);
 		if (ret)
 			error = true;
+
+		mutex_unlock(&dev->mode_config.mutex);
 	}
 	return error;
 }
@@ -359,15 +398,21 @@ static bool drm_fb_helper_is_bound(struct drm_fb_helper *fb_helper)
 	struct drm_crtc *crtc;
 	int bound = 0, crtcs_bound = 0;
 
+	/* Sometimes user space wants everything disabled, so don't steal the
+	 * display if there's a master. */
+	if (dev->primary->master)
+		return false;
+
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
-		if (crtc->fb)
+		if (crtc->primary->fb)
 			crtcs_bound++;
-		if (crtc->fb == fb_helper->fb)
+		if (crtc->primary->fb == fb_helper->fb)
 			bound++;
 	}
 
 	if (bound < crtcs_bound)
 		return false;
+
 	return true;
 }
 
@@ -510,6 +555,9 @@ int drm_fb_helper_init(struct drm_device *dev,
 	struct drm_crtc *crtc;
 	int i;
 
+	if (!max_conn_count)
+		return -EINVAL;
+
 	fb_helper->dev = dev;
 
 	INIT_LIST_HEAD(&fb_helper->kernel_fb_list);
@@ -801,25 +849,14 @@ EXPORT_SYMBOL(drm_fb_helper_check_var);
 int drm_fb_helper_set_par(struct fb_info *info)
 {
 	struct drm_fb_helper *fb_helper = info->par;
-	struct drm_device *dev = fb_helper->dev;
 	struct fb_var_screeninfo *var = &info->var;
-	int ret;
-	int i;
 
 	if (var->pixclock != 0) {
 		DRM_ERROR("PIXEL CLOCK SET\n");
 		return -EINVAL;
 	}
 
-	drm_modeset_lock_all(dev);
-	for (i = 0; i < fb_helper->crtc_count; i++) {
-		ret = drm_mode_set_config_internal(&fb_helper->crtc_info[i].mode_set);
-		if (ret) {
-			drm_modeset_unlock_all(dev);
-			return ret;
-		}
-	}
-	drm_modeset_unlock_all(dev);
+	drm_fb_helper_restore_fbdev_mode_unlocked(fb_helper);
 
 	if (fb_helper->delayed_hotplug) {
 		fb_helper->delayed_hotplug = false;
@@ -1130,19 +1167,20 @@ static int drm_fb_helper_probe_connector_modes(struct drm_fb_helper *fb_helper,
 	return count;
 }
 
-static struct drm_display_mode *drm_has_preferred_mode(struct drm_fb_helper_connector *fb_connector, int width, int height)
+struct drm_display_mode *drm_has_preferred_mode(struct drm_fb_helper_connector *fb_connector, int width, int height)
 {
 	struct drm_display_mode *mode;
 
 	list_for_each_entry(mode, &fb_connector->connector->modes, head) {
-		if (drm_mode_width(mode) > width ||
-		    drm_mode_height(mode) > height)
+		if (mode->hdisplay > width ||
+		    mode->vdisplay > height)
 			continue;
 		if (mode->type & DRM_MODE_TYPE_PREFERRED)
 			return mode;
 	}
 	return NULL;
 }
+EXPORT_SYMBOL(drm_has_preferred_mode);
 
 static bool drm_has_cmdline_mode(struct drm_fb_helper_connector *fb_connector)
 {
@@ -1151,11 +1189,12 @@ static bool drm_has_cmdline_mode(struct drm_fb_helper_connector *fb_connector)
 	return cmdline_mode->specified;
 }
 
-static struct drm_display_mode *drm_pick_cmdline_mode(struct drm_fb_helper_connector *fb_helper_conn,
+struct drm_display_mode *drm_pick_cmdline_mode(struct drm_fb_helper_connector *fb_helper_conn,
 						      int width, int height)
 {
 	struct drm_cmdline_mode *cmdline_mode;
 	struct drm_display_mode *mode = NULL;
+	bool prefer_non_interlace;
 
 	cmdline_mode = &fb_helper_conn->cmdline_mode;
 	if (cmdline_mode->specified == false)
@@ -1167,6 +1206,8 @@ static struct drm_display_mode *drm_pick_cmdline_mode(struct drm_fb_helper_conne
 	if (cmdline_mode->rb || cmdline_mode->margins)
 		goto create_mode;
 
+	prefer_non_interlace = !cmdline_mode->interlace;
+ again:
 	list_for_each_entry(mode, &fb_helper_conn->connector->modes, head) {
 		/* check width/height */
 		if (mode->hdisplay != cmdline_mode->xres ||
@@ -1181,16 +1222,25 @@ static struct drm_display_mode *drm_pick_cmdline_mode(struct drm_fb_helper_conne
 		if (cmdline_mode->interlace) {
 			if (!(mode->flags & DRM_MODE_FLAG_INTERLACE))
 				continue;
+		} else if (prefer_non_interlace) {
+			if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+				continue;
 		}
 		return mode;
 	}
 
+	if (prefer_non_interlace) {
+		prefer_non_interlace = false;
+		goto again;
+	}
+
 create_mode:
 	mode = drm_mode_create_from_cmdline_mode(fb_helper_conn->connector->dev,
 						 cmdline_mode);
 	list_add(&mode->head, &fb_helper_conn->connector->modes);
 	return mode;
 }
+EXPORT_SYMBOL(drm_pick_cmdline_mode);
 
 static bool drm_connector_enabled(struct drm_connector *connector, bool strict)
 {
@@ -1533,9 +1583,11 @@ bool drm_fb_helper_initial_config(struct drm_fb_helper *fb_helper, int bpp_sel)
 
 	drm_fb_helper_parse_command_line(fb_helper);
 
+	mutex_lock(&dev->mode_config.mutex);
 	count = drm_fb_helper_probe_connector_modes(fb_helper,
 						    dev->mode_config.max_width,
 						    dev->mode_config.max_height);
+	mutex_unlock(&dev->mode_config.mutex);
 	/*
 	 * we shouldn't end up with no modes here.
 	 */
diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c
index c5b929c3f77..021fe5d11df 100644
--- a/drivers/gpu/drm/drm_fops.c
+++ b/drivers/gpu/drm/drm_fops.c
@@ -39,12 +39,11 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 
-/* from BKL pushdown: note that nothing else serializes idr_find() */
+/* from BKL pushdown */
 DEFINE_MUTEX(drm_global_mutex);
 EXPORT_SYMBOL(drm_global_mutex);
 
-static int drm_open_helper(struct inode *inode, struct file *filp,
-			   struct drm_device * dev);
+static int drm_open_helper(struct file *filp, struct drm_minor *minor);
 
 static int drm_setup(struct drm_device * dev)
 {
@@ -79,38 +78,23 @@ static int drm_setup(struct drm_device * dev)
  */
 int drm_open(struct inode *inode, struct file *filp)
 {
-	struct drm_device *dev = NULL;
-	int minor_id = iminor(inode);
+	struct drm_device *dev;
 	struct drm_minor *minor;
-	int retcode = 0;
+	int retcode;
 	int need_setup = 0;
-	struct address_space *old_mapping;
-	struct address_space *old_imapping;
-
-	minor = idr_find(&drm_minors_idr, minor_id);
-	if (!minor)
-		return -ENODEV;
-
-	if (!(dev = minor->dev))
-		return -ENODEV;
 
-	if (drm_device_is_unplugged(dev))
-		return -ENODEV;
+	minor = drm_minor_acquire(iminor(inode));
+	if (IS_ERR(minor))
+		return PTR_ERR(minor);
 
+	dev = minor->dev;
 	if (!dev->open_count++)
 		need_setup = 1;
-	mutex_lock(&dev->struct_mutex);
-	old_imapping = inode->i_mapping;
-	old_mapping = dev->dev_mapping;
-	if (old_mapping == NULL)
-		dev->dev_mapping = &inode->i_data;
-	/* ihold ensures nobody can remove inode with our i_data */
-	ihold(container_of(dev->dev_mapping, struct inode, i_data));
-	inode->i_mapping = dev->dev_mapping;
-	filp->f_mapping = dev->dev_mapping;
-	mutex_unlock(&dev->struct_mutex);
 
-	retcode = drm_open_helper(inode, filp, dev);
+	/* share address_space across all char-devs of a single device */
+	filp->f_mapping = dev->anon_inode->i_mapping;
+
+	retcode = drm_open_helper(filp, minor);
 	if (retcode)
 		goto err_undo;
 	if (need_setup) {
@@ -121,13 +105,8 @@ int drm_open(struct inode *inode, struct file *filp)
 	return 0;
 
 err_undo:
-	mutex_lock(&dev->struct_mutex);
-	filp->f_mapping = old_imapping;
-	inode->i_mapping = old_imapping;
-	iput(container_of(dev->dev_mapping, struct inode, i_data));
-	dev->dev_mapping = old_mapping;
-	mutex_unlock(&dev->struct_mutex);
 	dev->open_count--;
+	drm_minor_release(minor);
 	return retcode;
 }
 EXPORT_SYMBOL(drm_open);
@@ -143,33 +122,30 @@ EXPORT_SYMBOL(drm_open);
  */
 int drm_stub_open(struct inode *inode, struct file *filp)
 {
-	struct drm_device *dev = NULL;
+	struct drm_device *dev;
 	struct drm_minor *minor;
-	int minor_id = iminor(inode);
 	int err = -ENODEV;
 	const struct file_operations *new_fops;
 
 	DRM_DEBUG("\n");
 
 	mutex_lock(&drm_global_mutex);
-	minor = idr_find(&drm_minors_idr, minor_id);
-	if (!minor)
-		goto out;
-
-	if (!(dev = minor->dev))
-		goto out;
-
-	if (drm_device_is_unplugged(dev))
-		goto out;
+	minor = drm_minor_acquire(iminor(inode));
+	if (IS_ERR(minor))
+		goto out_unlock;
 
+	dev = minor->dev;
 	new_fops = fops_get(dev->driver->fops);
 	if (!new_fops)
-		goto out;
+		goto out_release;
 
 	replace_fops(filp, new_fops);
 	if (filp->f_op->open)
 		err = filp->f_op->open(inode, filp);
-out:
+
+out_release:
+	drm_minor_release(minor);
+out_unlock:
 	mutex_unlock(&drm_global_mutex);
 	return err;
 }
@@ -194,18 +170,16 @@ static int drm_cpu_valid(void)
 /**
  * Called whenever a process opens /dev/drm.
  *
- * \param inode device inode.
  * \param filp file pointer.
- * \param dev device.
+ * \param minor acquired minor-object.
  * \return zero on success or a negative number on failure.
  *
  * Creates and initializes a drm_file structure for the file private data in \p
  * filp and add it into the double linked list in \p dev.
  */
-static int drm_open_helper(struct inode *inode, struct file *filp,
-			   struct drm_device * dev)
+static int drm_open_helper(struct file *filp, struct drm_minor *minor)
 {
-	int minor_id = iminor(inode);
+	struct drm_device *dev = minor->dev;
 	struct drm_file *priv;
 	int ret;
 
@@ -216,7 +190,7 @@ static int drm_open_helper(struct inode *inode, struct file *filp,
 	if (dev->switch_power_state != DRM_SWITCH_POWER_ON && dev->switch_power_state != DRM_SWITCH_POWER_DYNAMIC_OFF)
 		return -EINVAL;
 
-	DRM_DEBUG("pid = %d, minor = %d\n", task_pid_nr(current), minor_id);
+	DRM_DEBUG("pid = %d, minor = %d\n", task_pid_nr(current), minor->index);
 
 	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
 	if (!priv)
@@ -226,13 +200,8 @@ static int drm_open_helper(struct inode *inode, struct file *filp,
 	priv->filp = filp;
 	priv->uid = current_euid();
 	priv->pid = get_pid(task_pid(current));
-	priv->minor = idr_find(&drm_minors_idr, minor_id);
-	if (!priv->minor) {
-		ret = -ENODEV;
-		goto out_put_pid;
-	}
+	priv->minor = minor;
 
-	priv->ioctl_count = 0;
 	/* for compatibility root is always authenticated */
 	priv->always_authenticated = capable(CAP_SYS_ADMIN);
 	priv->authenticated = priv->always_authenticated;
@@ -259,12 +228,11 @@ static int drm_open_helper(struct inode *inode, struct file *filp,
 
 	/* if there is no current master make this fd it, but do not create
 	 * any master object for render clients */
-	mutex_lock(&dev->struct_mutex);
-	if (!priv->minor->master && !drm_is_render_client(priv)) {
+	mutex_lock(&dev->master_mutex);
+	if (drm_is_primary_client(priv) && !priv->minor->master) {
 		/* create a new master */
 		priv->minor->master = drm_master_create(priv->minor);
 		if (!priv->minor->master) {
-			mutex_unlock(&dev->struct_mutex);
 			ret = -ENOMEM;
 			goto out_close;
 		}
@@ -272,37 +240,31 @@ static int drm_open_helper(struct inode *inode, struct file *filp,
 		priv->is_master = 1;
 		/* take another reference for the copy in the local file priv */
 		priv->master = drm_master_get(priv->minor->master);
-
 		priv->authenticated = 1;
 
-		mutex_unlock(&dev->struct_mutex);
 		if (dev->driver->master_create) {
 			ret = dev->driver->master_create(dev, priv->master);
 			if (ret) {
-				mutex_lock(&dev->struct_mutex);
 				/* drop both references if this fails */
 				drm_master_put(&priv->minor->master);
 				drm_master_put(&priv->master);
-				mutex_unlock(&dev->struct_mutex);
 				goto out_close;
 			}
 		}
-		mutex_lock(&dev->struct_mutex);
 		if (dev->driver->master_set) {
 			ret = dev->driver->master_set(dev, priv, true);
 			if (ret) {
 				/* drop both references if this fails */
 				drm_master_put(&priv->minor->master);
 				drm_master_put(&priv->master);
-				mutex_unlock(&dev->struct_mutex);
 				goto out_close;
 			}
 		}
-	} else if (!drm_is_render_client(priv)) {
+	} else if (drm_is_primary_client(priv)) {
 		/* get a reference to the master */
 		priv->master = drm_master_get(priv->minor->master);
 	}
-	mutex_unlock(&dev->struct_mutex);
+	mutex_unlock(&dev->master_mutex);
 
 	mutex_lock(&dev->struct_mutex);
 	list_add(&priv->lhead, &dev->filelist);
@@ -320,7 +282,8 @@ static int drm_open_helper(struct inode *inode, struct file *filp,
 			pci_dev_put(pci_dev);
 		}
 		if (!dev->hose) {
-			struct pci_bus *b = pci_bus_b(pci_root_buses.next);
+			struct pci_bus *b = list_entry(pci_root_buses.next,
+				struct pci_bus, node);
 			if (b)
 				dev->hose = b->sysdata;
 		}
@@ -330,6 +293,7 @@ static int drm_open_helper(struct inode *inode, struct file *filp,
 	return 0;
 
 out_close:
+	mutex_unlock(&dev->master_mutex);
 	if (dev->driver->postclose)
 		dev->driver->postclose(dev, priv);
 out_prime_destroy:
@@ -337,7 +301,6 @@ out_prime_destroy:
 		drm_prime_destroy_file_private(&priv->prime);
 	if (dev->driver->driver_features & DRIVER_GEM)
 		drm_gem_release(dev, priv);
-out_put_pid:
 	put_pid(priv->pid);
 	kfree(priv);
 	filp->private_data = NULL;
@@ -392,9 +355,6 @@ static void drm_legacy_dev_reinit(struct drm_device *dev)
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return;
 
-	atomic_set(&dev->ioctl_count, 0);
-	atomic_set(&dev->vma_count, 0);
-
 	dev->sigdata.lock = NULL;
 
 	dev->context_flag = 0;
@@ -438,7 +398,6 @@ int drm_lastclose(struct drm_device * dev)
 
 	drm_legacy_dma_takedown(dev);
 
-	dev->dev_mapping = NULL;
 	mutex_unlock(&dev->struct_mutex);
 
 	drm_legacy_dev_reinit(dev);
@@ -462,7 +421,8 @@ int drm_lastclose(struct drm_device * dev)
 int drm_release(struct inode *inode, struct file *filp)
 {
 	struct drm_file *file_priv = filp->private_data;
-	struct drm_device *dev = file_priv->minor->dev;
+	struct drm_minor *minor = file_priv->minor;
+	struct drm_device *dev = minor->dev;
 	int retcode = 0;
 
 	mutex_lock(&drm_global_mutex);
@@ -478,7 +438,7 @@ int drm_release(struct inode *inode, struct file *filp)
 
 	DRM_DEBUG("pid = %d, device = 0x%lx, open_count = %d\n",
 		  task_pid_nr(current),
-		  (long)old_encode_dev(file_priv->minor->device),
+		  (long)old_encode_dev(file_priv->minor->kdev->devt),
 		  dev->open_count);
 
 	/* Release any auth tokens that might point to this file_priv,
@@ -521,11 +481,13 @@ int drm_release(struct inode *inode, struct file *filp)
 	}
 	mutex_unlock(&dev->ctxlist_mutex);
 
-	mutex_lock(&dev->struct_mutex);
+	mutex_lock(&dev->master_mutex);
 
 	if (file_priv->is_master) {
 		struct drm_master *master = file_priv->master;
 		struct drm_file *temp;
+
+		mutex_lock(&dev->struct_mutex);
 		list_for_each_entry(temp, &dev->filelist, lhead) {
 			if ((temp->master == file_priv->master) &&
 			    (temp != file_priv))
@@ -544,6 +506,7 @@ int drm_release(struct inode *inode, struct file *filp)
 			master->lock.file_priv = NULL;
 			wake_up_interruptible_all(&master->lock.lock_queue);
 		}
+		mutex_unlock(&dev->struct_mutex);
 
 		if (file_priv->minor->master == file_priv->master) {
 			/* drop the reference held my the minor */
@@ -553,13 +516,13 @@ int drm_release(struct inode *inode, struct file *filp)
 		}
 	}
 
-	BUG_ON(dev->dev_mapping == NULL);
-	iput(container_of(dev->dev_mapping, struct inode, i_data));
-
-	/* drop the reference held my the file priv */
+	/* drop the master reference held by the file priv */
 	if (file_priv->master)
 		drm_master_put(&file_priv->master);
 	file_priv->is_master = 0;
+	mutex_unlock(&dev->master_mutex);
+
+	mutex_lock(&dev->struct_mutex);
 	list_del(&file_priv->lhead);
 	mutex_unlock(&dev->struct_mutex);
 
@@ -578,17 +541,14 @@ int drm_release(struct inode *inode, struct file *filp)
 	 */
 
 	if (!--dev->open_count) {
-		if (atomic_read(&dev->ioctl_count)) {
-			DRM_ERROR("Device busy: %d\n",
-				  atomic_read(&dev->ioctl_count));
-			retcode = -EBUSY;
-		} else
-			retcode = drm_lastclose(dev);
+		retcode = drm_lastclose(dev);
 		if (drm_device_is_unplugged(dev))
 			drm_put_dev(dev);
 	}
 	mutex_unlock(&drm_global_mutex);
 
+	drm_minor_release(minor);
+
 	return retcode;
 }
 EXPORT_SYMBOL(drm_release);
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
index 4761adedad2..f7d71190aad 100644
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -85,25 +85,25 @@
 #endif
 
 /**
- * Initialize the GEM device fields
+ * drm_gem_init - Initialize the GEM device fields
+ * @dev: drm_devic structure to initialize
  */
-
 int
 drm_gem_init(struct drm_device *dev)
 {
-	struct drm_gem_mm *mm;
+	struct drm_vma_offset_manager *vma_offset_manager;
 
 	mutex_init(&dev->object_name_lock);
 	idr_init(&dev->object_name_idr);
 
-	mm = kzalloc(sizeof(struct drm_gem_mm), GFP_KERNEL);
-	if (!mm) {
+	vma_offset_manager = kzalloc(sizeof(*vma_offset_manager), GFP_KERNEL);
+	if (!vma_offset_manager) {
 		DRM_ERROR("out of memory\n");
 		return -ENOMEM;
 	}
 
-	dev->mm_private = mm;
-	drm_vma_offset_manager_init(&mm->vma_manager,
+	dev->vma_offset_manager = vma_offset_manager;
+	drm_vma_offset_manager_init(vma_offset_manager,
 				    DRM_FILE_PAGE_OFFSET_START,
 				    DRM_FILE_PAGE_OFFSET_SIZE);
 
@@ -113,14 +113,18 @@ drm_gem_init(struct drm_device *dev)
 void
 drm_gem_destroy(struct drm_device *dev)
 {
-	struct drm_gem_mm *mm = dev->mm_private;
 
-	drm_vma_offset_manager_destroy(&mm->vma_manager);
-	kfree(mm);
-	dev->mm_private = NULL;
+	drm_vma_offset_manager_destroy(dev->vma_offset_manager);
+	kfree(dev->vma_offset_manager);
+	dev->vma_offset_manager = NULL;
 }
 
 /**
+ * drm_gem_object_init - initialize an allocated shmem-backed GEM object
+ * @dev: drm_device the object should be initialized for
+ * @obj: drm_gem_object to initialize
+ * @size: object size
+ *
  * Initialize an already allocated GEM object of the specified size with
  * shmfs backing store.
  */
@@ -129,11 +133,12 @@ int drm_gem_object_init(struct drm_device *dev,
 {
 	struct file *filp;
 
+	drm_gem_private_object_init(dev, obj, size);
+
 	filp = shmem_file_setup("drm mm object", size, VM_NORESERVE);
 	if (IS_ERR(filp))
 		return PTR_ERR(filp);
 
-	drm_gem_private_object_init(dev, obj, size);
 	obj->filp = filp;
 
 	return 0;
@@ -141,6 +146,11 @@ int drm_gem_object_init(struct drm_device *dev,
 EXPORT_SYMBOL(drm_gem_object_init);
 
 /**
+ * drm_gem_object_init - initialize an allocated private GEM object
+ * @dev: drm_device the object should be initialized for
+ * @obj: drm_gem_object to initialize
+ * @size: object size
+ *
  * Initialize an already allocated GEM object of the specified size with
  * no GEM provided backing store. Instead the caller is responsible for
  * backing the object and handling it.
@@ -175,12 +185,10 @@ drm_gem_remove_prime_handles(struct drm_gem_object *obj, struct drm_file *filp)
 	mutex_unlock(&filp->prime.lock);
 }
 
-static void drm_gem_object_ref_bug(struct kref *list_kref)
-{
-	BUG();
-}
-
 /**
+ * drm_gem_object_free - release resources bound to userspace handles
+ * @obj: GEM object to clean up.
+ *
  * Called after the last handle to the object has been closed
  *
  * Removes any name for the object. Note that this must be
@@ -195,13 +203,6 @@ static void drm_gem_object_handle_free(struct drm_gem_object *obj)
 	if (obj->name) {
 		idr_remove(&dev->object_name_idr, obj->name);
 		obj->name = 0;
-		/*
-		 * The object name held a reference to this object, drop
-		 * that now.
-		*
-		* This cannot be the last reference, since the handle holds one too.
-		 */
-		kref_put(&obj->refcount, drm_gem_object_ref_bug);
 	}
 }
 
@@ -237,7 +238,12 @@ drm_gem_object_handle_unreference_unlocked(struct drm_gem_object *obj)
 }
 
 /**
- * Removes the mapping from handle to filp for this object.
+ * drm_gem_handle_delete - deletes the given file-private handle
+ * @filp: drm file-private structure to use for the handle look up
+ * @handle: userspace handle to delete
+ *
+ * Removes the GEM handle from the @filp lookup table and if this is the last
+ * handle also cleans up linked resources like GEM names.
  */
 int
 drm_gem_handle_delete(struct drm_file *filp, u32 handle)
@@ -282,6 +288,9 @@ EXPORT_SYMBOL(drm_gem_handle_delete);
 
 /**
  * drm_gem_dumb_destroy - dumb fb callback helper for gem based drivers
+ * @file: drm file-private structure to remove the dumb handle from
+ * @dev: corresponding drm_device
+ * @handle: the dumb handle to remove
  * 
  * This implements the ->dumb_destroy kms driver callback for drivers which use
  * gem to manage their backing storage.
@@ -296,6 +305,9 @@ EXPORT_SYMBOL(drm_gem_dumb_destroy);
 
 /**
  * drm_gem_handle_create_tail - internal functions to create a handle
+ * @file_priv: drm file-private structure to register the handle for
+ * @obj: object to register
+ * @handlep: pionter to return the created handle to the caller
  * 
  * This expects the dev->object_name_lock to be held already and will drop it
  * before returning. Used to avoid races in establishing new handles when
@@ -348,6 +360,11 @@ drm_gem_handle_create_tail(struct drm_file *file_priv,
 }
 
 /**
+ * gem_handle_create - create a gem handle for an object
+ * @file_priv: drm file-private structure to register the handle for
+ * @obj: object to register
+ * @handlep: pionter to return the created handle to the caller
+ *
  * Create a handle for this object. This adds a handle reference
  * to the object, which includes a regular reference count. Callers
  * will likely want to dereference the object afterwards.
@@ -374,9 +391,8 @@ void
 drm_gem_free_mmap_offset(struct drm_gem_object *obj)
 {
 	struct drm_device *dev = obj->dev;
-	struct drm_gem_mm *mm = dev->mm_private;
 
-	drm_vma_offset_remove(&mm->vma_manager, &obj->vma_node);
+	drm_vma_offset_remove(dev->vma_offset_manager, &obj->vma_node);
 }
 EXPORT_SYMBOL(drm_gem_free_mmap_offset);
 
@@ -398,9 +414,8 @@ int
 drm_gem_create_mmap_offset_size(struct drm_gem_object *obj, size_t size)
 {
 	struct drm_device *dev = obj->dev;
-	struct drm_gem_mm *mm = dev->mm_private;
 
-	return drm_vma_offset_add(&mm->vma_manager, &obj->vma_node,
+	return drm_vma_offset_add(dev->vma_offset_manager, &obj->vma_node,
 				  size / PAGE_SIZE);
 }
 EXPORT_SYMBOL(drm_gem_create_mmap_offset_size);
@@ -459,21 +474,10 @@ struct page **drm_gem_get_pages(struct drm_gem_object *obj, gfp_t gfpmask)
 			goto fail;
 		pages[i] = p;
 
-		/* There is a hypothetical issue w/ drivers that require
-		 * buffer memory in the low 4GB.. if the pages are un-
-		 * pinned, and swapped out, they can end up swapped back
-		 * in above 4GB.  If pages are already in memory, then
-		 * shmem_read_mapping_page_gfp will ignore the gfpmask,
-		 * even if the already in-memory page disobeys the mask.
-		 *
-		 * It is only a theoretical issue today, because none of
-		 * the devices with this limitation can be populated with
-		 * enough memory to trigger the issue.  But this BUG_ON()
-		 * is here as a reminder in case the problem with
-		 * shmem_read_mapping_page_gfp() isn't solved by the time
-		 * it does become a real issue.
-		 *
-		 * See this thread: http://lkml.org/lkml/2011/7/11/238
+		/* Make sure shmem keeps __GFP_DMA32 allocated pages in the
+		 * correct region during swapin. Note that this requires
+		 * __GFP_DMA32 to be set in mapping_gfp_mask(inode->i_mapping)
+		 * so shmem can relocate pages during swapin if required.
 		 */
 		BUG_ON((gfpmask & __GFP_DMA32) &&
 				(page_to_pfn(p) >= 0x00100000UL));
@@ -550,6 +554,11 @@ drm_gem_object_lookup(struct drm_device *dev, struct drm_file *filp,
 EXPORT_SYMBOL(drm_gem_object_lookup);
 
 /**
+ * drm_gem_close_ioctl - implementation of the GEM_CLOSE ioctl
+ * @dev: drm_device
+ * @data: ioctl data
+ * @file_priv: drm file-private structure
+ *
  * Releases the handle to an mm object.
  */
 int
@@ -568,6 +577,11 @@ drm_gem_close_ioctl(struct drm_device *dev, void *data,
 }
 
 /**
+ * drm_gem_flink_ioctl - implementation of the GEM_FLINK ioctl
+ * @dev: drm_device
+ * @data: ioctl data
+ * @file_priv: drm file-private structure
+ *
  * Create a global name for an object, returning the name.
  *
  * Note that the name does not hold a reference; when the object
@@ -602,9 +616,6 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
 			goto err;
 
 		obj->name = ret;
-
-		/* Allocate a reference for the name table.  */
-		drm_gem_object_reference(obj);
 	}
 
 	args->name = (uint64_t) obj->name;
@@ -618,6 +629,11 @@ err:
 }
 
 /**
+ * drm_gem_open - implementation of the GEM_OPEN ioctl
+ * @dev: drm_device
+ * @data: ioctl data
+ * @file_priv: drm file-private structure
+ *
  * Open an object using the global name, returning a handle and the size.
  *
  * This handle (of course) holds a reference to the object, so the object
@@ -657,6 +673,10 @@ drm_gem_open_ioctl(struct drm_device *dev, void *data,
 }
 
 /**
+ * gem_gem_open - initalizes GEM file-private structures at devnode open time
+ * @dev: drm_device which is being opened by userspace
+ * @file_private: drm file-private structure to set up
+ *
  * Called at device open time, sets up the structure for handling refcounting
  * of mm objects.
  */
@@ -667,7 +687,7 @@ drm_gem_open(struct drm_device *dev, struct drm_file *file_private)
 	spin_lock_init(&file_private->table_lock);
 }
 
-/**
+/*
  * Called at device close to release the file's
  * handle references on objects.
  */
@@ -691,6 +711,10 @@ drm_gem_object_release_handle(int id, void *ptr, void *data)
 }
 
 /**
+ * drm_gem_release - release file-private GEM resources
+ * @dev: drm_device which is being closed by userspace
+ * @file_private: drm file-private structure to clean up
+ *
  * Called at close time when the filp is going away.
  *
  * Releases any remaining references on objects by this filp.
@@ -709,11 +733,16 @@ drm_gem_object_release(struct drm_gem_object *obj)
 	WARN_ON(obj->dma_buf);
 
 	if (obj->filp)
-	    fput(obj->filp);
+		fput(obj->filp);
+
+	drm_gem_free_mmap_offset(obj);
 }
 EXPORT_SYMBOL(drm_gem_object_release);
 
 /**
+ * drm_gem_object_free - free a GEM object
+ * @kref: kref of the object to free
+ *
  * Called after the last reference to the object has been lost.
  * Must be called holding struct_ mutex
  *
@@ -799,7 +828,7 @@ int drm_gem_mmap_obj(struct drm_gem_object *obj, unsigned long obj_size,
 	vma->vm_flags |= VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP;
 	vma->vm_ops = dev->driver->gem_vm_ops;
 	vma->vm_private_data = obj;
-	vma->vm_page_prot =  pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
+	vma->vm_page_prot = pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
 
 	/* Take a ref for this mapping of the object, so that the fault
 	 * handler can dereference the mmap offset's pointer to the object.
@@ -833,17 +862,17 @@ int drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 {
 	struct drm_file *priv = filp->private_data;
 	struct drm_device *dev = priv->minor->dev;
-	struct drm_gem_mm *mm = dev->mm_private;
 	struct drm_gem_object *obj;
 	struct drm_vma_offset_node *node;
-	int ret = 0;
+	int ret;
 
 	if (drm_device_is_unplugged(dev))
 		return -ENODEV;
 
 	mutex_lock(&dev->struct_mutex);
 
-	node = drm_vma_offset_exact_lookup(&mm->vma_manager, vma->vm_pgoff,
+	node = drm_vma_offset_exact_lookup(dev->vma_offset_manager,
+					   vma->vm_pgoff,
 					   vma_pages(vma));
 	if (!node) {
 		mutex_unlock(&dev->struct_mutex);
diff --git a/drivers/gpu/drm/drm_gem_cma_helper.c b/drivers/gpu/drm/drm_gem_cma_helper.c
index 6b51bf90df0..05c97c5350a 100644
--- a/drivers/gpu/drm/drm_gem_cma_helper.c
+++ b/drivers/gpu/drm/drm_gem_cma_helper.c
@@ -79,7 +79,6 @@ struct drm_gem_cma_object *drm_gem_cma_create(struct drm_device *drm,
 		unsigned int size)
 {
 	struct drm_gem_cma_object *cma_obj;
-	struct sg_table *sgt = NULL;
 	int ret;
 
 	size = round_up(size, PAGE_SIZE);
@@ -97,23 +96,9 @@ struct drm_gem_cma_object *drm_gem_cma_create(struct drm_device *drm,
 		goto error;
 	}
 
-	sgt = kzalloc(sizeof(*cma_obj->sgt), GFP_KERNEL);
-	if (sgt == NULL) {
-		ret = -ENOMEM;
-		goto error;
-	}
-
-	ret = dma_get_sgtable(drm->dev, sgt, cma_obj->vaddr,
-			      cma_obj->paddr, size);
-	if (ret < 0)
-		goto error;
-
-	cma_obj->sgt = sgt;
-
 	return cma_obj;
 
 error:
-	kfree(sgt);
 	drm_gem_cma_free_object(&cma_obj->base);
 	return ERR_PTR(ret);
 }
@@ -175,10 +160,6 @@ void drm_gem_cma_free_object(struct drm_gem_object *gem_obj)
 	if (cma_obj->vaddr) {
 		dma_free_writecombine(gem_obj->dev->dev, cma_obj->base.size,
 				      cma_obj->vaddr, cma_obj->paddr);
-		if (cma_obj->sgt) {
-			sg_free_table(cma_obj->sgt);
-			kfree(cma_obj->sgt);
-		}
 	} else if (gem_obj->import_attach) {
 		drm_prime_gem_destroy(gem_obj, cma_obj->sgt);
 	}
@@ -253,8 +234,17 @@ static int drm_gem_cma_mmap_obj(struct drm_gem_cma_object *cma_obj,
 {
 	int ret;
 
-	ret = remap_pfn_range(vma, vma->vm_start, cma_obj->paddr >> PAGE_SHIFT,
-			vma->vm_end - vma->vm_start, vma->vm_page_prot);
+	/*
+	 * Clear the VM_PFNMAP flag that was set by drm_gem_mmap(), and set the
+	 * vm_pgoff (used as a fake buffer offset by DRM) to 0 as we want to map
+	 * the whole buffer.
+	 */
+	vma->vm_flags &= ~VM_PFNMAP;
+	vma->vm_pgoff = 0;
+
+	ret = dma_mmap_writecombine(cma_obj->base.dev->dev, vma,
+				    cma_obj->vaddr, cma_obj->paddr,
+				    vma->vm_end - vma->vm_start);
 	if (ret)
 		drm_gem_vm_close(vma);
 
@@ -292,9 +282,9 @@ void drm_gem_cma_describe(struct drm_gem_cma_object *cma_obj, struct seq_file *m
 
 	off = drm_vma_node_start(&obj->vma_node);
 
-	seq_printf(m, "%2d (%2d) %08llx %08Zx %p %d",
+	seq_printf(m, "%2d (%2d) %08llx %pad %p %d",
 			obj->name, obj->refcount.refcount.counter,
-			off, cma_obj->paddr, cma_obj->vaddr, obj->size);
+			off, &cma_obj->paddr, cma_obj->vaddr, obj->size);
 
 	seq_printf(m, "\n");
 }
@@ -342,7 +332,7 @@ drm_gem_cma_prime_import_sg_table(struct drm_device *dev, size_t size,
 	cma_obj->paddr = sg_dma_address(sgt->sgl);
 	cma_obj->sgt = sgt;
 
-	DRM_DEBUG_PRIME("dma_addr = 0x%x, size = %zu\n", cma_obj->paddr, size);
+	DRM_DEBUG_PRIME("dma_addr = %pad, size = %zu\n", &cma_obj->paddr, size);
 
 	return &cma_obj->base;
 }
diff --git a/drivers/gpu/drm/drm_info.c b/drivers/gpu/drm/drm_info.c
index 7d5a152eeb0..86feedd5e6f 100644
--- a/drivers/gpu/drm/drm_info.c
+++ b/drivers/gpu/drm/drm_info.c
@@ -47,18 +47,16 @@ int drm_name_info(struct seq_file *m, void *data)
 	struct drm_minor *minor = node->minor;
 	struct drm_device *dev = minor->dev;
 	struct drm_master *master = minor->master;
-	const char *bus_name;
 	if (!master)
 		return 0;
 
-	bus_name = dev->driver->bus->get_name(dev);
 	if (master->unique) {
 		seq_printf(m, "%s %s %s\n",
-			   bus_name,
+			   dev->driver->name,
 			   dev_name(dev->dev), master->unique);
 	} else {
 		seq_printf(m, "%s %s\n",
-			   bus_name, dev_name(dev->dev));
+			   dev->driver->name, dev_name(dev->dev));
 	}
 	return 0;
 }
@@ -186,14 +184,14 @@ int drm_clients_info(struct seq_file *m, void *data)
 	struct drm_file *priv;
 
 	mutex_lock(&dev->struct_mutex);
-	seq_printf(m, "a dev	pid    uid	magic	  ioctls\n\n");
+	seq_printf(m, "a dev	pid    uid	magic\n\n");
 	list_for_each_entry(priv, &dev->filelist, lhead) {
-		seq_printf(m, "%c %3d %5d %5d %10u %10lu\n",
+		seq_printf(m, "%c %3d %5d %5d %10u\n",
 			   priv->authenticated ? 'y' : 'n',
 			   priv->minor->index,
 			   pid_vnr(priv->pid),
 			   from_kuid_munged(seq_user_ns(m), priv->uid),
-			   priv->magic, priv->ioctl_count);
+			   priv->magic);
 	}
 	mutex_unlock(&dev->struct_mutex);
 	return 0;
@@ -234,14 +232,18 @@ int drm_vma_info(struct seq_file *m, void *data)
 	struct drm_device *dev = node->minor->dev;
 	struct drm_vma_entry *pt;
 	struct vm_area_struct *vma;
+	unsigned long vma_count = 0;
 #if defined(__i386__)
 	unsigned int pgprot;
 #endif
 
 	mutex_lock(&dev->struct_mutex);
-	seq_printf(m, "vma use count: %d, high_memory = %pK, 0x%pK\n",
-		   atomic_read(&dev->vma_count),
-		   high_memory, (void *)(unsigned long)virt_to_phys(high_memory));
+	list_for_each_entry(pt, &dev->vmalist, head)
+		vma_count++;
+
+	seq_printf(m, "vma use count: %lu, high_memory = %pK, 0x%pK\n",
+		   vma_count, high_memory,
+		   (void *)(unsigned long)virt_to_phys(high_memory));
 
 	list_for_each_entry(pt, &dev->vmalist, head) {
 		vma = pt->vma;
diff --git a/drivers/gpu/drm/drm_ioctl.c b/drivers/gpu/drm/drm_ioctl.c
index dffc836144c..69c61f392e6 100644
--- a/drivers/gpu/drm/drm_ioctl.c
+++ b/drivers/gpu/drm/drm_ioctl.c
@@ -72,9 +72,6 @@ static void
 drm_unset_busid(struct drm_device *dev,
 		struct drm_master *master)
 {
-	kfree(dev->devname);
-	dev->devname = NULL;
-
 	kfree(master->unique);
 	master->unique = NULL;
 	master->unique_len = 0;
@@ -93,7 +90,8 @@ drm_unset_busid(struct drm_device *dev,
  * Copies the bus id from userspace into drm_device::unique, and verifies that
  * it matches the device this DRM is attached to (EINVAL otherwise).  Deprecated
  * in interface version 1.1 and will return EBUSY when setversion has requested
- * version 1.1 or greater.
+ * version 1.1 or greater. Also note that KMS is all version 1.1 and later and
+ * UMS was only ever supported on pci devices.
  */
 int drm_setunique(struct drm_device *dev, void *data,
 		  struct drm_file *file_priv)
@@ -108,10 +106,13 @@ int drm_setunique(struct drm_device *dev, void *data,
 	if (!u->unique_len || u->unique_len > 1024)
 		return -EINVAL;
 
-	if (!dev->driver->bus->set_unique)
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
+		return 0;
+
+	if (WARN_ON(!dev->pdev))
 		return -EINVAL;
 
-	ret = dev->driver->bus->set_unique(dev, master, u);
+	ret = drm_pci_set_unique(dev, master, u);
 	if (ret)
 		goto err;
 
@@ -130,13 +131,25 @@ static int drm_set_busid(struct drm_device *dev, struct drm_file *file_priv)
 	if (master->unique != NULL)
 		drm_unset_busid(dev, master);
 
-	ret = dev->driver->bus->set_busid(dev, master);
-	if (ret)
-		goto err;
+	if (dev->driver->bus && dev->driver->bus->set_busid) {
+		ret = dev->driver->bus->set_busid(dev, master);
+		if (ret) {
+			drm_unset_busid(dev, master);
+			return ret;
+		}
+	} else {
+		if (WARN(dev->unique == NULL,
+			 "No drm_bus.set_busid() implementation provided by "
+			 "%ps. Use drm_dev_set_unique() to set the unique "
+			 "name explicitly.", dev->driver))
+			return -EINVAL;
+
+		master->unique = kstrdup(dev->unique, GFP_KERNEL);
+		if (master->unique)
+			master->unique_len = strlen(dev->unique);
+	}
+
 	return 0;
-err:
-	drm_unset_busid(dev, master);
-	return ret;
 }
 
 /**
@@ -296,6 +309,18 @@ int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 	case DRM_CAP_ASYNC_PAGE_FLIP:
 		req->value = dev->mode_config.async_page_flip;
 		break;
+	case DRM_CAP_CURSOR_WIDTH:
+		if (dev->mode_config.cursor_width)
+			req->value = dev->mode_config.cursor_width;
+		else
+			req->value = 64;
+		break;
+	case DRM_CAP_CURSOR_HEIGHT:
+		if (dev->mode_config.cursor_height)
+			req->value = dev->mode_config.cursor_height;
+		else
+			req->value = 64;
+		break;
 	default:
 		return -EINVAL;
 	}
@@ -316,6 +341,13 @@ drm_setclientcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 			return -EINVAL;
 		file_priv->stereo_allowed = req->value;
 		break;
+	case DRM_CLIENT_CAP_UNIVERSAL_PLANES:
+		if (!drm_universal_planes)
+			return -EINVAL;
+		if (req->value > 1)
+			return -EINVAL;
+		file_priv->universal_planes = req->value;
+		break;
 	default:
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
index 64c34d5876f..0de123afdb3 100644
--- a/drivers/gpu/drm/drm_irq.c
+++ b/drivers/gpu/drm/drm_irq.c
@@ -1,6 +1,5 @@
-/**
- * \file drm_irq.c
- * IRQ support
+/*
+ * drm_irq.c IRQ and vblank support
  *
  * \author Rickard E. (Rik) Faith <faith@valinux.com>
  * \author Gareth Hughes <gareth@valinux.com>
@@ -56,33 +55,6 @@
  */
 #define DRM_REDUNDANT_VBLIRQ_THRESH_NS 1000000
 
-/**
- * Get interrupt from bus id.
- *
- * \param inode device inode.
- * \param file_priv DRM file private.
- * \param cmd command.
- * \param arg user argument, pointing to a drm_irq_busid structure.
- * \return zero on success or a negative number on failure.
- *
- * Finds the PCI device with the specified bus id and gets its IRQ number.
- * This IOCTL is deprecated, and will now return EINVAL for any busid not equal
- * to that of the device that this DRM instance attached to.
- */
-int drm_irq_by_busid(struct drm_device *dev, void *data,
-		     struct drm_file *file_priv)
-{
-	struct drm_irq_busid *p = data;
-
-	if (!dev->driver->bus->irq_by_busid)
-		return -EINVAL;
-
-	if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
-		return -EINVAL;
-
-	return dev->driver->bus->irq_by_busid(dev, p);
-}
-
 /*
  * Clear vblank timestamp buffer for a crtc.
  */
@@ -156,7 +128,7 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 	 */
 	if ((vblrc > 0) && (abs64(diff_ns) > 1000000)) {
 		atomic_inc(&dev->vblank[crtc].count);
-		smp_mb__after_atomic_inc();
+		smp_mb__after_atomic();
 	}
 
 	/* Invalidate all timestamps while vblank irq's are off. */
@@ -167,33 +139,40 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
 
 static void vblank_disable_fn(unsigned long arg)
 {
-	struct drm_device *dev = (struct drm_device *)arg;
+	struct drm_vblank_crtc *vblank = (void *)arg;
+	struct drm_device *dev = vblank->dev;
 	unsigned long irqflags;
-	int i;
+	int crtc = vblank->crtc;
 
 	if (!dev->vblank_disable_allowed)
 		return;
 
-	for (i = 0; i < dev->num_crtcs; i++) {
-		spin_lock_irqsave(&dev->vbl_lock, irqflags);
-		if (atomic_read(&dev->vblank[i].refcount) == 0 &&
-		    dev->vblank[i].enabled) {
-			DRM_DEBUG("disabling vblank on crtc %d\n", i);
-			vblank_disable_and_save(dev, i);
-		}
-		spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
+	spin_lock_irqsave(&dev->vbl_lock, irqflags);
+	if (atomic_read(&vblank->refcount) == 0 && vblank->enabled) {
+		DRM_DEBUG("disabling vblank on crtc %d\n", crtc);
+		vblank_disable_and_save(dev, crtc);
 	}
+	spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
 }
 
+/**
+ * drm_vblank_cleanup - cleanup vblank support
+ * @dev: DRM device
+ *
+ * This function cleans up any resources allocated in drm_vblank_init.
+ */
 void drm_vblank_cleanup(struct drm_device *dev)
 {
+	int crtc;
+
 	/* Bail if the driver didn't call drm_vblank_init() */
 	if (dev->num_crtcs == 0)
 		return;
 
-	del_timer_sync(&dev->vblank_disable_timer);
-
-	vblank_disable_fn((unsigned long)dev);
+	for (crtc = 0; crtc < dev->num_crtcs; crtc++) {
+		del_timer_sync(&dev->vblank[crtc].disable_timer);
+		vblank_disable_fn((unsigned long)&dev->vblank[crtc]);
+	}
 
 	kfree(dev->vblank);
 
@@ -201,12 +180,20 @@ void drm_vblank_cleanup(struct drm_device *dev)
 }
 EXPORT_SYMBOL(drm_vblank_cleanup);
 
+/**
+ * drm_vblank_init - initialize vblank support
+ * @dev: drm_device
+ * @num_crtcs: number of crtcs supported by @dev
+ *
+ * This function initializes vblank support for @num_crtcs display pipelines.
+ *
+ * Returns:
+ * Zero on success or a negative error code on failure.
+ */
 int drm_vblank_init(struct drm_device *dev, int num_crtcs)
 {
 	int i, ret = -ENOMEM;
 
-	setup_timer(&dev->vblank_disable_timer, vblank_disable_fn,
-		    (unsigned long)dev);
 	spin_lock_init(&dev->vbl_lock);
 	spin_lock_init(&dev->vblank_time_lock);
 
@@ -216,8 +203,13 @@ int drm_vblank_init(struct drm_device *dev, int num_crtcs)
 	if (!dev->vblank)
 		goto err;
 
-	for (i = 0; i < num_crtcs; i++)
+	for (i = 0; i < num_crtcs; i++) {
+		dev->vblank[i].dev = dev;
+		dev->vblank[i].crtc = i;
 		init_waitqueue_head(&dev->vblank[i].queue);
+		setup_timer(&dev->vblank[i].disable_timer, vblank_disable_fn,
+			    (unsigned long)&dev->vblank[i]);
+	}
 
 	DRM_INFO("Supports vblank timestamp caching Rev 2 (21.10.2013).\n");
 
@@ -261,42 +253,42 @@ static void drm_irq_vgaarb_nokms(void *cookie, bool state)
 }
 
 /**
- * Install IRQ handler.
- *
- * \param dev DRM device.
+ * drm_irq_install - install IRQ handler
+ * @dev: DRM device
+ * @irq: IRQ number to install the handler for
  *
  * Initializes the IRQ related data. Installs the handler, calling the driver
- * \c irq_preinstall() and \c irq_postinstall() functions
- * before and after the installation.
+ * irq_preinstall() and irq_postinstall() functions before and after the
+ * installation.
+ *
+ * This is the simplified helper interface provided for drivers with no special
+ * needs. Drivers which need to install interrupt handlers for multiple
+ * interrupts must instead set drm_device->irq_enabled to signal the DRM core
+ * that vblank interrupts are available.
+ *
+ * Returns:
+ * Zero on success or a negative error code on failure.
  */
-int drm_irq_install(struct drm_device *dev)
+int drm_irq_install(struct drm_device *dev, int irq)
 {
 	int ret;
 	unsigned long sh_flags = 0;
-	char *irqname;
 
 	if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
 		return -EINVAL;
 
-	if (drm_dev_to_irq(dev) == 0)
+	if (irq == 0)
 		return -EINVAL;
 
-	mutex_lock(&dev->struct_mutex);
-
 	/* Driver must have been initialized */
-	if (!dev->dev_private) {
-		mutex_unlock(&dev->struct_mutex);
+	if (!dev->dev_private)
 		return -EINVAL;
-	}
 
-	if (dev->irq_enabled) {
-		mutex_unlock(&dev->struct_mutex);
+	if (dev->irq_enabled)
 		return -EBUSY;
-	}
 	dev->irq_enabled = true;
-	mutex_unlock(&dev->struct_mutex);
 
-	DRM_DEBUG("irq=%d\n", drm_dev_to_irq(dev));
+	DRM_DEBUG("irq=%d\n", irq);
 
 	/* Before installing handler */
 	if (dev->driver->irq_preinstall)
@@ -306,18 +298,11 @@ int drm_irq_install(struct drm_device *dev)
 	if (drm_core_check_feature(dev, DRIVER_IRQ_SHARED))
 		sh_flags = IRQF_SHARED;
 
-	if (dev->devname)
-		irqname = dev->devname;
-	else
-		irqname = dev->driver->name;
-
-	ret = request_irq(drm_dev_to_irq(dev), dev->driver->irq_handler,
-			  sh_flags, irqname, dev);
+	ret = request_irq(irq, dev->driver->irq_handler,
+			  sh_flags, dev->driver->name, dev);
 
 	if (ret < 0) {
-		mutex_lock(&dev->struct_mutex);
 		dev->irq_enabled = false;
-		mutex_unlock(&dev->struct_mutex);
 		return ret;
 	}
 
@@ -329,12 +314,12 @@ int drm_irq_install(struct drm_device *dev)
 		ret = dev->driver->irq_postinstall(dev);
 
 	if (ret < 0) {
-		mutex_lock(&dev->struct_mutex);
 		dev->irq_enabled = false;
-		mutex_unlock(&dev->struct_mutex);
 		if (!drm_core_check_feature(dev, DRIVER_MODESET))
 			vga_client_register(dev->pdev, NULL, NULL, NULL);
-		free_irq(drm_dev_to_irq(dev), dev);
+		free_irq(irq, dev);
+	} else {
+		dev->irq = irq;
 	}
 
 	return ret;
@@ -342,11 +327,20 @@ int drm_irq_install(struct drm_device *dev)
 EXPORT_SYMBOL(drm_irq_install);
 
 /**
- * Uninstall the IRQ handler.
+ * drm_irq_uninstall - uninstall the IRQ handler
+ * @dev: DRM device
+ *
+ * Calls the driver's irq_uninstall() function and unregisters the IRQ handler.
+ * This should only be called by drivers which used drm_irq_install() to set up
+ * their interrupt handler. Other drivers must only reset
+ * drm_device->irq_enabled to false.
  *
- * \param dev DRM device.
+ * Note that for kernel modesetting drivers it is a bug if this function fails.
+ * The sanity checks are only to catch buggy user modesetting drivers which call
+ * the same function through an ioctl.
  *
- * Calls the driver's \c irq_uninstall() function, and stops the irq.
+ * Returns:
+ * Zero on success or a negative error code on failure.
  */
 int drm_irq_uninstall(struct drm_device *dev)
 {
@@ -357,10 +351,8 @@ int drm_irq_uninstall(struct drm_device *dev)
 	if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
 		return -EINVAL;
 
-	mutex_lock(&dev->struct_mutex);
 	irq_enabled = dev->irq_enabled;
 	dev->irq_enabled = false;
-	mutex_unlock(&dev->struct_mutex);
 
 	/*
 	 * Wake up any waiters so they don't hang.
@@ -368,7 +360,7 @@ int drm_irq_uninstall(struct drm_device *dev)
 	if (dev->num_crtcs) {
 		spin_lock_irqsave(&dev->vbl_lock, irqflags);
 		for (i = 0; i < dev->num_crtcs; i++) {
-			DRM_WAKEUP(&dev->vblank[i].queue);
+			wake_up(&dev->vblank[i].queue);
 			dev->vblank[i].enabled = false;
 			dev->vblank[i].last =
 				dev->driver->get_vblank_counter(dev, i);
@@ -379,7 +371,7 @@ int drm_irq_uninstall(struct drm_device *dev)
 	if (!irq_enabled)
 		return -EINVAL;
 
-	DRM_DEBUG("irq=%d\n", drm_dev_to_irq(dev));
+	DRM_DEBUG("irq=%d\n", dev->irq);
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		vga_client_register(dev->pdev, NULL, NULL, NULL);
@@ -387,13 +379,13 @@ int drm_irq_uninstall(struct drm_device *dev)
 	if (dev->driver->irq_uninstall)
 		dev->driver->irq_uninstall(dev);
 
-	free_irq(drm_dev_to_irq(dev), dev);
+	free_irq(dev->irq, dev);
 
 	return 0;
 }
 EXPORT_SYMBOL(drm_irq_uninstall);
 
-/**
+/*
  * IRQ control ioctl.
  *
  * \param inode device inode.
@@ -408,73 +400,78 @@ int drm_control(struct drm_device *dev, void *data,
 		struct drm_file *file_priv)
 {
 	struct drm_control *ctl = data;
+	int ret = 0, irq;
 
 	/* if we haven't irq we fallback for compatibility reasons -
 	 * this used to be a separate function in drm_dma.h
 	 */
 
+	if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
+		return 0;
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
+		return 0;
+	/* UMS was only ever support on pci devices. */
+	if (WARN_ON(!dev->pdev))
+		return -EINVAL;
 
 	switch (ctl->func) {
 	case DRM_INST_HANDLER:
-		if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
-			return 0;
-		if (drm_core_check_feature(dev, DRIVER_MODESET))
-			return 0;
+		irq = dev->pdev->irq;
+
 		if (dev->if_version < DRM_IF_VERSION(1, 2) &&
-		    ctl->irq != drm_dev_to_irq(dev))
+		    ctl->irq != irq)
 			return -EINVAL;
-		return drm_irq_install(dev);
+		mutex_lock(&dev->struct_mutex);
+		ret = drm_irq_install(dev, irq);
+		mutex_unlock(&dev->struct_mutex);
+
+		return ret;
 	case DRM_UNINST_HANDLER:
-		if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
-			return 0;
-		if (drm_core_check_feature(dev, DRIVER_MODESET))
-			return 0;
-		return drm_irq_uninstall(dev);
+		mutex_lock(&dev->struct_mutex);
+		ret = drm_irq_uninstall(dev);
+		mutex_unlock(&dev->struct_mutex);
+
+		return ret;
 	default:
 		return -EINVAL;
 	}
 }
 
 /**
- * drm_calc_timestamping_constants - Calculate and
- * store various constants which are later needed by
- * vblank and swap-completion timestamping, e.g, by
- * drm_calc_vbltimestamp_from_scanoutpos().
- * They are derived from crtc's true scanout timing,
- * so they take things like panel scaling or other
- * adjustments into account.
- *
- * @crtc drm_crtc whose timestamp constants should be updated.
+ * drm_calc_timestamping_constants - calculate vblank timestamp constants
+ * @crtc: drm_crtc whose timestamp constants should be updated.
+ * @mode: display mode containing the scanout timings
  *
+ * Calculate and store various constants which are later
+ * needed by vblank and swap-completion timestamping, e.g,
+ * by drm_calc_vbltimestamp_from_scanoutpos(). They are
+ * derived from CRTC's true scanout timing, so they take
+ * things like panel scaling or other adjustments into account.
  */
-void drm_calc_timestamping_constants(struct drm_crtc *crtc)
+void drm_calc_timestamping_constants(struct drm_crtc *crtc,
+				     const struct drm_display_mode *mode)
 {
-	s64 linedur_ns = 0, pixeldur_ns = 0, framedur_ns = 0;
-	u64 dotclock;
-
-	/* Dot clock in Hz: */
-	dotclock = (u64) crtc->hwmode.clock * 1000;
-
-	/* Fields of interlaced scanout modes are only half a frame duration.
-	 * Double the dotclock to get half the frame-/line-/pixelduration.
-	 */
-	if (crtc->hwmode.flags & DRM_MODE_FLAG_INTERLACE)
-		dotclock *= 2;
+	int linedur_ns = 0, pixeldur_ns = 0, framedur_ns = 0;
+	int dotclock = mode->crtc_clock;
 
 	/* Valid dotclock? */
 	if (dotclock > 0) {
-		int frame_size;
-		/* Convert scanline length in pixels and video dot clock to
-		 * line duration, frame duration and pixel duration in
-		 * nanoseconds:
+		int frame_size = mode->crtc_htotal * mode->crtc_vtotal;
+
+		/*
+		 * Convert scanline length in pixels and video
+		 * dot clock to line duration, frame duration
+		 * and pixel duration in nanoseconds:
 		 */
-		pixeldur_ns = (s64) div64_u64(1000000000, dotclock);
-		linedur_ns  = (s64) div64_u64(((u64) crtc->hwmode.crtc_htotal *
-					      1000000000), dotclock);
-		frame_size = crtc->hwmode.crtc_htotal *
-				crtc->hwmode.crtc_vtotal;
-		framedur_ns = (s64) div64_u64((u64) frame_size * 1000000000,
-					      dotclock);
+		pixeldur_ns = 1000000 / dotclock;
+		linedur_ns  = div_u64((u64) mode->crtc_htotal * 1000000, dotclock);
+		framedur_ns = div_u64((u64) frame_size * 1000000, dotclock);
+
+		/*
+		 * Fields of interlaced scanout modes are only half a frame duration.
+		 */
+		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+			framedur_ns /= 2;
 	} else
 		DRM_ERROR("crtc %d: Can't calculate constants, dotclock = 0!\n",
 			  crtc->base.id);
@@ -484,20 +481,31 @@ void drm_calc_timestamping_constants(struct drm_crtc *crtc)
 	crtc->framedur_ns = framedur_ns;
 
 	DRM_DEBUG("crtc %d: hwmode: htotal %d, vtotal %d, vdisplay %d\n",
-		  crtc->base.id, crtc->hwmode.crtc_htotal,
-		  crtc->hwmode.crtc_vtotal, crtc->hwmode.crtc_vdisplay);
+		  crtc->base.id, mode->crtc_htotal,
+		  mode->crtc_vtotal, mode->crtc_vdisplay);
 	DRM_DEBUG("crtc %d: clock %d kHz framedur %d linedur %d, pixeldur %d\n",
-		  crtc->base.id, (int) dotclock/1000, (int) framedur_ns,
-		  (int) linedur_ns, (int) pixeldur_ns);
+		  crtc->base.id, dotclock, framedur_ns,
+		  linedur_ns, pixeldur_ns);
 }
 EXPORT_SYMBOL(drm_calc_timestamping_constants);
 
 /**
- * drm_calc_vbltimestamp_from_scanoutpos - helper routine for kms
- * drivers. Implements calculation of exact vblank timestamps from
- * given drm_display_mode timings and current video scanout position
- * of a crtc. This can be called from within get_vblank_timestamp()
- * implementation of a kms driver to implement the actual timestamping.
+ * drm_calc_vbltimestamp_from_scanoutpos - precise vblank timestamp helper
+ * @dev: DRM device
+ * @crtc: Which CRTC's vblank timestamp to retrieve
+ * @max_error: Desired maximum allowable error in timestamps (nanosecs)
+ *             On return contains true maximum error of timestamp
+ * @vblank_time: Pointer to struct timeval which should receive the timestamp
+ * @flags: Flags to pass to driver:
+ *         0 = Default,
+ *         DRM_CALLED_FROM_VBLIRQ = If function is called from vbl IRQ handler
+ * @refcrtc: CRTC which defines scanout timing
+ * @mode: mode which defines the scanout timings
+ *
+ * Implements calculation of exact vblank timestamps from given drm_display_mode
+ * timings and current video scanout position of a CRTC. This can be called from
+ * within get_vblank_timestamp() implementation of a kms driver to implement the
+ * actual timestamping.
  *
  * Should return timestamps conforming to the OML_sync_control OpenML
  * extension specification. The timestamp corresponds to the end of
@@ -512,20 +520,11 @@ EXPORT_SYMBOL(drm_calc_timestamping_constants);
  * returns as no operation if a doublescan or interlaced video mode is
  * active. Higher level code is expected to handle this.
  *
- * @dev: DRM device.
- * @crtc: Which crtc's vblank timestamp to retrieve.
- * @max_error: Desired maximum allowable error in timestamps (nanosecs).
- *             On return contains true maximum error of timestamp.
- * @vblank_time: Pointer to struct timeval which should receive the timestamp.
- * @flags: Flags to pass to driver:
- *         0 = Default.
- *         DRM_CALLED_FROM_VBLIRQ = If function is called from vbl irq handler.
- * @refcrtc: drm_crtc* of crtc which defines scanout timing.
- *
- * Returns negative value on error, failure or if not supported in current
+ * Returns:
+ * Negative value on error, failure or if not supported in current
  * video mode:
  *
- * -EINVAL   - Invalid crtc.
+ * -EINVAL   - Invalid CRTC.
  * -EAGAIN   - Temporary unavailable, e.g., called before initial modeset.
  * -ENOTSUPP - Function not supported in current display mode.
  * -EIO      - Failed, e.g., due to failed scanout position query.
@@ -540,14 +539,14 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 					  int *max_error,
 					  struct timeval *vblank_time,
 					  unsigned flags,
-					  struct drm_crtc *refcrtc)
+					  const struct drm_crtc *refcrtc,
+					  const struct drm_display_mode *mode)
 {
 	ktime_t stime, etime, mono_time_offset;
 	struct timeval tv_etime;
-	struct drm_display_mode *mode;
-	int vbl_status, vtotal, vdisplay;
+	int vbl_status;
 	int vpos, hpos, i;
-	s64 framedur_ns, linedur_ns, pixeldur_ns, delta_ns, duration_ns;
+	int framedur_ns, linedur_ns, pixeldur_ns, delta_ns, duration_ns;
 	bool invbl;
 
 	if (crtc < 0 || crtc >= dev->num_crtcs) {
@@ -561,10 +560,6 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 		return -EIO;
 	}
 
-	mode = &refcrtc->hwmode;
-	vtotal = mode->crtc_vtotal;
-	vdisplay = mode->crtc_vdisplay;
-
 	/* Durations of frames, lines, pixels in nanoseconds. */
 	framedur_ns = refcrtc->framedur_ns;
 	linedur_ns  = refcrtc->linedur_ns;
@@ -573,7 +568,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 	/* If mode timing undefined, just return as no-op:
 	 * Happens during initial modesetting of a crtc.
 	 */
-	if (vtotal <= 0 || vdisplay <= 0 || framedur_ns == 0) {
+	if (framedur_ns == 0) {
 		DRM_DEBUG("crtc %d: Noop due to uninitialized mode.\n", crtc);
 		return -EAGAIN;
 	}
@@ -590,7 +585,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 		 * Get vertical and horizontal scanout position vpos, hpos,
 		 * and bounding timestamps stime, etime, pre/post query.
 		 */
-		vbl_status = dev->driver->get_scanout_position(dev, crtc, &vpos,
+		vbl_status = dev->driver->get_scanout_position(dev, crtc, flags, &vpos,
 							       &hpos, &stime, &etime);
 
 		/*
@@ -611,18 +606,18 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 		duration_ns = ktime_to_ns(etime) - ktime_to_ns(stime);
 
 		/* Accept result with <  max_error nsecs timing uncertainty. */
-		if (duration_ns <= (s64) *max_error)
+		if (duration_ns <= *max_error)
 			break;
 	}
 
 	/* Noisy system timing? */
 	if (i == DRM_TIMESTAMP_MAXRETRIES) {
 		DRM_DEBUG("crtc %d: Noisy timestamp %d us > %d us [%d reps].\n",
-			  crtc, (int) duration_ns/1000, *max_error/1000, i);
+			  crtc, duration_ns/1000, *max_error/1000, i);
 	}
 
 	/* Return upper bound of timestamp precision error. */
-	*max_error = (int) duration_ns;
+	*max_error = duration_ns;
 
 	/* Check if in vblank area:
 	 * vpos is >=0 in video scanout area, but negative
@@ -635,25 +630,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 	 * since start of scanout at first display scanline. delta_ns
 	 * can be negative if start of scanout hasn't happened yet.
 	 */
-	delta_ns = (s64) vpos * linedur_ns + (s64) hpos * pixeldur_ns;
-
-	/* Is vpos outside nominal vblank area, but less than
-	 * 1/100 of a frame height away from start of vblank?
-	 * If so, assume this isn't a massively delayed vblank
-	 * interrupt, but a vblank interrupt that fired a few
-	 * microseconds before true start of vblank. Compensate
-	 * by adding a full frame duration to the final timestamp.
-	 * Happens, e.g., on ATI R500, R600.
-	 *
-	 * We only do this if DRM_CALLED_FROM_VBLIRQ.
-	 */
-	if ((flags & DRM_CALLED_FROM_VBLIRQ) && !invbl &&
-	    ((vdisplay - vpos) < vtotal / 100)) {
-		delta_ns = delta_ns - framedur_ns;
-
-		/* Signal this correction as "applied". */
-		vbl_status |= 0x8;
-	}
+	delta_ns = vpos * linedur_ns + hpos * pixeldur_ns;
 
 	if (!drm_timestamp_monotonic)
 		etime = ktime_sub(etime, mono_time_offset);
@@ -673,7 +650,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 		  crtc, (int)vbl_status, hpos, vpos,
 		  (long)tv_etime.tv_sec, (long)tv_etime.tv_usec,
 		  (long)vblank_time->tv_sec, (long)vblank_time->tv_usec,
-		  (int)duration_ns/1000, i);
+		  duration_ns/1000, i);
 
 	vbl_status = DRM_VBLANKTIME_SCANOUTPOS_METHOD;
 	if (invbl)
@@ -696,23 +673,23 @@ static struct timeval get_drm_timestamp(void)
 
 /**
  * drm_get_last_vbltimestamp - retrieve raw timestamp for the most recent
- * vblank interval.
- *
+ * 			       vblank interval
  * @dev: DRM device
- * @crtc: which crtc's vblank timestamp to retrieve
+ * @crtc: which CRTC's vblank timestamp to retrieve
  * @tvblank: Pointer to target struct timeval which should receive the timestamp
  * @flags: Flags to pass to driver:
- *         0 = Default.
- *         DRM_CALLED_FROM_VBLIRQ = If function is called from vbl irq handler.
+ *         0 = Default,
+ *         DRM_CALLED_FROM_VBLIRQ = If function is called from vbl IRQ handler
  *
  * Fetches the system timestamp corresponding to the time of the most recent
- * vblank interval on specified crtc. May call into kms-driver to
+ * vblank interval on specified CRTC. May call into kms-driver to
  * compute the timestamp with a high-precision GPU specific method.
  *
  * Returns zero if timestamp originates from uncorrected do_gettimeofday()
  * call, i.e., it isn't very precisely locked to the true vblank.
  *
- * Returns non-zero if timestamp is considered to be very precise.
+ * Returns:
+ * Non-zero if timestamp is considered to be very precise, zero otherwise.
  */
 u32 drm_get_last_vbltimestamp(struct drm_device *dev, int crtc,
 			      struct timeval *tvblank, unsigned flags)
@@ -747,6 +724,9 @@ EXPORT_SYMBOL(drm_get_last_vbltimestamp);
  * Fetches the "cooked" vblank count value that represents the number of
  * vblank events since the system was booted, including lost events due to
  * modesetting activity.
+ *
+ * Returns:
+ * The software vblank counter.
  */
 u32 drm_vblank_count(struct drm_device *dev, int crtc)
 {
@@ -765,8 +745,7 @@ EXPORT_SYMBOL(drm_vblank_count);
  * Fetches the "cooked" vblank count value that represents the number of
  * vblank events since the system was booted, including lost events due to
  * modesetting activity. Returns corresponding system timestamp of the time
- * of the vblank interval that corresponds to the current value vblank counter
- * value.
+ * of the vblank interval that corresponds to the current vblank counter value.
  */
 u32 drm_vblank_count_and_time(struct drm_device *dev, int crtc,
 			      struct timeval *vblanktime)
@@ -889,9 +868,45 @@ static void drm_update_vblank_count(struct drm_device *dev, int crtc)
 		vblanktimestamp(dev, crtc, tslot) = t_vblank;
 	}
 
-	smp_mb__before_atomic_inc();
+	smp_mb__before_atomic();
 	atomic_add(diff, &dev->vblank[crtc].count);
-	smp_mb__after_atomic_inc();
+	smp_mb__after_atomic();
+}
+
+/**
+ * drm_vblank_enable - enable the vblank interrupt on a CRTC
+ * @dev: DRM device
+ * @crtc: CRTC in question
+ */
+static int drm_vblank_enable(struct drm_device *dev, int crtc)
+{
+	int ret = 0;
+
+	assert_spin_locked(&dev->vbl_lock);
+
+	spin_lock(&dev->vblank_time_lock);
+
+	if (!dev->vblank[crtc].enabled) {
+		/*
+		 * Enable vblank irqs under vblank_time_lock protection.
+		 * All vblank count & timestamp updates are held off
+		 * until we are done reinitializing master counter and
+		 * timestamps. Filtercode in drm_handle_vblank() will
+		 * prevent double-accounting of same vblank interval.
+		 */
+		ret = dev->driver->enable_vblank(dev, crtc);
+		DRM_DEBUG("enabling vblank on crtc %d, ret: %d\n", crtc, ret);
+		if (ret)
+			atomic_dec(&dev->vblank[crtc].refcount);
+		else {
+			dev->vblank[crtc].enabled = true;
+			drm_update_vblank_count(dev, crtc);
+		}
+	}
+
+	spin_unlock(&dev->vblank_time_lock);
+
+	return ret;
 }
 
 /**
@@ -902,36 +917,20 @@ static void drm_update_vblank_count(struct drm_device *dev, int crtc)
  * Acquire a reference count on vblank events to avoid having them disabled
  * while in use.
  *
- * RETURNS
+ * This is the legacy version of drm_crtc_vblank_get().
+ *
+ * Returns:
  * Zero on success, nonzero on failure.
  */
 int drm_vblank_get(struct drm_device *dev, int crtc)
 {
-	unsigned long irqflags, irqflags2;
+	unsigned long irqflags;
 	int ret = 0;
 
 	spin_lock_irqsave(&dev->vbl_lock, irqflags);
 	/* Going from 0->1 means we have to enable interrupts again */
 	if (atomic_add_return(1, &dev->vblank[crtc].refcount) == 1) {
-		spin_lock_irqsave(&dev->vblank_time_lock, irqflags2);
-		if (!dev->vblank[crtc].enabled) {
-			/* Enable vblank irqs under vblank_time_lock protection.
-			 * All vblank count & timestamp updates are held off
-			 * until we are done reinitializing master counter and
-			 * timestamps. Filtercode in drm_handle_vblank() will
-			 * prevent double-accounting of same vblank interval.
-			 */
-			ret = dev->driver->enable_vblank(dev, crtc);
-			DRM_DEBUG("enabling vblank on crtc %d, ret: %d\n",
-				  crtc, ret);
-			if (ret)
-				atomic_dec(&dev->vblank[crtc].refcount);
-			else {
-				dev->vblank[crtc].enabled = true;
-				drm_update_vblank_count(dev, crtc);
-			}
-		}
-		spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags2);
+		ret = drm_vblank_enable(dev, crtc);
 	} else {
 		if (!dev->vblank[crtc].enabled) {
 			atomic_dec(&dev->vblank[crtc].refcount);
@@ -945,12 +944,32 @@ int drm_vblank_get(struct drm_device *dev, int crtc)
 EXPORT_SYMBOL(drm_vblank_get);
 
 /**
+ * drm_crtc_vblank_get - get a reference count on vblank events
+ * @crtc: which CRTC to own
+ *
+ * Acquire a reference count on vblank events to avoid having them disabled
+ * while in use.
+ *
+ * This is the native kms version of drm_vblank_off().
+ *
+ * Returns:
+ * Zero on success, nonzero on failure.
+ */
+int drm_crtc_vblank_get(struct drm_crtc *crtc)
+{
+	return drm_vblank_get(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_vblank_get);
+
+/**
  * drm_vblank_put - give up ownership of vblank events
  * @dev: DRM device
  * @crtc: which counter to give up
  *
  * Release ownership of a given vblank counter, turning off interrupts
  * if possible. Disable interrupts after drm_vblank_offdelay milliseconds.
+ *
+ * This is the legacy version of drm_crtc_vblank_put().
  */
 void drm_vblank_put(struct drm_device *dev, int crtc)
 {
@@ -959,17 +978,39 @@ void drm_vblank_put(struct drm_device *dev, int crtc)
 	/* Last user schedules interrupt disable */
 	if (atomic_dec_and_test(&dev->vblank[crtc].refcount) &&
 	    (drm_vblank_offdelay > 0))
-		mod_timer(&dev->vblank_disable_timer,
-			  jiffies + ((drm_vblank_offdelay * DRM_HZ)/1000));
+		mod_timer(&dev->vblank[crtc].disable_timer,
+			  jiffies + ((drm_vblank_offdelay * HZ)/1000));
 }
 EXPORT_SYMBOL(drm_vblank_put);
 
 /**
+ * drm_crtc_vblank_put - give up ownership of vblank events
+ * @crtc: which counter to give up
+ *
+ * Release ownership of a given vblank counter, turning off interrupts
+ * if possible. Disable interrupts after drm_vblank_offdelay milliseconds.
+ *
+ * This is the native kms version of drm_vblank_put().
+ */
+void drm_crtc_vblank_put(struct drm_crtc *crtc)
+{
+	drm_vblank_put(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_vblank_put);
+
+/**
  * drm_vblank_off - disable vblank events on a CRTC
  * @dev: DRM device
  * @crtc: CRTC in question
  *
- * Caller must hold event lock.
+ * Drivers can use this function to shut down the vblank interrupt handling when
+ * disabling a crtc. This function ensures that the latest vblank frame count is
+ * stored so that drm_vblank_on() can restore it again.
+ *
+ * Drivers must use this function when the hardware vblank counter can get
+ * reset, e.g. when suspending.
+ *
+ * This is the legacy version of drm_crtc_vblank_off().
  */
 void drm_vblank_off(struct drm_device *dev, int crtc)
 {
@@ -980,7 +1021,7 @@ void drm_vblank_off(struct drm_device *dev, int crtc)
 
 	spin_lock_irqsave(&dev->vbl_lock, irqflags);
 	vblank_disable_and_save(dev, crtc);
-	DRM_WAKEUP(&dev->vblank[crtc].queue);
+	wake_up(&dev->vblank[crtc].queue);
 
 	/* Send any queued vblank events, lest the natives grow disquiet */
 	seq = drm_vblank_count_and_time(dev, crtc, &now);
@@ -1003,12 +1044,87 @@ void drm_vblank_off(struct drm_device *dev, int crtc)
 EXPORT_SYMBOL(drm_vblank_off);
 
 /**
+ * drm_crtc_vblank_off - disable vblank events on a CRTC
+ * @crtc: CRTC in question
+ *
+ * Drivers can use this function to shut down the vblank interrupt handling when
+ * disabling a crtc. This function ensures that the latest vblank frame count is
+ * stored so that drm_vblank_on can restore it again.
+ *
+ * Drivers must use this function when the hardware vblank counter can get
+ * reset, e.g. when suspending.
+ *
+ * This is the native kms version of drm_vblank_off().
+ */
+void drm_crtc_vblank_off(struct drm_crtc *crtc)
+{
+	drm_vblank_off(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_vblank_off);
+
+/**
+ * drm_vblank_on - enable vblank events on a CRTC
+ * @dev: DRM device
+ * @crtc: CRTC in question
+ *
+ * This functions restores the vblank interrupt state captured with
+ * drm_vblank_off() again. Note that calls to drm_vblank_on() and
+ * drm_vblank_off() can be unbalanced and so can also be unconditionaly called
+ * in driver load code to reflect the current hardware state of the crtc.
+ *
+ * This is the legacy version of drm_crtc_vblank_on().
+ */
+void drm_vblank_on(struct drm_device *dev, int crtc)
+{
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&dev->vbl_lock, irqflags);
+	/* re-enable interrupts if there's are users left */
+	if (atomic_read(&dev->vblank[crtc].refcount) != 0)
+		WARN_ON(drm_vblank_enable(dev, crtc));
+	spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
+}
+EXPORT_SYMBOL(drm_vblank_on);
+
+/**
+ * drm_crtc_vblank_on - enable vblank events on a CRTC
+ * @crtc: CRTC in question
+ *
+ * This functions restores the vblank interrupt state captured with
+ * drm_vblank_off() again. Note that calls to drm_vblank_on() and
+ * drm_vblank_off() can be unbalanced and so can also be unconditionaly called
+ * in driver load code to reflect the current hardware state of the crtc.
+ *
+ * This is the native kms version of drm_vblank_on().
+ */
+void drm_crtc_vblank_on(struct drm_crtc *crtc)
+{
+	drm_vblank_on(crtc->dev, drm_crtc_index(crtc));
+}
+EXPORT_SYMBOL(drm_crtc_vblank_on);
+
+/**
  * drm_vblank_pre_modeset - account for vblanks across mode sets
  * @dev: DRM device
  * @crtc: CRTC in question
  *
  * Account for vblank events across mode setting events, which will likely
  * reset the hardware frame counter.
+ *
+ * This is done by grabbing a temporary vblank reference to ensure that the
+ * vblank interrupt keeps running across the modeset sequence. With this the
+ * software-side vblank frame counting will ensure that there are no jumps or
+ * discontinuities.
+ *
+ * Unfortunately this approach is racy and also doesn't work when the vblank
+ * interrupt stops running, e.g. across system suspend resume. It is therefore
+ * highly recommended that drivers use the newer drm_vblank_off() and
+ * drm_vblank_on() instead. drm_vblank_pre_modeset() only works correctly when
+ * using "cooked" software vblank frame counters and not relying on any hardware
+ * counters.
+ *
+ * Drivers must call drm_vblank_post_modeset() when re-enabling the same crtc
+ * again.
  */
 void drm_vblank_pre_modeset(struct drm_device *dev, int crtc)
 {
@@ -1030,6 +1146,14 @@ void drm_vblank_pre_modeset(struct drm_device *dev, int crtc)
 }
 EXPORT_SYMBOL(drm_vblank_pre_modeset);
 
+/**
+ * drm_vblank_post_modeset - undo drm_vblank_pre_modeset changes
+ * @dev: DRM device
+ * @crtc: CRTC in question
+ *
+ * This function again drops the temporary vblank reference acquired in
+ * drm_vblank_pre_modeset.
+ */
 void drm_vblank_post_modeset(struct drm_device *dev, int crtc)
 {
 	unsigned long irqflags;
@@ -1051,7 +1175,7 @@ void drm_vblank_post_modeset(struct drm_device *dev, int crtc)
 }
 EXPORT_SYMBOL(drm_vblank_post_modeset);
 
-/**
+/*
  * drm_modeset_ctl - handle vblank event counter changes across mode switch
  * @DRM_IOCTL_ARGS: standard ioctl arguments
  *
@@ -1164,7 +1288,7 @@ err_put:
 	return ret;
 }
 
-/**
+/*
  * Wait for VBLANK.
  *
  * \param inode device inode.
@@ -1175,7 +1299,7 @@ err_put:
  *
  * This function enables the vblank interrupt on the pipe requested, then
  * sleeps waiting for the requested sequence number to occur, and drops
- * the vblank interrupt refcount afterwards. (vblank irq disable follows that
+ * the vblank interrupt refcount afterwards. (vblank IRQ disable follows that
  * after a timeout with no further vblank waits scheduled).
  */
 int drm_wait_vblank(struct drm_device *dev, void *data,
@@ -1185,9 +1309,8 @@ int drm_wait_vblank(struct drm_device *dev, void *data,
 	int ret;
 	unsigned int flags, seq, crtc, high_crtc;
 
-	if (drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
-		if ((!drm_dev_to_irq(dev)) || (!dev->irq_enabled))
-			return -EINVAL;
+	if (!dev->irq_enabled)
+		return -EINVAL;
 
 	if (vblwait->request.type & _DRM_VBLANK_SIGNAL)
 		return -EINVAL;
@@ -1244,9 +1367,10 @@ int drm_wait_vblank(struct drm_device *dev, void *data,
 	DRM_DEBUG("waiting on vblank count %d, crtc %d\n",
 		  vblwait->request.sequence, crtc);
 	dev->vblank[crtc].last_wait = vblwait->request.sequence;
-	DRM_WAIT_ON(ret, dev->vblank[crtc].queue, 3 * DRM_HZ,
+	DRM_WAIT_ON(ret, dev->vblank[crtc].queue, 3 * HZ,
 		    (((drm_vblank_count(dev, crtc) -
 		       vblwait->request.sequence) <= (1 << 23)) ||
+		     !dev->vblank[crtc].enabled ||
 		     !dev->irq_enabled));
 
 	if (ret != -EINTR) {
@@ -1355,15 +1479,15 @@ bool drm_handle_vblank(struct drm_device *dev, int crtc)
 		/* Increment cooked vblank count. This also atomically commits
 		 * the timestamp computed above.
 		 */
-		smp_mb__before_atomic_inc();
+		smp_mb__before_atomic();
 		atomic_inc(&dev->vblank[crtc].count);
-		smp_mb__after_atomic_inc();
+		smp_mb__after_atomic();
 	} else {
 		DRM_DEBUG("crtc %d: Redundant vblirq ignored. diff_ns = %d\n",
 			  crtc, (int) diff_ns);
 	}
 
-	DRM_WAKEUP(&dev->vblank[crtc].queue);
+	wake_up(&dev->vblank[crtc].queue);
 	drm_handle_vblank_events(dev, crtc);
 
 	spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
diff --git a/drivers/gpu/drm/drm_memory.c b/drivers/gpu/drm/drm_memory.c
index 64e44fad8ae..00c67c0f238 100644
--- a/drivers/gpu/drm/drm_memory.c
+++ b/drivers/gpu/drm/drm_memory.c
@@ -82,19 +82,19 @@ static void *agp_remap(unsigned long offset, unsigned long size,
 }
 
 /** Wrapper around agp_free_memory() */
-void drm_free_agp(DRM_AGP_MEM * handle, int pages)
+void drm_free_agp(struct agp_memory * handle, int pages)
 {
 	agp_free_memory(handle);
 }
 
 /** Wrapper around agp_bind_memory() */
-int drm_bind_agp(DRM_AGP_MEM * handle, unsigned int start)
+int drm_bind_agp(struct agp_memory * handle, unsigned int start)
 {
 	return agp_bind_memory(handle, start);
 }
 
 /** Wrapper around agp_unbind_memory() */
-int drm_unbind_agp(DRM_AGP_MEM * handle)
+int drm_unbind_agp(struct agp_memory * handle)
 {
 	return agp_unbind_memory(handle);
 }
@@ -110,8 +110,7 @@ static inline void *agp_remap(unsigned long offset, unsigned long size,
 
 void drm_core_ioremap(struct drm_local_map *map, struct drm_device *dev)
 {
-	if (drm_core_has_AGP(dev) &&
-	    dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
+	if (dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
 		map->handle = agp_remap(map->offset, map->size, dev);
 	else
 		map->handle = ioremap(map->offset, map->size);
@@ -120,8 +119,7 @@ EXPORT_SYMBOL(drm_core_ioremap);
 
 void drm_core_ioremap_wc(struct drm_local_map *map, struct drm_device *dev)
 {
-	if (drm_core_has_AGP(dev) &&
-	    dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
+	if (dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
 		map->handle = agp_remap(map->offset, map->size, dev);
 	else
 		map->handle = ioremap_wc(map->offset, map->size);
@@ -133,8 +131,7 @@ void drm_core_ioremapfree(struct drm_local_map *map, struct drm_device *dev)
 	if (!map->handle || !map->size)
 		return;
 
-	if (drm_core_has_AGP(dev) &&
-	    dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
+	if (dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
 		vunmap(map->handle);
 	else
 		iounmap(map->handle);
diff --git a/drivers/gpu/drm/drm_mipi_dsi.c b/drivers/gpu/drm/drm_mipi_dsi.c
new file mode 100644
index 00000000000..e633df2f68d
--- /dev/null
+++ b/drivers/gpu/drm/drm_mipi_dsi.c
@@ -0,0 +1,329 @@
+/*
+ * MIPI DSI Bus
+ *
+ * Copyright (C) 2012-2013, Samsung Electronics, Co., Ltd.
+ * Andrzej Hajda <a.hajda@samsung.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <drm/drm_mipi_dsi.h>
+
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/slab.h>
+
+#include <video/mipi_display.h>
+
+static int mipi_dsi_device_match(struct device *dev, struct device_driver *drv)
+{
+	return of_driver_match_device(dev, drv);
+}
+
+static const struct dev_pm_ops mipi_dsi_device_pm_ops = {
+	.runtime_suspend = pm_generic_runtime_suspend,
+	.runtime_resume = pm_generic_runtime_resume,
+	.suspend = pm_generic_suspend,
+	.resume = pm_generic_resume,
+	.freeze = pm_generic_freeze,
+	.thaw = pm_generic_thaw,
+	.poweroff = pm_generic_poweroff,
+	.restore = pm_generic_restore,
+};
+
+static struct bus_type mipi_dsi_bus_type = {
+	.name = "mipi-dsi",
+	.match = mipi_dsi_device_match,
+	.pm = &mipi_dsi_device_pm_ops,
+};
+
+static void mipi_dsi_dev_release(struct device *dev)
+{
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
+
+	of_node_put(dev->of_node);
+	kfree(dsi);
+}
+
+static const struct device_type mipi_dsi_device_type = {
+	.release = mipi_dsi_dev_release,
+};
+
+static struct mipi_dsi_device *mipi_dsi_device_alloc(struct mipi_dsi_host *host)
+{
+	struct mipi_dsi_device *dsi;
+
+	dsi = kzalloc(sizeof(*dsi), GFP_KERNEL);
+	if (!dsi)
+		return ERR_PTR(-ENOMEM);
+
+	dsi->host = host;
+	dsi->dev.bus = &mipi_dsi_bus_type;
+	dsi->dev.parent = host->dev;
+	dsi->dev.type = &mipi_dsi_device_type;
+
+	device_initialize(&dsi->dev);
+
+	return dsi;
+}
+
+static int mipi_dsi_device_add(struct mipi_dsi_device *dsi)
+{
+	struct mipi_dsi_host *host = dsi->host;
+
+	dev_set_name(&dsi->dev, "%s.%d", dev_name(host->dev),  dsi->channel);
+
+	return device_add(&dsi->dev);
+}
+
+static struct mipi_dsi_device *
+of_mipi_dsi_device_add(struct mipi_dsi_host *host, struct device_node *node)
+{
+	struct mipi_dsi_device *dsi;
+	struct device *dev = host->dev;
+	int ret;
+	u32 reg;
+
+	ret = of_property_read_u32(node, "reg", &reg);
+	if (ret) {
+		dev_err(dev, "device node %s has no valid reg property: %d\n",
+			node->full_name, ret);
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (reg > 3) {
+		dev_err(dev, "device node %s has invalid reg property: %u\n",
+			node->full_name, reg);
+		return ERR_PTR(-EINVAL);
+	}
+
+	dsi = mipi_dsi_device_alloc(host);
+	if (IS_ERR(dsi)) {
+		dev_err(dev, "failed to allocate DSI device %s: %ld\n",
+			node->full_name, PTR_ERR(dsi));
+		return dsi;
+	}
+
+	dsi->dev.of_node = of_node_get(node);
+	dsi->channel = reg;
+
+	ret = mipi_dsi_device_add(dsi);
+	if (ret) {
+		dev_err(dev, "failed to add DSI device %s: %d\n",
+			node->full_name, ret);
+		kfree(dsi);
+		return ERR_PTR(ret);
+	}
+
+	return dsi;
+}
+
+int mipi_dsi_host_register(struct mipi_dsi_host *host)
+{
+	struct device_node *node;
+
+	for_each_available_child_of_node(host->dev->of_node, node) {
+		/* skip nodes without reg property */
+		if (!of_find_property(node, "reg", NULL))
+			continue;
+		of_mipi_dsi_device_add(host, node);
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_host_register);
+
+static int mipi_dsi_remove_device_fn(struct device *dev, void *priv)
+{
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
+
+	device_unregister(&dsi->dev);
+
+	return 0;
+}
+
+void mipi_dsi_host_unregister(struct mipi_dsi_host *host)
+{
+	device_for_each_child(host->dev, NULL, mipi_dsi_remove_device_fn);
+}
+EXPORT_SYMBOL(mipi_dsi_host_unregister);
+
+/**
+ * mipi_dsi_attach - attach a DSI device to its DSI host
+ * @dsi: DSI peripheral
+ */
+int mipi_dsi_attach(struct mipi_dsi_device *dsi)
+{
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+
+	if (!ops || !ops->attach)
+		return -ENOSYS;
+
+	return ops->attach(dsi->host, dsi);
+}
+EXPORT_SYMBOL(mipi_dsi_attach);
+
+/**
+ * mipi_dsi_detach - detach a DSI device from its DSI host
+ * @dsi: DSI peripheral
+ */
+int mipi_dsi_detach(struct mipi_dsi_device *dsi)
+{
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+
+	if (!ops || !ops->detach)
+		return -ENOSYS;
+
+	return ops->detach(dsi->host, dsi);
+}
+EXPORT_SYMBOL(mipi_dsi_detach);
+
+/**
+ * mipi_dsi_dcs_write - send DCS write command
+ * @dsi: DSI device
+ * @channel: virtual channel
+ * @data: pointer to the command followed by parameters
+ * @len: length of @data
+ */
+int mipi_dsi_dcs_write(struct mipi_dsi_device *dsi, unsigned int channel,
+		       const void *data, size_t len)
+{
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+	struct mipi_dsi_msg msg = {
+		.channel = channel,
+		.tx_buf = data,
+		.tx_len = len
+	};
+
+	if (!ops || !ops->transfer)
+		return -ENOSYS;
+
+	switch (len) {
+	case 0:
+		return -EINVAL;
+	case 1:
+		msg.type = MIPI_DSI_DCS_SHORT_WRITE;
+		break;
+	case 2:
+		msg.type = MIPI_DSI_DCS_SHORT_WRITE_PARAM;
+		break;
+	default:
+		msg.type = MIPI_DSI_DCS_LONG_WRITE;
+		break;
+	}
+
+	return ops->transfer(dsi->host, &msg);
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_write);
+
+/**
+ * mipi_dsi_dcs_read - send DCS read request command
+ * @dsi: DSI device
+ * @channel: virtual channel
+ * @cmd: DCS read command
+ * @data: pointer to read buffer
+ * @len: length of @data
+ *
+ * Function returns number of read bytes or error code.
+ */
+ssize_t mipi_dsi_dcs_read(struct mipi_dsi_device *dsi, unsigned int channel,
+			  u8 cmd, void *data, size_t len)
+{
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+	struct mipi_dsi_msg msg = {
+		.channel = channel,
+		.type = MIPI_DSI_DCS_READ,
+		.tx_buf = &cmd,
+		.tx_len = 1,
+		.rx_buf = data,
+		.rx_len = len
+	};
+
+	if (!ops || !ops->transfer)
+		return -ENOSYS;
+
+	return ops->transfer(dsi->host, &msg);
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_read);
+
+static int mipi_dsi_drv_probe(struct device *dev)
+{
+	struct mipi_dsi_driver *drv = to_mipi_dsi_driver(dev->driver);
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
+
+	return drv->probe(dsi);
+}
+
+static int mipi_dsi_drv_remove(struct device *dev)
+{
+	struct mipi_dsi_driver *drv = to_mipi_dsi_driver(dev->driver);
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
+
+	return drv->remove(dsi);
+}
+
+static void mipi_dsi_drv_shutdown(struct device *dev)
+{
+	struct mipi_dsi_driver *drv = to_mipi_dsi_driver(dev->driver);
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
+
+	drv->shutdown(dsi);
+}
+
+/**
+ * mipi_dsi_driver_register - register a driver for DSI devices
+ * @drv: DSI driver structure
+ */
+int mipi_dsi_driver_register(struct mipi_dsi_driver *drv)
+{
+	drv->driver.bus = &mipi_dsi_bus_type;
+	if (drv->probe)
+		drv->driver.probe = mipi_dsi_drv_probe;
+	if (drv->remove)
+		drv->driver.remove = mipi_dsi_drv_remove;
+	if (drv->shutdown)
+		drv->driver.shutdown = mipi_dsi_drv_shutdown;
+
+	return driver_register(&drv->driver);
+}
+EXPORT_SYMBOL(mipi_dsi_driver_register);
+
+/**
+ * mipi_dsi_driver_unregister - unregister a driver for DSI devices
+ * @drv: DSI driver structure
+ */
+void mipi_dsi_driver_unregister(struct mipi_dsi_driver *drv)
+{
+	driver_unregister(&drv->driver);
+}
+EXPORT_SYMBOL(mipi_dsi_driver_unregister);
+
+static int __init mipi_dsi_bus_init(void)
+{
+	return bus_register(&mipi_dsi_bus_type);
+}
+postcore_initcall(mipi_dsi_bus_init);
+
+MODULE_AUTHOR("Andrzej Hajda <a.hajda@samsung.com>");
+MODULE_DESCRIPTION("MIPI DSI Bus");
+MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/drm_mm.c b/drivers/gpu/drm/drm_mm.c
index af93cc55259..04a209e2b66 100644
--- a/drivers/gpu/drm/drm_mm.c
+++ b/drivers/gpu/drm/drm_mm.c
@@ -47,7 +47,48 @@
 #include <linux/seq_file.h>
 #include <linux/export.h>
 
-#define MM_UNUSED_TARGET 4
+/**
+ * DOC: Overview
+ *
+ * drm_mm provides a simple range allocator. The drivers are free to use the
+ * resource allocator from the linux core if it suits them, the upside of drm_mm
+ * is that it's in the DRM core. Which means that it's easier to extend for
+ * some of the crazier special purpose needs of gpus.
+ *
+ * The main data struct is &drm_mm, allocations are tracked in &drm_mm_node.
+ * Drivers are free to embed either of them into their own suitable
+ * datastructures. drm_mm itself will not do any allocations of its own, so if
+ * drivers choose not to embed nodes they need to still allocate them
+ * themselves.
+ *
+ * The range allocator also supports reservation of preallocated blocks. This is
+ * useful for taking over initial mode setting configurations from the firmware,
+ * where an object needs to be created which exactly matches the firmware's
+ * scanout target. As long as the range is still free it can be inserted anytime
+ * after the allocator is initialized, which helps with avoiding looped
+ * depencies in the driver load sequence.
+ *
+ * drm_mm maintains a stack of most recently freed holes, which of all
+ * simplistic datastructures seems to be a fairly decent approach to clustering
+ * allocations and avoiding too much fragmentation. This means free space
+ * searches are O(num_holes). Given that all the fancy features drm_mm supports
+ * something better would be fairly complex and since gfx thrashing is a fairly
+ * steep cliff not a real concern. Removing a node again is O(1).
+ *
+ * drm_mm supports a few features: Alignment and range restrictions can be
+ * supplied. Further more every &drm_mm_node has a color value (which is just an
+ * opaqua unsigned long) which in conjunction with a driver callback can be used
+ * to implement sophisticated placement restrictions. The i915 DRM driver uses
+ * this to implement guard pages between incompatible caching domains in the
+ * graphics TT.
+ *
+ * Two behaviors are supported for searching and allocating: bottom-up and top-down.
+ * The default is bottom-up. Top-down allocation can be used if the memory area
+ * has different restrictions, or just to reduce fragmentation.
+ *
+ * Finally iteration helpers to walk all nodes and all holes are provided as are
+ * some basic allocator dumpers for debugging.
+ */
 
 static struct drm_mm_node *drm_mm_search_free_generic(const struct drm_mm *mm,
 						unsigned long size,
@@ -65,7 +106,8 @@ static struct drm_mm_node *drm_mm_search_free_in_range_generic(const struct drm_
 static void drm_mm_insert_helper(struct drm_mm_node *hole_node,
 				 struct drm_mm_node *node,
 				 unsigned long size, unsigned alignment,
-				 unsigned long color)
+				 unsigned long color,
+				 enum drm_mm_allocator_flags flags)
 {
 	struct drm_mm *mm = hole_node->mm;
 	unsigned long hole_start = drm_mm_hole_node_start(hole_node);
@@ -78,12 +120,22 @@ static void drm_mm_insert_helper(struct drm_mm_node *hole_node,
 	if (mm->color_adjust)
 		mm->color_adjust(hole_node, color, &adj_start, &adj_end);
 
+	if (flags & DRM_MM_CREATE_TOP)
+		adj_start = adj_end - size;
+
 	if (alignment) {
 		unsigned tmp = adj_start % alignment;
-		if (tmp)
-			adj_start += alignment - tmp;
+		if (tmp) {
+			if (flags & DRM_MM_CREATE_TOP)
+				adj_start -= tmp;
+			else
+				adj_start += alignment - tmp;
+		}
 	}
 
+	BUG_ON(adj_start < hole_start);
+	BUG_ON(adj_end > hole_end);
+
 	if (adj_start == hole_start) {
 		hole_node->hole_follows = 0;
 		list_del(&hole_node->hole_stack);
@@ -107,6 +159,20 @@ static void drm_mm_insert_helper(struct drm_mm_node *hole_node,
 	}
 }
 
+/**
+ * drm_mm_reserve_node - insert an pre-initialized node
+ * @mm: drm_mm allocator to insert @node into
+ * @node: drm_mm_node to insert
+ *
+ * This functions inserts an already set-up drm_mm_node into the allocator,
+ * meaning that start, size and color must be set by the caller. This is useful
+ * to initialize the allocator with preallocated objects which must be set-up
+ * before the range allocator can be set-up, e.g. when taking over a firmware
+ * framebuffer.
+ *
+ * Returns:
+ * 0 on success, -ENOSPC if there's no hole where @node is.
+ */
 int drm_mm_reserve_node(struct drm_mm *mm, struct drm_mm_node *node)
 {
 	struct drm_mm_node *hole;
@@ -141,30 +207,39 @@ int drm_mm_reserve_node(struct drm_mm *mm, struct drm_mm_node *node)
 		return 0;
 	}
 
-	WARN(1, "no hole found for node 0x%lx + 0x%lx\n",
-	     node->start, node->size);
 	return -ENOSPC;
 }
 EXPORT_SYMBOL(drm_mm_reserve_node);
 
 /**
- * Search for free space and insert a preallocated memory node. Returns
- * -ENOSPC if no suitable free area is available. The preallocated memory node
- * must be cleared.
+ * drm_mm_insert_node_generic - search for space and insert @node
+ * @mm: drm_mm to allocate from
+ * @node: preallocate node to insert
+ * @size: size of the allocation
+ * @alignment: alignment of the allocation
+ * @color: opaque tag value to use for this node
+ * @sflags: flags to fine-tune the allocation search
+ * @aflags: flags to fine-tune the allocation behavior
+ *
+ * The preallocated node must be cleared to 0.
+ *
+ * Returns:
+ * 0 on success, -ENOSPC if there's no suitable hole.
  */
 int drm_mm_insert_node_generic(struct drm_mm *mm, struct drm_mm_node *node,
 			       unsigned long size, unsigned alignment,
 			       unsigned long color,
-			       enum drm_mm_search_flags flags)
+			       enum drm_mm_search_flags sflags,
+			       enum drm_mm_allocator_flags aflags)
 {
 	struct drm_mm_node *hole_node;
 
 	hole_node = drm_mm_search_free_generic(mm, size, alignment,
-					       color, flags);
+					       color, sflags);
 	if (!hole_node)
 		return -ENOSPC;
 
-	drm_mm_insert_helper(hole_node, node, size, alignment, color);
+	drm_mm_insert_helper(hole_node, node, size, alignment, color, aflags);
 	return 0;
 }
 EXPORT_SYMBOL(drm_mm_insert_node_generic);
@@ -173,7 +248,8 @@ static void drm_mm_insert_helper_range(struct drm_mm_node *hole_node,
 				       struct drm_mm_node *node,
 				       unsigned long size, unsigned alignment,
 				       unsigned long color,
-				       unsigned long start, unsigned long end)
+				       unsigned long start, unsigned long end,
+				       enum drm_mm_allocator_flags flags)
 {
 	struct drm_mm *mm = hole_node->mm;
 	unsigned long hole_start = drm_mm_hole_node_start(hole_node);
@@ -188,13 +264,20 @@ static void drm_mm_insert_helper_range(struct drm_mm_node *hole_node,
 	if (adj_end > end)
 		adj_end = end;
 
+	if (flags & DRM_MM_CREATE_TOP)
+		adj_start = adj_end - size;
+
 	if (mm->color_adjust)
 		mm->color_adjust(hole_node, color, &adj_start, &adj_end);
 
 	if (alignment) {
 		unsigned tmp = adj_start % alignment;
-		if (tmp)
-			adj_start += alignment - tmp;
+		if (tmp) {
+			if (flags & DRM_MM_CREATE_TOP)
+				adj_start -= tmp;
+			else
+				adj_start += alignment - tmp;
+		}
 	}
 
 	if (adj_start == hole_start) {
@@ -211,6 +294,8 @@ static void drm_mm_insert_helper_range(struct drm_mm_node *hole_node,
 	INIT_LIST_HEAD(&node->hole_stack);
 	list_add(&node->node_list, &hole_node->node_list);
 
+	BUG_ON(node->start < start);
+	BUG_ON(node->start < adj_start);
 	BUG_ON(node->start + node->size > adj_end);
 	BUG_ON(node->start + node->size > end);
 
@@ -222,32 +307,51 @@ static void drm_mm_insert_helper_range(struct drm_mm_node *hole_node,
 }
 
 /**
- * Search for free space and insert a preallocated memory node. Returns
- * -ENOSPC if no suitable free area is available. This is for range
- * restricted allocations. The preallocated memory node must be cleared.
+ * drm_mm_insert_node_in_range_generic - ranged search for space and insert @node
+ * @mm: drm_mm to allocate from
+ * @node: preallocate node to insert
+ * @size: size of the allocation
+ * @alignment: alignment of the allocation
+ * @color: opaque tag value to use for this node
+ * @start: start of the allowed range for this node
+ * @end: end of the allowed range for this node
+ * @sflags: flags to fine-tune the allocation search
+ * @aflags: flags to fine-tune the allocation behavior
+ *
+ * The preallocated node must be cleared to 0.
+ *
+ * Returns:
+ * 0 on success, -ENOSPC if there's no suitable hole.
  */
 int drm_mm_insert_node_in_range_generic(struct drm_mm *mm, struct drm_mm_node *node,
-					unsigned long size, unsigned alignment, unsigned long color,
+					unsigned long size, unsigned alignment,
+					unsigned long color,
 					unsigned long start, unsigned long end,
-					enum drm_mm_search_flags flags)
+					enum drm_mm_search_flags sflags,
+					enum drm_mm_allocator_flags aflags)
 {
 	struct drm_mm_node *hole_node;
 
 	hole_node = drm_mm_search_free_in_range_generic(mm,
 							size, alignment, color,
-							start, end, flags);
+							start, end, sflags);
 	if (!hole_node)
 		return -ENOSPC;
 
 	drm_mm_insert_helper_range(hole_node, node,
 				   size, alignment, color,
-				   start, end);
+				   start, end, aflags);
 	return 0;
 }
 EXPORT_SYMBOL(drm_mm_insert_node_in_range_generic);
 
 /**
- * Remove a memory node from the allocator.
+ * drm_mm_remove_node - Remove a memory node from the allocator.
+ * @node: drm_mm_node to remove
+ *
+ * This just removes a node from its drm_mm allocator. The node does not need to
+ * be cleared again before it can be re-inserted into this or any other drm_mm
+ * allocator. It is a bug to call this function on a un-allocated node.
  */
 void drm_mm_remove_node(struct drm_mm_node *node)
 {
@@ -315,7 +419,10 @@ static struct drm_mm_node *drm_mm_search_free_generic(const struct drm_mm *mm,
 	best = NULL;
 	best_size = ~0UL;
 
-	drm_mm_for_each_hole(entry, mm, adj_start, adj_end) {
+	__drm_mm_for_each_hole(entry, mm, adj_start, adj_end,
+			       flags & DRM_MM_SEARCH_BELOW) {
+		unsigned long hole_size = adj_end - adj_start;
+
 		if (mm->color_adjust) {
 			mm->color_adjust(entry, color, &adj_start, &adj_end);
 			if (adj_end <= adj_start)
@@ -328,9 +435,9 @@ static struct drm_mm_node *drm_mm_search_free_generic(const struct drm_mm *mm,
 		if (!(flags & DRM_MM_SEARCH_BEST))
 			return entry;
 
-		if (entry->size < best_size) {
+		if (hole_size < best_size) {
 			best = entry;
-			best_size = entry->size;
+			best_size = hole_size;
 		}
 	}
 
@@ -356,7 +463,10 @@ static struct drm_mm_node *drm_mm_search_free_in_range_generic(const struct drm_
 	best = NULL;
 	best_size = ~0UL;
 
-	drm_mm_for_each_hole(entry, mm, adj_start, adj_end) {
+	__drm_mm_for_each_hole(entry, mm, adj_start, adj_end,
+			       flags & DRM_MM_SEARCH_BELOW) {
+		unsigned long hole_size = adj_end - adj_start;
+
 		if (adj_start < start)
 			adj_start = start;
 		if (adj_end > end)
@@ -374,9 +484,9 @@ static struct drm_mm_node *drm_mm_search_free_in_range_generic(const struct drm_
 		if (!(flags & DRM_MM_SEARCH_BEST))
 			return entry;
 
-		if (entry->size < best_size) {
+		if (hole_size < best_size) {
 			best = entry;
-			best_size = entry->size;
+			best_size = hole_size;
 		}
 	}
 
@@ -384,7 +494,13 @@ static struct drm_mm_node *drm_mm_search_free_in_range_generic(const struct drm_
 }
 
 /**
- * Moves an allocation. To be used with embedded struct drm_mm_node.
+ * drm_mm_replace_node - move an allocation from @old to @new
+ * @old: drm_mm_node to remove from the allocator
+ * @new: drm_mm_node which should inherit @old's allocation
+ *
+ * This is useful for when drivers embed the drm_mm_node structure and hence
+ * can't move allocations by reassigning pointers. It's a combination of remove
+ * and insert with the guarantee that the allocation start will match.
  */
 void drm_mm_replace_node(struct drm_mm_node *old, struct drm_mm_node *new)
 {
@@ -402,12 +518,46 @@ void drm_mm_replace_node(struct drm_mm_node *old, struct drm_mm_node *new)
 EXPORT_SYMBOL(drm_mm_replace_node);
 
 /**
- * Initializa lru scanning.
+ * DOC: lru scan roaster
+ *
+ * Very often GPUs need to have continuous allocations for a given object. When
+ * evicting objects to make space for a new one it is therefore not most
+ * efficient when we simply start to select all objects from the tail of an LRU
+ * until there's a suitable hole: Especially for big objects or nodes that
+ * otherwise have special allocation constraints there's a good chance we evict
+ * lots of (smaller) objects unecessarily.
+ *
+ * The DRM range allocator supports this use-case through the scanning
+ * interfaces. First a scan operation needs to be initialized with
+ * drm_mm_init_scan() or drm_mm_init_scan_with_range(). The the driver adds
+ * objects to the roaster (probably by walking an LRU list, but this can be
+ * freely implemented) until a suitable hole is found or there's no further
+ * evitable object.
+ *
+ * The the driver must walk through all objects again in exactly the reverse
+ * order to restore the allocator state. Note that while the allocator is used
+ * in the scan mode no other operation is allowed.
+ *
+ * Finally the driver evicts all objects selected in the scan. Adding and
+ * removing an object is O(1), and since freeing a node is also O(1) the overall
+ * complexity is O(scanned_objects). So like the free stack which needs to be
+ * walked before a scan operation even begins this is linear in the number of
+ * objects. It doesn't seem to hurt badly.
+ */
+
+/**
+ * drm_mm_init_scan - initialize lru scanning
+ * @mm: drm_mm to scan
+ * @size: size of the allocation
+ * @alignment: alignment of the allocation
+ * @color: opaque tag value to use for the allocation
  *
  * This simply sets up the scanning routines with the parameters for the desired
- * hole.
+ * hole. Note that there's no need to specify allocation flags, since they only
+ * change the place a node is allocated from within a suitable hole.
  *
- * Warning: As long as the scan list is non-empty, no other operations than
+ * Warning:
+ * As long as the scan list is non-empty, no other operations than
  * adding/removing nodes to/from the scan list are allowed.
  */
 void drm_mm_init_scan(struct drm_mm *mm,
@@ -427,12 +577,20 @@ void drm_mm_init_scan(struct drm_mm *mm,
 EXPORT_SYMBOL(drm_mm_init_scan);
 
 /**
- * Initializa lru scanning.
+ * drm_mm_init_scan - initialize range-restricted lru scanning
+ * @mm: drm_mm to scan
+ * @size: size of the allocation
+ * @alignment: alignment of the allocation
+ * @color: opaque tag value to use for the allocation
+ * @start: start of the allowed range for the allocation
+ * @end: end of the allowed range for the allocation
  *
  * This simply sets up the scanning routines with the parameters for the desired
- * hole. This version is for range-restricted scans.
+ * hole. Note that there's no need to specify allocation flags, since they only
+ * change the place a node is allocated from within a suitable hole.
  *
- * Warning: As long as the scan list is non-empty, no other operations than
+ * Warning:
+ * As long as the scan list is non-empty, no other operations than
  * adding/removing nodes to/from the scan list are allowed.
  */
 void drm_mm_init_scan_with_range(struct drm_mm *mm,
@@ -456,12 +614,16 @@ void drm_mm_init_scan_with_range(struct drm_mm *mm,
 EXPORT_SYMBOL(drm_mm_init_scan_with_range);
 
 /**
+ * drm_mm_scan_add_block - add a node to the scan list
+ * @node: drm_mm_node to add
+ *
  * Add a node to the scan list that might be freed to make space for the desired
  * hole.
  *
- * Returns non-zero, if a hole has been found, zero otherwise.
+ * Returns:
+ * True if a hole has been found, false otherwise.
  */
-int drm_mm_scan_add_block(struct drm_mm_node *node)
+bool drm_mm_scan_add_block(struct drm_mm_node *node)
 {
 	struct drm_mm *mm = node->mm;
 	struct drm_mm_node *prev_node;
@@ -501,15 +663,16 @@ int drm_mm_scan_add_block(struct drm_mm_node *node)
 			    mm->scan_size, mm->scan_alignment)) {
 		mm->scan_hit_start = hole_start;
 		mm->scan_hit_end = hole_end;
-		return 1;
+		return true;
 	}
 
-	return 0;
+	return false;
 }
 EXPORT_SYMBOL(drm_mm_scan_add_block);
 
 /**
- * Remove a node from the scan list.
+ * drm_mm_scan_remove_block - remove a node from the scan list
+ * @node: drm_mm_node to remove
  *
  * Nodes _must_ be removed in the exact same order from the scan list as they
  * have been added, otherwise the internal state of the memory manager will be
@@ -519,10 +682,11 @@ EXPORT_SYMBOL(drm_mm_scan_add_block);
  * immediately following drm_mm_search_free with !DRM_MM_SEARCH_BEST will then
  * return the just freed block (because its at the top of the free_stack list).
  *
- * Returns one if this block should be evicted, zero otherwise. Will always
- * return zero when no hole has been found.
+ * Returns:
+ * True if this block should be evicted, false otherwise. Will always
+ * return false when no hole has been found.
  */
-int drm_mm_scan_remove_block(struct drm_mm_node *node)
+bool drm_mm_scan_remove_block(struct drm_mm_node *node)
 {
 	struct drm_mm *mm = node->mm;
 	struct drm_mm_node *prev_node;
@@ -543,7 +707,15 @@ int drm_mm_scan_remove_block(struct drm_mm_node *node)
 }
 EXPORT_SYMBOL(drm_mm_scan_remove_block);
 
-int drm_mm_clean(struct drm_mm * mm)
+/**
+ * drm_mm_clean - checks whether an allocator is clean
+ * @mm: drm_mm allocator to check
+ *
+ * Returns:
+ * True if the allocator is completely free, false if there's still a node
+ * allocated in it.
+ */
+bool drm_mm_clean(struct drm_mm * mm)
 {
 	struct list_head *head = &mm->head_node.node_list;
 
@@ -551,6 +723,14 @@ int drm_mm_clean(struct drm_mm * mm)
 }
 EXPORT_SYMBOL(drm_mm_clean);
 
+/**
+ * drm_mm_init - initialize a drm-mm allocator
+ * @mm: the drm_mm structure to initialize
+ * @start: start of the range managed by @mm
+ * @size: end of the range managed by @mm
+ *
+ * Note that @mm must be cleared to 0 before calling this function.
+ */
 void drm_mm_init(struct drm_mm * mm, unsigned long start, unsigned long size)
 {
 	INIT_LIST_HEAD(&mm->hole_stack);
@@ -572,6 +752,13 @@ void drm_mm_init(struct drm_mm * mm, unsigned long start, unsigned long size)
 }
 EXPORT_SYMBOL(drm_mm_init);
 
+/**
+ * drm_mm_takedown - clean up a drm_mm allocator
+ * @mm: drm_mm allocator to clean up
+ *
+ * Note that it is a bug to call this function on an allocator which is not
+ * clean.
+ */
 void drm_mm_takedown(struct drm_mm * mm)
 {
 	WARN(!list_empty(&mm->head_node.node_list),
@@ -597,6 +784,11 @@ static unsigned long drm_mm_debug_hole(struct drm_mm_node *entry,
 	return 0;
 }
 
+/**
+ * drm_mm_debug_table - dump allocator state to dmesg
+ * @mm: drm_mm allocator to dump
+ * @prefix: prefix to use for dumping to dmesg
+ */
 void drm_mm_debug_table(struct drm_mm *mm, const char *prefix)
 {
 	struct drm_mm_node *entry;
@@ -635,6 +827,11 @@ static unsigned long drm_mm_dump_hole(struct seq_file *m, struct drm_mm_node *en
 	return 0;
 }
 
+/**
+ * drm_mm_dump_table - dump allocator state to a seq_file
+ * @m: seq_file to dump to
+ * @mm: drm_mm allocator to dump
+ */
 int drm_mm_dump_table(struct seq_file *m, struct drm_mm *mm)
 {
 	struct drm_mm_node *entry;
diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c
index b0733153dfd..bedf1894e17 100644
--- a/drivers/gpu/drm/drm_modes.c
+++ b/drivers/gpu/drm/drm_modes.c
@@ -37,15 +37,14 @@
 #include <drm/drm_crtc.h>
 #include <video/of_videomode.h>
 #include <video/videomode.h>
+#include <drm/drm_modes.h>
+
+#include "drm_crtc_internal.h"
 
 /**
- * drm_mode_debug_printmodeline - debug print a mode
- * @dev: DRM device
+ * drm_mode_debug_printmodeline - print a mode to dmesg
  * @mode: mode to print
  *
- * LOCKING:
- * None.
- *
  * Describe @mode using DRM_DEBUG.
  */
 void drm_mode_debug_printmodeline(const struct drm_display_mode *mode)
@@ -61,18 +60,77 @@ void drm_mode_debug_printmodeline(const struct drm_display_mode *mode)
 EXPORT_SYMBOL(drm_mode_debug_printmodeline);
 
 /**
- * drm_cvt_mode -create a modeline based on CVT algorithm
+ * drm_mode_create - create a new display mode
  * @dev: DRM device
- * @hdisplay: hdisplay size
- * @vdisplay: vdisplay size
- * @vrefresh  : vrefresh rate
- * @reduced : Whether the GTF calculation is simplified
- * @interlaced:Whether the interlace is supported
  *
- * LOCKING:
- * none.
+ * Create a new, cleared drm_display_mode with kzalloc, allocate an ID for it
+ * and return it.
  *
- * return the modeline based on CVT algorithm
+ * Returns:
+ * Pointer to new mode on success, NULL on error.
+ */
+struct drm_display_mode *drm_mode_create(struct drm_device *dev)
+{
+	struct drm_display_mode *nmode;
+
+	nmode = kzalloc(sizeof(struct drm_display_mode), GFP_KERNEL);
+	if (!nmode)
+		return NULL;
+
+	if (drm_mode_object_get(dev, &nmode->base, DRM_MODE_OBJECT_MODE)) {
+		kfree(nmode);
+		return NULL;
+	}
+
+	return nmode;
+}
+EXPORT_SYMBOL(drm_mode_create);
+
+/**
+ * drm_mode_destroy - remove a mode
+ * @dev: DRM device
+ * @mode: mode to remove
+ *
+ * Release @mode's unique ID, then free it @mode structure itself using kfree.
+ */
+void drm_mode_destroy(struct drm_device *dev, struct drm_display_mode *mode)
+{
+	if (!mode)
+		return;
+
+	drm_mode_object_put(dev, &mode->base);
+
+	kfree(mode);
+}
+EXPORT_SYMBOL(drm_mode_destroy);
+
+/**
+ * drm_mode_probed_add - add a mode to a connector's probed_mode list
+ * @connector: connector the new mode
+ * @mode: mode data
+ *
+ * Add @mode to @connector's probed_mode list for later use. This list should
+ * then in a second step get filtered and all the modes actually supported by
+ * the hardware moved to the @connector's modes list.
+ */
+void drm_mode_probed_add(struct drm_connector *connector,
+			 struct drm_display_mode *mode)
+{
+	WARN_ON(!mutex_is_locked(&connector->dev->mode_config.mutex));
+
+	list_add_tail(&mode->head, &connector->probed_modes);
+}
+EXPORT_SYMBOL(drm_mode_probed_add);
+
+/**
+ * drm_cvt_mode -create a modeline based on the CVT algorithm
+ * @dev: drm device
+ * @hdisplay: hdisplay size
+ * @vdisplay: vdisplay size
+ * @vrefresh: vrefresh rate
+ * @reduced: whether to use reduced blanking
+ * @interlaced: whether to compute an interlaced mode
+ * @margins: whether to add margins (borders)
  *
  * This function is called to generate the modeline based on CVT algorithm
  * according to the hdisplay, vdisplay, vrefresh.
@@ -82,12 +140,17 @@ EXPORT_SYMBOL(drm_mode_debug_printmodeline);
  *
  * And it is copied from xf86CVTmode in xserver/hw/xfree86/modes/xf86cvt.c.
  * What I have done is to translate it by using integer calculation.
+ *
+ * Returns:
+ * The modeline based on the CVT algorithm stored in a drm_display_mode object.
+ * The display mode object is allocated with drm_mode_create(). Returns NULL
+ * when no mode could be allocated.
  */
-#define HV_FACTOR			1000
 struct drm_display_mode *drm_cvt_mode(struct drm_device *dev, int hdisplay,
 				      int vdisplay, int vrefresh,
 				      bool reduced, bool interlaced, bool margins)
 {
+#define HV_FACTOR			1000
 	/* 1) top/bottom margin size (% of height) - default: 1.8, */
 #define	CVT_MARGIN_PERCENTAGE		18
 	/* 2) character cell horizontal granularity (pixels) - default 8 */
@@ -281,23 +344,25 @@ struct drm_display_mode *drm_cvt_mode(struct drm_device *dev, int hdisplay,
 EXPORT_SYMBOL(drm_cvt_mode);
 
 /**
- * drm_gtf_mode_complex - create the modeline based on full GTF algorithm
- *
- * @dev		:drm device
- * @hdisplay	:hdisplay size
- * @vdisplay	:vdisplay size
- * @vrefresh	:vrefresh rate.
- * @interlaced	:whether the interlace is supported
- * @margins	:desired margin size
- * @GTF_[MCKJ]  :extended GTF formula parameters
- *
- * LOCKING.
- * none.
- *
- * return the modeline based on full GTF algorithm.
+ * drm_gtf_mode_complex - create the modeline based on the full GTF algorithm
+ * @dev: drm device
+ * @hdisplay: hdisplay size
+ * @vdisplay: vdisplay size
+ * @vrefresh: vrefresh rate.
+ * @interlaced: whether to compute an interlaced mode
+ * @margins: desired margin (borders) size
+ * @GTF_M: extended GTF formula parameters
+ * @GTF_2C: extended GTF formula parameters
+ * @GTF_K: extended GTF formula parameters
+ * @GTF_2J: extended GTF formula parameters
  *
  * GTF feature blocks specify C and J in multiples of 0.5, so we pass them
  * in here multiplied by two.  For a C of 40, pass in 80.
+ *
+ * Returns:
+ * The modeline based on the full GTF algorithm stored in a drm_display_mode object.
+ * The display mode object is allocated with drm_mode_create(). Returns NULL
+ * when no mode could be allocated.
  */
 struct drm_display_mode *
 drm_gtf_mode_complex(struct drm_device *dev, int hdisplay, int vdisplay,
@@ -467,17 +532,13 @@ drm_gtf_mode_complex(struct drm_device *dev, int hdisplay, int vdisplay,
 EXPORT_SYMBOL(drm_gtf_mode_complex);
 
 /**
- * drm_gtf_mode - create the modeline based on GTF algorithm
- *
- * @dev		:drm device
- * @hdisplay	:hdisplay size
- * @vdisplay	:vdisplay size
- * @vrefresh	:vrefresh rate.
- * @interlaced	:whether the interlace is supported
- * @margins	:whether the margin is supported
- *
- * LOCKING.
- * none.
+ * drm_gtf_mode - create the modeline based on the GTF algorithm
+ * @dev: drm device
+ * @hdisplay: hdisplay size
+ * @vdisplay: vdisplay size
+ * @vrefresh: vrefresh rate.
+ * @interlaced: whether to compute an interlaced mode
+ * @margins: desired margin (borders) size
  *
  * return the modeline based on GTF algorithm
  *
@@ -496,19 +557,32 @@ EXPORT_SYMBOL(drm_gtf_mode_complex);
  * C = 40
  * K = 128
  * J = 20
+ *
+ * Returns:
+ * The modeline based on the GTF algorithm stored in a drm_display_mode object.
+ * The display mode object is allocated with drm_mode_create(). Returns NULL
+ * when no mode could be allocated.
  */
 struct drm_display_mode *
 drm_gtf_mode(struct drm_device *dev, int hdisplay, int vdisplay, int vrefresh,
-	     bool lace, int margins)
+	     bool interlaced, int margins)
 {
-	return drm_gtf_mode_complex(dev, hdisplay, vdisplay, vrefresh, lace,
-				    margins, 600, 40 * 2, 128, 20 * 2);
+	return drm_gtf_mode_complex(dev, hdisplay, vdisplay, vrefresh,
+				    interlaced, margins,
+				    600, 40 * 2, 128, 20 * 2);
 }
 EXPORT_SYMBOL(drm_gtf_mode);
 
 #ifdef CONFIG_VIDEOMODE_HELPERS
-int drm_display_mode_from_videomode(const struct videomode *vm,
-				    struct drm_display_mode *dmode)
+/**
+ * drm_display_mode_from_videomode - fill in @dmode using @vm,
+ * @vm: videomode structure to use as source
+ * @dmode: drm_display_mode structure to use as destination
+ *
+ * Fills out @dmode using the display mode specified in @vm.
+ */
+void drm_display_mode_from_videomode(const struct videomode *vm,
+				     struct drm_display_mode *dmode)
 {
 	dmode->hdisplay = vm->hactive;
 	dmode->hsync_start = dmode->hdisplay + vm->hfront_porch;
@@ -538,8 +612,6 @@ int drm_display_mode_from_videomode(const struct videomode *vm,
 	if (vm->flags & DISPLAY_FLAGS_DOUBLECLK)
 		dmode->flags |= DRM_MODE_FLAG_DBLCLK;
 	drm_mode_set_name(dmode);
-
-	return 0;
 }
 EXPORT_SYMBOL_GPL(drm_display_mode_from_videomode);
 
@@ -553,6 +625,9 @@ EXPORT_SYMBOL_GPL(drm_display_mode_from_videomode);
  * This function is expensive and should only be used, if only one mode is to be
  * read from DT. To get multiple modes start with of_get_display_timings and
  * work with that instead.
+ *
+ * Returns:
+ * 0 on success, a negative errno code when no of videomode node was found.
  */
 int of_get_drm_display_mode(struct device_node *np,
 			    struct drm_display_mode *dmode, int index)
@@ -580,10 +655,8 @@ EXPORT_SYMBOL_GPL(of_get_drm_display_mode);
  * drm_mode_set_name - set the name on a mode
  * @mode: name will be set in this mode
  *
- * LOCKING:
- * None.
- *
- * Set the name of @mode to a standard format.
+ * Set the name of @mode to a standard format which is <hdisplay>x<vdisplay>
+ * with an optional 'i' suffix for interlaced modes.
  */
 void drm_mode_set_name(struct drm_display_mode *mode)
 {
@@ -595,54 +668,12 @@ void drm_mode_set_name(struct drm_display_mode *mode)
 }
 EXPORT_SYMBOL(drm_mode_set_name);
 
-/**
- * drm_mode_width - get the width of a mode
- * @mode: mode
- *
- * LOCKING:
- * None.
- *
- * Return @mode's width (hdisplay) value.
- *
- * FIXME: is this needed?
- *
- * RETURNS:
- * @mode->hdisplay
- */
-int drm_mode_width(const struct drm_display_mode *mode)
-{
-	return mode->hdisplay;
-
-}
-EXPORT_SYMBOL(drm_mode_width);
-
-/**
- * drm_mode_height - get the height of a mode
- * @mode: mode
- *
- * LOCKING:
- * None.
- *
- * Return @mode's height (vdisplay) value.
- *
- * FIXME: is this needed?
- *
- * RETURNS:
- * @mode->vdisplay
- */
-int drm_mode_height(const struct drm_display_mode *mode)
-{
-	return mode->vdisplay;
-}
-EXPORT_SYMBOL(drm_mode_height);
-
 /** drm_mode_hsync - get the hsync of a mode
  * @mode: mode
  *
- * LOCKING:
- * None.
- *
- * Return @modes's hsync rate in kHz, rounded to the nearest int.
+ * Returns:
+ * @modes's hsync rate in kHz, rounded to the nearest integer. Calculates the
+ * value first if it is not yet set.
  */
 int drm_mode_hsync(const struct drm_display_mode *mode)
 {
@@ -666,17 +697,9 @@ EXPORT_SYMBOL(drm_mode_hsync);
  * drm_mode_vrefresh - get the vrefresh of a mode
  * @mode: mode
  *
- * LOCKING:
- * None.
- *
- * Return @mode's vrefresh rate in Hz or calculate it if necessary.
- *
- * FIXME: why is this needed?  shouldn't vrefresh be set already?
- *
- * RETURNS:
- * Vertical refresh rate. It will be the result of actual value plus 0.5.
- * If it is 70.288, it will return 70Hz.
- * If it is 59.6, it will return 60Hz.
+ * Returns:
+ * @modes's vrefresh rate in Hz, rounded to the nearest integer. Calculates the
+ * value first if it is not yet set.
  */
 int drm_mode_vrefresh(const struct drm_display_mode *mode)
 {
@@ -705,14 +728,11 @@ int drm_mode_vrefresh(const struct drm_display_mode *mode)
 EXPORT_SYMBOL(drm_mode_vrefresh);
 
 /**
- * drm_mode_set_crtcinfo - set CRTC modesetting parameters
+ * drm_mode_set_crtcinfo - set CRTC modesetting timing parameters
  * @p: mode
  * @adjust_flags: a combination of adjustment flags
  *
- * LOCKING:
- * None.
- *
- * Setup the CRTC modesetting parameters for @p, adjusting if necessary.
+ * Setup the CRTC modesetting timing parameters for @p, adjusting if necessary.
  *
  * - The CRTC_INTERLACE_HALVE_V flag can be used to halve vertical timings of
  *   interlaced modes.
@@ -780,15 +800,11 @@ void drm_mode_set_crtcinfo(struct drm_display_mode *p, int adjust_flags)
 }
 EXPORT_SYMBOL(drm_mode_set_crtcinfo);
 
-
 /**
  * drm_mode_copy - copy the mode
  * @dst: mode to overwrite
  * @src: mode to copy
  *
- * LOCKING:
- * None.
- *
  * Copy an existing mode into another mode, preserving the object id and
  * list head of the destination mode.
  */
@@ -805,13 +821,14 @@ EXPORT_SYMBOL(drm_mode_copy);
 
 /**
  * drm_mode_duplicate - allocate and duplicate an existing mode
- * @m: mode to duplicate
- *
- * LOCKING:
- * None.
+ * @dev: drm_device to allocate the duplicated mode for
+ * @mode: mode to duplicate
  *
  * Just allocate a new mode, copy the existing mode into it, and return
  * a pointer to it.  Used to create new instances of established modes.
+ *
+ * Returns:
+ * Pointer to duplicated mode on success, NULL on error.
  */
 struct drm_display_mode *drm_mode_duplicate(struct drm_device *dev,
 					    const struct drm_display_mode *mode)
@@ -833,12 +850,9 @@ EXPORT_SYMBOL(drm_mode_duplicate);
  * @mode1: first mode
  * @mode2: second mode
  *
- * LOCKING:
- * None.
- *
  * Check to see if @mode1 and @mode2 are equivalent.
  *
- * RETURNS:
+ * Returns:
  * True if the modes are equal, false otherwise.
  */
 bool drm_mode_equal(const struct drm_display_mode *mode1, const struct drm_display_mode *mode2)
@@ -864,13 +878,10 @@ EXPORT_SYMBOL(drm_mode_equal);
  * @mode1: first mode
  * @mode2: second mode
  *
- * LOCKING:
- * None.
- *
  * Check to see if @mode1 and @mode2 are equivalent, but
  * don't check the pixel clocks nor the stereo layout.
  *
- * RETURNS:
+ * Returns:
  * True if the modes are equal, false otherwise.
  */
 bool drm_mode_equal_no_clocks_no_stereo(const struct drm_display_mode *mode1,
@@ -900,25 +911,19 @@ EXPORT_SYMBOL(drm_mode_equal_no_clocks_no_stereo);
  * @mode_list: list of modes to check
  * @maxX: maximum width
  * @maxY: maximum height
- * @maxPitch: max pitch
  *
- * LOCKING:
- * Caller must hold a lock protecting @mode_list.
- *
- * The DRM device (@dev) has size and pitch limits.  Here we validate the
- * modes we probed for @dev against those limits and set their status as
- * necessary.
+ * This function is a helper which can be used to validate modes against size
+ * limitations of the DRM device/connector. If a mode is too big its status
+ * memeber is updated with the appropriate validation failure code. The list
+ * itself is not changed.
  */
 void drm_mode_validate_size(struct drm_device *dev,
 			    struct list_head *mode_list,
-			    int maxX, int maxY, int maxPitch)
+			    int maxX, int maxY)
 {
 	struct drm_display_mode *mode;
 
 	list_for_each_entry(mode, mode_list, head) {
-		if (maxPitch > 0 && mode->hdisplay > maxPitch)
-			mode->status = MODE_BAD_WIDTH;
-
 		if (maxX > 0 && mode->hdisplay > maxX)
 			mode->status = MODE_VIRTUAL_X;
 
@@ -934,12 +939,10 @@ EXPORT_SYMBOL(drm_mode_validate_size);
  * @mode_list: list of modes to check
  * @verbose: be verbose about it
  *
- * LOCKING:
- * Caller must hold a lock protecting @mode_list.
- *
- * Once mode list generation is complete, a caller can use this routine to
- * remove invalid modes from a mode list.  If any of the modes have a
- * status other than %MODE_OK, they are removed from @mode_list and freed.
+ * This helper function can be used to prune a display mode list after
+ * validation has been completed. All modes who's status is not MODE_OK will be
+ * removed from the list, and if @verbose the status code and mode name is also
+ * printed to dmesg.
  */
 void drm_mode_prune_invalid(struct drm_device *dev,
 			    struct list_head *mode_list, bool verbose)
@@ -966,13 +969,10 @@ EXPORT_SYMBOL(drm_mode_prune_invalid);
  * @lh_a: list_head for first mode
  * @lh_b: list_head for second mode
  *
- * LOCKING:
- * None.
- *
  * Compare two modes, given by @lh_a and @lh_b, returning a value indicating
  * which is better.
  *
- * RETURNS:
+ * Returns:
  * Negative if @lh_a is better than @lh_b, zero if they're equivalent, or
  * positive if @lh_b is better than @lh_a.
  */
@@ -1000,12 +1000,9 @@ static int drm_mode_compare(void *priv, struct list_head *lh_a, struct list_head
 
 /**
  * drm_mode_sort - sort mode list
- * @mode_list: list to sort
+ * @mode_list: list of drm_display_mode structures to sort
  *
- * LOCKING:
- * Caller must hold a lock protecting @mode_list.
- *
- * Sort @mode_list by favorability, putting good modes first.
+ * Sort @mode_list by favorability, moving good modes to the head of the list.
  */
 void drm_mode_sort(struct list_head *mode_list)
 {
@@ -1016,21 +1013,24 @@ EXPORT_SYMBOL(drm_mode_sort);
 /**
  * drm_mode_connector_list_update - update the mode list for the connector
  * @connector: the connector to update
- *
- * LOCKING:
- * Caller must hold a lock protecting @mode_list.
+ * @merge_type_bits: whether to merge or overright type bits.
  *
  * This moves the modes from the @connector probed_modes list
  * to the actual mode list. It compares the probed mode against the current
- * list and only adds different modes. All modes unverified after this point
- * will be removed by the prune invalid modes.
+ * list and only adds different/new modes.
+ *
+ * This is just a helper functions doesn't validate any modes itself and also
+ * doesn't prune any invalid modes. Callers need to do that themselves.
  */
-void drm_mode_connector_list_update(struct drm_connector *connector)
+void drm_mode_connector_list_update(struct drm_connector *connector,
+				    bool merge_type_bits)
 {
 	struct drm_display_mode *mode;
 	struct drm_display_mode *pmode, *pt;
 	int found_it;
 
+	WARN_ON(!mutex_is_locked(&connector->dev->mode_config.mutex));
+
 	list_for_each_entry_safe(pmode, pt, &connector->probed_modes,
 				 head) {
 		found_it = 0;
@@ -1041,7 +1041,10 @@ void drm_mode_connector_list_update(struct drm_connector *connector)
 				/* if equal delete the probed mode */
 				mode->status = pmode->status;
 				/* Merge type bits together */
-				mode->type |= pmode->type;
+				if (merge_type_bits)
+					mode->type |= pmode->type;
+				else
+					mode->type = pmode->type;
 				list_del(&pmode->head);
 				drm_mode_destroy(connector->dev, pmode);
 				break;
@@ -1056,17 +1059,25 @@ void drm_mode_connector_list_update(struct drm_connector *connector)
 EXPORT_SYMBOL(drm_mode_connector_list_update);
 
 /**
- * drm_mode_parse_command_line_for_connector - parse command line for connector
- * @mode_option - per connector mode option
- * @connector - connector to parse line for
+ * drm_mode_parse_command_line_for_connector - parse command line modeline for connector
+ * @mode_option: optional per connector mode option
+ * @connector: connector to parse modeline for
+ * @mode: preallocated drm_cmdline_mode structure to fill out
+ *
+ * This parses @mode_option command line modeline for modes and options to
+ * configure the connector. If @mode_option is NULL the default command line
+ * modeline in fb_mode_option will be parsed instead.
  *
- * This parses the connector specific then generic command lines for
- * modes and options to configure the connector.
+ * This uses the same parameters as the fb modedb.c, except for an extra
+ * force-enable, force-enable-digital and force-disable bit at the end:
  *
- * This uses the same parameters as the fb modedb.c, except for extra
  *	<xres>x<yres>[M][R][-<bpp>][@<refresh>][i][m][eDd]
  *
- * enable/enable Digital/disable bit at the end
+ * The intermediate drm_cmdline_mode structure is required to store additional
+ * options from the command line modline like the force-enabel/disable flag.
+ *
+ * Returns:
+ * True if a valid modeline has been parsed, false otherwise.
  */
 bool drm_mode_parse_command_line_for_connector(const char *mode_option,
 					       struct drm_connector *connector,
@@ -1219,6 +1230,14 @@ done:
 }
 EXPORT_SYMBOL(drm_mode_parse_command_line_for_connector);
 
+/**
+ * drm_mode_create_from_cmdline_mode - convert a command line modeline into a DRM display mode
+ * @dev: DRM device to create the new mode for
+ * @cmd: input command line modeline
+ *
+ * Returns:
+ * Pointer to converted mode on success, NULL on error.
+ */
 struct drm_display_mode *
 drm_mode_create_from_cmdline_mode(struct drm_device *dev,
 				  struct drm_cmdline_mode *cmd)
diff --git a/drivers/gpu/drm/drm_modeset_lock.c b/drivers/gpu/drm/drm_modeset_lock.c
new file mode 100644
index 00000000000..0dc57d5ecd1
--- /dev/null
+++ b/drivers/gpu/drm/drm_modeset_lock.c
@@ -0,0 +1,248 @@
+/*
+ * Copyright (C) 2014 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_modeset_lock.h>
+
+/**
+ * DOC: kms locking
+ *
+ * As KMS moves toward more fine grained locking, and atomic ioctl where
+ * userspace can indirectly control locking order, it becomes necessary
+ * to use ww_mutex and acquire-contexts to avoid deadlocks.  But because
+ * the locking is more distributed around the driver code, we want a bit
+ * of extra utility/tracking out of our acquire-ctx.  This is provided
+ * by drm_modeset_lock / drm_modeset_acquire_ctx.
+ *
+ * For basic principles of ww_mutex, see: Documentation/ww-mutex-design.txt
+ *
+ * The basic usage pattern is to:
+ *
+ *     drm_modeset_acquire_init(&ctx)
+ *   retry:
+ *     foreach (lock in random_ordered_set_of_locks) {
+ *       ret = drm_modeset_lock(lock, &ctx)
+ *       if (ret == -EDEADLK) {
+ *          drm_modeset_backoff(&ctx);
+ *          goto retry;
+ *       }
+ *     }
+ *
+ *     ... do stuff ...
+ *
+ *     drm_modeset_drop_locks(&ctx);
+ *     drm_modeset_acquire_fini(&ctx);
+ */
+
+
+/**
+ * drm_modeset_acquire_init - initialize acquire context
+ * @ctx: the acquire context
+ * @flags: for future
+ */
+void drm_modeset_acquire_init(struct drm_modeset_acquire_ctx *ctx,
+		uint32_t flags)
+{
+	memset(ctx, 0, sizeof(*ctx));
+	ww_acquire_init(&ctx->ww_ctx, &crtc_ww_class);
+	INIT_LIST_HEAD(&ctx->locked);
+}
+EXPORT_SYMBOL(drm_modeset_acquire_init);
+
+/**
+ * drm_modeset_acquire_fini - cleanup acquire context
+ * @ctx: the acquire context
+ */
+void drm_modeset_acquire_fini(struct drm_modeset_acquire_ctx *ctx)
+{
+	ww_acquire_fini(&ctx->ww_ctx);
+}
+EXPORT_SYMBOL(drm_modeset_acquire_fini);
+
+/**
+ * drm_modeset_drop_locks - drop all locks
+ * @ctx: the acquire context
+ *
+ * Drop all locks currently held against this acquire context.
+ */
+void drm_modeset_drop_locks(struct drm_modeset_acquire_ctx *ctx)
+{
+	WARN_ON(ctx->contended);
+	while (!list_empty(&ctx->locked)) {
+		struct drm_modeset_lock *lock;
+
+		lock = list_first_entry(&ctx->locked,
+				struct drm_modeset_lock, head);
+
+		drm_modeset_unlock(lock);
+	}
+}
+EXPORT_SYMBOL(drm_modeset_drop_locks);
+
+static inline int modeset_lock(struct drm_modeset_lock *lock,
+		struct drm_modeset_acquire_ctx *ctx,
+		bool interruptible, bool slow)
+{
+	int ret;
+
+	WARN_ON(ctx->contended);
+
+	if (interruptible && slow) {
+		ret = ww_mutex_lock_slow_interruptible(&lock->mutex, &ctx->ww_ctx);
+	} else if (interruptible) {
+		ret = ww_mutex_lock_interruptible(&lock->mutex, &ctx->ww_ctx);
+	} else if (slow) {
+		ww_mutex_lock_slow(&lock->mutex, &ctx->ww_ctx);
+		ret = 0;
+	} else {
+		ret = ww_mutex_lock(&lock->mutex, &ctx->ww_ctx);
+	}
+	if (!ret) {
+		WARN_ON(!list_empty(&lock->head));
+		list_add(&lock->head, &ctx->locked);
+	} else if (ret == -EALREADY) {
+		/* we already hold the lock.. this is fine.  For atomic
+		 * we will need to be able to drm_modeset_lock() things
+		 * without having to keep track of what is already locked
+		 * or not.
+		 */
+		ret = 0;
+	} else if (ret == -EDEADLK) {
+		ctx->contended = lock;
+	}
+
+	return ret;
+}
+
+static int modeset_backoff(struct drm_modeset_acquire_ctx *ctx,
+		bool interruptible)
+{
+	struct drm_modeset_lock *contended = ctx->contended;
+
+	ctx->contended = NULL;
+
+	if (WARN_ON(!contended))
+		return 0;
+
+	drm_modeset_drop_locks(ctx);
+
+	return modeset_lock(contended, ctx, interruptible, true);
+}
+
+/**
+ * drm_modeset_backoff - deadlock avoidance backoff
+ * @ctx: the acquire context
+ *
+ * If deadlock is detected (ie. drm_modeset_lock() returns -EDEADLK),
+ * you must call this function to drop all currently held locks and
+ * block until the contended lock becomes available.
+ */
+void drm_modeset_backoff(struct drm_modeset_acquire_ctx *ctx)
+{
+	modeset_backoff(ctx, false);
+}
+EXPORT_SYMBOL(drm_modeset_backoff);
+
+/**
+ * drm_modeset_backoff_interruptible - deadlock avoidance backoff
+ * @ctx: the acquire context
+ *
+ * Interruptible version of drm_modeset_backoff()
+ */
+int drm_modeset_backoff_interruptible(struct drm_modeset_acquire_ctx *ctx)
+{
+	return modeset_backoff(ctx, true);
+}
+EXPORT_SYMBOL(drm_modeset_backoff_interruptible);
+
+/**
+ * drm_modeset_lock - take modeset lock
+ * @lock: lock to take
+ * @ctx: acquire ctx
+ *
+ * If ctx is not NULL, then its ww acquire context is used and the
+ * lock will be tracked by the context and can be released by calling
+ * drm_modeset_drop_locks().  If -EDEADLK is returned, this means a
+ * deadlock scenario has been detected and it is an error to attempt
+ * to take any more locks without first calling drm_modeset_backoff().
+ */
+int drm_modeset_lock(struct drm_modeset_lock *lock,
+		struct drm_modeset_acquire_ctx *ctx)
+{
+	if (ctx)
+		return modeset_lock(lock, ctx, false, false);
+
+	ww_mutex_lock(&lock->mutex, NULL);
+	return 0;
+}
+EXPORT_SYMBOL(drm_modeset_lock);
+
+/**
+ * drm_modeset_lock_interruptible - take modeset lock
+ * @lock: lock to take
+ * @ctx: acquire ctx
+ *
+ * Interruptible version of drm_modeset_lock()
+ */
+int drm_modeset_lock_interruptible(struct drm_modeset_lock *lock,
+		struct drm_modeset_acquire_ctx *ctx)
+{
+	if (ctx)
+		return modeset_lock(lock, ctx, true, false);
+
+	return ww_mutex_lock_interruptible(&lock->mutex, NULL);
+}
+EXPORT_SYMBOL(drm_modeset_lock_interruptible);
+
+/**
+ * drm_modeset_unlock - drop modeset lock
+ * @lock: lock to release
+ */
+void drm_modeset_unlock(struct drm_modeset_lock *lock)
+{
+	list_del_init(&lock->head);
+	ww_mutex_unlock(&lock->mutex);
+}
+EXPORT_SYMBOL(drm_modeset_unlock);
+
+/* Temporary.. until we have sufficiently fine grained locking, there
+ * are a couple scenarios where it is convenient to grab all crtc locks.
+ * It is planned to remove this:
+ */
+int drm_modeset_lock_all_crtcs(struct drm_device *dev,
+		struct drm_modeset_acquire_ctx *ctx)
+{
+	struct drm_mode_config *config = &dev->mode_config;
+	struct drm_crtc *crtc;
+	int ret = 0;
+
+	list_for_each_entry(crtc, &config->crtc_list, head) {
+		ret = drm_modeset_lock(&crtc->mutex, ctx);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_modeset_lock_all_crtcs);
diff --git a/drivers/gpu/drm/drm_panel.c b/drivers/gpu/drm/drm_panel.c
new file mode 100644
index 00000000000..2ef988e037b
--- /dev/null
+++ b/drivers/gpu/drm/drm_panel.c
@@ -0,0 +1,100 @@
+/*
+ * Copyright (C) 2013, NVIDIA Corporation.  All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sub license,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/err.h>
+#include <linux/module.h>
+
+#include <drm/drm_crtc.h>
+#include <drm/drm_panel.h>
+
+static DEFINE_MUTEX(panel_lock);
+static LIST_HEAD(panel_list);
+
+void drm_panel_init(struct drm_panel *panel)
+{
+	INIT_LIST_HEAD(&panel->list);
+}
+EXPORT_SYMBOL(drm_panel_init);
+
+int drm_panel_add(struct drm_panel *panel)
+{
+	mutex_lock(&panel_lock);
+	list_add_tail(&panel->list, &panel_list);
+	mutex_unlock(&panel_lock);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_panel_add);
+
+void drm_panel_remove(struct drm_panel *panel)
+{
+	mutex_lock(&panel_lock);
+	list_del_init(&panel->list);
+	mutex_unlock(&panel_lock);
+}
+EXPORT_SYMBOL(drm_panel_remove);
+
+int drm_panel_attach(struct drm_panel *panel, struct drm_connector *connector)
+{
+	if (panel->connector)
+		return -EBUSY;
+
+	panel->connector = connector;
+	panel->drm = connector->dev;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_panel_attach);
+
+int drm_panel_detach(struct drm_panel *panel)
+{
+	panel->connector = NULL;
+	panel->drm = NULL;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_panel_detach);
+
+#ifdef CONFIG_OF
+struct drm_panel *of_drm_find_panel(struct device_node *np)
+{
+	struct drm_panel *panel;
+
+	mutex_lock(&panel_lock);
+
+	list_for_each_entry(panel, &panel_list, list) {
+		if (panel->dev->of_node == np) {
+			mutex_unlock(&panel_lock);
+			return panel;
+		}
+	}
+
+	mutex_unlock(&panel_lock);
+	return NULL;
+}
+EXPORT_SYMBOL(of_drm_find_panel);
+#endif
+
+MODULE_AUTHOR("Thierry Reding <treding@nvidia.com>");
+MODULE_DESCRIPTION("DRM panel infrastructure");
+MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/drm_pci.c b/drivers/gpu/drm/drm_pci.c
index 02679793c9e..020cfd93485 100644
--- a/drivers/gpu/drm/drm_pci.c
+++ b/drivers/gpu/drm/drm_pci.c
@@ -1,17 +1,3 @@
-/* drm_pci.h -- PCI DMA memory management wrappers for DRM -*- linux-c -*- */
-/**
- * \file drm_pci.c
- * \brief Functions and ioctls to manage PCI memory
- *
- * \warning These interfaces aren't stable yet.
- *
- * \todo Implement the remaining ioctl's for the PCI pools.
- * \todo The wrappers here are so thin that they would be better off inlined..
- *
- * \author José Fonseca <jrfonseca@tungstengraphics.com>
- * \author Leif Delgass <ldelgass@retinalburn.net>
- */
-
 /*
  * Copyright 2003 José Fonseca.
  * Copyright 2003 Leif Delgass.
@@ -42,12 +28,14 @@
 #include <linux/export.h>
 #include <drm/drmP.h>
 
-/**********************************************************************/
-/** \name PCI memory */
-/*@{*/
-
 /**
- * \brief Allocate a PCI consistent memory block, for DMA.
+ * drm_pci_alloc - Allocate a PCI consistent memory block, for DMA.
+ * @dev: DRM device
+ * @size: size of block to allocate
+ * @align: alignment of block
+ *
+ * Return: A handle to the allocated memory block on success or NULL on
+ * failure.
  */
 drm_dma_handle_t *drm_pci_alloc(struct drm_device * dev, size_t size, size_t align)
 {
@@ -88,8 +76,8 @@ drm_dma_handle_t *drm_pci_alloc(struct drm_device * dev, size_t size, size_t ali
 
 EXPORT_SYMBOL(drm_pci_alloc);
 
-/**
- * \brief Free a PCI consistent memory block without freeing its descriptor.
+/*
+ * Free a PCI consistent memory block without freeing its descriptor.
  *
  * This function is for internal use in the Linux-specific DRM core code.
  */
@@ -111,7 +99,9 @@ void __drm_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
 }
 
 /**
- * \brief Free a PCI consistent memory block
+ * drm_pci_free - Free a PCI consistent memory block
+ * @dev: DRM device
+ * @dmah: handle to memory block
  */
 void drm_pci_free(struct drm_device * dev, drm_dma_handle_t * dmah)
 {
@@ -137,21 +127,9 @@ static int drm_get_pci_domain(struct drm_device *dev)
 	return pci_domain_nr(dev->pdev->bus);
 }
 
-static int drm_pci_get_irq(struct drm_device *dev)
-{
-	return dev->pdev->irq;
-}
-
-static const char *drm_pci_get_name(struct drm_device *dev)
-{
-	struct pci_driver *pdriver = dev->driver->kdriver.pci;
-	return pdriver->name;
-}
-
 static int drm_pci_set_busid(struct drm_device *dev, struct drm_master *master)
 {
 	int len, ret;
-	struct pci_driver *pdriver = dev->driver->kdriver.pci;
 	master->unique_len = 40;
 	master->unique_size = master->unique_len;
 	master->unique = kmalloc(master->unique_size, GFP_KERNEL);
@@ -173,29 +151,16 @@ static int drm_pci_set_busid(struct drm_device *dev, struct drm_master *master)
 	} else
 		master->unique_len = len;
 
-	dev->devname =
-		kmalloc(strlen(pdriver->name) +
-			master->unique_len + 2, GFP_KERNEL);
-
-	if (dev->devname == NULL) {
-		ret = -ENOMEM;
-		goto err;
-	}
-
-	sprintf(dev->devname, "%s@%s", pdriver->name,
-		master->unique);
-
 	return 0;
 err:
 	return ret;
 }
 
-static int drm_pci_set_unique(struct drm_device *dev,
-			      struct drm_master *master,
-			      struct drm_unique *u)
+int drm_pci_set_unique(struct drm_device *dev,
+		       struct drm_master *master,
+		       struct drm_unique *u)
 {
 	int domain, bus, slot, func, ret;
-	const char *bus_name;
 
 	master->unique_len = u->unique_len;
 	master->unique_size = u->unique_len + 1;
@@ -212,17 +177,6 @@ static int drm_pci_set_unique(struct drm_device *dev,
 
 	master->unique[master->unique_len] = '\0';
 
-	bus_name = dev->driver->bus->get_name(dev);
-	dev->devname = kmalloc(strlen(bus_name) +
-			       strlen(master->unique) + 2, GFP_KERNEL);
-	if (!dev->devname) {
-		ret = -ENOMEM;
-		goto err;
-	}
-
-	sprintf(dev->devname, "%s@%s", bus_name,
-		master->unique);
-
 	/* Return error if the busid submitted doesn't match the device's actual
 	 * busid.
 	 */
@@ -247,7 +201,6 @@ err:
 	return ret;
 }
 
-
 static int drm_pci_irq_by_busid(struct drm_device *dev, struct drm_irq_busid *p)
 {
 	if ((p->busnum >> 8) != drm_get_pci_domain(dev) ||
@@ -262,16 +215,41 @@ static int drm_pci_irq_by_busid(struct drm_device *dev, struct drm_irq_busid *p)
 	return 0;
 }
 
-static int drm_pci_agp_init(struct drm_device *dev)
+/**
+ * drm_irq_by_busid - Get interrupt from bus ID
+ * @dev: DRM device
+ * @data: IOCTL parameter pointing to a drm_irq_busid structure
+ * @file_priv: DRM file private.
+ *
+ * Finds the PCI device with the specified bus id and gets its IRQ number.
+ * This IOCTL is deprecated, and will now return EINVAL for any busid not equal
+ * to that of the device that this DRM instance attached to.
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int drm_irq_by_busid(struct drm_device *dev, void *data,
+		     struct drm_file *file_priv)
+{
+	struct drm_irq_busid *p = data;
+
+	if (drm_core_check_feature(dev, DRIVER_MODESET))
+		return -EINVAL;
+
+	/* UMS was only ever support on PCI devices. */
+	if (WARN_ON(!dev->pdev))
+		return -EINVAL;
+
+	if (!drm_core_check_feature(dev, DRIVER_HAVE_IRQ))
+		return -EINVAL;
+
+	return drm_pci_irq_by_busid(dev, p);
+}
+
+static void drm_pci_agp_init(struct drm_device *dev)
 {
-	if (drm_core_has_AGP(dev)) {
+	if (drm_core_check_feature(dev, DRIVER_USE_AGP)) {
 		if (drm_pci_device_is_agp(dev))
 			dev->agp = drm_agp_init(dev);
-		if (drm_core_check_feature(dev, DRIVER_REQUIRE_AGP)
-		    && (dev->agp == NULL)) {
-			DRM_ERROR("Cannot initialize the agpgart module.\n");
-			return -EINVAL;
-		}
 		if (dev->agp) {
 			dev->agp->agp_mtrr = arch_phys_wc_add(
 				dev->agp->agp_info.aper_base,
@@ -279,40 +257,33 @@ static int drm_pci_agp_init(struct drm_device *dev)
 				1024 * 1024);
 		}
 	}
-	return 0;
 }
 
-static void drm_pci_agp_destroy(struct drm_device *dev)
+void drm_pci_agp_destroy(struct drm_device *dev)
 {
-	if (drm_core_has_AGP(dev) && dev->agp) {
+	if (dev->agp) {
 		arch_phys_wc_del(dev->agp->agp_mtrr);
 		drm_agp_clear(dev);
-		drm_agp_destroy(dev->agp);
+		kfree(dev->agp);
 		dev->agp = NULL;
 	}
 }
 
 static struct drm_bus drm_pci_bus = {
-	.bus_type = DRIVER_BUS_PCI,
-	.get_irq = drm_pci_get_irq,
-	.get_name = drm_pci_get_name,
 	.set_busid = drm_pci_set_busid,
-	.set_unique = drm_pci_set_unique,
-	.irq_by_busid = drm_pci_irq_by_busid,
-	.agp_init = drm_pci_agp_init,
-	.agp_destroy = drm_pci_agp_destroy,
 };
 
 /**
- * Register.
- *
- * \param pdev - PCI device structure
- * \param ent entry from the PCI ID table with device type flags
- * \return zero on success or a negative number on failure.
+ * drm_get_pci_dev - Register a PCI device with the DRM subsystem
+ * @pdev: PCI device
+ * @ent: entry from the PCI ID table that matches @pdev
+ * @driver: DRM device driver
  *
  * Attempt to gets inter module "drm" information. If we are first
  * then register the character device and inter module information.
  * Try and register, if we fail to register, backout previous work.
+ *
+ * Return: 0 on success or a negative error code on failure.
  */
 int drm_get_pci_dev(struct pci_dev *pdev, const struct pci_device_id *ent,
 		    struct drm_driver *driver)
@@ -338,34 +309,41 @@ int drm_get_pci_dev(struct pci_dev *pdev, const struct pci_device_id *ent,
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		pci_set_drvdata(pdev, dev);
 
+	drm_pci_agp_init(dev);
+
 	ret = drm_dev_register(dev, ent->driver_data);
 	if (ret)
-		goto err_pci;
+		goto err_agp;
 
 	DRM_INFO("Initialized %s %d.%d.%d %s for %s on minor %d\n",
 		 driver->name, driver->major, driver->minor, driver->patchlevel,
 		 driver->date, pci_name(pdev), dev->primary->index);
 
+	/* No locking needed since shadow-attach is single-threaded since it may
+	 * only be called from the per-driver module init hook. */
+	if (!drm_core_check_feature(dev, DRIVER_MODESET))
+		list_add_tail(&dev->legacy_dev_list, &driver->legacy_dev_list);
+
 	return 0;
 
-err_pci:
+err_agp:
+	drm_pci_agp_destroy(dev);
 	pci_disable_device(pdev);
 err_free:
-	drm_dev_free(dev);
+	drm_dev_unref(dev);
 	return ret;
 }
 EXPORT_SYMBOL(drm_get_pci_dev);
 
 /**
- * PCI device initialization. Called direct from modules at load time.
+ * drm_pci_init - Register matching PCI devices with the DRM subsystem
+ * @driver: DRM device driver
+ * @pdriver: PCI device driver
  *
- * \return zero on success or a negative number on failure.
+ * Initializes a drm_device structures, registering the stubs and initializing
+ * the AGP device.
  *
- * Initializes a drm_device structures,registering the
- * stubs and initializing the AGP device.
- *
- * Expands the \c DRIVER_PREINIT and \c DRIVER_POST_INIT macros before and
- * after the initialization for driver customization.
+ * Return: 0 on success or a negative error code on failure.
  */
 int drm_pci_init(struct drm_driver *driver, struct pci_driver *pdriver)
 {
@@ -375,14 +353,13 @@ int drm_pci_init(struct drm_driver *driver, struct pci_driver *pdriver)
 
 	DRM_DEBUG("\n");
 
-	INIT_LIST_HEAD(&driver->device_list);
-	driver->kdriver.pci = pdriver;
 	driver->bus = &drm_pci_bus;
 
 	if (driver->driver_features & DRIVER_MODESET)
 		return pci_register_driver(pdriver);
 
 	/* If not using KMS, fall back to stealth mode manual scanning. */
+	INIT_LIST_HEAD(&driver->legacy_dev_list);
 	for (i = 0; pdriver->id_table[i].vendor != 0; i++) {
 		pid = &pdriver->id_table[i];
 
@@ -452,11 +429,32 @@ int drm_pci_init(struct drm_driver *driver, struct pci_driver *pdriver)
 	return -1;
 }
 
+void drm_pci_agp_destroy(struct drm_device *dev) {}
+
+int drm_irq_by_busid(struct drm_device *dev, void *data,
+		     struct drm_file *file_priv)
+{
+	return -EINVAL;
+}
+
+int drm_pci_set_unique(struct drm_device *dev,
+		       struct drm_master *master,
+		       struct drm_unique *u)
+{
+	return -EINVAL;
+}
 #endif
 
 EXPORT_SYMBOL(drm_pci_init);
 
-/*@}*/
+/**
+ * drm_pci_exit - Unregister matching PCI devices from the DRM subsystem
+ * @driver: DRM device driver
+ * @pdriver: PCI device driver
+ *
+ * Unregisters one or more devices matched by a PCI driver from the DRM
+ * subsystem.
+ */
 void drm_pci_exit(struct drm_driver *driver, struct pci_driver *pdriver)
 {
 	struct drm_device *dev, *tmp;
@@ -465,8 +463,11 @@ void drm_pci_exit(struct drm_driver *driver, struct pci_driver *pdriver)
 	if (driver->driver_features & DRIVER_MODESET) {
 		pci_unregister_driver(pdriver);
 	} else {
-		list_for_each_entry_safe(dev, tmp, &driver->device_list, driver_item)
+		list_for_each_entry_safe(dev, tmp, &driver->legacy_dev_list,
+					 legacy_dev_list) {
+			list_del(&dev->legacy_dev_list);
 			drm_put_dev(dev);
+		}
 	}
 	DRM_INFO("Module unloaded\n");
 }
diff --git a/drivers/gpu/drm/drm_plane_helper.c b/drivers/gpu/drm/drm_plane_helper.c
new file mode 100644
index 00000000000..6d133149cc7
--- /dev/null
+++ b/drivers/gpu/drm/drm_plane_helper.c
@@ -0,0 +1,370 @@
+/*
+ * Copyright (C) 2014 Intel Corporation
+ *
+ * DRM universal plane helper functions
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <linux/list.h>
+#include <drm/drmP.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_rect.h>
+#include <drm/drm_plane_helper.h>
+
+#define SUBPIXEL_MASK 0xffff
+
+/*
+ * This is the minimal list of formats that seem to be safe for modeset use
+ * with all current DRM drivers.  Most hardware can actually support more
+ * formats than this and drivers may specify a more accurate list when
+ * creating the primary plane.  However drivers that still call
+ * drm_plane_init() will use this minimal format list as the default.
+ */
+static const uint32_t safe_modeset_formats[] = {
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+};
+
+/*
+ * Returns the connectors currently associated with a CRTC.  This function
+ * should be called twice:  once with a NULL connector list to retrieve
+ * the list size, and once with the properly allocated list to be filled in.
+ */
+static int get_connectors_for_crtc(struct drm_crtc *crtc,
+				   struct drm_connector **connector_list,
+				   int num_connectors)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_connector *connector;
+	int count = 0;
+
+	/*
+	 * Note: Once we change the plane hooks to more fine-grained locking we
+	 * need to grab the connection_mutex here to be able to make these
+	 * checks.
+	 */
+	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
+		if (connector->encoder && connector->encoder->crtc == crtc) {
+			if (connector_list != NULL && count < num_connectors)
+				*(connector_list++) = connector;
+
+			count++;
+		}
+
+	return count;
+}
+
+/**
+ * drm_plane_helper_check_update() - Check plane update for validity
+ * @plane: plane object to update
+ * @crtc: owning CRTC of owning plane
+ * @fb: framebuffer to flip onto plane
+ * @src: source coordinates in 16.16 fixed point
+ * @dest: integer destination coordinates
+ * @clip: integer clipping coordinates
+ * @min_scale: minimum @src:@dest scaling factor in 16.16 fixed point
+ * @max_scale: maximum @src:@dest scaling factor in 16.16 fixed point
+ * @can_position: is it legal to position the plane such that it
+ *                doesn't cover the entire crtc?  This will generally
+ *                only be false for primary planes.
+ * @can_update_disabled: can the plane be updated while the crtc
+ *                       is disabled?
+ * @visible: output parameter indicating whether plane is still visible after
+ *           clipping
+ *
+ * Checks that a desired plane update is valid.  Drivers that provide
+ * their own plane handling rather than helper-provided implementations may
+ * still wish to call this function to avoid duplication of error checking
+ * code.
+ *
+ * RETURNS:
+ * Zero if update appears valid, error code on failure
+ */
+int drm_plane_helper_check_update(struct drm_plane *plane,
+				    struct drm_crtc *crtc,
+				    struct drm_framebuffer *fb,
+				    struct drm_rect *src,
+				    struct drm_rect *dest,
+				    const struct drm_rect *clip,
+				    int min_scale,
+				    int max_scale,
+				    bool can_position,
+				    bool can_update_disabled,
+				    bool *visible)
+{
+	int hscale, vscale;
+
+	if (!crtc->enabled && !can_update_disabled) {
+		DRM_DEBUG_KMS("Cannot update plane of a disabled CRTC.\n");
+		return -EINVAL;
+	}
+
+	/* Check scaling */
+	hscale = drm_rect_calc_hscale(src, dest, min_scale, max_scale);
+	vscale = drm_rect_calc_vscale(src, dest, min_scale, max_scale);
+	if (hscale < 0 || vscale < 0) {
+		DRM_DEBUG_KMS("Invalid scaling of plane\n");
+		return -ERANGE;
+	}
+
+	*visible = drm_rect_clip_scaled(src, dest, clip, hscale, vscale);
+	if (!*visible)
+		/*
+		 * Plane isn't visible; some drivers can handle this
+		 * so we just return success here.  Drivers that can't
+		 * (including those that use the primary plane helper's
+		 * update function) will return an error from their
+		 * update_plane handler.
+		 */
+		return 0;
+
+	if (!can_position && !drm_rect_equals(dest, clip)) {
+		DRM_DEBUG_KMS("Plane must cover entire CRTC\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_plane_helper_check_update);
+
+/**
+ * drm_primary_helper_update() - Helper for primary plane update
+ * @plane: plane object to update
+ * @crtc: owning CRTC of owning plane
+ * @fb: framebuffer to flip onto plane
+ * @crtc_x: x offset of primary plane on crtc
+ * @crtc_y: y offset of primary plane on crtc
+ * @crtc_w: width of primary plane rectangle on crtc
+ * @crtc_h: height of primary plane rectangle on crtc
+ * @src_x: x offset of @fb for panning
+ * @src_y: y offset of @fb for panning
+ * @src_w: width of source rectangle in @fb
+ * @src_h: height of source rectangle in @fb
+ *
+ * Provides a default plane update handler for primary planes.  This is handler
+ * is called in response to a userspace SetPlane operation on the plane with a
+ * non-NULL framebuffer.  We call the driver's modeset handler to update the
+ * framebuffer.
+ *
+ * SetPlane() on a primary plane of a disabled CRTC is not supported, and will
+ * return an error.
+ *
+ * Note that we make some assumptions about hardware limitations that may not be
+ * true for all hardware --
+ *   1) Primary plane cannot be repositioned.
+ *   2) Primary plane cannot be scaled.
+ *   3) Primary plane must cover the entire CRTC.
+ *   4) Subpixel positioning is not supported.
+ * Drivers for hardware that don't have these restrictions can provide their
+ * own implementation rather than using this helper.
+ *
+ * RETURNS:
+ * Zero on success, error code on failure
+ */
+int drm_primary_helper_update(struct drm_plane *plane, struct drm_crtc *crtc,
+			      struct drm_framebuffer *fb,
+			      int crtc_x, int crtc_y,
+			      unsigned int crtc_w, unsigned int crtc_h,
+			      uint32_t src_x, uint32_t src_y,
+			      uint32_t src_w, uint32_t src_h)
+{
+	struct drm_mode_set set = {
+		.crtc = crtc,
+		.fb = fb,
+		.mode = &crtc->mode,
+		.x = src_x >> 16,
+		.y = src_y >> 16,
+	};
+	struct drm_rect src = {
+		.x1 = src_x,
+		.y1 = src_y,
+		.x2 = src_x + src_w,
+		.y2 = src_y + src_h,
+	};
+	struct drm_rect dest = {
+		.x1 = crtc_x,
+		.y1 = crtc_y,
+		.x2 = crtc_x + crtc_w,
+		.y2 = crtc_y + crtc_h,
+	};
+	const struct drm_rect clip = {
+		.x2 = crtc->mode.hdisplay,
+		.y2 = crtc->mode.vdisplay,
+	};
+	struct drm_connector **connector_list;
+	int num_connectors, ret;
+	bool visible;
+
+	ret = drm_plane_helper_check_update(plane, crtc, fb,
+					    &src, &dest, &clip,
+					    DRM_PLANE_HELPER_NO_SCALING,
+					    DRM_PLANE_HELPER_NO_SCALING,
+					    false, false, &visible);
+	if (ret)
+		return ret;
+
+	if (!visible)
+		/*
+		 * Primary plane isn't visible.  Note that unless a driver
+		 * provides their own disable function, this will just
+		 * wind up returning -EINVAL to userspace.
+		 */
+		return plane->funcs->disable_plane(plane);
+
+	/* Find current connectors for CRTC */
+	num_connectors = get_connectors_for_crtc(crtc, NULL, 0);
+	BUG_ON(num_connectors == 0);
+	connector_list = kzalloc(num_connectors * sizeof(*connector_list),
+				 GFP_KERNEL);
+	if (!connector_list)
+		return -ENOMEM;
+	get_connectors_for_crtc(crtc, connector_list, num_connectors);
+
+	set.connectors = connector_list;
+	set.num_connectors = num_connectors;
+
+	/*
+	 * We call set_config() directly here rather than using
+	 * drm_mode_set_config_internal.  We're reprogramming the same
+	 * connectors that were already in use, so we shouldn't need the extra
+	 * cross-CRTC fb refcounting to accomodate stealing connectors.
+	 * drm_mode_setplane() already handles the basic refcounting for the
+	 * framebuffers involved in this operation.
+	 */
+	ret = crtc->funcs->set_config(&set);
+
+	kfree(connector_list);
+	return ret;
+}
+EXPORT_SYMBOL(drm_primary_helper_update);
+
+/**
+ * drm_primary_helper_disable() - Helper for primary plane disable
+ * @plane: plane to disable
+ *
+ * Provides a default plane disable handler for primary planes.  This is handler
+ * is called in response to a userspace SetPlane operation on the plane with a
+ * NULL framebuffer parameter.  It unconditionally fails the disable call with
+ * -EINVAL the only way to disable the primary plane without driver support is
+ * to disable the entier CRTC. Which does not match the plane ->disable hook.
+ *
+ * Note that some hardware may be able to disable the primary plane without
+ * disabling the whole CRTC.  Drivers for such hardware should provide their
+ * own disable handler that disables just the primary plane (and they'll likely
+ * need to provide their own update handler as well to properly re-enable a
+ * disabled primary plane).
+ *
+ * RETURNS:
+ * Unconditionally returns -EINVAL.
+ */
+int drm_primary_helper_disable(struct drm_plane *plane)
+{
+	return -EINVAL;
+}
+EXPORT_SYMBOL(drm_primary_helper_disable);
+
+/**
+ * drm_primary_helper_destroy() - Helper for primary plane destruction
+ * @plane: plane to destroy
+ *
+ * Provides a default plane destroy handler for primary planes.  This handler
+ * is called during CRTC destruction.  We disable the primary plane, remove
+ * it from the DRM plane list, and deallocate the plane structure.
+ */
+void drm_primary_helper_destroy(struct drm_plane *plane)
+{
+	drm_plane_cleanup(plane);
+	kfree(plane);
+}
+EXPORT_SYMBOL(drm_primary_helper_destroy);
+
+const struct drm_plane_funcs drm_primary_helper_funcs = {
+	.update_plane = drm_primary_helper_update,
+	.disable_plane = drm_primary_helper_disable,
+	.destroy = drm_primary_helper_destroy,
+};
+EXPORT_SYMBOL(drm_primary_helper_funcs);
+
+/**
+ * drm_primary_helper_create_plane() - Create a generic primary plane
+ * @dev: drm device
+ * @formats: pixel formats supported, or NULL for a default safe list
+ * @num_formats: size of @formats; ignored if @formats is NULL
+ *
+ * Allocates and initializes a primary plane that can be used with the primary
+ * plane helpers.  Drivers that wish to use driver-specific plane structures or
+ * provide custom handler functions may perform their own allocation and
+ * initialization rather than calling this function.
+ */
+struct drm_plane *drm_primary_helper_create_plane(struct drm_device *dev,
+						  const uint32_t *formats,
+						  int num_formats)
+{
+	struct drm_plane *primary;
+	int ret;
+
+	primary = kzalloc(sizeof(*primary), GFP_KERNEL);
+	if (primary == NULL) {
+		DRM_DEBUG_KMS("Failed to allocate primary plane\n");
+		return NULL;
+	}
+
+	if (formats == NULL) {
+		formats = safe_modeset_formats;
+		num_formats = ARRAY_SIZE(safe_modeset_formats);
+	}
+
+	/* possible_crtc's will be filled in later by crtc_init */
+	ret = drm_plane_init(dev, primary, 0, &drm_primary_helper_funcs,
+			     formats, num_formats,
+			     DRM_PLANE_TYPE_PRIMARY);
+	if (ret) {
+		kfree(primary);
+		primary = NULL;
+	}
+
+	return primary;
+}
+EXPORT_SYMBOL(drm_primary_helper_create_plane);
+
+/**
+ * drm_crtc_init - Legacy CRTC initialization function
+ * @dev: DRM device
+ * @crtc: CRTC object to init
+ * @funcs: callbacks for the new CRTC
+ *
+ * Initialize a CRTC object with a default helper-provided primary plane and no
+ * cursor plane.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int drm_crtc_init(struct drm_device *dev, struct drm_crtc *crtc,
+		  const struct drm_crtc_funcs *funcs)
+{
+	struct drm_plane *primary;
+
+	primary = drm_primary_helper_create_plane(dev, NULL, 0);
+	return drm_crtc_init_with_planes(dev, crtc, primary, NULL, funcs);
+}
+EXPORT_SYMBOL(drm_crtc_init);
diff --git a/drivers/gpu/drm/drm_platform.c b/drivers/gpu/drm/drm_platform.c
index fc24fee8ec8..d5b76f148c1 100644
--- a/drivers/gpu/drm/drm_platform.c
+++ b/drivers/gpu/drm/drm_platform.c
@@ -64,20 +64,10 @@ static int drm_get_platform_dev(struct platform_device *platdev,
 	return 0;
 
 err_free:
-	drm_dev_free(dev);
+	drm_dev_unref(dev);
 	return ret;
 }
 
-static int drm_platform_get_irq(struct drm_device *dev)
-{
-	return platform_get_irq(dev->platformdev, 0);
-}
-
-static const char *drm_platform_get_name(struct drm_device *dev)
-{
-	return dev->platformdev->name;
-}
-
 static int drm_platform_set_busid(struct drm_device *dev, struct drm_master *master)
 {
 	int len, ret, id;
@@ -106,59 +96,31 @@ static int drm_platform_set_busid(struct drm_device *dev, struct drm_master *mas
 		goto err;
 	}
 
-	dev->devname =
-		kmalloc(strlen(dev->platformdev->name) +
-			master->unique_len + 2, GFP_KERNEL);
-
-	if (dev->devname == NULL) {
-		ret = -ENOMEM;
-		goto err;
-	}
-
-	sprintf(dev->devname, "%s@%s", dev->platformdev->name,
-		master->unique);
 	return 0;
 err:
 	return ret;
 }
 
 static struct drm_bus drm_platform_bus = {
-	.bus_type = DRIVER_BUS_PLATFORM,
-	.get_irq = drm_platform_get_irq,
-	.get_name = drm_platform_get_name,
 	.set_busid = drm_platform_set_busid,
 };
 
 /**
- * Platform device initialization. Called direct from modules.
- *
- * \return zero on success or a negative number on failure.
+ * drm_platform_init - Register a platform device with the DRM subsystem
+ * @driver: DRM device driver
+ * @platform_device: platform device to register
  *
- * Initializes a drm_device structures,registering the
- * stubs
+ * Registers the specified DRM device driver and platform device with the DRM
+ * subsystem, initializing a drm_device structure and calling the driver's
+ * .load() function.
  *
- * Expands the \c DRIVER_PREINIT and \c DRIVER_POST_INIT macros before and
- * after the initialization for driver customization.
+ * Return: 0 on success or a negative error code on failure.
  */
-
 int drm_platform_init(struct drm_driver *driver, struct platform_device *platform_device)
 {
 	DRM_DEBUG("\n");
 
-	driver->kdriver.platform_device = platform_device;
 	driver->bus = &drm_platform_bus;
-	INIT_LIST_HEAD(&driver->device_list);
 	return drm_get_platform_dev(platform_device, driver);
 }
 EXPORT_SYMBOL(drm_platform_init);
-
-void drm_platform_exit(struct drm_driver *driver, struct platform_device *platform_device)
-{
-	struct drm_device *dev, *tmp;
-	DRM_DEBUG("\n");
-
-	list_for_each_entry_safe(dev, tmp, &driver->device_list, driver_item)
-		drm_put_dev(dev);
-	DRM_INFO("Module unloaded\n");
-}
-EXPORT_SYMBOL(drm_platform_exit);
diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c
index 56805c39c90..304ca8cacbc 100644
--- a/drivers/gpu/drm/drm_prime.c
+++ b/drivers/gpu/drm/drm_prime.c
@@ -68,7 +68,8 @@ struct drm_prime_attachment {
 	enum dma_data_direction dir;
 };
 
-static int drm_prime_add_buf_handle(struct drm_prime_file_private *prime_fpriv, struct dma_buf *dma_buf, uint32_t handle)
+static int drm_prime_add_buf_handle(struct drm_prime_file_private *prime_fpriv,
+				    struct dma_buf *dma_buf, uint32_t handle)
 {
 	struct drm_prime_member *member;
 
@@ -174,7 +175,7 @@ void drm_prime_remove_buf_handle_locked(struct drm_prime_file_private *prime_fpr
 }
 
 static struct sg_table *drm_gem_map_dma_buf(struct dma_buf_attachment *attach,
-		enum dma_data_direction dir)
+					    enum dma_data_direction dir)
 {
 	struct drm_prime_attachment *prime_attach = attach->priv;
 	struct drm_gem_object *obj = attach->dmabuf->priv;
@@ -211,11 +212,19 @@ static struct sg_table *drm_gem_map_dma_buf(struct dma_buf_attachment *attach,
 }
 
 static void drm_gem_unmap_dma_buf(struct dma_buf_attachment *attach,
-		struct sg_table *sgt, enum dma_data_direction dir)
+				  struct sg_table *sgt,
+				  enum dma_data_direction dir)
 {
 	/* nothing to be done here */
 }
 
+/**
+ * drm_gem_dmabuf_release - dma_buf release implementation for GEM
+ * @dma_buf: buffer to be released
+ *
+ * Generic release function for dma_bufs exported as PRIME buffers. GEM drivers
+ * must use this in their dma_buf ops structure as the release callback.
+ */
 void drm_gem_dmabuf_release(struct dma_buf *dma_buf)
 {
 	struct drm_gem_object *obj = dma_buf->priv;
@@ -242,30 +251,30 @@ static void drm_gem_dmabuf_vunmap(struct dma_buf *dma_buf, void *vaddr)
 }
 
 static void *drm_gem_dmabuf_kmap_atomic(struct dma_buf *dma_buf,
-		unsigned long page_num)
+					unsigned long page_num)
 {
 	return NULL;
 }
 
 static void drm_gem_dmabuf_kunmap_atomic(struct dma_buf *dma_buf,
-		unsigned long page_num, void *addr)
+					 unsigned long page_num, void *addr)
 {
 
 }
 static void *drm_gem_dmabuf_kmap(struct dma_buf *dma_buf,
-		unsigned long page_num)
+				 unsigned long page_num)
 {
 	return NULL;
 }
 
 static void drm_gem_dmabuf_kunmap(struct dma_buf *dma_buf,
-		unsigned long page_num, void *addr)
+				  unsigned long page_num, void *addr)
 {
 
 }
 
 static int drm_gem_dmabuf_mmap(struct dma_buf *dma_buf,
-		struct vm_area_struct *vma)
+			       struct vm_area_struct *vma)
 {
 	struct drm_gem_object *obj = dma_buf->priv;
 	struct drm_device *dev = obj->dev;
@@ -315,6 +324,15 @@ static const struct dma_buf_ops drm_gem_prime_dmabuf_ops =  {
  *    driver's scatter/gather table
  */
 
+/**
+ * drm_gem_prime_export - helper library implemention of the export callback
+ * @dev: drm_device to export from
+ * @obj: GEM object to export
+ * @flags: flags like DRM_CLOEXEC
+ *
+ * This is the implementation of the gem_prime_export functions for GEM drivers
+ * using the PRIME helpers.
+ */
 struct dma_buf *drm_gem_prime_export(struct drm_device *dev,
 				     struct drm_gem_object *obj, int flags)
 {
@@ -355,9 +373,23 @@ static struct dma_buf *export_and_register_object(struct drm_device *dev,
 	return dmabuf;
 }
 
+/**
+ * drm_gem_prime_handle_to_fd - PRIME export function for GEM drivers
+ * @dev: dev to export the buffer from
+ * @file_priv: drm file-private structure
+ * @handle: buffer handle to export
+ * @flags: flags like DRM_CLOEXEC
+ * @prime_fd: pointer to storage for the fd id of the create dma-buf
+ *
+ * This is the PRIME export function which must be used mandatorily by GEM
+ * drivers to ensure correct lifetime management of the underlying GEM object.
+ * The actual exporting from GEM object to a dma-buf is done through the
+ * gem_prime_export driver callback.
+ */
 int drm_gem_prime_handle_to_fd(struct drm_device *dev,
-		struct drm_file *file_priv, uint32_t handle, uint32_t flags,
-		int *prime_fd)
+			       struct drm_file *file_priv, uint32_t handle,
+			       uint32_t flags,
+			       int *prime_fd)
 {
 	struct drm_gem_object *obj;
 	int ret = 0;
@@ -441,6 +473,14 @@ out_unlock:
 }
 EXPORT_SYMBOL(drm_gem_prime_handle_to_fd);
 
+/**
+ * drm_gem_prime_import - helper library implemention of the import callback
+ * @dev: drm_device to import into
+ * @dma_buf: dma-buf object to import
+ *
+ * This is the implementation of the gem_prime_import functions for GEM drivers
+ * using the PRIME helpers.
+ */
 struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev,
 					    struct dma_buf *dma_buf)
 {
@@ -471,7 +511,7 @@ struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev,
 	get_dma_buf(dma_buf);
 
 	sgt = dma_buf_map_attachment(attach, DMA_BIDIRECTIONAL);
-	if (IS_ERR_OR_NULL(sgt)) {
+	if (IS_ERR(sgt)) {
 		ret = PTR_ERR(sgt);
 		goto fail_detach;
 	}
@@ -496,8 +536,21 @@ fail_detach:
 }
 EXPORT_SYMBOL(drm_gem_prime_import);
 
+/**
+ * drm_gem_prime_fd_to_handle - PRIME import function for GEM drivers
+ * @dev: dev to export the buffer from
+ * @file_priv: drm file-private structure
+ * @prime_fd: fd id of the dma-buf which should be imported
+ * @handle: pointer to storage for the handle of the imported buffer object
+ *
+ * This is the PRIME import function which must be used mandatorily by GEM
+ * drivers to ensure correct lifetime management of the underlying GEM object.
+ * The actual importing of GEM object from the dma-buf is done through the
+ * gem_import_export driver callback.
+ */
 int drm_gem_prime_fd_to_handle(struct drm_device *dev,
-		struct drm_file *file_priv, int prime_fd, uint32_t *handle)
+			       struct drm_file *file_priv, int prime_fd,
+			       uint32_t *handle)
 {
 	struct dma_buf *dma_buf;
 	struct drm_gem_object *obj;
@@ -598,12 +651,14 @@ int drm_prime_fd_to_handle_ioctl(struct drm_device *dev, void *data,
 			args->fd, &args->handle);
 }
 
-/*
- * drm_prime_pages_to_sg
+/**
+ * drm_prime_pages_to_sg - converts a page array into an sg list
+ * @pages: pointer to the array of page pointers to convert
+ * @nr_pages: length of the page vector
  *
- * this helper creates an sg table object from a set of pages
+ * This helper creates an sg table object from a set of pages
  * the driver is responsible for mapping the pages into the
- * importers address space
+ * importers address space for use with dma_buf itself.
  */
 struct sg_table *drm_prime_pages_to_sg(struct page **pages, int nr_pages)
 {
@@ -628,9 +683,16 @@ out:
 }
 EXPORT_SYMBOL(drm_prime_pages_to_sg);
 
-/* export an sg table into an array of pages and addresses
-   this is currently required by the TTM driver in order to do correct fault
-   handling */
+/**
+ * drm_prime_sg_to_page_addr_arrays - convert an sg table into a page array
+ * @sgt: scatter-gather table to convert
+ * @pages: array of page pointers to store the page array in
+ * @addrs: optional array to store the dma bus address of each page
+ * @max_pages: size of both the passed-in arrays
+ *
+ * Exports an sg table into an array of pages and addresses. This is currently
+ * required by the TTM driver in order to do correct fault handling.
+ */
 int drm_prime_sg_to_page_addr_arrays(struct sg_table *sgt, struct page **pages,
 				     dma_addr_t *addrs, int max_pages)
 {
@@ -663,7 +725,15 @@ int drm_prime_sg_to_page_addr_arrays(struct sg_table *sgt, struct page **pages,
 	return 0;
 }
 EXPORT_SYMBOL(drm_prime_sg_to_page_addr_arrays);
-/* helper function to cleanup a GEM/prime object */
+
+/**
+ * drm_prime_gem_destroy - helper to clean up a PRIME-imported GEM object
+ * @obj: GEM object which was created from a dma-buf
+ * @sg: the sg-table which was pinned at import time
+ *
+ * This is the cleanup functions which GEM drivers need to call when they use
+ * @drm_gem_prime_import to import dma-bufs.
+ */
 void drm_prime_gem_destroy(struct drm_gem_object *obj, struct sg_table *sg)
 {
 	struct dma_buf_attachment *attach;
@@ -683,11 +753,9 @@ void drm_prime_init_file_private(struct drm_prime_file_private *prime_fpriv)
 	INIT_LIST_HEAD(&prime_fpriv->head);
 	mutex_init(&prime_fpriv->lock);
 }
-EXPORT_SYMBOL(drm_prime_init_file_private);
 
 void drm_prime_destroy_file_private(struct drm_prime_file_private *prime_fpriv)
 {
 	/* by now drm_gem_release should've made sure the list is empty */
 	WARN_ON(!list_empty(&prime_fpriv->head));
 }
-EXPORT_SYMBOL(drm_prime_destroy_file_private);
diff --git a/drivers/gpu/drm/drm_probe_helper.c b/drivers/gpu/drm/drm_probe_helper.c
new file mode 100644
index 00000000000..d22676b89cb
--- /dev/null
+++ b/drivers/gpu/drm/drm_probe_helper.c
@@ -0,0 +1,448 @@
+/*
+ * Copyright (c) 2006-2008 Intel Corporation
+ * Copyright (c) 2007 Dave Airlie <airlied@linux.ie>
+ *
+ * DRM core CRTC related functions
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission.  The copyright holders make no representations
+ * about the suitability of this software for any purpose.  It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ *
+ * Authors:
+ *      Keith Packard
+ *	Eric Anholt <eric@anholt.net>
+ *      Dave Airlie <airlied@linux.ie>
+ *      Jesse Barnes <jesse.barnes@intel.com>
+ */
+
+#include <linux/export.h>
+#include <linux/moduleparam.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_fourcc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_edid.h>
+
+/**
+ * DOC: output probing helper overview
+ *
+ * This library provides some helper code for output probing. It provides an
+ * implementation of the core connector->fill_modes interface with
+ * drm_helper_probe_single_connector_modes.
+ *
+ * It also provides support for polling connectors with a work item and for
+ * generic hotplug interrupt handling where the driver doesn't or cannot keep
+ * track of a per-connector hpd interrupt.
+ *
+ * This helper library can be used independently of the modeset helper library.
+ * Drivers can also overwrite different parts e.g. use their own hotplug
+ * handling code to avoid probing unrelated outputs.
+ */
+
+static bool drm_kms_helper_poll = true;
+module_param_named(poll, drm_kms_helper_poll, bool, 0600);
+
+static void drm_mode_validate_flag(struct drm_connector *connector,
+				   int flags)
+{
+	struct drm_display_mode *mode;
+
+	if (flags == (DRM_MODE_FLAG_DBLSCAN | DRM_MODE_FLAG_INTERLACE |
+		      DRM_MODE_FLAG_3D_MASK))
+		return;
+
+	list_for_each_entry(mode, &connector->modes, head) {
+		if ((mode->flags & DRM_MODE_FLAG_INTERLACE) &&
+				!(flags & DRM_MODE_FLAG_INTERLACE))
+			mode->status = MODE_NO_INTERLACE;
+		if ((mode->flags & DRM_MODE_FLAG_DBLSCAN) &&
+				!(flags & DRM_MODE_FLAG_DBLSCAN))
+			mode->status = MODE_NO_DBLESCAN;
+		if ((mode->flags & DRM_MODE_FLAG_3D_MASK) &&
+				!(flags & DRM_MODE_FLAG_3D_MASK))
+			mode->status = MODE_NO_STEREO;
+	}
+
+	return;
+}
+
+static int drm_helper_probe_single_connector_modes_merge_bits(struct drm_connector *connector,
+							      uint32_t maxX, uint32_t maxY, bool merge_type_bits)
+{
+	struct drm_device *dev = connector->dev;
+	struct drm_display_mode *mode;
+	struct drm_connector_helper_funcs *connector_funcs =
+		connector->helper_private;
+	int count = 0;
+	int mode_flags = 0;
+	bool verbose_prune = true;
+
+	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n", connector->base.id,
+			connector->name);
+	/* set all modes to the unverified state */
+	list_for_each_entry(mode, &connector->modes, head)
+		mode->status = MODE_UNVERIFIED;
+
+	if (connector->force) {
+		if (connector->force == DRM_FORCE_ON)
+			connector->status = connector_status_connected;
+		else
+			connector->status = connector_status_disconnected;
+		if (connector->funcs->force)
+			connector->funcs->force(connector);
+	} else {
+		connector->status = connector->funcs->detect(connector, true);
+	}
+
+	/* Re-enable polling in case the global poll config changed. */
+	if (drm_kms_helper_poll != dev->mode_config.poll_running)
+		drm_kms_helper_poll_enable(dev);
+
+	dev->mode_config.poll_running = drm_kms_helper_poll;
+
+	if (connector->status == connector_status_disconnected) {
+		DRM_DEBUG_KMS("[CONNECTOR:%d:%s] disconnected\n",
+			connector->base.id, connector->name);
+		drm_mode_connector_update_edid_property(connector, NULL);
+		verbose_prune = false;
+		goto prune;
+	}
+
+#ifdef CONFIG_DRM_LOAD_EDID_FIRMWARE
+	count = drm_load_edid_firmware(connector);
+	if (count == 0)
+#endif
+		count = (*connector_funcs->get_modes)(connector);
+
+	if (count == 0 && connector->status == connector_status_connected)
+		count = drm_add_modes_noedid(connector, 1024, 768);
+	if (count == 0)
+		goto prune;
+
+	drm_mode_connector_list_update(connector, merge_type_bits);
+
+	if (maxX && maxY)
+		drm_mode_validate_size(dev, &connector->modes, maxX, maxY);
+
+	if (connector->interlace_allowed)
+		mode_flags |= DRM_MODE_FLAG_INTERLACE;
+	if (connector->doublescan_allowed)
+		mode_flags |= DRM_MODE_FLAG_DBLSCAN;
+	if (connector->stereo_allowed)
+		mode_flags |= DRM_MODE_FLAG_3D_MASK;
+	drm_mode_validate_flag(connector, mode_flags);
+
+	list_for_each_entry(mode, &connector->modes, head) {
+		if (mode->status == MODE_OK && connector_funcs->mode_valid)
+			mode->status = connector_funcs->mode_valid(connector,
+								   mode);
+	}
+
+prune:
+	drm_mode_prune_invalid(dev, &connector->modes, verbose_prune);
+
+	if (list_empty(&connector->modes))
+		return 0;
+
+	list_for_each_entry(mode, &connector->modes, head)
+		mode->vrefresh = drm_mode_vrefresh(mode);
+
+	drm_mode_sort(&connector->modes);
+
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s] probed modes :\n", connector->base.id,
+			connector->name);
+	list_for_each_entry(mode, &connector->modes, head) {
+		drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V);
+		drm_mode_debug_printmodeline(mode);
+	}
+
+	return count;
+}
+
+/**
+ * drm_helper_probe_single_connector_modes - get complete set of display modes
+ * @connector: connector to probe
+ * @maxX: max width for modes
+ * @maxY: max height for modes
+ *
+ * Based on the helper callbacks implemented by @connector try to detect all
+ * valid modes.  Modes will first be added to the connector's probed_modes list,
+ * then culled (based on validity and the @maxX, @maxY parameters) and put into
+ * the normal modes list.
+ *
+ * Intended to be use as a generic implementation of the ->fill_modes()
+ * @connector vfunc for drivers that use the crtc helpers for output mode
+ * filtering and detection.
+ *
+ * Returns:
+ * The number of modes found on @connector.
+ */
+int drm_helper_probe_single_connector_modes(struct drm_connector *connector,
+					    uint32_t maxX, uint32_t maxY)
+{
+	return drm_helper_probe_single_connector_modes_merge_bits(connector, maxX, maxY, true);
+}
+EXPORT_SYMBOL(drm_helper_probe_single_connector_modes);
+
+/**
+ * drm_helper_probe_single_connector_modes_nomerge - get complete set of display modes
+ * @connector: connector to probe
+ * @maxX: max width for modes
+ * @maxY: max height for modes
+ *
+ * This operates like drm_hehlper_probe_single_connector_modes except it
+ * replaces the mode bits instead of merging them for preferred modes.
+ */
+int drm_helper_probe_single_connector_modes_nomerge(struct drm_connector *connector,
+					    uint32_t maxX, uint32_t maxY)
+{
+	return drm_helper_probe_single_connector_modes_merge_bits(connector, maxX, maxY, false);
+}
+EXPORT_SYMBOL(drm_helper_probe_single_connector_modes_nomerge);
+
+/**
+ * drm_kms_helper_hotplug_event - fire off KMS hotplug events
+ * @dev: drm_device whose connector state changed
+ *
+ * This function fires off the uevent for userspace and also calls the
+ * output_poll_changed function, which is most commonly used to inform the fbdev
+ * emulation code and allow it to update the fbcon output configuration.
+ *
+ * Drivers should call this from their hotplug handling code when a change is
+ * detected. Note that this function does not do any output detection of its
+ * own, like drm_helper_hpd_irq_event() does - this is assumed to be done by the
+ * driver already.
+ *
+ * This function must be called from process context with no mode
+ * setting locks held.
+ */
+void drm_kms_helper_hotplug_event(struct drm_device *dev)
+{
+	/* send a uevent + call fbdev */
+	drm_sysfs_hotplug_event(dev);
+	if (dev->mode_config.funcs->output_poll_changed)
+		dev->mode_config.funcs->output_poll_changed(dev);
+}
+EXPORT_SYMBOL(drm_kms_helper_hotplug_event);
+
+#define DRM_OUTPUT_POLL_PERIOD (10*HZ)
+static void output_poll_execute(struct work_struct *work)
+{
+	struct delayed_work *delayed_work = to_delayed_work(work);
+	struct drm_device *dev = container_of(delayed_work, struct drm_device, mode_config.output_poll_work);
+	struct drm_connector *connector;
+	enum drm_connector_status old_status;
+	bool repoll = false, changed = false;
+
+	if (!drm_kms_helper_poll)
+		return;
+
+	mutex_lock(&dev->mode_config.mutex);
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+
+		/* Ignore forced connectors. */
+		if (connector->force)
+			continue;
+
+		/* Ignore HPD capable connectors and connectors where we don't
+		 * want any hotplug detection at all for polling. */
+		if (!connector->polled || connector->polled == DRM_CONNECTOR_POLL_HPD)
+			continue;
+
+		repoll = true;
+
+		old_status = connector->status;
+		/* if we are connected and don't want to poll for disconnect
+		   skip it */
+		if (old_status == connector_status_connected &&
+		    !(connector->polled & DRM_CONNECTOR_POLL_DISCONNECT))
+			continue;
+
+		connector->status = connector->funcs->detect(connector, false);
+		if (old_status != connector->status) {
+			const char *old, *new;
+
+			old = drm_get_connector_status_name(old_status);
+			new = drm_get_connector_status_name(connector->status);
+
+			DRM_DEBUG_KMS("[CONNECTOR:%d:%s] "
+				      "status updated from %s to %s\n",
+				      connector->base.id,
+				      connector->name,
+				      old, new);
+
+			changed = true;
+		}
+	}
+
+	mutex_unlock(&dev->mode_config.mutex);
+
+	if (changed)
+		drm_kms_helper_hotplug_event(dev);
+
+	if (repoll)
+		schedule_delayed_work(delayed_work, DRM_OUTPUT_POLL_PERIOD);
+}
+
+/**
+ * drm_kms_helper_poll_disable - disable output polling
+ * @dev: drm_device
+ *
+ * This function disables the output polling work.
+ *
+ * Drivers can call this helper from their device suspend implementation. It is
+ * not an error to call this even when output polling isn't enabled or arlready
+ * disabled.
+ */
+void drm_kms_helper_poll_disable(struct drm_device *dev)
+{
+	if (!dev->mode_config.poll_enabled)
+		return;
+	cancel_delayed_work_sync(&dev->mode_config.output_poll_work);
+}
+EXPORT_SYMBOL(drm_kms_helper_poll_disable);
+
+/**
+ * drm_kms_helper_poll_enable - re-enable output polling.
+ * @dev: drm_device
+ *
+ * This function re-enables the output polling work.
+ *
+ * Drivers can call this helper from their device resume implementation. It is
+ * an error to call this when the output polling support has not yet been set
+ * up.
+ */
+void drm_kms_helper_poll_enable(struct drm_device *dev)
+{
+	bool poll = false;
+	struct drm_connector *connector;
+
+	if (!dev->mode_config.poll_enabled || !drm_kms_helper_poll)
+		return;
+
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		if (connector->polled & (DRM_CONNECTOR_POLL_CONNECT |
+					 DRM_CONNECTOR_POLL_DISCONNECT))
+			poll = true;
+	}
+
+	if (poll)
+		schedule_delayed_work(&dev->mode_config.output_poll_work, DRM_OUTPUT_POLL_PERIOD);
+}
+EXPORT_SYMBOL(drm_kms_helper_poll_enable);
+
+/**
+ * drm_kms_helper_poll_init - initialize and enable output polling
+ * @dev: drm_device
+ *
+ * This function intializes and then also enables output polling support for
+ * @dev. Drivers which do not have reliable hotplug support in hardware can use
+ * this helper infrastructure to regularly poll such connectors for changes in
+ * their connection state.
+ *
+ * Drivers can control which connectors are polled by setting the
+ * DRM_CONNECTOR_POLL_CONNECT and DRM_CONNECTOR_POLL_DISCONNECT flags. On
+ * connectors where probing live outputs can result in visual distortion drivers
+ * should not set the DRM_CONNECTOR_POLL_DISCONNECT flag to avoid this.
+ * Connectors which have no flag or only DRM_CONNECTOR_POLL_HPD set are
+ * completely ignored by the polling logic.
+ *
+ * Note that a connector can be both polled and probed from the hotplug handler,
+ * in case the hotplug interrupt is known to be unreliable.
+ */
+void drm_kms_helper_poll_init(struct drm_device *dev)
+{
+	INIT_DELAYED_WORK(&dev->mode_config.output_poll_work, output_poll_execute);
+	dev->mode_config.poll_enabled = true;
+
+	drm_kms_helper_poll_enable(dev);
+}
+EXPORT_SYMBOL(drm_kms_helper_poll_init);
+
+/**
+ * drm_kms_helper_poll_fini - disable output polling and clean it up
+ * @dev: drm_device
+ */
+void drm_kms_helper_poll_fini(struct drm_device *dev)
+{
+	drm_kms_helper_poll_disable(dev);
+}
+EXPORT_SYMBOL(drm_kms_helper_poll_fini);
+
+/**
+ * drm_helper_hpd_irq_event - hotplug processing
+ * @dev: drm_device
+ *
+ * Drivers can use this helper function to run a detect cycle on all connectors
+ * which have the DRM_CONNECTOR_POLL_HPD flag set in their &polled member. All
+ * other connectors are ignored, which is useful to avoid reprobing fixed
+ * panels.
+ *
+ * This helper function is useful for drivers which can't or don't track hotplug
+ * interrupts for each connector.
+ *
+ * Drivers which support hotplug interrupts for each connector individually and
+ * which have a more fine-grained detect logic should bypass this code and
+ * directly call drm_kms_helper_hotplug_event() in case the connector state
+ * changed.
+ *
+ * This function must be called from process context with no mode
+ * setting locks held.
+ *
+ * Note that a connector can be both polled and probed from the hotplug handler,
+ * in case the hotplug interrupt is known to be unreliable.
+ */
+bool drm_helper_hpd_irq_event(struct drm_device *dev)
+{
+	struct drm_connector *connector;
+	enum drm_connector_status old_status;
+	bool changed = false;
+
+	if (!dev->mode_config.poll_enabled)
+		return false;
+
+	mutex_lock(&dev->mode_config.mutex);
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+
+		/* Only handle HPD capable connectors. */
+		if (!(connector->polled & DRM_CONNECTOR_POLL_HPD))
+			continue;
+
+		old_status = connector->status;
+
+		connector->status = connector->funcs->detect(connector, false);
+		DRM_DEBUG_KMS("[CONNECTOR:%d:%s] status updated from %s to %s\n",
+			      connector->base.id,
+			      connector->name,
+			      drm_get_connector_status_name(old_status),
+			      drm_get_connector_status_name(connector->status));
+		if (old_status != connector->status)
+			changed = true;
+	}
+
+	mutex_unlock(&dev->mode_config.mutex);
+
+	if (changed)
+		drm_kms_helper_hotplug_event(dev);
+
+	return changed;
+}
+EXPORT_SYMBOL(drm_helper_hpd_irq_event);
diff --git a/drivers/gpu/drm/drm_stub.c b/drivers/gpu/drm/drm_stub.c
index 66dd3a001cf..14d16464000 100644
--- a/drivers/gpu/drm/drm_stub.c
+++ b/drivers/gpu/drm/drm_stub.c
@@ -1,16 +1,11 @@
-/**
- * \file drm_stub.h
- * Stub support
- *
- * \author Rickard E. (Rik) Faith <faith@valinux.com>
- */
-
 /*
  * Created: Fri Jan 19 10:48:35 2001 by faith@acm.org
  *
  * Copyright 2001 VA Linux Systems, Inc., Sunnyvale, California.
  * All Rights Reserved.
  *
+ * Author Rickard E. (Rik) Faith <faith@valinux.com>
+ *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
  * to deal in the Software without restriction, including without limitation
@@ -31,8 +26,10 @@
  * DEALINGS IN THE SOFTWARE.
  */
 
+#include <linux/fs.h>
 #include <linux/module.h>
 #include <linux/moduleparam.h>
+#include <linux/mount.h>
 #include <linux/slab.h>
 #include <drm/drmP.h>
 #include <drm/drm_core.h>
@@ -43,6 +40,10 @@ EXPORT_SYMBOL(drm_debug);
 unsigned int drm_rnodes = 0;	/* 1 to enable experimental render nodes API */
 EXPORT_SYMBOL(drm_rnodes);
 
+/* 1 to allow user space to request universal planes (experimental) */
+unsigned int drm_universal_planes = 0;
+EXPORT_SYMBOL(drm_universal_planes);
+
 unsigned int drm_vblank_offdelay = 5000;    /* Default to 5000 msecs. */
 EXPORT_SYMBOL(drm_vblank_offdelay);
 
@@ -66,10 +67,12 @@ MODULE_PARM_DESC(timestamp_monotonic, "Use monotonic timestamps");
 
 module_param_named(debug, drm_debug, int, 0600);
 module_param_named(rnodes, drm_rnodes, int, 0600);
+module_param_named(universal_planes, drm_universal_planes, int, 0600);
 module_param_named(vblankoffdelay, drm_vblank_offdelay, int, 0600);
 module_param_named(timestamp_precision_usec, drm_timestamp_precision, int, 0600);
 module_param_named(timestamp_monotonic, drm_timestamp_monotonic, int, 0600);
 
+static DEFINE_SPINLOCK(drm_minor_lock);
 struct idr drm_minors_idr;
 
 struct class *drm_class;
@@ -94,42 +97,20 @@ int drm_err(const char *func, const char *format, ...)
 }
 EXPORT_SYMBOL(drm_err);
 
-void drm_ut_debug_printk(unsigned int request_level,
-			 const char *prefix,
-			 const char *function_name,
-			 const char *format, ...)
+void drm_ut_debug_printk(const char *function_name, const char *format, ...)
 {
+	struct va_format vaf;
 	va_list args;
 
-	if (drm_debug & request_level) {
-		if (function_name)
-			printk(KERN_DEBUG "[%s:%s], ", prefix, function_name);
-		va_start(args, format);
-		vprintk(format, args);
-		va_end(args);
-	}
-}
-EXPORT_SYMBOL(drm_ut_debug_printk);
-
-static int drm_minor_get_id(struct drm_device *dev, int type)
-{
-	int ret;
-	int base = 0, limit = 63;
-
-	if (type == DRM_MINOR_CONTROL) {
-		base += 64;
-		limit = base + 63;
-	} else if (type == DRM_MINOR_RENDER) {
-		base += 128;
-		limit = base + 63;
-	}
+	va_start(args, format);
+	vaf.fmt = format;
+	vaf.va = &args;
 
-	mutex_lock(&dev->struct_mutex);
-	ret = idr_alloc(&drm_minors_idr, NULL, base, limit, GFP_KERNEL);
-	mutex_unlock(&dev->struct_mutex);
+	printk(KERN_DEBUG "[" DRM_NAME ":%s] %pV", function_name, &vaf);
 
-	return ret == -ENOSPC ? -EINVAL : ret;
+	va_end(args);
 }
+EXPORT_SYMBOL(drm_ut_debug_printk);
 
 struct drm_master *drm_master_create(struct drm_minor *minor)
 {
@@ -142,12 +123,13 @@ struct drm_master *drm_master_create(struct drm_minor *minor)
 	kref_init(&master->refcount);
 	spin_lock_init(&master->lock.spinlock);
 	init_waitqueue_head(&master->lock.lock_queue);
-	drm_ht_create(&master->magiclist, DRM_MAGIC_HASH_ORDER);
+	if (drm_ht_create(&master->magiclist, DRM_MAGIC_HASH_ORDER)) {
+		kfree(master);
+		return NULL;
+	}
 	INIT_LIST_HEAD(&master->magicfree);
 	master->minor = minor;
 
-	list_add_tail(&master->head, &minor->master_list);
-
 	return master;
 }
 
@@ -165,8 +147,7 @@ static void drm_master_destroy(struct kref *kref)
 	struct drm_device *dev = master->minor->dev;
 	struct drm_map_list *r_list, *list_temp;
 
-	list_del(&master->head);
-
+	mutex_lock(&dev->struct_mutex);
 	if (dev->driver->master_destroy)
 		dev->driver->master_destroy(dev, master);
 
@@ -183,9 +164,6 @@ static void drm_master_destroy(struct kref *kref)
 		master->unique_len = 0;
 	}
 
-	kfree(dev->devname);
-	dev->devname = NULL;
-
 	list_for_each_entry_safe(pt, next, &master->magicfree, head) {
 		list_del(&pt->head);
 		drm_ht_remove_item(&master->magiclist, &pt->hash_item);
@@ -194,6 +172,7 @@ static void drm_master_destroy(struct kref *kref)
 
 	drm_ht_remove(&master->magiclist);
 
+	mutex_unlock(&dev->struct_mutex);
 	kfree(master);
 }
 
@@ -209,19 +188,20 @@ int drm_setmaster_ioctl(struct drm_device *dev, void *data,
 {
 	int ret = 0;
 
+	mutex_lock(&dev->master_mutex);
 	if (file_priv->is_master)
-		return 0;
+		goto out_unlock;
 
-	if (file_priv->minor->master && file_priv->minor->master != file_priv->master)
-		return -EINVAL;
-
-	if (!file_priv->master)
-		return -EINVAL;
+	if (file_priv->minor->master) {
+		ret = -EINVAL;
+		goto out_unlock;
+	}
 
-	if (file_priv->minor->master)
-		return -EINVAL;
+	if (!file_priv->master) {
+		ret = -EINVAL;
+		goto out_unlock;
+	}
 
-	mutex_lock(&dev->struct_mutex);
 	file_priv->minor->master = drm_master_get(file_priv->master);
 	file_priv->is_master = 1;
 	if (dev->driver->master_set) {
@@ -231,150 +211,224 @@ int drm_setmaster_ioctl(struct drm_device *dev, void *data,
 			drm_master_put(&file_priv->minor->master);
 		}
 	}
-	mutex_unlock(&dev->struct_mutex);
 
+out_unlock:
+	mutex_unlock(&dev->master_mutex);
 	return ret;
 }
 
 int drm_dropmaster_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv)
 {
+	int ret = -EINVAL;
+
+	mutex_lock(&dev->master_mutex);
 	if (!file_priv->is_master)
-		return -EINVAL;
+		goto out_unlock;
 
 	if (!file_priv->minor->master)
-		return -EINVAL;
+		goto out_unlock;
 
-	mutex_lock(&dev->struct_mutex);
+	ret = 0;
 	if (dev->driver->master_drop)
 		dev->driver->master_drop(dev, file_priv, false);
 	drm_master_put(&file_priv->minor->master);
 	file_priv->is_master = 0;
-	mutex_unlock(&dev->struct_mutex);
-	return 0;
+
+out_unlock:
+	mutex_unlock(&dev->master_mutex);
+	return ret;
 }
 
-/**
- * drm_get_minor - Allocate and register new DRM minor
- * @dev: DRM device
- * @minor: Pointer to where new minor is stored
- * @type: Type of minor
- *
- * Allocate a new minor of the given type and register it. A pointer to the new
- * minor is returned in @minor.
- * Caller must hold the global DRM mutex.
+/*
+ * DRM Minors
+ * A DRM device can provide several char-dev interfaces on the DRM-Major. Each
+ * of them is represented by a drm_minor object. Depending on the capabilities
+ * of the device-driver, different interfaces are registered.
  *
- * RETURNS:
- * 0 on success, negative error code on failure.
+ * Minors can be accessed via dev->$minor_name. This pointer is either
+ * NULL or a valid drm_minor pointer and stays valid as long as the device is
+ * valid. This means, DRM minors have the same life-time as the underlying
+ * device. However, this doesn't mean that the minor is active. Minors are
+ * registered and unregistered dynamically according to device-state.
  */
-static int drm_get_minor(struct drm_device *dev, struct drm_minor **minor,
-			 int type)
+
+static struct drm_minor **drm_minor_get_slot(struct drm_device *dev,
+					     unsigned int type)
+{
+	switch (type) {
+	case DRM_MINOR_LEGACY:
+		return &dev->primary;
+	case DRM_MINOR_RENDER:
+		return &dev->render;
+	case DRM_MINOR_CONTROL:
+		return &dev->control;
+	default:
+		return NULL;
+	}
+}
+
+static int drm_minor_alloc(struct drm_device *dev, unsigned int type)
+{
+	struct drm_minor *minor;
+
+	minor = kzalloc(sizeof(*minor), GFP_KERNEL);
+	if (!minor)
+		return -ENOMEM;
+
+	minor->type = type;
+	minor->dev = dev;
+
+	*drm_minor_get_slot(dev, type) = minor;
+	return 0;
+}
+
+static void drm_minor_free(struct drm_device *dev, unsigned int type)
+{
+	struct drm_minor **slot;
+
+	slot = drm_minor_get_slot(dev, type);
+	if (*slot) {
+		drm_mode_group_destroy(&(*slot)->mode_group);
+		kfree(*slot);
+		*slot = NULL;
+	}
+}
+
+static int drm_minor_register(struct drm_device *dev, unsigned int type)
 {
 	struct drm_minor *new_minor;
+	unsigned long flags;
 	int ret;
 	int minor_id;
 
 	DRM_DEBUG("\n");
 
-	minor_id = drm_minor_get_id(dev, type);
+	new_minor = *drm_minor_get_slot(dev, type);
+	if (!new_minor)
+		return 0;
+
+	idr_preload(GFP_KERNEL);
+	spin_lock_irqsave(&drm_minor_lock, flags);
+	minor_id = idr_alloc(&drm_minors_idr,
+			     NULL,
+			     64 * type,
+			     64 * (type + 1),
+			     GFP_NOWAIT);
+	spin_unlock_irqrestore(&drm_minor_lock, flags);
+	idr_preload_end();
+
 	if (minor_id < 0)
 		return minor_id;
 
-	new_minor = kzalloc(sizeof(struct drm_minor), GFP_KERNEL);
-	if (!new_minor) {
-		ret = -ENOMEM;
-		goto err_idr;
-	}
-
-	new_minor->type = type;
-	new_minor->device = MKDEV(DRM_MAJOR, minor_id);
-	new_minor->dev = dev;
 	new_minor->index = minor_id;
-	INIT_LIST_HEAD(&new_minor->master_list);
 
-	idr_replace(&drm_minors_idr, new_minor, minor_id);
-
-#if defined(CONFIG_DEBUG_FS)
 	ret = drm_debugfs_init(new_minor, minor_id, drm_debugfs_root);
 	if (ret) {
 		DRM_ERROR("DRM: Failed to initialize /sys/kernel/debug/dri.\n");
-		goto err_mem;
+		goto err_id;
 	}
-#endif
 
 	ret = drm_sysfs_device_add(new_minor);
 	if (ret) {
-		printk(KERN_ERR
-		       "DRM: Error sysfs_device_add.\n");
+		DRM_ERROR("DRM: Error sysfs_device_add.\n");
 		goto err_debugfs;
 	}
-	*minor = new_minor;
+
+	/* replace NULL with @minor so lookups will succeed from now on */
+	spin_lock_irqsave(&drm_minor_lock, flags);
+	idr_replace(&drm_minors_idr, new_minor, new_minor->index);
+	spin_unlock_irqrestore(&drm_minor_lock, flags);
 
 	DRM_DEBUG("new minor assigned %d\n", minor_id);
 	return 0;
 
-
 err_debugfs:
-#if defined(CONFIG_DEBUG_FS)
 	drm_debugfs_cleanup(new_minor);
-err_mem:
-#endif
-	kfree(new_minor);
-err_idr:
+err_id:
+	spin_lock_irqsave(&drm_minor_lock, flags);
 	idr_remove(&drm_minors_idr, minor_id);
-	*minor = NULL;
+	spin_unlock_irqrestore(&drm_minor_lock, flags);
+	new_minor->index = 0;
 	return ret;
 }
 
-/**
- * drm_unplug_minor - Unplug DRM minor
- * @minor: Minor to unplug
- *
- * Unplugs the given DRM minor but keeps the object. So after this returns,
- * minor->dev is still valid so existing open-files can still access it to get
- * device information from their drm_file ojects.
- * If the minor is already unplugged or if @minor is NULL, nothing is done.
- * The global DRM mutex must be held by the caller.
- */
-static void drm_unplug_minor(struct drm_minor *minor)
+static void drm_minor_unregister(struct drm_device *dev, unsigned int type)
 {
+	struct drm_minor *minor;
+	unsigned long flags;
+
+	minor = *drm_minor_get_slot(dev, type);
 	if (!minor || !minor->kdev)
 		return;
 
-#if defined(CONFIG_DEBUG_FS)
-	drm_debugfs_cleanup(minor);
-#endif
+	spin_lock_irqsave(&drm_minor_lock, flags);
+	idr_remove(&drm_minors_idr, minor->index);
+	spin_unlock_irqrestore(&drm_minor_lock, flags);
+	minor->index = 0;
 
+	drm_debugfs_cleanup(minor);
 	drm_sysfs_device_remove(minor);
-	idr_remove(&drm_minors_idr, minor->index);
 }
 
 /**
- * drm_put_minor - Destroy DRM minor
- * @minor: Minor to destroy
+ * drm_minor_acquire - Acquire a DRM minor
+ * @minor_id: Minor ID of the DRM-minor
  *
- * This calls drm_unplug_minor() on the given minor and then frees it. Nothing
- * is done if @minor is NULL. It is fine to call this on already unplugged
- * minors.
- * The global DRM mutex must be held by the caller.
+ * Looks up the given minor-ID and returns the respective DRM-minor object. The
+ * refence-count of the underlying device is increased so you must release this
+ * object with drm_minor_release().
+ *
+ * As long as you hold this minor, it is guaranteed that the object and the
+ * minor->dev pointer will stay valid! However, the device may get unplugged and
+ * unregistered while you hold the minor.
+ *
+ * Returns:
+ * Pointer to minor-object with increased device-refcount, or PTR_ERR on
+ * failure.
  */
-static void drm_put_minor(struct drm_minor *minor)
+struct drm_minor *drm_minor_acquire(unsigned int minor_id)
 {
-	if (!minor)
-		return;
+	struct drm_minor *minor;
+	unsigned long flags;
+
+	spin_lock_irqsave(&drm_minor_lock, flags);
+	minor = idr_find(&drm_minors_idr, minor_id);
+	if (minor)
+		drm_dev_ref(minor->dev);
+	spin_unlock_irqrestore(&drm_minor_lock, flags);
+
+	if (!minor) {
+		return ERR_PTR(-ENODEV);
+	} else if (drm_device_is_unplugged(minor->dev)) {
+		drm_dev_unref(minor->dev);
+		return ERR_PTR(-ENODEV);
+	}
 
-	DRM_DEBUG("release secondary minor %d\n", minor->index);
+	return minor;
+}
 
-	drm_unplug_minor(minor);
-	kfree(minor);
+/**
+ * drm_minor_release - Release DRM minor
+ * @minor: Pointer to DRM minor object
+ *
+ * Release a minor that was previously acquired via drm_minor_acquire().
+ */
+void drm_minor_release(struct drm_minor *minor)
+{
+	drm_dev_unref(minor->dev);
 }
 
 /**
- * Called via drm_exit() at module unload time or when pci device is
- * unplugged.
+ * drm_put_dev - Unregister and release a DRM device
+ * @dev: DRM device
  *
- * Cleans up all DRM device, calling drm_lastclose().
+ * Called at module unload time or when a PCI device is unplugged.
+ *
+ * Use of this function is discouraged. It will eventually go away completely.
+ * Please use drm_dev_unregister() and drm_dev_unref() explicitly instead.
  *
+ * Cleans up all DRM device, calling drm_lastclose().
  */
 void drm_put_dev(struct drm_device *dev)
 {
@@ -386,18 +440,16 @@ void drm_put_dev(struct drm_device *dev)
 	}
 
 	drm_dev_unregister(dev);
-	drm_dev_free(dev);
+	drm_dev_unref(dev);
 }
 EXPORT_SYMBOL(drm_put_dev);
 
 void drm_unplug_dev(struct drm_device *dev)
 {
 	/* for a USB device */
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
-		drm_unplug_minor(dev->control);
-	if (dev->render)
-		drm_unplug_minor(dev->render);
-	drm_unplug_minor(dev->primary);
+	drm_minor_unregister(dev, DRM_MINOR_LEGACY);
+	drm_minor_unregister(dev, DRM_MINOR_RENDER);
+	drm_minor_unregister(dev, DRM_MINOR_CONTROL);
 
 	mutex_lock(&drm_global_mutex);
 
@@ -410,8 +462,80 @@ void drm_unplug_dev(struct drm_device *dev)
 }
 EXPORT_SYMBOL(drm_unplug_dev);
 
+/*
+ * DRM internal mount
+ * We want to be able to allocate our own "struct address_space" to control
+ * memory-mappings in VRAM (or stolen RAM, ...). However, core MM does not allow
+ * stand-alone address_space objects, so we need an underlying inode. As there
+ * is no way to allocate an independent inode easily, we need a fake internal
+ * VFS mount-point.
+ *
+ * The drm_fs_inode_new() function allocates a new inode, drm_fs_inode_free()
+ * frees it again. You are allowed to use iget() and iput() to get references to
+ * the inode. But each drm_fs_inode_new() call must be paired with exactly one
+ * drm_fs_inode_free() call (which does not have to be the last iput()).
+ * We use drm_fs_inode_*() to manage our internal VFS mount-point and share it
+ * between multiple inode-users. You could, technically, call
+ * iget() + drm_fs_inode_free() directly after alloc and sometime later do an
+ * iput(), but this way you'd end up with a new vfsmount for each inode.
+ */
+
+static int drm_fs_cnt;
+static struct vfsmount *drm_fs_mnt;
+
+static const struct dentry_operations drm_fs_dops = {
+	.d_dname	= simple_dname,
+};
+
+static const struct super_operations drm_fs_sops = {
+	.statfs		= simple_statfs,
+};
+
+static struct dentry *drm_fs_mount(struct file_system_type *fs_type, int flags,
+				   const char *dev_name, void *data)
+{
+	return mount_pseudo(fs_type,
+			    "drm:",
+			    &drm_fs_sops,
+			    &drm_fs_dops,
+			    0x010203ff);
+}
+
+static struct file_system_type drm_fs_type = {
+	.name		= "drm",
+	.owner		= THIS_MODULE,
+	.mount		= drm_fs_mount,
+	.kill_sb	= kill_anon_super,
+};
+
+static struct inode *drm_fs_inode_new(void)
+{
+	struct inode *inode;
+	int r;
+
+	r = simple_pin_fs(&drm_fs_type, &drm_fs_mnt, &drm_fs_cnt);
+	if (r < 0) {
+		DRM_ERROR("Cannot mount pseudo fs: %d\n", r);
+		return ERR_PTR(r);
+	}
+
+	inode = alloc_anon_inode(drm_fs_mnt->mnt_sb);
+	if (IS_ERR(inode))
+		simple_release_fs(&drm_fs_mnt, &drm_fs_cnt);
+
+	return inode;
+}
+
+static void drm_fs_inode_free(struct inode *inode)
+{
+	if (inode) {
+		iput(inode);
+		simple_release_fs(&drm_fs_mnt, &drm_fs_cnt);
+	}
+}
+
 /**
- * drm_dev_alloc - Allocate new drm device
+ * drm_dev_alloc - Allocate new DRM device
  * @driver: DRM driver to allocate device for
  * @parent: Parent device object
  *
@@ -419,6 +543,9 @@ EXPORT_SYMBOL(drm_unplug_dev);
  * Call drm_dev_register() to advertice the device to user space and register it
  * with other core subsystems.
  *
+ * The initial ref-count of the object is 1. Use drm_dev_ref() and
+ * drm_dev_unref() to take and drop further ref-counts.
+ *
  * RETURNS:
  * Pointer to new DRM device, or NULL if out of memory.
  */
@@ -432,6 +559,7 @@ struct drm_device *drm_dev_alloc(struct drm_driver *driver,
 	if (!dev)
 		return NULL;
 
+	kref_init(&dev->ref);
 	dev->dev = parent;
 	dev->driver = driver;
 
@@ -441,13 +569,37 @@ struct drm_device *drm_dev_alloc(struct drm_driver *driver,
 	INIT_LIST_HEAD(&dev->maplist);
 	INIT_LIST_HEAD(&dev->vblank_event_list);
 
-	spin_lock_init(&dev->count_lock);
+	spin_lock_init(&dev->buf_lock);
 	spin_lock_init(&dev->event_lock);
 	mutex_init(&dev->struct_mutex);
 	mutex_init(&dev->ctxlist_mutex);
+	mutex_init(&dev->master_mutex);
 
-	if (drm_ht_create(&dev->map_hash, 12))
+	dev->anon_inode = drm_fs_inode_new();
+	if (IS_ERR(dev->anon_inode)) {
+		ret = PTR_ERR(dev->anon_inode);
+		DRM_ERROR("Cannot allocate anonymous inode: %d\n", ret);
 		goto err_free;
+	}
+
+	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
+		ret = drm_minor_alloc(dev, DRM_MINOR_CONTROL);
+		if (ret)
+			goto err_minors;
+	}
+
+	if (drm_core_check_feature(dev, DRIVER_RENDER) && drm_rnodes) {
+		ret = drm_minor_alloc(dev, DRM_MINOR_RENDER);
+		if (ret)
+			goto err_minors;
+	}
+
+	ret = drm_minor_alloc(dev, DRM_MINOR_LEGACY);
+	if (ret)
+		goto err_minors;
+
+	if (drm_ht_create(&dev->map_hash, 12))
+		goto err_minors;
 
 	ret = drm_ctxbitmap_init(dev);
 	if (ret) {
@@ -469,42 +621,75 @@ err_ctxbitmap:
 	drm_ctxbitmap_cleanup(dev);
 err_ht:
 	drm_ht_remove(&dev->map_hash);
+err_minors:
+	drm_minor_free(dev, DRM_MINOR_LEGACY);
+	drm_minor_free(dev, DRM_MINOR_RENDER);
+	drm_minor_free(dev, DRM_MINOR_CONTROL);
+	drm_fs_inode_free(dev->anon_inode);
 err_free:
+	mutex_destroy(&dev->master_mutex);
 	kfree(dev);
 	return NULL;
 }
 EXPORT_SYMBOL(drm_dev_alloc);
 
-/**
- * drm_dev_free - Free DRM device
- * @dev: DRM device to free
- *
- * Free a DRM device that has previously been allocated via drm_dev_alloc().
- * You must not use kfree() instead or you will leak memory.
- *
- * This must not be called once the device got registered. Use drm_put_dev()
- * instead, which then calls drm_dev_free().
- */
-void drm_dev_free(struct drm_device *dev)
+static void drm_dev_release(struct kref *ref)
 {
-	drm_put_minor(dev->control);
-	drm_put_minor(dev->render);
-	drm_put_minor(dev->primary);
+	struct drm_device *dev = container_of(ref, struct drm_device, ref);
 
 	if (dev->driver->driver_features & DRIVER_GEM)
 		drm_gem_destroy(dev);
 
 	drm_ctxbitmap_cleanup(dev);
 	drm_ht_remove(&dev->map_hash);
+	drm_fs_inode_free(dev->anon_inode);
+
+	drm_minor_free(dev, DRM_MINOR_LEGACY);
+	drm_minor_free(dev, DRM_MINOR_RENDER);
+	drm_minor_free(dev, DRM_MINOR_CONTROL);
 
-	kfree(dev->devname);
+	mutex_destroy(&dev->master_mutex);
+	kfree(dev->unique);
 	kfree(dev);
 }
-EXPORT_SYMBOL(drm_dev_free);
+
+/**
+ * drm_dev_ref - Take reference of a DRM device
+ * @dev: device to take reference of or NULL
+ *
+ * This increases the ref-count of @dev by one. You *must* already own a
+ * reference when calling this. Use drm_dev_unref() to drop this reference
+ * again.
+ *
+ * This function never fails. However, this function does not provide *any*
+ * guarantee whether the device is alive or running. It only provides a
+ * reference to the object and the memory associated with it.
+ */
+void drm_dev_ref(struct drm_device *dev)
+{
+	if (dev)
+		kref_get(&dev->ref);
+}
+EXPORT_SYMBOL(drm_dev_ref);
+
+/**
+ * drm_dev_unref - Drop reference of a DRM device
+ * @dev: device to drop reference of or NULL
+ *
+ * This decreases the ref-count of @dev by one. The device is destroyed if the
+ * ref-count drops to zero.
+ */
+void drm_dev_unref(struct drm_device *dev)
+{
+	if (dev)
+		kref_put(&dev->ref, drm_dev_release);
+}
+EXPORT_SYMBOL(drm_dev_unref);
 
 /**
  * drm_dev_register - Register DRM device
  * @dev: Device to register
+ * @flags: Flags passed to the driver's .load() function
  *
  * Register the DRM device @dev with the system, advertise device to user-space
  * and start normal device operation. @dev must be allocated via drm_dev_alloc()
@@ -521,32 +706,22 @@ int drm_dev_register(struct drm_device *dev, unsigned long flags)
 
 	mutex_lock(&drm_global_mutex);
 
-	if (dev->driver->bus->agp_init) {
-		ret = dev->driver->bus->agp_init(dev);
-		if (ret)
-			goto out_unlock;
-	}
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
-		ret = drm_get_minor(dev, &dev->control, DRM_MINOR_CONTROL);
-		if (ret)
-			goto err_agp;
-	}
+	ret = drm_minor_register(dev, DRM_MINOR_CONTROL);
+	if (ret)
+		goto err_minors;
 
-	if (drm_core_check_feature(dev, DRIVER_RENDER) && drm_rnodes) {
-		ret = drm_get_minor(dev, &dev->render, DRM_MINOR_RENDER);
-		if (ret)
-			goto err_control_node;
-	}
+	ret = drm_minor_register(dev, DRM_MINOR_RENDER);
+	if (ret)
+		goto err_minors;
 
-	ret = drm_get_minor(dev, &dev->primary, DRM_MINOR_LEGACY);
+	ret = drm_minor_register(dev, DRM_MINOR_LEGACY);
 	if (ret)
-		goto err_render_node;
+		goto err_minors;
 
 	if (dev->driver->load) {
 		ret = dev->driver->load(dev, flags);
 		if (ret)
-			goto err_primary_node;
+			goto err_minors;
 	}
 
 	/* setup grouping for legacy outputs */
@@ -557,23 +732,16 @@ int drm_dev_register(struct drm_device *dev, unsigned long flags)
 			goto err_unload;
 	}
 
-	list_add_tail(&dev->driver_item, &dev->driver->device_list);
-
 	ret = 0;
 	goto out_unlock;
 
 err_unload:
 	if (dev->driver->unload)
 		dev->driver->unload(dev);
-err_primary_node:
-	drm_unplug_minor(dev->primary);
-err_render_node:
-	drm_unplug_minor(dev->render);
-err_control_node:
-	drm_unplug_minor(dev->control);
-err_agp:
-	if (dev->driver->bus->agp_destroy)
-		dev->driver->bus->agp_destroy(dev);
+err_minors:
+	drm_minor_unregister(dev, DRM_MINOR_LEGACY);
+	drm_minor_unregister(dev, DRM_MINOR_RENDER);
+	drm_minor_unregister(dev, DRM_MINOR_CONTROL);
 out_unlock:
 	mutex_unlock(&drm_global_mutex);
 	return ret;
@@ -586,7 +754,7 @@ EXPORT_SYMBOL(drm_dev_register);
  *
  * Unregister the DRM device from the system. This does the reverse of
  * drm_dev_register() but does not deallocate the device. The caller must call
- * drm_dev_free() to free all resources.
+ * drm_dev_unref() to drop their final reference.
  */
 void drm_dev_unregister(struct drm_device *dev)
 {
@@ -597,18 +765,41 @@ void drm_dev_unregister(struct drm_device *dev)
 	if (dev->driver->unload)
 		dev->driver->unload(dev);
 
-	if (dev->driver->bus->agp_destroy)
-		dev->driver->bus->agp_destroy(dev);
+	if (dev->agp)
+		drm_pci_agp_destroy(dev);
 
 	drm_vblank_cleanup(dev);
 
 	list_for_each_entry_safe(r_list, list_temp, &dev->maplist, head)
 		drm_rmmap(dev, r_list->map);
 
-	drm_unplug_minor(dev->control);
-	drm_unplug_minor(dev->render);
-	drm_unplug_minor(dev->primary);
-
-	list_del(&dev->driver_item);
+	drm_minor_unregister(dev, DRM_MINOR_LEGACY);
+	drm_minor_unregister(dev, DRM_MINOR_RENDER);
+	drm_minor_unregister(dev, DRM_MINOR_CONTROL);
 }
 EXPORT_SYMBOL(drm_dev_unregister);
+
+/**
+ * drm_dev_set_unique - Set the unique name of a DRM device
+ * @dev: device of which to set the unique name
+ * @fmt: format string for unique name
+ *
+ * Sets the unique name of a DRM device using the specified format string and
+ * a variable list of arguments. Drivers can use this at driver probe time if
+ * the unique name of the devices they drive is static.
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
+int drm_dev_set_unique(struct drm_device *dev, const char *fmt, ...)
+{
+	va_list ap;
+
+	kfree(dev->unique);
+
+	va_start(ap, fmt);
+	dev->unique = kvasprintf(GFP_KERNEL, fmt, ap);
+	va_end(ap);
+
+	return dev->unique ? 0 : -ENOMEM;
+}
+EXPORT_SYMBOL(drm_dev_set_unique);
diff --git a/drivers/gpu/drm/drm_sysfs.c b/drivers/gpu/drm/drm_sysfs.c
index c22c3097c3e..369b26278e7 100644
--- a/drivers/gpu/drm/drm_sysfs.c
+++ b/drivers/gpu/drm/drm_sysfs.c
@@ -380,9 +380,9 @@ int drm_sysfs_connector_add(struct drm_connector *connector)
 
 	connector->kdev = device_create(drm_class, dev->primary->kdev,
 					0, connector, "card%d-%s",
-					dev->primary->index, drm_get_connector_name(connector));
+					dev->primary->index, connector->name);
 	DRM_DEBUG("adding \"%s\" to sysfs\n",
-		  drm_get_connector_name(connector));
+		  connector->name);
 
 	if (IS_ERR(connector->kdev)) {
 		DRM_ERROR("failed to register connector device: %ld\n", PTR_ERR(connector->kdev));
@@ -460,7 +460,7 @@ void drm_sysfs_connector_remove(struct drm_connector *connector)
 	if (!connector->kdev)
 		return;
 	DRM_DEBUG("removing \"%s\" from sysfs\n",
-		  drm_get_connector_name(connector));
+		  connector->name);
 
 	for (i = 0; i < ARRAY_SIZE(connector_attrs); i++)
 		device_remove_file(connector->kdev, &connector_attrs[i]);
diff --git a/drivers/gpu/drm/drm_usb.c b/drivers/gpu/drm/drm_usb.c
index b179b70e785..f2fe94aab90 100644
--- a/drivers/gpu/drm/drm_usb.c
+++ b/drivers/gpu/drm/drm_usb.c
@@ -1,4 +1,5 @@
 #include <drm/drmP.h>
+#include <drm/drm_usb.h>
 #include <linux/usb.h>
 #include <linux/module.h>
 
@@ -29,22 +30,12 @@ int drm_get_usb_dev(struct usb_interface *interface,
 	return 0;
 
 err_free:
-	drm_dev_free(dev);
+	drm_dev_unref(dev);
 	return ret;
 
 }
 EXPORT_SYMBOL(drm_get_usb_dev);
 
-static int drm_usb_get_irq(struct drm_device *dev)
-{
-	return 0;
-}
-
-static const char *drm_usb_get_name(struct drm_device *dev)
-{
-	return "USB";
-}
-
 static int drm_usb_set_busid(struct drm_device *dev,
 			       struct drm_master *master)
 {
@@ -52,19 +43,24 @@ static int drm_usb_set_busid(struct drm_device *dev,
 }
 
 static struct drm_bus drm_usb_bus = {
-	.bus_type = DRIVER_BUS_USB,
-	.get_irq = drm_usb_get_irq,
-	.get_name = drm_usb_get_name,
 	.set_busid = drm_usb_set_busid,
 };
-    
+
+/**
+ * drm_usb_init - Register matching USB devices with the DRM subsystem
+ * @driver: DRM device driver
+ * @udriver: USB device driver
+ *
+ * Registers one or more devices matched by a USB driver with the DRM
+ * subsystem.
+ *
+ * Return: 0 on success or a negative error code on failure.
+ */
 int drm_usb_init(struct drm_driver *driver, struct usb_driver *udriver)
 {
 	int res;
 	DRM_DEBUG("\n");
 
-	INIT_LIST_HEAD(&driver->device_list);
-	driver->kdriver.usb = udriver;
 	driver->bus = &drm_usb_bus;
 
 	res = usb_register(udriver);
@@ -72,6 +68,14 @@ int drm_usb_init(struct drm_driver *driver, struct usb_driver *udriver)
 }
 EXPORT_SYMBOL(drm_usb_init);
 
+/**
+ * drm_usb_exit - Unregister matching USB devices from the DRM subsystem
+ * @driver: DRM device driver
+ * @udriver: USB device driver
+ *
+ * Unregisters one or more devices matched by a USB driver from the DRM
+ * subsystem.
+ */
 void drm_usb_exit(struct drm_driver *driver,
 		  struct usb_driver *udriver)
 {
diff --git a/drivers/gpu/drm/drm_vm.c b/drivers/gpu/drm/drm_vm.c
index 93e95d7efd5..24e045c4f53 100644
--- a/drivers/gpu/drm/drm_vm.c
+++ b/drivers/gpu/drm/drm_vm.c
@@ -101,7 +101,7 @@ static int drm_do_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	/*
 	 * Find the right map
 	 */
-	if (!drm_core_has_AGP(dev))
+	if (!dev->agp)
 		goto vm_fault_error;
 
 	if (!dev->agp || !dev->agp->cant_use_aperture)
@@ -220,7 +220,6 @@ static void drm_vm_shm_close(struct vm_area_struct *vma)
 
 	DRM_DEBUG("0x%08lx,0x%08lx\n",
 		  vma->vm_start, vma->vm_end - vma->vm_start);
-	atomic_dec(&dev->vma_count);
 
 	map = vma->vm_private_data;
 
@@ -266,9 +265,6 @@ static void drm_vm_shm_close(struct vm_area_struct *vma)
 				dmah.size = map->size;
 				__drm_pci_free(dev, &dmah);
 				break;
-			case _DRM_GEM:
-				DRM_ERROR("tried to rmmap GEM object\n");
-				break;
 			}
 			kfree(map);
 		}
@@ -408,7 +404,6 @@ void drm_vm_open_locked(struct drm_device *dev,
 
 	DRM_DEBUG("0x%08lx,0x%08lx\n",
 		  vma->vm_start, vma->vm_end - vma->vm_start);
-	atomic_inc(&dev->vma_count);
 
 	vma_entry = kmalloc(sizeof(*vma_entry), GFP_KERNEL);
 	if (vma_entry) {
@@ -436,7 +431,6 @@ void drm_vm_close_locked(struct drm_device *dev,
 
 	DRM_DEBUG("0x%08lx,0x%08lx\n",
 		  vma->vm_start, vma->vm_end - vma->vm_start);
-	atomic_dec(&dev->vma_count);
 
 	list_for_each_entry_safe(pt, temp, &dev->vmalist, head) {
 		if (pt->vma == vma) {
@@ -595,7 +589,7 @@ int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
 	switch (map->type) {
 #if !defined(__arm__)
 	case _DRM_AGP:
-		if (drm_core_has_AGP(dev) && dev->agp->cant_use_aperture) {
+		if (dev->agp && dev->agp->cant_use_aperture) {
 			/*
 			 * On some platforms we can't talk to bus dma address from the CPU, so for
 			 * memory of type DRM_AGP, we'll deal with sorting out the real physical
diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index f227f544aa3..178d2a9672a 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -26,11 +26,35 @@ config DRM_EXYNOS_DMABUF
 
 config DRM_EXYNOS_FIMD
 	bool "Exynos DRM FIMD"
-	depends on DRM_EXYNOS && !FB_S3C && !ARCH_MULTIPLATFORM
+	depends on DRM_EXYNOS && !FB_S3C
 	select FB_MODE_HELPERS
 	help
 	  Choose this option if you want to use Exynos FIMD for DRM.
 
+config DRM_EXYNOS_DPI
+	bool "EXYNOS DRM parallel output support"
+	depends on DRM_EXYNOS_FIMD
+	select DRM_PANEL
+	default n
+	help
+	  This enables support for Exynos parallel output.
+
+config DRM_EXYNOS_DSI
+	bool "EXYNOS DRM MIPI-DSI driver support"
+	depends on DRM_EXYNOS_FIMD
+	select DRM_MIPI_DSI
+	select DRM_PANEL
+	default n
+	help
+	  This enables support for Exynos MIPI-DSI device.
+
+config DRM_EXYNOS_DP
+	bool "EXYNOS DRM DP driver support"
+	depends on DRM_EXYNOS_FIMD && ARCH_EXYNOS && (DRM_PTN3460=n || DRM_PTN3460=y || DRM_PTN3460=DRM_EXYNOS)
+	default DRM_EXYNOS
+	help
+	  This enables support for DP device.
+
 config DRM_EXYNOS_HDMI
 	bool "Exynos DRM HDMI"
 	depends on DRM_EXYNOS && !VIDEO_SAMSUNG_S5P_TV
@@ -51,7 +75,7 @@ config DRM_EXYNOS_G2D
 
 config DRM_EXYNOS_IPP
 	bool "Exynos DRM IPP"
-	depends on DRM_EXYNOS && !ARCH_MULTIPLATFORM
+	depends on DRM_EXYNOS
 	help
 	  Choose this option if you want to use IPP feature for DRM.
 
@@ -69,6 +93,6 @@ config DRM_EXYNOS_ROTATOR
 
 config DRM_EXYNOS_GSC
 	bool "Exynos DRM GSC"
-	depends on DRM_EXYNOS_IPP && ARCH_EXYNOS5
+	depends on DRM_EXYNOS_IPP && ARCH_EXYNOS5 && !ARCH_MULTIPLATFORM
 	help
 	  Choose this option if you want to use Exynos GSC for DRM.
diff --git a/drivers/gpu/drm/exynos/Makefile b/drivers/gpu/drm/exynos/Makefile
index 639b49e1ec0..33ae3652b8d 100644
--- a/drivers/gpu/drm/exynos/Makefile
+++ b/drivers/gpu/drm/exynos/Makefile
@@ -3,7 +3,7 @@
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
 ccflags-y := -Iinclude/drm -Idrivers/gpu/drm/exynos
-exynosdrm-y := exynos_drm_drv.o exynos_drm_encoder.o exynos_drm_connector.o \
+exynosdrm-y := exynos_drm_drv.o exynos_drm_encoder.o \
 		exynos_drm_crtc.o exynos_drm_fbdev.o exynos_drm_fb.o \
 		exynos_drm_buf.o exynos_drm_gem.o exynos_drm_core.o \
 		exynos_drm_plane.o
@@ -11,9 +11,10 @@ exynosdrm-y := exynos_drm_drv.o exynos_drm_encoder.o exynos_drm_connector.o \
 exynosdrm-$(CONFIG_DRM_EXYNOS_IOMMU) += exynos_drm_iommu.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_DMABUF) += exynos_drm_dmabuf.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_FIMD)	+= exynos_drm_fimd.o
-exynosdrm-$(CONFIG_DRM_EXYNOS_HDMI)	+= exynos_hdmi.o exynos_mixer.o \
-					   exynos_ddc.o exynos_hdmiphy.o \
-					   exynos_drm_hdmi.o
+exynosdrm-$(CONFIG_DRM_EXYNOS_DPI)	+= exynos_drm_dpi.o
+exynosdrm-$(CONFIG_DRM_EXYNOS_DSI)	+= exynos_drm_dsi.o
+exynosdrm-$(CONFIG_DRM_EXYNOS_DP)	+= exynos_dp_core.o exynos_dp_reg.o
+exynosdrm-$(CONFIG_DRM_EXYNOS_HDMI)	+= exynos_hdmi.o exynos_mixer.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_VIDI)	+= exynos_drm_vidi.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_G2D)	+= exynos_drm_g2d.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_IPP)	+= exynos_drm_ipp.o
diff --git a/drivers/gpu/drm/exynos/exynos_ddc.c b/drivers/gpu/drm/exynos/exynos_ddc.c
deleted file mode 100644
index 6a8c84e7c83..00000000000
--- a/drivers/gpu/drm/exynos/exynos_ddc.c
+++ /dev/null
@@ -1,63 +0,0 @@
-/*
- * Copyright (C) 2011 Samsung Electronics Co.Ltd
- * Authors:
- *	Seung-Woo Kim <sw0312.kim@samsung.com>
- *	Inki Dae <inki.dae@samsung.com>
- *
- * This program is free software; you can redistribute  it and/or modify it
- * under  the terms of  the GNU General  Public License as published by the
- * Free Software Foundation;  either version 2 of the  License, or (at your
- * option) any later version.
- *
- */
-
-#include <drm/drmP.h>
-
-#include <linux/kernel.h>
-#include <linux/i2c.h>
-#include <linux/of.h>
-
-#include "exynos_drm_drv.h"
-#include "exynos_hdmi.h"
-
-static int s5p_ddc_probe(struct i2c_client *client,
-			const struct i2c_device_id *dev_id)
-{
-	hdmi_attach_ddc_client(client);
-
-	dev_info(&client->adapter->dev,
-		"attached %s into i2c adapter successfully\n",
-		client->name);
-
-	return 0;
-}
-
-static int s5p_ddc_remove(struct i2c_client *client)
-{
-	dev_info(&client->adapter->dev,
-		"detached %s from i2c adapter successfully\n",
-		client->name);
-
-	return 0;
-}
-
-static struct of_device_id hdmiddc_match_types[] = {
-	{
-		.compatible = "samsung,exynos5-hdmiddc",
-	}, {
-		.compatible = "samsung,exynos4210-hdmiddc",
-	}, {
-		/* end node */
-	}
-};
-
-struct i2c_driver ddc_driver = {
-	.driver = {
-		.name = "exynos-hdmiddc",
-		.owner = THIS_MODULE,
-		.of_match_table = hdmiddc_match_types,
-	},
-	.probe		= s5p_ddc_probe,
-	.remove		= s5p_ddc_remove,
-	.command		= NULL,
-};
diff --git a/drivers/gpu/drm/exynos/exynos_dp_core.c b/drivers/gpu/drm/exynos/exynos_dp_core.c
new file mode 100644
index 00000000000..a8ffc8c1477
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_dp_core.c
@@ -0,0 +1,1393 @@
+/*
+ * Samsung SoC DP (Display Port) interface driver.
+ *
+ * Copyright (C) 2012 Samsung Electronics Co., Ltd.
+ * Author: Jingoo Han <jg1.han@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ */
+
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>
+#include <linux/gpio.h>
+#include <linux/component.h>
+#include <linux/phy/phy.h>
+#include <video/of_display_timing.h>
+#include <video/of_videomode.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/bridge/ptn3460.h>
+
+#include "exynos_drm_drv.h"
+#include "exynos_dp_core.h"
+
+#define ctx_from_connector(c)	container_of(c, struct exynos_dp_device, \
+					connector)
+
+struct bridge_init {
+	struct i2c_client *client;
+	struct device_node *node;
+};
+
+static int exynos_dp_init_dp(struct exynos_dp_device *dp)
+{
+	exynos_dp_reset(dp);
+
+	exynos_dp_swreset(dp);
+
+	exynos_dp_init_analog_param(dp);
+	exynos_dp_init_interrupt(dp);
+
+	/* SW defined function Normal operation */
+	exynos_dp_enable_sw_function(dp);
+
+	exynos_dp_config_interrupt(dp);
+	exynos_dp_init_analog_func(dp);
+
+	exynos_dp_init_hpd(dp);
+	exynos_dp_init_aux(dp);
+
+	return 0;
+}
+
+static int exynos_dp_detect_hpd(struct exynos_dp_device *dp)
+{
+	int timeout_loop = 0;
+
+	while (exynos_dp_get_plug_in_status(dp) != 0) {
+		timeout_loop++;
+		if (DP_TIMEOUT_LOOP_COUNT < timeout_loop) {
+			dev_err(dp->dev, "failed to get hpd plug status\n");
+			return -ETIMEDOUT;
+		}
+		usleep_range(10, 11);
+	}
+
+	return 0;
+}
+
+static unsigned char exynos_dp_calc_edid_check_sum(unsigned char *edid_data)
+{
+	int i;
+	unsigned char sum = 0;
+
+	for (i = 0; i < EDID_BLOCK_LENGTH; i++)
+		sum = sum + edid_data[i];
+
+	return sum;
+}
+
+static int exynos_dp_read_edid(struct exynos_dp_device *dp)
+{
+	unsigned char edid[EDID_BLOCK_LENGTH * 2];
+	unsigned int extend_block = 0;
+	unsigned char sum;
+	unsigned char test_vector;
+	int retval;
+
+	/*
+	 * EDID device address is 0x50.
+	 * However, if necessary, you must have set upper address
+	 * into E-EDID in I2C device, 0x30.
+	 */
+
+	/* Read Extension Flag, Number of 128-byte EDID extension blocks */
+	retval = exynos_dp_read_byte_from_i2c(dp, I2C_EDID_DEVICE_ADDR,
+				EDID_EXTENSION_FLAG,
+				&extend_block);
+	if (retval)
+		return retval;
+
+	if (extend_block > 0) {
+		dev_dbg(dp->dev, "EDID data includes a single extension!\n");
+
+		/* Read EDID data */
+		retval = exynos_dp_read_bytes_from_i2c(dp, I2C_EDID_DEVICE_ADDR,
+						EDID_HEADER_PATTERN,
+						EDID_BLOCK_LENGTH,
+						&edid[EDID_HEADER_PATTERN]);
+		if (retval != 0) {
+			dev_err(dp->dev, "EDID Read failed!\n");
+			return -EIO;
+		}
+		sum = exynos_dp_calc_edid_check_sum(edid);
+		if (sum != 0) {
+			dev_err(dp->dev, "EDID bad checksum!\n");
+			return -EIO;
+		}
+
+		/* Read additional EDID data */
+		retval = exynos_dp_read_bytes_from_i2c(dp,
+				I2C_EDID_DEVICE_ADDR,
+				EDID_BLOCK_LENGTH,
+				EDID_BLOCK_LENGTH,
+				&edid[EDID_BLOCK_LENGTH]);
+		if (retval != 0) {
+			dev_err(dp->dev, "EDID Read failed!\n");
+			return -EIO;
+		}
+		sum = exynos_dp_calc_edid_check_sum(&edid[EDID_BLOCK_LENGTH]);
+		if (sum != 0) {
+			dev_err(dp->dev, "EDID bad checksum!\n");
+			return -EIO;
+		}
+
+		exynos_dp_read_byte_from_dpcd(dp, DP_TEST_REQUEST,
+					&test_vector);
+		if (test_vector & DP_TEST_LINK_EDID_READ) {
+			exynos_dp_write_byte_to_dpcd(dp,
+				DP_TEST_EDID_CHECKSUM,
+				edid[EDID_BLOCK_LENGTH + EDID_CHECKSUM]);
+			exynos_dp_write_byte_to_dpcd(dp,
+				DP_TEST_RESPONSE,
+				DP_TEST_EDID_CHECKSUM_WRITE);
+		}
+	} else {
+		dev_info(dp->dev, "EDID data does not include any extensions.\n");
+
+		/* Read EDID data */
+		retval = exynos_dp_read_bytes_from_i2c(dp,
+				I2C_EDID_DEVICE_ADDR,
+				EDID_HEADER_PATTERN,
+				EDID_BLOCK_LENGTH,
+				&edid[EDID_HEADER_PATTERN]);
+		if (retval != 0) {
+			dev_err(dp->dev, "EDID Read failed!\n");
+			return -EIO;
+		}
+		sum = exynos_dp_calc_edid_check_sum(edid);
+		if (sum != 0) {
+			dev_err(dp->dev, "EDID bad checksum!\n");
+			return -EIO;
+		}
+
+		exynos_dp_read_byte_from_dpcd(dp,
+			DP_TEST_REQUEST,
+			&test_vector);
+		if (test_vector & DP_TEST_LINK_EDID_READ) {
+			exynos_dp_write_byte_to_dpcd(dp,
+				DP_TEST_EDID_CHECKSUM,
+				edid[EDID_CHECKSUM]);
+			exynos_dp_write_byte_to_dpcd(dp,
+				DP_TEST_RESPONSE,
+				DP_TEST_EDID_CHECKSUM_WRITE);
+		}
+	}
+
+	dev_err(dp->dev, "EDID Read success!\n");
+	return 0;
+}
+
+static int exynos_dp_handle_edid(struct exynos_dp_device *dp)
+{
+	u8 buf[12];
+	int i;
+	int retval;
+
+	/* Read DPCD DP_DPCD_REV~RECEIVE_PORT1_CAP_1 */
+	retval = exynos_dp_read_bytes_from_dpcd(dp, DP_DPCD_REV,
+				12, buf);
+	if (retval)
+		return retval;
+
+	/* Read EDID */
+	for (i = 0; i < 3; i++) {
+		retval = exynos_dp_read_edid(dp);
+		if (!retval)
+			break;
+	}
+
+	return retval;
+}
+
+static void exynos_dp_enable_rx_to_enhanced_mode(struct exynos_dp_device *dp,
+						bool enable)
+{
+	u8 data;
+
+	exynos_dp_read_byte_from_dpcd(dp, DP_LANE_COUNT_SET, &data);
+
+	if (enable)
+		exynos_dp_write_byte_to_dpcd(dp, DP_LANE_COUNT_SET,
+			DP_LANE_COUNT_ENHANCED_FRAME_EN |
+			DPCD_LANE_COUNT_SET(data));
+	else
+		exynos_dp_write_byte_to_dpcd(dp, DP_LANE_COUNT_SET,
+			DPCD_LANE_COUNT_SET(data));
+}
+
+static int exynos_dp_is_enhanced_mode_available(struct exynos_dp_device *dp)
+{
+	u8 data;
+	int retval;
+
+	exynos_dp_read_byte_from_dpcd(dp, DP_MAX_LANE_COUNT, &data);
+	retval = DPCD_ENHANCED_FRAME_CAP(data);
+
+	return retval;
+}
+
+static void exynos_dp_set_enhanced_mode(struct exynos_dp_device *dp)
+{
+	u8 data;
+
+	data = exynos_dp_is_enhanced_mode_available(dp);
+	exynos_dp_enable_rx_to_enhanced_mode(dp, data);
+	exynos_dp_enable_enhanced_mode(dp, data);
+}
+
+static void exynos_dp_training_pattern_dis(struct exynos_dp_device *dp)
+{
+	exynos_dp_set_training_pattern(dp, DP_NONE);
+
+	exynos_dp_write_byte_to_dpcd(dp,
+		DP_TRAINING_PATTERN_SET,
+		DP_TRAINING_PATTERN_DISABLE);
+}
+
+static void exynos_dp_set_lane_lane_pre_emphasis(struct exynos_dp_device *dp,
+					int pre_emphasis, int lane)
+{
+	switch (lane) {
+	case 0:
+		exynos_dp_set_lane0_pre_emphasis(dp, pre_emphasis);
+		break;
+	case 1:
+		exynos_dp_set_lane1_pre_emphasis(dp, pre_emphasis);
+		break;
+
+	case 2:
+		exynos_dp_set_lane2_pre_emphasis(dp, pre_emphasis);
+		break;
+
+	case 3:
+		exynos_dp_set_lane3_pre_emphasis(dp, pre_emphasis);
+		break;
+	}
+}
+
+static int exynos_dp_link_start(struct exynos_dp_device *dp)
+{
+	u8 buf[4];
+	int lane, lane_count, pll_tries, retval;
+
+	lane_count = dp->link_train.lane_count;
+
+	dp->link_train.lt_state = CLOCK_RECOVERY;
+	dp->link_train.eq_loop = 0;
+
+	for (lane = 0; lane < lane_count; lane++)
+		dp->link_train.cr_loop[lane] = 0;
+
+	/* Set link rate and count as you want to establish*/
+	exynos_dp_set_link_bandwidth(dp, dp->link_train.link_rate);
+	exynos_dp_set_lane_count(dp, dp->link_train.lane_count);
+
+	/* Setup RX configuration */
+	buf[0] = dp->link_train.link_rate;
+	buf[1] = dp->link_train.lane_count;
+	retval = exynos_dp_write_bytes_to_dpcd(dp, DP_LINK_BW_SET,
+				2, buf);
+	if (retval)
+		return retval;
+
+	/* Set TX pre-emphasis to minimum */
+	for (lane = 0; lane < lane_count; lane++)
+		exynos_dp_set_lane_lane_pre_emphasis(dp,
+			PRE_EMPHASIS_LEVEL_0, lane);
+
+	/* Wait for PLL lock */
+	pll_tries = 0;
+	while (exynos_dp_get_pll_lock_status(dp) == PLL_UNLOCKED) {
+		if (pll_tries == DP_TIMEOUT_LOOP_COUNT) {
+			dev_err(dp->dev, "Wait for PLL lock timed out\n");
+			return -ETIMEDOUT;
+		}
+
+		pll_tries++;
+		usleep_range(90, 120);
+	}
+
+	/* Set training pattern 1 */
+	exynos_dp_set_training_pattern(dp, TRAINING_PTN1);
+
+	/* Set RX training pattern */
+	retval = exynos_dp_write_byte_to_dpcd(dp,
+			DP_TRAINING_PATTERN_SET,
+			DP_LINK_SCRAMBLING_DISABLE | DP_TRAINING_PATTERN_1);
+	if (retval)
+		return retval;
+
+	for (lane = 0; lane < lane_count; lane++)
+		buf[lane] = DP_TRAIN_PRE_EMPHASIS_0 |
+			    DP_TRAIN_VOLTAGE_SWING_400;
+
+	retval = exynos_dp_write_bytes_to_dpcd(dp, DP_TRAINING_LANE0_SET,
+			lane_count, buf);
+
+	return retval;
+}
+
+static unsigned char exynos_dp_get_lane_status(u8 link_status[2], int lane)
+{
+	int shift = (lane & 1) * 4;
+	u8 link_value = link_status[lane>>1];
+
+	return (link_value >> shift) & 0xf;
+}
+
+static int exynos_dp_clock_recovery_ok(u8 link_status[2], int lane_count)
+{
+	int lane;
+	u8 lane_status;
+
+	for (lane = 0; lane < lane_count; lane++) {
+		lane_status = exynos_dp_get_lane_status(link_status, lane);
+		if ((lane_status & DP_LANE_CR_DONE) == 0)
+			return -EINVAL;
+	}
+	return 0;
+}
+
+static int exynos_dp_channel_eq_ok(u8 link_status[2], u8 link_align,
+				int lane_count)
+{
+	int lane;
+	u8 lane_status;
+
+	if ((link_align & DP_INTERLANE_ALIGN_DONE) == 0)
+		return -EINVAL;
+
+	for (lane = 0; lane < lane_count; lane++) {
+		lane_status = exynos_dp_get_lane_status(link_status, lane);
+		lane_status &= DP_CHANNEL_EQ_BITS;
+		if (lane_status != DP_CHANNEL_EQ_BITS)
+			return -EINVAL;
+	}
+
+	return 0;
+}
+
+static unsigned char exynos_dp_get_adjust_request_voltage(u8 adjust_request[2],
+							int lane)
+{
+	int shift = (lane & 1) * 4;
+	u8 link_value = adjust_request[lane>>1];
+
+	return (link_value >> shift) & 0x3;
+}
+
+static unsigned char exynos_dp_get_adjust_request_pre_emphasis(
+					u8 adjust_request[2],
+					int lane)
+{
+	int shift = (lane & 1) * 4;
+	u8 link_value = adjust_request[lane>>1];
+
+	return ((link_value >> shift) & 0xc) >> 2;
+}
+
+static void exynos_dp_set_lane_link_training(struct exynos_dp_device *dp,
+					u8 training_lane_set, int lane)
+{
+	switch (lane) {
+	case 0:
+		exynos_dp_set_lane0_link_training(dp, training_lane_set);
+		break;
+	case 1:
+		exynos_dp_set_lane1_link_training(dp, training_lane_set);
+		break;
+
+	case 2:
+		exynos_dp_set_lane2_link_training(dp, training_lane_set);
+		break;
+
+	case 3:
+		exynos_dp_set_lane3_link_training(dp, training_lane_set);
+		break;
+	}
+}
+
+static unsigned int exynos_dp_get_lane_link_training(
+				struct exynos_dp_device *dp,
+				int lane)
+{
+	u32 reg;
+
+	switch (lane) {
+	case 0:
+		reg = exynos_dp_get_lane0_link_training(dp);
+		break;
+	case 1:
+		reg = exynos_dp_get_lane1_link_training(dp);
+		break;
+	case 2:
+		reg = exynos_dp_get_lane2_link_training(dp);
+		break;
+	case 3:
+		reg = exynos_dp_get_lane3_link_training(dp);
+		break;
+	default:
+		WARN_ON(1);
+		return 0;
+	}
+
+	return reg;
+}
+
+static void exynos_dp_reduce_link_rate(struct exynos_dp_device *dp)
+{
+	exynos_dp_training_pattern_dis(dp);
+	exynos_dp_set_enhanced_mode(dp);
+
+	dp->link_train.lt_state = FAILED;
+}
+
+static void exynos_dp_get_adjust_training_lane(struct exynos_dp_device *dp,
+					u8 adjust_request[2])
+{
+	int lane, lane_count;
+	u8 voltage_swing, pre_emphasis, training_lane;
+
+	lane_count = dp->link_train.lane_count;
+	for (lane = 0; lane < lane_count; lane++) {
+		voltage_swing = exynos_dp_get_adjust_request_voltage(
+						adjust_request, lane);
+		pre_emphasis = exynos_dp_get_adjust_request_pre_emphasis(
+						adjust_request, lane);
+		training_lane = DPCD_VOLTAGE_SWING_SET(voltage_swing) |
+				DPCD_PRE_EMPHASIS_SET(pre_emphasis);
+
+		if (voltage_swing == VOLTAGE_LEVEL_3)
+			training_lane |= DP_TRAIN_MAX_SWING_REACHED;
+		if (pre_emphasis == PRE_EMPHASIS_LEVEL_3)
+			training_lane |= DP_TRAIN_MAX_PRE_EMPHASIS_REACHED;
+
+		dp->link_train.training_lane[lane] = training_lane;
+	}
+}
+
+static int exynos_dp_process_clock_recovery(struct exynos_dp_device *dp)
+{
+	int lane, lane_count, retval;
+	u8 voltage_swing, pre_emphasis, training_lane;
+	u8 link_status[2], adjust_request[2];
+
+	usleep_range(100, 101);
+
+	lane_count = dp->link_train.lane_count;
+
+	retval =  exynos_dp_read_bytes_from_dpcd(dp,
+			DP_LANE0_1_STATUS, 2, link_status);
+	if (retval)
+		return retval;
+
+	retval =  exynos_dp_read_bytes_from_dpcd(dp,
+			DP_ADJUST_REQUEST_LANE0_1, 2, adjust_request);
+	if (retval)
+		return retval;
+
+	if (exynos_dp_clock_recovery_ok(link_status, lane_count) == 0) {
+		/* set training pattern 2 for EQ */
+		exynos_dp_set_training_pattern(dp, TRAINING_PTN2);
+
+		retval = exynos_dp_write_byte_to_dpcd(dp,
+				DP_TRAINING_PATTERN_SET,
+				DP_LINK_SCRAMBLING_DISABLE |
+				DP_TRAINING_PATTERN_2);
+		if (retval)
+			return retval;
+
+		dev_info(dp->dev, "Link Training Clock Recovery success\n");
+		dp->link_train.lt_state = EQUALIZER_TRAINING;
+	} else {
+		for (lane = 0; lane < lane_count; lane++) {
+			training_lane = exynos_dp_get_lane_link_training(
+							dp, lane);
+			voltage_swing = exynos_dp_get_adjust_request_voltage(
+							adjust_request, lane);
+			pre_emphasis = exynos_dp_get_adjust_request_pre_emphasis(
+							adjust_request, lane);
+
+			if (DPCD_VOLTAGE_SWING_GET(training_lane) ==
+					voltage_swing &&
+			    DPCD_PRE_EMPHASIS_GET(training_lane) ==
+					pre_emphasis)
+				dp->link_train.cr_loop[lane]++;
+
+			if (dp->link_train.cr_loop[lane] == MAX_CR_LOOP ||
+			    voltage_swing == VOLTAGE_LEVEL_3 ||
+			    pre_emphasis == PRE_EMPHASIS_LEVEL_3) {
+				dev_err(dp->dev, "CR Max reached (%d,%d,%d)\n",
+					dp->link_train.cr_loop[lane],
+					voltage_swing, pre_emphasis);
+				exynos_dp_reduce_link_rate(dp);
+				return -EIO;
+			}
+		}
+	}
+
+	exynos_dp_get_adjust_training_lane(dp, adjust_request);
+
+	for (lane = 0; lane < lane_count; lane++)
+		exynos_dp_set_lane_link_training(dp,
+			dp->link_train.training_lane[lane], lane);
+
+	retval = exynos_dp_write_bytes_to_dpcd(dp,
+			DP_TRAINING_LANE0_SET, lane_count,
+			dp->link_train.training_lane);
+	if (retval)
+		return retval;
+
+	return retval;
+}
+
+static int exynos_dp_process_equalizer_training(struct exynos_dp_device *dp)
+{
+	int lane, lane_count, retval;
+	u32 reg;
+	u8 link_align, link_status[2], adjust_request[2];
+
+	usleep_range(400, 401);
+
+	lane_count = dp->link_train.lane_count;
+
+	retval = exynos_dp_read_bytes_from_dpcd(dp,
+			DP_LANE0_1_STATUS, 2, link_status);
+	if (retval)
+		return retval;
+
+	if (exynos_dp_clock_recovery_ok(link_status, lane_count)) {
+		exynos_dp_reduce_link_rate(dp);
+		return -EIO;
+	}
+
+	retval = exynos_dp_read_bytes_from_dpcd(dp,
+			DP_ADJUST_REQUEST_LANE0_1, 2, adjust_request);
+	if (retval)
+		return retval;
+
+	retval = exynos_dp_read_byte_from_dpcd(dp,
+			DP_LANE_ALIGN_STATUS_UPDATED, &link_align);
+	if (retval)
+		return retval;
+
+	exynos_dp_get_adjust_training_lane(dp, adjust_request);
+
+	if (!exynos_dp_channel_eq_ok(link_status, link_align, lane_count)) {
+		/* traing pattern Set to Normal */
+		exynos_dp_training_pattern_dis(dp);
+
+		dev_info(dp->dev, "Link Training success!\n");
+
+		exynos_dp_get_link_bandwidth(dp, &reg);
+		dp->link_train.link_rate = reg;
+		dev_dbg(dp->dev, "final bandwidth = %.2x\n",
+			dp->link_train.link_rate);
+
+		exynos_dp_get_lane_count(dp, &reg);
+		dp->link_train.lane_count = reg;
+		dev_dbg(dp->dev, "final lane count = %.2x\n",
+			dp->link_train.lane_count);
+
+		/* set enhanced mode if available */
+		exynos_dp_set_enhanced_mode(dp);
+		dp->link_train.lt_state = FINISHED;
+
+		return 0;
+	}
+
+	/* not all locked */
+	dp->link_train.eq_loop++;
+
+	if (dp->link_train.eq_loop > MAX_EQ_LOOP) {
+		dev_err(dp->dev, "EQ Max loop\n");
+		exynos_dp_reduce_link_rate(dp);
+		return -EIO;
+	}
+
+	for (lane = 0; lane < lane_count; lane++)
+		exynos_dp_set_lane_link_training(dp,
+			dp->link_train.training_lane[lane], lane);
+
+	retval = exynos_dp_write_bytes_to_dpcd(dp, DP_TRAINING_LANE0_SET,
+			lane_count, dp->link_train.training_lane);
+
+	return retval;
+}
+
+static void exynos_dp_get_max_rx_bandwidth(struct exynos_dp_device *dp,
+					u8 *bandwidth)
+{
+	u8 data;
+
+	/*
+	 * For DP rev.1.1, Maximum link rate of Main Link lanes
+	 * 0x06 = 1.62 Gbps, 0x0a = 2.7 Gbps
+	 */
+	exynos_dp_read_byte_from_dpcd(dp, DP_MAX_LINK_RATE, &data);
+	*bandwidth = data;
+}
+
+static void exynos_dp_get_max_rx_lane_count(struct exynos_dp_device *dp,
+					u8 *lane_count)
+{
+	u8 data;
+
+	/*
+	 * For DP rev.1.1, Maximum number of Main Link lanes
+	 * 0x01 = 1 lane, 0x02 = 2 lanes, 0x04 = 4 lanes
+	 */
+	exynos_dp_read_byte_from_dpcd(dp, DP_MAX_LANE_COUNT, &data);
+	*lane_count = DPCD_MAX_LANE_COUNT(data);
+}
+
+static void exynos_dp_init_training(struct exynos_dp_device *dp,
+			enum link_lane_count_type max_lane,
+			enum link_rate_type max_rate)
+{
+	/*
+	 * MACRO_RST must be applied after the PLL_LOCK to avoid
+	 * the DP inter pair skew issue for at least 10 us
+	 */
+	exynos_dp_reset_macro(dp);
+
+	/* Initialize by reading RX's DPCD */
+	exynos_dp_get_max_rx_bandwidth(dp, &dp->link_train.link_rate);
+	exynos_dp_get_max_rx_lane_count(dp, &dp->link_train.lane_count);
+
+	if ((dp->link_train.link_rate != LINK_RATE_1_62GBPS) &&
+	   (dp->link_train.link_rate != LINK_RATE_2_70GBPS)) {
+		dev_err(dp->dev, "Rx Max Link Rate is abnormal :%x !\n",
+			dp->link_train.link_rate);
+		dp->link_train.link_rate = LINK_RATE_1_62GBPS;
+	}
+
+	if (dp->link_train.lane_count == 0) {
+		dev_err(dp->dev, "Rx Max Lane count is abnormal :%x !\n",
+			dp->link_train.lane_count);
+		dp->link_train.lane_count = (u8)LANE_COUNT1;
+	}
+
+	/* Setup TX lane count & rate */
+	if (dp->link_train.lane_count > max_lane)
+		dp->link_train.lane_count = max_lane;
+	if (dp->link_train.link_rate > max_rate)
+		dp->link_train.link_rate = max_rate;
+
+	/* All DP analog module power up */
+	exynos_dp_set_analog_power_down(dp, POWER_ALL, 0);
+}
+
+static int exynos_dp_sw_link_training(struct exynos_dp_device *dp)
+{
+	int retval = 0, training_finished = 0;
+
+	dp->link_train.lt_state = START;
+
+	/* Process here */
+	while (!retval && !training_finished) {
+		switch (dp->link_train.lt_state) {
+		case START:
+			retval = exynos_dp_link_start(dp);
+			if (retval)
+				dev_err(dp->dev, "LT link start failed!\n");
+			break;
+		case CLOCK_RECOVERY:
+			retval = exynos_dp_process_clock_recovery(dp);
+			if (retval)
+				dev_err(dp->dev, "LT CR failed!\n");
+			break;
+		case EQUALIZER_TRAINING:
+			retval = exynos_dp_process_equalizer_training(dp);
+			if (retval)
+				dev_err(dp->dev, "LT EQ failed!\n");
+			break;
+		case FINISHED:
+			training_finished = 1;
+			break;
+		case FAILED:
+			return -EREMOTEIO;
+		}
+	}
+	if (retval)
+		dev_err(dp->dev, "eDP link training failed (%d)\n", retval);
+
+	return retval;
+}
+
+static int exynos_dp_set_link_train(struct exynos_dp_device *dp,
+				u32 count,
+				u32 bwtype)
+{
+	int i;
+	int retval;
+
+	for (i = 0; i < DP_TIMEOUT_LOOP_COUNT; i++) {
+		exynos_dp_init_training(dp, count, bwtype);
+		retval = exynos_dp_sw_link_training(dp);
+		if (retval == 0)
+			break;
+
+		usleep_range(100, 110);
+	}
+
+	return retval;
+}
+
+static int exynos_dp_config_video(struct exynos_dp_device *dp)
+{
+	int retval = 0;
+	int timeout_loop = 0;
+	int done_count = 0;
+
+	exynos_dp_config_video_slave_mode(dp);
+
+	exynos_dp_set_video_color_format(dp);
+
+	if (exynos_dp_get_pll_lock_status(dp) == PLL_UNLOCKED) {
+		dev_err(dp->dev, "PLL is not locked yet.\n");
+		return -EINVAL;
+	}
+
+	for (;;) {
+		timeout_loop++;
+		if (exynos_dp_is_slave_video_stream_clock_on(dp) == 0)
+			break;
+		if (DP_TIMEOUT_LOOP_COUNT < timeout_loop) {
+			dev_err(dp->dev, "Timeout of video streamclk ok\n");
+			return -ETIMEDOUT;
+		}
+
+		usleep_range(1, 2);
+	}
+
+	/* Set to use the register calculated M/N video */
+	exynos_dp_set_video_cr_mn(dp, CALCULATED_M, 0, 0);
+
+	/* For video bist, Video timing must be generated by register */
+	exynos_dp_set_video_timing_mode(dp, VIDEO_TIMING_FROM_CAPTURE);
+
+	/* Disable video mute */
+	exynos_dp_enable_video_mute(dp, 0);
+
+	/* Configure video slave mode */
+	exynos_dp_enable_video_master(dp, 0);
+
+	/* Enable video */
+	exynos_dp_start_video(dp);
+
+	timeout_loop = 0;
+
+	for (;;) {
+		timeout_loop++;
+		if (exynos_dp_is_video_stream_on(dp) == 0) {
+			done_count++;
+			if (done_count > 10)
+				break;
+		} else if (done_count) {
+			done_count = 0;
+		}
+		if (DP_TIMEOUT_LOOP_COUNT < timeout_loop) {
+			dev_err(dp->dev, "Timeout of video streamclk ok\n");
+			return -ETIMEDOUT;
+		}
+
+		usleep_range(1000, 1001);
+	}
+
+	if (retval != 0)
+		dev_err(dp->dev, "Video stream is not detected!\n");
+
+	return retval;
+}
+
+static void exynos_dp_enable_scramble(struct exynos_dp_device *dp, bool enable)
+{
+	u8 data;
+
+	if (enable) {
+		exynos_dp_enable_scrambling(dp);
+
+		exynos_dp_read_byte_from_dpcd(dp,
+			DP_TRAINING_PATTERN_SET,
+			&data);
+		exynos_dp_write_byte_to_dpcd(dp,
+			DP_TRAINING_PATTERN_SET,
+			(u8)(data & ~DP_LINK_SCRAMBLING_DISABLE));
+	} else {
+		exynos_dp_disable_scrambling(dp);
+
+		exynos_dp_read_byte_from_dpcd(dp,
+			DP_TRAINING_PATTERN_SET,
+			&data);
+		exynos_dp_write_byte_to_dpcd(dp,
+			DP_TRAINING_PATTERN_SET,
+			(u8)(data | DP_LINK_SCRAMBLING_DISABLE));
+	}
+}
+
+static irqreturn_t exynos_dp_irq_handler(int irq, void *arg)
+{
+	struct exynos_dp_device *dp = arg;
+
+	enum dp_irq_type irq_type;
+
+	irq_type = exynos_dp_get_irq_type(dp);
+	switch (irq_type) {
+	case DP_IRQ_TYPE_HP_CABLE_IN:
+		dev_dbg(dp->dev, "Received irq - cable in\n");
+		schedule_work(&dp->hotplug_work);
+		exynos_dp_clear_hotplug_interrupts(dp);
+		break;
+	case DP_IRQ_TYPE_HP_CABLE_OUT:
+		dev_dbg(dp->dev, "Received irq - cable out\n");
+		exynos_dp_clear_hotplug_interrupts(dp);
+		break;
+	case DP_IRQ_TYPE_HP_CHANGE:
+		/*
+		 * We get these change notifications once in a while, but there
+		 * is nothing we can do with them. Just ignore it for now and
+		 * only handle cable changes.
+		 */
+		dev_dbg(dp->dev, "Received irq - hotplug change; ignoring.\n");
+		exynos_dp_clear_hotplug_interrupts(dp);
+		break;
+	default:
+		dev_err(dp->dev, "Received irq - unknown type!\n");
+		break;
+	}
+	return IRQ_HANDLED;
+}
+
+static void exynos_dp_hotplug(struct work_struct *work)
+{
+	struct exynos_dp_device *dp;
+	int ret;
+
+	dp = container_of(work, struct exynos_dp_device, hotplug_work);
+
+	ret = exynos_dp_detect_hpd(dp);
+	if (ret) {
+		/* Cable has been disconnected, we're done */
+		return;
+	}
+
+	ret = exynos_dp_handle_edid(dp);
+	if (ret) {
+		dev_err(dp->dev, "unable to handle edid\n");
+		return;
+	}
+
+	ret = exynos_dp_set_link_train(dp, dp->video_info->lane_count,
+					dp->video_info->link_rate);
+	if (ret) {
+		dev_err(dp->dev, "unable to do link train\n");
+		return;
+	}
+
+	exynos_dp_enable_scramble(dp, 1);
+	exynos_dp_enable_rx_to_enhanced_mode(dp, 1);
+	exynos_dp_enable_enhanced_mode(dp, 1);
+
+	exynos_dp_set_lane_count(dp, dp->video_info->lane_count);
+	exynos_dp_set_link_bandwidth(dp, dp->video_info->link_rate);
+
+	exynos_dp_init_video(dp);
+	ret = exynos_dp_config_video(dp);
+	if (ret)
+		dev_err(dp->dev, "unable to config video\n");
+}
+
+static enum drm_connector_status exynos_dp_detect(
+				struct drm_connector *connector, bool force)
+{
+	return connector_status_connected;
+}
+
+static void exynos_dp_connector_destroy(struct drm_connector *connector)
+{
+}
+
+static struct drm_connector_funcs exynos_dp_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = exynos_dp_detect,
+	.destroy = exynos_dp_connector_destroy,
+};
+
+static int exynos_dp_get_modes(struct drm_connector *connector)
+{
+	struct exynos_dp_device *dp = ctx_from_connector(connector);
+	struct drm_display_mode *mode;
+
+	mode = drm_mode_create(connector->dev);
+	if (!mode) {
+		DRM_ERROR("failed to create a new display mode.\n");
+		return 0;
+	}
+
+	drm_display_mode_from_videomode(&dp->panel.vm, mode);
+	mode->width_mm = dp->panel.width_mm;
+	mode->height_mm = dp->panel.height_mm;
+	connector->display_info.width_mm = mode->width_mm;
+	connector->display_info.height_mm = mode->height_mm;
+
+	mode->type = DRM_MODE_TYPE_DRIVER | DRM_MODE_TYPE_PREFERRED;
+	drm_mode_set_name(mode);
+	drm_mode_probed_add(connector, mode);
+
+	return 1;
+}
+
+static struct drm_encoder *exynos_dp_best_encoder(
+			struct drm_connector *connector)
+{
+	struct exynos_dp_device *dp = ctx_from_connector(connector);
+
+	return dp->encoder;
+}
+
+static struct drm_connector_helper_funcs exynos_dp_connector_helper_funcs = {
+	.get_modes = exynos_dp_get_modes,
+	.best_encoder = exynos_dp_best_encoder,
+};
+
+static bool find_bridge(const char *compat, struct bridge_init *bridge)
+{
+	bridge->client = NULL;
+	bridge->node = of_find_compatible_node(NULL, NULL, compat);
+	if (!bridge->node)
+		return false;
+
+	bridge->client = of_find_i2c_device_by_node(bridge->node);
+	if (!bridge->client)
+		return false;
+
+	return true;
+}
+
+/* returns the number of bridges attached */
+static int exynos_drm_attach_lcd_bridge(struct drm_device *dev,
+		struct drm_encoder *encoder)
+{
+	struct bridge_init bridge;
+	int ret;
+
+	if (find_bridge("nxp,ptn3460", &bridge)) {
+		ret = ptn3460_init(dev, encoder, bridge.client, bridge.node);
+		if (!ret)
+			return 1;
+	}
+	return 0;
+}
+
+static int exynos_dp_create_connector(struct exynos_drm_display *display,
+				struct drm_encoder *encoder)
+{
+	struct exynos_dp_device *dp = display->ctx;
+	struct drm_connector *connector = &dp->connector;
+	int ret;
+
+	dp->encoder = encoder;
+
+	/* Pre-empt DP connector creation if there's a bridge */
+	ret = exynos_drm_attach_lcd_bridge(dp->drm_dev, encoder);
+	if (ret)
+		return 0;
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	ret = drm_connector_init(dp->drm_dev, connector,
+			&exynos_dp_connector_funcs, DRM_MODE_CONNECTOR_eDP);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	drm_connector_helper_add(connector, &exynos_dp_connector_helper_funcs);
+	drm_sysfs_connector_add(connector);
+	drm_mode_connector_attach_encoder(connector, encoder);
+
+	return 0;
+}
+
+static void exynos_dp_phy_init(struct exynos_dp_device *dp)
+{
+	if (dp->phy) {
+		phy_power_on(dp->phy);
+	} else if (dp->phy_addr) {
+		u32 reg;
+
+		reg = __raw_readl(dp->phy_addr);
+		reg |= dp->enable_mask;
+		__raw_writel(reg, dp->phy_addr);
+	}
+}
+
+static void exynos_dp_phy_exit(struct exynos_dp_device *dp)
+{
+	if (dp->phy) {
+		phy_power_off(dp->phy);
+	} else if (dp->phy_addr) {
+		u32 reg;
+
+		reg = __raw_readl(dp->phy_addr);
+		reg &= ~(dp->enable_mask);
+		__raw_writel(reg, dp->phy_addr);
+	}
+}
+
+static void exynos_dp_poweron(struct exynos_dp_device *dp)
+{
+	if (dp->dpms_mode == DRM_MODE_DPMS_ON)
+		return;
+
+	clk_prepare_enable(dp->clock);
+	exynos_dp_phy_init(dp);
+	exynos_dp_init_dp(dp);
+	enable_irq(dp->irq);
+}
+
+static void exynos_dp_poweroff(struct exynos_dp_device *dp)
+{
+	if (dp->dpms_mode != DRM_MODE_DPMS_ON)
+		return;
+
+	disable_irq(dp->irq);
+	flush_work(&dp->hotplug_work);
+	exynos_dp_phy_exit(dp);
+	clk_disable_unprepare(dp->clock);
+}
+
+static void exynos_dp_dpms(struct exynos_drm_display *display, int mode)
+{
+	struct exynos_dp_device *dp = display->ctx;
+
+	switch (mode) {
+	case DRM_MODE_DPMS_ON:
+		exynos_dp_poweron(dp);
+		break;
+	case DRM_MODE_DPMS_STANDBY:
+	case DRM_MODE_DPMS_SUSPEND:
+	case DRM_MODE_DPMS_OFF:
+		exynos_dp_poweroff(dp);
+		break;
+	default:
+		break;
+	}
+	dp->dpms_mode = mode;
+}
+
+static struct exynos_drm_display_ops exynos_dp_display_ops = {
+	.create_connector = exynos_dp_create_connector,
+	.dpms = exynos_dp_dpms,
+};
+
+static struct exynos_drm_display exynos_dp_display = {
+	.type = EXYNOS_DISPLAY_TYPE_LCD,
+	.ops = &exynos_dp_display_ops,
+};
+
+static struct video_info *exynos_dp_dt_parse_pdata(struct device *dev)
+{
+	struct device_node *dp_node = dev->of_node;
+	struct video_info *dp_video_config;
+
+	dp_video_config = devm_kzalloc(dev,
+				sizeof(*dp_video_config), GFP_KERNEL);
+	if (!dp_video_config)
+		return ERR_PTR(-ENOMEM);
+
+	dp_video_config->h_sync_polarity =
+		of_property_read_bool(dp_node, "hsync-active-high");
+
+	dp_video_config->v_sync_polarity =
+		of_property_read_bool(dp_node, "vsync-active-high");
+
+	dp_video_config->interlaced =
+		of_property_read_bool(dp_node, "interlaced");
+
+	if (of_property_read_u32(dp_node, "samsung,color-space",
+				&dp_video_config->color_space)) {
+		dev_err(dev, "failed to get color-space\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (of_property_read_u32(dp_node, "samsung,dynamic-range",
+				&dp_video_config->dynamic_range)) {
+		dev_err(dev, "failed to get dynamic-range\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (of_property_read_u32(dp_node, "samsung,ycbcr-coeff",
+				&dp_video_config->ycbcr_coeff)) {
+		dev_err(dev, "failed to get ycbcr-coeff\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (of_property_read_u32(dp_node, "samsung,color-depth",
+				&dp_video_config->color_depth)) {
+		dev_err(dev, "failed to get color-depth\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (of_property_read_u32(dp_node, "samsung,link-rate",
+				&dp_video_config->link_rate)) {
+		dev_err(dev, "failed to get link-rate\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (of_property_read_u32(dp_node, "samsung,lane-count",
+				&dp_video_config->lane_count)) {
+		dev_err(dev, "failed to get lane-count\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	return dp_video_config;
+}
+
+static int exynos_dp_dt_parse_phydata(struct exynos_dp_device *dp)
+{
+	struct device_node *dp_phy_node = of_node_get(dp->dev->of_node);
+	u32 phy_base;
+	int ret = 0;
+
+	dp_phy_node = of_find_node_by_name(dp_phy_node, "dptx-phy");
+	if (!dp_phy_node) {
+		dp->phy = devm_phy_get(dp->dev, "dp");
+		return PTR_ERR_OR_ZERO(dp->phy);
+	}
+
+	if (of_property_read_u32(dp_phy_node, "reg", &phy_base)) {
+		dev_err(dp->dev, "failed to get reg for dptx-phy\n");
+		ret = -EINVAL;
+		goto err;
+	}
+
+	if (of_property_read_u32(dp_phy_node, "samsung,enable-mask",
+				&dp->enable_mask)) {
+		dev_err(dp->dev, "failed to get enable-mask for dptx-phy\n");
+		ret = -EINVAL;
+		goto err;
+	}
+
+	dp->phy_addr = ioremap(phy_base, SZ_4);
+	if (!dp->phy_addr) {
+		dev_err(dp->dev, "failed to ioremap dp-phy\n");
+		ret = -ENOMEM;
+		goto err;
+	}
+
+err:
+	of_node_put(dp_phy_node);
+
+	return ret;
+}
+
+static int exynos_dp_dt_parse_panel(struct exynos_dp_device *dp)
+{
+	int ret;
+
+	ret = of_get_videomode(dp->dev->of_node, &dp->panel.vm,
+			OF_USE_NATIVE_MODE);
+	if (ret) {
+		DRM_ERROR("failed: of_get_videomode() : %d\n", ret);
+		return ret;
+	}
+	return 0;
+}
+
+static int exynos_dp_bind(struct device *dev, struct device *master, void *data)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct drm_device *drm_dev = data;
+	struct resource *res;
+	struct exynos_dp_device *dp;
+	unsigned int irq_flags;
+
+	int ret = 0;
+
+	dp = devm_kzalloc(&pdev->dev, sizeof(struct exynos_dp_device),
+				GFP_KERNEL);
+	if (!dp)
+		return -ENOMEM;
+
+	dp->dev = &pdev->dev;
+	dp->dpms_mode = DRM_MODE_DPMS_OFF;
+
+	dp->video_info = exynos_dp_dt_parse_pdata(&pdev->dev);
+	if (IS_ERR(dp->video_info))
+		return PTR_ERR(dp->video_info);
+
+	ret = exynos_dp_dt_parse_phydata(dp);
+	if (ret)
+		return ret;
+
+	ret = exynos_dp_dt_parse_panel(dp);
+	if (ret)
+		return ret;
+
+	dp->clock = devm_clk_get(&pdev->dev, "dp");
+	if (IS_ERR(dp->clock)) {
+		dev_err(&pdev->dev, "failed to get clock\n");
+		return PTR_ERR(dp->clock);
+	}
+
+	clk_prepare_enable(dp->clock);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+
+	dp->reg_base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(dp->reg_base))
+		return PTR_ERR(dp->reg_base);
+
+	dp->hpd_gpio = of_get_named_gpio(dev->of_node, "samsung,hpd-gpio", 0);
+
+	if (gpio_is_valid(dp->hpd_gpio)) {
+		/*
+		 * Set up the hotplug GPIO from the device tree as an interrupt.
+		 * Simply specifying a different interrupt in the device tree
+		 * doesn't work since we handle hotplug rather differently when
+		 * using a GPIO.  We also need the actual GPIO specifier so
+		 * that we can get the current state of the GPIO.
+		 */
+		ret = devm_gpio_request_one(&pdev->dev, dp->hpd_gpio, GPIOF_IN,
+					    "hpd_gpio");
+		if (ret) {
+			dev_err(&pdev->dev, "failed to get hpd gpio\n");
+			return ret;
+		}
+		dp->irq = gpio_to_irq(dp->hpd_gpio);
+		irq_flags = IRQF_TRIGGER_RISING | IRQF_TRIGGER_FALLING;
+	} else {
+		dp->hpd_gpio = -ENODEV;
+		dp->irq = platform_get_irq(pdev, 0);
+		irq_flags = 0;
+	}
+
+	if (dp->irq == -ENXIO) {
+		dev_err(&pdev->dev, "failed to get irq\n");
+		return -ENODEV;
+	}
+
+	INIT_WORK(&dp->hotplug_work, exynos_dp_hotplug);
+
+	exynos_dp_phy_init(dp);
+
+	exynos_dp_init_dp(dp);
+
+	ret = devm_request_irq(&pdev->dev, dp->irq, exynos_dp_irq_handler,
+			irq_flags, "exynos-dp", dp);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to request irq\n");
+		return ret;
+	}
+	disable_irq(dp->irq);
+
+	dp->drm_dev = drm_dev;
+	exynos_dp_display.ctx = dp;
+
+	platform_set_drvdata(pdev, &exynos_dp_display);
+
+	return exynos_drm_create_enc_conn(drm_dev, &exynos_dp_display);
+}
+
+static void exynos_dp_unbind(struct device *dev, struct device *master,
+				void *data)
+{
+	struct exynos_drm_display *display = dev_get_drvdata(dev);
+	struct exynos_dp_device *dp = display->ctx;
+	struct drm_encoder *encoder = dp->encoder;
+
+	exynos_dp_dpms(display, DRM_MODE_DPMS_OFF);
+
+	encoder->funcs->destroy(encoder);
+	drm_connector_cleanup(&dp->connector);
+}
+
+static const struct component_ops exynos_dp_ops = {
+	.bind	= exynos_dp_bind,
+	.unbind	= exynos_dp_unbind,
+};
+
+static int exynos_dp_probe(struct platform_device *pdev)
+{
+	int ret;
+
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
+					exynos_dp_display.type);
+	if (ret)
+		return ret;
+
+	ret = component_add(&pdev->dev, &exynos_dp_ops);
+	if (ret)
+		exynos_drm_component_del(&pdev->dev,
+						EXYNOS_DEVICE_TYPE_CONNECTOR);
+
+	return ret;
+}
+
+static int exynos_dp_remove(struct platform_device *pdev)
+{
+	component_del(&pdev->dev, &exynos_dp_ops);
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM_SLEEP
+static int exynos_dp_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct exynos_drm_display *display = platform_get_drvdata(pdev);
+
+	exynos_dp_dpms(display, DRM_MODE_DPMS_OFF);
+	return 0;
+}
+
+static int exynos_dp_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct exynos_drm_display *display = platform_get_drvdata(pdev);
+
+	exynos_dp_dpms(display, DRM_MODE_DPMS_ON);
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops exynos_dp_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(exynos_dp_suspend, exynos_dp_resume)
+};
+
+static const struct of_device_id exynos_dp_match[] = {
+	{ .compatible = "samsung,exynos5-dp" },
+	{},
+};
+
+struct platform_driver dp_driver = {
+	.probe		= exynos_dp_probe,
+	.remove		= exynos_dp_remove,
+	.driver		= {
+		.name	= "exynos-dp",
+		.owner	= THIS_MODULE,
+		.pm	= &exynos_dp_pm_ops,
+		.of_match_table = exynos_dp_match,
+	},
+};
+
+MODULE_AUTHOR("Jingoo Han <jg1.han@samsung.com>");
+MODULE_DESCRIPTION("Samsung SoC DP Driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/exynos/exynos_dp_core.h b/drivers/gpu/drm/exynos/exynos_dp_core.h
new file mode 100644
index 00000000000..02cc4f9ab90
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_dp_core.h
@@ -0,0 +1,279 @@
+/*
+ * Header file for Samsung DP (Display Port) interface driver.
+ *
+ * Copyright (C) 2012 Samsung Electronics Co., Ltd.
+ * Author: Jingoo Han <jg1.han@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ */
+
+#ifndef _EXYNOS_DP_CORE_H
+#define _EXYNOS_DP_CORE_H
+
+#include <drm/drm_crtc.h>
+#include <drm/drm_dp_helper.h>
+#include <drm/exynos_drm.h>
+
+#define DP_TIMEOUT_LOOP_COUNT 100
+#define MAX_CR_LOOP 5
+#define MAX_EQ_LOOP 5
+
+enum link_rate_type {
+	LINK_RATE_1_62GBPS = 0x06,
+	LINK_RATE_2_70GBPS = 0x0a
+};
+
+enum link_lane_count_type {
+	LANE_COUNT1 = 1,
+	LANE_COUNT2 = 2,
+	LANE_COUNT4 = 4
+};
+
+enum link_training_state {
+	START,
+	CLOCK_RECOVERY,
+	EQUALIZER_TRAINING,
+	FINISHED,
+	FAILED
+};
+
+enum voltage_swing_level {
+	VOLTAGE_LEVEL_0,
+	VOLTAGE_LEVEL_1,
+	VOLTAGE_LEVEL_2,
+	VOLTAGE_LEVEL_3,
+};
+
+enum pre_emphasis_level {
+	PRE_EMPHASIS_LEVEL_0,
+	PRE_EMPHASIS_LEVEL_1,
+	PRE_EMPHASIS_LEVEL_2,
+	PRE_EMPHASIS_LEVEL_3,
+};
+
+enum pattern_set {
+	PRBS7,
+	D10_2,
+	TRAINING_PTN1,
+	TRAINING_PTN2,
+	DP_NONE
+};
+
+enum color_space {
+	COLOR_RGB,
+	COLOR_YCBCR422,
+	COLOR_YCBCR444
+};
+
+enum color_depth {
+	COLOR_6,
+	COLOR_8,
+	COLOR_10,
+	COLOR_12
+};
+
+enum color_coefficient {
+	COLOR_YCBCR601,
+	COLOR_YCBCR709
+};
+
+enum dynamic_range {
+	VESA,
+	CEA
+};
+
+enum pll_status {
+	PLL_UNLOCKED,
+	PLL_LOCKED
+};
+
+enum clock_recovery_m_value_type {
+	CALCULATED_M,
+	REGISTER_M
+};
+
+enum video_timing_recognition_type {
+	VIDEO_TIMING_FROM_CAPTURE,
+	VIDEO_TIMING_FROM_REGISTER
+};
+
+enum analog_power_block {
+	AUX_BLOCK,
+	CH0_BLOCK,
+	CH1_BLOCK,
+	CH2_BLOCK,
+	CH3_BLOCK,
+	ANALOG_TOTAL,
+	POWER_ALL
+};
+
+enum dp_irq_type {
+	DP_IRQ_TYPE_HP_CABLE_IN,
+	DP_IRQ_TYPE_HP_CABLE_OUT,
+	DP_IRQ_TYPE_HP_CHANGE,
+	DP_IRQ_TYPE_UNKNOWN,
+};
+
+struct video_info {
+	char *name;
+
+	bool h_sync_polarity;
+	bool v_sync_polarity;
+	bool interlaced;
+
+	enum color_space color_space;
+	enum dynamic_range dynamic_range;
+	enum color_coefficient ycbcr_coeff;
+	enum color_depth color_depth;
+
+	enum link_rate_type link_rate;
+	enum link_lane_count_type lane_count;
+};
+
+struct link_train {
+	int eq_loop;
+	int cr_loop[4];
+
+	u8 link_rate;
+	u8 lane_count;
+	u8 training_lane[4];
+
+	enum link_training_state lt_state;
+};
+
+struct exynos_dp_device {
+	struct device		*dev;
+	struct drm_device	*drm_dev;
+	struct drm_connector	connector;
+	struct drm_encoder	*encoder;
+	struct clk		*clock;
+	unsigned int		irq;
+	void __iomem		*reg_base;
+	void __iomem		*phy_addr;
+	unsigned int		enable_mask;
+
+	struct video_info	*video_info;
+	struct link_train	link_train;
+	struct work_struct	hotplug_work;
+	struct phy		*phy;
+	int			dpms_mode;
+	int			hpd_gpio;
+
+	struct exynos_drm_panel_info panel;
+};
+
+/* exynos_dp_reg.c */
+void exynos_dp_enable_video_mute(struct exynos_dp_device *dp, bool enable);
+void exynos_dp_stop_video(struct exynos_dp_device *dp);
+void exynos_dp_lane_swap(struct exynos_dp_device *dp, bool enable);
+void exynos_dp_init_analog_param(struct exynos_dp_device *dp);
+void exynos_dp_init_interrupt(struct exynos_dp_device *dp);
+void exynos_dp_reset(struct exynos_dp_device *dp);
+void exynos_dp_swreset(struct exynos_dp_device *dp);
+void exynos_dp_config_interrupt(struct exynos_dp_device *dp);
+enum pll_status exynos_dp_get_pll_lock_status(struct exynos_dp_device *dp);
+void exynos_dp_set_pll_power_down(struct exynos_dp_device *dp, bool enable);
+void exynos_dp_set_analog_power_down(struct exynos_dp_device *dp,
+				enum analog_power_block block,
+				bool enable);
+void exynos_dp_init_analog_func(struct exynos_dp_device *dp);
+void exynos_dp_init_hpd(struct exynos_dp_device *dp);
+enum dp_irq_type exynos_dp_get_irq_type(struct exynos_dp_device *dp);
+void exynos_dp_clear_hotplug_interrupts(struct exynos_dp_device *dp);
+void exynos_dp_reset_aux(struct exynos_dp_device *dp);
+void exynos_dp_init_aux(struct exynos_dp_device *dp);
+int exynos_dp_get_plug_in_status(struct exynos_dp_device *dp);
+void exynos_dp_enable_sw_function(struct exynos_dp_device *dp);
+int exynos_dp_start_aux_transaction(struct exynos_dp_device *dp);
+int exynos_dp_write_byte_to_dpcd(struct exynos_dp_device *dp,
+				unsigned int reg_addr,
+				unsigned char data);
+int exynos_dp_read_byte_from_dpcd(struct exynos_dp_device *dp,
+				unsigned int reg_addr,
+				unsigned char *data);
+int exynos_dp_write_bytes_to_dpcd(struct exynos_dp_device *dp,
+				unsigned int reg_addr,
+				unsigned int count,
+				unsigned char data[]);
+int exynos_dp_read_bytes_from_dpcd(struct exynos_dp_device *dp,
+				unsigned int reg_addr,
+				unsigned int count,
+				unsigned char data[]);
+int exynos_dp_select_i2c_device(struct exynos_dp_device *dp,
+				unsigned int device_addr,
+				unsigned int reg_addr);
+int exynos_dp_read_byte_from_i2c(struct exynos_dp_device *dp,
+				unsigned int device_addr,
+				unsigned int reg_addr,
+				unsigned int *data);
+int exynos_dp_read_bytes_from_i2c(struct exynos_dp_device *dp,
+				unsigned int device_addr,
+				unsigned int reg_addr,
+				unsigned int count,
+				unsigned char edid[]);
+void exynos_dp_set_link_bandwidth(struct exynos_dp_device *dp, u32 bwtype);
+void exynos_dp_get_link_bandwidth(struct exynos_dp_device *dp, u32 *bwtype);
+void exynos_dp_set_lane_count(struct exynos_dp_device *dp, u32 count);
+void exynos_dp_get_lane_count(struct exynos_dp_device *dp, u32 *count);
+void exynos_dp_enable_enhanced_mode(struct exynos_dp_device *dp, bool enable);
+void exynos_dp_set_training_pattern(struct exynos_dp_device *dp,
+				 enum pattern_set pattern);
+void exynos_dp_set_lane0_pre_emphasis(struct exynos_dp_device *dp, u32 level);
+void exynos_dp_set_lane1_pre_emphasis(struct exynos_dp_device *dp, u32 level);
+void exynos_dp_set_lane2_pre_emphasis(struct exynos_dp_device *dp, u32 level);
+void exynos_dp_set_lane3_pre_emphasis(struct exynos_dp_device *dp, u32 level);
+void exynos_dp_set_lane0_link_training(struct exynos_dp_device *dp,
+				u32 training_lane);
+void exynos_dp_set_lane1_link_training(struct exynos_dp_device *dp,
+				u32 training_lane);
+void exynos_dp_set_lane2_link_training(struct exynos_dp_device *dp,
+				u32 training_lane);
+void exynos_dp_set_lane3_link_training(struct exynos_dp_device *dp,
+				u32 training_lane);
+u32 exynos_dp_get_lane0_link_training(struct exynos_dp_device *dp);
+u32 exynos_dp_get_lane1_link_training(struct exynos_dp_device *dp);
+u32 exynos_dp_get_lane2_link_training(struct exynos_dp_device *dp);
+u32 exynos_dp_get_lane3_link_training(struct exynos_dp_device *dp);
+void exynos_dp_reset_macro(struct exynos_dp_device *dp);
+void exynos_dp_init_video(struct exynos_dp_device *dp);
+
+void exynos_dp_set_video_color_format(struct exynos_dp_device *dp);
+int exynos_dp_is_slave_video_stream_clock_on(struct exynos_dp_device *dp);
+void exynos_dp_set_video_cr_mn(struct exynos_dp_device *dp,
+			enum clock_recovery_m_value_type type,
+			u32 m_value,
+			u32 n_value);
+void exynos_dp_set_video_timing_mode(struct exynos_dp_device *dp, u32 type);
+void exynos_dp_enable_video_master(struct exynos_dp_device *dp, bool enable);
+void exynos_dp_start_video(struct exynos_dp_device *dp);
+int exynos_dp_is_video_stream_on(struct exynos_dp_device *dp);
+void exynos_dp_config_video_slave_mode(struct exynos_dp_device *dp);
+void exynos_dp_enable_scrambling(struct exynos_dp_device *dp);
+void exynos_dp_disable_scrambling(struct exynos_dp_device *dp);
+
+/* I2C EDID Chip ID, Slave Address */
+#define I2C_EDID_DEVICE_ADDR			0x50
+#define I2C_E_EDID_DEVICE_ADDR			0x30
+
+#define EDID_BLOCK_LENGTH			0x80
+#define EDID_HEADER_PATTERN			0x00
+#define EDID_EXTENSION_FLAG			0x7e
+#define EDID_CHECKSUM				0x7f
+
+/* DP_MAX_LANE_COUNT */
+#define DPCD_ENHANCED_FRAME_CAP(x)		(((x) >> 7) & 0x1)
+#define DPCD_MAX_LANE_COUNT(x)			((x) & 0x1f)
+
+/* DP_LANE_COUNT_SET */
+#define DPCD_LANE_COUNT_SET(x)			((x) & 0x1f)
+
+/* DP_TRAINING_LANE0_SET */
+#define DPCD_PRE_EMPHASIS_SET(x)		(((x) & 0x3) << 3)
+#define DPCD_PRE_EMPHASIS_GET(x)		(((x) >> 3) & 0x3)
+#define DPCD_VOLTAGE_SWING_SET(x)		(((x) & 0x3) << 0)
+#define DPCD_VOLTAGE_SWING_GET(x)		(((x) >> 0) & 0x3)
+
+#endif /* _EXYNOS_DP_CORE_H */
diff --git a/drivers/gpu/drm/exynos/exynos_dp_reg.c b/drivers/gpu/drm/exynos/exynos_dp_reg.c
new file mode 100644
index 00000000000..c1f87a2a928
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_dp_reg.c
@@ -0,0 +1,1263 @@
+/*
+ * Samsung DP (Display port) register interface driver.
+ *
+ * Copyright (C) 2012 Samsung Electronics Co., Ltd.
+ * Author: Jingoo Han <jg1.han@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ */
+
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/delay.h>
+#include <linux/gpio.h>
+
+#include "exynos_dp_core.h"
+#include "exynos_dp_reg.h"
+
+#define COMMON_INT_MASK_1	0
+#define COMMON_INT_MASK_2	0
+#define COMMON_INT_MASK_3	0
+#define COMMON_INT_MASK_4	(HOTPLUG_CHG | HPD_LOST | PLUG)
+#define INT_STA_MASK		INT_HPD
+
+void exynos_dp_enable_video_mute(struct exynos_dp_device *dp, bool enable)
+{
+	u32 reg;
+
+	if (enable) {
+		reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_1);
+		reg |= HDCP_VIDEO_MUTE;
+		writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_1);
+	} else {
+		reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_1);
+		reg &= ~HDCP_VIDEO_MUTE;
+		writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_1);
+	}
+}
+
+void exynos_dp_stop_video(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_1);
+	reg &= ~VIDEO_EN;
+	writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_1);
+}
+
+void exynos_dp_lane_swap(struct exynos_dp_device *dp, bool enable)
+{
+	u32 reg;
+
+	if (enable)
+		reg = LANE3_MAP_LOGIC_LANE_0 | LANE2_MAP_LOGIC_LANE_1 |
+			LANE1_MAP_LOGIC_LANE_2 | LANE0_MAP_LOGIC_LANE_3;
+	else
+		reg = LANE3_MAP_LOGIC_LANE_3 | LANE2_MAP_LOGIC_LANE_2 |
+			LANE1_MAP_LOGIC_LANE_1 | LANE0_MAP_LOGIC_LANE_0;
+
+	writel(reg, dp->reg_base + EXYNOS_DP_LANE_MAP);
+}
+
+void exynos_dp_init_analog_param(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = TX_TERMINAL_CTRL_50_OHM;
+	writel(reg, dp->reg_base + EXYNOS_DP_ANALOG_CTL_1);
+
+	reg = SEL_24M | TX_DVDD_BIT_1_0625V;
+	writel(reg, dp->reg_base + EXYNOS_DP_ANALOG_CTL_2);
+
+	reg = DRIVE_DVDD_BIT_1_0625V | VCO_BIT_600_MICRO;
+	writel(reg, dp->reg_base + EXYNOS_DP_ANALOG_CTL_3);
+
+	reg = PD_RING_OSC | AUX_TERMINAL_CTRL_50_OHM |
+		TX_CUR1_2X | TX_CUR_16_MA;
+	writel(reg, dp->reg_base + EXYNOS_DP_PLL_FILTER_CTL_1);
+
+	reg = CH3_AMP_400_MV | CH2_AMP_400_MV |
+		CH1_AMP_400_MV | CH0_AMP_400_MV;
+	writel(reg, dp->reg_base + EXYNOS_DP_TX_AMP_TUNING_CTL);
+}
+
+void exynos_dp_init_interrupt(struct exynos_dp_device *dp)
+{
+	/* Set interrupt pin assertion polarity as high */
+	writel(INT_POL1 | INT_POL0, dp->reg_base + EXYNOS_DP_INT_CTL);
+
+	/* Clear pending regisers */
+	writel(0xff, dp->reg_base + EXYNOS_DP_COMMON_INT_STA_1);
+	writel(0x4f, dp->reg_base + EXYNOS_DP_COMMON_INT_STA_2);
+	writel(0xe0, dp->reg_base + EXYNOS_DP_COMMON_INT_STA_3);
+	writel(0xe7, dp->reg_base + EXYNOS_DP_COMMON_INT_STA_4);
+	writel(0x63, dp->reg_base + EXYNOS_DP_INT_STA);
+
+	/* 0:mask,1: unmask */
+	writel(0x00, dp->reg_base + EXYNOS_DP_COMMON_INT_MASK_1);
+	writel(0x00, dp->reg_base + EXYNOS_DP_COMMON_INT_MASK_2);
+	writel(0x00, dp->reg_base + EXYNOS_DP_COMMON_INT_MASK_3);
+	writel(0x00, dp->reg_base + EXYNOS_DP_COMMON_INT_MASK_4);
+	writel(0x00, dp->reg_base + EXYNOS_DP_INT_STA_MASK);
+}
+
+void exynos_dp_reset(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	exynos_dp_stop_video(dp);
+	exynos_dp_enable_video_mute(dp, 0);
+
+	reg = MASTER_VID_FUNC_EN_N | SLAVE_VID_FUNC_EN_N |
+		AUD_FIFO_FUNC_EN_N | AUD_FUNC_EN_N |
+		HDCP_FUNC_EN_N | SW_FUNC_EN_N;
+	writel(reg, dp->reg_base + EXYNOS_DP_FUNC_EN_1);
+
+	reg = SSC_FUNC_EN_N | AUX_FUNC_EN_N |
+		SERDES_FIFO_FUNC_EN_N |
+		LS_CLK_DOMAIN_FUNC_EN_N;
+	writel(reg, dp->reg_base + EXYNOS_DP_FUNC_EN_2);
+
+	usleep_range(20, 30);
+
+	exynos_dp_lane_swap(dp, 0);
+
+	writel(0x0, dp->reg_base + EXYNOS_DP_SYS_CTL_1);
+	writel(0x40, dp->reg_base + EXYNOS_DP_SYS_CTL_2);
+	writel(0x0, dp->reg_base + EXYNOS_DP_SYS_CTL_3);
+	writel(0x0, dp->reg_base + EXYNOS_DP_SYS_CTL_4);
+
+	writel(0x0, dp->reg_base + EXYNOS_DP_PKT_SEND_CTL);
+	writel(0x0, dp->reg_base + EXYNOS_DP_HDCP_CTL);
+
+	writel(0x5e, dp->reg_base + EXYNOS_DP_HPD_DEGLITCH_L);
+	writel(0x1a, dp->reg_base + EXYNOS_DP_HPD_DEGLITCH_H);
+
+	writel(0x10, dp->reg_base + EXYNOS_DP_LINK_DEBUG_CTL);
+
+	writel(0x0, dp->reg_base + EXYNOS_DP_PHY_TEST);
+
+	writel(0x0, dp->reg_base + EXYNOS_DP_VIDEO_FIFO_THRD);
+	writel(0x20, dp->reg_base + EXYNOS_DP_AUDIO_MARGIN);
+
+	writel(0x4, dp->reg_base + EXYNOS_DP_M_VID_GEN_FILTER_TH);
+	writel(0x2, dp->reg_base + EXYNOS_DP_M_AUD_GEN_FILTER_TH);
+
+	writel(0x00000101, dp->reg_base + EXYNOS_DP_SOC_GENERAL_CTL);
+}
+
+void exynos_dp_swreset(struct exynos_dp_device *dp)
+{
+	writel(RESET_DP_TX, dp->reg_base + EXYNOS_DP_TX_SW_RESET);
+}
+
+void exynos_dp_config_interrupt(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	/* 0: mask, 1: unmask */
+	reg = COMMON_INT_MASK_1;
+	writel(reg, dp->reg_base + EXYNOS_DP_COMMON_INT_MASK_1);
+
+	reg = COMMON_INT_MASK_2;
+	writel(reg, dp->reg_base + EXYNOS_DP_COMMON_INT_MASK_2);
+
+	reg = COMMON_INT_MASK_3;
+	writel(reg, dp->reg_base + EXYNOS_DP_COMMON_INT_MASK_3);
+
+	reg = COMMON_INT_MASK_4;
+	writel(reg, dp->reg_base + EXYNOS_DP_COMMON_INT_MASK_4);
+
+	reg = INT_STA_MASK;
+	writel(reg, dp->reg_base + EXYNOS_DP_INT_STA_MASK);
+}
+
+enum pll_status exynos_dp_get_pll_lock_status(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_DEBUG_CTL);
+	if (reg & PLL_LOCK)
+		return PLL_LOCKED;
+	else
+		return PLL_UNLOCKED;
+}
+
+void exynos_dp_set_pll_power_down(struct exynos_dp_device *dp, bool enable)
+{
+	u32 reg;
+
+	if (enable) {
+		reg = readl(dp->reg_base + EXYNOS_DP_PLL_CTL);
+		reg |= DP_PLL_PD;
+		writel(reg, dp->reg_base + EXYNOS_DP_PLL_CTL);
+	} else {
+		reg = readl(dp->reg_base + EXYNOS_DP_PLL_CTL);
+		reg &= ~DP_PLL_PD;
+		writel(reg, dp->reg_base + EXYNOS_DP_PLL_CTL);
+	}
+}
+
+void exynos_dp_set_analog_power_down(struct exynos_dp_device *dp,
+				enum analog_power_block block,
+				bool enable)
+{
+	u32 reg;
+
+	switch (block) {
+	case AUX_BLOCK:
+		if (enable) {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg |= AUX_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		} else {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg &= ~AUX_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		}
+		break;
+	case CH0_BLOCK:
+		if (enable) {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg |= CH0_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		} else {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg &= ~CH0_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		}
+		break;
+	case CH1_BLOCK:
+		if (enable) {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg |= CH1_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		} else {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg &= ~CH1_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		}
+		break;
+	case CH2_BLOCK:
+		if (enable) {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg |= CH2_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		} else {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg &= ~CH2_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		}
+		break;
+	case CH3_BLOCK:
+		if (enable) {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg |= CH3_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		} else {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg &= ~CH3_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		}
+		break;
+	case ANALOG_TOTAL:
+		if (enable) {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg |= DP_PHY_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		} else {
+			reg = readl(dp->reg_base + EXYNOS_DP_PHY_PD);
+			reg &= ~DP_PHY_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		}
+		break;
+	case POWER_ALL:
+		if (enable) {
+			reg = DP_PHY_PD | AUX_PD | CH3_PD | CH2_PD |
+				CH1_PD | CH0_PD;
+			writel(reg, dp->reg_base + EXYNOS_DP_PHY_PD);
+		} else {
+			writel(0x00, dp->reg_base + EXYNOS_DP_PHY_PD);
+		}
+		break;
+	default:
+		break;
+	}
+}
+
+void exynos_dp_init_analog_func(struct exynos_dp_device *dp)
+{
+	u32 reg;
+	int timeout_loop = 0;
+
+	exynos_dp_set_analog_power_down(dp, POWER_ALL, 0);
+
+	reg = PLL_LOCK_CHG;
+	writel(reg, dp->reg_base + EXYNOS_DP_COMMON_INT_STA_1);
+
+	reg = readl(dp->reg_base + EXYNOS_DP_DEBUG_CTL);
+	reg &= ~(F_PLL_LOCK | PLL_LOCK_CTRL);
+	writel(reg, dp->reg_base + EXYNOS_DP_DEBUG_CTL);
+
+	/* Power up PLL */
+	if (exynos_dp_get_pll_lock_status(dp) == PLL_UNLOCKED) {
+		exynos_dp_set_pll_power_down(dp, 0);
+
+		while (exynos_dp_get_pll_lock_status(dp) == PLL_UNLOCKED) {
+			timeout_loop++;
+			if (DP_TIMEOUT_LOOP_COUNT < timeout_loop) {
+				dev_err(dp->dev, "failed to get pll lock status\n");
+				return;
+			}
+			usleep_range(10, 20);
+		}
+	}
+
+	/* Enable Serdes FIFO function and Link symbol clock domain module */
+	reg = readl(dp->reg_base + EXYNOS_DP_FUNC_EN_2);
+	reg &= ~(SERDES_FIFO_FUNC_EN_N | LS_CLK_DOMAIN_FUNC_EN_N
+		| AUX_FUNC_EN_N);
+	writel(reg, dp->reg_base + EXYNOS_DP_FUNC_EN_2);
+}
+
+void exynos_dp_clear_hotplug_interrupts(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	if (gpio_is_valid(dp->hpd_gpio))
+		return;
+
+	reg = HOTPLUG_CHG | HPD_LOST | PLUG;
+	writel(reg, dp->reg_base + EXYNOS_DP_COMMON_INT_STA_4);
+
+	reg = INT_HPD;
+	writel(reg, dp->reg_base + EXYNOS_DP_INT_STA);
+}
+
+void exynos_dp_init_hpd(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	if (gpio_is_valid(dp->hpd_gpio))
+		return;
+
+	exynos_dp_clear_hotplug_interrupts(dp);
+
+	reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_3);
+	reg &= ~(F_HPD | HPD_CTRL);
+	writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_3);
+}
+
+enum dp_irq_type exynos_dp_get_irq_type(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	if (gpio_is_valid(dp->hpd_gpio)) {
+		reg = gpio_get_value(dp->hpd_gpio);
+		if (reg)
+			return DP_IRQ_TYPE_HP_CABLE_IN;
+		else
+			return DP_IRQ_TYPE_HP_CABLE_OUT;
+	} else {
+		/* Parse hotplug interrupt status register */
+		reg = readl(dp->reg_base + EXYNOS_DP_COMMON_INT_STA_4);
+
+		if (reg & PLUG)
+			return DP_IRQ_TYPE_HP_CABLE_IN;
+
+		if (reg & HPD_LOST)
+			return DP_IRQ_TYPE_HP_CABLE_OUT;
+
+		if (reg & HOTPLUG_CHG)
+			return DP_IRQ_TYPE_HP_CHANGE;
+
+		return DP_IRQ_TYPE_UNKNOWN;
+	}
+}
+
+void exynos_dp_reset_aux(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	/* Disable AUX channel module */
+	reg = readl(dp->reg_base + EXYNOS_DP_FUNC_EN_2);
+	reg |= AUX_FUNC_EN_N;
+	writel(reg, dp->reg_base + EXYNOS_DP_FUNC_EN_2);
+}
+
+void exynos_dp_init_aux(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	/* Clear inerrupts related to AUX channel */
+	reg = RPLY_RECEIV | AUX_ERR;
+	writel(reg, dp->reg_base + EXYNOS_DP_INT_STA);
+
+	exynos_dp_reset_aux(dp);
+
+	/* Disable AUX transaction H/W retry */
+	reg = AUX_BIT_PERIOD_EXPECTED_DELAY(3) | AUX_HW_RETRY_COUNT_SEL(0)|
+		AUX_HW_RETRY_INTERVAL_600_MICROSECONDS;
+	writel(reg, dp->reg_base + EXYNOS_DP_AUX_HW_RETRY_CTL);
+
+	/* Receive AUX Channel DEFER commands equal to DEFFER_COUNT*64 */
+	reg = DEFER_CTRL_EN | DEFER_COUNT(1);
+	writel(reg, dp->reg_base + EXYNOS_DP_AUX_CH_DEFER_CTL);
+
+	/* Enable AUX channel module */
+	reg = readl(dp->reg_base + EXYNOS_DP_FUNC_EN_2);
+	reg &= ~AUX_FUNC_EN_N;
+	writel(reg, dp->reg_base + EXYNOS_DP_FUNC_EN_2);
+}
+
+int exynos_dp_get_plug_in_status(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	if (gpio_is_valid(dp->hpd_gpio)) {
+		if (gpio_get_value(dp->hpd_gpio))
+			return 0;
+	} else {
+		reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_3);
+		if (reg & HPD_STATUS)
+			return 0;
+	}
+
+	return -EINVAL;
+}
+
+void exynos_dp_enable_sw_function(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_FUNC_EN_1);
+	reg &= ~SW_FUNC_EN_N;
+	writel(reg, dp->reg_base + EXYNOS_DP_FUNC_EN_1);
+}
+
+int exynos_dp_start_aux_transaction(struct exynos_dp_device *dp)
+{
+	int reg;
+	int retval = 0;
+	int timeout_loop = 0;
+
+	/* Enable AUX CH operation */
+	reg = readl(dp->reg_base + EXYNOS_DP_AUX_CH_CTL_2);
+	reg |= AUX_EN;
+	writel(reg, dp->reg_base + EXYNOS_DP_AUX_CH_CTL_2);
+
+	/* Is AUX CH command reply received? */
+	reg = readl(dp->reg_base + EXYNOS_DP_INT_STA);
+	while (!(reg & RPLY_RECEIV)) {
+		timeout_loop++;
+		if (DP_TIMEOUT_LOOP_COUNT < timeout_loop) {
+			dev_err(dp->dev, "AUX CH command reply failed!\n");
+			return -ETIMEDOUT;
+		}
+		reg = readl(dp->reg_base + EXYNOS_DP_INT_STA);
+		usleep_range(10, 11);
+	}
+
+	/* Clear interrupt source for AUX CH command reply */
+	writel(RPLY_RECEIV, dp->reg_base + EXYNOS_DP_INT_STA);
+
+	/* Clear interrupt source for AUX CH access error */
+	reg = readl(dp->reg_base + EXYNOS_DP_INT_STA);
+	if (reg & AUX_ERR) {
+		writel(AUX_ERR, dp->reg_base + EXYNOS_DP_INT_STA);
+		return -EREMOTEIO;
+	}
+
+	/* Check AUX CH error access status */
+	reg = readl(dp->reg_base + EXYNOS_DP_AUX_CH_STA);
+	if ((reg & AUX_STATUS_MASK) != 0) {
+		dev_err(dp->dev, "AUX CH error happens: %d\n\n",
+			reg & AUX_STATUS_MASK);
+		return -EREMOTEIO;
+	}
+
+	return retval;
+}
+
+int exynos_dp_write_byte_to_dpcd(struct exynos_dp_device *dp,
+				unsigned int reg_addr,
+				unsigned char data)
+{
+	u32 reg;
+	int i;
+	int retval;
+
+	for (i = 0; i < 3; i++) {
+		/* Clear AUX CH data buffer */
+		reg = BUF_CLR;
+		writel(reg, dp->reg_base + EXYNOS_DP_BUFFER_DATA_CTL);
+
+		/* Select DPCD device address */
+		reg = AUX_ADDR_7_0(reg_addr);
+		writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_7_0);
+		reg = AUX_ADDR_15_8(reg_addr);
+		writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_15_8);
+		reg = AUX_ADDR_19_16(reg_addr);
+		writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_19_16);
+
+		/* Write data buffer */
+		reg = (unsigned int)data;
+		writel(reg, dp->reg_base + EXYNOS_DP_BUF_DATA_0);
+
+		/*
+		 * Set DisplayPort transaction and write 1 byte
+		 * If bit 3 is 1, DisplayPort transaction.
+		 * If Bit 3 is 0, I2C transaction.
+		 */
+		reg = AUX_TX_COMM_DP_TRANSACTION | AUX_TX_COMM_WRITE;
+		writel(reg, dp->reg_base + EXYNOS_DP_AUX_CH_CTL_1);
+
+		/* Start AUX transaction */
+		retval = exynos_dp_start_aux_transaction(dp);
+		if (retval == 0)
+			break;
+		else
+			dev_dbg(dp->dev, "%s: Aux Transaction fail!\n",
+				__func__);
+	}
+
+	return retval;
+}
+
+int exynos_dp_read_byte_from_dpcd(struct exynos_dp_device *dp,
+				unsigned int reg_addr,
+				unsigned char *data)
+{
+	u32 reg;
+	int i;
+	int retval;
+
+	for (i = 0; i < 3; i++) {
+		/* Clear AUX CH data buffer */
+		reg = BUF_CLR;
+		writel(reg, dp->reg_base + EXYNOS_DP_BUFFER_DATA_CTL);
+
+		/* Select DPCD device address */
+		reg = AUX_ADDR_7_0(reg_addr);
+		writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_7_0);
+		reg = AUX_ADDR_15_8(reg_addr);
+		writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_15_8);
+		reg = AUX_ADDR_19_16(reg_addr);
+		writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_19_16);
+
+		/*
+		 * Set DisplayPort transaction and read 1 byte
+		 * If bit 3 is 1, DisplayPort transaction.
+		 * If Bit 3 is 0, I2C transaction.
+		 */
+		reg = AUX_TX_COMM_DP_TRANSACTION | AUX_TX_COMM_READ;
+		writel(reg, dp->reg_base + EXYNOS_DP_AUX_CH_CTL_1);
+
+		/* Start AUX transaction */
+		retval = exynos_dp_start_aux_transaction(dp);
+		if (retval == 0)
+			break;
+		else
+			dev_dbg(dp->dev, "%s: Aux Transaction fail!\n",
+				__func__);
+	}
+
+	/* Read data buffer */
+	reg = readl(dp->reg_base + EXYNOS_DP_BUF_DATA_0);
+	*data = (unsigned char)(reg & 0xff);
+
+	return retval;
+}
+
+int exynos_dp_write_bytes_to_dpcd(struct exynos_dp_device *dp,
+				unsigned int reg_addr,
+				unsigned int count,
+				unsigned char data[])
+{
+	u32 reg;
+	unsigned int start_offset;
+	unsigned int cur_data_count;
+	unsigned int cur_data_idx;
+	int i;
+	int retval = 0;
+
+	/* Clear AUX CH data buffer */
+	reg = BUF_CLR;
+	writel(reg, dp->reg_base + EXYNOS_DP_BUFFER_DATA_CTL);
+
+	start_offset = 0;
+	while (start_offset < count) {
+		/* Buffer size of AUX CH is 16 * 4bytes */
+		if ((count - start_offset) > 16)
+			cur_data_count = 16;
+		else
+			cur_data_count = count - start_offset;
+
+		for (i = 0; i < 3; i++) {
+			/* Select DPCD device address */
+			reg = AUX_ADDR_7_0(reg_addr + start_offset);
+			writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_7_0);
+			reg = AUX_ADDR_15_8(reg_addr + start_offset);
+			writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_15_8);
+			reg = AUX_ADDR_19_16(reg_addr + start_offset);
+			writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_19_16);
+
+			for (cur_data_idx = 0; cur_data_idx < cur_data_count;
+			     cur_data_idx++) {
+				reg = data[start_offset + cur_data_idx];
+				writel(reg, dp->reg_base + EXYNOS_DP_BUF_DATA_0
+							  + 4 * cur_data_idx);
+			}
+
+			/*
+			 * Set DisplayPort transaction and write
+			 * If bit 3 is 1, DisplayPort transaction.
+			 * If Bit 3 is 0, I2C transaction.
+			 */
+			reg = AUX_LENGTH(cur_data_count) |
+				AUX_TX_COMM_DP_TRANSACTION | AUX_TX_COMM_WRITE;
+			writel(reg, dp->reg_base + EXYNOS_DP_AUX_CH_CTL_1);
+
+			/* Start AUX transaction */
+			retval = exynos_dp_start_aux_transaction(dp);
+			if (retval == 0)
+				break;
+			else
+				dev_dbg(dp->dev, "%s: Aux Transaction fail!\n",
+					__func__);
+		}
+
+		start_offset += cur_data_count;
+	}
+
+	return retval;
+}
+
+int exynos_dp_read_bytes_from_dpcd(struct exynos_dp_device *dp,
+				unsigned int reg_addr,
+				unsigned int count,
+				unsigned char data[])
+{
+	u32 reg;
+	unsigned int start_offset;
+	unsigned int cur_data_count;
+	unsigned int cur_data_idx;
+	int i;
+	int retval = 0;
+
+	/* Clear AUX CH data buffer */
+	reg = BUF_CLR;
+	writel(reg, dp->reg_base + EXYNOS_DP_BUFFER_DATA_CTL);
+
+	start_offset = 0;
+	while (start_offset < count) {
+		/* Buffer size of AUX CH is 16 * 4bytes */
+		if ((count - start_offset) > 16)
+			cur_data_count = 16;
+		else
+			cur_data_count = count - start_offset;
+
+		/* AUX CH Request Transaction process */
+		for (i = 0; i < 3; i++) {
+			/* Select DPCD device address */
+			reg = AUX_ADDR_7_0(reg_addr + start_offset);
+			writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_7_0);
+			reg = AUX_ADDR_15_8(reg_addr + start_offset);
+			writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_15_8);
+			reg = AUX_ADDR_19_16(reg_addr + start_offset);
+			writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_19_16);
+
+			/*
+			 * Set DisplayPort transaction and read
+			 * If bit 3 is 1, DisplayPort transaction.
+			 * If Bit 3 is 0, I2C transaction.
+			 */
+			reg = AUX_LENGTH(cur_data_count) |
+				AUX_TX_COMM_DP_TRANSACTION | AUX_TX_COMM_READ;
+			writel(reg, dp->reg_base + EXYNOS_DP_AUX_CH_CTL_1);
+
+			/* Start AUX transaction */
+			retval = exynos_dp_start_aux_transaction(dp);
+			if (retval == 0)
+				break;
+			else
+				dev_dbg(dp->dev, "%s: Aux Transaction fail!\n",
+					__func__);
+		}
+
+		for (cur_data_idx = 0; cur_data_idx < cur_data_count;
+		    cur_data_idx++) {
+			reg = readl(dp->reg_base + EXYNOS_DP_BUF_DATA_0
+						 + 4 * cur_data_idx);
+			data[start_offset + cur_data_idx] =
+				(unsigned char)reg;
+		}
+
+		start_offset += cur_data_count;
+	}
+
+	return retval;
+}
+
+int exynos_dp_select_i2c_device(struct exynos_dp_device *dp,
+				unsigned int device_addr,
+				unsigned int reg_addr)
+{
+	u32 reg;
+	int retval;
+
+	/* Set EDID device address */
+	reg = device_addr;
+	writel(reg, dp->reg_base + EXYNOS_DP_AUX_ADDR_7_0);
+	writel(0x0, dp->reg_base + EXYNOS_DP_AUX_ADDR_15_8);
+	writel(0x0, dp->reg_base + EXYNOS_DP_AUX_ADDR_19_16);
+
+	/* Set offset from base address of EDID device */
+	writel(reg_addr, dp->reg_base + EXYNOS_DP_BUF_DATA_0);
+
+	/*
+	 * Set I2C transaction and write address
+	 * If bit 3 is 1, DisplayPort transaction.
+	 * If Bit 3 is 0, I2C transaction.
+	 */
+	reg = AUX_TX_COMM_I2C_TRANSACTION | AUX_TX_COMM_MOT |
+		AUX_TX_COMM_WRITE;
+	writel(reg, dp->reg_base + EXYNOS_DP_AUX_CH_CTL_1);
+
+	/* Start AUX transaction */
+	retval = exynos_dp_start_aux_transaction(dp);
+	if (retval != 0)
+		dev_dbg(dp->dev, "%s: Aux Transaction fail!\n", __func__);
+
+	return retval;
+}
+
+int exynos_dp_read_byte_from_i2c(struct exynos_dp_device *dp,
+				unsigned int device_addr,
+				unsigned int reg_addr,
+				unsigned int *data)
+{
+	u32 reg;
+	int i;
+	int retval;
+
+	for (i = 0; i < 3; i++) {
+		/* Clear AUX CH data buffer */
+		reg = BUF_CLR;
+		writel(reg, dp->reg_base + EXYNOS_DP_BUFFER_DATA_CTL);
+
+		/* Select EDID device */
+		retval = exynos_dp_select_i2c_device(dp, device_addr, reg_addr);
+		if (retval != 0)
+			continue;
+
+		/*
+		 * Set I2C transaction and read data
+		 * If bit 3 is 1, DisplayPort transaction.
+		 * If Bit 3 is 0, I2C transaction.
+		 */
+		reg = AUX_TX_COMM_I2C_TRANSACTION |
+			AUX_TX_COMM_READ;
+		writel(reg, dp->reg_base + EXYNOS_DP_AUX_CH_CTL_1);
+
+		/* Start AUX transaction */
+		retval = exynos_dp_start_aux_transaction(dp);
+		if (retval == 0)
+			break;
+		else
+			dev_dbg(dp->dev, "%s: Aux Transaction fail!\n",
+				__func__);
+	}
+
+	/* Read data */
+	if (retval == 0)
+		*data = readl(dp->reg_base + EXYNOS_DP_BUF_DATA_0);
+
+	return retval;
+}
+
+int exynos_dp_read_bytes_from_i2c(struct exynos_dp_device *dp,
+				unsigned int device_addr,
+				unsigned int reg_addr,
+				unsigned int count,
+				unsigned char edid[])
+{
+	u32 reg;
+	unsigned int i, j;
+	unsigned int cur_data_idx;
+	unsigned int defer = 0;
+	int retval = 0;
+
+	for (i = 0; i < count; i += 16) {
+		for (j = 0; j < 3; j++) {
+			/* Clear AUX CH data buffer */
+			reg = BUF_CLR;
+			writel(reg, dp->reg_base + EXYNOS_DP_BUFFER_DATA_CTL);
+
+			/* Set normal AUX CH command */
+			reg = readl(dp->reg_base + EXYNOS_DP_AUX_CH_CTL_2);
+			reg &= ~ADDR_ONLY;
+			writel(reg, dp->reg_base + EXYNOS_DP_AUX_CH_CTL_2);
+
+			/*
+			 * If Rx sends defer, Tx sends only reads
+			 * request without sending address
+			 */
+			if (!defer)
+				retval = exynos_dp_select_i2c_device(dp,
+						device_addr, reg_addr + i);
+			else
+				defer = 0;
+
+			if (retval == 0) {
+				/*
+				 * Set I2C transaction and write data
+				 * If bit 3 is 1, DisplayPort transaction.
+				 * If Bit 3 is 0, I2C transaction.
+				 */
+				reg = AUX_LENGTH(16) |
+					AUX_TX_COMM_I2C_TRANSACTION |
+					AUX_TX_COMM_READ;
+				writel(reg, dp->reg_base +
+					EXYNOS_DP_AUX_CH_CTL_1);
+
+				/* Start AUX transaction */
+				retval = exynos_dp_start_aux_transaction(dp);
+				if (retval == 0)
+					break;
+				else
+					dev_dbg(dp->dev,
+						"%s: Aux Transaction fail!\n",
+						__func__);
+			}
+			/* Check if Rx sends defer */
+			reg = readl(dp->reg_base + EXYNOS_DP_AUX_RX_COMM);
+			if (reg == AUX_RX_COMM_AUX_DEFER ||
+				reg == AUX_RX_COMM_I2C_DEFER) {
+				dev_err(dp->dev, "Defer: %d\n\n", reg);
+				defer = 1;
+			}
+		}
+
+		for (cur_data_idx = 0; cur_data_idx < 16; cur_data_idx++) {
+			reg = readl(dp->reg_base + EXYNOS_DP_BUF_DATA_0
+						 + 4 * cur_data_idx);
+			edid[i + cur_data_idx] = (unsigned char)reg;
+		}
+	}
+
+	return retval;
+}
+
+void exynos_dp_set_link_bandwidth(struct exynos_dp_device *dp, u32 bwtype)
+{
+	u32 reg;
+
+	reg = bwtype;
+	if ((bwtype == LINK_RATE_2_70GBPS) || (bwtype == LINK_RATE_1_62GBPS))
+		writel(reg, dp->reg_base + EXYNOS_DP_LINK_BW_SET);
+}
+
+void exynos_dp_get_link_bandwidth(struct exynos_dp_device *dp, u32 *bwtype)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LINK_BW_SET);
+	*bwtype = reg;
+}
+
+void exynos_dp_set_lane_count(struct exynos_dp_device *dp, u32 count)
+{
+	u32 reg;
+
+	reg = count;
+	writel(reg, dp->reg_base + EXYNOS_DP_LANE_COUNT_SET);
+}
+
+void exynos_dp_get_lane_count(struct exynos_dp_device *dp, u32 *count)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LANE_COUNT_SET);
+	*count = reg;
+}
+
+void exynos_dp_enable_enhanced_mode(struct exynos_dp_device *dp, bool enable)
+{
+	u32 reg;
+
+	if (enable) {
+		reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_4);
+		reg |= ENHANCED;
+		writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_4);
+	} else {
+		reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_4);
+		reg &= ~ENHANCED;
+		writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_4);
+	}
+}
+
+void exynos_dp_set_training_pattern(struct exynos_dp_device *dp,
+				 enum pattern_set pattern)
+{
+	u32 reg;
+
+	switch (pattern) {
+	case PRBS7:
+		reg = SCRAMBLING_ENABLE | LINK_QUAL_PATTERN_SET_PRBS7;
+		writel(reg, dp->reg_base + EXYNOS_DP_TRAINING_PTN_SET);
+		break;
+	case D10_2:
+		reg = SCRAMBLING_ENABLE | LINK_QUAL_PATTERN_SET_D10_2;
+		writel(reg, dp->reg_base + EXYNOS_DP_TRAINING_PTN_SET);
+		break;
+	case TRAINING_PTN1:
+		reg = SCRAMBLING_DISABLE | SW_TRAINING_PATTERN_SET_PTN1;
+		writel(reg, dp->reg_base + EXYNOS_DP_TRAINING_PTN_SET);
+		break;
+	case TRAINING_PTN2:
+		reg = SCRAMBLING_DISABLE | SW_TRAINING_PATTERN_SET_PTN2;
+		writel(reg, dp->reg_base + EXYNOS_DP_TRAINING_PTN_SET);
+		break;
+	case DP_NONE:
+		reg = SCRAMBLING_ENABLE |
+			LINK_QUAL_PATTERN_SET_DISABLE |
+			SW_TRAINING_PATTERN_SET_NORMAL;
+		writel(reg, dp->reg_base + EXYNOS_DP_TRAINING_PTN_SET);
+		break;
+	default:
+		break;
+	}
+}
+
+void exynos_dp_set_lane0_pre_emphasis(struct exynos_dp_device *dp, u32 level)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LN0_LINK_TRAINING_CTL);
+	reg &= ~PRE_EMPHASIS_SET_MASK;
+	reg |= level << PRE_EMPHASIS_SET_SHIFT;
+	writel(reg, dp->reg_base + EXYNOS_DP_LN0_LINK_TRAINING_CTL);
+}
+
+void exynos_dp_set_lane1_pre_emphasis(struct exynos_dp_device *dp, u32 level)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LN1_LINK_TRAINING_CTL);
+	reg &= ~PRE_EMPHASIS_SET_MASK;
+	reg |= level << PRE_EMPHASIS_SET_SHIFT;
+	writel(reg, dp->reg_base + EXYNOS_DP_LN1_LINK_TRAINING_CTL);
+}
+
+void exynos_dp_set_lane2_pre_emphasis(struct exynos_dp_device *dp, u32 level)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LN2_LINK_TRAINING_CTL);
+	reg &= ~PRE_EMPHASIS_SET_MASK;
+	reg |= level << PRE_EMPHASIS_SET_SHIFT;
+	writel(reg, dp->reg_base + EXYNOS_DP_LN2_LINK_TRAINING_CTL);
+}
+
+void exynos_dp_set_lane3_pre_emphasis(struct exynos_dp_device *dp, u32 level)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LN3_LINK_TRAINING_CTL);
+	reg &= ~PRE_EMPHASIS_SET_MASK;
+	reg |= level << PRE_EMPHASIS_SET_SHIFT;
+	writel(reg, dp->reg_base + EXYNOS_DP_LN3_LINK_TRAINING_CTL);
+}
+
+void exynos_dp_set_lane0_link_training(struct exynos_dp_device *dp,
+					u32 training_lane)
+{
+	u32 reg;
+
+	reg = training_lane;
+	writel(reg, dp->reg_base + EXYNOS_DP_LN0_LINK_TRAINING_CTL);
+}
+
+void exynos_dp_set_lane1_link_training(struct exynos_dp_device *dp,
+					u32 training_lane)
+{
+	u32 reg;
+
+	reg = training_lane;
+	writel(reg, dp->reg_base + EXYNOS_DP_LN1_LINK_TRAINING_CTL);
+}
+
+void exynos_dp_set_lane2_link_training(struct exynos_dp_device *dp,
+					u32 training_lane)
+{
+	u32 reg;
+
+	reg = training_lane;
+	writel(reg, dp->reg_base + EXYNOS_DP_LN2_LINK_TRAINING_CTL);
+}
+
+void exynos_dp_set_lane3_link_training(struct exynos_dp_device *dp,
+					u32 training_lane)
+{
+	u32 reg;
+
+	reg = training_lane;
+	writel(reg, dp->reg_base + EXYNOS_DP_LN3_LINK_TRAINING_CTL);
+}
+
+u32 exynos_dp_get_lane0_link_training(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LN0_LINK_TRAINING_CTL);
+	return reg;
+}
+
+u32 exynos_dp_get_lane1_link_training(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LN1_LINK_TRAINING_CTL);
+	return reg;
+}
+
+u32 exynos_dp_get_lane2_link_training(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LN2_LINK_TRAINING_CTL);
+	return reg;
+}
+
+u32 exynos_dp_get_lane3_link_training(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_LN3_LINK_TRAINING_CTL);
+	return reg;
+}
+
+void exynos_dp_reset_macro(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_PHY_TEST);
+	reg |= MACRO_RST;
+	writel(reg, dp->reg_base + EXYNOS_DP_PHY_TEST);
+
+	/* 10 us is the minimum reset time. */
+	usleep_range(10, 20);
+
+	reg &= ~MACRO_RST;
+	writel(reg, dp->reg_base + EXYNOS_DP_PHY_TEST);
+}
+
+void exynos_dp_init_video(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = VSYNC_DET | VID_FORMAT_CHG | VID_CLK_CHG;
+	writel(reg, dp->reg_base + EXYNOS_DP_COMMON_INT_STA_1);
+
+	reg = 0x0;
+	writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_1);
+
+	reg = CHA_CRI(4) | CHA_CTRL;
+	writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_2);
+
+	reg = 0x0;
+	writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_3);
+
+	reg = VID_HRES_TH(2) | VID_VRES_TH(0);
+	writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_8);
+}
+
+void exynos_dp_set_video_color_format(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	/* Configure the input color depth, color space, dynamic range */
+	reg = (dp->video_info->dynamic_range << IN_D_RANGE_SHIFT) |
+		(dp->video_info->color_depth << IN_BPC_SHIFT) |
+		(dp->video_info->color_space << IN_COLOR_F_SHIFT);
+	writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_2);
+
+	/* Set Input Color YCbCr Coefficients to ITU601 or ITU709 */
+	reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_3);
+	reg &= ~IN_YC_COEFFI_MASK;
+	if (dp->video_info->ycbcr_coeff)
+		reg |= IN_YC_COEFFI_ITU709;
+	else
+		reg |= IN_YC_COEFFI_ITU601;
+	writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_3);
+}
+
+int exynos_dp_is_slave_video_stream_clock_on(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_1);
+	writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_1);
+
+	reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_1);
+
+	if (!(reg & DET_STA)) {
+		dev_dbg(dp->dev, "Input stream clock not detected.\n");
+		return -EINVAL;
+	}
+
+	reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_2);
+	writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_2);
+
+	reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_2);
+	dev_dbg(dp->dev, "wait SYS_CTL_2.\n");
+
+	if (reg & CHA_STA) {
+		dev_dbg(dp->dev, "Input stream clk is changing\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+void exynos_dp_set_video_cr_mn(struct exynos_dp_device *dp,
+		enum clock_recovery_m_value_type type,
+		u32 m_value,
+		u32 n_value)
+{
+	u32 reg;
+
+	if (type == REGISTER_M) {
+		reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_4);
+		reg |= FIX_M_VID;
+		writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_4);
+		reg = m_value & 0xff;
+		writel(reg, dp->reg_base + EXYNOS_DP_M_VID_0);
+		reg = (m_value >> 8) & 0xff;
+		writel(reg, dp->reg_base + EXYNOS_DP_M_VID_1);
+		reg = (m_value >> 16) & 0xff;
+		writel(reg, dp->reg_base + EXYNOS_DP_M_VID_2);
+
+		reg = n_value & 0xff;
+		writel(reg, dp->reg_base + EXYNOS_DP_N_VID_0);
+		reg = (n_value >> 8) & 0xff;
+		writel(reg, dp->reg_base + EXYNOS_DP_N_VID_1);
+		reg = (n_value >> 16) & 0xff;
+		writel(reg, dp->reg_base + EXYNOS_DP_N_VID_2);
+	} else  {
+		reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_4);
+		reg &= ~FIX_M_VID;
+		writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_4);
+
+		writel(0x00, dp->reg_base + EXYNOS_DP_N_VID_0);
+		writel(0x80, dp->reg_base + EXYNOS_DP_N_VID_1);
+		writel(0x00, dp->reg_base + EXYNOS_DP_N_VID_2);
+	}
+}
+
+void exynos_dp_set_video_timing_mode(struct exynos_dp_device *dp, u32 type)
+{
+	u32 reg;
+
+	if (type == VIDEO_TIMING_FROM_CAPTURE) {
+		reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+		reg &= ~FORMAT_SEL;
+		writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+	} else {
+		reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+		reg |= FORMAT_SEL;
+		writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+	}
+}
+
+void exynos_dp_enable_video_master(struct exynos_dp_device *dp, bool enable)
+{
+	u32 reg;
+
+	if (enable) {
+		reg = readl(dp->reg_base + EXYNOS_DP_SOC_GENERAL_CTL);
+		reg &= ~VIDEO_MODE_MASK;
+		reg |= VIDEO_MASTER_MODE_EN | VIDEO_MODE_MASTER_MODE;
+		writel(reg, dp->reg_base + EXYNOS_DP_SOC_GENERAL_CTL);
+	} else {
+		reg = readl(dp->reg_base + EXYNOS_DP_SOC_GENERAL_CTL);
+		reg &= ~VIDEO_MODE_MASK;
+		reg |= VIDEO_MODE_SLAVE_MODE;
+		writel(reg, dp->reg_base + EXYNOS_DP_SOC_GENERAL_CTL);
+	}
+}
+
+void exynos_dp_start_video(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_1);
+	reg |= VIDEO_EN;
+	writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_1);
+}
+
+int exynos_dp_is_video_stream_on(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_3);
+	writel(reg, dp->reg_base + EXYNOS_DP_SYS_CTL_3);
+
+	reg = readl(dp->reg_base + EXYNOS_DP_SYS_CTL_3);
+	if (!(reg & STRM_VALID)) {
+		dev_dbg(dp->dev, "Input video stream is not detected.\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+void exynos_dp_config_video_slave_mode(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_FUNC_EN_1);
+	reg &= ~(MASTER_VID_FUNC_EN_N|SLAVE_VID_FUNC_EN_N);
+	reg |= MASTER_VID_FUNC_EN_N;
+	writel(reg, dp->reg_base + EXYNOS_DP_FUNC_EN_1);
+
+	reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+	reg &= ~INTERACE_SCAN_CFG;
+	reg |= (dp->video_info->interlaced << 2);
+	writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+
+	reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+	reg &= ~VSYNC_POLARITY_CFG;
+	reg |= (dp->video_info->v_sync_polarity << 1);
+	writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+
+	reg = readl(dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+	reg &= ~HSYNC_POLARITY_CFG;
+	reg |= (dp->video_info->h_sync_polarity << 0);
+	writel(reg, dp->reg_base + EXYNOS_DP_VIDEO_CTL_10);
+
+	reg = AUDIO_MODE_SPDIF_MODE | VIDEO_MODE_SLAVE_MODE;
+	writel(reg, dp->reg_base + EXYNOS_DP_SOC_GENERAL_CTL);
+}
+
+void exynos_dp_enable_scrambling(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_TRAINING_PTN_SET);
+	reg &= ~SCRAMBLING_DISABLE;
+	writel(reg, dp->reg_base + EXYNOS_DP_TRAINING_PTN_SET);
+}
+
+void exynos_dp_disable_scrambling(struct exynos_dp_device *dp)
+{
+	u32 reg;
+
+	reg = readl(dp->reg_base + EXYNOS_DP_TRAINING_PTN_SET);
+	reg |= SCRAMBLING_DISABLE;
+	writel(reg, dp->reg_base + EXYNOS_DP_TRAINING_PTN_SET);
+}
diff --git a/drivers/gpu/drm/exynos/exynos_dp_reg.h b/drivers/gpu/drm/exynos/exynos_dp_reg.h
new file mode 100644
index 00000000000..2e9bd0e0b9f
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_dp_reg.h
@@ -0,0 +1,366 @@
+/*
+ * Register definition file for Samsung DP driver
+ *
+ * Copyright (C) 2012 Samsung Electronics Co., Ltd.
+ * Author: Jingoo Han <jg1.han@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef _EXYNOS_DP_REG_H
+#define _EXYNOS_DP_REG_H
+
+#define EXYNOS_DP_TX_SW_RESET			0x14
+#define EXYNOS_DP_FUNC_EN_1			0x18
+#define EXYNOS_DP_FUNC_EN_2			0x1C
+#define EXYNOS_DP_VIDEO_CTL_1			0x20
+#define EXYNOS_DP_VIDEO_CTL_2			0x24
+#define EXYNOS_DP_VIDEO_CTL_3			0x28
+
+#define EXYNOS_DP_VIDEO_CTL_8			0x3C
+#define EXYNOS_DP_VIDEO_CTL_10			0x44
+
+#define EXYNOS_DP_LANE_MAP			0x35C
+
+#define EXYNOS_DP_ANALOG_CTL_1			0x370
+#define EXYNOS_DP_ANALOG_CTL_2			0x374
+#define EXYNOS_DP_ANALOG_CTL_3			0x378
+#define EXYNOS_DP_PLL_FILTER_CTL_1		0x37C
+#define EXYNOS_DP_TX_AMP_TUNING_CTL		0x380
+
+#define EXYNOS_DP_AUX_HW_RETRY_CTL		0x390
+
+#define EXYNOS_DP_COMMON_INT_STA_1		0x3C4
+#define EXYNOS_DP_COMMON_INT_STA_2		0x3C8
+#define EXYNOS_DP_COMMON_INT_STA_3		0x3CC
+#define EXYNOS_DP_COMMON_INT_STA_4		0x3D0
+#define EXYNOS_DP_INT_STA			0x3DC
+#define EXYNOS_DP_COMMON_INT_MASK_1		0x3E0
+#define EXYNOS_DP_COMMON_INT_MASK_2		0x3E4
+#define EXYNOS_DP_COMMON_INT_MASK_3		0x3E8
+#define EXYNOS_DP_COMMON_INT_MASK_4		0x3EC
+#define EXYNOS_DP_INT_STA_MASK			0x3F8
+#define EXYNOS_DP_INT_CTL			0x3FC
+
+#define EXYNOS_DP_SYS_CTL_1			0x600
+#define EXYNOS_DP_SYS_CTL_2			0x604
+#define EXYNOS_DP_SYS_CTL_3			0x608
+#define EXYNOS_DP_SYS_CTL_4			0x60C
+
+#define EXYNOS_DP_PKT_SEND_CTL			0x640
+#define EXYNOS_DP_HDCP_CTL			0x648
+
+#define EXYNOS_DP_LINK_BW_SET			0x680
+#define EXYNOS_DP_LANE_COUNT_SET		0x684
+#define EXYNOS_DP_TRAINING_PTN_SET		0x688
+#define EXYNOS_DP_LN0_LINK_TRAINING_CTL		0x68C
+#define EXYNOS_DP_LN1_LINK_TRAINING_CTL		0x690
+#define EXYNOS_DP_LN2_LINK_TRAINING_CTL		0x694
+#define EXYNOS_DP_LN3_LINK_TRAINING_CTL		0x698
+
+#define EXYNOS_DP_DEBUG_CTL			0x6C0
+#define EXYNOS_DP_HPD_DEGLITCH_L		0x6C4
+#define EXYNOS_DP_HPD_DEGLITCH_H		0x6C8
+#define EXYNOS_DP_LINK_DEBUG_CTL		0x6E0
+
+#define EXYNOS_DP_M_VID_0			0x700
+#define EXYNOS_DP_M_VID_1			0x704
+#define EXYNOS_DP_M_VID_2			0x708
+#define EXYNOS_DP_N_VID_0			0x70C
+#define EXYNOS_DP_N_VID_1			0x710
+#define EXYNOS_DP_N_VID_2			0x714
+
+#define EXYNOS_DP_PLL_CTL			0x71C
+#define EXYNOS_DP_PHY_PD			0x720
+#define EXYNOS_DP_PHY_TEST			0x724
+
+#define EXYNOS_DP_VIDEO_FIFO_THRD		0x730
+#define EXYNOS_DP_AUDIO_MARGIN			0x73C
+
+#define EXYNOS_DP_M_VID_GEN_FILTER_TH		0x764
+#define EXYNOS_DP_M_AUD_GEN_FILTER_TH		0x778
+#define EXYNOS_DP_AUX_CH_STA			0x780
+#define EXYNOS_DP_AUX_CH_DEFER_CTL		0x788
+#define EXYNOS_DP_AUX_RX_COMM			0x78C
+#define EXYNOS_DP_BUFFER_DATA_CTL		0x790
+#define EXYNOS_DP_AUX_CH_CTL_1			0x794
+#define EXYNOS_DP_AUX_ADDR_7_0			0x798
+#define EXYNOS_DP_AUX_ADDR_15_8			0x79C
+#define EXYNOS_DP_AUX_ADDR_19_16		0x7A0
+#define EXYNOS_DP_AUX_CH_CTL_2			0x7A4
+
+#define EXYNOS_DP_BUF_DATA_0			0x7C0
+
+#define EXYNOS_DP_SOC_GENERAL_CTL		0x800
+
+/* EXYNOS_DP_TX_SW_RESET */
+#define RESET_DP_TX				(0x1 << 0)
+
+/* EXYNOS_DP_FUNC_EN_1 */
+#define MASTER_VID_FUNC_EN_N			(0x1 << 7)
+#define SLAVE_VID_FUNC_EN_N			(0x1 << 5)
+#define AUD_FIFO_FUNC_EN_N			(0x1 << 4)
+#define AUD_FUNC_EN_N				(0x1 << 3)
+#define HDCP_FUNC_EN_N				(0x1 << 2)
+#define CRC_FUNC_EN_N				(0x1 << 1)
+#define SW_FUNC_EN_N				(0x1 << 0)
+
+/* EXYNOS_DP_FUNC_EN_2 */
+#define SSC_FUNC_EN_N				(0x1 << 7)
+#define AUX_FUNC_EN_N				(0x1 << 2)
+#define SERDES_FIFO_FUNC_EN_N			(0x1 << 1)
+#define LS_CLK_DOMAIN_FUNC_EN_N			(0x1 << 0)
+
+/* EXYNOS_DP_VIDEO_CTL_1 */
+#define VIDEO_EN				(0x1 << 7)
+#define HDCP_VIDEO_MUTE				(0x1 << 6)
+
+/* EXYNOS_DP_VIDEO_CTL_1 */
+#define IN_D_RANGE_MASK				(0x1 << 7)
+#define IN_D_RANGE_SHIFT			(7)
+#define IN_D_RANGE_CEA				(0x1 << 7)
+#define IN_D_RANGE_VESA				(0x0 << 7)
+#define IN_BPC_MASK				(0x7 << 4)
+#define IN_BPC_SHIFT				(4)
+#define IN_BPC_12_BITS				(0x3 << 4)
+#define IN_BPC_10_BITS				(0x2 << 4)
+#define IN_BPC_8_BITS				(0x1 << 4)
+#define IN_BPC_6_BITS				(0x0 << 4)
+#define IN_COLOR_F_MASK				(0x3 << 0)
+#define IN_COLOR_F_SHIFT			(0)
+#define IN_COLOR_F_YCBCR444			(0x2 << 0)
+#define IN_COLOR_F_YCBCR422			(0x1 << 0)
+#define IN_COLOR_F_RGB				(0x0 << 0)
+
+/* EXYNOS_DP_VIDEO_CTL_3 */
+#define IN_YC_COEFFI_MASK			(0x1 << 7)
+#define IN_YC_COEFFI_SHIFT			(7)
+#define IN_YC_COEFFI_ITU709			(0x1 << 7)
+#define IN_YC_COEFFI_ITU601			(0x0 << 7)
+#define VID_CHK_UPDATE_TYPE_MASK		(0x1 << 4)
+#define VID_CHK_UPDATE_TYPE_SHIFT		(4)
+#define VID_CHK_UPDATE_TYPE_1			(0x1 << 4)
+#define VID_CHK_UPDATE_TYPE_0			(0x0 << 4)
+
+/* EXYNOS_DP_VIDEO_CTL_8 */
+#define VID_HRES_TH(x)				(((x) & 0xf) << 4)
+#define VID_VRES_TH(x)				(((x) & 0xf) << 0)
+
+/* EXYNOS_DP_VIDEO_CTL_10 */
+#define FORMAT_SEL				(0x1 << 4)
+#define INTERACE_SCAN_CFG			(0x1 << 2)
+#define VSYNC_POLARITY_CFG			(0x1 << 1)
+#define HSYNC_POLARITY_CFG			(0x1 << 0)
+
+/* EXYNOS_DP_LANE_MAP */
+#define LANE3_MAP_LOGIC_LANE_0			(0x0 << 6)
+#define LANE3_MAP_LOGIC_LANE_1			(0x1 << 6)
+#define LANE3_MAP_LOGIC_LANE_2			(0x2 << 6)
+#define LANE3_MAP_LOGIC_LANE_3			(0x3 << 6)
+#define LANE2_MAP_LOGIC_LANE_0			(0x0 << 4)
+#define LANE2_MAP_LOGIC_LANE_1			(0x1 << 4)
+#define LANE2_MAP_LOGIC_LANE_2			(0x2 << 4)
+#define LANE2_MAP_LOGIC_LANE_3			(0x3 << 4)
+#define LANE1_MAP_LOGIC_LANE_0			(0x0 << 2)
+#define LANE1_MAP_LOGIC_LANE_1			(0x1 << 2)
+#define LANE1_MAP_LOGIC_LANE_2			(0x2 << 2)
+#define LANE1_MAP_LOGIC_LANE_3			(0x3 << 2)
+#define LANE0_MAP_LOGIC_LANE_0			(0x0 << 0)
+#define LANE0_MAP_LOGIC_LANE_1			(0x1 << 0)
+#define LANE0_MAP_LOGIC_LANE_2			(0x2 << 0)
+#define LANE0_MAP_LOGIC_LANE_3			(0x3 << 0)
+
+/* EXYNOS_DP_ANALOG_CTL_1 */
+#define TX_TERMINAL_CTRL_50_OHM			(0x1 << 4)
+
+/* EXYNOS_DP_ANALOG_CTL_2 */
+#define SEL_24M					(0x1 << 3)
+#define TX_DVDD_BIT_1_0625V			(0x4 << 0)
+
+/* EXYNOS_DP_ANALOG_CTL_3 */
+#define DRIVE_DVDD_BIT_1_0625V			(0x4 << 5)
+#define VCO_BIT_600_MICRO			(0x5 << 0)
+
+/* EXYNOS_DP_PLL_FILTER_CTL_1 */
+#define PD_RING_OSC				(0x1 << 6)
+#define AUX_TERMINAL_CTRL_50_OHM		(0x2 << 4)
+#define TX_CUR1_2X				(0x1 << 2)
+#define TX_CUR_16_MA				(0x3 << 0)
+
+/* EXYNOS_DP_TX_AMP_TUNING_CTL */
+#define CH3_AMP_400_MV				(0x0 << 24)
+#define CH2_AMP_400_MV				(0x0 << 16)
+#define CH1_AMP_400_MV				(0x0 << 8)
+#define CH0_AMP_400_MV				(0x0 << 0)
+
+/* EXYNOS_DP_AUX_HW_RETRY_CTL */
+#define AUX_BIT_PERIOD_EXPECTED_DELAY(x)	(((x) & 0x7) << 8)
+#define AUX_HW_RETRY_INTERVAL_MASK		(0x3 << 3)
+#define AUX_HW_RETRY_INTERVAL_600_MICROSECONDS	(0x0 << 3)
+#define AUX_HW_RETRY_INTERVAL_800_MICROSECONDS	(0x1 << 3)
+#define AUX_HW_RETRY_INTERVAL_1000_MICROSECONDS	(0x2 << 3)
+#define AUX_HW_RETRY_INTERVAL_1800_MICROSECONDS	(0x3 << 3)
+#define AUX_HW_RETRY_COUNT_SEL(x)		(((x) & 0x7) << 0)
+
+/* EXYNOS_DP_COMMON_INT_STA_1 */
+#define VSYNC_DET				(0x1 << 7)
+#define PLL_LOCK_CHG				(0x1 << 6)
+#define SPDIF_ERR				(0x1 << 5)
+#define SPDIF_UNSTBL				(0x1 << 4)
+#define VID_FORMAT_CHG				(0x1 << 3)
+#define AUD_CLK_CHG				(0x1 << 2)
+#define VID_CLK_CHG				(0x1 << 1)
+#define SW_INT					(0x1 << 0)
+
+/* EXYNOS_DP_COMMON_INT_STA_2 */
+#define ENC_EN_CHG				(0x1 << 6)
+#define HW_BKSV_RDY				(0x1 << 3)
+#define HW_SHA_DONE				(0x1 << 2)
+#define HW_AUTH_STATE_CHG			(0x1 << 1)
+#define HW_AUTH_DONE				(0x1 << 0)
+
+/* EXYNOS_DP_COMMON_INT_STA_3 */
+#define AFIFO_UNDER				(0x1 << 7)
+#define AFIFO_OVER				(0x1 << 6)
+#define R0_CHK_FLAG				(0x1 << 5)
+
+/* EXYNOS_DP_COMMON_INT_STA_4 */
+#define PSR_ACTIVE				(0x1 << 7)
+#define PSR_INACTIVE				(0x1 << 6)
+#define SPDIF_BI_PHASE_ERR			(0x1 << 5)
+#define HOTPLUG_CHG				(0x1 << 2)
+#define HPD_LOST				(0x1 << 1)
+#define PLUG					(0x1 << 0)
+
+/* EXYNOS_DP_INT_STA */
+#define INT_HPD					(0x1 << 6)
+#define HW_TRAINING_FINISH			(0x1 << 5)
+#define RPLY_RECEIV				(0x1 << 1)
+#define AUX_ERR					(0x1 << 0)
+
+/* EXYNOS_DP_INT_CTL */
+#define SOFT_INT_CTRL				(0x1 << 2)
+#define INT_POL1				(0x1 << 1)
+#define INT_POL0				(0x1 << 0)
+
+/* EXYNOS_DP_SYS_CTL_1 */
+#define DET_STA					(0x1 << 2)
+#define FORCE_DET				(0x1 << 1)
+#define DET_CTRL				(0x1 << 0)
+
+/* EXYNOS_DP_SYS_CTL_2 */
+#define CHA_CRI(x)				(((x) & 0xf) << 4)
+#define CHA_STA					(0x1 << 2)
+#define FORCE_CHA				(0x1 << 1)
+#define CHA_CTRL				(0x1 << 0)
+
+/* EXYNOS_DP_SYS_CTL_3 */
+#define HPD_STATUS				(0x1 << 6)
+#define F_HPD					(0x1 << 5)
+#define HPD_CTRL				(0x1 << 4)
+#define HDCP_RDY				(0x1 << 3)
+#define STRM_VALID				(0x1 << 2)
+#define F_VALID					(0x1 << 1)
+#define VALID_CTRL				(0x1 << 0)
+
+/* EXYNOS_DP_SYS_CTL_4 */
+#define FIX_M_AUD				(0x1 << 4)
+#define ENHANCED				(0x1 << 3)
+#define FIX_M_VID				(0x1 << 2)
+#define M_VID_UPDATE_CTRL			(0x3 << 0)
+
+/* EXYNOS_DP_TRAINING_PTN_SET */
+#define SCRAMBLER_TYPE				(0x1 << 9)
+#define HW_LINK_TRAINING_PATTERN		(0x1 << 8)
+#define SCRAMBLING_DISABLE			(0x1 << 5)
+#define SCRAMBLING_ENABLE			(0x0 << 5)
+#define LINK_QUAL_PATTERN_SET_MASK		(0x3 << 2)
+#define LINK_QUAL_PATTERN_SET_PRBS7		(0x3 << 2)
+#define LINK_QUAL_PATTERN_SET_D10_2		(0x1 << 2)
+#define LINK_QUAL_PATTERN_SET_DISABLE		(0x0 << 2)
+#define SW_TRAINING_PATTERN_SET_MASK		(0x3 << 0)
+#define SW_TRAINING_PATTERN_SET_PTN2		(0x2 << 0)
+#define SW_TRAINING_PATTERN_SET_PTN1		(0x1 << 0)
+#define SW_TRAINING_PATTERN_SET_NORMAL		(0x0 << 0)
+
+/* EXYNOS_DP_LN0_LINK_TRAINING_CTL */
+#define PRE_EMPHASIS_SET_MASK			(0x3 << 3)
+#define PRE_EMPHASIS_SET_SHIFT			(3)
+
+/* EXYNOS_DP_DEBUG_CTL */
+#define PLL_LOCK				(0x1 << 4)
+#define F_PLL_LOCK				(0x1 << 3)
+#define PLL_LOCK_CTRL				(0x1 << 2)
+#define PN_INV					(0x1 << 0)
+
+/* EXYNOS_DP_PLL_CTL */
+#define DP_PLL_PD				(0x1 << 7)
+#define DP_PLL_RESET				(0x1 << 6)
+#define DP_PLL_LOOP_BIT_DEFAULT			(0x1 << 4)
+#define DP_PLL_REF_BIT_1_1250V			(0x5 << 0)
+#define DP_PLL_REF_BIT_1_2500V			(0x7 << 0)
+
+/* EXYNOS_DP_PHY_PD */
+#define DP_PHY_PD				(0x1 << 5)
+#define AUX_PD					(0x1 << 4)
+#define CH3_PD					(0x1 << 3)
+#define CH2_PD					(0x1 << 2)
+#define CH1_PD					(0x1 << 1)
+#define CH0_PD					(0x1 << 0)
+
+/* EXYNOS_DP_PHY_TEST */
+#define MACRO_RST				(0x1 << 5)
+#define CH1_TEST				(0x1 << 1)
+#define CH0_TEST				(0x1 << 0)
+
+/* EXYNOS_DP_AUX_CH_STA */
+#define AUX_BUSY				(0x1 << 4)
+#define AUX_STATUS_MASK				(0xf << 0)
+
+/* EXYNOS_DP_AUX_CH_DEFER_CTL */
+#define DEFER_CTRL_EN				(0x1 << 7)
+#define DEFER_COUNT(x)				(((x) & 0x7f) << 0)
+
+/* EXYNOS_DP_AUX_RX_COMM */
+#define AUX_RX_COMM_I2C_DEFER			(0x2 << 2)
+#define AUX_RX_COMM_AUX_DEFER			(0x2 << 0)
+
+/* EXYNOS_DP_BUFFER_DATA_CTL */
+#define BUF_CLR					(0x1 << 7)
+#define BUF_DATA_COUNT(x)			(((x) & 0x1f) << 0)
+
+/* EXYNOS_DP_AUX_CH_CTL_1 */
+#define AUX_LENGTH(x)				(((x - 1) & 0xf) << 4)
+#define AUX_TX_COMM_MASK			(0xf << 0)
+#define AUX_TX_COMM_DP_TRANSACTION		(0x1 << 3)
+#define AUX_TX_COMM_I2C_TRANSACTION		(0x0 << 3)
+#define AUX_TX_COMM_MOT				(0x1 << 2)
+#define AUX_TX_COMM_WRITE			(0x0 << 0)
+#define AUX_TX_COMM_READ			(0x1 << 0)
+
+/* EXYNOS_DP_AUX_ADDR_7_0 */
+#define AUX_ADDR_7_0(x)				(((x) >> 0) & 0xff)
+
+/* EXYNOS_DP_AUX_ADDR_15_8 */
+#define AUX_ADDR_15_8(x)			(((x) >> 8) & 0xff)
+
+/* EXYNOS_DP_AUX_ADDR_19_16 */
+#define AUX_ADDR_19_16(x)			(((x) >> 16) & 0x0f)
+
+/* EXYNOS_DP_AUX_CH_CTL_2 */
+#define ADDR_ONLY				(0x1 << 1)
+#define AUX_EN					(0x1 << 0)
+
+/* EXYNOS_DP_SOC_GENERAL_CTL */
+#define AUDIO_MODE_SPDIF_MODE			(0x1 << 8)
+#define AUDIO_MODE_MASTER_MODE			(0x0 << 8)
+#define MASTER_VIDEO_INTERLACE_EN		(0x1 << 4)
+#define VIDEO_MASTER_CLK_SEL			(0x1 << 2)
+#define VIDEO_MASTER_MODE_EN			(0x1 << 1)
+#define VIDEO_MODE_MASK				(0x1 << 0)
+#define VIDEO_MODE_SLAVE_MODE			(0x1 << 0)
+#define VIDEO_MODE_MASTER_MODE			(0x0 << 0)
+
+#endif /* _EXYNOS_DP_REG_H */
diff --git a/drivers/gpu/drm/exynos/exynos_drm_connector.c b/drivers/gpu/drm/exynos/exynos_drm_connector.c
index e082efb2fec..9a16dbe121d 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_connector.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_connector.c
@@ -23,27 +23,20 @@
 				drm_connector)
 
 struct exynos_drm_connector {
-	struct drm_connector	drm_connector;
-	uint32_t		encoder_id;
-	struct exynos_drm_manager *manager;
-	uint32_t		dpms;
+	struct drm_connector		drm_connector;
+	uint32_t			encoder_id;
+	struct exynos_drm_display	*display;
 };
 
 static int exynos_drm_connector_get_modes(struct drm_connector *connector)
 {
 	struct exynos_drm_connector *exynos_connector =
 					to_exynos_connector(connector);
-	struct exynos_drm_manager *manager = exynos_connector->manager;
-	struct exynos_drm_display_ops *display_ops = manager->display_ops;
+	struct exynos_drm_display *display = exynos_connector->display;
 	struct edid *edid = NULL;
 	unsigned int count = 0;
 	int ret;
 
-	if (!display_ops) {
-		DRM_DEBUG_KMS("display_ops is null.\n");
-		return 0;
-	}
-
 	/*
 	 * if get_edid() exists then get_edid() callback of hdmi side
 	 * is called to get edid data through i2c interface else
@@ -52,8 +45,8 @@ static int exynos_drm_connector_get_modes(struct drm_connector *connector)
 	 * P.S. in case of lcd panel, count is always 1 if success
 	 * because lcd panel has only one mode.
 	 */
-	if (display_ops->get_edid) {
-		edid = display_ops->get_edid(manager->dev, connector);
+	if (display->ops->get_edid) {
+		edid = display->ops->get_edid(display, connector);
 		if (IS_ERR_OR_NULL(edid)) {
 			ret = PTR_ERR(edid);
 			edid = NULL;
@@ -76,8 +69,8 @@ static int exynos_drm_connector_get_modes(struct drm_connector *connector)
 			return 0;
 		}
 
-		if (display_ops->get_panel)
-			panel = display_ops->get_panel(manager->dev);
+		if (display->ops->get_panel)
+			panel = display->ops->get_panel(display);
 		else {
 			drm_mode_destroy(connector->dev, mode);
 			return 0;
@@ -106,20 +99,20 @@ static int exynos_drm_connector_mode_valid(struct drm_connector *connector,
 {
 	struct exynos_drm_connector *exynos_connector =
 					to_exynos_connector(connector);
-	struct exynos_drm_manager *manager = exynos_connector->manager;
-	struct exynos_drm_display_ops *display_ops = manager->display_ops;
+	struct exynos_drm_display *display = exynos_connector->display;
 	int ret = MODE_BAD;
 
 	DRM_DEBUG_KMS("%s\n", __FILE__);
 
-	if (display_ops && display_ops->check_mode)
-		if (!display_ops->check_mode(manager->dev, mode))
+	if (display->ops->check_mode)
+		if (!display->ops->check_mode(display, mode))
 			ret = MODE_OK;
 
 	return ret;
 }
 
-struct drm_encoder *exynos_drm_best_encoder(struct drm_connector *connector)
+static struct drm_encoder *exynos_drm_best_encoder(
+		struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
 	struct exynos_drm_connector *exynos_connector =
@@ -146,48 +139,12 @@ static struct drm_connector_helper_funcs exynos_connector_helper_funcs = {
 	.best_encoder	= exynos_drm_best_encoder,
 };
 
-void exynos_drm_display_power(struct drm_connector *connector, int mode)
-{
-	struct drm_encoder *encoder = exynos_drm_best_encoder(connector);
-	struct exynos_drm_connector *exynos_connector;
-	struct exynos_drm_manager *manager = exynos_drm_get_manager(encoder);
-	struct exynos_drm_display_ops *display_ops = manager->display_ops;
-
-	exynos_connector = to_exynos_connector(connector);
-
-	if (exynos_connector->dpms == mode) {
-		DRM_DEBUG_KMS("desired dpms mode is same as previous one.\n");
-		return;
-	}
-
-	if (display_ops && display_ops->power_on)
-		display_ops->power_on(manager->dev, mode);
-
-	exynos_connector->dpms = mode;
-}
-
-static void exynos_drm_connector_dpms(struct drm_connector *connector,
-					int mode)
-{
-	/*
-	 * in case that drm_crtc_helper_set_mode() is called,
-	 * encoder/crtc->funcs->dpms() will be just returned
-	 * because they already were DRM_MODE_DPMS_ON so only
-	 * exynos_drm_display_power() will be called.
-	 */
-	drm_helper_connector_dpms(connector, mode);
-
-	exynos_drm_display_power(connector, mode);
-
-}
-
 static int exynos_drm_connector_fill_modes(struct drm_connector *connector,
 				unsigned int max_width, unsigned int max_height)
 {
 	struct exynos_drm_connector *exynos_connector =
 					to_exynos_connector(connector);
-	struct exynos_drm_manager *manager = exynos_connector->manager;
-	struct exynos_drm_manager_ops *ops = manager->ops;
+	struct exynos_drm_display *display = exynos_connector->display;
 	unsigned int width, height;
 
 	width = max_width;
@@ -197,8 +154,8 @@ static int exynos_drm_connector_fill_modes(struct drm_connector *connector,
 	 * if specific driver want to find desired_mode using maxmum
 	 * resolution then get max width and height from that driver.
 	 */
-	if (ops && ops->get_max_resol)
-		ops->get_max_resol(manager->dev, &width, &height);
+	if (display->ops->get_max_resol)
+		display->ops->get_max_resol(display, &width, &height);
 
 	return drm_helper_probe_single_connector_modes(connector, width,
 							height);
@@ -210,13 +167,11 @@ exynos_drm_connector_detect(struct drm_connector *connector, bool force)
 {
 	struct exynos_drm_connector *exynos_connector =
 					to_exynos_connector(connector);
-	struct exynos_drm_manager *manager = exynos_connector->manager;
-	struct exynos_drm_display_ops *display_ops =
-					manager->display_ops;
+	struct exynos_drm_display *display = exynos_connector->display;
 	enum drm_connector_status status = connector_status_disconnected;
 
-	if (display_ops && display_ops->is_connected) {
-		if (display_ops->is_connected(manager->dev))
+	if (display->ops->is_connected) {
+		if (display->ops->is_connected(display))
 			status = connector_status_connected;
 		else
 			status = connector_status_disconnected;
@@ -236,7 +191,7 @@ static void exynos_drm_connector_destroy(struct drm_connector *connector)
 }
 
 static struct drm_connector_funcs exynos_connector_funcs = {
-	.dpms		= exynos_drm_connector_dpms,
+	.dpms		= drm_helper_connector_dpms,
 	.fill_modes	= exynos_drm_connector_fill_modes,
 	.detect		= exynos_drm_connector_detect,
 	.destroy	= exynos_drm_connector_destroy,
@@ -246,7 +201,7 @@ struct drm_connector *exynos_drm_connector_create(struct drm_device *dev,
 						   struct drm_encoder *encoder)
 {
 	struct exynos_drm_connector *exynos_connector;
-	struct exynos_drm_manager *manager = exynos_drm_get_manager(encoder);
+	struct exynos_drm_display *display = exynos_drm_get_display(encoder);
 	struct drm_connector *connector;
 	int type;
 	int err;
@@ -257,7 +212,7 @@ struct drm_connector *exynos_drm_connector_create(struct drm_device *dev,
 
 	connector = &exynos_connector->drm_connector;
 
-	switch (manager->display_ops->type) {
+	switch (display->type) {
 	case EXYNOS_DISPLAY_TYPE_HDMI:
 		type = DRM_MODE_CONNECTOR_HDMIA;
 		connector->interlace_allowed = true;
@@ -280,8 +235,7 @@ struct drm_connector *exynos_drm_connector_create(struct drm_device *dev,
 		goto err_connector;
 
 	exynos_connector->encoder_id = encoder->base.id;
-	exynos_connector->manager = manager;
-	exynos_connector->dpms = DRM_MODE_DPMS_OFF;
+	exynos_connector->display = display;
 	connector->dpms = DRM_MODE_DPMS_OFF;
 	connector->encoder = encoder;
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_connector.h b/drivers/gpu/drm/exynos/exynos_drm_connector.h
index 547c6b59035..4eb20d78379 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_connector.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_connector.h
@@ -17,8 +17,4 @@
 struct drm_connector *exynos_drm_connector_create(struct drm_device *dev,
 						   struct drm_encoder *encoder);
 
-struct drm_encoder *exynos_drm_best_encoder(struct drm_connector *connector);
-
-void exynos_drm_display_power(struct drm_connector *connector, int mode);
-
 #endif
diff --git a/drivers/gpu/drm/exynos/exynos_drm_core.c b/drivers/gpu/drm/exynos/exynos_drm_core.c
index 1bef6dc7747..4c9f972eaa0 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_core.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_core.c
@@ -14,43 +14,40 @@
 
 #include <drm/drmP.h>
 #include "exynos_drm_drv.h"
+#include "exynos_drm_crtc.h"
 #include "exynos_drm_encoder.h"
-#include "exynos_drm_connector.h"
 #include "exynos_drm_fbdev.h"
 
 static LIST_HEAD(exynos_drm_subdrv_list);
 
-static int exynos_drm_create_enc_conn(struct drm_device *dev,
-					struct exynos_drm_subdrv *subdrv)
+int exynos_drm_create_enc_conn(struct drm_device *dev,
+					struct exynos_drm_display *display)
 {
 	struct drm_encoder *encoder;
-	struct drm_connector *connector;
 	int ret;
+	unsigned long possible_crtcs = 0;
 
-	subdrv->manager->dev = subdrv->dev;
+	ret = exynos_drm_crtc_get_pipe_from_type(dev, display->type);
+	if (ret < 0)
+		return ret;
+
+	possible_crtcs |= 1 << ret;
 
 	/* create and initialize a encoder for this sub driver. */
-	encoder = exynos_drm_encoder_create(dev, subdrv->manager,
-			(1 << MAX_CRTC) - 1);
+	encoder = exynos_drm_encoder_create(dev, display, possible_crtcs);
 	if (!encoder) {
 		DRM_ERROR("failed to create encoder\n");
 		return -EFAULT;
 	}
 
-	/*
-	 * create and initialize a connector for this sub driver and
-	 * attach the encoder created above to the connector.
-	 */
-	connector = exynos_drm_connector_create(dev, encoder);
-	if (!connector) {
-		DRM_ERROR("failed to create connector\n");
-		ret = -EFAULT;
+	display->encoder = encoder;
+
+	ret = display->ops->create_connector(display, encoder);
+	if (ret) {
+		DRM_ERROR("failed to create connector ret = %d\n", ret);
 		goto err_destroy_encoder;
 	}
 
-	subdrv->encoder = encoder;
-	subdrv->connector = connector;
-
 	return 0;
 
 err_destroy_encoder:
@@ -58,97 +55,59 @@ err_destroy_encoder:
 	return ret;
 }
 
-static void exynos_drm_destroy_enc_conn(struct exynos_drm_subdrv *subdrv)
+int exynos_drm_subdrv_register(struct exynos_drm_subdrv *subdrv)
 {
-	if (subdrv->encoder) {
-		struct drm_encoder *encoder = subdrv->encoder;
-		encoder->funcs->destroy(encoder);
-		subdrv->encoder = NULL;
-	}
-
-	if (subdrv->connector) {
-		struct drm_connector *connector = subdrv->connector;
-		connector->funcs->destroy(connector);
-		subdrv->connector = NULL;
-	}
-}
+	if (!subdrv)
+		return -EINVAL;
 
-static int exynos_drm_subdrv_probe(struct drm_device *dev,
-					struct exynos_drm_subdrv *subdrv)
-{
-	if (subdrv->probe) {
-		int ret;
-
-		subdrv->drm_dev = dev;
-
-		/*
-		 * this probe callback would be called by sub driver
-		 * after setting of all resources to this sub driver,
-		 * such as clock, irq and register map are done or by load()
-		 * of exynos drm driver.
-		 *
-		 * P.S. note that this driver is considered for modularization.
-		 */
-		ret = subdrv->probe(dev, subdrv->dev);
-		if (ret)
-			return ret;
-	}
+	list_add_tail(&subdrv->list, &exynos_drm_subdrv_list);
 
 	return 0;
 }
+EXPORT_SYMBOL_GPL(exynos_drm_subdrv_register);
 
-static void exynos_drm_subdrv_remove(struct drm_device *dev,
-				      struct exynos_drm_subdrv *subdrv)
+int exynos_drm_subdrv_unregister(struct exynos_drm_subdrv *subdrv)
 {
-	if (subdrv->remove)
-		subdrv->remove(dev, subdrv->dev);
+	if (!subdrv)
+		return -EINVAL;
+
+	list_del(&subdrv->list);
+
+	return 0;
 }
+EXPORT_SYMBOL_GPL(exynos_drm_subdrv_unregister);
 
-int exynos_drm_device_register(struct drm_device *dev)
+int exynos_drm_device_subdrv_probe(struct drm_device *dev)
 {
 	struct exynos_drm_subdrv *subdrv, *n;
-	unsigned int fine_cnt = 0;
 	int err;
 
 	if (!dev)
 		return -EINVAL;
 
 	list_for_each_entry_safe(subdrv, n, &exynos_drm_subdrv_list, list) {
-		err = exynos_drm_subdrv_probe(dev, subdrv);
-		if (err) {
-			DRM_DEBUG("exynos drm subdrv probe failed.\n");
-			list_del(&subdrv->list);
-			continue;
-		}
-
-		/*
-		 * if manager is null then it means that this sub driver
-		 * doesn't need encoder and connector.
-		 */
-		if (!subdrv->manager) {
-			fine_cnt++;
-			continue;
-		}
-
-		err = exynos_drm_create_enc_conn(dev, subdrv);
-		if (err) {
-			DRM_DEBUG("failed to create encoder and connector.\n");
-			exynos_drm_subdrv_remove(dev, subdrv);
-			list_del(&subdrv->list);
-			continue;
+		if (subdrv->probe) {
+			subdrv->drm_dev = dev;
+
+			/*
+			 * this probe callback would be called by sub driver
+			 * after setting of all resources to this sub driver,
+			 * such as clock, irq and register map are done.
+			 */
+			err = subdrv->probe(dev, subdrv->dev);
+			if (err) {
+				DRM_DEBUG("exynos drm subdrv probe failed.\n");
+				list_del(&subdrv->list);
+				continue;
+			}
 		}
-
-		fine_cnt++;
 	}
 
-	if (!fine_cnt)
-		return -EINVAL;
-
 	return 0;
 }
-EXPORT_SYMBOL_GPL(exynos_drm_device_register);
+EXPORT_SYMBOL_GPL(exynos_drm_device_subdrv_probe);
 
-int exynos_drm_device_unregister(struct drm_device *dev)
+int exynos_drm_device_subdrv_remove(struct drm_device *dev)
 {
 	struct exynos_drm_subdrv *subdrv;
 
@@ -158,35 +117,13 @@ int exynos_drm_device_unregister(struct drm_device *dev)
 	}
 
 	list_for_each_entry(subdrv, &exynos_drm_subdrv_list, list) {
-		exynos_drm_subdrv_remove(dev, subdrv);
-		exynos_drm_destroy_enc_conn(subdrv);
+		if (subdrv->remove)
+			subdrv->remove(dev, subdrv->dev);
 	}
 
 	return 0;
 }
-EXPORT_SYMBOL_GPL(exynos_drm_device_unregister);
-
-int exynos_drm_subdrv_register(struct exynos_drm_subdrv *subdrv)
-{
-	if (!subdrv)
-		return -EINVAL;
-
-	list_add_tail(&subdrv->list, &exynos_drm_subdrv_list);
-
-	return 0;
-}
-EXPORT_SYMBOL_GPL(exynos_drm_subdrv_register);
-
-int exynos_drm_subdrv_unregister(struct exynos_drm_subdrv *subdrv)
-{
-	if (!subdrv)
-		return -EINVAL;
-
-	list_del(&subdrv->list);
-
-	return 0;
-}
-EXPORT_SYMBOL_GPL(exynos_drm_subdrv_unregister);
+EXPORT_SYMBOL_GPL(exynos_drm_device_subdrv_remove);
 
 int exynos_drm_subdrv_open(struct drm_device *dev, struct drm_file *file)
 {
diff --git a/drivers/gpu/drm/exynos/exynos_drm_crtc.c b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
index 6f3400f3978..95c9435d026 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_crtc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
@@ -33,6 +33,7 @@ enum exynos_crtc_mode {
  *
  * @drm_crtc: crtc object.
  * @drm_plane: pointer of private plane object for this crtc
+ * @manager: the manager associated with this crtc
  * @pipe: a crtc index created at load() with a new crtc object creation
  *	and the crtc object would be set to private->crtc array
  *	to get a crtc object corresponding to this pipe from private->crtc
@@ -46,6 +47,7 @@ enum exynos_crtc_mode {
 struct exynos_drm_crtc {
 	struct drm_crtc			drm_crtc;
 	struct drm_plane		*plane;
+	struct exynos_drm_manager	*manager;
 	unsigned int			pipe;
 	unsigned int			dpms;
 	enum exynos_crtc_mode		mode;
@@ -56,6 +58,7 @@ struct exynos_drm_crtc {
 static void exynos_drm_crtc_dpms(struct drm_crtc *crtc, int mode)
 {
 	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
+	struct exynos_drm_manager *manager = exynos_crtc->manager;
 
 	DRM_DEBUG_KMS("crtc[%d] mode[%d]\n", crtc->base.id, mode);
 
@@ -71,7 +74,9 @@ static void exynos_drm_crtc_dpms(struct drm_crtc *crtc, int mode)
 		drm_vblank_off(crtc->dev, exynos_crtc->pipe);
 	}
 
-	exynos_drm_fn_encoder(crtc, &mode, exynos_drm_encoder_crtc_dpms);
+	if (manager->ops->dpms)
+		manager->ops->dpms(manager, mode);
+
 	exynos_crtc->dpms = mode;
 }
 
@@ -83,9 +88,15 @@ static void exynos_drm_crtc_prepare(struct drm_crtc *crtc)
 static void exynos_drm_crtc_commit(struct drm_crtc *crtc)
 {
 	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
+	struct exynos_drm_manager *manager = exynos_crtc->manager;
 
 	exynos_drm_crtc_dpms(crtc, DRM_MODE_DPMS_ON);
+
 	exynos_plane_commit(exynos_crtc->plane);
+
+	if (manager->ops->commit)
+		manager->ops->commit(manager);
+
 	exynos_plane_dpms(exynos_crtc->plane, DRM_MODE_DPMS_ON);
 }
 
@@ -94,7 +105,12 @@ exynos_drm_crtc_mode_fixup(struct drm_crtc *crtc,
 			    const struct drm_display_mode *mode,
 			    struct drm_display_mode *adjusted_mode)
 {
-	/* drm framework doesn't check NULL */
+	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
+	struct exynos_drm_manager *manager = exynos_crtc->manager;
+
+	if (manager->ops->mode_fixup)
+		return manager->ops->mode_fixup(manager, mode, adjusted_mode);
+
 	return true;
 }
 
@@ -104,10 +120,10 @@ exynos_drm_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode,
 			  struct drm_framebuffer *old_fb)
 {
 	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
+	struct exynos_drm_manager *manager = exynos_crtc->manager;
 	struct drm_plane *plane = exynos_crtc->plane;
 	unsigned int crtc_w;
 	unsigned int crtc_h;
-	int pipe = exynos_crtc->pipe;
 	int ret;
 
 	/*
@@ -116,18 +132,20 @@ exynos_drm_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode,
 	 */
 	memcpy(&crtc->mode, adjusted_mode, sizeof(*adjusted_mode));
 
-	crtc_w = crtc->fb->width - x;
-	crtc_h = crtc->fb->height - y;
+	crtc_w = crtc->primary->fb->width - x;
+	crtc_h = crtc->primary->fb->height - y;
 
-	ret = exynos_plane_mode_set(plane, crtc, crtc->fb, 0, 0, crtc_w, crtc_h,
+	if (manager->ops->mode_set)
+		manager->ops->mode_set(manager, &crtc->mode);
+
+	ret = exynos_plane_mode_set(plane, crtc, crtc->primary->fb, 0, 0, crtc_w, crtc_h,
 				    x, y, crtc_w, crtc_h);
 	if (ret)
 		return ret;
 
 	plane->crtc = crtc;
-	plane->fb = crtc->fb;
-
-	exynos_drm_fn_encoder(crtc, &pipe, exynos_drm_encoder_crtc_pipe);
+	plane->fb = crtc->primary->fb;
+	drm_framebuffer_reference(plane->fb);
 
 	return 0;
 }
@@ -147,10 +165,10 @@ static int exynos_drm_crtc_mode_set_commit(struct drm_crtc *crtc, int x, int y,
 		return -EPERM;
 	}
 
-	crtc_w = crtc->fb->width - x;
-	crtc_h = crtc->fb->height - y;
+	crtc_w = crtc->primary->fb->width - x;
+	crtc_h = crtc->primary->fb->height - y;
 
-	ret = exynos_plane_mode_set(plane, crtc, crtc->fb, 0, 0, crtc_w, crtc_h,
+	ret = exynos_plane_mode_set(plane, crtc, crtc->primary->fb, 0, 0, crtc_w, crtc_h,
 				    x, y, crtc_w, crtc_h);
 	if (ret)
 		return ret;
@@ -168,10 +186,19 @@ static int exynos_drm_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 
 static void exynos_drm_crtc_disable(struct drm_crtc *crtc)
 {
-	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
+	struct drm_plane *plane;
+	int ret;
 
-	exynos_plane_dpms(exynos_crtc->plane, DRM_MODE_DPMS_OFF);
 	exynos_drm_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
+
+	drm_for_each_legacy_plane(plane, &crtc->dev->mode_config.plane_list) {
+		if (plane->crtc != crtc)
+			continue;
+
+		ret = plane->funcs->disable_plane(plane);
+		if (ret)
+			DRM_ERROR("Failed to disable plane %d\n", ret);
+	}
 }
 
 static struct drm_crtc_helper_funcs exynos_crtc_helper_funcs = {
@@ -192,7 +219,7 @@ static int exynos_drm_crtc_page_flip(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct exynos_drm_private *dev_priv = dev->dev_private;
 	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(crtc);
-	struct drm_framebuffer *old_fb = crtc->fb;
+	struct drm_framebuffer *old_fb = crtc->primary->fb;
 	int ret = -EINVAL;
 
 	/* when the page flip is requested, crtc's dpms should be on */
@@ -223,11 +250,11 @@ static int exynos_drm_crtc_page_flip(struct drm_crtc *crtc,
 		atomic_set(&exynos_crtc->pending_flip, 1);
 		spin_unlock_irq(&dev->event_lock);
 
-		crtc->fb = fb;
+		crtc->primary->fb = fb;
 		ret = exynos_drm_crtc_mode_set_commit(crtc, crtc->x, crtc->y,
 						    NULL);
 		if (ret) {
-			crtc->fb = old_fb;
+			crtc->primary->fb = old_fb;
 
 			spin_lock_irq(&dev->event_lock);
 			drm_vblank_put(dev, exynos_crtc->pipe);
@@ -318,31 +345,35 @@ static void exynos_drm_crtc_attach_mode_property(struct drm_crtc *crtc)
 	drm_object_attach_property(&crtc->base, prop, 0);
 }
 
-int exynos_drm_crtc_create(struct drm_device *dev, unsigned int nr)
+int exynos_drm_crtc_create(struct exynos_drm_manager *manager)
 {
 	struct exynos_drm_crtc *exynos_crtc;
-	struct exynos_drm_private *private = dev->dev_private;
+	struct exynos_drm_private *private = manager->drm_dev->dev_private;
 	struct drm_crtc *crtc;
 
 	exynos_crtc = kzalloc(sizeof(*exynos_crtc), GFP_KERNEL);
 	if (!exynos_crtc)
 		return -ENOMEM;
 
-	exynos_crtc->pipe = nr;
-	exynos_crtc->dpms = DRM_MODE_DPMS_OFF;
 	init_waitqueue_head(&exynos_crtc->pending_flip_queue);
 	atomic_set(&exynos_crtc->pending_flip, 0);
-	exynos_crtc->plane = exynos_plane_init(dev, 1 << nr, true);
+
+	exynos_crtc->dpms = DRM_MODE_DPMS_OFF;
+	exynos_crtc->manager = manager;
+	exynos_crtc->pipe = manager->pipe;
+	exynos_crtc->plane = exynos_plane_init(manager->drm_dev,
+				1 << manager->pipe, true);
 	if (!exynos_crtc->plane) {
 		kfree(exynos_crtc);
 		return -ENOMEM;
 	}
 
+	manager->crtc = &exynos_crtc->drm_crtc;
 	crtc = &exynos_crtc->drm_crtc;
 
-	private->crtc[nr] = crtc;
+	private->crtc[manager->pipe] = crtc;
 
-	drm_crtc_init(dev, crtc, &exynos_crtc_funcs);
+	drm_crtc_init(manager->drm_dev, crtc, &exynos_crtc_funcs);
 	drm_crtc_helper_add(crtc, &exynos_crtc_helper_funcs);
 
 	exynos_drm_crtc_attach_mode_property(crtc);
@@ -350,39 +381,41 @@ int exynos_drm_crtc_create(struct drm_device *dev, unsigned int nr)
 	return 0;
 }
 
-int exynos_drm_crtc_enable_vblank(struct drm_device *dev, int crtc)
+int exynos_drm_crtc_enable_vblank(struct drm_device *dev, int pipe)
 {
 	struct exynos_drm_private *private = dev->dev_private;
 	struct exynos_drm_crtc *exynos_crtc =
-		to_exynos_crtc(private->crtc[crtc]);
+		to_exynos_crtc(private->crtc[pipe]);
+	struct exynos_drm_manager *manager = exynos_crtc->manager;
 
 	if (exynos_crtc->dpms != DRM_MODE_DPMS_ON)
 		return -EPERM;
 
-	exynos_drm_fn_encoder(private->crtc[crtc], &crtc,
-			exynos_drm_enable_vblank);
+	if (manager->ops->enable_vblank)
+		manager->ops->enable_vblank(manager);
 
 	return 0;
 }
 
-void exynos_drm_crtc_disable_vblank(struct drm_device *dev, int crtc)
+void exynos_drm_crtc_disable_vblank(struct drm_device *dev, int pipe)
 {
 	struct exynos_drm_private *private = dev->dev_private;
 	struct exynos_drm_crtc *exynos_crtc =
-		to_exynos_crtc(private->crtc[crtc]);
+		to_exynos_crtc(private->crtc[pipe]);
+	struct exynos_drm_manager *manager = exynos_crtc->manager;
 
 	if (exynos_crtc->dpms != DRM_MODE_DPMS_ON)
 		return;
 
-	exynos_drm_fn_encoder(private->crtc[crtc], &crtc,
-			exynos_drm_disable_vblank);
+	if (manager->ops->disable_vblank)
+		manager->ops->disable_vblank(manager);
 }
 
-void exynos_drm_crtc_finish_pageflip(struct drm_device *dev, int crtc)
+void exynos_drm_crtc_finish_pageflip(struct drm_device *dev, int pipe)
 {
 	struct exynos_drm_private *dev_priv = dev->dev_private;
 	struct drm_pending_vblank_event *e, *t;
-	struct drm_crtc *drm_crtc = dev_priv->crtc[crtc];
+	struct drm_crtc *drm_crtc = dev_priv->crtc[pipe];
 	struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(drm_crtc);
 	unsigned long flags;
 
@@ -391,15 +424,87 @@ void exynos_drm_crtc_finish_pageflip(struct drm_device *dev, int crtc)
 	list_for_each_entry_safe(e, t, &dev_priv->pageflip_event_list,
 			base.link) {
 		/* if event's pipe isn't same as crtc then ignore it. */
-		if (crtc != e->pipe)
+		if (pipe != e->pipe)
 			continue;
 
 		list_del(&e->base.link);
 		drm_send_vblank_event(dev, -1, e);
-		drm_vblank_put(dev, crtc);
+		drm_vblank_put(dev, pipe);
 		atomic_set(&exynos_crtc->pending_flip, 0);
 		wake_up(&exynos_crtc->pending_flip_queue);
 	}
 
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 }
+
+void exynos_drm_crtc_plane_mode_set(struct drm_crtc *crtc,
+			struct exynos_drm_overlay *overlay)
+{
+	struct exynos_drm_manager *manager = to_exynos_crtc(crtc)->manager;
+
+	if (manager->ops->win_mode_set)
+		manager->ops->win_mode_set(manager, overlay);
+}
+
+void exynos_drm_crtc_plane_commit(struct drm_crtc *crtc, int zpos)
+{
+	struct exynos_drm_manager *manager = to_exynos_crtc(crtc)->manager;
+
+	if (manager->ops->win_commit)
+		manager->ops->win_commit(manager, zpos);
+}
+
+void exynos_drm_crtc_plane_enable(struct drm_crtc *crtc, int zpos)
+{
+	struct exynos_drm_manager *manager = to_exynos_crtc(crtc)->manager;
+
+	if (manager->ops->win_enable)
+		manager->ops->win_enable(manager, zpos);
+}
+
+void exynos_drm_crtc_plane_disable(struct drm_crtc *crtc, int zpos)
+{
+	struct exynos_drm_manager *manager = to_exynos_crtc(crtc)->manager;
+
+	if (manager->ops->win_disable)
+		manager->ops->win_disable(manager, zpos);
+}
+
+void exynos_drm_crtc_complete_scanout(struct drm_framebuffer *fb)
+{
+	struct exynos_drm_manager *manager;
+	struct drm_device *dev = fb->dev;
+	struct drm_crtc *crtc;
+
+	/*
+	 * make sure that overlay data are updated to real hardware
+	 * for all encoders.
+	 */
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+		manager = to_exynos_crtc(crtc)->manager;
+
+		/*
+		 * wait for vblank interrupt
+		 * - this makes sure that overlay data are updated to
+		 *	real hardware.
+		 */
+		if (manager->ops->wait_for_vblank)
+			manager->ops->wait_for_vblank(manager);
+	}
+}
+
+int exynos_drm_crtc_get_pipe_from_type(struct drm_device *drm_dev,
+					unsigned int out_type)
+{
+	struct drm_crtc *crtc;
+
+	list_for_each_entry(crtc, &drm_dev->mode_config.crtc_list, head) {
+		struct exynos_drm_crtc *exynos_crtc;
+
+		exynos_crtc = to_exynos_crtc(crtc);
+		if (exynos_crtc->manager->type == out_type)
+			return exynos_crtc->manager->pipe;
+	}
+
+	return -EPERM;
+}
diff --git a/drivers/gpu/drm/exynos/exynos_drm_crtc.h b/drivers/gpu/drm/exynos/exynos_drm_crtc.h
index 3e197e6ae7d..9f74b10a8a0 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_crtc.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_crtc.h
@@ -15,9 +15,25 @@
 #ifndef _EXYNOS_DRM_CRTC_H_
 #define _EXYNOS_DRM_CRTC_H_
 
-int exynos_drm_crtc_create(struct drm_device *dev, unsigned int nr);
-int exynos_drm_crtc_enable_vblank(struct drm_device *dev, int crtc);
-void exynos_drm_crtc_disable_vblank(struct drm_device *dev, int crtc);
-void exynos_drm_crtc_finish_pageflip(struct drm_device *dev, int crtc);
+struct drm_device;
+struct drm_crtc;
+struct exynos_drm_manager;
+struct exynos_drm_overlay;
+
+int exynos_drm_crtc_create(struct exynos_drm_manager *manager);
+int exynos_drm_crtc_enable_vblank(struct drm_device *dev, int pipe);
+void exynos_drm_crtc_disable_vblank(struct drm_device *dev, int pipe);
+void exynos_drm_crtc_finish_pageflip(struct drm_device *dev, int pipe);
+void exynos_drm_crtc_complete_scanout(struct drm_framebuffer *fb);
+
+void exynos_drm_crtc_plane_mode_set(struct drm_crtc *crtc,
+			struct exynos_drm_overlay *overlay);
+void exynos_drm_crtc_plane_commit(struct drm_crtc *crtc, int zpos);
+void exynos_drm_crtc_plane_enable(struct drm_crtc *crtc, int zpos);
+void exynos_drm_crtc_plane_disable(struct drm_crtc *crtc, int zpos);
+
+/* This function gets pipe value to crtc device matched with out_type. */
+int exynos_drm_crtc_get_pipe_from_type(struct drm_device *drm_dev,
+					unsigned int out_type);
 
 #endif
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dmabuf.c b/drivers/gpu/drm/exynos/exynos_drm_dmabuf.c
index 59827cc5e77..2a3ad24276f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dmabuf.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dmabuf.c
@@ -224,7 +224,7 @@ struct drm_gem_object *exynos_dmabuf_prime_import(struct drm_device *drm_dev,
 	get_dma_buf(dma_buf);
 
 	sgt = dma_buf_map_attachment(attach, DMA_BIDIRECTIONAL);
-	if (IS_ERR_OR_NULL(sgt)) {
+	if (IS_ERR(sgt)) {
 		ret = PTR_ERR(sgt);
 		goto err_buf_detach;
 	}
@@ -263,7 +263,7 @@ struct drm_gem_object *exynos_dmabuf_prime_import(struct drm_device *drm_dev,
 	buffer->sgt = sgt;
 	exynos_gem_obj->base.import_attach = attach;
 
-	DRM_DEBUG_PRIME("dma_addr = 0x%x, size = 0x%lx\n", buffer->dma_addr,
+	DRM_DEBUG_PRIME("dma_addr = %pad, size = 0x%lx\n", &buffer->dma_addr,
 								buffer->size);
 
 	return &exynos_gem_obj->base;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dpi.c b/drivers/gpu/drm/exynos/exynos_drm_dpi.c
new file mode 100644
index 00000000000..9e530f205ad
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_drm_dpi.c
@@ -0,0 +1,347 @@
+/*
+ * Exynos DRM Parallel output support.
+ *
+ * Copyright (c) 2014 Samsung Electronics Co., Ltd
+ *
+ * Contacts: Andrzej Hajda <a.hajda@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+*/
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_panel.h>
+
+#include <linux/regulator/consumer.h>
+
+#include <video/of_videomode.h>
+#include <video/videomode.h>
+
+#include "exynos_drm_drv.h"
+
+struct exynos_dpi {
+	struct device *dev;
+	struct device_node *panel_node;
+
+	struct drm_panel *panel;
+	struct drm_connector connector;
+	struct drm_encoder *encoder;
+
+	struct videomode *vm;
+	int dpms_mode;
+};
+
+#define connector_to_dpi(c) container_of(c, struct exynos_dpi, connector)
+
+static enum drm_connector_status
+exynos_dpi_detect(struct drm_connector *connector, bool force)
+{
+	struct exynos_dpi *ctx = connector_to_dpi(connector);
+
+	if (ctx->panel && !ctx->panel->connector)
+		drm_panel_attach(ctx->panel, &ctx->connector);
+
+	return connector_status_connected;
+}
+
+static void exynos_dpi_connector_destroy(struct drm_connector *connector)
+{
+	drm_sysfs_connector_remove(connector);
+	drm_connector_cleanup(connector);
+}
+
+static struct drm_connector_funcs exynos_dpi_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.detect = exynos_dpi_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = exynos_dpi_connector_destroy,
+};
+
+static int exynos_dpi_get_modes(struct drm_connector *connector)
+{
+	struct exynos_dpi *ctx = connector_to_dpi(connector);
+
+	/* fimd timings gets precedence over panel modes */
+	if (ctx->vm) {
+		struct drm_display_mode *mode;
+
+		mode = drm_mode_create(connector->dev);
+		if (!mode) {
+			DRM_ERROR("failed to create a new display mode\n");
+			return 0;
+		}
+		drm_display_mode_from_videomode(ctx->vm, mode);
+		mode->type = DRM_MODE_TYPE_DRIVER | DRM_MODE_TYPE_PREFERRED;
+		drm_mode_probed_add(connector, mode);
+		return 1;
+	}
+
+	if (ctx->panel)
+		return ctx->panel->funcs->get_modes(ctx->panel);
+
+	return 0;
+}
+
+static struct drm_encoder *
+exynos_dpi_best_encoder(struct drm_connector *connector)
+{
+	struct exynos_dpi *ctx = connector_to_dpi(connector);
+
+	return ctx->encoder;
+}
+
+static struct drm_connector_helper_funcs exynos_dpi_connector_helper_funcs = {
+	.get_modes = exynos_dpi_get_modes,
+	.best_encoder = exynos_dpi_best_encoder,
+};
+
+static int exynos_dpi_create_connector(struct exynos_drm_display *display,
+				       struct drm_encoder *encoder)
+{
+	struct exynos_dpi *ctx = display->ctx;
+	struct drm_connector *connector = &ctx->connector;
+	int ret;
+
+	ctx->encoder = encoder;
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	ret = drm_connector_init(encoder->dev, connector,
+				 &exynos_dpi_connector_funcs,
+				 DRM_MODE_CONNECTOR_VGA);
+	if (ret) {
+		DRM_ERROR("failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	drm_connector_helper_add(connector, &exynos_dpi_connector_helper_funcs);
+	drm_sysfs_connector_add(connector);
+	drm_mode_connector_attach_encoder(connector, encoder);
+
+	return 0;
+}
+
+static void exynos_dpi_poweron(struct exynos_dpi *ctx)
+{
+	if (ctx->panel)
+		drm_panel_enable(ctx->panel);
+}
+
+static void exynos_dpi_poweroff(struct exynos_dpi *ctx)
+{
+	if (ctx->panel)
+		drm_panel_disable(ctx->panel);
+}
+
+static void exynos_dpi_dpms(struct exynos_drm_display *display, int mode)
+{
+	struct exynos_dpi *ctx = display->ctx;
+
+	switch (mode) {
+	case DRM_MODE_DPMS_ON:
+		if (ctx->dpms_mode != DRM_MODE_DPMS_ON)
+				exynos_dpi_poweron(ctx);
+			break;
+	case DRM_MODE_DPMS_STANDBY:
+	case DRM_MODE_DPMS_SUSPEND:
+	case DRM_MODE_DPMS_OFF:
+		if (ctx->dpms_mode == DRM_MODE_DPMS_ON)
+			exynos_dpi_poweroff(ctx);
+		break;
+	default:
+		break;
+	}
+	ctx->dpms_mode = mode;
+}
+
+static struct exynos_drm_display_ops exynos_dpi_display_ops = {
+	.create_connector = exynos_dpi_create_connector,
+	.dpms = exynos_dpi_dpms
+};
+
+static struct exynos_drm_display exynos_dpi_display = {
+	.type = EXYNOS_DISPLAY_TYPE_LCD,
+	.ops = &exynos_dpi_display_ops,
+};
+
+/* of_* functions will be removed after merge of of_graph patches */
+static struct device_node *
+of_get_child_by_name_reg(struct device_node *parent, const char *name, u32 reg)
+{
+	struct device_node *np;
+
+	for_each_child_of_node(parent, np) {
+		u32 r;
+
+		if (!np->name || of_node_cmp(np->name, name))
+			continue;
+
+		if (of_property_read_u32(np, "reg", &r) < 0)
+			r = 0;
+
+		if (reg == r)
+			break;
+	}
+
+	return np;
+}
+
+static struct device_node *of_graph_get_port_by_reg(struct device_node *parent,
+						    u32 reg)
+{
+	struct device_node *ports, *port;
+
+	ports = of_get_child_by_name(parent, "ports");
+	if (ports)
+		parent = ports;
+
+	port = of_get_child_by_name_reg(parent, "port", reg);
+
+	of_node_put(ports);
+
+	return port;
+}
+
+static struct device_node *
+of_graph_get_endpoint_by_reg(struct device_node *port, u32 reg)
+{
+	return of_get_child_by_name_reg(port, "endpoint", reg);
+}
+
+static struct device_node *
+of_graph_get_remote_port_parent(const struct device_node *node)
+{
+	struct device_node *np;
+	unsigned int depth;
+
+	np = of_parse_phandle(node, "remote-endpoint", 0);
+
+	/* Walk 3 levels up only if there is 'ports' node. */
+	for (depth = 3; depth && np; depth--) {
+		np = of_get_next_parent(np);
+		if (depth == 2 && of_node_cmp(np->name, "ports"))
+			break;
+	}
+	return np;
+}
+
+enum {
+	FIMD_PORT_IN0,
+	FIMD_PORT_IN1,
+	FIMD_PORT_IN2,
+	FIMD_PORT_RGB,
+	FIMD_PORT_WRB,
+};
+
+static struct device_node *exynos_dpi_of_find_panel_node(struct device *dev)
+{
+	struct device_node *np, *ep;
+
+	np = of_graph_get_port_by_reg(dev->of_node, FIMD_PORT_RGB);
+	if (!np)
+		return NULL;
+
+	ep = of_graph_get_endpoint_by_reg(np, 0);
+	of_node_put(np);
+	if (!ep)
+		return NULL;
+
+	np = of_graph_get_remote_port_parent(ep);
+	of_node_put(ep);
+
+	return np;
+}
+
+static int exynos_dpi_parse_dt(struct exynos_dpi *ctx)
+{
+	struct device *dev = ctx->dev;
+	struct device_node *dn = dev->of_node;
+	struct device_node *np;
+
+	ctx->panel_node = exynos_dpi_of_find_panel_node(dev);
+
+	np = of_get_child_by_name(dn, "display-timings");
+	if (np) {
+		struct videomode *vm;
+		int ret;
+
+		of_node_put(np);
+
+		vm = devm_kzalloc(dev, sizeof(*ctx->vm), GFP_KERNEL);
+		if (!vm)
+			return -ENOMEM;
+
+		ret = of_get_videomode(dn, vm, 0);
+		if (ret < 0) {
+			devm_kfree(dev, vm);
+			return ret;
+		}
+
+		ctx->vm = vm;
+
+		return 0;
+	}
+
+	if (!ctx->panel_node)
+		return -EINVAL;
+
+	return 0;
+}
+
+struct exynos_drm_display *exynos_dpi_probe(struct device *dev)
+{
+	struct exynos_dpi *ctx;
+	int ret;
+
+	ret = exynos_drm_component_add(dev,
+					EXYNOS_DEVICE_TYPE_CONNECTOR,
+					exynos_dpi_display.type);
+	if (ret)
+		return ERR_PTR(ret);
+
+	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
+	if (!ctx)
+		goto err_del_component;
+
+	ctx->dev = dev;
+	exynos_dpi_display.ctx = ctx;
+	ctx->dpms_mode = DRM_MODE_DPMS_OFF;
+
+	ret = exynos_dpi_parse_dt(ctx);
+	if (ret < 0) {
+		devm_kfree(dev, ctx);
+		goto err_del_component;
+	}
+
+	if (ctx->panel_node) {
+		ctx->panel = of_drm_find_panel(ctx->panel_node);
+		if (!ctx->panel) {
+			exynos_drm_component_del(dev,
+						EXYNOS_DEVICE_TYPE_CONNECTOR);
+			return ERR_PTR(-EPROBE_DEFER);
+		}
+	}
+
+	return &exynos_dpi_display;
+
+err_del_component:
+	exynos_drm_component_del(dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
+
+	return NULL;
+}
+
+int exynos_dpi_remove(struct device *dev)
+{
+	struct drm_encoder *encoder = exynos_dpi_display.encoder;
+	struct exynos_dpi *ctx = exynos_dpi_display.ctx;
+
+	exynos_dpi_dpms(&exynos_dpi_display, DRM_MODE_DPMS_OFF);
+	encoder->funcs->destroy(encoder);
+	drm_connector_cleanup(&ctx->connector);
+
+	exynos_drm_component_del(dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index 22b8f5eced8..ab7d182063c 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -11,9 +11,13 @@
  * option) any later version.
  */
 
+#include <linux/pm_runtime.h>
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 
+#include <linux/anon_inodes.h>
+#include <linux/component.h>
+
 #include <drm/exynos_drm.h>
 
 #include "exynos_drm_drv.h"
@@ -37,9 +41,19 @@
 
 #define VBLANK_OFF_DELAY	50000
 
-/* platform device pointer for eynos drm device. */
 static struct platform_device *exynos_drm_pdev;
 
+static DEFINE_MUTEX(drm_component_lock);
+static LIST_HEAD(drm_component_list);
+
+struct component_dev {
+	struct list_head list;
+	struct device *crtc_dev;
+	struct device *conn_dev;
+	enum exynos_drm_output_type out_type;
+	unsigned int dev_type_flag;
+};
+
 static int exynos_drm_load(struct drm_device *dev, unsigned long flags)
 {
 	struct exynos_drm_private *private;
@@ -51,6 +65,7 @@ static int exynos_drm_load(struct drm_device *dev, unsigned long flags)
 		return -ENOMEM;
 
 	INIT_LIST_HEAD(&private->pageflip_event_list);
+	dev_set_drvdata(dev->dev, dev);
 	dev->dev_private = (void *)private;
 
 	/*
@@ -62,73 +77,59 @@ static int exynos_drm_load(struct drm_device *dev, unsigned long flags)
 	ret = drm_create_iommu_mapping(dev);
 	if (ret < 0) {
 		DRM_ERROR("failed to create iommu mapping.\n");
-		goto err_crtc;
+		goto err_free_private;
 	}
 
 	drm_mode_config_init(dev);
 
-	/* init kms poll for handling hpd */
-	drm_kms_helper_poll_init(dev);
-
 	exynos_drm_mode_config_init(dev);
 
-	/*
-	 * EXYNOS4 is enough to have two CRTCs and each crtc would be used
-	 * without dependency of hardware.
-	 */
-	for (nr = 0; nr < MAX_CRTC; nr++) {
-		ret = exynos_drm_crtc_create(dev, nr);
-		if (ret)
-			goto err_release_iommu_mapping;
-	}
-
 	for (nr = 0; nr < MAX_PLANE; nr++) {
 		struct drm_plane *plane;
-		unsigned int possible_crtcs = (1 << MAX_CRTC) - 1;
+		unsigned long possible_crtcs = (1 << MAX_CRTC) - 1;
 
 		plane = exynos_plane_init(dev, possible_crtcs, false);
 		if (!plane)
-			goto err_release_iommu_mapping;
+			goto err_mode_config_cleanup;
 	}
 
-	ret = drm_vblank_init(dev, MAX_CRTC);
-	if (ret)
-		goto err_release_iommu_mapping;
+	/* init kms poll for handling hpd */
+	drm_kms_helper_poll_init(dev);
 
-	/*
-	 * probe sub drivers such as display controller and hdmi driver,
-	 * that were registered at probe() of platform driver
-	 * to the sub driver and create encoder and connector for them.
-	 */
-	ret = exynos_drm_device_register(dev);
+	ret = drm_vblank_init(dev, MAX_CRTC);
 	if (ret)
-		goto err_vblank;
+		goto err_mode_config_cleanup;
 
 	/* setup possible_clones. */
 	exynos_drm_encoder_setup(dev);
 
-	/*
-	 * create and configure fb helper and also exynos specific
-	 * fbdev object.
-	 */
-	ret = exynos_drm_fbdev_init(dev);
-	if (ret) {
-		DRM_ERROR("failed to initialize drm fbdev\n");
-		goto err_drm_device;
-	}
-
 	drm_vblank_offdelay = VBLANK_OFF_DELAY;
 
+	platform_set_drvdata(dev->platformdev, dev);
+
+	/* Try to bind all sub drivers. */
+	ret = component_bind_all(dev->dev, dev);
+	if (ret)
+		goto err_cleanup_vblank;
+
+	/* Probe non kms sub drivers and virtual display driver. */
+	ret = exynos_drm_device_subdrv_probe(dev);
+	if (ret)
+		goto err_unbind_all;
+
+	/* force connectors detection */
+	drm_helper_hpd_irq_event(dev);
+
 	return 0;
 
-err_drm_device:
-	exynos_drm_device_unregister(dev);
-err_vblank:
+err_unbind_all:
+	component_unbind_all(dev->dev, dev);
+err_cleanup_vblank:
 	drm_vblank_cleanup(dev);
-err_release_iommu_mapping:
-	drm_release_iommu_mapping(dev);
-err_crtc:
+err_mode_config_cleanup:
 	drm_mode_config_cleanup(dev);
+	drm_release_iommu_mapping(dev);
+err_free_private:
 	kfree(private);
 
 	return ret;
@@ -136,8 +137,9 @@ err_crtc:
 
 static int exynos_drm_unload(struct drm_device *dev)
 {
+	exynos_drm_device_subdrv_remove(dev);
+
 	exynos_drm_fbdev_fini(dev);
-	exynos_drm_device_unregister(dev);
 	drm_vblank_cleanup(dev);
 	drm_kms_helper_poll_fini(dev);
 	drm_mode_config_cleanup(dev);
@@ -145,14 +147,55 @@ static int exynos_drm_unload(struct drm_device *dev)
 	drm_release_iommu_mapping(dev);
 	kfree(dev->dev_private);
 
+	component_unbind_all(dev->dev, dev);
 	dev->dev_private = NULL;
 
 	return 0;
 }
 
+static const struct file_operations exynos_drm_gem_fops = {
+	.mmap = exynos_drm_gem_mmap_buffer,
+};
+
+static int exynos_drm_suspend(struct drm_device *dev, pm_message_t state)
+{
+	struct drm_connector *connector;
+
+	drm_modeset_lock_all(dev);
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		int old_dpms = connector->dpms;
+
+		if (connector->funcs->dpms)
+			connector->funcs->dpms(connector, DRM_MODE_DPMS_OFF);
+
+		/* Set the old mode back to the connector for resume */
+		connector->dpms = old_dpms;
+	}
+	drm_modeset_unlock_all(dev);
+
+	return 0;
+}
+
+static int exynos_drm_resume(struct drm_device *dev)
+{
+	struct drm_connector *connector;
+
+	drm_modeset_lock_all(dev);
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		if (connector->funcs->dpms)
+			connector->funcs->dpms(connector, connector->dpms);
+	}
+	drm_modeset_unlock_all(dev);
+
+	drm_helper_resume_force_mode(dev);
+
+	return 0;
+}
+
 static int exynos_drm_open(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_exynos_file_private *file_priv;
+	struct file *anon_filp;
 	int ret;
 
 	file_priv = kzalloc(sizeof(*file_priv), GFP_KERNEL);
@@ -162,11 +205,27 @@ static int exynos_drm_open(struct drm_device *dev, struct drm_file *file)
 	file->driver_priv = file_priv;
 
 	ret = exynos_drm_subdrv_open(dev, file);
-	if (ret) {
-		kfree(file_priv);
-		file->driver_priv = NULL;
+	if (ret)
+		goto err_file_priv_free;
+
+	anon_filp = anon_inode_getfile("exynos_gem", &exynos_drm_gem_fops,
+					NULL, 0);
+	if (IS_ERR(anon_filp)) {
+		ret = PTR_ERR(anon_filp);
+		goto err_subdrv_close;
 	}
 
+	anon_filp->f_mode = FMODE_READ | FMODE_WRITE;
+	file_priv->anon_filp = anon_filp;
+
+	return ret;
+
+err_subdrv_close:
+	exynos_drm_subdrv_close(dev, file);
+
+err_file_priv_free:
+	kfree(file_priv);
+	file->driver_priv = NULL;
 	return ret;
 }
 
@@ -179,6 +238,7 @@ static void exynos_drm_preclose(struct drm_device *dev,
 static void exynos_drm_postclose(struct drm_device *dev, struct drm_file *file)
 {
 	struct exynos_drm_private *private = dev->dev_private;
+	struct drm_exynos_file_private *file_priv;
 	struct drm_pending_vblank_event *v, *vt;
 	struct drm_pending_event *e, *et;
 	unsigned long flags;
@@ -204,6 +264,9 @@ static void exynos_drm_postclose(struct drm_device *dev, struct drm_file *file)
 	}
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
+	file_priv = file->driver_priv;
+	if (file_priv->anon_filp)
+		fput(file_priv->anon_filp);
 
 	kfree(file->driver_priv);
 	file->driver_priv = NULL;
@@ -262,10 +325,11 @@ static const struct file_operations exynos_drm_driver_fops = {
 };
 
 static struct drm_driver exynos_drm_driver = {
-	.driver_features	= DRIVER_HAVE_IRQ | DRIVER_MODESET |
-					DRIVER_GEM | DRIVER_PRIME,
+	.driver_features	= DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME,
 	.load			= exynos_drm_load,
 	.unload			= exynos_drm_unload,
+	.suspend		= exynos_drm_suspend,
+	.resume			= exynos_drm_resume,
 	.open			= exynos_drm_open,
 	.preclose		= exynos_drm_preclose,
 	.lastclose		= exynos_drm_lastclose,
@@ -292,172 +356,343 @@ static struct drm_driver exynos_drm_driver = {
 	.minor	= DRIVER_MINOR,
 };
 
-static int exynos_drm_platform_probe(struct platform_device *pdev)
+#ifdef CONFIG_PM_SLEEP
+static int exynos_drm_sys_suspend(struct device *dev)
 {
-	int ret;
+	struct drm_device *drm_dev = dev_get_drvdata(dev);
+	pm_message_t message;
 
-	ret = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32));
-	if (ret)
-		return ret;
+	if (pm_runtime_suspended(dev))
+		return 0;
 
-	return drm_platform_init(&exynos_drm_driver, pdev);
+	message.event = PM_EVENT_SUSPEND;
+	return exynos_drm_suspend(drm_dev, message);
 }
 
-static int exynos_drm_platform_remove(struct platform_device *pdev)
+static int exynos_drm_sys_resume(struct device *dev)
+{
+	struct drm_device *drm_dev = dev_get_drvdata(dev);
+
+	if (pm_runtime_suspended(dev))
+		return 0;
+
+	return exynos_drm_resume(drm_dev);
+}
+#endif
+
+static const struct dev_pm_ops exynos_drm_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(exynos_drm_sys_suspend, exynos_drm_sys_resume)
+};
+
+int exynos_drm_component_add(struct device *dev,
+				enum exynos_drm_device_type dev_type,
+				enum exynos_drm_output_type out_type)
 {
-	drm_platform_exit(&exynos_drm_driver, pdev);
+	struct component_dev *cdev;
+
+	if (dev_type != EXYNOS_DEVICE_TYPE_CRTC &&
+			dev_type != EXYNOS_DEVICE_TYPE_CONNECTOR) {
+		DRM_ERROR("invalid device type.\n");
+		return -EINVAL;
+	}
+
+	mutex_lock(&drm_component_lock);
+
+	/*
+	 * Make sure to check if there is a component which has two device
+	 * objects, for connector and for encoder/connector.
+	 * It should make sure that crtc and encoder/connector drivers are
+	 * ready before exynos drm core binds them.
+	 */
+	list_for_each_entry(cdev, &drm_component_list, list) {
+		if (cdev->out_type == out_type) {
+			/*
+			 * If crtc and encoder/connector device objects are
+			 * added already just return.
+			 */
+			if (cdev->dev_type_flag == (EXYNOS_DEVICE_TYPE_CRTC |
+						EXYNOS_DEVICE_TYPE_CONNECTOR)) {
+				mutex_unlock(&drm_component_lock);
+				return 0;
+			}
+
+			if (dev_type == EXYNOS_DEVICE_TYPE_CRTC) {
+				cdev->crtc_dev = dev;
+				cdev->dev_type_flag |= dev_type;
+			}
+
+			if (dev_type == EXYNOS_DEVICE_TYPE_CONNECTOR) {
+				cdev->conn_dev = dev;
+				cdev->dev_type_flag |= dev_type;
+			}
+
+			mutex_unlock(&drm_component_lock);
+			return 0;
+		}
+	}
+
+	mutex_unlock(&drm_component_lock);
+
+	cdev = kzalloc(sizeof(*cdev), GFP_KERNEL);
+	if (!cdev)
+		return -ENOMEM;
+
+	if (dev_type == EXYNOS_DEVICE_TYPE_CRTC)
+		cdev->crtc_dev = dev;
+	if (dev_type == EXYNOS_DEVICE_TYPE_CONNECTOR)
+		cdev->conn_dev = dev;
+
+	cdev->out_type = out_type;
+	cdev->dev_type_flag = dev_type;
+
+	mutex_lock(&drm_component_lock);
+	list_add_tail(&cdev->list, &drm_component_list);
+	mutex_unlock(&drm_component_lock);
 
 	return 0;
 }
 
-static struct platform_driver exynos_drm_platform_driver = {
-	.probe		= exynos_drm_platform_probe,
-	.remove		= exynos_drm_platform_remove,
-	.driver		= {
-		.owner	= THIS_MODULE,
-		.name	= "exynos-drm",
-	},
+void exynos_drm_component_del(struct device *dev,
+				enum exynos_drm_device_type dev_type)
+{
+	struct component_dev *cdev, *next;
+
+	mutex_lock(&drm_component_lock);
+
+	list_for_each_entry_safe(cdev, next, &drm_component_list, list) {
+		if (dev_type == EXYNOS_DEVICE_TYPE_CRTC) {
+			if (cdev->crtc_dev == dev) {
+				cdev->crtc_dev = NULL;
+				cdev->dev_type_flag &= ~dev_type;
+			}
+		}
+
+		if (dev_type == EXYNOS_DEVICE_TYPE_CONNECTOR) {
+			if (cdev->conn_dev == dev) {
+				cdev->conn_dev = NULL;
+				cdev->dev_type_flag &= ~dev_type;
+			}
+		}
+
+		/*
+		 * Release cdev object only in case that both of crtc and
+		 * encoder/connector device objects are NULL.
+		 */
+		if (!cdev->crtc_dev && !cdev->conn_dev) {
+			list_del(&cdev->list);
+			kfree(cdev);
+		}
+
+		break;
+	}
+
+	mutex_unlock(&drm_component_lock);
+}
+
+static int compare_of(struct device *dev, void *data)
+{
+	return dev == (struct device *)data;
+}
+
+static int exynos_drm_add_components(struct device *dev, struct master *m)
+{
+	struct component_dev *cdev;
+	unsigned int attach_cnt = 0;
+
+	mutex_lock(&drm_component_lock);
+
+	list_for_each_entry(cdev, &drm_component_list, list) {
+		int ret;
+
+		/*
+		 * Add components to master only in case that crtc and
+		 * encoder/connector device objects exist.
+		 */
+		if (!cdev->crtc_dev || !cdev->conn_dev)
+			continue;
+
+		attach_cnt++;
+
+		mutex_unlock(&drm_component_lock);
+
+		/*
+		 * fimd and dpi modules have same device object so add
+		 * only crtc device object in this case.
+		 *
+		 * TODO. if dpi module follows driver-model driver then
+		 * below codes can be removed.
+		 */
+		if (cdev->crtc_dev == cdev->conn_dev) {
+			ret = component_master_add_child(m, compare_of,
+					cdev->crtc_dev);
+			if (ret < 0)
+				return ret;
+
+			goto out_lock;
+		}
+
+		/*
+		 * Do not chage below call order.
+		 * crtc device first should be added to master because
+		 * connector/encoder need pipe number of crtc when they
+		 * are created.
+		 */
+		ret = component_master_add_child(m, compare_of, cdev->crtc_dev);
+		ret |= component_master_add_child(m, compare_of,
+							cdev->conn_dev);
+		if (ret < 0)
+			return ret;
+
+out_lock:
+		mutex_lock(&drm_component_lock);
+	}
+
+	mutex_unlock(&drm_component_lock);
+
+	return attach_cnt ? 0 : -ENODEV;
+}
+
+static int exynos_drm_bind(struct device *dev)
+{
+	return drm_platform_init(&exynos_drm_driver, to_platform_device(dev));
+}
+
+static void exynos_drm_unbind(struct device *dev)
+{
+	drm_put_dev(dev_get_drvdata(dev));
+}
+
+static const struct component_master_ops exynos_drm_ops = {
+	.add_components = exynos_drm_add_components,
+	.bind		= exynos_drm_bind,
+	.unbind		= exynos_drm_unbind,
 };
 
-static int __init exynos_drm_init(void)
+static int exynos_drm_platform_probe(struct platform_device *pdev)
 {
 	int ret;
 
+	pdev->dev.coherent_dma_mask = DMA_BIT_MASK(32);
+	exynos_drm_driver.num_ioctls = ARRAY_SIZE(exynos_ioctls);
+
 #ifdef CONFIG_DRM_EXYNOS_FIMD
 	ret = platform_driver_register(&fimd_driver);
 	if (ret < 0)
-		goto out_fimd;
+		return ret;
 #endif
 
-#ifdef CONFIG_DRM_EXYNOS_HDMI
-	ret = platform_driver_register(&hdmi_driver);
-	if (ret < 0)
-		goto out_hdmi;
-	ret = platform_driver_register(&mixer_driver);
-	if (ret < 0)
-		goto out_mixer;
-	ret = platform_driver_register(&exynos_drm_common_hdmi_driver);
+#ifdef CONFIG_DRM_EXYNOS_DP
+	ret = platform_driver_register(&dp_driver);
 	if (ret < 0)
-		goto out_common_hdmi;
+		goto err_unregister_fimd_drv;
+#endif
 
-	ret = exynos_platform_device_hdmi_register();
+#ifdef CONFIG_DRM_EXYNOS_DSI
+	ret = platform_driver_register(&dsi_driver);
 	if (ret < 0)
-		goto out_common_hdmi_dev;
+		goto err_unregister_dp_drv;
 #endif
 
-#ifdef CONFIG_DRM_EXYNOS_VIDI
-	ret = platform_driver_register(&vidi_driver);
+#ifdef CONFIG_DRM_EXYNOS_HDMI
+	ret = platform_driver_register(&mixer_driver);
 	if (ret < 0)
-		goto out_vidi;
+		goto err_unregister_dsi_drv;
+	ret = platform_driver_register(&hdmi_driver);
+	if (ret < 0)
+		goto err_unregister_mixer_drv;
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_G2D
 	ret = platform_driver_register(&g2d_driver);
 	if (ret < 0)
-		goto out_g2d;
+		goto err_unregister_hdmi_drv;
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_FIMC
 	ret = platform_driver_register(&fimc_driver);
 	if (ret < 0)
-		goto out_fimc;
+		goto err_unregister_g2d_drv;
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_ROTATOR
 	ret = platform_driver_register(&rotator_driver);
 	if (ret < 0)
-		goto out_rotator;
+		goto err_unregister_fimc_drv;
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_GSC
 	ret = platform_driver_register(&gsc_driver);
 	if (ret < 0)
-		goto out_gsc;
+		goto err_unregister_rotator_drv;
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_IPP
 	ret = platform_driver_register(&ipp_driver);
 	if (ret < 0)
-		goto out_ipp;
+		goto err_unregister_gsc_drv;
 
 	ret = exynos_platform_device_ipp_register();
 	if (ret < 0)
-		goto out_ipp_dev;
+		goto err_unregister_ipp_drv;
 #endif
 
-	ret = platform_driver_register(&exynos_drm_platform_driver);
+	ret = component_master_add(&pdev->dev, &exynos_drm_ops);
 	if (ret < 0)
-		goto out_drm;
-
-	exynos_drm_pdev = platform_device_register_simple("exynos-drm", -1,
-				NULL, 0);
-	if (IS_ERR(exynos_drm_pdev)) {
-		ret = PTR_ERR(exynos_drm_pdev);
-		goto out;
-	}
+		DRM_DEBUG_KMS("re-tried by last sub driver probed later.\n");
 
 	return 0;
 
-out:
-	platform_driver_unregister(&exynos_drm_platform_driver);
-
-out_drm:
 #ifdef CONFIG_DRM_EXYNOS_IPP
-	exynos_platform_device_ipp_unregister();
-out_ipp_dev:
+err_unregister_ipp_drv:
 	platform_driver_unregister(&ipp_driver);
-out_ipp:
+err_unregister_gsc_drv:
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_GSC
 	platform_driver_unregister(&gsc_driver);
-out_gsc:
+err_unregister_rotator_drv:
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_ROTATOR
 	platform_driver_unregister(&rotator_driver);
-out_rotator:
+err_unregister_fimc_drv:
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_FIMC
 	platform_driver_unregister(&fimc_driver);
-out_fimc:
+err_unregister_g2d_drv:
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_G2D
 	platform_driver_unregister(&g2d_driver);
-out_g2d:
-#endif
-
-#ifdef CONFIG_DRM_EXYNOS_VIDI
-	platform_driver_unregister(&vidi_driver);
-out_vidi:
+err_unregister_hdmi_drv:
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_HDMI
-	exynos_platform_device_hdmi_unregister();
-out_common_hdmi_dev:
-	platform_driver_unregister(&exynos_drm_common_hdmi_driver);
-out_common_hdmi:
-	platform_driver_unregister(&mixer_driver);
-out_mixer:
 	platform_driver_unregister(&hdmi_driver);
-out_hdmi:
+err_unregister_mixer_drv:
+	platform_driver_unregister(&mixer_driver);
+err_unregister_dsi_drv:
+#endif
+
+#ifdef CONFIG_DRM_EXYNOS_DSI
+	platform_driver_unregister(&dsi_driver);
+err_unregister_dp_drv:
+#endif
+
+#ifdef CONFIG_DRM_EXYNOS_DP
+	platform_driver_unregister(&dp_driver);
+err_unregister_fimd_drv:
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_FIMD
 	platform_driver_unregister(&fimd_driver);
-out_fimd:
 #endif
 	return ret;
 }
 
-static void __exit exynos_drm_exit(void)
+static int exynos_drm_platform_remove(struct platform_device *pdev)
 {
-	platform_device_unregister(exynos_drm_pdev);
-
-	platform_driver_unregister(&exynos_drm_platform_driver);
-
 #ifdef CONFIG_DRM_EXYNOS_IPP
 	exynos_platform_device_ipp_unregister();
 	platform_driver_unregister(&ipp_driver);
@@ -480,19 +715,74 @@ static void __exit exynos_drm_exit(void)
 #endif
 
 #ifdef CONFIG_DRM_EXYNOS_HDMI
-	exynos_platform_device_hdmi_unregister();
-	platform_driver_unregister(&exynos_drm_common_hdmi_driver);
 	platform_driver_unregister(&mixer_driver);
 	platform_driver_unregister(&hdmi_driver);
 #endif
 
+#ifdef CONFIG_DRM_EXYNOS_FIMD
+	platform_driver_unregister(&fimd_driver);
+#endif
+
+#ifdef CONFIG_DRM_EXYNOS_DSI
+	platform_driver_unregister(&dsi_driver);
+#endif
+
+#ifdef CONFIG_DRM_EXYNOS_DP
+	platform_driver_unregister(&dp_driver);
+#endif
+	component_master_del(&pdev->dev, &exynos_drm_ops);
+	return 0;
+}
+
+static struct platform_driver exynos_drm_platform_driver = {
+	.probe	= exynos_drm_platform_probe,
+	.remove	= exynos_drm_platform_remove,
+	.driver	= {
+		.owner	= THIS_MODULE,
+		.name	= "exynos-drm",
+		.pm	= &exynos_drm_pm_ops,
+	},
+};
+
+static int exynos_drm_init(void)
+{
+	int ret;
+
+	exynos_drm_pdev = platform_device_register_simple("exynos-drm", -1,
+								NULL, 0);
+	if (IS_ERR(exynos_drm_pdev))
+		return PTR_ERR(exynos_drm_pdev);
+
 #ifdef CONFIG_DRM_EXYNOS_VIDI
-	platform_driver_unregister(&vidi_driver);
+	ret = exynos_drm_probe_vidi();
+	if (ret < 0)
+		goto err_unregister_pd;
 #endif
 
-#ifdef CONFIG_DRM_EXYNOS_FIMD
-	platform_driver_unregister(&fimd_driver);
+	ret = platform_driver_register(&exynos_drm_platform_driver);
+	if (ret)
+		goto err_remove_vidi;
+
+	return 0;
+
+err_remove_vidi:
+#ifdef CONFIG_DRM_EXYNOS_VIDI
+	exynos_drm_remove_vidi();
+
+err_unregister_pd:
+#endif
+	platform_device_unregister(exynos_drm_pdev);
+
+	return ret;
+}
+
+static void exynos_drm_exit(void)
+{
+	platform_driver_unregister(&exynos_drm_platform_driver);
+#ifdef CONFIG_DRM_EXYNOS_VIDI
+	exynos_drm_remove_vidi();
 #endif
+	platform_device_unregister(exynos_drm_pdev);
 }
 
 module_init(exynos_drm_init);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.h b/drivers/gpu/drm/exynos/exynos_drm_drv.h
index eaa19668bf0..06cde450627 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.h
@@ -42,6 +42,13 @@ struct drm_connector;
 
 extern unsigned int drm_vblank_offdelay;
 
+/* This enumerates device type. */
+enum exynos_drm_device_type {
+	EXYNOS_DEVICE_TYPE_NONE,
+	EXYNOS_DEVICE_TYPE_CRTC,
+	EXYNOS_DEVICE_TYPE_CONNECTOR,
+};
+
 /* this enumerates display type. */
 enum exynos_drm_output_type {
 	EXYNOS_DISPLAY_TYPE_NONE,
@@ -54,22 +61,6 @@ enum exynos_drm_output_type {
 };
 
 /*
- * Exynos drm overlay ops structure.
- *
- * @mode_set: copy drm overlay info to hw specific overlay info.
- * @commit: apply hardware specific overlay data to registers.
- * @enable: enable hardware specific overlay.
- * @disable: disable hardware specific overlay.
- */
-struct exynos_drm_overlay_ops {
-	void (*mode_set)(struct device *subdrv_dev,
-			 struct exynos_drm_overlay *overlay);
-	void (*commit)(struct device *subdrv_dev, int zpos);
-	void (*enable)(struct device *subdrv_dev, int zpos);
-	void (*disable)(struct device *subdrv_dev, int zpos);
-};
-
-/*
  * Exynos drm common overlay structure.
  *
  * @fb_x: offset x on a framebuffer to be displayed.
@@ -138,77 +129,104 @@ struct exynos_drm_overlay {
  * Exynos DRM Display Structure.
  *	- this structure is common to analog tv, digital tv and lcd panel.
  *
- * @type: one of EXYNOS_DISPLAY_TYPE_LCD and HDMI.
- * @is_connected: check for that display is connected or not.
- * @get_edid: get edid modes from display driver.
- * @get_panel: get panel object from display driver.
+ * @remove: cleans up the display for removal
+ * @mode_fixup: fix mode data comparing to hw specific display mode.
+ * @mode_set: convert drm_display_mode to hw specific display mode and
+ *	      would be called by encoder->mode_set().
  * @check_mode: check if mode is valid or not.
- * @power_on: display device on or off.
+ * @dpms: display device on or off.
+ * @commit: apply changes to hw
  */
+struct exynos_drm_display;
 struct exynos_drm_display_ops {
+	int (*create_connector)(struct exynos_drm_display *display,
+				struct drm_encoder *encoder);
+	void (*remove)(struct exynos_drm_display *display);
+	void (*mode_fixup)(struct exynos_drm_display *display,
+				struct drm_connector *connector,
+				const struct drm_display_mode *mode,
+				struct drm_display_mode *adjusted_mode);
+	void (*mode_set)(struct exynos_drm_display *display,
+				struct drm_display_mode *mode);
+	int (*check_mode)(struct exynos_drm_display *display,
+				struct drm_display_mode *mode);
+	void (*dpms)(struct exynos_drm_display *display, int mode);
+	void (*commit)(struct exynos_drm_display *display);
+};
+
+/*
+ * Exynos drm display structure, maps 1:1 with an encoder/connector
+ *
+ * @list: the list entry for this manager
+ * @type: one of EXYNOS_DISPLAY_TYPE_LCD and HDMI.
+ * @encoder: encoder object this display maps to
+ * @connector: connector object this display maps to
+ * @ops: pointer to callbacks for exynos drm specific functionality
+ * @ctx: A pointer to the display's implementation specific context
+ */
+struct exynos_drm_display {
+	struct list_head list;
 	enum exynos_drm_output_type type;
-	bool (*is_connected)(struct device *dev);
-	struct edid *(*get_edid)(struct device *dev,
-			struct drm_connector *connector);
-	void *(*get_panel)(struct device *dev);
-	int (*check_mode)(struct device *dev, struct drm_display_mode *mode);
-	int (*power_on)(struct device *dev, int mode);
+	struct drm_encoder *encoder;
+	struct drm_connector *connector;
+	struct exynos_drm_display_ops *ops;
+	void *ctx;
 };
 
 /*
  * Exynos drm manager ops
  *
  * @dpms: control device power.
- * @apply: set timing, vblank and overlay data to registers.
- * @mode_fixup: fix mode data comparing to hw specific display mode.
- * @mode_set: convert drm_display_mode to hw specific display mode and
- *	      would be called by encoder->mode_set().
- * @get_max_resol: get maximum resolution to specific hardware.
+ * @mode_fixup: fix mode data before applying it
+ * @mode_set: set the given mode to the manager
  * @commit: set current hw specific display mode to hw.
  * @enable_vblank: specific driver callback for enabling vblank interrupt.
  * @disable_vblank: specific driver callback for disabling vblank interrupt.
  * @wait_for_vblank: wait for vblank interrupt to make sure that
  *	hardware overlay is updated.
+ * @win_mode_set: copy drm overlay info to hw specific overlay info.
+ * @win_commit: apply hardware specific overlay data to registers.
+ * @win_enable: enable hardware specific overlay.
+ * @win_disable: disable hardware specific overlay.
  */
+struct exynos_drm_manager;
 struct exynos_drm_manager_ops {
-	void (*dpms)(struct device *subdrv_dev, int mode);
-	void (*apply)(struct device *subdrv_dev);
-	void (*mode_fixup)(struct device *subdrv_dev,
-				struct drm_connector *connector,
+	void (*dpms)(struct exynos_drm_manager *mgr, int mode);
+	bool (*mode_fixup)(struct exynos_drm_manager *mgr,
 				const struct drm_display_mode *mode,
 				struct drm_display_mode *adjusted_mode);
-	void (*mode_set)(struct device *subdrv_dev, void *mode);
-	void (*get_max_resol)(struct device *subdrv_dev, unsigned int *width,
-				unsigned int *height);
-	void (*commit)(struct device *subdrv_dev);
-	int (*enable_vblank)(struct device *subdrv_dev);
-	void (*disable_vblank)(struct device *subdrv_dev);
-	void (*wait_for_vblank)(struct device *subdrv_dev);
+	void (*mode_set)(struct exynos_drm_manager *mgr,
+				const struct drm_display_mode *mode);
+	void (*commit)(struct exynos_drm_manager *mgr);
+	int (*enable_vblank)(struct exynos_drm_manager *mgr);
+	void (*disable_vblank)(struct exynos_drm_manager *mgr);
+	void (*wait_for_vblank)(struct exynos_drm_manager *mgr);
+	void (*win_mode_set)(struct exynos_drm_manager *mgr,
+				struct exynos_drm_overlay *overlay);
+	void (*win_commit)(struct exynos_drm_manager *mgr, int zpos);
+	void (*win_enable)(struct exynos_drm_manager *mgr, int zpos);
+	void (*win_disable)(struct exynos_drm_manager *mgr, int zpos);
 };
 
 /*
- * Exynos drm common manager structure.
+ * Exynos drm common manager structure, maps 1:1 with a crtc
  *
- * @dev: pointer to device object for subdrv device driver.
- *	sub drivers such as display controller or hdmi driver,
- *	have their own device object.
- * @ops: pointer to callbacks for exynos drm specific framebuffer.
- *	these callbacks should be set by specific drivers such fimd
- *	or hdmi driver and are used to control hardware global registers.
- * @overlay_ops: pointer to callbacks for exynos drm specific framebuffer.
- *	these callbacks should be set by specific drivers such fimd
- *	or hdmi driver and are used to control hardware overlay reigsters.
- * @display: pointer to callbacks for exynos drm specific framebuffer.
- *	these callbacks should be set by specific drivers such fimd
- *	or hdmi driver and are used to control display devices such as
- *	analog tv, digital tv and lcd panel and also get timing data for them.
+ * @list: the list entry for this manager
+ * @type: one of EXYNOS_DISPLAY_TYPE_LCD and HDMI.
+ * @drm_dev: pointer to the drm device
+ * @crtc: crtc object.
+ * @pipe: the pipe number for this crtc/manager
+ * @ops: pointer to callbacks for exynos drm specific functionality
+ * @ctx: A pointer to the manager's implementation specific context
  */
 struct exynos_drm_manager {
-	struct device *dev;
+	struct list_head list;
+	enum exynos_drm_output_type type;
+	struct drm_device *drm_dev;
+	struct drm_crtc *crtc;
 	int pipe;
 	struct exynos_drm_manager_ops *ops;
-	struct exynos_drm_overlay_ops *overlay_ops;
-	struct exynos_drm_display_ops *display_ops;
+	void *ctx;
 };
 
 struct exynos_drm_g2d_private {
@@ -226,6 +244,7 @@ struct exynos_drm_ipp_private {
 struct drm_exynos_file_private {
 	struct exynos_drm_g2d_private	*g2d_priv;
 	struct exynos_drm_ipp_private	*ipp_priv;
+	struct file			*anon_filp;
 };
 
 /*
@@ -236,7 +255,7 @@ struct drm_exynos_file_private {
  *	otherwise default one.
  * @da_space_size: size of device address space.
  *	if 0 then default value is used for it.
- * @da_space_order: order to device address space.
+ * @pipe: the pipe number for this crtc/manager.
  */
 struct exynos_drm_private {
 	struct drm_fb_helper *fb_helper;
@@ -254,7 +273,8 @@ struct exynos_drm_private {
 
 	unsigned long da_start;
 	unsigned long da_space_size;
-	unsigned long da_space_order;
+
+	unsigned int pipe;
 };
 
 /*
@@ -265,21 +285,18 @@ struct exynos_drm_private {
  * @drm_dev: pointer to drm_device and this pointer would be set
  *	when sub driver calls exynos_drm_subdrv_register().
  * @manager: subdrv has its own manager to control a hardware appropriately
- *	and we can access a hardware drawing on this manager.
+ *     and we can access a hardware drawing on this manager.
  * @probe: this callback would be called by exynos drm driver after
- *	subdrv is registered to it.
+ *     subdrv is registered to it.
  * @remove: this callback is used to release resources created
- *	by probe callback.
+ *     by probe callback.
  * @open: this would be called with drm device file open.
  * @close: this would be called with drm device file close.
- * @encoder: encoder object owned by this sub driver.
- * @connector: connector object owned by this sub driver.
  */
 struct exynos_drm_subdrv {
 	struct list_head list;
 	struct device *dev;
 	struct drm_device *drm_dev;
-	struct exynos_drm_manager *manager;
 
 	int (*probe)(struct drm_device *drm_dev, struct device *dev);
 	void (*remove)(struct drm_device *drm_dev, struct device *dev);
@@ -287,34 +304,16 @@ struct exynos_drm_subdrv {
 			struct drm_file *file);
 	void (*close)(struct drm_device *drm_dev, struct device *dev,
 			struct drm_file *file);
-
-	struct drm_encoder *encoder;
-	struct drm_connector *connector;
 };
 
-/*
- * this function calls a probe callback registered to sub driver list and
- * create its own encoder and connector and then set drm_device object
- * to global one.
- */
-int exynos_drm_device_register(struct drm_device *dev);
-/*
- * this function calls a remove callback registered to sub driver list and
- * destroy its own encoder and connetor.
- */
-int exynos_drm_device_unregister(struct drm_device *dev);
-
-/*
- * this function would be called by sub drivers such as display controller
- * or hdmi driver to register this sub driver object to exynos drm driver
- * and when a sub driver is registered to exynos drm driver a probe callback
- * of the sub driver is called and creates its own encoder and connector.
- */
+ /* This function would be called by non kms drivers such as g2d and ipp. */
 int exynos_drm_subdrv_register(struct exynos_drm_subdrv *drm_subdrv);
 
 /* this function removes subdrv list from exynos drm driver */
 int exynos_drm_subdrv_unregister(struct exynos_drm_subdrv *drm_subdrv);
 
+int exynos_drm_device_subdrv_probe(struct drm_device *dev);
+int exynos_drm_device_subdrv_remove(struct drm_device *dev);
 int exynos_drm_subdrv_open(struct drm_device *dev, struct drm_file *file);
 void exynos_drm_subdrv_close(struct drm_device *dev, struct drm_file *file);
 
@@ -339,9 +338,41 @@ int exynos_platform_device_ipp_register(void);
  */
 void exynos_platform_device_ipp_unregister(void);
 
+#ifdef CONFIG_DRM_EXYNOS_DPI
+struct exynos_drm_display * exynos_dpi_probe(struct device *dev);
+int exynos_dpi_remove(struct device *dev);
+#else
+static inline struct exynos_drm_display *
+exynos_dpi_probe(struct device *dev) { return NULL; }
+static inline int exynos_dpi_remove(struct device *dev) { return 0; }
+#endif
+
+/*
+ * this function registers exynos drm vidi platform device/driver.
+ */
+int exynos_drm_probe_vidi(void);
+
+/*
+ * this function unregister exynos drm vidi platform device/driver.
+ */
+void exynos_drm_remove_vidi(void);
+
+/* This function creates a encoder and a connector, and initializes them. */
+int exynos_drm_create_enc_conn(struct drm_device *dev,
+				struct exynos_drm_display *display);
+
+int exynos_drm_component_add(struct device *dev,
+				enum exynos_drm_device_type dev_type,
+				enum exynos_drm_output_type out_type);
+
+void exynos_drm_component_del(struct device *dev,
+				enum exynos_drm_device_type dev_type);
+
 extern struct platform_driver fimd_driver;
-extern struct platform_driver hdmi_driver;
+extern struct platform_driver dp_driver;
+extern struct platform_driver dsi_driver;
 extern struct platform_driver mixer_driver;
+extern struct platform_driver hdmi_driver;
 extern struct platform_driver exynos_drm_common_hdmi_driver;
 extern struct platform_driver vidi_driver;
 extern struct platform_driver g2d_driver;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dsi.c b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
new file mode 100644
index 00000000000..6302aa64f6c
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
@@ -0,0 +1,1546 @@
+/*
+ * Samsung SoC MIPI DSI Master driver.
+ *
+ * Copyright (c) 2014 Samsung Electronics Co., Ltd
+ *
+ * Contacts: Tomasz Figa <t.figa@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+*/
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_panel.h>
+
+#include <linux/clk.h>
+#include <linux/irq.h>
+#include <linux/phy/phy.h>
+#include <linux/regulator/consumer.h>
+#include <linux/component.h>
+
+#include <video/mipi_display.h>
+#include <video/videomode.h>
+
+#include "exynos_drm_drv.h"
+
+/* returns true iff both arguments logically differs */
+#define NEQV(a, b) (!(a) ^ !(b))
+
+#define DSIM_STATUS_REG		0x0	/* Status register */
+#define DSIM_SWRST_REG		0x4	/* Software reset register */
+#define DSIM_CLKCTRL_REG	0x8	/* Clock control register */
+#define DSIM_TIMEOUT_REG	0xc	/* Time out register */
+#define DSIM_CONFIG_REG		0x10	/* Configuration register */
+#define DSIM_ESCMODE_REG	0x14	/* Escape mode register */
+
+/* Main display image resolution register */
+#define DSIM_MDRESOL_REG	0x18
+#define DSIM_MVPORCH_REG	0x1c	/* Main display Vporch register */
+#define DSIM_MHPORCH_REG	0x20	/* Main display Hporch register */
+#define DSIM_MSYNC_REG		0x24	/* Main display sync area register */
+
+/* Sub display image resolution register */
+#define DSIM_SDRESOL_REG	0x28
+#define DSIM_INTSRC_REG		0x2c	/* Interrupt source register */
+#define DSIM_INTMSK_REG		0x30	/* Interrupt mask register */
+#define DSIM_PKTHDR_REG		0x34	/* Packet Header FIFO register */
+#define DSIM_PAYLOAD_REG	0x38	/* Payload FIFO register */
+#define DSIM_RXFIFO_REG		0x3c	/* Read FIFO register */
+#define DSIM_FIFOTHLD_REG	0x40	/* FIFO threshold level register */
+#define DSIM_FIFOCTRL_REG	0x44	/* FIFO status and control register */
+
+/* FIFO memory AC characteristic register */
+#define DSIM_PLLCTRL_REG	0x4c	/* PLL control register */
+#define DSIM_PLLTMR_REG		0x50	/* PLL timer register */
+#define DSIM_PHYACCHR_REG	0x54	/* D-PHY AC characteristic register */
+#define DSIM_PHYACCHR1_REG	0x58	/* D-PHY AC characteristic register1 */
+
+/* DSIM_STATUS */
+#define DSIM_STOP_STATE_DAT(x)		(((x) & 0xf) << 0)
+#define DSIM_STOP_STATE_CLK		(1 << 8)
+#define DSIM_TX_READY_HS_CLK		(1 << 10)
+#define DSIM_PLL_STABLE			(1 << 31)
+
+/* DSIM_SWRST */
+#define DSIM_FUNCRST			(1 << 16)
+#define DSIM_SWRST			(1 << 0)
+
+/* DSIM_TIMEOUT */
+#define DSIM_LPDR_TIMEOUT(x)		((x) << 0)
+#define DSIM_BTA_TIMEOUT(x)		((x) << 16)
+
+/* DSIM_CLKCTRL */
+#define DSIM_ESC_PRESCALER(x)		(((x) & 0xffff) << 0)
+#define DSIM_ESC_PRESCALER_MASK		(0xffff << 0)
+#define DSIM_LANE_ESC_CLK_EN_CLK	(1 << 19)
+#define DSIM_LANE_ESC_CLK_EN_DATA(x)	(((x) & 0xf) << 20)
+#define DSIM_LANE_ESC_CLK_EN_DATA_MASK	(0xf << 20)
+#define DSIM_BYTE_CLKEN			(1 << 24)
+#define DSIM_BYTE_CLK_SRC(x)		(((x) & 0x3) << 25)
+#define DSIM_BYTE_CLK_SRC_MASK		(0x3 << 25)
+#define DSIM_PLL_BYPASS			(1 << 27)
+#define DSIM_ESC_CLKEN			(1 << 28)
+#define DSIM_TX_REQUEST_HSCLK		(1 << 31)
+
+/* DSIM_CONFIG */
+#define DSIM_LANE_EN_CLK		(1 << 0)
+#define DSIM_LANE_EN(x)			(((x) & 0xf) << 1)
+#define DSIM_NUM_OF_DATA_LANE(x)	(((x) & 0x3) << 5)
+#define DSIM_SUB_PIX_FORMAT(x)		(((x) & 0x7) << 8)
+#define DSIM_MAIN_PIX_FORMAT_MASK	(0x7 << 12)
+#define DSIM_MAIN_PIX_FORMAT_RGB888	(0x7 << 12)
+#define DSIM_MAIN_PIX_FORMAT_RGB666	(0x6 << 12)
+#define DSIM_MAIN_PIX_FORMAT_RGB666_P	(0x5 << 12)
+#define DSIM_MAIN_PIX_FORMAT_RGB565	(0x4 << 12)
+#define DSIM_SUB_VC			(((x) & 0x3) << 16)
+#define DSIM_MAIN_VC			(((x) & 0x3) << 18)
+#define DSIM_HSA_MODE			(1 << 20)
+#define DSIM_HBP_MODE			(1 << 21)
+#define DSIM_HFP_MODE			(1 << 22)
+#define DSIM_HSE_MODE			(1 << 23)
+#define DSIM_AUTO_MODE			(1 << 24)
+#define DSIM_VIDEO_MODE			(1 << 25)
+#define DSIM_BURST_MODE			(1 << 26)
+#define DSIM_SYNC_INFORM		(1 << 27)
+#define DSIM_EOT_DISABLE		(1 << 28)
+#define DSIM_MFLUSH_VS			(1 << 29)
+
+/* DSIM_ESCMODE */
+#define DSIM_TX_TRIGGER_RST		(1 << 4)
+#define DSIM_TX_LPDT_LP			(1 << 6)
+#define DSIM_CMD_LPDT_LP		(1 << 7)
+#define DSIM_FORCE_BTA			(1 << 16)
+#define DSIM_FORCE_STOP_STATE		(1 << 20)
+#define DSIM_STOP_STATE_CNT(x)		(((x) & 0x7ff) << 21)
+#define DSIM_STOP_STATE_CNT_MASK	(0x7ff << 21)
+
+/* DSIM_MDRESOL */
+#define DSIM_MAIN_STAND_BY		(1 << 31)
+#define DSIM_MAIN_VRESOL(x)		(((x) & 0x7ff) << 16)
+#define DSIM_MAIN_HRESOL(x)		(((x) & 0X7ff) << 0)
+
+/* DSIM_MVPORCH */
+#define DSIM_CMD_ALLOW(x)		((x) << 28)
+#define DSIM_STABLE_VFP(x)		((x) << 16)
+#define DSIM_MAIN_VBP(x)		((x) << 0)
+#define DSIM_CMD_ALLOW_MASK		(0xf << 28)
+#define DSIM_STABLE_VFP_MASK		(0x7ff << 16)
+#define DSIM_MAIN_VBP_MASK		(0x7ff << 0)
+
+/* DSIM_MHPORCH */
+#define DSIM_MAIN_HFP(x)		((x) << 16)
+#define DSIM_MAIN_HBP(x)		((x) << 0)
+#define DSIM_MAIN_HFP_MASK		((0xffff) << 16)
+#define DSIM_MAIN_HBP_MASK		((0xffff) << 0)
+
+/* DSIM_MSYNC */
+#define DSIM_MAIN_VSA(x)		((x) << 22)
+#define DSIM_MAIN_HSA(x)		((x) << 0)
+#define DSIM_MAIN_VSA_MASK		((0x3ff) << 22)
+#define DSIM_MAIN_HSA_MASK		((0xffff) << 0)
+
+/* DSIM_SDRESOL */
+#define DSIM_SUB_STANDY(x)		((x) << 31)
+#define DSIM_SUB_VRESOL(x)		((x) << 16)
+#define DSIM_SUB_HRESOL(x)		((x) << 0)
+#define DSIM_SUB_STANDY_MASK		((0x1) << 31)
+#define DSIM_SUB_VRESOL_MASK		((0x7ff) << 16)
+#define DSIM_SUB_HRESOL_MASK		((0x7ff) << 0)
+
+/* DSIM_INTSRC */
+#define DSIM_INT_PLL_STABLE		(1 << 31)
+#define DSIM_INT_SW_RST_RELEASE		(1 << 30)
+#define DSIM_INT_SFR_FIFO_EMPTY		(1 << 29)
+#define DSIM_INT_BTA			(1 << 25)
+#define DSIM_INT_FRAME_DONE		(1 << 24)
+#define DSIM_INT_RX_TIMEOUT		(1 << 21)
+#define DSIM_INT_BTA_TIMEOUT		(1 << 20)
+#define DSIM_INT_RX_DONE		(1 << 18)
+#define DSIM_INT_RX_TE			(1 << 17)
+#define DSIM_INT_RX_ACK			(1 << 16)
+#define DSIM_INT_RX_ECC_ERR		(1 << 15)
+#define DSIM_INT_RX_CRC_ERR		(1 << 14)
+
+/* DSIM_FIFOCTRL */
+#define DSIM_RX_DATA_FULL		(1 << 25)
+#define DSIM_RX_DATA_EMPTY		(1 << 24)
+#define DSIM_SFR_HEADER_FULL		(1 << 23)
+#define DSIM_SFR_HEADER_EMPTY		(1 << 22)
+#define DSIM_SFR_PAYLOAD_FULL		(1 << 21)
+#define DSIM_SFR_PAYLOAD_EMPTY		(1 << 20)
+#define DSIM_I80_HEADER_FULL		(1 << 19)
+#define DSIM_I80_HEADER_EMPTY		(1 << 18)
+#define DSIM_I80_PAYLOAD_FULL		(1 << 17)
+#define DSIM_I80_PAYLOAD_EMPTY		(1 << 16)
+#define DSIM_SD_HEADER_FULL		(1 << 15)
+#define DSIM_SD_HEADER_EMPTY		(1 << 14)
+#define DSIM_SD_PAYLOAD_FULL		(1 << 13)
+#define DSIM_SD_PAYLOAD_EMPTY		(1 << 12)
+#define DSIM_MD_HEADER_FULL		(1 << 11)
+#define DSIM_MD_HEADER_EMPTY		(1 << 10)
+#define DSIM_MD_PAYLOAD_FULL		(1 << 9)
+#define DSIM_MD_PAYLOAD_EMPTY		(1 << 8)
+#define DSIM_RX_FIFO			(1 << 4)
+#define DSIM_SFR_FIFO			(1 << 3)
+#define DSIM_I80_FIFO			(1 << 2)
+#define DSIM_SD_FIFO			(1 << 1)
+#define DSIM_MD_FIFO			(1 << 0)
+
+/* DSIM_PHYACCHR */
+#define DSIM_AFC_EN			(1 << 14)
+#define DSIM_AFC_CTL(x)			(((x) & 0x7) << 5)
+
+/* DSIM_PLLCTRL */
+#define DSIM_FREQ_BAND(x)		((x) << 24)
+#define DSIM_PLL_EN			(1 << 23)
+#define DSIM_PLL_P(x)			((x) << 13)
+#define DSIM_PLL_M(x)			((x) << 4)
+#define DSIM_PLL_S(x)			((x) << 1)
+
+#define DSI_MAX_BUS_WIDTH		4
+#define DSI_NUM_VIRTUAL_CHANNELS	4
+#define DSI_TX_FIFO_SIZE		2048
+#define DSI_RX_FIFO_SIZE		256
+#define DSI_XFER_TIMEOUT_MS		100
+#define DSI_RX_FIFO_EMPTY		0x30800002
+
+enum exynos_dsi_transfer_type {
+	EXYNOS_DSI_TX,
+	EXYNOS_DSI_RX,
+};
+
+struct exynos_dsi_transfer {
+	struct list_head list;
+	struct completion completed;
+	int result;
+	u8 data_id;
+	u8 data[2];
+	u16 flags;
+
+	const u8 *tx_payload;
+	u16 tx_len;
+	u16 tx_done;
+
+	u8 *rx_payload;
+	u16 rx_len;
+	u16 rx_done;
+};
+
+#define DSIM_STATE_ENABLED		BIT(0)
+#define DSIM_STATE_INITIALIZED		BIT(1)
+#define DSIM_STATE_CMD_LPM		BIT(2)
+
+struct exynos_dsi {
+	struct mipi_dsi_host dsi_host;
+	struct drm_connector connector;
+	struct drm_encoder *encoder;
+	struct device_node *panel_node;
+	struct drm_panel *panel;
+	struct device *dev;
+
+	void __iomem *reg_base;
+	struct phy *phy;
+	struct clk *pll_clk;
+	struct clk *bus_clk;
+	struct regulator_bulk_data supplies[2];
+	int irq;
+
+	u32 pll_clk_rate;
+	u32 burst_clk_rate;
+	u32 esc_clk_rate;
+	u32 lanes;
+	u32 mode_flags;
+	u32 format;
+	struct videomode vm;
+
+	int state;
+	struct drm_property *brightness;
+	struct completion completed;
+
+	spinlock_t transfer_lock; /* protects transfer_list */
+	struct list_head transfer_list;
+};
+
+#define host_to_dsi(host) container_of(host, struct exynos_dsi, dsi_host)
+#define connector_to_dsi(c) container_of(c, struct exynos_dsi, connector)
+
+static void exynos_dsi_wait_for_reset(struct exynos_dsi *dsi)
+{
+	if (wait_for_completion_timeout(&dsi->completed, msecs_to_jiffies(300)))
+		return;
+
+	dev_err(dsi->dev, "timeout waiting for reset\n");
+}
+
+static void exynos_dsi_reset(struct exynos_dsi *dsi)
+{
+	reinit_completion(&dsi->completed);
+	writel(DSIM_SWRST, dsi->reg_base + DSIM_SWRST_REG);
+}
+
+#ifndef MHZ
+#define MHZ	(1000*1000)
+#endif
+
+static unsigned long exynos_dsi_pll_find_pms(struct exynos_dsi *dsi,
+		unsigned long fin, unsigned long fout, u8 *p, u16 *m, u8 *s)
+{
+	unsigned long best_freq = 0;
+	u32 min_delta = 0xffffffff;
+	u8 p_min, p_max;
+	u8 _p, uninitialized_var(best_p);
+	u16 _m, uninitialized_var(best_m);
+	u8 _s, uninitialized_var(best_s);
+
+	p_min = DIV_ROUND_UP(fin, (12 * MHZ));
+	p_max = fin / (6 * MHZ);
+
+	for (_p = p_min; _p <= p_max; ++_p) {
+		for (_s = 0; _s <= 5; ++_s) {
+			u64 tmp;
+			u32 delta;
+
+			tmp = (u64)fout * (_p << _s);
+			do_div(tmp, fin);
+			_m = tmp;
+			if (_m < 41 || _m > 125)
+				continue;
+
+			tmp = (u64)_m * fin;
+			do_div(tmp, _p);
+			if (tmp < 500 * MHZ || tmp > 1000 * MHZ)
+				continue;
+
+			tmp = (u64)_m * fin;
+			do_div(tmp, _p << _s);
+
+			delta = abs(fout - tmp);
+			if (delta < min_delta) {
+				best_p = _p;
+				best_m = _m;
+				best_s = _s;
+				min_delta = delta;
+				best_freq = tmp;
+			}
+		}
+	}
+
+	if (best_freq) {
+		*p = best_p;
+		*m = best_m;
+		*s = best_s;
+	}
+
+	return best_freq;
+}
+
+static unsigned long exynos_dsi_set_pll(struct exynos_dsi *dsi,
+					unsigned long freq)
+{
+	static const unsigned long freq_bands[] = {
+		100 * MHZ, 120 * MHZ, 160 * MHZ, 200 * MHZ,
+		270 * MHZ, 320 * MHZ, 390 * MHZ, 450 * MHZ,
+		510 * MHZ, 560 * MHZ, 640 * MHZ, 690 * MHZ,
+		770 * MHZ, 870 * MHZ, 950 * MHZ,
+	};
+	unsigned long fin, fout;
+	int timeout, band;
+	u8 p, s;
+	u16 m;
+	u32 reg;
+
+	clk_set_rate(dsi->pll_clk, dsi->pll_clk_rate);
+
+	fin = clk_get_rate(dsi->pll_clk);
+	if (!fin) {
+		dev_err(dsi->dev, "failed to get PLL clock frequency\n");
+		return 0;
+	}
+
+	dev_dbg(dsi->dev, "PLL input frequency: %lu\n", fin);
+
+	fout = exynos_dsi_pll_find_pms(dsi, fin, freq, &p, &m, &s);
+	if (!fout) {
+		dev_err(dsi->dev,
+			"failed to find PLL PMS for requested frequency\n");
+		return -EFAULT;
+	}
+
+	for (band = 0; band < ARRAY_SIZE(freq_bands); ++band)
+		if (fout < freq_bands[band])
+			break;
+
+	dev_dbg(dsi->dev, "PLL freq %lu, (p %d, m %d, s %d), band %d\n", fout,
+		p, m, s, band);
+
+	writel(500, dsi->reg_base + DSIM_PLLTMR_REG);
+
+	reg = DSIM_FREQ_BAND(band) | DSIM_PLL_EN
+			| DSIM_PLL_P(p) | DSIM_PLL_M(m) | DSIM_PLL_S(s);
+	writel(reg, dsi->reg_base + DSIM_PLLCTRL_REG);
+
+	timeout = 1000;
+	do {
+		if (timeout-- == 0) {
+			dev_err(dsi->dev, "PLL failed to stabilize\n");
+			return -EFAULT;
+		}
+		reg = readl(dsi->reg_base + DSIM_STATUS_REG);
+	} while ((reg & DSIM_PLL_STABLE) == 0);
+
+	return fout;
+}
+
+static int exynos_dsi_enable_clock(struct exynos_dsi *dsi)
+{
+	unsigned long hs_clk, byte_clk, esc_clk;
+	unsigned long esc_div;
+	u32 reg;
+
+	hs_clk = exynos_dsi_set_pll(dsi, dsi->burst_clk_rate);
+	if (!hs_clk) {
+		dev_err(dsi->dev, "failed to configure DSI PLL\n");
+		return -EFAULT;
+	}
+
+	byte_clk = hs_clk / 8;
+	esc_div = DIV_ROUND_UP(byte_clk, dsi->esc_clk_rate);
+	esc_clk = byte_clk / esc_div;
+
+	if (esc_clk > 20 * MHZ) {
+		++esc_div;
+		esc_clk = byte_clk / esc_div;
+	}
+
+	dev_dbg(dsi->dev, "hs_clk = %lu, byte_clk = %lu, esc_clk = %lu\n",
+		hs_clk, byte_clk, esc_clk);
+
+	reg = readl(dsi->reg_base + DSIM_CLKCTRL_REG);
+	reg &= ~(DSIM_ESC_PRESCALER_MASK | DSIM_LANE_ESC_CLK_EN_CLK
+			| DSIM_LANE_ESC_CLK_EN_DATA_MASK | DSIM_PLL_BYPASS
+			| DSIM_BYTE_CLK_SRC_MASK);
+	reg |= DSIM_ESC_CLKEN | DSIM_BYTE_CLKEN
+			| DSIM_ESC_PRESCALER(esc_div)
+			| DSIM_LANE_ESC_CLK_EN_CLK
+			| DSIM_LANE_ESC_CLK_EN_DATA(BIT(dsi->lanes) - 1)
+			| DSIM_BYTE_CLK_SRC(0)
+			| DSIM_TX_REQUEST_HSCLK;
+	writel(reg, dsi->reg_base + DSIM_CLKCTRL_REG);
+
+	return 0;
+}
+
+static void exynos_dsi_disable_clock(struct exynos_dsi *dsi)
+{
+	u32 reg;
+
+	reg = readl(dsi->reg_base + DSIM_CLKCTRL_REG);
+	reg &= ~(DSIM_LANE_ESC_CLK_EN_CLK | DSIM_LANE_ESC_CLK_EN_DATA_MASK
+			| DSIM_ESC_CLKEN | DSIM_BYTE_CLKEN);
+	writel(reg, dsi->reg_base + DSIM_CLKCTRL_REG);
+
+	reg = readl(dsi->reg_base + DSIM_PLLCTRL_REG);
+	reg &= ~DSIM_PLL_EN;
+	writel(reg, dsi->reg_base + DSIM_PLLCTRL_REG);
+}
+
+static int exynos_dsi_init_link(struct exynos_dsi *dsi)
+{
+	int timeout;
+	u32 reg;
+	u32 lanes_mask;
+
+	/* Initialize FIFO pointers */
+	reg = readl(dsi->reg_base + DSIM_FIFOCTRL_REG);
+	reg &= ~0x1f;
+	writel(reg, dsi->reg_base + DSIM_FIFOCTRL_REG);
+
+	usleep_range(9000, 11000);
+
+	reg |= 0x1f;
+	writel(reg, dsi->reg_base + DSIM_FIFOCTRL_REG);
+
+	usleep_range(9000, 11000);
+
+	/* DSI configuration */
+	reg = 0;
+
+	if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO) {
+		reg |= DSIM_VIDEO_MODE;
+
+		if (!(dsi->mode_flags & MIPI_DSI_MODE_VSYNC_FLUSH))
+			reg |= DSIM_MFLUSH_VS;
+		if (!(dsi->mode_flags & MIPI_DSI_MODE_EOT_PACKET))
+			reg |= DSIM_EOT_DISABLE;
+		if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO_SYNC_PULSE)
+			reg |= DSIM_SYNC_INFORM;
+		if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO_BURST)
+			reg |= DSIM_BURST_MODE;
+		if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO_AUTO_VERT)
+			reg |= DSIM_AUTO_MODE;
+		if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO_HSE)
+			reg |= DSIM_HSE_MODE;
+		if (!(dsi->mode_flags & MIPI_DSI_MODE_VIDEO_HFP))
+			reg |= DSIM_HFP_MODE;
+		if (!(dsi->mode_flags & MIPI_DSI_MODE_VIDEO_HBP))
+			reg |= DSIM_HBP_MODE;
+		if (!(dsi->mode_flags & MIPI_DSI_MODE_VIDEO_HSA))
+			reg |= DSIM_HSA_MODE;
+	}
+
+	switch (dsi->format) {
+	case MIPI_DSI_FMT_RGB888:
+		reg |= DSIM_MAIN_PIX_FORMAT_RGB888;
+		break;
+	case MIPI_DSI_FMT_RGB666:
+		reg |= DSIM_MAIN_PIX_FORMAT_RGB666;
+		break;
+	case MIPI_DSI_FMT_RGB666_PACKED:
+		reg |= DSIM_MAIN_PIX_FORMAT_RGB666_P;
+		break;
+	case MIPI_DSI_FMT_RGB565:
+		reg |= DSIM_MAIN_PIX_FORMAT_RGB565;
+		break;
+	default:
+		dev_err(dsi->dev, "invalid pixel format\n");
+		return -EINVAL;
+	}
+
+	reg |= DSIM_NUM_OF_DATA_LANE(dsi->lanes - 1);
+
+	writel(reg, dsi->reg_base + DSIM_CONFIG_REG);
+
+	reg |= DSIM_LANE_EN_CLK;
+	writel(reg, dsi->reg_base + DSIM_CONFIG_REG);
+
+	lanes_mask = BIT(dsi->lanes) - 1;
+	reg |= DSIM_LANE_EN(lanes_mask);
+	writel(reg, dsi->reg_base + DSIM_CONFIG_REG);
+
+	/* Check clock and data lane state are stop state */
+	timeout = 100;
+	do {
+		if (timeout-- == 0) {
+			dev_err(dsi->dev, "waiting for bus lanes timed out\n");
+			return -EFAULT;
+		}
+
+		reg = readl(dsi->reg_base + DSIM_STATUS_REG);
+		if ((reg & DSIM_STOP_STATE_DAT(lanes_mask))
+		    != DSIM_STOP_STATE_DAT(lanes_mask))
+			continue;
+	} while (!(reg & (DSIM_STOP_STATE_CLK | DSIM_TX_READY_HS_CLK)));
+
+	reg = readl(dsi->reg_base + DSIM_ESCMODE_REG);
+	reg &= ~DSIM_STOP_STATE_CNT_MASK;
+	reg |= DSIM_STOP_STATE_CNT(0xf);
+	writel(reg, dsi->reg_base + DSIM_ESCMODE_REG);
+
+	reg = DSIM_BTA_TIMEOUT(0xff) | DSIM_LPDR_TIMEOUT(0xffff);
+	writel(reg, dsi->reg_base + DSIM_TIMEOUT_REG);
+
+	return 0;
+}
+
+static void exynos_dsi_set_display_mode(struct exynos_dsi *dsi)
+{
+	struct videomode *vm = &dsi->vm;
+	u32 reg;
+
+	if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO) {
+		reg = DSIM_CMD_ALLOW(0xf)
+			| DSIM_STABLE_VFP(vm->vfront_porch)
+			| DSIM_MAIN_VBP(vm->vback_porch);
+		writel(reg, dsi->reg_base + DSIM_MVPORCH_REG);
+
+		reg = DSIM_MAIN_HFP(vm->hfront_porch)
+			| DSIM_MAIN_HBP(vm->hback_porch);
+		writel(reg, dsi->reg_base + DSIM_MHPORCH_REG);
+
+		reg = DSIM_MAIN_VSA(vm->vsync_len)
+			| DSIM_MAIN_HSA(vm->hsync_len);
+		writel(reg, dsi->reg_base + DSIM_MSYNC_REG);
+	}
+
+	reg = DSIM_MAIN_HRESOL(vm->hactive) | DSIM_MAIN_VRESOL(vm->vactive);
+	writel(reg, dsi->reg_base + DSIM_MDRESOL_REG);
+
+	dev_dbg(dsi->dev, "LCD size = %dx%d\n", vm->hactive, vm->vactive);
+}
+
+static void exynos_dsi_set_display_enable(struct exynos_dsi *dsi, bool enable)
+{
+	u32 reg;
+
+	reg = readl(dsi->reg_base + DSIM_MDRESOL_REG);
+	if (enable)
+		reg |= DSIM_MAIN_STAND_BY;
+	else
+		reg &= ~DSIM_MAIN_STAND_BY;
+	writel(reg, dsi->reg_base + DSIM_MDRESOL_REG);
+}
+
+static int exynos_dsi_wait_for_hdr_fifo(struct exynos_dsi *dsi)
+{
+	int timeout = 2000;
+
+	do {
+		u32 reg = readl(dsi->reg_base + DSIM_FIFOCTRL_REG);
+
+		if (!(reg & DSIM_SFR_HEADER_FULL))
+			return 0;
+
+		if (!cond_resched())
+			usleep_range(950, 1050);
+	} while (--timeout);
+
+	return -ETIMEDOUT;
+}
+
+static void exynos_dsi_set_cmd_lpm(struct exynos_dsi *dsi, bool lpm)
+{
+	u32 v = readl(dsi->reg_base + DSIM_ESCMODE_REG);
+
+	if (lpm)
+		v |= DSIM_CMD_LPDT_LP;
+	else
+		v &= ~DSIM_CMD_LPDT_LP;
+
+	writel(v, dsi->reg_base + DSIM_ESCMODE_REG);
+}
+
+static void exynos_dsi_force_bta(struct exynos_dsi *dsi)
+{
+	u32 v = readl(dsi->reg_base + DSIM_ESCMODE_REG);
+
+	v |= DSIM_FORCE_BTA;
+	writel(v, dsi->reg_base + DSIM_ESCMODE_REG);
+}
+
+static void exynos_dsi_send_to_fifo(struct exynos_dsi *dsi,
+					struct exynos_dsi_transfer *xfer)
+{
+	struct device *dev = dsi->dev;
+	const u8 *payload = xfer->tx_payload + xfer->tx_done;
+	u16 length = xfer->tx_len - xfer->tx_done;
+	bool first = !xfer->tx_done;
+	u32 reg;
+
+	dev_dbg(dev, "< xfer %p: tx len %u, done %u, rx len %u, done %u\n",
+		xfer, xfer->tx_len, xfer->tx_done, xfer->rx_len, xfer->rx_done);
+
+	if (length > DSI_TX_FIFO_SIZE)
+		length = DSI_TX_FIFO_SIZE;
+
+	xfer->tx_done += length;
+
+	/* Send payload */
+	while (length >= 4) {
+		reg = (payload[3] << 24) | (payload[2] << 16)
+					| (payload[1] << 8) | payload[0];
+		writel(reg, dsi->reg_base + DSIM_PAYLOAD_REG);
+		payload += 4;
+		length -= 4;
+	}
+
+	reg = 0;
+	switch (length) {
+	case 3:
+		reg |= payload[2] << 16;
+		/* Fall through */
+	case 2:
+		reg |= payload[1] << 8;
+		/* Fall through */
+	case 1:
+		reg |= payload[0];
+		writel(reg, dsi->reg_base + DSIM_PAYLOAD_REG);
+		break;
+	case 0:
+		/* Do nothing */
+		break;
+	}
+
+	/* Send packet header */
+	if (!first)
+		return;
+
+	reg = (xfer->data[1] << 16) | (xfer->data[0] << 8) | xfer->data_id;
+	if (exynos_dsi_wait_for_hdr_fifo(dsi)) {
+		dev_err(dev, "waiting for header FIFO timed out\n");
+		return;
+	}
+
+	if (NEQV(xfer->flags & MIPI_DSI_MSG_USE_LPM,
+		 dsi->state & DSIM_STATE_CMD_LPM)) {
+		exynos_dsi_set_cmd_lpm(dsi, xfer->flags & MIPI_DSI_MSG_USE_LPM);
+		dsi->state ^= DSIM_STATE_CMD_LPM;
+	}
+
+	writel(reg, dsi->reg_base + DSIM_PKTHDR_REG);
+
+	if (xfer->flags & MIPI_DSI_MSG_REQ_ACK)
+		exynos_dsi_force_bta(dsi);
+}
+
+static void exynos_dsi_read_from_fifo(struct exynos_dsi *dsi,
+					struct exynos_dsi_transfer *xfer)
+{
+	u8 *payload = xfer->rx_payload + xfer->rx_done;
+	bool first = !xfer->rx_done;
+	struct device *dev = dsi->dev;
+	u16 length;
+	u32 reg;
+
+	if (first) {
+		reg = readl(dsi->reg_base + DSIM_RXFIFO_REG);
+
+		switch (reg & 0x3f) {
+		case MIPI_DSI_RX_GENERIC_SHORT_READ_RESPONSE_2BYTE:
+		case MIPI_DSI_RX_DCS_SHORT_READ_RESPONSE_2BYTE:
+			if (xfer->rx_len >= 2) {
+				payload[1] = reg >> 16;
+				++xfer->rx_done;
+			}
+			/* Fall through */
+		case MIPI_DSI_RX_GENERIC_SHORT_READ_RESPONSE_1BYTE:
+		case MIPI_DSI_RX_DCS_SHORT_READ_RESPONSE_1BYTE:
+			payload[0] = reg >> 8;
+			++xfer->rx_done;
+			xfer->rx_len = xfer->rx_done;
+			xfer->result = 0;
+			goto clear_fifo;
+		case MIPI_DSI_RX_ACKNOWLEDGE_AND_ERROR_REPORT:
+			dev_err(dev, "DSI Error Report: 0x%04x\n",
+				(reg >> 8) & 0xffff);
+			xfer->result = 0;
+			goto clear_fifo;
+		}
+
+		length = (reg >> 8) & 0xffff;
+		if (length > xfer->rx_len) {
+			dev_err(dev,
+				"response too long (%u > %u bytes), stripping\n",
+				xfer->rx_len, length);
+			length = xfer->rx_len;
+		} else if (length < xfer->rx_len)
+			xfer->rx_len = length;
+	}
+
+	length = xfer->rx_len - xfer->rx_done;
+	xfer->rx_done += length;
+
+	/* Receive payload */
+	while (length >= 4) {
+		reg = readl(dsi->reg_base + DSIM_RXFIFO_REG);
+		payload[0] = (reg >>  0) & 0xff;
+		payload[1] = (reg >>  8) & 0xff;
+		payload[2] = (reg >> 16) & 0xff;
+		payload[3] = (reg >> 24) & 0xff;
+		payload += 4;
+		length -= 4;
+	}
+
+	if (length) {
+		reg = readl(dsi->reg_base + DSIM_RXFIFO_REG);
+		switch (length) {
+		case 3:
+			payload[2] = (reg >> 16) & 0xff;
+			/* Fall through */
+		case 2:
+			payload[1] = (reg >> 8) & 0xff;
+			/* Fall through */
+		case 1:
+			payload[0] = reg & 0xff;
+		}
+	}
+
+	if (xfer->rx_done == xfer->rx_len)
+		xfer->result = 0;
+
+clear_fifo:
+	length = DSI_RX_FIFO_SIZE / 4;
+	do {
+		reg = readl(dsi->reg_base + DSIM_RXFIFO_REG);
+		if (reg == DSI_RX_FIFO_EMPTY)
+			break;
+	} while (--length);
+}
+
+static void exynos_dsi_transfer_start(struct exynos_dsi *dsi)
+{
+	unsigned long flags;
+	struct exynos_dsi_transfer *xfer;
+	bool start = false;
+
+again:
+	spin_lock_irqsave(&dsi->transfer_lock, flags);
+
+	if (list_empty(&dsi->transfer_list)) {
+		spin_unlock_irqrestore(&dsi->transfer_lock, flags);
+		return;
+	}
+
+	xfer = list_first_entry(&dsi->transfer_list,
+					struct exynos_dsi_transfer, list);
+
+	spin_unlock_irqrestore(&dsi->transfer_lock, flags);
+
+	if (xfer->tx_len && xfer->tx_done == xfer->tx_len)
+		/* waiting for RX */
+		return;
+
+	exynos_dsi_send_to_fifo(dsi, xfer);
+
+	if (xfer->tx_len || xfer->rx_len)
+		return;
+
+	xfer->result = 0;
+	complete(&xfer->completed);
+
+	spin_lock_irqsave(&dsi->transfer_lock, flags);
+
+	list_del_init(&xfer->list);
+	start = !list_empty(&dsi->transfer_list);
+
+	spin_unlock_irqrestore(&dsi->transfer_lock, flags);
+
+	if (start)
+		goto again;
+}
+
+static bool exynos_dsi_transfer_finish(struct exynos_dsi *dsi)
+{
+	struct exynos_dsi_transfer *xfer;
+	unsigned long flags;
+	bool start = true;
+
+	spin_lock_irqsave(&dsi->transfer_lock, flags);
+
+	if (list_empty(&dsi->transfer_list)) {
+		spin_unlock_irqrestore(&dsi->transfer_lock, flags);
+		return false;
+	}
+
+	xfer = list_first_entry(&dsi->transfer_list,
+					struct exynos_dsi_transfer, list);
+
+	spin_unlock_irqrestore(&dsi->transfer_lock, flags);
+
+	dev_dbg(dsi->dev,
+		"> xfer %p, tx_len %u, tx_done %u, rx_len %u, rx_done %u\n",
+		xfer, xfer->tx_len, xfer->tx_done, xfer->rx_len, xfer->rx_done);
+
+	if (xfer->tx_done != xfer->tx_len)
+		return true;
+
+	if (xfer->rx_done != xfer->rx_len)
+		exynos_dsi_read_from_fifo(dsi, xfer);
+
+	if (xfer->rx_done != xfer->rx_len)
+		return true;
+
+	spin_lock_irqsave(&dsi->transfer_lock, flags);
+
+	list_del_init(&xfer->list);
+	start = !list_empty(&dsi->transfer_list);
+
+	spin_unlock_irqrestore(&dsi->transfer_lock, flags);
+
+	if (!xfer->rx_len)
+		xfer->result = 0;
+	complete(&xfer->completed);
+
+	return start;
+}
+
+static void exynos_dsi_remove_transfer(struct exynos_dsi *dsi,
+					struct exynos_dsi_transfer *xfer)
+{
+	unsigned long flags;
+	bool start;
+
+	spin_lock_irqsave(&dsi->transfer_lock, flags);
+
+	if (!list_empty(&dsi->transfer_list) &&
+	    xfer == list_first_entry(&dsi->transfer_list,
+				     struct exynos_dsi_transfer, list)) {
+		list_del_init(&xfer->list);
+		start = !list_empty(&dsi->transfer_list);
+		spin_unlock_irqrestore(&dsi->transfer_lock, flags);
+		if (start)
+			exynos_dsi_transfer_start(dsi);
+		return;
+	}
+
+	list_del_init(&xfer->list);
+
+	spin_unlock_irqrestore(&dsi->transfer_lock, flags);
+}
+
+static int exynos_dsi_transfer(struct exynos_dsi *dsi,
+					struct exynos_dsi_transfer *xfer)
+{
+	unsigned long flags;
+	bool stopped;
+
+	xfer->tx_done = 0;
+	xfer->rx_done = 0;
+	xfer->result = -ETIMEDOUT;
+	init_completion(&xfer->completed);
+
+	spin_lock_irqsave(&dsi->transfer_lock, flags);
+
+	stopped = list_empty(&dsi->transfer_list);
+	list_add_tail(&xfer->list, &dsi->transfer_list);
+
+	spin_unlock_irqrestore(&dsi->transfer_lock, flags);
+
+	if (stopped)
+		exynos_dsi_transfer_start(dsi);
+
+	wait_for_completion_timeout(&xfer->completed,
+				    msecs_to_jiffies(DSI_XFER_TIMEOUT_MS));
+	if (xfer->result == -ETIMEDOUT) {
+		exynos_dsi_remove_transfer(dsi, xfer);
+		dev_err(dsi->dev, "xfer timed out: %*ph %*ph\n", 2, xfer->data,
+			xfer->tx_len, xfer->tx_payload);
+		return -ETIMEDOUT;
+	}
+
+	/* Also covers hardware timeout condition */
+	return xfer->result;
+}
+
+static irqreturn_t exynos_dsi_irq(int irq, void *dev_id)
+{
+	struct exynos_dsi *dsi = dev_id;
+	u32 status;
+
+	status = readl(dsi->reg_base + DSIM_INTSRC_REG);
+	if (!status) {
+		static unsigned long int j;
+		if (printk_timed_ratelimit(&j, 500))
+			dev_warn(dsi->dev, "spurious interrupt\n");
+		return IRQ_HANDLED;
+	}
+	writel(status, dsi->reg_base + DSIM_INTSRC_REG);
+
+	if (status & DSIM_INT_SW_RST_RELEASE) {
+		u32 mask = ~(DSIM_INT_RX_DONE | DSIM_INT_SFR_FIFO_EMPTY);
+		writel(mask, dsi->reg_base + DSIM_INTMSK_REG);
+		complete(&dsi->completed);
+		return IRQ_HANDLED;
+	}
+
+	if (!(status & (DSIM_INT_RX_DONE | DSIM_INT_SFR_FIFO_EMPTY)))
+		return IRQ_HANDLED;
+
+	if (exynos_dsi_transfer_finish(dsi))
+		exynos_dsi_transfer_start(dsi);
+
+	return IRQ_HANDLED;
+}
+
+static int exynos_dsi_init(struct exynos_dsi *dsi)
+{
+	exynos_dsi_enable_clock(dsi);
+	exynos_dsi_reset(dsi);
+	enable_irq(dsi->irq);
+	exynos_dsi_wait_for_reset(dsi);
+	exynos_dsi_init_link(dsi);
+
+	return 0;
+}
+
+static int exynos_dsi_host_attach(struct mipi_dsi_host *host,
+				  struct mipi_dsi_device *device)
+{
+	struct exynos_dsi *dsi = host_to_dsi(host);
+
+	dsi->lanes = device->lanes;
+	dsi->format = device->format;
+	dsi->mode_flags = device->mode_flags;
+	dsi->panel_node = device->dev.of_node;
+
+	if (dsi->connector.dev)
+		drm_helper_hpd_irq_event(dsi->connector.dev);
+
+	return 0;
+}
+
+static int exynos_dsi_host_detach(struct mipi_dsi_host *host,
+				  struct mipi_dsi_device *device)
+{
+	struct exynos_dsi *dsi = host_to_dsi(host);
+
+	dsi->panel_node = NULL;
+
+	if (dsi->connector.dev)
+		drm_helper_hpd_irq_event(dsi->connector.dev);
+
+	return 0;
+}
+
+/* distinguish between short and long DSI packet types */
+static bool exynos_dsi_is_short_dsi_type(u8 type)
+{
+	return (type & 0x0f) <= 8;
+}
+
+static ssize_t exynos_dsi_host_transfer(struct mipi_dsi_host *host,
+				       struct mipi_dsi_msg *msg)
+{
+	struct exynos_dsi *dsi = host_to_dsi(host);
+	struct exynos_dsi_transfer xfer;
+	int ret;
+
+	if (!(dsi->state & DSIM_STATE_INITIALIZED)) {
+		ret = exynos_dsi_init(dsi);
+		if (ret)
+			return ret;
+		dsi->state |= DSIM_STATE_INITIALIZED;
+	}
+
+	if (msg->tx_len == 0)
+		return -EINVAL;
+
+	xfer.data_id = msg->type | (msg->channel << 6);
+
+	if (exynos_dsi_is_short_dsi_type(msg->type)) {
+		const char *tx_buf = msg->tx_buf;
+
+		if (msg->tx_len > 2)
+			return -EINVAL;
+		xfer.tx_len = 0;
+		xfer.data[0] = tx_buf[0];
+		xfer.data[1] = (msg->tx_len == 2) ? tx_buf[1] : 0;
+	} else {
+		xfer.tx_len = msg->tx_len;
+		xfer.data[0] = msg->tx_len & 0xff;
+		xfer.data[1] = msg->tx_len >> 8;
+		xfer.tx_payload = msg->tx_buf;
+	}
+
+	xfer.rx_len = msg->rx_len;
+	xfer.rx_payload = msg->rx_buf;
+	xfer.flags = msg->flags;
+
+	ret = exynos_dsi_transfer(dsi, &xfer);
+	return (ret < 0) ? ret : xfer.rx_done;
+}
+
+static const struct mipi_dsi_host_ops exynos_dsi_ops = {
+	.attach = exynos_dsi_host_attach,
+	.detach = exynos_dsi_host_detach,
+	.transfer = exynos_dsi_host_transfer,
+};
+
+static int exynos_dsi_poweron(struct exynos_dsi *dsi)
+{
+	int ret;
+
+	ret = regulator_bulk_enable(ARRAY_SIZE(dsi->supplies), dsi->supplies);
+	if (ret < 0) {
+		dev_err(dsi->dev, "cannot enable regulators %d\n", ret);
+		return ret;
+	}
+
+	ret = clk_prepare_enable(dsi->bus_clk);
+	if (ret < 0) {
+		dev_err(dsi->dev, "cannot enable bus clock %d\n", ret);
+		goto err_bus_clk;
+	}
+
+	ret = clk_prepare_enable(dsi->pll_clk);
+	if (ret < 0) {
+		dev_err(dsi->dev, "cannot enable pll clock %d\n", ret);
+		goto err_pll_clk;
+	}
+
+	ret = phy_power_on(dsi->phy);
+	if (ret < 0) {
+		dev_err(dsi->dev, "cannot enable phy %d\n", ret);
+		goto err_phy;
+	}
+
+	return 0;
+
+err_phy:
+	clk_disable_unprepare(dsi->pll_clk);
+err_pll_clk:
+	clk_disable_unprepare(dsi->bus_clk);
+err_bus_clk:
+	regulator_bulk_disable(ARRAY_SIZE(dsi->supplies), dsi->supplies);
+
+	return ret;
+}
+
+static void exynos_dsi_poweroff(struct exynos_dsi *dsi)
+{
+	int ret;
+
+	usleep_range(10000, 20000);
+
+	if (dsi->state & DSIM_STATE_INITIALIZED) {
+		dsi->state &= ~DSIM_STATE_INITIALIZED;
+
+		exynos_dsi_disable_clock(dsi);
+
+		disable_irq(dsi->irq);
+	}
+
+	dsi->state &= ~DSIM_STATE_CMD_LPM;
+
+	phy_power_off(dsi->phy);
+
+	clk_disable_unprepare(dsi->pll_clk);
+	clk_disable_unprepare(dsi->bus_clk);
+
+	ret = regulator_bulk_disable(ARRAY_SIZE(dsi->supplies), dsi->supplies);
+	if (ret < 0)
+		dev_err(dsi->dev, "cannot disable regulators %d\n", ret);
+}
+
+static int exynos_dsi_enable(struct exynos_dsi *dsi)
+{
+	int ret;
+
+	if (dsi->state & DSIM_STATE_ENABLED)
+		return 0;
+
+	ret = exynos_dsi_poweron(dsi);
+	if (ret < 0)
+		return ret;
+
+	ret = drm_panel_enable(dsi->panel);
+	if (ret < 0) {
+		exynos_dsi_poweroff(dsi);
+		return ret;
+	}
+
+	exynos_dsi_set_display_mode(dsi);
+	exynos_dsi_set_display_enable(dsi, true);
+
+	dsi->state |= DSIM_STATE_ENABLED;
+
+	return 0;
+}
+
+static void exynos_dsi_disable(struct exynos_dsi *dsi)
+{
+	if (!(dsi->state & DSIM_STATE_ENABLED))
+		return;
+
+	exynos_dsi_set_display_enable(dsi, false);
+	drm_panel_disable(dsi->panel);
+	exynos_dsi_poweroff(dsi);
+
+	dsi->state &= ~DSIM_STATE_ENABLED;
+}
+
+static void exynos_dsi_dpms(struct exynos_drm_display *display, int mode)
+{
+	struct exynos_dsi *dsi = display->ctx;
+
+	if (dsi->panel) {
+		switch (mode) {
+		case DRM_MODE_DPMS_ON:
+			exynos_dsi_enable(dsi);
+			break;
+		case DRM_MODE_DPMS_STANDBY:
+		case DRM_MODE_DPMS_SUSPEND:
+		case DRM_MODE_DPMS_OFF:
+			exynos_dsi_disable(dsi);
+			break;
+		default:
+			break;
+		}
+	}
+}
+
+static enum drm_connector_status
+exynos_dsi_detect(struct drm_connector *connector, bool force)
+{
+	struct exynos_dsi *dsi = connector_to_dsi(connector);
+
+	if (!dsi->panel) {
+		dsi->panel = of_drm_find_panel(dsi->panel_node);
+		if (dsi->panel)
+			drm_panel_attach(dsi->panel, &dsi->connector);
+	} else if (!dsi->panel_node) {
+		struct exynos_drm_display *display;
+
+		display = platform_get_drvdata(to_platform_device(dsi->dev));
+		exynos_dsi_dpms(display, DRM_MODE_DPMS_OFF);
+		drm_panel_detach(dsi->panel);
+		dsi->panel = NULL;
+	}
+
+	if (dsi->panel)
+		return connector_status_connected;
+
+	return connector_status_disconnected;
+}
+
+static void exynos_dsi_connector_destroy(struct drm_connector *connector)
+{
+}
+
+static struct drm_connector_funcs exynos_dsi_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.detect = exynos_dsi_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = exynos_dsi_connector_destroy,
+};
+
+static int exynos_dsi_get_modes(struct drm_connector *connector)
+{
+	struct exynos_dsi *dsi = connector_to_dsi(connector);
+
+	if (dsi->panel)
+		return dsi->panel->funcs->get_modes(dsi->panel);
+
+	return 0;
+}
+
+static int exynos_dsi_mode_valid(struct drm_connector *connector,
+				 struct drm_display_mode *mode)
+{
+	return MODE_OK;
+}
+
+static struct drm_encoder *
+exynos_dsi_best_encoder(struct drm_connector *connector)
+{
+	struct exynos_dsi *dsi = connector_to_dsi(connector);
+
+	return dsi->encoder;
+}
+
+static struct drm_connector_helper_funcs exynos_dsi_connector_helper_funcs = {
+	.get_modes = exynos_dsi_get_modes,
+	.mode_valid = exynos_dsi_mode_valid,
+	.best_encoder = exynos_dsi_best_encoder,
+};
+
+static int exynos_dsi_create_connector(struct exynos_drm_display *display,
+				       struct drm_encoder *encoder)
+{
+	struct exynos_dsi *dsi = display->ctx;
+	struct drm_connector *connector = &dsi->connector;
+	int ret;
+
+	dsi->encoder = encoder;
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	ret = drm_connector_init(encoder->dev, connector,
+				 &exynos_dsi_connector_funcs,
+				 DRM_MODE_CONNECTOR_DSI);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	drm_connector_helper_add(connector, &exynos_dsi_connector_helper_funcs);
+	drm_sysfs_connector_add(connector);
+	drm_mode_connector_attach_encoder(connector, encoder);
+
+	return 0;
+}
+
+static void exynos_dsi_mode_set(struct exynos_drm_display *display,
+			 struct drm_display_mode *mode)
+{
+	struct exynos_dsi *dsi = display->ctx;
+	struct videomode *vm = &dsi->vm;
+
+	vm->hactive = mode->hdisplay;
+	vm->vactive = mode->vdisplay;
+	vm->vfront_porch = mode->vsync_start - mode->vdisplay;
+	vm->vback_porch = mode->vtotal - mode->vsync_end;
+	vm->vsync_len = mode->vsync_end - mode->vsync_start;
+	vm->hfront_porch = mode->hsync_start - mode->hdisplay;
+	vm->hback_porch = mode->htotal - mode->hsync_end;
+	vm->hsync_len = mode->hsync_end - mode->hsync_start;
+}
+
+static struct exynos_drm_display_ops exynos_dsi_display_ops = {
+	.create_connector = exynos_dsi_create_connector,
+	.mode_set = exynos_dsi_mode_set,
+	.dpms = exynos_dsi_dpms
+};
+
+static struct exynos_drm_display exynos_dsi_display = {
+	.type = EXYNOS_DISPLAY_TYPE_LCD,
+	.ops = &exynos_dsi_display_ops,
+};
+
+/* of_* functions will be removed after merge of of_graph patches */
+static struct device_node *
+of_get_child_by_name_reg(struct device_node *parent, const char *name, u32 reg)
+{
+	struct device_node *np;
+
+	for_each_child_of_node(parent, np) {
+		u32 r;
+
+		if (!np->name || of_node_cmp(np->name, name))
+			continue;
+
+		if (of_property_read_u32(np, "reg", &r) < 0)
+			r = 0;
+
+		if (reg == r)
+			break;
+	}
+
+	return np;
+}
+
+static struct device_node *of_graph_get_port_by_reg(struct device_node *parent,
+						    u32 reg)
+{
+	struct device_node *ports, *port;
+
+	ports = of_get_child_by_name(parent, "ports");
+	if (ports)
+		parent = ports;
+
+	port = of_get_child_by_name_reg(parent, "port", reg);
+
+	of_node_put(ports);
+
+	return port;
+}
+
+static struct device_node *
+of_graph_get_endpoint_by_reg(struct device_node *port, u32 reg)
+{
+	return of_get_child_by_name_reg(port, "endpoint", reg);
+}
+
+static int exynos_dsi_of_read_u32(const struct device_node *np,
+				  const char *propname, u32 *out_value)
+{
+	int ret = of_property_read_u32(np, propname, out_value);
+
+	if (ret < 0)
+		pr_err("%s: failed to get '%s' property\n", np->full_name,
+		       propname);
+
+	return ret;
+}
+
+enum {
+	DSI_PORT_IN,
+	DSI_PORT_OUT
+};
+
+static int exynos_dsi_parse_dt(struct exynos_dsi *dsi)
+{
+	struct device *dev = dsi->dev;
+	struct device_node *node = dev->of_node;
+	struct device_node *port, *ep;
+	int ret;
+
+	ret = exynos_dsi_of_read_u32(node, "samsung,pll-clock-frequency",
+				     &dsi->pll_clk_rate);
+	if (ret < 0)
+		return ret;
+
+	port = of_graph_get_port_by_reg(node, DSI_PORT_OUT);
+	if (!port) {
+		dev_err(dev, "no output port specified\n");
+		return -EINVAL;
+	}
+
+	ep = of_graph_get_endpoint_by_reg(port, 0);
+	of_node_put(port);
+	if (!ep) {
+		dev_err(dev, "no endpoint specified in output port\n");
+		return -EINVAL;
+	}
+
+	ret = exynos_dsi_of_read_u32(ep, "samsung,burst-clock-frequency",
+				     &dsi->burst_clk_rate);
+	if (ret < 0)
+		goto end;
+
+	ret = exynos_dsi_of_read_u32(ep, "samsung,esc-clock-frequency",
+				     &dsi->esc_clk_rate);
+
+end:
+	of_node_put(ep);
+
+	return ret;
+}
+
+static int exynos_dsi_bind(struct device *dev, struct device *master,
+				void *data)
+{
+	struct drm_device *drm_dev = data;
+	struct exynos_dsi *dsi;
+	int ret;
+
+	ret = exynos_drm_create_enc_conn(drm_dev, &exynos_dsi_display);
+	if (ret) {
+		DRM_ERROR("Encoder create [%d] failed with %d\n",
+				exynos_dsi_display.type, ret);
+		return ret;
+	}
+
+	dsi = exynos_dsi_display.ctx;
+
+	return mipi_dsi_host_register(&dsi->dsi_host);
+}
+
+static void exynos_dsi_unbind(struct device *dev, struct device *master,
+				void *data)
+{
+	struct exynos_dsi *dsi = exynos_dsi_display.ctx;
+	struct drm_encoder *encoder = dsi->encoder;
+
+	exynos_dsi_dpms(&exynos_dsi_display, DRM_MODE_DPMS_OFF);
+
+	mipi_dsi_host_unregister(&dsi->dsi_host);
+
+	encoder->funcs->destroy(encoder);
+	drm_connector_cleanup(&dsi->connector);
+}
+
+static const struct component_ops exynos_dsi_component_ops = {
+	.bind	= exynos_dsi_bind,
+	.unbind	= exynos_dsi_unbind,
+};
+
+static int exynos_dsi_probe(struct platform_device *pdev)
+{
+	struct resource *res;
+	struct exynos_dsi *dsi;
+	int ret;
+
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
+					exynos_dsi_display.type);
+	if (ret)
+		return ret;
+
+	dsi = devm_kzalloc(&pdev->dev, sizeof(*dsi), GFP_KERNEL);
+	if (!dsi) {
+		dev_err(&pdev->dev, "failed to allocate dsi object.\n");
+		ret = -ENOMEM;
+		goto err_del_component;
+	}
+
+	init_completion(&dsi->completed);
+	spin_lock_init(&dsi->transfer_lock);
+	INIT_LIST_HEAD(&dsi->transfer_list);
+
+	dsi->dsi_host.ops = &exynos_dsi_ops;
+	dsi->dsi_host.dev = &pdev->dev;
+
+	dsi->dev = &pdev->dev;
+
+	ret = exynos_dsi_parse_dt(dsi);
+	if (ret)
+		goto err_del_component;
+
+	dsi->supplies[0].supply = "vddcore";
+	dsi->supplies[1].supply = "vddio";
+	ret = devm_regulator_bulk_get(&pdev->dev, ARRAY_SIZE(dsi->supplies),
+				      dsi->supplies);
+	if (ret) {
+		dev_info(&pdev->dev, "failed to get regulators: %d\n", ret);
+		return -EPROBE_DEFER;
+	}
+
+	dsi->pll_clk = devm_clk_get(&pdev->dev, "pll_clk");
+	if (IS_ERR(dsi->pll_clk)) {
+		dev_info(&pdev->dev, "failed to get dsi pll input clock\n");
+		ret = PTR_ERR(dsi->pll_clk);
+		goto err_del_component;
+	}
+
+	dsi->bus_clk = devm_clk_get(&pdev->dev, "bus_clk");
+	if (IS_ERR(dsi->bus_clk)) {
+		dev_info(&pdev->dev, "failed to get dsi bus clock\n");
+		ret = PTR_ERR(dsi->bus_clk);
+		goto err_del_component;
+	}
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dsi->reg_base = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(dsi->reg_base)) {
+		dev_err(&pdev->dev, "failed to remap io region\n");
+		ret = PTR_ERR(dsi->reg_base);
+		goto err_del_component;
+	}
+
+	dsi->phy = devm_phy_get(&pdev->dev, "dsim");
+	if (IS_ERR(dsi->phy)) {
+		dev_info(&pdev->dev, "failed to get dsim phy\n");
+		ret = PTR_ERR(dsi->phy);
+		goto err_del_component;
+	}
+
+	dsi->irq = platform_get_irq(pdev, 0);
+	if (dsi->irq < 0) {
+		dev_err(&pdev->dev, "failed to request dsi irq resource\n");
+		ret = dsi->irq;
+		goto err_del_component;
+	}
+
+	irq_set_status_flags(dsi->irq, IRQ_NOAUTOEN);
+	ret = devm_request_threaded_irq(&pdev->dev, dsi->irq, NULL,
+					exynos_dsi_irq, IRQF_ONESHOT,
+					dev_name(&pdev->dev), dsi);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to request dsi irq\n");
+		goto err_del_component;
+	}
+
+	exynos_dsi_display.ctx = dsi;
+
+	platform_set_drvdata(pdev, &exynos_dsi_display);
+
+	ret = component_add(&pdev->dev, &exynos_dsi_component_ops);
+	if (ret)
+		goto err_del_component;
+
+	return ret;
+
+err_del_component:
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
+	return ret;
+}
+
+static int exynos_dsi_remove(struct platform_device *pdev)
+{
+	component_del(&pdev->dev, &exynos_dsi_component_ops);
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
+
+	return 0;
+}
+
+static struct of_device_id exynos_dsi_of_match[] = {
+	{ .compatible = "samsung,exynos4210-mipi-dsi" },
+	{ }
+};
+
+struct platform_driver dsi_driver = {
+	.probe = exynos_dsi_probe,
+	.remove = exynos_dsi_remove,
+	.driver = {
+		   .name = "exynos-dsi",
+		   .owner = THIS_MODULE,
+		   .of_match_table = exynos_dsi_of_match,
+	},
+};
+
+MODULE_AUTHOR("Tomasz Figa <t.figa@samsung.com>");
+MODULE_AUTHOR("Andrzej Hajda <a.hajda@samsung.com>");
+MODULE_DESCRIPTION("Samsung SoC MIPI DSI Master");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/exynos/exynos_drm_encoder.c b/drivers/gpu/drm/exynos/exynos_drm_encoder.c
index 06f1b2a09da..7e282e3d603 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_encoder.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_encoder.c
@@ -17,7 +17,6 @@
 
 #include "exynos_drm_drv.h"
 #include "exynos_drm_encoder.h"
-#include "exynos_drm_connector.h"
 
 #define to_exynos_encoder(x)	container_of(x, struct exynos_drm_encoder,\
 				drm_encoder)
@@ -26,72 +25,22 @@
  * exynos specific encoder structure.
  *
  * @drm_encoder: encoder object.
- * @manager: specific encoder has its own manager to control a hardware
- *	appropriately and we can access a hardware drawing on this manager.
- * @dpms: store the encoder dpms value.
- * @updated: indicate whether overlay data updating is needed or not.
+ * @display: the display structure that maps to this encoder
  */
 struct exynos_drm_encoder {
-	struct drm_crtc			*old_crtc;
 	struct drm_encoder		drm_encoder;
-	struct exynos_drm_manager	*manager;
-	int				dpms;
-	bool				updated;
+	struct exynos_drm_display	*display;
 };
 
-static void exynos_drm_connector_power(struct drm_encoder *encoder, int mode)
-{
-	struct drm_device *dev = encoder->dev;
-	struct drm_connector *connector;
-
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		if (exynos_drm_best_encoder(connector) == encoder) {
-			DRM_DEBUG_KMS("connector[%d] dpms[%d]\n",
-					connector->base.id, mode);
-
-			exynos_drm_display_power(connector, mode);
-		}
-	}
-}
-
 static void exynos_drm_encoder_dpms(struct drm_encoder *encoder, int mode)
 {
-	struct drm_device *dev = encoder->dev;
-	struct exynos_drm_manager *manager = exynos_drm_get_manager(encoder);
-	struct exynos_drm_manager_ops *manager_ops = manager->ops;
 	struct exynos_drm_encoder *exynos_encoder = to_exynos_encoder(encoder);
+	struct exynos_drm_display *display = exynos_encoder->display;
 
 	DRM_DEBUG_KMS("encoder dpms: %d\n", mode);
 
-	if (exynos_encoder->dpms == mode) {
-		DRM_DEBUG_KMS("desired dpms mode is same as previous one.\n");
-		return;
-	}
-
-	mutex_lock(&dev->struct_mutex);
-
-	switch (mode) {
-	case DRM_MODE_DPMS_ON:
-		if (manager_ops && manager_ops->apply)
-			if (!exynos_encoder->updated)
-				manager_ops->apply(manager->dev);
-
-		exynos_drm_connector_power(encoder, mode);
-		exynos_encoder->dpms = mode;
-		break;
-	case DRM_MODE_DPMS_STANDBY:
-	case DRM_MODE_DPMS_SUSPEND:
-	case DRM_MODE_DPMS_OFF:
-		exynos_drm_connector_power(encoder, mode);
-		exynos_encoder->dpms = mode;
-		exynos_encoder->updated = false;
-		break;
-	default:
-		DRM_ERROR("unspecified mode %d\n", mode);
-		break;
-	}
-
-	mutex_unlock(&dev->struct_mutex);
+	if (display->ops->dpms)
+		display->ops->dpms(display, mode);
 }
 
 static bool
@@ -100,87 +49,31 @@ exynos_drm_encoder_mode_fixup(struct drm_encoder *encoder,
 			       struct drm_display_mode *adjusted_mode)
 {
 	struct drm_device *dev = encoder->dev;
+	struct exynos_drm_encoder *exynos_encoder = to_exynos_encoder(encoder);
+	struct exynos_drm_display *display = exynos_encoder->display;
 	struct drm_connector *connector;
-	struct exynos_drm_manager *manager = exynos_drm_get_manager(encoder);
-	struct exynos_drm_manager_ops *manager_ops = manager->ops;
 
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		if (connector->encoder == encoder)
-			if (manager_ops && manager_ops->mode_fixup)
-				manager_ops->mode_fixup(manager->dev, connector,
-							mode, adjusted_mode);
+		if (connector->encoder != encoder)
+			continue;
+
+		if (display->ops->mode_fixup)
+			display->ops->mode_fixup(display, connector, mode,
+					adjusted_mode);
 	}
 
 	return true;
 }
 
-static void disable_plane_to_crtc(struct drm_device *dev,
-						struct drm_crtc *old_crtc,
-						struct drm_crtc *new_crtc)
-{
-	struct drm_plane *plane;
-
-	/*
-	 * if old_crtc isn't same as encoder->crtc then it means that
-	 * user changed crtc id to another one so the plane to old_crtc
-	 * should be disabled and plane->crtc should be set to new_crtc
-	 * (encoder->crtc)
-	 */
-	list_for_each_entry(plane, &dev->mode_config.plane_list, head) {
-		if (plane->crtc == old_crtc) {
-			/*
-			 * do not change below call order.
-			 *
-			 * plane->funcs->disable_plane call checks
-			 * if encoder->crtc is same as plane->crtc and if same
-			 * then overlay_ops->disable callback will be called
-			 * to diasble current hw overlay so plane->crtc should
-			 * have new_crtc because new_crtc was set to
-			 * encoder->crtc in advance.
-			 */
-			plane->crtc = new_crtc;
-			plane->funcs->disable_plane(plane);
-		}
-	}
-}
-
 static void exynos_drm_encoder_mode_set(struct drm_encoder *encoder,
 					 struct drm_display_mode *mode,
 					 struct drm_display_mode *adjusted_mode)
 {
-	struct drm_device *dev = encoder->dev;
-	struct drm_connector *connector;
-	struct exynos_drm_manager *manager;
-	struct exynos_drm_manager_ops *manager_ops;
-
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		if (connector->encoder == encoder) {
-			struct exynos_drm_encoder *exynos_encoder;
-
-			exynos_encoder = to_exynos_encoder(encoder);
-
-			if (exynos_encoder->old_crtc != encoder->crtc &&
-					exynos_encoder->old_crtc) {
-
-				/*
-				 * disable a plane to old crtc and change
-				 * crtc of the plane to new one.
-				 */
-				disable_plane_to_crtc(dev,
-						exynos_encoder->old_crtc,
-						encoder->crtc);
-			}
-
-			manager = exynos_drm_get_manager(encoder);
-			manager_ops = manager->ops;
-
-			if (manager_ops && manager_ops->mode_set)
-				manager_ops->mode_set(manager->dev,
-							adjusted_mode);
+	struct exynos_drm_encoder *exynos_encoder = to_exynos_encoder(encoder);
+	struct exynos_drm_display *display = exynos_encoder->display;
 
-			exynos_encoder->old_crtc = encoder->crtc;
-		}
-	}
+	if (display->ops->mode_set)
+		display->ops->mode_set(display, adjusted_mode);
 }
 
 static void exynos_drm_encoder_prepare(struct drm_encoder *encoder)
@@ -191,53 +84,15 @@ static void exynos_drm_encoder_prepare(struct drm_encoder *encoder)
 static void exynos_drm_encoder_commit(struct drm_encoder *encoder)
 {
 	struct exynos_drm_encoder *exynos_encoder = to_exynos_encoder(encoder);
-	struct exynos_drm_manager *manager = exynos_encoder->manager;
-	struct exynos_drm_manager_ops *manager_ops = manager->ops;
-
-	if (manager_ops && manager_ops->commit)
-		manager_ops->commit(manager->dev);
-
-	/*
-	 * this will avoid one issue that overlay data is updated to
-	 * real hardware two times.
-	 * And this variable will be used to check if the data was
-	 * already updated or not by exynos_drm_encoder_dpms function.
-	 */
-	exynos_encoder->updated = true;
-
-	/*
-	 * In case of setcrtc, there is no way to update encoder's dpms
-	 * so update it here.
-	 */
-	exynos_encoder->dpms = DRM_MODE_DPMS_ON;
-}
+	struct exynos_drm_display *display = exynos_encoder->display;
 
-void exynos_drm_encoder_complete_scanout(struct drm_framebuffer *fb)
-{
-	struct exynos_drm_encoder *exynos_encoder;
-	struct exynos_drm_manager_ops *ops;
-	struct drm_device *dev = fb->dev;
-	struct drm_encoder *encoder;
+	if (display->ops->dpms)
+		display->ops->dpms(display, DRM_MODE_DPMS_ON);
 
-	/*
-	 * make sure that overlay data are updated to real hardware
-	 * for all encoders.
-	 */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		exynos_encoder = to_exynos_encoder(encoder);
-		ops = exynos_encoder->manager->ops;
-
-		/*
-		 * wait for vblank interrupt
-		 * - this makes sure that overlay data are updated to
-		 *	real hardware.
-		 */
-		if (ops->wait_for_vblank)
-			ops->wait_for_vblank(exynos_encoder->manager->dev);
-	}
+	if (display->ops->commit)
+		display->ops->commit(display);
 }
 
-
 static void exynos_drm_encoder_disable(struct drm_encoder *encoder)
 {
 	struct drm_plane *plane;
@@ -246,7 +101,7 @@ static void exynos_drm_encoder_disable(struct drm_encoder *encoder)
 	exynos_drm_encoder_dpms(encoder, DRM_MODE_DPMS_OFF);
 
 	/* all planes connected to this encoder should be also disabled. */
-	list_for_each_entry(plane, &dev->mode_config.plane_list, head) {
+	drm_for_each_legacy_plane(plane, &dev->mode_config.plane_list) {
 		if (plane->crtc == encoder->crtc)
 			plane->funcs->disable_plane(plane);
 	}
@@ -263,10 +118,7 @@ static struct drm_encoder_helper_funcs exynos_encoder_helper_funcs = {
 
 static void exynos_drm_encoder_destroy(struct drm_encoder *encoder)
 {
-	struct exynos_drm_encoder *exynos_encoder =
-		to_exynos_encoder(encoder);
-
-	exynos_encoder->manager->pipe = -1;
+	struct exynos_drm_encoder *exynos_encoder = to_exynos_encoder(encoder);
 
 	drm_encoder_cleanup(encoder);
 	kfree(exynos_encoder);
@@ -281,13 +133,12 @@ static unsigned int exynos_drm_encoder_clones(struct drm_encoder *encoder)
 	struct drm_encoder *clone;
 	struct drm_device *dev = encoder->dev;
 	struct exynos_drm_encoder *exynos_encoder = to_exynos_encoder(encoder);
-	struct exynos_drm_display_ops *display_ops =
-				exynos_encoder->manager->display_ops;
+	struct exynos_drm_display *display = exynos_encoder->display;
 	unsigned int clone_mask = 0;
 	int cnt = 0;
 
 	list_for_each_entry(clone, &dev->mode_config.encoder_list, head) {
-		switch (display_ops->type) {
+		switch (display->type) {
 		case EXYNOS_DISPLAY_TYPE_LCD:
 		case EXYNOS_DISPLAY_TYPE_HDMI:
 		case EXYNOS_DISPLAY_TYPE_VIDI:
@@ -311,24 +162,20 @@ void exynos_drm_encoder_setup(struct drm_device *dev)
 
 struct drm_encoder *
 exynos_drm_encoder_create(struct drm_device *dev,
-			   struct exynos_drm_manager *manager,
-			   unsigned int possible_crtcs)
+			   struct exynos_drm_display *display,
+			   unsigned long possible_crtcs)
 {
 	struct drm_encoder *encoder;
 	struct exynos_drm_encoder *exynos_encoder;
 
-	if (!manager || !possible_crtcs)
-		return NULL;
-
-	if (!manager->dev)
+	if (!possible_crtcs)
 		return NULL;
 
 	exynos_encoder = kzalloc(sizeof(*exynos_encoder), GFP_KERNEL);
 	if (!exynos_encoder)
 		return NULL;
 
-	exynos_encoder->dpms = DRM_MODE_DPMS_OFF;
-	exynos_encoder->manager = manager;
+	exynos_encoder->display = display;
 	encoder = &exynos_encoder->drm_encoder;
 	encoder->possible_crtcs = possible_crtcs;
 
@@ -344,149 +191,7 @@ exynos_drm_encoder_create(struct drm_device *dev,
 	return encoder;
 }
 
-struct exynos_drm_manager *exynos_drm_get_manager(struct drm_encoder *encoder)
-{
-	return to_exynos_encoder(encoder)->manager;
-}
-
-void exynos_drm_fn_encoder(struct drm_crtc *crtc, void *data,
-			    void (*fn)(struct drm_encoder *, void *))
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_encoder *encoder;
-	struct exynos_drm_private *private = dev->dev_private;
-	struct exynos_drm_manager *manager;
-
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		/*
-		 * if crtc is detached from encoder, check pipe,
-		 * otherwise check crtc attached to encoder
-		 */
-		if (!encoder->crtc) {
-			manager = to_exynos_encoder(encoder)->manager;
-			if (manager->pipe < 0 ||
-					private->crtc[manager->pipe] != crtc)
-				continue;
-		} else {
-			if (encoder->crtc != crtc)
-				continue;
-		}
-
-		fn(encoder, data);
-	}
-}
-
-void exynos_drm_enable_vblank(struct drm_encoder *encoder, void *data)
-{
-	struct exynos_drm_manager *manager =
-		to_exynos_encoder(encoder)->manager;
-	struct exynos_drm_manager_ops *manager_ops = manager->ops;
-	int crtc = *(int *)data;
-
-	if (manager->pipe != crtc)
-		return;
-
-	if (manager_ops->enable_vblank)
-		manager_ops->enable_vblank(manager->dev);
-}
-
-void exynos_drm_disable_vblank(struct drm_encoder *encoder, void *data)
-{
-	struct exynos_drm_manager *manager =
-		to_exynos_encoder(encoder)->manager;
-	struct exynos_drm_manager_ops *manager_ops = manager->ops;
-	int crtc = *(int *)data;
-
-	if (manager->pipe != crtc)
-		return;
-
-	if (manager_ops->disable_vblank)
-		manager_ops->disable_vblank(manager->dev);
-}
-
-void exynos_drm_encoder_crtc_dpms(struct drm_encoder *encoder, void *data)
-{
-	struct exynos_drm_encoder *exynos_encoder = to_exynos_encoder(encoder);
-	struct exynos_drm_manager *manager = exynos_encoder->manager;
-	struct exynos_drm_manager_ops *manager_ops = manager->ops;
-	int mode = *(int *)data;
-
-	if (manager_ops && manager_ops->dpms)
-		manager_ops->dpms(manager->dev, mode);
-
-	/*
-	 * if this condition is ok then it means that the crtc is already
-	 * detached from encoder and last function for detaching is properly
-	 * done, so clear pipe from manager to prevent repeated call.
-	 */
-	if (mode > DRM_MODE_DPMS_ON) {
-		if (!encoder->crtc)
-			manager->pipe = -1;
-	}
-}
-
-void exynos_drm_encoder_crtc_pipe(struct drm_encoder *encoder, void *data)
-{
-	struct exynos_drm_manager *manager =
-		to_exynos_encoder(encoder)->manager;
-	int pipe = *(int *)data;
-
-	/*
-	 * when crtc is detached from encoder, this pipe is used
-	 * to select manager operation
-	 */
-	manager->pipe = pipe;
-}
-
-void exynos_drm_encoder_plane_mode_set(struct drm_encoder *encoder, void *data)
-{
-	struct exynos_drm_manager *manager =
-		to_exynos_encoder(encoder)->manager;
-	struct exynos_drm_overlay_ops *overlay_ops = manager->overlay_ops;
-	struct exynos_drm_overlay *overlay = data;
-
-	if (overlay_ops && overlay_ops->mode_set)
-		overlay_ops->mode_set(manager->dev, overlay);
-}
-
-void exynos_drm_encoder_plane_commit(struct drm_encoder *encoder, void *data)
+struct exynos_drm_display *exynos_drm_get_display(struct drm_encoder *encoder)
 {
-	struct exynos_drm_manager *manager =
-		to_exynos_encoder(encoder)->manager;
-	struct exynos_drm_overlay_ops *overlay_ops = manager->overlay_ops;
-	int zpos = DEFAULT_ZPOS;
-
-	if (data)
-		zpos = *(int *)data;
-
-	if (overlay_ops && overlay_ops->commit)
-		overlay_ops->commit(manager->dev, zpos);
-}
-
-void exynos_drm_encoder_plane_enable(struct drm_encoder *encoder, void *data)
-{
-	struct exynos_drm_manager *manager =
-		to_exynos_encoder(encoder)->manager;
-	struct exynos_drm_overlay_ops *overlay_ops = manager->overlay_ops;
-	int zpos = DEFAULT_ZPOS;
-
-	if (data)
-		zpos = *(int *)data;
-
-	if (overlay_ops && overlay_ops->enable)
-		overlay_ops->enable(manager->dev, zpos);
-}
-
-void exynos_drm_encoder_plane_disable(struct drm_encoder *encoder, void *data)
-{
-	struct exynos_drm_manager *manager =
-		to_exynos_encoder(encoder)->manager;
-	struct exynos_drm_overlay_ops *overlay_ops = manager->overlay_ops;
-	int zpos = DEFAULT_ZPOS;
-
-	if (data)
-		zpos = *(int *)data;
-
-	if (overlay_ops && overlay_ops->disable)
-		overlay_ops->disable(manager->dev, zpos);
+	return to_exynos_encoder(encoder)->display;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_encoder.h b/drivers/gpu/drm/exynos/exynos_drm_encoder.h
index 89e2fb0770a..b7a1620a7e7 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_encoder.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_encoder.h
@@ -18,20 +18,8 @@ struct exynos_drm_manager;
 
 void exynos_drm_encoder_setup(struct drm_device *dev);
 struct drm_encoder *exynos_drm_encoder_create(struct drm_device *dev,
-					       struct exynos_drm_manager *mgr,
-					       unsigned int possible_crtcs);
-struct exynos_drm_manager *
-exynos_drm_get_manager(struct drm_encoder *encoder);
-void exynos_drm_fn_encoder(struct drm_crtc *crtc, void *data,
-			    void (*fn)(struct drm_encoder *, void *));
-void exynos_drm_enable_vblank(struct drm_encoder *encoder, void *data);
-void exynos_drm_disable_vblank(struct drm_encoder *encoder, void *data);
-void exynos_drm_encoder_crtc_dpms(struct drm_encoder *encoder, void *data);
-void exynos_drm_encoder_crtc_pipe(struct drm_encoder *encoder, void *data);
-void exynos_drm_encoder_plane_mode_set(struct drm_encoder *encoder, void *data);
-void exynos_drm_encoder_plane_commit(struct drm_encoder *encoder, void *data);
-void exynos_drm_encoder_plane_enable(struct drm_encoder *encoder, void *data);
-void exynos_drm_encoder_plane_disable(struct drm_encoder *encoder, void *data);
-void exynos_drm_encoder_complete_scanout(struct drm_framebuffer *fb);
+			struct exynos_drm_display *mgr,
+			unsigned long possible_crtcs);
+struct exynos_drm_display *exynos_drm_get_display(struct drm_encoder *encoder);
 
 #endif
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fb.c b/drivers/gpu/drm/exynos/exynos_drm_fb.c
index ea39e0ef2ae..65a22cad7b3 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fb.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fb.c
@@ -20,9 +20,10 @@
 
 #include "exynos_drm_drv.h"
 #include "exynos_drm_fb.h"
+#include "exynos_drm_fbdev.h"
 #include "exynos_drm_gem.h"
 #include "exynos_drm_iommu.h"
-#include "exynos_drm_encoder.h"
+#include "exynos_drm_crtc.h"
 
 #define to_exynos_fb(x)	container_of(x, struct exynos_drm_fb, fb)
 
@@ -71,7 +72,7 @@ static void exynos_drm_fb_destroy(struct drm_framebuffer *fb)
 	unsigned int i;
 
 	/* make sure that overlay data are updated before relesing fb. */
-	exynos_drm_encoder_complete_scanout(fb);
+	exynos_drm_crtc_complete_scanout(fb);
 
 	drm_framebuffer_cleanup(fb);
 
@@ -300,6 +301,8 @@ static void exynos_drm_output_poll_changed(struct drm_device *dev)
 
 	if (fb_helper)
 		drm_fb_helper_hotplug_event(fb_helper);
+	else
+		exynos_drm_fbdev_init(dev);
 }
 
 static const struct drm_mode_config_funcs exynos_drm_mode_config_funcs = {
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fbdev.c b/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
index e7c2f2d07f1..d771b467cf0 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fbdev.c
@@ -90,7 +90,7 @@ static int exynos_drm_fbdev_update(struct drm_fb_helper *helper,
 	/* RGB formats use only one buffer */
 	buffer = exynos_drm_fb_buffer(fb, 0);
 	if (!buffer) {
-		DRM_LOG_KMS("buffer is null.\n");
+		DRM_DEBUG_KMS("buffer is null.\n");
 		return -EFAULT;
 	}
 
@@ -121,16 +121,8 @@ static int exynos_drm_fbdev_update(struct drm_fb_helper *helper,
 	offset = fbi->var.xoffset * (fb->bits_per_pixel >> 3);
 	offset += fbi->var.yoffset * fb->pitches[0];
 
-	dev->mode_config.fb_base = (resource_size_t)buffer->dma_addr;
 	fbi->screen_base = buffer->kvaddr + offset;
-	if (is_drm_iommu_supported(dev))
-		fbi->fix.smem_start = (unsigned long)
-			(page_to_phys(sg_page(buffer->sgt->sgl)) + offset);
-	else
-		fbi->fix.smem_start = (unsigned long)buffer->dma_addr;
-
 	fbi->screen_size = size;
-	fbi->fix.smem_len = size;
 
 	return 0;
 }
@@ -237,6 +229,24 @@ static struct drm_fb_helper_funcs exynos_drm_fb_helper_funcs = {
 	.fb_probe =	exynos_drm_fbdev_create,
 };
 
+static bool exynos_drm_fbdev_is_anything_connected(struct drm_device *dev)
+{
+	struct drm_connector *connector;
+	bool ret = false;
+
+	mutex_lock(&dev->mode_config.mutex);
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		if (connector->status != connector_status_connected)
+			continue;
+
+		ret = true;
+		break;
+	}
+	mutex_unlock(&dev->mode_config.mutex);
+
+	return ret;
+}
+
 int exynos_drm_fbdev_init(struct drm_device *dev)
 {
 	struct exynos_drm_fbdev *fbdev;
@@ -248,6 +258,9 @@ int exynos_drm_fbdev_init(struct drm_device *dev)
 	if (!dev->mode_config.num_crtc || !dev->mode_config.num_connector)
 		return 0;
 
+	if (!exynos_drm_fbdev_is_anything_connected(dev))
+		return 0;
+
 	fbdev = kzalloc(sizeof(*fbdev), GFP_KERNEL);
 	if (!fbdev)
 		return -ENOMEM;
@@ -354,7 +367,5 @@ void exynos_drm_fbdev_restore_mode(struct drm_device *dev)
 	if (!private || !private->fb_helper)
 		return;
 
-	drm_modeset_lock_all(dev);
-	drm_fb_helper_restore_fbdev_mode(private->fb_helper);
-	drm_modeset_unlock_all(dev);
+	drm_fb_helper_restore_fbdev_mode_unlocked(private->fb_helper);
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimc.c b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
index 30d76b2ff9c..831dde9034c 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
@@ -18,6 +18,7 @@
 #include <linux/clk.h>
 #include <linux/pm_runtime.h>
 #include <linux/of.h>
+#include <linux/spinlock.h>
 
 #include <drm/drmP.h>
 #include <drm/exynos_drm.h>
@@ -57,7 +58,6 @@
 #define FIMC_SHFACTOR	10
 #define FIMC_BUF_STOP	1
 #define FIMC_BUF_START	2
-#define FIMC_REG_SZ		32
 #define FIMC_WIDTH_ITU_709	1280
 #define FIMC_REFRESH_MAX	60
 #define FIMC_REFRESH_MIN	12
@@ -69,9 +69,6 @@
 #define get_fimc_context(dev)	platform_get_drvdata(to_platform_device(dev))
 #define get_ctx_from_ippdrv(ippdrv)	container_of(ippdrv,\
 					struct fimc_context, ippdrv);
-#define fimc_read(offset)		readl(ctx->regs + (offset))
-#define fimc_write(cfg, offset)	writel(cfg, ctx->regs + (offset))
-
 enum fimc_wb {
 	FIMC_WB_NONE,
 	FIMC_WB_A,
@@ -161,7 +158,7 @@ struct fimc_context {
 	struct exynos_drm_ippdrv	ippdrv;
 	struct resource	*regs_res;
 	void __iomem	*regs;
-	struct mutex	lock;
+	spinlock_t	lock;
 	struct clk	*clocks[FIMC_CLKS_MAX];
 	u32		clk_frequency;
 	struct regmap	*sysreg;
@@ -172,39 +169,53 @@ struct fimc_context {
 	bool	suspended;
 };
 
+static u32 fimc_read(struct fimc_context *ctx, u32 reg)
+{
+	return readl(ctx->regs + reg);
+}
+
+static void fimc_write(struct fimc_context *ctx, u32 val, u32 reg)
+{
+	writel(val, ctx->regs + reg);
+}
+
+static void fimc_set_bits(struct fimc_context *ctx, u32 reg, u32 bits)
+{
+	void __iomem *r = ctx->regs + reg;
+
+	writel(readl(r) | bits, r);
+}
+
+static void fimc_clear_bits(struct fimc_context *ctx, u32 reg, u32 bits)
+{
+	void __iomem *r = ctx->regs + reg;
+
+	writel(readl(r) & ~bits, r);
+}
+
 static void fimc_sw_reset(struct fimc_context *ctx)
 {
 	u32 cfg;
 
 	/* stop dma operation */
-	cfg = fimc_read(EXYNOS_CISTATUS);
-	if (EXYNOS_CISTATUS_GET_ENVID_STATUS(cfg)) {
-		cfg = fimc_read(EXYNOS_MSCTRL);
-		cfg &= ~EXYNOS_MSCTRL_ENVID;
-		fimc_write(cfg, EXYNOS_MSCTRL);
-	}
+	cfg = fimc_read(ctx, EXYNOS_CISTATUS);
+	if (EXYNOS_CISTATUS_GET_ENVID_STATUS(cfg))
+		fimc_clear_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
 
-	cfg = fimc_read(EXYNOS_CISRCFMT);
-	cfg |= EXYNOS_CISRCFMT_ITU601_8BIT;
-	fimc_write(cfg, EXYNOS_CISRCFMT);
+	fimc_set_bits(ctx, EXYNOS_CISRCFMT, EXYNOS_CISRCFMT_ITU601_8BIT);
 
 	/* disable image capture */
-	cfg = fimc_read(EXYNOS_CIIMGCPT);
-	cfg &= ~(EXYNOS_CIIMGCPT_IMGCPTEN_SC | EXYNOS_CIIMGCPT_IMGCPTEN);
-	fimc_write(cfg, EXYNOS_CIIMGCPT);
+	fimc_clear_bits(ctx, EXYNOS_CIIMGCPT,
+		EXYNOS_CIIMGCPT_IMGCPTEN_SC | EXYNOS_CIIMGCPT_IMGCPTEN);
 
 	/* s/w reset */
-	cfg = fimc_read(EXYNOS_CIGCTRL);
-	cfg |= (EXYNOS_CIGCTRL_SWRST);
-	fimc_write(cfg, EXYNOS_CIGCTRL);
+	fimc_set_bits(ctx, EXYNOS_CIGCTRL, EXYNOS_CIGCTRL_SWRST);
 
 	/* s/w reset complete */
-	cfg = fimc_read(EXYNOS_CIGCTRL);
-	cfg &= ~EXYNOS_CIGCTRL_SWRST;
-	fimc_write(cfg, EXYNOS_CIGCTRL);
+	fimc_clear_bits(ctx, EXYNOS_CIGCTRL, EXYNOS_CIGCTRL_SWRST);
 
 	/* reset sequence */
-	fimc_write(0x0, EXYNOS_CIFCNTSEQ);
+	fimc_write(ctx, 0x0, EXYNOS_CIFCNTSEQ);
 }
 
 static int fimc_set_camblk_fimd0_wb(struct fimc_context *ctx)
@@ -220,7 +231,7 @@ static void fimc_set_type_ctrl(struct fimc_context *ctx, enum fimc_wb wb)
 
 	DRM_DEBUG_KMS("wb[%d]\n", wb);
 
-	cfg = fimc_read(EXYNOS_CIGCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CIGCTRL);
 	cfg &= ~(EXYNOS_CIGCTRL_TESTPATTERN_MASK |
 		EXYNOS_CIGCTRL_SELCAM_ITU_MASK |
 		EXYNOS_CIGCTRL_SELCAM_MIPI_MASK |
@@ -246,7 +257,7 @@ static void fimc_set_type_ctrl(struct fimc_context *ctx, enum fimc_wb wb)
 		break;
 	}
 
-	fimc_write(cfg, EXYNOS_CIGCTRL);
+	fimc_write(ctx, cfg, EXYNOS_CIGCTRL);
 }
 
 static void fimc_set_polarity(struct fimc_context *ctx,
@@ -259,7 +270,7 @@ static void fimc_set_polarity(struct fimc_context *ctx,
 	DRM_DEBUG_KMS("inv_href[%d]inv_hsync[%d]\n",
 		pol->inv_href, pol->inv_hsync);
 
-	cfg = fimc_read(EXYNOS_CIGCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CIGCTRL);
 	cfg &= ~(EXYNOS_CIGCTRL_INVPOLPCLK | EXYNOS_CIGCTRL_INVPOLVSYNC |
 		 EXYNOS_CIGCTRL_INVPOLHREF | EXYNOS_CIGCTRL_INVPOLHSYNC);
 
@@ -272,7 +283,7 @@ static void fimc_set_polarity(struct fimc_context *ctx,
 	if (pol->inv_hsync)
 		cfg |= EXYNOS_CIGCTRL_INVPOLHSYNC;
 
-	fimc_write(cfg, EXYNOS_CIGCTRL);
+	fimc_write(ctx, cfg, EXYNOS_CIGCTRL);
 }
 
 static void fimc_handle_jpeg(struct fimc_context *ctx, bool enable)
@@ -281,70 +292,54 @@ static void fimc_handle_jpeg(struct fimc_context *ctx, bool enable)
 
 	DRM_DEBUG_KMS("enable[%d]\n", enable);
 
-	cfg = fimc_read(EXYNOS_CIGCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CIGCTRL);
 	if (enable)
 		cfg |= EXYNOS_CIGCTRL_CAM_JPEG;
 	else
 		cfg &= ~EXYNOS_CIGCTRL_CAM_JPEG;
 
-	fimc_write(cfg, EXYNOS_CIGCTRL);
+	fimc_write(ctx, cfg, EXYNOS_CIGCTRL);
 }
 
-static void fimc_handle_irq(struct fimc_context *ctx, bool enable,
-		bool overflow, bool level)
+static void fimc_mask_irq(struct fimc_context *ctx, bool enable)
 {
 	u32 cfg;
 
-	DRM_DEBUG_KMS("enable[%d]overflow[%d]level[%d]\n",
-			enable, overflow, level);
+	DRM_DEBUG_KMS("enable[%d]\n", enable);
 
-	cfg = fimc_read(EXYNOS_CIGCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CIGCTRL);
 	if (enable) {
-		cfg &= ~(EXYNOS_CIGCTRL_IRQ_OVFEN | EXYNOS_CIGCTRL_IRQ_LEVEL);
-		cfg |= EXYNOS_CIGCTRL_IRQ_ENABLE;
-		if (overflow)
-			cfg |= EXYNOS_CIGCTRL_IRQ_OVFEN;
-		if (level)
-			cfg |= EXYNOS_CIGCTRL_IRQ_LEVEL;
+		cfg &= ~EXYNOS_CIGCTRL_IRQ_OVFEN;
+		cfg |= EXYNOS_CIGCTRL_IRQ_ENABLE | EXYNOS_CIGCTRL_IRQ_LEVEL;
 	} else
-		cfg &= ~(EXYNOS_CIGCTRL_IRQ_OVFEN | EXYNOS_CIGCTRL_IRQ_ENABLE);
-
-	fimc_write(cfg, EXYNOS_CIGCTRL);
+		cfg &= ~EXYNOS_CIGCTRL_IRQ_ENABLE;
+	fimc_write(ctx, cfg, EXYNOS_CIGCTRL);
 }
 
 static void fimc_clear_irq(struct fimc_context *ctx)
 {
-	u32 cfg;
-
-	cfg = fimc_read(EXYNOS_CIGCTRL);
-	cfg |= EXYNOS_CIGCTRL_IRQ_CLR;
-	fimc_write(cfg, EXYNOS_CIGCTRL);
+	fimc_set_bits(ctx, EXYNOS_CIGCTRL, EXYNOS_CIGCTRL_IRQ_CLR);
 }
 
 static bool fimc_check_ovf(struct fimc_context *ctx)
 {
 	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	u32 cfg, status, flag;
+	u32 status, flag;
 
-	status = fimc_read(EXYNOS_CISTATUS);
+	status = fimc_read(ctx, EXYNOS_CISTATUS);
 	flag = EXYNOS_CISTATUS_OVFIY | EXYNOS_CISTATUS_OVFICB |
 		EXYNOS_CISTATUS_OVFICR;
 
 	DRM_DEBUG_KMS("flag[0x%x]\n", flag);
 
 	if (status & flag) {
-		cfg = fimc_read(EXYNOS_CIWDOFST);
-		cfg |= (EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
+		fimc_set_bits(ctx, EXYNOS_CIWDOFST,
+			EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
 			EXYNOS_CIWDOFST_CLROVFICR);
-
-		fimc_write(cfg, EXYNOS_CIWDOFST);
-
-		cfg = fimc_read(EXYNOS_CIWDOFST);
-		cfg &= ~(EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
+		fimc_clear_bits(ctx, EXYNOS_CIWDOFST,
+			EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
 			EXYNOS_CIWDOFST_CLROVFICR);
 
-		fimc_write(cfg, EXYNOS_CIWDOFST);
-
 		dev_err(ippdrv->dev, "occurred overflow at %d, status 0x%x.\n",
 			ctx->id, status);
 		return true;
@@ -357,7 +352,7 @@ static bool fimc_check_frame_end(struct fimc_context *ctx)
 {
 	u32 cfg;
 
-	cfg = fimc_read(EXYNOS_CISTATUS);
+	cfg = fimc_read(ctx, EXYNOS_CISTATUS);
 
 	DRM_DEBUG_KMS("cfg[0x%x]\n", cfg);
 
@@ -365,7 +360,7 @@ static bool fimc_check_frame_end(struct fimc_context *ctx)
 		return false;
 
 	cfg &= ~(EXYNOS_CISTATUS_FRAMEEND);
-	fimc_write(cfg, EXYNOS_CISTATUS);
+	fimc_write(ctx, cfg, EXYNOS_CISTATUS);
 
 	return true;
 }
@@ -375,7 +370,7 @@ static int fimc_get_buf_id(struct fimc_context *ctx)
 	u32 cfg;
 	int frame_cnt, buf_id;
 
-	cfg = fimc_read(EXYNOS_CISTATUS2);
+	cfg = fimc_read(ctx, EXYNOS_CISTATUS2);
 	frame_cnt = EXYNOS_CISTATUS2_GET_FRAMECOUNT_BEFORE(cfg);
 
 	if (frame_cnt == 0)
@@ -402,13 +397,13 @@ static void fimc_handle_lastend(struct fimc_context *ctx, bool enable)
 
 	DRM_DEBUG_KMS("enable[%d]\n", enable);
 
-	cfg = fimc_read(EXYNOS_CIOCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CIOCTRL);
 	if (enable)
 		cfg |= EXYNOS_CIOCTRL_LASTENDEN;
 	else
 		cfg &= ~EXYNOS_CIOCTRL_LASTENDEN;
 
-	fimc_write(cfg, EXYNOS_CIOCTRL);
+	fimc_write(ctx, cfg, EXYNOS_CIOCTRL);
 }
 
 
@@ -420,18 +415,18 @@ static int fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
 
 	/* RGB */
-	cfg = fimc_read(EXYNOS_CISCCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CISCCTRL);
 	cfg &= ~EXYNOS_CISCCTRL_INRGB_FMT_RGB_MASK;
 
 	switch (fmt) {
 	case DRM_FORMAT_RGB565:
 		cfg |= EXYNOS_CISCCTRL_INRGB_FMT_RGB565;
-		fimc_write(cfg, EXYNOS_CISCCTRL);
+		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
 		return 0;
 	case DRM_FORMAT_RGB888:
 	case DRM_FORMAT_XRGB8888:
 		cfg |= EXYNOS_CISCCTRL_INRGB_FMT_RGB888;
-		fimc_write(cfg, EXYNOS_CISCCTRL);
+		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
 		return 0;
 	default:
 		/* bypass */
@@ -439,7 +434,7 @@ static int fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 	}
 
 	/* YUV */
-	cfg = fimc_read(EXYNOS_MSCTRL);
+	cfg = fimc_read(ctx, EXYNOS_MSCTRL);
 	cfg &= ~(EXYNOS_MSCTRL_ORDER2P_SHIFT_MASK |
 		EXYNOS_MSCTRL_C_INT_IN_2PLANE |
 		EXYNOS_MSCTRL_ORDER422_YCBYCR);
@@ -479,7 +474,7 @@ static int fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 		return -EINVAL;
 	}
 
-	fimc_write(cfg, EXYNOS_MSCTRL);
+	fimc_write(ctx, cfg, EXYNOS_MSCTRL);
 
 	return 0;
 }
@@ -492,7 +487,7 @@ static int fimc_src_set_fmt(struct device *dev, u32 fmt)
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
 
-	cfg = fimc_read(EXYNOS_MSCTRL);
+	cfg = fimc_read(ctx, EXYNOS_MSCTRL);
 	cfg &= ~EXYNOS_MSCTRL_INFORMAT_RGB;
 
 	switch (fmt) {
@@ -527,9 +522,9 @@ static int fimc_src_set_fmt(struct device *dev, u32 fmt)
 		return -EINVAL;
 	}
 
-	fimc_write(cfg, EXYNOS_MSCTRL);
+	fimc_write(ctx, cfg, EXYNOS_MSCTRL);
 
-	cfg = fimc_read(EXYNOS_CIDMAPARAM);
+	cfg = fimc_read(ctx, EXYNOS_CIDMAPARAM);
 	cfg &= ~EXYNOS_CIDMAPARAM_R_MODE_MASK;
 
 	if (fmt == DRM_FORMAT_NV12MT)
@@ -537,7 +532,7 @@ static int fimc_src_set_fmt(struct device *dev, u32 fmt)
 	else
 		cfg |= EXYNOS_CIDMAPARAM_R_MODE_LINEAR;
 
-	fimc_write(cfg, EXYNOS_CIDMAPARAM);
+	fimc_write(ctx, cfg, EXYNOS_CIDMAPARAM);
 
 	return fimc_src_set_fmt_order(ctx, fmt);
 }
@@ -552,11 +547,11 @@ static int fimc_src_set_transf(struct device *dev,
 
 	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
 
-	cfg1 = fimc_read(EXYNOS_MSCTRL);
+	cfg1 = fimc_read(ctx, EXYNOS_MSCTRL);
 	cfg1 &= ~(EXYNOS_MSCTRL_FLIP_X_MIRROR |
 		EXYNOS_MSCTRL_FLIP_Y_MIRROR);
 
-	cfg2 = fimc_read(EXYNOS_CITRGFMT);
+	cfg2 = fimc_read(ctx, EXYNOS_CITRGFMT);
 	cfg2 &= ~EXYNOS_CITRGFMT_INROT90_CLOCKWISE;
 
 	switch (degree) {
@@ -595,8 +590,8 @@ static int fimc_src_set_transf(struct device *dev,
 		return -EINVAL;
 	}
 
-	fimc_write(cfg1, EXYNOS_MSCTRL);
-	fimc_write(cfg2, EXYNOS_CITRGFMT);
+	fimc_write(ctx, cfg1, EXYNOS_MSCTRL);
+	fimc_write(ctx, cfg2, EXYNOS_CITRGFMT);
 	*swap = (cfg2 & EXYNOS_CITRGFMT_INROT90_CLOCKWISE) ? 1 : 0;
 
 	return 0;
@@ -621,17 +616,17 @@ static int fimc_set_window(struct fimc_context *ctx,
 	 * set window offset 1, 2 size
 	 * check figure 43-21 in user manual
 	 */
-	cfg = fimc_read(EXYNOS_CIWDOFST);
+	cfg = fimc_read(ctx, EXYNOS_CIWDOFST);
 	cfg &= ~(EXYNOS_CIWDOFST_WINHOROFST_MASK |
 		EXYNOS_CIWDOFST_WINVEROFST_MASK);
 	cfg |= (EXYNOS_CIWDOFST_WINHOROFST(h1) |
 		EXYNOS_CIWDOFST_WINVEROFST(v1));
 	cfg |= EXYNOS_CIWDOFST_WINOFSEN;
-	fimc_write(cfg, EXYNOS_CIWDOFST);
+	fimc_write(ctx, cfg, EXYNOS_CIWDOFST);
 
 	cfg = (EXYNOS_CIWDOFST2_WINHOROFST2(h2) |
 		EXYNOS_CIWDOFST2_WINVEROFST2(v2));
-	fimc_write(cfg, EXYNOS_CIWDOFST2);
+	fimc_write(ctx, cfg, EXYNOS_CIWDOFST2);
 
 	return 0;
 }
@@ -651,7 +646,7 @@ static int fimc_src_set_size(struct device *dev, int swap,
 	cfg = (EXYNOS_ORGISIZE_HORIZONTAL(img_sz.hsize) |
 		EXYNOS_ORGISIZE_VERTICAL(img_sz.vsize));
 
-	fimc_write(cfg, EXYNOS_ORGISIZE);
+	fimc_write(ctx, cfg, EXYNOS_ORGISIZE);
 
 	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", pos->x, pos->y, pos->w, pos->h);
 
@@ -663,12 +658,12 @@ static int fimc_src_set_size(struct device *dev, int swap,
 	}
 
 	/* set input DMA image size */
-	cfg = fimc_read(EXYNOS_CIREAL_ISIZE);
+	cfg = fimc_read(ctx, EXYNOS_CIREAL_ISIZE);
 	cfg &= ~(EXYNOS_CIREAL_ISIZE_HEIGHT_MASK |
 		EXYNOS_CIREAL_ISIZE_WIDTH_MASK);
 	cfg |= (EXYNOS_CIREAL_ISIZE_WIDTH(img_pos.w) |
 		EXYNOS_CIREAL_ISIZE_HEIGHT(img_pos.h));
-	fimc_write(cfg, EXYNOS_CIREAL_ISIZE);
+	fimc_write(ctx, cfg, EXYNOS_CIREAL_ISIZE);
 
 	/*
 	 * set input FIFO image size
@@ -677,18 +672,18 @@ static int fimc_src_set_size(struct device *dev, int swap,
 	cfg = (EXYNOS_CISRCFMT_ITU601_8BIT |
 		EXYNOS_CISRCFMT_SOURCEHSIZE(img_sz.hsize) |
 		EXYNOS_CISRCFMT_SOURCEVSIZE(img_sz.vsize));
-	fimc_write(cfg, EXYNOS_CISRCFMT);
+	fimc_write(ctx, cfg, EXYNOS_CISRCFMT);
 
 	/* offset Y(RGB), Cb, Cr */
 	cfg = (EXYNOS_CIIYOFF_HORIZONTAL(img_pos.x) |
 		EXYNOS_CIIYOFF_VERTICAL(img_pos.y));
-	fimc_write(cfg, EXYNOS_CIIYOFF);
+	fimc_write(ctx, cfg, EXYNOS_CIIYOFF);
 	cfg = (EXYNOS_CIICBOFF_HORIZONTAL(img_pos.x) |
 		EXYNOS_CIICBOFF_VERTICAL(img_pos.y));
-	fimc_write(cfg, EXYNOS_CIICBOFF);
+	fimc_write(ctx, cfg, EXYNOS_CIICBOFF);
 	cfg = (EXYNOS_CIICROFF_HORIZONTAL(img_pos.x) |
 		EXYNOS_CIICROFF_VERTICAL(img_pos.y));
-	fimc_write(cfg, EXYNOS_CIICROFF);
+	fimc_write(ctx, cfg, EXYNOS_CIICROFF);
 
 	return fimc_set_window(ctx, &img_pos, &img_sz);
 }
@@ -722,25 +717,25 @@ static int fimc_src_set_addr(struct device *dev,
 	switch (buf_type) {
 	case IPP_BUF_ENQUEUE:
 		config = &property->config[EXYNOS_DRM_OPS_SRC];
-		fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_Y],
+		fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_Y],
 			EXYNOS_CIIYSA(buf_id));
 
 		if (config->fmt == DRM_FORMAT_YVU420) {
-			fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_CR],
+			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
 				EXYNOS_CIICBSA(buf_id));
-			fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_CB],
+			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
 				EXYNOS_CIICRSA(buf_id));
 		} else {
-			fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_CB],
+			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
 				EXYNOS_CIICBSA(buf_id));
-			fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_CR],
+			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
 				EXYNOS_CIICRSA(buf_id));
 		}
 		break;
 	case IPP_BUF_DEQUEUE:
-		fimc_write(0x0, EXYNOS_CIIYSA(buf_id));
-		fimc_write(0x0, EXYNOS_CIICBSA(buf_id));
-		fimc_write(0x0, EXYNOS_CIICRSA(buf_id));
+		fimc_write(ctx, 0x0, EXYNOS_CIIYSA(buf_id));
+		fimc_write(ctx, 0x0, EXYNOS_CIICBSA(buf_id));
+		fimc_write(ctx, 0x0, EXYNOS_CIICRSA(buf_id));
 		break;
 	default:
 		/* bypass */
@@ -765,22 +760,22 @@ static int fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
 
 	/* RGB */
-	cfg = fimc_read(EXYNOS_CISCCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CISCCTRL);
 	cfg &= ~EXYNOS_CISCCTRL_OUTRGB_FMT_RGB_MASK;
 
 	switch (fmt) {
 	case DRM_FORMAT_RGB565:
 		cfg |= EXYNOS_CISCCTRL_OUTRGB_FMT_RGB565;
-		fimc_write(cfg, EXYNOS_CISCCTRL);
+		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
 		return 0;
 	case DRM_FORMAT_RGB888:
 		cfg |= EXYNOS_CISCCTRL_OUTRGB_FMT_RGB888;
-		fimc_write(cfg, EXYNOS_CISCCTRL);
+		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
 		return 0;
 	case DRM_FORMAT_XRGB8888:
 		cfg |= (EXYNOS_CISCCTRL_OUTRGB_FMT_RGB888 |
 			EXYNOS_CISCCTRL_EXTRGB_EXTENSION);
-		fimc_write(cfg, EXYNOS_CISCCTRL);
+		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
 		break;
 	default:
 		/* bypass */
@@ -788,7 +783,7 @@ static int fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 	}
 
 	/* YUV */
-	cfg = fimc_read(EXYNOS_CIOCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CIOCTRL);
 	cfg &= ~(EXYNOS_CIOCTRL_ORDER2P_MASK |
 		EXYNOS_CIOCTRL_ORDER422_MASK |
 		EXYNOS_CIOCTRL_YCBCR_PLANE_MASK);
@@ -830,7 +825,7 @@ static int fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 		return -EINVAL;
 	}
 
-	fimc_write(cfg, EXYNOS_CIOCTRL);
+	fimc_write(ctx, cfg, EXYNOS_CIOCTRL);
 
 	return 0;
 }
@@ -843,16 +838,16 @@ static int fimc_dst_set_fmt(struct device *dev, u32 fmt)
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
 
-	cfg = fimc_read(EXYNOS_CIEXTEN);
+	cfg = fimc_read(ctx, EXYNOS_CIEXTEN);
 
 	if (fmt == DRM_FORMAT_AYUV) {
 		cfg |= EXYNOS_CIEXTEN_YUV444_OUT;
-		fimc_write(cfg, EXYNOS_CIEXTEN);
+		fimc_write(ctx, cfg, EXYNOS_CIEXTEN);
 	} else {
 		cfg &= ~EXYNOS_CIEXTEN_YUV444_OUT;
-		fimc_write(cfg, EXYNOS_CIEXTEN);
+		fimc_write(ctx, cfg, EXYNOS_CIEXTEN);
 
-		cfg = fimc_read(EXYNOS_CITRGFMT);
+		cfg = fimc_read(ctx, EXYNOS_CITRGFMT);
 		cfg &= ~EXYNOS_CITRGFMT_OUTFORMAT_MASK;
 
 		switch (fmt) {
@@ -885,10 +880,10 @@ static int fimc_dst_set_fmt(struct device *dev, u32 fmt)
 			return -EINVAL;
 		}
 
-		fimc_write(cfg, EXYNOS_CITRGFMT);
+		fimc_write(ctx, cfg, EXYNOS_CITRGFMT);
 	}
 
-	cfg = fimc_read(EXYNOS_CIDMAPARAM);
+	cfg = fimc_read(ctx, EXYNOS_CIDMAPARAM);
 	cfg &= ~EXYNOS_CIDMAPARAM_W_MODE_MASK;
 
 	if (fmt == DRM_FORMAT_NV12MT)
@@ -896,7 +891,7 @@ static int fimc_dst_set_fmt(struct device *dev, u32 fmt)
 	else
 		cfg |= EXYNOS_CIDMAPARAM_W_MODE_LINEAR;
 
-	fimc_write(cfg, EXYNOS_CIDMAPARAM);
+	fimc_write(ctx, cfg, EXYNOS_CIDMAPARAM);
 
 	return fimc_dst_set_fmt_order(ctx, fmt);
 }
@@ -911,7 +906,7 @@ static int fimc_dst_set_transf(struct device *dev,
 
 	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
 
-	cfg = fimc_read(EXYNOS_CITRGFMT);
+	cfg = fimc_read(ctx, EXYNOS_CITRGFMT);
 	cfg &= ~EXYNOS_CITRGFMT_FLIP_MASK;
 	cfg &= ~EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE;
 
@@ -951,53 +946,23 @@ static int fimc_dst_set_transf(struct device *dev,
 		return -EINVAL;
 	}
 
-	fimc_write(cfg, EXYNOS_CITRGFMT);
+	fimc_write(ctx, cfg, EXYNOS_CITRGFMT);
 	*swap = (cfg & EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE) ? 1 : 0;
 
 	return 0;
 }
 
-static int fimc_get_ratio_shift(u32 src, u32 dst, u32 *ratio, u32 *shift)
-{
-	DRM_DEBUG_KMS("src[%d]dst[%d]\n", src, dst);
-
-	if (src >= dst * 64) {
-		DRM_ERROR("failed to make ratio and shift.\n");
-		return -EINVAL;
-	} else if (src >= dst * 32) {
-		*ratio = 32;
-		*shift = 5;
-	} else if (src >= dst * 16) {
-		*ratio = 16;
-		*shift = 4;
-	} else if (src >= dst * 8) {
-		*ratio = 8;
-		*shift = 3;
-	} else if (src >= dst * 4) {
-		*ratio = 4;
-		*shift = 2;
-	} else if (src >= dst * 2) {
-		*ratio = 2;
-		*shift = 1;
-	} else {
-		*ratio = 1;
-		*shift = 0;
-	}
-
-	return 0;
-}
-
 static int fimc_set_prescaler(struct fimc_context *ctx, struct fimc_scaler *sc,
 		struct drm_exynos_pos *src, struct drm_exynos_pos *dst)
 {
 	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg, cfg_ext, shfactor;
 	u32 pre_dst_width, pre_dst_height;
-	u32 pre_hratio, hfactor, pre_vratio, vfactor;
+	u32 hfactor, vfactor;
 	int ret = 0;
 	u32 src_w, src_h, dst_w, dst_h;
 
-	cfg_ext = fimc_read(EXYNOS_CITRGFMT);
+	cfg_ext = fimc_read(ctx, EXYNOS_CITRGFMT);
 	if (cfg_ext & EXYNOS_CITRGFMT_INROT90_CLOCKWISE) {
 		src_w = src->h;
 		src_h = src->w;
@@ -1014,24 +979,24 @@ static int fimc_set_prescaler(struct fimc_context *ctx, struct fimc_scaler *sc,
 		dst_h = dst->h;
 	}
 
-	ret = fimc_get_ratio_shift(src_w, dst_w, &pre_hratio, &hfactor);
-	if (ret) {
+	/* fimc_ippdrv_check_property assures that dividers are not null */
+	hfactor = fls(src_w / dst_w / 2);
+	if (hfactor > FIMC_SHFACTOR / 2) {
 		dev_err(ippdrv->dev, "failed to get ratio horizontal.\n");
-		return ret;
+		return -EINVAL;
 	}
 
-	ret = fimc_get_ratio_shift(src_h, dst_h, &pre_vratio, &vfactor);
-	if (ret) {
+	vfactor = fls(src_h / dst_h / 2);
+	if (vfactor > FIMC_SHFACTOR / 2) {
 		dev_err(ippdrv->dev, "failed to get ratio vertical.\n");
-		return ret;
+		return -EINVAL;
 	}
 
-	pre_dst_width = src_w / pre_hratio;
-	pre_dst_height = src_h / pre_vratio;
+	pre_dst_width = src_w >> hfactor;
+	pre_dst_height = src_h >> vfactor;
 	DRM_DEBUG_KMS("pre_dst_width[%d]pre_dst_height[%d]\n",
 		pre_dst_width, pre_dst_height);
-	DRM_DEBUG_KMS("pre_hratio[%d]hfactor[%d]pre_vratio[%d]vfactor[%d]\n",
-		pre_hratio, hfactor, pre_vratio, vfactor);
+	DRM_DEBUG_KMS("hfactor[%d]vfactor[%d]\n", hfactor, vfactor);
 
 	sc->hratio = (src_w << 14) / (dst_w << hfactor);
 	sc->vratio = (src_h << 14) / (dst_h << vfactor);
@@ -1044,13 +1009,13 @@ static int fimc_set_prescaler(struct fimc_context *ctx, struct fimc_scaler *sc,
 	DRM_DEBUG_KMS("shfactor[%d]\n", shfactor);
 
 	cfg = (EXYNOS_CISCPRERATIO_SHFACTOR(shfactor) |
-		EXYNOS_CISCPRERATIO_PREHORRATIO(pre_hratio) |
-		EXYNOS_CISCPRERATIO_PREVERRATIO(pre_vratio));
-	fimc_write(cfg, EXYNOS_CISCPRERATIO);
+		EXYNOS_CISCPRERATIO_PREHORRATIO(1 << hfactor) |
+		EXYNOS_CISCPRERATIO_PREVERRATIO(1 << vfactor));
+	fimc_write(ctx, cfg, EXYNOS_CISCPRERATIO);
 
 	cfg = (EXYNOS_CISCPREDST_PREDSTWIDTH(pre_dst_width) |
 		EXYNOS_CISCPREDST_PREDSTHEIGHT(pre_dst_height));
-	fimc_write(cfg, EXYNOS_CISCPREDST);
+	fimc_write(ctx, cfg, EXYNOS_CISCPREDST);
 
 	return ret;
 }
@@ -1064,7 +1029,7 @@ static void fimc_set_scaler(struct fimc_context *ctx, struct fimc_scaler *sc)
 	DRM_DEBUG_KMS("hratio[%d]vratio[%d]\n",
 		sc->hratio, sc->vratio);
 
-	cfg = fimc_read(EXYNOS_CISCCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CISCCTRL);
 	cfg &= ~(EXYNOS_CISCCTRL_SCALERBYPASS |
 		EXYNOS_CISCCTRL_SCALEUP_H | EXYNOS_CISCCTRL_SCALEUP_V |
 		EXYNOS_CISCCTRL_MAIN_V_RATIO_MASK |
@@ -1084,14 +1049,14 @@ static void fimc_set_scaler(struct fimc_context *ctx, struct fimc_scaler *sc)
 
 	cfg |= (EXYNOS_CISCCTRL_MAINHORRATIO((sc->hratio >> 6)) |
 		EXYNOS_CISCCTRL_MAINVERRATIO((sc->vratio >> 6)));
-	fimc_write(cfg, EXYNOS_CISCCTRL);
+	fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
 
-	cfg_ext = fimc_read(EXYNOS_CIEXTEN);
+	cfg_ext = fimc_read(ctx, EXYNOS_CIEXTEN);
 	cfg_ext &= ~EXYNOS_CIEXTEN_MAINHORRATIO_EXT_MASK;
 	cfg_ext &= ~EXYNOS_CIEXTEN_MAINVERRATIO_EXT_MASK;
 	cfg_ext |= (EXYNOS_CIEXTEN_MAINHORRATIO_EXT(sc->hratio) |
 		EXYNOS_CIEXTEN_MAINVERRATIO_EXT(sc->vratio));
-	fimc_write(cfg_ext, EXYNOS_CIEXTEN);
+	fimc_write(ctx, cfg_ext, EXYNOS_CIEXTEN);
 }
 
 static int fimc_dst_set_size(struct device *dev, int swap,
@@ -1109,12 +1074,12 @@ static int fimc_dst_set_size(struct device *dev, int swap,
 	cfg = (EXYNOS_ORGOSIZE_HORIZONTAL(img_sz.hsize) |
 		EXYNOS_ORGOSIZE_VERTICAL(img_sz.vsize));
 
-	fimc_write(cfg, EXYNOS_ORGOSIZE);
+	fimc_write(ctx, cfg, EXYNOS_ORGOSIZE);
 
 	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", pos->x, pos->y, pos->w, pos->h);
 
 	/* CSC ITU */
-	cfg = fimc_read(EXYNOS_CIGCTRL);
+	cfg = fimc_read(ctx, EXYNOS_CIGCTRL);
 	cfg &= ~EXYNOS_CIGCTRL_CSC_MASK;
 
 	if (sz->hsize >= FIMC_WIDTH_ITU_709)
@@ -1122,7 +1087,7 @@ static int fimc_dst_set_size(struct device *dev, int swap,
 	else
 		cfg |= EXYNOS_CIGCTRL_CSC_ITU601;
 
-	fimc_write(cfg, EXYNOS_CIGCTRL);
+	fimc_write(ctx, cfg, EXYNOS_CIGCTRL);
 
 	if (swap) {
 		img_pos.w = pos->h;
@@ -1132,41 +1097,38 @@ static int fimc_dst_set_size(struct device *dev, int swap,
 	}
 
 	/* target image size */
-	cfg = fimc_read(EXYNOS_CITRGFMT);
+	cfg = fimc_read(ctx, EXYNOS_CITRGFMT);
 	cfg &= ~(EXYNOS_CITRGFMT_TARGETH_MASK |
 		EXYNOS_CITRGFMT_TARGETV_MASK);
 	cfg |= (EXYNOS_CITRGFMT_TARGETHSIZE(img_pos.w) |
 		EXYNOS_CITRGFMT_TARGETVSIZE(img_pos.h));
-	fimc_write(cfg, EXYNOS_CITRGFMT);
+	fimc_write(ctx, cfg, EXYNOS_CITRGFMT);
 
 	/* target area */
 	cfg = EXYNOS_CITAREA_TARGET_AREA(img_pos.w * img_pos.h);
-	fimc_write(cfg, EXYNOS_CITAREA);
+	fimc_write(ctx, cfg, EXYNOS_CITAREA);
 
 	/* offset Y(RGB), Cb, Cr */
 	cfg = (EXYNOS_CIOYOFF_HORIZONTAL(img_pos.x) |
 		EXYNOS_CIOYOFF_VERTICAL(img_pos.y));
-	fimc_write(cfg, EXYNOS_CIOYOFF);
+	fimc_write(ctx, cfg, EXYNOS_CIOYOFF);
 	cfg = (EXYNOS_CIOCBOFF_HORIZONTAL(img_pos.x) |
 		EXYNOS_CIOCBOFF_VERTICAL(img_pos.y));
-	fimc_write(cfg, EXYNOS_CIOCBOFF);
+	fimc_write(ctx, cfg, EXYNOS_CIOCBOFF);
 	cfg = (EXYNOS_CIOCROFF_HORIZONTAL(img_pos.x) |
 		EXYNOS_CIOCROFF_VERTICAL(img_pos.y));
-	fimc_write(cfg, EXYNOS_CIOCROFF);
+	fimc_write(ctx, cfg, EXYNOS_CIOCROFF);
 
 	return 0;
 }
 
-static int fimc_dst_get_buf_seq(struct fimc_context *ctx)
+static int fimc_dst_get_buf_count(struct fimc_context *ctx)
 {
-	u32 cfg, i, buf_num = 0;
-	u32 mask = 0x00000001;
+	u32 cfg, buf_num;
 
-	cfg = fimc_read(EXYNOS_CIFCNTSEQ);
+	cfg = fimc_read(ctx, EXYNOS_CIFCNTSEQ);
 
-	for (i = 0; i < FIMC_REG_SZ; i++)
-		if (cfg & (mask << i))
-			buf_num++;
+	buf_num = hweight32(cfg);
 
 	DRM_DEBUG_KMS("buf_num[%d]\n", buf_num);
 
@@ -1181,13 +1143,14 @@ static int fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
 	u32 cfg;
 	u32 mask = 0x00000001 << buf_id;
 	int ret = 0;
+	unsigned long flags;
 
 	DRM_DEBUG_KMS("buf_id[%d]buf_type[%d]\n", buf_id, buf_type);
 
-	mutex_lock(&ctx->lock);
+	spin_lock_irqsave(&ctx->lock, flags);
 
 	/* mask register set */
-	cfg = fimc_read(EXYNOS_CIFCNTSEQ);
+	cfg = fimc_read(ctx, EXYNOS_CIFCNTSEQ);
 
 	switch (buf_type) {
 	case IPP_BUF_ENQUEUE:
@@ -1205,20 +1168,20 @@ static int fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
 	/* sequence id */
 	cfg &= ~mask;
 	cfg |= (enable << buf_id);
-	fimc_write(cfg, EXYNOS_CIFCNTSEQ);
+	fimc_write(ctx, cfg, EXYNOS_CIFCNTSEQ);
 
 	/* interrupt enable */
 	if (buf_type == IPP_BUF_ENQUEUE &&
-	    fimc_dst_get_buf_seq(ctx) >= FIMC_BUF_START)
-		fimc_handle_irq(ctx, true, false, true);
+	    fimc_dst_get_buf_count(ctx) >= FIMC_BUF_START)
+		fimc_mask_irq(ctx, true);
 
 	/* interrupt disable */
 	if (buf_type == IPP_BUF_DEQUEUE &&
-	    fimc_dst_get_buf_seq(ctx) <= FIMC_BUF_STOP)
-		fimc_handle_irq(ctx, false, false, true);
+	    fimc_dst_get_buf_count(ctx) <= FIMC_BUF_STOP)
+		fimc_mask_irq(ctx, false);
 
 err_unlock:
-	mutex_unlock(&ctx->lock);
+	spin_unlock_irqrestore(&ctx->lock, flags);
 	return ret;
 }
 
@@ -1252,25 +1215,25 @@ static int fimc_dst_set_addr(struct device *dev,
 	case IPP_BUF_ENQUEUE:
 		config = &property->config[EXYNOS_DRM_OPS_DST];
 
-		fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_Y],
+		fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_Y],
 			EXYNOS_CIOYSA(buf_id));
 
 		if (config->fmt == DRM_FORMAT_YVU420) {
-			fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_CR],
+			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
 				EXYNOS_CIOCBSA(buf_id));
-			fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_CB],
+			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
 				EXYNOS_CIOCRSA(buf_id));
 		} else {
-			fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_CB],
+			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
 				EXYNOS_CIOCBSA(buf_id));
-			fimc_write(buf_info->base[EXYNOS_DRM_PLANAR_CR],
+			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
 				EXYNOS_CIOCRSA(buf_id));
 		}
 		break;
 	case IPP_BUF_DEQUEUE:
-		fimc_write(0x0, EXYNOS_CIOYSA(buf_id));
-		fimc_write(0x0, EXYNOS_CIOCBSA(buf_id));
-		fimc_write(0x0, EXYNOS_CIOCRSA(buf_id));
+		fimc_write(ctx, 0x0, EXYNOS_CIOYSA(buf_id));
+		fimc_write(ctx, 0x0, EXYNOS_CIOCBSA(buf_id));
+		fimc_write(ctx, 0x0, EXYNOS_CIOCRSA(buf_id));
 		break;
 	default:
 		/* bypass */
@@ -1342,11 +1305,7 @@ static irqreturn_t fimc_irq_handler(int irq, void *dev_id)
 
 static int fimc_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
 {
-	struct drm_exynos_ipp_prop_list *prop_list;
-
-	prop_list = devm_kzalloc(ippdrv->dev, sizeof(*prop_list), GFP_KERNEL);
-	if (!prop_list)
-		return -ENOMEM;
+	struct drm_exynos_ipp_prop_list *prop_list = &ippdrv->prop_list;
 
 	prop_list->version = 1;
 	prop_list->writeback = 1;
@@ -1371,8 +1330,6 @@ static int fimc_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
 	prop_list->scale_min.hsize = FIMC_SCALE_MIN;
 	prop_list->scale_min.vsize = FIMC_SCALE_MIN;
 
-	ippdrv->prop_list = prop_list;
-
 	return 0;
 }
 
@@ -1395,7 +1352,7 @@ static int fimc_ippdrv_check_property(struct device *dev,
 {
 	struct fimc_context *ctx = get_fimc_context(dev);
 	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_prop_list *pp = ippdrv->prop_list;
+	struct drm_exynos_ipp_prop_list *pp = &ippdrv->prop_list;
 	struct drm_exynos_ipp_config *config;
 	struct drm_exynos_pos *pos;
 	struct drm_exynos_sz *sz;
@@ -1508,15 +1465,15 @@ static void fimc_clear_addr(struct fimc_context *ctx)
 	int i;
 
 	for (i = 0; i < FIMC_MAX_SRC; i++) {
-		fimc_write(0, EXYNOS_CIIYSA(i));
-		fimc_write(0, EXYNOS_CIICBSA(i));
-		fimc_write(0, EXYNOS_CIICRSA(i));
+		fimc_write(ctx, 0, EXYNOS_CIIYSA(i));
+		fimc_write(ctx, 0, EXYNOS_CIICBSA(i));
+		fimc_write(ctx, 0, EXYNOS_CIICRSA(i));
 	}
 
 	for (i = 0; i < FIMC_MAX_DST; i++) {
-		fimc_write(0, EXYNOS_CIOYSA(i));
-		fimc_write(0, EXYNOS_CIOCBSA(i));
-		fimc_write(0, EXYNOS_CIOCRSA(i));
+		fimc_write(ctx, 0, EXYNOS_CIOYSA(i));
+		fimc_write(ctx, 0, EXYNOS_CIOCBSA(i));
+		fimc_write(ctx, 0, EXYNOS_CIOCRSA(i));
 	}
 }
 
@@ -1556,7 +1513,7 @@ static int fimc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 
 	property = &c_node->property;
 
-	fimc_handle_irq(ctx, true, false, true);
+	fimc_mask_irq(ctx, true);
 
 	for_each_ipp_ops(i) {
 		config = &property->config[i];
@@ -1582,10 +1539,10 @@ static int fimc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 		fimc_handle_lastend(ctx, false);
 
 		/* setup dma */
-		cfg0 = fimc_read(EXYNOS_MSCTRL);
+		cfg0 = fimc_read(ctx, EXYNOS_MSCTRL);
 		cfg0 &= ~EXYNOS_MSCTRL_INPUT_MASK;
 		cfg0 |= EXYNOS_MSCTRL_INPUT_MEMORY;
-		fimc_write(cfg0, EXYNOS_MSCTRL);
+		fimc_write(ctx, cfg0, EXYNOS_MSCTRL);
 		break;
 	case IPP_CMD_WB:
 		fimc_set_type_ctrl(ctx, FIMC_WB_A);
@@ -1610,41 +1567,33 @@ static int fimc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 	}
 
 	/* Reset status */
-	fimc_write(0x0, EXYNOS_CISTATUS);
+	fimc_write(ctx, 0x0, EXYNOS_CISTATUS);
 
-	cfg0 = fimc_read(EXYNOS_CIIMGCPT);
+	cfg0 = fimc_read(ctx, EXYNOS_CIIMGCPT);
 	cfg0 &= ~EXYNOS_CIIMGCPT_IMGCPTEN_SC;
 	cfg0 |= EXYNOS_CIIMGCPT_IMGCPTEN_SC;
 
 	/* Scaler */
-	cfg1 = fimc_read(EXYNOS_CISCCTRL);
+	cfg1 = fimc_read(ctx, EXYNOS_CISCCTRL);
 	cfg1 &= ~EXYNOS_CISCCTRL_SCAN_MASK;
 	cfg1 |= (EXYNOS_CISCCTRL_PROGRESSIVE |
 		EXYNOS_CISCCTRL_SCALERSTART);
 
-	fimc_write(cfg1, EXYNOS_CISCCTRL);
+	fimc_write(ctx, cfg1, EXYNOS_CISCCTRL);
 
 	/* Enable image capture*/
 	cfg0 |= EXYNOS_CIIMGCPT_IMGCPTEN;
-	fimc_write(cfg0, EXYNOS_CIIMGCPT);
+	fimc_write(ctx, cfg0, EXYNOS_CIIMGCPT);
 
 	/* Disable frame end irq */
-	cfg0 = fimc_read(EXYNOS_CIGCTRL);
-	cfg0 &= ~EXYNOS_CIGCTRL_IRQ_END_DISABLE;
-	fimc_write(cfg0, EXYNOS_CIGCTRL);
+	fimc_clear_bits(ctx, EXYNOS_CIGCTRL, EXYNOS_CIGCTRL_IRQ_END_DISABLE);
 
-	cfg0 = fimc_read(EXYNOS_CIOCTRL);
-	cfg0 &= ~EXYNOS_CIOCTRL_WEAVE_MASK;
-	fimc_write(cfg0, EXYNOS_CIOCTRL);
+	fimc_clear_bits(ctx, EXYNOS_CIOCTRL, EXYNOS_CIOCTRL_WEAVE_MASK);
 
 	if (cmd == IPP_CMD_M2M) {
-		cfg0 = fimc_read(EXYNOS_MSCTRL);
-		cfg0 |= EXYNOS_MSCTRL_ENVID;
-		fimc_write(cfg0, EXYNOS_MSCTRL);
+		fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
 
-		cfg0 = fimc_read(EXYNOS_MSCTRL);
-		cfg0 |= EXYNOS_MSCTRL_ENVID;
-		fimc_write(cfg0, EXYNOS_MSCTRL);
+		fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
 	}
 
 	return 0;
@@ -1661,10 +1610,10 @@ static void fimc_ippdrv_stop(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 	switch (cmd) {
 	case IPP_CMD_M2M:
 		/* Source clear */
-		cfg = fimc_read(EXYNOS_MSCTRL);
+		cfg = fimc_read(ctx, EXYNOS_MSCTRL);
 		cfg &= ~EXYNOS_MSCTRL_INPUT_MASK;
 		cfg &= ~EXYNOS_MSCTRL_ENVID;
-		fimc_write(cfg, EXYNOS_MSCTRL);
+		fimc_write(ctx, cfg, EXYNOS_MSCTRL);
 		break;
 	case IPP_CMD_WB:
 		exynos_drm_ippnb_send_event(IPP_SET_WRITEBACK, (void *)&set_wb);
@@ -1675,25 +1624,20 @@ static void fimc_ippdrv_stop(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 		break;
 	}
 
-	fimc_handle_irq(ctx, false, false, true);
+	fimc_mask_irq(ctx, false);
 
 	/* reset sequence */
-	fimc_write(0x0, EXYNOS_CIFCNTSEQ);
+	fimc_write(ctx, 0x0, EXYNOS_CIFCNTSEQ);
 
 	/* Scaler disable */
-	cfg = fimc_read(EXYNOS_CISCCTRL);
-	cfg &= ~EXYNOS_CISCCTRL_SCALERSTART;
-	fimc_write(cfg, EXYNOS_CISCCTRL);
+	fimc_clear_bits(ctx, EXYNOS_CISCCTRL, EXYNOS_CISCCTRL_SCALERSTART);
 
 	/* Disable image capture */
-	cfg = fimc_read(EXYNOS_CIIMGCPT);
-	cfg &= ~(EXYNOS_CIIMGCPT_IMGCPTEN_SC | EXYNOS_CIIMGCPT_IMGCPTEN);
-	fimc_write(cfg, EXYNOS_CIIMGCPT);
+	fimc_clear_bits(ctx, EXYNOS_CIIMGCPT,
+		EXYNOS_CIIMGCPT_IMGCPTEN_SC | EXYNOS_CIIMGCPT_IMGCPTEN);
 
 	/* Enable frame end irq */
-	cfg = fimc_read(EXYNOS_CIGCTRL);
-	cfg |= EXYNOS_CIGCTRL_IRQ_END_DISABLE;
-	fimc_write(cfg, EXYNOS_CIGCTRL);
+	fimc_set_bits(ctx, EXYNOS_CIGCTRL, EXYNOS_CIGCTRL_IRQ_END_DISABLE);
 }
 
 static void fimc_put_clocks(struct fimc_context *ctx)
@@ -1848,7 +1792,7 @@ static int fimc_probe(struct platform_device *pdev)
 
 	DRM_DEBUG_KMS("id[%d]ippdrv[0x%x]\n", ctx->id, (int)ippdrv);
 
-	mutex_init(&ctx->lock);
+	spin_lock_init(&ctx->lock);
 	platform_set_drvdata(pdev, ctx);
 
 	pm_runtime_set_active(dev);
@@ -1879,7 +1823,6 @@ static int fimc_remove(struct platform_device *pdev)
 	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 
 	exynos_drm_ippdrv_unregister(ippdrv);
-	mutex_destroy(&ctx->lock);
 
 	fimc_put_clocks(ctx);
 	pm_runtime_set_suspended(dev);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
index a61878bf5dc..33161ad3820 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
@@ -19,6 +19,7 @@
 #include <linux/of.h>
 #include <linux/of_device.h>
 #include <linux/pm_runtime.h>
+#include <linux/component.h>
 
 #include <video/of_display_timing.h>
 #include <video/of_videomode.h>
@@ -38,6 +39,7 @@
  */
 
 #define FIMD_DEFAULT_FRAMERATE 60
+#define MIN_FB_WIDTH_FOR_16WORD_BURST 128
 
 /* position control register for hardware window 0, 2 ~ 4.*/
 #define VIDOSD_A(win)		(VIDOSD_BASE + 0x00 + (win) * 16)
@@ -62,7 +64,7 @@
 /* FIMD has totally five hardware windows. */
 #define WINDOWS_NR	5
 
-#define get_fimd_context(dev)	platform_get_drvdata(to_platform_device(dev))
+#define get_fimd_manager(mgr)	platform_get_drvdata(to_platform_device(dev))
 
 struct fimd_driver_data {
 	unsigned int timing_base;
@@ -105,25 +107,24 @@ struct fimd_win_data {
 };
 
 struct fimd_context {
-	struct exynos_drm_subdrv	subdrv;
-	int				irq;
-	struct drm_crtc			*crtc;
+	struct device			*dev;
+	struct drm_device		*drm_dev;
 	struct clk			*bus_clk;
 	struct clk			*lcd_clk;
 	void __iomem			*regs;
+	struct drm_display_mode		mode;
 	struct fimd_win_data		win_data[WINDOWS_NR];
-	unsigned int			clkdiv;
 	unsigned int			default_win;
 	unsigned long			irq_flags;
-	u32				vidcon0;
 	u32				vidcon1;
 	bool				suspended;
-	struct mutex			lock;
+	int				pipe;
 	wait_queue_head_t		wait_vsync_queue;
 	atomic_t			wait_vsync_event;
 
 	struct exynos_drm_panel_info panel;
 	struct fimd_driver_data *driver_data;
+	struct exynos_drm_display *display;
 };
 
 static const struct of_device_id fimd_driver_dt_match[] = {
@@ -145,153 +146,197 @@ static inline struct fimd_driver_data *drm_fimd_get_driver_data(
 	return (struct fimd_driver_data *)of_id->data;
 }
 
-static bool fimd_display_is_connected(struct device *dev)
+static void fimd_wait_for_vblank(struct exynos_drm_manager *mgr)
 {
-	/* TODO. */
+	struct fimd_context *ctx = mgr->ctx;
 
-	return true;
+	if (ctx->suspended)
+		return;
+
+	atomic_set(&ctx->wait_vsync_event, 1);
+
+	/*
+	 * wait for FIMD to signal VSYNC interrupt or return after
+	 * timeout which is set to 50ms (refresh rate of 20).
+	 */
+	if (!wait_event_timeout(ctx->wait_vsync_queue,
+				!atomic_read(&ctx->wait_vsync_event),
+				HZ/20))
+		DRM_DEBUG_KMS("vblank wait timed out.\n");
 }
 
-static void *fimd_get_panel(struct device *dev)
+
+static void fimd_clear_channel(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
+	struct fimd_context *ctx = mgr->ctx;
+	int win, ch_enabled = 0;
+
+	DRM_DEBUG_KMS("%s\n", __FILE__);
+
+	/* Check if any channel is enabled. */
+	for (win = 0; win < WINDOWS_NR; win++) {
+		u32 val = readl(ctx->regs + SHADOWCON);
+		if (val & SHADOWCON_CHx_ENABLE(win)) {
+			val &= ~SHADOWCON_CHx_ENABLE(win);
+			writel(val, ctx->regs + SHADOWCON);
+			ch_enabled = 1;
+		}
+	}
 
-	return &ctx->panel;
+	/* Wait for vsync, as disable channel takes effect at next vsync */
+	if (ch_enabled)
+		fimd_wait_for_vblank(mgr);
 }
 
-static int fimd_check_mode(struct device *dev, struct drm_display_mode *mode)
+static int fimd_mgr_initialize(struct exynos_drm_manager *mgr,
+			struct drm_device *drm_dev)
 {
-	/* TODO. */
+	struct fimd_context *ctx = mgr->ctx;
+	struct exynos_drm_private *priv;
+	priv = drm_dev->dev_private;
+
+	mgr->drm_dev = ctx->drm_dev = drm_dev;
+	mgr->pipe = ctx->pipe = priv->pipe++;
+
+	/*
+	 * enable drm irq mode.
+	 * - with irq_enabled = true, we can use the vblank feature.
+	 *
+	 * P.S. note that we wouldn't use drm irq handler but
+	 *	just specific driver own one instead because
+	 *	drm framework supports only one irq handler.
+	 */
+	drm_dev->irq_enabled = true;
+
+	/*
+	 * with vblank_disable_allowed = true, vblank interrupt will be disabled
+	 * by drm timer once a current process gives up ownership of
+	 * vblank event.(after drm_vblank_put function is called)
+	 */
+	drm_dev->vblank_disable_allowed = true;
+
+	/* attach this sub driver to iommu mapping if supported. */
+	if (is_drm_iommu_supported(ctx->drm_dev)) {
+		/*
+		 * If any channel is already active, iommu will throw
+		 * a PAGE FAULT when enabled. So clear any channel if enabled.
+		 */
+		fimd_clear_channel(mgr);
+		drm_iommu_attach_device(ctx->drm_dev, ctx->dev);
+	}
 
 	return 0;
 }
 
-static int fimd_display_power_on(struct device *dev, int mode)
+static void fimd_mgr_remove(struct exynos_drm_manager *mgr)
 {
-	/* TODO */
+	struct fimd_context *ctx = mgr->ctx;
 
-	return 0;
+	/* detach this sub driver from iommu mapping if supported. */
+	if (is_drm_iommu_supported(ctx->drm_dev))
+		drm_iommu_detach_device(ctx->drm_dev, ctx->dev);
 }
 
-static struct exynos_drm_display_ops fimd_display_ops = {
-	.type = EXYNOS_DISPLAY_TYPE_LCD,
-	.is_connected = fimd_display_is_connected,
-	.get_panel = fimd_get_panel,
-	.check_mode = fimd_check_mode,
-	.power_on = fimd_display_power_on,
-};
-
-static void fimd_dpms(struct device *subdrv_dev, int mode)
+static u32 fimd_calc_clkdiv(struct fimd_context *ctx,
+		const struct drm_display_mode *mode)
 {
-	struct fimd_context *ctx = get_fimd_context(subdrv_dev);
+	unsigned long ideal_clk = mode->htotal * mode->vtotal * mode->vrefresh;
+	u32 clkdiv;
 
-	DRM_DEBUG_KMS("%d\n", mode);
+	/* Find the clock divider value that gets us closest to ideal_clk */
+	clkdiv = DIV_ROUND_UP(clk_get_rate(ctx->lcd_clk), ideal_clk);
 
-	mutex_lock(&ctx->lock);
+	return (clkdiv < 0x100) ? clkdiv : 0xff;
+}
 
-	switch (mode) {
-	case DRM_MODE_DPMS_ON:
-		/*
-		 * enable fimd hardware only if suspended status.
-		 *
-		 * P.S. fimd_dpms function would be called at booting time so
-		 * clk_enable could be called double time.
-		 */
-		if (ctx->suspended)
-			pm_runtime_get_sync(subdrv_dev);
-		break;
-	case DRM_MODE_DPMS_STANDBY:
-	case DRM_MODE_DPMS_SUSPEND:
-	case DRM_MODE_DPMS_OFF:
-		if (!ctx->suspended)
-			pm_runtime_put_sync(subdrv_dev);
-		break;
-	default:
-		DRM_DEBUG_KMS("unspecified mode %d\n", mode);
-		break;
-	}
+static bool fimd_mode_fixup(struct exynos_drm_manager *mgr,
+		const struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	if (adjusted_mode->vrefresh == 0)
+		adjusted_mode->vrefresh = FIMD_DEFAULT_FRAMERATE;
 
-	mutex_unlock(&ctx->lock);
+	return true;
 }
 
-static void fimd_apply(struct device *subdrv_dev)
+static void fimd_mode_set(struct exynos_drm_manager *mgr,
+		const struct drm_display_mode *in_mode)
 {
-	struct fimd_context *ctx = get_fimd_context(subdrv_dev);
-	struct exynos_drm_manager *mgr = ctx->subdrv.manager;
-	struct exynos_drm_manager_ops *mgr_ops = mgr->ops;
-	struct exynos_drm_overlay_ops *ovl_ops = mgr->overlay_ops;
-	struct fimd_win_data *win_data;
-	int i;
+	struct fimd_context *ctx = mgr->ctx;
 
-	for (i = 0; i < WINDOWS_NR; i++) {
-		win_data = &ctx->win_data[i];
-		if (win_data->enabled && (ovl_ops && ovl_ops->commit))
-			ovl_ops->commit(subdrv_dev, i);
-	}
-
-	if (mgr_ops && mgr_ops->commit)
-		mgr_ops->commit(subdrv_dev);
+	drm_mode_copy(&ctx->mode, in_mode);
 }
 
-static void fimd_commit(struct device *dev)
+static void fimd_commit(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
-	struct exynos_drm_panel_info *panel = &ctx->panel;
-	struct videomode *vm = &panel->vm;
+	struct fimd_context *ctx = mgr->ctx;
+	struct drm_display_mode *mode = &ctx->mode;
 	struct fimd_driver_data *driver_data;
-	u32 val;
+	u32 val, clkdiv, vidcon1;
+	int vsync_len, vbpd, vfpd, hsync_len, hbpd, hfpd;
 
 	driver_data = ctx->driver_data;
 	if (ctx->suspended)
 		return;
 
-	/* setup polarity values from machine code. */
-	writel(ctx->vidcon1, ctx->regs + driver_data->timing_base + VIDCON1);
+	/* nothing to do if we haven't set the mode yet */
+	if (mode->htotal == 0 || mode->vtotal == 0)
+		return;
+
+	/* setup polarity values */
+	vidcon1 = ctx->vidcon1;
+	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+		vidcon1 |= VIDCON1_INV_VSYNC;
+	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+		vidcon1 |= VIDCON1_INV_HSYNC;
+	writel(vidcon1, ctx->regs + driver_data->timing_base + VIDCON1);
 
 	/* setup vertical timing values. */
-	val = VIDTCON0_VBPD(vm->vback_porch - 1) |
-	       VIDTCON0_VFPD(vm->vfront_porch - 1) |
-	       VIDTCON0_VSPW(vm->vsync_len - 1);
+	vsync_len = mode->crtc_vsync_end - mode->crtc_vsync_start;
+	vbpd = mode->crtc_vtotal - mode->crtc_vsync_end;
+	vfpd = mode->crtc_vsync_start - mode->crtc_vdisplay;
+
+	val = VIDTCON0_VBPD(vbpd - 1) |
+		VIDTCON0_VFPD(vfpd - 1) |
+		VIDTCON0_VSPW(vsync_len - 1);
 	writel(val, ctx->regs + driver_data->timing_base + VIDTCON0);
 
 	/* setup horizontal timing values.  */
-	val = VIDTCON1_HBPD(vm->hback_porch - 1) |
-	       VIDTCON1_HFPD(vm->hfront_porch - 1) |
-	       VIDTCON1_HSPW(vm->hsync_len - 1);
+	hsync_len = mode->crtc_hsync_end - mode->crtc_hsync_start;
+	hbpd = mode->crtc_htotal - mode->crtc_hsync_end;
+	hfpd = mode->crtc_hsync_start - mode->crtc_hdisplay;
+
+	val = VIDTCON1_HBPD(hbpd - 1) |
+		VIDTCON1_HFPD(hfpd - 1) |
+		VIDTCON1_HSPW(hsync_len - 1);
 	writel(val, ctx->regs + driver_data->timing_base + VIDTCON1);
 
 	/* setup horizontal and vertical display size. */
-	val = VIDTCON2_LINEVAL(vm->vactive - 1) |
-	       VIDTCON2_HOZVAL(vm->hactive - 1) |
-	       VIDTCON2_LINEVAL_E(vm->vactive - 1) |
-	       VIDTCON2_HOZVAL_E(vm->hactive - 1);
+	val = VIDTCON2_LINEVAL(mode->vdisplay - 1) |
+	       VIDTCON2_HOZVAL(mode->hdisplay - 1) |
+	       VIDTCON2_LINEVAL_E(mode->vdisplay - 1) |
+	       VIDTCON2_HOZVAL_E(mode->hdisplay - 1);
 	writel(val, ctx->regs + driver_data->timing_base + VIDTCON2);
 
-	/* setup clock source, clock divider, enable dma. */
-	val = ctx->vidcon0;
-	val &= ~(VIDCON0_CLKVAL_F_MASK | VIDCON0_CLKDIR);
-
-	if (ctx->driver_data->has_clksel) {
-		val &= ~VIDCON0_CLKSEL_MASK;
-		val |= VIDCON0_CLKSEL_LCD;
-	}
-
-	if (ctx->clkdiv > 1)
-		val |= VIDCON0_CLKVAL_F(ctx->clkdiv - 1) | VIDCON0_CLKDIR;
-	else
-		val &= ~VIDCON0_CLKDIR;	/* 1:1 clock */
-
 	/*
 	 * fields of register with prefix '_F' would be updated
 	 * at vsync(same as dma start)
 	 */
-	val |= VIDCON0_ENVID | VIDCON0_ENVID_F;
+	val = VIDCON0_ENVID | VIDCON0_ENVID_F;
+
+	if (ctx->driver_data->has_clksel)
+		val |= VIDCON0_CLKSEL_LCD;
+
+	clkdiv = fimd_calc_clkdiv(ctx, mode);
+	if (clkdiv > 1)
+		val |= VIDCON0_CLKVAL_F(clkdiv - 1) | VIDCON0_CLKDIR;
+
 	writel(val, ctx->regs + VIDCON0);
 }
 
-static int fimd_enable_vblank(struct device *dev)
+static int fimd_enable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
+	struct fimd_context *ctx = mgr->ctx;
 	u32 val;
 
 	if (ctx->suspended)
@@ -314,9 +359,9 @@ static int fimd_enable_vblank(struct device *dev)
 	return 0;
 }
 
-static void fimd_disable_vblank(struct device *dev)
+static void fimd_disable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
+	struct fimd_context *ctx = mgr->ctx;
 	u32 val;
 
 	if (ctx->suspended)
@@ -332,44 +377,16 @@ static void fimd_disable_vblank(struct device *dev)
 	}
 }
 
-static void fimd_wait_for_vblank(struct device *dev)
-{
-	struct fimd_context *ctx = get_fimd_context(dev);
-
-	if (ctx->suspended)
-		return;
-
-	atomic_set(&ctx->wait_vsync_event, 1);
-
-	/*
-	 * wait for FIMD to signal VSYNC interrupt or return after
-	 * timeout which is set to 50ms (refresh rate of 20).
-	 */
-	if (!wait_event_timeout(ctx->wait_vsync_queue,
-				!atomic_read(&ctx->wait_vsync_event),
-				DRM_HZ/20))
-		DRM_DEBUG_KMS("vblank wait timed out.\n");
-}
-
-static struct exynos_drm_manager_ops fimd_manager_ops = {
-	.dpms = fimd_dpms,
-	.apply = fimd_apply,
-	.commit = fimd_commit,
-	.enable_vblank = fimd_enable_vblank,
-	.disable_vblank = fimd_disable_vblank,
-	.wait_for_vblank = fimd_wait_for_vblank,
-};
-
-static void fimd_win_mode_set(struct device *dev,
-			      struct exynos_drm_overlay *overlay)
+static void fimd_win_mode_set(struct exynos_drm_manager *mgr,
+			struct exynos_drm_overlay *overlay)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
+	struct fimd_context *ctx = mgr->ctx;
 	struct fimd_win_data *win_data;
 	int win;
 	unsigned long offset;
 
 	if (!overlay) {
-		dev_err(dev, "overlay is NULL\n");
+		DRM_ERROR("overlay is NULL\n");
 		return;
 	}
 
@@ -409,9 +426,8 @@ static void fimd_win_mode_set(struct device *dev,
 			overlay->fb_width, overlay->crtc_width);
 }
 
-static void fimd_win_set_pixfmt(struct device *dev, unsigned int win)
+static void fimd_win_set_pixfmt(struct fimd_context *ctx, unsigned int win)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
 	struct fimd_win_data *win_data = &ctx->win_data[win];
 	unsigned long val;
 
@@ -464,12 +480,24 @@ static void fimd_win_set_pixfmt(struct device *dev, unsigned int win)
 
 	DRM_DEBUG_KMS("bpp = %d\n", win_data->bpp);
 
+	/*
+	 * In case of exynos, setting dma-burst to 16Word causes permanent
+	 * tearing for very small buffers, e.g. cursor buffer. Burst Mode
+	 * switching which is based on overlay size is not recommended as
+	 * overlay size varies alot towards the end of the screen and rapid
+	 * movement causes unstable DMA which results into iommu crash/tear.
+	 */
+
+	if (win_data->fb_width < MIN_FB_WIDTH_FOR_16WORD_BURST) {
+		val &= ~WINCONx_BURSTLEN_MASK;
+		val |= WINCONx_BURSTLEN_4WORD;
+	}
+
 	writel(val, ctx->regs + WINCON(win));
 }
 
-static void fimd_win_set_colkey(struct device *dev, unsigned int win)
+static void fimd_win_set_colkey(struct fimd_context *ctx, unsigned int win)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
 	unsigned int keycon0 = 0, keycon1 = 0;
 
 	keycon0 = ~(WxKEYCON0_KEYBL_EN | WxKEYCON0_KEYEN_F |
@@ -508,9 +536,9 @@ static void fimd_shadow_protect_win(struct fimd_context *ctx,
 	writel(val, ctx->regs + reg);
 }
 
-static void fimd_win_commit(struct device *dev, int zpos)
+static void fimd_win_commit(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
+	struct fimd_context *ctx = mgr->ctx;
 	struct fimd_win_data *win_data;
 	int win = zpos;
 	unsigned long val, alpha, size;
@@ -528,6 +556,12 @@ static void fimd_win_commit(struct device *dev, int zpos)
 
 	win_data = &ctx->win_data[win];
 
+	/* If suspended, enable this on resume */
+	if (ctx->suspended) {
+		win_data->resume = true;
+		return;
+	}
+
 	/*
 	 * SHADOWCON/PRTCON register is used for enabling timing.
 	 *
@@ -605,11 +639,11 @@ static void fimd_win_commit(struct device *dev, int zpos)
 		DRM_DEBUG_KMS("osd size = 0x%x\n", (unsigned int)val);
 	}
 
-	fimd_win_set_pixfmt(dev, win);
+	fimd_win_set_pixfmt(ctx, win);
 
 	/* hardware window 0 doesn't support color key. */
 	if (win != 0)
-		fimd_win_set_colkey(dev, win);
+		fimd_win_set_colkey(ctx, win);
 
 	/* wincon */
 	val = readl(ctx->regs + WINCON(win));
@@ -628,9 +662,9 @@ static void fimd_win_commit(struct device *dev, int zpos)
 	win_data->enabled = true;
 }
 
-static void fimd_win_disable(struct device *dev, int zpos)
+static void fimd_win_disable(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
+	struct fimd_context *ctx = mgr->ctx;
 	struct fimd_win_data *win_data;
 	int win = zpos;
 	u32 val;
@@ -669,408 +703,333 @@ static void fimd_win_disable(struct device *dev, int zpos)
 	win_data->enabled = false;
 }
 
-static struct exynos_drm_overlay_ops fimd_overlay_ops = {
-	.mode_set = fimd_win_mode_set,
-	.commit = fimd_win_commit,
-	.disable = fimd_win_disable,
-};
-
-static struct exynos_drm_manager fimd_manager = {
-	.pipe		= -1,
-	.ops		= &fimd_manager_ops,
-	.overlay_ops	= &fimd_overlay_ops,
-	.display_ops	= &fimd_display_ops,
-};
-
-static irqreturn_t fimd_irq_handler(int irq, void *dev_id)
+static void fimd_window_suspend(struct exynos_drm_manager *mgr)
 {
-	struct fimd_context *ctx = (struct fimd_context *)dev_id;
-	struct exynos_drm_subdrv *subdrv = &ctx->subdrv;
-	struct drm_device *drm_dev = subdrv->drm_dev;
-	struct exynos_drm_manager *manager = subdrv->manager;
-	u32 val;
-
-	val = readl(ctx->regs + VIDINTCON1);
-
-	if (val & VIDINTCON1_INT_FRAME)
-		/* VSYNC interrupt */
-		writel(VIDINTCON1_INT_FRAME, ctx->regs + VIDINTCON1);
+	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_win_data *win_data;
+	int i;
 
-	/* check the crtc is detached already from encoder */
-	if (manager->pipe < 0)
-		goto out;
+	for (i = 0; i < WINDOWS_NR; i++) {
+		win_data = &ctx->win_data[i];
+		win_data->resume = win_data->enabled;
+		if (win_data->enabled)
+			fimd_win_disable(mgr, i);
+	}
+	fimd_wait_for_vblank(mgr);
+}
 
-	drm_handle_vblank(drm_dev, manager->pipe);
-	exynos_drm_crtc_finish_pageflip(drm_dev, manager->pipe);
+static void fimd_window_resume(struct exynos_drm_manager *mgr)
+{
+	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_win_data *win_data;
+	int i;
 
-	/* set wait vsync event to zero and wake up queue. */
-	if (atomic_read(&ctx->wait_vsync_event)) {
-		atomic_set(&ctx->wait_vsync_event, 0);
-		DRM_WAKEUP(&ctx->wait_vsync_queue);
+	for (i = 0; i < WINDOWS_NR; i++) {
+		win_data = &ctx->win_data[i];
+		win_data->enabled = win_data->resume;
+		win_data->resume = false;
 	}
-out:
-	return IRQ_HANDLED;
 }
 
-static int fimd_subdrv_probe(struct drm_device *drm_dev, struct device *dev)
+static void fimd_apply(struct exynos_drm_manager *mgr)
 {
-	/*
-	 * enable drm irq mode.
-	 * - with irq_enabled = true, we can use the vblank feature.
-	 *
-	 * P.S. note that we wouldn't use drm irq handler but
-	 *	just specific driver own one instead because
-	 *	drm framework supports only one irq handler.
-	 */
-	drm_dev->irq_enabled = true;
-
-	/*
-	 * with vblank_disable_allowed = true, vblank interrupt will be disabled
-	 * by drm timer once a current process gives up ownership of
-	 * vblank event.(after drm_vblank_put function is called)
-	 */
-	drm_dev->vblank_disable_allowed = true;
+	struct fimd_context *ctx = mgr->ctx;
+	struct fimd_win_data *win_data;
+	int i;
 
-	/* attach this sub driver to iommu mapping if supported. */
-	if (is_drm_iommu_supported(drm_dev))
-		drm_iommu_attach_device(drm_dev, dev);
+	for (i = 0; i < WINDOWS_NR; i++) {
+		win_data = &ctx->win_data[i];
+		if (win_data->enabled)
+			fimd_win_commit(mgr, i);
+		else
+			fimd_win_disable(mgr, i);
+	}
 
-	return 0;
+	fimd_commit(mgr);
 }
 
-static void fimd_subdrv_remove(struct drm_device *drm_dev, struct device *dev)
+static int fimd_poweron(struct exynos_drm_manager *mgr)
 {
-	/* detach this sub driver from iommu mapping if supported. */
-	if (is_drm_iommu_supported(drm_dev))
-		drm_iommu_detach_device(drm_dev, dev);
-}
+	struct fimd_context *ctx = mgr->ctx;
+	int ret;
 
-static int fimd_configure_clocks(struct fimd_context *ctx, struct device *dev)
-{
-	struct videomode *vm = &ctx->panel.vm;
-	unsigned long clk;
+	if (!ctx->suspended)
+		return 0;
 
-	ctx->bus_clk = devm_clk_get(dev, "fimd");
-	if (IS_ERR(ctx->bus_clk)) {
-		dev_err(dev, "failed to get bus clock\n");
-		return PTR_ERR(ctx->bus_clk);
-	}
+	ctx->suspended = false;
 
-	ctx->lcd_clk = devm_clk_get(dev, "sclk_fimd");
-	if (IS_ERR(ctx->lcd_clk)) {
-		dev_err(dev, "failed to get lcd clock\n");
-		return PTR_ERR(ctx->lcd_clk);
+	pm_runtime_get_sync(ctx->dev);
+
+	ret = clk_prepare_enable(ctx->bus_clk);
+	if (ret < 0) {
+		DRM_ERROR("Failed to prepare_enable the bus clk [%d]\n", ret);
+		goto bus_clk_err;
 	}
 
-	clk = clk_get_rate(ctx->lcd_clk);
-	if (clk == 0) {
-		dev_err(dev, "error getting sclk_fimd clock rate\n");
-		return -EINVAL;
+	ret = clk_prepare_enable(ctx->lcd_clk);
+	if  (ret < 0) {
+		DRM_ERROR("Failed to prepare_enable the lcd clk [%d]\n", ret);
+		goto lcd_clk_err;
 	}
 
-	if (vm->pixelclock == 0) {
-		unsigned long c;
-		c = vm->hactive + vm->hback_porch + vm->hfront_porch +
-		    vm->hsync_len;
-		c *= vm->vactive + vm->vback_porch + vm->vfront_porch +
-		     vm->vsync_len;
-		vm->pixelclock = c * FIMD_DEFAULT_FRAMERATE;
-		if (vm->pixelclock == 0) {
-			dev_err(dev, "incorrect display timings\n");
-			return -EINVAL;
+	/* if vblank was enabled status, enable it again. */
+	if (test_and_clear_bit(0, &ctx->irq_flags)) {
+		ret = fimd_enable_vblank(mgr);
+		if (ret) {
+			DRM_ERROR("Failed to re-enable vblank [%d]\n", ret);
+			goto enable_vblank_err;
 		}
-		dev_warn(dev, "pixel clock recalculated to %luHz (%dHz frame rate)\n",
-			 vm->pixelclock, FIMD_DEFAULT_FRAMERATE);
 	}
-	ctx->clkdiv = DIV_ROUND_UP(clk, vm->pixelclock);
-	if (ctx->clkdiv > 256) {
-		dev_warn(dev, "calculated pixel clock divider too high (%u), lowered to 256\n",
-			 ctx->clkdiv);
-		ctx->clkdiv = 256;
-	}
-	vm->pixelclock = clk / ctx->clkdiv;
-	DRM_DEBUG_KMS("pixel clock = %lu, clkdiv = %d\n", vm->pixelclock,
-		      ctx->clkdiv);
 
-	return 0;
-}
+	fimd_window_resume(mgr);
 
-static void fimd_clear_win(struct fimd_context *ctx, int win)
-{
-	writel(0, ctx->regs + WINCON(win));
-	writel(0, ctx->regs + VIDOSD_A(win));
-	writel(0, ctx->regs + VIDOSD_B(win));
-	writel(0, ctx->regs + VIDOSD_C(win));
+	fimd_apply(mgr);
 
-	if (win == 1 || win == 2)
-		writel(0, ctx->regs + VIDOSD_D(win));
+	return 0;
 
-	fimd_shadow_protect_win(ctx, win, false);
+enable_vblank_err:
+	clk_disable_unprepare(ctx->lcd_clk);
+lcd_clk_err:
+	clk_disable_unprepare(ctx->bus_clk);
+bus_clk_err:
+	ctx->suspended = true;
+	return ret;
 }
 
-static int fimd_clock(struct fimd_context *ctx, bool enable)
+static int fimd_poweroff(struct exynos_drm_manager *mgr)
 {
-	if (enable) {
-		int ret;
+	struct fimd_context *ctx = mgr->ctx;
 
-		ret = clk_prepare_enable(ctx->bus_clk);
-		if (ret < 0)
-			return ret;
+	if (ctx->suspended)
+		return 0;
 
-		ret = clk_prepare_enable(ctx->lcd_clk);
-		if  (ret < 0) {
-			clk_disable_unprepare(ctx->bus_clk);
-			return ret;
-		}
-	} else {
-		clk_disable_unprepare(ctx->lcd_clk);
-		clk_disable_unprepare(ctx->bus_clk);
-	}
+	/*
+	 * We need to make sure that all windows are disabled before we
+	 * suspend that connector. Otherwise we might try to scan from
+	 * a destroyed buffer later.
+	 */
+	fimd_window_suspend(mgr);
+
+	clk_disable_unprepare(ctx->lcd_clk);
+	clk_disable_unprepare(ctx->bus_clk);
 
+	pm_runtime_put_sync(ctx->dev);
+
+	ctx->suspended = true;
 	return 0;
 }
 
-static void fimd_window_suspend(struct device *dev)
+static void fimd_dpms(struct exynos_drm_manager *mgr, int mode)
 {
-	struct fimd_context *ctx = get_fimd_context(dev);
-	struct fimd_win_data *win_data;
-	int i;
+	DRM_DEBUG_KMS("%s, %d\n", __FILE__, mode);
 
-	for (i = 0; i < WINDOWS_NR; i++) {
-		win_data = &ctx->win_data[i];
-		win_data->resume = win_data->enabled;
-		fimd_win_disable(dev, i);
+	switch (mode) {
+	case DRM_MODE_DPMS_ON:
+		fimd_poweron(mgr);
+		break;
+	case DRM_MODE_DPMS_STANDBY:
+	case DRM_MODE_DPMS_SUSPEND:
+	case DRM_MODE_DPMS_OFF:
+		fimd_poweroff(mgr);
+		break;
+	default:
+		DRM_DEBUG_KMS("unspecified mode %d\n", mode);
+		break;
 	}
-	fimd_wait_for_vblank(dev);
 }
 
-static void fimd_window_resume(struct device *dev)
-{
-	struct fimd_context *ctx = get_fimd_context(dev);
-	struct fimd_win_data *win_data;
-	int i;
+static struct exynos_drm_manager_ops fimd_manager_ops = {
+	.dpms = fimd_dpms,
+	.mode_fixup = fimd_mode_fixup,
+	.mode_set = fimd_mode_set,
+	.commit = fimd_commit,
+	.enable_vblank = fimd_enable_vblank,
+	.disable_vblank = fimd_disable_vblank,
+	.wait_for_vblank = fimd_wait_for_vblank,
+	.win_mode_set = fimd_win_mode_set,
+	.win_commit = fimd_win_commit,
+	.win_disable = fimd_win_disable,
+};
 
-	for (i = 0; i < WINDOWS_NR; i++) {
-		win_data = &ctx->win_data[i];
-		win_data->enabled = win_data->resume;
-		win_data->resume = false;
-	}
-}
+static struct exynos_drm_manager fimd_manager = {
+	.type = EXYNOS_DISPLAY_TYPE_LCD,
+	.ops = &fimd_manager_ops,
+};
 
-static int fimd_activate(struct fimd_context *ctx, bool enable)
+static irqreturn_t fimd_irq_handler(int irq, void *dev_id)
 {
-	struct device *dev = ctx->subdrv.dev;
-	if (enable) {
-		int ret;
+	struct fimd_context *ctx = (struct fimd_context *)dev_id;
+	u32 val;
 
-		ret = fimd_clock(ctx, true);
-		if (ret < 0)
-			return ret;
+	val = readl(ctx->regs + VIDINTCON1);
 
-		ctx->suspended = false;
+	if (val & VIDINTCON1_INT_FRAME)
+		/* VSYNC interrupt */
+		writel(VIDINTCON1_INT_FRAME, ctx->regs + VIDINTCON1);
 
-		/* if vblank was enabled status, enable it again. */
-		if (test_and_clear_bit(0, &ctx->irq_flags))
-			fimd_enable_vblank(dev);
+	/* check the crtc is detached already from encoder */
+	if (ctx->pipe < 0 || !ctx->drm_dev)
+		goto out;
 
-		fimd_window_resume(dev);
-	} else {
-		fimd_window_suspend(dev);
+	drm_handle_vblank(ctx->drm_dev, ctx->pipe);
+	exynos_drm_crtc_finish_pageflip(ctx->drm_dev, ctx->pipe);
 
-		fimd_clock(ctx, false);
-		ctx->suspended = true;
+	/* set wait vsync event to zero and wake up queue. */
+	if (atomic_read(&ctx->wait_vsync_event)) {
+		atomic_set(&ctx->wait_vsync_event, 0);
+		wake_up(&ctx->wait_vsync_queue);
 	}
+out:
+	return IRQ_HANDLED;
+}
+
+static int fimd_bind(struct device *dev, struct device *master, void *data)
+{
+	struct fimd_context *ctx = fimd_manager.ctx;
+	struct drm_device *drm_dev = data;
+
+	fimd_mgr_initialize(&fimd_manager, drm_dev);
+	exynos_drm_crtc_create(&fimd_manager);
+	if (ctx->display)
+		exynos_drm_create_enc_conn(drm_dev, ctx->display);
 
 	return 0;
+
 }
 
-static int fimd_get_platform_data(struct fimd_context *ctx, struct device *dev)
+static void fimd_unbind(struct device *dev, struct device *master,
+			void *data)
 {
-	struct videomode *vm;
-	int ret;
+	struct exynos_drm_manager *mgr = dev_get_drvdata(dev);
+	struct fimd_context *ctx = fimd_manager.ctx;
+	struct drm_crtc *crtc = mgr->crtc;
 
-	vm = &ctx->panel.vm;
-	ret = of_get_videomode(dev->of_node, vm, OF_USE_NATIVE_MODE);
-	if (ret) {
-		DRM_ERROR("failed: of_get_videomode() : %d\n", ret);
-		return ret;
-	}
+	fimd_dpms(mgr, DRM_MODE_DPMS_OFF);
 
-	if (vm->flags & DISPLAY_FLAGS_VSYNC_LOW)
-		ctx->vidcon1 |= VIDCON1_INV_VSYNC;
-	if (vm->flags & DISPLAY_FLAGS_HSYNC_LOW)
-		ctx->vidcon1 |= VIDCON1_INV_HSYNC;
-	if (vm->flags & DISPLAY_FLAGS_DE_LOW)
-		ctx->vidcon1 |= VIDCON1_INV_VDEN;
-	if (vm->flags & DISPLAY_FLAGS_PIXDATA_NEGEDGE)
-		ctx->vidcon1 |= VIDCON1_INV_VCLK;
+	if (ctx->display)
+		exynos_dpi_remove(dev);
 
-	return 0;
+	fimd_mgr_remove(mgr);
+
+	crtc->funcs->destroy(crtc);
 }
 
+static const struct component_ops fimd_component_ops = {
+	.bind	= fimd_bind,
+	.unbind = fimd_unbind,
+};
+
 static int fimd_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	struct fimd_context *ctx;
-	struct exynos_drm_subdrv *subdrv;
 	struct resource *res;
-	int win;
 	int ret = -EINVAL;
 
-	if (!dev->of_node)
-		return -ENODEV;
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC,
+					fimd_manager.type);
+	if (ret)
+		return ret;
+
+	if (!dev->of_node) {
+		ret = -ENODEV;
+		goto err_del_component;
+	}
 
 	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
-	if (!ctx)
-		return -ENOMEM;
+	if (!ctx) {
+		ret = -ENOMEM;
+		goto err_del_component;
+	}
 
-	ret = fimd_get_platform_data(ctx, dev);
-	if (ret)
-		return ret;
+	ctx->dev = dev;
+	ctx->suspended = true;
 
-	ret = fimd_configure_clocks(ctx, dev);
-	if (ret)
-		return ret;
+	if (of_property_read_bool(dev->of_node, "samsung,invert-vden"))
+		ctx->vidcon1 |= VIDCON1_INV_VDEN;
+	if (of_property_read_bool(dev->of_node, "samsung,invert-vclk"))
+		ctx->vidcon1 |= VIDCON1_INV_VCLK;
+
+	ctx->bus_clk = devm_clk_get(dev, "fimd");
+	if (IS_ERR(ctx->bus_clk)) {
+		dev_err(dev, "failed to get bus clock\n");
+		ret = PTR_ERR(ctx->bus_clk);
+		goto err_del_component;
+	}
+
+	ctx->lcd_clk = devm_clk_get(dev, "sclk_fimd");
+	if (IS_ERR(ctx->lcd_clk)) {
+		dev_err(dev, "failed to get lcd clock\n");
+		ret = PTR_ERR(ctx->lcd_clk);
+		goto err_del_component;
+	}
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 
 	ctx->regs = devm_ioremap_resource(dev, res);
-	if (IS_ERR(ctx->regs))
-		return PTR_ERR(ctx->regs);
+	if (IS_ERR(ctx->regs)) {
+		ret = PTR_ERR(ctx->regs);
+		goto err_del_component;
+	}
 
 	res = platform_get_resource_byname(pdev, IORESOURCE_IRQ, "vsync");
 	if (!res) {
 		dev_err(dev, "irq request failed.\n");
-		return -ENXIO;
+		ret = -ENXIO;
+		goto err_del_component;
 	}
 
-	ctx->irq = res->start;
-
-	ret = devm_request_irq(dev, ctx->irq, fimd_irq_handler,
+	ret = devm_request_irq(dev, res->start, fimd_irq_handler,
 							0, "drm_fimd", ctx);
 	if (ret) {
 		dev_err(dev, "irq request failed.\n");
-		return ret;
+		goto err_del_component;
 	}
 
 	ctx->driver_data = drm_fimd_get_driver_data(pdev);
-	DRM_INIT_WAITQUEUE(&ctx->wait_vsync_queue);
+	init_waitqueue_head(&ctx->wait_vsync_queue);
 	atomic_set(&ctx->wait_vsync_event, 0);
 
-	subdrv = &ctx->subdrv;
+	platform_set_drvdata(pdev, &fimd_manager);
 
-	subdrv->dev = dev;
-	subdrv->manager = &fimd_manager;
-	subdrv->probe = fimd_subdrv_probe;
-	subdrv->remove = fimd_subdrv_remove;
+	fimd_manager.ctx = ctx;
 
-	mutex_init(&ctx->lock);
+	ctx->display = exynos_dpi_probe(dev);
+	if (IS_ERR(ctx->display))
+		return PTR_ERR(ctx->display);
 
-	platform_set_drvdata(pdev, ctx);
+	pm_runtime_enable(&pdev->dev);
 
-	pm_runtime_enable(dev);
-	pm_runtime_get_sync(dev);
+	ret = component_add(&pdev->dev, &fimd_component_ops);
+	if (ret)
+		goto err_disable_pm_runtime;
 
-	for (win = 0; win < WINDOWS_NR; win++)
-		fimd_clear_win(ctx, win);
+	return ret;
 
-	exynos_drm_subdrv_register(subdrv);
+err_disable_pm_runtime:
+	pm_runtime_disable(&pdev->dev);
 
-	return 0;
+err_del_component:
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
+	return ret;
 }
 
 static int fimd_remove(struct platform_device *pdev)
 {
-	struct device *dev = &pdev->dev;
-	struct fimd_context *ctx = platform_get_drvdata(pdev);
+	pm_runtime_disable(&pdev->dev);
 
-	exynos_drm_subdrv_unregister(&ctx->subdrv);
-
-	if (ctx->suspended)
-		goto out;
-
-	pm_runtime_set_suspended(dev);
-	pm_runtime_put_sync(dev);
-
-out:
-	pm_runtime_disable(dev);
+	component_del(&pdev->dev, &fimd_component_ops);
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
 
 	return 0;
 }
 
-#ifdef CONFIG_PM_SLEEP
-static int fimd_suspend(struct device *dev)
-{
-	struct fimd_context *ctx = get_fimd_context(dev);
-
-	/*
-	 * do not use pm_runtime_suspend(). if pm_runtime_suspend() is
-	 * called here, an error would be returned by that interface
-	 * because the usage_count of pm runtime is more than 1.
-	 */
-	if (!pm_runtime_suspended(dev))
-		return fimd_activate(ctx, false);
-
-	return 0;
-}
-
-static int fimd_resume(struct device *dev)
-{
-	struct fimd_context *ctx = get_fimd_context(dev);
-
-	/*
-	 * if entered to sleep when lcd panel was on, the usage_count
-	 * of pm runtime would still be 1 so in this case, fimd driver
-	 * should be on directly not drawing on pm runtime interface.
-	 */
-	if (!pm_runtime_suspended(dev)) {
-		int ret;
-
-		ret = fimd_activate(ctx, true);
-		if (ret < 0)
-			return ret;
-
-		/*
-		 * in case of dpms on(standby), fimd_apply function will
-		 * be called by encoder's dpms callback to update fimd's
-		 * registers but in case of sleep wakeup, it's not.
-		 * so fimd_apply function should be called at here.
-		 */
-		fimd_apply(dev);
-	}
-
-	return 0;
-}
-#endif
-
-#ifdef CONFIG_PM_RUNTIME
-static int fimd_runtime_suspend(struct device *dev)
-{
-	struct fimd_context *ctx = get_fimd_context(dev);
-
-	return fimd_activate(ctx, false);
-}
-
-static int fimd_runtime_resume(struct device *dev)
-{
-	struct fimd_context *ctx = get_fimd_context(dev);
-
-	return fimd_activate(ctx, true);
-}
-#endif
-
-static const struct dev_pm_ops fimd_pm_ops = {
-	SET_SYSTEM_SLEEP_PM_OPS(fimd_suspend, fimd_resume)
-	SET_RUNTIME_PM_OPS(fimd_runtime_suspend, fimd_runtime_resume, NULL)
-};
-
 struct platform_driver fimd_driver = {
 	.probe		= fimd_probe,
 	.remove		= fimd_remove,
 	.driver		= {
 		.name	= "exynos4-fb",
 		.owner	= THIS_MODULE,
-		.pm	= &fimd_pm_ops,
 		.of_match_table = fimd_driver_dt_match,
 	},
 };
diff --git a/drivers/gpu/drm/exynos/exynos_drm_g2d.c b/drivers/gpu/drm/exynos/exynos_drm_g2d.c
index 380aec28840..80015871447 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_g2d.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_g2d.c
@@ -467,14 +467,17 @@ static dma_addr_t *g2d_userptr_get_dma_addr(struct drm_device *drm_dev,
 		goto err_free;
 	}
 
+	down_read(&current->mm->mmap_sem);
 	vma = find_vma(current->mm, userptr);
 	if (!vma) {
+		up_read(&current->mm->mmap_sem);
 		DRM_ERROR("failed to get vm region.\n");
 		ret = -EFAULT;
 		goto err_free_pages;
 	}
 
 	if (vma->vm_end < userptr + size) {
+		up_read(&current->mm->mmap_sem);
 		DRM_ERROR("vma is too small.\n");
 		ret = -EFAULT;
 		goto err_free_pages;
@@ -482,6 +485,7 @@ static dma_addr_t *g2d_userptr_get_dma_addr(struct drm_device *drm_dev,
 
 	g2d_userptr->vma = exynos_gem_get_vma(vma);
 	if (!g2d_userptr->vma) {
+		up_read(&current->mm->mmap_sem);
 		DRM_ERROR("failed to copy vma.\n");
 		ret = -ENOMEM;
 		goto err_free_pages;
@@ -492,10 +496,12 @@ static dma_addr_t *g2d_userptr_get_dma_addr(struct drm_device *drm_dev,
 	ret = exynos_gem_get_pages_from_userptr(start & PAGE_MASK,
 						npages, pages, vma);
 	if (ret < 0) {
+		up_read(&current->mm->mmap_sem);
 		DRM_ERROR("failed to get user pages from userptr.\n");
 		goto err_put_vma;
 	}
 
+	up_read(&current->mm->mmap_sem);
 	g2d_userptr->pages = pages;
 
 	sgt = kzalloc(sizeof(*sgt), GFP_KERNEL);
@@ -607,7 +613,7 @@ static enum g2d_reg_type g2d_get_reg_type(int reg_offset)
 		reg_type = REG_TYPE_NONE;
 		DRM_ERROR("Unknown register offset![%d]\n", reg_offset);
 		break;
-	};
+	}
 
 	return reg_type;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c
index be59d50d8b1..163a054922c 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
@@ -338,46 +338,22 @@ int exynos_drm_gem_map_offset_ioctl(struct drm_device *dev, void *data,
 			&args->offset);
 }
 
-static struct drm_file *exynos_drm_find_drm_file(struct drm_device *drm_dev,
-							struct file *filp)
-{
-	struct drm_file *file_priv;
-
-	/* find current process's drm_file from filelist. */
-	list_for_each_entry(file_priv, &drm_dev->filelist, lhead)
-		if (file_priv->filp == filp)
-			return file_priv;
-
-	WARN_ON(1);
-
-	return ERR_PTR(-EFAULT);
-}
-
-static int exynos_drm_gem_mmap_buffer(struct file *filp,
+int exynos_drm_gem_mmap_buffer(struct file *filp,
 				      struct vm_area_struct *vma)
 {
 	struct drm_gem_object *obj = filp->private_data;
 	struct exynos_drm_gem_obj *exynos_gem_obj = to_exynos_gem_obj(obj);
 	struct drm_device *drm_dev = obj->dev;
 	struct exynos_drm_gem_buf *buffer;
-	struct drm_file *file_priv;
 	unsigned long vm_size;
 	int ret;
 
+	WARN_ON(!mutex_is_locked(&obj->dev->struct_mutex));
+
 	vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP;
 	vma->vm_private_data = obj;
 	vma->vm_ops = drm_dev->driver->gem_vm_ops;
 
-	/* restore it to driver's fops. */
-	filp->f_op = fops_get(drm_dev->driver->fops);
-
-	file_priv = exynos_drm_find_drm_file(drm_dev, filp);
-	if (IS_ERR(file_priv))
-		return PTR_ERR(file_priv);
-
-	/* restore it to drm_file. */
-	filp->private_data = file_priv;
-
 	update_vm_cache_attr(exynos_gem_obj, vma);
 
 	vm_size = vma->vm_end - vma->vm_start;
@@ -411,15 +387,13 @@ static int exynos_drm_gem_mmap_buffer(struct file *filp,
 	return 0;
 }
 
-static const struct file_operations exynos_drm_gem_fops = {
-	.mmap = exynos_drm_gem_mmap_buffer,
-};
-
 int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
 			      struct drm_file *file_priv)
 {
+	struct drm_exynos_file_private *exynos_file_priv;
 	struct drm_exynos_gem_mmap *args = data;
 	struct drm_gem_object *obj;
+	struct file *anon_filp;
 	unsigned long addr;
 
 	if (!(dev->driver->driver_features & DRIVER_GEM)) {
@@ -427,47 +401,25 @@ int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
 		return -ENODEV;
 	}
 
+	mutex_lock(&dev->struct_mutex);
+
 	obj = drm_gem_object_lookup(dev, file_priv, args->handle);
 	if (!obj) {
 		DRM_ERROR("failed to lookup gem object.\n");
+		mutex_unlock(&dev->struct_mutex);
 		return -EINVAL;
 	}
 
-	/*
-	 * We have to use gem object and its fops for specific mmaper,
-	 * but vm_mmap() can deliver only filp. So we have to change
-	 * filp->f_op and filp->private_data temporarily, then restore
-	 * again. So it is important to keep lock until restoration the
-	 * settings to prevent others from misuse of filp->f_op or
-	 * filp->private_data.
-	 */
-	mutex_lock(&dev->struct_mutex);
-
-	/*
-	 * Set specific mmper's fops. And it will be restored by
-	 * exynos_drm_gem_mmap_buffer to dev->driver->fops.
-	 * This is used to call specific mapper temporarily.
-	 */
-	file_priv->filp->f_op = &exynos_drm_gem_fops;
-
-	/*
-	 * Set gem object to private_data so that specific mmaper
-	 * can get the gem object. And it will be restored by
-	 * exynos_drm_gem_mmap_buffer to drm_file.
-	 */
-	file_priv->filp->private_data = obj;
+	exynos_file_priv = file_priv->driver_priv;
+	anon_filp = exynos_file_priv->anon_filp;
+	anon_filp->private_data = obj;
 
-	addr = vm_mmap(file_priv->filp, 0, args->size,
-			PROT_READ | PROT_WRITE, MAP_SHARED, 0);
+	addr = vm_mmap(anon_filp, 0, args->size, PROT_READ | PROT_WRITE,
+			MAP_SHARED, 0);
 
 	drm_gem_object_unreference(obj);
 
 	if (IS_ERR_VALUE(addr)) {
-		/* check filp->f_op, filp->private_data are restored */
-		if (file_priv->filp->f_op == &exynos_drm_gem_fops) {
-			file_priv->filp->f_op = fops_get(dev->driver->fops);
-			file_priv->filp->private_data = file_priv;
-		}
 		mutex_unlock(&dev->struct_mutex);
 		return (int)addr;
 	}
@@ -660,22 +612,20 @@ int exynos_drm_gem_dumb_create(struct drm_file *file_priv,
 	args->pitch = args->width * ((args->bpp + 7) / 8);
 	args->size = args->pitch * args->height;
 
-	exynos_gem_obj = exynos_drm_gem_create(dev, EXYNOS_BO_CONTIG |
-						EXYNOS_BO_WC, args->size);
-	/*
-	 * If physically contiguous memory allocation fails and if IOMMU is
-	 * supported then try to get buffer from non physically contiguous
-	 * memory area.
-	 */
-	if (IS_ERR(exynos_gem_obj) && is_drm_iommu_supported(dev)) {
-		dev_warn(dev->dev, "contiguous FB allocation failed, falling back to non-contiguous\n");
+	if (is_drm_iommu_supported(dev)) {
 		exynos_gem_obj = exynos_drm_gem_create(dev,
-					EXYNOS_BO_NONCONTIG | EXYNOS_BO_WC,
-					args->size);
+			EXYNOS_BO_NONCONTIG | EXYNOS_BO_WC,
+			args->size);
+	} else {
+		exynos_gem_obj = exynos_drm_gem_create(dev,
+			EXYNOS_BO_CONTIG | EXYNOS_BO_WC,
+			args->size);
 	}
 
-	if (IS_ERR(exynos_gem_obj))
+	if (IS_ERR(exynos_gem_obj)) {
+		dev_warn(dev->dev, "FB allocation failed.\n");
 		return PTR_ERR(exynos_gem_obj);
+	}
 
 	ret = exynos_drm_gem_handle_create(&exynos_gem_obj->base, file_priv,
 			&args->handle);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.h b/drivers/gpu/drm/exynos/exynos_drm_gem.h
index b8c818ba2ff..1592c0ba7de 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.h
@@ -122,6 +122,9 @@ int exynos_drm_gem_map_offset_ioctl(struct drm_device *dev, void *data,
 int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
 			      struct drm_file *file_priv);
 
+int exynos_drm_gem_mmap_buffer(struct file *filp,
+				      struct vm_area_struct *vma);
+
 /* map user space allocated by malloc to pages. */
 int exynos_drm_gem_userptr_ioctl(struct drm_device *dev, void *data,
 				      struct drm_file *file_priv);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gsc.c b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
index fa75059a610..9e3ff167296 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gsc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
@@ -1335,11 +1335,7 @@ static irqreturn_t gsc_irq_handler(int irq, void *dev_id)
 
 static int gsc_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
 {
-	struct drm_exynos_ipp_prop_list *prop_list;
-
-	prop_list = devm_kzalloc(ippdrv->dev, sizeof(*prop_list), GFP_KERNEL);
-	if (!prop_list)
-		return -ENOMEM;
+	struct drm_exynos_ipp_prop_list *prop_list = &ippdrv->prop_list;
 
 	prop_list->version = 1;
 	prop_list->writeback = 1;
@@ -1363,8 +1359,6 @@ static int gsc_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
 	prop_list->scale_min.hsize = GSC_SCALE_MIN;
 	prop_list->scale_min.vsize = GSC_SCALE_MIN;
 
-	ippdrv->prop_list = prop_list;
-
 	return 0;
 }
 
@@ -1387,7 +1381,7 @@ static int gsc_ippdrv_check_property(struct device *dev,
 {
 	struct gsc_context *ctx = get_gsc_context(dev);
 	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_prop_list *pp = ippdrv->prop_list;
+	struct drm_exynos_ipp_prop_list *pp = &ippdrv->prop_list;
 	struct drm_exynos_ipp_config *config;
 	struct drm_exynos_pos *pos;
 	struct drm_exynos_sz *sz;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_hdmi.c b/drivers/gpu/drm/exynos/exynos_drm_hdmi.c
deleted file mode 100644
index 8548b974bd5..00000000000
--- a/drivers/gpu/drm/exynos/exynos_drm_hdmi.c
+++ /dev/null
@@ -1,439 +0,0 @@
-/*
- * Copyright (C) 2011 Samsung Electronics Co.Ltd
- * Authors:
- *	Inki Dae <inki.dae@samsung.com>
- *	Seung-Woo Kim <sw0312.kim@samsung.com>
- *
- * This program is free software; you can redistribute  it and/or modify it
- * under  the terms of  the GNU General  Public License as published by the
- * Free Software Foundation;  either version 2 of the  License, or (at your
- * option) any later version.
- *
- */
-
-#include <drm/drmP.h>
-
-#include <linux/kernel.h>
-#include <linux/wait.h>
-#include <linux/platform_device.h>
-#include <linux/pm_runtime.h>
-
-#include <drm/exynos_drm.h>
-
-#include "exynos_drm_drv.h"
-#include "exynos_drm_hdmi.h"
-
-#define to_context(dev)		platform_get_drvdata(to_platform_device(dev))
-#define to_subdrv(dev)		to_context(dev)
-#define get_ctx_from_subdrv(subdrv)	container_of(subdrv,\
-					struct drm_hdmi_context, subdrv);
-
-/* platform device pointer for common drm hdmi device. */
-static struct platform_device *exynos_drm_hdmi_pdev;
-
-/* Common hdmi subdrv needs to access the hdmi and mixer though context.
-* These should be initialied by the repective drivers */
-static struct exynos_drm_hdmi_context *hdmi_ctx;
-static struct exynos_drm_hdmi_context *mixer_ctx;
-
-/* these callback points shoud be set by specific drivers. */
-static struct exynos_hdmi_ops *hdmi_ops;
-static struct exynos_mixer_ops *mixer_ops;
-
-struct drm_hdmi_context {
-	struct exynos_drm_subdrv	subdrv;
-	struct exynos_drm_hdmi_context	*hdmi_ctx;
-	struct exynos_drm_hdmi_context	*mixer_ctx;
-
-	bool	enabled[MIXER_WIN_NR];
-};
-
-int exynos_platform_device_hdmi_register(void)
-{
-	struct platform_device *pdev;
-
-	if (exynos_drm_hdmi_pdev)
-		return -EEXIST;
-
-	pdev = platform_device_register_simple(
-			"exynos-drm-hdmi", -1, NULL, 0);
-	if (IS_ERR(pdev))
-		return PTR_ERR(pdev);
-
-	exynos_drm_hdmi_pdev = pdev;
-
-	return 0;
-}
-
-void exynos_platform_device_hdmi_unregister(void)
-{
-	if (exynos_drm_hdmi_pdev) {
-		platform_device_unregister(exynos_drm_hdmi_pdev);
-		exynos_drm_hdmi_pdev = NULL;
-	}
-}
-
-void exynos_hdmi_drv_attach(struct exynos_drm_hdmi_context *ctx)
-{
-	if (ctx)
-		hdmi_ctx = ctx;
-}
-
-void exynos_mixer_drv_attach(struct exynos_drm_hdmi_context *ctx)
-{
-	if (ctx)
-		mixer_ctx = ctx;
-}
-
-void exynos_hdmi_ops_register(struct exynos_hdmi_ops *ops)
-{
-	if (ops)
-		hdmi_ops = ops;
-}
-
-void exynos_mixer_ops_register(struct exynos_mixer_ops *ops)
-{
-	if (ops)
-		mixer_ops = ops;
-}
-
-static bool drm_hdmi_is_connected(struct device *dev)
-{
-	struct drm_hdmi_context *ctx = to_context(dev);
-
-	if (hdmi_ops && hdmi_ops->is_connected)
-		return hdmi_ops->is_connected(ctx->hdmi_ctx->ctx);
-
-	return false;
-}
-
-static struct edid *drm_hdmi_get_edid(struct device *dev,
-			struct drm_connector *connector)
-{
-	struct drm_hdmi_context *ctx = to_context(dev);
-
-	if (hdmi_ops && hdmi_ops->get_edid)
-		return hdmi_ops->get_edid(ctx->hdmi_ctx->ctx, connector);
-
-	return NULL;
-}
-
-static int drm_hdmi_check_mode(struct device *dev,
-		struct drm_display_mode *mode)
-{
-	struct drm_hdmi_context *ctx = to_context(dev);
-	int ret = 0;
-
-	/*
-	* Both, mixer and hdmi should be able to handle the requested mode.
-	* If any of the two fails, return mode as BAD.
-	*/
-
-	if (mixer_ops && mixer_ops->check_mode)
-		ret = mixer_ops->check_mode(ctx->mixer_ctx->ctx, mode);
-
-	if (ret)
-		return ret;
-
-	if (hdmi_ops && hdmi_ops->check_mode)
-		return hdmi_ops->check_mode(ctx->hdmi_ctx->ctx, mode);
-
-	return 0;
-}
-
-static int drm_hdmi_power_on(struct device *dev, int mode)
-{
-	struct drm_hdmi_context *ctx = to_context(dev);
-
-	if (hdmi_ops && hdmi_ops->power_on)
-		return hdmi_ops->power_on(ctx->hdmi_ctx->ctx, mode);
-
-	return 0;
-}
-
-static struct exynos_drm_display_ops drm_hdmi_display_ops = {
-	.type = EXYNOS_DISPLAY_TYPE_HDMI,
-	.is_connected = drm_hdmi_is_connected,
-	.get_edid = drm_hdmi_get_edid,
-	.check_mode = drm_hdmi_check_mode,
-	.power_on = drm_hdmi_power_on,
-};
-
-static int drm_hdmi_enable_vblank(struct device *subdrv_dev)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-	struct exynos_drm_subdrv *subdrv = &ctx->subdrv;
-	struct exynos_drm_manager *manager = subdrv->manager;
-
-	if (mixer_ops && mixer_ops->enable_vblank)
-		return mixer_ops->enable_vblank(ctx->mixer_ctx->ctx,
-						manager->pipe);
-
-	return 0;
-}
-
-static void drm_hdmi_disable_vblank(struct device *subdrv_dev)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-
-	if (mixer_ops && mixer_ops->disable_vblank)
-		return mixer_ops->disable_vblank(ctx->mixer_ctx->ctx);
-}
-
-static void drm_hdmi_wait_for_vblank(struct device *subdrv_dev)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-
-	if (mixer_ops && mixer_ops->wait_for_vblank)
-		mixer_ops->wait_for_vblank(ctx->mixer_ctx->ctx);
-}
-
-static void drm_hdmi_mode_fixup(struct device *subdrv_dev,
-				struct drm_connector *connector,
-				const struct drm_display_mode *mode,
-				struct drm_display_mode *adjusted_mode)
-{
-	struct drm_display_mode *m;
-	int mode_ok;
-
-	drm_mode_set_crtcinfo(adjusted_mode, 0);
-
-	mode_ok = drm_hdmi_check_mode(subdrv_dev, adjusted_mode);
-
-	/* just return if user desired mode exists. */
-	if (mode_ok == 0)
-		return;
-
-	/*
-	 * otherwise, find the most suitable mode among modes and change it
-	 * to adjusted_mode.
-	 */
-	list_for_each_entry(m, &connector->modes, head) {
-		mode_ok = drm_hdmi_check_mode(subdrv_dev, m);
-
-		if (mode_ok == 0) {
-			struct drm_mode_object base;
-			struct list_head head;
-
-			DRM_INFO("desired mode doesn't exist so\n");
-			DRM_INFO("use the most suitable mode among modes.\n");
-
-			DRM_DEBUG_KMS("Adjusted Mode: [%d]x[%d] [%d]Hz\n",
-				m->hdisplay, m->vdisplay, m->vrefresh);
-
-			/* preserve display mode header while copying. */
-			head = adjusted_mode->head;
-			base = adjusted_mode->base;
-			memcpy(adjusted_mode, m, sizeof(*m));
-			adjusted_mode->head = head;
-			adjusted_mode->base = base;
-			break;
-		}
-	}
-}
-
-static void drm_hdmi_mode_set(struct device *subdrv_dev, void *mode)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-
-	if (hdmi_ops && hdmi_ops->mode_set)
-		hdmi_ops->mode_set(ctx->hdmi_ctx->ctx, mode);
-}
-
-static void drm_hdmi_get_max_resol(struct device *subdrv_dev,
-				unsigned int *width, unsigned int *height)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-
-	if (hdmi_ops && hdmi_ops->get_max_resol)
-		hdmi_ops->get_max_resol(ctx->hdmi_ctx->ctx, width, height);
-}
-
-static void drm_hdmi_commit(struct device *subdrv_dev)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-
-	if (hdmi_ops && hdmi_ops->commit)
-		hdmi_ops->commit(ctx->hdmi_ctx->ctx);
-}
-
-static void drm_hdmi_dpms(struct device *subdrv_dev, int mode)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-
-	if (mixer_ops && mixer_ops->dpms)
-		mixer_ops->dpms(ctx->mixer_ctx->ctx, mode);
-
-	if (hdmi_ops && hdmi_ops->dpms)
-		hdmi_ops->dpms(ctx->hdmi_ctx->ctx, mode);
-}
-
-static void drm_hdmi_apply(struct device *subdrv_dev)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-	int i;
-
-	for (i = 0; i < MIXER_WIN_NR; i++) {
-		if (!ctx->enabled[i])
-			continue;
-		if (mixer_ops && mixer_ops->win_commit)
-			mixer_ops->win_commit(ctx->mixer_ctx->ctx, i);
-	}
-
-	if (hdmi_ops && hdmi_ops->commit)
-		hdmi_ops->commit(ctx->hdmi_ctx->ctx);
-}
-
-static struct exynos_drm_manager_ops drm_hdmi_manager_ops = {
-	.dpms = drm_hdmi_dpms,
-	.apply = drm_hdmi_apply,
-	.enable_vblank = drm_hdmi_enable_vblank,
-	.disable_vblank = drm_hdmi_disable_vblank,
-	.wait_for_vblank = drm_hdmi_wait_for_vblank,
-	.mode_fixup = drm_hdmi_mode_fixup,
-	.mode_set = drm_hdmi_mode_set,
-	.get_max_resol = drm_hdmi_get_max_resol,
-	.commit = drm_hdmi_commit,
-};
-
-static void drm_mixer_mode_set(struct device *subdrv_dev,
-		struct exynos_drm_overlay *overlay)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-
-	if (mixer_ops && mixer_ops->win_mode_set)
-		mixer_ops->win_mode_set(ctx->mixer_ctx->ctx, overlay);
-}
-
-static void drm_mixer_commit(struct device *subdrv_dev, int zpos)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-	int win = (zpos == DEFAULT_ZPOS) ? MIXER_DEFAULT_WIN : zpos;
-
-	if (win < 0 || win >= MIXER_WIN_NR) {
-		DRM_ERROR("mixer window[%d] is wrong\n", win);
-		return;
-	}
-
-	if (mixer_ops && mixer_ops->win_commit)
-		mixer_ops->win_commit(ctx->mixer_ctx->ctx, win);
-
-	ctx->enabled[win] = true;
-}
-
-static void drm_mixer_disable(struct device *subdrv_dev, int zpos)
-{
-	struct drm_hdmi_context *ctx = to_context(subdrv_dev);
-	int win = (zpos == DEFAULT_ZPOS) ? MIXER_DEFAULT_WIN : zpos;
-
-	if (win < 0 || win >= MIXER_WIN_NR) {
-		DRM_ERROR("mixer window[%d] is wrong\n", win);
-		return;
-	}
-
-	if (mixer_ops && mixer_ops->win_disable)
-		mixer_ops->win_disable(ctx->mixer_ctx->ctx, win);
-
-	ctx->enabled[win] = false;
-}
-
-static struct exynos_drm_overlay_ops drm_hdmi_overlay_ops = {
-	.mode_set = drm_mixer_mode_set,
-	.commit = drm_mixer_commit,
-	.disable = drm_mixer_disable,
-};
-
-static struct exynos_drm_manager hdmi_manager = {
-	.pipe		= -1,
-	.ops		= &drm_hdmi_manager_ops,
-	.overlay_ops	= &drm_hdmi_overlay_ops,
-	.display_ops	= &drm_hdmi_display_ops,
-};
-
-static int hdmi_subdrv_probe(struct drm_device *drm_dev,
-		struct device *dev)
-{
-	struct exynos_drm_subdrv *subdrv = to_subdrv(dev);
-	struct drm_hdmi_context *ctx;
-
-	if (!hdmi_ctx) {
-		DRM_ERROR("hdmi context not initialized.\n");
-		return -EFAULT;
-	}
-
-	if (!mixer_ctx) {
-		DRM_ERROR("mixer context not initialized.\n");
-		return -EFAULT;
-	}
-
-	ctx = get_ctx_from_subdrv(subdrv);
-
-	if (!ctx) {
-		DRM_ERROR("no drm hdmi context.\n");
-		return -EFAULT;
-	}
-
-	ctx->hdmi_ctx = hdmi_ctx;
-	ctx->mixer_ctx = mixer_ctx;
-
-	ctx->hdmi_ctx->drm_dev = drm_dev;
-	ctx->mixer_ctx->drm_dev = drm_dev;
-
-	if (mixer_ops->iommu_on)
-		mixer_ops->iommu_on(ctx->mixer_ctx->ctx, true);
-
-	return 0;
-}
-
-static void hdmi_subdrv_remove(struct drm_device *drm_dev, struct device *dev)
-{
-	struct drm_hdmi_context *ctx;
-	struct exynos_drm_subdrv *subdrv = to_subdrv(dev);
-
-	ctx = get_ctx_from_subdrv(subdrv);
-
-	if (mixer_ops->iommu_on)
-		mixer_ops->iommu_on(ctx->mixer_ctx->ctx, false);
-}
-
-static int exynos_drm_hdmi_probe(struct platform_device *pdev)
-{
-	struct device *dev = &pdev->dev;
-	struct exynos_drm_subdrv *subdrv;
-	struct drm_hdmi_context *ctx;
-
-	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
-	if (!ctx)
-		return -ENOMEM;
-
-	subdrv = &ctx->subdrv;
-
-	subdrv->dev = dev;
-	subdrv->manager = &hdmi_manager;
-	subdrv->probe = hdmi_subdrv_probe;
-	subdrv->remove = hdmi_subdrv_remove;
-
-	platform_set_drvdata(pdev, subdrv);
-
-	exynos_drm_subdrv_register(subdrv);
-
-	return 0;
-}
-
-static int exynos_drm_hdmi_remove(struct platform_device *pdev)
-{
-	struct drm_hdmi_context *ctx = platform_get_drvdata(pdev);
-
-	exynos_drm_subdrv_unregister(&ctx->subdrv);
-
-	return 0;
-}
-
-struct platform_driver exynos_drm_common_hdmi_driver = {
-	.probe		= exynos_drm_hdmi_probe,
-	.remove		= exynos_drm_hdmi_remove,
-	.driver		= {
-		.name	= "exynos-drm-hdmi",
-		.owner	= THIS_MODULE,
-	},
-};
diff --git a/drivers/gpu/drm/exynos/exynos_drm_hdmi.h b/drivers/gpu/drm/exynos/exynos_drm_hdmi.h
deleted file mode 100644
index 724cab18197..00000000000
--- a/drivers/gpu/drm/exynos/exynos_drm_hdmi.h
+++ /dev/null
@@ -1,67 +0,0 @@
-/* exynos_drm_hdmi.h
- *
- * Copyright (c) 2011 Samsung Electronics Co., Ltd.
- * Authoer: Inki Dae <inki.dae@samsung.com>
- *
- * This program is free software; you can redistribute  it and/or modify it
- * under  the terms of  the GNU General  Public License as published by the
- * Free Software Foundation;  either version 2 of the  License, or (at your
- * option) any later version.
- */
-
-#ifndef _EXYNOS_DRM_HDMI_H_
-#define _EXYNOS_DRM_HDMI_H_
-
-#define MIXER_WIN_NR		3
-#define MIXER_DEFAULT_WIN	0
-
-/*
- * exynos hdmi common context structure.
- *
- * @drm_dev: pointer to drm_device.
- * @ctx: pointer to the context of specific device driver.
- *	this context should be hdmi_context or mixer_context.
- */
-struct exynos_drm_hdmi_context {
-	struct drm_device	*drm_dev;
-	void			*ctx;
-};
-
-struct exynos_hdmi_ops {
-	/* display */
-	bool (*is_connected)(void *ctx);
-	struct edid *(*get_edid)(void *ctx,
-			struct drm_connector *connector);
-	int (*check_mode)(void *ctx, struct drm_display_mode *mode);
-	int (*power_on)(void *ctx, int mode);
-
-	/* manager */
-	void (*mode_set)(void *ctx, struct drm_display_mode *mode);
-	void (*get_max_resol)(void *ctx, unsigned int *width,
-				unsigned int *height);
-	void (*commit)(void *ctx);
-	void (*dpms)(void *ctx, int mode);
-};
-
-struct exynos_mixer_ops {
-	/* manager */
-	int (*iommu_on)(void *ctx, bool enable);
-	int (*enable_vblank)(void *ctx, int pipe);
-	void (*disable_vblank)(void *ctx);
-	void (*wait_for_vblank)(void *ctx);
-	void (*dpms)(void *ctx, int mode);
-
-	/* overlay */
-	void (*win_mode_set)(void *ctx, struct exynos_drm_overlay *overlay);
-	void (*win_commit)(void *ctx, int zpos);
-	void (*win_disable)(void *ctx, int zpos);
-
-	/* display */
-	int (*check_mode)(void *ctx, struct drm_display_mode *mode);
-};
-
-void exynos_hdmi_drv_attach(struct exynos_drm_hdmi_context *ctx);
-void exynos_mixer_drv_attach(struct exynos_drm_hdmi_context *ctx);
-void exynos_hdmi_ops_register(struct exynos_hdmi_ops *ops);
-void exynos_mixer_ops_register(struct exynos_mixer_ops *ops);
-#endif
diff --git a/drivers/gpu/drm/exynos/exynos_drm_iommu.c b/drivers/gpu/drm/exynos/exynos_drm_iommu.c
index fb8db037827..b32b291f88f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_iommu.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_iommu.c
@@ -36,12 +36,10 @@ int drm_create_iommu_mapping(struct drm_device *drm_dev)
 		priv->da_start = EXYNOS_DEV_ADDR_START;
 	if (!priv->da_space_size)
 		priv->da_space_size = EXYNOS_DEV_ADDR_SIZE;
-	if (!priv->da_space_order)
-		priv->da_space_order = EXYNOS_DEV_ADDR_ORDER;
 
 	mapping = arm_iommu_create_mapping(&platform_bus_type, priv->da_start,
-						priv->da_space_size,
-						priv->da_space_order);
+						priv->da_space_size);
+
 	if (IS_ERR(mapping))
 		return PTR_ERR(mapping);
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_iommu.h b/drivers/gpu/drm/exynos/exynos_drm_iommu.h
index 598e60f57d4..72376d41c51 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_iommu.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_iommu.h
@@ -14,7 +14,6 @@
 
 #define EXYNOS_DEV_ADDR_START	0x20000000
 #define EXYNOS_DEV_ADDR_SIZE	0x40000000
-#define EXYNOS_DEV_ADDR_ORDER	0x0
 
 #ifdef CONFIG_DRM_EXYNOS_IOMMU
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.c b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
index d519a4e5fe4..a1888e128f1 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_ipp.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
@@ -16,7 +16,6 @@
 #include <linux/types.h>
 #include <linux/clk.h>
 #include <linux/pm_runtime.h>
-#include <plat/map-base.h>
 
 #include <drm/drmP.h>
 #include <drm/exynos_drm.h>
@@ -168,6 +167,13 @@ static int ipp_create_id(struct idr *id_idr, struct mutex *lock, void *obj,
 	return 0;
 }
 
+static void ipp_remove_id(struct idr *id_idr, struct mutex *lock, u32 id)
+{
+	mutex_lock(lock);
+	idr_remove(id_idr, id);
+	mutex_unlock(lock);
+}
+
 static void *ipp_find_obj(struct idr *id_idr, struct mutex *lock, u32 id)
 {
 	void *obj;
@@ -277,24 +283,22 @@ static struct exynos_drm_ippdrv *ipp_find_drv_by_handle(u32 prop_id)
 
 	DRM_DEBUG_KMS("prop_id[%d]\n", prop_id);
 
-	if (list_empty(&exynos_drm_ippdrv_list)) {
-		DRM_DEBUG_KMS("ippdrv_list is empty.\n");
-		return ERR_PTR(-ENODEV);
-	}
-
 	/*
 	 * This case is search ipp driver by prop_id handle.
 	 * sometimes, ipp subsystem find driver by prop_id.
-	 * e.g PAUSE state, queue buf, command contro.
+	 * e.g PAUSE state, queue buf, command control.
 	 */
 	list_for_each_entry(ippdrv, &exynos_drm_ippdrv_list, drv_list) {
 		DRM_DEBUG_KMS("count[%d]ippdrv[0x%x]\n", count++, (int)ippdrv);
 
-		if (!list_empty(&ippdrv->cmd_list)) {
-			list_for_each_entry(c_node, &ippdrv->cmd_list, list)
-				if (c_node->property.prop_id == prop_id)
-					return ippdrv;
+		mutex_lock(&ippdrv->cmd_lock);
+		list_for_each_entry(c_node, &ippdrv->cmd_list, list) {
+			if (c_node->property.prop_id == prop_id) {
+				mutex_unlock(&ippdrv->cmd_lock);
+				return ippdrv;
+			}
 		}
+		mutex_unlock(&ippdrv->cmd_lock);
 	}
 
 	return ERR_PTR(-ENODEV);
@@ -326,6 +330,7 @@ int exynos_drm_ipp_get_property(struct drm_device *drm_dev, void *data,
 	if (!prop_list->ipp_id) {
 		list_for_each_entry(ippdrv, &exynos_drm_ippdrv_list, drv_list)
 			count++;
+
 		/*
 		 * Supports ippdrv list count for user application.
 		 * First step user application getting ippdrv count.
@@ -347,7 +352,7 @@ int exynos_drm_ipp_get_property(struct drm_device *drm_dev, void *data,
 			return PTR_ERR(ippdrv);
 		}
 
-		prop_list = ippdrv->prop_list;
+		*prop_list = ippdrv->prop_list;
 	}
 
 	return 0;
@@ -387,9 +392,11 @@ static int ipp_find_and_set_property(struct drm_exynos_ipp_property *property)
 	 * when we find this command no using prop_id.
 	 * return property information set in this command node.
 	 */
+	mutex_lock(&ippdrv->cmd_lock);
 	list_for_each_entry(c_node, &ippdrv->cmd_list, list) {
 		if ((c_node->property.prop_id == prop_id) &&
 		    (c_node->state == IPP_STATE_STOP)) {
+			mutex_unlock(&ippdrv->cmd_lock);
 			DRM_DEBUG_KMS("found cmd[%d]ippdrv[0x%x]\n",
 				property->cmd, (int)ippdrv);
 
@@ -397,6 +404,7 @@ static int ipp_find_and_set_property(struct drm_exynos_ipp_property *property)
 			return 0;
 		}
 	}
+	mutex_unlock(&ippdrv->cmd_lock);
 
 	DRM_ERROR("failed to search property.\n");
 
@@ -500,7 +508,7 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 	c_node->start_work = ipp_create_cmd_work();
 	if (IS_ERR(c_node->start_work)) {
 		DRM_ERROR("failed to create start work.\n");
-		goto err_clear;
+		goto err_remove_id;
 	}
 
 	c_node->stop_work = ipp_create_cmd_work();
@@ -515,7 +523,7 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 		goto err_free_stop;
 	}
 
-	mutex_init(&c_node->cmd_lock);
+	mutex_init(&c_node->lock);
 	mutex_init(&c_node->mem_lock);
 	mutex_init(&c_node->event_lock);
 
@@ -527,7 +535,9 @@ int exynos_drm_ipp_set_property(struct drm_device *drm_dev, void *data,
 
 	INIT_LIST_HEAD(&c_node->event_list);
 	list_splice_init(&priv->event_list, &c_node->event_list);
+	mutex_lock(&ippdrv->cmd_lock);
 	list_add_tail(&c_node->list, &ippdrv->cmd_list);
+	mutex_unlock(&ippdrv->cmd_lock);
 
 	/* make dedicated state without m2m */
 	if (!ipp_is_m2m_cmd(property->cmd))
@@ -539,18 +549,24 @@ err_free_stop:
 	kfree(c_node->stop_work);
 err_free_start:
 	kfree(c_node->start_work);
+err_remove_id:
+	ipp_remove_id(&ctx->prop_idr, &ctx->prop_lock, property->prop_id);
 err_clear:
 	kfree(c_node);
 	return ret;
 }
 
-static void ipp_clean_cmd_node(struct drm_exynos_ipp_cmd_node *c_node)
+static void ipp_clean_cmd_node(struct ipp_context *ctx,
+				struct drm_exynos_ipp_cmd_node *c_node)
 {
 	/* delete list */
 	list_del(&c_node->list);
 
+	ipp_remove_id(&ctx->prop_idr, &ctx->prop_lock,
+			c_node->property.prop_id);
+
 	/* destroy mutex */
-	mutex_destroy(&c_node->cmd_lock);
+	mutex_destroy(&c_node->lock);
 	mutex_destroy(&c_node->mem_lock);
 	mutex_destroy(&c_node->event_lock);
 
@@ -568,17 +584,10 @@ static int ipp_check_mem_list(struct drm_exynos_ipp_cmd_node *c_node)
 	struct list_head *head;
 	int ret, i, count[EXYNOS_DRM_OPS_MAX] = { 0, };
 
-	mutex_lock(&c_node->mem_lock);
-
 	for_each_ipp_ops(i) {
 		/* source/destination memory list */
 		head = &c_node->mem_list[i];
 
-		if (list_empty(head)) {
-			DRM_DEBUG_KMS("%s memory empty.\n", i ? "dst" : "src");
-			continue;
-		}
-
 		/* find memory node entry */
 		list_for_each_entry(m_node, head, list) {
 			DRM_DEBUG_KMS("%s,count[%d]m_node[0x%x]\n",
@@ -603,8 +612,6 @@ static int ipp_check_mem_list(struct drm_exynos_ipp_cmd_node *c_node)
 		ret = max(count[EXYNOS_DRM_OPS_SRC],
 			count[EXYNOS_DRM_OPS_DST]);
 
-	mutex_unlock(&c_node->mem_lock);
-
 	return ret;
 }
 
@@ -647,16 +654,13 @@ static int ipp_set_mem_node(struct exynos_drm_ippdrv *ippdrv,
 		return -EFAULT;
 	}
 
-	mutex_lock(&c_node->mem_lock);
-
 	DRM_DEBUG_KMS("ops_id[%d]\n", m_node->ops_id);
 
 	/* get operations callback */
 	ops = ippdrv->ops[m_node->ops_id];
 	if (!ops) {
 		DRM_ERROR("not support ops.\n");
-		ret = -EFAULT;
-		goto err_unlock;
+		return -EFAULT;
 	}
 
 	/* set address and enable irq */
@@ -665,12 +669,10 @@ static int ipp_set_mem_node(struct exynos_drm_ippdrv *ippdrv,
 			m_node->buf_id, IPP_BUF_ENQUEUE);
 		if (ret) {
 			DRM_ERROR("failed to set addr.\n");
-			goto err_unlock;
+			return ret;
 		}
 	}
 
-err_unlock:
-	mutex_unlock(&c_node->mem_lock);
 	return ret;
 }
 
@@ -685,11 +687,9 @@ static struct drm_exynos_ipp_mem_node
 	void *addr;
 	int i;
 
-	mutex_lock(&c_node->mem_lock);
-
 	m_node = kzalloc(sizeof(*m_node), GFP_KERNEL);
 	if (!m_node)
-		goto err_unlock;
+		return ERR_PTR(-ENOMEM);
 
 	/* clear base address for error handling */
 	memset(&buf_info, 0x0, sizeof(buf_info));
@@ -723,15 +723,14 @@ static struct drm_exynos_ipp_mem_node
 
 	m_node->filp = file;
 	m_node->buf_info = buf_info;
+	mutex_lock(&c_node->mem_lock);
 	list_add_tail(&m_node->list, &c_node->mem_list[qbuf->ops_id]);
-
 	mutex_unlock(&c_node->mem_lock);
+
 	return m_node;
 
 err_clear:
 	kfree(m_node);
-err_unlock:
-	mutex_unlock(&c_node->mem_lock);
 	return ERR_PTR(-EFAULT);
 }
 
@@ -748,13 +747,6 @@ static int ipp_put_mem_node(struct drm_device *drm_dev,
 		return -EFAULT;
 	}
 
-	if (list_empty(&m_node->list)) {
-		DRM_ERROR("empty memory node.\n");
-		return -ENOMEM;
-	}
-
-	mutex_lock(&c_node->mem_lock);
-
 	DRM_DEBUG_KMS("ops_id[%d]\n", m_node->ops_id);
 
 	/* put gem buffer */
@@ -769,8 +761,6 @@ static int ipp_put_mem_node(struct drm_device *drm_dev,
 	list_del(&m_node->list);
 	kfree(m_node);
 
-	mutex_unlock(&c_node->mem_lock);
-
 	return 0;
 }
 
@@ -806,7 +796,9 @@ static int ipp_get_event(struct drm_device *drm_dev,
 	e->base.event = &e->event.base;
 	e->base.file_priv = file;
 	e->base.destroy = ipp_free_event;
+	mutex_lock(&c_node->event_lock);
 	list_add_tail(&e->base.link, &c_node->event_list);
+	mutex_unlock(&c_node->event_lock);
 
 	return 0;
 }
@@ -817,16 +809,12 @@ static void ipp_put_event(struct drm_exynos_ipp_cmd_node *c_node,
 	struct drm_exynos_ipp_send_event *e, *te;
 	int count = 0;
 
-	if (list_empty(&c_node->event_list)) {
-		DRM_DEBUG_KMS("event_list is empty.\n");
-		return;
-	}
-
+	mutex_lock(&c_node->event_lock);
 	list_for_each_entry_safe(e, te, &c_node->event_list, base.link) {
 		DRM_DEBUG_KMS("count[%d]e[0x%x]\n", count++, (int)e);
 
 		/*
-		 * quf == NULL condition means all event deletion.
+		 * qbuf == NULL condition means all event deletion.
 		 * stop operations want to delete all event list.
 		 * another case delete only same buf id.
 		 */
@@ -842,9 +830,13 @@ static void ipp_put_event(struct drm_exynos_ipp_cmd_node *c_node,
 			/* delete list */
 			list_del(&e->base.link);
 			kfree(e);
-			return;
+			goto out_unlock;
 		}
 	}
+
+out_unlock:
+	mutex_unlock(&c_node->event_lock);
+	return;
 }
 
 static void ipp_handle_cmd_work(struct device *dev,
@@ -888,7 +880,9 @@ static int ipp_queue_buf_with_run(struct device *dev,
 		return 0;
 	}
 
+	mutex_lock(&c_node->mem_lock);
 	if (!ipp_check_mem_list(c_node)) {
+		mutex_unlock(&c_node->mem_lock);
 		DRM_DEBUG_KMS("empty memory.\n");
 		return 0;
 	}
@@ -905,10 +899,12 @@ static int ipp_queue_buf_with_run(struct device *dev,
 	} else {
 		ret = ipp_set_mem_node(ippdrv, c_node, m_node);
 		if (ret) {
+			mutex_unlock(&c_node->mem_lock);
 			DRM_ERROR("failed to set m node.\n");
 			return ret;
 		}
 	}
+	mutex_unlock(&c_node->mem_lock);
 
 	return 0;
 }
@@ -919,15 +915,15 @@ static void ipp_clean_queue_buf(struct drm_device *drm_dev,
 {
 	struct drm_exynos_ipp_mem_node *m_node, *tm_node;
 
-	if (!list_empty(&c_node->mem_list[qbuf->ops_id])) {
-		/* delete list */
-		list_for_each_entry_safe(m_node, tm_node,
-			&c_node->mem_list[qbuf->ops_id], list) {
-			if (m_node->buf_id == qbuf->buf_id &&
-			    m_node->ops_id == qbuf->ops_id)
-				ipp_put_mem_node(drm_dev, c_node, m_node);
-		}
+	/* delete list */
+	mutex_lock(&c_node->mem_lock);
+	list_for_each_entry_safe(m_node, tm_node,
+		&c_node->mem_list[qbuf->ops_id], list) {
+		if (m_node->buf_id == qbuf->buf_id &&
+		    m_node->ops_id == qbuf->ops_id)
+			ipp_put_mem_node(drm_dev, c_node, m_node);
 	}
+	mutex_unlock(&c_node->mem_lock);
 }
 
 int exynos_drm_ipp_queue_buf(struct drm_device *drm_dev, void *data,
@@ -999,7 +995,7 @@ int exynos_drm_ipp_queue_buf(struct drm_device *drm_dev, void *data,
 		}
 		break;
 	case IPP_BUF_DEQUEUE:
-		mutex_lock(&c_node->cmd_lock);
+		mutex_lock(&c_node->lock);
 
 		/* put event for destination buffer */
 		if (qbuf->ops_id == EXYNOS_DRM_OPS_DST)
@@ -1007,7 +1003,7 @@ int exynos_drm_ipp_queue_buf(struct drm_device *drm_dev, void *data,
 
 		ipp_clean_queue_buf(drm_dev, c_node, qbuf);
 
-		mutex_unlock(&c_node->cmd_lock);
+		mutex_unlock(&c_node->lock);
 		break;
 	default:
 		DRM_ERROR("invalid buffer control.\n");
@@ -1110,12 +1106,12 @@ int exynos_drm_ipp_cmd_ctrl(struct drm_device *drm_dev, void *data,
 	case IPP_CTRL_PLAY:
 		if (pm_runtime_suspended(ippdrv->dev))
 			pm_runtime_get_sync(ippdrv->dev);
+
 		c_node->state = IPP_STATE_START;
 
 		cmd_work = c_node->start_work;
 		cmd_work->ctrl = cmd_ctrl->ctrl;
 		ipp_handle_cmd_work(dev, ippdrv, cmd_work, c_node);
-		c_node->state = IPP_STATE_START;
 		break;
 	case IPP_CTRL_STOP:
 		cmd_work = c_node->stop_work;
@@ -1130,10 +1126,12 @@ int exynos_drm_ipp_cmd_ctrl(struct drm_device *drm_dev, void *data,
 
 		c_node->state = IPP_STATE_STOP;
 		ippdrv->dedicated = false;
-		ipp_clean_cmd_node(c_node);
+		mutex_lock(&ippdrv->cmd_lock);
+		ipp_clean_cmd_node(ctx, c_node);
 
 		if (list_empty(&ippdrv->cmd_list))
 			pm_runtime_put_sync(ippdrv->dev);
+		mutex_unlock(&ippdrv->cmd_lock);
 		break;
 	case IPP_CTRL_PAUSE:
 		cmd_work = c_node->stop_work;
@@ -1261,9 +1259,11 @@ static int ipp_start_property(struct exynos_drm_ippdrv *ippdrv,
 	/* store command info in ippdrv */
 	ippdrv->c_node = c_node;
 
+	mutex_lock(&c_node->mem_lock);
 	if (!ipp_check_mem_list(c_node)) {
 		DRM_DEBUG_KMS("empty memory.\n");
-		return -ENOMEM;
+		ret = -ENOMEM;
+		goto err_unlock;
 	}
 
 	/* set current property in ippdrv */
@@ -1271,7 +1271,7 @@ static int ipp_start_property(struct exynos_drm_ippdrv *ippdrv,
 	if (ret) {
 		DRM_ERROR("failed to set property.\n");
 		ippdrv->c_node = NULL;
-		return ret;
+		goto err_unlock;
 	}
 
 	/* check command */
@@ -1286,7 +1286,7 @@ static int ipp_start_property(struct exynos_drm_ippdrv *ippdrv,
 			if (!m_node) {
 				DRM_ERROR("failed to get node.\n");
 				ret = -EFAULT;
-				return ret;
+				goto err_unlock;
 			}
 
 			DRM_DEBUG_KMS("m_node[0x%x]\n", (int)m_node);
@@ -1294,7 +1294,7 @@ static int ipp_start_property(struct exynos_drm_ippdrv *ippdrv,
 			ret = ipp_set_mem_node(ippdrv, c_node, m_node);
 			if (ret) {
 				DRM_ERROR("failed to set m node.\n");
-				return ret;
+				goto err_unlock;
 			}
 		}
 		break;
@@ -1306,7 +1306,7 @@ static int ipp_start_property(struct exynos_drm_ippdrv *ippdrv,
 			ret = ipp_set_mem_node(ippdrv, c_node, m_node);
 			if (ret) {
 				DRM_ERROR("failed to set m node.\n");
-				return ret;
+				goto err_unlock;
 			}
 		}
 		break;
@@ -1318,14 +1318,16 @@ static int ipp_start_property(struct exynos_drm_ippdrv *ippdrv,
 			ret = ipp_set_mem_node(ippdrv, c_node, m_node);
 			if (ret) {
 				DRM_ERROR("failed to set m node.\n");
-				return ret;
+				goto err_unlock;
 			}
 		}
 		break;
 	default:
 		DRM_ERROR("invalid operations.\n");
-		return -EINVAL;
+		ret = -EINVAL;
+		goto err_unlock;
 	}
+	mutex_unlock(&c_node->mem_lock);
 
 	DRM_DEBUG_KMS("cmd[%d]\n", property->cmd);
 
@@ -1334,11 +1336,17 @@ static int ipp_start_property(struct exynos_drm_ippdrv *ippdrv,
 		ret = ippdrv->start(ippdrv->dev, property->cmd);
 		if (ret) {
 			DRM_ERROR("failed to start ops.\n");
+			ippdrv->c_node = NULL;
 			return ret;
 		}
 	}
 
 	return 0;
+
+err_unlock:
+	mutex_unlock(&c_node->mem_lock);
+	ippdrv->c_node = NULL;
+	return ret;
 }
 
 static int ipp_stop_property(struct drm_device *drm_dev,
@@ -1355,6 +1363,8 @@ static int ipp_stop_property(struct drm_device *drm_dev,
 	/* put event */
 	ipp_put_event(c_node, NULL);
 
+	mutex_lock(&c_node->mem_lock);
+
 	/* check command */
 	switch (property->cmd) {
 	case IPP_CMD_M2M:
@@ -1362,11 +1372,6 @@ static int ipp_stop_property(struct drm_device *drm_dev,
 			/* source/destination memory list */
 			head = &c_node->mem_list[i];
 
-			if (list_empty(head)) {
-				DRM_DEBUG_KMS("mem_list is empty.\n");
-				break;
-			}
-
 			list_for_each_entry_safe(m_node, tm_node,
 				head, list) {
 				ret = ipp_put_mem_node(drm_dev, c_node,
@@ -1382,11 +1387,6 @@ static int ipp_stop_property(struct drm_device *drm_dev,
 		/* destination memory list */
 		head = &c_node->mem_list[EXYNOS_DRM_OPS_DST];
 
-		if (list_empty(head)) {
-			DRM_DEBUG_KMS("mem_list is empty.\n");
-			break;
-		}
-
 		list_for_each_entry_safe(m_node, tm_node, head, list) {
 			ret = ipp_put_mem_node(drm_dev, c_node, m_node);
 			if (ret) {
@@ -1399,11 +1399,6 @@ static int ipp_stop_property(struct drm_device *drm_dev,
 		/* source memory list */
 		head = &c_node->mem_list[EXYNOS_DRM_OPS_SRC];
 
-		if (list_empty(head)) {
-			DRM_DEBUG_KMS("mem_list is empty.\n");
-			break;
-		}
-
 		list_for_each_entry_safe(m_node, tm_node, head, list) {
 			ret = ipp_put_mem_node(drm_dev, c_node, m_node);
 			if (ret) {
@@ -1419,6 +1414,8 @@ static int ipp_stop_property(struct drm_device *drm_dev,
 	}
 
 err_clear:
+	mutex_unlock(&c_node->mem_lock);
+
 	/* stop operations */
 	if (ippdrv->stop)
 		ippdrv->stop(ippdrv->dev, property->cmd);
@@ -1447,7 +1444,7 @@ void ipp_sched_cmd(struct work_struct *work)
 		return;
 	}
 
-	mutex_lock(&c_node->cmd_lock);
+	mutex_lock(&c_node->lock);
 
 	property = &c_node->property;
 
@@ -1495,7 +1492,7 @@ void ipp_sched_cmd(struct work_struct *work)
 	DRM_DEBUG_KMS("ctrl[%d] done.\n", cmd_work->ctrl);
 
 err_unlock:
-	mutex_unlock(&c_node->cmd_lock);
+	mutex_unlock(&c_node->lock);
 }
 
 static int ipp_send_event(struct exynos_drm_ippdrv *ippdrv,
@@ -1525,14 +1522,18 @@ static int ipp_send_event(struct exynos_drm_ippdrv *ippdrv,
 		return -EINVAL;
 	}
 
+	mutex_lock(&c_node->event_lock);
 	if (list_empty(&c_node->event_list)) {
 		DRM_DEBUG_KMS("event list is empty.\n");
-		return 0;
+		ret = 0;
+		goto err_event_unlock;
 	}
 
+	mutex_lock(&c_node->mem_lock);
 	if (!ipp_check_mem_list(c_node)) {
 		DRM_DEBUG_KMS("empty memory.\n");
-		return 0;
+		ret = 0;
+		goto err_mem_unlock;
 	}
 
 	/* check command */
@@ -1546,7 +1547,8 @@ static int ipp_send_event(struct exynos_drm_ippdrv *ippdrv,
 				struct drm_exynos_ipp_mem_node, list);
 			if (!m_node) {
 				DRM_ERROR("empty memory node.\n");
-				return -ENOMEM;
+				ret = -ENOMEM;
+				goto err_mem_unlock;
 			}
 
 			tbuf_id[i] = m_node->buf_id;
@@ -1568,7 +1570,8 @@ static int ipp_send_event(struct exynos_drm_ippdrv *ippdrv,
 		m_node = ipp_find_mem_node(c_node, &qbuf);
 		if (!m_node) {
 			DRM_ERROR("empty memory node.\n");
-			return -ENOMEM;
+			ret = -ENOMEM;
+			goto err_mem_unlock;
 		}
 
 		tbuf_id[EXYNOS_DRM_OPS_DST] = m_node->buf_id;
@@ -1585,7 +1588,8 @@ static int ipp_send_event(struct exynos_drm_ippdrv *ippdrv,
 			struct drm_exynos_ipp_mem_node, list);
 		if (!m_node) {
 			DRM_ERROR("empty memory node.\n");
-			return -ENOMEM;
+			ret = -ENOMEM;
+			goto err_mem_unlock;
 		}
 
 		tbuf_id[EXYNOS_DRM_OPS_SRC] = m_node->buf_id;
@@ -1596,8 +1600,10 @@ static int ipp_send_event(struct exynos_drm_ippdrv *ippdrv,
 		break;
 	default:
 		DRM_ERROR("invalid operations.\n");
-		return -EINVAL;
+		ret = -EINVAL;
+		goto err_mem_unlock;
 	}
+	mutex_unlock(&c_node->mem_lock);
 
 	if (tbuf_id[EXYNOS_DRM_OPS_DST] != buf_id[EXYNOS_DRM_OPS_DST])
 		DRM_ERROR("failed to match buf_id[%d %d]prop_id[%d]\n",
@@ -1612,11 +1618,6 @@ static int ipp_send_event(struct exynos_drm_ippdrv *ippdrv,
 	e = list_first_entry(&c_node->event_list,
 		struct drm_exynos_ipp_send_event, base.link);
 
-	if (!e) {
-		DRM_ERROR("empty event.\n");
-		return -EINVAL;
-	}
-
 	do_gettimeofday(&now);
 	DRM_DEBUG_KMS("tv_sec[%ld]tv_usec[%ld]\n", now.tv_sec, now.tv_usec);
 	e->event.tv_sec = now.tv_sec;
@@ -1631,11 +1632,18 @@ static int ipp_send_event(struct exynos_drm_ippdrv *ippdrv,
 	list_move_tail(&e->base.link, &e->base.file_priv->event_list);
 	wake_up_interruptible(&e->base.file_priv->event_wait);
 	spin_unlock_irqrestore(&drm_dev->event_lock, flags);
+	mutex_unlock(&c_node->event_lock);
 
 	DRM_DEBUG_KMS("done cmd[%d]prop_id[%d]buf_id[%d]\n",
 		property->cmd, property->prop_id, tbuf_id[EXYNOS_DRM_OPS_DST]);
 
 	return 0;
+
+err_mem_unlock:
+	mutex_unlock(&c_node->mem_lock);
+err_event_unlock:
+	mutex_unlock(&c_node->event_lock);
+	return ret;
 }
 
 void ipp_sched_event(struct work_struct *work)
@@ -1677,8 +1685,6 @@ void ipp_sched_event(struct work_struct *work)
 		goto err_completion;
 	}
 
-	mutex_lock(&c_node->event_lock);
-
 	ret = ipp_send_event(ippdrv, c_node, event_work->buf_id);
 	if (ret) {
 		DRM_ERROR("failed to send event.\n");
@@ -1688,8 +1694,6 @@ void ipp_sched_event(struct work_struct *work)
 err_completion:
 	if (ipp_is_m2m_cmd(c_node->property.cmd))
 		complete(&c_node->start_complete);
-
-	mutex_unlock(&c_node->event_lock);
 }
 
 static int ipp_subdrv_probe(struct drm_device *drm_dev, struct device *dev)
@@ -1700,23 +1704,21 @@ static int ipp_subdrv_probe(struct drm_device *drm_dev, struct device *dev)
 
 	/* get ipp driver entry */
 	list_for_each_entry(ippdrv, &exynos_drm_ippdrv_list, drv_list) {
+		u32 ipp_id;
+
 		ippdrv->drm_dev = drm_dev;
 
 		ret = ipp_create_id(&ctx->ipp_idr, &ctx->ipp_lock, ippdrv,
-			&ippdrv->ipp_id);
-		if (ret) {
+				    &ipp_id);
+		if (ret || ipp_id == 0) {
 			DRM_ERROR("failed to create id.\n");
-			goto err_idr;
+			goto err;
 		}
 
 		DRM_DEBUG_KMS("count[%d]ippdrv[0x%x]ipp_id[%d]\n",
-			count++, (int)ippdrv, ippdrv->ipp_id);
+			count++, (int)ippdrv, ipp_id);
 
-		if (ippdrv->ipp_id == 0) {
-			DRM_ERROR("failed to get ipp_id[%d]\n",
-				ippdrv->ipp_id);
-			goto err_idr;
-		}
+		ippdrv->prop_list.ipp_id = ipp_id;
 
 		/* store parent device for node */
 		ippdrv->parent_dev = dev;
@@ -1725,39 +1727,46 @@ static int ipp_subdrv_probe(struct drm_device *drm_dev, struct device *dev)
 		ippdrv->event_workq = ctx->event_workq;
 		ippdrv->sched_event = ipp_sched_event;
 		INIT_LIST_HEAD(&ippdrv->cmd_list);
+		mutex_init(&ippdrv->cmd_lock);
 
 		if (is_drm_iommu_supported(drm_dev)) {
 			ret = drm_iommu_attach_device(drm_dev, ippdrv->dev);
 			if (ret) {
 				DRM_ERROR("failed to activate iommu\n");
-				goto err_iommu;
+				goto err;
 			}
 		}
 	}
 
 	return 0;
 
-err_iommu:
+err:
 	/* get ipp driver entry */
-	list_for_each_entry_reverse(ippdrv, &exynos_drm_ippdrv_list, drv_list)
+	list_for_each_entry_continue_reverse(ippdrv, &exynos_drm_ippdrv_list,
+						drv_list) {
 		if (is_drm_iommu_supported(drm_dev))
 			drm_iommu_detach_device(drm_dev, ippdrv->dev);
 
-err_idr:
-	idr_destroy(&ctx->ipp_idr);
-	idr_destroy(&ctx->prop_idr);
+		ipp_remove_id(&ctx->ipp_idr, &ctx->ipp_lock,
+				ippdrv->prop_list.ipp_id);
+	}
+
 	return ret;
 }
 
 static void ipp_subdrv_remove(struct drm_device *drm_dev, struct device *dev)
 {
 	struct exynos_drm_ippdrv *ippdrv;
+	struct ipp_context *ctx = get_ipp_context(dev);
 
 	/* get ipp driver entry */
 	list_for_each_entry(ippdrv, &exynos_drm_ippdrv_list, drv_list) {
 		if (is_drm_iommu_supported(drm_dev))
 			drm_iommu_detach_device(drm_dev, ippdrv->dev);
 
+		ipp_remove_id(&ctx->ipp_idr, &ctx->ipp_lock,
+				ippdrv->prop_list.ipp_id);
+
 		ippdrv->drm_dev = NULL;
 		exynos_drm_ippdrv_unregister(ippdrv);
 	}
@@ -1788,20 +1797,14 @@ static void ipp_subdrv_close(struct drm_device *drm_dev, struct device *dev,
 	struct drm_exynos_file_private *file_priv = file->driver_priv;
 	struct exynos_drm_ipp_private *priv = file_priv->ipp_priv;
 	struct exynos_drm_ippdrv *ippdrv = NULL;
+	struct ipp_context *ctx = get_ipp_context(dev);
 	struct drm_exynos_ipp_cmd_node *c_node, *tc_node;
 	int count = 0;
 
 	DRM_DEBUG_KMS("for priv[0x%x]\n", (int)priv);
 
-	if (list_empty(&exynos_drm_ippdrv_list)) {
-		DRM_DEBUG_KMS("ippdrv_list is empty.\n");
-		goto err_clear;
-	}
-
 	list_for_each_entry(ippdrv, &exynos_drm_ippdrv_list, drv_list) {
-		if (list_empty(&ippdrv->cmd_list))
-			continue;
-
+		mutex_lock(&ippdrv->cmd_lock);
 		list_for_each_entry_safe(c_node, tc_node,
 			&ippdrv->cmd_list, list) {
 			DRM_DEBUG_KMS("count[%d]ippdrv[0x%x]\n",
@@ -1821,14 +1824,14 @@ static void ipp_subdrv_close(struct drm_device *drm_dev, struct device *dev,
 				}
 
 				ippdrv->dedicated = false;
-				ipp_clean_cmd_node(c_node);
+				ipp_clean_cmd_node(ctx, c_node);
 				if (list_empty(&ippdrv->cmd_list))
 					pm_runtime_put_sync(ippdrv->dev);
 			}
 		}
+		mutex_unlock(&ippdrv->cmd_lock);
 	}
 
-err_clear:
 	kfree(priv);
 	return;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.h b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
index ab1634befc0..7aaeaae757c 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_ipp.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
@@ -52,7 +52,7 @@ struct drm_exynos_ipp_cmd_work {
  * @list: list head to command queue information.
  * @event_list: list head of event.
  * @mem_list: list head to source,destination memory queue information.
- * @cmd_lock: lock for synchronization of access to ioctl.
+ * @lock: lock for synchronization of access to ioctl.
  * @mem_lock: lock for synchronization of access to memory nodes.
  * @event_lock: lock for synchronization of access to scheduled event.
  * @start_complete: completion of start of command.
@@ -68,7 +68,7 @@ struct drm_exynos_ipp_cmd_node {
 	struct list_head	list;
 	struct list_head	event_list;
 	struct list_head	mem_list[EXYNOS_DRM_OPS_MAX];
-	struct mutex	cmd_lock;
+	struct mutex	lock;
 	struct mutex	mem_lock;
 	struct mutex	event_lock;
 	struct completion	start_complete;
@@ -83,7 +83,7 @@ struct drm_exynos_ipp_cmd_node {
 /*
  * A structure of buffer information.
  *
- * @gem_objs: Y, Cb, Cr each gem object.
+ * @handles: Y, Cb, Cr each gem object handle.
  * @base: Y, Cb, Cr each planar address.
  */
 struct drm_exynos_ipp_buf_info {
@@ -142,12 +142,12 @@ struct exynos_drm_ipp_ops {
  * @parent_dev: parent device information.
  * @dev: platform device.
  * @drm_dev: drm device.
- * @ipp_id: id of ipp driver.
  * @dedicated: dedicated ipp device.
  * @ops: source, destination operations.
  * @event_workq: event work queue.
  * @c_node: current command information.
  * @cmd_list: list head for command information.
+ * @cmd_lock: lock for synchronization of access to cmd_list.
  * @prop_list: property informations of current ipp driver.
  * @check_property: check property about format, size, buffer.
  * @reset: reset ipp block.
@@ -160,13 +160,13 @@ struct exynos_drm_ippdrv {
 	struct device	*parent_dev;
 	struct device	*dev;
 	struct drm_device	*drm_dev;
-	u32	ipp_id;
 	bool	dedicated;
 	struct exynos_drm_ipp_ops	*ops[EXYNOS_DRM_OPS_MAX];
 	struct workqueue_struct	*event_workq;
 	struct drm_exynos_ipp_cmd_node *c_node;
 	struct list_head	cmd_list;
-	struct drm_exynos_ipp_prop_list *prop_list;
+	struct mutex	cmd_lock;
+	struct drm_exynos_ipp_prop_list prop_list;
 
 	int (*check_property)(struct device *dev,
 		struct drm_exynos_ipp_property *property);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_plane.c b/drivers/gpu/drm/exynos/exynos_drm_plane.c
index fcb0652e77d..8371cbd7631 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_plane.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_plane.c
@@ -13,7 +13,7 @@
 
 #include <drm/exynos_drm.h>
 #include "exynos_drm_drv.h"
-#include "exynos_drm_encoder.h"
+#include "exynos_drm_crtc.h"
 #include "exynos_drm_fb.h"
 #include "exynos_drm_gem.h"
 #include "exynos_drm_plane.h"
@@ -87,7 +87,7 @@ int exynos_plane_mode_set(struct drm_plane *plane, struct drm_crtc *crtc,
 		struct exynos_drm_gem_buf *buffer = exynos_drm_fb_buffer(fb, i);
 
 		if (!buffer) {
-			DRM_LOG_KMS("buffer is null\n");
+			DRM_DEBUG_KMS("buffer is null\n");
 			return -EFAULT;
 		}
 
@@ -139,7 +139,7 @@ int exynos_plane_mode_set(struct drm_plane *plane, struct drm_crtc *crtc,
 			overlay->crtc_x, overlay->crtc_y,
 			overlay->crtc_width, overlay->crtc_height);
 
-	exynos_drm_fn_encoder(crtc, overlay, exynos_drm_encoder_plane_mode_set);
+	exynos_drm_crtc_plane_mode_set(crtc, overlay);
 
 	return 0;
 }
@@ -149,8 +149,7 @@ void exynos_plane_commit(struct drm_plane *plane)
 	struct exynos_plane *exynos_plane = to_exynos_plane(plane);
 	struct exynos_drm_overlay *overlay = &exynos_plane->overlay;
 
-	exynos_drm_fn_encoder(plane->crtc, &overlay->zpos,
-			exynos_drm_encoder_plane_commit);
+	exynos_drm_crtc_plane_commit(plane->crtc, overlay->zpos);
 }
 
 void exynos_plane_dpms(struct drm_plane *plane, int mode)
@@ -162,17 +161,13 @@ void exynos_plane_dpms(struct drm_plane *plane, int mode)
 		if (exynos_plane->enabled)
 			return;
 
-		exynos_drm_fn_encoder(plane->crtc, &overlay->zpos,
-				exynos_drm_encoder_plane_enable);
-
+		exynos_drm_crtc_plane_enable(plane->crtc, overlay->zpos);
 		exynos_plane->enabled = true;
 	} else {
 		if (!exynos_plane->enabled)
 			return;
 
-		exynos_drm_fn_encoder(plane->crtc, &overlay->zpos,
-				exynos_drm_encoder_plane_disable);
-
+		exynos_drm_crtc_plane_disable(plane->crtc, overlay->zpos);
 		exynos_plane->enabled = false;
 	}
 }
@@ -259,7 +254,7 @@ static void exynos_plane_attach_zpos_property(struct drm_plane *plane)
 }
 
 struct drm_plane *exynos_plane_init(struct drm_device *dev,
-				    unsigned int possible_crtcs, bool priv)
+				    unsigned long possible_crtcs, bool priv)
 {
 	struct exynos_plane *exynos_plane;
 	int err;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_plane.h b/drivers/gpu/drm/exynos/exynos_drm_plane.h
index 88312458580..84d464c90d3 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_plane.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_plane.h
@@ -17,4 +17,4 @@ int exynos_plane_mode_set(struct drm_plane *plane, struct drm_crtc *crtc,
 void exynos_plane_commit(struct drm_plane *plane);
 void exynos_plane_dpms(struct drm_plane *plane, int mode);
 struct drm_plane *exynos_plane_init(struct drm_device *dev,
-				    unsigned int possible_crtcs, bool priv);
+				    unsigned long possible_crtcs, bool priv);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_rotator.c b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
index 7b901688def..f01fbb6dc1f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_rotator.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
@@ -158,8 +158,9 @@ static irqreturn_t rotator_irq_handler(int irq, void *arg)
 			rot->cur_buf_id[EXYNOS_DRM_OPS_DST];
 		queue_work(ippdrv->event_workq,
 			(struct work_struct *)event_work);
-	} else
+	} else {
 		DRM_ERROR("the SFR is set illegally\n");
+	}
 
 	return IRQ_HANDLED;
 }
@@ -469,11 +470,7 @@ static struct exynos_drm_ipp_ops rot_dst_ops = {
 
 static int rotator_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
 {
-	struct drm_exynos_ipp_prop_list *prop_list;
-
-	prop_list = devm_kzalloc(ippdrv->dev, sizeof(*prop_list), GFP_KERNEL);
-	if (!prop_list)
-		return -ENOMEM;
+	struct drm_exynos_ipp_prop_list *prop_list = &ippdrv->prop_list;
 
 	prop_list->version = 1;
 	prop_list->flip = (1 << EXYNOS_DRM_FLIP_VERTICAL) |
@@ -486,8 +483,6 @@ static int rotator_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
 	prop_list->crop = 0;
 	prop_list->scale = 0;
 
-	ippdrv->prop_list = prop_list;
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_vidi.c b/drivers/gpu/drm/exynos/exynos_drm_vidi.c
index ddaaedde173..2fb8705d646 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_vidi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_vidi.c
@@ -28,7 +28,9 @@
 /* vidi has totally three virtual windows. */
 #define WINDOWS_NR		3
 
-#define get_vidi_context(dev)	platform_get_drvdata(to_platform_device(dev))
+#define get_vidi_mgr(dev)	platform_get_drvdata(to_platform_device(dev))
+#define ctx_from_connector(c)	container_of(c, struct vidi_context, \
+					connector)
 
 struct vidi_win_data {
 	unsigned int		offset_x;
@@ -45,8 +47,11 @@ struct vidi_win_data {
 };
 
 struct vidi_context {
-	struct exynos_drm_subdrv	subdrv;
+	struct drm_device		*drm_dev;
 	struct drm_crtc			*crtc;
+	struct drm_encoder		*encoder;
+	struct drm_connector		connector;
+	struct exynos_drm_subdrv	subdrv;
 	struct vidi_win_data		win_data[WINDOWS_NR];
 	struct edid			*raw_edid;
 	unsigned int			clkdiv;
@@ -58,6 +63,7 @@ struct vidi_context {
 	bool				direct_vblank;
 	struct work_struct		work;
 	struct mutex			lock;
+	int				pipe;
 };
 
 static const char fake_edid_info[] = {
@@ -85,126 +91,34 @@ static const char fake_edid_info[] = {
 	0x00, 0x00, 0x00, 0x06
 };
 
-static bool vidi_display_is_connected(struct device *dev)
-{
-	struct vidi_context *ctx = get_vidi_context(dev);
-
-	/*
-	 * connection request would come from user side
-	 * to do hotplug through specific ioctl.
-	 */
-	return ctx->connected ? true : false;
-}
-
-static struct edid *vidi_get_edid(struct device *dev,
-			struct drm_connector *connector)
-{
-	struct vidi_context *ctx = get_vidi_context(dev);
-	struct edid *edid;
-
-	/*
-	 * the edid data comes from user side and it would be set
-	 * to ctx->raw_edid through specific ioctl.
-	 */
-	if (!ctx->raw_edid) {
-		DRM_DEBUG_KMS("raw_edid is null.\n");
-		return ERR_PTR(-EFAULT);
-	}
-
-	edid = drm_edid_duplicate(ctx->raw_edid);
-	if (!edid) {
-		DRM_DEBUG_KMS("failed to allocate edid\n");
-		return ERR_PTR(-ENOMEM);
-	}
-
-	return edid;
-}
-
-static void *vidi_get_panel(struct device *dev)
-{
-	/* TODO. */
-
-	return NULL;
-}
-
-static int vidi_check_mode(struct device *dev, struct drm_display_mode *mode)
+static void vidi_apply(struct exynos_drm_manager *mgr)
 {
-	/* TODO. */
-
-	return 0;
-}
-
-static int vidi_display_power_on(struct device *dev, int mode)
-{
-	/* TODO */
-
-	return 0;
-}
-
-static struct exynos_drm_display_ops vidi_display_ops = {
-	.type = EXYNOS_DISPLAY_TYPE_VIDI,
-	.is_connected = vidi_display_is_connected,
-	.get_edid = vidi_get_edid,
-	.get_panel = vidi_get_panel,
-	.check_mode = vidi_check_mode,
-	.power_on = vidi_display_power_on,
-};
-
-static void vidi_dpms(struct device *subdrv_dev, int mode)
-{
-	struct vidi_context *ctx = get_vidi_context(subdrv_dev);
-
-	DRM_DEBUG_KMS("%d\n", mode);
-
-	mutex_lock(&ctx->lock);
-
-	switch (mode) {
-	case DRM_MODE_DPMS_ON:
-		/* TODO. */
-		break;
-	case DRM_MODE_DPMS_STANDBY:
-	case DRM_MODE_DPMS_SUSPEND:
-	case DRM_MODE_DPMS_OFF:
-		/* TODO. */
-		break;
-	default:
-		DRM_DEBUG_KMS("unspecified mode %d\n", mode);
-		break;
-	}
-
-	mutex_unlock(&ctx->lock);
-}
-
-static void vidi_apply(struct device *subdrv_dev)
-{
-	struct vidi_context *ctx = get_vidi_context(subdrv_dev);
-	struct exynos_drm_manager *mgr = ctx->subdrv.manager;
+	struct vidi_context *ctx = mgr->ctx;
 	struct exynos_drm_manager_ops *mgr_ops = mgr->ops;
-	struct exynos_drm_overlay_ops *ovl_ops = mgr->overlay_ops;
 	struct vidi_win_data *win_data;
 	int i;
 
 	for (i = 0; i < WINDOWS_NR; i++) {
 		win_data = &ctx->win_data[i];
-		if (win_data->enabled && (ovl_ops && ovl_ops->commit))
-			ovl_ops->commit(subdrv_dev, i);
+		if (win_data->enabled && (mgr_ops && mgr_ops->win_commit))
+			mgr_ops->win_commit(mgr, i);
 	}
 
 	if (mgr_ops && mgr_ops->commit)
-		mgr_ops->commit(subdrv_dev);
+		mgr_ops->commit(mgr);
 }
 
-static void vidi_commit(struct device *dev)
+static void vidi_commit(struct exynos_drm_manager *mgr)
 {
-	struct vidi_context *ctx = get_vidi_context(dev);
+	struct vidi_context *ctx = mgr->ctx;
 
 	if (ctx->suspended)
 		return;
 }
 
-static int vidi_enable_vblank(struct device *dev)
+static int vidi_enable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct vidi_context *ctx = get_vidi_context(dev);
+	struct vidi_context *ctx = mgr->ctx;
 
 	if (ctx->suspended)
 		return -EPERM;
@@ -217,16 +131,16 @@ static int vidi_enable_vblank(struct device *dev)
 	/*
 	 * in case of page flip request, vidi_finish_pageflip function
 	 * will not be called because direct_vblank is true and then
-	 * that function will be called by overlay_ops->commit callback
+	 * that function will be called by manager_ops->win_commit callback
 	 */
 	schedule_work(&ctx->work);
 
 	return 0;
 }
 
-static void vidi_disable_vblank(struct device *dev)
+static void vidi_disable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct vidi_context *ctx = get_vidi_context(dev);
+	struct vidi_context *ctx = mgr->ctx;
 
 	if (ctx->suspended)
 		return;
@@ -235,24 +149,16 @@ static void vidi_disable_vblank(struct device *dev)
 		ctx->vblank_on = false;
 }
 
-static struct exynos_drm_manager_ops vidi_manager_ops = {
-	.dpms = vidi_dpms,
-	.apply = vidi_apply,
-	.commit = vidi_commit,
-	.enable_vblank = vidi_enable_vblank,
-	.disable_vblank = vidi_disable_vblank,
-};
-
-static void vidi_win_mode_set(struct device *dev,
-			      struct exynos_drm_overlay *overlay)
+static void vidi_win_mode_set(struct exynos_drm_manager *mgr,
+			struct exynos_drm_overlay *overlay)
 {
-	struct vidi_context *ctx = get_vidi_context(dev);
+	struct vidi_context *ctx = mgr->ctx;
 	struct vidi_win_data *win_data;
 	int win;
 	unsigned long offset;
 
 	if (!overlay) {
-		dev_err(dev, "overlay is NULL\n");
+		DRM_ERROR("overlay is NULL\n");
 		return;
 	}
 
@@ -296,9 +202,9 @@ static void vidi_win_mode_set(struct device *dev,
 			overlay->fb_width, overlay->crtc_width);
 }
 
-static void vidi_win_commit(struct device *dev, int zpos)
+static void vidi_win_commit(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct vidi_context *ctx = get_vidi_context(dev);
+	struct vidi_context *ctx = mgr->ctx;
 	struct vidi_win_data *win_data;
 	int win = zpos;
 
@@ -315,15 +221,15 @@ static void vidi_win_commit(struct device *dev, int zpos)
 
 	win_data->enabled = true;
 
-	DRM_DEBUG_KMS("dma_addr = 0x%x\n", win_data->dma_addr);
+	DRM_DEBUG_KMS("dma_addr = %pad\n", &win_data->dma_addr);
 
 	if (ctx->vblank_on)
 		schedule_work(&ctx->work);
 }
 
-static void vidi_win_disable(struct device *dev, int zpos)
+static void vidi_win_disable(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct vidi_context *ctx = get_vidi_context(dev);
+	struct vidi_context *ctx = mgr->ctx;
 	struct vidi_win_data *win_data;
 	int win = zpos;
 
@@ -339,98 +245,130 @@ static void vidi_win_disable(struct device *dev, int zpos)
 	/* TODO. */
 }
 
-static struct exynos_drm_overlay_ops vidi_overlay_ops = {
-	.mode_set = vidi_win_mode_set,
-	.commit = vidi_win_commit,
-	.disable = vidi_win_disable,
-};
+static int vidi_power_on(struct exynos_drm_manager *mgr, bool enable)
+{
+	struct vidi_context *ctx = mgr->ctx;
 
-static struct exynos_drm_manager vidi_manager = {
-	.pipe		= -1,
-	.ops		= &vidi_manager_ops,
-	.overlay_ops	= &vidi_overlay_ops,
-	.display_ops	= &vidi_display_ops,
-};
+	DRM_DEBUG_KMS("%s\n", __FILE__);
 
-static void vidi_fake_vblank_handler(struct work_struct *work)
-{
-	struct vidi_context *ctx = container_of(work, struct vidi_context,
-					work);
-	struct exynos_drm_subdrv *subdrv = &ctx->subdrv;
-	struct exynos_drm_manager *manager = subdrv->manager;
+	if (enable != false && enable != true)
+		return -EINVAL;
 
-	if (manager->pipe < 0)
-		return;
+	if (enable) {
+		ctx->suspended = false;
 
-	/* refresh rate is about 50Hz. */
-	usleep_range(16000, 20000);
+		/* if vblank was enabled status, enable it again. */
+		if (test_and_clear_bit(0, &ctx->irq_flags))
+			vidi_enable_vblank(mgr);
+
+		vidi_apply(mgr);
+	} else {
+		ctx->suspended = true;
+	}
+
+	return 0;
+}
+
+static void vidi_dpms(struct exynos_drm_manager *mgr, int mode)
+{
+	struct vidi_context *ctx = mgr->ctx;
+
+	DRM_DEBUG_KMS("%d\n", mode);
 
 	mutex_lock(&ctx->lock);
 
-	if (ctx->direct_vblank) {
-		drm_handle_vblank(subdrv->drm_dev, manager->pipe);
-		ctx->direct_vblank = false;
-		mutex_unlock(&ctx->lock);
-		return;
+	switch (mode) {
+	case DRM_MODE_DPMS_ON:
+		vidi_power_on(mgr, true);
+		break;
+	case DRM_MODE_DPMS_STANDBY:
+	case DRM_MODE_DPMS_SUSPEND:
+	case DRM_MODE_DPMS_OFF:
+		vidi_power_on(mgr, false);
+		break;
+	default:
+		DRM_DEBUG_KMS("unspecified mode %d\n", mode);
+		break;
 	}
 
 	mutex_unlock(&ctx->lock);
-
-	exynos_drm_crtc_finish_pageflip(subdrv->drm_dev, manager->pipe);
 }
 
-static int vidi_subdrv_probe(struct drm_device *drm_dev, struct device *dev)
+static int vidi_mgr_initialize(struct exynos_drm_manager *mgr,
+			struct drm_device *drm_dev)
 {
+	struct vidi_context *ctx = mgr->ctx;
+	struct exynos_drm_private *priv = drm_dev->dev_private;
+
+	mgr->drm_dev = ctx->drm_dev = drm_dev;
+	mgr->pipe = ctx->pipe = priv->pipe++;
+
 	/*
 	 * enable drm irq mode.
-	 * - with irq_enabled = true, we can use the vblank feature.
+	 * - with irq_enabled = 1, we can use the vblank feature.
 	 *
 	 * P.S. note that we wouldn't use drm irq handler but
 	 *	just specific driver own one instead because
 	 *	drm framework supports only one irq handler.
 	 */
-	drm_dev->irq_enabled = true;
+	drm_dev->irq_enabled = 1;
 
 	/*
-	 * with vblank_disable_allowed = true, vblank interrupt will be disabled
+	 * with vblank_disable_allowed = 1, vblank interrupt will be disabled
 	 * by drm timer once a current process gives up ownership of
 	 * vblank event.(after drm_vblank_put function is called)
 	 */
-	drm_dev->vblank_disable_allowed = true;
+	drm_dev->vblank_disable_allowed = 1;
 
 	return 0;
 }
 
-static void vidi_subdrv_remove(struct drm_device *drm_dev, struct device *dev)
-{
-	/* TODO. */
-}
+static struct exynos_drm_manager_ops vidi_manager_ops = {
+	.dpms = vidi_dpms,
+	.commit = vidi_commit,
+	.enable_vblank = vidi_enable_vblank,
+	.disable_vblank = vidi_disable_vblank,
+	.win_mode_set = vidi_win_mode_set,
+	.win_commit = vidi_win_commit,
+	.win_disable = vidi_win_disable,
+};
+
+static struct exynos_drm_manager vidi_manager = {
+	.type = EXYNOS_DISPLAY_TYPE_VIDI,
+	.ops = &vidi_manager_ops,
+};
 
-static int vidi_power_on(struct vidi_context *ctx, bool enable)
+static void vidi_fake_vblank_handler(struct work_struct *work)
 {
-	struct exynos_drm_subdrv *subdrv = &ctx->subdrv;
-	struct device *dev = subdrv->dev;
+	struct vidi_context *ctx = container_of(work, struct vidi_context,
+					work);
 
-	if (enable) {
-		ctx->suspended = false;
+	if (ctx->pipe < 0)
+		return;
 
-		/* if vblank was enabled status, enable it again. */
-		if (test_and_clear_bit(0, &ctx->irq_flags))
-			vidi_enable_vblank(dev);
+	/* refresh rate is about 50Hz. */
+	usleep_range(16000, 20000);
 
-		vidi_apply(dev);
-	} else {
-		ctx->suspended = true;
+	mutex_lock(&ctx->lock);
+
+	if (ctx->direct_vblank) {
+		drm_handle_vblank(ctx->drm_dev, ctx->pipe);
+		ctx->direct_vblank = false;
+		mutex_unlock(&ctx->lock);
+		return;
 	}
 
-	return 0;
+	mutex_unlock(&ctx->lock);
+
+	exynos_drm_crtc_finish_pageflip(ctx->drm_dev, ctx->pipe);
 }
 
 static int vidi_show_connection(struct device *dev,
 				struct device_attribute *attr, char *buf)
 {
 	int rc;
-	struct vidi_context *ctx = get_vidi_context(dev);
+	struct exynos_drm_manager *mgr = get_vidi_mgr(dev);
+	struct vidi_context *ctx = mgr->ctx;
 
 	mutex_lock(&ctx->lock);
 
@@ -445,7 +383,8 @@ static int vidi_store_connection(struct device *dev,
 				struct device_attribute *attr,
 				const char *buf, size_t len)
 {
-	struct vidi_context *ctx = get_vidi_context(dev);
+	struct exynos_drm_manager *mgr = get_vidi_mgr(dev);
+	struct vidi_context *ctx = mgr->ctx;
 	int ret;
 
 	ret = kstrtoint(buf, 0, &ctx->connected);
@@ -467,7 +406,7 @@ static int vidi_store_connection(struct device *dev,
 
 	DRM_DEBUG_KMS("requested connection.\n");
 
-	drm_helper_hpd_irq_event(ctx->subdrv.drm_dev);
+	drm_helper_hpd_irq_event(ctx->drm_dev);
 
 	return len;
 }
@@ -480,8 +419,7 @@ int vidi_connection_ioctl(struct drm_device *drm_dev, void *data,
 {
 	struct vidi_context *ctx = NULL;
 	struct drm_encoder *encoder;
-	struct exynos_drm_manager *manager;
-	struct exynos_drm_display_ops *display_ops;
+	struct exynos_drm_display *display;
 	struct drm_exynos_vidi_connection *vidi = data;
 
 	if (!vidi) {
@@ -496,11 +434,10 @@ int vidi_connection_ioctl(struct drm_device *drm_dev, void *data,
 
 	list_for_each_entry(encoder, &drm_dev->mode_config.encoder_list,
 								head) {
-		manager = exynos_drm_get_manager(encoder);
-		display_ops = manager->display_ops;
+		display = exynos_drm_get_display(encoder);
 
-		if (display_ops->type == EXYNOS_DISPLAY_TYPE_VIDI) {
-			ctx = get_vidi_context(manager->dev);
+		if (display->type == EXYNOS_DISPLAY_TYPE_VIDI) {
+			ctx = display->ctx;
 			break;
 		}
 	}
@@ -539,19 +476,140 @@ int vidi_connection_ioctl(struct drm_device *drm_dev, void *data,
 	}
 
 	ctx->connected = vidi->connection;
-	drm_helper_hpd_irq_event(ctx->subdrv.drm_dev);
+	drm_helper_hpd_irq_event(ctx->drm_dev);
+
+	return 0;
+}
+
+static enum drm_connector_status vidi_detect(struct drm_connector *connector,
+			bool force)
+{
+	struct vidi_context *ctx = ctx_from_connector(connector);
+
+	/*
+	 * connection request would come from user side
+	 * to do hotplug through specific ioctl.
+	 */
+	return ctx->connected ? connector_status_connected :
+			connector_status_disconnected;
+}
+
+static void vidi_connector_destroy(struct drm_connector *connector)
+{
+}
+
+static struct drm_connector_funcs vidi_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = vidi_detect,
+	.destroy = vidi_connector_destroy,
+};
+
+static int vidi_get_modes(struct drm_connector *connector)
+{
+	struct vidi_context *ctx = ctx_from_connector(connector);
+	struct edid *edid;
+	int edid_len;
+
+	/*
+	 * the edid data comes from user side and it would be set
+	 * to ctx->raw_edid through specific ioctl.
+	 */
+	if (!ctx->raw_edid) {
+		DRM_DEBUG_KMS("raw_edid is null.\n");
+		return -EFAULT;
+	}
+
+	edid_len = (1 + ctx->raw_edid->extensions) * EDID_LENGTH;
+	edid = kmemdup(ctx->raw_edid, edid_len, GFP_KERNEL);
+	if (!edid) {
+		DRM_DEBUG_KMS("failed to allocate edid\n");
+		return -ENOMEM;
+	}
+
+	drm_mode_connector_update_edid_property(connector, edid);
+
+	return drm_add_edid_modes(connector, edid);
+}
+
+static struct drm_encoder *vidi_best_encoder(struct drm_connector *connector)
+{
+	struct vidi_context *ctx = ctx_from_connector(connector);
+
+	return ctx->encoder;
+}
+
+static struct drm_connector_helper_funcs vidi_connector_helper_funcs = {
+	.get_modes = vidi_get_modes,
+	.best_encoder = vidi_best_encoder,
+};
+
+static int vidi_create_connector(struct exynos_drm_display *display,
+				struct drm_encoder *encoder)
+{
+	struct vidi_context *ctx = display->ctx;
+	struct drm_connector *connector = &ctx->connector;
+	int ret;
+
+	ctx->encoder = encoder;
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	ret = drm_connector_init(ctx->drm_dev, connector,
+			&vidi_connector_funcs, DRM_MODE_CONNECTOR_VIRTUAL);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
+		return ret;
+	}
+
+	drm_connector_helper_add(connector, &vidi_connector_helper_funcs);
+	drm_sysfs_connector_add(connector);
+	drm_mode_connector_attach_encoder(connector, encoder);
+
+	return 0;
+}
+
+
+static struct exynos_drm_display_ops vidi_display_ops = {
+	.create_connector = vidi_create_connector,
+};
+
+static struct exynos_drm_display vidi_display = {
+	.type = EXYNOS_DISPLAY_TYPE_VIDI,
+	.ops = &vidi_display_ops,
+};
+
+static int vidi_subdrv_probe(struct drm_device *drm_dev, struct device *dev)
+{
+	struct exynos_drm_manager *mgr = get_vidi_mgr(dev);
+	struct vidi_context *ctx = mgr->ctx;
+	struct drm_crtc *crtc = ctx->crtc;
+	int ret;
+
+	vidi_mgr_initialize(mgr, drm_dev);
+
+	ret = exynos_drm_crtc_create(&vidi_manager);
+	if (ret) {
+		DRM_ERROR("failed to create crtc.\n");
+		return ret;
+	}
+
+	ret = exynos_drm_create_enc_conn(drm_dev, &vidi_display);
+	if (ret) {
+		crtc->funcs->destroy(crtc);
+		DRM_ERROR("failed to create encoder and connector.\n");
+		return ret;
+	}
 
 	return 0;
 }
 
 static int vidi_probe(struct platform_device *pdev)
 {
-	struct device *dev = &pdev->dev;
-	struct vidi_context *ctx;
 	struct exynos_drm_subdrv *subdrv;
+	struct vidi_context *ctx;
 	int ret;
 
-	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
+	ctx = devm_kzalloc(&pdev->dev, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)
 		return -ENOMEM;
 
@@ -559,58 +617,52 @@ static int vidi_probe(struct platform_device *pdev)
 
 	INIT_WORK(&ctx->work, vidi_fake_vblank_handler);
 
-	subdrv = &ctx->subdrv;
-	subdrv->dev = dev;
-	subdrv->manager = &vidi_manager;
-	subdrv->probe = vidi_subdrv_probe;
-	subdrv->remove = vidi_subdrv_remove;
+	vidi_manager.ctx = ctx;
+	vidi_display.ctx = ctx;
 
 	mutex_init(&ctx->lock);
 
-	platform_set_drvdata(pdev, ctx);
+	platform_set_drvdata(pdev, &vidi_manager);
 
-	ret = device_create_file(dev, &dev_attr_connection);
-	if (ret < 0)
-		DRM_INFO("failed to create connection sysfs.\n");
+	subdrv = &ctx->subdrv;
+	subdrv->dev = &pdev->dev;
+	subdrv->probe = vidi_subdrv_probe;
 
-	exynos_drm_subdrv_register(subdrv);
+	ret = exynos_drm_subdrv_register(subdrv);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to register drm vidi device\n");
+		return ret;
+	}
+
+	ret = device_create_file(&pdev->dev, &dev_attr_connection);
+	if (ret < 0) {
+		exynos_drm_subdrv_unregister(subdrv);
+		DRM_INFO("failed to create connection sysfs.\n");
+	}
 
 	return 0;
 }
 
 static int vidi_remove(struct platform_device *pdev)
 {
-	struct vidi_context *ctx = platform_get_drvdata(pdev);
-
-	exynos_drm_subdrv_unregister(&ctx->subdrv);
+	struct exynos_drm_manager *mgr = platform_get_drvdata(pdev);
+	struct vidi_context *ctx = mgr->ctx;
+	struct drm_encoder *encoder = ctx->encoder;
+	struct drm_crtc *crtc = mgr->crtc;
 
 	if (ctx->raw_edid != (struct edid *)fake_edid_info) {
 		kfree(ctx->raw_edid);
 		ctx->raw_edid = NULL;
-	}
 
-	return 0;
-}
-
-#ifdef CONFIG_PM_SLEEP
-static int vidi_suspend(struct device *dev)
-{
-	struct vidi_context *ctx = get_vidi_context(dev);
-
-	return vidi_power_on(ctx, false);
-}
+		return -EINVAL;
+	}
 
-static int vidi_resume(struct device *dev)
-{
-	struct vidi_context *ctx = get_vidi_context(dev);
+	crtc->funcs->destroy(crtc);
+	encoder->funcs->destroy(encoder);
+	drm_connector_cleanup(&ctx->connector);
 
-	return vidi_power_on(ctx, true);
+	return 0;
 }
-#endif
-
-static const struct dev_pm_ops vidi_pm_ops = {
-	SET_SYSTEM_SLEEP_PM_OPS(vidi_suspend, vidi_resume)
-};
 
 struct platform_driver vidi_driver = {
 	.probe		= vidi_probe,
@@ -618,6 +670,33 @@ struct platform_driver vidi_driver = {
 	.driver		= {
 		.name	= "exynos-drm-vidi",
 		.owner	= THIS_MODULE,
-		.pm	= &vidi_pm_ops,
 	},
 };
+
+int exynos_drm_probe_vidi(void)
+{
+	struct platform_device *pdev;
+	int ret;
+
+	pdev = platform_device_register_simple("exynos-drm-vidi", -1, NULL, 0);
+	if (IS_ERR(pdev))
+		return PTR_ERR(pdev);
+
+	ret = platform_driver_register(&vidi_driver);
+	if (ret) {
+		platform_device_unregister(pdev);
+		return ret;
+	}
+
+	return ret;
+}
+
+void exynos_drm_remove_vidi(void)
+{
+	struct vidi_context *ctx = vidi_manager.ctx;
+	struct exynos_drm_subdrv *subdrv = &ctx->subdrv;
+	struct platform_device *pdev = to_platform_device(subdrv->dev);
+
+	platform_driver_unregister(&vidi_driver);
+	platform_device_unregister(pdev);
+}
diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.c b/drivers/gpu/drm/exynos/exynos_hdmi.c
index a0e10aeb0e6..aa259b0a873 100644
--- a/drivers/gpu/drm/exynos/exynos_hdmi.c
+++ b/drivers/gpu/drm/exynos/exynos_hdmi.c
@@ -33,56 +33,55 @@
 #include <linux/regulator/consumer.h>
 #include <linux/io.h>
 #include <linux/of.h>
+#include <linux/of_address.h>
 #include <linux/of_gpio.h>
+#include <linux/hdmi.h>
+#include <linux/component.h>
+#include <linux/mfd/syscon.h>
+#include <linux/regmap.h>
 
 #include <drm/exynos_drm.h>
 
 #include "exynos_drm_drv.h"
-#include "exynos_drm_hdmi.h"
-
-#include "exynos_hdmi.h"
+#include "exynos_drm_crtc.h"
+#include "exynos_mixer.h"
 
 #include <linux/gpio.h>
 #include <media/s5p_hdmi.h>
 
-#define MAX_WIDTH		1920
-#define MAX_HEIGHT		1080
-#define get_hdmi_context(dev)	platform_get_drvdata(to_platform_device(dev))
+#define get_hdmi_display(dev)	platform_get_drvdata(to_platform_device(dev))
+#define ctx_from_connector(c)	container_of(c, struct hdmi_context, connector)
+
+#define HOTPLUG_DEBOUNCE_MS		1100
 
 /* AVI header and aspect ratio */
 #define HDMI_AVI_VERSION		0x02
 #define HDMI_AVI_LENGTH		0x0D
-#define AVI_PIC_ASPECT_RATIO_16_9	(2 << 4)
-#define AVI_SAME_AS_PIC_ASPECT_RATIO	8
 
 /* AUI header info */
 #define HDMI_AUI_VERSION	0x01
 #define HDMI_AUI_LENGTH	0x0A
-
-/* HDMI infoframe to configure HDMI out packet header, AUI and AVI */
-enum HDMI_PACKET_TYPE {
-	/* refer to Table 5-8 Packet Type in HDMI specification v1.4a */
-	/* InfoFrame packet type */
-	HDMI_PACKET_TYPE_INFOFRAME = 0x80,
-	/* Vendor-Specific InfoFrame */
-	HDMI_PACKET_TYPE_VSI = HDMI_PACKET_TYPE_INFOFRAME + 1,
-	/* Auxiliary Video information InfoFrame */
-	HDMI_PACKET_TYPE_AVI = HDMI_PACKET_TYPE_INFOFRAME + 2,
-	/* Audio information InfoFrame */
-	HDMI_PACKET_TYPE_AUI = HDMI_PACKET_TYPE_INFOFRAME + 4
-};
+#define AVI_SAME_AS_PIC_ASPECT_RATIO 0x8
+#define AVI_4_3_CENTER_RATIO	0x9
+#define AVI_16_9_CENTER_RATIO	0xa
 
 enum hdmi_type {
 	HDMI_TYPE13,
 	HDMI_TYPE14,
 };
 
+struct hdmi_driver_data {
+	unsigned int type;
+	const struct hdmiphy_config *phy_confs;
+	unsigned int phy_conf_count;
+	unsigned int is_apb_phy:1;
+};
+
 struct hdmi_resources {
 	struct clk			*hdmi;
 	struct clk			*sclk_hdmi;
 	struct clk			*sclk_pixel;
 	struct clk			*sclk_hdmiphy;
-	struct clk			*hdmiphy;
 	struct clk			*mout_hdmi;
 	struct regulator_bulk_data	*regul_bulk;
 	int				regul_count;
@@ -174,6 +173,7 @@ struct hdmi_v14_conf {
 struct hdmi_conf_regs {
 	int pixel_clock;
 	int cea_video_id;
+	enum hdmi_picture_aspect aspect_ratio;
 	union {
 		struct hdmi_v13_conf v13_conf;
 		struct hdmi_v14_conf v14_conf;
@@ -183,25 +183,32 @@ struct hdmi_conf_regs {
 struct hdmi_context {
 	struct device			*dev;
 	struct drm_device		*drm_dev;
+	struct drm_connector		connector;
+	struct drm_encoder		*encoder;
 	bool				hpd;
 	bool				powered;
 	bool				dvi_mode;
 	struct mutex			hdmi_mutex;
 
 	void __iomem			*regs;
-	void				*parent_ctx;
 	int				irq;
+	struct delayed_work		hotplug_work;
 
-	struct i2c_client		*ddc_port;
+	struct i2c_adapter		*ddc_adpt;
 	struct i2c_client		*hdmiphy_port;
 
 	/* current hdmiphy conf regs */
+	struct drm_display_mode		current_mode;
 	struct hdmi_conf_regs		mode_conf;
 
 	struct hdmi_resources		res;
 
 	int				hpd_gpio;
+	void __iomem			*regs_hdmiphy;
+	const struct hdmiphy_config		*phy_confs;
+	unsigned int			phy_conf_count;
 
+	struct regmap			*pmureg;
 	enum hdmi_type			type;
 };
 
@@ -315,6 +322,24 @@ static const struct hdmiphy_config hdmiphy_v14_configs[] = {
 		},
 	},
 	{
+		.pixel_clock = 71000000,
+		.conf = {
+			0x01, 0xd1, 0x3b, 0x35, 0x40, 0x0c, 0x04, 0x08,
+			0x85, 0xa0, 0x63, 0xd9, 0x45, 0xa0, 0xac, 0x80,
+			0x08, 0x80, 0x11, 0x04, 0x02, 0x22, 0x44, 0x86,
+			0x54, 0xad, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 73250000,
+		.conf = {
+			0x01, 0xd1, 0x3d, 0x35, 0x40, 0x18, 0x02, 0x08,
+			0x83, 0xa0, 0x6e, 0xd9, 0x45, 0xa0, 0xac, 0x80,
+			0x08, 0x80, 0x11, 0x04, 0x02, 0x22, 0x44, 0x86,
+			0x54, 0xa8, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
 		.pixel_clock = 74176000,
 		.conf = {
 			0x01, 0xd1, 0x3e, 0x35, 0x40, 0x5b, 0xde, 0x08,
@@ -360,6 +385,24 @@ static const struct hdmiphy_config hdmiphy_v14_configs[] = {
 		},
 	},
 	{
+		.pixel_clock = 115500000,
+		.conf = {
+			0x01, 0xd1, 0x30, 0x12, 0x40, 0x40, 0x10, 0x08,
+			0x80, 0x80, 0x21, 0xd9, 0x45, 0xa0, 0xac, 0x80,
+			0x08, 0x80, 0x11, 0x04, 0x02, 0x22, 0x44, 0x86,
+			0x54, 0xaa, 0x25, 0x03, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 119000000,
+		.conf = {
+			0x01, 0xd1, 0x32, 0x1a, 0x40, 0x30, 0xd8, 0x08,
+			0x04, 0xa0, 0x2a, 0xd9, 0x45, 0xa0, 0xac, 0x80,
+			0x08, 0x80, 0x11, 0x04, 0x02, 0x22, 0x44, 0x86,
+			0x54, 0x9d, 0x25, 0x03, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
 		.pixel_clock = 146250000,
 		.conf = {
 			0x01, 0xd1, 0x3d, 0x15, 0x40, 0x18, 0xfd, 0x08,
@@ -379,10 +422,181 @@ static const struct hdmiphy_config hdmiphy_v14_configs[] = {
 	},
 };
 
-struct hdmi_infoframe {
-	enum HDMI_PACKET_TYPE type;
-	u8 ver;
-	u8 len;
+static const struct hdmiphy_config hdmiphy_5420_configs[] = {
+	{
+		.pixel_clock = 25200000,
+		.conf = {
+			0x01, 0x52, 0x3F, 0x55, 0x40, 0x01, 0x00, 0xC8,
+			0x82, 0xC8, 0xBD, 0xD8, 0x45, 0xA0, 0xAC, 0x80,
+			0x06, 0x80, 0x01, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0xF4, 0x24, 0x00, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 27000000,
+		.conf = {
+			0x01, 0xD1, 0x22, 0x51, 0x40, 0x08, 0xFC, 0xE0,
+			0x98, 0xE8, 0xCB, 0xD8, 0x45, 0xA0, 0xAC, 0x80,
+			0x06, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0xE4, 0x24, 0x00, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 27027000,
+		.conf = {
+			0x01, 0xD1, 0x2D, 0x72, 0x40, 0x64, 0x12, 0xC8,
+			0x43, 0xE8, 0x0E, 0xD9, 0x45, 0xA0, 0xAC, 0x80,
+			0x26, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0xE3, 0x24, 0x00, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 36000000,
+		.conf = {
+			0x01, 0x51, 0x2D, 0x55, 0x40, 0x40, 0x00, 0xC8,
+			0x02, 0xC8, 0x0E, 0xD9, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0xAB, 0x24, 0x00, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 40000000,
+		.conf = {
+			0x01, 0xD1, 0x21, 0x31, 0x40, 0x3C, 0x28, 0xC8,
+			0x87, 0xE8, 0xC8, 0xD8, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0x9A, 0x24, 0x00, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 65000000,
+		.conf = {
+			0x01, 0xD1, 0x36, 0x34, 0x40, 0x0C, 0x04, 0xC8,
+			0x82, 0xE8, 0x45, 0xD9, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0xBD, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 71000000,
+		.conf = {
+			0x01, 0xD1, 0x3B, 0x35, 0x40, 0x0C, 0x04, 0xC8,
+			0x85, 0xE8, 0x63, 0xD9, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0x57, 0x24, 0x00, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 73250000,
+		.conf = {
+			0x01, 0xD1, 0x1F, 0x10, 0x40, 0x78, 0x8D, 0xC8,
+			0x81, 0xE8, 0xB7, 0xD8, 0x45, 0xA0, 0xAC, 0x80,
+			0x56, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0xA8, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 74176000,
+		.conf = {
+			0x01, 0xD1, 0x1F, 0x10, 0x40, 0x5B, 0xEF, 0xC8,
+			0x81, 0xE8, 0xB9, 0xD8, 0x45, 0xA0, 0xAC, 0x80,
+			0x56, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0xA6, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 74250000,
+		.conf = {
+			0x01, 0xD1, 0x1F, 0x10, 0x40, 0x40, 0xF8, 0x08,
+			0x81, 0xE8, 0xBA, 0xD8, 0x45, 0xA0, 0xAC, 0x80,
+			0x26, 0x80, 0x09, 0x84, 0x05, 0x22, 0x24, 0x66,
+			0x54, 0xA5, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 83500000,
+		.conf = {
+			0x01, 0xD1, 0x23, 0x11, 0x40, 0x0C, 0xFB, 0xC8,
+			0x85, 0xE8, 0xD1, 0xD8, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0x4A, 0x24, 0x00, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 88750000,
+		.conf = {
+			0x01, 0xD1, 0x25, 0x11, 0x40, 0x18, 0xFF, 0xC8,
+			0x83, 0xE8, 0xDE, 0xD8, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0x45, 0x24, 0x00, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 106500000,
+		.conf = {
+			0x01, 0xD1, 0x2C, 0x12, 0x40, 0x0C, 0x09, 0xC8,
+			0x84, 0xE8, 0x0A, 0xD9, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0x73, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 108000000,
+		.conf = {
+			0x01, 0x51, 0x2D, 0x15, 0x40, 0x01, 0x00, 0xC8,
+			0x82, 0xC8, 0x0E, 0xD9, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0xC7, 0x25, 0x03, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 115500000,
+		.conf = {
+			0x01, 0xD1, 0x30, 0x14, 0x40, 0x0C, 0x03, 0xC8,
+			0x88, 0xE8, 0x21, 0xD9, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0x6A, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 146250000,
+		.conf = {
+			0x01, 0xD1, 0x3D, 0x15, 0x40, 0x18, 0xFD, 0xC8,
+			0x83, 0xE8, 0x6E, 0xD9, 0x45, 0xA0, 0xAC, 0x80,
+			0x08, 0x80, 0x09, 0x84, 0x05, 0x02, 0x24, 0x66,
+			0x54, 0x54, 0x24, 0x01, 0x00, 0x00, 0x01, 0x80,
+		},
+	},
+	{
+		.pixel_clock = 148500000,
+		.conf = {
+			0x01, 0xD1, 0x1F, 0x00, 0x40, 0x40, 0xF8, 0x08,
+			0x81, 0xE8, 0xBA, 0xD8, 0x45, 0xA0, 0xAC, 0x80,
+			0x26, 0x80, 0x09, 0x84, 0x05, 0x22, 0x24, 0x66,
+			0x54, 0x4B, 0x25, 0x03, 0x00, 0x80, 0x01, 0x80,
+		},
+	},
+};
+
+static struct hdmi_driver_data exynos5420_hdmi_driver_data = {
+	.type		= HDMI_TYPE14,
+	.phy_confs	= hdmiphy_5420_configs,
+	.phy_conf_count	= ARRAY_SIZE(hdmiphy_5420_configs),
+	.is_apb_phy	= 1,
+};
+
+static struct hdmi_driver_data exynos4212_hdmi_driver_data = {
+	.type		= HDMI_TYPE14,
+	.phy_confs	= hdmiphy_v14_configs,
+	.phy_conf_count	= ARRAY_SIZE(hdmiphy_v14_configs),
+	.is_apb_phy	= 0,
+};
+
+static struct hdmi_driver_data exynos5_hdmi_driver_data = {
+	.type		= HDMI_TYPE14,
+	.phy_confs	= hdmiphy_v13_configs,
+	.phy_conf_count	= ARRAY_SIZE(hdmiphy_v13_configs),
+	.is_apb_phy	= 0,
 };
 
 static inline u32 hdmi_reg_read(struct hdmi_context *hdata, u32 reg_id)
@@ -404,6 +618,48 @@ static inline void hdmi_reg_writemask(struct hdmi_context *hdata,
 	writel(value, hdata->regs + reg_id);
 }
 
+static int hdmiphy_reg_writeb(struct hdmi_context *hdata,
+			u32 reg_offset, u8 value)
+{
+	if (hdata->hdmiphy_port) {
+		u8 buffer[2];
+		int ret;
+
+		buffer[0] = reg_offset;
+		buffer[1] = value;
+
+		ret = i2c_master_send(hdata->hdmiphy_port, buffer, 2);
+		if (ret == 2)
+			return 0;
+		return ret;
+	} else {
+		writeb(value, hdata->regs_hdmiphy + (reg_offset<<2));
+		return 0;
+	}
+}
+
+static int hdmiphy_reg_write_buf(struct hdmi_context *hdata,
+			u32 reg_offset, const u8 *buf, u32 len)
+{
+	if ((reg_offset + len) > 32)
+		return -EINVAL;
+
+	if (hdata->hdmiphy_port) {
+		int ret;
+
+		ret = i2c_master_send(hdata->hdmiphy_port, buf, len);
+		if (ret == len)
+			return 0;
+		return ret;
+	} else {
+		int i;
+		for (i = 0; i < len; i++)
+			writeb(buf[i], hdata->regs_hdmiphy +
+				((reg_offset + i)<<2));
+		return 0;
+	}
+}
+
 static void hdmi_v13_regs_dump(struct hdmi_context *hdata, char *prefix)
 {
 #define DUMPREG(reg_id) \
@@ -682,11 +938,10 @@ static u8 hdmi_chksum(struct hdmi_context *hdata,
 }
 
 static void hdmi_reg_infoframe(struct hdmi_context *hdata,
-			struct hdmi_infoframe *infoframe)
+			union hdmi_infoframe *infoframe)
 {
 	u32 hdr_sum;
 	u8 chksum;
-	u32 aspect_ratio;
 	u32 mod;
 	u32 vic;
 
@@ -700,40 +955,62 @@ static void hdmi_reg_infoframe(struct hdmi_context *hdata,
 		return;
 	}
 
-	switch (infoframe->type) {
-	case HDMI_PACKET_TYPE_AVI:
+	switch (infoframe->any.type) {
+	case HDMI_INFOFRAME_TYPE_AVI:
 		hdmi_reg_writeb(hdata, HDMI_AVI_CON, HDMI_AVI_CON_EVERY_VSYNC);
-		hdmi_reg_writeb(hdata, HDMI_AVI_HEADER0, infoframe->type);
-		hdmi_reg_writeb(hdata, HDMI_AVI_HEADER1, infoframe->ver);
-		hdmi_reg_writeb(hdata, HDMI_AVI_HEADER2, infoframe->len);
-		hdr_sum = infoframe->type + infoframe->ver + infoframe->len;
+		hdmi_reg_writeb(hdata, HDMI_AVI_HEADER0, infoframe->any.type);
+		hdmi_reg_writeb(hdata, HDMI_AVI_HEADER1,
+				infoframe->any.version);
+		hdmi_reg_writeb(hdata, HDMI_AVI_HEADER2, infoframe->any.length);
+		hdr_sum = infoframe->any.type + infoframe->any.version +
+			  infoframe->any.length;
 
 		/* Output format zero hardcoded ,RGB YBCR selection */
 		hdmi_reg_writeb(hdata, HDMI_AVI_BYTE(1), 0 << 5 |
 			AVI_ACTIVE_FORMAT_VALID |
 			AVI_UNDERSCANNED_DISPLAY_VALID);
 
-		aspect_ratio = AVI_PIC_ASPECT_RATIO_16_9;
-
-		hdmi_reg_writeb(hdata, HDMI_AVI_BYTE(2), aspect_ratio |
-				AVI_SAME_AS_PIC_ASPECT_RATIO);
+		/*
+		 * Set the aspect ratio as per the mode, mentioned in
+		 * Table 9 AVI InfoFrame Data Byte 2 of CEA-861-D Standard
+		 */
+		switch (hdata->mode_conf.aspect_ratio) {
+		case HDMI_PICTURE_ASPECT_4_3:
+			hdmi_reg_writeb(hdata, HDMI_AVI_BYTE(2),
+					hdata->mode_conf.aspect_ratio |
+					AVI_4_3_CENTER_RATIO);
+			break;
+		case HDMI_PICTURE_ASPECT_16_9:
+			hdmi_reg_writeb(hdata, HDMI_AVI_BYTE(2),
+					hdata->mode_conf.aspect_ratio |
+					AVI_16_9_CENTER_RATIO);
+			break;
+		case HDMI_PICTURE_ASPECT_NONE:
+		default:
+			hdmi_reg_writeb(hdata, HDMI_AVI_BYTE(2),
+					hdata->mode_conf.aspect_ratio |
+					AVI_SAME_AS_PIC_ASPECT_RATIO);
+			break;
+		}
 
 		vic = hdata->mode_conf.cea_video_id;
 		hdmi_reg_writeb(hdata, HDMI_AVI_BYTE(4), vic);
 
 		chksum = hdmi_chksum(hdata, HDMI_AVI_BYTE(1),
-					infoframe->len, hdr_sum);
+					infoframe->any.length, hdr_sum);
 		DRM_DEBUG_KMS("AVI checksum = 0x%x\n", chksum);
 		hdmi_reg_writeb(hdata, HDMI_AVI_CHECK_SUM, chksum);
 		break;
-	case HDMI_PACKET_TYPE_AUI:
+	case HDMI_INFOFRAME_TYPE_AUDIO:
 		hdmi_reg_writeb(hdata, HDMI_AUI_CON, 0x02);
-		hdmi_reg_writeb(hdata, HDMI_AUI_HEADER0, infoframe->type);
-		hdmi_reg_writeb(hdata, HDMI_AUI_HEADER1, infoframe->ver);
-		hdmi_reg_writeb(hdata, HDMI_AUI_HEADER2, infoframe->len);
-		hdr_sum = infoframe->type + infoframe->ver + infoframe->len;
+		hdmi_reg_writeb(hdata, HDMI_AUI_HEADER0, infoframe->any.type);
+		hdmi_reg_writeb(hdata, HDMI_AUI_HEADER1,
+				infoframe->any.version);
+		hdmi_reg_writeb(hdata, HDMI_AUI_HEADER2, infoframe->any.length);
+		hdr_sum = infoframe->any.type + infoframe->any.version +
+			  infoframe->any.length;
 		chksum = hdmi_chksum(hdata, HDMI_AUI_BYTE(1),
-					infoframe->len, hdr_sum);
+					infoframe->any.length, hdr_sum);
 		DRM_DEBUG_KMS("AUI checksum = 0x%x\n", chksum);
 		hdmi_reg_writeb(hdata, HDMI_AUI_CHECK_SUM, chksum);
 		break;
@@ -742,58 +1019,66 @@ static void hdmi_reg_infoframe(struct hdmi_context *hdata,
 	}
 }
 
-static bool hdmi_is_connected(void *ctx)
+static enum drm_connector_status hdmi_detect(struct drm_connector *connector,
+				bool force)
 {
-	struct hdmi_context *hdata = ctx;
+	struct hdmi_context *hdata = ctx_from_connector(connector);
+
+	hdata->hpd = gpio_get_value(hdata->hpd_gpio);
 
-	return hdata->hpd;
+	return hdata->hpd ? connector_status_connected :
+			connector_status_disconnected;
 }
 
-static struct edid *hdmi_get_edid(void *ctx, struct drm_connector *connector)
+static void hdmi_connector_destroy(struct drm_connector *connector)
+{
+}
+
+static struct drm_connector_funcs hdmi_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = hdmi_detect,
+	.destroy = hdmi_connector_destroy,
+};
+
+static int hdmi_get_modes(struct drm_connector *connector)
 {
-	struct edid *raw_edid;
-	struct hdmi_context *hdata = ctx;
+	struct hdmi_context *hdata = ctx_from_connector(connector);
+	struct edid *edid;
 
-	if (!hdata->ddc_port)
-		return ERR_PTR(-ENODEV);
+	if (!hdata->ddc_adpt)
+		return -ENODEV;
 
-	raw_edid = drm_get_edid(connector, hdata->ddc_port->adapter);
-	if (!raw_edid)
-		return ERR_PTR(-ENODEV);
+	edid = drm_get_edid(connector, hdata->ddc_adpt);
+	if (!edid)
+		return -ENODEV;
 
-	hdata->dvi_mode = !drm_detect_hdmi_monitor(raw_edid);
+	hdata->dvi_mode = !drm_detect_hdmi_monitor(edid);
 	DRM_DEBUG_KMS("%s : width[%d] x height[%d]\n",
 		(hdata->dvi_mode ? "dvi monitor" : "hdmi monitor"),
-		raw_edid->width_cm, raw_edid->height_cm);
+		edid->width_cm, edid->height_cm);
+
+	drm_mode_connector_update_edid_property(connector, edid);
 
-	return raw_edid;
+	return drm_add_edid_modes(connector, edid);
 }
 
 static int hdmi_find_phy_conf(struct hdmi_context *hdata, u32 pixel_clock)
 {
-	const struct hdmiphy_config *confs;
-	int count, i;
-
-	if (hdata->type == HDMI_TYPE13) {
-		confs = hdmiphy_v13_configs;
-		count = ARRAY_SIZE(hdmiphy_v13_configs);
-	} else if (hdata->type == HDMI_TYPE14) {
-		confs = hdmiphy_v14_configs;
-		count = ARRAY_SIZE(hdmiphy_v14_configs);
-	} else
-		return -EINVAL;
+	int i;
 
-	for (i = 0; i < count; i++)
-		if (confs[i].pixel_clock == pixel_clock)
+	for (i = 0; i < hdata->phy_conf_count; i++)
+		if (hdata->phy_confs[i].pixel_clock == pixel_clock)
 			return i;
 
 	DRM_DEBUG_KMS("Could not find phy config for %d\n", pixel_clock);
 	return -EINVAL;
 }
 
-static int hdmi_check_mode(void *ctx, struct drm_display_mode *mode)
+static int hdmi_mode_valid(struct drm_connector *connector,
+			struct drm_display_mode *mode)
 {
-	struct hdmi_context *hdata = ctx;
+	struct hdmi_context *hdata = ctx_from_connector(connector);
 	int ret;
 
 	DRM_DEBUG_KMS("xres=%d, yres=%d, refresh=%d, intl=%d clock=%d\n",
@@ -801,12 +1086,93 @@ static int hdmi_check_mode(void *ctx, struct drm_display_mode *mode)
 		(mode->flags & DRM_MODE_FLAG_INTERLACE) ? true :
 		false, mode->clock * 1000);
 
+	ret = mixer_check_mode(mode);
+	if (ret)
+		return MODE_BAD;
+
 	ret = hdmi_find_phy_conf(hdata, mode->clock * 1000);
 	if (ret < 0)
+		return MODE_BAD;
+
+	return MODE_OK;
+}
+
+static struct drm_encoder *hdmi_best_encoder(struct drm_connector *connector)
+{
+	struct hdmi_context *hdata = ctx_from_connector(connector);
+
+	return hdata->encoder;
+}
+
+static struct drm_connector_helper_funcs hdmi_connector_helper_funcs = {
+	.get_modes = hdmi_get_modes,
+	.mode_valid = hdmi_mode_valid,
+	.best_encoder = hdmi_best_encoder,
+};
+
+static int hdmi_create_connector(struct exynos_drm_display *display,
+			struct drm_encoder *encoder)
+{
+	struct hdmi_context *hdata = display->ctx;
+	struct drm_connector *connector = &hdata->connector;
+	int ret;
+
+	hdata->encoder = encoder;
+	connector->interlace_allowed = true;
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	ret = drm_connector_init(hdata->drm_dev, connector,
+			&hdmi_connector_funcs, DRM_MODE_CONNECTOR_HDMIA);
+	if (ret) {
+		DRM_ERROR("Failed to initialize connector with drm\n");
 		return ret;
+	}
+
+	drm_connector_helper_add(connector, &hdmi_connector_helper_funcs);
+	drm_sysfs_connector_add(connector);
+	drm_mode_connector_attach_encoder(connector, encoder);
+
 	return 0;
 }
 
+static void hdmi_mode_fixup(struct exynos_drm_display *display,
+				struct drm_connector *connector,
+				const struct drm_display_mode *mode,
+				struct drm_display_mode *adjusted_mode)
+{
+	struct drm_display_mode *m;
+	int mode_ok;
+
+	DRM_DEBUG_KMS("%s\n", __FILE__);
+
+	drm_mode_set_crtcinfo(adjusted_mode, 0);
+
+	mode_ok = hdmi_mode_valid(connector, adjusted_mode);
+
+	/* just return if user desired mode exists. */
+	if (mode_ok == MODE_OK)
+		return;
+
+	/*
+	 * otherwise, find the most suitable mode among modes and change it
+	 * to adjusted_mode.
+	 */
+	list_for_each_entry(m, &connector->modes, head) {
+		mode_ok = hdmi_mode_valid(connector, m);
+
+		if (mode_ok == MODE_OK) {
+			DRM_INFO("desired mode doesn't exist so\n");
+			DRM_INFO("use the most suitable mode among modes.\n");
+
+			DRM_DEBUG_KMS("Adjusted Mode: [%d]x[%d] [%d]Hz\n",
+				m->hdisplay, m->vdisplay, m->vrefresh);
+
+			drm_mode_copy(adjusted_mode, m);
+			break;
+		}
+	}
+}
+
 static void hdmi_set_acr(u32 freq, u8 *acr)
 {
 	u32 n, cts;
@@ -967,25 +1333,20 @@ static void hdmi_audio_control(struct hdmi_context *hdata, bool onoff)
 			HDMI_ASP_EN : HDMI_ASP_DIS, HDMI_ASP_MASK);
 }
 
-static void hdmi_conf_reset(struct hdmi_context *hdata)
+static void hdmi_start(struct hdmi_context *hdata, bool start)
 {
-	u32 reg;
+	u32 val = start ? HDMI_TG_EN : 0;
 
-	if (hdata->type == HDMI_TYPE13)
-		reg = HDMI_V13_CORE_RSTOUT;
-	else
-		reg = HDMI_CORE_RSTOUT;
+	if (hdata->current_mode.flags & DRM_MODE_FLAG_INTERLACE)
+		val |= HDMI_FIELD_EN;
 
-	/* resetting HDMI core */
-	hdmi_reg_writemask(hdata, reg,  0, HDMI_CORE_SW_RSTOUT);
-	usleep_range(10000, 12000);
-	hdmi_reg_writemask(hdata, reg, ~0, HDMI_CORE_SW_RSTOUT);
-	usleep_range(10000, 12000);
+	hdmi_reg_writemask(hdata, HDMI_CON_0, val, HDMI_EN);
+	hdmi_reg_writemask(hdata, HDMI_TG_CMD, val, HDMI_TG_EN | HDMI_FIELD_EN);
 }
 
 static void hdmi_conf_init(struct hdmi_context *hdata)
 {
-	struct hdmi_infoframe infoframe;
+	union hdmi_infoframe infoframe;
 
 	/* disable HPD interrupts from HDMI IP block, use GPIO instead */
 	hdmi_reg_writemask(hdata, HDMI_INTC_CON, 0, HDMI_INTC_EN_GLOBAL |
@@ -994,6 +1355,8 @@ static void hdmi_conf_init(struct hdmi_context *hdata)
 	/* choose HDMI mode */
 	hdmi_reg_writemask(hdata, HDMI_MODE_SEL,
 		HDMI_MODE_HDMI_EN, HDMI_MODE_MASK);
+	/* Apply Video preable and Guard band in HDMI mode only */
+	hdmi_reg_writeb(hdata, HDMI_CON_2, 0);
 	/* disable bluescreen */
 	hdmi_reg_writemask(hdata, HDMI_CON_0, 0, HDMI_BLUE_SCR_EN);
 
@@ -1021,14 +1384,14 @@ static void hdmi_conf_init(struct hdmi_context *hdata)
 		hdmi_reg_writeb(hdata, HDMI_V13_AUI_CON, 0x02);
 		hdmi_reg_writeb(hdata, HDMI_V13_ACR_CON, 0x04);
 	} else {
-		infoframe.type = HDMI_PACKET_TYPE_AVI;
-		infoframe.ver = HDMI_AVI_VERSION;
-		infoframe.len = HDMI_AVI_LENGTH;
+		infoframe.any.type = HDMI_INFOFRAME_TYPE_AVI;
+		infoframe.any.version = HDMI_AVI_VERSION;
+		infoframe.any.length = HDMI_AVI_LENGTH;
 		hdmi_reg_infoframe(hdata, &infoframe);
 
-		infoframe.type = HDMI_PACKET_TYPE_AUI;
-		infoframe.ver = HDMI_AUI_VERSION;
-		infoframe.len = HDMI_AUI_LENGTH;
+		infoframe.any.type = HDMI_INFOFRAME_TYPE_AUDIO;
+		infoframe.any.version = HDMI_AUI_VERSION;
+		infoframe.any.length = HDMI_AUI_LENGTH;
 		hdmi_reg_infoframe(hdata, &infoframe);
 
 		/* enable AVI packet every vsync, fixes purple line problem */
@@ -1117,12 +1480,7 @@ static void hdmi_v13_mode_apply(struct hdmi_context *hdata)
 	clk_prepare_enable(hdata->res.sclk_hdmi);
 
 	/* enable HDMI and timing generator */
-	hdmi_reg_writemask(hdata, HDMI_CON_0, ~0, HDMI_EN);
-	if (core->int_pro_mode[0])
-		hdmi_reg_writemask(hdata, HDMI_TG_CMD, ~0, HDMI_TG_EN |
-				HDMI_FIELD_EN);
-	else
-		hdmi_reg_writemask(hdata, HDMI_TG_CMD, ~0, HDMI_TG_EN);
+	hdmi_start(hdata, true);
 }
 
 static void hdmi_v14_mode_apply(struct hdmi_context *hdata)
@@ -1284,12 +1642,7 @@ static void hdmi_v14_mode_apply(struct hdmi_context *hdata)
 	clk_prepare_enable(hdata->res.sclk_hdmi);
 
 	/* enable HDMI and timing generator */
-	hdmi_reg_writemask(hdata, HDMI_CON_0, ~0, HDMI_EN);
-	if (core->int_pro_mode[0])
-		hdmi_reg_writemask(hdata, HDMI_TG_CMD, ~0, HDMI_TG_EN |
-				HDMI_FIELD_EN);
-	else
-		hdmi_reg_writemask(hdata, HDMI_TG_CMD, ~0, HDMI_TG_EN);
+	hdmi_start(hdata, true);
 }
 
 static void hdmi_mode_apply(struct hdmi_context *hdata)
@@ -1330,32 +1683,51 @@ static void hdmiphy_conf_reset(struct hdmi_context *hdata)
 
 static void hdmiphy_poweron(struct hdmi_context *hdata)
 {
-	if (hdata->type == HDMI_TYPE14)
-		hdmi_reg_writemask(hdata, HDMI_PHY_CON_0, 0,
-			HDMI_PHY_POWER_OFF_EN);
+	if (hdata->type != HDMI_TYPE14)
+		return;
+
+	DRM_DEBUG_KMS("\n");
+
+	/* For PHY Mode Setting */
+	hdmiphy_reg_writeb(hdata, HDMIPHY_MODE_SET_DONE,
+				HDMI_PHY_ENABLE_MODE_SET);
+	/* Phy Power On */
+	hdmiphy_reg_writeb(hdata, HDMIPHY_POWER,
+				HDMI_PHY_POWER_ON);
+	/* For PHY Mode Setting */
+	hdmiphy_reg_writeb(hdata, HDMIPHY_MODE_SET_DONE,
+				HDMI_PHY_DISABLE_MODE_SET);
+	/* PHY SW Reset */
+	hdmiphy_conf_reset(hdata);
 }
 
 static void hdmiphy_poweroff(struct hdmi_context *hdata)
 {
-	if (hdata->type == HDMI_TYPE14)
-		hdmi_reg_writemask(hdata, HDMI_PHY_CON_0, ~0,
-			HDMI_PHY_POWER_OFF_EN);
+	if (hdata->type != HDMI_TYPE14)
+		return;
+
+	DRM_DEBUG_KMS("\n");
+
+	/* PHY SW Reset */
+	hdmiphy_conf_reset(hdata);
+	/* For PHY Mode Setting */
+	hdmiphy_reg_writeb(hdata, HDMIPHY_MODE_SET_DONE,
+				HDMI_PHY_ENABLE_MODE_SET);
+
+	/* PHY Power Off */
+	hdmiphy_reg_writeb(hdata, HDMIPHY_POWER,
+				HDMI_PHY_POWER_OFF);
+
+	/* For PHY Mode Setting */
+	hdmiphy_reg_writeb(hdata, HDMIPHY_MODE_SET_DONE,
+				HDMI_PHY_DISABLE_MODE_SET);
 }
 
 static void hdmiphy_conf_apply(struct hdmi_context *hdata)
 {
-	const u8 *hdmiphy_data;
-	u8 buffer[32];
-	u8 operation[2];
-	u8 read_buffer[32] = {0, };
 	int ret;
 	int i;
 
-	if (!hdata->hdmiphy_port) {
-		DRM_ERROR("hdmiphy is not attached\n");
-		return;
-	}
-
 	/* pixel clock */
 	i = hdmi_find_phy_conf(hdata, hdata->mode_conf.pixel_clock);
 	if (i < 0) {
@@ -1363,39 +1735,21 @@ static void hdmiphy_conf_apply(struct hdmi_context *hdata)
 		return;
 	}
 
-	if (hdata->type == HDMI_TYPE13)
-		hdmiphy_data = hdmiphy_v13_configs[i].conf;
-	else
-		hdmiphy_data = hdmiphy_v14_configs[i].conf;
-
-	memcpy(buffer, hdmiphy_data, 32);
-	ret = i2c_master_send(hdata->hdmiphy_port, buffer, 32);
-	if (ret != 32) {
-		DRM_ERROR("failed to configure HDMIPHY via I2C\n");
+	ret = hdmiphy_reg_write_buf(hdata, 0, hdata->phy_confs[i].conf, 32);
+	if (ret) {
+		DRM_ERROR("failed to configure hdmiphy\n");
 		return;
 	}
 
 	usleep_range(10000, 12000);
 
-	/* operation mode */
-	operation[0] = 0x1f;
-	operation[1] = 0x80;
-
-	ret = i2c_master_send(hdata->hdmiphy_port, operation, 2);
-	if (ret != 2) {
+	ret = hdmiphy_reg_writeb(hdata, HDMIPHY_MODE_SET_DONE,
+				HDMI_PHY_DISABLE_MODE_SET);
+	if (ret) {
 		DRM_ERROR("failed to enable hdmiphy\n");
 		return;
 	}
 
-	ret = i2c_master_recv(hdata->hdmiphy_port, read_buffer, 32);
-	if (ret < 0) {
-		DRM_ERROR("failed to read hdmiphy config\n");
-		return;
-	}
-
-	for (i = 0; i < ret; i++)
-		DRM_DEBUG_KMS("hdmiphy[0x%02x] write[0x%02x] - "
-			"recv [0x%02x]\n", i, buffer[i], read_buffer[i]);
 }
 
 static void hdmi_conf_apply(struct hdmi_context *hdata)
@@ -1404,7 +1758,7 @@ static void hdmi_conf_apply(struct hdmi_context *hdata)
 	hdmiphy_conf_apply(hdata);
 
 	mutex_lock(&hdata->hdmi_mutex);
-	hdmi_conf_reset(hdata);
+	hdmi_start(hdata, false);
 	hdmi_conf_init(hdata);
 	mutex_unlock(&hdata->hdmi_mutex);
 
@@ -1435,6 +1789,7 @@ static void hdmi_v13_mode_set(struct hdmi_context *hdata,
 	hdata->mode_conf.cea_video_id =
 		drm_match_cea_mode((struct drm_display_mode *)m);
 	hdata->mode_conf.pixel_clock = m->clock * 1000;
+	hdata->mode_conf.aspect_ratio = m->picture_aspect_ratio;
 
 	hdmi_set_reg(core->h_blank, 2, m->htotal - m->hdisplay);
 	hdmi_set_reg(core->h_v_line, 3, (m->htotal << 12) | m->vtotal);
@@ -1531,6 +1886,7 @@ static void hdmi_v14_mode_set(struct hdmi_context *hdata,
 	hdata->mode_conf.cea_video_id =
 		drm_match_cea_mode((struct drm_display_mode *)m);
 	hdata->mode_conf.pixel_clock = m->clock * 1000;
+	hdata->mode_conf.aspect_ratio = m->picture_aspect_ratio;
 
 	hdmi_set_reg(core->h_blank, 2, m->htotal - m->hdisplay);
 	hdmi_set_reg(core->v_line, 2, m->vtotal);
@@ -1632,9 +1988,10 @@ static void hdmi_v14_mode_set(struct hdmi_context *hdata,
 	hdmi_set_reg(tg->tg_3d, 1, 0x0);
 }
 
-static void hdmi_mode_set(void *ctx, struct drm_display_mode *mode)
+static void hdmi_mode_set(struct exynos_drm_display *display,
+			struct drm_display_mode *mode)
 {
-	struct hdmi_context *hdata = ctx;
+	struct hdmi_context *hdata = display->ctx;
 	struct drm_display_mode *m = mode;
 
 	DRM_DEBUG_KMS("xres=%d, yres=%d, refresh=%d, intl=%s\n",
@@ -1642,22 +1999,18 @@ static void hdmi_mode_set(void *ctx, struct drm_display_mode *mode)
 		m->vrefresh, (m->flags & DRM_MODE_FLAG_INTERLACE) ?
 		"INTERLACED" : "PROGERESSIVE");
 
+	/* preserve mode information for later use. */
+	drm_mode_copy(&hdata->current_mode, mode);
+
 	if (hdata->type == HDMI_TYPE13)
 		hdmi_v13_mode_set(hdata, mode);
 	else
 		hdmi_v14_mode_set(hdata, mode);
 }
 
-static void hdmi_get_max_resol(void *ctx, unsigned int *width,
-					unsigned int *height)
+static void hdmi_commit(struct exynos_drm_display *display)
 {
-	*width = MAX_WIDTH;
-	*height = MAX_HEIGHT;
-}
-
-static void hdmi_commit(void *ctx)
-{
-	struct hdmi_context *hdata = ctx;
+	struct hdmi_context *hdata = display->ctx;
 
 	mutex_lock(&hdata->hdmi_mutex);
 	if (!hdata->powered) {
@@ -1669,8 +2022,9 @@ static void hdmi_commit(void *ctx)
 	hdmi_conf_apply(hdata);
 }
 
-static void hdmi_poweron(struct hdmi_context *hdata)
+static void hdmi_poweron(struct exynos_drm_display *display)
 {
+	struct hdmi_context *hdata = display->ctx;
 	struct hdmi_resources *res = &hdata->res;
 
 	mutex_lock(&hdata->hdmi_mutex);
@@ -1683,18 +2037,25 @@ static void hdmi_poweron(struct hdmi_context *hdata)
 
 	mutex_unlock(&hdata->hdmi_mutex);
 
+	pm_runtime_get_sync(hdata->dev);
+
 	if (regulator_bulk_enable(res->regul_count, res->regul_bulk))
 		DRM_DEBUG_KMS("failed to enable regulator bulk\n");
 
-	clk_prepare_enable(res->hdmiphy);
+	/* set pmu hdmiphy control bit to enable hdmiphy */
+	regmap_update_bits(hdata->pmureg, PMU_HDMI_PHY_CONTROL,
+			PMU_HDMI_PHY_ENABLE_BIT, 1);
+
 	clk_prepare_enable(res->hdmi);
 	clk_prepare_enable(res->sclk_hdmi);
 
 	hdmiphy_poweron(hdata);
+	hdmi_commit(display);
 }
 
-static void hdmi_poweroff(struct hdmi_context *hdata)
+static void hdmi_poweroff(struct exynos_drm_display *display)
 {
+	struct hdmi_context *hdata = display->ctx;
 	struct hdmi_resources *res = &hdata->res;
 
 	mutex_lock(&hdata->hdmi_mutex);
@@ -1702,42 +2063,62 @@ static void hdmi_poweroff(struct hdmi_context *hdata)
 		goto out;
 	mutex_unlock(&hdata->hdmi_mutex);
 
-	/*
-	 * The TV power domain needs any condition of hdmiphy to turn off and
-	 * its reset state seems to meet the condition.
-	 */
-	hdmiphy_conf_reset(hdata);
+	/* HDMI System Disable */
+	hdmi_reg_writemask(hdata, HDMI_CON_0, 0, HDMI_EN);
+
 	hdmiphy_poweroff(hdata);
 
+	cancel_delayed_work(&hdata->hotplug_work);
+
 	clk_disable_unprepare(res->sclk_hdmi);
 	clk_disable_unprepare(res->hdmi);
-	clk_disable_unprepare(res->hdmiphy);
+
+	/* reset pmu hdmiphy control bit to disable hdmiphy */
+	regmap_update_bits(hdata->pmureg, PMU_HDMI_PHY_CONTROL,
+			PMU_HDMI_PHY_ENABLE_BIT, 0);
+
 	regulator_bulk_disable(res->regul_count, res->regul_bulk);
 
-	mutex_lock(&hdata->hdmi_mutex);
+	pm_runtime_put_sync(hdata->dev);
 
+	mutex_lock(&hdata->hdmi_mutex);
 	hdata->powered = false;
 
 out:
 	mutex_unlock(&hdata->hdmi_mutex);
 }
 
-static void hdmi_dpms(void *ctx, int mode)
+static void hdmi_dpms(struct exynos_drm_display *display, int mode)
 {
-	struct hdmi_context *hdata = ctx;
+	struct hdmi_context *hdata = display->ctx;
+	struct drm_encoder *encoder = hdata->encoder;
+	struct drm_crtc *crtc = encoder->crtc;
+	struct drm_crtc_helper_funcs *funcs = NULL;
 
 	DRM_DEBUG_KMS("mode %d\n", mode);
 
 	switch (mode) {
 	case DRM_MODE_DPMS_ON:
-		if (pm_runtime_suspended(hdata->dev))
-			pm_runtime_get_sync(hdata->dev);
+		hdmi_poweron(display);
 		break;
 	case DRM_MODE_DPMS_STANDBY:
 	case DRM_MODE_DPMS_SUSPEND:
 	case DRM_MODE_DPMS_OFF:
-		if (!pm_runtime_suspended(hdata->dev))
-			pm_runtime_put_sync(hdata->dev);
+		/*
+		 * The SFRs of VP and Mixer are updated by Vertical Sync of
+		 * Timing generator which is a part of HDMI so the sequence
+		 * to disable TV Subsystem should be as following,
+		 *	VP -> Mixer -> HDMI
+		 *
+		 * Below codes will try to disable Mixer and VP(if used)
+		 * prior to disabling HDMI.
+		 */
+		if (crtc)
+			funcs = crtc->helper_private;
+		if (funcs && funcs->dpms)
+			(*funcs->dpms)(crtc, mode);
+
+		hdmi_poweroff(display);
 		break;
 	default:
 		DRM_DEBUG_KMS("unknown dpms mode: %d\n", mode);
@@ -1745,30 +2126,39 @@ static void hdmi_dpms(void *ctx, int mode)
 	}
 }
 
-static struct exynos_hdmi_ops hdmi_ops = {
-	/* display */
-	.is_connected	= hdmi_is_connected,
-	.get_edid	= hdmi_get_edid,
-	.check_mode	= hdmi_check_mode,
-
-	/* manager */
+static struct exynos_drm_display_ops hdmi_display_ops = {
+	.create_connector = hdmi_create_connector,
+	.mode_fixup	= hdmi_mode_fixup,
 	.mode_set	= hdmi_mode_set,
-	.get_max_resol	= hdmi_get_max_resol,
-	.commit		= hdmi_commit,
 	.dpms		= hdmi_dpms,
+	.commit		= hdmi_commit,
 };
 
-static irqreturn_t hdmi_irq_thread(int irq, void *arg)
+static struct exynos_drm_display hdmi_display = {
+	.type = EXYNOS_DISPLAY_TYPE_HDMI,
+	.ops = &hdmi_display_ops,
+};
+
+static void hdmi_hotplug_work_func(struct work_struct *work)
 {
-	struct exynos_drm_hdmi_context *ctx = arg;
-	struct hdmi_context *hdata = ctx->ctx;
+	struct hdmi_context *hdata;
+
+	hdata = container_of(work, struct hdmi_context, hotplug_work.work);
 
 	mutex_lock(&hdata->hdmi_mutex);
 	hdata->hpd = gpio_get_value(hdata->hpd_gpio);
 	mutex_unlock(&hdata->hdmi_mutex);
 
-	if (ctx->drm_dev)
-		drm_helper_hpd_irq_event(ctx->drm_dev);
+	if (hdata->drm_dev)
+		drm_helper_hpd_irq_event(hdata->drm_dev);
+}
+
+static irqreturn_t hdmi_irq_thread(int irq, void *arg)
+{
+	struct hdmi_context *hdata = arg;
+
+	mod_delayed_work(system_wq, &hdata->hotplug_work,
+			msecs_to_jiffies(HOTPLUG_DEBOUNCE_MS));
 
 	return IRQ_HANDLED;
 }
@@ -1787,37 +2177,35 @@ static int hdmi_resources_init(struct hdmi_context *hdata)
 
 	DRM_DEBUG_KMS("HDMI resource init\n");
 
-	memset(res, 0, sizeof(*res));
-
 	/* get clocks, power */
 	res->hdmi = devm_clk_get(dev, "hdmi");
 	if (IS_ERR(res->hdmi)) {
 		DRM_ERROR("failed to get clock 'hdmi'\n");
+		ret = PTR_ERR(res->hdmi);
 		goto fail;
 	}
 	res->sclk_hdmi = devm_clk_get(dev, "sclk_hdmi");
 	if (IS_ERR(res->sclk_hdmi)) {
 		DRM_ERROR("failed to get clock 'sclk_hdmi'\n");
+		ret = PTR_ERR(res->sclk_hdmi);
 		goto fail;
 	}
 	res->sclk_pixel = devm_clk_get(dev, "sclk_pixel");
 	if (IS_ERR(res->sclk_pixel)) {
 		DRM_ERROR("failed to get clock 'sclk_pixel'\n");
+		ret = PTR_ERR(res->sclk_pixel);
 		goto fail;
 	}
 	res->sclk_hdmiphy = devm_clk_get(dev, "sclk_hdmiphy");
 	if (IS_ERR(res->sclk_hdmiphy)) {
 		DRM_ERROR("failed to get clock 'sclk_hdmiphy'\n");
-		goto fail;
-	}
-	res->hdmiphy = devm_clk_get(dev, "hdmiphy");
-	if (IS_ERR(res->hdmiphy)) {
-		DRM_ERROR("failed to get clock 'hdmiphy'\n");
+		ret = PTR_ERR(res->sclk_hdmiphy);
 		goto fail;
 	}
 	res->mout_hdmi = devm_clk_get(dev, "mout_hdmi");
 	if (IS_ERR(res->mout_hdmi)) {
 		DRM_ERROR("failed to get clock 'mout_hdmi'\n");
+		ret = PTR_ERR(res->mout_hdmi);
 		goto fail;
 	}
 
@@ -1825,8 +2213,10 @@ static int hdmi_resources_init(struct hdmi_context *hdata)
 
 	res->regul_bulk = devm_kzalloc(dev, ARRAY_SIZE(supply) *
 		sizeof(res->regul_bulk[0]), GFP_KERNEL);
-	if (!res->regul_bulk)
+	if (!res->regul_bulk) {
+		ret = -ENOMEM;
 		goto fail;
+	}
 	for (i = 0; i < ARRAY_SIZE(supply); ++i) {
 		res->regul_bulk[i].supply = supply[i];
 		res->regul_bulk[i].consumer = NULL;
@@ -1834,28 +2224,14 @@ static int hdmi_resources_init(struct hdmi_context *hdata)
 	ret = devm_regulator_bulk_get(dev, ARRAY_SIZE(supply), res->regul_bulk);
 	if (ret) {
 		DRM_ERROR("failed to get regulators\n");
-		goto fail;
+		return ret;
 	}
 	res->regul_count = ARRAY_SIZE(supply);
 
-	return 0;
+	return ret;
 fail:
 	DRM_ERROR("HDMI resource init - failed\n");
-	return -ENODEV;
-}
-
-static struct i2c_client *hdmi_ddc, *hdmi_hdmiphy;
-
-void hdmi_attach_ddc_client(struct i2c_client *ddc)
-{
-	if (ddc)
-		hdmi_ddc = ddc;
-}
-
-void hdmi_attach_hdmiphy_client(struct i2c_client *hdmiphy)
-{
-	if (hdmiphy)
-		hdmi_hdmiphy = hdmiphy;
+	return ret;
 }
 
 static struct s5p_hdmi_platform_data *drm_hdmi_dt_parse_pdata
@@ -1885,51 +2261,110 @@ err_data:
 static struct of_device_id hdmi_match_types[] = {
 	{
 		.compatible = "samsung,exynos5-hdmi",
-		.data	= (void	*)HDMI_TYPE14,
+		.data = &exynos5_hdmi_driver_data,
 	}, {
 		.compatible = "samsung,exynos4212-hdmi",
-		.data	= (void	*)HDMI_TYPE14,
+		.data = &exynos4212_hdmi_driver_data,
+	}, {
+		.compatible = "samsung,exynos5420-hdmi",
+		.data = &exynos5420_hdmi_driver_data,
 	}, {
 		/* end node */
 	}
 };
 
+static int hdmi_bind(struct device *dev, struct device *master, void *data)
+{
+	struct drm_device *drm_dev = data;
+	struct hdmi_context *hdata;
+
+	hdata = hdmi_display.ctx;
+	hdata->drm_dev = drm_dev;
+
+	return exynos_drm_create_enc_conn(drm_dev, &hdmi_display);
+}
+
+static void hdmi_unbind(struct device *dev, struct device *master, void *data)
+{
+	struct exynos_drm_display *display = get_hdmi_display(dev);
+	struct drm_encoder *encoder = display->encoder;
+	struct hdmi_context *hdata = display->ctx;
+
+	encoder->funcs->destroy(encoder);
+	drm_connector_cleanup(&hdata->connector);
+}
+
+static const struct component_ops hdmi_component_ops = {
+	.bind	= hdmi_bind,
+	.unbind = hdmi_unbind,
+};
+
+static struct device_node *hdmi_legacy_ddc_dt_binding(struct device *dev)
+{
+	const char *compatible_str = "samsung,exynos4210-hdmiddc";
+	struct device_node *np;
+
+	np = of_find_compatible_node(NULL, NULL, compatible_str);
+	if (np)
+		return of_get_next_parent(np);
+
+	return NULL;
+}
+
+static struct device_node *hdmi_legacy_phy_dt_binding(struct device *dev)
+{
+	const char *compatible_str = "samsung,exynos4212-hdmiphy";
+
+	return of_find_compatible_node(NULL, NULL, compatible_str);
+}
+
 static int hdmi_probe(struct platform_device *pdev)
 {
+	struct device_node *ddc_node, *phy_node;
+	struct s5p_hdmi_platform_data *pdata;
+	struct hdmi_driver_data *drv_data;
+	const struct of_device_id *match;
 	struct device *dev = &pdev->dev;
-	struct exynos_drm_hdmi_context *drm_hdmi_ctx;
 	struct hdmi_context *hdata;
-	struct s5p_hdmi_platform_data *pdata;
 	struct resource *res;
-	const struct of_device_id *match;
 	int ret;
 
-	 if (!dev->of_node)
-		return -ENODEV;
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR,
+					hdmi_display.type);
+	if (ret)
+		return ret;
 
-	pdata = drm_hdmi_dt_parse_pdata(dev);
-	if (!pdata)
-		return -EINVAL;
+	if (!dev->of_node) {
+		ret = -ENODEV;
+		goto err_del_component;
+	}
 
-	drm_hdmi_ctx = devm_kzalloc(dev, sizeof(*drm_hdmi_ctx), GFP_KERNEL);
-	if (!drm_hdmi_ctx)
-		return -ENOMEM;
+	pdata = drm_hdmi_dt_parse_pdata(dev);
+	if (!pdata) {
+		ret = -EINVAL;
+		goto err_del_component;
+	}
 
 	hdata = devm_kzalloc(dev, sizeof(struct hdmi_context), GFP_KERNEL);
-	if (!hdata)
-		return -ENOMEM;
+	if (!hdata) {
+		ret = -ENOMEM;
+		goto err_del_component;
+	}
 
 	mutex_init(&hdata->hdmi_mutex);
 
-	drm_hdmi_ctx->ctx = (void *)hdata;
-	hdata->parent_ctx = (void *)drm_hdmi_ctx;
-
-	platform_set_drvdata(pdev, drm_hdmi_ctx);
+	platform_set_drvdata(pdev, &hdmi_display);
 
 	match = of_match_node(hdmi_match_types, dev->of_node);
-	if (!match)
-		return -ENODEV;
-	hdata->type = (enum hdmi_type)match->data;
+	if (!match) {
+		ret = -ENODEV;
+		goto err_del_component;
+	}
+
+	drv_data = (struct hdmi_driver_data *)match->data;
+	hdata->type = drv_data->type;
+	hdata->phy_confs = drv_data->phy_confs;
+	hdata->phy_conf_count = drv_data->phy_conf_count;
 
 	hdata->hpd_gpio = pdata->hpd_gpio;
 	hdata->dev = dev;
@@ -1937,36 +2372,69 @@ static int hdmi_probe(struct platform_device *pdev)
 	ret = hdmi_resources_init(hdata);
 	if (ret) {
 		DRM_ERROR("hdmi_resources_init failed\n");
-		return -EINVAL;
+		return ret;
 	}
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	hdata->regs = devm_ioremap_resource(dev, res);
-	if (IS_ERR(hdata->regs))
-		return PTR_ERR(hdata->regs);
+	if (IS_ERR(hdata->regs)) {
+		ret = PTR_ERR(hdata->regs);
+		goto err_del_component;
+	}
 
 	ret = devm_gpio_request(dev, hdata->hpd_gpio, "HPD");
 	if (ret) {
 		DRM_ERROR("failed to request HPD gpio\n");
-		return ret;
+		goto err_del_component;
 	}
 
+	ddc_node = hdmi_legacy_ddc_dt_binding(dev);
+	if (ddc_node)
+		goto out_get_ddc_adpt;
+
 	/* DDC i2c driver */
-	if (i2c_add_driver(&ddc_driver)) {
-		DRM_ERROR("failed to register ddc i2c driver\n");
-		return -ENOENT;
+	ddc_node = of_parse_phandle(dev->of_node, "ddc", 0);
+	if (!ddc_node) {
+		DRM_ERROR("Failed to find ddc node in device tree\n");
+		ret = -ENODEV;
+		goto err_del_component;
 	}
 
-	hdata->ddc_port = hdmi_ddc;
+out_get_ddc_adpt:
+	hdata->ddc_adpt = of_find_i2c_adapter_by_node(ddc_node);
+	if (!hdata->ddc_adpt) {
+		DRM_ERROR("Failed to get ddc i2c adapter by node\n");
+		return -EPROBE_DEFER;
+	}
+
+	phy_node = hdmi_legacy_phy_dt_binding(dev);
+	if (phy_node)
+		goto out_get_phy_port;
 
 	/* hdmiphy i2c driver */
-	if (i2c_add_driver(&hdmiphy_driver)) {
-		DRM_ERROR("failed to register hdmiphy i2c driver\n");
-		ret = -ENOENT;
+	phy_node = of_parse_phandle(dev->of_node, "phy", 0);
+	if (!phy_node) {
+		DRM_ERROR("Failed to find hdmiphy node in device tree\n");
+		ret = -ENODEV;
 		goto err_ddc;
 	}
 
-	hdata->hdmiphy_port = hdmi_hdmiphy;
+out_get_phy_port:
+	if (drv_data->is_apb_phy) {
+		hdata->regs_hdmiphy = of_iomap(phy_node, 0);
+		if (!hdata->regs_hdmiphy) {
+			DRM_ERROR("failed to ioremap hdmi phy\n");
+			ret = -ENOMEM;
+			goto err_ddc;
+		}
+	} else {
+		hdata->hdmiphy_port = of_find_i2c_device_by_node(phy_node);
+		if (!hdata->hdmiphy_port) {
+			DRM_ERROR("Failed to get hdmi phy i2c client\n");
+			ret = -EPROBE_DEFER;
+			goto err_ddc;
+		}
+	}
 
 	hdata->irq = gpio_to_irq(hdata->hpd_gpio);
 	if (hdata->irq < 0) {
@@ -1977,114 +2445,64 @@ static int hdmi_probe(struct platform_device *pdev)
 
 	hdata->hpd = gpio_get_value(hdata->hpd_gpio);
 
+	INIT_DELAYED_WORK(&hdata->hotplug_work, hdmi_hotplug_work_func);
+
 	ret = devm_request_threaded_irq(dev, hdata->irq, NULL,
 			hdmi_irq_thread, IRQF_TRIGGER_RISING |
 			IRQF_TRIGGER_FALLING | IRQF_ONESHOT,
-			"hdmi", drm_hdmi_ctx);
+			"hdmi", hdata);
 	if (ret) {
 		DRM_ERROR("failed to register hdmi interrupt\n");
 		goto err_hdmiphy;
 	}
 
-	/* Attach HDMI Driver to common hdmi. */
-	exynos_hdmi_drv_attach(drm_hdmi_ctx);
-
-	/* register specific callbacks to common hdmi. */
-	exynos_hdmi_ops_register(&hdmi_ops);
+	hdata->pmureg = syscon_regmap_lookup_by_phandle(dev->of_node,
+			"samsung,syscon-phandle");
+	if (IS_ERR(hdata->pmureg)) {
+		DRM_ERROR("syscon regmap lookup failed.\n");
+		ret = -EPROBE_DEFER;
+		goto err_hdmiphy;
+	}
 
 	pm_runtime_enable(dev);
+	hdmi_display.ctx = hdata;
 
-	return 0;
+	ret = component_add(&pdev->dev, &hdmi_component_ops);
+	if (ret)
+		goto err_disable_pm_runtime;
 
-err_hdmiphy:
-	i2c_del_driver(&hdmiphy_driver);
-err_ddc:
-	i2c_del_driver(&ddc_driver);
 	return ret;
-}
-
-static int hdmi_remove(struct platform_device *pdev)
-{
-	struct device *dev = &pdev->dev;
 
+err_disable_pm_runtime:
 	pm_runtime_disable(dev);
 
-	/* hdmiphy i2c driver */
-	i2c_del_driver(&hdmiphy_driver);
-	/* DDC i2c driver */
-	i2c_del_driver(&ddc_driver);
-
-	return 0;
-}
-
-#ifdef CONFIG_PM_SLEEP
-static int hdmi_suspend(struct device *dev)
-{
-	struct exynos_drm_hdmi_context *ctx = get_hdmi_context(dev);
-	struct hdmi_context *hdata = ctx->ctx;
-
-	disable_irq(hdata->irq);
-
-	hdata->hpd = false;
-	if (ctx->drm_dev)
-		drm_helper_hpd_irq_event(ctx->drm_dev);
-
-	if (pm_runtime_suspended(dev)) {
-		DRM_DEBUG_KMS("Already suspended\n");
-		return 0;
-	}
-
-	hdmi_poweroff(hdata);
-
-	return 0;
-}
-
-static int hdmi_resume(struct device *dev)
-{
-	struct exynos_drm_hdmi_context *ctx = get_hdmi_context(dev);
-	struct hdmi_context *hdata = ctx->ctx;
-
-	hdata->hpd = gpio_get_value(hdata->hpd_gpio);
-
-	enable_irq(hdata->irq);
-
-	if (!pm_runtime_suspended(dev)) {
-		DRM_DEBUG_KMS("Already resumed\n");
-		return 0;
-	}
+err_hdmiphy:
+	if (hdata->hdmiphy_port)
+		put_device(&hdata->hdmiphy_port->dev);
+err_ddc:
+	put_device(&hdata->ddc_adpt->dev);
 
-	hdmi_poweron(hdata);
+err_del_component:
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
 
-	return 0;
+	return ret;
 }
-#endif
 
-#ifdef CONFIG_PM_RUNTIME
-static int hdmi_runtime_suspend(struct device *dev)
+static int hdmi_remove(struct platform_device *pdev)
 {
-	struct exynos_drm_hdmi_context *ctx = get_hdmi_context(dev);
-	struct hdmi_context *hdata = ctx->ctx;
+	struct hdmi_context *hdata = hdmi_display.ctx;
 
-	hdmi_poweroff(hdata);
+	cancel_delayed_work_sync(&hdata->hotplug_work);
 
-	return 0;
-}
-
-static int hdmi_runtime_resume(struct device *dev)
-{
-	struct exynos_drm_hdmi_context *ctx = get_hdmi_context(dev);
-	struct hdmi_context *hdata = ctx->ctx;
+	put_device(&hdata->hdmiphy_port->dev);
+	put_device(&hdata->ddc_adpt->dev);
 
-	hdmi_poweron(hdata);
+	pm_runtime_disable(&pdev->dev);
+	component_del(&pdev->dev, &hdmi_component_ops);
 
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CONNECTOR);
 	return 0;
 }
-#endif
-
-static const struct dev_pm_ops hdmi_pm_ops = {
-	SET_SYSTEM_SLEEP_PM_OPS(hdmi_suspend, hdmi_resume)
-	SET_RUNTIME_PM_OPS(hdmi_runtime_suspend, hdmi_runtime_resume, NULL)
-};
 
 struct platform_driver hdmi_driver = {
 	.probe		= hdmi_probe,
@@ -2092,7 +2510,6 @@ struct platform_driver hdmi_driver = {
 	.driver		= {
 		.name	= "exynos-hdmi",
 		.owner	= THIS_MODULE,
-		.pm	= &hdmi_pm_ops,
 		.of_match_table = hdmi_match_types,
 	},
 };
diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.h b/drivers/gpu/drm/exynos/exynos_hdmi.h
deleted file mode 100644
index 0ddf3957de1..00000000000
--- a/drivers/gpu/drm/exynos/exynos_hdmi.h
+++ /dev/null
@@ -1,23 +0,0 @@
-/*
- *
- * Copyright (c) 2011 Samsung Electronics Co., Ltd.
- * Authors:
- *	Inki Dae <inki.dae@samsung.com>
- *	Seung-Woo Kim <sw0312.kim@samsung.com>
- *
- * This program is free software; you can redistribute  it and/or modify it
- * under  the terms of  the GNU General  Public License as published by the
- * Free Software Foundation;  either version 2 of the  License, or (at your
- * option) any later version.
- */
-
-#ifndef _EXYNOS_HDMI_H_
-#define _EXYNOS_HDMI_H_
-
-void hdmi_attach_ddc_client(struct i2c_client *ddc);
-void hdmi_attach_hdmiphy_client(struct i2c_client *hdmiphy);
-
-extern struct i2c_driver hdmiphy_driver;
-extern struct i2c_driver ddc_driver;
-
-#endif
diff --git a/drivers/gpu/drm/exynos/exynos_hdmiphy.c b/drivers/gpu/drm/exynos/exynos_hdmiphy.c
deleted file mode 100644
index 59abb1494ce..00000000000
--- a/drivers/gpu/drm/exynos/exynos_hdmiphy.c
+++ /dev/null
@@ -1,65 +0,0 @@
-/*
- * Copyright (C) 2011 Samsung Electronics Co.Ltd
- * Authors:
- *	Seung-Woo Kim <sw0312.kim@samsung.com>
- *	Inki Dae <inki.dae@samsung.com>
- *
- * This program is free software; you can redistribute  it and/or modify it
- * under  the terms of  the GNU General  Public License as published by the
- * Free Software Foundation;  either version 2 of the  License, or (at your
- * option) any later version.
- *
- */
-
-#include <drm/drmP.h>
-
-#include <linux/kernel.h>
-#include <linux/i2c.h>
-#include <linux/of.h>
-
-#include "exynos_drm_drv.h"
-#include "exynos_hdmi.h"
-
-
-static int hdmiphy_probe(struct i2c_client *client,
-	const struct i2c_device_id *id)
-{
-	hdmi_attach_hdmiphy_client(client);
-
-	dev_info(&client->adapter->dev, "attached s5p_hdmiphy "
-		"into i2c adapter successfully\n");
-
-	return 0;
-}
-
-static int hdmiphy_remove(struct i2c_client *client)
-{
-	dev_info(&client->adapter->dev, "detached s5p_hdmiphy "
-		"from i2c adapter successfully\n");
-
-	return 0;
-}
-
-static struct of_device_id hdmiphy_match_types[] = {
-	{
-		.compatible = "samsung,exynos5-hdmiphy",
-	}, {
-		.compatible = "samsung,exynos4210-hdmiphy",
-	}, {
-		.compatible = "samsung,exynos4212-hdmiphy",
-	}, {
-		/* end node */
-	}
-};
-
-struct i2c_driver hdmiphy_driver = {
-	.driver = {
-		.name	= "exynos-hdmiphy",
-		.owner	= THIS_MODULE,
-		.of_match_table = hdmiphy_match_types,
-	},
-	.probe		= hdmiphy_probe,
-	.remove		= hdmiphy_remove,
-	.command		= NULL,
-};
-EXPORT_SYMBOL(hdmiphy_driver);
diff --git a/drivers/gpu/drm/exynos/exynos_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c
index 63bc5f92fbb..7529946d0a7 100644
--- a/drivers/gpu/drm/exynos/exynos_mixer.c
+++ b/drivers/gpu/drm/exynos/exynos_mixer.c
@@ -31,15 +31,19 @@
 #include <linux/clk.h>
 #include <linux/regulator/consumer.h>
 #include <linux/of.h>
+#include <linux/component.h>
 
 #include <drm/exynos_drm.h>
 
 #include "exynos_drm_drv.h"
 #include "exynos_drm_crtc.h"
-#include "exynos_drm_hdmi.h"
 #include "exynos_drm_iommu.h"
+#include "exynos_mixer.h"
 
-#define get_mixer_context(dev)	platform_get_drvdata(to_platform_device(dev))
+#define get_mixer_manager(dev)	platform_get_drvdata(to_platform_device(dev))
+
+#define MIXER_WIN_NR		3
+#define MIXER_DEFAULT_WIN	0
 
 struct hdmi_win_data {
 	dma_addr_t		dma_addr;
@@ -82,6 +86,7 @@ enum mixer_version_id {
 };
 
 struct mixer_context {
+	struct platform_device *pdev;
 	struct device		*dev;
 	struct drm_device	*drm_dev;
 	int			pipe;
@@ -94,7 +99,6 @@ struct mixer_context {
 	struct mixer_resources	mixer_res;
 	struct hdmi_win_data	win_data[MIXER_WIN_NR];
 	enum mixer_version_id	mxr_ver;
-	void			*parent_ctx;
 	wait_queue_head_t	wait_vsync_queue;
 	atomic_t		wait_vsync_event;
 };
@@ -373,6 +377,20 @@ static void mixer_run(struct mixer_context *ctx)
 	mixer_regs_dump(ctx);
 }
 
+static void mixer_stop(struct mixer_context *ctx)
+{
+	struct mixer_resources *res = &ctx->mixer_res;
+	int timeout = 20;
+
+	mixer_reg_writemask(res, MXR_STATUS, 0, MXR_STATUS_REG_RUN);
+
+	while (!(mixer_reg_read(res, MXR_STATUS) & MXR_STATUS_REG_IDLE) &&
+			--timeout)
+		usleep_range(10000, 12000);
+
+	mixer_regs_dump(ctx);
+}
+
 static void vp_video_buffer(struct mixer_context *ctx, int win)
 {
 	struct mixer_resources *res = &ctx->mixer_res;
@@ -493,13 +511,8 @@ static void vp_video_buffer(struct mixer_context *ctx, int win)
 static void mixer_layer_update(struct mixer_context *ctx)
 {
 	struct mixer_resources *res = &ctx->mixer_res;
-	u32 val;
-
-	val = mixer_reg_read(res, MXR_CFG);
 
-	/* allow one update per vsync only */
-	if (!(val & MXR_CFG_LAYER_UPDATE_COUNT_MASK))
-		mixer_reg_writemask(res, MXR_CFG, ~0, MXR_CFG_LAYER_UPDATE);
+	mixer_reg_writemask(res, MXR_CFG, ~0, MXR_CFG_LAYER_UPDATE);
 }
 
 static void mixer_graph_buffer(struct mixer_context *ctx, int win)
@@ -685,30 +698,197 @@ static void mixer_win_reset(struct mixer_context *ctx)
 	spin_unlock_irqrestore(&res->reg_slock, flags);
 }
 
-static int mixer_iommu_on(void *ctx, bool enable)
+static irqreturn_t mixer_irq_handler(int irq, void *arg)
+{
+	struct mixer_context *ctx = arg;
+	struct mixer_resources *res = &ctx->mixer_res;
+	u32 val, base, shadow;
+
+	spin_lock(&res->reg_slock);
+
+	/* read interrupt status for handling and clearing flags for VSYNC */
+	val = mixer_reg_read(res, MXR_INT_STATUS);
+
+	/* handling VSYNC */
+	if (val & MXR_INT_STATUS_VSYNC) {
+		/* interlace scan need to check shadow register */
+		if (ctx->interlace) {
+			base = mixer_reg_read(res, MXR_GRAPHIC_BASE(0));
+			shadow = mixer_reg_read(res, MXR_GRAPHIC_BASE_S(0));
+			if (base != shadow)
+				goto out;
+
+			base = mixer_reg_read(res, MXR_GRAPHIC_BASE(1));
+			shadow = mixer_reg_read(res, MXR_GRAPHIC_BASE_S(1));
+			if (base != shadow)
+				goto out;
+		}
+
+		drm_handle_vblank(ctx->drm_dev, ctx->pipe);
+		exynos_drm_crtc_finish_pageflip(ctx->drm_dev, ctx->pipe);
+
+		/* set wait vsync event to zero and wake up queue. */
+		if (atomic_read(&ctx->wait_vsync_event)) {
+			atomic_set(&ctx->wait_vsync_event, 0);
+			wake_up(&ctx->wait_vsync_queue);
+		}
+	}
+
+out:
+	/* clear interrupts */
+	if (~val & MXR_INT_EN_VSYNC) {
+		/* vsync interrupt use different bit for read and clear */
+		val &= ~MXR_INT_EN_VSYNC;
+		val |= MXR_INT_CLEAR_VSYNC;
+	}
+	mixer_reg_write(res, MXR_INT_STATUS, val);
+
+	spin_unlock(&res->reg_slock);
+
+	return IRQ_HANDLED;
+}
+
+static int mixer_resources_init(struct mixer_context *mixer_ctx)
 {
-	struct exynos_drm_hdmi_context *drm_hdmi_ctx;
-	struct mixer_context *mdata = ctx;
-	struct drm_device *drm_dev;
+	struct device *dev = &mixer_ctx->pdev->dev;
+	struct mixer_resources *mixer_res = &mixer_ctx->mixer_res;
+	struct resource *res;
+	int ret;
+
+	spin_lock_init(&mixer_res->reg_slock);
+
+	mixer_res->mixer = devm_clk_get(dev, "mixer");
+	if (IS_ERR(mixer_res->mixer)) {
+		dev_err(dev, "failed to get clock 'mixer'\n");
+		return -ENODEV;
+	}
+
+	mixer_res->sclk_hdmi = devm_clk_get(dev, "sclk_hdmi");
+	if (IS_ERR(mixer_res->sclk_hdmi)) {
+		dev_err(dev, "failed to get clock 'sclk_hdmi'\n");
+		return -ENODEV;
+	}
+	res = platform_get_resource(mixer_ctx->pdev, IORESOURCE_MEM, 0);
+	if (res == NULL) {
+		dev_err(dev, "get memory resource failed.\n");
+		return -ENXIO;
+	}
 
-	drm_hdmi_ctx = mdata->parent_ctx;
-	drm_dev = drm_hdmi_ctx->drm_dev;
+	mixer_res->mixer_regs = devm_ioremap(dev, res->start,
+							resource_size(res));
+	if (mixer_res->mixer_regs == NULL) {
+		dev_err(dev, "register mapping failed.\n");
+		return -ENXIO;
+	}
 
-	if (is_drm_iommu_supported(drm_dev)) {
-		if (enable)
-			return drm_iommu_attach_device(drm_dev, mdata->dev);
+	res = platform_get_resource(mixer_ctx->pdev, IORESOURCE_IRQ, 0);
+	if (res == NULL) {
+		dev_err(dev, "get interrupt resource failed.\n");
+		return -ENXIO;
+	}
 
-		drm_iommu_detach_device(drm_dev, mdata->dev);
+	ret = devm_request_irq(dev, res->start, mixer_irq_handler,
+						0, "drm_mixer", mixer_ctx);
+	if (ret) {
+		dev_err(dev, "request interrupt failed.\n");
+		return ret;
 	}
+	mixer_res->irq = res->start;
+
 	return 0;
 }
 
-static int mixer_enable_vblank(void *ctx, int pipe)
+static int vp_resources_init(struct mixer_context *mixer_ctx)
 {
-	struct mixer_context *mixer_ctx = ctx;
+	struct device *dev = &mixer_ctx->pdev->dev;
+	struct mixer_resources *mixer_res = &mixer_ctx->mixer_res;
+	struct resource *res;
+
+	mixer_res->vp = devm_clk_get(dev, "vp");
+	if (IS_ERR(mixer_res->vp)) {
+		dev_err(dev, "failed to get clock 'vp'\n");
+		return -ENODEV;
+	}
+	mixer_res->sclk_mixer = devm_clk_get(dev, "sclk_mixer");
+	if (IS_ERR(mixer_res->sclk_mixer)) {
+		dev_err(dev, "failed to get clock 'sclk_mixer'\n");
+		return -ENODEV;
+	}
+	mixer_res->sclk_dac = devm_clk_get(dev, "sclk_dac");
+	if (IS_ERR(mixer_res->sclk_dac)) {
+		dev_err(dev, "failed to get clock 'sclk_dac'\n");
+		return -ENODEV;
+	}
+
+	if (mixer_res->sclk_hdmi)
+		clk_set_parent(mixer_res->sclk_mixer, mixer_res->sclk_hdmi);
+
+	res = platform_get_resource(mixer_ctx->pdev, IORESOURCE_MEM, 1);
+	if (res == NULL) {
+		dev_err(dev, "get memory resource failed.\n");
+		return -ENXIO;
+	}
+
+	mixer_res->vp_regs = devm_ioremap(dev, res->start,
+							resource_size(res));
+	if (mixer_res->vp_regs == NULL) {
+		dev_err(dev, "register mapping failed.\n");
+		return -ENXIO;
+	}
+
+	return 0;
+}
+
+static int mixer_initialize(struct exynos_drm_manager *mgr,
+			struct drm_device *drm_dev)
+{
+	int ret;
+	struct mixer_context *mixer_ctx = mgr->ctx;
+	struct exynos_drm_private *priv;
+	priv = drm_dev->dev_private;
+
+	mgr->drm_dev = mixer_ctx->drm_dev = drm_dev;
+	mgr->pipe = mixer_ctx->pipe = priv->pipe++;
+
+	/* acquire resources: regs, irqs, clocks */
+	ret = mixer_resources_init(mixer_ctx);
+	if (ret) {
+		DRM_ERROR("mixer_resources_init failed ret=%d\n", ret);
+		return ret;
+	}
+
+	if (mixer_ctx->vp_enabled) {
+		/* acquire vp resources: regs, irqs, clocks */
+		ret = vp_resources_init(mixer_ctx);
+		if (ret) {
+			DRM_ERROR("vp_resources_init failed ret=%d\n", ret);
+			return ret;
+		}
+	}
+
+	if (!is_drm_iommu_supported(mixer_ctx->drm_dev))
+		return 0;
+
+	return drm_iommu_attach_device(mixer_ctx->drm_dev, mixer_ctx->dev);
+}
+
+static void mixer_mgr_remove(struct exynos_drm_manager *mgr)
+{
+	struct mixer_context *mixer_ctx = mgr->ctx;
+
+	if (is_drm_iommu_supported(mixer_ctx->drm_dev))
+		drm_iommu_detach_device(mixer_ctx->drm_dev, mixer_ctx->dev);
+}
+
+static int mixer_enable_vblank(struct exynos_drm_manager *mgr)
+{
+	struct mixer_context *mixer_ctx = mgr->ctx;
 	struct mixer_resources *res = &mixer_ctx->mixer_res;
 
-	mixer_ctx->pipe = pipe;
+	if (!mixer_ctx->powered) {
+		mixer_ctx->int_en |= MXR_INT_EN_VSYNC;
+		return 0;
+	}
 
 	/* enable vsync interrupt */
 	mixer_reg_writemask(res, MXR_INT_EN, MXR_INT_EN_VSYNC,
@@ -717,19 +897,19 @@ static int mixer_enable_vblank(void *ctx, int pipe)
 	return 0;
 }
 
-static void mixer_disable_vblank(void *ctx)
+static void mixer_disable_vblank(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *mixer_ctx = ctx;
+	struct mixer_context *mixer_ctx = mgr->ctx;
 	struct mixer_resources *res = &mixer_ctx->mixer_res;
 
 	/* disable vsync interrupt */
 	mixer_reg_writemask(res, MXR_INT_EN, 0, MXR_INT_EN_VSYNC);
 }
 
-static void mixer_win_mode_set(void *ctx,
-			      struct exynos_drm_overlay *overlay)
+static void mixer_win_mode_set(struct exynos_drm_manager *mgr,
+			struct exynos_drm_overlay *overlay)
 {
-	struct mixer_context *mixer_ctx = ctx;
+	struct mixer_context *mixer_ctx = mgr->ctx;
 	struct hdmi_win_data *win_data;
 	int win;
 
@@ -778,9 +958,10 @@ static void mixer_win_mode_set(void *ctx,
 	win_data->scan_flags = overlay->scan_flag;
 }
 
-static void mixer_win_commit(void *ctx, int win)
+static void mixer_win_commit(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct mixer_context *mixer_ctx = ctx;
+	struct mixer_context *mixer_ctx = mgr->ctx;
+	int win = zpos == DEFAULT_ZPOS ? MIXER_DEFAULT_WIN : zpos;
 
 	DRM_DEBUG_KMS("win: %d\n", win);
 
@@ -799,10 +980,11 @@ static void mixer_win_commit(void *ctx, int win)
 	mixer_ctx->win_data[win].enabled = true;
 }
 
-static void mixer_win_disable(void *ctx, int win)
+static void mixer_win_disable(struct exynos_drm_manager *mgr, int zpos)
 {
-	struct mixer_context *mixer_ctx = ctx;
+	struct mixer_context *mixer_ctx = mgr->ctx;
 	struct mixer_resources *res = &mixer_ctx->mixer_res;
+	int win = zpos == DEFAULT_ZPOS ? MIXER_DEFAULT_WIN : zpos;
 	unsigned long flags;
 
 	DRM_DEBUG_KMS("win: %d\n", win);
@@ -826,32 +1008,9 @@ static void mixer_win_disable(void *ctx, int win)
 	mixer_ctx->win_data[win].enabled = false;
 }
 
-static int mixer_check_mode(void *ctx, struct drm_display_mode *mode)
+static void mixer_wait_for_vblank(struct exynos_drm_manager *mgr)
 {
-	struct mixer_context *mixer_ctx = ctx;
-	u32 w, h;
-
-	w = mode->hdisplay;
-	h = mode->vdisplay;
-
-	DRM_DEBUG_KMS("xres=%d, yres=%d, refresh=%d, intl=%d\n",
-		mode->hdisplay, mode->vdisplay, mode->vrefresh,
-		(mode->flags & DRM_MODE_FLAG_INTERLACE) ? 1 : 0);
-
-	if (mixer_ctx->mxr_ver == MXR_VER_0_0_0_16 ||
-		mixer_ctx->mxr_ver == MXR_VER_128_0_0_184)
-		return 0;
-
-	if ((w >= 464 && w <= 720 && h >= 261 && h <= 576) ||
-		(w >= 1024 && w <= 1280 && h >= 576 && h <= 720) ||
-		(w >= 1664 && w <= 1920 && h >= 936 && h <= 1080))
-		return 0;
-
-	return -EINVAL;
-}
-static void mixer_wait_for_vblank(void *ctx)
-{
-	struct mixer_context *mixer_ctx = ctx;
+	struct mixer_context *mixer_ctx = mgr->ctx;
 
 	mutex_lock(&mixer_ctx->mixer_mutex);
 	if (!mixer_ctx->powered) {
@@ -860,6 +1019,8 @@ static void mixer_wait_for_vblank(void *ctx)
 	}
 	mutex_unlock(&mixer_ctx->mixer_mutex);
 
+	drm_vblank_get(mgr->crtc->dev, mixer_ctx->pipe);
+
 	atomic_set(&mixer_ctx->wait_vsync_event, 1);
 
 	/*
@@ -868,25 +1029,29 @@ static void mixer_wait_for_vblank(void *ctx)
 	 */
 	if (!wait_event_timeout(mixer_ctx->wait_vsync_queue,
 				!atomic_read(&mixer_ctx->wait_vsync_event),
-				DRM_HZ/20))
+				HZ/20))
 		DRM_DEBUG_KMS("vblank wait timed out.\n");
+
+	drm_vblank_put(mgr->crtc->dev, mixer_ctx->pipe);
 }
 
-static void mixer_window_suspend(struct mixer_context *ctx)
+static void mixer_window_suspend(struct exynos_drm_manager *mgr)
 {
+	struct mixer_context *ctx = mgr->ctx;
 	struct hdmi_win_data *win_data;
 	int i;
 
 	for (i = 0; i < MIXER_WIN_NR; i++) {
 		win_data = &ctx->win_data[i];
 		win_data->resume = win_data->enabled;
-		mixer_win_disable(ctx, i);
+		mixer_win_disable(mgr, i);
 	}
-	mixer_wait_for_vblank(ctx);
+	mixer_wait_for_vblank(mgr);
 }
 
-static void mixer_window_resume(struct mixer_context *ctx)
+static void mixer_window_resume(struct exynos_drm_manager *mgr)
 {
+	struct mixer_context *ctx = mgr->ctx;
 	struct hdmi_win_data *win_data;
 	int i;
 
@@ -894,11 +1059,14 @@ static void mixer_window_resume(struct mixer_context *ctx)
 		win_data = &ctx->win_data[i];
 		win_data->enabled = win_data->resume;
 		win_data->resume = false;
+		if (win_data->enabled)
+			mixer_win_commit(mgr, i);
 	}
 }
 
-static void mixer_poweron(struct mixer_context *ctx)
+static void mixer_poweron(struct exynos_drm_manager *mgr)
 {
+	struct mixer_context *ctx = mgr->ctx;
 	struct mixer_resources *res = &ctx->mixer_res;
 
 	mutex_lock(&ctx->mixer_mutex);
@@ -906,61 +1074,69 @@ static void mixer_poweron(struct mixer_context *ctx)
 		mutex_unlock(&ctx->mixer_mutex);
 		return;
 	}
-	ctx->powered = true;
+
 	mutex_unlock(&ctx->mixer_mutex);
 
+	pm_runtime_get_sync(ctx->dev);
+
 	clk_prepare_enable(res->mixer);
 	if (ctx->vp_enabled) {
 		clk_prepare_enable(res->vp);
 		clk_prepare_enable(res->sclk_mixer);
 	}
 
+	mutex_lock(&ctx->mixer_mutex);
+	ctx->powered = true;
+	mutex_unlock(&ctx->mixer_mutex);
+
+	mixer_reg_writemask(res, MXR_STATUS, ~0, MXR_STATUS_SOFT_RESET);
+
 	mixer_reg_write(res, MXR_INT_EN, ctx->int_en);
 	mixer_win_reset(ctx);
 
-	mixer_window_resume(ctx);
+	mixer_window_resume(mgr);
 }
 
-static void mixer_poweroff(struct mixer_context *ctx)
+static void mixer_poweroff(struct exynos_drm_manager *mgr)
 {
+	struct mixer_context *ctx = mgr->ctx;
 	struct mixer_resources *res = &ctx->mixer_res;
 
 	mutex_lock(&ctx->mixer_mutex);
-	if (!ctx->powered)
-		goto out;
+	if (!ctx->powered) {
+		mutex_unlock(&ctx->mixer_mutex);
+		return;
+	}
 	mutex_unlock(&ctx->mixer_mutex);
 
-	mixer_window_suspend(ctx);
+	mixer_stop(ctx);
+	mixer_window_suspend(mgr);
 
 	ctx->int_en = mixer_reg_read(res, MXR_INT_EN);
 
+	mutex_lock(&ctx->mixer_mutex);
+	ctx->powered = false;
+	mutex_unlock(&ctx->mixer_mutex);
+
 	clk_disable_unprepare(res->mixer);
 	if (ctx->vp_enabled) {
 		clk_disable_unprepare(res->vp);
 		clk_disable_unprepare(res->sclk_mixer);
 	}
 
-	mutex_lock(&ctx->mixer_mutex);
-	ctx->powered = false;
-
-out:
-	mutex_unlock(&ctx->mixer_mutex);
+	pm_runtime_put_sync(ctx->dev);
 }
 
-static void mixer_dpms(void *ctx, int mode)
+static void mixer_dpms(struct exynos_drm_manager *mgr, int mode)
 {
-	struct mixer_context *mixer_ctx = ctx;
-
 	switch (mode) {
 	case DRM_MODE_DPMS_ON:
-		if (pm_runtime_suspended(mixer_ctx->dev))
-			pm_runtime_get_sync(mixer_ctx->dev);
+		mixer_poweron(mgr);
 		break;
 	case DRM_MODE_DPMS_STANDBY:
 	case DRM_MODE_DPMS_SUSPEND:
 	case DRM_MODE_DPMS_OFF:
-		if (!pm_runtime_suspended(mixer_ctx->dev))
-			pm_runtime_put_sync(mixer_ctx->dev);
+		mixer_poweroff(mgr);
 		break;
 	default:
 		DRM_DEBUG_KMS("unknown dpms mode: %d\n", mode);
@@ -968,169 +1144,40 @@ static void mixer_dpms(void *ctx, int mode)
 	}
 }
 
-static struct exynos_mixer_ops mixer_ops = {
-	/* manager */
-	.iommu_on		= mixer_iommu_on,
-	.enable_vblank		= mixer_enable_vblank,
-	.disable_vblank		= mixer_disable_vblank,
-	.wait_for_vblank	= mixer_wait_for_vblank,
-	.dpms			= mixer_dpms,
-
-	/* overlay */
-	.win_mode_set		= mixer_win_mode_set,
-	.win_commit		= mixer_win_commit,
-	.win_disable		= mixer_win_disable,
-
-	/* display */
-	.check_mode		= mixer_check_mode,
-};
-
-static irqreturn_t mixer_irq_handler(int irq, void *arg)
+/* Only valid for Mixer version 16.0.33.0 */
+int mixer_check_mode(struct drm_display_mode *mode)
 {
-	struct exynos_drm_hdmi_context *drm_hdmi_ctx = arg;
-	struct mixer_context *ctx = drm_hdmi_ctx->ctx;
-	struct mixer_resources *res = &ctx->mixer_res;
-	u32 val, base, shadow;
-
-	spin_lock(&res->reg_slock);
-
-	/* read interrupt status for handling and clearing flags for VSYNC */
-	val = mixer_reg_read(res, MXR_INT_STATUS);
-
-	/* handling VSYNC */
-	if (val & MXR_INT_STATUS_VSYNC) {
-		/* interlace scan need to check shadow register */
-		if (ctx->interlace) {
-			base = mixer_reg_read(res, MXR_GRAPHIC_BASE(0));
-			shadow = mixer_reg_read(res, MXR_GRAPHIC_BASE_S(0));
-			if (base != shadow)
-				goto out;
-
-			base = mixer_reg_read(res, MXR_GRAPHIC_BASE(1));
-			shadow = mixer_reg_read(res, MXR_GRAPHIC_BASE_S(1));
-			if (base != shadow)
-				goto out;
-		}
-
-		drm_handle_vblank(drm_hdmi_ctx->drm_dev, ctx->pipe);
-		exynos_drm_crtc_finish_pageflip(drm_hdmi_ctx->drm_dev,
-				ctx->pipe);
-
-		/* set wait vsync event to zero and wake up queue. */
-		if (atomic_read(&ctx->wait_vsync_event)) {
-			atomic_set(&ctx->wait_vsync_event, 0);
-			DRM_WAKEUP(&ctx->wait_vsync_queue);
-		}
-	}
-
-out:
-	/* clear interrupts */
-	if (~val & MXR_INT_EN_VSYNC) {
-		/* vsync interrupt use different bit for read and clear */
-		val &= ~MXR_INT_EN_VSYNC;
-		val |= MXR_INT_CLEAR_VSYNC;
-	}
-	mixer_reg_write(res, MXR_INT_STATUS, val);
-
-	spin_unlock(&res->reg_slock);
-
-	return IRQ_HANDLED;
-}
-
-static int mixer_resources_init(struct exynos_drm_hdmi_context *ctx,
-				struct platform_device *pdev)
-{
-	struct mixer_context *mixer_ctx = ctx->ctx;
-	struct device *dev = &pdev->dev;
-	struct mixer_resources *mixer_res = &mixer_ctx->mixer_res;
-	struct resource *res;
-	int ret;
-
-	spin_lock_init(&mixer_res->reg_slock);
-
-	mixer_res->mixer = devm_clk_get(dev, "mixer");
-	if (IS_ERR(mixer_res->mixer)) {
-		dev_err(dev, "failed to get clock 'mixer'\n");
-		return -ENODEV;
-	}
-
-	mixer_res->sclk_hdmi = devm_clk_get(dev, "sclk_hdmi");
-	if (IS_ERR(mixer_res->sclk_hdmi)) {
-		dev_err(dev, "failed to get clock 'sclk_hdmi'\n");
-		return -ENODEV;
-	}
-	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (res == NULL) {
-		dev_err(dev, "get memory resource failed.\n");
-		return -ENXIO;
-	}
+	u32 w, h;
 
-	mixer_res->mixer_regs = devm_ioremap(dev, res->start,
-							resource_size(res));
-	if (mixer_res->mixer_regs == NULL) {
-		dev_err(dev, "register mapping failed.\n");
-		return -ENXIO;
-	}
+	w = mode->hdisplay;
+	h = mode->vdisplay;
 
-	res = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
-	if (res == NULL) {
-		dev_err(dev, "get interrupt resource failed.\n");
-		return -ENXIO;
-	}
+	DRM_DEBUG_KMS("xres=%d, yres=%d, refresh=%d, intl=%d\n",
+		mode->hdisplay, mode->vdisplay, mode->vrefresh,
+		(mode->flags & DRM_MODE_FLAG_INTERLACE) ? 1 : 0);
 
-	ret = devm_request_irq(dev, res->start, mixer_irq_handler,
-							0, "drm_mixer", ctx);
-	if (ret) {
-		dev_err(dev, "request interrupt failed.\n");
-		return ret;
-	}
-	mixer_res->irq = res->start;
+	if ((w >= 464 && w <= 720 && h >= 261 && h <= 576) ||
+		(w >= 1024 && w <= 1280 && h >= 576 && h <= 720) ||
+		(w >= 1664 && w <= 1920 && h >= 936 && h <= 1080))
+		return 0;
 
-	return 0;
+	return -EINVAL;
 }
 
-static int vp_resources_init(struct exynos_drm_hdmi_context *ctx,
-			     struct platform_device *pdev)
-{
-	struct mixer_context *mixer_ctx = ctx->ctx;
-	struct device *dev = &pdev->dev;
-	struct mixer_resources *mixer_res = &mixer_ctx->mixer_res;
-	struct resource *res;
-
-	mixer_res->vp = devm_clk_get(dev, "vp");
-	if (IS_ERR(mixer_res->vp)) {
-		dev_err(dev, "failed to get clock 'vp'\n");
-		return -ENODEV;
-	}
-	mixer_res->sclk_mixer = devm_clk_get(dev, "sclk_mixer");
-	if (IS_ERR(mixer_res->sclk_mixer)) {
-		dev_err(dev, "failed to get clock 'sclk_mixer'\n");
-		return -ENODEV;
-	}
-	mixer_res->sclk_dac = devm_clk_get(dev, "sclk_dac");
-	if (IS_ERR(mixer_res->sclk_dac)) {
-		dev_err(dev, "failed to get clock 'sclk_dac'\n");
-		return -ENODEV;
-	}
-
-	if (mixer_res->sclk_hdmi)
-		clk_set_parent(mixer_res->sclk_mixer, mixer_res->sclk_hdmi);
-
-	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
-	if (res == NULL) {
-		dev_err(dev, "get memory resource failed.\n");
-		return -ENXIO;
-	}
-
-	mixer_res->vp_regs = devm_ioremap(dev, res->start,
-							resource_size(res));
-	if (mixer_res->vp_regs == NULL) {
-		dev_err(dev, "register mapping failed.\n");
-		return -ENXIO;
-	}
+static struct exynos_drm_manager_ops mixer_manager_ops = {
+	.dpms			= mixer_dpms,
+	.enable_vblank		= mixer_enable_vblank,
+	.disable_vblank		= mixer_disable_vblank,
+	.wait_for_vblank	= mixer_wait_for_vblank,
+	.win_mode_set		= mixer_win_mode_set,
+	.win_commit		= mixer_win_commit,
+	.win_disable		= mixer_win_disable,
+};
 
-	return 0;
-}
+static struct exynos_drm_manager mixer_manager = {
+	.type			= EXYNOS_DISPLAY_TYPE_HDMI,
+	.ops			= &mixer_manager_ops,
+};
 
 static struct mixer_drv_data exynos5420_mxr_drv_data = {
 	.version = MXR_VER_128_0_0_184,
@@ -1174,24 +1221,21 @@ static struct of_device_id mixer_match_types[] = {
 	}
 };
 
-static int mixer_probe(struct platform_device *pdev)
+static int mixer_bind(struct device *dev, struct device *manager, void *data)
 {
-	struct device *dev = &pdev->dev;
-	struct exynos_drm_hdmi_context *drm_hdmi_ctx;
+	struct platform_device *pdev = to_platform_device(dev);
+	struct drm_device *drm_dev = data;
 	struct mixer_context *ctx;
 	struct mixer_drv_data *drv;
 	int ret;
 
 	dev_info(dev, "probe start\n");
 
-	drm_hdmi_ctx = devm_kzalloc(dev, sizeof(*drm_hdmi_ctx),
-								GFP_KERNEL);
-	if (!drm_hdmi_ctx)
-		return -ENOMEM;
-
-	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
-	if (!ctx)
+	ctx = devm_kzalloc(&pdev->dev, sizeof(*ctx), GFP_KERNEL);
+	if (!ctx) {
+		DRM_ERROR("failed to alloc mixer context.\n");
 		return -ENOMEM;
+	}
 
 	mutex_init(&ctx->mixer_mutex);
 
@@ -1204,121 +1248,77 @@ static int mixer_probe(struct platform_device *pdev)
 			platform_get_device_id(pdev)->driver_data;
 	}
 
+	ctx->pdev = pdev;
 	ctx->dev = dev;
-	ctx->parent_ctx = (void *)drm_hdmi_ctx;
-	drm_hdmi_ctx->ctx = (void *)ctx;
 	ctx->vp_enabled = drv->is_vp_enabled;
 	ctx->mxr_ver = drv->version;
-	DRM_INIT_WAITQUEUE(&ctx->wait_vsync_queue);
+	init_waitqueue_head(&ctx->wait_vsync_queue);
 	atomic_set(&ctx->wait_vsync_event, 0);
 
-	platform_set_drvdata(pdev, drm_hdmi_ctx);
+	mixer_manager.ctx = ctx;
+	ret = mixer_initialize(&mixer_manager, drm_dev);
+	if (ret)
+		return ret;
 
-	/* acquire resources: regs, irqs, clocks */
-	ret = mixer_resources_init(drm_hdmi_ctx, pdev);
+	platform_set_drvdata(pdev, &mixer_manager);
+	ret = exynos_drm_crtc_create(&mixer_manager);
 	if (ret) {
-		DRM_ERROR("mixer_resources_init failed\n");
-		goto fail;
-	}
-
-	if (ctx->vp_enabled) {
-		/* acquire vp resources: regs, irqs, clocks */
-		ret = vp_resources_init(drm_hdmi_ctx, pdev);
-		if (ret) {
-			DRM_ERROR("vp_resources_init failed\n");
-			goto fail;
-		}
+		mixer_mgr_remove(&mixer_manager);
+		return ret;
 	}
 
-	/* attach mixer driver to common hdmi. */
-	exynos_mixer_drv_attach(drm_hdmi_ctx);
-
-	/* register specific callback point to common hdmi. */
-	exynos_mixer_ops_register(&mixer_ops);
-
 	pm_runtime_enable(dev);
 
 	return 0;
-
-
-fail:
-	dev_info(dev, "probe failed\n");
-	return ret;
 }
 
-static int mixer_remove(struct platform_device *pdev)
+static void mixer_unbind(struct device *dev, struct device *master, void *data)
 {
-	dev_info(&pdev->dev, "remove successful\n");
+	struct exynos_drm_manager *mgr = dev_get_drvdata(dev);
+	struct drm_crtc *crtc = mgr->crtc;
 
-	pm_runtime_disable(&pdev->dev);
+	dev_info(dev, "remove successful\n");
 
-	return 0;
-}
+	mixer_mgr_remove(mgr);
 
-#ifdef CONFIG_PM_SLEEP
-static int mixer_suspend(struct device *dev)
-{
-	struct exynos_drm_hdmi_context *drm_hdmi_ctx = get_mixer_context(dev);
-	struct mixer_context *ctx = drm_hdmi_ctx->ctx;
+	pm_runtime_disable(dev);
 
-	if (pm_runtime_suspended(dev)) {
-		DRM_DEBUG_KMS("Already suspended\n");
-		return 0;
-	}
-
-	mixer_poweroff(ctx);
-
-	return 0;
+	crtc->funcs->destroy(crtc);
 }
 
-static int mixer_resume(struct device *dev)
-{
-	struct exynos_drm_hdmi_context *drm_hdmi_ctx = get_mixer_context(dev);
-	struct mixer_context *ctx = drm_hdmi_ctx->ctx;
-
-	if (!pm_runtime_suspended(dev)) {
-		DRM_DEBUG_KMS("Already resumed\n");
-		return 0;
-	}
-
-	mixer_poweron(ctx);
-
-	return 0;
-}
-#endif
+static const struct component_ops mixer_component_ops = {
+	.bind	= mixer_bind,
+	.unbind	= mixer_unbind,
+};
 
-#ifdef CONFIG_PM_RUNTIME
-static int mixer_runtime_suspend(struct device *dev)
+static int mixer_probe(struct platform_device *pdev)
 {
-	struct exynos_drm_hdmi_context *drm_hdmi_ctx = get_mixer_context(dev);
-	struct mixer_context *ctx = drm_hdmi_ctx->ctx;
+	int ret;
 
-	mixer_poweroff(ctx);
+	ret = exynos_drm_component_add(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC,
+					mixer_manager.type);
+	if (ret)
+		return ret;
 
-	return 0;
+	ret = component_add(&pdev->dev, &mixer_component_ops);
+	if (ret)
+		exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
+
+	return ret;
 }
 
-static int mixer_runtime_resume(struct device *dev)
+static int mixer_remove(struct platform_device *pdev)
 {
-	struct exynos_drm_hdmi_context *drm_hdmi_ctx = get_mixer_context(dev);
-	struct mixer_context *ctx = drm_hdmi_ctx->ctx;
-
-	mixer_poweron(ctx);
+	component_del(&pdev->dev, &mixer_component_ops);
+	exynos_drm_component_del(&pdev->dev, EXYNOS_DEVICE_TYPE_CRTC);
 
 	return 0;
 }
-#endif
-
-static const struct dev_pm_ops mixer_pm_ops = {
-	SET_SYSTEM_SLEEP_PM_OPS(mixer_suspend, mixer_resume)
-	SET_RUNTIME_PM_OPS(mixer_runtime_suspend, mixer_runtime_resume, NULL)
-};
 
 struct platform_driver mixer_driver = {
 	.driver = {
 		.name = "exynos-mixer",
 		.owner = THIS_MODULE,
-		.pm = &mixer_pm_ops,
 		.of_match_table = mixer_match_types,
 	},
 	.probe = mixer_probe,
diff --git a/drivers/gpu/drm/exynos/exynos_mixer.h b/drivers/gpu/drm/exynos/exynos_mixer.h
new file mode 100644
index 00000000000..3811e417f0e
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_mixer.h
@@ -0,0 +1,20 @@
+/*
+ * Copyright (C) 2013 Google, Inc.
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef _EXYNOS_MIXER_H_
+#define _EXYNOS_MIXER_H_
+
+/* This function returns 0 if the given timing is valid for the mixer */
+int mixer_check_mode(struct drm_display_mode *mode);
+
+#endif
diff --git a/drivers/gpu/drm/exynos/regs-hdmi.h b/drivers/gpu/drm/exynos/regs-hdmi.h
index ef1b3eb3ba6..3f35ac6d8a4 100644
--- a/drivers/gpu/drm/exynos/regs-hdmi.h
+++ b/drivers/gpu/drm/exynos/regs-hdmi.h
@@ -578,4 +578,20 @@
 #define HDMI_TG_VACT_ST4_H		HDMI_TG_BASE(0x0074)
 #define HDMI_TG_3D			HDMI_TG_BASE(0x00F0)
 
+/* HDMI PHY Registers Offsets*/
+#define HDMIPHY_POWER		(0x74 >> 2)
+#define HDMIPHY_MODE_SET_DONE		(0x7c >> 2)
+
+/* HDMI PHY Values */
+#define HDMI_PHY_POWER_ON              0x80
+#define HDMI_PHY_POWER_OFF             0xff
+
+/* HDMI PHY Values */
+#define HDMI_PHY_DISABLE_MODE_SET	0x80
+#define HDMI_PHY_ENABLE_MODE_SET	0x00
+
+/* PMU Registers for PHY */
+#define PMU_HDMI_PHY_CONTROL		0x700
+#define PMU_HDMI_PHY_ENABLE_BIT		BIT(0)
+
 #endif /* SAMSUNG_REGS_HDMI_H */
diff --git a/drivers/gpu/drm/exynos/regs-mixer.h b/drivers/gpu/drm/exynos/regs-mixer.h
index 4537026bc38..5f32e1a2941 100644
--- a/drivers/gpu/drm/exynos/regs-mixer.h
+++ b/drivers/gpu/drm/exynos/regs-mixer.h
@@ -78,6 +78,7 @@
 #define MXR_STATUS_BIG_ENDIAN		(1 << 3)
 #define MXR_STATUS_ENDIAN_MASK		(1 << 3)
 #define MXR_STATUS_SYNC_ENABLE		(1 << 2)
+#define MXR_STATUS_REG_IDLE		(1 << 1)
 #define MXR_STATUS_REG_RUN		(1 << 0)
 
 /* bits for MXR_CFG */
diff --git a/drivers/gpu/drm/gma500/Kconfig b/drivers/gpu/drm/gma500/Kconfig
index 508cf99a292..17f928ec84e 100644
--- a/drivers/gpu/drm/gma500/Kconfig
+++ b/drivers/gpu/drm/gma500/Kconfig
@@ -10,7 +10,6 @@ config DRM_GMA500
 	# GMA500 depends on ACPI_VIDEO when ACPI is enabled, just like i915
 	select ACPI_VIDEO if ACPI
 	select BACKLIGHT_CLASS_DEVICE if ACPI
-	select VIDEO_OUTPUT_CONTROL if ACPI
 	select INPUT if ACPI
 	help
 	  Say yes for an experimental 2D KMS framebuffer driver for the
diff --git a/drivers/gpu/drm/gma500/Makefile b/drivers/gpu/drm/gma500/Makefile
index e9064dd9045..b1531557637 100644
--- a/drivers/gpu/drm/gma500/Makefile
+++ b/drivers/gpu/drm/gma500/Makefile
@@ -13,9 +13,11 @@ gma500_gfx-y += \
 	  intel_i2c.o \
 	  intel_gmbus.o \
 	  mmu.o \
+	  blitter.o \
 	  power.o \
 	  psb_drv.o \
 	  gma_display.o \
+	  gma_device.o \
 	  psb_intel_display.o \
 	  psb_intel_lvds.o \
 	  psb_intel_modes.o \
diff --git a/drivers/gpu/drm/gma500/accel_2d.c b/drivers/gpu/drm/gma500/accel_2d.c
index d5ef1a5793c..de6f62a6ceb 100644
--- a/drivers/gpu/drm/gma500/accel_2d.c
+++ b/drivers/gpu/drm/gma500/accel_2d.c
@@ -326,7 +326,7 @@ int psbfb_sync(struct fb_info *info)
 	struct psb_framebuffer *psbfb = &fbdev->pfb;
 	struct drm_device *dev = psbfb->base.dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
-	unsigned long _end = jiffies + DRM_HZ;
+	unsigned long _end = jiffies + HZ;
 	int busy = 0;
 	unsigned long flags;
 
diff --git a/drivers/gpu/drm/gma500/backlight.c b/drivers/gpu/drm/gma500/backlight.c
index 143eba3309c..ea7dfc59d79 100644
--- a/drivers/gpu/drm/gma500/backlight.c
+++ b/drivers/gpu/drm/gma500/backlight.c
@@ -26,13 +26,13 @@
 #include "intel_bios.h"
 #include "power.h"
 
+#ifdef CONFIG_BACKLIGHT_CLASS_DEVICE
 static void do_gma_backlight_set(struct drm_device *dev)
 {
-#ifdef CONFIG_BACKLIGHT_CLASS_DEVICE
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	backlight_update_status(dev_priv->backlight_device);
-#endif	
 }
+#endif
 
 void gma_backlight_enable(struct drm_device *dev)
 {
diff --git a/drivers/gpu/drm/gma500/blitter.c b/drivers/gpu/drm/gma500/blitter.c
new file mode 100644
index 00000000000..9cd54a6fb89
--- /dev/null
+++ b/drivers/gpu/drm/gma500/blitter.c
@@ -0,0 +1,51 @@
+/*
+ * Copyright (c) 2014, Patrik Jakobsson
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * Authors: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>
+ */
+
+#include "psb_drv.h"
+
+#include "blitter.h"
+#include "psb_reg.h"
+
+/* Wait for the blitter to be completely idle */
+int gma_blt_wait_idle(struct drm_psb_private *dev_priv)
+{
+	unsigned long stop = jiffies + HZ;
+	int busy = 1;
+
+	/* NOP for Cedarview */
+	if (IS_CDV(dev_priv->dev))
+		return 0;
+
+	/* First do a quick check */
+	if ((PSB_RSGX32(PSB_CR_2D_SOCIF) == _PSB_C2_SOCIF_EMPTY) &&
+	    ((PSB_RSGX32(PSB_CR_2D_BLIT_STATUS) & _PSB_C2B_STATUS_BUSY) == 0))
+		return 0;
+
+	do {
+		busy = (PSB_RSGX32(PSB_CR_2D_SOCIF) != _PSB_C2_SOCIF_EMPTY);
+	} while (busy && !time_after_eq(jiffies, stop));
+
+	if (busy)
+		return -EBUSY;
+
+	do {
+		busy = ((PSB_RSGX32(PSB_CR_2D_BLIT_STATUS) &
+			_PSB_C2B_STATUS_BUSY) != 0);
+	} while (busy && !time_after_eq(jiffies, stop));
+
+	/* If still busy, we probably have a hang */
+	return (busy) ? -EBUSY : 0;
+}
diff --git a/drivers/gpu/drm/gma500/blitter.h b/drivers/gpu/drm/gma500/blitter.h
new file mode 100644
index 00000000000..b83648df590
--- /dev/null
+++ b/drivers/gpu/drm/gma500/blitter.h
@@ -0,0 +1,22 @@
+/*
+ * Copyright (c) 2014, Patrik Jakobsson
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * Authors: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>
+ */
+
+#ifndef __BLITTER_H
+#define __BLITTER_H
+
+extern int gma_blt_wait_idle(struct drm_psb_private *dev_priv);
+
+#endif
diff --git a/drivers/gpu/drm/gma500/cdv_device.c b/drivers/gpu/drm/gma500/cdv_device.c
index 5a9a6a3063a..3531f90e53d 100644
--- a/drivers/gpu/drm/gma500/cdv_device.c
+++ b/drivers/gpu/drm/gma500/cdv_device.c
@@ -26,6 +26,7 @@
 #include "psb_intel_reg.h"
 #include "intel_bios.h"
 #include "cdv_device.h"
+#include "gma_device.h"
 
 #define VGA_SR_INDEX		0x3c4
 #define VGA_SR_DATA		0x3c5
@@ -426,43 +427,6 @@ static int cdv_power_up(struct drm_device *dev)
 	return 0;
 }
 
-/* FIXME ? - shared with Poulsbo */
-static void cdv_get_core_freq(struct drm_device *dev)
-{
-	uint32_t clock;
-	struct pci_dev *pci_root = pci_get_bus_and_slot(0, 0);
-	struct drm_psb_private *dev_priv = dev->dev_private;
-
-	pci_write_config_dword(pci_root, 0xD0, 0xD0050300);
-	pci_read_config_dword(pci_root, 0xD4, &clock);
-	pci_dev_put(pci_root);
-
-	switch (clock & 0x07) {
-	case 0:
-		dev_priv->core_freq = 100;
-		break;
-	case 1:
-		dev_priv->core_freq = 133;
-		break;
-	case 2:
-		dev_priv->core_freq = 150;
-		break;
-	case 3:
-		dev_priv->core_freq = 178;
-		break;
-	case 4:
-		dev_priv->core_freq = 200;
-		break;
-	case 5:
-	case 6:
-	case 7:
-		dev_priv->core_freq = 266;
-		break;
-	default:
-		dev_priv->core_freq = 0;
-	}
-}
-
 static void cdv_hotplug_work_func(struct work_struct *work)
 {
         struct drm_psb_private *dev_priv = container_of(work, struct drm_psb_private,
@@ -618,7 +582,7 @@ static int cdv_chip_setup(struct drm_device *dev)
 	if (pci_enable_msi(dev->pdev))
 		dev_warn(dev->dev, "Enabling MSI failed!\n");
 	dev_priv->regmap = cdv_regmap;
-	cdv_get_core_freq(dev);
+	gma_get_core_freq(dev);
 	psb_intel_opregion_init(dev);
 	psb_intel_init_bios(dev);
 	cdv_hotplug_enable(dev, false);
diff --git a/drivers/gpu/drm/gma500/cdv_intel_crt.c b/drivers/gpu/drm/gma500/cdv_intel_crt.c
index 661af492173..c18268cd516 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_crt.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_crt.c
@@ -81,13 +81,6 @@ static int cdv_intel_crt_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
-static bool cdv_intel_crt_mode_fixup(struct drm_encoder *encoder,
-				 const struct drm_display_mode *mode,
-				 struct drm_display_mode *adjusted_mode)
-{
-	return true;
-}
-
 static void cdv_intel_crt_mode_set(struct drm_encoder *encoder,
 			       struct drm_display_mode *mode,
 			       struct drm_display_mode *adjusted_mode)
@@ -224,7 +217,7 @@ static int cdv_intel_crt_set_property(struct drm_connector *connector,
 
 static const struct drm_encoder_helper_funcs cdv_intel_crt_helper_funcs = {
 	.dpms = cdv_intel_crt_dpms,
-	.mode_fixup = cdv_intel_crt_mode_fixup,
+	.mode_fixup = gma_encoder_mode_fixup,
 	.prepare = gma_encoder_prepare,
 	.commit = gma_encoder_commit,
 	.mode_set = cdv_intel_crt_mode_set,
diff --git a/drivers/gpu/drm/gma500/cdv_intel_display.c b/drivers/gpu/drm/gma500/cdv_intel_display.c
index 8fbfa06da62..66727328832 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_display.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_display.c
@@ -412,8 +412,11 @@ static bool cdv_intel_find_dp_pll(const struct gma_limit_t *limit,
 				  int refclk,
 				  struct gma_clock_t *best_clock)
 {
+	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
 	struct gma_clock_t clock;
-	if (refclk == 27000) {
+
+	switch (refclk) {
+	case 27000:
 		if (target < 200000) {
 			clock.p1 = 2;
 			clock.p2 = 10;
@@ -427,7 +430,9 @@ static bool cdv_intel_find_dp_pll(const struct gma_limit_t *limit,
 			clock.m1 = 0;
 			clock.m2 = 98;
 		}
-	} else if (refclk == 100000) {
+		break;
+
+	case 100000:
 		if (target < 200000) {
 			clock.p1 = 2;
 			clock.p2 = 10;
@@ -441,12 +446,13 @@ static bool cdv_intel_find_dp_pll(const struct gma_limit_t *limit,
 			clock.m1 = 0;
 			clock.m2 = 133;
 		}
-	} else
+		break;
+
+	default:
 		return false;
-	clock.m = clock.m2 + 2;
-	clock.p = clock.p1 * clock.p2;
-	clock.vco = (refclk * clock.m) / clock.n;
-	clock.dot = clock.vco / clock.p;
+	}
+
+	gma_crtc->clock_funcs->clock(refclk, &clock);
 	memcpy(best_clock, &clock, sizeof(struct gma_clock_t));
 	return true;
 }
@@ -463,54 +469,11 @@ static bool cdv_intel_pipe_enabled(struct drm_device *dev, int pipe)
 	crtc = dev_priv->pipe_to_crtc_mapping[pipe];
 	gma_crtc = to_gma_crtc(crtc);
 
-	if (crtc->fb == NULL || !gma_crtc->active)
+	if (crtc->primary->fb == NULL || !gma_crtc->active)
 		return false;
 	return true;
 }
 
-static bool cdv_intel_single_pipe_active (struct drm_device *dev)
-{
-	uint32_t pipe_enabled = 0;
-
-	if (cdv_intel_pipe_enabled(dev, 0))
-		pipe_enabled |= FIFO_PIPEA;
-
-	if (cdv_intel_pipe_enabled(dev, 1))
-		pipe_enabled |= FIFO_PIPEB;
-
-
-	DRM_DEBUG_KMS("pipe enabled %x\n", pipe_enabled);
-
-	if (pipe_enabled == FIFO_PIPEA || pipe_enabled == FIFO_PIPEB)
-		return true;
-	else
-		return false;
-}
-
-static bool is_pipeb_lvds(struct drm_device *dev, struct drm_crtc *crtc)
-{
-	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
-	struct drm_mode_config *mode_config = &dev->mode_config;
-	struct drm_connector *connector;
-
-	if (gma_crtc->pipe != 1)
-		return false;
-
-	list_for_each_entry(connector, &mode_config->connector_list, head) {
-		struct gma_encoder *gma_encoder =
-					gma_attached_encoder(connector);
-
-		if (!connector->encoder
-		    || connector->encoder->crtc != crtc)
-			continue;
-
-		if (gma_encoder->type == INTEL_OUTPUT_LVDS)
-			return true;
-	}
-
-	return false;
-}
-
 void cdv_disable_sr(struct drm_device *dev)
 {
 	if (REG_READ(FW_BLC_SELF) & FW_BLC_SELF_EN) {
@@ -535,8 +498,10 @@ void cdv_disable_sr(struct drm_device *dev)
 void cdv_update_wm(struct drm_device *dev, struct drm_crtc *crtc)
 {
 	struct drm_psb_private *dev_priv = dev->dev_private;
+	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
 
-	if (cdv_intel_single_pipe_active(dev)) {
+	/* Is only one pipe enabled? */
+	if (cdv_intel_pipe_enabled(dev, 0) ^ cdv_intel_pipe_enabled(dev, 1)) {
 		u32 fw;
 
 		fw = REG_READ(DSPFW1);
@@ -557,7 +522,9 @@ void cdv_update_wm(struct drm_device *dev, struct drm_crtc *crtc)
 
 		/* ignore FW4 */
 
-		if (is_pipeb_lvds(dev, crtc)) {
+		/* Is pipe b lvds ? */
+		if (gma_crtc->pipe == 1 &&
+		    gma_pipe_has_type(crtc, INTEL_OUTPUT_LVDS)) {
 			REG_WRITE(DSPFW5, 0x00040330);
 		} else {
 			fw = (3 << DSP_PLANE_B_FIFO_WM1_SHIFT) |
diff --git a/drivers/gpu/drm/gma500/cdv_intel_dp.c b/drivers/gpu/drm/gma500/cdv_intel_dp.c
index f88a1815d87..9ff30c2efad 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_dp.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_dp.c
@@ -483,7 +483,7 @@ cdv_intel_dp_aux_native_write(struct gma_encoder *encoder,
 
 	if (send_bytes > 16)
 		return -1;
-	msg[0] = AUX_NATIVE_WRITE << 4;
+	msg[0] = DP_AUX_NATIVE_WRITE << 4;
 	msg[1] = address >> 8;
 	msg[2] = address & 0xff;
 	msg[3] = send_bytes - 1;
@@ -493,9 +493,10 @@ cdv_intel_dp_aux_native_write(struct gma_encoder *encoder,
 		ret = cdv_intel_dp_aux_ch(encoder, msg, msg_bytes, &ack, 1);
 		if (ret < 0)
 			return ret;
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK)
+		ack >>= 4;
+		if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_ACK)
 			break;
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
+		else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER)
 			udelay(100);
 		else
 			return -EIO;
@@ -523,7 +524,7 @@ cdv_intel_dp_aux_native_read(struct gma_encoder *encoder,
 	uint8_t ack;
 	int ret;
 
-	msg[0] = AUX_NATIVE_READ << 4;
+	msg[0] = DP_AUX_NATIVE_READ << 4;
 	msg[1] = address >> 8;
 	msg[2] = address & 0xff;
 	msg[3] = recv_bytes - 1;
@@ -538,12 +539,12 @@ cdv_intel_dp_aux_native_read(struct gma_encoder *encoder,
 			return -EPROTO;
 		if (ret < 0)
 			return ret;
-		ack = reply[0];
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK) {
+		ack = reply[0] >> 4;
+		if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_ACK) {
 			memcpy(recv, reply + 1, ret - 1);
 			return ret - 1;
 		}
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
+		else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER)
 			udelay(100);
 		else
 			return -EIO;
@@ -569,12 +570,12 @@ cdv_intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 
 	/* Set up the command byte */
 	if (mode & MODE_I2C_READ)
-		msg[0] = AUX_I2C_READ << 4;
+		msg[0] = DP_AUX_I2C_READ << 4;
 	else
-		msg[0] = AUX_I2C_WRITE << 4;
+		msg[0] = DP_AUX_I2C_WRITE << 4;
 
 	if (!(mode & MODE_I2C_STOP))
-		msg[0] |= AUX_I2C_MOT << 4;
+		msg[0] |= DP_AUX_I2C_MOT << 4;
 
 	msg[1] = address >> 8;
 	msg[2] = address;
@@ -606,16 +607,16 @@ cdv_intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 			return ret;
 		}
 
-		switch (reply[0] & AUX_NATIVE_REPLY_MASK) {
-		case AUX_NATIVE_REPLY_ACK:
+		switch ((reply[0] >> 4) & DP_AUX_NATIVE_REPLY_MASK) {
+		case DP_AUX_NATIVE_REPLY_ACK:
 			/* I2C-over-AUX Reply field is only valid
 			 * when paired with AUX ACK.
 			 */
 			break;
-		case AUX_NATIVE_REPLY_NACK:
+		case DP_AUX_NATIVE_REPLY_NACK:
 			DRM_DEBUG_KMS("aux_ch native nack\n");
 			return -EREMOTEIO;
-		case AUX_NATIVE_REPLY_DEFER:
+		case DP_AUX_NATIVE_REPLY_DEFER:
 			udelay(100);
 			continue;
 		default:
@@ -624,16 +625,16 @@ cdv_intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 			return -EREMOTEIO;
 		}
 
-		switch (reply[0] & AUX_I2C_REPLY_MASK) {
-		case AUX_I2C_REPLY_ACK:
+		switch ((reply[0] >> 4) & DP_AUX_I2C_REPLY_MASK) {
+		case DP_AUX_I2C_REPLY_ACK:
 			if (mode == MODE_I2C_READ) {
 				*read_byte = reply[1];
 			}
 			return reply_bytes - 1;
-		case AUX_I2C_REPLY_NACK:
+		case DP_AUX_I2C_REPLY_NACK:
 			DRM_DEBUG_KMS("aux_i2c nack\n");
 			return -EREMOTEIO;
-		case AUX_I2C_REPLY_DEFER:
+		case DP_AUX_I2C_REPLY_DEFER:
 			DRM_DEBUG_KMS("aux_i2c defer\n");
 			udelay(100);
 			break;
@@ -677,7 +678,7 @@ cdv_intel_dp_i2c_init(struct gma_connector *connector,
 	return ret;
 }
 
-void cdv_intel_fixed_panel_mode(struct drm_display_mode *fixed_mode,
+static void cdv_intel_fixed_panel_mode(struct drm_display_mode *fixed_mode,
 	struct drm_display_mode *adjusted_mode)
 {
 	adjusted_mode->hdisplay = fixed_mode->hdisplay;
@@ -1692,7 +1693,7 @@ done:
 		struct drm_crtc *crtc = encoder->base.crtc;
 		drm_crtc_helper_set_mode(crtc, &crtc->mode,
 					 crtc->x, crtc->y,
-					 crtc->fb);
+					 crtc->primary->fb);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/gma500/cdv_intel_hdmi.c b/drivers/gpu/drm/gma500/cdv_intel_hdmi.c
index 1c0d723b8d2..b99084b3f70 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_hdmi.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_hdmi.c
@@ -89,13 +89,6 @@ static void cdv_hdmi_mode_set(struct drm_encoder *encoder,
 	REG_READ(hdmi_priv->hdmi_reg);
 }
 
-static bool cdv_hdmi_mode_fixup(struct drm_encoder *encoder,
-				  const struct drm_display_mode *mode,
-				  struct drm_display_mode *adjusted_mode)
-{
-	return true;
-}
-
 static void cdv_hdmi_dpms(struct drm_encoder *encoder, int mode)
 {
 	struct drm_device *dev = encoder->dev;
@@ -199,7 +192,7 @@ static int cdv_hdmi_set_property(struct drm_connector *connector,
 		    crtc->saved_mode.vdisplay != 0) {
 			if (centre) {
 				if (!drm_crtc_helper_set_mode(encoder->crtc, &crtc->saved_mode,
-					    encoder->crtc->x, encoder->crtc->y, encoder->crtc->fb))
+					    encoder->crtc->x, encoder->crtc->y, encoder->crtc->primary->fb))
 					return -1;
 			} else {
 				struct drm_encoder_helper_funcs *helpers
@@ -262,7 +255,7 @@ static void cdv_hdmi_destroy(struct drm_connector *connector)
 
 static const struct drm_encoder_helper_funcs cdv_hdmi_helper_funcs = {
 	.dpms = cdv_hdmi_dpms,
-	.mode_fixup = cdv_hdmi_mode_fixup,
+	.mode_fixup = gma_encoder_mode_fixup,
 	.prepare = gma_encoder_prepare,
 	.mode_set = cdv_hdmi_mode_set,
 	.commit = gma_encoder_commit,
diff --git a/drivers/gpu/drm/gma500/cdv_intel_lvds.c b/drivers/gpu/drm/gma500/cdv_intel_lvds.c
index 20e08e65d46..8ecc920fc26 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_lvds.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_lvds.c
@@ -494,7 +494,7 @@ static int cdv_intel_lvds_set_property(struct drm_connector *connector,
 						      &crtc->saved_mode,
 						      encoder->crtc->x,
 						      encoder->crtc->y,
-						      encoder->crtc->fb))
+						      encoder->crtc->primary->fb))
 				return -1;
 		}
 	} else if (!strcmp(property->name, "backlight") && encoder) {
@@ -712,6 +712,7 @@ void cdv_intel_lvds_init(struct drm_device *dev,
 	 * Attempt to get the fixed panel mode from DDC.  Assume that the
 	 * preferred mode is the right one.
 	 */
+	mutex_lock(&dev->mode_config.mutex);
 	psb_intel_ddc_get_modes(connector,
 				&gma_encoder->ddc_bus->adapter);
 	list_for_each_entry(scan, &connector->probed_modes, head) {
@@ -772,10 +773,12 @@ void cdv_intel_lvds_init(struct drm_device *dev,
 	}
 
 out:
+	mutex_unlock(&dev->mode_config.mutex);
 	drm_sysfs_connector_add(connector);
 	return;
 
 failed_find:
+	mutex_unlock(&dev->mode_config.mutex);
 	printk(KERN_ERR "Failed find\n");
 	if (gma_encoder->ddc_bus)
 		psb_intel_i2c_destroy(gma_encoder->ddc_bus);
diff --git a/drivers/gpu/drm/gma500/framebuffer.c b/drivers/gpu/drm/gma500/framebuffer.c
index 94b3fec22c2..e7fcc148f33 100644
--- a/drivers/gpu/drm/gma500/framebuffer.c
+++ b/drivers/gpu/drm/gma500/framebuffer.c
@@ -319,7 +319,7 @@ static struct gtt_range *psbfb_alloc(struct drm_device *dev, int aligned_size)
 {
 	struct gtt_range *backing;
 	/* Begin by trying to use stolen memory backing */
-	backing = psb_gtt_alloc_range(dev, aligned_size, "fb", 1);
+	backing = psb_gtt_alloc_range(dev, aligned_size, "fb", 1, PAGE_SIZE);
 	if (backing) {
 		drm_gem_private_object_init(dev, &backing->gem, aligned_size);
 		return backing;
diff --git a/drivers/gpu/drm/gma500/gem.c b/drivers/gpu/drm/gma500/gem.c
index e2db48a81ed..c707fa6fca8 100644
--- a/drivers/gpu/drm/gma500/gem.c
+++ b/drivers/gpu/drm/gma500/gem.c
@@ -62,9 +62,6 @@ int psb_gem_dumb_map_gtt(struct drm_file *file, struct drm_device *dev,
 	int ret = 0;
 	struct drm_gem_object *obj;
 
-	if (!(dev->driver->driver_features & DRIVER_GEM))
-		return -ENODEV;
-
 	mutex_lock(&dev->struct_mutex);
 
 	/* GEM does all our handle to object mapping */
@@ -98,8 +95,8 @@ unlock:
  *	it so that userspace can speak about it. This does the core work
  *	for the various methods that do/will create GEM objects for things
  */
-static int psb_gem_create(struct drm_file *file,
-	struct drm_device *dev, uint64_t size, uint32_t *handlep)
+int psb_gem_create(struct drm_file *file, struct drm_device *dev, u64 size,
+		   u32 *handlep, int stolen, u32 align)
 {
 	struct gtt_range *r;
 	int ret;
@@ -109,7 +106,7 @@ static int psb_gem_create(struct drm_file *file,
 
 	/* Allocate our object - for now a direct gtt range which is not
 	   stolen memory backed */
-	r = psb_gtt_alloc_range(dev, size, "gem", 0);
+	r = psb_gtt_alloc_range(dev, size, "gem", 0, PAGE_SIZE);
 	if (r == NULL) {
 		dev_err(dev->dev, "no memory for %lld byte GEM object\n", size);
 		return -ENOSPC;
@@ -153,7 +150,8 @@ int psb_gem_dumb_create(struct drm_file *file, struct drm_device *dev,
 {
 	args->pitch = ALIGN(args->width * ((args->bpp + 7) / 8), 64);
 	args->size = args->pitch * args->height;
-	return psb_gem_create(file, dev, args->size, &args->handle);
+	return psb_gem_create(file, dev, args->size, &args->handle, 0,
+			      PAGE_SIZE);
 }
 
 /**
@@ -229,47 +227,3 @@ fail:
 		return VM_FAULT_SIGBUS;
 	}
 }
-
-static int psb_gem_create_stolen(struct drm_file *file, struct drm_device *dev,
-						int size, u32 *handle)
-{
-	struct gtt_range *gtt = psb_gtt_alloc_range(dev, size, "gem", 1);
-	if (gtt == NULL)
-		return -ENOMEM;
-
-	drm_gem_private_object_init(dev, &gtt->gem, size);
-	if (drm_gem_handle_create(file, &gtt->gem, handle) == 0)
-		return 0;
-
-	drm_gem_object_release(&gtt->gem);
-	psb_gtt_free_range(dev, gtt);
-	return -ENOMEM;
-}
-
-/*
- *	GEM interfaces for our specific client
- */
-int psb_gem_create_ioctl(struct drm_device *dev, void *data,
-					struct drm_file *file)
-{
-	struct drm_psb_gem_create *args = data;
-	int ret;
-	if (args->flags & GMA_GEM_CREATE_STOLEN) {
-		ret = psb_gem_create_stolen(file, dev, args->size,
-							&args->handle);
-		if (ret == 0)
-			return 0;
-		/* Fall throguh */
-		args->flags &= ~GMA_GEM_CREATE_STOLEN;
-	}
-	return psb_gem_create(file, dev, args->size, &args->handle);
-}
-
-int psb_gem_mmap_ioctl(struct drm_device *dev, void *data,
-					struct drm_file *file)
-{
-	struct drm_psb_gem_mmap *args = data;
-	return dev->driver->dumb_map_offset(file, dev,
-						args->handle, &args->offset);
-}
-
diff --git a/drivers/gpu/drm/gma500/gem.h b/drivers/gpu/drm/gma500/gem.h
new file mode 100644
index 00000000000..1381c5190f4
--- /dev/null
+++ b/drivers/gpu/drm/gma500/gem.h
@@ -0,0 +1,21 @@
+/**************************************************************************
+ * Copyright (c) 2014 Patrik Jakobsson
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ **************************************************************************/
+
+#ifndef _GEM_H
+#define _GEM_H
+
+extern int psb_gem_create(struct drm_file *file, struct drm_device *dev,
+			  u64 size, u32 *handlep, int stolen, u32 align);
+#endif
diff --git a/drivers/gpu/drm/gma500/gma_device.c b/drivers/gpu/drm/gma500/gma_device.c
new file mode 100644
index 00000000000..4a295f9ba06
--- /dev/null
+++ b/drivers/gpu/drm/gma500/gma_device.c
@@ -0,0 +1,56 @@
+/**************************************************************************
+ * Copyright (c) 2011, Intel Corporation.
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ **************************************************************************/
+
+#include <drm/drmP.h>
+#include "psb_drv.h"
+
+void gma_get_core_freq(struct drm_device *dev)
+{
+	uint32_t clock;
+	struct pci_dev *pci_root = pci_get_bus_and_slot(0, 0);
+	struct drm_psb_private *dev_priv = dev->dev_private;
+
+	/*pci_write_config_dword(pci_root, 0xD4, 0x00C32004);*/
+	/*pci_write_config_dword(pci_root, 0xD0, 0xE0033000);*/
+
+	pci_write_config_dword(pci_root, 0xD0, 0xD0050300);
+	pci_read_config_dword(pci_root, 0xD4, &clock);
+	pci_dev_put(pci_root);
+
+	switch (clock & 0x07) {
+	case 0:
+		dev_priv->core_freq = 100;
+		break;
+	case 1:
+		dev_priv->core_freq = 133;
+		break;
+	case 2:
+		dev_priv->core_freq = 150;
+		break;
+	case 3:
+		dev_priv->core_freq = 178;
+		break;
+	case 4:
+		dev_priv->core_freq = 200;
+		break;
+	case 5:
+	case 6:
+	case 7:
+		dev_priv->core_freq = 266;
+		break;
+	default:
+		dev_priv->core_freq = 0;
+	}
+}
diff --git a/drivers/gpu/drm/gma500/gma_device.h b/drivers/gpu/drm/gma500/gma_device.h
new file mode 100644
index 00000000000..e1dbb007b82
--- /dev/null
+++ b/drivers/gpu/drm/gma500/gma_device.h
@@ -0,0 +1,21 @@
+/**************************************************************************
+ * Copyright (c) 2011, Intel Corporation.
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ **************************************************************************/
+
+#ifndef _GMA_DEVICE_H
+#define _GMA_DEVICE_H
+
+extern void gma_get_core_freq(struct drm_device *dev);
+
+#endif
diff --git a/drivers/gpu/drm/gma500/gma_display.c b/drivers/gpu/drm/gma500/gma_display.c
index 24e8af3d22b..9bb9bddd881 100644
--- a/drivers/gpu/drm/gma500/gma_display.c
+++ b/drivers/gpu/drm/gma500/gma_display.c
@@ -59,7 +59,7 @@ int gma_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	struct drm_device *dev = crtc->dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
-	struct psb_framebuffer *psbfb = to_psb_fb(crtc->fb);
+	struct psb_framebuffer *psbfb = to_psb_fb(crtc->primary->fb);
 	int pipe = gma_crtc->pipe;
 	const struct psb_offset *map = &dev_priv->regmap[pipe];
 	unsigned long start, offset;
@@ -70,7 +70,7 @@ int gma_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 		return 0;
 
 	/* no fb bound */
-	if (!crtc->fb) {
+	if (!crtc->primary->fb) {
 		dev_err(dev->dev, "No FB bound\n");
 		goto gma_pipe_cleaner;
 	}
@@ -81,19 +81,19 @@ int gma_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	if (ret < 0)
 		goto gma_pipe_set_base_exit;
 	start = psbfb->gtt->offset;
-	offset = y * crtc->fb->pitches[0] + x * (crtc->fb->bits_per_pixel / 8);
+	offset = y * crtc->primary->fb->pitches[0] + x * (crtc->primary->fb->bits_per_pixel / 8);
 
-	REG_WRITE(map->stride, crtc->fb->pitches[0]);
+	REG_WRITE(map->stride, crtc->primary->fb->pitches[0]);
 
 	dspcntr = REG_READ(map->cntr);
 	dspcntr &= ~DISPPLANE_PIXFORMAT_MASK;
 
-	switch (crtc->fb->bits_per_pixel) {
+	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
 		dspcntr |= DISPPLANE_8BPP;
 		break;
 	case 16:
-		if (crtc->fb->depth == 15)
+		if (crtc->primary->fb->depth == 15)
 			dspcntr |= DISPPLANE_15_16BPP;
 		else
 			dspcntr |= DISPPLANE_16BPP;
@@ -349,6 +349,7 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 	/* If we didn't get a handle then turn the cursor off */
 	if (!handle) {
 		temp = CURSOR_MODE_DISABLE;
+		mutex_lock(&dev->struct_mutex);
 
 		if (gma_power_begin(dev, false)) {
 			REG_WRITE(control, temp);
@@ -365,6 +366,7 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 			gma_crtc->cursor_obj = NULL;
 		}
 
+		mutex_unlock(&dev->struct_mutex);
 		return 0;
 	}
 
@@ -374,9 +376,12 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 		return -EINVAL;
 	}
 
+	mutex_lock(&dev->struct_mutex);
 	obj = drm_gem_object_lookup(dev, file_priv, handle);
-	if (!obj)
-		return -ENOENT;
+	if (!obj) {
+		ret = -ENOENT;
+		goto unlock;
+	}
 
 	if (obj->size < width * height * 4) {
 		dev_dbg(dev->dev, "Buffer is too small\n");
@@ -440,10 +445,13 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 	}
 
 	gma_crtc->cursor_obj = obj;
+unlock:
+	mutex_unlock(&dev->struct_mutex);
 	return ret;
 
 unref_cursor:
 	drm_gem_object_unreference(obj);
+	mutex_unlock(&dev->struct_mutex);
 	return ret;
 }
 
@@ -477,6 +485,13 @@ int gma_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
 	return 0;
 }
 
+bool gma_encoder_mode_fixup(struct drm_encoder *encoder,
+			    const struct drm_display_mode *mode,
+			    struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
 bool gma_crtc_mode_fixup(struct drm_crtc *crtc,
 			 const struct drm_display_mode *mode,
 			 struct drm_display_mode *adjusted_mode)
@@ -503,8 +518,8 @@ void gma_crtc_disable(struct drm_crtc *crtc)
 
 	crtc_funcs->dpms(crtc, DRM_MODE_DPMS_OFF);
 
-	if (crtc->fb) {
-		gt = to_psb_fb(crtc->fb)->gtt;
+	if (crtc->primary->fb) {
+		gt = to_psb_fb(crtc->primary->fb)->gtt;
 		psb_gtt_unpin(gt);
 	}
 }
diff --git a/drivers/gpu/drm/gma500/gma_display.h b/drivers/gpu/drm/gma500/gma_display.h
index 78b9f986a6e..ed569d8a6af 100644
--- a/drivers/gpu/drm/gma500/gma_display.h
+++ b/drivers/gpu/drm/gma500/gma_display.h
@@ -90,6 +90,9 @@ extern void gma_crtc_restore(struct drm_crtc *crtc);
 extern void gma_encoder_prepare(struct drm_encoder *encoder);
 extern void gma_encoder_commit(struct drm_encoder *encoder);
 extern void gma_encoder_destroy(struct drm_encoder *encoder);
+extern bool gma_encoder_mode_fixup(struct drm_encoder *encoder,
+				   const struct drm_display_mode *mode,
+				   struct drm_display_mode *adjusted_mode);
 
 /* Common clock related functions */
 extern const struct gma_limit_t *gma_limit(struct drm_crtc *crtc, int refclk);
diff --git a/drivers/gpu/drm/gma500/gtt.c b/drivers/gpu/drm/gma500/gtt.c
index 2db731f0093..592d205a008 100644
--- a/drivers/gpu/drm/gma500/gtt.c
+++ b/drivers/gpu/drm/gma500/gtt.c
@@ -22,6 +22,7 @@
 #include <drm/drmP.h>
 #include <linux/shmem_fs.h>
 #include "psb_drv.h"
+#include "blitter.h"
 
 
 /*
@@ -105,11 +106,13 @@ static int psb_gtt_insert(struct drm_device *dev, struct gtt_range *r,
 
 	/* Write our page entries into the GTT itself */
 	for (i = r->roll; i < r->npage; i++) {
-		pte = psb_gtt_mask_pte(page_to_pfn(r->pages[i]), 0);
+		pte = psb_gtt_mask_pte(page_to_pfn(r->pages[i]),
+				       PSB_MMU_CACHED_MEMORY);
 		iowrite32(pte, gtt_slot++);
 	}
 	for (i = 0; i < r->roll; i++) {
-		pte = psb_gtt_mask_pte(page_to_pfn(r->pages[i]), 0);
+		pte = psb_gtt_mask_pte(page_to_pfn(r->pages[i]),
+				       PSB_MMU_CACHED_MEMORY);
 		iowrite32(pte, gtt_slot++);
 	}
 	/* Make sure all the entries are set before we return */
@@ -127,7 +130,7 @@ static int psb_gtt_insert(struct drm_device *dev, struct gtt_range *r,
  *	page table entries with the dummy page. This is protected via the gtt
  *	mutex which the caller must hold.
  */
-static void psb_gtt_remove(struct drm_device *dev, struct gtt_range *r)
+void psb_gtt_remove(struct drm_device *dev, struct gtt_range *r)
 {
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	u32 __iomem *gtt_slot;
@@ -137,7 +140,8 @@ static void psb_gtt_remove(struct drm_device *dev, struct gtt_range *r)
 	WARN_ON(r->stolen);
 
 	gtt_slot = psb_gtt_entry(dev, r);
-	pte = psb_gtt_mask_pte(page_to_pfn(dev_priv->scratch_page), 0);
+	pte = psb_gtt_mask_pte(page_to_pfn(dev_priv->scratch_page),
+			       PSB_MMU_CACHED_MEMORY);
 
 	for (i = 0; i < r->npage; i++)
 		iowrite32(pte, gtt_slot++);
@@ -176,11 +180,13 @@ void psb_gtt_roll(struct drm_device *dev, struct gtt_range *r, int roll)
 	gtt_slot = psb_gtt_entry(dev, r);
 
 	for (i = r->roll; i < r->npage; i++) {
-		pte = psb_gtt_mask_pte(page_to_pfn(r->pages[i]), 0);
+		pte = psb_gtt_mask_pte(page_to_pfn(r->pages[i]),
+				       PSB_MMU_CACHED_MEMORY);
 		iowrite32(pte, gtt_slot++);
 	}
 	for (i = 0; i < r->roll; i++) {
-		pte = psb_gtt_mask_pte(page_to_pfn(r->pages[i]), 0);
+		pte = psb_gtt_mask_pte(page_to_pfn(r->pages[i]),
+				       PSB_MMU_CACHED_MEMORY);
 		iowrite32(pte, gtt_slot++);
 	}
 	ioread32(gtt_slot - 1);
@@ -240,6 +246,7 @@ int psb_gtt_pin(struct gtt_range *gt)
 	int ret = 0;
 	struct drm_device *dev = gt->gem.dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
+	u32 gpu_base = dev_priv->gtt.gatt_start;
 
 	mutex_lock(&dev_priv->gtt_mutex);
 
@@ -252,6 +259,9 @@ int psb_gtt_pin(struct gtt_range *gt)
 			psb_gtt_detach_pages(gt);
 			goto out;
 		}
+		psb_mmu_insert_pages(psb_mmu_get_default_pd(dev_priv->mmu),
+				     gt->pages, (gpu_base + gt->offset),
+				     gt->npage, 0, 0, PSB_MMU_CACHED_MEMORY);
 	}
 	gt->in_gart++;
 out:
@@ -274,16 +284,30 @@ void psb_gtt_unpin(struct gtt_range *gt)
 {
 	struct drm_device *dev = gt->gem.dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
+	u32 gpu_base = dev_priv->gtt.gatt_start;
+	int ret;
 
+	/* While holding the gtt_mutex no new blits can be initiated */
 	mutex_lock(&dev_priv->gtt_mutex);
 
+	/* Wait for any possible usage of the memory to be finished */
+	ret = gma_blt_wait_idle(dev_priv);
+	if (ret) {
+		DRM_ERROR("Failed to idle the blitter, unpin failed!");
+		goto out;
+	}
+
 	WARN_ON(!gt->in_gart);
 
 	gt->in_gart--;
 	if (gt->in_gart == 0 && gt->stolen == 0) {
+		psb_mmu_remove_pages(psb_mmu_get_default_pd(dev_priv->mmu),
+				     (gpu_base + gt->offset), gt->npage, 0, 0);
 		psb_gtt_remove(dev, gt);
 		psb_gtt_detach_pages(gt);
 	}
+
+out:
 	mutex_unlock(&dev_priv->gtt_mutex);
 }
 
@@ -306,7 +330,7 @@ void psb_gtt_unpin(struct gtt_range *gt)
  *	as in use.
  */
 struct gtt_range *psb_gtt_alloc_range(struct drm_device *dev, int len,
-						const char *name, int backed)
+				      const char *name, int backed, u32 align)
 {
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	struct gtt_range *gt;
@@ -334,7 +358,7 @@ struct gtt_range *psb_gtt_alloc_range(struct drm_device *dev, int len,
 	/* Ensure this is set for non GEM objects */
 	gt->gem.dev = dev;
 	ret = allocate_resource(dev_priv->gtt_mem, &gt->resource,
-				len, start, end, PAGE_SIZE, NULL, NULL);
+				len, start, end, align, NULL, NULL);
 	if (ret == 0) {
 		gt->offset = gt->resource.start - r->start;
 		return gt;
@@ -497,6 +521,7 @@ int psb_gtt_init(struct drm_device *dev, int resume)
 	if (!resume)
 		dev_priv->vram_addr = ioremap_wc(dev_priv->stolen_base,
 						 stolen_size);
+
 	if (!dev_priv->vram_addr) {
 		dev_err(dev->dev, "Failure to map stolen base.\n");
 		ret = -ENOMEM;
@@ -512,7 +537,7 @@ int psb_gtt_init(struct drm_device *dev, int resume)
 	dev_dbg(dev->dev, "Set up %d stolen pages starting at 0x%08x, GTT offset %dK\n",
 		num_pages, pfn_base << PAGE_SHIFT, 0);
 	for (i = 0; i < num_pages; ++i) {
-		pte = psb_gtt_mask_pte(pfn_base + i, 0);
+		pte = psb_gtt_mask_pte(pfn_base + i, PSB_MMU_CACHED_MEMORY);
 		iowrite32(pte, dev_priv->gtt_map + i);
 	}
 
@@ -521,7 +546,7 @@ int psb_gtt_init(struct drm_device *dev, int resume)
 	 */
 
 	pfn_base = page_to_pfn(dev_priv->scratch_page);
-	pte = psb_gtt_mask_pte(pfn_base, 0);
+	pte = psb_gtt_mask_pte(pfn_base, PSB_MMU_CACHED_MEMORY);
 	for (; i < gtt_pages; ++i)
 		iowrite32(pte, dev_priv->gtt_map + i);
 
diff --git a/drivers/gpu/drm/gma500/gtt.h b/drivers/gpu/drm/gma500/gtt.h
index 6191d10acf3..f5860a739bd 100644
--- a/drivers/gpu/drm/gma500/gtt.h
+++ b/drivers/gpu/drm/gma500/gtt.h
@@ -53,7 +53,8 @@ struct gtt_range {
 };
 
 extern struct gtt_range *psb_gtt_alloc_range(struct drm_device *dev, int len,
-						const char *name, int backed);
+					     const char *name, int backed,
+					     u32 align);
 extern void psb_gtt_kref_put(struct gtt_range *gt);
 extern void psb_gtt_free_range(struct drm_device *dev, struct gtt_range *gt);
 extern int psb_gtt_pin(struct gtt_range *gt);
diff --git a/drivers/gpu/drm/gma500/mdfld_dsi_output.c b/drivers/gpu/drm/gma500/mdfld_dsi_output.c
index 860a4ee9baa..6e91b20ce2e 100644
--- a/drivers/gpu/drm/gma500/mdfld_dsi_output.c
+++ b/drivers/gpu/drm/gma500/mdfld_dsi_output.c
@@ -287,7 +287,7 @@ static int mdfld_dsi_connector_set_property(struct drm_connector *connector,
 						&gma_crtc->saved_mode,
 						encoder->crtc->x,
 						encoder->crtc->y,
-						encoder->crtc->fb))
+						encoder->crtc->primary->fb))
 					goto set_prop_error;
 			} else {
 				struct drm_encoder_helper_funcs *funcs =
diff --git a/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c b/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c
index 489ffd2c66e..87885d8c06e 100644
--- a/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c
+++ b/drivers/gpu/drm/gma500/mdfld_dsi_pkg_sender.c
@@ -148,7 +148,7 @@ static int handle_dsi_error(struct mdfld_dsi_pkg_sender *sender, u32 mask)
 		break;
 	case BIT(14):
 		/*wait for all fifo empty*/
-		/*wait_for_all_fifos_empty(sender)*/;
+		/*wait_for_all_fifos_empty(sender)*/
 		break;
 	case BIT(15):
 		dev_dbg(sender->dev->dev, "No Action required\n");
diff --git a/drivers/gpu/drm/gma500/mdfld_intel_display.c b/drivers/gpu/drm/gma500/mdfld_intel_display.c
index 321c00a944e..8cc8a5abbc7 100644
--- a/drivers/gpu/drm/gma500/mdfld_intel_display.c
+++ b/drivers/gpu/drm/gma500/mdfld_intel_display.c
@@ -166,7 +166,7 @@ static int mdfld__intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	struct drm_device *dev = crtc->dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
-	struct psb_framebuffer *psbfb = to_psb_fb(crtc->fb);
+	struct psb_framebuffer *psbfb = to_psb_fb(crtc->primary->fb);
 	int pipe = gma_crtc->pipe;
 	const struct psb_offset *map = &dev_priv->regmap[pipe];
 	unsigned long start, offset;
@@ -178,12 +178,12 @@ static int mdfld__intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	dev_dbg(dev->dev, "pipe = 0x%x.\n", pipe);
 
 	/* no fb bound */
-	if (!crtc->fb) {
+	if (!crtc->primary->fb) {
 		dev_dbg(dev->dev, "No FB bound\n");
 		return 0;
 	}
 
-	ret = check_fb(crtc->fb);
+	ret = check_fb(crtc->primary->fb);
 	if (ret)
 		return ret;
 
@@ -196,18 +196,18 @@ static int mdfld__intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 		return 0;
 
 	start = psbfb->gtt->offset;
-	offset = y * crtc->fb->pitches[0] + x * (crtc->fb->bits_per_pixel / 8);
+	offset = y * crtc->primary->fb->pitches[0] + x * (crtc->primary->fb->bits_per_pixel / 8);
 
-	REG_WRITE(map->stride, crtc->fb->pitches[0]);
+	REG_WRITE(map->stride, crtc->primary->fb->pitches[0]);
 	dspcntr = REG_READ(map->cntr);
 	dspcntr &= ~DISPPLANE_PIXFORMAT_MASK;
 
-	switch (crtc->fb->bits_per_pixel) {
+	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
 		dspcntr |= DISPPLANE_8BPP;
 		break;
 	case 16:
-		if (crtc->fb->depth == 15)
+		if (crtc->primary->fb->depth == 15)
 			dspcntr |= DISPPLANE_15_16BPP;
 		else
 			dspcntr |= DISPPLANE_16BPP;
@@ -700,7 +700,7 @@ static int mdfld_crtc_mode_set(struct drm_crtc *crtc,
 	}
 #endif
 
-	ret = check_fb(crtc->fb);
+	ret = check_fb(crtc->primary->fb);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/gma500/mmu.c b/drivers/gpu/drm/gma500/mmu.c
index 49bac41beef..0eaf11c1993 100644
--- a/drivers/gpu/drm/gma500/mmu.c
+++ b/drivers/gpu/drm/gma500/mmu.c
@@ -18,6 +18,7 @@
 #include <drm/drmP.h>
 #include "psb_drv.h"
 #include "psb_reg.h"
+#include "mmu.h"
 
 /*
  * Code for the SGX MMU:
@@ -47,51 +48,6 @@
  * but on average it should be fast.
  */
 
-struct psb_mmu_driver {
-	/* protects driver- and pd structures. Always take in read mode
-	 * before taking the page table spinlock.
-	 */
-	struct rw_semaphore sem;
-
-	/* protects page tables, directory tables and pt tables.
-	 * and pt structures.
-	 */
-	spinlock_t lock;
-
-	atomic_t needs_tlbflush;
-
-	uint8_t __iomem *register_map;
-	struct psb_mmu_pd *default_pd;
-	/*uint32_t bif_ctrl;*/
-	int has_clflush;
-	int clflush_add;
-	unsigned long clflush_mask;
-
-	struct drm_psb_private *dev_priv;
-};
-
-struct psb_mmu_pd;
-
-struct psb_mmu_pt {
-	struct psb_mmu_pd *pd;
-	uint32_t index;
-	uint32_t count;
-	struct page *p;
-	uint32_t *v;
-};
-
-struct psb_mmu_pd {
-	struct psb_mmu_driver *driver;
-	int hw_context;
-	struct psb_mmu_pt **tables;
-	struct page *p;
-	struct page *dummy_pt;
-	struct page *dummy_page;
-	uint32_t pd_mask;
-	uint32_t invalid_pde;
-	uint32_t invalid_pte;
-};
-
 static inline uint32_t psb_mmu_pt_index(uint32_t offset)
 {
 	return (offset >> PSB_PTE_SHIFT) & 0x3FF;
@@ -102,13 +58,13 @@ static inline uint32_t psb_mmu_pd_index(uint32_t offset)
 	return offset >> PSB_PDE_SHIFT;
 }
 
+#if defined(CONFIG_X86)
 static inline void psb_clflush(void *addr)
 {
 	__asm__ __volatile__("clflush (%0)\n" : : "r"(addr) : "memory");
 }
 
-static inline void psb_mmu_clflush(struct psb_mmu_driver *driver,
-				   void *addr)
+static inline void psb_mmu_clflush(struct psb_mmu_driver *driver, void *addr)
 {
 	if (!driver->has_clflush)
 		return;
@@ -117,62 +73,77 @@ static inline void psb_mmu_clflush(struct psb_mmu_driver *driver,
 	psb_clflush(addr);
 	mb();
 }
+#else
 
-static void psb_page_clflush(struct psb_mmu_driver *driver, struct page* page)
-{
-	uint32_t clflush_add = driver->clflush_add >> PAGE_SHIFT;
-	uint32_t clflush_count = PAGE_SIZE / clflush_add;
-	int i;
-	uint8_t *clf;
-
-	clf = kmap_atomic(page);
-	mb();
-	for (i = 0; i < clflush_count; ++i) {
-		psb_clflush(clf);
-		clf += clflush_add;
-	}
-	mb();
-	kunmap_atomic(clf);
+static inline void psb_mmu_clflush(struct psb_mmu_driver *driver, void *addr)
+{;
 }
 
-static void psb_pages_clflush(struct psb_mmu_driver *driver,
-				struct page *page[], unsigned long num_pages)
-{
-	int i;
-
-	if (!driver->has_clflush)
-		return ;
+#endif
 
-	for (i = 0; i < num_pages; i++)
-		psb_page_clflush(driver, *page++);
-}
-
-static void psb_mmu_flush_pd_locked(struct psb_mmu_driver *driver,
-				    int force)
+static void psb_mmu_flush_pd_locked(struct psb_mmu_driver *driver, int force)
 {
+	struct drm_device *dev = driver->dev;
+	struct drm_psb_private *dev_priv = dev->dev_private;
+
+	if (atomic_read(&driver->needs_tlbflush) || force) {
+		uint32_t val = PSB_RSGX32(PSB_CR_BIF_CTRL);
+		PSB_WSGX32(val | _PSB_CB_CTRL_INVALDC, PSB_CR_BIF_CTRL);
+
+		/* Make sure data cache is turned off before enabling it */
+		wmb();
+		PSB_WSGX32(val & ~_PSB_CB_CTRL_INVALDC, PSB_CR_BIF_CTRL);
+		(void)PSB_RSGX32(PSB_CR_BIF_CTRL);
+		if (driver->msvdx_mmu_invaldc)
+			atomic_set(driver->msvdx_mmu_invaldc, 1);
+	}
 	atomic_set(&driver->needs_tlbflush, 0);
 }
 
+#if 0
 static void psb_mmu_flush_pd(struct psb_mmu_driver *driver, int force)
 {
 	down_write(&driver->sem);
 	psb_mmu_flush_pd_locked(driver, force);
 	up_write(&driver->sem);
 }
+#endif
 
-void psb_mmu_flush(struct psb_mmu_driver *driver, int rc_prot)
+void psb_mmu_flush(struct psb_mmu_driver *driver)
 {
-	if (rc_prot)
-		down_write(&driver->sem);
-	if (rc_prot)
-		up_write(&driver->sem);
+	struct drm_device *dev = driver->dev;
+	struct drm_psb_private *dev_priv = dev->dev_private;
+	uint32_t val;
+
+	down_write(&driver->sem);
+	val = PSB_RSGX32(PSB_CR_BIF_CTRL);
+	if (atomic_read(&driver->needs_tlbflush))
+		PSB_WSGX32(val | _PSB_CB_CTRL_INVALDC, PSB_CR_BIF_CTRL);
+	else
+		PSB_WSGX32(val | _PSB_CB_CTRL_FLUSH, PSB_CR_BIF_CTRL);
+
+	/* Make sure data cache is turned off and MMU is flushed before
+	   restoring bank interface control register */
+	wmb();
+	PSB_WSGX32(val & ~(_PSB_CB_CTRL_FLUSH | _PSB_CB_CTRL_INVALDC),
+		   PSB_CR_BIF_CTRL);
+	(void)PSB_RSGX32(PSB_CR_BIF_CTRL);
+
+	atomic_set(&driver->needs_tlbflush, 0);
+	if (driver->msvdx_mmu_invaldc)
+		atomic_set(driver->msvdx_mmu_invaldc, 1);
+	up_write(&driver->sem);
 }
 
 void psb_mmu_set_pd_context(struct psb_mmu_pd *pd, int hw_context)
 {
-	/*ttm_tt_cache_flush(&pd->p, 1);*/
-	psb_pages_clflush(pd->driver, &pd->p, 1);
+	struct drm_device *dev = pd->driver->dev;
+	struct drm_psb_private *dev_priv = dev->dev_private;
+	uint32_t offset = (hw_context == 0) ? PSB_CR_BIF_DIR_LIST_BASE0 :
+			  PSB_CR_BIF_DIR_LIST_BASE1 + hw_context * 4;
+
 	down_write(&pd->driver->sem);
+	PSB_WSGX32(page_to_pfn(pd->p) << PAGE_SHIFT, offset);
 	wmb();
 	psb_mmu_flush_pd_locked(pd->driver, 1);
 	pd->hw_context = hw_context;
@@ -183,7 +154,6 @@ void psb_mmu_set_pd_context(struct psb_mmu_pd *pd, int hw_context)
 static inline unsigned long psb_pd_addr_end(unsigned long addr,
 					    unsigned long end)
 {
-
 	addr = (addr + PSB_PDE_MASK + 1) & ~PSB_PDE_MASK;
 	return (addr < end) ? addr : end;
 }
@@ -223,12 +193,10 @@ struct psb_mmu_pd *psb_mmu_alloc_pd(struct psb_mmu_driver *driver,
 		goto out_err3;
 
 	if (!trap_pagefaults) {
-		pd->invalid_pde =
-		    psb_mmu_mask_pte(page_to_pfn(pd->dummy_pt),
-				     invalid_type);
-		pd->invalid_pte =
-		    psb_mmu_mask_pte(page_to_pfn(pd->dummy_page),
-				     invalid_type);
+		pd->invalid_pde = psb_mmu_mask_pte(page_to_pfn(pd->dummy_pt),
+						   invalid_type);
+		pd->invalid_pte = psb_mmu_mask_pte(page_to_pfn(pd->dummy_page),
+						   invalid_type);
 	} else {
 		pd->invalid_pde = 0;
 		pd->invalid_pte = 0;
@@ -279,12 +247,16 @@ static void psb_mmu_free_pt(struct psb_mmu_pt *pt)
 void psb_mmu_free_pagedir(struct psb_mmu_pd *pd)
 {
 	struct psb_mmu_driver *driver = pd->driver;
+	struct drm_device *dev = driver->dev;
+	struct drm_psb_private *dev_priv = dev->dev_private;
 	struct psb_mmu_pt *pt;
 	int i;
 
 	down_write(&driver->sem);
-	if (pd->hw_context != -1)
+	if (pd->hw_context != -1) {
+		PSB_WSGX32(0, PSB_CR_BIF_DIR_LIST_BASE0 + pd->hw_context * 4);
 		psb_mmu_flush_pd_locked(driver, 1);
+	}
 
 	/* Should take the spinlock here, but we don't need to do that
 	   since we have the semaphore in write mode. */
@@ -331,7 +303,7 @@ static struct psb_mmu_pt *psb_mmu_alloc_pt(struct psb_mmu_pd *pd)
 	for (i = 0; i < (PAGE_SIZE / sizeof(uint32_t)); ++i)
 		*ptes++ = pd->invalid_pte;
 
-
+#if defined(CONFIG_X86)
 	if (pd->driver->has_clflush && pd->hw_context != -1) {
 		mb();
 		for (i = 0; i < clflush_count; ++i) {
@@ -340,7 +312,7 @@ static struct psb_mmu_pt *psb_mmu_alloc_pt(struct psb_mmu_pd *pd)
 		}
 		mb();
 	}
-
+#endif
 	kunmap_atomic(v);
 	spin_unlock(lock);
 
@@ -351,7 +323,7 @@ static struct psb_mmu_pt *psb_mmu_alloc_pt(struct psb_mmu_pd *pd)
 	return pt;
 }
 
-static struct psb_mmu_pt *psb_mmu_pt_alloc_map_lock(struct psb_mmu_pd *pd,
+struct psb_mmu_pt *psb_mmu_pt_alloc_map_lock(struct psb_mmu_pd *pd,
 					     unsigned long addr)
 {
 	uint32_t index = psb_mmu_pd_index(addr);
@@ -383,7 +355,7 @@ static struct psb_mmu_pt *psb_mmu_pt_alloc_map_lock(struct psb_mmu_pd *pd,
 		kunmap_atomic((void *) v);
 
 		if (pd->hw_context != -1) {
-			psb_mmu_clflush(pd->driver, (void *) &v[index]);
+			psb_mmu_clflush(pd->driver, (void *)&v[index]);
 			atomic_set(&pd->driver->needs_tlbflush, 1);
 		}
 	}
@@ -420,8 +392,7 @@ static void psb_mmu_pt_unmap_unlock(struct psb_mmu_pt *pt)
 		pd->tables[pt->index] = NULL;
 
 		if (pd->hw_context != -1) {
-			psb_mmu_clflush(pd->driver,
-					(void *) &v[pt->index]);
+			psb_mmu_clflush(pd->driver, (void *)&v[pt->index]);
 			atomic_set(&pd->driver->needs_tlbflush, 1);
 		}
 		kunmap_atomic(pt->v);
@@ -432,8 +403,8 @@ static void psb_mmu_pt_unmap_unlock(struct psb_mmu_pt *pt)
 	spin_unlock(&pd->driver->lock);
 }
 
-static inline void psb_mmu_set_pte(struct psb_mmu_pt *pt,
-				   unsigned long addr, uint32_t pte)
+static inline void psb_mmu_set_pte(struct psb_mmu_pt *pt, unsigned long addr,
+				   uint32_t pte)
 {
 	pt->v[psb_mmu_pt_index(addr)] = pte;
 }
@@ -444,69 +415,50 @@ static inline void psb_mmu_invalidate_pte(struct psb_mmu_pt *pt,
 	pt->v[psb_mmu_pt_index(addr)] = pt->pd->invalid_pte;
 }
 
-
-void psb_mmu_mirror_gtt(struct psb_mmu_pd *pd,
-			uint32_t mmu_offset, uint32_t gtt_start,
-			uint32_t gtt_pages)
+struct psb_mmu_pd *psb_mmu_get_default_pd(struct psb_mmu_driver *driver)
 {
-	uint32_t *v;
-	uint32_t start = psb_mmu_pd_index(mmu_offset);
-	struct psb_mmu_driver *driver = pd->driver;
-	int num_pages = gtt_pages;
+	struct psb_mmu_pd *pd;
 
 	down_read(&driver->sem);
-	spin_lock(&driver->lock);
-
-	v = kmap_atomic(pd->p);
-	v += start;
-
-	while (gtt_pages--) {
-		*v++ = gtt_start | pd->pd_mask;
-		gtt_start += PAGE_SIZE;
-	}
-
-	/*ttm_tt_cache_flush(&pd->p, num_pages);*/
-	psb_pages_clflush(pd->driver, &pd->p, num_pages);
-	kunmap_atomic(v);
-	spin_unlock(&driver->lock);
-
-	if (pd->hw_context != -1)
-		atomic_set(&pd->driver->needs_tlbflush, 1);
+	pd = driver->default_pd;
+	up_read(&driver->sem);
 
-	up_read(&pd->driver->sem);
-	psb_mmu_flush_pd(pd->driver, 0);
+	return pd;
 }
 
-struct psb_mmu_pd *psb_mmu_get_default_pd(struct psb_mmu_driver *driver)
+/* Returns the physical address of the PD shared by sgx/msvdx */
+uint32_t psb_get_default_pd_addr(struct psb_mmu_driver *driver)
 {
 	struct psb_mmu_pd *pd;
 
-	/* down_read(&driver->sem); */
-	pd = driver->default_pd;
-	/* up_read(&driver->sem); */
-
-	return pd;
+	pd = psb_mmu_get_default_pd(driver);
+	return page_to_pfn(pd->p) << PAGE_SHIFT;
 }
 
 void psb_mmu_driver_takedown(struct psb_mmu_driver *driver)
 {
+	struct drm_device *dev = driver->dev;
+	struct drm_psb_private *dev_priv = dev->dev_private;
+
+	PSB_WSGX32(driver->bif_ctrl, PSB_CR_BIF_CTRL);
 	psb_mmu_free_pagedir(driver->default_pd);
 	kfree(driver);
 }
 
-struct psb_mmu_driver *psb_mmu_driver_init(uint8_t __iomem * registers,
-					int trap_pagefaults,
-					int invalid_type,
-					struct drm_psb_private *dev_priv)
+struct psb_mmu_driver *psb_mmu_driver_init(struct drm_device *dev,
+					   int trap_pagefaults,
+					   int invalid_type,
+					   atomic_t *msvdx_mmu_invaldc)
 {
 	struct psb_mmu_driver *driver;
+	struct drm_psb_private *dev_priv = dev->dev_private;
 
 	driver = kmalloc(sizeof(*driver), GFP_KERNEL);
 
 	if (!driver)
 		return NULL;
-	driver->dev_priv = dev_priv;
 
+	driver->dev = dev;
 	driver->default_pd = psb_mmu_alloc_pd(driver, trap_pagefaults,
 					      invalid_type);
 	if (!driver->default_pd)
@@ -515,17 +467,24 @@ struct psb_mmu_driver *psb_mmu_driver_init(uint8_t __iomem * registers,
 	spin_lock_init(&driver->lock);
 	init_rwsem(&driver->sem);
 	down_write(&driver->sem);
-	driver->register_map = registers;
 	atomic_set(&driver->needs_tlbflush, 1);
+	driver->msvdx_mmu_invaldc = msvdx_mmu_invaldc;
+
+	driver->bif_ctrl = PSB_RSGX32(PSB_CR_BIF_CTRL);
+	PSB_WSGX32(driver->bif_ctrl | _PSB_CB_CTRL_CLEAR_FAULT,
+		   PSB_CR_BIF_CTRL);
+	PSB_WSGX32(driver->bif_ctrl & ~_PSB_CB_CTRL_CLEAR_FAULT,
+		   PSB_CR_BIF_CTRL);
 
 	driver->has_clflush = 0;
 
-	if (boot_cpu_has(X86_FEATURE_CLFLSH)) {
+#if defined(CONFIG_X86)
+	if (boot_cpu_has(X86_FEATURE_CLFLUSH)) {
 		uint32_t tfms, misc, cap0, cap4, clflush_size;
 
 		/*
-		 * clflush size is determined at kernel setup for x86_64
-		 *  but not for i386. We have to do it here.
+		 * clflush size is determined at kernel setup for x86_64 but not
+		 * for i386. We have to do it here.
 		 */
 
 		cpuid(0x00000001, &tfms, &misc, &cap0, &cap4);
@@ -536,6 +495,7 @@ struct psb_mmu_driver *psb_mmu_driver_init(uint8_t __iomem * registers,
 		driver->clflush_mask = driver->clflush_add - 1;
 		driver->clflush_mask = ~driver->clflush_mask;
 	}
+#endif
 
 	up_write(&driver->sem);
 	return driver;
@@ -545,9 +505,9 @@ out_err1:
 	return NULL;
 }
 
-static void psb_mmu_flush_ptes(struct psb_mmu_pd *pd,
-			       unsigned long address, uint32_t num_pages,
-			       uint32_t desired_tile_stride,
+#if defined(CONFIG_X86)
+static void psb_mmu_flush_ptes(struct psb_mmu_pd *pd, unsigned long address,
+			       uint32_t num_pages, uint32_t desired_tile_stride,
 			       uint32_t hw_tile_stride)
 {
 	struct psb_mmu_pt *pt;
@@ -561,11 +521,8 @@ static void psb_mmu_flush_ptes(struct psb_mmu_pd *pd,
 	unsigned long clflush_add = pd->driver->clflush_add;
 	unsigned long clflush_mask = pd->driver->clflush_mask;
 
-	if (!pd->driver->has_clflush) {
-		/*ttm_tt_cache_flush(&pd->p, num_pages);*/
-		psb_pages_clflush(pd->driver, &pd->p, num_pages);
+	if (!pd->driver->has_clflush)
 		return;
-	}
 
 	if (hw_tile_stride)
 		rows = num_pages / desired_tile_stride;
@@ -586,10 +543,8 @@ static void psb_mmu_flush_ptes(struct psb_mmu_pd *pd,
 			if (!pt)
 				continue;
 			do {
-				psb_clflush(&pt->v
-					    [psb_mmu_pt_index(addr)]);
-			} while (addr +=
-				 clflush_add,
+				psb_clflush(&pt->v[psb_mmu_pt_index(addr)]);
+			} while (addr += clflush_add,
 				 (addr & clflush_mask) < next);
 
 			psb_mmu_pt_unmap_unlock(pt);
@@ -598,6 +553,14 @@ static void psb_mmu_flush_ptes(struct psb_mmu_pd *pd,
 	}
 	mb();
 }
+#else
+static void psb_mmu_flush_ptes(struct psb_mmu_pd *pd, unsigned long address,
+			       uint32_t num_pages, uint32_t desired_tile_stride,
+			       uint32_t hw_tile_stride)
+{
+	drm_ttm_cache_flush();
+}
+#endif
 
 void psb_mmu_remove_pfn_sequence(struct psb_mmu_pd *pd,
 				 unsigned long address, uint32_t num_pages)
@@ -633,7 +596,7 @@ out:
 	up_read(&pd->driver->sem);
 
 	if (pd->hw_context != -1)
-		psb_mmu_flush(pd->driver, 0);
+		psb_mmu_flush(pd->driver);
 
 	return;
 }
@@ -660,7 +623,7 @@ void psb_mmu_remove_pages(struct psb_mmu_pd *pd, unsigned long address,
 	add = desired_tile_stride << PAGE_SHIFT;
 	row_add = hw_tile_stride << PAGE_SHIFT;
 
-	/* down_read(&pd->driver->sem); */
+	down_read(&pd->driver->sem);
 
 	/* Make sure we only need to flush this processor's cache */
 
@@ -688,10 +651,10 @@ void psb_mmu_remove_pages(struct psb_mmu_pd *pd, unsigned long address,
 		psb_mmu_flush_ptes(pd, f_address, num_pages,
 				   desired_tile_stride, hw_tile_stride);
 
-	/* up_read(&pd->driver->sem); */
+	up_read(&pd->driver->sem);
 
 	if (pd->hw_context != -1)
-		psb_mmu_flush(pd->driver, 0);
+		psb_mmu_flush(pd->driver);
 }
 
 int psb_mmu_insert_pfn_sequence(struct psb_mmu_pd *pd, uint32_t start_pfn,
@@ -704,7 +667,7 @@ int psb_mmu_insert_pfn_sequence(struct psb_mmu_pd *pd, uint32_t start_pfn,
 	unsigned long end;
 	unsigned long next;
 	unsigned long f_address = address;
-	int ret = 0;
+	int ret = -ENOMEM;
 
 	down_read(&pd->driver->sem);
 
@@ -726,6 +689,7 @@ int psb_mmu_insert_pfn_sequence(struct psb_mmu_pd *pd, uint32_t start_pfn,
 		psb_mmu_pt_unmap_unlock(pt);
 
 	} while (addr = next, next != end);
+	ret = 0;
 
 out:
 	if (pd->hw_context != -1)
@@ -734,15 +698,15 @@ out:
 	up_read(&pd->driver->sem);
 
 	if (pd->hw_context != -1)
-		psb_mmu_flush(pd->driver, 1);
+		psb_mmu_flush(pd->driver);
 
-	return ret;
+	return 0;
 }
 
 int psb_mmu_insert_pages(struct psb_mmu_pd *pd, struct page **pages,
 			 unsigned long address, uint32_t num_pages,
-			 uint32_t desired_tile_stride,
-			 uint32_t hw_tile_stride, int type)
+			 uint32_t desired_tile_stride, uint32_t hw_tile_stride,
+			 int type)
 {
 	struct psb_mmu_pt *pt;
 	uint32_t rows = 1;
@@ -754,7 +718,7 @@ int psb_mmu_insert_pages(struct psb_mmu_pd *pd, struct page **pages,
 	unsigned long add;
 	unsigned long row_add;
 	unsigned long f_address = address;
-	int ret = 0;
+	int ret = -ENOMEM;
 
 	if (hw_tile_stride) {
 		if (num_pages % desired_tile_stride != 0)
@@ -777,14 +741,11 @@ int psb_mmu_insert_pages(struct psb_mmu_pd *pd, struct page **pages,
 		do {
 			next = psb_pd_addr_end(addr, end);
 			pt = psb_mmu_pt_alloc_map_lock(pd, addr);
-			if (!pt) {
-				ret = -ENOMEM;
+			if (!pt)
 				goto out;
-			}
 			do {
-				pte =
-				    psb_mmu_mask_pte(page_to_pfn(*pages++),
-						     type);
+				pte = psb_mmu_mask_pte(page_to_pfn(*pages++),
+						       type);
 				psb_mmu_set_pte(pt, addr, pte);
 				pt->count++;
 			} while (addr += PAGE_SIZE, addr < next);
@@ -794,6 +755,8 @@ int psb_mmu_insert_pages(struct psb_mmu_pd *pd, struct page **pages,
 
 		address += row_add;
 	}
+
+	ret = 0;
 out:
 	if (pd->hw_context != -1)
 		psb_mmu_flush_ptes(pd, f_address, num_pages,
@@ -802,7 +765,7 @@ out:
 	up_read(&pd->driver->sem);
 
 	if (pd->hw_context != -1)
-		psb_mmu_flush(pd->driver, 1);
+		psb_mmu_flush(pd->driver);
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/gma500/mmu.h b/drivers/gpu/drm/gma500/mmu.h
new file mode 100644
index 00000000000..e89abec6209
--- /dev/null
+++ b/drivers/gpu/drm/gma500/mmu.h
@@ -0,0 +1,93 @@
+/**************************************************************************
+ * Copyright (c) 2007-2011, Intel Corporation.
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ **************************************************************************/
+
+#ifndef __MMU_H
+#define __MMU_H
+
+struct psb_mmu_driver {
+	/* protects driver- and pd structures. Always take in read mode
+	 * before taking the page table spinlock.
+	 */
+	struct rw_semaphore sem;
+
+	/* protects page tables, directory tables and pt tables.
+	 * and pt structures.
+	 */
+	spinlock_t lock;
+
+	atomic_t needs_tlbflush;
+	atomic_t *msvdx_mmu_invaldc;
+	struct psb_mmu_pd *default_pd;
+	uint32_t bif_ctrl;
+	int has_clflush;
+	int clflush_add;
+	unsigned long clflush_mask;
+
+	struct drm_device *dev;
+};
+
+struct psb_mmu_pd;
+
+struct psb_mmu_pt {
+	struct psb_mmu_pd *pd;
+	uint32_t index;
+	uint32_t count;
+	struct page *p;
+	uint32_t *v;
+};
+
+struct psb_mmu_pd {
+	struct psb_mmu_driver *driver;
+	int hw_context;
+	struct psb_mmu_pt **tables;
+	struct page *p;
+	struct page *dummy_pt;
+	struct page *dummy_page;
+	uint32_t pd_mask;
+	uint32_t invalid_pde;
+	uint32_t invalid_pte;
+};
+
+extern struct psb_mmu_driver *psb_mmu_driver_init(struct drm_device *dev,
+						  int trap_pagefaults,
+						  int invalid_type,
+						  atomic_t *msvdx_mmu_invaldc);
+extern void psb_mmu_driver_takedown(struct psb_mmu_driver *driver);
+extern struct psb_mmu_pd *psb_mmu_get_default_pd(struct psb_mmu_driver
+						 *driver);
+extern struct psb_mmu_pd *psb_mmu_alloc_pd(struct psb_mmu_driver *driver,
+					   int trap_pagefaults,
+					   int invalid_type);
+extern void psb_mmu_free_pagedir(struct psb_mmu_pd *pd);
+extern void psb_mmu_flush(struct psb_mmu_driver *driver);
+extern void psb_mmu_remove_pfn_sequence(struct psb_mmu_pd *pd,
+					unsigned long address,
+					uint32_t num_pages);
+extern int psb_mmu_insert_pfn_sequence(struct psb_mmu_pd *pd,
+				       uint32_t start_pfn,
+				       unsigned long address,
+				       uint32_t num_pages, int type);
+extern int psb_mmu_virtual_to_pfn(struct psb_mmu_pd *pd, uint32_t virtual,
+				  unsigned long *pfn);
+extern void psb_mmu_set_pd_context(struct psb_mmu_pd *pd, int hw_context);
+extern int psb_mmu_insert_pages(struct psb_mmu_pd *pd, struct page **pages,
+				unsigned long address, uint32_t num_pages,
+				uint32_t desired_tile_stride,
+				uint32_t hw_tile_stride, int type);
+extern void psb_mmu_remove_pages(struct psb_mmu_pd *pd,
+				 unsigned long address, uint32_t num_pages,
+				 uint32_t desired_tile_stride,
+				 uint32_t hw_tile_stride);
+
+#endif
diff --git a/drivers/gpu/drm/gma500/oaktrail_crtc.c b/drivers/gpu/drm/gma500/oaktrail_crtc.c
index 8195e859210..2de216c2374 100644
--- a/drivers/gpu/drm/gma500/oaktrail_crtc.c
+++ b/drivers/gpu/drm/gma500/oaktrail_crtc.c
@@ -599,7 +599,7 @@ static int oaktrail_pipe_set_base(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
-	struct psb_framebuffer *psbfb = to_psb_fb(crtc->fb);
+	struct psb_framebuffer *psbfb = to_psb_fb(crtc->primary->fb);
 	int pipe = gma_crtc->pipe;
 	const struct psb_offset *map = &dev_priv->regmap[pipe];
 	unsigned long start, offset;
@@ -608,7 +608,7 @@ static int oaktrail_pipe_set_base(struct drm_crtc *crtc,
 	int ret = 0;
 
 	/* no fb bound */
-	if (!crtc->fb) {
+	if (!crtc->primary->fb) {
 		dev_dbg(dev->dev, "No FB bound\n");
 		return 0;
 	}
@@ -617,19 +617,19 @@ static int oaktrail_pipe_set_base(struct drm_crtc *crtc,
 		return 0;
 
 	start = psbfb->gtt->offset;
-	offset = y * crtc->fb->pitches[0] + x * (crtc->fb->bits_per_pixel / 8);
+	offset = y * crtc->primary->fb->pitches[0] + x * (crtc->primary->fb->bits_per_pixel / 8);
 
-	REG_WRITE(map->stride, crtc->fb->pitches[0]);
+	REG_WRITE(map->stride, crtc->primary->fb->pitches[0]);
 
 	dspcntr = REG_READ(map->cntr);
 	dspcntr &= ~DISPPLANE_PIXFORMAT_MASK;
 
-	switch (crtc->fb->bits_per_pixel) {
+	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
 		dspcntr |= DISPPLANE_8BPP;
 		break;
 	case 16:
-		if (crtc->fb->depth == 15)
+		if (crtc->primary->fb->depth == 15)
 			dspcntr |= DISPPLANE_15_16BPP;
 		else
 			dspcntr |= DISPPLANE_16BPP;
diff --git a/drivers/gpu/drm/gma500/oaktrail_hdmi.c b/drivers/gpu/drm/gma500/oaktrail_hdmi.c
index 38153143ed8..cf018ddcc5a 100644
--- a/drivers/gpu/drm/gma500/oaktrail_hdmi.c
+++ b/drivers/gpu/drm/gma500/oaktrail_hdmi.c
@@ -523,13 +523,6 @@ static int oaktrail_hdmi_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
-static bool oaktrail_hdmi_mode_fixup(struct drm_encoder *encoder,
-				 const struct drm_display_mode *mode,
-				 struct drm_display_mode *adjusted_mode)
-{
-	return true;
-}
-
 static enum drm_connector_status
 oaktrail_hdmi_detect(struct drm_connector *connector, bool force)
 {
@@ -608,7 +601,7 @@ static void oaktrail_hdmi_destroy(struct drm_connector *connector)
 
 static const struct drm_encoder_helper_funcs oaktrail_hdmi_helper_funcs = {
 	.dpms = oaktrail_hdmi_dpms,
-	.mode_fixup = oaktrail_hdmi_mode_fixup,
+	.mode_fixup = gma_encoder_mode_fixup,
 	.prepare = gma_encoder_prepare,
 	.mode_set = oaktrail_hdmi_mode_set,
 	.commit = gma_encoder_commit,
diff --git a/drivers/gpu/drm/gma500/oaktrail_lvds.c b/drivers/gpu/drm/gma500/oaktrail_lvds.c
index 5e069786273..9b099468a5d 100644
--- a/drivers/gpu/drm/gma500/oaktrail_lvds.c
+++ b/drivers/gpu/drm/gma500/oaktrail_lvds.c
@@ -359,6 +359,7 @@ void oaktrail_lvds_init(struct drm_device *dev,
 	 *    if closed, act like it's not there for now
 	 */
 
+	mutex_lock(&dev->mode_config.mutex);
 	i2c_adap = i2c_get_adapter(dev_priv->ops->i2c_bus);
 	if (i2c_adap == NULL)
 		dev_err(dev->dev, "No ddc adapter available!\n");
@@ -401,10 +402,14 @@ void oaktrail_lvds_init(struct drm_device *dev,
 	}
 
 out:
+	mutex_unlock(&dev->mode_config.mutex);
+
 	drm_sysfs_connector_add(connector);
 	return;
 
 failed_find:
+	mutex_unlock(&dev->mode_config.mutex);
+
 	dev_dbg(dev->dev, "No LVDS modes found, disabling.\n");
 	if (gma_encoder->ddc_bus)
 		psb_intel_i2c_destroy(gma_encoder->ddc_bus);
diff --git a/drivers/gpu/drm/gma500/opregion.c b/drivers/gpu/drm/gma500/opregion.c
index ad0d6de938f..ab696ca7eee 100644
--- a/drivers/gpu/drm/gma500/opregion.c
+++ b/drivers/gpu/drm/gma500/opregion.c
@@ -22,7 +22,6 @@
  *
  */
 #include <linux/acpi.h>
-#include <linux/acpi_io.h>
 #include "psb_drv.h"
 #include "psb_intel_reg.h"
 
@@ -174,10 +173,13 @@ static u32 asle_set_backlight(struct drm_device *dev, u32 bclp)
 	return 0;
 }
 
-void psb_intel_opregion_asle_intr(struct drm_device *dev)
+static void psb_intel_opregion_asle_work(struct work_struct *work)
 {
-	struct drm_psb_private *dev_priv = dev->dev_private;
-	struct opregion_asle *asle = dev_priv->opregion.asle;
+	struct psb_intel_opregion *opregion =
+		container_of(work, struct psb_intel_opregion, asle_work);
+	struct drm_psb_private *dev_priv =
+		container_of(opregion, struct drm_psb_private, opregion);
+	struct opregion_asle *asle = opregion->asle;
 	u32 asle_stat = 0;
 	u32 asle_req;
 
@@ -191,9 +193,18 @@ void psb_intel_opregion_asle_intr(struct drm_device *dev)
 	}
 
 	if (asle_req & ASLE_SET_BACKLIGHT)
-		asle_stat |= asle_set_backlight(dev, asle->bclp);
+		asle_stat |= asle_set_backlight(dev_priv->dev, asle->bclp);
 
 	asle->aslc = asle_stat;
+
+}
+
+void psb_intel_opregion_asle_intr(struct drm_device *dev)
+{
+	struct drm_psb_private *dev_priv = dev->dev_private;
+
+	if (dev_priv->opregion.asle)
+		schedule_work(&dev_priv->opregion.asle_work);
 }
 
 #define ASLE_ALS_EN    (1<<0)
@@ -283,6 +294,8 @@ void psb_intel_opregion_fini(struct drm_device *dev)
 		unregister_acpi_notifier(&psb_intel_opregion_notifier);
 	}
 
+	cancel_work_sync(&opregion->asle_work);
+
 	/* just clear all opregion memory pointers now */
 	iounmap(opregion->header);
 	opregion->header = NULL;
@@ -305,6 +318,9 @@ int psb_intel_opregion_setup(struct drm_device *dev)
 		DRM_DEBUG_DRIVER("ACPI Opregion not supported\n");
 		return -ENOTSUPP;
 	}
+
+	INIT_WORK(&opregion->asle_work, psb_intel_opregion_asle_work);
+
 	DRM_DEBUG("OpRegion detected at 0x%8x\n", opregion_phy);
 	base = acpi_os_ioremap(opregion_phy, 8*1024);
 	if (!base)
diff --git a/drivers/gpu/drm/gma500/psb_device.c b/drivers/gpu/drm/gma500/psb_device.c
index 23fb33f1471..07df7d4eea7 100644
--- a/drivers/gpu/drm/gma500/psb_device.c
+++ b/drivers/gpu/drm/gma500/psb_device.c
@@ -26,6 +26,7 @@
 #include "psb_intel_reg.h"
 #include "intel_bios.h"
 #include "psb_device.h"
+#include "gma_device.h"
 
 static int psb_output_init(struct drm_device *dev)
 {
@@ -257,45 +258,6 @@ static int psb_power_up(struct drm_device *dev)
 	return 0;
 }
 
-static void psb_get_core_freq(struct drm_device *dev)
-{
-	uint32_t clock;
-	struct pci_dev *pci_root = pci_get_bus_and_slot(0, 0);
-	struct drm_psb_private *dev_priv = dev->dev_private;
-
-	/*pci_write_config_dword(pci_root, 0xD4, 0x00C32004);*/
-	/*pci_write_config_dword(pci_root, 0xD0, 0xE0033000);*/
-
-	pci_write_config_dword(pci_root, 0xD0, 0xD0050300);
-	pci_read_config_dword(pci_root, 0xD4, &clock);
-	pci_dev_put(pci_root);
-
-	switch (clock & 0x07) {
-	case 0:
-		dev_priv->core_freq = 100;
-		break;
-	case 1:
-		dev_priv->core_freq = 133;
-		break;
-	case 2:
-		dev_priv->core_freq = 150;
-		break;
-	case 3:
-		dev_priv->core_freq = 178;
-		break;
-	case 4:
-		dev_priv->core_freq = 200;
-		break;
-	case 5:
-	case 6:
-	case 7:
-		dev_priv->core_freq = 266;
-		break;
-	default:
-		dev_priv->core_freq = 0;
-	}
-}
-
 /* Poulsbo */
 static const struct psb_offset psb_regmap[2] = {
 	{
@@ -352,7 +314,7 @@ static int psb_chip_setup(struct drm_device *dev)
 {
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	dev_priv->regmap = psb_regmap;
-	psb_get_core_freq(dev);
+	gma_get_core_freq(dev);
 	gma_intel_setup_gmbus(dev);
 	psb_intel_opregion_init(dev);
 	psb_intel_init_bios(dev);
diff --git a/drivers/gpu/drm/gma500/psb_drv.c b/drivers/gpu/drm/gma500/psb_drv.c
index 1199180667c..6e8fe9ec02b 100644
--- a/drivers/gpu/drm/gma500/psb_drv.c
+++ b/drivers/gpu/drm/gma500/psb_drv.c
@@ -21,7 +21,6 @@
 
 #include <drm/drmP.h>
 #include <drm/drm.h>
-#include <drm/gma_drm.h>
 #include "psb_drv.h"
 #include "framebuffer.h"
 #include "psb_reg.h"
@@ -37,56 +36,65 @@
 #include <acpi/video.h>
 #include <linux/module.h>
 
-static int drm_psb_trap_pagefaults;
-
-static int psb_probe(struct pci_dev *pdev, const struct pci_device_id *ent);
-
-MODULE_PARM_DESC(trap_pagefaults, "Error and reset on MMU pagefaults");
-module_param_named(trap_pagefaults, drm_psb_trap_pagefaults, int, 0600);
-
+static struct drm_driver driver;
+static int psb_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent);
 
+/*
+ * The table below contains a mapping of the PCI vendor ID and the PCI Device ID
+ * to the different groups of PowerVR 5-series chip designs
+ *
+ * 0x8086 = Intel Corporation
+ *
+ * PowerVR SGX535    - Poulsbo    - Intel GMA 500, Intel Atom Z5xx
+ * PowerVR SGX535    - Moorestown - Intel GMA 600
+ * PowerVR SGX535    - Oaktrail   - Intel GMA 600, Intel Atom Z6xx, E6xx
+ * PowerVR SGX540    - Medfield   - Intel Atom Z2460
+ * PowerVR SGX544MP2 - Medfield   -
+ * PowerVR SGX545    - Cedartrail - Intel GMA 3600, Intel Atom D2500, N2600
+ * PowerVR SGX545    - Cedartrail - Intel GMA 3650, Intel Atom D2550, D2700,
+ *                                  N2800
+ */
 static DEFINE_PCI_DEVICE_TABLE(pciidlist) = {
 	{ 0x8086, 0x8108, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &psb_chip_ops },
 	{ 0x8086, 0x8109, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &psb_chip_ops },
 #if defined(CONFIG_DRM_GMA600)
-	{ 0x8086, 0x4100, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops},
-	{ 0x8086, 0x4101, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops},
-	{ 0x8086, 0x4102, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops},
-	{ 0x8086, 0x4103, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops},
-	{ 0x8086, 0x4104, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops},
-	{ 0x8086, 0x4105, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops},
-	{ 0x8086, 0x4106, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops},
-	{ 0x8086, 0x4107, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops},
-	/* Atom E620 */
-	{ 0x8086, 0x4108, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops},
+	{ 0x8086, 0x4100, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops },
+	{ 0x8086, 0x4101, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops },
+	{ 0x8086, 0x4102, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops },
+	{ 0x8086, 0x4103, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops },
+	{ 0x8086, 0x4104, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops },
+	{ 0x8086, 0x4105, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops },
+	{ 0x8086, 0x4106, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops },
+	{ 0x8086, 0x4107, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops },
+	{ 0x8086, 0x4108, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &oaktrail_chip_ops },
 #endif
 #if defined(CONFIG_DRM_MEDFIELD)
-	{0x8086, 0x0130, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops},
-	{0x8086, 0x0131, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops},
-	{0x8086, 0x0132, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops},
-	{0x8086, 0x0133, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops},
-	{0x8086, 0x0134, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops},
-	{0x8086, 0x0135, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops},
-	{0x8086, 0x0136, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops},
-	{0x8086, 0x0137, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops},
+	{ 0x8086, 0x0130, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops },
+	{ 0x8086, 0x0131, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops },
+	{ 0x8086, 0x0132, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops },
+	{ 0x8086, 0x0133, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops },
+	{ 0x8086, 0x0134, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops },
+	{ 0x8086, 0x0135, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops },
+	{ 0x8086, 0x0136, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops },
+	{ 0x8086, 0x0137, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &mdfld_chip_ops },
 #endif
 #if defined(CONFIG_DRM_GMA3600)
-	{ 0x8086, 0x0be0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0be1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0be2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0be3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0be4, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0be5, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0be6, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0be7, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0be8, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0be9, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0bea, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0beb, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0bec, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0bed, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0bee, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
-	{ 0x8086, 0x0bef, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops},
+	{ 0x8086, 0x0be0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0be1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0be2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0be3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0be4, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0be5, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0be6, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0be7, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0be8, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0be9, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0bea, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0beb, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0bec, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0bed, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0bee, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
+	{ 0x8086, 0x0bef, PCI_ANY_ID, PCI_ANY_ID, 0, 0, (long) &cdv_chip_ops },
 #endif
 	{ 0, }
 };
@@ -95,69 +103,18 @@ MODULE_DEVICE_TABLE(pci, pciidlist);
 /*
  * Standard IOCTLs.
  */
-
-#define DRM_IOCTL_GMA_ADB	\
-		DRM_IOWR(DRM_GMA_ADB + DRM_COMMAND_BASE, uint32_t)
-#define DRM_IOCTL_GMA_MODE_OPERATION	\
-		DRM_IOWR(DRM_GMA_MODE_OPERATION + DRM_COMMAND_BASE, \
-			 struct drm_psb_mode_operation_arg)
-#define DRM_IOCTL_GMA_STOLEN_MEMORY	\
-		DRM_IOWR(DRM_GMA_STOLEN_MEMORY + DRM_COMMAND_BASE, \
-			 struct drm_psb_stolen_memory_arg)
-#define DRM_IOCTL_GMA_GAMMA	\
-		DRM_IOWR(DRM_GMA_GAMMA + DRM_COMMAND_BASE, \
-			 struct drm_psb_dpst_lut_arg)
-#define DRM_IOCTL_GMA_DPST_BL	\
-		DRM_IOWR(DRM_GMA_DPST_BL + DRM_COMMAND_BASE, \
-			 uint32_t)
-#define DRM_IOCTL_GMA_GET_PIPE_FROM_CRTC_ID	\
-		DRM_IOWR(DRM_GMA_GET_PIPE_FROM_CRTC_ID + DRM_COMMAND_BASE, \
-			 struct drm_psb_get_pipe_from_crtc_id_arg)
-#define DRM_IOCTL_GMA_GEM_CREATE	\
-		DRM_IOWR(DRM_GMA_GEM_CREATE + DRM_COMMAND_BASE, \
-			 struct drm_psb_gem_create)
-#define DRM_IOCTL_GMA_GEM_MMAP	\
-		DRM_IOWR(DRM_GMA_GEM_MMAP + DRM_COMMAND_BASE, \
-			 struct drm_psb_gem_mmap)
-
-static int psb_adb_ioctl(struct drm_device *dev, void *data,
-			 struct drm_file *file_priv);
-static int psb_mode_operation_ioctl(struct drm_device *dev, void *data,
-				    struct drm_file *file_priv);
-static int psb_stolen_memory_ioctl(struct drm_device *dev, void *data,
-				   struct drm_file *file_priv);
-static int psb_gamma_ioctl(struct drm_device *dev, void *data,
-			   struct drm_file *file_priv);
-static int psb_dpst_bl_ioctl(struct drm_device *dev, void *data,
-			     struct drm_file *file_priv);
-
 static const struct drm_ioctl_desc psb_ioctls[] = {
-	DRM_IOCTL_DEF_DRV(GMA_ADB, psb_adb_ioctl, DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(GMA_MODE_OPERATION, psb_mode_operation_ioctl,
-		      DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(GMA_STOLEN_MEMORY, psb_stolen_memory_ioctl,
-		      DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(GMA_GAMMA, psb_gamma_ioctl, DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(GMA_DPST_BL, psb_dpst_bl_ioctl, DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(GMA_GET_PIPE_FROM_CRTC_ID,
-					psb_intel_get_pipe_from_crtc_id, 0),
-	DRM_IOCTL_DEF_DRV(GMA_GEM_CREATE, psb_gem_create_ioctl,
-						DRM_UNLOCKED | DRM_AUTH),
-	DRM_IOCTL_DEF_DRV(GMA_GEM_MMAP, psb_gem_mmap_ioctl,
-						DRM_UNLOCKED | DRM_AUTH),
 };
 
-static void psb_lastclose(struct drm_device *dev)
+static void psb_driver_lastclose(struct drm_device *dev)
 {
 	int ret;
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	struct psb_fbdev *fbdev = dev_priv->fbdev;
 
-	drm_modeset_lock_all(dev);
-	ret = drm_fb_helper_restore_fbdev_mode(&fbdev->psb_fb_helper);
+	ret = drm_fb_helper_restore_fbdev_mode_unlocked(&fbdev->psb_fb_helper);
 	if (ret)
 		DRM_DEBUG("failed to restore crtc mode\n");
-	drm_modeset_unlock_all(dev);
 
 	return;
 }
@@ -169,19 +126,14 @@ static int psb_do_init(struct drm_device *dev)
 
 	uint32_t stolen_gtt;
 
-	int ret = -ENOMEM;
-
 	if (pg->mmu_gatt_start & 0x0FFFFFFF) {
 		dev_err(dev->dev, "Gatt must be 256M aligned. This is a bug.\n");
-		ret = -EINVAL;
-		goto out_err;
+		return -EINVAL;
 	}
 
-
 	stolen_gtt = (pg->stolen_size >> PAGE_SHIFT) * 4;
 	stolen_gtt = (stolen_gtt + PAGE_SIZE - 1) >> PAGE_SHIFT;
-	stolen_gtt =
-	    (stolen_gtt < pg->gtt_pages) ? stolen_gtt : pg->gtt_pages;
+	stolen_gtt = (stolen_gtt < pg->gtt_pages) ? stolen_gtt : pg->gtt_pages;
 
 	dev_priv->gatt_free_offset = pg->mmu_gatt_start +
 	    (stolen_gtt << PAGE_SHIFT) * 1024;
@@ -192,23 +144,26 @@ static int psb_do_init(struct drm_device *dev)
 	PSB_WSGX32(0x00000000, PSB_CR_BIF_BANK0);
 	PSB_WSGX32(0x00000000, PSB_CR_BIF_BANK1);
 	PSB_RSGX32(PSB_CR_BIF_BANK1);
-	PSB_WSGX32(PSB_RSGX32(PSB_CR_BIF_CTRL) | _PSB_MMU_ER_MASK,
-							PSB_CR_BIF_CTRL);
+
+	/* Do not bypass any MMU access, let them pagefault instead */
+	PSB_WSGX32((PSB_RSGX32(PSB_CR_BIF_CTRL) & ~_PSB_MMU_ER_MASK),
+		   PSB_CR_BIF_CTRL);
+	PSB_RSGX32(PSB_CR_BIF_CTRL);
+
 	psb_spank(dev_priv);
 
 	/* mmu_gatt ?? */
 	PSB_WSGX32(pg->gatt_start, PSB_CR_BIF_TWOD_REQ_BASE);
+	PSB_RSGX32(PSB_CR_BIF_TWOD_REQ_BASE); /* Post */
+
 	return 0;
-out_err:
-	return ret;
 }
 
 static int psb_driver_unload(struct drm_device *dev)
 {
 	struct drm_psb_private *dev_priv = dev->dev_private;
 
-	/* Kill vblank etc here */
-
+	/* TODO: Kill vblank etc here */
 
 	if (dev_priv) {
 		if (dev_priv->backlight_device)
@@ -268,8 +223,7 @@ static int psb_driver_unload(struct drm_device *dev)
 	return 0;
 }
 
-
-static int psb_driver_load(struct drm_device *dev, unsigned long chipset)
+static int psb_driver_load(struct drm_device *dev, unsigned long flags)
 {
 	struct drm_psb_private *dev_priv;
 	unsigned long resource_start, resource_len;
@@ -277,15 +231,19 @@ static int psb_driver_load(struct drm_device *dev, unsigned long chipset)
 	int ret = -ENOMEM;
 	struct drm_connector *connector;
 	struct gma_encoder *gma_encoder;
+	struct psb_gtt *pg;
 
+	/* allocating and initializing driver private data */
 	dev_priv = kzalloc(sizeof(*dev_priv), GFP_KERNEL);
 	if (dev_priv == NULL)
 		return -ENOMEM;
 
-	dev_priv->ops = (struct psb_ops *)chipset;
+	dev_priv->ops = (struct psb_ops *)flags;
 	dev_priv->dev = dev;
 	dev->dev_private = (void *) dev_priv;
 
+	pg = &dev_priv->gtt;
+
 	pci_set_master(dev->pdev);
 
 	dev_priv->num_pipe = dev_priv->ops->pipes;
@@ -347,9 +305,7 @@ static int psb_driver_load(struct drm_device *dev, unsigned long chipset)
 	if (ret)
 		goto out_err;
 
-	dev_priv->mmu = psb_mmu_driver_init((void *)0,
-					drm_psb_trap_pagefaults, 0,
-					dev_priv);
+	dev_priv->mmu = psb_mmu_driver_init(dev, 1, 0, 0);
 	if (!dev_priv->mmu)
 		goto out_err;
 
@@ -357,18 +313,27 @@ static int psb_driver_load(struct drm_device *dev, unsigned long chipset)
 	if (!dev_priv->pf_pd)
 		goto out_err;
 
-	psb_mmu_set_pd_context(psb_mmu_get_default_pd(dev_priv->mmu), 0);
-	psb_mmu_set_pd_context(dev_priv->pf_pd, 1);
-
 	ret = psb_do_init(dev);
 	if (ret)
 		return ret;
 
+	/* Add stolen memory to SGX MMU */
+	down_read(&pg->sem);
+	ret = psb_mmu_insert_pfn_sequence(psb_mmu_get_default_pd(dev_priv->mmu),
+					  dev_priv->stolen_base >> PAGE_SHIFT,
+					  pg->gatt_start,
+					  pg->stolen_size >> PAGE_SHIFT, 0);
+	up_read(&pg->sem);
+
+	psb_mmu_set_pd_context(psb_mmu_get_default_pd(dev_priv->mmu), 0);
+	psb_mmu_set_pd_context(dev_priv->pf_pd, 1);
+
 	PSB_WSGX32(0x20000000, PSB_CR_PDS_EXEC_BASE);
 	PSB_WSGX32(0x30000000, PSB_CR_BIF_3D_REQ_BASE);
 
 	acpi_video_register();
 
+	/* Setup vertical blanking handling */
 	ret = drm_vblank_init(dev, dev_priv->num_pipe);
 	if (ret)
 		goto out_err;
@@ -387,12 +352,10 @@ static int psb_driver_load(struct drm_device *dev, unsigned long chipset)
 	PSB_WVDC32(0xFFFFFFFF, PSB_INT_MASK_R);
 	spin_unlock_irqrestore(&dev_priv->irqmask_lock, irqflags);
 
-	drm_irq_install(dev);
+	drm_irq_install(dev, dev->pdev->irq);
 
 	dev->vblank_disable_allowed = true;
-
 	dev->max_vblank_count = 0xffffff; /* only 24 bits of frame count */
-
 	dev->driver->get_vblank_counter = psb_get_vblank_counter;
 
 	psb_modeset_init(dev);
@@ -416,11 +379,11 @@ static int psb_driver_load(struct drm_device *dev, unsigned long chipset)
 		return ret;
 	psb_intel_opregion_enable_asle(dev);
 #if 0
-	/*enable runtime pm at last*/
+	/* Enable runtime pm at last */
 	pm_runtime_enable(&dev->pdev->dev);
 	pm_runtime_set_active(&dev->pdev->dev);
 #endif
-	/*Intel drm driver load is done, continue doing pvr load*/
+	/* Intel drm driver load is done, continue doing pvr load */
 	return 0;
 out_err:
 	psb_driver_unload(dev);
@@ -442,161 +405,6 @@ static inline void get_brightness(struct backlight_device *bd)
 #endif
 }
 
-static int psb_dpst_bl_ioctl(struct drm_device *dev, void *data,
-		       struct drm_file *file_priv)
-{
-	struct drm_psb_private *dev_priv = psb_priv(dev);
-	uint32_t *arg = data;
-
-	dev_priv->blc_adj2 = *arg;
-	get_brightness(dev_priv->backlight_device);
-	return 0;
-}
-
-static int psb_adb_ioctl(struct drm_device *dev, void *data,
-			struct drm_file *file_priv)
-{
-	struct drm_psb_private *dev_priv = psb_priv(dev);
-	uint32_t *arg = data;
-
-	dev_priv->blc_adj1 = *arg;
-	get_brightness(dev_priv->backlight_device);
-	return 0;
-}
-
-static int psb_gamma_ioctl(struct drm_device *dev, void *data,
-			   struct drm_file *file_priv)
-{
-	struct drm_psb_dpst_lut_arg *lut_arg = data;
-	struct drm_mode_object *obj;
-	struct drm_crtc *crtc;
-	struct drm_connector *connector;
-	struct gma_crtc *gma_crtc;
-	int i = 0;
-	int32_t obj_id;
-
-	obj_id = lut_arg->output_id;
-	obj = drm_mode_object_find(dev, obj_id, DRM_MODE_OBJECT_CONNECTOR);
-	if (!obj) {
-		dev_dbg(dev->dev, "Invalid Connector object.\n");
-		return -ENOENT;
-	}
-
-	connector = obj_to_connector(obj);
-	crtc = connector->encoder->crtc;
-	gma_crtc = to_gma_crtc(crtc);
-
-	for (i = 0; i < 256; i++)
-		gma_crtc->lut_adj[i] = lut_arg->lut[i];
-
-	gma_crtc_load_lut(crtc);
-
-	return 0;
-}
-
-static int psb_mode_operation_ioctl(struct drm_device *dev, void *data,
-				struct drm_file *file_priv)
-{
-	uint32_t obj_id;
-	uint16_t op;
-	struct drm_mode_modeinfo *umode;
-	struct drm_display_mode *mode = NULL;
-	struct drm_psb_mode_operation_arg *arg;
-	struct drm_mode_object *obj;
-	struct drm_connector *connector;
-	struct drm_connector_helper_funcs *connector_funcs;
-	int ret = 0;
-	int resp = MODE_OK;
-
-	arg = (struct drm_psb_mode_operation_arg *)data;
-	obj_id = arg->obj_id;
-	op = arg->operation;
-
-	switch (op) {
-	case PSB_MODE_OPERATION_MODE_VALID:
-		umode = &arg->mode;
-
-		drm_modeset_lock_all(dev);
-
-		obj = drm_mode_object_find(dev, obj_id,
-					DRM_MODE_OBJECT_CONNECTOR);
-		if (!obj) {
-			ret = -ENOENT;
-			goto mode_op_out;
-		}
-
-		connector = obj_to_connector(obj);
-
-		mode = drm_mode_create(dev);
-		if (!mode) {
-			ret = -ENOMEM;
-			goto mode_op_out;
-		}
-
-		/* drm_crtc_convert_umode(mode, umode); */
-		{
-			mode->clock = umode->clock;
-			mode->hdisplay = umode->hdisplay;
-			mode->hsync_start = umode->hsync_start;
-			mode->hsync_end = umode->hsync_end;
-			mode->htotal = umode->htotal;
-			mode->hskew = umode->hskew;
-			mode->vdisplay = umode->vdisplay;
-			mode->vsync_start = umode->vsync_start;
-			mode->vsync_end = umode->vsync_end;
-			mode->vtotal = umode->vtotal;
-			mode->vscan = umode->vscan;
-			mode->vrefresh = umode->vrefresh;
-			mode->flags = umode->flags;
-			mode->type = umode->type;
-			strncpy(mode->name, umode->name, DRM_DISPLAY_MODE_LEN);
-			mode->name[DRM_DISPLAY_MODE_LEN-1] = 0;
-		}
-
-		connector_funcs = (struct drm_connector_helper_funcs *)
-				   connector->helper_private;
-
-		if (connector_funcs->mode_valid) {
-			resp = connector_funcs->mode_valid(connector, mode);
-			arg->data = resp;
-		}
-
-		/*do some clean up work*/
-		if (mode)
-			drm_mode_destroy(dev, mode);
-mode_op_out:
-		drm_modeset_unlock_all(dev);
-		return ret;
-
-	default:
-		dev_dbg(dev->dev, "Unsupported psb mode operation\n");
-		return -EOPNOTSUPP;
-	}
-
-	return 0;
-}
-
-static int psb_stolen_memory_ioctl(struct drm_device *dev, void *data,
-				   struct drm_file *file_priv)
-{
-	struct drm_psb_private *dev_priv = psb_priv(dev);
-	struct drm_psb_stolen_memory_arg *arg = data;
-
-	arg->base = dev_priv->stolen_base;
-	arg->size = dev_priv->vram_stolen_size;
-
-	return 0;
-}
-
-static int psb_driver_open(struct drm_device *dev, struct drm_file *priv)
-{
-	return 0;
-}
-
-static void psb_driver_close(struct drm_device *dev, struct drm_file *priv)
-{
-}
-
 static long psb_unlocked_ioctl(struct file *filp, unsigned int cmd,
 			       unsigned long arg)
 {
@@ -614,15 +422,21 @@ static long psb_unlocked_ioctl(struct file *filp, unsigned int cmd,
 	/* FIXME: do we need to wrap the other side of this */
 }
 
-
-/* When a client dies:
+/*
+ * When a client dies:
  *    - Check for and clean up flipped page state
  */
 static void psb_driver_preclose(struct drm_device *dev, struct drm_file *priv)
 {
 }
 
-static void psb_remove(struct pci_dev *pdev)
+static int psb_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	return drm_get_pci_dev(pdev, ent, &driver);
+}
+
+
+static void psb_pci_remove(struct pci_dev *pdev)
 {
 	struct drm_device *dev = pci_get_drvdata(pdev);
 	drm_put_dev(dev);
@@ -657,12 +471,13 @@ static const struct file_operations psb_gem_fops = {
 
 static struct drm_driver driver = {
 	.driver_features = DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED | \
-			   DRIVER_MODESET | DRIVER_GEM ,
+			   DRIVER_MODESET | DRIVER_GEM,
 	.load = psb_driver_load,
 	.unload = psb_driver_unload,
+	.lastclose = psb_driver_lastclose,
+	.preclose = psb_driver_preclose,
 
-	.ioctls = psb_ioctls,
-	.num_ioctls = DRM_ARRAY_SIZE(psb_ioctls),
+	.num_ioctls = ARRAY_SIZE(psb_ioctls),
 	.device_is_agp = psb_driver_device_is_agp,
 	.irq_preinstall = psb_irq_preinstall,
 	.irq_postinstall = psb_irq_postinstall,
@@ -671,40 +486,31 @@ static struct drm_driver driver = {
 	.enable_vblank = psb_enable_vblank,
 	.disable_vblank = psb_disable_vblank,
 	.get_vblank_counter = psb_get_vblank_counter,
-	.lastclose = psb_lastclose,
-	.open = psb_driver_open,
-	.preclose = psb_driver_preclose,
-	.postclose = psb_driver_close,
 
 	.gem_free_object = psb_gem_free_object,
 	.gem_vm_ops = &psb_gem_vm_ops,
+
 	.dumb_create = psb_gem_dumb_create,
 	.dumb_map_offset = psb_gem_dumb_map_gtt,
 	.dumb_destroy = drm_gem_dumb_destroy,
+	.ioctls = psb_ioctls,
 	.fops = &psb_gem_fops,
 	.name = DRIVER_NAME,
 	.desc = DRIVER_DESC,
-	.date = PSB_DRM_DRIVER_DATE,
-	.major = PSB_DRM_DRIVER_MAJOR,
-	.minor = PSB_DRM_DRIVER_MINOR,
-	.patchlevel = PSB_DRM_DRIVER_PATCHLEVEL
+	.date = DRIVER_DATE,
+	.major = DRIVER_MAJOR,
+	.minor = DRIVER_MINOR,
+	.patchlevel = DRIVER_PATCHLEVEL
 };
 
 static struct pci_driver psb_pci_driver = {
 	.name = DRIVER_NAME,
 	.id_table = pciidlist,
-	.probe = psb_probe,
-	.remove = psb_remove,
-	.driver = {
-		.pm = &psb_pm_ops,
-	}
+	.probe = psb_pci_probe,
+	.remove = psb_pci_remove,
+	.driver.pm = &psb_pm_ops,
 };
 
-static int psb_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
-{
-	return drm_get_pci_dev(pdev, ent, &driver);
-}
-
 static int __init psb_init(void)
 {
 	return drm_pci_init(&driver, &psb_pci_driver);
@@ -718,6 +524,6 @@ static void __exit psb_exit(void)
 late_initcall(psb_init);
 module_exit(psb_exit);
 
-MODULE_AUTHOR("Alan Cox <alan@linux.intel.com> and others");
+MODULE_AUTHOR(DRIVER_AUTHOR);
 MODULE_DESCRIPTION(DRIVER_DESC);
-MODULE_LICENSE("GPL");
+MODULE_LICENSE(DRIVER_LICENSE);
diff --git a/drivers/gpu/drm/gma500/psb_drv.h b/drivers/gpu/drm/gma500/psb_drv.h
index b59e6588c34..55ebe2bd88d 100644
--- a/drivers/gpu/drm/gma500/psb_drv.h
+++ b/drivers/gpu/drm/gma500/psb_drv.h
@@ -33,6 +33,18 @@
 #include "power.h"
 #include "opregion.h"
 #include "oaktrail.h"
+#include "mmu.h"
+
+#define DRIVER_AUTHOR "Alan Cox <alan@linux.intel.com> and others"
+#define DRIVER_LICENSE "GPL"
+
+#define DRIVER_NAME "gma500"
+#define DRIVER_DESC "DRM driver for the Intel GMA500, GMA600, GMA3600, GMA3650"
+#define DRIVER_DATE "20140314"
+
+#define DRIVER_MAJOR 1
+#define DRIVER_MINOR 0
+#define DRIVER_PATCHLEVEL 0
 
 /* Append new drm mode definition here, align with libdrm definition */
 #define DRM_MODE_SCALE_NO_SCALE   	2
@@ -49,21 +61,7 @@ enum {
 #define IS_MFLD(dev) (((dev)->pdev->device & 0xfff8) == 0x0130)
 #define IS_CDV(dev) (((dev)->pdev->device & 0xfff0) == 0x0be0)
 
-/*
- * Driver definitions
- */
-
-#define DRIVER_NAME "gma500"
-#define DRIVER_DESC "DRM driver for the Intel GMA500"
-
-#define PSB_DRM_DRIVER_DATE "2011-06-06"
-#define PSB_DRM_DRIVER_MAJOR 1
-#define PSB_DRM_DRIVER_MINOR 0
-#define PSB_DRM_DRIVER_PATCHLEVEL 0
-
-/*
- *	Hardware offsets
- */
+/* Hardware offsets */
 #define PSB_VDC_OFFSET		 0x00000000
 #define PSB_VDC_SIZE		 0x000080000
 #define MRST_MMIO_SIZE		 0x0000C0000
@@ -71,16 +69,14 @@ enum {
 #define PSB_SGX_SIZE		 0x8000
 #define PSB_SGX_OFFSET		 0x00040000
 #define MRST_SGX_OFFSET		 0x00080000
-/*
- *	PCI resource identifiers
- */
+
+/* PCI resource identifiers */
 #define PSB_MMIO_RESOURCE	 0
 #define PSB_AUX_RESOURCE	 0
 #define PSB_GATT_RESOURCE	 2
 #define PSB_GTT_RESOURCE	 3
-/*
- *	PCI configuration
- */
+
+/* PCI configuration */
 #define PSB_GMCH_CTRL		 0x52
 #define PSB_BSM			 0x5C
 #define _PSB_GMCH_ENABLED	 0x4
@@ -88,37 +84,29 @@ enum {
 #define _PSB_PGETBL_ENABLED	 0x00000001
 #define PSB_SGX_2D_SLAVE_PORT	 0x4000
 
-/* To get rid of */
+/* TODO: To get rid of */
 #define PSB_TT_PRIV0_LIMIT	 (256*1024*1024)
 #define PSB_TT_PRIV0_PLIMIT	 (PSB_TT_PRIV0_LIMIT >> PAGE_SHIFT)
 
-/*
- *	SGX side MMU definitions (these can probably go)
- */
+/* SGX side MMU definitions (these can probably go) */
 
-/*
- *	Flags for external memory type field.
- */
+/* Flags for external memory type field */
 #define PSB_MMU_CACHED_MEMORY	  0x0001	/* Bind to MMU only */
 #define PSB_MMU_RO_MEMORY	  0x0002	/* MMU RO memory */
 #define PSB_MMU_WO_MEMORY	  0x0004	/* MMU WO memory */
-/*
- *	PTE's and PDE's
- */
+
+/* PTE's and PDE's */
 #define PSB_PDE_MASK		  0x003FFFFF
 #define PSB_PDE_SHIFT		  22
 #define PSB_PTE_SHIFT		  12
-/*
- *	Cache control
- */
+
+/* Cache control */
 #define PSB_PTE_VALID		  0x0001	/* PTE / PDE valid */
 #define PSB_PTE_WO		  0x0002	/* Write only */
 #define PSB_PTE_RO		  0x0004	/* Read only */
 #define PSB_PTE_CACHED		  0x0008	/* CPU cache coherent */
 
-/*
- *	VDC registers and bits
- */
+/* VDC registers and bits */
 #define PSB_MSVDX_CLOCKGATING	  0x2064
 #define PSB_TOPAZ_CLOCKGATING	  0x2068
 #define PSB_HWSTAM		  0x2098
@@ -212,8 +200,8 @@ enum {
 #define PSB_HIGH_REG_OFFS 0x0600
 
 #define PSB_NUM_VBLANKS 2
-#define PSB_WATCHDOG_DELAY (DRM_HZ * 2)
-#define PSB_LID_DELAY (DRM_HZ / 10)
+#define PSB_WATCHDOG_DELAY (HZ * 2)
+#define PSB_LID_DELAY (HZ / 10)
 
 #define MDFLD_PNW_B0 0x04
 #define MDFLD_PNW_C0 0x08
@@ -232,7 +220,7 @@ enum {
 #define MDFLD_DSR_RR		45
 #define MDFLD_DPU_ENABLE 	(1 << 31)
 #define MDFLD_DSR_FULLSCREEN 	(1 << 30)
-#define MDFLD_DSR_DELAY		(DRM_HZ / MDFLD_DSR_RR)
+#define MDFLD_DSR_DELAY		(HZ / MDFLD_DSR_RR)
 
 #define PSB_PWR_STATE_ON		1
 #define PSB_PWR_STATE_OFF		2
@@ -265,6 +253,7 @@ struct psb_intel_opregion {
 	struct opregion_asle *asle;
 	void *vbt;
 	u32 __iomem *lid_state;
+	struct work_struct asle_work;
 };
 
 struct sdvo_device_mapping {
@@ -283,10 +272,7 @@ struct intel_gmbus {
 	u32 reg0;
 };
 
-/*
- *	Register offset maps
- */
-
+/* Register offset maps */
 struct psb_offset {
 	u32	fp0;
 	u32	fp1;
@@ -320,9 +306,7 @@ struct psb_offset {
  *	update the register cache instead.
  */
 
-/*
- *	Common status for pipes.
- */
+/* Common status for pipes */
 struct psb_pipe {
 	u32	fp0;
 	u32	fp1;
@@ -482,35 +466,24 @@ struct drm_psb_private {
 	struct psb_mmu_driver *mmu;
 	struct psb_mmu_pd *pf_pd;
 
-	/*
-	 * Register base
-	 */
-
+	/* Register base */
 	uint8_t __iomem *sgx_reg;
 	uint8_t __iomem *vdc_reg;
 	uint8_t __iomem *aux_reg; /* Auxillary vdc pipe regs */
 	uint32_t gatt_free_offset;
 
-	/*
-	 * Fencing / irq.
-	 */
-
+	/* Fencing / irq */
 	uint32_t vdc_irq_mask;
 	uint32_t pipestat[PSB_NUM_PIPE];
 
 	spinlock_t irqmask_lock;
 
-	/*
-	 * Power
-	 */
-
+	/* Power */
 	bool suspended;
 	bool display_power;
 	int display_count;
 
-	/*
-	 * Modesetting
-	 */
+	/* Modesetting */
 	struct psb_intel_mode_device mode_dev;
 	bool modeset;	/* true if we have done the mode_device setup */
 
@@ -518,15 +491,10 @@ struct drm_psb_private {
 	struct drm_crtc *pipe_to_crtc_mapping[PSB_NUM_PIPE];
 	uint32_t num_pipe;
 
-	/*
-	 * OSPM info (Power management base) (can go ?)
-	 */
+	/* OSPM info (Power management base) (TODO: can go ?) */
 	uint32_t ospm_base;
 
-	/*
-	 * Sizes info
-	 */
-
+	/* Sizes info */
 	u32 fuse_reg_value;
 	u32 video_device_fuse;
 
@@ -546,9 +514,7 @@ struct drm_psb_private {
 	struct drm_property *broadcast_rgb_property;
 	struct drm_property *force_audio_property;
 
-	/*
-	 * LVDS info
-	 */
+	/* LVDS info */
 	int backlight_duty_cycle;	/* restore backlight to this value */
 	bool panel_wants_dither;
 	struct drm_display_mode *panel_fixed_mode;
@@ -582,34 +548,23 @@ struct drm_psb_private {
 	/* Oaktrail HDMI state */
 	struct oaktrail_hdmi_dev *hdmi_priv;
 	
-	/*
-	 * Register state
-	 */
-
+	/* Register state */
 	struct psb_save_area regs;
 
 	/* MSI reg save */
 	uint32_t msi_addr;
 	uint32_t msi_data;
 
-	/*
-	 * Hotplug handling
-	 */
-
+	/* Hotplug handling */
 	struct work_struct hotplug_work;
 
-	/*
-	 * LID-Switch
-	 */
+	/* LID-Switch */
 	spinlock_t lid_lock;
 	struct timer_list lid_timer;
 	struct psb_intel_opregion opregion;
 	u32 lid_last_state;
 
-	/*
-	 * Watchdog
-	 */
-
+	/* Watchdog */
 	uint32_t apm_reg;
 	uint16_t apm_base;
 
@@ -629,9 +584,7 @@ struct drm_psb_private {
 	/* 2D acceleration */
 	spinlock_t lock_2d;
 
-	/*
-	 * Panel brightness
-	 */
+	/* Panel brightness */
 	int brightness;
 	int brightness_adjusted;
 
@@ -664,10 +617,7 @@ struct drm_psb_private {
 };
 
 
-/*
- *	Operations for each board type
- */
- 
+/* Operations for each board type */
 struct psb_ops {
 	const char *name;
 	unsigned int accel_2d:1;
@@ -713,8 +663,6 @@ struct psb_ops {
 
 
 
-struct psb_mmu_driver;
-
 extern int drm_crtc_probe_output_modes(struct drm_device *dev, int, int);
 extern int drm_pick_crtcs(struct drm_device *dev);
 
@@ -723,53 +671,8 @@ static inline struct drm_psb_private *psb_priv(struct drm_device *dev)
 	return (struct drm_psb_private *) dev->dev_private;
 }
 
-/*
- * MMU stuff.
- */
-
-extern struct psb_mmu_driver *psb_mmu_driver_init(uint8_t __iomem * registers,
-					int trap_pagefaults,
-					int invalid_type,
-					struct drm_psb_private *dev_priv);
-extern void psb_mmu_driver_takedown(struct psb_mmu_driver *driver);
-extern struct psb_mmu_pd *psb_mmu_get_default_pd(struct psb_mmu_driver
-						 *driver);
-extern void psb_mmu_mirror_gtt(struct psb_mmu_pd *pd, uint32_t mmu_offset,
-			       uint32_t gtt_start, uint32_t gtt_pages);
-extern struct psb_mmu_pd *psb_mmu_alloc_pd(struct psb_mmu_driver *driver,
-					   int trap_pagefaults,
-					   int invalid_type);
-extern void psb_mmu_free_pagedir(struct psb_mmu_pd *pd);
-extern void psb_mmu_flush(struct psb_mmu_driver *driver, int rc_prot);
-extern void psb_mmu_remove_pfn_sequence(struct psb_mmu_pd *pd,
-					unsigned long address,
-					uint32_t num_pages);
-extern int psb_mmu_insert_pfn_sequence(struct psb_mmu_pd *pd,
-				       uint32_t start_pfn,
-				       unsigned long address,
-				       uint32_t num_pages, int type);
-extern int psb_mmu_virtual_to_pfn(struct psb_mmu_pd *pd, uint32_t virtual,
-				  unsigned long *pfn);
-
-/*
- * Enable / disable MMU for different requestors.
- */
-
-
-extern void psb_mmu_set_pd_context(struct psb_mmu_pd *pd, int hw_context);
-extern int psb_mmu_insert_pages(struct psb_mmu_pd *pd, struct page **pages,
-				unsigned long address, uint32_t num_pages,
-				uint32_t desired_tile_stride,
-				uint32_t hw_tile_stride, int type);
-extern void psb_mmu_remove_pages(struct psb_mmu_pd *pd,
-				 unsigned long address, uint32_t num_pages,
-				 uint32_t desired_tile_stride,
-				 uint32_t hw_tile_stride);
-/*
- *psb_irq.c
- */
-
-extern irqreturn_t psb_irq_handler(DRM_IRQ_ARGS);
+/* psb_irq.c */
+extern irqreturn_t psb_irq_handler(int irq, void *arg);
 extern int psb_irq_enable_dpst(struct drm_device *dev);
 extern int psb_irq_disable_dpst(struct drm_device *dev);
 extern void psb_irq_preinstall(struct drm_device *dev);
@@ -791,24 +694,17 @@ psb_disable_pipestat(struct drm_psb_private *dev_priv, int pipe, u32 mask);
 
 extern u32 psb_get_vblank_counter(struct drm_device *dev, int crtc);
 
-/*
- * framebuffer.c
- */
+/* framebuffer.c */
 extern int psbfb_probed(struct drm_device *dev);
 extern int psbfb_remove(struct drm_device *dev,
 			struct drm_framebuffer *fb);
-/*
- * accel_2d.c
- */
+/* accel_2d.c */
 extern void psbfb_copyarea(struct fb_info *info,
 					const struct fb_copyarea *region);
 extern int psbfb_sync(struct fb_info *info);
 extern void psb_spank(struct drm_psb_private *dev_priv);
 
-/*
- * psb_reset.c
- */
-
+/* psb_reset.c */
 extern void psb_lid_timer_init(struct drm_psb_private *dev_priv);
 extern void psb_lid_timer_takedown(struct drm_psb_private *dev_priv);
 extern void psb_print_pagefault(struct drm_psb_private *dev_priv);
@@ -867,9 +763,7 @@ extern const struct psb_ops mdfld_chip_ops;
 /* cdv_device.c */
 extern const struct psb_ops cdv_chip_ops;
 
-/*
- * Debug print bits setting
- */
+/* Debug print bits setting */
 #define PSB_D_GENERAL (1 << 0)
 #define PSB_D_INIT    (1 << 1)
 #define PSB_D_IRQ     (1 << 2)
@@ -885,10 +779,7 @@ extern const struct psb_ops cdv_chip_ops;
 
 extern int drm_idle_check_interval;
 
-/*
- *	Utilities
- */
-
+/* Utilities */
 static inline u32 MRST_MSG_READ32(uint port, uint offset)
 {
 	int mcr = (0xD0<<24) | (port << 16) | (offset << 8);
diff --git a/drivers/gpu/drm/gma500/psb_intel_display.c b/drivers/gpu/drm/gma500/psb_intel_display.c
index c8841ac6c8f..87b50ba64ed 100644
--- a/drivers/gpu/drm/gma500/psb_intel_display.c
+++ b/drivers/gpu/drm/gma500/psb_intel_display.c
@@ -120,7 +120,7 @@ static int psb_intel_crtc_mode_set(struct drm_crtc *crtc,
 	const struct gma_limit_t *limit;
 
 	/* No scan out no play */
-	if (crtc->fb == NULL) {
+	if (crtc->primary->fb == NULL) {
 		crtc_funcs->mode_set_base(crtc, x, y, old_fb);
 		return 0;
 	}
@@ -469,7 +469,8 @@ static void psb_intel_cursor_init(struct drm_device *dev,
 		/* Allocate 4 pages of stolen mem for a hardware cursor. That
 		 * is enough for the 64 x 64 ARGB cursors we support.
 		 */
-		cursor_gt = psb_gtt_alloc_range(dev, 4 * PAGE_SIZE, "cursor", 1);
+		cursor_gt = psb_gtt_alloc_range(dev, 4 * PAGE_SIZE, "cursor", 1,
+						PAGE_SIZE);
 		if (!cursor_gt) {
 			gma_crtc->cursor_gt = NULL;
 			goto out;
@@ -554,33 +555,6 @@ void psb_intel_crtc_init(struct drm_device *dev, int pipe,
 	gma_crtc->active = true;
 }
 
-int psb_intel_get_pipe_from_crtc_id(struct drm_device *dev, void *data,
-				struct drm_file *file_priv)
-{
-	struct drm_psb_private *dev_priv = dev->dev_private;
-	struct drm_psb_get_pipe_from_crtc_id_arg *pipe_from_crtc_id = data;
-	struct drm_mode_object *drmmode_obj;
-	struct gma_crtc *crtc;
-
-	if (!dev_priv) {
-		dev_err(dev->dev, "called with no initialization\n");
-		return -EINVAL;
-	}
-
-	drmmode_obj = drm_mode_object_find(dev, pipe_from_crtc_id->crtc_id,
-			DRM_MODE_OBJECT_CRTC);
-
-	if (!drmmode_obj) {
-		dev_err(dev->dev, "no such CRTC id\n");
-		return -ENOENT;
-	}
-
-	crtc = to_gma_crtc(obj_to_crtc(drmmode_obj));
-	pipe_from_crtc_id->pipe = crtc->pipe;
-
-	return 0;
-}
-
 struct drm_crtc *psb_intel_get_crtc_from_pipe(struct drm_device *dev, int pipe)
 {
 	struct drm_crtc *crtc = NULL;
diff --git a/drivers/gpu/drm/gma500/psb_intel_drv.h b/drivers/gpu/drm/gma500/psb_intel_drv.h
index bde27fdb41b..336bd3aa1a0 100644
--- a/drivers/gpu/drm/gma500/psb_intel_drv.h
+++ b/drivers/gpu/drm/gma500/psb_intel_drv.h
@@ -238,8 +238,6 @@ static inline struct gma_encoder *gma_attached_encoder(
 
 extern struct drm_display_mode *psb_intel_crtc_mode_get(struct drm_device *dev,
 						    struct drm_crtc *crtc);
-extern int psb_intel_get_pipe_from_crtc_id(struct drm_device *dev, void *data,
-				struct drm_file *file_priv);
 extern struct drm_crtc *psb_intel_get_crtc_from_pipe(struct drm_device *dev,
 						 int pipe);
 extern struct drm_connector *psb_intel_sdvo_find(struct drm_device *dev,
@@ -250,11 +248,6 @@ extern void psb_intel_sdvo_set_hotplug(struct drm_connector *connector,
 extern int intelfb_probe(struct drm_device *dev);
 extern int intelfb_remove(struct drm_device *dev,
 			  struct drm_framebuffer *fb);
-extern struct drm_framebuffer *psb_intel_framebuffer_create(struct drm_device
-							*dev, struct
-							drm_mode_fb_cmd
-							*mode_cmd,
-							void *mm_private);
 extern bool psb_intel_lvds_mode_fixup(struct drm_encoder *encoder,
 				      const struct drm_display_mode *mode,
 				      struct drm_display_mode *adjusted_mode);
diff --git a/drivers/gpu/drm/gma500/psb_intel_lvds.c b/drivers/gpu/drm/gma500/psb_intel_lvds.c
index 32342f6990d..d7778d0472c 100644
--- a/drivers/gpu/drm/gma500/psb_intel_lvds.c
+++ b/drivers/gpu/drm/gma500/psb_intel_lvds.c
@@ -614,7 +614,7 @@ int psb_intel_lvds_set_property(struct drm_connector *connector,
 						      &crtc->saved_mode,
 						      encoder->crtc->x,
 						      encoder->crtc->y,
-						      encoder->crtc->fb))
+						      encoder->crtc->primary->fb))
 				goto set_prop_error;
 		}
 	} else if (!strcmp(property->name, "backlight")) {
@@ -777,6 +777,7 @@ void psb_intel_lvds_init(struct drm_device *dev,
 	 * Attempt to get the fixed panel mode from DDC.  Assume that the
 	 * preferred mode is the right one.
 	 */
+	mutex_lock(&dev->mode_config.mutex);
 	psb_intel_ddc_get_modes(connector, &lvds_priv->ddc_bus->adapter);
 	list_for_each_entry(scan, &connector->probed_modes, head) {
 		if (scan->type & DRM_MODE_TYPE_PREFERRED) {
@@ -827,10 +828,12 @@ void psb_intel_lvds_init(struct drm_device *dev,
 	 * actually having one.
 	 */
 out:
+	mutex_unlock(&dev->mode_config.mutex);
 	drm_sysfs_connector_add(connector);
 	return;
 
 failed_find:
+	mutex_unlock(&dev->mode_config.mutex);
 	if (lvds_priv->ddc_bus)
 		psb_intel_i2c_destroy(lvds_priv->ddc_bus);
 failed_ddc:
diff --git a/drivers/gpu/drm/gma500/psb_intel_sdvo.c b/drivers/gpu/drm/gma500/psb_intel_sdvo.c
index 07d3a9e6d79..deeb0829b12 100644
--- a/drivers/gpu/drm/gma500/psb_intel_sdvo.c
+++ b/drivers/gpu/drm/gma500/psb_intel_sdvo.c
@@ -406,18 +406,18 @@ static void psb_intel_sdvo_debug_write(struct psb_intel_sdvo *psb_intel_sdvo, u8
 	DRM_DEBUG_KMS("%s: W: %02X ",
 				SDVO_NAME(psb_intel_sdvo), cmd);
 	for (i = 0; i < args_len; i++)
-		DRM_LOG_KMS("%02X ", ((u8 *)args)[i]);
+		DRM_DEBUG_KMS("%02X ", ((u8 *)args)[i]);
 	for (; i < 8; i++)
-		DRM_LOG_KMS("   ");
+		DRM_DEBUG_KMS("   ");
 	for (i = 0; i < ARRAY_SIZE(sdvo_cmd_names); i++) {
 		if (cmd == sdvo_cmd_names[i].cmd) {
-			DRM_LOG_KMS("(%s)", sdvo_cmd_names[i].name);
+			DRM_DEBUG_KMS("(%s)", sdvo_cmd_names[i].name);
 			break;
 		}
 	}
 	if (i == ARRAY_SIZE(sdvo_cmd_names))
-		DRM_LOG_KMS("(%02X)", cmd);
-	DRM_LOG_KMS("\n");
+		DRM_DEBUG_KMS("(%02X)", cmd);
+	DRM_DEBUG_KMS("\n");
 }
 
 static const char *cmd_status_names[] = {
@@ -512,9 +512,9 @@ static bool psb_intel_sdvo_read_response(struct psb_intel_sdvo *psb_intel_sdvo,
 	}
 
 	if (status <= SDVO_CMD_STATUS_SCALING_NOT_SUPP)
-		DRM_LOG_KMS("(%s)", cmd_status_names[status]);
+		DRM_DEBUG_KMS("(%s)", cmd_status_names[status]);
 	else
-		DRM_LOG_KMS("(??? %d)", status);
+		DRM_DEBUG_KMS("(??? %d)", status);
 
 	if (status != SDVO_CMD_STATUS_SUCCESS)
 		goto log_fail;
@@ -525,13 +525,13 @@ static bool psb_intel_sdvo_read_response(struct psb_intel_sdvo *psb_intel_sdvo,
 					  SDVO_I2C_RETURN_0 + i,
 					  &((u8 *)response)[i]))
 			goto log_fail;
-		DRM_LOG_KMS(" %02X", ((u8 *)response)[i]);
+		DRM_DEBUG_KMS(" %02X", ((u8 *)response)[i]);
 	}
-	DRM_LOG_KMS("\n");
+	DRM_DEBUG_KMS("\n");
 	return true;
 
 log_fail:
-	DRM_LOG_KMS("... failed\n");
+	DRM_DEBUG_KMS("... failed\n");
 	return false;
 }
 
@@ -1844,7 +1844,7 @@ done:
 	if (psb_intel_sdvo->base.base.crtc) {
 		struct drm_crtc *crtc = psb_intel_sdvo->base.base.crtc;
 		drm_crtc_helper_set_mode(crtc, &crtc->mode, crtc->x,
-					 crtc->y, crtc->fb);
+					 crtc->y, crtc->primary->fb);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/gma500/psb_irq.c b/drivers/gpu/drm/gma500/psb_irq.c
index ba4830342d3..624eb36511c 100644
--- a/drivers/gpu/drm/gma500/psb_irq.c
+++ b/drivers/gpu/drm/gma500/psb_irq.c
@@ -200,11 +200,64 @@ static void psb_vdc_interrupt(struct drm_device *dev, uint32_t vdc_stat)
 		mid_pipe_event_handler(dev, 1);
 }
 
-irqreturn_t psb_irq_handler(DRM_IRQ_ARGS)
+/*
+ * SGX interrupt handler
+ */
+static void psb_sgx_interrupt(struct drm_device *dev, u32 stat_1, u32 stat_2)
+{
+	struct drm_psb_private *dev_priv = dev->dev_private;
+	u32 val, addr;
+	int error = false;
+
+	if (stat_1 & _PSB_CE_TWOD_COMPLETE)
+		val = PSB_RSGX32(PSB_CR_2D_BLIT_STATUS);
+
+	if (stat_2 & _PSB_CE2_BIF_REQUESTER_FAULT) {
+		val = PSB_RSGX32(PSB_CR_BIF_INT_STAT);
+		addr = PSB_RSGX32(PSB_CR_BIF_FAULT);
+		if (val) {
+			if (val & _PSB_CBI_STAT_PF_N_RW)
+				DRM_ERROR("SGX MMU page fault:");
+			else
+				DRM_ERROR("SGX MMU read / write protection fault:");
+
+			if (val & _PSB_CBI_STAT_FAULT_CACHE)
+				DRM_ERROR("\tCache requestor");
+			if (val & _PSB_CBI_STAT_FAULT_TA)
+				DRM_ERROR("\tTA requestor");
+			if (val & _PSB_CBI_STAT_FAULT_VDM)
+				DRM_ERROR("\tVDM requestor");
+			if (val & _PSB_CBI_STAT_FAULT_2D)
+				DRM_ERROR("\t2D requestor");
+			if (val & _PSB_CBI_STAT_FAULT_PBE)
+				DRM_ERROR("\tPBE requestor");
+			if (val & _PSB_CBI_STAT_FAULT_TSP)
+				DRM_ERROR("\tTSP requestor");
+			if (val & _PSB_CBI_STAT_FAULT_ISP)
+				DRM_ERROR("\tISP requestor");
+			if (val & _PSB_CBI_STAT_FAULT_USSEPDS)
+				DRM_ERROR("\tUSSEPDS requestor");
+			if (val & _PSB_CBI_STAT_FAULT_HOST)
+				DRM_ERROR("\tHost requestor");
+
+			DRM_ERROR("\tMMU failing address is 0x%08x.\n",
+				  (unsigned int)addr);
+			error = true;
+		}
+	}
+
+	/* Clear bits */
+	PSB_WSGX32(stat_1, PSB_CR_EVENT_HOST_CLEAR);
+	PSB_WSGX32(stat_2, PSB_CR_EVENT_HOST_CLEAR2);
+	PSB_RSGX32(PSB_CR_EVENT_HOST_CLEAR2);
+}
+
+irqreturn_t psb_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = arg;
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	uint32_t vdc_stat, dsp_int = 0, sgx_int = 0, hotplug_int = 0;
+	u32 sgx_stat_1, sgx_stat_2;
 	int handled = 0;
 
 	spin_lock(&dev_priv->irqmask_lock);
@@ -233,14 +286,9 @@ irqreturn_t psb_irq_handler(DRM_IRQ_ARGS)
 	}
 
 	if (sgx_int) {
-		/* Not expected - we have it masked, shut it up */
-		u32 s, s2;
-		s = PSB_RSGX32(PSB_CR_EVENT_STATUS);
-		s2 = PSB_RSGX32(PSB_CR_EVENT_STATUS2);
-		PSB_WSGX32(s, PSB_CR_EVENT_HOST_CLEAR);
-		PSB_WSGX32(s2, PSB_CR_EVENT_HOST_CLEAR2);
-		/* if s & _PSB_CE_TWOD_COMPLETE we have 2D done but
-		   we may as well poll even if we add that ! */
+		sgx_stat_1 = PSB_RSGX32(PSB_CR_EVENT_STATUS);
+		sgx_stat_2 = PSB_RSGX32(PSB_CR_EVENT_STATUS2);
+		psb_sgx_interrupt(dev, sgx_stat_1, sgx_stat_2);
 		handled = 1;
 	}
 
@@ -253,7 +301,7 @@ irqreturn_t psb_irq_handler(DRM_IRQ_ARGS)
 
 	PSB_WVDC32(vdc_stat, PSB_INT_IDENTITY_R);
 	(void) PSB_RVDC32(PSB_INT_IDENTITY_R);
-	DRM_READMEMORYBARRIER();
+	rmb();
 
 	if (!handled)
 		return IRQ_NONE;
@@ -269,8 +317,13 @@ void psb_irq_preinstall(struct drm_device *dev)
 
 	spin_lock_irqsave(&dev_priv->irqmask_lock, irqflags);
 
-	if (gma_power_is_on(dev))
+	if (gma_power_is_on(dev)) {
 		PSB_WVDC32(0xFFFFFFFF, PSB_HWSTAM);
+		PSB_WVDC32(0x00000000, PSB_INT_MASK_R);
+		PSB_WVDC32(0x00000000, PSB_INT_ENABLE_R);
+		PSB_WSGX32(0x00000000, PSB_CR_EVENT_HOST_ENABLE);
+		PSB_RSGX32(PSB_CR_EVENT_HOST_ENABLE);
+	}
 	if (dev->vblank[0].enabled)
 		dev_priv->vdc_irq_mask |= _PSB_VSYNC_PIPEA_FLAG;
 	if (dev->vblank[1].enabled)
@@ -286,7 +339,7 @@ void psb_irq_preinstall(struct drm_device *dev)
 	/* Revisit this area - want per device masks ? */
 	if (dev_priv->ops->hotplug)
 		dev_priv->vdc_irq_mask |= _PSB_IRQ_DISP_HOTSYNC;
-	dev_priv->vdc_irq_mask |= _PSB_IRQ_ASLE;
+	dev_priv->vdc_irq_mask |= _PSB_IRQ_ASLE | _PSB_IRQ_SGX_FLAG;
 
 	/* This register is safe even if display island is off */
 	PSB_WVDC32(~dev_priv->vdc_irq_mask, PSB_INT_MASK_R);
@@ -295,12 +348,16 @@ void psb_irq_preinstall(struct drm_device *dev)
 
 int psb_irq_postinstall(struct drm_device *dev)
 {
-	struct drm_psb_private *dev_priv =
-	    (struct drm_psb_private *) dev->dev_private;
+	struct drm_psb_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
 
 	spin_lock_irqsave(&dev_priv->irqmask_lock, irqflags);
 
+	/* Enable 2D and MMU fault interrupts */
+	PSB_WSGX32(_PSB_CE2_BIF_REQUESTER_FAULT, PSB_CR_EVENT_HOST_ENABLE2);
+	PSB_WSGX32(_PSB_CE_TWOD_COMPLETE, PSB_CR_EVENT_HOST_ENABLE);
+	PSB_RSGX32(PSB_CR_EVENT_HOST_ENABLE); /* Post */
+
 	/* This register is safe even if display island is off */
 	PSB_WVDC32(dev_priv->vdc_irq_mask, PSB_INT_ENABLE_R);
 	PSB_WVDC32(0xFFFFFFFF, PSB_HWSTAM);
@@ -450,21 +507,6 @@ int psb_irq_disable_dpst(struct drm_device *dev)
 	return 0;
 }
 
-#ifdef PSB_FIXME
-static int psb_vblank_do_wait(struct drm_device *dev,
-			      unsigned int *sequence, atomic_t *counter)
-{
-	unsigned int cur_vblank;
-	int ret = 0;
-	DRM_WAIT_ON(ret, dev->vblank.queue, 3 * DRM_HZ,
-		    (((cur_vblank = atomic_read(counter))
-		      - *sequence) <= (1 << 23)));
-	*sequence = cur_vblank;
-
-	return ret;
-}
-#endif
-
 /*
  * It is used to enable VBLANK interrupt
  */
diff --git a/drivers/gpu/drm/gma500/psb_irq.h b/drivers/gpu/drm/gma500/psb_irq.h
index debb7f190c0..d0b45ffa112 100644
--- a/drivers/gpu/drm/gma500/psb_irq.h
+++ b/drivers/gpu/drm/gma500/psb_irq.h
@@ -32,7 +32,7 @@ void sysirq_uninit(struct drm_device *dev);
 void psb_irq_preinstall(struct drm_device *dev);
 int  psb_irq_postinstall(struct drm_device *dev);
 void psb_irq_uninstall(struct drm_device *dev);
-irqreturn_t psb_irq_handler(DRM_IRQ_ARGS);
+irqreturn_t psb_irq_handler(int irq, void *arg);
 
 int psb_irq_enable_dpst(struct drm_device *dev);
 int psb_irq_disable_dpst(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index 400b0c4a10f..ac357b02bd3 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -19,6 +19,8 @@
 
 #include <linux/hdmi.h>
 #include <linux/module.h>
+#include <linux/irq.h>
+#include <sound/asoundef.h>
 
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
@@ -30,6 +32,7 @@
 
 struct tda998x_priv {
 	struct i2c_client *cec;
+	struct i2c_client *hdmi;
 	uint16_t rev;
 	uint8_t current_page;
 	int dpms;
@@ -38,6 +41,10 @@ struct tda998x_priv {
 	u8 vip_cntrl_1;
 	u8 vip_cntrl_2;
 	struct tda998x_encoder_params params;
+
+	wait_queue_head_t wq_edid;
+	volatile int wq_edid_wait;
+	struct drm_encoder *encoder;
 };
 
 #define to_tda998x_priv(x)  ((struct tda998x_priv *)to_encoder_slave(x)->slave_priv)
@@ -120,6 +127,8 @@ struct tda998x_priv {
 # define VIP_CNTRL_5_CKCASE       (1 << 0)
 # define VIP_CNTRL_5_SP_CNT(x)    (((x) & 3) << 1)
 #define REG_MUX_AP                REG(0x00, 0x26)     /* read/write */
+# define MUX_AP_SELECT_I2S	  0x64
+# define MUX_AP_SELECT_SPDIF	  0x40
 #define REG_MUX_VP_VIP_OUT        REG(0x00, 0x27)     /* read/write */
 #define REG_MAT_CONTRL            REG(0x00, 0x80)     /* write */
 # define MAT_CONTRL_MAT_SC(x)     (((x) & 3) << 0)
@@ -197,10 +206,11 @@ struct tda998x_priv {
 #define REG_I2S_FORMAT            REG(0x00, 0xfc)     /* read/write */
 # define I2S_FORMAT(x)            (((x) & 3) << 0)
 #define REG_AIP_CLKSEL            REG(0x00, 0xfd)     /* write */
-# define AIP_CLKSEL_FS(x)         (((x) & 3) << 0)
-# define AIP_CLKSEL_CLK_POL(x)    (((x) & 1) << 2)
-# define AIP_CLKSEL_AIP(x)        (((x) & 7) << 3)
-
+# define AIP_CLKSEL_AIP_SPDIF	  (0 << 3)
+# define AIP_CLKSEL_AIP_I2S	  (1 << 3)
+# define AIP_CLKSEL_FS_ACLK	  (0 << 0)
+# define AIP_CLKSEL_FS_MCLK	  (1 << 0)
+# define AIP_CLKSEL_FS_FS64SPDIF  (2 << 0)
 
 /* Page 02h: PLL settings */
 #define REG_PLL_SERIAL_1          REG(0x02, 0x00)     /* read/write */
@@ -208,7 +218,7 @@ struct tda998x_priv {
 # define PLL_SERIAL_1_SRL_IZ(x)   (((x) & 3) << 1)
 # define PLL_SERIAL_1_SRL_MAN_IZ  (1 << 6)
 #define REG_PLL_SERIAL_2          REG(0x02, 0x01)     /* read/write */
-# define PLL_SERIAL_2_SRL_NOSC(x) (((x) & 3) << 0)
+# define PLL_SERIAL_2_SRL_NOSC(x) ((x) << 0)
 # define PLL_SERIAL_2_SRL_PR(x)   (((x) & 0xf) << 4)
 #define REG_PLL_SERIAL_3          REG(0x02, 0x02)     /* read/write */
 # define PLL_SERIAL_3_SRL_CCIR    (1 << 0)
@@ -304,11 +314,16 @@ struct tda998x_priv {
 
 /* CEC registers: (not paged)
  */
+#define REG_CEC_INTSTATUS	  0xee		      /* read */
+# define CEC_INTSTATUS_CEC	  (1 << 0)
+# define CEC_INTSTATUS_HDMI	  (1 << 1)
 #define REG_CEC_FRO_IM_CLK_CTRL   0xfb                /* read/write */
 # define CEC_FRO_IM_CLK_CTRL_GHOST_DIS (1 << 7)
 # define CEC_FRO_IM_CLK_CTRL_ENA_OTP   (1 << 6)
 # define CEC_FRO_IM_CLK_CTRL_IMCLK_SEL (1 << 1)
 # define CEC_FRO_IM_CLK_CTRL_FRO_DIV   (1 << 0)
+#define REG_CEC_RXSHPDINTENA	  0xfc		      /* read/write */
+#define REG_CEC_RXSHPDINT	  0xfd		      /* read */
 #define REG_CEC_RXSHPDLEV         0xfe                /* read */
 # define CEC_RXSHPDLEV_RXSENS     (1 << 0)
 # define CEC_RXSHPDLEV_HPD        (1 << 1)
@@ -328,21 +343,21 @@ struct tda998x_priv {
 #define TDA19988                  0x0301
 
 static void
-cec_write(struct drm_encoder *encoder, uint16_t addr, uint8_t val)
+cec_write(struct tda998x_priv *priv, uint16_t addr, uint8_t val)
 {
-	struct i2c_client *client = to_tda998x_priv(encoder)->cec;
+	struct i2c_client *client = priv->cec;
 	uint8_t buf[] = {addr, val};
 	int ret;
 
-	ret = i2c_master_send(client, buf, ARRAY_SIZE(buf));
+	ret = i2c_master_send(client, buf, sizeof(buf));
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to cec:0x%x\n", ret, addr);
 }
 
 static uint8_t
-cec_read(struct drm_encoder *encoder, uint8_t addr)
+cec_read(struct tda998x_priv *priv, uint8_t addr)
 {
-	struct i2c_client *client = to_tda998x_priv(encoder)->cec;
+	struct i2c_client *client = priv->cec;
 	uint8_t val;
 	int ret;
 
@@ -361,32 +376,36 @@ fail:
 	return 0;
 }
 
-static void
-set_page(struct drm_encoder *encoder, uint16_t reg)
+static int
+set_page(struct tda998x_priv *priv, uint16_t reg)
 {
-	struct tda998x_priv *priv = to_tda998x_priv(encoder);
-
 	if (REG2PAGE(reg) != priv->current_page) {
-		struct i2c_client *client = drm_i2c_encoder_get_client(encoder);
+		struct i2c_client *client = priv->hdmi;
 		uint8_t buf[] = {
 				REG_CURPAGE, REG2PAGE(reg)
 		};
 		int ret = i2c_master_send(client, buf, sizeof(buf));
-		if (ret < 0)
-			dev_err(&client->dev, "Error %d writing to REG_CURPAGE\n", ret);
+		if (ret < 0) {
+			dev_err(&client->dev, "setpage %04x err %d\n",
+					reg, ret);
+			return ret;
+		}
 
 		priv->current_page = REG2PAGE(reg);
 	}
+	return 0;
 }
 
 static int
-reg_read_range(struct drm_encoder *encoder, uint16_t reg, char *buf, int cnt)
+reg_read_range(struct tda998x_priv *priv, uint16_t reg, char *buf, int cnt)
 {
-	struct i2c_client *client = drm_i2c_encoder_get_client(encoder);
+	struct i2c_client *client = priv->hdmi;
 	uint8_t addr = REG2ADDR(reg);
 	int ret;
 
-	set_page(encoder, reg);
+	ret = set_page(priv, reg);
+	if (ret < 0)
+		return ret;
 
 	ret = i2c_master_send(client, &addr, sizeof(addr));
 	if (ret < 0)
@@ -404,200 +423,244 @@ fail:
 }
 
 static void
-reg_write_range(struct drm_encoder *encoder, uint16_t reg, uint8_t *p, int cnt)
+reg_write_range(struct tda998x_priv *priv, uint16_t reg, uint8_t *p, int cnt)
 {
-	struct i2c_client *client = drm_i2c_encoder_get_client(encoder);
+	struct i2c_client *client = priv->hdmi;
 	uint8_t buf[cnt+1];
 	int ret;
 
 	buf[0] = REG2ADDR(reg);
 	memcpy(&buf[1], p, cnt);
 
-	set_page(encoder, reg);
+	ret = set_page(priv, reg);
+	if (ret < 0)
+		return;
 
 	ret = i2c_master_send(client, buf, cnt + 1);
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to 0x%x\n", ret, reg);
 }
 
-static uint8_t
-reg_read(struct drm_encoder *encoder, uint16_t reg)
+static int
+reg_read(struct tda998x_priv *priv, uint16_t reg)
 {
 	uint8_t val = 0;
-	reg_read_range(encoder, reg, &val, sizeof(val));
+	int ret;
+
+	ret = reg_read_range(priv, reg, &val, sizeof(val));
+	if (ret < 0)
+		return ret;
 	return val;
 }
 
 static void
-reg_write(struct drm_encoder *encoder, uint16_t reg, uint8_t val)
+reg_write(struct tda998x_priv *priv, uint16_t reg, uint8_t val)
 {
-	struct i2c_client *client = drm_i2c_encoder_get_client(encoder);
+	struct i2c_client *client = priv->hdmi;
 	uint8_t buf[] = {REG2ADDR(reg), val};
 	int ret;
 
-	set_page(encoder, reg);
+	ret = set_page(priv, reg);
+	if (ret < 0)
+		return;
 
-	ret = i2c_master_send(client, buf, ARRAY_SIZE(buf));
+	ret = i2c_master_send(client, buf, sizeof(buf));
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to 0x%x\n", ret, reg);
 }
 
 static void
-reg_write16(struct drm_encoder *encoder, uint16_t reg, uint16_t val)
+reg_write16(struct tda998x_priv *priv, uint16_t reg, uint16_t val)
 {
-	struct i2c_client *client = drm_i2c_encoder_get_client(encoder);
+	struct i2c_client *client = priv->hdmi;
 	uint8_t buf[] = {REG2ADDR(reg), val >> 8, val};
 	int ret;
 
-	set_page(encoder, reg);
+	ret = set_page(priv, reg);
+	if (ret < 0)
+		return;
 
-	ret = i2c_master_send(client, buf, ARRAY_SIZE(buf));
+	ret = i2c_master_send(client, buf, sizeof(buf));
 	if (ret < 0)
 		dev_err(&client->dev, "Error %d writing to 0x%x\n", ret, reg);
 }
 
 static void
-reg_set(struct drm_encoder *encoder, uint16_t reg, uint8_t val)
+reg_set(struct tda998x_priv *priv, uint16_t reg, uint8_t val)
 {
-	reg_write(encoder, reg, reg_read(encoder, reg) | val);
+	int old_val;
+
+	old_val = reg_read(priv, reg);
+	if (old_val >= 0)
+		reg_write(priv, reg, old_val | val);
 }
 
 static void
-reg_clear(struct drm_encoder *encoder, uint16_t reg, uint8_t val)
+reg_clear(struct tda998x_priv *priv, uint16_t reg, uint8_t val)
 {
-	reg_write(encoder, reg, reg_read(encoder, reg) & ~val);
+	int old_val;
+
+	old_val = reg_read(priv, reg);
+	if (old_val >= 0)
+		reg_write(priv, reg, old_val & ~val);
 }
 
 static void
-tda998x_reset(struct drm_encoder *encoder)
+tda998x_reset(struct tda998x_priv *priv)
 {
 	/* reset audio and i2c master: */
-	reg_set(encoder, REG_SOFTRESET, SOFTRESET_AUDIO | SOFTRESET_I2C_MASTER);
+	reg_write(priv, REG_SOFTRESET, SOFTRESET_AUDIO | SOFTRESET_I2C_MASTER);
 	msleep(50);
-	reg_clear(encoder, REG_SOFTRESET, SOFTRESET_AUDIO | SOFTRESET_I2C_MASTER);
+	reg_write(priv, REG_SOFTRESET, 0);
 	msleep(50);
 
 	/* reset transmitter: */
-	reg_set(encoder, REG_MAIN_CNTRL0, MAIN_CNTRL0_SR);
-	reg_clear(encoder, REG_MAIN_CNTRL0, MAIN_CNTRL0_SR);
+	reg_set(priv, REG_MAIN_CNTRL0, MAIN_CNTRL0_SR);
+	reg_clear(priv, REG_MAIN_CNTRL0, MAIN_CNTRL0_SR);
 
 	/* PLL registers common configuration */
-	reg_write(encoder, REG_PLL_SERIAL_1, 0x00);
-	reg_write(encoder, REG_PLL_SERIAL_2, PLL_SERIAL_2_SRL_NOSC(1));
-	reg_write(encoder, REG_PLL_SERIAL_3, 0x00);
-	reg_write(encoder, REG_SERIALIZER,   0x00);
-	reg_write(encoder, REG_BUFFER_OUT,   0x00);
-	reg_write(encoder, REG_PLL_SCG1,     0x00);
-	reg_write(encoder, REG_AUDIO_DIV,    AUDIO_DIV_SERCLK_8);
-	reg_write(encoder, REG_SEL_CLK,      SEL_CLK_SEL_CLK1 | SEL_CLK_ENA_SC_CLK);
-	reg_write(encoder, REG_PLL_SCGN1,    0xfa);
-	reg_write(encoder, REG_PLL_SCGN2,    0x00);
-	reg_write(encoder, REG_PLL_SCGR1,    0x5b);
-	reg_write(encoder, REG_PLL_SCGR2,    0x00);
-	reg_write(encoder, REG_PLL_SCG2,     0x10);
+	reg_write(priv, REG_PLL_SERIAL_1, 0x00);
+	reg_write(priv, REG_PLL_SERIAL_2, PLL_SERIAL_2_SRL_NOSC(1));
+	reg_write(priv, REG_PLL_SERIAL_3, 0x00);
+	reg_write(priv, REG_SERIALIZER,   0x00);
+	reg_write(priv, REG_BUFFER_OUT,   0x00);
+	reg_write(priv, REG_PLL_SCG1,     0x00);
+	reg_write(priv, REG_AUDIO_DIV,    AUDIO_DIV_SERCLK_8);
+	reg_write(priv, REG_SEL_CLK,      SEL_CLK_SEL_CLK1 | SEL_CLK_ENA_SC_CLK);
+	reg_write(priv, REG_PLL_SCGN1,    0xfa);
+	reg_write(priv, REG_PLL_SCGN2,    0x00);
+	reg_write(priv, REG_PLL_SCGR1,    0x5b);
+	reg_write(priv, REG_PLL_SCGR2,    0x00);
+	reg_write(priv, REG_PLL_SCG2,     0x10);
 
 	/* Write the default value MUX register */
-	reg_write(encoder, REG_MUX_VP_VIP_OUT, 0x24);
+	reg_write(priv, REG_MUX_VP_VIP_OUT, 0x24);
+}
+
+/*
+ * only 2 interrupts may occur: screen plug/unplug and EDID read
+ */
+static irqreturn_t tda998x_irq_thread(int irq, void *data)
+{
+	struct tda998x_priv *priv = data;
+	u8 sta, cec, lvl, flag0, flag1, flag2;
+
+	if (!priv)
+		return IRQ_HANDLED;
+	sta = cec_read(priv, REG_CEC_INTSTATUS);
+	cec = cec_read(priv, REG_CEC_RXSHPDINT);
+	lvl = cec_read(priv, REG_CEC_RXSHPDLEV);
+	flag0 = reg_read(priv, REG_INT_FLAGS_0);
+	flag1 = reg_read(priv, REG_INT_FLAGS_1);
+	flag2 = reg_read(priv, REG_INT_FLAGS_2);
+	DRM_DEBUG_DRIVER(
+		"tda irq sta %02x cec %02x lvl %02x f0 %02x f1 %02x f2 %02x\n",
+		sta, cec, lvl, flag0, flag1, flag2);
+	if ((flag2 & INT_FLAGS_2_EDID_BLK_RD) && priv->wq_edid_wait) {
+		priv->wq_edid_wait = 0;
+		wake_up(&priv->wq_edid);
+	} else if (cec != 0) {			/* HPD change */
+		if (priv->encoder && priv->encoder->dev)
+			drm_helper_hpd_irq_event(priv->encoder->dev);
+	}
+	return IRQ_HANDLED;
 }
 
 static uint8_t tda998x_cksum(uint8_t *buf, size_t bytes)
 {
-	uint8_t sum = 0;
+	int sum = 0;
 
 	while (bytes--)
-		sum += *buf++;
-	return (255 - sum) + 1;
+		sum -= *buf++;
+	return sum;
 }
 
 #define HB(x) (x)
 #define PB(x) (HB(2) + 1 + (x))
 
 static void
-tda998x_write_if(struct drm_encoder *encoder, uint8_t bit, uint16_t addr,
+tda998x_write_if(struct tda998x_priv *priv, uint8_t bit, uint16_t addr,
 		 uint8_t *buf, size_t size)
 {
 	buf[PB(0)] = tda998x_cksum(buf, size);
 
-	reg_clear(encoder, REG_DIP_IF_FLAGS, bit);
-	reg_write_range(encoder, addr, buf, size);
-	reg_set(encoder, REG_DIP_IF_FLAGS, bit);
+	reg_clear(priv, REG_DIP_IF_FLAGS, bit);
+	reg_write_range(priv, addr, buf, size);
+	reg_set(priv, REG_DIP_IF_FLAGS, bit);
 }
 
 static void
-tda998x_write_aif(struct drm_encoder *encoder, struct tda998x_encoder_params *p)
+tda998x_write_aif(struct tda998x_priv *priv, struct tda998x_encoder_params *p)
 {
-	uint8_t buf[PB(5) + 1];
+	u8 buf[PB(HDMI_AUDIO_INFOFRAME_SIZE) + 1];
 
-	buf[HB(0)] = 0x84;
+	memset(buf, 0, sizeof(buf));
+	buf[HB(0)] = HDMI_INFOFRAME_TYPE_AUDIO;
 	buf[HB(1)] = 0x01;
-	buf[HB(2)] = 10;
-	buf[PB(0)] = 0;
+	buf[HB(2)] = HDMI_AUDIO_INFOFRAME_SIZE;
 	buf[PB(1)] = p->audio_frame[1] & 0x07; /* CC */
 	buf[PB(2)] = p->audio_frame[2] & 0x1c; /* SF */
 	buf[PB(4)] = p->audio_frame[4];
 	buf[PB(5)] = p->audio_frame[5] & 0xf8; /* DM_INH + LSV */
 
-	tda998x_write_if(encoder, DIP_IF_FLAGS_IF4, REG_IF4_HB0, buf,
+	tda998x_write_if(priv, DIP_IF_FLAGS_IF4, REG_IF4_HB0, buf,
 			 sizeof(buf));
 }
 
 static void
-tda998x_write_avi(struct drm_encoder *encoder, struct drm_display_mode *mode)
+tda998x_write_avi(struct tda998x_priv *priv, struct drm_display_mode *mode)
 {
-	uint8_t buf[PB(13) + 1];
+	u8 buf[PB(HDMI_AVI_INFOFRAME_SIZE) + 1];
 
 	memset(buf, 0, sizeof(buf));
-	buf[HB(0)] = 0x82;
+	buf[HB(0)] = HDMI_INFOFRAME_TYPE_AVI;
 	buf[HB(1)] = 0x02;
-	buf[HB(2)] = 13;
+	buf[HB(2)] = HDMI_AVI_INFOFRAME_SIZE;
 	buf[PB(1)] = HDMI_SCAN_MODE_UNDERSCAN;
+	buf[PB(2)] = HDMI_ACTIVE_ASPECT_PICTURE;
 	buf[PB(3)] = HDMI_QUANTIZATION_RANGE_FULL << 2;
 	buf[PB(4)] = drm_match_cea_mode(mode);
 
-	tda998x_write_if(encoder, DIP_IF_FLAGS_IF2, REG_IF2_HB0, buf,
+	tda998x_write_if(priv, DIP_IF_FLAGS_IF2, REG_IF2_HB0, buf,
 			 sizeof(buf));
 }
 
-static void tda998x_audio_mute(struct drm_encoder *encoder, bool on)
+static void tda998x_audio_mute(struct tda998x_priv *priv, bool on)
 {
 	if (on) {
-		reg_set(encoder, REG_SOFTRESET, SOFTRESET_AUDIO);
-		reg_clear(encoder, REG_SOFTRESET, SOFTRESET_AUDIO);
-		reg_set(encoder, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_FIFO);
+		reg_set(priv, REG_SOFTRESET, SOFTRESET_AUDIO);
+		reg_clear(priv, REG_SOFTRESET, SOFTRESET_AUDIO);
+		reg_set(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_FIFO);
 	} else {
-		reg_clear(encoder, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_FIFO);
+		reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_FIFO);
 	}
 }
 
 static void
-tda998x_configure_audio(struct drm_encoder *encoder,
+tda998x_configure_audio(struct tda998x_priv *priv,
 		struct drm_display_mode *mode, struct tda998x_encoder_params *p)
 {
-	uint8_t buf[6], clksel_aip, clksel_fs, ca_i2s, cts_n, adiv;
+	uint8_t buf[6], clksel_aip, clksel_fs, cts_n, adiv;
 	uint32_t n;
 
 	/* Enable audio ports */
-	reg_write(encoder, REG_ENA_AP, p->audio_cfg);
-	reg_write(encoder, REG_ENA_ACLK, p->audio_clk_cfg);
+	reg_write(priv, REG_ENA_AP, p->audio_cfg);
+	reg_write(priv, REG_ENA_ACLK, p->audio_clk_cfg);
 
 	/* Set audio input source */
 	switch (p->audio_format) {
 	case AFMT_SPDIF:
-		reg_write(encoder, REG_MUX_AP, 0x40);
-		clksel_aip = AIP_CLKSEL_AIP(0);
-		/* FS64SPDIF */
-		clksel_fs = AIP_CLKSEL_FS(2);
+		reg_write(priv, REG_MUX_AP, MUX_AP_SELECT_SPDIF);
+		clksel_aip = AIP_CLKSEL_AIP_SPDIF;
+		clksel_fs = AIP_CLKSEL_FS_FS64SPDIF;
 		cts_n = CTS_N_M(3) | CTS_N_K(3);
-		ca_i2s = 0;
 		break;
 
 	case AFMT_I2S:
-		reg_write(encoder, REG_MUX_AP, 0x64);
-		clksel_aip = AIP_CLKSEL_AIP(1);
-		/* ACLK */
-		clksel_fs = AIP_CLKSEL_FS(0);
+		reg_write(priv, REG_MUX_AP, MUX_AP_SELECT_I2S);
+		clksel_aip = AIP_CLKSEL_AIP_I2S;
+		clksel_fs = AIP_CLKSEL_FS_ACLK;
 		cts_n = CTS_N_M(3) | CTS_N_K(3);
-		ca_i2s = CA_I2S_CA_I2S(0);
 		break;
 
 	default:
@@ -605,12 +668,10 @@ tda998x_configure_audio(struct drm_encoder *encoder,
 		return;
 	}
 
-	reg_write(encoder, REG_AIP_CLKSEL, clksel_aip);
-	reg_clear(encoder, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT);
-
-	/* Enable automatic CTS generation */
-	reg_clear(encoder, REG_AIP_CNTRL_0, AIP_CNTRL_0_ACR_MAN);
-	reg_write(encoder, REG_CTS_N, cts_n);
+	reg_write(priv, REG_AIP_CLKSEL, clksel_aip);
+	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |
+					AIP_CNTRL_0_ACR_MAN);	/* auto CTS */
+	reg_write(priv, REG_CTS_N, cts_n);
 
 	/*
 	 * Audio input somehow depends on HDMI line rate which is
@@ -619,11 +680,15 @@ tda998x_configure_audio(struct drm_encoder *encoder,
 	 * There is no detailed info in the datasheet, so we just
 	 * assume 100MHz requires larger divider.
 	 */
+	adiv = AUDIO_DIV_SERCLK_8;
 	if (mode->clock > 100000)
-		adiv = AUDIO_DIV_SERCLK_16;
-	else
-		adiv = AUDIO_DIV_SERCLK_8;
-	reg_write(encoder, REG_AUDIO_DIV, adiv);
+		adiv++;			/* AUDIO_DIV_SERCLK_16 */
+
+	/* S/PDIF asks for a larger divider */
+	if (p->audio_format == AFMT_SPDIF)
+		adiv++;			/* AUDIO_DIV_SERCLK_16 or _32 */
+
+	reg_write(priv, REG_AUDIO_DIV, adiv);
 
 	/*
 	 * This is the approximate value of N, which happens to be
@@ -638,28 +703,29 @@ tda998x_configure_audio(struct drm_encoder *encoder,
 	buf[3] = n;
 	buf[4] = n >> 8;
 	buf[5] = n >> 16;
-	reg_write_range(encoder, REG_ACR_CTS_0, buf, 6);
+	reg_write_range(priv, REG_ACR_CTS_0, buf, 6);
 
 	/* Set CTS clock reference */
-	reg_write(encoder, REG_AIP_CLKSEL, clksel_aip | clksel_fs);
+	reg_write(priv, REG_AIP_CLKSEL, clksel_aip | clksel_fs);
 
 	/* Reset CTS generator */
-	reg_set(encoder, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_CTS);
-	reg_clear(encoder, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_CTS);
+	reg_set(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_CTS);
+	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_CTS);
 
 	/* Write the channel status */
-	buf[0] = 0x04;
+	buf[0] = IEC958_AES0_CON_NOT_COPYRIGHT;
 	buf[1] = 0x00;
-	buf[2] = 0x00;
-	buf[3] = 0xf1;
-	reg_write_range(encoder, REG_CH_STAT_B(0), buf, 4);
+	buf[2] = IEC958_AES3_CON_FS_NOTID;
+	buf[3] = IEC958_AES4_CON_ORIGFS_NOTID |
+			IEC958_AES4_CON_MAX_WORDLEN_24;
+	reg_write_range(priv, REG_CH_STAT_B(0), buf, 4);
 
-	tda998x_audio_mute(encoder, true);
-	mdelay(20);
-	tda998x_audio_mute(encoder, false);
+	tda998x_audio_mute(priv, true);
+	msleep(20);
+	tda998x_audio_mute(priv, false);
 
 	/* Write the audio information packet */
-	tda998x_write_aif(encoder, p);
+	tda998x_write_aif(priv, p);
 }
 
 /* DRM encoder functions */
@@ -701,19 +767,19 @@ tda998x_encoder_dpms(struct drm_encoder *encoder, int mode)
 	switch (mode) {
 	case DRM_MODE_DPMS_ON:
 		/* enable video ports, audio will be enabled later */
-		reg_write(encoder, REG_ENA_VP_0, 0xff);
-		reg_write(encoder, REG_ENA_VP_1, 0xff);
-		reg_write(encoder, REG_ENA_VP_2, 0xff);
+		reg_write(priv, REG_ENA_VP_0, 0xff);
+		reg_write(priv, REG_ENA_VP_1, 0xff);
+		reg_write(priv, REG_ENA_VP_2, 0xff);
 		/* set muxing after enabling ports: */
-		reg_write(encoder, REG_VIP_CNTRL_0, priv->vip_cntrl_0);
-		reg_write(encoder, REG_VIP_CNTRL_1, priv->vip_cntrl_1);
-		reg_write(encoder, REG_VIP_CNTRL_2, priv->vip_cntrl_2);
+		reg_write(priv, REG_VIP_CNTRL_0, priv->vip_cntrl_0);
+		reg_write(priv, REG_VIP_CNTRL_1, priv->vip_cntrl_1);
+		reg_write(priv, REG_VIP_CNTRL_2, priv->vip_cntrl_2);
 		break;
 	case DRM_MODE_DPMS_OFF:
 		/* disable video ports */
-		reg_write(encoder, REG_ENA_VP_0, 0x00);
-		reg_write(encoder, REG_ENA_VP_1, 0x00);
-		reg_write(encoder, REG_ENA_VP_2, 0x00);
+		reg_write(priv, REG_ENA_VP_0, 0x00);
+		reg_write(priv, REG_ENA_VP_1, 0x00);
+		reg_write(priv, REG_ENA_VP_2, 0x00);
 		break;
 	}
 
@@ -744,6 +810,12 @@ static int
 tda998x_encoder_mode_valid(struct drm_encoder *encoder,
 			  struct drm_display_mode *mode)
 {
+	if (mode->clock > 150000)
+		return MODE_CLOCK_HIGH;
+	if (mode->htotal >= BIT(13))
+		return MODE_BAD_HVALUE;
+	if (mode->vtotal >= BIT(11))
+		return MODE_BAD_VVALUE;
 	return MODE_OK;
 }
 
@@ -824,112 +896,117 @@ tda998x_encoder_mode_set(struct drm_encoder *encoder,
 	}
 
 	div = 148500 / mode->clock;
+	if (div != 0) {
+		div--;
+		if (div > 3)
+			div = 3;
+	}
 
 	/* mute the audio FIFO: */
-	reg_set(encoder, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_FIFO);
+	reg_set(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_FIFO);
 
 	/* set HDMI HDCP mode off: */
-	reg_set(encoder, REG_TBG_CNTRL_1, TBG_CNTRL_1_DWIN_DIS);
-	reg_clear(encoder, REG_TX33, TX33_HDMI);
+	reg_write(priv, REG_TBG_CNTRL_1, TBG_CNTRL_1_DWIN_DIS);
+	reg_clear(priv, REG_TX33, TX33_HDMI);
+	reg_write(priv, REG_ENC_CNTRL, ENC_CNTRL_CTL_CODE(0));
 
-	reg_write(encoder, REG_ENC_CNTRL, ENC_CNTRL_CTL_CODE(0));
 	/* no pre-filter or interpolator: */
-	reg_write(encoder, REG_HVF_CNTRL_0, HVF_CNTRL_0_PREFIL(0) |
+	reg_write(priv, REG_HVF_CNTRL_0, HVF_CNTRL_0_PREFIL(0) |
 			HVF_CNTRL_0_INTPOL(0));
-	reg_write(encoder, REG_VIP_CNTRL_5, VIP_CNTRL_5_SP_CNT(0));
-	reg_write(encoder, REG_VIP_CNTRL_4, VIP_CNTRL_4_BLANKIT(0) |
+	reg_write(priv, REG_VIP_CNTRL_5, VIP_CNTRL_5_SP_CNT(0));
+	reg_write(priv, REG_VIP_CNTRL_4, VIP_CNTRL_4_BLANKIT(0) |
 			VIP_CNTRL_4_BLC(0));
-	reg_clear(encoder, REG_PLL_SERIAL_3, PLL_SERIAL_3_SRL_CCIR);
 
-	reg_clear(encoder, REG_PLL_SERIAL_1, PLL_SERIAL_1_SRL_MAN_IZ);
-	reg_clear(encoder, REG_PLL_SERIAL_3, PLL_SERIAL_3_SRL_DE);
-	reg_write(encoder, REG_SERIALIZER, 0);
-	reg_write(encoder, REG_HVF_CNTRL_1, HVF_CNTRL_1_VQR(0));
+	reg_clear(priv, REG_PLL_SERIAL_1, PLL_SERIAL_1_SRL_MAN_IZ);
+	reg_clear(priv, REG_PLL_SERIAL_3, PLL_SERIAL_3_SRL_CCIR |
+					  PLL_SERIAL_3_SRL_DE);
+	reg_write(priv, REG_SERIALIZER, 0);
+	reg_write(priv, REG_HVF_CNTRL_1, HVF_CNTRL_1_VQR(0));
 
 	/* TODO enable pixel repeat for pixel rates less than 25Msamp/s */
 	rep = 0;
-	reg_write(encoder, REG_RPT_CNTRL, 0);
-	reg_write(encoder, REG_SEL_CLK, SEL_CLK_SEL_VRF_CLK(0) |
+	reg_write(priv, REG_RPT_CNTRL, 0);
+	reg_write(priv, REG_SEL_CLK, SEL_CLK_SEL_VRF_CLK(0) |
 			SEL_CLK_SEL_CLK1 | SEL_CLK_ENA_SC_CLK);
 
-	reg_write(encoder, REG_PLL_SERIAL_2, PLL_SERIAL_2_SRL_NOSC(div) |
+	reg_write(priv, REG_PLL_SERIAL_2, PLL_SERIAL_2_SRL_NOSC(div) |
 			PLL_SERIAL_2_SRL_PR(rep));
 
 	/* set color matrix bypass flag: */
-	reg_set(encoder, REG_MAT_CONTRL, MAT_CONTRL_MAT_BP);
+	reg_write(priv, REG_MAT_CONTRL, MAT_CONTRL_MAT_BP |
+				MAT_CONTRL_MAT_SC(1));
 
 	/* set BIAS tmds value: */
-	reg_write(encoder, REG_ANA_GENERAL, 0x09);
-
-	reg_clear(encoder, REG_TBG_CNTRL_0, TBG_CNTRL_0_SYNC_MTHD);
+	reg_write(priv, REG_ANA_GENERAL, 0x09);
 
 	/*
 	 * Sync on rising HSYNC/VSYNC
 	 */
-	reg_write(encoder, REG_VIP_CNTRL_3, 0);
-	reg_set(encoder, REG_VIP_CNTRL_3, VIP_CNTRL_3_SYNC_HS);
+	reg = VIP_CNTRL_3_SYNC_HS;
 
 	/*
 	 * TDA19988 requires high-active sync at input stage,
 	 * so invert low-active sync provided by master encoder here
 	 */
 	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-		reg_set(encoder, REG_VIP_CNTRL_3, VIP_CNTRL_3_H_TGL);
+		reg |= VIP_CNTRL_3_H_TGL;
 	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-		reg_set(encoder, REG_VIP_CNTRL_3, VIP_CNTRL_3_V_TGL);
+		reg |= VIP_CNTRL_3_V_TGL;
+	reg_write(priv, REG_VIP_CNTRL_3, reg);
+
+	reg_write(priv, REG_VIDFORMAT, 0x00);
+	reg_write16(priv, REG_REFPIX_MSB, ref_pix);
+	reg_write16(priv, REG_REFLINE_MSB, ref_line);
+	reg_write16(priv, REG_NPIX_MSB, n_pix);
+	reg_write16(priv, REG_NLINE_MSB, n_line);
+	reg_write16(priv, REG_VS_LINE_STRT_1_MSB, vs1_line_s);
+	reg_write16(priv, REG_VS_PIX_STRT_1_MSB, vs1_pix_s);
+	reg_write16(priv, REG_VS_LINE_END_1_MSB, vs1_line_e);
+	reg_write16(priv, REG_VS_PIX_END_1_MSB, vs1_pix_e);
+	reg_write16(priv, REG_VS_LINE_STRT_2_MSB, vs2_line_s);
+	reg_write16(priv, REG_VS_PIX_STRT_2_MSB, vs2_pix_s);
+	reg_write16(priv, REG_VS_LINE_END_2_MSB, vs2_line_e);
+	reg_write16(priv, REG_VS_PIX_END_2_MSB, vs2_pix_e);
+	reg_write16(priv, REG_HS_PIX_START_MSB, hs_pix_s);
+	reg_write16(priv, REG_HS_PIX_STOP_MSB, hs_pix_e);
+	reg_write16(priv, REG_VWIN_START_1_MSB, vwin1_line_s);
+	reg_write16(priv, REG_VWIN_END_1_MSB, vwin1_line_e);
+	reg_write16(priv, REG_VWIN_START_2_MSB, vwin2_line_s);
+	reg_write16(priv, REG_VWIN_END_2_MSB, vwin2_line_e);
+	reg_write16(priv, REG_DE_START_MSB, de_pix_s);
+	reg_write16(priv, REG_DE_STOP_MSB, de_pix_e);
+
+	if (priv->rev == TDA19988) {
+		/* let incoming pixels fill the active space (if any) */
+		reg_write(priv, REG_ENABLE_SPACE, 0x00);
+	}
 
 	/*
 	 * Always generate sync polarity relative to input sync and
 	 * revert input stage toggled sync at output stage
 	 */
-	reg = TBG_CNTRL_1_TGL_EN;
+	reg = TBG_CNTRL_1_DWIN_DIS | TBG_CNTRL_1_TGL_EN;
 	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
 		reg |= TBG_CNTRL_1_H_TGL;
 	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
 		reg |= TBG_CNTRL_1_V_TGL;
-	reg_write(encoder, REG_TBG_CNTRL_1, reg);
-
-	reg_write(encoder, REG_VIDFORMAT, 0x00);
-	reg_write16(encoder, REG_REFPIX_MSB, ref_pix);
-	reg_write16(encoder, REG_REFLINE_MSB, ref_line);
-	reg_write16(encoder, REG_NPIX_MSB, n_pix);
-	reg_write16(encoder, REG_NLINE_MSB, n_line);
-	reg_write16(encoder, REG_VS_LINE_STRT_1_MSB, vs1_line_s);
-	reg_write16(encoder, REG_VS_PIX_STRT_1_MSB, vs1_pix_s);
-	reg_write16(encoder, REG_VS_LINE_END_1_MSB, vs1_line_e);
-	reg_write16(encoder, REG_VS_PIX_END_1_MSB, vs1_pix_e);
-	reg_write16(encoder, REG_VS_LINE_STRT_2_MSB, vs2_line_s);
-	reg_write16(encoder, REG_VS_PIX_STRT_2_MSB, vs2_pix_s);
-	reg_write16(encoder, REG_VS_LINE_END_2_MSB, vs2_line_e);
-	reg_write16(encoder, REG_VS_PIX_END_2_MSB, vs2_pix_e);
-	reg_write16(encoder, REG_HS_PIX_START_MSB, hs_pix_s);
-	reg_write16(encoder, REG_HS_PIX_STOP_MSB, hs_pix_e);
-	reg_write16(encoder, REG_VWIN_START_1_MSB, vwin1_line_s);
-	reg_write16(encoder, REG_VWIN_END_1_MSB, vwin1_line_e);
-	reg_write16(encoder, REG_VWIN_START_2_MSB, vwin2_line_s);
-	reg_write16(encoder, REG_VWIN_END_2_MSB, vwin2_line_e);
-	reg_write16(encoder, REG_DE_START_MSB, de_pix_s);
-	reg_write16(encoder, REG_DE_STOP_MSB, de_pix_e);
-
-	if (priv->rev == TDA19988) {
-		/* let incoming pixels fill the active space (if any) */
-		reg_write(encoder, REG_ENABLE_SPACE, 0x01);
-	}
+	reg_write(priv, REG_TBG_CNTRL_1, reg);
 
 	/* must be last register set: */
-	reg_clear(encoder, REG_TBG_CNTRL_0, TBG_CNTRL_0_SYNC_ONCE);
+	reg_write(priv, REG_TBG_CNTRL_0, 0);
 
 	/* Only setup the info frames if the sink is HDMI */
 	if (priv->is_hdmi_sink) {
 		/* We need to turn HDMI HDCP stuff on to get audio through */
-		reg_clear(encoder, REG_TBG_CNTRL_1, TBG_CNTRL_1_DWIN_DIS);
-		reg_write(encoder, REG_ENC_CNTRL, ENC_CNTRL_CTL_CODE(1));
-		reg_set(encoder, REG_TX33, TX33_HDMI);
+		reg &= ~TBG_CNTRL_1_DWIN_DIS;
+		reg_write(priv, REG_TBG_CNTRL_1, reg);
+		reg_write(priv, REG_ENC_CNTRL, ENC_CNTRL_CTL_CODE(1));
+		reg_set(priv, REG_TX33, TX33_HDMI);
 
-		tda998x_write_avi(encoder, adjusted_mode);
+		tda998x_write_avi(priv, adjusted_mode);
 
 		if (priv->params.audio_cfg)
-			tda998x_configure_audio(encoder, adjusted_mode,
+			tda998x_configure_audio(priv, adjusted_mode,
 						&priv->params);
 	}
 }
@@ -938,7 +1015,9 @@ static enum drm_connector_status
 tda998x_encoder_detect(struct drm_encoder *encoder,
 		      struct drm_connector *connector)
 {
-	uint8_t val = cec_read(encoder, REG_CEC_RXSHPDLEV);
+	struct tda998x_priv *priv = to_tda998x_priv(encoder);
+	uint8_t val = cec_read(priv, REG_CEC_RXSHPDLEV);
+
 	return (val & CEC_RXSHPDLEV_HPD) ? connector_status_connected :
 			connector_status_disconnected;
 }
@@ -946,46 +1025,57 @@ tda998x_encoder_detect(struct drm_encoder *encoder,
 static int
 read_edid_block(struct drm_encoder *encoder, uint8_t *buf, int blk)
 {
+	struct tda998x_priv *priv = to_tda998x_priv(encoder);
 	uint8_t offset, segptr;
 	int ret, i;
 
-	/* enable EDID read irq: */
-	reg_set(encoder, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD);
-
 	offset = (blk & 1) ? 128 : 0;
 	segptr = blk / 2;
 
-	reg_write(encoder, REG_DDC_ADDR, 0xa0);
-	reg_write(encoder, REG_DDC_OFFS, offset);
-	reg_write(encoder, REG_DDC_SEGM_ADDR, 0x60);
-	reg_write(encoder, REG_DDC_SEGM, segptr);
+	reg_write(priv, REG_DDC_ADDR, 0xa0);
+	reg_write(priv, REG_DDC_OFFS, offset);
+	reg_write(priv, REG_DDC_SEGM_ADDR, 0x60);
+	reg_write(priv, REG_DDC_SEGM, segptr);
 
 	/* enable reading EDID: */
-	reg_write(encoder, REG_EDID_CTRL, 0x1);
+	priv->wq_edid_wait = 1;
+	reg_write(priv, REG_EDID_CTRL, 0x1);
 
 	/* flag must be cleared by sw: */
-	reg_write(encoder, REG_EDID_CTRL, 0x0);
+	reg_write(priv, REG_EDID_CTRL, 0x0);
 
 	/* wait for block read to complete: */
-	for (i = 100; i > 0; i--) {
-		uint8_t val = reg_read(encoder, REG_INT_FLAGS_2);
-		if (val & INT_FLAGS_2_EDID_BLK_RD)
-			break;
-		msleep(1);
+	if (priv->hdmi->irq) {
+		i = wait_event_timeout(priv->wq_edid,
+					!priv->wq_edid_wait,
+					msecs_to_jiffies(100));
+		if (i < 0) {
+			dev_err(&priv->hdmi->dev, "read edid wait err %d\n", i);
+			return i;
+		}
+	} else {
+		for (i = 100; i > 0; i--) {
+			msleep(1);
+			ret = reg_read(priv, REG_INT_FLAGS_2);
+			if (ret < 0)
+				return ret;
+			if (ret & INT_FLAGS_2_EDID_BLK_RD)
+				break;
+		}
 	}
 
-	if (i == 0)
+	if (i == 0) {
+		dev_err(&priv->hdmi->dev, "read edid timeout\n");
 		return -ETIMEDOUT;
+	}
 
-	ret = reg_read_range(encoder, REG_EDID_DATA_0, buf, EDID_LENGTH);
+	ret = reg_read_range(priv, REG_EDID_DATA_0, buf, EDID_LENGTH);
 	if (ret != EDID_LENGTH) {
-		dev_err(encoder->dev->dev, "failed to read edid block %d: %d",
-				blk, ret);
+		dev_err(&priv->hdmi->dev, "failed to read edid block %d: %d\n",
+			blk, ret);
 		return ret;
 	}
 
-	reg_clear(encoder, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD);
-
 	return 0;
 }
 
@@ -993,7 +1083,7 @@ static uint8_t *
 do_get_edid(struct drm_encoder *encoder)
 {
 	struct tda998x_priv *priv = to_tda998x_priv(encoder);
-	int j = 0, valid_extensions = 0;
+	int j, valid_extensions = 0;
 	uint8_t *block, *new;
 	bool print_bad_edid = drm_debug & DRM_UT_KMS;
 
@@ -1001,7 +1091,7 @@ do_get_edid(struct drm_encoder *encoder)
 		return NULL;
 
 	if (priv->rev == TDA19988)
-		reg_clear(encoder, REG_TX4, TX4_PD_RAM);
+		reg_clear(priv, REG_TX4, TX4_PD_RAM);
 
 	/* base block fetch */
 	if (read_edid_block(encoder, block, 0))
@@ -1041,14 +1131,14 @@ do_get_edid(struct drm_encoder *encoder)
 
 done:
 	if (priv->rev == TDA19988)
-		reg_set(encoder, REG_TX4, TX4_PD_RAM);
+		reg_set(priv, REG_TX4, TX4_PD_RAM);
 
 	return block;
 
 fail:
 	if (priv->rev == TDA19988)
-		reg_set(encoder, REG_TX4, TX4_PD_RAM);
-	dev_warn(encoder->dev->dev, "failed to read EDID\n");
+		reg_set(priv, REG_TX4, TX4_PD_RAM);
+	dev_warn(&priv->hdmi->dev, "failed to read EDID\n");
 	kfree(block);
 	return NULL;
 }
@@ -1075,7 +1165,13 @@ static int
 tda998x_encoder_create_resources(struct drm_encoder *encoder,
 				struct drm_connector *connector)
 {
-	DBG("");
+	struct tda998x_priv *priv = to_tda998x_priv(encoder);
+
+	if (priv->hdmi->irq)
+		connector->polled = DRM_CONNECTOR_POLL_HPD;
+	else
+		connector->polled = DRM_CONNECTOR_POLL_CONNECT |
+			DRM_CONNECTOR_POLL_DISCONNECT;
 	return 0;
 }
 
@@ -1093,6 +1189,15 @@ static void
 tda998x_encoder_destroy(struct drm_encoder *encoder)
 {
 	struct tda998x_priv *priv = to_tda998x_priv(encoder);
+
+	/* disable all IRQs and free the IRQ handler */
+	cec_write(priv, REG_CEC_RXSHPDINTENA, 0);
+	reg_clear(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD);
+	if (priv->hdmi->irq)
+		free_irq(priv->hdmi->irq, priv);
+
+	if (priv->cec)
+		i2c_unregister_device(priv->cec);
 	drm_i2c_encoder_destroy(encoder);
 	kfree(priv);
 }
@@ -1131,8 +1236,10 @@ tda998x_encoder_init(struct i2c_client *client,
 		    struct drm_device *dev,
 		    struct drm_encoder_slave *encoder_slave)
 {
-	struct drm_encoder *encoder = &encoder_slave->base;
 	struct tda998x_priv *priv;
+	struct device_node *np = client->dev.of_node;
+	u32 video;
+	int rev_lo, rev_hi, ret;
 
 	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
 	if (!priv)
@@ -1142,49 +1249,114 @@ tda998x_encoder_init(struct i2c_client *client,
 	priv->vip_cntrl_1 = VIP_CNTRL_1_SWAP_C(0) | VIP_CNTRL_1_SWAP_D(1);
 	priv->vip_cntrl_2 = VIP_CNTRL_2_SWAP_E(4) | VIP_CNTRL_2_SWAP_F(5);
 
-	priv->current_page = 0;
+	priv->current_page = 0xff;
+	priv->hdmi = client;
 	priv->cec = i2c_new_dummy(client->adapter, 0x34);
+	if (!priv->cec) {
+		kfree(priv);
+		return -ENODEV;
+	}
+
+	priv->encoder = &encoder_slave->base;
 	priv->dpms = DRM_MODE_DPMS_OFF;
 
 	encoder_slave->slave_priv = priv;
 	encoder_slave->slave_funcs = &tda998x_encoder_funcs;
 
 	/* wake up the device: */
-	cec_write(encoder, REG_CEC_ENAMODS,
+	cec_write(priv, REG_CEC_ENAMODS,
 			CEC_ENAMODS_EN_RXSENS | CEC_ENAMODS_EN_HDMI);
 
-	tda998x_reset(encoder);
+	tda998x_reset(priv);
 
 	/* read version: */
-	priv->rev = reg_read(encoder, REG_VERSION_LSB) |
-			reg_read(encoder, REG_VERSION_MSB) << 8;
+	rev_lo = reg_read(priv, REG_VERSION_LSB);
+	rev_hi = reg_read(priv, REG_VERSION_MSB);
+	if (rev_lo < 0 || rev_hi < 0) {
+		ret = rev_lo < 0 ? rev_lo : rev_hi;
+		goto fail;
+	}
+
+	priv->rev = rev_lo | rev_hi << 8;
 
 	/* mask off feature bits: */
 	priv->rev &= ~0x30; /* not-hdcp and not-scalar bit */
 
 	switch (priv->rev) {
-	case TDA9989N2:  dev_info(dev->dev, "found TDA9989 n2");  break;
-	case TDA19989:   dev_info(dev->dev, "found TDA19989");    break;
-	case TDA19989N2: dev_info(dev->dev, "found TDA19989 n2"); break;
-	case TDA19988:   dev_info(dev->dev, "found TDA19988");    break;
+	case TDA9989N2:
+		dev_info(&client->dev, "found TDA9989 n2");
+		break;
+	case TDA19989:
+		dev_info(&client->dev, "found TDA19989");
+		break;
+	case TDA19989N2:
+		dev_info(&client->dev, "found TDA19989 n2");
+		break;
+	case TDA19988:
+		dev_info(&client->dev, "found TDA19988");
+		break;
 	default:
-		DBG("found unsupported device: %04x", priv->rev);
+		dev_err(&client->dev, "found unsupported device: %04x\n",
+			priv->rev);
 		goto fail;
 	}
 
 	/* after reset, enable DDC: */
-	reg_write(encoder, REG_DDC_DISABLE, 0x00);
+	reg_write(priv, REG_DDC_DISABLE, 0x00);
 
 	/* set clock on DDC channel: */
-	reg_write(encoder, REG_TX3, 39);
+	reg_write(priv, REG_TX3, 39);
 
 	/* if necessary, disable multi-master: */
 	if (priv->rev == TDA19989)
-		reg_set(encoder, REG_I2C_MASTER, I2C_MASTER_DIS_MM);
+		reg_set(priv, REG_I2C_MASTER, I2C_MASTER_DIS_MM);
 
-	cec_write(encoder, REG_CEC_FRO_IM_CLK_CTRL,
+	cec_write(priv, REG_CEC_FRO_IM_CLK_CTRL,
 			CEC_FRO_IM_CLK_CTRL_GHOST_DIS | CEC_FRO_IM_CLK_CTRL_IMCLK_SEL);
 
+	/* initialize the optional IRQ */
+	if (client->irq) {
+		int irqf_trigger;
+
+		/* init read EDID waitqueue */
+		init_waitqueue_head(&priv->wq_edid);
+
+		/* clear pending interrupts */
+		reg_read(priv, REG_INT_FLAGS_0);
+		reg_read(priv, REG_INT_FLAGS_1);
+		reg_read(priv, REG_INT_FLAGS_2);
+
+		irqf_trigger =
+			irqd_get_trigger_type(irq_get_irq_data(client->irq));
+		ret = request_threaded_irq(client->irq, NULL,
+					   tda998x_irq_thread,
+					   irqf_trigger | IRQF_ONESHOT,
+					   "tda998x", priv);
+		if (ret) {
+			dev_err(&client->dev,
+				"failed to request IRQ#%u: %d\n",
+				client->irq, ret);
+			goto fail;
+		}
+
+		/* enable HPD irq */
+		cec_write(priv, REG_CEC_RXSHPDINTENA, CEC_RXSHPDLEV_HPD);
+	}
+
+	/* enable EDID read irq: */
+	reg_set(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD);
+
+	if (!np)
+		return 0;		/* non-DT */
+
+	/* get the optional video properties */
+	ret = of_property_read_u32(np, "video-ports", &video);
+	if (ret == 0) {
+		priv->vip_cntrl_0 = video >> 16;
+		priv->vip_cntrl_1 = video >> 8;
+		priv->vip_cntrl_2 = video;
+	}
+
 	return 0;
 
 fail:
@@ -1199,6 +1371,14 @@ fail:
 	return -ENXIO;
 }
 
+#ifdef CONFIG_OF
+static const struct of_device_id tda998x_dt_ids[] = {
+	{ .compatible = "nxp,tda998x", },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, tda998x_dt_ids);
+#endif
+
 static struct i2c_device_id tda998x_ids[] = {
 	{ "tda998x", 0 },
 	{ }
@@ -1211,6 +1391,7 @@ static struct drm_i2c_encoder_driver tda998x_driver = {
 		.remove = tda998x_remove,
 		.driver = {
 			.name = "tda998x",
+			.of_match_table = of_match_ptr(tda998x_dt_ids),
 		},
 		.id_table = tda998x_ids,
 	},
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index 249fdff305c..e88bac1d781 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -1193,6 +1193,10 @@ static int i810_flip_bufs(struct drm_device *dev, void *data,
 
 int i810_driver_load(struct drm_device *dev, unsigned long flags)
 {
+	/* Our userspace depends upon the agp mapping support. */
+	if (!dev->agp)
+		return -EINVAL;
+
 	pci_set_master(dev->pdev);
 
 	return 0;
@@ -1247,7 +1251,7 @@ const struct drm_ioctl_desc i810_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(I810_FLIP, i810_flip_bufs, DRM_AUTH|DRM_UNLOCKED),
 };
 
-int i810_max_ioctl = DRM_ARRAY_SIZE(i810_ioctls);
+int i810_max_ioctl = ARRAY_SIZE(i810_ioctls);
 
 /**
  * Determine if the device really is AGP or not.
diff --git a/drivers/gpu/drm/i810/i810_drv.c b/drivers/gpu/drm/i810/i810_drv.c
index d8180d22ced..441ccf8f5bd 100644
--- a/drivers/gpu/drm/i810/i810_drv.c
+++ b/drivers/gpu/drm/i810/i810_drv.c
@@ -57,7 +57,7 @@ static const struct file_operations i810_driver_fops = {
 
 static struct drm_driver driver = {
 	.driver_features =
-	    DRIVER_USE_AGP | DRIVER_REQUIRE_AGP |
+	    DRIVER_USE_AGP |
 	    DRIVER_HAVE_DMA,
 	.dev_priv_size = sizeof(drm_i810_buf_priv_t),
 	.load = i810_driver_load,
diff --git a/drivers/gpu/drm/i915/Kconfig b/drivers/gpu/drm/i915/Kconfig
index 6199d0b5b95..437e1824d0b 100644
--- a/drivers/gpu/drm/i915/Kconfig
+++ b/drivers/gpu/drm/i915/Kconfig
@@ -1,8 +1,11 @@
 config DRM_I915
 	tristate "Intel 8xx/9xx/G3x/G4x/HD Graphics"
 	depends on DRM
-	depends on AGP
-	depends on AGP_INTEL
+	depends on X86 && PCI
+	depends on (AGP || AGP=n)
+	select INTEL_GTT
+	select AGP_INTEL if AGP
+	select INTERVAL_TREE
 	# we need shmfs for the swappable backing store, and in particular
 	# the shmem_readpage() which depends upon tmpfs
 	select SHMEM
@@ -12,7 +15,6 @@ config DRM_I915
 	# but for select to work, need to select ACPI_VIDEO's dependencies, ick
 	select BACKLIGHT_LCD_SUPPORT if ACPI
 	select BACKLIGHT_CLASS_DEVICE if ACPI
-	select VIDEO_OUTPUT_CONTROL if ACPI
 	select INPUT if ACPI
 	select ACPI_VIDEO if ACPI
 	select ACPI_BUTTON if ACPI
@@ -35,15 +37,14 @@ config DRM_I915
 config DRM_I915_KMS
 	bool "Enable modesetting on intel by default"
 	depends on DRM_I915
+	default y
 	help
-	  Choose this option if you want kernel modesetting enabled by default,
-	  and you have a new enough userspace to support this. Running old
-	  userspaces with this enabled will cause pain.  Note that this causes
-	  the driver to bind to PCI devices, which precludes loading things
-	  like intelfb.
+	  Choose this option if you want kernel modesetting enabled by default.
+
+	  If in doubt, say "Y".
 
 config DRM_I915_FBDEV
-	bool "Enable legacy fbdev support for the modesettting intel driver"
+	bool "Enable legacy fbdev support for the modesetting intel driver"
 	depends on DRM_I915
 	select DRM_KMS_FB_HELPER
 	select FB_CFB_FILLRECT
@@ -55,9 +56,12 @@ config DRM_I915_FBDEV
 	  support. Note that this support also provide the linux console
 	  support on top of the intel modesetting driver.
 
+	  If in doubt, say "Y".
+
 config DRM_I915_PRELIMINARY_HW_SUPPORT
 	bool "Enable preliminary support for prerelease Intel hardware by default"
 	depends on DRM_I915
+	default n
 	help
 	  Choose this option if you have prerelease Intel hardware and want the
 	  i915 driver to support it by default.  You can enable such support at
@@ -65,3 +69,15 @@ config DRM_I915_PRELIMINARY_HW_SUPPORT
 	  option changes the default for that module option.
 
 	  If in doubt, say "N".
+
+config DRM_I915_UMS
+	bool "Enable userspace modesetting on Intel hardware (DEPRECATED)"
+	depends on DRM_I915 && BROKEN
+	default n
+	help
+	  Choose this option if you still need userspace modesetting.
+
+	  Userspace modesetting is deprecated for quite some time now, so
+	  enable this only if you have ancient versions of the DDX drivers.
+
+	  If in doubt, say "N".
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 41838eaa799..cad1683d8bb 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -3,57 +3,77 @@
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
 ccflags-y := -Iinclude/drm
-i915-y := i915_drv.o i915_dma.o i915_irq.o \
-	  i915_debugfs.o \
-	  i915_gpu_error.o \
+
+# Please keep these build lists sorted!
+
+# core driver code
+i915-y := i915_drv.o \
+	  i915_params.o \
           i915_suspend.o \
-	  i915_gem.o \
+	  i915_sysfs.o \
+	  intel_pm.o
+i915-$(CONFIG_COMPAT)   += i915_ioc32.o
+i915-$(CONFIG_DEBUG_FS) += i915_debugfs.o
+
+# GEM code
+i915-y += i915_cmd_parser.o \
 	  i915_gem_context.o \
+	  i915_gem_render_state.o \
 	  i915_gem_debug.o \
+	  i915_gem_dmabuf.o \
 	  i915_gem_evict.o \
 	  i915_gem_execbuffer.o \
 	  i915_gem_gtt.o \
+	  i915_gem.o \
 	  i915_gem_stolen.o \
 	  i915_gem_tiling.o \
-	  i915_sysfs.o \
+	  i915_gem_userptr.o \
+	  i915_gpu_error.o \
+	  i915_irq.o \
 	  i915_trace_points.o \
-	  i915_ums.o \
+	  intel_ringbuffer.o \
+	  intel_uncore.o
+
+# autogenerated null render state
+i915-y += intel_renderstate_gen6.o \
+	  intel_renderstate_gen7.o \
+	  intel_renderstate_gen8.o
+
+# modesetting core code
+i915-y += intel_bios.o \
 	  intel_display.o \
-	  intel_crt.o \
-	  intel_lvds.o \
-	  intel_dsi.o \
-	  intel_dsi_cmd.o \
-	  intel_dsi_pll.o \
-	  intel_bios.o \
-	  intel_ddi.o \
-	  intel_dp.o \
-	  intel_hdmi.o \
-	  intel_sdvo.o \
 	  intel_modes.o \
-	  intel_panel.o \
-	  intel_pm.o \
-	  intel_i2c.o \
-	  intel_tv.o \
-	  intel_dvo.o \
-	  intel_ringbuffer.o \
 	  intel_overlay.o \
-	  intel_sprite.o \
-	  intel_opregion.o \
 	  intel_sideband.o \
-	  intel_uncore.o \
+	  intel_sprite.o
+i915-$(CONFIG_ACPI)		+= intel_acpi.o intel_opregion.o
+i915-$(CONFIG_DRM_I915_FBDEV)	+= intel_fbdev.o
+
+# modesetting output/encoder code
+i915-y += dvo_ch7017.o \
 	  dvo_ch7xxx.o \
-	  dvo_ch7017.o \
 	  dvo_ivch.o \
-	  dvo_tfp410.o \
-	  dvo_sil164.o \
 	  dvo_ns2501.o \
-	  i915_gem_dmabuf.o
-
-i915-$(CONFIG_COMPAT)   += i915_ioc32.o
-
-i915-$(CONFIG_ACPI)	+= intel_acpi.o
+	  dvo_sil164.o \
+	  dvo_tfp410.o \
+	  intel_crt.o \
+	  intel_ddi.o \
+	  intel_dp.o \
+	  intel_dsi_cmd.o \
+	  intel_dsi.o \
+	  intel_dsi_pll.o \
+	  intel_dsi_panel_vbt.o \
+	  intel_dvo.o \
+	  intel_hdmi.o \
+	  intel_i2c.o \
+	  intel_lvds.o \
+	  intel_panel.o \
+	  intel_sdvo.o \
+	  intel_tv.o
 
-i915-$(CONFIG_DRM_I915_FBDEV) += intel_fbdev.o
+# legacy horrors
+i915-y += i915_dma.o \
+	  i915_ums.o
 
 obj-$(CONFIG_DRM_I915)  += i915.o
 
diff --git a/drivers/gpu/drm/i915/dvo_ch7xxx.c b/drivers/gpu/drm/i915/dvo_ch7xxx.c
index af42e94f684..80449f47596 100644
--- a/drivers/gpu/drm/i915/dvo_ch7xxx.c
+++ b/drivers/gpu/drm/i915/dvo_ch7xxx.c
@@ -160,7 +160,7 @@ static bool ch7xxx_readb(struct intel_dvo_device *dvo, int addr, uint8_t *ch)
 	if (i2c_transfer(adapter, msgs, 2) == 2) {
 		*ch = in_buf[0];
 		return true;
-	};
+	}
 
 	if (!ch7xxx->quiet) {
 		DRM_DEBUG_KMS("Unable to read register 0x%02x from %s:%02x.\n",
@@ -340,9 +340,9 @@ static void ch7xxx_dump_regs(struct intel_dvo_device *dvo)
 	for (i = 0; i < CH7xxx_NUM_REGS; i++) {
 		uint8_t val;
 		if ((i % 8) == 0)
-			DRM_LOG_KMS("\n %02X: ", i);
+			DRM_DEBUG_KMS("\n %02X: ", i);
 		ch7xxx_readb(dvo, i, &val);
-		DRM_LOG_KMS("%02X ", val);
+		DRM_DEBUG_KMS("%02X ", val);
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/dvo_ivch.c b/drivers/gpu/drm/i915/dvo_ivch.c
index baaf65bf0bd..0f2587ff347 100644
--- a/drivers/gpu/drm/i915/dvo_ivch.c
+++ b/drivers/gpu/drm/i915/dvo_ivch.c
@@ -195,7 +195,7 @@ static bool ivch_read(struct intel_dvo_device *dvo, int addr, uint16_t *data)
 	if (i2c_transfer(adapter, msgs, 3) == 3) {
 		*data = (in_buf[1] << 8) | in_buf[0];
 		return true;
-	};
+	}
 
 	if (!priv->quiet) {
 		DRM_DEBUG_KMS("Unable to read register 0x%02x from "
@@ -377,41 +377,41 @@ static void ivch_dump_regs(struct intel_dvo_device *dvo)
 	uint16_t val;
 
 	ivch_read(dvo, VR00, &val);
-	DRM_LOG_KMS("VR00: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR00: 0x%04x\n", val);
 	ivch_read(dvo, VR01, &val);
-	DRM_LOG_KMS("VR01: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR01: 0x%04x\n", val);
 	ivch_read(dvo, VR30, &val);
-	DRM_LOG_KMS("VR30: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR30: 0x%04x\n", val);
 	ivch_read(dvo, VR40, &val);
-	DRM_LOG_KMS("VR40: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR40: 0x%04x\n", val);
 
 	/* GPIO registers */
 	ivch_read(dvo, VR80, &val);
-	DRM_LOG_KMS("VR80: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR80: 0x%04x\n", val);
 	ivch_read(dvo, VR81, &val);
-	DRM_LOG_KMS("VR81: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR81: 0x%04x\n", val);
 	ivch_read(dvo, VR82, &val);
-	DRM_LOG_KMS("VR82: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR82: 0x%04x\n", val);
 	ivch_read(dvo, VR83, &val);
-	DRM_LOG_KMS("VR83: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR83: 0x%04x\n", val);
 	ivch_read(dvo, VR84, &val);
-	DRM_LOG_KMS("VR84: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR84: 0x%04x\n", val);
 	ivch_read(dvo, VR85, &val);
-	DRM_LOG_KMS("VR85: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR85: 0x%04x\n", val);
 	ivch_read(dvo, VR86, &val);
-	DRM_LOG_KMS("VR86: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR86: 0x%04x\n", val);
 	ivch_read(dvo, VR87, &val);
-	DRM_LOG_KMS("VR87: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR87: 0x%04x\n", val);
 	ivch_read(dvo, VR88, &val);
-	DRM_LOG_KMS("VR88: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR88: 0x%04x\n", val);
 
 	/* Scratch register 0 - AIM Panel type */
 	ivch_read(dvo, VR8E, &val);
-	DRM_LOG_KMS("VR8E: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR8E: 0x%04x\n", val);
 
 	/* Scratch register 1 - Status register */
 	ivch_read(dvo, VR8F, &val);
-	DRM_LOG_KMS("VR8F: 0x%04x\n", val);
+	DRM_DEBUG_KMS("VR8F: 0x%04x\n", val);
 }
 
 static void ivch_destroy(struct intel_dvo_device *dvo)
diff --git a/drivers/gpu/drm/i915/dvo_ns2501.c b/drivers/gpu/drm/i915/dvo_ns2501.c
index c4a255be697..74f2af7c2d3 100644
--- a/drivers/gpu/drm/i915/dvo_ns2501.c
+++ b/drivers/gpu/drm/i915/dvo_ns2501.c
@@ -87,49 +87,6 @@ struct ns2501_priv {
  * when switching the resolution.
  */
 
-static void enable_dvo(struct intel_dvo_device *dvo)
-{
-	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
-	struct i2c_adapter *adapter = dvo->i2c_bus;
-	struct intel_gmbus *bus = container_of(adapter,
-					       struct intel_gmbus,
-					       adapter);
-	struct drm_i915_private *dev_priv = bus->dev_priv;
-
-	DRM_DEBUG_KMS("%s: Trying to re-enable the DVO\n", __FUNCTION__);
-
-	ns->dvoc = I915_READ(DVO_C);
-	ns->pll_a = I915_READ(_DPLL_A);
-	ns->srcdim = I915_READ(DVOC_SRCDIM);
-	ns->fw_blc = I915_READ(FW_BLC);
-
-	I915_WRITE(DVOC, 0x10004084);
-	I915_WRITE(_DPLL_A, 0xd0820000);
-	I915_WRITE(DVOC_SRCDIM, 0x400300);	// 1024x768
-	I915_WRITE(FW_BLC, 0x1080304);
-
-	I915_WRITE(DVOC, 0x90004084);
-}
-
-/*
- * Restore the I915 registers modified by the above
- * trigger function.
- */
-static void restore_dvo(struct intel_dvo_device *dvo)
-{
-	struct i2c_adapter *adapter = dvo->i2c_bus;
-	struct intel_gmbus *bus = container_of(adapter,
-					       struct intel_gmbus,
-					       adapter);
-	struct drm_i915_private *dev_priv = bus->dev_priv;
-	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
-
-	I915_WRITE(DVOC, ns->dvoc);
-	I915_WRITE(_DPLL_A, ns->pll_a);
-	I915_WRITE(DVOC_SRCDIM, ns->srcdim);
-	I915_WRITE(FW_BLC, ns->fw_blc);
-}
-
 /*
 ** Read a register from the ns2501.
 ** Returns true if successful, false otherwise.
@@ -164,7 +121,7 @@ static bool ns2501_readb(struct intel_dvo_device *dvo, int addr, uint8_t * ch)
 	if (i2c_transfer(adapter, msgs, 2) == 2) {
 		*ch = in_buf[0];
 		return true;
-	};
+	}
 
 	if (!ns->quiet) {
 		DRM_DEBUG_KMS
@@ -276,9 +233,8 @@ static enum drm_mode_status ns2501_mode_valid(struct intel_dvo_device *dvo,
 					      struct drm_display_mode *mode)
 {
 	DRM_DEBUG_KMS
-	    ("%s: is mode valid (hdisplay=%d,htotal=%d,vdisplay=%d,vtotal=%d)\n",
-	     __FUNCTION__, mode->hdisplay, mode->htotal, mode->vdisplay,
-	     mode->vtotal);
+	    ("is mode valid (hdisplay=%d,htotal=%d,vdisplay=%d,vtotal=%d)\n",
+	     mode->hdisplay, mode->htotal, mode->vdisplay, mode->vtotal);
 
 	/*
 	 * Currently, these are all the modes I have data from.
@@ -300,13 +256,12 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 			    struct drm_display_mode *adjusted_mode)
 {
 	bool ok;
-	bool restore = false;
+	int retries = 10;
 	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
 
 	DRM_DEBUG_KMS
-	    ("%s: set mode (hdisplay=%d,htotal=%d,vdisplay=%d,vtotal=%d).\n",
-	     __FUNCTION__, mode->hdisplay, mode->htotal, mode->vdisplay,
-	     mode->vtotal);
+	    ("set mode (hdisplay=%d,htotal=%d,vdisplay=%d,vtotal=%d).\n",
+	     mode->hdisplay, mode->htotal, mode->vdisplay, mode->vtotal);
 
 	/*
 	 * Where do I find the native resolution for which scaling is not required???
@@ -320,8 +275,7 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 		if (mode->hdisplay == 800 && mode->vdisplay == 600) {
 			/* mode 277 */
 			ns->reg_8_shadow &= ~NS2501_8_BPAS;
-			DRM_DEBUG_KMS("%s: switching to 800x600\n",
-				      __FUNCTION__);
+			DRM_DEBUG_KMS("switching to 800x600\n");
 
 			/*
 			 * No, I do not know where this data comes from.
@@ -384,8 +338,7 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 
 		} else if (mode->hdisplay == 640 && mode->vdisplay == 480) {
 			/* mode 274 */
-			DRM_DEBUG_KMS("%s: switching to 640x480\n",
-				      __FUNCTION__);
+			DRM_DEBUG_KMS("switching to 640x480\n");
 			/*
 			 * No, I do not know where this data comes from.
 			 * It is just what the video bios left in the DVO, so
@@ -449,8 +402,7 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 
 		} else if (mode->hdisplay == 1024 && mode->vdisplay == 768) {
 			/* mode 280 */
-			DRM_DEBUG_KMS("%s: switching to 1024x768\n",
-				      __FUNCTION__);
+			DRM_DEBUG_KMS("switching to 1024x768\n");
 			/*
 			 * This might or might not work, actually. I'm silently
 			 * assuming here that the native panel resolution is
@@ -476,20 +428,7 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 			ns->reg_8_shadow |= NS2501_8_BPAS;
 		}
 		ok &= ns2501_writeb(dvo, NS2501_REG8, ns->reg_8_shadow);
-
-		if (!ok) {
-			if (restore)
-				restore_dvo(dvo);
-			enable_dvo(dvo);
-			restore = true;
-		}
-	} while (!ok);
-	/*
-	 * Restore the old i915 registers before
-	 * forcing the ns2501 on.
-	 */
-	if (restore)
-		restore_dvo(dvo);
+	} while (!ok && retries--);
 }
 
 /* set the NS2501 power state */
@@ -510,12 +449,11 @@ static bool ns2501_get_hw_state(struct intel_dvo_device *dvo)
 static void ns2501_dpms(struct intel_dvo_device *dvo, bool enable)
 {
 	bool ok;
-	bool restore = false;
+	int retries = 10;
 	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
 	unsigned char ch;
 
-	DRM_DEBUG_KMS("%s: Trying set the dpms of the DVO to %i\n",
-		      __FUNCTION__, enable);
+	DRM_DEBUG_KMS("Trying set the dpms of the DVO to %i\n", enable);
 
 	ch = ns->reg_8_shadow;
 
@@ -537,16 +475,7 @@ static void ns2501_dpms(struct intel_dvo_device *dvo, bool enable)
 			ok &=
 			    ns2501_writeb(dvo, 0x35,
 					  enable ? 0xff : 0x00);
-			if (!ok) {
-				if (restore)
-					restore_dvo(dvo);
-				enable_dvo(dvo);
-				restore = true;
-			}
-		} while (!ok);
-
-		if (restore)
-			restore_dvo(dvo);
+		} while (!ok && retries--);
 	}
 }
 
@@ -555,15 +484,15 @@ static void ns2501_dump_regs(struct intel_dvo_device *dvo)
 	uint8_t val;
 
 	ns2501_readb(dvo, NS2501_FREQ_LO, &val);
-	DRM_LOG_KMS("NS2501_FREQ_LO: 0x%02x\n", val);
+	DRM_DEBUG_KMS("NS2501_FREQ_LO: 0x%02x\n", val);
 	ns2501_readb(dvo, NS2501_FREQ_HI, &val);
-	DRM_LOG_KMS("NS2501_FREQ_HI: 0x%02x\n", val);
+	DRM_DEBUG_KMS("NS2501_FREQ_HI: 0x%02x\n", val);
 	ns2501_readb(dvo, NS2501_REG8, &val);
-	DRM_LOG_KMS("NS2501_REG8: 0x%02x\n", val);
+	DRM_DEBUG_KMS("NS2501_REG8: 0x%02x\n", val);
 	ns2501_readb(dvo, NS2501_REG9, &val);
-	DRM_LOG_KMS("NS2501_REG9: 0x%02x\n", val);
+	DRM_DEBUG_KMS("NS2501_REG9: 0x%02x\n", val);
 	ns2501_readb(dvo, NS2501_REGC, &val);
-	DRM_LOG_KMS("NS2501_REGC: 0x%02x\n", val);
+	DRM_DEBUG_KMS("NS2501_REGC: 0x%02x\n", val);
 }
 
 static void ns2501_destroy(struct intel_dvo_device *dvo)
diff --git a/drivers/gpu/drm/i915/dvo_sil164.c b/drivers/gpu/drm/i915/dvo_sil164.c
index 4debd32e3e4..fa011496707 100644
--- a/drivers/gpu/drm/i915/dvo_sil164.c
+++ b/drivers/gpu/drm/i915/dvo_sil164.c
@@ -93,7 +93,7 @@ static bool sil164_readb(struct intel_dvo_device *dvo, int addr, uint8_t *ch)
 	if (i2c_transfer(adapter, msgs, 2) == 2) {
 		*ch = in_buf[0];
 		return true;
-	};
+	}
 
 	if (!sil->quiet) {
 		DRM_DEBUG_KMS("Unable to read register 0x%02x from %s:%02x.\n",
@@ -246,15 +246,15 @@ static void sil164_dump_regs(struct intel_dvo_device *dvo)
 	uint8_t val;
 
 	sil164_readb(dvo, SIL164_FREQ_LO, &val);
-	DRM_LOG_KMS("SIL164_FREQ_LO: 0x%02x\n", val);
+	DRM_DEBUG_KMS("SIL164_FREQ_LO: 0x%02x\n", val);
 	sil164_readb(dvo, SIL164_FREQ_HI, &val);
-	DRM_LOG_KMS("SIL164_FREQ_HI: 0x%02x\n", val);
+	DRM_DEBUG_KMS("SIL164_FREQ_HI: 0x%02x\n", val);
 	sil164_readb(dvo, SIL164_REG8, &val);
-	DRM_LOG_KMS("SIL164_REG8: 0x%02x\n", val);
+	DRM_DEBUG_KMS("SIL164_REG8: 0x%02x\n", val);
 	sil164_readb(dvo, SIL164_REG9, &val);
-	DRM_LOG_KMS("SIL164_REG9: 0x%02x\n", val);
+	DRM_DEBUG_KMS("SIL164_REG9: 0x%02x\n", val);
 	sil164_readb(dvo, SIL164_REGC, &val);
-	DRM_LOG_KMS("SIL164_REGC: 0x%02x\n", val);
+	DRM_DEBUG_KMS("SIL164_REGC: 0x%02x\n", val);
 }
 
 static void sil164_destroy(struct intel_dvo_device *dvo)
diff --git a/drivers/gpu/drm/i915/dvo_tfp410.c b/drivers/gpu/drm/i915/dvo_tfp410.c
index e17f1b07e91..7853719a0e8 100644
--- a/drivers/gpu/drm/i915/dvo_tfp410.c
+++ b/drivers/gpu/drm/i915/dvo_tfp410.c
@@ -118,7 +118,7 @@ static bool tfp410_readb(struct intel_dvo_device *dvo, int addr, uint8_t *ch)
 	if (i2c_transfer(adapter, msgs, 2) == 2) {
 		*ch = in_buf[0];
 		return true;
-	};
+	}
 
 	if (!tfp->quiet) {
 		DRM_DEBUG_KMS("Unable to read register 0x%02x from %s:%02x.\n",
@@ -267,33 +267,33 @@ static void tfp410_dump_regs(struct intel_dvo_device *dvo)
 	uint8_t val, val2;
 
 	tfp410_readb(dvo, TFP410_REV, &val);
-	DRM_LOG_KMS("TFP410_REV: 0x%02X\n", val);
+	DRM_DEBUG_KMS("TFP410_REV: 0x%02X\n", val);
 	tfp410_readb(dvo, TFP410_CTL_1, &val);
-	DRM_LOG_KMS("TFP410_CTL1: 0x%02X\n", val);
+	DRM_DEBUG_KMS("TFP410_CTL1: 0x%02X\n", val);
 	tfp410_readb(dvo, TFP410_CTL_2, &val);
-	DRM_LOG_KMS("TFP410_CTL2: 0x%02X\n", val);
+	DRM_DEBUG_KMS("TFP410_CTL2: 0x%02X\n", val);
 	tfp410_readb(dvo, TFP410_CTL_3, &val);
-	DRM_LOG_KMS("TFP410_CTL3: 0x%02X\n", val);
+	DRM_DEBUG_KMS("TFP410_CTL3: 0x%02X\n", val);
 	tfp410_readb(dvo, TFP410_USERCFG, &val);
-	DRM_LOG_KMS("TFP410_USERCFG: 0x%02X\n", val);
+	DRM_DEBUG_KMS("TFP410_USERCFG: 0x%02X\n", val);
 	tfp410_readb(dvo, TFP410_DE_DLY, &val);
-	DRM_LOG_KMS("TFP410_DE_DLY: 0x%02X\n", val);
+	DRM_DEBUG_KMS("TFP410_DE_DLY: 0x%02X\n", val);
 	tfp410_readb(dvo, TFP410_DE_CTL, &val);
-	DRM_LOG_KMS("TFP410_DE_CTL: 0x%02X\n", val);
+	DRM_DEBUG_KMS("TFP410_DE_CTL: 0x%02X\n", val);
 	tfp410_readb(dvo, TFP410_DE_TOP, &val);
-	DRM_LOG_KMS("TFP410_DE_TOP: 0x%02X\n", val);
+	DRM_DEBUG_KMS("TFP410_DE_TOP: 0x%02X\n", val);
 	tfp410_readb(dvo, TFP410_DE_CNT_LO, &val);
 	tfp410_readb(dvo, TFP410_DE_CNT_HI, &val2);
-	DRM_LOG_KMS("TFP410_DE_CNT: 0x%02X%02X\n", val2, val);
+	DRM_DEBUG_KMS("TFP410_DE_CNT: 0x%02X%02X\n", val2, val);
 	tfp410_readb(dvo, TFP410_DE_LIN_LO, &val);
 	tfp410_readb(dvo, TFP410_DE_LIN_HI, &val2);
-	DRM_LOG_KMS("TFP410_DE_LIN: 0x%02X%02X\n", val2, val);
+	DRM_DEBUG_KMS("TFP410_DE_LIN: 0x%02X%02X\n", val2, val);
 	tfp410_readb(dvo, TFP410_H_RES_LO, &val);
 	tfp410_readb(dvo, TFP410_H_RES_HI, &val2);
-	DRM_LOG_KMS("TFP410_H_RES: 0x%02X%02X\n", val2, val);
+	DRM_DEBUG_KMS("TFP410_H_RES: 0x%02X%02X\n", val2, val);
 	tfp410_readb(dvo, TFP410_V_RES_LO, &val);
 	tfp410_readb(dvo, TFP410_V_RES_HI, &val2);
-	DRM_LOG_KMS("TFP410_V_RES: 0x%02X%02X\n", val2, val);
+	DRM_DEBUG_KMS("TFP410_V_RES: 0x%02X%02X\n", val2, val);
 }
 
 static void tfp410_destroy(struct intel_dvo_device *dvo)
diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
new file mode 100644
index 00000000000..9d7954366bd
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
@@ -0,0 +1,1061 @@
+/*
+ * Copyright © 2013 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ * Authors:
+ *    Brad Volkin <bradley.d.volkin@intel.com>
+ *
+ */
+
+#include "i915_drv.h"
+
+/**
+ * DOC: batch buffer command parser
+ *
+ * Motivation:
+ * Certain OpenGL features (e.g. transform feedback, performance monitoring)
+ * require userspace code to submit batches containing commands such as
+ * MI_LOAD_REGISTER_IMM to access various registers. Unfortunately, some
+ * generations of the hardware will noop these commands in "unsecure" batches
+ * (which includes all userspace batches submitted via i915) even though the
+ * commands may be safe and represent the intended programming model of the
+ * device.
+ *
+ * The software command parser is similar in operation to the command parsing
+ * done in hardware for unsecure batches. However, the software parser allows
+ * some operations that would be noop'd by hardware, if the parser determines
+ * the operation is safe, and submits the batch as "secure" to prevent hardware
+ * parsing.
+ *
+ * Threats:
+ * At a high level, the hardware (and software) checks attempt to prevent
+ * granting userspace undue privileges. There are three categories of privilege.
+ *
+ * First, commands which are explicitly defined as privileged or which should
+ * only be used by the kernel driver. The parser generally rejects such
+ * commands, though it may allow some from the drm master process.
+ *
+ * Second, commands which access registers. To support correct/enhanced
+ * userspace functionality, particularly certain OpenGL extensions, the parser
+ * provides a whitelist of registers which userspace may safely access (for both
+ * normal and drm master processes).
+ *
+ * Third, commands which access privileged memory (i.e. GGTT, HWS page, etc).
+ * The parser always rejects such commands.
+ *
+ * The majority of the problematic commands fall in the MI_* range, with only a
+ * few specific commands on each ring (e.g. PIPE_CONTROL and MI_FLUSH_DW).
+ *
+ * Implementation:
+ * Each ring maintains tables of commands and registers which the parser uses in
+ * scanning batch buffers submitted to that ring.
+ *
+ * Since the set of commands that the parser must check for is significantly
+ * smaller than the number of commands supported, the parser tables contain only
+ * those commands required by the parser. This generally works because command
+ * opcode ranges have standard command length encodings. So for commands that
+ * the parser does not need to check, it can easily skip them. This is
+ * implementated via a per-ring length decoding vfunc.
+ *
+ * Unfortunately, there are a number of commands that do not follow the standard
+ * length encoding for their opcode range, primarily amongst the MI_* commands.
+ * To handle this, the parser provides a way to define explicit "skip" entries
+ * in the per-ring command tables.
+ *
+ * Other command table entries map fairly directly to high level categories
+ * mentioned above: rejected, master-only, register whitelist. The parser
+ * implements a number of checks, including the privileged memory checks, via a
+ * general bitmasking mechanism.
+ */
+
+#define STD_MI_OPCODE_MASK  0xFF800000
+#define STD_3D_OPCODE_MASK  0xFFFF0000
+#define STD_2D_OPCODE_MASK  0xFFC00000
+#define STD_MFX_OPCODE_MASK 0xFFFF0000
+
+#define CMD(op, opm, f, lm, fl, ...)				\
+	{							\
+		.flags = (fl) | ((f) ? CMD_DESC_FIXED : 0),	\
+		.cmd = { (op), (opm) }, 			\
+		.length = { (lm) },				\
+		__VA_ARGS__					\
+	}
+
+/* Convenience macros to compress the tables */
+#define SMI STD_MI_OPCODE_MASK
+#define S3D STD_3D_OPCODE_MASK
+#define S2D STD_2D_OPCODE_MASK
+#define SMFX STD_MFX_OPCODE_MASK
+#define F true
+#define S CMD_DESC_SKIP
+#define R CMD_DESC_REJECT
+#define W CMD_DESC_REGISTER
+#define B CMD_DESC_BITMASK
+#define M CMD_DESC_MASTER
+
+/*            Command                          Mask   Fixed Len   Action
+	      ---------------------------------------------------------- */
+static const struct drm_i915_cmd_descriptor common_cmds[] = {
+	CMD(  MI_NOOP,                          SMI,    F,  1,      S  ),
+	CMD(  MI_USER_INTERRUPT,                SMI,    F,  1,      R  ),
+	CMD(  MI_WAIT_FOR_EVENT,                SMI,    F,  1,      M  ),
+	CMD(  MI_ARB_CHECK,                     SMI,    F,  1,      S  ),
+	CMD(  MI_REPORT_HEAD,                   SMI,    F,  1,      S  ),
+	CMD(  MI_SUSPEND_FLUSH,                 SMI,    F,  1,      S  ),
+	CMD(  MI_SEMAPHORE_MBOX,                SMI,   !F,  0xFF,   R  ),
+	CMD(  MI_STORE_DWORD_INDEX,             SMI,   !F,  0xFF,   R  ),
+	CMD(  MI_LOAD_REGISTER_IMM(1),          SMI,   !F,  0xFF,   W,
+	      .reg = { .offset = 1, .mask = 0x007FFFFC }               ),
+	CMD(  MI_STORE_REGISTER_MEM(1),         SMI,   !F,  0xFF,   W | B,
+	      .reg = { .offset = 1, .mask = 0x007FFFFC },
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+	CMD(  MI_LOAD_REGISTER_MEM,             SMI,   !F,  0xFF,   W | B,
+	      .reg = { .offset = 1, .mask = 0x007FFFFC },
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+	CMD(  MI_BATCH_BUFFER_START,            SMI,   !F,  0xFF,   S  ),
+};
+
+static const struct drm_i915_cmd_descriptor render_cmds[] = {
+	CMD(  MI_FLUSH,                         SMI,    F,  1,      S  ),
+	CMD(  MI_ARB_ON_OFF,                    SMI,    F,  1,      R  ),
+	CMD(  MI_PREDICATE,                     SMI,    F,  1,      S  ),
+	CMD(  MI_TOPOLOGY_FILTER,               SMI,    F,  1,      S  ),
+	CMD(  MI_DISPLAY_FLIP,                  SMI,   !F,  0xFF,   R  ),
+	CMD(  MI_SET_CONTEXT,                   SMI,   !F,  0xFF,   R  ),
+	CMD(  MI_URB_CLEAR,                     SMI,   !F,  0xFF,   S  ),
+	CMD(  MI_STORE_DWORD_IMM,               SMI,   !F,  0x3F,   B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+	CMD(  MI_UPDATE_GTT,                    SMI,   !F,  0xFF,   R  ),
+	CMD(  MI_CLFLUSH,                       SMI,   !F,  0x3FF,  B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+	CMD(  MI_REPORT_PERF_COUNT,             SMI,   !F,  0x3F,   B,
+	      .bits = {{
+			.offset = 1,
+			.mask = MI_REPORT_PERF_COUNT_GGTT,
+			.expected = 0,
+	      }},						       ),
+	CMD(  MI_CONDITIONAL_BATCH_BUFFER_END,  SMI,   !F,  0xFF,   B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+	CMD(  GFX_OP_3DSTATE_VF_STATISTICS,     S3D,    F,  1,      S  ),
+	CMD(  PIPELINE_SELECT,                  S3D,    F,  1,      S  ),
+	CMD(  MEDIA_VFE_STATE,			S3D,   !F,  0xFFFF, B,
+	      .bits = {{
+			.offset = 2,
+			.mask = MEDIA_VFE_STATE_MMIO_ACCESS_MASK,
+			.expected = 0,
+	      }},						       ),
+	CMD(  GPGPU_OBJECT,                     S3D,   !F,  0xFF,   S  ),
+	CMD(  GPGPU_WALKER,                     S3D,   !F,  0xFF,   S  ),
+	CMD(  GFX_OP_3DSTATE_SO_DECL_LIST,      S3D,   !F,  0x1FF,  S  ),
+	CMD(  GFX_OP_PIPE_CONTROL(5),           S3D,   !F,  0xFF,   B,
+	      .bits = {{
+			.offset = 1,
+			.mask = (PIPE_CONTROL_MMIO_WRITE | PIPE_CONTROL_NOTIFY),
+			.expected = 0,
+	      },
+	      {
+			.offset = 1,
+		        .mask = (PIPE_CONTROL_GLOBAL_GTT_IVB |
+				 PIPE_CONTROL_STORE_DATA_INDEX),
+			.expected = 0,
+			.condition_offset = 1,
+			.condition_mask = PIPE_CONTROL_POST_SYNC_OP_MASK,
+	      }},						       ),
+};
+
+static const struct drm_i915_cmd_descriptor hsw_render_cmds[] = {
+	CMD(  MI_SET_PREDICATE,                 SMI,    F,  1,      S  ),
+	CMD(  MI_RS_CONTROL,                    SMI,    F,  1,      S  ),
+	CMD(  MI_URB_ATOMIC_ALLOC,              SMI,    F,  1,      S  ),
+	CMD(  MI_RS_CONTEXT,                    SMI,    F,  1,      S  ),
+	CMD(  MI_LOAD_SCAN_LINES_INCL,          SMI,   !F,  0x3F,   M  ),
+	CMD(  MI_LOAD_SCAN_LINES_EXCL,          SMI,   !F,  0x3F,   R  ),
+	CMD(  MI_LOAD_REGISTER_REG,             SMI,   !F,  0xFF,   R  ),
+	CMD(  MI_RS_STORE_DATA_IMM,             SMI,   !F,  0xFF,   S  ),
+	CMD(  MI_LOAD_URB_MEM,                  SMI,   !F,  0xFF,   S  ),
+	CMD(  MI_STORE_URB_MEM,                 SMI,   !F,  0xFF,   S  ),
+	CMD(  GFX_OP_3DSTATE_DX9_CONSTANTF_VS,  S3D,   !F,  0x7FF,  S  ),
+	CMD(  GFX_OP_3DSTATE_DX9_CONSTANTF_PS,  S3D,   !F,  0x7FF,  S  ),
+
+	CMD(  GFX_OP_3DSTATE_BINDING_TABLE_EDIT_VS,  S3D,   !F,  0x1FF,  S  ),
+	CMD(  GFX_OP_3DSTATE_BINDING_TABLE_EDIT_GS,  S3D,   !F,  0x1FF,  S  ),
+	CMD(  GFX_OP_3DSTATE_BINDING_TABLE_EDIT_HS,  S3D,   !F,  0x1FF,  S  ),
+	CMD(  GFX_OP_3DSTATE_BINDING_TABLE_EDIT_DS,  S3D,   !F,  0x1FF,  S  ),
+	CMD(  GFX_OP_3DSTATE_BINDING_TABLE_EDIT_PS,  S3D,   !F,  0x1FF,  S  ),
+};
+
+static const struct drm_i915_cmd_descriptor video_cmds[] = {
+	CMD(  MI_ARB_ON_OFF,                    SMI,    F,  1,      R  ),
+	CMD(  MI_STORE_DWORD_IMM,               SMI,   !F,  0xFF,   B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+	CMD(  MI_UPDATE_GTT,                    SMI,   !F,  0x3F,   R  ),
+	CMD(  MI_FLUSH_DW,                      SMI,   !F,  0x3F,   B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_FLUSH_DW_NOTIFY,
+			.expected = 0,
+	      },
+	      {
+			.offset = 1,
+			.mask = MI_FLUSH_DW_USE_GTT,
+			.expected = 0,
+			.condition_offset = 0,
+			.condition_mask = MI_FLUSH_DW_OP_MASK,
+	      },
+	      {
+			.offset = 0,
+			.mask = MI_FLUSH_DW_STORE_INDEX,
+			.expected = 0,
+			.condition_offset = 0,
+			.condition_mask = MI_FLUSH_DW_OP_MASK,
+	      }},						       ),
+	CMD(  MI_CONDITIONAL_BATCH_BUFFER_END,  SMI,   !F,  0xFF,   B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+	/*
+	 * MFX_WAIT doesn't fit the way we handle length for most commands.
+	 * It has a length field but it uses a non-standard length bias.
+	 * It is always 1 dword though, so just treat it as fixed length.
+	 */
+	CMD(  MFX_WAIT,                         SMFX,   F,  1,      S  ),
+};
+
+static const struct drm_i915_cmd_descriptor vecs_cmds[] = {
+	CMD(  MI_ARB_ON_OFF,                    SMI,    F,  1,      R  ),
+	CMD(  MI_STORE_DWORD_IMM,               SMI,   !F,  0xFF,   B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+	CMD(  MI_UPDATE_GTT,                    SMI,   !F,  0x3F,   R  ),
+	CMD(  MI_FLUSH_DW,                      SMI,   !F,  0x3F,   B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_FLUSH_DW_NOTIFY,
+			.expected = 0,
+	      },
+	      {
+			.offset = 1,
+			.mask = MI_FLUSH_DW_USE_GTT,
+			.expected = 0,
+			.condition_offset = 0,
+			.condition_mask = MI_FLUSH_DW_OP_MASK,
+	      },
+	      {
+			.offset = 0,
+			.mask = MI_FLUSH_DW_STORE_INDEX,
+			.expected = 0,
+			.condition_offset = 0,
+			.condition_mask = MI_FLUSH_DW_OP_MASK,
+	      }},						       ),
+	CMD(  MI_CONDITIONAL_BATCH_BUFFER_END,  SMI,   !F,  0xFF,   B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+};
+
+static const struct drm_i915_cmd_descriptor blt_cmds[] = {
+	CMD(  MI_DISPLAY_FLIP,                  SMI,   !F,  0xFF,   R  ),
+	CMD(  MI_STORE_DWORD_IMM,               SMI,   !F,  0x3FF,  B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_GLOBAL_GTT,
+			.expected = 0,
+	      }},						       ),
+	CMD(  MI_UPDATE_GTT,                    SMI,   !F,  0x3F,   R  ),
+	CMD(  MI_FLUSH_DW,                      SMI,   !F,  0x3F,   B,
+	      .bits = {{
+			.offset = 0,
+			.mask = MI_FLUSH_DW_NOTIFY,
+			.expected = 0,
+	      },
+	      {
+			.offset = 1,
+			.mask = MI_FLUSH_DW_USE_GTT,
+			.expected = 0,
+			.condition_offset = 0,
+			.condition_mask = MI_FLUSH_DW_OP_MASK,
+	      },
+	      {
+			.offset = 0,
+			.mask = MI_FLUSH_DW_STORE_INDEX,
+			.expected = 0,
+			.condition_offset = 0,
+			.condition_mask = MI_FLUSH_DW_OP_MASK,
+	      }},						       ),
+	CMD(  COLOR_BLT,                        S2D,   !F,  0x3F,   S  ),
+	CMD(  SRC_COPY_BLT,                     S2D,   !F,  0x3F,   S  ),
+};
+
+static const struct drm_i915_cmd_descriptor hsw_blt_cmds[] = {
+	CMD(  MI_LOAD_SCAN_LINES_INCL,          SMI,   !F,  0x3F,   M  ),
+	CMD(  MI_LOAD_SCAN_LINES_EXCL,          SMI,   !F,  0x3F,   R  ),
+};
+
+#undef CMD
+#undef SMI
+#undef S3D
+#undef S2D
+#undef SMFX
+#undef F
+#undef S
+#undef R
+#undef W
+#undef B
+#undef M
+
+static const struct drm_i915_cmd_table gen7_render_cmds[] = {
+	{ common_cmds, ARRAY_SIZE(common_cmds) },
+	{ render_cmds, ARRAY_SIZE(render_cmds) },
+};
+
+static const struct drm_i915_cmd_table hsw_render_ring_cmds[] = {
+	{ common_cmds, ARRAY_SIZE(common_cmds) },
+	{ render_cmds, ARRAY_SIZE(render_cmds) },
+	{ hsw_render_cmds, ARRAY_SIZE(hsw_render_cmds) },
+};
+
+static const struct drm_i915_cmd_table gen7_video_cmds[] = {
+	{ common_cmds, ARRAY_SIZE(common_cmds) },
+	{ video_cmds, ARRAY_SIZE(video_cmds) },
+};
+
+static const struct drm_i915_cmd_table hsw_vebox_cmds[] = {
+	{ common_cmds, ARRAY_SIZE(common_cmds) },
+	{ vecs_cmds, ARRAY_SIZE(vecs_cmds) },
+};
+
+static const struct drm_i915_cmd_table gen7_blt_cmds[] = {
+	{ common_cmds, ARRAY_SIZE(common_cmds) },
+	{ blt_cmds, ARRAY_SIZE(blt_cmds) },
+};
+
+static const struct drm_i915_cmd_table hsw_blt_ring_cmds[] = {
+	{ common_cmds, ARRAY_SIZE(common_cmds) },
+	{ blt_cmds, ARRAY_SIZE(blt_cmds) },
+	{ hsw_blt_cmds, ARRAY_SIZE(hsw_blt_cmds) },
+};
+
+/*
+ * Register whitelists, sorted by increasing register offset.
+ *
+ * Some registers that userspace accesses are 64 bits. The register
+ * access commands only allow 32-bit accesses. Hence, we have to include
+ * entries for both halves of the 64-bit registers.
+ */
+
+/* Convenience macro for adding 64-bit registers */
+#define REG64(addr) (addr), (addr + sizeof(u32))
+
+static const u32 gen7_render_regs[] = {
+	REG64(HS_INVOCATION_COUNT),
+	REG64(DS_INVOCATION_COUNT),
+	REG64(IA_VERTICES_COUNT),
+	REG64(IA_PRIMITIVES_COUNT),
+	REG64(VS_INVOCATION_COUNT),
+	REG64(GS_INVOCATION_COUNT),
+	REG64(GS_PRIMITIVES_COUNT),
+	REG64(CL_INVOCATION_COUNT),
+	REG64(CL_PRIMITIVES_COUNT),
+	REG64(PS_INVOCATION_COUNT),
+	REG64(PS_DEPTH_COUNT),
+	OACONTROL, /* Only allowed for LRI and SRM. See below. */
+	GEN7_3DPRIM_END_OFFSET,
+	GEN7_3DPRIM_START_VERTEX,
+	GEN7_3DPRIM_VERTEX_COUNT,
+	GEN7_3DPRIM_INSTANCE_COUNT,
+	GEN7_3DPRIM_START_INSTANCE,
+	GEN7_3DPRIM_BASE_VERTEX,
+	REG64(GEN7_SO_NUM_PRIMS_WRITTEN(0)),
+	REG64(GEN7_SO_NUM_PRIMS_WRITTEN(1)),
+	REG64(GEN7_SO_NUM_PRIMS_WRITTEN(2)),
+	REG64(GEN7_SO_NUM_PRIMS_WRITTEN(3)),
+	REG64(GEN7_SO_PRIM_STORAGE_NEEDED(0)),
+	REG64(GEN7_SO_PRIM_STORAGE_NEEDED(1)),
+	REG64(GEN7_SO_PRIM_STORAGE_NEEDED(2)),
+	REG64(GEN7_SO_PRIM_STORAGE_NEEDED(3)),
+	GEN7_SO_WRITE_OFFSET(0),
+	GEN7_SO_WRITE_OFFSET(1),
+	GEN7_SO_WRITE_OFFSET(2),
+	GEN7_SO_WRITE_OFFSET(3),
+};
+
+static const u32 gen7_blt_regs[] = {
+	BCS_SWCTRL,
+};
+
+static const u32 ivb_master_regs[] = {
+	FORCEWAKE_MT,
+	DERRMR,
+	GEN7_PIPE_DE_LOAD_SL(PIPE_A),
+	GEN7_PIPE_DE_LOAD_SL(PIPE_B),
+	GEN7_PIPE_DE_LOAD_SL(PIPE_C),
+};
+
+static const u32 hsw_master_regs[] = {
+	FORCEWAKE_MT,
+	DERRMR,
+};
+
+#undef REG64
+
+static u32 gen7_render_get_cmd_length_mask(u32 cmd_header)
+{
+	u32 client = (cmd_header & INSTR_CLIENT_MASK) >> INSTR_CLIENT_SHIFT;
+	u32 subclient =
+		(cmd_header & INSTR_SUBCLIENT_MASK) >> INSTR_SUBCLIENT_SHIFT;
+
+	if (client == INSTR_MI_CLIENT)
+		return 0x3F;
+	else if (client == INSTR_RC_CLIENT) {
+		if (subclient == INSTR_MEDIA_SUBCLIENT)
+			return 0xFFFF;
+		else
+			return 0xFF;
+	}
+
+	DRM_DEBUG_DRIVER("CMD: Abnormal rcs cmd length! 0x%08X\n", cmd_header);
+	return 0;
+}
+
+static u32 gen7_bsd_get_cmd_length_mask(u32 cmd_header)
+{
+	u32 client = (cmd_header & INSTR_CLIENT_MASK) >> INSTR_CLIENT_SHIFT;
+	u32 subclient =
+		(cmd_header & INSTR_SUBCLIENT_MASK) >> INSTR_SUBCLIENT_SHIFT;
+
+	if (client == INSTR_MI_CLIENT)
+		return 0x3F;
+	else if (client == INSTR_RC_CLIENT) {
+		if (subclient == INSTR_MEDIA_SUBCLIENT)
+			return 0xFFF;
+		else
+			return 0xFF;
+	}
+
+	DRM_DEBUG_DRIVER("CMD: Abnormal bsd cmd length! 0x%08X\n", cmd_header);
+	return 0;
+}
+
+static u32 gen7_blt_get_cmd_length_mask(u32 cmd_header)
+{
+	u32 client = (cmd_header & INSTR_CLIENT_MASK) >> INSTR_CLIENT_SHIFT;
+
+	if (client == INSTR_MI_CLIENT)
+		return 0x3F;
+	else if (client == INSTR_BC_CLIENT)
+		return 0xFF;
+
+	DRM_DEBUG_DRIVER("CMD: Abnormal blt cmd length! 0x%08X\n", cmd_header);
+	return 0;
+}
+
+static bool validate_cmds_sorted(struct intel_engine_cs *ring,
+				 const struct drm_i915_cmd_table *cmd_tables,
+				 int cmd_table_count)
+{
+	int i;
+	bool ret = true;
+
+	if (!cmd_tables || cmd_table_count == 0)
+		return true;
+
+	for (i = 0; i < cmd_table_count; i++) {
+		const struct drm_i915_cmd_table *table = &cmd_tables[i];
+		u32 previous = 0;
+		int j;
+
+		for (j = 0; j < table->count; j++) {
+			const struct drm_i915_cmd_descriptor *desc =
+				&table->table[i];
+			u32 curr = desc->cmd.value & desc->cmd.mask;
+
+			if (curr < previous) {
+				DRM_ERROR("CMD: table not sorted ring=%d table=%d entry=%d cmd=0x%08X prev=0x%08X\n",
+					  ring->id, i, j, curr, previous);
+				ret = false;
+			}
+
+			previous = curr;
+		}
+	}
+
+	return ret;
+}
+
+static bool check_sorted(int ring_id, const u32 *reg_table, int reg_count)
+{
+	int i;
+	u32 previous = 0;
+	bool ret = true;
+
+	for (i = 0; i < reg_count; i++) {
+		u32 curr = reg_table[i];
+
+		if (curr < previous) {
+			DRM_ERROR("CMD: table not sorted ring=%d entry=%d reg=0x%08X prev=0x%08X\n",
+				  ring_id, i, curr, previous);
+			ret = false;
+		}
+
+		previous = curr;
+	}
+
+	return ret;
+}
+
+static bool validate_regs_sorted(struct intel_engine_cs *ring)
+{
+	return check_sorted(ring->id, ring->reg_table, ring->reg_count) &&
+		check_sorted(ring->id, ring->master_reg_table,
+			     ring->master_reg_count);
+}
+
+struct cmd_node {
+	const struct drm_i915_cmd_descriptor *desc;
+	struct hlist_node node;
+};
+
+/*
+ * Different command ranges have different numbers of bits for the opcode. For
+ * example, MI commands use bits 31:23 while 3D commands use bits 31:16. The
+ * problem is that, for example, MI commands use bits 22:16 for other fields
+ * such as GGTT vs PPGTT bits. If we include those bits in the mask then when
+ * we mask a command from a batch it could hash to the wrong bucket due to
+ * non-opcode bits being set. But if we don't include those bits, some 3D
+ * commands may hash to the same bucket due to not including opcode bits that
+ * make the command unique. For now, we will risk hashing to the same bucket.
+ *
+ * If we attempt to generate a perfect hash, we should be able to look at bits
+ * 31:29 of a command from a batch buffer and use the full mask for that
+ * client. The existing INSTR_CLIENT_MASK/SHIFT defines can be used for this.
+ */
+#define CMD_HASH_MASK STD_MI_OPCODE_MASK
+
+static int init_hash_table(struct intel_engine_cs *ring,
+			   const struct drm_i915_cmd_table *cmd_tables,
+			   int cmd_table_count)
+{
+	int i, j;
+
+	hash_init(ring->cmd_hash);
+
+	for (i = 0; i < cmd_table_count; i++) {
+		const struct drm_i915_cmd_table *table = &cmd_tables[i];
+
+		for (j = 0; j < table->count; j++) {
+			const struct drm_i915_cmd_descriptor *desc =
+				&table->table[j];
+			struct cmd_node *desc_node =
+				kmalloc(sizeof(*desc_node), GFP_KERNEL);
+
+			if (!desc_node)
+				return -ENOMEM;
+
+			desc_node->desc = desc;
+			hash_add(ring->cmd_hash, &desc_node->node,
+				 desc->cmd.value & CMD_HASH_MASK);
+		}
+	}
+
+	return 0;
+}
+
+static void fini_hash_table(struct intel_engine_cs *ring)
+{
+	struct hlist_node *tmp;
+	struct cmd_node *desc_node;
+	int i;
+
+	hash_for_each_safe(ring->cmd_hash, i, tmp, desc_node, node) {
+		hash_del(&desc_node->node);
+		kfree(desc_node);
+	}
+}
+
+/**
+ * i915_cmd_parser_init_ring() - set cmd parser related fields for a ringbuffer
+ * @ring: the ringbuffer to initialize
+ *
+ * Optionally initializes fields related to batch buffer command parsing in the
+ * struct intel_engine_cs based on whether the platform requires software
+ * command parsing.
+ *
+ * Return: non-zero if initialization fails
+ */
+int i915_cmd_parser_init_ring(struct intel_engine_cs *ring)
+{
+	const struct drm_i915_cmd_table *cmd_tables;
+	int cmd_table_count;
+	int ret;
+
+	if (!IS_GEN7(ring->dev))
+		return 0;
+
+	switch (ring->id) {
+	case RCS:
+		if (IS_HASWELL(ring->dev)) {
+			cmd_tables = hsw_render_ring_cmds;
+			cmd_table_count =
+				ARRAY_SIZE(hsw_render_ring_cmds);
+		} else {
+			cmd_tables = gen7_render_cmds;
+			cmd_table_count = ARRAY_SIZE(gen7_render_cmds);
+		}
+
+		ring->reg_table = gen7_render_regs;
+		ring->reg_count = ARRAY_SIZE(gen7_render_regs);
+
+		if (IS_HASWELL(ring->dev)) {
+			ring->master_reg_table = hsw_master_regs;
+			ring->master_reg_count = ARRAY_SIZE(hsw_master_regs);
+		} else {
+			ring->master_reg_table = ivb_master_regs;
+			ring->master_reg_count = ARRAY_SIZE(ivb_master_regs);
+		}
+
+		ring->get_cmd_length_mask = gen7_render_get_cmd_length_mask;
+		break;
+	case VCS:
+		cmd_tables = gen7_video_cmds;
+		cmd_table_count = ARRAY_SIZE(gen7_video_cmds);
+		ring->get_cmd_length_mask = gen7_bsd_get_cmd_length_mask;
+		break;
+	case BCS:
+		if (IS_HASWELL(ring->dev)) {
+			cmd_tables = hsw_blt_ring_cmds;
+			cmd_table_count = ARRAY_SIZE(hsw_blt_ring_cmds);
+		} else {
+			cmd_tables = gen7_blt_cmds;
+			cmd_table_count = ARRAY_SIZE(gen7_blt_cmds);
+		}
+
+		ring->reg_table = gen7_blt_regs;
+		ring->reg_count = ARRAY_SIZE(gen7_blt_regs);
+
+		if (IS_HASWELL(ring->dev)) {
+			ring->master_reg_table = hsw_master_regs;
+			ring->master_reg_count = ARRAY_SIZE(hsw_master_regs);
+		} else {
+			ring->master_reg_table = ivb_master_regs;
+			ring->master_reg_count = ARRAY_SIZE(ivb_master_regs);
+		}
+
+		ring->get_cmd_length_mask = gen7_blt_get_cmd_length_mask;
+		break;
+	case VECS:
+		cmd_tables = hsw_vebox_cmds;
+		cmd_table_count = ARRAY_SIZE(hsw_vebox_cmds);
+		/* VECS can use the same length_mask function as VCS */
+		ring->get_cmd_length_mask = gen7_bsd_get_cmd_length_mask;
+		break;
+	default:
+		DRM_ERROR("CMD: cmd_parser_init with unknown ring: %d\n",
+			  ring->id);
+		BUG();
+	}
+
+	BUG_ON(!validate_cmds_sorted(ring, cmd_tables, cmd_table_count));
+	BUG_ON(!validate_regs_sorted(ring));
+
+	ret = init_hash_table(ring, cmd_tables, cmd_table_count);
+	if (ret) {
+		DRM_ERROR("CMD: cmd_parser_init failed!\n");
+		fini_hash_table(ring);
+		return ret;
+	}
+
+	ring->needs_cmd_parser = true;
+
+	return 0;
+}
+
+/**
+ * i915_cmd_parser_fini_ring() - clean up cmd parser related fields
+ * @ring: the ringbuffer to clean up
+ *
+ * Releases any resources related to command parsing that may have been
+ * initialized for the specified ring.
+ */
+void i915_cmd_parser_fini_ring(struct intel_engine_cs *ring)
+{
+	if (!ring->needs_cmd_parser)
+		return;
+
+	fini_hash_table(ring);
+}
+
+static const struct drm_i915_cmd_descriptor*
+find_cmd_in_table(struct intel_engine_cs *ring,
+		  u32 cmd_header)
+{
+	struct cmd_node *desc_node;
+
+	hash_for_each_possible(ring->cmd_hash, desc_node, node,
+			       cmd_header & CMD_HASH_MASK) {
+		const struct drm_i915_cmd_descriptor *desc = desc_node->desc;
+		u32 masked_cmd = desc->cmd.mask & cmd_header;
+		u32 masked_value = desc->cmd.value & desc->cmd.mask;
+
+		if (masked_cmd == masked_value)
+			return desc;
+	}
+
+	return NULL;
+}
+
+/*
+ * Returns a pointer to a descriptor for the command specified by cmd_header.
+ *
+ * The caller must supply space for a default descriptor via the default_desc
+ * parameter. If no descriptor for the specified command exists in the ring's
+ * command parser tables, this function fills in default_desc based on the
+ * ring's default length encoding and returns default_desc.
+ */
+static const struct drm_i915_cmd_descriptor*
+find_cmd(struct intel_engine_cs *ring,
+	 u32 cmd_header,
+	 struct drm_i915_cmd_descriptor *default_desc)
+{
+	const struct drm_i915_cmd_descriptor *desc;
+	u32 mask;
+
+	desc = find_cmd_in_table(ring, cmd_header);
+	if (desc)
+		return desc;
+
+	mask = ring->get_cmd_length_mask(cmd_header);
+	if (!mask)
+		return NULL;
+
+	BUG_ON(!default_desc);
+	default_desc->flags = CMD_DESC_SKIP;
+	default_desc->length.mask = mask;
+
+	return default_desc;
+}
+
+static bool valid_reg(const u32 *table, int count, u32 addr)
+{
+	if (table && count != 0) {
+		int i;
+
+		for (i = 0; i < count; i++) {
+			if (table[i] == addr)
+				return true;
+		}
+	}
+
+	return false;
+}
+
+static u32 *vmap_batch(struct drm_i915_gem_object *obj)
+{
+	int i;
+	void *addr = NULL;
+	struct sg_page_iter sg_iter;
+	struct page **pages;
+
+	pages = drm_malloc_ab(obj->base.size >> PAGE_SHIFT, sizeof(*pages));
+	if (pages == NULL) {
+		DRM_DEBUG_DRIVER("Failed to get space for pages\n");
+		goto finish;
+	}
+
+	i = 0;
+	for_each_sg_page(obj->pages->sgl, &sg_iter, obj->pages->nents, 0) {
+		pages[i] = sg_page_iter_page(&sg_iter);
+		i++;
+	}
+
+	addr = vmap(pages, i, 0, PAGE_KERNEL);
+	if (addr == NULL) {
+		DRM_DEBUG_DRIVER("Failed to vmap pages\n");
+		goto finish;
+	}
+
+finish:
+	if (pages)
+		drm_free_large(pages);
+	return (u32*)addr;
+}
+
+/**
+ * i915_needs_cmd_parser() - should a given ring use software command parsing?
+ * @ring: the ring in question
+ *
+ * Only certain platforms require software batch buffer command parsing, and
+ * only when enabled via module paramter.
+ *
+ * Return: true if the ring requires software command parsing
+ */
+bool i915_needs_cmd_parser(struct intel_engine_cs *ring)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+
+	if (!ring->needs_cmd_parser)
+		return false;
+
+	/*
+	 * XXX: VLV is Gen7 and therefore has cmd_tables, but has PPGTT
+	 * disabled. That will cause all of the parser's PPGTT checks to
+	 * fail. For now, disable parsing when PPGTT is off.
+	 */
+	if (!dev_priv->mm.aliasing_ppgtt)
+		return false;
+
+	return (i915.enable_cmd_parser == 1);
+}
+
+static bool check_cmd(const struct intel_engine_cs *ring,
+		      const struct drm_i915_cmd_descriptor *desc,
+		      const u32 *cmd,
+		      const bool is_master,
+		      bool *oacontrol_set)
+{
+	if (desc->flags & CMD_DESC_REJECT) {
+		DRM_DEBUG_DRIVER("CMD: Rejected command: 0x%08X\n", *cmd);
+		return false;
+	}
+
+	if ((desc->flags & CMD_DESC_MASTER) && !is_master) {
+		DRM_DEBUG_DRIVER("CMD: Rejected master-only command: 0x%08X\n",
+				 *cmd);
+		return false;
+	}
+
+	if (desc->flags & CMD_DESC_REGISTER) {
+		u32 reg_addr = cmd[desc->reg.offset] & desc->reg.mask;
+
+		/*
+		 * OACONTROL requires some special handling for writes. We
+		 * want to make sure that any batch which enables OA also
+		 * disables it before the end of the batch. The goal is to
+		 * prevent one process from snooping on the perf data from
+		 * another process. To do that, we need to check the value
+		 * that will be written to the register. Hence, limit
+		 * OACONTROL writes to only MI_LOAD_REGISTER_IMM commands.
+		 */
+		if (reg_addr == OACONTROL) {
+			if (desc->cmd.value == MI_LOAD_REGISTER_MEM)
+				return false;
+
+			if (desc->cmd.value == MI_LOAD_REGISTER_IMM(1))
+				*oacontrol_set = (cmd[2] != 0);
+		}
+
+		if (!valid_reg(ring->reg_table,
+			       ring->reg_count, reg_addr)) {
+			if (!is_master ||
+			    !valid_reg(ring->master_reg_table,
+				       ring->master_reg_count,
+				       reg_addr)) {
+				DRM_DEBUG_DRIVER("CMD: Rejected register 0x%08X in command: 0x%08X (ring=%d)\n",
+						 reg_addr,
+						 *cmd,
+						 ring->id);
+				return false;
+			}
+		}
+	}
+
+	if (desc->flags & CMD_DESC_BITMASK) {
+		int i;
+
+		for (i = 0; i < MAX_CMD_DESC_BITMASKS; i++) {
+			u32 dword;
+
+			if (desc->bits[i].mask == 0)
+				break;
+
+			if (desc->bits[i].condition_mask != 0) {
+				u32 offset =
+					desc->bits[i].condition_offset;
+				u32 condition = cmd[offset] &
+					desc->bits[i].condition_mask;
+
+				if (condition == 0)
+					continue;
+			}
+
+			dword = cmd[desc->bits[i].offset] &
+				desc->bits[i].mask;
+
+			if (dword != desc->bits[i].expected) {
+				DRM_DEBUG_DRIVER("CMD: Rejected command 0x%08X for bitmask 0x%08X (exp=0x%08X act=0x%08X) (ring=%d)\n",
+						 *cmd,
+						 desc->bits[i].mask,
+						 desc->bits[i].expected,
+						 dword, ring->id);
+				return false;
+			}
+		}
+	}
+
+	return true;
+}
+
+#define LENGTH_BIAS 2
+
+/**
+ * i915_parse_cmds() - parse a submitted batch buffer for privilege violations
+ * @ring: the ring on which the batch is to execute
+ * @batch_obj: the batch buffer in question
+ * @batch_start_offset: byte offset in the batch at which execution starts
+ * @is_master: is the submitting process the drm master?
+ *
+ * Parses the specified batch buffer looking for privilege violations as
+ * described in the overview.
+ *
+ * Return: non-zero if the parser finds violations or otherwise fails
+ */
+int i915_parse_cmds(struct intel_engine_cs *ring,
+		    struct drm_i915_gem_object *batch_obj,
+		    u32 batch_start_offset,
+		    bool is_master)
+{
+	int ret = 0;
+	u32 *cmd, *batch_base, *batch_end;
+	struct drm_i915_cmd_descriptor default_desc = { 0 };
+	int needs_clflush = 0;
+	bool oacontrol_set = false; /* OACONTROL tracking. See check_cmd() */
+
+	ret = i915_gem_obj_prepare_shmem_read(batch_obj, &needs_clflush);
+	if (ret) {
+		DRM_DEBUG_DRIVER("CMD: failed to prep read\n");
+		return ret;
+	}
+
+	batch_base = vmap_batch(batch_obj);
+	if (!batch_base) {
+		DRM_DEBUG_DRIVER("CMD: Failed to vmap batch\n");
+		i915_gem_object_unpin_pages(batch_obj);
+		return -ENOMEM;
+	}
+
+	if (needs_clflush)
+		drm_clflush_virt_range((char *)batch_base, batch_obj->base.size);
+
+	cmd = batch_base + (batch_start_offset / sizeof(*cmd));
+	batch_end = cmd + (batch_obj->base.size / sizeof(*batch_end));
+
+	while (cmd < batch_end) {
+		const struct drm_i915_cmd_descriptor *desc;
+		u32 length;
+
+		if (*cmd == MI_BATCH_BUFFER_END)
+			break;
+
+		desc = find_cmd(ring, *cmd, &default_desc);
+		if (!desc) {
+			DRM_DEBUG_DRIVER("CMD: Unrecognized command: 0x%08X\n",
+					 *cmd);
+			ret = -EINVAL;
+			break;
+		}
+
+		if (desc->flags & CMD_DESC_FIXED)
+			length = desc->length.fixed;
+		else
+			length = ((*cmd & desc->length.mask) + LENGTH_BIAS);
+
+		if ((batch_end - cmd) < length) {
+			DRM_DEBUG_DRIVER("CMD: Command length exceeds batch length: 0x%08X length=%u batchlen=%td\n",
+					 *cmd,
+					 length,
+					 batch_end - cmd);
+			ret = -EINVAL;
+			break;
+		}
+
+		if (!check_cmd(ring, desc, cmd, is_master, &oacontrol_set)) {
+			ret = -EINVAL;
+			break;
+		}
+
+		cmd += length;
+	}
+
+	if (oacontrol_set) {
+		DRM_DEBUG_DRIVER("CMD: batch set OACONTROL but did not clear it\n");
+		ret = -EINVAL;
+	}
+
+	if (cmd >= batch_end) {
+		DRM_DEBUG_DRIVER("CMD: Got to the end of the buffer w/o a BBE cmd!\n");
+		ret = -EINVAL;
+	}
+
+	vunmap(batch_base);
+
+	i915_gem_object_unpin_pages(batch_obj);
+
+	return ret;
+}
+
+/**
+ * i915_cmd_parser_get_version() - get the cmd parser version number
+ *
+ * The cmd parser maintains a simple increasing integer version number suitable
+ * for passing to userspace clients to determine what operations are permitted.
+ *
+ * Return: the current version number of the cmd parser
+ */
+int i915_cmd_parser_get_version(void)
+{
+	/*
+	 * Command parser version history
+	 *
+	 * 1. Initial version. Checks batches and reports violations, but leaves
+	 *    hardware parsing enabled (so does not allow new use cases).
+	 */
+	return 1;
+}
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 6ed45a98423..b8c689202c4 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -40,8 +40,6 @@
 #include <drm/i915_drm.h>
 #include "i915_drv.h"
 
-#if defined(CONFIG_DEBUG_FS)
-
 enum {
 	ACTIVE_LIST,
 	INACTIVE_LIST,
@@ -81,7 +79,7 @@ drm_add_fake_info_node(struct drm_minor *minor,
 
 static int i915_capabilities(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	const struct intel_device_info *info = INTEL_INFO(dev);
 
@@ -100,7 +98,7 @@ static const char *get_pin_flag(struct drm_i915_gem_object *obj)
 {
 	if (obj->user_pin_count > 0)
 		return "P";
-	else if (obj->pin_count > 0)
+	else if (i915_gem_obj_is_pinned(obj))
 		return "p";
 	else
 		return " ";
@@ -125,6 +123,8 @@ static void
 describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 {
 	struct i915_vma *vma;
+	int pin_count = 0;
+
 	seq_printf(m, "%pK: %s%s%s %8zdKiB %02x %02x %u %u %u%s%s%s",
 		   &obj->base,
 		   get_pin_flag(obj),
@@ -141,8 +141,10 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 		   obj->madv == I915_MADV_DONTNEED ? " purgeable" : "");
 	if (obj->base.name)
 		seq_printf(m, " (name: %d)", obj->base.name);
-	if (obj->pin_count)
-		seq_printf(m, " (pinned x %d)", obj->pin_count);
+	list_for_each_entry(vma, &obj->vma_list, vma_link)
+		if (vma->pin_count > 0)
+			pin_count++;
+		seq_printf(m, " (pinned x %d)", pin_count);
 	if (obj->pin_display)
 		seq_printf(m, " (display)");
 	if (obj->fence_reg != I915_FENCE_REG_NONE)
@@ -170,7 +172,7 @@ describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 		seq_printf(m, " (%s)", obj->ring->name);
 }
 
-static void describe_ctx(struct seq_file *m, struct i915_hw_context *ctx)
+static void describe_ctx(struct seq_file *m, struct intel_context *ctx)
 {
 	seq_putc(m, ctx->is_initialized ? 'I' : 'i');
 	seq_putc(m, ctx->remap_slice ? 'R' : 'r');
@@ -179,7 +181,7 @@ static void describe_ctx(struct seq_file *m, struct i915_hw_context *ctx)
 
 static int i915_gem_object_list_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	uintptr_t list = (uintptr_t) node->info_ent->data;
 	struct list_head *head;
 	struct drm_device *dev = node->minor->dev;
@@ -237,7 +239,7 @@ static int obj_rank_by_stolen(void *priv,
 
 static int i915_gem_stolen_list_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj;
@@ -297,28 +299,62 @@ static int i915_gem_stolen_list_info(struct seq_file *m, void *data)
 } while (0)
 
 struct file_stats {
+	struct drm_i915_file_private *file_priv;
 	int count;
-	size_t total, active, inactive, unbound;
+	size_t total, unbound;
+	size_t global, shared;
+	size_t active, inactive;
 };
 
 static int per_file_stats(int id, void *ptr, void *data)
 {
 	struct drm_i915_gem_object *obj = ptr;
 	struct file_stats *stats = data;
+	struct i915_vma *vma;
 
 	stats->count++;
 	stats->total += obj->base.size;
 
-	if (i915_gem_obj_ggtt_bound(obj)) {
-		if (!list_empty(&obj->ring_list))
-			stats->active += obj->base.size;
-		else
-			stats->inactive += obj->base.size;
+	if (obj->base.name || obj->base.dma_buf)
+		stats->shared += obj->base.size;
+
+	if (USES_FULL_PPGTT(obj->base.dev)) {
+		list_for_each_entry(vma, &obj->vma_list, vma_link) {
+			struct i915_hw_ppgtt *ppgtt;
+
+			if (!drm_mm_node_allocated(&vma->node))
+				continue;
+
+			if (i915_is_ggtt(vma->vm)) {
+				stats->global += obj->base.size;
+				continue;
+			}
+
+			ppgtt = container_of(vma->vm, struct i915_hw_ppgtt, base);
+			if (ppgtt->ctx && ppgtt->ctx->file_priv != stats->file_priv)
+				continue;
+
+			if (obj->ring) /* XXX per-vma statistic */
+				stats->active += obj->base.size;
+			else
+				stats->inactive += obj->base.size;
+
+			return 0;
+		}
 	} else {
-		if (!list_empty(&obj->global_list))
-			stats->unbound += obj->base.size;
+		if (i915_gem_obj_ggtt_bound(obj)) {
+			stats->global += obj->base.size;
+			if (obj->ring)
+				stats->active += obj->base.size;
+			else
+				stats->inactive += obj->base.size;
+			return 0;
+		}
 	}
 
+	if (!list_empty(&obj->global_list))
+		stats->unbound += obj->base.size;
+
 	return 0;
 }
 
@@ -335,7 +371,7 @@ static int per_file_stats(int id, void *ptr, void *data)
 
 static int i915_gem_object_info(struct seq_file *m, void* data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 count, mappable_count, purgeable_count;
@@ -406,16 +442,31 @@ static int i915_gem_object_info(struct seq_file *m, void* data)
 	seq_putc(m, '\n');
 	list_for_each_entry_reverse(file, &dev->filelist, lhead) {
 		struct file_stats stats;
+		struct task_struct *task;
 
 		memset(&stats, 0, sizeof(stats));
+		stats.file_priv = file->driver_priv;
+		spin_lock(&file->table_lock);
 		idr_for_each(&file->object_idr, per_file_stats, &stats);
-		seq_printf(m, "%s: %u objects, %zu bytes (%zu active, %zu inactive, %zu unbound)\n",
-			   get_pid_task(file->pid, PIDTYPE_PID)->comm,
+		spin_unlock(&file->table_lock);
+		/*
+		 * Although we have a valid reference on file->pid, that does
+		 * not guarantee that the task_struct who called get_pid() is
+		 * still alive (e.g. get_pid(current) => fork() => exit()).
+		 * Therefore, we need to protect this ->comm access using RCU.
+		 */
+		rcu_read_lock();
+		task = pid_task(file->pid, PIDTYPE_PID);
+		seq_printf(m, "%s: %u objects, %zu bytes (%zu active, %zu inactive, %zu global, %zu shared, %zu unbound)\n",
+			   task ? task->comm : "<unknown>",
 			   stats.count,
 			   stats.total,
 			   stats.active,
 			   stats.inactive,
+			   stats.global,
+			   stats.shared,
 			   stats.unbound);
+		rcu_read_unlock();
 	}
 
 	mutex_unlock(&dev->struct_mutex);
@@ -425,7 +476,7 @@ static int i915_gem_object_info(struct seq_file *m, void* data)
 
 static int i915_gem_gtt_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	uintptr_t list = (uintptr_t) node->info_ent->data;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -439,7 +490,7 @@ static int i915_gem_gtt_info(struct seq_file *m, void *data)
 
 	total_obj_size = total_gtt_size = count = 0;
 	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
-		if (list == PINNED_LIST && obj->pin_count == 0)
+		if (list == PINNED_LIST && !i915_gem_obj_is_pinned(obj))
 			continue;
 
 		seq_puts(m, "   ");
@@ -460,12 +511,12 @@ static int i915_gem_gtt_info(struct seq_file *m, void *data)
 
 static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	unsigned long flags;
 	struct intel_crtc *crtc;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, base.head) {
+	for_each_intel_crtc(dev, crtc) {
 		const char pipe = pipe_name(crtc->pipe);
 		const char plane = plane_name(crtc->plane);
 		struct intel_unpin_work *work;
@@ -510,10 +561,10 @@ static int i915_gem_pageflip_info(struct seq_file *m, void *data)
 
 static int i915_gem_request_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
 	struct drm_i915_gem_request *gem_request;
 	int ret, count, i;
 
@@ -545,7 +596,7 @@ static int i915_gem_request_info(struct seq_file *m, void *data)
 }
 
 static void i915_ring_seqno_info(struct seq_file *m,
-				 struct intel_ring_buffer *ring)
+				 struct intel_engine_cs *ring)
 {
 	if (ring->get_seqno) {
 		seq_printf(m, "Current sequence (%s): %u\n",
@@ -555,19 +606,21 @@ static void i915_ring_seqno_info(struct seq_file *m,
 
 static int i915_gem_seqno_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
 	int ret, i;
 
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	for_each_ring(ring, dev_priv, i)
 		i915_ring_seqno_info(m, ring);
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -576,21 +629,61 @@ static int i915_gem_seqno_info(struct seq_file *m, void *data)
 
 static int i915_interrupt_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
 	int ret, i, pipe;
 
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
-	if (INTEL_INFO(dev)->gen >= 8) {
+	if (IS_CHERRYVIEW(dev)) {
 		int i;
 		seq_printf(m, "Master Interrupt Control:\t%08x\n",
 			   I915_READ(GEN8_MASTER_IRQ));
 
+		seq_printf(m, "Display IER:\t%08x\n",
+			   I915_READ(VLV_IER));
+		seq_printf(m, "Display IIR:\t%08x\n",
+			   I915_READ(VLV_IIR));
+		seq_printf(m, "Display IIR_RW:\t%08x\n",
+			   I915_READ(VLV_IIR_RW));
+		seq_printf(m, "Display IMR:\t%08x\n",
+			   I915_READ(VLV_IMR));
+		for_each_pipe(pipe)
+			seq_printf(m, "Pipe %c stat:\t%08x\n",
+				   pipe_name(pipe),
+				   I915_READ(PIPESTAT(pipe)));
+
+		seq_printf(m, "Port hotplug:\t%08x\n",
+			   I915_READ(PORT_HOTPLUG_EN));
+		seq_printf(m, "DPFLIPSTAT:\t%08x\n",
+			   I915_READ(VLV_DPFLIPSTAT));
+		seq_printf(m, "DPINVGTT:\t%08x\n",
+			   I915_READ(DPINVGTT));
+
+		for (i = 0; i < 4; i++) {
+			seq_printf(m, "GT Interrupt IMR %d:\t%08x\n",
+				   i, I915_READ(GEN8_GT_IMR(i)));
+			seq_printf(m, "GT Interrupt IIR %d:\t%08x\n",
+				   i, I915_READ(GEN8_GT_IIR(i)));
+			seq_printf(m, "GT Interrupt IER %d:\t%08x\n",
+				   i, I915_READ(GEN8_GT_IER(i)));
+		}
+
+		seq_printf(m, "PCU interrupt mask:\t%08x\n",
+			   I915_READ(GEN8_PCU_IMR));
+		seq_printf(m, "PCU interrupt identity:\t%08x\n",
+			   I915_READ(GEN8_PCU_IIR));
+		seq_printf(m, "PCU interrupt enable:\t%08x\n",
+			   I915_READ(GEN8_PCU_IER));
+	} else if (INTEL_INFO(dev)->gen >= 8) {
+		seq_printf(m, "Master Interrupt Control:\t%08x\n",
+			   I915_READ(GEN8_MASTER_IRQ));
+
 		for (i = 0; i < 4; i++) {
 			seq_printf(m, "GT Interrupt IMR %d:\t%08x\n",
 				   i, I915_READ(GEN8_GT_IMR(i)));
@@ -600,16 +693,16 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 				   i, I915_READ(GEN8_GT_IER(i)));
 		}
 
-		for_each_pipe(i) {
+		for_each_pipe(pipe) {
 			seq_printf(m, "Pipe %c IMR:\t%08x\n",
-				   pipe_name(i),
-				   I915_READ(GEN8_DE_PIPE_IMR(i)));
+				   pipe_name(pipe),
+				   I915_READ(GEN8_DE_PIPE_IMR(pipe)));
 			seq_printf(m, "Pipe %c IIR:\t%08x\n",
-				   pipe_name(i),
-				   I915_READ(GEN8_DE_PIPE_IIR(i)));
+				   pipe_name(pipe),
+				   I915_READ(GEN8_DE_PIPE_IIR(pipe)));
 			seq_printf(m, "Pipe %c IER:\t%08x\n",
-				   pipe_name(i),
-				   I915_READ(GEN8_DE_PIPE_IER(i)));
+				   pipe_name(pipe),
+				   I915_READ(GEN8_DE_PIPE_IER(pipe)));
 		}
 
 		seq_printf(m, "Display Engine port interrupt mask:\t%08x\n",
@@ -701,8 +794,6 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 		seq_printf(m, "Graphics Interrupt mask:		%08x\n",
 			   I915_READ(GTIMR));
 	}
-	seq_printf(m, "Interrupts received: %d\n",
-		   atomic_read(&dev_priv->irq_received));
 	for_each_ring(ring, dev_priv, i) {
 		if (INTEL_INFO(dev)->gen >= 6) {
 			seq_printf(m,
@@ -711,6 +802,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 		}
 		i915_ring_seqno_info(m, ring);
 	}
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -718,9 +810,9 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 
 static int i915_gem_fence_regs_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i, ret;
 
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
@@ -747,10 +839,10 @@ static int i915_gem_fence_regs_info(struct seq_file *m, void *data)
 
 static int i915_hws_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
 	const u32 *hws;
 	int i;
 
@@ -860,7 +952,7 @@ static int
 i915_next_seqno_get(void *data, u64 *val)
 {
 	struct drm_device *dev = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
@@ -895,18 +987,20 @@ DEFINE_SIMPLE_ATTRIBUTE(i915_next_seqno_fops,
 
 static int i915_rstdby_delays(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u16 crstanddelay;
 	int ret;
 
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	crstanddelay = I915_READ16(CRSTANDVID);
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	seq_printf(m, "w/ctx: %d, w/o ctx: %d\n", (crstanddelay >> 8) & 0x3f, (crstanddelay & 0x3f));
@@ -914,12 +1008,14 @@ static int i915_rstdby_delays(struct seq_file *m, void *unused)
 	return 0;
 }
 
-static int i915_cur_delayinfo(struct seq_file *m, void *unused)
+static int i915_frequency_info(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	int ret;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret = 0;
+
+	intel_runtime_pm_get(dev_priv);
 
 	flush_delayed_work(&dev_priv->rps.delayed_resume_work);
 
@@ -937,6 +1033,7 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 		u32 gt_perf_status = I915_READ(GEN6_GT_PERF_STATUS);
 		u32 rp_state_limits = I915_READ(GEN6_RP_STATE_LIMITS);
 		u32 rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
+		u32 rpmodectl, rpinclimit, rpdeclimit;
 		u32 rpstat, cagf, reqf;
 		u32 rpupei, rpcurup, rpprevup;
 		u32 rpdownei, rpcurdown, rpprevdown;
@@ -945,9 +1042,9 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 		/* RPSTAT1 is in the GT power well */
 		ret = mutex_lock_interruptible(&dev->struct_mutex);
 		if (ret)
-			return ret;
+			goto out;
 
-		gen6_gt_force_wake_get(dev_priv);
+		gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 		reqf = I915_READ(GEN6_RPNSWREQ);
 		reqf &= ~GEN6_TURBO_DISABLE;
@@ -957,6 +1054,10 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 			reqf >>= 25;
 		reqf *= GT_FREQUENCY_MULTIPLIER;
 
+		rpmodectl = I915_READ(GEN6_RP_CONTROL);
+		rpinclimit = I915_READ(GEN6_RP_UP_THRESHOLD);
+		rpdeclimit = I915_READ(GEN6_RP_DOWN_THRESHOLD);
+
 		rpstat = I915_READ(GEN6_RPSTAT1);
 		rpupei = I915_READ(GEN6_RP_CUR_UP_EI);
 		rpcurup = I915_READ(GEN6_RP_CUR_UP);
@@ -970,17 +1071,26 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 			cagf = (rpstat & GEN6_CAGF_MASK) >> GEN6_CAGF_SHIFT;
 		cagf *= GT_FREQUENCY_MULTIPLIER;
 
-		gen6_gt_force_wake_put(dev_priv);
+		gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 		mutex_unlock(&dev->struct_mutex);
 
+		seq_printf(m, "PM IER=0x%08x IMR=0x%08x ISR=0x%08x IIR=0x%08x, MASK=0x%08x\n",
+			   I915_READ(GEN6_PMIER),
+			   I915_READ(GEN6_PMIMR),
+			   I915_READ(GEN6_PMISR),
+			   I915_READ(GEN6_PMIIR),
+			   I915_READ(GEN6_PMINTRMSK));
 		seq_printf(m, "GT_PERF_STATUS: 0x%08x\n", gt_perf_status);
-		seq_printf(m, "RPSTAT1: 0x%08x\n", rpstat);
 		seq_printf(m, "Render p-state ratio: %d\n",
 			   (gt_perf_status & 0xff00) >> 8);
 		seq_printf(m, "Render p-state VID: %d\n",
 			   gt_perf_status & 0xff);
 		seq_printf(m, "Render p-state limit: %d\n",
 			   rp_state_limits & 0xff);
+		seq_printf(m, "RPSTAT1: 0x%08x\n", rpstat);
+		seq_printf(m, "RPMODECTL: 0x%08x\n", rpmodectl);
+		seq_printf(m, "RPINCLIMIT: 0x%08x\n", rpinclimit);
+		seq_printf(m, "RPDECLIMIT: 0x%08x\n", rpdeclimit);
 		seq_printf(m, "RPNSWREQ: %dMHz\n", reqf);
 		seq_printf(m, "CAGF: %dMHz\n", cagf);
 		seq_printf(m, "RP CUR UP EI: %dus\n", rpupei &
@@ -1009,7 +1119,7 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 			   max_freq * GT_FREQUENCY_MULTIPLIER);
 
 		seq_printf(m, "Max overclocked frequency: %dMHz\n",
-			   dev_priv->rps.hw_max * GT_FREQUENCY_MULTIPLIER);
+			   dev_priv->rps.max_freq * GT_FREQUENCY_MULTIPLIER);
 	} else if (IS_VALLEYVIEW(dev)) {
 		u32 freq_sts, val;
 
@@ -1018,36 +1128,38 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 		seq_printf(m, "PUNIT_REG_GPU_FREQ_STS: 0x%08x\n", freq_sts);
 		seq_printf(m, "DDR freq: %d MHz\n", dev_priv->mem_freq);
 
-		val = vlv_punit_read(dev_priv, PUNIT_FUSE_BUS1);
+		val = valleyview_rps_max_freq(dev_priv);
 		seq_printf(m, "max GPU freq: %d MHz\n",
-			   vlv_gpu_freq(dev_priv->mem_freq, val));
+			   vlv_gpu_freq(dev_priv, val));
 
-		val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_LFM);
+		val = valleyview_rps_min_freq(dev_priv);
 		seq_printf(m, "min GPU freq: %d MHz\n",
-			   vlv_gpu_freq(dev_priv->mem_freq, val));
+			   vlv_gpu_freq(dev_priv, val));
 
 		seq_printf(m, "current GPU freq: %d MHz\n",
-			   vlv_gpu_freq(dev_priv->mem_freq,
-					(freq_sts >> 8) & 0xff));
+			   vlv_gpu_freq(dev_priv, (freq_sts >> 8) & 0xff));
 		mutex_unlock(&dev_priv->rps.hw_lock);
 	} else {
 		seq_puts(m, "no P-state info available\n");
 	}
 
-	return 0;
+out:
+	intel_runtime_pm_put(dev_priv);
+	return ret;
 }
 
 static int i915_delayfreq_table(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 delayfreq;
 	int ret, i;
 
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	for (i = 0; i < 16; i++) {
 		delayfreq = I915_READ(PXVFREQ_BASE + i * 4);
@@ -1055,6 +1167,8 @@ static int i915_delayfreq_table(struct seq_file *m, void *unused)
 			   (delayfreq & PXVFREQ_PX_MASK) >> PXVFREQ_PX_SHIFT);
 	}
 
+	intel_runtime_pm_put(dev_priv);
+
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -1067,21 +1181,23 @@ static inline int MAP_TO_MV(int map)
 
 static int i915_inttoext_table(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 inttoext;
 	int ret, i;
 
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	for (i = 1; i <= 32; i++) {
 		inttoext = I915_READ(INTTOEXT_BASE_ILK + i * 4);
 		seq_printf(m, "INTTOEXT%02d: 0x%08x\n", i, inttoext);
 	}
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -1089,9 +1205,9 @@ static int i915_inttoext_table(struct seq_file *m, void *unused)
 
 static int ironlake_drpc_info(struct seq_file *m)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 rgvmodectl, rstdbyctl;
 	u16 crstandvid;
 	int ret;
@@ -1099,11 +1215,13 @@ static int ironlake_drpc_info(struct seq_file *m)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	rgvmodectl = I915_READ(MEMMODECTL);
 	rstdbyctl = I915_READ(RSTDBYCTL);
 	crstandvid = I915_READ16(CRSTANDVID);
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	seq_printf(m, "HD boost: %s\n", (rgvmodectl & MEMMODE_BOOST_EN) ?
@@ -1154,10 +1272,63 @@ static int ironlake_drpc_info(struct seq_file *m)
 	return 0;
 }
 
+static int vlv_drpc_info(struct seq_file *m)
+{
+
+	struct drm_info_node *node = m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 rpmodectl1, rcctl1;
+	unsigned fw_rendercount = 0, fw_mediacount = 0;
+
+	intel_runtime_pm_get(dev_priv);
+
+	rpmodectl1 = I915_READ(GEN6_RP_CONTROL);
+	rcctl1 = I915_READ(GEN6_RC_CONTROL);
+
+	intel_runtime_pm_put(dev_priv);
+
+	seq_printf(m, "Video Turbo Mode: %s\n",
+		   yesno(rpmodectl1 & GEN6_RP_MEDIA_TURBO));
+	seq_printf(m, "Turbo enabled: %s\n",
+		   yesno(rpmodectl1 & GEN6_RP_ENABLE));
+	seq_printf(m, "HW control enabled: %s\n",
+		   yesno(rpmodectl1 & GEN6_RP_ENABLE));
+	seq_printf(m, "SW control enabled: %s\n",
+		   yesno((rpmodectl1 & GEN6_RP_MEDIA_MODE_MASK) ==
+			  GEN6_RP_MEDIA_SW_MODE));
+	seq_printf(m, "RC6 Enabled: %s\n",
+		   yesno(rcctl1 & (GEN7_RC_CTL_TO_MODE |
+					GEN6_RC_CTL_EI_MODE(1))));
+	seq_printf(m, "Render Power Well: %s\n",
+			(I915_READ(VLV_GTLC_PW_STATUS) &
+				VLV_GTLC_PW_RENDER_STATUS_MASK) ? "Up" : "Down");
+	seq_printf(m, "Media Power Well: %s\n",
+			(I915_READ(VLV_GTLC_PW_STATUS) &
+				VLV_GTLC_PW_MEDIA_STATUS_MASK) ? "Up" : "Down");
+
+	seq_printf(m, "Render RC6 residency since boot: %u\n",
+		   I915_READ(VLV_GT_RENDER_RC6));
+	seq_printf(m, "Media RC6 residency since boot: %u\n",
+		   I915_READ(VLV_GT_MEDIA_RC6));
+
+	spin_lock_irq(&dev_priv->uncore.lock);
+	fw_rendercount = dev_priv->uncore.fw_rendercount;
+	fw_mediacount = dev_priv->uncore.fw_mediacount;
+	spin_unlock_irq(&dev_priv->uncore.lock);
+
+	seq_printf(m, "Forcewake Render Count = %u\n", fw_rendercount);
+	seq_printf(m, "Forcewake Media Count = %u\n", fw_mediacount);
+
+
+	return 0;
+}
+
+
 static int gen6_drpc_info(struct seq_file *m)
 {
 
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 rpmodectl1, gt_core_status, rcctl1, rc6vids = 0;
@@ -1167,6 +1338,7 @@ static int gen6_drpc_info(struct seq_file *m)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	spin_lock_irq(&dev_priv->uncore.lock);
 	forcewake_count = dev_priv->uncore.forcewake_count;
@@ -1192,6 +1364,8 @@ static int gen6_drpc_info(struct seq_file *m)
 	sandybridge_pcode_read(dev_priv, GEN6_PCODE_READ_RC6VIDS, &rc6vids);
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
+	intel_runtime_pm_put(dev_priv);
+
 	seq_printf(m, "Video Turbo Mode: %s\n",
 		   yesno(rpmodectl1 & GEN6_RP_MEDIA_TURBO));
 	seq_printf(m, "HW control enabled: %s\n",
@@ -1253,10 +1427,12 @@ static int gen6_drpc_info(struct seq_file *m)
 
 static int i915_drpc_info(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 
-	if (IS_GEN6(dev) || IS_GEN7(dev))
+	if (IS_VALLEYVIEW(dev))
+		return vlv_drpc_info(m);
+	else if (IS_GEN6(dev) || IS_GEN7(dev))
 		return gen6_drpc_info(m);
 	else
 		return ironlake_drpc_info(m);
@@ -1264,15 +1440,17 @@ static int i915_drpc_info(struct seq_file *m, void *unused)
 
 static int i915_fbc_status(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (!I915_HAS_FBC(dev)) {
+	if (!HAS_FBC(dev)) {
 		seq_puts(m, "FBC unsupported on this chipset\n");
 		return 0;
 	}
 
+	intel_runtime_pm_get(dev_priv);
+
 	if (intel_fbc_enabled(dev)) {
 		seq_puts(m, "FBC enabled\n");
 	} else {
@@ -1316,12 +1494,15 @@ static int i915_fbc_status(struct seq_file *m, void *unused)
 		}
 		seq_putc(m, '\n');
 	}
+
+	intel_runtime_pm_put(dev_priv);
+
 	return 0;
 }
 
 static int i915_ips_status(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
@@ -1330,21 +1511,27 @@ static int i915_ips_status(struct seq_file *m, void *unused)
 		return 0;
 	}
 
-	if (I915_READ(IPS_CTL) & IPS_ENABLE)
+	intel_runtime_pm_get(dev_priv);
+
+	if (IS_BROADWELL(dev) || I915_READ(IPS_CTL) & IPS_ENABLE)
 		seq_puts(m, "enabled\n");
 	else
 		seq_puts(m, "disabled\n");
 
+	intel_runtime_pm_put(dev_priv);
+
 	return 0;
 }
 
 static int i915_sr_status(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	bool sr_enabled = false;
 
+	intel_runtime_pm_get(dev_priv);
+
 	if (HAS_PCH_SPLIT(dev))
 		sr_enabled = I915_READ(WM1_LP_ILK) & WM1_LP_SR_EN;
 	else if (IS_CRESTLINE(dev) || IS_I945G(dev) || IS_I945GM(dev))
@@ -1354,6 +1541,8 @@ static int i915_sr_status(struct seq_file *m, void *unused)
 	else if (IS_PINEVIEW(dev))
 		sr_enabled = I915_READ(DSPFW3) & PINEVIEW_SELF_REFRESH_EN;
 
+	intel_runtime_pm_put(dev_priv);
+
 	seq_printf(m, "self-refresh: %s\n",
 		   sr_enabled ? "enabled" : "disabled");
 
@@ -1362,9 +1551,9 @@ static int i915_sr_status(struct seq_file *m, void *unused)
 
 static int i915_emon_status(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long temp, chipset, gfx;
 	int ret;
 
@@ -1390,10 +1579,10 @@ static int i915_emon_status(struct seq_file *m, void *unused)
 
 static int i915_ring_freq_table(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	int ret;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret = 0;
 	int gpu_freq, ia_freq;
 
 	if (!(IS_GEN6(dev) || IS_GEN7(dev))) {
@@ -1401,16 +1590,18 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused)
 		return 0;
 	}
 
+	intel_runtime_pm_get(dev_priv);
+
 	flush_delayed_work(&dev_priv->rps.delayed_resume_work);
 
 	ret = mutex_lock_interruptible(&dev_priv->rps.hw_lock);
 	if (ret)
-		return ret;
+		goto out;
 
 	seq_puts(m, "GPU freq (MHz)\tEffective CPU freq (MHz)\tEffective Ring freq (MHz)\n");
 
-	for (gpu_freq = dev_priv->rps.min_delay;
-	     gpu_freq <= dev_priv->rps.max_delay;
+	for (gpu_freq = dev_priv->rps.min_freq_softlimit;
+	     gpu_freq <= dev_priv->rps.max_freq_softlimit;
 	     gpu_freq++) {
 		ia_freq = gpu_freq;
 		sandybridge_pcode_read(dev_priv,
@@ -1424,21 +1615,25 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused)
 
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
-	return 0;
+out:
+	intel_runtime_pm_put(dev_priv);
+	return ret;
 }
 
 static int i915_gfxec(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	seq_printf(m, "GFXEC: %ld\n", (unsigned long)I915_READ(0x112f4));
+	intel_runtime_pm_put(dev_priv);
 
 	mutex_unlock(&dev->struct_mutex);
 
@@ -1447,9 +1642,9 @@ static int i915_gfxec(struct seq_file *m, void *unused)
 
 static int i915_opregion(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_opregion *opregion = &dev_priv->opregion;
 	void *data = kmalloc(OPREGION_SIZE, GFP_KERNEL);
 	int ret;
@@ -1475,7 +1670,7 @@ out:
 
 static int i915_gem_framebuffer_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct intel_fbdev *ifbdev = NULL;
 	struct intel_framebuffer *fb;
@@ -1521,11 +1716,11 @@ static int i915_gem_framebuffer_info(struct seq_file *m, void *data)
 
 static int i915_context_status(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
-	struct i915_hw_context *ctx;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	struct intel_context *ctx;
 	int ret, i;
 
 	ret = mutex_lock_interruptible(&dev->mode_config.mutex);
@@ -1545,6 +1740,9 @@ static int i915_context_status(struct seq_file *m, void *unused)
 	}
 
 	list_for_each_entry(ctx, &dev_priv->context_list, link) {
+		if (ctx->obj == NULL)
+			continue;
+
 		seq_puts(m, "HW context ");
 		describe_ctx(m, ctx);
 		for_each_ring(ring, dev_priv, i)
@@ -1562,16 +1760,24 @@ static int i915_context_status(struct seq_file *m, void *unused)
 
 static int i915_gen6_forcewake_count_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned forcewake_count;
+	unsigned forcewake_count = 0, fw_rendercount = 0, fw_mediacount = 0;
 
 	spin_lock_irq(&dev_priv->uncore.lock);
-	forcewake_count = dev_priv->uncore.forcewake_count;
+	if (IS_VALLEYVIEW(dev)) {
+		fw_rendercount = dev_priv->uncore.fw_rendercount;
+		fw_mediacount = dev_priv->uncore.fw_mediacount;
+	} else
+		forcewake_count = dev_priv->uncore.forcewake_count;
 	spin_unlock_irq(&dev_priv->uncore.lock);
 
-	seq_printf(m, "forcewake count = %u\n", forcewake_count);
+	if (IS_VALLEYVIEW(dev)) {
+		seq_printf(m, "fw_rendercount = %u\n", fw_rendercount);
+		seq_printf(m, "fw_mediacount = %u\n", fw_mediacount);
+	} else
+		seq_printf(m, "forcewake count = %u\n", forcewake_count);
 
 	return 0;
 }
@@ -1602,7 +1808,7 @@ static const char *swizzle_string(unsigned swizzle)
 
 static int i915_swizzle_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
@@ -1610,6 +1816,7 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	seq_printf(m, "bit6 swizzle for X-tiling = %s\n",
 		   swizzle_string(dev_priv->mm.bit_6_swizzle_x));
@@ -1641,15 +1848,31 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
 		seq_printf(m, "DISP_ARB_CTL = 0x%08x\n",
 			   I915_READ(DISP_ARB_CTL));
 	}
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
 }
 
+static int per_file_ctx(int id, void *ptr, void *data)
+{
+	struct intel_context *ctx = ptr;
+	struct seq_file *m = data;
+	struct i915_hw_ppgtt *ppgtt = ctx_to_ppgtt(ctx);
+
+	if (i915_gem_context_is_default(ctx))
+		seq_puts(m, "  default context:\n");
+	else
+		seq_printf(m, "  context %d:\n", ctx->id);
+	ppgtt->debug_dump(ppgtt, m);
+
+	return 0;
+}
+
 static void gen8_ppgtt_info(struct seq_file *m, struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
 	int unused, i;
 
@@ -1657,7 +1880,7 @@ static void gen8_ppgtt_info(struct seq_file *m, struct drm_device *dev)
 		return;
 
 	seq_printf(m, "Page directories: %d\n", ppgtt->num_pd_pages);
-	seq_printf(m, "Page tables: %d\n", ppgtt->num_pt_pages);
+	seq_printf(m, "Page tables: %d\n", ppgtt->num_pd_entries);
 	for_each_ring(ring, dev_priv, unused) {
 		seq_printf(m, "%s\n", ring->name);
 		for (i = 0; i < 4; i++) {
@@ -1665,8 +1888,7 @@ static void gen8_ppgtt_info(struct seq_file *m, struct drm_device *dev)
 			u64 pdp = I915_READ(ring->mmio_base + offset + 4);
 			pdp <<= 32;
 			pdp |= I915_READ(ring->mmio_base + offset);
-			for (i = 0; i < 4; i++)
-				seq_printf(m, "\tPDP%d 0x%016llx\n", i, pdp);
+			seq_printf(m, "\tPDP%d 0x%016llx\n", i, pdp);
 		}
 	}
 }
@@ -1674,7 +1896,8 @@ static void gen8_ppgtt_info(struct seq_file *m, struct drm_device *dev)
 static void gen6_ppgtt_info(struct seq_file *m, struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
+	struct drm_file *file;
 	int i;
 
 	if (INTEL_INFO(dev)->gen == 6)
@@ -1693,79 +1916,46 @@ static void gen6_ppgtt_info(struct seq_file *m, struct drm_device *dev)
 
 		seq_puts(m, "aliasing PPGTT:\n");
 		seq_printf(m, "pd gtt offset: 0x%08x\n", ppgtt->pd_offset);
+
+		ppgtt->debug_dump(ppgtt, m);
+	} else
+		return;
+
+	list_for_each_entry_reverse(file, &dev->filelist, lhead) {
+		struct drm_i915_file_private *file_priv = file->driver_priv;
+
+		seq_printf(m, "proc: %s\n",
+			   get_pid_task(file->pid, PIDTYPE_PID)->comm);
+		idr_for_each(&file_priv->context_idr, per_file_ctx, m);
 	}
 	seq_printf(m, "ECOCHK: 0x%08x\n", I915_READ(GAM_ECOCHK));
 }
 
 static int i915_ppgtt_info(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	int ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	if (INTEL_INFO(dev)->gen >= 8)
 		gen8_ppgtt_info(m, dev);
 	else if (INTEL_INFO(dev)->gen >= 6)
 		gen6_ppgtt_info(m, dev);
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
 }
 
-static int i915_dpio_info(struct seq_file *m, void *data)
-{
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
-	struct drm_device *dev = node->minor->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int ret;
-
-
-	if (!IS_VALLEYVIEW(dev)) {
-		seq_puts(m, "unsupported\n");
-		return 0;
-	}
-
-	ret = mutex_lock_interruptible(&dev_priv->dpio_lock);
-	if (ret)
-		return ret;
-
-	seq_printf(m, "DPIO_CTL: 0x%08x\n", I915_READ(DPIO_CTL));
-
-	seq_printf(m, "DPIO_DIV_A: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_DIV_A));
-	seq_printf(m, "DPIO_DIV_B: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_DIV_B));
-
-	seq_printf(m, "DPIO_REFSFR_A: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_REFSFR_A));
-	seq_printf(m, "DPIO_REFSFR_B: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_REFSFR_B));
-
-	seq_printf(m, "DPIO_CORE_CLK_A: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_CORE_CLK_A));
-	seq_printf(m, "DPIO_CORE_CLK_B: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_CORE_CLK_B));
-
-	seq_printf(m, "DPIO_LPF_COEFF_A: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_LPF_COEFF_A));
-	seq_printf(m, "DPIO_LPF_COEFF_B: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_LPF_COEFF_B));
-
-	seq_printf(m, "DPIO_FASTCLK_DISABLE: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, DPIO_FASTCLK_DISABLE));
-
-	mutex_unlock(&dev_priv->dpio_lock);
-
-	return 0;
-}
-
 static int i915_llc(struct seq_file *m, void *data)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
@@ -1784,6 +1974,8 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 	u32 psrperf = 0;
 	bool enabled = false;
 
+	intel_runtime_pm_get(dev_priv);
+
 	seq_printf(m, "Sink_Support: %s\n", yesno(dev_priv->psr.sink_support));
 	seq_printf(m, "Source_OK: %s\n", yesno(dev_priv->psr.source_ok));
 
@@ -1796,9 +1988,51 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 			EDP_PSR_PERF_CNT_MASK;
 	seq_printf(m, "Performance_Counter: %u\n", psrperf);
 
+	intel_runtime_pm_put(dev_priv);
 	return 0;
 }
 
+static int i915_sink_crc(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct intel_encoder *encoder;
+	struct intel_connector *connector;
+	struct intel_dp *intel_dp = NULL;
+	int ret;
+	u8 crc[6];
+
+	drm_modeset_lock_all(dev);
+	list_for_each_entry(connector, &dev->mode_config.connector_list,
+			    base.head) {
+
+		if (connector->base.dpms != DRM_MODE_DPMS_ON)
+			continue;
+
+		if (!connector->base.encoder)
+			continue;
+
+		encoder = to_intel_encoder(connector->base.encoder);
+		if (encoder->type != INTEL_OUTPUT_EDP)
+			continue;
+
+		intel_dp = enc_to_intel_dp(&encoder->base);
+
+		ret = intel_dp_sink_crc(intel_dp, crc);
+		if (ret)
+			goto out;
+
+		seq_printf(m, "%02x%02x%02x%02x%02x%02x\n",
+			   crc[0], crc[1], crc[2],
+			   crc[3], crc[4], crc[5]);
+		goto out;
+	}
+	ret = -ENODEV;
+out:
+	drm_modeset_unlock_all(dev);
+	return ret;
+}
+
 static int i915_energy_uJ(struct seq_file *m, void *data)
 {
 	struct drm_info_node *node = m->private;
@@ -1810,12 +2044,16 @@ static int i915_energy_uJ(struct seq_file *m, void *data)
 	if (INTEL_INFO(dev)->gen < 6)
 		return -ENODEV;
 
+	intel_runtime_pm_get(dev_priv);
+
 	rdmsrl(MSR_RAPL_POWER_UNIT, power);
 	power = (power & 0x1f00) >> 8;
 	units = 1000000 / (1 << power); /* convert to uJ */
 	power = I915_READ(MCH_SECP_NRG_STTS);
 	power *= units;
 
+	intel_runtime_pm_put(dev_priv);
+
 	seq_printf(m, "%llu", (long long unsigned)power);
 
 	return 0;
@@ -1823,24 +2061,318 @@ static int i915_energy_uJ(struct seq_file *m, void *data)
 
 static int i915_pc8_status(struct seq_file *m, void *unused)
 {
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_info_node *node = m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (!IS_HASWELL(dev)) {
+	if (!IS_HASWELL(dev) && !IS_BROADWELL(dev)) {
 		seq_puts(m, "not supported\n");
 		return 0;
 	}
 
-	mutex_lock(&dev_priv->pc8.lock);
-	seq_printf(m, "Requirements met: %s\n",
-		   yesno(dev_priv->pc8.requirements_met));
-	seq_printf(m, "GPU idle: %s\n", yesno(dev_priv->pc8.gpu_idle));
-	seq_printf(m, "Disable count: %d\n", dev_priv->pc8.disable_count);
+	seq_printf(m, "GPU idle: %s\n", yesno(!dev_priv->mm.busy));
 	seq_printf(m, "IRQs disabled: %s\n",
-		   yesno(dev_priv->pc8.irqs_disabled));
-	seq_printf(m, "Enabled: %s\n", yesno(dev_priv->pc8.enabled));
-	mutex_unlock(&dev_priv->pc8.lock);
+		   yesno(dev_priv->pm.irqs_disabled));
+
+	return 0;
+}
+
+static const char *power_domain_str(enum intel_display_power_domain domain)
+{
+	switch (domain) {
+	case POWER_DOMAIN_PIPE_A:
+		return "PIPE_A";
+	case POWER_DOMAIN_PIPE_B:
+		return "PIPE_B";
+	case POWER_DOMAIN_PIPE_C:
+		return "PIPE_C";
+	case POWER_DOMAIN_PIPE_A_PANEL_FITTER:
+		return "PIPE_A_PANEL_FITTER";
+	case POWER_DOMAIN_PIPE_B_PANEL_FITTER:
+		return "PIPE_B_PANEL_FITTER";
+	case POWER_DOMAIN_PIPE_C_PANEL_FITTER:
+		return "PIPE_C_PANEL_FITTER";
+	case POWER_DOMAIN_TRANSCODER_A:
+		return "TRANSCODER_A";
+	case POWER_DOMAIN_TRANSCODER_B:
+		return "TRANSCODER_B";
+	case POWER_DOMAIN_TRANSCODER_C:
+		return "TRANSCODER_C";
+	case POWER_DOMAIN_TRANSCODER_EDP:
+		return "TRANSCODER_EDP";
+	case POWER_DOMAIN_PORT_DDI_A_2_LANES:
+		return "PORT_DDI_A_2_LANES";
+	case POWER_DOMAIN_PORT_DDI_A_4_LANES:
+		return "PORT_DDI_A_4_LANES";
+	case POWER_DOMAIN_PORT_DDI_B_2_LANES:
+		return "PORT_DDI_B_2_LANES";
+	case POWER_DOMAIN_PORT_DDI_B_4_LANES:
+		return "PORT_DDI_B_4_LANES";
+	case POWER_DOMAIN_PORT_DDI_C_2_LANES:
+		return "PORT_DDI_C_2_LANES";
+	case POWER_DOMAIN_PORT_DDI_C_4_LANES:
+		return "PORT_DDI_C_4_LANES";
+	case POWER_DOMAIN_PORT_DDI_D_2_LANES:
+		return "PORT_DDI_D_2_LANES";
+	case POWER_DOMAIN_PORT_DDI_D_4_LANES:
+		return "PORT_DDI_D_4_LANES";
+	case POWER_DOMAIN_PORT_DSI:
+		return "PORT_DSI";
+	case POWER_DOMAIN_PORT_CRT:
+		return "PORT_CRT";
+	case POWER_DOMAIN_PORT_OTHER:
+		return "PORT_OTHER";
+	case POWER_DOMAIN_VGA:
+		return "VGA";
+	case POWER_DOMAIN_AUDIO:
+		return "AUDIO";
+	case POWER_DOMAIN_INIT:
+		return "INIT";
+	default:
+		WARN_ON(1);
+		return "?";
+	}
+}
+
+static int i915_power_domain_info(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_power_domains *power_domains = &dev_priv->power_domains;
+	int i;
+
+	mutex_lock(&power_domains->lock);
+
+	seq_printf(m, "%-25s %s\n", "Power well/domain", "Use count");
+	for (i = 0; i < power_domains->power_well_count; i++) {
+		struct i915_power_well *power_well;
+		enum intel_display_power_domain power_domain;
+
+		power_well = &power_domains->power_wells[i];
+		seq_printf(m, "%-25s %d\n", power_well->name,
+			   power_well->count);
+
+		for (power_domain = 0; power_domain < POWER_DOMAIN_NUM;
+		     power_domain++) {
+			if (!(BIT(power_domain) & power_well->domains))
+				continue;
+
+			seq_printf(m, "  %-23s %d\n",
+				 power_domain_str(power_domain),
+				 power_domains->domain_use_count[power_domain]);
+		}
+	}
+
+	mutex_unlock(&power_domains->lock);
+
+	return 0;
+}
+
+static void intel_seq_print_mode(struct seq_file *m, int tabs,
+				 struct drm_display_mode *mode)
+{
+	int i;
+
+	for (i = 0; i < tabs; i++)
+		seq_putc(m, '\t');
+
+	seq_printf(m, "id %d:\"%s\" freq %d clock %d hdisp %d hss %d hse %d htot %d vdisp %d vss %d vse %d vtot %d type 0x%x flags 0x%x\n",
+		   mode->base.id, mode->name,
+		   mode->vrefresh, mode->clock,
+		   mode->hdisplay, mode->hsync_start,
+		   mode->hsync_end, mode->htotal,
+		   mode->vdisplay, mode->vsync_start,
+		   mode->vsync_end, mode->vtotal,
+		   mode->type, mode->flags);
+}
+
+static void intel_encoder_info(struct seq_file *m,
+			       struct intel_crtc *intel_crtc,
+			       struct intel_encoder *intel_encoder)
+{
+	struct drm_info_node *node = m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_crtc *crtc = &intel_crtc->base;
+	struct intel_connector *intel_connector;
+	struct drm_encoder *encoder;
+
+	encoder = &intel_encoder->base;
+	seq_printf(m, "\tencoder %d: type: %s, connectors:\n",
+		   encoder->base.id, encoder->name);
+	for_each_connector_on_encoder(dev, encoder, intel_connector) {
+		struct drm_connector *connector = &intel_connector->base;
+		seq_printf(m, "\t\tconnector %d: type: %s, status: %s",
+			   connector->base.id,
+			   connector->name,
+			   drm_get_connector_status_name(connector->status));
+		if (connector->status == connector_status_connected) {
+			struct drm_display_mode *mode = &crtc->mode;
+			seq_printf(m, ", mode:\n");
+			intel_seq_print_mode(m, 2, mode);
+		} else {
+			seq_putc(m, '\n');
+		}
+	}
+}
+
+static void intel_crtc_info(struct seq_file *m, struct intel_crtc *intel_crtc)
+{
+	struct drm_info_node *node = m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_crtc *crtc = &intel_crtc->base;
+	struct intel_encoder *intel_encoder;
+
+	seq_printf(m, "\tfb: %d, pos: %dx%d, size: %dx%d\n",
+		   crtc->primary->fb->base.id, crtc->x, crtc->y,
+		   crtc->primary->fb->width, crtc->primary->fb->height);
+	for_each_encoder_on_crtc(dev, crtc, intel_encoder)
+		intel_encoder_info(m, intel_crtc, intel_encoder);
+}
+
+static void intel_panel_info(struct seq_file *m, struct intel_panel *panel)
+{
+	struct drm_display_mode *mode = panel->fixed_mode;
+
+	seq_printf(m, "\tfixed mode:\n");
+	intel_seq_print_mode(m, 2, mode);
+}
+
+static void intel_dp_info(struct seq_file *m,
+			  struct intel_connector *intel_connector)
+{
+	struct intel_encoder *intel_encoder = intel_connector->encoder;
+	struct intel_dp *intel_dp = enc_to_intel_dp(&intel_encoder->base);
+
+	seq_printf(m, "\tDPCD rev: %x\n", intel_dp->dpcd[DP_DPCD_REV]);
+	seq_printf(m, "\taudio support: %s\n", intel_dp->has_audio ? "yes" :
+		   "no");
+	if (intel_encoder->type == INTEL_OUTPUT_EDP)
+		intel_panel_info(m, &intel_connector->panel);
+}
+
+static void intel_hdmi_info(struct seq_file *m,
+			    struct intel_connector *intel_connector)
+{
+	struct intel_encoder *intel_encoder = intel_connector->encoder;
+	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&intel_encoder->base);
+
+	seq_printf(m, "\taudio support: %s\n", intel_hdmi->has_audio ? "yes" :
+		   "no");
+}
+
+static void intel_lvds_info(struct seq_file *m,
+			    struct intel_connector *intel_connector)
+{
+	intel_panel_info(m, &intel_connector->panel);
+}
+
+static void intel_connector_info(struct seq_file *m,
+				 struct drm_connector *connector)
+{
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_encoder *intel_encoder = intel_connector->encoder;
+	struct drm_display_mode *mode;
+
+	seq_printf(m, "connector %d: type %s, status: %s\n",
+		   connector->base.id, connector->name,
+		   drm_get_connector_status_name(connector->status));
+	if (connector->status == connector_status_connected) {
+		seq_printf(m, "\tname: %s\n", connector->display_info.name);
+		seq_printf(m, "\tphysical dimensions: %dx%dmm\n",
+			   connector->display_info.width_mm,
+			   connector->display_info.height_mm);
+		seq_printf(m, "\tsubpixel order: %s\n",
+			   drm_get_subpixel_order_name(connector->display_info.subpixel_order));
+		seq_printf(m, "\tCEA rev: %d\n",
+			   connector->display_info.cea_rev);
+	}
+	if (intel_encoder->type == INTEL_OUTPUT_DISPLAYPORT ||
+	    intel_encoder->type == INTEL_OUTPUT_EDP)
+		intel_dp_info(m, intel_connector);
+	else if (intel_encoder->type == INTEL_OUTPUT_HDMI)
+		intel_hdmi_info(m, intel_connector);
+	else if (intel_encoder->type == INTEL_OUTPUT_LVDS)
+		intel_lvds_info(m, intel_connector);
+
+	seq_printf(m, "\tmodes:\n");
+	list_for_each_entry(mode, &connector->modes, head)
+		intel_seq_print_mode(m, 2, mode);
+}
+
+static bool cursor_active(struct drm_device *dev, int pipe)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 state;
+
+	if (IS_845G(dev) || IS_I865G(dev))
+		state = I915_READ(_CURACNTR) & CURSOR_ENABLE;
+	else
+		state = I915_READ(CURCNTR(pipe)) & CURSOR_MODE;
+
+	return state;
+}
+
+static bool cursor_position(struct drm_device *dev, int pipe, int *x, int *y)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 pos;
+
+	pos = I915_READ(CURPOS(pipe));
+
+	*x = (pos >> CURSOR_X_SHIFT) & CURSOR_POS_MASK;
+	if (pos & (CURSOR_POS_SIGN << CURSOR_X_SHIFT))
+		*x = -*x;
+
+	*y = (pos >> CURSOR_Y_SHIFT) & CURSOR_POS_MASK;
+	if (pos & (CURSOR_POS_SIGN << CURSOR_Y_SHIFT))
+		*y = -*y;
+
+	return cursor_active(dev, pipe);
+}
+
+static int i915_display_info(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *crtc;
+	struct drm_connector *connector;
+
+	intel_runtime_pm_get(dev_priv);
+	drm_modeset_lock_all(dev);
+	seq_printf(m, "CRTC info\n");
+	seq_printf(m, "---------\n");
+	for_each_intel_crtc(dev, crtc) {
+		bool active;
+		int x, y;
+
+		seq_printf(m, "CRTC %d: pipe: %c, active: %s\n",
+			   crtc->base.base.id, pipe_name(crtc->pipe),
+			   yesno(crtc->active));
+		if (crtc->active) {
+			intel_crtc_info(m, crtc);
+
+			active = cursor_position(dev, crtc->pipe, &x, &y);
+			seq_printf(m, "\tcursor visible? %s, position (%d, %d), addr 0x%08x, active? %s\n",
+				   yesno(crtc->cursor_base),
+				   x, y, crtc->cursor_addr,
+				   yesno(active));
+		}
+
+		seq_printf(m, "\tunderrun reporting: cpu=%s pch=%s \n",
+			   yesno(!crtc->cpu_fifo_underrun_disabled),
+			   yesno(!crtc->pch_fifo_underrun_disabled));
+	}
+
+	seq_printf(m, "\n");
+	seq_printf(m, "Connector info\n");
+	seq_printf(m, "--------------\n");
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		intel_connector_info(m, connector);
+	}
+	drm_modeset_unlock_all(dev);
+	intel_runtime_pm_put(dev_priv);
 
 	return 0;
 }
@@ -1857,6 +2389,9 @@ static int i915_pipe_crc_open(struct inode *inode, struct file *filep)
 	struct drm_i915_private *dev_priv = info->dev->dev_private;
 	struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[info->pipe];
 
+	if (info->pipe >= INTEL_INFO(info->dev)->num_pipes)
+		return -ENODEV;
+
 	spin_lock_irq(&pipe_crc->lock);
 
 	if (pipe_crc->opened) {
@@ -2005,8 +2540,8 @@ static int i915_pipe_crc_create(struct dentry *root, struct drm_minor *minor,
 	info->dev = dev;
 	ent = debugfs_create_file(info->name, S_IRUGO, root, info,
 				  &i915_pipe_crc_fops);
-	if (IS_ERR(ent))
-		return PTR_ERR(ent);
+	if (!ent)
+		return -ENOMEM;
 
 	return drm_add_fake_info_node(minor, ent, info);
 }
@@ -2080,7 +2615,7 @@ static int i9xx_pipe_crc_auto_source(struct drm_device *dev, enum pipe pipe,
 
 	*source = INTEL_PIPE_CRC_SOURCE_PIPE;
 
-	mutex_lock(&dev->mode_config.mutex);
+	drm_modeset_lock_all(dev);
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
 			    base.head) {
 		if (!encoder->base.crtc)
@@ -2116,7 +2651,7 @@ static int i9xx_pipe_crc_auto_source(struct drm_device *dev, enum pipe pipe,
 			break;
 		}
 	}
-	mutex_unlock(&dev->mode_config.mutex);
+	drm_modeset_unlock_all(dev);
 
 	return ret;
 }
@@ -2166,8 +2701,6 @@ static int vlv_pipe_crc_ctl_reg(struct drm_device *dev,
 	if (need_stable_symbols) {
 		uint32_t tmp = I915_READ(PORT_DFT2_G4X);
 
-		WARN_ON(!IS_G4X(dev));
-
 		tmp |= DC_BALANCE_RESET_VLV;
 		if (pipe == PIPE_A)
 			tmp |= PIPE_A_SCRAMBLE_RESET;
@@ -2347,7 +2880,7 @@ static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe,
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
-	u32 val;
+	u32 val = 0; /* shut up gcc */
 	int ret;
 
 	if (pipe_crc->source == source)
@@ -2590,11 +3123,179 @@ static const struct file_operations i915_display_crc_ctl_fops = {
 	.write = display_crc_ctl_write
 };
 
+static void wm_latency_show(struct seq_file *m, const uint16_t wm[5])
+{
+	struct drm_device *dev = m->private;
+	int num_levels = ilk_wm_max_level(dev) + 1;
+	int level;
+
+	drm_modeset_lock_all(dev);
+
+	for (level = 0; level < num_levels; level++) {
+		unsigned int latency = wm[level];
+
+		/* WM1+ latency values in 0.5us units */
+		if (level > 0)
+			latency *= 5;
+
+		seq_printf(m, "WM%d %u (%u.%u usec)\n",
+			   level, wm[level],
+			   latency / 10, latency % 10);
+	}
+
+	drm_modeset_unlock_all(dev);
+}
+
+static int pri_wm_latency_show(struct seq_file *m, void *data)
+{
+	struct drm_device *dev = m->private;
+
+	wm_latency_show(m, to_i915(dev)->wm.pri_latency);
+
+	return 0;
+}
+
+static int spr_wm_latency_show(struct seq_file *m, void *data)
+{
+	struct drm_device *dev = m->private;
+
+	wm_latency_show(m, to_i915(dev)->wm.spr_latency);
+
+	return 0;
+}
+
+static int cur_wm_latency_show(struct seq_file *m, void *data)
+{
+	struct drm_device *dev = m->private;
+
+	wm_latency_show(m, to_i915(dev)->wm.cur_latency);
+
+	return 0;
+}
+
+static int pri_wm_latency_open(struct inode *inode, struct file *file)
+{
+	struct drm_device *dev = inode->i_private;
+
+	if (!HAS_PCH_SPLIT(dev))
+		return -ENODEV;
+
+	return single_open(file, pri_wm_latency_show, dev);
+}
+
+static int spr_wm_latency_open(struct inode *inode, struct file *file)
+{
+	struct drm_device *dev = inode->i_private;
+
+	if (!HAS_PCH_SPLIT(dev))
+		return -ENODEV;
+
+	return single_open(file, spr_wm_latency_show, dev);
+}
+
+static int cur_wm_latency_open(struct inode *inode, struct file *file)
+{
+	struct drm_device *dev = inode->i_private;
+
+	if (!HAS_PCH_SPLIT(dev))
+		return -ENODEV;
+
+	return single_open(file, cur_wm_latency_show, dev);
+}
+
+static ssize_t wm_latency_write(struct file *file, const char __user *ubuf,
+				size_t len, loff_t *offp, uint16_t wm[5])
+{
+	struct seq_file *m = file->private_data;
+	struct drm_device *dev = m->private;
+	uint16_t new[5] = { 0 };
+	int num_levels = ilk_wm_max_level(dev) + 1;
+	int level;
+	int ret;
+	char tmp[32];
+
+	if (len >= sizeof(tmp))
+		return -EINVAL;
+
+	if (copy_from_user(tmp, ubuf, len))
+		return -EFAULT;
+
+	tmp[len] = '\0';
+
+	ret = sscanf(tmp, "%hu %hu %hu %hu %hu", &new[0], &new[1], &new[2], &new[3], &new[4]);
+	if (ret != num_levels)
+		return -EINVAL;
+
+	drm_modeset_lock_all(dev);
+
+	for (level = 0; level < num_levels; level++)
+		wm[level] = new[level];
+
+	drm_modeset_unlock_all(dev);
+
+	return len;
+}
+
+
+static ssize_t pri_wm_latency_write(struct file *file, const char __user *ubuf,
+				    size_t len, loff_t *offp)
+{
+	struct seq_file *m = file->private_data;
+	struct drm_device *dev = m->private;
+
+	return wm_latency_write(file, ubuf, len, offp, to_i915(dev)->wm.pri_latency);
+}
+
+static ssize_t spr_wm_latency_write(struct file *file, const char __user *ubuf,
+				    size_t len, loff_t *offp)
+{
+	struct seq_file *m = file->private_data;
+	struct drm_device *dev = m->private;
+
+	return wm_latency_write(file, ubuf, len, offp, to_i915(dev)->wm.spr_latency);
+}
+
+static ssize_t cur_wm_latency_write(struct file *file, const char __user *ubuf,
+				    size_t len, loff_t *offp)
+{
+	struct seq_file *m = file->private_data;
+	struct drm_device *dev = m->private;
+
+	return wm_latency_write(file, ubuf, len, offp, to_i915(dev)->wm.cur_latency);
+}
+
+static const struct file_operations i915_pri_wm_latency_fops = {
+	.owner = THIS_MODULE,
+	.open = pri_wm_latency_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+	.write = pri_wm_latency_write
+};
+
+static const struct file_operations i915_spr_wm_latency_fops = {
+	.owner = THIS_MODULE,
+	.open = spr_wm_latency_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+	.write = spr_wm_latency_write
+};
+
+static const struct file_operations i915_cur_wm_latency_fops = {
+	.owner = THIS_MODULE,
+	.open = cur_wm_latency_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+	.write = cur_wm_latency_write
+};
+
 static int
 i915_wedged_get(void *data, u64 *val)
 {
 	struct drm_device *dev = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	*val = atomic_read(&dev_priv->gpu_error.reset_counter);
 
@@ -2605,9 +3306,14 @@ static int
 i915_wedged_set(void *data, u64 val)
 {
 	struct drm_device *dev = data;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	intel_runtime_pm_get(dev_priv);
 
-	DRM_INFO("Manually setting wedged to %llu\n", val);
-	i915_handle_error(dev, val);
+	i915_handle_error(dev, val,
+			  "Manually setting wedged to %llu", val);
+
+	intel_runtime_pm_put(dev_priv);
 
 	return 0;
 }
@@ -2620,7 +3326,7 @@ static int
 i915_ring_stop_get(void *data, u64 *val)
 {
 	struct drm_device *dev = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	*val = dev_priv->gpu_error.stop_rings;
 
@@ -2742,7 +3448,7 @@ i915_drop_caches_set(void *data, u64 val)
 	struct i915_vma *vma, *x;
 	int ret;
 
-	DRM_DEBUG_DRIVER("Dropping caches: 0x%08llx\n", val);
+	DRM_DEBUG("Dropping caches: 0x%08llx\n", val);
 
 	/* No need to check and wait for gpu resets, only libdrm auto-restarts
 	 * on ioctls on -EAGAIN. */
@@ -2763,7 +3469,7 @@ i915_drop_caches_set(void *data, u64 val)
 		list_for_each_entry(vm, &dev_priv->vm_list, global_link) {
 			list_for_each_entry_safe(vma, x, &vm->inactive_list,
 						 mm_list) {
-				if (vma->obj->pin_count)
+				if (vma->pin_count)
 					continue;
 
 				ret = i915_vma_unbind(vma);
@@ -2797,7 +3503,7 @@ static int
 i915_max_freq_get(void *data, u64 *val)
 {
 	struct drm_device *dev = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
 	if (!(IS_GEN6(dev) || IS_GEN7(dev)))
@@ -2810,10 +3516,9 @@ i915_max_freq_get(void *data, u64 *val)
 		return ret;
 
 	if (IS_VALLEYVIEW(dev))
-		*val = vlv_gpu_freq(dev_priv->mem_freq,
-				    dev_priv->rps.max_delay);
+		*val = vlv_gpu_freq(dev_priv, dev_priv->rps.max_freq_softlimit);
 	else
-		*val = dev_priv->rps.max_delay * GT_FREQUENCY_MULTIPLIER;
+		*val = dev_priv->rps.max_freq_softlimit * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return 0;
@@ -2824,6 +3529,7 @@ i915_max_freq_set(void *data, u64 val)
 {
 	struct drm_device *dev = data;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 rp_state_cap, hw_max, hw_min;
 	int ret;
 
 	if (!(IS_GEN6(dev) || IS_GEN7(dev)))
@@ -2841,15 +3547,30 @@ i915_max_freq_set(void *data, u64 val)
 	 * Turbo will still be enabled, but won't go above the set value.
 	 */
 	if (IS_VALLEYVIEW(dev)) {
-		val = vlv_freq_opcode(dev_priv->mem_freq, val);
-		dev_priv->rps.max_delay = val;
-		gen6_set_rps(dev, val);
+		val = vlv_freq_opcode(dev_priv, val);
+
+		hw_max = valleyview_rps_max_freq(dev_priv);
+		hw_min = valleyview_rps_min_freq(dev_priv);
 	} else {
 		do_div(val, GT_FREQUENCY_MULTIPLIER);
-		dev_priv->rps.max_delay = val;
-		gen6_set_rps(dev, val);
+
+		rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
+		hw_max = dev_priv->rps.max_freq;
+		hw_min = (rp_state_cap >> 16) & 0xff;
+	}
+
+	if (val < hw_min || val > hw_max || val < dev_priv->rps.min_freq_softlimit) {
+		mutex_unlock(&dev_priv->rps.hw_lock);
+		return -EINVAL;
 	}
 
+	dev_priv->rps.max_freq_softlimit = val;
+
+	if (IS_VALLEYVIEW(dev))
+		valleyview_set_rps(dev, val);
+	else
+		gen6_set_rps(dev, val);
+
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return 0;
@@ -2863,7 +3584,7 @@ static int
 i915_min_freq_get(void *data, u64 *val)
 {
 	struct drm_device *dev = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
 	if (!(IS_GEN6(dev) || IS_GEN7(dev)))
@@ -2876,10 +3597,9 @@ i915_min_freq_get(void *data, u64 *val)
 		return ret;
 
 	if (IS_VALLEYVIEW(dev))
-		*val = vlv_gpu_freq(dev_priv->mem_freq,
-				    dev_priv->rps.min_delay);
+		*val = vlv_gpu_freq(dev_priv, dev_priv->rps.min_freq_softlimit);
 	else
-		*val = dev_priv->rps.min_delay * GT_FREQUENCY_MULTIPLIER;
+		*val = dev_priv->rps.min_freq_softlimit * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return 0;
@@ -2890,6 +3610,7 @@ i915_min_freq_set(void *data, u64 val)
 {
 	struct drm_device *dev = data;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 rp_state_cap, hw_max, hw_min;
 	int ret;
 
 	if (!(IS_GEN6(dev) || IS_GEN7(dev)))
@@ -2907,14 +3628,30 @@ i915_min_freq_set(void *data, u64 val)
 	 * Turbo will still be enabled, but won't go below the set value.
 	 */
 	if (IS_VALLEYVIEW(dev)) {
-		val = vlv_freq_opcode(dev_priv->mem_freq, val);
-		dev_priv->rps.min_delay = val;
-		valleyview_set_rps(dev, val);
+		val = vlv_freq_opcode(dev_priv, val);
+
+		hw_max = valleyview_rps_max_freq(dev_priv);
+		hw_min = valleyview_rps_min_freq(dev_priv);
 	} else {
 		do_div(val, GT_FREQUENCY_MULTIPLIER);
-		dev_priv->rps.min_delay = val;
-		gen6_set_rps(dev, val);
+
+		rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
+		hw_max = dev_priv->rps.max_freq;
+		hw_min = (rp_state_cap >> 16) & 0xff;
+	}
+
+	if (val < hw_min || val > hw_max || val > dev_priv->rps.max_freq_softlimit) {
+		mutex_unlock(&dev_priv->rps.hw_lock);
+		return -EINVAL;
 	}
+
+	dev_priv->rps.min_freq_softlimit = val;
+
+	if (IS_VALLEYVIEW(dev))
+		valleyview_set_rps(dev, val);
+	else
+		gen6_set_rps(dev, val);
+
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return 0;
@@ -2928,7 +3665,7 @@ static int
 i915_cache_sharing_get(void *data, u64 *val)
 {
 	struct drm_device *dev = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 snpcr;
 	int ret;
 
@@ -2938,8 +3675,11 @@ i915_cache_sharing_get(void *data, u64 *val)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	snpcr = I915_READ(GEN6_MBCUNIT_SNPCR);
+
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev_priv->dev->struct_mutex);
 
 	*val = (snpcr & GEN6_MBC_SNPCR_MASK) >> GEN6_MBC_SNPCR_SHIFT;
@@ -2960,6 +3700,7 @@ i915_cache_sharing_set(void *data, u64 val)
 	if (val > 3)
 		return -EINVAL;
 
+	intel_runtime_pm_get(dev_priv);
 	DRM_DEBUG_DRIVER("Manually setting uncore sharing to %llu\n", val);
 
 	/* Update the cache sharing policy here as well */
@@ -2968,6 +3709,7 @@ i915_cache_sharing_set(void *data, u64 val)
 	snpcr |= (val << GEN6_MBC_SNPCR_SHIFT);
 	I915_WRITE(GEN6_MBCUNIT_SNPCR, snpcr);
 
+	intel_runtime_pm_put(dev_priv);
 	return 0;
 }
 
@@ -2983,7 +3725,7 @@ static int i915_forcewake_open(struct inode *inode, struct file *file)
 	if (INTEL_INFO(dev)->gen < 6)
 		return 0;
 
-	gen6_gt_force_wake_get(dev_priv);
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	return 0;
 }
@@ -2996,7 +3738,7 @@ static int i915_forcewake_release(struct inode *inode, struct file *file)
 	if (INTEL_INFO(dev)->gen < 6)
 		return 0;
 
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 
 	return 0;
 }
@@ -3016,8 +3758,8 @@ static int i915_forcewake_create(struct dentry *root, struct drm_minor *minor)
 				  S_IRUSR,
 				  root, dev,
 				  &i915_forcewake_fops);
-	if (IS_ERR(ent))
-		return PTR_ERR(ent);
+	if (!ent)
+		return -ENOMEM;
 
 	return drm_add_fake_info_node(minor, ent, &i915_forcewake_fops);
 }
@@ -3034,8 +3776,8 @@ static int i915_debugfs_create(struct dentry *root,
 				  S_IRUGO | S_IWUSR,
 				  root, dev,
 				  fops);
-	if (IS_ERR(ent))
-		return PTR_ERR(ent);
+	if (!ent)
+		return -ENOMEM;
 
 	return drm_add_fake_info_node(minor, ent, fops);
 }
@@ -3058,7 +3800,7 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_gem_hws_bsd", i915_hws_info, 0, (void *)VCS},
 	{"i915_gem_hws_vebox", i915_hws_info, 0, (void *)VECS},
 	{"i915_rstdby_delays", i915_rstdby_delays, 0},
-	{"i915_cur_delayinfo", i915_cur_delayinfo, 0},
+	{"i915_frequency_info", i915_frequency_info, 0},
 	{"i915_delayfreq_table", i915_delayfreq_table, 0},
 	{"i915_inttoext_table", i915_inttoext_table, 0},
 	{"i915_drpc_info", i915_drpc_info, 0},
@@ -3074,11 +3816,13 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_gen6_forcewake_count", i915_gen6_forcewake_count_info, 0},
 	{"i915_swizzle_info", i915_swizzle_info, 0},
 	{"i915_ppgtt_info", i915_ppgtt_info, 0},
-	{"i915_dpio", i915_dpio_info, 0},
 	{"i915_llc", i915_llc, 0},
 	{"i915_edp_psr_status", i915_edp_psr_status, 0},
+	{"i915_sink_crc_eDP1", i915_sink_crc, 0},
 	{"i915_energy_uJ", i915_energy_uJ, 0},
 	{"i915_pc8_status", i915_pc8_status, 0},
+	{"i915_power_domain_info", i915_power_domain_info, 0},
+	{"i915_display_info", i915_display_info, 0},
 };
 #define I915_DEBUGFS_ENTRIES ARRAY_SIZE(i915_debugfs_list)
 
@@ -3097,15 +3841,18 @@ static const struct i915_debugfs_files {
 	{"i915_error_state", &i915_error_state_fops},
 	{"i915_next_seqno", &i915_next_seqno_fops},
 	{"i915_display_crc_ctl", &i915_display_crc_ctl_fops},
+	{"i915_pri_wm_latency", &i915_pri_wm_latency_fops},
+	{"i915_spr_wm_latency", &i915_spr_wm_latency_fops},
+	{"i915_cur_wm_latency", &i915_cur_wm_latency_fops},
 };
 
 void intel_display_crc_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int i;
+	enum pipe pipe;
 
-	for (i = 0; i < INTEL_INFO(dev)->num_pipes; i++) {
-		struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[i];
+	for_each_pipe(pipe) {
+		struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
 
 		pipe_crc->opened = false;
 		spin_lock_init(&pipe_crc->lock);
@@ -3164,5 +3911,3 @@ void i915_debugfs_cleanup(struct drm_minor *minor)
 		drm_debugfs_remove_files(info_list, 1, minor);
 	}
 }
-
-#endif /* CONFIG_DEBUG_FS */
diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 5c648425c1e..d4434414062 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -36,12 +36,17 @@
 #include "i915_drv.h"
 #include "i915_trace.h"
 #include <linux/pci.h>
+#include <linux/console.h>
+#include <linux/vt.h>
 #include <linux/vgaarb.h>
 #include <linux/acpi.h>
 #include <linux/pnp.h>
 #include <linux/vga_switcheroo.h>
 #include <linux/slab.h>
 #include <acpi/video.h>
+#include <linux/pm.h>
+#include <linux/pm_runtime.h>
+#include <linux/oom.h>
 
 #define LP_RING(d) (&((struct drm_i915_private *)(d))->ring[RCS])
 
@@ -61,7 +66,7 @@
  * has access to the ring.
  */
 #define RING_LOCK_TEST_WITH_RETURN(dev, file) do {			\
-	if (LP_RING(dev->dev_private)->obj == NULL)			\
+	if (LP_RING(dev->dev_private)->buffer->obj == NULL)			\
 		LOCK_TEST_WITH_RETURN(dev, file);			\
 } while (0)
 
@@ -80,7 +85,7 @@ intel_read_legacy_status_page(struct drm_i915_private *dev_priv, int reg)
 
 void i915_update_dri1_breadcrumb(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_master_private *master_priv;
 
 	/*
@@ -101,7 +106,7 @@ void i915_update_dri1_breadcrumb(struct drm_device *dev)
 
 static void i915_write_hws_pga(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 addr;
 
 	addr = dev_priv->status_page_dmah->busaddr;
@@ -116,8 +121,8 @@ static void i915_write_hws_pga(struct drm_device *dev)
  */
 static void i915_free_hws(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = LP_RING(dev_priv);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = LP_RING(dev_priv);
 
 	if (dev_priv->status_page_dmah) {
 		drm_pci_free(dev, dev_priv->status_page_dmah);
@@ -135,9 +140,10 @@ static void i915_free_hws(struct drm_device *dev)
 
 void i915_kernel_lost_context(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_master_private *master_priv;
-	struct intel_ring_buffer *ring = LP_RING(dev_priv);
+	struct intel_engine_cs *ring = LP_RING(dev_priv);
+	struct intel_ringbuffer *ringbuf = ring->buffer;
 
 	/*
 	 * We should never lose context on the ring with modesetting
@@ -146,23 +152,23 @@ void i915_kernel_lost_context(struct drm_device * dev)
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return;
 
-	ring->head = I915_READ_HEAD(ring) & HEAD_ADDR;
-	ring->tail = I915_READ_TAIL(ring) & TAIL_ADDR;
-	ring->space = ring->head - (ring->tail + I915_RING_FREE_SPACE);
-	if (ring->space < 0)
-		ring->space += ring->size;
+	ringbuf->head = I915_READ_HEAD(ring) & HEAD_ADDR;
+	ringbuf->tail = I915_READ_TAIL(ring) & TAIL_ADDR;
+	ringbuf->space = ringbuf->head - (ringbuf->tail + I915_RING_FREE_SPACE);
+	if (ringbuf->space < 0)
+		ringbuf->space += ringbuf->size;
 
 	if (!dev->primary->master)
 		return;
 
 	master_priv = dev->primary->master->driver_priv;
-	if (ring->head == ring->tail && master_priv->sarea_priv)
+	if (ringbuf->head == ringbuf->tail && master_priv->sarea_priv)
 		master_priv->sarea_priv->perf_boxes |= I915_BOX_RING_EMPTY;
 }
 
 static int i915_dma_cleanup(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 
 	/* Make sure interrupts are disabled here because the uninstall ioctl
@@ -186,7 +192,7 @@ static int i915_dma_cleanup(struct drm_device * dev)
 
 static int i915_initialize(struct drm_device * dev, drm_i915_init_t * init)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
 	int ret;
 
@@ -199,7 +205,7 @@ static int i915_initialize(struct drm_device * dev, drm_i915_init_t * init)
 	}
 
 	if (init->ring_size != 0) {
-		if (LP_RING(dev_priv)->obj != NULL) {
+		if (LP_RING(dev_priv)->buffer->obj != NULL) {
 			i915_dma_cleanup(dev);
 			DRM_ERROR("Client tried to initialize ringbuffer in "
 				  "GEM mode\n");
@@ -231,12 +237,12 @@ static int i915_initialize(struct drm_device * dev, drm_i915_init_t * init)
 
 static int i915_dma_resume(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
-	struct intel_ring_buffer *ring = LP_RING(dev_priv);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = LP_RING(dev_priv);
 
 	DRM_DEBUG_DRIVER("%s\n", __func__);
 
-	if (ring->virtual_start == NULL) {
+	if (ring->buffer->virtual_start == NULL) {
 		DRM_ERROR("can not ioremap virtual address for"
 			  " ring buffer\n");
 		return -ENOMEM;
@@ -355,10 +361,10 @@ static int validate_cmd(int cmd)
 
 static int i915_emit_cmds(struct drm_device * dev, int *buffer, int dwords)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i, ret;
 
-	if ((dwords+1) * sizeof(int) >= LP_RING(dev_priv)->size - 8)
+	if ((dwords+1) * sizeof(int) >= LP_RING(dev_priv)->buffer->size - 8)
 		return -EINVAL;
 
 	for (i = 0; i < dwords;) {
@@ -429,7 +435,7 @@ i915_emit_box(struct drm_device *dev,
 
 static void i915_emit_breadcrumb(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
 
 	dev_priv->dri1.counter++;
@@ -545,7 +551,7 @@ static int i915_dispatch_batchbuffer(struct drm_device * dev,
 
 static int i915_dispatch_flip(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_master_private *master_priv =
 		dev->primary->master->driver_priv;
 	int ret;
@@ -623,10 +629,9 @@ static int i915_flush_ioctl(struct drm_device *dev, void *data,
 static int i915_batchbuffer(struct drm_device *dev, void *data,
 			    struct drm_file *file_priv)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
-	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
-	drm_i915_sarea_t *sarea_priv = (drm_i915_sarea_t *)
-	    master_priv->sarea_priv;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_master_private *master_priv;
+	drm_i915_sarea_t *sarea_priv;
 	drm_i915_batchbuffer_t *batch = data;
 	int ret;
 	struct drm_clip_rect *cliprects = NULL;
@@ -634,6 +639,9 @@ static int i915_batchbuffer(struct drm_device *dev, void *data,
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return -ENODEV;
 
+	master_priv = dev->primary->master->driver_priv;
+	sarea_priv = (drm_i915_sarea_t *) master_priv->sarea_priv;
+
 	if (!dev_priv->dri1.allow_batchbuffer) {
 		DRM_ERROR("Batchbuffer ioctl disabled\n");
 		return -EINVAL;
@@ -679,10 +687,9 @@ fail_free:
 static int i915_cmdbuffer(struct drm_device *dev, void *data,
 			  struct drm_file *file_priv)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
-	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
-	drm_i915_sarea_t *sarea_priv = (drm_i915_sarea_t *)
-	    master_priv->sarea_priv;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_master_private *master_priv;
+	drm_i915_sarea_t *sarea_priv;
 	drm_i915_cmdbuffer_t *cmdbuf = data;
 	struct drm_clip_rect *cliprects = NULL;
 	void *batch_data;
@@ -694,6 +701,9 @@ static int i915_cmdbuffer(struct drm_device *dev, void *data,
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return -ENODEV;
 
+	master_priv = dev->primary->master->driver_priv;
+	sarea_priv = (drm_i915_sarea_t *) master_priv->sarea_priv;
+
 	RING_LOCK_TEST_WITH_RETURN(dev, file_priv);
 
 	if (cmdbuf->num_cliprects < 0)
@@ -747,7 +757,7 @@ fail_batch_free:
 
 static int i915_emit_irq(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
 
 	i915_kernel_lost_context(dev);
@@ -773,10 +783,10 @@ static int i915_emit_irq(struct drm_device * dev)
 
 static int i915_wait_irq(struct drm_device * dev, int irq_nr)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
 	int ret = 0;
-	struct intel_ring_buffer *ring = LP_RING(dev_priv);
+	struct intel_engine_cs *ring = LP_RING(dev_priv);
 
 	DRM_DEBUG_DRIVER("irq_nr=%d breadcrumb=%d\n", irq_nr,
 		  READ_BREADCRUMB(dev_priv));
@@ -791,7 +801,7 @@ static int i915_wait_irq(struct drm_device * dev, int irq_nr)
 		master_priv->sarea_priv->perf_boxes |= I915_BOX_WAIT;
 
 	if (ring->irq_get(ring)) {
-		DRM_WAIT_ON(ret, ring->irq_queue, 3 * DRM_HZ,
+		DRM_WAIT_ON(ret, ring->irq_queue, 3 * HZ,
 			    READ_BREADCRUMB(dev_priv) >= irq_nr);
 		ring->irq_put(ring);
 	} else if (wait_for(READ_BREADCRUMB(dev_priv) >= irq_nr, 3000))
@@ -810,14 +820,14 @@ static int i915_wait_irq(struct drm_device * dev, int irq_nr)
 static int i915_irq_emit(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	drm_i915_irq_emit_t *emit = data;
 	int result;
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return -ENODEV;
 
-	if (!dev_priv || !LP_RING(dev_priv)->virtual_start) {
+	if (!dev_priv || !LP_RING(dev_priv)->buffer->virtual_start) {
 		DRM_ERROR("called with no initialization\n");
 		return -EINVAL;
 	}
@@ -828,7 +838,7 @@ static int i915_irq_emit(struct drm_device *dev, void *data,
 	result = i915_emit_irq(dev);
 	mutex_unlock(&dev->struct_mutex);
 
-	if (DRM_COPY_TO_USER(emit->irq_seq, &result, sizeof(int))) {
+	if (copy_to_user(emit->irq_seq, &result, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
@@ -841,7 +851,7 @@ static int i915_irq_emit(struct drm_device *dev, void *data,
 static int i915_irq_wait(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	drm_i915_irq_wait_t *irqwait = data;
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
@@ -858,7 +868,7 @@ static int i915_irq_wait(struct drm_device *dev, void *data,
 static int i915_vblank_pipe_get(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	drm_i915_vblank_pipe_t *pipe = data;
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
@@ -919,7 +929,7 @@ static int i915_flip_bufs(struct drm_device *dev, void *data,
 static int i915_getparam(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	drm_i915_getparam_t *param = data;
 	int value;
 
@@ -988,7 +998,7 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		value = HAS_WT(dev);
 		break;
 	case I915_PARAM_HAS_ALIASING_PPGTT:
-		value = dev_priv->mm.aliasing_ppgtt ? 1 : 0;
+		value = dev_priv->mm.aliasing_ppgtt || USES_FULL_PPGTT(dev);
 		break;
 	case I915_PARAM_HAS_WAIT_TIMEOUT:
 		value = 1;
@@ -1011,13 +1021,16 @@ static int i915_getparam(struct drm_device *dev, void *data,
 	case I915_PARAM_HAS_EXEC_HANDLE_LUT:
 		value = 1;
 		break;
+	case I915_PARAM_CMD_PARSER_VERSION:
+		value = i915_cmd_parser_get_version();
+		break;
 	default:
 		DRM_DEBUG("Unknown parameter %d\n", param->param);
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_TO_USER(param->value, &value, sizeof(int))) {
-		DRM_ERROR("DRM_COPY_TO_USER failed\n");
+	if (copy_to_user(param->value, &value, sizeof(int))) {
+		DRM_ERROR("copy_to_user failed\n");
 		return -EFAULT;
 	}
 
@@ -1027,7 +1040,7 @@ static int i915_getparam(struct drm_device *dev, void *data,
 static int i915_setparam(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	drm_i915_setparam_t *param = data;
 
 	if (!dev_priv) {
@@ -1062,9 +1075,9 @@ static int i915_setparam(struct drm_device *dev, void *data,
 static int i915_set_status_page(struct drm_device *dev, void *data,
 				struct drm_file *file_priv)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	drm_i915_hws_addr_t *hws = data;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return -ENODEV;
@@ -1130,7 +1143,7 @@ static int i915_get_bridge_dev(struct drm_device *dev)
 static int
 intel_alloc_mchbar_resource(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int reg = INTEL_INFO(dev)->gen >= 4 ? MCHBAR_I965 : MCHBAR_I915;
 	u32 temp_lo, temp_hi = 0;
 	u64 mchbar_addr;
@@ -1176,11 +1189,14 @@ intel_alloc_mchbar_resource(struct drm_device *dev)
 static void
 intel_setup_mchbar(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int mchbar_reg = INTEL_INFO(dev)->gen >= 4 ? MCHBAR_I965 : MCHBAR_I915;
 	u32 temp;
 	bool enabled;
 
+	if (IS_VALLEYVIEW(dev))
+		return;
+
 	dev_priv->mchbar_need_disable = false;
 
 	if (IS_I915G(dev) || IS_I915GM(dev)) {
@@ -1213,7 +1229,7 @@ intel_setup_mchbar(struct drm_device *dev)
 static void
 intel_teardown_mchbar(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int mchbar_reg = INTEL_INFO(dev)->gen >= 4 ? MCHBAR_I965 : MCHBAR_I915;
 	u32 temp;
 
@@ -1268,12 +1284,13 @@ static void i915_switcheroo_set_state(struct pci_dev *pdev, enum vga_switcheroo_
 static bool i915_switcheroo_can_switch(struct pci_dev *pdev)
 {
 	struct drm_device *dev = pci_get_drvdata(pdev);
-	bool can_switch;
 
-	spin_lock(&dev->count_lock);
-	can_switch = (dev->open_count == 0);
-	spin_unlock(&dev->count_lock);
-	return can_switch;
+	/*
+	 * FIXME: open_count is protected by drm_global_mutex but that would lead to
+	 * locking inversion with the driver load path. And the access here is
+	 * completely racy anyway. So don't bother with locking for now.
+	 */
+	return dev->open_count == 0;
 }
 
 static const struct vga_switcheroo_client_ops i915_switcheroo_ops = {
@@ -1315,19 +1332,19 @@ static int i915_load_modeset_init(struct drm_device *dev)
 	if (ret)
 		goto cleanup_vga_switcheroo;
 
-	ret = drm_irq_install(dev);
+	intel_power_domains_init_hw(dev_priv);
+
+	ret = drm_irq_install(dev, dev->pdev->irq);
 	if (ret)
 		goto cleanup_gem_stolen;
 
-	intel_power_domains_init_hw(dev);
-
 	/* Important: The output setup functions called by modeset_init need
 	 * working irqs for e.g. gmbus and dp aux transfers. */
 	intel_modeset_init(dev);
 
 	ret = i915_gem_init(dev);
 	if (ret)
-		goto cleanup_power;
+		goto cleanup_irq;
 
 	INIT_WORK(&dev_priv->console_resume_work, intel_console_resume);
 
@@ -1336,10 +1353,8 @@ static int i915_load_modeset_init(struct drm_device *dev)
 	/* Always safe in the mode setting case. */
 	/* FIXME: do pre/post-mode set stuff in core KMS code */
 	dev->vblank_disable_allowed = true;
-	if (INTEL_INFO(dev)->num_pipes == 0) {
-		intel_display_power_put(dev, POWER_DOMAIN_VGA);
+	if (INTEL_INFO(dev)->num_pipes == 0)
 		return 0;
-	}
 
 	ret = intel_fbdev_init(dev);
 	if (ret)
@@ -1372,10 +1387,8 @@ cleanup_gem:
 	i915_gem_cleanup_ringbuffer(dev);
 	i915_gem_context_fini(dev);
 	mutex_unlock(&dev->struct_mutex);
-	i915_gem_cleanup_aliasing_ppgtt(dev);
-	drm_mm_takedown(&dev_priv->gtt.base.mm);
-cleanup_power:
-	intel_display_power_put(dev, POWER_DOMAIN_VGA);
+	WARN_ON(dev_priv->mm.aliasing_ppgtt);
+cleanup_irq:
 	drm_irq_uninstall(dev);
 cleanup_gem_stolen:
 	i915_gem_cleanup_stolen(dev);
@@ -1411,7 +1424,7 @@ void i915_master_destroy(struct drm_device *dev, struct drm_master *master)
 	master->driver_priv = NULL;
 }
 
-#ifdef CONFIG_DRM_I915_FBDEV
+#if IS_ENABLED(CONFIG_FB)
 static void i915_kick_out_firmware_fb(struct drm_i915_private *dev_priv)
 {
 	struct apertures_struct *ap;
@@ -1438,9 +1451,42 @@ static void i915_kick_out_firmware_fb(struct drm_i915_private *dev_priv)
 }
 #endif
 
+#if !defined(CONFIG_VGA_CONSOLE)
+static int i915_kick_out_vgacon(struct drm_i915_private *dev_priv)
+{
+	return 0;
+}
+#elif !defined(CONFIG_DUMMY_CONSOLE)
+static int i915_kick_out_vgacon(struct drm_i915_private *dev_priv)
+{
+	return -ENODEV;
+}
+#else
+static int i915_kick_out_vgacon(struct drm_i915_private *dev_priv)
+{
+	int ret = 0;
+
+	DRM_INFO("Replacing VGA console driver\n");
+
+	console_lock();
+	if (con_is_bound(&vga_con))
+		ret = do_take_over_console(&dummy_con, 0, MAX_NR_CONSOLES - 1, 1);
+	if (ret == 0) {
+		ret = do_unregister_con_driver(&vga_con);
+
+		/* Ignore "already unregistered". */
+		if (ret == -ENODEV)
+			ret = 0;
+	}
+	console_unlock();
+
+	return ret;
+}
+#endif
+
 static void i915_dump_device_info(struct drm_i915_private *dev_priv)
 {
-	const struct intel_device_info *info = dev_priv->info;
+	const struct intel_device_info *info = &dev_priv->info;
 
 #define PRINT_S(name) "%s"
 #define SEP_EMPTY
@@ -1457,6 +1503,62 @@ static void i915_dump_device_info(struct drm_i915_private *dev_priv)
 #undef SEP_COMMA
 }
 
+/*
+ * Determine various intel_device_info fields at runtime.
+ *
+ * Use it when either:
+ *   - it's judged too laborious to fill n static structures with the limit
+ *     when a simple if statement does the job,
+ *   - run-time checks (eg read fuse/strap registers) are needed.
+ *
+ * This function needs to be called:
+ *   - after the MMIO has been setup as we are reading registers,
+ *   - after the PCH has been detected,
+ *   - before the first usage of the fields it can tweak.
+ */
+static void intel_device_info_runtime_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_device_info *info;
+	enum pipe pipe;
+
+	info = (struct intel_device_info *)&dev_priv->info;
+
+	if (IS_VALLEYVIEW(dev))
+		for_each_pipe(pipe)
+			info->num_sprites[pipe] = 2;
+	else
+		for_each_pipe(pipe)
+			info->num_sprites[pipe] = 1;
+
+	if (i915.disable_display) {
+		DRM_INFO("Display disabled (module parameter)\n");
+		info->num_pipes = 0;
+	} else if (info->num_pipes > 0 &&
+		   (INTEL_INFO(dev)->gen == 7 || INTEL_INFO(dev)->gen == 8) &&
+		   !IS_VALLEYVIEW(dev)) {
+		u32 fuse_strap = I915_READ(FUSE_STRAP);
+		u32 sfuse_strap = I915_READ(SFUSE_STRAP);
+
+		/*
+		 * SFUSE_STRAP is supposed to have a bit signalling the display
+		 * is fused off. Unfortunately it seems that, at least in
+		 * certain cases, fused off display means that PCH display
+		 * reads don't land anywhere. In that case, we read 0s.
+		 *
+		 * On CPT/PPT, we can detect this case as SFUSE_STRAP_FUSE_LOCK
+		 * should be set when taking over after the firmware.
+		 */
+		if (fuse_strap & ILK_INTERNAL_DISPLAY_DISABLE ||
+		    sfuse_strap & SFUSE_STRAP_DISPLAY_DISABLED ||
+		    (dev_priv->pch_type == PCH_CPT &&
+		     !(sfuse_strap & SFUSE_STRAP_FUSE_LOCK))) {
+			DRM_INFO("Display fused off, disabling\n");
+			info->num_pipes = 0;
+		}
+	}
+}
+
 /**
  * i915_driver_load - setup chip and create an initial config
  * @dev: DRM device
@@ -1471,7 +1573,7 @@ static void i915_dump_device_info(struct drm_i915_private *dev_priv)
 int i915_driver_load(struct drm_device *dev, unsigned long flags)
 {
 	struct drm_i915_private *dev_priv;
-	struct intel_device_info *info;
+	struct intel_device_info *info, *device_info;
 	int ret = 0, mmio_bar, mmio_size;
 	uint32_t aperture_size;
 
@@ -1484,17 +1586,24 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 		return -ENODEV;
 	}
 
+	/* UMS needs agp support. */
+	if (!drm_core_check_feature(dev, DRIVER_MODESET) && !dev->agp)
+		return -EINVAL;
+
 	dev_priv = kzalloc(sizeof(*dev_priv), GFP_KERNEL);
 	if (dev_priv == NULL)
 		return -ENOMEM;
 
 	dev->dev_private = (void *)dev_priv;
 	dev_priv->dev = dev;
-	dev_priv->info = info;
+
+	/* copy initial configuration to dev_priv->info */
+	device_info = (struct intel_device_info *)&dev_priv->info;
+	*device_info = *info;
 
 	spin_lock_init(&dev_priv->irq_lock);
 	spin_lock_init(&dev_priv->gpu_error.lock);
-	spin_lock_init(&dev_priv->backlight.lock);
+	spin_lock_init(&dev_priv->backlight_lock);
 	spin_lock_init(&dev_priv->uncore.lock);
 	spin_lock_init(&dev_priv->mm.object_stat_lock);
 	mutex_init(&dev_priv->dpio_lock);
@@ -1539,8 +1648,6 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 		goto put_bridge;
 	}
 
-	intel_uncore_early_sanitize(dev);
-
 	/* This must be called before any calls to HAS_PCH_* */
 	intel_detect_pch(dev);
 
@@ -1550,8 +1657,15 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	if (ret)
 		goto out_regs;
 
-	if (drm_core_check_feature(dev, DRIVER_MODESET))
+	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
+		ret = i915_kick_out_vgacon(dev_priv);
+		if (ret) {
+			DRM_ERROR("failed to remove conflicting VGA console\n");
+			goto out_gtt;
+		}
+
 		i915_kick_out_firmware_fb(dev_priv);
+	}
 
 	pci_set_master(dev->pdev);
 
@@ -1629,9 +1743,7 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	if (!IS_I945G(dev) && !IS_I945GM(dev))
 		pci_enable_msi(dev->pdev);
 
-	dev_priv->num_plane = 1;
-	if (IS_VALLEYVIEW(dev))
-		dev_priv->num_plane = 2;
+	intel_device_info_runtime_init(dev);
 
 	if (INTEL_INFO(dev)->num_pipes) {
 		ret = drm_vblank_init(dev, INTEL_INFO(dev)->num_pipes);
@@ -1639,8 +1751,7 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 			goto out_gem_unload;
 	}
 
-	if (HAS_POWER_WELL(dev))
-		intel_power_domains_init(dev);
+	intel_power_domains_init(dev_priv);
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		ret = i915_load_modeset_init(dev);
@@ -1664,28 +1775,28 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	if (IS_GEN5(dev))
 		intel_gpu_ips_init(dev_priv);
 
+	intel_init_runtime_pm(dev_priv);
+
 	return 0;
 
 out_power_well:
-	if (HAS_POWER_WELL(dev))
-		intel_power_domains_remove(dev);
+	intel_power_domains_remove(dev_priv);
 	drm_vblank_cleanup(dev);
 out_gem_unload:
-	if (dev_priv->mm.inactive_shrinker.scan_objects)
-		unregister_shrinker(&dev_priv->mm.inactive_shrinker);
+	WARN_ON(unregister_oom_notifier(&dev_priv->mm.oom_notifier));
+	unregister_shrinker(&dev_priv->mm.shrinker);
 
 	if (dev->pdev->msi_enabled)
 		pci_disable_msi(dev->pdev);
 
 	intel_teardown_gmbus(dev);
 	intel_teardown_mchbar(dev);
+	pm_qos_remove_request(&dev_priv->pm_qos);
 	destroy_workqueue(dev_priv->wq);
 out_mtrrfree:
 	arch_phys_wc_del(dev_priv->gtt.mtrr);
 	io_mapping_free(dev_priv->gtt.mappable);
 out_gtt:
-	list_del(&dev_priv->gtt.base.global_link);
-	drm_mm_takedown(&dev_priv->gtt.base.mm);
 	dev_priv->gtt.base.cleanup(&dev_priv->gtt.base);
 out_regs:
 	intel_uncore_fini(dev);
@@ -1704,24 +1815,26 @@ int i915_driver_unload(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
+	ret = i915_gem_suspend(dev);
+	if (ret) {
+		DRM_ERROR("failed to idle hardware: %d\n", ret);
+		return ret;
+	}
+
+	intel_fini_runtime_pm(dev_priv);
+
 	intel_gpu_ips_teardown();
 
-	if (HAS_POWER_WELL(dev)) {
-		/* The i915.ko module is still not prepared to be loaded when
-		 * the power well is not enabled, so just enable it in case
-		 * we're going to unload/reload. */
-		intel_display_set_init_power(dev, true);
-		intel_power_domains_remove(dev);
-	}
+	/* The i915.ko module is still not prepared to be loaded when
+	 * the power well is not enabled, so just enable it in case
+	 * we're going to unload/reload. */
+	intel_display_set_init_power(dev_priv, true);
+	intel_power_domains_remove(dev_priv);
 
 	i915_teardown_sysfs(dev);
 
-	if (dev_priv->mm.inactive_shrinker.scan_objects)
-		unregister_shrinker(&dev_priv->mm.inactive_shrinker);
-
-	ret = i915_gem_suspend(dev);
-	if (ret)
-		DRM_ERROR("failed to idle hardware: %d\n", ret);
+	WARN_ON(unregister_oom_notifier(&dev_priv->mm.oom_notifier));
+	unregister_shrinker(&dev_priv->mm.shrinker);
 
 	io_mapping_free(dev_priv->gtt.mappable);
 	arch_phys_wc_del(dev_priv->gtt.mtrr);
@@ -1752,8 +1865,6 @@ int i915_driver_unload(struct drm_device *dev)
 	cancel_work_sync(&dev_priv->gpu_error.work);
 	i915_destroy_error_state(dev);
 
-	cancel_delayed_work_sync(&dev_priv->pc8.enable_work);
-
 	if (dev->pdev->msi_enabled)
 		pci_disable_msi(dev->pdev);
 
@@ -1764,20 +1875,17 @@ int i915_driver_unload(struct drm_device *dev)
 		flush_workqueue(dev_priv->wq);
 
 		mutex_lock(&dev->struct_mutex);
-		i915_gem_free_all_phys_object(dev);
 		i915_gem_cleanup_ringbuffer(dev);
 		i915_gem_context_fini(dev);
+		WARN_ON(dev_priv->mm.aliasing_ppgtt);
 		mutex_unlock(&dev->struct_mutex);
-		i915_gem_cleanup_aliasing_ppgtt(dev);
 		i915_gem_cleanup_stolen(dev);
 
 		if (!I915_NEED_GFX_HWS(dev))
 			i915_free_hws(dev);
 	}
 
-	list_del(&dev_priv->gtt.base.global_link);
 	WARN_ON(!list_empty(&dev_priv->vm_list));
-	drm_mm_takedown(&dev_priv->gtt.base.mm);
 
 	drm_vblank_cleanup(dev);
 
@@ -1797,7 +1905,7 @@ int i915_driver_unload(struct drm_device *dev)
 		kmem_cache_destroy(dev_priv->slab);
 
 	pci_dev_put(dev_priv->bridge_dev);
-	kfree(dev->dev_private);
+	kfree(dev_priv);
 
 	return 0;
 }
@@ -1827,7 +1935,7 @@ int i915_driver_open(struct drm_device *dev, struct drm_file *file)
  */
 void i915_driver_lastclose(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	/* On gen6+ we refuse to init without kms enabled, but then the drm core
 	 * goes right around and calls lastclose. Check for this and don't clean
@@ -1858,6 +1966,8 @@ void i915_driver_postclose(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_i915_file_private *file_priv = file->driver_priv;
 
+	if (file_priv && file_priv->bsd_ring)
+		file_priv->bsd_ring = NULL;
 	kfree(file_priv);
 }
 
@@ -1910,9 +2020,11 @@ const struct drm_ioctl_desc i915_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_CREATE, i915_gem_context_create_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_DESTROY, i915_gem_context_destroy_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_REG_READ, i915_reg_read_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(I915_GET_RESET_STATS, i915_get_reset_stats_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(I915_GEM_USERPTR, i915_gem_userptr_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 };
 
-int i915_max_ioctl = DRM_ARRAY_SIZE(i915_ioctls);
+int i915_max_ioctl = ARRAY_SIZE(i915_ioctls);
 
 /*
  * This is really ugly: Because old userspace abused the linux agp interface to
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 5b7b7e06cb3..651e65e051c 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -36,175 +36,105 @@
 
 #include <linux/console.h>
 #include <linux/module.h>
+#include <linux/pm_runtime.h>
 #include <drm/drm_crtc_helper.h>
 
-static int i915_modeset __read_mostly = -1;
-module_param_named(modeset, i915_modeset, int, 0400);
-MODULE_PARM_DESC(modeset,
-		"Use kernel modesetting [KMS] (0=DRM_I915_KMS from .config, "
-		"1=on, -1=force vga console preference [default])");
-
-unsigned int i915_fbpercrtc __always_unused = 0;
-module_param_named(fbpercrtc, i915_fbpercrtc, int, 0400);
-
-int i915_panel_ignore_lid __read_mostly = 1;
-module_param_named(panel_ignore_lid, i915_panel_ignore_lid, int, 0600);
-MODULE_PARM_DESC(panel_ignore_lid,
-		"Override lid status (0=autodetect, 1=autodetect disabled [default], "
-		"-1=force lid closed, -2=force lid open)");
-
-unsigned int i915_powersave __read_mostly = 1;
-module_param_named(powersave, i915_powersave, int, 0600);
-MODULE_PARM_DESC(powersave,
-		"Enable powersavings, fbc, downclocking, etc. (default: true)");
-
-int i915_semaphores __read_mostly = -1;
-module_param_named(semaphores, i915_semaphores, int, 0600);
-MODULE_PARM_DESC(semaphores,
-		"Use semaphores for inter-ring sync (default: -1 (use per-chip defaults))");
-
-int i915_enable_rc6 __read_mostly = -1;
-module_param_named(i915_enable_rc6, i915_enable_rc6, int, 0400);
-MODULE_PARM_DESC(i915_enable_rc6,
-		"Enable power-saving render C-state 6. "
-		"Different stages can be selected via bitmask values "
-		"(0 = disable; 1 = enable rc6; 2 = enable deep rc6; 4 = enable deepest rc6). "
-		"For example, 3 would enable rc6 and deep rc6, and 7 would enable everything. "
-		"default: -1 (use per-chip default)");
-
-int i915_enable_fbc __read_mostly = -1;
-module_param_named(i915_enable_fbc, i915_enable_fbc, int, 0600);
-MODULE_PARM_DESC(i915_enable_fbc,
-		"Enable frame buffer compression for power savings "
-		"(default: -1 (use per-chip default))");
-
-unsigned int i915_lvds_downclock __read_mostly = 0;
-module_param_named(lvds_downclock, i915_lvds_downclock, int, 0400);
-MODULE_PARM_DESC(lvds_downclock,
-		"Use panel (LVDS/eDP) downclocking for power savings "
-		"(default: false)");
-
-int i915_lvds_channel_mode __read_mostly;
-module_param_named(lvds_channel_mode, i915_lvds_channel_mode, int, 0600);
-MODULE_PARM_DESC(lvds_channel_mode,
-		 "Specify LVDS channel mode "
-		 "(0=probe BIOS [default], 1=single-channel, 2=dual-channel)");
-
-int i915_panel_use_ssc __read_mostly = -1;
-module_param_named(lvds_use_ssc, i915_panel_use_ssc, int, 0600);
-MODULE_PARM_DESC(lvds_use_ssc,
-		"Use Spread Spectrum Clock with panels [LVDS/eDP] "
-		"(default: auto from VBT)");
-
-int i915_vbt_sdvo_panel_type __read_mostly = -1;
-module_param_named(vbt_sdvo_panel_type, i915_vbt_sdvo_panel_type, int, 0600);
-MODULE_PARM_DESC(vbt_sdvo_panel_type,
-		"Override/Ignore selection of SDVO panel mode in the VBT "
-		"(-2=ignore, -1=auto [default], index in VBT BIOS table)");
-
-static bool i915_try_reset __read_mostly = true;
-module_param_named(reset, i915_try_reset, bool, 0600);
-MODULE_PARM_DESC(reset, "Attempt GPU resets (default: true)");
-
-bool i915_enable_hangcheck __read_mostly = true;
-module_param_named(enable_hangcheck, i915_enable_hangcheck, bool, 0644);
-MODULE_PARM_DESC(enable_hangcheck,
-		"Periodically check GPU activity for detecting hangs. "
-		"WARNING: Disabling this can cause system wide hangs. "
-		"(default: true)");
-
-int i915_enable_ppgtt __read_mostly = -1;
-module_param_named(i915_enable_ppgtt, i915_enable_ppgtt, int, 0600);
-MODULE_PARM_DESC(i915_enable_ppgtt,
-		"Enable PPGTT (default: true)");
-
-int i915_enable_psr __read_mostly = 0;
-module_param_named(enable_psr, i915_enable_psr, int, 0600);
-MODULE_PARM_DESC(enable_psr, "Enable PSR (default: false)");
-
-unsigned int i915_preliminary_hw_support __read_mostly = IS_ENABLED(CONFIG_DRM_I915_PRELIMINARY_HW_SUPPORT);
-module_param_named(preliminary_hw_support, i915_preliminary_hw_support, int, 0600);
-MODULE_PARM_DESC(preliminary_hw_support,
-		"Enable preliminary hardware support.");
-
-int i915_disable_power_well __read_mostly = 1;
-module_param_named(disable_power_well, i915_disable_power_well, int, 0600);
-MODULE_PARM_DESC(disable_power_well,
-		 "Disable the power well when possible (default: true)");
-
-int i915_enable_ips __read_mostly = 1;
-module_param_named(enable_ips, i915_enable_ips, int, 0600);
-MODULE_PARM_DESC(enable_ips, "Enable IPS (default: true)");
-
-bool i915_fastboot __read_mostly = 0;
-module_param_named(fastboot, i915_fastboot, bool, 0600);
-MODULE_PARM_DESC(fastboot, "Try to skip unnecessary mode sets at boot time "
-		 "(default: false)");
-
-int i915_enable_pc8 __read_mostly = 1;
-module_param_named(enable_pc8, i915_enable_pc8, int, 0600);
-MODULE_PARM_DESC(enable_pc8, "Enable support for low power package C states (PC8+) (default: true)");
-
-int i915_pc8_timeout __read_mostly = 5000;
-module_param_named(pc8_timeout, i915_pc8_timeout, int, 0600);
-MODULE_PARM_DESC(pc8_timeout, "Number of msecs of idleness required to enter PC8+ (default: 5000)");
-
-bool i915_prefault_disable __read_mostly;
-module_param_named(prefault_disable, i915_prefault_disable, bool, 0600);
-MODULE_PARM_DESC(prefault_disable,
-		"Disable page prefaulting for pread/pwrite/reloc (default:false). For developers only.");
-
 static struct drm_driver driver;
-extern int intel_agp_enabled;
+
+#define GEN_DEFAULT_PIPEOFFSETS \
+	.pipe_offsets = { PIPE_A_OFFSET, PIPE_B_OFFSET, \
+			  PIPE_C_OFFSET, PIPE_EDP_OFFSET }, \
+	.trans_offsets = { TRANSCODER_A_OFFSET, TRANSCODER_B_OFFSET, \
+			   TRANSCODER_C_OFFSET, TRANSCODER_EDP_OFFSET }, \
+	.dpll_offsets = { DPLL_A_OFFSET, DPLL_B_OFFSET }, \
+	.dpll_md_offsets = { DPLL_A_MD_OFFSET, DPLL_B_MD_OFFSET }, \
+	.palette_offsets = { PALETTE_A_OFFSET, PALETTE_B_OFFSET }
+
+#define GEN_CHV_PIPEOFFSETS \
+	.pipe_offsets = { PIPE_A_OFFSET, PIPE_B_OFFSET, \
+			  CHV_PIPE_C_OFFSET }, \
+	.trans_offsets = { TRANSCODER_A_OFFSET, TRANSCODER_B_OFFSET, \
+			   CHV_TRANSCODER_C_OFFSET, }, \
+	.dpll_offsets = { DPLL_A_OFFSET, DPLL_B_OFFSET, \
+			  CHV_DPLL_C_OFFSET }, \
+	.dpll_md_offsets = { DPLL_A_MD_OFFSET, DPLL_B_MD_OFFSET, \
+			     CHV_DPLL_C_MD_OFFSET }, \
+	.palette_offsets = { PALETTE_A_OFFSET, PALETTE_B_OFFSET, \
+			     CHV_PALETTE_C_OFFSET }
+
+#define CURSOR_OFFSETS \
+	.cursor_offsets = { CURSOR_A_OFFSET, CURSOR_B_OFFSET, CHV_CURSOR_C_OFFSET }
+
+#define IVB_CURSOR_OFFSETS \
+	.cursor_offsets = { CURSOR_A_OFFSET, IVB_CURSOR_B_OFFSET, IVB_CURSOR_C_OFFSET }
 
 static const struct intel_device_info intel_i830_info = {
 	.gen = 2, .is_mobile = 1, .cursor_needs_physical = 1, .num_pipes = 2,
 	.has_overlay = 1, .overlay_needs_physical = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_845g_info = {
 	.gen = 2, .num_pipes = 1,
 	.has_overlay = 1, .overlay_needs_physical = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_i85x_info = {
 	.gen = 2, .is_i85x = 1, .is_mobile = 1, .num_pipes = 2,
 	.cursor_needs_physical = 1,
 	.has_overlay = 1, .overlay_needs_physical = 1,
+	.has_fbc = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_i865g_info = {
 	.gen = 2, .num_pipes = 1,
 	.has_overlay = 1, .overlay_needs_physical = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_i915g_info = {
 	.gen = 3, .is_i915g = 1, .cursor_needs_physical = 1, .num_pipes = 2,
 	.has_overlay = 1, .overlay_needs_physical = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 static const struct intel_device_info intel_i915gm_info = {
 	.gen = 3, .is_mobile = 1, .num_pipes = 2,
 	.cursor_needs_physical = 1,
 	.has_overlay = 1, .overlay_needs_physical = 1,
 	.supports_tv = 1,
+	.has_fbc = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 static const struct intel_device_info intel_i945g_info = {
 	.gen = 3, .has_hotplug = 1, .cursor_needs_physical = 1, .num_pipes = 2,
 	.has_overlay = 1, .overlay_needs_physical = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 static const struct intel_device_info intel_i945gm_info = {
 	.gen = 3, .is_i945gm = 1, .is_mobile = 1, .num_pipes = 2,
 	.has_hotplug = 1, .cursor_needs_physical = 1,
 	.has_overlay = 1, .overlay_needs_physical = 1,
 	.supports_tv = 1,
+	.has_fbc = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_i965g_info = {
@@ -212,6 +142,8 @@ static const struct intel_device_info intel_i965g_info = {
 	.has_hotplug = 1,
 	.has_overlay = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_i965gm_info = {
@@ -220,6 +152,8 @@ static const struct intel_device_info intel_i965gm_info = {
 	.has_overlay = 1,
 	.supports_tv = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_g33_info = {
@@ -227,12 +161,16 @@ static const struct intel_device_info intel_g33_info = {
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_overlay = 1,
 	.ring_mask = RENDER_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_g45_info = {
 	.gen = 4, .is_g4x = 1, .need_gfx_hws = 1, .num_pipes = 2,
 	.has_pipe_cxsr = 1, .has_hotplug = 1,
 	.ring_mask = RENDER_RING | BSD_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_gm45_info = {
@@ -241,18 +179,24 @@ static const struct intel_device_info intel_gm45_info = {
 	.has_pipe_cxsr = 1, .has_hotplug = 1,
 	.supports_tv = 1,
 	.ring_mask = RENDER_RING | BSD_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_pineview_info = {
 	.gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_overlay = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_ironlake_d_info = {
 	.gen = 5, .num_pipes = 2,
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.ring_mask = RENDER_RING | BSD_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_ironlake_m_info = {
@@ -260,13 +204,18 @@ static const struct intel_device_info intel_ironlake_m_info = {
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.has_fbc = 1,
 	.ring_mask = RENDER_RING | BSD_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_sandybridge_d_info = {
 	.gen = 6, .num_pipes = 2,
 	.need_gfx_hws = 1, .has_hotplug = 1,
+	.has_fbc = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING,
 	.has_llc = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_sandybridge_m_info = {
@@ -275,30 +224,38 @@ static const struct intel_device_info intel_sandybridge_m_info = {
 	.has_fbc = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING,
 	.has_llc = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 #define GEN7_FEATURES  \
 	.gen = 7, .num_pipes = 3, \
 	.need_gfx_hws = 1, .has_hotplug = 1, \
+	.has_fbc = 1, \
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING, \
 	.has_llc = 1
 
 static const struct intel_device_info intel_ivybridge_d_info = {
 	GEN7_FEATURES,
 	.is_ivybridge = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_ivybridge_m_info = {
 	GEN7_FEATURES,
 	.is_ivybridge = 1,
 	.is_mobile = 1,
-	.has_fbc = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_ivybridge_q_info = {
 	GEN7_FEATURES,
 	.is_ivybridge = 1,
 	.num_pipes = 0, /* legal, last one wins */
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_valleyview_m_info = {
@@ -307,7 +264,10 @@ static const struct intel_device_info intel_valleyview_m_info = {
 	.num_pipes = 2,
 	.is_valleyview = 1,
 	.display_mmio_offset = VLV_DISPLAY_BASE,
+	.has_fbc = 0, /* legal, last one wins */
 	.has_llc = 0, /* legal, last one wins */
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_valleyview_d_info = {
@@ -315,7 +275,10 @@ static const struct intel_device_info intel_valleyview_d_info = {
 	.num_pipes = 2,
 	.is_valleyview = 1,
 	.display_mmio_offset = VLV_DISPLAY_BASE,
+	.has_fbc = 0, /* legal, last one wins */
 	.has_llc = 0, /* legal, last one wins */
+	GEN_DEFAULT_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_haswell_d_info = {
@@ -324,6 +287,8 @@ static const struct intel_device_info intel_haswell_d_info = {
 	.has_ddi = 1,
 	.has_fpga_dbg = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_haswell_m_info = {
@@ -332,26 +297,64 @@ static const struct intel_device_info intel_haswell_m_info = {
 	.is_mobile = 1,
 	.has_ddi = 1,
 	.has_fpga_dbg = 1,
-	.has_fbc = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_broadwell_d_info = {
-	.is_preliminary = 1,
 	.gen = 8, .num_pipes = 3,
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
 	.has_llc = 1,
 	.has_ddi = 1,
+	.has_fbc = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
 };
 
 static const struct intel_device_info intel_broadwell_m_info = {
-	.is_preliminary = 1,
 	.gen = 8, .is_mobile = 1, .num_pipes = 3,
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
 	.has_llc = 1,
 	.has_ddi = 1,
+	.has_fbc = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
+};
+
+static const struct intel_device_info intel_broadwell_gt3d_info = {
+	.gen = 8, .num_pipes = 3,
+	.need_gfx_hws = 1, .has_hotplug = 1,
+	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
+	.has_llc = 1,
+	.has_ddi = 1,
+	.has_fbc = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
+};
+
+static const struct intel_device_info intel_broadwell_gt3m_info = {
+	.gen = 8, .is_mobile = 1, .num_pipes = 3,
+	.need_gfx_hws = 1, .has_hotplug = 1,
+	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
+	.has_llc = 1,
+	.has_ddi = 1,
+	.has_fbc = 1,
+	GEN_DEFAULT_PIPEOFFSETS,
+	IVB_CURSOR_OFFSETS,
+};
+
+static const struct intel_device_info intel_cherryview_info = {
+	.is_preliminary = 1,
+	.gen = 8, .num_pipes = 3,
+	.need_gfx_hws = 1, .has_hotplug = 1,
+	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
+	.is_valleyview = 1,
+	.display_mmio_offset = VLV_DISPLAY_BASE,
+	GEN_CHV_PIPEOFFSETS,
+	CURSOR_OFFSETS,
 };
 
 /*
@@ -386,8 +389,11 @@ static const struct intel_device_info intel_broadwell_m_info = {
 	INTEL_HSW_M_IDS(&intel_haswell_m_info), \
 	INTEL_VLV_M_IDS(&intel_valleyview_m_info),	\
 	INTEL_VLV_D_IDS(&intel_valleyview_d_info),	\
-	INTEL_BDW_M_IDS(&intel_broadwell_m_info),	\
-	INTEL_BDW_D_IDS(&intel_broadwell_d_info)
+	INTEL_BDW_GT12M_IDS(&intel_broadwell_m_info),	\
+	INTEL_BDW_GT12D_IDS(&intel_broadwell_d_info),	\
+	INTEL_BDW_GT3M_IDS(&intel_broadwell_gt3m_info),	\
+	INTEL_BDW_GT3D_IDS(&intel_broadwell_gt3d_info), \
+	INTEL_CHV_IDS(&intel_cherryview_info)
 
 static const struct pci_device_id pciidlist[] = {		/* aka */
 	INTEL_PCI_IDS,
@@ -401,7 +407,7 @@ MODULE_DEVICE_TABLE(pci, pciidlist);
 void intel_detect_pch(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct pci_dev *pch;
+	struct pci_dev *pch = NULL;
 
 	/* In all current cases, num_pipes is equivalent to the PCH_NOP setting
 	 * (which really amounts to a PCH but no South Display).
@@ -422,12 +428,9 @@ void intel_detect_pch(struct drm_device *dev)
 	 * all the ISA bridge devices and check for the first match, instead
 	 * of only checking the first one.
 	 */
-	pch = pci_get_class(PCI_CLASS_BRIDGE_ISA << 8, NULL);
-	while (pch) {
-		struct pci_dev *curr = pch;
+	while ((pch = pci_get_class(PCI_CLASS_BRIDGE_ISA << 8, pch))) {
 		if (pch->vendor == PCI_VENDOR_ID_INTEL) {
-			unsigned short id;
-			id = pch->device & INTEL_PCH_DEVICE_ID_MASK;
+			unsigned short id = pch->device & INTEL_PCH_DEVICE_ID_MASK;
 			dev_priv->pch_id = id;
 
 			if (id == INTEL_PCH_IBX_DEVICE_ID_TYPE) {
@@ -459,33 +462,29 @@ void intel_detect_pch(struct drm_device *dev)
 				DRM_DEBUG_KMS("Found LynxPoint LP PCH\n");
 				WARN_ON(!IS_HASWELL(dev));
 				WARN_ON(!IS_ULT(dev));
-			} else {
-				goto check_next;
-			}
-			pci_dev_put(pch);
+			} else
+				continue;
+
 			break;
 		}
-check_next:
-		pch = pci_get_class(PCI_CLASS_BRIDGE_ISA << 8, curr);
-		pci_dev_put(curr);
 	}
 	if (!pch)
-		DRM_DEBUG_KMS("No PCH found?\n");
+		DRM_DEBUG_KMS("No PCH found.\n");
+
+	pci_dev_put(pch);
 }
 
 bool i915_semaphore_is_enabled(struct drm_device *dev)
 {
 	if (INTEL_INFO(dev)->gen < 6)
-		return 0;
+		return false;
 
-	/* Until we get further testing... */
-	if (IS_GEN8(dev)) {
-		WARN_ON(!i915_preliminary_hw_support);
-		return 0;
-	}
+	if (i915.semaphores >= 0)
+		return i915.semaphores;
 
-	if (i915_semaphores >= 0)
-		return i915_semaphores;
+	/* Until we get further testing... */
+	if (IS_GEN8(dev))
+		return false;
 
 #ifdef CONFIG_INTEL_IOMMU
 	/* Enable semaphores on SNB when IO remapping is off */
@@ -493,7 +492,7 @@ bool i915_semaphore_is_enabled(struct drm_device *dev)
 		return false;
 #endif
 
-	return 1;
+	return true;
 }
 
 static int i915_drm_freeze(struct drm_device *dev)
@@ -501,6 +500,8 @@ static int i915_drm_freeze(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc;
 
+	intel_runtime_pm_get(dev_priv);
+
 	/* ignore lid events during suspend */
 	mutex_lock(&dev_priv->modeset_restore_lock);
 	dev_priv->modeset_restore = MODESET_SUSPENDED;
@@ -508,8 +509,7 @@ static int i915_drm_freeze(struct drm_device *dev)
 
 	/* We do a lot of poking in a lot of registers, make sure they work
 	 * properly. */
-	hsw_disable_package_c8(dev_priv);
-	intel_display_set_init_power(dev, true);
+	intel_display_set_init_power(dev_priv, true);
 
 	drm_kms_helper_poll_disable(dev);
 
@@ -526,18 +526,20 @@ static int i915_drm_freeze(struct drm_device *dev)
 			return error;
 		}
 
-		cancel_delayed_work_sync(&dev_priv->rps.delayed_resume_work);
-
 		drm_irq_uninstall(dev);
 		dev_priv->enable_hotplug_processing = false;
+
+		intel_disable_gt_powersave(dev);
+
 		/*
 		 * Disable CRTCs directly since we want to preserve sw state
 		 * for _thaw.
 		 */
-		mutex_lock(&dev->mode_config.mutex);
-		list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
+		drm_modeset_lock_all(dev);
+		for_each_crtc(dev, crtc) {
 			dev_priv->display.crtc_disable(crtc);
-		mutex_unlock(&dev->mode_config.mutex);
+		}
+		drm_modeset_unlock_all(dev);
 
 		intel_modeset_suspend_hw(dev);
 	}
@@ -547,11 +549,14 @@ static int i915_drm_freeze(struct drm_device *dev)
 	i915_save_state(dev);
 
 	intel_opregion_fini(dev);
+	intel_uncore_fini(dev);
 
 	console_lock();
 	intel_fbdev_set_suspend(dev, FBINFO_STATE_SUSPENDED);
 	console_unlock();
 
+	dev_priv->suspend_count++;
+
 	return 0;
 }
 
@@ -597,32 +602,20 @@ void intel_console_resume(struct work_struct *work)
 	console_unlock();
 }
 
-static void intel_resume_hotplug(struct drm_device *dev)
+static int i915_drm_thaw_early(struct drm_device *dev)
 {
-	struct drm_mode_config *mode_config = &dev->mode_config;
-	struct intel_encoder *encoder;
-
-	mutex_lock(&mode_config->mutex);
-	DRM_DEBUG_KMS("running encoder hotplug functions\n");
-
-	list_for_each_entry(encoder, &mode_config->encoder_list, base.head)
-		if (encoder->hot_plug)
-			encoder->hot_plug(encoder);
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	mutex_unlock(&mode_config->mutex);
+	intel_uncore_early_sanitize(dev);
+	intel_uncore_sanitize(dev);
+	intel_power_domains_init_hw(dev_priv);
 
-	/* Just fire off a uevent and let userspace tell us what to do */
-	drm_helper_hpd_irq_event(dev);
+	return 0;
 }
 
 static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int error = 0;
-
-	intel_uncore_early_sanitize(dev);
-
-	intel_uncore_sanitize(dev);
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET) &&
 	    restore_gtt_mappings) {
@@ -631,27 +624,27 @@ static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
 		mutex_unlock(&dev->struct_mutex);
 	}
 
-	intel_power_domains_init_hw(dev);
-
 	i915_restore_state(dev);
 	intel_opregion_setup(dev);
 
 	/* KMS EnterVT equivalent */
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		intel_init_pch_refclk(dev);
+		drm_mode_config_reset(dev);
 
 		mutex_lock(&dev->struct_mutex);
-
-		error = i915_gem_init_hw(dev);
+		if (i915_gem_init_hw(dev)) {
+			DRM_ERROR("failed to re-initialize GPU, declaring wedged!\n");
+			atomic_set_mask(I915_WEDGED, &dev_priv->gpu_error.reset_counter);
+		}
 		mutex_unlock(&dev->struct_mutex);
 
 		/* We need working interrupts for modeset enabling ... */
-		drm_irq_install(dev);
+		drm_irq_install(dev, dev->pdev->irq);
 
 		intel_modeset_init_hw(dev);
 
 		drm_modeset_lock_all(dev);
-		drm_mode_config_reset(dev);
 		intel_modeset_setup_hw_state(dev, true);
 		drm_modeset_unlock_all(dev);
 
@@ -664,7 +657,7 @@ static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
 		intel_hpd_init(dev);
 		dev_priv->enable_hotplug_processing = true;
 		/* Config may have changed between suspend and resume */
-		intel_resume_hotplug(dev);
+		drm_helper_hpd_irq_event(dev);
 	}
 
 	intel_opregion_init(dev);
@@ -681,14 +674,12 @@ static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
 		schedule_work(&dev_priv->console_resume_work);
 	}
 
-	/* Undo what we did at i915_drm_freeze so the refcount goes back to the
-	 * expected level. */
-	hsw_enable_package_c8(dev_priv);
-
 	mutex_lock(&dev_priv->modeset_restore_lock);
 	dev_priv->modeset_restore = MODESET_DONE;
 	mutex_unlock(&dev_priv->modeset_restore_lock);
-	return error;
+
+	intel_runtime_pm_put(dev_priv);
+	return 0;
 }
 
 static int i915_drm_thaw(struct drm_device *dev)
@@ -699,19 +690,33 @@ static int i915_drm_thaw(struct drm_device *dev)
 	return __i915_drm_thaw(dev, true);
 }
 
-int i915_resume(struct drm_device *dev)
+static int i915_resume_early(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int ret;
-
 	if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 		return 0;
 
+	/*
+	 * We have a resume ordering issue with the snd-hda driver also
+	 * requiring our device to be power up. Due to the lack of a
+	 * parent/child relationship we currently solve this with an early
+	 * resume hook.
+	 *
+	 * FIXME: This should be solved with a special hdmi sink device or
+	 * similar so that power domains can be employed.
+	 */
 	if (pci_enable_device(dev->pdev))
 		return -EIO;
 
 	pci_set_master(dev->pdev);
 
+	return i915_drm_thaw_early(dev);
+}
+
+int i915_resume(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
 	/*
 	 * Platforms with opregion should have sane BIOS, older ones (gen3 and
 	 * earlier) need to restore the GTT mappings since the BIOS might clear
@@ -725,6 +730,14 @@ int i915_resume(struct drm_device *dev)
 	return 0;
 }
 
+static int i915_resume_legacy(struct drm_device *dev)
+{
+	i915_resume_early(dev);
+	i915_resume(dev);
+
+	return 0;
+}
+
 /**
  * i915_reset - reset chip after a hang
  * @dev: drm device to reset
@@ -742,11 +755,11 @@ int i915_resume(struct drm_device *dev)
  */
 int i915_reset(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	bool simulated;
 	int ret;
 
-	if (!i915_try_reset)
+	if (!i915.reset)
 		return 0;
 
 	mutex_lock(&dev->struct_mutex);
@@ -762,14 +775,14 @@ int i915_reset(struct drm_device *dev)
 		DRM_INFO("Simulated gpu hang, resetting stop_rings\n");
 		dev_priv->gpu_error.stop_rings = 0;
 		if (ret == -ENODEV) {
-			DRM_ERROR("Reset not implemented, but ignoring "
-				  "error for simulated gpu hangs\n");
+			DRM_INFO("Reset not implemented, but ignoring "
+				 "error for simulated gpu hangs\n");
 			ret = 0;
 		}
 	}
 
 	if (ret) {
-		DRM_ERROR("Failed to reset chip.\n");
+		DRM_ERROR("Failed to reset chip: %i\n", ret);
 		mutex_unlock(&dev->struct_mutex);
 		return ret;
 	}
@@ -790,20 +803,30 @@ int i915_reset(struct drm_device *dev)
 	 */
 	if (drm_core_check_feature(dev, DRIVER_MODESET) ||
 			!dev_priv->ums.mm_suspended) {
-		bool hw_contexts_disabled = dev_priv->hw_contexts_disabled;
 		dev_priv->ums.mm_suspended = 0;
 
 		ret = i915_gem_init_hw(dev);
-		if (!hw_contexts_disabled && dev_priv->hw_contexts_disabled)
-			DRM_ERROR("HW contexts didn't survive reset\n");
 		mutex_unlock(&dev->struct_mutex);
 		if (ret) {
 			DRM_ERROR("Failed hw init on reset %d\n", ret);
 			return ret;
 		}
 
-		drm_irq_uninstall(dev);
-		drm_irq_install(dev);
+		/*
+		 * FIXME: This races pretty badly against concurrent holders of
+		 * ring interrupts. This is possible since we've started to drop
+		 * dev->struct_mutex in select places when waiting for the gpu.
+		 */
+
+		/*
+		 * rps/rc6 re-init is necessary to restore state lost after the
+		 * reset and the re-install of gt irqs. Skip for ironlake per
+		 * previous concerns that it doesn't respond well to some forms
+		 * of re-init after reset.
+		 */
+		if (INTEL_INFO(dev)->gen > 5)
+			intel_reset_gt_powersave(dev);
+
 		intel_hpd_init(dev);
 	} else {
 		mutex_unlock(&dev->struct_mutex);
@@ -817,7 +840,7 @@ static int i915_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	struct intel_device_info *intel_info =
 		(struct intel_device_info *) ent->driver_data;
 
-	if (IS_PRELIMINARY_HW(intel_info) && !i915_preliminary_hw_support) {
+	if (IS_PRELIMINARY_HW(intel_info) && !i915.preliminary_hw_support) {
 		DRM_INFO("This hardware requires preliminary hardware support.\n"
 			 "See CONFIG_DRM_I915_PRELIMINARY_HW_SUPPORT, and/or modparam preliminary_hw_support\n");
 		return -ENODEV;
@@ -831,17 +854,7 @@ static int i915_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (PCI_FUNC(pdev->devfn))
 		return -ENODEV;
 
-	/* We've managed to ship a kms-enabled ddx that shipped with an XvMC
-	 * implementation for gen3 (and only gen3) that used legacy drm maps
-	 * (gasp!) to share buffers between X and the client. Hence we need to
-	 * keep around the fake agp stuff for gen3, even when kms is enabled. */
-	if (intel_info->gen != 3) {
-		driver.driver_features &=
-			~(DRIVER_USE_AGP | DRIVER_REQUIRE_AGP);
-	} else if (!intel_agp_enabled) {
-		DRM_ERROR("drm/i915 can't work without intel_agp module!\n");
-		return -ENODEV;
-	}
+	driver.driver_features &= ~(DRIVER_USE_AGP);
 
 	return drm_get_pci_dev(pdev, ent, &driver);
 }
@@ -858,7 +871,6 @@ static int i915_pm_suspend(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
-	int error;
 
 	if (!drm_dev || !drm_dev->dev_private) {
 		dev_err(dev, "DRM not initialized, aborting suspend.\n");
@@ -868,9 +880,25 @@ static int i915_pm_suspend(struct device *dev)
 	if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
 		return 0;
 
-	error = i915_drm_freeze(drm_dev);
-	if (error)
-		return error;
+	return i915_drm_freeze(drm_dev);
+}
+
+static int i915_pm_suspend_late(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+
+	/*
+	 * We have a suspedn ordering issue with the snd-hda driver also
+	 * requiring our device to be power up. Due to the lack of a
+	 * parent/child relationship we currently solve this with an late
+	 * suspend hook.
+	 *
+	 * FIXME: This should be solved with a special hdmi sink device or
+	 * similar so that power domains can be employed.
+	 */
+	if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+		return 0;
 
 	pci_disable_device(pdev);
 	pci_set_power_state(pdev, PCI_D3hot);
@@ -878,6 +906,14 @@ static int i915_pm_suspend(struct device *dev)
 	return 0;
 }
 
+static int i915_pm_resume_early(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+
+	return i915_resume_early(drm_dev);
+}
+
 static int i915_pm_resume(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
@@ -899,6 +935,14 @@ static int i915_pm_freeze(struct device *dev)
 	return i915_drm_freeze(drm_dev);
 }
 
+static int i915_pm_thaw_early(struct device *dev)
+{
+	struct pci_dev *pdev = to_pci_dev(dev);
+	struct drm_device *drm_dev = pci_get_drvdata(pdev);
+
+	return i915_drm_thaw_early(drm_dev);
+}
+
 static int i915_pm_thaw(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
@@ -915,13 +959,526 @@ static int i915_pm_poweroff(struct device *dev)
 	return i915_drm_freeze(drm_dev);
 }
 
+static int hsw_runtime_suspend(struct drm_i915_private *dev_priv)
+{
+	hsw_enable_pc8(dev_priv);
+
+	return 0;
+}
+
+static int snb_runtime_resume(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+
+	intel_init_pch_refclk(dev);
+
+	return 0;
+}
+
+static int hsw_runtime_resume(struct drm_i915_private *dev_priv)
+{
+	hsw_disable_pc8(dev_priv);
+
+	return 0;
+}
+
+/*
+ * Save all Gunit registers that may be lost after a D3 and a subsequent
+ * S0i[R123] transition. The list of registers needing a save/restore is
+ * defined in the VLV2_S0IXRegs document. This documents marks all Gunit
+ * registers in the following way:
+ * - Driver: saved/restored by the driver
+ * - Punit : saved/restored by the Punit firmware
+ * - No, w/o marking: no need to save/restore, since the register is R/O or
+ *                    used internally by the HW in a way that doesn't depend
+ *                    keeping the content across a suspend/resume.
+ * - Debug : used for debugging
+ *
+ * We save/restore all registers marked with 'Driver', with the following
+ * exceptions:
+ * - Registers out of use, including also registers marked with 'Debug'.
+ *   These have no effect on the driver's operation, so we don't save/restore
+ *   them to reduce the overhead.
+ * - Registers that are fully setup by an initialization function called from
+ *   the resume path. For example many clock gating and RPS/RC6 registers.
+ * - Registers that provide the right functionality with their reset defaults.
+ *
+ * TODO: Except for registers that based on the above 3 criteria can be safely
+ * ignored, we save/restore all others, practically treating the HW context as
+ * a black-box for the driver. Further investigation is needed to reduce the
+ * saved/restored registers even further, by following the same 3 criteria.
+ */
+static void vlv_save_gunit_s0ix_state(struct drm_i915_private *dev_priv)
+{
+	struct vlv_s0ix_state *s = &dev_priv->vlv_s0ix_state;
+	int i;
+
+	/* GAM 0x4000-0x4770 */
+	s->wr_watermark		= I915_READ(GEN7_WR_WATERMARK);
+	s->gfx_prio_ctrl	= I915_READ(GEN7_GFX_PRIO_CTRL);
+	s->arb_mode		= I915_READ(ARB_MODE);
+	s->gfx_pend_tlb0	= I915_READ(GEN7_GFX_PEND_TLB0);
+	s->gfx_pend_tlb1	= I915_READ(GEN7_GFX_PEND_TLB1);
+
+	for (i = 0; i < ARRAY_SIZE(s->lra_limits); i++)
+		s->lra_limits[i] = I915_READ(GEN7_LRA_LIMITS_BASE + i * 4);
+
+	s->media_max_req_count	= I915_READ(GEN7_MEDIA_MAX_REQ_COUNT);
+	s->gfx_max_req_count	= I915_READ(GEN7_MEDIA_MAX_REQ_COUNT);
+
+	s->render_hwsp		= I915_READ(RENDER_HWS_PGA_GEN7);
+	s->ecochk		= I915_READ(GAM_ECOCHK);
+	s->bsd_hwsp		= I915_READ(BSD_HWS_PGA_GEN7);
+	s->blt_hwsp		= I915_READ(BLT_HWS_PGA_GEN7);
+
+	s->tlb_rd_addr		= I915_READ(GEN7_TLB_RD_ADDR);
+
+	/* MBC 0x9024-0x91D0, 0x8500 */
+	s->g3dctl		= I915_READ(VLV_G3DCTL);
+	s->gsckgctl		= I915_READ(VLV_GSCKGCTL);
+	s->mbctl		= I915_READ(GEN6_MBCTL);
+
+	/* GCP 0x9400-0x9424, 0x8100-0x810C */
+	s->ucgctl1		= I915_READ(GEN6_UCGCTL1);
+	s->ucgctl3		= I915_READ(GEN6_UCGCTL3);
+	s->rcgctl1		= I915_READ(GEN6_RCGCTL1);
+	s->rcgctl2		= I915_READ(GEN6_RCGCTL2);
+	s->rstctl		= I915_READ(GEN6_RSTCTL);
+	s->misccpctl		= I915_READ(GEN7_MISCCPCTL);
+
+	/* GPM 0xA000-0xAA84, 0x8000-0x80FC */
+	s->gfxpause		= I915_READ(GEN6_GFXPAUSE);
+	s->rpdeuhwtc		= I915_READ(GEN6_RPDEUHWTC);
+	s->rpdeuc		= I915_READ(GEN6_RPDEUC);
+	s->ecobus		= I915_READ(ECOBUS);
+	s->pwrdwnupctl		= I915_READ(VLV_PWRDWNUPCTL);
+	s->rp_down_timeout	= I915_READ(GEN6_RP_DOWN_TIMEOUT);
+	s->rp_deucsw		= I915_READ(GEN6_RPDEUCSW);
+	s->rcubmabdtmr		= I915_READ(GEN6_RCUBMABDTMR);
+	s->rcedata		= I915_READ(VLV_RCEDATA);
+	s->spare2gh		= I915_READ(VLV_SPAREG2H);
+
+	/* Display CZ domain, 0x4400C-0x4402C, 0x4F000-0x4F11F */
+	s->gt_imr		= I915_READ(GTIMR);
+	s->gt_ier		= I915_READ(GTIER);
+	s->pm_imr		= I915_READ(GEN6_PMIMR);
+	s->pm_ier		= I915_READ(GEN6_PMIER);
+
+	for (i = 0; i < ARRAY_SIZE(s->gt_scratch); i++)
+		s->gt_scratch[i] = I915_READ(GEN7_GT_SCRATCH_BASE + i * 4);
+
+	/* GT SA CZ domain, 0x100000-0x138124 */
+	s->tilectl		= I915_READ(TILECTL);
+	s->gt_fifoctl		= I915_READ(GTFIFOCTL);
+	s->gtlc_wake_ctrl	= I915_READ(VLV_GTLC_WAKE_CTRL);
+	s->gtlc_survive		= I915_READ(VLV_GTLC_SURVIVABILITY_REG);
+	s->pmwgicz		= I915_READ(VLV_PMWGICZ);
+
+	/* Gunit-Display CZ domain, 0x182028-0x1821CF */
+	s->gu_ctl0		= I915_READ(VLV_GU_CTL0);
+	s->gu_ctl1		= I915_READ(VLV_GU_CTL1);
+	s->clock_gate_dis2	= I915_READ(VLV_GUNIT_CLOCK_GATE2);
+
+	/*
+	 * Not saving any of:
+	 * DFT,		0x9800-0x9EC0
+	 * SARB,	0xB000-0xB1FC
+	 * GAC,		0x5208-0x524C, 0x14000-0x14C000
+	 * PCI CFG
+	 */
+}
+
+static void vlv_restore_gunit_s0ix_state(struct drm_i915_private *dev_priv)
+{
+	struct vlv_s0ix_state *s = &dev_priv->vlv_s0ix_state;
+	u32 val;
+	int i;
+
+	/* GAM 0x4000-0x4770 */
+	I915_WRITE(GEN7_WR_WATERMARK,	s->wr_watermark);
+	I915_WRITE(GEN7_GFX_PRIO_CTRL,	s->gfx_prio_ctrl);
+	I915_WRITE(ARB_MODE,		s->arb_mode | (0xffff << 16));
+	I915_WRITE(GEN7_GFX_PEND_TLB0,	s->gfx_pend_tlb0);
+	I915_WRITE(GEN7_GFX_PEND_TLB1,	s->gfx_pend_tlb1);
+
+	for (i = 0; i < ARRAY_SIZE(s->lra_limits); i++)
+		I915_WRITE(GEN7_LRA_LIMITS_BASE + i * 4, s->lra_limits[i]);
+
+	I915_WRITE(GEN7_MEDIA_MAX_REQ_COUNT, s->media_max_req_count);
+	I915_WRITE(GEN7_MEDIA_MAX_REQ_COUNT, s->gfx_max_req_count);
+
+	I915_WRITE(RENDER_HWS_PGA_GEN7,	s->render_hwsp);
+	I915_WRITE(GAM_ECOCHK,		s->ecochk);
+	I915_WRITE(BSD_HWS_PGA_GEN7,	s->bsd_hwsp);
+	I915_WRITE(BLT_HWS_PGA_GEN7,	s->blt_hwsp);
+
+	I915_WRITE(GEN7_TLB_RD_ADDR,	s->tlb_rd_addr);
+
+	/* MBC 0x9024-0x91D0, 0x8500 */
+	I915_WRITE(VLV_G3DCTL,		s->g3dctl);
+	I915_WRITE(VLV_GSCKGCTL,	s->gsckgctl);
+	I915_WRITE(GEN6_MBCTL,		s->mbctl);
+
+	/* GCP 0x9400-0x9424, 0x8100-0x810C */
+	I915_WRITE(GEN6_UCGCTL1,	s->ucgctl1);
+	I915_WRITE(GEN6_UCGCTL3,	s->ucgctl3);
+	I915_WRITE(GEN6_RCGCTL1,	s->rcgctl1);
+	I915_WRITE(GEN6_RCGCTL2,	s->rcgctl2);
+	I915_WRITE(GEN6_RSTCTL,		s->rstctl);
+	I915_WRITE(GEN7_MISCCPCTL,	s->misccpctl);
+
+	/* GPM 0xA000-0xAA84, 0x8000-0x80FC */
+	I915_WRITE(GEN6_GFXPAUSE,	s->gfxpause);
+	I915_WRITE(GEN6_RPDEUHWTC,	s->rpdeuhwtc);
+	I915_WRITE(GEN6_RPDEUC,		s->rpdeuc);
+	I915_WRITE(ECOBUS,		s->ecobus);
+	I915_WRITE(VLV_PWRDWNUPCTL,	s->pwrdwnupctl);
+	I915_WRITE(GEN6_RP_DOWN_TIMEOUT,s->rp_down_timeout);
+	I915_WRITE(GEN6_RPDEUCSW,	s->rp_deucsw);
+	I915_WRITE(GEN6_RCUBMABDTMR,	s->rcubmabdtmr);
+	I915_WRITE(VLV_RCEDATA,		s->rcedata);
+	I915_WRITE(VLV_SPAREG2H,	s->spare2gh);
+
+	/* Display CZ domain, 0x4400C-0x4402C, 0x4F000-0x4F11F */
+	I915_WRITE(GTIMR,		s->gt_imr);
+	I915_WRITE(GTIER,		s->gt_ier);
+	I915_WRITE(GEN6_PMIMR,		s->pm_imr);
+	I915_WRITE(GEN6_PMIER,		s->pm_ier);
+
+	for (i = 0; i < ARRAY_SIZE(s->gt_scratch); i++)
+		I915_WRITE(GEN7_GT_SCRATCH_BASE + i * 4, s->gt_scratch[i]);
+
+	/* GT SA CZ domain, 0x100000-0x138124 */
+	I915_WRITE(TILECTL,			s->tilectl);
+	I915_WRITE(GTFIFOCTL,			s->gt_fifoctl);
+	/*
+	 * Preserve the GT allow wake and GFX force clock bit, they are not
+	 * be restored, as they are used to control the s0ix suspend/resume
+	 * sequence by the caller.
+	 */
+	val = I915_READ(VLV_GTLC_WAKE_CTRL);
+	val &= VLV_GTLC_ALLOWWAKEREQ;
+	val |= s->gtlc_wake_ctrl & ~VLV_GTLC_ALLOWWAKEREQ;
+	I915_WRITE(VLV_GTLC_WAKE_CTRL, val);
+
+	val = I915_READ(VLV_GTLC_SURVIVABILITY_REG);
+	val &= VLV_GFX_CLK_FORCE_ON_BIT;
+	val |= s->gtlc_survive & ~VLV_GFX_CLK_FORCE_ON_BIT;
+	I915_WRITE(VLV_GTLC_SURVIVABILITY_REG, val);
+
+	I915_WRITE(VLV_PMWGICZ,			s->pmwgicz);
+
+	/* Gunit-Display CZ domain, 0x182028-0x1821CF */
+	I915_WRITE(VLV_GU_CTL0,			s->gu_ctl0);
+	I915_WRITE(VLV_GU_CTL1,			s->gu_ctl1);
+	I915_WRITE(VLV_GUNIT_CLOCK_GATE2,	s->clock_gate_dis2);
+}
+
+int vlv_force_gfx_clock(struct drm_i915_private *dev_priv, bool force_on)
+{
+	u32 val;
+	int err;
+
+	val = I915_READ(VLV_GTLC_SURVIVABILITY_REG);
+	WARN_ON(!!(val & VLV_GFX_CLK_FORCE_ON_BIT) == force_on);
+
+#define COND (I915_READ(VLV_GTLC_SURVIVABILITY_REG) & VLV_GFX_CLK_STATUS_BIT)
+	/* Wait for a previous force-off to settle */
+	if (force_on) {
+		err = wait_for(!COND, 20);
+		if (err) {
+			DRM_ERROR("timeout waiting for GFX clock force-off (%08x)\n",
+				  I915_READ(VLV_GTLC_SURVIVABILITY_REG));
+			return err;
+		}
+	}
+
+	val = I915_READ(VLV_GTLC_SURVIVABILITY_REG);
+	val &= ~VLV_GFX_CLK_FORCE_ON_BIT;
+	if (force_on)
+		val |= VLV_GFX_CLK_FORCE_ON_BIT;
+	I915_WRITE(VLV_GTLC_SURVIVABILITY_REG, val);
+
+	if (!force_on)
+		return 0;
+
+	err = wait_for(COND, 20);
+	if (err)
+		DRM_ERROR("timeout waiting for GFX clock force-on (%08x)\n",
+			  I915_READ(VLV_GTLC_SURVIVABILITY_REG));
+
+	return err;
+#undef COND
+}
+
+static int vlv_allow_gt_wake(struct drm_i915_private *dev_priv, bool allow)
+{
+	u32 val;
+	int err = 0;
+
+	val = I915_READ(VLV_GTLC_WAKE_CTRL);
+	val &= ~VLV_GTLC_ALLOWWAKEREQ;
+	if (allow)
+		val |= VLV_GTLC_ALLOWWAKEREQ;
+	I915_WRITE(VLV_GTLC_WAKE_CTRL, val);
+	POSTING_READ(VLV_GTLC_WAKE_CTRL);
+
+#define COND (!!(I915_READ(VLV_GTLC_PW_STATUS) & VLV_GTLC_ALLOWWAKEACK) == \
+	      allow)
+	err = wait_for(COND, 1);
+	if (err)
+		DRM_ERROR("timeout disabling GT waking\n");
+	return err;
+#undef COND
+}
+
+static int vlv_wait_for_gt_wells(struct drm_i915_private *dev_priv,
+				 bool wait_for_on)
+{
+	u32 mask;
+	u32 val;
+	int err;
+
+	mask = VLV_GTLC_PW_MEDIA_STATUS_MASK | VLV_GTLC_PW_RENDER_STATUS_MASK;
+	val = wait_for_on ? mask : 0;
+#define COND ((I915_READ(VLV_GTLC_PW_STATUS) & mask) == val)
+	if (COND)
+		return 0;
+
+	DRM_DEBUG_KMS("waiting for GT wells to go %s (%08x)\n",
+			wait_for_on ? "on" : "off",
+			I915_READ(VLV_GTLC_PW_STATUS));
+
+	/*
+	 * RC6 transitioning can be delayed up to 2 msec (see
+	 * valleyview_enable_rps), use 3 msec for safety.
+	 */
+	err = wait_for(COND, 3);
+	if (err)
+		DRM_ERROR("timeout waiting for GT wells to go %s\n",
+			  wait_for_on ? "on" : "off");
+
+	return err;
+#undef COND
+}
+
+static void vlv_check_no_gt_access(struct drm_i915_private *dev_priv)
+{
+	if (!(I915_READ(VLV_GTLC_PW_STATUS) & VLV_GTLC_ALLOWWAKEERR))
+		return;
+
+	DRM_ERROR("GT register access while GT waking disabled\n");
+	I915_WRITE(VLV_GTLC_PW_STATUS, VLV_GTLC_ALLOWWAKEERR);
+}
+
+static int vlv_runtime_suspend(struct drm_i915_private *dev_priv)
+{
+	u32 mask;
+	int err;
+
+	/*
+	 * Bspec defines the following GT well on flags as debug only, so
+	 * don't treat them as hard failures.
+	 */
+	(void)vlv_wait_for_gt_wells(dev_priv, false);
+
+	mask = VLV_GTLC_RENDER_CTX_EXISTS | VLV_GTLC_MEDIA_CTX_EXISTS;
+	WARN_ON((I915_READ(VLV_GTLC_WAKE_CTRL) & mask) != mask);
+
+	vlv_check_no_gt_access(dev_priv);
+
+	err = vlv_force_gfx_clock(dev_priv, true);
+	if (err)
+		goto err1;
+
+	err = vlv_allow_gt_wake(dev_priv, false);
+	if (err)
+		goto err2;
+	vlv_save_gunit_s0ix_state(dev_priv);
+
+	err = vlv_force_gfx_clock(dev_priv, false);
+	if (err)
+		goto err2;
+
+	return 0;
+
+err2:
+	/* For safety always re-enable waking and disable gfx clock forcing */
+	vlv_allow_gt_wake(dev_priv, true);
+err1:
+	vlv_force_gfx_clock(dev_priv, false);
+
+	return err;
+}
+
+static int vlv_runtime_resume(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	int err;
+	int ret;
+
+	/*
+	 * If any of the steps fail just try to continue, that's the best we
+	 * can do at this point. Return the first error code (which will also
+	 * leave RPM permanently disabled).
+	 */
+	ret = vlv_force_gfx_clock(dev_priv, true);
+
+	vlv_restore_gunit_s0ix_state(dev_priv);
+
+	err = vlv_allow_gt_wake(dev_priv, true);
+	if (!ret)
+		ret = err;
+
+	err = vlv_force_gfx_clock(dev_priv, false);
+	if (!ret)
+		ret = err;
+
+	vlv_check_no_gt_access(dev_priv);
+
+	intel_init_clock_gating(dev);
+	i915_gem_restore_fences(dev);
+
+	return ret;
+}
+
+static int intel_runtime_suspend(struct device *device)
+{
+	struct pci_dev *pdev = to_pci_dev(device);
+	struct drm_device *dev = pci_get_drvdata(pdev);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	if (WARN_ON_ONCE(!(dev_priv->rps.enabled && intel_enable_rc6(dev))))
+		return -ENODEV;
+
+	WARN_ON(!HAS_RUNTIME_PM(dev));
+	assert_force_wake_inactive(dev_priv);
+
+	DRM_DEBUG_KMS("Suspending device\n");
+
+	/*
+	 * We could deadlock here in case another thread holding struct_mutex
+	 * calls RPM suspend concurrently, since the RPM suspend will wait
+	 * first for this RPM suspend to finish. In this case the concurrent
+	 * RPM resume will be followed by its RPM suspend counterpart. Still
+	 * for consistency return -EAGAIN, which will reschedule this suspend.
+	 */
+	if (!mutex_trylock(&dev->struct_mutex)) {
+		DRM_DEBUG_KMS("device lock contention, deffering suspend\n");
+		/*
+		 * Bump the expiration timestamp, otherwise the suspend won't
+		 * be rescheduled.
+		 */
+		pm_runtime_mark_last_busy(device);
+
+		return -EAGAIN;
+	}
+	/*
+	 * We are safe here against re-faults, since the fault handler takes
+	 * an RPM reference.
+	 */
+	i915_gem_release_all_mmaps(dev_priv);
+	mutex_unlock(&dev->struct_mutex);
+
+	/*
+	 * rps.work can't be rearmed here, since we get here only after making
+	 * sure the GPU is idle and the RPS freq is set to the minimum. See
+	 * intel_mark_idle().
+	 */
+	cancel_work_sync(&dev_priv->rps.work);
+	intel_runtime_pm_disable_interrupts(dev);
+
+	if (IS_GEN6(dev)) {
+		ret = 0;
+	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+		ret = hsw_runtime_suspend(dev_priv);
+	} else if (IS_VALLEYVIEW(dev)) {
+		ret = vlv_runtime_suspend(dev_priv);
+	} else {
+		ret = -ENODEV;
+		WARN_ON(1);
+	}
+
+	if (ret) {
+		DRM_ERROR("Runtime suspend failed, disabling it (%d)\n", ret);
+		intel_runtime_pm_restore_interrupts(dev);
+
+		return ret;
+	}
+
+	del_timer_sync(&dev_priv->gpu_error.hangcheck_timer);
+	dev_priv->pm.suspended = true;
+
+	/*
+	 * current versions of firmware which depend on this opregion
+	 * notification have repurposed the D1 definition to mean
+	 * "runtime suspended" vs. what you would normally expect (D3)
+	 * to distinguish it from notifications that might be sent
+	 * via the suspend path.
+	 */
+	intel_opregion_notify_adapter(dev, PCI_D1);
+
+	DRM_DEBUG_KMS("Device suspended\n");
+	return 0;
+}
+
+static int intel_runtime_resume(struct device *device)
+{
+	struct pci_dev *pdev = to_pci_dev(device);
+	struct drm_device *dev = pci_get_drvdata(pdev);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	WARN_ON(!HAS_RUNTIME_PM(dev));
+
+	DRM_DEBUG_KMS("Resuming device\n");
+
+	intel_opregion_notify_adapter(dev, PCI_D0);
+	dev_priv->pm.suspended = false;
+
+	if (IS_GEN6(dev)) {
+		ret = snb_runtime_resume(dev_priv);
+	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+		ret = hsw_runtime_resume(dev_priv);
+	} else if (IS_VALLEYVIEW(dev)) {
+		ret = vlv_runtime_resume(dev_priv);
+	} else {
+		WARN_ON(1);
+		ret = -ENODEV;
+	}
+
+	/*
+	 * No point of rolling back things in case of an error, as the best
+	 * we can do is to hope that things will still work (and disable RPM).
+	 */
+	i915_gem_init_swizzling(dev);
+	gen6_update_ring_freq(dev);
+
+	intel_runtime_pm_restore_interrupts(dev);
+	intel_reset_gt_powersave(dev);
+
+	if (ret)
+		DRM_ERROR("Runtime resume failed, disabling it (%d)\n", ret);
+	else
+		DRM_DEBUG_KMS("Device resumed\n");
+
+	return ret;
+}
+
 static const struct dev_pm_ops i915_pm_ops = {
 	.suspend = i915_pm_suspend,
+	.suspend_late = i915_pm_suspend_late,
+	.resume_early = i915_pm_resume_early,
 	.resume = i915_pm_resume,
 	.freeze = i915_pm_freeze,
+	.thaw_early = i915_pm_thaw_early,
 	.thaw = i915_pm_thaw,
 	.poweroff = i915_pm_poweroff,
+	.restore_early = i915_pm_resume_early,
 	.restore = i915_pm_resume,
+	.runtime_suspend = intel_runtime_suspend,
+	.runtime_resume = intel_runtime_resume,
 };
 
 static const struct vm_operations_struct i915_gem_vm_ops = {
@@ -949,7 +1506,7 @@ static struct drm_driver driver = {
 	 * deal with them for Intel hardware.
 	 */
 	.driver_features =
-	    DRIVER_USE_AGP | DRIVER_REQUIRE_AGP |
+	    DRIVER_USE_AGP |
 	    DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED | DRIVER_GEM | DRIVER_PRIME |
 	    DRIVER_RENDER,
 	.load = i915_driver_load,
@@ -961,7 +1518,7 @@ static struct drm_driver driver = {
 
 	/* Used in place of i915_pm_ops for non-DRIVER_MODESET */
 	.suspend = i915_suspend,
-	.resume = i915_resume,
+	.resume = i915_resume_legacy,
 
 	.device_is_agp = i915_driver_device_is_agp,
 	.master_create = i915_master_create,
@@ -1013,25 +1570,36 @@ static int __init i915_init(void)
 	 * the default behavior.
 	 */
 #if defined(CONFIG_DRM_I915_KMS)
-	if (i915_modeset != 0)
+	if (i915.modeset != 0)
 		driver.driver_features |= DRIVER_MODESET;
 #endif
-	if (i915_modeset == 1)
+	if (i915.modeset == 1)
 		driver.driver_features |= DRIVER_MODESET;
 
 #ifdef CONFIG_VGA_CONSOLE
-	if (vgacon_text_force() && i915_modeset == -1)
+	if (vgacon_text_force() && i915.modeset == -1)
 		driver.driver_features &= ~DRIVER_MODESET;
 #endif
 
-	if (!(driver.driver_features & DRIVER_MODESET))
+	if (!(driver.driver_features & DRIVER_MODESET)) {
 		driver.get_vblank_timestamp = NULL;
+#ifndef CONFIG_DRM_I915_UMS
+		/* Silently fail loading to not upset userspace. */
+		DRM_DEBUG_DRIVER("KMS and UMS disabled.\n");
+		return 0;
+#endif
+	}
 
 	return drm_pci_init(&driver, &i915_pci_driver);
 }
 
 static void __exit i915_exit(void)
 {
+#ifndef CONFIG_DRM_I915_UMS
+	if (!(driver.driver_features & DRIVER_MODESET))
+		return; /* Never loaded a driver. */
+#endif
+
 	drm_pci_exit(&driver, &i915_pci_driver);
 }
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 90fcccba17b..374f964323a 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -35,11 +35,13 @@
 #include "i915_reg.h"
 #include "intel_bios.h"
 #include "intel_ringbuffer.h"
+#include "i915_gem_gtt.h"
 #include <linux/io-mapping.h>
 #include <linux/i2c.h>
 #include <linux/i2c-algo-bit.h>
 #include <drm/intel-gtt.h>
 #include <linux/backlight.h>
+#include <linux/hashtable.h>
 #include <linux/intel-iommu.h>
 #include <linux/kref.h>
 #include <linux/pm_qos.h>
@@ -58,7 +60,8 @@ enum pipe {
 	PIPE_A = 0,
 	PIPE_B,
 	PIPE_C,
-	I915_MAX_PIPES
+	_PIPE_EDP,
+	I915_MAX_PIPES = _PIPE_EDP
 };
 #define pipe_name(p) ((p) + 'A')
 
@@ -66,7 +69,8 @@ enum transcoder {
 	TRANSCODER_A = 0,
 	TRANSCODER_B,
 	TRANSCODER_C,
-	TRANSCODER_EDP = 0xF,
+	TRANSCODER_EDP,
+	I915_MAX_TRANSCODERS
 };
 #define transcoder_name(t) ((t) + 'A')
 
@@ -77,7 +81,7 @@ enum plane {
 };
 #define plane_name(p) ((p) + 'A')
 
-#define sprite_name(p, s) ((p) * dev_priv->num_plane + (s) + 'A')
+#define sprite_name(p, s) ((p) * INTEL_INFO(dev)->num_sprites[(p)] + (s) + 'A')
 
 enum port {
 	PORT_A = 0,
@@ -89,6 +93,18 @@ enum port {
 };
 #define port_name(p) ((p) + 'A')
 
+#define I915_NUM_PHYS_VLV 2
+
+enum dpio_channel {
+	DPIO_CH0,
+	DPIO_CH1
+};
+
+enum dpio_phy {
+	DPIO_PHY0,
+	DPIO_PHY1
+};
+
 enum intel_display_power_domain {
 	POWER_DOMAIN_PIPE_A,
 	POWER_DOMAIN_PIPE_B,
@@ -100,14 +116,24 @@ enum intel_display_power_domain {
 	POWER_DOMAIN_TRANSCODER_B,
 	POWER_DOMAIN_TRANSCODER_C,
 	POWER_DOMAIN_TRANSCODER_EDP,
+	POWER_DOMAIN_PORT_DDI_A_2_LANES,
+	POWER_DOMAIN_PORT_DDI_A_4_LANES,
+	POWER_DOMAIN_PORT_DDI_B_2_LANES,
+	POWER_DOMAIN_PORT_DDI_B_4_LANES,
+	POWER_DOMAIN_PORT_DDI_C_2_LANES,
+	POWER_DOMAIN_PORT_DDI_C_4_LANES,
+	POWER_DOMAIN_PORT_DDI_D_2_LANES,
+	POWER_DOMAIN_PORT_DDI_D_4_LANES,
+	POWER_DOMAIN_PORT_DSI,
+	POWER_DOMAIN_PORT_CRT,
+	POWER_DOMAIN_PORT_OTHER,
 	POWER_DOMAIN_VGA,
+	POWER_DOMAIN_AUDIO,
 	POWER_DOMAIN_INIT,
 
 	POWER_DOMAIN_NUM,
 };
 
-#define POWER_DOMAIN_MASK (BIT(POWER_DOMAIN_NUM) - 1)
-
 #define POWER_DOMAIN_PIPE(pipe) ((pipe) + POWER_DOMAIN_PIPE_A)
 #define POWER_DOMAIN_PIPE_PANEL_FITTER(pipe) \
 		((pipe) + POWER_DOMAIN_PIPE_A_PANEL_FITTER)
@@ -115,14 +141,6 @@ enum intel_display_power_domain {
 	((tran) == TRANSCODER_EDP ? POWER_DOMAIN_TRANSCODER_EDP : \
 	 (tran) + POWER_DOMAIN_TRANSCODER_A)
 
-#define HSW_ALWAYS_ON_POWER_DOMAINS (		\
-	BIT(POWER_DOMAIN_PIPE_A) |		\
-	BIT(POWER_DOMAIN_TRANSCODER_EDP))
-#define BDW_ALWAYS_ON_POWER_DOMAINS (		\
-	BIT(POWER_DOMAIN_PIPE_A) |		\
-	BIT(POWER_DOMAIN_TRANSCODER_EDP) |	\
-	BIT(POWER_DOMAIN_PIPE_A_PANEL_FITTER))
-
 enum hpd_pin {
 	HPD_NONE = 0,
 	HPD_PORT_A = HPD_NONE, /* PORT_A is internal */
@@ -144,12 +162,24 @@ enum hpd_pin {
 	 I915_GEM_DOMAIN_VERTEX)
 
 #define for_each_pipe(p) for ((p) = 0; (p) < INTEL_INFO(dev)->num_pipes; (p)++)
+#define for_each_sprite(p, s) for ((s) = 0; (s) < INTEL_INFO(dev)->num_sprites[(p)]; (s)++)
+
+#define for_each_crtc(dev, crtc) \
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
+
+#define for_each_intel_crtc(dev, intel_crtc) \
+	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list, base.head)
 
 #define for_each_encoder_on_crtc(dev, __crtc, intel_encoder) \
 	list_for_each_entry((intel_encoder), &(dev)->mode_config.encoder_list, base.head) \
 		if ((intel_encoder)->base.crtc == (__crtc))
 
+#define for_each_connector_on_encoder(dev, __encoder, intel_connector) \
+	list_for_each_entry((intel_connector), &(dev)->mode_config.connector_list, base.head) \
+		if ((intel_connector)->base.encoder == (__encoder))
+
 struct drm_i915_private;
+struct i915_mmu_object;
 
 enum intel_dpll_id {
 	DPLL_ID_PRIVATE = -1, /* non-shared dpll in use */
@@ -221,18 +251,6 @@ struct intel_ddi_plls {
 #define WATCH_LISTS	0
 #define WATCH_GTT	0
 
-#define I915_GEM_PHYS_CURSOR_0 1
-#define I915_GEM_PHYS_CURSOR_1 2
-#define I915_GEM_PHYS_OVERLAY_REGS 3
-#define I915_MAX_PHYS_OBJECT (I915_GEM_PHYS_OVERLAY_REGS)
-
-struct drm_i915_gem_phys_object {
-	int id;
-	struct page **page_list;
-	drm_dma_handle_t *handle;
-	struct drm_i915_gem_object *cur_obj;
-};
-
 struct opregion_header;
 struct opregion_acpi;
 struct opregion_swsci;
@@ -282,52 +300,86 @@ struct intel_display_error_state;
 
 struct drm_i915_error_state {
 	struct kref ref;
+	struct timeval time;
+
+	char error_msg[128];
+	u32 reset_count;
+	u32 suspend_count;
+
+	/* Generic register state */
 	u32 eir;
 	u32 pgtbl_er;
 	u32 ier;
 	u32 ccid;
 	u32 derrmr;
 	u32 forcewake;
-	bool waiting[I915_NUM_RINGS];
-	u32 pipestat[I915_MAX_PIPES];
-	u32 tail[I915_NUM_RINGS];
-	u32 head[I915_NUM_RINGS];
-	u32 ctl[I915_NUM_RINGS];
-	u32 ipeir[I915_NUM_RINGS];
-	u32 ipehr[I915_NUM_RINGS];
-	u32 instdone[I915_NUM_RINGS];
-	u32 acthd[I915_NUM_RINGS];
-	u32 semaphore_mboxes[I915_NUM_RINGS][I915_NUM_RINGS - 1];
-	u32 semaphore_seqno[I915_NUM_RINGS][I915_NUM_RINGS - 1];
-	u32 rc_psmi[I915_NUM_RINGS]; /* sleep state */
-	/* our own tracking of ring head and tail */
-	u32 cpu_ring_head[I915_NUM_RINGS];
-	u32 cpu_ring_tail[I915_NUM_RINGS];
 	u32 error; /* gen6+ */
 	u32 err_int; /* gen7 */
-	u32 bbstate[I915_NUM_RINGS];
-	u32 instpm[I915_NUM_RINGS];
-	u32 instps[I915_NUM_RINGS];
-	u32 extra_instdone[I915_NUM_INSTDONE_REG];
-	u32 seqno[I915_NUM_RINGS];
-	u64 bbaddr;
-	u32 fault_reg[I915_NUM_RINGS];
 	u32 done_reg;
-	u32 faddr[I915_NUM_RINGS];
+	u32 gac_eco;
+	u32 gam_ecochk;
+	u32 gab_ctl;
+	u32 gfx_mode;
+	u32 extra_instdone[I915_NUM_INSTDONE_REG];
 	u64 fence[I915_MAX_NUM_FENCES];
-	struct timeval time;
+	struct intel_overlay_error_state *overlay;
+	struct intel_display_error_state *display;
+
 	struct drm_i915_error_ring {
+		bool valid;
+		/* Software tracked state */
+		bool waiting;
+		int hangcheck_score;
+		enum intel_ring_hangcheck_action hangcheck_action;
+		int num_requests;
+
+		/* our own tracking of ring head and tail */
+		u32 cpu_ring_head;
+		u32 cpu_ring_tail;
+
+		u32 semaphore_seqno[I915_NUM_RINGS - 1];
+
+		/* Register state */
+		u32 tail;
+		u32 head;
+		u32 ctl;
+		u32 hws;
+		u32 ipeir;
+		u32 ipehr;
+		u32 instdone;
+		u32 bbstate;
+		u32 instpm;
+		u32 instps;
+		u32 seqno;
+		u64 bbaddr;
+		u64 acthd;
+		u32 fault_reg;
+		u64 faddr;
+		u32 rc_psmi; /* sleep state */
+		u32 semaphore_mboxes[I915_NUM_RINGS - 1];
+
 		struct drm_i915_error_object {
 			int page_count;
 			u32 gtt_offset;
 			u32 *pages[0];
-		} *ringbuffer, *batchbuffer, *ctx;
+		} *ringbuffer, *batchbuffer, *wa_batchbuffer, *ctx, *hws_page;
+
 		struct drm_i915_error_request {
 			long jiffies;
 			u32 seqno;
 			u32 tail;
 		} *requests;
-		int num_requests;
+
+		struct {
+			u32 gfx_mode;
+			union {
+				u64 pdp[4];
+				u32 pp_dir_base;
+			};
+		} vm_info;
+
+		pid_t pid;
+		char comm[TASK_COMM_LEN];
 	} ring[I915_NUM_RINGS];
 	struct drm_i915_error_buffer {
 		u32 size;
@@ -341,24 +393,24 @@ struct drm_i915_error_state {
 		u32 tiling:2;
 		u32 dirty:1;
 		u32 purgeable:1;
+		u32 userptr:1;
 		s32 ring:4;
 		u32 cache_level:3;
 	} **active_bo, **pinned_bo;
+
 	u32 *active_bo_count, *pinned_bo_count;
-	struct intel_overlay_error_state *overlay;
-	struct intel_display_error_state *display;
-	int hangcheck_score[I915_NUM_RINGS];
-	enum intel_ring_hangcheck_action hangcheck_action[I915_NUM_RINGS];
 };
 
+struct intel_connector;
 struct intel_crtc_config;
+struct intel_plane_config;
 struct intel_crtc;
 struct intel_limit;
 struct dpll;
 
 struct drm_i915_display_funcs {
 	bool (*fbc_enabled)(struct drm_device *dev);
-	void (*enable_fbc)(struct drm_crtc *crtc, unsigned long interval);
+	void (*enable_fbc)(struct drm_crtc *crtc);
 	void (*disable_fbc)(struct drm_device *dev);
 	int (*get_display_clock_speed)(struct drm_device *dev);
 	int (*get_fifo_size)(struct drm_device *dev, int plane);
@@ -390,6 +442,8 @@ struct drm_i915_display_funcs {
 	 * fills out the pipe-config with the hw state. */
 	bool (*get_pipe_config)(struct intel_crtc *,
 				struct intel_crtc_config *);
+	void (*get_plane_config)(struct intel_crtc *,
+				 struct intel_plane_config *);
 	int (*crtc_mode_set)(struct drm_crtc *crtc,
 			     int x, int y,
 			     struct drm_framebuffer *old_fb);
@@ -404,20 +458,31 @@ struct drm_i915_display_funcs {
 	int (*queue_flip)(struct drm_device *dev, struct drm_crtc *crtc,
 			  struct drm_framebuffer *fb,
 			  struct drm_i915_gem_object *obj,
+			  struct intel_engine_cs *ring,
 			  uint32_t flags);
-	int (*update_plane)(struct drm_crtc *crtc, struct drm_framebuffer *fb,
-			    int x, int y);
+	void (*update_primary_plane)(struct drm_crtc *crtc,
+				     struct drm_framebuffer *fb,
+				     int x, int y);
 	void (*hpd_irq_setup)(struct drm_device *dev);
 	/* clock updates for mode set */
 	/* cursor updates */
 	/* render clock increase/decrease */
 	/* display clock increase/decrease */
 	/* pll clock increase/decrease */
+
+	int (*setup_backlight)(struct intel_connector *connector);
+	uint32_t (*get_backlight)(struct intel_connector *connector);
+	void (*set_backlight)(struct intel_connector *connector,
+			      uint32_t level);
+	void (*disable_backlight)(struct intel_connector *connector);
+	void (*enable_backlight)(struct intel_connector *connector);
 };
 
 struct intel_uncore_funcs {
-	void (*force_wake_get)(struct drm_i915_private *dev_priv);
-	void (*force_wake_put)(struct drm_i915_private *dev_priv);
+	void (*force_wake_get)(struct drm_i915_private *dev_priv,
+							int fw_engine);
+	void (*force_wake_put)(struct drm_i915_private *dev_priv,
+							int fw_engine);
 
 	uint8_t  (*mmio_readb)(struct drm_i915_private *dev_priv, off_t offset, bool trace);
 	uint16_t (*mmio_readw)(struct drm_i915_private *dev_priv, off_t offset, bool trace);
@@ -442,7 +507,10 @@ struct intel_uncore {
 	unsigned fifo_count;
 	unsigned forcewake_count;
 
-	struct delayed_work force_wake_work;
+	unsigned fw_rendercount;
+	unsigned fw_mediacount;
+
+	struct timer_list force_wake_timer;
 };
 
 #define DEV_INFO_FOR_EACH_FLAG(func, sep) \
@@ -477,9 +545,17 @@ struct intel_uncore {
 struct intel_device_info {
 	u32 display_mmio_offset;
 	u8 num_pipes:3;
+	u8 num_sprites[I915_MAX_PIPES];
 	u8 gen;
 	u8 ring_mask; /* Rings supported by the HW */
 	DEV_INFO_FOR_EACH_FLAG(DEFINE_FLAG, SEP_SEMICOLON);
+	/* Register offsets for the various display pipes and transcoders */
+	int pipe_offsets[I915_MAX_TRANSCODERS];
+	int trans_offsets[I915_MAX_TRANSCODERS];
+	int dpll_offsets[I915_MAX_PIPES];
+	int dpll_md_offsets[I915_MAX_PIPES];
+	int palette_offsets[I915_MAX_PIPES];
+	int cursor_offsets[I915_MAX_PIPES];
 };
 
 #undef DEFINE_FLAG
@@ -495,138 +571,6 @@ enum i915_cache_level {
 	I915_CACHE_WT, /* hsw:gt3e WriteThrough for scanouts */
 };
 
-typedef uint32_t gen6_gtt_pte_t;
-
-struct i915_address_space {
-	struct drm_mm mm;
-	struct drm_device *dev;
-	struct list_head global_link;
-	unsigned long start;		/* Start offset always 0 for dri2 */
-	size_t total;		/* size addr space maps (ex. 2GB for ggtt) */
-
-	struct {
-		dma_addr_t addr;
-		struct page *page;
-	} scratch;
-
-	/**
-	 * List of objects currently involved in rendering.
-	 *
-	 * Includes buffers having the contents of their GPU caches
-	 * flushed, not necessarily primitives.  last_rendering_seqno
-	 * represents when the rendering involved will be completed.
-	 *
-	 * A reference is held on the buffer while on this list.
-	 */
-	struct list_head active_list;
-
-	/**
-	 * LRU list of objects which are not in the ringbuffer and
-	 * are ready to unbind, but are still in the GTT.
-	 *
-	 * last_rendering_seqno is 0 while an object is in this list.
-	 *
-	 * A reference is not held on the buffer while on this list,
-	 * as merely being GTT-bound shouldn't prevent its being
-	 * freed, and we'll pull it off the list in the free path.
-	 */
-	struct list_head inactive_list;
-
-	/* FIXME: Need a more generic return type */
-	gen6_gtt_pte_t (*pte_encode)(dma_addr_t addr,
-				     enum i915_cache_level level,
-				     bool valid); /* Create a valid PTE */
-	void (*clear_range)(struct i915_address_space *vm,
-			    unsigned int first_entry,
-			    unsigned int num_entries,
-			    bool use_scratch);
-	void (*insert_entries)(struct i915_address_space *vm,
-			       struct sg_table *st,
-			       unsigned int first_entry,
-			       enum i915_cache_level cache_level);
-	void (*cleanup)(struct i915_address_space *vm);
-};
-
-/* The Graphics Translation Table is the way in which GEN hardware translates a
- * Graphics Virtual Address into a Physical Address. In addition to the normal
- * collateral associated with any va->pa translations GEN hardware also has a
- * portion of the GTT which can be mapped by the CPU and remain both coherent
- * and correct (in cases like swizzling). That region is referred to as GMADR in
- * the spec.
- */
-struct i915_gtt {
-	struct i915_address_space base;
-	size_t stolen_size;		/* Total size of stolen memory */
-
-	unsigned long mappable_end;	/* End offset that we can CPU map */
-	struct io_mapping *mappable;	/* Mapping to our CPU mappable region */
-	phys_addr_t mappable_base;	/* PA of our GMADR */
-
-	/** "Graphics Stolen Memory" holds the global PTEs */
-	void __iomem *gsm;
-
-	bool do_idle_maps;
-
-	int mtrr;
-
-	/* global gtt ops */
-	int (*gtt_probe)(struct drm_device *dev, size_t *gtt_total,
-			  size_t *stolen, phys_addr_t *mappable_base,
-			  unsigned long *mappable_end);
-};
-#define gtt_total_entries(gtt) ((gtt).base.total >> PAGE_SHIFT)
-
-struct i915_hw_ppgtt {
-	struct i915_address_space base;
-	unsigned num_pd_entries;
-	union {
-		struct page **pt_pages;
-		struct page *gen8_pt_pages;
-	};
-	struct page *pd_pages;
-	int num_pd_pages;
-	int num_pt_pages;
-	union {
-		uint32_t pd_offset;
-		dma_addr_t pd_dma_addr[4];
-	};
-	union {
-		dma_addr_t *pt_dma_addr;
-		dma_addr_t *gen8_pt_dma_addr[4];
-	};
-	int (*enable)(struct drm_device *dev);
-};
-
-/**
- * A VMA represents a GEM BO that is bound into an address space. Therefore, a
- * VMA's presence cannot be guaranteed before binding, or after unbinding the
- * object into/from the address space.
- *
- * To make things as simple as possible (ie. no refcounting), a VMA's lifetime
- * will always be <= an objects lifetime. So object refcounting should cover us.
- */
-struct i915_vma {
-	struct drm_mm_node node;
-	struct drm_i915_gem_object *obj;
-	struct i915_address_space *vm;
-
-	/** This object's place on the active/inactive lists */
-	struct list_head mm_list;
-
-	struct list_head vma_link; /* Link in the object's VMA list */
-
-	/** This vma's place in the batchbuffer or on the eviction list */
-	struct list_head exec_list;
-
-	/**
-	 * Used for performing relocations during execbuffer insertion.
-	 */
-	struct hlist_node exec_node;
-	unsigned long exec_handle;
-	struct drm_i915_gem_exec_object2 *exec_entry;
-
-};
-
 struct i915_ctx_hang_stats {
 	/* This context had batch pending when hang was declared */
 	unsigned batch_pending;
@@ -643,15 +587,16 @@ struct i915_ctx_hang_stats {
 
 /* This must match up with the value previously used for execbuf2.rsvd1. */
 #define DEFAULT_CONTEXT_ID 0
-struct i915_hw_context {
+struct intel_context {
 	struct kref ref;
 	int id;
 	bool is_initialized;
 	uint8_t remap_slice;
 	struct drm_i915_file_private *file_priv;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *last_ring;
 	struct drm_i915_gem_object *obj;
 	struct i915_ctx_hang_stats hang_stats;
+	struct i915_address_space *vm;
 
 	struct list_head link;
 };
@@ -669,7 +614,6 @@ struct i915_fbc {
 		struct delayed_work work;
 		struct drm_crtc *crtc;
 		struct drm_framebuffer *fb;
-		int interval;
 	} *fbc_work;
 
 	enum no_fbc_reason {
@@ -687,6 +631,10 @@ struct i915_fbc {
 	} no_fbc_reason;
 };
 
+struct i915_drrs {
+	struct intel_connector *connector;
+};
+
 struct i915_psr {
 	bool sink_support;
 	bool source_ok;
@@ -708,7 +656,7 @@ enum intel_sbi_destination {
 #define QUIRK_PIPEA_FORCE (1<<0)
 #define QUIRK_LVDS_SSC_DISABLE (1<<1)
 #define QUIRK_INVERT_BRIGHTNESS (1<<2)
-#define QUIRK_NO_PCH_PWM_ENABLE (1<<3)
+#define QUIRK_BACKLIGHT_PRESENT (1<<3)
 
 struct intel_fbdev;
 struct intel_fbc_work;
@@ -761,8 +709,6 @@ struct i915_suspend_saved_registers {
 	u32 saveBLC_PWM_CTL;
 	u32 saveBLC_PWM_CTL2;
 	u32 saveBLC_HIST_CTL_B;
-	u32 saveBLC_PWM_CTL_B;
-	u32 saveBLC_PWM_CTL2_B;
 	u32 saveBLC_CPU_PWM_CTL;
 	u32 saveBLC_CPU_PWM_CTL2;
 	u32 saveFPB0;
@@ -808,11 +754,7 @@ struct i915_suspend_saved_registers {
 	u32 savePFIT_CONTROL;
 	u32 save_palette_a[256];
 	u32 save_palette_b[256];
-	u32 saveDPFC_CB_BASE;
-	u32 saveFBC_CFB_BASE;
-	u32 saveFBC_LL_BASE;
 	u32 saveFBC_CONTROL;
-	u32 saveFBC_CONTROL2;
 	u32 saveIER;
 	u32 saveIIR;
 	u32 saveIMR;
@@ -877,20 +819,90 @@ struct i915_suspend_saved_registers {
 	u32 savePCH_PORT_HOTPLUG;
 };
 
+struct vlv_s0ix_state {
+	/* GAM */
+	u32 wr_watermark;
+	u32 gfx_prio_ctrl;
+	u32 arb_mode;
+	u32 gfx_pend_tlb0;
+	u32 gfx_pend_tlb1;
+	u32 lra_limits[GEN7_LRA_LIMITS_REG_NUM];
+	u32 media_max_req_count;
+	u32 gfx_max_req_count;
+	u32 render_hwsp;
+	u32 ecochk;
+	u32 bsd_hwsp;
+	u32 blt_hwsp;
+	u32 tlb_rd_addr;
+
+	/* MBC */
+	u32 g3dctl;
+	u32 gsckgctl;
+	u32 mbctl;
+
+	/* GCP */
+	u32 ucgctl1;
+	u32 ucgctl3;
+	u32 rcgctl1;
+	u32 rcgctl2;
+	u32 rstctl;
+	u32 misccpctl;
+
+	/* GPM */
+	u32 gfxpause;
+	u32 rpdeuhwtc;
+	u32 rpdeuc;
+	u32 ecobus;
+	u32 pwrdwnupctl;
+	u32 rp_down_timeout;
+	u32 rp_deucsw;
+	u32 rcubmabdtmr;
+	u32 rcedata;
+	u32 spare2gh;
+
+	/* Display 1 CZ domain */
+	u32 gt_imr;
+	u32 gt_ier;
+	u32 pm_imr;
+	u32 pm_ier;
+	u32 gt_scratch[GEN7_GT_SCRATCH_REG_NUM];
+
+	/* GT SA CZ domain */
+	u32 tilectl;
+	u32 gt_fifoctl;
+	u32 gtlc_wake_ctrl;
+	u32 gtlc_survive;
+	u32 pmwgicz;
+
+	/* Display 2 CZ domain */
+	u32 gu_ctl0;
+	u32 gu_ctl1;
+	u32 clock_gate_dis2;
+};
+
 struct intel_gen6_power_mgmt {
 	/* work and pm_iir are protected by dev_priv->irq_lock */
 	struct work_struct work;
 	u32 pm_iir;
 
-	/* The below variables an all the rps hw state are protected by
-	 * dev->struct mutext. */
-	u8 cur_delay;
-	u8 min_delay;
-	u8 max_delay;
-	u8 rpe_delay;
-	u8 rp1_delay;
-	u8 rp0_delay;
-	u8 hw_max;
+	/* Frequencies are stored in potentially platform dependent multiples.
+	 * In other words, *_freq needs to be multiplied by X to be interesting.
+	 * Soft limits are those which are used for the dynamic reclocking done
+	 * by the driver (raise frequencies under heavy loads, and lower for
+	 * lighter loads). Hard limits are those imposed by the hardware.
+	 *
+	 * A distinction is made for overclocking, which is never enabled by
+	 * default, and is considered to be above the hard limit if it's
+	 * possible at all.
+	 */
+	u8 cur_freq;		/* Current frequency (cached, may not == HW) */
+	u8 min_freq_softlimit;	/* Minimum frequency permitted by the driver */
+	u8 max_freq_softlimit;	/* Max frequency permitted by the driver */
+	u8 max_freq;		/* Maximum frequency, RP0 if not overclocking */
+	u8 min_freq;		/* AKA RPn. Minimum frequency */
+	u8 efficient_freq;	/* AKA RPe. Pre-determined balanced frequency */
+	u8 rp1_freq;		/* "less than" RP0 power/freqency */
+	u8 rp0_freq;		/* Non-overclocked max frequency. */
 
 	int last_adj;
 	enum { LOW_POWER, BETWEEN, HIGH_POWER } power;
@@ -930,23 +942,61 @@ struct intel_ilk_power_mgmt {
 	struct drm_i915_gem_object *renderctx;
 };
 
+struct drm_i915_private;
+struct i915_power_well;
+
+struct i915_power_well_ops {
+	/*
+	 * Synchronize the well's hw state to match the current sw state, for
+	 * example enable/disable it based on the current refcount. Called
+	 * during driver init and resume time, possibly after first calling
+	 * the enable/disable handlers.
+	 */
+	void (*sync_hw)(struct drm_i915_private *dev_priv,
+			struct i915_power_well *power_well);
+	/*
+	 * Enable the well and resources that depend on it (for example
+	 * interrupts located on the well). Called after the 0->1 refcount
+	 * transition.
+	 */
+	void (*enable)(struct drm_i915_private *dev_priv,
+		       struct i915_power_well *power_well);
+	/*
+	 * Disable the well and resources that depend on it. Called after
+	 * the 1->0 refcount transition.
+	 */
+	void (*disable)(struct drm_i915_private *dev_priv,
+			struct i915_power_well *power_well);
+	/* Returns the hw enabled state. */
+	bool (*is_enabled)(struct drm_i915_private *dev_priv,
+			   struct i915_power_well *power_well);
+};
+
 /* Power well structure for haswell */
 struct i915_power_well {
+	const char *name;
+	bool always_on;
 	/* power well enable/disable usage count */
 	int count;
+	/* cached hw enabled state */
+	bool hw_enabled;
+	unsigned long domains;
+	unsigned long data;
+	const struct i915_power_well_ops *ops;
 };
 
-#define I915_MAX_POWER_WELLS 1
-
 struct i915_power_domains {
 	/*
 	 * Power wells needed for initialization at driver init and suspend
 	 * time are on. They are kept on until after the first modeset.
 	 */
 	bool init_power_on;
+	bool initializing;
+	int power_well_count;
 
 	struct mutex lock;
-	struct i915_power_well power_wells[I915_MAX_POWER_WELLS];
+	int domain_use_count[POWER_DOMAIN_NUM];
+	struct i915_power_well *power_wells;
 };
 
 struct i915_dri1_state {
@@ -1000,7 +1050,8 @@ struct i915_gem_mm {
 	/** PPGTT used for aliasing the PPGTT with the GTT */
 	struct i915_hw_ppgtt *aliasing_ppgtt;
 
-	struct shrinker inactive_shrinker;
+	struct notifier_block oom_notifier;
+	struct shrinker shrinker;
 	bool shrinker_no_lock_stealing;
 
 	/** LRU list of objects with fence regs on them. */
@@ -1030,14 +1081,22 @@ struct i915_gem_mm {
 	 */
 	bool interruptible;
 
+	/**
+	 * Is the GPU currently considered idle, or busy executing userspace
+	 * requests?  Whilst idle, we attempt to power down the hardware and
+	 * display clocks. In order to reduce the effect on performance, there
+	 * is a slight delay before we do so.
+	 */
+	bool busy;
+
+	/* the indicator for dispatch video commands on two BSD rings */
+	int bsd_ring_dispatch_index;
+
 	/** Bit 6 swizzling required for X tiling */
 	uint32_t bit_6_swizzle_x;
 	/** Bit 6 swizzling required for Y tiling */
 	uint32_t bit_6_swizzle_y;
 
-	/* storage for physical objects */
-	struct drm_i915_gem_phys_object *phys_objs[I915_MAX_PHYS_OBJECT];
-
 	/* accounting, useful for userland debugging */
 	spinlock_t object_stat_lock;
 	size_t object_memory;
@@ -1077,34 +1136,30 @@ struct i915_gpu_error {
 	unsigned long missed_irq_rings;
 
 	/**
-	 * State variable and reset counter controlling the reset flow
+	 * State variable controlling the reset flow and count
+	 *
+	 * This is a counter which gets incremented when reset is triggered,
+	 * and again when reset has been handled. So odd values (lowest bit set)
+	 * means that reset is in progress and even values that
+	 * (reset_counter >> 1):th reset was successfully completed.
+	 *
+	 * If reset is not completed succesfully, the I915_WEDGE bit is
+	 * set meaning that hardware is terminally sour and there is no
+	 * recovery. All waiters on the reset_queue will be woken when
+	 * that happens.
 	 *
-	 * Upper bits are for the reset counter.  This counter is used by the
-	 * wait_seqno code to race-free noticed that a reset event happened and
-	 * that it needs to restart the entire ioctl (since most likely the
-	 * seqno it waited for won't ever signal anytime soon).
+	 * This counter is used by the wait_seqno code to notice that reset
+	 * event happened and it needs to restart the entire ioctl (since most
+	 * likely the seqno it waited for won't ever signal anytime soon).
 	 *
 	 * This is important for lock-free wait paths, where no contended lock
 	 * naturally enforces the correct ordering between the bail-out of the
 	 * waiter and the gpu reset work code.
-	 *
-	 * Lowest bit controls the reset state machine: Set means a reset is in
-	 * progress. This state will (presuming we don't have any bugs) decay
-	 * into either unset (successful reset) or the special WEDGED value (hw
-	 * terminally sour). All waiters on the reset_queue will be woken when
-	 * that happens.
 	 */
 	atomic_t reset_counter;
 
-	/**
-	 * Special values/flags for reset_counter
-	 *
-	 * Note that the code relies on
-	 * 	I915_WEDGED & I915_RESET_IN_PROGRESS_FLAG
-	 * being true.
-	 */
 #define I915_RESET_IN_PROGRESS_FLAG	1
-#define I915_WEDGED			0xffffffff
+#define I915_WEDGED			(1 << 31)
 
 	/**
 	 * Waitqueue to signal when the reset has completed. Used by clients
@@ -1112,8 +1167,12 @@ struct i915_gpu_error {
 	 */
 	wait_queue_head_t reset_queue;
 
-	/* For gpu hang simulation. */
-	unsigned int stop_rings;
+	/* Userspace knobs for gpu hang simulation;
+	 * combines both a ring mask, and extra flags
+	 */
+	u32 stop_rings;
+#define I915_STOP_RING_ALLOW_BAN       (1 << 31)
+#define I915_STOP_RING_ALLOW_WARN      (1 << 30)
 
 	/* For missed irq/seqno simulation. */
 	unsigned int test_irq_rings;
@@ -1133,6 +1192,12 @@ struct ddi_vbt_port_info {
 	uint8_t supports_dp:1;
 };
 
+enum drrs_support_type {
+	DRRS_NOT_SUPPORTED = 0,
+	STATIC_DRRS_SUPPORT = 1,
+	SEAMLESS_DRRS_SUPPORT = 2
+};
+
 struct intel_vbt_data {
 	struct drm_display_mode *lfp_lvds_vbt_mode; /* if any */
 	struct drm_display_mode *sdvo_lvds_vbt_mode; /* if any */
@@ -1145,9 +1210,12 @@ struct intel_vbt_data {
 	unsigned int lvds_use_ssc:1;
 	unsigned int display_clock_mode:1;
 	unsigned int fdi_rx_polarity_inverted:1;
+	unsigned int has_mipi:1;
 	int lvds_ssc_freq;
 	unsigned int bios_lvds_val; /* initial [PCH_]LVDS reg val in VBIOS */
 
+	enum drrs_support_type drrs_type;
+
 	/* eDP */
 	int edp_rate;
 	int edp_lanes;
@@ -1158,9 +1226,22 @@ struct intel_vbt_data {
 	int edp_bpp;
 	struct edp_power_seq edp_pps;
 
+	struct {
+		u16 pwm_freq_hz;
+		bool present;
+		bool active_low_pwm;
+	} backlight;
+
 	/* MIPI DSI */
 	struct {
+		u16 port;
 		u16 panel_id;
+		struct mipi_config *config;
+		struct mipi_pps_data *pps;
+		u8 seq_version;
+		u32 size;
+		u8 *data;
+		u8 *sequence[MIPI_SEQ_MAX];
 	} dsi;
 
 	int crt_ddc_pin;
@@ -1184,7 +1265,7 @@ struct intel_wm_level {
 	uint32_t fbc_val;
 };
 
-struct hsw_wm_values {
+struct ilk_wm_values {
 	uint32_t wm_pipe[3];
 	uint32_t wm_lp[3];
 	uint32_t wm_lp_spr[3];
@@ -1194,72 +1275,31 @@ struct hsw_wm_values {
 };
 
 /*
- * This struct tracks the state needed for the Package C8+ feature.
- *
- * Package states C8 and deeper are really deep PC states that can only be
- * reached when all the devices on the system allow it, so even if the graphics
- * device allows PC8+, it doesn't mean the system will actually get to these
- * states.
+ * This struct helps tracking the state needed for runtime PM, which puts the
+ * device in PCI D3 state. Notice that when this happens, nothing on the
+ * graphics device works, even register access, so we don't get interrupts nor
+ * anything else.
  *
- * Our driver only allows PC8+ when all the outputs are disabled, the power well
- * is disabled and the GPU is idle. When these conditions are met, we manually
- * do the other conditions: disable the interrupts, clocks and switch LCPLL
- * refclk to Fclk.
+ * Every piece of our code that needs to actually touch the hardware needs to
+ * either call intel_runtime_pm_get or call intel_display_power_get with the
+ * appropriate power domain.
  *
- * When we really reach PC8 or deeper states (not just when we allow it) we lose
- * the state of some registers, so when we come back from PC8+ we need to
- * restore this state. We don't get into PC8+ if we're not in RC6, so we don't
- * need to take care of the registers kept by RC6.
- *
- * The interrupt disabling is part of the requirements. We can only leave the
- * PCH HPD interrupts enabled. If we're in PC8+ and we get another interrupt we
- * can lock the machine.
- *
- * Ideally every piece of our code that needs PC8+ disabled would call
- * hsw_disable_package_c8, which would increment disable_count and prevent the
- * system from reaching PC8+. But we don't have a symmetric way to do this for
- * everything, so we have the requirements_met and gpu_idle variables. When we
- * switch requirements_met or gpu_idle to true we decrease disable_count, and
- * increase it in the opposite case. The requirements_met variable is true when
- * all the CRTCs, encoders and the power well are disabled. The gpu_idle
- * variable is true when the GPU is idle.
- *
- * In addition to everything, we only actually enable PC8+ if disable_count
- * stays at zero for at least some seconds. This is implemented with the
- * enable_work variable. We do this so we don't enable/disable PC8 dozens of
- * consecutive times when all screens are disabled and some background app
- * queries the state of our connectors, or we have some application constantly
- * waking up to use the GPU. Only after the enable_work function actually
- * enables PC8+ the "enable" variable will become true, which means that it can
- * be false even if disable_count is 0.
+ * Our driver uses the autosuspend delay feature, which means we'll only really
+ * suspend if we stay with zero refcount for a certain amount of time. The
+ * default value is currently very conservative (see intel_init_runtime_pm), but
+ * it can be changed with the standard runtime PM files from sysfs.
  *
  * The irqs_disabled variable becomes true exactly after we disable the IRQs and
  * goes back to false exactly before we reenable the IRQs. We use this variable
  * to check if someone is trying to enable/disable IRQs while they're supposed
  * to be disabled. This shouldn't happen and we'll print some error messages in
- * case it happens, but if it actually happens we'll also update the variables
- * inside struct regsave so when we restore the IRQs they will contain the
- * latest expected values.
+ * case it happens.
  *
- * For more, read "Display Sequences for Package C8" on our documentation.
+ * For more, read the Documentation/power/runtime_pm.txt.
  */
-struct i915_package_c8 {
-	bool requirements_met;
-	bool gpu_idle;
+struct i915_runtime_pm {
+	bool suspended;
 	bool irqs_disabled;
-	/* Only true after the delayed work task actually enables it. */
-	bool enabled;
-	int disable_count;
-	struct mutex lock;
-	struct delayed_work enable_work;
-
-	struct {
-		uint32_t deimr;
-		uint32_t sdeimr;
-		uint32_t gtimr;
-		uint32_t gtier;
-		uint32_t gen6_pmimr;
-	} regsave;
 };
 
 enum intel_pipe_crc_source {
@@ -1292,11 +1332,11 @@ struct intel_pipe_crc {
 	wait_queue_head_t wq;
 };
 
-typedef struct drm_i915_private {
+struct drm_i915_private {
 	struct drm_device *dev;
 	struct kmem_cache *slab;
 
-	const struct intel_device_info *info;
+	const struct intel_device_info info;
 
 	int relative_constants_mode;
 
@@ -1316,20 +1356,23 @@ typedef struct drm_i915_private {
 	 */
 	uint32_t gpio_mmio_base;
 
+	/* MMIO base address for MIPI regs */
+	uint32_t mipi_mmio_base;
+
 	wait_queue_head_t gmbus_wait_queue;
 
 	struct pci_dev *bridge_dev;
-	struct intel_ring_buffer ring[I915_NUM_RINGS];
+	struct intel_engine_cs ring[I915_NUM_RINGS];
 	uint32_t last_seqno, next_seqno;
 
 	drm_dma_handle_t *status_page_dmah;
 	struct resource mch_res;
 
-	atomic_t irq_received;
-
 	/* protects the irq masks */
 	spinlock_t irq_lock;
 
+	bool display_irqs_enabled;
+
 	/* To control wakeup latency, e.g. for irq-driven dp aux transfers. */
 	struct pm_qos_request pm_qos;
 
@@ -1343,6 +1386,8 @@ typedef struct drm_i915_private {
 	};
 	u32 gt_irq_mask;
 	u32 pm_irq_mask;
+	u32 pm_rps_events;
+	u32 pipestat_irq_mask[I915_MAX_PIPES];
 
 	struct work_struct hotplug_work;
 	bool enable_hotplug_processing;
@@ -1358,23 +1403,16 @@ typedef struct drm_i915_private {
 	u32 hpd_event_bits;
 	struct timer_list hotplug_reenable_timer;
 
-	int num_plane;
-
 	struct i915_fbc fbc;
+	struct i915_drrs drrs;
 	struct intel_opregion opregion;
 	struct intel_vbt_data vbt;
 
 	/* overlay */
 	struct intel_overlay *overlay;
-	unsigned int sprite_scaling_enabled;
 
-	/* backlight */
-	struct {
-		int level;
-		bool enabled;
-		spinlock_t lock; /* bl registers and the above bl fields */
-		struct backlight_device *device;
-	} backlight;
+	/* backlight registers and fields in struct intel_panel */
+	spinlock_t backlight_lock;
 
 	/* LVDS info */
 	bool no_aux_handshake;
@@ -1384,6 +1422,7 @@ typedef struct drm_i915_private {
 	int num_fence_regs; /* 8 on pre-965, 16 otherwise */
 
 	unsigned int fsb_freq, mem_freq, is_ddr3;
+	unsigned int vlv_cdclk_freq;
 
 	/**
 	 * wq - Driver workqueue for GEM.
@@ -1407,16 +1446,19 @@ typedef struct drm_i915_private {
 	struct mutex modeset_restore_lock;
 
 	struct list_head vm_list; /* Global list of all address spaces */
-	struct i915_gtt gtt; /* VMA representing the global address space */
+	struct i915_gtt gtt; /* VM representing the global address space */
 
 	struct i915_gem_mm mm;
+#if defined(CONFIG_MMU_NOTIFIER)
+	DECLARE_HASHTABLE(mmu_notifiers, 7);
+#endif
 
 	/* Kernel Modesetting */
 
 	struct sdvo_device_mapping sdvo_mappings[2];
 
-	struct drm_crtc *plane_to_crtc_mapping[3];
-	struct drm_crtc *pipe_to_crtc_mapping[3];
+	struct drm_crtc *plane_to_crtc_mapping[I915_MAX_PIPES];
+	struct drm_crtc *pipe_to_crtc_mapping[I915_MAX_PIPES];
 	wait_queue_head_t pending_flip_queue;
 
 #ifdef CONFIG_DEBUG_FS
@@ -1426,6 +1468,7 @@ typedef struct drm_i915_private {
 	int num_shared_dpll;
 	struct intel_shared_dpll shared_dplls[I915_NUM_PLLS];
 	struct intel_ddi_plls ddi_plls;
+	int dpio_phy_iosf_port[I915_NUM_PHYS_VLV];
 
 	/* Reclocking support */
 	bool render_reclock_avail;
@@ -1470,13 +1513,14 @@ typedef struct drm_i915_private {
 	struct drm_property *broadcast_rgb_property;
 	struct drm_property *force_audio_property;
 
-	bool hw_contexts_disabled;
 	uint32_t hw_context_size;
 	struct list_head context_list;
 
 	u32 fdi_rx_config;
 
+	u32 suspend_count;
 	struct i915_suspend_saved_registers regfile;
+	struct vlv_s0ix_state vlv_s0ix_state;
 
 	struct {
 		/*
@@ -1492,17 +1536,22 @@ typedef struct drm_i915_private {
 		uint16_t cur_latency[5];
 
 		/* current hardware state */
-		struct hsw_wm_values hw;
+		struct ilk_wm_values hw;
 	} wm;
 
-	struct i915_package_c8 pc8;
+	struct i915_runtime_pm pm;
 
 	/* Old dri1 support infrastructure, beware the dragons ya fools entering
 	 * here! */
 	struct i915_dri1_state dri1;
 	/* Old ums support infrastructure, same warning applies. */
 	struct i915_ums_state ums;
-} drm_i915_private_t;
+
+	/*
+	 * NOTE: This is the dri1/ums dungeon, don't add stuff here. Your patch
+	 * will be rejected. Instead look for a better place.
+	 */
+};
 
 static inline struct drm_i915_private *to_i915(const struct drm_device *dev)
 {
@@ -1539,6 +1588,8 @@ struct drm_i915_gem_object_ops {
 	 */
 	int (*get_pages)(struct drm_i915_gem_object *);
 	void (*put_pages)(struct drm_i915_gem_object *);
+	int (*dmabuf_export)(struct drm_i915_gem_object *);
+	void (*release)(struct drm_i915_gem_object *);
 };
 
 struct drm_i915_gem_object {
@@ -1595,18 +1646,6 @@ struct drm_i915_gem_object {
 	 */
 	unsigned int fence_dirty:1;
 
-	/** How many users have pinned this object in GTT space. The following
-	 * users can each hold at most one reference: pwrite/pread, pin_ioctl
-	 * (via user_pin_count), execbuffer (objects are not allowed multiple
-	 * times for the same batchbuffer), and the framebuffer code. When
-	 * switching/pageflipping, the framebuffer code has at most two buffers
-	 * pinned per crtc.
-	 *
-	 * In the worst case this is 1 + 1 + 1 + 2*2 = 7. That would fit into 3
-	 * bits with absolutely no headroom. So use 4 bits. */
-	unsigned int pin_count:4;
-#define DRM_I915_GEM_OBJECT_MAX_PIN_COUNT 0xf
-
 	/**
 	 * Is the object at the current location in the gtt mappable and
 	 * fenceable? Used to avoid costly recalculations.
@@ -1641,7 +1680,7 @@ struct drm_i915_gem_object {
 	void *dma_buf_vmapping;
 	int vmapping_count;
 
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 
 	/** Breadcrumb of last rendering to the buffer. */
 	uint32_t last_read_seqno;
@@ -1663,10 +1702,21 @@ struct drm_i915_gem_object {
 	struct drm_file *pin_filp;
 
 	/** for phy allocated objects */
-	struct drm_i915_gem_phys_object *phys_obj;
-};
-#define to_gem_object(obj) (&((struct drm_i915_gem_object *)(obj))->base)
+	drm_dma_handle_t *phys_handle;
 
+	union {
+		struct i915_gem_userptr {
+			uintptr_t ptr;
+			unsigned read_only :1;
+			unsigned workers :4;
+#define I915_GEM_USERPTR_MAX_WORKERS 15
+
+			struct mm_struct *mm;
+			struct i915_mmu_object *mn;
+			struct work_struct *work;
+		} userptr;
+	};
+};
 #define to_intel_bo(x) container_of(x, struct drm_i915_gem_object, base)
 
 /**
@@ -1681,7 +1731,7 @@ struct drm_i915_gem_object {
  */
 struct drm_i915_gem_request {
 	/** On Which ring this request was generated */
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 
 	/** GEM sequence number associated with this request. */
 	uint32_t seqno;
@@ -1693,7 +1743,7 @@ struct drm_i915_gem_request {
 	u32 tail;
 
 	/** Context related to this request */
-	struct i915_hw_context *ctx;
+	struct intel_context *ctx;
 
 	/** Batch buffer related to this request if any */
 	struct drm_i915_gem_object *batch_obj;
@@ -1711,6 +1761,7 @@ struct drm_i915_gem_request {
 
 struct drm_i915_file_private {
 	struct drm_i915_private *dev_priv;
+	struct drm_file *file;
 
 	struct {
 		spinlock_t lock;
@@ -1719,11 +1770,101 @@ struct drm_i915_file_private {
 	} mm;
 	struct idr context_idr;
 
-	struct i915_ctx_hang_stats hang_stats;
 	atomic_t rps_wait_boost;
+	struct  intel_engine_cs *bsd_ring;
 };
 
-#define INTEL_INFO(dev)	(to_i915(dev)->info)
+/*
+ * A command that requires special handling by the command parser.
+ */
+struct drm_i915_cmd_descriptor {
+	/*
+	 * Flags describing how the command parser processes the command.
+	 *
+	 * CMD_DESC_FIXED: The command has a fixed length if this is set,
+	 *                 a length mask if not set
+	 * CMD_DESC_SKIP: The command is allowed but does not follow the
+	 *                standard length encoding for the opcode range in
+	 *                which it falls
+	 * CMD_DESC_REJECT: The command is never allowed
+	 * CMD_DESC_REGISTER: The command should be checked against the
+	 *                    register whitelist for the appropriate ring
+	 * CMD_DESC_MASTER: The command is allowed if the submitting process
+	 *                  is the DRM master
+	 */
+	u32 flags;
+#define CMD_DESC_FIXED    (1<<0)
+#define CMD_DESC_SKIP     (1<<1)
+#define CMD_DESC_REJECT   (1<<2)
+#define CMD_DESC_REGISTER (1<<3)
+#define CMD_DESC_BITMASK  (1<<4)
+#define CMD_DESC_MASTER   (1<<5)
+
+	/*
+	 * The command's unique identification bits and the bitmask to get them.
+	 * This isn't strictly the opcode field as defined in the spec and may
+	 * also include type, subtype, and/or subop fields.
+	 */
+	struct {
+		u32 value;
+		u32 mask;
+	} cmd;
+
+	/*
+	 * The command's length. The command is either fixed length (i.e. does
+	 * not include a length field) or has a length field mask. The flag
+	 * CMD_DESC_FIXED indicates a fixed length. Otherwise, the command has
+	 * a length mask. All command entries in a command table must include
+	 * length information.
+	 */
+	union {
+		u32 fixed;
+		u32 mask;
+	} length;
+
+	/*
+	 * Describes where to find a register address in the command to check
+	 * against the ring's register whitelist. Only valid if flags has the
+	 * CMD_DESC_REGISTER bit set.
+	 */
+	struct {
+		u32 offset;
+		u32 mask;
+	} reg;
+
+#define MAX_CMD_DESC_BITMASKS 3
+	/*
+	 * Describes command checks where a particular dword is masked and
+	 * compared against an expected value. If the command does not match
+	 * the expected value, the parser rejects it. Only valid if flags has
+	 * the CMD_DESC_BITMASK bit set. Only entries where mask is non-zero
+	 * are valid.
+	 *
+	 * If the check specifies a non-zero condition_mask then the parser
+	 * only performs the check when the bits specified by condition_mask
+	 * are non-zero.
+	 */
+	struct {
+		u32 offset;
+		u32 mask;
+		u32 expected;
+		u32 condition_offset;
+		u32 condition_mask;
+	} bits[MAX_CMD_DESC_BITMASKS];
+};
+
+/*
+ * A table of commands requiring special handling by the command parser.
+ *
+ * Each ring has an array of tables. Each table consists of an array of command
+ * descriptors, which must be sorted with command opcodes in ascending order.
+ */
+struct drm_i915_cmd_table {
+	const struct drm_i915_cmd_descriptor *table;
+	int count;
+};
+
+#define INTEL_INFO(dev)	(&to_i915(dev)->info)
 
 #define IS_I830(dev)		((dev)->pdev->device == 0x3577)
 #define IS_845G(dev)		((dev)->pdev->device == 0x2562)
@@ -1750,8 +1891,9 @@ struct drm_i915_file_private {
 				 (dev)->pdev->device == 0x0106 || \
 				 (dev)->pdev->device == 0x010A)
 #define IS_VALLEYVIEW(dev)	(INTEL_INFO(dev)->is_valleyview)
+#define IS_CHERRYVIEW(dev)	(INTEL_INFO(dev)->is_valleyview && IS_GEN8(dev))
 #define IS_HASWELL(dev)	(INTEL_INFO(dev)->is_haswell)
-#define IS_BROADWELL(dev)	(INTEL_INFO(dev)->gen == 8)
+#define IS_BROADWELL(dev)	(!INTEL_INFO(dev)->is_valleyview && IS_GEN8(dev))
 #define IS_MOBILE(dev)		(INTEL_INFO(dev)->is_mobile)
 #define IS_HSW_EARLY_SDV(dev)	(IS_HASWELL(dev) && \
 				 ((dev)->pdev->device & 0xFF00) == 0x0C00)
@@ -1764,6 +1906,9 @@ struct drm_i915_file_private {
 #define IS_ULT(dev)		(IS_HSW_ULT(dev) || IS_BDW_ULT(dev))
 #define IS_HSW_GT3(dev)		(IS_HASWELL(dev) && \
 				 ((dev)->pdev->device & 0x00F0) == 0x0020)
+/* ULX machines are also considered ULT. */
+#define IS_HSW_ULX(dev)		((dev)->pdev->device == 0x0A0E || \
+				 (dev)->pdev->device == 0x0A1E)
 #define IS_PRELIMINARY_HW(intel_info) ((intel_info)->is_preliminary)
 
 /*
@@ -1784,21 +1929,37 @@ struct drm_i915_file_private {
 #define BSD_RING		(1<<VCS)
 #define BLT_RING		(1<<BCS)
 #define VEBOX_RING		(1<<VECS)
-#define HAS_BSD(dev)            (INTEL_INFO(dev)->ring_mask & BSD_RING)
-#define HAS_BLT(dev)            (INTEL_INFO(dev)->ring_mask & BLT_RING)
-#define HAS_VEBOX(dev)            (INTEL_INFO(dev)->ring_mask & VEBOX_RING)
-#define HAS_LLC(dev)            (INTEL_INFO(dev)->has_llc)
-#define HAS_WT(dev)            (IS_HASWELL(dev) && to_i915(dev)->ellc_size)
+#define BSD2_RING		(1<<VCS2)
+#define HAS_BSD(dev)		(INTEL_INFO(dev)->ring_mask & BSD_RING)
+#define HAS_BSD2(dev)		(INTEL_INFO(dev)->ring_mask & BSD2_RING)
+#define HAS_BLT(dev)		(INTEL_INFO(dev)->ring_mask & BLT_RING)
+#define HAS_VEBOX(dev)		(INTEL_INFO(dev)->ring_mask & VEBOX_RING)
+#define HAS_LLC(dev)		(INTEL_INFO(dev)->has_llc)
+#define HAS_WT(dev)		((IS_HASWELL(dev) || IS_BROADWELL(dev)) && \
+				 to_i915(dev)->ellc_size)
 #define I915_NEED_GFX_HWS(dev)	(INTEL_INFO(dev)->need_gfx_hws)
 
 #define HAS_HW_CONTEXTS(dev)	(INTEL_INFO(dev)->gen >= 6)
-#define HAS_ALIASING_PPGTT(dev)	(INTEL_INFO(dev)->gen >=6 && !IS_VALLEYVIEW(dev))
+#define HAS_ALIASING_PPGTT(dev)	(INTEL_INFO(dev)->gen >= 6 && \
+				 (!IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)))
+#define HAS_PPGTT(dev)		(INTEL_INFO(dev)->gen >= 7 \
+				 && !IS_GEN8(dev))
+#define USES_PPGTT(dev)		intel_enable_ppgtt(dev, false)
+#define USES_FULL_PPGTT(dev)	intel_enable_ppgtt(dev, true)
 
 #define HAS_OVERLAY(dev)		(INTEL_INFO(dev)->has_overlay)
 #define OVERLAY_NEEDS_PHYSICAL(dev)	(INTEL_INFO(dev)->overlay_needs_physical)
 
 /* Early gen2 have a totally busted CS tlb and require pinned batches. */
 #define HAS_BROKEN_CS_TLB(dev)		(IS_I830(dev) || IS_845G(dev))
+/*
+ * dp aux and gmbus irq on gen4 seems to be able to generate legacy interrupts
+ * even when in MSI mode. This results in spurious interrupt warnings if the
+ * legacy irq no. is shared with another device. The kernel then disables that
+ * interrupt source and so prevents the other device from working properly.
+ */
+#define HAS_AUX_IRQ(dev) (INTEL_INFO(dev)->gen >= 5)
+#define HAS_GMBUS_IRQ(dev) (INTEL_INFO(dev)->gen >= 5)
 
 /* With the 945 and later, Y tiling got adjusted so that it was 32 128-byte
  * rows, which changed the alignment requirements and fence programming.
@@ -1813,15 +1974,15 @@ struct drm_i915_file_private {
 
 #define HAS_FW_BLC(dev) (INTEL_INFO(dev)->gen > 2)
 #define HAS_PIPE_CXSR(dev) (INTEL_INFO(dev)->has_pipe_cxsr)
-#define I915_HAS_FBC(dev) (INTEL_INFO(dev)->has_fbc)
+#define HAS_FBC(dev) (INTEL_INFO(dev)->has_fbc)
 
 #define HAS_IPS(dev)		(IS_ULT(dev) || IS_BROADWELL(dev))
 
 #define HAS_DDI(dev)		(INTEL_INFO(dev)->has_ddi)
-#define HAS_POWER_WELL(dev)	(IS_HASWELL(dev) || IS_BROADWELL(dev))
 #define HAS_FPGA_DBG_UNCLAIMED(dev)	(INTEL_INFO(dev)->has_fpga_dbg)
 #define HAS_PSR(dev)		(IS_HASWELL(dev) || IS_BROADWELL(dev))
-#define HAS_PC8(dev)		(IS_HASWELL(dev)) /* XXX HSW:ULX */
+#define HAS_RUNTIME_PM(dev)	(IS_GEN6(dev) || IS_HASWELL(dev) || \
+				 IS_BROADWELL(dev) || IS_VALLEYVIEW(dev))
 
 #define INTEL_PCH_DEVICE_ID_MASK		0xff00
 #define INTEL_PCH_IBX_DEVICE_ID_TYPE		0x3b00
@@ -1847,32 +2008,41 @@ struct drm_i915_file_private {
 
 extern const struct drm_ioctl_desc i915_ioctls[];
 extern int i915_max_ioctl;
-extern unsigned int i915_fbpercrtc __always_unused;
-extern int i915_panel_ignore_lid __read_mostly;
-extern unsigned int i915_powersave __read_mostly;
-extern int i915_semaphores __read_mostly;
-extern unsigned int i915_lvds_downclock __read_mostly;
-extern int i915_lvds_channel_mode __read_mostly;
-extern int i915_panel_use_ssc __read_mostly;
-extern int i915_vbt_sdvo_panel_type __read_mostly;
-extern int i915_enable_rc6 __read_mostly;
-extern int i915_enable_fbc __read_mostly;
-extern bool i915_enable_hangcheck __read_mostly;
-extern int i915_enable_ppgtt __read_mostly;
-extern int i915_enable_psr __read_mostly;
-extern unsigned int i915_preliminary_hw_support __read_mostly;
-extern int i915_disable_power_well __read_mostly;
-extern int i915_enable_ips __read_mostly;
-extern bool i915_fastboot __read_mostly;
-extern int i915_enable_pc8 __read_mostly;
-extern int i915_pc8_timeout __read_mostly;
-extern bool i915_prefault_disable __read_mostly;
 
 extern int i915_suspend(struct drm_device *dev, pm_message_t state);
 extern int i915_resume(struct drm_device *dev);
 extern int i915_master_create(struct drm_device *dev, struct drm_master *master);
 extern void i915_master_destroy(struct drm_device *dev, struct drm_master *master);
 
+/* i915_params.c */
+struct i915_params {
+	int modeset;
+	int panel_ignore_lid;
+	unsigned int powersave;
+	int semaphores;
+	unsigned int lvds_downclock;
+	int lvds_channel_mode;
+	int panel_use_ssc;
+	int vbt_sdvo_panel_type;
+	int enable_rc6;
+	int enable_fbc;
+	int enable_ppgtt;
+	int enable_psr;
+	unsigned int preliminary_hw_support;
+	int disable_power_well;
+	int enable_ips;
+	int invert_brightness;
+	int enable_cmd_parser;
+	/* leave bools at the end to not create holes */
+	bool enable_hangcheck;
+	bool fastboot;
+	bool prefault_disable;
+	bool reset;
+	bool disable_display;
+	bool disable_vtd_wa;
+};
+extern struct i915_params i915 __read_mostly;
+
 				/* i915_dma.c */
 void i915_update_dri1_breadcrumb(struct drm_device *dev);
 extern void i915_kernel_lost_context(struct drm_device * dev);
@@ -1898,28 +2068,37 @@ extern unsigned long i915_chipset_val(struct drm_i915_private *dev_priv);
 extern unsigned long i915_mch_val(struct drm_i915_private *dev_priv);
 extern unsigned long i915_gfx_val(struct drm_i915_private *dev_priv);
 extern void i915_update_gfx_val(struct drm_i915_private *dev_priv);
+int vlv_force_gfx_clock(struct drm_i915_private *dev_priv, bool on);
 
 extern void intel_console_resume(struct work_struct *work);
 
 /* i915_irq.c */
 void i915_queue_hangcheck(struct drm_device *dev);
-void i915_handle_error(struct drm_device *dev, bool wedged);
+__printf(3, 4)
+void i915_handle_error(struct drm_device *dev, bool wedged,
+		       const char *fmt, ...);
 
+void gen6_set_pm_mask(struct drm_i915_private *dev_priv, u32 pm_iir,
+							int new_delay);
 extern void intel_irq_init(struct drm_device *dev);
 extern void intel_hpd_init(struct drm_device *dev);
 
 extern void intel_uncore_sanitize(struct drm_device *dev);
 extern void intel_uncore_early_sanitize(struct drm_device *dev);
 extern void intel_uncore_init(struct drm_device *dev);
-extern void intel_uncore_clear_errors(struct drm_device *dev);
 extern void intel_uncore_check_errors(struct drm_device *dev);
 extern void intel_uncore_fini(struct drm_device *dev);
 
 void
-i915_enable_pipestat(drm_i915_private_t *dev_priv, enum pipe pipe, u32 mask);
+i915_enable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
+		     u32 status_mask);
 
 void
-i915_disable_pipestat(drm_i915_private_t *dev_priv, enum pipe pipe, u32 mask);
+i915_disable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
+		      u32 status_mask);
+
+void valleyview_enable_display_irqs(struct drm_i915_private *dev_priv);
+void valleyview_disable_display_irqs(struct drm_i915_private *dev_priv);
 
 /* i915_gem.c */
 int i915_gem_init_ioctl(struct drm_device *dev, void *data,
@@ -1964,6 +2143,9 @@ int i915_gem_set_tiling(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
 int i915_gem_get_tiling(struct drm_device *dev, void *data,
 			struct drm_file *file_priv);
+int i915_gem_init_userptr(struct drm_device *dev);
+int i915_gem_userptr_ioctl(struct drm_device *dev, void *data,
+			   struct drm_file *file);
 int i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 				struct drm_file *file_priv);
 int i915_gem_wait_ioctl(struct drm_device *dev, void *data,
@@ -1975,21 +2157,29 @@ void i915_gem_object_init(struct drm_i915_gem_object *obj,
 			 const struct drm_i915_gem_object_ops *ops);
 struct drm_i915_gem_object *i915_gem_alloc_object(struct drm_device *dev,
 						  size_t size);
+void i915_init_vm(struct drm_i915_private *dev_priv,
+		  struct i915_address_space *vm);
 void i915_gem_free_object(struct drm_gem_object *obj);
 void i915_gem_vma_destroy(struct i915_vma *vma);
 
+#define PIN_MAPPABLE 0x1
+#define PIN_NONBLOCK 0x2
+#define PIN_GLOBAL 0x4
+#define PIN_OFFSET_BIAS 0x8
+#define PIN_OFFSET_MASK (~4095)
 int __must_check i915_gem_object_pin(struct drm_i915_gem_object *obj,
 				     struct i915_address_space *vm,
 				     uint32_t alignment,
-				     bool map_and_fenceable,
-				     bool nonblocking);
-void i915_gem_object_unpin(struct drm_i915_gem_object *obj);
+				     uint64_t flags);
 int __must_check i915_vma_unbind(struct i915_vma *vma);
-int __must_check i915_gem_object_ggtt_unbind(struct drm_i915_gem_object *obj);
 int i915_gem_object_put_pages(struct drm_i915_gem_object *obj);
+void i915_gem_release_all_mmaps(struct drm_i915_private *dev_priv);
 void i915_gem_release_mmap(struct drm_i915_gem_object *obj);
 void i915_gem_lastclose(struct drm_device *dev);
 
+int i915_gem_obj_prepare_shmem_read(struct drm_i915_gem_object *obj,
+				    int *needs_clflush);
+
 int __must_check i915_gem_object_get_pages(struct drm_i915_gem_object *obj);
 static inline struct page *i915_gem_object_get_page(struct drm_i915_gem_object *obj, int n)
 {
@@ -2013,9 +2203,9 @@ static inline void i915_gem_object_unpin_pages(struct drm_i915_gem_object *obj)
 
 int __must_check i915_mutex_lock_interruptible(struct drm_device *dev);
 int i915_gem_object_sync(struct drm_i915_gem_object *obj,
-			 struct intel_ring_buffer *to);
+			 struct intel_engine_cs *to);
 void i915_vma_move_to_active(struct i915_vma *vma,
-			     struct intel_ring_buffer *ring);
+			     struct intel_engine_cs *ring);
 int i915_gem_dumb_create(struct drm_file *file_priv,
 			 struct drm_device *dev,
 			 struct drm_mode_create_dumb *args);
@@ -2035,40 +2225,42 @@ int __must_check i915_gem_set_seqno(struct drm_device *dev, u32 seqno);
 int __must_check i915_gem_object_get_fence(struct drm_i915_gem_object *obj);
 int __must_check i915_gem_object_put_fence(struct drm_i915_gem_object *obj);
 
-static inline bool
-i915_gem_object_pin_fence(struct drm_i915_gem_object *obj)
-{
-	if (obj->fence_reg != I915_FENCE_REG_NONE) {
-		struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
-		dev_priv->fence_regs[obj->fence_reg].pin_count++;
-		return true;
-	} else
-		return false;
-}
+bool i915_gem_object_pin_fence(struct drm_i915_gem_object *obj);
+void i915_gem_object_unpin_fence(struct drm_i915_gem_object *obj);
 
-static inline void
-i915_gem_object_unpin_fence(struct drm_i915_gem_object *obj)
-{
-	if (obj->fence_reg != I915_FENCE_REG_NONE) {
-		struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
-		WARN_ON(dev_priv->fence_regs[obj->fence_reg].pin_count <= 0);
-		dev_priv->fence_regs[obj->fence_reg].pin_count--;
-	}
-}
+struct drm_i915_gem_request *
+i915_gem_find_active_request(struct intel_engine_cs *ring);
 
 bool i915_gem_retire_requests(struct drm_device *dev);
-void i915_gem_retire_requests_ring(struct intel_ring_buffer *ring);
+void i915_gem_retire_requests_ring(struct intel_engine_cs *ring);
 int __must_check i915_gem_check_wedge(struct i915_gpu_error *error,
 				      bool interruptible);
 static inline bool i915_reset_in_progress(struct i915_gpu_error *error)
 {
 	return unlikely(atomic_read(&error->reset_counter)
-			& I915_RESET_IN_PROGRESS_FLAG);
+			& (I915_RESET_IN_PROGRESS_FLAG | I915_WEDGED));
 }
 
 static inline bool i915_terminally_wedged(struct i915_gpu_error *error)
 {
-	return atomic_read(&error->reset_counter) == I915_WEDGED;
+	return atomic_read(&error->reset_counter) & I915_WEDGED;
+}
+
+static inline u32 i915_reset_count(struct i915_gpu_error *error)
+{
+	return ((atomic_read(&error->reset_counter) & ~I915_WEDGED) + 1) / 2;
+}
+
+static inline bool i915_stop_ring_allow_ban(struct drm_i915_private *dev_priv)
+{
+	return dev_priv->gpu_error.stop_rings == 0 ||
+		dev_priv->gpu_error.stop_rings & I915_STOP_RING_ALLOW_BAN;
+}
+
+static inline bool i915_stop_ring_allow_warn(struct drm_i915_private *dev_priv)
+{
+	return dev_priv->gpu_error.stop_rings == 0 ||
+		dev_priv->gpu_error.stop_rings & I915_STOP_RING_ALLOW_WARN;
 }
 
 void i915_gem_reset(struct drm_device *dev);
@@ -2076,18 +2268,18 @@ bool i915_gem_clflush_object(struct drm_i915_gem_object *obj, bool force);
 int __must_check i915_gem_object_finish_gpu(struct drm_i915_gem_object *obj);
 int __must_check i915_gem_init(struct drm_device *dev);
 int __must_check i915_gem_init_hw(struct drm_device *dev);
-int i915_gem_l3_remap(struct intel_ring_buffer *ring, int slice);
+int i915_gem_l3_remap(struct intel_engine_cs *ring, int slice);
 void i915_gem_init_swizzling(struct drm_device *dev);
 void i915_gem_cleanup_ringbuffer(struct drm_device *dev);
 int __must_check i915_gpu_idle(struct drm_device *dev);
 int __must_check i915_gem_suspend(struct drm_device *dev);
-int __i915_add_request(struct intel_ring_buffer *ring,
+int __i915_add_request(struct intel_engine_cs *ring,
 		       struct drm_file *file,
 		       struct drm_i915_gem_object *batch_obj,
 		       u32 *seqno);
 #define i915_add_request(ring, seqno) \
 	__i915_add_request(ring, NULL, NULL, seqno)
-int __must_check i915_wait_seqno(struct intel_ring_buffer *ring,
+int __must_check i915_wait_seqno(struct intel_engine_cs *ring,
 				 uint32_t seqno);
 int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
 int __must_check
@@ -2098,15 +2290,10 @@ i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write);
 int __must_check
 i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 				     u32 alignment,
-				     struct intel_ring_buffer *pipelined);
+				     struct intel_engine_cs *pipelined);
 void i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj);
-int i915_gem_attach_phys_object(struct drm_device *dev,
-				struct drm_i915_gem_object *obj,
-				int id,
+int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj,
 				int align);
-void i915_gem_detach_phys_object(struct drm_device *dev,
-				 struct drm_i915_gem_object *obj);
-void i915_gem_free_all_phys_object(struct drm_device *dev);
 int i915_gem_open(struct drm_device *dev, struct drm_file *file);
 void i915_gem_release(struct drm_device *dev, struct drm_file *file);
 
@@ -2141,6 +2328,13 @@ i915_gem_obj_lookup_or_create_vma(struct drm_i915_gem_object *obj,
 				  struct i915_address_space *vm);
 
 struct i915_vma *i915_gem_obj_to_ggtt(struct drm_i915_gem_object *obj);
+static inline bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj) {
+	struct i915_vma *vma;
+	list_for_each_entry(vma, &obj->vma_list, vma_link)
+		if (vma->pin_count > 0)
+			return true;
+	return false;
+}
 
 /* Some GGTT VM helpers */
 #define obj_to_ggtt(obj) \
@@ -2172,77 +2366,73 @@ i915_gem_obj_ggtt_size(struct drm_i915_gem_object *obj)
 static inline int __must_check
 i915_gem_obj_ggtt_pin(struct drm_i915_gem_object *obj,
 		      uint32_t alignment,
-		      bool map_and_fenceable,
-		      bool nonblocking)
+		      unsigned flags)
 {
-	return i915_gem_object_pin(obj, obj_to_ggtt(obj), alignment,
-				   map_and_fenceable, nonblocking);
+	return i915_gem_object_pin(obj, obj_to_ggtt(obj), alignment, flags | PIN_GLOBAL);
 }
 
+static inline int
+i915_gem_object_ggtt_unbind(struct drm_i915_gem_object *obj)
+{
+	return i915_vma_unbind(i915_gem_obj_to_ggtt(obj));
+}
+
+void i915_gem_object_ggtt_unpin(struct drm_i915_gem_object *obj);
+
 /* i915_gem_context.c */
-void i915_gem_context_init(struct drm_device *dev);
+#define ctx_to_ppgtt(ctx) container_of((ctx)->vm, struct i915_hw_ppgtt, base)
+int __must_check i915_gem_context_init(struct drm_device *dev);
 void i915_gem_context_fini(struct drm_device *dev);
+void i915_gem_context_reset(struct drm_device *dev);
+int i915_gem_context_open(struct drm_device *dev, struct drm_file *file);
+int i915_gem_context_enable(struct drm_i915_private *dev_priv);
 void i915_gem_context_close(struct drm_device *dev, struct drm_file *file);
-int i915_switch_context(struct intel_ring_buffer *ring,
-			struct drm_file *file, int to_id);
+int i915_switch_context(struct intel_engine_cs *ring,
+			struct intel_context *to);
+struct intel_context *
+i915_gem_context_get(struct drm_i915_file_private *file_priv, u32 id);
 void i915_gem_context_free(struct kref *ctx_ref);
-static inline void i915_gem_context_reference(struct i915_hw_context *ctx)
+static inline void i915_gem_context_reference(struct intel_context *ctx)
 {
 	kref_get(&ctx->ref);
 }
 
-static inline void i915_gem_context_unreference(struct i915_hw_context *ctx)
+static inline void i915_gem_context_unreference(struct intel_context *ctx)
 {
 	kref_put(&ctx->ref, i915_gem_context_free);
 }
 
-struct i915_ctx_hang_stats * __must_check
-i915_gem_context_get_hang_stats(struct drm_device *dev,
-				struct drm_file *file,
-				u32 id);
+static inline bool i915_gem_context_is_default(const struct intel_context *c)
+{
+	return c->id == DEFAULT_CONTEXT_ID;
+}
+
 int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file);
 int i915_gem_context_destroy_ioctl(struct drm_device *dev, void *data,
 				   struct drm_file *file);
 
-/* i915_gem_gtt.c */
-void i915_gem_cleanup_aliasing_ppgtt(struct drm_device *dev);
-void i915_ppgtt_bind_object(struct i915_hw_ppgtt *ppgtt,
-			    struct drm_i915_gem_object *obj,
-			    enum i915_cache_level cache_level);
-void i915_ppgtt_unbind_object(struct i915_hw_ppgtt *ppgtt,
-			      struct drm_i915_gem_object *obj);
-
-void i915_check_and_clear_faults(struct drm_device *dev);
-void i915_gem_suspend_gtt_mappings(struct drm_device *dev);
-void i915_gem_restore_gtt_mappings(struct drm_device *dev);
-int __must_check i915_gem_gtt_prepare_object(struct drm_i915_gem_object *obj);
-void i915_gem_gtt_bind_object(struct drm_i915_gem_object *obj,
-				enum i915_cache_level cache_level);
-void i915_gem_gtt_unbind_object(struct drm_i915_gem_object *obj);
-void i915_gem_gtt_finish_object(struct drm_i915_gem_object *obj);
-void i915_gem_init_global_gtt(struct drm_device *dev);
-void i915_gem_setup_global_gtt(struct drm_device *dev, unsigned long start,
-			       unsigned long mappable_end, unsigned long end);
-int i915_gem_gtt_init(struct drm_device *dev);
-static inline void i915_gem_chipset_flush(struct drm_device *dev)
-{
-	if (INTEL_INFO(dev)->gen < 6)
-		intel_gtt_chipset_flush();
-}
-
-
+/* i915_gem_render_state.c */
+int i915_gem_render_state_init(struct intel_engine_cs *ring);
 /* i915_gem_evict.c */
 int __must_check i915_gem_evict_something(struct drm_device *dev,
 					  struct i915_address_space *vm,
 					  int min_size,
 					  unsigned alignment,
 					  unsigned cache_level,
-					  bool mappable,
-					  bool nonblock);
+					  unsigned long start,
+					  unsigned long end,
+					  unsigned flags);
 int i915_gem_evict_vm(struct i915_address_space *vm, bool do_idle);
 int i915_gem_evict_everything(struct drm_device *dev);
 
+/* belongs in i915_gem_gtt.h */
+static inline void i915_gem_chipset_flush(struct drm_device *dev)
+{
+	if (INTEL_INFO(dev)->gen < 6)
+		intel_gtt_chipset_flush();
+}
+
 /* i915_gem_stolen.c */
 int i915_gem_init_stolen(struct drm_device *dev);
 int i915_gem_stolen_setup_compression(struct drm_device *dev, int size);
@@ -2260,7 +2450,7 @@ void i915_gem_object_release_stolen(struct drm_i915_gem_object *obj);
 /* i915_gem_tiling.c */
 static inline bool i915_gem_object_needs_bit17_swizzle(struct drm_i915_gem_object *obj)
 {
-	drm_i915_private_t *dev_priv = obj->base.dev->dev_private;
+	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
 
 	return dev_priv->mm.bit_6_swizzle_x == I915_BIT_6_SWIZZLE_9_10_17 &&
 		obj->tiling_mode != I915_TILING_NONE;
@@ -2298,7 +2488,8 @@ static inline void i915_error_state_buf_release(
 {
 	kfree(eb->buf);
 }
-void i915_capture_error_state(struct drm_device *dev);
+void i915_capture_error_state(struct drm_device *dev, bool wedge,
+			      const char *error_msg);
 void i915_error_state_get(struct drm_device *dev,
 			  struct i915_error_state_file_priv *error_priv);
 void i915_error_state_put(struct i915_error_state_file_priv *error_priv);
@@ -2307,6 +2498,16 @@ void i915_destroy_error_state(struct drm_device *dev);
 void i915_get_extra_instdone(struct drm_device *dev, uint32_t *instdone);
 const char *i915_cache_level_str(int type);
 
+/* i915_cmd_parser.c */
+int i915_cmd_parser_get_version(void);
+int i915_cmd_parser_init_ring(struct intel_engine_cs *ring);
+void i915_cmd_parser_fini_ring(struct intel_engine_cs *ring);
+bool i915_needs_cmd_parser(struct intel_engine_cs *ring);
+int i915_parse_cmds(struct intel_engine_cs *ring,
+		    struct drm_i915_gem_object *batch_obj,
+		    u32 batch_start_offset,
+		    bool is_master);
+
 /* i915_suspend.c */
 extern int i915_save_state(struct drm_device *dev);
 extern int i915_restore_state(struct drm_device *dev);
@@ -2339,8 +2540,8 @@ extern void intel_i2c_reset(struct drm_device *dev);
 
 /* intel_opregion.c */
 struct intel_encoder;
-extern int intel_opregion_setup(struct drm_device *dev);
 #ifdef CONFIG_ACPI
+extern int intel_opregion_setup(struct drm_device *dev);
 extern void intel_opregion_init(struct drm_device *dev);
 extern void intel_opregion_fini(struct drm_device *dev);
 extern void intel_opregion_asle_intr(struct drm_device *dev);
@@ -2349,6 +2550,7 @@ extern int intel_opregion_notify_encoder(struct intel_encoder *intel_encoder,
 extern int intel_opregion_notify_adapter(struct drm_device *dev,
 					 pci_power_t state);
 #else
+static inline int intel_opregion_setup(struct drm_device *dev) { return 0; }
 static inline void intel_opregion_init(struct drm_device *dev) { return; }
 static inline void intel_opregion_fini(struct drm_device *dev) { return; }
 static inline void intel_opregion_asle_intr(struct drm_device *dev) { return; }
@@ -2379,10 +2581,12 @@ extern void intel_modeset_suspend_hw(struct drm_device *dev);
 extern void intel_modeset_init(struct drm_device *dev);
 extern void intel_modeset_gem_init(struct drm_device *dev);
 extern void intel_modeset_cleanup(struct drm_device *dev);
+extern void intel_connector_unregister(struct intel_connector *);
 extern int intel_modeset_vga_set_state(struct drm_device *dev, bool state);
 extern void intel_modeset_setup_hw_state(struct drm_device *dev,
 					 bool force_restore);
 extern void i915_redisable_vga(struct drm_device *dev);
+extern void i915_redisable_vga_power_on(struct drm_device *dev);
 extern bool intel_fbc_enabled(struct drm_device *dev);
 extern void intel_disable_fbc(struct drm_device *dev);
 extern bool ironlake_set_drps(struct drm_device *dev, u8 val);
@@ -2398,6 +2602,8 @@ extern int intel_enable_rc6(const struct drm_device *dev);
 extern bool i915_semaphore_is_enabled(struct drm_device *dev);
 int i915_reg_read_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file);
+int i915_get_reset_stats_ioctl(struct drm_device *dev, void *data,
+			       struct drm_file *file);
 
 /* overlay */
 extern struct intel_overlay_error_state *intel_overlay_capture_error_state(struct drm_device *dev);
@@ -2413,8 +2619,9 @@ extern void intel_display_print_error_state(struct drm_i915_error_state_buf *e,
  * must be set to prevent GT core from power down and stale values being
  * returned.
  */
-void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv);
-void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv);
+void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine);
+void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine);
+void assert_force_wake_inactive(struct drm_i915_private *dev_priv);
 
 int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u8 mbox, u32 *val);
 int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val);
@@ -2429,6 +2636,8 @@ u32 vlv_cck_read(struct drm_i915_private *dev_priv, u32 reg);
 void vlv_cck_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
 u32 vlv_ccu_read(struct drm_i915_private *dev_priv, u32 reg);
 void vlv_ccu_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
+u32 vlv_bunit_read(struct drm_i915_private *dev_priv, u32 reg);
+void vlv_bunit_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
 u32 vlv_gps_core_read(struct drm_i915_private *dev_priv, u32 reg);
 void vlv_gps_core_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
 u32 vlv_dpio_read(struct drm_i915_private *dev_priv, enum pipe pipe, int reg);
@@ -2437,9 +2646,16 @@ u32 intel_sbi_read(struct drm_i915_private *dev_priv, u16 reg,
 		   enum intel_sbi_destination destination);
 void intel_sbi_write(struct drm_i915_private *dev_priv, u16 reg, u32 value,
 		     enum intel_sbi_destination destination);
+u32 vlv_flisdsi_read(struct drm_i915_private *dev_priv, u32 reg);
+void vlv_flisdsi_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
+
+int vlv_gpu_freq(struct drm_i915_private *dev_priv, int val);
+int vlv_freq_opcode(struct drm_i915_private *dev_priv, int val);
+
+#define FORCEWAKE_RENDER	(1 << 0)
+#define FORCEWAKE_MEDIA		(1 << 1)
+#define FORCEWAKE_ALL		(FORCEWAKE_RENDER | FORCEWAKE_MEDIA)
 
-int vlv_gpu_freq(int ddr_freq, int val);
-int vlv_freq_opcode(int ddr_freq, int val);
 
 #define I915_READ8(reg)		dev_priv->uncore.funcs.mmio_readb(dev_priv, (reg), true)
 #define I915_WRITE8(reg, val)	dev_priv->uncore.funcs.mmio_writeb(dev_priv, (reg), (val), true)
@@ -2454,9 +2670,26 @@ int vlv_freq_opcode(int ddr_freq, int val);
 #define I915_READ_NOTRACE(reg)		dev_priv->uncore.funcs.mmio_readl(dev_priv, (reg), false)
 #define I915_WRITE_NOTRACE(reg, val)	dev_priv->uncore.funcs.mmio_writel(dev_priv, (reg), (val), false)
 
+/* Be very careful with read/write 64-bit values. On 32-bit machines, they
+ * will be implemented using 2 32-bit writes in an arbitrary order with
+ * an arbitrary delay between them. This can cause the hardware to
+ * act upon the intermediate value, possibly leading to corruption and
+ * machine death. You have been warned.
+ */
 #define I915_WRITE64(reg, val)	dev_priv->uncore.funcs.mmio_writeq(dev_priv, (reg), (val), true)
 #define I915_READ64(reg)	dev_priv->uncore.funcs.mmio_readq(dev_priv, (reg), true)
 
+#define I915_READ64_2x32(lower_reg, upper_reg) ({			\
+		u32 upper = I915_READ(upper_reg);			\
+		u32 lower = I915_READ(lower_reg);			\
+		u32 tmp = I915_READ(upper_reg);				\
+		if (upper != tmp) {					\
+			upper = tmp;					\
+			lower = I915_READ(lower_reg);			\
+			WARN_ON(I915_READ(upper_reg) != upper);		\
+		}							\
+		(u64)upper << 32 | lower; })
+
 #define POSTING_READ(reg)	(void)I915_READ_NOTRACE(reg)
 #define POSTING_READ16(reg)	(void)I915_READ16_NOTRACE(reg)
 
@@ -2495,4 +2728,31 @@ timespec_to_jiffies_timeout(const struct timespec *value)
 	return min_t(unsigned long, MAX_JIFFY_OFFSET, j + 1);
 }
 
+/*
+ * If you need to wait X milliseconds between events A and B, but event B
+ * doesn't happen exactly after event A, you record the timestamp (jiffies) of
+ * when event A happened, then just before event B you call this function and
+ * pass the timestamp as the first argument, and X as the second argument.
+ */
+static inline void
+wait_remaining_ms_from_jiffies(unsigned long timestamp_jiffies, int to_wait_ms)
+{
+	unsigned long target_jiffies, tmp_jiffies, remaining_jiffies;
+
+	/*
+	 * Don't re-read the value of "jiffies" every time since it may change
+	 * behind our back and break the math.
+	 */
+	tmp_jiffies = jiffies;
+	target_jiffies = timestamp_jiffies +
+			 msecs_to_jiffies_timeout(to_wait_ms);
+
+	if (time_after(target_jiffies, tmp_jiffies)) {
+		remaining_jiffies = target_jiffies - tmp_jiffies;
+		while (remaining_jiffies)
+			remaining_jiffies =
+			    schedule_timeout_uninterruptible(remaining_jiffies);
+	}
+}
+
 #endif
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 76d3d1ab73c..d893e4da5dc 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -31,6 +31,7 @@
 #include "i915_drv.h"
 #include "i915_trace.h"
 #include "intel_drv.h"
+#include <linux/oom.h>
 #include <linux/shmem_fs.h>
 #include <linux/slab.h>
 #include <linux/swap.h>
@@ -43,16 +44,8 @@ static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *o
 static __must_check int
 i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
 			       bool readonly);
-static __must_check int
-i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj,
-			   struct i915_address_space *vm,
-			   unsigned alignment,
-			   bool map_and_fenceable,
-			   bool nonblocking);
-static int i915_gem_phys_pwrite(struct drm_device *dev,
-				struct drm_i915_gem_object *obj,
-				struct drm_i915_gem_pwrite *args,
-				struct drm_file *file);
+static void
+i915_gem_object_retire(struct drm_i915_gem_object *obj);
 
 static void i915_gem_write_fence(struct drm_device *dev, int reg,
 				 struct drm_i915_gem_object *obj);
@@ -60,13 +53,15 @@ static void i915_gem_object_update_fence(struct drm_i915_gem_object *obj,
 					 struct drm_i915_fence_reg *fence,
 					 bool enable);
 
-static unsigned long i915_gem_inactive_count(struct shrinker *shrinker,
+static unsigned long i915_gem_shrinker_count(struct shrinker *shrinker,
 					     struct shrink_control *sc);
-static unsigned long i915_gem_inactive_scan(struct shrinker *shrinker,
+static unsigned long i915_gem_shrinker_scan(struct shrinker *shrinker,
 					    struct shrink_control *sc);
+static int i915_gem_shrinker_oom(struct notifier_block *nb,
+				 unsigned long event,
+				 void *ptr);
 static unsigned long i915_gem_purge(struct drm_i915_private *dev_priv, long target);
 static unsigned long i915_gem_shrink_all(struct drm_i915_private *dev_priv);
-static void i915_gem_object_truncate(struct drm_i915_gem_object *obj);
 
 static bool cpu_cache_is_coherent(struct drm_device *dev,
 				  enum i915_cache_level level)
@@ -204,7 +199,7 @@ i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 	pinned = 0;
 	mutex_lock(&dev->struct_mutex);
 	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list)
-		if (obj->pin_count)
+		if (i915_gem_obj_is_pinned(obj))
 			pinned += i915_gem_obj_ggtt_size(obj);
 	mutex_unlock(&dev->struct_mutex);
 
@@ -214,6 +209,128 @@ i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 	return 0;
 }
 
+static void i915_gem_object_detach_phys(struct drm_i915_gem_object *obj)
+{
+	drm_dma_handle_t *phys = obj->phys_handle;
+
+	if (!phys)
+		return;
+
+	if (obj->madv == I915_MADV_WILLNEED) {
+		struct address_space *mapping = file_inode(obj->base.filp)->i_mapping;
+		char *vaddr = phys->vaddr;
+		int i;
+
+		for (i = 0; i < obj->base.size / PAGE_SIZE; i++) {
+			struct page *page = shmem_read_mapping_page(mapping, i);
+			if (!IS_ERR(page)) {
+				char *dst = kmap_atomic(page);
+				memcpy(dst, vaddr, PAGE_SIZE);
+				drm_clflush_virt_range(dst, PAGE_SIZE);
+				kunmap_atomic(dst);
+
+				set_page_dirty(page);
+				mark_page_accessed(page);
+				page_cache_release(page);
+			}
+			vaddr += PAGE_SIZE;
+		}
+		i915_gem_chipset_flush(obj->base.dev);
+	}
+
+#ifdef CONFIG_X86
+	set_memory_wb((unsigned long)phys->vaddr, phys->size / PAGE_SIZE);
+#endif
+	drm_pci_free(obj->base.dev, phys);
+	obj->phys_handle = NULL;
+}
+
+int
+i915_gem_object_attach_phys(struct drm_i915_gem_object *obj,
+			    int align)
+{
+	drm_dma_handle_t *phys;
+	struct address_space *mapping;
+	char *vaddr;
+	int i;
+
+	if (obj->phys_handle) {
+		if ((unsigned long)obj->phys_handle->vaddr & (align -1))
+			return -EBUSY;
+
+		return 0;
+	}
+
+	if (obj->madv != I915_MADV_WILLNEED)
+		return -EFAULT;
+
+	if (obj->base.filp == NULL)
+		return -EINVAL;
+
+	/* create a new object */
+	phys = drm_pci_alloc(obj->base.dev, obj->base.size, align);
+	if (!phys)
+		return -ENOMEM;
+
+	vaddr = phys->vaddr;
+#ifdef CONFIG_X86
+	set_memory_wc((unsigned long)vaddr, phys->size / PAGE_SIZE);
+#endif
+	mapping = file_inode(obj->base.filp)->i_mapping;
+	for (i = 0; i < obj->base.size / PAGE_SIZE; i++) {
+		struct page *page;
+		char *src;
+
+		page = shmem_read_mapping_page(mapping, i);
+		if (IS_ERR(page)) {
+#ifdef CONFIG_X86
+			set_memory_wb((unsigned long)phys->vaddr, phys->size / PAGE_SIZE);
+#endif
+			drm_pci_free(obj->base.dev, phys);
+			return PTR_ERR(page);
+		}
+
+		src = kmap_atomic(page);
+		memcpy(vaddr, src, PAGE_SIZE);
+		kunmap_atomic(src);
+
+		mark_page_accessed(page);
+		page_cache_release(page);
+
+		vaddr += PAGE_SIZE;
+	}
+
+	obj->phys_handle = phys;
+	return 0;
+}
+
+static int
+i915_gem_phys_pwrite(struct drm_i915_gem_object *obj,
+		     struct drm_i915_gem_pwrite *args,
+		     struct drm_file *file_priv)
+{
+	struct drm_device *dev = obj->base.dev;
+	void *vaddr = obj->phys_handle->vaddr + args->offset;
+	char __user *user_data = to_user_ptr(args->data_ptr);
+
+	if (__copy_from_user_inatomic_nocache(vaddr, user_data, args->size)) {
+		unsigned long unwritten;
+
+		/* The physical object once assigned is fixed for the lifetime
+		 * of the obj, so we can safely drop the lock and continue
+		 * to access vaddr.
+		 */
+		mutex_unlock(&dev->struct_mutex);
+		unwritten = copy_from_user(vaddr, user_data, args->size);
+		mutex_lock(&dev->struct_mutex);
+		if (unwritten)
+			return -EFAULT;
+	}
+
+	i915_gem_chipset_flush(dev);
+	return 0;
+}
+
 void *i915_gem_object_alloc(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -332,6 +449,44 @@ __copy_from_user_swizzled(char *gpu_vaddr, int gpu_offset,
 	return 0;
 }
 
+/*
+ * Pins the specified object's pages and synchronizes the object with
+ * GPU accesses. Sets needs_clflush to non-zero if the caller should
+ * flush the object from the CPU cache.
+ */
+int i915_gem_obj_prepare_shmem_read(struct drm_i915_gem_object *obj,
+				    int *needs_clflush)
+{
+	int ret;
+
+	*needs_clflush = 0;
+
+	if (!obj->base.filp)
+		return -EINVAL;
+
+	if (!(obj->base.read_domains & I915_GEM_DOMAIN_CPU)) {
+		/* If we're not in the cpu read domain, set ourself into the gtt
+		 * read domain and manually flush cachelines (if required). This
+		 * optimizes for the case when the gpu will dirty the data
+		 * anyway again before the next pread happens. */
+		*needs_clflush = !cpu_cache_is_coherent(obj->base.dev,
+							obj->cache_level);
+		ret = i915_gem_object_wait_rendering(obj, true);
+		if (ret)
+			return ret;
+
+		i915_gem_object_retire(obj);
+	}
+
+	ret = i915_gem_object_get_pages(obj);
+	if (ret)
+		return ret;
+
+	i915_gem_object_pin_pages(obj);
+
+	return ret;
+}
+
 /* Per-page copy function for the shmem pread fastpath.
  * Flushes invalid cachelines before reading the target if
  * needs_clflush is set. */
@@ -429,23 +584,10 @@ i915_gem_shmem_pread(struct drm_device *dev,
 
 	obj_do_bit17_swizzling = i915_gem_object_needs_bit17_swizzle(obj);
 
-	if (!(obj->base.read_domains & I915_GEM_DOMAIN_CPU)) {
-		/* If we're not in the cpu read domain, set ourself into the gtt
-		 * read domain and manually flush cachelines (if required). This
-		 * optimizes for the case when the gpu will dirty the data
-		 * anyway again before the next pread happens. */
-		needs_clflush = !cpu_cache_is_coherent(dev, obj->cache_level);
-		ret = i915_gem_object_wait_rendering(obj, true);
-		if (ret)
-			return ret;
-	}
-
-	ret = i915_gem_object_get_pages(obj);
+	ret = i915_gem_obj_prepare_shmem_read(obj, &needs_clflush);
 	if (ret)
 		return ret;
 
-	i915_gem_object_pin_pages(obj);
-
 	offset = args->offset;
 
 	for_each_sg_page(obj->pages->sgl, &sg_iter, obj->pages->nents,
@@ -476,7 +618,7 @@ i915_gem_shmem_pread(struct drm_device *dev,
 
 		mutex_unlock(&dev->struct_mutex);
 
-		if (likely(!i915_prefault_disable) && !prefaulted) {
+		if (likely(!i915.prefault_disable) && !prefaulted) {
 			ret = fault_in_multipages_writeable(user_data, remain);
 			/* Userspace is tricking us, but we've already clobbered
 			 * its pages with the prefault and promised to write the
@@ -492,12 +634,10 @@ i915_gem_shmem_pread(struct drm_device *dev,
 
 		mutex_lock(&dev->struct_mutex);
 
-next_page:
-		mark_page_accessed(page);
-
 		if (ret)
 			goto out;
 
+next_page:
 		remain -= page_length;
 		user_data += page_length;
 		offset += page_length;
@@ -599,13 +739,13 @@ i915_gem_gtt_pwrite_fast(struct drm_device *dev,
 			 struct drm_i915_gem_pwrite *args,
 			 struct drm_file *file)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	ssize_t remain;
 	loff_t offset, page_base;
 	char __user *user_data;
 	int page_offset, page_length, ret;
 
-	ret = i915_gem_obj_ggtt_pin(obj, 0, true, true);
+	ret = i915_gem_obj_ggtt_pin(obj, 0, PIN_MAPPABLE | PIN_NONBLOCK);
 	if (ret)
 		goto out;
 
@@ -651,7 +791,7 @@ i915_gem_gtt_pwrite_fast(struct drm_device *dev,
 	}
 
 out_unpin:
-	i915_gem_object_unpin(obj);
+	i915_gem_object_ggtt_unpin(obj);
 out:
 	return ret;
 }
@@ -677,9 +817,8 @@ shmem_pwrite_fast(struct page *page, int shmem_page_offset, int page_length,
 	if (needs_clflush_before)
 		drm_clflush_virt_range(vaddr + shmem_page_offset,
 				       page_length);
-	ret = __copy_from_user_inatomic_nocache(vaddr + shmem_page_offset,
-						user_data,
-						page_length);
+	ret = __copy_from_user_inatomic(vaddr + shmem_page_offset,
+					user_data, page_length);
 	if (needs_clflush_after)
 		drm_clflush_virt_range(vaddr + shmem_page_offset,
 				       page_length);
@@ -752,6 +891,8 @@ i915_gem_shmem_pwrite(struct drm_device *dev,
 		ret = i915_gem_object_wait_rendering(obj, false);
 		if (ret)
 			return ret;
+
+		i915_gem_object_retire(obj);
 	}
 	/* Same trick applies to invalidate partially written cachelines read
 	 * before writing. */
@@ -813,13 +954,10 @@ i915_gem_shmem_pwrite(struct drm_device *dev,
 
 		mutex_lock(&dev->struct_mutex);
 
-next_page:
-		set_page_dirty(page);
-		mark_page_accessed(page);
-
 		if (ret)
 			goto out;
 
+next_page:
 		remain -= page_length;
 		user_data += page_length;
 		offset += page_length;
@@ -868,7 +1006,7 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 		       args->size))
 		return -EFAULT;
 
-	if (likely(!i915_prefault_disable)) {
+	if (likely(!i915.prefault_disable)) {
 		ret = fault_in_multipages_readable(to_user_ptr(args->data_ptr),
 						   args->size);
 		if (ret)
@@ -909,8 +1047,8 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 	 * pread/pwrite currently are reading and writing from the CPU
 	 * perspective, requiring manual detiling by the client.
 	 */
-	if (obj->phys_obj) {
-		ret = i915_gem_phys_pwrite(dev, obj, args, file);
+	if (obj->phys_handle) {
+		ret = i915_gem_phys_pwrite(obj, args, file);
 		goto out;
 	}
 
@@ -958,7 +1096,7 @@ i915_gem_check_wedge(struct i915_gpu_error *error,
  * equal.
  */
 static int
-i915_gem_check_olr(struct intel_ring_buffer *ring, u32 seqno)
+i915_gem_check_olr(struct intel_engine_cs *ring, u32 seqno)
 {
 	int ret;
 
@@ -977,7 +1115,7 @@ static void fake_irq(unsigned long data)
 }
 
 static bool missed_irq(struct drm_i915_private *dev_priv,
-		       struct intel_ring_buffer *ring)
+		       struct intel_engine_cs *ring)
 {
 	return test_bit(ring->id, &dev_priv->gpu_error.missed_irq_rings);
 }
@@ -1008,26 +1146,29 @@ static bool can_wait_boost(struct drm_i915_file_private *file_priv)
  * Returns 0 if the seqno was found within the alloted time. Else returns the
  * errno with remaining time filled in timeout argument.
  */
-static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
+static int __wait_seqno(struct intel_engine_cs *ring, u32 seqno,
 			unsigned reset_counter,
 			bool interruptible,
 			struct timespec *timeout,
 			struct drm_i915_file_private *file_priv)
 {
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	const bool irq_test_in_progress =
+		ACCESS_ONCE(dev_priv->gpu_error.test_irq_rings) & intel_ring_flag(ring);
 	struct timespec before, now;
 	DEFINE_WAIT(wait);
-	long timeout_jiffies;
+	unsigned long timeout_expire;
 	int ret;
 
-	WARN(dev_priv->pc8.irqs_disabled, "IRQs disabled\n");
+	WARN(dev_priv->pm.irqs_disabled, "IRQs disabled\n");
 
 	if (i915_seqno_passed(ring->get_seqno(ring, true), seqno))
 		return 0;
 
-	timeout_jiffies = timeout ? timespec_to_jiffies_timeout(timeout) : 1;
+	timeout_expire = timeout ? jiffies + timespec_to_jiffies_timeout(timeout) : 0;
 
-	if (dev_priv->info->gen >= 6 && can_wait_boost(file_priv)) {
+	if (INTEL_INFO(dev)->gen >= 6 && can_wait_boost(file_priv)) {
 		gen6_rps_boost(dev_priv);
 		if (file_priv)
 			mod_delayed_work(dev_priv->wq,
@@ -1035,8 +1176,7 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 					 msecs_to_jiffies(100));
 	}
 
-	if (!(dev_priv->gpu_error.test_irq_rings & intel_ring_flag(ring)) &&
-	    WARN_ON(!ring->irq_get(ring)))
+	if (!irq_test_in_progress && WARN_ON(!ring->irq_get(ring)))
 		return -ENODEV;
 
 	/* Record current time in case interrupted by signal, or wedged */
@@ -1044,7 +1184,6 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 	getrawmonotonic(&before);
 	for (;;) {
 		struct timer_list timer;
-		unsigned long expire;
 
 		prepare_to_wait(&ring->irq_queue, &wait,
 				interruptible ? TASK_INTERRUPTIBLE : TASK_UNINTERRUPTIBLE);
@@ -1070,23 +1209,22 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 			break;
 		}
 
-		if (timeout_jiffies <= 0) {
+		if (timeout && time_after_eq(jiffies, timeout_expire)) {
 			ret = -ETIME;
 			break;
 		}
 
 		timer.function = NULL;
 		if (timeout || missed_irq(dev_priv, ring)) {
+			unsigned long expire;
+
 			setup_timer_on_stack(&timer, fake_irq, (unsigned long)current);
-			expire = jiffies + (missed_irq(dev_priv, ring) ? 1: timeout_jiffies);
+			expire = missed_irq(dev_priv, ring) ? jiffies + 1 : timeout_expire;
 			mod_timer(&timer, expire);
 		}
 
 		io_schedule();
 
-		if (timeout)
-			timeout_jiffies = expire - jiffies;
-
 		if (timer.function) {
 			del_singleshot_timer_sync(&timer);
 			destroy_timer_on_stack(&timer);
@@ -1095,7 +1233,8 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 	getrawmonotonic(&now);
 	trace_i915_gem_request_wait_end(ring, seqno);
 
-	ring->irq_put(ring);
+	if (!irq_test_in_progress)
+		ring->irq_put(ring);
 
 	finish_wait(&ring->irq_queue, &wait);
 
@@ -1114,7 +1253,7 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
  * request and object lists appropriately for that event.
  */
 int
-i915_wait_seqno(struct intel_ring_buffer *ring, uint32_t seqno)
+i915_wait_seqno(struct intel_engine_cs *ring, uint32_t seqno)
 {
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1139,9 +1278,10 @@ i915_wait_seqno(struct intel_ring_buffer *ring, uint32_t seqno)
 
 static int
 i915_gem_object_wait_rendering__tail(struct drm_i915_gem_object *obj,
-				     struct intel_ring_buffer *ring)
+				     struct intel_engine_cs *ring)
 {
-	i915_gem_retire_requests_ring(ring);
+	if (!obj->active)
+		return 0;
 
 	/* Manually manage the write flush as we may have not yet
 	 * retired the buffer.
@@ -1151,7 +1291,6 @@ i915_gem_object_wait_rendering__tail(struct drm_i915_gem_object *obj,
 	 * we know we have passed the last write.
 	 */
 	obj->last_write_seqno = 0;
-	obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
 
 	return 0;
 }
@@ -1164,7 +1303,7 @@ static __must_check int
 i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
 			       bool readonly)
 {
-	struct intel_ring_buffer *ring = obj->ring;
+	struct intel_engine_cs *ring = obj->ring;
 	u32 seqno;
 	int ret;
 
@@ -1184,12 +1323,12 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
  */
 static __must_check int
 i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
-					    struct drm_file *file,
+					    struct drm_i915_file_private *file_priv,
 					    bool readonly)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = obj->ring;
+	struct intel_engine_cs *ring = obj->ring;
 	unsigned reset_counter;
 	u32 seqno;
 	int ret;
@@ -1211,7 +1350,7 @@ i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
 
 	reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
 	mutex_unlock(&dev->struct_mutex);
-	ret = __wait_seqno(ring, seqno, reset_counter, true, NULL, file->driver_priv);
+	ret = __wait_seqno(ring, seqno, reset_counter, true, NULL, file_priv);
 	mutex_lock(&dev->struct_mutex);
 	if (ret)
 		return ret;
@@ -1260,7 +1399,9 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 	 * We will repeat the flush holding the lock in the normal manner
 	 * to catch cases where we are gazumped.
 	 */
-	ret = i915_gem_object_wait_rendering__nonblocking(obj, file, !write_domain);
+	ret = i915_gem_object_wait_rendering__nonblocking(obj,
+							  file->driver_priv,
+							  !write_domain);
 	if (ret)
 		goto unref;
 
@@ -1374,12 +1515,14 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 {
 	struct drm_i915_gem_object *obj = to_intel_bo(vma->vm_private_data);
 	struct drm_device *dev = obj->base.dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	pgoff_t page_offset;
 	unsigned long pfn;
 	int ret = 0;
 	bool write = !!(vmf->flags & FAULT_FLAG_WRITE);
 
+	intel_runtime_pm_get(dev_priv);
+
 	/* We don't use vmf->pgoff since that has the fake offset */
 	page_offset = ((unsigned long)vmf->virtual_address - vma->vm_start) >>
 		PAGE_SHIFT;
@@ -1390,14 +1533,23 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 
 	trace_i915_gem_object_fault(obj, page_offset, true, write);
 
+	/* Try to flush the object off the GPU first without holding the lock.
+	 * Upon reacquiring the lock, we will perform our sanity checks and then
+	 * repeat the flush holding the lock in the normal manner to catch cases
+	 * where we are gazumped.
+	 */
+	ret = i915_gem_object_wait_rendering__nonblocking(obj, NULL, !write);
+	if (ret)
+		goto unlock;
+
 	/* Access to snoopable pages through the GTT is incoherent. */
 	if (obj->cache_level != I915_CACHE_NONE && !HAS_LLC(dev)) {
-		ret = -EINVAL;
+		ret = -EFAULT;
 		goto unlock;
 	}
 
 	/* Now bind it into the GTT if needed */
-	ret = i915_gem_obj_ggtt_pin(obj,  0, true, false);
+	ret = i915_gem_obj_ggtt_pin(obj, 0, PIN_MAPPABLE);
 	if (ret)
 		goto unlock;
 
@@ -1418,7 +1570,7 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	/* Finally, remap it using the new GTT offset */
 	ret = vm_insert_pfn(vma, (unsigned long)vmf->virtual_address, pfn);
 unpin:
-	i915_gem_object_unpin(obj);
+	i915_gem_object_ggtt_unpin(obj);
 unlock:
 	mutex_unlock(&dev->struct_mutex);
 out:
@@ -1427,8 +1579,10 @@ out:
 		/* If this -EIO is due to a gpu hang, give the reset code a
 		 * chance to clean up the mess. Otherwise return the proper
 		 * SIGBUS. */
-		if (i915_terminally_wedged(&dev_priv->gpu_error))
-			return VM_FAULT_SIGBUS;
+		if (i915_terminally_wedged(&dev_priv->gpu_error)) {
+			ret = VM_FAULT_SIGBUS;
+			break;
+		}
 	case -EAGAIN:
 		/*
 		 * EAGAIN means the gpu is hung and we'll wait for the error
@@ -1443,15 +1597,23 @@ out:
 		 * EBUSY is ok: this just means that another thread
 		 * already did the job.
 		 */
-		return VM_FAULT_NOPAGE;
+		ret = VM_FAULT_NOPAGE;
+		break;
 	case -ENOMEM:
-		return VM_FAULT_OOM;
+		ret = VM_FAULT_OOM;
+		break;
 	case -ENOSPC:
-		return VM_FAULT_SIGBUS;
+	case -EFAULT:
+		ret = VM_FAULT_SIGBUS;
+		break;
 	default:
 		WARN_ONCE(ret, "unhandled error in i915_gem_fault: %i\n", ret);
-		return VM_FAULT_SIGBUS;
+		ret = VM_FAULT_SIGBUS;
+		break;
 	}
+
+	intel_runtime_pm_put(dev_priv);
+	return ret;
 }
 
 /**
@@ -1474,10 +1636,20 @@ i915_gem_release_mmap(struct drm_i915_gem_object *obj)
 	if (!obj->fault_mappable)
 		return;
 
-	drm_vma_node_unmap(&obj->base.vma_node, obj->base.dev->dev_mapping);
+	drm_vma_node_unmap(&obj->base.vma_node,
+			   obj->base.dev->anon_inode->i_mapping);
 	obj->fault_mappable = false;
 }
 
+void
+i915_gem_release_all_mmaps(struct drm_i915_private *dev_priv)
+{
+	struct drm_i915_gem_object *obj;
+
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list)
+		i915_gem_release_mmap(obj);
+}
+
 uint32_t
 i915_gem_get_gtt_size(struct drm_device *dev, uint32_t size, int tiling_mode)
 {
@@ -1590,8 +1762,8 @@ i915_gem_mmap_gtt(struct drm_file *file,
 	}
 
 	if (obj->madv != I915_MADV_WILLNEED) {
-		DRM_ERROR("Attempting to mmap a purgeable buffer\n");
-		ret = -EINVAL;
+		DRM_DEBUG("Attempting to mmap a purgeable buffer\n");
+		ret = -EFAULT;
 		goto out;
 	}
 
@@ -1632,12 +1804,16 @@ i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data,
 	return i915_gem_mmap_gtt(file, dev, args->handle, &args->offset);
 }
 
+static inline int
+i915_gem_object_is_purgeable(struct drm_i915_gem_object *obj)
+{
+	return obj->madv == I915_MADV_DONTNEED;
+}
+
 /* Immediately discard the backing storage */
 static void
 i915_gem_object_truncate(struct drm_i915_gem_object *obj)
 {
-	struct inode *inode;
-
 	i915_gem_object_free_mmap_offset(obj);
 
 	if (obj->base.filp == NULL)
@@ -1648,16 +1824,28 @@ i915_gem_object_truncate(struct drm_i915_gem_object *obj)
 	 * To do this we must instruct the shmfs to drop all of its
 	 * backing pages, *now*.
 	 */
-	inode = file_inode(obj->base.filp);
-	shmem_truncate_range(inode, 0, (loff_t)-1);
-
+	shmem_truncate_range(file_inode(obj->base.filp), 0, (loff_t)-1);
 	obj->madv = __I915_MADV_PURGED;
 }
 
-static inline int
-i915_gem_object_is_purgeable(struct drm_i915_gem_object *obj)
+/* Try to discard unwanted pages */
+static void
+i915_gem_object_invalidate(struct drm_i915_gem_object *obj)
 {
-	return obj->madv == I915_MADV_DONTNEED;
+	struct address_space *mapping;
+
+	switch (obj->madv) {
+	case I915_MADV_DONTNEED:
+		i915_gem_object_truncate(obj);
+	case __I915_MADV_PURGED:
+		return;
+	}
+
+	if (obj->base.filp == NULL)
+		return;
+
+	mapping = file_inode(obj->base.filp)->i_mapping,
+	invalidate_mapping_pages(mapping, 0, (loff_t)-1);
 }
 
 static void
@@ -1722,8 +1910,7 @@ i915_gem_object_put_pages(struct drm_i915_gem_object *obj)
 	ops->put_pages(obj);
 	obj->pages = NULL;
 
-	if (i915_gem_object_is_purgeable(obj))
-		i915_gem_object_truncate(obj);
+	i915_gem_object_invalidate(obj);
 
 	return 0;
 }
@@ -1732,58 +1919,58 @@ static unsigned long
 __i915_gem_shrink(struct drm_i915_private *dev_priv, long target,
 		  bool purgeable_only)
 {
-	struct list_head still_bound_list;
-	struct drm_i915_gem_object *obj, *next;
+	struct list_head still_in_list;
+	struct drm_i915_gem_object *obj;
 	unsigned long count = 0;
 
-	list_for_each_entry_safe(obj, next,
-				 &dev_priv->mm.unbound_list,
-				 global_list) {
-		if ((i915_gem_object_is_purgeable(obj) || !purgeable_only) &&
-		    i915_gem_object_put_pages(obj) == 0) {
-			count += obj->base.size >> PAGE_SHIFT;
-			if (count >= target)
-				return count;
-		}
-	}
-
 	/*
-	 * As we may completely rewrite the bound list whilst unbinding
+	 * As we may completely rewrite the (un)bound list whilst unbinding
 	 * (due to retiring requests) we have to strictly process only
 	 * one element of the list at the time, and recheck the list
 	 * on every iteration.
+	 *
+	 * In particular, we must hold a reference whilst removing the
+	 * object as we may end up waiting for and/or retiring the objects.
+	 * This might release the final reference (held by the active list)
+	 * and result in the object being freed from under us. This is
+	 * similar to the precautions the eviction code must take whilst
+	 * removing objects.
+	 *
+	 * Also note that although these lists do not hold a reference to
+	 * the object we can safely grab one here: The final object
+	 * unreferencing and the bound_list are both protected by the
+	 * dev->struct_mutex and so we won't ever be able to observe an
+	 * object on the bound_list with a reference count equals 0.
 	 */
-	INIT_LIST_HEAD(&still_bound_list);
+	INIT_LIST_HEAD(&still_in_list);
+	while (count < target && !list_empty(&dev_priv->mm.unbound_list)) {
+		obj = list_first_entry(&dev_priv->mm.unbound_list,
+				       typeof(*obj), global_list);
+		list_move_tail(&obj->global_list, &still_in_list);
+
+		if (!i915_gem_object_is_purgeable(obj) && purgeable_only)
+			continue;
+
+		drm_gem_object_reference(&obj->base);
+
+		if (i915_gem_object_put_pages(obj) == 0)
+			count += obj->base.size >> PAGE_SHIFT;
+
+		drm_gem_object_unreference(&obj->base);
+	}
+	list_splice(&still_in_list, &dev_priv->mm.unbound_list);
+
+	INIT_LIST_HEAD(&still_in_list);
 	while (count < target && !list_empty(&dev_priv->mm.bound_list)) {
 		struct i915_vma *vma, *v;
 
 		obj = list_first_entry(&dev_priv->mm.bound_list,
 				       typeof(*obj), global_list);
-		list_move_tail(&obj->global_list, &still_bound_list);
+		list_move_tail(&obj->global_list, &still_in_list);
 
 		if (!i915_gem_object_is_purgeable(obj) && purgeable_only)
 			continue;
 
-		/*
-		 * Hold a reference whilst we unbind this object, as we may
-		 * end up waiting for and retiring requests. This might
-		 * release the final reference (held by the active list)
-		 * and result in the object being freed from under us.
-		 * in this object being freed.
-		 *
-		 * Note 1: Shrinking the bound list is special since only active
-		 * (and hence bound objects) can contain such limbo objects, so
-		 * we don't need special tricks for shrinking the unbound list.
-		 * The only other place where we have to be careful with active
-		 * objects suddenly disappearing due to retiring requests is the
-		 * eviction code.
-		 *
-		 * Note 2: Even though the bound list doesn't hold a reference
-		 * to the object we can safely grab one here: The final object
-		 * unreferencing and the bound_list are both protected by the
-		 * dev->struct_mutex and so we won't ever be able to observe an
-		 * object on the bound_list with a reference count equals 0.
-		 */
 		drm_gem_object_reference(&obj->base);
 
 		list_for_each_entry_safe(vma, v, &obj->vma_list, vma_link)
@@ -1795,7 +1982,7 @@ __i915_gem_shrink(struct drm_i915_private *dev_priv, long target,
 
 		drm_gem_object_unreference(&obj->base);
 	}
-	list_splice(&still_bound_list, &dev_priv->mm.bound_list);
+	list_splice(&still_in_list, &dev_priv->mm.bound_list);
 
 	return count;
 }
@@ -1809,17 +1996,8 @@ i915_gem_purge(struct drm_i915_private *dev_priv, long target)
 static unsigned long
 i915_gem_shrink_all(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_gem_object *obj, *next;
-	long freed = 0;
-
 	i915_gem_evict_everything(dev_priv->dev);
-
-	list_for_each_entry_safe(obj, next, &dev_priv->mm.unbound_list,
-				 global_list) {
-		if (i915_gem_object_put_pages(obj) == 0)
-			freed += obj->base.size >> PAGE_SHIFT;
-	}
-	return freed;
+	return __i915_gem_shrink(dev_priv, LONG_MAX, false);
 }
 
 static int
@@ -1923,7 +2101,19 @@ err_pages:
 		page_cache_release(sg_page_iter_page(&sg_iter));
 	sg_free_table(st);
 	kfree(st);
-	return PTR_ERR(page);
+
+	/* shmemfs first checks if there is enough memory to allocate the page
+	 * and reports ENOSPC should there be insufficient, along with the usual
+	 * ENOMEM for a genuine allocation failure.
+	 *
+	 * We use ENOSPC in our driver to mean that we have run out of aperture
+	 * space and so want to translate the error from shmemfs back to our
+	 * usual understanding of ENOMEM.
+	 */
+	if (PTR_ERR(page) == -ENOSPC)
+		return -ENOMEM;
+	else
+		return PTR_ERR(page);
 }
 
 /* Ensure that the associated pages are gathered from the backing storage
@@ -1944,8 +2134,8 @@ i915_gem_object_get_pages(struct drm_i915_gem_object *obj)
 		return 0;
 
 	if (obj->madv != I915_MADV_WILLNEED) {
-		DRM_ERROR("Attempting to obtain a purgeable object\n");
-		return -EINVAL;
+		DRM_DEBUG("Attempting to obtain a purgeable object\n");
+		return -EFAULT;
 	}
 
 	BUG_ON(obj->pages_pin_count);
@@ -1960,7 +2150,7 @@ i915_gem_object_get_pages(struct drm_i915_gem_object *obj)
 
 static void
 i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
-			       struct intel_ring_buffer *ring)
+			       struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1998,7 +2188,7 @@ i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
 }
 
 void i915_vma_move_to_active(struct i915_vma *vma,
-			     struct intel_ring_buffer *ring)
+			     struct intel_engine_cs *ring)
 {
 	list_move_tail(&vma->mm_list, &vma->vm->active_list);
 	return i915_gem_object_move_to_active(vma->obj, ring);
@@ -2008,13 +2198,17 @@ static void
 i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
 {
 	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
-	struct i915_address_space *ggtt_vm = &dev_priv->gtt.base;
-	struct i915_vma *vma = i915_gem_obj_to_vma(obj, ggtt_vm);
+	struct i915_address_space *vm;
+	struct i915_vma *vma;
 
 	BUG_ON(obj->base.write_domain & ~I915_GEM_GPU_DOMAINS);
 	BUG_ON(!obj->active);
 
-	list_move_tail(&vma->mm_list, &ggtt_vm->inactive_list);
+	list_for_each_entry(vm, &dev_priv->vm_list, global_link) {
+		vma = i915_gem_obj_to_vma(obj, vm);
+		if (vma && !list_empty(&vma->mm_list))
+			list_move_tail(&vma->mm_list, &vm->inactive_list);
+	}
 
 	list_del_init(&obj->ring_list);
 	obj->ring = NULL;
@@ -2032,11 +2226,24 @@ i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
 	WARN_ON(i915_verify_lists(dev));
 }
 
+static void
+i915_gem_object_retire(struct drm_i915_gem_object *obj)
+{
+	struct intel_engine_cs *ring = obj->ring;
+
+	if (ring == NULL)
+		return;
+
+	if (i915_seqno_passed(ring->get_seqno(ring, true),
+			      obj->last_read_seqno))
+		i915_gem_object_move_to_inactive(obj);
+}
+
 static int
 i915_gem_init_seqno(struct drm_device *dev, u32 seqno)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	int ret, i, j;
 
 	/* Carefully retire all requests without writing to the rings */
@@ -2051,8 +2258,8 @@ i915_gem_init_seqno(struct drm_device *dev, u32 seqno)
 	for_each_ring(ring, dev_priv, i) {
 		intel_ring_init_seqno(ring, seqno);
 
-		for (j = 0; j < ARRAY_SIZE(ring->sync_seqno); j++)
-			ring->sync_seqno[j] = 0;
+		for (j = 0; j < ARRAY_SIZE(ring->semaphore.sync_seqno); j++)
+			ring->semaphore.sync_seqno[j] = 0;
 	}
 
 	return 0;
@@ -2102,15 +2309,14 @@ i915_gem_get_seqno(struct drm_device *dev, u32 *seqno)
 	return 0;
 }
 
-int __i915_add_request(struct intel_ring_buffer *ring,
+int __i915_add_request(struct intel_engine_cs *ring,
 		       struct drm_file *file,
 		       struct drm_i915_gem_object *obj,
 		       u32 *out_seqno)
 {
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	struct drm_i915_gem_request *request;
 	u32 request_ring_position, request_start;
-	int was_empty;
 	int ret;
 
 	request_start = intel_ring_get_tail(ring);
@@ -2161,7 +2367,6 @@ int __i915_add_request(struct intel_ring_buffer *ring,
 		i915_gem_context_reference(request->ctx);
 
 	request->emitted_jiffies = jiffies;
-	was_empty = list_empty(&ring->request_list);
 	list_add_tail(&request->list, &ring->request_list);
 	request->file_priv = NULL;
 
@@ -2182,13 +2387,11 @@ int __i915_add_request(struct intel_ring_buffer *ring,
 	if (!dev_priv->ums.mm_suspended) {
 		i915_queue_hangcheck(ring->dev);
 
-		if (was_empty) {
-			cancel_delayed_work_sync(&dev_priv->mm.idle_work);
-			queue_delayed_work(dev_priv->wq,
-					   &dev_priv->mm.retire_work,
-					   round_jiffies_up_relative(HZ));
-			intel_mark_busy(dev_priv->dev);
-		}
+		cancel_delayed_work_sync(&dev_priv->mm.idle_work);
+		queue_delayed_work(dev_priv->wq,
+				   &dev_priv->mm.retire_work,
+				   round_jiffies_up_relative(HZ));
+		intel_mark_busy(dev_priv->dev);
 	}
 
 	if (out_seqno)
@@ -2210,125 +2413,47 @@ i915_gem_request_remove_from_client(struct drm_i915_gem_request *request)
 	spin_unlock(&file_priv->mm.lock);
 }
 
-static bool i915_head_inside_object(u32 acthd, struct drm_i915_gem_object *obj,
-				    struct i915_address_space *vm)
+static bool i915_context_is_banned(struct drm_i915_private *dev_priv,
+				   const struct intel_context *ctx)
 {
-	if (acthd >= i915_gem_obj_offset(obj, vm) &&
-	    acthd < i915_gem_obj_offset(obj, vm) + obj->base.size)
-		return true;
+	unsigned long elapsed;
 
-	return false;
-}
+	elapsed = get_seconds() - ctx->hang_stats.guilty_ts;
 
-static bool i915_head_inside_request(const u32 acthd_unmasked,
-				     const u32 request_start,
-				     const u32 request_end)
-{
-	const u32 acthd = acthd_unmasked & HEAD_ADDR;
+	if (ctx->hang_stats.banned)
+		return true;
 
-	if (request_start < request_end) {
-		if (acthd >= request_start && acthd < request_end)
+	if (elapsed <= DRM_I915_CTX_BAN_PERIOD) {
+		if (!i915_gem_context_is_default(ctx)) {
+			DRM_DEBUG("context hanging too fast, banning!\n");
 			return true;
-	} else if (request_start > request_end) {
-		if (acthd >= request_start || acthd < request_end)
-			return true;
-	}
-
-	return false;
-}
-
-static struct i915_address_space *
-request_to_vm(struct drm_i915_gem_request *request)
-{
-	struct drm_i915_private *dev_priv = request->ring->dev->dev_private;
-	struct i915_address_space *vm;
-
-	vm = &dev_priv->gtt.base;
-
-	return vm;
-}
-
-static bool i915_request_guilty(struct drm_i915_gem_request *request,
-				const u32 acthd, bool *inside)
-{
-	/* There is a possibility that unmasked head address
-	 * pointing inside the ring, matches the batch_obj address range.
-	 * However this is extremely unlikely.
-	 */
-	if (request->batch_obj) {
-		if (i915_head_inside_object(acthd, request->batch_obj,
-					    request_to_vm(request))) {
-			*inside = true;
+		} else if (i915_stop_ring_allow_ban(dev_priv)) {
+			if (i915_stop_ring_allow_warn(dev_priv))
+				DRM_ERROR("gpu hanging too fast, banning!\n");
 			return true;
 		}
 	}
 
-	if (i915_head_inside_request(acthd, request->head, request->tail)) {
-		*inside = false;
-		return true;
-	}
-
 	return false;
 }
 
-static bool i915_context_is_banned(const struct i915_ctx_hang_stats *hs)
+static void i915_set_reset_status(struct drm_i915_private *dev_priv,
+				  struct intel_context *ctx,
+				  const bool guilty)
 {
-	const unsigned long elapsed = get_seconds() - hs->guilty_ts;
+	struct i915_ctx_hang_stats *hs;
 
-	if (hs->banned)
-		return true;
-
-	if (elapsed <= DRM_I915_CTX_BAN_PERIOD) {
-		DRM_ERROR("context hanging too fast, declaring banned!\n");
-		return true;
-	}
-
-	return false;
-}
-
-static void i915_set_reset_status(struct intel_ring_buffer *ring,
-				  struct drm_i915_gem_request *request,
-				  u32 acthd)
-{
-	struct i915_ctx_hang_stats *hs = NULL;
-	bool inside, guilty;
-	unsigned long offset = 0;
-
-	/* Innocent until proven guilty */
-	guilty = false;
-
-	if (request->batch_obj)
-		offset = i915_gem_obj_offset(request->batch_obj,
-					     request_to_vm(request));
-
-	if (ring->hangcheck.action != HANGCHECK_WAIT &&
-	    i915_request_guilty(request, acthd, &inside)) {
-		DRM_ERROR("%s hung %s bo (0x%lx ctx %d) at 0x%x\n",
-			  ring->name,
-			  inside ? "inside" : "flushing",
-			  offset,
-			  request->ctx ? request->ctx->id : 0,
-			  acthd);
+	if (WARN_ON(!ctx))
+		return;
 
-		guilty = true;
-	}
+	hs = &ctx->hang_stats;
 
-	/* If contexts are disabled or this is the default context, use
-	 * file_priv->reset_state
-	 */
-	if (request->ctx && request->ctx->id != DEFAULT_CONTEXT_ID)
-		hs = &request->ctx->hang_stats;
-	else if (request->file_priv)
-		hs = &request->file_priv->hang_stats;
-
-	if (hs) {
-		if (guilty) {
-			hs->banned = i915_context_is_banned(hs);
-			hs->batch_active++;
-			hs->guilty_ts = get_seconds();
-		} else {
-			hs->batch_pending++;
-		}
+	if (guilty) {
+		hs->banned = i915_context_is_banned(dev_priv, ctx);
+		hs->batch_active++;
+		hs->guilty_ts = get_seconds();
+	} else {
+		hs->batch_pending++;
 	}
 }
 
@@ -2343,34 +2468,46 @@ static void i915_gem_free_request(struct drm_i915_gem_request *request)
 	kfree(request);
 }
 
-static void i915_gem_reset_ring_status(struct drm_i915_private *dev_priv,
-				       struct intel_ring_buffer *ring)
+struct drm_i915_gem_request *
+i915_gem_find_active_request(struct intel_engine_cs *ring)
 {
-	u32 completed_seqno = ring->get_seqno(ring, false);
-	u32 acthd = intel_ring_get_active_head(ring);
 	struct drm_i915_gem_request *request;
+	u32 completed_seqno;
+
+	completed_seqno = ring->get_seqno(ring, false);
 
 	list_for_each_entry(request, &ring->request_list, list) {
 		if (i915_seqno_passed(completed_seqno, request->seqno))
 			continue;
 
-		i915_set_reset_status(ring, request, acthd);
+		return request;
 	}
+
+	return NULL;
 }
 
-static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
-					struct intel_ring_buffer *ring)
+static void i915_gem_reset_ring_status(struct drm_i915_private *dev_priv,
+				       struct intel_engine_cs *ring)
 {
-	while (!list_empty(&ring->request_list)) {
-		struct drm_i915_gem_request *request;
+	struct drm_i915_gem_request *request;
+	bool ring_hung;
 
-		request = list_first_entry(&ring->request_list,
-					   struct drm_i915_gem_request,
-					   list);
+	request = i915_gem_find_active_request(ring);
 
-		i915_gem_free_request(request);
-	}
+	if (request == NULL)
+		return;
 
+	ring_hung = ring->hangcheck.score >= HANGCHECK_SCORE_RING_HUNG;
+
+	i915_set_reset_status(dev_priv, request->ctx, ring_hung);
+
+	list_for_each_entry_continue(request, &ring->request_list, list)
+		i915_set_reset_status(dev_priv, request->ctx, false);
+}
+
+static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
+					struct intel_engine_cs *ring)
+{
 	while (!list_empty(&ring->active_list)) {
 		struct drm_i915_gem_object *obj;
 
@@ -2380,6 +2517,28 @@ static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
 
 		i915_gem_object_move_to_inactive(obj);
 	}
+
+	/*
+	 * We must free the requests after all the corresponding objects have
+	 * been moved off active lists. Which is the same order as the normal
+	 * retire_requests function does. This is important if object hold
+	 * implicit references on things like e.g. ppgtt address spaces through
+	 * the request.
+	 */
+	while (!list_empty(&ring->request_list)) {
+		struct drm_i915_gem_request *request;
+
+		request = list_first_entry(&ring->request_list,
+					   struct drm_i915_gem_request,
+					   list);
+
+		i915_gem_free_request(request);
+	}
+
+	/* These may not have been flush before the reset, do so now */
+	kfree(ring->preallocated_lazy_request);
+	ring->preallocated_lazy_request = NULL;
+	ring->outstanding_lazy_seqno = 0;
 }
 
 void i915_gem_restore_fences(struct drm_device *dev)
@@ -2406,7 +2565,7 @@ void i915_gem_restore_fences(struct drm_device *dev)
 void i915_gem_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	int i;
 
 	/*
@@ -2420,7 +2579,7 @@ void i915_gem_reset(struct drm_device *dev)
 	for_each_ring(ring, dev_priv, i)
 		i915_gem_reset_ring_cleanup(dev_priv, ring);
 
-	i915_gem_cleanup_ringbuffer(dev);
+	i915_gem_context_reset(dev);
 
 	i915_gem_restore_fences(dev);
 }
@@ -2429,7 +2588,7 @@ void i915_gem_reset(struct drm_device *dev)
  * This function clears the request list as sequence numbers are passed.
  */
 void
-i915_gem_retire_requests_ring(struct intel_ring_buffer *ring)
+i915_gem_retire_requests_ring(struct intel_engine_cs *ring)
 {
 	uint32_t seqno;
 
@@ -2440,6 +2599,24 @@ i915_gem_retire_requests_ring(struct intel_ring_buffer *ring)
 
 	seqno = ring->get_seqno(ring, true);
 
+	/* Move any buffers on the active list that are no longer referenced
+	 * by the ringbuffer to the flushing/inactive lists as appropriate,
+	 * before we free the context associated with the requests.
+	 */
+	while (!list_empty(&ring->active_list)) {
+		struct drm_i915_gem_object *obj;
+
+		obj = list_first_entry(&ring->active_list,
+				      struct drm_i915_gem_object,
+				      ring_list);
+
+		if (!i915_seqno_passed(seqno, obj->last_read_seqno))
+			break;
+
+		i915_gem_object_move_to_inactive(obj);
+	}
+
+
 	while (!list_empty(&ring->request_list)) {
 		struct drm_i915_gem_request *request;
 
@@ -2456,27 +2633,11 @@ i915_gem_retire_requests_ring(struct intel_ring_buffer *ring)
 		 * of tail of the request to update the last known position
 		 * of the GPU head.
 		 */
-		ring->last_retired_head = request->tail;
+		ring->buffer->last_retired_head = request->tail;
 
 		i915_gem_free_request(request);
 	}
 
-	/* Move any buffers on the active list that are no longer referenced
-	 * by the ringbuffer to the flushing/inactive lists as appropriate.
-	 */
-	while (!list_empty(&ring->active_list)) {
-		struct drm_i915_gem_object *obj;
-
-		obj = list_first_entry(&ring->active_list,
-				      struct drm_i915_gem_object,
-				      ring_list);
-
-		if (!i915_seqno_passed(seqno, obj->last_read_seqno))
-			break;
-
-		i915_gem_object_move_to_inactive(obj);
-	}
-
 	if (unlikely(ring->trace_irq_seqno &&
 		     i915_seqno_passed(seqno, ring->trace_irq_seqno))) {
 		ring->irq_put(ring);
@@ -2489,8 +2650,8 @@ i915_gem_retire_requests_ring(struct intel_ring_buffer *ring)
 bool
 i915_gem_retire_requests(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
 	bool idle = true;
 	int i;
 
@@ -2581,10 +2742,10 @@ i915_gem_object_flush_active(struct drm_i915_gem_object *obj)
 int
 i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_wait *args = data;
 	struct drm_i915_gem_object *obj;
-	struct intel_ring_buffer *ring = NULL;
+	struct intel_engine_cs *ring = NULL;
 	struct timespec timeout_stack, *timeout = NULL;
 	unsigned reset_counter;
 	u32 seqno = 0;
@@ -2655,9 +2816,9 @@ out:
  */
 int
 i915_gem_object_sync(struct drm_i915_gem_object *obj,
-		     struct intel_ring_buffer *to)
+		     struct intel_engine_cs *to)
 {
-	struct intel_ring_buffer *from = obj->ring;
+	struct intel_engine_cs *from = obj->ring;
 	u32 seqno;
 	int ret, idx;
 
@@ -2670,7 +2831,7 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
 	idx = intel_ring_sync_index(from, to);
 
 	seqno = obj->last_read_seqno;
-	if (seqno <= from->sync_seqno[idx])
+	if (seqno <= from->semaphore.sync_seqno[idx])
 		return 0;
 
 	ret = i915_gem_check_olr(obj->ring, seqno);
@@ -2678,13 +2839,13 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
 		return ret;
 
 	trace_i915_gem_ring_sync_to(from, to, seqno);
-	ret = to->sync_to(to, from, seqno);
+	ret = to->semaphore.sync_to(to, from, seqno);
 	if (!ret)
 		/* We use last_read_seqno because sync_to()
 		 * might have just caused seqno wrap under
 		 * the radar.
 		 */
-		from->sync_seqno[idx] = obj->last_read_seqno;
+		from->semaphore.sync_seqno[idx] = obj->last_read_seqno;
 
 	return ret;
 }
@@ -2716,22 +2877,18 @@ static void i915_gem_object_finish_gtt(struct drm_i915_gem_object *obj)
 int i915_vma_unbind(struct i915_vma *vma)
 {
 	struct drm_i915_gem_object *obj = vma->obj;
-	drm_i915_private_t *dev_priv = obj->base.dev->dev_private;
+	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
 	int ret;
 
-	/* For now we only ever use 1 vma per object */
-	WARN_ON(!list_is_singular(&obj->vma_list));
-
 	if (list_empty(&vma->vma_link))
 		return 0;
 
 	if (!drm_mm_node_allocated(&vma->node)) {
 		i915_gem_vma_destroy(vma);
-
 		return 0;
 	}
 
-	if (obj->pin_count)
+	if (vma->pin_count)
 		return -EBUSY;
 
 	BUG_ON(obj->pages == NULL);
@@ -2744,31 +2901,27 @@ int i915_vma_unbind(struct i915_vma *vma)
 	 * cause memory corruption through use-after-free.
 	 */
 
-	i915_gem_object_finish_gtt(obj);
+	if (i915_is_ggtt(vma->vm)) {
+		i915_gem_object_finish_gtt(obj);
 
-	/* release the fence reg _after_ flushing */
-	ret = i915_gem_object_put_fence(obj);
-	if (ret)
-		return ret;
+		/* release the fence reg _after_ flushing */
+		ret = i915_gem_object_put_fence(obj);
+		if (ret)
+			return ret;
+	}
 
 	trace_i915_vma_unbind(vma);
 
-	if (obj->has_global_gtt_mapping)
-		i915_gem_gtt_unbind_object(obj);
-	if (obj->has_aliasing_ppgtt_mapping) {
-		i915_ppgtt_unbind_object(dev_priv->mm.aliasing_ppgtt, obj);
-		obj->has_aliasing_ppgtt_mapping = 0;
-	}
+	vma->unbind_vma(vma);
+
 	i915_gem_gtt_finish_object(obj);
-	i915_gem_object_unpin_pages(obj);
 
-	list_del(&vma->mm_list);
+	list_del_init(&vma->mm_list);
 	/* Avoid an unnecessary call to unbind on rebind. */
 	if (i915_is_ggtt(vma->vm))
 		obj->map_and_fenceable = true;
 
 	drm_mm_remove_node(&vma->node);
-
 	i915_gem_vma_destroy(vma);
 
 	/* Since the unbound list is global, only move to that list if
@@ -2776,38 +2929,24 @@ int i915_vma_unbind(struct i915_vma *vma)
 	if (list_empty(&obj->vma_list))
 		list_move_tail(&obj->global_list, &dev_priv->mm.unbound_list);
 
-	return 0;
-}
-
-/**
- * Unbinds an object from the global GTT aperture.
- */
-int
-i915_gem_object_ggtt_unbind(struct drm_i915_gem_object *obj)
-{
-	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
-	struct i915_address_space *ggtt = &dev_priv->gtt.base;
-
-	if (!i915_gem_obj_ggtt_bound(obj))
-		return 0;
-
-	if (obj->pin_count)
-		return -EBUSY;
-
-	BUG_ON(obj->pages == NULL);
+	/* And finally now the object is completely decoupled from this vma,
+	 * we can drop its hold on the backing storage and allow it to be
+	 * reaped by the shrinker.
+	 */
+	i915_gem_object_unpin_pages(obj);
 
-	return i915_vma_unbind(i915_gem_obj_to_vma(obj, ggtt));
+	return 0;
 }
 
 int i915_gpu_idle(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
 	int ret, i;
 
 	/* Flush everything onto the inactive list. */
 	for_each_ring(ring, dev_priv, i) {
-		ret = i915_switch_context(ring, NULL, DEFAULT_CONTEXT_ID);
+		ret = i915_switch_context(ring, ring->default_context);
 		if (ret)
 			return ret;
 
@@ -2822,7 +2961,7 @@ int i915_gpu_idle(struct drm_device *dev)
 static void i965_write_fence_reg(struct drm_device *dev, int reg,
 				 struct drm_i915_gem_object *obj)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int fence_reg;
 	int fence_pitch_shift;
 
@@ -2874,7 +3013,7 @@ static void i965_write_fence_reg(struct drm_device *dev, int reg,
 static void i915_write_fence_reg(struct drm_device *dev, int reg,
 				 struct drm_i915_gem_object *obj)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 val;
 
 	if (obj) {
@@ -2918,7 +3057,7 @@ static void i915_write_fence_reg(struct drm_device *dev, int reg,
 static void i830_write_fence_reg(struct drm_device *dev, int reg,
 				struct drm_i915_gem_object *obj)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t val;
 
 	if (obj) {
@@ -3043,6 +3182,9 @@ i915_gem_object_put_fence(struct drm_i915_gem_object *obj)
 
 	fence = &dev_priv->fence_regs[obj->fence_reg];
 
+	if (WARN_ON(fence->pin_count))
+		return -EBUSY;
+
 	i915_gem_object_fence_lost(obj);
 	i915_gem_object_update_fence(obj, fence, false);
 
@@ -3068,7 +3210,7 @@ i915_find_fence_reg(struct drm_device *dev)
 	}
 
 	if (avail == NULL)
-		return NULL;
+		goto deadlock;
 
 	/* None available, try to steal one or wait for a user to finish */
 	list_for_each_entry(reg, &dev_priv->mm.fence_list, lru_list) {
@@ -3078,7 +3220,12 @@ i915_find_fence_reg(struct drm_device *dev)
 		return reg;
 	}
 
-	return NULL;
+deadlock:
+	/* Wait for completion of pending flips which consume fences */
+	if (intel_has_pending_fb_unpin(dev))
+		return ERR_PTR(-EAGAIN);
+
+	return ERR_PTR(-EDEADLK);
 }
 
 /**
@@ -3123,8 +3270,8 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj)
 		}
 	} else if (enable) {
 		reg = i915_find_fence_reg(dev);
-		if (reg == NULL)
-			return -EDEADLK;
+		if (IS_ERR(reg))
+			return PTR_ERR(reg);
 
 		if (reg->obj) {
 			struct drm_i915_gem_object *old = reg->obj;
@@ -3216,18 +3363,19 @@ static void i915_gem_verify_gtt(struct drm_device *dev)
 /**
  * Finds free space in the GTT aperture and binds the object there.
  */
-static int
+static struct i915_vma *
 i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj,
 			   struct i915_address_space *vm,
 			   unsigned alignment,
-			   bool map_and_fenceable,
-			   bool nonblocking)
+			   uint64_t flags)
 {
 	struct drm_device *dev = obj->base.dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 size, fence_size, fence_alignment, unfenced_alignment;
-	size_t gtt_max =
-		map_and_fenceable ? dev_priv->gtt.mappable_end : vm->total;
+	unsigned long start =
+		flags & PIN_OFFSET_BIAS ? flags & PIN_OFFSET_MASK : 0;
+	unsigned long end =
+		flags & PIN_MAPPABLE ? dev_priv->gtt.mappable_end : vm->total;
 	struct i915_vma *vma;
 	int ret;
 
@@ -3239,57 +3387,52 @@ i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj,
 						     obj->tiling_mode, true);
 	unfenced_alignment =
 		i915_gem_get_gtt_alignment(dev,
-						    obj->base.size,
-						    obj->tiling_mode, false);
+					   obj->base.size,
+					   obj->tiling_mode, false);
 
 	if (alignment == 0)
-		alignment = map_and_fenceable ? fence_alignment :
+		alignment = flags & PIN_MAPPABLE ? fence_alignment :
 						unfenced_alignment;
-	if (map_and_fenceable && alignment & (fence_alignment - 1)) {
-		DRM_ERROR("Invalid object alignment requested %u\n", alignment);
-		return -EINVAL;
+	if (flags & PIN_MAPPABLE && alignment & (fence_alignment - 1)) {
+		DRM_DEBUG("Invalid object alignment requested %u\n", alignment);
+		return ERR_PTR(-EINVAL);
 	}
 
-	size = map_and_fenceable ? fence_size : obj->base.size;
+	size = flags & PIN_MAPPABLE ? fence_size : obj->base.size;
 
 	/* If the object is bigger than the entire aperture, reject it early
 	 * before evicting everything in a vain attempt to find space.
 	 */
-	if (obj->base.size > gtt_max) {
-		DRM_ERROR("Attempting to bind an object larger than the aperture: object=%zd > %s aperture=%zu\n",
+	if (obj->base.size > end) {
+		DRM_DEBUG("Attempting to bind an object larger than the aperture: object=%zd > %s aperture=%lu\n",
 			  obj->base.size,
-			  map_and_fenceable ? "mappable" : "total",
-			  gtt_max);
-		return -E2BIG;
+			  flags & PIN_MAPPABLE ? "mappable" : "total",
+			  end);
+		return ERR_PTR(-E2BIG);
 	}
 
 	ret = i915_gem_object_get_pages(obj);
 	if (ret)
-		return ret;
+		return ERR_PTR(ret);
 
 	i915_gem_object_pin_pages(obj);
 
-	BUG_ON(!i915_is_ggtt(vm));
-
 	vma = i915_gem_obj_lookup_or_create_vma(obj, vm);
-	if (IS_ERR(vma)) {
-		ret = PTR_ERR(vma);
+	if (IS_ERR(vma))
 		goto err_unpin;
-	}
-
-	/* For now we only ever use 1 vma per object */
-	WARN_ON(!list_is_singular(&obj->vma_list));
 
 search_free:
 	ret = drm_mm_insert_node_in_range_generic(&vm->mm, &vma->node,
 						  size, alignment,
-						  obj->cache_level, 0, gtt_max,
-						  DRM_MM_SEARCH_DEFAULT);
+						  obj->cache_level,
+						  start, end,
+						  DRM_MM_SEARCH_DEFAULT,
+						  DRM_MM_CREATE_DEFAULT);
 	if (ret) {
 		ret = i915_gem_evict_something(dev, vm, size, alignment,
 					       obj->cache_level,
-					       map_and_fenceable,
-					       nonblocking);
+					       start, end,
+					       flags);
 		if (ret == 0)
 			goto search_free;
 
@@ -3320,19 +3463,23 @@ search_free:
 		obj->map_and_fenceable = mappable && fenceable;
 	}
 
-	WARN_ON(map_and_fenceable && !obj->map_and_fenceable);
+	WARN_ON(flags & PIN_MAPPABLE && !obj->map_and_fenceable);
+
+	trace_i915_vma_bind(vma, flags);
+	vma->bind_vma(vma, obj->cache_level,
+		      flags & (PIN_MAPPABLE | PIN_GLOBAL) ? GLOBAL_BIND : 0);
 
-	trace_i915_vma_bind(vma, map_and_fenceable);
 	i915_gem_verify_gtt(dev);
-	return 0;
+	return vma;
 
 err_remove_node:
 	drm_mm_remove_node(&vma->node);
 err_free_vma:
 	i915_gem_vma_destroy(vma);
+	vma = ERR_PTR(ret);
 err_unpin:
 	i915_gem_object_unpin_pages(obj);
-	return ret;
+	return vma;
 }
 
 bool
@@ -3427,7 +3574,7 @@ i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj,
 int
 i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 {
-	drm_i915_private_t *dev_priv = obj->base.dev->dev_private;
+	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
 	uint32_t old_write_domain, old_read_domains;
 	int ret;
 
@@ -3442,6 +3589,7 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 	if (ret)
 		return ret;
 
+	i915_gem_object_retire(obj);
 	i915_gem_object_flush_cpu_write_domain(obj, false);
 
 	/* Serialise direct access to this object with the barriers for
@@ -3485,25 +3633,22 @@ int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 				    enum i915_cache_level cache_level)
 {
 	struct drm_device *dev = obj->base.dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct i915_vma *vma;
+	struct i915_vma *vma, *next;
 	int ret;
 
 	if (obj->cache_level == cache_level)
 		return 0;
 
-	if (obj->pin_count) {
+	if (i915_gem_obj_is_pinned(obj)) {
 		DRM_DEBUG("can not change the cache level of pinned objects\n");
 		return -EBUSY;
 	}
 
-	list_for_each_entry(vma, &obj->vma_list, vma_link) {
+	list_for_each_entry_safe(vma, next, &obj->vma_list, vma_link) {
 		if (!i915_gem_valid_gtt_space(dev, &vma->node, cache_level)) {
 			ret = i915_vma_unbind(vma);
 			if (ret)
 				return ret;
-
-			break;
 		}
 	}
 
@@ -3524,11 +3669,10 @@ int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 				return ret;
 		}
 
-		if (obj->has_global_gtt_mapping)
-			i915_gem_gtt_bind_object(obj, cache_level);
-		if (obj->has_aliasing_ppgtt_mapping)
-			i915_ppgtt_bind_object(dev_priv->mm.aliasing_ppgtt,
-					       obj, cache_level);
+		list_for_each_entry(vma, &obj->vma_list, vma_link)
+			if (drm_mm_node_allocated(&vma->node))
+				vma->bind_vma(vma, cache_level,
+					      obj->has_global_gtt_mapping ? GLOBAL_BIND : 0);
 	}
 
 	list_for_each_entry(vma, &obj->vma_list, vma_link)
@@ -3544,6 +3688,7 @@ int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 		 * in obj->write_domain and have been skipping the clflushes.
 		 * Just set it to the CPU cache for now.
 		 */
+		i915_gem_object_retire(obj);
 		WARN_ON(obj->base.write_domain & ~I915_GEM_DOMAIN_CPU);
 
 		old_read_domains = obj->base.read_domains;
@@ -3641,6 +3786,15 @@ unlock:
 
 static bool is_pin_display(struct drm_i915_gem_object *obj)
 {
+	struct i915_vma *vma;
+
+	if (list_empty(&obj->vma_list))
+		return false;
+
+	vma = i915_gem_obj_to_ggtt(obj);
+	if (!vma)
+		return false;
+
 	/* There are 3 sources that pin objects:
 	 *   1. The display engine (scanouts, sprites, cursors);
 	 *   2. Reservations for execbuffer;
@@ -3652,7 +3806,7 @@ static bool is_pin_display(struct drm_i915_gem_object *obj)
 	 * subtracting the potential reference by the user, any pin_count
 	 * remains, it must be due to another use by the display engine.
 	 */
-	return obj->pin_count - !!obj->user_pin_count;
+	return vma->pin_count - !!obj->user_pin_count;
 }
 
 /*
@@ -3663,9 +3817,10 @@ static bool is_pin_display(struct drm_i915_gem_object *obj)
 int
 i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 				     u32 alignment,
-				     struct intel_ring_buffer *pipelined)
+				     struct intel_engine_cs *pipelined)
 {
 	u32 old_read_domains, old_write_domain;
+	bool was_pin_display;
 	int ret;
 
 	if (pipelined != obj->ring) {
@@ -3677,6 +3832,7 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 	/* Mark the pin_display early so that we account for the
 	 * display coherency whilst setting up the cache domains.
 	 */
+	was_pin_display = obj->pin_display;
 	obj->pin_display = true;
 
 	/* The display engine is not coherent with the LLC cache on gen6.  As
@@ -3697,7 +3853,7 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 	 * (e.g. libkms for the bootup splash), we have to ensure that we
 	 * always use map_and_fenceable for all scanout buffers.
 	 */
-	ret = i915_gem_obj_ggtt_pin(obj, alignment, true, false);
+	ret = i915_gem_obj_ggtt_pin(obj, alignment, PIN_MAPPABLE);
 	if (ret)
 		goto err_unpin_display;
 
@@ -3719,14 +3875,15 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 	return 0;
 
 err_unpin_display:
-	obj->pin_display = is_pin_display(obj);
+	WARN_ON(was_pin_display != is_pin_display(obj));
+	obj->pin_display = was_pin_display;
 	return ret;
 }
 
 void
 i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj)
 {
-	i915_gem_object_unpin(obj);
+	i915_gem_object_ggtt_unpin(obj);
 	obj->pin_display = is_pin_display(obj);
 }
 
@@ -3766,6 +3923,7 @@ i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write)
 	if (ret)
 		return ret;
 
+	i915_gem_object_retire(obj);
 	i915_gem_object_flush_gtt_write_domain(obj);
 
 	old_write_domain = obj->base.write_domain;
@@ -3815,7 +3973,7 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
 	struct drm_i915_file_private *file_priv = file->driver_priv;
 	unsigned long recent_enough = jiffies - msecs_to_jiffies(20);
 	struct drm_i915_gem_request *request;
-	struct intel_ring_buffer *ring = NULL;
+	struct intel_engine_cs *ring = NULL;
 	unsigned reset_counter;
 	u32 seqno = 0;
 	int ret;
@@ -3849,72 +4007,117 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
 	return ret;
 }
 
+static bool
+i915_vma_misplaced(struct i915_vma *vma, uint32_t alignment, uint64_t flags)
+{
+	struct drm_i915_gem_object *obj = vma->obj;
+
+	if (alignment &&
+	    vma->node.start & (alignment - 1))
+		return true;
+
+	if (flags & PIN_MAPPABLE && !obj->map_and_fenceable)
+		return true;
+
+	if (flags & PIN_OFFSET_BIAS &&
+	    vma->node.start < (flags & PIN_OFFSET_MASK))
+		return true;
+
+	return false;
+}
+
 int
 i915_gem_object_pin(struct drm_i915_gem_object *obj,
 		    struct i915_address_space *vm,
 		    uint32_t alignment,
-		    bool map_and_fenceable,
-		    bool nonblocking)
+		    uint64_t flags)
 {
+	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
 	struct i915_vma *vma;
 	int ret;
 
-	if (WARN_ON(obj->pin_count == DRM_I915_GEM_OBJECT_MAX_PIN_COUNT))
-		return -EBUSY;
+	if (WARN_ON(vm == &dev_priv->mm.aliasing_ppgtt->base))
+		return -ENODEV;
 
-	WARN_ON(map_and_fenceable && !i915_is_ggtt(vm));
+	if (WARN_ON(flags & (PIN_GLOBAL | PIN_MAPPABLE) && !i915_is_ggtt(vm)))
+		return -EINVAL;
 
 	vma = i915_gem_obj_to_vma(obj, vm);
-
 	if (vma) {
-		if ((alignment &&
-		     vma->node.start & (alignment - 1)) ||
-		    (map_and_fenceable && !obj->map_and_fenceable)) {
-			WARN(obj->pin_count,
+		if (WARN_ON(vma->pin_count == DRM_I915_GEM_OBJECT_MAX_PIN_COUNT))
+			return -EBUSY;
+
+		if (i915_vma_misplaced(vma, alignment, flags)) {
+			WARN(vma->pin_count,
 			     "bo is already pinned with incorrect alignment:"
 			     " offset=%lx, req.alignment=%x, req.map_and_fenceable=%d,"
 			     " obj->map_and_fenceable=%d\n",
 			     i915_gem_obj_offset(obj, vm), alignment,
-			     map_and_fenceable,
+			     !!(flags & PIN_MAPPABLE),
 			     obj->map_and_fenceable);
 			ret = i915_vma_unbind(vma);
 			if (ret)
 				return ret;
+
+			vma = NULL;
 		}
 	}
 
-	if (!i915_gem_obj_bound(obj, vm)) {
-		struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
-
-		ret = i915_gem_object_bind_to_vm(obj, vm, alignment,
-						 map_and_fenceable,
-						 nonblocking);
-		if (ret)
-			return ret;
-
-		if (!dev_priv->mm.aliasing_ppgtt)
-			i915_gem_gtt_bind_object(obj, obj->cache_level);
+	if (vma == NULL || !drm_mm_node_allocated(&vma->node)) {
+		vma = i915_gem_object_bind_to_vm(obj, vm, alignment, flags);
+		if (IS_ERR(vma))
+			return PTR_ERR(vma);
 	}
 
-	if (!obj->has_global_gtt_mapping && map_and_fenceable)
-		i915_gem_gtt_bind_object(obj, obj->cache_level);
+	if (flags & PIN_GLOBAL && !obj->has_global_gtt_mapping)
+		vma->bind_vma(vma, obj->cache_level, GLOBAL_BIND);
 
-	obj->pin_count++;
-	obj->pin_mappable |= map_and_fenceable;
+	vma->pin_count++;
+	if (flags & PIN_MAPPABLE)
+		obj->pin_mappable |= true;
 
 	return 0;
 }
 
 void
-i915_gem_object_unpin(struct drm_i915_gem_object *obj)
+i915_gem_object_ggtt_unpin(struct drm_i915_gem_object *obj)
 {
-	BUG_ON(obj->pin_count == 0);
-	BUG_ON(!i915_gem_obj_bound_any(obj));
+	struct i915_vma *vma = i915_gem_obj_to_ggtt(obj);
 
-	if (--obj->pin_count == 0)
+	BUG_ON(!vma);
+	BUG_ON(vma->pin_count == 0);
+	BUG_ON(!i915_gem_obj_ggtt_bound(obj));
+
+	if (--vma->pin_count == 0)
 		obj->pin_mappable = false;
 }
 
+bool
+i915_gem_object_pin_fence(struct drm_i915_gem_object *obj)
+{
+	if (obj->fence_reg != I915_FENCE_REG_NONE) {
+		struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
+		struct i915_vma *ggtt_vma = i915_gem_obj_to_ggtt(obj);
+
+		WARN_ON(!ggtt_vma ||
+			dev_priv->fence_regs[obj->fence_reg].pin_count >
+			ggtt_vma->pin_count);
+		dev_priv->fence_regs[obj->fence_reg].pin_count++;
+		return true;
+	} else
+		return false;
+}
+
+void
+i915_gem_object_unpin_fence(struct drm_i915_gem_object *obj)
+{
+	if (obj->fence_reg != I915_FENCE_REG_NONE) {
+		struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
+		WARN_ON(dev_priv->fence_regs[obj->fence_reg].pin_count <= 0);
+		dev_priv->fence_regs[obj->fence_reg].pin_count--;
+	}
+}
+
 int
 i915_gem_pin_ioctl(struct drm_device *dev, void *data,
 		   struct drm_file *file)
@@ -3923,6 +4126,9 @@ i915_gem_pin_ioctl(struct drm_device *dev, void *data,
 	struct drm_i915_gem_object *obj;
 	int ret;
 
+	if (INTEL_INFO(dev)->gen >= 6)
+		return -ENODEV;
+
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
 		return ret;
@@ -3934,13 +4140,13 @@ i915_gem_pin_ioctl(struct drm_device *dev, void *data,
 	}
 
 	if (obj->madv != I915_MADV_WILLNEED) {
-		DRM_ERROR("Attempting to pin a purgeable buffer\n");
-		ret = -EINVAL;
+		DRM_DEBUG("Attempting to pin a purgeable buffer\n");
+		ret = -EFAULT;
 		goto out;
 	}
 
 	if (obj->pin_filp != NULL && obj->pin_filp != file) {
-		DRM_ERROR("Already pinned in i915_gem_pin_ioctl(): %d\n",
+		DRM_DEBUG("Already pinned in i915_gem_pin_ioctl(): %d\n",
 			  args->handle);
 		ret = -EINVAL;
 		goto out;
@@ -3952,7 +4158,7 @@ i915_gem_pin_ioctl(struct drm_device *dev, void *data,
 	}
 
 	if (obj->user_pin_count == 0) {
-		ret = i915_gem_obj_ggtt_pin(obj, args->alignment, true, false);
+		ret = i915_gem_obj_ggtt_pin(obj, args->alignment, PIN_MAPPABLE);
 		if (ret)
 			goto out;
 	}
@@ -3987,7 +4193,7 @@ i915_gem_unpin_ioctl(struct drm_device *dev, void *data,
 	}
 
 	if (obj->pin_filp != file) {
-		DRM_ERROR("Not pinned by caller in i915_gem_pin_ioctl(): %d\n",
+		DRM_DEBUG("Not pinned by caller in i915_gem_pin_ioctl(): %d\n",
 			  args->handle);
 		ret = -EINVAL;
 		goto out;
@@ -3995,7 +4201,7 @@ i915_gem_unpin_ioctl(struct drm_device *dev, void *data,
 	obj->user_pin_count--;
 	if (obj->user_pin_count == 0) {
 		obj->pin_filp = NULL;
-		i915_gem_object_unpin(obj);
+		i915_gem_object_ggtt_unpin(obj);
 	}
 
 out:
@@ -4075,7 +4281,7 @@ i915_gem_madvise_ioctl(struct drm_device *dev, void *data,
 		goto unlock;
 	}
 
-	if (obj->pin_count) {
+	if (i915_gem_obj_is_pinned(obj)) {
 		ret = -EINVAL;
 		goto out;
 	}
@@ -4172,24 +4378,46 @@ struct drm_i915_gem_object *i915_gem_alloc_object(struct drm_device *dev,
 	return obj;
 }
 
+static bool discard_backing_storage(struct drm_i915_gem_object *obj)
+{
+	/* If we are the last user of the backing storage (be it shmemfs
+	 * pages or stolen etc), we know that the pages are going to be
+	 * immediately released. In this case, we can then skip copying
+	 * back the contents from the GPU.
+	 */
+
+	if (obj->madv != I915_MADV_WILLNEED)
+		return false;
+
+	if (obj->base.filp == NULL)
+		return true;
+
+	/* At first glance, this looks racy, but then again so would be
+	 * userspace racing mmap against close. However, the first external
+	 * reference to the filp can only be obtained through the
+	 * i915_gem_mmap_ioctl() which safeguards us against the user
+	 * acquiring such a reference whilst we are in the middle of
+	 * freeing the object.
+	 */
+	return atomic_long_read(&obj->base.filp->f_count) == 1;
+}
+
 void i915_gem_free_object(struct drm_gem_object *gem_obj)
 {
 	struct drm_i915_gem_object *obj = to_intel_bo(gem_obj);
 	struct drm_device *dev = obj->base.dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_vma *vma, *next;
 
-	trace_i915_gem_object_destroy(obj);
+	intel_runtime_pm_get(dev_priv);
 
-	if (obj->phys_obj)
-		i915_gem_detach_phys_object(dev, obj);
+	trace_i915_gem_object_destroy(obj);
 
-	obj->pin_count = 0;
-	/* NB: 0 or 1 elements */
-	WARN_ON(!list_empty(&obj->vma_list) &&
-		!list_is_singular(&obj->vma_list));
 	list_for_each_entry_safe(vma, next, &obj->vma_list, vma_link) {
-		int ret = i915_vma_unbind(vma);
+		int ret;
+
+		vma->pin_count = 0;
+		ret = i915_vma_unbind(vma);
 		if (WARN_ON(ret == -ERESTARTSYS)) {
 			bool was_interruptible;
 
@@ -4202,6 +4430,8 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 		}
 	}
 
+	i915_gem_object_detach_phys(obj);
+
 	/* Stolen objects don't hold a ref, but do hold pin count. Fix that up
 	 * before progressing. */
 	if (obj->stolen)
@@ -4209,6 +4439,8 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 
 	if (WARN_ON(obj->pages_pin_count))
 		obj->pages_pin_count = 0;
+	if (discard_backing_storage(obj))
+		obj->madv = I915_MADV_DONTNEED;
 	i915_gem_object_put_pages(obj);
 	i915_gem_object_free_mmap_offset(obj);
 	i915_gem_object_release_stolen(obj);
@@ -4218,11 +4450,16 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 	if (obj->base.import_attach)
 		drm_prime_gem_destroy(&obj->base, NULL);
 
+	if (obj->ops->release)
+		obj->ops->release(obj);
+
 	drm_gem_object_release(&obj->base);
 	i915_gem_info_remove_obj(dev_priv, obj->base.size);
 
 	kfree(obj->bit_17);
 	i915_gem_object_free(obj);
+
+	intel_runtime_pm_put(dev_priv);
 }
 
 struct i915_vma *i915_gem_obj_to_vma(struct drm_i915_gem_object *obj,
@@ -4236,41 +4473,6 @@ struct i915_vma *i915_gem_obj_to_vma(struct drm_i915_gem_object *obj,
 	return NULL;
 }
 
-static struct i915_vma *__i915_gem_vma_create(struct drm_i915_gem_object *obj,
-					      struct i915_address_space *vm)
-{
-	struct i915_vma *vma = kzalloc(sizeof(*vma), GFP_KERNEL);
-	if (vma == NULL)
-		return ERR_PTR(-ENOMEM);
-
-	INIT_LIST_HEAD(&vma->vma_link);
-	INIT_LIST_HEAD(&vma->mm_list);
-	INIT_LIST_HEAD(&vma->exec_list);
-	vma->vm = vm;
-	vma->obj = obj;
-
-	/* Keep GGTT vmas first to make debug easier */
-	if (i915_is_ggtt(vm))
-		list_add(&vma->vma_link, &obj->vma_list);
-	else
-		list_add_tail(&vma->vma_link, &obj->vma_list);
-
-	return vma;
-}
-
-struct i915_vma *
-i915_gem_obj_lookup_or_create_vma(struct drm_i915_gem_object *obj,
-				  struct i915_address_space *vm)
-{
-	struct i915_vma *vma;
-
-	vma = i915_gem_obj_to_vma(obj, vm);
-	if (!vma)
-		vma = __i915_gem_vma_create(obj, vm);
-
-	return vma;
-}
-
 void i915_gem_vma_destroy(struct i915_vma *vma)
 {
 	WARN_ON(vma->node.allocated);
@@ -4284,10 +4486,21 @@ void i915_gem_vma_destroy(struct i915_vma *vma)
 	kfree(vma);
 }
 
+static void
+i915_gem_stop_ringbuffers(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	int i;
+
+	for_each_ring(ring, dev_priv, i)
+		intel_stop_ring_buffer(ring);
+}
+
 int
 i915_gem_suspend(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret = 0;
 
 	mutex_lock(&dev->struct_mutex);
@@ -4305,7 +4518,7 @@ i915_gem_suspend(struct drm_device *dev)
 		i915_gem_evict_everything(dev);
 
 	i915_kernel_lost_context(dev);
-	i915_gem_cleanup_ringbuffer(dev);
+	i915_gem_stop_ringbuffers(dev);
 
 	/* Hack!  Don't let anybody do execbuf while we don't control the chip.
 	 * We need to replace this with a semaphore, or something.
@@ -4326,10 +4539,10 @@ err:
 	return ret;
 }
 
-int i915_gem_l3_remap(struct intel_ring_buffer *ring, int slice)
+int i915_gem_l3_remap(struct intel_engine_cs *ring, int slice)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 reg_base = GEN7_L3LOG_BASE + (slice * 0x200);
 	u32 *remap_info = dev_priv->l3_parity.remap_info[slice];
 	int i, ret;
@@ -4359,7 +4572,7 @@ int i915_gem_l3_remap(struct intel_ring_buffer *ring, int slice)
 
 void i915_gem_init_swizzling(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	if (INTEL_INFO(dev)->gen < 5 ||
 	    dev_priv->mm.bit_6_swizzle_x == I915_BIT_6_SWIZZLE_NONE)
@@ -4425,13 +4638,20 @@ static int i915_gem_init_rings(struct drm_device *dev)
 			goto cleanup_blt_ring;
 	}
 
+	if (HAS_BSD2(dev)) {
+		ret = intel_init_bsd2_ring_buffer(dev);
+		if (ret)
+			goto cleanup_vebox_ring;
+	}
 
 	ret = i915_gem_set_seqno(dev, ((u32)~0 - 0x1000));
 	if (ret)
-		goto cleanup_vebox_ring;
+		goto cleanup_bsd2_ring;
 
 	return 0;
 
+cleanup_bsd2_ring:
+	intel_cleanup_ring_buffer(&dev_priv->ring[VCS2]);
 cleanup_vebox_ring:
 	intel_cleanup_ring_buffer(&dev_priv->ring[VECS]);
 cleanup_blt_ring:
@@ -4447,7 +4667,7 @@ cleanup_render_ring:
 int
 i915_gem_init_hw(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret, i;
 
 	if (INTEL_INFO(dev)->gen < 6 && !intel_enable_gtt())
@@ -4461,9 +4681,15 @@ i915_gem_init_hw(struct drm_device *dev)
 			   LOWER_SLICE_ENABLED : LOWER_SLICE_DISABLED);
 
 	if (HAS_PCH_NOP(dev)) {
-		u32 temp = I915_READ(GEN7_MSG_CTL);
-		temp &= ~(WAIT_FOR_PCH_FLR_ACK | WAIT_FOR_PCH_RESET_ACK);
-		I915_WRITE(GEN7_MSG_CTL, temp);
+		if (IS_IVYBRIDGE(dev)) {
+			u32 temp = I915_READ(GEN7_MSG_CTL);
+			temp &= ~(WAIT_FOR_PCH_FLR_ACK | WAIT_FOR_PCH_RESET_ACK);
+			I915_WRITE(GEN7_MSG_CTL, temp);
+		} else if (INTEL_INFO(dev)->gen >= 7) {
+			u32 temp = I915_READ(HSW_NDE_RSTWRN_OPT);
+			temp &= ~RESET_PCH_HANDSHAKE_ENABLE;
+			I915_WRITE(HSW_NDE_RSTWRN_OPT, temp);
+		}
 	}
 
 	i915_gem_init_swizzling(dev);
@@ -4476,19 +4702,19 @@ i915_gem_init_hw(struct drm_device *dev)
 		i915_gem_l3_remap(&dev_priv->ring[RCS], i);
 
 	/*
-	 * XXX: There was some w/a described somewhere suggesting loading
-	 * contexts before PPGTT.
+	 * XXX: Contexts should only be initialized once. Doing a switch to the
+	 * default context switch however is something we'd like to do after
+	 * reset or thaw (the latter may not actually be necessary for HW, but
+	 * goes with our code better). Context switching requires rings (for
+	 * the do_switch), but before enabling PPGTT. So don't move this.
 	 */
-	i915_gem_context_init(dev);
-	if (dev_priv->mm.aliasing_ppgtt) {
-		ret = dev_priv->mm.aliasing_ppgtt->enable(dev);
-		if (ret) {
-			i915_gem_cleanup_aliasing_ppgtt(dev);
-			DRM_INFO("PPGTT enable failed. This is not fatal, but unexpected\n");
-		}
+	ret = i915_gem_context_enable(dev_priv);
+	if (ret && ret != -EIO) {
+		DRM_ERROR("Context enable failed %d\n", ret);
+		i915_gem_cleanup_ringbuffer(dev);
 	}
 
-	return 0;
+	return ret;
 }
 
 int i915_gem_init(struct drm_device *dev)
@@ -4500,31 +4726,44 @@ int i915_gem_init(struct drm_device *dev)
 
 	if (IS_VALLEYVIEW(dev)) {
 		/* VLVA0 (potential hack), BIOS isn't actually waking us */
-		I915_WRITE(VLV_GTLC_WAKE_CTRL, 1);
-		if (wait_for((I915_READ(VLV_GTLC_PW_STATUS) & 1) == 1, 10))
+		I915_WRITE(VLV_GTLC_WAKE_CTRL, VLV_GTLC_ALLOWWAKEREQ);
+		if (wait_for((I915_READ(VLV_GTLC_PW_STATUS) &
+			      VLV_GTLC_ALLOWWAKEACK), 10))
 			DRM_DEBUG_DRIVER("allow wake ack timed out\n");
 	}
 
+	i915_gem_init_userptr(dev);
 	i915_gem_init_global_gtt(dev);
 
-	ret = i915_gem_init_hw(dev);
-	mutex_unlock(&dev->struct_mutex);
+	ret = i915_gem_context_init(dev);
 	if (ret) {
-		i915_gem_cleanup_aliasing_ppgtt(dev);
+		mutex_unlock(&dev->struct_mutex);
 		return ret;
 	}
 
+	ret = i915_gem_init_hw(dev);
+	if (ret == -EIO) {
+		/* Allow ring initialisation to fail by marking the GPU as
+		 * wedged. But we only want to do this where the GPU is angry,
+		 * for all other failure, such as an allocation failure, bail.
+		 */
+		DRM_ERROR("Failed to initialize GPU, declaring it wedged\n");
+		atomic_set_mask(I915_WEDGED, &dev_priv->gpu_error.reset_counter);
+		ret = 0;
+	}
+	mutex_unlock(&dev->struct_mutex);
+
 	/* Allow hardware batchbuffers unless told otherwise, but not for KMS. */
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		dev_priv->dri1.allow_batchbuffer = 1;
-	return 0;
+	return ret;
 }
 
 void
 i915_gem_cleanup_ringbuffer(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
 	int i;
 
 	for_each_ring(ring, dev_priv, i)
@@ -4556,16 +4795,15 @@ i915_gem_entervt_ioctl(struct drm_device *dev, void *data,
 	}
 
 	BUG_ON(!list_empty(&dev_priv->gtt.base.active_list));
-	mutex_unlock(&dev->struct_mutex);
 
-	ret = drm_irq_install(dev);
+	ret = drm_irq_install(dev, dev->pdev->irq);
 	if (ret)
 		goto cleanup_ringbuffer;
+	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
 
 cleanup_ringbuffer:
-	mutex_lock(&dev->struct_mutex);
 	i915_gem_cleanup_ringbuffer(dev);
 	dev_priv->ums.mm_suspended = 1;
 	mutex_unlock(&dev->struct_mutex);
@@ -4580,7 +4818,9 @@ i915_gem_leavevt_ioctl(struct drm_device *dev, void *data,
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return 0;
 
+	mutex_lock(&dev->struct_mutex);
 	drm_irq_uninstall(dev);
+	mutex_unlock(&dev->struct_mutex);
 
 	return i915_gem_suspend(dev);
 }
@@ -4599,26 +4839,28 @@ i915_gem_lastclose(struct drm_device *dev)
 }
 
 static void
-init_ring_lists(struct intel_ring_buffer *ring)
+init_ring_lists(struct intel_engine_cs *ring)
 {
 	INIT_LIST_HEAD(&ring->active_list);
 	INIT_LIST_HEAD(&ring->request_list);
 }
 
-static void i915_init_vm(struct drm_i915_private *dev_priv,
-			 struct i915_address_space *vm)
+void i915_init_vm(struct drm_i915_private *dev_priv,
+		  struct i915_address_space *vm)
 {
+	if (!i915_is_ggtt(vm))
+		drm_mm_init(&vm->mm, vm->start, vm->total);
 	vm->dev = dev_priv->dev;
 	INIT_LIST_HEAD(&vm->active_list);
 	INIT_LIST_HEAD(&vm->inactive_list);
 	INIT_LIST_HEAD(&vm->global_link);
-	list_add(&vm->global_link, &dev_priv->vm_list);
+	list_add_tail(&vm->global_link, &dev_priv->vm_list);
 }
 
 void
 i915_gem_load(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 
 	dev_priv->slab =
@@ -4645,7 +4887,7 @@ i915_gem_load(struct drm_device *dev)
 	init_waitqueue_head(&dev_priv->gpu_error.reset_queue);
 
 	/* On GEN3 we really need to make sure the ARB C3 LP bit is set */
-	if (IS_GEN3(dev)) {
+	if (!drm_core_check_feature(dev, DRIVER_MODESET) && IS_GEN3(dev)) {
 		I915_WRITE(MI_ARB_STATE,
 			   _MASKED_BIT_ENABLE(MI_ARB_C3_LP_WRITE_ENABLE));
 	}
@@ -4672,194 +4914,13 @@ i915_gem_load(struct drm_device *dev)
 
 	dev_priv->mm.interruptible = true;
 
-	dev_priv->mm.inactive_shrinker.scan_objects = i915_gem_inactive_scan;
-	dev_priv->mm.inactive_shrinker.count_objects = i915_gem_inactive_count;
-	dev_priv->mm.inactive_shrinker.seeks = DEFAULT_SEEKS;
-	register_shrinker(&dev_priv->mm.inactive_shrinker);
-}
-
-/*
- * Create a physically contiguous memory object for this object
- * e.g. for cursor + overlay regs
- */
-static int i915_gem_init_phys_object(struct drm_device *dev,
-				     int id, int size, int align)
-{
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct drm_i915_gem_phys_object *phys_obj;
-	int ret;
-
-	if (dev_priv->mm.phys_objs[id - 1] || !size)
-		return 0;
-
-	phys_obj = kzalloc(sizeof(*phys_obj), GFP_KERNEL);
-	if (!phys_obj)
-		return -ENOMEM;
-
-	phys_obj->id = id;
-
-	phys_obj->handle = drm_pci_alloc(dev, size, align);
-	if (!phys_obj->handle) {
-		ret = -ENOMEM;
-		goto kfree_obj;
-	}
-#ifdef CONFIG_X86
-	set_memory_wc((unsigned long)phys_obj->handle->vaddr, phys_obj->handle->size / PAGE_SIZE);
-#endif
-
-	dev_priv->mm.phys_objs[id - 1] = phys_obj;
-
-	return 0;
-kfree_obj:
-	kfree(phys_obj);
-	return ret;
-}
-
-static void i915_gem_free_phys_object(struct drm_device *dev, int id)
-{
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct drm_i915_gem_phys_object *phys_obj;
-
-	if (!dev_priv->mm.phys_objs[id - 1])
-		return;
-
-	phys_obj = dev_priv->mm.phys_objs[id - 1];
-	if (phys_obj->cur_obj) {
-		i915_gem_detach_phys_object(dev, phys_obj->cur_obj);
-	}
-
-#ifdef CONFIG_X86
-	set_memory_wb((unsigned long)phys_obj->handle->vaddr, phys_obj->handle->size / PAGE_SIZE);
-#endif
-	drm_pci_free(dev, phys_obj->handle);
-	kfree(phys_obj);
-	dev_priv->mm.phys_objs[id - 1] = NULL;
-}
-
-void i915_gem_free_all_phys_object(struct drm_device *dev)
-{
-	int i;
-
-	for (i = I915_GEM_PHYS_CURSOR_0; i <= I915_MAX_PHYS_OBJECT; i++)
-		i915_gem_free_phys_object(dev, i);
-}
-
-void i915_gem_detach_phys_object(struct drm_device *dev,
-				 struct drm_i915_gem_object *obj)
-{
-	struct address_space *mapping = file_inode(obj->base.filp)->i_mapping;
-	char *vaddr;
-	int i;
-	int page_count;
-
-	if (!obj->phys_obj)
-		return;
-	vaddr = obj->phys_obj->handle->vaddr;
-
-	page_count = obj->base.size / PAGE_SIZE;
-	for (i = 0; i < page_count; i++) {
-		struct page *page = shmem_read_mapping_page(mapping, i);
-		if (!IS_ERR(page)) {
-			char *dst = kmap_atomic(page);
-			memcpy(dst, vaddr + i*PAGE_SIZE, PAGE_SIZE);
-			kunmap_atomic(dst);
-
-			drm_clflush_pages(&page, 1);
+	dev_priv->mm.shrinker.scan_objects = i915_gem_shrinker_scan;
+	dev_priv->mm.shrinker.count_objects = i915_gem_shrinker_count;
+	dev_priv->mm.shrinker.seeks = DEFAULT_SEEKS;
+	register_shrinker(&dev_priv->mm.shrinker);
 
-			set_page_dirty(page);
-			mark_page_accessed(page);
-			page_cache_release(page);
-		}
-	}
-	i915_gem_chipset_flush(dev);
-
-	obj->phys_obj->cur_obj = NULL;
-	obj->phys_obj = NULL;
-}
-
-int
-i915_gem_attach_phys_object(struct drm_device *dev,
-			    struct drm_i915_gem_object *obj,
-			    int id,
-			    int align)
-{
-	struct address_space *mapping = file_inode(obj->base.filp)->i_mapping;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	int ret = 0;
-	int page_count;
-	int i;
-
-	if (id > I915_MAX_PHYS_OBJECT)
-		return -EINVAL;
-
-	if (obj->phys_obj) {
-		if (obj->phys_obj->id == id)
-			return 0;
-		i915_gem_detach_phys_object(dev, obj);
-	}
-
-	/* create a new object */
-	if (!dev_priv->mm.phys_objs[id - 1]) {
-		ret = i915_gem_init_phys_object(dev, id,
-						obj->base.size, align);
-		if (ret) {
-			DRM_ERROR("failed to init phys object %d size: %zu\n",
-				  id, obj->base.size);
-			return ret;
-		}
-	}
-
-	/* bind to the object */
-	obj->phys_obj = dev_priv->mm.phys_objs[id - 1];
-	obj->phys_obj->cur_obj = obj;
-
-	page_count = obj->base.size / PAGE_SIZE;
-
-	for (i = 0; i < page_count; i++) {
-		struct page *page;
-		char *dst, *src;
-
-		page = shmem_read_mapping_page(mapping, i);
-		if (IS_ERR(page))
-			return PTR_ERR(page);
-
-		src = kmap_atomic(page);
-		dst = obj->phys_obj->handle->vaddr + (i * PAGE_SIZE);
-		memcpy(dst, src, PAGE_SIZE);
-		kunmap_atomic(src);
-
-		mark_page_accessed(page);
-		page_cache_release(page);
-	}
-
-	return 0;
-}
-
-static int
-i915_gem_phys_pwrite(struct drm_device *dev,
-		     struct drm_i915_gem_object *obj,
-		     struct drm_i915_gem_pwrite *args,
-		     struct drm_file *file_priv)
-{
-	void *vaddr = obj->phys_obj->handle->vaddr + args->offset;
-	char __user *user_data = to_user_ptr(args->data_ptr);
-
-	if (__copy_from_user_inatomic_nocache(vaddr, user_data, args->size)) {
-		unsigned long unwritten;
-
-		/* The physical object once assigned is fixed for the lifetime
-		 * of the obj, so we can safely drop the lock and continue
-		 * to access vaddr.
-		 */
-		mutex_unlock(&dev->struct_mutex);
-		unwritten = copy_from_user(vaddr, user_data, args->size);
-		mutex_lock(&dev->struct_mutex);
-		if (unwritten)
-			return -EFAULT;
-	}
-
-	i915_gem_chipset_flush(dev);
-	return 0;
+	dev_priv->mm.oom_notifier.notifier_call = i915_gem_shrinker_oom;
+	register_oom_notifier(&dev_priv->mm.oom_notifier);
 }
 
 void i915_gem_release(struct drm_device *dev, struct drm_file *file)
@@ -4897,6 +4958,7 @@ i915_gem_file_idle_work_handler(struct work_struct *work)
 int i915_gem_open(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_i915_file_private *file_priv;
+	int ret;
 
 	DRM_DEBUG_DRIVER("\n");
 
@@ -4906,15 +4968,18 @@ int i915_gem_open(struct drm_device *dev, struct drm_file *file)
 
 	file->driver_priv = file_priv;
 	file_priv->dev_priv = dev->dev_private;
+	file_priv->file = file;
 
 	spin_lock_init(&file_priv->mm.lock);
 	INIT_LIST_HEAD(&file_priv->mm.request_list);
 	INIT_DELAYED_WORK(&file_priv->mm.idle_work,
 			  i915_gem_file_idle_work_handler);
 
-	idr_init(&file_priv->context_idr);
+	ret = i915_gem_context_open(dev, file);
+	if (ret)
+		kfree(file_priv);
 
-	return 0;
+	return ret;
 }
 
 static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task)
@@ -4930,27 +4995,46 @@ static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task)
 #endif
 }
 
+static bool i915_gem_shrinker_lock(struct drm_device *dev, bool *unlock)
+{
+	if (!mutex_trylock(&dev->struct_mutex)) {
+		if (!mutex_is_locked_by(&dev->struct_mutex, current))
+			return false;
+
+		if (to_i915(dev)->mm.shrinker_no_lock_stealing)
+			return false;
+
+		*unlock = false;
+	} else
+		*unlock = true;
+
+	return true;
+}
+
+static int num_vma_bound(struct drm_i915_gem_object *obj)
+{
+	struct i915_vma *vma;
+	int count = 0;
+
+	list_for_each_entry(vma, &obj->vma_list, vma_link)
+		if (drm_mm_node_allocated(&vma->node))
+			count++;
+
+	return count;
+}
+
 static unsigned long
-i915_gem_inactive_count(struct shrinker *shrinker, struct shrink_control *sc)
+i915_gem_shrinker_count(struct shrinker *shrinker, struct shrink_control *sc)
 {
 	struct drm_i915_private *dev_priv =
-		container_of(shrinker,
-			     struct drm_i915_private,
-			     mm.inactive_shrinker);
+		container_of(shrinker, struct drm_i915_private, mm.shrinker);
 	struct drm_device *dev = dev_priv->dev;
 	struct drm_i915_gem_object *obj;
-	bool unlock = true;
 	unsigned long count;
+	bool unlock;
 
-	if (!mutex_trylock(&dev->struct_mutex)) {
-		if (!mutex_is_locked_by(&dev->struct_mutex, current))
-			return 0;
-
-		if (dev_priv->mm.shrinker_no_lock_stealing)
-			return 0;
-
-		unlock = false;
-	}
+	if (!i915_gem_shrinker_lock(dev, &unlock))
+		return 0;
 
 	count = 0;
 	list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list)
@@ -4958,10 +5042,8 @@ i915_gem_inactive_count(struct shrinker *shrinker, struct shrink_control *sc)
 			count += obj->base.size >> PAGE_SHIFT;
 
 	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
-		if (obj->active)
-			continue;
-
-		if (obj->pin_count == 0 && obj->pages_pin_count == 0)
+		if (!i915_gem_obj_is_pinned(obj) &&
+		    obj->pages_pin_count == num_vma_bound(obj))
 			count += obj->base.size >> PAGE_SHIFT;
 	}
 
@@ -4978,7 +5060,8 @@ unsigned long i915_gem_obj_offset(struct drm_i915_gem_object *o,
 	struct drm_i915_private *dev_priv = o->base.dev->dev_private;
 	struct i915_vma *vma;
 
-	if (vm == &dev_priv->mm.aliasing_ppgtt->base)
+	if (!dev_priv->mm.aliasing_ppgtt ||
+	    vm == &dev_priv->mm.aliasing_ppgtt->base)
 		vm = &dev_priv->gtt.base;
 
 	BUG_ON(list_empty(&o->vma_list));
@@ -5019,7 +5102,8 @@ unsigned long i915_gem_obj_size(struct drm_i915_gem_object *o,
 	struct drm_i915_private *dev_priv = o->base.dev->dev_private;
 	struct i915_vma *vma;
 
-	if (vm == &dev_priv->mm.aliasing_ppgtt->base)
+	if (!dev_priv->mm.aliasing_ppgtt ||
+	    vm == &dev_priv->mm.aliasing_ppgtt->base)
 		vm = &dev_priv->gtt.base;
 
 	BUG_ON(list_empty(&o->vma_list));
@@ -5032,49 +5116,104 @@ unsigned long i915_gem_obj_size(struct drm_i915_gem_object *o,
 }
 
 static unsigned long
-i915_gem_inactive_scan(struct shrinker *shrinker, struct shrink_control *sc)
+i915_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc)
 {
 	struct drm_i915_private *dev_priv =
-		container_of(shrinker,
-			     struct drm_i915_private,
-			     mm.inactive_shrinker);
+		container_of(shrinker, struct drm_i915_private, mm.shrinker);
 	struct drm_device *dev = dev_priv->dev;
 	unsigned long freed;
-	bool unlock = true;
-
-	if (!mutex_trylock(&dev->struct_mutex)) {
-		if (!mutex_is_locked_by(&dev->struct_mutex, current))
-			return SHRINK_STOP;
-
-		if (dev_priv->mm.shrinker_no_lock_stealing)
-			return SHRINK_STOP;
+	bool unlock;
 
-		unlock = false;
-	}
+	if (!i915_gem_shrinker_lock(dev, &unlock))
+		return SHRINK_STOP;
 
 	freed = i915_gem_purge(dev_priv, sc->nr_to_scan);
 	if (freed < sc->nr_to_scan)
 		freed += __i915_gem_shrink(dev_priv,
 					   sc->nr_to_scan - freed,
 					   false);
-	if (freed < sc->nr_to_scan)
-		freed += i915_gem_shrink_all(dev_priv);
-
 	if (unlock)
 		mutex_unlock(&dev->struct_mutex);
 
 	return freed;
 }
 
+static int
+i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr)
+{
+	struct drm_i915_private *dev_priv =
+		container_of(nb, struct drm_i915_private, mm.oom_notifier);
+	struct drm_device *dev = dev_priv->dev;
+	struct drm_i915_gem_object *obj;
+	unsigned long timeout = msecs_to_jiffies(5000) + 1;
+	unsigned long pinned, bound, unbound, freed;
+	bool was_interruptible;
+	bool unlock;
+
+	while (!i915_gem_shrinker_lock(dev, &unlock) && --timeout)
+		schedule_timeout_killable(1);
+	if (timeout == 0) {
+		pr_err("Unable to purge GPU memory due lock contention.\n");
+		return NOTIFY_DONE;
+	}
+
+	was_interruptible = dev_priv->mm.interruptible;
+	dev_priv->mm.interruptible = false;
+
+	freed = i915_gem_shrink_all(dev_priv);
+
+	dev_priv->mm.interruptible = was_interruptible;
+
+	/* Because we may be allocating inside our own driver, we cannot
+	 * assert that there are no objects with pinned pages that are not
+	 * being pointed to by hardware.
+	 */
+	unbound = bound = pinned = 0;
+	list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) {
+		if (!obj->base.filp) /* not backed by a freeable object */
+			continue;
+
+		if (obj->pages_pin_count)
+			pinned += obj->base.size;
+		else
+			unbound += obj->base.size;
+	}
+	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
+		if (!obj->base.filp)
+			continue;
+
+		if (obj->pages_pin_count)
+			pinned += obj->base.size;
+		else
+			bound += obj->base.size;
+	}
+
+	if (unlock)
+		mutex_unlock(&dev->struct_mutex);
+
+	pr_info("Purging GPU memory, %lu bytes freed, %lu bytes still pinned.\n",
+		freed, pinned);
+	if (unbound || bound)
+		pr_err("%lu and %lu bytes still available in the "
+		       "bound and unbound GPU page lists.\n",
+		       bound, unbound);
+
+	*(unsigned long *)ptr += freed;
+	return NOTIFY_DONE;
+}
+
 struct i915_vma *i915_gem_obj_to_ggtt(struct drm_i915_gem_object *obj)
 {
 	struct i915_vma *vma;
 
+	/* This WARN has probably outlived its usefulness (callers already
+	 * WARN if they don't find the GGTT vma they expect). When removing,
+	 * remember to remove the pre-check in is_pin_display() as well */
 	if (WARN_ON(list_empty(&obj->vma_list)))
 		return NULL;
 
 	vma = list_first_entry(&obj->vma_list, typeof(*vma), vma_link);
-	if (WARN_ON(vma->vm != obj_to_ggtt(obj)))
+	if (vma->vm != obj_to_ggtt(obj))
 		return NULL;
 
 	return vma;
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index b0f42b9ca03..a5ddf3bce9c 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -93,11 +93,60 @@
  * I've seen in a spec to date, and that was a workaround for a non-shipping
  * part. It should be safe to decrease this, but it's more future proof as is.
  */
-#define CONTEXT_ALIGN (64<<10)
+#define GEN6_CONTEXT_ALIGN (64<<10)
+#define GEN7_CONTEXT_ALIGN 4096
 
-static struct i915_hw_context *
-i915_gem_context_get(struct drm_i915_file_private *file_priv, u32 id);
-static int do_switch(struct i915_hw_context *to);
+static void do_ppgtt_cleanup(struct i915_hw_ppgtt *ppgtt)
+{
+	struct drm_device *dev = ppgtt->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_address_space *vm = &ppgtt->base;
+
+	if (ppgtt == dev_priv->mm.aliasing_ppgtt ||
+	    (list_empty(&vm->active_list) && list_empty(&vm->inactive_list))) {
+		ppgtt->base.cleanup(&ppgtt->base);
+		return;
+	}
+
+	/*
+	 * Make sure vmas are unbound before we take down the drm_mm
+	 *
+	 * FIXME: Proper refcounting should take care of this, this shouldn't be
+	 * needed at all.
+	 */
+	if (!list_empty(&vm->active_list)) {
+		struct i915_vma *vma;
+
+		list_for_each_entry(vma, &vm->active_list, mm_list)
+			if (WARN_ON(list_empty(&vma->vma_link) ||
+				    list_is_singular(&vma->vma_link)))
+				break;
+
+		i915_gem_evict_vm(&ppgtt->base, true);
+	} else {
+		i915_gem_retire_requests(dev);
+		i915_gem_evict_vm(&ppgtt->base, false);
+	}
+
+	ppgtt->base.cleanup(&ppgtt->base);
+}
+
+static void ppgtt_release(struct kref *kref)
+{
+	struct i915_hw_ppgtt *ppgtt =
+		container_of(kref, struct i915_hw_ppgtt, ref);
+
+	do_ppgtt_cleanup(ppgtt);
+	kfree(ppgtt);
+}
+
+static size_t get_context_alignment(struct drm_device *dev)
+{
+	if (IS_GEN6(dev))
+		return GEN6_CONTEXT_ALIGN;
+
+	return GEN7_CONTEXT_ALIGN;
+}
 
 static int get_context_size(struct drm_device *dev)
 {
@@ -129,20 +178,52 @@ static int get_context_size(struct drm_device *dev)
 
 void i915_gem_context_free(struct kref *ctx_ref)
 {
-	struct i915_hw_context *ctx = container_of(ctx_ref,
+	struct intel_context *ctx = container_of(ctx_ref,
 						   typeof(*ctx), ref);
+	struct i915_hw_ppgtt *ppgtt = NULL;
+
+	if (ctx->obj) {
+		/* We refcount even the aliasing PPGTT to keep the code symmetric */
+		if (USES_PPGTT(ctx->obj->base.dev))
+			ppgtt = ctx_to_ppgtt(ctx);
+
+		/* XXX: Free up the object before tearing down the address space, in
+		 * case we're bound in the PPGTT */
+		drm_gem_object_unreference(&ctx->obj->base);
+	}
 
+	if (ppgtt)
+		kref_put(&ppgtt->ref, ppgtt_release);
 	list_del(&ctx->link);
-	drm_gem_object_unreference(&ctx->obj->base);
 	kfree(ctx);
 }
 
-static struct i915_hw_context *
-create_hw_context(struct drm_device *dev,
+static struct i915_hw_ppgtt *
+create_vm_for_ctx(struct drm_device *dev, struct intel_context *ctx)
+{
+	struct i915_hw_ppgtt *ppgtt;
+	int ret;
+
+	ppgtt = kzalloc(sizeof(*ppgtt), GFP_KERNEL);
+	if (!ppgtt)
+		return ERR_PTR(-ENOMEM);
+
+	ret = i915_gem_init_ppgtt(dev, ppgtt);
+	if (ret) {
+		kfree(ppgtt);
+		return ERR_PTR(ret);
+	}
+
+	ppgtt->ctx = ctx;
+	return ppgtt;
+}
+
+static struct intel_context *
+__create_hw_context(struct drm_device *dev,
 		  struct drm_i915_file_private *file_priv)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct i915_hw_context *ctx;
+	struct intel_context *ctx;
 	int ret;
 
 	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
@@ -150,37 +231,40 @@ create_hw_context(struct drm_device *dev,
 		return ERR_PTR(-ENOMEM);
 
 	kref_init(&ctx->ref);
-	ctx->obj = i915_gem_alloc_object(dev, dev_priv->hw_context_size);
-	INIT_LIST_HEAD(&ctx->link);
-	if (ctx->obj == NULL) {
-		kfree(ctx);
-		DRM_DEBUG_DRIVER("Context object allocated failed\n");
-		return ERR_PTR(-ENOMEM);
-	}
+	list_add_tail(&ctx->link, &dev_priv->context_list);
 
-	if (INTEL_INFO(dev)->gen >= 7) {
-		ret = i915_gem_object_set_cache_level(ctx->obj,
-						      I915_CACHE_L3_LLC);
-		/* Failure shouldn't ever happen this early */
-		if (WARN_ON(ret))
+	if (dev_priv->hw_context_size) {
+		ctx->obj = i915_gem_alloc_object(dev, dev_priv->hw_context_size);
+		if (ctx->obj == NULL) {
+			ret = -ENOMEM;
 			goto err_out;
+		}
+
+		/*
+		 * Try to make the context utilize L3 as well as LLC.
+		 *
+		 * On VLV we don't have L3 controls in the PTEs so we
+		 * shouldn't touch the cache level, especially as that
+		 * would make the object snooped which might have a
+		 * negative performance impact.
+		 */
+		if (INTEL_INFO(dev)->gen >= 7 && !IS_VALLEYVIEW(dev)) {
+			ret = i915_gem_object_set_cache_level(ctx->obj,
+							      I915_CACHE_L3_LLC);
+			/* Failure shouldn't ever happen this early */
+			if (WARN_ON(ret))
+				goto err_out;
+		}
 	}
 
-	/* The ring associated with the context object is handled by the normal
-	 * object tracking code. We give an initial ring value simple to pass an
-	 * assertion in the context switch code.
-	 */
-	ctx->ring = &dev_priv->ring[RCS];
-	list_add_tail(&ctx->link, &dev_priv->context_list);
-
 	/* Default context will never have a file_priv */
-	if (file_priv == NULL)
-		return ctx;
-
-	ret = idr_alloc(&file_priv->context_idr, ctx, DEFAULT_CONTEXT_ID + 1, 0,
-			GFP_KERNEL);
-	if (ret < 0)
-		goto err_out;
+	if (file_priv != NULL) {
+		ret = idr_alloc(&file_priv->context_idr, ctx,
+				DEFAULT_CONTEXT_ID, 0, GFP_KERNEL);
+		if (ret < 0)
+			goto err_out;
+	} else
+		ret = DEFAULT_CONTEXT_ID;
 
 	ctx->file_priv = file_priv;
 	ctx->id = ret;
@@ -196,151 +280,244 @@ err_out:
 	return ERR_PTR(ret);
 }
 
-static inline bool is_default_context(struct i915_hw_context *ctx)
-{
-	return (ctx == ctx->ring->default_context);
-}
-
 /**
  * The default context needs to exist per ring that uses contexts. It stores the
  * context state of the GPU for applications that don't utilize HW contexts, as
  * well as an idle case.
  */
-static int create_default_context(struct drm_i915_private *dev_priv)
+static struct intel_context *
+i915_gem_create_context(struct drm_device *dev,
+			struct drm_i915_file_private *file_priv,
+			bool create_vm)
 {
-	struct i915_hw_context *ctx;
-	int ret;
+	const bool is_global_default_ctx = file_priv == NULL;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_context *ctx;
+	int ret = 0;
 
-	BUG_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex));
+	BUG_ON(!mutex_is_locked(&dev->struct_mutex));
 
-	ctx = create_hw_context(dev_priv->dev, NULL);
+	ctx = __create_hw_context(dev, file_priv);
 	if (IS_ERR(ctx))
-		return PTR_ERR(ctx);
-
-	/* We may need to do things with the shrinker which require us to
-	 * immediately switch back to the default context. This can cause a
-	 * problem as pinning the default context also requires GTT space which
-	 * may not be available. To avoid this we always pin the
-	 * default context.
-	 */
-	ret = i915_gem_obj_ggtt_pin(ctx->obj, CONTEXT_ALIGN, false, false);
-	if (ret) {
-		DRM_DEBUG_DRIVER("Couldn't pin %d\n", ret);
-		goto err_destroy;
-	}
+		return ctx;
 
-	ret = do_switch(ctx);
-	if (ret) {
-		DRM_DEBUG_DRIVER("Switch failed %d\n", ret);
-		goto err_unpin;
+	if (is_global_default_ctx && ctx->obj) {
+		/* We may need to do things with the shrinker which
+		 * require us to immediately switch back to the default
+		 * context. This can cause a problem as pinning the
+		 * default context also requires GTT space which may not
+		 * be available. To avoid this we always pin the default
+		 * context.
+		 */
+		ret = i915_gem_obj_ggtt_pin(ctx->obj,
+					    get_context_alignment(dev), 0);
+		if (ret) {
+			DRM_DEBUG_DRIVER("Couldn't pin %d\n", ret);
+			goto err_destroy;
+		}
 	}
 
-	dev_priv->ring[RCS].default_context = ctx;
+	if (create_vm) {
+		struct i915_hw_ppgtt *ppgtt = create_vm_for_ctx(dev, ctx);
+
+		if (IS_ERR_OR_NULL(ppgtt)) {
+			DRM_DEBUG_DRIVER("PPGTT setup failed (%ld)\n",
+					 PTR_ERR(ppgtt));
+			ret = PTR_ERR(ppgtt);
+			goto err_unpin;
+		} else
+			ctx->vm = &ppgtt->base;
+
+		/* This case is reserved for the global default context and
+		 * should only happen once. */
+		if (is_global_default_ctx) {
+			if (WARN_ON(dev_priv->mm.aliasing_ppgtt)) {
+				ret = -EEXIST;
+				goto err_unpin;
+			}
+
+			dev_priv->mm.aliasing_ppgtt = ppgtt;
+		}
+	} else if (USES_PPGTT(dev)) {
+		/* For platforms which only have aliasing PPGTT, we fake the
+		 * address space and refcounting. */
+		ctx->vm = &dev_priv->mm.aliasing_ppgtt->base;
+		kref_get(&dev_priv->mm.aliasing_ppgtt->ref);
+	} else
+		ctx->vm = &dev_priv->gtt.base;
 
-	DRM_DEBUG_DRIVER("Default HW context loaded\n");
-	return 0;
+	return ctx;
 
 err_unpin:
-	i915_gem_object_unpin(ctx->obj);
+	if (is_global_default_ctx && ctx->obj)
+		i915_gem_object_ggtt_unpin(ctx->obj);
 err_destroy:
 	i915_gem_context_unreference(ctx);
-	return ret;
+	return ERR_PTR(ret);
 }
 
-void i915_gem_context_init(struct drm_device *dev)
+void i915_gem_context_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	int i;
 
-	if (!HAS_HW_CONTEXTS(dev)) {
-		dev_priv->hw_contexts_disabled = true;
-		DRM_DEBUG_DRIVER("Disabling HW Contexts; old hardware\n");
-		return;
+	/* Prevent the hardware from restoring the last context (which hung) on
+	 * the next switch */
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		struct intel_engine_cs *ring = &dev_priv->ring[i];
+		struct intel_context *dctx = ring->default_context;
+
+		/* Do a fake switch to the default context */
+		if (ring->last_context == dctx)
+			continue;
+
+		if (!ring->last_context)
+			continue;
+
+		if (dctx->obj && i == RCS) {
+			WARN_ON(i915_gem_obj_ggtt_pin(dctx->obj,
+						      get_context_alignment(dev), 0));
+			/* Fake a finish/inactive */
+			dctx->obj->base.write_domain = 0;
+			dctx->obj->active = 0;
+		}
+
+		i915_gem_context_unreference(ring->last_context);
+		i915_gem_context_reference(dctx);
+		ring->last_context = dctx;
 	}
+}
 
-	/* If called from reset, or thaw... we've been here already */
-	if (dev_priv->hw_contexts_disabled ||
-	    dev_priv->ring[RCS].default_context)
-		return;
+int i915_gem_context_init(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_context *ctx;
+	int i;
 
-	dev_priv->hw_context_size = round_up(get_context_size(dev), 4096);
+	/* Init should only be called once per module load. Eventually the
+	 * restriction on the context_disabled check can be loosened. */
+	if (WARN_ON(dev_priv->ring[RCS].default_context))
+		return 0;
 
-	if (dev_priv->hw_context_size > (1<<20)) {
-		dev_priv->hw_contexts_disabled = true;
-		DRM_DEBUG_DRIVER("Disabling HW Contexts; invalid size\n");
-		return;
+	if (HAS_HW_CONTEXTS(dev)) {
+		dev_priv->hw_context_size = round_up(get_context_size(dev), 4096);
+		if (dev_priv->hw_context_size > (1<<20)) {
+			DRM_DEBUG_DRIVER("Disabling HW Contexts; invalid size %d\n",
+					 dev_priv->hw_context_size);
+			dev_priv->hw_context_size = 0;
+		}
 	}
 
-	if (create_default_context(dev_priv)) {
-		dev_priv->hw_contexts_disabled = true;
-		DRM_DEBUG_DRIVER("Disabling HW Contexts; create failed\n");
-		return;
+	ctx = i915_gem_create_context(dev, NULL, USES_PPGTT(dev));
+	if (IS_ERR(ctx)) {
+		DRM_ERROR("Failed to create default global context (error %ld)\n",
+			  PTR_ERR(ctx));
+		return PTR_ERR(ctx);
 	}
 
-	DRM_DEBUG_DRIVER("HW context support initialized\n");
+	/* NB: RCS will hold a ref for all rings */
+	for (i = 0; i < I915_NUM_RINGS; i++)
+		dev_priv->ring[i].default_context = ctx;
+
+	DRM_DEBUG_DRIVER("%s context support initialized\n", dev_priv->hw_context_size ? "HW" : "fake");
+	return 0;
 }
 
 void i915_gem_context_fini(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct i915_hw_context *dctx = dev_priv->ring[RCS].default_context;
+	struct intel_context *dctx = dev_priv->ring[RCS].default_context;
+	int i;
+
+	if (dctx->obj) {
+		/* The only known way to stop the gpu from accessing the hw context is
+		 * to reset it. Do this as the very last operation to avoid confusing
+		 * other code, leading to spurious errors. */
+		intel_gpu_reset(dev);
+
+		/* When default context is created and switched to, base object refcount
+		 * will be 2 (+1 from object creation and +1 from do_switch()).
+		 * i915_gem_context_fini() will be called after gpu_idle() has switched
+		 * to default context. So we need to unreference the base object once
+		 * to offset the do_switch part, so that i915_gem_context_unreference()
+		 * can then free the base object correctly. */
+		WARN_ON(!dev_priv->ring[RCS].last_context);
+		if (dev_priv->ring[RCS].last_context == dctx) {
+			/* Fake switch to NULL context */
+			WARN_ON(dctx->obj->active);
+			i915_gem_object_ggtt_unpin(dctx->obj);
+			i915_gem_context_unreference(dctx);
+			dev_priv->ring[RCS].last_context = NULL;
+		}
+
+		i915_gem_object_ggtt_unpin(dctx->obj);
+	}
 
-	if (dev_priv->hw_contexts_disabled)
-		return;
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		struct intel_engine_cs *ring = &dev_priv->ring[i];
 
-	/* The only known way to stop the gpu from accessing the hw context is
-	 * to reset it. Do this as the very last operation to avoid confusing
-	 * other code, leading to spurious errors. */
-	intel_gpu_reset(dev);
-
-	/* When default context is created and switched to, base object refcount
-	 * will be 2 (+1 from object creation and +1 from do_switch()).
-	 * i915_gem_context_fini() will be called after gpu_idle() has switched
-	 * to default context. So we need to unreference the base object once
-	 * to offset the do_switch part, so that i915_gem_context_unreference()
-	 * can then free the base object correctly. */
-	WARN_ON(!dev_priv->ring[RCS].last_context);
-	if (dev_priv->ring[RCS].last_context == dctx) {
-		/* Fake switch to NULL context */
-		WARN_ON(dctx->obj->active);
-		i915_gem_object_unpin(dctx->obj);
-		i915_gem_context_unreference(dctx);
+		if (ring->last_context)
+			i915_gem_context_unreference(ring->last_context);
+
+		ring->default_context = NULL;
+		ring->last_context = NULL;
 	}
 
-	i915_gem_object_unpin(dctx->obj);
 	i915_gem_context_unreference(dctx);
-	dev_priv->ring[RCS].default_context = NULL;
-	dev_priv->ring[RCS].last_context = NULL;
 }
 
-static int context_idr_cleanup(int id, void *p, void *data)
+int i915_gem_context_enable(struct drm_i915_private *dev_priv)
 {
-	struct i915_hw_context *ctx = p;
+	struct intel_engine_cs *ring;
+	int ret, i;
+
+	/* This is the only place the aliasing PPGTT gets enabled, which means
+	 * it has to happen before we bail on reset */
+	if (dev_priv->mm.aliasing_ppgtt) {
+		struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
+		ppgtt->enable(ppgtt);
+	}
+
+	/* FIXME: We should make this work, even in reset */
+	if (i915_reset_in_progress(&dev_priv->gpu_error))
+		return 0;
 
-	BUG_ON(id == DEFAULT_CONTEXT_ID);
+	BUG_ON(!dev_priv->ring[RCS].default_context);
+
+	for_each_ring(ring, dev_priv, i) {
+		ret = i915_switch_context(ring, ring->default_context);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static int context_idr_cleanup(int id, void *p, void *data)
+{
+	struct intel_context *ctx = p;
 
 	i915_gem_context_unreference(ctx);
 	return 0;
 }
 
-struct i915_ctx_hang_stats *
-i915_gem_context_get_hang_stats(struct drm_device *dev,
-				struct drm_file *file,
-				u32 id)
+int i915_gem_context_open(struct drm_device *dev, struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_file_private *file_priv = file->driver_priv;
-	struct i915_hw_context *ctx;
+	struct intel_context *ctx;
 
-	if (id == DEFAULT_CONTEXT_ID)
-		return &file_priv->hang_stats;
+	idr_init(&file_priv->context_idr);
 
-	ctx = NULL;
-	if (!dev_priv->hw_contexts_disabled)
-		ctx = i915_gem_context_get(file->driver_priv, id);
-	if (ctx == NULL)
-		return ERR_PTR(-ENOENT);
+	mutex_lock(&dev->struct_mutex);
+	ctx = i915_gem_create_context(dev, file_priv, USES_FULL_PPGTT(dev));
+	mutex_unlock(&dev->struct_mutex);
+
+	if (IS_ERR(ctx)) {
+		idr_destroy(&file_priv->context_idr);
+		return PTR_ERR(ctx);
+	}
 
-	return &ctx->hang_stats;
+	return 0;
 }
 
 void i915_gem_context_close(struct drm_device *dev, struct drm_file *file)
@@ -351,15 +528,21 @@ void i915_gem_context_close(struct drm_device *dev, struct drm_file *file)
 	idr_destroy(&file_priv->context_idr);
 }
 
-static struct i915_hw_context *
+struct intel_context *
 i915_gem_context_get(struct drm_i915_file_private *file_priv, u32 id)
 {
-	return (struct i915_hw_context *)idr_find(&file_priv->context_idr, id);
+	struct intel_context *ctx;
+
+	ctx = (struct intel_context *)idr_find(&file_priv->context_idr, id);
+	if (!ctx)
+		return ERR_PTR(-ENOENT);
+
+	return ctx;
 }
 
 static inline int
-mi_set_context(struct intel_ring_buffer *ring,
-	       struct i915_hw_context *new_context,
+mi_set_context(struct intel_engine_cs *ring,
+	       struct intel_context *new_context,
 	       u32 hw_flags)
 {
 	int ret;
@@ -369,7 +552,7 @@ mi_set_context(struct intel_ring_buffer *ring,
 	 * explicitly, so we rely on the value at ring init, stored in
 	 * itlb_before_ctx_switch.
 	 */
-	if (IS_GEN6(ring->dev) && ring->itlb_before_ctx_switch) {
+	if (IS_GEN6(ring->dev)) {
 		ret = ring->flush(ring, I915_GEM_GPU_DOMAINS, 0);
 		if (ret)
 			return ret;
@@ -379,8 +562,8 @@ mi_set_context(struct intel_ring_buffer *ring,
 	if (ret)
 		return ret;
 
-	/* WaProgramMiArbOnOffAroundMiSetContext:ivb,vlv,hsw */
-	if (IS_GEN7(ring->dev))
+	/* WaProgramMiArbOnOffAroundMiSetContext:ivb,vlv,hsw,bdw,chv */
+	if (INTEL_INFO(ring->dev)->gen >= 7)
 		intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_DISABLE);
 	else
 		intel_ring_emit(ring, MI_NOOP);
@@ -392,10 +575,13 @@ mi_set_context(struct intel_ring_buffer *ring,
 			MI_SAVE_EXT_STATE_EN |
 			MI_RESTORE_EXT_STATE_EN |
 			hw_flags);
-	/* w/a: MI_SET_CONTEXT must always be followed by MI_NOOP */
+	/*
+	 * w/a: MI_SET_CONTEXT must always be followed by MI_NOOP
+	 * WaMiSetContext_Hang:snb,ivb,vlv
+	 */
 	intel_ring_emit(ring, MI_NOOP);
 
-	if (IS_GEN7(ring->dev))
+	if (INTEL_INFO(ring->dev)->gen >= 7)
 		intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_ENABLE);
 	else
 		intel_ring_emit(ring, MI_NOOP);
@@ -405,21 +591,31 @@ mi_set_context(struct intel_ring_buffer *ring,
 	return ret;
 }
 
-static int do_switch(struct i915_hw_context *to)
+static int do_switch(struct intel_engine_cs *ring,
+		     struct intel_context *to)
 {
-	struct intel_ring_buffer *ring = to->ring;
-	struct i915_hw_context *from = ring->last_context;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	struct intel_context *from = ring->last_context;
+	struct i915_hw_ppgtt *ppgtt = ctx_to_ppgtt(to);
 	u32 hw_flags = 0;
+	bool uninitialized = false;
 	int ret, i;
 
-	BUG_ON(from != NULL && from->obj != NULL && from->obj->pin_count == 0);
+	if (from != NULL && ring == &dev_priv->ring[RCS]) {
+		BUG_ON(from->obj == NULL);
+		BUG_ON(!i915_gem_obj_is_pinned(from->obj));
+	}
 
-	if (from == to && !to->remap_slice)
+	if (from == to && from->last_ring == ring && !to->remap_slice)
 		return 0;
 
-	ret = i915_gem_obj_ggtt_pin(to->obj, CONTEXT_ALIGN, false, false);
-	if (ret)
-		return ret;
+	/* Trying to pin first makes error handling easier. */
+	if (ring == &dev_priv->ring[RCS]) {
+		ret = i915_gem_obj_ggtt_pin(to->obj,
+					    get_context_alignment(ring->dev), 0);
+		if (ret)
+			return ret;
+	}
 
 	/*
 	 * Pin can switch back to the default context if we end up calling into
@@ -428,6 +624,18 @@ static int do_switch(struct i915_hw_context *to)
 	 */
 	from = ring->last_context;
 
+	if (USES_FULL_PPGTT(ring->dev)) {
+		ret = ppgtt->switch_mm(ppgtt, ring, false);
+		if (ret)
+			goto unpin_out;
+	}
+
+	if (ring != &dev_priv->ring[RCS]) {
+		if (from)
+			i915_gem_context_unreference(from);
+		goto done;
+	}
+
 	/*
 	 * Clear this page out of any CPU caches for coherent swap-in/out. Note
 	 * that thanks to write = false in this call and us not setting any gpu
@@ -437,22 +645,21 @@ static int do_switch(struct i915_hw_context *to)
 	 * XXX: We need a real interface to do this instead of trickery.
 	 */
 	ret = i915_gem_object_set_to_gtt_domain(to->obj, false);
-	if (ret) {
-		i915_gem_object_unpin(to->obj);
-		return ret;
-	}
+	if (ret)
+		goto unpin_out;
 
-	if (!to->obj->has_global_gtt_mapping)
-		i915_gem_gtt_bind_object(to->obj, to->obj->cache_level);
+	if (!to->obj->has_global_gtt_mapping) {
+		struct i915_vma *vma = i915_gem_obj_to_vma(to->obj,
+							   &dev_priv->gtt.base);
+		vma->bind_vma(vma, to->obj->cache_level, GLOBAL_BIND);
+	}
 
-	if (!to->is_initialized || is_default_context(to))
+	if (!to->is_initialized || i915_gem_context_is_default(to))
 		hw_flags |= MI_RESTORE_INHIBIT;
 
 	ret = mi_set_context(ring, to, hw_flags);
-	if (ret) {
-		i915_gem_object_unpin(to->obj);
-		return ret;
-	}
+	if (ret)
+		goto unpin_out;
 
 	for (i = 0; i < MAX_L3_SLICES; i++) {
 		if (!(to->remap_slice & (1<<i)))
@@ -486,79 +693,83 @@ static int do_switch(struct i915_hw_context *to)
 		BUG_ON(from->obj->ring != ring);
 
 		/* obj is kept alive until the next request by its active ref */
-		i915_gem_object_unpin(from->obj);
+		i915_gem_object_ggtt_unpin(from->obj);
 		i915_gem_context_unreference(from);
 	}
 
+	uninitialized = !to->is_initialized && from == NULL;
+	to->is_initialized = true;
+
+done:
 	i915_gem_context_reference(to);
 	ring->last_context = to;
-	to->is_initialized = true;
+	to->last_ring = ring;
+
+	if (uninitialized) {
+		ret = i915_gem_render_state_init(ring);
+		if (ret)
+			DRM_ERROR("init render state: %d\n", ret);
+	}
 
 	return 0;
+
+unpin_out:
+	if (ring->id == RCS)
+		i915_gem_object_ggtt_unpin(to->obj);
+	return ret;
 }
 
 /**
  * i915_switch_context() - perform a GPU context switch.
  * @ring: ring for which we'll execute the context switch
- * @file_priv: file_priv associated with the context, may be NULL
- * @id: context id number
- * @seqno: sequence number by which the new context will be switched to
- * @flags:
+ * @to: the context to switch to
  *
  * The context life cycle is simple. The context refcount is incremented and
  * decremented by 1 and create and destroy. If the context is in use by the GPU,
  * it will have a refoucnt > 1. This allows us to destroy the context abstract
  * object while letting the normal object tracking destroy the backing BO.
  */
-int i915_switch_context(struct intel_ring_buffer *ring,
-			struct drm_file *file,
-			int to_id)
+int i915_switch_context(struct intel_engine_cs *ring,
+			struct intel_context *to)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-	struct i915_hw_context *to;
-
-	if (dev_priv->hw_contexts_disabled)
-		return 0;
 
 	WARN_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex));
 
-	if (ring != &dev_priv->ring[RCS])
+	if (to->obj == NULL) { /* We have the fake context */
+		if (to != ring->last_context) {
+			i915_gem_context_reference(to);
+			if (ring->last_context)
+				i915_gem_context_unreference(ring->last_context);
+			ring->last_context = to;
+		}
 		return 0;
-
-	if (to_id == DEFAULT_CONTEXT_ID) {
-		to = ring->default_context;
-	} else {
-		if (file == NULL)
-			return -EINVAL;
-
-		to = i915_gem_context_get(file->driver_priv, to_id);
-		if (to == NULL)
-			return -ENOENT;
 	}
 
-	return do_switch(to);
+	return do_switch(ring, to);
+}
+
+static bool hw_context_enabled(struct drm_device *dev)
+{
+	return to_i915(dev)->hw_context_size;
 }
 
 int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_context_create *args = data;
 	struct drm_i915_file_private *file_priv = file->driver_priv;
-	struct i915_hw_context *ctx;
+	struct intel_context *ctx;
 	int ret;
 
-	if (!(dev->driver->driver_features & DRIVER_GEM))
-		return -ENODEV;
-
-	if (dev_priv->hw_contexts_disabled)
+	if (!hw_context_enabled(dev))
 		return -ENODEV;
 
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
 		return ret;
 
-	ctx = create_hw_context(dev, file_priv);
+	ctx = i915_gem_create_context(dev, file_priv, USES_FULL_PPGTT(dev));
 	mutex_unlock(&dev->struct_mutex);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
@@ -574,20 +785,20 @@ int i915_gem_context_destroy_ioctl(struct drm_device *dev, void *data,
 {
 	struct drm_i915_gem_context_destroy *args = data;
 	struct drm_i915_file_private *file_priv = file->driver_priv;
-	struct i915_hw_context *ctx;
+	struct intel_context *ctx;
 	int ret;
 
-	if (!(dev->driver->driver_features & DRIVER_GEM))
-		return -ENODEV;
+	if (args->ctx_id == DEFAULT_CONTEXT_ID)
+		return -ENOENT;
 
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
 		return ret;
 
 	ctx = i915_gem_context_get(file_priv, args->ctx_id);
-	if (!ctx) {
+	if (IS_ERR(ctx)) {
 		mutex_unlock(&dev->struct_mutex);
-		return -ENOENT;
+		return PTR_ERR(ctx);
 	}
 
 	idr_remove(&ctx->file_priv->context_idr, ctx->id);
diff --git a/drivers/gpu/drm/i915/i915_gem_debug.c b/drivers/gpu/drm/i915/i915_gem_debug.c
index 775d506b320..f462d1b51d9 100644
--- a/drivers/gpu/drm/i915/i915_gem_debug.c
+++ b/drivers/gpu/drm/i915/i915_gem_debug.c
@@ -34,7 +34,7 @@ int
 i915_verify_lists(struct drm_device *dev)
 {
 	static int warned;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj;
 	int err = 0;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
index 9bb533e0d76..580aa42443e 100644
--- a/drivers/gpu/drm/i915/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
@@ -161,12 +161,8 @@ static void i915_gem_dmabuf_vunmap(struct dma_buf *dma_buf, void *vaddr)
 {
 	struct drm_i915_gem_object *obj = dma_buf_to_obj(dma_buf);
 	struct drm_device *dev = obj->base.dev;
-	int ret;
-
-	ret = i915_mutex_lock_interruptible(dev);
-	if (ret)
-		return;
 
+	mutex_lock(&dev->struct_mutex);
 	if (--obj->vmapping_count == 0) {
 		vunmap(obj->dma_buf_vmapping);
 		obj->dma_buf_vmapping = NULL;
@@ -233,6 +229,14 @@ static const struct dma_buf_ops i915_dmabuf_ops =  {
 struct dma_buf *i915_gem_prime_export(struct drm_device *dev,
 				      struct drm_gem_object *gem_obj, int flags)
 {
+	struct drm_i915_gem_object *obj = to_intel_bo(gem_obj);
+
+	if (obj->ops->dmabuf_export) {
+		int ret = obj->ops->dmabuf_export(obj);
+		if (ret)
+			return ERR_PTR(ret);
+	}
+
 	return dma_buf_export(gem_obj, &i915_dmabuf_ops, gem_obj->size, flags);
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index 8f3adc7d0dc..bbf4b12d842 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -27,14 +27,16 @@
  */
 
 #include <drm/drmP.h>
-#include "i915_drv.h"
 #include <drm/i915_drm.h>
+
+#include "i915_drv.h"
+#include "intel_drv.h"
 #include "i915_trace.h"
 
 static bool
 mark_free(struct i915_vma *vma, struct list_head *unwind)
 {
-	if (vma->obj->pin_count)
+	if (vma->pin_count)
 		return false;
 
 	if (WARN_ON(!list_empty(&vma->exec_list)))
@@ -44,17 +46,37 @@ mark_free(struct i915_vma *vma, struct list_head *unwind)
 	return drm_mm_scan_add_block(&vma->node);
 }
 
+/**
+ * i915_gem_evict_something - Evict vmas to make room for binding a new one
+ * @dev: drm_device
+ * @vm: address space to evict from
+ * @size: size of the desired free space
+ * @alignment: alignment constraint of the desired free space
+ * @cache_level: cache_level for the desired space
+ * @mappable: whether the free space must be mappable
+ * @nonblocking: whether evicting active objects is allowed or not
+ *
+ * This function will try to evict vmas until a free space satisfying the
+ * requirements is found. Callers must check first whether any such hole exists
+ * already before calling this function.
+ *
+ * This function is used by the object/vma binding code.
+ *
+ * To clarify: This is for freeing up virtual address space, not for freeing
+ * memory in e.g. the shrinker.
+ */
 int
 i915_gem_evict_something(struct drm_device *dev, struct i915_address_space *vm,
 			 int min_size, unsigned alignment, unsigned cache_level,
-			 bool mappable, bool nonblocking)
+			 unsigned long start, unsigned long end,
+			 unsigned flags)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
 	struct list_head eviction_list, unwind_list;
 	struct i915_vma *vma;
 	int ret = 0;
+	int pass = 0;
 
-	trace_i915_gem_evict(dev, min_size, alignment, mappable);
+	trace_i915_gem_evict(dev, min_size, alignment, flags);
 
 	/*
 	 * The goal is to evict objects and amalgamate space in LRU order.
@@ -80,11 +102,10 @@ i915_gem_evict_something(struct drm_device *dev, struct i915_address_space *vm,
 	 */
 
 	INIT_LIST_HEAD(&unwind_list);
-	if (mappable) {
-		BUG_ON(!i915_is_ggtt(vm));
+	if (start != 0 || end != vm->total) {
 		drm_mm_init_scan_with_range(&vm->mm, min_size,
-					    alignment, cache_level, 0,
-					    dev_priv->gtt.mappable_end);
+					    alignment, cache_level,
+					    start, end);
 	} else
 		drm_mm_init_scan(&vm->mm, min_size, alignment, cache_level);
 
@@ -95,7 +116,7 @@ search_again:
 			goto found;
 	}
 
-	if (nonblocking)
+	if (flags & PIN_NONBLOCK)
 		goto none;
 
 	/* Now merge in the soon-to-be-expired objects... */
@@ -119,14 +140,24 @@ none:
 	/* Can we unpin some objects such as idle hw contents,
 	 * or pending flips?
 	 */
-	ret = nonblocking ? -ENOSPC : i915_gpu_idle(dev);
-	if (ret)
-		return ret;
+	if (flags & PIN_NONBLOCK)
+		return -ENOSPC;
 
 	/* Only idle the GPU and repeat the search once */
-	i915_gem_retire_requests(dev);
-	nonblocking = true;
-	goto search_again;
+	if (pass++ == 0) {
+		ret = i915_gpu_idle(dev);
+		if (ret)
+			return ret;
+
+		i915_gem_retire_requests(dev);
+		goto search_again;
+	}
+
+	/* If we still have pending pageflip completions, drop
+	 * back to userspace to give our workqueues time to
+	 * acquire our locks and unpin the old scanouts.
+	 */
+	return intel_has_pending_fb_unpin(dev) ? -EAGAIN : -ENOSPC;
 
 found:
 	/* drm_mm doesn't allow any other other operations while
@@ -164,19 +195,19 @@ found:
 }
 
 /**
- * i915_gem_evict_vm - Try to free up VM space
+ * i915_gem_evict_vm - Evict all idle vmas from a vm
  *
- * @vm: Address space to evict from
+ * @vm: Address space to cleanse
  * @do_idle: Boolean directing whether to idle first.
  *
- * VM eviction is about freeing up virtual address space. If one wants fine
- * grained eviction, they should see evict something for more details. In terms
- * of freeing up actual system memory, this function may not accomplish the
- * desired result. An object may be shared in multiple address space, and this
- * function will not assert those objects be freed.
+ * This function evicts all idles vmas from a vm. If all unpinned vmas should be
+ * evicted the @do_idle needs to be set to true.
+ *
+ * This is used by the execbuf code as a last-ditch effort to defragment the
+ * address space.
  *
- * Using do_idle will result in a more complete eviction because it retires, and
- * inactivates current BOs.
+ * To clarify: This is for freeing up virtual address space, not for freeing
+ * memory in e.g. the shrinker.
  */
 int i915_gem_evict_vm(struct i915_address_space *vm, bool do_idle)
 {
@@ -194,16 +225,24 @@ int i915_gem_evict_vm(struct i915_address_space *vm, bool do_idle)
 	}
 
 	list_for_each_entry_safe(vma, next, &vm->inactive_list, mm_list)
-		if (vma->obj->pin_count == 0)
+		if (vma->pin_count == 0)
 			WARN_ON(i915_vma_unbind(vma));
 
 	return 0;
 }
 
+/**
+ * i915_gem_evict_everything - Try to evict all objects
+ * @dev: Device to evict objects for
+ *
+ * This functions tries to evict all gem objects from all address spaces. Used
+ * by the shrinker as a last-ditch effort and for suspend, before releasing the
+ * backing storage of all unbound objects.
+ */
 int
 i915_gem_evict_everything(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_address_space *vm;
 	bool lists_empty = true;
 	int ret;
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index a3ba9a8cd68..3a30133f93e 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -35,6 +35,9 @@
 
 #define  __EXEC_OBJECT_HAS_PIN (1<<31)
 #define  __EXEC_OBJECT_HAS_FENCE (1<<30)
+#define  __EXEC_OBJECT_NEEDS_BIAS (1<<28)
+
+#define BATCH_OFFSET_BIAS (256*1024)
 
 struct eb_vmas {
 	struct list_head vmas;
@@ -46,7 +49,7 @@ struct eb_vmas {
 };
 
 static struct eb_vmas *
-eb_create(struct drm_i915_gem_execbuffer2 *args, struct i915_address_space *vm)
+eb_create(struct drm_i915_gem_execbuffer2 *args)
 {
 	struct eb_vmas *eb = NULL;
 
@@ -91,6 +94,7 @@ eb_lookup_vmas(struct eb_vmas *eb,
 	       struct i915_address_space *vm,
 	       struct drm_file *file)
 {
+	struct drm_i915_private *dev_priv = vm->dev->dev_private;
 	struct drm_i915_gem_object *obj;
 	struct list_head objects;
 	int i, ret;
@@ -125,6 +129,20 @@ eb_lookup_vmas(struct eb_vmas *eb,
 	i = 0;
 	while (!list_empty(&objects)) {
 		struct i915_vma *vma;
+		struct i915_address_space *bind_vm = vm;
+
+		if (exec[i].flags & EXEC_OBJECT_NEEDS_GTT &&
+		    USES_FULL_PPGTT(vm->dev)) {
+			ret = -EINVAL;
+			goto err;
+		}
+
+		/* If we have secure dispatch, or the userspace assures us that
+		 * they know what they're doing, use the GGTT VM.
+		 */
+		if (((args->flags & I915_EXEC_SECURE) &&
+		    (i == (args->buffer_count - 1))))
+			bind_vm = &dev_priv->gtt.base;
 
 		obj = list_first_entry(&objects,
 				       struct drm_i915_gem_object,
@@ -138,7 +156,7 @@ eb_lookup_vmas(struct eb_vmas *eb,
 		 * from the (obj, vm) we don't run the risk of creating
 		 * duplicated vmas for the same vm.
 		 */
-		vma = i915_gem_obj_lookup_or_create_vma(obj, vm);
+		vma = i915_gem_obj_lookup_or_create_vma(obj, bind_vm);
 		if (IS_ERR(vma)) {
 			DRM_DEBUG("Failed to lookup VMA\n");
 			ret = PTR_ERR(vma);
@@ -217,7 +235,7 @@ i915_gem_execbuffer_unreserve_vma(struct i915_vma *vma)
 		i915_gem_object_unpin_fence(obj);
 
 	if (entry->flags & __EXEC_OBJECT_HAS_PIN)
-		i915_gem_object_unpin(obj);
+		vma->pin_count--;
 
 	entry->flags &= ~(__EXEC_OBJECT_HAS_FENCE | __EXEC_OBJECT_HAS_PIN);
 }
@@ -247,12 +265,14 @@ static inline int use_cpu_reloc(struct drm_i915_gem_object *obj)
 
 static int
 relocate_entry_cpu(struct drm_i915_gem_object *obj,
-		   struct drm_i915_gem_relocation_entry *reloc)
+		   struct drm_i915_gem_relocation_entry *reloc,
+		   uint64_t target_offset)
 {
 	struct drm_device *dev = obj->base.dev;
 	uint32_t page_offset = offset_in_page(reloc->offset);
+	uint64_t delta = reloc->delta + target_offset;
 	char *vaddr;
-	int ret = -EINVAL;
+	int ret;
 
 	ret = i915_gem_object_set_to_cpu_domain(obj, true);
 	if (ret)
@@ -260,7 +280,7 @@ relocate_entry_cpu(struct drm_i915_gem_object *obj,
 
 	vaddr = kmap_atomic(i915_gem_object_get_page(obj,
 				reloc->offset >> PAGE_SHIFT));
-	*(uint32_t *)(vaddr + page_offset) = reloc->delta;
+	*(uint32_t *)(vaddr + page_offset) = lower_32_bits(delta);
 
 	if (INTEL_INFO(dev)->gen >= 8) {
 		page_offset = offset_in_page(page_offset + sizeof(uint32_t));
@@ -271,7 +291,7 @@ relocate_entry_cpu(struct drm_i915_gem_object *obj,
 			    (reloc->offset + sizeof(uint32_t)) >> PAGE_SHIFT));
 		}
 
-		*(uint32_t *)(vaddr + page_offset) = 0;
+		*(uint32_t *)(vaddr + page_offset) = upper_32_bits(delta);
 	}
 
 	kunmap_atomic(vaddr);
@@ -281,13 +301,15 @@ relocate_entry_cpu(struct drm_i915_gem_object *obj,
 
 static int
 relocate_entry_gtt(struct drm_i915_gem_object *obj,
-		   struct drm_i915_gem_relocation_entry *reloc)
+		   struct drm_i915_gem_relocation_entry *reloc,
+		   uint64_t target_offset)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	uint64_t delta = reloc->delta + target_offset;
 	uint32_t __iomem *reloc_entry;
 	void __iomem *reloc_page;
-	int ret = -EINVAL;
+	int ret;
 
 	ret = i915_gem_object_set_to_gtt_domain(obj, true);
 	if (ret)
@@ -303,7 +325,7 @@ relocate_entry_gtt(struct drm_i915_gem_object *obj,
 			reloc->offset & PAGE_MASK);
 	reloc_entry = (uint32_t __iomem *)
 		(reloc_page + offset_in_page(reloc->offset));
-	iowrite32(reloc->delta, reloc_entry);
+	iowrite32(lower_32_bits(delta), reloc_entry);
 
 	if (INTEL_INFO(dev)->gen >= 8) {
 		reloc_entry += 1;
@@ -316,7 +338,7 @@ relocate_entry_gtt(struct drm_i915_gem_object *obj,
 			reloc_entry = reloc_page;
 		}
 
-		iowrite32(0, reloc_entry);
+		iowrite32(upper_32_bits(delta), reloc_entry);
 	}
 
 	io_mapping_unmap_atomic(reloc_page);
@@ -327,15 +349,14 @@ relocate_entry_gtt(struct drm_i915_gem_object *obj,
 static int
 i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 				   struct eb_vmas *eb,
-				   struct drm_i915_gem_relocation_entry *reloc,
-				   struct i915_address_space *vm)
+				   struct drm_i915_gem_relocation_entry *reloc)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_gem_object *target_obj;
 	struct drm_i915_gem_object *target_i915_obj;
 	struct i915_vma *target_vma;
-	uint32_t target_offset;
-	int ret = -EINVAL;
+	uint64_t target_offset;
+	int ret;
 
 	/* we've already hold a reference to all valid objects */
 	target_vma = eb_get_vma(eb, reloc->target_handle);
@@ -344,7 +365,7 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 	target_i915_obj = target_vma->obj;
 	target_obj = &target_vma->obj->base;
 
-	target_offset = i915_gem_obj_ggtt_offset(target_i915_obj);
+	target_offset = target_vma->node.start;
 
 	/* Sandybridge PPGTT errata: We need a global gtt mapping for MI and
 	 * pipe_control writes because the gpu doesn't properly redirect them
@@ -352,8 +373,10 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 	if (unlikely(IS_GEN6(dev) &&
 	    reloc->write_domain == I915_GEM_DOMAIN_INSTRUCTION &&
 	    !target_i915_obj->has_global_gtt_mapping)) {
-		i915_gem_gtt_bind_object(target_i915_obj,
-					 target_i915_obj->cache_level);
+		struct i915_vma *vma =
+			list_first_entry(&target_i915_obj->vma_list,
+					 typeof(*vma), vma_link);
+		vma->bind_vma(vma, target_i915_obj->cache_level, GLOBAL_BIND);
 	}
 
 	/* Validate that the target is in a valid r/w GPU domain */
@@ -365,7 +388,7 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 			  (int) reloc->offset,
 			  reloc->read_domains,
 			  reloc->write_domain);
-		return ret;
+		return -EINVAL;
 	}
 	if (unlikely((reloc->write_domain | reloc->read_domains)
 		     & ~I915_GEM_GPU_DOMAINS)) {
@@ -376,7 +399,7 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 			  (int) reloc->offset,
 			  reloc->read_domains,
 			  reloc->write_domain);
-		return ret;
+		return -EINVAL;
 	}
 
 	target_obj->pending_read_domains |= reloc->read_domains;
@@ -396,25 +419,24 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 			  obj, reloc->target_handle,
 			  (int) reloc->offset,
 			  (int) obj->base.size);
-		return ret;
+		return -EINVAL;
 	}
 	if (unlikely(reloc->offset & 3)) {
 		DRM_DEBUG("Relocation not 4-byte aligned: "
 			  "obj %p target %d offset %d.\n",
 			  obj, reloc->target_handle,
 			  (int) reloc->offset);
-		return ret;
+		return -EINVAL;
 	}
 
 	/* We can't wait for rendering with pagefaults disabled */
 	if (obj->active && in_atomic())
 		return -EFAULT;
 
-	reloc->delta += target_offset;
 	if (use_cpu_reloc(obj))
-		ret = relocate_entry_cpu(obj, reloc);
+		ret = relocate_entry_cpu(obj, reloc, target_offset);
 	else
-		ret = relocate_entry_gtt(obj, reloc);
+		ret = relocate_entry_gtt(obj, reloc, target_offset);
 
 	if (ret)
 		return ret;
@@ -451,8 +473,7 @@ i915_gem_execbuffer_relocate_vma(struct i915_vma *vma,
 		do {
 			u64 offset = r->presumed_offset;
 
-			ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, r,
-								 vma->vm);
+			ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, r);
 			if (ret)
 				return ret;
 
@@ -481,8 +502,7 @@ i915_gem_execbuffer_relocate_vma_slow(struct i915_vma *vma,
 	int i, ret;
 
 	for (i = 0; i < entry->relocation_count; i++) {
-		ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, &relocs[i],
-							 vma->vm);
+		ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, &relocs[i]);
 		if (ret)
 			return ret;
 	}
@@ -491,8 +511,7 @@ i915_gem_execbuffer_relocate_vma_slow(struct i915_vma *vma,
 }
 
 static int
-i915_gem_execbuffer_relocate(struct eb_vmas *eb,
-			     struct i915_address_space *vm)
+i915_gem_execbuffer_relocate(struct eb_vmas *eb)
 {
 	struct i915_vma *vma;
 	int ret = 0;
@@ -525,24 +544,31 @@ need_reloc_mappable(struct i915_vma *vma)
 
 static int
 i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
-				struct intel_ring_buffer *ring,
+				struct intel_engine_cs *ring,
 				bool *need_reloc)
 {
-	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	struct drm_i915_gem_object *obj = vma->obj;
 	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
 	bool has_fenced_gpu_access = INTEL_INFO(ring->dev)->gen < 4;
-	bool need_fence, need_mappable;
-	struct drm_i915_gem_object *obj = vma->obj;
+	bool need_fence;
+	uint64_t flags;
 	int ret;
 
+	flags = 0;
+
 	need_fence =
 		has_fenced_gpu_access &&
 		entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
 		obj->tiling_mode != I915_TILING_NONE;
-	need_mappable = need_fence || need_reloc_mappable(vma);
+	if (need_fence || need_reloc_mappable(vma))
+		flags |= PIN_MAPPABLE;
 
-	ret = i915_gem_object_pin(obj, vma->vm, entry->alignment, need_mappable,
-				  false);
+	if (entry->flags & EXEC_OBJECT_NEEDS_GTT)
+		flags |= PIN_GLOBAL;
+	if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS)
+		flags |= BATCH_OFFSET_BIAS | PIN_OFFSET_BIAS;
+
+	ret = i915_gem_object_pin(obj, vma->vm, entry->alignment, flags);
 	if (ret)
 		return ret;
 
@@ -561,14 +587,6 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 		}
 	}
 
-	/* Ensure ppgtt mapping exists if needed */
-	if (dev_priv->mm.aliasing_ppgtt && !obj->has_aliasing_ppgtt_mapping) {
-		i915_ppgtt_bind_object(dev_priv->mm.aliasing_ppgtt,
-				       obj, obj->cache_level);
-
-		obj->has_aliasing_ppgtt_mapping = 1;
-	}
-
 	if (entry->offset != vma->node.start) {
 		entry->offset = vma->node.start;
 		*need_reloc = true;
@@ -579,15 +597,41 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 		obj->base.pending_write_domain = I915_GEM_DOMAIN_RENDER;
 	}
 
-	if (entry->flags & EXEC_OBJECT_NEEDS_GTT &&
-	    !obj->has_global_gtt_mapping)
-		i915_gem_gtt_bind_object(obj, obj->cache_level);
-
 	return 0;
 }
 
+static bool
+eb_vma_misplaced(struct i915_vma *vma, bool has_fenced_gpu_access)
+{
+	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
+	struct drm_i915_gem_object *obj = vma->obj;
+	bool need_fence, need_mappable;
+
+	need_fence =
+		has_fenced_gpu_access &&
+		entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
+		obj->tiling_mode != I915_TILING_NONE;
+	need_mappable = need_fence || need_reloc_mappable(vma);
+
+	WARN_ON((need_mappable || need_fence) &&
+	       !i915_is_ggtt(vma->vm));
+
+	if (entry->alignment &&
+	    vma->node.start & (entry->alignment - 1))
+		return true;
+
+	if (need_mappable && !obj->map_and_fenceable)
+		return true;
+
+	if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS &&
+	    vma->node.start < BATCH_OFFSET_BIAS)
+		return true;
+
+	return false;
+}
+
 static int
-i915_gem_execbuffer_reserve(struct intel_ring_buffer *ring,
+i915_gem_execbuffer_reserve(struct intel_engine_cs *ring,
 			    struct list_head *vmas,
 			    bool *need_relocs)
 {
@@ -601,6 +645,8 @@ i915_gem_execbuffer_reserve(struct intel_ring_buffer *ring,
 	if (list_empty(vmas))
 		return 0;
 
+	i915_gem_retire_requests_ring(ring);
+
 	vm = list_first_entry(vmas, struct i915_vma, exec_list)->vm;
 
 	INIT_LIST_HEAD(&ordered_vmas);
@@ -647,26 +693,10 @@ i915_gem_execbuffer_reserve(struct intel_ring_buffer *ring,
 
 		/* Unbind any ill-fitting objects or pin. */
 		list_for_each_entry(vma, vmas, exec_list) {
-			struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
-			bool need_fence, need_mappable;
-
-			obj = vma->obj;
-
 			if (!drm_mm_node_allocated(&vma->node))
 				continue;
 
-			need_fence =
-				has_fenced_gpu_access &&
-				entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
-				obj->tiling_mode != I915_TILING_NONE;
-			need_mappable = need_fence || need_reloc_mappable(vma);
-
-			WARN_ON((need_mappable || need_fence) &&
-			       !i915_is_ggtt(vma->vm));
-
-			if ((entry->alignment &&
-			     vma->node.start & (entry->alignment - 1)) ||
-			    (need_mappable && !obj->map_and_fenceable))
+			if (eb_vma_misplaced(vma, has_fenced_gpu_access))
 				ret = i915_vma_unbind(vma);
 			else
 				ret = i915_gem_execbuffer_reserve_vma(vma, ring, need_relocs);
@@ -702,7 +732,7 @@ static int
 i915_gem_execbuffer_relocate_slow(struct drm_device *dev,
 				  struct drm_i915_gem_execbuffer2 *args,
 				  struct drm_file *file,
-				  struct intel_ring_buffer *ring,
+				  struct intel_engine_cs *ring,
 				  struct eb_vmas *eb,
 				  struct drm_i915_gem_exec_object2 *exec)
 {
@@ -767,9 +797,9 @@ i915_gem_execbuffer_relocate_slow(struct drm_device *dev,
 		 * relocations were valid.
 		 */
 		for (j = 0; j < exec[i].relocation_count; j++) {
-			if (copy_to_user(&user_relocs[j].presumed_offset,
-					 &invalid_offset,
-					 sizeof(invalid_offset))) {
+			if (__copy_to_user(&user_relocs[j].presumed_offset,
+					   &invalid_offset,
+					   sizeof(invalid_offset))) {
 				ret = -EFAULT;
 				mutex_lock(&dev->struct_mutex);
 				goto err;
@@ -818,7 +848,7 @@ err:
 }
 
 static int
-i915_gem_execbuffer_move_to_gpu(struct intel_ring_buffer *ring,
+i915_gem_execbuffer_move_to_gpu(struct intel_engine_cs *ring,
 				struct list_head *vmas)
 {
 	struct i915_vma *vma;
@@ -892,7 +922,7 @@ validate_exec_list(struct drm_i915_gem_exec_object2 *exec,
 		if (!access_ok(VERIFY_WRITE, ptr, length))
 			return -EFAULT;
 
-		if (likely(!i915_prefault_disable)) {
+		if (likely(!i915.prefault_disable)) {
 			if (fault_in_multipages_readable(ptr, length))
 				return -EFAULT;
 		}
@@ -901,9 +931,32 @@ validate_exec_list(struct drm_i915_gem_exec_object2 *exec,
 	return 0;
 }
 
+static struct intel_context *
+i915_gem_validate_context(struct drm_device *dev, struct drm_file *file,
+			  struct intel_engine_cs *ring, const u32 ctx_id)
+{
+	struct intel_context *ctx = NULL;
+	struct i915_ctx_hang_stats *hs;
+
+	if (ring->id != RCS && ctx_id != DEFAULT_CONTEXT_ID)
+		return ERR_PTR(-EINVAL);
+
+	ctx = i915_gem_context_get(file->driver_priv, ctx_id);
+	if (IS_ERR(ctx))
+		return ctx;
+
+	hs = &ctx->hang_stats;
+	if (hs->banned) {
+		DRM_DEBUG("Context %u tried to submit while banned\n", ctx_id);
+		return ERR_PTR(-EIO);
+	}
+
+	return ctx;
+}
+
 static void
 i915_gem_execbuffer_move_to_active(struct list_head *vmas,
-				   struct intel_ring_buffer *ring)
+				   struct intel_engine_cs *ring)
 {
 	struct i915_vma *vma;
 
@@ -922,8 +975,13 @@ i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 		if (obj->base.write_domain) {
 			obj->dirty = 1;
 			obj->last_write_seqno = intel_ring_get_seqno(ring);
-			if (obj->pin_count) /* check for potential scanout */
+			/* check for potential scanout */
+			if (i915_gem_obj_ggtt_bound(obj) &&
+			    i915_gem_obj_to_ggtt(obj)->pin_count)
 				intel_mark_fb_busy(obj, ring);
+
+			/* update for the implicit flush after a batch */
+			obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
 		}
 
 		trace_i915_gem_object_change_domain(obj, old_read, old_write);
@@ -933,7 +991,7 @@ i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 static void
 i915_gem_execbuffer_retire_commands(struct drm_device *dev,
 				    struct drm_file *file,
-				    struct intel_ring_buffer *ring,
+				    struct intel_engine_cs *ring,
 				    struct drm_i915_gem_object *obj)
 {
 	/* Unconditionally force add_request to emit a full flush. */
@@ -945,13 +1003,15 @@ i915_gem_execbuffer_retire_commands(struct drm_device *dev,
 
 static int
 i915_reset_gen7_sol_offsets(struct drm_device *dev,
-			    struct intel_ring_buffer *ring)
+			    struct intel_engine_cs *ring)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret, i;
 
-	if (!IS_GEN7(dev) || ring != &dev_priv->ring[RCS])
-		return 0;
+	if (!IS_GEN7(dev) || ring != &dev_priv->ring[RCS]) {
+		DRM_DEBUG("sol reset is gen7/rcs only\n");
+		return -EINVAL;
+	}
 
 	ret = intel_ring_begin(ring, 4 * 3);
 	if (ret)
@@ -968,21 +1028,71 @@ i915_reset_gen7_sol_offsets(struct drm_device *dev,
 	return 0;
 }
 
+/**
+ * Find one BSD ring to dispatch the corresponding BSD command.
+ * The Ring ID is returned.
+ */
+static int gen8_dispatch_bsd_ring(struct drm_device *dev,
+				  struct drm_file *file)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_file_private *file_priv = file->driver_priv;
+
+	/* Check whether the file_priv is using one ring */
+	if (file_priv->bsd_ring)
+		return file_priv->bsd_ring->id;
+	else {
+		/* If no, use the ping-pong mechanism to select one ring */
+		int ring_id;
+
+		mutex_lock(&dev->struct_mutex);
+		if (dev_priv->mm.bsd_ring_dispatch_index == 0) {
+			ring_id = VCS;
+			dev_priv->mm.bsd_ring_dispatch_index = 1;
+		} else {
+			ring_id = VCS2;
+			dev_priv->mm.bsd_ring_dispatch_index = 0;
+		}
+		file_priv->bsd_ring = &dev_priv->ring[ring_id];
+		mutex_unlock(&dev->struct_mutex);
+		return ring_id;
+	}
+}
+
+static struct drm_i915_gem_object *
+eb_get_batch(struct eb_vmas *eb)
+{
+	struct i915_vma *vma = list_entry(eb->vmas.prev, typeof(*vma), exec_list);
+
+	/*
+	 * SNA is doing fancy tricks with compressing batch buffers, which leads
+	 * to negative relocation deltas. Usually that works out ok since the
+	 * relocate address is still positive, except when the batch is placed
+	 * very low in the GTT. Ensure this doesn't happen.
+	 *
+	 * Note that actual hangs have only been observed on gen7, but for
+	 * paranoia do it everywhere.
+	 */
+	vma->exec_entry->flags |= __EXEC_OBJECT_NEEDS_BIAS;
+
+	return vma->obj;
+}
+
 static int
 i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 		       struct drm_file *file,
 		       struct drm_i915_gem_execbuffer2 *args,
-		       struct drm_i915_gem_exec_object2 *exec,
-		       struct i915_address_space *vm)
+		       struct drm_i915_gem_exec_object2 *exec)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct eb_vmas *eb;
 	struct drm_i915_gem_object *batch_obj;
 	struct drm_clip_rect *cliprects = NULL;
-	struct intel_ring_buffer *ring;
-	struct i915_ctx_hang_stats *hs;
-	u32 ctx_id = i915_execbuffer2_get_context_id(*args);
-	u32 exec_start, exec_len;
+	struct intel_engine_cs *ring;
+	struct intel_context *ctx;
+	struct i915_address_space *vm;
+	const u32 ctx_id = i915_execbuffer2_get_context_id(*args);
+	u64 exec_start = args->batch_start_offset, exec_len;
 	u32 mask, flags;
 	int ret, mode, i;
 	bool need_relocs;
@@ -1004,41 +1114,24 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	if (args->flags & I915_EXEC_IS_PINNED)
 		flags |= I915_DISPATCH_PINNED;
 
-	switch (args->flags & I915_EXEC_RING_MASK) {
-	case I915_EXEC_DEFAULT:
-	case I915_EXEC_RENDER:
-		ring = &dev_priv->ring[RCS];
-		break;
-	case I915_EXEC_BSD:
-		ring = &dev_priv->ring[VCS];
-		if (ctx_id != DEFAULT_CONTEXT_ID) {
-			DRM_DEBUG("Ring %s doesn't support contexts\n",
-				  ring->name);
-			return -EPERM;
-		}
-		break;
-	case I915_EXEC_BLT:
-		ring = &dev_priv->ring[BCS];
-		if (ctx_id != DEFAULT_CONTEXT_ID) {
-			DRM_DEBUG("Ring %s doesn't support contexts\n",
-				  ring->name);
-			return -EPERM;
-		}
-		break;
-	case I915_EXEC_VEBOX:
-		ring = &dev_priv->ring[VECS];
-		if (ctx_id != DEFAULT_CONTEXT_ID) {
-			DRM_DEBUG("Ring %s doesn't support contexts\n",
-				  ring->name);
-			return -EPERM;
-		}
-		break;
-
-	default:
+	if ((args->flags & I915_EXEC_RING_MASK) > LAST_USER_RING) {
 		DRM_DEBUG("execbuf with unknown ring: %d\n",
 			  (int)(args->flags & I915_EXEC_RING_MASK));
 		return -EINVAL;
 	}
+
+	if ((args->flags & I915_EXEC_RING_MASK) == I915_EXEC_DEFAULT)
+		ring = &dev_priv->ring[RCS];
+	else if ((args->flags & I915_EXEC_RING_MASK) == I915_EXEC_BSD) {
+		if (HAS_BSD2(dev)) {
+			int ring_id;
+			ring_id = gen8_dispatch_bsd_ring(dev, file);
+			ring = &dev_priv->ring[ring_id];
+		} else
+			ring = &dev_priv->ring[VCS];
+	} else
+		ring = &dev_priv->ring[(args->flags & I915_EXEC_RING_MASK) - 1];
+
 	if (!intel_ring_initialized(ring)) {
 		DRM_DEBUG("execbuf with invalid ring: %d\n",
 			  (int)(args->flags & I915_EXEC_RING_MASK));
@@ -1051,14 +1144,22 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	case I915_EXEC_CONSTANTS_REL_GENERAL:
 	case I915_EXEC_CONSTANTS_ABSOLUTE:
 	case I915_EXEC_CONSTANTS_REL_SURFACE:
-		if (ring == &dev_priv->ring[RCS] &&
-		    mode != dev_priv->relative_constants_mode) {
-			if (INTEL_INFO(dev)->gen < 4)
+		if (mode != 0 && ring != &dev_priv->ring[RCS]) {
+			DRM_DEBUG("non-0 rel constants mode on non-RCS\n");
+			return -EINVAL;
+		}
+
+		if (mode != dev_priv->relative_constants_mode) {
+			if (INTEL_INFO(dev)->gen < 4) {
+				DRM_DEBUG("no rel constants on pre-gen4\n");
 				return -EINVAL;
+			}
 
 			if (INTEL_INFO(dev)->gen > 5 &&
-			    mode == I915_EXEC_CONSTANTS_REL_SURFACE)
+			    mode == I915_EXEC_CONSTANTS_REL_SURFACE) {
+				DRM_DEBUG("rel surface constants mode invalid on gen5+\n");
 				return -EINVAL;
+			}
 
 			/* The HW changed the meaning on this bit on gen6 */
 			if (INTEL_INFO(dev)->gen >= 6)
@@ -1106,8 +1207,20 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 			ret = -EFAULT;
 			goto pre_mutex_err;
 		}
+	} else {
+		if (args->DR4 == 0xffffffff) {
+			DRM_DEBUG("UXA submitting garbage DR4, fixing up\n");
+			args->DR4 = 0;
+		}
+
+		if (args->DR1 || args->DR4 || args->cliprects_ptr) {
+			DRM_DEBUG("0 cliprects but dirt in cliprects fields\n");
+			return -EINVAL;
+		}
 	}
 
+	intel_runtime_pm_get(dev_priv);
+
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
 		goto pre_mutex_err;
@@ -1118,8 +1231,22 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 		goto pre_mutex_err;
 	}
 
-	eb = eb_create(args, vm);
+	ctx = i915_gem_validate_context(dev, file, ring, ctx_id);
+	if (IS_ERR(ctx)) {
+		mutex_unlock(&dev->struct_mutex);
+		ret = PTR_ERR(ctx);
+		goto pre_mutex_err;
+	}
+
+	i915_gem_context_reference(ctx);
+
+	vm = ctx->vm;
+	if (!USES_FULL_PPGTT(dev))
+		vm = &dev_priv->gtt.base;
+
+	eb = eb_create(args);
 	if (eb == NULL) {
+		i915_gem_context_unreference(ctx);
 		mutex_unlock(&dev->struct_mutex);
 		ret = -ENOMEM;
 		goto pre_mutex_err;
@@ -1131,7 +1258,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 		goto err;
 
 	/* take note of the batch buffer before we might reorder the lists */
-	batch_obj = list_entry(eb->vmas.prev, struct i915_vma, exec_list)->obj;
+	batch_obj = eb_get_batch(eb);
 
 	/* Move the objects en-masse into the GTT, evicting if necessary. */
 	need_relocs = (args->flags & I915_EXEC_NO_RELOC) == 0;
@@ -1141,7 +1268,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 
 	/* The objects are in their final locations, apply the relocations. */
 	if (need_relocs)
-		ret = i915_gem_execbuffer_relocate(eb, vm);
+		ret = i915_gem_execbuffer_relocate(eb);
 	if (ret) {
 		if (ret == -EFAULT) {
 			ret = i915_gem_execbuffer_relocate_slow(dev, args, file, ring,
@@ -1160,28 +1287,46 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	}
 	batch_obj->base.pending_read_domains |= I915_GEM_DOMAIN_COMMAND;
 
+	if (i915_needs_cmd_parser(ring)) {
+		ret = i915_parse_cmds(ring,
+				      batch_obj,
+				      args->batch_start_offset,
+				      file->is_master);
+		if (ret)
+			goto err;
+
+		/*
+		 * XXX: Actually do this when enabling batch copy...
+		 *
+		 * Set the DISPATCH_SECURE bit to remove the NON_SECURE bit
+		 * from MI_BATCH_BUFFER_START commands issued in the
+		 * dispatch_execbuffer implementations. We specifically don't
+		 * want that set when the command parser is enabled.
+		 */
+	}
+
 	/* snb/ivb/vlv conflate the "batch in ppgtt" bit with the "non-secure
 	 * batch" bit. Hence we need to pin secure batches into the global gtt.
 	 * hsw should have this fixed, but bdw mucks it up again. */
-	if (flags & I915_DISPATCH_SECURE && !batch_obj->has_global_gtt_mapping)
-		i915_gem_gtt_bind_object(batch_obj, batch_obj->cache_level);
+	if (flags & I915_DISPATCH_SECURE &&
+	    !batch_obj->has_global_gtt_mapping) {
+		/* When we have multiple VMs, we'll need to make sure that we
+		 * allocate space first */
+		struct i915_vma *vma = i915_gem_obj_to_ggtt(batch_obj);
+		BUG_ON(!vma);
+		vma->bind_vma(vma, batch_obj->cache_level, GLOBAL_BIND);
+	}
+
+	if (flags & I915_DISPATCH_SECURE)
+		exec_start += i915_gem_obj_ggtt_offset(batch_obj);
+	else
+		exec_start += i915_gem_obj_offset(batch_obj, vm);
 
 	ret = i915_gem_execbuffer_move_to_gpu(ring, &eb->vmas);
 	if (ret)
 		goto err;
 
-	hs = i915_gem_context_get_hang_stats(dev, file, ctx_id);
-	if (IS_ERR(hs)) {
-		ret = PTR_ERR(hs);
-		goto err;
-	}
-
-	if (hs->banned) {
-		ret = -EIO;
-		goto err;
-	}
-
-	ret = i915_switch_context(ring, file, ctx_id);
+	ret = i915_switch_context(ring, ctx);
 	if (ret)
 		goto err;
 
@@ -1206,8 +1351,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 			goto err;
 	}
 
-	exec_start = i915_gem_obj_offset(batch_obj, vm) +
-		args->batch_start_offset;
+
 	exec_len = args->batch_len;
 	if (cliprects) {
 		for (i = 0; i < args->num_cliprects; i++) {
@@ -1236,12 +1380,18 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	i915_gem_execbuffer_retire_commands(dev, file, ring, batch_obj);
 
 err:
+	/* the request owns the ref now */
+	i915_gem_context_unreference(ctx);
 	eb_destroy(eb);
 
 	mutex_unlock(&dev->struct_mutex);
 
 pre_mutex_err:
 	kfree(cliprects);
+
+	/* intel_gpu_busy should also get a ref, so it will free when the device
+	 * is really idle. */
+	intel_runtime_pm_put(dev_priv);
 	return ret;
 }
 
@@ -1253,7 +1403,6 @@ int
 i915_gem_execbuffer(struct drm_device *dev, void *data,
 		    struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_execbuffer *args = data;
 	struct drm_i915_gem_execbuffer2 exec2;
 	struct drm_i915_gem_exec_object *exec_list = NULL;
@@ -1309,21 +1458,23 @@ i915_gem_execbuffer(struct drm_device *dev, void *data,
 	exec2.flags = I915_EXEC_RENDER;
 	i915_execbuffer2_set_context_id(exec2, 0);
 
-	ret = i915_gem_do_execbuffer(dev, data, file, &exec2, exec2_list,
-				     &dev_priv->gtt.base);
+	ret = i915_gem_do_execbuffer(dev, data, file, &exec2, exec2_list);
 	if (!ret) {
+		struct drm_i915_gem_exec_object __user *user_exec_list =
+			to_user_ptr(args->buffers_ptr);
+
 		/* Copy the new buffer offsets back to the user's exec list. */
-		for (i = 0; i < args->buffer_count; i++)
-			exec_list[i].offset = exec2_list[i].offset;
-		/* ... and back out to userspace */
-		ret = copy_to_user(to_user_ptr(args->buffers_ptr),
-				   exec_list,
-				   sizeof(*exec_list) * args->buffer_count);
-		if (ret) {
-			ret = -EFAULT;
-			DRM_DEBUG("failed to copy %d exec entries "
-				  "back to user (%d)\n",
-				  args->buffer_count, ret);
+		for (i = 0; i < args->buffer_count; i++) {
+			ret = __copy_to_user(&user_exec_list[i].offset,
+					     &exec2_list[i].offset,
+					     sizeof(user_exec_list[i].offset));
+			if (ret) {
+				ret = -EFAULT;
+				DRM_DEBUG("failed to copy %d exec entries "
+					  "back to user (%d)\n",
+					  args->buffer_count, ret);
+				break;
+			}
 		}
 	}
 
@@ -1336,7 +1487,6 @@ int
 i915_gem_execbuffer2(struct drm_device *dev, void *data,
 		     struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_execbuffer2 *args = data;
 	struct drm_i915_gem_exec_object2 *exec2_list = NULL;
 	int ret;
@@ -1347,6 +1497,11 @@ i915_gem_execbuffer2(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
+	if (args->rsvd2 != 0) {
+		DRM_DEBUG("dirty rvsd2 field\n");
+		return -EINVAL;
+	}
+
 	exec2_list = kmalloc(sizeof(*exec2_list)*args->buffer_count,
 			     GFP_TEMPORARY | __GFP_NOWARN | __GFP_NORETRY);
 	if (exec2_list == NULL)
@@ -1367,18 +1522,24 @@ i915_gem_execbuffer2(struct drm_device *dev, void *data,
 		return -EFAULT;
 	}
 
-	ret = i915_gem_do_execbuffer(dev, data, file, args, exec2_list,
-				     &dev_priv->gtt.base);
+	ret = i915_gem_do_execbuffer(dev, data, file, args, exec2_list);
 	if (!ret) {
 		/* Copy the new buffer offsets back to the user's exec list. */
-		ret = copy_to_user(to_user_ptr(args->buffers_ptr),
-				   exec2_list,
-				   sizeof(*exec2_list) * args->buffer_count);
-		if (ret) {
-			ret = -EFAULT;
-			DRM_DEBUG("failed to copy %d exec entries "
-				  "back to user (%d)\n",
-				  args->buffer_count, ret);
+		struct drm_i915_gem_exec_object2 *user_exec_list =
+				   to_user_ptr(args->buffers_ptr);
+		int i;
+
+		for (i = 0; i < args->buffer_count; i++) {
+			ret = __copy_to_user(&user_exec_list[i].offset,
+					     &exec2_list[i].offset,
+					     sizeof(user_exec_list[i].offset));
+			if (ret) {
+				ret = -EFAULT;
+				DRM_DEBUG("failed to copy %d exec entries "
+					  "back to user\n",
+					  args->buffer_count);
+				break;
+			}
 		}
 	}
 
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 3540569948d..8b3cde70336 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -1,5 +1,6 @@
 /*
  * Copyright © 2010 Daniel Vetter
+ * Copyright © 2011-2014 Intel Corporation
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -22,53 +23,55 @@
  *
  */
 
+#include <linux/seq_file.h>
 #include <drm/drmP.h>
 #include <drm/i915_drm.h>
 #include "i915_drv.h"
 #include "i915_trace.h"
 #include "intel_drv.h"
 
-#define GEN6_PPGTT_PD_ENTRIES 512
-#define I915_PPGTT_PT_ENTRIES (PAGE_SIZE / sizeof(gen6_gtt_pte_t))
-typedef uint64_t gen8_gtt_pte_t;
-typedef gen8_gtt_pte_t gen8_ppgtt_pde_t;
-
-/* PPGTT stuff */
-#define GEN6_GTT_ADDR_ENCODE(addr)	((addr) | (((addr) >> 28) & 0xff0))
-#define HSW_GTT_ADDR_ENCODE(addr)	((addr) | (((addr) >> 28) & 0x7f0))
-
-#define GEN6_PDE_VALID			(1 << 0)
-/* gen6+ has bit 11-4 for physical addr bit 39-32 */
-#define GEN6_PDE_ADDR_ENCODE(addr)	GEN6_GTT_ADDR_ENCODE(addr)
-
-#define GEN6_PTE_VALID			(1 << 0)
-#define GEN6_PTE_UNCACHED		(1 << 1)
-#define HSW_PTE_UNCACHED		(0)
-#define GEN6_PTE_CACHE_LLC		(2 << 1)
-#define GEN7_PTE_CACHE_L3_LLC		(3 << 1)
-#define GEN6_PTE_ADDR_ENCODE(addr)	GEN6_GTT_ADDR_ENCODE(addr)
-#define HSW_PTE_ADDR_ENCODE(addr)	HSW_GTT_ADDR_ENCODE(addr)
-
-/* Cacheability Control is a 4-bit value. The low three bits are stored in *
- * bits 3:1 of the PTE, while the fourth bit is stored in bit 11 of the PTE.
- */
-#define HSW_CACHEABILITY_CONTROL(bits)	((((bits) & 0x7) << 1) | \
-					 (((bits) & 0x8) << (11 - 3)))
-#define HSW_WB_LLC_AGE3			HSW_CACHEABILITY_CONTROL(0x2)
-#define HSW_WB_LLC_AGE0			HSW_CACHEABILITY_CONTROL(0x3)
-#define HSW_WB_ELLC_LLC_AGE0		HSW_CACHEABILITY_CONTROL(0xb)
-#define HSW_WB_ELLC_LLC_AGE3		HSW_CACHEABILITY_CONTROL(0x8)
-#define HSW_WT_ELLC_LLC_AGE0		HSW_CACHEABILITY_CONTROL(0x6)
-#define HSW_WT_ELLC_LLC_AGE3		HSW_CACHEABILITY_CONTROL(0x7)
-
-#define GEN8_PTES_PER_PAGE		(PAGE_SIZE / sizeof(gen8_gtt_pte_t))
-#define GEN8_PDES_PER_PAGE		(PAGE_SIZE / sizeof(gen8_ppgtt_pde_t))
-#define GEN8_LEGACY_PDPS		4
-
-#define PPAT_UNCACHED_INDEX		(_PAGE_PWT | _PAGE_PCD)
-#define PPAT_CACHED_PDE_INDEX		0 /* WB LLC */
-#define PPAT_CACHED_INDEX		_PAGE_PAT /* WB LLCeLLC */
-#define PPAT_DISPLAY_ELLC_INDEX		_PAGE_PCD /* WT eLLC */
+static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv);
+static void chv_setup_private_ppat(struct drm_i915_private *dev_priv);
+
+bool intel_enable_ppgtt(struct drm_device *dev, bool full)
+{
+	if (i915.enable_ppgtt == 0)
+		return false;
+
+	if (i915.enable_ppgtt == 1 && full)
+		return false;
+
+	return true;
+}
+
+static int sanitize_enable_ppgtt(struct drm_device *dev, int enable_ppgtt)
+{
+	if (enable_ppgtt == 0 || !HAS_ALIASING_PPGTT(dev))
+		return 0;
+
+	if (enable_ppgtt == 1)
+		return 1;
+
+	if (enable_ppgtt == 2 && HAS_PPGTT(dev))
+		return 2;
+
+#ifdef CONFIG_INTEL_IOMMU
+	/* Disable ppgtt on SNB if VT-d is on. */
+	if (INTEL_INFO(dev)->gen == 6 && intel_iommu_gfx_mapped) {
+		DRM_INFO("Disabling PPGTT because VT-d is on\n");
+		return 0;
+	}
+#endif
+
+	return HAS_ALIASING_PPGTT(dev) ? 1 : 0;
+}
+
+
+static void ppgtt_bind_vma(struct i915_vma *vma,
+			   enum i915_cache_level cache_level,
+			   u32 flags);
+static void ppgtt_unbind_vma(struct i915_vma *vma);
+static int gen8_ppgtt_enable(struct i915_hw_ppgtt *ppgtt);
 
 static inline gen8_gtt_pte_t gen8_pte_encode(dma_addr_t addr,
 					     enum i915_cache_level level,
@@ -76,10 +79,19 @@ static inline gen8_gtt_pte_t gen8_pte_encode(dma_addr_t addr,
 {
 	gen8_gtt_pte_t pte = valid ? _PAGE_PRESENT | _PAGE_RW : 0;
 	pte |= addr;
-	if (level != I915_CACHE_NONE)
-		pte |= PPAT_CACHED_INDEX;
-	else
+
+	switch (level) {
+	case I915_CACHE_NONE:
 		pte |= PPAT_UNCACHED_INDEX;
+		break;
+	case I915_CACHE_WT:
+		pte |= PPAT_DISPLAY_ELLC_INDEX;
+		break;
+	default:
+		pte |= PPAT_CACHED_INDEX;
+		break;
+	}
+
 	return pte;
 }
 
@@ -142,9 +154,6 @@ static gen6_gtt_pte_t ivb_pte_encode(dma_addr_t addr,
 	return pte;
 }
 
-#define BYT_PTE_WRITEABLE		(1 << 1)
-#define BYT_PTE_SNOOPED_BY_CPU_CACHES	(1 << 2)
-
 static gen6_gtt_pte_t byt_pte_encode(dma_addr_t addr,
 				     enum i915_cache_level level,
 				     bool valid)
@@ -198,13 +207,20 @@ static gen6_gtt_pte_t iris_pte_encode(dma_addr_t addr,
 }
 
 /* Broadwell Page Directory Pointer Descriptors */
-static int gen8_write_pdp(struct intel_ring_buffer *ring, unsigned entry,
-			   uint64_t val)
+static int gen8_write_pdp(struct intel_engine_cs *ring, unsigned entry,
+			   uint64_t val, bool synchronous)
 {
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	int ret;
 
 	BUG_ON(entry >= 4);
 
+	if (synchronous) {
+		I915_WRITE(GEN8_RING_PDP_UDW(ring, entry), val >> 32);
+		I915_WRITE(GEN8_RING_PDP_LDW(ring, entry), (u32)val);
+		return 0;
+	}
+
 	ret = intel_ring_begin(ring, 6);
 	if (ret)
 		return ret;
@@ -220,206 +236,364 @@ static int gen8_write_pdp(struct intel_ring_buffer *ring, unsigned entry,
 	return 0;
 }
 
-static int gen8_ppgtt_enable(struct drm_device *dev)
+static int gen8_mm_switch(struct i915_hw_ppgtt *ppgtt,
+			  struct intel_engine_cs *ring,
+			  bool synchronous)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
-	struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
-	int i, j, ret;
+	int i, ret;
 
 	/* bit of a hack to find the actual last used pd */
 	int used_pd = ppgtt->num_pd_entries / GEN8_PDES_PER_PAGE;
 
-	for_each_ring(ring, dev_priv, j) {
-		I915_WRITE(RING_MODE_GEN7(ring),
-			   _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
-	}
-
 	for (i = used_pd - 1; i >= 0; i--) {
 		dma_addr_t addr = ppgtt->pd_dma_addr[i];
-		for_each_ring(ring, dev_priv, j) {
-			ret = gen8_write_pdp(ring, i, addr);
-			if (ret)
-				return ret;
-		}
+		ret = gen8_write_pdp(ring, i, addr, synchronous);
+		if (ret)
+			return ret;
 	}
+
 	return 0;
 }
 
 static void gen8_ppgtt_clear_range(struct i915_address_space *vm,
-				   unsigned first_entry,
-				   unsigned num_entries,
+				   uint64_t start,
+				   uint64_t length,
 				   bool use_scratch)
 {
 	struct i915_hw_ppgtt *ppgtt =
 		container_of(vm, struct i915_hw_ppgtt, base);
 	gen8_gtt_pte_t *pt_vaddr, scratch_pte;
-	unsigned act_pt = first_entry / GEN8_PTES_PER_PAGE;
-	unsigned first_pte = first_entry % GEN8_PTES_PER_PAGE;
+	unsigned pdpe = start >> GEN8_PDPE_SHIFT & GEN8_PDPE_MASK;
+	unsigned pde = start >> GEN8_PDE_SHIFT & GEN8_PDE_MASK;
+	unsigned pte = start >> GEN8_PTE_SHIFT & GEN8_PTE_MASK;
+	unsigned num_entries = length >> PAGE_SHIFT;
 	unsigned last_pte, i;
 
 	scratch_pte = gen8_pte_encode(ppgtt->base.scratch.addr,
 				      I915_CACHE_LLC, use_scratch);
 
 	while (num_entries) {
-		struct page *page_table = &ppgtt->gen8_pt_pages[act_pt];
+		struct page *page_table = ppgtt->gen8_pt_pages[pdpe][pde];
 
-		last_pte = first_pte + num_entries;
+		last_pte = pte + num_entries;
 		if (last_pte > GEN8_PTES_PER_PAGE)
 			last_pte = GEN8_PTES_PER_PAGE;
 
 		pt_vaddr = kmap_atomic(page_table);
 
-		for (i = first_pte; i < last_pte; i++)
+		for (i = pte; i < last_pte; i++) {
 			pt_vaddr[i] = scratch_pte;
+			num_entries--;
+		}
 
+		if (!HAS_LLC(ppgtt->base.dev))
+			drm_clflush_virt_range(pt_vaddr, PAGE_SIZE);
 		kunmap_atomic(pt_vaddr);
 
-		num_entries -= last_pte - first_pte;
-		first_pte = 0;
-		act_pt++;
+		pte = 0;
+		if (++pde == GEN8_PDES_PER_PAGE) {
+			pdpe++;
+			pde = 0;
+		}
 	}
 }
 
 static void gen8_ppgtt_insert_entries(struct i915_address_space *vm,
 				      struct sg_table *pages,
-				      unsigned first_entry,
+				      uint64_t start,
 				      enum i915_cache_level cache_level)
 {
 	struct i915_hw_ppgtt *ppgtt =
 		container_of(vm, struct i915_hw_ppgtt, base);
 	gen8_gtt_pte_t *pt_vaddr;
-	unsigned act_pt = first_entry / GEN8_PTES_PER_PAGE;
-	unsigned act_pte = first_entry % GEN8_PTES_PER_PAGE;
+	unsigned pdpe = start >> GEN8_PDPE_SHIFT & GEN8_PDPE_MASK;
+	unsigned pde = start >> GEN8_PDE_SHIFT & GEN8_PDE_MASK;
+	unsigned pte = start >> GEN8_PTE_SHIFT & GEN8_PTE_MASK;
 	struct sg_page_iter sg_iter;
 
-	pt_vaddr = kmap_atomic(&ppgtt->gen8_pt_pages[act_pt]);
-	for_each_sg_page(pages->sgl, &sg_iter, pages->nents, 0) {
-		dma_addr_t page_addr;
+	pt_vaddr = NULL;
 
-		page_addr = sg_dma_address(sg_iter.sg) +
-				(sg_iter.sg_pgoffset << PAGE_SHIFT);
-		pt_vaddr[act_pte] = gen8_pte_encode(page_addr, cache_level,
-						    true);
-		if (++act_pte == GEN8_PTES_PER_PAGE) {
+	for_each_sg_page(pages->sgl, &sg_iter, pages->nents, 0) {
+		if (WARN_ON(pdpe >= GEN8_LEGACY_PDPS))
+			break;
+
+		if (pt_vaddr == NULL)
+			pt_vaddr = kmap_atomic(ppgtt->gen8_pt_pages[pdpe][pde]);
+
+		pt_vaddr[pte] =
+			gen8_pte_encode(sg_page_iter_dma_address(&sg_iter),
+					cache_level, true);
+		if (++pte == GEN8_PTES_PER_PAGE) {
+			if (!HAS_LLC(ppgtt->base.dev))
+				drm_clflush_virt_range(pt_vaddr, PAGE_SIZE);
 			kunmap_atomic(pt_vaddr);
-			act_pt++;
-			pt_vaddr = kmap_atomic(&ppgtt->gen8_pt_pages[act_pt]);
-			act_pte = 0;
+			pt_vaddr = NULL;
+			if (++pde == GEN8_PDES_PER_PAGE) {
+				pdpe++;
+				pde = 0;
+			}
+			pte = 0;
+		}
+	}
+	if (pt_vaddr) {
+		if (!HAS_LLC(ppgtt->base.dev))
+			drm_clflush_virt_range(pt_vaddr, PAGE_SIZE);
+		kunmap_atomic(pt_vaddr);
+	}
+}
 
+static void gen8_free_page_tables(struct page **pt_pages)
+{
+	int i;
+
+	if (pt_pages == NULL)
+		return;
+
+	for (i = 0; i < GEN8_PDES_PER_PAGE; i++)
+		if (pt_pages[i])
+			__free_pages(pt_pages[i], 0);
+}
+
+static void gen8_ppgtt_free(const struct i915_hw_ppgtt *ppgtt)
+{
+	int i;
+
+	for (i = 0; i < ppgtt->num_pd_pages; i++) {
+		gen8_free_page_tables(ppgtt->gen8_pt_pages[i]);
+		kfree(ppgtt->gen8_pt_pages[i]);
+		kfree(ppgtt->gen8_pt_dma_addr[i]);
+	}
+
+	__free_pages(ppgtt->pd_pages, get_order(ppgtt->num_pd_pages << PAGE_SHIFT));
+}
+
+static void gen8_ppgtt_unmap_pages(struct i915_hw_ppgtt *ppgtt)
+{
+	struct pci_dev *hwdev = ppgtt->base.dev->pdev;
+	int i, j;
+
+	for (i = 0; i < ppgtt->num_pd_pages; i++) {
+		/* TODO: In the future we'll support sparse mappings, so this
+		 * will have to change. */
+		if (!ppgtt->pd_dma_addr[i])
+			continue;
+
+		pci_unmap_page(hwdev, ppgtt->pd_dma_addr[i], PAGE_SIZE,
+			       PCI_DMA_BIDIRECTIONAL);
+
+		for (j = 0; j < GEN8_PDES_PER_PAGE; j++) {
+			dma_addr_t addr = ppgtt->gen8_pt_dma_addr[i][j];
+			if (addr)
+				pci_unmap_page(hwdev, addr, PAGE_SIZE,
+					       PCI_DMA_BIDIRECTIONAL);
 		}
 	}
-	kunmap_atomic(pt_vaddr);
 }
 
 static void gen8_ppgtt_cleanup(struct i915_address_space *vm)
 {
 	struct i915_hw_ppgtt *ppgtt =
 		container_of(vm, struct i915_hw_ppgtt, base);
-	int i, j;
 
-	for (i = 0; i < ppgtt->num_pd_pages ; i++) {
-		if (ppgtt->pd_dma_addr[i]) {
-			pci_unmap_page(ppgtt->base.dev->pdev,
-				       ppgtt->pd_dma_addr[i],
-				       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
+	list_del(&vm->global_link);
+	drm_mm_takedown(&vm->mm);
 
-			for (j = 0; j < GEN8_PDES_PER_PAGE; j++) {
-				dma_addr_t addr = ppgtt->gen8_pt_dma_addr[i][j];
-				if (addr)
-					pci_unmap_page(ppgtt->base.dev->pdev,
-						       addr,
-						       PAGE_SIZE,
-						       PCI_DMA_BIDIRECTIONAL);
+	gen8_ppgtt_unmap_pages(ppgtt);
+	gen8_ppgtt_free(ppgtt);
+}
 
-			}
-		}
-		kfree(ppgtt->gen8_pt_dma_addr[i]);
+static struct page **__gen8_alloc_page_tables(void)
+{
+	struct page **pt_pages;
+	int i;
+
+	pt_pages = kcalloc(GEN8_PDES_PER_PAGE, sizeof(struct page *), GFP_KERNEL);
+	if (!pt_pages)
+		return ERR_PTR(-ENOMEM);
+
+	for (i = 0; i < GEN8_PDES_PER_PAGE; i++) {
+		pt_pages[i] = alloc_page(GFP_KERNEL);
+		if (!pt_pages[i])
+			goto bail;
 	}
 
-	__free_pages(ppgtt->gen8_pt_pages, get_order(ppgtt->num_pt_pages << PAGE_SHIFT));
-	__free_pages(ppgtt->pd_pages, get_order(ppgtt->num_pd_pages << PAGE_SHIFT));
+	return pt_pages;
+
+bail:
+	gen8_free_page_tables(pt_pages);
+	kfree(pt_pages);
+	return ERR_PTR(-ENOMEM);
 }
 
-/**
- * GEN8 legacy ppgtt programming is accomplished through 4 PDP registers with a
- * net effect resembling a 2-level page table in normal x86 terms. Each PDP
- * represents 1GB of memory
- * 4 * 512 * 512 * 4096 = 4GB legacy 32b address space.
- *
- * TODO: Do something with the size parameter
- **/
-static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size)
+static int gen8_ppgtt_allocate_page_tables(struct i915_hw_ppgtt *ppgtt,
+					   const int max_pdp)
 {
-	struct page *pt_pages;
-	int i, j, ret = -ENOMEM;
-	const int max_pdp = DIV_ROUND_UP(size, 1 << 30);
-	const int num_pt_pages = GEN8_PDES_PER_PAGE * max_pdp;
+	struct page **pt_pages[GEN8_LEGACY_PDPS];
+	int i, ret;
 
-	if (size % (1<<30))
-		DRM_INFO("Pages will be wasted unless GTT size (%llu) is divisible by 1GB\n", size);
+	for (i = 0; i < max_pdp; i++) {
+		pt_pages[i] = __gen8_alloc_page_tables();
+		if (IS_ERR(pt_pages[i])) {
+			ret = PTR_ERR(pt_pages[i]);
+			goto unwind_out;
+		}
+	}
 
-	/* FIXME: split allocation into smaller pieces. For now we only ever do
-	 * this once, but with full PPGTT, the multiple contiguous allocations
-	 * will be bad.
+	/* NB: Avoid touching gen8_pt_pages until last to keep the allocation,
+	 * "atomic" - for cleanup purposes.
 	 */
+	for (i = 0; i < max_pdp; i++)
+		ppgtt->gen8_pt_pages[i] = pt_pages[i];
+
+	return 0;
+
+unwind_out:
+	while (i--) {
+		gen8_free_page_tables(pt_pages[i]);
+		kfree(pt_pages[i]);
+	}
+
+	return ret;
+}
+
+static int gen8_ppgtt_allocate_dma(struct i915_hw_ppgtt *ppgtt)
+{
+	int i;
+
+	for (i = 0; i < ppgtt->num_pd_pages; i++) {
+		ppgtt->gen8_pt_dma_addr[i] = kcalloc(GEN8_PDES_PER_PAGE,
+						     sizeof(dma_addr_t),
+						     GFP_KERNEL);
+		if (!ppgtt->gen8_pt_dma_addr[i])
+			return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static int gen8_ppgtt_allocate_page_directories(struct i915_hw_ppgtt *ppgtt,
+						const int max_pdp)
+{
 	ppgtt->pd_pages = alloc_pages(GFP_KERNEL, get_order(max_pdp << PAGE_SHIFT));
 	if (!ppgtt->pd_pages)
 		return -ENOMEM;
 
-	pt_pages = alloc_pages(GFP_KERNEL, get_order(num_pt_pages << PAGE_SHIFT));
-	if (!pt_pages) {
+	ppgtt->num_pd_pages = 1 << get_order(max_pdp << PAGE_SHIFT);
+	BUG_ON(ppgtt->num_pd_pages > GEN8_LEGACY_PDPS);
+
+	return 0;
+}
+
+static int gen8_ppgtt_alloc(struct i915_hw_ppgtt *ppgtt,
+			    const int max_pdp)
+{
+	int ret;
+
+	ret = gen8_ppgtt_allocate_page_directories(ppgtt, max_pdp);
+	if (ret)
+		return ret;
+
+	ret = gen8_ppgtt_allocate_page_tables(ppgtt, max_pdp);
+	if (ret) {
 		__free_pages(ppgtt->pd_pages, get_order(max_pdp << PAGE_SHIFT));
-		return -ENOMEM;
+		return ret;
 	}
 
-	ppgtt->gen8_pt_pages = pt_pages;
-	ppgtt->num_pd_pages = 1 << get_order(max_pdp << PAGE_SHIFT);
-	ppgtt->num_pt_pages = 1 << get_order(num_pt_pages << PAGE_SHIFT);
 	ppgtt->num_pd_entries = max_pdp * GEN8_PDES_PER_PAGE;
-	ppgtt->enable = gen8_ppgtt_enable;
-	ppgtt->base.clear_range = gen8_ppgtt_clear_range;
-	ppgtt->base.insert_entries = gen8_ppgtt_insert_entries;
-	ppgtt->base.cleanup = gen8_ppgtt_cleanup;
 
-	BUG_ON(ppgtt->num_pd_pages > GEN8_LEGACY_PDPS);
+	ret = gen8_ppgtt_allocate_dma(ppgtt);
+	if (ret)
+		gen8_ppgtt_free(ppgtt);
 
-	/*
-	 * - Create a mapping for the page directories.
-	 * - For each page directory:
-	 *      allocate space for page table mappings.
-	 *      map each page table
-	 */
-	for (i = 0; i < max_pdp; i++) {
-		dma_addr_t temp;
-		temp = pci_map_page(ppgtt->base.dev->pdev,
-				    &ppgtt->pd_pages[i], 0,
-				    PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
-		if (pci_dma_mapping_error(ppgtt->base.dev->pdev, temp))
-			goto err_out;
+	return ret;
+}
 
-		ppgtt->pd_dma_addr[i] = temp;
+static int gen8_ppgtt_setup_page_directories(struct i915_hw_ppgtt *ppgtt,
+					     const int pd)
+{
+	dma_addr_t pd_addr;
+	int ret;
 
-		ppgtt->gen8_pt_dma_addr[i] = kmalloc(sizeof(dma_addr_t) * GEN8_PDES_PER_PAGE, GFP_KERNEL);
-		if (!ppgtt->gen8_pt_dma_addr[i])
-			goto err_out;
+	pd_addr = pci_map_page(ppgtt->base.dev->pdev,
+			       &ppgtt->pd_pages[pd], 0,
+			       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
 
-		for (j = 0; j < GEN8_PDES_PER_PAGE; j++) {
-			struct page *p = &pt_pages[i * GEN8_PDES_PER_PAGE + j];
-			temp = pci_map_page(ppgtt->base.dev->pdev,
-					    p, 0, PAGE_SIZE,
-					    PCI_DMA_BIDIRECTIONAL);
+	ret = pci_dma_mapping_error(ppgtt->base.dev->pdev, pd_addr);
+	if (ret)
+		return ret;
+
+	ppgtt->pd_dma_addr[pd] = pd_addr;
+
+	return 0;
+}
+
+static int gen8_ppgtt_setup_page_tables(struct i915_hw_ppgtt *ppgtt,
+					const int pd,
+					const int pt)
+{
+	dma_addr_t pt_addr;
+	struct page *p;
+	int ret;
 
-			if (pci_dma_mapping_error(ppgtt->base.dev->pdev, temp))
-				goto err_out;
+	p = ppgtt->gen8_pt_pages[pd][pt];
+	pt_addr = pci_map_page(ppgtt->base.dev->pdev,
+			       p, 0, PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
+	ret = pci_dma_mapping_error(ppgtt->base.dev->pdev, pt_addr);
+	if (ret)
+		return ret;
+
+	ppgtt->gen8_pt_dma_addr[pd][pt] = pt_addr;
+
+	return 0;
+}
+
+/**
+ * GEN8 legacy ppgtt programming is accomplished through a max 4 PDP registers
+ * with a net effect resembling a 2-level page table in normal x86 terms. Each
+ * PDP represents 1GB of memory 4 * 512 * 512 * 4096 = 4GB legacy 32b address
+ * space.
+ *
+ * FIXME: split allocation into smaller pieces. For now we only ever do this
+ * once, but with full PPGTT, the multiple contiguous allocations will be bad.
+ * TODO: Do something with the size parameter
+ */
+static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size)
+{
+	const int max_pdp = DIV_ROUND_UP(size, 1 << 30);
+	const int min_pt_pages = GEN8_PDES_PER_PAGE * max_pdp;
+	int i, j, ret;
+
+	if (size % (1<<30))
+		DRM_INFO("Pages will be wasted unless GTT size (%llu) is divisible by 1GB\n", size);
 
-			ppgtt->gen8_pt_dma_addr[i][j] = temp;
+	/* 1. Do all our allocations for page directories and page tables. */
+	ret = gen8_ppgtt_alloc(ppgtt, max_pdp);
+	if (ret)
+		return ret;
+
+	/*
+	 * 2. Create DMA mappings for the page directories and page tables.
+	 */
+	for (i = 0; i < max_pdp; i++) {
+		ret = gen8_ppgtt_setup_page_directories(ppgtt, i);
+		if (ret)
+			goto bail;
+
+		for (j = 0; j < GEN8_PDES_PER_PAGE; j++) {
+			ret = gen8_ppgtt_setup_page_tables(ppgtt, i, j);
+			if (ret)
+				goto bail;
 		}
 	}
 
-	/* For now, the PPGTT helper functions all require that the PDEs are
+	/*
+	 * 3. Map all the page directory entires to point to the page tables
+	 * we've allocated.
+	 *
+	 * For now, the PPGTT helper functions all require that the PDEs are
 	 * plugged in correctly. So we do that now/here. For aliasing PPGTT, we
-	 * will never need to touch the PDEs again */
+	 * will never need to touch the PDEs again.
+	 */
 	for (i = 0; i < max_pdp; i++) {
 		gen8_ppgtt_pde_t *pd_vaddr;
 		pd_vaddr = kmap_atomic(&ppgtt->pd_pages[i]);
@@ -428,26 +602,90 @@ static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size)
 			pd_vaddr[j] = gen8_pde_encode(ppgtt->base.dev, addr,
 						      I915_CACHE_LLC);
 		}
+		if (!HAS_LLC(ppgtt->base.dev))
+			drm_clflush_virt_range(pd_vaddr, PAGE_SIZE);
 		kunmap_atomic(pd_vaddr);
 	}
 
-	ppgtt->base.clear_range(&ppgtt->base, 0,
-				ppgtt->num_pd_entries * GEN8_PTES_PER_PAGE,
-				true);
+	ppgtt->enable = gen8_ppgtt_enable;
+	ppgtt->switch_mm = gen8_mm_switch;
+	ppgtt->base.clear_range = gen8_ppgtt_clear_range;
+	ppgtt->base.insert_entries = gen8_ppgtt_insert_entries;
+	ppgtt->base.cleanup = gen8_ppgtt_cleanup;
+	ppgtt->base.start = 0;
+	ppgtt->base.total = ppgtt->num_pd_entries * GEN8_PTES_PER_PAGE * PAGE_SIZE;
+
+	ppgtt->base.clear_range(&ppgtt->base, 0, ppgtt->base.total, true);
 
 	DRM_DEBUG_DRIVER("Allocated %d pages for page directories (%d wasted)\n",
 			 ppgtt->num_pd_pages, ppgtt->num_pd_pages - max_pdp);
 	DRM_DEBUG_DRIVER("Allocated %d pages for page tables (%lld wasted)\n",
-			 ppgtt->num_pt_pages,
-			 (ppgtt->num_pt_pages - num_pt_pages) +
-			 size % (1<<30));
+			 ppgtt->num_pd_entries,
+			 (ppgtt->num_pd_entries - min_pt_pages) + size % (1<<30));
 	return 0;
 
-err_out:
-	ppgtt->base.cleanup(&ppgtt->base);
+bail:
+	gen8_ppgtt_unmap_pages(ppgtt);
+	gen8_ppgtt_free(ppgtt);
 	return ret;
 }
 
+static void gen6_dump_ppgtt(struct i915_hw_ppgtt *ppgtt, struct seq_file *m)
+{
+	struct drm_i915_private *dev_priv = ppgtt->base.dev->dev_private;
+	struct i915_address_space *vm = &ppgtt->base;
+	gen6_gtt_pte_t __iomem *pd_addr;
+	gen6_gtt_pte_t scratch_pte;
+	uint32_t pd_entry;
+	int pte, pde;
+
+	scratch_pte = vm->pte_encode(vm->scratch.addr, I915_CACHE_LLC, true);
+
+	pd_addr = (gen6_gtt_pte_t __iomem *)dev_priv->gtt.gsm +
+		ppgtt->pd_offset / sizeof(gen6_gtt_pte_t);
+
+	seq_printf(m, "  VM %p (pd_offset %x-%x):\n", vm,
+		   ppgtt->pd_offset, ppgtt->pd_offset + ppgtt->num_pd_entries);
+	for (pde = 0; pde < ppgtt->num_pd_entries; pde++) {
+		u32 expected;
+		gen6_gtt_pte_t *pt_vaddr;
+		dma_addr_t pt_addr = ppgtt->pt_dma_addr[pde];
+		pd_entry = readl(pd_addr + pde);
+		expected = (GEN6_PDE_ADDR_ENCODE(pt_addr) | GEN6_PDE_VALID);
+
+		if (pd_entry != expected)
+			seq_printf(m, "\tPDE #%d mismatch: Actual PDE: %x Expected PDE: %x\n",
+				   pde,
+				   pd_entry,
+				   expected);
+		seq_printf(m, "\tPDE: %x\n", pd_entry);
+
+		pt_vaddr = kmap_atomic(ppgtt->pt_pages[pde]);
+		for (pte = 0; pte < I915_PPGTT_PT_ENTRIES; pte+=4) {
+			unsigned long va =
+				(pde * PAGE_SIZE * I915_PPGTT_PT_ENTRIES) +
+				(pte * PAGE_SIZE);
+			int i;
+			bool found = false;
+			for (i = 0; i < 4; i++)
+				if (pt_vaddr[pte + i] != scratch_pte)
+					found = true;
+			if (!found)
+				continue;
+
+			seq_printf(m, "\t\t0x%lx [%03d,%04d]: =", va, pde, pte);
+			for (i = 0; i < 4; i++) {
+				if (pt_vaddr[pte + i] != scratch_pte)
+					seq_printf(m, " %08x", pt_vaddr[pte + i]);
+				else
+					seq_puts(m, "  SCRATCH ");
+			}
+			seq_puts(m, "\n");
+		}
+		kunmap_atomic(pt_vaddr);
+	}
+}
+
 static void gen6_write_pdes(struct i915_hw_ppgtt *ppgtt)
 {
 	struct drm_i915_private *dev_priv = ppgtt->base.dev->dev_private;
@@ -470,73 +708,235 @@ static void gen6_write_pdes(struct i915_hw_ppgtt *ppgtt)
 	readl(pd_addr);
 }
 
-static int gen6_ppgtt_enable(struct drm_device *dev)
+static uint32_t get_pd_offset(struct i915_hw_ppgtt *ppgtt)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	uint32_t pd_offset;
-	struct intel_ring_buffer *ring;
-	struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
-	int i;
-
 	BUG_ON(ppgtt->pd_offset & 0x3f);
 
-	gen6_write_pdes(ppgtt);
+	return (ppgtt->pd_offset / 64) << 16;
+}
+
+static int hsw_mm_switch(struct i915_hw_ppgtt *ppgtt,
+			 struct intel_engine_cs *ring,
+			 bool synchronous)
+{
+	struct drm_device *dev = ppgtt->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	/* If we're in reset, we can assume the GPU is sufficiently idle to
+	 * manually frob these bits. Ideally we could use the ring functions,
+	 * except our error handling makes it quite difficult (can't use
+	 * intel_ring_begin, ring->flush, or intel_ring_advance)
+	 *
+	 * FIXME: We should try not to special case reset
+	 */
+	if (synchronous ||
+	    i915_reset_in_progress(&dev_priv->gpu_error)) {
+		WARN_ON(ppgtt != dev_priv->mm.aliasing_ppgtt);
+		I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
+		I915_WRITE(RING_PP_DIR_BASE(ring), get_pd_offset(ppgtt));
+		POSTING_READ(RING_PP_DIR_BASE(ring));
+		return 0;
+	}
+
+	/* NB: TLBs must be flushed and invalidated before a switch */
+	ret = ring->flush(ring, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
+	if (ret)
+		return ret;
+
+	ret = intel_ring_begin(ring, 6);
+	if (ret)
+		return ret;
 
-	pd_offset = ppgtt->pd_offset;
-	pd_offset /= 64; /* in cachelines, */
-	pd_offset <<= 16;
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(2));
+	intel_ring_emit(ring, RING_PP_DIR_DCLV(ring));
+	intel_ring_emit(ring, PP_DIR_DCLV_2G);
+	intel_ring_emit(ring, RING_PP_DIR_BASE(ring));
+	intel_ring_emit(ring, get_pd_offset(ppgtt));
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
 
-	if (INTEL_INFO(dev)->gen == 6) {
-		uint32_t ecochk, gab_ctl, ecobits;
+	return 0;
+}
 
-		ecobits = I915_READ(GAC_ECO_BITS);
-		I915_WRITE(GAC_ECO_BITS, ecobits | ECOBITS_SNB_BIT |
-					 ECOBITS_PPGTT_CACHE64B);
+static int gen7_mm_switch(struct i915_hw_ppgtt *ppgtt,
+			  struct intel_engine_cs *ring,
+			  bool synchronous)
+{
+	struct drm_device *dev = ppgtt->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
 
-		gab_ctl = I915_READ(GAB_CTL);
-		I915_WRITE(GAB_CTL, gab_ctl | GAB_CTL_CONT_AFTER_PAGEFAULT);
+	/* If we're in reset, we can assume the GPU is sufficiently idle to
+	 * manually frob these bits. Ideally we could use the ring functions,
+	 * except our error handling makes it quite difficult (can't use
+	 * intel_ring_begin, ring->flush, or intel_ring_advance)
+	 *
+	 * FIXME: We should try not to special case reset
+	 */
+	if (synchronous ||
+	    i915_reset_in_progress(&dev_priv->gpu_error)) {
+		WARN_ON(ppgtt != dev_priv->mm.aliasing_ppgtt);
+		I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
+		I915_WRITE(RING_PP_DIR_BASE(ring), get_pd_offset(ppgtt));
+		POSTING_READ(RING_PP_DIR_BASE(ring));
+		return 0;
+	}
 
-		ecochk = I915_READ(GAM_ECOCHK);
-		I915_WRITE(GAM_ECOCHK, ecochk | ECOCHK_SNB_BIT |
-				       ECOCHK_PPGTT_CACHE64B);
-		I915_WRITE(GFX_MODE, _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
-	} else if (INTEL_INFO(dev)->gen >= 7) {
-		uint32_t ecochk, ecobits;
+	/* NB: TLBs must be flushed and invalidated before a switch */
+	ret = ring->flush(ring, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
+	if (ret)
+		return ret;
 
-		ecobits = I915_READ(GAC_ECO_BITS);
-		I915_WRITE(GAC_ECO_BITS, ecobits | ECOBITS_PPGTT_CACHE64B);
+	ret = intel_ring_begin(ring, 6);
+	if (ret)
+		return ret;
 
-		ecochk = I915_READ(GAM_ECOCHK);
-		if (IS_HASWELL(dev)) {
-			ecochk |= ECOCHK_PPGTT_WB_HSW;
-		} else {
-			ecochk |= ECOCHK_PPGTT_LLC_IVB;
-			ecochk &= ~ECOCHK_PPGTT_GFDT_IVB;
-		}
-		I915_WRITE(GAM_ECOCHK, ecochk);
-		/* GFX_MODE is per-ring on gen7+ */
+	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(2));
+	intel_ring_emit(ring, RING_PP_DIR_DCLV(ring));
+	intel_ring_emit(ring, PP_DIR_DCLV_2G);
+	intel_ring_emit(ring, RING_PP_DIR_BASE(ring));
+	intel_ring_emit(ring, get_pd_offset(ppgtt));
+	intel_ring_emit(ring, MI_NOOP);
+	intel_ring_advance(ring);
+
+	/* XXX: RCS is the only one to auto invalidate the TLBs? */
+	if (ring->id != RCS) {
+		ret = ring->flush(ring, I915_GEM_GPU_DOMAINS, I915_GEM_GPU_DOMAINS);
+		if (ret)
+			return ret;
 	}
 
+	return 0;
+}
+
+static int gen6_mm_switch(struct i915_hw_ppgtt *ppgtt,
+			  struct intel_engine_cs *ring,
+			  bool synchronous)
+{
+	struct drm_device *dev = ppgtt->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (!synchronous)
+		return 0;
+
+	I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
+	I915_WRITE(RING_PP_DIR_BASE(ring), get_pd_offset(ppgtt));
+
+	POSTING_READ(RING_PP_DIR_DCLV(ring));
+
+	return 0;
+}
+
+static int gen8_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
+{
+	struct drm_device *dev = ppgtt->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	int j, ret;
+
+	for_each_ring(ring, dev_priv, j) {
+		I915_WRITE(RING_MODE_GEN7(ring),
+			   _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
+
+		/* We promise to do a switch later with FULL PPGTT. If this is
+		 * aliasing, this is the one and only switch we'll do */
+		if (USES_FULL_PPGTT(dev))
+			continue;
+
+		ret = ppgtt->switch_mm(ppgtt, ring, true);
+		if (ret)
+			goto err_out;
+	}
+
+	return 0;
+
+err_out:
+	for_each_ring(ring, dev_priv, j)
+		I915_WRITE(RING_MODE_GEN7(ring),
+			   _MASKED_BIT_DISABLE(GFX_PPGTT_ENABLE));
+	return ret;
+}
+
+static int gen7_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
+{
+	struct drm_device *dev = ppgtt->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	uint32_t ecochk, ecobits;
+	int i;
+
+	ecobits = I915_READ(GAC_ECO_BITS);
+	I915_WRITE(GAC_ECO_BITS, ecobits | ECOBITS_PPGTT_CACHE64B);
+
+	ecochk = I915_READ(GAM_ECOCHK);
+	if (IS_HASWELL(dev)) {
+		ecochk |= ECOCHK_PPGTT_WB_HSW;
+	} else {
+		ecochk |= ECOCHK_PPGTT_LLC_IVB;
+		ecochk &= ~ECOCHK_PPGTT_GFDT_IVB;
+	}
+	I915_WRITE(GAM_ECOCHK, ecochk);
+
 	for_each_ring(ring, dev_priv, i) {
-		if (INTEL_INFO(dev)->gen >= 7)
-			I915_WRITE(RING_MODE_GEN7(ring),
-				   _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
+		int ret;
+		/* GFX_MODE is per-ring on gen7+ */
+		I915_WRITE(RING_MODE_GEN7(ring),
+			   _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
 
-		I915_WRITE(RING_PP_DIR_DCLV(ring), PP_DIR_DCLV_2G);
-		I915_WRITE(RING_PP_DIR_BASE(ring), pd_offset);
+		/* We promise to do a switch later with FULL PPGTT. If this is
+		 * aliasing, this is the one and only switch we'll do */
+		if (USES_FULL_PPGTT(dev))
+			continue;
+
+		ret = ppgtt->switch_mm(ppgtt, ring, true);
+		if (ret)
+			return ret;
 	}
+
+	return 0;
+}
+
+static int gen6_ppgtt_enable(struct i915_hw_ppgtt *ppgtt)
+{
+	struct drm_device *dev = ppgtt->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
+	uint32_t ecochk, gab_ctl, ecobits;
+	int i;
+
+	ecobits = I915_READ(GAC_ECO_BITS);
+	I915_WRITE(GAC_ECO_BITS, ecobits | ECOBITS_SNB_BIT |
+		   ECOBITS_PPGTT_CACHE64B);
+
+	gab_ctl = I915_READ(GAB_CTL);
+	I915_WRITE(GAB_CTL, gab_ctl | GAB_CTL_CONT_AFTER_PAGEFAULT);
+
+	ecochk = I915_READ(GAM_ECOCHK);
+	I915_WRITE(GAM_ECOCHK, ecochk | ECOCHK_SNB_BIT | ECOCHK_PPGTT_CACHE64B);
+
+	I915_WRITE(GFX_MODE, _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
+
+	for_each_ring(ring, dev_priv, i) {
+		int ret = ppgtt->switch_mm(ppgtt, ring, true);
+		if (ret)
+			return ret;
+	}
+
 	return 0;
 }
 
 /* PPGTT support for Sandybdrige/Gen6 and later */
 static void gen6_ppgtt_clear_range(struct i915_address_space *vm,
-				   unsigned first_entry,
-				   unsigned num_entries,
+				   uint64_t start,
+				   uint64_t length,
 				   bool use_scratch)
 {
 	struct i915_hw_ppgtt *ppgtt =
 		container_of(vm, struct i915_hw_ppgtt, base);
 	gen6_gtt_pte_t *pt_vaddr, scratch_pte;
+	unsigned first_entry = start >> PAGE_SHIFT;
+	unsigned num_entries = length >> PAGE_SHIFT;
 	unsigned act_pt = first_entry / I915_PPGTT_PT_ENTRIES;
 	unsigned first_pte = first_entry % I915_PPGTT_PT_ENTRIES;
 	unsigned last_pte, i;
@@ -563,90 +963,159 @@ static void gen6_ppgtt_clear_range(struct i915_address_space *vm,
 
 static void gen6_ppgtt_insert_entries(struct i915_address_space *vm,
 				      struct sg_table *pages,
-				      unsigned first_entry,
+				      uint64_t start,
 				      enum i915_cache_level cache_level)
 {
 	struct i915_hw_ppgtt *ppgtt =
 		container_of(vm, struct i915_hw_ppgtt, base);
 	gen6_gtt_pte_t *pt_vaddr;
+	unsigned first_entry = start >> PAGE_SHIFT;
 	unsigned act_pt = first_entry / I915_PPGTT_PT_ENTRIES;
 	unsigned act_pte = first_entry % I915_PPGTT_PT_ENTRIES;
 	struct sg_page_iter sg_iter;
 
-	pt_vaddr = kmap_atomic(ppgtt->pt_pages[act_pt]);
+	pt_vaddr = NULL;
 	for_each_sg_page(pages->sgl, &sg_iter, pages->nents, 0) {
-		dma_addr_t page_addr;
+		if (pt_vaddr == NULL)
+			pt_vaddr = kmap_atomic(ppgtt->pt_pages[act_pt]);
 
-		page_addr = sg_page_iter_dma_address(&sg_iter);
-		pt_vaddr[act_pte] = vm->pte_encode(page_addr, cache_level, true);
+		pt_vaddr[act_pte] =
+			vm->pte_encode(sg_page_iter_dma_address(&sg_iter),
+				       cache_level, true);
 		if (++act_pte == I915_PPGTT_PT_ENTRIES) {
 			kunmap_atomic(pt_vaddr);
+			pt_vaddr = NULL;
 			act_pt++;
-			pt_vaddr = kmap_atomic(ppgtt->pt_pages[act_pt]);
 			act_pte = 0;
-
 		}
 	}
-	kunmap_atomic(pt_vaddr);
+	if (pt_vaddr)
+		kunmap_atomic(pt_vaddr);
 }
 
-static void gen6_ppgtt_cleanup(struct i915_address_space *vm)
+static void gen6_ppgtt_unmap_pages(struct i915_hw_ppgtt *ppgtt)
 {
-	struct i915_hw_ppgtt *ppgtt =
-		container_of(vm, struct i915_hw_ppgtt, base);
 	int i;
 
-	drm_mm_takedown(&ppgtt->base.mm);
-
 	if (ppgtt->pt_dma_addr) {
 		for (i = 0; i < ppgtt->num_pd_entries; i++)
 			pci_unmap_page(ppgtt->base.dev->pdev,
 				       ppgtt->pt_dma_addr[i],
 				       4096, PCI_DMA_BIDIRECTIONAL);
 	}
+}
+
+static void gen6_ppgtt_free(struct i915_hw_ppgtt *ppgtt)
+{
+	int i;
 
 	kfree(ppgtt->pt_dma_addr);
 	for (i = 0; i < ppgtt->num_pd_entries; i++)
 		__free_page(ppgtt->pt_pages[i]);
 	kfree(ppgtt->pt_pages);
-	kfree(ppgtt);
 }
 
-static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt)
+static void gen6_ppgtt_cleanup(struct i915_address_space *vm)
+{
+	struct i915_hw_ppgtt *ppgtt =
+		container_of(vm, struct i915_hw_ppgtt, base);
+
+	list_del(&vm->global_link);
+	drm_mm_takedown(&ppgtt->base.mm);
+	drm_mm_remove_node(&ppgtt->node);
+
+	gen6_ppgtt_unmap_pages(ppgtt);
+	gen6_ppgtt_free(ppgtt);
+}
+
+static int gen6_ppgtt_allocate_page_directories(struct i915_hw_ppgtt *ppgtt)
 {
 	struct drm_device *dev = ppgtt->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned first_pd_entry_in_global_pt;
-	int i;
-	int ret = -ENOMEM;
+	bool retried = false;
+	int ret;
 
-	/* ppgtt PDEs reside in the global gtt pagetable, which has 512*1024
-	 * entries. For aliasing ppgtt support we just steal them at the end for
-	 * now. */
-	first_pd_entry_in_global_pt = gtt_total_entries(dev_priv->gtt);
+	/* PPGTT PDEs reside in the GGTT and consists of 512 entries. The
+	 * allocator works in address space sizes, so it's multiplied by page
+	 * size. We allocate at the top of the GTT to avoid fragmentation.
+	 */
+	BUG_ON(!drm_mm_initialized(&dev_priv->gtt.base.mm));
+alloc:
+	ret = drm_mm_insert_node_in_range_generic(&dev_priv->gtt.base.mm,
+						  &ppgtt->node, GEN6_PD_SIZE,
+						  GEN6_PD_ALIGN, 0,
+						  0, dev_priv->gtt.base.total,
+						  DRM_MM_TOPDOWN);
+	if (ret == -ENOSPC && !retried) {
+		ret = i915_gem_evict_something(dev, &dev_priv->gtt.base,
+					       GEN6_PD_SIZE, GEN6_PD_ALIGN,
+					       I915_CACHE_NONE,
+					       0, dev_priv->gtt.base.total,
+					       0);
+		if (ret)
+			return ret;
+
+		retried = true;
+		goto alloc;
+	}
+
+	if (ppgtt->node.start < dev_priv->gtt.mappable_end)
+		DRM_DEBUG("Forced to use aperture for PDEs\n");
 
-	ppgtt->base.pte_encode = dev_priv->gtt.base.pte_encode;
 	ppgtt->num_pd_entries = GEN6_PPGTT_PD_ENTRIES;
-	ppgtt->enable = gen6_ppgtt_enable;
-	ppgtt->base.clear_range = gen6_ppgtt_clear_range;
-	ppgtt->base.insert_entries = gen6_ppgtt_insert_entries;
-	ppgtt->base.cleanup = gen6_ppgtt_cleanup;
-	ppgtt->base.scratch = dev_priv->gtt.base.scratch;
+	return ret;
+}
+
+static int gen6_ppgtt_allocate_page_tables(struct i915_hw_ppgtt *ppgtt)
+{
+	int i;
+
 	ppgtt->pt_pages = kcalloc(ppgtt->num_pd_entries, sizeof(struct page *),
 				  GFP_KERNEL);
+
 	if (!ppgtt->pt_pages)
 		return -ENOMEM;
 
 	for (i = 0; i < ppgtt->num_pd_entries; i++) {
 		ppgtt->pt_pages[i] = alloc_page(GFP_KERNEL);
-		if (!ppgtt->pt_pages[i])
-			goto err_pt_alloc;
+		if (!ppgtt->pt_pages[i]) {
+			gen6_ppgtt_free(ppgtt);
+			return -ENOMEM;
+		}
+	}
+
+	return 0;
+}
+
+static int gen6_ppgtt_alloc(struct i915_hw_ppgtt *ppgtt)
+{
+	int ret;
+
+	ret = gen6_ppgtt_allocate_page_directories(ppgtt);
+	if (ret)
+		return ret;
+
+	ret = gen6_ppgtt_allocate_page_tables(ppgtt);
+	if (ret) {
+		drm_mm_remove_node(&ppgtt->node);
+		return ret;
 	}
 
 	ppgtt->pt_dma_addr = kcalloc(ppgtt->num_pd_entries, sizeof(dma_addr_t),
 				     GFP_KERNEL);
-	if (!ppgtt->pt_dma_addr)
-		goto err_pt_alloc;
+	if (!ppgtt->pt_dma_addr) {
+		drm_mm_remove_node(&ppgtt->node);
+		gen6_ppgtt_free(ppgtt);
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static int gen6_ppgtt_setup_page_tables(struct i915_hw_ppgtt *ppgtt)
+{
+	struct drm_device *dev = ppgtt->base.dev;
+	int i;
 
 	for (i = 0; i < ppgtt->num_pd_entries; i++) {
 		dma_addr_t pt_addr;
@@ -655,48 +1124,71 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt)
 				       PCI_DMA_BIDIRECTIONAL);
 
 		if (pci_dma_mapping_error(dev->pdev, pt_addr)) {
-			ret = -EIO;
-			goto err_pd_pin;
-
+			gen6_ppgtt_unmap_pages(ppgtt);
+			return -EIO;
 		}
+
 		ppgtt->pt_dma_addr[i] = pt_addr;
 	}
 
-	ppgtt->base.clear_range(&ppgtt->base, 0,
-				ppgtt->num_pd_entries * I915_PPGTT_PT_ENTRIES, true);
+	return 0;
+}
 
-	ppgtt->pd_offset = first_pd_entry_in_global_pt * sizeof(gen6_gtt_pte_t);
+static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt)
+{
+	struct drm_device *dev = ppgtt->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
 
-	return 0;
+	ppgtt->base.pte_encode = dev_priv->gtt.base.pte_encode;
+	if (IS_GEN6(dev)) {
+		ppgtt->enable = gen6_ppgtt_enable;
+		ppgtt->switch_mm = gen6_mm_switch;
+	} else if (IS_HASWELL(dev)) {
+		ppgtt->enable = gen7_ppgtt_enable;
+		ppgtt->switch_mm = hsw_mm_switch;
+	} else if (IS_GEN7(dev)) {
+		ppgtt->enable = gen7_ppgtt_enable;
+		ppgtt->switch_mm = gen7_mm_switch;
+	} else
+		BUG();
 
-err_pd_pin:
-	if (ppgtt->pt_dma_addr) {
-		for (i--; i >= 0; i--)
-			pci_unmap_page(dev->pdev, ppgtt->pt_dma_addr[i],
-				       4096, PCI_DMA_BIDIRECTIONAL);
-	}
-err_pt_alloc:
-	kfree(ppgtt->pt_dma_addr);
-	for (i = 0; i < ppgtt->num_pd_entries; i++) {
-		if (ppgtt->pt_pages[i])
-			__free_page(ppgtt->pt_pages[i]);
+	ret = gen6_ppgtt_alloc(ppgtt);
+	if (ret)
+		return ret;
+
+	ret = gen6_ppgtt_setup_page_tables(ppgtt);
+	if (ret) {
+		gen6_ppgtt_free(ppgtt);
+		return ret;
 	}
-	kfree(ppgtt->pt_pages);
 
-	return ret;
+	ppgtt->base.clear_range = gen6_ppgtt_clear_range;
+	ppgtt->base.insert_entries = gen6_ppgtt_insert_entries;
+	ppgtt->base.cleanup = gen6_ppgtt_cleanup;
+	ppgtt->base.start = 0;
+	ppgtt->base.total =  ppgtt->num_pd_entries * I915_PPGTT_PT_ENTRIES * PAGE_SIZE;
+	ppgtt->debug_dump = gen6_dump_ppgtt;
+
+	ppgtt->pd_offset =
+		ppgtt->node.start / PAGE_SIZE * sizeof(gen6_gtt_pte_t);
+
+	ppgtt->base.clear_range(&ppgtt->base, 0, ppgtt->base.total, true);
+
+	DRM_DEBUG_DRIVER("Allocated pde space (%ldM) at GTT entry: %lx\n",
+			 ppgtt->node.size >> 20,
+			 ppgtt->node.start / PAGE_SIZE);
+
+	return 0;
 }
 
-static int i915_gem_init_aliasing_ppgtt(struct drm_device *dev)
+int i915_gem_init_ppgtt(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct i915_hw_ppgtt *ppgtt;
-	int ret;
-
-	ppgtt = kzalloc(sizeof(*ppgtt), GFP_KERNEL);
-	if (!ppgtt)
-		return -ENOMEM;
+	int ret = 0;
 
 	ppgtt->base.dev = dev;
+	ppgtt->base.scratch = dev_priv->gtt.base.scratch;
 
 	if (INTEL_INFO(dev)->gen < 8)
 		ret = gen6_ppgtt_init(ppgtt);
@@ -705,45 +1197,37 @@ static int i915_gem_init_aliasing_ppgtt(struct drm_device *dev)
 	else
 		BUG();
 
-	if (ret)
-		kfree(ppgtt);
-	else {
-		dev_priv->mm.aliasing_ppgtt = ppgtt;
+	if (!ret) {
+		struct drm_i915_private *dev_priv = dev->dev_private;
+		kref_init(&ppgtt->ref);
 		drm_mm_init(&ppgtt->base.mm, ppgtt->base.start,
 			    ppgtt->base.total);
+		i915_init_vm(dev_priv, &ppgtt->base);
+		if (INTEL_INFO(dev)->gen < 8) {
+			gen6_write_pdes(ppgtt);
+			DRM_DEBUG("Adding PPGTT at offset %x\n",
+				  ppgtt->pd_offset << 10);
+		}
 	}
 
 	return ret;
 }
 
-void i915_gem_cleanup_aliasing_ppgtt(struct drm_device *dev)
+static void
+ppgtt_bind_vma(struct i915_vma *vma,
+	       enum i915_cache_level cache_level,
+	       u32 flags)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
-
-	if (!ppgtt)
-		return;
-
-	ppgtt->base.cleanup(&ppgtt->base);
-	dev_priv->mm.aliasing_ppgtt = NULL;
+	vma->vm->insert_entries(vma->vm, vma->obj->pages, vma->node.start,
+				cache_level);
 }
 
-void i915_ppgtt_bind_object(struct i915_hw_ppgtt *ppgtt,
-			    struct drm_i915_gem_object *obj,
-			    enum i915_cache_level cache_level)
+static void ppgtt_unbind_vma(struct i915_vma *vma)
 {
-	ppgtt->base.insert_entries(&ppgtt->base, obj->pages,
-				   i915_gem_obj_ggtt_offset(obj) >> PAGE_SHIFT,
-				   cache_level);
-}
-
-void i915_ppgtt_unbind_object(struct i915_hw_ppgtt *ppgtt,
-			      struct drm_i915_gem_object *obj)
-{
-	ppgtt->base.clear_range(&ppgtt->base,
-				i915_gem_obj_ggtt_offset(obj) >> PAGE_SHIFT,
-				obj->base.size >> PAGE_SHIFT,
-				true);
+	vma->vm->clear_range(vma->vm,
+			     vma->node.start,
+			     vma->obj->base.size,
+			     true);
 }
 
 extern int intel_iommu_gfx_mapped;
@@ -787,7 +1271,7 @@ static void undo_idling(struct drm_i915_private *dev_priv, bool interruptible)
 void i915_check_and_clear_faults(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	int i;
 
 	if (INTEL_INFO(dev)->gen < 6)
@@ -826,27 +1310,59 @@ void i915_gem_suspend_gtt_mappings(struct drm_device *dev)
 	i915_check_and_clear_faults(dev);
 
 	dev_priv->gtt.base.clear_range(&dev_priv->gtt.base,
-				       dev_priv->gtt.base.start / PAGE_SIZE,
-				       dev_priv->gtt.base.total / PAGE_SIZE,
-				       false);
+				       dev_priv->gtt.base.start,
+				       dev_priv->gtt.base.total,
+				       true);
 }
 
 void i915_gem_restore_gtt_mappings(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj;
+	struct i915_address_space *vm;
 
 	i915_check_and_clear_faults(dev);
 
 	/* First fill our portion of the GTT with scratch pages */
 	dev_priv->gtt.base.clear_range(&dev_priv->gtt.base,
-				       dev_priv->gtt.base.start / PAGE_SIZE,
-				       dev_priv->gtt.base.total / PAGE_SIZE,
+				       dev_priv->gtt.base.start,
+				       dev_priv->gtt.base.total,
 				       true);
 
 	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
+		struct i915_vma *vma = i915_gem_obj_to_vma(obj,
+							   &dev_priv->gtt.base);
+		if (!vma)
+			continue;
+
 		i915_gem_clflush_object(obj, obj->pin_display);
-		i915_gem_gtt_bind_object(obj, obj->cache_level);
+		/* The bind_vma code tries to be smart about tracking mappings.
+		 * Unfortunately above, we've just wiped out the mappings
+		 * without telling our object about it. So we need to fake it.
+		 */
+		obj->has_global_gtt_mapping = 0;
+		vma->bind_vma(vma, obj->cache_level, GLOBAL_BIND);
+	}
+
+
+	if (INTEL_INFO(dev)->gen >= 8) {
+		if (IS_CHERRYVIEW(dev))
+			chv_setup_private_ppat(dev_priv);
+		else
+			bdw_setup_private_ppat(dev_priv);
+
+		return;
+	}
+
+	list_for_each_entry(vm, &dev_priv->vm_list, global_link) {
+		/* TODO: Perhaps it shouldn't be gen6 specific */
+		if (i915_is_ggtt(vm)) {
+			if (dev_priv->mm.aliasing_ppgtt)
+				gen6_write_pdes(dev_priv->mm.aliasing_ppgtt);
+			continue;
+		}
+
+		gen6_write_pdes(container_of(vm, struct i915_hw_ppgtt, base));
 	}
 
 	i915_gem_chipset_flush(dev);
@@ -877,15 +1393,16 @@ static inline void gen8_set_pte(void __iomem *addr, gen8_gtt_pte_t pte)
 
 static void gen8_ggtt_insert_entries(struct i915_address_space *vm,
 				     struct sg_table *st,
-				     unsigned int first_entry,
+				     uint64_t start,
 				     enum i915_cache_level level)
 {
 	struct drm_i915_private *dev_priv = vm->dev->dev_private;
+	unsigned first_entry = start >> PAGE_SHIFT;
 	gen8_gtt_pte_t __iomem *gtt_entries =
 		(gen8_gtt_pte_t __iomem *)dev_priv->gtt.gsm + first_entry;
 	int i = 0;
 	struct sg_page_iter sg_iter;
-	dma_addr_t addr;
+	dma_addr_t addr = 0;
 
 	for_each_sg_page(st->sgl, &sg_iter, st->nents, 0) {
 		addr = sg_dma_address(sg_iter.sg) +
@@ -922,10 +1439,11 @@ static void gen8_ggtt_insert_entries(struct i915_address_space *vm,
  */
 static void gen6_ggtt_insert_entries(struct i915_address_space *vm,
 				     struct sg_table *st,
-				     unsigned int first_entry,
+				     uint64_t start,
 				     enum i915_cache_level level)
 {
 	struct drm_i915_private *dev_priv = vm->dev->dev_private;
+	unsigned first_entry = start >> PAGE_SHIFT;
 	gen6_gtt_pte_t __iomem *gtt_entries =
 		(gen6_gtt_pte_t __iomem *)dev_priv->gtt.gsm + first_entry;
 	int i = 0;
@@ -957,11 +1475,13 @@ static void gen6_ggtt_insert_entries(struct i915_address_space *vm,
 }
 
 static void gen8_ggtt_clear_range(struct i915_address_space *vm,
-				  unsigned int first_entry,
-				  unsigned int num_entries,
+				  uint64_t start,
+				  uint64_t length,
 				  bool use_scratch)
 {
 	struct drm_i915_private *dev_priv = vm->dev->dev_private;
+	unsigned first_entry = start >> PAGE_SHIFT;
+	unsigned num_entries = length >> PAGE_SHIFT;
 	gen8_gtt_pte_t scratch_pte, __iomem *gtt_base =
 		(gen8_gtt_pte_t __iomem *) dev_priv->gtt.gsm + first_entry;
 	const int max_entries = gtt_total_entries(dev_priv->gtt) - first_entry;
@@ -981,11 +1501,13 @@ static void gen8_ggtt_clear_range(struct i915_address_space *vm,
 }
 
 static void gen6_ggtt_clear_range(struct i915_address_space *vm,
-				  unsigned int first_entry,
-				  unsigned int num_entries,
+				  uint64_t start,
+				  uint64_t length,
 				  bool use_scratch)
 {
 	struct drm_i915_private *dev_priv = vm->dev->dev_private;
+	unsigned first_entry = start >> PAGE_SHIFT;
+	unsigned num_entries = length >> PAGE_SHIFT;
 	gen6_gtt_pte_t scratch_pte, __iomem *gtt_base =
 		(gen6_gtt_pte_t __iomem *) dev_priv->gtt.gsm + first_entry;
 	const int max_entries = gtt_total_entries(dev_priv->gtt) - first_entry;
@@ -1003,53 +1525,103 @@ static void gen6_ggtt_clear_range(struct i915_address_space *vm,
 	readl(gtt_base);
 }
 
-static void i915_ggtt_insert_entries(struct i915_address_space *vm,
-				     struct sg_table *st,
-				     unsigned int pg_start,
-				     enum i915_cache_level cache_level)
+
+static void i915_ggtt_bind_vma(struct i915_vma *vma,
+			       enum i915_cache_level cache_level,
+			       u32 unused)
 {
+	const unsigned long entry = vma->node.start >> PAGE_SHIFT;
 	unsigned int flags = (cache_level == I915_CACHE_NONE) ?
 		AGP_USER_MEMORY : AGP_USER_CACHED_MEMORY;
 
-	intel_gtt_insert_sg_entries(st, pg_start, flags);
-
+	BUG_ON(!i915_is_ggtt(vma->vm));
+	intel_gtt_insert_sg_entries(vma->obj->pages, entry, flags);
+	vma->obj->has_global_gtt_mapping = 1;
 }
 
 static void i915_ggtt_clear_range(struct i915_address_space *vm,
-				  unsigned int first_entry,
-				  unsigned int num_entries,
+				  uint64_t start,
+				  uint64_t length,
 				  bool unused)
 {
+	unsigned first_entry = start >> PAGE_SHIFT;
+	unsigned num_entries = length >> PAGE_SHIFT;
 	intel_gtt_clear_range(first_entry, num_entries);
 }
 
+static void i915_ggtt_unbind_vma(struct i915_vma *vma)
+{
+	const unsigned int first = vma->node.start >> PAGE_SHIFT;
+	const unsigned int size = vma->obj->base.size >> PAGE_SHIFT;
 
-void i915_gem_gtt_bind_object(struct drm_i915_gem_object *obj,
-			      enum i915_cache_level cache_level)
+	BUG_ON(!i915_is_ggtt(vma->vm));
+	vma->obj->has_global_gtt_mapping = 0;
+	intel_gtt_clear_range(first, size);
+}
+
+static void ggtt_bind_vma(struct i915_vma *vma,
+			  enum i915_cache_level cache_level,
+			  u32 flags)
 {
-	struct drm_device *dev = obj->base.dev;
+	struct drm_device *dev = vma->vm->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	const unsigned long entry = i915_gem_obj_ggtt_offset(obj) >> PAGE_SHIFT;
+	struct drm_i915_gem_object *obj = vma->obj;
 
-	dev_priv->gtt.base.insert_entries(&dev_priv->gtt.base, obj->pages,
-					  entry,
-					  cache_level);
+	/* If there is no aliasing PPGTT, or the caller needs a global mapping,
+	 * or we have a global mapping already but the cacheability flags have
+	 * changed, set the global PTEs.
+	 *
+	 * If there is an aliasing PPGTT it is anecdotally faster, so use that
+	 * instead if none of the above hold true.
+	 *
+	 * NB: A global mapping should only be needed for special regions like
+	 * "gtt mappable", SNB errata, or if specified via special execbuf
+	 * flags. At all other times, the GPU will use the aliasing PPGTT.
+	 */
+	if (!dev_priv->mm.aliasing_ppgtt || flags & GLOBAL_BIND) {
+		if (!obj->has_global_gtt_mapping ||
+		    (cache_level != obj->cache_level)) {
+			vma->vm->insert_entries(vma->vm, obj->pages,
+						vma->node.start,
+						cache_level);
+			obj->has_global_gtt_mapping = 1;
+		}
+	}
 
-	obj->has_global_gtt_mapping = 1;
+	if (dev_priv->mm.aliasing_ppgtt &&
+	    (!obj->has_aliasing_ppgtt_mapping ||
+	     (cache_level != obj->cache_level))) {
+		struct i915_hw_ppgtt *appgtt = dev_priv->mm.aliasing_ppgtt;
+		appgtt->base.insert_entries(&appgtt->base,
+					    vma->obj->pages,
+					    vma->node.start,
+					    cache_level);
+		vma->obj->has_aliasing_ppgtt_mapping = 1;
+	}
 }
 
-void i915_gem_gtt_unbind_object(struct drm_i915_gem_object *obj)
+static void ggtt_unbind_vma(struct i915_vma *vma)
 {
-	struct drm_device *dev = obj->base.dev;
+	struct drm_device *dev = vma->vm->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	const unsigned long entry = i915_gem_obj_ggtt_offset(obj) >> PAGE_SHIFT;
-
-	dev_priv->gtt.base.clear_range(&dev_priv->gtt.base,
-				       entry,
-				       obj->base.size >> PAGE_SHIFT,
-				       true);
+	struct drm_i915_gem_object *obj = vma->obj;
+
+	if (obj->has_global_gtt_mapping) {
+		vma->vm->clear_range(vma->vm,
+				     vma->node.start,
+				     obj->base.size,
+				     true);
+		obj->has_global_gtt_mapping = 0;
+	}
 
-	obj->has_global_gtt_mapping = 0;
+	if (obj->has_aliasing_ppgtt_mapping) {
+		struct i915_hw_ppgtt *appgtt = dev_priv->mm.aliasing_ppgtt;
+		appgtt->base.clear_range(&appgtt->base,
+					 vma->node.start,
+					 obj->base.size,
+					 true);
+		obj->has_aliasing_ppgtt_mapping = 0;
+	}
 }
 
 void i915_gem_gtt_finish_object(struct drm_i915_gem_object *obj)
@@ -1124,7 +1696,6 @@ void i915_gem_setup_global_gtt(struct drm_device *dev,
 		if (ret)
 			DRM_DEBUG_KMS("Reservation failed\n");
 		obj->has_global_gtt_mapping = 1;
-		list_add(&vma->vma_link, &obj->vma_list);
 	}
 
 	dev_priv->gtt.base.start = start;
@@ -1132,29 +1703,14 @@ void i915_gem_setup_global_gtt(struct drm_device *dev,
 
 	/* Clear any non-preallocated blocks */
 	drm_mm_for_each_hole(entry, &ggtt_vm->mm, hole_start, hole_end) {
-		const unsigned long count = (hole_end - hole_start) / PAGE_SIZE;
 		DRM_DEBUG_KMS("clearing unused GTT space: [%lx, %lx]\n",
 			      hole_start, hole_end);
-		ggtt_vm->clear_range(ggtt_vm, hole_start / PAGE_SIZE, count, true);
+		ggtt_vm->clear_range(ggtt_vm, hole_start,
+				     hole_end - hole_start, true);
 	}
 
 	/* And finally clear the reserved guard page */
-	ggtt_vm->clear_range(ggtt_vm, end / PAGE_SIZE - 1, 1, true);
-}
-
-static bool
-intel_enable_ppgtt(struct drm_device *dev)
-{
-	if (i915_enable_ppgtt >= 0)
-		return i915_enable_ppgtt;
-
-#ifdef CONFIG_INTEL_IOMMU
-	/* Disable ppgtt on SNB if VT-d is on. */
-	if (INTEL_INFO(dev)->gen == 6 && intel_iommu_gfx_mapped)
-		return false;
-#endif
-
-	return true;
+	ggtt_vm->clear_range(ggtt_vm, end - PAGE_SIZE, PAGE_SIZE, true);
 }
 
 void i915_gem_init_global_gtt(struct drm_device *dev)
@@ -1165,26 +1721,6 @@ void i915_gem_init_global_gtt(struct drm_device *dev)
 	gtt_size = dev_priv->gtt.base.total;
 	mappable_size = dev_priv->gtt.mappable_end;
 
-	if (intel_enable_ppgtt(dev) && HAS_ALIASING_PPGTT(dev)) {
-		int ret;
-
-		if (INTEL_INFO(dev)->gen <= 7) {
-			/* PPGTT pdes are stolen from global gtt ptes, so shrink the
-			 * aperture accordingly when using aliasing ppgtt. */
-			gtt_size -= GEN6_PPGTT_PD_ENTRIES * PAGE_SIZE;
-		}
-
-		i915_gem_setup_global_gtt(dev, 0, mappable_size, gtt_size);
-
-		ret = i915_gem_init_aliasing_ppgtt(dev);
-		if (!ret)
-			return;
-
-		DRM_ERROR("Aliased PPGTT setup failed %d\n", ret);
-		drm_mm_takedown(&dev_priv->gtt.base.mm);
-		if (INTEL_INFO(dev)->gen < 8)
-			gtt_size += GEN6_PPGTT_PD_ENTRIES*PAGE_SIZE;
-	}
 	i915_gem_setup_global_gtt(dev, 0, mappable_size, gtt_size);
 }
 
@@ -1239,14 +1775,27 @@ static inline unsigned int gen8_get_total_gtt_size(u16 bdw_gmch_ctl)
 	bdw_gmch_ctl &= BDW_GMCH_GGMS_MASK;
 	if (bdw_gmch_ctl)
 		bdw_gmch_ctl = 1 << bdw_gmch_ctl;
-	if (bdw_gmch_ctl > 4) {
-		WARN_ON(!i915_preliminary_hw_support);
-		return 4<<20;
-	}
+
+#ifdef CONFIG_X86_32
+	/* Limit 32b platforms to a 2GB GGTT: 4 << 20 / pte size * PAGE_SIZE */
+	if (bdw_gmch_ctl > 4)
+		bdw_gmch_ctl = 4;
+#endif
 
 	return bdw_gmch_ctl << 20;
 }
 
+static inline unsigned int chv_get_total_gtt_size(u16 gmch_ctrl)
+{
+	gmch_ctrl >>= SNB_GMCH_GGMS_SHIFT;
+	gmch_ctrl &= SNB_GMCH_GGMS_MASK;
+
+	if (gmch_ctrl)
+		return 1 << (20 + gmch_ctrl);
+
+	return 0;
+}
+
 static inline size_t gen6_get_stolen_size(u16 snb_gmch_ctl)
 {
 	snb_gmch_ctl >>= SNB_GMCH_GMS_SHIFT;
@@ -1261,6 +1810,24 @@ static inline size_t gen8_get_stolen_size(u16 bdw_gmch_ctl)
 	return bdw_gmch_ctl << 25; /* 32 MB units */
 }
 
+static size_t chv_get_stolen_size(u16 gmch_ctrl)
+{
+	gmch_ctrl >>= SNB_GMCH_GMS_SHIFT;
+	gmch_ctrl &= SNB_GMCH_GMS_MASK;
+
+	/*
+	 * 0x0  to 0x10: 32MB increments starting at 0MB
+	 * 0x11 to 0x16: 4MB increments starting at 8MB
+	 * 0x17 to 0x1d: 4MB increments start at 36MB
+	 */
+	if (gmch_ctrl < 0x11)
+		return gmch_ctrl << 25;
+	else if (gmch_ctrl < 0x17)
+		return (gmch_ctrl - 0x11 + 2) << 22;
+	else
+		return (gmch_ctrl - 0x17 + 9) << 22;
+}
+
 static int ggtt_probe_common(struct drm_device *dev,
 			     size_t gtt_size)
 {
@@ -1291,19 +1858,8 @@ static int ggtt_probe_common(struct drm_device *dev,
 /* The GGTT and PPGTT need a private PPAT setup in order to handle cacheability
  * bits. When using advanced contexts each context stores its own PAT, but
  * writing this data shouldn't be harmful even in those cases. */
-static void gen8_setup_private_ppat(struct drm_i915_private *dev_priv)
-{
-#define GEN8_PPAT_UC		(0<<0)
-#define GEN8_PPAT_WC		(1<<0)
-#define GEN8_PPAT_WT		(2<<0)
-#define GEN8_PPAT_WB		(3<<0)
-#define GEN8_PPAT_ELLC_OVERRIDE	(0<<2)
-/* FIXME(BDW): Bspec is completely confused about cache control bits. */
-#define GEN8_PPAT_LLC		(1<<2)
-#define GEN8_PPAT_LLCELLC	(2<<2)
-#define GEN8_PPAT_LLCeLLC	(3<<2)
-#define GEN8_PPAT_AGE(x)	(x<<4)
-#define GEN8_PPAT(i, x) ((uint64_t) (x) << ((i) * 8))
+static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv)
+{
 	uint64_t pat;
 
 	pat = GEN8_PPAT(0, GEN8_PPAT_WB | GEN8_PPAT_LLC)     | /* for normal objects, no eLLC */
@@ -1321,6 +1877,33 @@ static void gen8_setup_private_ppat(struct drm_i915_private *dev_priv)
 	I915_WRITE(GEN8_PRIVATE_PAT + 4, pat >> 32);
 }
 
+static void chv_setup_private_ppat(struct drm_i915_private *dev_priv)
+{
+	uint64_t pat;
+
+	/*
+	 * Map WB on BDW to snooped on CHV.
+	 *
+	 * Only the snoop bit has meaning for CHV, the rest is
+	 * ignored.
+	 *
+	 * Note that the harware enforces snooping for all page
+	 * table accesses. The snoop bit is actually ignored for
+	 * PDEs.
+	 */
+	pat = GEN8_PPAT(0, CHV_PPAT_SNOOP) |
+	      GEN8_PPAT(1, 0) |
+	      GEN8_PPAT(2, 0) |
+	      GEN8_PPAT(3, 0) |
+	      GEN8_PPAT(4, CHV_PPAT_SNOOP) |
+	      GEN8_PPAT(5, CHV_PPAT_SNOOP) |
+	      GEN8_PPAT(6, CHV_PPAT_SNOOP) |
+	      GEN8_PPAT(7, CHV_PPAT_SNOOP);
+
+	I915_WRITE(GEN8_PRIVATE_PAT, pat);
+	I915_WRITE(GEN8_PRIVATE_PAT + 4, pat >> 32);
+}
+
 static int gen8_gmch_probe(struct drm_device *dev,
 			   size_t *gtt_total,
 			   size_t *stolen,
@@ -1341,12 +1924,20 @@ static int gen8_gmch_probe(struct drm_device *dev,
 
 	pci_read_config_word(dev->pdev, SNB_GMCH_CTRL, &snb_gmch_ctl);
 
-	*stolen = gen8_get_stolen_size(snb_gmch_ctl);
+	if (IS_CHERRYVIEW(dev)) {
+		*stolen = chv_get_stolen_size(snb_gmch_ctl);
+		gtt_size = chv_get_total_gtt_size(snb_gmch_ctl);
+	} else {
+		*stolen = gen8_get_stolen_size(snb_gmch_ctl);
+		gtt_size = gen8_get_total_gtt_size(snb_gmch_ctl);
+	}
 
-	gtt_size = gen8_get_total_gtt_size(snb_gmch_ctl);
 	*gtt_total = (gtt_size / sizeof(gen8_gtt_pte_t)) << PAGE_SHIFT;
 
-	gen8_setup_private_ppat(dev_priv);
+	if (IS_CHERRYVIEW(dev))
+		chv_setup_private_ppat(dev_priv);
+	else
+		bdw_setup_private_ppat(dev_priv);
 
 	ret = ggtt_probe_common(dev, gtt_size);
 
@@ -1400,6 +1991,11 @@ static void gen6_gmch_remove(struct i915_address_space *vm)
 {
 
 	struct i915_gtt *gtt = container_of(vm, struct i915_gtt, base);
+
+	if (drm_mm_initialized(&vm->mm)) {
+		drm_mm_takedown(&vm->mm);
+		list_del(&vm->global_link);
+	}
 	iounmap(gtt->gsm);
 	teardown_scratch_page(vm->dev);
 }
@@ -1423,13 +2019,19 @@ static int i915_gmch_probe(struct drm_device *dev,
 
 	dev_priv->gtt.do_idle_maps = needs_idle_maps(dev_priv->dev);
 	dev_priv->gtt.base.clear_range = i915_ggtt_clear_range;
-	dev_priv->gtt.base.insert_entries = i915_ggtt_insert_entries;
+
+	if (unlikely(dev_priv->gtt.do_idle_maps))
+		DRM_INFO("applying Ironlake quirks for intel_iommu\n");
 
 	return 0;
 }
 
 static void i915_gmch_remove(struct i915_address_space *vm)
 {
+	if (drm_mm_initialized(&vm->mm)) {
+		drm_mm_takedown(&vm->mm);
+		list_del(&vm->global_link);
+	}
 	intel_gmch_remove();
 }
 
@@ -1472,6 +2074,77 @@ int i915_gem_gtt_init(struct drm_device *dev)
 		 gtt->base.total >> 20);
 	DRM_DEBUG_DRIVER("GMADR size = %ldM\n", gtt->mappable_end >> 20);
 	DRM_DEBUG_DRIVER("GTT stolen size = %zdM\n", gtt->stolen_size >> 20);
+#ifdef CONFIG_INTEL_IOMMU
+	if (intel_iommu_gfx_mapped)
+		DRM_INFO("VT-d active for gfx access\n");
+#endif
+	/*
+	 * i915.enable_ppgtt is read-only, so do an early pass to validate the
+	 * user's requested state against the hardware/driver capabilities.  We
+	 * do this now so that we can print out any log messages once rather
+	 * than every time we check intel_enable_ppgtt().
+	 */
+	i915.enable_ppgtt = sanitize_enable_ppgtt(dev, i915.enable_ppgtt);
+	DRM_DEBUG_DRIVER("ppgtt mode: %i\n", i915.enable_ppgtt);
 
 	return 0;
 }
+
+static struct i915_vma *__i915_gem_vma_create(struct drm_i915_gem_object *obj,
+					      struct i915_address_space *vm)
+{
+	struct i915_vma *vma = kzalloc(sizeof(*vma), GFP_KERNEL);
+	if (vma == NULL)
+		return ERR_PTR(-ENOMEM);
+
+	INIT_LIST_HEAD(&vma->vma_link);
+	INIT_LIST_HEAD(&vma->mm_list);
+	INIT_LIST_HEAD(&vma->exec_list);
+	vma->vm = vm;
+	vma->obj = obj;
+
+	switch (INTEL_INFO(vm->dev)->gen) {
+	case 8:
+	case 7:
+	case 6:
+		if (i915_is_ggtt(vm)) {
+			vma->unbind_vma = ggtt_unbind_vma;
+			vma->bind_vma = ggtt_bind_vma;
+		} else {
+			vma->unbind_vma = ppgtt_unbind_vma;
+			vma->bind_vma = ppgtt_bind_vma;
+		}
+		break;
+	case 5:
+	case 4:
+	case 3:
+	case 2:
+		BUG_ON(!i915_is_ggtt(vm));
+		vma->unbind_vma = i915_ggtt_unbind_vma;
+		vma->bind_vma = i915_ggtt_bind_vma;
+		break;
+	default:
+		BUG();
+	}
+
+	/* Keep GGTT vmas first to make debug easier */
+	if (i915_is_ggtt(vm))
+		list_add(&vma->vma_link, &obj->vma_list);
+	else
+		list_add_tail(&vma->vma_link, &obj->vma_list);
+
+	return vma;
+}
+
+struct i915_vma *
+i915_gem_obj_lookup_or_create_vma(struct drm_i915_gem_object *obj,
+				  struct i915_address_space *vm)
+{
+	struct i915_vma *vma;
+
+	vma = i915_gem_obj_to_vma(obj, vm);
+	if (!vma)
+		vma = __i915_gem_vma_create(obj, vm);
+
+	return vma;
+}
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
new file mode 100644
index 00000000000..1b96a06be3c
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -0,0 +1,284 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ * Please try to maintain the following order within this file unless it makes
+ * sense to do otherwise. From top to bottom:
+ * 1. typedefs
+ * 2. #defines, and macros
+ * 3. structure definitions
+ * 4. function prototypes
+ *
+ * Within each section, please try to order by generation in ascending order,
+ * from top to bottom (ie. gen6 on the top, gen8 on the bottom).
+ */
+
+#ifndef __I915_GEM_GTT_H__
+#define __I915_GEM_GTT_H__
+
+typedef uint32_t gen6_gtt_pte_t;
+typedef uint64_t gen8_gtt_pte_t;
+typedef gen8_gtt_pte_t gen8_ppgtt_pde_t;
+
+#define gtt_total_entries(gtt) ((gtt).base.total >> PAGE_SHIFT)
+
+#define I915_PPGTT_PT_ENTRIES		(PAGE_SIZE / sizeof(gen6_gtt_pte_t))
+/* gen6-hsw has bit 11-4 for physical addr bit 39-32 */
+#define GEN6_GTT_ADDR_ENCODE(addr)	((addr) | (((addr) >> 28) & 0xff0))
+#define GEN6_PTE_ADDR_ENCODE(addr)	GEN6_GTT_ADDR_ENCODE(addr)
+#define GEN6_PDE_ADDR_ENCODE(addr)	GEN6_GTT_ADDR_ENCODE(addr)
+#define GEN6_PTE_CACHE_LLC		(2 << 1)
+#define GEN6_PTE_UNCACHED		(1 << 1)
+#define GEN6_PTE_VALID			(1 << 0)
+
+#define GEN6_PPGTT_PD_ENTRIES		512
+#define GEN6_PD_SIZE			(GEN6_PPGTT_PD_ENTRIES * PAGE_SIZE)
+#define GEN6_PD_ALIGN			(PAGE_SIZE * 16)
+#define GEN6_PDE_VALID			(1 << 0)
+
+#define GEN7_PTE_CACHE_L3_LLC		(3 << 1)
+
+#define BYT_PTE_SNOOPED_BY_CPU_CACHES	(1 << 2)
+#define BYT_PTE_WRITEABLE		(1 << 1)
+
+/* Cacheability Control is a 4-bit value. The low three bits are stored in bits
+ * 3:1 of the PTE, while the fourth bit is stored in bit 11 of the PTE.
+ */
+#define HSW_CACHEABILITY_CONTROL(bits)	((((bits) & 0x7) << 1) | \
+					 (((bits) & 0x8) << (11 - 3)))
+#define HSW_WB_LLC_AGE3			HSW_CACHEABILITY_CONTROL(0x2)
+#define HSW_WB_LLC_AGE0			HSW_CACHEABILITY_CONTROL(0x3)
+#define HSW_WB_ELLC_LLC_AGE3		HSW_CACHEABILITY_CONTROL(0x8)
+#define HSW_WB_ELLC_LLC_AGE0		HSW_CACHEABILITY_CONTROL(0xb)
+#define HSW_WT_ELLC_LLC_AGE3		HSW_CACHEABILITY_CONTROL(0x7)
+#define HSW_WT_ELLC_LLC_AGE0		HSW_CACHEABILITY_CONTROL(0x6)
+#define HSW_PTE_UNCACHED		(0)
+#define HSW_GTT_ADDR_ENCODE(addr)	((addr) | (((addr) >> 28) & 0x7f0))
+#define HSW_PTE_ADDR_ENCODE(addr)	HSW_GTT_ADDR_ENCODE(addr)
+
+/* GEN8 legacy style address is defined as a 3 level page table:
+ * 31:30 | 29:21 | 20:12 |  11:0
+ * PDPE  |  PDE  |  PTE  | offset
+ * The difference as compared to normal x86 3 level page table is the PDPEs are
+ * programmed via register.
+ */
+#define GEN8_PDPE_SHIFT			30
+#define GEN8_PDPE_MASK			0x3
+#define GEN8_PDE_SHIFT			21
+#define GEN8_PDE_MASK			0x1ff
+#define GEN8_PTE_SHIFT			12
+#define GEN8_PTE_MASK			0x1ff
+#define GEN8_LEGACY_PDPS		4
+#define GEN8_PTES_PER_PAGE		(PAGE_SIZE / sizeof(gen8_gtt_pte_t))
+#define GEN8_PDES_PER_PAGE		(PAGE_SIZE / sizeof(gen8_ppgtt_pde_t))
+
+#define PPAT_UNCACHED_INDEX		(_PAGE_PWT | _PAGE_PCD)
+#define PPAT_CACHED_PDE_INDEX		0 /* WB LLC */
+#define PPAT_CACHED_INDEX		_PAGE_PAT /* WB LLCeLLC */
+#define PPAT_DISPLAY_ELLC_INDEX		_PAGE_PCD /* WT eLLC */
+
+#define CHV_PPAT_SNOOP			(1<<6)
+#define GEN8_PPAT_AGE(x)		(x<<4)
+#define GEN8_PPAT_LLCeLLC		(3<<2)
+#define GEN8_PPAT_LLCELLC		(2<<2)
+#define GEN8_PPAT_LLC			(1<<2)
+#define GEN8_PPAT_WB			(3<<0)
+#define GEN8_PPAT_WT			(2<<0)
+#define GEN8_PPAT_WC			(1<<0)
+#define GEN8_PPAT_UC			(0<<0)
+#define GEN8_PPAT_ELLC_OVERRIDE		(0<<2)
+#define GEN8_PPAT(i, x)			((uint64_t) (x) << ((i) * 8))
+
+enum i915_cache_level;
+/**
+ * A VMA represents a GEM BO that is bound into an address space. Therefore, a
+ * VMA's presence cannot be guaranteed before binding, or after unbinding the
+ * object into/from the address space.
+ *
+ * To make things as simple as possible (ie. no refcounting), a VMA's lifetime
+ * will always be <= an objects lifetime. So object refcounting should cover us.
+ */
+struct i915_vma {
+	struct drm_mm_node node;
+	struct drm_i915_gem_object *obj;
+	struct i915_address_space *vm;
+
+	/** This object's place on the active/inactive lists */
+	struct list_head mm_list;
+
+	struct list_head vma_link; /* Link in the object's VMA list */
+
+	/** This vma's place in the batchbuffer or on the eviction list */
+	struct list_head exec_list;
+
+	/**
+	 * Used for performing relocations during execbuffer insertion.
+	 */
+	struct hlist_node exec_node;
+	unsigned long exec_handle;
+	struct drm_i915_gem_exec_object2 *exec_entry;
+
+	/**
+	 * How many users have pinned this object in GTT space. The following
+	 * users can each hold at most one reference: pwrite/pread, pin_ioctl
+	 * (via user_pin_count), execbuffer (objects are not allowed multiple
+	 * times for the same batchbuffer), and the framebuffer code. When
+	 * switching/pageflipping, the framebuffer code has at most two buffers
+	 * pinned per crtc.
+	 *
+	 * In the worst case this is 1 + 1 + 1 + 2*2 = 7. That would fit into 3
+	 * bits with absolutely no headroom. So use 4 bits. */
+	unsigned int pin_count:4;
+#define DRM_I915_GEM_OBJECT_MAX_PIN_COUNT 0xf
+
+	/** Unmap an object from an address space. This usually consists of
+	 * setting the valid PTE entries to a reserved scratch page. */
+	void (*unbind_vma)(struct i915_vma *vma);
+	/* Map an object into an address space with the given cache flags. */
+#define GLOBAL_BIND (1<<0)
+	void (*bind_vma)(struct i915_vma *vma,
+			 enum i915_cache_level cache_level,
+			 u32 flags);
+};
+
+struct i915_address_space {
+	struct drm_mm mm;
+	struct drm_device *dev;
+	struct list_head global_link;
+	unsigned long start;		/* Start offset always 0 for dri2 */
+	size_t total;		/* size addr space maps (ex. 2GB for ggtt) */
+
+	struct {
+		dma_addr_t addr;
+		struct page *page;
+	} scratch;
+
+	/**
+	 * List of objects currently involved in rendering.
+	 *
+	 * Includes buffers having the contents of their GPU caches
+	 * flushed, not necessarily primitives.  last_rendering_seqno
+	 * represents when the rendering involved will be completed.
+	 *
+	 * A reference is held on the buffer while on this list.
+	 */
+	struct list_head active_list;
+
+	/**
+	 * LRU list of objects which are not in the ringbuffer and
+	 * are ready to unbind, but are still in the GTT.
+	 *
+	 * last_rendering_seqno is 0 while an object is in this list.
+	 *
+	 * A reference is not held on the buffer while on this list,
+	 * as merely being GTT-bound shouldn't prevent its being
+	 * freed, and we'll pull it off the list in the free path.
+	 */
+	struct list_head inactive_list;
+
+	/* FIXME: Need a more generic return type */
+	gen6_gtt_pte_t (*pte_encode)(dma_addr_t addr,
+				     enum i915_cache_level level,
+				     bool valid); /* Create a valid PTE */
+	void (*clear_range)(struct i915_address_space *vm,
+			    uint64_t start,
+			    uint64_t length,
+			    bool use_scratch);
+	void (*insert_entries)(struct i915_address_space *vm,
+			       struct sg_table *st,
+			       uint64_t start,
+			       enum i915_cache_level cache_level);
+	void (*cleanup)(struct i915_address_space *vm);
+};
+
+/* The Graphics Translation Table is the way in which GEN hardware translates a
+ * Graphics Virtual Address into a Physical Address. In addition to the normal
+ * collateral associated with any va->pa translations GEN hardware also has a
+ * portion of the GTT which can be mapped by the CPU and remain both coherent
+ * and correct (in cases like swizzling). That region is referred to as GMADR in
+ * the spec.
+ */
+struct i915_gtt {
+	struct i915_address_space base;
+	size_t stolen_size;		/* Total size of stolen memory */
+
+	unsigned long mappable_end;	/* End offset that we can CPU map */
+	struct io_mapping *mappable;	/* Mapping to our CPU mappable region */
+	phys_addr_t mappable_base;	/* PA of our GMADR */
+
+	/** "Graphics Stolen Memory" holds the global PTEs */
+	void __iomem *gsm;
+
+	bool do_idle_maps;
+
+	int mtrr;
+
+	/* global gtt ops */
+	int (*gtt_probe)(struct drm_device *dev, size_t *gtt_total,
+			  size_t *stolen, phys_addr_t *mappable_base,
+			  unsigned long *mappable_end);
+};
+
+struct i915_hw_ppgtt {
+	struct i915_address_space base;
+	struct kref ref;
+	struct drm_mm_node node;
+	unsigned num_pd_entries;
+	unsigned num_pd_pages; /* gen8+ */
+	union {
+		struct page **pt_pages;
+		struct page **gen8_pt_pages[GEN8_LEGACY_PDPS];
+	};
+	struct page *pd_pages;
+	union {
+		uint32_t pd_offset;
+		dma_addr_t pd_dma_addr[GEN8_LEGACY_PDPS];
+	};
+	union {
+		dma_addr_t *pt_dma_addr;
+		dma_addr_t *gen8_pt_dma_addr[4];
+	};
+
+	struct intel_context *ctx;
+
+	int (*enable)(struct i915_hw_ppgtt *ppgtt);
+	int (*switch_mm)(struct i915_hw_ppgtt *ppgtt,
+			 struct intel_engine_cs *ring,
+			 bool synchronous);
+	void (*debug_dump)(struct i915_hw_ppgtt *ppgtt, struct seq_file *m);
+};
+
+int i915_gem_gtt_init(struct drm_device *dev);
+void i915_gem_init_global_gtt(struct drm_device *dev);
+void i915_gem_setup_global_gtt(struct drm_device *dev, unsigned long start,
+			       unsigned long mappable_end, unsigned long end);
+
+bool intel_enable_ppgtt(struct drm_device *dev, bool full);
+int i915_gem_init_ppgtt(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt);
+
+void i915_check_and_clear_faults(struct drm_device *dev);
+void i915_gem_suspend_gtt_mappings(struct drm_device *dev);
+void i915_gem_restore_gtt_mappings(struct drm_device *dev);
+
+int __must_check i915_gem_gtt_prepare_object(struct drm_i915_gem_object *obj);
+void i915_gem_gtt_finish_object(struct drm_i915_gem_object *obj);
+
+#endif
diff --git a/drivers/gpu/drm/i915/i915_gem_render_state.c b/drivers/gpu/drm/i915/i915_gem_render_state.c
new file mode 100644
index 00000000000..34894b57306
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem_render_state.c
@@ -0,0 +1,198 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ * Authors:
+ *    Mika Kuoppala <mika.kuoppala@intel.com>
+ *
+ */
+
+#include "i915_drv.h"
+#include "intel_renderstate.h"
+
+struct i915_render_state {
+	struct drm_i915_gem_object *obj;
+	unsigned long ggtt_offset;
+	u32 *batch;
+	u32 size;
+	u32 len;
+};
+
+static struct i915_render_state *render_state_alloc(struct drm_device *dev)
+{
+	struct i915_render_state *so;
+	struct page *page;
+	int ret;
+
+	so = kzalloc(sizeof(*so), GFP_KERNEL);
+	if (!so)
+		return ERR_PTR(-ENOMEM);
+
+	so->obj = i915_gem_alloc_object(dev, 4096);
+	if (so->obj == NULL) {
+		ret = -ENOMEM;
+		goto free;
+	}
+	so->size = 4096;
+
+	ret = i915_gem_obj_ggtt_pin(so->obj, 4096, 0);
+	if (ret)
+		goto free_gem;
+
+	BUG_ON(so->obj->pages->nents != 1);
+	page = sg_page(so->obj->pages->sgl);
+
+	so->batch = kmap(page);
+	if (!so->batch) {
+		ret = -ENOMEM;
+		goto unpin;
+	}
+
+	so->ggtt_offset = i915_gem_obj_ggtt_offset(so->obj);
+
+	return so;
+unpin:
+	i915_gem_object_ggtt_unpin(so->obj);
+free_gem:
+	drm_gem_object_unreference(&so->obj->base);
+free:
+	kfree(so);
+	return ERR_PTR(ret);
+}
+
+static void render_state_free(struct i915_render_state *so)
+{
+	kunmap(kmap_to_page(so->batch));
+	i915_gem_object_ggtt_unpin(so->obj);
+	drm_gem_object_unreference(&so->obj->base);
+	kfree(so);
+}
+
+static const struct intel_renderstate_rodata *
+render_state_get_rodata(struct drm_device *dev, const int gen)
+{
+	switch (gen) {
+	case 6:
+		return &gen6_null_state;
+	case 7:
+		return &gen7_null_state;
+	case 8:
+		return &gen8_null_state;
+	}
+
+	return NULL;
+}
+
+static int render_state_setup(const int gen,
+			      const struct intel_renderstate_rodata *rodata,
+			      struct i915_render_state *so)
+{
+	const u64 goffset = i915_gem_obj_ggtt_offset(so->obj);
+	u32 reloc_index = 0;
+	u32 * const d = so->batch;
+	unsigned int i = 0;
+	int ret;
+
+	if (!rodata || rodata->batch_items * 4 > so->size)
+		return -EINVAL;
+
+	ret = i915_gem_object_set_to_cpu_domain(so->obj, true);
+	if (ret)
+		return ret;
+
+	while (i < rodata->batch_items) {
+		u32 s = rodata->batch[i];
+
+		if (reloc_index < rodata->reloc_items &&
+		    i * 4  == rodata->reloc[reloc_index]) {
+
+			s += goffset & 0xffffffff;
+
+			/* We keep batch offsets max 32bit */
+			if (gen >= 8) {
+				if (i + 1 >= rodata->batch_items ||
+				    rodata->batch[i + 1] != 0)
+					return -EINVAL;
+
+				d[i] = s;
+				i++;
+				s = (goffset & 0xffffffff00000000ull) >> 32;
+			}
+
+			reloc_index++;
+		}
+
+		d[i] = s;
+		i++;
+	}
+
+	ret = i915_gem_object_set_to_gtt_domain(so->obj, false);
+	if (ret)
+		return ret;
+
+	if (rodata->reloc_items != reloc_index) {
+		DRM_ERROR("not all relocs resolved, %d out of %d\n",
+			  reloc_index, rodata->reloc_items);
+		return -EINVAL;
+	}
+
+	so->len = rodata->batch_items * 4;
+
+	return 0;
+}
+
+int i915_gem_render_state_init(struct intel_engine_cs *ring)
+{
+	const int gen = INTEL_INFO(ring->dev)->gen;
+	struct i915_render_state *so;
+	const struct intel_renderstate_rodata *rodata;
+	int ret;
+
+	if (WARN_ON(ring->id != RCS))
+		return -ENOENT;
+
+	rodata = render_state_get_rodata(ring->dev, gen);
+	if (rodata == NULL)
+		return 0;
+
+	so = render_state_alloc(ring->dev);
+	if (IS_ERR(so))
+		return PTR_ERR(so);
+
+	ret = render_state_setup(gen, rodata, so);
+	if (ret)
+		goto out;
+
+	ret = ring->dispatch_execbuffer(ring,
+					i915_gem_obj_ggtt_offset(so->obj),
+					so->len,
+					I915_DISPATCH_SECURE);
+	if (ret)
+		goto out;
+
+	i915_vma_move_to_active(i915_gem_obj_to_ggtt(so->obj), ring);
+
+	ret = __i915_add_request(ring, NULL, so->obj, NULL);
+	/* __i915_add_request moves object to inactive if it fails */
+out:
+	render_state_free(so);
+	return ret;
+}
diff --git a/drivers/gpu/drm/i915/i915_gem_stolen.c b/drivers/gpu/drm/i915/i915_gem_stolen.c
index d284d892ed9..7465ab0fd39 100644
--- a/drivers/gpu/drm/i915/i915_gem_stolen.c
+++ b/drivers/gpu/drm/i915/i915_gem_stolen.c
@@ -74,6 +74,50 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev)
 	if (base == 0)
 		return 0;
 
+	/* make sure we don't clobber the GTT if it's within stolen memory */
+	if (INTEL_INFO(dev)->gen <= 4 && !IS_G33(dev) && !IS_G4X(dev)) {
+		struct {
+			u32 start, end;
+		} stolen[2] = {
+			{ .start = base, .end = base + dev_priv->gtt.stolen_size, },
+			{ .start = base, .end = base + dev_priv->gtt.stolen_size, },
+		};
+		u64 gtt_start, gtt_end;
+
+		gtt_start = I915_READ(PGTBL_CTL);
+		if (IS_GEN4(dev))
+			gtt_start = (gtt_start & PGTBL_ADDRESS_LO_MASK) |
+				(gtt_start & PGTBL_ADDRESS_HI_MASK) << 28;
+		else
+			gtt_start &= PGTBL_ADDRESS_LO_MASK;
+		gtt_end = gtt_start + gtt_total_entries(dev_priv->gtt) * 4;
+
+		if (gtt_start >= stolen[0].start && gtt_start < stolen[0].end)
+			stolen[0].end = gtt_start;
+		if (gtt_end > stolen[1].start && gtt_end <= stolen[1].end)
+			stolen[1].start = gtt_end;
+
+		/* pick the larger of the two chunks */
+		if (stolen[0].end - stolen[0].start >
+		    stolen[1].end - stolen[1].start) {
+			base = stolen[0].start;
+			dev_priv->gtt.stolen_size = stolen[0].end - stolen[0].start;
+		} else {
+			base = stolen[1].start;
+			dev_priv->gtt.stolen_size = stolen[1].end - stolen[1].start;
+		}
+
+		if (stolen[0].start != stolen[1].start ||
+		    stolen[0].end != stolen[1].end) {
+			DRM_DEBUG_KMS("GTT within stolen memory at 0x%llx-0x%llx\n",
+				      (unsigned long long) gtt_start,
+				      (unsigned long long) gtt_end - 1);
+			DRM_DEBUG_KMS("Stolen memory adjusted to 0x%x-0x%x\n",
+				      base, base + (u32) dev_priv->gtt.stolen_size - 1);
+		}
+	}
+
+
 	/* Verify that nothing else uses this physical address. Stolen
 	 * memory should be reserved by the BIOS and hidden from the
 	 * kernel. So if the region is already marked as busy, something
@@ -82,9 +126,22 @@ static unsigned long i915_stolen_to_physical(struct drm_device *dev)
 	r = devm_request_mem_region(dev->dev, base, dev_priv->gtt.stolen_size,
 				    "Graphics Stolen Memory");
 	if (r == NULL) {
-		DRM_ERROR("conflict detected with stolen region: [0x%08x - 0x%08x]\n",
-			  base, base + (uint32_t)dev_priv->gtt.stolen_size);
-		base = 0;
+		/*
+		 * One more attempt but this time requesting region from
+		 * base + 1, as we have seen that this resolves the region
+		 * conflict with the PCI Bus.
+		 * This is a BIOS w/a: Some BIOS wrap stolen in the root
+		 * PCI bus, but have an off-by-one error. Hence retry the
+		 * reservation starting from 1 instead of 0.
+		 */
+		r = devm_request_mem_region(dev->dev, base + 1,
+					    dev_priv->gtt.stolen_size - 1,
+					    "Graphics Stolen Memory");
+		if (r == NULL) {
+			DRM_ERROR("conflict detected with stolen region: [0x%08x - 0x%08x]\n",
+				  base, base + (uint32_t)dev_priv->gtt.stolen_size);
+			base = 0;
+		}
 	}
 
 	return base;
@@ -201,6 +258,13 @@ int i915_gem_init_stolen(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int bios_reserved = 0;
 
+#ifdef CONFIG_INTEL_IOMMU
+	if (intel_iommu_gfx_mapped && INTEL_INFO(dev)->gen < 8) {
+		DRM_INFO("DMAR active, disabling use of stolen memory\n");
+		return 0;
+	}
+#endif
+
 	if (dev_priv->gtt.stolen_size == 0)
 		return 0;
 
@@ -250,7 +314,7 @@ i915_pages_create_for_stolen(struct drm_device *dev,
 	}
 
 	sg = st->sgl;
-	sg->offset = offset;
+	sg->offset = 0;
 	sg->length = size;
 
 	sg_dma_address(sg) = (dma_addr_t)dev_priv->mm.stolen_base + offset;
@@ -420,6 +484,7 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_device *dev,
 
 	list_add_tail(&obj->global_list, &dev_priv->mm.bound_list);
 	list_add_tail(&vma->mm_list, &ggtt->inactive_list);
+	i915_gem_object_pin_pages(obj);
 
 	return obj;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_tiling.c b/drivers/gpu/drm/i915/i915_gem_tiling.c
index b1390534804..cb150e8b433 100644
--- a/drivers/gpu/drm/i915/i915_gem_tiling.c
+++ b/drivers/gpu/drm/i915/i915_gem_tiling.c
@@ -87,7 +87,7 @@
 void
 i915_gem_detect_bit_6_swizzle(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
 	uint32_t swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
 
@@ -294,7 +294,7 @@ i915_gem_set_tiling(struct drm_device *dev, void *data,
 		   struct drm_file *file)
 {
 	struct drm_i915_gem_set_tiling *args = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj;
 	int ret = 0;
 
@@ -308,7 +308,7 @@ i915_gem_set_tiling(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
-	if (obj->pin_count || obj->framebuffer_references) {
+	if (i915_gem_obj_is_pinned(obj) || obj->framebuffer_references) {
 		drm_gem_object_unreference_unlocked(&obj->base);
 		return -EBUSY;
 	}
@@ -415,7 +415,7 @@ i915_gem_get_tiling(struct drm_device *dev, void *data,
 		   struct drm_file *file)
 {
 	struct drm_i915_gem_get_tiling *args = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj;
 
 	obj = to_intel_bo(drm_gem_object_lookup(dev, file, args->handle));
diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c
new file mode 100644
index 00000000000..21ea92886a5
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_gem_userptr.c
@@ -0,0 +1,711 @@
+/*
+ * Copyright © 2012-2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include "drmP.h"
+#include "i915_drm.h"
+#include "i915_drv.h"
+#include "i915_trace.h"
+#include "intel_drv.h"
+#include <linux/mmu_context.h>
+#include <linux/mmu_notifier.h>
+#include <linux/mempolicy.h>
+#include <linux/swap.h>
+
+#if defined(CONFIG_MMU_NOTIFIER)
+#include <linux/interval_tree.h>
+
+struct i915_mmu_notifier {
+	spinlock_t lock;
+	struct hlist_node node;
+	struct mmu_notifier mn;
+	struct rb_root objects;
+	struct drm_device *dev;
+	struct mm_struct *mm;
+	struct work_struct work;
+	unsigned long count;
+	unsigned long serial;
+};
+
+struct i915_mmu_object {
+	struct i915_mmu_notifier *mmu;
+	struct interval_tree_node it;
+	struct drm_i915_gem_object *obj;
+};
+
+static void i915_gem_userptr_mn_invalidate_range_start(struct mmu_notifier *_mn,
+						       struct mm_struct *mm,
+						       unsigned long start,
+						       unsigned long end)
+{
+	struct i915_mmu_notifier *mn = container_of(_mn, struct i915_mmu_notifier, mn);
+	struct interval_tree_node *it = NULL;
+	unsigned long serial = 0;
+
+	end--; /* interval ranges are inclusive, but invalidate range is exclusive */
+	while (start < end) {
+		struct drm_i915_gem_object *obj;
+
+		obj = NULL;
+		spin_lock(&mn->lock);
+		if (serial == mn->serial)
+			it = interval_tree_iter_next(it, start, end);
+		else
+			it = interval_tree_iter_first(&mn->objects, start, end);
+		if (it != NULL) {
+			obj = container_of(it, struct i915_mmu_object, it)->obj;
+			drm_gem_object_reference(&obj->base);
+			serial = mn->serial;
+		}
+		spin_unlock(&mn->lock);
+		if (obj == NULL)
+			return;
+
+		mutex_lock(&mn->dev->struct_mutex);
+		/* Cancel any active worker and force us to re-evaluate gup */
+		obj->userptr.work = NULL;
+
+		if (obj->pages != NULL) {
+			struct drm_i915_private *dev_priv = to_i915(mn->dev);
+			struct i915_vma *vma, *tmp;
+			bool was_interruptible;
+
+			was_interruptible = dev_priv->mm.interruptible;
+			dev_priv->mm.interruptible = false;
+
+			list_for_each_entry_safe(vma, tmp, &obj->vma_list, vma_link) {
+				int ret = i915_vma_unbind(vma);
+				WARN_ON(ret && ret != -EIO);
+			}
+			WARN_ON(i915_gem_object_put_pages(obj));
+
+			dev_priv->mm.interruptible = was_interruptible;
+		}
+
+		start = obj->userptr.ptr + obj->base.size;
+
+		drm_gem_object_unreference(&obj->base);
+		mutex_unlock(&mn->dev->struct_mutex);
+	}
+}
+
+static const struct mmu_notifier_ops i915_gem_userptr_notifier = {
+	.invalidate_range_start = i915_gem_userptr_mn_invalidate_range_start,
+};
+
+static struct i915_mmu_notifier *
+__i915_mmu_notifier_lookup(struct drm_device *dev, struct mm_struct *mm)
+{
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct i915_mmu_notifier *mmu;
+
+	/* Protected by dev->struct_mutex */
+	hash_for_each_possible(dev_priv->mmu_notifiers, mmu, node, (unsigned long)mm)
+		if (mmu->mm == mm)
+			return mmu;
+
+	return NULL;
+}
+
+static struct i915_mmu_notifier *
+i915_mmu_notifier_get(struct drm_device *dev, struct mm_struct *mm)
+{
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct i915_mmu_notifier *mmu;
+	int ret;
+
+	lockdep_assert_held(&dev->struct_mutex);
+
+	mmu = __i915_mmu_notifier_lookup(dev, mm);
+	if (mmu)
+		return mmu;
+
+	mmu = kmalloc(sizeof(*mmu), GFP_KERNEL);
+	if (mmu == NULL)
+		return ERR_PTR(-ENOMEM);
+
+	spin_lock_init(&mmu->lock);
+	mmu->dev = dev;
+	mmu->mn.ops = &i915_gem_userptr_notifier;
+	mmu->mm = mm;
+	mmu->objects = RB_ROOT;
+	mmu->count = 0;
+	mmu->serial = 0;
+
+	/* Protected by mmap_sem (write-lock) */
+	ret = __mmu_notifier_register(&mmu->mn, mm);
+	if (ret) {
+		kfree(mmu);
+		return ERR_PTR(ret);
+	}
+
+	/* Protected by dev->struct_mutex */
+	hash_add(dev_priv->mmu_notifiers, &mmu->node, (unsigned long)mm);
+	return mmu;
+}
+
+static void
+__i915_mmu_notifier_destroy_worker(struct work_struct *work)
+{
+	struct i915_mmu_notifier *mmu = container_of(work, typeof(*mmu), work);
+	mmu_notifier_unregister(&mmu->mn, mmu->mm);
+	kfree(mmu);
+}
+
+static void
+__i915_mmu_notifier_destroy(struct i915_mmu_notifier *mmu)
+{
+	lockdep_assert_held(&mmu->dev->struct_mutex);
+
+	/* Protected by dev->struct_mutex */
+	hash_del(&mmu->node);
+
+	/* Our lock ordering is: mmap_sem, mmu_notifier_scru, struct_mutex.
+	 * We enter the function holding struct_mutex, therefore we need
+	 * to drop our mutex prior to calling mmu_notifier_unregister in
+	 * order to prevent lock inversion (and system-wide deadlock)
+	 * between the mmap_sem and struct-mutex. Hence we defer the
+	 * unregistration to a workqueue where we hold no locks.
+	 */
+	INIT_WORK(&mmu->work, __i915_mmu_notifier_destroy_worker);
+	schedule_work(&mmu->work);
+}
+
+static void __i915_mmu_notifier_update_serial(struct i915_mmu_notifier *mmu)
+{
+	if (++mmu->serial == 0)
+		mmu->serial = 1;
+}
+
+static void
+i915_mmu_notifier_del(struct i915_mmu_notifier *mmu,
+		      struct i915_mmu_object *mn)
+{
+	lockdep_assert_held(&mmu->dev->struct_mutex);
+
+	spin_lock(&mmu->lock);
+	interval_tree_remove(&mn->it, &mmu->objects);
+	__i915_mmu_notifier_update_serial(mmu);
+	spin_unlock(&mmu->lock);
+
+	/* Protected against _add() by dev->struct_mutex */
+	if (--mmu->count == 0)
+		__i915_mmu_notifier_destroy(mmu);
+}
+
+static int
+i915_mmu_notifier_add(struct i915_mmu_notifier *mmu,
+		      struct i915_mmu_object *mn)
+{
+	struct interval_tree_node *it;
+	int ret;
+
+	ret = i915_mutex_lock_interruptible(mmu->dev);
+	if (ret)
+		return ret;
+
+	/* Make sure we drop the final active reference (and thereby
+	 * remove the objects from the interval tree) before we do
+	 * the check for overlapping objects.
+	 */
+	i915_gem_retire_requests(mmu->dev);
+
+	/* Disallow overlapping userptr objects */
+	spin_lock(&mmu->lock);
+	it = interval_tree_iter_first(&mmu->objects,
+				      mn->it.start, mn->it.last);
+	if (it) {
+		struct drm_i915_gem_object *obj;
+
+		/* We only need to check the first object in the range as it
+		 * either has cancelled gup work queued and we need to
+		 * return back to the user to give time for the gup-workers
+		 * to flush their object references upon which the object will
+		 * be removed from the interval-tree, or the the range is
+		 * still in use by another client and the overlap is invalid.
+		 */
+
+		obj = container_of(it, struct i915_mmu_object, it)->obj;
+		ret = obj->userptr.workers ? -EAGAIN : -EINVAL;
+	} else {
+		interval_tree_insert(&mn->it, &mmu->objects);
+		__i915_mmu_notifier_update_serial(mmu);
+		ret = 0;
+	}
+	spin_unlock(&mmu->lock);
+	mutex_unlock(&mmu->dev->struct_mutex);
+
+	return ret;
+}
+
+static void
+i915_gem_userptr_release__mmu_notifier(struct drm_i915_gem_object *obj)
+{
+	struct i915_mmu_object *mn;
+
+	mn = obj->userptr.mn;
+	if (mn == NULL)
+		return;
+
+	i915_mmu_notifier_del(mn->mmu, mn);
+	obj->userptr.mn = NULL;
+}
+
+static int
+i915_gem_userptr_init__mmu_notifier(struct drm_i915_gem_object *obj,
+				    unsigned flags)
+{
+	struct i915_mmu_notifier *mmu;
+	struct i915_mmu_object *mn;
+	int ret;
+
+	if (flags & I915_USERPTR_UNSYNCHRONIZED)
+		return capable(CAP_SYS_ADMIN) ? 0 : -EPERM;
+
+	down_write(&obj->userptr.mm->mmap_sem);
+	ret = i915_mutex_lock_interruptible(obj->base.dev);
+	if (ret == 0) {
+		mmu = i915_mmu_notifier_get(obj->base.dev, obj->userptr.mm);
+		if (!IS_ERR(mmu))
+			mmu->count++; /* preemptive add to act as a refcount */
+		else
+			ret = PTR_ERR(mmu);
+		mutex_unlock(&obj->base.dev->struct_mutex);
+	}
+	up_write(&obj->userptr.mm->mmap_sem);
+	if (ret)
+		return ret;
+
+	mn = kzalloc(sizeof(*mn), GFP_KERNEL);
+	if (mn == NULL) {
+		ret = -ENOMEM;
+		goto destroy_mmu;
+	}
+
+	mn->mmu = mmu;
+	mn->it.start = obj->userptr.ptr;
+	mn->it.last = mn->it.start + obj->base.size - 1;
+	mn->obj = obj;
+
+	ret = i915_mmu_notifier_add(mmu, mn);
+	if (ret)
+		goto free_mn;
+
+	obj->userptr.mn = mn;
+	return 0;
+
+free_mn:
+	kfree(mn);
+destroy_mmu:
+	mutex_lock(&obj->base.dev->struct_mutex);
+	if (--mmu->count == 0)
+		__i915_mmu_notifier_destroy(mmu);
+	mutex_unlock(&obj->base.dev->struct_mutex);
+	return ret;
+}
+
+#else
+
+static void
+i915_gem_userptr_release__mmu_notifier(struct drm_i915_gem_object *obj)
+{
+}
+
+static int
+i915_gem_userptr_init__mmu_notifier(struct drm_i915_gem_object *obj,
+				    unsigned flags)
+{
+	if ((flags & I915_USERPTR_UNSYNCHRONIZED) == 0)
+		return -ENODEV;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	return 0;
+}
+#endif
+
+struct get_pages_work {
+	struct work_struct work;
+	struct drm_i915_gem_object *obj;
+	struct task_struct *task;
+};
+
+
+#if IS_ENABLED(CONFIG_SWIOTLB)
+#define swiotlb_active() swiotlb_nr_tbl()
+#else
+#define swiotlb_active() 0
+#endif
+
+static int
+st_set_pages(struct sg_table **st, struct page **pvec, int num_pages)
+{
+	struct scatterlist *sg;
+	int ret, n;
+
+	*st = kmalloc(sizeof(**st), GFP_KERNEL);
+	if (*st == NULL)
+		return -ENOMEM;
+
+	if (swiotlb_active()) {
+		ret = sg_alloc_table(*st, num_pages, GFP_KERNEL);
+		if (ret)
+			goto err;
+
+		for_each_sg((*st)->sgl, sg, num_pages, n)
+			sg_set_page(sg, pvec[n], PAGE_SIZE, 0);
+	} else {
+		ret = sg_alloc_table_from_pages(*st, pvec, num_pages,
+						0, num_pages << PAGE_SHIFT,
+						GFP_KERNEL);
+		if (ret)
+			goto err;
+	}
+
+	return 0;
+
+err:
+	kfree(*st);
+	*st = NULL;
+	return ret;
+}
+
+static void
+__i915_gem_userptr_get_pages_worker(struct work_struct *_work)
+{
+	struct get_pages_work *work = container_of(_work, typeof(*work), work);
+	struct drm_i915_gem_object *obj = work->obj;
+	struct drm_device *dev = obj->base.dev;
+	const int num_pages = obj->base.size >> PAGE_SHIFT;
+	struct page **pvec;
+	int pinned, ret;
+
+	ret = -ENOMEM;
+	pinned = 0;
+
+	pvec = kmalloc(num_pages*sizeof(struct page *),
+		       GFP_TEMPORARY | __GFP_NOWARN | __GFP_NORETRY);
+	if (pvec == NULL)
+		pvec = drm_malloc_ab(num_pages, sizeof(struct page *));
+	if (pvec != NULL) {
+		struct mm_struct *mm = obj->userptr.mm;
+
+		down_read(&mm->mmap_sem);
+		while (pinned < num_pages) {
+			ret = get_user_pages(work->task, mm,
+					     obj->userptr.ptr + pinned * PAGE_SIZE,
+					     num_pages - pinned,
+					     !obj->userptr.read_only, 0,
+					     pvec + pinned, NULL);
+			if (ret < 0)
+				break;
+
+			pinned += ret;
+		}
+		up_read(&mm->mmap_sem);
+	}
+
+	mutex_lock(&dev->struct_mutex);
+	if (obj->userptr.work != &work->work) {
+		ret = 0;
+	} else if (pinned == num_pages) {
+		ret = st_set_pages(&obj->pages, pvec, num_pages);
+		if (ret == 0) {
+			list_add_tail(&obj->global_list, &to_i915(dev)->mm.unbound_list);
+			pinned = 0;
+		}
+	}
+
+	obj->userptr.work = ERR_PTR(ret);
+	obj->userptr.workers--;
+	drm_gem_object_unreference(&obj->base);
+	mutex_unlock(&dev->struct_mutex);
+
+	release_pages(pvec, pinned, 0);
+	drm_free_large(pvec);
+
+	put_task_struct(work->task);
+	kfree(work);
+}
+
+static int
+i915_gem_userptr_get_pages(struct drm_i915_gem_object *obj)
+{
+	const int num_pages = obj->base.size >> PAGE_SHIFT;
+	struct page **pvec;
+	int pinned, ret;
+
+	/* If userspace should engineer that these pages are replaced in
+	 * the vma between us binding this page into the GTT and completion
+	 * of rendering... Their loss. If they change the mapping of their
+	 * pages they need to create a new bo to point to the new vma.
+	 *
+	 * However, that still leaves open the possibility of the vma
+	 * being copied upon fork. Which falls under the same userspace
+	 * synchronisation issue as a regular bo, except that this time
+	 * the process may not be expecting that a particular piece of
+	 * memory is tied to the GPU.
+	 *
+	 * Fortunately, we can hook into the mmu_notifier in order to
+	 * discard the page references prior to anything nasty happening
+	 * to the vma (discard or cloning) which should prevent the more
+	 * egregious cases from causing harm.
+	 */
+
+	pvec = NULL;
+	pinned = 0;
+	if (obj->userptr.mm == current->mm) {
+		pvec = kmalloc(num_pages*sizeof(struct page *),
+			       GFP_TEMPORARY | __GFP_NOWARN | __GFP_NORETRY);
+		if (pvec == NULL) {
+			pvec = drm_malloc_ab(num_pages, sizeof(struct page *));
+			if (pvec == NULL)
+				return -ENOMEM;
+		}
+
+		pinned = __get_user_pages_fast(obj->userptr.ptr, num_pages,
+					       !obj->userptr.read_only, pvec);
+	}
+	if (pinned < num_pages) {
+		if (pinned < 0) {
+			ret = pinned;
+			pinned = 0;
+		} else {
+			/* Spawn a worker so that we can acquire the
+			 * user pages without holding our mutex. Access
+			 * to the user pages requires mmap_sem, and we have
+			 * a strict lock ordering of mmap_sem, struct_mutex -
+			 * we already hold struct_mutex here and so cannot
+			 * call gup without encountering a lock inversion.
+			 *
+			 * Userspace will keep on repeating the operation
+			 * (thanks to EAGAIN) until either we hit the fast
+			 * path or the worker completes. If the worker is
+			 * cancelled or superseded, the task is still run
+			 * but the results ignored. (This leads to
+			 * complications that we may have a stray object
+			 * refcount that we need to be wary of when
+			 * checking for existing objects during creation.)
+			 * If the worker encounters an error, it reports
+			 * that error back to this function through
+			 * obj->userptr.work = ERR_PTR.
+			 */
+			ret = -EAGAIN;
+			if (obj->userptr.work == NULL &&
+			    obj->userptr.workers < I915_GEM_USERPTR_MAX_WORKERS) {
+				struct get_pages_work *work;
+
+				work = kmalloc(sizeof(*work), GFP_KERNEL);
+				if (work != NULL) {
+					obj->userptr.work = &work->work;
+					obj->userptr.workers++;
+
+					work->obj = obj;
+					drm_gem_object_reference(&obj->base);
+
+					work->task = current;
+					get_task_struct(work->task);
+
+					INIT_WORK(&work->work, __i915_gem_userptr_get_pages_worker);
+					schedule_work(&work->work);
+				} else
+					ret = -ENOMEM;
+			} else {
+				if (IS_ERR(obj->userptr.work)) {
+					ret = PTR_ERR(obj->userptr.work);
+					obj->userptr.work = NULL;
+				}
+			}
+		}
+	} else {
+		ret = st_set_pages(&obj->pages, pvec, num_pages);
+		if (ret == 0) {
+			obj->userptr.work = NULL;
+			pinned = 0;
+		}
+	}
+
+	release_pages(pvec, pinned, 0);
+	drm_free_large(pvec);
+	return ret;
+}
+
+static void
+i915_gem_userptr_put_pages(struct drm_i915_gem_object *obj)
+{
+	struct scatterlist *sg;
+	int i;
+
+	BUG_ON(obj->userptr.work != NULL);
+
+	if (obj->madv != I915_MADV_WILLNEED)
+		obj->dirty = 0;
+
+	for_each_sg(obj->pages->sgl, sg, obj->pages->nents, i) {
+		struct page *page = sg_page(sg);
+
+		if (obj->dirty)
+			set_page_dirty(page);
+
+		mark_page_accessed(page);
+		page_cache_release(page);
+	}
+	obj->dirty = 0;
+
+	sg_free_table(obj->pages);
+	kfree(obj->pages);
+}
+
+static void
+i915_gem_userptr_release(struct drm_i915_gem_object *obj)
+{
+	i915_gem_userptr_release__mmu_notifier(obj);
+
+	if (obj->userptr.mm) {
+		mmput(obj->userptr.mm);
+		obj->userptr.mm = NULL;
+	}
+}
+
+static int
+i915_gem_userptr_dmabuf_export(struct drm_i915_gem_object *obj)
+{
+	if (obj->userptr.mn)
+		return 0;
+
+	return i915_gem_userptr_init__mmu_notifier(obj, 0);
+}
+
+static const struct drm_i915_gem_object_ops i915_gem_userptr_ops = {
+	.dmabuf_export = i915_gem_userptr_dmabuf_export,
+	.get_pages = i915_gem_userptr_get_pages,
+	.put_pages = i915_gem_userptr_put_pages,
+	.release = i915_gem_userptr_release,
+};
+
+/**
+ * Creates a new mm object that wraps some normal memory from the process
+ * context - user memory.
+ *
+ * We impose several restrictions upon the memory being mapped
+ * into the GPU.
+ * 1. It must be page aligned (both start/end addresses, i.e ptr and size).
+ * 2. It cannot overlap any other userptr object in the same address space.
+ * 3. It must be normal system memory, not a pointer into another map of IO
+ *    space (e.g. it must not be a GTT mmapping of another object).
+ * 4. We only allow a bo as large as we could in theory map into the GTT,
+ *    that is we limit the size to the total size of the GTT.
+ * 5. The bo is marked as being snoopable. The backing pages are left
+ *    accessible directly by the CPU, but reads and writes by the GPU may
+ *    incur the cost of a snoop (unless you have an LLC architecture).
+ *
+ * Synchronisation between multiple users and the GPU is left to userspace
+ * through the normal set-domain-ioctl. The kernel will enforce that the
+ * GPU relinquishes the VMA before it is returned back to the system
+ * i.e. upon free(), munmap() or process termination. However, the userspace
+ * malloc() library may not immediately relinquish the VMA after free() and
+ * instead reuse it whilst the GPU is still reading and writing to the VMA.
+ * Caveat emptor.
+ *
+ * Also note, that the object created here is not currently a "first class"
+ * object, in that several ioctls are banned. These are the CPU access
+ * ioctls: mmap(), pwrite and pread. In practice, you are expected to use
+ * direct access via your pointer rather than use those ioctls.
+ *
+ * If you think this is a good interface to use to pass GPU memory between
+ * drivers, please use dma-buf instead. In fact, wherever possible use
+ * dma-buf instead.
+ */
+int
+i915_gem_userptr_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_gem_userptr *args = data;
+	struct drm_i915_gem_object *obj;
+	int ret;
+	u32 handle;
+
+	if (args->flags & ~(I915_USERPTR_READ_ONLY |
+			    I915_USERPTR_UNSYNCHRONIZED))
+		return -EINVAL;
+
+	if (offset_in_page(args->user_ptr | args->user_size))
+		return -EINVAL;
+
+	if (args->user_size > dev_priv->gtt.base.total)
+		return -E2BIG;
+
+	if (!access_ok(args->flags & I915_USERPTR_READ_ONLY ? VERIFY_READ : VERIFY_WRITE,
+		       (char __user *)(unsigned long)args->user_ptr, args->user_size))
+		return -EFAULT;
+
+	if (args->flags & I915_USERPTR_READ_ONLY) {
+		/* On almost all of the current hw, we cannot tell the GPU that a
+		 * page is readonly, so this is just a placeholder in the uAPI.
+		 */
+		return -ENODEV;
+	}
+
+	/* Allocate the new object */
+	obj = i915_gem_object_alloc(dev);
+	if (obj == NULL)
+		return -ENOMEM;
+
+	drm_gem_private_object_init(dev, &obj->base, args->user_size);
+	i915_gem_object_init(obj, &i915_gem_userptr_ops);
+	obj->cache_level = I915_CACHE_LLC;
+	obj->base.write_domain = I915_GEM_DOMAIN_CPU;
+	obj->base.read_domains = I915_GEM_DOMAIN_CPU;
+
+	obj->userptr.ptr = args->user_ptr;
+	obj->userptr.read_only = !!(args->flags & I915_USERPTR_READ_ONLY);
+
+	/* And keep a pointer to the current->mm for resolving the user pages
+	 * at binding. This means that we need to hook into the mmu_notifier
+	 * in order to detect if the mmu is destroyed.
+	 */
+	ret = -ENOMEM;
+	if ((obj->userptr.mm = get_task_mm(current)))
+		ret = i915_gem_userptr_init__mmu_notifier(obj, args->flags);
+	if (ret == 0)
+		ret = drm_gem_handle_create(file, &obj->base, &handle);
+
+	/* drop reference from allocate - handle holds it now */
+	drm_gem_object_unreference_unlocked(&obj->base);
+	if (ret)
+		return ret;
+
+	args->handle = handle;
+	return 0;
+}
+
+int
+i915_gem_init_userptr(struct drm_device *dev)
+{
+#if defined(CONFIG_MMU_NOTIFIER)
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	hash_init(dev_priv->mmu_notifiers);
+#endif
+	return 0;
+}
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 79dcb8f896c..66cf41765bf 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -42,6 +42,7 @@ static const char *ring_str(int ring)
 	case VCS: return "bsd";
 	case BCS: return "blt";
 	case VECS: return "vebox";
+	case VCS2: return "bsd2";
 	default: return "";
 	}
 }
@@ -146,7 +147,10 @@ static void i915_error_vprintf(struct drm_i915_error_state_buf *e,
 		va_list tmp;
 
 		va_copy(tmp, args);
-		if (!__i915_error_seek(e, vsnprintf(NULL, 0, f, tmp)))
+		len = vsnprintf(NULL, 0, f, tmp);
+		va_end(tmp);
+
+		if (!__i915_error_seek(e, len))
 			return;
 	}
 
@@ -201,6 +205,7 @@ static void print_error_buffers(struct drm_i915_error_state_buf *m,
 		err_puts(m, tiling_flag(err->tiling));
 		err_puts(m, dirty_flag(err->dirty));
 		err_puts(m, purgeable_flag(err->purgeable));
+		err_puts(m, err->userptr ? " userptr" : "");
 		err_puts(m, err->ring != -1 ? " " : "");
 		err_puts(m, ring_str(err->ring));
 		err_puts(m, i915_cache_level_str(err->cache_level));
@@ -235,48 +240,62 @@ static const char *hangcheck_action_to_str(enum intel_ring_hangcheck_action a)
 
 static void i915_ring_error_state(struct drm_i915_error_state_buf *m,
 				  struct drm_device *dev,
-				  struct drm_i915_error_state *error,
-				  unsigned ring)
+				  struct drm_i915_error_ring *ring)
 {
-	BUG_ON(ring >= I915_NUM_RINGS); /* shut up confused gcc */
-	err_printf(m, "%s command stream:\n", ring_str(ring));
-	err_printf(m, "  HEAD: 0x%08x\n", error->head[ring]);
-	err_printf(m, "  TAIL: 0x%08x\n", error->tail[ring]);
-	err_printf(m, "  CTL: 0x%08x\n", error->ctl[ring]);
-	err_printf(m, "  ACTHD: 0x%08x\n", error->acthd[ring]);
-	err_printf(m, "  IPEIR: 0x%08x\n", error->ipeir[ring]);
-	err_printf(m, "  IPEHR: 0x%08x\n", error->ipehr[ring]);
-	err_printf(m, "  INSTDONE: 0x%08x\n", error->instdone[ring]);
-	if (ring == RCS && INTEL_INFO(dev)->gen >= 4)
-		err_printf(m, "  BBADDR: 0x%08llx\n", error->bbaddr);
-	if (INTEL_INFO(dev)->gen >= 4)
-		err_printf(m, "  BB_STATE: 0x%08x\n", error->bbstate[ring]);
-	if (INTEL_INFO(dev)->gen >= 4)
-		err_printf(m, "  INSTPS: 0x%08x\n", error->instps[ring]);
-	err_printf(m, "  INSTPM: 0x%08x\n", error->instpm[ring]);
-	err_printf(m, "  FADDR: 0x%08x\n", error->faddr[ring]);
+	if (!ring->valid)
+		return;
+
+	err_printf(m, "  HEAD: 0x%08x\n", ring->head);
+	err_printf(m, "  TAIL: 0x%08x\n", ring->tail);
+	err_printf(m, "  CTL: 0x%08x\n", ring->ctl);
+	err_printf(m, "  HWS: 0x%08x\n", ring->hws);
+	err_printf(m, "  ACTHD: 0x%08x %08x\n", (u32)(ring->acthd>>32), (u32)ring->acthd);
+	err_printf(m, "  IPEIR: 0x%08x\n", ring->ipeir);
+	err_printf(m, "  IPEHR: 0x%08x\n", ring->ipehr);
+	err_printf(m, "  INSTDONE: 0x%08x\n", ring->instdone);
+	if (INTEL_INFO(dev)->gen >= 4) {
+		err_printf(m, "  BBADDR: 0x%08x %08x\n", (u32)(ring->bbaddr>>32), (u32)ring->bbaddr);
+		err_printf(m, "  BB_STATE: 0x%08x\n", ring->bbstate);
+		err_printf(m, "  INSTPS: 0x%08x\n", ring->instps);
+	}
+	err_printf(m, "  INSTPM: 0x%08x\n", ring->instpm);
+	err_printf(m, "  FADDR: 0x%08x %08x\n", upper_32_bits(ring->faddr),
+		   lower_32_bits(ring->faddr));
 	if (INTEL_INFO(dev)->gen >= 6) {
-		err_printf(m, "  RC PSMI: 0x%08x\n", error->rc_psmi[ring]);
-		err_printf(m, "  FAULT_REG: 0x%08x\n", error->fault_reg[ring]);
+		err_printf(m, "  RC PSMI: 0x%08x\n", ring->rc_psmi);
+		err_printf(m, "  FAULT_REG: 0x%08x\n", ring->fault_reg);
 		err_printf(m, "  SYNC_0: 0x%08x [last synced 0x%08x]\n",
-			   error->semaphore_mboxes[ring][0],
-			   error->semaphore_seqno[ring][0]);
+			   ring->semaphore_mboxes[0],
+			   ring->semaphore_seqno[0]);
 		err_printf(m, "  SYNC_1: 0x%08x [last synced 0x%08x]\n",
-			   error->semaphore_mboxes[ring][1],
-			   error->semaphore_seqno[ring][1]);
+			   ring->semaphore_mboxes[1],
+			   ring->semaphore_seqno[1]);
 		if (HAS_VEBOX(dev)) {
 			err_printf(m, "  SYNC_2: 0x%08x [last synced 0x%08x]\n",
-				   error->semaphore_mboxes[ring][2],
-				   error->semaphore_seqno[ring][2]);
+				   ring->semaphore_mboxes[2],
+				   ring->semaphore_seqno[2]);
 		}
 	}
-	err_printf(m, "  seqno: 0x%08x\n", error->seqno[ring]);
-	err_printf(m, "  waiting: %s\n", yesno(error->waiting[ring]));
-	err_printf(m, "  ring->head: 0x%08x\n", error->cpu_ring_head[ring]);
-	err_printf(m, "  ring->tail: 0x%08x\n", error->cpu_ring_tail[ring]);
+	if (USES_PPGTT(dev)) {
+		err_printf(m, "  GFX_MODE: 0x%08x\n", ring->vm_info.gfx_mode);
+
+		if (INTEL_INFO(dev)->gen >= 8) {
+			int i;
+			for (i = 0; i < 4; i++)
+				err_printf(m, "  PDP%d: 0x%016llx\n",
+					   i, ring->vm_info.pdp[i]);
+		} else {
+			err_printf(m, "  PP_DIR_BASE: 0x%08x\n",
+				   ring->vm_info.pp_dir_base);
+		}
+	}
+	err_printf(m, "  seqno: 0x%08x\n", ring->seqno);
+	err_printf(m, "  waiting: %s\n", yesno(ring->waiting));
+	err_printf(m, "  ring->head: 0x%08x\n", ring->cpu_ring_head);
+	err_printf(m, "  ring->tail: 0x%08x\n", ring->cpu_ring_tail);
 	err_printf(m, "  hangcheck: %s [%d]\n",
-		   hangcheck_action_to_str(error->hangcheck_action[ring]),
-		   error->hangcheck_score[ring]);
+		   hangcheck_action_to_str(ring->hangcheck_action),
+		   ring->hangcheck_score);
 }
 
 void i915_error_printf(struct drm_i915_error_state_buf *e, const char *f, ...)
@@ -288,23 +307,54 @@ void i915_error_printf(struct drm_i915_error_state_buf *e, const char *f, ...)
 	va_end(args);
 }
 
+static void print_error_obj(struct drm_i915_error_state_buf *m,
+			    struct drm_i915_error_object *obj)
+{
+	int page, offset, elt;
+
+	for (page = offset = 0; page < obj->page_count; page++) {
+		for (elt = 0; elt < PAGE_SIZE/4; elt++) {
+			err_printf(m, "%08x :  %08x\n", offset,
+				   obj->pages[page][elt]);
+			offset += 4;
+		}
+	}
+}
+
 int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 			    const struct i915_error_state_file_priv *error_priv)
 {
 	struct drm_device *dev = error_priv->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_error_state *error = error_priv->error;
-	struct intel_ring_buffer *ring;
-	int i, j, page, offset, elt;
+	int i, j, offset, elt;
+	int max_hangcheck_score;
 
 	if (!error) {
 		err_printf(m, "no error state collected\n");
 		goto out;
 	}
 
+	err_printf(m, "%s\n", error->error_msg);
 	err_printf(m, "Time: %ld s %ld us\n", error->time.tv_sec,
 		   error->time.tv_usec);
 	err_printf(m, "Kernel: " UTS_RELEASE "\n");
+	max_hangcheck_score = 0;
+	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
+		if (error->ring[i].hangcheck_score > max_hangcheck_score)
+			max_hangcheck_score = error->ring[i].hangcheck_score;
+	}
+	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
+		if (error->ring[i].hangcheck_score == max_hangcheck_score &&
+		    error->ring[i].pid != -1) {
+			err_printf(m, "Active process (on ring %s): %s [%d]\n",
+				   ring_str(i),
+				   error->ring[i].comm,
+				   error->ring[i].pid);
+		}
+	}
+	err_printf(m, "Reset count: %u\n", error->reset_count);
+	err_printf(m, "Suspend count: %u\n", error->suspend_count);
 	err_printf(m, "PCI ID: 0x%04x\n", dev->pdev->device);
 	err_printf(m, "EIR: 0x%08x\n", error->eir);
 	err_printf(m, "IER: 0x%08x\n", error->ier);
@@ -329,8 +379,10 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 	if (INTEL_INFO(dev)->gen == 7)
 		err_printf(m, "ERR_INT: 0x%08x\n", error->err_int);
 
-	for_each_ring(ring, dev_priv, i)
-		i915_ring_error_state(m, dev, error, i);
+	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
+		err_printf(m, "%s command stream:\n", ring_str(i));
+		i915_ring_error_state(m, dev, &error->ring[i]);
+	}
 
 	if (error->active_bo)
 		print_error_buffers(m, "Active",
@@ -345,18 +397,23 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
 		struct drm_i915_error_object *obj;
 
-		if ((obj = error->ring[i].batchbuffer)) {
-			err_printf(m, "%s --- gtt_offset = 0x%08x\n",
-				   dev_priv->ring[i].name,
+		obj = error->ring[i].batchbuffer;
+		if (obj) {
+			err_puts(m, dev_priv->ring[i].name);
+			if (error->ring[i].pid != -1)
+				err_printf(m, " (submitted by %s [%d])",
+					   error->ring[i].comm,
+					   error->ring[i].pid);
+			err_printf(m, " --- gtt_offset = 0x%08x\n",
 				   obj->gtt_offset);
-			offset = 0;
-			for (page = 0; page < obj->page_count; page++) {
-				for (elt = 0; elt < PAGE_SIZE/4; elt++) {
-					err_printf(m, "%08x :  %08x\n", offset,
-						   obj->pages[page][elt]);
-					offset += 4;
-				}
-			}
+			print_error_obj(m, obj);
+		}
+
+		obj = error->ring[i].wa_batchbuffer;
+		if (obj) {
+			err_printf(m, "%s (w/a) --- gtt_offset = 0x%08x\n",
+				   dev_priv->ring[i].name, obj->gtt_offset);
+			print_error_obj(m, obj);
 		}
 
 		if (error->ring[i].num_requests) {
@@ -375,20 +432,11 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 			err_printf(m, "%s --- ringbuffer = 0x%08x\n",
 				   dev_priv->ring[i].name,
 				   obj->gtt_offset);
-			offset = 0;
-			for (page = 0; page < obj->page_count; page++) {
-				for (elt = 0; elt < PAGE_SIZE/4; elt++) {
-					err_printf(m, "%08x :  %08x\n",
-						   offset,
-						   obj->pages[page][elt]);
-					offset += 4;
-				}
-			}
+			print_error_obj(m, obj);
 		}
 
-		obj = error->ring[i].ctx;
-		if (obj) {
-			err_printf(m, "%s --- HW Context = 0x%08x\n",
+		if ((obj = error->ring[i].hws_page)) {
+			err_printf(m, "%s --- HW Status = 0x%08x\n",
 				   dev_priv->ring[i].name,
 				   obj->gtt_offset);
 			offset = 0;
@@ -402,6 +450,13 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 					offset += 16;
 			}
 		}
+
+		if ((obj = error->ring[i].ctx)) {
+			err_printf(m, "%s --- HW Context = 0x%08x\n",
+				   dev_priv->ring[i].name,
+				   obj->gtt_offset);
+			print_error_obj(m, obj);
+		}
 	}
 
 	if (error->overlay)
@@ -469,6 +524,7 @@ static void i915_error_state_free(struct kref *error_ref)
 	for (i = 0; i < ARRAY_SIZE(error->ring); i++) {
 		i915_error_object_free(error->ring[i].batchbuffer);
 		i915_error_object_free(error->ring[i].ringbuffer);
+		i915_error_object_free(error->ring[i].hws_page);
 		i915_error_object_free(error->ring[i].ctx);
 		kfree(error->ring[i].requests);
 	}
@@ -482,6 +538,7 @@ static void i915_error_state_free(struct kref *error_ref)
 static struct drm_i915_error_object *
 i915_error_object_create_sized(struct drm_i915_private *dev_priv,
 			       struct drm_i915_gem_object *src,
+			       struct i915_address_space *vm,
 			       const int num_pages)
 {
 	struct drm_i915_error_object *dst;
@@ -495,7 +552,7 @@ i915_error_object_create_sized(struct drm_i915_private *dev_priv,
 	if (dst == NULL)
 		return NULL;
 
-	reloc_offset = dst->gtt_offset = i915_gem_obj_ggtt_offset(src);
+	reloc_offset = dst->gtt_offset = i915_gem_obj_offset(src, vm);
 	for (i = 0; i < num_pages; i++) {
 		unsigned long flags;
 		void *d;
@@ -505,8 +562,10 @@ i915_error_object_create_sized(struct drm_i915_private *dev_priv,
 			goto unwind;
 
 		local_irq_save(flags);
-		if (reloc_offset < dev_priv->gtt.mappable_end &&
-		    src->has_global_gtt_mapping) {
+		if (src->cache_level == I915_CACHE_NONE &&
+		    reloc_offset < dev_priv->gtt.mappable_end &&
+		    src->has_global_gtt_mapping &&
+		    i915_is_ggtt(vm)) {
 			void __iomem *s;
 
 			/* Simply ignore tiling or any overlapping fence.
@@ -556,8 +615,12 @@ unwind:
 	kfree(dst);
 	return NULL;
 }
-#define i915_error_object_create(dev_priv, src) \
-	i915_error_object_create_sized((dev_priv), (src), \
+#define i915_error_object_create(dev_priv, src, vm) \
+	i915_error_object_create_sized((dev_priv), (src), (vm), \
+				       (src)->base.size>>PAGE_SHIFT)
+
+#define i915_error_ggtt_object_create(dev_priv, src) \
+	i915_error_object_create_sized((dev_priv), (src), &(dev_priv)->gtt.base, \
 				       (src)->base.size>>PAGE_SHIFT)
 
 static void capture_bo(struct drm_i915_error_buffer *err,
@@ -572,13 +635,14 @@ static void capture_bo(struct drm_i915_error_buffer *err,
 	err->write_domain = obj->base.write_domain;
 	err->fence_reg = obj->fence_reg;
 	err->pinned = 0;
-	if (obj->pin_count > 0)
+	if (i915_gem_obj_is_pinned(obj))
 		err->pinned = 1;
 	if (obj->user_pin_count > 0)
 		err->pinned = -1;
 	err->tiling = obj->tiling_mode;
 	err->dirty = obj->dirty;
 	err->purgeable = obj->madv != I915_MADV_WILLNEED;
+	err->userptr = obj->userptr.mm != NULL;
 	err->ring = obj->ring ? obj->ring->id : -1;
 	err->cache_level = obj->cache_level;
 }
@@ -605,7 +669,7 @@ static u32 capture_pinned_bo(struct drm_i915_error_buffer *err,
 	int i = 0;
 
 	list_for_each_entry(obj, head, global_list) {
-		if (obj->pin_count == 0)
+		if (!i915_gem_obj_is_pinned(obj))
 			continue;
 
 		capture_bo(err++, obj);
@@ -616,6 +680,39 @@ static u32 capture_pinned_bo(struct drm_i915_error_buffer *err,
 	return i;
 }
 
+/* Generate a semi-unique error code. The code is not meant to have meaning, The
+ * code's only purpose is to try to prevent false duplicated bug reports by
+ * grossly estimating a GPU error state.
+ *
+ * TODO Ideally, hashing the batchbuffer would be a very nice way to determine
+ * the hang if we could strip the GTT offset information from it.
+ *
+ * It's only a small step better than a random number in its current form.
+ */
+static uint32_t i915_error_generate_code(struct drm_i915_private *dev_priv,
+					 struct drm_i915_error_state *error,
+					 int *ring_id)
+{
+	uint32_t error_code = 0;
+	int i;
+
+	/* IPEHR would be an ideal way to detect errors, as it's the gross
+	 * measure of "the command that hung." However, has some very common
+	 * synchronization commands which almost always appear in the case
+	 * strictly a client bug. Use instdone to differentiate those some.
+	 */
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		if (error->ring[i].hangcheck_action == HANGCHECK_HUNG) {
+			if (ring_id)
+				*ring_id = i;
+
+			return error->ring[i].ipehr ^ error->ring[i].instdone;
+		}
+	}
+
+	return error_code;
+}
+
 static void i915_gem_record_fences(struct drm_device *dev,
 				   struct drm_i915_error_state *error)
 {
@@ -649,109 +746,120 @@ static void i915_gem_record_fences(struct drm_device *dev,
 	}
 }
 
-static struct drm_i915_error_object *
-i915_error_first_batchbuffer(struct drm_i915_private *dev_priv,
-			     struct intel_ring_buffer *ring)
-{
-	struct i915_address_space *vm;
-	struct i915_vma *vma;
-	struct drm_i915_gem_object *obj;
-	u32 seqno;
-
-	if (!ring->get_seqno)
-		return NULL;
-
-	if (HAS_BROKEN_CS_TLB(dev_priv->dev)) {
-		u32 acthd = I915_READ(ACTHD);
-
-		if (WARN_ON(ring->id != RCS))
-			return NULL;
-
-		obj = ring->scratch.obj;
-		if (acthd >= i915_gem_obj_ggtt_offset(obj) &&
-		    acthd < i915_gem_obj_ggtt_offset(obj) + obj->base.size)
-			return i915_error_object_create(dev_priv, obj);
-	}
-
-	seqno = ring->get_seqno(ring, false);
-	list_for_each_entry(vm, &dev_priv->vm_list, global_link) {
-		list_for_each_entry(vma, &vm->active_list, mm_list) {
-			obj = vma->obj;
-			if (obj->ring != ring)
-				continue;
-
-			if (i915_seqno_passed(seqno, obj->last_read_seqno))
-				continue;
-
-			if ((obj->base.read_domains & I915_GEM_DOMAIN_COMMAND) == 0)
-				continue;
-
-			/* We need to copy these to an anonymous buffer as the simplest
-			 * method to avoid being overwritten by userspace.
-			 */
-			return i915_error_object_create(dev_priv, obj);
-		}
-	}
-
-	return NULL;
-}
-
 static void i915_record_ring_state(struct drm_device *dev,
-				   struct drm_i915_error_state *error,
-				   struct intel_ring_buffer *ring)
+				   struct intel_engine_cs *ring,
+				   struct drm_i915_error_ring *ering)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	if (INTEL_INFO(dev)->gen >= 6) {
-		error->rc_psmi[ring->id] = I915_READ(ring->mmio_base + 0x50);
-		error->fault_reg[ring->id] = I915_READ(RING_FAULT_REG(ring));
-		error->semaphore_mboxes[ring->id][0]
+		ering->rc_psmi = I915_READ(ring->mmio_base + 0x50);
+		ering->fault_reg = I915_READ(RING_FAULT_REG(ring));
+		ering->semaphore_mboxes[0]
 			= I915_READ(RING_SYNC_0(ring->mmio_base));
-		error->semaphore_mboxes[ring->id][1]
+		ering->semaphore_mboxes[1]
 			= I915_READ(RING_SYNC_1(ring->mmio_base));
-		error->semaphore_seqno[ring->id][0] = ring->sync_seqno[0];
-		error->semaphore_seqno[ring->id][1] = ring->sync_seqno[1];
+		ering->semaphore_seqno[0] = ring->semaphore.sync_seqno[0];
+		ering->semaphore_seqno[1] = ring->semaphore.sync_seqno[1];
 	}
 
 	if (HAS_VEBOX(dev)) {
-		error->semaphore_mboxes[ring->id][2] =
+		ering->semaphore_mboxes[2] =
 			I915_READ(RING_SYNC_2(ring->mmio_base));
-		error->semaphore_seqno[ring->id][2] = ring->sync_seqno[2];
+		ering->semaphore_seqno[2] = ring->semaphore.sync_seqno[2];
 	}
 
 	if (INTEL_INFO(dev)->gen >= 4) {
-		error->faddr[ring->id] = I915_READ(RING_DMA_FADD(ring->mmio_base));
-		error->ipeir[ring->id] = I915_READ(RING_IPEIR(ring->mmio_base));
-		error->ipehr[ring->id] = I915_READ(RING_IPEHR(ring->mmio_base));
-		error->instdone[ring->id] = I915_READ(RING_INSTDONE(ring->mmio_base));
-		error->instps[ring->id] = I915_READ(RING_INSTPS(ring->mmio_base));
-		if (ring->id == RCS)
-			error->bbaddr = I915_READ64(BB_ADDR);
-		error->bbstate[ring->id] = I915_READ(RING_BBSTATE(ring->mmio_base));
+		ering->faddr = I915_READ(RING_DMA_FADD(ring->mmio_base));
+		ering->ipeir = I915_READ(RING_IPEIR(ring->mmio_base));
+		ering->ipehr = I915_READ(RING_IPEHR(ring->mmio_base));
+		ering->instdone = I915_READ(RING_INSTDONE(ring->mmio_base));
+		ering->instps = I915_READ(RING_INSTPS(ring->mmio_base));
+		ering->bbaddr = I915_READ(RING_BBADDR(ring->mmio_base));
+		if (INTEL_INFO(dev)->gen >= 8) {
+			ering->faddr |= (u64) I915_READ(RING_DMA_FADD_UDW(ring->mmio_base)) << 32;
+			ering->bbaddr |= (u64) I915_READ(RING_BBADDR_UDW(ring->mmio_base)) << 32;
+		}
+		ering->bbstate = I915_READ(RING_BBSTATE(ring->mmio_base));
 	} else {
-		error->faddr[ring->id] = I915_READ(DMA_FADD_I8XX);
-		error->ipeir[ring->id] = I915_READ(IPEIR);
-		error->ipehr[ring->id] = I915_READ(IPEHR);
-		error->instdone[ring->id] = I915_READ(INSTDONE);
+		ering->faddr = I915_READ(DMA_FADD_I8XX);
+		ering->ipeir = I915_READ(IPEIR);
+		ering->ipehr = I915_READ(IPEHR);
+		ering->instdone = I915_READ(INSTDONE);
+	}
+
+	ering->waiting = waitqueue_active(&ring->irq_queue);
+	ering->instpm = I915_READ(RING_INSTPM(ring->mmio_base));
+	ering->seqno = ring->get_seqno(ring, false);
+	ering->acthd = intel_ring_get_active_head(ring);
+	ering->head = I915_READ_HEAD(ring);
+	ering->tail = I915_READ_TAIL(ring);
+	ering->ctl = I915_READ_CTL(ring);
+
+	if (I915_NEED_GFX_HWS(dev)) {
+		int mmio;
+
+		if (IS_GEN7(dev)) {
+			switch (ring->id) {
+			default:
+			case RCS:
+				mmio = RENDER_HWS_PGA_GEN7;
+				break;
+			case BCS:
+				mmio = BLT_HWS_PGA_GEN7;
+				break;
+			case VCS:
+				mmio = BSD_HWS_PGA_GEN7;
+				break;
+			case VECS:
+				mmio = VEBOX_HWS_PGA_GEN7;
+				break;
+			}
+		} else if (IS_GEN6(ring->dev)) {
+			mmio = RING_HWS_PGA_GEN6(ring->mmio_base);
+		} else {
+			/* XXX: gen8 returns to sanity */
+			mmio = RING_HWS_PGA(ring->mmio_base);
+		}
+
+		ering->hws = I915_READ(mmio);
 	}
 
-	error->waiting[ring->id] = waitqueue_active(&ring->irq_queue);
-	error->instpm[ring->id] = I915_READ(RING_INSTPM(ring->mmio_base));
-	error->seqno[ring->id] = ring->get_seqno(ring, false);
-	error->acthd[ring->id] = intel_ring_get_active_head(ring);
-	error->head[ring->id] = I915_READ_HEAD(ring);
-	error->tail[ring->id] = I915_READ_TAIL(ring);
-	error->ctl[ring->id] = I915_READ_CTL(ring);
+	ering->cpu_ring_head = ring->buffer->head;
+	ering->cpu_ring_tail = ring->buffer->tail;
+
+	ering->hangcheck_score = ring->hangcheck.score;
+	ering->hangcheck_action = ring->hangcheck.action;
+
+	if (USES_PPGTT(dev)) {
+		int i;
 
-	error->cpu_ring_head[ring->id] = ring->head;
-	error->cpu_ring_tail[ring->id] = ring->tail;
+		ering->vm_info.gfx_mode = I915_READ(RING_MODE_GEN7(ring));
 
-	error->hangcheck_score[ring->id] = ring->hangcheck.score;
-	error->hangcheck_action[ring->id] = ring->hangcheck.action;
+		switch (INTEL_INFO(dev)->gen) {
+		case 8:
+			for (i = 0; i < 4; i++) {
+				ering->vm_info.pdp[i] =
+					I915_READ(GEN8_RING_PDP_UDW(ring, i));
+				ering->vm_info.pdp[i] <<= 32;
+				ering->vm_info.pdp[i] |=
+					I915_READ(GEN8_RING_PDP_LDW(ring, i));
+			}
+			break;
+		case 7:
+			ering->vm_info.pp_dir_base =
+				I915_READ(RING_PP_DIR_BASE(ring));
+			break;
+		case 6:
+			ering->vm_info.pp_dir_base =
+				I915_READ(RING_PP_DIR_BASE_READ(ring));
+			break;
+		}
+	}
 }
 
 
-static void i915_gem_record_active_context(struct intel_ring_buffer *ring,
+static void i915_gem_record_active_context(struct intel_engine_cs *ring,
 					   struct drm_i915_error_state *error,
 					   struct drm_i915_error_ring *ering)
 {
@@ -764,8 +872,7 @@ static void i915_gem_record_active_context(struct intel_ring_buffer *ring,
 
 	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
 		if ((error->ccid & PAGE_MASK) == i915_gem_obj_ggtt_offset(obj)) {
-			ering->ctx = i915_error_object_create_sized(dev_priv,
-								    obj, 1);
+			ering->ctx = i915_error_ggtt_object_create(dev_priv, obj);
 			break;
 		}
 	}
@@ -775,19 +882,60 @@ static void i915_gem_record_rings(struct drm_device *dev,
 				  struct drm_i915_error_state *error)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
 	struct drm_i915_gem_request *request;
 	int i, count;
 
-	for_each_ring(ring, dev_priv, i) {
-		i915_record_ring_state(dev, error, ring);
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		struct intel_engine_cs *ring = &dev_priv->ring[i];
+
+		error->ring[i].pid = -1;
+
+		if (ring->dev == NULL)
+			continue;
 
-		error->ring[i].batchbuffer =
-			i915_error_first_batchbuffer(dev_priv, ring);
+		error->ring[i].valid = true;
+
+		i915_record_ring_state(dev, ring, &error->ring[i]);
+
+		request = i915_gem_find_active_request(ring);
+		if (request) {
+			/* We need to copy these to an anonymous buffer
+			 * as the simplest method to avoid being overwritten
+			 * by userspace.
+			 */
+			error->ring[i].batchbuffer =
+				i915_error_object_create(dev_priv,
+							 request->batch_obj,
+							 request->ctx ?
+							 request->ctx->vm :
+							 &dev_priv->gtt.base);
+
+			if (HAS_BROKEN_CS_TLB(dev_priv->dev) &&
+			    ring->scratch.obj)
+				error->ring[i].wa_batchbuffer =
+					i915_error_ggtt_object_create(dev_priv,
+							     ring->scratch.obj);
+
+			if (request->file_priv) {
+				struct task_struct *task;
+
+				rcu_read_lock();
+				task = pid_task(request->file_priv->file->pid,
+						PIDTYPE_PID);
+				if (task) {
+					strcpy(error->ring[i].comm, task->comm);
+					error->ring[i].pid = task->pid;
+				}
+				rcu_read_unlock();
+			}
+		}
 
 		error->ring[i].ringbuffer =
-			i915_error_object_create(dev_priv, ring->obj);
+			i915_error_ggtt_object_create(dev_priv, ring->buffer->obj);
 
+		if (ring->status_page.obj)
+			error->ring[i].hws_page =
+				i915_error_ggtt_object_create(dev_priv, ring->status_page.obj);
 
 		i915_gem_record_active_context(ring, error, &error->ring[i]);
 
@@ -834,7 +982,7 @@ static void i915_gem_capture_vm(struct drm_i915_private *dev_priv,
 		i++;
 	error->active_bo_count[ndx] = i;
 	list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list)
-		if (obj->pin_count)
+		if (i915_gem_obj_is_pinned(obj))
 			i++;
 	error->pinned_bo_count[ndx] = i - error->active_bo_count[ndx];
 
@@ -868,11 +1016,6 @@ static void i915_gem_capture_buffers(struct drm_i915_private *dev_priv,
 	list_for_each_entry(vm, &dev_priv->vm_list, global_link)
 		cnt++;
 
-	if (WARN(cnt > 1, "Multiple VMs not yet supported\n"))
-		cnt = 1;
-
-	vm = &dev_priv->gtt.base;
-
 	error->active_bo = kcalloc(cnt, sizeof(*error->active_bo), GFP_ATOMIC);
 	error->pinned_bo = kcalloc(cnt, sizeof(*error->pinned_bo), GFP_ATOMIC);
 	error->active_bo_count = kcalloc(cnt, sizeof(*error->active_bo_count),
@@ -884,6 +1027,105 @@ static void i915_gem_capture_buffers(struct drm_i915_private *dev_priv,
 		i915_gem_capture_vm(dev_priv, error, vm, i++);
 }
 
+/* Capture all registers which don't fit into another category. */
+static void i915_capture_reg_state(struct drm_i915_private *dev_priv,
+				   struct drm_i915_error_state *error)
+{
+	struct drm_device *dev = dev_priv->dev;
+
+	/* General organization
+	 * 1. Registers specific to a single generation
+	 * 2. Registers which belong to multiple generations
+	 * 3. Feature specific registers.
+	 * 4. Everything else
+	 * Please try to follow the order.
+	 */
+
+	/* 1: Registers specific to a single generation */
+	if (IS_VALLEYVIEW(dev)) {
+		error->ier = I915_READ(GTIER) | I915_READ(VLV_IER);
+		error->forcewake = I915_READ(FORCEWAKE_VLV);
+	}
+
+	if (IS_GEN7(dev))
+		error->err_int = I915_READ(GEN7_ERR_INT);
+
+	if (IS_GEN6(dev)) {
+		error->forcewake = I915_READ(FORCEWAKE);
+		error->gab_ctl = I915_READ(GAB_CTL);
+		error->gfx_mode = I915_READ(GFX_MODE);
+	}
+
+	/* 2: Registers which belong to multiple generations */
+	if (INTEL_INFO(dev)->gen >= 7)
+		error->forcewake = I915_READ(FORCEWAKE_MT);
+
+	if (INTEL_INFO(dev)->gen >= 6) {
+		error->derrmr = I915_READ(DERRMR);
+		error->error = I915_READ(ERROR_GEN6);
+		error->done_reg = I915_READ(DONE_REG);
+	}
+
+	/* 3: Feature specific registers */
+	if (IS_GEN6(dev) || IS_GEN7(dev)) {
+		error->gam_ecochk = I915_READ(GAM_ECOCHK);
+		error->gac_eco = I915_READ(GAC_ECO_BITS);
+	}
+
+	/* 4: Everything else */
+	if (HAS_HW_CONTEXTS(dev))
+		error->ccid = I915_READ(CCID);
+
+	if (HAS_PCH_SPLIT(dev))
+		error->ier = I915_READ(DEIER) | I915_READ(GTIER);
+	else {
+		if (IS_GEN2(dev))
+			error->ier = I915_READ16(IER);
+		else
+			error->ier = I915_READ(IER);
+	}
+
+	/* 4: Everything else */
+	error->eir = I915_READ(EIR);
+	error->pgtbl_er = I915_READ(PGTBL_ER);
+
+	i915_get_extra_instdone(dev, error->extra_instdone);
+}
+
+static void i915_error_capture_msg(struct drm_device *dev,
+				   struct drm_i915_error_state *error,
+				   bool wedged,
+				   const char *error_msg)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 ecode;
+	int ring_id = -1, len;
+
+	ecode = i915_error_generate_code(dev_priv, error, &ring_id);
+
+	len = scnprintf(error->error_msg, sizeof(error->error_msg),
+			"GPU HANG: ecode %d:0x%08x", ring_id, ecode);
+
+	if (ring_id != -1 && error->ring[ring_id].pid != -1)
+		len += scnprintf(error->error_msg + len,
+				 sizeof(error->error_msg) - len,
+				 ", in %s [%d]",
+				 error->ring[ring_id].comm,
+				 error->ring[ring_id].pid);
+
+	scnprintf(error->error_msg + len, sizeof(error->error_msg) - len,
+		  ", reason: %s, action: %s",
+		  error_msg,
+		  wedged ? "reset" : "continue");
+}
+
+static void i915_capture_gen_state(struct drm_i915_private *dev_priv,
+				   struct drm_i915_error_state *error)
+{
+	error->reset_count = i915_reset_count(&dev_priv->gpu_error);
+	error->suspend_count = dev_priv->suspend_count;
+}
+
 /**
  * i915_capture_error_state - capture an error record for later analysis
  * @dev: drm device
@@ -893,18 +1135,13 @@ static void i915_gem_capture_buffers(struct drm_i915_private *dev_priv,
  * out a structure which becomes available in debugfs for user level tools
  * to pick up.
  */
-void i915_capture_error_state(struct drm_device *dev)
+void i915_capture_error_state(struct drm_device *dev, bool wedged,
+			      const char *error_msg)
 {
+	static bool warned;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_error_state *error;
 	unsigned long flags;
-	int pipe;
-
-	spin_lock_irqsave(&dev_priv->gpu_error.lock, flags);
-	error = dev_priv->gpu_error.first_error;
-	spin_unlock_irqrestore(&dev_priv->gpu_error.lock, flags);
-	if (error)
-		return;
 
 	/* Account for pipe specific data like PIPE*STAT */
 	error = kzalloc(sizeof(*error), GFP_ATOMIC);
@@ -913,52 +1150,10 @@ void i915_capture_error_state(struct drm_device *dev)
 		return;
 	}
 
-	DRM_INFO("GPU crash dump saved to /sys/class/drm/card%d/error\n",
-		 dev->primary->index);
-	DRM_INFO("GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.\n");
-	DRM_INFO("Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel\n");
-	DRM_INFO("drm/i915 developers can then reassign to the right component if it's not a kernel issue.\n");
-	DRM_INFO("The gpu crash dump is required to analyze gpu hangs, so please always attach it.\n");
-
 	kref_init(&error->ref);
-	error->eir = I915_READ(EIR);
-	error->pgtbl_er = I915_READ(PGTBL_ER);
-	if (HAS_HW_CONTEXTS(dev))
-		error->ccid = I915_READ(CCID);
-
-	if (HAS_PCH_SPLIT(dev))
-		error->ier = I915_READ(DEIER) | I915_READ(GTIER);
-	else if (IS_VALLEYVIEW(dev))
-		error->ier = I915_READ(GTIER) | I915_READ(VLV_IER);
-	else if (IS_GEN2(dev))
-		error->ier = I915_READ16(IER);
-	else
-		error->ier = I915_READ(IER);
-
-	if (INTEL_INFO(dev)->gen >= 6)
-		error->derrmr = I915_READ(DERRMR);
-
-	if (IS_VALLEYVIEW(dev))
-		error->forcewake = I915_READ(FORCEWAKE_VLV);
-	else if (INTEL_INFO(dev)->gen >= 7)
-		error->forcewake = I915_READ(FORCEWAKE_MT);
-	else if (INTEL_INFO(dev)->gen == 6)
-		error->forcewake = I915_READ(FORCEWAKE);
-
-	if (!HAS_PCH_SPLIT(dev))
-		for_each_pipe(pipe)
-			error->pipestat[pipe] = I915_READ(PIPESTAT(pipe));
-
-	if (INTEL_INFO(dev)->gen >= 6) {
-		error->error = I915_READ(ERROR_GEN6);
-		error->done_reg = I915_READ(DONE_REG);
-	}
-
-	if (INTEL_INFO(dev)->gen == 7)
-		error->err_int = I915_READ(GEN7_ERR_INT);
-
-	i915_get_extra_instdone(dev, error->extra_instdone);
 
+	i915_capture_gen_state(dev_priv, error);
+	i915_capture_reg_state(dev_priv, error);
 	i915_gem_capture_buffers(dev_priv, error);
 	i915_gem_record_fences(dev, error);
 	i915_gem_record_rings(dev, error);
@@ -968,6 +1163,9 @@ void i915_capture_error_state(struct drm_device *dev)
 	error->overlay = intel_overlay_capture_error_state(dev);
 	error->display = intel_display_capture_error_state(dev);
 
+	i915_error_capture_msg(dev, error, wedged, error_msg);
+	DRM_INFO("%s\n", error->error_msg);
+
 	spin_lock_irqsave(&dev_priv->gpu_error.lock, flags);
 	if (dev_priv->gpu_error.first_error == NULL) {
 		dev_priv->gpu_error.first_error = error;
@@ -975,8 +1173,19 @@ void i915_capture_error_state(struct drm_device *dev)
 	}
 	spin_unlock_irqrestore(&dev_priv->gpu_error.lock, flags);
 
-	if (error)
+	if (error) {
 		i915_error_state_free(&error->ref);
+		return;
+	}
+
+	if (!warned) {
+		DRM_INFO("GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.\n");
+		DRM_INFO("Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel\n");
+		DRM_INFO("drm/i915 developers can then reassign to the right component if it's not a kernel issue.\n");
+		DRM_INFO("The gpu crash dump is required to analyze gpu hangs, so please always attach it.\n");
+		DRM_INFO("GPU crash dump saved to /sys/class/drm/card%d/error\n", dev->primary->index);
+		warned = true;
+	}
 }
 
 void i915_error_state_get(struct drm_device *dev,
diff --git a/drivers/gpu/drm/i915/i915_ioc32.c b/drivers/gpu/drm/i915/i915_ioc32.c
index 3c59584161c..2e0613e2625 100644
--- a/drivers/gpu/drm/i915/i915_ioc32.c
+++ b/drivers/gpu/drm/i915/i915_ioc32.c
@@ -208,7 +208,7 @@ long i915_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	if (nr < DRM_COMMAND_BASE)
 		return drm_compat_ioctl(filp, cmd, arg);
 
-	if (nr < DRM_COMMAND_BASE + DRM_ARRAY_SIZE(i915_compat_ioctls))
+	if (nr < DRM_COMMAND_BASE + ARRAY_SIZE(i915_compat_ioctls))
 		fn = i915_compat_ioctls[nr - DRM_COMMAND_BASE];
 
 	if (fn != NULL)
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index f13d5edc39d..c05c84f3f09 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -62,7 +62,7 @@ static const u32 hpd_mask_i915[] = {
 	[HPD_PORT_D] = PORTD_HOTPLUG_INT_EN
 };
 
-static const u32 hpd_status_gen4[] = {
+static const u32 hpd_status_g4x[] = {
 	[HPD_CRT] = CRT_HOTPLUG_INT_STATUS,
 	[HPD_SDVO_B] = SDVOB_HOTPLUG_INT_STATUS_G4X,
 	[HPD_SDVO_C] = SDVOC_HOTPLUG_INT_STATUS_G4X,
@@ -80,17 +80,64 @@ static const u32 hpd_status_i915[] = { /* i915 and valleyview are the same */
 	[HPD_PORT_D] = PORTD_HOTPLUG_INT_STATUS
 };
 
+/* IIR can theoretically queue up two events. Be paranoid. */
+#define GEN8_IRQ_RESET_NDX(type, which) do { \
+	I915_WRITE(GEN8_##type##_IMR(which), 0xffffffff); \
+	POSTING_READ(GEN8_##type##_IMR(which)); \
+	I915_WRITE(GEN8_##type##_IER(which), 0); \
+	I915_WRITE(GEN8_##type##_IIR(which), 0xffffffff); \
+	POSTING_READ(GEN8_##type##_IIR(which)); \
+	I915_WRITE(GEN8_##type##_IIR(which), 0xffffffff); \
+	POSTING_READ(GEN8_##type##_IIR(which)); \
+} while (0)
+
+#define GEN5_IRQ_RESET(type) do { \
+	I915_WRITE(type##IMR, 0xffffffff); \
+	POSTING_READ(type##IMR); \
+	I915_WRITE(type##IER, 0); \
+	I915_WRITE(type##IIR, 0xffffffff); \
+	POSTING_READ(type##IIR); \
+	I915_WRITE(type##IIR, 0xffffffff); \
+	POSTING_READ(type##IIR); \
+} while (0)
+
+/*
+ * We should clear IMR at preinstall/uninstall, and just check at postinstall.
+ */
+#define GEN5_ASSERT_IIR_IS_ZERO(reg) do { \
+	u32 val = I915_READ(reg); \
+	if (val) { \
+		WARN(1, "Interrupt register 0x%x is not zero: 0x%08x\n", \
+		     (reg), val); \
+		I915_WRITE((reg), 0xffffffff); \
+		POSTING_READ(reg); \
+		I915_WRITE((reg), 0xffffffff); \
+		POSTING_READ(reg); \
+	} \
+} while (0)
+
+#define GEN8_IRQ_INIT_NDX(type, which, imr_val, ier_val) do { \
+	GEN5_ASSERT_IIR_IS_ZERO(GEN8_##type##_IIR(which)); \
+	I915_WRITE(GEN8_##type##_IMR(which), (imr_val)); \
+	I915_WRITE(GEN8_##type##_IER(which), (ier_val)); \
+	POSTING_READ(GEN8_##type##_IER(which)); \
+} while (0)
+
+#define GEN5_IRQ_INIT(type, imr_val, ier_val) do { \
+	GEN5_ASSERT_IIR_IS_ZERO(type##IIR); \
+	I915_WRITE(type##IMR, (imr_val)); \
+	I915_WRITE(type##IER, (ier_val)); \
+	POSTING_READ(type##IER); \
+} while (0)
+
 /* For display hotplug interrupt */
 static void
-ironlake_enable_display_irq(drm_i915_private_t *dev_priv, u32 mask)
+ironlake_enable_display_irq(struct drm_i915_private *dev_priv, u32 mask)
 {
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if (dev_priv->pc8.irqs_disabled) {
-		WARN(1, "IRQs disabled\n");
-		dev_priv->pc8.regsave.deimr &= ~mask;
+	if (WARN_ON(dev_priv->pm.irqs_disabled))
 		return;
-	}
 
 	if ((dev_priv->irq_mask & mask) != 0) {
 		dev_priv->irq_mask &= ~mask;
@@ -100,15 +147,12 @@ ironlake_enable_display_irq(drm_i915_private_t *dev_priv, u32 mask)
 }
 
 static void
-ironlake_disable_display_irq(drm_i915_private_t *dev_priv, u32 mask)
+ironlake_disable_display_irq(struct drm_i915_private *dev_priv, u32 mask)
 {
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if (dev_priv->pc8.irqs_disabled) {
-		WARN(1, "IRQs disabled\n");
-		dev_priv->pc8.regsave.deimr |= mask;
+	if (WARN_ON(dev_priv->pm.irqs_disabled))
 		return;
-	}
 
 	if ((dev_priv->irq_mask & mask) != mask) {
 		dev_priv->irq_mask |= mask;
@@ -129,13 +173,8 @@ static void ilk_update_gt_irq(struct drm_i915_private *dev_priv,
 {
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if (dev_priv->pc8.irqs_disabled) {
-		WARN(1, "IRQs disabled\n");
-		dev_priv->pc8.regsave.gtimr &= ~interrupt_mask;
-		dev_priv->pc8.regsave.gtimr |= (~enabled_irq_mask &
-						interrupt_mask);
+	if (WARN_ON(dev_priv->pm.irqs_disabled))
 		return;
-	}
 
 	dev_priv->gt_irq_mask &= ~interrupt_mask;
 	dev_priv->gt_irq_mask |= (~enabled_irq_mask & interrupt_mask);
@@ -167,13 +206,8 @@ static void snb_update_pm_irq(struct drm_i915_private *dev_priv,
 
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if (dev_priv->pc8.irqs_disabled) {
-		WARN(1, "IRQs disabled\n");
-		dev_priv->pc8.regsave.gen6_pmimr &= ~interrupt_mask;
-		dev_priv->pc8.regsave.gen6_pmimr |= (~enabled_irq_mask &
-						     interrupt_mask);
+	if (WARN_ON(dev_priv->pm.irqs_disabled))
 		return;
-	}
 
 	new_val = dev_priv->pm_irq_mask;
 	new_val &= ~interrupt_mask;
@@ -214,6 +248,46 @@ static bool ivb_can_enable_err_int(struct drm_device *dev)
 	return true;
 }
 
+/**
+  * bdw_update_pm_irq - update GT interrupt 2
+  * @dev_priv: driver private
+  * @interrupt_mask: mask of interrupt bits to update
+  * @enabled_irq_mask: mask of interrupt bits to enable
+  *
+  * Copied from the snb function, updated with relevant register offsets
+  */
+static void bdw_update_pm_irq(struct drm_i915_private *dev_priv,
+			      uint32_t interrupt_mask,
+			      uint32_t enabled_irq_mask)
+{
+	uint32_t new_val;
+
+	assert_spin_locked(&dev_priv->irq_lock);
+
+	if (WARN_ON(dev_priv->pm.irqs_disabled))
+		return;
+
+	new_val = dev_priv->pm_irq_mask;
+	new_val &= ~interrupt_mask;
+	new_val |= (~enabled_irq_mask & interrupt_mask);
+
+	if (new_val != dev_priv->pm_irq_mask) {
+		dev_priv->pm_irq_mask = new_val;
+		I915_WRITE(GEN8_GT_IMR(2), dev_priv->pm_irq_mask);
+		POSTING_READ(GEN8_GT_IMR(2));
+	}
+}
+
+void bdw_enable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask)
+{
+	bdw_update_pm_irq(dev_priv, mask, mask);
+}
+
+void bdw_disable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask)
+{
+	bdw_update_pm_irq(dev_priv, mask, 0);
+}
+
 static bool cpt_can_enable_serr_int(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -232,6 +306,53 @@ static bool cpt_can_enable_serr_int(struct drm_device *dev)
 	return true;
 }
 
+void i9xx_check_fifo_underruns(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *crtc;
+	unsigned long flags;
+
+	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+
+	for_each_intel_crtc(dev, crtc) {
+		u32 reg = PIPESTAT(crtc->pipe);
+		u32 pipestat;
+
+		if (crtc->cpu_fifo_underrun_disabled)
+			continue;
+
+		pipestat = I915_READ(reg) & 0xffff0000;
+		if ((pipestat & PIPE_FIFO_UNDERRUN_STATUS) == 0)
+			continue;
+
+		I915_WRITE(reg, pipestat | PIPE_FIFO_UNDERRUN_STATUS);
+		POSTING_READ(reg);
+
+		DRM_ERROR("pipe %c underrun\n", pipe_name(crtc->pipe));
+	}
+
+	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+}
+
+static void i9xx_set_fifo_underrun_reporting(struct drm_device *dev,
+					     enum pipe pipe,
+					     bool enable, bool old)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 reg = PIPESTAT(pipe);
+	u32 pipestat = I915_READ(reg) & 0xffff0000;
+
+	assert_spin_locked(&dev_priv->irq_lock);
+
+	if (enable) {
+		I915_WRITE(reg, pipestat | PIPE_FIFO_UNDERRUN_STATUS);
+		POSTING_READ(reg);
+	} else {
+		if (old && pipestat & PIPE_FIFO_UNDERRUN_STATUS)
+			DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
+	}
+}
+
 static void ironlake_set_fifo_underrun_reporting(struct drm_device *dev,
 						 enum pipe pipe, bool enable)
 {
@@ -246,7 +367,8 @@ static void ironlake_set_fifo_underrun_reporting(struct drm_device *dev,
 }
 
 static void ivybridge_set_fifo_underrun_reporting(struct drm_device *dev,
-						  enum pipe pipe, bool enable)
+						  enum pipe pipe,
+						  bool enable, bool old)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	if (enable) {
@@ -257,15 +379,12 @@ static void ivybridge_set_fifo_underrun_reporting(struct drm_device *dev,
 
 		ironlake_enable_display_irq(dev_priv, DE_ERR_INT_IVB);
 	} else {
-		bool was_enabled = !(I915_READ(DEIMR) & DE_ERR_INT_IVB);
-
-		/* Change the state _after_ we've read out the current one. */
 		ironlake_disable_display_irq(dev_priv, DE_ERR_INT_IVB);
 
-		if (!was_enabled &&
-		    (I915_READ(GEN7_ERR_INT) & ERR_INT_FIFO_UNDERRUN(pipe))) {
-			DRM_DEBUG_KMS("uncleared fifo underrun on pipe %c\n",
-				      pipe_name(pipe));
+		if (old &&
+		    I915_READ(GEN7_ERR_INT) & ERR_INT_FIFO_UNDERRUN(pipe)) {
+			DRM_ERROR("uncleared fifo underrun on pipe %c\n",
+				  pipe_name(pipe));
 		}
 	}
 }
@@ -301,14 +420,8 @@ static void ibx_display_interrupt_update(struct drm_i915_private *dev_priv,
 
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if (dev_priv->pc8.irqs_disabled &&
-	    (interrupt_mask & SDE_HOTPLUG_MASK_CPT)) {
-		WARN(1, "IRQs disabled\n");
-		dev_priv->pc8.regsave.sdeimr &= ~interrupt_mask;
-		dev_priv->pc8.regsave.sdeimr |= (~enabled_irq_mask &
-						 interrupt_mask);
+	if (WARN_ON(dev_priv->pm.irqs_disabled))
 		return;
-	}
 
 	I915_WRITE(SDEIMR, sdeimr);
 	POSTING_READ(SDEIMR);
@@ -334,7 +447,7 @@ static void ibx_set_fifo_underrun_reporting(struct drm_device *dev,
 
 static void cpt_set_fifo_underrun_reporting(struct drm_device *dev,
 					    enum transcoder pch_transcoder,
-					    bool enable)
+					    bool enable, bool old)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
@@ -347,16 +460,12 @@ static void cpt_set_fifo_underrun_reporting(struct drm_device *dev,
 
 		ibx_enable_display_interrupt(dev_priv, SDE_ERROR_CPT);
 	} else {
-		uint32_t tmp = I915_READ(SERR_INT);
-		bool was_enabled = !(I915_READ(SDEIMR) & SDE_ERROR_CPT);
-
-		/* Change the state _after_ we've read out the current one. */
 		ibx_disable_display_interrupt(dev_priv, SDE_ERROR_CPT);
 
-		if (!was_enabled &&
-		    (tmp & SERR_INT_TRANS_FIFO_UNDERRUN(pch_transcoder))) {
-			DRM_DEBUG_KMS("uncleared pch fifo underrun on pch transcoder %c\n",
-				      transcoder_name(pch_transcoder));
+		if (old && I915_READ(SERR_INT) &
+		    SERR_INT_TRANS_FIFO_UNDERRUN(pch_transcoder)) {
+			DRM_ERROR("uncleared pch fifo underrun on pch transcoder %c\n",
+				  transcoder_name(pch_transcoder));
 		}
 	}
 }
@@ -375,36 +484,55 @@ static void cpt_set_fifo_underrun_reporting(struct drm_device *dev,
  *
  * Returns the previous state of underrun reporting.
  */
-bool intel_set_cpu_fifo_underrun_reporting(struct drm_device *dev,
-					   enum pipe pipe, bool enable)
+static bool __intel_set_cpu_fifo_underrun_reporting(struct drm_device *dev,
+						    enum pipe pipe, bool enable)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	unsigned long flags;
-	bool ret;
-
-	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+	bool old;
 
-	ret = !intel_crtc->cpu_fifo_underrun_disabled;
-
-	if (enable == ret)
-		goto done;
+	assert_spin_locked(&dev_priv->irq_lock);
 
+	old = !intel_crtc->cpu_fifo_underrun_disabled;
 	intel_crtc->cpu_fifo_underrun_disabled = !enable;
 
-	if (IS_GEN5(dev) || IS_GEN6(dev))
+	if (INTEL_INFO(dev)->gen < 5 || IS_VALLEYVIEW(dev))
+		i9xx_set_fifo_underrun_reporting(dev, pipe, enable, old);
+	else if (IS_GEN5(dev) || IS_GEN6(dev))
 		ironlake_set_fifo_underrun_reporting(dev, pipe, enable);
 	else if (IS_GEN7(dev))
-		ivybridge_set_fifo_underrun_reporting(dev, pipe, enable);
+		ivybridge_set_fifo_underrun_reporting(dev, pipe, enable, old);
 	else if (IS_GEN8(dev))
 		broadwell_set_fifo_underrun_reporting(dev, pipe, enable);
 
-done:
+	return old;
+}
+
+bool intel_set_cpu_fifo_underrun_reporting(struct drm_device *dev,
+					   enum pipe pipe, bool enable)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long flags;
+	bool ret;
+
+	spin_lock_irqsave(&dev_priv->irq_lock, flags);
+	ret = __intel_set_cpu_fifo_underrun_reporting(dev, pipe, enable);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+
 	return ret;
 }
 
+static bool __cpu_fifo_underrun_reporting_enabled(struct drm_device *dev,
+						  enum pipe pipe)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+
+	return !intel_crtc->cpu_fifo_underrun_disabled;
+}
+
 /**
  * intel_set_pch_fifo_underrun_reporting - enable/disable FIFO underrun messages
  * @dev: drm device
@@ -427,7 +555,7 @@ bool intel_set_pch_fifo_underrun_reporting(struct drm_device *dev,
 	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pch_transcoder];
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	unsigned long flags;
-	bool ret;
+	bool old;
 
 	/*
 	 * NOTE: Pre-LPT has a fixed cpu pipe -> pch transcoder mapping, but LPT
@@ -440,63 +568,132 @@ bool intel_set_pch_fifo_underrun_reporting(struct drm_device *dev,
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
 
-	ret = !intel_crtc->pch_fifo_underrun_disabled;
-
-	if (enable == ret)
-		goto done;
-
+	old = !intel_crtc->pch_fifo_underrun_disabled;
 	intel_crtc->pch_fifo_underrun_disabled = !enable;
 
 	if (HAS_PCH_IBX(dev))
 		ibx_set_fifo_underrun_reporting(dev, pch_transcoder, enable);
 	else
-		cpt_set_fifo_underrun_reporting(dev, pch_transcoder, enable);
+		cpt_set_fifo_underrun_reporting(dev, pch_transcoder, enable, old);
 
-done:
 	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
-	return ret;
+	return old;
 }
 
 
-void
-i915_enable_pipestat(drm_i915_private_t *dev_priv, enum pipe pipe, u32 mask)
+static void
+__i915_enable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
+		       u32 enable_mask, u32 status_mask)
 {
 	u32 reg = PIPESTAT(pipe);
-	u32 pipestat = I915_READ(reg) & 0x7fff0000;
+	u32 pipestat = I915_READ(reg) & PIPESTAT_INT_ENABLE_MASK;
 
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if ((pipestat & mask) == mask)
+	if (WARN_ONCE(enable_mask & ~PIPESTAT_INT_ENABLE_MASK ||
+		      status_mask & ~PIPESTAT_INT_STATUS_MASK,
+		      "pipe %c: enable_mask=0x%x, status_mask=0x%x\n",
+		      pipe_name(pipe), enable_mask, status_mask))
 		return;
 
+	if ((pipestat & enable_mask) == enable_mask)
+		return;
+
+	dev_priv->pipestat_irq_mask[pipe] |= status_mask;
+
 	/* Enable the interrupt, clear any pending status */
-	pipestat |= mask | (mask >> 16);
+	pipestat |= enable_mask | status_mask;
 	I915_WRITE(reg, pipestat);
 	POSTING_READ(reg);
 }
 
-void
-i915_disable_pipestat(drm_i915_private_t *dev_priv, enum pipe pipe, u32 mask)
+static void
+__i915_disable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
+		        u32 enable_mask, u32 status_mask)
 {
 	u32 reg = PIPESTAT(pipe);
-	u32 pipestat = I915_READ(reg) & 0x7fff0000;
+	u32 pipestat = I915_READ(reg) & PIPESTAT_INT_ENABLE_MASK;
 
 	assert_spin_locked(&dev_priv->irq_lock);
 
-	if ((pipestat & mask) == 0)
+	if (WARN_ONCE(enable_mask & ~PIPESTAT_INT_ENABLE_MASK ||
+		      status_mask & ~PIPESTAT_INT_STATUS_MASK,
+		      "pipe %c: enable_mask=0x%x, status_mask=0x%x\n",
+		      pipe_name(pipe), enable_mask, status_mask))
+		return;
+
+	if ((pipestat & enable_mask) == 0)
 		return;
 
-	pipestat &= ~mask;
+	dev_priv->pipestat_irq_mask[pipe] &= ~status_mask;
+
+	pipestat &= ~enable_mask;
 	I915_WRITE(reg, pipestat);
 	POSTING_READ(reg);
 }
 
+static u32 vlv_get_pipestat_enable_mask(struct drm_device *dev, u32 status_mask)
+{
+	u32 enable_mask = status_mask << 16;
+
+	/*
+	 * On pipe A we don't support the PSR interrupt yet,
+	 * on pipe B and C the same bit MBZ.
+	 */
+	if (WARN_ON_ONCE(status_mask & PIPE_A_PSR_STATUS_VLV))
+		return 0;
+	/*
+	 * On pipe B and C we don't support the PSR interrupt yet, on pipe
+	 * A the same bit is for perf counters which we don't use either.
+	 */
+	if (WARN_ON_ONCE(status_mask & PIPE_B_PSR_STATUS_VLV))
+		return 0;
+
+	enable_mask &= ~(PIPE_FIFO_UNDERRUN_STATUS |
+			 SPRITE0_FLIP_DONE_INT_EN_VLV |
+			 SPRITE1_FLIP_DONE_INT_EN_VLV);
+	if (status_mask & SPRITE0_FLIP_DONE_INT_STATUS_VLV)
+		enable_mask |= SPRITE0_FLIP_DONE_INT_EN_VLV;
+	if (status_mask & SPRITE1_FLIP_DONE_INT_STATUS_VLV)
+		enable_mask |= SPRITE1_FLIP_DONE_INT_EN_VLV;
+
+	return enable_mask;
+}
+
+void
+i915_enable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
+		     u32 status_mask)
+{
+	u32 enable_mask;
+
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		enable_mask = vlv_get_pipestat_enable_mask(dev_priv->dev,
+							   status_mask);
+	else
+		enable_mask = status_mask << 16;
+	__i915_enable_pipestat(dev_priv, pipe, enable_mask, status_mask);
+}
+
+void
+i915_disable_pipestat(struct drm_i915_private *dev_priv, enum pipe pipe,
+		      u32 status_mask)
+{
+	u32 enable_mask;
+
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		enable_mask = vlv_get_pipestat_enable_mask(dev_priv->dev,
+							   status_mask);
+	else
+		enable_mask = status_mask << 16;
+	__i915_disable_pipestat(dev_priv, pipe, enable_mask, status_mask);
+}
+
 /**
  * i915_enable_asle_pipestat - enable ASLE pipestat for OpRegion
  */
 static void i915_enable_asle_pipestat(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
 
 	if (!dev_priv->opregion.asle || !IS_MOBILE(dev))
@@ -504,10 +701,10 @@ static void i915_enable_asle_pipestat(struct drm_device *dev)
 
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 
-	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_LEGACY_BLC_EVENT_ENABLE);
+	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_LEGACY_BLC_EVENT_STATUS);
 	if (INTEL_INFO(dev)->gen >= 4)
 		i915_enable_pipestat(dev_priv, PIPE_A,
-				     PIPE_LEGACY_BLC_EVENT_ENABLE);
+				     PIPE_LEGACY_BLC_EVENT_STATUS);
 
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 }
@@ -524,7 +721,7 @@ static void i915_enable_asle_pipestat(struct drm_device *dev)
 static int
 i915_pipe_enabled(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		/* Locking is horribly broken here, but whatever. */
@@ -537,6 +734,56 @@ i915_pipe_enabled(struct drm_device *dev, int pipe)
 	}
 }
 
+/*
+ * This timing diagram depicts the video signal in and
+ * around the vertical blanking period.
+ *
+ * Assumptions about the fictitious mode used in this example:
+ *  vblank_start >= 3
+ *  vsync_start = vblank_start + 1
+ *  vsync_end = vblank_start + 2
+ *  vtotal = vblank_start + 3
+ *
+ *           start of vblank:
+ *           latch double buffered registers
+ *           increment frame counter (ctg+)
+ *           generate start of vblank interrupt (gen4+)
+ *           |
+ *           |          frame start:
+ *           |          generate frame start interrupt (aka. vblank interrupt) (gmch)
+ *           |          may be shifted forward 1-3 extra lines via PIPECONF
+ *           |          |
+ *           |          |  start of vsync:
+ *           |          |  generate vsync interrupt
+ *           |          |  |
+ * ___xxxx___    ___xxxx___    ___xxxx___    ___xxxx___    ___xxxx___    ___xxxx
+ *       .   \hs/   .      \hs/          \hs/          \hs/   .      \hs/
+ * ----va---> <-----------------vb--------------------> <--------va-------------
+ *       |          |       <----vs----->                     |
+ * -vbs-----> <---vbs+1---> <---vbs+2---> <-----0-----> <-----1-----> <-----2--- (scanline counter gen2)
+ * -vbs-2---> <---vbs-1---> <---vbs-----> <---vbs+1---> <---vbs+2---> <-----0--- (scanline counter gen3+)
+ * -vbs-2---> <---vbs-2---> <---vbs-1---> <---vbs-----> <---vbs+1---> <---vbs+2- (scanline counter hsw+ hdmi)
+ *       |          |                                         |
+ *       last visible pixel                                   first visible pixel
+ *                  |                                         increment frame counter (gen3/4)
+ *                  pixel counter = vblank_start * htotal     pixel counter = 0 (gen3/4)
+ *
+ * x  = horizontal active
+ * _  = horizontal blanking
+ * hs = horizontal sync
+ * va = vertical active
+ * vb = vertical blanking
+ * vs = vertical sync
+ * vbs = vblank_start (number)
+ *
+ * Summary:
+ * - most events happen at the start of horizontal sync
+ * - frame start happens at the start of horizontal blank, 1-4 lines
+ *   (depending on PIPECONF settings) after the start of vblank
+ * - gen3/4 pixel and frame counter are synchronized with the start
+ *   of horizontal active on the first line of vertical active
+ */
+
 static u32 i8xx_get_vblank_counter(struct drm_device *dev, int pipe)
 {
 	/* Gen2 doesn't have a hardware frame counter */
@@ -548,10 +795,10 @@ static u32 i8xx_get_vblank_counter(struct drm_device *dev, int pipe)
  */
 static u32 i915_get_vblank_counter(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long high_frame;
 	unsigned long low_frame;
-	u32 high1, high2, low, pixel, vbl_start;
+	u32 high1, high2, low, pixel, vbl_start, hsync_start, htotal;
 
 	if (!i915_pipe_enabled(dev, pipe)) {
 		DRM_DEBUG_DRIVER("trying to get vblank count for disabled "
@@ -565,18 +812,28 @@ static u32 i915_get_vblank_counter(struct drm_device *dev, int pipe)
 		const struct drm_display_mode *mode =
 			&intel_crtc->config.adjusted_mode;
 
-		vbl_start = mode->crtc_vblank_start * mode->crtc_htotal;
+		htotal = mode->crtc_htotal;
+		hsync_start = mode->crtc_hsync_start;
+		vbl_start = mode->crtc_vblank_start;
+		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+			vbl_start = DIV_ROUND_UP(vbl_start, 2);
 	} else {
-		enum transcoder cpu_transcoder =
-			intel_pipe_to_cpu_transcoder(dev_priv, pipe);
-		u32 htotal;
+		enum transcoder cpu_transcoder = (enum transcoder) pipe;
 
 		htotal = ((I915_READ(HTOTAL(cpu_transcoder)) >> 16) & 0x1fff) + 1;
+		hsync_start = (I915_READ(HSYNC(cpu_transcoder))  & 0x1fff) + 1;
 		vbl_start = (I915_READ(VBLANK(cpu_transcoder)) & 0x1fff) + 1;
-
-		vbl_start *= htotal;
+		if ((I915_READ(PIPECONF(cpu_transcoder)) &
+		     PIPECONF_INTERLACE_MASK) != PIPECONF_PROGRESSIVE)
+			vbl_start = DIV_ROUND_UP(vbl_start, 2);
 	}
 
+	/* Convert to pixel count */
+	vbl_start *= htotal;
+
+	/* Start of vblank event occurs at start of hsync */
+	vbl_start -= htotal - hsync_start;
+
 	high_frame = PIPEFRAME(pipe);
 	low_frame = PIPEFRAMEPIXEL(pipe);
 
@@ -600,12 +857,12 @@ static u32 i915_get_vblank_counter(struct drm_device *dev, int pipe)
 	 * Cook up a vblank counter by also checking the pixel
 	 * counter against vblank start.
 	 */
-	return ((high1 << 8) | low) + (pixel >= vbl_start);
+	return (((high1 << 8) | low) + (pixel >= vbl_start)) & 0xffffff;
 }
 
 static u32 gm45_get_vblank_counter(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int reg = PIPE_FRMCOUNT_GM45(pipe);
 
 	if (!i915_pipe_enabled(dev, pipe)) {
@@ -619,70 +876,41 @@ static u32 gm45_get_vblank_counter(struct drm_device *dev, int pipe)
 
 /* raw reads, only for fast reads of display block, no need for forcewake etc. */
 #define __raw_i915_read32(dev_priv__, reg__) readl((dev_priv__)->regs + (reg__))
-#define __raw_i915_read16(dev_priv__, reg__) readw((dev_priv__)->regs + (reg__))
 
-static bool intel_pipe_in_vblank_locked(struct drm_device *dev, enum pipe pipe)
+static int __intel_get_crtc_scanline(struct intel_crtc *crtc)
 {
+	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t status;
-	int reg;
-
-	if (IS_VALLEYVIEW(dev)) {
-		status = pipe == PIPE_A ?
-			I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT :
-			I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
+	const struct drm_display_mode *mode = &crtc->config.adjusted_mode;
+	enum pipe pipe = crtc->pipe;
+	int position, vtotal;
 
-		reg = VLV_ISR;
-	} else if (IS_GEN2(dev)) {
-		status = pipe == PIPE_A ?
-			I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT :
-			I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
-
-		reg = ISR;
-	} else if (INTEL_INFO(dev)->gen < 5) {
-		status = pipe == PIPE_A ?
-			I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT :
-			I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
-
-		reg = ISR;
-	} else if (INTEL_INFO(dev)->gen < 7) {
-		status = pipe == PIPE_A ?
-			DE_PIPEA_VBLANK :
-			DE_PIPEB_VBLANK;
-
-		reg = DEISR;
-	} else {
-		switch (pipe) {
-		default:
-		case PIPE_A:
-			status = DE_PIPEA_VBLANK_IVB;
-			break;
-		case PIPE_B:
-			status = DE_PIPEB_VBLANK_IVB;
-			break;
-		case PIPE_C:
-			status = DE_PIPEC_VBLANK_IVB;
-			break;
-		}
-
-		reg = DEISR;
-	}
+	vtotal = mode->crtc_vtotal;
+	if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+		vtotal /= 2;
 
 	if (IS_GEN2(dev))
-		return __raw_i915_read16(dev_priv, reg) & status;
+		position = __raw_i915_read32(dev_priv, PIPEDSL(pipe)) & DSL_LINEMASK_GEN2;
 	else
-		return __raw_i915_read32(dev_priv, reg) & status;
+		position = __raw_i915_read32(dev_priv, PIPEDSL(pipe)) & DSL_LINEMASK_GEN3;
+
+	/*
+	 * See update_scanline_offset() for the details on the
+	 * scanline_offset adjustment.
+	 */
+	return (position + crtc->scanline_offset) % vtotal;
 }
 
 static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
-			     int *vpos, int *hpos, ktime_t *stime, ktime_t *etime)
+				    unsigned int flags, int *vpos, int *hpos,
+				    ktime_t *stime, ktime_t *etime)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	const struct drm_display_mode *mode = &intel_crtc->config.adjusted_mode;
 	int position;
-	int vbl_start, vbl_end, htotal, vtotal;
+	int vbl_start, vbl_end, hsync_start, htotal, vtotal;
 	bool in_vbl = true;
 	int ret = 0;
 	unsigned long irqflags;
@@ -694,10 +922,17 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
 	}
 
 	htotal = mode->crtc_htotal;
+	hsync_start = mode->crtc_hsync_start;
 	vtotal = mode->crtc_vtotal;
 	vbl_start = mode->crtc_vblank_start;
 	vbl_end = mode->crtc_vblank_end;
 
+	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
+		vbl_start = DIV_ROUND_UP(vbl_start, 2);
+		vbl_end /= 2;
+		vtotal /= 2;
+	}
+
 	ret |= DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE;
 
 	/*
@@ -706,7 +941,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
 	 * following code must not block on uncore.lock.
 	 */
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
-	
+
 	/* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */
 
 	/* Get optional system timestamp before query. */
@@ -717,22 +952,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
 		/* No obvious pixelcount register. Only query vertical
 		 * scanout position from Display scan line register.
 		 */
-		if (IS_GEN2(dev))
-			position = __raw_i915_read32(dev_priv, PIPEDSL(pipe)) & DSL_LINEMASK_GEN2;
-		else
-			position = __raw_i915_read32(dev_priv, PIPEDSL(pipe)) & DSL_LINEMASK_GEN3;
-
-		/*
-		 * The scanline counter increments at the leading edge
-		 * of hsync, ie. it completely misses the active portion
-		 * of the line. Fix up the counter at both edges of vblank
-		 * to get a more accurate picture whether we're in vblank
-		 * or not.
-		 */
-		in_vbl = intel_pipe_in_vblank_locked(dev, pipe);
-		if ((in_vbl && position == vbl_start - 1) ||
-		    (!in_vbl && position == vbl_end - 1))
-			position = (position + 1) % vtotal;
+		position = __intel_get_crtc_scanline(intel_crtc);
 	} else {
 		/* Have access to pixelcount since start of frame.
 		 * We can split this into vertical and horizontal
@@ -744,6 +964,29 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
 		vbl_start *= htotal;
 		vbl_end *= htotal;
 		vtotal *= htotal;
+
+		/*
+		 * In interlaced modes, the pixel counter counts all pixels,
+		 * so one field will have htotal more pixels. In order to avoid
+		 * the reported position from jumping backwards when the pixel
+		 * counter is beyond the length of the shorter field, just
+		 * clamp the position the length of the shorter field. This
+		 * matches how the scanline counter based position works since
+		 * the scanline counter doesn't count the two half lines.
+		 */
+		if (position >= vtotal)
+			position = vtotal - 1;
+
+		/*
+		 * Start of vblank interrupt is triggered at start of hsync,
+		 * just prior to the first active line of vblank. However we
+		 * consider lines to start at the leading edge of horizontal
+		 * active. So, should we get here before we've crossed into
+		 * the horizontal active of the first line in vblank, we would
+		 * not set the DRM_SCANOUTPOS_INVBL flag. In order to fix that,
+		 * always add htotal-hsync_start to the current pixel position.
+		 */
+		position = (position + htotal - hsync_start) % vtotal;
 	}
 
 	/* Get optional system timestamp after query. */
@@ -782,6 +1025,19 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
 	return ret;
 }
 
+int intel_get_crtc_scanline(struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
+	unsigned long irqflags;
+	int position;
+
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+	position = __intel_get_crtc_scanline(crtc);
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+
+	return position;
+}
+
 static int i915_get_vblank_timestamp(struct drm_device *dev, int pipe,
 			      int *max_error,
 			      struct timeval *vblank_time,
@@ -809,7 +1065,8 @@ static int i915_get_vblank_timestamp(struct drm_device *dev, int pipe,
 	/* Helper routine in DRM core does all the work: */
 	return drm_calc_vbltimestamp_from_scanoutpos(dev, pipe, max_error,
 						     vblank_time, flags,
-						     crtc);
+						     crtc,
+						     &to_intel_crtc(crtc)->config.adjusted_mode);
 }
 
 static bool intel_hpd_irq_event(struct drm_device *dev,
@@ -826,7 +1083,7 @@ static bool intel_hpd_irq_event(struct drm_device *dev,
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s] status updated from %s to %s\n",
 		      connector->base.id,
-		      drm_get_connector_name(connector),
+		      connector->name,
 		      drm_get_connector_status_name(old_status),
 		      drm_get_connector_status_name(connector->status));
 
@@ -840,8 +1097,8 @@ static bool intel_hpd_irq_event(struct drm_device *dev,
 
 static void i915_hotplug_work_func(struct work_struct *work)
 {
-	drm_i915_private_t *dev_priv = container_of(work, drm_i915_private_t,
-						    hotplug_work);
+	struct drm_i915_private *dev_priv =
+		container_of(work, struct drm_i915_private, hotplug_work);
 	struct drm_device *dev = dev_priv->dev;
 	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_connector *intel_connector;
@@ -871,7 +1128,7 @@ static void i915_hotplug_work_func(struct work_struct *work)
 		    connector->polled == DRM_CONNECTOR_POLL_HPD) {
 			DRM_INFO("HPD interrupt storm detected on connector %s: "
 				 "switching from hotplug detection to polling\n",
-				drm_get_connector_name(connector));
+				connector->name);
 			dev_priv->hpd_stats[intel_encoder->hpd_pin].hpd_mark = HPD_DISABLED;
 			connector->polled = DRM_CONNECTOR_POLL_CONNECT
 				| DRM_CONNECTOR_POLL_DISCONNECT;
@@ -879,7 +1136,7 @@ static void i915_hotplug_work_func(struct work_struct *work)
 		}
 		if (hpd_event_bits & (1 << intel_encoder->hpd_pin)) {
 			DRM_DEBUG_KMS("Connector %s (pin %i) received hotplug event.\n",
-				      drm_get_connector_name(connector), intel_encoder->hpd_pin);
+				      connector->name, intel_encoder->hpd_pin);
 		}
 	}
 	 /* if there were no outputs to poll, poll was disabled,
@@ -909,9 +1166,14 @@ static void i915_hotplug_work_func(struct work_struct *work)
 		drm_kms_helper_hotplug_event(dev);
 }
 
+static void intel_hpd_irq_uninstall(struct drm_i915_private *dev_priv)
+{
+	del_timer_sync(&dev_priv->hotplug_reenable_timer);
+}
+
 static void ironlake_rps_change_irq_handler(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 busy_up, busy_down, max_avg, min_avg;
 	u8 new_delay;
 
@@ -949,9 +1211,9 @@ static void ironlake_rps_change_irq_handler(struct drm_device *dev)
 }
 
 static void notify_ring(struct drm_device *dev,
-			struct intel_ring_buffer *ring)
+			struct intel_engine_cs *ring)
 {
-	if (ring->obj == NULL)
+	if (!intel_ring_initialized(ring))
 		return;
 
 	trace_i915_gem_request_complete(ring);
@@ -962,22 +1224,26 @@ static void notify_ring(struct drm_device *dev,
 
 static void gen6_pm_rps_work(struct work_struct *work)
 {
-	drm_i915_private_t *dev_priv = container_of(work, drm_i915_private_t,
-						    rps.work);
+	struct drm_i915_private *dev_priv =
+		container_of(work, struct drm_i915_private, rps.work);
 	u32 pm_iir;
 	int new_delay, adj;
 
 	spin_lock_irq(&dev_priv->irq_lock);
 	pm_iir = dev_priv->rps.pm_iir;
 	dev_priv->rps.pm_iir = 0;
-	/* Make sure not to corrupt PMIMR state used by ringbuffer code */
-	snb_enable_pm_irq(dev_priv, GEN6_PM_RPS_EVENTS);
+	if (IS_BROADWELL(dev_priv->dev))
+		bdw_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
+	else {
+		/* Make sure not to corrupt PMIMR state used by ringbuffer */
+		snb_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
+	}
 	spin_unlock_irq(&dev_priv->irq_lock);
 
 	/* Make sure we didn't queue anything we're not going to process. */
-	WARN_ON(pm_iir & ~GEN6_PM_RPS_EVENTS);
+	WARN_ON(pm_iir & ~dev_priv->pm_rps_events);
 
-	if ((pm_iir & GEN6_PM_RPS_EVENTS) == 0)
+	if ((pm_iir & dev_priv->pm_rps_events) == 0)
 		return;
 
 	mutex_lock(&dev_priv->rps.hw_lock);
@@ -988,38 +1254,38 @@ static void gen6_pm_rps_work(struct work_struct *work)
 			adj *= 2;
 		else
 			adj = 1;
-		new_delay = dev_priv->rps.cur_delay + adj;
+		new_delay = dev_priv->rps.cur_freq + adj;
 
 		/*
 		 * For better performance, jump directly
 		 * to RPe if we're below it.
 		 */
-		if (new_delay < dev_priv->rps.rpe_delay)
-			new_delay = dev_priv->rps.rpe_delay;
+		if (new_delay < dev_priv->rps.efficient_freq)
+			new_delay = dev_priv->rps.efficient_freq;
 	} else if (pm_iir & GEN6_PM_RP_DOWN_TIMEOUT) {
-		if (dev_priv->rps.cur_delay > dev_priv->rps.rpe_delay)
-			new_delay = dev_priv->rps.rpe_delay;
+		if (dev_priv->rps.cur_freq > dev_priv->rps.efficient_freq)
+			new_delay = dev_priv->rps.efficient_freq;
 		else
-			new_delay = dev_priv->rps.min_delay;
+			new_delay = dev_priv->rps.min_freq_softlimit;
 		adj = 0;
 	} else if (pm_iir & GEN6_PM_RP_DOWN_THRESHOLD) {
 		if (adj < 0)
 			adj *= 2;
 		else
 			adj = -1;
-		new_delay = dev_priv->rps.cur_delay + adj;
+		new_delay = dev_priv->rps.cur_freq + adj;
 	} else { /* unknown event */
-		new_delay = dev_priv->rps.cur_delay;
+		new_delay = dev_priv->rps.cur_freq;
 	}
 
 	/* sysfs frequency interfaces may have snuck in while servicing the
 	 * interrupt
 	 */
-	if (new_delay < (int)dev_priv->rps.min_delay)
-		new_delay = dev_priv->rps.min_delay;
-	if (new_delay > (int)dev_priv->rps.max_delay)
-		new_delay = dev_priv->rps.max_delay;
-	dev_priv->rps.last_adj = new_delay - dev_priv->rps.cur_delay;
+	new_delay = clamp_t(int, new_delay,
+			    dev_priv->rps.min_freq_softlimit,
+			    dev_priv->rps.max_freq_softlimit);
+
+	dev_priv->rps.last_adj = new_delay - dev_priv->rps.cur_freq;
 
 	if (IS_VALLEYVIEW(dev_priv->dev))
 		valleyview_set_rps(dev_priv->dev, new_delay);
@@ -1041,8 +1307,8 @@ static void gen6_pm_rps_work(struct work_struct *work)
  */
 static void ivybridge_parity_work(struct work_struct *work)
 {
-	drm_i915_private_t *dev_priv = container_of(work, drm_i915_private_t,
-						    l3_parity.error_work);
+	struct drm_i915_private *dev_priv =
+		container_of(work, struct drm_i915_private, l3_parity.error_work);
 	u32 error_status, row, bank, subbank;
 	char *parity_event[6];
 	uint32_t misccpctl;
@@ -1114,7 +1380,7 @@ out:
 
 static void ivybridge_parity_error_irq_handler(struct drm_device *dev, u32 iir)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	if (!HAS_L3_DPF(dev))
 		return;
@@ -1160,14 +1426,27 @@ static void snb_gt_irq_handler(struct drm_device *dev,
 	if (gt_iir & (GT_BLT_CS_ERROR_INTERRUPT |
 		      GT_BSD_CS_ERROR_INTERRUPT |
 		      GT_RENDER_CS_MASTER_ERROR_INTERRUPT)) {
-		DRM_ERROR("GT error interrupt 0x%08x\n", gt_iir);
-		i915_handle_error(dev, false);
+		i915_handle_error(dev, false, "GT error interrupt 0x%08x",
+				  gt_iir);
 	}
 
 	if (gt_iir & GT_PARITY_ERROR(dev))
 		ivybridge_parity_error_irq_handler(dev, gt_iir);
 }
 
+static void gen8_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir)
+{
+	if ((pm_iir & dev_priv->pm_rps_events) == 0)
+		return;
+
+	spin_lock(&dev_priv->irq_lock);
+	dev_priv->rps.pm_iir |= pm_iir & dev_priv->pm_rps_events;
+	bdw_disable_pm_irq(dev_priv, pm_iir & dev_priv->pm_rps_events);
+	spin_unlock(&dev_priv->irq_lock);
+
+	queue_work(dev_priv->wq, &dev_priv->rps.work);
+}
+
 static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 				       struct drm_i915_private *dev_priv,
 				       u32 master_ctl)
@@ -1191,18 +1470,32 @@ static irqreturn_t gen8_gt_irq_handler(struct drm_device *dev,
 			DRM_ERROR("The master control interrupt lied (GT0)!\n");
 	}
 
-	if (master_ctl & GEN8_GT_VCS1_IRQ) {
+	if (master_ctl & (GEN8_GT_VCS1_IRQ | GEN8_GT_VCS2_IRQ)) {
 		tmp = I915_READ(GEN8_GT_IIR(1));
 		if (tmp) {
 			ret = IRQ_HANDLED;
 			vcs = tmp >> GEN8_VCS1_IRQ_SHIFT;
 			if (vcs & GT_RENDER_USER_INTERRUPT)
 				notify_ring(dev, &dev_priv->ring[VCS]);
+			vcs = tmp >> GEN8_VCS2_IRQ_SHIFT;
+			if (vcs & GT_RENDER_USER_INTERRUPT)
+				notify_ring(dev, &dev_priv->ring[VCS2]);
 			I915_WRITE(GEN8_GT_IIR(1), tmp);
 		} else
 			DRM_ERROR("The master control interrupt lied (GT1)!\n");
 	}
 
+	if (master_ctl & GEN8_GT_PM_IRQ) {
+		tmp = I915_READ(GEN8_GT_IIR(2));
+		if (tmp & dev_priv->pm_rps_events) {
+			ret = IRQ_HANDLED;
+			gen8_rps_irq_handler(dev_priv, tmp);
+			I915_WRITE(GEN8_GT_IIR(2),
+				   tmp & dev_priv->pm_rps_events);
+		} else
+			DRM_ERROR("The master control interrupt lied (PM)!\n");
+	}
+
 	if (master_ctl & GEN8_GT_VECS_IRQ) {
 		tmp = I915_READ(GEN8_GT_IIR(3));
 		if (tmp) {
@@ -1225,19 +1518,33 @@ static inline void intel_hpd_irq_handler(struct drm_device *dev,
 					 u32 hotplug_trigger,
 					 const u32 *hpd)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 	bool storm_detected = false;
 
 	if (!hotplug_trigger)
 		return;
 
+	DRM_DEBUG_DRIVER("hotplug event received, stat 0x%08x\n",
+			  hotplug_trigger);
+
 	spin_lock(&dev_priv->irq_lock);
 	for (i = 1; i < HPD_NUM_PINS; i++) {
 
-		WARN(((hpd[i] & hotplug_trigger) &&
-		      dev_priv->hpd_stats[i].hpd_mark != HPD_ENABLED),
-		     "Received HPD interrupt although disabled\n");
+		if (hpd[i] & hotplug_trigger &&
+		    dev_priv->hpd_stats[i].hpd_mark == HPD_DISABLED) {
+			/*
+			 * On GMCH platforms the interrupt mask bits only
+			 * prevent irq generation, not the setting of the
+			 * hotplug bits itself. So only WARN about unexpected
+			 * interrupts on saner platforms.
+			 */
+			WARN_ONCE(INTEL_INFO(dev)->gen >= 5 && !IS_VALLEYVIEW(dev),
+				  "Received HPD interrupt (0x%08x) on pin %d (0x%08x) although disabled\n",
+				  hotplug_trigger, i, hpd[i]);
+
+			continue;
+		}
 
 		if (!(hpd[i] & hotplug_trigger) ||
 		    dev_priv->hpd_stats[i].hpd_mark != HPD_ENABLED)
@@ -1277,14 +1584,14 @@ static inline void intel_hpd_irq_handler(struct drm_device *dev,
 
 static void gmbus_irq_handler(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	wake_up_all(&dev_priv->gmbus_wait_queue);
 }
 
 static void dp_aux_irq_handler(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	wake_up_all(&dev_priv->gmbus_wait_queue);
 }
@@ -1390,10 +1697,10 @@ static void i9xx_pipe_crc_irq_handler(struct drm_device *dev, enum pipe pipe)
  * the work queue. */
 static void gen6_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir)
 {
-	if (pm_iir & GEN6_PM_RPS_EVENTS) {
+	if (pm_iir & dev_priv->pm_rps_events) {
 		spin_lock(&dev_priv->irq_lock);
-		dev_priv->rps.pm_iir |= pm_iir & GEN6_PM_RPS_EVENTS;
-		snb_disable_pm_irq(dev_priv, pm_iir & GEN6_PM_RPS_EVENTS);
+		dev_priv->rps.pm_iir |= pm_iir & dev_priv->pm_rps_events;
+		snb_disable_pm_irq(dev_priv, pm_iir & dev_priv->pm_rps_events);
 		spin_unlock(&dev_priv->irq_lock);
 
 		queue_work(dev_priv->wq, &dev_priv->rps.work);
@@ -1404,23 +1711,132 @@ static void gen6_rps_irq_handler(struct drm_i915_private *dev_priv, u32 pm_iir)
 			notify_ring(dev_priv->dev, &dev_priv->ring[VECS]);
 
 		if (pm_iir & PM_VEBOX_CS_ERROR_INTERRUPT) {
-			DRM_ERROR("VEBOX CS error interrupt 0x%08x\n", pm_iir);
-			i915_handle_error(dev_priv->dev, false);
+			i915_handle_error(dev_priv->dev, false,
+					  "VEBOX CS error interrupt 0x%08x",
+					  pm_iir);
 		}
 	}
 }
 
+static bool intel_pipe_handle_vblank(struct drm_device *dev, enum pipe pipe)
+{
+	struct intel_crtc *crtc;
+
+	if (!drm_handle_vblank(dev, pipe))
+		return false;
+
+	crtc = to_intel_crtc(intel_get_crtc_for_pipe(dev, pipe));
+	wake_up(&crtc->vbl_wait);
+
+	return true;
+}
+
+static void valleyview_pipestat_irq_handler(struct drm_device *dev, u32 iir)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 pipe_stats[I915_MAX_PIPES] = { };
+	int pipe;
+
+	spin_lock(&dev_priv->irq_lock);
+	for_each_pipe(pipe) {
+		int reg;
+		u32 mask, iir_bit = 0;
+
+		/*
+		 * PIPESTAT bits get signalled even when the interrupt is
+		 * disabled with the mask bits, and some of the status bits do
+		 * not generate interrupts at all (like the underrun bit). Hence
+		 * we need to be careful that we only handle what we want to
+		 * handle.
+		 */
+		mask = 0;
+		if (__cpu_fifo_underrun_reporting_enabled(dev, pipe))
+			mask |= PIPE_FIFO_UNDERRUN_STATUS;
+
+		switch (pipe) {
+		case PIPE_A:
+			iir_bit = I915_DISPLAY_PIPE_A_EVENT_INTERRUPT;
+			break;
+		case PIPE_B:
+			iir_bit = I915_DISPLAY_PIPE_B_EVENT_INTERRUPT;
+			break;
+		case PIPE_C:
+			iir_bit = I915_DISPLAY_PIPE_C_EVENT_INTERRUPT;
+			break;
+		}
+		if (iir & iir_bit)
+			mask |= dev_priv->pipestat_irq_mask[pipe];
+
+		if (!mask)
+			continue;
+
+		reg = PIPESTAT(pipe);
+		mask |= PIPESTAT_INT_ENABLE_MASK;
+		pipe_stats[pipe] = I915_READ(reg) & mask;
+
+		/*
+		 * Clear the PIPE*STAT regs before the IIR
+		 */
+		if (pipe_stats[pipe] & (PIPE_FIFO_UNDERRUN_STATUS |
+					PIPESTAT_INT_STATUS_MASK))
+			I915_WRITE(reg, pipe_stats[pipe]);
+	}
+	spin_unlock(&dev_priv->irq_lock);
+
+	for_each_pipe(pipe) {
+		if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS)
+			intel_pipe_handle_vblank(dev, pipe);
+
+		if (pipe_stats[pipe] & PLANE_FLIP_DONE_INT_STATUS_VLV) {
+			intel_prepare_page_flip(dev, pipe);
+			intel_finish_page_flip(dev, pipe);
+		}
+
+		if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
+			i9xx_pipe_crc_irq_handler(dev, pipe);
+
+		if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS &&
+		    intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
+			DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
+	}
+
+	if (pipe_stats[0] & PIPE_GMBUS_INTERRUPT_STATUS)
+		gmbus_irq_handler(dev);
+}
+
+static void i9xx_hpd_irq_handler(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT);
+
+	if (IS_G4X(dev)) {
+		u32 hotplug_trigger = hotplug_status & HOTPLUG_INT_STATUS_G4X;
+
+		intel_hpd_irq_handler(dev, hotplug_trigger, hpd_status_g4x);
+	} else {
+		u32 hotplug_trigger = hotplug_status & HOTPLUG_INT_STATUS_I915;
+
+		intel_hpd_irq_handler(dev, hotplug_trigger, hpd_status_i915);
+	}
+
+	if ((IS_G4X(dev) || IS_VALLEYVIEW(dev)) &&
+	    hotplug_status & DP_AUX_CHANNEL_MASK_INT_STATUS_G4X)
+		dp_aux_irq_handler(dev);
+
+	I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status);
+	/*
+	 * Make sure hotplug status is cleared before we clear IIR, or else we
+	 * may miss hotplug events.
+	 */
+	POSTING_READ(PORT_HOTPLUG_STAT);
+}
+
 static irqreturn_t valleyview_irq_handler(int irq, void *arg)
 {
-	struct drm_device *dev = (struct drm_device *) arg;
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_device *dev = arg;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 iir, gt_iir, pm_iir;
 	irqreturn_t ret = IRQ_NONE;
-	unsigned long irqflags;
-	int pipe;
-	u32 pipe_stats[I915_MAX_PIPES];
-
-	atomic_inc(&dev_priv->irq_received);
 
 	while (true) {
 		iir = I915_READ(VLV_IIR);
@@ -1434,68 +1850,61 @@ static irqreturn_t valleyview_irq_handler(int irq, void *arg)
 
 		snb_gt_irq_handler(dev, dev_priv, gt_iir);
 
-		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-		for_each_pipe(pipe) {
-			int reg = PIPESTAT(pipe);
-			pipe_stats[pipe] = I915_READ(reg);
+		valleyview_pipestat_irq_handler(dev, iir);
 
-			/*
-			 * Clear the PIPE*STAT regs before the IIR
-			 */
-			if (pipe_stats[pipe] & 0x8000ffff) {
-				if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
-					DRM_DEBUG_DRIVER("pipe %c underrun\n",
-							 pipe_name(pipe));
-				I915_WRITE(reg, pipe_stats[pipe]);
-			}
-		}
-		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+		/* Consume port.  Then clear IIR or we'll miss events */
+		if (iir & I915_DISPLAY_PORT_INTERRUPT)
+			i9xx_hpd_irq_handler(dev);
 
-		for_each_pipe(pipe) {
-			if (pipe_stats[pipe] & PIPE_START_VBLANK_INTERRUPT_STATUS)
-				drm_handle_vblank(dev, pipe);
+		if (pm_iir)
+			gen6_rps_irq_handler(dev_priv, pm_iir);
 
-			if (pipe_stats[pipe] & PLANE_FLIPDONE_INT_STATUS_VLV) {
-				intel_prepare_page_flip(dev, pipe);
-				intel_finish_page_flip(dev, pipe);
-			}
+		I915_WRITE(GTIIR, gt_iir);
+		I915_WRITE(GEN6_PMIIR, pm_iir);
+		I915_WRITE(VLV_IIR, iir);
+	}
 
-			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
-				i9xx_pipe_crc_irq_handler(dev, pipe);
-		}
+out:
+	return ret;
+}
 
-		/* Consume port.  Then clear IIR or we'll miss events */
-		if (iir & I915_DISPLAY_PORT_INTERRUPT) {
-			u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT);
-			u32 hotplug_trigger = hotplug_status & HOTPLUG_INT_STATUS_I915;
+static irqreturn_t cherryview_irq_handler(int irq, void *arg)
+{
+	struct drm_device *dev = arg;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 master_ctl, iir;
+	irqreturn_t ret = IRQ_NONE;
 
-			DRM_DEBUG_DRIVER("hotplug event received, stat 0x%08x\n",
-					 hotplug_status);
+	for (;;) {
+		master_ctl = I915_READ(GEN8_MASTER_IRQ) & ~GEN8_MASTER_IRQ_CONTROL;
+		iir = I915_READ(VLV_IIR);
 
-			intel_hpd_irq_handler(dev, hotplug_trigger, hpd_status_i915);
+		if (master_ctl == 0 && iir == 0)
+			break;
 
-			I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status);
-			I915_READ(PORT_HOTPLUG_STAT);
-		}
+		I915_WRITE(GEN8_MASTER_IRQ, 0);
 
-		if (pipe_stats[0] & PIPE_GMBUS_INTERRUPT_STATUS)
-			gmbus_irq_handler(dev);
+		gen8_gt_irq_handler(dev, dev_priv, master_ctl);
 
-		if (pm_iir)
-			gen6_rps_irq_handler(dev_priv, pm_iir);
+		valleyview_pipestat_irq_handler(dev, iir);
+
+		/* Consume port.  Then clear IIR or we'll miss events */
+		i9xx_hpd_irq_handler(dev);
 
-		I915_WRITE(GTIIR, gt_iir);
-		I915_WRITE(GEN6_PMIIR, pm_iir);
 		I915_WRITE(VLV_IIR, iir);
+
+		I915_WRITE(GEN8_MASTER_IRQ, DE_MASTER_IRQ_CONTROL);
+		POSTING_READ(GEN8_MASTER_IRQ);
+
+		ret = IRQ_HANDLED;
 	}
 
-out:
 	return ret;
 }
 
 static void ibx_irq_handler(struct drm_device *dev, u32 pch_iir)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 	u32 hotplug_trigger = pch_iir & SDE_HOTPLUG_MASK;
 
@@ -1538,12 +1947,12 @@ static void ibx_irq_handler(struct drm_device *dev, u32 pch_iir)
 	if (pch_iir & SDE_TRANSA_FIFO_UNDER)
 		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_A,
 							  false))
-			DRM_DEBUG_DRIVER("PCH transcoder A FIFO underrun\n");
+			DRM_ERROR("PCH transcoder A FIFO underrun\n");
 
 	if (pch_iir & SDE_TRANSB_FIFO_UNDER)
 		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_B,
 							  false))
-			DRM_DEBUG_DRIVER("PCH transcoder B FIFO underrun\n");
+			DRM_ERROR("PCH transcoder B FIFO underrun\n");
 }
 
 static void ivb_err_int_handler(struct drm_device *dev)
@@ -1559,8 +1968,8 @@ static void ivb_err_int_handler(struct drm_device *dev)
 		if (err_int & ERR_INT_FIFO_UNDERRUN(pipe)) {
 			if (intel_set_cpu_fifo_underrun_reporting(dev, pipe,
 								  false))
-				DRM_DEBUG_DRIVER("Pipe %c FIFO underrun\n",
-						 pipe_name(pipe));
+				DRM_ERROR("Pipe %c FIFO underrun\n",
+					  pipe_name(pipe));
 		}
 
 		if (err_int & ERR_INT_PIPE_CRC_DONE(pipe)) {
@@ -1585,24 +1994,24 @@ static void cpt_serr_int_handler(struct drm_device *dev)
 	if (serr_int & SERR_INT_TRANS_A_FIFO_UNDERRUN)
 		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_A,
 							  false))
-			DRM_DEBUG_DRIVER("PCH transcoder A FIFO underrun\n");
+			DRM_ERROR("PCH transcoder A FIFO underrun\n");
 
 	if (serr_int & SERR_INT_TRANS_B_FIFO_UNDERRUN)
 		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_B,
 							  false))
-			DRM_DEBUG_DRIVER("PCH transcoder B FIFO underrun\n");
+			DRM_ERROR("PCH transcoder B FIFO underrun\n");
 
 	if (serr_int & SERR_INT_TRANS_C_FIFO_UNDERRUN)
 		if (intel_set_pch_fifo_underrun_reporting(dev, TRANSCODER_C,
 							  false))
-			DRM_DEBUG_DRIVER("PCH transcoder C FIFO underrun\n");
+			DRM_ERROR("PCH transcoder C FIFO underrun\n");
 
 	I915_WRITE(SERR_INT, serr_int);
 }
 
 static void cpt_irq_handler(struct drm_device *dev, u32 pch_iir)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 	u32 hotplug_trigger = pch_iir & SDE_HOTPLUG_MASK_CPT;
 
@@ -1653,12 +2062,12 @@ static void ilk_display_irq_handler(struct drm_device *dev, u32 de_iir)
 
 	for_each_pipe(pipe) {
 		if (de_iir & DE_PIPE_VBLANK(pipe))
-			drm_handle_vblank(dev, pipe);
+			intel_pipe_handle_vblank(dev, pipe);
 
 		if (de_iir & DE_PIPE_FIFO_UNDERRUN(pipe))
 			if (intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
-				DRM_DEBUG_DRIVER("Pipe %c FIFO underrun\n",
-						 pipe_name(pipe));
+				DRM_ERROR("Pipe %c FIFO underrun\n",
+					  pipe_name(pipe));
 
 		if (de_iir & DE_PIPE_CRC_DONE(pipe))
 			i9xx_pipe_crc_irq_handler(dev, pipe);
@@ -1690,7 +2099,7 @@ static void ilk_display_irq_handler(struct drm_device *dev, u32 de_iir)
 static void ivb_display_irq_handler(struct drm_device *dev, u32 de_iir)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum pipe i;
+	enum pipe pipe;
 
 	if (de_iir & DE_ERR_INT_IVB)
 		ivb_err_int_handler(dev);
@@ -1701,14 +2110,14 @@ static void ivb_display_irq_handler(struct drm_device *dev, u32 de_iir)
 	if (de_iir & DE_GSE_IVB)
 		intel_opregion_asle_intr(dev);
 
-	for_each_pipe(i) {
-		if (de_iir & (DE_PIPE_VBLANK_IVB(i)))
-			drm_handle_vblank(dev, i);
+	for_each_pipe(pipe) {
+		if (de_iir & (DE_PIPE_VBLANK_IVB(pipe)))
+			intel_pipe_handle_vblank(dev, pipe);
 
 		/* plane/pipes map 1:1 on ilk+ */
-		if (de_iir & DE_PLANE_FLIP_DONE_IVB(i)) {
-			intel_prepare_page_flip(dev, i);
-			intel_finish_page_flip_plane(dev, i);
+		if (de_iir & DE_PLANE_FLIP_DONE_IVB(pipe)) {
+			intel_prepare_page_flip(dev, pipe);
+			intel_finish_page_flip_plane(dev, pipe);
 		}
 	}
 
@@ -1725,13 +2134,11 @@ static void ivb_display_irq_handler(struct drm_device *dev, u32 de_iir)
 
 static irqreturn_t ironlake_irq_handler(int irq, void *arg)
 {
-	struct drm_device *dev = (struct drm_device *) arg;
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_device *dev = arg;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 de_iir, gt_iir, de_ier, sde_ier = 0;
 	irqreturn_t ret = IRQ_NONE;
 
-	atomic_inc(&dev_priv->irq_received);
-
 	/* We get interrupts on unclaimed registers, so check for this before we
 	 * do any I915_{READ,WRITE}. */
 	intel_uncore_check_errors(dev);
@@ -1800,8 +2207,6 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 	uint32_t tmp = 0;
 	enum pipe pipe;
 
-	atomic_inc(&dev_priv->irq_received);
-
 	master_ctl = I915_READ(GEN8_MASTER_IRQ);
 	master_ctl &= ~GEN8_MASTER_IRQ_CONTROL;
 	if (!master_ctl)
@@ -1850,9 +2255,9 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 
 		pipe_iir = I915_READ(GEN8_DE_PIPE_IIR(pipe));
 		if (pipe_iir & GEN8_PIPE_VBLANK)
-			drm_handle_vblank(dev, pipe);
+			intel_pipe_handle_vblank(dev, pipe);
 
-		if (pipe_iir & GEN8_PIPE_FLIP_DONE) {
+		if (pipe_iir & GEN8_PIPE_PRIMARY_FLIP_DONE) {
 			intel_prepare_page_flip(dev, pipe);
 			intel_finish_page_flip_plane(dev, pipe);
 		}
@@ -1863,8 +2268,8 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 		if (pipe_iir & GEN8_PIPE_FIFO_UNDERRUN) {
 			if (intel_set_cpu_fifo_underrun_reporting(dev, pipe,
 								  false))
-				DRM_DEBUG_DRIVER("Pipe %c FIFO underrun\n",
-						 pipe_name(pipe));
+				DRM_ERROR("Pipe %c FIFO underrun\n",
+					  pipe_name(pipe));
 		}
 
 		if (pipe_iir & GEN8_DE_PIPE_IRQ_FAULT_ERRORS) {
@@ -1905,7 +2310,7 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg)
 static void i915_error_wake_up(struct drm_i915_private *dev_priv,
 			       bool reset_completed)
 {
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	int i;
 
 	/*
@@ -1941,8 +2346,8 @@ static void i915_error_work_func(struct work_struct *work)
 {
 	struct i915_gpu_error *error = container_of(work, struct i915_gpu_error,
 						    work);
-	drm_i915_private_t *dev_priv = container_of(error, drm_i915_private_t,
-						    gpu_error);
+	struct drm_i915_private *dev_priv =
+		container_of(error, struct drm_i915_private, gpu_error);
 	struct drm_device *dev = dev_priv->dev;
 	char *error_event[] = { I915_ERROR_UEVENT "=1", NULL };
 	char *reset_event[] = { I915_RESET_UEVENT "=1", NULL };
@@ -1967,6 +2372,14 @@ static void i915_error_work_func(struct work_struct *work)
 				   reset_event);
 
 		/*
+		 * In most cases it's guaranteed that we get here with an RPM
+		 * reference held, for example because there is a pending GPU
+		 * request that won't finish until the reset is done. This
+		 * isn't the case at least when we get here by doing a
+		 * simulated reset via debugs, so get an RPM reference.
+		 */
+		intel_runtime_pm_get(dev_priv);
+		/*
 		 * All state reset _must_ be completed before we update the
 		 * reset counter, for otherwise waiters might miss the reset
 		 * pending state and not properly drop locks, resulting in
@@ -1976,6 +2389,8 @@ static void i915_error_work_func(struct work_struct *work)
 
 		intel_display_handle_reset(dev);
 
+		intel_runtime_pm_put(dev_priv);
+
 		if (ret == 0) {
 			/*
 			 * After all the gem state is reset, increment the reset
@@ -1987,13 +2402,13 @@ static void i915_error_work_func(struct work_struct *work)
 			 * updates before
 			 * the counter increment.
 			 */
-			smp_mb__before_atomic_inc();
+			smp_mb__before_atomic();
 			atomic_inc(&dev_priv->gpu_error.reset_counter);
 
 			kobject_uevent_env(&dev->primary->kdev->kobj,
 					   KOBJ_CHANGE, reset_done_event);
 		} else {
-			atomic_set(&error->reset_counter, I915_WEDGED);
+			atomic_set_mask(I915_WEDGED, &error->reset_counter);
 		}
 
 		/*
@@ -2106,11 +2521,18 @@ static void i915_report_and_clear_eir(struct drm_device *dev)
  * so userspace knows something bad happened (should trigger collection
  * of a ring dump etc.).
  */
-void i915_handle_error(struct drm_device *dev, bool wedged)
+void i915_handle_error(struct drm_device *dev, bool wedged,
+		       const char *fmt, ...)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	va_list args;
+	char error_msg[80];
 
-	i915_capture_error_state(dev);
+	va_start(args, fmt);
+	vscnprintf(error_msg, sizeof(error_msg), fmt, args);
+	va_end(args);
+
+	i915_capture_error_state(dev, wedged, error_msg);
 	i915_report_and_clear_eir(dev);
 
 	if (wedged) {
@@ -2144,7 +2566,7 @@ void i915_handle_error(struct drm_device *dev, bool wedged)
 
 static void __always_unused i915_pageflip_stall_check(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct drm_i915_gem_object *obj;
@@ -2176,8 +2598,8 @@ static void __always_unused i915_pageflip_stall_check(struct drm_device *dev, in
 	} else {
 		int dspaddr = DSPADDR(intel_crtc->plane);
 		stall_detected = I915_READ(dspaddr) == (i915_gem_obj_ggtt_offset(obj) +
-							crtc->y * crtc->fb->pitches[0] +
-							crtc->x * crtc->fb->bits_per_pixel/8);
+							crtc->y * crtc->primary->fb->pitches[0] +
+							crtc->x * crtc->primary->fb->bits_per_pixel/8);
 	}
 
 	spin_unlock_irqrestore(&dev->event_lock, flags);
@@ -2193,7 +2615,7 @@ static void __always_unused i915_pageflip_stall_check(struct drm_device *dev, in
  */
 static int i915_enable_vblank(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
 
 	if (!i915_pipe_enabled(dev, pipe))
@@ -2202,14 +2624,10 @@ static int i915_enable_vblank(struct drm_device *dev, int pipe)
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 	if (INTEL_INFO(dev)->gen >= 4)
 		i915_enable_pipestat(dev_priv, pipe,
-				     PIPE_START_VBLANK_INTERRUPT_ENABLE);
+				     PIPE_START_VBLANK_INTERRUPT_STATUS);
 	else
 		i915_enable_pipestat(dev_priv, pipe,
-				     PIPE_VBLANK_INTERRUPT_ENABLE);
-
-	/* maintain vblank delivery even in deep C-states */
-	if (dev_priv->info->gen == 3)
-		I915_WRITE(INSTPM, _MASKED_BIT_DISABLE(INSTPM_AGPBUSY_DIS));
+				     PIPE_VBLANK_INTERRUPT_STATUS);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
 	return 0;
@@ -2217,7 +2635,7 @@ static int i915_enable_vblank(struct drm_device *dev, int pipe)
 
 static int ironlake_enable_vblank(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
 	uint32_t bit = (INTEL_INFO(dev)->gen >= 7) ? DE_PIPE_VBLANK_IVB(pipe) :
 						     DE_PIPE_VBLANK(pipe);
@@ -2234,22 +2652,15 @@ static int ironlake_enable_vblank(struct drm_device *dev, int pipe)
 
 static int valleyview_enable_vblank(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
-	u32 imr;
 
 	if (!i915_pipe_enabled(dev, pipe))
 		return -EINVAL;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-	imr = I915_READ(VLV_IMR);
-	if (pipe == PIPE_A)
-		imr &= ~I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT;
-	else
-		imr &= ~I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
-	I915_WRITE(VLV_IMR, imr);
 	i915_enable_pipestat(dev_priv, pipe,
-			     PIPE_START_VBLANK_INTERRUPT_ENABLE);
+			     PIPE_START_VBLANK_INTERRUPT_STATUS);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
 	return 0;
@@ -2276,22 +2687,19 @@ static int gen8_enable_vblank(struct drm_device *dev, int pipe)
  */
 static void i915_disable_vblank(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-	if (dev_priv->info->gen == 3)
-		I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_AGPBUSY_DIS));
-
 	i915_disable_pipestat(dev_priv, pipe,
-			      PIPE_VBLANK_INTERRUPT_ENABLE |
-			      PIPE_START_VBLANK_INTERRUPT_ENABLE);
+			      PIPE_VBLANK_INTERRUPT_STATUS |
+			      PIPE_START_VBLANK_INTERRUPT_STATUS);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 }
 
 static void ironlake_disable_vblank(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
 	uint32_t bit = (INTEL_INFO(dev)->gen >= 7) ? DE_PIPE_VBLANK_IVB(pipe) :
 						     DE_PIPE_VBLANK(pipe);
@@ -2303,19 +2711,12 @@ static void ironlake_disable_vblank(struct drm_device *dev, int pipe)
 
 static void valleyview_disable_vblank(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
-	u32 imr;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 	i915_disable_pipestat(dev_priv, pipe,
-			      PIPE_START_VBLANK_INTERRUPT_ENABLE);
-	imr = I915_READ(VLV_IMR);
-	if (pipe == PIPE_A)
-		imr |= I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT;
-	else
-		imr |= I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
-	I915_WRITE(VLV_IMR, imr);
+			      PIPE_START_VBLANK_INTERRUPT_STATUS);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 }
 
@@ -2335,80 +2736,149 @@ static void gen8_disable_vblank(struct drm_device *dev, int pipe)
 }
 
 static u32
-ring_last_seqno(struct intel_ring_buffer *ring)
+ring_last_seqno(struct intel_engine_cs *ring)
 {
 	return list_entry(ring->request_list.prev,
 			  struct drm_i915_gem_request, list)->seqno;
 }
 
 static bool
-ring_idle(struct intel_ring_buffer *ring, u32 seqno)
+ring_idle(struct intel_engine_cs *ring, u32 seqno)
 {
 	return (list_empty(&ring->request_list) ||
 		i915_seqno_passed(seqno, ring_last_seqno(ring)));
 }
 
-static struct intel_ring_buffer *
-semaphore_waits_for(struct intel_ring_buffer *ring, u32 *seqno)
+static bool
+ipehr_is_semaphore_wait(struct drm_device *dev, u32 ipehr)
+{
+	if (INTEL_INFO(dev)->gen >= 8) {
+		/*
+		 * FIXME: gen8 semaphore support - currently we don't emit
+		 * semaphores on bdw anyway, but this needs to be addressed when
+		 * we merge that code.
+		 */
+		return false;
+	} else {
+		ipehr &= ~MI_SEMAPHORE_SYNC_MASK;
+		return ipehr == (MI_SEMAPHORE_MBOX | MI_SEMAPHORE_COMPARE |
+				 MI_SEMAPHORE_REGISTER);
+	}
+}
+
+static struct intel_engine_cs *
+semaphore_wait_to_signaller_ring(struct intel_engine_cs *ring, u32 ipehr)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-	u32 cmd, ipehr, acthd, acthd_min;
+	struct intel_engine_cs *signaller;
+	int i;
+
+	if (INTEL_INFO(dev_priv->dev)->gen >= 8) {
+		/*
+		 * FIXME: gen8 semaphore support - currently we don't emit
+		 * semaphores on bdw anyway, but this needs to be addressed when
+		 * we merge that code.
+		 */
+		return NULL;
+	} else {
+		u32 sync_bits = ipehr & MI_SEMAPHORE_SYNC_MASK;
+
+		for_each_ring(signaller, dev_priv, i) {
+			if(ring == signaller)
+				continue;
+
+			if (sync_bits == signaller->semaphore.mbox.wait[ring->id])
+				return signaller;
+		}
+	}
+
+	DRM_ERROR("No signaller ring found for ring %i, ipehr 0x%08x\n",
+		  ring->id, ipehr);
+
+	return NULL;
+}
+
+static struct intel_engine_cs *
+semaphore_waits_for(struct intel_engine_cs *ring, u32 *seqno)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	u32 cmd, ipehr, head;
+	int i;
 
 	ipehr = I915_READ(RING_IPEHR(ring->mmio_base));
-	if ((ipehr & ~(0x3 << 16)) !=
-	    (MI_SEMAPHORE_MBOX | MI_SEMAPHORE_COMPARE | MI_SEMAPHORE_REGISTER))
+	if (!ipehr_is_semaphore_wait(ring->dev, ipehr))
 		return NULL;
 
-	/* ACTHD is likely pointing to the dword after the actual command,
-	 * so scan backwards until we find the MBOX.
+	/*
+	 * HEAD is likely pointing to the dword after the actual command,
+	 * so scan backwards until we find the MBOX. But limit it to just 3
+	 * dwords. Note that we don't care about ACTHD here since that might
+	 * point at at batch, and semaphores are always emitted into the
+	 * ringbuffer itself.
 	 */
-	acthd = intel_ring_get_active_head(ring) & HEAD_ADDR;
-	acthd_min = max((int)acthd - 3 * 4, 0);
-	do {
-		cmd = ioread32(ring->virtual_start + acthd);
+	head = I915_READ_HEAD(ring) & HEAD_ADDR;
+
+	for (i = 4; i; --i) {
+		/*
+		 * Be paranoid and presume the hw has gone off into the wild -
+		 * our ring is smaller than what the hardware (and hence
+		 * HEAD_ADDR) allows. Also handles wrap-around.
+		 */
+		head &= ring->buffer->size - 1;
+
+		/* This here seems to blow up */
+		cmd = ioread32(ring->buffer->virtual_start + head);
 		if (cmd == ipehr)
 			break;
 
-		acthd -= 4;
-		if (acthd < acthd_min)
-			return NULL;
-	} while (1);
+		head -= 4;
+	}
+
+	if (!i)
+		return NULL;
 
-	*seqno = ioread32(ring->virtual_start+acthd+4)+1;
-	return &dev_priv->ring[(ring->id + (((ipehr >> 17) & 1) + 1)) % 3];
+	*seqno = ioread32(ring->buffer->virtual_start + head + 4) + 1;
+	return semaphore_wait_to_signaller_ring(ring, ipehr);
 }
 
-static int semaphore_passed(struct intel_ring_buffer *ring)
+static int semaphore_passed(struct intel_engine_cs *ring)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-	struct intel_ring_buffer *signaller;
-	u32 seqno, ctl;
+	struct intel_engine_cs *signaller;
+	u32 seqno;
 
-	ring->hangcheck.deadlock = true;
+	ring->hangcheck.deadlock++;
 
 	signaller = semaphore_waits_for(ring, &seqno);
-	if (signaller == NULL || signaller->hangcheck.deadlock)
+	if (signaller == NULL)
 		return -1;
 
+	/* Prevent pathological recursion due to driver bugs */
+	if (signaller->hangcheck.deadlock >= I915_NUM_RINGS)
+		return -1;
+
+	if (i915_seqno_passed(signaller->get_seqno(signaller, false), seqno))
+		return 1;
+
 	/* cursory check for an unkickable deadlock */
-	ctl = I915_READ_CTL(signaller);
-	if (ctl & RING_WAIT_SEMAPHORE && semaphore_passed(signaller) < 0)
+	if (I915_READ_CTL(signaller) & RING_WAIT_SEMAPHORE &&
+	    semaphore_passed(signaller) < 0)
 		return -1;
 
-	return i915_seqno_passed(signaller->get_seqno(signaller, false), seqno);
+	return 0;
 }
 
 static void semaphore_clear_deadlocks(struct drm_i915_private *dev_priv)
 {
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	int i;
 
 	for_each_ring(ring, dev_priv, i)
-		ring->hangcheck.deadlock = false;
+		ring->hangcheck.deadlock = 0;
 }
 
 static enum intel_ring_hangcheck_action
-ring_stuck(struct intel_ring_buffer *ring, u32 acthd)
+ring_stuck(struct intel_engine_cs *ring, u64 acthd)
 {
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -2427,9 +2897,9 @@ ring_stuck(struct intel_ring_buffer *ring, u32 acthd)
 	 */
 	tmp = I915_READ_CTL(ring);
 	if (tmp & RING_WAIT) {
-		DRM_ERROR("Kicking stuck wait on %s\n",
-			  ring->name);
-		i915_handle_error(dev, false);
+		i915_handle_error(dev, false,
+				  "Kicking stuck wait on %s",
+				  ring->name);
 		I915_WRITE_CTL(ring, tmp);
 		return HANGCHECK_KICK;
 	}
@@ -2439,9 +2909,9 @@ ring_stuck(struct intel_ring_buffer *ring, u32 acthd)
 		default:
 			return HANGCHECK_HUNG;
 		case 1:
-			DRM_ERROR("Kicking stuck semaphore on %s\n",
-				  ring->name);
-			i915_handle_error(dev, false);
+			i915_handle_error(dev, false,
+					  "Kicking stuck semaphore on %s",
+					  ring->name);
 			I915_WRITE_CTL(ring, tmp);
 			return HANGCHECK_KICK;
 		case 0:
@@ -2463,21 +2933,21 @@ ring_stuck(struct intel_ring_buffer *ring, u32 acthd)
 static void i915_hangcheck_elapsed(unsigned long data)
 {
 	struct drm_device *dev = (struct drm_device *)data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring;
 	int i;
 	int busy_count = 0, rings_hung = 0;
 	bool stuck[I915_NUM_RINGS] = { 0 };
 #define BUSY 1
 #define KICK 5
 #define HUNG 20
-#define FIRE 30
 
-	if (!i915_enable_hangcheck)
+	if (!i915.enable_hangcheck)
 		return;
 
 	for_each_ring(ring, dev_priv, i) {
-		u32 seqno, acthd;
+		u64 acthd;
+		u32 seqno;
 		bool busy = true;
 
 		semaphore_clear_deadlocks(dev_priv);
@@ -2555,7 +3025,7 @@ static void i915_hangcheck_elapsed(unsigned long data)
 	}
 
 	for_each_ring(ring, dev_priv, i) {
-		if (ring->hangcheck.score > FIRE) {
+		if (ring->hangcheck.score >= HANGCHECK_SCORE_RING_HUNG) {
 			DRM_INFO("%s on %s\n",
 				 stuck[i] ? "stuck" : "no progress",
 				 ring->name);
@@ -2564,7 +3034,7 @@ static void i915_hangcheck_elapsed(unsigned long data)
 	}
 
 	if (rings_hung)
-		return i915_handle_error(dev, true);
+		return i915_handle_error(dev, true, "Ring hung");
 
 	if (busy_count)
 		/* Reset timer case chip hangs without another request
@@ -2575,75 +3045,77 @@ static void i915_hangcheck_elapsed(unsigned long data)
 void i915_queue_hangcheck(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	if (!i915_enable_hangcheck)
+	if (!i915.enable_hangcheck)
 		return;
 
 	mod_timer(&dev_priv->gpu_error.hangcheck_timer,
 		  round_jiffies_up(jiffies + DRM_I915_HANGCHECK_JIFFIES));
 }
 
-static void ibx_irq_preinstall(struct drm_device *dev)
+static void ibx_irq_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	if (HAS_PCH_NOP(dev))
 		return;
 
-	/* south display irq */
-	I915_WRITE(SDEIMR, 0xffffffff);
-	/*
-	 * SDEIER is also touched by the interrupt handler to work around missed
-	 * PCH interrupts. Hence we can't update it after the interrupt handler
-	 * is enabled - instead we unconditionally enable all PCH interrupt
-	 * sources here, but then only unmask them as needed with SDEIMR.
-	 */
+	GEN5_IRQ_RESET(SDE);
+
+	if (HAS_PCH_CPT(dev) || HAS_PCH_LPT(dev))
+		I915_WRITE(SERR_INT, 0xffffffff);
+}
+
+/*
+ * SDEIER is also touched by the interrupt handler to work around missed PCH
+ * interrupts. Hence we can't update it after the interrupt handler is enabled -
+ * instead we unconditionally enable all PCH interrupt sources here, but then
+ * only unmask them as needed with SDEIMR.
+ *
+ * This function needs to be called before interrupts are enabled.
+ */
+static void ibx_irq_pre_postinstall(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (HAS_PCH_NOP(dev))
+		return;
+
+	WARN_ON(I915_READ(SDEIER) != 0);
 	I915_WRITE(SDEIER, 0xffffffff);
 	POSTING_READ(SDEIER);
 }
 
-static void gen5_gt_irq_preinstall(struct drm_device *dev)
+static void gen5_gt_irq_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	/* and GT */
-	I915_WRITE(GTIMR, 0xffffffff);
-	I915_WRITE(GTIER, 0x0);
-	POSTING_READ(GTIER);
-
-	if (INTEL_INFO(dev)->gen >= 6) {
-		/* and PM */
-		I915_WRITE(GEN6_PMIMR, 0xffffffff);
-		I915_WRITE(GEN6_PMIER, 0x0);
-		POSTING_READ(GEN6_PMIER);
-	}
+	GEN5_IRQ_RESET(GT);
+	if (INTEL_INFO(dev)->gen >= 6)
+		GEN5_IRQ_RESET(GEN6_PM);
 }
 
 /* drm_dma.h hooks
 */
-static void ironlake_irq_preinstall(struct drm_device *dev)
+static void ironlake_irq_reset(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
-
-	atomic_set(&dev_priv->irq_received, 0);
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	I915_WRITE(HWSTAM, 0xeffe);
+	I915_WRITE(HWSTAM, 0xffffffff);
 
-	I915_WRITE(DEIMR, 0xffffffff);
-	I915_WRITE(DEIER, 0x0);
-	POSTING_READ(DEIER);
+	GEN5_IRQ_RESET(DE);
+	if (IS_GEN7(dev))
+		I915_WRITE(GEN7_ERR_INT, 0xffffffff);
 
-	gen5_gt_irq_preinstall(dev);
+	gen5_gt_irq_reset(dev);
 
-	ibx_irq_preinstall(dev);
+	ibx_irq_reset(dev);
 }
 
 static void valleyview_irq_preinstall(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	atomic_set(&dev_priv->irq_received, 0);
-
 	/* VLV magic */
 	I915_WRITE(VLV_IMR, 0);
 	I915_WRITE(RING_IMR(RENDER_RING_BASE), 0);
@@ -2654,7 +3126,7 @@ static void valleyview_irq_preinstall(struct drm_device *dev)
 	I915_WRITE(GTIIR, I915_READ(GTIIR));
 	I915_WRITE(GTIIR, I915_READ(GTIIR));
 
-	gen5_gt_irq_preinstall(dev);
+	gen5_gt_irq_reset(dev);
 
 	I915_WRITE(DPINVGTT, 0xff);
 
@@ -2668,58 +3140,65 @@ static void valleyview_irq_preinstall(struct drm_device *dev)
 	POSTING_READ(VLV_IER);
 }
 
-static void gen8_irq_preinstall(struct drm_device *dev)
+static void gen8_gt_irq_reset(struct drm_i915_private *dev_priv)
+{
+	GEN8_IRQ_RESET_NDX(GT, 0);
+	GEN8_IRQ_RESET_NDX(GT, 1);
+	GEN8_IRQ_RESET_NDX(GT, 2);
+	GEN8_IRQ_RESET_NDX(GT, 3);
+}
+
+static void gen8_irq_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	atomic_set(&dev_priv->irq_received, 0);
-
 	I915_WRITE(GEN8_MASTER_IRQ, 0);
 	POSTING_READ(GEN8_MASTER_IRQ);
 
-	/* IIR can theoretically queue up two events. Be paranoid */
-#define GEN8_IRQ_INIT_NDX(type, which) do { \
-		I915_WRITE(GEN8_##type##_IMR(which), 0xffffffff); \
-		POSTING_READ(GEN8_##type##_IMR(which)); \
-		I915_WRITE(GEN8_##type##_IER(which), 0); \
-		I915_WRITE(GEN8_##type##_IIR(which), 0xffffffff); \
-		POSTING_READ(GEN8_##type##_IIR(which)); \
-		I915_WRITE(GEN8_##type##_IIR(which), 0xffffffff); \
-	} while (0)
-
-#define GEN8_IRQ_INIT(type) do { \
-		I915_WRITE(GEN8_##type##_IMR, 0xffffffff); \
-		POSTING_READ(GEN8_##type##_IMR); \
-		I915_WRITE(GEN8_##type##_IER, 0); \
-		I915_WRITE(GEN8_##type##_IIR, 0xffffffff); \
-		POSTING_READ(GEN8_##type##_IIR); \
-		I915_WRITE(GEN8_##type##_IIR, 0xffffffff); \
-	} while (0)
-
-	GEN8_IRQ_INIT_NDX(GT, 0);
-	GEN8_IRQ_INIT_NDX(GT, 1);
-	GEN8_IRQ_INIT_NDX(GT, 2);
-	GEN8_IRQ_INIT_NDX(GT, 3);
+	gen8_gt_irq_reset(dev_priv);
 
-	for_each_pipe(pipe) {
-		GEN8_IRQ_INIT_NDX(DE_PIPE, pipe);
-	}
+	for_each_pipe(pipe)
+		GEN8_IRQ_RESET_NDX(DE_PIPE, pipe);
+
+	GEN5_IRQ_RESET(GEN8_DE_PORT_);
+	GEN5_IRQ_RESET(GEN8_DE_MISC_);
+	GEN5_IRQ_RESET(GEN8_PCU_);
 
-	GEN8_IRQ_INIT(DE_PORT);
-	GEN8_IRQ_INIT(DE_MISC);
-	GEN8_IRQ_INIT(PCU);
-#undef GEN8_IRQ_INIT
-#undef GEN8_IRQ_INIT_NDX
+	ibx_irq_reset(dev);
+}
+
+static void cherryview_irq_preinstall(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe;
+
+	I915_WRITE(GEN8_MASTER_IRQ, 0);
+	POSTING_READ(GEN8_MASTER_IRQ);
+
+	gen8_gt_irq_reset(dev_priv);
+
+	GEN5_IRQ_RESET(GEN8_PCU_);
 
 	POSTING_READ(GEN8_PCU_IIR);
 
-	ibx_irq_preinstall(dev);
+	I915_WRITE(DPINVGTT, DPINVGTT_STATUS_MASK_CHV);
+
+	I915_WRITE(PORT_HOTPLUG_EN, 0);
+	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
+
+	for_each_pipe(pipe)
+		I915_WRITE(PIPESTAT(pipe), 0xffff);
+
+	I915_WRITE(VLV_IMR, 0xffffffff);
+	I915_WRITE(VLV_IER, 0x0);
+	I915_WRITE(VLV_IIR, 0xffffffff);
+	POSTING_READ(VLV_IIR);
 }
 
 static void ibx_hpd_irq_setup(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *intel_encoder;
 	u32 hotplug_irqs, hotplug, enabled_irqs = 0;
@@ -2754,22 +3233,18 @@ static void ibx_hpd_irq_setup(struct drm_device *dev)
 
 static void ibx_irq_postinstall(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 mask;
 
 	if (HAS_PCH_NOP(dev))
 		return;
 
-	if (HAS_PCH_IBX(dev)) {
-		mask = SDE_GMBUS | SDE_AUX_MASK | SDE_TRANSB_FIFO_UNDER |
-		       SDE_TRANSA_FIFO_UNDER | SDE_POISON;
-	} else {
-		mask = SDE_GMBUS_CPT | SDE_AUX_MASK_CPT | SDE_ERROR_CPT;
-
-		I915_WRITE(SERR_INT, I915_READ(SERR_INT));
-	}
+	if (HAS_PCH_IBX(dev))
+		mask = SDE_GMBUS | SDE_AUX_MASK | SDE_POISON;
+	else
+		mask = SDE_GMBUS_CPT | SDE_AUX_MASK_CPT;
 
-	I915_WRITE(SDEIIR, I915_READ(SDEIIR));
+	GEN5_ASSERT_IIR_IS_ZERO(SDEIIR);
 	I915_WRITE(SDEIMR, ~mask);
 }
 
@@ -2795,58 +3270,49 @@ static void gen5_gt_irq_postinstall(struct drm_device *dev)
 		gt_irqs |= GT_BLT_USER_INTERRUPT | GT_BSD_USER_INTERRUPT;
 	}
 
-	I915_WRITE(GTIIR, I915_READ(GTIIR));
-	I915_WRITE(GTIMR, dev_priv->gt_irq_mask);
-	I915_WRITE(GTIER, gt_irqs);
-	POSTING_READ(GTIER);
+	GEN5_IRQ_INIT(GT, dev_priv->gt_irq_mask, gt_irqs);
 
 	if (INTEL_INFO(dev)->gen >= 6) {
-		pm_irqs |= GEN6_PM_RPS_EVENTS;
+		pm_irqs |= dev_priv->pm_rps_events;
 
 		if (HAS_VEBOX(dev))
 			pm_irqs |= PM_VEBOX_USER_INTERRUPT;
 
 		dev_priv->pm_irq_mask = 0xffffffff;
-		I915_WRITE(GEN6_PMIIR, I915_READ(GEN6_PMIIR));
-		I915_WRITE(GEN6_PMIMR, dev_priv->pm_irq_mask);
-		I915_WRITE(GEN6_PMIER, pm_irqs);
-		POSTING_READ(GEN6_PMIER);
+		GEN5_IRQ_INIT(GEN6_PM, dev_priv->pm_irq_mask, pm_irqs);
 	}
 }
 
 static int ironlake_irq_postinstall(struct drm_device *dev)
 {
 	unsigned long irqflags;
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 display_mask, extra_mask;
 
 	if (INTEL_INFO(dev)->gen >= 7) {
 		display_mask = (DE_MASTER_IRQ_CONTROL | DE_GSE_IVB |
 				DE_PCH_EVENT_IVB | DE_PLANEC_FLIP_DONE_IVB |
 				DE_PLANEB_FLIP_DONE_IVB |
-				DE_PLANEA_FLIP_DONE_IVB | DE_AUX_CHANNEL_A_IVB |
-				DE_ERR_INT_IVB);
+				DE_PLANEA_FLIP_DONE_IVB | DE_AUX_CHANNEL_A_IVB);
 		extra_mask = (DE_PIPEC_VBLANK_IVB | DE_PIPEB_VBLANK_IVB |
-			      DE_PIPEA_VBLANK_IVB);
-
-		I915_WRITE(GEN7_ERR_INT, I915_READ(GEN7_ERR_INT));
+			      DE_PIPEA_VBLANK_IVB | DE_ERR_INT_IVB);
 	} else {
 		display_mask = (DE_MASTER_IRQ_CONTROL | DE_GSE | DE_PCH_EVENT |
 				DE_PLANEA_FLIP_DONE | DE_PLANEB_FLIP_DONE |
 				DE_AUX_CHANNEL_A |
-				DE_PIPEB_FIFO_UNDERRUN | DE_PIPEA_FIFO_UNDERRUN |
 				DE_PIPEB_CRC_DONE | DE_PIPEA_CRC_DONE |
 				DE_POISON);
-		extra_mask = DE_PIPEA_VBLANK | DE_PIPEB_VBLANK | DE_PCU_EVENT;
+		extra_mask = DE_PIPEA_VBLANK | DE_PIPEB_VBLANK | DE_PCU_EVENT |
+				DE_PIPEB_FIFO_UNDERRUN | DE_PIPEA_FIFO_UNDERRUN;
 	}
 
 	dev_priv->irq_mask = ~display_mask;
 
-	/* should always can generate irq */
-	I915_WRITE(DEIIR, I915_READ(DEIIR));
-	I915_WRITE(DEIMR, dev_priv->irq_mask);
-	I915_WRITE(DEIER, display_mask | extra_mask);
-	POSTING_READ(DEIER);
+	I915_WRITE(HWSTAM, 0xeffe);
+
+	ibx_irq_pre_postinstall(dev);
+
+	GEN5_IRQ_INIT(DE, dev_priv->irq_mask, display_mask | extra_mask);
 
 	gen5_gt_irq_postinstall(dev);
 
@@ -2866,44 +3332,113 @@ static int ironlake_irq_postinstall(struct drm_device *dev)
 	return 0;
 }
 
+static void valleyview_display_irqs_install(struct drm_i915_private *dev_priv)
+{
+	u32 pipestat_mask;
+	u32 iir_mask;
+
+	pipestat_mask = PIPESTAT_INT_STATUS_MASK |
+			PIPE_FIFO_UNDERRUN_STATUS;
+
+	I915_WRITE(PIPESTAT(PIPE_A), pipestat_mask);
+	I915_WRITE(PIPESTAT(PIPE_B), pipestat_mask);
+	POSTING_READ(PIPESTAT(PIPE_A));
+
+	pipestat_mask = PLANE_FLIP_DONE_INT_STATUS_VLV |
+			PIPE_CRC_DONE_INTERRUPT_STATUS;
+
+	i915_enable_pipestat(dev_priv, PIPE_A, pipestat_mask |
+					       PIPE_GMBUS_INTERRUPT_STATUS);
+	i915_enable_pipestat(dev_priv, PIPE_B, pipestat_mask);
+
+	iir_mask = I915_DISPLAY_PORT_INTERRUPT |
+		   I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
+		   I915_DISPLAY_PIPE_B_EVENT_INTERRUPT;
+	dev_priv->irq_mask &= ~iir_mask;
+
+	I915_WRITE(VLV_IIR, iir_mask);
+	I915_WRITE(VLV_IIR, iir_mask);
+	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
+	I915_WRITE(VLV_IER, ~dev_priv->irq_mask);
+	POSTING_READ(VLV_IER);
+}
+
+static void valleyview_display_irqs_uninstall(struct drm_i915_private *dev_priv)
+{
+	u32 pipestat_mask;
+	u32 iir_mask;
+
+	iir_mask = I915_DISPLAY_PORT_INTERRUPT |
+		   I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
+		   I915_DISPLAY_PIPE_B_EVENT_INTERRUPT;
+
+	dev_priv->irq_mask |= iir_mask;
+	I915_WRITE(VLV_IER, ~dev_priv->irq_mask);
+	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
+	I915_WRITE(VLV_IIR, iir_mask);
+	I915_WRITE(VLV_IIR, iir_mask);
+	POSTING_READ(VLV_IIR);
+
+	pipestat_mask = PLANE_FLIP_DONE_INT_STATUS_VLV |
+			PIPE_CRC_DONE_INTERRUPT_STATUS;
+
+	i915_disable_pipestat(dev_priv, PIPE_A, pipestat_mask |
+					        PIPE_GMBUS_INTERRUPT_STATUS);
+	i915_disable_pipestat(dev_priv, PIPE_B, pipestat_mask);
+
+	pipestat_mask = PIPESTAT_INT_STATUS_MASK |
+			PIPE_FIFO_UNDERRUN_STATUS;
+	I915_WRITE(PIPESTAT(PIPE_A), pipestat_mask);
+	I915_WRITE(PIPESTAT(PIPE_B), pipestat_mask);
+	POSTING_READ(PIPESTAT(PIPE_A));
+}
+
+void valleyview_enable_display_irqs(struct drm_i915_private *dev_priv)
+{
+	assert_spin_locked(&dev_priv->irq_lock);
+
+	if (dev_priv->display_irqs_enabled)
+		return;
+
+	dev_priv->display_irqs_enabled = true;
+
+	if (dev_priv->dev->irq_enabled)
+		valleyview_display_irqs_install(dev_priv);
+}
+
+void valleyview_disable_display_irqs(struct drm_i915_private *dev_priv)
+{
+	assert_spin_locked(&dev_priv->irq_lock);
+
+	if (!dev_priv->display_irqs_enabled)
+		return;
+
+	dev_priv->display_irqs_enabled = false;
+
+	if (dev_priv->dev->irq_enabled)
+		valleyview_display_irqs_uninstall(dev_priv);
+}
+
 static int valleyview_irq_postinstall(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
-	u32 enable_mask;
-	u32 pipestat_enable = PLANE_FLIP_DONE_INT_EN_VLV |
-		PIPE_CRC_DONE_ENABLE;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
 
-	enable_mask = I915_DISPLAY_PORT_INTERRUPT;
-	enable_mask |= I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
-		I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT |
-		I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
-		I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
-
-	/*
-	 *Leave vblank interrupts masked initially.  enable/disable will
-	 * toggle them based on usage.
-	 */
-	dev_priv->irq_mask = (~enable_mask) |
-		I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT |
-		I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
+	dev_priv->irq_mask = ~0;
 
 	I915_WRITE(PORT_HOTPLUG_EN, 0);
 	POSTING_READ(PORT_HOTPLUG_EN);
 
 	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
-	I915_WRITE(VLV_IER, enable_mask);
+	I915_WRITE(VLV_IER, ~dev_priv->irq_mask);
 	I915_WRITE(VLV_IIR, 0xffffffff);
-	I915_WRITE(PIPESTAT(0), 0xffff);
-	I915_WRITE(PIPESTAT(1), 0xffff);
 	POSTING_READ(VLV_IER);
 
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-	i915_enable_pipestat(dev_priv, PIPE_A, pipestat_enable);
-	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_EVENT_ENABLE);
-	i915_enable_pipestat(dev_priv, PIPE_B, pipestat_enable);
+	if (dev_priv->display_irqs_enabled)
+		valleyview_display_irqs_install(dev_priv);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
 	I915_WRITE(VLV_IIR, 0xffffffff);
@@ -2937,49 +3472,38 @@ static void gen8_gt_irq_postinstall(struct drm_i915_private *dev_priv)
 		GT_RENDER_USER_INTERRUPT << GEN8_VECS_IRQ_SHIFT
 		};
 
-	for (i = 0; i < ARRAY_SIZE(gt_interrupts); i++) {
-		u32 tmp = I915_READ(GEN8_GT_IIR(i));
-		if (tmp)
-			DRM_ERROR("Interrupt (%d) should have been masked in pre-install 0x%08x\n",
-				  i, tmp);
-		I915_WRITE(GEN8_GT_IMR(i), ~gt_interrupts[i]);
-		I915_WRITE(GEN8_GT_IER(i), gt_interrupts[i]);
-	}
-	POSTING_READ(GEN8_GT_IER(0));
+	for (i = 0; i < ARRAY_SIZE(gt_interrupts); i++)
+		GEN8_IRQ_INIT_NDX(GT, i, ~gt_interrupts[i], gt_interrupts[i]);
+
+	dev_priv->pm_irq_mask = 0xffffffff;
 }
 
 static void gen8_de_irq_postinstall(struct drm_i915_private *dev_priv)
 {
 	struct drm_device *dev = dev_priv->dev;
-	uint32_t de_pipe_masked = GEN8_PIPE_FLIP_DONE |
+	uint32_t de_pipe_masked = GEN8_PIPE_PRIMARY_FLIP_DONE |
 		GEN8_PIPE_CDCLK_CRC_DONE |
-		GEN8_PIPE_FIFO_UNDERRUN |
 		GEN8_DE_PIPE_IRQ_FAULT_ERRORS;
-	uint32_t de_pipe_enables = de_pipe_masked | GEN8_PIPE_VBLANK;
+	uint32_t de_pipe_enables = de_pipe_masked | GEN8_PIPE_VBLANK |
+		GEN8_PIPE_FIFO_UNDERRUN;
 	int pipe;
 	dev_priv->de_irq_mask[PIPE_A] = ~de_pipe_masked;
 	dev_priv->de_irq_mask[PIPE_B] = ~de_pipe_masked;
 	dev_priv->de_irq_mask[PIPE_C] = ~de_pipe_masked;
 
-	for_each_pipe(pipe) {
-		u32 tmp = I915_READ(GEN8_DE_PIPE_IIR(pipe));
-		if (tmp)
-			DRM_ERROR("Interrupt (%d) should have been masked in pre-install 0x%08x\n",
-				  pipe, tmp);
-		I915_WRITE(GEN8_DE_PIPE_IMR(pipe), dev_priv->de_irq_mask[pipe]);
-		I915_WRITE(GEN8_DE_PIPE_IER(pipe), de_pipe_enables);
-	}
-	POSTING_READ(GEN8_DE_PIPE_ISR(0));
+	for_each_pipe(pipe)
+		GEN8_IRQ_INIT_NDX(DE_PIPE, pipe, dev_priv->de_irq_mask[pipe],
+				  de_pipe_enables);
 
-	I915_WRITE(GEN8_DE_PORT_IMR, ~GEN8_AUX_CHANNEL_A);
-	I915_WRITE(GEN8_DE_PORT_IER, GEN8_AUX_CHANNEL_A);
-	POSTING_READ(GEN8_DE_PORT_IER);
+	GEN5_IRQ_INIT(GEN8_DE_PORT_, ~GEN8_AUX_CHANNEL_A, GEN8_AUX_CHANNEL_A);
 }
 
 static int gen8_irq_postinstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
+	ibx_irq_pre_postinstall(dev);
+
 	gen8_gt_irq_postinstall(dev_priv);
 	gen8_de_irq_postinstall(dev_priv);
 
@@ -2991,57 +3515,69 @@ static int gen8_irq_postinstall(struct drm_device *dev)
 	return 0;
 }
 
-static void gen8_irq_uninstall(struct drm_device *dev)
+static int cherryview_irq_postinstall(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 enable_mask = I915_DISPLAY_PORT_INTERRUPT |
+		I915_DISPLAY_PIPE_A_EVENT_INTERRUPT |
+		I915_DISPLAY_PIPE_B_EVENT_INTERRUPT |
+		I915_DISPLAY_PIPE_C_EVENT_INTERRUPT;
+	u32 pipestat_enable = PLANE_FLIP_DONE_INT_STATUS_VLV |
+		PIPE_CRC_DONE_INTERRUPT_STATUS;
+	unsigned long irqflags;
 	int pipe;
 
-	if (!dev_priv)
-		return;
+	/*
+	 * Leave vblank interrupts masked initially.  enable/disable will
+	 * toggle them based on usage.
+	 */
+	dev_priv->irq_mask = ~enable_mask;
 
-	atomic_set(&dev_priv->irq_received, 0);
+	for_each_pipe(pipe)
+		I915_WRITE(PIPESTAT(pipe), 0xffff);
 
-	I915_WRITE(GEN8_MASTER_IRQ, 0);
+	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
+	for_each_pipe(pipe)
+		i915_enable_pipestat(dev_priv, pipe, pipestat_enable);
+	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
-#define GEN8_IRQ_FINI_NDX(type, which) do { \
-		I915_WRITE(GEN8_##type##_IMR(which), 0xffffffff); \
-		I915_WRITE(GEN8_##type##_IER(which), 0); \
-		I915_WRITE(GEN8_##type##_IIR(which), 0xffffffff); \
-	} while (0)
+	I915_WRITE(VLV_IIR, 0xffffffff);
+	I915_WRITE(VLV_IMR, dev_priv->irq_mask);
+	I915_WRITE(VLV_IER, enable_mask);
 
-#define GEN8_IRQ_FINI(type) do { \
-		I915_WRITE(GEN8_##type##_IMR, 0xffffffff); \
-		I915_WRITE(GEN8_##type##_IER, 0); \
-		I915_WRITE(GEN8_##type##_IIR, 0xffffffff); \
-	} while (0)
+	gen8_gt_irq_postinstall(dev_priv);
 
-	GEN8_IRQ_FINI_NDX(GT, 0);
-	GEN8_IRQ_FINI_NDX(GT, 1);
-	GEN8_IRQ_FINI_NDX(GT, 2);
-	GEN8_IRQ_FINI_NDX(GT, 3);
+	I915_WRITE(GEN8_MASTER_IRQ, MASTER_INTERRUPT_ENABLE);
+	POSTING_READ(GEN8_MASTER_IRQ);
 
-	for_each_pipe(pipe) {
-		GEN8_IRQ_FINI_NDX(DE_PIPE, pipe);
-	}
+	return 0;
+}
 
-	GEN8_IRQ_FINI(DE_PORT);
-	GEN8_IRQ_FINI(DE_MISC);
-	GEN8_IRQ_FINI(PCU);
-#undef GEN8_IRQ_FINI
-#undef GEN8_IRQ_FINI_NDX
+static void gen8_irq_uninstall(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	POSTING_READ(GEN8_PCU_IIR);
+	if (!dev_priv)
+		return;
+
+	intel_hpd_irq_uninstall(dev_priv);
+
+	gen8_irq_reset(dev);
 }
 
 static void valleyview_irq_uninstall(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long irqflags;
 	int pipe;
 
 	if (!dev_priv)
 		return;
 
-	del_timer_sync(&dev_priv->hotplug_reenable_timer);
+	I915_WRITE(VLV_MASTER_IER, 0);
+
+	intel_hpd_irq_uninstall(dev_priv);
 
 	for_each_pipe(pipe)
 		I915_WRITE(PIPESTAT(pipe), 0xffff);
@@ -3049,52 +3585,88 @@ static void valleyview_irq_uninstall(struct drm_device *dev)
 	I915_WRITE(HWSTAM, 0xffffffff);
 	I915_WRITE(PORT_HOTPLUG_EN, 0);
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
-	for_each_pipe(pipe)
-		I915_WRITE(PIPESTAT(pipe), 0xffff);
+
+	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+	if (dev_priv->display_irqs_enabled)
+		valleyview_display_irqs_uninstall(dev_priv);
+	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+
+	dev_priv->irq_mask = 0;
+
 	I915_WRITE(VLV_IIR, 0xffffffff);
 	I915_WRITE(VLV_IMR, 0xffffffff);
 	I915_WRITE(VLV_IER, 0x0);
 	POSTING_READ(VLV_IER);
 }
 
-static void ironlake_irq_uninstall(struct drm_device *dev)
+static void cherryview_irq_uninstall(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe;
 
 	if (!dev_priv)
 		return;
 
-	del_timer_sync(&dev_priv->hotplug_reenable_timer);
+	I915_WRITE(GEN8_MASTER_IRQ, 0);
+	POSTING_READ(GEN8_MASTER_IRQ);
 
-	I915_WRITE(HWSTAM, 0xffffffff);
+#define GEN8_IRQ_FINI_NDX(type, which)				\
+do {								\
+	I915_WRITE(GEN8_##type##_IMR(which), 0xffffffff);	\
+	I915_WRITE(GEN8_##type##_IER(which), 0);		\
+	I915_WRITE(GEN8_##type##_IIR(which), 0xffffffff);	\
+	POSTING_READ(GEN8_##type##_IIR(which));			\
+	I915_WRITE(GEN8_##type##_IIR(which), 0xffffffff);	\
+} while (0)
+
+#define GEN8_IRQ_FINI(type)				\
+do {							\
+	I915_WRITE(GEN8_##type##_IMR, 0xffffffff);	\
+	I915_WRITE(GEN8_##type##_IER, 0);		\
+	I915_WRITE(GEN8_##type##_IIR, 0xffffffff);	\
+	POSTING_READ(GEN8_##type##_IIR);		\
+	I915_WRITE(GEN8_##type##_IIR, 0xffffffff);	\
+} while (0)
 
-	I915_WRITE(DEIMR, 0xffffffff);
-	I915_WRITE(DEIER, 0x0);
-	I915_WRITE(DEIIR, I915_READ(DEIIR));
-	if (IS_GEN7(dev))
-		I915_WRITE(GEN7_ERR_INT, I915_READ(GEN7_ERR_INT));
+	GEN8_IRQ_FINI_NDX(GT, 0);
+	GEN8_IRQ_FINI_NDX(GT, 1);
+	GEN8_IRQ_FINI_NDX(GT, 2);
+	GEN8_IRQ_FINI_NDX(GT, 3);
 
-	I915_WRITE(GTIMR, 0xffffffff);
-	I915_WRITE(GTIER, 0x0);
-	I915_WRITE(GTIIR, I915_READ(GTIIR));
+	GEN8_IRQ_FINI(PCU);
 
-	if (HAS_PCH_NOP(dev))
+#undef GEN8_IRQ_FINI
+#undef GEN8_IRQ_FINI_NDX
+
+	I915_WRITE(PORT_HOTPLUG_EN, 0);
+	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
+
+	for_each_pipe(pipe)
+		I915_WRITE(PIPESTAT(pipe), 0xffff);
+
+	I915_WRITE(VLV_IMR, 0xffffffff);
+	I915_WRITE(VLV_IER, 0x0);
+	I915_WRITE(VLV_IIR, 0xffffffff);
+	POSTING_READ(VLV_IIR);
+}
+
+static void ironlake_irq_uninstall(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (!dev_priv)
 		return;
 
-	I915_WRITE(SDEIMR, 0xffffffff);
-	I915_WRITE(SDEIER, 0x0);
-	I915_WRITE(SDEIIR, I915_READ(SDEIIR));
-	if (HAS_PCH_CPT(dev) || HAS_PCH_LPT(dev))
-		I915_WRITE(SERR_INT, I915_READ(SERR_INT));
+	intel_hpd_irq_uninstall(dev_priv);
+
+	ironlake_irq_reset(dev);
 }
 
 static void i8xx_irq_preinstall(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	atomic_set(&dev_priv->irq_received, 0);
-
 	for_each_pipe(pipe)
 		I915_WRITE(PIPESTAT(pipe), 0);
 	I915_WRITE16(IMR, 0xffff);
@@ -3104,7 +3676,7 @@ static void i8xx_irq_preinstall(struct drm_device * dev)
 
 static int i8xx_irq_postinstall(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
 
 	I915_WRITE16(EMR,
@@ -3129,8 +3701,8 @@ static int i8xx_irq_postinstall(struct drm_device *dev)
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_ENABLE);
-	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_ENABLE);
+	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_INTERRUPT_STATUS);
+	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_INTERRUPT_STATUS);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
 	return 0;
@@ -3140,18 +3712,18 @@ static int i8xx_irq_postinstall(struct drm_device *dev)
  * Returns true when a page flip has completed.
  */
 static bool i8xx_handle_vblank(struct drm_device *dev,
-			       int pipe, u16 iir)
+			       int plane, int pipe, u32 iir)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	u16 flip_pending = DISPLAY_PLANE_FLIP_PENDING(pipe);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u16 flip_pending = DISPLAY_PLANE_FLIP_PENDING(plane);
 
-	if (!drm_handle_vblank(dev, pipe))
+	if (!intel_pipe_handle_vblank(dev, pipe))
 		return false;
 
 	if ((iir & flip_pending) == 0)
 		return false;
 
-	intel_prepare_page_flip(dev, pipe);
+	intel_prepare_page_flip(dev, plane);
 
 	/* We detect FlipDone by looking for the change in PendingFlip from '1'
 	 * to '0' on the following vblank, i.e. IIR has the Pendingflip
@@ -3169,8 +3741,8 @@ static bool i8xx_handle_vblank(struct drm_device *dev,
 
 static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 {
-	struct drm_device *dev = (struct drm_device *) arg;
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_device *dev = arg;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u16 iir, new_iir;
 	u32 pipe_stats[2];
 	unsigned long irqflags;
@@ -3179,8 +3751,6 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 		I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT |
 		I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT;
 
-	atomic_inc(&dev_priv->irq_received);
-
 	iir = I915_READ16(IIR);
 	if (iir == 0)
 		return IRQ_NONE;
@@ -3193,7 +3763,9 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 		 */
 		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
-			i915_handle_error(dev, false);
+			i915_handle_error(dev, false,
+					  "Command parser error, iir 0x%08x",
+					  iir);
 
 		for_each_pipe(pipe) {
 			int reg = PIPESTAT(pipe);
@@ -3202,12 +3774,8 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 			/*
 			 * Clear the PIPE*STAT regs before the IIR
 			 */
-			if (pipe_stats[pipe] & 0x8000ffff) {
-				if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
-					DRM_DEBUG_DRIVER("pipe %c underrun\n",
-							 pipe_name(pipe));
+			if (pipe_stats[pipe] & 0x8000ffff)
 				I915_WRITE(reg, pipe_stats[pipe]);
-			}
 		}
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
@@ -3220,12 +3788,20 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 			notify_ring(dev, &dev_priv->ring[RCS]);
 
 		for_each_pipe(pipe) {
+			int plane = pipe;
+			if (HAS_FBC(dev))
+				plane = !plane;
+
 			if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS &&
-			    i8xx_handle_vblank(dev, pipe, iir))
-				flip_mask &= ~DISPLAY_PLANE_FLIP_PENDING(pipe);
+			    i8xx_handle_vblank(dev, plane, pipe, iir))
+				flip_mask &= ~DISPLAY_PLANE_FLIP_PENDING(plane);
 
 			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 				i9xx_pipe_crc_irq_handler(dev, pipe);
+
+			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS &&
+			    intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
+				DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
 		}
 
 		iir = new_iir;
@@ -3236,7 +3812,7 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 
 static void i8xx_irq_uninstall(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
 	for_each_pipe(pipe) {
@@ -3251,11 +3827,9 @@ static void i8xx_irq_uninstall(struct drm_device * dev)
 
 static void i915_irq_preinstall(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	atomic_set(&dev_priv->irq_received, 0);
-
 	if (I915_HAS_HOTPLUG(dev)) {
 		I915_WRITE(PORT_HOTPLUG_EN, 0);
 		I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
@@ -3271,7 +3845,7 @@ static void i915_irq_preinstall(struct drm_device * dev)
 
 static int i915_irq_postinstall(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 enable_mask;
 	unsigned long irqflags;
 
@@ -3312,8 +3886,8 @@ static int i915_irq_postinstall(struct drm_device *dev)
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_ENABLE);
-	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_ENABLE);
+	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_INTERRUPT_STATUS);
+	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_INTERRUPT_STATUS);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
 	return 0;
@@ -3325,10 +3899,10 @@ static int i915_irq_postinstall(struct drm_device *dev)
 static bool i915_handle_vblank(struct drm_device *dev,
 			       int plane, int pipe, u32 iir)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 flip_pending = DISPLAY_PLANE_FLIP_PENDING(plane);
 
-	if (!drm_handle_vblank(dev, pipe))
+	if (!intel_pipe_handle_vblank(dev, pipe))
 		return false;
 
 	if ((iir & flip_pending) == 0)
@@ -3352,8 +3926,8 @@ static bool i915_handle_vblank(struct drm_device *dev,
 
 static irqreturn_t i915_irq_handler(int irq, void *arg)
 {
-	struct drm_device *dev = (struct drm_device *) arg;
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_device *dev = arg;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 iir, new_iir, pipe_stats[I915_MAX_PIPES];
 	unsigned long irqflags;
 	u32 flip_mask =
@@ -3361,8 +3935,6 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 		I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT;
 	int pipe, ret = IRQ_NONE;
 
-	atomic_inc(&dev_priv->irq_received);
-
 	iir = I915_READ(IIR);
 	do {
 		bool irq_received = (iir & ~flip_mask) != 0;
@@ -3375,7 +3947,9 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 		 */
 		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
-			i915_handle_error(dev, false);
+			i915_handle_error(dev, false,
+					  "Command parser error, iir 0x%08x",
+					  iir);
 
 		for_each_pipe(pipe) {
 			int reg = PIPESTAT(pipe);
@@ -3383,9 +3957,6 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 
 			/* Clear the PIPE*STAT regs before the IIR */
 			if (pipe_stats[pipe] & 0x8000ffff) {
-				if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
-					DRM_DEBUG_DRIVER("pipe %c underrun\n",
-							 pipe_name(pipe));
 				I915_WRITE(reg, pipe_stats[pipe]);
 				irq_received = true;
 			}
@@ -3396,19 +3967,9 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 			break;
 
 		/* Consume port.  Then clear IIR or we'll miss events */
-		if ((I915_HAS_HOTPLUG(dev)) &&
-		    (iir & I915_DISPLAY_PORT_INTERRUPT)) {
-			u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT);
-			u32 hotplug_trigger = hotplug_status & HOTPLUG_INT_STATUS_I915;
-
-			DRM_DEBUG_DRIVER("hotplug event received, stat 0x%08x\n",
-				  hotplug_status);
-
-			intel_hpd_irq_handler(dev, hotplug_trigger, hpd_status_i915);
-
-			I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status);
-			POSTING_READ(PORT_HOTPLUG_STAT);
-		}
+		if (I915_HAS_HOTPLUG(dev) &&
+		    iir & I915_DISPLAY_PORT_INTERRUPT)
+			i9xx_hpd_irq_handler(dev);
 
 		I915_WRITE(IIR, iir & ~flip_mask);
 		new_iir = I915_READ(IIR); /* Flush posted writes */
@@ -3418,7 +3979,7 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 
 		for_each_pipe(pipe) {
 			int plane = pipe;
-			if (IS_MOBILE(dev))
+			if (HAS_FBC(dev))
 				plane = !plane;
 
 			if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS &&
@@ -3430,6 +3991,10 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 
 			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 				i9xx_pipe_crc_irq_handler(dev, pipe);
+
+			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS &&
+			    intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
+				DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
 		}
 
 		if (blc_event || (iir & I915_ASLE_INTERRUPT))
@@ -3461,10 +4026,10 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 
 static void i915_irq_uninstall(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	del_timer_sync(&dev_priv->hotplug_reenable_timer);
+	intel_hpd_irq_uninstall(dev_priv);
 
 	if (I915_HAS_HOTPLUG(dev)) {
 		I915_WRITE(PORT_HOTPLUG_EN, 0);
@@ -3485,11 +4050,9 @@ static void i915_irq_uninstall(struct drm_device * dev)
 
 static void i965_irq_preinstall(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
-	atomic_set(&dev_priv->irq_received, 0);
-
 	I915_WRITE(PORT_HOTPLUG_EN, 0);
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
 
@@ -3503,7 +4066,7 @@ static void i965_irq_preinstall(struct drm_device * dev)
 
 static int i965_irq_postinstall(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 enable_mask;
 	u32 error_mask;
 	unsigned long irqflags;
@@ -3528,9 +4091,9 @@ static int i965_irq_postinstall(struct drm_device *dev)
 	/* Interrupt setup is already guaranteed to be single-threaded, this is
 	 * just to make the assert_spin_locked check happy. */
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_EVENT_ENABLE);
-	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_ENABLE);
-	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_ENABLE);
+	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_GMBUS_INTERRUPT_STATUS);
+	i915_enable_pipestat(dev_priv, PIPE_A, PIPE_CRC_DONE_INTERRUPT_STATUS);
+	i915_enable_pipestat(dev_priv, PIPE_B, PIPE_CRC_DONE_INTERRUPT_STATUS);
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 
 	/*
@@ -3562,7 +4125,7 @@ static int i965_irq_postinstall(struct drm_device *dev)
 
 static void i915_hpd_irq_setup(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_mode_config *mode_config = &dev->mode_config;
 	struct intel_encoder *intel_encoder;
 	u32 hotplug_en;
@@ -3593,26 +4156,22 @@ static void i915_hpd_irq_setup(struct drm_device *dev)
 
 static irqreturn_t i965_irq_handler(int irq, void *arg)
 {
-	struct drm_device *dev = (struct drm_device *) arg;
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_device *dev = arg;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 iir, new_iir;
 	u32 pipe_stats[I915_MAX_PIPES];
 	unsigned long irqflags;
-	int irq_received;
 	int ret = IRQ_NONE, pipe;
 	u32 flip_mask =
 		I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT |
 		I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT;
 
-	atomic_inc(&dev_priv->irq_received);
-
 	iir = I915_READ(IIR);
 
 	for (;;) {
+		bool irq_received = (iir & ~flip_mask) != 0;
 		bool blc_event = false;
 
-		irq_received = (iir & ~flip_mask) != 0;
-
 		/* Can't rely on pipestat interrupt bit in iir as it might
 		 * have been cleared after the pipestat interrupt was received.
 		 * It doesn't set the bit in iir again, but it still produces
@@ -3620,7 +4179,9 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 		 */
 		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 		if (iir & I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT)
-			i915_handle_error(dev, false);
+			i915_handle_error(dev, false,
+					  "Command parser error, iir 0x%08x",
+					  iir);
 
 		for_each_pipe(pipe) {
 			int reg = PIPESTAT(pipe);
@@ -3630,11 +4191,8 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 			 * Clear the PIPE*STAT regs before the IIR
 			 */
 			if (pipe_stats[pipe] & 0x8000ffff) {
-				if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS)
-					DRM_DEBUG_DRIVER("pipe %c underrun\n",
-							 pipe_name(pipe));
 				I915_WRITE(reg, pipe_stats[pipe]);
-				irq_received = 1;
+				irq_received = true;
 			}
 		}
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
@@ -3645,21 +4203,8 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 		ret = IRQ_HANDLED;
 
 		/* Consume port.  Then clear IIR or we'll miss events */
-		if (iir & I915_DISPLAY_PORT_INTERRUPT) {
-			u32 hotplug_status = I915_READ(PORT_HOTPLUG_STAT);
-			u32 hotplug_trigger = hotplug_status & (IS_G4X(dev) ?
-								  HOTPLUG_INT_STATUS_G4X :
-								  HOTPLUG_INT_STATUS_I915);
-
-			DRM_DEBUG_DRIVER("hotplug event received, stat 0x%08x\n",
-				  hotplug_status);
-
-			intel_hpd_irq_handler(dev, hotplug_trigger,
-					      IS_G4X(dev) ? hpd_status_gen4 : hpd_status_i915);
-
-			I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status);
-			I915_READ(PORT_HOTPLUG_STAT);
-		}
+		if (iir & I915_DISPLAY_PORT_INTERRUPT)
+			i9xx_hpd_irq_handler(dev);
 
 		I915_WRITE(IIR, iir & ~flip_mask);
 		new_iir = I915_READ(IIR); /* Flush posted writes */
@@ -3679,8 +4224,11 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 
 			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 				i9xx_pipe_crc_irq_handler(dev, pipe);
-		}
 
+			if (pipe_stats[pipe] & PIPE_FIFO_UNDERRUN_STATUS &&
+			    intel_set_cpu_fifo_underrun_reporting(dev, pipe, false))
+				DRM_ERROR("pipe %c underrun\n", pipe_name(pipe));
+		}
 
 		if (blc_event || (iir & I915_ASLE_INTERRUPT))
 			intel_opregion_asle_intr(dev);
@@ -3713,13 +4261,13 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 
 static void i965_irq_uninstall(struct drm_device * dev)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe;
 
 	if (!dev_priv)
 		return;
 
-	del_timer_sync(&dev_priv->hotplug_reenable_timer);
+	intel_hpd_irq_uninstall(dev_priv);
 
 	I915_WRITE(PORT_HOTPLUG_EN, 0);
 	I915_WRITE(PORT_HOTPLUG_STAT, I915_READ(PORT_HOTPLUG_STAT));
@@ -3736,9 +4284,9 @@ static void i965_irq_uninstall(struct drm_device * dev)
 	I915_WRITE(IIR, I915_READ(IIR));
 }
 
-static void i915_reenable_hotplug_timer_func(unsigned long data)
+static void intel_hpd_irq_reenable(unsigned long data)
 {
-	drm_i915_private_t *dev_priv = (drm_i915_private_t *)data;
+	struct drm_i915_private *dev_priv = (struct drm_i915_private *)data;
 	struct drm_device *dev = dev_priv->dev;
 	struct drm_mode_config *mode_config = &dev->mode_config;
 	unsigned long irqflags;
@@ -3759,7 +4307,7 @@ static void i915_reenable_hotplug_timer_func(unsigned long data)
 			if (intel_connector->encoder->hpd_pin == i) {
 				if (connector->polled != intel_connector->polled)
 					DRM_DEBUG_DRIVER("Reenabling HPD on connector %s\n",
-							 drm_get_connector_name(connector));
+							 connector->name);
 				connector->polled = intel_connector->polled;
 				if (!connector->polled)
 					connector->polled = DRM_CONNECTOR_POLL_HPD;
@@ -3780,10 +4328,13 @@ void intel_irq_init(struct drm_device *dev)
 	INIT_WORK(&dev_priv->rps.work, gen6_pm_rps_work);
 	INIT_WORK(&dev_priv->l3_parity.error_work, ivybridge_parity_work);
 
+	/* Let's track the enabled rps events */
+	dev_priv->pm_rps_events = GEN6_PM_RPS_EVENTS;
+
 	setup_timer(&dev_priv->gpu_error.hangcheck_timer,
 		    i915_hangcheck_elapsed,
 		    (unsigned long) dev);
-	setup_timer(&dev_priv->hotplug_reenable_timer, i915_reenable_hotplug_timer_func,
+	setup_timer(&dev_priv->hotplug_reenable_timer, intel_hpd_irq_reenable,
 		    (unsigned long) dev_priv);
 
 	pm_qos_add_request(&dev_priv->pm_qos, PM_QOS_CPU_DMA_LATENCY, PM_QOS_DEFAULT_VALUE);
@@ -3804,7 +4355,15 @@ void intel_irq_init(struct drm_device *dev)
 		dev->driver->get_scanout_position = i915_get_crtc_scanoutpos;
 	}
 
-	if (IS_VALLEYVIEW(dev)) {
+	if (IS_CHERRYVIEW(dev)) {
+		dev->driver->irq_handler = cherryview_irq_handler;
+		dev->driver->irq_preinstall = cherryview_irq_preinstall;
+		dev->driver->irq_postinstall = cherryview_irq_postinstall;
+		dev->driver->irq_uninstall = cherryview_irq_uninstall;
+		dev->driver->enable_vblank = valleyview_enable_vblank;
+		dev->driver->disable_vblank = valleyview_disable_vblank;
+		dev_priv->display.hpd_irq_setup = i915_hpd_irq_setup;
+	} else if (IS_VALLEYVIEW(dev)) {
 		dev->driver->irq_handler = valleyview_irq_handler;
 		dev->driver->irq_preinstall = valleyview_irq_preinstall;
 		dev->driver->irq_postinstall = valleyview_irq_postinstall;
@@ -3814,7 +4373,7 @@ void intel_irq_init(struct drm_device *dev)
 		dev_priv->display.hpd_irq_setup = i915_hpd_irq_setup;
 	} else if (IS_GEN8(dev)) {
 		dev->driver->irq_handler = gen8_irq_handler;
-		dev->driver->irq_preinstall = gen8_irq_preinstall;
+		dev->driver->irq_preinstall = gen8_irq_reset;
 		dev->driver->irq_postinstall = gen8_irq_postinstall;
 		dev->driver->irq_uninstall = gen8_irq_uninstall;
 		dev->driver->enable_vblank = gen8_enable_vblank;
@@ -3822,7 +4381,7 @@ void intel_irq_init(struct drm_device *dev)
 		dev_priv->display.hpd_irq_setup = ibx_hpd_irq_setup;
 	} else if (HAS_PCH_SPLIT(dev)) {
 		dev->driver->irq_handler = ironlake_irq_handler;
-		dev->driver->irq_preinstall = ironlake_irq_preinstall;
+		dev->driver->irq_preinstall = ironlake_irq_reset;
 		dev->driver->irq_postinstall = ironlake_irq_postinstall;
 		dev->driver->irq_uninstall = ironlake_irq_uninstall;
 		dev->driver->enable_vblank = ironlake_enable_vblank;
@@ -3879,66 +4438,21 @@ void intel_hpd_init(struct drm_device *dev)
 	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 }
 
-/* Disable interrupts so we can allow Package C8+. */
-void hsw_pc8_disable_interrupts(struct drm_device *dev)
+/* Disable interrupts so we can allow runtime PM. */
+void intel_runtime_pm_disable_interrupts(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long irqflags;
 
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-
-	dev_priv->pc8.regsave.deimr = I915_READ(DEIMR);
-	dev_priv->pc8.regsave.sdeimr = I915_READ(SDEIMR);
-	dev_priv->pc8.regsave.gtimr = I915_READ(GTIMR);
-	dev_priv->pc8.regsave.gtier = I915_READ(GTIER);
-	dev_priv->pc8.regsave.gen6_pmimr = I915_READ(GEN6_PMIMR);
-
-	ironlake_disable_display_irq(dev_priv, ~DE_PCH_EVENT_IVB);
-	ibx_disable_display_interrupt(dev_priv, ~SDE_HOTPLUG_MASK_CPT);
-	ilk_disable_gt_irq(dev_priv, 0xffffffff);
-	snb_disable_pm_irq(dev_priv, 0xffffffff);
-
-	dev_priv->pc8.irqs_disabled = true;
-
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	dev->driver->irq_uninstall(dev);
+	dev_priv->pm.irqs_disabled = true;
 }
 
-/* Restore interrupts so we can recover from Package C8+. */
-void hsw_pc8_restore_interrupts(struct drm_device *dev)
+/* Restore interrupts so we can recover from runtime PM. */
+void intel_runtime_pm_restore_interrupts(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long irqflags;
-	uint32_t val, expected;
-
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-
-	val = I915_READ(DEIMR);
-	expected = ~DE_PCH_EVENT_IVB;
-	WARN(val != expected, "DEIMR is 0x%08x, not 0x%08x\n", val, expected);
-
-	val = I915_READ(SDEIMR) & ~SDE_HOTPLUG_MASK_CPT;
-	expected = ~SDE_HOTPLUG_MASK_CPT;
-	WARN(val != expected, "SDEIMR non-HPD bits are 0x%08x, not 0x%08x\n",
-	     val, expected);
 
-	val = I915_READ(GTIMR);
-	expected = 0xffffffff;
-	WARN(val != expected, "GTIMR is 0x%08x, not 0x%08x\n", val, expected);
-
-	val = I915_READ(GEN6_PMIMR);
-	expected = 0xffffffff;
-	WARN(val != expected, "GEN6_PMIMR is 0x%08x, not 0x%08x\n", val,
-	     expected);
-
-	dev_priv->pc8.irqs_disabled = false;
-
-	ironlake_enable_display_irq(dev_priv, ~dev_priv->pc8.regsave.deimr);
-	ibx_enable_display_interrupt(dev_priv,
-				     ~dev_priv->pc8.regsave.sdeimr &
-				     ~SDE_HOTPLUG_MASK_CPT);
-	ilk_enable_gt_irq(dev_priv, ~dev_priv->pc8.regsave.gtimr);
-	snb_enable_pm_irq(dev_priv, ~dev_priv->pc8.regsave.gen6_pmimr);
-	I915_WRITE(GTIER, dev_priv->pc8.regsave.gtier);
-
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	dev_priv->pm.irqs_disabled = false;
+	dev->driver->irq_preinstall(dev);
+	dev->driver->irq_postinstall(dev);
 }
diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c
new file mode 100644
index 00000000000..d05a2afa17d
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_params.c
@@ -0,0 +1,158 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include "i915_drv.h"
+
+struct i915_params i915 __read_mostly = {
+	.modeset = -1,
+	.panel_ignore_lid = 1,
+	.powersave = 1,
+	.semaphores = -1,
+	.lvds_downclock = 0,
+	.lvds_channel_mode = 0,
+	.panel_use_ssc = -1,
+	.vbt_sdvo_panel_type = -1,
+	.enable_rc6 = -1,
+	.enable_fbc = -1,
+	.enable_hangcheck = true,
+	.enable_ppgtt = -1,
+	.enable_psr = 0,
+	.preliminary_hw_support = IS_ENABLED(CONFIG_DRM_I915_PRELIMINARY_HW_SUPPORT),
+	.disable_power_well = 1,
+	.enable_ips = 1,
+	.fastboot = 0,
+	.prefault_disable = 0,
+	.reset = true,
+	.invert_brightness = 0,
+	.disable_display = 0,
+	.enable_cmd_parser = 1,
+	.disable_vtd_wa = 0,
+};
+
+module_param_named(modeset, i915.modeset, int, 0400);
+MODULE_PARM_DESC(modeset,
+	"Use kernel modesetting [KMS] (0=DRM_I915_KMS from .config, "
+	"1=on, -1=force vga console preference [default])");
+
+module_param_named(panel_ignore_lid, i915.panel_ignore_lid, int, 0600);
+MODULE_PARM_DESC(panel_ignore_lid,
+	"Override lid status (0=autodetect, 1=autodetect disabled [default], "
+	"-1=force lid closed, -2=force lid open)");
+
+module_param_named(powersave, i915.powersave, int, 0600);
+MODULE_PARM_DESC(powersave,
+	"Enable powersavings, fbc, downclocking, etc. (default: true)");
+
+module_param_named(semaphores, i915.semaphores, int, 0400);
+MODULE_PARM_DESC(semaphores,
+	"Use semaphores for inter-ring sync "
+	"(default: -1 (use per-chip defaults))");
+
+module_param_named(enable_rc6, i915.enable_rc6, int, 0400);
+MODULE_PARM_DESC(enable_rc6,
+	"Enable power-saving render C-state 6. "
+	"Different stages can be selected via bitmask values "
+	"(0 = disable; 1 = enable rc6; 2 = enable deep rc6; 4 = enable deepest rc6). "
+	"For example, 3 would enable rc6 and deep rc6, and 7 would enable everything. "
+	"default: -1 (use per-chip default)");
+
+module_param_named(enable_fbc, i915.enable_fbc, int, 0600);
+MODULE_PARM_DESC(enable_fbc,
+	"Enable frame buffer compression for power savings "
+	"(default: -1 (use per-chip default))");
+
+module_param_named(lvds_downclock, i915.lvds_downclock, int, 0400);
+MODULE_PARM_DESC(lvds_downclock,
+	"Use panel (LVDS/eDP) downclocking for power savings "
+	"(default: false)");
+
+module_param_named(lvds_channel_mode, i915.lvds_channel_mode, int, 0600);
+MODULE_PARM_DESC(lvds_channel_mode,
+	 "Specify LVDS channel mode "
+	 "(0=probe BIOS [default], 1=single-channel, 2=dual-channel)");
+
+module_param_named(lvds_use_ssc, i915.panel_use_ssc, int, 0600);
+MODULE_PARM_DESC(lvds_use_ssc,
+	"Use Spread Spectrum Clock with panels [LVDS/eDP] "
+	"(default: auto from VBT)");
+
+module_param_named(vbt_sdvo_panel_type, i915.vbt_sdvo_panel_type, int, 0600);
+MODULE_PARM_DESC(vbt_sdvo_panel_type,
+	"Override/Ignore selection of SDVO panel mode in the VBT "
+	"(-2=ignore, -1=auto [default], index in VBT BIOS table)");
+
+module_param_named(reset, i915.reset, bool, 0600);
+MODULE_PARM_DESC(reset, "Attempt GPU resets (default: true)");
+
+module_param_named(enable_hangcheck, i915.enable_hangcheck, bool, 0644);
+MODULE_PARM_DESC(enable_hangcheck,
+	"Periodically check GPU activity for detecting hangs. "
+	"WARNING: Disabling this can cause system wide hangs. "
+	"(default: true)");
+
+module_param_named(enable_ppgtt, i915.enable_ppgtt, int, 0400);
+MODULE_PARM_DESC(enable_ppgtt,
+	"Override PPGTT usage. "
+	"(-1=auto [default], 0=disabled, 1=aliasing, 2=full)");
+
+module_param_named(enable_psr, i915.enable_psr, int, 0600);
+MODULE_PARM_DESC(enable_psr, "Enable PSR (default: false)");
+
+module_param_named(preliminary_hw_support, i915.preliminary_hw_support, int, 0600);
+MODULE_PARM_DESC(preliminary_hw_support,
+	"Enable preliminary hardware support.");
+
+module_param_named(disable_power_well, i915.disable_power_well, int, 0600);
+MODULE_PARM_DESC(disable_power_well,
+	"Disable the power well when possible (default: true)");
+
+module_param_named(enable_ips, i915.enable_ips, int, 0600);
+MODULE_PARM_DESC(enable_ips, "Enable IPS (default: true)");
+
+module_param_named(fastboot, i915.fastboot, bool, 0600);
+MODULE_PARM_DESC(fastboot,
+	"Try to skip unnecessary mode sets at boot time (default: false)");
+
+module_param_named(prefault_disable, i915.prefault_disable, bool, 0600);
+MODULE_PARM_DESC(prefault_disable,
+	"Disable page prefaulting for pread/pwrite/reloc (default:false). "
+	"For developers only.");
+
+module_param_named(invert_brightness, i915.invert_brightness, int, 0600);
+MODULE_PARM_DESC(invert_brightness,
+	"Invert backlight brightness "
+	"(-1 force normal, 0 machine defaults, 1 force inversion), please "
+	"report PCI device ID, subsystem vendor and subsystem device ID "
+	"to dri-devel@lists.freedesktop.org, if your machine needs it. "
+	"It will then be included in an upcoming module version.");
+
+module_param_named(disable_display, i915.disable_display, bool, 0600);
+MODULE_PARM_DESC(disable_display, "Disable display (default: false)");
+
+module_param_named(disable_vtd_wa, i915.disable_vtd_wa, bool, 0600);
+MODULE_PARM_DESC(disable_vtd_wa, "Disable all VT-d workarounds (default: false)");
+
+module_param_named(enable_cmd_parser, i915.enable_cmd_parser, int, 0600);
+MODULE_PARM_DESC(enable_cmd_parser,
+		 "Enable command parsing (1=enabled [default], 0=disabled)");
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index ee2742122a0..a5bab61bfc0 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -26,10 +26,11 @@
 #define _I915_REG_H_
 
 #define _PIPE(pipe, a, b) ((a) + (pipe)*((b)-(a)))
-#define _PIPE_INC(pipe, base, inc) ((base) + (pipe)*(inc))
 #define _TRANSCODER(tran, a, b) ((a) + (tran)*((b)-(a)))
 
 #define _PORT(port, a, b) ((a) + (port)*((b)-(a)))
+#define _PIPE3(pipe, a, b, c) (pipe < 2 ? _PIPE(pipe, a, b) : c)
+#define _PORT3(port, a, b, c) (port < 2 ? _PORT(port, a, b) : c)
 
 #define _MASKED_BIT_ENABLE(a) (((a) << 16) | (a))
 #define _MASKED_BIT_DISABLE(a) ((a) << 16)
@@ -73,17 +74,24 @@
 #define   I915_GC_RENDER_CLOCK_166_MHZ	(0 << 0)
 #define   I915_GC_RENDER_CLOCK_200_MHZ	(1 << 0)
 #define   I915_GC_RENDER_CLOCK_333_MHZ	(4 << 0)
-#define LBB	0xf4
+#define PCI_LBPC 0xf4 /* legacy/combination backlight modes, also called LBB */
+
 
 /* Graphics reset regs */
 #define I965_GDRST 0xc0 /* PCI config register */
-#define ILK_GDSR 0x2ca4 /* MCHBAR offset */
 #define  GRDOM_FULL	(0<<2)
 #define  GRDOM_RENDER	(1<<2)
 #define  GRDOM_MEDIA	(3<<2)
 #define  GRDOM_MASK	(3<<2)
 #define  GRDOM_RESET_ENABLE (1<<0)
 
+#define ILK_GDSR 0x2ca4 /* MCHBAR offset */
+#define  ILK_GRDOM_FULL		(0<<1)
+#define  ILK_GRDOM_RENDER	(1<<1)
+#define  ILK_GRDOM_MEDIA	(3<<1)
+#define  ILK_GRDOM_MASK		(3<<1)
+#define  ILK_GRDOM_RESET_ENABLE (1<<0)
+
 #define GEN6_MBCUNIT_SNPCR	0x900c /* for LLC config */
 #define   GEN6_MBC_SNPCR_SHIFT	21
 #define   GEN6_MBC_SNPCR_MASK	(3<<21)
@@ -92,6 +100,9 @@
 #define   GEN6_MBC_SNPCR_LOW	(2<<21)
 #define   GEN6_MBC_SNPCR_MIN	(3<<21) /* only 1/16th of the cache is shared */
 
+#define VLV_G3DCTL		0x9024
+#define VLV_GSCKGCTL		0x9028
+
 #define GEN6_MBCTL		0x0907c
 #define   GEN6_MBCTL_ENABLE_BOOT_FETCH	(1 << 4)
 #define   GEN6_MBCTL_CTX_FETCH_NEEDED	(1 << 3)
@@ -175,9 +186,23 @@
 #define VGA_CR_DATA_CGA 0x3d5
 
 /*
+ * Instruction field definitions used by the command parser
+ */
+#define INSTR_CLIENT_SHIFT      29
+#define INSTR_CLIENT_MASK       0xE0000000
+#define   INSTR_MI_CLIENT       0x0
+#define   INSTR_BC_CLIENT       0x2
+#define   INSTR_RC_CLIENT       0x3
+#define INSTR_SUBCLIENT_SHIFT   27
+#define INSTR_SUBCLIENT_MASK    0x18000000
+#define   INSTR_MEDIA_SUBCLIENT 0x2
+
+/*
  * Memory interface instructions used by the kernel
  */
 #define MI_INSTR(opcode, flags) (((opcode) << 23) | (flags))
+/* Many MI commands use bit 22 of the header dword for GGTT vs PPGTT */
+#define  MI_GLOBAL_GTT    (1<<22)
 
 #define MI_NOOP			MI_INSTR(0, 0)
 #define MI_USER_INTERRUPT	MI_INSTR(0x02, 0)
@@ -193,10 +218,13 @@
 #define   MI_SCENE_COUNT	(1 << 3) /* just increment scene count */
 #define   MI_END_SCENE		(1 << 4) /* flush binner and incr scene count */
 #define   MI_INVALIDATE_ISP	(1 << 5) /* invalidate indirect state pointers */
+#define MI_REPORT_HEAD		MI_INSTR(0x07, 0)
+#define MI_ARB_ON_OFF		MI_INSTR(0x08, 0)
+#define   MI_ARB_ENABLE			(1<<0)
+#define   MI_ARB_DISABLE		(0<<0)
 #define MI_BATCH_BUFFER_END	MI_INSTR(0x0a, 0)
 #define MI_SUSPEND_FLUSH	MI_INSTR(0x0b, 0)
 #define   MI_SUSPEND_FLUSH_EN	(1<<0)
-#define MI_REPORT_HEAD		MI_INSTR(0x07, 0)
 #define MI_OVERLAY_FLIP		MI_INSTR(0x11, 0)
 #define   MI_OVERLAY_CONTINUE	(0x0<<21)
 #define   MI_OVERLAY_ON		(0x1<<21)
@@ -212,10 +240,25 @@
 #define   MI_DISPLAY_FLIP_IVB_SPRITE_B (3 << 19)
 #define   MI_DISPLAY_FLIP_IVB_PLANE_C  (4 << 19)
 #define   MI_DISPLAY_FLIP_IVB_SPRITE_C (5 << 19)
-#define MI_ARB_ON_OFF		MI_INSTR(0x08, 0)
-#define   MI_ARB_ENABLE			(1<<0)
-#define   MI_ARB_DISABLE		(0<<0)
-
+#define MI_SEMAPHORE_MBOX	MI_INSTR(0x16, 1) /* gen6+ */
+#define   MI_SEMAPHORE_GLOBAL_GTT    (1<<22)
+#define   MI_SEMAPHORE_UPDATE	    (1<<21)
+#define   MI_SEMAPHORE_COMPARE	    (1<<20)
+#define   MI_SEMAPHORE_REGISTER	    (1<<18)
+#define   MI_SEMAPHORE_SYNC_VR	    (0<<16) /* RCS  wait for VCS  (RVSYNC) */
+#define   MI_SEMAPHORE_SYNC_VER	    (1<<16) /* RCS  wait for VECS (RVESYNC) */
+#define   MI_SEMAPHORE_SYNC_BR	    (2<<16) /* RCS  wait for BCS  (RBSYNC) */
+#define   MI_SEMAPHORE_SYNC_BV	    (0<<16) /* VCS  wait for BCS  (VBSYNC) */
+#define   MI_SEMAPHORE_SYNC_VEV	    (1<<16) /* VCS  wait for VECS (VVESYNC) */
+#define   MI_SEMAPHORE_SYNC_RV	    (2<<16) /* VCS  wait for RCS  (VRSYNC) */
+#define   MI_SEMAPHORE_SYNC_RB	    (0<<16) /* BCS  wait for RCS  (BRSYNC) */
+#define   MI_SEMAPHORE_SYNC_VEB	    (1<<16) /* BCS  wait for VECS (BVESYNC) */
+#define   MI_SEMAPHORE_SYNC_VB	    (2<<16) /* BCS  wait for VCS  (BVSYNC) */
+#define   MI_SEMAPHORE_SYNC_BVE	    (0<<16) /* VECS wait for BCS  (VEBSYNC) */
+#define   MI_SEMAPHORE_SYNC_VVE	    (1<<16) /* VECS wait for VCS  (VEVSYNC) */
+#define   MI_SEMAPHORE_SYNC_RVE	    (2<<16) /* VECS wait for RCS  (VERSYNC) */
+#define   MI_SEMAPHORE_SYNC_INVALID (3<<16)
+#define   MI_SEMAPHORE_SYNC_MASK    (3<<16)
 #define MI_SET_CONTEXT		MI_INSTR(0x18, 0)
 #define   MI_MM_SPACE_GTT		(1<<8)
 #define   MI_MM_SPACE_PHYSICAL		(0<<8)
@@ -233,43 +276,29 @@
  * - One can actually load arbitrary many arbitrary registers: Simply issue x
  *   address/value pairs. Don't overdue it, though, x <= 2^4 must hold!
  */
-#define MI_LOAD_REGISTER_IMM(x)	MI_INSTR(0x22, 2*x-1)
-#define MI_STORE_REGISTER_MEM(x) MI_INSTR(0x24, 2*x-1)
-#define  MI_SRM_LRM_GLOBAL_GTT		(1<<22)
+#define MI_LOAD_REGISTER_IMM(x)	MI_INSTR(0x22, 2*(x)-1)
+#define MI_STORE_REGISTER_MEM(x) MI_INSTR(0x24, 2*(x)-1)
+#define MI_STORE_REGISTER_MEM_GEN8(x) MI_INSTR(0x24, 3*(x)-1)
+#define   MI_SRM_LRM_GLOBAL_GTT		(1<<22)
 #define MI_FLUSH_DW		MI_INSTR(0x26, 1) /* for GEN6 */
 #define   MI_FLUSH_DW_STORE_INDEX	(1<<21)
 #define   MI_INVALIDATE_TLB		(1<<18)
 #define   MI_FLUSH_DW_OP_STOREDW	(1<<14)
+#define   MI_FLUSH_DW_OP_MASK		(3<<14)
+#define   MI_FLUSH_DW_NOTIFY		(1<<8)
 #define   MI_INVALIDATE_BSD		(1<<7)
 #define   MI_FLUSH_DW_USE_GTT		(1<<2)
 #define   MI_FLUSH_DW_USE_PPGTT		(0<<2)
 #define MI_BATCH_BUFFER		MI_INSTR(0x30, 1)
 #define   MI_BATCH_NON_SECURE		(1)
 /* for snb/ivb/vlv this also means "batch in ppgtt" when ppgtt is enabled. */
-#define   MI_BATCH_NON_SECURE_I965 	(1<<8)
+#define   MI_BATCH_NON_SECURE_I965	(1<<8)
 #define   MI_BATCH_PPGTT_HSW		(1<<8)
-#define   MI_BATCH_NON_SECURE_HSW 	(1<<13)
+#define   MI_BATCH_NON_SECURE_HSW	(1<<13)
 #define MI_BATCH_BUFFER_START	MI_INSTR(0x31, 0)
 #define   MI_BATCH_GTT		    (2<<6) /* aliased with (1<<7) on gen4 */
 #define MI_BATCH_BUFFER_START_GEN8	MI_INSTR(0x31, 1)
-#define MI_SEMAPHORE_MBOX	MI_INSTR(0x16, 1) /* gen6+ */
-#define  MI_SEMAPHORE_GLOBAL_GTT    (1<<22)
-#define  MI_SEMAPHORE_UPDATE	    (1<<21)
-#define  MI_SEMAPHORE_COMPARE	    (1<<20)
-#define  MI_SEMAPHORE_REGISTER	    (1<<18)
-#define  MI_SEMAPHORE_SYNC_VR	    (0<<16) /* RCS  wait for VCS  (RVSYNC) */
-#define  MI_SEMAPHORE_SYNC_VER	    (1<<16) /* RCS  wait for VECS (RVESYNC) */
-#define  MI_SEMAPHORE_SYNC_BR	    (2<<16) /* RCS  wait for BCS  (RBSYNC) */
-#define  MI_SEMAPHORE_SYNC_BV	    (0<<16) /* VCS  wait for BCS  (VBSYNC) */
-#define  MI_SEMAPHORE_SYNC_VEV	    (1<<16) /* VCS  wait for VECS (VVESYNC) */
-#define  MI_SEMAPHORE_SYNC_RV	    (2<<16) /* VCS  wait for RCS  (VRSYNC) */
-#define  MI_SEMAPHORE_SYNC_RB	    (0<<16) /* BCS  wait for RCS  (BRSYNC) */
-#define  MI_SEMAPHORE_SYNC_VEB	    (1<<16) /* BCS  wait for VECS (BVESYNC) */
-#define  MI_SEMAPHORE_SYNC_VB	    (2<<16) /* BCS  wait for VCS  (BVSYNC) */
-#define  MI_SEMAPHORE_SYNC_BVE	    (0<<16) /* VECS wait for BCS  (VEBSYNC) */
-#define  MI_SEMAPHORE_SYNC_VVE	    (1<<16) /* VECS wait for VCS  (VEVSYNC) */
-#define  MI_SEMAPHORE_SYNC_RVE	    (2<<16) /* VECS wait for RCS  (VERSYNC) */
-#define  MI_SEMAPHORE_SYNC_INVALID  (3<<16)
+
 
 #define MI_PREDICATE_RESULT_2	(0x2214)
 #define  LOWER_SLICE_ENABLED	(1<<0)
@@ -318,9 +347,12 @@
 #define   DISPLAY_PLANE_B           (1<<20)
 #define GFX_OP_PIPE_CONTROL(len)	((0x3<<29)|(0x3<<27)|(0x2<<24)|(len-2))
 #define   PIPE_CONTROL_GLOBAL_GTT_IVB			(1<<24) /* gen7+ */
+#define   PIPE_CONTROL_MMIO_WRITE			(1<<23)
+#define   PIPE_CONTROL_STORE_DATA_INDEX			(1<<21)
 #define   PIPE_CONTROL_CS_STALL				(1<<20)
 #define   PIPE_CONTROL_TLB_INVALIDATE			(1<<18)
 #define   PIPE_CONTROL_QW_WRITE				(1<<14)
+#define   PIPE_CONTROL_POST_SYNC_OP_MASK                (3<<14)
 #define   PIPE_CONTROL_DEPTH_STALL			(1<<13)
 #define   PIPE_CONTROL_WRITE_FLUSH			(1<<12)
 #define   PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH	(1<<12) /* gen6+ */
@@ -335,6 +367,94 @@
 #define   PIPE_CONTROL_DEPTH_CACHE_FLUSH		(1<<0)
 #define   PIPE_CONTROL_GLOBAL_GTT (1<<2) /* in addr dword */
 
+/*
+ * Commands used only by the command parser
+ */
+#define MI_SET_PREDICATE        MI_INSTR(0x01, 0)
+#define MI_ARB_CHECK            MI_INSTR(0x05, 0)
+#define MI_RS_CONTROL           MI_INSTR(0x06, 0)
+#define MI_URB_ATOMIC_ALLOC     MI_INSTR(0x09, 0)
+#define MI_PREDICATE            MI_INSTR(0x0C, 0)
+#define MI_RS_CONTEXT           MI_INSTR(0x0F, 0)
+#define MI_TOPOLOGY_FILTER      MI_INSTR(0x0D, 0)
+#define MI_LOAD_SCAN_LINES_EXCL MI_INSTR(0x13, 0)
+#define MI_URB_CLEAR            MI_INSTR(0x19, 0)
+#define MI_UPDATE_GTT           MI_INSTR(0x23, 0)
+#define MI_CLFLUSH              MI_INSTR(0x27, 0)
+#define MI_REPORT_PERF_COUNT    MI_INSTR(0x28, 0)
+#define   MI_REPORT_PERF_COUNT_GGTT (1<<0)
+#define MI_LOAD_REGISTER_MEM    MI_INSTR(0x29, 0)
+#define MI_LOAD_REGISTER_REG    MI_INSTR(0x2A, 0)
+#define MI_RS_STORE_DATA_IMM    MI_INSTR(0x2B, 0)
+#define MI_LOAD_URB_MEM         MI_INSTR(0x2C, 0)
+#define MI_STORE_URB_MEM        MI_INSTR(0x2D, 0)
+#define MI_CONDITIONAL_BATCH_BUFFER_END MI_INSTR(0x36, 0)
+
+#define PIPELINE_SELECT                ((0x3<<29)|(0x1<<27)|(0x1<<24)|(0x4<<16))
+#define GFX_OP_3DSTATE_VF_STATISTICS   ((0x3<<29)|(0x1<<27)|(0x0<<24)|(0xB<<16))
+#define MEDIA_VFE_STATE                ((0x3<<29)|(0x2<<27)|(0x0<<24)|(0x0<<16))
+#define  MEDIA_VFE_STATE_MMIO_ACCESS_MASK (0x18)
+#define GPGPU_OBJECT                   ((0x3<<29)|(0x2<<27)|(0x1<<24)|(0x4<<16))
+#define GPGPU_WALKER                   ((0x3<<29)|(0x2<<27)|(0x1<<24)|(0x5<<16))
+#define GFX_OP_3DSTATE_DX9_CONSTANTF_VS \
+	((0x3<<29)|(0x3<<27)|(0x0<<24)|(0x39<<16))
+#define GFX_OP_3DSTATE_DX9_CONSTANTF_PS \
+	((0x3<<29)|(0x3<<27)|(0x0<<24)|(0x3A<<16))
+#define GFX_OP_3DSTATE_SO_DECL_LIST \
+	((0x3<<29)|(0x3<<27)|(0x1<<24)|(0x17<<16))
+
+#define GFX_OP_3DSTATE_BINDING_TABLE_EDIT_VS \
+	((0x3<<29)|(0x3<<27)|(0x0<<24)|(0x43<<16))
+#define GFX_OP_3DSTATE_BINDING_TABLE_EDIT_GS \
+	((0x3<<29)|(0x3<<27)|(0x0<<24)|(0x44<<16))
+#define GFX_OP_3DSTATE_BINDING_TABLE_EDIT_HS \
+	((0x3<<29)|(0x3<<27)|(0x0<<24)|(0x45<<16))
+#define GFX_OP_3DSTATE_BINDING_TABLE_EDIT_DS \
+	((0x3<<29)|(0x3<<27)|(0x0<<24)|(0x46<<16))
+#define GFX_OP_3DSTATE_BINDING_TABLE_EDIT_PS \
+	((0x3<<29)|(0x3<<27)|(0x0<<24)|(0x47<<16))
+
+#define MFX_WAIT  ((0x3<<29)|(0x1<<27)|(0x0<<16))
+
+#define COLOR_BLT     ((0x2<<29)|(0x40<<22))
+#define SRC_COPY_BLT  ((0x2<<29)|(0x43<<22))
+
+/*
+ * Registers used only by the command parser
+ */
+#define BCS_SWCTRL 0x22200
+
+#define HS_INVOCATION_COUNT 0x2300
+#define DS_INVOCATION_COUNT 0x2308
+#define IA_VERTICES_COUNT   0x2310
+#define IA_PRIMITIVES_COUNT 0x2318
+#define VS_INVOCATION_COUNT 0x2320
+#define GS_INVOCATION_COUNT 0x2328
+#define GS_PRIMITIVES_COUNT 0x2330
+#define CL_INVOCATION_COUNT 0x2338
+#define CL_PRIMITIVES_COUNT 0x2340
+#define PS_INVOCATION_COUNT 0x2348
+#define PS_DEPTH_COUNT      0x2350
+
+/* There are the 4 64-bit counter registers, one for each stream output */
+#define GEN7_SO_NUM_PRIMS_WRITTEN(n) (0x5200 + (n) * 8)
+
+#define GEN7_SO_PRIM_STORAGE_NEEDED(n)  (0x5240 + (n) * 8)
+
+#define GEN7_3DPRIM_END_OFFSET          0x2420
+#define GEN7_3DPRIM_START_VERTEX        0x2430
+#define GEN7_3DPRIM_VERTEX_COUNT        0x2434
+#define GEN7_3DPRIM_INSTANCE_COUNT      0x2438
+#define GEN7_3DPRIM_START_INSTANCE      0x243C
+#define GEN7_3DPRIM_BASE_VERTEX         0x2440
+
+#define OACONTROL 0x2360
+
+#define _GEN7_PIPEA_DE_LOAD_SL	0x70068
+#define _GEN7_PIPEB_DE_LOAD_SL	0x71068
+#define GEN7_PIPE_DE_LOAD_SL(pipe) _PIPE(pipe, \
+					 _GEN7_PIPEA_DE_LOAD_SL, \
+					 _GEN7_PIPEB_DE_LOAD_SL)
 
 /*
  * Reset registers
@@ -354,27 +474,51 @@
 #define   IOSF_BYTE_ENABLES_SHIFT		4
 #define   IOSF_BAR_SHIFT			1
 #define   IOSF_SB_BUSY				(1<<0)
+#define   IOSF_PORT_BUNIT			0x3
 #define   IOSF_PORT_PUNIT			0x4
 #define   IOSF_PORT_NC				0x11
 #define   IOSF_PORT_DPIO			0x12
+#define   IOSF_PORT_DPIO_2			0x1a
 #define   IOSF_PORT_GPIO_NC			0x13
 #define   IOSF_PORT_CCK				0x14
 #define   IOSF_PORT_CCU				0xA9
 #define   IOSF_PORT_GPS_CORE			0x48
+#define   IOSF_PORT_FLISDSI			0x1B
 #define VLV_IOSF_DATA				(VLV_DISPLAY_BASE + 0x2104)
 #define VLV_IOSF_ADDR				(VLV_DISPLAY_BASE + 0x2108)
 
-#define PUNIT_OPCODE_REG_READ			6
-#define PUNIT_OPCODE_REG_WRITE			7
+/* See configdb bunit SB addr map */
+#define BUNIT_REG_BISOC				0x11
+
+#define PUNIT_REG_DSPFREQ			0x36
+#define   DSPFREQSTAT_SHIFT			30
+#define   DSPFREQSTAT_MASK			(0x3 << DSPFREQSTAT_SHIFT)
+#define   DSPFREQGUAR_SHIFT			14
+#define   DSPFREQGUAR_MASK			(0x3 << DSPFREQGUAR_SHIFT)
+
+/* See the PUNIT HAS v0.8 for the below bits */
+enum punit_power_well {
+	PUNIT_POWER_WELL_RENDER			= 0,
+	PUNIT_POWER_WELL_MEDIA			= 1,
+	PUNIT_POWER_WELL_DISP2D			= 3,
+	PUNIT_POWER_WELL_DPIO_CMN_BC		= 5,
+	PUNIT_POWER_WELL_DPIO_TX_B_LANES_01	= 6,
+	PUNIT_POWER_WELL_DPIO_TX_B_LANES_23	= 7,
+	PUNIT_POWER_WELL_DPIO_TX_C_LANES_01	= 8,
+	PUNIT_POWER_WELL_DPIO_TX_C_LANES_23	= 9,
+	PUNIT_POWER_WELL_DPIO_RX0		= 10,
+	PUNIT_POWER_WELL_DPIO_RX1		= 11,
+
+	PUNIT_POWER_WELL_NUM,
+};
 
 #define PUNIT_REG_PWRGT_CTRL			0x60
 #define PUNIT_REG_PWRGT_STATUS			0x61
-#define	  PUNIT_CLK_GATE			1
-#define	  PUNIT_PWR_RESET			2
-#define	  PUNIT_PWR_GATE			3
-#define	  RENDER_PWRGT				(PUNIT_PWR_GATE << 0)
-#define	  MEDIA_PWRGT				(PUNIT_PWR_GATE << 2)
-#define	  DISP2D_PWRGT				(PUNIT_PWR_GATE << 6)
+#define   PUNIT_PWRGT_MASK(power_well)		(3 << ((power_well) * 2))
+#define   PUNIT_PWRGT_PWR_ON(power_well)	(0 << ((power_well) * 2))
+#define   PUNIT_PWRGT_CLK_GATE(power_well)	(1 << ((power_well) * 2))
+#define   PUNIT_PWRGT_RESET(power_well)		(2 << ((power_well) * 2))
+#define   PUNIT_PWRGT_PWR_GATE(power_well)	(3 << ((power_well) * 2))
 
 #define PUNIT_REG_GPU_LFM			0xd3
 #define PUNIT_REG_GPU_FREQ_REQ			0xd4
@@ -429,17 +573,93 @@
 #define  DSI_PLL_N1_DIV_MASK			(3 << 16)
 #define  DSI_PLL_M1_DIV_SHIFT			0
 #define  DSI_PLL_M1_DIV_MASK			(0x1ff << 0)
+#define CCK_DISPLAY_CLOCK_CONTROL		0x6b
 
-/*
- * DPIO - a special bus for various display related registers to hide behind
+/**
+ * DOC: DPIO
+ *
+ * VLV and CHV have slightly peculiar display PHYs for driving DP/HDMI
+ * ports. DPIO is the name given to such a display PHY. These PHYs
+ * don't follow the standard programming model using direct MMIO
+ * registers, and instead their registers must be accessed trough IOSF
+ * sideband. VLV has one such PHY for driving ports B and C, and CHV
+ * adds another PHY for driving port D. Each PHY responds to specific
+ * IOSF-SB port.
+ *
+ * Each display PHY is made up of one or two channels. Each channel
+ * houses a common lane part which contains the PLL and other common
+ * logic. CH0 common lane also contains the IOSF-SB logic for the
+ * Common Register Interface (CRI) ie. the DPIO registers. CRI clock
+ * must be running when any DPIO registers are accessed.
+ *
+ * In addition to having their own registers, the PHYs are also
+ * controlled through some dedicated signals from the display
+ * controller. These include PLL reference clock enable, PLL enable,
+ * and CRI clock selection, for example.
+ *
+ * Eeach channel also has two splines (also called data lanes), and
+ * each spline is made up of one Physical Access Coding Sub-Layer
+ * (PCS) block and two TX lanes. So each channel has two PCS blocks
+ * and four TX lanes. The TX lanes are used as DP lanes or TMDS
+ * data/clock pairs depending on the output type.
+ *
+ * Additionally the PHY also contains an AUX lane with AUX blocks
+ * for each channel. This is used for DP AUX communication, but
+ * this fact isn't really relevant for the driver since AUX is
+ * controlled from the display controller side. No DPIO registers
+ * need to be accessed during AUX communication,
+ *
+ * Generally the common lane corresponds to the pipe and
+ * the spline (PCS/TX) correponds to the port.
+ *
+ * For dual channel PHY (VLV/CHV):
+ *
+ *  pipe A == CMN/PLL/REF CH0
+ *
+ *  pipe B == CMN/PLL/REF CH1
+ *
+ *  port B == PCS/TX CH0
+ *
+ *  port C == PCS/TX CH1
+ *
+ * This is especially important when we cross the streams
+ * ie. drive port B with pipe B, or port C with pipe A.
  *
- * DPIO is VLV only.
+ * For single channel PHY (CHV):
  *
- * Note: digital port B is DDI0, digital pot C is DDI1
+ *  pipe C == CMN/PLL/REF CH0
+ *
+ *  port D == PCS/TX CH0
+ *
+ * Note: digital port B is DDI0, digital port C is DDI1,
+ * digital port D is DDI2
+ */
+/*
+ * Dual channel PHY (VLV/CHV)
+ * ---------------------------------
+ * |      CH0      |      CH1      |
+ * |  CMN/PLL/REF  |  CMN/PLL/REF  |
+ * |---------------|---------------| Display PHY
+ * | PCS01 | PCS23 | PCS01 | PCS23 |
+ * |-------|-------|-------|-------|
+ * |TX0|TX1|TX2|TX3|TX0|TX1|TX2|TX3|
+ * ---------------------------------
+ * |     DDI0      |     DDI1      | DP/HDMI ports
+ * ---------------------------------
+ *
+ * Single channel PHY (CHV)
+ * -----------------
+ * |      CH0      |
+ * |  CMN/PLL/REF  |
+ * |---------------| Display PHY
+ * | PCS01 | PCS23 |
+ * |-------|-------|
+ * |TX0|TX1|TX2|TX3|
+ * -----------------
+ * |     DDI2      | DP/HDMI port
+ * -----------------
  */
 #define DPIO_DEVFN			0
-#define DPIO_OPCODE_REG_WRITE		1
-#define DPIO_OPCODE_REG_READ		0
 
 #define DPIO_CTL			(VLV_DISPLAY_BASE + 0x2110)
 #define  DPIO_MODSEL1			(1<<3) /* if ref clk b == 27 */
@@ -447,15 +667,13 @@
 #define  DPIO_SFR_BYPASS		(1<<1)
 #define  DPIO_CMNRST			(1<<0)
 
-#define _DPIO_TX3_SWING_CTL4_A		0x690
-#define _DPIO_TX3_SWING_CTL4_B		0x2a90
-#define DPIO_TX3_SWING_CTL4(pipe) _PIPE(pipe, _DPIO_TX3_SWING_CTL4_A, \
-					_DPIO_TX3_SWING_CTL4_B)
+#define DPIO_PHY(pipe)			((pipe) >> 1)
+#define DPIO_PHY_IOSF_PORT(phy)		(dev_priv->dpio_phy_iosf_port[phy])
 
 /*
  * Per pipe/PLL DPIO regs
  */
-#define _DPIO_DIV_A			0x800c
+#define _VLV_PLL_DW3_CH0		0x800c
 #define   DPIO_POST_DIV_SHIFT		(28) /* 3 bits */
 #define   DPIO_POST_DIV_DAC		0
 #define   DPIO_POST_DIV_HDMIDP		1 /* DAC 225-400M rate */
@@ -468,10 +686,10 @@
 #define   DPIO_ENABLE_CALIBRATION	(1<<11)
 #define   DPIO_M1DIV_SHIFT		(8) /* 3 bits */
 #define   DPIO_M2DIV_MASK		0xff
-#define _DPIO_DIV_B			0x802c
-#define DPIO_DIV(pipe) _PIPE(pipe, _DPIO_DIV_A, _DPIO_DIV_B)
+#define _VLV_PLL_DW3_CH1		0x802c
+#define VLV_PLL_DW3(ch) _PIPE(ch, _VLV_PLL_DW3_CH0, _VLV_PLL_DW3_CH1)
 
-#define _DPIO_REFSFR_A			0x8014
+#define _VLV_PLL_DW5_CH0		0x8014
 #define   DPIO_REFSEL_OVERRIDE		27
 #define   DPIO_PLL_MODESEL_SHIFT	24 /* 3 bits */
 #define   DPIO_BIAS_CURRENT_CTL_SHIFT	21 /* 3 bits, always 0x7 */
@@ -479,119 +697,215 @@
 #define   DPIO_PLL_REFCLK_SEL_MASK	3
 #define   DPIO_DRIVER_CTL_SHIFT		12 /* always set to 0x8 */
 #define   DPIO_CLK_BIAS_CTL_SHIFT	8 /* always set to 0x5 */
-#define _DPIO_REFSFR_B			0x8034
-#define DPIO_REFSFR(pipe) _PIPE(pipe, _DPIO_REFSFR_A, _DPIO_REFSFR_B)
+#define _VLV_PLL_DW5_CH1		0x8034
+#define VLV_PLL_DW5(ch) _PIPE(ch, _VLV_PLL_DW5_CH0, _VLV_PLL_DW5_CH1)
 
-#define _DPIO_CORE_CLK_A		0x801c
-#define _DPIO_CORE_CLK_B		0x803c
-#define DPIO_CORE_CLK(pipe) _PIPE(pipe, _DPIO_CORE_CLK_A, _DPIO_CORE_CLK_B)
+#define _VLV_PLL_DW7_CH0		0x801c
+#define _VLV_PLL_DW7_CH1		0x803c
+#define VLV_PLL_DW7(ch) _PIPE(ch, _VLV_PLL_DW7_CH0, _VLV_PLL_DW7_CH1)
 
-#define _DPIO_IREF_CTL_A		0x8040
-#define _DPIO_IREF_CTL_B		0x8060
-#define DPIO_IREF_CTL(pipe) _PIPE(pipe, _DPIO_IREF_CTL_A, _DPIO_IREF_CTL_B)
+#define _VLV_PLL_DW8_CH0		0x8040
+#define _VLV_PLL_DW8_CH1		0x8060
+#define VLV_PLL_DW8(ch) _PIPE(ch, _VLV_PLL_DW8_CH0, _VLV_PLL_DW8_CH1)
 
-#define DPIO_IREF_BCAST			0xc044
-#define _DPIO_IREF_A			0x8044
-#define _DPIO_IREF_B			0x8064
-#define DPIO_IREF(pipe) _PIPE(pipe, _DPIO_IREF_A, _DPIO_IREF_B)
+#define VLV_PLL_DW9_BCAST		0xc044
+#define _VLV_PLL_DW9_CH0		0x8044
+#define _VLV_PLL_DW9_CH1		0x8064
+#define VLV_PLL_DW9(ch) _PIPE(ch, _VLV_PLL_DW9_CH0, _VLV_PLL_DW9_CH1)
 
-#define _DPIO_PLL_CML_A			0x804c
-#define _DPIO_PLL_CML_B			0x806c
-#define DPIO_PLL_CML(pipe) _PIPE(pipe, _DPIO_PLL_CML_A, _DPIO_PLL_CML_B)
+#define _VLV_PLL_DW10_CH0		0x8048
+#define _VLV_PLL_DW10_CH1		0x8068
+#define VLV_PLL_DW10(ch) _PIPE(ch, _VLV_PLL_DW10_CH0, _VLV_PLL_DW10_CH1)
 
-#define _DPIO_LPF_COEFF_A		0x8048
-#define _DPIO_LPF_COEFF_B		0x8068
-#define DPIO_LPF_COEFF(pipe) _PIPE(pipe, _DPIO_LPF_COEFF_A, _DPIO_LPF_COEFF_B)
+#define _VLV_PLL_DW11_CH0		0x804c
+#define _VLV_PLL_DW11_CH1		0x806c
+#define VLV_PLL_DW11(ch) _PIPE(ch, _VLV_PLL_DW11_CH0, _VLV_PLL_DW11_CH1)
 
-#define DPIO_CALIBRATION		0x80ac
+/* Spec for ref block start counts at DW10 */
+#define VLV_REF_DW13			0x80ac
 
-#define DPIO_FASTCLK_DISABLE		0x8100
+#define VLV_CMN_DW0			0x8100
 
 /*
  * Per DDI channel DPIO regs
  */
 
-#define _DPIO_PCS_TX_0			0x8200
-#define _DPIO_PCS_TX_1			0x8400
+#define _VLV_PCS_DW0_CH0		0x8200
+#define _VLV_PCS_DW0_CH1		0x8400
 #define   DPIO_PCS_TX_LANE2_RESET	(1<<16)
 #define   DPIO_PCS_TX_LANE1_RESET	(1<<7)
-#define DPIO_PCS_TX(port) _PORT(port, _DPIO_PCS_TX_0, _DPIO_PCS_TX_1)
-
-#define _DPIO_PCS_CLK_0			0x8204
-#define _DPIO_PCS_CLK_1			0x8404
+#define VLV_PCS_DW0(ch) _PORT(ch, _VLV_PCS_DW0_CH0, _VLV_PCS_DW0_CH1)
+
+#define _VLV_PCS01_DW0_CH0		0x200
+#define _VLV_PCS23_DW0_CH0		0x400
+#define _VLV_PCS01_DW0_CH1		0x2600
+#define _VLV_PCS23_DW0_CH1		0x2800
+#define VLV_PCS01_DW0(ch) _PORT(ch, _VLV_PCS01_DW0_CH0, _VLV_PCS01_DW0_CH1)
+#define VLV_PCS23_DW0(ch) _PORT(ch, _VLV_PCS23_DW0_CH0, _VLV_PCS23_DW0_CH1)
+
+#define _VLV_PCS_DW1_CH0		0x8204
+#define _VLV_PCS_DW1_CH1		0x8404
+#define   CHV_PCS_REQ_SOFTRESET_EN	(1<<23)
 #define   DPIO_PCS_CLK_CRI_RXEB_EIOS_EN	(1<<22)
 #define   DPIO_PCS_CLK_CRI_RXDIGFILTSG_EN (1<<21)
 #define   DPIO_PCS_CLK_DATAWIDTH_SHIFT	(6)
 #define   DPIO_PCS_CLK_SOFT_RESET	(1<<5)
-#define DPIO_PCS_CLK(port) _PORT(port, _DPIO_PCS_CLK_0, _DPIO_PCS_CLK_1)
-
-#define _DPIO_PCS_CTL_OVR1_A		0x8224
-#define _DPIO_PCS_CTL_OVR1_B		0x8424
-#define DPIO_PCS_CTL_OVER1(port) _PORT(port, _DPIO_PCS_CTL_OVR1_A, \
-				       _DPIO_PCS_CTL_OVR1_B)
-
-#define _DPIO_PCS_STAGGER0_A		0x822c
-#define _DPIO_PCS_STAGGER0_B		0x842c
-#define DPIO_PCS_STAGGER0(port) _PORT(port, _DPIO_PCS_STAGGER0_A, \
-				      _DPIO_PCS_STAGGER0_B)
-
-#define _DPIO_PCS_STAGGER1_A		0x8230
-#define _DPIO_PCS_STAGGER1_B		0x8430
-#define DPIO_PCS_STAGGER1(port) _PORT(port, _DPIO_PCS_STAGGER1_A, \
-				      _DPIO_PCS_STAGGER1_B)
-
-#define _DPIO_PCS_CLOCKBUF0_A		0x8238
-#define _DPIO_PCS_CLOCKBUF0_B		0x8438
-#define DPIO_PCS_CLOCKBUF0(port) _PORT(port, _DPIO_PCS_CLOCKBUF0_A, \
-				       _DPIO_PCS_CLOCKBUF0_B)
-
-#define _DPIO_PCS_CLOCKBUF8_A		0x825c
-#define _DPIO_PCS_CLOCKBUF8_B		0x845c
-#define DPIO_PCS_CLOCKBUF8(port) _PORT(port, _DPIO_PCS_CLOCKBUF8_A, \
-				       _DPIO_PCS_CLOCKBUF8_B)
-
-#define _DPIO_TX_SWING_CTL2_A		0x8288
-#define _DPIO_TX_SWING_CTL2_B		0x8488
-#define DPIO_TX_SWING_CTL2(port) _PORT(port, _DPIO_TX_SWING_CTL2_A, \
-				       _DPIO_TX_SWING_CTL2_B)
-
-#define _DPIO_TX_SWING_CTL3_A		0x828c
-#define _DPIO_TX_SWING_CTL3_B		0x848c
-#define DPIO_TX_SWING_CTL3(port) _PORT(port, _DPIO_TX_SWING_CTL3_A, \
-				       _DPIO_TX_SWING_CTL3_B)
-
-#define _DPIO_TX_SWING_CTL4_A		0x8290
-#define _DPIO_TX_SWING_CTL4_B		0x8490
-#define DPIO_TX_SWING_CTL4(port) _PORT(port, _DPIO_TX_SWING_CTL4_A, \
-				       _DPIO_TX_SWING_CTL4_B)
-
-#define _DPIO_TX_OCALINIT_0		0x8294
-#define _DPIO_TX_OCALINIT_1		0x8494
+#define VLV_PCS_DW1(ch) _PORT(ch, _VLV_PCS_DW1_CH0, _VLV_PCS_DW1_CH1)
+
+#define _VLV_PCS01_DW1_CH0		0x204
+#define _VLV_PCS23_DW1_CH0		0x404
+#define _VLV_PCS01_DW1_CH1		0x2604
+#define _VLV_PCS23_DW1_CH1		0x2804
+#define VLV_PCS01_DW1(ch) _PORT(ch, _VLV_PCS01_DW1_CH0, _VLV_PCS01_DW1_CH1)
+#define VLV_PCS23_DW1(ch) _PORT(ch, _VLV_PCS23_DW1_CH0, _VLV_PCS23_DW1_CH1)
+
+#define _VLV_PCS_DW8_CH0		0x8220
+#define _VLV_PCS_DW8_CH1		0x8420
+#define VLV_PCS_DW8(ch) _PORT(ch, _VLV_PCS_DW8_CH0, _VLV_PCS_DW8_CH1)
+
+#define _VLV_PCS01_DW8_CH0		0x0220
+#define _VLV_PCS23_DW8_CH0		0x0420
+#define _VLV_PCS01_DW8_CH1		0x2620
+#define _VLV_PCS23_DW8_CH1		0x2820
+#define VLV_PCS01_DW8(port) _PORT(port, _VLV_PCS01_DW8_CH0, _VLV_PCS01_DW8_CH1)
+#define VLV_PCS23_DW8(port) _PORT(port, _VLV_PCS23_DW8_CH0, _VLV_PCS23_DW8_CH1)
+
+#define _VLV_PCS_DW9_CH0		0x8224
+#define _VLV_PCS_DW9_CH1		0x8424
+#define	VLV_PCS_DW9(ch) _PORT(ch, _VLV_PCS_DW9_CH0, _VLV_PCS_DW9_CH1)
+
+#define _CHV_PCS_DW10_CH0		0x8228
+#define _CHV_PCS_DW10_CH1		0x8428
+#define   DPIO_PCS_SWING_CALC_TX0_TX2	(1<<30)
+#define   DPIO_PCS_SWING_CALC_TX1_TX3	(1<<31)
+#define CHV_PCS_DW10(ch) _PORT(ch, _CHV_PCS_DW10_CH0, _CHV_PCS_DW10_CH1)
+
+#define _VLV_PCS01_DW10_CH0		0x0228
+#define _VLV_PCS23_DW10_CH0		0x0428
+#define _VLV_PCS01_DW10_CH1		0x2628
+#define _VLV_PCS23_DW10_CH1		0x2828
+#define VLV_PCS01_DW10(port) _PORT(port, _VLV_PCS01_DW10_CH0, _VLV_PCS01_DW10_CH1)
+#define VLV_PCS23_DW10(port) _PORT(port, _VLV_PCS23_DW10_CH0, _VLV_PCS23_DW10_CH1)
+
+#define _VLV_PCS_DW11_CH0		0x822c
+#define _VLV_PCS_DW11_CH1		0x842c
+#define VLV_PCS_DW11(ch) _PORT(ch, _VLV_PCS_DW11_CH0, _VLV_PCS_DW11_CH1)
+
+#define _VLV_PCS_DW12_CH0		0x8230
+#define _VLV_PCS_DW12_CH1		0x8430
+#define VLV_PCS_DW12(ch) _PORT(ch, _VLV_PCS_DW12_CH0, _VLV_PCS_DW12_CH1)
+
+#define _VLV_PCS_DW14_CH0		0x8238
+#define _VLV_PCS_DW14_CH1		0x8438
+#define	VLV_PCS_DW14(ch) _PORT(ch, _VLV_PCS_DW14_CH0, _VLV_PCS_DW14_CH1)
+
+#define _VLV_PCS_DW23_CH0		0x825c
+#define _VLV_PCS_DW23_CH1		0x845c
+#define VLV_PCS_DW23(ch) _PORT(ch, _VLV_PCS_DW23_CH0, _VLV_PCS_DW23_CH1)
+
+#define _VLV_TX_DW2_CH0			0x8288
+#define _VLV_TX_DW2_CH1			0x8488
+#define   DPIO_SWING_MARGIN_SHIFT	16
+#define   DPIO_SWING_MARGIN_MASK	(0xff << DPIO_SWING_MARGIN_SHIFT)
+#define   DPIO_UNIQ_TRANS_SCALE_SHIFT	8
+#define VLV_TX_DW2(ch) _PORT(ch, _VLV_TX_DW2_CH0, _VLV_TX_DW2_CH1)
+
+#define _VLV_TX_DW3_CH0			0x828c
+#define _VLV_TX_DW3_CH1			0x848c
+/* The following bit for CHV phy */
+#define   DPIO_TX_UNIQ_TRANS_SCALE_EN	(1<<27)
+#define VLV_TX_DW3(ch) _PORT(ch, _VLV_TX_DW3_CH0, _VLV_TX_DW3_CH1)
+
+#define _VLV_TX_DW4_CH0			0x8290
+#define _VLV_TX_DW4_CH1			0x8490
+#define   DPIO_SWING_DEEMPH9P5_SHIFT	24
+#define   DPIO_SWING_DEEMPH9P5_MASK	(0xff << DPIO_SWING_DEEMPH9P5_SHIFT)
+#define VLV_TX_DW4(ch) _PORT(ch, _VLV_TX_DW4_CH0, _VLV_TX_DW4_CH1)
+
+#define _VLV_TX3_DW4_CH0		0x690
+#define _VLV_TX3_DW4_CH1		0x2a90
+#define VLV_TX3_DW4(ch) _PORT(ch, _VLV_TX3_DW4_CH0, _VLV_TX3_DW4_CH1)
+
+#define _VLV_TX_DW5_CH0			0x8294
+#define _VLV_TX_DW5_CH1			0x8494
 #define   DPIO_TX_OCALINIT_EN		(1<<31)
-#define DPIO_TX_OCALINIT(port) _PORT(port, _DPIO_TX_OCALINIT_0, \
-				     _DPIO_TX_OCALINIT_1)
-
-#define _DPIO_TX_CTL_0			0x82ac
-#define _DPIO_TX_CTL_1			0x84ac
-#define DPIO_TX_CTL(port) _PORT(port, _DPIO_TX_CTL_0, _DPIO_TX_CTL_1)
-
-#define _DPIO_TX_LANE_0			0x82b8
-#define _DPIO_TX_LANE_1			0x84b8
-#define DPIO_TX_LANE(port) _PORT(port, _DPIO_TX_LANE_0, _DPIO_TX_LANE_1)
-
-#define _DPIO_DATA_CHANNEL1		0x8220
-#define _DPIO_DATA_CHANNEL2		0x8420
-#define DPIO_DATA_CHANNEL(port) _PORT(port, _DPIO_DATA_CHANNEL1, _DPIO_DATA_CHANNEL2)
-
-#define _DPIO_PORT0_PCS0		0x0220
-#define _DPIO_PORT0_PCS1		0x0420
-#define _DPIO_PORT1_PCS2		0x2620
-#define _DPIO_PORT1_PCS3		0x2820
-#define DPIO_DATA_LANE_A(port) _PORT(port, _DPIO_PORT0_PCS0, _DPIO_PORT1_PCS2)
-#define DPIO_DATA_LANE_B(port) _PORT(port, _DPIO_PORT0_PCS1, _DPIO_PORT1_PCS3)
-#define DPIO_DATA_CHANNEL1              0x8220
-#define DPIO_DATA_CHANNEL2              0x8420
-
+#define VLV_TX_DW5(ch) _PORT(ch, _VLV_TX_DW5_CH0, _VLV_TX_DW5_CH1)
+
+#define _VLV_TX_DW11_CH0		0x82ac
+#define _VLV_TX_DW11_CH1		0x84ac
+#define VLV_TX_DW11(ch) _PORT(ch, _VLV_TX_DW11_CH0, _VLV_TX_DW11_CH1)
+
+#define _VLV_TX_DW14_CH0		0x82b8
+#define _VLV_TX_DW14_CH1		0x84b8
+#define VLV_TX_DW14(ch) _PORT(ch, _VLV_TX_DW14_CH0, _VLV_TX_DW14_CH1)
+
+/* CHV dpPhy registers */
+#define _CHV_PLL_DW0_CH0		0x8000
+#define _CHV_PLL_DW0_CH1		0x8180
+#define CHV_PLL_DW0(ch) _PIPE(ch, _CHV_PLL_DW0_CH0, _CHV_PLL_DW0_CH1)
+
+#define _CHV_PLL_DW1_CH0		0x8004
+#define _CHV_PLL_DW1_CH1		0x8184
+#define   DPIO_CHV_N_DIV_SHIFT		8
+#define   DPIO_CHV_M1_DIV_BY_2		(0 << 0)
+#define CHV_PLL_DW1(ch) _PIPE(ch, _CHV_PLL_DW1_CH0, _CHV_PLL_DW1_CH1)
+
+#define _CHV_PLL_DW2_CH0		0x8008
+#define _CHV_PLL_DW2_CH1		0x8188
+#define CHV_PLL_DW2(ch) _PIPE(ch, _CHV_PLL_DW2_CH0, _CHV_PLL_DW2_CH1)
+
+#define _CHV_PLL_DW3_CH0		0x800c
+#define _CHV_PLL_DW3_CH1		0x818c
+#define  DPIO_CHV_FRAC_DIV_EN		(1 << 16)
+#define  DPIO_CHV_FIRST_MOD		(0 << 8)
+#define  DPIO_CHV_SECOND_MOD		(1 << 8)
+#define  DPIO_CHV_FEEDFWD_GAIN_SHIFT	0
+#define CHV_PLL_DW3(ch) _PIPE(ch, _CHV_PLL_DW3_CH0, _CHV_PLL_DW3_CH1)
+
+#define _CHV_PLL_DW6_CH0		0x8018
+#define _CHV_PLL_DW6_CH1		0x8198
+#define   DPIO_CHV_GAIN_CTRL_SHIFT	16
+#define	  DPIO_CHV_INT_COEFF_SHIFT	8
+#define   DPIO_CHV_PROP_COEFF_SHIFT	0
+#define CHV_PLL_DW6(ch) _PIPE(ch, _CHV_PLL_DW6_CH0, _CHV_PLL_DW6_CH1)
+
+#define _CHV_CMN_DW13_CH0		0x8134
+#define _CHV_CMN_DW0_CH1		0x8080
+#define   DPIO_CHV_S1_DIV_SHIFT		21
+#define   DPIO_CHV_P1_DIV_SHIFT		13 /* 3 bits */
+#define   DPIO_CHV_P2_DIV_SHIFT		8  /* 5 bits */
+#define   DPIO_CHV_K_DIV_SHIFT		4
+#define   DPIO_PLL_FREQLOCK		(1 << 1)
+#define   DPIO_PLL_LOCK			(1 << 0)
+#define CHV_CMN_DW13(ch) _PIPE(ch, _CHV_CMN_DW13_CH0, _CHV_CMN_DW0_CH1)
+
+#define _CHV_CMN_DW14_CH0		0x8138
+#define _CHV_CMN_DW1_CH1		0x8084
+#define   DPIO_AFC_RECAL		(1 << 14)
+#define   DPIO_DCLKP_EN			(1 << 13)
+#define CHV_CMN_DW14(ch) _PIPE(ch, _CHV_CMN_DW14_CH0, _CHV_CMN_DW1_CH1)
+
+#define CHV_CMN_DW30			0x8178
+#define   DPIO_LRC_BYPASS		(1 << 3)
+
+#define _TXLANE(ch, lane, offset) ((ch ? 0x2400 : 0) + \
+					(lane) * 0x200 + (offset))
+
+#define CHV_TX_DW0(ch, lane) _TXLANE(ch, lane, 0x80)
+#define CHV_TX_DW1(ch, lane) _TXLANE(ch, lane, 0x84)
+#define CHV_TX_DW2(ch, lane) _TXLANE(ch, lane, 0x88)
+#define CHV_TX_DW3(ch, lane) _TXLANE(ch, lane, 0x8c)
+#define CHV_TX_DW4(ch, lane) _TXLANE(ch, lane, 0x90)
+#define CHV_TX_DW5(ch, lane) _TXLANE(ch, lane, 0x94)
+#define CHV_TX_DW6(ch, lane) _TXLANE(ch, lane, 0x98)
+#define CHV_TX_DW7(ch, lane) _TXLANE(ch, lane, 0x9c)
+#define CHV_TX_DW8(ch, lane) _TXLANE(ch, lane, 0xa0)
+#define CHV_TX_DW9(ch, lane) _TXLANE(ch, lane, 0xa4)
+#define CHV_TX_DW10(ch, lane) _TXLANE(ch, lane, 0xa8)
+#define CHV_TX_DW11(ch, lane) _TXLANE(ch, lane, 0xac)
+#define   DPIO_FRC_LATENCY_SHFIT	8
+#define CHV_TX_DW14(ch, lane) _TXLANE(ch, lane, 0xb8)
+#define   DPIO_UPAR_SHIFT		30
 /*
  * Fence registers
  */
@@ -628,10 +942,14 @@
 /*
  * Instruction and interrupt control regs
  */
+#define PGTBL_CTL	0x02020
+#define   PGTBL_ADDRESS_LO_MASK	0xfffff000 /* bits [31:12] */
+#define   PGTBL_ADDRESS_HI_MASK	0x000000f0 /* bits [35:32] (gen4) */
 #define PGTBL_ER	0x02024
 #define RENDER_RING_BASE	0x02000
 #define BSD_RING_BASE		0x04000
 #define GEN6_BSD_RING_BASE	0x12000
+#define GEN8_BSD2_RING_BASE	0x1c000
 #define VEBOX_RING_BASE		0x1a000
 #define BLT_RING_BASE		0x22000
 #define RING_TAIL(base)		((base)+0x30)
@@ -657,9 +975,20 @@
 #define RING_MAX_IDLE(base)	((base)+0x54)
 #define RING_HWS_PGA(base)	((base)+0x80)
 #define RING_HWS_PGA_GEN6(base)	((base)+0x2080)
-#define ARB_MODE		0x04030
+
+#define GEN7_WR_WATERMARK	0x4028
+#define GEN7_GFX_PRIO_CTRL	0x402C
+#define ARB_MODE		0x4030
 #define   ARB_MODE_SWIZZLE_SNB	(1<<4)
 #define   ARB_MODE_SWIZZLE_IVB	(1<<5)
+#define GEN7_GFX_PEND_TLB0	0x4034
+#define GEN7_GFX_PEND_TLB1	0x4038
+/* L3, CVS, ZTLB, RCC, CASC LRA min, max values */
+#define GEN7_LRA_LIMITS_BASE	0x403C
+#define GEN7_LRA_LIMITS_REG_NUM	13
+#define GEN7_MEDIA_MAX_REQ_COUNT	0x4070
+#define GEN7_GFX_MAX_REQ_COUNT		0x4074
+
 #define GAMTARBMODE		0x04a08
 #define   ARB_MODE_BWGTLB_DISABLE (1<<9)
 #define   ARB_MODE_SWIZZLE_BDW	(1<<1)
@@ -675,6 +1004,7 @@
 #define BLT_HWS_PGA_GEN7	(0x04280)
 #define VEBOX_HWS_PGA_GEN7	(0x04380)
 #define RING_ACTHD(base)	((base)+0x74)
+#define RING_ACTHD_UDW(base)	((base)+0x5c)
 #define RING_NOPID(base)	((base)+0x94)
 #define RING_IMR(base)		((base)+0xa8)
 #define RING_TIMESTAMP(base)	((base)+0x358)
@@ -693,6 +1023,9 @@
 #define   RING_WAIT_I8XX	(1<<0) /* gen2, PRBx_HEAD */
 #define   RING_WAIT		(1<<11) /* gen3+, PRBx_CTL */
 #define   RING_WAIT_SEMAPHORE	(1<<10) /* gen6+ */
+
+#define GEN7_TLB_RD_ADDR	0x4700
+
 #if 0
 #define PRB0_TAIL	0x02030
 #define PRB0_HEAD	0x02034
@@ -716,7 +1049,9 @@
 #define RING_INSTDONE(base)	((base)+0x6c)
 #define RING_INSTPS(base)	((base)+0x70)
 #define RING_DMA_FADD(base)	((base)+0x78)
+#define RING_DMA_FADD_UDW(base)	((base)+0x60) /* gen8+ */
 #define RING_INSTPM(base)	((base)+0xc0)
+#define RING_MI_MODE(base)	((base)+0x9c)
 #define INSTPS		0x02070 /* 965+ only */
 #define INSTDONE1	0x0207c /* 965+ only */
 #define ACTHD_I965	0x02074
@@ -732,6 +1067,8 @@
 #define HWSTAM		0x02098
 #define DMA_FADD_I8XX	0x020d0
 #define RING_BBSTATE(base)	((base)+0x110)
+#define RING_BBADDR(base)	((base)+0x140)
+#define RING_BBADDR_UDW(base)	((base)+0x168) /* gen8+ */
 
 #define ERROR_GEN6	0x040a0
 #define GEN7_ERR_INT	0x44040
@@ -784,36 +1121,49 @@
 #define _3D_CHICKEN3	0x02090
 #define  _3D_CHICKEN_SF_DISABLE_OBJEND_CULL		(1 << 10)
 #define  _3D_CHICKEN3_SF_DISABLE_FASTCLIP_CULL		(1 << 5)
-#define  _3D_CHICKEN_SDE_LIMIT_FIFO_POLY_DEPTH(x)	((x)<<1)
+#define  _3D_CHICKEN_SDE_LIMIT_FIFO_POLY_DEPTH(x)	((x)<<1) /* gen8+ */
+#define  _3D_CHICKEN3_SF_DISABLE_PIPELINED_ATTR_FETCH	(1 << 1) /* gen6 */
 
 #define MI_MODE		0x0209c
 # define VS_TIMER_DISPATCH				(1 << 6)
 # define MI_FLUSH_ENABLE				(1 << 12)
 # define ASYNC_FLIP_PERF_DISABLE			(1 << 14)
+# define MODE_IDLE					(1 << 9)
+# define STOP_RING					(1 << 8)
 
 #define GEN6_GT_MODE	0x20d0
-#define   GEN6_GT_MODE_HI				(1 << 9)
+#define GEN7_GT_MODE	0x7008
+#define   GEN6_WIZ_HASHING(hi, lo)			(((hi) << 9) | ((lo) << 7))
+#define   GEN6_WIZ_HASHING_8x8				GEN6_WIZ_HASHING(0, 0)
+#define   GEN6_WIZ_HASHING_8x4				GEN6_WIZ_HASHING(0, 1)
+#define   GEN6_WIZ_HASHING_16x4				GEN6_WIZ_HASHING(1, 0)
+#define   GEN6_WIZ_HASHING_MASK				(GEN6_WIZ_HASHING(1, 1) << 16)
 #define   GEN6_TD_FOUR_ROW_DISPATCH_DISABLE		(1 << 5)
 
 #define GFX_MODE	0x02520
 #define GFX_MODE_GEN7	0x0229c
 #define RING_MODE_GEN7(ring)	((ring)->mmio_base+0x29c)
 #define   GFX_RUN_LIST_ENABLE		(1<<15)
-#define   GFX_TLB_INVALIDATE_ALWAYS	(1<<13)
+#define   GFX_TLB_INVALIDATE_EXPLICIT	(1<<13)
 #define   GFX_SURFACE_FAULT_ENABLE	(1<<12)
 #define   GFX_REPLAY_MODE		(1<<11)
 #define   GFX_PSMI_GRANULARITY		(1<<10)
 #define   GFX_PPGTT_ENABLE		(1<<9)
 
 #define VLV_DISPLAY_BASE 0x180000
+#define VLV_MIPI_BASE VLV_DISPLAY_BASE
 
+#define VLV_GU_CTL0	(VLV_DISPLAY_BASE + 0x2030)
+#define VLV_GU_CTL1	(VLV_DISPLAY_BASE + 0x2034)
 #define SCPD0		0x0209c /* 915+ only */
 #define IER		0x020a0
 #define IIR		0x020a4
 #define IMR		0x020a8
 #define ISR		0x020ac
 #define VLV_GUNIT_CLOCK_GATE	(VLV_DISPLAY_BASE + 0x2060)
+#define   GINT_DIS		(1<<22)
 #define   GCFG_DIS		(1<<8)
+#define VLV_GUNIT_CLOCK_GATE2	(VLV_DISPLAY_BASE + 0x2064)
 #define VLV_IIR_RW	(VLV_DISPLAY_BASE + 0x2084)
 #define VLV_IER		(VLV_DISPLAY_BASE + 0x20a0)
 #define VLV_IIR		(VLV_DISPLAY_BASE + 0x20a4)
@@ -832,7 +1182,7 @@
 #define   I915_ERROR_INSTRUCTION			(1<<0)
 #define INSTPM	        0x020c0
 #define   INSTPM_SELF_EN (1<<12) /* 915GM only */
-#define   INSTPM_AGPBUSY_DIS (1<<11) /* gen3: when disabled, pending interrupts
+#define   INSTPM_AGPBUSY_INT_EN (1<<11) /* gen3: when disabled, pending interrupts
 					will not assert AGPBUSY# and will only
 					be delivered when out of C3. */
 #define   INSTPM_FORCE_ORDERING				(1<<7) /* GEN6+ */
@@ -913,6 +1263,10 @@
 #define   MI_ARB_DISPLAY_PRIORITY_A_B		(0 << 0)	/* display A > display B */
 #define   MI_ARB_DISPLAY_PRIORITY_B_A		(1 << 0)	/* display B > display A */
 
+#define MI_STATE	0x020e4 /* gen2 only */
+#define   MI_AGPBUSY_INT_EN			(1 << 1) /* 85x only */
+#define   MI_AGPBUSY_830_MODE			(1 << 0) /* 85x only */
+
 #define CACHE_MODE_0	0x02120 /* 915+ only */
 #define   CM0_PIPELINED_RENDER_FLUSH_DISABLE (1<<8)
 #define   CM0_IZ_OPT_DISABLE      (1<<6)
@@ -922,7 +1276,6 @@
 #define   CM0_COLOR_EVICT_DISABLE (1<<3)
 #define   CM0_DEPTH_WRITE_DISABLE (1<<1)
 #define   CM0_RC_OP_FLUSH_DISABLE (1<<0)
-#define BB_ADDR		0x02140 /* 8 bytes */
 #define GFX_FLSH_CNTL	0x02170 /* 915+ only */
 #define GFX_FLSH_CNTL_GEN6	0x101008
 #define   GFX_FLSH_CNTL_EN	(1<<0)
@@ -930,13 +1283,21 @@
 #define   ECO_GATING_CX_ONLY	(1<<3)
 #define   ECO_FLIP_DONE		(1<<0)
 
+#define CACHE_MODE_0_GEN7	0x7000 /* IVB+ */
+#define RC_OP_FLUSH_ENABLE (1<<0)
+#define   HIZ_RAW_STALL_OPT_DISABLE (1<<2)
 #define CACHE_MODE_1		0x7004 /* IVB+ */
-#define   PIXEL_SUBSPAN_COLLECT_OPT_DISABLE (1<<6)
+#define   PIXEL_SUBSPAN_COLLECT_OPT_DISABLE	(1<<6)
+#define   GEN8_4x4_STC_OPTIMIZATION_DISABLE	(1<<6)
 
 #define GEN6_BLITTER_ECOSKPD	0x221d0
 #define   GEN6_BLITTER_LOCK_SHIFT			16
 #define   GEN6_BLITTER_FBC_NOTIFY			(1<<3)
 
+#define GEN6_RC_SLEEP_PSMI_CONTROL	0x2050
+#define   GEN8_RC_SEMA_IDLE_MSG_DISABLE	(1 << 12)
+#define   GEN8_FF_DOP_CLOCK_GATE_DISABLE	(1<<10)
+
 #define GEN6_BSD_SLEEP_PSMI_CONTROL	0x12050
 #define   GEN6_BSD_SLEEP_MSG_DISABLE	(1 << 0)
 #define   GEN6_BSD_SLEEP_FLUSH_DISABLE	(1 << 2)
@@ -976,29 +1337,49 @@
 
 /* These are all the "old" interrupts */
 #define ILK_BSD_USER_INTERRUPT				(1<<5)
+
+#define I915_PM_INTERRUPT				(1<<31)
+#define I915_ISP_INTERRUPT				(1<<22)
+#define I915_LPE_PIPE_B_INTERRUPT			(1<<21)
+#define I915_LPE_PIPE_A_INTERRUPT			(1<<20)
+#define I915_MIPIB_INTERRUPT				(1<<19)
+#define I915_MIPIA_INTERRUPT				(1<<18)
 #define I915_PIPE_CONTROL_NOTIFY_INTERRUPT		(1<<18)
 #define I915_DISPLAY_PORT_INTERRUPT			(1<<17)
+#define I915_DISPLAY_PIPE_C_HBLANK_INTERRUPT		(1<<16)
+#define I915_MASTER_ERROR_INTERRUPT			(1<<15)
 #define I915_RENDER_COMMAND_PARSER_ERROR_INTERRUPT	(1<<15)
+#define I915_DISPLAY_PIPE_B_HBLANK_INTERRUPT		(1<<14)
 #define I915_GMCH_THERMAL_SENSOR_EVENT_INTERRUPT	(1<<14) /* p-state */
+#define I915_DISPLAY_PIPE_A_HBLANK_INTERRUPT		(1<<13)
 #define I915_HWB_OOM_INTERRUPT				(1<<13)
+#define I915_LPE_PIPE_C_INTERRUPT			(1<<12)
 #define I915_SYNC_STATUS_INTERRUPT			(1<<12)
+#define I915_MISC_INTERRUPT				(1<<11)
 #define I915_DISPLAY_PLANE_A_FLIP_PENDING_INTERRUPT	(1<<11)
+#define I915_DISPLAY_PIPE_C_VBLANK_INTERRUPT		(1<<10)
 #define I915_DISPLAY_PLANE_B_FLIP_PENDING_INTERRUPT	(1<<10)
+#define I915_DISPLAY_PIPE_C_EVENT_INTERRUPT		(1<<9)
 #define I915_OVERLAY_PLANE_FLIP_PENDING_INTERRUPT	(1<<9)
+#define I915_DISPLAY_PIPE_C_DPBM_INTERRUPT		(1<<8)
 #define I915_DISPLAY_PLANE_C_FLIP_PENDING_INTERRUPT	(1<<8)
 #define I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT		(1<<7)
 #define I915_DISPLAY_PIPE_A_EVENT_INTERRUPT		(1<<6)
 #define I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT		(1<<5)
 #define I915_DISPLAY_PIPE_B_EVENT_INTERRUPT		(1<<4)
+#define I915_DISPLAY_PIPE_A_DPBM_INTERRUPT		(1<<3)
+#define I915_DISPLAY_PIPE_B_DPBM_INTERRUPT		(1<<2)
 #define I915_DEBUG_INTERRUPT				(1<<2)
+#define I915_WINVALID_INTERRUPT				(1<<1)
 #define I915_USER_INTERRUPT				(1<<1)
 #define I915_ASLE_INTERRUPT				(1<<0)
-#define I915_BSD_USER_INTERRUPT				(1 << 25)
+#define I915_BSD_USER_INTERRUPT				(1<<25)
 
 #define GEN6_BSD_RNCID			0x12198
 
 #define GEN7_FF_THREAD_MODE		0x20a0
 #define   GEN7_FF_SCHED_MASK		0x0077070
+#define   GEN8_FF_DS_REF_CNT_FFME	(1 << 19)
 #define   GEN7_FF_TS_SCHED_HS1		(0x5<<16)
 #define   GEN7_FF_TS_SCHED_HS0		(0x3<<16)
 #define   GEN7_FF_TS_SCHED_LOAD_BALANCE	(0x1<<16)
@@ -1026,14 +1407,14 @@
 #define   FBC_CTL_UNCOMPRESSIBLE (1<<14)
 #define   FBC_CTL_C3_IDLE	(1<<13)
 #define   FBC_CTL_STRIDE_SHIFT	(5)
-#define   FBC_CTL_FENCENO	(1<<0)
+#define   FBC_CTL_FENCENO_SHIFT	(0)
 #define FBC_COMMAND		0x0320c
 #define   FBC_CMD_COMPRESS	(1<<0)
 #define FBC_STATUS		0x03210
 #define   FBC_STAT_COMPRESSING	(1<<31)
 #define   FBC_STAT_COMPRESSED	(1<<30)
 #define   FBC_STAT_MODIFIED	(1<<29)
-#define   FBC_STAT_CURRENT_LINE	(1<<0)
+#define   FBC_STAT_CURRENT_LINE_SHIFT	(0)
 #define FBC_CONTROL2		0x03214
 #define   FBC_CTL_FENCE_DBL	(0<<4)
 #define   FBC_CTL_IDLE_IMM	(0<<2)
@@ -1041,9 +1422,8 @@
 #define   FBC_CTL_IDLE_LINE	(2<<2)
 #define   FBC_CTL_IDLE_DEBUG	(3<<2)
 #define   FBC_CTL_CPU_FENCE	(1<<1)
-#define   FBC_CTL_PLANEA	(0<<0)
-#define   FBC_CTL_PLANEB	(1<<0)
-#define FBC_FENCE_OFF		0x0321b
+#define   FBC_CTL_PLANE(plane)	((plane)<<0)
+#define FBC_FENCE_OFF		0x03218 /* BSpec typo has 321Bh */
 #define FBC_TAG			0x03300
 
 #define FBC_LL_SIZE		(1536)
@@ -1052,9 +1432,8 @@
 #define DPFC_CB_BASE		0x3200
 #define DPFC_CONTROL		0x3208
 #define   DPFC_CTL_EN		(1<<31)
-#define   DPFC_CTL_PLANEA	(0<<30)
-#define   DPFC_CTL_PLANEB	(1<<30)
-#define   IVB_DPFC_CTL_PLANE_SHIFT	(29)
+#define   DPFC_CTL_PLANE(plane)	((plane)<<30)
+#define   IVB_DPFC_CTL_PLANE(plane)	((plane)<<29)
 #define   DPFC_CTL_FENCE_EN	(1<<29)
 #define   IVB_DPFC_CTL_FENCE_EN	(1<<28)
 #define   DPFC_CTL_PERSISTENT_MODE	(1<<25)
@@ -1115,13 +1494,6 @@
 #define   FBC_REND_NUKE		(1<<2)
 #define   FBC_REND_CACHE_CLEAN	(1<<1)
 
-#define _HSW_PIPE_SLICE_CHICKEN_1_A	0x420B0
-#define _HSW_PIPE_SLICE_CHICKEN_1_B	0x420B4
-#define   HSW_BYPASS_FBC_QUEUE		(1<<22)
-#define HSW_PIPE_SLICE_CHICKEN_1(pipe) _PIPE(pipe, + \
-					     _HSW_PIPE_SLICE_CHICKEN_1_A, + \
-					     _HSW_PIPE_SLICE_CHICKEN_1_B)
-
 /*
  * GPIO regs
  */
@@ -1158,6 +1530,7 @@
 #define   GMBUS_PORT_SSC	1
 #define   GMBUS_PORT_VGADDC	2
 #define   GMBUS_PORT_PANEL	3
+#define   GMBUS_PORT_DPD_CHV	3 /* HDMID_CHV */
 #define   GMBUS_PORT_DPC	4 /* HDMIC */
 #define   GMBUS_PORT_DPB	5 /* SDVO, HDMIB */
 #define   GMBUS_PORT_DPD	6 /* HDMID */
@@ -1197,6 +1570,11 @@
 /*
  * Clock control & power management
  */
+#define DPLL_A_OFFSET 0x6014
+#define DPLL_B_OFFSET 0x6018
+#define CHV_DPLL_C_OFFSET 0x6030
+#define DPLL(pipe) (dev_priv->info.dpll_offsets[pipe] + \
+		    dev_priv->info.display_mmio_offset)
 
 #define VGA0	0x6000
 #define VGA1	0x6004
@@ -1209,9 +1587,6 @@
 #define   VGA1_PD_P1_DIV_2	(1 << 13)
 #define   VGA1_PD_P1_SHIFT	8
 #define   VGA1_PD_P1_MASK	(0x1f << 8)
-#define _DPLL_A	(dev_priv->info->display_mmio_offset + 0x6014)
-#define _DPLL_B	(dev_priv->info->display_mmio_offset + 0x6018)
-#define DPLL(pipe) _PIPE(pipe, _DPLL_A, _DPLL_B)
 #define   DPLL_VCO_ENABLE		(1 << 31)
 #define   DPLL_SDVO_HIGH_SPEED		(1 << 30)
 #define   DPLL_DVO_2X_MODE		(1 << 30)
@@ -1232,10 +1607,23 @@
 #define   DPLL_LOCK_VLV			(1<<15)
 #define   DPLL_INTEGRATED_CRI_CLK_VLV	(1<<14)
 #define   DPLL_INTEGRATED_CLOCK_VLV	(1<<13)
+#define   DPLL_SSC_REF_CLOCK_CHV	(1<<13)
 #define   DPLL_PORTC_READY_MASK		(0xf << 4)
 #define   DPLL_PORTB_READY_MASK		(0xf)
 
 #define   DPLL_FPA01_P1_POST_DIV_MASK_I830	0x001f0000
+
+/* Additional CHV pll/phy registers */
+#define DPIO_PHY_STATUS			(VLV_DISPLAY_BASE + 0x6240)
+#define   DPLL_PORTD_READY_MASK		(0xf)
+#define DISPLAY_PHY_CONTROL (VLV_DISPLAY_BASE + 0x60100)
+#define   PHY_COM_LANE_RESET_DEASSERT(phy, val) \
+				((phy == DPIO_PHY0) ? (val | 1) : (val | 2))
+#define   PHY_COM_LANE_RESET_ASSERT(phy, val) \
+				((phy == DPIO_PHY0) ? (val & ~1) : (val & ~2))
+#define DISPLAY_PHY_STATUS (VLV_DISPLAY_BASE + 0x60104)
+#define   PHY_POWERGOOD(phy)	((phy == DPIO_PHY0) ? (1<<31) : (1<<30))
+
 /*
  * The i830 generation, in LVDS mode, defines P1 as the bit number set within
  * this field (only one bit may be set).
@@ -1273,7 +1661,13 @@
 #define   SDVO_MULTIPLIER_MASK			0x000000ff
 #define   SDVO_MULTIPLIER_SHIFT_HIRES		4
 #define   SDVO_MULTIPLIER_SHIFT_VGA		0
-#define _DPLL_A_MD (dev_priv->info->display_mmio_offset + 0x601c) /* 965+ only */
+
+#define DPLL_A_MD_OFFSET 0x601c /* 965+ only */
+#define DPLL_B_MD_OFFSET 0x6020 /* 965+ only */
+#define CHV_DPLL_C_MD_OFFSET 0x603c
+#define DPLL_MD(pipe) (dev_priv->info.dpll_md_offsets[pipe] + \
+		       dev_priv->info.display_mmio_offset)
+
 /*
  * UDI pixel divider, controlling how many pixels are stuffed into a packet.
  *
@@ -1310,8 +1704,6 @@
  */
 #define   DPLL_MD_VGA_UDI_MULTIPLIER_MASK	0x0000003f
 #define   DPLL_MD_VGA_UDI_MULTIPLIER_SHIFT	0
-#define _DPLL_B_MD (dev_priv->info->display_mmio_offset + 0x6020) /* 965+ only */
-#define DPLL_MD(pipe) _PIPE(pipe, _DPLL_A_MD, _DPLL_B_MD)
 
 #define _FPA0	0x06040
 #define _FPA1	0x06044
@@ -1343,7 +1735,7 @@
 #define  DSTATE_PLL_D3_OFF			(1<<3)
 #define  DSTATE_GFX_CLOCK_GATING		(1<<1)
 #define  DSTATE_DOT_CLOCK_GATING		(1<<0)
-#define DSPCLK_GATE_D	(dev_priv->info->display_mmio_offset + 0x6200)
+#define DSPCLK_GATE_D	(dev_priv->info.display_mmio_offset + 0x6200)
 # define DPUNIT_B_CLOCK_GATE_DISABLE		(1 << 30) /* 965 */
 # define VSUNIT_CLOCK_GATE_DISABLE		(1 << 29) /* 965 */
 # define VRHUNIT_CLOCK_GATE_DISABLE		(1 << 28) /* 965 */
@@ -1372,7 +1764,7 @@
 # define DPIOUNIT_CLOCK_GATE_DISABLE		(1 << 6) /* 915-945 */
 # define OVFUNIT_CLOCK_GATE_DISABLE		(1 << 5)
 # define OVBUNIT_CLOCK_GATE_DISABLE		(1 << 4)
-/**
+/*
  * This bit must be set on the 830 to prevent hangs when turning off the
  * overlay scaler.
  */
@@ -1392,12 +1784,12 @@
 # define COLOR_CALCULATOR_CLOCK_GATE_DISABLE	(1 << 7)
 # define MOTION_COMP_CLOCK_GATE_DISABLE		(1 << 6)
 # define MAG_CLOCK_GATE_DISABLE			(1 << 5)
-/** This bit must be unset on 855,865 */
+/* This bit must be unset on 855,865 */
 # define MECI_CLOCK_GATE_DISABLE		(1 << 4)
 # define DCMP_CLOCK_GATE_DISABLE		(1 << 3)
 # define MEC_CLOCK_GATE_DISABLE			(1 << 2)
 # define MECO_CLOCK_GATE_DISABLE		(1 << 1)
-/** This bit must be set on 855,865. */
+/* This bit must be set on 855,865. */
 # define SV_CLOCK_GATE_DISABLE			(1 << 0)
 # define I915_MPEG_CLOCK_GATE_DISABLE		(1 << 16)
 # define I915_VLD_IP_PR_CLOCK_GATE_DISABLE	(1 << 15)
@@ -1418,14 +1810,14 @@
 # define I915_BY_CLOCK_GATE_DISABLE		(1 << 0)
 
 # define I965_RCZ_CLOCK_GATE_DISABLE		(1 << 30)
-/** This bit must always be set on 965G/965GM */
+/* This bit must always be set on 965G/965GM */
 # define I965_RCC_CLOCK_GATE_DISABLE		(1 << 29)
 # define I965_RCPB_CLOCK_GATE_DISABLE		(1 << 28)
 # define I965_DAP_CLOCK_GATE_DISABLE		(1 << 27)
 # define I965_ROC_CLOCK_GATE_DISABLE		(1 << 26)
 # define I965_GW_CLOCK_GATE_DISABLE		(1 << 25)
 # define I965_TD_CLOCK_GATE_DISABLE		(1 << 24)
-/** This bit must always be set on 965G */
+/* This bit must always be set on 965G */
 # define I965_ISC_CLOCK_GATE_DISABLE		(1 << 23)
 # define I965_IC_CLOCK_GATE_DISABLE		(1 << 22)
 # define I965_EU_CLOCK_GATE_DISABLE		(1 << 21)
@@ -1450,6 +1842,10 @@
 #define VF_UNIT_CLOCK_GATE_DISABLE		(1 << 9)
 #define GS_UNIT_CLOCK_GATE_DISABLE		(1 << 7)
 #define CL_UNIT_CLOCK_GATE_DISABLE		(1 << 6)
+
+#define VDECCLK_GATE_D		0x620C		/* g4x only */
+#define  VCP_UNIT_CLOCK_GATE_DISABLE		(1 << 4)
+
 #define RAMCLK_GATE_D		0x6210		/* CRL only */
 #define DEUC			0x6214          /* CRL only */
 
@@ -1467,10 +1863,11 @@
 /*
  * Palette regs
  */
-
-#define _PALETTE_A		(dev_priv->info->display_mmio_offset + 0xa000)
-#define _PALETTE_B		(dev_priv->info->display_mmio_offset + 0xa800)
-#define PALETTE(pipe) _PIPE(pipe, _PALETTE_A, _PALETTE_B)
+#define PALETTE_A_OFFSET 0xa000
+#define PALETTE_B_OFFSET 0xa800
+#define CHV_PALETTE_C_OFFSET 0xc000
+#define PALETTE(pipe) (dev_priv->info.palette_offsets[pipe] + \
+		       dev_priv->info.display_mmio_offset)
 
 /* MCH MMIO space */
 
@@ -1491,7 +1888,7 @@
 /* Memory controller frequency in MCHBAR for Haswell (possible SNB+) */
 #define DCLK (MCHBAR_MIRROR_BASE_SNB + 0x5e04)
 
-/** 915-945 and GM965 MCH register controlling DRAM channel access */
+/* 915-945 and GM965 MCH register controlling DRAM channel access */
 #define DCC			0x10200
 #define DCC_ADDRESSING_MODE_SINGLE_CHANNEL		(0 << 0)
 #define DCC_ADDRESSING_MODE_DUAL_CHANNEL_ASYMMETRIC	(1 << 0)
@@ -1500,15 +1897,15 @@
 #define DCC_CHANNEL_XOR_DISABLE				(1 << 10)
 #define DCC_CHANNEL_XOR_BIT_17				(1 << 9)
 
-/** Pineview MCH register contains DDR3 setting */
+/* Pineview MCH register contains DDR3 setting */
 #define CSHRDDR3CTL            0x101a8
 #define CSHRDDR3CTL_DDR3       (1 << 2)
 
-/** 965 MCH register controlling DRAM channel configuration */
+/* 965 MCH register controlling DRAM channel configuration */
 #define C0DRB3			0x10206
 #define C1DRB3			0x10606
 
-/** snb MCH registers for reading the DRAM channel configuration */
+/* snb MCH registers for reading the DRAM channel configuration */
 #define MAD_DIMM_C0			(MCHBAR_MIRROR_BASE_SNB + 0x5004)
 #define MAD_DIMM_C1			(MCHBAR_MIRROR_BASE_SNB + 0x5008)
 #define MAD_DIMM_C2			(MCHBAR_MIRROR_BASE_SNB + 0x500C)
@@ -1530,7 +1927,7 @@
 #define   MAD_DIMM_A_SIZE_SHIFT		0
 #define   MAD_DIMM_A_SIZE_MASK		(0xff << MAD_DIMM_A_SIZE_SHIFT)
 
-/** snb MCH registers for priority tuning */
+/* snb MCH registers for priority tuning */
 #define MCH_SSKPD			(MCHBAR_MIRROR_BASE_SNB + 0x5d10)
 #define   MCH_SSKPD_WM0_MASK		0x3f
 #define   MCH_SSKPD_WM0_VAL		0xc
@@ -1857,7 +2254,7 @@
  */
 
 /* Pipe A CRC regs */
-#define _PIPE_CRC_CTL_A		(dev_priv->info->display_mmio_offset + 0x60050)
+#define _PIPE_CRC_CTL_A			0x60050
 #define   PIPE_CRC_ENABLE		(1 << 31)
 /* ivb+ source selection */
 #define   PIPE_CRC_SOURCE_PRIMARY_IVB	(0 << 29)
@@ -1897,11 +2294,11 @@
 #define _PIPE_CRC_RES_4_A_IVB		0x60070
 #define _PIPE_CRC_RES_5_A_IVB		0x60074
 
-#define _PIPE_CRC_RES_RED_A		(dev_priv->info->display_mmio_offset + 0x60060)
-#define _PIPE_CRC_RES_GREEN_A		(dev_priv->info->display_mmio_offset + 0x60064)
-#define _PIPE_CRC_RES_BLUE_A		(dev_priv->info->display_mmio_offset + 0x60068)
-#define _PIPE_CRC_RES_RES1_A_I915	(dev_priv->info->display_mmio_offset + 0x6006c)
-#define _PIPE_CRC_RES_RES2_A_G4X	(dev_priv->info->display_mmio_offset + 0x60080)
+#define _PIPE_CRC_RES_RED_A		0x60060
+#define _PIPE_CRC_RES_GREEN_A		0x60064
+#define _PIPE_CRC_RES_BLUE_A		0x60068
+#define _PIPE_CRC_RES_RES1_A_I915	0x6006c
+#define _PIPE_CRC_RES_RES2_A_G4X	0x60080
 
 /* Pipe B CRC regs */
 #define _PIPE_CRC_RES_1_B_IVB		0x61064
@@ -1910,59 +2307,70 @@
 #define _PIPE_CRC_RES_4_B_IVB		0x61070
 #define _PIPE_CRC_RES_5_B_IVB		0x61074
 
-#define PIPE_CRC_CTL(pipe)	_PIPE_INC(pipe, _PIPE_CRC_CTL_A, 0x01000)
+#define PIPE_CRC_CTL(pipe) _TRANSCODER2(pipe, _PIPE_CRC_CTL_A)
 #define PIPE_CRC_RES_1_IVB(pipe)	\
-	_PIPE(pipe, _PIPE_CRC_RES_1_A_IVB, _PIPE_CRC_RES_1_B_IVB)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_1_A_IVB)
 #define PIPE_CRC_RES_2_IVB(pipe)	\
-	_PIPE(pipe, _PIPE_CRC_RES_2_A_IVB, _PIPE_CRC_RES_2_B_IVB)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_2_A_IVB)
 #define PIPE_CRC_RES_3_IVB(pipe)	\
-	_PIPE(pipe, _PIPE_CRC_RES_3_A_IVB, _PIPE_CRC_RES_3_B_IVB)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_3_A_IVB)
 #define PIPE_CRC_RES_4_IVB(pipe)	\
-	_PIPE(pipe, _PIPE_CRC_RES_4_A_IVB, _PIPE_CRC_RES_4_B_IVB)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_4_A_IVB)
 #define PIPE_CRC_RES_5_IVB(pipe)	\
-	_PIPE(pipe, _PIPE_CRC_RES_5_A_IVB, _PIPE_CRC_RES_5_B_IVB)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_5_A_IVB)
 
 #define PIPE_CRC_RES_RED(pipe) \
-	_PIPE_INC(pipe, _PIPE_CRC_RES_RED_A, 0x01000)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_RED_A)
 #define PIPE_CRC_RES_GREEN(pipe) \
-	_PIPE_INC(pipe, _PIPE_CRC_RES_GREEN_A, 0x01000)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_GREEN_A)
 #define PIPE_CRC_RES_BLUE(pipe) \
-	_PIPE_INC(pipe, _PIPE_CRC_RES_BLUE_A, 0x01000)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_BLUE_A)
 #define PIPE_CRC_RES_RES1_I915(pipe) \
-	_PIPE_INC(pipe, _PIPE_CRC_RES_RES1_A_I915, 0x01000)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_RES1_A_I915)
 #define PIPE_CRC_RES_RES2_G4X(pipe) \
-	_PIPE_INC(pipe, _PIPE_CRC_RES_RES2_A_G4X, 0x01000)
+	_TRANSCODER2(pipe, _PIPE_CRC_RES_RES2_A_G4X)
 
 /* Pipe A timing regs */
-#define _HTOTAL_A	(dev_priv->info->display_mmio_offset + 0x60000)
-#define _HBLANK_A	(dev_priv->info->display_mmio_offset + 0x60004)
-#define _HSYNC_A	(dev_priv->info->display_mmio_offset + 0x60008)
-#define _VTOTAL_A	(dev_priv->info->display_mmio_offset + 0x6000c)
-#define _VBLANK_A	(dev_priv->info->display_mmio_offset + 0x60010)
-#define _VSYNC_A	(dev_priv->info->display_mmio_offset + 0x60014)
-#define _PIPEASRC	(dev_priv->info->display_mmio_offset + 0x6001c)
-#define _BCLRPAT_A	(dev_priv->info->display_mmio_offset + 0x60020)
-#define _VSYNCSHIFT_A	(dev_priv->info->display_mmio_offset + 0x60028)
+#define _HTOTAL_A	0x60000
+#define _HBLANK_A	0x60004
+#define _HSYNC_A	0x60008
+#define _VTOTAL_A	0x6000c
+#define _VBLANK_A	0x60010
+#define _VSYNC_A	0x60014
+#define _PIPEASRC	0x6001c
+#define _BCLRPAT_A	0x60020
+#define _VSYNCSHIFT_A	0x60028
 
 /* Pipe B timing regs */
-#define _HTOTAL_B	(dev_priv->info->display_mmio_offset + 0x61000)
-#define _HBLANK_B	(dev_priv->info->display_mmio_offset + 0x61004)
-#define _HSYNC_B	(dev_priv->info->display_mmio_offset + 0x61008)
-#define _VTOTAL_B	(dev_priv->info->display_mmio_offset + 0x6100c)
-#define _VBLANK_B	(dev_priv->info->display_mmio_offset + 0x61010)
-#define _VSYNC_B	(dev_priv->info->display_mmio_offset + 0x61014)
-#define _PIPEBSRC	(dev_priv->info->display_mmio_offset + 0x6101c)
-#define _BCLRPAT_B	(dev_priv->info->display_mmio_offset + 0x61020)
-#define _VSYNCSHIFT_B	(dev_priv->info->display_mmio_offset + 0x61028)
-
-#define HTOTAL(trans) _TRANSCODER(trans, _HTOTAL_A, _HTOTAL_B)
-#define HBLANK(trans) _TRANSCODER(trans, _HBLANK_A, _HBLANK_B)
-#define HSYNC(trans) _TRANSCODER(trans, _HSYNC_A, _HSYNC_B)
-#define VTOTAL(trans) _TRANSCODER(trans, _VTOTAL_A, _VTOTAL_B)
-#define VBLANK(trans) _TRANSCODER(trans, _VBLANK_A, _VBLANK_B)
-#define VSYNC(trans) _TRANSCODER(trans, _VSYNC_A, _VSYNC_B)
-#define BCLRPAT(pipe) _PIPE(pipe, _BCLRPAT_A, _BCLRPAT_B)
-#define VSYNCSHIFT(trans) _TRANSCODER(trans, _VSYNCSHIFT_A, _VSYNCSHIFT_B)
+#define _HTOTAL_B	0x61000
+#define _HBLANK_B	0x61004
+#define _HSYNC_B	0x61008
+#define _VTOTAL_B	0x6100c
+#define _VBLANK_B	0x61010
+#define _VSYNC_B	0x61014
+#define _PIPEBSRC	0x6101c
+#define _BCLRPAT_B	0x61020
+#define _VSYNCSHIFT_B	0x61028
+
+#define TRANSCODER_A_OFFSET 0x60000
+#define TRANSCODER_B_OFFSET 0x61000
+#define TRANSCODER_C_OFFSET 0x62000
+#define CHV_TRANSCODER_C_OFFSET 0x63000
+#define TRANSCODER_EDP_OFFSET 0x6f000
+
+#define _TRANSCODER2(pipe, reg) (dev_priv->info.trans_offsets[(pipe)] - \
+	dev_priv->info.trans_offsets[TRANSCODER_A] + (reg) + \
+	dev_priv->info.display_mmio_offset)
+
+#define HTOTAL(trans) _TRANSCODER2(trans, _HTOTAL_A)
+#define HBLANK(trans) _TRANSCODER2(trans, _HBLANK_A)
+#define HSYNC(trans) _TRANSCODER2(trans, _HSYNC_A)
+#define VTOTAL(trans) _TRANSCODER2(trans, _VTOTAL_A)
+#define VBLANK(trans) _TRANSCODER2(trans, _VBLANK_A)
+#define VSYNC(trans) _TRANSCODER2(trans, _VSYNC_A)
+#define BCLRPAT(trans) _TRANSCODER2(trans, _BCLRPAT_A)
+#define VSYNCSHIFT(trans) _TRANSCODER2(trans, _VSYNCSHIFT_A)
+#define PIPESRC(trans) _TRANSCODER2(trans, _PIPEASRC)
 
 /* HSW+ eDP PSR registers */
 #define EDP_PSR_BASE(dev)                       (IS_HASWELL(dev) ? 0x64800 : 0x6f800)
@@ -2079,7 +2487,7 @@
 
 
 /* Hotplug control (945+ only) */
-#define PORT_HOTPLUG_EN		(dev_priv->info->display_mmio_offset + 0x61110)
+#define PORT_HOTPLUG_EN		(dev_priv->info.display_mmio_offset + 0x61110)
 #define   PORTB_HOTPLUG_INT_EN			(1 << 29)
 #define   PORTC_HOTPLUG_INT_EN			(1 << 28)
 #define   PORTD_HOTPLUG_INT_EN			(1 << 27)
@@ -2109,7 +2517,7 @@
 #define CRT_HOTPLUG_DETECT_VOLTAGE_325MV	(0 << 2)
 #define CRT_HOTPLUG_DETECT_VOLTAGE_475MV	(1 << 2)
 
-#define PORT_HOTPLUG_STAT	(dev_priv->info->display_mmio_offset + 0x61114)
+#define PORT_HOTPLUG_STAT	(dev_priv->info.display_mmio_offset + 0x61114)
 /*
  * HDMI/DP bits are gen4+
  *
@@ -2117,9 +2525,13 @@
  * Please check the detailed lore in the commit message for for experimental
  * evidence.
  */
-#define   PORTD_HOTPLUG_LIVE_STATUS               (1 << 29)
-#define   PORTC_HOTPLUG_LIVE_STATUS               (1 << 28)
-#define   PORTB_HOTPLUG_LIVE_STATUS               (1 << 27)
+#define   PORTD_HOTPLUG_LIVE_STATUS_G4X		(1 << 29)
+#define   PORTC_HOTPLUG_LIVE_STATUS_G4X		(1 << 28)
+#define   PORTB_HOTPLUG_LIVE_STATUS_G4X		(1 << 27)
+/* VLV DP/HDMI bits again match Bspec */
+#define   PORTD_HOTPLUG_LIVE_STATUS_VLV		(1 << 27)
+#define   PORTC_HOTPLUG_LIVE_STATUS_VLV		(1 << 28)
+#define   PORTB_HOTPLUG_LIVE_STATUS_VLV		(1 << 29)
 #define   PORTD_HOTPLUG_INT_STATUS		(3 << 21)
 #define   PORTC_HOTPLUG_INT_STATUS		(3 << 19)
 #define   PORTB_HOTPLUG_INT_STATUS		(3 << 17)
@@ -2130,6 +2542,11 @@
 #define   CRT_HOTPLUG_MONITOR_COLOR		(3 << 8)
 #define   CRT_HOTPLUG_MONITOR_MONO		(2 << 8)
 #define   CRT_HOTPLUG_MONITOR_NONE		(0 << 8)
+#define   DP_AUX_CHANNEL_D_INT_STATUS_G4X	(1 << 6)
+#define   DP_AUX_CHANNEL_C_INT_STATUS_G4X	(1 << 5)
+#define   DP_AUX_CHANNEL_B_INT_STATUS_G4X	(1 << 4)
+#define   DP_AUX_CHANNEL_MASK_INT_STATUS_G4X	(7 << 4)
+
 /* SDVO is different across gen3/4 */
 #define   SDVOC_HOTPLUG_INT_STATUS_G4X		(1 << 3)
 #define   SDVOB_HOTPLUG_INT_STATUS_G4X		(1 << 2)
@@ -2163,6 +2580,7 @@
 #define GEN3_SDVOC	0x61160
 #define GEN4_HDMIB	GEN3_SDVOB
 #define GEN4_HDMIC	GEN3_SDVOC
+#define CHV_HDMID	0x6116C
 #define PCH_SDVOB	0xe1140
 #define PCH_HDMIB	PCH_SDVOB
 #define PCH_HDMIC	0xe1150
@@ -2183,7 +2601,7 @@
 #define   SDVO_PIPE_B_SELECT			(1 << 30)
 #define   SDVO_STALL_SELECT			(1 << 29)
 #define   SDVO_INTERRUPT_ENABLE			(1 << 26)
-/**
+/*
  * 915G/GM SDVO pixel multiplier.
  * Programmed value is multiplier - 1, up to 5x.
  * \sa DPLL_MD_UDI_MULTIPLIER_MASK
@@ -2223,6 +2641,10 @@
 #define   SDVO_PIPE_SEL_CPT(pipe)		((pipe) << 29)
 #define   SDVO_PIPE_SEL_MASK_CPT		(3 << 29)
 
+/* CHV SDVO/HDMI bits: */
+#define   SDVO_PIPE_SEL_CHV(pipe)		((pipe) << 24)
+#define   SDVO_PIPE_SEL_MASK_CHV		(3 << 24)
+
 
 /* DVO port control */
 #define DVOA			0x61120
@@ -2318,9 +2740,7 @@
 #define VIDEO_DIP_CTL		0x61170
 /* Pre HSW: */
 #define   VIDEO_DIP_ENABLE		(1 << 31)
-#define   VIDEO_DIP_PORT_B		(1 << 29)
-#define   VIDEO_DIP_PORT_C		(2 << 29)
-#define   VIDEO_DIP_PORT_D		(3 << 29)
+#define   VIDEO_DIP_PORT(port)		((port) << 29)
 #define   VIDEO_DIP_PORT_MASK		(3 << 29)
 #define   VIDEO_DIP_ENABLE_GCP		(1 << 25)
 #define   VIDEO_DIP_ENABLE_AVI		(1 << 21)
@@ -2377,7 +2797,7 @@
 #define PP_DIVISOR	0x61210
 
 /* Panel fitting */
-#define PFIT_CONTROL	(dev_priv->info->display_mmio_offset + 0x61230)
+#define PFIT_CONTROL	(dev_priv->info.display_mmio_offset + 0x61230)
 #define   PFIT_ENABLE		(1 << 31)
 #define   PFIT_PIPE_MASK	(3 << 29)
 #define   PFIT_PIPE_SHIFT	29
@@ -2395,7 +2815,7 @@
 #define   PFIT_SCALING_PROGRAMMED (1 << 26)
 #define   PFIT_SCALING_PILLAR	(2 << 26)
 #define   PFIT_SCALING_LETTER	(3 << 26)
-#define PFIT_PGM_RATIOS	(dev_priv->info->display_mmio_offset + 0x61234)
+#define PFIT_PGM_RATIOS	(dev_priv->info.display_mmio_offset + 0x61234)
 /* Pre-965 */
 #define		PFIT_VERT_SCALE_SHIFT		20
 #define		PFIT_VERT_SCALE_MASK		0xfff00000
@@ -2407,25 +2827,25 @@
 #define		PFIT_HORIZ_SCALE_SHIFT_965	0
 #define		PFIT_HORIZ_SCALE_MASK_965	0x00001fff
 
-#define PFIT_AUTO_RATIOS (dev_priv->info->display_mmio_offset + 0x61238)
+#define PFIT_AUTO_RATIOS (dev_priv->info.display_mmio_offset + 0x61238)
 
-#define _VLV_BLC_PWM_CTL2_A (dev_priv->info->display_mmio_offset + 0x61250)
-#define _VLV_BLC_PWM_CTL2_B (dev_priv->info->display_mmio_offset + 0x61350)
+#define _VLV_BLC_PWM_CTL2_A (dev_priv->info.display_mmio_offset + 0x61250)
+#define _VLV_BLC_PWM_CTL2_B (dev_priv->info.display_mmio_offset + 0x61350)
 #define VLV_BLC_PWM_CTL2(pipe) _PIPE(pipe, _VLV_BLC_PWM_CTL2_A, \
 				     _VLV_BLC_PWM_CTL2_B)
 
-#define _VLV_BLC_PWM_CTL_A (dev_priv->info->display_mmio_offset + 0x61254)
-#define _VLV_BLC_PWM_CTL_B (dev_priv->info->display_mmio_offset + 0x61354)
+#define _VLV_BLC_PWM_CTL_A (dev_priv->info.display_mmio_offset + 0x61254)
+#define _VLV_BLC_PWM_CTL_B (dev_priv->info.display_mmio_offset + 0x61354)
 #define VLV_BLC_PWM_CTL(pipe) _PIPE(pipe, _VLV_BLC_PWM_CTL_A, \
 				    _VLV_BLC_PWM_CTL_B)
 
-#define _VLV_BLC_HIST_CTL_A (dev_priv->info->display_mmio_offset + 0x61260)
-#define _VLV_BLC_HIST_CTL_B (dev_priv->info->display_mmio_offset + 0x61360)
+#define _VLV_BLC_HIST_CTL_A (dev_priv->info.display_mmio_offset + 0x61260)
+#define _VLV_BLC_HIST_CTL_B (dev_priv->info.display_mmio_offset + 0x61360)
 #define VLV_BLC_HIST_CTL(pipe) _PIPE(pipe, _VLV_BLC_HIST_CTL_A, \
 				     _VLV_BLC_HIST_CTL_B)
 
 /* Backlight control */
-#define BLC_PWM_CTL2	(dev_priv->info->display_mmio_offset + 0x61250) /* 965+ only */
+#define BLC_PWM_CTL2	(dev_priv->info.display_mmio_offset + 0x61250) /* 965+ only */
 #define   BLM_PWM_ENABLE		(1 << 31)
 #define   BLM_COMBINATION_MODE		(1 << 30) /* gen4 only */
 #define   BLM_PIPE_SELECT		(1 << 29)
@@ -2448,7 +2868,7 @@
 #define   BLM_PHASE_IN_COUNT_MASK	(0xff << 8)
 #define   BLM_PHASE_IN_INCR_SHIFT	(0)
 #define   BLM_PHASE_IN_INCR_MASK	(0xff << 0)
-#define BLC_PWM_CTL	(dev_priv->info->display_mmio_offset + 0x61254)
+#define BLC_PWM_CTL	(dev_priv->info.display_mmio_offset + 0x61254)
 /*
  * This is the most significant 15 bits of the number of backlight cycles in a
  * complete cycle of the modulated backlight control.
@@ -2470,7 +2890,7 @@
 #define   BACKLIGHT_DUTY_CYCLE_MASK_PNV		(0xfffe)
 #define   BLM_POLARITY_PNV			(1 << 0) /* pnv only */
 
-#define BLC_HIST_CTL	(dev_priv->info->display_mmio_offset + 0x61260)
+#define BLC_HIST_CTL	(dev_priv->info.display_mmio_offset + 0x61260)
 
 /* New registers for PCH-split platforms. Safe where new bits show up, the
  * register layout machtes with gen4 BLC_PWM_CTL[12]. */
@@ -2495,65 +2915,65 @@
 
 /* TV port control */
 #define TV_CTL			0x68000
-/** Enables the TV encoder */
+/* Enables the TV encoder */
 # define TV_ENC_ENABLE			(1 << 31)
-/** Sources the TV encoder input from pipe B instead of A. */
+/* Sources the TV encoder input from pipe B instead of A. */
 # define TV_ENC_PIPEB_SELECT		(1 << 30)
-/** Outputs composite video (DAC A only) */
+/* Outputs composite video (DAC A only) */
 # define TV_ENC_OUTPUT_COMPOSITE	(0 << 28)
-/** Outputs SVideo video (DAC B/C) */
+/* Outputs SVideo video (DAC B/C) */
 # define TV_ENC_OUTPUT_SVIDEO		(1 << 28)
-/** Outputs Component video (DAC A/B/C) */
+/* Outputs Component video (DAC A/B/C) */
 # define TV_ENC_OUTPUT_COMPONENT	(2 << 28)
-/** Outputs Composite and SVideo (DAC A/B/C) */
+/* Outputs Composite and SVideo (DAC A/B/C) */
 # define TV_ENC_OUTPUT_SVIDEO_COMPOSITE	(3 << 28)
 # define TV_TRILEVEL_SYNC		(1 << 21)
-/** Enables slow sync generation (945GM only) */
+/* Enables slow sync generation (945GM only) */
 # define TV_SLOW_SYNC			(1 << 20)
-/** Selects 4x oversampling for 480i and 576p */
+/* Selects 4x oversampling for 480i and 576p */
 # define TV_OVERSAMPLE_4X		(0 << 18)
-/** Selects 2x oversampling for 720p and 1080i */
+/* Selects 2x oversampling for 720p and 1080i */
 # define TV_OVERSAMPLE_2X		(1 << 18)
-/** Selects no oversampling for 1080p */
+/* Selects no oversampling for 1080p */
 # define TV_OVERSAMPLE_NONE		(2 << 18)
-/** Selects 8x oversampling */
+/* Selects 8x oversampling */
 # define TV_OVERSAMPLE_8X		(3 << 18)
-/** Selects progressive mode rather than interlaced */
+/* Selects progressive mode rather than interlaced */
 # define TV_PROGRESSIVE			(1 << 17)
-/** Sets the colorburst to PAL mode.  Required for non-M PAL modes. */
+/* Sets the colorburst to PAL mode.  Required for non-M PAL modes. */
 # define TV_PAL_BURST			(1 << 16)
-/** Field for setting delay of Y compared to C */
+/* Field for setting delay of Y compared to C */
 # define TV_YC_SKEW_MASK		(7 << 12)
-/** Enables a fix for 480p/576p standard definition modes on the 915GM only */
+/* Enables a fix for 480p/576p standard definition modes on the 915GM only */
 # define TV_ENC_SDP_FIX			(1 << 11)
-/**
+/*
  * Enables a fix for the 915GM only.
  *
  * Not sure what it does.
  */
 # define TV_ENC_C0_FIX			(1 << 10)
-/** Bits that must be preserved by software */
+/* Bits that must be preserved by software */
 # define TV_CTL_SAVE			((1 << 11) | (3 << 9) | (7 << 6) | 0xf)
 # define TV_FUSE_STATE_MASK		(3 << 4)
-/** Read-only state that reports all features enabled */
+/* Read-only state that reports all features enabled */
 # define TV_FUSE_STATE_ENABLED		(0 << 4)
-/** Read-only state that reports that Macrovision is disabled in hardware*/
+/* Read-only state that reports that Macrovision is disabled in hardware*/
 # define TV_FUSE_STATE_NO_MACROVISION	(1 << 4)
-/** Read-only state that reports that TV-out is disabled in hardware. */
+/* Read-only state that reports that TV-out is disabled in hardware. */
 # define TV_FUSE_STATE_DISABLED		(2 << 4)
-/** Normal operation */
+/* Normal operation */
 # define TV_TEST_MODE_NORMAL		(0 << 0)
-/** Encoder test pattern 1 - combo pattern */
+/* Encoder test pattern 1 - combo pattern */
 # define TV_TEST_MODE_PATTERN_1		(1 << 0)
-/** Encoder test pattern 2 - full screen vertical 75% color bars */
+/* Encoder test pattern 2 - full screen vertical 75% color bars */
 # define TV_TEST_MODE_PATTERN_2		(2 << 0)
-/** Encoder test pattern 3 - full screen horizontal 75% color bars */
+/* Encoder test pattern 3 - full screen horizontal 75% color bars */
 # define TV_TEST_MODE_PATTERN_3		(3 << 0)
-/** Encoder test pattern 4 - random noise */
+/* Encoder test pattern 4 - random noise */
 # define TV_TEST_MODE_PATTERN_4		(4 << 0)
-/** Encoder test pattern 5 - linear color ramps */
+/* Encoder test pattern 5 - linear color ramps */
 # define TV_TEST_MODE_PATTERN_5		(5 << 0)
-/**
+/*
  * This test mode forces the DACs to 50% of full output.
  *
  * This is used for load detection in combination with TVDAC_SENSE_MASK
@@ -2563,35 +2983,35 @@
 
 #define TV_DAC			0x68004
 # define TV_DAC_SAVE		0x00ffff00
-/**
+/*
  * Reports that DAC state change logic has reported change (RO).
  *
  * This gets cleared when TV_DAC_STATE_EN is cleared
 */
 # define TVDAC_STATE_CHG		(1 << 31)
 # define TVDAC_SENSE_MASK		(7 << 28)
-/** Reports that DAC A voltage is above the detect threshold */
+/* Reports that DAC A voltage is above the detect threshold */
 # define TVDAC_A_SENSE			(1 << 30)
-/** Reports that DAC B voltage is above the detect threshold */
+/* Reports that DAC B voltage is above the detect threshold */
 # define TVDAC_B_SENSE			(1 << 29)
-/** Reports that DAC C voltage is above the detect threshold */
+/* Reports that DAC C voltage is above the detect threshold */
 # define TVDAC_C_SENSE			(1 << 28)
-/**
+/*
  * Enables DAC state detection logic, for load-based TV detection.
  *
  * The PLL of the chosen pipe (in TV_CTL) must be running, and the encoder set
  * to off, for load detection to work.
  */
 # define TVDAC_STATE_CHG_EN		(1 << 27)
-/** Sets the DAC A sense value to high */
+/* Sets the DAC A sense value to high */
 # define TVDAC_A_SENSE_CTL		(1 << 26)
-/** Sets the DAC B sense value to high */
+/* Sets the DAC B sense value to high */
 # define TVDAC_B_SENSE_CTL		(1 << 25)
-/** Sets the DAC C sense value to high */
+/* Sets the DAC C sense value to high */
 # define TVDAC_C_SENSE_CTL		(1 << 24)
-/** Overrides the ENC_ENABLE and DAC voltage levels */
+/* Overrides the ENC_ENABLE and DAC voltage levels */
 # define DAC_CTL_OVERRIDE		(1 << 7)
-/** Sets the slew rate.  Must be preserved in software */
+/* Sets the slew rate.  Must be preserved in software */
 # define ENC_TVDAC_SLEW_FAST		(1 << 6)
 # define DAC_A_1_3_V			(0 << 4)
 # define DAC_A_1_1_V			(1 << 4)
@@ -2606,7 +3026,7 @@
 # define DAC_C_0_7_V			(2 << 0)
 # define DAC_C_MASK			(3 << 0)
 
-/**
+/*
  * CSC coefficients are stored in a floating point format with 9 bits of
  * mantissa and 2 or 3 bits of exponent.  The exponent is represented as 2**-n,
  * where 2-bit exponents are unsigned n, and 3-bit exponents are signed n with
@@ -2621,7 +3041,7 @@
 #define TV_CSC_Y2		0x68014
 # define TV_BY_MASK			0x07ff0000
 # define TV_BY_SHIFT			16
-/**
+/*
  * Y attenuation for component video.
  *
  * Stored in 1.9 fixed point.
@@ -2638,7 +3058,7 @@
 #define TV_CSC_U2		0x6801c
 # define TV_BU_MASK			0x07ff0000
 # define TV_BU_SHIFT			16
-/**
+/*
  * U attenuation for component video.
  *
  * Stored in 1.9 fixed point.
@@ -2655,7 +3075,7 @@
 #define TV_CSC_V2		0x68024
 # define TV_BV_MASK			0x07ff0000
 # define TV_BV_SHIFT			16
-/**
+/*
  * V attenuation for component video.
  *
  * Stored in 1.9 fixed point.
@@ -2664,74 +3084,74 @@
 # define TV_AV_SHIFT			0
 
 #define TV_CLR_KNOBS		0x68028
-/** 2s-complement brightness adjustment */
+/* 2s-complement brightness adjustment */
 # define TV_BRIGHTNESS_MASK		0xff000000
 # define TV_BRIGHTNESS_SHIFT		24
-/** Contrast adjustment, as a 2.6 unsigned floating point number */
+/* Contrast adjustment, as a 2.6 unsigned floating point number */
 # define TV_CONTRAST_MASK		0x00ff0000
 # define TV_CONTRAST_SHIFT		16
-/** Saturation adjustment, as a 2.6 unsigned floating point number */
+/* Saturation adjustment, as a 2.6 unsigned floating point number */
 # define TV_SATURATION_MASK		0x0000ff00
 # define TV_SATURATION_SHIFT		8
-/** Hue adjustment, as an integer phase angle in degrees */
+/* Hue adjustment, as an integer phase angle in degrees */
 # define TV_HUE_MASK			0x000000ff
 # define TV_HUE_SHIFT			0
 
 #define TV_CLR_LEVEL		0x6802c
-/** Controls the DAC level for black */
+/* Controls the DAC level for black */
 # define TV_BLACK_LEVEL_MASK		0x01ff0000
 # define TV_BLACK_LEVEL_SHIFT		16
-/** Controls the DAC level for blanking */
+/* Controls the DAC level for blanking */
 # define TV_BLANK_LEVEL_MASK		0x000001ff
 # define TV_BLANK_LEVEL_SHIFT		0
 
 #define TV_H_CTL_1		0x68030
-/** Number of pixels in the hsync. */
+/* Number of pixels in the hsync. */
 # define TV_HSYNC_END_MASK		0x1fff0000
 # define TV_HSYNC_END_SHIFT		16
-/** Total number of pixels minus one in the line (display and blanking). */
+/* Total number of pixels minus one in the line (display and blanking). */
 # define TV_HTOTAL_MASK			0x00001fff
 # define TV_HTOTAL_SHIFT		0
 
 #define TV_H_CTL_2		0x68034
-/** Enables the colorburst (needed for non-component color) */
+/* Enables the colorburst (needed for non-component color) */
 # define TV_BURST_ENA			(1 << 31)
-/** Offset of the colorburst from the start of hsync, in pixels minus one. */
+/* Offset of the colorburst from the start of hsync, in pixels minus one. */
 # define TV_HBURST_START_SHIFT		16
 # define TV_HBURST_START_MASK		0x1fff0000
-/** Length of the colorburst */
+/* Length of the colorburst */
 # define TV_HBURST_LEN_SHIFT		0
 # define TV_HBURST_LEN_MASK		0x0001fff
 
 #define TV_H_CTL_3		0x68038
-/** End of hblank, measured in pixels minus one from start of hsync */
+/* End of hblank, measured in pixels minus one from start of hsync */
 # define TV_HBLANK_END_SHIFT		16
 # define TV_HBLANK_END_MASK		0x1fff0000
-/** Start of hblank, measured in pixels minus one from start of hsync */
+/* Start of hblank, measured in pixels minus one from start of hsync */
 # define TV_HBLANK_START_SHIFT		0
 # define TV_HBLANK_START_MASK		0x0001fff
 
 #define TV_V_CTL_1		0x6803c
-/** XXX */
+/* XXX */
 # define TV_NBR_END_SHIFT		16
 # define TV_NBR_END_MASK		0x07ff0000
-/** XXX */
+/* XXX */
 # define TV_VI_END_F1_SHIFT		8
 # define TV_VI_END_F1_MASK		0x00003f00
-/** XXX */
+/* XXX */
 # define TV_VI_END_F2_SHIFT		0
 # define TV_VI_END_F2_MASK		0x0000003f
 
 #define TV_V_CTL_2		0x68040
-/** Length of vsync, in half lines */
+/* Length of vsync, in half lines */
 # define TV_VSYNC_LEN_MASK		0x07ff0000
 # define TV_VSYNC_LEN_SHIFT		16
-/** Offset of the start of vsync in field 1, measured in one less than the
+/* Offset of the start of vsync in field 1, measured in one less than the
  * number of half lines.
  */
 # define TV_VSYNC_START_F1_MASK		0x00007f00
 # define TV_VSYNC_START_F1_SHIFT	8
-/**
+/*
  * Offset of the start of vsync in field 2, measured in one less than the
  * number of half lines.
  */
@@ -2739,17 +3159,17 @@
 # define TV_VSYNC_START_F2_SHIFT	0
 
 #define TV_V_CTL_3		0x68044
-/** Enables generation of the equalization signal */
+/* Enables generation of the equalization signal */
 # define TV_EQUAL_ENA			(1 << 31)
-/** Length of vsync, in half lines */
+/* Length of vsync, in half lines */
 # define TV_VEQ_LEN_MASK		0x007f0000
 # define TV_VEQ_LEN_SHIFT		16
-/** Offset of the start of equalization in field 1, measured in one less than
+/* Offset of the start of equalization in field 1, measured in one less than
  * the number of half lines.
  */
 # define TV_VEQ_START_F1_MASK		0x0007f00
 # define TV_VEQ_START_F1_SHIFT		8
-/**
+/*
  * Offset of the start of equalization in field 2, measured in one less than
  * the number of half lines.
  */
@@ -2757,13 +3177,13 @@
 # define TV_VEQ_START_F2_SHIFT		0
 
 #define TV_V_CTL_4		0x68048
-/**
+/*
  * Offset to start of vertical colorburst, measured in one less than the
  * number of lines from vertical start.
  */
 # define TV_VBURST_START_F1_MASK	0x003f0000
 # define TV_VBURST_START_F1_SHIFT	16
-/**
+/*
  * Offset to the end of vertical colorburst, measured in one less than the
  * number of lines from the start of NBR.
  */
@@ -2771,13 +3191,13 @@
 # define TV_VBURST_END_F1_SHIFT		0
 
 #define TV_V_CTL_5		0x6804c
-/**
+/*
  * Offset to start of vertical colorburst, measured in one less than the
  * number of lines from vertical start.
  */
 # define TV_VBURST_START_F2_MASK	0x003f0000
 # define TV_VBURST_START_F2_SHIFT	16
-/**
+/*
  * Offset to the end of vertical colorburst, measured in one less than the
  * number of lines from the start of NBR.
  */
@@ -2785,13 +3205,13 @@
 # define TV_VBURST_END_F2_SHIFT		0
 
 #define TV_V_CTL_6		0x68050
-/**
+/*
  * Offset to start of vertical colorburst, measured in one less than the
  * number of lines from vertical start.
  */
 # define TV_VBURST_START_F3_MASK	0x003f0000
 # define TV_VBURST_START_F3_SHIFT	16
-/**
+/*
  * Offset to the end of vertical colorburst, measured in one less than the
  * number of lines from the start of NBR.
  */
@@ -2799,13 +3219,13 @@
 # define TV_VBURST_END_F3_SHIFT		0
 
 #define TV_V_CTL_7		0x68054
-/**
+/*
  * Offset to start of vertical colorburst, measured in one less than the
  * number of lines from vertical start.
  */
 # define TV_VBURST_START_F4_MASK	0x003f0000
 # define TV_VBURST_START_F4_SHIFT	16
-/**
+/*
  * Offset to the end of vertical colorburst, measured in one less than the
  * number of lines from the start of NBR.
  */
@@ -2813,56 +3233,56 @@
 # define TV_VBURST_END_F4_SHIFT		0
 
 #define TV_SC_CTL_1		0x68060
-/** Turns on the first subcarrier phase generation DDA */
+/* Turns on the first subcarrier phase generation DDA */
 # define TV_SC_DDA1_EN			(1 << 31)
-/** Turns on the first subcarrier phase generation DDA */
+/* Turns on the first subcarrier phase generation DDA */
 # define TV_SC_DDA2_EN			(1 << 30)
-/** Turns on the first subcarrier phase generation DDA */
+/* Turns on the first subcarrier phase generation DDA */
 # define TV_SC_DDA3_EN			(1 << 29)
-/** Sets the subcarrier DDA to reset frequency every other field */
+/* Sets the subcarrier DDA to reset frequency every other field */
 # define TV_SC_RESET_EVERY_2		(0 << 24)
-/** Sets the subcarrier DDA to reset frequency every fourth field */
+/* Sets the subcarrier DDA to reset frequency every fourth field */
 # define TV_SC_RESET_EVERY_4		(1 << 24)
-/** Sets the subcarrier DDA to reset frequency every eighth field */
+/* Sets the subcarrier DDA to reset frequency every eighth field */
 # define TV_SC_RESET_EVERY_8		(2 << 24)
-/** Sets the subcarrier DDA to never reset the frequency */
+/* Sets the subcarrier DDA to never reset the frequency */
 # define TV_SC_RESET_NEVER		(3 << 24)
-/** Sets the peak amplitude of the colorburst.*/
+/* Sets the peak amplitude of the colorburst.*/
 # define TV_BURST_LEVEL_MASK		0x00ff0000
 # define TV_BURST_LEVEL_SHIFT		16
-/** Sets the increment of the first subcarrier phase generation DDA */
+/* Sets the increment of the first subcarrier phase generation DDA */
 # define TV_SCDDA1_INC_MASK		0x00000fff
 # define TV_SCDDA1_INC_SHIFT		0
 
 #define TV_SC_CTL_2		0x68064
-/** Sets the rollover for the second subcarrier phase generation DDA */
+/* Sets the rollover for the second subcarrier phase generation DDA */
 # define TV_SCDDA2_SIZE_MASK		0x7fff0000
 # define TV_SCDDA2_SIZE_SHIFT		16
-/** Sets the increent of the second subcarrier phase generation DDA */
+/* Sets the increent of the second subcarrier phase generation DDA */
 # define TV_SCDDA2_INC_MASK		0x00007fff
 # define TV_SCDDA2_INC_SHIFT		0
 
 #define TV_SC_CTL_3		0x68068
-/** Sets the rollover for the third subcarrier phase generation DDA */
+/* Sets the rollover for the third subcarrier phase generation DDA */
 # define TV_SCDDA3_SIZE_MASK		0x7fff0000
 # define TV_SCDDA3_SIZE_SHIFT		16
-/** Sets the increent of the third subcarrier phase generation DDA */
+/* Sets the increent of the third subcarrier phase generation DDA */
 # define TV_SCDDA3_INC_MASK		0x00007fff
 # define TV_SCDDA3_INC_SHIFT		0
 
 #define TV_WIN_POS		0x68070
-/** X coordinate of the display from the start of horizontal active */
+/* X coordinate of the display from the start of horizontal active */
 # define TV_XPOS_MASK			0x1fff0000
 # define TV_XPOS_SHIFT			16
-/** Y coordinate of the display from the start of vertical active (NBR) */
+/* Y coordinate of the display from the start of vertical active (NBR) */
 # define TV_YPOS_MASK			0x00000fff
 # define TV_YPOS_SHIFT			0
 
 #define TV_WIN_SIZE		0x68074
-/** Horizontal size of the display window, measured in pixels*/
+/* Horizontal size of the display window, measured in pixels*/
 # define TV_XSIZE_MASK			0x1fff0000
 # define TV_XSIZE_SHIFT			16
-/**
+/*
  * Vertical size of the display window, measured in pixels.
  *
  * Must be even for interlaced modes.
@@ -2871,28 +3291,28 @@
 # define TV_YSIZE_SHIFT			0
 
 #define TV_FILTER_CTL_1		0x68080
-/**
+/*
  * Enables automatic scaling calculation.
  *
  * If set, the rest of the registers are ignored, and the calculated values can
  * be read back from the register.
  */
 # define TV_AUTO_SCALE			(1 << 31)
-/**
+/*
  * Disables the vertical filter.
  *
  * This is required on modes more than 1024 pixels wide */
 # define TV_V_FILTER_BYPASS		(1 << 29)
-/** Enables adaptive vertical filtering */
+/* Enables adaptive vertical filtering */
 # define TV_VADAPT			(1 << 28)
 # define TV_VADAPT_MODE_MASK		(3 << 26)
-/** Selects the least adaptive vertical filtering mode */
+/* Selects the least adaptive vertical filtering mode */
 # define TV_VADAPT_MODE_LEAST		(0 << 26)
-/** Selects the moderately adaptive vertical filtering mode */
+/* Selects the moderately adaptive vertical filtering mode */
 # define TV_VADAPT_MODE_MODERATE	(1 << 26)
-/** Selects the most adaptive vertical filtering mode */
+/* Selects the most adaptive vertical filtering mode */
 # define TV_VADAPT_MODE_MOST		(3 << 26)
-/**
+/*
  * Sets the horizontal scaling factor.
  *
  * This should be the fractional part of the horizontal scaling factor divided
@@ -2904,14 +3324,14 @@
 # define TV_HSCALE_FRAC_SHIFT		0
 
 #define TV_FILTER_CTL_2		0x68084
-/**
+/*
  * Sets the integer part of the 3.15 fixed-point vertical scaling factor.
  *
  * TV_VSCALE should be (src height - 1) / ((interlace * dest height) - 1)
  */
 # define TV_VSCALE_INT_MASK		0x00038000
 # define TV_VSCALE_INT_SHIFT		15
-/**
+/*
  * Sets the fractional part of the 3.15 fixed-point vertical scaling factor.
  *
  * \sa TV_VSCALE_INT_MASK
@@ -2920,7 +3340,7 @@
 # define TV_VSCALE_FRAC_SHIFT		0
 
 #define TV_FILTER_CTL_3		0x68088
-/**
+/*
  * Sets the integer part of the 3.15 fixed-point vertical scaling factor.
  *
  * TV_VSCALE should be (src height - 1) / (1/4 * (dest height - 1))
@@ -2929,7 +3349,7 @@
  */
 # define TV_VSCALE_IP_INT_MASK		0x00038000
 # define TV_VSCALE_IP_INT_SHIFT		15
-/**
+/*
  * Sets the fractional part of the 3.15 fixed-point vertical scaling factor.
  *
  * For progressive modes, TV_VSCALE_IP_INT should be set to zeroes.
@@ -2941,26 +3361,26 @@
 
 #define TV_CC_CONTROL		0x68090
 # define TV_CC_ENABLE			(1 << 31)
-/**
+/*
  * Specifies which field to send the CC data in.
  *
  * CC data is usually sent in field 0.
  */
 # define TV_CC_FID_MASK			(1 << 27)
 # define TV_CC_FID_SHIFT		27
-/** Sets the horizontal position of the CC data.  Usually 135. */
+/* Sets the horizontal position of the CC data.  Usually 135. */
 # define TV_CC_HOFF_MASK		0x03ff0000
 # define TV_CC_HOFF_SHIFT		16
-/** Sets the vertical position of the CC data.  Usually 21 */
+/* Sets the vertical position of the CC data.  Usually 21 */
 # define TV_CC_LINE_MASK		0x0000003f
 # define TV_CC_LINE_SHIFT		0
 
 #define TV_CC_DATA		0x68094
 # define TV_CC_RDY			(1 << 31)
-/** Second word of CC data to be transmitted. */
+/* Second word of CC data to be transmitted. */
 # define TV_CC_DATA_2_MASK		0x007f0000
 # define TV_CC_DATA_2_SHIFT		16
-/** First word of CC data to be transmitted. */
+/* First word of CC data to be transmitted. */
 # define TV_CC_DATA_1_MASK		0x0000007f
 # define TV_CC_DATA_1_SHIFT		0
 
@@ -2982,6 +3402,8 @@
 #define   DP_PORT_EN			(1 << 31)
 #define   DP_PIPEB_SELECT		(1 << 30)
 #define   DP_PIPE_MASK			(1 << 30)
+#define   DP_PIPE_SELECT_CHV(pipe)	((pipe) << 16)
+#define   DP_PIPE_MASK_CHV		(3 << 16)
 
 /* Link training mode - select a suitable mode for each stage */
 #define   DP_LINK_TRAIN_PAT_1		(0 << 28)
@@ -3029,32 +3451,32 @@
 #define   DP_PLL_FREQ_160MHZ		(1 << 16)
 #define   DP_PLL_FREQ_MASK		(3 << 16)
 
-/** locked once port is enabled */
+/* locked once port is enabled */
 #define   DP_PORT_REVERSAL		(1 << 15)
 
 /* eDP */
 #define   DP_PLL_ENABLE			(1 << 14)
 
-/** sends the clock on lane 15 of the PEG for debug */
+/* sends the clock on lane 15 of the PEG for debug */
 #define   DP_CLOCK_OUTPUT_ENABLE	(1 << 13)
 
 #define   DP_SCRAMBLING_DISABLE		(1 << 12)
 #define   DP_SCRAMBLING_DISABLE_IRONLAKE	(1 << 7)
 
-/** limit RGB values to avoid confusing TVs */
+/* limit RGB values to avoid confusing TVs */
 #define   DP_COLOR_RANGE_16_235		(1 << 8)
 
-/** Turn on the audio link */
+/* Turn on the audio link */
 #define   DP_AUDIO_OUTPUT_ENABLE	(1 << 6)
 
-/** vs and hs sync polarity */
+/* vs and hs sync polarity */
 #define   DP_SYNC_VS_HIGH		(1 << 4)
 #define   DP_SYNC_HS_HIGH		(1 << 3)
 
-/** A fantasy */
+/* A fantasy */
 #define   DP_DETECTED			(1 << 2)
 
-/** The aux channel provides a way to talk to the
+/* The aux channel provides a way to talk to the
  * signal sink for DDC etc. Max packet size supported
  * is 20 bytes in each direction, hence the 5 fixed
  * data registers
@@ -3164,10 +3586,10 @@
 /* Display & cursor control */
 
 /* Pipe A */
-#define _PIPEADSL		(dev_priv->info->display_mmio_offset + 0x70000)
+#define _PIPEADSL		0x70000
 #define   DSL_LINEMASK_GEN2	0x00000fff
 #define   DSL_LINEMASK_GEN3	0x00001fff
-#define _PIPEACONF		(dev_priv->info->display_mmio_offset + 0x70008)
+#define _PIPEACONF		0x70008
 #define   PIPECONF_ENABLE	(1<<31)
 #define   PIPECONF_DISABLE	0
 #define   PIPECONF_DOUBLE_WIDE	(1<<30)
@@ -3197,6 +3619,7 @@
 #define   PIPECONF_INTERLACED_DBL_ILK		(4 << 21) /* ilk/snb only */
 #define   PIPECONF_PFIT_PF_INTERLACED_DBL_ILK	(5 << 21) /* ilk/snb only */
 #define   PIPECONF_INTERLACE_MODE_MASK		(7 << 21)
+#define   PIPECONF_EDP_RR_MODE_SWITCH		(1 << 20)
 #define   PIPECONF_CXSR_DOWNCLOCK	(1<<16)
 #define   PIPECONF_COLOR_RANGE_SELECT	(1 << 13)
 #define   PIPECONF_BPC_MASK	(0x7 << 5)
@@ -3210,11 +3633,12 @@
 #define   PIPECONF_DITHER_TYPE_ST1 (1<<2)
 #define   PIPECONF_DITHER_TYPE_ST2 (2<<2)
 #define   PIPECONF_DITHER_TYPE_TEMP (3<<2)
-#define _PIPEASTAT		(dev_priv->info->display_mmio_offset + 0x70024)
+#define _PIPEASTAT		0x70024
 #define   PIPE_FIFO_UNDERRUN_STATUS		(1UL<<31)
-#define   SPRITE1_FLIPDONE_INT_EN_VLV		(1UL<<30)
+#define   SPRITE1_FLIP_DONE_INT_EN_VLV		(1UL<<30)
 #define   PIPE_CRC_ERROR_ENABLE			(1UL<<29)
 #define   PIPE_CRC_DONE_ENABLE			(1UL<<28)
+#define   PERF_COUNTER2_INTERRUPT_EN		(1UL<<27)
 #define   PIPE_GMBUS_EVENT_ENABLE		(1UL<<27)
 #define   PLANE_FLIP_DONE_INT_EN_VLV		(1UL<<26)
 #define   PIPE_HOTPLUG_INTERRUPT_ENABLE		(1UL<<26)
@@ -3225,35 +3649,63 @@
 #define   PIPE_LEGACY_BLC_EVENT_ENABLE		(1UL<<22)
 #define   PIPE_ODD_FIELD_INTERRUPT_ENABLE	(1UL<<21)
 #define   PIPE_EVEN_FIELD_INTERRUPT_ENABLE	(1UL<<20)
+#define   PIPE_B_PSR_INTERRUPT_ENABLE_VLV	(1UL<<19)
+#define   PERF_COUNTER_INTERRUPT_EN		(1UL<<19)
 #define   PIPE_HOTPLUG_TV_INTERRUPT_ENABLE	(1UL<<18) /* pre-965 */
 #define   PIPE_START_VBLANK_INTERRUPT_ENABLE	(1UL<<18) /* 965 or later */
+#define   PIPE_FRAMESTART_INTERRUPT_ENABLE	(1UL<<17)
 #define   PIPE_VBLANK_INTERRUPT_ENABLE		(1UL<<17)
 #define   PIPEA_HBLANK_INT_EN_VLV		(1UL<<16)
 #define   PIPE_OVERLAY_UPDATED_ENABLE		(1UL<<16)
-#define   SPRITE1_FLIPDONE_INT_STATUS_VLV	(1UL<<15)
-#define   SPRITE0_FLIPDONE_INT_STATUS_VLV	(1UL<<14)
+#define   SPRITE1_FLIP_DONE_INT_STATUS_VLV	(1UL<<15)
+#define   SPRITE0_FLIP_DONE_INT_STATUS_VLV	(1UL<<14)
 #define   PIPE_CRC_ERROR_INTERRUPT_STATUS	(1UL<<13)
 #define   PIPE_CRC_DONE_INTERRUPT_STATUS	(1UL<<12)
+#define   PERF_COUNTER2_INTERRUPT_STATUS	(1UL<<11)
 #define   PIPE_GMBUS_INTERRUPT_STATUS		(1UL<<11)
-#define   PLANE_FLIPDONE_INT_STATUS_VLV		(1UL<<10)
+#define   PLANE_FLIP_DONE_INT_STATUS_VLV	(1UL<<10)
 #define   PIPE_HOTPLUG_INTERRUPT_STATUS		(1UL<<10)
 #define   PIPE_VSYNC_INTERRUPT_STATUS		(1UL<<9)
 #define   PIPE_DISPLAY_LINE_COMPARE_STATUS	(1UL<<8)
 #define   PIPE_DPST_EVENT_STATUS		(1UL<<7)
 #define   PIPE_LEGACY_BLC_EVENT_STATUS		(1UL<<6)
+#define   PIPE_A_PSR_STATUS_VLV			(1UL<<6)
+#define   PIPE_LEGACY_BLC_EVENT_STATUS		(1UL<<6)
 #define   PIPE_ODD_FIELD_INTERRUPT_STATUS	(1UL<<5)
 #define   PIPE_EVEN_FIELD_INTERRUPT_STATUS	(1UL<<4)
+#define   PIPE_B_PSR_STATUS_VLV			(1UL<<3)
+#define   PERF_COUNTER_INTERRUPT_STATUS		(1UL<<3)
 #define   PIPE_HOTPLUG_TV_INTERRUPT_STATUS	(1UL<<2) /* pre-965 */
 #define   PIPE_START_VBLANK_INTERRUPT_STATUS	(1UL<<2) /* 965 or later */
+#define   PIPE_FRAMESTART_INTERRUPT_STATUS	(1UL<<1)
 #define   PIPE_VBLANK_INTERRUPT_STATUS		(1UL<<1)
+#define   PIPE_HBLANK_INT_STATUS		(1UL<<0)
 #define   PIPE_OVERLAY_UPDATED_STATUS		(1UL<<0)
 
-#define PIPESRC(pipe) _PIPE(pipe, _PIPEASRC, _PIPEBSRC)
-#define PIPECONF(tran) _TRANSCODER(tran, _PIPEACONF, _PIPEBCONF)
-#define PIPEDSL(pipe)  _PIPE(pipe, _PIPEADSL, _PIPEBDSL)
-#define PIPEFRAME(pipe) _PIPE(pipe, _PIPEAFRAMEHIGH, _PIPEBFRAMEHIGH)
-#define PIPEFRAMEPIXEL(pipe)  _PIPE(pipe, _PIPEAFRAMEPIXEL, _PIPEBFRAMEPIXEL)
-#define PIPESTAT(pipe) _PIPE(pipe, _PIPEASTAT, _PIPEBSTAT)
+#define PIPESTAT_INT_ENABLE_MASK		0x7fff0000
+#define PIPESTAT_INT_STATUS_MASK		0x0000ffff
+
+#define PIPE_A_OFFSET		0x70000
+#define PIPE_B_OFFSET		0x71000
+#define PIPE_C_OFFSET		0x72000
+#define CHV_PIPE_C_OFFSET	0x74000
+/*
+ * There's actually no pipe EDP. Some pipe registers have
+ * simply shifted from the pipe to the transcoder, while
+ * keeping their original offset. Thus we need PIPE_EDP_OFFSET
+ * to access such registers in transcoder EDP.
+ */
+#define PIPE_EDP_OFFSET	0x7f000
+
+#define _PIPE2(pipe, reg) (dev_priv->info.pipe_offsets[pipe] - \
+	dev_priv->info.pipe_offsets[PIPE_A] + (reg) + \
+	dev_priv->info.display_mmio_offset)
+
+#define PIPECONF(pipe) _PIPE2(pipe, _PIPEACONF)
+#define PIPEDSL(pipe)  _PIPE2(pipe, _PIPEADSL)
+#define PIPEFRAME(pipe) _PIPE2(pipe, _PIPEAFRAMEHIGH)
+#define PIPEFRAMEPIXEL(pipe)  _PIPE2(pipe, _PIPEAFRAMEPIXEL)
+#define PIPESTAT(pipe) _PIPE2(pipe, _PIPEASTAT)
 
 #define _PIPE_MISC_A			0x70030
 #define _PIPE_MISC_B			0x71030
@@ -3265,23 +3717,34 @@
 #define   PIPEMISC_DITHER_ENABLE	(1<<4)
 #define   PIPEMISC_DITHER_TYPE_MASK	(3<<2)
 #define   PIPEMISC_DITHER_TYPE_SP	(0<<2)
-#define PIPEMISC(pipe) _PIPE(pipe, _PIPE_MISC_A, _PIPE_MISC_B)
+#define PIPEMISC(pipe) _PIPE2(pipe, _PIPE_MISC_A)
 
 #define VLV_DPFLIPSTAT				(VLV_DISPLAY_BASE + 0x70028)
 #define   PIPEB_LINE_COMPARE_INT_EN		(1<<29)
 #define   PIPEB_HLINE_INT_EN			(1<<28)
 #define   PIPEB_VBLANK_INT_EN			(1<<27)
-#define   SPRITED_FLIPDONE_INT_EN		(1<<26)
-#define   SPRITEC_FLIPDONE_INT_EN		(1<<25)
-#define   PLANEB_FLIPDONE_INT_EN		(1<<24)
+#define   SPRITED_FLIP_DONE_INT_EN		(1<<26)
+#define   SPRITEC_FLIP_DONE_INT_EN		(1<<25)
+#define   PLANEB_FLIP_DONE_INT_EN		(1<<24)
+#define   PIPE_PSR_INT_EN			(1<<22)
 #define   PIPEA_LINE_COMPARE_INT_EN		(1<<21)
 #define   PIPEA_HLINE_INT_EN			(1<<20)
 #define   PIPEA_VBLANK_INT_EN			(1<<19)
-#define   SPRITEB_FLIPDONE_INT_EN		(1<<18)
-#define   SPRITEA_FLIPDONE_INT_EN		(1<<17)
+#define   SPRITEB_FLIP_DONE_INT_EN		(1<<18)
+#define   SPRITEA_FLIP_DONE_INT_EN		(1<<17)
 #define   PLANEA_FLIPDONE_INT_EN		(1<<16)
-
-#define DPINVGTT				(VLV_DISPLAY_BASE + 0x7002c) /* VLV only */
+#define   PIPEC_LINE_COMPARE_INT_EN		(1<<13)
+#define   PIPEC_HLINE_INT_EN			(1<<12)
+#define   PIPEC_VBLANK_INT_EN			(1<<11)
+#define   SPRITEF_FLIPDONE_INT_EN		(1<<10)
+#define   SPRITEE_FLIPDONE_INT_EN		(1<<9)
+#define   PLANEC_FLIPDONE_INT_EN		(1<<8)
+
+#define DPINVGTT				(VLV_DISPLAY_BASE + 0x7002c) /* VLV/CHV only */
+#define   SPRITEF_INVALID_GTT_INT_EN		(1<<27)
+#define   SPRITEE_INVALID_GTT_INT_EN		(1<<26)
+#define   PLANEC_INVALID_GTT_INT_EN		(1<<25)
+#define   CURSORC_INVALID_GTT_INT_EN		(1<<24)
 #define   CURSORB_INVALID_GTT_INT_EN		(1<<23)
 #define   CURSORA_INVALID_GTT_INT_EN		(1<<22)
 #define   SPRITED_INVALID_GTT_INT_EN		(1<<21)
@@ -3291,6 +3754,11 @@
 #define   SPRITEA_INVALID_GTT_INT_EN		(1<<17)
 #define   PLANEA_INVALID_GTT_INT_EN		(1<<16)
 #define   DPINVGTT_EN_MASK			0xff0000
+#define   DPINVGTT_EN_MASK_CHV			0xfff0000
+#define   SPRITEF_INVALID_GTT_STATUS		(1<<11)
+#define   SPRITEE_INVALID_GTT_STATUS		(1<<10)
+#define   PLANEC_INVALID_GTT_STATUS		(1<<9)
+#define   CURSORC_INVALID_GTT_STATUS		(1<<8)
 #define   CURSORB_INVALID_GTT_STATUS		(1<<7)
 #define   CURSORA_INVALID_GTT_STATUS		(1<<6)
 #define   SPRITED_INVALID_GTT_STATUS		(1<<5)
@@ -3300,6 +3768,7 @@
 #define   SPRITEA_INVALID_GTT_STATUS		(1<<1)
 #define   PLANEA_INVALID_GTT_STATUS		(1<<0)
 #define   DPINVGTT_STATUS_MASK			0xff
+#define   DPINVGTT_STATUS_MASK_CHV		0xfff
 
 #define DSPARB			0x70030
 #define   DSPARB_CSTART_MASK	(0x7f << 7)
@@ -3309,7 +3778,7 @@
 #define   DSPARB_BEND_SHIFT	9 /* on 855 */
 #define   DSPARB_AEND_SHIFT	0
 
-#define DSPFW1			(dev_priv->info->display_mmio_offset + 0x70034)
+#define DSPFW1			(dev_priv->info.display_mmio_offset + 0x70034)
 #define   DSPFW_SR_SHIFT	23
 #define   DSPFW_SR_MASK		(0x1ff<<23)
 #define   DSPFW_CURSORB_SHIFT	16
@@ -3317,11 +3786,11 @@
 #define   DSPFW_PLANEB_SHIFT	8
 #define   DSPFW_PLANEB_MASK	(0x7f<<8)
 #define   DSPFW_PLANEA_MASK	(0x7f)
-#define DSPFW2			(dev_priv->info->display_mmio_offset + 0x70038)
+#define DSPFW2			(dev_priv->info.display_mmio_offset + 0x70038)
 #define   DSPFW_CURSORA_MASK	0x00003f00
 #define   DSPFW_CURSORA_SHIFT	8
 #define   DSPFW_PLANEC_MASK	(0x7f)
-#define DSPFW3			(dev_priv->info->display_mmio_offset + 0x7003c)
+#define DSPFW3			(dev_priv->info.display_mmio_offset + 0x7003c)
 #define   DSPFW_HPLL_SR_EN	(1<<31)
 #define   DSPFW_CURSOR_SR_SHIFT	24
 #define   PINEVIEW_SELF_REFRESH_EN	(1<<30)
@@ -3329,8 +3798,8 @@
 #define   DSPFW_HPLL_CURSOR_SHIFT	16
 #define   DSPFW_HPLL_CURSOR_MASK	(0x3f<<16)
 #define   DSPFW_HPLL_SR_MASK		(0x1ff)
-#define DSPFW4			(dev_priv->info->display_mmio_offset + 0x70070)
-#define DSPFW7			(dev_priv->info->display_mmio_offset + 0x7007c)
+#define DSPFW4			(dev_priv->info.display_mmio_offset + 0x70070)
+#define DSPFW7			(dev_priv->info.display_mmio_offset + 0x7007c)
 
 /* drain latency register values*/
 #define DRAIN_LATENCY_PRECISION_32	32
@@ -3339,14 +3808,43 @@
 #define DDL_CURSORA_PRECISION_32	(1<<31)
 #define DDL_CURSORA_PRECISION_16	(0<<31)
 #define DDL_CURSORA_SHIFT		24
+#define DDL_SPRITEB_PRECISION_32	(1<<23)
+#define DDL_SPRITEB_PRECISION_16	(0<<23)
+#define DDL_SPRITEB_SHIFT		16
+#define DDL_SPRITEA_PRECISION_32	(1<<15)
+#define DDL_SPRITEA_PRECISION_16	(0<<15)
+#define DDL_SPRITEA_SHIFT		8
 #define DDL_PLANEA_PRECISION_32		(1<<7)
 #define DDL_PLANEA_PRECISION_16		(0<<7)
+#define DDL_PLANEA_SHIFT		0
+
 #define VLV_DDL2			(VLV_DISPLAY_BASE + 0x70054)
 #define DDL_CURSORB_PRECISION_32	(1<<31)
 #define DDL_CURSORB_PRECISION_16	(0<<31)
 #define DDL_CURSORB_SHIFT		24
+#define DDL_SPRITED_PRECISION_32	(1<<23)
+#define DDL_SPRITED_PRECISION_16	(0<<23)
+#define DDL_SPRITED_SHIFT		16
+#define DDL_SPRITEC_PRECISION_32	(1<<15)
+#define DDL_SPRITEC_PRECISION_16	(0<<15)
+#define DDL_SPRITEC_SHIFT		8
 #define DDL_PLANEB_PRECISION_32		(1<<7)
 #define DDL_PLANEB_PRECISION_16		(0<<7)
+#define DDL_PLANEB_SHIFT		0
+
+#define VLV_DDL3			(VLV_DISPLAY_BASE + 0x70058)
+#define DDL_CURSORC_PRECISION_32	(1<<31)
+#define DDL_CURSORC_PRECISION_16	(0<<31)
+#define DDL_CURSORC_SHIFT		24
+#define DDL_SPRITEF_PRECISION_32	(1<<23)
+#define DDL_SPRITEF_PRECISION_16	(0<<23)
+#define DDL_SPRITEF_SHIFT		16
+#define DDL_SPRITEE_PRECISION_32	(1<<15)
+#define DDL_SPRITEE_PRECISION_16	(0<<15)
+#define DDL_SPRITEE_SHIFT		8
+#define DDL_PLANEC_PRECISION_32		(1<<7)
+#define DDL_PLANEC_PRECISION_16		(0<<7)
+#define DDL_PLANEC_SHIFT		0
 
 /* FIFO watermark sizes etc */
 #define G4X_FIFO_LINE_SIZE	64
@@ -3421,42 +3919,6 @@
 /* the unit of memory self-refresh latency time is 0.5us */
 #define  ILK_SRLT_MASK		0x3f
 
-/* define the fifo size on Ironlake */
-#define ILK_DISPLAY_FIFO	128
-#define ILK_DISPLAY_MAXWM	64
-#define ILK_DISPLAY_DFTWM	8
-#define ILK_CURSOR_FIFO		32
-#define ILK_CURSOR_MAXWM	16
-#define ILK_CURSOR_DFTWM	8
-
-#define ILK_DISPLAY_SR_FIFO	512
-#define ILK_DISPLAY_MAX_SRWM	0x1ff
-#define ILK_DISPLAY_DFT_SRWM	0x3f
-#define ILK_CURSOR_SR_FIFO	64
-#define ILK_CURSOR_MAX_SRWM	0x3f
-#define ILK_CURSOR_DFT_SRWM	8
-
-#define ILK_FIFO_LINE_SIZE	64
-
-/* define the WM info on Sandybridge */
-#define SNB_DISPLAY_FIFO	128
-#define SNB_DISPLAY_MAXWM	0x7f	/* bit 16:22 */
-#define SNB_DISPLAY_DFTWM	8
-#define SNB_CURSOR_FIFO		32
-#define SNB_CURSOR_MAXWM	0x1f	/* bit 4:0 */
-#define SNB_CURSOR_DFTWM	8
-
-#define SNB_DISPLAY_SR_FIFO	512
-#define SNB_DISPLAY_MAX_SRWM	0x1ff	/* bit 16:8 */
-#define SNB_DISPLAY_DFT_SRWM	0x3f
-#define SNB_CURSOR_SR_FIFO	64
-#define SNB_CURSOR_MAX_SRWM	0x3f	/* bit 5:0 */
-#define SNB_CURSOR_DFT_SRWM	8
-
-#define SNB_FBC_MAX_SRWM	0xf	/* bit 23:20 */
-
-#define SNB_FIFO_LINE_SIZE	64
-
 
 /* the address where we get all kinds of latency value */
 #define SSKPD			0x5d10
@@ -3490,12 +3952,13 @@
 #define   PIPE_PIXEL_MASK         0x00ffffff
 #define   PIPE_PIXEL_SHIFT        0
 /* GM45+ just has to be different */
-#define _PIPEA_FRMCOUNT_GM45	(dev_priv->info->display_mmio_offset + 0x70040)
-#define _PIPEA_FLIPCOUNT_GM45	(dev_priv->info->display_mmio_offset + 0x70044)
-#define PIPE_FRMCOUNT_GM45(pipe) _PIPE(pipe, _PIPEA_FRMCOUNT_GM45, _PIPEB_FRMCOUNT_GM45)
+#define _PIPEA_FRMCOUNT_GM45	0x70040
+#define _PIPEA_FLIPCOUNT_GM45	0x70044
+#define PIPE_FRMCOUNT_GM45(pipe) _PIPE2(pipe, _PIPEA_FRMCOUNT_GM45)
+#define PIPE_FLIPCOUNT_GM45(pipe) _PIPE2(pipe, _PIPEA_FLIPCOUNT_GM45)
 
 /* Cursor A & B regs */
-#define _CURACNTR		(dev_priv->info->display_mmio_offset + 0x70080)
+#define _CURACNTR		0x70080
 /* Old style CUR*CNTR flags (desktop 8xx) */
 #define   CURSOR_ENABLE		0x80000000
 #define   CURSOR_GAMMA_ENABLE	0x40000000
@@ -3511,38 +3974,48 @@
 /* New style CUR*CNTR flags */
 #define   CURSOR_MODE		0x27
 #define   CURSOR_MODE_DISABLE   0x00
+#define   CURSOR_MODE_128_32B_AX 0x02
+#define   CURSOR_MODE_256_32B_AX 0x03
 #define   CURSOR_MODE_64_32B_AX 0x07
+#define   CURSOR_MODE_128_ARGB_AX ((1 << 5) | CURSOR_MODE_128_32B_AX)
+#define   CURSOR_MODE_256_ARGB_AX ((1 << 5) | CURSOR_MODE_256_32B_AX)
 #define   CURSOR_MODE_64_ARGB_AX ((1 << 5) | CURSOR_MODE_64_32B_AX)
 #define   MCURSOR_PIPE_SELECT	(1 << 28)
 #define   MCURSOR_PIPE_A	0x00
 #define   MCURSOR_PIPE_B	(1 << 28)
 #define   MCURSOR_GAMMA_ENABLE  (1 << 26)
 #define   CURSOR_TRICKLE_FEED_DISABLE	(1 << 14)
-#define _CURABASE		(dev_priv->info->display_mmio_offset + 0x70084)
-#define _CURAPOS		(dev_priv->info->display_mmio_offset + 0x70088)
+#define _CURABASE		0x70084
+#define _CURAPOS		0x70088
 #define   CURSOR_POS_MASK       0x007FF
 #define   CURSOR_POS_SIGN       0x8000
 #define   CURSOR_X_SHIFT        0
 #define   CURSOR_Y_SHIFT        16
 #define CURSIZE			0x700a0
-#define _CURBCNTR		(dev_priv->info->display_mmio_offset + 0x700c0)
-#define _CURBBASE		(dev_priv->info->display_mmio_offset + 0x700c4)
-#define _CURBPOS		(dev_priv->info->display_mmio_offset + 0x700c8)
+#define _CURBCNTR		0x700c0
+#define _CURBBASE		0x700c4
+#define _CURBPOS		0x700c8
 
 #define _CURBCNTR_IVB		0x71080
 #define _CURBBASE_IVB		0x71084
 #define _CURBPOS_IVB		0x71088
 
-#define CURCNTR(pipe) _PIPE(pipe, _CURACNTR, _CURBCNTR)
-#define CURBASE(pipe) _PIPE(pipe, _CURABASE, _CURBBASE)
-#define CURPOS(pipe) _PIPE(pipe, _CURAPOS, _CURBPOS)
+#define _CURSOR2(pipe, reg) (dev_priv->info.cursor_offsets[(pipe)] - \
+	dev_priv->info.cursor_offsets[PIPE_A] + (reg) + \
+	dev_priv->info.display_mmio_offset)
+
+#define CURCNTR(pipe) _CURSOR2(pipe, _CURACNTR)
+#define CURBASE(pipe) _CURSOR2(pipe, _CURABASE)
+#define CURPOS(pipe) _CURSOR2(pipe, _CURAPOS)
 
-#define CURCNTR_IVB(pipe) _PIPE(pipe, _CURACNTR, _CURBCNTR_IVB)
-#define CURBASE_IVB(pipe) _PIPE(pipe, _CURABASE, _CURBBASE_IVB)
-#define CURPOS_IVB(pipe) _PIPE(pipe, _CURAPOS, _CURBPOS_IVB)
+#define CURSOR_A_OFFSET 0x70080
+#define CURSOR_B_OFFSET 0x700c0
+#define CHV_CURSOR_C_OFFSET 0x700e0
+#define IVB_CURSOR_B_OFFSET 0x71080
+#define IVB_CURSOR_C_OFFSET 0x72080
 
 /* Display A control */
-#define _DSPACNTR                (dev_priv->info->display_mmio_offset + 0x70180)
+#define _DSPACNTR				0x70180
 #define   DISPLAY_PLANE_ENABLE			(1<<31)
 #define   DISPLAY_PLANE_DISABLE			0
 #define   DISPPLANE_GAMMA_ENABLE		(1<<30)
@@ -3576,72 +4049,70 @@
 #define   DISPPLANE_STEREO_POLARITY_SECOND	(1<<18)
 #define   DISPPLANE_TRICKLE_FEED_DISABLE	(1<<14) /* Ironlake */
 #define   DISPPLANE_TILED			(1<<10)
-#define _DSPAADDR		(dev_priv->info->display_mmio_offset + 0x70184)
-#define _DSPASTRIDE		(dev_priv->info->display_mmio_offset + 0x70188)
-#define _DSPAPOS		(dev_priv->info->display_mmio_offset + 0x7018C) /* reserved */
-#define _DSPASIZE		(dev_priv->info->display_mmio_offset + 0x70190)
-#define _DSPASURF		(dev_priv->info->display_mmio_offset + 0x7019C) /* 965+ only */
-#define _DSPATILEOFF		(dev_priv->info->display_mmio_offset + 0x701A4) /* 965+ only */
-#define _DSPAOFFSET		(dev_priv->info->display_mmio_offset + 0x701A4) /* HSW */
-#define _DSPASURFLIVE		(dev_priv->info->display_mmio_offset + 0x701AC)
-
-#define DSPCNTR(plane) _PIPE(plane, _DSPACNTR, _DSPBCNTR)
-#define DSPADDR(plane) _PIPE(plane, _DSPAADDR, _DSPBADDR)
-#define DSPSTRIDE(plane) _PIPE(plane, _DSPASTRIDE, _DSPBSTRIDE)
-#define DSPPOS(plane) _PIPE(plane, _DSPAPOS, _DSPBPOS)
-#define DSPSIZE(plane) _PIPE(plane, _DSPASIZE, _DSPBSIZE)
-#define DSPSURF(plane) _PIPE(plane, _DSPASURF, _DSPBSURF)
-#define DSPTILEOFF(plane) _PIPE(plane, _DSPATILEOFF, _DSPBTILEOFF)
+#define _DSPAADDR				0x70184
+#define _DSPASTRIDE				0x70188
+#define _DSPAPOS				0x7018C /* reserved */
+#define _DSPASIZE				0x70190
+#define _DSPASURF				0x7019C /* 965+ only */
+#define _DSPATILEOFF				0x701A4 /* 965+ only */
+#define _DSPAOFFSET				0x701A4 /* HSW */
+#define _DSPASURFLIVE				0x701AC
+
+#define DSPCNTR(plane) _PIPE2(plane, _DSPACNTR)
+#define DSPADDR(plane) _PIPE2(plane, _DSPAADDR)
+#define DSPSTRIDE(plane) _PIPE2(plane, _DSPASTRIDE)
+#define DSPPOS(plane) _PIPE2(plane, _DSPAPOS)
+#define DSPSIZE(plane) _PIPE2(plane, _DSPASIZE)
+#define DSPSURF(plane) _PIPE2(plane, _DSPASURF)
+#define DSPTILEOFF(plane) _PIPE2(plane, _DSPATILEOFF)
 #define DSPLINOFF(plane) DSPADDR(plane)
-#define DSPOFFSET(plane) _PIPE(plane, _DSPAOFFSET, _DSPBOFFSET)
-#define DSPSURFLIVE(plane) _PIPE(plane, _DSPASURFLIVE, _DSPBSURFLIVE)
+#define DSPOFFSET(plane) _PIPE2(plane, _DSPAOFFSET)
+#define DSPSURFLIVE(plane) _PIPE2(plane, _DSPASURFLIVE)
 
 /* Display/Sprite base address macros */
 #define DISP_BASEADDR_MASK	(0xfffff000)
 #define I915_LO_DISPBASE(val)	(val & ~DISP_BASEADDR_MASK)
 #define I915_HI_DISPBASE(val)	(val & DISP_BASEADDR_MASK)
-#define I915_MODIFY_DISPBASE(reg, gfx_addr) \
-		(I915_WRITE((reg), (gfx_addr) | I915_LO_DISPBASE(I915_READ(reg))))
 
 /* VBIOS flags */
-#define SWF00			(dev_priv->info->display_mmio_offset + 0x71410)
-#define SWF01			(dev_priv->info->display_mmio_offset + 0x71414)
-#define SWF02			(dev_priv->info->display_mmio_offset + 0x71418)
-#define SWF03			(dev_priv->info->display_mmio_offset + 0x7141c)
-#define SWF04			(dev_priv->info->display_mmio_offset + 0x71420)
-#define SWF05			(dev_priv->info->display_mmio_offset + 0x71424)
-#define SWF06			(dev_priv->info->display_mmio_offset + 0x71428)
-#define SWF10			(dev_priv->info->display_mmio_offset + 0x70410)
-#define SWF11			(dev_priv->info->display_mmio_offset + 0x70414)
-#define SWF14			(dev_priv->info->display_mmio_offset + 0x71420)
-#define SWF30			(dev_priv->info->display_mmio_offset + 0x72414)
-#define SWF31			(dev_priv->info->display_mmio_offset + 0x72418)
-#define SWF32			(dev_priv->info->display_mmio_offset + 0x7241c)
+#define SWF00			(dev_priv->info.display_mmio_offset + 0x71410)
+#define SWF01			(dev_priv->info.display_mmio_offset + 0x71414)
+#define SWF02			(dev_priv->info.display_mmio_offset + 0x71418)
+#define SWF03			(dev_priv->info.display_mmio_offset + 0x7141c)
+#define SWF04			(dev_priv->info.display_mmio_offset + 0x71420)
+#define SWF05			(dev_priv->info.display_mmio_offset + 0x71424)
+#define SWF06			(dev_priv->info.display_mmio_offset + 0x71428)
+#define SWF10			(dev_priv->info.display_mmio_offset + 0x70410)
+#define SWF11			(dev_priv->info.display_mmio_offset + 0x70414)
+#define SWF14			(dev_priv->info.display_mmio_offset + 0x71420)
+#define SWF30			(dev_priv->info.display_mmio_offset + 0x72414)
+#define SWF31			(dev_priv->info.display_mmio_offset + 0x72418)
+#define SWF32			(dev_priv->info.display_mmio_offset + 0x7241c)
 
 /* Pipe B */
-#define _PIPEBDSL		(dev_priv->info->display_mmio_offset + 0x71000)
-#define _PIPEBCONF		(dev_priv->info->display_mmio_offset + 0x71008)
-#define _PIPEBSTAT		(dev_priv->info->display_mmio_offset + 0x71024)
+#define _PIPEBDSL		(dev_priv->info.display_mmio_offset + 0x71000)
+#define _PIPEBCONF		(dev_priv->info.display_mmio_offset + 0x71008)
+#define _PIPEBSTAT		(dev_priv->info.display_mmio_offset + 0x71024)
 #define _PIPEBFRAMEHIGH		0x71040
 #define _PIPEBFRAMEPIXEL	0x71044
-#define _PIPEB_FRMCOUNT_GM45	(dev_priv->info->display_mmio_offset + 0x71040)
-#define _PIPEB_FLIPCOUNT_GM45	(dev_priv->info->display_mmio_offset + 0x71044)
+#define _PIPEB_FRMCOUNT_GM45	(dev_priv->info.display_mmio_offset + 0x71040)
+#define _PIPEB_FLIPCOUNT_GM45	(dev_priv->info.display_mmio_offset + 0x71044)
 
 
 /* Display B control */
-#define _DSPBCNTR		(dev_priv->info->display_mmio_offset + 0x71180)
+#define _DSPBCNTR		(dev_priv->info.display_mmio_offset + 0x71180)
 #define   DISPPLANE_ALPHA_TRANS_ENABLE		(1<<15)
 #define   DISPPLANE_ALPHA_TRANS_DISABLE		0
 #define   DISPPLANE_SPRITE_ABOVE_DISPLAY	0
 #define   DISPPLANE_SPRITE_ABOVE_OVERLAY	(1)
-#define _DSPBADDR		(dev_priv->info->display_mmio_offset + 0x71184)
-#define _DSPBSTRIDE		(dev_priv->info->display_mmio_offset + 0x71188)
-#define _DSPBPOS		(dev_priv->info->display_mmio_offset + 0x7118C)
-#define _DSPBSIZE		(dev_priv->info->display_mmio_offset + 0x71190)
-#define _DSPBSURF		(dev_priv->info->display_mmio_offset + 0x7119C)
-#define _DSPBTILEOFF		(dev_priv->info->display_mmio_offset + 0x711A4)
-#define _DSPBOFFSET		(dev_priv->info->display_mmio_offset + 0x711A4)
-#define _DSPBSURFLIVE		(dev_priv->info->display_mmio_offset + 0x711AC)
+#define _DSPBADDR		(dev_priv->info.display_mmio_offset + 0x71184)
+#define _DSPBSTRIDE		(dev_priv->info.display_mmio_offset + 0x71188)
+#define _DSPBPOS		(dev_priv->info.display_mmio_offset + 0x7118C)
+#define _DSPBSIZE		(dev_priv->info.display_mmio_offset + 0x71190)
+#define _DSPBSURF		(dev_priv->info.display_mmio_offset + 0x7119C)
+#define _DSPBTILEOFF		(dev_priv->info.display_mmio_offset + 0x711A4)
+#define _DSPBOFFSET		(dev_priv->info.display_mmio_offset + 0x711A4)
+#define _DSPBSURFLIVE		(dev_priv->info.display_mmio_offset + 0x711AC)
 
 /* Sprite A control */
 #define _DVSACNTR		0x72180
@@ -3787,7 +4258,7 @@
 
 #define _SPACNTR		(VLV_DISPLAY_BASE + 0x72180)
 #define   SP_ENABLE			(1<<31)
-#define   SP_GEAMMA_ENABLE		(1<<30)
+#define   SP_GAMMA_ENABLE		(1<<30)
 #define   SP_PIXFORMAT_MASK		(0xf<<26)
 #define   SP_FORMAT_YUV422		(0<<26)
 #define   SP_FORMAT_BGR565		(5<<26)
@@ -3890,48 +4361,45 @@
 #define  FDI_PLL_FREQ_DISABLE_COUNT_LIMIT_MASK  0xff
 
 
-#define _PIPEA_DATA_M1           (dev_priv->info->display_mmio_offset + 0x60030)
+#define _PIPEA_DATA_M1		0x60030
 #define  PIPE_DATA_M1_OFFSET    0
-#define _PIPEA_DATA_N1           (dev_priv->info->display_mmio_offset + 0x60034)
+#define _PIPEA_DATA_N1		0x60034
 #define  PIPE_DATA_N1_OFFSET    0
 
-#define _PIPEA_DATA_M2           (dev_priv->info->display_mmio_offset + 0x60038)
+#define _PIPEA_DATA_M2		0x60038
 #define  PIPE_DATA_M2_OFFSET    0
-#define _PIPEA_DATA_N2           (dev_priv->info->display_mmio_offset + 0x6003c)
+#define _PIPEA_DATA_N2		0x6003c
 #define  PIPE_DATA_N2_OFFSET    0
 
-#define _PIPEA_LINK_M1           (dev_priv->info->display_mmio_offset + 0x60040)
+#define _PIPEA_LINK_M1		0x60040
 #define  PIPE_LINK_M1_OFFSET    0
-#define _PIPEA_LINK_N1           (dev_priv->info->display_mmio_offset + 0x60044)
+#define _PIPEA_LINK_N1		0x60044
 #define  PIPE_LINK_N1_OFFSET    0
 
-#define _PIPEA_LINK_M2           (dev_priv->info->display_mmio_offset + 0x60048)
+#define _PIPEA_LINK_M2		0x60048
 #define  PIPE_LINK_M2_OFFSET    0
-#define _PIPEA_LINK_N2           (dev_priv->info->display_mmio_offset + 0x6004c)
+#define _PIPEA_LINK_N2		0x6004c
 #define  PIPE_LINK_N2_OFFSET    0
 
 /* PIPEB timing regs are same start from 0x61000 */
 
-#define _PIPEB_DATA_M1           (dev_priv->info->display_mmio_offset + 0x61030)
-#define _PIPEB_DATA_N1           (dev_priv->info->display_mmio_offset + 0x61034)
-
-#define _PIPEB_DATA_M2           (dev_priv->info->display_mmio_offset + 0x61038)
-#define _PIPEB_DATA_N2           (dev_priv->info->display_mmio_offset + 0x6103c)
-
-#define _PIPEB_LINK_M1           (dev_priv->info->display_mmio_offset + 0x61040)
-#define _PIPEB_LINK_N1           (dev_priv->info->display_mmio_offset + 0x61044)
-
-#define _PIPEB_LINK_M2           (dev_priv->info->display_mmio_offset + 0x61048)
-#define _PIPEB_LINK_N2           (dev_priv->info->display_mmio_offset + 0x6104c)
-
-#define PIPE_DATA_M1(tran) _TRANSCODER(tran, _PIPEA_DATA_M1, _PIPEB_DATA_M1)
-#define PIPE_DATA_N1(tran) _TRANSCODER(tran, _PIPEA_DATA_N1, _PIPEB_DATA_N1)
-#define PIPE_DATA_M2(tran) _TRANSCODER(tran, _PIPEA_DATA_M2, _PIPEB_DATA_M2)
-#define PIPE_DATA_N2(tran) _TRANSCODER(tran, _PIPEA_DATA_N2, _PIPEB_DATA_N2)
-#define PIPE_LINK_M1(tran) _TRANSCODER(tran, _PIPEA_LINK_M1, _PIPEB_LINK_M1)
-#define PIPE_LINK_N1(tran) _TRANSCODER(tran, _PIPEA_LINK_N1, _PIPEB_LINK_N1)
-#define PIPE_LINK_M2(tran) _TRANSCODER(tran, _PIPEA_LINK_M2, _PIPEB_LINK_M2)
-#define PIPE_LINK_N2(tran) _TRANSCODER(tran, _PIPEA_LINK_N2, _PIPEB_LINK_N2)
+#define _PIPEB_DATA_M1		0x61030
+#define _PIPEB_DATA_N1		0x61034
+#define _PIPEB_DATA_M2		0x61038
+#define _PIPEB_DATA_N2		0x6103c
+#define _PIPEB_LINK_M1		0x61040
+#define _PIPEB_LINK_N1		0x61044
+#define _PIPEB_LINK_M2		0x61048
+#define _PIPEB_LINK_N2		0x6104c
+
+#define PIPE_DATA_M1(tran) _TRANSCODER2(tran, _PIPEA_DATA_M1)
+#define PIPE_DATA_N1(tran) _TRANSCODER2(tran, _PIPEA_DATA_N1)
+#define PIPE_DATA_M2(tran) _TRANSCODER2(tran, _PIPEA_DATA_M2)
+#define PIPE_DATA_N2(tran) _TRANSCODER2(tran, _PIPEA_DATA_N2)
+#define PIPE_LINK_M1(tran) _TRANSCODER2(tran, _PIPEA_LINK_M1)
+#define PIPE_LINK_N1(tran) _TRANSCODER2(tran, _PIPEA_LINK_N1)
+#define PIPE_LINK_M2(tran) _TRANSCODER2(tran, _PIPEA_LINK_M2)
+#define PIPE_LINK_N2(tran) _TRANSCODER2(tran, _PIPEA_LINK_N2)
 
 /* CPU panel fitter */
 /* IVB+ has 3 fitters, 0 is 7x5 capable, the other two only 3x3 */
@@ -4049,6 +4517,7 @@
 #define  GEN8_DE_PIPE_A_IRQ		(1<<16)
 #define  GEN8_DE_PIPE_IRQ(pipe)		(1<<(16+pipe))
 #define  GEN8_GT_VECS_IRQ		(1<<6)
+#define  GEN8_GT_PM_IRQ			(1<<4)
 #define  GEN8_GT_VCS2_IRQ		(1<<3)
 #define  GEN8_GT_VCS1_IRQ		(1<<2)
 #define  GEN8_GT_BCS_IRQ		(1<<1)
@@ -4076,7 +4545,7 @@
 #define  GEN8_PIPE_SPRITE_FAULT		(1 << 9)
 #define  GEN8_PIPE_PRIMARY_FAULT	(1 << 8)
 #define  GEN8_PIPE_SPRITE_FLIP_DONE	(1 << 5)
-#define  GEN8_PIPE_FLIP_DONE		(1 << 4)
+#define  GEN8_PIPE_PRIMARY_FLIP_DONE	(1 << 4)
 #define  GEN8_PIPE_SCAN_LINE_EVENT	(1 << 2)
 #define  GEN8_PIPE_VSYNC		(1 << 1)
 #define  GEN8_PIPE_VBLANK		(1 << 0)
@@ -4108,13 +4577,14 @@
 #define  ILK_ELPIN_409_SELECT	(1 << 25)
 #define  ILK_DPARB_GATE	(1<<22)
 #define  ILK_VSDPFD_FULL	(1<<21)
-#define ILK_DISPLAY_CHICKEN_FUSES	0x42014
-#define  ILK_INTERNAL_GRAPHICS_DISABLE	(1<<31)
-#define  ILK_INTERNAL_DISPLAY_DISABLE	(1<<30)
-#define  ILK_DISPLAY_DEBUG_DISABLE	(1<<29)
-#define  ILK_HDCP_DISABLE		(1<<25)
-#define  ILK_eDP_A_DISABLE		(1<<24)
-#define  ILK_DESKTOP			(1<<23)
+#define FUSE_STRAP			0x42014
+#define  ILK_INTERNAL_GRAPHICS_DISABLE	(1 << 31)
+#define  ILK_INTERNAL_DISPLAY_DISABLE	(1 << 30)
+#define  ILK_DISPLAY_DEBUG_DISABLE	(1 << 29)
+#define  ILK_HDCP_DISABLE		(1 << 25)
+#define  ILK_eDP_A_DISABLE		(1 << 24)
+#define  HSW_CDCLK_LIMIT		(1 << 24)
+#define  ILK_DESKTOP			(1 << 23)
 
 #define ILK_DSPCLK_GATE_D			0x42020
 #define   ILK_VRHUNIT_CLOCK_GATE_DISABLE	(1 << 28)
@@ -4133,15 +4603,20 @@
 
 #define _CHICKEN_PIPESL_1_A	0x420b0
 #define _CHICKEN_PIPESL_1_B	0x420b4
-#define  DPRS_MASK_VBLANK_SRD	(1 << 0)
+#define  HSW_FBCQ_DIS			(1 << 22)
+#define  BDW_DPRS_MASK_VBLANK_SRD	(1 << 0)
 #define CHICKEN_PIPESL_1(pipe) _PIPE(pipe, _CHICKEN_PIPESL_1_A, _CHICKEN_PIPESL_1_B)
 
 #define DISP_ARB_CTL	0x45000
 #define  DISP_TILE_SURFACE_SWIZZLING	(1<<13)
 #define  DISP_FBC_WM_DIS		(1<<15)
+#define DISP_ARB_CTL2	0x45004
+#define  DISP_DATA_PARTITION_5_6	(1<<6)
 #define GEN7_MSG_CTL	0x45010
 #define  WAIT_FOR_PCH_RESET_ACK		(1<<1)
 #define  WAIT_FOR_PCH_FLR_ACK		(1<<0)
+#define HSW_NDE_RSTWRN_OPT	0x46408
+#define  RESET_PCH_HANDSHAKE_ENABLE	(1<<4)
 
 /* GEN7 chicken */
 #define GEN7_COMMON_SLICE_CHICKEN1		0x7010
@@ -4149,8 +4624,11 @@
 #define COMMON_SLICE_CHICKEN2			0x7014
 # define GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE	(1<<0)
 
+#define GEN7_L3SQCREG1				0xB010
+#define  VLV_B0_WA_L3SQCREG1_VALUE		0x00D30000
+
 #define GEN7_L3CNTLREG1				0xB01C
-#define  GEN7_WA_FOR_GEN7_L3_CONTROL			0x3C4FFF8C
+#define  GEN7_WA_FOR_GEN7_L3_CONTROL			0x3C47FF8C
 #define  GEN7_L3AGDIS				(1<<19)
 
 #define GEN7_L3_CHICKEN_MODE_REGISTER		0xB030
@@ -4159,6 +4637,10 @@
 #define GEN7_L3SQCREG4				0xb034
 #define  L3SQ_URB_READ_CAM_MATCH_DISABLE	(1<<27)
 
+/* GEN8 chicken */
+#define HDC_CHICKEN0				0x7300
+#define  HDC_FORCE_NON_COHERENT			(1<<4)
+
 /* WaCatErrorRejectionIssue */
 #define GEN7_SQ_CHICKEN_MBCUNIT_CONFIG		0x9030
 #define  GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB	(1<<11)
@@ -4166,9 +4648,6 @@
 #define HSW_SCRATCH1				0xb038
 #define  HSW_SCRATCH1_L3_DATA_ATOMICS_DISABLE	(1<<27)
 
-#define HSW_FUSE_STRAP		0x42014
-#define  HSW_CDCLK_LIMIT	(1 << 24)
-
 /* PCH */
 
 /* south display engine interrupt: IBX */
@@ -4454,24 +4933,24 @@
 #define HSW_VIDEO_DIP_GCP_B		0x61210
 
 #define HSW_TVIDEO_DIP_CTL(trans) \
-	 _TRANSCODER(trans, HSW_VIDEO_DIP_CTL_A, HSW_VIDEO_DIP_CTL_B)
+	 _TRANSCODER2(trans, HSW_VIDEO_DIP_CTL_A)
 #define HSW_TVIDEO_DIP_AVI_DATA(trans) \
-	 _TRANSCODER(trans, HSW_VIDEO_DIP_AVI_DATA_A, HSW_VIDEO_DIP_AVI_DATA_B)
+	 _TRANSCODER2(trans, HSW_VIDEO_DIP_AVI_DATA_A)
 #define HSW_TVIDEO_DIP_VS_DATA(trans) \
-	 _TRANSCODER(trans, HSW_VIDEO_DIP_VS_DATA_A, HSW_VIDEO_DIP_VS_DATA_B)
+	 _TRANSCODER2(trans, HSW_VIDEO_DIP_VS_DATA_A)
 #define HSW_TVIDEO_DIP_SPD_DATA(trans) \
-	 _TRANSCODER(trans, HSW_VIDEO_DIP_SPD_DATA_A, HSW_VIDEO_DIP_SPD_DATA_B)
+	 _TRANSCODER2(trans, HSW_VIDEO_DIP_SPD_DATA_A)
 #define HSW_TVIDEO_DIP_GCP(trans) \
-	_TRANSCODER(trans, HSW_VIDEO_DIP_GCP_A, HSW_VIDEO_DIP_GCP_B)
+	_TRANSCODER2(trans, HSW_VIDEO_DIP_GCP_A)
 #define HSW_TVIDEO_DIP_VSC_DATA(trans) \
-	 _TRANSCODER(trans, HSW_VIDEO_DIP_VSC_DATA_A, HSW_VIDEO_DIP_VSC_DATA_B)
+	 _TRANSCODER2(trans, HSW_VIDEO_DIP_VSC_DATA_A)
 
 #define HSW_STEREO_3D_CTL_A	0x70020
 #define   S3D_ENABLE		(1<<31)
 #define HSW_STEREO_3D_CTL_B	0x71020
 
 #define HSW_STEREO_3D_CTL(trans) \
-	_TRANSCODER(trans, HSW_STEREO_3D_CTL_A, HSW_STEREO_3D_CTL_A)
+	_PIPE2(trans, HSW_STEREO_3D_CTL_A)
 
 #define _PCH_TRANS_HTOTAL_B          0xe1000
 #define _PCH_TRANS_HBLANK_B          0xe1004
@@ -4778,6 +5257,8 @@
 #define  PORT_TRANS_SEL_CPT(pipe)	((pipe) << 29)
 #define  PORT_TO_PIPE(val)	(((val) & (1<<30)) >> 30)
 #define  PORT_TO_PIPE_CPT(val)	(((val) & PORT_TRANS_SEL_MASK) >> 29)
+#define  SDVO_PORT_TO_PIPE_CHV(val)	(((val) & (3<<24)) >> 24)
+#define  DP_PORT_TO_PIPE_CHV(val)	(((val) & (3<<16)) >> 16)
 
 #define TRANS_DP_CTL_A		0xe0300
 #define TRANS_DP_CTL_B		0xe1300
@@ -4834,6 +5315,8 @@
 
 #define  EDP_LINK_TRAIN_VOL_EMP_MASK_IVB	(0x3f<<22)
 
+#define  VLV_PMWGICZ				0x1300a4
+
 #define  FORCEWAKE				0xA18C
 #define  FORCEWAKE_VLV				0x1300b0
 #define  FORCEWAKE_ACK_VLV			0x1300b4
@@ -4842,21 +5325,34 @@
 #define  FORCEWAKE_ACK_HSW			0x130044
 #define  FORCEWAKE_ACK				0x130090
 #define  VLV_GTLC_WAKE_CTRL			0x130090
+#define   VLV_GTLC_RENDER_CTX_EXISTS		(1 << 25)
+#define   VLV_GTLC_MEDIA_CTX_EXISTS		(1 << 24)
+#define   VLV_GTLC_ALLOWWAKEREQ			(1 << 0)
+
 #define  VLV_GTLC_PW_STATUS			0x130094
+#define   VLV_GTLC_ALLOWWAKEACK			(1 << 0)
+#define   VLV_GTLC_ALLOWWAKEERR			(1 << 1)
+#define   VLV_GTLC_PW_MEDIA_STATUS_MASK		(1 << 5)
+#define   VLV_GTLC_PW_RENDER_STATUS_MASK	(1 << 7)
 #define  FORCEWAKE_MT				0xa188 /* multi-threaded */
 #define   FORCEWAKE_KERNEL			0x1
 #define   FORCEWAKE_USER			0x2
 #define  FORCEWAKE_MT_ACK			0x130040
 #define  ECOBUS					0xa180
 #define    FORCEWAKE_MT_ENABLE			(1<<5)
+#define  VLV_SPAREG2H				0xA194
 
 #define  GTFIFODBG				0x120000
-#define    GT_FIFO_CPU_ERROR_MASK		7
+#define    GT_FIFO_SBDROPERR			(1<<6)
+#define    GT_FIFO_BLOBDROPERR			(1<<5)
+#define    GT_FIFO_SB_READ_ABORTERR		(1<<4)
+#define    GT_FIFO_DROPERR			(1<<3)
 #define    GT_FIFO_OVFERR			(1<<2)
 #define    GT_FIFO_IAWRERR			(1<<1)
 #define    GT_FIFO_IARDERR			(1<<0)
 
-#define  GT_FIFO_FREE_ENTRIES			0x120008
+#define  GTFIFOCTL				0x120008
+#define    GT_FIFO_FREE_ENTRIES_MASK		0x7f
 #define    GT_FIFO_NUM_RESERVED_ENTRIES		20
 
 #define  HSW_IDICR				0x9008
@@ -4864,6 +5360,7 @@
 #define  HSW_EDRAM_PRESENT			0x120010
 
 #define GEN6_UCGCTL1				0x9400
+# define GEN6_EU_TCUNIT_CLOCK_GATE_DISABLE		(1 << 16)
 # define GEN6_BLBUNIT_CLOCK_GATE_DISABLE		(1 << 5)
 # define GEN6_CSUNIT_CLOCK_GATE_DISABLE			(1 << 7)
 
@@ -4874,9 +5371,19 @@
 # define GEN6_RCPBUNIT_CLOCK_GATE_DISABLE		(1 << 12)
 # define GEN6_RCCUNIT_CLOCK_GATE_DISABLE		(1 << 11)
 
+#define GEN6_UCGCTL3				0x9408
+
 #define GEN7_UCGCTL4				0x940c
 #define  GEN7_L3BANK2X_CLOCK_GATE_DISABLE	(1<<25)
 
+#define GEN6_RCGCTL1				0x9410
+#define GEN6_RCGCTL2				0x9414
+#define GEN6_RSTCTL				0x9420
+
+#define GEN8_UCGCTL6				0x9430
+#define   GEN8_SDEUNIT_CLOCK_GATE_DISABLE	(1<<14)
+
+#define GEN6_GFXPAUSE				0xA000
 #define GEN6_RPNSWREQ				0xA008
 #define   GEN6_TURBO_DISABLE			(1<<31)
 #define   GEN6_FREQUENCY(x)			((x)<<25)
@@ -4890,6 +5397,7 @@
 #define   GEN6_RC_CTL_RC6_ENABLE		(1<<18)
 #define   GEN6_RC_CTL_RC1e_ENABLE		(1<<20)
 #define   GEN6_RC_CTL_RC7_ENABLE		(1<<22)
+#define   VLV_RC_CTL_CTX_RST_PARALLEL		(1<<24)
 #define   GEN7_RC_CTL_TO_MODE			(1<<28)
 #define   GEN6_RC_CTL_EI_MODE(x)		((x)<<27)
 #define   GEN6_RC_CTL_HW_ENABLE			(1<<31)
@@ -4928,6 +5436,9 @@
 #define GEN6_RP_UP_EI				0xA068
 #define GEN6_RP_DOWN_EI				0xA06C
 #define GEN6_RP_IDLE_HYSTERSIS			0xA070
+#define GEN6_RPDEUHWTC				0xA080
+#define GEN6_RPDEUC				0xA084
+#define GEN6_RPDEUCSW				0xA088
 #define GEN6_RC_STATE				0xA094
 #define GEN6_RC1_WAKE_RATE_LIMIT		0xA098
 #define GEN6_RC6_WAKE_RATE_LIMIT		0xA09C
@@ -4935,11 +5446,15 @@
 #define GEN6_RC_EVALUATION_INTERVAL		0xA0A8
 #define GEN6_RC_IDLE_HYSTERSIS			0xA0AC
 #define GEN6_RC_SLEEP				0xA0B0
+#define GEN6_RCUBMABDTMR			0xA0B0
 #define GEN6_RC1e_THRESHOLD			0xA0B4
 #define GEN6_RC6_THRESHOLD			0xA0B8
 #define GEN6_RC6p_THRESHOLD			0xA0BC
+#define VLV_RCEDATA				0xA0BC
 #define GEN6_RC6pp_THRESHOLD			0xA0C0
 #define GEN6_PMINTRMSK				0xA168
+#define GEN8_PMINTR_REDIRECT_TO_NON_DISP	(1<<31)
+#define VLV_PWRDWNUPCTL				0xA294
 
 #define GEN6_PMISR				0x44020
 #define GEN6_PMIMR				0x44024 /* rps_lock */
@@ -4956,12 +5471,22 @@
 						 GEN6_PM_RP_DOWN_THRESHOLD | \
 						 GEN6_PM_RP_DOWN_TIMEOUT)
 
+#define GEN7_GT_SCRATCH_BASE			0x4F100
+#define GEN7_GT_SCRATCH_REG_NUM			8
+
+#define VLV_GTLC_SURVIVABILITY_REG              0x130098
+#define VLV_GFX_CLK_STATUS_BIT			(1<<3)
+#define VLV_GFX_CLK_FORCE_ON_BIT		(1<<2)
+
 #define GEN6_GT_GFX_RC6_LOCKED			0x138104
 #define VLV_COUNTER_CONTROL			0x138104
 #define   VLV_COUNT_RANGE_HIGH			(1<<15)
 #define   VLV_MEDIA_RC6_COUNT_EN		(1<<1)
 #define   VLV_RENDER_RC6_COUNT_EN		(1<<0)
 #define GEN6_GT_GFX_RC6				0x138108
+#define VLV_GT_RENDER_RC6			0x138108
+#define VLV_GT_MEDIA_RC6			0x13810C
+
 #define GEN6_GT_GFX_RC6p			0x13810C
 #define GEN6_GT_GFX_RC6pp			0x138110
 
@@ -5017,6 +5542,10 @@
 #define   GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE	(1<<10)
 #define   GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE	(1<<3)
 
+#define GEN8_ROW_CHICKEN		0xe4f0
+#define   PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE	(1<<8)
+#define   STALL_DOP_GATING_DISABLE		(1<<5)
+
 #define GEN7_ROW_CHICKEN2		0xe4f4
 #define GEN7_ROW_CHICKEN2_GT2		0xf4f4
 #define   DOP_CLOCK_GATING_DISABLE	(1<<0)
@@ -5028,7 +5557,7 @@
 #define   GEN8_CENTROID_PIXEL_OPT_DIS	(1<<8)
 #define   GEN8_SAMPLER_POWER_BYPASS_DIS	(1<<1)
 
-#define G4X_AUD_VID_DID			(dev_priv->info->display_mmio_offset + 0x62020)
+#define G4X_AUD_VID_DID			(dev_priv->info.display_mmio_offset + 0x62020)
 #define INTEL_AUDIO_DEVCL		0x808629FB
 #define INTEL_AUDIO_DEVBLC		0x80862801
 #define INTEL_AUDIO_DEVCTG		0x80862802
@@ -5189,8 +5718,8 @@
 #define TRANS_DDI_FUNC_CTL_B		0x61400
 #define TRANS_DDI_FUNC_CTL_C		0x62400
 #define TRANS_DDI_FUNC_CTL_EDP		0x6F400
-#define TRANS_DDI_FUNC_CTL(tran) _TRANSCODER(tran, TRANS_DDI_FUNC_CTL_A, \
-						   TRANS_DDI_FUNC_CTL_B)
+#define TRANS_DDI_FUNC_CTL(tran) _TRANSCODER2(tran, TRANS_DDI_FUNC_CTL_A)
+
 #define  TRANS_DDI_FUNC_ENABLE		(1<<31)
 /* Those bits are ignored by pipe EDP since it can only connect to DDI A */
 #define  TRANS_DDI_PORT_MASK		(7<<28)
@@ -5322,8 +5851,12 @@
 #define  SPLL_PLL_ENABLE		(1<<31)
 #define  SPLL_PLL_SSC			(1<<28)
 #define  SPLL_PLL_NON_SSC		(2<<28)
+#define  SPLL_PLL_LCPLL			(3<<28)
+#define  SPLL_PLL_REF_MASK		(3<<28)
 #define  SPLL_PLL_FREQ_810MHz		(0<<26)
 #define  SPLL_PLL_FREQ_1350MHz		(1<<26)
+#define  SPLL_PLL_FREQ_2700MHz		(2<<26)
+#define  SPLL_PLL_FREQ_MASK		(3<<26)
 
 /* WRPLL */
 #define WRPLL_CTL1			0x46040
@@ -5334,8 +5867,13 @@
 #define  WRPLL_PLL_SELECT_LCPLL_2700	(0x03<<28)
 /* WRPLL divider programming */
 #define  WRPLL_DIVIDER_REFERENCE(x)	((x)<<0)
+#define  WRPLL_DIVIDER_REF_MASK		(0xff)
 #define  WRPLL_DIVIDER_POST(x)		((x)<<8)
+#define  WRPLL_DIVIDER_POST_MASK	(0x3f<<8)
+#define  WRPLL_DIVIDER_POST_SHIFT	8
 #define  WRPLL_DIVIDER_FEEDBACK(x)	((x)<<16)
+#define  WRPLL_DIVIDER_FB_SHIFT		16
+#define  WRPLL_DIVIDER_FB_MASK		(0xff<<16)
 
 /* Port clock selection */
 #define PORT_CLK_SEL_A			0x46100
@@ -5348,6 +5886,7 @@
 #define  PORT_CLK_SEL_WRPLL1		(4<<29)
 #define  PORT_CLK_SEL_WRPLL2		(5<<29)
 #define  PORT_CLK_SEL_NONE		(7<<29)
+#define  PORT_CLK_SEL_MASK		(7<<29)
 
 /* Transcoder clock selection */
 #define TRANS_CLK_SEL_A			0x46140
@@ -5357,10 +5896,12 @@
 #define  TRANS_CLK_SEL_DISABLED		(0x0<<29)
 #define  TRANS_CLK_SEL_PORT(x)		((x+1)<<29)
 
-#define _TRANSA_MSA_MISC		0x60410
-#define _TRANSB_MSA_MISC		0x61410
-#define TRANS_MSA_MISC(tran) _TRANSCODER(tran, _TRANSA_MSA_MISC, \
-					       _TRANSB_MSA_MISC)
+#define TRANSA_MSA_MISC			0x60410
+#define TRANSB_MSA_MISC			0x61410
+#define TRANSC_MSA_MISC			0x62410
+#define TRANS_EDP_MSA_MISC		0x6f410
+#define TRANS_MSA_MISC(tran) _TRANSCODER2(tran, TRANSA_MSA_MISC)
+
 #define  TRANS_MSA_SYNC_CLK		(1<<0)
 #define  TRANS_MSA_6_BPC		(0<<5)
 #define  TRANS_MSA_8_BPC		(1<<5)
@@ -5400,6 +5941,8 @@
 
 /* SFUSE_STRAP */
 #define SFUSE_STRAP			0xc2014
+#define  SFUSE_STRAP_FUSE_LOCK		(1<<13)
+#define  SFUSE_STRAP_DISPLAY_DISABLED	(1<<7)
 #define  SFUSE_STRAP_DDIB_DETECTED	(1<<2)
 #define  SFUSE_STRAP_DDIC_DETECTED	(1<<1)
 #define  SFUSE_STRAP_DDID_DETECTED	(1<<0)
@@ -5868,4 +6411,12 @@
 #define MIPI_READ_DATA_VALID(pipe)	_PIPE(pipe, _MIPIA_READ_DATA_VALID, _MIPIB_READ_DATA_VALID)
 #define  READ_DATA_VALID(n)				(1 << (n))
 
+/* For UMS only (deprecated): */
+#define _PALETTE_A (dev_priv->info.display_mmio_offset + 0xa000)
+#define _PALETTE_B (dev_priv->info.display_mmio_offset + 0xa800)
+#define _DPLL_A (dev_priv->info.display_mmio_offset + 0x6014)
+#define _DPLL_B (dev_priv->info.display_mmio_offset + 0x6018)
+#define _DPLL_A_MD (dev_priv->info.display_mmio_offset + 0x601c)
+#define _DPLL_B_MD (dev_priv->info.display_mmio_offset + 0x6020)
+
 #endif /* _I915_REG_H_ */
diff --git a/drivers/gpu/drm/i915/i915_suspend.c b/drivers/gpu/drm/i915/i915_suspend.c
index 98790c7cccb..043123c77a1 100644
--- a/drivers/gpu/drm/i915/i915_suspend.c
+++ b/drivers/gpu/drm/i915/i915_suspend.c
@@ -192,7 +192,6 @@ static void i915_restore_vga(struct drm_device *dev)
 static void i915_save_display(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long flags;
 
 	/* Display arbitration control */
 	if (INTEL_INFO(dev)->gen <= 4)
@@ -203,46 +202,27 @@ static void i915_save_display(struct drm_device *dev)
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		i915_save_display_reg(dev);
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
-
 	/* LVDS state */
 	if (HAS_PCH_SPLIT(dev)) {
 		dev_priv->regfile.savePP_CONTROL = I915_READ(PCH_PP_CONTROL);
-		dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_PCH_CTL1);
-		dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_PCH_CTL2);
-		dev_priv->regfile.saveBLC_CPU_PWM_CTL = I915_READ(BLC_PWM_CPU_CTL);
-		dev_priv->regfile.saveBLC_CPU_PWM_CTL2 = I915_READ(BLC_PWM_CPU_CTL2);
 		if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
 			dev_priv->regfile.saveLVDS = I915_READ(PCH_LVDS);
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->regfile.savePP_CONTROL = I915_READ(PP_CONTROL);
 		dev_priv->regfile.savePFIT_PGM_RATIOS = I915_READ(PFIT_PGM_RATIOS);
 
-		dev_priv->regfile.saveBLC_PWM_CTL =
-			I915_READ(VLV_BLC_PWM_CTL(PIPE_A));
 		dev_priv->regfile.saveBLC_HIST_CTL =
 			I915_READ(VLV_BLC_HIST_CTL(PIPE_A));
-		dev_priv->regfile.saveBLC_PWM_CTL2 =
-			I915_READ(VLV_BLC_PWM_CTL2(PIPE_A));
-		dev_priv->regfile.saveBLC_PWM_CTL_B =
-			I915_READ(VLV_BLC_PWM_CTL(PIPE_B));
 		dev_priv->regfile.saveBLC_HIST_CTL_B =
 			I915_READ(VLV_BLC_HIST_CTL(PIPE_B));
-		dev_priv->regfile.saveBLC_PWM_CTL2_B =
-			I915_READ(VLV_BLC_PWM_CTL2(PIPE_B));
 	} else {
 		dev_priv->regfile.savePP_CONTROL = I915_READ(PP_CONTROL);
 		dev_priv->regfile.savePFIT_PGM_RATIOS = I915_READ(PFIT_PGM_RATIOS);
-		dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_CTL);
 		dev_priv->regfile.saveBLC_HIST_CTL = I915_READ(BLC_HIST_CTL);
-		if (INTEL_INFO(dev)->gen >= 4)
-			dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_CTL2);
 		if (IS_MOBILE(dev) && !IS_I830(dev))
 			dev_priv->regfile.saveLVDS = I915_READ(LVDS);
 	}
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
-
 	if (!IS_I830(dev) && !IS_845G(dev) && !HAS_PCH_SPLIT(dev))
 		dev_priv->regfile.savePFIT_CONTROL = I915_READ(PFIT_CONTROL);
 
@@ -256,19 +236,9 @@ static void i915_save_display(struct drm_device *dev)
 		dev_priv->regfile.savePP_DIVISOR = I915_READ(PP_DIVISOR);
 	}
 
-	/* Only regfile.save FBC state on the platform that supports FBC */
-	if (I915_HAS_FBC(dev)) {
-		if (HAS_PCH_SPLIT(dev)) {
-			dev_priv->regfile.saveDPFC_CB_BASE = I915_READ(ILK_DPFC_CB_BASE);
-		} else if (IS_GM45(dev)) {
-			dev_priv->regfile.saveDPFC_CB_BASE = I915_READ(DPFC_CB_BASE);
-		} else {
-			dev_priv->regfile.saveFBC_CFB_BASE = I915_READ(FBC_CFB_BASE);
-			dev_priv->regfile.saveFBC_LL_BASE = I915_READ(FBC_LL_BASE);
-			dev_priv->regfile.saveFBC_CONTROL2 = I915_READ(FBC_CONTROL2);
-			dev_priv->regfile.saveFBC_CONTROL = I915_READ(FBC_CONTROL);
-		}
-	}
+	/* save FBC interval */
+	if (HAS_FBC(dev) && INTEL_INFO(dev)->gen <= 4 && !IS_G4X(dev))
+		dev_priv->regfile.saveFBC_CONTROL = I915_READ(FBC_CONTROL);
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		i915_save_vga(dev);
@@ -278,7 +248,6 @@ static void i915_restore_display(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 mask = 0xffffffff;
-	unsigned long flags;
 
 	/* Display arbitration */
 	if (INTEL_INFO(dev)->gen <= 4)
@@ -287,12 +256,6 @@ static void i915_restore_display(struct drm_device *dev)
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		i915_restore_display_reg(dev);
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
-
-	/* LVDS state */
-	if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev))
-		I915_WRITE(BLC_PWM_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
-
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		mask = ~LVDS_PORT_EN;
 
@@ -305,13 +268,6 @@ static void i915_restore_display(struct drm_device *dev)
 		I915_WRITE(PFIT_CONTROL, dev_priv->regfile.savePFIT_CONTROL);
 
 	if (HAS_PCH_SPLIT(dev)) {
-		I915_WRITE(BLC_PWM_PCH_CTL1, dev_priv->regfile.saveBLC_PWM_CTL);
-		I915_WRITE(BLC_PWM_PCH_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
-		/* NOTE: BLC_PWM_CPU_CTL must be written after BLC_PWM_CPU_CTL2;
-		 * otherwise we get blank eDP screen after S3 on some machines
-		 */
-		I915_WRITE(BLC_PWM_CPU_CTL2, dev_priv->regfile.saveBLC_CPU_PWM_CTL2);
-		I915_WRITE(BLC_PWM_CPU_CTL, dev_priv->regfile.saveBLC_CPU_PWM_CTL);
 		I915_WRITE(PCH_PP_ON_DELAYS, dev_priv->regfile.savePP_ON_DELAYS);
 		I915_WRITE(PCH_PP_OFF_DELAYS, dev_priv->regfile.savePP_OFF_DELAYS);
 		I915_WRITE(PCH_PP_DIVISOR, dev_priv->regfile.savePP_DIVISOR);
@@ -319,21 +275,12 @@ static void i915_restore_display(struct drm_device *dev)
 		I915_WRITE(RSTDBYCTL,
 			   dev_priv->regfile.saveMCHBAR_RENDER_STANDBY);
 	} else if (IS_VALLEYVIEW(dev)) {
-		I915_WRITE(VLV_BLC_PWM_CTL(PIPE_A),
-			   dev_priv->regfile.saveBLC_PWM_CTL);
 		I915_WRITE(VLV_BLC_HIST_CTL(PIPE_A),
 			   dev_priv->regfile.saveBLC_HIST_CTL);
-		I915_WRITE(VLV_BLC_PWM_CTL2(PIPE_A),
-			   dev_priv->regfile.saveBLC_PWM_CTL2);
-		I915_WRITE(VLV_BLC_PWM_CTL(PIPE_B),
-			   dev_priv->regfile.saveBLC_PWM_CTL);
 		I915_WRITE(VLV_BLC_HIST_CTL(PIPE_B),
 			   dev_priv->regfile.saveBLC_HIST_CTL);
-		I915_WRITE(VLV_BLC_PWM_CTL2(PIPE_B),
-			   dev_priv->regfile.saveBLC_PWM_CTL2);
 	} else {
 		I915_WRITE(PFIT_PGM_RATIOS, dev_priv->regfile.savePFIT_PGM_RATIOS);
-		I915_WRITE(BLC_PWM_CTL, dev_priv->regfile.saveBLC_PWM_CTL);
 		I915_WRITE(BLC_HIST_CTL, dev_priv->regfile.saveBLC_HIST_CTL);
 		I915_WRITE(PP_ON_DELAYS, dev_priv->regfile.savePP_ON_DELAYS);
 		I915_WRITE(PP_OFF_DELAYS, dev_priv->regfile.savePP_OFF_DELAYS);
@@ -341,22 +288,12 @@ static void i915_restore_display(struct drm_device *dev)
 		I915_WRITE(PP_CONTROL, dev_priv->regfile.savePP_CONTROL);
 	}
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
-
 	/* only restore FBC info on the platform that supports FBC*/
 	intel_disable_fbc(dev);
-	if (I915_HAS_FBC(dev)) {
-		if (HAS_PCH_SPLIT(dev)) {
-			I915_WRITE(ILK_DPFC_CB_BASE, dev_priv->regfile.saveDPFC_CB_BASE);
-		} else if (IS_GM45(dev)) {
-			I915_WRITE(DPFC_CB_BASE, dev_priv->regfile.saveDPFC_CB_BASE);
-		} else {
-			I915_WRITE(FBC_CFB_BASE, dev_priv->regfile.saveFBC_CFB_BASE);
-			I915_WRITE(FBC_LL_BASE, dev_priv->regfile.saveFBC_LL_BASE);
-			I915_WRITE(FBC_CONTROL2, dev_priv->regfile.saveFBC_CONTROL2);
-			I915_WRITE(FBC_CONTROL, dev_priv->regfile.saveFBC_CONTROL);
-		}
-	}
+
+	/* restore FBC interval */
+	if (HAS_FBC(dev) && INTEL_INFO(dev)->gen <= 4 && !IS_G4X(dev))
+		I915_WRITE(FBC_CONTROL, dev_priv->regfile.saveFBC_CONTROL);
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		i915_restore_vga(dev);
@@ -369,10 +306,6 @@ int i915_save_state(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 
-	if (INTEL_INFO(dev)->gen <= 4)
-		pci_read_config_byte(dev->pdev, LBB,
-				     &dev_priv->regfile.saveLBB);
-
 	mutex_lock(&dev->struct_mutex);
 
 	i915_save_display(dev);
@@ -395,8 +328,6 @@ int i915_save_state(struct drm_device *dev)
 		}
 	}
 
-	intel_disable_gt_powersave(dev);
-
 	/* Cache mode state */
 	if (INTEL_INFO(dev)->gen < 7)
 		dev_priv->regfile.saveCACHE_MODE_0 = I915_READ(CACHE_MODE_0);
@@ -422,10 +353,6 @@ int i915_restore_state(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 
-	if (INTEL_INFO(dev)->gen <= 4)
-		pci_write_config_byte(dev->pdev, LBB,
-				      dev_priv->regfile.saveLBB);
-
 	mutex_lock(&dev->struct_mutex);
 
 	i915_gem_restore_fences(dev);
diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c
index cef38fd320a..86ce39aad0f 100644
--- a/drivers/gpu/drm/i915/i915_sysfs.c
+++ b/drivers/gpu/drm/i915/i915_sysfs.c
@@ -40,10 +40,13 @@ static u32 calc_residency(struct drm_device *dev, const u32 reg)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u64 raw_time; /* 32b value may overflow during fixed point math */
 	u64 units = 128ULL, div = 100000ULL, bias = 100ULL;
+	u32 ret;
 
 	if (!intel_enable_rc6(dev))
 		return 0;
 
+	intel_runtime_pm_get(dev_priv);
+
 	/* On VLV, residency time is in CZ units rather than 1.28us */
 	if (IS_VALLEYVIEW(dev)) {
 		u32 clkctl2;
@@ -52,7 +55,8 @@ static u32 calc_residency(struct drm_device *dev, const u32 reg)
 			CLK_CTL2_CZCOUNT_30NS_SHIFT;
 		if (!clkctl2) {
 			WARN(!clkctl2, "bogus CZ count value");
-			return 0;
+			ret = 0;
+			goto out;
 		}
 		units = DIV_ROUND_UP_ULL(30ULL * bias, (u64)clkctl2);
 		if (I915_READ(VLV_COUNTER_CONTROL) & VLV_COUNT_RANGE_HIGH)
@@ -62,7 +66,11 @@ static u32 calc_residency(struct drm_device *dev, const u32 reg)
 	}
 
 	raw_time = I915_READ(reg) * units;
-	return DIV_ROUND_UP_ULL(raw_time, div);
+	ret = DIV_ROUND_UP_ULL(raw_time, div);
+
+out:
+	intel_runtime_pm_put(dev_priv);
+	return ret;
 }
 
 static ssize_t
@@ -178,18 +186,18 @@ i915_l3_write(struct file *filp, struct kobject *kobj,
 	struct drm_minor *dminor = dev_to_drm_minor(dev);
 	struct drm_device *drm_dev = dminor->dev;
 	struct drm_i915_private *dev_priv = drm_dev->dev_private;
-	struct i915_hw_context *ctx;
+	struct intel_context *ctx;
 	u32 *temp = NULL; /* Just here to make handling failures easy */
 	int slice = (int)(uintptr_t)attr->private;
 	int ret;
 
+	if (!HAS_HW_CONTEXTS(drm_dev))
+		return -ENXIO;
+
 	ret = l3_access_valid(drm_dev, offset);
 	if (ret)
 		return ret;
 
-	if (dev_priv->hw_contexts_disabled)
-		return -ENXIO;
-
 	ret = i915_mutex_lock_interruptible(drm_dev);
 	if (ret)
 		return ret;
@@ -255,16 +263,20 @@ static ssize_t gt_cur_freq_mhz_show(struct device *kdev,
 
 	flush_delayed_work(&dev_priv->rps.delayed_resume_work);
 
+	intel_runtime_pm_get(dev_priv);
+
 	mutex_lock(&dev_priv->rps.hw_lock);
 	if (IS_VALLEYVIEW(dev_priv->dev)) {
 		u32 freq;
 		freq = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
-		ret = vlv_gpu_freq(dev_priv->mem_freq, (freq >> 8) & 0xff);
+		ret = vlv_gpu_freq(dev_priv, (freq >> 8) & 0xff);
 	} else {
-		ret = dev_priv->rps.cur_delay * GT_FREQUENCY_MULTIPLIER;
+		ret = dev_priv->rps.cur_freq * GT_FREQUENCY_MULTIPLIER;
 	}
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
+	intel_runtime_pm_put(dev_priv);
+
 	return snprintf(buf, PAGE_SIZE, "%d\n", ret);
 }
 
@@ -276,8 +288,7 @@ static ssize_t vlv_rpe_freq_mhz_show(struct device *kdev,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
-			vlv_gpu_freq(dev_priv->mem_freq,
-				     dev_priv->rps.rpe_delay));
+			vlv_gpu_freq(dev_priv, dev_priv->rps.efficient_freq));
 }
 
 static ssize_t gt_max_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
@@ -291,9 +302,9 @@ static ssize_t gt_max_freq_mhz_show(struct device *kdev, struct device_attribute
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 	if (IS_VALLEYVIEW(dev_priv->dev))
-		ret = vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.max_delay);
+		ret = vlv_gpu_freq(dev_priv, dev_priv->rps.max_freq_softlimit);
 	else
-		ret = dev_priv->rps.max_delay * GT_FREQUENCY_MULTIPLIER;
+		ret = dev_priv->rps.max_freq_softlimit * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return snprintf(buf, PAGE_SIZE, "%d\n", ret);
@@ -306,7 +317,7 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev,
 	struct drm_minor *minor = dev_to_drm_minor(kdev);
 	struct drm_device *dev = minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val, rp_state_cap, hw_max, hw_min, non_oc_max;
+	u32 val;
 	ssize_t ret;
 
 	ret = kstrtou32(buf, 0, &val);
@@ -317,40 +328,36 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev,
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 
-	if (IS_VALLEYVIEW(dev_priv->dev)) {
-		val = vlv_freq_opcode(dev_priv->mem_freq, val);
-
-		hw_max = valleyview_rps_max_freq(dev_priv);
-		hw_min = valleyview_rps_min_freq(dev_priv);
-		non_oc_max = hw_max;
-	} else {
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		val = vlv_freq_opcode(dev_priv, val);
+	else
 		val /= GT_FREQUENCY_MULTIPLIER;
 
-		rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
-		hw_max = dev_priv->rps.hw_max;
-		non_oc_max = (rp_state_cap & 0xff);
-		hw_min = ((rp_state_cap & 0xff0000) >> 16);
-	}
-
-	if (val < hw_min || val > hw_max ||
-	    val < dev_priv->rps.min_delay) {
+	if (val < dev_priv->rps.min_freq ||
+	    val > dev_priv->rps.max_freq ||
+	    val < dev_priv->rps.min_freq_softlimit) {
 		mutex_unlock(&dev_priv->rps.hw_lock);
 		return -EINVAL;
 	}
 
-	if (val > non_oc_max)
+	if (val > dev_priv->rps.rp0_freq)
 		DRM_DEBUG("User requested overclocking to %d\n",
 			  val * GT_FREQUENCY_MULTIPLIER);
 
-	if (dev_priv->rps.cur_delay > val) {
-		if (IS_VALLEYVIEW(dev_priv->dev))
-			valleyview_set_rps(dev_priv->dev, val);
+	dev_priv->rps.max_freq_softlimit = val;
+
+	if (dev_priv->rps.cur_freq > val) {
+		if (IS_VALLEYVIEW(dev))
+			valleyview_set_rps(dev, val);
 		else
-			gen6_set_rps(dev_priv->dev, val);
+			gen6_set_rps(dev, val);
+	} else if (!IS_VALLEYVIEW(dev)) {
+		/* We still need gen6_set_rps to process the new max_delay and
+		 * update the interrupt limits even though frequency request is
+		 * unchanged. */
+		gen6_set_rps(dev, dev_priv->rps.cur_freq);
 	}
 
-	dev_priv->rps.max_delay = val;
-
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return count;
@@ -367,9 +374,9 @@ static ssize_t gt_min_freq_mhz_show(struct device *kdev, struct device_attribute
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 	if (IS_VALLEYVIEW(dev_priv->dev))
-		ret = vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.min_delay);
+		ret = vlv_gpu_freq(dev_priv, dev_priv->rps.min_freq_softlimit);
 	else
-		ret = dev_priv->rps.min_delay * GT_FREQUENCY_MULTIPLIER;
+		ret = dev_priv->rps.min_freq_softlimit * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return snprintf(buf, PAGE_SIZE, "%d\n", ret);
@@ -382,7 +389,7 @@ static ssize_t gt_min_freq_mhz_store(struct device *kdev,
 	struct drm_minor *minor = dev_to_drm_minor(kdev);
 	struct drm_device *dev = minor->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val, rp_state_cap, hw_max, hw_min;
+	u32 val;
 	ssize_t ret;
 
 	ret = kstrtou32(buf, 0, &val);
@@ -393,33 +400,32 @@ static ssize_t gt_min_freq_mhz_store(struct device *kdev,
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 
-	if (IS_VALLEYVIEW(dev)) {
-		val = vlv_freq_opcode(dev_priv->mem_freq, val);
-
-		hw_max = valleyview_rps_max_freq(dev_priv);
-		hw_min = valleyview_rps_min_freq(dev_priv);
-	} else {
+	if (IS_VALLEYVIEW(dev))
+		val = vlv_freq_opcode(dev_priv, val);
+	else
 		val /= GT_FREQUENCY_MULTIPLIER;
 
-		rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
-		hw_max = dev_priv->rps.hw_max;
-		hw_min = ((rp_state_cap & 0xff0000) >> 16);
-	}
-
-	if (val < hw_min || val > hw_max || val > dev_priv->rps.max_delay) {
+	if (val < dev_priv->rps.min_freq ||
+	    val > dev_priv->rps.max_freq ||
+	    val > dev_priv->rps.max_freq_softlimit) {
 		mutex_unlock(&dev_priv->rps.hw_lock);
 		return -EINVAL;
 	}
 
-	if (dev_priv->rps.cur_delay < val) {
+	dev_priv->rps.min_freq_softlimit = val;
+
+	if (dev_priv->rps.cur_freq < val) {
 		if (IS_VALLEYVIEW(dev))
 			valleyview_set_rps(dev, val);
 		else
-			gen6_set_rps(dev_priv->dev, val);
+			gen6_set_rps(dev, val);
+	} else if (!IS_VALLEYVIEW(dev)) {
+		/* We still need gen6_set_rps to process the new min_delay and
+		 * update the interrupt limits even though frequency request is
+		 * unchanged. */
+		gen6_set_rps(dev, dev_priv->rps.cur_freq);
 	}
 
-	dev_priv->rps.min_delay = val;
-
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return count;
@@ -449,7 +455,9 @@ static ssize_t gt_rp_mhz_show(struct device *kdev, struct device_attribute *attr
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	if (attr == &dev_attr_gt_RP0_freq_mhz) {
diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
index 6e580c98ded..f5aa0067755 100644
--- a/drivers/gpu/drm/i915/i915_trace.h
+++ b/drivers/gpu/drm/i915/i915_trace.h
@@ -7,6 +7,7 @@
 
 #include <drm/drmP.h>
 #include "i915_drv.h"
+#include "intel_drv.h"
 #include "intel_ringbuffer.h"
 
 #undef TRACE_SYSTEM
@@ -14,6 +15,80 @@
 #define TRACE_SYSTEM_STRING __stringify(TRACE_SYSTEM)
 #define TRACE_INCLUDE_FILE i915_trace
 
+/* pipe updates */
+
+TRACE_EVENT(i915_pipe_update_start,
+	    TP_PROTO(struct intel_crtc *crtc, u32 min, u32 max),
+	    TP_ARGS(crtc, min, max),
+
+	    TP_STRUCT__entry(
+			     __field(enum pipe, pipe)
+			     __field(u32, frame)
+			     __field(u32, scanline)
+			     __field(u32, min)
+			     __field(u32, max)
+			     ),
+
+	    TP_fast_assign(
+			   __entry->pipe = crtc->pipe;
+			   __entry->frame = crtc->base.dev->driver->get_vblank_counter(crtc->base.dev,
+										       crtc->pipe);
+			   __entry->scanline = intel_get_crtc_scanline(crtc);
+			   __entry->min = min;
+			   __entry->max = max;
+			   ),
+
+	    TP_printk("pipe %c, frame=%u, scanline=%u, min=%u, max=%u",
+		      pipe_name(__entry->pipe), __entry->frame,
+		       __entry->scanline, __entry->min, __entry->max)
+);
+
+TRACE_EVENT(i915_pipe_update_vblank_evaded,
+	    TP_PROTO(struct intel_crtc *crtc, u32 min, u32 max, u32 frame),
+	    TP_ARGS(crtc, min, max, frame),
+
+	    TP_STRUCT__entry(
+			     __field(enum pipe, pipe)
+			     __field(u32, frame)
+			     __field(u32, scanline)
+			     __field(u32, min)
+			     __field(u32, max)
+			     ),
+
+	    TP_fast_assign(
+			   __entry->pipe = crtc->pipe;
+			   __entry->frame = frame;
+			   __entry->scanline = intel_get_crtc_scanline(crtc);
+			   __entry->min = min;
+			   __entry->max = max;
+			   ),
+
+	    TP_printk("pipe %c, frame=%u, scanline=%u, min=%u, max=%u",
+		      pipe_name(__entry->pipe), __entry->frame,
+		       __entry->scanline, __entry->min, __entry->max)
+);
+
+TRACE_EVENT(i915_pipe_update_end,
+	    TP_PROTO(struct intel_crtc *crtc, u32 frame),
+	    TP_ARGS(crtc, frame),
+
+	    TP_STRUCT__entry(
+			     __field(enum pipe, pipe)
+			     __field(u32, frame)
+			     __field(u32, scanline)
+			     ),
+
+	    TP_fast_assign(
+			   __entry->pipe = crtc->pipe;
+			   __entry->frame = frame;
+			   __entry->scanline = intel_get_crtc_scanline(crtc);
+			   ),
+
+	    TP_printk("pipe %c, frame=%u, scanline=%u",
+		      pipe_name(__entry->pipe), __entry->frame,
+		      __entry->scanline)
+);
+
 /* object tracking */
 
 TRACE_EVENT(i915_gem_object_create,
@@ -34,15 +109,15 @@ TRACE_EVENT(i915_gem_object_create,
 );
 
 TRACE_EVENT(i915_vma_bind,
-	    TP_PROTO(struct i915_vma *vma, bool mappable),
-	    TP_ARGS(vma, mappable),
+	    TP_PROTO(struct i915_vma *vma, unsigned flags),
+	    TP_ARGS(vma, flags),
 
 	    TP_STRUCT__entry(
 			     __field(struct drm_i915_gem_object *, obj)
 			     __field(struct i915_address_space *, vm)
 			     __field(u32, offset)
 			     __field(u32, size)
-			     __field(bool, mappable)
+			     __field(unsigned, flags)
 			     ),
 
 	    TP_fast_assign(
@@ -50,12 +125,12 @@ TRACE_EVENT(i915_vma_bind,
 			   __entry->vm = vma->vm;
 			   __entry->offset = vma->node.start;
 			   __entry->size = vma->node.size;
-			   __entry->mappable = mappable;
+			   __entry->flags = flags;
 			   ),
 
 	    TP_printk("obj=%p, offset=%08x size=%x%s vm=%p",
 		      __entry->obj, __entry->offset, __entry->size,
-		      __entry->mappable ? ", mappable" : "",
+		      __entry->flags & PIN_MAPPABLE ? ", mappable" : "",
 		      __entry->vm)
 );
 
@@ -196,26 +271,26 @@ DEFINE_EVENT(i915_gem_object, i915_gem_object_destroy,
 );
 
 TRACE_EVENT(i915_gem_evict,
-	    TP_PROTO(struct drm_device *dev, u32 size, u32 align, bool mappable),
-	    TP_ARGS(dev, size, align, mappable),
+	    TP_PROTO(struct drm_device *dev, u32 size, u32 align, unsigned flags),
+	    TP_ARGS(dev, size, align, flags),
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
 			     __field(u32, size)
 			     __field(u32, align)
-			     __field(bool, mappable)
+			     __field(unsigned, flags)
 			    ),
 
 	    TP_fast_assign(
 			   __entry->dev = dev->primary->index;
 			   __entry->size = size;
 			   __entry->align = align;
-			   __entry->mappable = mappable;
+			   __entry->flags = flags;
 			  ),
 
 	    TP_printk("dev=%d, size=%d, align=%d %s",
 		      __entry->dev, __entry->size, __entry->align,
-		      __entry->mappable ? ", mappable" : "")
+		      __entry->flags & PIN_MAPPABLE ? ", mappable" : "")
 );
 
 TRACE_EVENT(i915_gem_evict_everything,
@@ -238,19 +313,21 @@ TRACE_EVENT(i915_gem_evict_vm,
 	    TP_ARGS(vm),
 
 	    TP_STRUCT__entry(
+			     __field(u32, dev)
 			     __field(struct i915_address_space *, vm)
 			    ),
 
 	    TP_fast_assign(
+			   __entry->dev = vm->dev->primary->index;
 			   __entry->vm = vm;
 			  ),
 
-	    TP_printk("dev=%d, vm=%p", __entry->vm->dev->primary->index, __entry->vm)
+	    TP_printk("dev=%d, vm=%p", __entry->dev, __entry->vm)
 );
 
 TRACE_EVENT(i915_gem_ring_sync_to,
-	    TP_PROTO(struct intel_ring_buffer *from,
-		     struct intel_ring_buffer *to,
+	    TP_PROTO(struct intel_engine_cs *from,
+		     struct intel_engine_cs *to,
 		     u32 seqno),
 	    TP_ARGS(from, to, seqno),
 
@@ -275,7 +352,7 @@ TRACE_EVENT(i915_gem_ring_sync_to,
 );
 
 TRACE_EVENT(i915_gem_ring_dispatch,
-	    TP_PROTO(struct intel_ring_buffer *ring, u32 seqno, u32 flags),
+	    TP_PROTO(struct intel_engine_cs *ring, u32 seqno, u32 flags),
 	    TP_ARGS(ring, seqno, flags),
 
 	    TP_STRUCT__entry(
@@ -298,7 +375,7 @@ TRACE_EVENT(i915_gem_ring_dispatch,
 );
 
 TRACE_EVENT(i915_gem_ring_flush,
-	    TP_PROTO(struct intel_ring_buffer *ring, u32 invalidate, u32 flush),
+	    TP_PROTO(struct intel_engine_cs *ring, u32 invalidate, u32 flush),
 	    TP_ARGS(ring, invalidate, flush),
 
 	    TP_STRUCT__entry(
@@ -321,7 +398,7 @@ TRACE_EVENT(i915_gem_ring_flush,
 );
 
 DECLARE_EVENT_CLASS(i915_gem_request,
-	    TP_PROTO(struct intel_ring_buffer *ring, u32 seqno),
+	    TP_PROTO(struct intel_engine_cs *ring, u32 seqno),
 	    TP_ARGS(ring, seqno),
 
 	    TP_STRUCT__entry(
@@ -341,12 +418,12 @@ DECLARE_EVENT_CLASS(i915_gem_request,
 );
 
 DEFINE_EVENT(i915_gem_request, i915_gem_request_add,
-	    TP_PROTO(struct intel_ring_buffer *ring, u32 seqno),
+	    TP_PROTO(struct intel_engine_cs *ring, u32 seqno),
 	    TP_ARGS(ring, seqno)
 );
 
 TRACE_EVENT(i915_gem_request_complete,
-	    TP_PROTO(struct intel_ring_buffer *ring),
+	    TP_PROTO(struct intel_engine_cs *ring),
 	    TP_ARGS(ring),
 
 	    TP_STRUCT__entry(
@@ -366,12 +443,12 @@ TRACE_EVENT(i915_gem_request_complete,
 );
 
 DEFINE_EVENT(i915_gem_request, i915_gem_request_retire,
-	    TP_PROTO(struct intel_ring_buffer *ring, u32 seqno),
+	    TP_PROTO(struct intel_engine_cs *ring, u32 seqno),
 	    TP_ARGS(ring, seqno)
 );
 
 TRACE_EVENT(i915_gem_request_wait_begin,
-	    TP_PROTO(struct intel_ring_buffer *ring, u32 seqno),
+	    TP_PROTO(struct intel_engine_cs *ring, u32 seqno),
 	    TP_ARGS(ring, seqno),
 
 	    TP_STRUCT__entry(
@@ -400,12 +477,12 @@ TRACE_EVENT(i915_gem_request_wait_begin,
 );
 
 DEFINE_EVENT(i915_gem_request, i915_gem_request_wait_end,
-	    TP_PROTO(struct intel_ring_buffer *ring, u32 seqno),
+	    TP_PROTO(struct intel_engine_cs *ring, u32 seqno),
 	    TP_ARGS(ring, seqno)
 );
 
 DECLARE_EVENT_CLASS(i915_ring,
-	    TP_PROTO(struct intel_ring_buffer *ring),
+	    TP_PROTO(struct intel_engine_cs *ring),
 	    TP_ARGS(ring),
 
 	    TP_STRUCT__entry(
@@ -422,12 +499,12 @@ DECLARE_EVENT_CLASS(i915_ring,
 );
 
 DEFINE_EVENT(i915_ring, i915_ring_wait_begin,
-	    TP_PROTO(struct intel_ring_buffer *ring),
+	    TP_PROTO(struct intel_engine_cs *ring),
 	    TP_ARGS(ring)
 );
 
 DEFINE_EVENT(i915_ring, i915_ring_wait_end,
-	    TP_PROTO(struct intel_ring_buffer *ring),
+	    TP_PROTO(struct intel_engine_cs *ring),
 	    TP_ARGS(ring)
 );
 
diff --git a/drivers/gpu/drm/i915/i915_ums.c b/drivers/gpu/drm/i915/i915_ums.c
index 967da4772c4..480da593e6c 100644
--- a/drivers/gpu/drm/i915/i915_ums.c
+++ b/drivers/gpu/drm/i915/i915_ums.c
@@ -270,6 +270,22 @@ void i915_save_display_reg(struct drm_device *dev)
 	}
 	/* FIXME: regfile.save TV & SDVO state */
 
+	/* Backlight */
+	if (INTEL_INFO(dev)->gen <= 4)
+		pci_read_config_byte(dev->pdev, PCI_LBPC,
+				     &dev_priv->regfile.saveLBB);
+
+	if (HAS_PCH_SPLIT(dev)) {
+		dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_PCH_CTL1);
+		dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_PCH_CTL2);
+		dev_priv->regfile.saveBLC_CPU_PWM_CTL = I915_READ(BLC_PWM_CPU_CTL);
+		dev_priv->regfile.saveBLC_CPU_PWM_CTL2 = I915_READ(BLC_PWM_CPU_CTL2);
+	} else {
+		dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_CTL);
+		if (INTEL_INFO(dev)->gen >= 4)
+			dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_CTL2);
+	}
+
 	return;
 }
 
@@ -280,6 +296,25 @@ void i915_restore_display_reg(struct drm_device *dev)
 	int dpll_b_reg, fpb0_reg, fpb1_reg;
 	int i;
 
+	/* Backlight */
+	if (INTEL_INFO(dev)->gen <= 4)
+		pci_write_config_byte(dev->pdev, PCI_LBPC,
+				      dev_priv->regfile.saveLBB);
+
+	if (HAS_PCH_SPLIT(dev)) {
+		I915_WRITE(BLC_PWM_PCH_CTL1, dev_priv->regfile.saveBLC_PWM_CTL);
+		I915_WRITE(BLC_PWM_PCH_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
+		/* NOTE: BLC_PWM_CPU_CTL must be written after BLC_PWM_CPU_CTL2;
+		 * otherwise we get blank eDP screen after S3 on some machines
+		 */
+		I915_WRITE(BLC_PWM_CPU_CTL2, dev_priv->regfile.saveBLC_CPU_PWM_CTL2);
+		I915_WRITE(BLC_PWM_CPU_CTL, dev_priv->regfile.saveBLC_CPU_PWM_CTL);
+	} else {
+		if (INTEL_INFO(dev)->gen >= 4)
+			I915_WRITE(BLC_PWM_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
+		I915_WRITE(BLC_PWM_CTL, dev_priv->regfile.saveBLC_PWM_CTL);
+	}
+
 	/* Display port ratios (must be done before clock is set) */
 	if (SUPPORTS_INTEGRATED_DP(dev)) {
 		I915_WRITE(_PIPEA_DATA_M_G4X, dev_priv->regfile.savePIPEA_GMCH_DATA_M);
diff --git a/drivers/gpu/drm/i915/intel_acpi.c b/drivers/gpu/drm/i915/intel_acpi.c
index dfff0907f70..d96eee1ae9c 100644
--- a/drivers/gpu/drm/i915/intel_acpi.c
+++ b/drivers/gpu/drm/i915/intel_acpi.c
@@ -6,14 +6,10 @@
 #include <linux/pci.h>
 #include <linux/acpi.h>
 #include <linux/vga_switcheroo.h>
-#include <acpi/acpi_drivers.h>
-
 #include <drm/drmP.h>
 #include "i915_drv.h"
 
 #define INTEL_DSM_REVISION_ID 1 /* For Calpella anyway... */
-
-#define INTEL_DSM_FN_SUPPORTED_FUNCTIONS 0 /* No args */
 #define INTEL_DSM_FN_PLATFORM_MUX_INFO 1 /* No args */
 
 static struct intel_dsm_priv {
@@ -28,61 +24,6 @@ static const u8 intel_dsm_guid[] = {
 	0x0f, 0x13, 0x17, 0xb0, 0x1c, 0x2c
 };
 
-static int intel_dsm(acpi_handle handle, int func)
-{
-	struct acpi_buffer output = { ACPI_ALLOCATE_BUFFER, NULL };
-	struct acpi_object_list input;
-	union acpi_object params[4];
-	union acpi_object *obj;
-	u32 result;
-	int ret = 0;
-
-	input.count = 4;
-	input.pointer = params;
-	params[0].type = ACPI_TYPE_BUFFER;
-	params[0].buffer.length = sizeof(intel_dsm_guid);
-	params[0].buffer.pointer = (char *)intel_dsm_guid;
-	params[1].type = ACPI_TYPE_INTEGER;
-	params[1].integer.value = INTEL_DSM_REVISION_ID;
-	params[2].type = ACPI_TYPE_INTEGER;
-	params[2].integer.value = func;
-	params[3].type = ACPI_TYPE_PACKAGE;
-	params[3].package.count = 0;
-	params[3].package.elements = NULL;
-
-	ret = acpi_evaluate_object(handle, "_DSM", &input, &output);
-	if (ret) {
-		DRM_DEBUG_DRIVER("failed to evaluate _DSM: %d\n", ret);
-		return ret;
-	}
-
-	obj = (union acpi_object *)output.pointer;
-
-	result = 0;
-	switch (obj->type) {
-	case ACPI_TYPE_INTEGER:
-		result = obj->integer.value;
-		break;
-
-	case ACPI_TYPE_BUFFER:
-		if (obj->buffer.length == 4) {
-			result = (obj->buffer.pointer[0] |
-				(obj->buffer.pointer[1] <<  8) |
-				(obj->buffer.pointer[2] << 16) |
-				(obj->buffer.pointer[3] << 24));
-			break;
-		}
-	default:
-		ret = -EINVAL;
-		break;
-	}
-	if (result == 0x80000002)
-		ret = -ENODEV;
-
-	kfree(output.pointer);
-	return ret;
-}
-
 static char *intel_dsm_port_name(u8 id)
 {
 	switch (id) {
@@ -137,83 +78,56 @@ static char *intel_dsm_mux_type(u8 type)
 
 static void intel_dsm_platform_mux_info(void)
 {
-	struct acpi_buffer output = { ACPI_ALLOCATE_BUFFER, NULL };
-	struct acpi_object_list input;
-	union acpi_object params[4];
-	union acpi_object *pkg;
-	int i, ret;
-
-	input.count = 4;
-	input.pointer = params;
-	params[0].type = ACPI_TYPE_BUFFER;
-	params[0].buffer.length = sizeof(intel_dsm_guid);
-	params[0].buffer.pointer = (char *)intel_dsm_guid;
-	params[1].type = ACPI_TYPE_INTEGER;
-	params[1].integer.value = INTEL_DSM_REVISION_ID;
-	params[2].type = ACPI_TYPE_INTEGER;
-	params[2].integer.value = INTEL_DSM_FN_PLATFORM_MUX_INFO;
-	params[3].type = ACPI_TYPE_PACKAGE;
-	params[3].package.count = 0;
-	params[3].package.elements = NULL;
-
-	ret = acpi_evaluate_object(intel_dsm_priv.dhandle, "_DSM", &input,
-				   &output);
-	if (ret) {
-		DRM_DEBUG_DRIVER("failed to evaluate _DSM: %d\n", ret);
-		goto out;
+	int i;
+	union acpi_object *pkg, *connector_count;
+
+	pkg = acpi_evaluate_dsm_typed(intel_dsm_priv.dhandle, intel_dsm_guid,
+			INTEL_DSM_REVISION_ID, INTEL_DSM_FN_PLATFORM_MUX_INFO,
+			NULL, ACPI_TYPE_PACKAGE);
+	if (!pkg) {
+		DRM_DEBUG_DRIVER("failed to evaluate _DSM\n");
+		return;
 	}
 
-	pkg = (union acpi_object *)output.pointer;
-
-	if (pkg->type == ACPI_TYPE_PACKAGE) {
-		union acpi_object *connector_count = &pkg->package.elements[0];
-		DRM_DEBUG_DRIVER("MUX info connectors: %lld\n",
-			  (unsigned long long)connector_count->integer.value);
-		for (i = 1; i < pkg->package.count; i++) {
-			union acpi_object *obj = &pkg->package.elements[i];
-			union acpi_object *connector_id =
-				&obj->package.elements[0];
-			union acpi_object *info = &obj->package.elements[1];
-			DRM_DEBUG_DRIVER("Connector id: 0x%016llx\n",
-				  (unsigned long long)connector_id->integer.value);
-			DRM_DEBUG_DRIVER("  port id: %s\n",
-			       intel_dsm_port_name(info->buffer.pointer[0]));
-			DRM_DEBUG_DRIVER("  display mux info: %s\n",
-			       intel_dsm_mux_type(info->buffer.pointer[1]));
-			DRM_DEBUG_DRIVER("  aux/dc mux info: %s\n",
-			       intel_dsm_mux_type(info->buffer.pointer[2]));
-			DRM_DEBUG_DRIVER("  hpd mux info: %s\n",
-			       intel_dsm_mux_type(info->buffer.pointer[3]));
-		}
+	connector_count = &pkg->package.elements[0];
+	DRM_DEBUG_DRIVER("MUX info connectors: %lld\n",
+		  (unsigned long long)connector_count->integer.value);
+	for (i = 1; i < pkg->package.count; i++) {
+		union acpi_object *obj = &pkg->package.elements[i];
+		union acpi_object *connector_id = &obj->package.elements[0];
+		union acpi_object *info = &obj->package.elements[1];
+		DRM_DEBUG_DRIVER("Connector id: 0x%016llx\n",
+			  (unsigned long long)connector_id->integer.value);
+		DRM_DEBUG_DRIVER("  port id: %s\n",
+		       intel_dsm_port_name(info->buffer.pointer[0]));
+		DRM_DEBUG_DRIVER("  display mux info: %s\n",
+		       intel_dsm_mux_type(info->buffer.pointer[1]));
+		DRM_DEBUG_DRIVER("  aux/dc mux info: %s\n",
+		       intel_dsm_mux_type(info->buffer.pointer[2]));
+		DRM_DEBUG_DRIVER("  hpd mux info: %s\n",
+		       intel_dsm_mux_type(info->buffer.pointer[3]));
 	}
 
-out:
-	kfree(output.pointer);
+	ACPI_FREE(pkg);
 }
 
 static bool intel_dsm_pci_probe(struct pci_dev *pdev)
 {
 	acpi_handle dhandle;
-	int ret;
 
 	dhandle = ACPI_HANDLE(&pdev->dev);
 	if (!dhandle)
 		return false;
 
-	if (!acpi_has_method(dhandle, "_DSM")) {
+	if (!acpi_check_dsm(dhandle, intel_dsm_guid, INTEL_DSM_REVISION_ID,
+			    1 << INTEL_DSM_FN_PLATFORM_MUX_INFO)) {
 		DRM_DEBUG_KMS("no _DSM method for intel device\n");
 		return false;
 	}
 
-	ret = intel_dsm(dhandle, INTEL_DSM_FN_SUPPORTED_FUNCTIONS);
-	if (ret < 0) {
-		DRM_DEBUG_KMS("failed to get supported _DSM functions\n");
-		return false;
-	}
-
 	intel_dsm_priv.dhandle = dhandle;
-
 	intel_dsm_platform_mux_info();
+
 	return true;
 }
 
diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c
index e4fba39631a..827498e081d 100644
--- a/drivers/gpu/drm/i915/intel_bios.c
+++ b/drivers/gpu/drm/i915/intel_bios.c
@@ -49,13 +49,19 @@ find_section(struct bdb_header *bdb, int section_id)
 	total = bdb->bdb_size;
 
 	/* walk the sections looking for section_id */
-	while (index < total) {
+	while (index + 3 < total) {
 		current_id = *(base + index);
 		index++;
+
 		current_size = *((u16 *)(base + index));
 		index += 2;
+
+		if (index + current_size > total)
+			return NULL;
+
 		if (current_id == section_id)
 			return base + index;
+
 		index += current_size;
 	}
 
@@ -206,7 +212,7 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 	const struct lvds_dvo_timing *panel_dvo_timing;
 	const struct lvds_fp_timing *fp_timing;
 	struct drm_display_mode *panel_fixed_mode;
-	int i, downclock;
+	int i, downclock, drrs_mode;
 
 	lvds_options = find_section(bdb, BDB_LVDS_OPTIONS);
 	if (!lvds_options)
@@ -218,6 +224,28 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 
 	panel_type = lvds_options->panel_type;
 
+	drrs_mode = (lvds_options->dps_panel_type_bits
+				>> (panel_type * 2)) & MODE_MASK;
+	/*
+	 * VBT has static DRRS = 0 and seamless DRRS = 2.
+	 * The below piece of code is required to adjust vbt.drrs_type
+	 * to match the enum drrs_support_type.
+	 */
+	switch (drrs_mode) {
+	case 0:
+		dev_priv->vbt.drrs_type = STATIC_DRRS_SUPPORT;
+		DRM_DEBUG_KMS("DRRS supported mode is static\n");
+		break;
+	case 2:
+		dev_priv->vbt.drrs_type = SEAMLESS_DRRS_SUPPORT;
+		DRM_DEBUG_KMS("DRRS supported mode is seamless\n");
+		break;
+	default:
+		dev_priv->vbt.drrs_type = DRRS_NOT_SUPPORTED;
+		DRM_DEBUG_KMS("DRRS not supported (VBT input)\n");
+		break;
+	}
+
 	lvds_lfp_data = find_section(bdb, BDB_LVDS_LFP_DATA);
 	if (!lvds_lfp_data)
 		return;
@@ -259,7 +287,7 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 			downclock = dvo_timing->clock;
 	}
 
-	if (downclock < panel_dvo_timing->clock && i915_lvds_downclock) {
+	if (downclock < panel_dvo_timing->clock && i915.lvds_downclock) {
 		dev_priv->lvds_downclock_avail = 1;
 		dev_priv->lvds_downclock = downclock * 10;
 		DRM_DEBUG_KMS("LVDS downclock is found in VBT. "
@@ -281,6 +309,41 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 	}
 }
 
+static void
+parse_lfp_backlight(struct drm_i915_private *dev_priv, struct bdb_header *bdb)
+{
+	const struct bdb_lfp_backlight_data *backlight_data;
+	const struct bdb_lfp_backlight_data_entry *entry;
+
+	backlight_data = find_section(bdb, BDB_LVDS_BACKLIGHT);
+	if (!backlight_data)
+		return;
+
+	if (backlight_data->entry_size != sizeof(backlight_data->data[0])) {
+		DRM_DEBUG_KMS("Unsupported backlight data entry size %u\n",
+			      backlight_data->entry_size);
+		return;
+	}
+
+	entry = &backlight_data->data[panel_type];
+
+	dev_priv->vbt.backlight.present = entry->type == BDB_BACKLIGHT_TYPE_PWM;
+	if (!dev_priv->vbt.backlight.present) {
+		DRM_DEBUG_KMS("PWM backlight not present in VBT (type %u)\n",
+			      entry->type);
+		return;
+	}
+
+	dev_priv->vbt.backlight.pwm_freq_hz = entry->pwm_freq_hz;
+	dev_priv->vbt.backlight.active_low_pwm = entry->active_low_pwm;
+	DRM_DEBUG_KMS("VBT backlight PWM modulation frequency %u Hz, "
+		      "active %s, min brightness %u, level %u\n",
+		      dev_priv->vbt.backlight.pwm_freq_hz,
+		      dev_priv->vbt.backlight.active_low_pwm ? "low" : "high",
+		      entry->min_brightness,
+		      backlight_data->level[panel_type]);
+}
+
 /* Try to find sdvo panel data */
 static void
 parse_sdvo_panel_data(struct drm_i915_private *dev_priv,
@@ -290,7 +353,7 @@ parse_sdvo_panel_data(struct drm_i915_private *dev_priv,
 	struct drm_display_mode *panel_fixed_mode;
 	int index;
 
-	index = i915_vbt_sdvo_panel_type;
+	index = i915.vbt_sdvo_panel_type;
 	if (index == -2) {
 		DRM_DEBUG_KMS("Ignore SDVO panel mode from BIOS VBT tables.\n");
 		return;
@@ -327,12 +390,12 @@ static int intel_bios_ssc_frequency(struct drm_device *dev,
 {
 	switch (INTEL_INFO(dev)->gen) {
 	case 2:
-		return alternate ? 66 : 48;
+		return alternate ? 66667 : 48000;
 	case 3:
 	case 4:
-		return alternate ? 100 : 96;
+		return alternate ? 100000 : 96000;
 	default:
-		return alternate ? 100 : 120;
+		return alternate ? 100000 : 120000;
 	}
 }
 
@@ -488,6 +551,16 @@ parse_driver_features(struct drm_i915_private *dev_priv,
 
 	if (driver->dual_frequency)
 		dev_priv->render_reclock_avail = true;
+
+	DRM_DEBUG_KMS("DRRS State Enabled:%d\n", driver->drrs_enabled);
+	/*
+	 * If DRRS is not supported, drrs_type has to be set to 0.
+	 * This is because, VBT is configured in such a way that
+	 * static DRRS is 0 and DRRS not supported is represented by
+	 * driver->drrs_enabled=false
+	 */
+	if (!driver->drrs_enabled)
+		dev_priv->vbt.drrs_type = DRRS_NOT_SUPPORTED;
 }
 
 static void
@@ -522,63 +595,289 @@ parse_edp(struct drm_i915_private *dev_priv, struct bdb_header *bdb)
 
 	dev_priv->vbt.edp_pps = *edp_pps;
 
-	dev_priv->vbt.edp_rate = edp_link_params->rate ? DP_LINK_BW_2_7 :
-		DP_LINK_BW_1_62;
+	switch (edp_link_params->rate) {
+	case EDP_RATE_1_62:
+		dev_priv->vbt.edp_rate = DP_LINK_BW_1_62;
+		break;
+	case EDP_RATE_2_7:
+		dev_priv->vbt.edp_rate = DP_LINK_BW_2_7;
+		break;
+	default:
+		DRM_DEBUG_KMS("VBT has unknown eDP link rate value %u\n",
+			      edp_link_params->rate);
+		break;
+	}
+
 	switch (edp_link_params->lanes) {
-	case 0:
+	case EDP_LANE_1:
 		dev_priv->vbt.edp_lanes = 1;
 		break;
-	case 1:
+	case EDP_LANE_2:
 		dev_priv->vbt.edp_lanes = 2;
 		break;
-	case 3:
-	default:
+	case EDP_LANE_4:
 		dev_priv->vbt.edp_lanes = 4;
 		break;
+	default:
+		DRM_DEBUG_KMS("VBT has unknown eDP lane count value %u\n",
+			      edp_link_params->lanes);
+		break;
 	}
+
 	switch (edp_link_params->preemphasis) {
-	case 0:
+	case EDP_PREEMPHASIS_NONE:
 		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_0;
 		break;
-	case 1:
+	case EDP_PREEMPHASIS_3_5dB:
 		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_3_5;
 		break;
-	case 2:
+	case EDP_PREEMPHASIS_6dB:
 		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_6;
 		break;
-	case 3:
+	case EDP_PREEMPHASIS_9_5dB:
 		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_9_5;
 		break;
+	default:
+		DRM_DEBUG_KMS("VBT has unknown eDP pre-emphasis value %u\n",
+			      edp_link_params->preemphasis);
+		break;
 	}
+
 	switch (edp_link_params->vswing) {
-	case 0:
+	case EDP_VSWING_0_4V:
 		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_400;
 		break;
-	case 1:
+	case EDP_VSWING_0_6V:
 		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_600;
 		break;
-	case 2:
+	case EDP_VSWING_0_8V:
 		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_800;
 		break;
-	case 3:
+	case EDP_VSWING_1_2V:
 		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_1200;
 		break;
+	default:
+		DRM_DEBUG_KMS("VBT has unknown eDP voltage swing value %u\n",
+			      edp_link_params->vswing);
+		break;
 	}
 }
 
+static u8 *goto_next_sequence(u8 *data, int *size)
+{
+	u16 len;
+	int tmp = *size;
+
+	if (--tmp < 0)
+		return NULL;
+
+	/* goto first element */
+	data++;
+	while (1) {
+		switch (*data) {
+		case MIPI_SEQ_ELEM_SEND_PKT:
+			/*
+			 * skip by this element payload size
+			 * skip elem id, command flag and data type
+			 */
+			tmp -= 5;
+			if (tmp < 0)
+				return NULL;
+
+			data += 3;
+			len = *((u16 *)data);
+
+			tmp -= len;
+			if (tmp < 0)
+				return NULL;
+
+			/* skip by len */
+			data = data + 2 + len;
+			break;
+		case MIPI_SEQ_ELEM_DELAY:
+			/* skip by elem id, and delay is 4 bytes */
+			tmp -= 5;
+			if (tmp < 0)
+				return NULL;
+
+			data += 5;
+			break;
+		case MIPI_SEQ_ELEM_GPIO:
+			tmp -= 3;
+			if (tmp < 0)
+				return NULL;
+
+			data += 3;
+			break;
+		default:
+			DRM_ERROR("Unknown element\n");
+			return NULL;
+		}
+
+		/* end of sequence ? */
+		if (*data == 0)
+			break;
+	}
+
+	/* goto next sequence or end of block byte */
+	if (--tmp < 0)
+		return NULL;
+
+	data++;
+
+	/* update amount of data left for the sequence block to be parsed */
+	*size = tmp;
+	return data;
+}
+
 static void
 parse_mipi(struct drm_i915_private *dev_priv, struct bdb_header *bdb)
 {
-	struct bdb_mipi *mipi;
+	struct bdb_mipi_config *start;
+	struct bdb_mipi_sequence *sequence;
+	struct mipi_config *config;
+	struct mipi_pps_data *pps;
+	u8 *data, *seq_data;
+	int i, panel_id, seq_size;
+	u16 block_size;
+
+	/* parse MIPI blocks only if LFP type is MIPI */
+	if (!dev_priv->vbt.has_mipi)
+		return;
+
+	/* Initialize this to undefined indicating no generic MIPI support */
+	dev_priv->vbt.dsi.panel_id = MIPI_DSI_UNDEFINED_PANEL_ID;
+
+	/* Block #40 is already parsed and panel_fixed_mode is
+	 * stored in dev_priv->lfp_lvds_vbt_mode
+	 * resuse this when needed
+	 */
+
+	/* Parse #52 for panel index used from panel_type already
+	 * parsed
+	 */
+	start = find_section(bdb, BDB_MIPI_CONFIG);
+	if (!start) {
+		DRM_DEBUG_KMS("No MIPI config BDB found");
+		return;
+	}
+
+	DRM_DEBUG_DRIVER("Found MIPI Config block, panel index = %d\n",
+								panel_type);
+
+	/*
+	 * get hold of the correct configuration block and pps data as per
+	 * the panel_type as index
+	 */
+	config = &start->config[panel_type];
+	pps = &start->pps[panel_type];
+
+	/* store as of now full data. Trim when we realise all is not needed */
+	dev_priv->vbt.dsi.config = kmemdup(config, sizeof(struct mipi_config), GFP_KERNEL);
+	if (!dev_priv->vbt.dsi.config)
+		return;
+
+	dev_priv->vbt.dsi.pps = kmemdup(pps, sizeof(struct mipi_pps_data), GFP_KERNEL);
+	if (!dev_priv->vbt.dsi.pps) {
+		kfree(dev_priv->vbt.dsi.config);
+		return;
+	}
+
+	/* We have mandatory mipi config blocks. Initialize as generic panel */
+	dev_priv->vbt.dsi.panel_id = MIPI_DSI_GENERIC_PANEL_ID;
+
+	/* Check if we have sequence block as well */
+	sequence = find_section(bdb, BDB_MIPI_SEQUENCE);
+	if (!sequence) {
+		DRM_DEBUG_KMS("No MIPI Sequence found, parsing complete\n");
+		return;
+	}
+
+	DRM_DEBUG_DRIVER("Found MIPI sequence block\n");
+
+	block_size = get_blocksize(sequence);
+
+	/*
+	 * parse the sequence block for individual sequences
+	 */
+	dev_priv->vbt.dsi.seq_version = sequence->version;
+
+	seq_data = &sequence->data[0];
+
+	/*
+	 * sequence block is variable length and hence we need to parse and
+	 * get the sequence data for specific panel id
+	 */
+	for (i = 0; i < MAX_MIPI_CONFIGURATIONS; i++) {
+		panel_id = *seq_data;
+		seq_size = *((u16 *) (seq_data + 1));
+		if (panel_id == panel_type)
+			break;
+
+		/* skip the sequence including seq header of 3 bytes */
+		seq_data = seq_data + 3 + seq_size;
+		if ((seq_data - &sequence->data[0]) > block_size) {
+			DRM_ERROR("Sequence start is beyond sequence block size, corrupted sequence block\n");
+			return;
+		}
+	}
+
+	if (i == MAX_MIPI_CONFIGURATIONS) {
+		DRM_ERROR("Sequence block detected but no valid configuration\n");
+		return;
+	}
+
+	/* check if found sequence is completely within the sequence block
+	 * just being paranoid */
+	if (seq_size > block_size) {
+		DRM_ERROR("Corrupted sequence/size, bailing out\n");
+		return;
+	}
 
-	mipi = find_section(bdb, BDB_MIPI);
-	if (!mipi) {
-		DRM_DEBUG_KMS("No MIPI BDB found");
+	/* skip the panel id(1 byte) and seq size(2 bytes) */
+	dev_priv->vbt.dsi.data = kmemdup(seq_data + 3, seq_size, GFP_KERNEL);
+	if (!dev_priv->vbt.dsi.data)
 		return;
+
+	/*
+	 * loop into the sequence data and split into multiple sequneces
+	 * There are only 5 types of sequences as of now
+	 */
+	data = dev_priv->vbt.dsi.data;
+	dev_priv->vbt.dsi.size = seq_size;
+
+	/* two consecutive 0x00 indicate end of all sequences */
+	while (1) {
+		int seq_id = *data;
+		if (MIPI_SEQ_MAX > seq_id && seq_id > MIPI_SEQ_UNDEFINED) {
+			dev_priv->vbt.dsi.sequence[seq_id] = data;
+			DRM_DEBUG_DRIVER("Found mipi sequence - %d\n", seq_id);
+		} else {
+			DRM_ERROR("undefined sequence\n");
+			goto err;
+		}
+
+		/* partial parsing to skip elements */
+		data = goto_next_sequence(data, &seq_size);
+
+		if (data == NULL) {
+			DRM_ERROR("Sequence elements going beyond block itself. Sequence block parsing failed\n");
+			goto err;
+		}
+
+		if (*data == 0)
+			break; /* end of sequence reached */
 	}
 
-	/* XXX: add more info */
-	dev_priv->vbt.dsi.panel_id = mipi->panel_id;
+	DRM_DEBUG_DRIVER("MIPI related vbt parsing complete\n");
+	return;
+err:
+	kfree(dev_priv->vbt.dsi.data);
+	dev_priv->vbt.dsi.data = NULL;
+
+	/* error during parsing so set all pointers to null
+	 * because of partial parsing */
+	memset(dev_priv->vbt.dsi.sequence, 0, MIPI_SEQ_MAX);
 }
 
 static void parse_ddi_port(struct drm_i915_private *dev_priv, enum port port,
@@ -761,6 +1060,15 @@ parse_device_mapping(struct drm_i915_private *dev_priv,
 			/* skip the device block if device type is invalid */
 			continue;
 		}
+
+		if (p_child->common.dvo_port >= DVO_PORT_MIPIA
+		    && p_child->common.dvo_port <= DVO_PORT_MIPID
+		    &&p_child->common.device_type & DEVICE_TYPE_MIPI_OUTPUT) {
+			DRM_DEBUG_KMS("Found MIPI as LFP\n");
+			dev_priv->vbt.has_mipi = 1;
+			dev_priv->vbt.dsi.port = p_child->common.dvo_port;
+		}
+
 		child_dev_ptr = dev_priv->vbt.child_dev + count;
 		count++;
 		memcpy((void *)child_dev_ptr, (void *)p_child,
@@ -777,6 +1085,9 @@ init_vbt_defaults(struct drm_i915_private *dev_priv)
 
 	dev_priv->vbt.crt_ddc_pin = GMBUS_PORT_VGADDC;
 
+	/* Default to having backlight */
+	dev_priv->vbt.backlight.present = true;
+
 	/* LFP panel data */
 	dev_priv->vbt.lvds_dither = 1;
 	dev_priv->vbt.lvds_vbt = 0;
@@ -796,7 +1107,7 @@ init_vbt_defaults(struct drm_i915_private *dev_priv)
 	 */
 	dev_priv->vbt.lvds_ssc_freq = intel_bios_ssc_frequency(dev,
 			!HAS_PCH_SPLIT(dev));
-	DRM_DEBUG_KMS("Set default to SSC at %dMHz\n", dev_priv->vbt.lvds_ssc_freq);
+	DRM_DEBUG_KMS("Set default to SSC at %d kHz\n", dev_priv->vbt.lvds_ssc_freq);
 
 	for (port = PORT_A; port < I915_MAX_PORTS; port++) {
 		struct ddi_vbt_port_info *info =
@@ -831,6 +1142,46 @@ static const struct dmi_system_id intel_no_opregion_vbt[] = {
 	{ }
 };
 
+static struct bdb_header *validate_vbt(char *base, size_t size,
+				       struct vbt_header *vbt,
+				       const char *source)
+{
+	size_t offset;
+	struct bdb_header *bdb;
+
+	if (vbt == NULL) {
+		DRM_DEBUG_DRIVER("VBT signature missing\n");
+		return NULL;
+	}
+
+	offset = (char *)vbt - base;
+	if (offset + sizeof(struct vbt_header) > size) {
+		DRM_DEBUG_DRIVER("VBT header incomplete\n");
+		return NULL;
+	}
+
+	if (memcmp(vbt->signature, "$VBT", 4)) {
+		DRM_DEBUG_DRIVER("VBT invalid signature\n");
+		return NULL;
+	}
+
+	offset += vbt->bdb_offset;
+	if (offset + sizeof(struct bdb_header) > size) {
+		DRM_DEBUG_DRIVER("BDB header incomplete\n");
+		return NULL;
+	}
+
+	bdb = (struct bdb_header *)(base + offset);
+	if (offset + bdb->bdb_size > size) {
+		DRM_DEBUG_DRIVER("BDB incomplete\n");
+		return NULL;
+	}
+
+	DRM_DEBUG_KMS("Using VBT from %s: %20s\n",
+		      source, vbt->signature);
+	return bdb;
+}
+
 /**
  * intel_parse_bios - find VBT and initialize settings from the BIOS
  * @dev: DRM device
@@ -854,20 +1205,13 @@ intel_parse_bios(struct drm_device *dev)
 	init_vbt_defaults(dev_priv);
 
 	/* XXX Should this validation be moved to intel_opregion.c? */
-	if (!dmi_check_system(intel_no_opregion_vbt) && dev_priv->opregion.vbt) {
-		struct vbt_header *vbt = dev_priv->opregion.vbt;
-		if (memcmp(vbt->signature, "$VBT", 4) == 0) {
-			DRM_DEBUG_KMS("Using VBT from OpRegion: %20s\n",
-					 vbt->signature);
-			bdb = (struct bdb_header *)((char *)vbt + vbt->bdb_offset);
-		} else
-			dev_priv->opregion.vbt = NULL;
-	}
+	if (!dmi_check_system(intel_no_opregion_vbt) && dev_priv->opregion.vbt)
+		bdb = validate_vbt((char *)dev_priv->opregion.header, OPREGION_SIZE,
+				   (struct vbt_header *)dev_priv->opregion.vbt,
+				   "OpRegion");
 
 	if (bdb == NULL) {
-		struct vbt_header *vbt = NULL;
-		size_t size;
-		int i;
+		size_t i, size;
 
 		bios = pci_map_rom(pdev, &size);
 		if (!bios)
@@ -875,25 +1219,25 @@ intel_parse_bios(struct drm_device *dev)
 
 		/* Scour memory looking for the VBT signature */
 		for (i = 0; i + 4 < size; i++) {
-			if (!memcmp(bios + i, "$VBT", 4)) {
-				vbt = (struct vbt_header *)(bios + i);
+			if (memcmp(bios + i, "$VBT", 4) == 0) {
+				bdb = validate_vbt(bios, size,
+						   (struct vbt_header *)(bios + i),
+						   "PCI ROM");
 				break;
 			}
 		}
 
-		if (!vbt) {
-			DRM_DEBUG_DRIVER("VBT signature missing\n");
+		if (!bdb) {
 			pci_unmap_rom(pdev, bios);
 			return -1;
 		}
-
-		bdb = (struct bdb_header *)(bios + i + vbt->bdb_offset);
 	}
 
 	/* Grab useful general definitions */
 	parse_general_features(dev_priv, bdb);
 	parse_general_definitions(dev_priv, bdb);
 	parse_lfp_panel_data(dev_priv, bdb);
+	parse_lfp_backlight(dev_priv, bdb);
 	parse_sdvo_panel_data(dev_priv, bdb);
 	parse_sdvo_device_mapping(dev_priv, bdb);
 	parse_device_mapping(dev_priv, bdb);
diff --git a/drivers/gpu/drm/i915/intel_bios.h b/drivers/gpu/drm/i915/intel_bios.h
index f580a2b0ddd..b9866779633 100644
--- a/drivers/gpu/drm/i915/intel_bios.h
+++ b/drivers/gpu/drm/i915/intel_bios.h
@@ -39,7 +39,7 @@ struct vbt_header {
 	u8 reserved0;
 	u32 bdb_offset;			/**< from beginning of VBT */
 	u32 aim_offset[4];		/**< from beginning of VBT */
-} __attribute__((packed));
+} __packed;
 
 struct bdb_header {
 	u8 signature[16];		/**< Always 'BIOS_DATA_BLOCK' */
@@ -65,7 +65,7 @@ struct vbios_data {
 	u8 rsvd4; /* popup memory size */
 	u8 resize_pci_bios;
 	u8 rsvd5; /* is crt already on ddc2 */
-} __attribute__((packed));
+} __packed;
 
 /*
  * There are several types of BIOS data blocks (BDBs), each block has
@@ -104,7 +104,8 @@ struct vbios_data {
 #define BDB_LVDS_LFP_DATA	 42
 #define BDB_LVDS_BACKLIGHT	 43
 #define BDB_LVDS_POWER		 44
-#define BDB_MIPI		 50
+#define BDB_MIPI_CONFIG		 52
+#define BDB_MIPI_SEQUENCE	 53
 #define BDB_SKIP		254 /* VBIOS private block, ignore */
 
 struct bdb_general_features {
@@ -142,7 +143,7 @@ struct bdb_general_features {
 	u8 dp_ssc_enb:1;	/* PCH attached eDP supports SSC */
 	u8 dp_ssc_freq:1;	/* SSC freq for PCH attached eDP */
 	u8 rsvd11:3; /* finish byte */
-} __attribute__((packed));
+} __packed;
 
 /* pre-915 */
 #define GPIO_PIN_DVI_LVDS	0x03 /* "DVI/LVDS DDC GPIO pins" */
@@ -225,7 +226,7 @@ struct old_child_dev_config {
 	u8  dvo2_wiring;
 	u16 extended_type;
 	u8  dvo_function;
-} __attribute__((packed));
+} __packed;
 
 /* This one contains field offsets that are known to be common for all BDB
  * versions. Notice that the meaning of the contents contents may still change,
@@ -238,7 +239,7 @@ struct common_child_dev_config {
 	u8 not_common2[2];
 	u8 ddc_pin;
 	u16 edid_ptr;
-} __attribute__((packed));
+} __packed;
 
 /* This field changes depending on the BDB version, so the most reliable way to
  * read it is by checking the BDB version and reading the raw pointer. */
@@ -279,7 +280,10 @@ struct bdb_general_definitions {
 	 *	     sizeof(child_device_config);
 	 */
 	union child_device_config devices[0];
-} __attribute__((packed));
+} __packed;
+
+/* Mask for DRRS / Panel Channel / SSC / BLT control bits extraction */
+#define MODE_MASK		0x3
 
 struct bdb_lvds_options {
 	u8 panel_type;
@@ -293,7 +297,19 @@ struct bdb_lvds_options {
 	u8 lvds_edid:1;
 	u8 rsvd2:1;
 	u8 rsvd4;
-} __attribute__((packed));
+	/* LVDS Panel channel bits stored here */
+	u32 lvds_panel_channel_bits;
+	/* LVDS SSC (Spread Spectrum Clock) bits stored here. */
+	u16 ssc_bits;
+	u16 ssc_freq;
+	u16 ssc_ddt;
+	/* Panel color depth defined here */
+	u16 panel_color_depth;
+	/* LVDS panel type bits stored here */
+	u32 dps_panel_type_bits;
+	/* LVDS backlight control type bits stored here */
+	u32 blt_control_type_bits;
+} __packed;
 
 /* LFP pointer table contains entries to the struct below */
 struct bdb_lvds_lfp_data_ptr {
@@ -303,12 +319,12 @@ struct bdb_lvds_lfp_data_ptr {
 	u8 dvo_table_size;
 	u16 panel_pnp_id_offset;
 	u8 pnp_table_size;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_lfp_data_ptrs {
 	u8 lvds_entries; /* followed by one or more lvds_data_ptr structs */
 	struct bdb_lvds_lfp_data_ptr ptr[16];
-} __attribute__((packed));
+} __packed;
 
 /* LFP data has 3 blocks per entry */
 struct lvds_fp_timing {
@@ -325,7 +341,7 @@ struct lvds_fp_timing {
 	u32 pfit_reg;
 	u32 pfit_reg_val;
 	u16 terminator;
-} __attribute__((packed));
+} __packed;
 
 struct lvds_dvo_timing {
 	u16 clock;		/**< In 10khz */
@@ -353,7 +369,7 @@ struct lvds_dvo_timing {
 	u8 vsync_positive:1;
 	u8 hsync_positive:1;
 	u8 rsvd2:1;
-} __attribute__((packed));
+} __packed;
 
 struct lvds_pnp_id {
 	u16 mfg_name;
@@ -361,17 +377,36 @@ struct lvds_pnp_id {
 	u32 serial;
 	u8 mfg_week;
 	u8 mfg_year;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_lfp_data_entry {
 	struct lvds_fp_timing fp_timing;
 	struct lvds_dvo_timing dvo_timing;
 	struct lvds_pnp_id pnp_id;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_lfp_data {
 	struct bdb_lvds_lfp_data_entry data[16];
-} __attribute__((packed));
+} __packed;
+
+#define BDB_BACKLIGHT_TYPE_NONE	0
+#define BDB_BACKLIGHT_TYPE_PWM	2
+
+struct bdb_lfp_backlight_data_entry {
+	u8 type:2;
+	u8 active_low_pwm:1;
+	u8 obsolete1:5;
+	u16 pwm_freq_hz;
+	u8 min_brightness;
+	u8 obsolete2;
+	u8 obsolete3;
+} __packed;
+
+struct bdb_lfp_backlight_data {
+	u8 entry_size;
+	struct bdb_lfp_backlight_data_entry data[16];
+	u8 level[16];
+} __packed;
 
 struct aimdb_header {
 	char signature[16];
@@ -379,12 +414,12 @@ struct aimdb_header {
 	u16 aimdb_version;
 	u16 aimdb_header_size;
 	u16 aimdb_size;
-} __attribute__((packed));
+} __packed;
 
 struct aimdb_block {
 	u8 aimdb_id;
 	u16 aimdb_size;
-} __attribute__((packed));
+} __packed;
 
 struct vch_panel_data {
 	u16 fp_timing_offset;
@@ -395,12 +430,12 @@ struct vch_panel_data {
 	u8 text_fitting_size;
 	u16 graphics_fitting_offset;
 	u8 graphics_fitting_size;
-} __attribute__((packed));
+} __packed;
 
 struct vch_bdb_22 {
 	struct aimdb_block aimdb_block;
 	struct vch_panel_data panels[16];
-} __attribute__((packed));
+} __packed;
 
 struct bdb_sdvo_lvds_options {
 	u8 panel_backlight;
@@ -416,7 +451,7 @@ struct bdb_sdvo_lvds_options {
 	u8 panel_misc_bits_2;
 	u8 panel_misc_bits_3;
 	u8 panel_misc_bits_4;
-} __attribute__((packed));
+} __packed;
 
 
 #define BDB_DRIVER_FEATURE_NO_LVDS		0
@@ -462,7 +497,21 @@ struct bdb_driver_features {
 
 	u8 hdmi_termination;
 	u8 custom_vbt_version;
-} __attribute__((packed));
+	/* Driver features data block */
+	u16 rmpm_enabled:1;
+	u16 s2ddt_enabled:1;
+	u16 dpst_enabled:1;
+	u16 bltclt_enabled:1;
+	u16 adb_enabled:1;
+	u16 drrs_enabled:1;
+	u16 grs_enabled:1;
+	u16 gpmt_enabled:1;
+	u16 tbt_enabled:1;
+	u16 psr_enabled:1;
+	u16 ips_enabled:1;
+	u16 reserved3:4;
+	u16 pc_feature_valid:1;
+} __packed;
 
 #define EDP_18BPP	0
 #define EDP_24BPP	1
@@ -487,14 +536,14 @@ struct edp_power_seq {
 	u16 t9;
 	u16 t10;
 	u16 t11_t12;
-} __attribute__ ((packed));
+} __packed;
 
 struct edp_link_params {
 	u8 rate:4;
 	u8 lanes:4;
 	u8 preemphasis:4;
 	u8 vswing:4;
-} __attribute__ ((packed));
+} __packed;
 
 struct bdb_edp {
 	struct edp_power_seq power_seqs[16];
@@ -505,7 +554,7 @@ struct bdb_edp {
 	/* ith bit indicates enabled/disabled for (i+1)th panel */
 	u16 edp_s3d_feature;
 	u16 edp_t3_optimization;
-} __attribute__ ((packed));
+} __packed;
 
 void intel_setup_bios(struct drm_device *dev);
 int intel_parse_bios(struct drm_device *dev);
@@ -694,45 +743,195 @@ int intel_parse_bios(struct drm_device *dev);
 #define DVO_PORT_DPC	8
 #define DVO_PORT_DPD	9
 #define DVO_PORT_DPA	10
+#define DVO_PORT_MIPIA	21
+#define DVO_PORT_MIPIB	22
+#define DVO_PORT_MIPIC	23
+#define DVO_PORT_MIPID	24
+
+/* Block 52 contains MIPI Panel info
+ * 6 such enteries will there. Index into correct
+ * entery is based on the panel_index in #40 LFP
+ */
+#define MAX_MIPI_CONFIGURATIONS	6
 
-/* MIPI DSI panel info */
-struct bdb_mipi {
+#define MIPI_DSI_UNDEFINED_PANEL_ID	0
+#define MIPI_DSI_GENERIC_PANEL_ID	1
+
+struct mipi_config {
 	u16 panel_id;
-	u16 bridge_revision;
 
-	/* General params */
-	u32 dithering:1;
-	u32 bpp_pixel_format:1;
+	/* General Params */
+	u32 enable_dithering:1;
 	u32 rsvd1:1;
-	u32 dphy_valid:1;
-	u32 resvd2:28;
-
-	u16 port_info;
-	u16 rsvd3:2;
-	u16 num_lanes:2;
-	u16 rsvd4:12;
-
-	/* DSI config */
-	u16 virt_ch_num:2;
-	u16 vtm:2;
-	u16 rsvd5:12;
-
-	u32 dsi_clock;
+	u32 is_bridge:1;
+
+	u32 panel_arch_type:2;
+	u32 is_cmd_mode:1;
+
+#define NON_BURST_SYNC_PULSE	0x1
+#define NON_BURST_SYNC_EVENTS	0x2
+#define BURST_MODE		0x3
+	u32 video_transfer_mode:2;
+
+	u32 cabc_supported:1;
+	u32 pwm_blc:1;
+
+	/* Bit 13:10 */
+#define PIXEL_FORMAT_RGB565			0x1
+#define PIXEL_FORMAT_RGB666			0x2
+#define PIXEL_FORMAT_RGB666_LOOSELY_PACKED	0x3
+#define PIXEL_FORMAT_RGB888			0x4
+	u32 videomode_color_format:4;
+
+	/* Bit 15:14 */
+#define ENABLE_ROTATION_0	0x0
+#define ENABLE_ROTATION_90	0x1
+#define ENABLE_ROTATION_180	0x2
+#define ENABLE_ROTATION_270	0x3
+	u32 rotation:2;
+	u32 bta_enabled:1;
+	u32 rsvd2:15;
+
+	/* 2 byte Port Description */
+#define DUAL_LINK_NOT_SUPPORTED	0
+#define DUAL_LINK_FRONT_BACK	1
+#define DUAL_LINK_PIXEL_ALT	2
+	u16 dual_link:2;
+	u16 lane_cnt:2;
+	u16 rsvd3:12;
+
+	u16 rsvd4;
+
+	u8 rsvd5[5];
+	u32 dsi_ddr_clk;
 	u32 bridge_ref_clk;
-	u16 rsvd_pwr;
 
-	/* Dphy Params */
-	u32 prepare_cnt:5;
-	u32 rsvd6:3;
+#define  BYTE_CLK_SEL_20MHZ		0
+#define  BYTE_CLK_SEL_10MHZ		1
+#define  BYTE_CLK_SEL_5MHZ		2
+	u8 byte_clk_sel:2;
+
+	u8 rsvd6:6;
+
+	/* DPHY Flags */
+	u16 dphy_param_valid:1;
+	u16 eot_pkt_disabled:1;
+	u16 enable_clk_stop:1;
+	u16 rsvd7:13;
+
+	u32 hs_tx_timeout;
+	u32 lp_rx_timeout;
+	u32 turn_around_timeout;
+	u32 device_reset_timer;
+	u32 master_init_timer;
+	u32 dbi_bw_timer;
+	u32 lp_byte_clk_val;
+
+	/*  4 byte Dphy Params */
+	u32 prepare_cnt:6;
+	u32 rsvd8:2;
 	u32 clk_zero_cnt:8;
 	u32 trail_cnt:5;
-	u32 rsvd7:3;
+	u32 rsvd9:3;
 	u32 exit_zero_cnt:6;
-	u32 rsvd8:2;
+	u32 rsvd10:2;
 
-	u32 hl_switch_cnt;
-	u32 lp_byte_clk;
 	u32 clk_lane_switch_cnt;
-} __attribute__((packed));
+	u32 hl_switch_cnt;
+
+	u32 rsvd11[6];
+
+	/* timings based on dphy spec */
+	u8 tclk_miss;
+	u8 tclk_post;
+	u8 rsvd12;
+	u8 tclk_pre;
+	u8 tclk_prepare;
+	u8 tclk_settle;
+	u8 tclk_term_enable;
+	u8 tclk_trail;
+	u16 tclk_prepare_clkzero;
+	u8 rsvd13;
+	u8 td_term_enable;
+	u8 teot;
+	u8 ths_exit;
+	u8 ths_prepare;
+	u16 ths_prepare_hszero;
+	u8 rsvd14;
+	u8 ths_settle;
+	u8 ths_skip;
+	u8 ths_trail;
+	u8 tinit;
+	u8 tlpx;
+	u8 rsvd15[3];
+
+	/* GPIOs */
+	u8 panel_enable;
+	u8 bl_enable;
+	u8 pwm_enable;
+	u8 reset_r_n;
+	u8 pwr_down_r;
+	u8 stdby_r_n;
+
+} __packed;
+
+/* Block 52 contains MIPI configuration block
+ * 6 * bdb_mipi_config, followed by 6 pps data
+ * block below
+ *
+ * all delays has a unit of 100us
+ */
+struct mipi_pps_data {
+	u16 panel_on_delay;
+	u16 bl_enable_delay;
+	u16 bl_disable_delay;
+	u16 panel_off_delay;
+	u16 panel_power_cycle_delay;
+};
+
+struct bdb_mipi_config {
+	struct mipi_config config[MAX_MIPI_CONFIGURATIONS];
+	struct mipi_pps_data pps[MAX_MIPI_CONFIGURATIONS];
+};
+
+/* Block 53 contains MIPI sequences as needed by the panel
+ * for enabling it. This block can be variable in size and
+ * can be maximum of 6 blocks
+ */
+struct bdb_mipi_sequence {
+	u8 version;
+	u8 data[0];
+};
+
+/* MIPI Sequnece Block definitions */
+enum mipi_seq {
+	MIPI_SEQ_UNDEFINED = 0,
+	MIPI_SEQ_ASSERT_RESET,
+	MIPI_SEQ_INIT_OTP,
+	MIPI_SEQ_DISPLAY_ON,
+	MIPI_SEQ_DISPLAY_OFF,
+	MIPI_SEQ_DEASSERT_RESET,
+	MIPI_SEQ_MAX
+};
+
+enum mipi_seq_element {
+	MIPI_SEQ_ELEM_UNDEFINED = 0,
+	MIPI_SEQ_ELEM_SEND_PKT,
+	MIPI_SEQ_ELEM_DELAY,
+	MIPI_SEQ_ELEM_GPIO,
+	MIPI_SEQ_ELEM_STATUS,
+	MIPI_SEQ_ELEM_MAX
+};
+
+enum mipi_gpio_pin_index {
+	MIPI_GPIO_UNDEFINED = 0,
+	MIPI_GPIO_PANEL_ENABLE,
+	MIPI_GPIO_BL_ENABLE,
+	MIPI_GPIO_PWM_ENABLE,
+	MIPI_GPIO_RESET_N,
+	MIPI_GPIO_PWR_DOWN_R,
+	MIPI_GPIO_STDBY_RST_N,
+	MIPI_GPIO_MAX
+};
 
 #endif /* _I830_BIOS_H_ */
diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index b5b1b9b23ad..5a045d3bd77 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -68,8 +68,13 @@ static bool intel_crt_get_hw_state(struct intel_encoder *encoder,
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crt *crt = intel_encoder_to_crt(encoder);
+	enum intel_display_power_domain power_domain;
 	u32 tmp;
 
+	power_domain = intel_display_port_power_domain(encoder);
+	if (!intel_display_power_enabled(dev_priv, power_domain))
+		return false;
+
 	tmp = I915_READ(crt->adpa_reg);
 
 	if (!(tmp & ADPA_DAC_ENABLE))
@@ -139,28 +144,49 @@ static void intel_crt_set_dpms(struct intel_encoder *encoder, int mode)
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crt *crt = intel_encoder_to_crt(encoder);
-	u32 temp;
+	struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
+	struct drm_display_mode *adjusted_mode = &crtc->config.adjusted_mode;
+	u32 adpa;
 
-	temp = I915_READ(crt->adpa_reg);
-	temp &= ~(ADPA_HSYNC_CNTL_DISABLE | ADPA_VSYNC_CNTL_DISABLE);
-	temp &= ~ADPA_DAC_ENABLE;
+	if (INTEL_INFO(dev)->gen >= 5)
+		adpa = ADPA_HOTPLUG_BITS;
+	else
+		adpa = 0;
+
+	if (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC)
+		adpa |= ADPA_HSYNC_ACTIVE_HIGH;
+	if (adjusted_mode->flags & DRM_MODE_FLAG_PVSYNC)
+		adpa |= ADPA_VSYNC_ACTIVE_HIGH;
+
+	/* For CPT allow 3 pipe config, for others just use A or B */
+	if (HAS_PCH_LPT(dev))
+		; /* Those bits don't exist here */
+	else if (HAS_PCH_CPT(dev))
+		adpa |= PORT_TRANS_SEL_CPT(crtc->pipe);
+	else if (crtc->pipe == 0)
+		adpa |= ADPA_PIPE_A_SELECT;
+	else
+		adpa |= ADPA_PIPE_B_SELECT;
+
+	if (!HAS_PCH_SPLIT(dev))
+		I915_WRITE(BCLRPAT(crtc->pipe), 0);
 
 	switch (mode) {
 	case DRM_MODE_DPMS_ON:
-		temp |= ADPA_DAC_ENABLE;
+		adpa |= ADPA_DAC_ENABLE;
 		break;
 	case DRM_MODE_DPMS_STANDBY:
-		temp |= ADPA_DAC_ENABLE | ADPA_HSYNC_CNTL_DISABLE;
+		adpa |= ADPA_DAC_ENABLE | ADPA_HSYNC_CNTL_DISABLE;
 		break;
 	case DRM_MODE_DPMS_SUSPEND:
-		temp |= ADPA_DAC_ENABLE | ADPA_VSYNC_CNTL_DISABLE;
+		adpa |= ADPA_DAC_ENABLE | ADPA_VSYNC_CNTL_DISABLE;
 		break;
 	case DRM_MODE_DPMS_OFF:
-		temp |= ADPA_HSYNC_CNTL_DISABLE | ADPA_VSYNC_CNTL_DISABLE;
+		adpa |= ADPA_HSYNC_CNTL_DISABLE | ADPA_VSYNC_CNTL_DISABLE;
 		break;
 	}
 
-	I915_WRITE(crt->adpa_reg, temp);
+	I915_WRITE(crt->adpa_reg, adpa);
 }
 
 static void intel_disable_crt(struct intel_encoder *encoder)
@@ -222,8 +248,9 @@ static void intel_crt_dpms(struct drm_connector *connector, int mode)
 	intel_modeset_check_state(connector->dev);
 }
 
-static int intel_crt_mode_valid(struct drm_connector *connector,
-				struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_crt_mode_valid(struct drm_connector *connector,
+		     struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
 
@@ -261,43 +288,11 @@ static bool intel_crt_compute_config(struct intel_encoder *encoder,
 	if (HAS_PCH_LPT(dev))
 		pipe_config->pipe_bpp = 24;
 
-	return true;
-}
-
-static void intel_crt_mode_set(struct intel_encoder *encoder)
-{
-
-	struct drm_device *dev = encoder->base.dev;
-	struct intel_crt *crt = intel_encoder_to_crt(encoder);
-	struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_display_mode *adjusted_mode = &crtc->config.adjusted_mode;
-	u32 adpa;
-
-	if (INTEL_INFO(dev)->gen >= 5)
-		adpa = ADPA_HOTPLUG_BITS;
-	else
-		adpa = 0;
-
-	if (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC)
-		adpa |= ADPA_HSYNC_ACTIVE_HIGH;
-	if (adjusted_mode->flags & DRM_MODE_FLAG_PVSYNC)
-		adpa |= ADPA_VSYNC_ACTIVE_HIGH;
-
-	/* For CPT allow 3 pipe config, for others just use A or B */
-	if (HAS_PCH_LPT(dev))
-		; /* Those bits don't exist here */
-	else if (HAS_PCH_CPT(dev))
-		adpa |= PORT_TRANS_SEL_CPT(crtc->pipe);
-	else if (crtc->pipe == 0)
-		adpa |= ADPA_PIPE_A_SELECT;
-	else
-		adpa |= ADPA_PIPE_B_SELECT;
-
-	if (!HAS_PCH_SPLIT(dev))
-		I915_WRITE(BCLRPAT(crtc->pipe), 0);
+	/* FDI must always be 2.7 GHz */
+	if (HAS_DDI(dev))
+		pipe_config->port_clock = 135000 * 2;
 
-	I915_WRITE(crt->adpa_reg, adpa);
+	return true;
 }
 
 static bool intel_ironlake_crt_detect_hotplug(struct drm_connector *connector)
@@ -629,14 +624,23 @@ static enum drm_connector_status
 intel_crt_detect(struct drm_connector *connector, bool force)
 {
 	struct drm_device *dev = connector->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crt *crt = intel_attached_crt(connector);
+	struct intel_encoder *intel_encoder = &crt->base;
+	enum intel_display_power_domain power_domain;
 	enum drm_connector_status status;
 	struct intel_load_detect_pipe tmp;
+	struct drm_modeset_acquire_ctx ctx;
+
+	intel_runtime_pm_get(dev_priv);
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s] force=%d\n",
-		      connector->base.id, drm_get_connector_name(connector),
+		      connector->base.id, connector->name,
 		      force);
 
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_get(dev_priv, power_domain);
+
 	if (I915_HAS_HOTPLUG(dev)) {
 		/* We can not rely on the HPD pin always being correctly wired
 		 * up, for example many KVM do not pass it through, and so
@@ -644,34 +648,45 @@ intel_crt_detect(struct drm_connector *connector, bool force)
 		 */
 		if (intel_crt_detect_hotplug(connector)) {
 			DRM_DEBUG_KMS("CRT detected via hotplug\n");
-			return connector_status_connected;
+			status = connector_status_connected;
+			goto out;
 		} else
 			DRM_DEBUG_KMS("CRT not detected via hotplug\n");
 	}
 
-	if (intel_crt_detect_ddc(connector))
-		return connector_status_connected;
+	if (intel_crt_detect_ddc(connector)) {
+		status = connector_status_connected;
+		goto out;
+	}
 
 	/* Load detection is broken on HPD capable machines. Whoever wants a
 	 * broken monitor (without edid) to work behind a broken kvm (that fails
 	 * to have the right resistors for HP detection) needs to fix this up.
 	 * For now just bail out. */
-	if (I915_HAS_HOTPLUG(dev))
-		return connector_status_disconnected;
+	if (I915_HAS_HOTPLUG(dev)) {
+		status = connector_status_disconnected;
+		goto out;
+	}
 
-	if (!force)
-		return connector->status;
+	if (!force) {
+		status = connector->status;
+		goto out;
+	}
 
 	/* for pre-945g platforms use load detect */
-	if (intel_get_load_detect_pipe(connector, NULL, &tmp)) {
+	if (intel_get_load_detect_pipe(connector, NULL, &tmp, &ctx)) {
 		if (intel_crt_detect_ddc(connector))
 			status = connector_status_connected;
 		else
 			status = intel_crt_load_detect(crt);
-		intel_release_load_detect_pipe(connector, &tmp);
+		intel_release_load_detect_pipe(connector, &tmp, &ctx);
 	} else
 		status = connector_status_unknown;
 
+out:
+	intel_display_power_put(dev_priv, power_domain);
+	intel_runtime_pm_put(dev_priv);
+
 	return status;
 }
 
@@ -685,17 +700,28 @@ static int intel_crt_get_modes(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crt *crt = intel_attached_crt(connector);
+	struct intel_encoder *intel_encoder = &crt->base;
+	enum intel_display_power_domain power_domain;
 	int ret;
 	struct i2c_adapter *i2c;
 
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_get(dev_priv, power_domain);
+
 	i2c = intel_gmbus_get_adapter(dev_priv, dev_priv->vbt.crt_ddc_pin);
 	ret = intel_crt_ddc_get_modes(connector, i2c);
 	if (ret || !IS_G4X(dev))
-		return ret;
+		goto out;
 
 	/* Try to probe digital port for output in DVI-I -> VGA mode. */
 	i2c = intel_gmbus_get_adapter(dev_priv, GMBUS_PORT_DPB);
-	return intel_crt_ddc_get_modes(connector, i2c);
+	ret = intel_crt_ddc_get_modes(connector, i2c);
+
+out:
+	intel_display_power_put(dev_priv, power_domain);
+
+	return ret;
 }
 
 static int intel_crt_set_property(struct drm_connector *connector,
@@ -764,6 +790,14 @@ static const struct dmi_system_id intel_no_crt[] = {
 			DMI_MATCH(DMI_PRODUCT_NAME, "ZGB"),
 		},
 	},
+	{
+		.callback = intel_no_crt_dmi_callback,
+		.ident = "DELL XPS 8700",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "XPS 8700"),
+		},
+	},
 	{ }
 };
 
@@ -799,7 +833,7 @@ void intel_crt_init(struct drm_device *dev)
 	intel_connector_attach_encoder(intel_connector, &crt->base);
 
 	crt->base.type = INTEL_OUTPUT_ANALOG;
-	crt->base.cloneable = true;
+	crt->base.cloneable = (1 << INTEL_OUTPUT_DVO) | (1 << INTEL_OUTPUT_HDMI);
 	if (IS_I830(dev))
 		crt->base.crtc_mask = (1 << 0);
 	else
@@ -819,7 +853,6 @@ void intel_crt_init(struct drm_device *dev)
 		crt->adpa_reg = ADPA;
 
 	crt->base.compute_config = intel_crt_compute_config;
-	crt->base.mode_set = intel_crt_mode_set;
 	crt->base.disable = intel_disable_crt;
 	crt->base.enable = intel_enable_crt;
 	if (I915_HAS_HOTPLUG(dev))
@@ -832,6 +865,7 @@ void intel_crt_init(struct drm_device *dev)
 		crt->base.get_hw_state = intel_crt_get_hw_state;
 	}
 	intel_connector->get_hw_state = intel_connector_get_hw_state;
+	intel_connector->unregister = intel_connector_unregister;
 
 	drm_connector_helper_add(connector, &intel_crt_connector_helper_funcs);
 
@@ -856,4 +890,6 @@ void intel_crt_init(struct drm_device *dev)
 
 		dev_priv->fdi_rx_config = I915_READ(_FDI_RXA_CTL) & fdi_config;
 	}
+
+	intel_crt_reset(connector);
 }
diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index b69dc3e66c1..b17b9c7c769 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -73,7 +73,7 @@ static const u32 hsw_ddi_translations_hdmi[] = {
 };
 
 static const u32 bdw_ddi_translations_edp[] = {
-	0x00FFFFFF, 0x00000012,		/* DP parameters */
+	0x00FFFFFF, 0x00000012,		/* eDP parameters */
 	0x00EBAFFF, 0x00020011,
 	0x00C71FFF, 0x0006000F,
 	0x00FFFFFF, 0x00020011,
@@ -364,55 +364,6 @@ void hsw_fdi_link_train(struct drm_crtc *crtc)
 	DRM_ERROR("FDI link training failed!\n");
 }
 
-static void intel_ddi_mode_set(struct intel_encoder *encoder)
-{
-	struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
-	int port = intel_ddi_get_encoder_port(encoder);
-	int pipe = crtc->pipe;
-	int type = encoder->type;
-	struct drm_display_mode *adjusted_mode = &crtc->config.adjusted_mode;
-
-	DRM_DEBUG_KMS("Preparing DDI mode on port %c, pipe %c\n",
-		      port_name(port), pipe_name(pipe));
-
-	crtc->eld_vld = false;
-	if (type == INTEL_OUTPUT_DISPLAYPORT || type == INTEL_OUTPUT_EDP) {
-		struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-		struct intel_digital_port *intel_dig_port =
-			enc_to_dig_port(&encoder->base);
-
-		intel_dp->DP = intel_dig_port->saved_port_bits |
-			       DDI_BUF_CTL_ENABLE | DDI_BUF_EMP_400MV_0DB_HSW;
-		intel_dp->DP |= DDI_PORT_WIDTH(intel_dp->lane_count);
-
-		if (intel_dp->has_audio) {
-			DRM_DEBUG_DRIVER("DP audio on pipe %c on DDI\n",
-					 pipe_name(crtc->pipe));
-
-			/* write eld */
-			DRM_DEBUG_DRIVER("DP audio: write eld information\n");
-			intel_write_eld(&encoder->base, adjusted_mode);
-		}
-	} else if (type == INTEL_OUTPUT_HDMI) {
-		struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
-
-		if (intel_hdmi->has_audio) {
-			/* Proper support for digital audio needs a new logic
-			 * and a new set of registers, so we leave it for future
-			 * patch bombing.
-			 */
-			DRM_DEBUG_DRIVER("HDMI audio on pipe %c on DDI\n",
-					 pipe_name(crtc->pipe));
-
-			/* write eld */
-			DRM_DEBUG_DRIVER("HDMI audio: write eld information\n");
-			intel_write_eld(&encoder->base, adjusted_mode);
-		}
-
-		intel_hdmi->set_infoframes(&encoder->base, adjusted_mode);
-	}
-}
-
 static struct intel_encoder *
 intel_ddi_get_crtc_encoder(struct drm_crtc *crtc)
 {
@@ -633,6 +584,97 @@ static void wrpll_update_rnp(uint64_t freq2k, unsigned budget,
 	/* Otherwise a < c && b >= d, do nothing */
 }
 
+static int intel_ddi_calc_wrpll_link(struct drm_i915_private *dev_priv,
+				     int reg)
+{
+	int refclk = LC_FREQ;
+	int n, p, r;
+	u32 wrpll;
+
+	wrpll = I915_READ(reg);
+	switch (wrpll & SPLL_PLL_REF_MASK) {
+	case SPLL_PLL_SSC:
+	case SPLL_PLL_NON_SSC:
+		/*
+		 * We could calculate spread here, but our checking
+		 * code only cares about 5% accuracy, and spread is a max of
+		 * 0.5% downspread.
+		 */
+		refclk = 135;
+		break;
+	case SPLL_PLL_LCPLL:
+		refclk = LC_FREQ;
+		break;
+	default:
+		WARN(1, "bad wrpll refclk\n");
+		return 0;
+	}
+
+	r = wrpll & WRPLL_DIVIDER_REF_MASK;
+	p = (wrpll & WRPLL_DIVIDER_POST_MASK) >> WRPLL_DIVIDER_POST_SHIFT;
+	n = (wrpll & WRPLL_DIVIDER_FB_MASK) >> WRPLL_DIVIDER_FB_SHIFT;
+
+	/* Convert to KHz, p & r have a fixed point portion */
+	return (refclk * n * 100) / (p * r);
+}
+
+static void intel_ddi_clock_get(struct intel_encoder *encoder,
+				struct intel_crtc_config *pipe_config)
+{
+	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	enum port port = intel_ddi_get_encoder_port(encoder);
+	int link_clock = 0;
+	u32 val, pll;
+
+	val = I915_READ(PORT_CLK_SEL(port));
+	switch (val & PORT_CLK_SEL_MASK) {
+	case PORT_CLK_SEL_LCPLL_810:
+		link_clock = 81000;
+		break;
+	case PORT_CLK_SEL_LCPLL_1350:
+		link_clock = 135000;
+		break;
+	case PORT_CLK_SEL_LCPLL_2700:
+		link_clock = 270000;
+		break;
+	case PORT_CLK_SEL_WRPLL1:
+		link_clock = intel_ddi_calc_wrpll_link(dev_priv, WRPLL_CTL1);
+		break;
+	case PORT_CLK_SEL_WRPLL2:
+		link_clock = intel_ddi_calc_wrpll_link(dev_priv, WRPLL_CTL2);
+		break;
+	case PORT_CLK_SEL_SPLL:
+		pll = I915_READ(SPLL_CTL) & SPLL_PLL_FREQ_MASK;
+		if (pll == SPLL_PLL_FREQ_810MHz)
+			link_clock = 81000;
+		else if (pll == SPLL_PLL_FREQ_1350MHz)
+			link_clock = 135000;
+		else if (pll == SPLL_PLL_FREQ_2700MHz)
+			link_clock = 270000;
+		else {
+			WARN(1, "bad spll freq\n");
+			return;
+		}
+		break;
+	default:
+		WARN(1, "bad port clock sel\n");
+		return;
+	}
+
+	pipe_config->port_clock = link_clock * 2;
+
+	if (pipe_config->has_pch_encoder)
+		pipe_config->adjusted_mode.crtc_clock =
+			intel_dotclock_calculate(pipe_config->port_clock,
+						 &pipe_config->fdi_m_n);
+	else if (pipe_config->has_dp_encoder)
+		pipe_config->adjusted_mode.crtc_clock =
+			intel_dotclock_calculate(pipe_config->port_clock,
+						 &pipe_config->dp_m_n);
+	else
+		pipe_config->adjusted_mode.crtc_clock = pipe_config->port_clock;
+}
+
 static void
 intel_ddi_calculate_wrpll(int clock /* in Hz */,
 			  unsigned *r2_out, unsigned *n2_out, unsigned *p_out)
@@ -696,25 +738,25 @@ intel_ddi_calculate_wrpll(int clock /* in Hz */,
 	*n2_out = best.n2;
 	*p_out = best.p;
 	*r2_out = best.r2;
-
-	DRM_DEBUG_KMS("WRPLL: %dHz refresh rate with p=%d, n2=%d r2=%d\n",
-		      clock, *p_out, *n2_out, *r2_out);
 }
 
-bool intel_ddi_pll_mode_set(struct drm_crtc *crtc)
+/*
+ * Tries to find a PLL for the CRTC. If it finds, it increases the refcount and
+ * stores it in intel_crtc->ddi_pll_sel, so other mode sets won't be able to
+ * steal the selected PLL. You need to call intel_ddi_pll_enable to actually
+ * enable the PLL.
+ */
+bool intel_ddi_pll_select(struct intel_crtc *intel_crtc)
 {
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct drm_crtc *crtc = &intel_crtc->base;
 	struct intel_encoder *intel_encoder = intel_ddi_get_crtc_encoder(crtc);
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_i915_private *dev_priv = crtc->dev->dev_private;
 	struct intel_ddi_plls *plls = &dev_priv->ddi_plls;
 	int type = intel_encoder->type;
 	enum pipe pipe = intel_crtc->pipe;
-	uint32_t reg, val;
 	int clock = intel_crtc->config.port_clock;
 
-	/* TODO: reuse PLLs when possible (compare values) */
-
 	intel_ddi_put_crtc_pll(crtc);
 
 	if (type == INTEL_OUTPUT_DISPLAYPORT || type == INTEL_OUTPUT_EDP) {
@@ -736,66 +778,145 @@ bool intel_ddi_pll_mode_set(struct drm_crtc *crtc)
 			return false;
 		}
 
-		/* We don't need to turn any PLL on because we'll use LCPLL. */
-		return true;
-
 	} else if (type == INTEL_OUTPUT_HDMI) {
+		uint32_t reg, val;
 		unsigned p, n2, r2;
 
-		if (plls->wrpll1_refcount == 0) {
+		intel_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
+
+		val = WRPLL_PLL_ENABLE | WRPLL_PLL_SELECT_LCPLL_2700 |
+		      WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) |
+		      WRPLL_DIVIDER_POST(p);
+
+		if (val == I915_READ(WRPLL_CTL1)) {
+			DRM_DEBUG_KMS("Reusing WRPLL 1 on pipe %c\n",
+				      pipe_name(pipe));
+			reg = WRPLL_CTL1;
+		} else if (val == I915_READ(WRPLL_CTL2)) {
+			DRM_DEBUG_KMS("Reusing WRPLL 2 on pipe %c\n",
+				      pipe_name(pipe));
+			reg = WRPLL_CTL2;
+		} else if (plls->wrpll1_refcount == 0) {
 			DRM_DEBUG_KMS("Using WRPLL 1 on pipe %c\n",
 				      pipe_name(pipe));
-			plls->wrpll1_refcount++;
 			reg = WRPLL_CTL1;
-			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_WRPLL1;
 		} else if (plls->wrpll2_refcount == 0) {
 			DRM_DEBUG_KMS("Using WRPLL 2 on pipe %c\n",
 				      pipe_name(pipe));
-			plls->wrpll2_refcount++;
 			reg = WRPLL_CTL2;
-			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_WRPLL2;
 		} else {
 			DRM_ERROR("No WRPLLs available!\n");
 			return false;
 		}
 
-		WARN(I915_READ(reg) & WRPLL_PLL_ENABLE,
-		     "WRPLL already enabled\n");
+		DRM_DEBUG_KMS("WRPLL: %dKHz refresh rate with p=%d, n2=%d r2=%d\n",
+			      clock, p, n2, r2);
 
-		intel_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
-
-		val = WRPLL_PLL_ENABLE | WRPLL_PLL_SELECT_LCPLL_2700 |
-		      WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) |
-		      WRPLL_DIVIDER_POST(p);
+		if (reg == WRPLL_CTL1) {
+			plls->wrpll1_refcount++;
+			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_WRPLL1;
+		} else {
+			plls->wrpll2_refcount++;
+			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_WRPLL2;
+		}
 
 	} else if (type == INTEL_OUTPUT_ANALOG) {
 		if (plls->spll_refcount == 0) {
 			DRM_DEBUG_KMS("Using SPLL on pipe %c\n",
 				      pipe_name(pipe));
 			plls->spll_refcount++;
-			reg = SPLL_CTL;
 			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_SPLL;
 		} else {
 			DRM_ERROR("SPLL already in use\n");
 			return false;
 		}
 
-		WARN(I915_READ(reg) & SPLL_PLL_ENABLE,
-		     "SPLL already enabled\n");
-
-		val = SPLL_PLL_ENABLE | SPLL_PLL_FREQ_1350MHz | SPLL_PLL_SSC;
-
 	} else {
 		WARN(1, "Invalid DDI encoder type %d\n", type);
 		return false;
 	}
 
-	I915_WRITE(reg, val);
-	udelay(20);
-
 	return true;
 }
 
+/*
+ * To be called after intel_ddi_pll_select(). That one selects the PLL to be
+ * used, this one actually enables the PLL.
+ */
+void intel_ddi_pll_enable(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_ddi_plls *plls = &dev_priv->ddi_plls;
+	int clock = crtc->config.port_clock;
+	uint32_t reg, cur_val, new_val;
+	int refcount;
+	const char *pll_name;
+	uint32_t enable_bit = (1 << 31);
+	unsigned int p, n2, r2;
+
+	BUILD_BUG_ON(enable_bit != SPLL_PLL_ENABLE);
+	BUILD_BUG_ON(enable_bit != WRPLL_PLL_ENABLE);
+
+	switch (crtc->ddi_pll_sel) {
+	case PORT_CLK_SEL_LCPLL_2700:
+	case PORT_CLK_SEL_LCPLL_1350:
+	case PORT_CLK_SEL_LCPLL_810:
+		/*
+		 * LCPLL should always be enabled at this point of the mode set
+		 * sequence, so nothing to do.
+		 */
+		return;
+
+	case PORT_CLK_SEL_SPLL:
+		pll_name = "SPLL";
+		reg = SPLL_CTL;
+		refcount = plls->spll_refcount;
+		new_val = SPLL_PLL_ENABLE | SPLL_PLL_FREQ_1350MHz |
+			  SPLL_PLL_SSC;
+		break;
+
+	case PORT_CLK_SEL_WRPLL1:
+	case PORT_CLK_SEL_WRPLL2:
+		if (crtc->ddi_pll_sel == PORT_CLK_SEL_WRPLL1) {
+			pll_name = "WRPLL1";
+			reg = WRPLL_CTL1;
+			refcount = plls->wrpll1_refcount;
+		} else {
+			pll_name = "WRPLL2";
+			reg = WRPLL_CTL2;
+			refcount = plls->wrpll2_refcount;
+		}
+
+		intel_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
+
+		new_val = WRPLL_PLL_ENABLE | WRPLL_PLL_SELECT_LCPLL_2700 |
+			  WRPLL_DIVIDER_REFERENCE(r2) |
+			  WRPLL_DIVIDER_FEEDBACK(n2) | WRPLL_DIVIDER_POST(p);
+
+		break;
+
+	case PORT_CLK_SEL_NONE:
+		WARN(1, "Bad selected pll: PORT_CLK_SEL_NONE\n");
+		return;
+	default:
+		WARN(1, "Bad selected pll: 0x%08x\n", crtc->ddi_pll_sel);
+		return;
+	}
+
+	cur_val = I915_READ(reg);
+
+	WARN(refcount < 1, "Bad %s refcount: %d\n", pll_name, refcount);
+	if (refcount == 1) {
+		WARN(cur_val & enable_bit, "%s already enabled\n", pll_name);
+		I915_WRITE(reg, new_val);
+		POSTING_READ(reg);
+		udelay(20);
+	} else {
+		WARN((cur_val & enable_bit) == 0, "%s disabled\n", pll_name);
+	}
+}
+
 void intel_ddi_set_pipe_settings(struct drm_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = crtc->dev->dev_private;
@@ -892,9 +1013,7 @@ void intel_ddi_enable_transcoder_func(struct drm_crtc *crtc)
 	}
 
 	if (type == INTEL_OUTPUT_HDMI) {
-		struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder);
-
-		if (intel_hdmi->has_hdmi_sink)
+		if (intel_crtc->config.has_hdmi_sink)
 			temp |= TRANS_DDI_MODE_SELECT_HDMI;
 		else
 			temp |= TRANS_DDI_MODE_SELECT_DVI;
@@ -938,8 +1057,13 @@ bool intel_ddi_connector_get_hw_state(struct intel_connector *intel_connector)
 	enum port port = intel_ddi_get_encoder_port(intel_encoder);
 	enum pipe pipe = 0;
 	enum transcoder cpu_transcoder;
+	enum intel_display_power_domain power_domain;
 	uint32_t tmp;
 
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	if (!intel_display_power_enabled(dev_priv, power_domain))
+		return false;
+
 	if (!intel_encoder->get_hw_state(intel_encoder, &pipe))
 		return false;
 
@@ -975,9 +1099,14 @@ bool intel_ddi_get_hw_state(struct intel_encoder *encoder,
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum port port = intel_ddi_get_encoder_port(encoder);
+	enum intel_display_power_domain power_domain;
 	u32 tmp;
 	int i;
 
+	power_domain = intel_display_port_power_domain(encoder);
+	if (!intel_display_power_enabled(dev_priv, power_domain))
+		return false;
+
 	tmp = I915_READ(DDI_BUF_CTL(port));
 
 	if (!(tmp & DDI_BUF_CTL_ENABLE))
@@ -1113,30 +1242,48 @@ void intel_ddi_disable_pipe_clock(struct intel_crtc *intel_crtc)
 static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder)
 {
 	struct drm_encoder *encoder = &intel_encoder->base;
-	struct drm_crtc *crtc = encoder->crtc;
 	struct drm_i915_private *dev_priv = encoder->dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct intel_crtc *crtc = to_intel_crtc(encoder->crtc);
 	enum port port = intel_ddi_get_encoder_port(intel_encoder);
 	int type = intel_encoder->type;
 
+	if (crtc->config.has_audio) {
+		DRM_DEBUG_DRIVER("Audio on pipe %c on DDI\n",
+				 pipe_name(crtc->pipe));
+
+		/* write eld */
+		DRM_DEBUG_DRIVER("DDI audio: write eld information\n");
+		intel_write_eld(encoder, &crtc->config.adjusted_mode);
+	}
+
 	if (type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
-		ironlake_edp_panel_vdd_on(intel_dp);
-		ironlake_edp_panel_on(intel_dp);
-		ironlake_edp_panel_vdd_off(intel_dp, true);
+		intel_edp_panel_on(intel_dp);
 	}
 
-	WARN_ON(intel_crtc->ddi_pll_sel == PORT_CLK_SEL_NONE);
-	I915_WRITE(PORT_CLK_SEL(port), intel_crtc->ddi_pll_sel);
+	WARN_ON(crtc->ddi_pll_sel == PORT_CLK_SEL_NONE);
+	I915_WRITE(PORT_CLK_SEL(port), crtc->ddi_pll_sel);
 
 	if (type == INTEL_OUTPUT_DISPLAYPORT || type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
+		struct intel_digital_port *intel_dig_port =
+			enc_to_dig_port(encoder);
+
+		intel_dp->DP = intel_dig_port->saved_port_bits |
+			       DDI_BUF_CTL_ENABLE | DDI_BUF_EMP_400MV_0DB_HSW;
+		intel_dp->DP |= DDI_PORT_WIDTH(intel_dp->lane_count);
 
 		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON);
 		intel_dp_start_link_train(intel_dp);
 		intel_dp_complete_link_train(intel_dp);
 		if (port != PORT_A)
 			intel_dp_stop_link_train(intel_dp);
+	} else if (type == INTEL_OUTPUT_HDMI) {
+		struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder);
+
+		intel_hdmi->set_infoframes(encoder,
+					   crtc->config.has_hdmi_sink,
+					   &crtc->config.adjusted_mode);
 	}
 }
 
@@ -1166,9 +1313,9 @@ static void intel_ddi_post_disable(struct intel_encoder *intel_encoder)
 
 	if (type == INTEL_OUTPUT_DISPLAYPORT || type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
-		ironlake_edp_panel_vdd_on(intel_dp);
 		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
-		ironlake_edp_panel_off(intel_dp);
+		intel_edp_panel_vdd_on(intel_dp);
+		intel_edp_panel_off(intel_dp);
 	}
 
 	I915_WRITE(PORT_CLK_SEL(port), PORT_CLK_SEL_NONE);
@@ -1203,11 +1350,12 @@ static void intel_enable_ddi(struct intel_encoder *intel_encoder)
 		if (port == PORT_A)
 			intel_dp_stop_link_train(intel_dp);
 
-		ironlake_edp_backlight_on(intel_dp);
+		intel_edp_backlight_on(intel_dp);
 		intel_edp_psr_enable(intel_dp);
 	}
 
-	if (intel_crtc->eld_vld && type != INTEL_OUTPUT_EDP) {
+	if (intel_crtc->config.has_audio) {
+		intel_display_power_get(dev_priv, POWER_DOMAIN_AUDIO);
 		tmp = I915_READ(HSW_AUD_PIN_ELD_CP_VLD);
 		tmp |= ((AUDIO_OUTPUT_ENABLE_A | AUDIO_ELD_VALID_A) << (pipe * 4));
 		I915_WRITE(HSW_AUD_PIN_ELD_CP_VLD, tmp);
@@ -1225,18 +1373,21 @@ static void intel_disable_ddi(struct intel_encoder *intel_encoder)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t tmp;
 
-	if (intel_crtc->eld_vld && type != INTEL_OUTPUT_EDP) {
+	/* We can't touch HSW_AUD_PIN_ELD_CP_VLD uncionditionally because this
+	 * register is part of the power well on Haswell. */
+	if (intel_crtc->config.has_audio) {
 		tmp = I915_READ(HSW_AUD_PIN_ELD_CP_VLD);
 		tmp &= ~((AUDIO_OUTPUT_ENABLE_A | AUDIO_ELD_VALID_A) <<
 			 (pipe * 4));
 		I915_WRITE(HSW_AUD_PIN_ELD_CP_VLD, tmp);
+		intel_display_power_put(dev_priv, POWER_DOMAIN_AUDIO);
 	}
 
 	if (type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
 
 		intel_edp_psr_disable(intel_dp);
-		ironlake_edp_backlight_off(intel_dp);
+		intel_edp_backlight_off(intel_dp);
 	}
 }
 
@@ -1248,7 +1399,7 @@ int intel_ddi_get_cdclk_freq(struct drm_i915_private *dev_priv)
 
 	if (lcpll & LCPLL_CD_SOURCE_FCLK) {
 		return 800000;
-	} else if (I915_READ(HSW_FUSE_STRAP) & HSW_CDCLK_LIMIT) {
+	} else if (I915_READ(FUSE_STRAP) & HSW_CDCLK_LIMIT) {
 		return 450000;
 	} else if (freq == LCPLL_CLK_FREQ_450) {
 		return 450000;
@@ -1402,6 +1553,7 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 
 	switch (temp & TRANS_DDI_MODE_SELECT_MASK) {
 	case TRANS_DDI_MODE_SELECT_HDMI:
+		pipe_config->has_hdmi_sink = true;
 	case TRANS_DDI_MODE_SELECT_DVI:
 	case TRANS_DDI_MODE_SELECT_FDI:
 		break;
@@ -1414,6 +1566,12 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 		break;
 	}
 
+	if (intel_display_power_enabled(dev_priv, POWER_DOMAIN_AUDIO)) {
+		temp = I915_READ(HSW_AUD_PIN_ELD_CP_VLD);
+		if (temp & (AUDIO_OUTPUT_ENABLE_A << (intel_crtc->pipe * 4)))
+			pipe_config->has_audio = true;
+	}
+
 	if (encoder->type == INTEL_OUTPUT_EDP && dev_priv->vbt.edp_bpp &&
 	    pipe_config->pipe_bpp > dev_priv->vbt.edp_bpp) {
 		/*
@@ -1433,6 +1591,8 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 			      pipe_config->pipe_bpp, dev_priv->vbt.edp_bpp);
 		dev_priv->vbt.edp_bpp = pipe_config->pipe_bpp;
 	}
+
+	intel_ddi_clock_get(encoder, pipe_config);
 }
 
 static void intel_ddi_destroy(struct drm_encoder *encoder)
@@ -1528,7 +1688,6 @@ void intel_ddi_init(struct drm_device *dev, enum port port)
 			 DRM_MODE_ENCODER_TMDS);
 
 	intel_encoder->compute_config = intel_ddi_compute_config;
-	intel_encoder->mode_set = intel_ddi_mode_set;
 	intel_encoder->enable = intel_enable_ddi;
 	intel_encoder->pre_enable = intel_ddi_pre_enable;
 	intel_encoder->disable = intel_disable_ddi;
@@ -1543,7 +1702,7 @@ void intel_ddi_init(struct drm_device *dev, enum port port)
 
 	intel_encoder->type = INTEL_OUTPUT_UNKNOWN;
 	intel_encoder->crtc_mask =  (1 << 0) | (1 << 1) | (1 << 2);
-	intel_encoder->cloneable = false;
+	intel_encoder->cloneable = 0;
 	intel_encoder->hot_plug = intel_ddi_hot_plug;
 
 	if (init_dp)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 2bde35d34eb..f0be855ddf4 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -41,6 +41,9 @@
 #include <drm/drm_crtc_helper.h>
 #include <linux/dma_remapping.h>
 
+#define DIV_ROUND_CLOSEST_ULL(ll, d)	\
+	({ unsigned long long _tmp = (ll)+(d)/2; do_div(_tmp, d); _tmp; })
+
 static void intel_increase_pllclock(struct drm_crtc *crtc);
 static void intel_crtc_update_cursor(struct drm_crtc *crtc, bool on);
 
@@ -51,7 +54,19 @@ static void ironlake_pch_clock_get(struct intel_crtc *crtc,
 
 static int intel_set_mode(struct drm_crtc *crtc, struct drm_display_mode *mode,
 			  int x, int y, struct drm_framebuffer *old_fb);
-
+static int intel_framebuffer_init(struct drm_device *dev,
+				  struct intel_framebuffer *ifb,
+				  struct drm_mode_fb_cmd2 *mode_cmd,
+				  struct drm_i915_gem_object *obj);
+static void intel_dp_set_m_n(struct intel_crtc *crtc);
+static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc);
+static void intel_set_pipe_timings(struct intel_crtc *intel_crtc);
+static void intel_cpu_transcoder_set_m_n(struct intel_crtc *crtc,
+					 struct intel_link_m_n *m_n);
+static void ironlake_set_pipeconf(struct drm_crtc *crtc);
+static void haswell_set_pipeconf(struct drm_crtc *crtc);
+static void intel_set_pipe_csc(struct drm_crtc *crtc);
+static void vlv_prepare_pll(struct intel_crtc *crtc);
 
 typedef struct {
 	int	min, max;
@@ -90,8 +105,8 @@ intel_fdi_link_freq(struct drm_device *dev)
 
 static const intel_limit_t intel_limits_i8xx_dac = {
 	.dot = { .min = 25000, .max = 350000 },
-	.vco = { .min = 930000, .max = 1400000 },
-	.n = { .min = 3, .max = 16 },
+	.vco = { .min = 908000, .max = 1512000 },
+	.n = { .min = 2, .max = 16 },
 	.m = { .min = 96, .max = 140 },
 	.m1 = { .min = 18, .max = 26 },
 	.m2 = { .min = 6, .max = 16 },
@@ -103,8 +118,8 @@ static const intel_limit_t intel_limits_i8xx_dac = {
 
 static const intel_limit_t intel_limits_i8xx_dvo = {
 	.dot = { .min = 25000, .max = 350000 },
-	.vco = { .min = 930000, .max = 1400000 },
-	.n = { .min = 3, .max = 16 },
+	.vco = { .min = 908000, .max = 1512000 },
+	.n = { .min = 2, .max = 16 },
 	.m = { .min = 96, .max = 140 },
 	.m1 = { .min = 18, .max = 26 },
 	.m2 = { .min = 6, .max = 16 },
@@ -116,8 +131,8 @@ static const intel_limit_t intel_limits_i8xx_dvo = {
 
 static const intel_limit_t intel_limits_i8xx_lvds = {
 	.dot = { .min = 25000, .max = 350000 },
-	.vco = { .min = 930000, .max = 1400000 },
-	.n = { .min = 3, .max = 16 },
+	.vco = { .min = 908000, .max = 1512000 },
+	.n = { .min = 2, .max = 16 },
 	.m = { .min = 96, .max = 140 },
 	.m1 = { .min = 18, .max = 26 },
 	.m2 = { .min = 6, .max = 16 },
@@ -325,10 +340,28 @@ static const intel_limit_t intel_limits_vlv = {
 	.p2 = { .p2_slow = 2, .p2_fast = 20 }, /* slow=min, fast=max */
 };
 
+static const intel_limit_t intel_limits_chv = {
+	/*
+	 * These are the data rate limits (measured in fast clocks)
+	 * since those are the strictest limits we have.  The fast
+	 * clock and actual rate limits are more relaxed, so checking
+	 * them would make no difference.
+	 */
+	.dot = { .min = 25000 * 5, .max = 540000 * 5},
+	.vco = { .min = 4860000, .max = 6700000 },
+	.n = { .min = 1, .max = 1 },
+	.m1 = { .min = 2, .max = 2 },
+	.m2 = { .min = 24 << 22, .max = 175 << 22 },
+	.p1 = { .min = 2, .max = 4 },
+	.p2 = {	.p2_slow = 1, .p2_fast = 14 },
+};
+
 static void vlv_clock(int refclk, intel_clock_t *clock)
 {
 	clock->m = clock->m1 * clock->m2;
 	clock->p = clock->p1 * clock->p2;
+	if (WARN_ON(clock->n == 0 || clock->p == 0))
+		return;
 	clock->vco = DIV_ROUND_CLOSEST(refclk * clock->m, clock->n);
 	clock->dot = DIV_ROUND_CLOSEST(clock->vco, clock->p);
 }
@@ -407,6 +440,8 @@ static const intel_limit_t *intel_limit(struct drm_crtc *crtc, int refclk)
 			limit = &intel_limits_pineview_lvds;
 		else
 			limit = &intel_limits_pineview_sdvo;
+	} else if (IS_CHERRYVIEW(dev)) {
+		limit = &intel_limits_chv;
 	} else if (IS_VALLEYVIEW(dev)) {
 		limit = &intel_limits_vlv;
 	} else if (!IS_GEN2(dev)) {
@@ -430,6 +465,8 @@ static void pineview_clock(int refclk, intel_clock_t *clock)
 {
 	clock->m = clock->m2 + 2;
 	clock->p = clock->p1 * clock->p2;
+	if (WARN_ON(clock->n == 0 || clock->p == 0))
+		return;
 	clock->vco = DIV_ROUND_CLOSEST(refclk * clock->m, clock->n);
 	clock->dot = DIV_ROUND_CLOSEST(clock->vco, clock->p);
 }
@@ -443,10 +480,23 @@ static void i9xx_clock(int refclk, intel_clock_t *clock)
 {
 	clock->m = i9xx_dpll_compute_m(clock);
 	clock->p = clock->p1 * clock->p2;
+	if (WARN_ON(clock->n + 2 == 0 || clock->p == 0))
+		return;
 	clock->vco = DIV_ROUND_CLOSEST(refclk * clock->m, clock->n + 2);
 	clock->dot = DIV_ROUND_CLOSEST(clock->vco, clock->p);
 }
 
+static void chv_clock(int refclk, intel_clock_t *clock)
+{
+	clock->m = clock->m1 * clock->m2;
+	clock->p = clock->p1 * clock->p2;
+	if (WARN_ON(clock->n == 0 || clock->p == 0))
+		return;
+	clock->vco = DIV_ROUND_CLOSEST_ULL((uint64_t)refclk * clock->m,
+			clock->n << 22);
+	clock->dot = DIV_ROUND_CLOSEST(clock->vco, clock->p);
+}
+
 #define INTELPllInvalid(s)   do { /* DRM_DEBUG(s); */ return false; } while (0)
 /**
  * Returns whether the given set of divisors are valid for a given refclk with
@@ -722,6 +772,58 @@ vlv_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
 	return found;
 }
 
+static bool
+chv_find_best_dpll(const intel_limit_t *limit, struct drm_crtc *crtc,
+		   int target, int refclk, intel_clock_t *match_clock,
+		   intel_clock_t *best_clock)
+{
+	struct drm_device *dev = crtc->dev;
+	intel_clock_t clock;
+	uint64_t m2;
+	int found = false;
+
+	memset(best_clock, 0, sizeof(*best_clock));
+
+	/*
+	 * Based on hardware doc, the n always set to 1, and m1 always
+	 * set to 2.  If requires to support 200Mhz refclk, we need to
+	 * revisit this because n may not 1 anymore.
+	 */
+	clock.n = 1, clock.m1 = 2;
+	target *= 5;	/* fast clock */
+
+	for (clock.p1 = limit->p1.max; clock.p1 >= limit->p1.min; clock.p1--) {
+		for (clock.p2 = limit->p2.p2_fast;
+				clock.p2 >= limit->p2.p2_slow;
+				clock.p2 -= clock.p2 > 10 ? 2 : 1) {
+
+			clock.p = clock.p1 * clock.p2;
+
+			m2 = DIV_ROUND_CLOSEST_ULL(((uint64_t)target * clock.p *
+					clock.n) << 22, refclk * clock.m1);
+
+			if (m2 > INT_MAX/clock.m1)
+				continue;
+
+			clock.m2 = m2;
+
+			chv_clock(refclk, &clock);
+
+			if (!intel_PLL_is_valid(dev, limit, &clock))
+				continue;
+
+			/* based on hardware requirement, prefer bigger p
+			 */
+			if (clock.p > best_clock->p) {
+				*best_clock = clock;
+				found = true;
+			}
+		}
+	}
+
+	return found;
+}
+
 bool intel_crtc_active(struct drm_crtc *crtc)
 {
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
@@ -732,10 +834,10 @@ bool intel_crtc_active(struct drm_crtc *crtc)
 	 * We can ditch the adjusted_mode.crtc_clock check as soon
 	 * as Haswell has gained clock readout/fastboot support.
 	 *
-	 * We can ditch the crtc->fb check as soon as we can
+	 * We can ditch the crtc->primary->fb check as soon as we can
 	 * properly reconstruct framebuffers.
 	 */
-	return intel_crtc->active && crtc->fb &&
+	return intel_crtc->active && crtc->primary->fb &&
 		intel_crtc->config.adjusted_mode.crtc_clock;
 }
 
@@ -748,15 +850,15 @@ enum transcoder intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
 	return intel_crtc->config.cpu_transcoder;
 }
 
-static void ironlake_wait_for_vblank(struct drm_device *dev, int pipe)
+static void g4x_wait_for_vblank(struct drm_device *dev, int pipe)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 frame, frame_reg = PIPEFRAME(pipe);
+	u32 frame, frame_reg = PIPE_FRMCOUNT_GM45(pipe);
 
 	frame = I915_READ(frame_reg);
 
 	if (wait_for(I915_READ_NOTRACE(frame_reg) != frame, 50))
-		DRM_DEBUG_KMS("vblank wait timed out\n");
+		WARN(1, "vblank wait timed out\n");
 }
 
 /**
@@ -772,8 +874,8 @@ void intel_wait_for_vblank(struct drm_device *dev, int pipe)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipestat_reg = PIPESTAT(pipe);
 
-	if (INTEL_INFO(dev)->gen >= 5) {
-		ironlake_wait_for_vblank(dev, pipe);
+	if (IS_G4X(dev) || INTEL_INFO(dev)->gen >= 5) {
+		g4x_wait_for_vblank(dev, pipe);
 		return;
 	}
 
@@ -869,7 +971,7 @@ bool ibx_digital_port_connected(struct drm_i915_private *dev_priv,
 	u32 bit;
 
 	if (HAS_PCH_IBX(dev_priv->dev)) {
-		switch(port->port) {
+		switch (port->port) {
 		case PORT_B:
 			bit = SDE_PORTB_HOTPLUG;
 			break;
@@ -883,7 +985,7 @@ bool ibx_digital_port_connected(struct drm_i915_private *dev_priv,
 			return true;
 		}
 	} else {
-		switch(port->port) {
+		switch (port->port) {
 		case PORT_B:
 			bit = SDE_PORTB_HOTPLUG_CPT;
 			break;
@@ -1024,7 +1126,7 @@ static void assert_fdi_tx_pll_enabled(struct drm_i915_private *dev_priv,
 	u32 val;
 
 	/* ILK FDI PLL is always enabled */
-	if (dev_priv->info->gen == 5)
+	if (INTEL_INFO(dev_priv->dev)->gen == 5)
 		return;
 
 	/* On Haswell, DDI ports are responsible for the FDI PLL setup */
@@ -1086,9 +1188,7 @@ static void assert_cursor(struct drm_i915_private *dev_priv,
 	struct drm_device *dev = dev_priv->dev;
 	bool cur_state;
 
-	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev))
-		cur_state = I915_READ(CURCNTR_IVB(pipe)) & CURSOR_MODE;
-	else if (IS_845G(dev) || IS_I865G(dev))
+	if (IS_845G(dev) || IS_I865G(dev))
 		cur_state = I915_READ(_CURACNTR) & CURSOR_ENABLE;
 	else
 		cur_state = I915_READ(CURCNTR(pipe)) & CURSOR_MODE;
@@ -1113,7 +1213,7 @@ void assert_pipe(struct drm_i915_private *dev_priv,
 	if (pipe == PIPE_A && dev_priv->quirks & QUIRK_PIPEA_FORCE)
 		state = true;
 
-	if (!intel_display_power_enabled(dev_priv->dev,
+	if (!intel_display_power_enabled(dev_priv,
 				POWER_DOMAIN_TRANSCODER(cpu_transcoder))) {
 		cur_state = false;
 	} else {
@@ -1157,7 +1257,7 @@ static void assert_planes_disabled(struct drm_i915_private *dev_priv,
 	if (INTEL_INFO(dev)->gen >= 4) {
 		reg = DSPCNTR(pipe);
 		val = I915_READ(reg);
-		WARN((val & DISPLAY_PLANE_ENABLE),
+		WARN(val & DISPLAY_PLANE_ENABLE,
 		     "plane %c assertion failure, should be disabled but not\n",
 		     plane_name(pipe));
 		return;
@@ -1179,41 +1279,38 @@ static void assert_sprites_disabled(struct drm_i915_private *dev_priv,
 				    enum pipe pipe)
 {
 	struct drm_device *dev = dev_priv->dev;
-	int reg, i;
+	int reg, sprite;
 	u32 val;
 
 	if (IS_VALLEYVIEW(dev)) {
-		for (i = 0; i < dev_priv->num_plane; i++) {
-			reg = SPCNTR(pipe, i);
+		for_each_sprite(pipe, sprite) {
+			reg = SPCNTR(pipe, sprite);
 			val = I915_READ(reg);
-			WARN((val & SP_ENABLE),
+			WARN(val & SP_ENABLE,
 			     "sprite %c assertion failure, should be off on pipe %c but is still active\n",
-			     sprite_name(pipe, i), pipe_name(pipe));
+			     sprite_name(pipe, sprite), pipe_name(pipe));
 		}
 	} else if (INTEL_INFO(dev)->gen >= 7) {
 		reg = SPRCTL(pipe);
 		val = I915_READ(reg);
-		WARN((val & SPRITE_ENABLE),
+		WARN(val & SPRITE_ENABLE,
 		     "sprite %c assertion failure, should be off on pipe %c but is still active\n",
 		     plane_name(pipe), pipe_name(pipe));
 	} else if (INTEL_INFO(dev)->gen >= 5) {
 		reg = DVSCNTR(pipe);
 		val = I915_READ(reg);
-		WARN((val & DVS_ENABLE),
+		WARN(val & DVS_ENABLE,
 		     "sprite %c assertion failure, should be off on pipe %c but is still active\n",
 		     plane_name(pipe), pipe_name(pipe));
 	}
 }
 
-static void assert_pch_refclk_enabled(struct drm_i915_private *dev_priv)
+static void ibx_assert_pch_refclk_enabled(struct drm_i915_private *dev_priv)
 {
 	u32 val;
 	bool enabled;
 
-	if (HAS_PCH_LPT(dev_priv->dev)) {
-		DRM_DEBUG_DRIVER("LPT does not has PCH refclk, skipping check\n");
-		return;
-	}
+	WARN_ON(!(HAS_PCH_IBX(dev_priv->dev) || HAS_PCH_CPT(dev_priv->dev)));
 
 	val = I915_READ(PCH_DREF_CONTROL);
 	enabled = !!(val & (DREF_SSC_SOURCE_MASK | DREF_NONSPREAD_SOURCE_MASK |
@@ -1247,6 +1344,9 @@ static bool dp_pipe_enabled(struct drm_i915_private *dev_priv,
 		u32	trans_dp_ctl = I915_READ(trans_dp_ctl_reg);
 		if ((trans_dp_ctl & TRANS_DP_PORT_SEL_MASK) != port_sel)
 			return false;
+	} else if (IS_CHERRYVIEW(dev_priv->dev)) {
+		if ((val & DP_PIPE_MASK_CHV) != DP_PIPE_SELECT_CHV(pipe))
+			return false;
 	} else {
 		if ((val & DP_PIPE_MASK) != (pipe << 30))
 			return false;
@@ -1263,6 +1363,9 @@ static bool hdmi_pipe_enabled(struct drm_i915_private *dev_priv,
 	if (HAS_PCH_CPT(dev_priv->dev)) {
 		if ((val & SDVO_PIPE_SEL_MASK_CPT) != SDVO_PIPE_SEL_CPT(pipe))
 			return false;
+	} else if (IS_CHERRYVIEW(dev_priv->dev)) {
+		if ((val & SDVO_PIPE_SEL_MASK_CHV) != SDVO_PIPE_SEL_CHV(pipe))
+			return false;
 	} else {
 		if ((val & SDVO_PIPE_SEL_MASK) != SDVO_PIPE_SEL(pipe))
 			return false;
@@ -1362,16 +1465,67 @@ static void intel_init_dpio(struct drm_device *dev)
 		return;
 
 	/*
-	 * From VLV2A0_DP_eDP_DPIO_driver_vbios_notes_10.docx -
-	 *  6.	De-assert cmn_reset/side_reset. Same as VLV X0.
-	 *   a.	GUnit 0x2110 bit[0] set to 1 (def 0)
-	 *   b.	The other bits such as sfr settings / modesel may all be set
-	 *      to 0.
-	 *
-	 * This should only be done on init and resume from S3 with both
-	 * PLLs disabled, or we risk losing DPIO and PLL synchronization.
+	 * IOSF_PORT_DPIO is used for VLV x2 PHY (DP/HDMI B and C),
+	 * CHV x1 PHY (DP/HDMI D)
+	 * IOSF_PORT_DPIO_2 is used for CHV x2 PHY (DP/HDMI B and C)
 	 */
-	I915_WRITE(DPIO_CTL, I915_READ(DPIO_CTL) | DPIO_CMNRST);
+	if (IS_CHERRYVIEW(dev)) {
+		DPIO_PHY_IOSF_PORT(DPIO_PHY0) = IOSF_PORT_DPIO_2;
+		DPIO_PHY_IOSF_PORT(DPIO_PHY1) = IOSF_PORT_DPIO;
+	} else {
+		DPIO_PHY_IOSF_PORT(DPIO_PHY0) = IOSF_PORT_DPIO;
+	}
+}
+
+static void intel_reset_dpio(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (!IS_VALLEYVIEW(dev))
+		return;
+
+	if (IS_CHERRYVIEW(dev)) {
+		enum dpio_phy phy;
+		u32 val;
+
+		for (phy = DPIO_PHY0; phy < I915_NUM_PHYS_VLV; phy++) {
+			/* Poll for phypwrgood signal */
+			if (wait_for(I915_READ(DISPLAY_PHY_STATUS) &
+						PHY_POWERGOOD(phy), 1))
+				DRM_ERROR("Display PHY %d is not power up\n", phy);
+
+			/*
+			 * Deassert common lane reset for PHY.
+			 *
+			 * This should only be done on init and resume from S3
+			 * with both PLLs disabled, or we risk losing DPIO and
+			 * PLL synchronization.
+			 */
+			val = I915_READ(DISPLAY_PHY_CONTROL);
+			I915_WRITE(DISPLAY_PHY_CONTROL,
+				PHY_COM_LANE_RESET_DEASSERT(phy, val));
+		}
+
+	} else {
+		/*
+		 * If DPIO has already been reset, e.g. by BIOS, just skip all
+		 * this.
+		 */
+		if (I915_READ(DPIO_CTL) & DPIO_CMNRST)
+			return;
+
+		/*
+		 * From VLV2A0_DP_eDP_HDMI_DPIO_driver_vbios_notes_11.docx:
+		 * Need to assert and de-assert PHY SB reset by gating the
+		 * common lane power, then un-gating it.
+		 * Simply ungating isn't enough to reset the PHY enough to get
+		 * ports and lanes running.
+		 */
+		__vlv_set_power_well(dev_priv, PUNIT_POWER_WELL_DPIO_CMN_BC,
+				     false);
+		__vlv_set_power_well(dev_priv, PUNIT_POWER_WELL_DPIO_CMN_BC,
+				     true);
+	}
 }
 
 static void vlv_enable_pll(struct intel_crtc *crtc)
@@ -1412,6 +1566,44 @@ static void vlv_enable_pll(struct intel_crtc *crtc)
 	udelay(150); /* wait for warmup */
 }
 
+static void chv_enable_pll(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe = crtc->pipe;
+	enum dpio_channel port = vlv_pipe_to_channel(pipe);
+	u32 tmp;
+
+	assert_pipe_disabled(dev_priv, crtc->pipe);
+
+	BUG_ON(!IS_CHERRYVIEW(dev_priv->dev));
+
+	mutex_lock(&dev_priv->dpio_lock);
+
+	/* Enable back the 10bit clock to display controller */
+	tmp = vlv_dpio_read(dev_priv, pipe, CHV_CMN_DW14(port));
+	tmp |= DPIO_DCLKP_EN;
+	vlv_dpio_write(dev_priv, pipe, CHV_CMN_DW14(port), tmp);
+
+	/*
+	 * Need to wait > 100ns between dclkp clock enable bit and PLL enable.
+	 */
+	udelay(1);
+
+	/* Enable PLL */
+	I915_WRITE(DPLL(pipe), crtc->config.dpll_hw_state.dpll);
+
+	/* Check PLL is locked */
+	if (wait_for(((I915_READ(DPLL(pipe)) & DPLL_LOCK_VLV) == DPLL_LOCK_VLV), 1))
+		DRM_ERROR("PLL %d failed to lock\n", pipe);
+
+	/* not sure when this should be written */
+	I915_WRITE(DPLL_MD(pipe), crtc->config.dpll_hw_state.dpll_md);
+	POSTING_READ(DPLL_MD(pipe));
+
+	mutex_unlock(&dev_priv->dpio_lock);
+}
+
 static void i9xx_enable_pll(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
@@ -1422,7 +1614,7 @@ static void i9xx_enable_pll(struct intel_crtc *crtc)
 	assert_pipe_disabled(dev_priv, crtc->pipe);
 
 	/* No really, not for ILK+ */
-	BUG_ON(dev_priv->info->gen >= 5);
+	BUG_ON(INTEL_INFO(dev)->gen >= 5);
 
 	/* PLL is protected by panel, make sure we can write it */
 	if (IS_MOBILE(dev) && !IS_I830(dev))
@@ -1487,42 +1679,100 @@ static void vlv_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
 	/* Make sure the pipe isn't still relying on us */
 	assert_pipe_disabled(dev_priv, pipe);
 
-	/* Leave integrated clock source enabled */
+	/*
+	 * Leave integrated clock source and reference clock enabled for pipe B.
+	 * The latter is needed for VGA hotplug / manual detection.
+	 */
 	if (pipe == PIPE_B)
-		val = DPLL_INTEGRATED_CRI_CLK_VLV;
+		val = DPLL_INTEGRATED_CRI_CLK_VLV | DPLL_REFA_CLK_ENABLE_VLV;
 	I915_WRITE(DPLL(pipe), val);
 	POSTING_READ(DPLL(pipe));
+
 }
 
-void vlv_wait_port_ready(struct drm_i915_private *dev_priv, int port)
+static void chv_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
+{
+	enum dpio_channel port = vlv_pipe_to_channel(pipe);
+	u32 val;
+
+	/* Make sure the pipe isn't still relying on us */
+	assert_pipe_disabled(dev_priv, pipe);
+
+	/* Set PLL en = 0 */
+	val = DPLL_SSC_REF_CLOCK_CHV;
+	if (pipe != PIPE_A)
+		val |= DPLL_INTEGRATED_CRI_CLK_VLV;
+	I915_WRITE(DPLL(pipe), val);
+	POSTING_READ(DPLL(pipe));
+
+	mutex_lock(&dev_priv->dpio_lock);
+
+	/* Disable 10bit clock to display controller */
+	val = vlv_dpio_read(dev_priv, pipe, CHV_CMN_DW14(port));
+	val &= ~DPIO_DCLKP_EN;
+	vlv_dpio_write(dev_priv, pipe, CHV_CMN_DW14(port), val);
+
+	mutex_unlock(&dev_priv->dpio_lock);
+}
+
+void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
+		struct intel_digital_port *dport)
 {
 	u32 port_mask;
+	int dpll_reg;
 
-	if (!port)
+	switch (dport->port) {
+	case PORT_B:
 		port_mask = DPLL_PORTB_READY_MASK;
-	else
+		dpll_reg = DPLL(0);
+		break;
+	case PORT_C:
 		port_mask = DPLL_PORTC_READY_MASK;
+		dpll_reg = DPLL(0);
+		break;
+	case PORT_D:
+		port_mask = DPLL_PORTD_READY_MASK;
+		dpll_reg = DPIO_PHY_STATUS;
+		break;
+	default:
+		BUG();
+	}
 
-	if (wait_for((I915_READ(DPLL(0)) & port_mask) == 0, 1000))
+	if (wait_for((I915_READ(dpll_reg) & port_mask) == 0, 1000))
 		WARN(1, "timed out waiting for port %c ready: 0x%08x\n",
-		     'B' + port, I915_READ(DPLL(0)));
+		     port_name(dport->port), I915_READ(dpll_reg));
+}
+
+static void intel_prepare_shared_dpll(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_shared_dpll *pll = intel_crtc_to_shared_dpll(crtc);
+
+	WARN_ON(!pll->refcount);
+	if (pll->active == 0) {
+		DRM_DEBUG_DRIVER("setting up %s\n", pll->name);
+		WARN_ON(pll->on);
+		assert_shared_dpll_disabled(dev_priv, pll);
+
+		pll->mode_set(dev_priv, pll);
+	}
 }
 
 /**
- * ironlake_enable_shared_dpll - enable PCH PLL
+ * intel_enable_shared_dpll - enable PCH PLL
  * @dev_priv: i915 private structure
  * @pipe: pipe PLL to enable
  *
  * The PCH PLL needs to be enabled before the PCH transcoder, since it
  * drives the transcoder clock.
  */
-static void ironlake_enable_shared_dpll(struct intel_crtc *crtc)
+static void intel_enable_shared_dpll(struct intel_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_shared_dpll *pll = intel_crtc_to_shared_dpll(crtc);
 
-	/* PCH PLLs only available on ILK, SNB and IVB */
-	BUG_ON(dev_priv->info->gen < 5);
 	if (WARN_ON(pll == NULL))
 		return;
 
@@ -1547,11 +1797,12 @@ static void ironlake_enable_shared_dpll(struct intel_crtc *crtc)
 
 static void intel_disable_shared_dpll(struct intel_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_shared_dpll *pll = intel_crtc_to_shared_dpll(crtc);
 
 	/* PCH only available on ILK+ */
-	BUG_ON(dev_priv->info->gen < 5);
+	BUG_ON(INTEL_INFO(dev)->gen < 5);
 	if (WARN_ON(pll == NULL))
 	       return;
 
@@ -1586,7 +1837,7 @@ static void ironlake_enable_pch_transcoder(struct drm_i915_private *dev_priv,
 	uint32_t reg, val, pipeconf_val;
 
 	/* PCH only available on ILK+ */
-	BUG_ON(dev_priv->info->gen < 5);
+	BUG_ON(INTEL_INFO(dev)->gen < 5);
 
 	/* Make sure PCH DPLL is enabled */
 	assert_shared_dpll_enabled(dev_priv,
@@ -1639,7 +1890,7 @@ static void lpt_enable_pch_transcoder(struct drm_i915_private *dev_priv,
 	u32 val, pipeconf_val;
 
 	/* PCH only available on ILK+ */
-	BUG_ON(dev_priv->info->gen < 5);
+	BUG_ON(INTEL_INFO(dev_priv->dev)->gen < 5);
 
 	/* FDI must be feeding us bits for PCH ports */
 	assert_fdi_tx_enabled(dev_priv, (enum pipe) cpu_transcoder);
@@ -1713,21 +1964,16 @@ static void lpt_disable_pch_transcoder(struct drm_i915_private *dev_priv)
 
 /**
  * intel_enable_pipe - enable a pipe, asserting requirements
- * @dev_priv: i915 private structure
- * @pipe: pipe to enable
- * @pch_port: on ILK+, is this pipe driving a PCH port or not
+ * @crtc: crtc responsible for the pipe
  *
- * Enable @pipe, making sure that various hardware specific requirements
+ * Enable @crtc's pipe, making sure that various hardware specific requirements
  * are met, if applicable, e.g. PLL enabled, LVDS pairs enabled, etc.
- *
- * @pipe should be %PIPE_A or %PIPE_B.
- *
- * Will wait until the pipe is actually running (i.e. first vblank) before
- * returning.
  */
-static void intel_enable_pipe(struct drm_i915_private *dev_priv, enum pipe pipe,
-			      bool pch_port, bool dsi)
+static void intel_enable_pipe(struct intel_crtc *crtc)
 {
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe = crtc->pipe;
 	enum transcoder cpu_transcoder = intel_pipe_to_cpu_transcoder(dev_priv,
 								      pipe);
 	enum pipe pch_transcoder;
@@ -1749,12 +1995,12 @@ static void intel_enable_pipe(struct drm_i915_private *dev_priv, enum pipe pipe,
 	 * need the check.
 	 */
 	if (!HAS_PCH_SPLIT(dev_priv->dev))
-		if (dsi)
+		if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DSI))
 			assert_dsi_pll_enabled(dev_priv);
 		else
 			assert_pll_enabled(dev_priv, pipe);
 	else {
-		if (pch_port) {
+		if (crtc->config.has_pch_encoder) {
 			/* if driving the PCH, we need FDI enabled */
 			assert_fdi_rx_pll_enabled(dev_priv, pch_transcoder);
 			assert_fdi_tx_pll_enabled(dev_priv,
@@ -1765,11 +2011,14 @@ static void intel_enable_pipe(struct drm_i915_private *dev_priv, enum pipe pipe,
 
 	reg = PIPECONF(cpu_transcoder);
 	val = I915_READ(reg);
-	if (val & PIPECONF_ENABLE)
+	if (val & PIPECONF_ENABLE) {
+		WARN_ON(!(pipe == PIPE_A &&
+			  dev_priv->quirks & QUIRK_PIPEA_FORCE));
 		return;
+	}
 
 	I915_WRITE(reg, val | PIPECONF_ENABLE);
-	intel_wait_for_vblank(dev_priv->dev, pipe);
+	POSTING_READ(reg);
 }
 
 /**
@@ -1820,23 +2069,25 @@ static void intel_disable_pipe(struct drm_i915_private *dev_priv,
 void intel_flush_primary_plane(struct drm_i915_private *dev_priv,
 			       enum plane plane)
 {
-	u32 reg = dev_priv->info->gen >= 4 ? DSPSURF(plane) : DSPADDR(plane);
+	struct drm_device *dev = dev_priv->dev;
+	u32 reg = INTEL_INFO(dev)->gen >= 4 ? DSPSURF(plane) : DSPADDR(plane);
 
 	I915_WRITE(reg, I915_READ(reg));
 	POSTING_READ(reg);
 }
 
 /**
- * intel_enable_primary_plane - enable the primary plane on a given pipe
+ * intel_enable_primary_hw_plane - enable the primary plane on a given pipe
  * @dev_priv: i915 private structure
  * @plane: plane to enable
  * @pipe: pipe being fed
  *
  * Enable @plane on @pipe, making sure that @pipe is running first.
  */
-static void intel_enable_primary_plane(struct drm_i915_private *dev_priv,
-				       enum plane plane, enum pipe pipe)
+static void intel_enable_primary_hw_plane(struct drm_i915_private *dev_priv,
+					  enum plane plane, enum pipe pipe)
 {
+	struct drm_device *dev = dev_priv->dev;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
 	int reg;
@@ -1845,48 +2096,54 @@ static void intel_enable_primary_plane(struct drm_i915_private *dev_priv,
 	/* If the pipe isn't enabled, we can't pump pixels and may hang */
 	assert_pipe_enabled(dev_priv, pipe);
 
-	WARN(intel_crtc->primary_enabled, "Primary plane already enabled\n");
+	if (intel_crtc->primary_enabled)
+		return;
 
 	intel_crtc->primary_enabled = true;
 
 	reg = DSPCNTR(plane);
 	val = I915_READ(reg);
-	if (val & DISPLAY_PLANE_ENABLE)
-		return;
+	WARN_ON(val & DISPLAY_PLANE_ENABLE);
 
 	I915_WRITE(reg, val | DISPLAY_PLANE_ENABLE);
 	intel_flush_primary_plane(dev_priv, plane);
-	intel_wait_for_vblank(dev_priv->dev, pipe);
+
+	/*
+	 * BDW signals flip done immediately if the plane
+	 * is disabled, even if the plane enable is already
+	 * armed to occur at the next vblank :(
+	 */
+	if (IS_BROADWELL(dev))
+		intel_wait_for_vblank(dev, intel_crtc->pipe);
 }
 
 /**
- * intel_disable_primary_plane - disable the primary plane
+ * intel_disable_primary_hw_plane - disable the primary hardware plane
  * @dev_priv: i915 private structure
  * @plane: plane to disable
  * @pipe: pipe consuming the data
  *
  * Disable @plane; should be an independent operation.
  */
-static void intel_disable_primary_plane(struct drm_i915_private *dev_priv,
-					enum plane plane, enum pipe pipe)
+static void intel_disable_primary_hw_plane(struct drm_i915_private *dev_priv,
+					   enum plane plane, enum pipe pipe)
 {
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
 	int reg;
 	u32 val;
 
-	WARN(!intel_crtc->primary_enabled, "Primary plane already disabled\n");
+	if (!intel_crtc->primary_enabled)
+		return;
 
 	intel_crtc->primary_enabled = false;
 
 	reg = DSPCNTR(plane);
 	val = I915_READ(reg);
-	if ((val & DISPLAY_PLANE_ENABLE) == 0)
-		return;
+	WARN_ON((val & DISPLAY_PLANE_ENABLE) == 0);
 
 	I915_WRITE(reg, val & ~DISPLAY_PLANE_ENABLE);
 	intel_flush_primary_plane(dev_priv, plane);
-	intel_wait_for_vblank(dev_priv->dev, pipe);
 }
 
 static bool need_vtd_wa(struct drm_device *dev)
@@ -1898,10 +2155,18 @@ static bool need_vtd_wa(struct drm_device *dev)
 	return false;
 }
 
+static int intel_align_height(struct drm_device *dev, int height, bool tiled)
+{
+	int tile_height;
+
+	tile_height = tiled ? (IS_GEN2(dev) ? 16 : 8) : 1;
+	return ALIGN(height, tile_height);
+}
+
 int
 intel_pin_and_fence_fb_obj(struct drm_device *dev,
 			   struct drm_i915_gem_object *obj,
-			   struct intel_ring_buffer *pipelined)
+			   struct intel_engine_cs *pipelined)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 alignment;
@@ -1994,8 +2259,114 @@ unsigned long intel_gen4_compute_page_offset(int *x, int *y,
 	}
 }
 
-static int i9xx_update_plane(struct drm_crtc *crtc, struct drm_framebuffer *fb,
-			     int x, int y)
+int intel_format_to_fourcc(int format)
+{
+	switch (format) {
+	case DISPPLANE_8BPP:
+		return DRM_FORMAT_C8;
+	case DISPPLANE_BGRX555:
+		return DRM_FORMAT_XRGB1555;
+	case DISPPLANE_BGRX565:
+		return DRM_FORMAT_RGB565;
+	default:
+	case DISPPLANE_BGRX888:
+		return DRM_FORMAT_XRGB8888;
+	case DISPPLANE_RGBX888:
+		return DRM_FORMAT_XBGR8888;
+	case DISPPLANE_BGRX101010:
+		return DRM_FORMAT_XRGB2101010;
+	case DISPPLANE_RGBX101010:
+		return DRM_FORMAT_XBGR2101010;
+	}
+}
+
+static bool intel_alloc_plane_obj(struct intel_crtc *crtc,
+				  struct intel_plane_config *plane_config)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_gem_object *obj = NULL;
+	struct drm_mode_fb_cmd2 mode_cmd = { 0 };
+	u32 base = plane_config->base;
+
+	if (plane_config->size == 0)
+		return false;
+
+	obj = i915_gem_object_create_stolen_for_preallocated(dev, base, base,
+							     plane_config->size);
+	if (!obj)
+		return false;
+
+	if (plane_config->tiled) {
+		obj->tiling_mode = I915_TILING_X;
+		obj->stride = crtc->base.primary->fb->pitches[0];
+	}
+
+	mode_cmd.pixel_format = crtc->base.primary->fb->pixel_format;
+	mode_cmd.width = crtc->base.primary->fb->width;
+	mode_cmd.height = crtc->base.primary->fb->height;
+	mode_cmd.pitches[0] = crtc->base.primary->fb->pitches[0];
+
+	mutex_lock(&dev->struct_mutex);
+
+	if (intel_framebuffer_init(dev, to_intel_framebuffer(crtc->base.primary->fb),
+				   &mode_cmd, obj)) {
+		DRM_DEBUG_KMS("intel fb init failed\n");
+		goto out_unref_obj;
+	}
+
+	mutex_unlock(&dev->struct_mutex);
+
+	DRM_DEBUG_KMS("plane fb obj %p\n", obj);
+	return true;
+
+out_unref_obj:
+	drm_gem_object_unreference(&obj->base);
+	mutex_unlock(&dev->struct_mutex);
+	return false;
+}
+
+static void intel_find_plane_obj(struct intel_crtc *intel_crtc,
+				 struct intel_plane_config *plane_config)
+{
+	struct drm_device *dev = intel_crtc->base.dev;
+	struct drm_crtc *c;
+	struct intel_crtc *i;
+	struct intel_framebuffer *fb;
+
+	if (!intel_crtc->base.primary->fb)
+		return;
+
+	if (intel_alloc_plane_obj(intel_crtc, plane_config))
+		return;
+
+	kfree(intel_crtc->base.primary->fb);
+	intel_crtc->base.primary->fb = NULL;
+
+	/*
+	 * Failed to alloc the obj, check to see if we should share
+	 * an fb with another CRTC instead
+	 */
+	for_each_crtc(dev, c) {
+		i = to_intel_crtc(c);
+
+		if (c == &intel_crtc->base)
+			continue;
+
+		if (!i->active || !c->primary->fb)
+			continue;
+
+		fb = to_intel_framebuffer(c->primary->fb);
+		if (i915_gem_obj_ggtt_offset(fb->obj) == plane_config->base) {
+			drm_framebuffer_reference(c->primary->fb);
+			intel_crtc->base.primary->fb = c->primary->fb;
+			break;
+		}
+	}
+}
+
+static void i9xx_update_primary_plane(struct drm_crtc *crtc,
+				      struct drm_framebuffer *fb,
+				      int x, int y)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -2007,15 +2378,6 @@ static int i9xx_update_plane(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	u32 dspcntr;
 	u32 reg;
 
-	switch (plane) {
-	case 0:
-	case 1:
-		break;
-	default:
-		DRM_ERROR("Can't update plane %c in SAREA\n", plane_name(plane));
-		return -EINVAL;
-	}
-
 	intel_fb = to_intel_framebuffer(fb);
 	obj = intel_fb->obj;
 
@@ -2083,19 +2445,18 @@ static int i9xx_update_plane(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 		      fb->pitches[0]);
 	I915_WRITE(DSPSTRIDE(plane), fb->pitches[0]);
 	if (INTEL_INFO(dev)->gen >= 4) {
-		I915_MODIFY_DISPBASE(DSPSURF(plane),
-				     i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+		I915_WRITE(DSPSURF(plane),
+			   i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
 		I915_WRITE(DSPTILEOFF(plane), (y << 16) | x);
 		I915_WRITE(DSPLINOFF(plane), linear_offset);
 	} else
 		I915_WRITE(DSPADDR(plane), i915_gem_obj_ggtt_offset(obj) + linear_offset);
 	POSTING_READ(reg);
-
-	return 0;
 }
 
-static int ironlake_update_plane(struct drm_crtc *crtc,
-				 struct drm_framebuffer *fb, int x, int y)
+static void ironlake_update_primary_plane(struct drm_crtc *crtc,
+					  struct drm_framebuffer *fb,
+					  int x, int y)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -2107,16 +2468,6 @@ static int ironlake_update_plane(struct drm_crtc *crtc,
 	u32 dspcntr;
 	u32 reg;
 
-	switch (plane) {
-	case 0:
-	case 1:
-	case 2:
-		break;
-	default:
-		DRM_ERROR("Can't update plane %c in SAREA\n", plane_name(plane));
-		return -EINVAL;
-	}
-
 	intel_fb = to_intel_framebuffer(fb);
 	obj = intel_fb->obj;
 
@@ -2174,8 +2525,8 @@ static int ironlake_update_plane(struct drm_crtc *crtc,
 		      i915_gem_obj_ggtt_offset(obj), linear_offset, x, y,
 		      fb->pitches[0]);
 	I915_WRITE(DSPSTRIDE(plane), fb->pitches[0]);
-	I915_MODIFY_DISPBASE(DSPSURF(plane),
-			     i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+	I915_WRITE(DSPSURF(plane),
+		   i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		I915_WRITE(DSPOFFSET(plane), (y << 16) | x);
 	} else {
@@ -2183,8 +2534,6 @@ static int ironlake_update_plane(struct drm_crtc *crtc,
 		I915_WRITE(DSPLINOFF(plane), linear_offset);
 	}
 	POSTING_READ(reg);
-
-	return 0;
 }
 
 /* Assume fb object is pinned & idle & fenced and just update base pointers */
@@ -2199,7 +2548,9 @@ intel_pipe_set_base_atomic(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 		dev_priv->display.disable_fbc(dev);
 	intel_increase_pllclock(crtc);
 
-	return dev_priv->display.update_plane(crtc, fb, x, y);
+	dev_priv->display.update_primary_plane(crtc, fb, x, y);
+
+	return 0;
 }
 
 void intel_display_handle_reset(struct drm_device *dev)
@@ -2221,7 +2572,7 @@ void intel_display_handle_reset(struct drm_device *dev)
 	 * pending_flip_queue really got woken up.
 	 */
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+	for_each_crtc(dev, crtc) {
 		struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 		enum plane plane = intel_crtc->plane;
 
@@ -2229,14 +2580,21 @@ void intel_display_handle_reset(struct drm_device *dev)
 		intel_finish_page_flip_plane(dev, plane);
 	}
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+	for_each_crtc(dev, crtc) {
 		struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 
-		mutex_lock(&crtc->mutex);
-		if (intel_crtc->active)
-			dev_priv->display.update_plane(crtc, crtc->fb,
-						       crtc->x, crtc->y);
-		mutex_unlock(&crtc->mutex);
+		drm_modeset_lock(&crtc->mutex, NULL);
+		/*
+		 * FIXME: Once we have proper support for primary planes (and
+		 * disabling them without disabling the entire crtc) allow again
+		 * a NULL crtc->primary->fb.
+		 */
+		if (intel_crtc->active && crtc->primary->fb)
+			dev_priv->display.update_primary_plane(crtc,
+							       crtc->primary->fb,
+							       crtc->x,
+							       crtc->y);
+		drm_modeset_unlock(&crtc->mutex);
 	}
 }
 
@@ -2263,31 +2621,23 @@ intel_finish_fb(struct drm_framebuffer *old_fb)
 	return ret;
 }
 
-static void intel_crtc_update_sarea_pos(struct drm_crtc *crtc, int x, int y)
+static bool intel_crtc_has_pending_flip(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_master_private *master_priv;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	unsigned long flags;
+	bool pending;
 
-	if (!dev->primary->master)
-		return;
+	if (i915_reset_in_progress(&dev_priv->gpu_error) ||
+	    intel_crtc->reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter))
+		return false;
 
-	master_priv = dev->primary->master->driver_priv;
-	if (!master_priv->sarea_priv)
-		return;
+	spin_lock_irqsave(&dev->event_lock, flags);
+	pending = to_intel_crtc(crtc)->unpin_work != NULL;
+	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	switch (intel_crtc->pipe) {
-	case 0:
-		master_priv->sarea_priv->pipeA_x = x;
-		master_priv->sarea_priv->pipeA_y = y;
-		break;
-	case 1:
-		master_priv->sarea_priv->pipeB_x = x;
-		master_priv->sarea_priv->pipeB_y = y;
-		break;
-	default:
-		break;
-	}
+	return pending;
 }
 
 static int
@@ -2300,6 +2650,11 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	struct drm_framebuffer *old_fb;
 	int ret;
 
+	if (intel_crtc_has_pending_flip(crtc)) {
+		DRM_ERROR("pipe is still busy with an old pageflip\n");
+		return -EBUSY;
+	}
+
 	/* no fb bound */
 	if (!fb) {
 		DRM_ERROR("No FB bound\n");
@@ -2317,8 +2672,8 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	ret = intel_pin_and_fence_fb_obj(dev,
 					 to_intel_framebuffer(fb)->obj,
 					 NULL);
+	mutex_unlock(&dev->struct_mutex);
 	if (ret != 0) {
-		mutex_unlock(&dev->struct_mutex);
 		DRM_ERROR("pin & fence failed\n");
 		return ret;
 	}
@@ -2336,7 +2691,7 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	 * whether the platform allows pfit disable with pipe active, and only
 	 * then update the pipesrc and pfit state, even on the flip path.
 	 */
-	if (i915_fastboot) {
+	if (i915.fastboot) {
 		const struct drm_display_mode *adjusted_mode =
 			&intel_crtc->config.adjusted_mode;
 
@@ -2350,33 +2705,30 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 			I915_WRITE(PF_WIN_POS(intel_crtc->pipe), 0);
 			I915_WRITE(PF_WIN_SZ(intel_crtc->pipe), 0);
 		}
+		intel_crtc->config.pipe_src_w = adjusted_mode->crtc_hdisplay;
+		intel_crtc->config.pipe_src_h = adjusted_mode->crtc_vdisplay;
 	}
 
-	ret = dev_priv->display.update_plane(crtc, fb, x, y);
-	if (ret) {
-		intel_unpin_fb_obj(to_intel_framebuffer(fb)->obj);
-		mutex_unlock(&dev->struct_mutex);
-		DRM_ERROR("failed to update base address\n");
-		return ret;
-	}
+	dev_priv->display.update_primary_plane(crtc, fb, x, y);
 
-	old_fb = crtc->fb;
-	crtc->fb = fb;
+	old_fb = crtc->primary->fb;
+	crtc->primary->fb = fb;
 	crtc->x = x;
 	crtc->y = y;
 
 	if (old_fb) {
 		if (intel_crtc->active && old_fb != fb)
 			intel_wait_for_vblank(dev, intel_crtc->pipe);
+		mutex_lock(&dev->struct_mutex);
 		intel_unpin_fb_obj(to_intel_framebuffer(old_fb)->obj);
+		mutex_unlock(&dev->struct_mutex);
 	}
 
+	mutex_lock(&dev->struct_mutex);
 	intel_update_fbc(dev);
 	intel_edp_psr_update(dev);
 	mutex_unlock(&dev->struct_mutex);
 
-	intel_crtc_update_sarea_pos(crtc, x, y);
-
 	return 0;
 }
 
@@ -2460,12 +2812,10 @@ static void ironlake_fdi_link_train(struct drm_crtc *crtc)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
 	u32 reg, temp, tries;
 
-	/* FDI needs bits from pipe & plane first */
+	/* FDI needs bits from pipe first */
 	assert_pipe_enabled(dev_priv, pipe);
-	assert_plane_enabled(dev_priv, plane);
 
 	/* Train 1: umask FDI RX Interrupt symbol_lock and bit_lock bit
 	   for train result */
@@ -2896,9 +3246,8 @@ static void ironlake_fdi_disable(struct drm_crtc *crtc)
 	udelay(100);
 
 	/* Ironlake workaround, disable clock pointer after downing FDI */
-	if (HAS_PCH_IBX(dev)) {
+	if (HAS_PCH_IBX(dev))
 		I915_WRITE(FDI_RX_CHICKEN(pipe), FDI_RX_PHASE_SYNC_POINTER_OVR);
-	}
 
 	/* still set train pattern 1 */
 	reg = FDI_TX_CTL(pipe);
@@ -2925,40 +3274,46 @@ static void ironlake_fdi_disable(struct drm_crtc *crtc)
 	udelay(100);
 }
 
-static bool intel_crtc_has_pending_flip(struct drm_crtc *crtc)
+bool intel_has_pending_fb_unpin(struct drm_device *dev)
 {
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	unsigned long flags;
-	bool pending;
+	struct intel_crtc *crtc;
 
-	if (i915_reset_in_progress(&dev_priv->gpu_error) ||
-	    intel_crtc->reset_counter != atomic_read(&dev_priv->gpu_error.reset_counter))
-		return false;
+	/* Note that we don't need to be called with mode_config.lock here
+	 * as our list of CRTC objects is static for the lifetime of the
+	 * device and so cannot disappear as we iterate. Similarly, we can
+	 * happily treat the predicates as racy, atomic checks as userspace
+	 * cannot claim and pin a new fb without at least acquring the
+	 * struct_mutex and so serialising with us.
+	 */
+	for_each_intel_crtc(dev, crtc) {
+		if (atomic_read(&crtc->unpin_work_count) == 0)
+			continue;
 
-	spin_lock_irqsave(&dev->event_lock, flags);
-	pending = to_intel_crtc(crtc)->unpin_work != NULL;
-	spin_unlock_irqrestore(&dev->event_lock, flags);
+		if (crtc->unpin_work)
+			intel_wait_for_vblank(dev, crtc->pipe);
 
-	return pending;
+		return true;
+	}
+
+	return false;
 }
 
-static void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc)
+void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (crtc->fb == NULL)
+	if (crtc->primary->fb == NULL)
 		return;
 
 	WARN_ON(waitqueue_active(&dev_priv->pending_flip_queue));
 
-	wait_event(dev_priv->pending_flip_queue,
-		   !intel_crtc_has_pending_flip(crtc));
+	WARN_ON(wait_event_timeout(dev_priv->pending_flip_queue,
+				   !intel_crtc_has_pending_flip(crtc),
+				   60*HZ) == 0);
 
 	mutex_lock(&dev->struct_mutex);
-	intel_finish_fb(crtc->fb);
+	intel_finish_fb(crtc->primary->fb);
 	mutex_unlock(&dev->struct_mutex);
 }
 
@@ -3168,7 +3523,7 @@ static void ironlake_pch_enable(struct drm_crtc *crtc)
 	 * Note that enable_shared_dpll tries to do the right thing, but
 	 * get_shared_dpll unconditionally resets the pll - we need that to have
 	 * the right LVDS enable sequence. */
-	ironlake_enable_shared_dpll(intel_crtc);
+	intel_enable_shared_dpll(intel_crtc);
 
 	/* set transcoder timing, panel must allow it */
 	assert_panel_unlocked(dev_priv, pipe);
@@ -3272,6 +3627,8 @@ static struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc)
 		DRM_DEBUG_KMS("CRTC:%d using pre-allocated %s\n",
 			      crtc->base.base.id, pll->name);
 
+		WARN_ON(pll->refcount);
+
 		goto found;
 	}
 
@@ -3305,20 +3662,13 @@ static struct intel_shared_dpll *intel_get_shared_dpll(struct intel_crtc *crtc)
 	return NULL;
 
 found:
+	if (pll->refcount == 0)
+		pll->hw_state = crtc->config.dpll_hw_state;
+
 	crtc->config.shared_dpll = i;
 	DRM_DEBUG_DRIVER("using %s for pipe %c\n", pll->name,
 			 pipe_name(crtc->pipe));
 
-	if (pll->active == 0) {
-		memcpy(&pll->hw_state, &crtc->config.dpll_hw_state,
-		       sizeof(pll->hw_state));
-
-		DRM_DEBUG_DRIVER("setting up %s\n", pll->name);
-		WARN_ON(pll->on);
-		assert_shared_dpll_disabled(dev_priv, pll);
-
-		pll->mode_set(dev_priv, pll);
-	}
 	pll->refcount++;
 
 	return pll;
@@ -3363,45 +3713,50 @@ static void intel_enable_planes(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	enum pipe pipe = to_intel_crtc(crtc)->pipe;
+	struct drm_plane *plane;
 	struct intel_plane *intel_plane;
 
-	list_for_each_entry(intel_plane, &dev->mode_config.plane_list, base.head)
+	drm_for_each_legacy_plane(plane, &dev->mode_config.plane_list) {
+		intel_plane = to_intel_plane(plane);
 		if (intel_plane->pipe == pipe)
 			intel_plane_restore(&intel_plane->base);
+	}
 }
 
 static void intel_disable_planes(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	enum pipe pipe = to_intel_crtc(crtc)->pipe;
+	struct drm_plane *plane;
 	struct intel_plane *intel_plane;
 
-	list_for_each_entry(intel_plane, &dev->mode_config.plane_list, base.head)
+	drm_for_each_legacy_plane(plane, &dev->mode_config.plane_list) {
+		intel_plane = to_intel_plane(plane);
 		if (intel_plane->pipe == pipe)
 			intel_plane_disable(&intel_plane->base);
+	}
 }
 
 void hsw_enable_ips(struct intel_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	if (!crtc->config.ips_enabled)
 		return;
 
-	/* We can only enable IPS after we enable a plane and wait for a vblank.
-	 * We guarantee that the plane is enabled by calling intel_enable_ips
-	 * only after intel_enable_plane. And intel_enable_plane already waits
-	 * for a vblank, so all we need to do here is to enable the IPS bit. */
+	/* We can only enable IPS after we enable a plane and wait for a vblank */
+	intel_wait_for_vblank(dev, crtc->pipe);
+
 	assert_plane_enabled(dev_priv, crtc->plane);
-	if (IS_BROADWELL(crtc->base.dev)) {
+	if (IS_BROADWELL(dev)) {
 		mutex_lock(&dev_priv->rps.hw_lock);
 		WARN_ON(sandybridge_pcode_write(dev_priv, DISPLAY_IPS_CONTROL, 0xc0000000));
 		mutex_unlock(&dev_priv->rps.hw_lock);
 		/* Quoting Art Runyan: "its not safe to expect any particular
 		 * value in IPS_CTL bit 31 after enabling IPS through the
-		 * mailbox." Therefore we need to defer waiting on the state
-		 * change.
-		 * TODO: need to fix this for state checker
+		 * mailbox." Moreover, the mailbox may return a bogus state,
+		 * so we need to just enable it and continue on.
 		 */
 	} else {
 		I915_WRITE(IPS_CTL, IPS_ENABLE);
@@ -3424,13 +3779,17 @@ void hsw_disable_ips(struct intel_crtc *crtc)
 		return;
 
 	assert_plane_enabled(dev_priv, crtc->plane);
-	if (IS_BROADWELL(crtc->base.dev)) {
+	if (IS_BROADWELL(dev)) {
 		mutex_lock(&dev_priv->rps.hw_lock);
 		WARN_ON(sandybridge_pcode_write(dev_priv, DISPLAY_IPS_CONTROL, 0));
 		mutex_unlock(&dev_priv->rps.hw_lock);
-	} else
+		/* wait for pcode to finish disabling IPS, which may take up to 42ms */
+		if (wait_for((I915_READ(IPS_CTL) & IPS_ENABLE) == 0, 42))
+			DRM_ERROR("Timed out waiting for IPS disable\n");
+	} else {
 		I915_WRITE(IPS_CTL, 0);
-	POSTING_READ(IPS_CTL);
+		POSTING_READ(IPS_CTL);
+	}
 
 	/* We need to wait for a vblank before we can disable the plane. */
 	intel_wait_for_vblank(dev, crtc->pipe);
@@ -3465,7 +3824,7 @@ static void intel_crtc_load_lut(struct drm_crtc *crtc)
 	/* Workaround : Do not read or write the pipe palette/gamma data while
 	 * GAMMA_MODE is configured for split gamma and IPS_CTL has IPS enabled.
 	 */
-	if (intel_crtc->config.ips_enabled &&
+	if (IS_HASWELL(dev) && intel_crtc->config.ips_enabled &&
 	    ((I915_READ(GAMMA_MODE(pipe)) & GAMMA_MODE_MODE_MASK) ==
 	     GAMMA_MODE_MODE_SPLIT)) {
 		hsw_disable_ips(intel_crtc);
@@ -3483,6 +3842,94 @@ static void intel_crtc_load_lut(struct drm_crtc *crtc)
 		hsw_enable_ips(intel_crtc);
 }
 
+static void intel_crtc_dpms_overlay(struct intel_crtc *intel_crtc, bool enable)
+{
+	if (!enable && intel_crtc->overlay) {
+		struct drm_device *dev = intel_crtc->base.dev;
+		struct drm_i915_private *dev_priv = dev->dev_private;
+
+		mutex_lock(&dev->struct_mutex);
+		dev_priv->mm.interruptible = false;
+		(void) intel_overlay_switch_off(intel_crtc->overlay);
+		dev_priv->mm.interruptible = true;
+		mutex_unlock(&dev->struct_mutex);
+	}
+
+	/* Let userspace switch the overlay on again. In most cases userspace
+	 * has to recompute where to put it anyway.
+	 */
+}
+
+/**
+ * i9xx_fixup_plane - ugly workaround for G45 to fire up the hardware
+ * cursor plane briefly if not already running after enabling the display
+ * plane.
+ * This workaround avoids occasional blank screens when self refresh is
+ * enabled.
+ */
+static void
+g4x_fixup_plane(struct drm_i915_private *dev_priv, enum pipe pipe)
+{
+	u32 cntl = I915_READ(CURCNTR(pipe));
+
+	if ((cntl & CURSOR_MODE) == 0) {
+		u32 fw_bcl_self = I915_READ(FW_BLC_SELF);
+
+		I915_WRITE(FW_BLC_SELF, fw_bcl_self & ~FW_BLC_SELF_EN);
+		I915_WRITE(CURCNTR(pipe), CURSOR_MODE_64_ARGB_AX);
+		intel_wait_for_vblank(dev_priv->dev, pipe);
+		I915_WRITE(CURCNTR(pipe), cntl);
+		I915_WRITE(CURBASE(pipe), I915_READ(CURBASE(pipe)));
+		I915_WRITE(FW_BLC_SELF, fw_bcl_self);
+	}
+}
+
+static void intel_crtc_enable_planes(struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	int pipe = intel_crtc->pipe;
+	int plane = intel_crtc->plane;
+
+	intel_enable_primary_hw_plane(dev_priv, plane, pipe);
+	intel_enable_planes(crtc);
+	/* The fixup needs to happen before cursor is enabled */
+	if (IS_G4X(dev))
+		g4x_fixup_plane(dev_priv, pipe);
+	intel_crtc_update_cursor(crtc, true);
+	intel_crtc_dpms_overlay(intel_crtc, true);
+
+	hsw_enable_ips(intel_crtc);
+
+	mutex_lock(&dev->struct_mutex);
+	intel_update_fbc(dev);
+	intel_edp_psr_update(dev);
+	mutex_unlock(&dev->struct_mutex);
+}
+
+static void intel_crtc_disable_planes(struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	int pipe = intel_crtc->pipe;
+	int plane = intel_crtc->plane;
+
+	intel_crtc_wait_for_pending_flips(crtc);
+	drm_crtc_vblank_off(crtc);
+
+	if (dev_priv->fbc.plane == plane)
+		intel_disable_fbc(dev);
+
+	hsw_disable_ips(intel_crtc);
+
+	intel_crtc_dpms_overlay(intel_crtc, false);
+	intel_crtc_update_cursor(crtc, false);
+	intel_disable_planes(crtc);
+	intel_disable_primary_hw_plane(dev_priv, plane, pipe);
+}
+
 static void ironlake_crtc_enable(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -3490,13 +3937,35 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
+	enum plane plane = intel_crtc->plane;
 
 	WARN_ON(!crtc->enabled);
 
 	if (intel_crtc->active)
 		return;
 
+	if (intel_crtc->config.has_pch_encoder)
+		intel_prepare_shared_dpll(intel_crtc);
+
+	if (intel_crtc->config.has_dp_encoder)
+		intel_dp_set_m_n(intel_crtc);
+
+	intel_set_pipe_timings(intel_crtc);
+
+	if (intel_crtc->config.has_pch_encoder) {
+		intel_cpu_transcoder_set_m_n(intel_crtc,
+					     &intel_crtc->config.fdi_m_n);
+	}
+
+	ironlake_set_pipeconf(crtc);
+
+	/* Set up the display plane register */
+	I915_WRITE(DSPCNTR(plane), DISPPLANE_GAMMA_ENABLE);
+	POSTING_READ(DSPCNTR(plane));
+
+	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
+					       crtc->x, crtc->y);
+
 	intel_crtc->active = true;
 
 	intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
@@ -3525,34 +3994,20 @@ static void ironlake_crtc_enable(struct drm_crtc *crtc)
 	intel_crtc_load_lut(crtc);
 
 	intel_update_watermarks(crtc);
-	intel_enable_pipe(dev_priv, pipe,
-			  intel_crtc->config.has_pch_encoder, false);
-	intel_enable_primary_plane(dev_priv, plane, pipe);
-	intel_enable_planes(crtc);
-	intel_crtc_update_cursor(crtc, true);
+	intel_enable_pipe(intel_crtc);
 
 	if (intel_crtc->config.has_pch_encoder)
 		ironlake_pch_enable(crtc);
 
-	mutex_lock(&dev->struct_mutex);
-	intel_update_fbc(dev);
-	mutex_unlock(&dev->struct_mutex);
-
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		encoder->enable(encoder);
 
 	if (HAS_PCH_CPT(dev))
 		cpt_verify_modeset(dev, intel_crtc->pipe);
 
-	/*
-	 * There seems to be a race in PCH platform hw (at least on some
-	 * outputs) where an enabled pipe still completes any pageflip right
-	 * away (as if the pipe is off) instead of waiting for vblank. As soon
-	 * as the first vblank happend, everything works as expected. Hence just
-	 * wait for one vblank before returning to avoid strange things
-	 * happening.
-	 */
-	intel_wait_for_vblank(dev, intel_crtc->pipe);
+	intel_crtc_enable_planes(crtc);
+
+	drm_crtc_vblank_on(crtc);
 }
 
 /* IPS only exists on ULT machines and is tied to pipe A. */
@@ -3561,47 +4016,6 @@ static bool hsw_crtc_supports_ips(struct intel_crtc *crtc)
 	return HAS_IPS(crtc->base.dev) && crtc->pipe == PIPE_A;
 }
 
-static void haswell_crtc_enable_planes(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
-
-	intel_enable_primary_plane(dev_priv, plane, pipe);
-	intel_enable_planes(crtc);
-	intel_crtc_update_cursor(crtc, true);
-
-	hsw_enable_ips(intel_crtc);
-
-	mutex_lock(&dev->struct_mutex);
-	intel_update_fbc(dev);
-	mutex_unlock(&dev->struct_mutex);
-}
-
-static void haswell_crtc_disable_planes(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
-
-	intel_crtc_wait_for_pending_flips(crtc);
-	drm_vblank_off(dev, pipe);
-
-	/* FBC must be disabled before disabling the plane on HSW. */
-	if (dev_priv->fbc.plane == plane)
-		intel_disable_fbc(dev);
-
-	hsw_disable_ips(intel_crtc);
-
-	intel_crtc_update_cursor(crtc, false);
-	intel_disable_planes(crtc);
-	intel_disable_primary_plane(dev_priv, plane, pipe);
-}
-
 /*
  * This implements the workaround described in the "notes" section of the mode
  * set sequence documentation. When going from no pipes or single pipe to
@@ -3615,7 +4029,7 @@ static void haswell_mode_set_planes_workaround(struct intel_crtc *crtc)
 
 	/* We want to get the other_active_crtc only if there's only 1 other
 	 * active crtc. */
-	list_for_each_entry(crtc_it, &dev->mode_config.crtc_list, base.head) {
+	for_each_intel_crtc(dev, crtc_it) {
 		if (!crtc_it->active || crtc_it == crtc)
 			continue;
 
@@ -3638,12 +4052,34 @@ static void haswell_crtc_enable(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
+	enum plane plane = intel_crtc->plane;
 
 	WARN_ON(!crtc->enabled);
 
 	if (intel_crtc->active)
 		return;
 
+	if (intel_crtc->config.has_dp_encoder)
+		intel_dp_set_m_n(intel_crtc);
+
+	intel_set_pipe_timings(intel_crtc);
+
+	if (intel_crtc->config.has_pch_encoder) {
+		intel_cpu_transcoder_set_m_n(intel_crtc,
+					     &intel_crtc->config.fdi_m_n);
+	}
+
+	haswell_set_pipeconf(crtc);
+
+	intel_set_pipe_csc(crtc);
+
+	/* Set up the display plane register */
+	I915_WRITE(DSPCNTR(plane), DISPPLANE_GAMMA_ENABLE | DISPPLANE_PIPE_CSC_ENABLE);
+	POSTING_READ(DSPCNTR(plane));
+
+	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
+					       crtc->x, crtc->y);
+
 	intel_crtc->active = true;
 
 	intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
@@ -3671,8 +4107,7 @@ static void haswell_crtc_enable(struct drm_crtc *crtc)
 	intel_ddi_enable_transcoder_func(crtc);
 
 	intel_update_watermarks(crtc);
-	intel_enable_pipe(dev_priv, pipe,
-			  intel_crtc->config.has_pch_encoder, false);
+	intel_enable_pipe(intel_crtc);
 
 	if (intel_crtc->config.has_pch_encoder)
 		lpt_pch_enable(crtc);
@@ -3685,17 +4120,9 @@ static void haswell_crtc_enable(struct drm_crtc *crtc)
 	/* If we change the relative order between pipe/planes enabling, we need
 	 * to change the workaround. */
 	haswell_mode_set_planes_workaround(intel_crtc);
-	haswell_crtc_enable_planes(crtc);
+	intel_crtc_enable_planes(crtc);
 
-	/*
-	 * There seems to be a race in PCH platform hw (at least on some
-	 * outputs) where an enabled pipe still completes any pageflip right
-	 * away (as if the pipe is off) instead of waiting for vblank. As soon
-	 * as the first vblank happend, everything works as expected. Hence just
-	 * wait for one vblank before returning to avoid strange things
-	 * happening.
-	 */
-	intel_wait_for_vblank(dev, intel_crtc->pipe);
+	drm_crtc_vblank_on(crtc);
 }
 
 static void ironlake_pfit_disable(struct intel_crtc *crtc)
@@ -3720,26 +4147,16 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
 	u32 reg, temp;
 
-
 	if (!intel_crtc->active)
 		return;
 
+	intel_crtc_disable_planes(crtc);
+
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		encoder->disable(encoder);
 
-	intel_crtc_wait_for_pending_flips(crtc);
-	drm_vblank_off(dev, pipe);
-
-	if (dev_priv->fbc.plane == plane)
-		intel_disable_fbc(dev);
-
-	intel_crtc_update_cursor(crtc, false);
-	intel_disable_planes(crtc);
-	intel_disable_primary_plane(dev_priv, plane, pipe);
-
 	if (intel_crtc->config.has_pch_encoder)
 		intel_set_pch_fifo_underrun_reporting(dev, pipe, false);
 
@@ -3783,6 +4200,7 @@ static void ironlake_crtc_disable(struct drm_crtc *crtc)
 
 	mutex_lock(&dev->struct_mutex);
 	intel_update_fbc(dev);
+	intel_edp_psr_update(dev);
 	mutex_unlock(&dev->struct_mutex);
 }
 
@@ -3798,7 +4216,7 @@ static void haswell_crtc_disable(struct drm_crtc *crtc)
 	if (!intel_crtc->active)
 		return;
 
-	haswell_crtc_disable_planes(crtc);
+	intel_crtc_disable_planes(crtc);
 
 	for_each_encoder_on_crtc(dev, crtc, encoder) {
 		intel_opregion_notify_encoder(encoder, false);
@@ -3830,6 +4248,7 @@ static void haswell_crtc_disable(struct drm_crtc *crtc)
 
 	mutex_lock(&dev->struct_mutex);
 	intel_update_fbc(dev);
+	intel_edp_psr_update(dev);
 	mutex_unlock(&dev->struct_mutex);
 }
 
@@ -3844,48 +4263,6 @@ static void haswell_crtc_off(struct drm_crtc *crtc)
 	intel_ddi_put_crtc_pll(crtc);
 }
 
-static void intel_crtc_dpms_overlay(struct intel_crtc *intel_crtc, bool enable)
-{
-	if (!enable && intel_crtc->overlay) {
-		struct drm_device *dev = intel_crtc->base.dev;
-		struct drm_i915_private *dev_priv = dev->dev_private;
-
-		mutex_lock(&dev->struct_mutex);
-		dev_priv->mm.interruptible = false;
-		(void) intel_overlay_switch_off(intel_crtc->overlay);
-		dev_priv->mm.interruptible = true;
-		mutex_unlock(&dev->struct_mutex);
-	}
-
-	/* Let userspace switch the overlay on again. In most cases userspace
-	 * has to recompute where to put it anyway.
-	 */
-}
-
-/**
- * i9xx_fixup_plane - ugly workaround for G45 to fire up the hardware
- * cursor plane briefly if not already running after enabling the display
- * plane.
- * This workaround avoids occasional blank screens when self refresh is
- * enabled.
- */
-static void
-g4x_fixup_plane(struct drm_i915_private *dev_priv, enum pipe pipe)
-{
-	u32 cntl = I915_READ(CURCNTR(pipe));
-
-	if ((cntl & CURSOR_MODE) == 0) {
-		u32 fw_bcl_self = I915_READ(FW_BLC_SELF);
-
-		I915_WRITE(FW_BLC_SELF, fw_bcl_self & ~FW_BLC_SELF_EN);
-		I915_WRITE(CURCNTR(pipe), CURSOR_MODE_64_ARGB_AX);
-		intel_wait_for_vblank(dev_priv->dev, pipe);
-		I915_WRITE(CURCNTR(pipe), cntl);
-		I915_WRITE(CURBASE(pipe), I915_READ(CURBASE(pipe)));
-		I915_WRITE(FW_BLC_SELF, fw_bcl_self);
-	}
-}
-
 static void i9xx_pfit_enable(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
@@ -3910,6 +4287,276 @@ static void i9xx_pfit_enable(struct intel_crtc *crtc)
 	I915_WRITE(BCLRPAT(crtc->pipe), 0);
 }
 
+#define for_each_power_domain(domain, mask)				\
+	for ((domain) = 0; (domain) < POWER_DOMAIN_NUM; (domain)++)	\
+		if ((1 << (domain)) & (mask))
+
+enum intel_display_power_domain
+intel_display_port_power_domain(struct intel_encoder *intel_encoder)
+{
+	struct drm_device *dev = intel_encoder->base.dev;
+	struct intel_digital_port *intel_dig_port;
+
+	switch (intel_encoder->type) {
+	case INTEL_OUTPUT_UNKNOWN:
+		/* Only DDI platforms should ever use this output type */
+		WARN_ON_ONCE(!HAS_DDI(dev));
+	case INTEL_OUTPUT_DISPLAYPORT:
+	case INTEL_OUTPUT_HDMI:
+	case INTEL_OUTPUT_EDP:
+		intel_dig_port = enc_to_dig_port(&intel_encoder->base);
+		switch (intel_dig_port->port) {
+		case PORT_A:
+			return POWER_DOMAIN_PORT_DDI_A_4_LANES;
+		case PORT_B:
+			return POWER_DOMAIN_PORT_DDI_B_4_LANES;
+		case PORT_C:
+			return POWER_DOMAIN_PORT_DDI_C_4_LANES;
+		case PORT_D:
+			return POWER_DOMAIN_PORT_DDI_D_4_LANES;
+		default:
+			WARN_ON_ONCE(1);
+			return POWER_DOMAIN_PORT_OTHER;
+		}
+	case INTEL_OUTPUT_ANALOG:
+		return POWER_DOMAIN_PORT_CRT;
+	case INTEL_OUTPUT_DSI:
+		return POWER_DOMAIN_PORT_DSI;
+	default:
+		return POWER_DOMAIN_PORT_OTHER;
+	}
+}
+
+static unsigned long get_crtc_power_domains(struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	struct intel_encoder *intel_encoder;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	enum pipe pipe = intel_crtc->pipe;
+	bool pfit_enabled = intel_crtc->config.pch_pfit.enabled;
+	unsigned long mask;
+	enum transcoder transcoder;
+
+	transcoder = intel_pipe_to_cpu_transcoder(dev->dev_private, pipe);
+
+	mask = BIT(POWER_DOMAIN_PIPE(pipe));
+	mask |= BIT(POWER_DOMAIN_TRANSCODER(transcoder));
+	if (pfit_enabled)
+		mask |= BIT(POWER_DOMAIN_PIPE_PANEL_FITTER(pipe));
+
+	for_each_encoder_on_crtc(dev, crtc, intel_encoder)
+		mask |= BIT(intel_display_port_power_domain(intel_encoder));
+
+	return mask;
+}
+
+void intel_display_set_init_power(struct drm_i915_private *dev_priv,
+				  bool enable)
+{
+	if (dev_priv->power_domains.init_power_on == enable)
+		return;
+
+	if (enable)
+		intel_display_power_get(dev_priv, POWER_DOMAIN_INIT);
+	else
+		intel_display_power_put(dev_priv, POWER_DOMAIN_INIT);
+
+	dev_priv->power_domains.init_power_on = enable;
+}
+
+static void modeset_update_crtc_power_domains(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long pipe_domains[I915_MAX_PIPES] = { 0, };
+	struct intel_crtc *crtc;
+
+	/*
+	 * First get all needed power domains, then put all unneeded, to avoid
+	 * any unnecessary toggling of the power wells.
+	 */
+	for_each_intel_crtc(dev, crtc) {
+		enum intel_display_power_domain domain;
+
+		if (!crtc->base.enabled)
+			continue;
+
+		pipe_domains[crtc->pipe] = get_crtc_power_domains(&crtc->base);
+
+		for_each_power_domain(domain, pipe_domains[crtc->pipe])
+			intel_display_power_get(dev_priv, domain);
+	}
+
+	for_each_intel_crtc(dev, crtc) {
+		enum intel_display_power_domain domain;
+
+		for_each_power_domain(domain, crtc->enabled_power_domains)
+			intel_display_power_put(dev_priv, domain);
+
+		crtc->enabled_power_domains = pipe_domains[crtc->pipe];
+	}
+
+	intel_display_set_init_power(dev_priv, false);
+}
+
+int valleyview_get_vco(struct drm_i915_private *dev_priv)
+{
+	int hpll_freq, vco_freq[] = { 800, 1600, 2000, 2400 };
+
+	/* Obtain SKU information */
+	mutex_lock(&dev_priv->dpio_lock);
+	hpll_freq = vlv_cck_read(dev_priv, CCK_FUSE_REG) &
+		CCK_FUSE_HPLL_FREQ_MASK;
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	return vco_freq[hpll_freq];
+}
+
+/* Adjust CDclk dividers to allow high res or save power if possible */
+static void valleyview_set_cdclk(struct drm_device *dev, int cdclk)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val, cmd;
+
+	WARN_ON(valleyview_cur_cdclk(dev_priv) != dev_priv->vlv_cdclk_freq);
+	dev_priv->vlv_cdclk_freq = cdclk;
+
+	if (cdclk >= 320) /* jump to highest voltage for 400MHz too */
+		cmd = 2;
+	else if (cdclk == 266)
+		cmd = 1;
+	else
+		cmd = 0;
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+	val = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ);
+	val &= ~DSPFREQGUAR_MASK;
+	val |= (cmd << DSPFREQGUAR_SHIFT);
+	vlv_punit_write(dev_priv, PUNIT_REG_DSPFREQ, val);
+	if (wait_for((vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) &
+		      DSPFREQSTAT_MASK) == (cmd << DSPFREQSTAT_SHIFT),
+		     50)) {
+		DRM_ERROR("timed out waiting for CDclk change\n");
+	}
+	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	if (cdclk == 400) {
+		u32 divider, vco;
+
+		vco = valleyview_get_vco(dev_priv);
+		divider = ((vco << 1) / cdclk) - 1;
+
+		mutex_lock(&dev_priv->dpio_lock);
+		/* adjust cdclk divider */
+		val = vlv_cck_read(dev_priv, CCK_DISPLAY_CLOCK_CONTROL);
+		val &= ~0xf;
+		val |= divider;
+		vlv_cck_write(dev_priv, CCK_DISPLAY_CLOCK_CONTROL, val);
+		mutex_unlock(&dev_priv->dpio_lock);
+	}
+
+	mutex_lock(&dev_priv->dpio_lock);
+	/* adjust self-refresh exit latency value */
+	val = vlv_bunit_read(dev_priv, BUNIT_REG_BISOC);
+	val &= ~0x7f;
+
+	/*
+	 * For high bandwidth configs, we set a higher latency in the bunit
+	 * so that the core display fetch happens in time to avoid underruns.
+	 */
+	if (cdclk == 400)
+		val |= 4500 / 250; /* 4.5 usec */
+	else
+		val |= 3000 / 250; /* 3.0 usec */
+	vlv_bunit_write(dev_priv, BUNIT_REG_BISOC, val);
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	/* Since we changed the CDclk, we need to update the GMBUSFREQ too */
+	intel_i2c_reset(dev);
+}
+
+int valleyview_cur_cdclk(struct drm_i915_private *dev_priv)
+{
+	int cur_cdclk, vco;
+	int divider;
+
+	vco = valleyview_get_vco(dev_priv);
+
+	mutex_lock(&dev_priv->dpio_lock);
+	divider = vlv_cck_read(dev_priv, CCK_DISPLAY_CLOCK_CONTROL);
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	divider &= 0xf;
+
+	cur_cdclk = (vco << 1) / (divider + 1);
+
+	return cur_cdclk;
+}
+
+static int valleyview_calc_cdclk(struct drm_i915_private *dev_priv,
+				 int max_pixclk)
+{
+	/*
+	 * Really only a few cases to deal with, as only 4 CDclks are supported:
+	 *   200MHz
+	 *   267MHz
+	 *   320MHz
+	 *   400MHz
+	 * So we check to see whether we're above 90% of the lower bin and
+	 * adjust if needed.
+	 */
+	if (max_pixclk > 288000) {
+		return 400;
+	} else if (max_pixclk > 240000) {
+		return 320;
+	} else
+		return 266;
+	/* Looks like the 200MHz CDclk freq doesn't work on some configs */
+}
+
+/* compute the max pixel clock for new configuration */
+static int intel_mode_max_pixclk(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct intel_crtc *intel_crtc;
+	int max_pixclk = 0;
+
+	for_each_intel_crtc(dev, intel_crtc) {
+		if (intel_crtc->new_enabled)
+			max_pixclk = max(max_pixclk,
+					 intel_crtc->new_config->adjusted_mode.crtc_clock);
+	}
+
+	return max_pixclk;
+}
+
+static void valleyview_modeset_global_pipes(struct drm_device *dev,
+					    unsigned *prepare_pipes)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc;
+	int max_pixclk = intel_mode_max_pixclk(dev_priv);
+
+	if (valleyview_calc_cdclk(dev_priv, max_pixclk) ==
+	    dev_priv->vlv_cdclk_freq)
+		return;
+
+	/* disable/enable all currently active pipes while we change cdclk */
+	for_each_intel_crtc(dev, intel_crtc)
+		if (intel_crtc->base.enabled)
+			*prepare_pipes |= (1 << intel_crtc->pipe);
+}
+
+static void valleyview_modeset_global_resources(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int max_pixclk = intel_mode_max_pixclk(dev_priv);
+	int req_cdclk = valleyview_calc_cdclk(dev_priv, max_pixclk);
+
+	if (req_cdclk != dev_priv->vlv_cdclk_freq)
+		valleyview_set_cdclk(dev, req_cdclk);
+	modeset_update_crtc_power_domains(dev);
+}
+
 static void valleyview_crtc_enable(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -3919,22 +4566,56 @@ static void valleyview_crtc_enable(struct drm_crtc *crtc)
 	int pipe = intel_crtc->pipe;
 	int plane = intel_crtc->plane;
 	bool is_dsi;
+	u32 dspcntr;
 
 	WARN_ON(!crtc->enabled);
 
 	if (intel_crtc->active)
 		return;
 
+	is_dsi = intel_pipe_has_type(crtc, INTEL_OUTPUT_DSI);
+
+	if (!is_dsi && !IS_CHERRYVIEW(dev))
+		vlv_prepare_pll(intel_crtc);
+
+	/* Set up the display plane register */
+	dspcntr = DISPPLANE_GAMMA_ENABLE;
+
+	if (intel_crtc->config.has_dp_encoder)
+		intel_dp_set_m_n(intel_crtc);
+
+	intel_set_pipe_timings(intel_crtc);
+
+	/* pipesrc and dspsize control the size that is scaled from,
+	 * which should always be the user's requested size.
+	 */
+	I915_WRITE(DSPSIZE(plane),
+		   ((intel_crtc->config.pipe_src_h - 1) << 16) |
+		   (intel_crtc->config.pipe_src_w - 1));
+	I915_WRITE(DSPPOS(plane), 0);
+
+	i9xx_set_pipeconf(intel_crtc);
+
+	I915_WRITE(DSPCNTR(plane), dspcntr);
+	POSTING_READ(DSPCNTR(plane));
+
+	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
+					       crtc->x, crtc->y);
+
 	intel_crtc->active = true;
 
+	intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
+
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		if (encoder->pre_pll_enable)
 			encoder->pre_pll_enable(encoder);
 
-	is_dsi = intel_pipe_has_type(crtc, INTEL_OUTPUT_DSI);
-
-	if (!is_dsi)
-		vlv_enable_pll(intel_crtc);
+	if (!is_dsi) {
+		if (IS_CHERRYVIEW(dev))
+			chv_enable_pll(intel_crtc);
+		else
+			vlv_enable_pll(intel_crtc);
+	}
 
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		if (encoder->pre_enable)
@@ -3945,15 +4626,26 @@ static void valleyview_crtc_enable(struct drm_crtc *crtc)
 	intel_crtc_load_lut(crtc);
 
 	intel_update_watermarks(crtc);
-	intel_enable_pipe(dev_priv, pipe, false, is_dsi);
-	intel_enable_primary_plane(dev_priv, plane, pipe);
-	intel_enable_planes(crtc);
-	intel_crtc_update_cursor(crtc, true);
-
-	intel_update_fbc(dev);
+	intel_enable_pipe(intel_crtc);
 
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		encoder->enable(encoder);
+
+	intel_crtc_enable_planes(crtc);
+
+	drm_crtc_vblank_on(crtc);
+
+	/* Underruns don't raise interrupts, so check manually. */
+	i9xx_check_fifo_underruns(dev);
+}
+
+static void i9xx_set_pll_dividers(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	I915_WRITE(FP0(crtc->pipe), crtc->config.dpll_hw_state.fp0);
+	I915_WRITE(FP1(crtc->pipe), crtc->config.dpll_hw_state.fp1);
 }
 
 static void i9xx_crtc_enable(struct drm_crtc *crtc)
@@ -3964,14 +4656,49 @@ static void i9xx_crtc_enable(struct drm_crtc *crtc)
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
 	int plane = intel_crtc->plane;
+	u32 dspcntr;
 
 	WARN_ON(!crtc->enabled);
 
 	if (intel_crtc->active)
 		return;
 
+	i9xx_set_pll_dividers(intel_crtc);
+
+	/* Set up the display plane register */
+	dspcntr = DISPPLANE_GAMMA_ENABLE;
+
+	if (pipe == 0)
+		dspcntr &= ~DISPPLANE_SEL_PIPE_MASK;
+	else
+		dspcntr |= DISPPLANE_SEL_PIPE_B;
+
+	if (intel_crtc->config.has_dp_encoder)
+		intel_dp_set_m_n(intel_crtc);
+
+	intel_set_pipe_timings(intel_crtc);
+
+	/* pipesrc and dspsize control the size that is scaled from,
+	 * which should always be the user's requested size.
+	 */
+	I915_WRITE(DSPSIZE(plane),
+		   ((intel_crtc->config.pipe_src_h - 1) << 16) |
+		   (intel_crtc->config.pipe_src_w - 1));
+	I915_WRITE(DSPPOS(plane), 0);
+
+	i9xx_set_pipeconf(intel_crtc);
+
+	I915_WRITE(DSPCNTR(plane), dspcntr);
+	POSTING_READ(DSPCNTR(plane));
+
+	dev_priv->display.update_primary_plane(crtc, crtc->primary->fb,
+					       crtc->x, crtc->y);
+
 	intel_crtc->active = true;
 
+	if (!IS_GEN2(dev))
+		intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
+
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		if (encoder->pre_enable)
 			encoder->pre_enable(encoder);
@@ -3983,21 +4710,27 @@ static void i9xx_crtc_enable(struct drm_crtc *crtc)
 	intel_crtc_load_lut(crtc);
 
 	intel_update_watermarks(crtc);
-	intel_enable_pipe(dev_priv, pipe, false, false);
-	intel_enable_primary_plane(dev_priv, plane, pipe);
-	intel_enable_planes(crtc);
-	/* The fixup needs to happen before cursor is enabled */
-	if (IS_G4X(dev))
-		g4x_fixup_plane(dev_priv, pipe);
-	intel_crtc_update_cursor(crtc, true);
-
-	/* Give the overlay scaler a chance to enable if it's on this pipe */
-	intel_crtc_dpms_overlay(intel_crtc, true);
-
-	intel_update_fbc(dev);
+	intel_enable_pipe(intel_crtc);
 
 	for_each_encoder_on_crtc(dev, crtc, encoder)
 		encoder->enable(encoder);
+
+	intel_crtc_enable_planes(crtc);
+
+	/*
+	 * Gen2 reports pipe underruns whenever all planes are disabled.
+	 * So don't enable underrun reporting before at least some planes
+	 * are enabled.
+	 * FIXME: Need to fix the logic to work when we turn off all planes
+	 * but leave the pipe running.
+	 */
+	if (IS_GEN2(dev))
+		intel_set_cpu_fifo_underrun_reporting(dev, pipe, true);
+
+	drm_crtc_vblank_on(crtc);
+
+	/* Underruns don't raise interrupts, so check manually. */
+	i9xx_check_fifo_underruns(dev);
 }
 
 static void i9xx_pfit_disable(struct intel_crtc *crtc)
@@ -4022,25 +4755,30 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_encoder *encoder;
 	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
 
 	if (!intel_crtc->active)
 		return;
 
-	for_each_encoder_on_crtc(dev, crtc, encoder)
-		encoder->disable(encoder);
+	/*
+	 * Gen2 reports pipe underruns whenever all planes are disabled.
+	 * So diasble underrun reporting before all the planes get disabled.
+	 * FIXME: Need to fix the logic to work when we turn off all planes
+	 * but leave the pipe running.
+	 */
+	if (IS_GEN2(dev))
+		intel_set_cpu_fifo_underrun_reporting(dev, pipe, false);
 
-	/* Give the overlay scaler a chance to disable if it's on this pipe */
-	intel_crtc_wait_for_pending_flips(crtc);
-	drm_vblank_off(dev, pipe);
+	intel_crtc_disable_planes(crtc);
 
-	if (dev_priv->fbc.plane == plane)
-		intel_disable_fbc(dev);
+	for_each_encoder_on_crtc(dev, crtc, encoder)
+		encoder->disable(encoder);
 
-	intel_crtc_dpms_overlay(intel_crtc, false);
-	intel_crtc_update_cursor(crtc, false);
-	intel_disable_planes(crtc);
-	intel_disable_primary_plane(dev_priv, plane, pipe);
+	/*
+	 * On gen2 planes are double buffered but the pipe isn't, so we must
+	 * wait for planes to fully turn off before disabling the pipe.
+	 */
+	if (IS_GEN2(dev))
+		intel_wait_for_vblank(dev, pipe);
 
 	intel_disable_pipe(dev_priv, pipe);
 
@@ -4050,15 +4788,25 @@ static void i9xx_crtc_disable(struct drm_crtc *crtc)
 		if (encoder->post_disable)
 			encoder->post_disable(encoder);
 
-	if (IS_VALLEYVIEW(dev) && !intel_pipe_has_type(crtc, INTEL_OUTPUT_DSI))
-		vlv_disable_pll(dev_priv, pipe);
-	else if (!IS_VALLEYVIEW(dev))
-		i9xx_disable_pll(dev_priv, pipe);
+	if (!intel_pipe_has_type(crtc, INTEL_OUTPUT_DSI)) {
+		if (IS_CHERRYVIEW(dev))
+			chv_disable_pll(dev_priv, pipe);
+		else if (IS_VALLEYVIEW(dev))
+			vlv_disable_pll(dev_priv, pipe);
+		else
+			i9xx_disable_pll(dev_priv, pipe);
+	}
+
+	if (!IS_GEN2(dev))
+		intel_set_cpu_fifo_underrun_reporting(dev, pipe, false);
 
 	intel_crtc->active = false;
 	intel_update_watermarks(crtc);
 
+	mutex_lock(&dev->struct_mutex);
 	intel_update_fbc(dev);
+	intel_edp_psr_update(dev);
+	mutex_unlock(&dev->struct_mutex);
 }
 
 static void i9xx_crtc_off(struct drm_crtc *crtc)
@@ -4121,13 +4869,11 @@ static void intel_crtc_disable(struct drm_crtc *crtc)
 	struct drm_device *dev = crtc->dev;
 	struct drm_connector *connector;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 
 	/* crtc should still be enabled when we disable it. */
 	WARN_ON(!crtc->enabled);
 
 	dev_priv->display.crtc_disable(crtc);
-	intel_crtc->eld_vld = false;
 	intel_crtc_update_sarea(crtc, false);
 	dev_priv->display.off(crtc);
 
@@ -4135,11 +4881,11 @@ static void intel_crtc_disable(struct drm_crtc *crtc)
 	assert_cursor_disabled(dev_priv, to_intel_crtc(crtc)->pipe);
 	assert_pipe_disabled(dev->dev_private, to_intel_crtc(crtc)->pipe);
 
-	if (crtc->fb) {
+	if (crtc->primary->fb) {
 		mutex_lock(&dev->struct_mutex);
-		intel_unpin_fb_obj(to_intel_framebuffer(crtc->fb)->obj);
+		intel_unpin_fb_obj(to_intel_framebuffer(crtc->primary->fb)->obj);
 		mutex_unlock(&dev->struct_mutex);
-		crtc->fb = NULL;
+		crtc->primary->fb = NULL;
 	}
 
 	/* Update computed state. */
@@ -4191,7 +4937,7 @@ static void intel_connector_check_state(struct intel_connector *connector)
 
 		DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
 			      connector->base.base.id,
-			      drm_get_connector_name(&connector->base));
+			      connector->base.name);
 
 		WARN(connector->base.dpms == DRM_MODE_DPMS_OFF,
 		     "wrong connector dpms state\n");
@@ -4353,7 +5099,7 @@ retry:
 static void hsw_compute_ips_config(struct intel_crtc *crtc,
 				   struct intel_crtc_config *pipe_config)
 {
-	pipe_config->ips_enabled = i915_enable_ips &&
+	pipe_config->ips_enabled = i915.enable_ips &&
 				   hsw_crtc_supports_ips(crtc) &&
 				   pipe_config->pipe_bpp <= 24;
 }
@@ -4554,8 +5300,8 @@ intel_link_compute_m_n(int bits_per_pixel, int nlanes,
 
 static inline bool intel_panel_use_ssc(struct drm_i915_private *dev_priv)
 {
-	if (i915_panel_use_ssc >= 0)
-		return i915_panel_use_ssc != 0;
+	if (i915.panel_use_ssc >= 0)
+		return i915.panel_use_ssc != 0;
 	return dev_priv->vbt.lvds_use_ssc
 		&& !(dev_priv->quirks & QUIRK_LVDS_SSC_DISABLE);
 }
@@ -4570,9 +5316,8 @@ static int i9xx_get_refclk(struct drm_crtc *crtc, int num_connectors)
 		refclk = 100000;
 	} else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) &&
 	    intel_panel_use_ssc(dev_priv) && num_connectors < 2) {
-		refclk = dev_priv->vbt.lvds_ssc_freq * 1000;
-		DRM_DEBUG_KMS("using SSC reference clock of %d MHz\n",
-			      refclk / 1000);
+		refclk = dev_priv->vbt.lvds_ssc_freq;
+		DRM_DEBUG_KMS("using SSC reference clock of %d kHz\n", refclk);
 	} else if (!IS_GEN2(dev)) {
 		refclk = 96000;
 	} else {
@@ -4596,8 +5341,6 @@ static void i9xx_update_pll_dividers(struct intel_crtc *crtc,
 				     intel_clock_t *reduced_clock)
 {
 	struct drm_device *dev = crtc->base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int pipe = crtc->pipe;
 	u32 fp, fp2 = 0;
 
 	if (IS_PINEVIEW(dev)) {
@@ -4610,17 +5353,14 @@ static void i9xx_update_pll_dividers(struct intel_crtc *crtc,
 			fp2 = i9xx_dpll_compute_fp(reduced_clock);
 	}
 
-	I915_WRITE(FP0(pipe), fp);
 	crtc->config.dpll_hw_state.fp0 = fp;
 
 	crtc->lowfreq_avail = false;
 	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_LVDS) &&
-	    reduced_clock && i915_powersave) {
-		I915_WRITE(FP1(pipe), fp2);
+	    reduced_clock && i915.powersave) {
 		crtc->config.dpll_hw_state.fp1 = fp2;
 		crtc->lowfreq_avail = true;
 	} else {
-		I915_WRITE(FP1(pipe), fp);
 		crtc->config.dpll_hw_state.fp1 = fp;
 	}
 }
@@ -4634,24 +5374,24 @@ static void vlv_pllb_recal_opamp(struct drm_i915_private *dev_priv, enum pipe
 	 * PLLB opamp always calibrates to max value of 0x3f, force enable it
 	 * and set it to a reasonable value instead.
 	 */
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_IREF(1));
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW9(1));
 	reg_val &= 0xffffff00;
 	reg_val |= 0x00000030;
-	vlv_dpio_write(dev_priv, pipe, DPIO_IREF(1), reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW9(1), reg_val);
 
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_CALIBRATION);
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_REF_DW13);
 	reg_val &= 0x8cffffff;
 	reg_val = 0x8c000000;
-	vlv_dpio_write(dev_priv, pipe, DPIO_CALIBRATION, reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_REF_DW13, reg_val);
 
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_IREF(1));
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW9(1));
 	reg_val &= 0xffffff00;
-	vlv_dpio_write(dev_priv, pipe, DPIO_IREF(1), reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW9(1), reg_val);
 
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_CALIBRATION);
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_REF_DW13);
 	reg_val &= 0x00ffffff;
 	reg_val |= 0xb0000000;
-	vlv_dpio_write(dev_priv, pipe, DPIO_CALIBRATION, reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_REF_DW13, reg_val);
 }
 
 static void intel_pch_transcoder_set_m_n(struct intel_crtc *crtc,
@@ -4698,12 +5438,34 @@ static void intel_dp_set_m_n(struct intel_crtc *crtc)
 
 static void vlv_update_pll(struct intel_crtc *crtc)
 {
+	u32 dpll, dpll_md;
+
+	/*
+	 * Enable DPIO clock input. We should never disable the reference
+	 * clock for pipe B, since VGA hotplug / manual detection depends
+	 * on it.
+	 */
+	dpll = DPLL_EXT_BUFFER_ENABLE_VLV | DPLL_REFA_CLK_ENABLE_VLV |
+		DPLL_VGA_MODE_DIS | DPLL_INTEGRATED_CLOCK_VLV;
+	/* We should never disable this, set it here for state tracking */
+	if (crtc->pipe == PIPE_B)
+		dpll |= DPLL_INTEGRATED_CRI_CLK_VLV;
+	dpll |= DPLL_VCO_ENABLE;
+	crtc->config.dpll_hw_state.dpll = dpll;
+
+	dpll_md = (crtc->config.pixel_multiplier - 1)
+		<< DPLL_MD_UDI_MULTIPLIER_SHIFT;
+	crtc->config.dpll_hw_state.dpll_md = dpll_md;
+}
+
+static void vlv_prepare_pll(struct intel_crtc *crtc)
+{
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipe = crtc->pipe;
-	u32 dpll, mdiv;
+	u32 mdiv;
 	u32 bestn, bestm1, bestm2, bestp1, bestp2;
-	u32 coreclk, reg_val, dpll_md;
+	u32 coreclk, reg_val;
 
 	mutex_lock(&dev_priv->dpio_lock);
 
@@ -4716,19 +5478,19 @@ static void vlv_update_pll(struct intel_crtc *crtc)
 	/* See eDP HDMI DPIO driver vbios notes doc */
 
 	/* PLL B needs special handling */
-	if (pipe)
+	if (pipe == PIPE_B)
 		vlv_pllb_recal_opamp(dev_priv, pipe);
 
 	/* Set up Tx target for periodic Rcomp update */
-	vlv_dpio_write(dev_priv, pipe, DPIO_IREF_BCAST, 0x0100000f);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW9_BCAST, 0x0100000f);
 
 	/* Disable target IRef on PLL */
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_IREF_CTL(pipe));
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW8(pipe));
 	reg_val &= 0x00ffffff;
-	vlv_dpio_write(dev_priv, pipe, DPIO_IREF_CTL(pipe), reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW8(pipe), reg_val);
 
 	/* Disable fast lock */
-	vlv_dpio_write(dev_priv, pipe, DPIO_FASTCLK_DISABLE, 0x610);
+	vlv_dpio_write(dev_priv, pipe, VLV_CMN_DW0, 0x610);
 
 	/* Set idtafcrecal before PLL is enabled */
 	mdiv = ((bestm1 << DPIO_M1DIV_SHIFT) | (bestm2 & DPIO_M2DIV_MASK));
@@ -4742,64 +5504,126 @@ static void vlv_update_pll(struct intel_crtc *crtc)
 	 * Note: don't use the DAC post divider as it seems unstable.
 	 */
 	mdiv |= (DPIO_POST_DIV_HDMIDP << DPIO_POST_DIV_SHIFT);
-	vlv_dpio_write(dev_priv, pipe, DPIO_DIV(pipe), mdiv);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW3(pipe), mdiv);
 
 	mdiv |= DPIO_ENABLE_CALIBRATION;
-	vlv_dpio_write(dev_priv, pipe, DPIO_DIV(pipe), mdiv);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW3(pipe), mdiv);
 
 	/* Set HBR and RBR LPF coefficients */
 	if (crtc->config.port_clock == 162000 ||
 	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_ANALOG) ||
 	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_HDMI))
-		vlv_dpio_write(dev_priv, pipe, DPIO_LPF_COEFF(pipe),
+		vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW10(pipe),
 				 0x009f0003);
 	else
-		vlv_dpio_write(dev_priv, pipe, DPIO_LPF_COEFF(pipe),
+		vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW10(pipe),
 				 0x00d0000f);
 
 	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_EDP) ||
 	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DISPLAYPORT)) {
 		/* Use SSC source */
-		if (!pipe)
-			vlv_dpio_write(dev_priv, pipe, DPIO_REFSFR(pipe),
+		if (pipe == PIPE_A)
+			vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW5(pipe),
 					 0x0df40000);
 		else
-			vlv_dpio_write(dev_priv, pipe, DPIO_REFSFR(pipe),
+			vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW5(pipe),
 					 0x0df70000);
 	} else { /* HDMI or VGA */
 		/* Use bend source */
-		if (!pipe)
-			vlv_dpio_write(dev_priv, pipe, DPIO_REFSFR(pipe),
+		if (pipe == PIPE_A)
+			vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW5(pipe),
 					 0x0df70000);
 		else
-			vlv_dpio_write(dev_priv, pipe, DPIO_REFSFR(pipe),
+			vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW5(pipe),
 					 0x0df40000);
 	}
 
-	coreclk = vlv_dpio_read(dev_priv, pipe, DPIO_CORE_CLK(pipe));
+	coreclk = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW7(pipe));
 	coreclk = (coreclk & 0x0000ff00) | 0x01c00000;
 	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DISPLAYPORT) ||
 	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_EDP))
 		coreclk |= 0x01000000;
-	vlv_dpio_write(dev_priv, pipe, DPIO_CORE_CLK(pipe), coreclk);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW7(pipe), coreclk);
 
-	vlv_dpio_write(dev_priv, pipe, DPIO_PLL_CML(pipe), 0x87871000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW11(pipe), 0x87871000);
+	mutex_unlock(&dev_priv->dpio_lock);
+}
 
-	/* Enable DPIO clock input */
-	dpll = DPLL_EXT_BUFFER_ENABLE_VLV | DPLL_REFA_CLK_ENABLE_VLV |
-		DPLL_VGA_MODE_DIS | DPLL_INTEGRATED_CLOCK_VLV;
-	/* We should never disable this, set it here for state tracking */
-	if (pipe == PIPE_B)
-		dpll |= DPLL_INTEGRATED_CRI_CLK_VLV;
-	dpll |= DPLL_VCO_ENABLE;
-	crtc->config.dpll_hw_state.dpll = dpll;
+static void chv_update_pll(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe = crtc->pipe;
+	int dpll_reg = DPLL(crtc->pipe);
+	enum dpio_channel port = vlv_pipe_to_channel(pipe);
+	u32 loopfilter, intcoeff;
+	u32 bestn, bestm1, bestm2, bestp1, bestp2, bestm2_frac;
+	int refclk;
 
-	dpll_md = (crtc->config.pixel_multiplier - 1)
-		<< DPLL_MD_UDI_MULTIPLIER_SHIFT;
-	crtc->config.dpll_hw_state.dpll_md = dpll_md;
+	crtc->config.dpll_hw_state.dpll = DPLL_SSC_REF_CLOCK_CHV |
+		DPLL_REFA_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS |
+		DPLL_VCO_ENABLE;
+	if (pipe != PIPE_A)
+		crtc->config.dpll_hw_state.dpll |= DPLL_INTEGRATED_CRI_CLK_VLV;
+
+	crtc->config.dpll_hw_state.dpll_md =
+		(crtc->config.pixel_multiplier - 1) << DPLL_MD_UDI_MULTIPLIER_SHIFT;
 
-	if (crtc->config.has_dp_encoder)
-		intel_dp_set_m_n(crtc);
+	bestn = crtc->config.dpll.n;
+	bestm2_frac = crtc->config.dpll.m2 & 0x3fffff;
+	bestm1 = crtc->config.dpll.m1;
+	bestm2 = crtc->config.dpll.m2 >> 22;
+	bestp1 = crtc->config.dpll.p1;
+	bestp2 = crtc->config.dpll.p2;
+
+	/*
+	 * Enable Refclk and SSC
+	 */
+	I915_WRITE(dpll_reg,
+		   crtc->config.dpll_hw_state.dpll & ~DPLL_VCO_ENABLE);
+
+	mutex_lock(&dev_priv->dpio_lock);
+
+	/* p1 and p2 divider */
+	vlv_dpio_write(dev_priv, pipe, CHV_CMN_DW13(port),
+			5 << DPIO_CHV_S1_DIV_SHIFT |
+			bestp1 << DPIO_CHV_P1_DIV_SHIFT |
+			bestp2 << DPIO_CHV_P2_DIV_SHIFT |
+			1 << DPIO_CHV_K_DIV_SHIFT);
+
+	/* Feedback post-divider - m2 */
+	vlv_dpio_write(dev_priv, pipe, CHV_PLL_DW0(port), bestm2);
+
+	/* Feedback refclk divider - n and m1 */
+	vlv_dpio_write(dev_priv, pipe, CHV_PLL_DW1(port),
+			DPIO_CHV_M1_DIV_BY_2 |
+			1 << DPIO_CHV_N_DIV_SHIFT);
+
+	/* M2 fraction division */
+	vlv_dpio_write(dev_priv, pipe, CHV_PLL_DW2(port), bestm2_frac);
+
+	/* M2 fraction division enable */
+	vlv_dpio_write(dev_priv, pipe, CHV_PLL_DW3(port),
+		       DPIO_CHV_FRAC_DIV_EN |
+		       (2 << DPIO_CHV_FEEDFWD_GAIN_SHIFT));
+
+	/* Loop filter */
+	refclk = i9xx_get_refclk(&crtc->base, 0);
+	loopfilter = 5 << DPIO_CHV_PROP_COEFF_SHIFT |
+		2 << DPIO_CHV_GAIN_CTRL_SHIFT;
+	if (refclk == 100000)
+		intcoeff = 11;
+	else if (refclk == 38400)
+		intcoeff = 10;
+	else
+		intcoeff = 9;
+	loopfilter |= intcoeff << DPIO_CHV_INT_COEFF_SHIFT;
+	vlv_dpio_write(dev_priv, pipe, CHV_PLL_DW6(port), loopfilter);
+
+	/* AFC Recal */
+	vlv_dpio_write(dev_priv, pipe, CHV_CMN_DW14(port),
+			vlv_dpio_read(dev_priv, pipe, CHV_CMN_DW14(port)) |
+			DPIO_AFC_RECAL);
 
 	mutex_unlock(&dev_priv->dpio_lock);
 }
@@ -4878,9 +5702,6 @@ static void i9xx_update_pll(struct intel_crtc *crtc,
 			<< DPLL_MD_UDI_MULTIPLIER_SHIFT;
 		crtc->config.dpll_hw_state.dpll_md = dpll_md;
 	}
-
-	if (crtc->config.has_dp_encoder)
-		intel_dp_set_m_n(crtc);
 }
 
 static void i8xx_update_pll(struct intel_crtc *crtc,
@@ -4928,21 +5749,26 @@ static void intel_set_pipe_timings(struct intel_crtc *intel_crtc)
 	enum transcoder cpu_transcoder = intel_crtc->config.cpu_transcoder;
 	struct drm_display_mode *adjusted_mode =
 		&intel_crtc->config.adjusted_mode;
-	uint32_t vsyncshift, crtc_vtotal, crtc_vblank_end;
+	uint32_t crtc_vtotal, crtc_vblank_end;
+	int vsyncshift = 0;
 
 	/* We need to be careful not to changed the adjusted mode, for otherwise
 	 * the hw state checker will get angry at the mismatch. */
 	crtc_vtotal = adjusted_mode->crtc_vtotal;
 	crtc_vblank_end = adjusted_mode->crtc_vblank_end;
 
-	if (!IS_GEN2(dev) && adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE) {
+	if (adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE) {
 		/* the chip adds 2 halflines automatically */
 		crtc_vtotal -= 1;
 		crtc_vblank_end -= 1;
-		vsyncshift = adjusted_mode->crtc_hsync_start
-			     - adjusted_mode->crtc_htotal / 2;
-	} else {
-		vsyncshift = 0;
+
+		if (intel_pipe_has_type(&intel_crtc->base, INTEL_OUTPUT_SDVO))
+			vsyncshift = (adjusted_mode->crtc_htotal - 1) / 2;
+		else
+			vsyncshift = adjusted_mode->crtc_hsync_start -
+				adjusted_mode->crtc_htotal / 2;
+		if (vsyncshift < 0)
+			vsyncshift += adjusted_mode->crtc_htotal;
 	}
 
 	if (INTEL_INFO(dev)->gen > 3)
@@ -5026,25 +5852,23 @@ static void intel_get_pipe_timings(struct intel_crtc *crtc,
 	pipe_config->requested_mode.hdisplay = pipe_config->pipe_src_w;
 }
 
-static void intel_crtc_mode_from_pipe_config(struct intel_crtc *intel_crtc,
-					     struct intel_crtc_config *pipe_config)
+void intel_mode_from_pipe_config(struct drm_display_mode *mode,
+				 struct intel_crtc_config *pipe_config)
 {
-	struct drm_crtc *crtc = &intel_crtc->base;
-
-	crtc->mode.hdisplay = pipe_config->adjusted_mode.crtc_hdisplay;
-	crtc->mode.htotal = pipe_config->adjusted_mode.crtc_htotal;
-	crtc->mode.hsync_start = pipe_config->adjusted_mode.crtc_hsync_start;
-	crtc->mode.hsync_end = pipe_config->adjusted_mode.crtc_hsync_end;
+	mode->hdisplay = pipe_config->adjusted_mode.crtc_hdisplay;
+	mode->htotal = pipe_config->adjusted_mode.crtc_htotal;
+	mode->hsync_start = pipe_config->adjusted_mode.crtc_hsync_start;
+	mode->hsync_end = pipe_config->adjusted_mode.crtc_hsync_end;
 
-	crtc->mode.vdisplay = pipe_config->adjusted_mode.crtc_vdisplay;
-	crtc->mode.vtotal = pipe_config->adjusted_mode.crtc_vtotal;
-	crtc->mode.vsync_start = pipe_config->adjusted_mode.crtc_vsync_start;
-	crtc->mode.vsync_end = pipe_config->adjusted_mode.crtc_vsync_end;
+	mode->vdisplay = pipe_config->adjusted_mode.crtc_vdisplay;
+	mode->vtotal = pipe_config->adjusted_mode.crtc_vtotal;
+	mode->vsync_start = pipe_config->adjusted_mode.crtc_vsync_start;
+	mode->vsync_end = pipe_config->adjusted_mode.crtc_vsync_end;
 
-	crtc->mode.flags = pipe_config->adjusted_mode.flags;
+	mode->flags = pipe_config->adjusted_mode.flags;
 
-	crtc->mode.clock = pipe_config->adjusted_mode.crtc_clock;
-	crtc->mode.flags |= pipe_config->adjusted_mode.flags;
+	mode->clock = pipe_config->adjusted_mode.crtc_clock;
+	mode->flags |= pipe_config->adjusted_mode.flags;
 }
 
 static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc)
@@ -5094,10 +5918,13 @@ static void i9xx_set_pipeconf(struct intel_crtc *intel_crtc)
 		}
 	}
 
-	if (!IS_GEN2(dev) &&
-	    intel_crtc->config.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE)
-		pipeconf |= PIPECONF_INTERLACE_W_FIELD_INDICATION;
-	else
+	if (intel_crtc->config.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE) {
+		if (INTEL_INFO(dev)->gen < 4 ||
+		    intel_pipe_has_type(&intel_crtc->base, INTEL_OUTPUT_SDVO))
+			pipeconf |= PIPECONF_INTERLACE_W_FIELD_INDICATION;
+		else
+			pipeconf |= PIPECONF_INTERLACE_W_SYNC_SHIFT;
+	} else
 		pipeconf |= PIPECONF_PROGRESSIVE;
 
 	if (IS_VALLEYVIEW(dev) && intel_crtc->config.limited_color_range)
@@ -5114,16 +5941,12 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
 	int refclk, num_connectors = 0;
 	intel_clock_t clock, reduced_clock;
-	u32 dspcntr;
 	bool ok, has_reduced_clock = false;
 	bool is_lvds = false, is_dsi = false;
 	struct intel_encoder *encoder;
 	const intel_limit_t *limit;
-	int ret;
 
 	for_each_encoder_on_crtc(dev, crtc, encoder) {
 		switch (encoder->type) {
@@ -5139,7 +5962,7 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
 	}
 
 	if (is_dsi)
-		goto skip_dpll;
+		return 0;
 
 	if (!intel_crtc->config.clock_set) {
 		refclk = i9xx_get_refclk(crtc, num_connectors);
@@ -5184,43 +6007,17 @@ static int i9xx_crtc_mode_set(struct drm_crtc *crtc,
 		i8xx_update_pll(intel_crtc,
 				has_reduced_clock ? &reduced_clock : NULL,
 				num_connectors);
+	} else if (IS_CHERRYVIEW(dev)) {
+		chv_update_pll(intel_crtc);
 	} else if (IS_VALLEYVIEW(dev)) {
 		vlv_update_pll(intel_crtc);
 	} else {
 		i9xx_update_pll(intel_crtc,
 				has_reduced_clock ? &reduced_clock : NULL,
-                                num_connectors);
-	}
-
-skip_dpll:
-	/* Set up the display plane register */
-	dspcntr = DISPPLANE_GAMMA_ENABLE;
-
-	if (!IS_VALLEYVIEW(dev)) {
-		if (pipe == 0)
-			dspcntr &= ~DISPPLANE_SEL_PIPE_MASK;
-		else
-			dspcntr |= DISPPLANE_SEL_PIPE_B;
+				num_connectors);
 	}
 
-	intel_set_pipe_timings(intel_crtc);
-
-	/* pipesrc and dspsize control the size that is scaled from,
-	 * which should always be the user's requested size.
-	 */
-	I915_WRITE(DSPSIZE(plane),
-		   ((intel_crtc->config.pipe_src_h - 1) << 16) |
-		   (intel_crtc->config.pipe_src_w - 1));
-	I915_WRITE(DSPPOS(plane), 0);
-
-	i9xx_set_pipeconf(intel_crtc);
-
-	I915_WRITE(DSPCNTR(plane), dspcntr);
-	POSTING_READ(DSPCNTR(plane));
-
-	ret = intel_pipe_set_base(crtc, x, y, fb);
-
-	return ret;
+	return 0;
 }
 
 static void i9xx_get_pfit_config(struct intel_crtc *crtc,
@@ -5230,6 +6027,9 @@ static void i9xx_get_pfit_config(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t tmp;
 
+	if (INTEL_INFO(dev)->gen <= 3 && (IS_I830(dev) || !IS_MOBILE(dev)))
+		return;
+
 	tmp = I915_READ(PFIT_CONTROL);
 	if (!(tmp & PFIT_ENABLE))
 		return;
@@ -5261,7 +6061,7 @@ static void vlv_crtc_clock_get(struct intel_crtc *crtc,
 	int refclk = 100000;
 
 	mutex_lock(&dev_priv->dpio_lock);
-	mdiv = vlv_dpio_read(dev_priv, pipe, DPIO_DIV(pipe));
+	mdiv = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW3(pipe));
 	mutex_unlock(&dev_priv->dpio_lock);
 
 	clock.m1 = (mdiv >> DPIO_M1DIV_SHIFT) & 7;
@@ -5276,6 +6076,97 @@ static void vlv_crtc_clock_get(struct intel_crtc *crtc,
 	pipe_config->port_clock = clock.dot / 5;
 }
 
+static void i9xx_get_plane_config(struct intel_crtc *crtc,
+				  struct intel_plane_config *plane_config)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val, base, offset;
+	int pipe = crtc->pipe, plane = crtc->plane;
+	int fourcc, pixel_format;
+	int aligned_height;
+
+	crtc->base.primary->fb = kzalloc(sizeof(struct intel_framebuffer), GFP_KERNEL);
+	if (!crtc->base.primary->fb) {
+		DRM_DEBUG_KMS("failed to alloc fb\n");
+		return;
+	}
+
+	val = I915_READ(DSPCNTR(plane));
+
+	if (INTEL_INFO(dev)->gen >= 4)
+		if (val & DISPPLANE_TILED)
+			plane_config->tiled = true;
+
+	pixel_format = val & DISPPLANE_PIXFORMAT_MASK;
+	fourcc = intel_format_to_fourcc(pixel_format);
+	crtc->base.primary->fb->pixel_format = fourcc;
+	crtc->base.primary->fb->bits_per_pixel =
+		drm_format_plane_cpp(fourcc, 0) * 8;
+
+	if (INTEL_INFO(dev)->gen >= 4) {
+		if (plane_config->tiled)
+			offset = I915_READ(DSPTILEOFF(plane));
+		else
+			offset = I915_READ(DSPLINOFF(plane));
+		base = I915_READ(DSPSURF(plane)) & 0xfffff000;
+	} else {
+		base = I915_READ(DSPADDR(plane));
+	}
+	plane_config->base = base;
+
+	val = I915_READ(PIPESRC(pipe));
+	crtc->base.primary->fb->width = ((val >> 16) & 0xfff) + 1;
+	crtc->base.primary->fb->height = ((val >> 0) & 0xfff) + 1;
+
+	val = I915_READ(DSPSTRIDE(pipe));
+	crtc->base.primary->fb->pitches[0] = val & 0xffffff80;
+
+	aligned_height = intel_align_height(dev, crtc->base.primary->fb->height,
+					    plane_config->tiled);
+
+	plane_config->size = ALIGN(crtc->base.primary->fb->pitches[0] *
+				   aligned_height, PAGE_SIZE);
+
+	DRM_DEBUG_KMS("pipe/plane %d/%d with fb: size=%dx%d@%d, offset=%x, pitch %d, size 0x%x\n",
+		      pipe, plane, crtc->base.primary->fb->width,
+		      crtc->base.primary->fb->height,
+		      crtc->base.primary->fb->bits_per_pixel, base,
+		      crtc->base.primary->fb->pitches[0],
+		      plane_config->size);
+
+}
+
+static void chv_crtc_clock_get(struct intel_crtc *crtc,
+			       struct intel_crtc_config *pipe_config)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int pipe = pipe_config->cpu_transcoder;
+	enum dpio_channel port = vlv_pipe_to_channel(pipe);
+	intel_clock_t clock;
+	u32 cmn_dw13, pll_dw0, pll_dw1, pll_dw2;
+	int refclk = 100000;
+
+	mutex_lock(&dev_priv->dpio_lock);
+	cmn_dw13 = vlv_dpio_read(dev_priv, pipe, CHV_CMN_DW13(port));
+	pll_dw0 = vlv_dpio_read(dev_priv, pipe, CHV_PLL_DW0(port));
+	pll_dw1 = vlv_dpio_read(dev_priv, pipe, CHV_PLL_DW1(port));
+	pll_dw2 = vlv_dpio_read(dev_priv, pipe, CHV_PLL_DW2(port));
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	clock.m1 = (pll_dw1 & 0x7) == DPIO_CHV_M1_DIV_BY_2 ? 2 : 0;
+	clock.m2 = ((pll_dw0 & 0xff) << 22) | (pll_dw2 & 0x3fffff);
+	clock.n = (pll_dw1 >> DPIO_CHV_N_DIV_SHIFT) & 0xf;
+	clock.p1 = (cmn_dw13 >> DPIO_CHV_P1_DIV_SHIFT) & 0x7;
+	clock.p2 = (cmn_dw13 >> DPIO_CHV_P2_DIV_SHIFT) & 0x1f;
+
+	chv_clock(refclk, &clock);
+
+	/* clock.dot is the fast clock */
+	pipe_config->port_clock = clock.dot / 5;
+}
+
 static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 				 struct intel_crtc_config *pipe_config)
 {
@@ -5283,6 +6174,10 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t tmp;
 
+	if (!intel_display_power_enabled(dev_priv,
+					 POWER_DOMAIN_PIPE(crtc->pipe)))
+		return false;
+
 	pipe_config->cpu_transcoder = (enum transcoder) crtc->pipe;
 	pipe_config->shared_dpll = DPLL_ID_PRIVATE;
 
@@ -5306,6 +6201,9 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 		}
 	}
 
+	if (IS_VALLEYVIEW(dev) && (tmp & PIPECONF_COLOR_RANGE_SELECT))
+		pipe_config->limited_color_range = true;
+
 	if (INTEL_INFO(dev)->gen < 4)
 		pipe_config->double_wide = tmp & PIPECONF_DOUBLE_WIDE;
 
@@ -5341,7 +6239,9 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 						     DPLL_PORTB_READY_MASK);
 	}
 
-	if (IS_VALLEYVIEW(dev))
+	if (IS_CHERRYVIEW(dev))
+		chv_crtc_clock_get(crtc, pipe_config);
+	else if (IS_VALLEYVIEW(dev))
 		vlv_crtc_clock_get(crtc, pipe_config);
 	else
 		i9xx_crtc_clock_get(crtc, pipe_config);
@@ -5462,8 +6362,7 @@ static void ironlake_init_pch_refclk(struct drm_device *dev)
 			if (intel_panel_use_ssc(dev_priv) && can_ssc) {
 				DRM_DEBUG_KMS("Using SSC on eDP\n");
 				val |= DREF_CPU_SOURCE_OUTPUT_DOWNSPREAD;
-			}
-			else
+			} else
 				val |= DREF_CPU_SOURCE_OUTPUT_NONSPREAD;
 		} else
 			val |= DREF_CPU_SOURCE_OUTPUT_DISABLE;
@@ -5718,9 +6617,9 @@ static int ironlake_get_refclk(struct drm_crtc *crtc)
 	}
 
 	if (is_lvds && intel_panel_use_ssc(dev_priv) && num_connectors < 2) {
-		DRM_DEBUG_KMS("using SSC reference clock of %d MHz\n",
+		DRM_DEBUG_KMS("using SSC reference clock of %d kHz\n",
 			      dev_priv->vbt.lvds_ssc_freq);
-		return dev_priv->vbt.lvds_ssc_freq * 1000;
+		return dev_priv->vbt.lvds_ssc_freq;
 	}
 
 	return 120000;
@@ -5944,7 +6843,7 @@ int ironlake_get_lanes_required(int target_clock, int link_bw, int bpp)
 	 * is 2.5%; use 5% for safety's sake.
 	 */
 	u32 bps = target_clock * bpp * 21 / 20;
-	return bps / (link_bw * 8) + 1;
+	return DIV_ROUND_UP(bps, link_bw * 8);
 }
 
 static bool ironlake_needs_fb_cb_tune(struct dpll *dpll, int factor)
@@ -5982,7 +6881,7 @@ static uint32_t ironlake_compute_dpll(struct intel_crtc *intel_crtc,
 	factor = 21;
 	if (is_lvds) {
 		if ((intel_panel_use_ssc(dev_priv) &&
-		     dev_priv->vbt.lvds_ssc_freq == 100) ||
+		     dev_priv->vbt.lvds_ssc_freq == 100000) ||
 		    (HAS_PCH_IBX(dev) && intel_is_dual_link_lvds(dev)))
 			factor = 25;
 	} else if (intel_crtc->config.sdvo_tv_clock)
@@ -6042,10 +6941,7 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc,
 				  struct drm_framebuffer *fb)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
-	int plane = intel_crtc->plane;
 	int num_connectors = 0;
 	intel_clock_t clock, reduced_clock;
 	u32 dpll = 0, fp = 0, fp2 = 0;
@@ -6053,7 +6949,6 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc,
 	bool is_lvds = false;
 	struct intel_encoder *encoder;
 	struct intel_shared_dpll *pll;
-	int ret;
 
 	for_each_encoder_on_crtc(dev, crtc, encoder) {
 		switch (encoder->type) {
@@ -6103,36 +6998,18 @@ static int ironlake_crtc_mode_set(struct drm_crtc *crtc,
 		pll = intel_get_shared_dpll(intel_crtc);
 		if (pll == NULL) {
 			DRM_DEBUG_DRIVER("failed to find PLL for pipe %c\n",
-					 pipe_name(pipe));
+					 pipe_name(intel_crtc->pipe));
 			return -EINVAL;
 		}
 	} else
 		intel_put_shared_dpll(intel_crtc);
 
-	if (intel_crtc->config.has_dp_encoder)
-		intel_dp_set_m_n(intel_crtc);
-
-	if (is_lvds && has_reduced_clock && i915_powersave)
+	if (is_lvds && has_reduced_clock && i915.powersave)
 		intel_crtc->lowfreq_avail = true;
 	else
 		intel_crtc->lowfreq_avail = false;
 
-	intel_set_pipe_timings(intel_crtc);
-
-	if (intel_crtc->config.has_pch_encoder) {
-		intel_cpu_transcoder_set_m_n(intel_crtc,
-					     &intel_crtc->config.fdi_m_n);
-	}
-
-	ironlake_set_pipeconf(crtc);
-
-	/* Set up the display plane register */
-	I915_WRITE(DSPCNTR(plane), DISPPLANE_GAMMA_ENABLE);
-	POSTING_READ(DSPCNTR(plane));
-
-	ret = intel_pipe_set_base(crtc, x, y, fb);
-
-	return ret;
+	return 0;
 }
 
 static void intel_pch_transcoder_get_m_n(struct intel_crtc *crtc,
@@ -6219,6 +7096,66 @@ static void ironlake_get_pfit_config(struct intel_crtc *crtc,
 	}
 }
 
+static void ironlake_get_plane_config(struct intel_crtc *crtc,
+				      struct intel_plane_config *plane_config)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val, base, offset;
+	int pipe = crtc->pipe, plane = crtc->plane;
+	int fourcc, pixel_format;
+	int aligned_height;
+
+	crtc->base.primary->fb = kzalloc(sizeof(struct intel_framebuffer), GFP_KERNEL);
+	if (!crtc->base.primary->fb) {
+		DRM_DEBUG_KMS("failed to alloc fb\n");
+		return;
+	}
+
+	val = I915_READ(DSPCNTR(plane));
+
+	if (INTEL_INFO(dev)->gen >= 4)
+		if (val & DISPPLANE_TILED)
+			plane_config->tiled = true;
+
+	pixel_format = val & DISPPLANE_PIXFORMAT_MASK;
+	fourcc = intel_format_to_fourcc(pixel_format);
+	crtc->base.primary->fb->pixel_format = fourcc;
+	crtc->base.primary->fb->bits_per_pixel =
+		drm_format_plane_cpp(fourcc, 0) * 8;
+
+	base = I915_READ(DSPSURF(plane)) & 0xfffff000;
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+		offset = I915_READ(DSPOFFSET(plane));
+	} else {
+		if (plane_config->tiled)
+			offset = I915_READ(DSPTILEOFF(plane));
+		else
+			offset = I915_READ(DSPLINOFF(plane));
+	}
+	plane_config->base = base;
+
+	val = I915_READ(PIPESRC(pipe));
+	crtc->base.primary->fb->width = ((val >> 16) & 0xfff) + 1;
+	crtc->base.primary->fb->height = ((val >> 0) & 0xfff) + 1;
+
+	val = I915_READ(DSPSTRIDE(pipe));
+	crtc->base.primary->fb->pitches[0] = val & 0xffffff80;
+
+	aligned_height = intel_align_height(dev, crtc->base.primary->fb->height,
+					    plane_config->tiled);
+
+	plane_config->size = ALIGN(crtc->base.primary->fb->pitches[0] *
+				   aligned_height, PAGE_SIZE);
+
+	DRM_DEBUG_KMS("pipe/plane %d/%d with fb: size=%dx%d@%d, offset=%x, pitch %d, size 0x%x\n",
+		      pipe, plane, crtc->base.primary->fb->width,
+		      crtc->base.primary->fb->height,
+		      crtc->base.primary->fb->bits_per_pixel, base,
+		      crtc->base.primary->fb->pitches[0],
+		      plane_config->size);
+}
+
 static bool ironlake_get_pipe_config(struct intel_crtc *crtc,
 				     struct intel_crtc_config *pipe_config)
 {
@@ -6250,6 +7187,9 @@ static bool ironlake_get_pipe_config(struct intel_crtc *crtc,
 		break;
 	}
 
+	if (tmp & PIPECONF_COLOR_RANGE_SELECT)
+		pipe_config->limited_color_range = true;
+
 	if (I915_READ(PCH_TRANSCONF(crtc->pipe)) & TRANS_ENABLE) {
 		struct intel_shared_dpll *pll;
 
@@ -6299,10 +7239,8 @@ static void assert_can_disable_lcpll(struct drm_i915_private *dev_priv)
 	struct drm_device *dev = dev_priv->dev;
 	struct intel_ddi_plls *plls = &dev_priv->ddi_plls;
 	struct intel_crtc *crtc;
-	unsigned long irqflags;
-	uint32_t val;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, base.head)
+	for_each_intel_crtc(dev, crtc)
 		WARN(crtc->active, "CRTC for pipe %c enabled\n",
 		     pipe_name(crtc->pipe));
 
@@ -6321,14 +7259,29 @@ static void assert_can_disable_lcpll(struct drm_i915_private *dev_priv)
 	     "Utility pin enabled\n");
 	WARN(I915_READ(PCH_GTC_CTL) & PCH_GTC_ENABLE, "PCH GTC enabled\n");
 
-	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-	val = I915_READ(DEIMR);
-	WARN((val & ~DE_PCH_EVENT_IVB) != val,
-	     "Unexpected DEIMR bits enabled: 0x%x\n", val);
-	val = I915_READ(SDEIMR);
-	WARN((val | SDE_HOTPLUG_MASK_CPT) != 0xffffffff,
-	     "Unexpected SDEIMR bits enabled: 0x%x\n", val);
-	spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	/*
+	 * In theory we can still leave IRQs enabled, as long as only the HPD
+	 * interrupts remain enabled. We used to check for that, but since it's
+	 * gen-specific and since we only disable LCPLL after we fully disable
+	 * the interrupts, the check below should be enough.
+	 */
+	WARN(!dev_priv->pm.irqs_disabled, "IRQs enabled\n");
+}
+
+static void hsw_write_dcomp(struct drm_i915_private *dev_priv, uint32_t val)
+{
+	struct drm_device *dev = dev_priv->dev;
+
+	if (IS_HASWELL(dev)) {
+		mutex_lock(&dev_priv->rps.hw_lock);
+		if (sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_D_COMP,
+					    val))
+			DRM_ERROR("Failed to disable D_COMP\n");
+		mutex_unlock(&dev_priv->rps.hw_lock);
+	} else {
+		I915_WRITE(D_COMP, val);
+	}
+	POSTING_READ(D_COMP);
 }
 
 /*
@@ -6368,11 +7321,7 @@ static void hsw_disable_lcpll(struct drm_i915_private *dev_priv,
 
 	val = I915_READ(D_COMP);
 	val |= D_COMP_COMP_DISABLE;
-	mutex_lock(&dev_priv->rps.hw_lock);
-	if (sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_D_COMP, val))
-		DRM_ERROR("Failed to disable D_COMP\n");
-	mutex_unlock(&dev_priv->rps.hw_lock);
-	POSTING_READ(D_COMP);
+	hsw_write_dcomp(dev_priv, val);
 	ndelay(100);
 
 	if (wait_for((I915_READ(D_COMP) & D_COMP_RCOMP_IN_PROGRESS) == 0, 1))
@@ -6393,6 +7342,7 @@ static void hsw_disable_lcpll(struct drm_i915_private *dev_priv,
 static void hsw_restore_lcpll(struct drm_i915_private *dev_priv)
 {
 	uint32_t val;
+	unsigned long irqflags;
 
 	val = I915_READ(LCPLL_CTL);
 
@@ -6400,9 +7350,22 @@ static void hsw_restore_lcpll(struct drm_i915_private *dev_priv)
 		    LCPLL_POWER_DOWN_ALLOW)) == LCPLL_PLL_LOCK)
 		return;
 
-	/* Make sure we're not on PC8 state before disabling PC8, otherwise
-	 * we'll hang the machine! */
-	gen6_gt_force_wake_get(dev_priv);
+	/*
+	 * Make sure we're not on PC8 state before disabling PC8, otherwise
+	 * we'll hang the machine. To prevent PC8 state, just enable force_wake.
+	 *
+	 * The other problem is that hsw_restore_lcpll() is called as part of
+	 * the runtime PM resume sequence, so we can't just call
+	 * gen6_gt_force_wake_get() because that function calls
+	 * intel_runtime_pm_get(), and we can't change the runtime PM refcount
+	 * while we are on the resume sequence. So to solve this problem we have
+	 * to call special forcewake code that doesn't touch runtime PM and
+	 * doesn't enable the forcewake delayed work.
+	 */
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+	if (dev_priv->uncore.forcewake_count++ == 0)
+		dev_priv->uncore.funcs.force_wake_get(dev_priv, FORCEWAKE_ALL);
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 
 	if (val & LCPLL_POWER_DOWN_ALLOW) {
 		val &= ~LCPLL_POWER_DOWN_ALLOW;
@@ -6413,11 +7376,7 @@ static void hsw_restore_lcpll(struct drm_i915_private *dev_priv)
 	val = I915_READ(D_COMP);
 	val |= D_COMP_COMP_FORCE;
 	val &= ~D_COMP_COMP_DISABLE;
-	mutex_lock(&dev_priv->rps.hw_lock);
-	if (sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_D_COMP, val))
-		DRM_ERROR("Failed to enable D_COMP\n");
-	mutex_unlock(&dev_priv->rps.hw_lock);
-	POSTING_READ(D_COMP);
+	hsw_write_dcomp(dev_priv, val);
 
 	val = I915_READ(LCPLL_CTL);
 	val &= ~LCPLL_PLL_DISABLE;
@@ -6436,24 +7395,43 @@ static void hsw_restore_lcpll(struct drm_i915_private *dev_priv)
 			DRM_ERROR("Switching back to LCPLL failed\n");
 	}
 
-	gen6_gt_force_wake_put(dev_priv);
+	/* See the big comment above. */
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+	if (--dev_priv->uncore.forcewake_count == 0)
+		dev_priv->uncore.funcs.force_wake_put(dev_priv, FORCEWAKE_ALL);
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
-void hsw_enable_pc8_work(struct work_struct *__work)
+/*
+ * Package states C8 and deeper are really deep PC states that can only be
+ * reached when all the devices on the system allow it, so even if the graphics
+ * device allows PC8+, it doesn't mean the system will actually get to these
+ * states. Our driver only allows PC8+ when going into runtime PM.
+ *
+ * The requirements for PC8+ are that all the outputs are disabled, the power
+ * well is disabled and most interrupts are disabled, and these are also
+ * requirements for runtime PM. When these conditions are met, we manually do
+ * the other conditions: disable the interrupts, clocks and switch LCPLL refclk
+ * to Fclk. If we're in PC8+ and we get an non-hotplug interrupt, we can hard
+ * hang the machine.
+ *
+ * When we really reach PC8 or deeper states (not just when we allow it) we lose
+ * the state of some registers, so when we come back from PC8+ we need to
+ * restore this state. We don't get into PC8+ if we're not in RC6, so we don't
+ * need to take care of the registers kept by RC6. Notice that this happens even
+ * if we don't put the device in PCI D3 state (which is what currently happens
+ * because of the runtime PM support).
+ *
+ * For more, read "Display Sequences for Package C8" on the hardware
+ * documentation.
+ */
+void hsw_enable_pc8(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv =
-		container_of(to_delayed_work(__work), struct drm_i915_private,
-			     pc8.enable_work);
 	struct drm_device *dev = dev_priv->dev;
 	uint32_t val;
 
-	if (dev_priv->pc8.enabled)
-		return;
-
 	DRM_DEBUG_KMS("Enabling package C8+\n");
 
-	dev_priv->pc8.enabled = true;
-
 	if (dev_priv->pch_id == INTEL_PCH_LPT_LP_DEVICE_ID_TYPE) {
 		val = I915_READ(SOUTH_DSPCLK_GATE_D);
 		val &= ~PCH_LP_PARTITION_LEVEL_DISABLE;
@@ -6461,45 +7439,17 @@ void hsw_enable_pc8_work(struct work_struct *__work)
 	}
 
 	lpt_disable_clkout_dp(dev);
-	hsw_pc8_disable_interrupts(dev);
 	hsw_disable_lcpll(dev_priv, true, true);
 }
 
-static void __hsw_enable_package_c8(struct drm_i915_private *dev_priv)
-{
-	WARN_ON(!mutex_is_locked(&dev_priv->pc8.lock));
-	WARN(dev_priv->pc8.disable_count < 1,
-	     "pc8.disable_count: %d\n", dev_priv->pc8.disable_count);
-
-	dev_priv->pc8.disable_count--;
-	if (dev_priv->pc8.disable_count != 0)
-		return;
-
-	schedule_delayed_work(&dev_priv->pc8.enable_work,
-			      msecs_to_jiffies(i915_pc8_timeout));
-}
-
-static void __hsw_disable_package_c8(struct drm_i915_private *dev_priv)
+void hsw_disable_pc8(struct drm_i915_private *dev_priv)
 {
 	struct drm_device *dev = dev_priv->dev;
 	uint32_t val;
 
-	WARN_ON(!mutex_is_locked(&dev_priv->pc8.lock));
-	WARN(dev_priv->pc8.disable_count < 0,
-	     "pc8.disable_count: %d\n", dev_priv->pc8.disable_count);
-
-	dev_priv->pc8.disable_count++;
-	if (dev_priv->pc8.disable_count != 1)
-		return;
-
-	cancel_delayed_work_sync(&dev_priv->pc8.enable_work);
-	if (!dev_priv->pc8.enabled)
-		return;
-
 	DRM_DEBUG_KMS("Disabling package C8+\n");
 
 	hsw_restore_lcpll(dev_priv);
-	hsw_pc8_restore_interrupts(dev);
 	lpt_init_pch_refclk(dev);
 
 	if (dev_priv->pch_id == INTEL_PCH_LPT_LP_DEVICE_ID_TYPE) {
@@ -6509,227 +7459,31 @@ static void __hsw_disable_package_c8(struct drm_i915_private *dev_priv)
 	}
 
 	intel_prepare_ddi(dev);
-	i915_gem_init_swizzling(dev);
-	mutex_lock(&dev_priv->rps.hw_lock);
-	gen6_update_ring_freq(dev);
-	mutex_unlock(&dev_priv->rps.hw_lock);
-	dev_priv->pc8.enabled = false;
-}
-
-void hsw_enable_package_c8(struct drm_i915_private *dev_priv)
-{
-	if (!HAS_PC8(dev_priv->dev))
-		return;
-
-	mutex_lock(&dev_priv->pc8.lock);
-	__hsw_enable_package_c8(dev_priv);
-	mutex_unlock(&dev_priv->pc8.lock);
-}
-
-void hsw_disable_package_c8(struct drm_i915_private *dev_priv)
-{
-	if (!HAS_PC8(dev_priv->dev))
-		return;
-
-	mutex_lock(&dev_priv->pc8.lock);
-	__hsw_disable_package_c8(dev_priv);
-	mutex_unlock(&dev_priv->pc8.lock);
-}
-
-static bool hsw_can_enable_package_c8(struct drm_i915_private *dev_priv)
-{
-	struct drm_device *dev = dev_priv->dev;
-	struct intel_crtc *crtc;
-	uint32_t val;
-
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, base.head)
-		if (crtc->base.enabled)
-			return false;
-
-	/* This case is still possible since we have the i915.disable_power_well
-	 * parameter and also the KVMr or something else might be requesting the
-	 * power well. */
-	val = I915_READ(HSW_PWR_WELL_DRIVER);
-	if (val != 0) {
-		DRM_DEBUG_KMS("Not enabling PC8: power well on\n");
-		return false;
-	}
-
-	return true;
-}
-
-/* Since we're called from modeset_global_resources there's no way to
- * symmetrically increase and decrease the refcount, so we use
- * dev_priv->pc8.requirements_met to track whether we already have the refcount
- * or not.
- */
-static void hsw_update_package_c8(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	bool allow;
-
-	if (!HAS_PC8(dev_priv->dev))
-		return;
-
-	if (!i915_enable_pc8)
-		return;
-
-	mutex_lock(&dev_priv->pc8.lock);
-
-	allow = hsw_can_enable_package_c8(dev_priv);
-
-	if (allow == dev_priv->pc8.requirements_met)
-		goto done;
-
-	dev_priv->pc8.requirements_met = allow;
-
-	if (allow)
-		__hsw_enable_package_c8(dev_priv);
-	else
-		__hsw_disable_package_c8(dev_priv);
-
-done:
-	mutex_unlock(&dev_priv->pc8.lock);
-}
-
-static void hsw_package_c8_gpu_idle(struct drm_i915_private *dev_priv)
-{
-	if (!HAS_PC8(dev_priv->dev))
-		return;
-
-	mutex_lock(&dev_priv->pc8.lock);
-	if (!dev_priv->pc8.gpu_idle) {
-		dev_priv->pc8.gpu_idle = true;
-		__hsw_enable_package_c8(dev_priv);
-	}
-	mutex_unlock(&dev_priv->pc8.lock);
-}
-
-static void hsw_package_c8_gpu_busy(struct drm_i915_private *dev_priv)
-{
-	if (!HAS_PC8(dev_priv->dev))
-		return;
-
-	mutex_lock(&dev_priv->pc8.lock);
-	if (dev_priv->pc8.gpu_idle) {
-		dev_priv->pc8.gpu_idle = false;
-		__hsw_disable_package_c8(dev_priv);
-	}
-	mutex_unlock(&dev_priv->pc8.lock);
-}
-
-#define for_each_power_domain(domain, mask)				\
-	for ((domain) = 0; (domain) < POWER_DOMAIN_NUM; (domain)++)	\
-		if ((1 << (domain)) & (mask))
-
-static unsigned long get_pipe_power_domains(struct drm_device *dev,
-					    enum pipe pipe, bool pfit_enabled)
-{
-	unsigned long mask;
-	enum transcoder transcoder;
-
-	transcoder = intel_pipe_to_cpu_transcoder(dev->dev_private, pipe);
-
-	mask = BIT(POWER_DOMAIN_PIPE(pipe));
-	mask |= BIT(POWER_DOMAIN_TRANSCODER(transcoder));
-	if (pfit_enabled)
-		mask |= BIT(POWER_DOMAIN_PIPE_PANEL_FITTER(pipe));
-
-	return mask;
 }
 
-void intel_display_set_init_power(struct drm_device *dev, bool enable)
+static void snb_modeset_global_resources(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	if (dev_priv->power_domains.init_power_on == enable)
-		return;
-
-	if (enable)
-		intel_display_power_get(dev, POWER_DOMAIN_INIT);
-	else
-		intel_display_power_put(dev, POWER_DOMAIN_INIT);
-
-	dev_priv->power_domains.init_power_on = enable;
-}
-
-static void modeset_update_power_wells(struct drm_device *dev)
-{
-	unsigned long pipe_domains[I915_MAX_PIPES] = { 0, };
-	struct intel_crtc *crtc;
-
-	/*
-	 * First get all needed power domains, then put all unneeded, to avoid
-	 * any unnecessary toggling of the power wells.
-	 */
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, base.head) {
-		enum intel_display_power_domain domain;
-
-		if (!crtc->base.enabled)
-			continue;
-
-		pipe_domains[crtc->pipe] = get_pipe_power_domains(dev,
-						crtc->pipe,
-						crtc->config.pch_pfit.enabled);
-
-		for_each_power_domain(domain, pipe_domains[crtc->pipe])
-			intel_display_power_get(dev, domain);
-	}
-
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, base.head) {
-		enum intel_display_power_domain domain;
-
-		for_each_power_domain(domain, crtc->enabled_power_domains)
-			intel_display_power_put(dev, domain);
-
-		crtc->enabled_power_domains = pipe_domains[crtc->pipe];
-	}
-
-	intel_display_set_init_power(dev, false);
+	modeset_update_crtc_power_domains(dev);
 }
 
 static void haswell_modeset_global_resources(struct drm_device *dev)
 {
-	modeset_update_power_wells(dev);
-	hsw_update_package_c8(dev);
+	modeset_update_crtc_power_domains(dev);
 }
 
 static int haswell_crtc_mode_set(struct drm_crtc *crtc,
 				 int x, int y,
 				 struct drm_framebuffer *fb)
 {
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int plane = intel_crtc->plane;
-	int ret;
 
-	if (!intel_ddi_pll_mode_set(crtc))
+	if (!intel_ddi_pll_select(intel_crtc))
 		return -EINVAL;
-
-	if (intel_crtc->config.has_dp_encoder)
-		intel_dp_set_m_n(intel_crtc);
+	intel_ddi_pll_enable(intel_crtc);
 
 	intel_crtc->lowfreq_avail = false;
 
-	intel_set_pipe_timings(intel_crtc);
-
-	if (intel_crtc->config.has_pch_encoder) {
-		intel_cpu_transcoder_set_m_n(intel_crtc,
-					     &intel_crtc->config.fdi_m_n);
-	}
-
-	haswell_set_pipeconf(crtc);
-
-	intel_set_pipe_csc(crtc);
-
-	/* Set up the display plane register */
-	I915_WRITE(DSPCNTR(plane), DISPPLANE_GAMMA_ENABLE | DISPPLANE_PIPE_CSC_ENABLE);
-	POSTING_READ(DSPCNTR(plane));
-
-	ret = intel_pipe_set_base(crtc, x, y, fb);
-
-	return ret;
+	return 0;
 }
 
 static bool haswell_get_pipe_config(struct intel_crtc *crtc,
@@ -6740,6 +7494,10 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 	enum intel_display_power_domain pfit_domain;
 	uint32_t tmp;
 
+	if (!intel_display_power_enabled(dev_priv,
+					 POWER_DOMAIN_PIPE(crtc->pipe)))
+		return false;
+
 	pipe_config->cpu_transcoder = (enum transcoder) crtc->pipe;
 	pipe_config->shared_dpll = DPLL_ID_PRIVATE;
 
@@ -6765,7 +7523,7 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 			pipe_config->cpu_transcoder = TRANSCODER_EDP;
 	}
 
-	if (!intel_display_power_enabled(dev,
+	if (!intel_display_power_enabled(dev_priv,
 			POWER_DOMAIN_TRANSCODER(pipe_config->cpu_transcoder)))
 		return false;
 
@@ -6793,49 +7551,18 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 	intel_get_pipe_timings(crtc, pipe_config);
 
 	pfit_domain = POWER_DOMAIN_PIPE_PANEL_FITTER(crtc->pipe);
-	if (intel_display_power_enabled(dev, pfit_domain))
+	if (intel_display_power_enabled(dev_priv, pfit_domain))
 		ironlake_get_pfit_config(crtc, pipe_config);
 
-	pipe_config->ips_enabled = hsw_crtc_supports_ips(crtc) &&
-				   (I915_READ(IPS_CTL) & IPS_ENABLE);
+	if (IS_HASWELL(dev))
+		pipe_config->ips_enabled = hsw_crtc_supports_ips(crtc) &&
+			(I915_READ(IPS_CTL) & IPS_ENABLE);
 
 	pipe_config->pixel_multiplier = 1;
 
 	return true;
 }
 
-static int intel_crtc_mode_set(struct drm_crtc *crtc,
-			       int x, int y,
-			       struct drm_framebuffer *fb)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_encoder *encoder;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_display_mode *mode = &intel_crtc->config.requested_mode;
-	int pipe = intel_crtc->pipe;
-	int ret;
-
-	drm_vblank_pre_modeset(dev, pipe);
-
-	ret = dev_priv->display.crtc_mode_set(crtc, x, y, fb);
-
-	drm_vblank_post_modeset(dev, pipe);
-
-	if (ret != 0)
-		return ret;
-
-	for_each_encoder_on_crtc(dev, crtc, encoder) {
-		DRM_DEBUG_KMS("[ENCODER:%d:%s] set [MODE:%d:%s]\n",
-			encoder->base.base.id,
-			drm_get_encoder_name(&encoder->base),
-			mode->base.id, mode->name);
-		encoder->mode_set(encoder);
-	}
-
-	return 0;
-}
-
 static struct {
 	int clock;
 	u32 config;
@@ -6950,8 +7677,6 @@ static void haswell_write_eld(struct drm_connector *connector,
 {
 	struct drm_i915_private *dev_priv = connector->dev->dev_private;
 	uint8_t *eld = connector->eld;
-	struct drm_device *dev = crtc->dev;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t eldv;
 	uint32_t i;
 	int len;
@@ -6963,17 +7688,14 @@ static void haswell_write_eld(struct drm_connector *connector,
 	int aud_config = HSW_AUD_CFG(pipe);
 	int aud_cntrl_st2 = HSW_AUD_PIN_ELD_CP_VLD;
 
-
-	DRM_DEBUG_DRIVER("HDMI: Haswell Audio initialize....\n");
-
 	/* Audio output enable */
 	DRM_DEBUG_DRIVER("HDMI audio: enable codec\n");
 	tmp = I915_READ(aud_cntrl_st2);
 	tmp |= (AUDIO_OUTPUT_ENABLE_A << (pipe * 4));
 	I915_WRITE(aud_cntrl_st2, tmp);
+	POSTING_READ(aud_cntrl_st2);
 
-	/* Wait for 1 vertical blank */
-	intel_wait_for_vblank(dev, pipe);
+	assert_pipe_disabled(dev_priv, to_intel_crtc(crtc)->pipe);
 
 	/* Set ELD valid state */
 	tmp = I915_READ(aud_cntrl_st2);
@@ -6993,7 +7715,6 @@ static void haswell_write_eld(struct drm_connector *connector,
 	DRM_DEBUG_DRIVER("ELD on pipe %c\n", pipe_name(pipe));
 
 	eldv = AUDIO_ELD_VALID_A << (pipe * 4);
-	intel_crtc->eld_vld = true;
 
 	if (intel_pipe_has_type(crtc, INTEL_OUTPUT_DISPLAYPORT)) {
 		DRM_DEBUG_DRIVER("ELD: DisplayPort detected\n");
@@ -7140,9 +7861,9 @@ void intel_write_eld(struct drm_encoder *encoder,
 
 	DRM_DEBUG_DRIVER("ELD on [CONNECTOR:%d:%s], [ENCODER:%d:%s]\n",
 			 connector->base.id,
-			 drm_get_connector_name(connector),
+			 connector->name,
 			 connector->encoder->base.id,
-			 drm_get_encoder_name(connector->encoder));
+			 connector->encoder->name);
 
 	connector->eld[6] = drm_av_sync_delay(connector, mode) / 2;
 
@@ -7155,29 +7876,33 @@ static void i845_update_cursor(struct drm_crtc *crtc, u32 base)
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	bool visible = base != 0;
-	u32 cntl;
+	uint32_t cntl;
 
-	if (intel_crtc->cursor_visible == visible)
-		return;
-
-	cntl = I915_READ(_CURACNTR);
-	if (visible) {
+	if (base != intel_crtc->cursor_base) {
 		/* On these chipsets we can only modify the base whilst
 		 * the cursor is disabled.
 		 */
+		if (intel_crtc->cursor_cntl) {
+			I915_WRITE(_CURACNTR, 0);
+			POSTING_READ(_CURACNTR);
+			intel_crtc->cursor_cntl = 0;
+		}
+
 		I915_WRITE(_CURABASE, base);
+		POSTING_READ(_CURABASE);
+	}
 
-		cntl &= ~(CURSOR_FORMAT_MASK);
-		/* XXX width must be 64, stride 256 => 0x00 << 28 */
-		cntl |= CURSOR_ENABLE |
+	/* XXX width must be 64, stride 256 => 0x00 << 28 */
+	cntl = 0;
+	if (base)
+		cntl = (CURSOR_ENABLE |
 			CURSOR_GAMMA_ENABLE |
-			CURSOR_FORMAT_ARGB;
-	} else
-		cntl &= ~(CURSOR_ENABLE | CURSOR_GAMMA_ENABLE);
-	I915_WRITE(_CURACNTR, cntl);
-
-	intel_crtc->cursor_visible = visible;
+			CURSOR_FORMAT_ARGB);
+	if (intel_crtc->cursor_cntl != cntl) {
+		I915_WRITE(_CURACNTR, cntl);
+		POSTING_READ(_CURACNTR);
+		intel_crtc->cursor_cntl = cntl;
+	}
 }
 
 static void i9xx_update_cursor(struct drm_crtc *crtc, u32 base)
@@ -7186,24 +7911,34 @@ static void i9xx_update_cursor(struct drm_crtc *crtc, u32 base)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_crtc->pipe;
-	bool visible = base != 0;
-
-	if (intel_crtc->cursor_visible != visible) {
-		uint32_t cntl = I915_READ(CURCNTR(pipe));
-		if (base) {
-			cntl &= ~(CURSOR_MODE | MCURSOR_PIPE_SELECT);
-			cntl |= CURSOR_MODE_64_ARGB_AX | MCURSOR_GAMMA_ENABLE;
-			cntl |= pipe << 28; /* Connect to correct pipe */
-		} else {
-			cntl &= ~(CURSOR_MODE | MCURSOR_GAMMA_ENABLE);
-			cntl |= CURSOR_MODE_DISABLE;
+	uint32_t cntl;
+
+	cntl = 0;
+	if (base) {
+		cntl = MCURSOR_GAMMA_ENABLE;
+		switch (intel_crtc->cursor_width) {
+			case 64:
+				cntl |= CURSOR_MODE_64_ARGB_AX;
+				break;
+			case 128:
+				cntl |= CURSOR_MODE_128_ARGB_AX;
+				break;
+			case 256:
+				cntl |= CURSOR_MODE_256_ARGB_AX;
+				break;
+			default:
+				WARN_ON(1);
+				return;
 		}
+		cntl |= pipe << 28; /* Connect to correct pipe */
+	}
+	if (intel_crtc->cursor_cntl != cntl) {
 		I915_WRITE(CURCNTR(pipe), cntl);
-
-		intel_crtc->cursor_visible = visible;
+		POSTING_READ(CURCNTR(pipe));
+		intel_crtc->cursor_cntl = cntl;
 	}
+
 	/* and commit changes on next vblank */
-	POSTING_READ(CURCNTR(pipe));
 	I915_WRITE(CURBASE(pipe), base);
 	POSTING_READ(CURBASE(pipe));
 }
@@ -7214,29 +7949,38 @@ static void ivb_update_cursor(struct drm_crtc *crtc, u32 base)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_crtc->pipe;
-	bool visible = base != 0;
-
-	if (intel_crtc->cursor_visible != visible) {
-		uint32_t cntl = I915_READ(CURCNTR_IVB(pipe));
-		if (base) {
-			cntl &= ~CURSOR_MODE;
-			cntl |= CURSOR_MODE_64_ARGB_AX | MCURSOR_GAMMA_ENABLE;
-		} else {
-			cntl &= ~(CURSOR_MODE | MCURSOR_GAMMA_ENABLE);
-			cntl |= CURSOR_MODE_DISABLE;
-		}
-		if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
-			cntl |= CURSOR_PIPE_CSC_ENABLE;
-			cntl &= ~CURSOR_TRICKLE_FEED_DISABLE;
+	uint32_t cntl;
+
+	cntl = 0;
+	if (base) {
+		cntl = MCURSOR_GAMMA_ENABLE;
+		switch (intel_crtc->cursor_width) {
+			case 64:
+				cntl |= CURSOR_MODE_64_ARGB_AX;
+				break;
+			case 128:
+				cntl |= CURSOR_MODE_128_ARGB_AX;
+				break;
+			case 256:
+				cntl |= CURSOR_MODE_256_ARGB_AX;
+				break;
+			default:
+				WARN_ON(1);
+				return;
 		}
-		I915_WRITE(CURCNTR_IVB(pipe), cntl);
+	}
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		cntl |= CURSOR_PIPE_CSC_ENABLE;
 
-		intel_crtc->cursor_visible = visible;
+	if (intel_crtc->cursor_cntl != cntl) {
+		I915_WRITE(CURCNTR(pipe), cntl);
+		POSTING_READ(CURCNTR(pipe));
+		intel_crtc->cursor_cntl = cntl;
 	}
+
 	/* and commit changes on next vblank */
-	POSTING_READ(CURCNTR_IVB(pipe));
-	I915_WRITE(CURBASE_IVB(pipe), base);
-	POSTING_READ(CURBASE_IVB(pipe));
+	I915_WRITE(CURBASE(pipe), base);
+	POSTING_READ(CURBASE(pipe));
 }
 
 /* If no-part of the cursor is visible on the framebuffer, then the GPU may hang... */
@@ -7250,7 +7994,6 @@ static void intel_crtc_update_cursor(struct drm_crtc *crtc,
 	int x = intel_crtc->cursor_x;
 	int y = intel_crtc->cursor_y;
 	u32 base = 0, pos = 0;
-	bool visible;
 
 	if (on)
 		base = intel_crtc->cursor_addr;
@@ -7279,20 +8022,18 @@ static void intel_crtc_update_cursor(struct drm_crtc *crtc,
 	}
 	pos |= y << CURSOR_Y_SHIFT;
 
-	visible = base != 0;
-	if (!visible && !intel_crtc->cursor_visible)
+	if (base == 0 && intel_crtc->cursor_base == 0)
 		return;
 
-	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_BROADWELL(dev)) {
-		I915_WRITE(CURPOS_IVB(pipe), pos);
+	I915_WRITE(CURPOS(pipe), pos);
+
+	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_BROADWELL(dev))
 		ivb_update_cursor(crtc, base);
-	} else {
-		I915_WRITE(CURPOS(pipe), pos);
-		if (IS_845G(dev) || IS_I865G(dev))
-			i845_update_cursor(crtc, base);
-		else
-			i9xx_update_cursor(crtc, base);
-	}
+	else if (IS_845G(dev) || IS_I865G(dev))
+		i845_update_cursor(crtc, base);
+	else
+		i9xx_update_cursor(crtc, base);
+	intel_crtc->cursor_base = base;
 }
 
 static int intel_crtc_cursor_set(struct drm_crtc *crtc,
@@ -7304,6 +8045,7 @@ static int intel_crtc_cursor_set(struct drm_crtc *crtc,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct drm_i915_gem_object *obj;
+	unsigned old_width;
 	uint32_t addr;
 	int ret;
 
@@ -7316,9 +8058,11 @@ static int intel_crtc_cursor_set(struct drm_crtc *crtc,
 		goto finish;
 	}
 
-	/* Currently we only support 64x64 cursors */
-	if (width != 64 || height != 64) {
-		DRM_ERROR("we currently only support 64x64 cursors\n");
+	/* Check for which cursor types we support */
+	if (!((width == 64 && height == 64) ||
+			(width == 128 && height == 128 && !IS_GEN2(dev)) ||
+			(width == 256 && height == 256 && !IS_GEN2(dev)))) {
+		DRM_DEBUG("Cursor dimension not supported\n");
 		return -EINVAL;
 	}
 
@@ -7327,18 +8071,18 @@ static int intel_crtc_cursor_set(struct drm_crtc *crtc,
 		return -ENOENT;
 
 	if (obj->base.size < width * height * 4) {
-		DRM_ERROR("buffer is to small\n");
+		DRM_DEBUG_KMS("buffer is to small\n");
 		ret = -ENOMEM;
 		goto fail;
 	}
 
 	/* we only need to pin inside GTT if cursor is non-phy */
 	mutex_lock(&dev->struct_mutex);
-	if (!dev_priv->info->cursor_needs_physical) {
+	if (!INTEL_INFO(dev)->cursor_needs_physical) {
 		unsigned alignment;
 
 		if (obj->tiling_mode) {
-			DRM_ERROR("cursor cannot be tiled\n");
+			DRM_DEBUG_KMS("cursor cannot be tiled\n");
 			ret = -EINVAL;
 			goto fail_locked;
 		}
@@ -7354,27 +8098,25 @@ static int intel_crtc_cursor_set(struct drm_crtc *crtc,
 
 		ret = i915_gem_object_pin_to_display_plane(obj, alignment, NULL);
 		if (ret) {
-			DRM_ERROR("failed to move cursor bo into the GTT\n");
+			DRM_DEBUG_KMS("failed to move cursor bo into the GTT\n");
 			goto fail_locked;
 		}
 
 		ret = i915_gem_object_put_fence(obj);
 		if (ret) {
-			DRM_ERROR("failed to release fence for cursor");
+			DRM_DEBUG_KMS("failed to release fence for cursor");
 			goto fail_unpin;
 		}
 
 		addr = i915_gem_obj_ggtt_offset(obj);
 	} else {
 		int align = IS_I830(dev) ? 16 * 1024 : 256;
-		ret = i915_gem_attach_phys_object(dev, obj,
-						  (intel_crtc->pipe == 0) ? I915_GEM_PHYS_CURSOR_0 : I915_GEM_PHYS_CURSOR_1,
-						  align);
+		ret = i915_gem_object_attach_phys(obj, align);
 		if (ret) {
-			DRM_ERROR("failed to attach phys object\n");
+			DRM_DEBUG_KMS("failed to attach phys object\n");
 			goto fail_locked;
 		}
-		addr = obj->phys_obj->handle->busaddr;
+		addr = obj->phys_handle->busaddr;
 	}
 
 	if (IS_GEN2(dev))
@@ -7382,23 +8124,25 @@ static int intel_crtc_cursor_set(struct drm_crtc *crtc,
 
  finish:
 	if (intel_crtc->cursor_bo) {
-		if (dev_priv->info->cursor_needs_physical) {
-			if (intel_crtc->cursor_bo != obj)
-				i915_gem_detach_phys_object(dev, intel_crtc->cursor_bo);
-		} else
+		if (!INTEL_INFO(dev)->cursor_needs_physical)
 			i915_gem_object_unpin_from_display_plane(intel_crtc->cursor_bo);
 		drm_gem_object_unreference(&intel_crtc->cursor_bo->base);
 	}
 
 	mutex_unlock(&dev->struct_mutex);
 
+	old_width = intel_crtc->cursor_width;
+
 	intel_crtc->cursor_addr = addr;
 	intel_crtc->cursor_bo = obj;
 	intel_crtc->cursor_width = width;
 	intel_crtc->cursor_height = height;
 
-	if (intel_crtc->active)
+	if (intel_crtc->active) {
+		if (old_width != width)
+			intel_update_watermarks(crtc);
 		intel_crtc_update_cursor(crtc, intel_crtc->cursor_bo != NULL);
+	}
 
 	return 0;
 fail_unpin:
@@ -7444,10 +8188,10 @@ static struct drm_display_mode load_detect_mode = {
 		 704, 832, 0, 480, 489, 491, 520, 0, DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 };
 
-static struct drm_framebuffer *
-intel_framebuffer_create(struct drm_device *dev,
-			 struct drm_mode_fb_cmd2 *mode_cmd,
-			 struct drm_i915_gem_object *obj)
+struct drm_framebuffer *
+__intel_framebuffer_create(struct drm_device *dev,
+			   struct drm_mode_fb_cmd2 *mode_cmd,
+			   struct drm_i915_gem_object *obj)
 {
 	struct intel_framebuffer *intel_fb;
 	int ret;
@@ -7458,12 +8202,7 @@ intel_framebuffer_create(struct drm_device *dev,
 		return ERR_PTR(-ENOMEM);
 	}
 
-	ret = i915_mutex_lock_interruptible(dev);
-	if (ret)
-		goto err;
-
 	ret = intel_framebuffer_init(dev, intel_fb, mode_cmd, obj);
-	mutex_unlock(&dev->struct_mutex);
 	if (ret)
 		goto err;
 
@@ -7475,6 +8214,23 @@ err:
 	return ERR_PTR(ret);
 }
 
+static struct drm_framebuffer *
+intel_framebuffer_create(struct drm_device *dev,
+			 struct drm_mode_fb_cmd2 *mode_cmd,
+			 struct drm_i915_gem_object *obj)
+{
+	struct drm_framebuffer *fb;
+	int ret;
+
+	ret = i915_mutex_lock_interruptible(dev);
+	if (ret)
+		return ERR_PTR(ret);
+	fb = __intel_framebuffer_create(dev, mode_cmd, obj);
+	mutex_unlock(&dev->struct_mutex);
+
+	return fb;
+}
+
 static u32
 intel_framebuffer_pitch_for_width(int width, int bpp)
 {
@@ -7520,14 +8276,16 @@ mode_fits_in_fbdev(struct drm_device *dev,
 	struct drm_i915_gem_object *obj;
 	struct drm_framebuffer *fb;
 
-	if (dev_priv->fbdev == NULL)
+	if (!dev_priv->fbdev)
 		return NULL;
 
-	obj = dev_priv->fbdev->ifb.obj;
-	if (obj == NULL)
+	if (!dev_priv->fbdev->fb)
 		return NULL;
 
-	fb = &dev_priv->fbdev->ifb.base;
+	obj = dev_priv->fbdev->fb->obj;
+	BUG_ON(!obj);
+
+	fb = &dev_priv->fbdev->fb->base;
 	if (fb->pitches[0] < intel_framebuffer_pitch_for_width(mode->hdisplay,
 							       fb->bits_per_pixel))
 		return NULL;
@@ -7543,7 +8301,8 @@ mode_fits_in_fbdev(struct drm_device *dev,
 
 bool intel_get_load_detect_pipe(struct drm_connector *connector,
 				struct drm_display_mode *mode,
-				struct intel_load_detect_pipe *old)
+				struct intel_load_detect_pipe *old,
+				struct drm_modeset_acquire_ctx *ctx)
 {
 	struct intel_crtc *intel_crtc;
 	struct intel_encoder *intel_encoder =
@@ -7553,11 +8312,19 @@ bool intel_get_load_detect_pipe(struct drm_connector *connector,
 	struct drm_crtc *crtc = NULL;
 	struct drm_device *dev = encoder->dev;
 	struct drm_framebuffer *fb;
-	int i = -1;
+	struct drm_mode_config *config = &dev->mode_config;
+	int ret, i = -1;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s], [ENCODER:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector),
-		      encoder->base.id, drm_get_encoder_name(encoder));
+		      connector->base.id, connector->name,
+		      encoder->base.id, encoder->name);
+
+	drm_modeset_acquire_init(ctx, 0);
+
+retry:
+	ret = drm_modeset_lock(&config->connection_mutex, ctx);
+	if (ret)
+		goto fail_unlock;
 
 	/*
 	 * Algorithm gets a little messy:
@@ -7573,7 +8340,9 @@ bool intel_get_load_detect_pipe(struct drm_connector *connector,
 	if (encoder->crtc) {
 		crtc = encoder->crtc;
 
-		mutex_lock(&crtc->mutex);
+		ret = drm_modeset_lock(&crtc->mutex, ctx);
+		if (ret)
+			goto fail_unlock;
 
 		old->dpms_mode = connector->dpms;
 		old->load_detect_temp = false;
@@ -7586,7 +8355,7 @@ bool intel_get_load_detect_pipe(struct drm_connector *connector,
 	}
 
 	/* Find an unused one (if possible) */
-	list_for_each_entry(possible_crtc, &dev->mode_config.crtc_list, head) {
+	for_each_crtc(dev, possible_crtc) {
 		i++;
 		if (!(encoder->possible_crtcs & (1 << i)))
 			continue;
@@ -7601,14 +8370,18 @@ bool intel_get_load_detect_pipe(struct drm_connector *connector,
 	 */
 	if (!crtc) {
 		DRM_DEBUG_KMS("no pipe available for load-detect\n");
-		return false;
+		goto fail_unlock;
 	}
 
-	mutex_lock(&crtc->mutex);
+	ret = drm_modeset_lock(&crtc->mutex, ctx);
+	if (ret)
+		goto fail_unlock;
 	intel_encoder->new_crtc = to_intel_crtc(crtc);
 	to_intel_connector(connector)->new_encoder = intel_encoder;
 
 	intel_crtc = to_intel_crtc(crtc);
+	intel_crtc->new_enabled = true;
+	intel_crtc->new_config = &intel_crtc->config;
 	old->dpms_mode = connector->dpms;
 	old->load_detect_temp = true;
 	old->release_fb = NULL;
@@ -7632,38 +8405,57 @@ bool intel_get_load_detect_pipe(struct drm_connector *connector,
 		DRM_DEBUG_KMS("reusing fbdev for load-detection framebuffer\n");
 	if (IS_ERR(fb)) {
 		DRM_DEBUG_KMS("failed to allocate framebuffer for load-detection\n");
-		mutex_unlock(&crtc->mutex);
-		return false;
+		goto fail;
 	}
 
 	if (intel_set_mode(crtc, mode, 0, 0, fb)) {
 		DRM_DEBUG_KMS("failed to set mode on load-detect pipe\n");
 		if (old->release_fb)
 			old->release_fb->funcs->destroy(old->release_fb);
-		mutex_unlock(&crtc->mutex);
-		return false;
+		goto fail;
 	}
 
 	/* let the connector get through one full cycle before testing */
 	intel_wait_for_vblank(dev, intel_crtc->pipe);
 	return true;
+
+ fail:
+	intel_crtc->new_enabled = crtc->enabled;
+	if (intel_crtc->new_enabled)
+		intel_crtc->new_config = &intel_crtc->config;
+	else
+		intel_crtc->new_config = NULL;
+fail_unlock:
+	if (ret == -EDEADLK) {
+		drm_modeset_backoff(ctx);
+		goto retry;
+	}
+
+	drm_modeset_drop_locks(ctx);
+	drm_modeset_acquire_fini(ctx);
+
+	return false;
 }
 
 void intel_release_load_detect_pipe(struct drm_connector *connector,
-				    struct intel_load_detect_pipe *old)
+				    struct intel_load_detect_pipe *old,
+				    struct drm_modeset_acquire_ctx *ctx)
 {
 	struct intel_encoder *intel_encoder =
 		intel_attached_encoder(connector);
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_crtc *crtc = encoder->crtc;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s], [ENCODER:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector),
-		      encoder->base.id, drm_get_encoder_name(encoder));
+		      connector->base.id, connector->name,
+		      encoder->base.id, encoder->name);
 
 	if (old->load_detect_temp) {
 		to_intel_connector(connector)->new_encoder = NULL;
 		intel_encoder->new_crtc = NULL;
+		intel_crtc->new_enabled = false;
+		intel_crtc->new_config = NULL;
 		intel_set_mode(crtc, NULL, 0, 0, NULL);
 
 		if (old->release_fb) {
@@ -7671,7 +8463,7 @@ void intel_release_load_detect_pipe(struct drm_connector *connector,
 			drm_framebuffer_unreference(old->release_fb);
 		}
 
-		mutex_unlock(&crtc->mutex);
+		goto unlock;
 		return;
 	}
 
@@ -7679,7 +8471,9 @@ void intel_release_load_detect_pipe(struct drm_connector *connector,
 	if (old->dpms_mode != DRM_MODE_DPMS_ON)
 		connector->funcs->dpms(connector, old->dpms_mode);
 
-	mutex_unlock(&crtc->mutex);
+unlock:
+	drm_modeset_drop_locks(ctx);
+	drm_modeset_acquire_fini(ctx);
 }
 
 static int i9xx_pll_refclk(struct drm_device *dev,
@@ -7689,7 +8483,7 @@ static int i9xx_pll_refclk(struct drm_device *dev,
 	u32 dpll = pipe_config->dpll_hw_state.dpll;
 
 	if ((dpll & PLL_REF_INPUT_MASK) == PLLB_REF_INPUT_SPREADSPECTRUMIN)
-		return dev_priv->vbt.lvds_ssc_freq * 1000;
+		return dev_priv->vbt.lvds_ssc_freq;
 	else if (HAS_PCH_SPLIT(dev))
 		return 120000;
 	else if (!IS_GEN2(dev))
@@ -7752,12 +8546,17 @@ static void i9xx_crtc_clock_get(struct intel_crtc *crtc,
 		else
 			i9xx_clock(refclk, &clock);
 	} else {
-		bool is_lvds = (pipe == 1) && (I915_READ(LVDS) & LVDS_PORT_EN);
+		u32 lvds = IS_I830(dev) ? 0 : I915_READ(LVDS);
+		bool is_lvds = (pipe == 1) && (lvds & LVDS_PORT_EN);
 
 		if (is_lvds) {
 			clock.p1 = ffs((dpll & DPLL_FPA01_P1_POST_DIV_MASK_I830_LVDS) >>
 				       DPLL_FPA01_P1_POST_DIV_SHIFT);
-			clock.p2 = 14;
+
+			if (lvds & LVDS_CLKB_POWER_UP)
+				clock.p2 = 7;
+			else
+				clock.p2 = 14;
 		} else {
 			if (dpll & PLL_P1_DIVIDE_BY_TWO)
 				clock.p1 = 2;
@@ -7871,7 +8670,7 @@ struct drm_display_mode *intel_crtc_mode_get(struct drm_device *dev,
 static void intel_increase_pllclock(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_crtc->pipe;
 	int dpll_reg = DPLL(pipe);
@@ -7902,7 +8701,7 @@ static void intel_increase_pllclock(struct drm_crtc *crtc)
 static void intel_decrease_pllclock(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 
 	if (HAS_PCH_SPLIT(dev))
@@ -7939,8 +8738,12 @@ void intel_mark_busy(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	hsw_package_c8_gpu_busy(dev_priv);
+	if (dev_priv->mm.busy)
+		return;
+
+	intel_runtime_pm_get(dev_priv);
 	i915_update_gfx_val(dev_priv);
+	dev_priv->mm.busy = true;
 }
 
 void intel_mark_idle(struct drm_device *dev)
@@ -7948,36 +8751,42 @@ void intel_mark_idle(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc;
 
-	hsw_package_c8_gpu_idle(dev_priv);
-
-	if (!i915_powersave)
+	if (!dev_priv->mm.busy)
 		return;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
-		if (!crtc->fb)
+	dev_priv->mm.busy = false;
+
+	if (!i915.powersave)
+		goto out;
+
+	for_each_crtc(dev, crtc) {
+		if (!crtc->primary->fb)
 			continue;
 
 		intel_decrease_pllclock(crtc);
 	}
 
-	if (dev_priv->info->gen >= 6)
+	if (INTEL_INFO(dev)->gen >= 6)
 		gen6_rps_idle(dev->dev_private);
+
+out:
+	intel_runtime_pm_put(dev_priv);
 }
 
 void intel_mark_fb_busy(struct drm_i915_gem_object *obj,
-			struct intel_ring_buffer *ring)
+			struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_crtc *crtc;
 
-	if (!i915_powersave)
+	if (!i915.powersave)
 		return;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
-		if (!crtc->fb)
+	for_each_crtc(dev, crtc) {
+		if (!crtc->primary->fb)
 			continue;
 
-		if (to_intel_framebuffer(crtc->fb)->obj != obj)
+		if (to_intel_framebuffer(crtc->primary->fb)->obj != obj)
 			continue;
 
 		intel_increase_pllclock(crtc);
@@ -8033,7 +8842,7 @@ static void intel_unpin_work_fn(struct work_struct *__work)
 static void do_intel_finish_page_flip(struct drm_device *dev,
 				      struct drm_crtc *crtc)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_unpin_work *work;
 	unsigned long flags;
@@ -8061,7 +8870,7 @@ static void do_intel_finish_page_flip(struct drm_device *dev,
 	if (work->event)
 		drm_send_vblank_event(dev, intel_crtc->pipe, work->event);
 
-	drm_vblank_put(dev, intel_crtc->pipe);
+	drm_crtc_vblank_put(crtc);
 
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
@@ -8074,7 +8883,7 @@ static void do_intel_finish_page_flip(struct drm_device *dev,
 
 void intel_finish_page_flip(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
 
 	do_intel_finish_page_flip(dev, crtc);
@@ -8082,15 +8891,57 @@ void intel_finish_page_flip(struct drm_device *dev, int pipe)
 
 void intel_finish_page_flip_plane(struct drm_device *dev, int plane)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc = dev_priv->plane_to_crtc_mapping[plane];
 
 	do_intel_finish_page_flip(dev, crtc);
 }
 
+/* Is 'a' after or equal to 'b'? */
+static bool g4x_flip_count_after_eq(u32 a, u32 b)
+{
+	return !((a - b) & 0x80000000);
+}
+
+static bool page_flip_finished(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	/*
+	 * The relevant registers doen't exist on pre-ctg.
+	 * As the flip done interrupt doesn't trigger for mmio
+	 * flips on gmch platforms, a flip count check isn't
+	 * really needed there. But since ctg has the registers,
+	 * include it in the check anyway.
+	 */
+	if (INTEL_INFO(dev)->gen < 5 && !IS_G4X(dev))
+		return true;
+
+	/*
+	 * A DSPSURFLIVE check isn't enough in case the mmio and CS flips
+	 * used the same base address. In that case the mmio flip might
+	 * have completed, but the CS hasn't even executed the flip yet.
+	 *
+	 * A flip count check isn't enough as the CS might have updated
+	 * the base address just after start of vblank, but before we
+	 * managed to process the interrupt. This means we'd complete the
+	 * CS flip too soon.
+	 *
+	 * Combining both checks should get us a good enough result. It may
+	 * still happen that the CS flip has been executed, but has not
+	 * yet actually completed. But in case the base address is the same
+	 * anyway, we don't really care.
+	 */
+	return (I915_READ(DSPSURFLIVE(crtc->plane)) & ~0xfff) ==
+		crtc->unpin_work->gtt_offset &&
+		g4x_flip_count_after_eq(I915_READ(PIPE_FLIPCOUNT_GM45(crtc->pipe)),
+				    crtc->unpin_work->flip_count);
+}
+
 void intel_prepare_page_flip(struct drm_device *dev, int plane)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(dev_priv->plane_to_crtc_mapping[plane]);
 	unsigned long flags;
@@ -8100,12 +8951,12 @@ void intel_prepare_page_flip(struct drm_device *dev, int plane)
 	 * is also accompanied by a spurious intel_prepare_page_flip().
 	 */
 	spin_lock_irqsave(&dev->event_lock, flags);
-	if (intel_crtc->unpin_work)
+	if (intel_crtc->unpin_work && page_flip_finished(intel_crtc))
 		atomic_inc_not_zero(&intel_crtc->unpin_work->pending);
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 }
 
-inline static void intel_mark_page_flip_active(struct intel_crtc *intel_crtc)
+static inline void intel_mark_page_flip_active(struct intel_crtc *intel_crtc)
 {
 	/* Ensure that the work item is consistent when activating it ... */
 	smp_wmb();
@@ -8118,21 +8969,16 @@ static int intel_gen2_queue_flip(struct drm_device *dev,
 				 struct drm_crtc *crtc,
 				 struct drm_framebuffer *fb,
 				 struct drm_i915_gem_object *obj,
+				 struct intel_engine_cs *ring,
 				 uint32_t flags)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 flip_mask;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
 	int ret;
 
-	ret = intel_pin_and_fence_fb_obj(dev, obj, ring);
-	if (ret)
-		goto err;
-
 	ret = intel_ring_begin(ring, 6);
 	if (ret)
-		goto err_unpin;
+		return ret;
 
 	/* Can't queue multiple flips, so wait for the previous
 	 * one to finish before executing the next.
@@ -8146,38 +8992,28 @@ static int intel_gen2_queue_flip(struct drm_device *dev,
 	intel_ring_emit(ring, MI_DISPLAY_FLIP |
 			MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
 	intel_ring_emit(ring, fb->pitches[0]);
-	intel_ring_emit(ring, i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+	intel_ring_emit(ring, intel_crtc->unpin_work->gtt_offset);
 	intel_ring_emit(ring, 0); /* aux display base address, unused */
 
 	intel_mark_page_flip_active(intel_crtc);
 	__intel_ring_advance(ring);
 	return 0;
-
-err_unpin:
-	intel_unpin_fb_obj(obj);
-err:
-	return ret;
 }
 
 static int intel_gen3_queue_flip(struct drm_device *dev,
 				 struct drm_crtc *crtc,
 				 struct drm_framebuffer *fb,
 				 struct drm_i915_gem_object *obj,
+				 struct intel_engine_cs *ring,
 				 uint32_t flags)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	u32 flip_mask;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
 	int ret;
 
-	ret = intel_pin_and_fence_fb_obj(dev, obj, ring);
-	if (ret)
-		goto err;
-
 	ret = intel_ring_begin(ring, 6);
 	if (ret)
-		goto err_unpin;
+		return ret;
 
 	if (intel_crtc->plane)
 		flip_mask = MI_WAIT_FOR_PLANE_B_FLIP;
@@ -8188,38 +9024,29 @@ static int intel_gen3_queue_flip(struct drm_device *dev,
 	intel_ring_emit(ring, MI_DISPLAY_FLIP_I915 |
 			MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
 	intel_ring_emit(ring, fb->pitches[0]);
-	intel_ring_emit(ring, i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+	intel_ring_emit(ring, intel_crtc->unpin_work->gtt_offset);
 	intel_ring_emit(ring, MI_NOOP);
 
 	intel_mark_page_flip_active(intel_crtc);
 	__intel_ring_advance(ring);
 	return 0;
-
-err_unpin:
-	intel_unpin_fb_obj(obj);
-err:
-	return ret;
 }
 
 static int intel_gen4_queue_flip(struct drm_device *dev,
 				 struct drm_crtc *crtc,
 				 struct drm_framebuffer *fb,
 				 struct drm_i915_gem_object *obj,
+				 struct intel_engine_cs *ring,
 				 uint32_t flags)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	uint32_t pf, pipesrc;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
 	int ret;
 
-	ret = intel_pin_and_fence_fb_obj(dev, obj, ring);
-	if (ret)
-		goto err;
-
 	ret = intel_ring_begin(ring, 4);
 	if (ret)
-		goto err_unpin;
+		return ret;
 
 	/* i965+ uses the linear or tiled offsets from the
 	 * Display Registers (which do not change across a page-flip)
@@ -8228,8 +9055,7 @@ static int intel_gen4_queue_flip(struct drm_device *dev,
 	intel_ring_emit(ring, MI_DISPLAY_FLIP |
 			MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
 	intel_ring_emit(ring, fb->pitches[0]);
-	intel_ring_emit(ring,
-			(i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset) |
+	intel_ring_emit(ring, intel_crtc->unpin_work->gtt_offset |
 			obj->tiling_mode);
 
 	/* XXX Enabling the panel-fitter across page-flip is so far
@@ -8243,37 +9069,28 @@ static int intel_gen4_queue_flip(struct drm_device *dev,
 	intel_mark_page_flip_active(intel_crtc);
 	__intel_ring_advance(ring);
 	return 0;
-
-err_unpin:
-	intel_unpin_fb_obj(obj);
-err:
-	return ret;
 }
 
 static int intel_gen6_queue_flip(struct drm_device *dev,
 				 struct drm_crtc *crtc,
 				 struct drm_framebuffer *fb,
 				 struct drm_i915_gem_object *obj,
+				 struct intel_engine_cs *ring,
 				 uint32_t flags)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
 	uint32_t pf, pipesrc;
 	int ret;
 
-	ret = intel_pin_and_fence_fb_obj(dev, obj, ring);
-	if (ret)
-		goto err;
-
 	ret = intel_ring_begin(ring, 4);
 	if (ret)
-		goto err_unpin;
+		return ret;
 
 	intel_ring_emit(ring, MI_DISPLAY_FLIP |
 			MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
 	intel_ring_emit(ring, fb->pitches[0] | obj->tiling_mode);
-	intel_ring_emit(ring, i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+	intel_ring_emit(ring, intel_crtc->unpin_work->gtt_offset);
 
 	/* Contrary to the suggestions in the documentation,
 	 * "Enable Panel Fitter" does not seem to be required when page
@@ -8288,34 +9105,20 @@ static int intel_gen6_queue_flip(struct drm_device *dev,
 	intel_mark_page_flip_active(intel_crtc);
 	__intel_ring_advance(ring);
 	return 0;
-
-err_unpin:
-	intel_unpin_fb_obj(obj);
-err:
-	return ret;
 }
 
 static int intel_gen7_queue_flip(struct drm_device *dev,
 				 struct drm_crtc *crtc,
 				 struct drm_framebuffer *fb,
 				 struct drm_i915_gem_object *obj,
+				 struct intel_engine_cs *ring,
 				 uint32_t flags)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct intel_ring_buffer *ring;
 	uint32_t plane_bit = 0;
 	int len, ret;
 
-	ring = obj->ring;
-	if (IS_VALLEYVIEW(dev) || ring == NULL || ring->id != RCS)
-		ring = &dev_priv->ring[BCS];
-
-	ret = intel_pin_and_fence_fb_obj(dev, obj, ring);
-	if (ret)
-		goto err;
-
-	switch(intel_crtc->plane) {
+	switch (intel_crtc->plane) {
 	case PLANE_A:
 		plane_bit = MI_DISPLAY_FLIP_IVB_PLANE_A;
 		break;
@@ -8327,17 +9130,38 @@ static int intel_gen7_queue_flip(struct drm_device *dev,
 		break;
 	default:
 		WARN_ONCE(1, "unknown plane in flip command\n");
-		ret = -ENODEV;
-		goto err_unpin;
+		return -ENODEV;
 	}
 
 	len = 4;
-	if (ring->id == RCS)
+	if (ring->id == RCS) {
 		len += 6;
+		/*
+		 * On Gen 8, SRM is now taking an extra dword to accommodate
+		 * 48bits addresses, and we need a NOOP for the batch size to
+		 * stay even.
+		 */
+		if (IS_GEN8(dev))
+			len += 2;
+	}
+
+	/*
+	 * BSpec MI_DISPLAY_FLIP for IVB:
+	 * "The full packet must be contained within the same cache line."
+	 *
+	 * Currently the LRI+SRM+MI_DISPLAY_FLIP all fit within the same
+	 * cacheline, if we ever start emitting more commands before
+	 * the MI_DISPLAY_FLIP we may need to first emit everything else,
+	 * then do the cacheline alignment, and finally emit the
+	 * MI_DISPLAY_FLIP.
+	 */
+	ret = intel_ring_cacheline_align(ring);
+	if (ret)
+		return ret;
 
 	ret = intel_ring_begin(ring, len);
 	if (ret)
-		goto err_unpin;
+		return ret;
 
 	/* Unmask the flip-done completion message. Note that the bspec says that
 	 * we should do this for both the BCS and RCS, and that we must not unmask
@@ -8354,31 +9178,35 @@ static int intel_gen7_queue_flip(struct drm_device *dev,
 		intel_ring_emit(ring, ~(DERRMR_PIPEA_PRI_FLIP_DONE |
 					DERRMR_PIPEB_PRI_FLIP_DONE |
 					DERRMR_PIPEC_PRI_FLIP_DONE));
-		intel_ring_emit(ring, MI_STORE_REGISTER_MEM(1) |
-				MI_SRM_LRM_GLOBAL_GTT);
+		if (IS_GEN8(dev))
+			intel_ring_emit(ring, MI_STORE_REGISTER_MEM_GEN8(1) |
+					      MI_SRM_LRM_GLOBAL_GTT);
+		else
+			intel_ring_emit(ring, MI_STORE_REGISTER_MEM(1) |
+					      MI_SRM_LRM_GLOBAL_GTT);
 		intel_ring_emit(ring, DERRMR);
 		intel_ring_emit(ring, ring->scratch.gtt_offset + 256);
+		if (IS_GEN8(dev)) {
+			intel_ring_emit(ring, 0);
+			intel_ring_emit(ring, MI_NOOP);
+		}
 	}
 
 	intel_ring_emit(ring, MI_DISPLAY_FLIP_I915 | plane_bit);
 	intel_ring_emit(ring, (fb->pitches[0] | obj->tiling_mode));
-	intel_ring_emit(ring, i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+	intel_ring_emit(ring, intel_crtc->unpin_work->gtt_offset);
 	intel_ring_emit(ring, (MI_NOOP));
 
 	intel_mark_page_flip_active(intel_crtc);
 	__intel_ring_advance(ring);
 	return 0;
-
-err_unpin:
-	intel_unpin_fb_obj(obj);
-err:
-	return ret;
 }
 
 static int intel_default_queue_flip(struct drm_device *dev,
 				    struct drm_crtc *crtc,
 				    struct drm_framebuffer *fb,
 				    struct drm_i915_gem_object *obj,
+				    struct intel_engine_cs *ring,
 				    uint32_t flags)
 {
 	return -ENODEV;
@@ -8391,15 +9219,16 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_framebuffer *old_fb = crtc->fb;
+	struct drm_framebuffer *old_fb = crtc->primary->fb;
 	struct drm_i915_gem_object *obj = to_intel_framebuffer(fb)->obj;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_unpin_work *work;
+	struct intel_engine_cs *ring;
 	unsigned long flags;
 	int ret;
 
 	/* Can't change pixel format via MI display flips. */
-	if (fb->pixel_format != crtc->fb->pixel_format)
+	if (fb->pixel_format != crtc->primary->fb->pixel_format)
 		return -EINVAL;
 
 	/*
@@ -8407,10 +9236,13 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 	 * Note that pitch changes could also affect these register.
 	 */
 	if (INTEL_INFO(dev)->gen > 3 &&
-	    (fb->offsets[0] != crtc->fb->offsets[0] ||
-	     fb->pitches[0] != crtc->fb->pitches[0]))
+	    (fb->offsets[0] != crtc->primary->fb->offsets[0] ||
+	     fb->pitches[0] != crtc->primary->fb->pitches[0]))
 		return -EINVAL;
 
+	if (i915_terminally_wedged(&dev_priv->gpu_error))
+		goto out_hang;
+
 	work = kzalloc(sizeof(*work), GFP_KERNEL);
 	if (work == NULL)
 		return -ENOMEM;
@@ -8420,7 +9252,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 	work->old_fb_obj = to_intel_framebuffer(old_fb)->obj;
 	INIT_WORK(&work->work, intel_unpin_work_fn);
 
-	ret = drm_vblank_get(dev, intel_crtc->pipe);
+	ret = drm_crtc_vblank_get(crtc);
 	if (ret)
 		goto free_work;
 
@@ -8429,7 +9261,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 	if (intel_crtc->unpin_work) {
 		spin_unlock_irqrestore(&dev->event_lock, flags);
 		kfree(work);
-		drm_vblank_put(dev, intel_crtc->pipe);
+		drm_crtc_vblank_put(crtc);
 
 		DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
 		return -EBUSY;
@@ -8448,7 +9280,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 	drm_gem_object_reference(&work->old_fb_obj->base);
 	drm_gem_object_reference(&obj->base);
 
-	crtc->fb = fb;
+	crtc->primary->fb = fb;
 
 	work->pending_flip_obj = obj;
 
@@ -8457,10 +9289,30 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 	atomic_inc(&intel_crtc->unpin_work_count);
 	intel_crtc->reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
 
-	ret = dev_priv->display.queue_flip(dev, crtc, fb, obj, page_flip_flags);
+	if (INTEL_INFO(dev)->gen >= 5 || IS_G4X(dev))
+		work->flip_count = I915_READ(PIPE_FLIPCOUNT_GM45(intel_crtc->pipe)) + 1;
+
+	if (IS_VALLEYVIEW(dev)) {
+		ring = &dev_priv->ring[BCS];
+	} else if (INTEL_INFO(dev)->gen >= 7) {
+		ring = obj->ring;
+		if (ring == NULL || ring->id != RCS)
+			ring = &dev_priv->ring[BCS];
+	} else {
+		ring = &dev_priv->ring[RCS];
+	}
+
+	ret = intel_pin_and_fence_fb_obj(dev, obj, ring);
 	if (ret)
 		goto cleanup_pending;
 
+	work->gtt_offset =
+		i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset;
+
+	ret = dev_priv->display.queue_flip(dev, crtc, fb, obj, ring, page_flip_flags);
+	if (ret)
+		goto cleanup_unpin;
+
 	intel_disable_fbc(dev);
 	intel_mark_fb_busy(obj, NULL);
 	mutex_unlock(&dev->struct_mutex);
@@ -8469,9 +9321,11 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
 
 	return 0;
 
+cleanup_unpin:
+	intel_unpin_fb_obj(obj);
 cleanup_pending:
 	atomic_dec(&intel_crtc->unpin_work_count);
-	crtc->fb = old_fb;
+	crtc->primary->fb = old_fb;
 	drm_gem_object_unreference(&work->old_fb_obj->base);
 	drm_gem_object_unreference(&obj->base);
 	mutex_unlock(&dev->struct_mutex);
@@ -8481,10 +9335,17 @@ cleanup:
 	intel_crtc->unpin_work = NULL;
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	drm_vblank_put(dev, intel_crtc->pipe);
+	drm_crtc_vblank_put(crtc);
 free_work:
 	kfree(work);
 
+	if (ret == -EIO) {
+out_hang:
+		intel_crtc_wait_for_pending_flips(crtc);
+		ret = intel_pipe_set_base(crtc, crtc->x, crtc->y, fb);
+		if (ret == 0 && event)
+			drm_send_vblank_event(dev, intel_crtc->pipe, event);
+	}
 	return ret;
 }
 
@@ -8493,28 +9354,6 @@ static struct drm_crtc_helper_funcs intel_helper_funcs = {
 	.load_lut = intel_crtc_load_lut,
 };
 
-static bool intel_encoder_crtc_ok(struct drm_encoder *encoder,
-				  struct drm_crtc *crtc)
-{
-	struct drm_device *dev;
-	struct drm_crtc *tmp;
-	int crtc_mask = 1;
-
-	WARN(!crtc, "checking null crtc?\n");
-
-	dev = crtc->dev;
-
-	list_for_each_entry(tmp, &dev->mode_config.crtc_list, head) {
-		if (tmp == crtc)
-			break;
-		crtc_mask <<= 1;
-	}
-
-	if (encoder->possible_crtcs & crtc_mask)
-		return true;
-	return false;
-}
-
 /**
  * intel_modeset_update_staged_output_state
  *
@@ -8523,6 +9362,7 @@ static bool intel_encoder_crtc_ok(struct drm_encoder *encoder,
  */
 static void intel_modeset_update_staged_output_state(struct drm_device *dev)
 {
+	struct intel_crtc *crtc;
 	struct intel_encoder *encoder;
 	struct intel_connector *connector;
 
@@ -8537,6 +9377,15 @@ static void intel_modeset_update_staged_output_state(struct drm_device *dev)
 		encoder->new_crtc =
 			to_intel_crtc(encoder->base.crtc);
 	}
+
+	for_each_intel_crtc(dev, crtc) {
+		crtc->new_enabled = crtc->base.enabled;
+
+		if (crtc->new_enabled)
+			crtc->new_config = &crtc->config;
+		else
+			crtc->new_config = NULL;
+	}
 }
 
 /**
@@ -8546,6 +9395,7 @@ static void intel_modeset_update_staged_output_state(struct drm_device *dev)
  */
 static void intel_modeset_commit_output_state(struct drm_device *dev)
 {
+	struct intel_crtc *crtc;
 	struct intel_encoder *encoder;
 	struct intel_connector *connector;
 
@@ -8558,17 +9408,21 @@ static void intel_modeset_commit_output_state(struct drm_device *dev)
 			    base.head) {
 		encoder->base.crtc = &encoder->new_crtc->base;
 	}
+
+	for_each_intel_crtc(dev, crtc) {
+		crtc->base.enabled = crtc->new_enabled;
+	}
 }
 
 static void
-connected_sink_compute_bpp(struct intel_connector * connector,
+connected_sink_compute_bpp(struct intel_connector *connector,
 			   struct intel_crtc_config *pipe_config)
 {
 	int bpp = pipe_config->pipe_bpp;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s] checking for sink bpp constrains\n",
 		connector->base.base.id,
-		drm_get_connector_name(&connector->base));
+		connector->base.name);
 
 	/* Don't use an invalid EDID bpc value */
 	if (connector->base.display_info.bpc &&
@@ -8698,23 +9552,47 @@ static void intel_dump_pipe_config(struct intel_crtc *crtc,
 	DRM_DEBUG_KMS("double wide: %i\n", pipe_config->double_wide);
 }
 
-static bool check_encoder_cloning(struct drm_crtc *crtc)
+static bool encoders_cloneable(const struct intel_encoder *a,
+			       const struct intel_encoder *b)
 {
-	int num_encoders = 0;
-	bool uncloneable_encoders = false;
+	/* masks could be asymmetric, so check both ways */
+	return a == b || (a->cloneable & (1 << b->type) &&
+			  b->cloneable & (1 << a->type));
+}
+
+static bool check_single_encoder_cloning(struct intel_crtc *crtc,
+					 struct intel_encoder *encoder)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct intel_encoder *source_encoder;
+
+	list_for_each_entry(source_encoder,
+			    &dev->mode_config.encoder_list, base.head) {
+		if (source_encoder->new_crtc != crtc)
+			continue;
+
+		if (!encoders_cloneable(encoder, source_encoder))
+			return false;
+	}
+
+	return true;
+}
+
+static bool check_encoder_cloning(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
 	struct intel_encoder *encoder;
 
-	list_for_each_entry(encoder, &crtc->dev->mode_config.encoder_list,
-			    base.head) {
-		if (&encoder->new_crtc->base != crtc)
+	list_for_each_entry(encoder,
+			    &dev->mode_config.encoder_list, base.head) {
+		if (encoder->new_crtc != crtc)
 			continue;
 
-		num_encoders++;
-		if (!encoder->cloneable)
-			uncloneable_encoders = true;
+		if (!check_single_encoder_cloning(crtc, encoder))
+			return false;
 	}
 
-	return !(num_encoders > 1 && uncloneable_encoders);
+	return true;
 }
 
 static struct intel_crtc_config *
@@ -8728,7 +9606,7 @@ intel_modeset_pipe_config(struct drm_crtc *crtc,
 	int plane_bpp, ret = -EINVAL;
 	bool retry = true;
 
-	if (!check_encoder_cloning(crtc)) {
+	if (!check_encoder_cloning(to_intel_crtc(crtc))) {
 		DRM_DEBUG_KMS("rejecting invalid cloning configuration\n");
 		return ERR_PTR(-EINVAL);
 	}
@@ -8884,29 +9762,21 @@ intel_modeset_affected_pipes(struct drm_crtc *crtc, unsigned *modeset_pipes,
 			*prepare_pipes |= 1 << encoder->new_crtc->pipe;
 	}
 
-	/* Check for any pipes that will be fully disabled ... */
-	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list,
-			    base.head) {
-		bool used = false;
-
-		/* Don't try to disable disabled crtcs. */
-		if (!intel_crtc->base.enabled)
+	/* Check for pipes that will be enabled/disabled ... */
+	for_each_intel_crtc(dev, intel_crtc) {
+		if (intel_crtc->base.enabled == intel_crtc->new_enabled)
 			continue;
 
-		list_for_each_entry(encoder, &dev->mode_config.encoder_list,
-				    base.head) {
-			if (encoder->new_crtc == intel_crtc)
-				used = true;
-		}
-
-		if (!used)
+		if (!intel_crtc->new_enabled)
 			*disable_pipes |= 1 << intel_crtc->pipe;
+		else
+			*prepare_pipes |= 1 << intel_crtc->pipe;
 	}
 
 
 	/* set_mode is also used to update properties on life display pipes. */
 	intel_crtc = to_intel_crtc(crtc);
-	if (crtc->enabled)
+	if (intel_crtc->new_enabled)
 		*prepare_pipes |= 1 << intel_crtc->pipe;
 
 	/*
@@ -8965,10 +9835,12 @@ intel_modeset_update_state(struct drm_device *dev, unsigned prepare_pipes)
 
 	intel_modeset_commit_output_state(dev);
 
-	/* Update computed state. */
-	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list,
-			    base.head) {
-		intel_crtc->base.enabled = intel_crtc_in_use(&intel_crtc->base);
+	/* Double check state. */
+	for_each_intel_crtc(dev, intel_crtc) {
+		WARN_ON(intel_crtc->base.enabled != intel_crtc_in_use(&intel_crtc->base));
+		WARN_ON(intel_crtc->new_config &&
+			intel_crtc->new_config != &intel_crtc->config);
+		WARN_ON(intel_crtc->base.enabled != !!intel_crtc->new_config);
 	}
 
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
@@ -9093,6 +9965,12 @@ intel_pipe_config_compare(struct drm_device *dev,
 	PIPE_CONF_CHECK_I(adjusted_mode.crtc_vsync_end);
 
 	PIPE_CONF_CHECK_I(pixel_multiplier);
+	PIPE_CONF_CHECK_I(has_hdmi_sink);
+	if ((INTEL_INFO(dev)->gen < 8 && !IS_HASWELL(dev)) ||
+	    IS_VALLEYVIEW(dev))
+		PIPE_CONF_CHECK_I(limited_color_range);
+
+	PIPE_CONF_CHECK_I(has_audio);
 
 	PIPE_CONF_CHECK_FLAGS(adjusted_mode.flags,
 			      DRM_MODE_FLAG_INTERLACE);
@@ -9111,18 +9989,31 @@ intel_pipe_config_compare(struct drm_device *dev,
 	PIPE_CONF_CHECK_I(pipe_src_w);
 	PIPE_CONF_CHECK_I(pipe_src_h);
 
-	PIPE_CONF_CHECK_I(gmch_pfit.control);
-	/* pfit ratios are autocomputed by the hw on gen4+ */
-	if (INTEL_INFO(dev)->gen < 4)
-		PIPE_CONF_CHECK_I(gmch_pfit.pgm_ratios);
-	PIPE_CONF_CHECK_I(gmch_pfit.lvds_border_bits);
+	/*
+	 * FIXME: BIOS likes to set up a cloned config with lvds+external
+	 * screen. Since we don't yet re-compute the pipe config when moving
+	 * just the lvds port away to another pipe the sw tracking won't match.
+	 *
+	 * Proper atomic modesets with recomputed global state will fix this.
+	 * Until then just don't check gmch state for inherited modes.
+	 */
+	if (!PIPE_CONF_QUIRK(PIPE_CONFIG_QUIRK_INHERITED_MODE)) {
+		PIPE_CONF_CHECK_I(gmch_pfit.control);
+		/* pfit ratios are autocomputed by the hw on gen4+ */
+		if (INTEL_INFO(dev)->gen < 4)
+			PIPE_CONF_CHECK_I(gmch_pfit.pgm_ratios);
+		PIPE_CONF_CHECK_I(gmch_pfit.lvds_border_bits);
+	}
+
 	PIPE_CONF_CHECK_I(pch_pfit.enabled);
 	if (current_config->pch_pfit.enabled) {
 		PIPE_CONF_CHECK_I(pch_pfit.pos);
 		PIPE_CONF_CHECK_I(pch_pfit.size);
 	}
 
-	PIPE_CONF_CHECK_I(ips_enabled);
+	/* BDW+ don't expose a synchronous way to read the state */
+	if (IS_HASWELL(dev))
+		PIPE_CONF_CHECK_I(ips_enabled);
 
 	PIPE_CONF_CHECK_I(double_wide);
 
@@ -9135,10 +10026,8 @@ intel_pipe_config_compare(struct drm_device *dev,
 	if (IS_G4X(dev) || INTEL_INFO(dev)->gen >= 5)
 		PIPE_CONF_CHECK_I(pipe_bpp);
 
-	if (!HAS_DDI(dev)) {
-		PIPE_CONF_CHECK_CLOCK_FUZZY(adjusted_mode.crtc_clock);
-		PIPE_CONF_CHECK_CLOCK_FUZZY(port_clock);
-	}
+	PIPE_CONF_CHECK_CLOCK_FUZZY(adjusted_mode.crtc_clock);
+	PIPE_CONF_CHECK_CLOCK_FUZZY(port_clock);
 
 #undef PIPE_CONF_CHECK_X
 #undef PIPE_CONF_CHECK_I
@@ -9179,7 +10068,7 @@ check_encoder_state(struct drm_device *dev)
 
 		DRM_DEBUG_KMS("[ENCODER:%d:%s]\n",
 			      encoder->base.base.id,
-			      drm_get_encoder_name(&encoder->base));
+			      encoder->base.name);
 
 		WARN(&encoder->new_crtc->base != encoder->base.crtc,
 		     "encoder's stage crtc doesn't match current crtc\n");
@@ -9226,13 +10115,12 @@ check_encoder_state(struct drm_device *dev)
 static void
 check_crtc_state(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *crtc;
 	struct intel_encoder *encoder;
 	struct intel_crtc_config pipe_config;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list,
-			    base.head) {
+	for_each_intel_crtc(dev, crtc) {
 		bool enabled = false;
 		bool active = false;
 
@@ -9294,7 +10182,7 @@ check_crtc_state(struct drm_device *dev)
 static void
 check_shared_dpll_state(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *crtc;
 	struct intel_dpll_hw_state dpll_hw_state;
 	int i;
@@ -9321,8 +10209,7 @@ check_shared_dpll_state(struct drm_device *dev)
 		     "pll on state mismatch (expected %i, found %i)\n",
 		     pll->on, active);
 
-		list_for_each_entry(crtc, &dev->mode_config.crtc_list,
-				    base.head) {
+		for_each_intel_crtc(dev, crtc) {
 			if (crtc->base.enabled && intel_crtc_to_shared_dpll(crtc) == pll)
 				enabled_crtcs++;
 			if (crtc->active && intel_crtc_to_shared_dpll(crtc) == pll)
@@ -9362,27 +10249,63 @@ void ironlake_check_encoder_dotclock(const struct intel_crtc_config *pipe_config
 	     pipe_config->adjusted_mode.crtc_clock, dotclock);
 }
 
+static void update_scanline_offset(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+
+	/*
+	 * The scanline counter increments at the leading edge of hsync.
+	 *
+	 * On most platforms it starts counting from vtotal-1 on the
+	 * first active line. That means the scanline counter value is
+	 * always one less than what we would expect. Ie. just after
+	 * start of vblank, which also occurs at start of hsync (on the
+	 * last active line), the scanline counter will read vblank_start-1.
+	 *
+	 * On gen2 the scanline counter starts counting from 1 instead
+	 * of vtotal-1, so we have to subtract one (or rather add vtotal-1
+	 * to keep the value positive), instead of adding one.
+	 *
+	 * On HSW+ the behaviour of the scanline counter depends on the output
+	 * type. For DP ports it behaves like most other platforms, but on HDMI
+	 * there's an extra 1 line difference. So we need to add two instead of
+	 * one to the value.
+	 */
+	if (IS_GEN2(dev)) {
+		const struct drm_display_mode *mode = &crtc->config.adjusted_mode;
+		int vtotal;
+
+		vtotal = mode->crtc_vtotal;
+		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+			vtotal /= 2;
+
+		crtc->scanline_offset = vtotal - 1;
+	} else if (HAS_DDI(dev) &&
+		   intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_HDMI)) {
+		crtc->scanline_offset = 2;
+	} else
+		crtc->scanline_offset = 1;
+}
+
 static int __intel_set_mode(struct drm_crtc *crtc,
 			    struct drm_display_mode *mode,
 			    int x, int y, struct drm_framebuffer *fb)
 {
 	struct drm_device *dev = crtc->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct drm_display_mode *saved_mode, *saved_hwmode;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_display_mode *saved_mode;
 	struct intel_crtc_config *pipe_config = NULL;
 	struct intel_crtc *intel_crtc;
 	unsigned disable_pipes, prepare_pipes, modeset_pipes;
 	int ret = 0;
 
-	saved_mode = kcalloc(2, sizeof(*saved_mode), GFP_KERNEL);
+	saved_mode = kmalloc(sizeof(*saved_mode), GFP_KERNEL);
 	if (!saved_mode)
 		return -ENOMEM;
-	saved_hwmode = saved_mode + 1;
 
 	intel_modeset_affected_pipes(crtc, &modeset_pipes,
 				     &prepare_pipes, &disable_pipes);
 
-	*saved_hwmode = crtc->hwmode;
 	*saved_mode = crtc->mode;
 
 	/* Hack: Because we don't (yet) support global modeset on multiple
@@ -9400,6 +10323,21 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 		}
 		intel_dump_pipe_config(to_intel_crtc(crtc), pipe_config,
 				       "[modeset]");
+		to_intel_crtc(crtc)->new_config = pipe_config;
+	}
+
+	/*
+	 * See if the config requires any additional preparation, e.g.
+	 * to adjust global state with pipes off.  We need to do this
+	 * here so we can get the modeset_pipe updated config for the new
+	 * mode set on this crtc.  For other crtcs we need to use the
+	 * adjusted_mode bits in the crtc directly.
+	 */
+	if (IS_VALLEYVIEW(dev)) {
+		valleyview_modeset_global_pipes(dev, &prepare_pipes);
+
+		/* may have added more to prepare_pipes than we should */
+		prepare_pipes &= ~disable_pipes;
 	}
 
 	for_each_intel_crtc_masked(dev, disable_pipes, intel_crtc)
@@ -9418,6 +10356,15 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 		/* mode_set/enable/disable functions rely on a correct pipe
 		 * config. */
 		to_intel_crtc(crtc)->config = *pipe_config;
+		to_intel_crtc(crtc)->new_config = &to_intel_crtc(crtc)->config;
+
+		/*
+		 * Calculate and store various constants which
+		 * are later needed by vblank and swap-completion
+		 * timestamping. They are derived from true hwmode.
+		 */
+		drm_calc_timestamping_constants(crtc,
+						&pipe_config->adjusted_mode);
 	}
 
 	/* Only after disabling all output pipelines that will be changed can we
@@ -9431,33 +10378,43 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 	 * on the DPLL.
 	 */
 	for_each_intel_crtc_masked(dev, modeset_pipes, intel_crtc) {
-		ret = intel_crtc_mode_set(&intel_crtc->base,
-					  x, y, fb);
+		struct drm_framebuffer *old_fb;
+
+		mutex_lock(&dev->struct_mutex);
+		ret = intel_pin_and_fence_fb_obj(dev,
+						 to_intel_framebuffer(fb)->obj,
+						 NULL);
+		if (ret != 0) {
+			DRM_ERROR("pin & fence failed\n");
+			mutex_unlock(&dev->struct_mutex);
+			goto done;
+		}
+		old_fb = crtc->primary->fb;
+		if (old_fb)
+			intel_unpin_fb_obj(to_intel_framebuffer(old_fb)->obj);
+		mutex_unlock(&dev->struct_mutex);
+
+		crtc->primary->fb = fb;
+		crtc->x = x;
+		crtc->y = y;
+
+		ret = dev_priv->display.crtc_mode_set(&intel_crtc->base,
+						      x, y, fb);
 		if (ret)
 			goto done;
 	}
 
 	/* Now enable the clocks, plane, pipe, and connectors that we set up. */
-	for_each_intel_crtc_masked(dev, prepare_pipes, intel_crtc)
-		dev_priv->display.crtc_enable(&intel_crtc->base);
-
-	if (modeset_pipes) {
-		/* Store real post-adjustment hardware mode. */
-		crtc->hwmode = pipe_config->adjusted_mode;
+	for_each_intel_crtc_masked(dev, prepare_pipes, intel_crtc) {
+		update_scanline_offset(intel_crtc);
 
-		/* Calculate and store various constants which
-		 * are later needed by vblank and swap-completion
-		 * timestamping. They are derived from true hwmode.
-		 */
-		drm_calc_timestamping_constants(crtc);
+		dev_priv->display.crtc_enable(&intel_crtc->base);
 	}
 
 	/* FIXME: add subpixel order */
 done:
-	if (ret && crtc->enabled) {
-		crtc->hwmode = *saved_hwmode;
+	if (ret && crtc->enabled)
 		crtc->mode = *saved_mode;
-	}
 
 out:
 	kfree(pipe_config);
@@ -9481,7 +10438,7 @@ static int intel_set_mode(struct drm_crtc *crtc,
 
 void intel_crtc_restore_mode(struct drm_crtc *crtc)
 {
-	intel_set_mode(crtc, &crtc->mode, crtc->x, crtc->y, crtc->fb);
+	intel_set_mode(crtc, &crtc->mode, crtc->x, crtc->y, crtc->primary->fb);
 }
 
 #undef for_each_intel_crtc_masked
@@ -9493,16 +10450,24 @@ static void intel_set_config_free(struct intel_set_config *config)
 
 	kfree(config->save_connector_encoders);
 	kfree(config->save_encoder_crtcs);
+	kfree(config->save_crtc_enabled);
 	kfree(config);
 }
 
 static int intel_set_config_save_state(struct drm_device *dev,
 				       struct intel_set_config *config)
 {
+	struct drm_crtc *crtc;
 	struct drm_encoder *encoder;
 	struct drm_connector *connector;
 	int count;
 
+	config->save_crtc_enabled =
+		kcalloc(dev->mode_config.num_crtc,
+			sizeof(bool), GFP_KERNEL);
+	if (!config->save_crtc_enabled)
+		return -ENOMEM;
+
 	config->save_encoder_crtcs =
 		kcalloc(dev->mode_config.num_encoder,
 			sizeof(struct drm_crtc *), GFP_KERNEL);
@@ -9520,6 +10485,11 @@ static int intel_set_config_save_state(struct drm_device *dev,
 	 * restored, not the drivers personal bookkeeping.
 	 */
 	count = 0;
+	for_each_crtc(dev, crtc) {
+		config->save_crtc_enabled[count++] = crtc->enabled;
+	}
+
+	count = 0;
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
 		config->save_encoder_crtcs[count++] = encoder->crtc;
 	}
@@ -9535,11 +10505,22 @@ static int intel_set_config_save_state(struct drm_device *dev,
 static void intel_set_config_restore_state(struct drm_device *dev,
 					   struct intel_set_config *config)
 {
+	struct intel_crtc *crtc;
 	struct intel_encoder *encoder;
 	struct intel_connector *connector;
 	int count;
 
 	count = 0;
+	for_each_intel_crtc(dev, crtc) {
+		crtc->new_enabled = config->save_crtc_enabled[count++];
+
+		if (crtc->new_enabled)
+			crtc->new_config = &crtc->config;
+		else
+			crtc->new_config = NULL;
+	}
+
+	count = 0;
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
 		encoder->new_crtc =
 			to_intel_crtc(config->save_encoder_crtcs[count++]);
@@ -9581,13 +10562,13 @@ intel_set_config_compute_mode_changes(struct drm_mode_set *set,
 	 * and then just flip_or_move it */
 	if (is_crtc_connector_off(set)) {
 		config->mode_changed = true;
-	} else if (set->crtc->fb != set->fb) {
+	} else if (set->crtc->primary->fb != set->fb) {
 		/* If we have no fb then treat it as a full mode set */
-		if (set->crtc->fb == NULL) {
+		if (set->crtc->primary->fb == NULL) {
 			struct intel_crtc *intel_crtc =
 				to_intel_crtc(set->crtc);
 
-			if (intel_crtc->active && i915_fastboot) {
+			if (intel_crtc->active && i915.fastboot) {
 				DRM_DEBUG_KMS("crtc has no fb, will flip\n");
 				config->fb_changed = true;
 			} else {
@@ -9597,7 +10578,7 @@ intel_set_config_compute_mode_changes(struct drm_mode_set *set,
 		} else if (set->fb == NULL) {
 			config->mode_changed = true;
 		} else if (set->fb->pixel_format !=
-			   set->crtc->fb->pixel_format) {
+			   set->crtc->primary->fb->pixel_format) {
 			config->mode_changed = true;
 		} else {
 			config->fb_changed = true;
@@ -9623,9 +10604,9 @@ intel_modeset_stage_output_state(struct drm_device *dev,
 				 struct drm_mode_set *set,
 				 struct intel_set_config *config)
 {
-	struct drm_crtc *new_crtc;
 	struct intel_connector *connector;
 	struct intel_encoder *encoder;
+	struct intel_crtc *crtc;
 	int ro;
 
 	/* The upper layers ensure that we either disable a crtc or have a list
@@ -9654,7 +10635,7 @@ intel_modeset_stage_output_state(struct drm_device *dev,
 
 			DRM_DEBUG_KMS("[CONNECTOR:%d:%s] to [NOCRTC]\n",
 				connector->base.base.id,
-				drm_get_connector_name(&connector->base));
+				connector->base.name);
 		}
 
 
@@ -9668,6 +10649,8 @@ intel_modeset_stage_output_state(struct drm_device *dev,
 	/* Update crtc of enabled connectors. */
 	list_for_each_entry(connector, &dev->mode_config.connector_list,
 			    base.head) {
+		struct drm_crtc *new_crtc;
+
 		if (!connector->new_encoder)
 			continue;
 
@@ -9679,32 +10662,36 @@ intel_modeset_stage_output_state(struct drm_device *dev,
 		}
 
 		/* Make sure the new CRTC will work with the encoder */
-		if (!intel_encoder_crtc_ok(&connector->new_encoder->base,
-					   new_crtc)) {
+		if (!drm_encoder_crtc_ok(&connector->new_encoder->base,
+					 new_crtc)) {
 			return -EINVAL;
 		}
 		connector->encoder->new_crtc = to_intel_crtc(new_crtc);
 
 		DRM_DEBUG_KMS("[CONNECTOR:%d:%s] to [CRTC:%d]\n",
 			connector->base.base.id,
-			drm_get_connector_name(&connector->base),
+			connector->base.name,
 			new_crtc->base.id);
 	}
 
 	/* Check for any encoders that needs to be disabled. */
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list,
 			    base.head) {
+		int num_connectors = 0;
 		list_for_each_entry(connector,
 				    &dev->mode_config.connector_list,
 				    base.head) {
 			if (connector->new_encoder == encoder) {
 				WARN_ON(!connector->new_encoder->new_crtc);
-
-				goto next_encoder;
+				num_connectors++;
 			}
 		}
-		encoder->new_crtc = NULL;
-next_encoder:
+
+		if (num_connectors == 0)
+			encoder->new_crtc = NULL;
+		else if (num_connectors > 1)
+			return -EINVAL;
+
 		/* Only now check for crtc changes so we don't miss encoders
 		 * that will be disabled. */
 		if (&encoder->new_crtc->base != encoder->base.crtc) {
@@ -9714,9 +10701,57 @@ next_encoder:
 	}
 	/* Now we've also updated encoder->new_crtc for all encoders. */
 
+	for_each_intel_crtc(dev, crtc) {
+		crtc->new_enabled = false;
+
+		list_for_each_entry(encoder,
+				    &dev->mode_config.encoder_list,
+				    base.head) {
+			if (encoder->new_crtc == crtc) {
+				crtc->new_enabled = true;
+				break;
+			}
+		}
+
+		if (crtc->new_enabled != crtc->base.enabled) {
+			DRM_DEBUG_KMS("crtc %sabled, full mode switch\n",
+				      crtc->new_enabled ? "en" : "dis");
+			config->mode_changed = true;
+		}
+
+		if (crtc->new_enabled)
+			crtc->new_config = &crtc->config;
+		else
+			crtc->new_config = NULL;
+	}
+
 	return 0;
 }
 
+static void disable_crtc_nofb(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct intel_encoder *encoder;
+	struct intel_connector *connector;
+
+	DRM_DEBUG_KMS("Trying to restore without FB -> disabling pipe %c\n",
+		      pipe_name(crtc->pipe));
+
+	list_for_each_entry(connector, &dev->mode_config.connector_list, base.head) {
+		if (connector->new_encoder &&
+		    connector->new_encoder->new_crtc == crtc)
+			connector->new_encoder = NULL;
+	}
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+		if (encoder->new_crtc == crtc)
+			encoder->new_crtc = NULL;
+	}
+
+	crtc->new_enabled = false;
+	crtc->new_config = NULL;
+}
+
 static int intel_crtc_set_config(struct drm_mode_set *set)
 {
 	struct drm_device *dev;
@@ -9755,7 +10790,7 @@ static int intel_crtc_set_config(struct drm_mode_set *set)
 	save_set.mode = &set->crtc->mode;
 	save_set.x = set->crtc->x;
 	save_set.y = set->crtc->y;
-	save_set.fb = set->crtc->fb;
+	save_set.fb = set->crtc->primary->fb;
 
 	/* Compute whether we need a full modeset, only an fb base update or no
 	 * change at all. In the future we might also check whether only the
@@ -9775,6 +10810,16 @@ static int intel_crtc_set_config(struct drm_mode_set *set)
 
 		ret = intel_pipe_set_base(set->crtc,
 					  set->x, set->y, set->fb);
+		/*
+		 * In the fastboot case this may be our only check of the
+		 * state after boot.  It would be better to only do it on
+		 * the first update, but we don't have a nice way of doing that
+		 * (and really, set_config isn't used much for high freq page
+		 * flipping, so increasing its cost here shouldn't be a big
+		 * deal).
+		 */
+		if (i915.fastboot && ret == 0)
+			intel_modeset_check_state(set->crtc->dev);
 	}
 
 	if (ret) {
@@ -9783,6 +10828,15 @@ static int intel_crtc_set_config(struct drm_mode_set *set)
 fail:
 		intel_set_config_restore_state(dev, config);
 
+		/*
+		 * HACK: if the pipe was on, but we didn't have a framebuffer,
+		 * force the pipe off to avoid oopsing in the modeset code
+		 * due to fb==NULL. This should only happen during boot since
+		 * we don't yet reconstruct the FB from the hardware state.
+		 */
+		if (to_intel_crtc(save_set.crtc)->new_enabled && !save_set.fb)
+			disable_crtc_nofb(to_intel_crtc(save_set.crtc));
+
 		/* Try to restore the config */
 		if (config->mode_changed &&
 		    intel_set_mode(save_set.crtc, save_set.mode,
@@ -9835,7 +10889,7 @@ static void ibx_pch_dpll_enable(struct drm_i915_private *dev_priv,
 				struct intel_shared_dpll *pll)
 {
 	/* PCH refclock must be enabled first */
-	assert_pch_refclk_enabled(dev_priv);
+	ibx_assert_pch_refclk_enabled(dev_priv);
 
 	I915_WRITE(PCH_DPLL(pll->id), pll->hw_state.dpll);
 
@@ -9860,7 +10914,7 @@ static void ibx_pch_dpll_disable(struct drm_i915_private *dev_priv,
 	struct intel_crtc *crtc;
 
 	/* Make sure no transcoder isn't still depending on us. */
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, base.head) {
+	for_each_intel_crtc(dev, crtc) {
 		if (intel_crtc_to_shared_dpll(crtc) == pll)
 			assert_pch_transcoder_disabled(dev_priv, crtc->pipe);
 	}
@@ -9903,13 +10957,11 @@ static void intel_shared_dpll_init(struct drm_device *dev)
 		dev_priv->num_shared_dpll = 0;
 
 	BUG_ON(dev_priv->num_shared_dpll > I915_NUM_PLLS);
-	DRM_DEBUG_KMS("%i shared PLLs initialized\n",
-		      dev_priv->num_shared_dpll);
 }
 
 static void intel_crtc_init(struct drm_device *dev, int pipe)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc;
 	int i;
 
@@ -9926,27 +10978,38 @@ static void intel_crtc_init(struct drm_device *dev, int pipe)
 		intel_crtc->lut_b[i] = i;
 	}
 
-	/* Swap pipes & planes for FBC on pre-965 */
+	/*
+	 * On gen2/3 only plane A can do fbc, but the panel fitter and lvds port
+	 * is hooked to plane B. Hence we want plane A feeding pipe B.
+	 */
 	intel_crtc->pipe = pipe;
 	intel_crtc->plane = pipe;
-	if (IS_MOBILE(dev) && IS_GEN3(dev)) {
+	if (HAS_FBC(dev) && INTEL_INFO(dev)->gen < 4) {
 		DRM_DEBUG_KMS("swapping pipes & planes for FBC\n");
 		intel_crtc->plane = !pipe;
 	}
 
+	intel_crtc->cursor_base = ~0;
+	intel_crtc->cursor_cntl = ~0;
+
+	init_waitqueue_head(&intel_crtc->vbl_wait);
+
 	BUG_ON(pipe >= ARRAY_SIZE(dev_priv->plane_to_crtc_mapping) ||
 	       dev_priv->plane_to_crtc_mapping[intel_crtc->plane] != NULL);
 	dev_priv->plane_to_crtc_mapping[intel_crtc->plane] = &intel_crtc->base;
 	dev_priv->pipe_to_crtc_mapping[intel_crtc->pipe] = &intel_crtc->base;
 
 	drm_crtc_helper_add(&intel_crtc->base, &intel_helper_funcs);
+
+	WARN_ON(drm_crtc_index(&intel_crtc->base) != intel_crtc->pipe);
 }
 
 enum pipe intel_get_pipe_from_connector(struct intel_connector *connector)
 {
 	struct drm_encoder *encoder = connector->base.encoder;
+	struct drm_device *dev = connector->base.dev;
 
-	WARN_ON(!mutex_is_locked(&connector->base.dev->mode_config.mutex));
+	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
 
 	if (!encoder)
 		return INVALID_PIPE;
@@ -9987,12 +11050,7 @@ static int intel_encoder_clones(struct intel_encoder *encoder)
 
 	list_for_each_entry(source_encoder,
 			    &dev->mode_config.encoder_list, base.head) {
-
-		if (encoder == source_encoder)
-			index_mask |= (1 << entry);
-
-		/* Intel hw has only one MUX where enocoders could be cloned. */
-		if (encoder->cloneable && source_encoder->cloneable)
+		if (encoders_cloneable(encoder, source_encoder))
 			index_mask |= (1 << entry);
 
 		entry++;
@@ -10011,8 +11069,45 @@ static bool has_edp_a(struct drm_device *dev)
 	if ((I915_READ(DP_A) & DP_DETECTED) == 0)
 		return false;
 
-	if (IS_GEN5(dev) &&
-	    (I915_READ(ILK_DISPLAY_CHICKEN_FUSES) & ILK_eDP_A_DISABLE))
+	if (IS_GEN5(dev) && (I915_READ(FUSE_STRAP) & ILK_eDP_A_DISABLE))
+		return false;
+
+	return true;
+}
+
+const char *intel_output_name(int output)
+{
+	static const char *names[] = {
+		[INTEL_OUTPUT_UNUSED] = "Unused",
+		[INTEL_OUTPUT_ANALOG] = "Analog",
+		[INTEL_OUTPUT_DVO] = "DVO",
+		[INTEL_OUTPUT_SDVO] = "SDVO",
+		[INTEL_OUTPUT_LVDS] = "LVDS",
+		[INTEL_OUTPUT_TVOUT] = "TV",
+		[INTEL_OUTPUT_HDMI] = "HDMI",
+		[INTEL_OUTPUT_DISPLAYPORT] = "DisplayPort",
+		[INTEL_OUTPUT_EDP] = "eDP",
+		[INTEL_OUTPUT_DSI] = "DSI",
+		[INTEL_OUTPUT_UNKNOWN] = "Unknown",
+	};
+
+	if (output < 0 || output >= ARRAY_SIZE(names) || !names[output])
+		return "Invalid";
+
+	return names[output];
+}
+
+static bool intel_crt_present(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (IS_ULT(dev))
+		return false;
+
+	if (IS_CHERRYVIEW(dev))
+		return false;
+
+	if (IS_VALLEYVIEW(dev) && !dev_priv->vbt.int_crt_support)
 		return false;
 
 	return true;
@@ -10026,7 +11121,7 @@ static void intel_setup_outputs(struct drm_device *dev)
 
 	intel_lvds_init(dev);
 
-	if (!IS_ULT(dev))
+	if (intel_crt_present(dev))
 		intel_crt_init(dev);
 
 	if (HAS_DDI(dev)) {
@@ -10090,6 +11185,15 @@ static void intel_setup_outputs(struct drm_device *dev)
 				intel_dp_init(dev, VLV_DISPLAY_BASE + DP_C, PORT_C);
 		}
 
+		if (IS_CHERRYVIEW(dev)) {
+			if (I915_READ(VLV_DISPLAY_BASE + CHV_HDMID) & SDVO_DETECTED) {
+				intel_hdmi_init(dev, VLV_DISPLAY_BASE + CHV_HDMID,
+						PORT_D);
+				if (I915_READ(VLV_DISPLAY_BASE + DP_D) & DP_DETECTED)
+					intel_dp_init(dev, VLV_DISPLAY_BASE + DP_D, PORT_D);
+			}
+		}
+
 		intel_dsi_init(dev);
 	} else if (SUPPORTS_DIGITAL_OUTPUTS(dev)) {
 		bool found = false;
@@ -10143,18 +11247,13 @@ static void intel_setup_outputs(struct drm_device *dev)
 	drm_helper_move_panel_connectors_to_head(dev);
 }
 
-void intel_framebuffer_fini(struct intel_framebuffer *fb)
-{
-	drm_framebuffer_cleanup(&fb->base);
-	WARN_ON(!fb->obj->framebuffer_references--);
-	drm_gem_object_unreference_unlocked(&fb->obj->base);
-}
-
 static void intel_user_framebuffer_destroy(struct drm_framebuffer *fb)
 {
 	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
 
-	intel_framebuffer_fini(intel_fb);
+	drm_framebuffer_cleanup(fb);
+	WARN_ON(!intel_fb->obj->framebuffer_references--);
+	drm_gem_object_unreference_unlocked(&intel_fb->obj->base);
 	kfree(intel_fb);
 }
 
@@ -10173,12 +11272,12 @@ static const struct drm_framebuffer_funcs intel_fb_funcs = {
 	.create_handle = intel_user_framebuffer_create_handle,
 };
 
-int intel_framebuffer_init(struct drm_device *dev,
-			   struct intel_framebuffer *intel_fb,
-			   struct drm_mode_fb_cmd2 *mode_cmd,
-			   struct drm_i915_gem_object *obj)
+static int intel_framebuffer_init(struct drm_device *dev,
+				  struct intel_framebuffer *intel_fb,
+				  struct drm_mode_fb_cmd2 *mode_cmd,
+				  struct drm_i915_gem_object *obj)
 {
-	int aligned_height, tile_height;
+	int aligned_height;
 	int pitch_limit;
 	int ret;
 
@@ -10272,9 +11371,8 @@ int intel_framebuffer_init(struct drm_device *dev,
 	if (mode_cmd->offsets[0] != 0)
 		return -EINVAL;
 
-	tile_height = IS_GEN2(dev) ? 16 : 8;
-	aligned_height = ALIGN(mode_cmd->height,
-			       obj->tiling_mode ? tile_height : 1);
+	aligned_height = intel_align_height(dev, mode_cmd->height,
+					    obj->tiling_mode);
 	/* FIXME drm helper for size checks (especially planar formats)? */
 	if (obj->base.size < aligned_height * mode_cmd->pitches[0])
 		return -EINVAL;
@@ -10325,6 +11423,8 @@ static void intel_init_display(struct drm_device *dev)
 
 	if (HAS_PCH_SPLIT(dev) || IS_G4X(dev))
 		dev_priv->display.find_dpll = g4x_find_best_dpll;
+	else if (IS_CHERRYVIEW(dev))
+		dev_priv->display.find_dpll = chv_find_best_dpll;
 	else if (IS_VALLEYVIEW(dev))
 		dev_priv->display.find_dpll = vlv_find_best_dpll;
 	else if (IS_PINEVIEW(dev))
@@ -10334,32 +11434,40 @@ static void intel_init_display(struct drm_device *dev)
 
 	if (HAS_DDI(dev)) {
 		dev_priv->display.get_pipe_config = haswell_get_pipe_config;
+		dev_priv->display.get_plane_config = ironlake_get_plane_config;
 		dev_priv->display.crtc_mode_set = haswell_crtc_mode_set;
 		dev_priv->display.crtc_enable = haswell_crtc_enable;
 		dev_priv->display.crtc_disable = haswell_crtc_disable;
 		dev_priv->display.off = haswell_crtc_off;
-		dev_priv->display.update_plane = ironlake_update_plane;
+		dev_priv->display.update_primary_plane =
+			ironlake_update_primary_plane;
 	} else if (HAS_PCH_SPLIT(dev)) {
 		dev_priv->display.get_pipe_config = ironlake_get_pipe_config;
+		dev_priv->display.get_plane_config = ironlake_get_plane_config;
 		dev_priv->display.crtc_mode_set = ironlake_crtc_mode_set;
 		dev_priv->display.crtc_enable = ironlake_crtc_enable;
 		dev_priv->display.crtc_disable = ironlake_crtc_disable;
 		dev_priv->display.off = ironlake_crtc_off;
-		dev_priv->display.update_plane = ironlake_update_plane;
+		dev_priv->display.update_primary_plane =
+			ironlake_update_primary_plane;
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->display.get_pipe_config = i9xx_get_pipe_config;
+		dev_priv->display.get_plane_config = i9xx_get_plane_config;
 		dev_priv->display.crtc_mode_set = i9xx_crtc_mode_set;
 		dev_priv->display.crtc_enable = valleyview_crtc_enable;
 		dev_priv->display.crtc_disable = i9xx_crtc_disable;
 		dev_priv->display.off = i9xx_crtc_off;
-		dev_priv->display.update_plane = i9xx_update_plane;
+		dev_priv->display.update_primary_plane =
+			i9xx_update_primary_plane;
 	} else {
 		dev_priv->display.get_pipe_config = i9xx_get_pipe_config;
+		dev_priv->display.get_plane_config = i9xx_get_plane_config;
 		dev_priv->display.crtc_mode_set = i9xx_crtc_mode_set;
 		dev_priv->display.crtc_enable = i9xx_crtc_enable;
 		dev_priv->display.crtc_disable = i9xx_crtc_disable;
 		dev_priv->display.off = i9xx_crtc_off;
-		dev_priv->display.update_plane = i9xx_update_plane;
+		dev_priv->display.update_primary_plane =
+			i9xx_update_primary_plane;
 	}
 
 	/* Returns the core display clock speed */
@@ -10398,6 +11506,8 @@ static void intel_init_display(struct drm_device *dev)
 		} else if (IS_GEN6(dev)) {
 			dev_priv->display.fdi_link_train = gen6_fdi_link_train;
 			dev_priv->display.write_eld = ironlake_write_eld;
+			dev_priv->display.modeset_global_resources =
+				snb_modeset_global_resources;
 		} else if (IS_IVYBRIDGE(dev)) {
 			/* FIXME: detect B0+ stepping and use auto training */
 			dev_priv->display.fdi_link_train = ivb_manual_fdi_link_train;
@@ -10412,8 +11522,11 @@ static void intel_init_display(struct drm_device *dev)
 		}
 	} else if (IS_G4X(dev)) {
 		dev_priv->display.write_eld = g4x_write_eld;
-	} else if (IS_VALLEYVIEW(dev))
+	} else if (IS_VALLEYVIEW(dev)) {
+		dev_priv->display.modeset_global_resources =
+			valleyview_modeset_global_resources;
 		dev_priv->display.write_eld = ironlake_write_eld;
+	}
 
 	/* Default just returns -ENODEV to indicate unsupported */
 	dev_priv->display.queue_flip = intel_default_queue_flip;
@@ -10440,6 +11553,8 @@ static void intel_init_display(struct drm_device *dev)
 		dev_priv->display.queue_flip = intel_gen7_queue_flip;
 		break;
 	}
+
+	intel_panel_init_backlight_funcs(dev);
 }
 
 /*
@@ -10476,15 +11591,12 @@ static void quirk_invert_brightness(struct drm_device *dev)
 	DRM_INFO("applying inverted panel brightness quirk\n");
 }
 
-/*
- * Some machines (Dell XPS13) suffer broken backlight controls if
- * BLM_PCH_PWM_ENABLE is set.
- */
-static void quirk_no_pcm_pwm_enable(struct drm_device *dev)
+/* Some VBT's incorrectly indicate no backlight is present */
+static void quirk_backlight_present(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	dev_priv->quirks |= QUIRK_NO_PCH_PWM_ENABLE;
-	DRM_INFO("applying no-PCH_PWM_ENABLE quirk\n");
+	dev_priv->quirks |= QUIRK_BACKLIGHT_PRESENT;
+	DRM_INFO("applying backlight present quirk\n");
 }
 
 struct intel_quirk {
@@ -10532,9 +11644,6 @@ static struct intel_quirk intel_quirks[] = {
 	/* ThinkPad T60 needs pipe A force quirk (bug #16494) */
 	{ 0x2782, 0x17aa, 0x201a, quirk_pipea_force },
 
-	/* 830 needs to leave pipe A & dpll A up */
-	{ 0x3577, PCI_ANY_ID, PCI_ANY_ID, quirk_pipea_force },
-
 	/* Lenovo U160 cannot use SSC on LVDS */
 	{ 0x0046, 0x17aa, 0x3920, quirk_ssc_force_disable },
 
@@ -10556,10 +11665,17 @@ static struct intel_quirk intel_quirks[] = {
 	/* Acer Aspire 4736Z */
 	{ 0x2a42, 0x1025, 0x0260, quirk_invert_brightness },
 
-	/* Dell XPS13 HD Sandy Bridge */
-	{ 0x0116, 0x1028, 0x052e, quirk_no_pcm_pwm_enable },
-	/* Dell XPS13 HD and XPS13 FHD Ivy Bridge */
-	{ 0x0166, 0x1028, 0x058b, quirk_no_pcm_pwm_enable },
+	/* Acer Aspire 5336 */
+	{ 0x2a42, 0x1025, 0x048a, quirk_invert_brightness },
+
+	/* Acer C720 and C720P Chromebooks (Celeron 2955U) have backlights */
+	{ 0x0a06, 0x1025, 0x0a11, quirk_backlight_present },
+
+	/* Toshiba CB35 Chromebook (Celeron 2955U) */
+	{ 0x0a06, 0x1179, 0x0a88, quirk_backlight_present },
+
+	/* HP Chromebook 14 (Celeron 2955U) */
+	{ 0x0a06, 0x103c, 0x21ed, quirk_backlight_present },
 };
 
 static void intel_init_quirks(struct drm_device *dev)
@@ -10590,6 +11706,7 @@ static void i915_disable_vga(struct drm_device *dev)
 	u8 sr1;
 	u32 vga_reg = i915_vgacntrl_reg(dev);
 
+	/* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
 	vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO);
 	outb(SR01, VGA_SR_INDEX);
 	sr1 = inb(VGA_SR_DATA);
@@ -10603,22 +11720,13 @@ static void i915_disable_vga(struct drm_device *dev)
 
 void intel_modeset_init_hw(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
 	intel_prepare_ddi(dev);
 
 	intel_init_clock_gating(dev);
 
-	/* Enable the CRI clock source so we can get at the display */
-	if (IS_VALLEYVIEW(dev))
-		I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
-			   DPLL_INTEGRATED_CRI_CLK_VLV);
-
-	intel_init_dpio(dev);
+	intel_reset_dpio(dev);
 
-	mutex_lock(&dev->struct_mutex);
 	intel_enable_gt_powersave(dev);
-	mutex_unlock(&dev->struct_mutex);
 }
 
 void intel_modeset_suspend_hw(struct drm_device *dev)
@@ -10629,7 +11737,9 @@ void intel_modeset_suspend_hw(struct drm_device *dev)
 void intel_modeset_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int i, j, ret;
+	int sprite, ret;
+	enum pipe pipe;
+	struct intel_crtc *crtc;
 
 	drm_mode_config_init(dev);
 
@@ -10660,22 +11770,34 @@ void intel_modeset_init(struct drm_device *dev)
 		dev->mode_config.max_width = 8192;
 		dev->mode_config.max_height = 8192;
 	}
+
+	if (IS_GEN2(dev)) {
+		dev->mode_config.cursor_width = GEN2_CURSOR_WIDTH;
+		dev->mode_config.cursor_height = GEN2_CURSOR_HEIGHT;
+	} else {
+		dev->mode_config.cursor_width = MAX_CURSOR_WIDTH;
+		dev->mode_config.cursor_height = MAX_CURSOR_HEIGHT;
+	}
+
 	dev->mode_config.fb_base = dev_priv->gtt.mappable_base;
 
 	DRM_DEBUG_KMS("%d display pipe%s available.\n",
 		      INTEL_INFO(dev)->num_pipes,
 		      INTEL_INFO(dev)->num_pipes > 1 ? "s" : "");
 
-	for_each_pipe(i) {
-		intel_crtc_init(dev, i);
-		for (j = 0; j < dev_priv->num_plane; j++) {
-			ret = intel_plane_init(dev, i, j);
+	for_each_pipe(pipe) {
+		intel_crtc_init(dev, pipe);
+		for_each_sprite(pipe, sprite) {
+			ret = intel_plane_init(dev, pipe, sprite);
 			if (ret)
 				DRM_DEBUG_KMS("pipe %c sprite %c init failed: %d\n",
-					      pipe_name(i), sprite_name(i, j), ret);
+					      pipe_name(pipe), sprite_name(pipe, sprite), ret);
 		}
 	}
 
+	intel_init_dpio(dev);
+	intel_reset_dpio(dev);
+
 	intel_cpu_pll_init(dev);
 	intel_shared_dpll_init(dev);
 
@@ -10685,15 +11807,32 @@ void intel_modeset_init(struct drm_device *dev)
 
 	/* Just in case the BIOS is doing something questionable. */
 	intel_disable_fbc(dev);
-}
 
-static void
-intel_connector_break_all_links(struct intel_connector *connector)
-{
-	connector->base.dpms = DRM_MODE_DPMS_OFF;
-	connector->base.encoder = NULL;
-	connector->encoder->connectors_active = false;
-	connector->encoder->base.crtc = NULL;
+	drm_modeset_lock_all(dev);
+	intel_modeset_setup_hw_state(dev, false);
+	drm_modeset_unlock_all(dev);
+
+	for_each_intel_crtc(dev, crtc) {
+		if (!crtc->active)
+			continue;
+
+		/*
+		 * Note that reserving the BIOS fb up front prevents us
+		 * from stuffing other stolen allocations like the ring
+		 * on top.  This prevents some ugliness at boot time, and
+		 * can even allow for smooth boot transitions if the BIOS
+		 * fb is large enough for the active pipe configuration.
+		 */
+		if (dev_priv->display.get_plane_config) {
+			dev_priv->display.get_plane_config(crtc,
+							   &crtc->plane_config);
+			/*
+			 * If the fb is shared between multiple heads, we'll
+			 * just get the first one.
+			 */
+			intel_find_plane_obj(crtc, &crtc->plane_config);
+		}
+	}
 }
 
 static void intel_enable_pipe_a(struct drm_device *dev)
@@ -10701,6 +11840,7 @@ static void intel_enable_pipe_a(struct drm_device *dev)
 	struct intel_connector *connector;
 	struct drm_connector *crt = NULL;
 	struct intel_load_detect_pipe load_detect_temp;
+	struct drm_modeset_acquire_ctx ctx;
 
 	/* We can't just switch on the pipe A, we need to set things up with a
 	 * proper mode and output configuration. As a gross hack, enable pipe A
@@ -10717,8 +11857,8 @@ static void intel_enable_pipe_a(struct drm_device *dev)
 	if (!crt)
 		return;
 
-	if (intel_get_load_detect_pipe(crt, NULL, &load_detect_temp))
-		intel_release_load_detect_pipe(crt, &load_detect_temp);
+	if (intel_get_load_detect_pipe(crt, NULL, &load_detect_temp, &ctx))
+		intel_release_load_detect_pipe(crt, &load_detect_temp, &ctx);
 
 
 }
@@ -10753,6 +11893,12 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 	reg = PIPECONF(crtc->config.cpu_transcoder);
 	I915_WRITE(reg, I915_READ(reg) & ~PIPECONF_FRAME_START_DELAY_MASK);
 
+	/* restore vblank interrupts to correct state */
+	if (crtc->active)
+		drm_vblank_on(dev, crtc->pipe);
+	else
+		drm_vblank_off(dev, crtc->pipe);
+
 	/* We need to sanitize the plane -> pipe mapping first because this will
 	 * disable the crtc (and hence change the state) if it is wrong. Note
 	 * that gen4+ has a fixed plane -> pipe mapping.  */
@@ -10768,6 +11914,7 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 		 * ...  */
 		plane = crtc->plane;
 		crtc->plane = !plane;
+		crtc->primary_enabled = true;
 		dev_priv->display.crtc_disable(&crtc->base);
 		crtc->plane = plane;
 
@@ -10777,8 +11924,17 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 			if (connector->encoder->base.crtc != &crtc->base)
 				continue;
 
-			intel_connector_break_all_links(connector);
+			connector->base.dpms = DRM_MODE_DPMS_OFF;
+			connector->base.encoder = NULL;
 		}
+		/* multiple connectors may have the same encoder:
+		 *  handle them and break crtc link separately */
+		list_for_each_entry(connector, &dev->mode_config.connector_list,
+				    base.head)
+			if (connector->encoder->base.crtc == &crtc->base) {
+				connector->encoder->base.crtc = NULL;
+				connector->encoder->connectors_active = false;
+			}
 
 		WARN_ON(crtc->active);
 		crtc->base.enabled = false;
@@ -10822,6 +11978,26 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 			encoder->base.crtc = NULL;
 		}
 	}
+
+	if (crtc->active || IS_VALLEYVIEW(dev) || INTEL_INFO(dev)->gen < 5) {
+		/*
+		 * We start out with underrun reporting disabled to avoid races.
+		 * For correct bookkeeping mark this on active crtcs.
+		 *
+		 * Also on gmch platforms we dont have any hardware bits to
+		 * disable the underrun reporting. Which means we need to start
+		 * out with underrun reporting disabled also on inactive pipes,
+		 * since otherwise we'll complain about the garbage we read when
+		 * e.g. coming up after runtime pm.
+		 *
+		 * No protection against concurrent access is required - at
+		 * worst a fifo underrun happens which also sets this to false.
+		 */
+		crtc->cpu_fifo_underrun_disabled = true;
+		crtc->pch_fifo_underrun_disabled = true;
+
+		update_scanline_offset(crtc);
+	}
 }
 
 static void intel_sanitize_encoder(struct intel_encoder *encoder)
@@ -10838,7 +12014,7 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder)
 	if (encoder->connectors_active && !has_active_crtc) {
 		DRM_DEBUG_KMS("[ENCODER:%d:%s] has active connectors but no active pipe!\n",
 			      encoder->base.base.id,
-			      drm_get_encoder_name(&encoder->base));
+			      encoder->base.name);
 
 		/* Connector is active, but has no active pipe. This is
 		 * fallout from our resume register restoring. Disable
@@ -10846,9 +12022,11 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder)
 		if (encoder->base.crtc) {
 			DRM_DEBUG_KMS("[ENCODER:%d:%s] manually disabled\n",
 				      encoder->base.base.id,
-				      drm_get_encoder_name(&encoder->base));
+				      encoder->base.name);
 			encoder->disable(encoder);
 		}
+		encoder->base.crtc = NULL;
+		encoder->connectors_active = false;
 
 		/* Inconsistent output/port/pipe state happens presumably due to
 		 * a bug in one of the get_hw_state functions. Or someplace else
@@ -10859,19 +12037,29 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder)
 				    base.head) {
 			if (connector->encoder != encoder)
 				continue;
-
-			intel_connector_break_all_links(connector);
+			connector->base.dpms = DRM_MODE_DPMS_OFF;
+			connector->base.encoder = NULL;
 		}
 	}
 	/* Enabled encoders without active connectors will be fixed in
 	 * the crtc fixup. */
 }
 
-void i915_redisable_vga(struct drm_device *dev)
+void i915_redisable_vga_power_on(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 vga_reg = i915_vgacntrl_reg(dev);
 
+	if (!(I915_READ(vga_reg) & VGA_DISP_DISABLE)) {
+		DRM_DEBUG_KMS("Something enabled VGA plane, disabling it\n");
+		i915_disable_vga(dev);
+	}
+}
+
+void i915_redisable_vga(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
 	/* This function can be called both from intel_modeset_setup_hw_state or
 	 * at a very early point in our resume sequence, where the power well
 	 * structures are not yet restored. Since this function is at a very
@@ -10879,14 +12067,20 @@ void i915_redisable_vga(struct drm_device *dev)
 	 * level, just check if the power well is enabled instead of trying to
 	 * follow the "don't touch the power well if we don't need it" policy
 	 * the rest of the driver uses. */
-	if (HAS_POWER_WELL(dev) &&
-	    (I915_READ(HSW_PWR_WELL_DRIVER) & HSW_PWR_WELL_STATE_ENABLED) == 0)
+	if (!intel_display_power_enabled(dev_priv, POWER_DOMAIN_VGA))
 		return;
 
-	if (!(I915_READ(vga_reg) & VGA_DISP_DISABLE)) {
-		DRM_DEBUG_KMS("Something enabled VGA plane, disabling it\n");
-		i915_disable_vga(dev);
-	}
+	i915_redisable_vga_power_on(dev);
+}
+
+static bool primary_get_hw_state(struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
+
+	if (!crtc->active)
+		return false;
+
+	return I915_READ(DSPCNTR(crtc->plane)) & DISPLAY_PLANE_ENABLE;
 }
 
 static void intel_modeset_readout_hw_state(struct drm_device *dev)
@@ -10898,15 +12092,16 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 	struct intel_connector *connector;
 	int i;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list,
-			    base.head) {
+	for_each_intel_crtc(dev, crtc) {
 		memset(&crtc->config, 0, sizeof(crtc->config));
 
+		crtc->config.quirks |= PIPE_CONFIG_QUIRK_INHERITED_MODE;
+
 		crtc->active = dev_priv->display.get_pipe_config(crtc,
 								 &crtc->config);
 
 		crtc->base.enabled = crtc->active;
-		crtc->primary_enabled = crtc->active;
+		crtc->primary_enabled = primary_get_hw_state(crtc);
 
 		DRM_DEBUG_KMS("[CRTC:%d] hw state readout: %s\n",
 			      crtc->base.base.id,
@@ -10922,8 +12117,7 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 
 		pll->on = pll->get_hw_state(dev_priv, pll, &pll->hw_state);
 		pll->active = 0;
-		list_for_each_entry(crtc, &dev->mode_config.crtc_list,
-				    base.head) {
+		for_each_intel_crtc(dev, crtc) {
 			if (crtc->active && intel_crtc_to_shared_dpll(crtc) == pll)
 				pll->active++;
 		}
@@ -10948,7 +12142,7 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 		encoder->connectors_active = false;
 		DRM_DEBUG_KMS("[ENCODER:%d:%s] hw state readout: %s, pipe %c\n",
 			      encoder->base.base.id,
-			      drm_get_encoder_name(&encoder->base),
+			      encoder->base.name,
 			      encoder->base.crtc ? "enabled" : "disabled",
 			      pipe_name(pipe));
 	}
@@ -10965,7 +12159,7 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 		}
 		DRM_DEBUG_KMS("[CONNECTOR:%d:%s] hw state readout: %s\n",
 			      connector->base.base.id,
-			      drm_get_connector_name(&connector->base),
+			      connector->base.name,
 			      connector->base.encoder ? "enabled" : "disabled");
 	}
 }
@@ -10988,11 +12182,9 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 	 * Note that this could go away if we move to using crtc_config
 	 * checking everywhere.
 	 */
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list,
-			    base.head) {
-		if (crtc->active && i915_fastboot) {
-			intel_crtc_mode_from_pipe_config(crtc, &crtc->config);
-
+	for_each_intel_crtc(dev, crtc) {
+		if (crtc->active && i915.fastboot) {
+			intel_mode_from_pipe_config(&crtc->base.mode, &crtc->config);
 			DRM_DEBUG_KMS("[CRTC:%d] found active mode: ",
 				      crtc->base.base.id);
 			drm_mode_debug_printmodeline(&crtc->base.mode);
@@ -11023,7 +12215,7 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 		pll->on = false;
 	}
 
-	if (IS_HASWELL(dev))
+	if (HAS_PCH_SPLIT(dev))
 		ilk_wm_get_hw_state(dev);
 
 	if (force_restore) {
@@ -11038,7 +12230,7 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 				dev_priv->pipe_to_crtc_mapping[pipe];
 
 			__intel_set_mode(crtc, &crtc->mode, crtc->x, crtc->y,
-					 crtc->fb);
+					 crtc->primary->fb);
 		}
 	} else {
 		intel_modeset_update_staged_output_state(dev);
@@ -11049,14 +12241,44 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 
 void intel_modeset_gem_init(struct drm_device *dev)
 {
+	struct drm_crtc *c;
+	struct intel_framebuffer *fb;
+
+	mutex_lock(&dev->struct_mutex);
+	intel_init_gt_powersave(dev);
+	mutex_unlock(&dev->struct_mutex);
+
 	intel_modeset_init_hw(dev);
 
 	intel_setup_overlay(dev);
 
-	mutex_lock(&dev->mode_config.mutex);
-	drm_mode_config_reset(dev);
-	intel_modeset_setup_hw_state(dev, false);
-	mutex_unlock(&dev->mode_config.mutex);
+	/*
+	 * Make sure any fbs we allocated at startup are properly
+	 * pinned & fenced.  When we do the allocation it's too early
+	 * for this.
+	 */
+	mutex_lock(&dev->struct_mutex);
+	for_each_crtc(dev, c) {
+		if (!c->primary->fb)
+			continue;
+
+		fb = to_intel_framebuffer(c->primary->fb);
+		if (intel_pin_and_fence_fb_obj(dev, fb->obj, NULL)) {
+			DRM_ERROR("failed to pin boot fb on pipe %d\n",
+				  to_intel_crtc(c)->pipe);
+			drm_framebuffer_unreference(c->primary->fb);
+			c->primary->fb = NULL;
+		}
+	}
+	mutex_unlock(&dev->struct_mutex);
+}
+
+void intel_connector_unregister(struct intel_connector *intel_connector)
+{
+	struct drm_connector *connector = &intel_connector->base;
+
+	intel_panel_destroy_backlight(connector);
+	drm_sysfs_connector_remove(connector);
 }
 
 void intel_modeset_cleanup(struct drm_device *dev)
@@ -11082,9 +12304,9 @@ void intel_modeset_cleanup(struct drm_device *dev)
 
 	intel_unregister_dsm_handler();
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+	for_each_crtc(dev, crtc) {
 		/* Skip inactive CRTCs */
-		if (!crtc->fb)
+		if (!crtc->primary->fb)
 			continue;
 
 		intel_increase_pllclock(crtc);
@@ -11101,16 +12323,21 @@ void intel_modeset_cleanup(struct drm_device *dev)
 	/* flush any delayed tasks or pending work */
 	flush_scheduled_work();
 
-	/* destroy backlight, if any, before the connectors */
-	intel_panel_destroy_backlight(dev);
+	/* destroy the backlight and sysfs files before encoders/connectors */
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		struct intel_connector *intel_connector;
 
-	/* destroy the sysfs files before encoders/connectors */
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
-		drm_sysfs_connector_remove(connector);
+		intel_connector = to_intel_connector(connector);
+		intel_connector->unregister(intel_connector);
+	}
 
 	drm_mode_config_cleanup(dev);
 
 	intel_cleanup_overlay(dev);
+
+	mutex_lock(&dev->struct_mutex);
+	intel_cleanup_gt_powersave(dev);
+	mutex_unlock(&dev->struct_mutex);
 }
 
 /*
@@ -11138,12 +12365,24 @@ int intel_modeset_vga_set_state(struct drm_device *dev, bool state)
 	unsigned reg = INTEL_INFO(dev)->gen >= 6 ? SNB_GMCH_CTRL : INTEL_GMCH_CTRL;
 	u16 gmch_ctrl;
 
-	pci_read_config_word(dev_priv->bridge_dev, reg, &gmch_ctrl);
+	if (pci_read_config_word(dev_priv->bridge_dev, reg, &gmch_ctrl)) {
+		DRM_ERROR("failed to read control word\n");
+		return -EIO;
+	}
+
+	if (!!(gmch_ctrl & INTEL_GMCH_VGA_DISABLE) == !state)
+		return 0;
+
 	if (state)
 		gmch_ctrl &= ~INTEL_GMCH_VGA_DISABLE;
 	else
 		gmch_ctrl |= INTEL_GMCH_VGA_DISABLE;
-	pci_write_config_word(dev_priv->bridge_dev, reg, gmch_ctrl);
+
+	if (pci_write_config_word(dev_priv->bridge_dev, reg, gmch_ctrl)) {
+		DRM_ERROR("failed to write control word\n");
+		return -EIO;
+	}
+
 	return 0;
 }
 
@@ -11161,7 +12400,9 @@ struct intel_display_error_state {
 	} cursor[I915_MAX_PIPES];
 
 	struct intel_pipe_error_state {
+		bool power_domain_on;
 		u32 source;
+		u32 stat;
 	} pipe[I915_MAX_PIPES];
 
 	struct intel_plane_error_state {
@@ -11175,6 +12416,7 @@ struct intel_display_error_state {
 	} plane[I915_MAX_PIPES];
 
 	struct intel_transcoder_error_state {
+		bool power_domain_on;
 		enum transcoder cpu_transcoder;
 
 		u32 conf;
@@ -11191,7 +12433,7 @@ struct intel_display_error_state {
 struct intel_display_error_state *
 intel_display_capture_error_state(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_display_error_state *error;
 	int transcoders[] = {
 		TRANSCODER_A,
@@ -11208,22 +12450,19 @@ intel_display_capture_error_state(struct drm_device *dev)
 	if (error == NULL)
 		return NULL;
 
-	if (HAS_POWER_WELL(dev))
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		error->power_well_driver = I915_READ(HSW_PWR_WELL_DRIVER);
 
 	for_each_pipe(i) {
-		if (!intel_display_power_enabled(dev, POWER_DOMAIN_PIPE(i)))
+		error->pipe[i].power_domain_on =
+			intel_display_power_enabled_unlocked(dev_priv,
+							   POWER_DOMAIN_PIPE(i));
+		if (!error->pipe[i].power_domain_on)
 			continue;
 
-		if (INTEL_INFO(dev)->gen <= 6 || IS_VALLEYVIEW(dev)) {
-			error->cursor[i].control = I915_READ(CURCNTR(i));
-			error->cursor[i].position = I915_READ(CURPOS(i));
-			error->cursor[i].base = I915_READ(CURBASE(i));
-		} else {
-			error->cursor[i].control = I915_READ(CURCNTR_IVB(i));
-			error->cursor[i].position = I915_READ(CURPOS_IVB(i));
-			error->cursor[i].base = I915_READ(CURBASE_IVB(i));
-		}
+		error->cursor[i].control = I915_READ(CURCNTR(i));
+		error->cursor[i].position = I915_READ(CURPOS(i));
+		error->cursor[i].base = I915_READ(CURBASE(i));
 
 		error->plane[i].control = I915_READ(DSPCNTR(i));
 		error->plane[i].stride = I915_READ(DSPSTRIDE(i));
@@ -11239,6 +12478,9 @@ intel_display_capture_error_state(struct drm_device *dev)
 		}
 
 		error->pipe[i].source = I915_READ(PIPESRC(i));
+
+		if (!HAS_PCH_SPLIT(dev))
+			error->pipe[i].stat = I915_READ(PIPESTAT(i));
 	}
 
 	error->num_transcoders = INTEL_INFO(dev)->num_pipes;
@@ -11248,8 +12490,10 @@ intel_display_capture_error_state(struct drm_device *dev)
 	for (i = 0; i < error->num_transcoders; i++) {
 		enum transcoder cpu_transcoder = transcoders[i];
 
-		if (!intel_display_power_enabled(dev,
-				POWER_DOMAIN_TRANSCODER(cpu_transcoder)))
+		error->transcoder[i].power_domain_on =
+			intel_display_power_enabled_unlocked(dev_priv,
+				POWER_DOMAIN_TRANSCODER(cpu_transcoder));
+		if (!error->transcoder[i].power_domain_on)
 			continue;
 
 		error->transcoder[i].cpu_transcoder = cpu_transcoder;
@@ -11279,12 +12523,15 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 		return;
 
 	err_printf(m, "Num Pipes: %d\n", INTEL_INFO(dev)->num_pipes);
-	if (HAS_POWER_WELL(dev))
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		err_printf(m, "PWR_WELL_CTL2: %08x\n",
 			   error->power_well_driver);
 	for_each_pipe(i) {
 		err_printf(m, "Pipe [%d]:\n", i);
+		err_printf(m, "  Power: %s\n",
+			   error->pipe[i].power_domain_on ? "on" : "off");
 		err_printf(m, "  SRC: %08x\n", error->pipe[i].source);
+		err_printf(m, "  STAT: %08x\n", error->pipe[i].stat);
 
 		err_printf(m, "Plane [%d]:\n", i);
 		err_printf(m, "  CNTR: %08x\n", error->plane[i].control);
@@ -11309,6 +12556,8 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 	for (i = 0; i < error->num_transcoders; i++) {
 		err_printf(m, "CPU transcoder: %c\n",
 			   transcoder_name(error->transcoder[i].cpu_transcoder));
+		err_printf(m, "  Power: %s\n",
+			   error->transcoder[i].power_domain_on ? "on" : "off");
 		err_printf(m, "  CONF: %08x\n", error->transcoder[i].conf);
 		err_printf(m, "  HTOTAL: %08x\n", error->transcoder[i].htotal);
 		err_printf(m, "  HBLANK: %08x\n", error->transcoder[i].hblank);
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 30c627c7b7b..8a1a4fbc06a 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -28,6 +28,8 @@
 #include <linux/i2c.h>
 #include <linux/slab.h>
 #include <linux/export.h>
+#include <linux/notifier.h>
+#include <linux/reboot.h>
 #include <drm/drmP.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_crtc_helper.h>
@@ -64,6 +66,24 @@ static const struct dp_link_dpll vlv_dpll[] = {
 		{ .p1 = 2, .p2 = 2, .n = 1, .m1 = 2, .m2 = 27 } }
 };
 
+/*
+ * CHV supports eDP 1.4 that have  more link rates.
+ * Below only provides the fixed rate but exclude variable rate.
+ */
+static const struct dp_link_dpll chv_dpll[] = {
+	/*
+	 * CHV requires to program fractional division for m2.
+	 * m2 is stored in fixed point format using formula below
+	 * (m2_int << 22) | m2_fraction
+	 */
+	{ DP_LINK_BW_1_62,	/* m2_int = 32, m2_fraction = 1677722 */
+		{ .p1 = 4, .p2 = 2, .n = 1, .m1 = 2, .m2 = 0x819999a } },
+	{ DP_LINK_BW_2_7,	/* m2_int = 27, m2_fraction = 0 */
+		{ .p1 = 4, .p2 = 1, .n = 1, .m1 = 2, .m2 = 0x6c00000 } },
+	{ DP_LINK_BW_5_4,	/* m2_int = 27, m2_fraction = 0 */
+		{ .p1 = 2, .p2 = 1, .n = 1, .m1 = 2, .m2 = 0x6c00000 } }
+};
+
 /**
  * is_edp - is the given port attached to an eDP panel (either CPU or PCH)
  * @intel_dp: DP struct
@@ -91,18 +111,26 @@ static struct intel_dp *intel_attached_dp(struct drm_connector *connector)
 }
 
 static void intel_dp_link_down(struct intel_dp *intel_dp);
+static bool _edp_panel_vdd_on(struct intel_dp *intel_dp);
+static void edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync);
 
 static int
 intel_dp_max_link_bw(struct intel_dp *intel_dp)
 {
 	int max_link_bw = intel_dp->dpcd[DP_MAX_LINK_RATE];
+	struct drm_device *dev = intel_dp->attached_connector->base.dev;
 
 	switch (max_link_bw) {
 	case DP_LINK_BW_1_62:
 	case DP_LINK_BW_2_7:
 		break;
 	case DP_LINK_BW_5_4: /* 1.2 capable displays may advertise higher bw */
-		max_link_bw = DP_LINK_BW_2_7;
+		if (((IS_HASWELL(dev) && !IS_HSW_ULX(dev)) ||
+		     INTEL_INFO(dev)->gen >= 8) &&
+		    intel_dp->dpcd[DP_DPCD_REV] >= 0x12)
+			max_link_bw = DP_LINK_BW_5_4;
+		else
+			max_link_bw = DP_LINK_BW_2_7;
 		break;
 	default:
 		WARN(1, "invalid max DP link bw val %x, using 1.62Gbps\n",
@@ -113,6 +141,22 @@ intel_dp_max_link_bw(struct intel_dp *intel_dp)
 	return max_link_bw;
 }
 
+static u8 intel_dp_max_lane_count(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	u8 source_max, sink_max;
+
+	source_max = 4;
+	if (HAS_DDI(dev) && intel_dig_port->port == PORT_A &&
+	    (intel_dig_port->saved_port_bits & DDI_A_4_LANES) == 0)
+		source_max = 2;
+
+	sink_max = drm_dp_max_lane_count(intel_dp->dpcd);
+
+	return min(source_max, sink_max);
+}
+
 /*
  * The units on the numbers in the next two are... bizarre.  Examples will
  * make it clearer; this one parallels an example in the eDP spec.
@@ -142,7 +186,7 @@ intel_dp_max_data_rate(int max_link_clock, int max_lanes)
 	return (max_link_clock * max_lanes * 8) / 10;
 }
 
-static int
+static enum drm_mode_status
 intel_dp_mode_valid(struct drm_connector *connector,
 		    struct drm_display_mode *mode)
 {
@@ -163,7 +207,7 @@ intel_dp_mode_valid(struct drm_connector *connector,
 	}
 
 	max_link_clock = drm_dp_bw_code_to_link_rate(intel_dp_max_link_bw(intel_dp));
-	max_lanes = drm_dp_max_lane_count(intel_dp->dpcd);
+	max_lanes = intel_dp_max_lane_count(intel_dp);
 
 	max_rate = intel_dp_max_data_rate(max_link_clock, max_lanes);
 	mode_rate = intel_dp_link_required(target_clock, 18);
@@ -294,7 +338,38 @@ static u32 _pp_stat_reg(struct intel_dp *intel_dp)
 		return VLV_PIPE_PP_STATUS(vlv_power_sequencer_pipe(intel_dp));
 }
 
-static bool ironlake_edp_have_panel_power(struct intel_dp *intel_dp)
+/* Reboot notifier handler to shutdown panel power to guarantee T12 timing
+   This function only applicable when panel PM state is not to be tracked */
+static int edp_notify_handler(struct notifier_block *this, unsigned long code,
+			      void *unused)
+{
+	struct intel_dp *intel_dp = container_of(this, typeof(* intel_dp),
+						 edp_notifier);
+	struct drm_device *dev = intel_dp_to_dev(intel_dp);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 pp_div;
+	u32 pp_ctrl_reg, pp_div_reg;
+	enum pipe pipe = vlv_power_sequencer_pipe(intel_dp);
+
+	if (!is_edp(intel_dp) || code != SYS_RESTART)
+		return 0;
+
+	if (IS_VALLEYVIEW(dev)) {
+		pp_ctrl_reg = VLV_PIPE_PP_CONTROL(pipe);
+		pp_div_reg  = VLV_PIPE_PP_DIVISOR(pipe);
+		pp_div = I915_READ(pp_div_reg);
+		pp_div &= PP_REFERENCE_DIVIDER_MASK;
+
+		/* 0x1F write to PP_DIV_REG sets max cycle delay */
+		I915_WRITE(pp_div_reg, pp_div | 0x1F);
+		I915_WRITE(pp_ctrl_reg, PANEL_UNLOCK_REGS | PANEL_POWER_OFF);
+		msleep(intel_dp->panel_power_cycle_delay);
+	}
+
+	return 0;
+}
+
+static bool edp_have_panel_power(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -302,12 +377,17 @@ static bool ironlake_edp_have_panel_power(struct intel_dp *intel_dp)
 	return (I915_READ(_pp_stat_reg(intel_dp)) & PP_ON) != 0;
 }
 
-static bool ironlake_edp_have_panel_vdd(struct intel_dp *intel_dp)
+static bool edp_have_panel_vdd(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *intel_encoder = &intel_dig_port->base;
+	enum intel_display_power_domain power_domain;
 
-	return (I915_READ(_pp_ctrl_reg(intel_dp)) & EDP_FORCE_VDD) != 0;
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	return intel_display_power_enabled(dev_priv, power_domain) &&
+	       (I915_READ(_pp_ctrl_reg(intel_dp)) & EDP_FORCE_VDD) != 0;
 }
 
 static void
@@ -319,7 +399,7 @@ intel_dp_check_edp(struct intel_dp *intel_dp)
 	if (!is_edp(intel_dp))
 		return;
 
-	if (!ironlake_edp_have_panel_power(intel_dp) && !ironlake_edp_have_panel_vdd(intel_dp)) {
+	if (!edp_have_panel_power(intel_dp) && !edp_have_panel_vdd(intel_dp)) {
 		WARN(1, "eDP powered off while attempting aux channel communication.\n");
 		DRM_DEBUG_KMS("Status 0x%08x Control 0x%08x\n",
 			      I915_READ(_pp_stat_reg(intel_dp)),
@@ -351,31 +431,46 @@ intel_dp_aux_wait_done(struct intel_dp *intel_dp, bool has_aux_irq)
 	return status;
 }
 
-static uint32_t get_aux_clock_divider(struct intel_dp *intel_dp,
-				      int index)
+static uint32_t i9xx_get_aux_clock_divider(struct intel_dp *intel_dp, int index)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct drm_device *dev = intel_dig_port->base.base.dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	/* The clock divider is based off the hrawclk,
-	 * and would like to run at 2MHz. So, take the
-	 * hrawclk value and divide by 2 and use that
-	 *
-	 * Note that PCH attached eDP panels should use a 125MHz input
-	 * clock divider.
+	/*
+	 * The clock divider is based off the hrawclk, and would like to run at
+	 * 2MHz.  So, take the hrawclk value and divide by 2 and use that
 	 */
-	if (IS_VALLEYVIEW(dev)) {
-		return index ? 0 : 100;
-	} else if (intel_dig_port->port == PORT_A) {
-		if (index)
-			return 0;
-		if (HAS_DDI(dev))
-			return DIV_ROUND_CLOSEST(intel_ddi_get_cdclk_freq(dev_priv), 2000);
-		else if (IS_GEN6(dev) || IS_GEN7(dev))
+	return index ? 0 : intel_hrawclk(dev) / 2;
+}
+
+static uint32_t ilk_get_aux_clock_divider(struct intel_dp *intel_dp, int index)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+
+	if (index)
+		return 0;
+
+	if (intel_dig_port->port == PORT_A) {
+		if (IS_GEN6(dev) || IS_GEN7(dev))
 			return 200; /* SNB & IVB eDP input clock at 400Mhz */
 		else
 			return 225; /* eDP input clock at 450Mhz */
+	} else {
+		return DIV_ROUND_UP(intel_pch_rawclk(dev), 2);
+	}
+}
+
+static uint32_t hsw_get_aux_clock_divider(struct intel_dp *intel_dp, int index)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (intel_dig_port->port == PORT_A) {
+		if (index)
+			return 0;
+		return DIV_ROUND_CLOSEST(intel_ddi_get_cdclk_freq(dev_priv), 2000);
 	} else if (dev_priv->pch_id == INTEL_PCH_LPT_DEVICE_ID_TYPE) {
 		/* Workaround for non-ULT HSW */
 		switch (index) {
@@ -383,13 +478,46 @@ static uint32_t get_aux_clock_divider(struct intel_dp *intel_dp,
 		case 1: return 72;
 		default: return 0;
 		}
-	} else if (HAS_PCH_SPLIT(dev)) {
+	} else  {
 		return index ? 0 : DIV_ROUND_UP(intel_pch_rawclk(dev), 2);
-	} else {
-		return index ? 0 :intel_hrawclk(dev) / 2;
 	}
 }
 
+static uint32_t vlv_get_aux_clock_divider(struct intel_dp *intel_dp, int index)
+{
+	return index ? 0 : 100;
+}
+
+static uint32_t i9xx_get_aux_send_ctl(struct intel_dp *intel_dp,
+				      bool has_aux_irq,
+				      int send_bytes,
+				      uint32_t aux_clock_divider)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	uint32_t precharge, timeout;
+
+	if (IS_GEN6(dev))
+		precharge = 3;
+	else
+		precharge = 5;
+
+	if (IS_BROADWELL(dev) && intel_dp->aux_ch_ctl_reg == DPA_AUX_CH_CTL)
+		timeout = DP_AUX_CH_CTL_TIME_OUT_600us;
+	else
+		timeout = DP_AUX_CH_CTL_TIME_OUT_400us;
+
+	return DP_AUX_CH_CTL_SEND_BUSY |
+	       DP_AUX_CH_CTL_DONE |
+	       (has_aux_irq ? DP_AUX_CH_CTL_INTERRUPT : 0) |
+	       DP_AUX_CH_CTL_TIME_OUT_ERROR |
+	       timeout |
+	       DP_AUX_CH_CTL_RECEIVE_ERROR |
+	       (send_bytes << DP_AUX_CH_CTL_MESSAGE_SIZE_SHIFT) |
+	       (precharge << DP_AUX_CH_CTL_PRECHARGE_2US_SHIFT) |
+	       (aux_clock_divider << DP_AUX_CH_CTL_BIT_CLOCK_2X_SHIFT);
+}
+
 static int
 intel_dp_aux_ch(struct intel_dp *intel_dp,
 		uint8_t *send, int send_bytes,
@@ -403,9 +531,11 @@ intel_dp_aux_ch(struct intel_dp *intel_dp,
 	uint32_t aux_clock_divider;
 	int i, ret, recv_bytes;
 	uint32_t status;
-	int try, precharge, clock = 0;
-	bool has_aux_irq = INTEL_INFO(dev)->gen >= 5 && !IS_VALLEYVIEW(dev);
-	uint32_t timeout;
+	int try, clock = 0;
+	bool has_aux_irq = HAS_AUX_IRQ(dev);
+	bool vdd;
+
+	vdd = _edp_panel_vdd_on(intel_dp);
 
 	/* dp aux is extremely sensitive to irq latency, hence request the
 	 * lowest possible wakeup latency and so prevent the cpu from going into
@@ -415,16 +545,6 @@ intel_dp_aux_ch(struct intel_dp *intel_dp,
 
 	intel_dp_check_edp(intel_dp);
 
-	if (IS_GEN6(dev))
-		precharge = 3;
-	else
-		precharge = 5;
-
-	if (IS_BROADWELL(dev) && ch_ctl == DPA_AUX_CH_CTL)
-		timeout = DP_AUX_CH_CTL_TIME_OUT_600us;
-	else
-		timeout = DP_AUX_CH_CTL_TIME_OUT_400us;
-
 	intel_aux_display_runtime_get(dev_priv);
 
 	/* Try to wait for any previous AUX channel activity */
@@ -448,7 +568,12 @@ intel_dp_aux_ch(struct intel_dp *intel_dp,
 		goto out;
 	}
 
-	while ((aux_clock_divider = get_aux_clock_divider(intel_dp, clock++))) {
+	while ((aux_clock_divider = intel_dp->get_aux_clock_divider(intel_dp, clock++))) {
+		u32 send_ctl = intel_dp->get_aux_send_ctl(intel_dp,
+							  has_aux_irq,
+							  send_bytes,
+							  aux_clock_divider);
+
 		/* Must try at least 3 times according to DP spec */
 		for (try = 0; try < 5; try++) {
 			/* Load the send data into the aux channel data registers */
@@ -457,16 +582,7 @@ intel_dp_aux_ch(struct intel_dp *intel_dp,
 					   pack_aux(send + i, send_bytes - i));
 
 			/* Send the command and wait for it to complete */
-			I915_WRITE(ch_ctl,
-				   DP_AUX_CH_CTL_SEND_BUSY |
-				   (has_aux_irq ? DP_AUX_CH_CTL_INTERRUPT : 0) |
-				   timeout |
-				   (send_bytes << DP_AUX_CH_CTL_MESSAGE_SIZE_SHIFT) |
-				   (precharge << DP_AUX_CH_CTL_PRECHARGE_2US_SHIFT) |
-				   (aux_clock_divider << DP_AUX_CH_CTL_BIT_CLOCK_2X_SHIFT) |
-				   DP_AUX_CH_CTL_DONE |
-				   DP_AUX_CH_CTL_TIME_OUT_ERROR |
-				   DP_AUX_CH_CTL_RECEIVE_ERROR);
+			I915_WRITE(ch_ctl, send_ctl);
 
 			status = intel_dp_aux_wait_done(intel_dp, has_aux_irq);
 
@@ -525,238 +641,141 @@ out:
 	pm_qos_update_request(&dev_priv->pm_qos, PM_QOS_DEFAULT_VALUE);
 	intel_aux_display_runtime_put(dev_priv);
 
+	if (vdd)
+		edp_panel_vdd_off(intel_dp, false);
+
 	return ret;
 }
 
-/* Write data to the aux channel in native mode */
-static int
-intel_dp_aux_native_write(struct intel_dp *intel_dp,
-			  uint16_t address, uint8_t *send, int send_bytes)
+#define BARE_ADDRESS_SIZE	3
+#define HEADER_SIZE		(BARE_ADDRESS_SIZE + 1)
+static ssize_t
+intel_dp_aux_transfer(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg)
 {
+	struct intel_dp *intel_dp = container_of(aux, struct intel_dp, aux);
+	uint8_t txbuf[20], rxbuf[20];
+	size_t txsize, rxsize;
 	int ret;
-	uint8_t	msg[20];
-	int msg_bytes;
-	uint8_t	ack;
 
-	if (WARN_ON(send_bytes > 16))
-		return -E2BIG;
+	txbuf[0] = msg->request << 4;
+	txbuf[1] = msg->address >> 8;
+	txbuf[2] = msg->address & 0xff;
+	txbuf[3] = msg->size - 1;
 
-	intel_dp_check_edp(intel_dp);
-	msg[0] = AUX_NATIVE_WRITE << 4;
-	msg[1] = address >> 8;
-	msg[2] = address & 0xff;
-	msg[3] = send_bytes - 1;
-	memcpy(&msg[4], send, send_bytes);
-	msg_bytes = send_bytes + 4;
-	for (;;) {
-		ret = intel_dp_aux_ch(intel_dp, msg, msg_bytes, &ack, 1);
-		if (ret < 0)
-			return ret;
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK)
-			break;
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
-			udelay(100);
-		else
-			return -EIO;
-	}
-	return send_bytes;
-}
+	switch (msg->request & ~DP_AUX_I2C_MOT) {
+	case DP_AUX_NATIVE_WRITE:
+	case DP_AUX_I2C_WRITE:
+		txsize = msg->size ? HEADER_SIZE + msg->size : BARE_ADDRESS_SIZE;
+		rxsize = 1;
 
-/* Write a single byte to the aux channel in native mode */
-static int
-intel_dp_aux_native_write_1(struct intel_dp *intel_dp,
-			    uint16_t address, uint8_t byte)
-{
-	return intel_dp_aux_native_write(intel_dp, address, &byte, 1);
-}
+		if (WARN_ON(txsize > 20))
+			return -E2BIG;
 
-/* read bytes from a native aux channel */
-static int
-intel_dp_aux_native_read(struct intel_dp *intel_dp,
-			 uint16_t address, uint8_t *recv, int recv_bytes)
-{
-	uint8_t msg[4];
-	int msg_bytes;
-	uint8_t reply[20];
-	int reply_bytes;
-	uint8_t ack;
-	int ret;
+		memcpy(txbuf + HEADER_SIZE, msg->buffer, msg->size);
 
-	if (WARN_ON(recv_bytes > 19))
-		return -E2BIG;
+		ret = intel_dp_aux_ch(intel_dp, txbuf, txsize, rxbuf, rxsize);
+		if (ret > 0) {
+			msg->reply = rxbuf[0] >> 4;
 
-	intel_dp_check_edp(intel_dp);
-	msg[0] = AUX_NATIVE_READ << 4;
-	msg[1] = address >> 8;
-	msg[2] = address & 0xff;
-	msg[3] = recv_bytes - 1;
+			/* Return payload size. */
+			ret = msg->size;
+		}
+		break;
 
-	msg_bytes = 4;
-	reply_bytes = recv_bytes + 1;
+	case DP_AUX_NATIVE_READ:
+	case DP_AUX_I2C_READ:
+		txsize = msg->size ? HEADER_SIZE : BARE_ADDRESS_SIZE;
+		rxsize = msg->size + 1;
 
-	for (;;) {
-		ret = intel_dp_aux_ch(intel_dp, msg, msg_bytes,
-				      reply, reply_bytes);
-		if (ret == 0)
-			return -EPROTO;
-		if (ret < 0)
-			return ret;
-		ack = reply[0];
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK) {
-			memcpy(recv, reply + 1, ret - 1);
-			return ret - 1;
+		if (WARN_ON(rxsize > 20))
+			return -E2BIG;
+
+		ret = intel_dp_aux_ch(intel_dp, txbuf, txsize, rxbuf, rxsize);
+		if (ret > 0) {
+			msg->reply = rxbuf[0] >> 4;
+			/*
+			 * Assume happy day, and copy the data. The caller is
+			 * expected to check msg->reply before touching it.
+			 *
+			 * Return payload size.
+			 */
+			ret--;
+			memcpy(msg->buffer, rxbuf + 1, ret);
 		}
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
-			udelay(100);
-		else
-			return -EIO;
+		break;
+
+	default:
+		ret = -EINVAL;
+		break;
 	}
+
+	return ret;
 }
 
-static int
-intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
-		    uint8_t write_byte, uint8_t *read_byte)
-{
-	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
-	struct intel_dp *intel_dp = container_of(adapter,
-						struct intel_dp,
-						adapter);
-	uint16_t address = algo_data->address;
-	uint8_t msg[5];
-	uint8_t reply[2];
-	unsigned retry;
-	int msg_bytes;
-	int reply_bytes;
+static void
+intel_dp_aux_init(struct intel_dp *intel_dp, struct intel_connector *connector)
+{
+	struct drm_device *dev = intel_dp_to_dev(intel_dp);
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	enum port port = intel_dig_port->port;
+	const char *name = NULL;
 	int ret;
 
-	ironlake_edp_panel_vdd_on(intel_dp);
-	intel_dp_check_edp(intel_dp);
-	/* Set up the command byte */
-	if (mode & MODE_I2C_READ)
-		msg[0] = AUX_I2C_READ << 4;
-	else
-		msg[0] = AUX_I2C_WRITE << 4;
-
-	if (!(mode & MODE_I2C_STOP))
-		msg[0] |= AUX_I2C_MOT << 4;
-
-	msg[1] = address >> 8;
-	msg[2] = address;
-
-	switch (mode) {
-	case MODE_I2C_WRITE:
-		msg[3] = 0;
-		msg[4] = write_byte;
-		msg_bytes = 5;
-		reply_bytes = 1;
+	switch (port) {
+	case PORT_A:
+		intel_dp->aux_ch_ctl_reg = DPA_AUX_CH_CTL;
+		name = "DPDDC-A";
 		break;
-	case MODE_I2C_READ:
-		msg[3] = 0;
-		msg_bytes = 4;
-		reply_bytes = 2;
+	case PORT_B:
+		intel_dp->aux_ch_ctl_reg = PCH_DPB_AUX_CH_CTL;
+		name = "DPDDC-B";
 		break;
-	default:
-		msg_bytes = 3;
-		reply_bytes = 1;
+	case PORT_C:
+		intel_dp->aux_ch_ctl_reg = PCH_DPC_AUX_CH_CTL;
+		name = "DPDDC-C";
 		break;
+	case PORT_D:
+		intel_dp->aux_ch_ctl_reg = PCH_DPD_AUX_CH_CTL;
+		name = "DPDDC-D";
+		break;
+	default:
+		BUG();
 	}
 
-	/*
-	 * DP1.2 sections 2.7.7.1.5.6.1 and 2.7.7.1.6.6.1: A DP Source device is
-	 * required to retry at least seven times upon receiving AUX_DEFER
-	 * before giving up the AUX transaction.
-	 */
-	for (retry = 0; retry < 7; retry++) {
-		ret = intel_dp_aux_ch(intel_dp,
-				      msg, msg_bytes,
-				      reply, reply_bytes);
-		if (ret < 0) {
-			DRM_DEBUG_KMS("aux_ch failed %d\n", ret);
-			goto out;
-		}
+	if (!HAS_DDI(dev))
+		intel_dp->aux_ch_ctl_reg = intel_dp->output_reg + 0x10;
 
-		switch (reply[0] & AUX_NATIVE_REPLY_MASK) {
-		case AUX_NATIVE_REPLY_ACK:
-			/* I2C-over-AUX Reply field is only valid
-			 * when paired with AUX ACK.
-			 */
-			break;
-		case AUX_NATIVE_REPLY_NACK:
-			DRM_DEBUG_KMS("aux_ch native nack\n");
-			ret = -EREMOTEIO;
-			goto out;
-		case AUX_NATIVE_REPLY_DEFER:
-			/*
-			 * For now, just give more slack to branch devices. We
-			 * could check the DPCD for I2C bit rate capabilities,
-			 * and if available, adjust the interval. We could also
-			 * be more careful with DP-to-Legacy adapters where a
-			 * long legacy cable may force very low I2C bit rates.
-			 */
-			if (intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] &
-			    DP_DWN_STRM_PORT_PRESENT)
-				usleep_range(500, 600);
-			else
-				usleep_range(300, 400);
-			continue;
-		default:
-			DRM_ERROR("aux_ch invalid native reply 0x%02x\n",
-				  reply[0]);
-			ret = -EREMOTEIO;
-			goto out;
-		}
+	intel_dp->aux.name = name;
+	intel_dp->aux.dev = dev->dev;
+	intel_dp->aux.transfer = intel_dp_aux_transfer;
 
-		switch (reply[0] & AUX_I2C_REPLY_MASK) {
-		case AUX_I2C_REPLY_ACK:
-			if (mode == MODE_I2C_READ) {
-				*read_byte = reply[1];
-			}
-			ret = reply_bytes - 1;
-			goto out;
-		case AUX_I2C_REPLY_NACK:
-			DRM_DEBUG_KMS("aux_i2c nack\n");
-			ret = -EREMOTEIO;
-			goto out;
-		case AUX_I2C_REPLY_DEFER:
-			DRM_DEBUG_KMS("aux_i2c defer\n");
-			udelay(100);
-			break;
-		default:
-			DRM_ERROR("aux_i2c invalid reply 0x%02x\n", reply[0]);
-			ret = -EREMOTEIO;
-			goto out;
-		}
-	}
+	DRM_DEBUG_KMS("registering %s bus for %s\n", name,
+		      connector->base.kdev->kobj.name);
 
-	DRM_ERROR("too many retries, giving up\n");
-	ret = -EREMOTEIO;
+	ret = drm_dp_aux_register(&intel_dp->aux);
+	if (ret < 0) {
+		DRM_ERROR("drm_dp_aux_register() for %s failed (%d)\n",
+			  name, ret);
+		return;
+	}
 
-out:
-	ironlake_edp_panel_vdd_off(intel_dp, false);
-	return ret;
+	ret = sysfs_create_link(&connector->base.kdev->kobj,
+				&intel_dp->aux.ddc.dev.kobj,
+				intel_dp->aux.ddc.dev.kobj.name);
+	if (ret < 0) {
+		DRM_ERROR("sysfs_create_link() for %s failed (%d)\n", name, ret);
+		drm_dp_aux_unregister(&intel_dp->aux);
+	}
 }
 
-static int
-intel_dp_i2c_init(struct intel_dp *intel_dp,
-		  struct intel_connector *intel_connector, const char *name)
+static void
+intel_dp_connector_unregister(struct intel_connector *intel_connector)
 {
-	int	ret;
-
-	DRM_DEBUG_KMS("i2c_init %s\n", name);
-	intel_dp->algo.running = false;
-	intel_dp->algo.address = 0;
-	intel_dp->algo.aux_ch = intel_dp_i2c_aux_ch;
-
-	memset(&intel_dp->adapter, '\0', sizeof(intel_dp->adapter));
-	intel_dp->adapter.owner = THIS_MODULE;
-	intel_dp->adapter.class = I2C_CLASS_DDC;
-	strncpy(intel_dp->adapter.name, name, sizeof(intel_dp->adapter.name) - 1);
-	intel_dp->adapter.name[sizeof(intel_dp->adapter.name) - 1] = '\0';
-	intel_dp->adapter.algo_data = &intel_dp->algo;
-	intel_dp->adapter.dev.parent = intel_connector->base.kdev;
+	struct intel_dp *intel_dp = intel_attached_dp(&intel_connector->base);
 
-	ret = i2c_dp_aux_add_bus(&intel_dp->adapter);
-	return ret;
+	sysfs_remove_link(&intel_connector->base.kdev->kobj,
+			  intel_dp->aux.ddc.dev.kobj.name);
+	intel_connector_unregister(intel_connector);
 }
 
 static void
@@ -775,6 +794,9 @@ intel_dp_set_clock(struct intel_encoder *encoder,
 	} else if (HAS_PCH_SPLIT(dev)) {
 		divisor = pch_dpll;
 		count = ARRAY_SIZE(pch_dpll);
+	} else if (IS_CHERRYVIEW(dev)) {
+		divisor = chv_dpll;
+		count = ARRAY_SIZE(chv_dpll);
 	} else if (IS_VALLEYVIEW(dev)) {
 		divisor = vlv_dpll;
 		count = ARRAY_SIZE(vlv_dpll);
@@ -791,6 +813,20 @@ intel_dp_set_clock(struct intel_encoder *encoder,
 	}
 }
 
+static void
+intel_dp_set_m2_n2(struct intel_crtc *crtc, struct intel_link_m_n *m_n)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum transcoder transcoder = crtc->config.cpu_transcoder;
+
+	I915_WRITE(PIPE_DATA_M2(transcoder),
+		TU_SIZE(m_n->tu) | m_n->gmch_m);
+	I915_WRITE(PIPE_DATA_N2(transcoder), m_n->gmch_n);
+	I915_WRITE(PIPE_LINK_M2(transcoder), m_n->link_m);
+	I915_WRITE(PIPE_LINK_N2(transcoder), m_n->link_n);
+}
+
 bool
 intel_dp_compute_config(struct intel_encoder *encoder,
 			struct intel_crtc_config *pipe_config)
@@ -803,16 +839,20 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	struct intel_crtc *intel_crtc = encoder->new_crtc;
 	struct intel_connector *intel_connector = intel_dp->attached_connector;
 	int lane_count, clock;
-	int max_lane_count = drm_dp_max_lane_count(intel_dp->dpcd);
-	int max_clock = intel_dp_max_link_bw(intel_dp) == DP_LINK_BW_2_7 ? 1 : 0;
+	int min_lane_count = 1;
+	int max_lane_count = intel_dp_max_lane_count(intel_dp);
+	/* Conveniently, the link BW constants become indices with a shift...*/
+	int min_clock = 0;
+	int max_clock = intel_dp_max_link_bw(intel_dp) >> 3;
 	int bpp, mode_rate;
-	static int bws[2] = { DP_LINK_BW_1_62, DP_LINK_BW_2_7 };
+	static int bws[] = { DP_LINK_BW_1_62, DP_LINK_BW_2_7, DP_LINK_BW_5_4 };
 	int link_avail, link_clock;
 
 	if (HAS_PCH_SPLIT(dev) && !HAS_DDI(dev) && port != PORT_A)
 		pipe_config->has_pch_encoder = true;
 
 	pipe_config->has_dp_encoder = true;
+	pipe_config->has_audio = intel_dp->has_audio;
 
 	if (is_edp(intel_dp) && intel_connector->panel.fixed_mode) {
 		intel_fixed_panel_mode(intel_connector->panel.fixed_mode,
@@ -836,19 +876,38 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	/* Walk through all bpp values. Luckily they're all nicely spaced with 2
 	 * bpc in between. */
 	bpp = pipe_config->pipe_bpp;
-	if (is_edp(intel_dp) && dev_priv->vbt.edp_bpp &&
-	    dev_priv->vbt.edp_bpp < bpp) {
-		DRM_DEBUG_KMS("clamping bpp for eDP panel to BIOS-provided %i\n",
-			      dev_priv->vbt.edp_bpp);
-		bpp = dev_priv->vbt.edp_bpp;
+	if (is_edp(intel_dp)) {
+		if (dev_priv->vbt.edp_bpp && dev_priv->vbt.edp_bpp < bpp) {
+			DRM_DEBUG_KMS("clamping bpp for eDP panel to BIOS-provided %i\n",
+				      dev_priv->vbt.edp_bpp);
+			bpp = dev_priv->vbt.edp_bpp;
+		}
+
+		if (IS_BROADWELL(dev)) {
+			/* Yes, it's an ugly hack. */
+			min_lane_count = max_lane_count;
+			DRM_DEBUG_KMS("forcing lane count to max (%u) on BDW\n",
+				      min_lane_count);
+		} else if (dev_priv->vbt.edp_lanes) {
+			min_lane_count = min(dev_priv->vbt.edp_lanes,
+					     max_lane_count);
+			DRM_DEBUG_KMS("using min %u lanes per VBT\n",
+				      min_lane_count);
+		}
+
+		if (dev_priv->vbt.edp_rate) {
+			min_clock = min(dev_priv->vbt.edp_rate >> 3, max_clock);
+			DRM_DEBUG_KMS("using min %02x link bw per VBT\n",
+				      bws[min_clock]);
+		}
 	}
 
 	for (; bpp >= 6*3; bpp -= 2*3) {
 		mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
 						   bpp);
 
-		for (clock = 0; clock <= max_clock; clock++) {
-			for (lane_count = 1; lane_count <= max_lane_count; lane_count <<= 1) {
+		for (clock = min_clock; clock <= max_clock; clock++) {
+			for (lane_count = min_lane_count; lane_count <= max_lane_count; lane_count <<= 1) {
 				link_clock = drm_dp_bw_code_to_link_rate(bws[clock]);
 				link_avail = intel_dp_max_data_rate(link_clock,
 								    lane_count);
@@ -894,6 +953,14 @@ found:
 			       pipe_config->port_clock,
 			       &pipe_config->dp_m_n);
 
+	if (intel_connector->panel.downclock_mode != NULL &&
+		intel_dp->drrs_state.type == SEAMLESS_DRRS_SUPPORT) {
+			intel_link_compute_m_n(bpp, lane_count,
+				intel_connector->panel.downclock_mode->clock,
+				pipe_config->port_clock,
+				&pipe_config->dp_m2_n2);
+	}
+
 	intel_dp_set_clock(encoder, pipe_config, intel_dp->link_bw);
 
 	return true;
@@ -929,7 +996,7 @@ static void ironlake_set_pll_cpu_edp(struct intel_dp *intel_dp)
 	udelay(500);
 }
 
-static void intel_dp_mode_set(struct intel_encoder *encoder)
+static void intel_dp_prepare(struct intel_encoder *encoder)
 {
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -964,7 +1031,7 @@ static void intel_dp_mode_set(struct intel_encoder *encoder)
 	intel_dp->DP |= DP_VOLTAGE_0_4 | DP_PRE_EMPHASIS_0;
 	intel_dp->DP |= DP_PORT_WIDTH(intel_dp->lane_count);
 
-	if (intel_dp->has_audio) {
+	if (crtc->config.has_audio) {
 		DRM_DEBUG_DRIVER("Enabling DP audio on pipe %c\n",
 				 pipe_name(crtc->pipe));
 		intel_dp->DP |= DP_AUDIO_OUTPUT_ENABLE;
@@ -997,26 +1064,27 @@ static void intel_dp_mode_set(struct intel_encoder *encoder)
 		if (drm_dp_enhanced_frame_cap(intel_dp->dpcd))
 			intel_dp->DP |= DP_ENHANCED_FRAMING;
 
-		if (crtc->pipe == 1)
-			intel_dp->DP |= DP_PIPEB_SELECT;
+		if (!IS_CHERRYVIEW(dev)) {
+			if (crtc->pipe == 1)
+				intel_dp->DP |= DP_PIPEB_SELECT;
+		} else {
+			intel_dp->DP |= DP_PIPE_SELECT_CHV(crtc->pipe);
+		}
 	} else {
 		intel_dp->DP |= DP_LINK_TRAIN_OFF_CPT;
 	}
-
-	if (port == PORT_A && !IS_VALLEYVIEW(dev))
-		ironlake_set_pll_cpu_edp(intel_dp);
 }
 
-#define IDLE_ON_MASK		(PP_ON | 0 	  | PP_SEQUENCE_MASK | 0                     | PP_SEQUENCE_STATE_MASK)
-#define IDLE_ON_VALUE   	(PP_ON | 0 	  | PP_SEQUENCE_NONE | 0                     | PP_SEQUENCE_STATE_ON_IDLE)
+#define IDLE_ON_MASK		(PP_ON | PP_SEQUENCE_MASK | 0                     | PP_SEQUENCE_STATE_MASK)
+#define IDLE_ON_VALUE   	(PP_ON | PP_SEQUENCE_NONE | 0                     | PP_SEQUENCE_STATE_ON_IDLE)
 
-#define IDLE_OFF_MASK		(PP_ON | 0        | PP_SEQUENCE_MASK | 0                     | PP_SEQUENCE_STATE_MASK)
-#define IDLE_OFF_VALUE		(0     | 0        | PP_SEQUENCE_NONE | 0                     | PP_SEQUENCE_STATE_OFF_IDLE)
+#define IDLE_OFF_MASK		(PP_ON | PP_SEQUENCE_MASK | 0                     | 0)
+#define IDLE_OFF_VALUE		(0     | PP_SEQUENCE_NONE | 0                     | 0)
 
-#define IDLE_CYCLE_MASK		(PP_ON | 0        | PP_SEQUENCE_MASK | PP_CYCLE_DELAY_ACTIVE | PP_SEQUENCE_STATE_MASK)
-#define IDLE_CYCLE_VALUE	(0     | 0        | PP_SEQUENCE_NONE | 0                     | PP_SEQUENCE_STATE_OFF_IDLE)
+#define IDLE_CYCLE_MASK		(PP_ON | PP_SEQUENCE_MASK | PP_CYCLE_DELAY_ACTIVE | PP_SEQUENCE_STATE_MASK)
+#define IDLE_CYCLE_VALUE	(0     | PP_SEQUENCE_NONE | 0                     | PP_SEQUENCE_STATE_OFF_IDLE)
 
-static void ironlake_wait_panel_status(struct intel_dp *intel_dp,
+static void wait_panel_status(struct intel_dp *intel_dp,
 				       u32 mask,
 				       u32 value)
 {
@@ -1037,26 +1105,45 @@ static void ironlake_wait_panel_status(struct intel_dp *intel_dp,
 				I915_READ(pp_stat_reg),
 				I915_READ(pp_ctrl_reg));
 	}
+
+	DRM_DEBUG_KMS("Wait complete\n");
 }
 
-static void ironlake_wait_panel_on(struct intel_dp *intel_dp)
+static void wait_panel_on(struct intel_dp *intel_dp)
 {
 	DRM_DEBUG_KMS("Wait for panel power on\n");
-	ironlake_wait_panel_status(intel_dp, IDLE_ON_MASK, IDLE_ON_VALUE);
+	wait_panel_status(intel_dp, IDLE_ON_MASK, IDLE_ON_VALUE);
 }
 
-static void ironlake_wait_panel_off(struct intel_dp *intel_dp)
+static void wait_panel_off(struct intel_dp *intel_dp)
 {
 	DRM_DEBUG_KMS("Wait for panel power off time\n");
-	ironlake_wait_panel_status(intel_dp, IDLE_OFF_MASK, IDLE_OFF_VALUE);
+	wait_panel_status(intel_dp, IDLE_OFF_MASK, IDLE_OFF_VALUE);
 }
 
-static void ironlake_wait_panel_power_cycle(struct intel_dp *intel_dp)
+static void wait_panel_power_cycle(struct intel_dp *intel_dp)
 {
 	DRM_DEBUG_KMS("Wait for panel power cycle\n");
-	ironlake_wait_panel_status(intel_dp, IDLE_CYCLE_MASK, IDLE_CYCLE_VALUE);
+
+	/* When we disable the VDD override bit last we have to do the manual
+	 * wait. */
+	wait_remaining_ms_from_jiffies(intel_dp->last_power_cycle,
+				       intel_dp->panel_power_cycle_delay);
+
+	wait_panel_status(intel_dp, IDLE_CYCLE_MASK, IDLE_CYCLE_VALUE);
+}
+
+static void wait_backlight_on(struct intel_dp *intel_dp)
+{
+	wait_remaining_ms_from_jiffies(intel_dp->last_power_on,
+				       intel_dp->backlight_on_delay);
 }
 
+static void edp_wait_backlight_off(struct intel_dp *intel_dp)
+{
+	wait_remaining_ms_from_jiffies(intel_dp->last_backlight_off,
+				       intel_dp->backlight_off_delay);
+}
 
 /* Read the current pp_control value, unlocking the register if it
  * is locked
@@ -1074,28 +1161,32 @@ static  u32 ironlake_get_pp_control(struct intel_dp *intel_dp)
 	return control;
 }
 
-void ironlake_edp_panel_vdd_on(struct intel_dp *intel_dp)
+static bool _edp_panel_vdd_on(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *intel_encoder = &intel_dig_port->base;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum intel_display_power_domain power_domain;
 	u32 pp;
 	u32 pp_stat_reg, pp_ctrl_reg;
+	bool need_to_disable = !intel_dp->want_panel_vdd;
 
 	if (!is_edp(intel_dp))
-		return;
-
-	WARN(intel_dp->want_panel_vdd,
-	     "eDP VDD already requested on\n");
+		return false;
 
 	intel_dp->want_panel_vdd = true;
 
-	if (ironlake_edp_have_panel_vdd(intel_dp))
-		return;
+	if (edp_have_panel_vdd(intel_dp))
+		return need_to_disable;
+
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_get(dev_priv, power_domain);
 
 	DRM_DEBUG_KMS("Turning eDP VDD on\n");
 
-	if (!ironlake_edp_have_panel_power(intel_dp))
-		ironlake_wait_panel_power_cycle(intel_dp);
+	if (!edp_have_panel_power(intel_dp))
+		wait_panel_power_cycle(intel_dp);
 
 	pp = ironlake_get_pp_control(intel_dp);
 	pp |= EDP_FORCE_VDD;
@@ -1110,22 +1201,38 @@ void ironlake_edp_panel_vdd_on(struct intel_dp *intel_dp)
 	/*
 	 * If the panel wasn't on, delay before accessing aux channel
 	 */
-	if (!ironlake_edp_have_panel_power(intel_dp)) {
+	if (!edp_have_panel_power(intel_dp)) {
 		DRM_DEBUG_KMS("eDP was not running\n");
 		msleep(intel_dp->panel_power_up_delay);
 	}
+
+	return need_to_disable;
 }
 
-static void ironlake_panel_vdd_off_sync(struct intel_dp *intel_dp)
+void intel_edp_panel_vdd_on(struct intel_dp *intel_dp)
+{
+	if (is_edp(intel_dp)) {
+		bool vdd = _edp_panel_vdd_on(intel_dp);
+
+		WARN(!vdd, "eDP VDD already requested on\n");
+	}
+}
+
+static void edp_panel_vdd_off_sync(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 pp;
 	u32 pp_stat_reg, pp_ctrl_reg;
 
-	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+
+	if (!intel_dp->want_panel_vdd && edp_have_panel_vdd(intel_dp)) {
+		struct intel_digital_port *intel_dig_port =
+						dp_to_dig_port(intel_dp);
+		struct intel_encoder *intel_encoder = &intel_dig_port->base;
+		enum intel_display_power_domain power_domain;
 
-	if (!intel_dp->want_panel_vdd && ironlake_edp_have_panel_vdd(intel_dp)) {
 		DRM_DEBUG_KMS("Turning eDP VDD off\n");
 
 		pp = ironlake_get_pp_control(intel_dp);
@@ -1140,22 +1247,27 @@ static void ironlake_panel_vdd_off_sync(struct intel_dp *intel_dp)
 		/* Make sure sequencer is idle before allowing subsequent activity */
 		DRM_DEBUG_KMS("PP_STATUS: 0x%08x PP_CONTROL: 0x%08x\n",
 		I915_READ(pp_stat_reg), I915_READ(pp_ctrl_reg));
-		msleep(intel_dp->panel_power_down_delay);
+
+		if ((pp & POWER_TARGET_ON) == 0)
+			intel_dp->last_power_cycle = jiffies;
+
+		power_domain = intel_display_port_power_domain(intel_encoder);
+		intel_display_power_put(dev_priv, power_domain);
 	}
 }
 
-static void ironlake_panel_vdd_work(struct work_struct *__work)
+static void edp_panel_vdd_work(struct work_struct *__work)
 {
 	struct intel_dp *intel_dp = container_of(to_delayed_work(__work),
 						 struct intel_dp, panel_vdd_work);
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 
-	mutex_lock(&dev->mode_config.mutex);
-	ironlake_panel_vdd_off_sync(intel_dp);
-	mutex_unlock(&dev->mode_config.mutex);
+	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+	edp_panel_vdd_off_sync(intel_dp);
+	drm_modeset_unlock(&dev->mode_config.connection_mutex);
 }
 
-void ironlake_edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
+static void edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
 {
 	if (!is_edp(intel_dp))
 		return;
@@ -1165,7 +1277,7 @@ void ironlake_edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
 	intel_dp->want_panel_vdd = false;
 
 	if (sync) {
-		ironlake_panel_vdd_off_sync(intel_dp);
+		edp_panel_vdd_off_sync(intel_dp);
 	} else {
 		/*
 		 * Queue the timer to fire a long
@@ -1177,7 +1289,7 @@ void ironlake_edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync)
 	}
 }
 
-void ironlake_edp_panel_on(struct intel_dp *intel_dp)
+void intel_edp_panel_on(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1189,12 +1301,12 @@ void ironlake_edp_panel_on(struct intel_dp *intel_dp)
 
 	DRM_DEBUG_KMS("Turn eDP power on\n");
 
-	if (ironlake_edp_have_panel_power(intel_dp)) {
+	if (edp_have_panel_power(intel_dp)) {
 		DRM_DEBUG_KMS("eDP power already on\n");
 		return;
 	}
 
-	ironlake_wait_panel_power_cycle(intel_dp);
+	wait_panel_power_cycle(intel_dp);
 
 	pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
 	pp = ironlake_get_pp_control(intel_dp);
@@ -1212,7 +1324,8 @@ void ironlake_edp_panel_on(struct intel_dp *intel_dp)
 	I915_WRITE(pp_ctrl_reg, pp);
 	POSTING_READ(pp_ctrl_reg);
 
-	ironlake_wait_panel_on(intel_dp);
+	wait_panel_on(intel_dp);
+	intel_dp->last_power_on = jiffies;
 
 	if (IS_GEN5(dev)) {
 		pp |= PANEL_POWER_RESET; /* restore panel reset bit */
@@ -1221,10 +1334,13 @@ void ironlake_edp_panel_on(struct intel_dp *intel_dp)
 	}
 }
 
-void ironlake_edp_panel_off(struct intel_dp *intel_dp)
+void intel_edp_panel_off(struct intel_dp *intel_dp)
 {
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *intel_encoder = &intel_dig_port->base;
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum intel_display_power_domain power_domain;
 	u32 pp;
 	u32 pp_ctrl_reg;
 
@@ -1233,24 +1349,32 @@ void ironlake_edp_panel_off(struct intel_dp *intel_dp)
 
 	DRM_DEBUG_KMS("Turn eDP power off\n");
 
+	edp_wait_backlight_off(intel_dp);
+
 	WARN(!intel_dp->want_panel_vdd, "Need VDD to turn off panel\n");
 
 	pp = ironlake_get_pp_control(intel_dp);
 	/* We need to switch off panel power _and_ force vdd, for otherwise some
 	 * panels get very unhappy and cease to work. */
-	pp &= ~(POWER_TARGET_ON | EDP_FORCE_VDD | PANEL_POWER_RESET | EDP_BLC_ENABLE);
+	pp &= ~(POWER_TARGET_ON | PANEL_POWER_RESET | EDP_FORCE_VDD |
+		EDP_BLC_ENABLE);
 
 	pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
 
+	intel_dp->want_panel_vdd = false;
+
 	I915_WRITE(pp_ctrl_reg, pp);
 	POSTING_READ(pp_ctrl_reg);
 
-	intel_dp->want_panel_vdd = false;
+	intel_dp->last_power_cycle = jiffies;
+	wait_panel_off(intel_dp);
 
-	ironlake_wait_panel_off(intel_dp);
+	/* We got a reference when we enabled the VDD. */
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_put(dev_priv, power_domain);
 }
 
-void ironlake_edp_backlight_on(struct intel_dp *intel_dp)
+void intel_edp_backlight_on(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct drm_device *dev = intel_dig_port->base.base.dev;
@@ -1268,7 +1392,7 @@ void ironlake_edp_backlight_on(struct intel_dp *intel_dp)
 	 * link.  So delay a bit to make sure the image is solid before
 	 * allowing it to appear.
 	 */
-	msleep(intel_dp->backlight_on_delay);
+	wait_backlight_on(intel_dp);
 	pp = ironlake_get_pp_control(intel_dp);
 	pp |= EDP_BLC_ENABLE;
 
@@ -1280,7 +1404,7 @@ void ironlake_edp_backlight_on(struct intel_dp *intel_dp)
 	intel_panel_enable_backlight(intel_dp->attached_connector);
 }
 
-void ironlake_edp_backlight_off(struct intel_dp *intel_dp)
+void intel_edp_backlight_off(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1300,7 +1424,7 @@ void ironlake_edp_backlight_off(struct intel_dp *intel_dp)
 
 	I915_WRITE(pp_ctrl_reg, pp);
 	POSTING_READ(pp_ctrl_reg);
-	msleep(intel_dp->backlight_off_delay);
+	intel_dp->last_backlight_off = jiffies;
 }
 
 static void ironlake_edp_pll_on(struct intel_dp *intel_dp)
@@ -1364,8 +1488,8 @@ void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode)
 		return;
 
 	if (mode != DRM_MODE_DPMS_ON) {
-		ret = intel_dp_aux_native_write_1(intel_dp, DP_SET_POWER,
-						  DP_SET_POWER_D3);
+		ret = drm_dp_dpcd_writeb(&intel_dp->aux, DP_SET_POWER,
+					 DP_SET_POWER_D3);
 		if (ret != 1)
 			DRM_DEBUG_DRIVER("failed to write sink power state\n");
 	} else {
@@ -1374,9 +1498,8 @@ void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode)
 		 * time to wake up.
 		 */
 		for (i = 0; i < 3; i++) {
-			ret = intel_dp_aux_native_write_1(intel_dp,
-							  DP_SET_POWER,
-							  DP_SET_POWER_D0);
+			ret = drm_dp_dpcd_writeb(&intel_dp->aux, DP_SET_POWER,
+						 DP_SET_POWER_D0);
 			if (ret == 1)
 				break;
 			msleep(1);
@@ -1391,13 +1514,22 @@ static bool intel_dp_get_hw_state(struct intel_encoder *encoder,
 	enum port port = dp_to_dig_port(intel_dp)->port;
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 tmp = I915_READ(intel_dp->output_reg);
+	enum intel_display_power_domain power_domain;
+	u32 tmp;
+
+	power_domain = intel_display_port_power_domain(encoder);
+	if (!intel_display_power_enabled(dev_priv, power_domain))
+		return false;
+
+	tmp = I915_READ(intel_dp->output_reg);
 
 	if (!(tmp & DP_PORT_EN))
 		return false;
 
 	if (port == PORT_A && IS_GEN7(dev) && !IS_VALLEYVIEW(dev)) {
 		*pipe = PORT_TO_PIPE_CPT(tmp);
+	} else if (IS_CHERRYVIEW(dev)) {
+		*pipe = DP_PORT_TO_PIPE_CHV(tmp);
 	} else if (!HAS_PCH_CPT(dev) || port == PORT_A) {
 		*pipe = PORT_TO_PIPE(tmp);
 	} else {
@@ -1445,8 +1577,11 @@ static void intel_dp_get_config(struct intel_encoder *encoder,
 	struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
 	int dotclock;
 
+	tmp = I915_READ(intel_dp->output_reg);
+	if (tmp & DP_AUDIO_OUTPUT_ENABLE)
+		pipe_config->has_audio = true;
+
 	if ((port == PORT_A) || !HAS_PCH_CPT(dev)) {
-		tmp = I915_READ(intel_dp->output_reg);
 		if (tmp & DP_SYNC_HS_HIGH)
 			flags |= DRM_MODE_FLAG_PHSYNC;
 		else
@@ -1585,19 +1720,19 @@ static void intel_edp_psr_enable_sink(struct intel_dp *intel_dp)
 {
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t aux_clock_divider = get_aux_clock_divider(intel_dp, 0);
+	uint32_t aux_clock_divider;
 	int precharge = 0x3;
 	int msg_size = 5;       /* Header(4) + Message(1) */
 
+	aux_clock_divider = intel_dp->get_aux_clock_divider(intel_dp, 0);
+
 	/* Enable PSR in sink */
 	if (intel_dp->psr_dpcd[1] & DP_PSR_NO_TRAIN_ON_EXIT)
-		intel_dp_aux_native_write_1(intel_dp, DP_PSR_EN_CFG,
-					    DP_PSR_ENABLE &
-					    ~DP_PSR_MAIN_LINK_ACTIVE);
+		drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_EN_CFG,
+				   DP_PSR_ENABLE & ~DP_PSR_MAIN_LINK_ACTIVE);
 	else
-		intel_dp_aux_native_write_1(intel_dp, DP_PSR_EN_CFG,
-					    DP_PSR_ENABLE |
-					    DP_PSR_MAIN_LINK_ACTIVE);
+		drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_EN_CFG,
+				   DP_PSR_ENABLE | DP_PSR_MAIN_LINK_ACTIVE);
 
 	/* Setup AUX registers */
 	I915_WRITE(EDP_PSR_AUX_DATA1(dev), EDP_PSR_DPCD_COMMAND);
@@ -1627,7 +1762,7 @@ static void intel_edp_psr_enable_source(struct intel_dp *intel_dp)
 		val |= EDP_PSR_LINK_DISABLE;
 
 	I915_WRITE(EDP_PSR_CTL(dev), val |
-		   IS_BROADWELL(dev) ? 0 : link_entry_time |
+		   (IS_BROADWELL(dev) ? 0 : link_entry_time) |
 		   max_sleep_time << EDP_PSR_MAX_SLEEP_TIME_SHIFT |
 		   idle_frames << EDP_PSR_IDLE_FRAME_SHIFT |
 		   EDP_PSR_ENABLE);
@@ -1640,7 +1775,7 @@ static bool intel_edp_psr_match_conditions(struct intel_dp *intel_dp)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc = dig_port->base.base.crtc;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_gem_object *obj = to_intel_framebuffer(crtc->fb)->obj;
+	struct drm_i915_gem_object *obj = to_intel_framebuffer(crtc->primary->fb)->obj;
 	struct intel_encoder *intel_encoder = &dp_to_dig_port(intel_dp)->base;
 
 	dev_priv->psr.source_ok = false;
@@ -1656,7 +1791,7 @@ static bool intel_edp_psr_match_conditions(struct intel_dp *intel_dp)
 		return false;
 	}
 
-	if (!i915_enable_psr) {
+	if (!i915.enable_psr) {
 		DRM_DEBUG_KMS("PSR disable by flag\n");
 		return false;
 	}
@@ -1673,7 +1808,7 @@ static bool intel_edp_psr_match_conditions(struct intel_dp *intel_dp)
 		return false;
 	}
 
-	obj = to_intel_framebuffer(crtc->fb)->obj;
+	obj = to_intel_framebuffer(crtc->primary->fb)->obj;
 	if (obj->tiling_mode != I915_TILING_X ||
 	    obj->fence_reg == I915_FENCE_REG_NONE) {
 		DRM_DEBUG_KMS("PSR condition failed: fb not tiled or fenced\n");
@@ -1772,27 +1907,69 @@ static void intel_disable_dp(struct intel_encoder *encoder)
 
 	/* Make sure the panel is off before trying to change the mode. But also
 	 * ensure that we have vdd while we switch off the panel. */
-	ironlake_edp_panel_vdd_on(intel_dp);
-	ironlake_edp_backlight_off(intel_dp);
+	intel_edp_panel_vdd_on(intel_dp);
+	intel_edp_backlight_off(intel_dp);
 	intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
-	ironlake_edp_panel_off(intel_dp);
+	intel_edp_panel_off(intel_dp);
 
 	/* cpu edp my only be disable _after_ the cpu pipe/plane is disabled. */
 	if (!(port == PORT_A || IS_VALLEYVIEW(dev)))
 		intel_dp_link_down(intel_dp);
 }
 
-static void intel_post_disable_dp(struct intel_encoder *encoder)
+static void g4x_post_disable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	enum port port = dp_to_dig_port(intel_dp)->port;
+
+	if (port != PORT_A)
+		return;
+
+	intel_dp_link_down(intel_dp);
+	ironlake_edp_pll_off(intel_dp);
+}
+
+static void vlv_post_disable_dp(struct intel_encoder *encoder)
+{
+	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+
+	intel_dp_link_down(intel_dp);
+}
+
+static void chv_post_disable_dp(struct intel_encoder *encoder)
+{
+	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+	struct intel_digital_port *dport = dp_to_dig_port(intel_dp);
 	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc =
+		to_intel_crtc(encoder->base.crtc);
+	enum dpio_channel ch = vlv_dport_to_channel(dport);
+	enum pipe pipe = intel_crtc->pipe;
+	u32 val;
 
-	if (port == PORT_A || IS_VALLEYVIEW(dev)) {
-		intel_dp_link_down(intel_dp);
-		if (!IS_VALLEYVIEW(dev))
-			ironlake_edp_pll_off(intel_dp);
-	}
+	intel_dp_link_down(intel_dp);
+
+	mutex_lock(&dev_priv->dpio_lock);
+
+	/* Propagate soft reset to data lane reset */
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW1(ch));
+	val |= CHV_PCS_REQ_SOFTRESET_EN;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW1(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW1(ch));
+	val |= CHV_PCS_REQ_SOFTRESET_EN;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW1(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW0(ch));
+	val &= ~(DPIO_PCS_TX_LANE2_RESET | DPIO_PCS_TX_LANE1_RESET);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW0(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW0(ch));
+	val &= ~(DPIO_PCS_TX_LANE2_RESET | DPIO_PCS_TX_LANE1_RESET);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW0(ch), val);
+
+	mutex_unlock(&dev_priv->dpio_lock);
 }
 
 static void intel_enable_dp(struct intel_encoder *encoder)
@@ -1805,11 +1982,11 @@ static void intel_enable_dp(struct intel_encoder *encoder)
 	if (WARN_ON(dp_reg & DP_PORT_EN))
 		return;
 
-	ironlake_edp_panel_vdd_on(intel_dp);
+	intel_edp_panel_vdd_on(intel_dp);
 	intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON);
 	intel_dp_start_link_train(intel_dp);
-	ironlake_edp_panel_on(intel_dp);
-	ironlake_edp_panel_vdd_off(intel_dp, true);
+	intel_edp_panel_on(intel_dp);
+	edp_panel_vdd_off(intel_dp, true);
 	intel_dp_complete_link_train(intel_dp);
 	intel_dp_stop_link_train(intel_dp);
 }
@@ -1819,14 +1996,14 @@ static void g4x_enable_dp(struct intel_encoder *encoder)
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 
 	intel_enable_dp(encoder);
-	ironlake_edp_backlight_on(intel_dp);
+	intel_edp_backlight_on(intel_dp);
 }
 
 static void vlv_enable_dp(struct intel_encoder *encoder)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 
-	ironlake_edp_backlight_on(intel_dp);
+	intel_edp_backlight_on(intel_dp);
 }
 
 static void g4x_pre_enable_dp(struct intel_encoder *encoder)
@@ -1834,8 +2011,13 @@ static void g4x_pre_enable_dp(struct intel_encoder *encoder)
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct intel_digital_port *dport = dp_to_dig_port(intel_dp);
 
-	if (dport->port == PORT_A)
+	intel_dp_prepare(encoder);
+
+	/* Only ilk+ has port A */
+	if (dport->port == PORT_A) {
+		ironlake_set_pll_cpu_edp(intel_dp);
 		ironlake_edp_pll_on(intel_dp);
+	}
 }
 
 static void vlv_pre_enable_dp(struct intel_encoder *encoder)
@@ -1845,34 +2027,36 @@ static void vlv_pre_enable_dp(struct intel_encoder *encoder)
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 	struct edp_power_seq power_seq;
 	u32 val;
 
 	mutex_lock(&dev_priv->dpio_lock);
 
-	val = vlv_dpio_read(dev_priv, pipe, DPIO_DATA_LANE_A(port));
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW8(port));
 	val = 0;
 	if (pipe)
 		val |= (1<<21);
 	else
 		val &= ~(1<<21);
 	val |= 0x001000c4;
-	vlv_dpio_write(dev_priv, pipe, DPIO_DATA_CHANNEL(port), val);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLOCKBUF0(port), 0x00760018);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLOCKBUF8(port), 0x00400888);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW8(port), val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW14(port), 0x00760018);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW23(port), 0x00400888);
 
 	mutex_unlock(&dev_priv->dpio_lock);
 
-	/* init power sequencer on this pipe and port */
-	intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
-	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
-						      &power_seq);
+	if (is_edp(intel_dp)) {
+		/* init power sequencer on this pipe and port */
+		intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
+		intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
+							      &power_seq);
+	}
 
 	intel_enable_dp(encoder);
 
-	vlv_wait_port_ready(dev_priv, port);
+	vlv_wait_port_ready(dev_priv, dport);
 }
 
 static void vlv_dp_pre_pll_enable(struct intel_encoder *encoder)
@@ -1882,50 +2066,114 @@ static void vlv_dp_pre_pll_enable(struct intel_encoder *encoder)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 
+	intel_dp_prepare(encoder);
+
 	/* Program Tx lane resets to default */
 	mutex_lock(&dev_priv->dpio_lock);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_TX(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW0(port),
 			 DPIO_PCS_TX_LANE2_RESET |
 			 DPIO_PCS_TX_LANE1_RESET);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLK(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW1(port),
 			 DPIO_PCS_CLK_CRI_RXEB_EIOS_EN |
 			 DPIO_PCS_CLK_CRI_RXDIGFILTSG_EN |
 			 (1<<DPIO_PCS_CLK_DATAWIDTH_SHIFT) |
 				 DPIO_PCS_CLK_SOFT_RESET);
 
 	/* Fix up inter-pair skew failure */
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_STAGGER1(port), 0x00750f00);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_CTL(port), 0x00001500);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_LANE(port), 0x40400000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW12(port), 0x00750f00);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW11(port), 0x00001500);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW14(port), 0x40400000);
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
+static void chv_pre_enable_dp(struct intel_encoder *encoder)
+{
+	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+	struct intel_digital_port *dport = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct edp_power_seq power_seq;
+	struct intel_crtc *intel_crtc =
+		to_intel_crtc(encoder->base.crtc);
+	enum dpio_channel ch = vlv_dport_to_channel(dport);
+	int pipe = intel_crtc->pipe;
+	int data, i;
+	u32 val;
+
+	mutex_lock(&dev_priv->dpio_lock);
+
+	/* Deassert soft data lane reset*/
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW1(ch));
+	val |= CHV_PCS_REQ_SOFTRESET_EN;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW1(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW1(ch));
+	val |= CHV_PCS_REQ_SOFTRESET_EN;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW1(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW0(ch));
+	val |= (DPIO_PCS_TX_LANE2_RESET | DPIO_PCS_TX_LANE1_RESET);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW0(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW0(ch));
+	val |= (DPIO_PCS_TX_LANE2_RESET | DPIO_PCS_TX_LANE1_RESET);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW0(ch), val);
+
+	/* Program Tx lane latency optimal setting*/
+	for (i = 0; i < 4; i++) {
+		/* Set the latency optimal bit */
+		data = (i == 1) ? 0x0 : 0x6;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW11(ch, i),
+				data << DPIO_FRC_LATENCY_SHFIT);
+
+		/* Set the upar bit */
+		data = (i == 1) ? 0x0 : 0x1;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW14(ch, i),
+				data << DPIO_UPAR_SHIFT);
+	}
+
+	/* Data lane stagger programming */
+	/* FIXME: Fix up value only after power analysis */
+
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	if (is_edp(intel_dp)) {
+		/* init power sequencer on this pipe and port */
+		intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
+		intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
+							      &power_seq);
+	}
+
+	intel_enable_dp(encoder);
+
+	vlv_wait_port_ready(dev_priv, dport);
+}
+
 /*
  * Native read with retry for link status and receiver capability reads for
  * cases where the sink may still be asleep.
+ *
+ * Sinks are *supposed* to come up within 1ms from an off state, but we're also
+ * supposed to retry 3 times per the spec.
  */
-static bool
-intel_dp_aux_native_read_retry(struct intel_dp *intel_dp, uint16_t address,
-			       uint8_t *recv, int recv_bytes)
+static ssize_t
+intel_dp_dpcd_read_wake(struct drm_dp_aux *aux, unsigned int offset,
+			void *buffer, size_t size)
 {
-	int ret, i;
+	ssize_t ret;
+	int i;
 
-	/*
-	 * Sinks are *supposed* to come up within 1ms from an off state,
-	 * but we're also supposed to retry 3 times per the spec.
-	 */
 	for (i = 0; i < 3; i++) {
-		ret = intel_dp_aux_native_read(intel_dp, address, recv,
-					       recv_bytes);
-		if (ret == recv_bytes)
-			return true;
+		ret = drm_dp_dpcd_read(aux, offset, buffer, size);
+		if (ret == size)
+			return ret;
 		msleep(1);
 	}
 
-	return false;
+	return ret;
 }
 
 /*
@@ -1935,24 +2183,12 @@ intel_dp_aux_native_read_retry(struct intel_dp *intel_dp, uint16_t address,
 static bool
 intel_dp_get_link_status(struct intel_dp *intel_dp, uint8_t link_status[DP_LINK_STATUS_SIZE])
 {
-	return intel_dp_aux_native_read_retry(intel_dp,
-					      DP_LANE0_1_STATUS,
-					      link_status,
-					      DP_LINK_STATUS_SIZE);
+	return intel_dp_dpcd_read_wake(&intel_dp->aux,
+				       DP_LANE0_1_STATUS,
+				       link_status,
+				       DP_LINK_STATUS_SIZE) == DP_LINK_STATUS_SIZE;
 }
 
-#if 0
-static char	*voltage_names[] = {
-	"0.4V", "0.6V", "0.8V", "1.2V"
-};
-static char	*pre_emph_names[] = {
-	"0dB", "3.5dB", "6dB", "9.5dB"
-};
-static char	*link_train_names[] = {
-	"pattern 1", "pattern 2", "idle", "off"
-};
-#endif
-
 /*
  * These are source-specific values; current Intel hardware supports
  * a maximum voltage of 800mV and a maximum pre-emphasis of 6dB
@@ -2050,7 +2286,7 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 	unsigned long demph_reg_value, preemph_reg_value,
 		uniqtranscale_reg_value;
 	uint8_t train_set = intel_dp->train_set[0];
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 
 	switch (train_set & DP_TRAIN_PRE_EMPHASIS_MASK) {
@@ -2127,14 +2363,174 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 	}
 
 	mutex_lock(&dev_priv->dpio_lock);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port), 0x00000000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL4(port), demph_reg_value);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL2(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), 0x00000000);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW4(port), demph_reg_value);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW2(port),
 			 uniqtranscale_reg_value);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL3(port), 0x0C782040);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_STAGGER0(port), 0x00030000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CTL_OVER1(port), preemph_reg_value);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port), 0x80000000);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW3(port), 0x0C782040);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW11(port), 0x00030000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW9(port), preemph_reg_value);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), 0x80000000);
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	return 0;
+}
+
+static uint32_t intel_chv_signal_levels(struct intel_dp *intel_dp)
+{
+	struct drm_device *dev = intel_dp_to_dev(intel_dp);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_digital_port *dport = dp_to_dig_port(intel_dp);
+	struct intel_crtc *intel_crtc = to_intel_crtc(dport->base.base.crtc);
+	u32 deemph_reg_value, margin_reg_value, val;
+	uint8_t train_set = intel_dp->train_set[0];
+	enum dpio_channel ch = vlv_dport_to_channel(dport);
+	enum pipe pipe = intel_crtc->pipe;
+	int i;
+
+	switch (train_set & DP_TRAIN_PRE_EMPHASIS_MASK) {
+	case DP_TRAIN_PRE_EMPHASIS_0:
+		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
+		case DP_TRAIN_VOLTAGE_SWING_400:
+			deemph_reg_value = 128;
+			margin_reg_value = 52;
+			break;
+		case DP_TRAIN_VOLTAGE_SWING_600:
+			deemph_reg_value = 128;
+			margin_reg_value = 77;
+			break;
+		case DP_TRAIN_VOLTAGE_SWING_800:
+			deemph_reg_value = 128;
+			margin_reg_value = 102;
+			break;
+		case DP_TRAIN_VOLTAGE_SWING_1200:
+			deemph_reg_value = 128;
+			margin_reg_value = 154;
+			/* FIXME extra to set for 1200 */
+			break;
+		default:
+			return 0;
+		}
+		break;
+	case DP_TRAIN_PRE_EMPHASIS_3_5:
+		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
+		case DP_TRAIN_VOLTAGE_SWING_400:
+			deemph_reg_value = 85;
+			margin_reg_value = 78;
+			break;
+		case DP_TRAIN_VOLTAGE_SWING_600:
+			deemph_reg_value = 85;
+			margin_reg_value = 116;
+			break;
+		case DP_TRAIN_VOLTAGE_SWING_800:
+			deemph_reg_value = 85;
+			margin_reg_value = 154;
+			break;
+		default:
+			return 0;
+		}
+		break;
+	case DP_TRAIN_PRE_EMPHASIS_6:
+		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
+		case DP_TRAIN_VOLTAGE_SWING_400:
+			deemph_reg_value = 64;
+			margin_reg_value = 104;
+			break;
+		case DP_TRAIN_VOLTAGE_SWING_600:
+			deemph_reg_value = 64;
+			margin_reg_value = 154;
+			break;
+		default:
+			return 0;
+		}
+		break;
+	case DP_TRAIN_PRE_EMPHASIS_9_5:
+		switch (train_set & DP_TRAIN_VOLTAGE_SWING_MASK) {
+		case DP_TRAIN_VOLTAGE_SWING_400:
+			deemph_reg_value = 43;
+			margin_reg_value = 154;
+			break;
+		default:
+			return 0;
+		}
+		break;
+	default:
+		return 0;
+	}
+
+	mutex_lock(&dev_priv->dpio_lock);
+
+	/* Clear calc init */
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW10(ch));
+	val &= ~(DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW10(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW10(ch));
+	val &= ~(DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW10(ch), val);
+
+	/* Program swing deemph */
+	for (i = 0; i < 4; i++) {
+		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW4(ch, i));
+		val &= ~DPIO_SWING_DEEMPH9P5_MASK;
+		val |= deemph_reg_value << DPIO_SWING_DEEMPH9P5_SHIFT;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW4(ch, i), val);
+	}
+
+	/* Program swing margin */
+	for (i = 0; i < 4; i++) {
+		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW2(ch, i));
+		val &= ~DPIO_SWING_MARGIN_MASK;
+		val |= margin_reg_value << DPIO_SWING_MARGIN_SHIFT;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW2(ch, i), val);
+	}
+
+	/* Disable unique transition scale */
+	for (i = 0; i < 4; i++) {
+		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW3(ch, i));
+		val &= ~DPIO_TX_UNIQ_TRANS_SCALE_EN;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW3(ch, i), val);
+	}
+
+	if (((train_set & DP_TRAIN_PRE_EMPHASIS_MASK)
+			== DP_TRAIN_PRE_EMPHASIS_0) &&
+		((train_set & DP_TRAIN_VOLTAGE_SWING_MASK)
+			== DP_TRAIN_VOLTAGE_SWING_1200)) {
+
+		/*
+		 * The document said it needs to set bit 27 for ch0 and bit 26
+		 * for ch1. Might be a typo in the doc.
+		 * For now, for this unique transition scale selection, set bit
+		 * 27 for ch0 and ch1.
+		 */
+		for (i = 0; i < 4; i++) {
+			val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW3(ch, i));
+			val |= DPIO_TX_UNIQ_TRANS_SCALE_EN;
+			vlv_dpio_write(dev_priv, pipe, CHV_TX_DW3(ch, i), val);
+		}
+
+		for (i = 0; i < 4; i++) {
+			val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW2(ch, i));
+			val &= ~(0xff << DPIO_UNIQ_TRANS_SCALE_SHIFT);
+			val |= (0x9a << DPIO_UNIQ_TRANS_SCALE_SHIFT);
+			vlv_dpio_write(dev_priv, pipe, CHV_TX_DW2(ch, i), val);
+		}
+	}
+
+	/* Start swing calculation */
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW10(ch));
+	val |= DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW10(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW10(ch));
+	val |= DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW10(ch), val);
+
+	/* LRC Bypass */
+	val = vlv_dpio_read(dev_priv, pipe, CHV_CMN_DW30);
+	val |= DPIO_LRC_BYPASS;
+	vlv_dpio_write(dev_priv, pipe, CHV_CMN_DW30, val);
+
 	mutex_unlock(&dev_priv->dpio_lock);
 
 	return 0;
@@ -2354,6 +2750,9 @@ intel_dp_set_signal_levels(struct intel_dp *intel_dp, uint32_t *DP)
 	} else if (IS_HASWELL(dev)) {
 		signal_levels = intel_hsw_signal_levels(train_set);
 		mask = DDI_BUF_EMP_MASK;
+	} else if (IS_CHERRYVIEW(dev)) {
+		signal_levels = intel_chv_signal_levels(intel_dp);
+		mask = 0;
 	} else if (IS_VALLEYVIEW(dev)) {
 		signal_levels = intel_vlv_signal_levels(intel_dp);
 		mask = 0;
@@ -2464,8 +2863,8 @@ intel_dp_set_link_train(struct intel_dp *intel_dp,
 		len = intel_dp->lane_count + 1;
 	}
 
-	ret = intel_dp_aux_native_write(intel_dp, DP_TRAINING_PATTERN_SET,
-					buf, len);
+	ret = drm_dp_dpcd_write(&intel_dp->aux, DP_TRAINING_PATTERN_SET,
+				buf, len);
 
 	return ret == len;
 }
@@ -2494,9 +2893,8 @@ intel_dp_update_link_train(struct intel_dp *intel_dp, uint32_t *DP,
 	I915_WRITE(intel_dp->output_reg, *DP);
 	POSTING_READ(intel_dp->output_reg);
 
-	ret = intel_dp_aux_native_write(intel_dp, DP_TRAINING_LANE0_SET,
-					intel_dp->train_set,
-					intel_dp->lane_count);
+	ret = drm_dp_dpcd_write(&intel_dp->aux, DP_TRAINING_LANE0_SET,
+				intel_dp->train_set, intel_dp->lane_count);
 
 	return ret == intel_dp->lane_count;
 }
@@ -2552,11 +2950,11 @@ intel_dp_start_link_train(struct intel_dp *intel_dp)
 	link_config[1] = intel_dp->lane_count;
 	if (drm_dp_enhanced_frame_cap(intel_dp->dpcd))
 		link_config[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
-	intel_dp_aux_native_write(intel_dp, DP_LINK_BW_SET, link_config, 2);
+	drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_BW_SET, link_config, 2);
 
 	link_config[0] = 0;
 	link_config[1] = DP_SET_ANSI_8B10B;
-	intel_dp_aux_native_write(intel_dp, DP_DOWNSPREAD_CTRL, link_config, 2);
+	drm_dp_dpcd_write(&intel_dp->aux, DP_DOWNSPREAD_CTRL, link_config, 2);
 
 	DP |= DP_PORT_EN;
 
@@ -2629,10 +3027,15 @@ intel_dp_complete_link_train(struct intel_dp *intel_dp)
 	bool channel_eq = false;
 	int tries, cr_tries;
 	uint32_t DP = intel_dp->DP;
+	uint32_t training_pattern = DP_TRAINING_PATTERN_2;
+
+	/* Training Pattern 3 for HBR2 ot 1.2 devices that support it*/
+	if (intel_dp->link_bw == DP_LINK_BW_5_4 || intel_dp->use_tps3)
+		training_pattern = DP_TRAINING_PATTERN_3;
 
 	/* channel equalization */
 	if (!intel_dp_set_link_train(intel_dp, &DP,
-				     DP_TRAINING_PATTERN_2 |
+				     training_pattern |
 				     DP_LINK_SCRAMBLING_DISABLE)) {
 		DRM_ERROR("failed to start channel equalization\n");
 		return;
@@ -2646,7 +3049,6 @@ intel_dp_complete_link_train(struct intel_dp *intel_dp)
 
 		if (cr_tries > 5) {
 			DRM_ERROR("failed to train DP, aborting\n");
-			intel_dp_link_down(intel_dp);
 			break;
 		}
 
@@ -2660,7 +3062,7 @@ intel_dp_complete_link_train(struct intel_dp *intel_dp)
 		if (!drm_dp_clock_recovery_ok(link_status, intel_dp->lane_count)) {
 			intel_dp_start_link_train(intel_dp);
 			intel_dp_set_link_train(intel_dp, &DP,
-						DP_TRAINING_PATTERN_2 |
+						training_pattern |
 						DP_LINK_SCRAMBLING_DISABLE);
 			cr_tries++;
 			continue;
@@ -2676,7 +3078,7 @@ intel_dp_complete_link_train(struct intel_dp *intel_dp)
 			intel_dp_link_down(intel_dp);
 			intel_dp_start_link_train(intel_dp);
 			intel_dp_set_link_train(intel_dp, &DP,
-						DP_TRAINING_PATTERN_2 |
+						training_pattern |
 						DP_LINK_SCRAMBLING_DISABLE);
 			tries = 0;
 			cr_tries++;
@@ -2717,22 +3119,7 @@ intel_dp_link_down(struct intel_dp *intel_dp)
 		to_intel_crtc(intel_dig_port->base.base.crtc);
 	uint32_t DP = intel_dp->DP;
 
-	/*
-	 * DDI code has a strict mode set sequence and we should try to respect
-	 * it, otherwise we might hang the machine in many different ways. So we
-	 * really should be disabling the port only on a complete crtc_disable
-	 * sequence. This function is just called under two conditions on DDI
-	 * code:
-	 * - Link train failed while doing crtc_enable, and on this case we
-	 *   really should respect the mode set sequence and wait for a
-	 *   crtc_disable.
-	 * - Someone turned the monitor off and intel_dp_check_link_status
-	 *   called us. We don't need to disable the whole port on this case, so
-	 *   when someone turns the monitor on again,
-	 *   intel_ddi_prepare_link_retrain will take care of redoing the link
-	 *   train.
-	 */
-	if (HAS_DDI(dev))
+	if (WARN_ON(HAS_DDI(dev)))
 		return;
 
 	if (WARN_ON((I915_READ(intel_dp->output_reg) & DP_PORT_EN) == 0))
@@ -2749,9 +3136,6 @@ intel_dp_link_down(struct intel_dp *intel_dp)
 	}
 	POSTING_READ(intel_dp->output_reg);
 
-	/* We don't really know why we're doing this */
-	intel_wait_for_vblank(dev, intel_crtc->pipe);
-
 	if (HAS_PCH_IBX(dev) &&
 	    I915_READ(intel_dp->output_reg) & DP_PIPEB_SELECT) {
 		struct drm_crtc *crtc = intel_dig_port->base.base.crtc;
@@ -2795,8 +3179,8 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 
 	char dpcd_hex_dump[sizeof(intel_dp->dpcd) * 3];
 
-	if (intel_dp_aux_native_read_retry(intel_dp, 0x000, intel_dp->dpcd,
-					   sizeof(intel_dp->dpcd)) == 0)
+	if (intel_dp_dpcd_read_wake(&intel_dp->aux, 0x000, intel_dp->dpcd,
+				    sizeof(intel_dp->dpcd)) < 0)
 		return false; /* aux transfer failed */
 
 	hex_dump_to_buffer(intel_dp->dpcd, sizeof(intel_dp->dpcd),
@@ -2809,15 +3193,23 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 	/* Check if the panel supports PSR */
 	memset(intel_dp->psr_dpcd, 0, sizeof(intel_dp->psr_dpcd));
 	if (is_edp(intel_dp)) {
-		intel_dp_aux_native_read_retry(intel_dp, DP_PSR_SUPPORT,
-					       intel_dp->psr_dpcd,
-					       sizeof(intel_dp->psr_dpcd));
+		intel_dp_dpcd_read_wake(&intel_dp->aux, DP_PSR_SUPPORT,
+					intel_dp->psr_dpcd,
+					sizeof(intel_dp->psr_dpcd));
 		if (intel_dp->psr_dpcd[0] & DP_PSR_IS_SUPPORTED) {
 			dev_priv->psr.sink_support = true;
 			DRM_DEBUG_KMS("Detected EDP PSR Panel.\n");
 		}
 	}
 
+	/* Training Pattern 3 support */
+	if (intel_dp->dpcd[DP_DPCD_REV] >= 0x12 &&
+	    intel_dp->dpcd[DP_MAX_LANE_COUNT] & DP_TPS3_SUPPORTED) {
+		intel_dp->use_tps3 = true;
+		DRM_DEBUG_KMS("Displayport TPS3 supported");
+	} else
+		intel_dp->use_tps3 = false;
+
 	if (!(intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] &
 	      DP_DWN_STRM_PORT_PRESENT))
 		return true; /* native DP sink */
@@ -2825,9 +3217,9 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 	if (intel_dp->dpcd[DP_DPCD_REV] == 0x10)
 		return true; /* no per-port downstream info */
 
-	if (intel_dp_aux_native_read_retry(intel_dp, DP_DOWNSTREAM_PORT_0,
-					   intel_dp->downstream_ports,
-					   DP_MAX_DOWNSTREAM_PORTS) == 0)
+	if (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_DOWNSTREAM_PORT_0,
+				    intel_dp->downstream_ports,
+				    DP_MAX_DOWNSTREAM_PORTS) < 0)
 		return false; /* downstream port status fetch failed */
 
 	return true;
@@ -2841,38 +3233,61 @@ intel_dp_probe_oui(struct intel_dp *intel_dp)
 	if (!(intel_dp->dpcd[DP_DOWN_STREAM_PORT_COUNT] & DP_OUI_SUPPORT))
 		return;
 
-	ironlake_edp_panel_vdd_on(intel_dp);
+	intel_edp_panel_vdd_on(intel_dp);
 
-	if (intel_dp_aux_native_read_retry(intel_dp, DP_SINK_OUI, buf, 3))
+	if (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_SINK_OUI, buf, 3) == 3)
 		DRM_DEBUG_KMS("Sink OUI: %02hx%02hx%02hx\n",
 			      buf[0], buf[1], buf[2]);
 
-	if (intel_dp_aux_native_read_retry(intel_dp, DP_BRANCH_OUI, buf, 3))
+	if (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_BRANCH_OUI, buf, 3) == 3)
 		DRM_DEBUG_KMS("Branch OUI: %02hx%02hx%02hx\n",
 			      buf[0], buf[1], buf[2]);
 
-	ironlake_edp_panel_vdd_off(intel_dp, false);
+	edp_panel_vdd_off(intel_dp, false);
 }
 
-static bool
-intel_dp_get_sink_irq(struct intel_dp *intel_dp, u8 *sink_irq_vector)
+int intel_dp_sink_crc(struct intel_dp *intel_dp, u8 *crc)
 {
-	int ret;
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct intel_crtc *intel_crtc =
+		to_intel_crtc(intel_dig_port->base.base.crtc);
+	u8 buf[1];
 
-	ret = intel_dp_aux_native_read_retry(intel_dp,
-					     DP_DEVICE_SERVICE_IRQ_VECTOR,
-					     sink_irq_vector, 1);
-	if (!ret)
-		return false;
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_SINK_MISC, buf) < 0)
+		return -EAGAIN;
 
-	return true;
+	if (!(buf[0] & DP_TEST_CRC_SUPPORTED))
+		return -ENOTTY;
+
+	if (drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_SINK,
+			       DP_TEST_SINK_START) < 0)
+		return -EAGAIN;
+
+	/* Wait 2 vblanks to be sure we will have the correct CRC value */
+	intel_wait_for_vblank(dev, intel_crtc->pipe);
+	intel_wait_for_vblank(dev, intel_crtc->pipe);
+
+	if (drm_dp_dpcd_read(&intel_dp->aux, DP_TEST_CRC_R_CR, crc, 6) < 0)
+		return -EAGAIN;
+
+	drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_SINK, 0);
+	return 0;
+}
+
+static bool
+intel_dp_get_sink_irq(struct intel_dp *intel_dp, u8 *sink_irq_vector)
+{
+	return intel_dp_dpcd_read_wake(&intel_dp->aux,
+				       DP_DEVICE_SERVICE_IRQ_VECTOR,
+				       sink_irq_vector, 1) == 1;
 }
 
 static void
 intel_dp_handle_test_request(struct intel_dp *intel_dp)
 {
 	/* NAK by default */
-	intel_dp_aux_native_write_1(intel_dp, DP_TEST_RESPONSE, DP_TEST_NAK);
+	drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_RESPONSE, DP_TEST_NAK);
 }
 
 /*
@@ -2891,6 +3306,7 @@ intel_dp_check_link_status(struct intel_dp *intel_dp)
 	u8 sink_irq_vector;
 	u8 link_status[DP_LINK_STATUS_SIZE];
 
+	/* FIXME: This access isn't protected by any locks. */
 	if (!intel_encoder->connectors_active)
 		return;
 
@@ -2899,13 +3315,11 @@ intel_dp_check_link_status(struct intel_dp *intel_dp)
 
 	/* Try to read receiver status if the link appears to be up */
 	if (!intel_dp_get_link_status(intel_dp, link_status)) {
-		intel_dp_link_down(intel_dp);
 		return;
 	}
 
 	/* Now read the DPCD to see if it's actually running */
 	if (!intel_dp_get_dpcd(intel_dp)) {
-		intel_dp_link_down(intel_dp);
 		return;
 	}
 
@@ -2913,9 +3327,9 @@ intel_dp_check_link_status(struct intel_dp *intel_dp)
 	if (intel_dp->dpcd[DP_DPCD_REV] >= 0x11 &&
 	    intel_dp_get_sink_irq(intel_dp, &sink_irq_vector)) {
 		/* Clear interrupt source */
-		intel_dp_aux_native_write_1(intel_dp,
-					    DP_DEVICE_SERVICE_IRQ_VECTOR,
-					    sink_irq_vector);
+		drm_dp_dpcd_writeb(&intel_dp->aux,
+				   DP_DEVICE_SERVICE_IRQ_VECTOR,
+				   sink_irq_vector);
 
 		if (sink_irq_vector & DP_AUTOMATED_TEST_REQUEST)
 			intel_dp_handle_test_request(intel_dp);
@@ -2925,7 +3339,7 @@ intel_dp_check_link_status(struct intel_dp *intel_dp)
 
 	if (!drm_dp_channel_eq_ok(link_status, intel_dp->lane_count)) {
 		DRM_DEBUG_KMS("%s: channel EQ not ok, retraining\n",
-			      drm_get_encoder_name(&intel_encoder->base));
+			      intel_encoder->base.name);
 		intel_dp_start_link_train(intel_dp);
 		intel_dp_complete_link_train(intel_dp);
 		intel_dp_stop_link_train(intel_dp);
@@ -2950,15 +3364,17 @@ intel_dp_detect_dpcd(struct intel_dp *intel_dp)
 	if (intel_dp->dpcd[DP_DPCD_REV] >= 0x11 &&
 	    intel_dp->downstream_ports[0] & DP_DS_PORT_HPD) {
 		uint8_t reg;
-		if (!intel_dp_aux_native_read_retry(intel_dp, DP_SINK_COUNT,
-						    &reg, 1))
+
+		if (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_SINK_COUNT,
+					    &reg, 1) < 0)
 			return connector_status_unknown;
+
 		return DP_GET_SINK_COUNT(reg) ? connector_status_connected
 					      : connector_status_disconnected;
 	}
 
 	/* If no HPD, poke DDC gently */
-	if (drm_probe_ddc(&intel_dp->adapter))
+	if (drm_probe_ddc(&intel_dp->aux.ddc))
 		return connector_status_connected;
 
 	/* Well we tried, say unknown for unreliable port types */
@@ -3020,18 +3436,34 @@ g4x_dp_detect(struct intel_dp *intel_dp)
 		return status;
 	}
 
-	switch (intel_dig_port->port) {
-	case PORT_B:
-		bit = PORTB_HOTPLUG_LIVE_STATUS;
-		break;
-	case PORT_C:
-		bit = PORTC_HOTPLUG_LIVE_STATUS;
-		break;
-	case PORT_D:
-		bit = PORTD_HOTPLUG_LIVE_STATUS;
-		break;
-	default:
-		return connector_status_unknown;
+	if (IS_VALLEYVIEW(dev)) {
+		switch (intel_dig_port->port) {
+		case PORT_B:
+			bit = PORTB_HOTPLUG_LIVE_STATUS_VLV;
+			break;
+		case PORT_C:
+			bit = PORTC_HOTPLUG_LIVE_STATUS_VLV;
+			break;
+		case PORT_D:
+			bit = PORTD_HOTPLUG_LIVE_STATUS_VLV;
+			break;
+		default:
+			return connector_status_unknown;
+		}
+	} else {
+		switch (intel_dig_port->port) {
+		case PORT_B:
+			bit = PORTB_HOTPLUG_LIVE_STATUS_G4X;
+			break;
+		case PORT_C:
+			bit = PORTC_HOTPLUG_LIVE_STATUS_G4X;
+			break;
+		case PORT_D:
+			bit = PORTD_HOTPLUG_LIVE_STATUS_G4X;
+			break;
+		default:
+			return connector_status_unknown;
+		}
 	}
 
 	if ((I915_READ(PORT_HOTPLUG_STAT) & bit) == 0)
@@ -3082,11 +3514,18 @@ intel_dp_detect(struct drm_connector *connector, bool force)
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct intel_encoder *intel_encoder = &intel_dig_port->base;
 	struct drm_device *dev = connector->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum drm_connector_status status;
+	enum intel_display_power_domain power_domain;
 	struct edid *edid = NULL;
 
+	intel_runtime_pm_get(dev_priv);
+
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_get(dev_priv, power_domain);
+
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector));
+		      connector->base.id, connector->name);
 
 	intel_dp->has_audio = false;
 
@@ -3096,14 +3535,14 @@ intel_dp_detect(struct drm_connector *connector, bool force)
 		status = g4x_dp_detect(intel_dp);
 
 	if (status != connector_status_connected)
-		return status;
+		goto out;
 
 	intel_dp_probe_oui(intel_dp);
 
 	if (intel_dp->force_audio != HDMI_AUDIO_AUTO) {
 		intel_dp->has_audio = (intel_dp->force_audio == HDMI_AUDIO_ON);
 	} else {
-		edid = intel_dp_get_edid(connector, &intel_dp->adapter);
+		edid = intel_dp_get_edid(connector, &intel_dp->aux.ddc);
 		if (edid) {
 			intel_dp->has_audio = drm_detect_monitor_audio(edid);
 			kfree(edid);
@@ -3112,20 +3551,35 @@ intel_dp_detect(struct drm_connector *connector, bool force)
 
 	if (intel_encoder->type != INTEL_OUTPUT_EDP)
 		intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
-	return connector_status_connected;
+	status = connector_status_connected;
+
+out:
+	intel_display_power_put(dev_priv, power_domain);
+
+	intel_runtime_pm_put(dev_priv);
+
+	return status;
 }
 
 static int intel_dp_get_modes(struct drm_connector *connector)
 {
 	struct intel_dp *intel_dp = intel_attached_dp(connector);
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *intel_encoder = &intel_dig_port->base;
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 	struct drm_device *dev = connector->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum intel_display_power_domain power_domain;
 	int ret;
 
 	/* We should parse the EDID data and find out if it has an audio sink
 	 */
 
-	ret = intel_dp_get_edid_modes(connector, &intel_dp->adapter);
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_get(dev_priv, power_domain);
+
+	ret = intel_dp_get_edid_modes(connector, &intel_dp->aux.ddc);
+	intel_display_power_put(dev_priv, power_domain);
 	if (ret)
 		return ret;
 
@@ -3146,15 +3600,25 @@ static bool
 intel_dp_detect_audio(struct drm_connector *connector)
 {
 	struct intel_dp *intel_dp = intel_attached_dp(connector);
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct intel_encoder *intel_encoder = &intel_dig_port->base;
+	struct drm_device *dev = connector->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum intel_display_power_domain power_domain;
 	struct edid *edid;
 	bool has_audio = false;
 
-	edid = intel_dp_get_edid(connector, &intel_dp->adapter);
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_get(dev_priv, power_domain);
+
+	edid = intel_dp_get_edid(connector, &intel_dp->aux.ddc);
 	if (edid) {
 		has_audio = drm_detect_monitor_audio(edid);
 		kfree(edid);
 	}
 
+	intel_display_power_put(dev_priv, power_domain);
+
 	return has_audio;
 }
 
@@ -3269,13 +3733,17 @@ void intel_dp_encoder_destroy(struct drm_encoder *encoder)
 	struct intel_dp *intel_dp = &intel_dig_port->dp;
 	struct drm_device *dev = intel_dp_to_dev(intel_dp);
 
-	i2c_del_adapter(&intel_dp->adapter);
+	drm_dp_aux_unregister(&intel_dp->aux);
 	drm_encoder_cleanup(encoder);
 	if (is_edp(intel_dp)) {
 		cancel_delayed_work_sync(&intel_dp->panel_vdd_work);
-		mutex_lock(&dev->mode_config.mutex);
-		ironlake_panel_vdd_off_sync(intel_dp);
-		mutex_unlock(&dev->mode_config.mutex);
+		drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+		edp_panel_vdd_off_sync(intel_dp);
+		drm_modeset_unlock(&dev->mode_config.connection_mutex);
+		if (intel_dp->edp_notifier.notifier_call) {
+			unregister_reboot_notifier(&intel_dp->edp_notifier);
+			intel_dp->edp_notifier.notifier_call = NULL;
+		}
 	}
 	kfree(intel_dig_port);
 }
@@ -3373,6 +3841,13 @@ intel_dp_add_properties(struct intel_dp *intel_dp, struct drm_connector *connect
 	}
 }
 
+static void intel_dp_init_panel_power_timestamps(struct intel_dp *intel_dp)
+{
+	intel_dp->last_power_cycle = jiffies;
+	intel_dp->last_power_on = jiffies;
+	intel_dp->last_backlight_off = jiffies;
+}
+
 static void
 intel_dp_init_panel_power_sequencer(struct drm_device *dev,
 				    struct intel_dp *intel_dp,
@@ -3495,10 +3970,17 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 		pp_div_reg = VLV_PIPE_PP_DIVISOR(pipe);
 	}
 
-	/* And finally store the new values in the power sequencer. */
+	/*
+	 * And finally store the new values in the power sequencer. The
+	 * backlight delays are set to 1 because we do manual waits on them. For
+	 * T8, even BSpec recommends doing it. For T9, if we don't do this,
+	 * we'll end up waiting for the backlight off delay twice: once when we
+	 * do the manual sleep, and once when we disable the panel and wait for
+	 * the PP_STATUS bit to become zero.
+	 */
 	pp_on = (seq->t1_t3 << PANEL_POWER_UP_DELAY_SHIFT) |
-		(seq->t8 << PANEL_LIGHT_ON_DELAY_SHIFT);
-	pp_off = (seq->t9 << PANEL_LIGHT_OFF_DELAY_SHIFT) |
+		(1 << PANEL_LIGHT_ON_DELAY_SHIFT);
+	pp_off = (1 << PANEL_LIGHT_OFF_DELAY_SHIFT) |
 		 (seq->t10 << PANEL_POWER_DOWN_DELAY_SHIFT);
 	/* Compute the divisor for the pp clock, simply match the Bspec
 	 * formula. */
@@ -3532,28 +4014,162 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 		      I915_READ(pp_div_reg));
 }
 
+void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_encoder *encoder;
+	struct intel_dp *intel_dp = NULL;
+	struct intel_crtc_config *config = NULL;
+	struct intel_crtc *intel_crtc = NULL;
+	struct intel_connector *intel_connector = dev_priv->drrs.connector;
+	u32 reg, val;
+	enum edp_drrs_refresh_rate_type index = DRRS_HIGH_RR;
+
+	if (refresh_rate <= 0) {
+		DRM_DEBUG_KMS("Refresh rate should be positive non-zero.\n");
+		return;
+	}
+
+	if (intel_connector == NULL) {
+		DRM_DEBUG_KMS("DRRS supported for eDP only.\n");
+		return;
+	}
+
+	if (INTEL_INFO(dev)->gen < 8 && intel_edp_is_psr_enabled(dev)) {
+		DRM_DEBUG_KMS("DRRS is disabled as PSR is enabled\n");
+		return;
+	}
+
+	encoder = intel_attached_encoder(&intel_connector->base);
+	intel_dp = enc_to_intel_dp(&encoder->base);
+	intel_crtc = encoder->new_crtc;
+
+	if (!intel_crtc) {
+		DRM_DEBUG_KMS("DRRS: intel_crtc not initialized\n");
+		return;
+	}
+
+	config = &intel_crtc->config;
+
+	if (intel_dp->drrs_state.type < SEAMLESS_DRRS_SUPPORT) {
+		DRM_DEBUG_KMS("Only Seamless DRRS supported.\n");
+		return;
+	}
+
+	if (intel_connector->panel.downclock_mode->vrefresh == refresh_rate)
+		index = DRRS_LOW_RR;
+
+	if (index == intel_dp->drrs_state.refresh_rate_type) {
+		DRM_DEBUG_KMS(
+			"DRRS requested for previously set RR...ignoring\n");
+		return;
+	}
+
+	if (!intel_crtc->active) {
+		DRM_DEBUG_KMS("eDP encoder disabled. CRTC not Active\n");
+		return;
+	}
+
+	if (INTEL_INFO(dev)->gen > 6 && INTEL_INFO(dev)->gen < 8) {
+		reg = PIPECONF(intel_crtc->config.cpu_transcoder);
+		val = I915_READ(reg);
+		if (index > DRRS_HIGH_RR) {
+			val |= PIPECONF_EDP_RR_MODE_SWITCH;
+			intel_dp_set_m2_n2(intel_crtc, &config->dp_m2_n2);
+		} else {
+			val &= ~PIPECONF_EDP_RR_MODE_SWITCH;
+		}
+		I915_WRITE(reg, val);
+	}
+
+	/*
+	 * mutex taken to ensure that there is no race between differnt
+	 * drrs calls trying to update refresh rate. This scenario may occur
+	 * in future when idleness detection based DRRS in kernel and
+	 * possible calls from user space to set differnt RR are made.
+	 */
+
+	mutex_lock(&intel_dp->drrs_state.mutex);
+
+	intel_dp->drrs_state.refresh_rate_type = index;
+
+	mutex_unlock(&intel_dp->drrs_state.mutex);
+
+	DRM_DEBUG_KMS("eDP Refresh Rate set to : %dHz\n", refresh_rate);
+}
+
+static struct drm_display_mode *
+intel_dp_drrs_init(struct intel_digital_port *intel_dig_port,
+			struct intel_connector *intel_connector,
+			struct drm_display_mode *fixed_mode)
+{
+	struct drm_connector *connector = &intel_connector->base;
+	struct intel_dp *intel_dp = &intel_dig_port->dp;
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_display_mode *downclock_mode = NULL;
+
+	if (INTEL_INFO(dev)->gen <= 6) {
+		DRM_DEBUG_KMS("DRRS supported for Gen7 and above\n");
+		return NULL;
+	}
+
+	if (dev_priv->vbt.drrs_type != SEAMLESS_DRRS_SUPPORT) {
+		DRM_INFO("VBT doesn't support DRRS\n");
+		return NULL;
+	}
+
+	downclock_mode = intel_find_panel_downclock
+					(dev, fixed_mode, connector);
+
+	if (!downclock_mode) {
+		DRM_INFO("DRRS not supported\n");
+		return NULL;
+	}
+
+	dev_priv->drrs.connector = intel_connector;
+
+	mutex_init(&intel_dp->drrs_state.mutex);
+
+	intel_dp->drrs_state.type = dev_priv->vbt.drrs_type;
+
+	intel_dp->drrs_state.refresh_rate_type = DRRS_HIGH_RR;
+	DRM_INFO("seamless DRRS supported for eDP panel.\n");
+	return downclock_mode;
+}
+
 static bool intel_edp_init_connector(struct intel_dp *intel_dp,
-				     struct intel_connector *intel_connector)
+				     struct intel_connector *intel_connector,
+				     struct edp_power_seq *power_seq)
 {
 	struct drm_connector *connector = &intel_connector->base;
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	struct drm_device *dev = intel_dig_port->base.base.dev;
+	struct intel_encoder *intel_encoder = &intel_dig_port->base;
+	struct drm_device *dev = intel_encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_display_mode *fixed_mode = NULL;
-	struct edp_power_seq power_seq = { 0 };
+	struct drm_display_mode *downclock_mode = NULL;
 	bool has_dpcd;
 	struct drm_display_mode *scan;
 	struct edid *edid;
 
+	intel_dp->drrs_state.type = DRRS_NOT_SUPPORTED;
+
 	if (!is_edp(intel_dp))
 		return true;
 
-	intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
+	/* The VDD bit needs a power domain reference, so if the bit is already
+	 * enabled when we boot, grab this reference. */
+	if (edp_have_panel_vdd(intel_dp)) {
+		enum intel_display_power_domain power_domain;
+		power_domain = intel_display_port_power_domain(intel_encoder);
+		intel_display_power_get(dev_priv, power_domain);
+	}
 
 	/* Cache DPCD and EDID for edp. */
-	ironlake_edp_panel_vdd_on(intel_dp);
+	intel_edp_panel_vdd_on(intel_dp);
 	has_dpcd = intel_dp_get_dpcd(intel_dp);
-	ironlake_edp_panel_vdd_off(intel_dp, false);
+	edp_panel_vdd_off(intel_dp, false);
 
 	if (has_dpcd) {
 		if (intel_dp->dpcd[DP_DPCD_REV] >= 0x11)
@@ -3567,10 +4183,10 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	}
 
 	/* We now know it's not a ghost, init power sequence regs. */
-	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp,
-						      &power_seq);
+	intel_dp_init_panel_power_sequencer_registers(dev, intel_dp, power_seq);
 
-	edid = drm_get_edid(connector, &intel_dp->adapter);
+	mutex_lock(&dev->mode_config.mutex);
+	edid = drm_get_edid(connector, &intel_dp->aux.ddc);
 	if (edid) {
 		if (drm_add_edid_modes(connector, edid)) {
 			drm_mode_connector_update_edid_property(connector,
@@ -3589,6 +4205,9 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	list_for_each_entry(scan, &connector->probed_modes, head) {
 		if ((scan->type & DRM_MODE_TYPE_PREFERRED)) {
 			fixed_mode = drm_mode_duplicate(dev, scan);
+			downclock_mode = intel_dp_drrs_init(
+						intel_dig_port,
+						intel_connector, fixed_mode);
 			break;
 		}
 	}
@@ -3600,8 +4219,14 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 		if (fixed_mode)
 			fixed_mode->type |= DRM_MODE_TYPE_PREFERRED;
 	}
+	mutex_unlock(&dev->mode_config.mutex);
 
-	intel_panel_init(&intel_connector->panel, fixed_mode);
+	if (IS_VALLEYVIEW(dev)) {
+		intel_dp->edp_notifier.notifier_call = edp_notify_handler;
+		register_reboot_notifier(&intel_dp->edp_notifier);
+	}
+
+	intel_panel_init(&intel_connector->panel, fixed_mode, downclock_mode);
 	intel_panel_setup_backlight(connector);
 
 	return true;
@@ -3617,8 +4242,20 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	struct drm_device *dev = intel_encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	enum port port = intel_dig_port->port;
-	const char *name = NULL;
-	int type, error;
+	struct edp_power_seq power_seq = { 0 };
+	int type;
+
+	/* intel_dp vfuncs */
+	if (IS_VALLEYVIEW(dev))
+		intel_dp->get_aux_clock_divider = vlv_get_aux_clock_divider;
+	else if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		intel_dp->get_aux_clock_divider = hsw_get_aux_clock_divider;
+	else if (HAS_PCH_SPLIT(dev))
+		intel_dp->get_aux_clock_divider = ilk_get_aux_clock_divider;
+	else
+		intel_dp->get_aux_clock_divider = i9xx_get_aux_clock_divider;
+
+	intel_dp->get_aux_send_ctl = i9xx_get_aux_send_ctl;
 
 	/* Preserve the current hw state. */
 	intel_dp->DP = I915_READ(intel_dp->output_reg);
@@ -3648,7 +4285,7 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	connector->doublescan_allowed = 0;
 
 	INIT_DELAYED_WORK(&intel_dp->panel_vdd_work,
-			  ironlake_panel_vdd_work);
+			  edp_panel_vdd_work);
 
 	intel_connector_attach_encoder(intel_connector, intel_encoder);
 	drm_sysfs_connector_add(connector);
@@ -3657,62 +4294,42 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 		intel_connector->get_hw_state = intel_ddi_connector_get_hw_state;
 	else
 		intel_connector->get_hw_state = intel_connector_get_hw_state;
+	intel_connector->unregister = intel_dp_connector_unregister;
 
-	intel_dp->aux_ch_ctl_reg = intel_dp->output_reg + 0x10;
-	if (HAS_DDI(dev)) {
-		switch (intel_dig_port->port) {
-		case PORT_A:
-			intel_dp->aux_ch_ctl_reg = DPA_AUX_CH_CTL;
-			break;
-		case PORT_B:
-			intel_dp->aux_ch_ctl_reg = PCH_DPB_AUX_CH_CTL;
-			break;
-		case PORT_C:
-			intel_dp->aux_ch_ctl_reg = PCH_DPC_AUX_CH_CTL;
-			break;
-		case PORT_D:
-			intel_dp->aux_ch_ctl_reg = PCH_DPD_AUX_CH_CTL;
-			break;
-		default:
-			BUG();
-		}
-	}
-
-	/* Set up the DDC bus. */
+	/* Set up the hotplug pin. */
 	switch (port) {
 	case PORT_A:
 		intel_encoder->hpd_pin = HPD_PORT_A;
-		name = "DPDDC-A";
 		break;
 	case PORT_B:
 		intel_encoder->hpd_pin = HPD_PORT_B;
-		name = "DPDDC-B";
 		break;
 	case PORT_C:
 		intel_encoder->hpd_pin = HPD_PORT_C;
-		name = "DPDDC-C";
 		break;
 	case PORT_D:
 		intel_encoder->hpd_pin = HPD_PORT_D;
-		name = "DPDDC-D";
 		break;
 	default:
 		BUG();
 	}
 
-	error = intel_dp_i2c_init(intel_dp, intel_connector, name);
-	WARN(error, "intel_dp_i2c_init failed with error %d for port %c\n",
-	     error, port_name(port));
+	if (is_edp(intel_dp)) {
+		intel_dp_init_panel_power_timestamps(intel_dp);
+		intel_dp_init_panel_power_sequencer(dev, intel_dp, &power_seq);
+	}
+
+	intel_dp_aux_init(intel_dp, intel_connector);
 
 	intel_dp->psr_setup_done = false;
 
-	if (!intel_edp_init_connector(intel_dp, intel_connector)) {
-		i2c_del_adapter(&intel_dp->adapter);
+	if (!intel_edp_init_connector(intel_dp, intel_connector, &power_seq)) {
+		drm_dp_aux_unregister(&intel_dp->aux);
 		if (is_edp(intel_dp)) {
 			cancel_delayed_work_sync(&intel_dp->panel_vdd_work);
-			mutex_lock(&dev->mode_config.mutex);
-			ironlake_panel_vdd_off_sync(intel_dp);
-			mutex_unlock(&dev->mode_config.mutex);
+			drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+			edp_panel_vdd_off_sync(intel_dp);
+			drm_modeset_unlock(&dev->mode_config.connection_mutex);
 		}
 		drm_sysfs_connector_remove(connector);
 		drm_connector_cleanup(connector);
@@ -3758,26 +4375,37 @@ intel_dp_init(struct drm_device *dev, int output_reg, enum port port)
 			 DRM_MODE_ENCODER_TMDS);
 
 	intel_encoder->compute_config = intel_dp_compute_config;
-	intel_encoder->mode_set = intel_dp_mode_set;
 	intel_encoder->disable = intel_disable_dp;
-	intel_encoder->post_disable = intel_post_disable_dp;
 	intel_encoder->get_hw_state = intel_dp_get_hw_state;
 	intel_encoder->get_config = intel_dp_get_config;
-	if (IS_VALLEYVIEW(dev)) {
+	if (IS_CHERRYVIEW(dev)) {
+		intel_encoder->pre_enable = chv_pre_enable_dp;
+		intel_encoder->enable = vlv_enable_dp;
+		intel_encoder->post_disable = chv_post_disable_dp;
+	} else if (IS_VALLEYVIEW(dev)) {
 		intel_encoder->pre_pll_enable = vlv_dp_pre_pll_enable;
 		intel_encoder->pre_enable = vlv_pre_enable_dp;
 		intel_encoder->enable = vlv_enable_dp;
+		intel_encoder->post_disable = vlv_post_disable_dp;
 	} else {
 		intel_encoder->pre_enable = g4x_pre_enable_dp;
 		intel_encoder->enable = g4x_enable_dp;
+		intel_encoder->post_disable = g4x_post_disable_dp;
 	}
 
 	intel_dig_port->port = port;
 	intel_dig_port->dp.output_reg = output_reg;
 
 	intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
-	intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
-	intel_encoder->cloneable = false;
+	if (IS_CHERRYVIEW(dev)) {
+		if (port == PORT_D)
+			intel_encoder->crtc_mask = 1 << 2;
+		else
+			intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
+	} else {
+		intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
+	}
+	intel_encoder->cloneable = 0;
 	intel_encoder->hot_plug = intel_dp_hot_plug;
 
 	if (!intel_dp_init_connector(intel_dig_port, intel_connector)) {
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 79f91f26e28..f67340ed2c1 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -65,8 +65,8 @@
 #define wait_for_atomic_us(COND, US) _wait_for((COND), \
 					       DIV_ROUND_UP((US), 1000), 0)
 
-#define KHz(x) (1000*x)
-#define MHz(x) KHz(1000*x)
+#define KHz(x) (1000 * (x))
+#define MHz(x) KHz(1000 * (x))
 
 /*
  * Display related stuff
@@ -78,6 +78,12 @@
 #define MAX_OUTPUTS 6
 /* maximum connectors per crtcs in the mode set */
 
+/* Maximum cursor sizes */
+#define GEN2_CURSOR_WIDTH 64
+#define GEN2_CURSOR_HEIGHT 64
+#define MAX_CURSOR_WIDTH 256
+#define MAX_CURSOR_HEIGHT 256
+
 #define INTEL_I2C_BUS_DVO 1
 #define INTEL_I2C_BUS_SDVO 2
 
@@ -100,8 +106,8 @@
 #define INTEL_DVO_CHIP_TMDS 2
 #define INTEL_DVO_CHIP_TVOUT 4
 
-#define INTEL_DSI_COMMAND_MODE	0
-#define INTEL_DSI_VIDEO_MODE	1
+#define INTEL_DSI_VIDEO_MODE	0
+#define INTEL_DSI_COMMAND_MODE	1
 
 struct intel_framebuffer {
 	struct drm_framebuffer base;
@@ -110,9 +116,10 @@ struct intel_framebuffer {
 
 struct intel_fbdev {
 	struct drm_fb_helper helper;
-	struct intel_framebuffer ifb;
+	struct intel_framebuffer *fb;
 	struct list_head fbdev_list;
 	struct drm_display_mode *our_mode;
+	int preferred_bpp;
 };
 
 struct intel_encoder {
@@ -124,11 +131,7 @@ struct intel_encoder {
 	struct intel_crtc *new_crtc;
 
 	int type;
-	/*
-	 * Intel hw has only one MUX where encoders could be clone, hence a
-	 * simple flag is enough to compute the possible_clones mask.
-	 */
-	bool cloneable;
+	unsigned int cloneable;
 	bool connectors_active;
 	void (*hot_plug)(struct intel_encoder *);
 	bool (*compute_config)(struct intel_encoder *,
@@ -155,7 +158,19 @@ struct intel_encoder {
 
 struct intel_panel {
 	struct drm_display_mode *fixed_mode;
+	struct drm_display_mode *downclock_mode;
 	int fitting_mode;
+
+	/* backlight */
+	struct {
+		bool present;
+		u32 level;
+		u32 max;
+		bool enabled;
+		bool combination_mode;	/* gen 2/4 only */
+		bool active_low_pwm;
+		struct backlight_device *device;
+	} backlight;
 };
 
 struct intel_connector {
@@ -175,6 +190,14 @@ struct intel_connector {
 	 * and active (i.e. dpms ON state). */
 	bool (*get_hw_state)(struct intel_connector *);
 
+	/*
+	 * Removes all interfaces through which the connector is accessible
+	 * - like sysfs, debugfs entries -, so that no new operations can be
+	 * started on the connector. Also makes sure all currently pending
+	 * operations finish before returing.
+	 */
+	void (*unregister)(struct intel_connector *);
+
 	/* Panel info for eDP and LVDS */
 	struct intel_panel panel;
 
@@ -198,6 +221,12 @@ typedef struct dpll {
 	int	p;
 } intel_clock_t;
 
+struct intel_plane_config {
+	bool tiled;
+	int size;
+	u32 base;
+};
+
 struct intel_crtc_config {
 	/**
 	 * quirks - bitfield with hw state readout quirks
@@ -207,7 +236,8 @@ struct intel_crtc_config {
 	 * tracked with quirk flags so that fastboot and state checker can act
 	 * accordingly.
 	 */
-#define PIPE_CONFIG_QUIRK_MODE_SYNC_FLAGS (1<<0) /* unreliable sync mode.flags */
+#define PIPE_CONFIG_QUIRK_MODE_SYNC_FLAGS	(1<<0) /* unreliable sync mode.flags */
+#define PIPE_CONFIG_QUIRK_INHERITED_MODE	(1<<1) /* mode inherited from firmware */
 	unsigned long quirks;
 
 	/* User requested mode, only valid as a starting point to
@@ -243,6 +273,13 @@ struct intel_crtc_config {
 	 * accordingly. */
 	bool has_dp_encoder;
 
+	/* Whether we should send NULL infoframes. Required for audio. */
+	bool has_hdmi_sink;
+
+	/* Audio enabled on this pipe. Only valid if either has_hdmi_sink or
+	 * has_dp_encoder is set. */
+	bool has_audio;
+
 	/*
 	 * Enable dithering, used when the selected pipe bpp doesn't match the
 	 * plane bpp.
@@ -276,6 +313,9 @@ struct intel_crtc_config {
 	int pipe_bpp;
 	struct intel_link_m_n dp_m_n;
 
+	/* m2_n2 for eDP downclock */
+	struct intel_link_m_n dp_m2_n2;
+
 	/*
 	 * Frequence the dpll for the port should run at. Differs from the
 	 * adjusted dotclock e.g. for DP or 12bpc hdmi mode. This is also
@@ -313,6 +353,9 @@ struct intel_pipe_wm {
 	struct intel_wm_level wm[5];
 	uint32_t linetime;
 	bool fbc_wm_enabled;
+	bool pipe_enabled;
+	bool sprites_enabled;
+	bool sprites_scaled;
 };
 
 struct intel_crtc {
@@ -327,7 +370,6 @@ struct intel_crtc {
 	 */
 	bool active;
 	unsigned long enabled_power_domains;
-	bool eld_vld;
 	bool primary_enabled; /* is the primary plane (partially) visible? */
 	bool lowfreq_avail;
 	struct intel_overlay *overlay;
@@ -344,9 +386,13 @@ struct intel_crtc {
 	uint32_t cursor_addr;
 	int16_t cursor_x, cursor_y;
 	int16_t cursor_width, cursor_height;
-	bool cursor_visible;
+	uint32_t cursor_cntl;
+	uint32_t cursor_base;
 
+	struct intel_plane_config plane_config;
 	struct intel_crtc_config config;
+	struct intel_crtc_config *new_config;
+	bool new_enabled;
 
 	uint32_t ddi_pll_sel;
 
@@ -362,6 +408,10 @@ struct intel_crtc {
 		/* watermarks currently being used  */
 		struct intel_pipe_wm active;
 	} wm;
+
+	wait_queue_head_t vbl_wait;
+
+	int scanline_offset;
 };
 
 struct intel_plane_wm_parameters {
@@ -443,13 +493,25 @@ struct intel_hdmi {
 	bool rgb_quant_range_selectable;
 	void (*write_infoframe)(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len);
+				const void *frame, ssize_t len);
 	void (*set_infoframes)(struct drm_encoder *encoder,
+			       bool enable,
 			       struct drm_display_mode *adjusted_mode);
 };
 
 #define DP_MAX_DOWNSTREAM_PORTS		0x10
 
+/**
+ * HIGH_RR is the highest eDP panel refresh rate read from EDID
+ * LOW_RR is the lowest eDP panel refresh rate found from EDID
+ * parsing for same resolution.
+ */
+enum edp_drrs_refresh_rate_type {
+	DRRS_HIGH_RR,
+	DRRS_LOW_RR,
+	DRRS_MAX_RR, /* RR count */
+};
+
 struct intel_dp {
 	uint32_t output_reg;
 	uint32_t aux_ch_ctl_reg;
@@ -463,8 +525,7 @@ struct intel_dp {
 	uint8_t dpcd[DP_RECEIVER_CAP_SIZE];
 	uint8_t psr_dpcd[EDP_PSR_RECEIVER_CAP_SIZE];
 	uint8_t downstream_ports[DP_MAX_DOWNSTREAM_PORTS];
-	struct i2c_adapter adapter;
-	struct i2c_algo_dp_aux_data algo;
+	struct drm_dp_aux aux;
 	uint8_t train_set[4];
 	int panel_power_up_delay;
 	int panel_power_down_delay;
@@ -473,8 +534,30 @@ struct intel_dp {
 	int backlight_off_delay;
 	struct delayed_work panel_vdd_work;
 	bool want_panel_vdd;
+	unsigned long last_power_cycle;
+	unsigned long last_power_on;
+	unsigned long last_backlight_off;
 	bool psr_setup_done;
+	struct notifier_block edp_notifier;
+
+	bool use_tps3;
 	struct intel_connector *attached_connector;
+
+	uint32_t (*get_aux_clock_divider)(struct intel_dp *dp, int index);
+	/*
+	 * This function returns the value we have to program the AUX_CTL
+	 * register with to kick off an AUX transaction.
+	 */
+	uint32_t (*get_aux_send_ctl)(struct intel_dp *dp,
+				     bool has_aux_irq,
+				     int send_bytes,
+				     uint32_t aux_clock_divider);
+	struct {
+		enum drrs_support_type type;
+		enum edp_drrs_refresh_rate_type refresh_rate_type;
+		struct mutex mutex;
+	} drrs_state;
+
 };
 
 struct intel_digital_port {
@@ -490,9 +573,24 @@ vlv_dport_to_channel(struct intel_digital_port *dport)
 {
 	switch (dport->port) {
 	case PORT_B:
-		return 0;
+	case PORT_D:
+		return DPIO_CH0;
 	case PORT_C:
-		return 1;
+		return DPIO_CH1;
+	default:
+		BUG();
+	}
+}
+
+static inline int
+vlv_pipe_to_channel(enum pipe pipe)
+{
+	switch (pipe) {
+	case PIPE_A:
+	case PIPE_C:
+		return DPIO_CH0;
+	case PIPE_B:
+		return DPIO_CH1;
 	default:
 		BUG();
 	}
@@ -522,12 +620,15 @@ struct intel_unpin_work {
 #define INTEL_FLIP_INACTIVE	0
 #define INTEL_FLIP_PENDING	1
 #define INTEL_FLIP_COMPLETE	2
+	u32 flip_count;
+	u32 gtt_offset;
 	bool enable_stall_check;
 };
 
 struct intel_set_config {
 	struct drm_encoder **save_connector_encoders;
 	struct drm_crtc **save_encoder_crtcs;
+	bool *save_crtc_enabled;
 
 	bool fb_changed;
 	bool mode_changed;
@@ -579,8 +680,12 @@ void ilk_enable_gt_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void ilk_disable_gt_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void snb_enable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask);
 void snb_disable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask);
-void hsw_pc8_disable_interrupts(struct drm_device *dev);
-void hsw_pc8_restore_interrupts(struct drm_device *dev);
+void bdw_enable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask);
+void bdw_disable_pm_irq(struct drm_i915_private *dev_priv, uint32_t mask);
+void intel_runtime_pm_disable_interrupts(struct drm_device *dev);
+void intel_runtime_pm_restore_interrupts(struct drm_device *dev);
+int intel_get_crtc_scanline(struct intel_crtc *crtc);
+void i9xx_check_fifo_underruns(struct drm_device *dev);
 
 
 /* intel_crt.c */
@@ -601,7 +706,8 @@ void intel_ddi_disable_transcoder_func(struct drm_i915_private *dev_priv,
 void intel_ddi_enable_pipe_clock(struct intel_crtc *intel_crtc);
 void intel_ddi_disable_pipe_clock(struct intel_crtc *intel_crtc);
 void intel_ddi_setup_hw_pll_state(struct drm_device *dev);
-bool intel_ddi_pll_mode_set(struct drm_crtc *crtc);
+bool intel_ddi_pll_select(struct intel_crtc *crtc);
+void intel_ddi_pll_enable(struct intel_crtc *crtc);
 void intel_ddi_put_crtc_pll(struct drm_crtc *crtc);
 void intel_ddi_set_pipe_settings(struct drm_crtc *crtc);
 void intel_ddi_prepare_link_retrain(struct drm_encoder *encoder);
@@ -612,10 +718,13 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 
 
 /* intel_display.c */
+const char *intel_output_name(int output);
+bool intel_has_pending_fb_unpin(struct drm_device *dev);
 int intel_pch_rawclk(struct drm_device *dev);
+int valleyview_cur_cdclk(struct drm_i915_private *dev_priv);
 void intel_mark_busy(struct drm_device *dev);
 void intel_mark_fb_busy(struct drm_i915_gem_object *obj,
-			struct intel_ring_buffer *ring);
+			struct intel_engine_cs *ring);
 void intel_mark_idle(struct drm_device *dev);
 void intel_crtc_restore_mode(struct drm_crtc *crtc);
 void intel_crtc_update_dpms(struct drm_crtc *crtc);
@@ -638,21 +747,23 @@ enum transcoder intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
 void intel_wait_for_vblank(struct drm_device *dev, int pipe);
 void intel_wait_for_pipe_off(struct drm_device *dev, int pipe);
 int ironlake_get_lanes_required(int target_clock, int link_bw, int bpp);
-void vlv_wait_port_ready(struct drm_i915_private *dev_priv, int port);
+void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
+			 struct intel_digital_port *dport);
 bool intel_get_load_detect_pipe(struct drm_connector *connector,
 				struct drm_display_mode *mode,
-				struct intel_load_detect_pipe *old);
+				struct intel_load_detect_pipe *old,
+				struct drm_modeset_acquire_ctx *ctx);
 void intel_release_load_detect_pipe(struct drm_connector *connector,
-				    struct intel_load_detect_pipe *old);
+				    struct intel_load_detect_pipe *old,
+				    struct drm_modeset_acquire_ctx *ctx);
 int intel_pin_and_fence_fb_obj(struct drm_device *dev,
 			       struct drm_i915_gem_object *obj,
-			       struct intel_ring_buffer *pipelined);
+			       struct intel_engine_cs *pipelined);
 void intel_unpin_fb_obj(struct drm_i915_gem_object *obj);
-int intel_framebuffer_init(struct drm_device *dev,
-			   struct intel_framebuffer *ifb,
+struct drm_framebuffer *
+__intel_framebuffer_create(struct drm_device *dev,
 			   struct drm_mode_fb_cmd2 *mode_cmd,
 			   struct drm_i915_gem_object *obj);
-void intel_framebuffer_fini(struct intel_framebuffer *fb);
 void intel_prepare_page_flip(struct drm_device *dev, int plane);
 void intel_finish_page_flip(struct drm_device *dev, int pipe);
 void intel_finish_page_flip_plane(struct drm_device *dev, int plane);
@@ -680,9 +791,8 @@ unsigned long intel_gen4_compute_page_offset(int *x, int *y,
 					     unsigned int bpp,
 					     unsigned int pitch);
 void intel_display_handle_reset(struct drm_device *dev);
-void hsw_enable_pc8_work(struct work_struct *__work);
-void hsw_enable_package_c8(struct drm_i915_private *dev_priv);
-void hsw_disable_package_c8(struct drm_i915_private *dev_priv);
+void hsw_enable_pc8(struct drm_i915_private *dev_priv);
+void hsw_disable_pc8(struct drm_i915_private *dev_priv);
 void intel_dp_get_m_n(struct intel_crtc *crtc,
 		      struct intel_crtc_config *pipe_config);
 int intel_dotclock_calculate(int link_freq, const struct intel_link_m_n *m_n);
@@ -690,10 +800,16 @@ void
 ironlake_check_encoder_dotclock(const struct intel_crtc_config *pipe_config,
 				int dotclock);
 bool intel_crtc_active(struct drm_crtc *crtc);
-void i915_disable_vga_mem(struct drm_device *dev);
 void hsw_enable_ips(struct intel_crtc *crtc);
 void hsw_disable_ips(struct intel_crtc *crtc);
-void intel_display_set_init_power(struct drm_device *dev, bool enable);
+void intel_display_set_init_power(struct drm_i915_private *dev, bool enable);
+enum intel_display_power_domain
+intel_display_port_power_domain(struct intel_encoder *intel_encoder);
+int valleyview_get_vco(struct drm_i915_private *dev_priv);
+void intel_mode_from_pipe_config(struct drm_display_mode *mode,
+				 struct intel_crtc_config *pipe_config);
+int intel_format_to_fourcc(int format);
+void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc);
 
 
 /* intel_dp.c */
@@ -706,19 +822,19 @@ void intel_dp_stop_link_train(struct intel_dp *intel_dp);
 void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode);
 void intel_dp_encoder_destroy(struct drm_encoder *encoder);
 void intel_dp_check_link_status(struct intel_dp *intel_dp);
+int intel_dp_sink_crc(struct intel_dp *intel_dp, u8 *crc);
 bool intel_dp_compute_config(struct intel_encoder *encoder,
 			     struct intel_crtc_config *pipe_config);
 bool intel_dp_is_edp(struct drm_device *dev, enum port port);
-void ironlake_edp_backlight_on(struct intel_dp *intel_dp);
-void ironlake_edp_backlight_off(struct intel_dp *intel_dp);
-void ironlake_edp_panel_on(struct intel_dp *intel_dp);
-void ironlake_edp_panel_off(struct intel_dp *intel_dp);
-void ironlake_edp_panel_vdd_on(struct intel_dp *intel_dp);
-void ironlake_edp_panel_vdd_off(struct intel_dp *intel_dp, bool sync);
+void intel_edp_backlight_on(struct intel_dp *intel_dp);
+void intel_edp_backlight_off(struct intel_dp *intel_dp);
+void intel_edp_panel_vdd_on(struct intel_dp *intel_dp);
+void intel_edp_panel_on(struct intel_dp *intel_dp);
+void intel_edp_panel_off(struct intel_dp *intel_dp);
 void intel_edp_psr_enable(struct intel_dp *intel_dp);
 void intel_edp_psr_disable(struct intel_dp *intel_dp);
 void intel_edp_psr_update(struct drm_device *dev);
-
+void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate);
 
 /* intel_dsi.c */
 bool intel_dsi_init(struct drm_device *dev);
@@ -793,7 +909,8 @@ int intel_overlay_attrs(struct drm_device *dev, void *data,
 
 /* intel_panel.c */
 int intel_panel_init(struct intel_panel *panel,
-		     struct drm_display_mode *fixed_mode);
+		     struct drm_display_mode *fixed_mode,
+		     struct drm_display_mode *downclock_mode);
 void intel_panel_fini(struct intel_panel *panel);
 void intel_fixed_panel_mode(const struct drm_display_mode *fixed_mode,
 			    struct drm_display_mode *adjusted_mode);
@@ -808,13 +925,18 @@ void intel_panel_set_backlight(struct intel_connector *connector, u32 level,
 int intel_panel_setup_backlight(struct drm_connector *connector);
 void intel_panel_enable_backlight(struct intel_connector *connector);
 void intel_panel_disable_backlight(struct intel_connector *connector);
-void intel_panel_destroy_backlight(struct drm_device *dev);
+void intel_panel_destroy_backlight(struct drm_connector *connector);
+void intel_panel_init_backlight_funcs(struct drm_device *dev);
 enum drm_connector_status intel_panel_detect(struct drm_device *dev);
-
+extern struct drm_display_mode *intel_find_panel_downclock(
+				struct drm_device *dev,
+				struct drm_display_mode *fixed_mode,
+				struct drm_connector *connector);
 
 /* intel_pm.c */
 void intel_init_clock_gating(struct drm_device *dev);
 void intel_suspend_hw(struct drm_device *dev);
+int ilk_wm_max_level(const struct drm_device *dev);
 void intel_update_watermarks(struct drm_crtc *crtc);
 void intel_update_sprite_watermarks(struct drm_plane *plane,
 				    struct drm_crtc *crtc,
@@ -826,26 +948,36 @@ bool intel_fbc_enabled(struct drm_device *dev);
 void intel_update_fbc(struct drm_device *dev);
 void intel_gpu_ips_init(struct drm_i915_private *dev_priv);
 void intel_gpu_ips_teardown(void);
-int intel_power_domains_init(struct drm_device *dev);
-void intel_power_domains_remove(struct drm_device *dev);
-bool intel_display_power_enabled(struct drm_device *dev,
+int intel_power_domains_init(struct drm_i915_private *);
+void intel_power_domains_remove(struct drm_i915_private *);
+bool intel_display_power_enabled(struct drm_i915_private *dev_priv,
 				 enum intel_display_power_domain domain);
-void intel_display_power_get(struct drm_device *dev,
+bool intel_display_power_enabled_unlocked(struct drm_i915_private *dev_priv,
+					  enum intel_display_power_domain domain);
+void intel_display_power_get(struct drm_i915_private *dev_priv,
 			     enum intel_display_power_domain domain);
-void intel_display_power_put(struct drm_device *dev,
+void intel_display_power_put(struct drm_i915_private *dev_priv,
 			     enum intel_display_power_domain domain);
-void intel_power_domains_init_hw(struct drm_device *dev);
-void intel_set_power_well(struct drm_device *dev, bool enable);
+void intel_power_domains_init_hw(struct drm_i915_private *dev_priv);
+void intel_init_gt_powersave(struct drm_device *dev);
+void intel_cleanup_gt_powersave(struct drm_device *dev);
 void intel_enable_gt_powersave(struct drm_device *dev);
 void intel_disable_gt_powersave(struct drm_device *dev);
+void intel_reset_gt_powersave(struct drm_device *dev);
 void ironlake_teardown_rc6(struct drm_device *dev);
 void gen6_update_ring_freq(struct drm_device *dev);
 void gen6_rps_idle(struct drm_i915_private *dev_priv);
 void gen6_rps_boost(struct drm_i915_private *dev_priv);
 void intel_aux_display_runtime_get(struct drm_i915_private *dev_priv);
 void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_get(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_get_noresume(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_put(struct drm_i915_private *dev_priv);
+void intel_init_runtime_pm(struct drm_i915_private *dev_priv);
+void intel_fini_runtime_pm(struct drm_i915_private *dev_priv);
 void ilk_wm_get_hw_state(struct drm_device *dev);
-
+void __vlv_set_power_well(struct drm_i915_private *dev_priv,
+			  enum punit_power_well power_well_id, bool enable);
 
 /* intel_sdvo.c */
 bool intel_sdvo_init(struct drm_device *dev, uint32_t sdvo_reg, bool is_sdvob);
diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c
index d257b093ca6..3fd082933c8 100644
--- a/drivers/gpu/drm/i915/intel_dsi.c
+++ b/drivers/gpu/drm/i915/intel_dsi.c
@@ -35,51 +35,25 @@
 
 /* the sub-encoders aka panel drivers */
 static const struct intel_dsi_device intel_dsi_devices[] = {
+	{
+		.panel_id = MIPI_DSI_GENERIC_PANEL_ID,
+		.name = "vbt-generic-dsi-vid-mode-display",
+		.dev_ops = &vbt_generic_dsi_display_ops,
+	},
 };
 
-
-static void vlv_cck_modify(struct drm_i915_private *dev_priv, u32 reg, u32 val,
-			   u32 mask)
-{
-	u32 tmp = vlv_cck_read(dev_priv, reg);
-	tmp &= ~mask;
-	tmp |= val;
-	vlv_cck_write(dev_priv, reg, tmp);
-}
-
-static void band_gap_wa(struct drm_i915_private *dev_priv)
+static void band_gap_reset(struct drm_i915_private *dev_priv)
 {
 	mutex_lock(&dev_priv->dpio_lock);
 
-	/* Enable bandgap fix in GOP driver */
-	vlv_cck_modify(dev_priv, 0x6D, 0x00010000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x6E, 0x00010000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x6F, 0x00010000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x00, 0x00008000, 0x00008000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x00, 0x00000000, 0x00008000);
-	msleep(20);
-
-	/* Turn Display Trunk on */
-	vlv_cck_modify(dev_priv, 0x6B, 0x00020000, 0x00030000);
-	msleep(20);
-
-	vlv_cck_modify(dev_priv, 0x6C, 0x00020000, 0x00030000);
-	msleep(20);
-
-	vlv_cck_modify(dev_priv, 0x6D, 0x00020000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x6E, 0x00020000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x6F, 0x00020000, 0x00030000);
+	vlv_flisdsi_write(dev_priv, 0x08, 0x0001);
+	vlv_flisdsi_write(dev_priv, 0x0F, 0x0005);
+	vlv_flisdsi_write(dev_priv, 0x0F, 0x0025);
+	udelay(150);
+	vlv_flisdsi_write(dev_priv, 0x0F, 0x0000);
+	vlv_flisdsi_write(dev_priv, 0x08, 0x0000);
 
 	mutex_unlock(&dev_priv->dpio_lock);
-
-	/* Need huge delay, otherwise clock is not stable */
-	msleep(100);
 }
 
 static struct intel_dsi *intel_attached_dsi(struct drm_connector *connector)
@@ -90,12 +64,12 @@ static struct intel_dsi *intel_attached_dsi(struct drm_connector *connector)
 
 static inline bool is_vid_mode(struct intel_dsi *intel_dsi)
 {
-	return intel_dsi->dev.type == INTEL_DSI_VIDEO_MODE;
+	return intel_dsi->operation_mode == INTEL_DSI_VIDEO_MODE;
 }
 
 static inline bool is_cmd_mode(struct intel_dsi *intel_dsi)
 {
-	return intel_dsi->dev.type == INTEL_DSI_COMMAND_MODE;
+	return intel_dsi->operation_mode == INTEL_DSI_COMMAND_MODE;
 }
 
 static void intel_dsi_hot_plug(struct intel_encoder *encoder)
@@ -125,21 +99,42 @@ static bool intel_dsi_compute_config(struct intel_encoder *encoder,
 	return true;
 }
 
-static void intel_dsi_pre_pll_enable(struct intel_encoder *encoder)
+static void intel_dsi_device_ready(struct intel_encoder *encoder)
 {
+	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
+	int pipe = intel_crtc->pipe;
+	u32 val;
+
 	DRM_DEBUG_KMS("\n");
 
-	vlv_enable_dsi_pll(encoder);
-}
+	mutex_lock(&dev_priv->dpio_lock);
+	/* program rcomp for compliance, reduce from 50 ohms to 45 ohms
+	 * needed everytime after power gate */
+	vlv_flisdsi_write(dev_priv, 0x04, 0x0004);
+	mutex_unlock(&dev_priv->dpio_lock);
 
-static void intel_dsi_pre_enable(struct intel_encoder *encoder)
-{
-	DRM_DEBUG_KMS("\n");
+	/* bandgap reset is needed after everytime we do power gate */
+	band_gap_reset(dev_priv);
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), ULPS_STATE_ENTER);
+	usleep_range(2500, 3000);
+
+	val = I915_READ(MIPI_PORT_CTRL(pipe));
+	I915_WRITE(MIPI_PORT_CTRL(pipe), val | LP_OUTPUT_HOLD);
+	usleep_range(1000, 1500);
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), ULPS_STATE_EXIT);
+	usleep_range(2500, 3000);
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), DEVICE_READY);
+	usleep_range(2500, 3000);
 }
 
 static void intel_dsi_enable(struct intel_encoder *encoder)
 {
-	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	int pipe = intel_crtc->pipe;
@@ -147,41 +142,88 @@ static void intel_dsi_enable(struct intel_encoder *encoder)
 
 	DRM_DEBUG_KMS("\n");
 
-	temp = I915_READ(MIPI_DEVICE_READY(pipe));
-	if ((temp & DEVICE_READY) == 0) {
-		temp &= ~ULPS_STATE_MASK;
-		I915_WRITE(MIPI_DEVICE_READY(pipe), temp | DEVICE_READY);
-	} else if (temp & ULPS_STATE_MASK) {
-		temp &= ~ULPS_STATE_MASK;
-		I915_WRITE(MIPI_DEVICE_READY(pipe), temp | ULPS_STATE_EXIT);
-		/*
-		 * We need to ensure that there is a minimum of 1 ms time
-		 * available before clearing the UPLS exit state.
-		 */
-		msleep(2);
-		I915_WRITE(MIPI_DEVICE_READY(pipe), temp);
-	}
-
 	if (is_cmd_mode(intel_dsi))
 		I915_WRITE(MIPI_MAX_RETURN_PKT_SIZE(pipe), 8 * 4);
-
-	if (is_vid_mode(intel_dsi)) {
+	else {
 		msleep(20); /* XXX */
-		dpi_send_cmd(intel_dsi, TURN_ON);
+		dpi_send_cmd(intel_dsi, TURN_ON, DPI_LP_MODE_EN);
 		msleep(100);
 
+		if (intel_dsi->dev.dev_ops->enable)
+			intel_dsi->dev.dev_ops->enable(&intel_dsi->dev);
+
 		/* assert ip_tg_enable signal */
-		temp = I915_READ(MIPI_PORT_CTRL(pipe));
+		temp = I915_READ(MIPI_PORT_CTRL(pipe)) & ~LANE_CONFIGURATION_MASK;
+		temp = temp | intel_dsi->port_bits;
 		I915_WRITE(MIPI_PORT_CTRL(pipe), temp | DPI_ENABLE);
 		POSTING_READ(MIPI_PORT_CTRL(pipe));
 	}
+}
 
-	intel_dsi->dev.dev_ops->enable(&intel_dsi->dev);
+static void intel_dsi_pre_enable(struct intel_encoder *encoder)
+{
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
+	enum pipe pipe = intel_crtc->pipe;
+	u32 tmp;
+
+	DRM_DEBUG_KMS("\n");
+
+	/* Disable DPOunit clock gating, can stall pipe
+	 * and we need DPLL REFA always enabled */
+	tmp = I915_READ(DPLL(pipe));
+	tmp |= DPLL_REFA_CLK_ENABLE_VLV;
+	I915_WRITE(DPLL(pipe), tmp);
+
+	tmp = I915_READ(DSPCLK_GATE_D);
+	tmp |= DPOUNIT_CLOCK_GATE_DISABLE;
+	I915_WRITE(DSPCLK_GATE_D, tmp);
+
+	/* put device in ready state */
+	intel_dsi_device_ready(encoder);
+
+	msleep(intel_dsi->panel_on_delay);
+
+	if (intel_dsi->dev.dev_ops->panel_reset)
+		intel_dsi->dev.dev_ops->panel_reset(&intel_dsi->dev);
+
+	if (intel_dsi->dev.dev_ops->send_otp_cmds)
+		intel_dsi->dev.dev_ops->send_otp_cmds(&intel_dsi->dev);
+
+	/* Enable port in pre-enable phase itself because as per hw team
+	 * recommendation, port should be enabled befor plane & pipe */
+	intel_dsi_enable(encoder);
+}
+
+static void intel_dsi_enable_nop(struct intel_encoder *encoder)
+{
+	DRM_DEBUG_KMS("\n");
+
+	/* for DSI port enable has to be done before pipe
+	 * and plane enable, so port enable is done in
+	 * pre_enable phase itself unlike other encoders
+	 */
+}
+
+static void intel_dsi_pre_disable(struct intel_encoder *encoder)
+{
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
+
+	DRM_DEBUG_KMS("\n");
+
+	if (is_vid_mode(intel_dsi)) {
+		/* Send Shutdown command to the panel in LP mode */
+		dpi_send_cmd(intel_dsi, SHUTDOWN, DPI_LP_MODE_EN);
+		msleep(10);
+	}
 }
 
 static void intel_dsi_disable(struct intel_encoder *encoder)
 {
-	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	int pipe = intel_crtc->pipe;
@@ -189,12 +231,7 @@ static void intel_dsi_disable(struct intel_encoder *encoder)
 
 	DRM_DEBUG_KMS("\n");
 
-	intel_dsi->dev.dev_ops->disable(&intel_dsi->dev);
-
 	if (is_vid_mode(intel_dsi)) {
-		dpi_send_cmd(intel_dsi, SHUTDOWN);
-		msleep(10);
-
 		/* de-assert ip_tg_enable signal */
 		temp = I915_READ(MIPI_PORT_CTRL(pipe));
 		I915_WRITE(MIPI_PORT_CTRL(pipe), temp & ~DPI_ENABLE);
@@ -203,30 +240,98 @@ static void intel_dsi_disable(struct intel_encoder *encoder)
 		msleep(2);
 	}
 
-	temp = I915_READ(MIPI_DEVICE_READY(pipe));
-	if (temp & DEVICE_READY) {
-		temp &= ~DEVICE_READY;
-		temp &= ~ULPS_STATE_MASK;
-		I915_WRITE(MIPI_DEVICE_READY(pipe), temp);
-	}
+	/* Panel commands can be sent when clock is in LP11 */
+	I915_WRITE(MIPI_DEVICE_READY(pipe), 0x0);
+
+	temp = I915_READ(MIPI_CTRL(pipe));
+	temp &= ~ESCAPE_CLOCK_DIVIDER_MASK;
+	I915_WRITE(MIPI_CTRL(pipe), temp |
+			intel_dsi->escape_clk_div <<
+			ESCAPE_CLOCK_DIVIDER_SHIFT);
+
+	I915_WRITE(MIPI_EOT_DISABLE(pipe), CLOCKSTOP);
+
+	temp = I915_READ(MIPI_DSI_FUNC_PRG(pipe));
+	temp &= ~VID_MODE_FORMAT_MASK;
+	I915_WRITE(MIPI_DSI_FUNC_PRG(pipe), temp);
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), 0x1);
+
+	/* if disable packets are sent before sending shutdown packet then in
+	 * some next enable sequence send turn on packet error is observed */
+	if (intel_dsi->dev.dev_ops->disable)
+		intel_dsi->dev.dev_ops->disable(&intel_dsi->dev);
 }
 
-static void intel_dsi_post_disable(struct intel_encoder *encoder)
+static void intel_dsi_clear_device_ready(struct intel_encoder *encoder)
 {
+	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
+	int pipe = intel_crtc->pipe;
+	u32 val;
+
 	DRM_DEBUG_KMS("\n");
 
+	I915_WRITE(MIPI_DEVICE_READY(pipe), DEVICE_READY | ULPS_STATE_ENTER);
+	usleep_range(2000, 2500);
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), DEVICE_READY | ULPS_STATE_EXIT);
+	usleep_range(2000, 2500);
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), DEVICE_READY | ULPS_STATE_ENTER);
+	usleep_range(2000, 2500);
+
+	if (wait_for(((I915_READ(MIPI_PORT_CTRL(pipe)) & AFE_LATCHOUT)
+					== 0x00000), 30))
+		DRM_ERROR("DSI LP not going Low\n");
+
+	val = I915_READ(MIPI_PORT_CTRL(pipe));
+	I915_WRITE(MIPI_PORT_CTRL(pipe), val & ~LP_OUTPUT_HOLD);
+	usleep_range(1000, 1500);
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), 0x00);
+	usleep_range(2000, 2500);
+
 	vlv_disable_dsi_pll(encoder);
 }
 
+static void intel_dsi_post_disable(struct intel_encoder *encoder)
+{
+	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
+	u32 val;
+
+	DRM_DEBUG_KMS("\n");
+
+	intel_dsi_disable(encoder);
+
+	intel_dsi_clear_device_ready(encoder);
+
+	val = I915_READ(DSPCLK_GATE_D);
+	val &= ~DPOUNIT_CLOCK_GATE_DISABLE;
+	I915_WRITE(DSPCLK_GATE_D, val);
+
+	if (intel_dsi->dev.dev_ops->disable_panel_power)
+		intel_dsi->dev.dev_ops->disable_panel_power(&intel_dsi->dev);
+
+	msleep(intel_dsi->panel_off_delay);
+	msleep(intel_dsi->panel_pwr_cycle_delay);
+}
+
 static bool intel_dsi_get_hw_state(struct intel_encoder *encoder,
 				   enum pipe *pipe)
 {
 	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	enum intel_display_power_domain power_domain;
 	u32 port, func;
 	enum pipe p;
 
 	DRM_DEBUG_KMS("\n");
 
+	power_domain = intel_display_port_power_domain(encoder);
+	if (!intel_display_power_enabled(dev_priv, power_domain))
+		return false;
+
 	/* XXX: this only works for one DSI output */
 	for (p = PIPE_A; p <= PIPE_B; p++) {
 		port = I915_READ(MIPI_PORT_CTRL(p));
@@ -251,8 +356,9 @@ static void intel_dsi_get_config(struct intel_encoder *encoder,
 	/* XXX: read flags, set to adjusted_mode */
 }
 
-static int intel_dsi_mode_valid(struct drm_connector *connector,
-				struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_dsi_mode_valid(struct drm_connector *connector,
+		     struct drm_display_mode *mode)
 {
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 	struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode;
@@ -337,7 +443,7 @@ static void set_dsi_timings(struct drm_encoder *encoder,
 	I915_WRITE(MIPI_VBP_COUNT(pipe), vbp);
 }
 
-static void intel_dsi_mode_set(struct intel_encoder *intel_encoder)
+static void intel_dsi_prepare(struct intel_encoder *intel_encoder)
 {
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_device *dev = encoder->dev;
@@ -352,12 +458,6 @@ static void intel_dsi_mode_set(struct intel_encoder *intel_encoder)
 
 	DRM_DEBUG_KMS("pipe %c\n", pipe_name(pipe));
 
-	/* Update the DSI PLL */
-	vlv_enable_dsi_pll(intel_encoder);
-
-	/* XXX: Location of the call */
-	band_gap_wa(dev_priv);
-
 	/* escape clock divider, 20MHz, shared for A and C. device ready must be
 	 * off when doing this! txclkesc? */
 	tmp = I915_READ(MIPI_CTRL(0));
@@ -373,11 +473,7 @@ static void intel_dsi_mode_set(struct intel_encoder *intel_encoder)
 	I915_WRITE(MIPI_INTR_STAT(pipe), 0xffffffff);
 	I915_WRITE(MIPI_INTR_EN(pipe), 0xffffffff);
 
-	I915_WRITE(MIPI_DPHY_PARAM(pipe),
-		   0x3c << EXIT_ZERO_COUNT_SHIFT |
-		   0x1f << TRAIL_COUNT_SHIFT |
-		   0xc5 << CLK_ZERO_COUNT_SHIFT |
-		   0x1f << PREPARE_COUNT_SHIFT);
+	I915_WRITE(MIPI_DPHY_PARAM(pipe), intel_dsi->dphy_reg);
 
 	I915_WRITE(MIPI_DPI_RESOLUTION(pipe),
 		   adjusted_mode->vdisplay << VERTICAL_ADDRESS_SHIFT |
@@ -425,54 +521,91 @@ static void intel_dsi_mode_set(struct intel_encoder *intel_encoder)
 				       adjusted_mode->htotal,
 				       bpp, intel_dsi->lane_count) + 1);
 	}
-	I915_WRITE(MIPI_LP_RX_TIMEOUT(pipe), 8309); /* max */
-	I915_WRITE(MIPI_TURN_AROUND_TIMEOUT(pipe), 0x14); /* max */
-	I915_WRITE(MIPI_DEVICE_RESET_TIMER(pipe), 0xffff); /* max */
+	I915_WRITE(MIPI_LP_RX_TIMEOUT(pipe), intel_dsi->lp_rx_timeout);
+	I915_WRITE(MIPI_TURN_AROUND_TIMEOUT(pipe), intel_dsi->turn_arnd_val);
+	I915_WRITE(MIPI_DEVICE_RESET_TIMER(pipe), intel_dsi->rst_timer_val);
 
 	/* dphy stuff */
 
 	/* in terms of low power clock */
-	I915_WRITE(MIPI_INIT_COUNT(pipe), txclkesc(ESCAPE_CLOCK_DIVIDER_1, 100));
+	I915_WRITE(MIPI_INIT_COUNT(pipe), txclkesc(intel_dsi->escape_clk_div, 100));
+
+	val = 0;
+	if (intel_dsi->eotp_pkt == 0)
+		val |= EOT_DISABLE;
+
+	if (intel_dsi->clock_stop)
+		val |= CLOCKSTOP;
 
 	/* recovery disables */
-	I915_WRITE(MIPI_EOT_DISABLE(pipe), intel_dsi->eot_disable);
+	I915_WRITE(MIPI_EOT_DISABLE(pipe), val);
+
+	/* in terms of low power clock */
+	I915_WRITE(MIPI_INIT_COUNT(pipe), intel_dsi->init_count);
 
 	/* in terms of txbyteclkhs. actual high to low switch +
 	 * MIPI_STOP_STATE_STALL * MIPI_LP_BYTECLK.
 	 *
 	 * XXX: write MIPI_STOP_STATE_STALL?
 	 */
-	I915_WRITE(MIPI_HIGH_LOW_SWITCH_COUNT(pipe), 0x46);
+	I915_WRITE(MIPI_HIGH_LOW_SWITCH_COUNT(pipe),
+						intel_dsi->hs_to_lp_count);
 
 	/* XXX: low power clock equivalence in terms of byte clock. the number
 	 * of byte clocks occupied in one low power clock. based on txbyteclkhs
 	 * and txclkesc. txclkesc time / txbyteclk time * (105 +
 	 * MIPI_STOP_STATE_STALL) / 105.???
 	 */
-	I915_WRITE(MIPI_LP_BYTECLK(pipe), 4);
+	I915_WRITE(MIPI_LP_BYTECLK(pipe), intel_dsi->lp_byte_clk);
 
 	/* the bw essential for transmitting 16 long packets containing 252
 	 * bytes meant for dcs write memory command is programmed in this
 	 * register in terms of byte clocks. based on dsi transfer rate and the
 	 * number of lanes configured the time taken to transmit 16 long packets
 	 * in a dsi stream varies. */
-	I915_WRITE(MIPI_DBI_BW_CTRL(pipe), 0x820);
+	I915_WRITE(MIPI_DBI_BW_CTRL(pipe), intel_dsi->bw_timer);
 
 	I915_WRITE(MIPI_CLK_LANE_SWITCH_TIME_CNT(pipe),
-		   0xa << LP_HS_SSW_CNT_SHIFT |
-		   0x14 << HS_LP_PWR_SW_CNT_SHIFT);
+		   intel_dsi->clk_lp_to_hs_count << LP_HS_SSW_CNT_SHIFT |
+		   intel_dsi->clk_hs_to_lp_count << HS_LP_PWR_SW_CNT_SHIFT);
 
 	if (is_vid_mode(intel_dsi))
+		/* Some panels might have resolution which is not a multiple of
+		 * 64 like 1366 x 768. Enable RANDOM resolution support for such
+		 * panels by default */
 		I915_WRITE(MIPI_VIDEO_MODE_FORMAT(pipe),
-			   intel_dsi->video_mode_format);
+				intel_dsi->video_frmt_cfg_bits |
+				intel_dsi->video_mode_format |
+				IP_TG_CONFIG |
+				RANDOM_DPI_DISPLAY_RESOLUTION);
+}
+
+static void intel_dsi_pre_pll_enable(struct intel_encoder *encoder)
+{
+	DRM_DEBUG_KMS("\n");
+
+	intel_dsi_prepare(encoder);
+
+	vlv_enable_dsi_pll(encoder);
 }
 
 static enum drm_connector_status
 intel_dsi_detect(struct drm_connector *connector, bool force)
 {
 	struct intel_dsi *intel_dsi = intel_attached_dsi(connector);
+	struct intel_encoder *intel_encoder = &intel_dsi->base;
+	enum intel_display_power_domain power_domain;
+	enum drm_connector_status connector_status;
+	struct drm_i915_private *dev_priv = intel_encoder->base.dev->dev_private;
+
 	DRM_DEBUG_KMS("\n");
-	return intel_dsi->dev.dev_ops->detect(&intel_dsi->dev);
+	power_domain = intel_display_port_power_domain(intel_encoder);
+
+	intel_display_power_get(dev_priv, power_domain);
+	connector_status = intel_dsi->dev.dev_ops->detect(&intel_dsi->dev);
+	intel_display_power_put(dev_priv, power_domain);
+
+	return connector_status;
 }
 
 static int intel_dsi_get_modes(struct drm_connector *connector)
@@ -533,11 +666,16 @@ bool intel_dsi_init(struct drm_device *dev)
 	struct intel_connector *intel_connector;
 	struct drm_connector *connector;
 	struct drm_display_mode *fixed_mode = NULL;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	const struct intel_dsi_device *dsi;
 	unsigned int i;
 
 	DRM_DEBUG_KMS("\n");
 
+	/* There is no detection method for MIPI so rely on VBT */
+	if (!dev_priv->vbt.has_mipi)
+		return false;
+
 	intel_dsi = kzalloc(sizeof(*intel_dsi), GFP_KERNEL);
 	if (!intel_dsi)
 		return false;
@@ -552,6 +690,13 @@ bool intel_dsi_init(struct drm_device *dev)
 	encoder = &intel_encoder->base;
 	intel_dsi->attached_connector = intel_connector;
 
+	if (IS_VALLEYVIEW(dev)) {
+		dev_priv->mipi_mmio_base = VLV_MIPI_BASE;
+	} else {
+		DRM_ERROR("Unsupported Mipi device to reg base");
+		return false;
+	}
+
 	connector = &intel_connector->base;
 
 	drm_encoder_init(dev, encoder, &intel_dsi_funcs, DRM_MODE_ENCODER_DSI);
@@ -561,14 +706,14 @@ bool intel_dsi_init(struct drm_device *dev)
 	intel_encoder->compute_config = intel_dsi_compute_config;
 	intel_encoder->pre_pll_enable = intel_dsi_pre_pll_enable;
 	intel_encoder->pre_enable = intel_dsi_pre_enable;
-	intel_encoder->enable = intel_dsi_enable;
-	intel_encoder->mode_set = intel_dsi_mode_set;
-	intel_encoder->disable = intel_dsi_disable;
+	intel_encoder->enable = intel_dsi_enable_nop;
+	intel_encoder->disable = intel_dsi_pre_disable;
 	intel_encoder->post_disable = intel_dsi_post_disable;
 	intel_encoder->get_hw_state = intel_dsi_get_hw_state;
 	intel_encoder->get_config = intel_dsi_get_config;
 
 	intel_connector->get_hw_state = intel_connector_get_hw_state;
+	intel_connector->unregister = intel_connector_unregister;
 
 	for (i = 0; i < ARRAY_SIZE(intel_dsi_devices); i++) {
 		dsi = &intel_dsi_devices[i];
@@ -586,7 +731,7 @@ bool intel_dsi_init(struct drm_device *dev)
 	intel_encoder->type = INTEL_OUTPUT_DSI;
 	intel_encoder->crtc_mask = (1 << 0); /* XXX */
 
-	intel_encoder->cloneable = false;
+	intel_encoder->cloneable = 0;
 	drm_connector_init(dev, connector, &intel_dsi_connector_funcs,
 			   DRM_MODE_CONNECTOR_DSI);
 
@@ -607,7 +752,7 @@ bool intel_dsi_init(struct drm_device *dev)
 	}
 
 	fixed_mode->type |= DRM_MODE_TYPE_PREFERRED;
-	intel_panel_init(&intel_connector->panel, fixed_mode);
+	intel_panel_init(&intel_connector->panel, fixed_mode, NULL);
 
 	return true;
 
diff --git a/drivers/gpu/drm/i915/intel_dsi.h b/drivers/gpu/drm/i915/intel_dsi.h
index c7765f33d52..31db33d3e5c 100644
--- a/drivers/gpu/drm/i915/intel_dsi.h
+++ b/drivers/gpu/drm/i915/intel_dsi.h
@@ -31,7 +31,6 @@
 struct intel_dsi_device {
 	unsigned int panel_id;
 	const char *name;
-	int type;
 	const struct intel_dsi_dev_ops *dev_ops;
 	void *dev_priv;
 };
@@ -39,6 +38,13 @@ struct intel_dsi_device {
 struct intel_dsi_dev_ops {
 	bool (*init)(struct intel_dsi_device *dsi);
 
+	void (*panel_reset)(struct intel_dsi_device *dsi);
+
+	void (*disable_panel_power)(struct intel_dsi_device *dsi);
+
+	/* one time programmable commands if needed */
+	void (*send_otp_cmds)(struct intel_dsi_device *dsi);
+
 	/* This callback must be able to assume DSI commands can be sent */
 	void (*enable)(struct intel_dsi_device *dsi);
 
@@ -78,6 +84,9 @@ struct intel_dsi {
 	/* virtual channel */
 	int channel;
 
+	/* Video mode or command mode */
+	u16 operation_mode;
+
 	/* number of DSI lanes */
 	unsigned int lane_count;
 
@@ -88,7 +97,32 @@ struct intel_dsi {
 	u32 video_mode_format;
 
 	/* eot for MIPI_EOT_DISABLE register */
-	u32 eot_disable;
+	u8 eotp_pkt;
+	u8 clock_stop;
+
+	u8 escape_clk_div;
+	u32 port_bits;
+	u32 bw_timer;
+	u32 dphy_reg;
+	u32 video_frmt_cfg_bits;
+	u16 lp_byte_clk;
+
+	/* timeouts in byte clocks */
+	u16 lp_rx_timeout;
+	u16 turn_arnd_val;
+	u16 rst_timer_val;
+	u16 hs_to_lp_count;
+	u16 clk_lp_to_hs_count;
+	u16 clk_hs_to_lp_count;
+
+	u16 init_count;
+
+	/* all delays in ms */
+	u16 backlight_off_delay;
+	u16 backlight_on_delay;
+	u16 panel_on_delay;
+	u16 panel_off_delay;
+	u16 panel_pwr_cycle_delay;
 };
 
 static inline struct intel_dsi *enc_to_intel_dsi(struct drm_encoder *encoder)
@@ -99,4 +133,6 @@ static inline struct intel_dsi *enc_to_intel_dsi(struct drm_encoder *encoder)
 extern void vlv_enable_dsi_pll(struct intel_encoder *encoder);
 extern void vlv_disable_dsi_pll(struct intel_encoder *encoder);
 
+extern struct intel_dsi_dev_ops vbt_generic_dsi_display_ops;
+
 #endif /* _INTEL_DSI_H */
diff --git a/drivers/gpu/drm/i915/intel_dsi_cmd.c b/drivers/gpu/drm/i915/intel_dsi_cmd.c
index 7c40f981d2c..933c8630523 100644
--- a/drivers/gpu/drm/i915/intel_dsi_cmd.c
+++ b/drivers/gpu/drm/i915/intel_dsi_cmd.c
@@ -389,7 +389,7 @@ int dsi_vc_generic_read(struct intel_dsi *intel_dsi, int channel,
  *
  * XXX: commands with data in MIPI_DPI_DATA?
  */
-int dpi_send_cmd(struct intel_dsi *intel_dsi, u32 cmd)
+int dpi_send_cmd(struct intel_dsi *intel_dsi, u32 cmd, bool hs)
 {
 	struct drm_encoder *encoder = &intel_dsi->base.base;
 	struct drm_device *dev = encoder->dev;
@@ -399,17 +399,11 @@ int dpi_send_cmd(struct intel_dsi *intel_dsi, u32 cmd)
 	u32 mask;
 
 	/* XXX: pipe, hs */
-	if (intel_dsi->hs)
+	if (hs)
 		cmd &= ~DPI_LP_MODE;
 	else
 		cmd |= DPI_LP_MODE;
 
-	/* DPI virtual channel?! */
-
-	mask = DPI_FIFO_EMPTY;
-	if (wait_for((I915_READ(MIPI_GEN_FIFO_STAT(pipe)) & mask) == mask, 50))
-		DRM_ERROR("Timeout waiting for DPI FIFO empty.\n");
-
 	/* clear bit */
 	I915_WRITE(MIPI_INTR_STAT(pipe), SPL_PKT_SENT_INTERRUPT);
 
diff --git a/drivers/gpu/drm/i915/intel_dsi_cmd.h b/drivers/gpu/drm/i915/intel_dsi_cmd.h
index 54c8a234a2e..9a18cbfa546 100644
--- a/drivers/gpu/drm/i915/intel_dsi_cmd.h
+++ b/drivers/gpu/drm/i915/intel_dsi_cmd.h
@@ -33,6 +33,9 @@
 #include "intel_drv.h"
 #include "intel_dsi.h"
 
+#define DPI_LP_MODE_EN	false
+#define DPI_HS_MODE_EN	true
+
 void dsi_hs_mode_enable(struct intel_dsi *intel_dsi, bool enable);
 
 int dsi_vc_dcs_write(struct intel_dsi *intel_dsi, int channel,
@@ -47,7 +50,7 @@ int dsi_vc_dcs_read(struct intel_dsi *intel_dsi, int channel, u8 dcs_cmd,
 int dsi_vc_generic_read(struct intel_dsi *intel_dsi, int channel,
 			u8 *reqdata, int reqlen, u8 *buf, int buflen);
 
-int dpi_send_cmd(struct intel_dsi *intel_dsi, u32 cmd);
+int dpi_send_cmd(struct intel_dsi *intel_dsi, u32 cmd, bool hs);
 
 /* XXX: questionable write helpers */
 static inline int dsi_vc_dcs_write_0(struct intel_dsi *intel_dsi,
diff --git a/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c b/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c
new file mode 100644
index 00000000000..21a0d348ced
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_dsi_panel_vbt.c
@@ -0,0 +1,589 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Shobhit Kumar <shobhit.kumar@intel.com>
+ *
+ */
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_edid.h>
+#include <drm/i915_drm.h>
+#include <linux/slab.h>
+#include <video/mipi_display.h>
+#include <asm/intel-mid.h>
+#include <video/mipi_display.h>
+#include "i915_drv.h"
+#include "intel_drv.h"
+#include "intel_dsi.h"
+#include "intel_dsi_cmd.h"
+
+#define MIPI_TRANSFER_MODE_SHIFT	0
+#define MIPI_VIRTUAL_CHANNEL_SHIFT	1
+#define MIPI_PORT_SHIFT			3
+
+#define PREPARE_CNT_MAX		0x3F
+#define EXIT_ZERO_CNT_MAX	0x3F
+#define CLK_ZERO_CNT_MAX	0xFF
+#define TRAIL_CNT_MAX		0x1F
+
+#define NS_KHZ_RATIO 1000000
+
+#define GPI0_NC_0_HV_DDI0_HPD           0x4130
+#define GPIO_NC_0_HV_DDI0_PAD           0x4138
+#define GPIO_NC_1_HV_DDI0_DDC_SDA       0x4120
+#define GPIO_NC_1_HV_DDI0_DDC_SDA_PAD   0x4128
+#define GPIO_NC_2_HV_DDI0_DDC_SCL       0x4110
+#define GPIO_NC_2_HV_DDI0_DDC_SCL_PAD   0x4118
+#define GPIO_NC_3_PANEL0_VDDEN          0x4140
+#define GPIO_NC_3_PANEL0_VDDEN_PAD      0x4148
+#define GPIO_NC_4_PANEL0_BLKEN          0x4150
+#define GPIO_NC_4_PANEL0_BLKEN_PAD      0x4158
+#define GPIO_NC_5_PANEL0_BLKCTL         0x4160
+#define GPIO_NC_5_PANEL0_BLKCTL_PAD     0x4168
+#define GPIO_NC_6_PCONF0                0x4180
+#define GPIO_NC_6_PAD                   0x4188
+#define GPIO_NC_7_PCONF0                0x4190
+#define GPIO_NC_7_PAD                   0x4198
+#define GPIO_NC_8_PCONF0                0x4170
+#define GPIO_NC_8_PAD                   0x4178
+#define GPIO_NC_9_PCONF0                0x4100
+#define GPIO_NC_9_PAD                   0x4108
+#define GPIO_NC_10_PCONF0               0x40E0
+#define GPIO_NC_10_PAD                  0x40E8
+#define GPIO_NC_11_PCONF0               0x40F0
+#define GPIO_NC_11_PAD                  0x40F8
+
+struct gpio_table {
+	u16 function_reg;
+	u16 pad_reg;
+	u8 init;
+};
+
+static struct gpio_table gtable[] = {
+	{ GPI0_NC_0_HV_DDI0_HPD, GPIO_NC_0_HV_DDI0_PAD, 0 },
+	{ GPIO_NC_1_HV_DDI0_DDC_SDA, GPIO_NC_1_HV_DDI0_DDC_SDA_PAD, 0 },
+	{ GPIO_NC_2_HV_DDI0_DDC_SCL, GPIO_NC_2_HV_DDI0_DDC_SCL_PAD, 0 },
+	{ GPIO_NC_3_PANEL0_VDDEN, GPIO_NC_3_PANEL0_VDDEN_PAD, 0 },
+	{ GPIO_NC_4_PANEL0_BLKEN, GPIO_NC_4_PANEL0_BLKEN_PAD, 0 },
+	{ GPIO_NC_5_PANEL0_BLKCTL, GPIO_NC_5_PANEL0_BLKCTL_PAD, 0 },
+	{ GPIO_NC_6_PCONF0, GPIO_NC_6_PAD, 0 },
+	{ GPIO_NC_7_PCONF0, GPIO_NC_7_PAD, 0 },
+	{ GPIO_NC_8_PCONF0, GPIO_NC_8_PAD, 0 },
+	{ GPIO_NC_9_PCONF0, GPIO_NC_9_PAD, 0 },
+	{ GPIO_NC_10_PCONF0, GPIO_NC_10_PAD, 0},
+	{ GPIO_NC_11_PCONF0, GPIO_NC_11_PAD, 0}
+};
+
+static u8 *mipi_exec_send_packet(struct intel_dsi *intel_dsi, u8 *data)
+{
+	u8 type, byte, mode, vc, port;
+	u16 len;
+
+	byte = *data++;
+	mode = (byte >> MIPI_TRANSFER_MODE_SHIFT) & 0x1;
+	vc = (byte >> MIPI_VIRTUAL_CHANNEL_SHIFT) & 0x3;
+	port = (byte >> MIPI_PORT_SHIFT) & 0x3;
+
+	/* LP or HS mode */
+	intel_dsi->hs = mode;
+
+	/* get packet type and increment the pointer */
+	type = *data++;
+
+	len = *((u16 *) data);
+	data += 2;
+
+	switch (type) {
+	case MIPI_DSI_GENERIC_SHORT_WRITE_0_PARAM:
+		dsi_vc_generic_write_0(intel_dsi, vc);
+		break;
+	case MIPI_DSI_GENERIC_SHORT_WRITE_1_PARAM:
+		dsi_vc_generic_write_1(intel_dsi, vc, *data);
+		break;
+	case MIPI_DSI_GENERIC_SHORT_WRITE_2_PARAM:
+		dsi_vc_generic_write_2(intel_dsi, vc, *data, *(data + 1));
+		break;
+	case MIPI_DSI_GENERIC_READ_REQUEST_0_PARAM:
+	case MIPI_DSI_GENERIC_READ_REQUEST_1_PARAM:
+	case MIPI_DSI_GENERIC_READ_REQUEST_2_PARAM:
+		DRM_DEBUG_DRIVER("Generic Read not yet implemented or used\n");
+		break;
+	case MIPI_DSI_GENERIC_LONG_WRITE:
+		dsi_vc_generic_write(intel_dsi, vc, data, len);
+		break;
+	case MIPI_DSI_DCS_SHORT_WRITE:
+		dsi_vc_dcs_write_0(intel_dsi, vc, *data);
+		break;
+	case MIPI_DSI_DCS_SHORT_WRITE_PARAM:
+		dsi_vc_dcs_write_1(intel_dsi, vc, *data, *(data + 1));
+		break;
+	case MIPI_DSI_DCS_READ:
+		DRM_DEBUG_DRIVER("DCS Read not yet implemented or used\n");
+		break;
+	case MIPI_DSI_DCS_LONG_WRITE:
+		dsi_vc_dcs_write(intel_dsi, vc, data, len);
+		break;
+	};
+
+	data += len;
+
+	return data;
+}
+
+static u8 *mipi_exec_delay(struct intel_dsi *intel_dsi, u8 *data)
+{
+	u32 delay = *((u32 *) data);
+
+	usleep_range(delay, delay + 10);
+	data += 4;
+
+	return data;
+}
+
+static u8 *mipi_exec_gpio(struct intel_dsi *intel_dsi, u8 *data)
+{
+	u8 gpio, action;
+	u16 function, pad;
+	u32 val;
+	struct drm_device *dev = intel_dsi->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	gpio = *data++;
+
+	/* pull up/down */
+	action = *data++;
+
+	function = gtable[gpio].function_reg;
+	pad = gtable[gpio].pad_reg;
+
+	mutex_lock(&dev_priv->dpio_lock);
+	if (!gtable[gpio].init) {
+		/* program the function */
+		/* FIXME: remove constant below */
+		vlv_gpio_nc_write(dev_priv, function, 0x2000CC00);
+		gtable[gpio].init = 1;
+	}
+
+	val = 0x4 | action;
+
+	/* pull up/down */
+	vlv_gpio_nc_write(dev_priv, pad, val);
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	return data;
+}
+
+typedef u8 * (*fn_mipi_elem_exec)(struct intel_dsi *intel_dsi, u8 *data);
+static const fn_mipi_elem_exec exec_elem[] = {
+	NULL, /* reserved */
+	mipi_exec_send_packet,
+	mipi_exec_delay,
+	mipi_exec_gpio,
+	NULL, /* status read; later */
+};
+
+/*
+ * MIPI Sequence from VBT #53 parsing logic
+ * We have already separated each seqence during bios parsing
+ * Following is generic execution function for any sequence
+ */
+
+static const char * const seq_name[] = {
+	"UNDEFINED",
+	"MIPI_SEQ_ASSERT_RESET",
+	"MIPI_SEQ_INIT_OTP",
+	"MIPI_SEQ_DISPLAY_ON",
+	"MIPI_SEQ_DISPLAY_OFF",
+	"MIPI_SEQ_DEASSERT_RESET"
+};
+
+static void generic_exec_sequence(struct intel_dsi *intel_dsi, char *sequence)
+{
+	u8 *data = sequence;
+	fn_mipi_elem_exec mipi_elem_exec;
+	int index;
+
+	if (!sequence)
+		return;
+
+	DRM_DEBUG_DRIVER("Starting MIPI sequence - %s\n", seq_name[*data]);
+
+	/* go to the first element of the sequence */
+	data++;
+
+	/* parse each byte till we reach end of sequence byte - 0x00 */
+	while (1) {
+		index = *data;
+		mipi_elem_exec = exec_elem[index];
+		if (!mipi_elem_exec) {
+			DRM_ERROR("Unsupported MIPI element, skipping sequence execution\n");
+			return;
+		}
+
+		/* goto element payload */
+		data++;
+
+		/* execute the element specific rotines */
+		data = mipi_elem_exec(intel_dsi, data);
+
+		/*
+		 * After processing the element, data should point to
+		 * next element or end of sequence
+		 * check if have we reached end of sequence
+		 */
+		if (*data == 0x00)
+			break;
+	}
+}
+
+static bool generic_init(struct intel_dsi_device *dsi)
+{
+	struct intel_dsi *intel_dsi = container_of(dsi, struct intel_dsi, dev);
+	struct drm_device *dev = intel_dsi->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct mipi_config *mipi_config = dev_priv->vbt.dsi.config;
+	struct mipi_pps_data *pps = dev_priv->vbt.dsi.pps;
+	struct drm_display_mode *mode = dev_priv->vbt.lfp_lvds_vbt_mode;
+	u32 bits_per_pixel = 24;
+	u32 tlpx_ns, extra_byte_count, bitrate, tlpx_ui;
+	u32 ui_num, ui_den;
+	u32 prepare_cnt, exit_zero_cnt, clk_zero_cnt, trail_cnt;
+	u32 ths_prepare_ns, tclk_trail_ns;
+	u32 tclk_prepare_clkzero, ths_prepare_hszero;
+	u32 lp_to_hs_switch, hs_to_lp_switch;
+
+	DRM_DEBUG_KMS("\n");
+
+	intel_dsi->eotp_pkt = mipi_config->eot_pkt_disabled ? 0 : 1;
+	intel_dsi->clock_stop = mipi_config->enable_clk_stop ? 1 : 0;
+	intel_dsi->lane_count = mipi_config->lane_cnt + 1;
+	intel_dsi->pixel_format = mipi_config->videomode_color_format << 7;
+
+	if (intel_dsi->pixel_format == VID_MODE_FORMAT_RGB666)
+		bits_per_pixel = 18;
+	else if (intel_dsi->pixel_format == VID_MODE_FORMAT_RGB565)
+		bits_per_pixel = 16;
+
+	bitrate = (mode->clock * bits_per_pixel) / intel_dsi->lane_count;
+
+	intel_dsi->operation_mode = mipi_config->is_cmd_mode;
+	intel_dsi->video_mode_format = mipi_config->video_transfer_mode;
+	intel_dsi->escape_clk_div = mipi_config->byte_clk_sel;
+	intel_dsi->lp_rx_timeout = mipi_config->lp_rx_timeout;
+	intel_dsi->turn_arnd_val = mipi_config->turn_around_timeout;
+	intel_dsi->rst_timer_val = mipi_config->device_reset_timer;
+	intel_dsi->init_count = mipi_config->master_init_timer;
+	intel_dsi->bw_timer = mipi_config->dbi_bw_timer;
+	intel_dsi->video_frmt_cfg_bits = mipi_config->bta_enabled ? DISABLE_VIDEO_BTA : 0;
+
+	switch (intel_dsi->escape_clk_div) {
+	case 0:
+		tlpx_ns = 50;
+		break;
+	case 1:
+		tlpx_ns = 100;
+		break;
+
+	case 2:
+		tlpx_ns = 200;
+		break;
+	default:
+		tlpx_ns = 50;
+		break;
+	}
+
+	switch (intel_dsi->lane_count) {
+	case 1:
+	case 2:
+		extra_byte_count = 2;
+		break;
+	case 3:
+		extra_byte_count = 4;
+		break;
+	case 4:
+	default:
+		extra_byte_count = 3;
+		break;
+	}
+
+	/*
+	 * ui(s) = 1/f [f in hz]
+	 * ui(ns) = 10^9 / (f*10^6) [f in Mhz] -> 10^3/f(Mhz)
+	 */
+
+	/* in Kbps */
+	ui_num = NS_KHZ_RATIO;
+	ui_den = bitrate;
+
+	tclk_prepare_clkzero = mipi_config->tclk_prepare_clkzero;
+	ths_prepare_hszero = mipi_config->ths_prepare_hszero;
+
+	/*
+	 * B060
+	 * LP byte clock = TLPX/ (8UI)
+	 */
+	intel_dsi->lp_byte_clk = DIV_ROUND_UP(tlpx_ns * ui_den, 8 * ui_num);
+
+	/* count values in UI = (ns value) * (bitrate / (2 * 10^6))
+	 *
+	 * Since txddrclkhs_i is 2xUI, all the count values programmed in
+	 * DPHY param register are divided by 2
+	 *
+	 * prepare count
+	 */
+	ths_prepare_ns = max(mipi_config->ths_prepare, mipi_config->tclk_prepare);
+	prepare_cnt = DIV_ROUND_UP(ths_prepare_ns * ui_den, ui_num * 2);
+
+	/* exit zero count */
+	exit_zero_cnt = DIV_ROUND_UP(
+				(ths_prepare_hszero - ths_prepare_ns) * ui_den,
+				ui_num * 2
+				);
+
+	/*
+	 * Exit zero  is unified val ths_zero and ths_exit
+	 * minimum value for ths_exit = 110ns
+	 * min (exit_zero_cnt * 2) = 110/UI
+	 * exit_zero_cnt = 55/UI
+	 */
+	 if (exit_zero_cnt < (55 * ui_den / ui_num))
+		if ((55 * ui_den) % ui_num)
+			exit_zero_cnt += 1;
+
+	/* clk zero count */
+	clk_zero_cnt = DIV_ROUND_UP(
+			(tclk_prepare_clkzero -	ths_prepare_ns)
+			* ui_den, 2 * ui_num);
+
+	/* trail count */
+	tclk_trail_ns = max(mipi_config->tclk_trail, mipi_config->ths_trail);
+	trail_cnt = DIV_ROUND_UP(tclk_trail_ns * ui_den, 2 * ui_num);
+
+	if (prepare_cnt > PREPARE_CNT_MAX ||
+		exit_zero_cnt > EXIT_ZERO_CNT_MAX ||
+		clk_zero_cnt > CLK_ZERO_CNT_MAX ||
+		trail_cnt > TRAIL_CNT_MAX)
+		DRM_DEBUG_DRIVER("Values crossing maximum limits, restricting to max values\n");
+
+	if (prepare_cnt > PREPARE_CNT_MAX)
+		prepare_cnt = PREPARE_CNT_MAX;
+
+	if (exit_zero_cnt > EXIT_ZERO_CNT_MAX)
+		exit_zero_cnt = EXIT_ZERO_CNT_MAX;
+
+	if (clk_zero_cnt > CLK_ZERO_CNT_MAX)
+		clk_zero_cnt = CLK_ZERO_CNT_MAX;
+
+	if (trail_cnt > TRAIL_CNT_MAX)
+		trail_cnt = TRAIL_CNT_MAX;
+
+	/* B080 */
+	intel_dsi->dphy_reg = exit_zero_cnt << 24 | trail_cnt << 16 |
+						clk_zero_cnt << 8 | prepare_cnt;
+
+	/*
+	 * LP to HS switch count = 4TLPX + PREP_COUNT * 2 + EXIT_ZERO_COUNT * 2
+	 *					+ 10UI + Extra Byte Count
+	 *
+	 * HS to LP switch count = THS-TRAIL + 2TLPX + Extra Byte Count
+	 * Extra Byte Count is calculated according to number of lanes.
+	 * High Low Switch Count is the Max of LP to HS and
+	 * HS to LP switch count
+	 *
+	 */
+	tlpx_ui = DIV_ROUND_UP(tlpx_ns * ui_den, ui_num);
+
+	/* B044 */
+	/* FIXME:
+	 * The comment above does not match with the code */
+	lp_to_hs_switch = DIV_ROUND_UP(4 * tlpx_ui + prepare_cnt * 2 +
+						exit_zero_cnt * 2 + 10, 8);
+
+	hs_to_lp_switch = DIV_ROUND_UP(mipi_config->ths_trail + 2 * tlpx_ui, 8);
+
+	intel_dsi->hs_to_lp_count = max(lp_to_hs_switch, hs_to_lp_switch);
+	intel_dsi->hs_to_lp_count += extra_byte_count;
+
+	/* B088 */
+	/* LP -> HS for clock lanes
+	 * LP clk sync + LP11 + LP01 + tclk_prepare + tclk_zero +
+	 *						extra byte count
+	 * 2TPLX + 1TLPX + 1 TPLX(in ns) + prepare_cnt * 2 + clk_zero_cnt *
+	 *					2(in UI) + extra byte count
+	 * In byteclks = (4TLPX + prepare_cnt * 2 + clk_zero_cnt *2 (in UI)) /
+	 *					8 + extra byte count
+	 */
+	intel_dsi->clk_lp_to_hs_count =
+		DIV_ROUND_UP(
+			4 * tlpx_ui + prepare_cnt * 2 +
+			clk_zero_cnt * 2,
+			8);
+
+	intel_dsi->clk_lp_to_hs_count += extra_byte_count;
+
+	/* HS->LP for Clock Lanes
+	 * Low Power clock synchronisations + 1Tx byteclk + tclk_trail +
+	 *						Extra byte count
+	 * 2TLPX + 8UI + (trail_count*2)(in UI) + Extra byte count
+	 * In byteclks = (2*TLpx(in UI) + trail_count*2 +8)(in UI)/8 +
+	 *						Extra byte count
+	 */
+	intel_dsi->clk_hs_to_lp_count =
+		DIV_ROUND_UP(2 * tlpx_ui + trail_cnt * 2 + 8,
+			8);
+	intel_dsi->clk_hs_to_lp_count += extra_byte_count;
+
+	DRM_DEBUG_KMS("Eot %s\n", intel_dsi->eotp_pkt ? "enabled" : "disabled");
+	DRM_DEBUG_KMS("Clockstop %s\n", intel_dsi->clock_stop ?
+						"disabled" : "enabled");
+	DRM_DEBUG_KMS("Mode %s\n", intel_dsi->operation_mode ? "command" : "video");
+	DRM_DEBUG_KMS("Pixel Format %d\n", intel_dsi->pixel_format);
+	DRM_DEBUG_KMS("TLPX %d\n", intel_dsi->escape_clk_div);
+	DRM_DEBUG_KMS("LP RX Timeout 0x%x\n", intel_dsi->lp_rx_timeout);
+	DRM_DEBUG_KMS("Turnaround Timeout 0x%x\n", intel_dsi->turn_arnd_val);
+	DRM_DEBUG_KMS("Init Count 0x%x\n", intel_dsi->init_count);
+	DRM_DEBUG_KMS("HS to LP Count 0x%x\n", intel_dsi->hs_to_lp_count);
+	DRM_DEBUG_KMS("LP Byte Clock %d\n", intel_dsi->lp_byte_clk);
+	DRM_DEBUG_KMS("DBI BW Timer 0x%x\n", intel_dsi->bw_timer);
+	DRM_DEBUG_KMS("LP to HS Clock Count 0x%x\n", intel_dsi->clk_lp_to_hs_count);
+	DRM_DEBUG_KMS("HS to LP Clock Count 0x%x\n", intel_dsi->clk_hs_to_lp_count);
+	DRM_DEBUG_KMS("BTA %s\n",
+			intel_dsi->video_frmt_cfg_bits & DISABLE_VIDEO_BTA ?
+			"disabled" : "enabled");
+
+	/* delays in VBT are in unit of 100us, so need to convert
+	 * here in ms
+	 * Delay (100us) * 100 /1000 = Delay / 10 (ms) */
+	intel_dsi->backlight_off_delay = pps->bl_disable_delay / 10;
+	intel_dsi->backlight_on_delay = pps->bl_enable_delay / 10;
+	intel_dsi->panel_on_delay = pps->panel_on_delay / 10;
+	intel_dsi->panel_off_delay = pps->panel_off_delay / 10;
+	intel_dsi->panel_pwr_cycle_delay = pps->panel_power_cycle_delay / 10;
+
+	return true;
+}
+
+static int generic_mode_valid(struct intel_dsi_device *dsi,
+		   struct drm_display_mode *mode)
+{
+	return MODE_OK;
+}
+
+static bool generic_mode_fixup(struct intel_dsi_device *dsi,
+		    const struct drm_display_mode *mode,
+		    struct drm_display_mode *adjusted_mode) {
+	return true;
+}
+
+static void generic_panel_reset(struct intel_dsi_device *dsi)
+{
+	struct intel_dsi *intel_dsi = container_of(dsi, struct intel_dsi, dev);
+	struct drm_device *dev = intel_dsi->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	char *sequence = dev_priv->vbt.dsi.sequence[MIPI_SEQ_ASSERT_RESET];
+
+	generic_exec_sequence(intel_dsi, sequence);
+}
+
+static void generic_disable_panel_power(struct intel_dsi_device *dsi)
+{
+	struct intel_dsi *intel_dsi = container_of(dsi, struct intel_dsi, dev);
+	struct drm_device *dev = intel_dsi->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	char *sequence = dev_priv->vbt.dsi.sequence[MIPI_SEQ_DEASSERT_RESET];
+
+	generic_exec_sequence(intel_dsi, sequence);
+}
+
+static void generic_send_otp_cmds(struct intel_dsi_device *dsi)
+{
+	struct intel_dsi *intel_dsi = container_of(dsi, struct intel_dsi, dev);
+	struct drm_device *dev = intel_dsi->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	char *sequence = dev_priv->vbt.dsi.sequence[MIPI_SEQ_INIT_OTP];
+
+	generic_exec_sequence(intel_dsi, sequence);
+}
+
+static void generic_enable(struct intel_dsi_device *dsi)
+{
+	struct intel_dsi *intel_dsi = container_of(dsi, struct intel_dsi, dev);
+	struct drm_device *dev = intel_dsi->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	char *sequence = dev_priv->vbt.dsi.sequence[MIPI_SEQ_DISPLAY_ON];
+
+	generic_exec_sequence(intel_dsi, sequence);
+}
+
+static void generic_disable(struct intel_dsi_device *dsi)
+{
+	struct intel_dsi *intel_dsi = container_of(dsi, struct intel_dsi, dev);
+	struct drm_device *dev = intel_dsi->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	char *sequence = dev_priv->vbt.dsi.sequence[MIPI_SEQ_DISPLAY_OFF];
+
+	generic_exec_sequence(intel_dsi, sequence);
+}
+
+static enum drm_connector_status generic_detect(struct intel_dsi_device *dsi)
+{
+	return connector_status_connected;
+}
+
+static bool generic_get_hw_state(struct intel_dsi_device *dev)
+{
+	return true;
+}
+
+static struct drm_display_mode *generic_get_modes(struct intel_dsi_device *dsi)
+{
+	struct intel_dsi *intel_dsi = container_of(dsi, struct intel_dsi, dev);
+	struct drm_device *dev = intel_dsi->base.base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	dev_priv->vbt.lfp_lvds_vbt_mode->type |= DRM_MODE_TYPE_PREFERRED;
+	return dev_priv->vbt.lfp_lvds_vbt_mode;
+}
+
+static void generic_destroy(struct intel_dsi_device *dsi) { }
+
+/* Callbacks. We might not need them all. */
+struct intel_dsi_dev_ops vbt_generic_dsi_display_ops = {
+	.init = generic_init,
+	.mode_valid = generic_mode_valid,
+	.mode_fixup = generic_mode_fixup,
+	.panel_reset = generic_panel_reset,
+	.disable_panel_power = generic_disable_panel_power,
+	.send_otp_cmds = generic_send_otp_cmds,
+	.enable = generic_enable,
+	.disable = generic_disable,
+	.detect = generic_detect,
+	.get_hw_state = generic_get_hw_state,
+	.get_modes = generic_get_modes,
+	.destroy = generic_destroy,
+};
diff --git a/drivers/gpu/drm/i915/intel_dsi_pll.c b/drivers/gpu/drm/i915/intel_dsi_pll.c
index 44279b2ade8..ba79ec19da3 100644
--- a/drivers/gpu/drm/i915/intel_dsi_pll.c
+++ b/drivers/gpu/drm/i915/intel_dsi_pll.c
@@ -50,6 +50,8 @@ static const u32 lfsr_converts[] = {
 	71, 35							/* 91 - 92 */
 };
 
+#ifdef DSI_CLK_FROM_RR
+
 static u32 dsi_rr_formula(const struct drm_display_mode *mode,
 			  int pixel_format, int video_mode_format,
 			  int lane_count, bool eotp)
@@ -121,7 +123,7 @@ static u32 dsi_rr_formula(const struct drm_display_mode *mode,
 
 	/* the dsi clock is divided by 2 in the hardware to get dsi ddr clock */
 	dsi_bit_clock_hz = bytes_per_x_frames_x_lanes * 8;
-	dsi_clk = dsi_bit_clock_hz / (1000 * 1000);
+	dsi_clk = dsi_bit_clock_hz / 1000;
 
 	if (eotp && video_mode_format == VIDEO_MODE_BURST)
 		dsi_clk *= 2;
@@ -129,64 +131,37 @@ static u32 dsi_rr_formula(const struct drm_display_mode *mode,
 	return dsi_clk;
 }
 
-#ifdef MNP_FROM_TABLE
-
-struct dsi_clock_table {
-	u32 freq;
-	u8 m;
-	u8 p;
-};
-
-static const struct dsi_clock_table dsi_clk_tbl[] = {
-	{300, 72, 6}, {313, 75, 6}, {323, 78, 6}, {333, 80, 6},
-	{343, 82, 6}, {353, 85, 6}, {363, 87, 6}, {373, 90, 6},
-	{383, 92, 6}, {390, 78, 5}, {393, 79, 5}, {400, 80, 5},
-	{401, 80, 5}, {402, 80, 5}, {403, 81, 5}, {404, 81, 5},
-	{405, 81, 5}, {406, 81, 5}, {407, 81, 5}, {408, 82, 5},
-	{409, 82, 5}, {410, 82, 5}, {411, 82, 5}, {412, 82, 5},
-	{413, 83, 5}, {414, 83, 5}, {415, 83, 5}, {416, 83, 5},
-	{417, 83, 5}, {418, 84, 5}, {419, 84, 5}, {420, 84, 5},
-	{430, 86, 5}, {440, 88, 5}, {450, 90, 5}, {460, 92, 5},
-	{470, 75, 4}, {480, 77, 4}, {490, 78, 4}, {500, 80, 4},
-	{510, 82, 4}, {520, 83, 4}, {530, 85, 4}, {540, 86, 4},
-	{550, 88, 4}, {560, 90, 4}, {570, 91, 4}, {580, 70, 3},
-	{590, 71, 3}, {600, 72, 3}, {610, 73, 3}, {620, 74, 3},
-	{630, 76, 3}, {640, 77, 3}, {650, 78, 3}, {660, 79, 3},
-	{670, 80, 3}, {680, 82, 3}, {690, 83, 3}, {700, 84, 3},
-	{710, 85, 3}, {720, 86, 3}, {730, 88, 3}, {740, 89, 3},
-	{750, 90, 3}, {760, 91, 3}, {770, 92, 3}, {780, 62, 2},
-	{790, 63, 2}, {800, 64, 2}, {880, 70, 2}, {900, 72, 2},
-	{1000, 80, 2},		/* dsi clock frequency in Mhz*/
-};
+#else
 
-static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
+/* Get DSI clock from pixel clock */
+static u32 dsi_clk_from_pclk(const struct drm_display_mode *mode,
+			  int pixel_format, int lane_count)
 {
-	unsigned int i;
-	u8 m;
-	u8 n;
-	u8 p;
-	u32 m_seed;
-
-	if (dsi_clk < 300 || dsi_clk > 1000)
-		return -ECHRNG;
+	u32 dsi_clk_khz;
+	u32 bpp;
 
-	for (i = 0; i <= ARRAY_SIZE(dsi_clk_tbl); i++) {
-		if (dsi_clk_tbl[i].freq > dsi_clk)
-			break;
+	switch (pixel_format) {
+	default:
+	case VID_MODE_FORMAT_RGB888:
+	case VID_MODE_FORMAT_RGB666_LOOSE:
+		bpp = 24;
+		break;
+	case VID_MODE_FORMAT_RGB666:
+		bpp = 18;
+		break;
+	case VID_MODE_FORMAT_RGB565:
+		bpp = 16;
+		break;
 	}
 
-	m = dsi_clk_tbl[i].m;
-	p = dsi_clk_tbl[i].p;
-	m_seed = lfsr_converts[m - 62];
-	n = 1;
-	dsi_mnp->dsi_pll_ctrl = 1 << (DSI_PLL_P1_POST_DIV_SHIFT + p - 2);
-	dsi_mnp->dsi_pll_div = (n - 1) << DSI_PLL_N1_DIV_SHIFT |
-		m_seed << DSI_PLL_M1_DIV_SHIFT;
+	/* DSI data rate = pixel clock * bits per pixel / lane count
+	   pixel clock is converted from KHz to Hz */
+	dsi_clk_khz = DIV_ROUND_CLOSEST(mode->clock * bpp, lane_count);
 
-	return 0;
+	return dsi_clk_khz;
 }
 
-#else
+#endif
 
 static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 {
@@ -194,36 +169,47 @@ static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 	u32 ref_clk;
 	u32 error;
 	u32 tmp_error;
-	u32 target_dsi_clk;
-	u32 calc_dsi_clk;
+	int target_dsi_clk;
+	int calc_dsi_clk;
 	u32 calc_m;
 	u32 calc_p;
 	u32 m_seed;
 
-	if (dsi_clk < 300 || dsi_clk > 1150) {
+	/* dsi_clk is expected in KHZ */
+	if (dsi_clk < 300000 || dsi_clk > 1150000) {
 		DRM_ERROR("DSI CLK Out of Range\n");
 		return -ECHRNG;
 	}
 
 	ref_clk = 25000;
-	target_dsi_clk = dsi_clk * 1000;
+	target_dsi_clk = dsi_clk;
 	error = 0xFFFFFFFF;
+	tmp_error = 0xFFFFFFFF;
 	calc_m = 0;
 	calc_p = 0;
 
 	for (m = 62; m <= 92; m++) {
 		for (p = 2; p <= 6; p++) {
-
+			/* Find the optimal m and p divisors
+			with minimal error +/- the required clock */
 			calc_dsi_clk = (m * ref_clk) / p;
-			if (calc_dsi_clk >= target_dsi_clk) {
-				tmp_error = calc_dsi_clk - target_dsi_clk;
-				if (tmp_error < error) {
-					error = tmp_error;
-					calc_m = m;
-					calc_p = p;
-				}
+			if (calc_dsi_clk == target_dsi_clk) {
+				calc_m = m;
+				calc_p = p;
+				error = 0;
+				break;
+			} else
+				tmp_error = abs(target_dsi_clk - calc_dsi_clk);
+
+			if (tmp_error < error) {
+				error = tmp_error;
+				calc_m = m;
+				calc_p = p;
 			}
 		}
+
+		if (error == 0)
+			break;
 	}
 
 	m_seed = lfsr_converts[calc_m - 62];
@@ -235,8 +221,6 @@ static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 	return 0;
 }
 
-#endif
-
 /*
  * XXX: The muxing and gating is hard coded for now. Need to add support for
  * sharing PLLs with two DSI outputs.
@@ -251,9 +235,8 @@ static void vlv_configure_dsi_pll(struct intel_encoder *encoder)
 	struct dsi_mnp dsi_mnp;
 	u32 dsi_clk;
 
-	dsi_clk = dsi_rr_formula(mode, intel_dsi->pixel_format,
-				 intel_dsi->video_mode_format,
-				 intel_dsi->lane_count, !intel_dsi->eot_disable);
+	dsi_clk = dsi_clk_from_pclk(mode, intel_dsi->pixel_format,
+						intel_dsi->lane_count);
 
 	ret = dsi_calc_mnp(dsi_clk, &dsi_mnp);
 	if (ret) {
diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c
index 3c773654685..a3631c0a5c2 100644
--- a/drivers/gpu/drm/i915/intel_dvo.c
+++ b/drivers/gpu/drm/i915/intel_dvo.c
@@ -234,8 +234,9 @@ static void intel_dvo_dpms(struct drm_connector *connector, int mode)
 	intel_modeset_check_state(connector->dev);
 }
 
-static int intel_dvo_mode_valid(struct drm_connector *connector,
-				struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_dvo_mode_valid(struct drm_connector *connector,
+		     struct drm_display_mode *mode)
 {
 	struct intel_dvo *intel_dvo = intel_attached_dvo(connector);
 
@@ -284,7 +285,7 @@ static bool intel_dvo_compute_config(struct intel_encoder *encoder,
 	return true;
 }
 
-static void intel_dvo_mode_set(struct intel_encoder *encoder)
+static void intel_dvo_pre_enable(struct intel_encoder *encoder)
 {
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -342,7 +343,7 @@ intel_dvo_detect(struct drm_connector *connector, bool force)
 {
 	struct intel_dvo *intel_dvo = intel_attached_dvo(connector);
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector));
+		      connector->base.id, connector->name);
 	return intel_dvo->dev.dev_ops->detect(&intel_dvo->dev);
 }
 
@@ -474,8 +475,9 @@ void intel_dvo_init(struct drm_device *dev)
 	intel_encoder->get_hw_state = intel_dvo_get_hw_state;
 	intel_encoder->get_config = intel_dvo_get_config;
 	intel_encoder->compute_config = intel_dvo_compute_config;
-	intel_encoder->mode_set = intel_dvo_mode_set;
+	intel_encoder->pre_enable = intel_dvo_pre_enable;
 	intel_connector->get_hw_state = intel_dvo_connector_get_hw_state;
+	intel_connector->unregister = intel_connector_unregister;
 
 	/* Now, try to find a controller */
 	for (i = 0; i < ARRAY_SIZE(intel_dvo_devices); i++) {
@@ -520,14 +522,15 @@ void intel_dvo_init(struct drm_device *dev)
 		intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
 		switch (dvo->type) {
 		case INTEL_DVO_CHIP_TMDS:
-			intel_encoder->cloneable = true;
+			intel_encoder->cloneable = (1 << INTEL_OUTPUT_ANALOG) |
+				(1 << INTEL_OUTPUT_DVO);
 			drm_connector_init(dev, connector,
 					   &intel_dvo_connector_funcs,
 					   DRM_MODE_CONNECTOR_DVII);
 			encoder_type = DRM_MODE_ENCODER_TMDS;
 			break;
 		case INTEL_DVO_CHIP_LVDS:
-			intel_encoder->cloneable = false;
+			intel_encoder->cloneable = 0;
 			drm_connector_init(dev, connector,
 					   &intel_dvo_connector_funcs,
 					   DRM_MODE_CONNECTOR_LVDS);
diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c
index 895fcb4fbd9..088fe9378a4 100644
--- a/drivers/gpu/drm/i915/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/intel_fbdev.c
@@ -57,18 +57,15 @@ static struct fb_ops intelfb_ops = {
 	.fb_debug_leave = drm_fb_helper_debug_leave,
 };
 
-static int intelfb_create(struct drm_fb_helper *helper,
-			  struct drm_fb_helper_surface_size *sizes)
+static int intelfb_alloc(struct drm_fb_helper *helper,
+			 struct drm_fb_helper_surface_size *sizes)
 {
 	struct intel_fbdev *ifbdev =
 		container_of(helper, struct intel_fbdev, helper);
-	struct drm_device *dev = helper->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct fb_info *info;
 	struct drm_framebuffer *fb;
+	struct drm_device *dev = helper->dev;
 	struct drm_mode_fb_cmd2 mode_cmd = {};
 	struct drm_i915_gem_object *obj;
-	struct device *device = &dev->pdev->dev;
 	int size, ret;
 
 	/* we don't do packed 24bpp */
@@ -94,16 +91,74 @@ static int intelfb_create(struct drm_fb_helper *helper,
 		goto out;
 	}
 
-	mutex_lock(&dev->struct_mutex);
-
 	/* Flush everything out, we'll be doing GTT only from now on */
 	ret = intel_pin_and_fence_fb_obj(dev, obj, NULL);
 	if (ret) {
-		DRM_ERROR("failed to pin fb: %d\n", ret);
+		DRM_ERROR("failed to pin obj: %d\n", ret);
 		goto out_unref;
 	}
 
-	info = framebuffer_alloc(0, device);
+	fb = __intel_framebuffer_create(dev, &mode_cmd, obj);
+	if (IS_ERR(fb)) {
+		ret = PTR_ERR(fb);
+		goto out_unpin;
+	}
+
+	ifbdev->fb = to_intel_framebuffer(fb);
+
+	return 0;
+
+out_unpin:
+	i915_gem_object_ggtt_unpin(obj);
+out_unref:
+	drm_gem_object_unreference(&obj->base);
+out:
+	return ret;
+}
+
+static int intelfb_create(struct drm_fb_helper *helper,
+			  struct drm_fb_helper_surface_size *sizes)
+{
+	struct intel_fbdev *ifbdev =
+		container_of(helper, struct intel_fbdev, helper);
+	struct intel_framebuffer *intel_fb = ifbdev->fb;
+	struct drm_device *dev = helper->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct fb_info *info;
+	struct drm_framebuffer *fb;
+	struct drm_i915_gem_object *obj;
+	int size, ret;
+	bool prealloc = false;
+
+	mutex_lock(&dev->struct_mutex);
+
+	if (intel_fb &&
+	    (sizes->fb_width > intel_fb->base.width ||
+	     sizes->fb_height > intel_fb->base.height)) {
+		DRM_DEBUG_KMS("BIOS fb too small (%dx%d), we require (%dx%d),"
+			      " releasing it\n",
+			      intel_fb->base.width, intel_fb->base.height,
+			      sizes->fb_width, sizes->fb_height);
+		drm_framebuffer_unreference(&intel_fb->base);
+		intel_fb = ifbdev->fb = NULL;
+	}
+	if (!intel_fb || WARN_ON(!intel_fb->obj)) {
+		DRM_DEBUG_KMS("no BIOS fb, allocating a new one\n");
+		ret = intelfb_alloc(helper, sizes);
+		if (ret)
+			goto out_unlock;
+		intel_fb = ifbdev->fb;
+	} else {
+		DRM_DEBUG_KMS("re-using BIOS fb\n");
+		prealloc = true;
+		sizes->fb_width = intel_fb->base.width;
+		sizes->fb_height = intel_fb->base.height;
+	}
+
+	obj = intel_fb->obj;
+	size = obj->base.size;
+
+	info = framebuffer_alloc(0, &dev->pdev->dev);
 	if (!info) {
 		ret = -ENOMEM;
 		goto out_unpin;
@@ -111,11 +166,7 @@ static int intelfb_create(struct drm_fb_helper *helper,
 
 	info->par = helper;
 
-	ret = intel_framebuffer_init(dev, &ifbdev->ifb, &mode_cmd, obj);
-	if (ret)
-		goto out_unpin;
-
-	fb = &ifbdev->ifb.base;
+	fb = &ifbdev->fb->base;
 
 	ifbdev->helper.fb = fb;
 	ifbdev->helper.fbdev = info;
@@ -161,7 +212,7 @@ static int intelfb_create(struct drm_fb_helper *helper,
 	 * If the object is stolen however, it will be full of whatever
 	 * garbage was left in there.
 	 */
-	if (ifbdev->ifb.obj->stolen)
+	if (ifbdev->fb->obj->stolen && !prealloc)
 		memset_io(info->screen_base, 0, info->screen_size);
 
 	/* Use default scratch pixmap (info->pixmap.flags = FB_PIXMAP_SYSTEM) */
@@ -170,17 +221,15 @@ static int intelfb_create(struct drm_fb_helper *helper,
 		      fb->width, fb->height,
 		      i915_gem_obj_ggtt_offset(obj), obj);
 
-
 	mutex_unlock(&dev->struct_mutex);
 	vga_switcheroo_client_fb_set(dev->pdev, info);
 	return 0;
 
 out_unpin:
-	i915_gem_object_unpin(obj);
-out_unref:
+	i915_gem_object_ggtt_unpin(obj);
 	drm_gem_object_unreference(&obj->base);
+out_unlock:
 	mutex_unlock(&dev->struct_mutex);
-out:
 	return ret;
 }
 
@@ -205,7 +254,206 @@ static void intel_crtc_fb_gamma_get(struct drm_crtc *crtc, u16 *red, u16 *green,
 	*blue = intel_crtc->lut_b[regno] << 8;
 }
 
+static struct drm_fb_helper_crtc *
+intel_fb_helper_crtc(struct drm_fb_helper *fb_helper, struct drm_crtc *crtc)
+{
+	int i;
+
+	for (i = 0; i < fb_helper->crtc_count; i++)
+		if (fb_helper->crtc_info[i].mode_set.crtc == crtc)
+			return &fb_helper->crtc_info[i];
+
+	return NULL;
+}
+
+/*
+ * Try to read the BIOS display configuration and use it for the initial
+ * fb configuration.
+ *
+ * The BIOS or boot loader will generally create an initial display
+ * configuration for us that includes some set of active pipes and displays.
+ * This routine tries to figure out which pipes and connectors are active
+ * and stuffs them into the crtcs and modes array given to us by the
+ * drm_fb_helper code.
+ *
+ * The overall sequence is:
+ *   intel_fbdev_init - from driver load
+ *     intel_fbdev_init_bios - initialize the intel_fbdev using BIOS data
+ *     drm_fb_helper_init - build fb helper structs
+ *     drm_fb_helper_single_add_all_connectors - more fb helper structs
+ *   intel_fbdev_initial_config - apply the config
+ *     drm_fb_helper_initial_config - call ->probe then register_framebuffer()
+ *         drm_setup_crtcs - build crtc config for fbdev
+ *           intel_fb_initial_config - find active connectors etc
+ *         drm_fb_helper_single_fb_probe - set up fbdev
+ *           intelfb_create - re-use or alloc fb, build out fbdev structs
+ *
+ * Note that we don't make special consideration whether we could actually
+ * switch to the selected modes without a full modeset. E.g. when the display
+ * is in VGA mode we need to recalculate watermarks and set a new high-res
+ * framebuffer anyway.
+ */
+static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
+				    struct drm_fb_helper_crtc **crtcs,
+				    struct drm_display_mode **modes,
+				    bool *enabled, int width, int height)
+{
+	struct drm_device *dev = fb_helper->dev;
+	int i, j;
+	bool *save_enabled;
+	bool fallback = true;
+	int num_connectors_enabled = 0;
+	int num_connectors_detected = 0;
+
+	/*
+	 * If the user specified any force options, just bail here
+	 * and use that config.
+	 */
+	for (i = 0; i < fb_helper->connector_count; i++) {
+		struct drm_fb_helper_connector *fb_conn;
+		struct drm_connector *connector;
+
+		fb_conn = fb_helper->connector_info[i];
+		connector = fb_conn->connector;
+
+		if (!enabled[i])
+			continue;
+
+		if (connector->force != DRM_FORCE_UNSPECIFIED)
+			return false;
+	}
+
+	save_enabled = kcalloc(dev->mode_config.num_connector, sizeof(bool),
+			       GFP_KERNEL);
+	if (!save_enabled)
+		return false;
+
+	memcpy(save_enabled, enabled, dev->mode_config.num_connector);
+
+	for (i = 0; i < fb_helper->connector_count; i++) {
+		struct drm_fb_helper_connector *fb_conn;
+		struct drm_connector *connector;
+		struct drm_encoder *encoder;
+		struct drm_fb_helper_crtc *new_crtc;
+
+		fb_conn = fb_helper->connector_info[i];
+		connector = fb_conn->connector;
+
+		if (connector->status == connector_status_connected)
+			num_connectors_detected++;
+
+		if (!enabled[i]) {
+			DRM_DEBUG_KMS("connector %s not enabled, skipping\n",
+				      connector->name);
+			continue;
+		}
+
+		encoder = connector->encoder;
+		if (!encoder || WARN_ON(!encoder->crtc)) {
+			DRM_DEBUG_KMS("connector %s has no encoder or crtc, skipping\n",
+				      connector->name);
+			enabled[i] = false;
+			continue;
+		}
+
+		num_connectors_enabled++;
+
+		new_crtc = intel_fb_helper_crtc(fb_helper, encoder->crtc);
+
+		/*
+		 * Make sure we're not trying to drive multiple connectors
+		 * with a single CRTC, since our cloning support may not
+		 * match the BIOS.
+		 */
+		for (j = 0; j < fb_helper->connector_count; j++) {
+			if (crtcs[j] == new_crtc) {
+				DRM_DEBUG_KMS("fallback: cloned configuration\n");
+				fallback = true;
+				goto out;
+			}
+		}
+
+		DRM_DEBUG_KMS("looking for cmdline mode on connector %s\n",
+			      connector->name);
+
+		/* go for command line mode first */
+		modes[i] = drm_pick_cmdline_mode(fb_conn, width, height);
+
+		/* try for preferred next */
+		if (!modes[i]) {
+			DRM_DEBUG_KMS("looking for preferred mode on connector %s\n",
+				      connector->name);
+			modes[i] = drm_has_preferred_mode(fb_conn, width,
+							  height);
+		}
+
+		/* No preferred mode marked by the EDID? Are there any modes? */
+		if (!modes[i] && !list_empty(&connector->modes)) {
+			DRM_DEBUG_KMS("using first mode listed on connector %s\n",
+				      connector->name);
+			modes[i] = list_first_entry(&connector->modes,
+						    struct drm_display_mode,
+						    head);
+		}
+
+		/* last resort: use current mode */
+		if (!modes[i]) {
+			/*
+			 * IMPORTANT: We want to use the adjusted mode (i.e.
+			 * after the panel fitter upscaling) as the initial
+			 * config, not the input mode, which is what crtc->mode
+			 * usually contains. But since our current fastboot
+			 * code puts a mode derived from the post-pfit timings
+			 * into crtc->mode this works out correctly. We don't
+			 * use hwmode anywhere right now, so use it for this
+			 * since the fb helper layer wants a pointer to
+			 * something we own.
+			 */
+			DRM_DEBUG_KMS("looking for current mode on connector %s\n",
+				      connector->name);
+			intel_mode_from_pipe_config(&encoder->crtc->hwmode,
+						    &to_intel_crtc(encoder->crtc)->config);
+			modes[i] = &encoder->crtc->hwmode;
+		}
+		crtcs[i] = new_crtc;
+
+		DRM_DEBUG_KMS("connector %s on pipe %d [CRTC:%d]: %dx%d%s\n",
+			      connector->name,
+			      pipe_name(to_intel_crtc(encoder->crtc)->pipe),
+			      encoder->crtc->base.id,
+			      modes[i]->hdisplay, modes[i]->vdisplay,
+			      modes[i]->flags & DRM_MODE_FLAG_INTERLACE ? "i" :"");
+
+		fallback = false;
+	}
+
+	/*
+	 * If the BIOS didn't enable everything it could, fall back to have the
+	 * same user experiencing of lighting up as much as possible like the
+	 * fbdev helper library.
+	 */
+	if (num_connectors_enabled != num_connectors_detected &&
+	    num_connectors_enabled < INTEL_INFO(dev)->num_pipes) {
+		DRM_DEBUG_KMS("fallback: Not all outputs enabled\n");
+		DRM_DEBUG_KMS("Enabled: %i, detected: %i\n", num_connectors_enabled,
+			      num_connectors_detected);
+		fallback = true;
+	}
+
+out:
+	if (fallback) {
+		DRM_DEBUG_KMS("Not using firmware configuration\n");
+		memcpy(enabled, save_enabled, dev->mode_config.num_connector);
+		kfree(save_enabled);
+		return false;
+	}
+
+	kfree(save_enabled);
+	return true;
+}
+
 static struct drm_fb_helper_funcs intel_fb_helper_funcs = {
+	.initial_config = intel_fb_initial_config,
 	.gamma_set = intel_crtc_fb_gamma_set,
 	.gamma_get = intel_crtc_fb_gamma_get,
 	.fb_probe = intelfb_create,
@@ -227,8 +475,139 @@ static void intel_fbdev_destroy(struct drm_device *dev,
 
 	drm_fb_helper_fini(&ifbdev->helper);
 
-	drm_framebuffer_unregister_private(&ifbdev->ifb.base);
-	intel_framebuffer_fini(&ifbdev->ifb);
+	drm_framebuffer_unregister_private(&ifbdev->fb->base);
+	drm_framebuffer_remove(&ifbdev->fb->base);
+}
+
+/*
+ * Build an intel_fbdev struct using a BIOS allocated framebuffer, if possible.
+ * The core display code will have read out the current plane configuration,
+ * so we use that to figure out if there's an object for us to use as the
+ * fb, and if so, we re-use it for the fbdev configuration.
+ *
+ * Note we only support a single fb shared across pipes for boot (mostly for
+ * fbcon), so we just find the biggest and use that.
+ */
+static bool intel_fbdev_init_bios(struct drm_device *dev,
+				 struct intel_fbdev *ifbdev)
+{
+	struct intel_framebuffer *fb = NULL;
+	struct drm_crtc *crtc;
+	struct intel_crtc *intel_crtc;
+	struct intel_plane_config *plane_config = NULL;
+	unsigned int max_size = 0;
+
+	if (!i915.fastboot)
+		return false;
+
+	/* Find the largest fb */
+	for_each_crtc(dev, crtc) {
+		intel_crtc = to_intel_crtc(crtc);
+
+		if (!intel_crtc->active || !crtc->primary->fb) {
+			DRM_DEBUG_KMS("pipe %c not active or no fb, skipping\n",
+				      pipe_name(intel_crtc->pipe));
+			continue;
+		}
+
+		if (intel_crtc->plane_config.size > max_size) {
+			DRM_DEBUG_KMS("found possible fb from plane %c\n",
+				      pipe_name(intel_crtc->pipe));
+			plane_config = &intel_crtc->plane_config;
+			fb = to_intel_framebuffer(crtc->primary->fb);
+			max_size = plane_config->size;
+		}
+	}
+
+	if (!fb) {
+		DRM_DEBUG_KMS("no active fbs found, not using BIOS config\n");
+		goto out;
+	}
+
+	/* Now make sure all the pipes will fit into it */
+	for_each_crtc(dev, crtc) {
+		unsigned int cur_size;
+
+		intel_crtc = to_intel_crtc(crtc);
+
+		if (!intel_crtc->active) {
+			DRM_DEBUG_KMS("pipe %c not active, skipping\n",
+				      pipe_name(intel_crtc->pipe));
+			continue;
+		}
+
+		DRM_DEBUG_KMS("checking plane %c for BIOS fb\n",
+			      pipe_name(intel_crtc->pipe));
+
+		/*
+		 * See if the plane fb we found above will fit on this
+		 * pipe.  Note we need to use the selected fb's pitch and bpp
+		 * rather than the current pipe's, since they differ.
+		 */
+		cur_size = intel_crtc->config.adjusted_mode.crtc_hdisplay;
+		cur_size = cur_size * fb->base.bits_per_pixel / 8;
+		if (fb->base.pitches[0] < cur_size) {
+			DRM_DEBUG_KMS("fb not wide enough for plane %c (%d vs %d)\n",
+				      pipe_name(intel_crtc->pipe),
+				      cur_size, fb->base.pitches[0]);
+			plane_config = NULL;
+			fb = NULL;
+			break;
+		}
+
+		cur_size = intel_crtc->config.adjusted_mode.crtc_vdisplay;
+		cur_size = ALIGN(cur_size, plane_config->tiled ? (IS_GEN2(dev) ? 16 : 8) : 1);
+		cur_size *= fb->base.pitches[0];
+		DRM_DEBUG_KMS("pipe %c area: %dx%d, bpp: %d, size: %d\n",
+			      pipe_name(intel_crtc->pipe),
+			      intel_crtc->config.adjusted_mode.crtc_hdisplay,
+			      intel_crtc->config.adjusted_mode.crtc_vdisplay,
+			      fb->base.bits_per_pixel,
+			      cur_size);
+
+		if (cur_size > max_size) {
+			DRM_DEBUG_KMS("fb not big enough for plane %c (%d vs %d)\n",
+				      pipe_name(intel_crtc->pipe),
+				      cur_size, max_size);
+			plane_config = NULL;
+			fb = NULL;
+			break;
+		}
+
+		DRM_DEBUG_KMS("fb big enough for plane %c (%d >= %d)\n",
+			      pipe_name(intel_crtc->pipe),
+			      max_size, cur_size);
+	}
+
+	if (!fb) {
+		DRM_DEBUG_KMS("BIOS fb not suitable for all pipes, not using\n");
+		goto out;
+	}
+
+	ifbdev->preferred_bpp = fb->base.bits_per_pixel;
+	ifbdev->fb = fb;
+
+	drm_framebuffer_reference(&ifbdev->fb->base);
+
+	/* Final pass to check if any active pipes don't have fbs */
+	for_each_crtc(dev, crtc) {
+		intel_crtc = to_intel_crtc(crtc);
+
+		if (!intel_crtc->active)
+			continue;
+
+		WARN(!crtc->primary->fb,
+		     "re-used BIOS config but lost an fb on crtc %d\n",
+		     crtc->base.id);
+	}
+
+
+	DRM_DEBUG_KMS("using BIOS fb for initial console\n");
+	return true;
+
+out:
+
+	return false;
 }
 
 int intel_fbdev_init(struct drm_device *dev)
@@ -237,21 +616,25 @@ int intel_fbdev_init(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
-	ifbdev = kzalloc(sizeof(*ifbdev), GFP_KERNEL);
-	if (!ifbdev)
+	if (WARN_ON(INTEL_INFO(dev)->num_pipes == 0))
+		return -ENODEV;
+
+	ifbdev = kzalloc(sizeof(struct intel_fbdev), GFP_KERNEL);
+	if (ifbdev == NULL)
 		return -ENOMEM;
 
-	dev_priv->fbdev = ifbdev;
 	ifbdev->helper.funcs = &intel_fb_helper_funcs;
+	if (!intel_fbdev_init_bios(dev, ifbdev))
+		ifbdev->preferred_bpp = 32;
 
 	ret = drm_fb_helper_init(dev, &ifbdev->helper,
-				 INTEL_INFO(dev)->num_pipes,
-				 4);
+				 INTEL_INFO(dev)->num_pipes, 4);
 	if (ret) {
 		kfree(ifbdev);
 		return ret;
 	}
 
+	dev_priv->fbdev = ifbdev;
 	drm_fb_helper_single_add_all_connectors(&ifbdev->helper);
 
 	return 0;
@@ -260,9 +643,10 @@ int intel_fbdev_init(struct drm_device *dev)
 void intel_fbdev_initial_config(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_fbdev *ifbdev = dev_priv->fbdev;
 
 	/* Due to peculiar init order wrt to hpd handling this is separate. */
-	drm_fb_helper_initial_config(&dev_priv->fbdev->helper, 32);
+	drm_fb_helper_initial_config(&ifbdev->helper, ifbdev->preferred_bpp);
 }
 
 void intel_fbdev_fini(struct drm_device *dev)
@@ -291,18 +675,17 @@ void intel_fbdev_set_suspend(struct drm_device *dev, int state)
 	 * been restored from swap. If the object is stolen however, it will be
 	 * full of whatever garbage was left in there.
 	 */
-	if (state == FBINFO_STATE_RUNNING && ifbdev->ifb.obj->stolen)
+	if (state == FBINFO_STATE_RUNNING && ifbdev->fb->obj->stolen)
 		memset_io(info->screen_base, 0, info->screen_size);
 
 	fb_set_suspend(info, state);
 }
 
-MODULE_LICENSE("GPL and additional rights");
-
 void intel_fbdev_output_poll_changed(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	drm_fb_helper_hotplug_event(&dev_priv->fbdev->helper);
+	if (dev_priv->fbdev)
+		drm_fb_helper_hotplug_event(&dev_priv->fbdev->helper);
 }
 
 void intel_fbdev_restore_mode(struct drm_device *dev)
@@ -310,14 +693,10 @@ void intel_fbdev_restore_mode(struct drm_device *dev)
 	int ret;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (INTEL_INFO(dev)->num_pipes == 0)
+	if (!dev_priv->fbdev)
 		return;
 
-	drm_modeset_lock_all(dev);
-
-	ret = drm_fb_helper_restore_fbdev_mode(&dev_priv->fbdev->helper);
+	ret = drm_fb_helper_restore_fbdev_mode_unlocked(&dev_priv->fbdev->helper);
 	if (ret)
 		DRM_DEBUG("failed to restore crtc mode\n");
-
-	drm_modeset_unlock_all(dev);
 }
diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c
index 03f9ca70530..eee2bbec295 100644
--- a/drivers/gpu/drm/i915/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/intel_hdmi.c
@@ -113,7 +113,8 @@ static u32 hsw_infoframe_enable(enum hdmi_infoframe_type type)
 }
 
 static u32 hsw_infoframe_data_reg(enum hdmi_infoframe_type type,
-				  enum transcoder cpu_transcoder)
+				  enum transcoder cpu_transcoder,
+				  struct drm_i915_private *dev_priv)
 {
 	switch (type) {
 	case HDMI_INFOFRAME_TYPE_AVI:
@@ -130,9 +131,9 @@ static u32 hsw_infoframe_data_reg(enum hdmi_infoframe_type type,
 
 static void g4x_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 val = I915_READ(VIDEO_DIP_CTL);
@@ -167,9 +168,9 @@ static void g4x_write_infoframe(struct drm_encoder *encoder,
 
 static void ibx_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
@@ -205,9 +206,9 @@ static void ibx_write_infoframe(struct drm_encoder *encoder,
 
 static void cpt_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
@@ -246,9 +247,9 @@ static void cpt_write_infoframe(struct drm_encoder *encoder,
 
 static void vlv_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
@@ -284,9 +285,9 @@ static void vlv_write_infoframe(struct drm_encoder *encoder,
 
 static void hsw_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
@@ -296,7 +297,8 @@ static void hsw_write_infoframe(struct drm_encoder *encoder,
 	u32 val = I915_READ(ctl_reg);
 
 	data_reg = hsw_infoframe_data_reg(type,
-					  intel_crtc->config.cpu_transcoder);
+					  intel_crtc->config.cpu_transcoder,
+					  dev_priv);
 	if (data_reg == 0)
 		return;
 
@@ -416,6 +418,7 @@ intel_hdmi_set_hdmi_infoframe(struct drm_encoder *encoder,
 }
 
 static void g4x_set_infoframes(struct drm_encoder *encoder,
+			       bool enable,
 			       struct drm_display_mode *adjusted_mode)
 {
 	struct drm_i915_private *dev_priv = encoder->dev->dev_private;
@@ -423,7 +426,7 @@ static void g4x_set_infoframes(struct drm_encoder *encoder,
 	struct intel_hdmi *intel_hdmi = &intel_dig_port->hdmi;
 	u32 reg = VIDEO_DIP_CTL;
 	u32 val = I915_READ(reg);
-	u32 port;
+	u32 port = VIDEO_DIP_PORT(intel_dig_port->port);
 
 	assert_hdmi_port_disabled(intel_hdmi);
 
@@ -438,7 +441,7 @@ static void g4x_set_infoframes(struct drm_encoder *encoder,
 	 * either. */
 	val |= VIDEO_DIP_SELECT_AVI | VIDEO_DIP_FREQ_VSYNC;
 
-	if (!intel_hdmi->has_hdmi_sink) {
+	if (!enable) {
 		if (!(val & VIDEO_DIP_ENABLE))
 			return;
 		val &= ~VIDEO_DIP_ENABLE;
@@ -447,18 +450,6 @@ static void g4x_set_infoframes(struct drm_encoder *encoder,
 		return;
 	}
 
-	switch (intel_dig_port->port) {
-	case PORT_B:
-		port = VIDEO_DIP_PORT_B;
-		break;
-	case PORT_C:
-		port = VIDEO_DIP_PORT_C;
-		break;
-	default:
-		BUG();
-		return;
-	}
-
 	if (port != (val & VIDEO_DIP_PORT_MASK)) {
 		if (val & VIDEO_DIP_ENABLE) {
 			val &= ~VIDEO_DIP_ENABLE;
@@ -481,6 +472,7 @@ static void g4x_set_infoframes(struct drm_encoder *encoder,
 }
 
 static void ibx_set_infoframes(struct drm_encoder *encoder,
+			       bool enable,
 			       struct drm_display_mode *adjusted_mode)
 {
 	struct drm_i915_private *dev_priv = encoder->dev->dev_private;
@@ -489,14 +481,14 @@ static void ibx_set_infoframes(struct drm_encoder *encoder,
 	struct intel_hdmi *intel_hdmi = &intel_dig_port->hdmi;
 	u32 reg = TVIDEO_DIP_CTL(intel_crtc->pipe);
 	u32 val = I915_READ(reg);
-	u32 port;
+	u32 port = VIDEO_DIP_PORT(intel_dig_port->port);
 
 	assert_hdmi_port_disabled(intel_hdmi);
 
 	/* See the big comment in g4x_set_infoframes() */
 	val |= VIDEO_DIP_SELECT_AVI | VIDEO_DIP_FREQ_VSYNC;
 
-	if (!intel_hdmi->has_hdmi_sink) {
+	if (!enable) {
 		if (!(val & VIDEO_DIP_ENABLE))
 			return;
 		val &= ~VIDEO_DIP_ENABLE;
@@ -505,21 +497,6 @@ static void ibx_set_infoframes(struct drm_encoder *encoder,
 		return;
 	}
 
-	switch (intel_dig_port->port) {
-	case PORT_B:
-		port = VIDEO_DIP_PORT_B;
-		break;
-	case PORT_C:
-		port = VIDEO_DIP_PORT_C;
-		break;
-	case PORT_D:
-		port = VIDEO_DIP_PORT_D;
-		break;
-	default:
-		BUG();
-		return;
-	}
-
 	if (port != (val & VIDEO_DIP_PORT_MASK)) {
 		if (val & VIDEO_DIP_ENABLE) {
 			val &= ~VIDEO_DIP_ENABLE;
@@ -543,6 +520,7 @@ static void ibx_set_infoframes(struct drm_encoder *encoder,
 }
 
 static void cpt_set_infoframes(struct drm_encoder *encoder,
+			       bool enable,
 			       struct drm_display_mode *adjusted_mode)
 {
 	struct drm_i915_private *dev_priv = encoder->dev->dev_private;
@@ -556,7 +534,7 @@ static void cpt_set_infoframes(struct drm_encoder *encoder,
 	/* See the big comment in g4x_set_infoframes() */
 	val |= VIDEO_DIP_SELECT_AVI | VIDEO_DIP_FREQ_VSYNC;
 
-	if (!intel_hdmi->has_hdmi_sink) {
+	if (!enable) {
 		if (!(val & VIDEO_DIP_ENABLE))
 			return;
 		val &= ~(VIDEO_DIP_ENABLE | VIDEO_DIP_ENABLE_AVI);
@@ -579,20 +557,23 @@ static void cpt_set_infoframes(struct drm_encoder *encoder,
 }
 
 static void vlv_set_infoframes(struct drm_encoder *encoder,
+			       bool enable,
 			       struct drm_display_mode *adjusted_mode)
 {
 	struct drm_i915_private *dev_priv = encoder->dev->dev_private;
+	struct intel_digital_port *intel_dig_port = enc_to_dig_port(encoder);
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder);
 	u32 reg = VLV_TVIDEO_DIP_CTL(intel_crtc->pipe);
 	u32 val = I915_READ(reg);
+	u32 port = VIDEO_DIP_PORT(intel_dig_port->port);
 
 	assert_hdmi_port_disabled(intel_hdmi);
 
 	/* See the big comment in g4x_set_infoframes() */
 	val |= VIDEO_DIP_SELECT_AVI | VIDEO_DIP_FREQ_VSYNC;
 
-	if (!intel_hdmi->has_hdmi_sink) {
+	if (!enable) {
 		if (!(val & VIDEO_DIP_ENABLE))
 			return;
 		val &= ~VIDEO_DIP_ENABLE;
@@ -601,9 +582,19 @@ static void vlv_set_infoframes(struct drm_encoder *encoder,
 		return;
 	}
 
+	if (port != (val & VIDEO_DIP_PORT_MASK)) {
+		if (val & VIDEO_DIP_ENABLE) {
+			val &= ~VIDEO_DIP_ENABLE;
+			I915_WRITE(reg, val);
+			POSTING_READ(reg);
+		}
+		val &= ~VIDEO_DIP_PORT_MASK;
+		val |= port;
+	}
+
 	val |= VIDEO_DIP_ENABLE;
-	val &= ~(VIDEO_DIP_ENABLE_VENDOR | VIDEO_DIP_ENABLE_GAMUT |
-		 VIDEO_DIP_ENABLE_GCP);
+	val &= ~(VIDEO_DIP_ENABLE_AVI | VIDEO_DIP_ENABLE_VENDOR |
+		 VIDEO_DIP_ENABLE_GAMUT | VIDEO_DIP_ENABLE_GCP);
 
 	I915_WRITE(reg, val);
 	POSTING_READ(reg);
@@ -614,6 +605,7 @@ static void vlv_set_infoframes(struct drm_encoder *encoder,
 }
 
 static void hsw_set_infoframes(struct drm_encoder *encoder,
+			       bool enable,
 			       struct drm_display_mode *adjusted_mode)
 {
 	struct drm_i915_private *dev_priv = encoder->dev->dev_private;
@@ -624,7 +616,7 @@ static void hsw_set_infoframes(struct drm_encoder *encoder,
 
 	assert_hdmi_port_disabled(intel_hdmi);
 
-	if (!intel_hdmi->has_hdmi_sink) {
+	if (!enable) {
 		I915_WRITE(reg, 0);
 		POSTING_READ(reg);
 		return;
@@ -641,7 +633,7 @@ static void hsw_set_infoframes(struct drm_encoder *encoder,
 	intel_hdmi_set_hdmi_infoframe(encoder, adjusted_mode);
 }
 
-static void intel_hdmi_mode_set(struct intel_encoder *encoder)
+static void intel_hdmi_prepare(struct intel_encoder *encoder)
 {
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -663,27 +655,26 @@ static void intel_hdmi_mode_set(struct intel_encoder *encoder)
 	else
 		hdmi_val |= SDVO_COLOR_FORMAT_8bpc;
 
-	/* Required on CPT */
-	if (intel_hdmi->has_hdmi_sink && HAS_PCH_CPT(dev))
+	if (crtc->config.has_hdmi_sink)
 		hdmi_val |= HDMI_MODE_SELECT_HDMI;
 
-	if (intel_hdmi->has_audio) {
+	if (crtc->config.has_audio) {
+		WARN_ON(!crtc->config.has_hdmi_sink);
 		DRM_DEBUG_DRIVER("Enabling HDMI audio on pipe %c\n",
 				 pipe_name(crtc->pipe));
 		hdmi_val |= SDVO_AUDIO_ENABLE;
-		hdmi_val |= HDMI_MODE_SELECT_HDMI;
 		intel_write_eld(&encoder->base, adjusted_mode);
 	}
 
 	if (HAS_PCH_CPT(dev))
 		hdmi_val |= SDVO_PIPE_SEL_CPT(crtc->pipe);
+	else if (IS_CHERRYVIEW(dev))
+		hdmi_val |= SDVO_PIPE_SEL_CHV(crtc->pipe);
 	else
 		hdmi_val |= SDVO_PIPE_SEL(crtc->pipe);
 
 	I915_WRITE(intel_hdmi->hdmi_reg, hdmi_val);
 	POSTING_READ(intel_hdmi->hdmi_reg);
-
-	intel_hdmi->set_infoframes(&encoder->base, adjusted_mode);
 }
 
 static bool intel_hdmi_get_hw_state(struct intel_encoder *encoder,
@@ -692,8 +683,13 @@ static bool intel_hdmi_get_hw_state(struct intel_encoder *encoder,
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
+	enum intel_display_power_domain power_domain;
 	u32 tmp;
 
+	power_domain = intel_display_port_power_domain(encoder);
+	if (!intel_display_power_enabled(dev_priv, power_domain))
+		return false;
+
 	tmp = I915_READ(intel_hdmi->hdmi_reg);
 
 	if (!(tmp & SDVO_ENABLE))
@@ -701,6 +697,8 @@ static bool intel_hdmi_get_hw_state(struct intel_encoder *encoder,
 
 	if (HAS_PCH_CPT(dev))
 		*pipe = PORT_TO_PIPE_CPT(tmp);
+	else if (IS_CHERRYVIEW(dev))
+		*pipe = SDVO_PORT_TO_PIPE_CHV(tmp);
 	else
 		*pipe = PORT_TO_PIPE(tmp);
 
@@ -727,6 +725,12 @@ static void intel_hdmi_get_config(struct intel_encoder *encoder,
 	else
 		flags |= DRM_MODE_FLAG_NVSYNC;
 
+	if (tmp & HDMI_MODE_SELECT_HDMI)
+		pipe_config->has_hdmi_sink = true;
+
+	if (tmp & HDMI_MODE_SELECT_HDMI)
+		pipe_config->has_audio = true;
+
 	pipe_config->adjusted_mode.flags |= flags;
 
 	if ((tmp & SDVO_COLOR_FORMAT_MASK) == HDMI_COLOR_FORMAT_12bpc)
@@ -749,7 +753,7 @@ static void intel_enable_hdmi(struct intel_encoder *encoder)
 	u32 temp;
 	u32 enable_bits = SDVO_ENABLE;
 
-	if (intel_hdmi->has_audio)
+	if (intel_crtc->config.has_audio)
 		enable_bits |= SDVO_AUDIO_ENABLE;
 
 	temp = I915_READ(intel_hdmi->hdmi_reg);
@@ -841,11 +845,11 @@ static void intel_disable_hdmi(struct intel_encoder *encoder)
 	}
 }
 
-static int hdmi_portclock_limit(struct intel_hdmi *hdmi)
+static int hdmi_portclock_limit(struct intel_hdmi *hdmi, bool respect_dvi_limit)
 {
 	struct drm_device *dev = intel_hdmi_to_dev(hdmi);
 
-	if (IS_G4X(dev))
+	if ((respect_dvi_limit && !hdmi->has_hdmi_sink) || IS_G4X(dev))
 		return 165000;
 	else if (IS_HASWELL(dev) || INTEL_INFO(dev)->gen >= 8)
 		return 300000;
@@ -853,10 +857,12 @@ static int hdmi_portclock_limit(struct intel_hdmi *hdmi)
 		return 225000;
 }
 
-static int intel_hdmi_mode_valid(struct drm_connector *connector,
-				 struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_hdmi_mode_valid(struct drm_connector *connector,
+		      struct drm_display_mode *mode)
 {
-	if (mode->clock > hdmi_portclock_limit(intel_attached_hdmi(connector)))
+	if (mode->clock > hdmi_portclock_limit(intel_attached_hdmi(connector),
+					       true))
 		return MODE_CLOCK_HIGH;
 	if (mode->clock < 20000)
 		return MODE_CLOCK_LOW;
@@ -867,6 +873,30 @@ static int intel_hdmi_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
+static bool hdmi_12bpc_possible(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct intel_encoder *encoder;
+	int count = 0, count_hdmi = 0;
+
+	if (!HAS_PCH_SPLIT(dev))
+		return false;
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, base.head) {
+		if (encoder->new_crtc != crtc)
+			continue;
+
+		count_hdmi += encoder->type == INTEL_OUTPUT_HDMI;
+		count++;
+	}
+
+	/*
+	 * HDMI 12bpc affects the clocks, so it's only possible
+	 * when not cloning with other encoder types.
+	 */
+	return count_hdmi > 0 && count_hdmi == count;
+}
+
 bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 			       struct intel_crtc_config *pipe_config)
 {
@@ -874,12 +904,14 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_display_mode *adjusted_mode = &pipe_config->adjusted_mode;
 	int clock_12bpc = pipe_config->adjusted_mode.crtc_clock * 3 / 2;
-	int portclock_limit = hdmi_portclock_limit(intel_hdmi);
+	int portclock_limit = hdmi_portclock_limit(intel_hdmi, false);
 	int desired_bpp;
 
+	pipe_config->has_hdmi_sink = intel_hdmi->has_hdmi_sink;
+
 	if (intel_hdmi->color_range_auto) {
 		/* See CEA-861-E - 5.1 Default Encoding Parameters */
-		if (intel_hdmi->has_hdmi_sink &&
+		if (pipe_config->has_hdmi_sink &&
 		    drm_match_cea_mode(adjusted_mode) > 1)
 			intel_hdmi->color_range = HDMI_COLOR_RANGE_16_235;
 		else
@@ -892,14 +924,18 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 	if (HAS_PCH_SPLIT(dev) && !HAS_DDI(dev))
 		pipe_config->has_pch_encoder = true;
 
+	if (pipe_config->has_hdmi_sink && intel_hdmi->has_audio)
+		pipe_config->has_audio = true;
+
 	/*
 	 * HDMI is either 12 or 8, so if the display lets 10bpc sneak
 	 * through, clamp it down. Note that g4x/vlv don't support 12bpc hdmi
 	 * outputs. We also need to check that the higher clock still fits
 	 * within limits.
 	 */
-	if (pipe_config->pipe_bpp > 8*3 && clock_12bpc <= portclock_limit
-	    && HAS_PCH_SPLIT(dev)) {
+	if (pipe_config->pipe_bpp > 8*3 && pipe_config->has_hdmi_sink &&
+	    clock_12bpc <= portclock_limit &&
+	    hdmi_12bpc_possible(encoder->new_crtc)) {
 		DRM_DEBUG_KMS("picking bpc to 12 for HDMI output\n");
 		desired_bpp = 12*3;
 
@@ -933,10 +969,14 @@ intel_hdmi_detect(struct drm_connector *connector, bool force)
 	struct intel_encoder *intel_encoder = &intel_dig_port->base;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct edid *edid;
+	enum intel_display_power_domain power_domain;
 	enum drm_connector_status status = connector_status_disconnected;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector));
+		      connector->base.id, connector->name);
+
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_get(dev_priv, power_domain);
 
 	intel_hdmi->has_hdmi_sink = false;
 	intel_hdmi->has_audio = false;
@@ -965,31 +1005,48 @@ intel_hdmi_detect(struct drm_connector *connector, bool force)
 		intel_encoder->type = INTEL_OUTPUT_HDMI;
 	}
 
+	intel_display_power_put(dev_priv, power_domain);
+
 	return status;
 }
 
 static int intel_hdmi_get_modes(struct drm_connector *connector)
 {
-	struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
+	struct intel_encoder *intel_encoder = intel_attached_encoder(connector);
+	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&intel_encoder->base);
 	struct drm_i915_private *dev_priv = connector->dev->dev_private;
+	enum intel_display_power_domain power_domain;
+	int ret;
 
 	/* We should parse the EDID data and find out if it's an HDMI sink so
 	 * we can send audio to it.
 	 */
 
-	return intel_ddc_get_modes(connector,
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_get(dev_priv, power_domain);
+
+	ret = intel_ddc_get_modes(connector,
 				   intel_gmbus_get_adapter(dev_priv,
 							   intel_hdmi->ddc_bus));
+
+	intel_display_power_put(dev_priv, power_domain);
+
+	return ret;
 }
 
 static bool
 intel_hdmi_detect_audio(struct drm_connector *connector)
 {
-	struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
+	struct intel_encoder *intel_encoder = intel_attached_encoder(connector);
+	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&intel_encoder->base);
 	struct drm_i915_private *dev_priv = connector->dev->dev_private;
+	enum intel_display_power_domain power_domain;
 	struct edid *edid;
 	bool has_audio = false;
 
+	power_domain = intel_display_port_power_domain(intel_encoder);
+	intel_display_power_get(dev_priv, power_domain);
+
 	edid = drm_get_edid(connector,
 			    intel_gmbus_get_adapter(dev_priv,
 						    intel_hdmi->ddc_bus));
@@ -999,6 +1056,8 @@ intel_hdmi_detect_audio(struct drm_connector *connector)
 		kfree(edid);
 	}
 
+	intel_display_power_put(dev_priv, power_domain);
+
 	return has_audio;
 }
 
@@ -1074,57 +1133,67 @@ done:
 	return 0;
 }
 
+static void intel_hdmi_pre_enable(struct intel_encoder *encoder)
+{
+	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
+	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
+	struct drm_display_mode *adjusted_mode =
+		&intel_crtc->config.adjusted_mode;
+
+	intel_hdmi_prepare(encoder);
+
+	intel_hdmi->set_infoframes(&encoder->base,
+				   intel_crtc->config.has_hdmi_sink,
+				   adjusted_mode);
+}
+
 static void vlv_hdmi_pre_enable(struct intel_encoder *encoder)
 {
 	struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
+	struct intel_hdmi *intel_hdmi = &dport->hdmi;
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
-	int port = vlv_dport_to_channel(dport);
+	struct drm_display_mode *adjusted_mode =
+		&intel_crtc->config.adjusted_mode;
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 	u32 val;
 
-	if (!IS_VALLEYVIEW(dev))
-		return;
-
 	/* Enable clock channels for this port */
 	mutex_lock(&dev_priv->dpio_lock);
-	val = vlv_dpio_read(dev_priv, pipe, DPIO_DATA_LANE_A(port));
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW8(port));
 	val = 0;
 	if (pipe)
 		val |= (1<<21);
 	else
 		val &= ~(1<<21);
 	val |= 0x001000c4;
-	vlv_dpio_write(dev_priv, pipe, DPIO_DATA_CHANNEL(port), val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW8(port), val);
 
 	/* HDMI 1.0V-2dB */
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port), 0);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL4(port),
-			 0x2b245f5f);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL2(port),
-			 0x5578b83a);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL3(port),
-			 0x0c782040);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX3_SWING_CTL4(port),
-			 0x2b247878);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_STAGGER0(port), 0x00030000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CTL_OVER1(port),
-			 0x00002000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port),
-			 DPIO_TX_OCALINIT_EN);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), 0);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW4(port), 0x2b245f5f);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW2(port), 0x5578b83a);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW3(port), 0x0c782040);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX3_DW4(port), 0x2b247878);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW11(port), 0x00030000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW9(port), 0x00002000);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), DPIO_TX_OCALINIT_EN);
 
 	/* Program lane clock */
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLOCKBUF0(port),
-			 0x00760018);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLOCKBUF8(port),
-			 0x00400888);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW14(port), 0x00760018);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW23(port), 0x00400888);
 	mutex_unlock(&dev_priv->dpio_lock);
 
+	intel_hdmi->set_infoframes(&encoder->base,
+				   intel_crtc->config.has_hdmi_sink,
+				   adjusted_mode);
+
 	intel_enable_hdmi(encoder);
 
-	vlv_wait_port_ready(dev_priv, port);
+	vlv_wait_port_ready(dev_priv, dport);
 }
 
 static void vlv_hdmi_pre_pll_enable(struct intel_encoder *encoder)
@@ -1134,32 +1203,29 @@ static void vlv_hdmi_pre_pll_enable(struct intel_encoder *encoder)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 
-	if (!IS_VALLEYVIEW(dev))
-		return;
+	intel_hdmi_prepare(encoder);
 
 	/* Program Tx lane resets to default */
 	mutex_lock(&dev_priv->dpio_lock);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_TX(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW0(port),
 			 DPIO_PCS_TX_LANE2_RESET |
 			 DPIO_PCS_TX_LANE1_RESET);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLK(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW1(port),
 			 DPIO_PCS_CLK_CRI_RXEB_EIOS_EN |
 			 DPIO_PCS_CLK_CRI_RXDIGFILTSG_EN |
 			 (1<<DPIO_PCS_CLK_DATAWIDTH_SHIFT) |
 			 DPIO_PCS_CLK_SOFT_RESET);
 
 	/* Fix up inter-pair skew failure */
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_STAGGER1(port), 0x00750f00);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_CTL(port), 0x00001500);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_LANE(port), 0x40400000);
-
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CTL_OVER1(port),
-			 0x00002000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port),
-			 DPIO_TX_OCALINIT_EN);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW12(port), 0x00750f00);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW11(port), 0x00001500);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW14(port), 0x40400000);
+
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW9(port), 0x00002000);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), DPIO_TX_OCALINIT_EN);
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
@@ -1169,14 +1235,160 @@ static void vlv_hdmi_post_disable(struct intel_encoder *encoder)
 	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 
 	/* Reset lanes to avoid HDMI flicker (VLV w/a) */
 	mutex_lock(&dev_priv->dpio_lock);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_TX(port), 0x00000000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLK(port), 0x00e00060);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW0(port), 0x00000000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW1(port), 0x00e00060);
+	mutex_unlock(&dev_priv->dpio_lock);
+}
+
+static void chv_hdmi_post_disable(struct intel_encoder *encoder)
+{
+	struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc =
+		to_intel_crtc(encoder->base.crtc);
+	enum dpio_channel ch = vlv_dport_to_channel(dport);
+	enum pipe pipe = intel_crtc->pipe;
+	u32 val;
+
+	mutex_lock(&dev_priv->dpio_lock);
+
+	/* Propagate soft reset to data lane reset */
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW1(ch));
+	val |= CHV_PCS_REQ_SOFTRESET_EN;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW1(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW1(ch));
+	val |= CHV_PCS_REQ_SOFTRESET_EN;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW1(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW0(ch));
+	val &= ~(DPIO_PCS_TX_LANE2_RESET | DPIO_PCS_TX_LANE1_RESET);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW0(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW0(ch));
+	val &= ~(DPIO_PCS_TX_LANE2_RESET | DPIO_PCS_TX_LANE1_RESET);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW0(ch), val);
+
+	mutex_unlock(&dev_priv->dpio_lock);
+}
+
+static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
+{
+	struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc =
+		to_intel_crtc(encoder->base.crtc);
+	enum dpio_channel ch = vlv_dport_to_channel(dport);
+	int pipe = intel_crtc->pipe;
+	int data, i;
+	u32 val;
+
+	mutex_lock(&dev_priv->dpio_lock);
+
+	/* Deassert soft data lane reset*/
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW1(ch));
+	val |= CHV_PCS_REQ_SOFTRESET_EN;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW1(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW1(ch));
+	val |= CHV_PCS_REQ_SOFTRESET_EN;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW1(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW0(ch));
+	val |= (DPIO_PCS_TX_LANE2_RESET | DPIO_PCS_TX_LANE1_RESET);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW0(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW0(ch));
+	val |= (DPIO_PCS_TX_LANE2_RESET | DPIO_PCS_TX_LANE1_RESET);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW0(ch), val);
+
+	/* Program Tx latency optimal setting */
+	for (i = 0; i < 4; i++) {
+		/* Set the latency optimal bit */
+		data = (i == 1) ? 0x0 : 0x6;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW11(ch, i),
+				data << DPIO_FRC_LATENCY_SHFIT);
+
+		/* Set the upar bit */
+		data = (i == 1) ? 0x0 : 0x1;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW14(ch, i),
+				data << DPIO_UPAR_SHIFT);
+	}
+
+	/* Data lane stagger programming */
+	/* FIXME: Fix up value only after power analysis */
+
+	/* Clear calc init */
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW10(ch));
+	val &= ~(DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW10(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW10(ch));
+	val &= ~(DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW10(ch), val);
+
+	/* FIXME: Program the support xxx V-dB */
+	/* Use 800mV-0dB */
+	for (i = 0; i < 4; i++) {
+		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW4(ch, i));
+		val &= ~DPIO_SWING_DEEMPH9P5_MASK;
+		val |= 128 << DPIO_SWING_DEEMPH9P5_SHIFT;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW4(ch, i), val);
+	}
+
+	for (i = 0; i < 4; i++) {
+		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW2(ch, i));
+		val &= ~DPIO_SWING_MARGIN_MASK;
+		val |= 102 << DPIO_SWING_MARGIN_SHIFT;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW2(ch, i), val);
+	}
+
+	/* Disable unique transition scale */
+	for (i = 0; i < 4; i++) {
+		val = vlv_dpio_read(dev_priv, pipe, CHV_TX_DW3(ch, i));
+		val &= ~DPIO_TX_UNIQ_TRANS_SCALE_EN;
+		vlv_dpio_write(dev_priv, pipe, CHV_TX_DW3(ch, i), val);
+	}
+
+	/* Additional steps for 1200mV-0dB */
+#if 0
+	val = vlv_dpio_read(dev_priv, pipe, VLV_TX_DW3(ch));
+	if (ch)
+		val |= DPIO_TX_UNIQ_TRANS_SCALE_CH1;
+	else
+		val |= DPIO_TX_UNIQ_TRANS_SCALE_CH0;
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW3(ch), val);
+
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW2(ch),
+			vlv_dpio_read(dev_priv, pipe, VLV_TX_DW2(ch)) |
+				(0x9a << DPIO_UNIQ_TRANS_SCALE_SHIFT));
+#endif
+	/* Start swing calculation */
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW10(ch));
+	val |= DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS01_DW10(ch), val);
+
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS23_DW10(ch));
+	val |= DPIO_PCS_SWING_CALC_TX0_TX2 | DPIO_PCS_SWING_CALC_TX1_TX3;
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS23_DW10(ch), val);
+
+	/* LRC Bypass */
+	val = vlv_dpio_read(dev_priv, pipe, CHV_CMN_DW30);
+	val |= DPIO_LRC_BYPASS;
+	vlv_dpio_write(dev_priv, pipe, CHV_CMN_DW30, val);
+
 	mutex_unlock(&dev_priv->dpio_lock);
+
+	intel_enable_hdmi(encoder);
+
+	vlv_wait_port_ready(dev_priv, dport);
 }
 
 static void intel_hdmi_destroy(struct drm_connector *connector)
@@ -1239,7 +1451,10 @@ void intel_hdmi_init_connector(struct intel_digital_port *intel_dig_port,
 		intel_encoder->hpd_pin = HPD_PORT_C;
 		break;
 	case PORT_D:
-		intel_hdmi->ddc_bus = GMBUS_PORT_DPD;
+		if (IS_CHERRYVIEW(dev))
+			intel_hdmi->ddc_bus = GMBUS_PORT_DPD_CHV;
+		else
+			intel_hdmi->ddc_bus = GMBUS_PORT_DPD;
 		intel_encoder->hpd_pin = HPD_PORT_D;
 		break;
 	case PORT_A:
@@ -1270,6 +1485,7 @@ void intel_hdmi_init_connector(struct intel_digital_port *intel_dig_port,
 		intel_connector->get_hw_state = intel_ddi_connector_get_hw_state;
 	else
 		intel_connector->get_hw_state = intel_connector_get_hw_state;
+	intel_connector->unregister = intel_connector_unregister;
 
 	intel_hdmi_add_properties(intel_hdmi, connector);
 
@@ -1308,22 +1524,40 @@ void intel_hdmi_init(struct drm_device *dev, int hdmi_reg, enum port port)
 			 DRM_MODE_ENCODER_TMDS);
 
 	intel_encoder->compute_config = intel_hdmi_compute_config;
-	intel_encoder->mode_set = intel_hdmi_mode_set;
 	intel_encoder->disable = intel_disable_hdmi;
 	intel_encoder->get_hw_state = intel_hdmi_get_hw_state;
 	intel_encoder->get_config = intel_hdmi_get_config;
-	if (IS_VALLEYVIEW(dev)) {
+	if (IS_CHERRYVIEW(dev)) {
+		intel_encoder->pre_enable = chv_hdmi_pre_enable;
+		intel_encoder->enable = vlv_enable_hdmi;
+		intel_encoder->post_disable = chv_hdmi_post_disable;
+	} else if (IS_VALLEYVIEW(dev)) {
 		intel_encoder->pre_pll_enable = vlv_hdmi_pre_pll_enable;
 		intel_encoder->pre_enable = vlv_hdmi_pre_enable;
 		intel_encoder->enable = vlv_enable_hdmi;
 		intel_encoder->post_disable = vlv_hdmi_post_disable;
 	} else {
+		intel_encoder->pre_enable = intel_hdmi_pre_enable;
 		intel_encoder->enable = intel_enable_hdmi;
 	}
 
 	intel_encoder->type = INTEL_OUTPUT_HDMI;
-	intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
-	intel_encoder->cloneable = false;
+	if (IS_CHERRYVIEW(dev)) {
+		if (port == PORT_D)
+			intel_encoder->crtc_mask = 1 << 2;
+		else
+			intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
+	} else {
+		intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
+	}
+	intel_encoder->cloneable = 1 << INTEL_OUTPUT_ANALOG;
+	/*
+	 * BSpec is unclear about HDMI+HDMI cloning on g4x, but it seems
+	 * to work on real hardware. And since g4x can send infoframes to
+	 * only one port anyway, nothing is lost by allowing it.
+	 */
+	if (IS_G4X(dev))
+		intel_encoder->cloneable |= 1 << INTEL_OUTPUT_HDMI;
 
 	intel_dig_port->port = port;
 	intel_dig_port->hdmi.hdmi_reg = hdmi_reg;
diff --git a/drivers/gpu/drm/i915/intel_i2c.c b/drivers/gpu/drm/i915/intel_i2c.c
index 2ca17b14b6c..d33b61d0dd3 100644
--- a/drivers/gpu/drm/i915/intel_i2c.c
+++ b/drivers/gpu/drm/i915/intel_i2c.c
@@ -82,20 +82,11 @@ static int get_disp_clk_div(struct drm_i915_private *dev_priv,
 
 static void gmbus_set_freq(struct drm_i915_private *dev_priv)
 {
-	int vco_freq[] = { 800, 1600, 2000, 2400 };
-	int gmbus_freq = 0, cdclk_div, hpll_freq;
+	int vco, gmbus_freq = 0, cdclk_div;
 
 	BUG_ON(!IS_VALLEYVIEW(dev_priv->dev));
 
-	/* Skip setting the gmbus freq if BIOS has already programmed it */
-	if (I915_READ(GMBUSFREQ_VLV) != 0xA0)
-		return;
-
-	/* Obtain SKU information */
-	mutex_lock(&dev_priv->dpio_lock);
-	hpll_freq =
-		vlv_cck_read(dev_priv, CCK_FUSE_REG) & CCK_FUSE_HPLL_FREQ_MASK;
-	mutex_unlock(&dev_priv->dpio_lock);
+	vco = valleyview_get_vco(dev_priv);
 
 	/* Get the CDCLK divide ratio */
 	cdclk_div = get_disp_clk_div(dev_priv, CDCLK);
@@ -106,7 +97,7 @@ static void gmbus_set_freq(struct drm_i915_private *dev_priv)
 	 * in fact 1MHz is the correct frequency.
 	 */
 	if (cdclk_div)
-		gmbus_freq = (vco_freq[hpll_freq] << 1) / cdclk_div;
+		gmbus_freq = (vco << 1) / cdclk_div;
 
 	if (WARN_ON(gmbus_freq == 0))
 		return;
@@ -267,13 +258,6 @@ intel_gpio_setup(struct intel_gmbus *bus, u32 pin)
 	algo->data = bus;
 }
 
-/*
- * gmbus on gen4 seems to be able to generate legacy interrupts even when in MSI
- * mode. This results in spurious interrupt warnings if the legacy irq no. is
- * shared with another device. The kernel then disables that interrupt source
- * and so prevents the other device from working properly.
- */
-#define HAS_GMBUS_IRQ(dev) (INTEL_INFO(dev)->gen >= 5)
 static int
 gmbus_wait_hw_status(struct drm_i915_private *dev_priv,
 		     u32 gmbus2_status,
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index c3b4da7895e..5e5a72fca5f 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -126,10 +126,6 @@ static void intel_lvds_get_config(struct intel_encoder *encoder,
 	pipe_config->adjusted_mode.crtc_clock = dotclock;
 }
 
-/* The LVDS pin pair needs to be on before the DPLLs are enabled.
- * This is an exception to the general rule that mode_set doesn't turn
- * things on.
- */
 static void intel_pre_enable_lvds(struct intel_encoder *encoder)
 {
 	struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
@@ -256,8 +252,9 @@ static void intel_disable_lvds(struct intel_encoder *encoder)
 	POSTING_READ(lvds_encoder->reg);
 }
 
-static int intel_lvds_mode_valid(struct drm_connector *connector,
-				 struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_lvds_mode_valid(struct drm_connector *connector,
+		      struct drm_display_mode *mode)
 {
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 	struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode;
@@ -330,15 +327,6 @@ static bool intel_lvds_compute_config(struct intel_encoder *intel_encoder,
 	return true;
 }
 
-static void intel_lvds_mode_set(struct intel_encoder *encoder)
-{
-	/*
-	 * We don't do anything here, the LVDS port is fully set up in the pre
-	 * enable hook - the ordering constraints for enabling the lvds port vs.
-	 * enabling the display pll are too strict.
-	 */
-}
-
 /**
  * Detect the LVDS connection.
  *
@@ -353,7 +341,7 @@ intel_lvds_detect(struct drm_connector *connector, bool force)
 	enum drm_connector_status status;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector));
+		      connector->base.id, connector->name);
 
 	status = intel_panel_detect(dev);
 	if (status != connector_status_unknown)
@@ -446,9 +434,19 @@ static int intel_lid_notify(struct notifier_block *nb, unsigned long val,
 	if (dev_priv->modeset_restore == MODESET_DONE)
 		goto exit;
 
-	drm_modeset_lock_all(dev);
-	intel_modeset_setup_hw_state(dev, true);
-	drm_modeset_unlock_all(dev);
+	/*
+	 * Some old platform's BIOS love to wreak havoc while the lid is closed.
+	 * We try to detect this here and undo any damage. The split for PCH
+	 * platforms is rather conservative and a bit arbitrary expect that on
+	 * those platforms VGA disabling requires actual legacy VGA I/O access,
+	 * and as part of the cleanup in the hw state restore we also redisable
+	 * the vga plane.
+	 */
+	if (!HAS_PCH_SPLIT(dev)) {
+		drm_modeset_lock_all(dev);
+		intel_modeset_setup_hw_state(dev, true);
+		drm_modeset_unlock_all(dev);
+	}
 
 	dev_priv->modeset_restore = MODESET_DONE;
 
@@ -744,57 +742,6 @@ static const struct dmi_system_id intel_no_lvds[] = {
 	{ }	/* terminating entry */
 };
 
-/**
- * intel_find_lvds_downclock - find the reduced downclock for LVDS in EDID
- * @dev: drm device
- * @connector: LVDS connector
- *
- * Find the reduced downclock for LVDS in EDID.
- */
-static void intel_find_lvds_downclock(struct drm_device *dev,
-				      struct drm_display_mode *fixed_mode,
-				      struct drm_connector *connector)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_display_mode *scan;
-	int temp_downclock;
-
-	temp_downclock = fixed_mode->clock;
-	list_for_each_entry(scan, &connector->probed_modes, head) {
-		/*
-		 * If one mode has the same resolution with the fixed_panel
-		 * mode while they have the different refresh rate, it means
-		 * that the reduced downclock is found for the LVDS. In such
-		 * case we can set the different FPx0/1 to dynamically select
-		 * between low and high frequency.
-		 */
-		if (scan->hdisplay == fixed_mode->hdisplay &&
-		    scan->hsync_start == fixed_mode->hsync_start &&
-		    scan->hsync_end == fixed_mode->hsync_end &&
-		    scan->htotal == fixed_mode->htotal &&
-		    scan->vdisplay == fixed_mode->vdisplay &&
-		    scan->vsync_start == fixed_mode->vsync_start &&
-		    scan->vsync_end == fixed_mode->vsync_end &&
-		    scan->vtotal == fixed_mode->vtotal) {
-			if (scan->clock < temp_downclock) {
-				/*
-				 * The downclock is already found. But we
-				 * expect to find the lower downclock.
-				 */
-				temp_downclock = scan->clock;
-			}
-		}
-	}
-	if (temp_downclock < fixed_mode->clock && i915_lvds_downclock) {
-		/* We found the downclock for LVDS. */
-		dev_priv->lvds_downclock_avail = 1;
-		dev_priv->lvds_downclock = temp_downclock;
-		DRM_DEBUG_KMS("LVDS downclock is found in EDID. "
-			      "Normal clock %dKhz, downclock %dKhz\n",
-			      fixed_mode->clock, temp_downclock);
-	}
-}
-
 /*
  * Enumerate the child dev array parsed from VBT to check whether
  * the LVDS is present.
@@ -888,8 +835,8 @@ static bool compute_is_dual_link_lvds(struct intel_lvds_encoder *lvds_encoder)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	/* use the module option value if specified */
-	if (i915_lvds_channel_mode > 0)
-		return i915_lvds_channel_mode == 2;
+	if (i915.lvds_channel_mode > 0)
+		return i915.lvds_channel_mode == 2;
 
 	if (dmi_check_system(intel_dual_link_lvds))
 		return true;
@@ -939,6 +886,7 @@ void intel_lvds_init(struct drm_device *dev)
 	struct drm_encoder *encoder;
 	struct drm_display_mode *scan; /* *modes, *bios_mode; */
 	struct drm_display_mode *fixed_mode = NULL;
+	struct drm_display_mode *downclock_mode = NULL;
 	struct edid *edid;
 	struct drm_crtc *crtc;
 	u32 lvds;
@@ -992,16 +940,16 @@ void intel_lvds_init(struct drm_device *dev)
 	intel_encoder->enable = intel_enable_lvds;
 	intel_encoder->pre_enable = intel_pre_enable_lvds;
 	intel_encoder->compute_config = intel_lvds_compute_config;
-	intel_encoder->mode_set = intel_lvds_mode_set;
 	intel_encoder->disable = intel_disable_lvds;
 	intel_encoder->get_hw_state = intel_lvds_get_hw_state;
 	intel_encoder->get_config = intel_lvds_get_config;
 	intel_connector->get_hw_state = intel_connector_get_hw_state;
+	intel_connector->unregister = intel_connector_unregister;
 
 	intel_connector_attach_encoder(intel_connector, intel_encoder);
 	intel_encoder->type = INTEL_OUTPUT_LVDS;
 
-	intel_encoder->cloneable = false;
+	intel_encoder->cloneable = 0;
 	if (HAS_PCH_SPLIT(dev))
 		intel_encoder->crtc_mask = (1 << 0) | (1 << 1) | (1 << 2);
 	else if (IS_GEN4(dev))
@@ -1040,6 +988,7 @@ void intel_lvds_init(struct drm_device *dev)
 	 * Attempt to get the fixed panel mode from DDC.  Assume that the
 	 * preferred mode is the right one.
 	 */
+	mutex_lock(&dev->mode_config.mutex);
 	edid = drm_get_edid(connector, intel_gmbus_get_adapter(dev_priv, pin));
 	if (edid) {
 		if (drm_add_edid_modes(connector, edid)) {
@@ -1072,8 +1021,21 @@ void intel_lvds_init(struct drm_device *dev)
 
 			fixed_mode = drm_mode_duplicate(dev, scan);
 			if (fixed_mode) {
-				intel_find_lvds_downclock(dev, fixed_mode,
-							  connector);
+				downclock_mode =
+					intel_find_panel_downclock(dev,
+					fixed_mode, connector);
+				if (downclock_mode != NULL &&
+					i915.lvds_downclock) {
+					/* We found the downclock for LVDS. */
+					dev_priv->lvds_downclock_avail = true;
+					dev_priv->lvds_downclock =
+						downclock_mode->clock;
+					DRM_DEBUG_KMS("LVDS downclock is found"
+					" in EDID. Normal clock %dKhz, "
+					"downclock %dKhz\n",
+					fixed_mode->clock,
+					dev_priv->lvds_downclock);
+				}
 				goto out;
 			}
 		}
@@ -1120,6 +1082,8 @@ void intel_lvds_init(struct drm_device *dev)
 		goto failed;
 
 out:
+	mutex_unlock(&dev->mode_config.mutex);
+
 	lvds_encoder->is_dual_link = compute_is_dual_link_lvds(lvds_encoder);
 	DRM_DEBUG_KMS("detected %s-link lvds configuration\n",
 		      lvds_encoder->is_dual_link ? "dual" : "single");
@@ -1142,17 +1106,17 @@ out:
 	}
 	drm_sysfs_connector_add(connector);
 
-	intel_panel_init(&intel_connector->panel, fixed_mode);
+	intel_panel_init(&intel_connector->panel, fixed_mode, downclock_mode);
 	intel_panel_setup_backlight(connector);
 
 	return;
 
 failed:
+	mutex_unlock(&dev->mode_config.mutex);
+
 	DRM_DEBUG_KMS("No LVDS modes found, disabling.\n");
 	drm_connector_cleanup(connector);
 	drm_encoder_cleanup(encoder);
-	if (fixed_mode)
-		drm_mode_destroy(dev, fixed_mode);
 	kfree(lvds_encoder);
 	kfree(lvds_connector);
 	return;
diff --git a/drivers/gpu/drm/i915/intel_opregion.c b/drivers/gpu/drm/i915/intel_opregion.c
index 6d69a9bad86..4f6b53998d7 100644
--- a/drivers/gpu/drm/i915/intel_opregion.c
+++ b/drivers/gpu/drm/i915/intel_opregion.c
@@ -28,7 +28,6 @@
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
 #include <linux/acpi.h>
-#include <linux/acpi_io.h>
 #include <acpi/video.h>
 
 #include <drm/drmP.h>
@@ -64,7 +63,7 @@ struct opregion_header {
 	u8 driver_ver[16];
 	u32 mboxes;
 	u8 reserved[164];
-} __attribute__((packed));
+} __packed;
 
 /* OpRegion mailbox #1: public ACPI methods */
 struct opregion_acpi {
@@ -86,7 +85,7 @@ struct opregion_acpi {
 	u32 cnot;       /* current OS notification */
 	u32 nrdy;       /* driver status */
 	u8 rsvd2[60];
-} __attribute__((packed));
+} __packed;
 
 /* OpRegion mailbox #2: SWSCI */
 struct opregion_swsci {
@@ -94,7 +93,7 @@ struct opregion_swsci {
 	u32 parm;       /* command parameters */
 	u32 dslp;       /* driver sleep time-out */
 	u8 rsvd[244];
-} __attribute__((packed));
+} __packed;
 
 /* OpRegion mailbox #3: ASLE */
 struct opregion_asle {
@@ -115,7 +114,7 @@ struct opregion_asle {
 	u32 srot;       /* supported rotation angles */
 	u32 iuer;       /* IUER events */
 	u8 rsvd[86];
-} __attribute__((packed));
+} __packed;
 
 /* Driver readiness indicator */
 #define ASLE_ARDY_READY		(1 << 0)
@@ -227,6 +226,8 @@ struct opregion_asle {
 #define ACPI_DIGITAL_OUTPUT (3<<8)
 #define ACPI_LVDS_OUTPUT (4<<8)
 
+#define MAX_DSLP	1500
+
 #ifdef CONFIG_ACPI
 static int swsci(struct drm_device *dev, u32 function, u32 parm, u32 *parm_out)
 {
@@ -261,10 +262,11 @@ static int swsci(struct drm_device *dev, u32 function, u32 parm, u32 *parm_out)
 		/* The spec says 2ms should be the default, but it's too small
 		 * for some machines. */
 		dslp = 50;
-	} else if (dslp > 500) {
+	} else if (dslp > MAX_DSLP) {
 		/* Hey bios, trust must be earned. */
-		WARN_ONCE(1, "excessive driver sleep timeout (DSPL) %u\n", dslp);
-		dslp = 500;
+		DRM_INFO_ONCE("ACPI BIOS requests an excessive sleep of %u ms, "
+			      "using %u ms instead\n", dslp, MAX_DSLP);
+		dslp = MAX_DSLP;
 	}
 
 	/* The spec tells us to do this, but we are the only user... */
@@ -396,16 +398,20 @@ int intel_opregion_notify_adapter(struct drm_device *dev, pci_power_t state)
 static u32 asle_set_backlight(struct drm_device *dev, u32 bclp)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_encoder *encoder;
-	struct drm_connector *connector;
-	struct intel_connector *intel_connector = NULL;
-	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[0];
+	struct intel_connector *intel_connector;
 	struct opregion_asle __iomem *asle = dev_priv->opregion.asle;
-	u32 ret = 0;
-	bool found = false;
 
 	DRM_DEBUG_DRIVER("bclp = 0x%08x\n", bclp);
 
+	/*
+	 * If the acpi_video interface is not supposed to be used, don't
+	 * bother processing backlight level change requests from firmware.
+	 */
+	if (!acpi_video_verify_backlight_support()) {
+		DRM_DEBUG_KMS("opregion backlight request ignored\n");
+		return 0;
+	}
+
 	if (!(bclp & ASLE_BCLP_VALID))
 		return ASLC_BACKLIGHT_FAILED;
 
@@ -413,39 +419,21 @@ static u32 asle_set_backlight(struct drm_device *dev, u32 bclp)
 	if (bclp > 255)
 		return ASLC_BACKLIGHT_FAILED;
 
-	mutex_lock(&dev->mode_config.mutex);
+	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+
 	/*
-	 * Could match the OpRegion connector here instead, but we'd also need
-	 * to verify the connector could handle a backlight call.
+	 * Update backlight on all connectors that support backlight (usually
+	 * only one).
 	 */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head)
-		if (encoder->crtc == crtc) {
-			found = true;
-			break;
-		}
-
-	if (!found) {
-		ret = ASLC_BACKLIGHT_FAILED;
-		goto out;
-	}
-
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
-		if (connector->encoder == encoder)
-			intel_connector = to_intel_connector(connector);
-
-	if (!intel_connector) {
-		ret = ASLC_BACKLIGHT_FAILED;
-		goto out;
-	}
-
 	DRM_DEBUG_KMS("updating opregion backlight %d/255\n", bclp);
-	intel_panel_set_backlight(intel_connector, bclp, 255);
+	list_for_each_entry(intel_connector, &dev->mode_config.connector_list, base.head)
+		intel_panel_set_backlight(intel_connector, bclp, 255);
 	iowrite32(DIV_ROUND_UP(bclp * 100, 255) | ASLE_CBLV_VALID, &asle->cblv);
 
-out:
-	mutex_unlock(&dev->mode_config.mutex);
+	drm_modeset_unlock(&dev->mode_config.connection_mutex);
 
-	return ret;
+
+	return 0;
 }
 
 static u32 asle_set_als_illum(struct drm_device *dev, u32 alsi)
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index a98a990fbab..daa118978ee 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -189,11 +189,11 @@ struct intel_overlay {
 static struct overlay_registers __iomem *
 intel_overlay_map_regs(struct intel_overlay *overlay)
 {
-	drm_i915_private_t *dev_priv = overlay->dev->dev_private;
+	struct drm_i915_private *dev_priv = overlay->dev->dev_private;
 	struct overlay_registers __iomem *regs;
 
 	if (OVERLAY_NEEDS_PHYSICAL(overlay->dev))
-		regs = (struct overlay_registers __iomem *)overlay->reg_bo->phys_obj->handle->vaddr;
+		regs = (struct overlay_registers __iomem *)overlay->reg_bo->phys_handle->vaddr;
 	else
 		regs = io_mapping_map_wc(dev_priv->gtt.mappable,
 					 i915_gem_obj_ggtt_offset(overlay->reg_bo));
@@ -212,8 +212,8 @@ static int intel_overlay_do_wait_request(struct intel_overlay *overlay,
 					 void (*tail)(struct intel_overlay *))
 {
 	struct drm_device *dev = overlay->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
 	int ret;
 
 	BUG_ON(overlay->last_flip_req);
@@ -236,7 +236,7 @@ static int intel_overlay_on(struct intel_overlay *overlay)
 {
 	struct drm_device *dev = overlay->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
 	int ret;
 
 	BUG_ON(overlay->active);
@@ -262,8 +262,8 @@ static int intel_overlay_continue(struct intel_overlay *overlay,
 				  bool load_polyphase_filter)
 {
 	struct drm_device *dev = overlay->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
 	u32 flip_addr = overlay->flip_addr;
 	u32 tmp;
 	int ret;
@@ -293,7 +293,7 @@ static void intel_overlay_release_old_vid_tail(struct intel_overlay *overlay)
 {
 	struct drm_i915_gem_object *obj = overlay->old_vid_bo;
 
-	i915_gem_object_unpin(obj);
+	i915_gem_object_ggtt_unpin(obj);
 	drm_gem_object_unreference(&obj->base);
 
 	overlay->old_vid_bo = NULL;
@@ -306,7 +306,7 @@ static void intel_overlay_off_tail(struct intel_overlay *overlay)
 	/* never have the overlay hw on without showing a frame */
 	BUG_ON(!overlay->vid_bo);
 
-	i915_gem_object_unpin(obj);
+	i915_gem_object_ggtt_unpin(obj);
 	drm_gem_object_unreference(&obj->base);
 	overlay->vid_bo = NULL;
 
@@ -320,7 +320,7 @@ static int intel_overlay_off(struct intel_overlay *overlay)
 {
 	struct drm_device *dev = overlay->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
 	u32 flip_addr = overlay->flip_addr;
 	int ret;
 
@@ -362,8 +362,8 @@ static int intel_overlay_off(struct intel_overlay *overlay)
 static int intel_overlay_recover_from_interrupt(struct intel_overlay *overlay)
 {
 	struct drm_device *dev = overlay->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
 	int ret;
 
 	if (overlay->last_flip_req == 0)
@@ -388,8 +388,8 @@ static int intel_overlay_recover_from_interrupt(struct intel_overlay *overlay)
 static int intel_overlay_release_old_vid(struct intel_overlay *overlay)
 {
 	struct drm_device *dev = overlay->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
 	int ret;
 
 	/* Only wait if there is actually an old frame to release to
@@ -606,14 +606,14 @@ static void update_colorkey(struct intel_overlay *overlay,
 {
 	u32 key = overlay->color_key;
 
-	switch (overlay->crtc->base.fb->bits_per_pixel) {
+	switch (overlay->crtc->base.primary->fb->bits_per_pixel) {
 	case 8:
 		iowrite32(0, &regs->DCLRKV);
 		iowrite32(CLK_RGB8I_MASK | DST_KEY_ENABLE, &regs->DCLRKM);
 		break;
 
 	case 16:
-		if (overlay->crtc->base.fb->depth == 15) {
+		if (overlay->crtc->base.primary->fb->depth == 15) {
 			iowrite32(RGB15_TO_COLORKEY(key), &regs->DCLRKV);
 			iowrite32(CLK_RGB15_MASK | DST_KEY_ENABLE,
 				  &regs->DCLRKM);
@@ -688,7 +688,7 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay,
 	u32 swidth, swidthsw, sheight, ostride;
 
 	BUG_ON(!mutex_is_locked(&dev->struct_mutex));
-	BUG_ON(!mutex_is_locked(&dev->mode_config.mutex));
+	BUG_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
 	BUG_ON(!overlay);
 
 	ret = intel_overlay_release_old_vid(overlay);
@@ -782,7 +782,7 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay,
 	return 0;
 
 out_unpin:
-	i915_gem_object_unpin(new_bo);
+	i915_gem_object_ggtt_unpin(new_bo);
 	return ret;
 }
 
@@ -793,7 +793,7 @@ int intel_overlay_switch_off(struct intel_overlay *overlay)
 	int ret;
 
 	BUG_ON(!mutex_is_locked(&dev->struct_mutex));
-	BUG_ON(!mutex_is_locked(&dev->mode_config.mutex));
+	BUG_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
 
 	ret = intel_overlay_recover_from_interrupt(overlay);
 	if (ret != 0)
@@ -834,7 +834,7 @@ static int check_overlay_possible_on_crtc(struct intel_overlay *overlay,
 static void update_pfit_vscale_ratio(struct intel_overlay *overlay)
 {
 	struct drm_device *dev = overlay->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 pfit_control = I915_READ(PFIT_CONTROL);
 	u32 ratio;
 
@@ -1005,7 +1005,7 @@ static int intel_panel_fitter_pipe(struct drm_device *dev)
 	u32  pfit_control;
 
 	/* i830 doesn't have a panel fitter */
-	if (IS_I830(dev))
+	if (INTEL_INFO(dev)->gen <= 3 && (IS_I830(dev) || !IS_MOBILE(dev)))
 		return -1;
 
 	pfit_control = I915_READ(PFIT_CONTROL);
@@ -1026,7 +1026,7 @@ int intel_overlay_put_image(struct drm_device *dev, void *data,
 			    struct drm_file *file_priv)
 {
 	struct drm_intel_overlay_put_image *put_image_rec = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_overlay *overlay;
 	struct drm_mode_object *drmmode_obj;
 	struct intel_crtc *crtc;
@@ -1076,7 +1076,7 @@ int intel_overlay_put_image(struct drm_device *dev, void *data,
 	mutex_lock(&dev->struct_mutex);
 
 	if (new_bo->tiling_mode) {
-		DRM_ERROR("buffer used for overlay image can not be tiled\n");
+		DRM_DEBUG_KMS("buffer used for overlay image can not be tiled\n");
 		ret = -EINVAL;
 		goto out_unlock;
 	}
@@ -1226,7 +1226,7 @@ int intel_overlay_attrs(struct drm_device *dev, void *data,
 			struct drm_file *file_priv)
 {
 	struct drm_intel_overlay_attrs *attrs = data;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_overlay *overlay;
 	struct overlay_registers __iomem *regs;
 	int ret;
@@ -1311,7 +1311,7 @@ out_unlock:
 
 void intel_setup_overlay(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_overlay *overlay;
 	struct drm_i915_gem_object *reg_bo;
 	struct overlay_registers __iomem *regs;
@@ -1340,16 +1340,14 @@ void intel_setup_overlay(struct drm_device *dev)
 	overlay->reg_bo = reg_bo;
 
 	if (OVERLAY_NEEDS_PHYSICAL(dev)) {
-		ret = i915_gem_attach_phys_object(dev, reg_bo,
-						  I915_GEM_PHYS_OVERLAY_REGS,
-						  PAGE_SIZE);
+		ret = i915_gem_object_attach_phys(reg_bo, PAGE_SIZE);
 		if (ret) {
 			DRM_ERROR("failed to attach phys overlay regs\n");
 			goto out_free_bo;
 		}
-		overlay->flip_addr = reg_bo->phys_obj->handle->busaddr;
+		overlay->flip_addr = reg_bo->phys_handle->busaddr;
 	} else {
-		ret = i915_gem_obj_ggtt_pin(reg_bo, PAGE_SIZE, true, false);
+		ret = i915_gem_obj_ggtt_pin(reg_bo, PAGE_SIZE, PIN_MAPPABLE);
 		if (ret) {
 			DRM_ERROR("failed to pin overlay register bo\n");
 			goto out_free_bo;
@@ -1386,7 +1384,7 @@ void intel_setup_overlay(struct drm_device *dev)
 
 out_unpin_bo:
 	if (!OVERLAY_NEEDS_PHYSICAL(dev))
-		i915_gem_object_unpin(reg_bo);
+		i915_gem_object_ggtt_unpin(reg_bo);
 out_free_bo:
 	drm_gem_object_unreference(&reg_bo->base);
 out_free:
@@ -1397,7 +1395,7 @@ out_free:
 
 void intel_cleanup_overlay(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	if (!dev_priv->overlay)
 		return;
@@ -1421,14 +1419,14 @@ struct intel_overlay_error_state {
 static struct overlay_registers __iomem *
 intel_overlay_map_regs_atomic(struct intel_overlay *overlay)
 {
-	drm_i915_private_t *dev_priv = overlay->dev->dev_private;
+	struct drm_i915_private *dev_priv = overlay->dev->dev_private;
 	struct overlay_registers __iomem *regs;
 
 	if (OVERLAY_NEEDS_PHYSICAL(overlay->dev))
 		/* Cast to make sparse happy, but it's wc memory anyway, so
 		 * equivalent to the wc io mapping on X86. */
 		regs = (struct overlay_registers __iomem *)
-			overlay->reg_bo->phys_obj->handle->vaddr;
+			overlay->reg_bo->phys_handle->vaddr;
 	else
 		regs = io_mapping_map_atomic_wc(dev_priv->gtt.mappable,
 						i915_gem_obj_ggtt_offset(overlay->reg_bo));
@@ -1447,7 +1445,7 @@ static void intel_overlay_unmap_regs_atomic(struct intel_overlay *overlay,
 struct intel_overlay_error_state *
 intel_overlay_capture_error_state(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_overlay *overlay = dev_priv->overlay;
 	struct intel_overlay_error_state *error;
 	struct overlay_registers __iomem *regs;
@@ -1462,7 +1460,7 @@ intel_overlay_capture_error_state(struct drm_device *dev)
 	error->dovsta = I915_READ(DOVSTA);
 	error->isr = I915_READ(ISR);
 	if (OVERLAY_NEEDS_PHYSICAL(overlay->dev))
-		error->base = (__force long)overlay->reg_bo->phys_obj->handle->vaddr;
+		error->base = (__force long)overlay->reg_bo->phys_handle->vaddr;
 	else
 		error->base = i915_gem_obj_ggtt_offset(overlay->reg_bo);
 
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index e6f782d1c66..12b02fe1d0a 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -33,8 +33,6 @@
 #include <linux/moduleparam.h>
 #include "intel_drv.h"
 
-#define PCI_LBPC 0xf4 /* legacy/combination backlight modes */
-
 void
 intel_fixed_panel_mode(const struct drm_display_mode *fixed_mode,
 		       struct drm_display_mode *adjusted_mode)
@@ -44,6 +42,59 @@ intel_fixed_panel_mode(const struct drm_display_mode *fixed_mode,
 	drm_mode_set_crtcinfo(adjusted_mode, 0);
 }
 
+/**
+ * intel_find_panel_downclock - find the reduced downclock for LVDS in EDID
+ * @dev: drm device
+ * @fixed_mode : panel native mode
+ * @connector: LVDS/eDP connector
+ *
+ * Return downclock_avail
+ * Find the reduced downclock for LVDS/eDP in EDID.
+ */
+struct drm_display_mode *
+intel_find_panel_downclock(struct drm_device *dev,
+			struct drm_display_mode *fixed_mode,
+			struct drm_connector *connector)
+{
+	struct drm_display_mode *scan, *tmp_mode;
+	int temp_downclock;
+
+	temp_downclock = fixed_mode->clock;
+	tmp_mode = NULL;
+
+	list_for_each_entry(scan, &connector->probed_modes, head) {
+		/*
+		 * If one mode has the same resolution with the fixed_panel
+		 * mode while they have the different refresh rate, it means
+		 * that the reduced downclock is found. In such
+		 * case we can set the different FPx0/1 to dynamically select
+		 * between low and high frequency.
+		 */
+		if (scan->hdisplay == fixed_mode->hdisplay &&
+		    scan->hsync_start == fixed_mode->hsync_start &&
+		    scan->hsync_end == fixed_mode->hsync_end &&
+		    scan->htotal == fixed_mode->htotal &&
+		    scan->vdisplay == fixed_mode->vdisplay &&
+		    scan->vsync_start == fixed_mode->vsync_start &&
+		    scan->vsync_end == fixed_mode->vsync_end &&
+		    scan->vtotal == fixed_mode->vtotal) {
+			if (scan->clock < temp_downclock) {
+				/*
+				 * The downclock is already found. But we
+				 * expect to find the lower downclock.
+				 */
+				temp_downclock = scan->clock;
+				tmp_mode = scan;
+			}
+		}
+	}
+
+	if (temp_downclock < fixed_mode->clock)
+		return drm_mode_duplicate(dev, tmp_mode);
+	else
+		return NULL;
+}
+
 /* adjusted_mode has been preset to be the panel's fixed mode */
 void
 intel_pch_panel_fitting(struct intel_crtc *intel_crtc,
@@ -325,214 +376,185 @@ out:
 	pipe_config->gmch_pfit.lvds_border_bits = border;
 }
 
-static int is_backlight_combination_mode(struct drm_device *dev)
+enum drm_connector_status
+intel_panel_detect(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (IS_GEN4(dev))
-		return I915_READ(BLC_PWM_CTL2) & BLM_COMBINATION_MODE;
-
-	if (IS_GEN2(dev))
-		return I915_READ(BLC_PWM_CTL) & BLM_LEGACY_MODE;
+	/* Assume that the BIOS does not lie through the OpRegion... */
+	if (!i915.panel_ignore_lid && dev_priv->opregion.lid_state) {
+		return ioread32(dev_priv->opregion.lid_state) & 0x1 ?
+			connector_status_connected :
+			connector_status_disconnected;
+	}
 
-	return 0;
+	switch (i915.panel_ignore_lid) {
+	case -2:
+		return connector_status_connected;
+	case -1:
+		return connector_status_disconnected;
+	default:
+		return connector_status_unknown;
+	}
 }
 
-/* XXX: query mode clock or hardware clock and program max PWM appropriately
- * when it's 0.
- */
-static u32 i915_read_blc_pwm_ctl(struct drm_device *dev, enum pipe pipe)
+static u32 intel_panel_compute_brightness(struct intel_connector *connector,
+					  u32 val)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val;
+	struct intel_panel *panel = &connector->panel;
 
-	WARN_ON_SMP(!spin_is_locked(&dev_priv->backlight.lock));
+	WARN_ON(panel->backlight.max == 0);
 
-	/* Restore the CTL value if it lost, e.g. GPU reset */
-
-	if (HAS_PCH_SPLIT(dev_priv->dev)) {
-		val = I915_READ(BLC_PWM_PCH_CTL2);
-		if (dev_priv->regfile.saveBLC_PWM_CTL2 == 0) {
-			dev_priv->regfile.saveBLC_PWM_CTL2 = val;
-		} else if (val == 0) {
-			val = dev_priv->regfile.saveBLC_PWM_CTL2;
-			I915_WRITE(BLC_PWM_PCH_CTL2, val);
-		}
-	} else if (IS_VALLEYVIEW(dev)) {
-		val = I915_READ(VLV_BLC_PWM_CTL(pipe));
-		if (dev_priv->regfile.saveBLC_PWM_CTL == 0) {
-			dev_priv->regfile.saveBLC_PWM_CTL = val;
-			dev_priv->regfile.saveBLC_PWM_CTL2 =
-				I915_READ(VLV_BLC_PWM_CTL2(pipe));
-		} else if (val == 0) {
-			val = dev_priv->regfile.saveBLC_PWM_CTL;
-			I915_WRITE(VLV_BLC_PWM_CTL(pipe), val);
-			I915_WRITE(VLV_BLC_PWM_CTL2(pipe),
-				   dev_priv->regfile.saveBLC_PWM_CTL2);
-		}
+	if (i915.invert_brightness < 0)
+		return val;
 
-		if (!val)
-			val = 0x0f42ffff;
-	} else {
-		val = I915_READ(BLC_PWM_CTL);
-		if (dev_priv->regfile.saveBLC_PWM_CTL == 0) {
-			dev_priv->regfile.saveBLC_PWM_CTL = val;
-			if (INTEL_INFO(dev)->gen >= 4)
-				dev_priv->regfile.saveBLC_PWM_CTL2 =
-					I915_READ(BLC_PWM_CTL2);
-		} else if (val == 0) {
-			val = dev_priv->regfile.saveBLC_PWM_CTL;
-			I915_WRITE(BLC_PWM_CTL, val);
-			if (INTEL_INFO(dev)->gen >= 4)
-				I915_WRITE(BLC_PWM_CTL2,
-					   dev_priv->regfile.saveBLC_PWM_CTL2);
-		}
+	if (i915.invert_brightness > 0 ||
+	    dev_priv->quirks & QUIRK_INVERT_BRIGHTNESS) {
+		return panel->backlight.max - val;
 	}
 
 	return val;
 }
 
-static u32 intel_panel_get_max_backlight(struct drm_device *dev,
-					 enum pipe pipe)
+static u32 bdw_get_backlight(struct intel_connector *connector)
 {
-	u32 max;
-
-	max = i915_read_blc_pwm_ctl(dev, pipe);
-
-	if (HAS_PCH_SPLIT(dev)) {
-		max >>= 16;
-	} else {
-		if (INTEL_INFO(dev)->gen < 4)
-			max >>= 17;
-		else
-			max >>= 16;
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
-		if (is_backlight_combination_mode(dev))
-			max *= 0xff;
-	}
+	return I915_READ(BLC_PWM_PCH_CTL2) & BACKLIGHT_DUTY_CYCLE_MASK;
+}
 
-	DRM_DEBUG_DRIVER("max backlight PWM = %d\n", max);
+static u32 pch_get_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	return max;
+	return I915_READ(BLC_PWM_CPU_CTL) & BACKLIGHT_DUTY_CYCLE_MASK;
 }
 
-static int i915_panel_invert_brightness;
-MODULE_PARM_DESC(invert_brightness, "Invert backlight brightness "
-	"(-1 force normal, 0 machine defaults, 1 force inversion), please "
-	"report PCI device ID, subsystem vendor and subsystem device ID "
-	"to dri-devel@lists.freedesktop.org, if your machine needs it. "
-	"It will then be included in an upcoming module version.");
-module_param_named(invert_brightness, i915_panel_invert_brightness, int, 0600);
-static u32 intel_panel_compute_brightness(struct drm_device *dev,
-					  enum pipe pipe, u32 val)
+static u32 i9xx_get_backlight(struct intel_connector *connector)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 val;
 
-	if (i915_panel_invert_brightness < 0)
-		return val;
+	val = I915_READ(BLC_PWM_CTL) & BACKLIGHT_DUTY_CYCLE_MASK;
+	if (INTEL_INFO(dev)->gen < 4)
+		val >>= 1;
 
-	if (i915_panel_invert_brightness > 0 ||
-	    dev_priv->quirks & QUIRK_INVERT_BRIGHTNESS) {
-		u32 max = intel_panel_get_max_backlight(dev, pipe);
-		if (max)
-			return max - val;
+	if (panel->backlight.combination_mode) {
+		u8 lbpc;
+
+		pci_read_config_byte(dev->pdev, PCI_LBPC, &lbpc);
+		val *= lbpc;
 	}
 
 	return val;
 }
 
-static u32 intel_panel_get_backlight(struct drm_device *dev,
-				     enum pipe pipe)
+static u32 _vlv_get_backlight(struct drm_device *dev, enum pipe pipe)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val;
-	unsigned long flags;
-	int reg;
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
+	return I915_READ(VLV_BLC_PWM_CTL(pipe)) & BACKLIGHT_DUTY_CYCLE_MASK;
+}
 
-	if (IS_BROADWELL(dev)) {
-		val = I915_READ(BLC_PWM_PCH_CTL2) & BACKLIGHT_DUTY_CYCLE_MASK;
-	} else if (HAS_PCH_SPLIT(dev)) {
-		val = I915_READ(BLC_PWM_CPU_CTL) & BACKLIGHT_DUTY_CYCLE_MASK;
-	} else {
-		if (IS_VALLEYVIEW(dev))
-			reg = VLV_BLC_PWM_CTL(pipe);
-		else
-			reg = BLC_PWM_CTL;
+static u32 vlv_get_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
 
-		val = I915_READ(reg) & BACKLIGHT_DUTY_CYCLE_MASK;
-		if (INTEL_INFO(dev)->gen < 4)
-			val >>= 1;
+	return _vlv_get_backlight(dev, pipe);
+}
 
-		if (is_backlight_combination_mode(dev)) {
-			u8 lbpc;
+static u32 intel_panel_get_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val;
+	unsigned long flags;
 
-			pci_read_config_byte(dev->pdev, PCI_LBPC, &lbpc);
-			val *= lbpc;
-		}
-	}
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
 
-	val = intel_panel_compute_brightness(dev, pipe, val);
+	val = dev_priv->display.get_backlight(connector);
+	val = intel_panel_compute_brightness(connector, val);
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
 
 	DRM_DEBUG_DRIVER("get backlight PWM = %d\n", val);
 	return val;
 }
 
-static void intel_bdw_panel_set_backlight(struct drm_device *dev, u32 level)
+static void bdw_set_backlight(struct intel_connector *connector, u32 level)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 val = I915_READ(BLC_PWM_PCH_CTL2) & ~BACKLIGHT_DUTY_CYCLE_MASK;
 	I915_WRITE(BLC_PWM_PCH_CTL2, val | level);
 }
 
-static void intel_pch_panel_set_backlight(struct drm_device *dev, u32 level)
+static void pch_set_backlight(struct intel_connector *connector, u32 level)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val = I915_READ(BLC_PWM_CPU_CTL) & ~BACKLIGHT_DUTY_CYCLE_MASK;
-	I915_WRITE(BLC_PWM_CPU_CTL, val | level);
+	u32 tmp;
+
+	tmp = I915_READ(BLC_PWM_CPU_CTL) & ~BACKLIGHT_DUTY_CYCLE_MASK;
+	I915_WRITE(BLC_PWM_CPU_CTL, tmp | level);
 }
 
-static void intel_panel_actually_set_backlight(struct drm_device *dev,
-					       enum pipe pipe, u32 level)
+static void i9xx_set_backlight(struct intel_connector *connector, u32 level)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 tmp;
-	int reg;
-
-	DRM_DEBUG_DRIVER("set backlight PWM = %d\n", level);
-	level = intel_panel_compute_brightness(dev, pipe, level);
+	struct intel_panel *panel = &connector->panel;
+	u32 tmp, mask;
 
-	if (IS_BROADWELL(dev))
-		return intel_bdw_panel_set_backlight(dev, level);
-	else if (HAS_PCH_SPLIT(dev))
-		return intel_pch_panel_set_backlight(dev, level);
+	WARN_ON(panel->backlight.max == 0);
 
-	if (is_backlight_combination_mode(dev)) {
-		u32 max = intel_panel_get_max_backlight(dev, pipe);
+	if (panel->backlight.combination_mode) {
 		u8 lbpc;
 
-		/* we're screwed, but keep behaviour backwards compatible */
-		if (!max)
-			max = 1;
-
-		lbpc = level * 0xfe / max + 1;
+		lbpc = level * 0xfe / panel->backlight.max + 1;
 		level /= lbpc;
 		pci_write_config_byte(dev->pdev, PCI_LBPC, lbpc);
 	}
 
-	if (IS_VALLEYVIEW(dev))
-		reg = VLV_BLC_PWM_CTL(pipe);
-	else
-		reg = BLC_PWM_CTL;
-
-	tmp = I915_READ(reg);
-	if (INTEL_INFO(dev)->gen < 4)
+	if (IS_GEN4(dev)) {
+		mask = BACKLIGHT_DUTY_CYCLE_MASK;
+	} else {
 		level <<= 1;
-	tmp &= ~BACKLIGHT_DUTY_CYCLE_MASK;
-	I915_WRITE(reg, tmp | level);
+		mask = BACKLIGHT_DUTY_CYCLE_MASK_PNV;
+	}
+
+	tmp = I915_READ(BLC_PWM_CTL) & ~mask;
+	I915_WRITE(BLC_PWM_CTL, tmp | level);
+}
+
+static void vlv_set_backlight(struct intel_connector *connector, u32 level)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	u32 tmp;
+
+	tmp = I915_READ(VLV_BLC_PWM_CTL(pipe)) & ~BACKLIGHT_DUTY_CYCLE_MASK;
+	I915_WRITE(VLV_BLC_PWM_CTL(pipe), tmp | level);
+}
+
+static void
+intel_panel_actually_set_backlight(struct intel_connector *connector, u32 level)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	DRM_DEBUG_DRIVER("set backlight PWM = %d\n", level);
+
+	level = intel_panel_compute_brightness(connector, level);
+	dev_priv->display.set_backlight(connector, level);
 }
 
 /* set backlight brightness to level in range [0..max] */
@@ -541,45 +563,89 @@ void intel_panel_set_backlight(struct intel_connector *connector, u32 level,
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	u32 freq;
 	unsigned long flags;
+	u64 n;
 
-	if (pipe == INVALID_PIPE)
+	if (!panel->backlight.present || pipe == INVALID_PIPE)
 		return;
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
 
-	freq = intel_panel_get_max_backlight(dev, pipe);
-	if (!freq) {
-		/* we are screwed, bail out */
-		goto out;
-	}
+	WARN_ON(panel->backlight.max == 0);
 
-	/* scale to hardware, but be careful to not overflow */
-	if (freq < max)
-		level = level * freq / max;
-	else
-		level = freq / max * level;
+	/* scale to hardware max, but be careful to not overflow */
+	freq = panel->backlight.max;
+	n = (u64)level * freq;
+	do_div(n, max);
+	level = n;
 
-	dev_priv->backlight.level = level;
-	if (dev_priv->backlight.device)
-		dev_priv->backlight.device->props.brightness = level;
+	panel->backlight.level = level;
+	if (panel->backlight.device)
+		panel->backlight.device->props.brightness = level;
 
-	if (dev_priv->backlight.enabled)
-		intel_panel_actually_set_backlight(dev, pipe, level);
-out:
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
+	if (panel->backlight.enabled)
+		intel_panel_actually_set_backlight(connector, level);
+
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+}
+
+static void pch_disable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 tmp;
+
+	intel_panel_actually_set_backlight(connector, 0);
+
+	tmp = I915_READ(BLC_PWM_CPU_CTL2);
+	I915_WRITE(BLC_PWM_CPU_CTL2, tmp & ~BLM_PWM_ENABLE);
+
+	tmp = I915_READ(BLC_PWM_PCH_CTL1);
+	I915_WRITE(BLC_PWM_PCH_CTL1, tmp & ~BLM_PCH_PWM_ENABLE);
+}
+
+static void i9xx_disable_backlight(struct intel_connector *connector)
+{
+	intel_panel_actually_set_backlight(connector, 0);
+}
+
+static void i965_disable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 tmp;
+
+	intel_panel_actually_set_backlight(connector, 0);
+
+	tmp = I915_READ(BLC_PWM_CTL2);
+	I915_WRITE(BLC_PWM_CTL2, tmp & ~BLM_PWM_ENABLE);
+}
+
+static void vlv_disable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	u32 tmp;
+
+	intel_panel_actually_set_backlight(connector, 0);
+
+	tmp = I915_READ(VLV_BLC_PWM_CTL2(pipe));
+	I915_WRITE(VLV_BLC_PWM_CTL2(pipe), tmp & ~BLM_PWM_ENABLE);
 }
 
 void intel_panel_disable_backlight(struct intel_connector *connector)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	unsigned long flags;
 
-	if (pipe == INVALID_PIPE)
+	if (!panel->backlight.present || pipe == INVALID_PIPE)
 		return;
 
 	/*
@@ -593,273 +659,550 @@ void intel_panel_disable_backlight(struct intel_connector *connector)
 		return;
 	}
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
 
-	dev_priv->backlight.enabled = false;
-	intel_panel_actually_set_backlight(dev, pipe, 0);
+	panel->backlight.enabled = false;
+	dev_priv->display.disable_backlight(connector);
 
-	if (INTEL_INFO(dev)->gen >= 4) {
-		uint32_t reg, tmp;
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+}
 
-		if (HAS_PCH_SPLIT(dev))
-			reg = BLC_PWM_CPU_CTL2;
-		else if (IS_VALLEYVIEW(dev))
-			reg = VLV_BLC_PWM_CTL2(pipe);
-		else
-			reg = BLC_PWM_CTL2;
+static void bdw_enable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 pch_ctl1, pch_ctl2;
+
+	pch_ctl1 = I915_READ(BLC_PWM_PCH_CTL1);
+	if (pch_ctl1 & BLM_PCH_PWM_ENABLE) {
+		DRM_DEBUG_KMS("pch backlight already enabled\n");
+		pch_ctl1 &= ~BLM_PCH_PWM_ENABLE;
+		I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1);
+	}
 
-		I915_WRITE(reg, I915_READ(reg) & ~BLM_PWM_ENABLE);
+	pch_ctl2 = panel->backlight.max << 16;
+	I915_WRITE(BLC_PWM_PCH_CTL2, pch_ctl2);
 
-		if (HAS_PCH_SPLIT(dev)) {
-			tmp = I915_READ(BLC_PWM_PCH_CTL1);
-			tmp &= ~BLM_PCH_PWM_ENABLE;
-			I915_WRITE(BLC_PWM_PCH_CTL1, tmp);
-		}
-	}
+	pch_ctl1 = 0;
+	if (panel->backlight.active_low_pwm)
+		pch_ctl1 |= BLM_PCH_POLARITY;
+
+	/* BDW always uses the pch pwm controls. */
+	pch_ctl1 |= BLM_PCH_OVERRIDE_ENABLE;
+
+	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1);
+	POSTING_READ(BLC_PWM_PCH_CTL1);
+	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1 | BLM_PCH_PWM_ENABLE);
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
+	/* This won't stick until the above enable. */
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
 }
 
-void intel_panel_enable_backlight(struct intel_connector *connector)
+static void pch_enable_backlight(struct intel_connector *connector)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	enum transcoder cpu_transcoder =
 		intel_pipe_to_cpu_transcoder(dev_priv, pipe);
-	unsigned long flags;
+	u32 cpu_ctl2, pch_ctl1, pch_ctl2;
 
-	if (pipe == INVALID_PIPE)
-		return;
-
-	DRM_DEBUG_KMS("pipe %c\n", pipe_name(pipe));
-
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
+	cpu_ctl2 = I915_READ(BLC_PWM_CPU_CTL2);
+	if (cpu_ctl2 & BLM_PWM_ENABLE) {
+		WARN(1, "cpu backlight already enabled\n");
+		cpu_ctl2 &= ~BLM_PWM_ENABLE;
+		I915_WRITE(BLC_PWM_CPU_CTL2, cpu_ctl2);
+	}
 
-	if (dev_priv->backlight.level == 0) {
-		dev_priv->backlight.level = intel_panel_get_max_backlight(dev,
-									  pipe);
-		if (dev_priv->backlight.device)
-			dev_priv->backlight.device->props.brightness =
-				dev_priv->backlight.level;
+	pch_ctl1 = I915_READ(BLC_PWM_PCH_CTL1);
+	if (pch_ctl1 & BLM_PCH_PWM_ENABLE) {
+		DRM_DEBUG_KMS("pch backlight already enabled\n");
+		pch_ctl1 &= ~BLM_PCH_PWM_ENABLE;
+		I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1);
 	}
 
-	if (INTEL_INFO(dev)->gen >= 4) {
-		uint32_t reg, tmp;
+	if (cpu_transcoder == TRANSCODER_EDP)
+		cpu_ctl2 = BLM_TRANSCODER_EDP;
+	else
+		cpu_ctl2 = BLM_PIPE(cpu_transcoder);
+	I915_WRITE(BLC_PWM_CPU_CTL2, cpu_ctl2);
+	POSTING_READ(BLC_PWM_CPU_CTL2);
+	I915_WRITE(BLC_PWM_CPU_CTL2, cpu_ctl2 | BLM_PWM_ENABLE);
 
-		if (HAS_PCH_SPLIT(dev))
-			reg = BLC_PWM_CPU_CTL2;
-		else if (IS_VALLEYVIEW(dev))
-			reg = VLV_BLC_PWM_CTL2(pipe);
-		else
-			reg = BLC_PWM_CTL2;
+	/* This won't stick until the above enable. */
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
 
-		tmp = I915_READ(reg);
+	pch_ctl2 = panel->backlight.max << 16;
+	I915_WRITE(BLC_PWM_PCH_CTL2, pch_ctl2);
 
-		/* Note that this can also get called through dpms changes. And
-		 * we don't track the backlight dpms state, hence check whether
-		 * we have to do anything first. */
-		if (tmp & BLM_PWM_ENABLE)
-			goto set_level;
+	pch_ctl1 = 0;
+	if (panel->backlight.active_low_pwm)
+		pch_ctl1 |= BLM_PCH_POLARITY;
 
-		if (INTEL_INFO(dev)->num_pipes == 3)
-			tmp &= ~BLM_PIPE_SELECT_IVB;
-		else
-			tmp &= ~BLM_PIPE_SELECT;
+	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1);
+	POSTING_READ(BLC_PWM_PCH_CTL1);
+	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1 | BLM_PCH_PWM_ENABLE);
+}
 
-		if (cpu_transcoder == TRANSCODER_EDP)
-			tmp |= BLM_TRANSCODER_EDP;
-		else
-			tmp |= BLM_PIPE(cpu_transcoder);
-		tmp &= ~BLM_PWM_ENABLE;
-
-		I915_WRITE(reg, tmp);
-		POSTING_READ(reg);
-		I915_WRITE(reg, tmp | BLM_PWM_ENABLE);
-
-		if (IS_BROADWELL(dev)) {
-			/*
-			 * Broadwell requires PCH override to drive the PCH
-			 * backlight pin. The above will configure the CPU
-			 * backlight pin, which we don't plan to use.
-			 */
-			tmp = I915_READ(BLC_PWM_PCH_CTL1);
-			tmp |= BLM_PCH_OVERRIDE_ENABLE | BLM_PCH_PWM_ENABLE;
-			I915_WRITE(BLC_PWM_PCH_CTL1, tmp);
-		} else if (HAS_PCH_SPLIT(dev) &&
-		    !(dev_priv->quirks & QUIRK_NO_PCH_PWM_ENABLE)) {
-			tmp = I915_READ(BLC_PWM_PCH_CTL1);
-			tmp |= BLM_PCH_PWM_ENABLE;
-			tmp &= ~BLM_PCH_OVERRIDE_ENABLE;
-			I915_WRITE(BLC_PWM_PCH_CTL1, tmp);
-		}
+static void i9xx_enable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 ctl, freq;
+
+	ctl = I915_READ(BLC_PWM_CTL);
+	if (ctl & BACKLIGHT_DUTY_CYCLE_MASK_PNV) {
+		WARN(1, "backlight already enabled\n");
+		I915_WRITE(BLC_PWM_CTL, 0);
 	}
 
-set_level:
-	/* Call below after setting BLC_PWM_CPU_CTL2 and BLC_PWM_PCH_CTL1.
-	 * BLC_PWM_CPU_CTL may be cleared to zero automatically when these
-	 * registers are set.
-	 */
-	dev_priv->backlight.enabled = true;
-	intel_panel_actually_set_backlight(dev, pipe,
-					   dev_priv->backlight.level);
+	freq = panel->backlight.max;
+	if (panel->backlight.combination_mode)
+		freq /= 0xff;
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
+	ctl = freq << 17;
+	if (panel->backlight.combination_mode)
+		ctl |= BLM_LEGACY_MODE;
+	if (IS_PINEVIEW(dev) && panel->backlight.active_low_pwm)
+		ctl |= BLM_POLARITY_PNV;
+
+	I915_WRITE(BLC_PWM_CTL, ctl);
+	POSTING_READ(BLC_PWM_CTL);
+
+	/* XXX: combine this into above write? */
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
 }
 
-/* FIXME: use VBT vals to init PWM_CTL and PWM_CTL2 correctly */
-static void intel_panel_init_backlight_regs(struct drm_device *dev)
+static void i965_enable_backlight(struct intel_connector *connector)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	u32 ctl, ctl2, freq;
+
+	ctl2 = I915_READ(BLC_PWM_CTL2);
+	if (ctl2 & BLM_PWM_ENABLE) {
+		WARN(1, "backlight already enabled\n");
+		ctl2 &= ~BLM_PWM_ENABLE;
+		I915_WRITE(BLC_PWM_CTL2, ctl2);
+	}
 
-	if (IS_VALLEYVIEW(dev)) {
-		enum pipe pipe;
+	freq = panel->backlight.max;
+	if (panel->backlight.combination_mode)
+		freq /= 0xff;
 
-		for_each_pipe(pipe) {
-			u32 cur_val = I915_READ(VLV_BLC_PWM_CTL(pipe));
+	ctl = freq << 16;
+	I915_WRITE(BLC_PWM_CTL, ctl);
 
-			/* Skip if the modulation freq is already set */
-			if (cur_val & ~BACKLIGHT_DUTY_CYCLE_MASK)
-				continue;
+	ctl2 = BLM_PIPE(pipe);
+	if (panel->backlight.combination_mode)
+		ctl2 |= BLM_COMBINATION_MODE;
+	if (panel->backlight.active_low_pwm)
+		ctl2 |= BLM_POLARITY_I965;
+	I915_WRITE(BLC_PWM_CTL2, ctl2);
+	POSTING_READ(BLC_PWM_CTL2);
+	I915_WRITE(BLC_PWM_CTL2, ctl2 | BLM_PWM_ENABLE);
 
-			cur_val &= BACKLIGHT_DUTY_CYCLE_MASK;
-			I915_WRITE(VLV_BLC_PWM_CTL(pipe), (0xf42 << 16) |
-				   cur_val);
-		}
-	}
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
 }
 
-static void intel_panel_init_backlight(struct drm_device *dev)
+static void vlv_enable_backlight(struct intel_connector *connector)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	u32 ctl, ctl2;
 
-	intel_panel_init_backlight_regs(dev);
+	ctl2 = I915_READ(VLV_BLC_PWM_CTL2(pipe));
+	if (ctl2 & BLM_PWM_ENABLE) {
+		WARN(1, "backlight already enabled\n");
+		ctl2 &= ~BLM_PWM_ENABLE;
+		I915_WRITE(VLV_BLC_PWM_CTL2(pipe), ctl2);
+	}
+
+	ctl = panel->backlight.max << 16;
+	I915_WRITE(VLV_BLC_PWM_CTL(pipe), ctl);
+
+	/* XXX: combine this into above write? */
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
 
-	dev_priv->backlight.level = intel_panel_get_backlight(dev, 0);
-	dev_priv->backlight.enabled = dev_priv->backlight.level != 0;
+	ctl2 = 0;
+	if (panel->backlight.active_low_pwm)
+		ctl2 |= BLM_POLARITY_I965;
+	I915_WRITE(VLV_BLC_PWM_CTL2(pipe), ctl2);
+	POSTING_READ(VLV_BLC_PWM_CTL2(pipe));
+	I915_WRITE(VLV_BLC_PWM_CTL2(pipe), ctl2 | BLM_PWM_ENABLE);
 }
 
-enum drm_connector_status
-intel_panel_detect(struct drm_device *dev)
+void intel_panel_enable_backlight(struct intel_connector *connector)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	unsigned long flags;
 
-	/* Assume that the BIOS does not lie through the OpRegion... */
-	if (!i915_panel_ignore_lid && dev_priv->opregion.lid_state) {
-		return ioread32(dev_priv->opregion.lid_state) & 0x1 ?
-			connector_status_connected :
-			connector_status_disconnected;
-	}
+	if (!panel->backlight.present || pipe == INVALID_PIPE)
+		return;
 
-	switch (i915_panel_ignore_lid) {
-	case -2:
-		return connector_status_connected;
-	case -1:
-		return connector_status_disconnected;
-	default:
-		return connector_status_unknown;
+	DRM_DEBUG_KMS("pipe %c\n", pipe_name(pipe));
+
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
+
+	WARN_ON(panel->backlight.max == 0);
+
+	if (panel->backlight.level == 0) {
+		panel->backlight.level = panel->backlight.max;
+		if (panel->backlight.device)
+			panel->backlight.device->props.brightness =
+				panel->backlight.level;
 	}
+
+	dev_priv->display.enable_backlight(connector);
+	panel->backlight.enabled = true;
+
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
 }
 
 #if IS_ENABLED(CONFIG_BACKLIGHT_CLASS_DEVICE)
-static int intel_panel_update_status(struct backlight_device *bd)
+static int intel_backlight_device_update_status(struct backlight_device *bd)
 {
 	struct intel_connector *connector = bl_get_data(bd);
 	struct drm_device *dev = connector->base.dev;
 
-	mutex_lock(&dev->mode_config.mutex);
+	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
 	DRM_DEBUG_KMS("updating intel_backlight, brightness=%d/%d\n",
 		      bd->props.brightness, bd->props.max_brightness);
 	intel_panel_set_backlight(connector, bd->props.brightness,
 				  bd->props.max_brightness);
-	mutex_unlock(&dev->mode_config.mutex);
+	drm_modeset_unlock(&dev->mode_config.connection_mutex);
 	return 0;
 }
 
-static int intel_panel_get_brightness(struct backlight_device *bd)
+static int intel_backlight_device_get_brightness(struct backlight_device *bd)
 {
 	struct intel_connector *connector = bl_get_data(bd);
 	struct drm_device *dev = connector->base.dev;
-	enum pipe pipe;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
 
-	mutex_lock(&dev->mode_config.mutex);
-	pipe = intel_get_pipe_from_connector(connector);
-	mutex_unlock(&dev->mode_config.mutex);
-	if (pipe == INVALID_PIPE)
-		return 0;
+	intel_runtime_pm_get(dev_priv);
+	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+	ret = intel_panel_get_backlight(connector);
+	drm_modeset_unlock(&dev->mode_config.connection_mutex);
+	intel_runtime_pm_put(dev_priv);
 
-	return intel_panel_get_backlight(connector->base.dev, pipe);
+	return ret;
 }
 
-static const struct backlight_ops intel_panel_bl_ops = {
-	.update_status = intel_panel_update_status,
-	.get_brightness = intel_panel_get_brightness,
+static const struct backlight_ops intel_backlight_device_ops = {
+	.update_status = intel_backlight_device_update_status,
+	.get_brightness = intel_backlight_device_get_brightness,
 };
 
-int intel_panel_setup_backlight(struct drm_connector *connector)
+static int intel_backlight_device_register(struct intel_connector *connector)
 {
-	struct drm_device *dev = connector->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
 	struct backlight_properties props;
-	unsigned long flags;
-
-	intel_panel_init_backlight(dev);
 
-	if (WARN_ON(dev_priv->backlight.device))
+	if (WARN_ON(panel->backlight.device))
 		return -ENODEV;
 
+	BUG_ON(panel->backlight.max == 0);
+
 	memset(&props, 0, sizeof(props));
 	props.type = BACKLIGHT_RAW;
-	props.brightness = dev_priv->backlight.level;
-
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
-	props.max_brightness = intel_panel_get_max_backlight(dev, 0);
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
+	props.brightness = panel->backlight.level;
+	props.max_brightness = panel->backlight.max;
 
-	if (props.max_brightness == 0) {
-		DRM_DEBUG_DRIVER("Failed to get maximum backlight value\n");
-		return -ENODEV;
-	}
-	dev_priv->backlight.device =
+	/*
+	 * Note: using the same name independent of the connector prevents
+	 * registration of multiple backlight devices in the driver.
+	 */
+	panel->backlight.device =
 		backlight_device_register("intel_backlight",
-					  connector->kdev,
-					  to_intel_connector(connector),
-					  &intel_panel_bl_ops, &props);
+					  connector->base.kdev,
+					  connector,
+					  &intel_backlight_device_ops, &props);
 
-	if (IS_ERR(dev_priv->backlight.device)) {
+	if (IS_ERR(panel->backlight.device)) {
 		DRM_ERROR("Failed to register backlight: %ld\n",
-			  PTR_ERR(dev_priv->backlight.device));
-		dev_priv->backlight.device = NULL;
+			  PTR_ERR(panel->backlight.device));
+		panel->backlight.device = NULL;
 		return -ENODEV;
 	}
 	return 0;
 }
 
-void intel_panel_destroy_backlight(struct drm_device *dev)
+static void intel_backlight_device_unregister(struct intel_connector *connector)
 {
+	struct intel_panel *panel = &connector->panel;
+
+	if (panel->backlight.device) {
+		backlight_device_unregister(panel->backlight.device);
+		panel->backlight.device = NULL;
+	}
+}
+#else /* CONFIG_BACKLIGHT_CLASS_DEVICE */
+static int intel_backlight_device_register(struct intel_connector *connector)
+{
+	return 0;
+}
+static void intel_backlight_device_unregister(struct intel_connector *connector)
+{
+}
+#endif /* CONFIG_BACKLIGHT_CLASS_DEVICE */
+
+/*
+ * Note: The setup hooks can't assume pipe is set!
+ *
+ * XXX: Query mode clock or hardware clock and program PWM modulation frequency
+ * appropriately when it's 0. Use VBT and/or sane defaults.
+ */
+static int bdw_setup_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	if (dev_priv->backlight.device) {
-		backlight_device_unregister(dev_priv->backlight.device);
-		dev_priv->backlight.device = NULL;
+	struct intel_panel *panel = &connector->panel;
+	u32 pch_ctl1, pch_ctl2, val;
+
+	pch_ctl1 = I915_READ(BLC_PWM_PCH_CTL1);
+	panel->backlight.active_low_pwm = pch_ctl1 & BLM_PCH_POLARITY;
+
+	pch_ctl2 = I915_READ(BLC_PWM_PCH_CTL2);
+	panel->backlight.max = pch_ctl2 >> 16;
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = bdw_get_backlight(connector);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	panel->backlight.enabled = (pch_ctl1 & BLM_PCH_PWM_ENABLE) &&
+		panel->backlight.level != 0;
+
+	return 0;
+}
+
+static int pch_setup_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 cpu_ctl2, pch_ctl1, pch_ctl2, val;
+
+	pch_ctl1 = I915_READ(BLC_PWM_PCH_CTL1);
+	panel->backlight.active_low_pwm = pch_ctl1 & BLM_PCH_POLARITY;
+
+	pch_ctl2 = I915_READ(BLC_PWM_PCH_CTL2);
+	panel->backlight.max = pch_ctl2 >> 16;
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = pch_get_backlight(connector);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	cpu_ctl2 = I915_READ(BLC_PWM_CPU_CTL2);
+	panel->backlight.enabled = (cpu_ctl2 & BLM_PWM_ENABLE) &&
+		(pch_ctl1 & BLM_PCH_PWM_ENABLE) && panel->backlight.level != 0;
+
+	return 0;
+}
+
+static int i9xx_setup_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 ctl, val;
+
+	ctl = I915_READ(BLC_PWM_CTL);
+
+	if (IS_GEN2(dev) || IS_I915GM(dev) || IS_I945GM(dev))
+		panel->backlight.combination_mode = ctl & BLM_LEGACY_MODE;
+
+	if (IS_PINEVIEW(dev))
+		panel->backlight.active_low_pwm = ctl & BLM_POLARITY_PNV;
+
+	panel->backlight.max = ctl >> 17;
+	if (panel->backlight.combination_mode)
+		panel->backlight.max *= 0xff;
+
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = i9xx_get_backlight(connector);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	panel->backlight.enabled = panel->backlight.level != 0;
+
+	return 0;
+}
+
+static int i965_setup_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 ctl, ctl2, val;
+
+	ctl2 = I915_READ(BLC_PWM_CTL2);
+	panel->backlight.combination_mode = ctl2 & BLM_COMBINATION_MODE;
+	panel->backlight.active_low_pwm = ctl2 & BLM_POLARITY_I965;
+
+	ctl = I915_READ(BLC_PWM_CTL);
+	panel->backlight.max = ctl >> 16;
+	if (panel->backlight.combination_mode)
+		panel->backlight.max *= 0xff;
+
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = i9xx_get_backlight(connector);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	panel->backlight.enabled = (ctl2 & BLM_PWM_ENABLE) &&
+		panel->backlight.level != 0;
+
+	return 0;
+}
+
+static int vlv_setup_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	enum pipe pipe;
+	u32 ctl, ctl2, val;
+
+	for_each_pipe(pipe) {
+		u32 cur_val = I915_READ(VLV_BLC_PWM_CTL(pipe));
+
+		/* Skip if the modulation freq is already set */
+		if (cur_val & ~BACKLIGHT_DUTY_CYCLE_MASK)
+			continue;
+
+		cur_val &= BACKLIGHT_DUTY_CYCLE_MASK;
+		I915_WRITE(VLV_BLC_PWM_CTL(pipe), (0xf42 << 16) |
+			   cur_val);
 	}
+
+	ctl2 = I915_READ(VLV_BLC_PWM_CTL2(PIPE_A));
+	panel->backlight.active_low_pwm = ctl2 & BLM_POLARITY_I965;
+
+	ctl = I915_READ(VLV_BLC_PWM_CTL(PIPE_A));
+	panel->backlight.max = ctl >> 16;
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = _vlv_get_backlight(dev, PIPE_A);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	panel->backlight.enabled = (ctl2 & BLM_PWM_ENABLE) &&
+		panel->backlight.level != 0;
+
+	return 0;
 }
-#else
+
 int intel_panel_setup_backlight(struct drm_connector *connector)
 {
-	intel_panel_init_backlight(connector->dev);
+	struct drm_device *dev = connector->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_panel *panel = &intel_connector->panel;
+	unsigned long flags;
+	int ret;
+
+	if (!dev_priv->vbt.backlight.present) {
+		if (dev_priv->quirks & QUIRK_BACKLIGHT_PRESENT) {
+			DRM_DEBUG_KMS("no backlight present per VBT, but present per quirk\n");
+		} else {
+			DRM_DEBUG_KMS("no backlight present per VBT\n");
+			return 0;
+		}
+	}
+
+	/* set level and max in panel struct */
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
+	ret = dev_priv->display.setup_backlight(intel_connector);
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+
+	if (ret) {
+		DRM_DEBUG_KMS("failed to setup backlight for connector %s\n",
+			      connector->name);
+		return ret;
+	}
+
+	intel_backlight_device_register(intel_connector);
+
+	panel->backlight.present = true;
+
+	DRM_DEBUG_KMS("backlight initialized, %s, brightness %u/%u, "
+		      "sysfs interface %sregistered\n",
+		      panel->backlight.enabled ? "enabled" : "disabled",
+		      panel->backlight.level, panel->backlight.max,
+		      panel->backlight.device ? "" : "not ");
+
 	return 0;
 }
 
-void intel_panel_destroy_backlight(struct drm_device *dev)
+void intel_panel_destroy_backlight(struct drm_connector *connector)
+{
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_panel *panel = &intel_connector->panel;
+
+	panel->backlight.present = false;
+	intel_backlight_device_unregister(intel_connector);
+}
+
+/* Set up chip specific backlight functions */
+void intel_panel_init_backlight_funcs(struct drm_device *dev)
 {
-	return;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (IS_BROADWELL(dev)) {
+		dev_priv->display.setup_backlight = bdw_setup_backlight;
+		dev_priv->display.enable_backlight = bdw_enable_backlight;
+		dev_priv->display.disable_backlight = pch_disable_backlight;
+		dev_priv->display.set_backlight = bdw_set_backlight;
+		dev_priv->display.get_backlight = bdw_get_backlight;
+	} else if (HAS_PCH_SPLIT(dev)) {
+		dev_priv->display.setup_backlight = pch_setup_backlight;
+		dev_priv->display.enable_backlight = pch_enable_backlight;
+		dev_priv->display.disable_backlight = pch_disable_backlight;
+		dev_priv->display.set_backlight = pch_set_backlight;
+		dev_priv->display.get_backlight = pch_get_backlight;
+	} else if (IS_VALLEYVIEW(dev)) {
+		dev_priv->display.setup_backlight = vlv_setup_backlight;
+		dev_priv->display.enable_backlight = vlv_enable_backlight;
+		dev_priv->display.disable_backlight = vlv_disable_backlight;
+		dev_priv->display.set_backlight = vlv_set_backlight;
+		dev_priv->display.get_backlight = vlv_get_backlight;
+	} else if (IS_GEN4(dev)) {
+		dev_priv->display.setup_backlight = i965_setup_backlight;
+		dev_priv->display.enable_backlight = i965_enable_backlight;
+		dev_priv->display.disable_backlight = i965_disable_backlight;
+		dev_priv->display.set_backlight = i9xx_set_backlight;
+		dev_priv->display.get_backlight = i9xx_get_backlight;
+	} else {
+		dev_priv->display.setup_backlight = i9xx_setup_backlight;
+		dev_priv->display.enable_backlight = i9xx_enable_backlight;
+		dev_priv->display.disable_backlight = i9xx_disable_backlight;
+		dev_priv->display.set_backlight = i9xx_set_backlight;
+		dev_priv->display.get_backlight = i9xx_get_backlight;
+	}
 }
-#endif
 
 int intel_panel_init(struct intel_panel *panel,
-		     struct drm_display_mode *fixed_mode)
+		     struct drm_display_mode *fixed_mode,
+		     struct drm_display_mode *downclock_mode)
 {
 	panel->fixed_mode = fixed_mode;
+	panel->downclock_mode = downclock_mode;
 
 	return 0;
 }
@@ -871,4 +1214,8 @@ void intel_panel_fini(struct intel_panel *panel)
 
 	if (panel->fixed_mode)
 		drm_mode_destroy(intel_connector->base.dev, panel->fixed_mode);
+
+	if (panel->downclock_mode)
+		drm_mode_destroy(intel_connector->base.dev,
+				panel->downclock_mode);
 }
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 26c29c17322..ee72807069e 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -30,7 +30,9 @@
 #include "intel_drv.h"
 #include "../../../platform/x86/intel_ips.h"
 #include <linux/module.h>
+#include <linux/vgaarb.h>
 #include <drm/i915_powerwell.h>
+#include <linux/pm_runtime.h>
 
 /**
  * RC6 is a special power stage which allows the GPU to enter an very
@@ -86,46 +88,53 @@ static void i8xx_disable_fbc(struct drm_device *dev)
 	DRM_DEBUG_KMS("disabled FBC\n");
 }
 
-static void i8xx_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void i8xx_enable_fbc(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_framebuffer *fb = crtc->fb;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
 	struct drm_i915_gem_object *obj = intel_fb->obj;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int cfb_pitch;
-	int plane, i;
-	u32 fbc_ctl, fbc_ctl2;
+	int i;
+	u32 fbc_ctl;
 
 	cfb_pitch = dev_priv->fbc.size / FBC_LL_SIZE;
 	if (fb->pitches[0] < cfb_pitch)
 		cfb_pitch = fb->pitches[0];
 
-	/* FBC_CTL wants 64B units */
-	cfb_pitch = (cfb_pitch / 64) - 1;
-	plane = intel_crtc->plane == 0 ? FBC_CTL_PLANEA : FBC_CTL_PLANEB;
+	/* FBC_CTL wants 32B or 64B units */
+	if (IS_GEN2(dev))
+		cfb_pitch = (cfb_pitch / 32) - 1;
+	else
+		cfb_pitch = (cfb_pitch / 64) - 1;
 
 	/* Clear old tags */
 	for (i = 0; i < (FBC_LL_SIZE / 32) + 1; i++)
 		I915_WRITE(FBC_TAG + (i * 4), 0);
 
-	/* Set it up... */
-	fbc_ctl2 = FBC_CTL_FENCE_DBL | FBC_CTL_IDLE_IMM | FBC_CTL_CPU_FENCE;
-	fbc_ctl2 |= plane;
-	I915_WRITE(FBC_CONTROL2, fbc_ctl2);
-	I915_WRITE(FBC_FENCE_OFF, crtc->y);
+	if (IS_GEN4(dev)) {
+		u32 fbc_ctl2;
+
+		/* Set it up... */
+		fbc_ctl2 = FBC_CTL_FENCE_DBL | FBC_CTL_IDLE_IMM | FBC_CTL_CPU_FENCE;
+		fbc_ctl2 |= FBC_CTL_PLANE(intel_crtc->plane);
+		I915_WRITE(FBC_CONTROL2, fbc_ctl2);
+		I915_WRITE(FBC_FENCE_OFF, crtc->y);
+	}
 
 	/* enable it... */
-	fbc_ctl = FBC_CTL_EN | FBC_CTL_PERIODIC;
+	fbc_ctl = I915_READ(FBC_CONTROL);
+	fbc_ctl &= 0x3fff << FBC_CTL_INTERVAL_SHIFT;
+	fbc_ctl |= FBC_CTL_EN | FBC_CTL_PERIODIC;
 	if (IS_I945GM(dev))
 		fbc_ctl |= FBC_CTL_C3_IDLE; /* 945 needs special SR handling */
 	fbc_ctl |= (cfb_pitch & 0xff) << FBC_CTL_STRIDE_SHIFT;
-	fbc_ctl |= (interval & 0x2fff) << FBC_CTL_INTERVAL_SHIFT;
 	fbc_ctl |= obj->fence_reg;
 	I915_WRITE(FBC_CONTROL, fbc_ctl);
 
-	DRM_DEBUG_KMS("enabled FBC, pitch %d, yoff %d, plane %c, ",
+	DRM_DEBUG_KMS("enabled FBC, pitch %d, yoff %d, plane %c\n",
 		      cfb_pitch, crtc->y, plane_name(intel_crtc->plane));
 }
 
@@ -136,29 +145,27 @@ static bool i8xx_fbc_enabled(struct drm_device *dev)
 	return I915_READ(FBC_CONTROL) & FBC_CTL_EN;
 }
 
-static void g4x_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void g4x_enable_fbc(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_framebuffer *fb = crtc->fb;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
 	struct drm_i915_gem_object *obj = intel_fb->obj;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int plane = intel_crtc->plane == 0 ? DPFC_CTL_PLANEA : DPFC_CTL_PLANEB;
-	unsigned long stall_watermark = 200;
 	u32 dpfc_ctl;
 
-	dpfc_ctl = plane | DPFC_SR_EN | DPFC_CTL_LIMIT_1X;
+	dpfc_ctl = DPFC_CTL_PLANE(intel_crtc->plane) | DPFC_SR_EN;
+	if (drm_format_plane_cpp(fb->pixel_format, 0) == 2)
+		dpfc_ctl |= DPFC_CTL_LIMIT_2X;
+	else
+		dpfc_ctl |= DPFC_CTL_LIMIT_1X;
 	dpfc_ctl |= DPFC_CTL_FENCE_EN | obj->fence_reg;
-	I915_WRITE(DPFC_CHICKEN, DPFC_HT_MODIFY);
 
-	I915_WRITE(DPFC_RECOMP_CTL, DPFC_RECOMP_STALL_EN |
-		   (stall_watermark << DPFC_RECOMP_STALL_WM_SHIFT) |
-		   (interval << DPFC_RECOMP_TIMER_COUNT_SHIFT));
 	I915_WRITE(DPFC_FENCE_YOFF, crtc->y);
 
 	/* enable it... */
-	I915_WRITE(DPFC_CONTROL, I915_READ(DPFC_CONTROL) | DPFC_CTL_EN);
+	I915_WRITE(DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);
 
 	DRM_DEBUG_KMS("enabled fbc on plane %c\n", plane_name(intel_crtc->plane));
 }
@@ -191,7 +198,11 @@ static void sandybridge_blit_fbc_update(struct drm_device *dev)
 	u32 blt_ecoskpd;
 
 	/* Make sure blitter notifies FBC of writes */
-	gen6_gt_force_wake_get(dev_priv);
+
+	/* Blitter is part of Media powerwell on VLV. No impact of
+	 * his param in other platforms for now */
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_MEDIA);
+
 	blt_ecoskpd = I915_READ(GEN6_BLITTER_ECOSKPD);
 	blt_ecoskpd |= GEN6_BLITTER_FBC_NOTIFY <<
 		GEN6_BLITTER_LOCK_SHIFT;
@@ -202,32 +213,29 @@ static void sandybridge_blit_fbc_update(struct drm_device *dev)
 			 GEN6_BLITTER_LOCK_SHIFT);
 	I915_WRITE(GEN6_BLITTER_ECOSKPD, blt_ecoskpd);
 	POSTING_READ(GEN6_BLITTER_ECOSKPD);
-	gen6_gt_force_wake_put(dev_priv);
+
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_MEDIA);
 }
 
-static void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void ironlake_enable_fbc(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_framebuffer *fb = crtc->fb;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
 	struct drm_i915_gem_object *obj = intel_fb->obj;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int plane = intel_crtc->plane == 0 ? DPFC_CTL_PLANEA : DPFC_CTL_PLANEB;
-	unsigned long stall_watermark = 200;
 	u32 dpfc_ctl;
 
-	dpfc_ctl = I915_READ(ILK_DPFC_CONTROL);
-	dpfc_ctl &= DPFC_RESERVED;
-	dpfc_ctl |= (plane | DPFC_CTL_LIMIT_1X);
-	/* Set persistent mode for front-buffer rendering, ala X. */
-	dpfc_ctl |= DPFC_CTL_PERSISTENT_MODE;
-	dpfc_ctl |= (DPFC_CTL_FENCE_EN | obj->fence_reg);
-	I915_WRITE(ILK_DPFC_CHICKEN, DPFC_HT_MODIFY);
-
-	I915_WRITE(ILK_DPFC_RECOMP_CTL, DPFC_RECOMP_STALL_EN |
-		   (stall_watermark << DPFC_RECOMP_STALL_WM_SHIFT) |
-		   (interval << DPFC_RECOMP_TIMER_COUNT_SHIFT));
+	dpfc_ctl = DPFC_CTL_PLANE(intel_crtc->plane);
+	if (drm_format_plane_cpp(fb->pixel_format, 0) == 2)
+		dpfc_ctl |= DPFC_CTL_LIMIT_2X;
+	else
+		dpfc_ctl |= DPFC_CTL_LIMIT_1X;
+	dpfc_ctl |= DPFC_CTL_FENCE_EN;
+	if (IS_GEN5(dev))
+		dpfc_ctl |= obj->fence_reg;
+
 	I915_WRITE(ILK_DPFC_FENCE_YOFF, crtc->y);
 	I915_WRITE(ILK_FBC_RT_BASE, i915_gem_obj_ggtt_offset(obj) | ILK_FBC_RT_VALID);
 	/* enable it... */
@@ -265,28 +273,35 @@ static bool ironlake_fbc_enabled(struct drm_device *dev)
 	return I915_READ(ILK_DPFC_CONTROL) & DPFC_CTL_EN;
 }
 
-static void gen7_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void gen7_enable_fbc(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_framebuffer *fb = crtc->fb;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
 	struct drm_i915_gem_object *obj = intel_fb->obj;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	u32 dpfc_ctl;
 
-	I915_WRITE(IVB_FBC_RT_BASE, i915_gem_obj_ggtt_offset(obj));
+	dpfc_ctl = IVB_DPFC_CTL_PLANE(intel_crtc->plane);
+	if (drm_format_plane_cpp(fb->pixel_format, 0) == 2)
+		dpfc_ctl |= DPFC_CTL_LIMIT_2X;
+	else
+		dpfc_ctl |= DPFC_CTL_LIMIT_1X;
+	dpfc_ctl |= IVB_DPFC_CTL_FENCE_EN;
 
-	I915_WRITE(ILK_DPFC_CONTROL, DPFC_CTL_EN | DPFC_CTL_LIMIT_1X |
-		   IVB_DPFC_CTL_FENCE_EN |
-		   intel_crtc->plane << IVB_DPFC_CTL_PLANE_SHIFT);
+	I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);
 
 	if (IS_IVYBRIDGE(dev)) {
 		/* WaFbcAsynchFlipDisableFbcQueue:ivb */
-		I915_WRITE(ILK_DISPLAY_CHICKEN1, ILK_FBCQ_DIS);
+		I915_WRITE(ILK_DISPLAY_CHICKEN1,
+			   I915_READ(ILK_DISPLAY_CHICKEN1) |
+			   ILK_FBCQ_DIS);
 	} else {
-		/* WaFbcAsynchFlipDisableFbcQueue:hsw */
-		I915_WRITE(HSW_PIPE_SLICE_CHICKEN_1(intel_crtc->pipe),
-			   HSW_BYPASS_FBC_QUEUE);
+		/* WaFbcAsynchFlipDisableFbcQueue:hsw,bdw */
+		I915_WRITE(CHICKEN_PIPESL_1(intel_crtc->pipe),
+			   I915_READ(CHICKEN_PIPESL_1(intel_crtc->pipe)) |
+			   HSW_FBCQ_DIS);
 	}
 
 	I915_WRITE(SNB_DPFC_CTL_SA,
@@ -295,7 +310,7 @@ static void gen7_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
 
 	sandybridge_blit_fbc_update(dev);
 
-	DRM_DEBUG_KMS("enabled fbc on plane %d\n", intel_crtc->plane);
+	DRM_DEBUG_KMS("enabled fbc on plane %c\n", plane_name(intel_crtc->plane));
 }
 
 bool intel_fbc_enabled(struct drm_device *dev)
@@ -321,12 +336,11 @@ static void intel_fbc_work_fn(struct work_struct *__work)
 		/* Double check that we haven't switched fb without cancelling
 		 * the prior work.
 		 */
-		if (work->crtc->fb == work->fb) {
-			dev_priv->display.enable_fbc(work->crtc,
-						     work->interval);
+		if (work->crtc->primary->fb == work->fb) {
+			dev_priv->display.enable_fbc(work->crtc);
 
 			dev_priv->fbc.plane = to_intel_crtc(work->crtc)->plane;
-			dev_priv->fbc.fb_id = work->crtc->fb->base.id;
+			dev_priv->fbc.fb_id = work->crtc->primary->fb->base.id;
 			dev_priv->fbc.y = work->crtc->y;
 		}
 
@@ -360,7 +374,7 @@ static void intel_cancel_fbc_work(struct drm_i915_private *dev_priv)
 	dev_priv->fbc.fbc_work = NULL;
 }
 
-static void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void intel_enable_fbc(struct drm_crtc *crtc)
 {
 	struct intel_fbc_work *work;
 	struct drm_device *dev = crtc->dev;
@@ -374,13 +388,12 @@ static void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
 	work = kzalloc(sizeof(*work), GFP_KERNEL);
 	if (work == NULL) {
 		DRM_ERROR("Failed to allocate FBC work structure\n");
-		dev_priv->display.enable_fbc(crtc, interval);
+		dev_priv->display.enable_fbc(crtc);
 		return;
 	}
 
 	work->crtc = crtc;
-	work->fb = crtc->fb;
-	work->interval = interval;
+	work->fb = crtc->primary->fb;
 	INIT_DELAYED_WORK(&work->work, intel_fbc_work_fn);
 
 	dev_priv->fbc.fbc_work = work;
@@ -454,12 +467,12 @@ void intel_update_fbc(struct drm_device *dev)
 	const struct drm_display_mode *adjusted_mode;
 	unsigned int max_width, max_height;
 
-	if (!I915_HAS_FBC(dev)) {
+	if (!HAS_FBC(dev)) {
 		set_no_fbc_reason(dev_priv, FBC_UNSUPPORTED);
 		return;
 	}
 
-	if (!i915_powersave) {
+	if (!i915.powersave) {
 		if (set_no_fbc_reason(dev_priv, FBC_MODULE_PARAM))
 			DRM_DEBUG_KMS("fbc disabled per module param\n");
 		return;
@@ -474,7 +487,7 @@ void intel_update_fbc(struct drm_device *dev)
 	 *   - new fb is too large to fit in compressed buffer
 	 *   - going to an unsupported config (interlace, pixel multiply, etc.)
 	 */
-	list_for_each_entry(tmp_crtc, &dev->mode_config.crtc_list, head) {
+	for_each_crtc(dev, tmp_crtc) {
 		if (intel_crtc_active(tmp_crtc) &&
 		    to_intel_crtc(tmp_crtc)->primary_enabled) {
 			if (crtc) {
@@ -486,25 +499,24 @@ void intel_update_fbc(struct drm_device *dev)
 		}
 	}
 
-	if (!crtc || crtc->fb == NULL) {
+	if (!crtc || crtc->primary->fb == NULL) {
 		if (set_no_fbc_reason(dev_priv, FBC_NO_OUTPUT))
 			DRM_DEBUG_KMS("no output, disabling\n");
 		goto out_disable;
 	}
 
 	intel_crtc = to_intel_crtc(crtc);
-	fb = crtc->fb;
+	fb = crtc->primary->fb;
 	intel_fb = to_intel_framebuffer(fb);
 	obj = intel_fb->obj;
 	adjusted_mode = &intel_crtc->config.adjusted_mode;
 
-	if (i915_enable_fbc < 0 &&
-	    INTEL_INFO(dev)->gen <= 7 && !IS_HASWELL(dev)) {
+	if (i915.enable_fbc < 0) {
 		if (set_no_fbc_reason(dev_priv, FBC_CHIP_DEFAULT))
 			DRM_DEBUG_KMS("disabled per chip default\n");
 		goto out_disable;
 	}
-	if (!i915_enable_fbc) {
+	if (!i915.enable_fbc) {
 		if (set_no_fbc_reason(dev_priv, FBC_MODULE_PARAM))
 			DRM_DEBUG_KMS("fbc disabled per module param\n");
 		goto out_disable;
@@ -530,10 +542,10 @@ void intel_update_fbc(struct drm_device *dev)
 			DRM_DEBUG_KMS("mode too large for compression, disabling\n");
 		goto out_disable;
 	}
-	if ((IS_I915GM(dev) || IS_I945GM(dev) || IS_HASWELL(dev)) &&
-	    intel_crtc->plane != 0) {
+	if ((INTEL_INFO(dev)->gen < 4 || HAS_DDI(dev)) &&
+	    intel_crtc->plane != PLANE_A) {
 		if (set_no_fbc_reason(dev_priv, FBC_BAD_PLANE))
-			DRM_DEBUG_KMS("plane not 0, disabling compression\n");
+			DRM_DEBUG_KMS("plane not A, disabling compression\n");
 		goto out_disable;
 	}
 
@@ -595,7 +607,7 @@ void intel_update_fbc(struct drm_device *dev)
 		intel_disable_fbc(dev);
 	}
 
-	intel_enable_fbc(crtc, 500);
+	intel_enable_fbc(crtc);
 	dev_priv->fbc.no_fbc_reason = FBC_OK;
 	return;
 
@@ -610,7 +622,7 @@ out_disable:
 
 static void i915_pineview_get_mem_freq(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 tmp;
 
 	tmp = I915_READ(CLKCFG);
@@ -649,7 +661,7 @@ static void i915_pineview_get_mem_freq(struct drm_device *dev)
 
 static void i915_ironlake_get_mem_freq(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	u16 ddrpll, csipll;
 
 	ddrpll = I915_READ16(DDRMPLL1);
@@ -817,7 +829,7 @@ static int i9xx_get_fifo_size(struct drm_device *dev, int plane)
 	return size;
 }
 
-static int i85x_get_fifo_size(struct drm_device *dev, int plane)
+static int i830_get_fifo_size(struct drm_device *dev, int plane)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t dsparb = I915_READ(DSPARB);
@@ -850,21 +862,6 @@ static int i845_get_fifo_size(struct drm_device *dev, int plane)
 	return size;
 }
 
-static int i830_get_fifo_size(struct drm_device *dev, int plane)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t dsparb = I915_READ(DSPARB);
-	int size;
-
-	size = dsparb & 0x7f;
-	size >>= 1; /* Convert to cachelines */
-
-	DRM_DEBUG_KMS("FIFO size - (0x%08x) %s: %d\n", dsparb,
-		      plane ? "B" : "A", size);
-
-	return size;
-}
-
 /* Pineview has different values for various configs */
 static const struct intel_watermark_params pineview_display_wm = {
 	PINEVIEW_DISPLAY_FIFO,
@@ -943,14 +940,14 @@ static const struct intel_watermark_params i915_wm_info = {
 	2,
 	I915_FIFO_LINE_SIZE
 };
-static const struct intel_watermark_params i855_wm_info = {
+static const struct intel_watermark_params i830_wm_info = {
 	I855GM_FIFO_SIZE,
 	I915_MAX_WM,
 	1,
 	2,
 	I830_FIFO_LINE_SIZE
 };
-static const struct intel_watermark_params i830_wm_info = {
+static const struct intel_watermark_params i845_wm_info = {
 	I830_FIFO_SIZE,
 	I915_MAX_WM,
 	1,
@@ -958,65 +955,6 @@ static const struct intel_watermark_params i830_wm_info = {
 	I830_FIFO_LINE_SIZE
 };
 
-static const struct intel_watermark_params ironlake_display_wm_info = {
-	ILK_DISPLAY_FIFO,
-	ILK_DISPLAY_MAXWM,
-	ILK_DISPLAY_DFTWM,
-	2,
-	ILK_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params ironlake_cursor_wm_info = {
-	ILK_CURSOR_FIFO,
-	ILK_CURSOR_MAXWM,
-	ILK_CURSOR_DFTWM,
-	2,
-	ILK_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params ironlake_display_srwm_info = {
-	ILK_DISPLAY_SR_FIFO,
-	ILK_DISPLAY_MAX_SRWM,
-	ILK_DISPLAY_DFT_SRWM,
-	2,
-	ILK_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params ironlake_cursor_srwm_info = {
-	ILK_CURSOR_SR_FIFO,
-	ILK_CURSOR_MAX_SRWM,
-	ILK_CURSOR_DFT_SRWM,
-	2,
-	ILK_FIFO_LINE_SIZE
-};
-
-static const struct intel_watermark_params sandybridge_display_wm_info = {
-	SNB_DISPLAY_FIFO,
-	SNB_DISPLAY_MAXWM,
-	SNB_DISPLAY_DFTWM,
-	2,
-	SNB_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params sandybridge_cursor_wm_info = {
-	SNB_CURSOR_FIFO,
-	SNB_CURSOR_MAXWM,
-	SNB_CURSOR_DFTWM,
-	2,
-	SNB_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params sandybridge_display_srwm_info = {
-	SNB_DISPLAY_SR_FIFO,
-	SNB_DISPLAY_MAX_SRWM,
-	SNB_DISPLAY_DFT_SRWM,
-	2,
-	SNB_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params sandybridge_cursor_srwm_info = {
-	SNB_CURSOR_SR_FIFO,
-	SNB_CURSOR_MAX_SRWM,
-	SNB_CURSOR_DFT_SRWM,
-	2,
-	SNB_FIFO_LINE_SIZE
-};
-
-
 /**
  * intel_calculate_wm - calculate watermark level
  * @clock_in_khz: pixel clock
@@ -1071,7 +1009,7 @@ static struct drm_crtc *single_enabled_crtc(struct drm_device *dev)
 {
 	struct drm_crtc *crtc, *enabled = NULL;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+	for_each_crtc(dev, crtc) {
 		if (intel_crtc_active(crtc)) {
 			if (enabled)
 				return NULL;
@@ -1102,7 +1040,7 @@ static void pineview_update_wm(struct drm_crtc *unused_crtc)
 	crtc = single_enabled_crtc(dev);
 	if (crtc) {
 		const struct drm_display_mode *adjusted_mode;
-		int pixel_size = crtc->fb->bits_per_pixel / 8;
+		int pixel_size = crtc->primary->fb->bits_per_pixel / 8;
 		int clock;
 
 		adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode;
@@ -1182,7 +1120,7 @@ static bool g4x_compute_wm0(struct drm_device *dev,
 	clock = adjusted_mode->crtc_clock;
 	htotal = adjusted_mode->crtc_htotal;
 	hdisplay = to_intel_crtc(crtc)->config.pipe_src_w;
-	pixel_size = crtc->fb->bits_per_pixel / 8;
+	pixel_size = crtc->primary->fb->bits_per_pixel / 8;
 
 	/* Use the small buffer method to calculate plane watermark */
 	entries = ((clock * pixel_size / 1000) * display_latency_ns) / 1000;
@@ -1195,9 +1133,9 @@ static bool g4x_compute_wm0(struct drm_device *dev,
 		*plane_wm = display->max_wm;
 
 	/* Use the large buffer method to calculate cursor watermark */
-	line_time_us = ((htotal * 1000) / clock);
+	line_time_us = max(htotal * 1000 / clock, 1);
 	line_count = (cursor_latency_ns / line_time_us + 1000) / 1000;
-	entries = line_count * 64 * pixel_size;
+	entries = line_count * to_intel_crtc(crtc)->cursor_width * pixel_size;
 	tlb_miss = cursor->fifo_size*cursor->cacheline_size - hdisplay * 8;
 	if (tlb_miss > 0)
 		entries += tlb_miss;
@@ -1269,9 +1207,9 @@ static bool g4x_compute_srwm(struct drm_device *dev,
 	clock = adjusted_mode->crtc_clock;
 	htotal = adjusted_mode->crtc_htotal;
 	hdisplay = to_intel_crtc(crtc)->config.pipe_src_w;
-	pixel_size = crtc->fb->bits_per_pixel / 8;
+	pixel_size = crtc->primary->fb->bits_per_pixel / 8;
 
-	line_time_us = (htotal * 1000) / clock;
+	line_time_us = max(htotal * 1000 / clock, 1);
 	line_count = (latency_ns / line_time_us + 1000) / 1000;
 	line_size = hdisplay * pixel_size;
 
@@ -1283,7 +1221,7 @@ static bool g4x_compute_srwm(struct drm_device *dev,
 	*display_wm = entries + display->guard_size;
 
 	/* calculate the self-refresh watermark for display cursor */
-	entries = line_count * pixel_size * 64;
+	entries = line_count * pixel_size * to_intel_crtc(crtc)->cursor_width;
 	entries = DIV_ROUND_UP(entries, cursor->cacheline_size);
 	*cursor_wm = entries + cursor->guard_size;
 
@@ -1308,7 +1246,7 @@ static bool vlv_compute_drain_latency(struct drm_device *dev,
 		return false;
 
 	clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock;
-	pixel_size = crtc->fb->bits_per_pixel / 8;	/* BPP */
+	pixel_size = crtc->primary->fb->bits_per_pixel / 8;	/* BPP */
 
 	entries = (clock / 1000) * pixel_size;
 	*plane_prec_mult = (entries > 256) ?
@@ -1500,11 +1438,11 @@ static void i965_update_wm(struct drm_crtc *unused_crtc)
 		int clock = adjusted_mode->crtc_clock;
 		int htotal = adjusted_mode->crtc_htotal;
 		int hdisplay = to_intel_crtc(crtc)->config.pipe_src_w;
-		int pixel_size = crtc->fb->bits_per_pixel / 8;
+		int pixel_size = crtc->primary->fb->bits_per_pixel / 8;
 		unsigned long line_time_us;
 		int entries;
 
-		line_time_us = ((htotal * 1000) / clock);
+		line_time_us = max(htotal * 1000 / clock, 1);
 
 		/* Use ns/us then divide to preserve precision */
 		entries = (((sr_latency_ns / line_time_us) + 1000) / 1000) *
@@ -1518,7 +1456,7 @@ static void i965_update_wm(struct drm_crtc *unused_crtc)
 			      entries, srwm);
 
 		entries = (((sr_latency_ns / line_time_us) + 1000) / 1000) *
-			pixel_size * 64;
+			pixel_size * to_intel_crtc(crtc)->cursor_width;
 		entries = DIV_ROUND_UP(entries,
 					  i965_cursor_wm_info.cacheline_size);
 		cursor_sr = i965_cursor_wm_info.fifo_size -
@@ -1567,13 +1505,13 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 	else if (!IS_GEN2(dev))
 		wm_info = &i915_wm_info;
 	else
-		wm_info = &i855_wm_info;
+		wm_info = &i830_wm_info;
 
 	fifo_size = dev_priv->display.get_fifo_size(dev, 0);
 	crtc = intel_get_crtc_for_plane(dev, 0);
 	if (intel_crtc_active(crtc)) {
 		const struct drm_display_mode *adjusted_mode;
-		int cpp = crtc->fb->bits_per_pixel / 8;
+		int cpp = crtc->primary->fb->bits_per_pixel / 8;
 		if (IS_GEN2(dev))
 			cpp = 4;
 
@@ -1589,7 +1527,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 	crtc = intel_get_crtc_for_plane(dev, 1);
 	if (intel_crtc_active(crtc)) {
 		const struct drm_display_mode *adjusted_mode;
-		int cpp = crtc->fb->bits_per_pixel / 8;
+		int cpp = crtc->primary->fb->bits_per_pixel / 8;
 		if (IS_GEN2(dev))
 			cpp = 4;
 
@@ -1606,6 +1544,16 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 
 	DRM_DEBUG_KMS("FIFO watermarks - A: %d, B: %d\n", planea_wm, planeb_wm);
 
+	if (IS_I915GM(dev) && enabled) {
+		struct intel_framebuffer *fb;
+
+		fb = to_intel_framebuffer(enabled->primary->fb);
+
+		/* self-refresh seems busted with untiled */
+		if (fb->obj->tiling_mode == I915_TILING_NONE)
+			enabled = NULL;
+	}
+
 	/*
 	 * Overlay gets an aggressive default since video jitter is bad.
 	 */
@@ -1615,7 +1563,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 	if (IS_I945G(dev) || IS_I945GM(dev))
 		I915_WRITE(FW_BLC_SELF, FW_BLC_SELF_EN_MASK | 0);
 	else if (IS_I915GM(dev))
-		I915_WRITE(INSTPM, I915_READ(INSTPM) & ~INSTPM_SELF_EN);
+		I915_WRITE(INSTPM, _MASKED_BIT_DISABLE(INSTPM_SELF_EN));
 
 	/* Calc sr entries for one plane configs */
 	if (HAS_FW_BLC(dev) && enabled) {
@@ -1626,11 +1574,11 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 		int clock = adjusted_mode->crtc_clock;
 		int htotal = adjusted_mode->crtc_htotal;
 		int hdisplay = to_intel_crtc(enabled)->config.pipe_src_w;
-		int pixel_size = enabled->fb->bits_per_pixel / 8;
+		int pixel_size = enabled->primary->fb->bits_per_pixel / 8;
 		unsigned long line_time_us;
 		int entries;
 
-		line_time_us = (htotal * 1000) / clock;
+		line_time_us = max(htotal * 1000 / clock, 1);
 
 		/* Use ns/us then divide to preserve precision */
 		entries = (((sr_latency_ns / line_time_us) + 1000) / 1000) *
@@ -1667,14 +1615,14 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 				I915_WRITE(FW_BLC_SELF,
 					   FW_BLC_SELF_EN_MASK | FW_BLC_SELF_EN);
 			else if (IS_I915GM(dev))
-				I915_WRITE(INSTPM, I915_READ(INSTPM) | INSTPM_SELF_EN);
+				I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_SELF_EN));
 			DRM_DEBUG_KMS("memory self refresh enabled\n");
 		} else
 			DRM_DEBUG_KMS("memory self refresh disabled\n");
 	}
 }
 
-static void i830_update_wm(struct drm_crtc *unused_crtc)
+static void i845_update_wm(struct drm_crtc *unused_crtc)
 {
 	struct drm_device *dev = unused_crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1689,7 +1637,7 @@ static void i830_update_wm(struct drm_crtc *unused_crtc)
 
 	adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode;
 	planea_wm = intel_calculate_wm(adjusted_mode->crtc_clock,
-				       &i830_wm_info,
+				       &i845_wm_info,
 				       dev_priv->display.get_fifo_size(dev, 0),
 				       4, latency_ns);
 	fwater_lo = I915_READ(FW_BLC) & ~0xfff;
@@ -1700,423 +1648,6 @@ static void i830_update_wm(struct drm_crtc *unused_crtc)
 	I915_WRITE(FW_BLC, fwater_lo);
 }
 
-/*
- * Check the wm result.
- *
- * If any calculated watermark values is larger than the maximum value that
- * can be programmed into the associated watermark register, that watermark
- * must be disabled.
- */
-static bool ironlake_check_srwm(struct drm_device *dev, int level,
-				int fbc_wm, int display_wm, int cursor_wm,
-				const struct intel_watermark_params *display,
-				const struct intel_watermark_params *cursor)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	DRM_DEBUG_KMS("watermark %d: display plane %d, fbc lines %d,"
-		      " cursor %d\n", level, display_wm, fbc_wm, cursor_wm);
-
-	if (fbc_wm > SNB_FBC_MAX_SRWM) {
-		DRM_DEBUG_KMS("fbc watermark(%d) is too large(%d), disabling wm%d+\n",
-			      fbc_wm, SNB_FBC_MAX_SRWM, level);
-
-		/* fbc has it's own way to disable FBC WM */
-		I915_WRITE(DISP_ARB_CTL,
-			   I915_READ(DISP_ARB_CTL) | DISP_FBC_WM_DIS);
-		return false;
-	} else if (INTEL_INFO(dev)->gen >= 6) {
-		/* enable FBC WM (except on ILK, where it must remain off) */
-		I915_WRITE(DISP_ARB_CTL,
-			   I915_READ(DISP_ARB_CTL) & ~DISP_FBC_WM_DIS);
-	}
-
-	if (display_wm > display->max_wm) {
-		DRM_DEBUG_KMS("display watermark(%d) is too large(%d), disabling wm%d+\n",
-			      display_wm, SNB_DISPLAY_MAX_SRWM, level);
-		return false;
-	}
-
-	if (cursor_wm > cursor->max_wm) {
-		DRM_DEBUG_KMS("cursor watermark(%d) is too large(%d), disabling wm%d+\n",
-			      cursor_wm, SNB_CURSOR_MAX_SRWM, level);
-		return false;
-	}
-
-	if (!(fbc_wm || display_wm || cursor_wm)) {
-		DRM_DEBUG_KMS("latency %d is 0, disabling wm%d+\n", level, level);
-		return false;
-	}
-
-	return true;
-}
-
-/*
- * Compute watermark values of WM[1-3],
- */
-static bool ironlake_compute_srwm(struct drm_device *dev, int level, int plane,
-				  int latency_ns,
-				  const struct intel_watermark_params *display,
-				  const struct intel_watermark_params *cursor,
-				  int *fbc_wm, int *display_wm, int *cursor_wm)
-{
-	struct drm_crtc *crtc;
-	const struct drm_display_mode *adjusted_mode;
-	unsigned long line_time_us;
-	int hdisplay, htotal, pixel_size, clock;
-	int line_count, line_size;
-	int small, large;
-	int entries;
-
-	if (!latency_ns) {
-		*fbc_wm = *display_wm = *cursor_wm = 0;
-		return false;
-	}
-
-	crtc = intel_get_crtc_for_plane(dev, plane);
-	adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode;
-	clock = adjusted_mode->crtc_clock;
-	htotal = adjusted_mode->crtc_htotal;
-	hdisplay = to_intel_crtc(crtc)->config.pipe_src_w;
-	pixel_size = crtc->fb->bits_per_pixel / 8;
-
-	line_time_us = (htotal * 1000) / clock;
-	line_count = (latency_ns / line_time_us + 1000) / 1000;
-	line_size = hdisplay * pixel_size;
-
-	/* Use the minimum of the small and large buffer method for primary */
-	small = ((clock * pixel_size / 1000) * latency_ns) / 1000;
-	large = line_count * line_size;
-
-	entries = DIV_ROUND_UP(min(small, large), display->cacheline_size);
-	*display_wm = entries + display->guard_size;
-
-	/*
-	 * Spec says:
-	 * FBC WM = ((Final Primary WM * 64) / number of bytes per line) + 2
-	 */
-	*fbc_wm = DIV_ROUND_UP(*display_wm * 64, line_size) + 2;
-
-	/* calculate the self-refresh watermark for display cursor */
-	entries = line_count * pixel_size * 64;
-	entries = DIV_ROUND_UP(entries, cursor->cacheline_size);
-	*cursor_wm = entries + cursor->guard_size;
-
-	return ironlake_check_srwm(dev, level,
-				   *fbc_wm, *display_wm, *cursor_wm,
-				   display, cursor);
-}
-
-static void ironlake_update_wm(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int fbc_wm, plane_wm, cursor_wm;
-	unsigned int enabled;
-
-	enabled = 0;
-	if (g4x_compute_wm0(dev, PIPE_A,
-			    &ironlake_display_wm_info,
-			    dev_priv->wm.pri_latency[0] * 100,
-			    &ironlake_cursor_wm_info,
-			    dev_priv->wm.cur_latency[0] * 100,
-			    &plane_wm, &cursor_wm)) {
-		I915_WRITE(WM0_PIPEA_ILK,
-			   (plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm);
-		DRM_DEBUG_KMS("FIFO watermarks For pipe A -"
-			      " plane %d, " "cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_A;
-	}
-
-	if (g4x_compute_wm0(dev, PIPE_B,
-			    &ironlake_display_wm_info,
-			    dev_priv->wm.pri_latency[0] * 100,
-			    &ironlake_cursor_wm_info,
-			    dev_priv->wm.cur_latency[0] * 100,
-			    &plane_wm, &cursor_wm)) {
-		I915_WRITE(WM0_PIPEB_ILK,
-			   (plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm);
-		DRM_DEBUG_KMS("FIFO watermarks For pipe B -"
-			      " plane %d, cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_B;
-	}
-
-	/*
-	 * Calculate and update the self-refresh watermark only when one
-	 * display plane is used.
-	 */
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
-
-	if (!single_plane_enabled(enabled))
-		return;
-	enabled = ffs(enabled) - 1;
-
-	/* WM1 */
-	if (!ironlake_compute_srwm(dev, 1, enabled,
-				   dev_priv->wm.pri_latency[1] * 500,
-				   &ironlake_display_srwm_info,
-				   &ironlake_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM1_LP_ILK,
-		   WM1_LP_SR_EN |
-		   (dev_priv->wm.pri_latency[1] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM2 */
-	if (!ironlake_compute_srwm(dev, 2, enabled,
-				   dev_priv->wm.pri_latency[2] * 500,
-				   &ironlake_display_srwm_info,
-				   &ironlake_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM2_LP_ILK,
-		   WM2_LP_EN |
-		   (dev_priv->wm.pri_latency[2] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/*
-	 * WM3 is unsupported on ILK, probably because we don't have latency
-	 * data for that power state
-	 */
-}
-
-static void sandybridge_update_wm(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int latency = dev_priv->wm.pri_latency[0] * 100;	/* In unit 0.1us */
-	u32 val;
-	int fbc_wm, plane_wm, cursor_wm;
-	unsigned int enabled;
-
-	enabled = 0;
-	if (g4x_compute_wm0(dev, PIPE_A,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEA_ILK);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEA_ILK, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe A -"
-			      " plane %d, " "cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_A;
-	}
-
-	if (g4x_compute_wm0(dev, PIPE_B,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEB_ILK);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEB_ILK, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe B -"
-			      " plane %d, cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_B;
-	}
-
-	/*
-	 * Calculate and update the self-refresh watermark only when one
-	 * display plane is used.
-	 *
-	 * SNB support 3 levels of watermark.
-	 *
-	 * WM1/WM2/WM2 watermarks have to be enabled in the ascending order,
-	 * and disabled in the descending order
-	 *
-	 */
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
-
-	if (!single_plane_enabled(enabled) ||
-	    dev_priv->sprite_scaling_enabled)
-		return;
-	enabled = ffs(enabled) - 1;
-
-	/* WM1 */
-	if (!ironlake_compute_srwm(dev, 1, enabled,
-				   dev_priv->wm.pri_latency[1] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM1_LP_ILK,
-		   WM1_LP_SR_EN |
-		   (dev_priv->wm.pri_latency[1] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM2 */
-	if (!ironlake_compute_srwm(dev, 2, enabled,
-				   dev_priv->wm.pri_latency[2] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM2_LP_ILK,
-		   WM2_LP_EN |
-		   (dev_priv->wm.pri_latency[2] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM3 */
-	if (!ironlake_compute_srwm(dev, 3, enabled,
-				   dev_priv->wm.pri_latency[3] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM3_LP_ILK,
-		   WM3_LP_EN |
-		   (dev_priv->wm.pri_latency[3] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-}
-
-static void ivybridge_update_wm(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int latency = dev_priv->wm.pri_latency[0] * 100;	/* In unit 0.1us */
-	u32 val;
-	int fbc_wm, plane_wm, cursor_wm;
-	int ignore_fbc_wm, ignore_plane_wm, ignore_cursor_wm;
-	unsigned int enabled;
-
-	enabled = 0;
-	if (g4x_compute_wm0(dev, PIPE_A,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEA_ILK);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEA_ILK, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe A -"
-			      " plane %d, " "cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_A;
-	}
-
-	if (g4x_compute_wm0(dev, PIPE_B,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEB_ILK);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEB_ILK, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe B -"
-			      " plane %d, cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_B;
-	}
-
-	if (g4x_compute_wm0(dev, PIPE_C,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEC_IVB);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEC_IVB, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe C -"
-			      " plane %d, cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_C;
-	}
-
-	/*
-	 * Calculate and update the self-refresh watermark only when one
-	 * display plane is used.
-	 *
-	 * SNB support 3 levels of watermark.
-	 *
-	 * WM1/WM2/WM2 watermarks have to be enabled in the ascending order,
-	 * and disabled in the descending order
-	 *
-	 */
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
-
-	if (!single_plane_enabled(enabled) ||
-	    dev_priv->sprite_scaling_enabled)
-		return;
-	enabled = ffs(enabled) - 1;
-
-	/* WM1 */
-	if (!ironlake_compute_srwm(dev, 1, enabled,
-				   dev_priv->wm.pri_latency[1] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM1_LP_ILK,
-		   WM1_LP_SR_EN |
-		   (dev_priv->wm.pri_latency[1] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM2 */
-	if (!ironlake_compute_srwm(dev, 2, enabled,
-				   dev_priv->wm.pri_latency[2] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM2_LP_ILK,
-		   WM2_LP_EN |
-		   (dev_priv->wm.pri_latency[2] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM3, note we have to correct the cursor latency */
-	if (!ironlake_compute_srwm(dev, 3, enabled,
-				   dev_priv->wm.pri_latency[3] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &ignore_cursor_wm) ||
-	    !ironlake_compute_srwm(dev, 3, enabled,
-				   dev_priv->wm.cur_latency[3] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &ignore_fbc_wm, &ignore_plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM3_LP_ILK,
-		   WM3_LP_EN |
-		   (dev_priv->wm.pri_latency[3] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-}
-
 static uint32_t ilk_pipe_pixel_rate(struct drm_device *dev,
 				    struct drm_crtc *crtc)
 {
@@ -2185,7 +1716,7 @@ static uint32_t ilk_wm_fbc(uint32_t pri_val, uint32_t horiz_pixels,
 	return DIV_ROUND_UP(pri_val * 64, horiz_pixels * bytes_per_pixel) + 2;
 }
 
-struct hsw_pipe_wm_parameters {
+struct ilk_pipe_wm_parameters {
 	bool active;
 	uint32_t pipe_htotal;
 	uint32_t pixel_rate;
@@ -2194,7 +1725,7 @@ struct hsw_pipe_wm_parameters {
 	struct intel_plane_wm_parameters cur;
 };
 
-struct hsw_wm_maximums {
+struct ilk_wm_maximums {
 	uint16_t pri;
 	uint16_t spr;
 	uint16_t cur;
@@ -2212,7 +1743,7 @@ struct intel_wm_config {
  * For both WM_PIPE and WM_LP.
  * mem_value must be in 0.1us units.
  */
-static uint32_t ilk_compute_pri_wm(const struct hsw_pipe_wm_parameters *params,
+static uint32_t ilk_compute_pri_wm(const struct ilk_pipe_wm_parameters *params,
 				   uint32_t mem_value,
 				   bool is_lp)
 {
@@ -2241,7 +1772,7 @@ static uint32_t ilk_compute_pri_wm(const struct hsw_pipe_wm_parameters *params,
  * For both WM_PIPE and WM_LP.
  * mem_value must be in 0.1us units.
  */
-static uint32_t ilk_compute_spr_wm(const struct hsw_pipe_wm_parameters *params,
+static uint32_t ilk_compute_spr_wm(const struct ilk_pipe_wm_parameters *params,
 				   uint32_t mem_value)
 {
 	uint32_t method1, method2;
@@ -2264,7 +1795,7 @@ static uint32_t ilk_compute_spr_wm(const struct hsw_pipe_wm_parameters *params,
  * For both WM_PIPE and WM_LP.
  * mem_value must be in 0.1us units.
  */
-static uint32_t ilk_compute_cur_wm(const struct hsw_pipe_wm_parameters *params,
+static uint32_t ilk_compute_cur_wm(const struct ilk_pipe_wm_parameters *params,
 				   uint32_t mem_value)
 {
 	if (!params->active || !params->cur.enabled)
@@ -2278,7 +1809,7 @@ static uint32_t ilk_compute_cur_wm(const struct hsw_pipe_wm_parameters *params,
 }
 
 /* Only for WM_LP. */
-static uint32_t ilk_compute_fbc_wm(const struct hsw_pipe_wm_parameters *params,
+static uint32_t ilk_compute_fbc_wm(const struct ilk_pipe_wm_parameters *params,
 				   uint32_t pri_val)
 {
 	if (!params->active || !params->pri.enabled)
@@ -2299,6 +1830,40 @@ static unsigned int ilk_display_fifo_size(const struct drm_device *dev)
 		return 512;
 }
 
+static unsigned int ilk_plane_wm_reg_max(const struct drm_device *dev,
+					 int level, bool is_sprite)
+{
+	if (INTEL_INFO(dev)->gen >= 8)
+		/* BDW primary/sprite plane watermarks */
+		return level == 0 ? 255 : 2047;
+	else if (INTEL_INFO(dev)->gen >= 7)
+		/* IVB/HSW primary/sprite plane watermarks */
+		return level == 0 ? 127 : 1023;
+	else if (!is_sprite)
+		/* ILK/SNB primary plane watermarks */
+		return level == 0 ? 127 : 511;
+	else
+		/* ILK/SNB sprite plane watermarks */
+		return level == 0 ? 63 : 255;
+}
+
+static unsigned int ilk_cursor_wm_reg_max(const struct drm_device *dev,
+					  int level)
+{
+	if (INTEL_INFO(dev)->gen >= 7)
+		return level == 0 ? 63 : 255;
+	else
+		return level == 0 ? 31 : 63;
+}
+
+static unsigned int ilk_fbc_wm_reg_max(const struct drm_device *dev)
+{
+	if (INTEL_INFO(dev)->gen >= 8)
+		return 31;
+	else
+		return 15;
+}
+
 /* Calculate the maximum primary/sprite plane watermark */
 static unsigned int ilk_plane_wm_max(const struct drm_device *dev,
 				     int level,
@@ -2307,7 +1872,6 @@ static unsigned int ilk_plane_wm_max(const struct drm_device *dev,
 				     bool is_sprite)
 {
 	unsigned int fifo_size = ilk_display_fifo_size(dev);
-	unsigned int max;
 
 	/* if sprites aren't enabled, sprites get nothing */
 	if (is_sprite && !config->sprites_enabled)
@@ -2338,19 +1902,7 @@ static unsigned int ilk_plane_wm_max(const struct drm_device *dev,
 	}
 
 	/* clamp to max that the registers can hold */
-	if (INTEL_INFO(dev)->gen >= 8)
-		max = level == 0 ? 255 : 2047;
-	else if (INTEL_INFO(dev)->gen >= 7)
-		/* IVB/HSW primary/sprite plane watermarks */
-		max = level == 0 ? 127 : 1023;
-	else if (!is_sprite)
-		/* ILK/SNB primary plane watermarks */
-		max = level == 0 ? 127 : 511;
-	else
-		/* ILK/SNB sprite plane watermarks */
-		max = level == 0 ? 63 : 255;
-
-	return min(fifo_size, max);
+	return min(fifo_size, ilk_plane_wm_reg_max(dev, level, is_sprite));
 }
 
 /* Calculate the maximum cursor plane watermark */
@@ -2363,36 +1915,33 @@ static unsigned int ilk_cursor_wm_max(const struct drm_device *dev,
 		return 64;
 
 	/* otherwise just report max that registers can hold */
-	if (INTEL_INFO(dev)->gen >= 7)
-		return level == 0 ? 63 : 255;
-	else
-		return level == 0 ? 31 : 63;
-}
-
-/* Calculate the maximum FBC watermark */
-static unsigned int ilk_fbc_wm_max(struct drm_device *dev)
-{
-	/* max that registers can hold */
-	if (INTEL_INFO(dev)->gen >= 8)
-		return 31;
-	else
-		return 15;
+	return ilk_cursor_wm_reg_max(dev, level);
 }
 
-static void ilk_compute_wm_maximums(struct drm_device *dev,
+static void ilk_compute_wm_maximums(const struct drm_device *dev,
 				    int level,
 				    const struct intel_wm_config *config,
 				    enum intel_ddb_partitioning ddb_partitioning,
-				    struct hsw_wm_maximums *max)
+				    struct ilk_wm_maximums *max)
 {
 	max->pri = ilk_plane_wm_max(dev, level, config, ddb_partitioning, false);
 	max->spr = ilk_plane_wm_max(dev, level, config, ddb_partitioning, true);
 	max->cur = ilk_cursor_wm_max(dev, level, config);
-	max->fbc = ilk_fbc_wm_max(dev);
+	max->fbc = ilk_fbc_wm_reg_max(dev);
+}
+
+static void ilk_compute_wm_reg_maximums(struct drm_device *dev,
+					int level,
+					struct ilk_wm_maximums *max)
+{
+	max->pri = ilk_plane_wm_reg_max(dev, level, false);
+	max->spr = ilk_plane_wm_reg_max(dev, level, true);
+	max->cur = ilk_cursor_wm_reg_max(dev, level);
+	max->fbc = ilk_fbc_wm_reg_max(dev);
 }
 
 static bool ilk_validate_wm_level(int level,
-				  const struct hsw_wm_maximums *max,
+				  const struct ilk_wm_maximums *max,
 				  struct intel_wm_level *result)
 {
 	bool ret;
@@ -2432,9 +1981,9 @@ static bool ilk_validate_wm_level(int level,
 	return ret;
 }
 
-static void ilk_compute_wm_level(struct drm_i915_private *dev_priv,
+static void ilk_compute_wm_level(const struct drm_i915_private *dev_priv,
 				 int level,
-				 const struct hsw_pipe_wm_parameters *p,
+				 const struct ilk_pipe_wm_parameters *p,
 				 struct intel_wm_level *result)
 {
 	uint16_t pri_latency = dev_priv->wm.pri_latency[level];
@@ -2482,7 +2031,7 @@ static void intel_read_wm_latency(struct drm_device *dev, uint16_t wm[5])
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (IS_HASWELL(dev)) {
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		uint64_t sskpd = I915_READ64(MCH_SSKPD);
 
 		wm[0] = (sskpd >> 56) & 0xFF;
@@ -2527,10 +2076,10 @@ static void intel_fixup_cur_wm_latency(struct drm_device *dev, uint16_t wm[5])
 		wm[3] *= 2;
 }
 
-static int ilk_wm_max_level(const struct drm_device *dev)
+int ilk_wm_max_level(const struct drm_device *dev)
 {
 	/* how many WM levels are we expecting */
-	if (IS_HASWELL(dev))
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		return 4;
 	else if (INTEL_INFO(dev)->gen >= 6)
 		return 3;
@@ -2563,7 +2112,44 @@ static void intel_print_wm_latency(struct drm_device *dev,
 	}
 }
 
-static void intel_setup_wm_latency(struct drm_device *dev)
+static bool ilk_increase_wm_latency(struct drm_i915_private *dev_priv,
+				    uint16_t wm[5], uint16_t min)
+{
+	int level, max_level = ilk_wm_max_level(dev_priv->dev);
+
+	if (wm[0] >= min)
+		return false;
+
+	wm[0] = max(wm[0], min);
+	for (level = 1; level <= max_level; level++)
+		wm[level] = max_t(uint16_t, wm[level], DIV_ROUND_UP(min, 5));
+
+	return true;
+}
+
+static void snb_wm_latency_quirk(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	bool changed;
+
+	/*
+	 * The BIOS provided WM memory latency values are often
+	 * inadequate for high resolution displays. Adjust them.
+	 */
+	changed = ilk_increase_wm_latency(dev_priv, dev_priv->wm.pri_latency, 12) |
+		ilk_increase_wm_latency(dev_priv, dev_priv->wm.spr_latency, 12) |
+		ilk_increase_wm_latency(dev_priv, dev_priv->wm.cur_latency, 12);
+
+	if (!changed)
+		return;
+
+	DRM_DEBUG_KMS("WM latency values increased to avoid potential underruns\n");
+	intel_print_wm_latency(dev, "Primary", dev_priv->wm.pri_latency);
+	intel_print_wm_latency(dev, "Sprite", dev_priv->wm.spr_latency);
+	intel_print_wm_latency(dev, "Cursor", dev_priv->wm.cur_latency);
+}
+
+static void ilk_setup_wm_latency(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
@@ -2580,51 +2166,68 @@ static void intel_setup_wm_latency(struct drm_device *dev)
 	intel_print_wm_latency(dev, "Primary", dev_priv->wm.pri_latency);
 	intel_print_wm_latency(dev, "Sprite", dev_priv->wm.spr_latency);
 	intel_print_wm_latency(dev, "Cursor", dev_priv->wm.cur_latency);
+
+	if (IS_GEN6(dev))
+		snb_wm_latency_quirk(dev);
 }
 
-static void hsw_compute_wm_parameters(struct drm_crtc *crtc,
-				      struct hsw_pipe_wm_parameters *p,
-				      struct intel_wm_config *config)
+static void ilk_compute_wm_parameters(struct drm_crtc *crtc,
+				      struct ilk_pipe_wm_parameters *p)
 {
 	struct drm_device *dev = crtc->dev;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	enum pipe pipe = intel_crtc->pipe;
 	struct drm_plane *plane;
 
-	p->active = intel_crtc_active(crtc);
-	if (p->active) {
-		p->pipe_htotal = intel_crtc->config.adjusted_mode.htotal;
-		p->pixel_rate = ilk_pipe_pixel_rate(dev, crtc);
-		p->pri.bytes_per_pixel = crtc->fb->bits_per_pixel / 8;
-		p->cur.bytes_per_pixel = 4;
-		p->pri.horiz_pixels = intel_crtc->config.pipe_src_w;
-		p->cur.horiz_pixels = 64;
-		/* TODO: for now, assume primary and cursor planes are always enabled. */
-		p->pri.enabled = true;
-		p->cur.enabled = true;
-	}
-
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
-		config->num_pipes_active += intel_crtc_active(crtc);
+	if (!intel_crtc_active(crtc))
+		return;
 
-	list_for_each_entry(plane, &dev->mode_config.plane_list, head) {
+	p->active = true;
+	p->pipe_htotal = intel_crtc->config.adjusted_mode.crtc_htotal;
+	p->pixel_rate = ilk_pipe_pixel_rate(dev, crtc);
+	p->pri.bytes_per_pixel = crtc->primary->fb->bits_per_pixel / 8;
+	p->cur.bytes_per_pixel = 4;
+	p->pri.horiz_pixels = intel_crtc->config.pipe_src_w;
+	p->cur.horiz_pixels = intel_crtc->cursor_width;
+	/* TODO: for now, assume primary and cursor planes are always enabled. */
+	p->pri.enabled = true;
+	p->cur.enabled = true;
+
+	drm_for_each_legacy_plane(plane, &dev->mode_config.plane_list) {
 		struct intel_plane *intel_plane = to_intel_plane(plane);
 
-		if (intel_plane->pipe == pipe)
+		if (intel_plane->pipe == pipe) {
 			p->spr = intel_plane->wm;
+			break;
+		}
+	}
+}
+
+static void ilk_compute_wm_config(struct drm_device *dev,
+				  struct intel_wm_config *config)
+{
+	struct intel_crtc *intel_crtc;
+
+	/* Compute the currently _active_ config */
+	for_each_intel_crtc(dev, intel_crtc) {
+		const struct intel_pipe_wm *wm = &intel_crtc->wm.active;
+
+		if (!wm->pipe_enabled)
+			continue;
 
-		config->sprites_enabled |= intel_plane->wm.enabled;
-		config->sprites_scaled |= intel_plane->wm.scaled;
+		config->sprites_enabled |= wm->sprites_enabled;
+		config->sprites_scaled |= wm->sprites_scaled;
+		config->num_pipes_active++;
 	}
 }
 
 /* Compute new watermarks for the pipe */
 static bool intel_compute_pipe_wm(struct drm_crtc *crtc,
-				  const struct hsw_pipe_wm_parameters *params,
+				  const struct ilk_pipe_wm_parameters *params,
 				  struct intel_pipe_wm *pipe_wm)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	const struct drm_i915_private *dev_priv = dev->dev_private;
 	int level, max_level = ilk_wm_max_level(dev);
 	/* LP0 watermark maximums depend on this pipe alone */
 	struct intel_wm_config config = {
@@ -2632,19 +2235,51 @@ static bool intel_compute_pipe_wm(struct drm_crtc *crtc,
 		.sprites_enabled = params->spr.enabled,
 		.sprites_scaled = params->spr.scaled,
 	};
-	struct hsw_wm_maximums max;
+	struct ilk_wm_maximums max;
+
+	pipe_wm->pipe_enabled = params->active;
+	pipe_wm->sprites_enabled = params->spr.enabled;
+	pipe_wm->sprites_scaled = params->spr.scaled;
+
+	/* ILK/SNB: LP2+ watermarks only w/o sprites */
+	if (INTEL_INFO(dev)->gen <= 6 && params->spr.enabled)
+		max_level = 1;
+
+	/* ILK/SNB/IVB: LP1+ watermarks only w/o scaling */
+	if (params->spr.scaled)
+		max_level = 0;
+
+	ilk_compute_wm_level(dev_priv, 0, params, &pipe_wm->wm[0]);
+
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		pipe_wm->linetime = hsw_compute_linetime_wm(dev, crtc);
 
 	/* LP0 watermarks always use 1/2 DDB partitioning */
 	ilk_compute_wm_maximums(dev, 0, &config, INTEL_DDB_PART_1_2, &max);
 
-	for (level = 0; level <= max_level; level++)
-		ilk_compute_wm_level(dev_priv, level, params,
-				     &pipe_wm->wm[level]);
+	/* At least LP0 must be valid */
+	if (!ilk_validate_wm_level(0, &max, &pipe_wm->wm[0]))
+		return false;
 
-	pipe_wm->linetime = hsw_compute_linetime_wm(dev, crtc);
+	ilk_compute_wm_reg_maximums(dev, 1, &max);
 
-	/* At least LP0 must be valid */
-	return ilk_validate_wm_level(0, &max, &pipe_wm->wm[0]);
+	for (level = 1; level <= max_level; level++) {
+		struct intel_wm_level wm = {};
+
+		ilk_compute_wm_level(dev_priv, level, params, &wm);
+
+		/*
+		 * Disable any watermark level that exceeds the
+		 * register maximums since such watermarks are
+		 * always invalid.
+		 */
+		if (!ilk_validate_wm_level(level, &max, &wm))
+			break;
+
+		pipe_wm->wm[level] = wm;
+	}
+
+	return true;
 }
 
 /*
@@ -2656,32 +2291,48 @@ static void ilk_merge_wm_level(struct drm_device *dev,
 {
 	const struct intel_crtc *intel_crtc;
 
-	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list, base.head) {
-		const struct intel_wm_level *wm =
-			&intel_crtc->wm.active.wm[level];
+	ret_wm->enable = true;
+
+	for_each_intel_crtc(dev, intel_crtc) {
+		const struct intel_pipe_wm *active = &intel_crtc->wm.active;
+		const struct intel_wm_level *wm = &active->wm[level];
+
+		if (!active->pipe_enabled)
+			continue;
 
+		/*
+		 * The watermark values may have been used in the past,
+		 * so we must maintain them in the registers for some
+		 * time even if the level is now disabled.
+		 */
 		if (!wm->enable)
-			return;
+			ret_wm->enable = false;
 
 		ret_wm->pri_val = max(ret_wm->pri_val, wm->pri_val);
 		ret_wm->spr_val = max(ret_wm->spr_val, wm->spr_val);
 		ret_wm->cur_val = max(ret_wm->cur_val, wm->cur_val);
 		ret_wm->fbc_val = max(ret_wm->fbc_val, wm->fbc_val);
 	}
-
-	ret_wm->enable = true;
 }
 
 /*
  * Merge all low power watermarks for all active pipes.
  */
 static void ilk_wm_merge(struct drm_device *dev,
-			 const struct hsw_wm_maximums *max,
+			 const struct intel_wm_config *config,
+			 const struct ilk_wm_maximums *max,
 			 struct intel_pipe_wm *merged)
 {
 	int level, max_level = ilk_wm_max_level(dev);
+	int last_enabled_level = max_level;
 
-	merged->fbc_wm_enabled = true;
+	/* ILK/SNB/IVB: LP1+ watermarks only w/ single pipe */
+	if ((INTEL_INFO(dev)->gen <= 6 || IS_IVYBRIDGE(dev)) &&
+	    config->num_pipes_active > 1)
+		return;
+
+	/* ILK: FBC WM must be disabled always */
+	merged->fbc_wm_enabled = INTEL_INFO(dev)->gen >= 6;
 
 	/* merge each WM1+ level */
 	for (level = 1; level <= max_level; level++) {
@@ -2689,18 +2340,36 @@ static void ilk_wm_merge(struct drm_device *dev,
 
 		ilk_merge_wm_level(dev, level, wm);
 
-		if (!ilk_validate_wm_level(level, max, wm))
-			break;
+		if (level > last_enabled_level)
+			wm->enable = false;
+		else if (!ilk_validate_wm_level(level, max, wm))
+			/* make sure all following levels get disabled */
+			last_enabled_level = level - 1;
 
 		/*
 		 * The spec says it is preferred to disable
 		 * FBC WMs instead of disabling a WM level.
 		 */
 		if (wm->fbc_val > max->fbc) {
-			merged->fbc_wm_enabled = false;
+			if (wm->enable)
+				merged->fbc_wm_enabled = false;
 			wm->fbc_val = 0;
 		}
 	}
+
+	/* ILK: LP2+ must be disabled when FBC WM is disabled but FBC enabled */
+	/*
+	 * FIXME this is racy. FBC might get enabled later.
+	 * What we should check here is whether FBC can be
+	 * enabled sometime later.
+	 */
+	if (IS_GEN5(dev) && !merged->fbc_wm_enabled && intel_fbc_enabled(dev)) {
+		for (level = 2; level <= max_level; level++) {
+			struct intel_wm_level *wm = &merged->wm[level];
+
+			wm->enable = false;
+		}
+	}
 }
 
 static int ilk_wm_lp_to_level(int wm_lp, const struct intel_pipe_wm *pipe_wm)
@@ -2709,10 +2378,21 @@ static int ilk_wm_lp_to_level(int wm_lp, const struct intel_pipe_wm *pipe_wm)
 	return wm_lp + (wm_lp >= 2 && pipe_wm->wm[4].enable);
 }
 
-static void hsw_compute_wm_results(struct drm_device *dev,
+/* The value we need to program into the WM_LPx latency field */
+static unsigned int ilk_wm_lp_latency(struct drm_device *dev, int level)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		return 2 * level;
+	else
+		return dev_priv->wm.pri_latency[level];
+}
+
+static void ilk_compute_wm_results(struct drm_device *dev,
 				   const struct intel_pipe_wm *merged,
 				   enum intel_ddb_partitioning partitioning,
-				   struct hsw_wm_values *results)
+				   struct ilk_wm_values *results)
 {
 	struct intel_crtc *intel_crtc;
 	int level, wm_lp;
@@ -2727,14 +2407,19 @@ static void hsw_compute_wm_results(struct drm_device *dev,
 		level = ilk_wm_lp_to_level(wm_lp, merged);
 
 		r = &merged->wm[level];
-		if (!r->enable)
-			break;
 
-		results->wm_lp[wm_lp - 1] = WM3_LP_EN |
-			((level * 2) << WM1_LP_LATENCY_SHIFT) |
+		/*
+		 * Maintain the watermark values even if the level is
+		 * disabled. Doing otherwise could cause underruns.
+		 */
+		results->wm_lp[wm_lp - 1] =
+			(ilk_wm_lp_latency(dev, level) << WM1_LP_LATENCY_SHIFT) |
 			(r->pri_val << WM1_LP_SR_SHIFT) |
 			r->cur_val;
 
+		if (r->enable)
+			results->wm_lp[wm_lp - 1] |= WM1_LP_SR_EN;
+
 		if (INTEL_INFO(dev)->gen >= 8)
 			results->wm_lp[wm_lp - 1] |=
 				r->fbc_val << WM1_LP_FBC_SHIFT_BDW;
@@ -2742,11 +2427,19 @@ static void hsw_compute_wm_results(struct drm_device *dev,
 			results->wm_lp[wm_lp - 1] |=
 				r->fbc_val << WM1_LP_FBC_SHIFT;
 
-		results->wm_lp_spr[wm_lp - 1] = r->spr_val;
+		/*
+		 * Always set WM1S_LP_EN when spr_val != 0, even if the
+		 * level is disabled. Doing otherwise could cause underruns.
+		 */
+		if (INTEL_INFO(dev)->gen <= 6 && r->spr_val) {
+			WARN_ON(wm_lp != 1);
+			results->wm_lp_spr[wm_lp - 1] = WM1S_LP_EN | r->spr_val;
+		} else
+			results->wm_lp_spr[wm_lp - 1] = r->spr_val;
 	}
 
 	/* LP0 register values */
-	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list, base.head) {
+	for_each_intel_crtc(dev, intel_crtc) {
 		enum pipe pipe = intel_crtc->pipe;
 		const struct intel_wm_level *r =
 			&intel_crtc->wm.active.wm[0];
@@ -2765,7 +2458,7 @@ static void hsw_compute_wm_results(struct drm_device *dev,
 
 /* Find the result with the highest level enabled. Check for enable_fbc_wm in
  * case both are at the same level. Prefer r1 in case they're the same. */
-static struct intel_pipe_wm *hsw_find_best_result(struct drm_device *dev,
+static struct intel_pipe_wm *ilk_find_best_result(struct drm_device *dev,
 						  struct intel_pipe_wm *r1,
 						  struct intel_pipe_wm *r2)
 {
@@ -2800,8 +2493,8 @@ static struct intel_pipe_wm *hsw_find_best_result(struct drm_device *dev,
 #define WM_DIRTY_DDB (1 << 25)
 
 static unsigned int ilk_compute_wm_dirty(struct drm_device *dev,
-					 const struct hsw_wm_values *old,
-					 const struct hsw_wm_values *new)
+					 const struct ilk_wm_values *old,
+					 const struct ilk_wm_values *new)
 {
 	unsigned int dirty = 0;
 	enum pipe pipe;
@@ -2851,27 +2544,53 @@ static unsigned int ilk_compute_wm_dirty(struct drm_device *dev,
 	return dirty;
 }
 
+static bool _ilk_disable_lp_wm(struct drm_i915_private *dev_priv,
+			       unsigned int dirty)
+{
+	struct ilk_wm_values *previous = &dev_priv->wm.hw;
+	bool changed = false;
+
+	if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] & WM1_LP_SR_EN) {
+		previous->wm_lp[2] &= ~WM1_LP_SR_EN;
+		I915_WRITE(WM3_LP_ILK, previous->wm_lp[2]);
+		changed = true;
+	}
+	if (dirty & WM_DIRTY_LP(2) && previous->wm_lp[1] & WM1_LP_SR_EN) {
+		previous->wm_lp[1] &= ~WM1_LP_SR_EN;
+		I915_WRITE(WM2_LP_ILK, previous->wm_lp[1]);
+		changed = true;
+	}
+	if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] & WM1_LP_SR_EN) {
+		previous->wm_lp[0] &= ~WM1_LP_SR_EN;
+		I915_WRITE(WM1_LP_ILK, previous->wm_lp[0]);
+		changed = true;
+	}
+
+	/*
+	 * Don't touch WM1S_LP_EN here.
+	 * Doing so could cause underruns.
+	 */
+
+	return changed;
+}
+
 /*
  * The spec says we shouldn't write when we don't need, because every write
  * causes WMs to be re-evaluated, expending some power.
  */
-static void hsw_write_wm_values(struct drm_i915_private *dev_priv,
-				struct hsw_wm_values *results)
+static void ilk_write_wm_values(struct drm_i915_private *dev_priv,
+				struct ilk_wm_values *results)
 {
-	struct hsw_wm_values *previous = &dev_priv->wm.hw;
+	struct drm_device *dev = dev_priv->dev;
+	struct ilk_wm_values *previous = &dev_priv->wm.hw;
 	unsigned int dirty;
 	uint32_t val;
 
-	dirty = ilk_compute_wm_dirty(dev_priv->dev, previous, results);
+	dirty = ilk_compute_wm_dirty(dev, previous, results);
 	if (!dirty)
 		return;
 
-	if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] != 0)
-		I915_WRITE(WM3_LP_ILK, 0);
-	if (dirty & WM_DIRTY_LP(2) && previous->wm_lp[1] != 0)
-		I915_WRITE(WM2_LP_ILK, 0);
-	if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] != 0)
-		I915_WRITE(WM1_LP_ILK, 0);
+	_ilk_disable_lp_wm(dev_priv, dirty);
 
 	if (dirty & WM_DIRTY_PIPE(PIPE_A))
 		I915_WRITE(WM0_PIPEA_ILK, results->wm_pipe[0]);
@@ -2888,12 +2607,21 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv,
 		I915_WRITE(PIPE_WM_LINETIME(PIPE_C), results->wm_linetime[2]);
 
 	if (dirty & WM_DIRTY_DDB) {
-		val = I915_READ(WM_MISC);
-		if (results->partitioning == INTEL_DDB_PART_1_2)
-			val &= ~WM_MISC_DATA_PARTITION_5_6;
-		else
-			val |= WM_MISC_DATA_PARTITION_5_6;
-		I915_WRITE(WM_MISC, val);
+		if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+			val = I915_READ(WM_MISC);
+			if (results->partitioning == INTEL_DDB_PART_1_2)
+				val &= ~WM_MISC_DATA_PARTITION_5_6;
+			else
+				val |= WM_MISC_DATA_PARTITION_5_6;
+			I915_WRITE(WM_MISC, val);
+		} else {
+			val = I915_READ(DISP_ARB_CTL2);
+			if (results->partitioning == INTEL_DDB_PART_1_2)
+				val &= ~DISP_DATA_PARTITION_5_6;
+			else
+				val |= DISP_DATA_PARTITION_5_6;
+			I915_WRITE(DISP_ARB_CTL2, val);
+		}
 	}
 
 	if (dirty & WM_DIRTY_FBC) {
@@ -2905,37 +2633,48 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv,
 		I915_WRITE(DISP_ARB_CTL, val);
 	}
 
-	if (dirty & WM_DIRTY_LP(1) && previous->wm_lp_spr[0] != results->wm_lp_spr[0])
+	if (dirty & WM_DIRTY_LP(1) &&
+	    previous->wm_lp_spr[0] != results->wm_lp_spr[0])
 		I915_WRITE(WM1S_LP_ILK, results->wm_lp_spr[0]);
-	if (dirty & WM_DIRTY_LP(2) && previous->wm_lp_spr[1] != results->wm_lp_spr[1])
-		I915_WRITE(WM2S_LP_IVB, results->wm_lp_spr[1]);
-	if (dirty & WM_DIRTY_LP(3) && previous->wm_lp_spr[2] != results->wm_lp_spr[2])
-		I915_WRITE(WM3S_LP_IVB, results->wm_lp_spr[2]);
 
-	if (dirty & WM_DIRTY_LP(1) && results->wm_lp[0] != 0)
+	if (INTEL_INFO(dev)->gen >= 7) {
+		if (dirty & WM_DIRTY_LP(2) && previous->wm_lp_spr[1] != results->wm_lp_spr[1])
+			I915_WRITE(WM2S_LP_IVB, results->wm_lp_spr[1]);
+		if (dirty & WM_DIRTY_LP(3) && previous->wm_lp_spr[2] != results->wm_lp_spr[2])
+			I915_WRITE(WM3S_LP_IVB, results->wm_lp_spr[2]);
+	}
+
+	if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] != results->wm_lp[0])
 		I915_WRITE(WM1_LP_ILK, results->wm_lp[0]);
-	if (dirty & WM_DIRTY_LP(2) && results->wm_lp[1] != 0)
+	if (dirty & WM_DIRTY_LP(2) && previous->wm_lp[1] != results->wm_lp[1])
 		I915_WRITE(WM2_LP_ILK, results->wm_lp[1]);
-	if (dirty & WM_DIRTY_LP(3) && results->wm_lp[2] != 0)
+	if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] != results->wm_lp[2])
 		I915_WRITE(WM3_LP_ILK, results->wm_lp[2]);
 
 	dev_priv->wm.hw = *results;
 }
 
-static void haswell_update_wm(struct drm_crtc *crtc)
+static bool ilk_disable_lp_wm(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	return _ilk_disable_lp_wm(dev_priv, WM_DIRTY_LP_ALL);
+}
+
+static void ilk_update_wm(struct drm_crtc *crtc)
 {
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct hsw_wm_maximums max;
-	struct hsw_pipe_wm_parameters params = {};
-	struct hsw_wm_values results = {};
+	struct ilk_wm_maximums max;
+	struct ilk_pipe_wm_parameters params = {};
+	struct ilk_wm_values results = {};
 	enum intel_ddb_partitioning partitioning;
 	struct intel_pipe_wm pipe_wm = {};
 	struct intel_pipe_wm lp_wm_1_2 = {}, lp_wm_5_6 = {}, *best_lp_wm;
 	struct intel_wm_config config = {};
 
-	hsw_compute_wm_parameters(crtc, &params, &config);
+	ilk_compute_wm_parameters(crtc, &params);
 
 	intel_compute_pipe_wm(crtc, &params, &pipe_wm);
 
@@ -2944,16 +2683,18 @@ static void haswell_update_wm(struct drm_crtc *crtc)
 
 	intel_crtc->wm.active = pipe_wm;
 
+	ilk_compute_wm_config(dev, &config);
+
 	ilk_compute_wm_maximums(dev, 1, &config, INTEL_DDB_PART_1_2, &max);
-	ilk_wm_merge(dev, &max, &lp_wm_1_2);
+	ilk_wm_merge(dev, &config, &max, &lp_wm_1_2);
 
 	/* 5/6 split only in single pipe config on IVB+ */
 	if (INTEL_INFO(dev)->gen >= 7 &&
 	    config.num_pipes_active == 1 && config.sprites_enabled) {
 		ilk_compute_wm_maximums(dev, 1, &config, INTEL_DDB_PART_5_6, &max);
-		ilk_wm_merge(dev, &max, &lp_wm_5_6);
+		ilk_wm_merge(dev, &config, &max, &lp_wm_5_6);
 
-		best_lp_wm = hsw_find_best_result(dev, &lp_wm_1_2, &lp_wm_5_6);
+		best_lp_wm = ilk_find_best_result(dev, &lp_wm_1_2, &lp_wm_5_6);
 	} else {
 		best_lp_wm = &lp_wm_1_2;
 	}
@@ -2961,16 +2702,17 @@ static void haswell_update_wm(struct drm_crtc *crtc)
 	partitioning = (best_lp_wm == &lp_wm_1_2) ?
 		       INTEL_DDB_PART_1_2 : INTEL_DDB_PART_5_6;
 
-	hsw_compute_wm_results(dev, best_lp_wm, partitioning, &results);
+	ilk_compute_wm_results(dev, best_lp_wm, partitioning, &results);
 
-	hsw_write_wm_values(dev_priv, &results);
+	ilk_write_wm_values(dev_priv, &results);
 }
 
-static void haswell_update_sprite_wm(struct drm_plane *plane,
+static void ilk_update_sprite_wm(struct drm_plane *plane,
 				     struct drm_crtc *crtc,
 				     uint32_t sprite_width, int pixel_size,
 				     bool enabled, bool scaled)
 {
+	struct drm_device *dev = plane->dev;
 	struct intel_plane *intel_plane = to_intel_plane(plane);
 
 	intel_plane->wm.enabled = enabled;
@@ -2978,176 +2720,24 @@ static void haswell_update_sprite_wm(struct drm_plane *plane,
 	intel_plane->wm.horiz_pixels = sprite_width;
 	intel_plane->wm.bytes_per_pixel = pixel_size;
 
-	haswell_update_wm(crtc);
-}
-
-static bool
-sandybridge_compute_sprite_wm(struct drm_device *dev, int plane,
-			      uint32_t sprite_width, int pixel_size,
-			      const struct intel_watermark_params *display,
-			      int display_latency_ns, int *sprite_wm)
-{
-	struct drm_crtc *crtc;
-	int clock;
-	int entries, tlb_miss;
-
-	crtc = intel_get_crtc_for_plane(dev, plane);
-	if (!intel_crtc_active(crtc)) {
-		*sprite_wm = display->guard_size;
-		return false;
-	}
-
-	clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock;
-
-	/* Use the small buffer method to calculate the sprite watermark */
-	entries = ((clock * pixel_size / 1000) * display_latency_ns) / 1000;
-	tlb_miss = display->fifo_size*display->cacheline_size -
-		sprite_width * 8;
-	if (tlb_miss > 0)
-		entries += tlb_miss;
-	entries = DIV_ROUND_UP(entries, display->cacheline_size);
-	*sprite_wm = entries + display->guard_size;
-	if (*sprite_wm > (int)display->max_wm)
-		*sprite_wm = display->max_wm;
-
-	return true;
-}
-
-static bool
-sandybridge_compute_sprite_srwm(struct drm_device *dev, int plane,
-				uint32_t sprite_width, int pixel_size,
-				const struct intel_watermark_params *display,
-				int latency_ns, int *sprite_wm)
-{
-	struct drm_crtc *crtc;
-	unsigned long line_time_us;
-	int clock;
-	int line_count, line_size;
-	int small, large;
-	int entries;
-
-	if (!latency_ns) {
-		*sprite_wm = 0;
-		return false;
-	}
-
-	crtc = intel_get_crtc_for_plane(dev, plane);
-	clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock;
-	if (!clock) {
-		*sprite_wm = 0;
-		return false;
-	}
-
-	line_time_us = (sprite_width * 1000) / clock;
-	if (!line_time_us) {
-		*sprite_wm = 0;
-		return false;
-	}
-
-	line_count = (latency_ns / line_time_us + 1000) / 1000;
-	line_size = sprite_width * pixel_size;
-
-	/* Use the minimum of the small and large buffer method for primary */
-	small = ((clock * pixel_size / 1000) * latency_ns) / 1000;
-	large = line_count * line_size;
-
-	entries = DIV_ROUND_UP(min(small, large), display->cacheline_size);
-	*sprite_wm = entries + display->guard_size;
-
-	return *sprite_wm > 0x3ff ? false : true;
-}
-
-static void sandybridge_update_sprite_wm(struct drm_plane *plane,
-					 struct drm_crtc *crtc,
-					 uint32_t sprite_width, int pixel_size,
-					 bool enabled, bool scaled)
-{
-	struct drm_device *dev = plane->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int pipe = to_intel_plane(plane)->pipe;
-	int latency = dev_priv->wm.spr_latency[0] * 100;	/* In unit 0.1us */
-	u32 val;
-	int sprite_wm, reg;
-	int ret;
-
-	if (!enabled)
-		return;
-
-	switch (pipe) {
-	case 0:
-		reg = WM0_PIPEA_ILK;
-		break;
-	case 1:
-		reg = WM0_PIPEB_ILK;
-		break;
-	case 2:
-		reg = WM0_PIPEC_IVB;
-		break;
-	default:
-		return; /* bad pipe */
-	}
-
-	ret = sandybridge_compute_sprite_wm(dev, pipe, sprite_width, pixel_size,
-					    &sandybridge_display_wm_info,
-					    latency, &sprite_wm);
-	if (!ret) {
-		DRM_DEBUG_KMS("failed to compute sprite wm for pipe %c\n",
-			      pipe_name(pipe));
-		return;
-	}
-
-	val = I915_READ(reg);
-	val &= ~WM0_PIPE_SPRITE_MASK;
-	I915_WRITE(reg, val | (sprite_wm << WM0_PIPE_SPRITE_SHIFT));
-	DRM_DEBUG_KMS("sprite watermarks For pipe %c - %d\n", pipe_name(pipe), sprite_wm);
-
-
-	ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width,
-					      pixel_size,
-					      &sandybridge_display_srwm_info,
-					      dev_priv->wm.spr_latency[1] * 500,
-					      &sprite_wm);
-	if (!ret) {
-		DRM_DEBUG_KMS("failed to compute sprite lp1 wm on pipe %c\n",
-			      pipe_name(pipe));
-		return;
-	}
-	I915_WRITE(WM1S_LP_ILK, sprite_wm);
-
-	/* Only IVB has two more LP watermarks for sprite */
-	if (!IS_IVYBRIDGE(dev))
-		return;
-
-	ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width,
-					      pixel_size,
-					      &sandybridge_display_srwm_info,
-					      dev_priv->wm.spr_latency[2] * 500,
-					      &sprite_wm);
-	if (!ret) {
-		DRM_DEBUG_KMS("failed to compute sprite lp2 wm on pipe %c\n",
-			      pipe_name(pipe));
-		return;
-	}
-	I915_WRITE(WM2S_LP_IVB, sprite_wm);
+	/*
+	 * IVB workaround: must disable low power watermarks for at least
+	 * one frame before enabling scaling.  LP watermarks can be re-enabled
+	 * when scaling is disabled.
+	 *
+	 * WaCxSRDisabledForSpriteScaling:ivb
+	 */
+	if (IS_IVYBRIDGE(dev) && scaled && ilk_disable_lp_wm(dev))
+		intel_wait_for_vblank(dev, intel_plane->pipe);
 
-	ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width,
-					      pixel_size,
-					      &sandybridge_display_srwm_info,
-					      dev_priv->wm.spr_latency[3] * 500,
-					      &sprite_wm);
-	if (!ret) {
-		DRM_DEBUG_KMS("failed to compute sprite lp3 wm on pipe %c\n",
-			      pipe_name(pipe));
-		return;
-	}
-	I915_WRITE(WM3S_LP_IVB, sprite_wm);
+	ilk_update_wm(crtc);
 }
 
 static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct hsw_wm_values *hw = &dev_priv->wm.hw;
+	struct ilk_wm_values *hw = &dev_priv->wm.hw;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_pipe_wm *active = &intel_crtc->wm.active;
 	enum pipe pipe = intel_crtc->pipe;
@@ -3158,9 +2748,12 @@ static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc)
 	};
 
 	hw->wm_pipe[pipe] = I915_READ(wm0_pipe_reg[pipe]);
-	hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe));
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe));
 
-	if (intel_crtc_active(crtc)) {
+	active->pipe_enabled = intel_crtc_active(crtc);
+
+	if (active->pipe_enabled) {
 		u32 tmp = hw->wm_pipe[pipe];
 
 		/*
@@ -3190,10 +2783,10 @@ static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc)
 void ilk_wm_get_hw_state(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct hsw_wm_values *hw = &dev_priv->wm.hw;
+	struct ilk_wm_values *hw = &dev_priv->wm.hw;
 	struct drm_crtc *crtc;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
+	for_each_crtc(dev, crtc)
 		ilk_pipe_wm_get_hw_state(crtc);
 
 	hw->wm_lp[0] = I915_READ(WM1_LP_ILK);
@@ -3201,11 +2794,17 @@ void ilk_wm_get_hw_state(struct drm_device *dev)
 	hw->wm_lp[2] = I915_READ(WM3_LP_ILK);
 
 	hw->wm_lp_spr[0] = I915_READ(WM1S_LP_ILK);
-	hw->wm_lp_spr[1] = I915_READ(WM2S_LP_IVB);
-	hw->wm_lp_spr[2] = I915_READ(WM3S_LP_IVB);
+	if (INTEL_INFO(dev)->gen >= 7) {
+		hw->wm_lp_spr[1] = I915_READ(WM2S_LP_IVB);
+		hw->wm_lp_spr[2] = I915_READ(WM3S_LP_IVB);
+	}
 
-	hw->partitioning = (I915_READ(WM_MISC) & WM_MISC_DATA_PARTITION_5_6) ?
-		INTEL_DDB_PART_5_6 : INTEL_DDB_PART_1_2;
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		hw->partitioning = (I915_READ(WM_MISC) & WM_MISC_DATA_PARTITION_5_6) ?
+			INTEL_DDB_PART_5_6 : INTEL_DDB_PART_1_2;
+	else if (IS_IVYBRIDGE(dev))
+		hw->partitioning = (I915_READ(DISP_ARB_CTL2) & DISP_DATA_PARTITION_5_6) ?
+			INTEL_DDB_PART_5_6 : INTEL_DDB_PART_1_2;
 
 	hw->enable_fbc_wm =
 		!(I915_READ(DISP_ARB_CTL) & DISP_FBC_WM_DIS);
@@ -3277,7 +2876,7 @@ intel_alloc_context_page(struct drm_device *dev)
 		return NULL;
 	}
 
-	ret = i915_gem_obj_ggtt_pin(ctx, 4096, true, false);
+	ret = i915_gem_obj_ggtt_pin(ctx, 4096, 0);
 	if (ret) {
 		DRM_ERROR("failed to pin power context: %d\n", ret);
 		goto err_unref;
@@ -3292,7 +2891,7 @@ intel_alloc_context_page(struct drm_device *dev)
 	return ctx;
 
 err_unpin:
-	i915_gem_object_unpin(ctx);
+	i915_gem_object_ggtt_unpin(ctx);
 err_unref:
 	drm_gem_object_unreference(&ctx->base);
 	return NULL;
@@ -3430,26 +3029,19 @@ static void ironlake_disable_drps(struct drm_device *dev)
  * ourselves, instead of doing a rmw cycle (which might result in us clearing
  * all limits and the gpu stuck at whatever frequency it is at atm).
  */
-static u32 gen6_rps_limits(struct drm_i915_private *dev_priv, u8 *val)
+static u32 gen6_rps_limits(struct drm_i915_private *dev_priv, u8 val)
 {
 	u32 limits;
 
-	limits = 0;
-
-	if (*val >= dev_priv->rps.max_delay)
-		*val = dev_priv->rps.max_delay;
-	limits |= dev_priv->rps.max_delay << 24;
-
 	/* Only set the down limit when we've reached the lowest level to avoid
 	 * getting more interrupts, otherwise leave this clear. This prevents a
 	 * race in the hw when coming out of rc6: There's a tiny window where
 	 * the hw runs at the minimal clock before selecting the desired
 	 * frequency, if the down threshold expires in that window we will not
 	 * receive a down interrupt. */
-	if (*val <= dev_priv->rps.min_delay) {
-		*val = dev_priv->rps.min_delay;
-		limits |= dev_priv->rps.min_delay << 16;
-	}
+	limits = dev_priv->rps.max_freq_softlimit << 24;
+	if (val <= dev_priv->rps.min_freq_softlimit)
+		limits |= dev_priv->rps.min_freq_softlimit << 16;
 
 	return limits;
 }
@@ -3461,26 +3053,26 @@ static void gen6_set_rps_thresholds(struct drm_i915_private *dev_priv, u8 val)
 	new_power = dev_priv->rps.power;
 	switch (dev_priv->rps.power) {
 	case LOW_POWER:
-		if (val > dev_priv->rps.rpe_delay + 1 && val > dev_priv->rps.cur_delay)
+		if (val > dev_priv->rps.efficient_freq + 1 && val > dev_priv->rps.cur_freq)
 			new_power = BETWEEN;
 		break;
 
 	case BETWEEN:
-		if (val <= dev_priv->rps.rpe_delay && val < dev_priv->rps.cur_delay)
+		if (val <= dev_priv->rps.efficient_freq && val < dev_priv->rps.cur_freq)
 			new_power = LOW_POWER;
-		else if (val >= dev_priv->rps.rp0_delay && val > dev_priv->rps.cur_delay)
+		else if (val >= dev_priv->rps.rp0_freq && val > dev_priv->rps.cur_freq)
 			new_power = HIGH_POWER;
 		break;
 
 	case HIGH_POWER:
-		if (val < (dev_priv->rps.rp1_delay + dev_priv->rps.rp0_delay) >> 1 && val < dev_priv->rps.cur_delay)
+		if (val < (dev_priv->rps.rp1_freq + dev_priv->rps.rp0_freq) >> 1 && val < dev_priv->rps.cur_freq)
 			new_power = BETWEEN;
 		break;
 	}
 	/* Max/min bins are special */
-	if (val == dev_priv->rps.min_delay)
+	if (val == dev_priv->rps.min_freq_softlimit)
 		new_power = LOW_POWER;
-	if (val == dev_priv->rps.max_delay)
+	if (val == dev_priv->rps.max_freq_softlimit)
 		new_power = HIGH_POWER;
 	if (new_power == dev_priv->rps.power)
 		return;
@@ -3546,49 +3138,122 @@ static void gen6_set_rps_thresholds(struct drm_i915_private *dev_priv, u8 val)
 	dev_priv->rps.last_adj = 0;
 }
 
+static u32 gen6_rps_pm_mask(struct drm_i915_private *dev_priv, u8 val)
+{
+	u32 mask = 0;
+
+	if (val > dev_priv->rps.min_freq_softlimit)
+		mask |= GEN6_PM_RP_DOWN_THRESHOLD | GEN6_PM_RP_DOWN_TIMEOUT;
+	if (val < dev_priv->rps.max_freq_softlimit)
+		mask |= GEN6_PM_RP_UP_THRESHOLD;
+
+	/* IVB and SNB hard hangs on looping batchbuffer
+	 * if GEN6_PM_UP_EI_EXPIRED is masked.
+	 */
+	if (INTEL_INFO(dev_priv->dev)->gen <= 7 && !IS_HASWELL(dev_priv->dev))
+		mask |= GEN6_PM_RP_UP_EI_EXPIRED;
+
+	if (IS_GEN8(dev_priv->dev))
+		mask |= GEN8_PMINTR_REDIRECT_TO_NON_DISP;
+
+	return ~mask;
+}
+
+/* gen6_set_rps is called to update the frequency request, but should also be
+ * called when the range (min_delay and max_delay) is modified so that we can
+ * update the GEN6_RP_INTERRUPT_LIMITS register accordingly. */
 void gen6_set_rps(struct drm_device *dev, u8 val)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 limits = gen6_rps_limits(dev_priv, &val);
 
 	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
-	WARN_ON(val > dev_priv->rps.max_delay);
-	WARN_ON(val < dev_priv->rps.min_delay);
-
-	if (val == dev_priv->rps.cur_delay)
-		return;
+	WARN_ON(val > dev_priv->rps.max_freq_softlimit);
+	WARN_ON(val < dev_priv->rps.min_freq_softlimit);
 
-	gen6_set_rps_thresholds(dev_priv, val);
+	/* min/max delay may still have been modified so be sure to
+	 * write the limits value.
+	 */
+	if (val != dev_priv->rps.cur_freq) {
+		gen6_set_rps_thresholds(dev_priv, val);
 
-	if (IS_HASWELL(dev))
-		I915_WRITE(GEN6_RPNSWREQ,
-			   HSW_FREQUENCY(val));
-	else
-		I915_WRITE(GEN6_RPNSWREQ,
-			   GEN6_FREQUENCY(val) |
-			   GEN6_OFFSET(0) |
-			   GEN6_AGGRESSIVE_TURBO);
+		if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+			I915_WRITE(GEN6_RPNSWREQ,
+				   HSW_FREQUENCY(val));
+		else
+			I915_WRITE(GEN6_RPNSWREQ,
+				   GEN6_FREQUENCY(val) |
+				   GEN6_OFFSET(0) |
+				   GEN6_AGGRESSIVE_TURBO);
+	}
 
 	/* Make sure we continue to get interrupts
 	 * until we hit the minimum or maximum frequencies.
 	 */
-	I915_WRITE(GEN6_RP_INTERRUPT_LIMITS, limits);
+	I915_WRITE(GEN6_RP_INTERRUPT_LIMITS, gen6_rps_limits(dev_priv, val));
+	I915_WRITE(GEN6_PMINTRMSK, gen6_rps_pm_mask(dev_priv, val));
 
 	POSTING_READ(GEN6_RPNSWREQ);
 
-	dev_priv->rps.cur_delay = val;
-
+	dev_priv->rps.cur_freq = val;
 	trace_intel_gpu_freq_change(val * 50);
 }
 
+/* vlv_set_rps_idle: Set the frequency to Rpn if Gfx clocks are down
+ *
+ * * If Gfx is Idle, then
+ * 1. Mask Turbo interrupts
+ * 2. Bring up Gfx clock
+ * 3. Change the freq to Rpn and wait till P-Unit updates freq
+ * 4. Clear the Force GFX CLK ON bit so that Gfx can down
+ * 5. Unmask Turbo interrupts
+*/
+static void vlv_set_rps_idle(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+
+	/* Latest VLV doesn't need to force the gfx clock */
+	if (dev->pdev->revision >= 0xd) {
+		valleyview_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit);
+		return;
+	}
+
+	/*
+	 * When we are idle.  Drop to min voltage state.
+	 */
+
+	if (dev_priv->rps.cur_freq <= dev_priv->rps.min_freq_softlimit)
+		return;
+
+	/* Mask turbo interrupt so that they will not come in between */
+	I915_WRITE(GEN6_PMINTRMSK, 0xffffffff);
+
+	vlv_force_gfx_clock(dev_priv, true);
+
+	dev_priv->rps.cur_freq = dev_priv->rps.min_freq_softlimit;
+
+	vlv_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ,
+					dev_priv->rps.min_freq_softlimit);
+
+	if (wait_for(((vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS))
+				& GENFREQSTATUS) == 0, 5))
+		DRM_ERROR("timed out waiting for Punit\n");
+
+	vlv_force_gfx_clock(dev_priv, false);
+
+	I915_WRITE(GEN6_PMINTRMSK,
+		   gen6_rps_pm_mask(dev_priv, dev_priv->rps.cur_freq));
+}
+
 void gen6_rps_idle(struct drm_i915_private *dev_priv)
 {
+	struct drm_device *dev = dev_priv->dev;
+
 	mutex_lock(&dev_priv->rps.hw_lock);
 	if (dev_priv->rps.enabled) {
-		if (dev_priv->info->is_valleyview)
-			valleyview_set_rps(dev_priv->dev, dev_priv->rps.min_delay);
+		if (IS_VALLEYVIEW(dev))
+			vlv_set_rps_idle(dev_priv);
 		else
-			gen6_set_rps(dev_priv->dev, dev_priv->rps.min_delay);
+			gen6_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit);
 		dev_priv->rps.last_adj = 0;
 	}
 	mutex_unlock(&dev_priv->rps.hw_lock);
@@ -3596,68 +3261,59 @@ void gen6_rps_idle(struct drm_i915_private *dev_priv)
 
 void gen6_rps_boost(struct drm_i915_private *dev_priv)
 {
+	struct drm_device *dev = dev_priv->dev;
+
 	mutex_lock(&dev_priv->rps.hw_lock);
 	if (dev_priv->rps.enabled) {
-		if (dev_priv->info->is_valleyview)
-			valleyview_set_rps(dev_priv->dev, dev_priv->rps.max_delay);
+		if (IS_VALLEYVIEW(dev))
+			valleyview_set_rps(dev_priv->dev, dev_priv->rps.max_freq_softlimit);
 		else
-			gen6_set_rps(dev_priv->dev, dev_priv->rps.max_delay);
+			gen6_set_rps(dev_priv->dev, dev_priv->rps.max_freq_softlimit);
 		dev_priv->rps.last_adj = 0;
 	}
 	mutex_unlock(&dev_priv->rps.hw_lock);
 }
 
-/*
- * Wait until the previous freq change has completed,
- * or the timeout elapsed, and then update our notion
- * of the current GPU frequency.
- */
-static void vlv_update_rps_cur_delay(struct drm_i915_private *dev_priv)
+void valleyview_set_rps(struct drm_device *dev, u8 val)
 {
-	u32 pval;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
+	WARN_ON(val > dev_priv->rps.max_freq_softlimit);
+	WARN_ON(val < dev_priv->rps.min_freq_softlimit);
 
-	if (wait_for(((pval = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS)) & GENFREQSTATUS) == 0, 10))
-		DRM_DEBUG_DRIVER("timed out waiting for Punit\n");
+	DRM_DEBUG_DRIVER("GPU freq request from %d MHz (%u) to %d MHz (%u)\n",
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.cur_freq),
+			 dev_priv->rps.cur_freq,
+			 vlv_gpu_freq(dev_priv, val), val);
 
-	pval >>= 8;
+	if (val != dev_priv->rps.cur_freq)
+		vlv_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ, val);
 
-	if (pval != dev_priv->rps.cur_delay)
-		DRM_DEBUG_DRIVER("Punit overrode GPU freq: %d MHz (%u) requested, but got %d Mhz (%u)\n",
-				 vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.cur_delay),
-				 dev_priv->rps.cur_delay,
-				 vlv_gpu_freq(dev_priv->mem_freq, pval), pval);
+	I915_WRITE(GEN6_PMINTRMSK, gen6_rps_pm_mask(dev_priv, val));
 
-	dev_priv->rps.cur_delay = pval;
+	dev_priv->rps.cur_freq = val;
+	trace_intel_gpu_freq_change(vlv_gpu_freq(dev_priv, val));
 }
 
-void valleyview_set_rps(struct drm_device *dev, u8 val)
+static void gen8_disable_rps_interrupts(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	gen6_rps_limits(dev_priv, &val);
-
-	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
-	WARN_ON(val > dev_priv->rps.max_delay);
-	WARN_ON(val < dev_priv->rps.min_delay);
-
-	vlv_update_rps_cur_delay(dev_priv);
-
-	DRM_DEBUG_DRIVER("GPU freq request from %d MHz (%u) to %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.cur_delay),
-			 dev_priv->rps.cur_delay,
-			 vlv_gpu_freq(dev_priv->mem_freq, val), val);
-
-	if (val == dev_priv->rps.cur_delay)
-		return;
-
-	vlv_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ, val);
+	I915_WRITE(GEN6_PMINTRMSK, ~GEN8_PMINTR_REDIRECT_TO_NON_DISP);
+	I915_WRITE(GEN8_GT_IER(2), I915_READ(GEN8_GT_IER(2)) &
+				   ~dev_priv->pm_rps_events);
+	/* Complete PM interrupt masking here doesn't race with the rps work
+	 * item again unmasking PM interrupts because that is using a different
+	 * register (GEN8_GT_IMR(2)) to mask PM interrupts. The only risk is in
+	 * leaving stale bits in GEN8_GT_IIR(2) and GEN8_GT_IMR(2) which
+	 * gen8_enable_rps will clean up. */
 
-	dev_priv->rps.cur_delay = val;
+	spin_lock_irq(&dev_priv->irq_lock);
+	dev_priv->rps.pm_iir = 0;
+	spin_unlock_irq(&dev_priv->irq_lock);
 
-	trace_intel_gpu_freq_change(vlv_gpu_freq(dev_priv->mem_freq, val));
+	I915_WRITE(GEN8_GT_IIR(2), dev_priv->pm_rps_events);
 }
 
 static void gen6_disable_rps_interrupts(struct drm_device *dev)
@@ -3665,7 +3321,8 @@ static void gen6_disable_rps_interrupts(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	I915_WRITE(GEN6_PMINTRMSK, 0xffffffff);
-	I915_WRITE(GEN6_PMIER, I915_READ(GEN6_PMIER) & ~GEN6_PM_RPS_EVENTS);
+	I915_WRITE(GEN6_PMIER, I915_READ(GEN6_PMIER) &
+				~dev_priv->pm_rps_events);
 	/* Complete PM interrupt masking here doesn't race with the rps work
 	 * item again unmasking PM interrupts because that is using a different
 	 * register (PMIMR) to mask PM interrupts. The only risk is in leaving
@@ -3675,7 +3332,7 @@ static void gen6_disable_rps_interrupts(struct drm_device *dev)
 	dev_priv->rps.pm_iir = 0;
 	spin_unlock_irq(&dev_priv->irq_lock);
 
-	I915_WRITE(GEN6_PMIIR, GEN6_PM_RPS_EVENTS);
+	I915_WRITE(GEN6_PMIIR, dev_priv->pm_rps_events);
 }
 
 static void gen6_disable_rps(struct drm_device *dev)
@@ -3685,7 +3342,10 @@ static void gen6_disable_rps(struct drm_device *dev)
 	I915_WRITE(GEN6_RC_CONTROL, 0);
 	I915_WRITE(GEN6_RPNSWREQ, 1 << 31);
 
-	gen6_disable_rps_interrupts(dev);
+	if (IS_BROADWELL(dev))
+		gen8_disable_rps_interrupts(dev);
+	else
+		gen6_disable_rps_interrupts(dev);
 }
 
 static void valleyview_disable_rps(struct drm_device *dev)
@@ -3695,78 +3355,111 @@ static void valleyview_disable_rps(struct drm_device *dev)
 	I915_WRITE(GEN6_RC_CONTROL, 0);
 
 	gen6_disable_rps_interrupts(dev);
-
-	if (dev_priv->vlv_pctx) {
-		drm_gem_object_unreference(&dev_priv->vlv_pctx->base);
-		dev_priv->vlv_pctx = NULL;
-	}
 }
 
 static void intel_print_rc6_info(struct drm_device *dev, u32 mode)
 {
-	if (IS_GEN6(dev))
-		DRM_DEBUG_DRIVER("Sandybridge: deep RC6 disabled\n");
-
-	if (IS_HASWELL(dev))
-		DRM_DEBUG_DRIVER("Haswell: only RC6 available\n");
-
+	if (IS_VALLEYVIEW(dev)) {
+		if (mode & (GEN7_RC_CTL_TO_MODE | GEN6_RC_CTL_EI_MODE(1)))
+			mode = GEN6_RC_CTL_RC6_ENABLE;
+		else
+			mode = 0;
+	}
 	DRM_INFO("Enabling RC6 states: RC6 %s, RC6p %s, RC6pp %s\n",
-			(mode & GEN6_RC_CTL_RC6_ENABLE) ? "on" : "off",
-			(mode & GEN6_RC_CTL_RC6p_ENABLE) ? "on" : "off",
-			(mode & GEN6_RC_CTL_RC6pp_ENABLE) ? "on" : "off");
+		 (mode & GEN6_RC_CTL_RC6_ENABLE) ? "on" : "off",
+		 (mode & GEN6_RC_CTL_RC6p_ENABLE) ? "on" : "off",
+		 (mode & GEN6_RC_CTL_RC6pp_ENABLE) ? "on" : "off");
 }
 
-int intel_enable_rc6(const struct drm_device *dev)
+static int sanitize_rc6_option(const struct drm_device *dev, int enable_rc6)
 {
 	/* No RC6 before Ironlake */
 	if (INTEL_INFO(dev)->gen < 5)
 		return 0;
 
+	/* RC6 is only on Ironlake mobile not on desktop */
+	if (INTEL_INFO(dev)->gen == 5 && !IS_IRONLAKE_M(dev))
+		return 0;
+
 	/* Respect the kernel parameter if it is set */
-	if (i915_enable_rc6 >= 0)
-		return i915_enable_rc6;
+	if (enable_rc6 >= 0) {
+		int mask;
+
+		if (INTEL_INFO(dev)->gen == 6 || IS_IVYBRIDGE(dev))
+			mask = INTEL_RC6_ENABLE | INTEL_RC6p_ENABLE |
+			       INTEL_RC6pp_ENABLE;
+		else
+			mask = INTEL_RC6_ENABLE;
+
+		if ((enable_rc6 & mask) != enable_rc6)
+			DRM_INFO("Adjusting RC6 mask to %d (requested %d, valid %d)\n",
+				 enable_rc6 & mask, enable_rc6, mask);
+
+		return enable_rc6 & mask;
+	}
 
 	/* Disable RC6 on Ironlake */
 	if (INTEL_INFO(dev)->gen == 5)
 		return 0;
 
-	if (IS_HASWELL(dev))
-		return INTEL_RC6_ENABLE;
+	if (IS_IVYBRIDGE(dev))
+		return (INTEL_RC6_ENABLE | INTEL_RC6p_ENABLE);
+
+	return INTEL_RC6_ENABLE;
+}
+
+int intel_enable_rc6(const struct drm_device *dev)
+{
+	return i915.enable_rc6;
+}
 
-	/* snb/ivb have more than one rc6 state. */
-	if (INTEL_INFO(dev)->gen == 6)
-		return INTEL_RC6_ENABLE;
+static void gen8_enable_rps_interrupts(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	return (INTEL_RC6_ENABLE | INTEL_RC6p_ENABLE);
+	spin_lock_irq(&dev_priv->irq_lock);
+	WARN_ON(dev_priv->rps.pm_iir);
+	bdw_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
+	I915_WRITE(GEN8_GT_IIR(2), dev_priv->pm_rps_events);
+	spin_unlock_irq(&dev_priv->irq_lock);
 }
 
 static void gen6_enable_rps_interrupts(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 enabled_intrs;
 
 	spin_lock_irq(&dev_priv->irq_lock);
 	WARN_ON(dev_priv->rps.pm_iir);
-	snb_enable_pm_irq(dev_priv, GEN6_PM_RPS_EVENTS);
-	I915_WRITE(GEN6_PMIIR, GEN6_PM_RPS_EVENTS);
+	snb_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
+	I915_WRITE(GEN6_PMIIR, dev_priv->pm_rps_events);
 	spin_unlock_irq(&dev_priv->irq_lock);
+}
 
-	/* only unmask PM interrupts we need. Mask all others. */
-	enabled_intrs = GEN6_PM_RPS_EVENTS;
+static void parse_rp_state_cap(struct drm_i915_private *dev_priv, u32 rp_state_cap)
+{
+	/* All of these values are in units of 50MHz */
+	dev_priv->rps.cur_freq		= 0;
+	/* static values from HW: RP0 < RPe < RP1 < RPn (min_freq) */
+	dev_priv->rps.rp1_freq		= (rp_state_cap >>  8) & 0xff;
+	dev_priv->rps.rp0_freq		= (rp_state_cap >>  0) & 0xff;
+	dev_priv->rps.min_freq		= (rp_state_cap >> 16) & 0xff;
+	/* XXX: only BYT has a special efficient freq */
+	dev_priv->rps.efficient_freq	= dev_priv->rps.rp1_freq;
+	/* hw_max = RP0 until we check for overclocking */
+	dev_priv->rps.max_freq		= dev_priv->rps.rp0_freq;
 
-	/* IVB and SNB hard hangs on looping batchbuffer
-	 * if GEN6_PM_UP_EI_EXPIRED is masked.
-	 */
-	if (INTEL_INFO(dev)->gen <= 7 && !IS_HASWELL(dev))
-		enabled_intrs |= GEN6_PM_RP_UP_EI_EXPIRED;
+	/* Preserve min/max settings in case of re-init */
+	if (dev_priv->rps.max_freq_softlimit == 0)
+		dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq;
 
-	I915_WRITE(GEN6_PMINTRMSK, ~enabled_intrs);
+	if (dev_priv->rps.min_freq_softlimit == 0)
+		dev_priv->rps.min_freq_softlimit = dev_priv->rps.min_freq;
 }
 
 static void gen8_enable_rps(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	uint32_t rc6_mask = 0, rp_state_cap;
 	int unused;
 
@@ -3775,12 +3468,13 @@ static void gen8_enable_rps(struct drm_device *dev)
 
 	/* 1c & 1d: Get forcewake during program sequence. Although the driver
 	 * hasn't enabled a state yet where we need forcewake, BIOS may have.*/
-	gen6_gt_force_wake_get(dev_priv);
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	/* 2a: Disable RC states. */
 	I915_WRITE(GEN6_RC_CONTROL, 0);
 
 	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
+	parse_rp_state_cap(dev_priv, rp_state_cap);
 
 	/* 2b: Program RC6 thresholds.*/
 	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16);
@@ -3794,21 +3488,23 @@ static void gen8_enable_rps(struct drm_device *dev)
 	/* 3: Enable RC6 */
 	if (intel_enable_rc6(dev) & INTEL_RC6_ENABLE)
 		rc6_mask = GEN6_RC_CTL_RC6_ENABLE;
-	DRM_INFO("RC6 %s\n", (rc6_mask & GEN6_RC_CTL_RC6_ENABLE) ? "on" : "off");
+	intel_print_rc6_info(dev, rc6_mask);
 	I915_WRITE(GEN6_RC_CONTROL, GEN6_RC_CTL_HW_ENABLE |
-			GEN6_RC_CTL_EI_MODE(1) |
-			rc6_mask);
+				    GEN6_RC_CTL_EI_MODE(1) |
+				    rc6_mask);
 
 	/* 4 Program defaults and thresholds for RPS*/
-	I915_WRITE(GEN6_RPNSWREQ, HSW_FREQUENCY(10)); /* Request 500 MHz */
-	I915_WRITE(GEN6_RC_VIDEO_FREQ, HSW_FREQUENCY(12)); /* Request 600 MHz */
+	I915_WRITE(GEN6_RPNSWREQ,
+		   HSW_FREQUENCY(dev_priv->rps.rp1_freq));
+	I915_WRITE(GEN6_RC_VIDEO_FREQ,
+		   HSW_FREQUENCY(dev_priv->rps.rp1_freq));
 	/* NB: Docs say 1s, and 1000000 - which aren't equivalent */
 	I915_WRITE(GEN6_RP_DOWN_TIMEOUT, 100000000 / 128); /* 1 second timeout */
 
 	/* Docs recommend 900MHz, and 300 MHz respectively */
 	I915_WRITE(GEN6_RP_INTERRUPT_LIMITS,
-		   dev_priv->rps.max_delay << 24 |
-		   dev_priv->rps.min_delay << 16);
+		   dev_priv->rps.max_freq_softlimit << 24 |
+		   dev_priv->rps.min_freq_softlimit << 16);
 
 	I915_WRITE(GEN6_RP_UP_THRESHOLD, 7600000 / 128); /* 76ms busyness per EI, 90% */
 	I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 31300000 / 128); /* 313ms busyness per EI, 70%*/
@@ -3830,18 +3526,18 @@ static void gen8_enable_rps(struct drm_device *dev)
 
 	gen6_set_rps(dev, (I915_READ(GEN6_GT_PERF_STATUS) & 0xff00) >> 8);
 
-	gen6_enable_rps_interrupts(dev);
+	gen8_enable_rps_interrupts(dev);
 
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
 static void gen6_enable_rps(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	u32 rp_state_cap;
 	u32 gt_perf_status;
-	u32 rc6vids, pcu_mbox, rc6_mask = 0;
+	u32 rc6vids, pcu_mbox = 0, rc6_mask = 0;
 	u32 gtfifodbg;
 	int rc6_mode;
 	int i, ret;
@@ -3862,18 +3558,12 @@ static void gen6_enable_rps(struct drm_device *dev)
 		I915_WRITE(GTFIFODBG, gtfifodbg);
 	}
 
-	gen6_gt_force_wake_get(dev_priv);
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
 	gt_perf_status = I915_READ(GEN6_GT_PERF_STATUS);
 
-	/* In units of 50MHz */
-	dev_priv->rps.hw_max = dev_priv->rps.max_delay = rp_state_cap & 0xff;
-	dev_priv->rps.min_delay = (rp_state_cap >> 16) & 0xff;
-	dev_priv->rps.rp1_delay = (rp_state_cap >>  8) & 0xff;
-	dev_priv->rps.rp0_delay = (rp_state_cap >>  0) & 0xff;
-	dev_priv->rps.rpe_delay = dev_priv->rps.rp1_delay;
-	dev_priv->rps.cur_delay = 0;
+	parse_rp_state_cap(dev_priv, rp_state_cap);
 
 	/* disable the counters and set deterministic thresholds */
 	I915_WRITE(GEN6_RC_CONTROL, 0);
@@ -3922,21 +3612,19 @@ static void gen6_enable_rps(struct drm_device *dev)
 	I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10);
 
 	ret = sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_MIN_FREQ_TABLE, 0);
-	if (!ret) {
-		pcu_mbox = 0;
-		ret = sandybridge_pcode_read(dev_priv, GEN6_READ_OC_PARAMS, &pcu_mbox);
-		if (!ret && (pcu_mbox & (1<<31))) { /* OC supported */
-			DRM_DEBUG_DRIVER("Overclocking supported. Max: %dMHz, Overclock max: %dMHz\n",
-					 (dev_priv->rps.max_delay & 0xff) * 50,
-					 (pcu_mbox & 0xff) * 50);
-			dev_priv->rps.hw_max = pcu_mbox & 0xff;
-		}
-	} else {
+	if (ret)
 		DRM_DEBUG_DRIVER("Failed to set the min frequency\n");
+
+	ret = sandybridge_pcode_read(dev_priv, GEN6_READ_OC_PARAMS, &pcu_mbox);
+	if (!ret && (pcu_mbox & (1<<31))) { /* OC supported */
+		DRM_DEBUG_DRIVER("Overclocking supported. Max: %dMHz, Overclock max: %dMHz\n",
+				 (dev_priv->rps.max_freq_softlimit & 0xff) * 50,
+				 (pcu_mbox & 0xff) * 50);
+		dev_priv->rps.max_freq = pcu_mbox & 0xff;
 	}
 
 	dev_priv->rps.power = HIGH_POWER; /* force a reset */
-	gen6_set_rps(dev_priv->dev, dev_priv->rps.min_delay);
+	gen6_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit);
 
 	gen6_enable_rps_interrupts(dev);
 
@@ -3954,10 +3642,10 @@ static void gen6_enable_rps(struct drm_device *dev)
 			DRM_ERROR("Couldn't fix incorrect rc6 voltage\n");
 	}
 
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
-void gen6_update_ring_freq(struct drm_device *dev)
+static void __gen6_update_ring_freq(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int min_freq = 15;
@@ -3992,9 +3680,9 @@ void gen6_update_ring_freq(struct drm_device *dev)
 	 * to use for memory access.  We do this by specifying the IA frequency
 	 * the PCU should use as a reference to determine the ring frequency.
 	 */
-	for (gpu_freq = dev_priv->rps.max_delay; gpu_freq >= dev_priv->rps.min_delay;
+	for (gpu_freq = dev_priv->rps.max_freq_softlimit; gpu_freq >= dev_priv->rps.min_freq_softlimit;
 	     gpu_freq--) {
-		int diff = dev_priv->rps.max_delay - gpu_freq;
+		int diff = dev_priv->rps.max_freq_softlimit - gpu_freq;
 		unsigned int ia_freq = 0, ring_freq = 0;
 
 		if (INTEL_INFO(dev)->gen >= 8) {
@@ -4027,6 +3715,18 @@ void gen6_update_ring_freq(struct drm_device *dev)
 	}
 }
 
+void gen6_update_ring_freq(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (INTEL_INFO(dev)->gen < 6 || IS_VALLEYVIEW(dev))
+		return;
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+	__gen6_update_ring_freq(dev);
+	mutex_unlock(&dev_priv->rps.hw_lock);
+}
+
 int valleyview_rps_max_freq(struct drm_i915_private *dev_priv)
 {
 	u32 val, rp0;
@@ -4057,6 +3757,15 @@ int valleyview_rps_min_freq(struct drm_i915_private *dev_priv)
 	return vlv_punit_read(dev_priv, PUNIT_REG_GPU_LFM) & 0xff;
 }
 
+/* Check that the pctx buffer wasn't move under us. */
+static void valleyview_check_pctx(struct drm_i915_private *dev_priv)
+{
+	unsigned long pctx_addr = I915_READ(VLV_PCBR) & ~4095;
+
+	WARN_ON(pctx_addr != dev_priv->mm.stolen_base +
+			     dev_priv->vlv_pctx->stolen->start);
+}
+
 static void valleyview_setup_pctx(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -4065,6 +3774,8 @@ static void valleyview_setup_pctx(struct drm_device *dev)
 	u32 pcbr;
 	int pctx_size = 24*1024;
 
+	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
+
 	pcbr = I915_READ(VLV_PCBR);
 	if (pcbr) {
 		/* BIOS set it up already, grab the pre-alloc'd space */
@@ -4099,24 +3810,75 @@ out:
 	dev_priv->vlv_pctx = pctx;
 }
 
+static void valleyview_cleanup_pctx(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (WARN_ON(!dev_priv->vlv_pctx))
+		return;
+
+	drm_gem_object_unreference(&dev_priv->vlv_pctx->base);
+	dev_priv->vlv_pctx = NULL;
+}
+
+static void valleyview_init_gt_powersave(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	valleyview_setup_pctx(dev);
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+
+	dev_priv->rps.max_freq = valleyview_rps_max_freq(dev_priv);
+	dev_priv->rps.rp0_freq = dev_priv->rps.max_freq;
+	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.max_freq),
+			 dev_priv->rps.max_freq);
+
+	dev_priv->rps.efficient_freq = valleyview_rps_rpe_freq(dev_priv);
+	DRM_DEBUG_DRIVER("RPe GPU freq: %d MHz (%u)\n",
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.efficient_freq),
+			 dev_priv->rps.efficient_freq);
+
+	dev_priv->rps.min_freq = valleyview_rps_min_freq(dev_priv);
+	DRM_DEBUG_DRIVER("min GPU freq: %d MHz (%u)\n",
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.min_freq),
+			 dev_priv->rps.min_freq);
+
+	/* Preserve min/max settings in case of re-init */
+	if (dev_priv->rps.max_freq_softlimit == 0)
+		dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq;
+
+	if (dev_priv->rps.min_freq_softlimit == 0)
+		dev_priv->rps.min_freq_softlimit = dev_priv->rps.min_freq;
+
+	mutex_unlock(&dev_priv->rps.hw_lock);
+}
+
+static void valleyview_cleanup_gt_powersave(struct drm_device *dev)
+{
+	valleyview_cleanup_pctx(dev);
+}
+
 static void valleyview_enable_rps(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	u32 gtfifodbg, val, rc6_mode = 0;
 	int i;
 
 	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
 
+	valleyview_check_pctx(dev_priv);
+
 	if ((gtfifodbg = I915_READ(GTFIFODBG))) {
 		DRM_DEBUG_DRIVER("GT fifo had a previous error %x\n",
 				 gtfifodbg);
 		I915_WRITE(GTFIFODBG, gtfifodbg);
 	}
 
-	valleyview_setup_pctx(dev);
-
-	gen6_gt_force_wake_get(dev_priv);
+	/* If VLV, Forcewake all wells, else re-direct to regular path */
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	I915_WRITE(GEN6_RP_UP_THRESHOLD, 59400);
 	I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 245000);
@@ -4140,7 +3902,7 @@ static void valleyview_enable_rps(struct drm_device *dev)
 	for_each_ring(ring, dev_priv, i)
 		I915_WRITE(RING_MAX_IDLE(ring->mmio_base), 10);
 
-	I915_WRITE(GEN6_RC6_THRESHOLD, 0xc350);
+	I915_WRITE(GEN6_RC6_THRESHOLD, 0x557);
 
 	/* allows RC6 residency counter to work */
 	I915_WRITE(VLV_COUNTER_CONTROL,
@@ -4148,65 +3910,31 @@ static void valleyview_enable_rps(struct drm_device *dev)
 				      VLV_MEDIA_RC6_COUNT_EN |
 				      VLV_RENDER_RC6_COUNT_EN));
 	if (intel_enable_rc6(dev) & INTEL_RC6_ENABLE)
-		rc6_mode = GEN7_RC_CTL_TO_MODE;
+		rc6_mode = GEN7_RC_CTL_TO_MODE | VLV_RC_CTL_CTX_RST_PARALLEL;
 
 	intel_print_rc6_info(dev, rc6_mode);
 
 	I915_WRITE(GEN6_RC_CONTROL, rc6_mode);
 
 	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
-	switch ((val >> 6) & 3) {
-	case 0:
-	case 1:
-		dev_priv->mem_freq = 800;
-		break;
-	case 2:
-		dev_priv->mem_freq = 1066;
-		break;
-	case 3:
-		dev_priv->mem_freq = 1333;
-		break;
-	}
-	DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq);
 
 	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & 0x10 ? "yes" : "no");
 	DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
 
-	dev_priv->rps.cur_delay = (val >> 8) & 0xff;
+	dev_priv->rps.cur_freq = (val >> 8) & 0xff;
 	DRM_DEBUG_DRIVER("current GPU freq: %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.cur_delay),
-			 dev_priv->rps.cur_delay);
-
-	dev_priv->rps.max_delay = valleyview_rps_max_freq(dev_priv);
-	dev_priv->rps.hw_max = dev_priv->rps.max_delay;
-	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.max_delay),
-			 dev_priv->rps.max_delay);
-
-	dev_priv->rps.rpe_delay = valleyview_rps_rpe_freq(dev_priv);
-	DRM_DEBUG_DRIVER("RPe GPU freq: %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.rpe_delay),
-			 dev_priv->rps.rpe_delay);
-
-	dev_priv->rps.min_delay = valleyview_rps_min_freq(dev_priv);
-	DRM_DEBUG_DRIVER("min GPU freq: %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.min_delay),
-			 dev_priv->rps.min_delay);
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.cur_freq),
+			 dev_priv->rps.cur_freq);
 
 	DRM_DEBUG_DRIVER("setting GPU freq to %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.rpe_delay),
-			 dev_priv->rps.rpe_delay);
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.efficient_freq),
+			 dev_priv->rps.efficient_freq);
 
-	valleyview_set_rps(dev_priv->dev, dev_priv->rps.rpe_delay);
+	valleyview_set_rps(dev_priv->dev, dev_priv->rps.efficient_freq);
 
 	gen6_enable_rps_interrupts(dev);
 
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
 void ironlake_teardown_rc6(struct drm_device *dev)
@@ -4214,13 +3942,13 @@ void ironlake_teardown_rc6(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	if (dev_priv->ips.renderctx) {
-		i915_gem_object_unpin(dev_priv->ips.renderctx);
+		i915_gem_object_ggtt_unpin(dev_priv->ips.renderctx);
 		drm_gem_object_unreference(&dev_priv->ips.renderctx->base);
 		dev_priv->ips.renderctx = NULL;
 	}
 
 	if (dev_priv->ips.pwrctx) {
-		i915_gem_object_unpin(dev_priv->ips.pwrctx);
+		i915_gem_object_ggtt_unpin(dev_priv->ips.pwrctx);
 		drm_gem_object_unreference(&dev_priv->ips.pwrctx->base);
 		dev_priv->ips.pwrctx = NULL;
 	}
@@ -4266,7 +3994,7 @@ static int ironlake_setup_rc6(struct drm_device *dev)
 static void ironlake_enable_rc6(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
 	bool was_interruptible;
 	int ret;
 
@@ -4324,7 +4052,7 @@ static void ironlake_enable_rc6(struct drm_device *dev)
 	I915_WRITE(PWRCTXA, i915_gem_obj_ggtt_offset(dev_priv->ips.pwrctx) | PWRCTX_EN);
 	I915_WRITE(RSTDBYCTL, I915_READ(RSTDBYCTL) & ~RCX_SW_EXIT);
 
-	intel_print_rc6_info(dev, INTEL_RC6_ENABLE);
+	intel_print_rc6_info(dev, GEN6_RC_CTL_RC6_ENABLE);
 }
 
 static unsigned long intel_pxfreq(u32 vidfreq)
@@ -4412,9 +4140,10 @@ static unsigned long __i915_chipset_val(struct drm_i915_private *dev_priv)
 
 unsigned long i915_chipset_val(struct drm_i915_private *dev_priv)
 {
+	struct drm_device *dev = dev_priv->dev;
 	unsigned long val;
 
-	if (dev_priv->info->gen != 5)
+	if (INTEL_INFO(dev)->gen != 5)
 		return 0;
 
 	spin_lock_irq(&mchdev_lock);
@@ -4443,6 +4172,7 @@ unsigned long i915_mch_val(struct drm_i915_private *dev_priv)
 
 static u16 pvid_to_extvid(struct drm_i915_private *dev_priv, u8 pxvid)
 {
+	struct drm_device *dev = dev_priv->dev;
 	static const struct v_table {
 		u16 vd; /* in .1 mil */
 		u16 vm; /* in .1 mil */
@@ -4576,7 +4306,7 @@ static u16 pvid_to_extvid(struct drm_i915_private *dev_priv, u8 pxvid)
 		{ 16000, 14875, },
 		{ 16125, 15000, },
 	};
-	if (dev_priv->info->is_mobile)
+	if (INTEL_INFO(dev)->is_mobile)
 		return v_table[pxvid].vm;
 	else
 		return v_table[pxvid].vd;
@@ -4619,7 +4349,9 @@ static void __i915_update_gfx_val(struct drm_i915_private *dev_priv)
 
 void i915_update_gfx_val(struct drm_i915_private *dev_priv)
 {
-	if (dev_priv->info->gen != 5)
+	struct drm_device *dev = dev_priv->dev;
+
+	if (INTEL_INFO(dev)->gen != 5)
 		return;
 
 	spin_lock_irq(&mchdev_lock);
@@ -4636,7 +4368,7 @@ static unsigned long __i915_gfx_val(struct drm_i915_private *dev_priv)
 
 	assert_spin_locked(&mchdev_lock);
 
-	pxvid = I915_READ(PXVFREQ_BASE + (dev_priv->rps.cur_delay * 4));
+	pxvid = I915_READ(PXVFREQ_BASE + (dev_priv->rps.cur_freq * 4));
 	pxvid = (pxvid >> 24) & 0x7f;
 	ext_v = pvid_to_extvid(dev_priv, pxvid);
 
@@ -4668,9 +4400,10 @@ static unsigned long __i915_gfx_val(struct drm_i915_private *dev_priv)
 
 unsigned long i915_gfx_val(struct drm_i915_private *dev_priv)
 {
+	struct drm_device *dev = dev_priv->dev;
 	unsigned long val;
 
-	if (dev_priv->info->gen != 5)
+	if (INTEL_INFO(dev)->gen != 5)
 		return 0;
 
 	spin_lock_irq(&mchdev_lock);
@@ -4773,7 +4506,7 @@ EXPORT_SYMBOL_GPL(i915_gpu_lower);
 bool i915_gpu_busy(void)
 {
 	struct drm_i915_private *dev_priv;
-	struct intel_ring_buffer *ring;
+	struct intel_engine_cs *ring;
 	bool ret = false;
 	int i;
 
@@ -4859,6 +4592,7 @@ void intel_gpu_ips_teardown(void)
 	i915_mch_dev = NULL;
 	spin_unlock_irq(&mchdev_lock);
 }
+
 static void intel_init_emon(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -4930,6 +4664,20 @@ static void intel_init_emon(struct drm_device *dev)
 	dev_priv->ips.corr = (lcfuse & LCFUSE_HIV_MASK);
 }
 
+void intel_init_gt_powersave(struct drm_device *dev)
+{
+	i915.enable_rc6 = sanitize_rc6_option(dev, i915.enable_rc6);
+
+	if (IS_VALLEYVIEW(dev))
+		valleyview_init_gt_powersave(dev);
+}
+
+void intel_cleanup_gt_powersave(struct drm_device *dev)
+{
+	if (IS_VALLEYVIEW(dev))
+		valleyview_cleanup_gt_powersave(dev);
+}
+
 void intel_disable_gt_powersave(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -4940,8 +4688,10 @@ void intel_disable_gt_powersave(struct drm_device *dev)
 	if (IS_IRONLAKE_M(dev)) {
 		ironlake_disable_drps(dev);
 		ironlake_disable_rc6(dev);
-	} else if (INTEL_INFO(dev)->gen >= 6) {
-		cancel_delayed_work_sync(&dev_priv->rps.delayed_resume_work);
+	} else if (IS_GEN6(dev) || IS_GEN7(dev) || IS_BROADWELL(dev)) {
+		if (cancel_delayed_work_sync(&dev_priv->rps.delayed_resume_work))
+			intel_runtime_pm_put(dev_priv);
+
 		cancel_work_sync(&dev_priv->rps.work);
 		mutex_lock(&dev_priv->rps.hw_lock);
 		if (IS_VALLEYVIEW(dev))
@@ -4966,13 +4716,15 @@ static void intel_gen6_powersave_work(struct work_struct *work)
 		valleyview_enable_rps(dev);
 	} else if (IS_BROADWELL(dev)) {
 		gen8_enable_rps(dev);
-		gen6_update_ring_freq(dev);
+		__gen6_update_ring_freq(dev);
 	} else {
 		gen6_enable_rps(dev);
-		gen6_update_ring_freq(dev);
+		__gen6_update_ring_freq(dev);
 	}
 	dev_priv->rps.enabled = true;
 	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	intel_runtime_pm_put(dev_priv);
 }
 
 void intel_enable_gt_powersave(struct drm_device *dev)
@@ -4980,20 +4732,38 @@ void intel_enable_gt_powersave(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	if (IS_IRONLAKE_M(dev)) {
+		mutex_lock(&dev->struct_mutex);
 		ironlake_enable_drps(dev);
 		ironlake_enable_rc6(dev);
 		intel_init_emon(dev);
-	} else if (IS_GEN6(dev) || IS_GEN7(dev)) {
+		mutex_unlock(&dev->struct_mutex);
+	} else if (IS_GEN6(dev) || IS_GEN7(dev) || IS_BROADWELL(dev)) {
 		/*
 		 * PCU communication is slow and this doesn't need to be
 		 * done at any specific time, so do this out of our fast path
 		 * to make resume and init faster.
+		 *
+		 * We depend on the HW RC6 power context save/restore
+		 * mechanism when entering D3 through runtime PM suspend. So
+		 * disable RPM until RPS/RC6 is properly setup. We can only
+		 * get here via the driver load/system resume/runtime resume
+		 * paths, so the _noresume version is enough (and in case of
+		 * runtime resume it's necessary).
 		 */
-		schedule_delayed_work(&dev_priv->rps.delayed_resume_work,
-				      round_jiffies_up_relative(HZ));
+		if (schedule_delayed_work(&dev_priv->rps.delayed_resume_work,
+					   round_jiffies_up_relative(HZ)))
+			intel_runtime_pm_get_noresume(dev_priv);
 	}
 }
 
+void intel_reset_gt_powersave(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	dev_priv->rps.enabled = false;
+	intel_enable_gt_powersave(dev);
+}
+
 static void ibx_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -5019,6 +4789,20 @@ static void g4x_disable_trickle_feed(struct drm_device *dev)
 	}
 }
 
+static void ilk_init_lp_watermarks(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	I915_WRITE(WM3_LP_ILK, I915_READ(WM3_LP_ILK) & ~WM1_LP_SR_EN);
+	I915_WRITE(WM2_LP_ILK, I915_READ(WM2_LP_ILK) & ~WM1_LP_SR_EN);
+	I915_WRITE(WM1_LP_ILK, I915_READ(WM1_LP_ILK) & ~WM1_LP_SR_EN);
+
+	/*
+	 * Don't touch WM1S_LP_EN here.
+	 * Doing so could cause underruns.
+	 */
+}
+
 static void ironlake_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -5052,9 +4836,8 @@ static void ironlake_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(DISP_ARB_CTL,
 		   (I915_READ(DISP_ARB_CTL) |
 		    DISP_FBC_WM_DIS));
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
+
+	ilk_init_lp_watermarks(dev);
 
 	/*
 	 * Based on the document from hardware guys the following bits
@@ -5086,6 +4869,9 @@ static void ironlake_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(CACHE_MODE_0,
 		   _MASKED_BIT_ENABLE(CM0_PIPELINED_RENDER_FLUSH_DISABLE));
 
+	/* WaDisable_RenderCache_OperationalFlush:ilk */
+	I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE));
+
 	g4x_disable_trickle_feed(dev);
 
 	ibx_init_clock_gating(dev);
@@ -5161,9 +4947,21 @@ static void gen6_init_clock_gating(struct drm_device *dev)
 		I915_WRITE(GEN6_GT_MODE,
 			   _MASKED_BIT_ENABLE(GEN6_TD_FOUR_ROW_DISPATCH_DISABLE));
 
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
+	/* WaDisable_RenderCache_OperationalFlush:snb */
+	I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE));
+
+	/*
+	 * BSpec recoomends 8x4 when MSAA is used,
+	 * however in practice 16x4 seems fastest.
+	 *
+	 * Note that PS/WM thread counts depend on the WIZ hashing
+	 * disable bit, which we don't touch here, but it's good
+	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
+	 */
+	I915_WRITE(GEN6_GT_MODE,
+		   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
+
+	ilk_init_lp_watermarks(dev);
 
 	I915_WRITE(CACHE_MODE_0,
 		   _MASKED_BIT_DISABLE(CM0_STC_EVICT_DISABLE_LRA_SNB));
@@ -5183,17 +4981,24 @@ static void gen6_init_clock_gating(struct drm_device *dev)
 	 * According to the spec, bit 11 (RCCUNIT) must also be set,
 	 * but we didn't debug actual testcases to find it out.
 	 *
-	 * Also apply WaDisableVDSUnitClockGating:snb and
-	 * WaDisableRCPBUnitClockGating:snb.
+	 * WaDisableRCCUnitClockGating:snb
+	 * WaDisableRCPBUnitClockGating:snb
 	 */
 	I915_WRITE(GEN6_UCGCTL2,
-		   GEN7_VDSUNIT_CLOCK_GATE_DISABLE |
 		   GEN6_RCPBUNIT_CLOCK_GATE_DISABLE |
 		   GEN6_RCCUNIT_CLOCK_GATE_DISABLE);
 
-	/* Bspec says we need to always set all mask bits. */
-	I915_WRITE(_3D_CHICKEN3, (0xFFFF << 16) |
-		   _3D_CHICKEN3_SF_DISABLE_FASTCLIP_CULL);
+	/* WaStripsFansDisableFastClipPerformanceFix:snb */
+	I915_WRITE(_3D_CHICKEN3,
+		   _MASKED_BIT_ENABLE(_3D_CHICKEN3_SF_DISABLE_FASTCLIP_CULL));
+
+	/*
+	 * Bspec says:
+	 * "This bit must be set if 3DSTATE_CLIP clip mode is set to normal and
+	 * 3DSTATE_SF number of SF output attributes is more than 16."
+	 */
+	I915_WRITE(_3D_CHICKEN3,
+		   _MASKED_BIT_ENABLE(_3D_CHICKEN3_SF_DISABLE_PIPELINED_ATTR_FETCH));
 
 	/*
 	 * According to the spec the following bits should be
@@ -5219,11 +5024,6 @@ static void gen6_init_clock_gating(struct drm_device *dev)
 
 	g4x_disable_trickle_feed(dev);
 
-	/* The default value should be 0x200 according to docs, but the two
-	 * platforms I checked have a 0 for this. (Maybe BIOS overrides?) */
-	I915_WRITE(GEN6_GT_MODE, _MASKED_BIT_DISABLE(0xffff));
-	I915_WRITE(GEN6_GT_MODE, _MASKED_BIT_ENABLE(GEN6_GT_MODE_HI));
-
 	cpt_init_clock_gating(dev);
 
 	gen6_check_mch_setup(dev);
@@ -5233,14 +5033,17 @@ static void gen7_setup_fixed_func_scheduler(struct drm_i915_private *dev_priv)
 {
 	uint32_t reg = I915_READ(GEN7_FF_THREAD_MODE);
 
+	/*
+	 * WaVSThreadDispatchOverride:ivb,vlv
+	 *
+	 * This actually overrides the dispatch
+	 * mode for all thread types.
+	 */
 	reg &= ~GEN7_FF_SCHED_MASK;
 	reg |= GEN7_FF_TS_SCHED_HW;
 	reg |= GEN7_FF_VS_SCHED_HW;
 	reg |= GEN7_FF_DS_SCHED_HW;
 
-	if (IS_HASWELL(dev_priv->dev))
-		reg &= ~GEN7_FF_VS_REF_CNT_FFME;
-
 	I915_WRITE(GEN7_FF_THREAD_MODE, reg);
 }
 
@@ -5278,7 +5081,7 @@ static void lpt_suspend_hw(struct drm_device *dev)
 static void gen8_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	enum pipe i;
+	enum pipe pipe;
 
 	I915_WRITE(WM3_LP_ILK, 0);
 	I915_WRITE(WM2_LP_ILK, 0);
@@ -5287,8 +5090,19 @@ static void gen8_init_clock_gating(struct drm_device *dev)
 	/* FIXME(BDW): Check all the w/a, some might only apply to
 	 * pre-production hw. */
 
-	WARN(!i915_preliminary_hw_support,
-	     "GEN8_CENTROID_PIXEL_OPT_DIS not be needed for production\n");
+	/* WaDisablePartialInstShootdown:bdw */
+	I915_WRITE(GEN8_ROW_CHICKEN,
+		   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE));
+
+	/* WaDisableThreadStallDopClockGating:bdw */
+	/* FIXME: Unclear whether we really need this on production bdw. */
+	I915_WRITE(GEN8_ROW_CHICKEN,
+		   _MASKED_BIT_ENABLE(STALL_DOP_GATING_DISABLE));
+
+	/*
+	 * This GEN8_CENTROID_PIXEL_OPT_DIS W/A is only needed for
+	 * pre-production hardware
+	 */
 	I915_WRITE(HALF_SLICE_CHICKEN3,
 		   _MASKED_BIT_ENABLE(GEN8_CENTROID_PIXEL_OPT_DIS));
 	I915_WRITE(HALF_SLICE_CHICKEN3,
@@ -5304,43 +5118,66 @@ static void gen8_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(GEN7_HALF_SLICE_CHICKEN1,
 		   _MASKED_BIT_ENABLE(GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE));
 
-	/* WaSwitchSolVfFArbitrationPriority */
+	/* WaDisableDopClockGating:bdw May not be needed for production */
+	I915_WRITE(GEN7_ROW_CHICKEN2,
+		   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
+
+	/* WaSwitchSolVfFArbitrationPriority:bdw */
 	I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL);
 
-	/* WaPsrDPAMaskVBlankInSRD */
+	/* WaPsrDPAMaskVBlankInSRD:bdw */
 	I915_WRITE(CHICKEN_PAR1_1,
 		   I915_READ(CHICKEN_PAR1_1) | DPA_MASK_VBLANK_SRD);
 
-	/* WaPsrDPRSUnmaskVBlankInSRD */
-	for_each_pipe(i) {
-		I915_WRITE(CHICKEN_PIPESL_1(i),
-			   I915_READ(CHICKEN_PIPESL_1(i) |
-				     DPRS_MASK_VBLANK_SRD));
+	/* WaPsrDPRSUnmaskVBlankInSRD:bdw */
+	for_each_pipe(pipe) {
+		I915_WRITE(CHICKEN_PIPESL_1(pipe),
+			   I915_READ(CHICKEN_PIPESL_1(pipe)) |
+			   BDW_DPRS_MASK_VBLANK_SRD);
 	}
-}
 
-static void haswell_init_clock_gating(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	/* Use Force Non-Coherent whenever executing a 3D context. This is a
+	 * workaround for for a possible hang in the unlikely event a TLB
+	 * invalidation occurs during a PSD flush.
+	 */
+	I915_WRITE(HDC_CHICKEN0,
+		   I915_READ(HDC_CHICKEN0) |
+		   _MASKED_BIT_ENABLE(HDC_FORCE_NON_COHERENT));
 
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
+	/* WaVSRefCountFullforceMissDisable:bdw */
+	/* WaDSRefCountFullforceMissDisable:bdw */
+	I915_WRITE(GEN7_FF_THREAD_MODE,
+		   I915_READ(GEN7_FF_THREAD_MODE) &
+		   ~(GEN8_FF_DS_REF_CNT_FFME | GEN7_FF_VS_REF_CNT_FFME));
 
-	/* According to the spec, bit 13 (RCZUNIT) must be set on IVB.
-	 * This implements the WaDisableRCZUnitClockGating:hsw workaround.
+	/*
+	 * BSpec recommends 8x4 when MSAA is used,
+	 * however in practice 16x4 seems fastest.
+	 *
+	 * Note that PS/WM thread counts depend on the WIZ hashing
+	 * disable bit, which we don't touch here, but it's good
+	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
 	 */
-	I915_WRITE(GEN6_UCGCTL2, GEN6_RCZUNIT_CLOCK_GATE_DISABLE);
+	I915_WRITE(GEN7_GT_MODE,
+		   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
 
-	/* Apply the WaDisableRHWOOptimizationForRenderHang:hsw workaround. */
-	I915_WRITE(GEN7_COMMON_SLICE_CHICKEN1,
-		   GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC);
+	I915_WRITE(GEN6_RC_SLEEP_PSMI_CONTROL,
+		   _MASKED_BIT_ENABLE(GEN8_RC_SEMA_IDLE_MSG_DISABLE));
 
-	/* WaApplyL3ControlAndL3ChickenMode:hsw */
-	I915_WRITE(GEN7_L3CNTLREG1,
-			GEN7_WA_FOR_GEN7_L3_CONTROL);
-	I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER,
-			GEN7_WA_L3_CHICKEN_MODE);
+	/* WaDisableSDEUnitClockGating:bdw */
+	I915_WRITE(GEN8_UCGCTL6, I915_READ(GEN8_UCGCTL6) |
+		   GEN8_SDEUNIT_CLOCK_GATE_DISABLE);
+
+	/* Wa4x4STCOptimizationDisable:bdw */
+	I915_WRITE(CACHE_MODE_1,
+		   _MASKED_BIT_ENABLE(GEN8_4x4_STC_OPTIMIZATION_DISABLE));
+}
+
+static void haswell_init_clock_gating(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	ilk_init_lp_watermarks(dev);
 
 	/* L3 caching of data atomics doesn't work -- disable it. */
 	I915_WRITE(HSW_SCRATCH1, HSW_SCRATCH1_L3_DATA_ATOMICS_DISABLE);
@@ -5353,12 +5190,31 @@ static void haswell_init_clock_gating(struct drm_device *dev)
 			GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB);
 
 	/* WaVSRefCountFullforceMissDisable:hsw */
-	gen7_setup_fixed_func_scheduler(dev_priv);
+	I915_WRITE(GEN7_FF_THREAD_MODE,
+		   I915_READ(GEN7_FF_THREAD_MODE) & ~GEN7_FF_VS_REF_CNT_FFME);
+
+	/* WaDisable_RenderCache_OperationalFlush:hsw */
+	I915_WRITE(CACHE_MODE_0_GEN7, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE));
+
+	/* enable HiZ Raw Stall Optimization */
+	I915_WRITE(CACHE_MODE_0_GEN7,
+		   _MASKED_BIT_DISABLE(HIZ_RAW_STALL_OPT_DISABLE));
 
 	/* WaDisable4x2SubspanOptimization:hsw */
 	I915_WRITE(CACHE_MODE_1,
 		   _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE));
 
+	/*
+	 * BSpec recommends 8x4 when MSAA is used,
+	 * however in practice 16x4 seems fastest.
+	 *
+	 * Note that PS/WM thread counts depend on the WIZ hashing
+	 * disable bit, which we don't touch here, but it's good
+	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
+	 */
+	I915_WRITE(GEN7_GT_MODE,
+		   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
+
 	/* WaSwitchSolVfFArbitrationPriority:hsw */
 	I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL);
 
@@ -5374,9 +5230,7 @@ static void ivybridge_init_clock_gating(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t snpcr;
 
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
+	ilk_init_lp_watermarks(dev);
 
 	I915_WRITE(ILK_DSPCLK_GATE_D, ILK_VRHUNIT_CLOCK_GATE_DISABLE);
 
@@ -5393,9 +5247,9 @@ static void ivybridge_init_clock_gating(struct drm_device *dev)
 	if (IS_IVB_GT1(dev))
 		I915_WRITE(GEN7_HALF_SLICE_CHICKEN1,
 			   _MASKED_BIT_ENABLE(GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE));
-	else
-		I915_WRITE(GEN7_HALF_SLICE_CHICKEN1_GT2,
-			   _MASKED_BIT_ENABLE(GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE));
+
+	/* WaDisable_RenderCache_OperationalFlush:ivb */
+	I915_WRITE(CACHE_MODE_0_GEN7, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE));
 
 	/* Apply the WaDisableRHWOOptimizationForRenderHang:ivb workaround. */
 	I915_WRITE(GEN7_COMMON_SLICE_CHICKEN1,
@@ -5409,31 +5263,24 @@ static void ivybridge_init_clock_gating(struct drm_device *dev)
 	if (IS_IVB_GT1(dev))
 		I915_WRITE(GEN7_ROW_CHICKEN2,
 			   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
-	else
+	else {
+		/* must write both registers */
+		I915_WRITE(GEN7_ROW_CHICKEN2,
+			   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
 		I915_WRITE(GEN7_ROW_CHICKEN2_GT2,
 			   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
-
+	}
 
 	/* WaForceL3Serialization:ivb */
 	I915_WRITE(GEN7_L3SQCREG4, I915_READ(GEN7_L3SQCREG4) &
 		   ~L3SQ_URB_READ_CAM_MATCH_DISABLE);
 
-	/* According to the BSpec vol1g, bit 12 (RCPBUNIT) clock
-	 * gating disable must be set.  Failure to set it results in
-	 * flickering pixels due to Z write ordering failures after
-	 * some amount of runtime in the Mesa "fire" demo, and Unigine
-	 * Sanctuary and Tropics, and apparently anything else with
-	 * alpha test or pixel discard.
-	 *
-	 * According to the spec, bit 11 (RCCUNIT) must also be set,
-	 * but we didn't debug actual testcases to find it out.
-	 *
+	/*
 	 * According to the spec, bit 13 (RCZUNIT) must be set on IVB.
 	 * This implements the WaDisableRCZUnitClockGating:ivb workaround.
 	 */
 	I915_WRITE(GEN6_UCGCTL2,
-		   GEN6_RCZUNIT_CLOCK_GATE_DISABLE |
-		   GEN6_RCCUNIT_CLOCK_GATE_DISABLE);
+		   GEN6_RCZUNIT_CLOCK_GATE_DISABLE);
 
 	/* This is required by WaCatErrorRejectionIssue:ivb */
 	I915_WRITE(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG,
@@ -5442,13 +5289,29 @@ static void ivybridge_init_clock_gating(struct drm_device *dev)
 
 	g4x_disable_trickle_feed(dev);
 
-	/* WaVSRefCountFullforceMissDisable:ivb */
 	gen7_setup_fixed_func_scheduler(dev_priv);
 
+	if (0) { /* causes HiZ corruption on ivb:gt1 */
+		/* enable HiZ Raw Stall Optimization */
+		I915_WRITE(CACHE_MODE_0_GEN7,
+			   _MASKED_BIT_DISABLE(HIZ_RAW_STALL_OPT_DISABLE));
+	}
+
 	/* WaDisable4x2SubspanOptimization:ivb */
 	I915_WRITE(CACHE_MODE_1,
 		   _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE));
 
+	/*
+	 * BSpec recommends 8x4 when MSAA is used,
+	 * however in practice 16x4 seems fastest.
+	 *
+	 * Note that PS/WM thread counts depend on the WIZ hashing
+	 * disable bit, which we don't touch here, but it's good
+	 * to keep in mind (see 3DSTATE_PS and 3DSTATE_WM).
+	 */
+	I915_WRITE(GEN7_GT_MODE,
+		   GEN6_WIZ_HASHING_MASK | GEN6_WIZ_HASHING_16x4);
+
 	snpcr = I915_READ(GEN6_MBCUNIT_SNPCR);
 	snpcr &= ~GEN6_MBC_SNPCR_MASK;
 	snpcr |= GEN6_MBC_SNPCR_MED;
@@ -5463,6 +5326,28 @@ static void ivybridge_init_clock_gating(struct drm_device *dev)
 static void valleyview_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val;
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+	val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
+	mutex_unlock(&dev_priv->rps.hw_lock);
+	switch ((val >> 6) & 3) {
+	case 0:
+	case 1:
+		dev_priv->mem_freq = 800;
+		break;
+	case 2:
+		dev_priv->mem_freq = 1066;
+		break;
+	case 3:
+		dev_priv->mem_freq = 1333;
+		break;
+	}
+	DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq);
+
+	dev_priv->vlv_cdclk_freq = valleyview_cur_cdclk(dev_priv);
+	DRM_DEBUG_DRIVER("Current CD clock rate: %d MHz",
+			 dev_priv->vlv_cdclk_freq);
 
 	I915_WRITE(DSPCLK_GATE_D, VRHUNIT_CLOCK_GATE_DISABLE);
 
@@ -5475,18 +5360,14 @@ static void valleyview_init_clock_gating(struct drm_device *dev)
 		   CHICKEN3_DGMG_REQ_OUT_FIX_DISABLE |
 		   CHICKEN3_DGMG_DONE_FIX_DISABLE);
 
+	/* WaPsdDispatchEnable:vlv */
 	/* WaDisablePSDDualDispatchEnable:vlv */
 	I915_WRITE(GEN7_HALF_SLICE_CHICKEN1,
 		   _MASKED_BIT_ENABLE(GEN7_MAX_PS_THREAD_DEP |
 				      GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE));
 
-	/* Apply the WaDisableRHWOOptimizationForRenderHang:vlv workaround. */
-	I915_WRITE(GEN7_COMMON_SLICE_CHICKEN1,
-		   GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC);
-
-	/* WaApplyL3ControlAndL3ChickenMode:vlv */
-	I915_WRITE(GEN7_L3CNTLREG1, I915_READ(GEN7_L3CNTLREG1) | GEN7_L3AGDIS);
-	I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER, GEN7_WA_L3_CHICKEN_MODE);
+	/* WaDisable_RenderCache_OperationalFlush:vlv */
+	I915_WRITE(CACHE_MODE_0_GEN7, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE));
 
 	/* WaForceL3Serialization:vlv */
 	I915_WRITE(GEN7_L3SQCREG4, I915_READ(GEN7_L3SQCREG4) &
@@ -5501,51 +5382,95 @@ static void valleyview_init_clock_gating(struct drm_device *dev)
 		   I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) |
 		   GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB);
 
-	/* According to the BSpec vol1g, bit 12 (RCPBUNIT) clock
-	 * gating disable must be set.  Failure to set it results in
-	 * flickering pixels due to Z write ordering failures after
-	 * some amount of runtime in the Mesa "fire" demo, and Unigine
-	 * Sanctuary and Tropics, and apparently anything else with
-	 * alpha test or pixel discard.
-	 *
-	 * According to the spec, bit 11 (RCCUNIT) must also be set,
-	 * but we didn't debug actual testcases to find it out.
-	 *
+	gen7_setup_fixed_func_scheduler(dev_priv);
+
+	/*
 	 * According to the spec, bit 13 (RCZUNIT) must be set on IVB.
 	 * This implements the WaDisableRCZUnitClockGating:vlv workaround.
-	 *
-	 * Also apply WaDisableVDSUnitClockGating:vlv and
-	 * WaDisableRCPBUnitClockGating:vlv.
 	 */
 	I915_WRITE(GEN6_UCGCTL2,
-		   GEN7_VDSUNIT_CLOCK_GATE_DISABLE |
-		   GEN7_TDLUNIT_CLOCK_GATE_DISABLE |
-		   GEN6_RCZUNIT_CLOCK_GATE_DISABLE |
-		   GEN6_RCPBUNIT_CLOCK_GATE_DISABLE |
-		   GEN6_RCCUNIT_CLOCK_GATE_DISABLE);
+		   GEN6_RCZUNIT_CLOCK_GATE_DISABLE);
 
-	I915_WRITE(GEN7_UCGCTL4, GEN7_L3BANK2X_CLOCK_GATE_DISABLE);
+	/* WaDisableL3Bank2xClockGate:vlv
+	 * Disabling L3 clock gating- MMIO 940c[25] = 1
+	 * Set bit 25, to disable L3_BANK_2x_CLK_GATING */
+	I915_WRITE(GEN7_UCGCTL4,
+		   I915_READ(GEN7_UCGCTL4) | GEN7_L3BANK2X_CLOCK_GATE_DISABLE);
 
 	I915_WRITE(MI_ARB_VLV, MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE);
 
+	/*
+	 * BSpec says this must be set, even though
+	 * WaDisable4x2SubspanOptimization isn't listed for VLV.
+	 */
 	I915_WRITE(CACHE_MODE_1,
 		   _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE));
 
 	/*
+	 * WaIncreaseL3CreditsForVLVB0:vlv
+	 * This is the hardware default actually.
+	 */
+	I915_WRITE(GEN7_L3SQCREG1, VLV_B0_WA_L3SQCREG1_VALUE);
+
+	/*
 	 * WaDisableVLVClockGating_VBIIssue:vlv
 	 * Disable clock gating on th GCFG unit to prevent a delay
 	 * in the reporting of vblank events.
 	 */
-	I915_WRITE(VLV_GUNIT_CLOCK_GATE, 0xffffffff);
+	I915_WRITE(VLV_GUNIT_CLOCK_GATE, GCFG_DIS);
+}
+
+static void cherryview_init_clock_gating(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	I915_WRITE(DSPCLK_GATE_D, VRHUNIT_CLOCK_GATE_DISABLE);
+
+	I915_WRITE(MI_ARB_VLV, MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE);
 
-	/* Conservative clock gating settings for now */
-	I915_WRITE(0x9400, 0xffffffff);
-	I915_WRITE(0x9404, 0xffffffff);
-	I915_WRITE(0x9408, 0xffffffff);
-	I915_WRITE(0x940c, 0xffffffff);
-	I915_WRITE(0x9410, 0xffffffff);
-	I915_WRITE(0x9414, 0xffffffff);
-	I915_WRITE(0x9418, 0xffffffff);
+	/* WaDisablePartialInstShootdown:chv */
+	I915_WRITE(GEN8_ROW_CHICKEN,
+		   _MASKED_BIT_ENABLE(PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE));
+
+	/* WaDisableThreadStallDopClockGating:chv */
+	I915_WRITE(GEN8_ROW_CHICKEN,
+		   _MASKED_BIT_ENABLE(STALL_DOP_GATING_DISABLE));
+
+	/* WaVSRefCountFullforceMissDisable:chv */
+	/* WaDSRefCountFullforceMissDisable:chv */
+	I915_WRITE(GEN7_FF_THREAD_MODE,
+		   I915_READ(GEN7_FF_THREAD_MODE) &
+		   ~(GEN8_FF_DS_REF_CNT_FFME | GEN7_FF_VS_REF_CNT_FFME));
+
+	/* WaDisableSemaphoreAndSyncFlipWait:chv */
+	I915_WRITE(GEN6_RC_SLEEP_PSMI_CONTROL,
+		   _MASKED_BIT_ENABLE(GEN8_RC_SEMA_IDLE_MSG_DISABLE));
+
+	/* WaDisableCSUnitClockGating:chv */
+	I915_WRITE(GEN6_UCGCTL1, I915_READ(GEN6_UCGCTL1) |
+		   GEN6_CSUNIT_CLOCK_GATE_DISABLE);
+
+	/* WaDisableSDEUnitClockGating:chv */
+	I915_WRITE(GEN8_UCGCTL6, I915_READ(GEN8_UCGCTL6) |
+		   GEN8_SDEUNIT_CLOCK_GATE_DISABLE);
+
+	/* WaDisableSamplerPowerBypass:chv (pre-production hw) */
+	I915_WRITE(HALF_SLICE_CHICKEN3,
+		   _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS));
+
+	/* WaDisableGunitClockGating:chv (pre-production hw) */
+	I915_WRITE(VLV_GUNIT_CLOCK_GATE, I915_READ(VLV_GUNIT_CLOCK_GATE) |
+		   GINT_DIS);
+
+	/* WaDisableFfDopClockGating:chv (pre-production hw) */
+	I915_WRITE(GEN6_RC_SLEEP_PSMI_CONTROL,
+		   _MASKED_BIT_ENABLE(GEN8_FF_DOP_CLOCK_GATE_DISABLE));
+
+	/* WaDisableDopClockGating:chv (pre-production hw) */
+	I915_WRITE(GEN7_ROW_CHICKEN2,
+		   _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
+	I915_WRITE(GEN6_UCGCTL1, I915_READ(GEN6_UCGCTL1) |
+		   GEN6_EU_TCUNIT_CLOCK_GATE_DISABLE);
 }
 
 static void g4x_init_clock_gating(struct drm_device *dev)
@@ -5569,6 +5494,9 @@ static void g4x_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(CACHE_MODE_0,
 		   _MASKED_BIT_ENABLE(CM0_PIPELINED_RENDER_FLUSH_DISABLE));
 
+	/* WaDisable_RenderCache_OperationalFlush:g4x */
+	I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE));
+
 	g4x_disable_trickle_feed(dev);
 }
 
@@ -5583,6 +5511,9 @@ static void crestline_init_clock_gating(struct drm_device *dev)
 	I915_WRITE16(DEUC, 0);
 	I915_WRITE(MI_ARB_STATE,
 		   _MASKED_BIT_ENABLE(MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE));
+
+	/* WaDisable_RenderCache_OperationalFlush:gen4 */
+	I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE));
 }
 
 static void broadwater_init_clock_gating(struct drm_device *dev)
@@ -5597,6 +5528,9 @@ static void broadwater_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(RENCLK_GATE_D2, 0);
 	I915_WRITE(MI_ARB_STATE,
 		   _MASKED_BIT_ENABLE(MI_ARB_DISPLAY_TRICKLE_FEED_DISABLE));
+
+	/* WaDisable_RenderCache_OperationalFlush:gen4 */
+	I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE));
 }
 
 static void gen3_init_clock_gating(struct drm_device *dev)
@@ -5613,6 +5547,12 @@ static void gen3_init_clock_gating(struct drm_device *dev)
 
 	/* IIR "flip pending" means done if this bit is set */
 	I915_WRITE(ECOSKPD, _MASKED_BIT_DISABLE(ECO_FLIP_DONE));
+
+	/* interrupts should cause a wake up from C3 */
+	I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_AGPBUSY_INT_EN));
+
+	/* On GEN3 we really need to make sure the ARB C3 LP bit is set */
+	I915_WRITE(MI_ARB_STATE, _MASKED_BIT_ENABLE(MI_ARB_C3_LP_WRITE_ENABLE));
 }
 
 static void i85x_init_clock_gating(struct drm_device *dev)
@@ -5620,6 +5560,10 @@ static void i85x_init_clock_gating(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	I915_WRITE(RENCLK_GATE_D1, SV_CLOCK_GATE_DISABLE);
+
+	/* interrupts should cause a wake up from C3 */
+	I915_WRITE(MI_STATE, _MASKED_BIT_ENABLE(MI_AGPBUSY_INT_EN) |
+		   _MASKED_BIT_DISABLE(MI_AGPBUSY_830_MODE));
 }
 
 static void i830_init_clock_gating(struct drm_device *dev)
@@ -5642,53 +5586,121 @@ void intel_suspend_hw(struct drm_device *dev)
 		lpt_suspend_hw(dev);
 }
 
-static bool is_always_on_power_domain(struct drm_device *dev,
-				      enum intel_display_power_domain domain)
-{
-	unsigned long always_on_domains;
+#define for_each_power_well(i, power_well, domain_mask, power_domains)	\
+	for (i = 0;							\
+	     i < (power_domains)->power_well_count &&			\
+		 ((power_well) = &(power_domains)->power_wells[i]);	\
+	     i++)							\
+		if ((power_well)->domains & (domain_mask))
 
-	BUG_ON(BIT(domain) & ~POWER_DOMAIN_MASK);
-
-	if (IS_BROADWELL(dev)) {
-		always_on_domains = BDW_ALWAYS_ON_POWER_DOMAINS;
-	} else if (IS_HASWELL(dev)) {
-		always_on_domains = HSW_ALWAYS_ON_POWER_DOMAINS;
-	} else {
-		WARN_ON(1);
-		return true;
-	}
-
-	return BIT(domain) & always_on_domains;
-}
+#define for_each_power_well_rev(i, power_well, domain_mask, power_domains) \
+	for (i = (power_domains)->power_well_count - 1;			 \
+	     i >= 0 && ((power_well) = &(power_domains)->power_wells[i]);\
+	     i--)							 \
+		if ((power_well)->domains & (domain_mask))
 
 /**
  * We should only use the power well if we explicitly asked the hardware to
  * enable it, so check if it's enabled and also check if we've requested it to
  * be enabled.
  */
-bool intel_display_power_enabled(struct drm_device *dev,
+static bool hsw_power_well_enabled(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	return I915_READ(HSW_PWR_WELL_DRIVER) ==
+		     (HSW_PWR_WELL_ENABLE_REQUEST | HSW_PWR_WELL_STATE_ENABLED);
+}
+
+bool intel_display_power_enabled_unlocked(struct drm_i915_private *dev_priv,
+					  enum intel_display_power_domain domain)
+{
+	struct i915_power_domains *power_domains;
+	struct i915_power_well *power_well;
+	bool is_enabled;
+	int i;
+
+	if (dev_priv->pm.suspended)
+		return false;
+
+	power_domains = &dev_priv->power_domains;
+
+	is_enabled = true;
+
+	for_each_power_well_rev(i, power_well, BIT(domain), power_domains) {
+		if (power_well->always_on)
+			continue;
+
+		if (!power_well->hw_enabled) {
+			is_enabled = false;
+			break;
+		}
+	}
+
+	return is_enabled;
+}
+
+bool intel_display_power_enabled(struct drm_i915_private *dev_priv,
 				 enum intel_display_power_domain domain)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_power_domains *power_domains;
+	bool ret;
 
-	if (!HAS_POWER_WELL(dev))
-		return true;
+	power_domains = &dev_priv->power_domains;
 
-	if (is_always_on_power_domain(dev, domain))
-		return true;
+	mutex_lock(&power_domains->lock);
+	ret = intel_display_power_enabled_unlocked(dev_priv, domain);
+	mutex_unlock(&power_domains->lock);
 
-	return I915_READ(HSW_PWR_WELL_DRIVER) ==
-		     (HSW_PWR_WELL_ENABLE_REQUEST | HSW_PWR_WELL_STATE_ENABLED);
+	return ret;
 }
 
-static void __intel_set_power_well(struct drm_device *dev, bool enable)
+/*
+ * Starting with Haswell, we have a "Power Down Well" that can be turned off
+ * when not needed anymore. We have 4 registers that can request the power well
+ * to be enabled, and it will only be disabled if none of the registers is
+ * requesting it to be enabled.
+ */
+static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	bool is_enabled, enable_requested;
+	struct drm_device *dev = dev_priv->dev;
 	unsigned long irqflags;
-	uint32_t tmp;
 
-	WARN_ON(dev_priv->pc8.enabled);
+	/*
+	 * After we re-enable the power well, if we touch VGA register 0x3d5
+	 * we'll get unclaimed register interrupts. This stops after we write
+	 * anything to the VGA MSR register. The vgacon module uses this
+	 * register all the time, so if we unbind our driver and, as a
+	 * consequence, bind vgacon, we'll get stuck in an infinite loop at
+	 * console_unlock(). So make here we touch the VGA MSR register, making
+	 * sure vgacon can keep working normally without triggering interrupts
+	 * and error messages.
+	 */
+	vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO);
+	outb(inb(VGA_MSR_READ), VGA_MSR_WRITE);
+	vga_put(dev->pdev, VGA_RSRC_LEGACY_IO);
+
+	if (IS_BROADWELL(dev)) {
+		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_B),
+			   dev_priv->de_irq_mask[PIPE_B]);
+		I915_WRITE(GEN8_DE_PIPE_IER(PIPE_B),
+			   ~dev_priv->de_irq_mask[PIPE_B] |
+			   GEN8_PIPE_VBLANK);
+		I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_C),
+			   dev_priv->de_irq_mask[PIPE_C]);
+		I915_WRITE(GEN8_DE_PIPE_IER(PIPE_C),
+			   ~dev_priv->de_irq_mask[PIPE_C] |
+			   GEN8_PIPE_VBLANK);
+		POSTING_READ(GEN8_DE_PIPE_IER(PIPE_C));
+		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	}
+}
+
+static void hsw_set_power_well(struct drm_i915_private *dev_priv,
+			       struct i915_power_well *power_well, bool enable)
+{
+	bool is_enabled, enable_requested;
+	uint32_t tmp;
 
 	tmp = I915_READ(HSW_PWR_WELL_DRIVER);
 	is_enabled = tmp & HSW_PWR_WELL_STATE_ENABLED;
@@ -5706,212 +5718,679 @@ static void __intel_set_power_well(struct drm_device *dev, bool enable)
 				DRM_ERROR("Timeout enabling power well\n");
 		}
 
-		if (IS_BROADWELL(dev)) {
-			spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-			I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_B),
-				   dev_priv->de_irq_mask[PIPE_B]);
-			I915_WRITE(GEN8_DE_PIPE_IER(PIPE_B),
-				   ~dev_priv->de_irq_mask[PIPE_B] |
-				   GEN8_PIPE_VBLANK);
-			I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_C),
-				   dev_priv->de_irq_mask[PIPE_C]);
-			I915_WRITE(GEN8_DE_PIPE_IER(PIPE_C),
-				   ~dev_priv->de_irq_mask[PIPE_C] |
-				   GEN8_PIPE_VBLANK);
-			POSTING_READ(GEN8_DE_PIPE_IER(PIPE_C));
-			spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
-		}
+		hsw_power_well_post_enable(dev_priv);
 	} else {
 		if (enable_requested) {
-			enum pipe p;
-
 			I915_WRITE(HSW_PWR_WELL_DRIVER, 0);
 			POSTING_READ(HSW_PWR_WELL_DRIVER);
 			DRM_DEBUG_KMS("Requesting to disable the power well\n");
+		}
+	}
+}
 
+static void hsw_power_well_sync_hw(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	hsw_set_power_well(dev_priv, power_well, power_well->count > 0);
+
+	/*
+	 * We're taking over the BIOS, so clear any requests made by it since
+	 * the driver is in charge now.
+	 */
+	if (I915_READ(HSW_PWR_WELL_BIOS) & HSW_PWR_WELL_ENABLE_REQUEST)
+		I915_WRITE(HSW_PWR_WELL_BIOS, 0);
+}
+
+static void hsw_power_well_enable(struct drm_i915_private *dev_priv,
+				  struct i915_power_well *power_well)
+{
+	hsw_set_power_well(dev_priv, power_well, true);
+}
+
+static void hsw_power_well_disable(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	hsw_set_power_well(dev_priv, power_well, false);
+}
+
+static void i9xx_always_on_power_well_noop(struct drm_i915_private *dev_priv,
+					   struct i915_power_well *power_well)
+{
+}
+
+static bool i9xx_always_on_power_well_enabled(struct drm_i915_private *dev_priv,
+					     struct i915_power_well *power_well)
+{
+	return true;
+}
+
+void __vlv_set_power_well(struct drm_i915_private *dev_priv,
+			  enum punit_power_well power_well_id, bool enable)
+{
+	struct drm_device *dev = dev_priv->dev;
+	u32 mask;
+	u32 state;
+	u32 ctrl;
+	enum pipe pipe;
+
+	if (power_well_id == PUNIT_POWER_WELL_DPIO_CMN_BC) {
+		if (enable) {
 			/*
-			 * After this, the registers on the pipes that are part
-			 * of the power well will become zero, so we have to
-			 * adjust our counters according to that.
-			 *
-			 * FIXME: Should we do this in general in
-			 * drm_vblank_post_modeset?
+			 * Enable the CRI clock source so we can get at the
+			 * display and the reference clock for VGA
+			 * hotplug / manual detection.
 			 */
-			spin_lock_irqsave(&dev->vbl_lock, irqflags);
-			for_each_pipe(p)
-				if (p != PIPE_A)
-					dev->vblank[p].last = 0;
-			spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
+			I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
+				   DPLL_REFA_CLK_ENABLE_VLV |
+				   DPLL_INTEGRATED_CRI_CLK_VLV);
+			udelay(1); /* >10ns for cmnreset, >0ns for sidereset */
+		} else {
+			for_each_pipe(pipe)
+				assert_pll_disabled(dev_priv, pipe);
+			/* Assert common reset */
+			I915_WRITE(DPIO_CTL, I915_READ(DPIO_CTL) &
+				   ~DPIO_CMNRST);
 		}
 	}
+
+	mask = PUNIT_PWRGT_MASK(power_well_id);
+	state = enable ? PUNIT_PWRGT_PWR_ON(power_well_id) :
+			 PUNIT_PWRGT_PWR_GATE(power_well_id);
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+
+#define COND \
+	((vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_STATUS) & mask) == state)
+
+	if (COND)
+		goto out;
+
+	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_CTRL);
+	ctrl &= ~mask;
+	ctrl |= state;
+	vlv_punit_write(dev_priv, PUNIT_REG_PWRGT_CTRL, ctrl);
+
+	if (wait_for(COND, 100))
+		DRM_ERROR("timout setting power well state %08x (%08x)\n",
+			  state,
+			  vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_CTRL));
+
+#undef COND
+
+out:
+	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	/*
+	 * From VLV2A0_DP_eDP_DPIO_driver_vbios_notes_10.docx -
+	 *  6.	De-assert cmn_reset/side_reset. Same as VLV X0.
+	 *   a.	GUnit 0x2110 bit[0] set to 1 (def 0)
+	 *   b.	The other bits such as sfr settings / modesel may all
+	 *	be set to 0.
+	 *
+	 * This should only be done on init and resume from S3 with
+	 * both PLLs disabled, or we risk losing DPIO and PLL
+	 * synchronization.
+	 */
+	if (power_well_id == PUNIT_POWER_WELL_DPIO_CMN_BC && enable)
+		I915_WRITE(DPIO_CTL, I915_READ(DPIO_CTL) | DPIO_CMNRST);
 }
 
-static void __intel_power_well_get(struct drm_device *dev,
+static void vlv_set_power_well(struct drm_i915_private *dev_priv,
+			       struct i915_power_well *power_well, bool enable)
+{
+	enum punit_power_well power_well_id = power_well->data;
+
+	__vlv_set_power_well(dev_priv, power_well_id, enable);
+}
+
+static void vlv_power_well_sync_hw(struct drm_i915_private *dev_priv,
 				   struct i915_power_well *power_well)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	vlv_set_power_well(dev_priv, power_well, power_well->count > 0);
+}
 
-	if (!power_well->count++) {
-		hsw_disable_package_c8(dev_priv);
-		__intel_set_power_well(dev, true);
-	}
+static void vlv_power_well_enable(struct drm_i915_private *dev_priv,
+				  struct i915_power_well *power_well)
+{
+	vlv_set_power_well(dev_priv, power_well, true);
 }
 
-static void __intel_power_well_put(struct drm_device *dev,
+static void vlv_power_well_disable(struct drm_i915_private *dev_priv,
 				   struct i915_power_well *power_well)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	vlv_set_power_well(dev_priv, power_well, false);
+}
 
-	WARN_ON(!power_well->count);
-	if (!--power_well->count && i915_disable_power_well) {
-		__intel_set_power_well(dev, false);
-		hsw_enable_package_c8(dev_priv);
-	}
+static bool vlv_power_well_enabled(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	int power_well_id = power_well->data;
+	bool enabled = false;
+	u32 mask;
+	u32 state;
+	u32 ctrl;
+
+	mask = PUNIT_PWRGT_MASK(power_well_id);
+	ctrl = PUNIT_PWRGT_PWR_ON(power_well_id);
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+
+	state = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_STATUS) & mask;
+	/*
+	 * We only ever set the power-on and power-gate states, anything
+	 * else is unexpected.
+	 */
+	WARN_ON(state != PUNIT_PWRGT_PWR_ON(power_well_id) &&
+		state != PUNIT_PWRGT_PWR_GATE(power_well_id));
+	if (state == ctrl)
+		enabled = true;
+
+	/*
+	 * A transient state at this point would mean some unexpected party
+	 * is poking at the power controls too.
+	 */
+	ctrl = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_CTRL) & mask;
+	WARN_ON(ctrl != state);
+
+	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	return enabled;
 }
 
-void intel_display_power_get(struct drm_device *dev,
-			     enum intel_display_power_domain domain)
+static void vlv_display_power_well_enable(struct drm_i915_private *dev_priv,
+					  struct i915_power_well *power_well)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct i915_power_domains *power_domains;
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DISP2D);
+
+	vlv_set_power_well(dev_priv, power_well, true);
+
+	spin_lock_irq(&dev_priv->irq_lock);
+	valleyview_enable_display_irqs(dev_priv);
+	spin_unlock_irq(&dev_priv->irq_lock);
 
-	if (!HAS_POWER_WELL(dev))
+	/*
+	 * During driver initialization/resume we can avoid restoring the
+	 * part of the HW/SW state that will be inited anyway explicitly.
+	 */
+	if (dev_priv->power_domains.initializing)
 		return;
 
-	if (is_always_on_power_domain(dev, domain))
+	intel_hpd_init(dev_priv->dev);
+
+	i915_redisable_vga_power_on(dev_priv->dev);
+}
+
+static void vlv_display_power_well_disable(struct drm_i915_private *dev_priv,
+					   struct i915_power_well *power_well)
+{
+	WARN_ON_ONCE(power_well->data != PUNIT_POWER_WELL_DISP2D);
+
+	spin_lock_irq(&dev_priv->irq_lock);
+	valleyview_disable_display_irqs(dev_priv);
+	spin_unlock_irq(&dev_priv->irq_lock);
+
+	vlv_set_power_well(dev_priv, power_well, false);
+}
+
+static void check_power_well_state(struct drm_i915_private *dev_priv,
+				   struct i915_power_well *power_well)
+{
+	bool enabled = power_well->ops->is_enabled(dev_priv, power_well);
+
+	if (power_well->always_on || !i915.disable_power_well) {
+		if (!enabled)
+			goto mismatch;
+
 		return;
+	}
+
+	if (enabled != (power_well->count > 0))
+		goto mismatch;
+
+	return;
+
+mismatch:
+	WARN(1, "state mismatch for '%s' (always_on %d hw state %d use-count %d disable_power_well %d\n",
+		  power_well->name, power_well->always_on, enabled,
+		  power_well->count, i915.disable_power_well);
+}
+
+void intel_display_power_get(struct drm_i915_private *dev_priv,
+			     enum intel_display_power_domain domain)
+{
+	struct i915_power_domains *power_domains;
+	struct i915_power_well *power_well;
+	int i;
+
+	intel_runtime_pm_get(dev_priv);
 
 	power_domains = &dev_priv->power_domains;
 
 	mutex_lock(&power_domains->lock);
-	__intel_power_well_get(dev, &power_domains->power_wells[0]);
+
+	for_each_power_well(i, power_well, BIT(domain), power_domains) {
+		if (!power_well->count++) {
+			DRM_DEBUG_KMS("enabling %s\n", power_well->name);
+			power_well->ops->enable(dev_priv, power_well);
+			power_well->hw_enabled = true;
+		}
+
+		check_power_well_state(dev_priv, power_well);
+	}
+
+	power_domains->domain_use_count[domain]++;
+
 	mutex_unlock(&power_domains->lock);
 }
 
-void intel_display_power_put(struct drm_device *dev,
+void intel_display_power_put(struct drm_i915_private *dev_priv,
 			     enum intel_display_power_domain domain)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_power_domains *power_domains;
-
-	if (!HAS_POWER_WELL(dev))
-		return;
-
-	if (is_always_on_power_domain(dev, domain))
-		return;
+	struct i915_power_well *power_well;
+	int i;
 
 	power_domains = &dev_priv->power_domains;
 
 	mutex_lock(&power_domains->lock);
-	__intel_power_well_put(dev, &power_domains->power_wells[0]);
+
+	WARN_ON(!power_domains->domain_use_count[domain]);
+	power_domains->domain_use_count[domain]--;
+
+	for_each_power_well_rev(i, power_well, BIT(domain), power_domains) {
+		WARN_ON(!power_well->count);
+
+		if (!--power_well->count && i915.disable_power_well) {
+			DRM_DEBUG_KMS("disabling %s\n", power_well->name);
+			power_well->hw_enabled = false;
+			power_well->ops->disable(dev_priv, power_well);
+		}
+
+		check_power_well_state(dev_priv, power_well);
+	}
+
 	mutex_unlock(&power_domains->lock);
+
+	intel_runtime_pm_put(dev_priv);
 }
 
 static struct i915_power_domains *hsw_pwr;
 
 /* Display audio driver power well request */
-void i915_request_power_well(void)
+int i915_request_power_well(void)
 {
 	struct drm_i915_private *dev_priv;
 
-	if (WARN_ON(!hsw_pwr))
-		return;
+	if (!hsw_pwr)
+		return -ENODEV;
 
 	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
 				power_domains);
-
-	mutex_lock(&hsw_pwr->lock);
-	__intel_power_well_get(dev_priv->dev, &hsw_pwr->power_wells[0]);
-	mutex_unlock(&hsw_pwr->lock);
+	intel_display_power_get(dev_priv, POWER_DOMAIN_AUDIO);
+	return 0;
 }
 EXPORT_SYMBOL_GPL(i915_request_power_well);
 
 /* Display audio driver power well release */
-void i915_release_power_well(void)
+int i915_release_power_well(void)
 {
 	struct drm_i915_private *dev_priv;
 
-	if (WARN_ON(!hsw_pwr))
-		return;
+	if (!hsw_pwr)
+		return -ENODEV;
 
 	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
 				power_domains);
-
-	mutex_lock(&hsw_pwr->lock);
-	__intel_power_well_put(dev_priv->dev, &hsw_pwr->power_wells[0]);
-	mutex_unlock(&hsw_pwr->lock);
+	intel_display_power_put(dev_priv, POWER_DOMAIN_AUDIO);
+	return 0;
 }
 EXPORT_SYMBOL_GPL(i915_release_power_well);
 
-int intel_power_domains_init(struct drm_device *dev)
+/*
+ * Private interface for the audio driver to get CDCLK in kHz.
+ *
+ * Caller must request power well using i915_request_power_well() prior to
+ * making the call.
+ */
+int i915_get_cdclk_freq(void)
+{
+	struct drm_i915_private *dev_priv;
+
+	if (!hsw_pwr)
+		return -ENODEV;
+
+	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
+				power_domains);
+
+	return intel_ddi_get_cdclk_freq(dev_priv);
+}
+EXPORT_SYMBOL_GPL(i915_get_cdclk_freq);
+
+
+#define POWER_DOMAIN_MASK (BIT(POWER_DOMAIN_NUM) - 1)
+
+#define HSW_ALWAYS_ON_POWER_DOMAINS (			\
+	BIT(POWER_DOMAIN_PIPE_A) |			\
+	BIT(POWER_DOMAIN_TRANSCODER_EDP) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_A_2_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_A_4_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_D_2_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_DDI_D_4_LANES) |		\
+	BIT(POWER_DOMAIN_PORT_CRT) |			\
+	BIT(POWER_DOMAIN_INIT))
+#define HSW_DISPLAY_POWER_DOMAINS (				\
+	(POWER_DOMAIN_MASK & ~HSW_ALWAYS_ON_POWER_DOMAINS) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define BDW_ALWAYS_ON_POWER_DOMAINS (			\
+	HSW_ALWAYS_ON_POWER_DOMAINS |			\
+	BIT(POWER_DOMAIN_PIPE_A_PANEL_FITTER))
+#define BDW_DISPLAY_POWER_DOMAINS (				\
+	(POWER_DOMAIN_MASK & ~BDW_ALWAYS_ON_POWER_DOMAINS) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_ALWAYS_ON_POWER_DOMAINS	BIT(POWER_DOMAIN_INIT)
+#define VLV_DISPLAY_POWER_DOMAINS	POWER_DOMAIN_MASK
+
+#define VLV_DPIO_CMN_BC_POWER_DOMAINS (		\
+	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_CRT) |		\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_B_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_2_LANES) |	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+#define VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS (	\
+	BIT(POWER_DOMAIN_PORT_DDI_C_4_LANES) |	\
+	BIT(POWER_DOMAIN_INIT))
+
+static const struct i915_power_well_ops i9xx_always_on_power_well_ops = {
+	.sync_hw = i9xx_always_on_power_well_noop,
+	.enable = i9xx_always_on_power_well_noop,
+	.disable = i9xx_always_on_power_well_noop,
+	.is_enabled = i9xx_always_on_power_well_enabled,
+};
+
+static struct i915_power_well i9xx_always_on_power_well[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = POWER_DOMAIN_MASK,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+};
+
+static const struct i915_power_well_ops hsw_power_well_ops = {
+	.sync_hw = hsw_power_well_sync_hw,
+	.enable = hsw_power_well_enable,
+	.disable = hsw_power_well_disable,
+	.is_enabled = hsw_power_well_enabled,
+};
+
+static struct i915_power_well hsw_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = HSW_ALWAYS_ON_POWER_DOMAINS,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+	{
+		.name = "display",
+		.domains = HSW_DISPLAY_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+	},
+};
+
+static struct i915_power_well bdw_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = BDW_ALWAYS_ON_POWER_DOMAINS,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+	{
+		.name = "display",
+		.domains = BDW_DISPLAY_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+	},
+};
+
+static const struct i915_power_well_ops vlv_display_power_well_ops = {
+	.sync_hw = vlv_power_well_sync_hw,
+	.enable = vlv_display_power_well_enable,
+	.disable = vlv_display_power_well_disable,
+	.is_enabled = vlv_power_well_enabled,
+};
+
+static const struct i915_power_well_ops vlv_dpio_power_well_ops = {
+	.sync_hw = vlv_power_well_sync_hw,
+	.enable = vlv_power_well_enable,
+	.disable = vlv_power_well_disable,
+	.is_enabled = vlv_power_well_enabled,
+};
+
+static struct i915_power_well vlv_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = VLV_ALWAYS_ON_POWER_DOMAINS,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+	{
+		.name = "display",
+		.domains = VLV_DISPLAY_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DISP2D,
+		.ops = &vlv_display_power_well_ops,
+	},
+	{
+		.name = "dpio-tx-b-01",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_01,
+	},
+	{
+		.name = "dpio-tx-b-23",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_B_LANES_23,
+	},
+	{
+		.name = "dpio-tx-c-01",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_01,
+	},
+	{
+		.name = "dpio-tx-c-23",
+		.domains = VLV_DPIO_TX_B_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_B_LANES_23_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_01_POWER_DOMAINS |
+			   VLV_DPIO_TX_C_LANES_23_POWER_DOMAINS,
+		.ops = &vlv_dpio_power_well_ops,
+		.data = PUNIT_POWER_WELL_DPIO_TX_C_LANES_23,
+	},
+	{
+		.name = "dpio-common",
+		.domains = VLV_DPIO_CMN_BC_POWER_DOMAINS,
+		.data = PUNIT_POWER_WELL_DPIO_CMN_BC,
+		.ops = &vlv_dpio_power_well_ops,
+	},
+};
+
+#define set_power_wells(power_domains, __power_wells) ({		\
+	(power_domains)->power_wells = (__power_wells);			\
+	(power_domains)->power_well_count = ARRAY_SIZE(__power_wells);	\
+})
+
+int intel_power_domains_init(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_power_domains *power_domains = &dev_priv->power_domains;
-	struct i915_power_well *power_well;
 
 	mutex_init(&power_domains->lock);
-	hsw_pwr = power_domains;
 
-	power_well = &power_domains->power_wells[0];
-	power_well->count = 0;
+	/*
+	 * The enabling order will be from lower to higher indexed wells,
+	 * the disabling order is reversed.
+	 */
+	if (IS_HASWELL(dev_priv->dev)) {
+		set_power_wells(power_domains, hsw_power_wells);
+		hsw_pwr = power_domains;
+	} else if (IS_BROADWELL(dev_priv->dev)) {
+		set_power_wells(power_domains, bdw_power_wells);
+		hsw_pwr = power_domains;
+	} else if (IS_VALLEYVIEW(dev_priv->dev)) {
+		set_power_wells(power_domains, vlv_power_wells);
+	} else {
+		set_power_wells(power_domains, i9xx_always_on_power_well);
+	}
 
 	return 0;
 }
 
-void intel_power_domains_remove(struct drm_device *dev)
+void intel_power_domains_remove(struct drm_i915_private *dev_priv)
 {
 	hsw_pwr = NULL;
 }
 
-static void intel_power_domains_resume(struct drm_device *dev)
+static void intel_power_domains_resume(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_power_domains *power_domains = &dev_priv->power_domains;
 	struct i915_power_well *power_well;
-
-	if (!HAS_POWER_WELL(dev))
-		return;
+	int i;
 
 	mutex_lock(&power_domains->lock);
+	for_each_power_well(i, power_well, POWER_DOMAIN_MASK, power_domains) {
+		power_well->ops->sync_hw(dev_priv, power_well);
+		power_well->hw_enabled = power_well->ops->is_enabled(dev_priv,
+								     power_well);
+	}
+	mutex_unlock(&power_domains->lock);
+}
 
-	power_well = &power_domains->power_wells[0];
-	__intel_set_power_well(dev, power_well->count > 0);
+void intel_power_domains_init_hw(struct drm_i915_private *dev_priv)
+{
+	struct i915_power_domains *power_domains = &dev_priv->power_domains;
 
-	mutex_unlock(&power_domains->lock);
+	power_domains->initializing = true;
+	/* For now, we need the power well to be always enabled. */
+	intel_display_set_init_power(dev_priv, true);
+	intel_power_domains_resume(dev_priv);
+	power_domains->initializing = false;
 }
 
-/*
- * Starting with Haswell, we have a "Power Down Well" that can be turned off
- * when not needed anymore. We have 4 registers that can request the power well
- * to be enabled, and it will only be disabled if none of the registers is
- * requesting it to be enabled.
- */
-void intel_power_domains_init_hw(struct drm_device *dev)
+void intel_aux_display_runtime_get(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	intel_runtime_pm_get(dev_priv);
+}
 
-	if (!HAS_POWER_WELL(dev))
+void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv)
+{
+	intel_runtime_pm_put(dev_priv);
+}
+
+void intel_runtime_pm_get(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
 		return;
 
-	/* For now, we need the power well to be always enabled. */
-	intel_display_set_init_power(dev, true);
-	intel_power_domains_resume(dev);
+	pm_runtime_get_sync(device);
+	WARN(dev_priv->pm.suspended, "Device still suspended.\n");
+}
 
-	/* We're taking over the BIOS, so clear any requests made by it since
-	 * the driver is in charge now. */
-	if (I915_READ(HSW_PWR_WELL_BIOS) & HSW_PWR_WELL_ENABLE_REQUEST)
-		I915_WRITE(HSW_PWR_WELL_BIOS, 0);
+void intel_runtime_pm_get_noresume(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	WARN(dev_priv->pm.suspended, "Getting nosync-ref while suspended.\n");
+	pm_runtime_get_noresume(device);
 }
 
-/* Disables PC8 so we can use the GMBUS and DP AUX interrupts. */
-void intel_aux_display_runtime_get(struct drm_i915_private *dev_priv)
+void intel_runtime_pm_put(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	pm_runtime_mark_last_busy(device);
+	pm_runtime_put_autosuspend(device);
+}
+
+void intel_init_runtime_pm(struct drm_i915_private *dev_priv)
 {
-	hsw_disable_package_c8(dev_priv);
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	pm_runtime_set_active(device);
+
+	/*
+	 * RPM depends on RC6 to save restore the GT HW context, so make RC6 a
+	 * requirement.
+	 */
+	if (!intel_enable_rc6(dev)) {
+		DRM_INFO("RC6 disabled, disabling runtime PM support\n");
+		return;
+	}
+
+	pm_runtime_set_autosuspend_delay(device, 10000); /* 10s */
+	pm_runtime_mark_last_busy(device);
+	pm_runtime_use_autosuspend(device);
+
+	pm_runtime_put_autosuspend(device);
 }
 
-void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv)
+void intel_fini_runtime_pm(struct drm_i915_private *dev_priv)
 {
-	hsw_enable_package_c8(dev_priv);
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	if (!intel_enable_rc6(dev))
+		return;
+
+	/* Make sure we're not suspended first. */
+	pm_runtime_get_sync(device);
+	pm_runtime_disable(device);
 }
 
 /* Set up chip specific power management-related functions */
@@ -5919,26 +6398,27 @@ void intel_init_pm(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (I915_HAS_FBC(dev)) {
-		if (HAS_PCH_SPLIT(dev)) {
+	if (HAS_FBC(dev)) {
+		if (INTEL_INFO(dev)->gen >= 7) {
 			dev_priv->display.fbc_enabled = ironlake_fbc_enabled;
-			if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev))
-				dev_priv->display.enable_fbc =
-					gen7_enable_fbc;
-			else
-				dev_priv->display.enable_fbc =
-					ironlake_enable_fbc;
+			dev_priv->display.enable_fbc = gen7_enable_fbc;
+			dev_priv->display.disable_fbc = ironlake_disable_fbc;
+		} else if (INTEL_INFO(dev)->gen >= 5) {
+			dev_priv->display.fbc_enabled = ironlake_fbc_enabled;
+			dev_priv->display.enable_fbc = ironlake_enable_fbc;
 			dev_priv->display.disable_fbc = ironlake_disable_fbc;
 		} else if (IS_GM45(dev)) {
 			dev_priv->display.fbc_enabled = g4x_fbc_enabled;
 			dev_priv->display.enable_fbc = g4x_enable_fbc;
 			dev_priv->display.disable_fbc = g4x_disable_fbc;
-		} else if (IS_CRESTLINE(dev)) {
+		} else {
 			dev_priv->display.fbc_enabled = i8xx_fbc_enabled;
 			dev_priv->display.enable_fbc = i8xx_enable_fbc;
 			dev_priv->display.disable_fbc = i8xx_disable_fbc;
+
+			/* This value was pulled out of someone's hat */
+			I915_WRITE(FBC_CONTROL, 500 << FBC_CTL_INTERVAL_SHIFT);
 		}
-		/* 855GM needs testing */
 	}
 
 	/* For cxsr */
@@ -5949,60 +6429,33 @@ void intel_init_pm(struct drm_device *dev)
 
 	/* For FIFO watermark updates */
 	if (HAS_PCH_SPLIT(dev)) {
-		intel_setup_wm_latency(dev);
-
-		if (IS_GEN5(dev)) {
-			if (dev_priv->wm.pri_latency[1] &&
-			    dev_priv->wm.spr_latency[1] &&
-			    dev_priv->wm.cur_latency[1])
-				dev_priv->display.update_wm = ironlake_update_wm;
-			else {
-				DRM_DEBUG_KMS("Failed to get proper latency. "
-					      "Disable CxSR\n");
-				dev_priv->display.update_wm = NULL;
-			}
+		ilk_setup_wm_latency(dev);
+
+		if ((IS_GEN5(dev) && dev_priv->wm.pri_latency[1] &&
+		     dev_priv->wm.spr_latency[1] && dev_priv->wm.cur_latency[1]) ||
+		    (!IS_GEN5(dev) && dev_priv->wm.pri_latency[0] &&
+		     dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0])) {
+			dev_priv->display.update_wm = ilk_update_wm;
+			dev_priv->display.update_sprite_wm = ilk_update_sprite_wm;
+		} else {
+			DRM_DEBUG_KMS("Failed to read display plane latency. "
+				      "Disable CxSR\n");
+		}
+
+		if (IS_GEN5(dev))
 			dev_priv->display.init_clock_gating = ironlake_init_clock_gating;
-		} else if (IS_GEN6(dev)) {
-			if (dev_priv->wm.pri_latency[0] &&
-			    dev_priv->wm.spr_latency[0] &&
-			    dev_priv->wm.cur_latency[0]) {
-				dev_priv->display.update_wm = sandybridge_update_wm;
-				dev_priv->display.update_sprite_wm = sandybridge_update_sprite_wm;
-			} else {
-				DRM_DEBUG_KMS("Failed to read display plane latency. "
-					      "Disable CxSR\n");
-				dev_priv->display.update_wm = NULL;
-			}
+		else if (IS_GEN6(dev))
 			dev_priv->display.init_clock_gating = gen6_init_clock_gating;
-		} else if (IS_IVYBRIDGE(dev)) {
-			if (dev_priv->wm.pri_latency[0] &&
-			    dev_priv->wm.spr_latency[0] &&
-			    dev_priv->wm.cur_latency[0]) {
-				dev_priv->display.update_wm = ivybridge_update_wm;
-				dev_priv->display.update_sprite_wm = sandybridge_update_sprite_wm;
-			} else {
-				DRM_DEBUG_KMS("Failed to read display plane latency. "
-					      "Disable CxSR\n");
-				dev_priv->display.update_wm = NULL;
-			}
+		else if (IS_IVYBRIDGE(dev))
 			dev_priv->display.init_clock_gating = ivybridge_init_clock_gating;
-		} else if (IS_HASWELL(dev)) {
-			if (dev_priv->wm.pri_latency[0] &&
-			    dev_priv->wm.spr_latency[0] &&
-			    dev_priv->wm.cur_latency[0]) {
-				dev_priv->display.update_wm = haswell_update_wm;
-				dev_priv->display.update_sprite_wm =
-					haswell_update_sprite_wm;
-			} else {
-				DRM_DEBUG_KMS("Failed to read display plane latency. "
-					      "Disable CxSR\n");
-				dev_priv->display.update_wm = NULL;
-			}
+		else if (IS_HASWELL(dev))
 			dev_priv->display.init_clock_gating = haswell_init_clock_gating;
-		} else if (INTEL_INFO(dev)->gen == 8) {
+		else if (INTEL_INFO(dev)->gen == 8)
 			dev_priv->display.init_clock_gating = gen8_init_clock_gating;
-		} else
-			dev_priv->display.update_wm = NULL;
+	} else if (IS_CHERRYVIEW(dev)) {
+		dev_priv->display.update_wm = valleyview_update_wm;
+		dev_priv->display.init_clock_gating =
+			cherryview_init_clock_gating;
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->display.update_wm = valleyview_update_wm;
 		dev_priv->display.init_clock_gating =
@@ -6036,21 +6489,21 @@ void intel_init_pm(struct drm_device *dev)
 		dev_priv->display.update_wm = i9xx_update_wm;
 		dev_priv->display.get_fifo_size = i9xx_get_fifo_size;
 		dev_priv->display.init_clock_gating = gen3_init_clock_gating;
-	} else if (IS_I865G(dev)) {
-		dev_priv->display.update_wm = i830_update_wm;
-		dev_priv->display.init_clock_gating = i85x_init_clock_gating;
-		dev_priv->display.get_fifo_size = i830_get_fifo_size;
-	} else if (IS_I85X(dev)) {
-		dev_priv->display.update_wm = i9xx_update_wm;
-		dev_priv->display.get_fifo_size = i85x_get_fifo_size;
-		dev_priv->display.init_clock_gating = i85x_init_clock_gating;
-	} else {
-		dev_priv->display.update_wm = i830_update_wm;
-		dev_priv->display.init_clock_gating = i830_init_clock_gating;
-		if (IS_845G(dev))
+	} else if (IS_GEN2(dev)) {
+		if (INTEL_INFO(dev)->num_pipes == 1) {
+			dev_priv->display.update_wm = i845_update_wm;
 			dev_priv->display.get_fifo_size = i845_get_fifo_size;
-		else
+		} else {
+			dev_priv->display.update_wm = i9xx_update_wm;
 			dev_priv->display.get_fifo_size = i830_get_fifo_size;
+		}
+
+		if (IS_I85X(dev) || IS_I865G(dev))
+			dev_priv->display.init_clock_gating = i85x_init_clock_gating;
+		else
+			dev_priv->display.init_clock_gating = i830_init_clock_gating;
+	} else {
+		DRM_ERROR("unexpected fall-through in intel_init_pm\n");
 	}
 }
 
@@ -6101,59 +6554,48 @@ int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val)
 	return 0;
 }
 
-int vlv_gpu_freq(int ddr_freq, int val)
+int vlv_gpu_freq(struct drm_i915_private *dev_priv, int val)
 {
-	int mult, base;
+	int div;
 
-	switch (ddr_freq) {
+	/* 4 x czclk */
+	switch (dev_priv->mem_freq) {
 	case 800:
-		mult = 20;
-		base = 120;
+		div = 10;
 		break;
 	case 1066:
-		mult = 22;
-		base = 133;
+		div = 12;
 		break;
 	case 1333:
-		mult = 21;
-		base = 125;
+		div = 16;
 		break;
 	default:
 		return -1;
 	}
 
-	return ((val - 0xbd) * mult) + base;
+	return DIV_ROUND_CLOSEST(dev_priv->mem_freq * (val + 6 - 0xbd), 4 * div);
 }
 
-int vlv_freq_opcode(int ddr_freq, int val)
+int vlv_freq_opcode(struct drm_i915_private *dev_priv, int val)
 {
-	int mult, base;
+	int mul;
 
-	switch (ddr_freq) {
+	/* 4 x czclk */
+	switch (dev_priv->mem_freq) {
 	case 800:
-		mult = 20;
-		base = 120;
+		mul = 10;
 		break;
 	case 1066:
-		mult = 22;
-		base = 133;
+		mul = 12;
 		break;
 	case 1333:
-		mult = 21;
-		base = 125;
+		mul = 16;
 		break;
 	default:
 		return -1;
 	}
 
-	val /= mult;
-	val -= base / mult;
-	val += 0xbd;
-
-	if (val > 0xea)
-		val = 0xea;
-
-	return val;
+	return DIV_ROUND_CLOSEST(4 * mul * val, dev_priv->mem_freq) + 0xbd - 6;
 }
 
 void intel_pm_setup(struct drm_device *dev)
@@ -6162,13 +6604,9 @@ void intel_pm_setup(struct drm_device *dev)
 
 	mutex_init(&dev_priv->rps.hw_lock);
 
-	mutex_init(&dev_priv->pc8.lock);
-	dev_priv->pc8.requirements_met = false;
-	dev_priv->pc8.gpu_idle = false;
-	dev_priv->pc8.irqs_disabled = false;
-	dev_priv->pc8.enabled = false;
-	dev_priv->pc8.disable_count = 2; /* requirements_met + gpu_idle */
-	INIT_DELAYED_WORK(&dev_priv->pc8.enable_work, hsw_enable_pc8_work);
 	INIT_DELAYED_WORK(&dev_priv->rps.delayed_resume_work,
 			  intel_gen6_powersave_work);
+
+	dev_priv->pm.suspended = false;
+	dev_priv->pm.irqs_disabled = false;
 }
diff --git a/drivers/gpu/drm/i915/intel_renderstate.h b/drivers/gpu/drm/i915/intel_renderstate.h
new file mode 100644
index 00000000000..a5e783a9928
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_renderstate.h
@@ -0,0 +1,48 @@
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef _INTEL_RENDERSTATE_H
+#define _INTEL_RENDERSTATE_H
+
+#include <linux/types.h>
+
+struct intel_renderstate_rodata {
+	const u32 *reloc;
+	const u32 reloc_items;
+	const u32 *batch;
+	const u32 batch_items;
+};
+
+extern const struct intel_renderstate_rodata gen6_null_state;
+extern const struct intel_renderstate_rodata gen7_null_state;
+extern const struct intel_renderstate_rodata gen8_null_state;
+
+#define RO_RENDERSTATE(_g)						\
+	const struct intel_renderstate_rodata gen ## _g ## _null_state = { \
+		.reloc = gen ## _g ## _null_state_relocs,		\
+		.reloc_items = sizeof(gen ## _g ## _null_state_relocs)/4, \
+		.batch = gen ## _g ## _null_state_batch,		\
+		.batch_items = sizeof(gen ## _g ## _null_state_batch)/4, \
+	}
+
+#endif /* INTEL_RENDERSTATE_H */
diff --git a/drivers/gpu/drm/i915/intel_renderstate_gen6.c b/drivers/gpu/drm/i915/intel_renderstate_gen6.c
new file mode 100644
index 00000000000..740538ad097
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_renderstate_gen6.c
@@ -0,0 +1,289 @@
+#include "intel_renderstate.h"
+
+static const u32 gen6_null_state_relocs[] = {
+	0x00000020,
+	0x00000024,
+	0x0000002c,
+	0x000001e0,
+	0x000001e4,
+};
+
+static const u32 gen6_null_state_batch[] = {
+	0x69040000,
+	0x790d0001,
+	0x00000000,
+	0x00000000,
+	0x78180000,
+	0x00000001,
+	0x61010008,
+	0x00000000,
+	0x00000001,	 /* reloc */
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00000001,
+	0x00000000,
+	0x00000001,
+	0x61020000,
+	0x00000000,
+	0x78050001,
+	0x00000018,
+	0x00000000,
+	0x780d1002,
+	0x00000000,
+	0x00000000,
+	0x00000420,
+	0x78150003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78100004,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78160003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78110005,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78120002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78170003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79050005,
+	0xe0040000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x79100000,
+	0x00000000,
+	0x79000002,
+	0xffffffff,
+	0x00000000,
+	0x00000000,
+	0x780e0002,
+	0x00000441,
+	0x00000401,
+	0x00000401,
+	0x78021002,
+	0x00000000,
+	0x00000000,
+	0x00000400,
+	0x78130012,
+	0x00400810,
+	0x00000000,
+	0x20000000,
+	0x04000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78140007,
+	0x00000280,
+	0x08080000,
+	0x00000000,
+	0x00060000,
+	0x4e080002,
+	0x00100400,
+	0x00000000,
+	0x00000000,
+	0x78090005,
+	0x02000000,
+	0x22220000,
+	0x02f60000,
+	0x11330000,
+	0x02850004,
+	0x11220000,
+	0x78011002,
+	0x00000000,
+	0x00000000,
+	0x00000200,
+	0x78080003,
+	0x00002000,
+	0x00000448,	 /* reloc */
+	0x00000448,	 /* reloc */
+	0x00000000,
+	0x05000000,	 /* cmds end */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000220,	 /* state start */
+	0x00000240,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0060005a,
+	0x204077be,
+	0x000000c0,
+	0x008d0040,
+	0x0060005a,
+	0x206077be,
+	0x000000c0,
+	0x008d0080,
+	0x0060005a,
+	0x208077be,
+	0x000000d0,
+	0x008d0040,
+	0x0060005a,
+	0x20a077be,
+	0x000000d0,
+	0x008d0080,
+	0x00000201,
+	0x20080061,
+	0x00000000,
+	0x00000000,
+	0x00600001,
+	0x20200022,
+	0x008d0000,
+	0x00000000,
+	0x02800031,
+	0x21c01cc9,
+	0x00000020,
+	0x0a8a0001,
+	0x00600001,
+	0x204003be,
+	0x008d01c0,
+	0x00000000,
+	0x00600001,
+	0x206003be,
+	0x008d01e0,
+	0x00000000,
+	0x00600001,
+	0x208003be,
+	0x008d0200,
+	0x00000000,
+	0x00600001,
+	0x20a003be,
+	0x008d0220,
+	0x00000000,
+	0x00600001,
+	0x20c003be,
+	0x008d0240,
+	0x00000000,
+	0x00600001,
+	0x20e003be,
+	0x008d0260,
+	0x00000000,
+	0x00600001,
+	0x210003be,
+	0x008d0280,
+	0x00000000,
+	0x00600001,
+	0x212003be,
+	0x008d02a0,
+	0x00000000,
+	0x05800031,
+	0x24001cc8,
+	0x00000040,
+	0x90019000,
+	0x0000007e,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0000007e,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0000007e,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0000007e,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0000007e,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0000007e,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0000007e,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0000007e,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x30000000,
+	0x00000124,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0xf99a130c,
+	0x799a130c,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x80000031,
+	0x00000003,
+	0x00000000,	 /* state end */
+};
+
+RO_RENDERSTATE(6);
diff --git a/drivers/gpu/drm/i915/intel_renderstate_gen7.c b/drivers/gpu/drm/i915/intel_renderstate_gen7.c
new file mode 100644
index 00000000000..6fa7ff2a129
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_renderstate_gen7.c
@@ -0,0 +1,253 @@
+#include "intel_renderstate.h"
+
+static const u32 gen7_null_state_relocs[] = {
+	0x0000000c,
+	0x00000010,
+	0x00000018,
+	0x000001ec,
+};
+
+static const u32 gen7_null_state_batch[] = {
+	0x69040000,
+	0x61010008,
+	0x00000000,
+	0x00000001,	 /* reloc */
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00000001,
+	0x00000000,
+	0x00000001,
+	0x790d0002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78180000,
+	0x00000001,
+	0x79160000,
+	0x00000008,
+	0x78300000,
+	0x02010040,
+	0x78310000,
+	0x04000000,
+	0x78320000,
+	0x04000000,
+	0x78330000,
+	0x02000000,
+	0x78100004,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781b0005,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781c0002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781d0004,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78110005,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78120002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78210000,
+	0x00000000,
+	0x78130005,
+	0x00000000,
+	0x20000000,
+	0x04000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78140001,
+	0x20000800,
+	0x00000000,
+	0x781e0001,
+	0x00000000,
+	0x00000000,
+	0x78050005,
+	0xe0040000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78040001,
+	0x00000000,
+	0x00000000,
+	0x78240000,
+	0x00000240,
+	0x78230000,
+	0x00000260,
+	0x782f0000,
+	0x00000280,
+	0x781f000c,
+	0x00400810,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78200006,
+	0x000002c0,
+	0x08080000,
+	0x00000000,
+	0x28000402,
+	0x00060000,
+	0x00000000,
+	0x00000000,
+	0x78090005,
+	0x02000000,
+	0x22220000,
+	0x02f60000,
+	0x11230000,
+	0x02f60004,
+	0x11230000,
+	0x78080003,
+	0x00006008,
+	0x00000340,	 /* reloc */
+	0xffffffff,
+	0x00000000,
+	0x782a0000,
+	0x00000360,
+	0x79000002,
+	0xffffffff,
+	0x00000000,
+	0x00000000,
+	0x7b000005,
+	0x0000000f,
+	0x00000003,
+	0x00000000,
+	0x00000001,
+	0x00000000,
+	0x00000000,
+	0x05000000,	 /* cmds end */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000031,	 /* state start */
+	0x00000003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0xf99a130c,
+	0x799a130c,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000492,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0080005a,
+	0x2e2077bd,
+	0x000000c0,
+	0x008d0040,
+	0x0080005a,
+	0x2e6077bd,
+	0x000000d0,
+	0x008d0040,
+	0x02800031,
+	0x21801fa9,
+	0x008d0e20,
+	0x08840001,
+	0x00800001,
+	0x2e2003bd,
+	0x008d0180,
+	0x00000000,
+	0x00800001,
+	0x2e6003bd,
+	0x008d01c0,
+	0x00000000,
+	0x00800001,
+	0x2ea003bd,
+	0x008d0200,
+	0x00000000,
+	0x00800001,
+	0x2ee003bd,
+	0x008d0240,
+	0x00000000,
+	0x05800031,
+	0x20001fa8,
+	0x008d0e20,
+	0x90031000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000380,
+	0x000003a0,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,	 /* state end */
+};
+
+RO_RENDERSTATE(7);
diff --git a/drivers/gpu/drm/i915/intel_renderstate_gen8.c b/drivers/gpu/drm/i915/intel_renderstate_gen8.c
new file mode 100644
index 00000000000..5c875615d42
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_renderstate_gen8.c
@@ -0,0 +1,479 @@
+#include "intel_renderstate.h"
+
+static const u32 gen8_null_state_relocs[] = {
+	0x00000048,
+	0x00000050,
+	0x00000060,
+	0x000003ec,
+};
+
+static const u32 gen8_null_state_batch[] = {
+	0x69040000,
+	0x61020001,
+	0x00000000,
+	0x00000000,
+	0x79120000,
+	0x00000000,
+	0x79130000,
+	0x00000000,
+	0x79140000,
+	0x00000000,
+	0x79150000,
+	0x00000000,
+	0x79160000,
+	0x00000000,
+	0x6101000e,
+	0x00000001,
+	0x00000000,
+	0x00000001,
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000001,	 /* reloc */
+	0x00000000,
+	0xfffff001,
+	0x00001001,
+	0xfffff001,
+	0x00001001,
+	0x78230000,
+	0x000006e0,
+	0x78210000,
+	0x00000700,
+	0x78300000,
+	0x08010040,
+	0x78330000,
+	0x08000000,
+	0x78310000,
+	0x08000000,
+	0x78320000,
+	0x08000000,
+	0x78240000,
+	0x00000641,
+	0x780e0000,
+	0x00000601,
+	0x780d0000,
+	0x00000000,
+	0x78180000,
+	0x00000001,
+	0x78520003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78190009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781b0007,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78270000,
+	0x00000000,
+	0x782c0000,
+	0x00000000,
+	0x781c0002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78160009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78110008,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78290000,
+	0x00000000,
+	0x782e0000,
+	0x00000000,
+	0x781a0009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781d0007,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78280000,
+	0x00000000,
+	0x782d0000,
+	0x00000000,
+	0x78260000,
+	0x00000000,
+	0x782b0000,
+	0x00000000,
+	0x78150009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78100007,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781e0003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78120002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x781f0002,
+	0x30400820,
+	0x00000000,
+	0x00000000,
+	0x78510009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78500003,
+	0x00210000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78130002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x782a0000,
+	0x00000480,
+	0x782f0000,
+	0x00000540,
+	0x78140000,
+	0x00000800,
+	0x78170009,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x7820000a,
+	0x00000580,
+	0x00000000,
+	0x08080000,
+	0x00000000,
+	0x00000000,
+	0x1f000002,
+	0x00060000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x784d0000,
+	0x40000000,
+	0x784f0000,
+	0x80000100,
+	0x780f0000,
+	0x00000740,
+	0x78050006,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78070003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78060003,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x78040001,
+	0x00000000,
+	0x00000001,
+	0x79000002,
+	0xffffffff,
+	0x00000000,
+	0x00000000,
+	0x78080003,
+	0x00006000,
+	0x000005e0,	 /* reloc */
+	0x00000000,
+	0x00000000,
+	0x78090005,
+	0x02000000,
+	0x22220000,
+	0x02f60000,
+	0x11230000,
+	0x02850004,
+	0x11230000,
+	0x784b0000,
+	0x0000000f,
+	0x78490001,
+	0x00000000,
+	0x00000000,
+	0x7b000005,
+	0x00000000,
+	0x00000003,
+	0x00000000,
+	0x00000001,
+	0x00000000,
+	0x00000000,
+	0x05000000,	 /* cmds end */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x000004c0,	 /* state start */
+	0x00000500,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000092,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x0060005a,
+	0x21403ae8,
+	0x3a0000c0,
+	0x008d0040,
+	0x0060005a,
+	0x21603ae8,
+	0x3a0000c0,
+	0x008d0080,
+	0x0060005a,
+	0x21803ae8,
+	0x3a0000d0,
+	0x008d0040,
+	0x0060005a,
+	0x21a03ae8,
+	0x3a0000d0,
+	0x008d0080,
+	0x02800031,
+	0x2e0022e8,
+	0x0e000140,
+	0x08840001,
+	0x05800031,
+	0x200022e0,
+	0x0e000e00,
+	0x90031000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x06200000,
+	0x00000002,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0xf99a130c,
+	0x799a130c,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x3f800000,
+	0x00000000,
+	0x3f800000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,	 /* state end */
+};
+
+RO_RENDERSTATE(8);
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index c2f09d45630..279488addf3 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -33,26 +33,44 @@
 #include "i915_trace.h"
 #include "intel_drv.h"
 
-static inline int ring_space(struct intel_ring_buffer *ring)
+/* Early gen2 devices have a cacheline of just 32 bytes, using 64 is overkill,
+ * but keeps the logic simple. Indeed, the whole purpose of this macro is just
+ * to give some inclination as to some of the magic values used in the various
+ * workarounds!
+ */
+#define CACHELINE_BYTES 64
+
+static inline int __ring_space(int head, int tail, int size)
 {
-	int space = (ring->head & HEAD_ADDR) - (ring->tail + I915_RING_FREE_SPACE);
+	int space = head - (tail + I915_RING_FREE_SPACE);
 	if (space < 0)
-		space += ring->size;
+		space += size;
 	return space;
 }
 
-void __intel_ring_advance(struct intel_ring_buffer *ring)
+static inline int ring_space(struct intel_engine_cs *ring)
+{
+	struct intel_ringbuffer *ringbuf = ring->buffer;
+	return __ring_space(ringbuf->head & HEAD_ADDR, ringbuf->tail, ringbuf->size);
+}
+
+static bool intel_ring_stopped(struct intel_engine_cs *ring)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	return dev_priv->gpu_error.stop_rings & intel_ring_flag(ring);
+}
 
-	ring->tail &= ring->size - 1;
-	if (dev_priv->gpu_error.stop_rings & intel_ring_flag(ring))
+void __intel_ring_advance(struct intel_engine_cs *ring)
+{
+	struct intel_ringbuffer *ringbuf = ring->buffer;
+	ringbuf->tail &= ringbuf->size - 1;
+	if (intel_ring_stopped(ring))
 		return;
-	ring->write_tail(ring, ring->tail);
+	ring->write_tail(ring, ringbuf->tail);
 }
 
 static int
-gen2_render_ring_flush(struct intel_ring_buffer *ring,
+gen2_render_ring_flush(struct intel_engine_cs *ring,
 		       u32	invalidate_domains,
 		       u32	flush_domains)
 {
@@ -78,7 +96,7 @@ gen2_render_ring_flush(struct intel_ring_buffer *ring,
 }
 
 static int
-gen4_render_ring_flush(struct intel_ring_buffer *ring,
+gen4_render_ring_flush(struct intel_engine_cs *ring,
 		       u32	invalidate_domains,
 		       u32	flush_domains)
 {
@@ -173,9 +191,9 @@ gen4_render_ring_flush(struct intel_ring_buffer *ring,
  * really our business.  That leaves only stall at scoreboard.
  */
 static int
-intel_emit_post_sync_nonzero_flush(struct intel_ring_buffer *ring)
+intel_emit_post_sync_nonzero_flush(struct intel_engine_cs *ring)
 {
-	u32 scratch_addr = ring->scratch.gtt_offset + 128;
+	u32 scratch_addr = ring->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
 
 
@@ -208,11 +226,11 @@ intel_emit_post_sync_nonzero_flush(struct intel_ring_buffer *ring)
 }
 
 static int
-gen6_render_ring_flush(struct intel_ring_buffer *ring,
+gen6_render_ring_flush(struct intel_engine_cs *ring,
                          u32 invalidate_domains, u32 flush_domains)
 {
 	u32 flags = 0;
-	u32 scratch_addr = ring->scratch.gtt_offset + 128;
+	u32 scratch_addr = ring->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
 
 	/* Force SNB workarounds for PIPE_CONTROL flushes */
@@ -260,7 +278,7 @@ gen6_render_ring_flush(struct intel_ring_buffer *ring,
 }
 
 static int
-gen7_render_ring_cs_stall_wa(struct intel_ring_buffer *ring)
+gen7_render_ring_cs_stall_wa(struct intel_engine_cs *ring)
 {
 	int ret;
 
@@ -278,21 +296,23 @@ gen7_render_ring_cs_stall_wa(struct intel_ring_buffer *ring)
 	return 0;
 }
 
-static int gen7_ring_fbc_flush(struct intel_ring_buffer *ring, u32 value)
+static int gen7_ring_fbc_flush(struct intel_engine_cs *ring, u32 value)
 {
 	int ret;
 
 	if (!ring->fbc_dirty)
 		return 0;
 
-	ret = intel_ring_begin(ring, 4);
+	ret = intel_ring_begin(ring, 6);
 	if (ret)
 		return ret;
-	intel_ring_emit(ring, MI_NOOP);
 	/* WaFbcNukeOn3DBlt:ivb/hsw */
 	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
 	intel_ring_emit(ring, MSG_FBC_REND_STATE);
 	intel_ring_emit(ring, value);
+	intel_ring_emit(ring, MI_STORE_REGISTER_MEM(1) | MI_SRM_LRM_GLOBAL_GTT);
+	intel_ring_emit(ring, MSG_FBC_REND_STATE);
+	intel_ring_emit(ring, ring->scratch.gtt_offset + 256);
 	intel_ring_advance(ring);
 
 	ring->fbc_dirty = false;
@@ -300,11 +320,11 @@ static int gen7_ring_fbc_flush(struct intel_ring_buffer *ring, u32 value)
 }
 
 static int
-gen7_render_ring_flush(struct intel_ring_buffer *ring,
+gen7_render_ring_flush(struct intel_engine_cs *ring,
 		       u32 invalidate_domains, u32 flush_domains)
 {
 	u32 flags = 0;
-	u32 scratch_addr = ring->scratch.gtt_offset + 128;
+	u32 scratch_addr = ring->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
 
 	/*
@@ -354,18 +374,18 @@ gen7_render_ring_flush(struct intel_ring_buffer *ring,
 	intel_ring_emit(ring, 0);
 	intel_ring_advance(ring);
 
-	if (flush_domains)
+	if (!invalidate_domains && flush_domains)
 		return gen7_ring_fbc_flush(ring, FBC_REND_NUKE);
 
 	return 0;
 }
 
 static int
-gen8_render_ring_flush(struct intel_ring_buffer *ring,
+gen8_render_ring_flush(struct intel_engine_cs *ring,
 		       u32 invalidate_domains, u32 flush_domains)
 {
 	u32 flags = 0;
-	u32 scratch_addr = ring->scratch.gtt_offset + 128;
+	u32 scratch_addr = ring->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
 
 	flags |= PIPE_CONTROL_CS_STALL;
@@ -401,23 +421,30 @@ gen8_render_ring_flush(struct intel_ring_buffer *ring,
 
 }
 
-static void ring_write_tail(struct intel_ring_buffer *ring,
+static void ring_write_tail(struct intel_engine_cs *ring,
 			    u32 value)
 {
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	I915_WRITE_TAIL(ring, value);
 }
 
-u32 intel_ring_get_active_head(struct intel_ring_buffer *ring)
+u64 intel_ring_get_active_head(struct intel_engine_cs *ring)
 {
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
-	u32 acthd_reg = INTEL_INFO(ring->dev)->gen >= 4 ?
-			RING_ACTHD(ring->mmio_base) : ACTHD;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	u64 acthd;
 
-	return I915_READ(acthd_reg);
+	if (INTEL_INFO(ring->dev)->gen >= 8)
+		acthd = I915_READ64_2x32(RING_ACTHD(ring->mmio_base),
+					 RING_ACTHD_UDW(ring->mmio_base));
+	else if (INTEL_INFO(ring->dev)->gen >= 4)
+		acthd = I915_READ(RING_ACTHD(ring->mmio_base));
+	else
+		acthd = I915_READ(ACTHD);
+
+	return acthd;
 }
 
-static void ring_setup_phys_status_page(struct intel_ring_buffer *ring)
+static void ring_setup_phys_status_page(struct intel_engine_cs *ring)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	u32 addr;
@@ -428,30 +455,42 @@ static void ring_setup_phys_status_page(struct intel_ring_buffer *ring)
 	I915_WRITE(HWS_PGA, addr);
 }
 
-static int init_ring_common(struct intel_ring_buffer *ring)
+static bool stop_ring(struct intel_engine_cs *ring)
 {
-	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct drm_i915_gem_object *obj = ring->obj;
-	int ret = 0;
-	u32 head;
-
-	gen6_gt_force_wake_get(dev_priv);
+	struct drm_i915_private *dev_priv = to_i915(ring->dev);
 
-	if (I915_NEED_GFX_HWS(dev))
-		intel_ring_setup_status_page(ring);
-	else
-		ring_setup_phys_status_page(ring);
+	if (!IS_GEN2(ring->dev)) {
+		I915_WRITE_MODE(ring, _MASKED_BIT_ENABLE(STOP_RING));
+		if (wait_for_atomic((I915_READ_MODE(ring) & MODE_IDLE) != 0, 1000)) {
+			DRM_ERROR("%s :timed out trying to stop ring\n", ring->name);
+			return false;
+		}
+	}
 
-	/* Stop the ring if it's running. */
 	I915_WRITE_CTL(ring, 0);
 	I915_WRITE_HEAD(ring, 0);
 	ring->write_tail(ring, 0);
 
-	head = I915_READ_HEAD(ring) & HEAD_ADDR;
+	if (!IS_GEN2(ring->dev)) {
+		(void)I915_READ_CTL(ring);
+		I915_WRITE_MODE(ring, _MASKED_BIT_DISABLE(STOP_RING));
+	}
 
-	/* G45 ring initialization fails to reset head to zero */
-	if (head != 0) {
+	return (I915_READ_HEAD(ring) & HEAD_ADDR) == 0;
+}
+
+static int init_ring_common(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_ringbuffer *ringbuf = ring->buffer;
+	struct drm_i915_gem_object *obj = ringbuf->obj;
+	int ret = 0;
+
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
+
+	if (!stop_ring(ring)) {
+		/* G45 ring initialization often fails to reset head to zero */
 		DRM_DEBUG_KMS("%s head not reset to zero "
 			      "ctl %08x head %08x tail %08x start %08x\n",
 			      ring->name,
@@ -460,9 +499,7 @@ static int init_ring_common(struct intel_ring_buffer *ring)
 			      I915_READ_TAIL(ring),
 			      I915_READ_START(ring));
 
-		I915_WRITE_HEAD(ring, 0);
-
-		if (I915_READ_HEAD(ring) & HEAD_ADDR) {
+		if (!stop_ring(ring)) {
 			DRM_ERROR("failed to set %s head to zero "
 				  "ctl %08x head %08x tail %08x start %08x\n",
 				  ring->name,
@@ -470,16 +507,23 @@ static int init_ring_common(struct intel_ring_buffer *ring)
 				  I915_READ_HEAD(ring),
 				  I915_READ_TAIL(ring),
 				  I915_READ_START(ring));
+			ret = -EIO;
+			goto out;
 		}
 	}
 
+	if (I915_NEED_GFX_HWS(dev))
+		intel_ring_setup_status_page(ring);
+	else
+		ring_setup_phys_status_page(ring);
+
 	/* Initialize the ring. This must happen _after_ we've cleared the ring
 	 * registers with the above sequence (the readback of the HEAD registers
 	 * also enforces ordering), otherwise the hw might lose the new ring
 	 * register values. */
 	I915_WRITE_START(ring, i915_gem_obj_ggtt_offset(obj));
 	I915_WRITE_CTL(ring,
-			((ring->size - PAGE_SIZE) & RING_NR_PAGES)
+			((ringbuf->size - PAGE_SIZE) & RING_NR_PAGES)
 			| RING_VALID);
 
 	/* If the head is still not zero, the ring is dead */
@@ -487,12 +531,11 @@ static int init_ring_common(struct intel_ring_buffer *ring)
 		     I915_READ_START(ring) == i915_gem_obj_ggtt_offset(obj) &&
 		     (I915_READ_HEAD(ring) & HEAD_ADDR) == 0, 50)) {
 		DRM_ERROR("%s initialization failed "
-				"ctl %08x head %08x tail %08x start %08x\n",
-				ring->name,
-				I915_READ_CTL(ring),
-				I915_READ_HEAD(ring),
-				I915_READ_TAIL(ring),
-				I915_READ_START(ring));
+			  "ctl %08x (valid? %d) head %08x tail %08x start %08x [expected %08lx]\n",
+			  ring->name,
+			  I915_READ_CTL(ring), I915_READ_CTL(ring) & RING_VALID,
+			  I915_READ_HEAD(ring), I915_READ_TAIL(ring),
+			  I915_READ_START(ring), (unsigned long)i915_gem_obj_ggtt_offset(obj));
 		ret = -EIO;
 		goto out;
 	}
@@ -500,22 +543,22 @@ static int init_ring_common(struct intel_ring_buffer *ring)
 	if (!drm_core_check_feature(ring->dev, DRIVER_MODESET))
 		i915_kernel_lost_context(ring->dev);
 	else {
-		ring->head = I915_READ_HEAD(ring);
-		ring->tail = I915_READ_TAIL(ring) & TAIL_ADDR;
-		ring->space = ring_space(ring);
-		ring->last_retired_head = -1;
+		ringbuf->head = I915_READ_HEAD(ring);
+		ringbuf->tail = I915_READ_TAIL(ring) & TAIL_ADDR;
+		ringbuf->space = ring_space(ring);
+		ringbuf->last_retired_head = -1;
 	}
 
 	memset(&ring->hangcheck, 0, sizeof(ring->hangcheck));
 
 out:
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 
 	return ret;
 }
 
 static int
-init_pipe_control(struct intel_ring_buffer *ring)
+init_pipe_control(struct intel_engine_cs *ring)
 {
 	int ret;
 
@@ -529,9 +572,11 @@ init_pipe_control(struct intel_ring_buffer *ring)
 		goto err;
 	}
 
-	i915_gem_object_set_cache_level(ring->scratch.obj, I915_CACHE_LLC);
+	ret = i915_gem_object_set_cache_level(ring->scratch.obj, I915_CACHE_LLC);
+	if (ret)
+		goto err_unref;
 
-	ret = i915_gem_obj_ggtt_pin(ring->scratch.obj, 4096, true, false);
+	ret = i915_gem_obj_ggtt_pin(ring->scratch.obj, 4096, 0);
 	if (ret)
 		goto err_unref;
 
@@ -547,39 +592,42 @@ init_pipe_control(struct intel_ring_buffer *ring)
 	return 0;
 
 err_unpin:
-	i915_gem_object_unpin(ring->scratch.obj);
+	i915_gem_object_ggtt_unpin(ring->scratch.obj);
 err_unref:
 	drm_gem_object_unreference(&ring->scratch.obj->base);
 err:
 	return ret;
 }
 
-static int init_render_ring(struct intel_ring_buffer *ring)
+static int init_render_ring(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret = init_ring_common(ring);
 
-	if (INTEL_INFO(dev)->gen > 3)
+	/* WaTimedSingleVertexDispatch:cl,bw,ctg,elk,ilk,snb */
+	if (INTEL_INFO(dev)->gen >= 4 && INTEL_INFO(dev)->gen < 7)
 		I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(VS_TIMER_DISPATCH));
 
 	/* We need to disable the AsyncFlip performance optimisations in order
 	 * to use MI_WAIT_FOR_EVENT within the CS. It should already be
 	 * programmed to '1' on all products.
 	 *
-	 * WaDisableAsyncFlipPerfMode:snb,ivb,hsw,vlv
+	 * WaDisableAsyncFlipPerfMode:snb,ivb,hsw,vlv,bdw,chv
 	 */
 	if (INTEL_INFO(dev)->gen >= 6)
 		I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(ASYNC_FLIP_PERF_DISABLE));
 
 	/* Required for the hardware to program scanline values for waiting */
+	/* WaEnableFlushTlbInvalidationMode:snb */
 	if (INTEL_INFO(dev)->gen == 6)
 		I915_WRITE(GFX_MODE,
-			   _MASKED_BIT_ENABLE(GFX_TLB_INVALIDATE_ALWAYS));
+			   _MASKED_BIT_ENABLE(GFX_TLB_INVALIDATE_EXPLICIT));
 
+	/* WaBCSVCSTlbInvalidationMode:ivb,vlv,hsw */
 	if (IS_GEN7(dev))
 		I915_WRITE(GFX_MODE_GEN7,
-			   _MASKED_BIT_DISABLE(GFX_TLB_INVALIDATE_ALWAYS) |
+			   _MASKED_BIT_ENABLE(GFX_TLB_INVALIDATE_EXPLICIT) |
 			   _MASKED_BIT_ENABLE(GFX_REPLAY_MODE));
 
 	if (INTEL_INFO(dev)->gen >= 5) {
@@ -596,13 +644,6 @@ static int init_render_ring(struct intel_ring_buffer *ring)
 		 */
 		I915_WRITE(CACHE_MODE_0,
 			   _MASKED_BIT_DISABLE(CM0_STC_EVICT_DISABLE_LRA_SNB));
-
-		/* This is not explicitly set for GEN6, so read the register.
-		 * see intel_ring_mi_set_context() for why we care.
-		 * TODO: consider explicitly setting the bit for GEN5
-		 */
-		ring->itlb_before_ctx_switch =
-			!!(I915_READ(GFX_MODE) & GFX_TLB_INVALIDATE_ALWAYS);
 	}
 
 	if (INTEL_INFO(dev)->gen >= 6)
@@ -614,7 +655,7 @@ static int init_render_ring(struct intel_ring_buffer *ring)
 	return ret;
 }
 
-static void render_ring_cleanup(struct intel_ring_buffer *ring)
+static void render_ring_cleanup(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
 
@@ -623,27 +664,53 @@ static void render_ring_cleanup(struct intel_ring_buffer *ring)
 
 	if (INTEL_INFO(dev)->gen >= 5) {
 		kunmap(sg_page(ring->scratch.obj->pages->sgl));
-		i915_gem_object_unpin(ring->scratch.obj);
+		i915_gem_object_ggtt_unpin(ring->scratch.obj);
 	}
 
 	drm_gem_object_unreference(&ring->scratch.obj->base);
 	ring->scratch.obj = NULL;
 }
 
-static void
-update_mboxes(struct intel_ring_buffer *ring,
-	      u32 mmio_offset)
+static int gen6_signal(struct intel_engine_cs *signaller,
+		       unsigned int num_dwords)
 {
-/* NB: In order to be able to do semaphore MBOX updates for varying number
- * of rings, it's easiest if we round up each individual update to a
- * multiple of 2 (since ring updates must always be a multiple of 2)
- * even though the actual update only requires 3 dwords.
- */
+	struct drm_device *dev = signaller->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *useless;
+	int i, ret;
+
+	/* NB: In order to be able to do semaphore MBOX updates for varying
+	 * number of rings, it's easiest if we round up each individual update
+	 * to a multiple of 2 (since ring updates must always be a multiple of
+	 * 2) even though the actual update only requires 3 dwords.
+	 */
 #define MBOX_UPDATE_DWORDS 4
-	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
-	intel_ring_emit(ring, mmio_offset);
-	intel_ring_emit(ring, ring->outstanding_lazy_seqno);
-	intel_ring_emit(ring, MI_NOOP);
+	if (i915_semaphore_is_enabled(dev))
+		num_dwords += ((I915_NUM_RINGS-1) * MBOX_UPDATE_DWORDS);
+	else
+		return intel_ring_begin(signaller, num_dwords);
+
+	ret = intel_ring_begin(signaller, num_dwords);
+	if (ret)
+		return ret;
+#undef MBOX_UPDATE_DWORDS
+
+	for_each_ring(useless, dev_priv, i) {
+		u32 mbox_reg = signaller->semaphore.mbox.signal[i];
+		if (mbox_reg != GEN6_NOSYNC) {
+			intel_ring_emit(signaller, MI_LOAD_REGISTER_IMM(1));
+			intel_ring_emit(signaller, mbox_reg);
+			intel_ring_emit(signaller, signaller->outstanding_lazy_seqno);
+			intel_ring_emit(signaller, MI_NOOP);
+		} else {
+			intel_ring_emit(signaller, MI_NOOP);
+			intel_ring_emit(signaller, MI_NOOP);
+			intel_ring_emit(signaller, MI_NOOP);
+			intel_ring_emit(signaller, MI_NOOP);
+		}
+	}
+
+	return 0;
 }
 
 /**
@@ -656,25 +723,13 @@ update_mboxes(struct intel_ring_buffer *ring,
  * This acts like a signal in the canonical semaphore.
  */
 static int
-gen6_add_request(struct intel_ring_buffer *ring)
+gen6_add_request(struct intel_engine_cs *ring)
 {
-	struct drm_device *dev = ring->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *useless;
-	int i, ret;
+	int ret;
 
-	ret = intel_ring_begin(ring, ((I915_NUM_RINGS-1) *
-				      MBOX_UPDATE_DWORDS) +
-				      4);
+	ret = ring->semaphore.signal(ring, 4);
 	if (ret)
 		return ret;
-#undef MBOX_UPDATE_DWORDS
-
-	for_each_ring(useless, dev_priv, i) {
-		u32 mbox_reg = ring->signal_mbox[i];
-		if (mbox_reg != GEN6_NOSYNC)
-			update_mboxes(ring, mbox_reg);
-	}
 
 	intel_ring_emit(ring, MI_STORE_DWORD_INDEX);
 	intel_ring_emit(ring, I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT);
@@ -700,14 +755,15 @@ static inline bool i915_gem_has_seqno_wrapped(struct drm_device *dev,
  * @seqno - seqno which the waiter will block on
  */
 static int
-gen6_ring_sync(struct intel_ring_buffer *waiter,
-	       struct intel_ring_buffer *signaller,
+gen6_ring_sync(struct intel_engine_cs *waiter,
+	       struct intel_engine_cs *signaller,
 	       u32 seqno)
 {
-	int ret;
 	u32 dw1 = MI_SEMAPHORE_MBOX |
 		  MI_SEMAPHORE_COMPARE |
 		  MI_SEMAPHORE_REGISTER;
+	u32 wait_mbox = signaller->semaphore.mbox.wait[waiter->id];
+	int ret;
 
 	/* Throughout all of the GEM code, seqno passed implies our current
 	 * seqno is >= the last seqno executed. However for hardware the
@@ -715,8 +771,7 @@ gen6_ring_sync(struct intel_ring_buffer *waiter,
 	 */
 	seqno -= 1;
 
-	WARN_ON(signaller->semaphore_register[waiter->id] ==
-		MI_SEMAPHORE_SYNC_INVALID);
+	WARN_ON(wait_mbox == MI_SEMAPHORE_SYNC_INVALID);
 
 	ret = intel_ring_begin(waiter, 4);
 	if (ret)
@@ -724,9 +779,7 @@ gen6_ring_sync(struct intel_ring_buffer *waiter,
 
 	/* If seqno wrap happened, omit the wait with no-ops */
 	if (likely(!i915_gem_has_seqno_wrapped(waiter->dev, seqno))) {
-		intel_ring_emit(waiter,
-				dw1 |
-				signaller->semaphore_register[waiter->id]);
+		intel_ring_emit(waiter, dw1 | wait_mbox);
 		intel_ring_emit(waiter, seqno);
 		intel_ring_emit(waiter, 0);
 		intel_ring_emit(waiter, MI_NOOP);
@@ -751,9 +804,9 @@ do {									\
 } while (0)
 
 static int
-pc_render_add_request(struct intel_ring_buffer *ring)
+pc_render_add_request(struct intel_engine_cs *ring)
 {
-	u32 scratch_addr = ring->scratch.gtt_offset + 128;
+	u32 scratch_addr = ring->scratch.gtt_offset + 2 * CACHELINE_BYTES;
 	int ret;
 
 	/* For Ironlake, MI_USER_INTERRUPT was deprecated and apparently
@@ -775,15 +828,15 @@ pc_render_add_request(struct intel_ring_buffer *ring)
 	intel_ring_emit(ring, ring->outstanding_lazy_seqno);
 	intel_ring_emit(ring, 0);
 	PIPE_CONTROL_FLUSH(ring, scratch_addr);
-	scratch_addr += 128; /* write to separate cachelines */
+	scratch_addr += 2 * CACHELINE_BYTES; /* write to separate cachelines */
 	PIPE_CONTROL_FLUSH(ring, scratch_addr);
-	scratch_addr += 128;
+	scratch_addr += 2 * CACHELINE_BYTES;
 	PIPE_CONTROL_FLUSH(ring, scratch_addr);
-	scratch_addr += 128;
+	scratch_addr += 2 * CACHELINE_BYTES;
 	PIPE_CONTROL_FLUSH(ring, scratch_addr);
-	scratch_addr += 128;
+	scratch_addr += 2 * CACHELINE_BYTES;
 	PIPE_CONTROL_FLUSH(ring, scratch_addr);
-	scratch_addr += 128;
+	scratch_addr += 2 * CACHELINE_BYTES;
 	PIPE_CONTROL_FLUSH(ring, scratch_addr);
 
 	intel_ring_emit(ring, GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE |
@@ -799,45 +852,48 @@ pc_render_add_request(struct intel_ring_buffer *ring)
 }
 
 static u32
-gen6_ring_get_seqno(struct intel_ring_buffer *ring, bool lazy_coherency)
+gen6_ring_get_seqno(struct intel_engine_cs *ring, bool lazy_coherency)
 {
 	/* Workaround to force correct ordering between irq and seqno writes on
 	 * ivb (and maybe also on snb) by reading from a CS register (like
 	 * ACTHD) before reading the status page. */
-	if (!lazy_coherency)
-		intel_ring_get_active_head(ring);
+	if (!lazy_coherency) {
+		struct drm_i915_private *dev_priv = ring->dev->dev_private;
+		POSTING_READ(RING_ACTHD(ring->mmio_base));
+	}
+
 	return intel_read_status_page(ring, I915_GEM_HWS_INDEX);
 }
 
 static u32
-ring_get_seqno(struct intel_ring_buffer *ring, bool lazy_coherency)
+ring_get_seqno(struct intel_engine_cs *ring, bool lazy_coherency)
 {
 	return intel_read_status_page(ring, I915_GEM_HWS_INDEX);
 }
 
 static void
-ring_set_seqno(struct intel_ring_buffer *ring, u32 seqno)
+ring_set_seqno(struct intel_engine_cs *ring, u32 seqno)
 {
 	intel_write_status_page(ring, I915_GEM_HWS_INDEX, seqno);
 }
 
 static u32
-pc_render_get_seqno(struct intel_ring_buffer *ring, bool lazy_coherency)
+pc_render_get_seqno(struct intel_engine_cs *ring, bool lazy_coherency)
 {
 	return ring->scratch.cpu_page[0];
 }
 
 static void
-pc_render_set_seqno(struct intel_ring_buffer *ring, u32 seqno)
+pc_render_set_seqno(struct intel_engine_cs *ring, u32 seqno)
 {
 	ring->scratch.cpu_page[0] = seqno;
 }
 
 static bool
-gen5_ring_get_irq(struct intel_ring_buffer *ring)
+gen5_ring_get_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
 	if (!dev->irq_enabled)
@@ -852,10 +908,10 @@ gen5_ring_get_irq(struct intel_ring_buffer *ring)
 }
 
 static void
-gen5_ring_put_irq(struct intel_ring_buffer *ring)
+gen5_ring_put_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
@@ -865,10 +921,10 @@ gen5_ring_put_irq(struct intel_ring_buffer *ring)
 }
 
 static bool
-i9xx_ring_get_irq(struct intel_ring_buffer *ring)
+i9xx_ring_get_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
 	if (!dev->irq_enabled)
@@ -886,10 +942,10 @@ i9xx_ring_get_irq(struct intel_ring_buffer *ring)
 }
 
 static void
-i9xx_ring_put_irq(struct intel_ring_buffer *ring)
+i9xx_ring_put_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
@@ -902,10 +958,10 @@ i9xx_ring_put_irq(struct intel_ring_buffer *ring)
 }
 
 static bool
-i8xx_ring_get_irq(struct intel_ring_buffer *ring)
+i8xx_ring_get_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
 	if (!dev->irq_enabled)
@@ -923,10 +979,10 @@ i8xx_ring_get_irq(struct intel_ring_buffer *ring)
 }
 
 static void
-i8xx_ring_put_irq(struct intel_ring_buffer *ring)
+i8xx_ring_put_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
@@ -938,10 +994,10 @@ i8xx_ring_put_irq(struct intel_ring_buffer *ring)
 	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
 }
 
-void intel_ring_setup_status_page(struct intel_ring_buffer *ring)
+void intel_ring_setup_status_page(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	u32 mmio = 0;
 
 	/* The ring status page addresses are no longer next to the rest of
@@ -955,6 +1011,11 @@ void intel_ring_setup_status_page(struct intel_ring_buffer *ring)
 		case BCS:
 			mmio = BLT_HWS_PGA_GEN7;
 			break;
+		/*
+		 * VCS2 actually doesn't exist on Gen7. Only shut up
+		 * gcc switch check warning
+		 */
+		case VCS2:
 		case VCS:
 			mmio = BSD_HWS_PGA_GEN7;
 			break;
@@ -972,9 +1033,19 @@ void intel_ring_setup_status_page(struct intel_ring_buffer *ring)
 	I915_WRITE(mmio, (u32)ring->status_page.gfx_addr);
 	POSTING_READ(mmio);
 
-	/* Flush the TLB for this page */
-	if (INTEL_INFO(dev)->gen >= 6) {
+	/*
+	 * Flush the TLB for this page
+	 *
+	 * FIXME: These two bits have disappeared on gen8, so a question
+	 * arises: do we still need this and if so how should we go about
+	 * invalidating the TLB?
+	 */
+	if (INTEL_INFO(dev)->gen >= 6 && INTEL_INFO(dev)->gen < 8) {
 		u32 reg = RING_INSTPM(ring->mmio_base);
+
+		/* ring should be idle before issuing a sync flush*/
+		WARN_ON((I915_READ_MODE(ring) & MODE_IDLE) == 0);
+
 		I915_WRITE(reg,
 			   _MASKED_BIT_ENABLE(INSTPM_TLB_INVALIDATE |
 					      INSTPM_SYNC_FLUSH));
@@ -986,7 +1057,7 @@ void intel_ring_setup_status_page(struct intel_ring_buffer *ring)
 }
 
 static int
-bsd_ring_flush(struct intel_ring_buffer *ring,
+bsd_ring_flush(struct intel_engine_cs *ring,
 	       u32     invalidate_domains,
 	       u32     flush_domains)
 {
@@ -1003,7 +1074,7 @@ bsd_ring_flush(struct intel_ring_buffer *ring,
 }
 
 static int
-i9xx_add_request(struct intel_ring_buffer *ring)
+i9xx_add_request(struct intel_engine_cs *ring)
 {
 	int ret;
 
@@ -1021,20 +1092,15 @@ i9xx_add_request(struct intel_ring_buffer *ring)
 }
 
 static bool
-gen6_ring_get_irq(struct intel_ring_buffer *ring)
+gen6_ring_get_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
 	if (!dev->irq_enabled)
 	       return false;
 
-	/* It looks like we need to prevent the gt from suspending while waiting
-	 * for an notifiy irq, otherwise irqs seem to get lost on at least the
-	 * blt/bsd rings on ivb. */
-	gen6_gt_force_wake_get(dev_priv);
-
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
 	if (ring->irq_refcount++ == 0) {
 		if (HAS_L3_DPF(dev) && ring->id == RCS)
@@ -1051,10 +1117,10 @@ gen6_ring_get_irq(struct intel_ring_buffer *ring)
 }
 
 static void
-gen6_ring_put_irq(struct intel_ring_buffer *ring)
+gen6_ring_put_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
-	drm_i915_private_t *dev_priv = dev->dev_private;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long flags;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
@@ -1066,12 +1132,10 @@ gen6_ring_put_irq(struct intel_ring_buffer *ring)
 		ilk_disable_gt_irq(dev_priv, ring->irq_enable_mask);
 	}
 	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
-
-	gen6_gt_force_wake_put(dev_priv);
 }
 
 static bool
-hsw_vebox_get_irq(struct intel_ring_buffer *ring)
+hsw_vebox_get_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1091,7 +1155,7 @@ hsw_vebox_get_irq(struct intel_ring_buffer *ring)
 }
 
 static void
-hsw_vebox_put_irq(struct intel_ring_buffer *ring)
+hsw_vebox_put_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1109,7 +1173,7 @@ hsw_vebox_put_irq(struct intel_ring_buffer *ring)
 }
 
 static bool
-gen8_ring_get_irq(struct intel_ring_buffer *ring)
+gen8_ring_get_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1135,7 +1199,7 @@ gen8_ring_get_irq(struct intel_ring_buffer *ring)
 }
 
 static void
-gen8_ring_put_irq(struct intel_ring_buffer *ring)
+gen8_ring_put_irq(struct intel_engine_cs *ring)
 {
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1155,8 +1219,8 @@ gen8_ring_put_irq(struct intel_ring_buffer *ring)
 }
 
 static int
-i965_dispatch_execbuffer(struct intel_ring_buffer *ring,
-			 u32 offset, u32 length,
+i965_dispatch_execbuffer(struct intel_engine_cs *ring,
+			 u64 offset, u32 length,
 			 unsigned flags)
 {
 	int ret;
@@ -1178,8 +1242,8 @@ i965_dispatch_execbuffer(struct intel_ring_buffer *ring,
 /* Just userspace ABI convention to limit the wa batch bo to a resonable size */
 #define I830_BATCH_LIMIT (256*1024)
 static int
-i830_dispatch_execbuffer(struct intel_ring_buffer *ring,
-				u32 offset, u32 len,
+i830_dispatch_execbuffer(struct intel_engine_cs *ring,
+				u64 offset, u32 len,
 				unsigned flags)
 {
 	int ret;
@@ -1229,8 +1293,8 @@ i830_dispatch_execbuffer(struct intel_ring_buffer *ring,
 }
 
 static int
-i915_dispatch_execbuffer(struct intel_ring_buffer *ring,
-			 u32 offset, u32 len,
+i915_dispatch_execbuffer(struct intel_engine_cs *ring,
+			 u64 offset, u32 len,
 			 unsigned flags)
 {
 	int ret;
@@ -1246,7 +1310,7 @@ i915_dispatch_execbuffer(struct intel_ring_buffer *ring,
 	return 0;
 }
 
-static void cleanup_status_page(struct intel_ring_buffer *ring)
+static void cleanup_status_page(struct intel_engine_cs *ring)
 {
 	struct drm_i915_gem_object *obj;
 
@@ -1255,54 +1319,49 @@ static void cleanup_status_page(struct intel_ring_buffer *ring)
 		return;
 
 	kunmap(sg_page(obj->pages->sgl));
-	i915_gem_object_unpin(obj);
+	i915_gem_object_ggtt_unpin(obj);
 	drm_gem_object_unreference(&obj->base);
 	ring->status_page.obj = NULL;
 }
 
-static int init_status_page(struct intel_ring_buffer *ring)
+static int init_status_page(struct intel_engine_cs *ring)
 {
-	struct drm_device *dev = ring->dev;
 	struct drm_i915_gem_object *obj;
-	int ret;
 
-	obj = i915_gem_alloc_object(dev, 4096);
-	if (obj == NULL) {
-		DRM_ERROR("Failed to allocate status page\n");
-		ret = -ENOMEM;
-		goto err;
-	}
+	if ((obj = ring->status_page.obj) == NULL) {
+		int ret;
+
+		obj = i915_gem_alloc_object(ring->dev, 4096);
+		if (obj == NULL) {
+			DRM_ERROR("Failed to allocate status page\n");
+			return -ENOMEM;
+		}
 
-	i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
+		ret = i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
+		if (ret)
+			goto err_unref;
 
-	ret = i915_gem_obj_ggtt_pin(obj, 4096, true, false);
-	if (ret != 0) {
-		goto err_unref;
+		ret = i915_gem_obj_ggtt_pin(obj, 4096, 0);
+		if (ret) {
+err_unref:
+			drm_gem_object_unreference(&obj->base);
+			return ret;
+		}
+
+		ring->status_page.obj = obj;
 	}
 
 	ring->status_page.gfx_addr = i915_gem_obj_ggtt_offset(obj);
 	ring->status_page.page_addr = kmap(sg_page(obj->pages->sgl));
-	if (ring->status_page.page_addr == NULL) {
-		ret = -ENOMEM;
-		goto err_unpin;
-	}
-	ring->status_page.obj = obj;
 	memset(ring->status_page.page_addr, 0, PAGE_SIZE);
 
 	DRM_DEBUG_DRIVER("%s hws offset: 0x%08x\n",
 			ring->name, ring->status_page.gfx_addr);
 
 	return 0;
-
-err_unpin:
-	i915_gem_object_unpin(obj);
-err_unref:
-	drm_gem_object_unreference(&obj->base);
-err:
-	return ret;
 }
 
-static int init_phys_status_page(struct intel_ring_buffer *ring)
+static int init_phys_status_page(struct intel_engine_cs *ring)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 
@@ -1319,46 +1378,26 @@ static int init_phys_status_page(struct intel_ring_buffer *ring)
 	return 0;
 }
 
-static int intel_init_ring_buffer(struct drm_device *dev,
-				  struct intel_ring_buffer *ring)
+static int allocate_ring_buffer(struct intel_engine_cs *ring)
 {
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_ringbuffer *ringbuf = ring->buffer;
 	struct drm_i915_gem_object *obj;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
-	ring->dev = dev;
-	INIT_LIST_HEAD(&ring->active_list);
-	INIT_LIST_HEAD(&ring->request_list);
-	ring->size = 32 * PAGE_SIZE;
-	memset(ring->sync_seqno, 0, sizeof(ring->sync_seqno));
-
-	init_waitqueue_head(&ring->irq_queue);
-
-	if (I915_NEED_GFX_HWS(dev)) {
-		ret = init_status_page(ring);
-		if (ret)
-			return ret;
-	} else {
-		BUG_ON(ring->id != RCS);
-		ret = init_phys_status_page(ring);
-		if (ret)
-			return ret;
-	}
+	if (intel_ring_initialized(ring))
+		return 0;
 
 	obj = NULL;
 	if (!HAS_LLC(dev))
-		obj = i915_gem_object_create_stolen(dev, ring->size);
+		obj = i915_gem_object_create_stolen(dev, ringbuf->size);
 	if (obj == NULL)
-		obj = i915_gem_alloc_object(dev, ring->size);
-	if (obj == NULL) {
-		DRM_ERROR("Failed to allocate ringbuffer\n");
-		ret = -ENOMEM;
-		goto err_hws;
-	}
-
-	ring->obj = obj;
+		obj = i915_gem_alloc_object(dev, ringbuf->size);
+	if (obj == NULL)
+		return -ENOMEM;
 
-	ret = i915_gem_obj_ggtt_pin(obj, PAGE_SIZE, true, false);
+	ret = i915_gem_obj_ggtt_pin(obj, PAGE_SIZE, PIN_MAPPABLE);
 	if (ret)
 		goto err_unref;
 
@@ -1366,63 +1405,102 @@ static int intel_init_ring_buffer(struct drm_device *dev,
 	if (ret)
 		goto err_unpin;
 
-	ring->virtual_start =
+	ringbuf->virtual_start =
 		ioremap_wc(dev_priv->gtt.mappable_base + i915_gem_obj_ggtt_offset(obj),
-			   ring->size);
-	if (ring->virtual_start == NULL) {
-		DRM_ERROR("Failed to map ringbuffer.\n");
+				ringbuf->size);
+	if (ringbuf->virtual_start == NULL) {
 		ret = -EINVAL;
 		goto err_unpin;
 	}
 
-	ret = ring->init(ring);
-	if (ret)
-		goto err_unmap;
+	ringbuf->obj = obj;
+	return 0;
+
+err_unpin:
+	i915_gem_object_ggtt_unpin(obj);
+err_unref:
+	drm_gem_object_unreference(&obj->base);
+	return ret;
+}
+
+static int intel_init_ring_buffer(struct drm_device *dev,
+				  struct intel_engine_cs *ring)
+{
+	struct intel_ringbuffer *ringbuf = ring->buffer;
+	int ret;
+
+	if (ringbuf == NULL) {
+		ringbuf = kzalloc(sizeof(*ringbuf), GFP_KERNEL);
+		if (!ringbuf)
+			return -ENOMEM;
+		ring->buffer = ringbuf;
+	}
+
+	ring->dev = dev;
+	INIT_LIST_HEAD(&ring->active_list);
+	INIT_LIST_HEAD(&ring->request_list);
+	ringbuf->size = 32 * PAGE_SIZE;
+	memset(ring->semaphore.sync_seqno, 0, sizeof(ring->semaphore.sync_seqno));
+
+	init_waitqueue_head(&ring->irq_queue);
+
+	if (I915_NEED_GFX_HWS(dev)) {
+		ret = init_status_page(ring);
+		if (ret)
+			goto error;
+	} else {
+		BUG_ON(ring->id != RCS);
+		ret = init_phys_status_page(ring);
+		if (ret)
+			goto error;
+	}
+
+	ret = allocate_ring_buffer(ring);
+	if (ret) {
+		DRM_ERROR("Failed to allocate ringbuffer %s: %d\n", ring->name, ret);
+		goto error;
+	}
 
 	/* Workaround an erratum on the i830 which causes a hang if
 	 * the TAIL pointer points to within the last 2 cachelines
 	 * of the buffer.
 	 */
-	ring->effective_size = ring->size;
-	if (IS_I830(ring->dev) || IS_845G(ring->dev))
-		ring->effective_size -= 128;
+	ringbuf->effective_size = ringbuf->size;
+	if (IS_I830(dev) || IS_845G(dev))
+		ringbuf->effective_size -= 2 * CACHELINE_BYTES;
+
+	ret = i915_cmd_parser_init_ring(ring);
+	if (ret)
+		goto error;
+
+	ret = ring->init(ring);
+	if (ret)
+		goto error;
 
 	return 0;
 
-err_unmap:
-	iounmap(ring->virtual_start);
-err_unpin:
-	i915_gem_object_unpin(obj);
-err_unref:
-	drm_gem_object_unreference(&obj->base);
-	ring->obj = NULL;
-err_hws:
-	cleanup_status_page(ring);
+error:
+	kfree(ringbuf);
+	ring->buffer = NULL;
 	return ret;
 }
 
-void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring)
+void intel_cleanup_ring_buffer(struct intel_engine_cs *ring)
 {
-	struct drm_i915_private *dev_priv;
-	int ret;
+	struct drm_i915_private *dev_priv = to_i915(ring->dev);
+	struct intel_ringbuffer *ringbuf = ring->buffer;
 
-	if (ring->obj == NULL)
+	if (!intel_ring_initialized(ring))
 		return;
 
-	/* Disable the ring buffer. The ring must be idle at this point */
-	dev_priv = ring->dev->dev_private;
-	ret = intel_ring_idle(ring);
-	if (ret && !i915_reset_in_progress(&dev_priv->gpu_error))
-		DRM_ERROR("failed to quiesce %s whilst cleaning up: %d\n",
-			  ring->name, ret);
-
-	I915_WRITE_CTL(ring, 0);
+	intel_stop_ring_buffer(ring);
+	WARN_ON(!IS_GEN2(ring->dev) && (I915_READ_MODE(ring) & MODE_IDLE) == 0);
 
-	iounmap(ring->virtual_start);
+	iounmap(ringbuf->virtual_start);
 
-	i915_gem_object_unpin(ring->obj);
-	drm_gem_object_unreference(&ring->obj->base);
-	ring->obj = NULL;
+	i915_gem_object_ggtt_unpin(ringbuf->obj);
+	drm_gem_object_unreference(&ringbuf->obj->base);
+	ringbuf->obj = NULL;
 	ring->preallocated_lazy_request = NULL;
 	ring->outstanding_lazy_seqno = 0;
 
@@ -1430,80 +1508,56 @@ void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring)
 		ring->cleanup(ring);
 
 	cleanup_status_page(ring);
-}
 
-static int intel_ring_wait_seqno(struct intel_ring_buffer *ring, u32 seqno)
-{
-	int ret;
+	i915_cmd_parser_fini_ring(ring);
 
-	ret = i915_wait_seqno(ring, seqno);
-	if (!ret)
-		i915_gem_retire_requests_ring(ring);
-
-	return ret;
+	kfree(ringbuf);
+	ring->buffer = NULL;
 }
 
-static int intel_ring_wait_request(struct intel_ring_buffer *ring, int n)
+static int intel_ring_wait_request(struct intel_engine_cs *ring, int n)
 {
+	struct intel_ringbuffer *ringbuf = ring->buffer;
 	struct drm_i915_gem_request *request;
 	u32 seqno = 0;
 	int ret;
 
-	i915_gem_retire_requests_ring(ring);
+	if (ringbuf->last_retired_head != -1) {
+		ringbuf->head = ringbuf->last_retired_head;
+		ringbuf->last_retired_head = -1;
 
-	if (ring->last_retired_head != -1) {
-		ring->head = ring->last_retired_head;
-		ring->last_retired_head = -1;
-		ring->space = ring_space(ring);
-		if (ring->space >= n)
+		ringbuf->space = ring_space(ring);
+		if (ringbuf->space >= n)
 			return 0;
 	}
 
 	list_for_each_entry(request, &ring->request_list, list) {
-		int space;
-
-		if (request->tail == -1)
-			continue;
-
-		space = request->tail - (ring->tail + I915_RING_FREE_SPACE);
-		if (space < 0)
-			space += ring->size;
-		if (space >= n) {
+		if (__ring_space(request->tail, ringbuf->tail, ringbuf->size) >= n) {
 			seqno = request->seqno;
 			break;
 		}
-
-		/* Consume this request in case we need more space than
-		 * is available and so need to prevent a race between
-		 * updating last_retired_head and direct reads of
-		 * I915_RING_HEAD. It also provides a nice sanity check.
-		 */
-		request->tail = -1;
 	}
 
 	if (seqno == 0)
 		return -ENOSPC;
 
-	ret = intel_ring_wait_seqno(ring, seqno);
+	ret = i915_wait_seqno(ring, seqno);
 	if (ret)
 		return ret;
 
-	if (WARN_ON(ring->last_retired_head == -1))
-		return -ENOSPC;
-
-	ring->head = ring->last_retired_head;
-	ring->last_retired_head = -1;
-	ring->space = ring_space(ring);
-	if (WARN_ON(ring->space < n))
-		return -ENOSPC;
+	i915_gem_retire_requests_ring(ring);
+	ringbuf->head = ringbuf->last_retired_head;
+	ringbuf->last_retired_head = -1;
 
+	ringbuf->space = ring_space(ring);
 	return 0;
 }
 
-static int ring_wait_for_space(struct intel_ring_buffer *ring, int n)
+static int ring_wait_for_space(struct intel_engine_cs *ring, int n)
 {
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_ringbuffer *ringbuf = ring->buffer;
 	unsigned long end;
 	int ret;
 
@@ -1514,7 +1568,6 @@ static int ring_wait_for_space(struct intel_ring_buffer *ring, int n)
 	/* force the tail write in case we have been skipping them */
 	__intel_ring_advance(ring);
 
-	trace_i915_ring_wait_begin(ring);
 	/* With GEM the hangcheck timer should kick us out of the loop,
 	 * leaving it early runs the risk of corrupting GEM state (due
 	 * to running on almost untested codepaths). But on resume
@@ -1522,15 +1575,17 @@ static int ring_wait_for_space(struct intel_ring_buffer *ring, int n)
 	 * case by choosing an insanely large timeout. */
 	end = jiffies + 60 * HZ;
 
+	trace_i915_ring_wait_begin(ring);
 	do {
-		ring->head = I915_READ_HEAD(ring);
-		ring->space = ring_space(ring);
-		if (ring->space >= n) {
-			trace_i915_ring_wait_end(ring);
-			return 0;
+		ringbuf->head = I915_READ_HEAD(ring);
+		ringbuf->space = ring_space(ring);
+		if (ringbuf->space >= n) {
+			ret = 0;
+			break;
 		}
 
-		if (dev->primary->master) {
+		if (!drm_core_check_feature(dev, DRIVER_MODESET) &&
+		    dev->primary->master) {
 			struct drm_i915_master_private *master_priv = dev->primary->master->driver_priv;
 			if (master_priv->sarea_priv)
 				master_priv->sarea_priv->perf_boxes |= I915_BOX_WAIT;
@@ -1538,38 +1593,49 @@ static int ring_wait_for_space(struct intel_ring_buffer *ring, int n)
 
 		msleep(1);
 
+		if (dev_priv->mm.interruptible && signal_pending(current)) {
+			ret = -ERESTARTSYS;
+			break;
+		}
+
 		ret = i915_gem_check_wedge(&dev_priv->gpu_error,
 					   dev_priv->mm.interruptible);
 		if (ret)
-			return ret;
-	} while (!time_after(jiffies, end));
+			break;
+
+		if (time_after(jiffies, end)) {
+			ret = -EBUSY;
+			break;
+		}
+	} while (1);
 	trace_i915_ring_wait_end(ring);
-	return -EBUSY;
+	return ret;
 }
 
-static int intel_wrap_ring_buffer(struct intel_ring_buffer *ring)
+static int intel_wrap_ring_buffer(struct intel_engine_cs *ring)
 {
 	uint32_t __iomem *virt;
-	int rem = ring->size - ring->tail;
+	struct intel_ringbuffer *ringbuf = ring->buffer;
+	int rem = ringbuf->size - ringbuf->tail;
 
-	if (ring->space < rem) {
+	if (ringbuf->space < rem) {
 		int ret = ring_wait_for_space(ring, rem);
 		if (ret)
 			return ret;
 	}
 
-	virt = ring->virtual_start + ring->tail;
+	virt = ringbuf->virtual_start + ringbuf->tail;
 	rem /= 4;
 	while (rem--)
 		iowrite32(MI_NOOP, virt++);
 
-	ring->tail = 0;
-	ring->space = ring_space(ring);
+	ringbuf->tail = 0;
+	ringbuf->space = ring_space(ring);
 
 	return 0;
 }
 
-int intel_ring_idle(struct intel_ring_buffer *ring)
+int intel_ring_idle(struct intel_engine_cs *ring)
 {
 	u32 seqno;
 	int ret;
@@ -1593,7 +1659,7 @@ int intel_ring_idle(struct intel_ring_buffer *ring)
 }
 
 static int
-intel_ring_alloc_seqno(struct intel_ring_buffer *ring)
+intel_ring_alloc_seqno(struct intel_engine_cs *ring)
 {
 	if (ring->outstanding_lazy_seqno)
 		return 0;
@@ -1611,31 +1677,31 @@ intel_ring_alloc_seqno(struct intel_ring_buffer *ring)
 	return i915_gem_get_seqno(ring->dev, &ring->outstanding_lazy_seqno);
 }
 
-static int __intel_ring_begin(struct intel_ring_buffer *ring,
-			      int bytes)
+static int __intel_ring_prepare(struct intel_engine_cs *ring,
+				int bytes)
 {
+	struct intel_ringbuffer *ringbuf = ring->buffer;
 	int ret;
 
-	if (unlikely(ring->tail + bytes > ring->effective_size)) {
+	if (unlikely(ringbuf->tail + bytes > ringbuf->effective_size)) {
 		ret = intel_wrap_ring_buffer(ring);
 		if (unlikely(ret))
 			return ret;
 	}
 
-	if (unlikely(ring->space < bytes)) {
+	if (unlikely(ringbuf->space < bytes)) {
 		ret = ring_wait_for_space(ring, bytes);
 		if (unlikely(ret))
 			return ret;
 	}
 
-	ring->space -= bytes;
 	return 0;
 }
 
-int intel_ring_begin(struct intel_ring_buffer *ring,
+int intel_ring_begin(struct intel_engine_cs *ring,
 		     int num_dwords)
 {
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	int ret;
 
 	ret = i915_gem_check_wedge(&dev_priv->gpu_error,
@@ -1643,15 +1709,42 @@ int intel_ring_begin(struct intel_ring_buffer *ring,
 	if (ret)
 		return ret;
 
+	ret = __intel_ring_prepare(ring, num_dwords * sizeof(uint32_t));
+	if (ret)
+		return ret;
+
 	/* Preallocate the olr before touching the ring */
 	ret = intel_ring_alloc_seqno(ring);
 	if (ret)
 		return ret;
 
-	return __intel_ring_begin(ring, num_dwords * sizeof(uint32_t));
+	ring->buffer->space -= num_dwords * sizeof(uint32_t);
+	return 0;
 }
 
-void intel_ring_init_seqno(struct intel_ring_buffer *ring, u32 seqno)
+/* Align the ring tail to a cacheline boundary */
+int intel_ring_cacheline_align(struct intel_engine_cs *ring)
+{
+	int num_dwords = (ring->buffer->tail & (CACHELINE_BYTES - 1)) / sizeof(uint32_t);
+	int ret;
+
+	if (num_dwords == 0)
+		return 0;
+
+	num_dwords = CACHELINE_BYTES / sizeof(uint32_t) - num_dwords;
+	ret = intel_ring_begin(ring, num_dwords);
+	if (ret)
+		return ret;
+
+	while (num_dwords--)
+		intel_ring_emit(ring, MI_NOOP);
+
+	intel_ring_advance(ring);
+
+	return 0;
+}
+
+void intel_ring_init_seqno(struct intel_engine_cs *ring, u32 seqno)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 
@@ -1668,10 +1761,10 @@ void intel_ring_init_seqno(struct intel_ring_buffer *ring, u32 seqno)
 	ring->hangcheck.seqno = seqno;
 }
 
-static void gen6_bsd_ring_write_tail(struct intel_ring_buffer *ring,
+static void gen6_bsd_ring_write_tail(struct intel_engine_cs *ring,
 				     u32 value)
 {
-	drm_i915_private_t *dev_priv = ring->dev->dev_private;
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 
        /* Every tail move must follow the sequence below */
 
@@ -1701,7 +1794,7 @@ static void gen6_bsd_ring_write_tail(struct intel_ring_buffer *ring,
 		   _MASKED_BIT_DISABLE(GEN6_BSD_SLEEP_MSG_DISABLE));
 }
 
-static int gen6_bsd_ring_flush(struct intel_ring_buffer *ring,
+static int gen6_bsd_ring_flush(struct intel_engine_cs *ring,
 			       u32 invalidate, u32 flush)
 {
 	uint32_t cmd;
@@ -1737,8 +1830,8 @@ static int gen6_bsd_ring_flush(struct intel_ring_buffer *ring,
 }
 
 static int
-gen8_ring_dispatch_execbuffer(struct intel_ring_buffer *ring,
-			      u32 offset, u32 len,
+gen8_ring_dispatch_execbuffer(struct intel_engine_cs *ring,
+			      u64 offset, u32 len,
 			      unsigned flags)
 {
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
@@ -1752,8 +1845,8 @@ gen8_ring_dispatch_execbuffer(struct intel_ring_buffer *ring,
 
 	/* FIXME(BDW): Address space and security selectors. */
 	intel_ring_emit(ring, MI_BATCH_BUFFER_START_GEN8 | (ppgtt<<8));
-	intel_ring_emit(ring, offset);
-	intel_ring_emit(ring, 0);
+	intel_ring_emit(ring, lower_32_bits(offset));
+	intel_ring_emit(ring, upper_32_bits(offset));
 	intel_ring_emit(ring, MI_NOOP);
 	intel_ring_advance(ring);
 
@@ -1761,8 +1854,8 @@ gen8_ring_dispatch_execbuffer(struct intel_ring_buffer *ring,
 }
 
 static int
-hsw_ring_dispatch_execbuffer(struct intel_ring_buffer *ring,
-			      u32 offset, u32 len,
+hsw_ring_dispatch_execbuffer(struct intel_engine_cs *ring,
+			      u64 offset, u32 len,
 			      unsigned flags)
 {
 	int ret;
@@ -1782,8 +1875,8 @@ hsw_ring_dispatch_execbuffer(struct intel_ring_buffer *ring,
 }
 
 static int
-gen6_ring_dispatch_execbuffer(struct intel_ring_buffer *ring,
-			      u32 offset, u32 len,
+gen6_ring_dispatch_execbuffer(struct intel_engine_cs *ring,
+			      u64 offset, u32 len,
 			      unsigned flags)
 {
 	int ret;
@@ -1804,7 +1897,7 @@ gen6_ring_dispatch_execbuffer(struct intel_ring_buffer *ring,
 
 /* Blitter support (SandyBridge+) */
 
-static int gen6_ring_flush(struct intel_ring_buffer *ring,
+static int gen6_ring_flush(struct intel_engine_cs *ring,
 			   u32 invalidate, u32 flush)
 {
 	struct drm_device *dev = ring->dev;
@@ -1838,7 +1931,7 @@ static int gen6_ring_flush(struct intel_ring_buffer *ring,
 	}
 	intel_ring_advance(ring);
 
-	if (IS_GEN7(dev) && flush)
+	if (IS_GEN7(dev) && !invalidate && flush)
 		return gen7_ring_fbc_flush(ring, FBC_REND_CACHE_CLEAN);
 
 	return 0;
@@ -1846,8 +1939,8 @@ static int gen6_ring_flush(struct intel_ring_buffer *ring,
 
 int intel_init_render_ring_buffer(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
 
 	ring->name = "render ring";
 	ring->id = RCS;
@@ -1869,15 +1962,24 @@ int intel_init_render_ring_buffer(struct drm_device *dev)
 		ring->irq_enable_mask = GT_RENDER_USER_INTERRUPT;
 		ring->get_seqno = gen6_ring_get_seqno;
 		ring->set_seqno = ring_set_seqno;
-		ring->sync_to = gen6_ring_sync;
-		ring->semaphore_register[RCS] = MI_SEMAPHORE_SYNC_INVALID;
-		ring->semaphore_register[VCS] = MI_SEMAPHORE_SYNC_RV;
-		ring->semaphore_register[BCS] = MI_SEMAPHORE_SYNC_RB;
-		ring->semaphore_register[VECS] = MI_SEMAPHORE_SYNC_RVE;
-		ring->signal_mbox[RCS] = GEN6_NOSYNC;
-		ring->signal_mbox[VCS] = GEN6_VRSYNC;
-		ring->signal_mbox[BCS] = GEN6_BRSYNC;
-		ring->signal_mbox[VECS] = GEN6_VERSYNC;
+		ring->semaphore.sync_to = gen6_ring_sync;
+		ring->semaphore.signal = gen6_signal;
+		/*
+		 * The current semaphore is only applied on pre-gen8 platform.
+		 * And there is no VCS2 ring on the pre-gen8 platform. So the
+		 * semaphore between RCS and VCS2 is initialized as INVALID.
+		 * Gen8 will initialize the sema between VCS2 and RCS later.
+		 */
+		ring->semaphore.mbox.wait[RCS] = MI_SEMAPHORE_SYNC_INVALID;
+		ring->semaphore.mbox.wait[VCS] = MI_SEMAPHORE_SYNC_RV;
+		ring->semaphore.mbox.wait[BCS] = MI_SEMAPHORE_SYNC_RB;
+		ring->semaphore.mbox.wait[VECS] = MI_SEMAPHORE_SYNC_RVE;
+		ring->semaphore.mbox.wait[VCS2] = MI_SEMAPHORE_SYNC_INVALID;
+		ring->semaphore.mbox.signal[RCS] = GEN6_NOSYNC;
+		ring->semaphore.mbox.signal[VCS] = GEN6_VRSYNC;
+		ring->semaphore.mbox.signal[BCS] = GEN6_BRSYNC;
+		ring->semaphore.mbox.signal[VECS] = GEN6_VERSYNC;
+		ring->semaphore.mbox.signal[VCS2] = GEN6_NOSYNC;
 	} else if (IS_GEN5(dev)) {
 		ring->add_request = pc_render_add_request;
 		ring->flush = gen4_render_ring_flush;
@@ -1931,7 +2033,7 @@ int intel_init_render_ring_buffer(struct drm_device *dev)
 			return -ENOMEM;
 		}
 
-		ret = i915_gem_obj_ggtt_pin(obj, 0, true, false);
+		ret = i915_gem_obj_ggtt_pin(obj, 0, 0);
 		if (ret != 0) {
 			drm_gem_object_unreference(&obj->base);
 			DRM_ERROR("Failed to ping batch bo\n");
@@ -1947,17 +2049,26 @@ int intel_init_render_ring_buffer(struct drm_device *dev)
 
 int intel_render_ring_init_dri(struct drm_device *dev, u64 start, u32 size)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[RCS];
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[RCS];
+	struct intel_ringbuffer *ringbuf = ring->buffer;
 	int ret;
 
+	if (ringbuf == NULL) {
+		ringbuf = kzalloc(sizeof(*ringbuf), GFP_KERNEL);
+		if (!ringbuf)
+			return -ENOMEM;
+		ring->buffer = ringbuf;
+	}
+
 	ring->name = "render ring";
 	ring->id = RCS;
 	ring->mmio_base = RENDER_RING_BASE;
 
 	if (INTEL_INFO(dev)->gen >= 6) {
 		/* non-kms not supported on gen6+ */
-		return -ENODEV;
+		ret = -ENODEV;
+		goto err_ringbuf;
 	}
 
 	/* Note: gem is not supported on gen5/ilk without kms (the corresponding
@@ -1992,31 +2103,39 @@ int intel_render_ring_init_dri(struct drm_device *dev, u64 start, u32 size)
 	INIT_LIST_HEAD(&ring->active_list);
 	INIT_LIST_HEAD(&ring->request_list);
 
-	ring->size = size;
-	ring->effective_size = ring->size;
+	ringbuf->size = size;
+	ringbuf->effective_size = ringbuf->size;
 	if (IS_I830(ring->dev) || IS_845G(ring->dev))
-		ring->effective_size -= 128;
+		ringbuf->effective_size -= 2 * CACHELINE_BYTES;
 
-	ring->virtual_start = ioremap_wc(start, size);
-	if (ring->virtual_start == NULL) {
+	ringbuf->virtual_start = ioremap_wc(start, size);
+	if (ringbuf->virtual_start == NULL) {
 		DRM_ERROR("can not ioremap virtual address for"
 			  " ring buffer\n");
-		return -ENOMEM;
+		ret = -ENOMEM;
+		goto err_ringbuf;
 	}
 
 	if (!I915_NEED_GFX_HWS(dev)) {
 		ret = init_phys_status_page(ring);
 		if (ret)
-			return ret;
+			goto err_vstart;
 	}
 
 	return 0;
+
+err_vstart:
+	iounmap(ringbuf->virtual_start);
+err_ringbuf:
+	kfree(ringbuf);
+	ring->buffer = NULL;
+	return ret;
 }
 
 int intel_init_bsd_ring_buffer(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[VCS];
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[VCS];
 
 	ring->name = "bsd ring";
 	ring->id = VCS;
@@ -2045,15 +2164,24 @@ int intel_init_bsd_ring_buffer(struct drm_device *dev)
 			ring->dispatch_execbuffer =
 				gen6_ring_dispatch_execbuffer;
 		}
-		ring->sync_to = gen6_ring_sync;
-		ring->semaphore_register[RCS] = MI_SEMAPHORE_SYNC_VR;
-		ring->semaphore_register[VCS] = MI_SEMAPHORE_SYNC_INVALID;
-		ring->semaphore_register[BCS] = MI_SEMAPHORE_SYNC_VB;
-		ring->semaphore_register[VECS] = MI_SEMAPHORE_SYNC_VVE;
-		ring->signal_mbox[RCS] = GEN6_RVSYNC;
-		ring->signal_mbox[VCS] = GEN6_NOSYNC;
-		ring->signal_mbox[BCS] = GEN6_BVSYNC;
-		ring->signal_mbox[VECS] = GEN6_VEVSYNC;
+		ring->semaphore.sync_to = gen6_ring_sync;
+		ring->semaphore.signal = gen6_signal;
+		/*
+		 * The current semaphore is only applied on pre-gen8 platform.
+		 * And there is no VCS2 ring on the pre-gen8 platform. So the
+		 * semaphore between VCS and VCS2 is initialized as INVALID.
+		 * Gen8 will initialize the sema between VCS2 and VCS later.
+		 */
+		ring->semaphore.mbox.wait[RCS] = MI_SEMAPHORE_SYNC_VR;
+		ring->semaphore.mbox.wait[VCS] = MI_SEMAPHORE_SYNC_INVALID;
+		ring->semaphore.mbox.wait[BCS] = MI_SEMAPHORE_SYNC_VB;
+		ring->semaphore.mbox.wait[VECS] = MI_SEMAPHORE_SYNC_VVE;
+		ring->semaphore.mbox.wait[VCS2] = MI_SEMAPHORE_SYNC_INVALID;
+		ring->semaphore.mbox.signal[RCS] = GEN6_RVSYNC;
+		ring->semaphore.mbox.signal[VCS] = GEN6_NOSYNC;
+		ring->semaphore.mbox.signal[BCS] = GEN6_BVSYNC;
+		ring->semaphore.mbox.signal[VECS] = GEN6_VEVSYNC;
+		ring->semaphore.mbox.signal[VCS2] = GEN6_NOSYNC;
 	} else {
 		ring->mmio_base = BSD_RING_BASE;
 		ring->flush = bsd_ring_flush;
@@ -2076,10 +2204,63 @@ int intel_init_bsd_ring_buffer(struct drm_device *dev)
 	return intel_init_ring_buffer(dev, ring);
 }
 
+/**
+ * Initialize the second BSD ring for Broadwell GT3.
+ * It is noted that this only exists on Broadwell GT3.
+ */
+int intel_init_bsd2_ring_buffer(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[VCS2];
+
+	if ((INTEL_INFO(dev)->gen != 8)) {
+		DRM_ERROR("No dual-BSD ring on non-BDW machine\n");
+		return -EINVAL;
+	}
+
+	ring->name = "bds2_ring";
+	ring->id = VCS2;
+
+	ring->write_tail = ring_write_tail;
+	ring->mmio_base = GEN8_BSD2_RING_BASE;
+	ring->flush = gen6_bsd_ring_flush;
+	ring->add_request = gen6_add_request;
+	ring->get_seqno = gen6_ring_get_seqno;
+	ring->set_seqno = ring_set_seqno;
+	ring->irq_enable_mask =
+			GT_RENDER_USER_INTERRUPT << GEN8_VCS2_IRQ_SHIFT;
+	ring->irq_get = gen8_ring_get_irq;
+	ring->irq_put = gen8_ring_put_irq;
+	ring->dispatch_execbuffer =
+			gen8_ring_dispatch_execbuffer;
+	ring->semaphore.sync_to = gen6_ring_sync;
+	ring->semaphore.signal = gen6_signal;
+	/*
+	 * The current semaphore is only applied on the pre-gen8. And there
+	 * is no bsd2 ring on the pre-gen8. So now the semaphore_register
+	 * between VCS2 and other ring is initialized as invalid.
+	 * Gen8 will initialize the sema between VCS2 and other ring later.
+	 */
+	ring->semaphore.mbox.wait[RCS] = MI_SEMAPHORE_SYNC_INVALID;
+	ring->semaphore.mbox.wait[VCS] = MI_SEMAPHORE_SYNC_INVALID;
+	ring->semaphore.mbox.wait[BCS] = MI_SEMAPHORE_SYNC_INVALID;
+	ring->semaphore.mbox.wait[VECS] = MI_SEMAPHORE_SYNC_INVALID;
+	ring->semaphore.mbox.wait[VCS2] = MI_SEMAPHORE_SYNC_INVALID;
+	ring->semaphore.mbox.signal[RCS] = GEN6_NOSYNC;
+	ring->semaphore.mbox.signal[VCS] = GEN6_NOSYNC;
+	ring->semaphore.mbox.signal[BCS] = GEN6_NOSYNC;
+	ring->semaphore.mbox.signal[VECS] = GEN6_NOSYNC;
+	ring->semaphore.mbox.signal[VCS2] = GEN6_NOSYNC;
+
+	ring->init = init_ring_common;
+
+	return intel_init_ring_buffer(dev, ring);
+}
+
 int intel_init_blt_ring_buffer(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[BCS];
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[BCS];
 
 	ring->name = "blitter ring";
 	ring->id = BCS;
@@ -2102,15 +2283,24 @@ int intel_init_blt_ring_buffer(struct drm_device *dev)
 		ring->irq_put = gen6_ring_put_irq;
 		ring->dispatch_execbuffer = gen6_ring_dispatch_execbuffer;
 	}
-	ring->sync_to = gen6_ring_sync;
-	ring->semaphore_register[RCS] = MI_SEMAPHORE_SYNC_BR;
-	ring->semaphore_register[VCS] = MI_SEMAPHORE_SYNC_BV;
-	ring->semaphore_register[BCS] = MI_SEMAPHORE_SYNC_INVALID;
-	ring->semaphore_register[VECS] = MI_SEMAPHORE_SYNC_BVE;
-	ring->signal_mbox[RCS] = GEN6_RBSYNC;
-	ring->signal_mbox[VCS] = GEN6_VBSYNC;
-	ring->signal_mbox[BCS] = GEN6_NOSYNC;
-	ring->signal_mbox[VECS] = GEN6_VEBSYNC;
+	ring->semaphore.sync_to = gen6_ring_sync;
+	ring->semaphore.signal = gen6_signal;
+	/*
+	 * The current semaphore is only applied on pre-gen8 platform. And
+	 * there is no VCS2 ring on the pre-gen8 platform. So the semaphore
+	 * between BCS and VCS2 is initialized as INVALID.
+	 * Gen8 will initialize the sema between BCS and VCS2 later.
+	 */
+	ring->semaphore.mbox.wait[RCS] = MI_SEMAPHORE_SYNC_BR;
+	ring->semaphore.mbox.wait[VCS] = MI_SEMAPHORE_SYNC_BV;
+	ring->semaphore.mbox.wait[BCS] = MI_SEMAPHORE_SYNC_INVALID;
+	ring->semaphore.mbox.wait[VECS] = MI_SEMAPHORE_SYNC_BVE;
+	ring->semaphore.mbox.wait[VCS2] = MI_SEMAPHORE_SYNC_INVALID;
+	ring->semaphore.mbox.signal[RCS] = GEN6_RBSYNC;
+	ring->semaphore.mbox.signal[VCS] = GEN6_VBSYNC;
+	ring->semaphore.mbox.signal[BCS] = GEN6_NOSYNC;
+	ring->semaphore.mbox.signal[VECS] = GEN6_VEBSYNC;
+	ring->semaphore.mbox.signal[VCS2] = GEN6_NOSYNC;
 	ring->init = init_ring_common;
 
 	return intel_init_ring_buffer(dev, ring);
@@ -2118,8 +2308,8 @@ int intel_init_blt_ring_buffer(struct drm_device *dev)
 
 int intel_init_vebox_ring_buffer(struct drm_device *dev)
 {
-	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring = &dev_priv->ring[VECS];
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_engine_cs *ring = &dev_priv->ring[VECS];
 
 	ring->name = "video enhancement ring";
 	ring->id = VECS;
@@ -2143,22 +2333,25 @@ int intel_init_vebox_ring_buffer(struct drm_device *dev)
 		ring->irq_put = hsw_vebox_put_irq;
 		ring->dispatch_execbuffer = gen6_ring_dispatch_execbuffer;
 	}
-	ring->sync_to = gen6_ring_sync;
-	ring->semaphore_register[RCS] = MI_SEMAPHORE_SYNC_VER;
-	ring->semaphore_register[VCS] = MI_SEMAPHORE_SYNC_VEV;
-	ring->semaphore_register[BCS] = MI_SEMAPHORE_SYNC_VEB;
-	ring->semaphore_register[VECS] = MI_SEMAPHORE_SYNC_INVALID;
-	ring->signal_mbox[RCS] = GEN6_RVESYNC;
-	ring->signal_mbox[VCS] = GEN6_VVESYNC;
-	ring->signal_mbox[BCS] = GEN6_BVESYNC;
-	ring->signal_mbox[VECS] = GEN6_NOSYNC;
+	ring->semaphore.sync_to = gen6_ring_sync;
+	ring->semaphore.signal = gen6_signal;
+	ring->semaphore.mbox.wait[RCS] = MI_SEMAPHORE_SYNC_VER;
+	ring->semaphore.mbox.wait[VCS] = MI_SEMAPHORE_SYNC_VEV;
+	ring->semaphore.mbox.wait[BCS] = MI_SEMAPHORE_SYNC_VEB;
+	ring->semaphore.mbox.wait[VECS] = MI_SEMAPHORE_SYNC_INVALID;
+	ring->semaphore.mbox.wait[VCS2] = MI_SEMAPHORE_SYNC_INVALID;
+	ring->semaphore.mbox.signal[RCS] = GEN6_RVESYNC;
+	ring->semaphore.mbox.signal[VCS] = GEN6_VVESYNC;
+	ring->semaphore.mbox.signal[BCS] = GEN6_BVESYNC;
+	ring->semaphore.mbox.signal[VECS] = GEN6_NOSYNC;
+	ring->semaphore.mbox.signal[VCS2] = GEN6_NOSYNC;
 	ring->init = init_ring_common;
 
 	return intel_init_ring_buffer(dev, ring);
 }
 
 int
-intel_ring_flush_all_caches(struct intel_ring_buffer *ring)
+intel_ring_flush_all_caches(struct intel_engine_cs *ring)
 {
 	int ret;
 
@@ -2176,7 +2369,7 @@ intel_ring_flush_all_caches(struct intel_ring_buffer *ring)
 }
 
 int
-intel_ring_invalidate_all_caches(struct intel_ring_buffer *ring)
+intel_ring_invalidate_all_caches(struct intel_engine_cs *ring)
 {
 	uint32_t flush_domains;
 	int ret;
@@ -2194,3 +2387,19 @@ intel_ring_invalidate_all_caches(struct intel_ring_buffer *ring)
 	ring->gpu_caches_dirty = false;
 	return 0;
 }
+
+void
+intel_stop_ring_buffer(struct intel_engine_cs *ring)
+{
+	int ret;
+
+	if (!intel_ring_initialized(ring))
+		return;
+
+	ret = intel_ring_idle(ring);
+	if (ret && !i915_reset_in_progress(&to_i915(ring->dev)->gpu_error))
+		DRM_ERROR("failed to quiesce %s whilst cleaning up: %d\n",
+			  ring->name, ret);
+
+	stop_ring(ring);
+}
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 71a73f4fe25..e72017bdcd7 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -1,6 +1,10 @@
 #ifndef _INTEL_RINGBUFFER_H_
 #define _INTEL_RINGBUFFER_H_
 
+#include <linux/hashtable.h>
+
+#define I915_CMD_HASH_ORDER 9
+
 /*
  * Gen2 BSpec "1. Programming Environment" / 1.4.4.6 "Ring Buffer Use"
  * Gen3 BSpec "vol1c Memory Interface Functions" / 2.3.4.5 "Ring Buffer Use"
@@ -33,6 +37,9 @@ struct  intel_hw_status_page {
 #define I915_READ_IMR(ring) I915_READ(RING_IMR((ring)->mmio_base))
 #define I915_WRITE_IMR(ring, val) I915_WRITE(RING_IMR((ring)->mmio_base), val)
 
+#define I915_READ_MODE(ring) I915_READ(RING_MI_MODE((ring)->mmio_base))
+#define I915_WRITE_MODE(ring, val) I915_WRITE(RING_MI_MODE((ring)->mmio_base), val)
+
 enum intel_ring_hangcheck_action {
 	HANGCHECK_IDLE = 0,
 	HANGCHECK_WAIT,
@@ -41,84 +48,103 @@ enum intel_ring_hangcheck_action {
 	HANGCHECK_HUNG,
 };
 
+#define HANGCHECK_SCORE_RING_HUNG 31
+
 struct intel_ring_hangcheck {
-	bool deadlock;
+	u64 acthd;
 	u32 seqno;
-	u32 acthd;
 	int score;
 	enum intel_ring_hangcheck_action action;
+	int deadlock;
 };
 
-struct  intel_ring_buffer {
+struct intel_ringbuffer {
+	struct drm_i915_gem_object *obj;
+	void __iomem *virtual_start;
+
+	u32 head;
+	u32 tail;
+	int space;
+	int size;
+	int effective_size;
+
+	/** We track the position of the requests in the ring buffer, and
+	 * when each is retired we increment last_retired_head as the GPU
+	 * must have finished processing the request and so we know we
+	 * can advance the ringbuffer up to that position.
+	 *
+	 * last_retired_head is set to -1 after the value is consumed so
+	 * we can detect new retirements.
+	 */
+	u32 last_retired_head;
+};
+
+struct  intel_engine_cs {
 	const char	*name;
 	enum intel_ring_id {
 		RCS = 0x0,
 		VCS,
 		BCS,
 		VECS,
+		VCS2
 	} id;
-#define I915_NUM_RINGS 4
+#define I915_NUM_RINGS 5
+#define LAST_USER_RING (VECS + 1)
 	u32		mmio_base;
-	void		__iomem *virtual_start;
 	struct		drm_device *dev;
-	struct		drm_i915_gem_object *obj;
+	struct intel_ringbuffer *buffer;
 
-	u32		head;
-	u32		tail;
-	int		space;
-	int		size;
-	int		effective_size;
 	struct intel_hw_status_page status_page;
 
-	/** We track the position of the requests in the ring buffer, and
-	 * when each is retired we increment last_retired_head as the GPU
-	 * must have finished processing the request and so we know we
-	 * can advance the ringbuffer up to that position.
-	 *
-	 * last_retired_head is set to -1 after the value is consumed so
-	 * we can detect new retirements.
-	 */
-	u32		last_retired_head;
-
 	unsigned irq_refcount; /* protected by dev_priv->irq_lock */
 	u32		irq_enable_mask;	/* bitmask to enable ring interrupt */
 	u32		trace_irq_seqno;
-	u32		sync_seqno[I915_NUM_RINGS-1];
-	bool __must_check (*irq_get)(struct intel_ring_buffer *ring);
-	void		(*irq_put)(struct intel_ring_buffer *ring);
+	bool __must_check (*irq_get)(struct intel_engine_cs *ring);
+	void		(*irq_put)(struct intel_engine_cs *ring);
 
-	int		(*init)(struct intel_ring_buffer *ring);
+	int		(*init)(struct intel_engine_cs *ring);
 
-	void		(*write_tail)(struct intel_ring_buffer *ring,
+	void		(*write_tail)(struct intel_engine_cs *ring,
 				      u32 value);
-	int __must_check (*flush)(struct intel_ring_buffer *ring,
+	int __must_check (*flush)(struct intel_engine_cs *ring,
 				  u32	invalidate_domains,
 				  u32	flush_domains);
-	int		(*add_request)(struct intel_ring_buffer *ring);
+	int		(*add_request)(struct intel_engine_cs *ring);
 	/* Some chipsets are not quite as coherent as advertised and need
 	 * an expensive kick to force a true read of the up-to-date seqno.
 	 * However, the up-to-date seqno is not always required and the last
 	 * seen value is good enough. Note that the seqno will always be
 	 * monotonic, even if not coherent.
 	 */
-	u32		(*get_seqno)(struct intel_ring_buffer *ring,
+	u32		(*get_seqno)(struct intel_engine_cs *ring,
 				     bool lazy_coherency);
-	void		(*set_seqno)(struct intel_ring_buffer *ring,
+	void		(*set_seqno)(struct intel_engine_cs *ring,
 				     u32 seqno);
-	int		(*dispatch_execbuffer)(struct intel_ring_buffer *ring,
-					       u32 offset, u32 length,
+	int		(*dispatch_execbuffer)(struct intel_engine_cs *ring,
+					       u64 offset, u32 length,
 					       unsigned flags);
 #define I915_DISPATCH_SECURE 0x1
 #define I915_DISPATCH_PINNED 0x2
-	void		(*cleanup)(struct intel_ring_buffer *ring);
-	int		(*sync_to)(struct intel_ring_buffer *ring,
-				   struct intel_ring_buffer *to,
-				   u32 seqno);
+	void		(*cleanup)(struct intel_engine_cs *ring);
 
-	/* our mbox written by others */
-	u32		semaphore_register[I915_NUM_RINGS];
-	/* mboxes this ring signals to */
-	u32		signal_mbox[I915_NUM_RINGS];
+	struct {
+		u32	sync_seqno[I915_NUM_RINGS-1];
+
+		struct {
+			/* our mbox written by others */
+			u32		wait[I915_NUM_RINGS];
+			/* mboxes this ring signals to */
+			u32		signal[I915_NUM_RINGS];
+		} mbox;
+
+		/* AKA wait() */
+		int	(*sync_to)(struct intel_engine_cs *ring,
+				   struct intel_engine_cs *to,
+				   u32 seqno);
+		int	(*signal)(struct intel_engine_cs *signaller,
+				  /* num_dwords needed by caller */
+				  unsigned int num_dwords);
+	} semaphore;
 
 	/**
 	 * List of objects currently involved in rendering from the
@@ -148,12 +174,8 @@ struct  intel_ring_buffer {
 
 	wait_queue_head_t irq_queue;
 
-	/**
-	 * Do an explicit TLB flush before MI_SET_CONTEXT
-	 */
-	bool itlb_before_ctx_switch;
-	struct i915_hw_context *default_context;
-	struct i915_hw_context *last_context;
+	struct intel_context *default_context;
+	struct intel_context *last_context;
 
 	struct intel_ring_hangcheck hangcheck;
 
@@ -162,23 +184,56 @@ struct  intel_ring_buffer {
 		u32 gtt_offset;
 		volatile u32 *cpu_page;
 	} scratch;
+
+	bool needs_cmd_parser;
+
+	/*
+	 * Table of commands the command parser needs to know about
+	 * for this ring.
+	 */
+	DECLARE_HASHTABLE(cmd_hash, I915_CMD_HASH_ORDER);
+
+	/*
+	 * Table of registers allowed in commands that read/write registers.
+	 */
+	const u32 *reg_table;
+	int reg_count;
+
+	/*
+	 * Table of registers allowed in commands that read/write registers, but
+	 * only from the DRM master.
+	 */
+	const u32 *master_reg_table;
+	int master_reg_count;
+
+	/*
+	 * Returns the bitmask for the length field of the specified command.
+	 * Return 0 for an unrecognized/invalid command.
+	 *
+	 * If the command parser finds an entry for a command in the ring's
+	 * cmd_tables, it gets the command's length based on the table entry.
+	 * If not, it calls this function to determine the per-ring length field
+	 * encoding for the command (i.e. certain opcode ranges use certain bits
+	 * to encode the command length in the header).
+	 */
+	u32 (*get_cmd_length_mask)(u32 cmd_header);
 };
 
 static inline bool
-intel_ring_initialized(struct intel_ring_buffer *ring)
+intel_ring_initialized(struct intel_engine_cs *ring)
 {
-	return ring->obj != NULL;
+	return ring->buffer && ring->buffer->obj;
 }
 
 static inline unsigned
-intel_ring_flag(struct intel_ring_buffer *ring)
+intel_ring_flag(struct intel_engine_cs *ring)
 {
 	return 1 << ring->id;
 }
 
 static inline u32
-intel_ring_sync_index(struct intel_ring_buffer *ring,
-		      struct intel_ring_buffer *other)
+intel_ring_sync_index(struct intel_engine_cs *ring,
+		      struct intel_engine_cs *other)
 {
 	int idx;
 
@@ -196,7 +251,7 @@ intel_ring_sync_index(struct intel_ring_buffer *ring,
 }
 
 static inline u32
-intel_read_status_page(struct intel_ring_buffer *ring,
+intel_read_status_page(struct intel_engine_cs *ring,
 		       int reg)
 {
 	/* Ensure that the compiler doesn't optimize away the load. */
@@ -205,7 +260,7 @@ intel_read_status_page(struct intel_ring_buffer *ring,
 }
 
 static inline void
-intel_write_status_page(struct intel_ring_buffer *ring,
+intel_write_status_page(struct intel_engine_cs *ring,
 			int reg, u32 value)
 {
 	ring->status_page.page_addr[reg] = value;
@@ -230,46 +285,51 @@ intel_write_status_page(struct intel_ring_buffer *ring,
 #define I915_GEM_HWS_SCRATCH_INDEX	0x30
 #define I915_GEM_HWS_SCRATCH_ADDR (I915_GEM_HWS_SCRATCH_INDEX << MI_STORE_DWORD_INDEX_SHIFT)
 
-void intel_cleanup_ring_buffer(struct intel_ring_buffer *ring);
+void intel_stop_ring_buffer(struct intel_engine_cs *ring);
+void intel_cleanup_ring_buffer(struct intel_engine_cs *ring);
 
-int __must_check intel_ring_begin(struct intel_ring_buffer *ring, int n);
-static inline void intel_ring_emit(struct intel_ring_buffer *ring,
+int __must_check intel_ring_begin(struct intel_engine_cs *ring, int n);
+int __must_check intel_ring_cacheline_align(struct intel_engine_cs *ring);
+static inline void intel_ring_emit(struct intel_engine_cs *ring,
 				   u32 data)
 {
-	iowrite32(data, ring->virtual_start + ring->tail);
-	ring->tail += 4;
+	struct intel_ringbuffer *ringbuf = ring->buffer;
+	iowrite32(data, ringbuf->virtual_start + ringbuf->tail);
+	ringbuf->tail += 4;
 }
-static inline void intel_ring_advance(struct intel_ring_buffer *ring)
+static inline void intel_ring_advance(struct intel_engine_cs *ring)
 {
-	ring->tail &= ring->size - 1;
+	struct intel_ringbuffer *ringbuf = ring->buffer;
+	ringbuf->tail &= ringbuf->size - 1;
 }
-void __intel_ring_advance(struct intel_ring_buffer *ring);
+void __intel_ring_advance(struct intel_engine_cs *ring);
 
-int __must_check intel_ring_idle(struct intel_ring_buffer *ring);
-void intel_ring_init_seqno(struct intel_ring_buffer *ring, u32 seqno);
-int intel_ring_flush_all_caches(struct intel_ring_buffer *ring);
-int intel_ring_invalidate_all_caches(struct intel_ring_buffer *ring);
+int __must_check intel_ring_idle(struct intel_engine_cs *ring);
+void intel_ring_init_seqno(struct intel_engine_cs *ring, u32 seqno);
+int intel_ring_flush_all_caches(struct intel_engine_cs *ring);
+int intel_ring_invalidate_all_caches(struct intel_engine_cs *ring);
 
 int intel_init_render_ring_buffer(struct drm_device *dev);
 int intel_init_bsd_ring_buffer(struct drm_device *dev);
+int intel_init_bsd2_ring_buffer(struct drm_device *dev);
 int intel_init_blt_ring_buffer(struct drm_device *dev);
 int intel_init_vebox_ring_buffer(struct drm_device *dev);
 
-u32 intel_ring_get_active_head(struct intel_ring_buffer *ring);
-void intel_ring_setup_status_page(struct intel_ring_buffer *ring);
+u64 intel_ring_get_active_head(struct intel_engine_cs *ring);
+void intel_ring_setup_status_page(struct intel_engine_cs *ring);
 
-static inline u32 intel_ring_get_tail(struct intel_ring_buffer *ring)
+static inline u32 intel_ring_get_tail(struct intel_engine_cs *ring)
 {
-	return ring->tail;
+	return ring->buffer->tail;
 }
 
-static inline u32 intel_ring_get_seqno(struct intel_ring_buffer *ring)
+static inline u32 intel_ring_get_seqno(struct intel_engine_cs *ring)
 {
 	BUG_ON(ring->outstanding_lazy_seqno == 0);
 	return ring->outstanding_lazy_seqno;
 }
 
-static inline void i915_trace_irq_get(struct intel_ring_buffer *ring, u32 seqno)
+static inline void i915_trace_irq_get(struct intel_engine_cs *ring, u32 seqno)
 {
 	if (ring->trace_irq_seqno == 0 && ring->irq_get(ring))
 		ring->trace_irq_seqno = seqno;
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index a583e8f718a..20375cc7f82 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -413,23 +413,34 @@ static const struct _sdvo_cmd_name {
 static void intel_sdvo_debug_write(struct intel_sdvo *intel_sdvo, u8 cmd,
 				   const void *args, int args_len)
 {
-	int i;
+	int i, pos = 0;
+#define BUF_LEN 256
+	char buffer[BUF_LEN];
+
+#define BUF_PRINT(args...) \
+	pos += snprintf(buffer + pos, max_t(int, BUF_LEN - pos, 0), args)
 
-	DRM_DEBUG_KMS("%s: W: %02X ",
-				SDVO_NAME(intel_sdvo), cmd);
-	for (i = 0; i < args_len; i++)
-		DRM_LOG_KMS("%02X ", ((u8 *)args)[i]);
-	for (; i < 8; i++)
-		DRM_LOG_KMS("   ");
+
+	for (i = 0; i < args_len; i++) {
+		BUF_PRINT("%02X ", ((u8 *)args)[i]);
+	}
+	for (; i < 8; i++) {
+		BUF_PRINT("   ");
+	}
 	for (i = 0; i < ARRAY_SIZE(sdvo_cmd_names); i++) {
 		if (cmd == sdvo_cmd_names[i].cmd) {
-			DRM_LOG_KMS("(%s)", sdvo_cmd_names[i].name);
+			BUF_PRINT("(%s)", sdvo_cmd_names[i].name);
 			break;
 		}
 	}
-	if (i == ARRAY_SIZE(sdvo_cmd_names))
-		DRM_LOG_KMS("(%02X)", cmd);
-	DRM_LOG_KMS("\n");
+	if (i == ARRAY_SIZE(sdvo_cmd_names)) {
+		BUF_PRINT("(%02X)", cmd);
+	}
+	BUG_ON(pos >= BUF_LEN - 1);
+#undef BUF_PRINT
+#undef BUF_LEN
+
+	DRM_DEBUG_KMS("%s: W: %02X %s\n", SDVO_NAME(intel_sdvo), cmd, buffer);
 }
 
 static const char *cmd_status_names[] = {
@@ -512,9 +523,10 @@ static bool intel_sdvo_read_response(struct intel_sdvo *intel_sdvo,
 {
 	u8 retry = 15; /* 5 quick checks, followed by 10 long checks */
 	u8 status;
-	int i;
+	int i, pos = 0;
+#define BUF_LEN 256
+	char buffer[BUF_LEN];
 
-	DRM_DEBUG_KMS("%s: R: ", SDVO_NAME(intel_sdvo));
 
 	/*
 	 * The documentation states that all commands will be
@@ -551,10 +563,13 @@ static bool intel_sdvo_read_response(struct intel_sdvo *intel_sdvo,
 			goto log_fail;
 	}
 
+#define BUF_PRINT(args...) \
+	pos += snprintf(buffer + pos, max_t(int, BUF_LEN - pos, 0), args)
+
 	if (status <= SDVO_CMD_STATUS_SCALING_NOT_SUPP)
-		DRM_LOG_KMS("(%s)", cmd_status_names[status]);
+		BUF_PRINT("(%s)", cmd_status_names[status]);
 	else
-		DRM_LOG_KMS("(??? %d)", status);
+		BUF_PRINT("(??? %d)", status);
 
 	if (status != SDVO_CMD_STATUS_SUCCESS)
 		goto log_fail;
@@ -565,13 +580,17 @@ static bool intel_sdvo_read_response(struct intel_sdvo *intel_sdvo,
 					  SDVO_I2C_RETURN_0 + i,
 					  &((u8 *)response)[i]))
 			goto log_fail;
-		DRM_LOG_KMS(" %02X", ((u8 *)response)[i]);
+		BUF_PRINT(" %02X", ((u8 *)response)[i]);
 	}
-	DRM_LOG_KMS("\n");
+	BUG_ON(pos >= BUF_LEN - 1);
+#undef BUF_PRINT
+#undef BUF_LEN
+
+	DRM_DEBUG_KMS("%s: R: %s\n", SDVO_NAME(intel_sdvo), buffer);
 	return true;
 
 log_fail:
-	DRM_LOG_KMS("... failed\n");
+	DRM_DEBUG_KMS("%s: R: ... failed\n", SDVO_NAME(intel_sdvo));
 	return false;
 }
 
@@ -933,7 +952,7 @@ static void intel_sdvo_dump_hdmi_buf(struct intel_sdvo *intel_sdvo)
 
 static bool intel_sdvo_write_infoframe(struct intel_sdvo *intel_sdvo,
 				       unsigned if_index, uint8_t tx_rate,
-				       uint8_t *data, unsigned length)
+				       const uint8_t *data, unsigned length)
 {
 	uint8_t set_buf_index[2] = { if_index, 0 };
 	uint8_t hbuf_size, tmp[8];
@@ -1134,20 +1153,21 @@ static bool intel_sdvo_compute_config(struct intel_encoder *encoder,
 	pipe_config->pixel_multiplier =
 		intel_sdvo_get_pixel_multiplier(adjusted_mode);
 
+	pipe_config->has_hdmi_sink = intel_sdvo->has_hdmi_monitor;
+
 	if (intel_sdvo->color_range_auto) {
 		/* See CEA-861-E - 5.1 Default Encoding Parameters */
 		/* FIXME: This bit is only valid when using TMDS encoding and 8
 		 * bit per color mode. */
-		if (intel_sdvo->has_hdmi_monitor &&
+		if (pipe_config->has_hdmi_sink &&
 		    drm_match_cea_mode(adjusted_mode) > 1)
-			intel_sdvo->color_range = HDMI_COLOR_RANGE_16_235;
-		else
-			intel_sdvo->color_range = 0;
+			pipe_config->limited_color_range = true;
+	} else {
+		if (pipe_config->has_hdmi_sink &&
+		    intel_sdvo->color_range == HDMI_COLOR_RANGE_16_235)
+			pipe_config->limited_color_range = true;
 	}
 
-	if (intel_sdvo->color_range)
-		pipe_config->limited_color_range = true;
-
 	/* Clock computation needs to happen after pixel multiplier. */
 	if (intel_sdvo->is_tv)
 		i9xx_adjust_sdvo_tv_clock(pipe_config);
@@ -1155,7 +1175,7 @@ static bool intel_sdvo_compute_config(struct intel_encoder *encoder,
 	return true;
 }
 
-static void intel_sdvo_mode_set(struct intel_encoder *intel_encoder)
+static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder)
 {
 	struct drm_device *dev = intel_encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1204,7 +1224,7 @@ static void intel_sdvo_mode_set(struct intel_encoder *intel_encoder)
 	if (!intel_sdvo_set_target_input(intel_sdvo))
 		return;
 
-	if (intel_sdvo->has_hdmi_monitor) {
+	if (crtc->config.has_hdmi_sink) {
 		intel_sdvo_set_encode(intel_sdvo, SDVO_ENCODE_HDMI);
 		intel_sdvo_set_colorimetry(intel_sdvo,
 					   SDVO_COLORIMETRY_RGB256);
@@ -1239,8 +1259,8 @@ static void intel_sdvo_mode_set(struct intel_encoder *intel_encoder)
 		/* The real mode polarity is set by the SDVO commands, using
 		 * struct intel_sdvo_dtd. */
 		sdvox = SDVO_VSYNC_ACTIVE_HIGH | SDVO_HSYNC_ACTIVE_HIGH;
-		if (!HAS_PCH_SPLIT(dev) && intel_sdvo->is_hdmi)
-			sdvox |= intel_sdvo->color_range;
+		if (!HAS_PCH_SPLIT(dev) && crtc->config.limited_color_range)
+			sdvox |= HDMI_COLOR_RANGE_16_235;
 		if (INTEL_INFO(dev)->gen < 5)
 			sdvox |= SDVO_BORDER_ENABLE;
 	} else {
@@ -1330,6 +1350,8 @@ static void intel_sdvo_get_config(struct intel_encoder *encoder,
 	u8 val;
 	bool ret;
 
+	sdvox = I915_READ(intel_sdvo->sdvo_reg);
+
 	ret = intel_sdvo_get_input_timing(intel_sdvo, &dtd);
 	if (!ret) {
 		/* Some sdvo encoders are not spec compliant and don't
@@ -1358,13 +1380,14 @@ static void intel_sdvo_get_config(struct intel_encoder *encoder,
 	 * other platfroms.
 	 */
 	if (IS_I915G(dev) || IS_I915GM(dev)) {
-		sdvox = I915_READ(intel_sdvo->sdvo_reg);
 		pipe_config->pixel_multiplier =
 			((sdvox & SDVO_PORT_MULTIPLY_MASK)
 			 >> SDVO_PORT_MULTIPLY_SHIFT) + 1;
 	}
 
-	dotclock = pipe_config->port_clock / pipe_config->pixel_multiplier;
+	dotclock = pipe_config->port_clock;
+	if (pipe_config->pixel_multiplier)
+		dotclock /= pipe_config->pixel_multiplier;
 
 	if (HAS_PCH_SPLIT(dev))
 		ironlake_check_encoder_dotclock(pipe_config, dotclock);
@@ -1387,6 +1410,15 @@ static void intel_sdvo_get_config(struct intel_encoder *encoder,
 		}
 	}
 
+	if (sdvox & HDMI_COLOR_RANGE_16_235)
+		pipe_config->limited_color_range = true;
+
+	if (intel_sdvo_get_value(intel_sdvo, SDVO_CMD_GET_ENCODE,
+				 &val, 1)) {
+		if (val == SDVO_ENCODE_HDMI)
+			pipe_config->has_hdmi_sink = true;
+	}
+
 	WARN(encoder_pixel_multiplier != pipe_config->pixel_multiplier,
 	     "SDVO pixel multiplier mismatch, port: %i, encoder: %i\n",
 	     pipe_config->pixel_multiplier, encoder_pixel_multiplier);
@@ -1442,7 +1474,7 @@ static void intel_enable_sdvo(struct intel_encoder *encoder)
 	u32 temp;
 	bool input1, input2;
 	int i;
-	u8 status;
+	bool success;
 
 	temp = I915_READ(intel_sdvo->sdvo_reg);
 	if ((temp & SDVO_ENABLE) == 0) {
@@ -1456,12 +1488,12 @@ static void intel_enable_sdvo(struct intel_encoder *encoder)
 	for (i = 0; i < 2; i++)
 		intel_wait_for_vblank(dev, intel_crtc->pipe);
 
-	status = intel_sdvo_get_trained_inputs(intel_sdvo, &input1, &input2);
+	success = intel_sdvo_get_trained_inputs(intel_sdvo, &input1, &input2);
 	/* Warn if the device reported failure to sync.
 	 * A lot of SDVO devices fail to notify of sync, but it's
 	 * a given it the status is a success, we succeeded.
 	 */
-	if (status == SDVO_CMD_STATUS_SUCCESS && !input1) {
+	if (success && !input1) {
 		DRM_DEBUG_KMS("First %s output reported failure to "
 				"sync\n", SDVO_NAME(intel_sdvo));
 	}
@@ -1517,8 +1549,9 @@ static void intel_sdvo_dpms(struct drm_connector *connector, int mode)
 	intel_modeset_check_state(connector->dev);
 }
 
-static int intel_sdvo_mode_valid(struct drm_connector *connector,
-				 struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_sdvo_mode_valid(struct drm_connector *connector,
+		      struct drm_display_mode *mode)
 {
 	struct intel_sdvo *intel_sdvo = intel_attached_sdvo(connector);
 
@@ -1712,7 +1745,7 @@ intel_sdvo_detect(struct drm_connector *connector, bool force)
 	enum drm_connector_status ret;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector));
+		      connector->base.id, connector->name);
 
 	if (!intel_sdvo_get_value(intel_sdvo,
 				  SDVO_CMD_GET_ATTACHED_DISPLAYS,
@@ -1774,7 +1807,7 @@ static void intel_sdvo_get_ddc_modes(struct drm_connector *connector)
 	struct edid *edid;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector));
+		      connector->base.id, connector->name);
 
 	/* set the bus switch and get the modes */
 	edid = intel_sdvo_get_edid(connector);
@@ -1872,7 +1905,7 @@ static void intel_sdvo_get_tv_modes(struct drm_connector *connector)
 	int i;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector));
+		      connector->base.id, connector->name);
 
 	/* Read the list of supported input resolutions for the selected TV
 	 * format.
@@ -1909,7 +1942,7 @@ static void intel_sdvo_get_lvds_modes(struct drm_connector *connector)
 	struct drm_display_mode *newmode;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
-		      connector->base.id, drm_get_connector_name(connector));
+		      connector->base.id, connector->name);
 
 	/*
 	 * Fetch modes from VBT. For SDVO prefer the VBT mode since some
@@ -2362,24 +2395,62 @@ intel_sdvo_get_slave_addr(struct drm_device *dev, struct intel_sdvo *sdvo)
 }
 
 static void
+intel_sdvo_connector_unregister(struct intel_connector *intel_connector)
+{
+	struct drm_connector *drm_connector;
+	struct intel_sdvo *sdvo_encoder;
+
+	drm_connector = &intel_connector->base;
+	sdvo_encoder = intel_attached_sdvo(&intel_connector->base);
+
+	sysfs_remove_link(&drm_connector->kdev->kobj,
+			  sdvo_encoder->ddc.dev.kobj.name);
+	intel_connector_unregister(intel_connector);
+}
+
+static int
 intel_sdvo_connector_init(struct intel_sdvo_connector *connector,
 			  struct intel_sdvo *encoder)
 {
-	drm_connector_init(encoder->base.base.dev,
-			   &connector->base.base,
+	struct drm_connector *drm_connector;
+	int ret;
+
+	drm_connector = &connector->base.base;
+	ret = drm_connector_init(encoder->base.base.dev,
+			   drm_connector,
 			   &intel_sdvo_connector_funcs,
 			   connector->base.base.connector_type);
+	if (ret < 0)
+		return ret;
 
-	drm_connector_helper_add(&connector->base.base,
+	drm_connector_helper_add(drm_connector,
 				 &intel_sdvo_connector_helper_funcs);
 
 	connector->base.base.interlace_allowed = 1;
 	connector->base.base.doublescan_allowed = 0;
 	connector->base.base.display_info.subpixel_order = SubPixelHorizontalRGB;
 	connector->base.get_hw_state = intel_sdvo_connector_get_hw_state;
+	connector->base.unregister = intel_sdvo_connector_unregister;
 
 	intel_connector_attach_encoder(&connector->base, &encoder->base);
-	drm_sysfs_connector_add(&connector->base.base);
+	ret = drm_sysfs_connector_add(drm_connector);
+	if (ret < 0)
+		goto err1;
+
+	ret = sysfs_create_link(&drm_connector->kdev->kobj,
+				&encoder->ddc.dev.kobj,
+				encoder->ddc.dev.kobj.name);
+	if (ret < 0)
+		goto err2;
+
+	return 0;
+
+err2:
+	drm_sysfs_connector_remove(drm_connector);
+err1:
+	drm_connector_cleanup(drm_connector);
+
+	return ret;
 }
 
 static void
@@ -2439,7 +2510,11 @@ intel_sdvo_dvi_init(struct intel_sdvo *intel_sdvo, int device)
 		intel_sdvo->is_hdmi = true;
 	}
 
-	intel_sdvo_connector_init(intel_sdvo_connector, intel_sdvo);
+	if (intel_sdvo_connector_init(intel_sdvo_connector, intel_sdvo) < 0) {
+		kfree(intel_sdvo_connector);
+		return false;
+	}
+
 	if (intel_sdvo->is_hdmi)
 		intel_sdvo_add_hdmi_properties(intel_sdvo, intel_sdvo_connector);
 
@@ -2470,7 +2545,10 @@ intel_sdvo_tv_init(struct intel_sdvo *intel_sdvo, int type)
 
 	intel_sdvo->is_tv = true;
 
-	intel_sdvo_connector_init(intel_sdvo_connector, intel_sdvo);
+	if (intel_sdvo_connector_init(intel_sdvo_connector, intel_sdvo) < 0) {
+		kfree(intel_sdvo_connector);
+		return false;
+	}
 
 	if (!intel_sdvo_tv_create_property(intel_sdvo, intel_sdvo_connector, type))
 		goto err;
@@ -2514,8 +2592,11 @@ intel_sdvo_analog_init(struct intel_sdvo *intel_sdvo, int device)
 		intel_sdvo_connector->output_flag = SDVO_OUTPUT_RGB1;
 	}
 
-	intel_sdvo_connector_init(intel_sdvo_connector,
-				  intel_sdvo);
+	if (intel_sdvo_connector_init(intel_sdvo_connector, intel_sdvo) < 0) {
+		kfree(intel_sdvo_connector);
+		return false;
+	}
+
 	return true;
 }
 
@@ -2546,7 +2627,11 @@ intel_sdvo_lvds_init(struct intel_sdvo *intel_sdvo, int device)
 		intel_sdvo_connector->output_flag = SDVO_OUTPUT_LVDS1;
 	}
 
-	intel_sdvo_connector_init(intel_sdvo_connector, intel_sdvo);
+	if (intel_sdvo_connector_init(intel_sdvo_connector, intel_sdvo) < 0) {
+		kfree(intel_sdvo_connector);
+		return false;
+	}
+
 	if (!intel_sdvo_create_enhance_property(intel_sdvo, intel_sdvo_connector))
 		goto err;
 
@@ -2927,7 +3012,7 @@ bool intel_sdvo_init(struct drm_device *dev, uint32_t sdvo_reg, bool is_sdvob)
 
 	intel_encoder->compute_config = intel_sdvo_compute_config;
 	intel_encoder->disable = intel_disable_sdvo;
-	intel_encoder->mode_set = intel_sdvo_mode_set;
+	intel_encoder->pre_enable = intel_sdvo_pre_enable;
 	intel_encoder->enable = intel_enable_sdvo;
 	intel_encoder->get_hw_state = intel_sdvo_get_hw_state;
 	intel_encoder->get_config = intel_sdvo_get_config;
@@ -2960,7 +3045,7 @@ bool intel_sdvo_init(struct drm_device *dev, uint32_t sdvo_reg, bool is_sdvob)
 	 * simplistic anyway to express such constraints, so just give up on
 	 * cloning for SDVO encoders.
 	 */
-	intel_sdvo->base.cloneable = false;
+	intel_sdvo->base.cloneable = 0;
 
 	intel_sdvo_select_ddc_bus(dev_priv, intel_sdvo, sdvo_reg);
 
diff --git a/drivers/gpu/drm/i915/intel_sdvo_regs.h b/drivers/gpu/drm/i915/intel_sdvo_regs.h
index 770bdd6ecd9..2e2d4eb4a00 100644
--- a/drivers/gpu/drm/i915/intel_sdvo_regs.h
+++ b/drivers/gpu/drm/i915/intel_sdvo_regs.h
@@ -59,7 +59,7 @@ struct intel_sdvo_caps {
 	unsigned int stall_support:1;
 	unsigned int pad:1;
 	u16 output_flags;
-} __attribute__((packed));
+} __packed;
 
 /* Note: SDVO detailed timing flags match EDID misc flags. */
 #define DTD_FLAG_HSYNC_POSITIVE (1 << 1)
@@ -94,12 +94,12 @@ struct intel_sdvo_dtd {
 		u8 v_sync_off_high;
 		u8 reserved;
 	} part2;
-} __attribute__((packed));
+} __packed;
 
 struct intel_sdvo_pixel_clock_range {
 	u16 min;	/**< pixel clock, in 10kHz units */
 	u16 max;	/**< pixel clock, in 10kHz units */
-} __attribute__((packed));
+} __packed;
 
 struct intel_sdvo_preferred_input_timing_args {
 	u16 clock;
@@ -108,7 +108,7 @@ struct intel_sdvo_preferred_input_timing_args {
 	u8	interlace:1;
 	u8	scaled:1;
 	u8	pad:6;
-} __attribute__((packed));
+} __packed;
 
 /* I2C registers for SDVO */
 #define SDVO_I2C_ARG_0				0x07
@@ -162,7 +162,7 @@ struct intel_sdvo_get_trained_inputs_response {
 	unsigned int input0_trained:1;
 	unsigned int input1_trained:1;
 	unsigned int pad:6;
-} __attribute__((packed));
+} __packed;
 
 /** Returns a struct intel_sdvo_output_flags of active outputs. */
 #define SDVO_CMD_GET_ACTIVE_OUTPUTS			0x04
@@ -219,7 +219,7 @@ struct intel_sdvo_get_interrupt_event_source_response {
 	unsigned int ambient_light_interrupt:1;
 	unsigned int hdmi_audio_encrypt_change:1;
 	unsigned int pad:6;
-} __attribute__((packed));
+} __packed;
 
 /**
  * Selects which input is affected by future input commands.
@@ -232,7 +232,7 @@ struct intel_sdvo_get_interrupt_event_source_response {
 struct intel_sdvo_set_target_input_args {
 	unsigned int target_1:1;
 	unsigned int pad:7;
-} __attribute__((packed));
+} __packed;
 
 /**
  * Takes a struct intel_sdvo_output_flags of which outputs are targeted by
@@ -370,7 +370,7 @@ struct intel_sdvo_tv_format {
 	unsigned int hdtv_std_eia_7702a_480i_60:1;
 	unsigned int hdtv_std_eia_7702a_480p_60:1;
 	unsigned int pad:3;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_TV_FORMAT				0x28
 
@@ -401,7 +401,7 @@ struct intel_sdvo_sdtv_resolution_request {
 	unsigned int secam_l:1;
 	unsigned int secam_60:1;
 	unsigned int pad:5;
-} __attribute__((packed));
+} __packed;
 
 struct intel_sdvo_sdtv_resolution_reply {
 	unsigned int res_320x200:1;
@@ -426,7 +426,7 @@ struct intel_sdvo_sdtv_resolution_reply {
 	unsigned int res_1024x768:1;
 	unsigned int res_1280x1024:1;
 	unsigned int pad:5;
-} __attribute__((packed));
+} __packed;
 
 /* Get supported resolution with squire pixel aspect ratio that can be
    scaled for the requested HDTV format */
@@ -463,7 +463,7 @@ struct intel_sdvo_hdtv_resolution_request {
 	unsigned int hdtv_std_eia_7702a_480i_60:1;
 	unsigned int hdtv_std_eia_7702a_480p_60:1;
 	unsigned int pad:6;
-} __attribute__((packed));
+} __packed;
 
 struct intel_sdvo_hdtv_resolution_reply {
 	unsigned int res_640x480:1;
@@ -517,7 +517,7 @@ struct intel_sdvo_hdtv_resolution_reply {
 
 	unsigned int res_1280x768:1;
 	unsigned int pad5:7;
-} __attribute__((packed));
+} __packed;
 
 /* Get supported power state returns info for encoder and monitor, rely on
    last SetTargetInput and SetTargetOutput calls */
@@ -557,13 +557,13 @@ struct sdvo_panel_power_sequencing {
 
 	unsigned int t4_high:2;
 	unsigned int pad:6;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_MAX_BACKLIGHT_LEVEL		0x30
 struct sdvo_max_backlight_reply {
 	u8 max_value;
 	u8 default_value;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_BACKLIGHT_LEVEL			0x31
 #define SDVO_CMD_SET_BACKLIGHT_LEVEL			0x32
@@ -573,14 +573,14 @@ struct sdvo_get_ambient_light_reply {
 	u16 trip_low;
 	u16 trip_high;
 	u16 value;
-} __attribute__((packed));
+} __packed;
 #define SDVO_CMD_SET_AMBIENT_LIGHT			0x34
 struct sdvo_set_ambient_light_reply {
 	u16 trip_low;
 	u16 trip_high;
 	unsigned int enable:1;
 	unsigned int pad:7;
-} __attribute__((packed));
+} __packed;
 
 /* Set display power state */
 #define SDVO_CMD_SET_DISPLAY_POWER_STATE		0x7d
@@ -608,7 +608,7 @@ struct intel_sdvo_enhancements_reply {
 	unsigned int dither:1;
 	unsigned int tv_chroma_filter:1;
 	unsigned int tv_luma_filter:1;
-} __attribute__((packed));
+} __packed;
 
 /* Picture enhancement limits below are dependent on the current TV format,
  * and thus need to be queried and set after it.
@@ -630,7 +630,7 @@ struct intel_sdvo_enhancements_reply {
 struct intel_sdvo_enhancement_limits_reply {
 	u16 max_value;
 	u16 default_value;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_LVDS_PANEL_INFORMATION		0x7f
 #define SDVO_CMD_SET_LVDS_PANEL_INFORMATION		0x80
@@ -671,7 +671,7 @@ struct intel_sdvo_enhancement_limits_reply {
 #define SDVO_CMD_SET_TV_LUMA_FILTER			0x79
 struct intel_sdvo_enhancements_arg {
 	u16 value;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_DOT_CRAWL				0x70
 #define SDVO_CMD_SET_DOT_CRAWL				0x71
@@ -727,4 +727,4 @@ struct intel_sdvo_enhancements_arg {
 struct intel_sdvo_encode {
 	u8 dvi_rev;
 	u8 hdmi_rev;
-} __attribute__ ((packed));
+} __packed;
diff --git a/drivers/gpu/drm/i915/intel_sideband.c b/drivers/gpu/drm/i915/intel_sideband.c
index 9944d8135e8..01d841ea314 100644
--- a/drivers/gpu/drm/i915/intel_sideband.c
+++ b/drivers/gpu/drm/i915/intel_sideband.c
@@ -29,12 +29,21 @@
  * IOSF sideband, see VLV2_SidebandMsg_HAS.docx and
  * VLV_VLV2_PUNIT_HAS_0.8.docx
  */
+
+/* Standard MMIO read, non-posted */
+#define SB_MRD_NP	0x00
+/* Standard MMIO write, non-posted */
+#define SB_MWR_NP	0x01
+/* Private register read, double-word addressing, non-posted */
+#define SB_CRRDDA_NP	0x06
+/* Private register write, double-word addressing, non-posted */
+#define SB_CRWRDA_NP	0x07
+
 static int vlv_sideband_rw(struct drm_i915_private *dev_priv, u32 devfn,
 			   u32 port, u32 opcode, u32 addr, u32 *val)
 {
 	u32 cmd, be = 0xf, bar = 0;
-	bool is_read = (opcode == PUNIT_OPCODE_REG_READ ||
-			opcode == DPIO_OPCODE_REG_READ);
+	bool is_read = (opcode == SB_MRD_NP || opcode == SB_CRRDDA_NP);
 
 	cmd = (devfn << IOSF_DEVFN_SHIFT) | (opcode << IOSF_OPCODE_SHIFT) |
 		(port << IOSF_PORT_SHIFT) | (be << IOSF_BYTE_ENABLES_SHIFT) |
@@ -74,7 +83,7 @@ u32 vlv_punit_read(struct drm_i915_private *dev_priv, u8 addr)
 
 	mutex_lock(&dev_priv->dpio_lock);
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_PUNIT,
-			PUNIT_OPCODE_REG_READ, addr, &val);
+			SB_CRRDDA_NP, addr, &val);
 	mutex_unlock(&dev_priv->dpio_lock);
 
 	return val;
@@ -86,10 +95,26 @@ void vlv_punit_write(struct drm_i915_private *dev_priv, u8 addr, u32 val)
 
 	mutex_lock(&dev_priv->dpio_lock);
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_PUNIT,
-			PUNIT_OPCODE_REG_WRITE, addr, &val);
+			SB_CRWRDA_NP, addr, &val);
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
+u32 vlv_bunit_read(struct drm_i915_private *dev_priv, u32 reg)
+{
+	u32 val = 0;
+
+	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_BUNIT,
+			SB_CRRDDA_NP, reg, &val);
+
+	return val;
+}
+
+void vlv_bunit_write(struct drm_i915_private *dev_priv, u32 reg, u32 val)
+{
+	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_BUNIT,
+			SB_CRWRDA_NP, reg, &val);
+}
+
 u32 vlv_nc_read(struct drm_i915_private *dev_priv, u8 addr)
 {
 	u32 val = 0;
@@ -98,7 +123,7 @@ u32 vlv_nc_read(struct drm_i915_private *dev_priv, u8 addr)
 
 	mutex_lock(&dev_priv->dpio_lock);
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_NC,
-			PUNIT_OPCODE_REG_READ, addr, &val);
+			SB_CRRDDA_NP, addr, &val);
 	mutex_unlock(&dev_priv->dpio_lock);
 
 	return val;
@@ -108,80 +133,79 @@ u32 vlv_gpio_nc_read(struct drm_i915_private *dev_priv, u32 reg)
 {
 	u32 val = 0;
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_GPIO_NC,
-			PUNIT_OPCODE_REG_READ, reg, &val);
+			SB_CRRDDA_NP, reg, &val);
 	return val;
 }
 
 void vlv_gpio_nc_write(struct drm_i915_private *dev_priv, u32 reg, u32 val)
 {
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_GPIO_NC,
-			PUNIT_OPCODE_REG_WRITE, reg, &val);
+			SB_CRWRDA_NP, reg, &val);
 }
 
 u32 vlv_cck_read(struct drm_i915_private *dev_priv, u32 reg)
 {
 	u32 val = 0;
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_CCK,
-			PUNIT_OPCODE_REG_READ, reg, &val);
+			SB_CRRDDA_NP, reg, &val);
 	return val;
 }
 
 void vlv_cck_write(struct drm_i915_private *dev_priv, u32 reg, u32 val)
 {
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_CCK,
-			PUNIT_OPCODE_REG_WRITE, reg, &val);
+			SB_CRWRDA_NP, reg, &val);
 }
 
 u32 vlv_ccu_read(struct drm_i915_private *dev_priv, u32 reg)
 {
 	u32 val = 0;
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_CCU,
-			PUNIT_OPCODE_REG_READ, reg, &val);
+			SB_CRRDDA_NP, reg, &val);
 	return val;
 }
 
 void vlv_ccu_write(struct drm_i915_private *dev_priv, u32 reg, u32 val)
 {
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_CCU,
-			PUNIT_OPCODE_REG_WRITE, reg, &val);
+			SB_CRWRDA_NP, reg, &val);
 }
 
 u32 vlv_gps_core_read(struct drm_i915_private *dev_priv, u32 reg)
 {
 	u32 val = 0;
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_GPS_CORE,
-			PUNIT_OPCODE_REG_READ, reg, &val);
+			SB_CRRDDA_NP, reg, &val);
 	return val;
 }
 
 void vlv_gps_core_write(struct drm_i915_private *dev_priv, u32 reg, u32 val)
 {
 	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_GPS_CORE,
-			PUNIT_OPCODE_REG_WRITE, reg, &val);
-}
-
-static u32 vlv_get_phy_port(enum pipe pipe)
-{
-	u32 port = IOSF_PORT_DPIO;
-
-	WARN_ON ((pipe != PIPE_A) && (pipe != PIPE_B));
-
-	return port;
+			SB_CRWRDA_NP, reg, &val);
 }
 
 u32 vlv_dpio_read(struct drm_i915_private *dev_priv, enum pipe pipe, int reg)
 {
 	u32 val = 0;
 
-	vlv_sideband_rw(dev_priv, DPIO_DEVFN, vlv_get_phy_port(pipe),
-			DPIO_OPCODE_REG_READ, reg, &val);
+	vlv_sideband_rw(dev_priv, DPIO_DEVFN, DPIO_PHY_IOSF_PORT(DPIO_PHY(pipe)),
+			SB_MRD_NP, reg, &val);
+
+	/*
+	 * FIXME: There might be some registers where all 1's is a valid value,
+	 * so ideally we should check the register offset instead...
+	 */
+	WARN(val == 0xffffffff, "DPIO read pipe %c reg 0x%x == 0x%x\n",
+	     pipe_name(pipe), reg, val);
+
 	return val;
 }
 
 void vlv_dpio_write(struct drm_i915_private *dev_priv, enum pipe pipe, int reg, u32 val)
 {
-	vlv_sideband_rw(dev_priv, DPIO_DEVFN, vlv_get_phy_port(pipe),
-			DPIO_OPCODE_REG_WRITE, reg, &val);
+	vlv_sideband_rw(dev_priv, DPIO_DEVFN, DPIO_PHY_IOSF_PORT(DPIO_PHY(pipe)),
+			SB_MWR_NP, reg, &val);
 }
 
 /* SBI access */
@@ -242,3 +266,17 @@ void intel_sbi_write(struct drm_i915_private *dev_priv, u16 reg, u32 value,
 		return;
 	}
 }
+
+u32 vlv_flisdsi_read(struct drm_i915_private *dev_priv, u32 reg)
+{
+	u32 val = 0;
+	vlv_sideband_rw(dev_priv, DPIO_DEVFN, IOSF_PORT_FLISDSI, SB_CRRDDA_NP,
+			reg, &val);
+	return val;
+}
+
+void vlv_flisdsi_write(struct drm_i915_private *dev_priv, u32 reg, u32 val)
+{
+	vlv_sideband_rw(dev_priv, DPIO_DEVFN, IOSF_PORT_FLISDSI, SB_CRWRDA_NP,
+			reg, &val);
+}
diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c
index b9fabf826f7..9a17b4e92ef 100644
--- a/drivers/gpu/drm/i915/intel_sprite.c
+++ b/drivers/gpu/drm/i915/intel_sprite.c
@@ -37,6 +37,106 @@
 #include <drm/i915_drm.h>
 #include "i915_drv.h"
 
+static int usecs_to_scanlines(const struct drm_display_mode *mode, int usecs)
+{
+	/* paranoia */
+	if (!mode->crtc_htotal)
+		return 1;
+
+	return DIV_ROUND_UP(usecs * mode->crtc_clock, 1000 * mode->crtc_htotal);
+}
+
+static bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl_count)
+{
+	struct drm_device *dev = crtc->base.dev;
+	const struct drm_display_mode *mode = &crtc->config.adjusted_mode;
+	enum pipe pipe = crtc->pipe;
+	long timeout = msecs_to_jiffies_timeout(1);
+	int scanline, min, max, vblank_start;
+	DEFINE_WAIT(wait);
+
+	WARN_ON(!drm_modeset_is_locked(&crtc->base.mutex));
+
+	vblank_start = mode->crtc_vblank_start;
+	if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+		vblank_start = DIV_ROUND_UP(vblank_start, 2);
+
+	/* FIXME needs to be calibrated sensibly */
+	min = vblank_start - usecs_to_scanlines(mode, 100);
+	max = vblank_start - 1;
+
+	if (min <= 0 || max <= 0)
+		return false;
+
+	if (WARN_ON(drm_vblank_get(dev, pipe)))
+		return false;
+
+	local_irq_disable();
+
+	trace_i915_pipe_update_start(crtc, min, max);
+
+	for (;;) {
+		/*
+		 * prepare_to_wait() has a memory barrier, which guarantees
+		 * other CPUs can see the task state update by the time we
+		 * read the scanline.
+		 */
+		prepare_to_wait(&crtc->vbl_wait, &wait, TASK_UNINTERRUPTIBLE);
+
+		scanline = intel_get_crtc_scanline(crtc);
+		if (scanline < min || scanline > max)
+			break;
+
+		if (timeout <= 0) {
+			DRM_ERROR("Potential atomic update failure on pipe %c\n",
+				  pipe_name(crtc->pipe));
+			break;
+		}
+
+		local_irq_enable();
+
+		timeout = schedule_timeout(timeout);
+
+		local_irq_disable();
+	}
+
+	finish_wait(&crtc->vbl_wait, &wait);
+
+	drm_vblank_put(dev, pipe);
+
+	*start_vbl_count = dev->driver->get_vblank_counter(dev, pipe);
+
+	trace_i915_pipe_update_vblank_evaded(crtc, min, max, *start_vbl_count);
+
+	return true;
+}
+
+static void intel_pipe_update_end(struct intel_crtc *crtc, u32 start_vbl_count)
+{
+	struct drm_device *dev = crtc->base.dev;
+	enum pipe pipe = crtc->pipe;
+	u32 end_vbl_count = dev->driver->get_vblank_counter(dev, pipe);
+
+	trace_i915_pipe_update_end(crtc, end_vbl_count);
+
+	local_irq_enable();
+
+	if (start_vbl_count != end_vbl_count)
+		DRM_ERROR("Atomic update failure on pipe %c (start=%u end=%u)\n",
+			  pipe_name(pipe), start_vbl_count, end_vbl_count);
+}
+
+static void intel_update_primary_plane(struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
+	int reg = DSPCNTR(crtc->plane);
+
+	if (crtc->primary_enabled)
+		I915_WRITE(reg, I915_READ(reg) | DISPLAY_PLANE_ENABLE);
+	else
+		I915_WRITE(reg, I915_READ(reg) & ~DISPLAY_PLANE_ENABLE);
+}
+
 static void
 vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 		 struct drm_framebuffer *fb,
@@ -48,11 +148,14 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 	struct drm_device *dev = dplane->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_plane *intel_plane = to_intel_plane(dplane);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_plane->pipe;
 	int plane = intel_plane->plane;
 	u32 sprctl;
 	unsigned long sprsurf_offset, linear_offset;
 	int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
+	u32 start_vbl_count;
+	bool atomic_update;
 
 	sprctl = I915_READ(SPCNTR(pipe, plane));
 
@@ -104,6 +207,12 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 		break;
 	}
 
+	/*
+	 * Enable gamma to match primary/cursor plane behaviour.
+	 * FIXME should be user controllable via propertiesa.
+	 */
+	sprctl |= SP_GAMMA_ENABLE;
+
 	if (obj->tiling_mode != I915_TILING_NONE)
 		sprctl |= SP_TILED;
 
@@ -118,9 +227,6 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 	crtc_w--;
 	crtc_h--;
 
-	I915_WRITE(SPSTRIDE(pipe, plane), fb->pitches[0]);
-	I915_WRITE(SPPOS(pipe, plane), (crtc_y << 16) | crtc_x);
-
 	linear_offset = y * fb->pitches[0] + x * pixel_size;
 	sprsurf_offset = intel_gen4_compute_page_offset(&x, &y,
 							obj->tiling_mode,
@@ -128,6 +234,13 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 							fb->pitches[0]);
 	linear_offset -= sprsurf_offset;
 
+	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
+
+	intel_update_primary_plane(intel_crtc);
+
+	I915_WRITE(SPSTRIDE(pipe, plane), fb->pitches[0]);
+	I915_WRITE(SPPOS(pipe, plane), (crtc_y << 16) | crtc_x);
+
 	if (obj->tiling_mode != I915_TILING_NONE)
 		I915_WRITE(SPTILEOFF(pipe, plane), (y << 16) | x);
 	else
@@ -135,9 +248,13 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 
 	I915_WRITE(SPSIZE(pipe, plane), (crtc_h << 16) | crtc_w);
 	I915_WRITE(SPCNTR(pipe, plane), sprctl);
-	I915_MODIFY_DISPBASE(SPSURF(pipe, plane), i915_gem_obj_ggtt_offset(obj) +
-			     sprsurf_offset);
-	POSTING_READ(SPSURF(pipe, plane));
+	I915_WRITE(SPSURF(pipe, plane), i915_gem_obj_ggtt_offset(obj) +
+		   sprsurf_offset);
+
+	intel_flush_primary_plane(dev_priv, intel_crtc->plane);
+
+	if (atomic_update)
+		intel_pipe_update_end(intel_crtc, start_vbl_count);
 }
 
 static void
@@ -146,14 +263,25 @@ vlv_disable_plane(struct drm_plane *dplane, struct drm_crtc *crtc)
 	struct drm_device *dev = dplane->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_plane *intel_plane = to_intel_plane(dplane);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_plane->pipe;
 	int plane = intel_plane->plane;
+	u32 start_vbl_count;
+	bool atomic_update;
+
+	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
+
+	intel_update_primary_plane(intel_crtc);
 
 	I915_WRITE(SPCNTR(pipe, plane), I915_READ(SPCNTR(pipe, plane)) &
 		   ~SP_ENABLE);
 	/* Activate double buffered register update */
-	I915_MODIFY_DISPBASE(SPSURF(pipe, plane), 0);
-	POSTING_READ(SPSURF(pipe, plane));
+	I915_WRITE(SPSURF(pipe, plane), 0);
+
+	intel_flush_primary_plane(dev_priv, intel_crtc->plane);
+
+	if (atomic_update)
+		intel_pipe_update_end(intel_crtc, start_vbl_count);
 
 	intel_update_sprite_watermarks(dplane, crtc, 0, 0, false, false);
 }
@@ -220,11 +348,13 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	struct drm_device *dev = plane->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_plane *intel_plane = to_intel_plane(plane);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_plane->pipe;
 	u32 sprctl, sprscale = 0;
 	unsigned long sprsurf_offset, linear_offset;
 	int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
-	bool scaling_was_enabled = dev_priv->sprite_scaling_enabled;
+	u32 start_vbl_count;
+	bool atomic_update;
 
 	sprctl = I915_READ(SPRCTL(pipe));
 
@@ -257,6 +387,12 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		BUG();
 	}
 
+	/*
+	 * Enable gamma to match primary/cursor plane behaviour.
+	 * FIXME should be user controllable via propertiesa.
+	 */
+	sprctl |= SPRITE_GAMMA_ENABLE;
+
 	if (obj->tiling_mode != I915_TILING_NONE)
 		sprctl |= SPRITE_TILED;
 
@@ -279,24 +415,8 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	crtc_w--;
 	crtc_h--;
 
-	/*
-	 * IVB workaround: must disable low power watermarks for at least
-	 * one frame before enabling scaling.  LP watermarks can be re-enabled
-	 * when scaling is disabled.
-	 */
-	if (crtc_w != src_w || crtc_h != src_h) {
-		dev_priv->sprite_scaling_enabled |= 1 << pipe;
-
-		if (!scaling_was_enabled) {
-			intel_update_watermarks(crtc);
-			intel_wait_for_vblank(dev, pipe);
-		}
+	if (crtc_w != src_w || crtc_h != src_h)
 		sprscale = SPRITE_SCALE_ENABLE | (src_w << 16) | src_h;
-	} else
-		dev_priv->sprite_scaling_enabled &= ~(1 << pipe);
-
-	I915_WRITE(SPRSTRIDE(pipe), fb->pitches[0]);
-	I915_WRITE(SPRPOS(pipe), (crtc_y << 16) | crtc_x);
 
 	linear_offset = y * fb->pitches[0] + x * pixel_size;
 	sprsurf_offset =
@@ -304,6 +424,13 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 					       pixel_size, fb->pitches[0]);
 	linear_offset -= sprsurf_offset;
 
+	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
+
+	intel_update_primary_plane(intel_crtc);
+
+	I915_WRITE(SPRSTRIDE(pipe), fb->pitches[0]);
+	I915_WRITE(SPRPOS(pipe), (crtc_y << 16) | crtc_x);
+
 	/* HSW consolidates SPRTILEOFF and SPRLINOFF into a single SPROFFSET
 	 * register */
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
@@ -317,13 +444,13 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	if (intel_plane->can_scale)
 		I915_WRITE(SPRSCALE(pipe), sprscale);
 	I915_WRITE(SPRCTL(pipe), sprctl);
-	I915_MODIFY_DISPBASE(SPRSURF(pipe),
-			     i915_gem_obj_ggtt_offset(obj) + sprsurf_offset);
-	POSTING_READ(SPRSURF(pipe));
+	I915_WRITE(SPRSURF(pipe),
+		   i915_gem_obj_ggtt_offset(obj) + sprsurf_offset);
 
-	/* potentially re-enable LP watermarks */
-	if (scaling_was_enabled && !dev_priv->sprite_scaling_enabled)
-		intel_update_watermarks(crtc);
+	intel_flush_primary_plane(dev_priv, intel_crtc->plane);
+
+	if (atomic_update)
+		intel_pipe_update_end(intel_crtc, start_vbl_count);
 }
 
 static void
@@ -332,24 +459,34 @@ ivb_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc)
 	struct drm_device *dev = plane->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_plane *intel_plane = to_intel_plane(plane);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_plane->pipe;
-	bool scaling_was_enabled = dev_priv->sprite_scaling_enabled;
+	u32 start_vbl_count;
+	bool atomic_update;
+
+	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
+
+	intel_update_primary_plane(intel_crtc);
 
 	I915_WRITE(SPRCTL(pipe), I915_READ(SPRCTL(pipe)) & ~SPRITE_ENABLE);
 	/* Can't leave the scaler enabled... */
 	if (intel_plane->can_scale)
 		I915_WRITE(SPRSCALE(pipe), 0);
 	/* Activate double buffered register update */
-	I915_MODIFY_DISPBASE(SPRSURF(pipe), 0);
-	POSTING_READ(SPRSURF(pipe));
+	I915_WRITE(SPRSURF(pipe), 0);
 
-	dev_priv->sprite_scaling_enabled &= ~(1 << pipe);
+	intel_flush_primary_plane(dev_priv, intel_crtc->plane);
 
-	intel_update_sprite_watermarks(plane, crtc, 0, 0, false, false);
+	if (atomic_update)
+		intel_pipe_update_end(intel_crtc, start_vbl_count);
+
+	/*
+	 * Avoid underruns when disabling the sprite.
+	 * FIXME remove once watermark updates are done properly.
+	 */
+	intel_wait_for_vblank(dev, pipe);
 
-	/* potentially re-enable LP watermarks */
-	if (scaling_was_enabled && !dev_priv->sprite_scaling_enabled)
-		intel_update_watermarks(crtc);
+	intel_update_sprite_watermarks(plane, crtc, 0, 0, false, false);
 }
 
 static int
@@ -417,10 +554,13 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	struct drm_device *dev = plane->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_plane *intel_plane = to_intel_plane(plane);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_plane->pipe;
 	unsigned long dvssurf_offset, linear_offset;
 	u32 dvscntr, dvsscale;
 	int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
+	u32 start_vbl_count;
+	bool atomic_update;
 
 	dvscntr = I915_READ(DVSCNTR(pipe));
 
@@ -453,6 +593,12 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		BUG();
 	}
 
+	/*
+	 * Enable gamma to match primary/cursor plane behaviour.
+	 * FIXME should be user controllable via propertiesa.
+	 */
+	dvscntr |= DVS_GAMMA_ENABLE;
+
 	if (obj->tiling_mode != I915_TILING_NONE)
 		dvscntr |= DVS_TILED;
 
@@ -470,18 +616,22 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	crtc_h--;
 
 	dvsscale = 0;
-	if (IS_GEN5(dev) || crtc_w != src_w || crtc_h != src_h)
+	if (crtc_w != src_w || crtc_h != src_h)
 		dvsscale = DVS_SCALE_ENABLE | (src_w << 16) | src_h;
 
-	I915_WRITE(DVSSTRIDE(pipe), fb->pitches[0]);
-	I915_WRITE(DVSPOS(pipe), (crtc_y << 16) | crtc_x);
-
 	linear_offset = y * fb->pitches[0] + x * pixel_size;
 	dvssurf_offset =
 		intel_gen4_compute_page_offset(&x, &y, obj->tiling_mode,
 					       pixel_size, fb->pitches[0]);
 	linear_offset -= dvssurf_offset;
 
+	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
+
+	intel_update_primary_plane(intel_crtc);
+
+	I915_WRITE(DVSSTRIDE(pipe), fb->pitches[0]);
+	I915_WRITE(DVSPOS(pipe), (crtc_y << 16) | crtc_x);
+
 	if (obj->tiling_mode != I915_TILING_NONE)
 		I915_WRITE(DVSTILEOFF(pipe), (y << 16) | x);
 	else
@@ -490,9 +640,13 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	I915_WRITE(DVSSIZE(pipe), (crtc_h << 16) | crtc_w);
 	I915_WRITE(DVSSCALE(pipe), dvsscale);
 	I915_WRITE(DVSCNTR(pipe), dvscntr);
-	I915_MODIFY_DISPBASE(DVSSURF(pipe),
-			     i915_gem_obj_ggtt_offset(obj) + dvssurf_offset);
-	POSTING_READ(DVSSURF(pipe));
+	I915_WRITE(DVSSURF(pipe),
+		   i915_gem_obj_ggtt_offset(obj) + dvssurf_offset);
+
+	intel_flush_primary_plane(dev_priv, intel_crtc->plane);
+
+	if (atomic_update)
+		intel_pipe_update_end(intel_crtc, start_vbl_count);
 }
 
 static void
@@ -501,33 +655,48 @@ ilk_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc)
 	struct drm_device *dev = plane->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_plane *intel_plane = to_intel_plane(plane);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int pipe = intel_plane->pipe;
+	u32 start_vbl_count;
+	bool atomic_update;
+
+	atomic_update = intel_pipe_update_start(intel_crtc, &start_vbl_count);
+
+	intel_update_primary_plane(intel_crtc);
 
 	I915_WRITE(DVSCNTR(pipe), I915_READ(DVSCNTR(pipe)) & ~DVS_ENABLE);
 	/* Disable the scaler */
 	I915_WRITE(DVSSCALE(pipe), 0);
 	/* Flush double buffered register updates */
-	I915_MODIFY_DISPBASE(DVSSURF(pipe), 0);
-	POSTING_READ(DVSSURF(pipe));
+	I915_WRITE(DVSSURF(pipe), 0);
+
+	intel_flush_primary_plane(dev_priv, intel_crtc->plane);
+
+	if (atomic_update)
+		intel_pipe_update_end(intel_crtc, start_vbl_count);
+
+	/*
+	 * Avoid underruns when disabling the sprite.
+	 * FIXME remove once watermark updates are done properly.
+	 */
+	intel_wait_for_vblank(dev, pipe);
 
 	intel_update_sprite_watermarks(plane, crtc, 0, 0, false, false);
 }
 
 static void
-intel_enable_primary(struct drm_crtc *crtc)
+intel_post_enable_primary(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int reg = DSPCNTR(intel_crtc->plane);
 
-	if (intel_crtc->primary_enabled)
-		return;
-
-	intel_crtc->primary_enabled = true;
-
-	I915_WRITE(reg, I915_READ(reg) | DISPLAY_PLANE_ENABLE);
-	intel_flush_primary_plane(dev_priv, intel_crtc->plane);
+	/*
+	 * BDW signals flip done immediately if the plane
+	 * is disabled, even if the plane enable is already
+	 * armed to occur at the next vblank :(
+	 */
+	if (IS_BROADWELL(dev))
+		intel_wait_for_vblank(dev, intel_crtc->pipe);
 
 	/*
 	 * FIXME IPS should be fine as long as one plane is
@@ -535,10 +704,7 @@ intel_enable_primary(struct drm_crtc *crtc)
 	 * when going from primary only to sprite only and vice
 	 * versa.
 	 */
-	if (intel_crtc->config.ips_enabled) {
-		intel_wait_for_vblank(dev, intel_crtc->pipe);
-		hsw_enable_ips(intel_crtc);
-	}
+	hsw_enable_ips(intel_crtc);
 
 	mutex_lock(&dev->struct_mutex);
 	intel_update_fbc(dev);
@@ -546,17 +712,11 @@ intel_enable_primary(struct drm_crtc *crtc)
 }
 
 static void
-intel_disable_primary(struct drm_crtc *crtc)
+intel_pre_disable_primary(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int reg = DSPCNTR(intel_crtc->plane);
-
-	if (!intel_crtc->primary_enabled)
-		return;
-
-	intel_crtc->primary_enabled = false;
 
 	mutex_lock(&dev->struct_mutex);
 	if (dev_priv->fbc.plane == intel_crtc->plane)
@@ -570,9 +730,6 @@ intel_disable_primary(struct drm_crtc *crtc)
 	 * versa.
 	 */
 	hsw_disable_ips(intel_crtc);
-
-	I915_WRITE(reg, I915_READ(reg) & ~DISPLAY_PLANE_ENABLE);
-	intel_flush_primary_plane(dev_priv, intel_crtc->plane);
 }
 
 static int
@@ -643,6 +800,15 @@ format_is_yuv(uint32_t format)
 	}
 }
 
+static bool colorkey_enabled(struct intel_plane *intel_plane)
+{
+	struct drm_intel_sprite_colorkey key;
+
+	intel_plane->get_colorkey(&intel_plane->base, &key);
+
+	return key.flags != I915_SET_COLORKEY_NONE;
+}
+
 static int
 intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		   struct drm_framebuffer *fb, int crtc_x, int crtc_y,
@@ -657,7 +823,7 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	struct drm_i915_gem_object *obj = intel_fb->obj;
 	struct drm_i915_gem_object *old_obj = intel_plane->obj;
 	int ret;
-	bool disable_primary = false;
+	bool primary_enabled;
 	bool visible;
 	int hscale, vscale;
 	int max_scale, min_scale;
@@ -828,8 +994,8 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	 * If the sprite is completely covering the primary plane,
 	 * we can disable the primary and save power.
 	 */
-	disable_primary = drm_rect_equals(&dst, &clip);
-	WARN_ON(disable_primary && !visible && intel_crtc->active);
+	primary_enabled = !drm_rect_equals(&dst, &clip) || colorkey_enabled(intel_plane);
+	WARN_ON(!primary_enabled && !visible && intel_crtc->active);
 
 	mutex_lock(&dev->struct_mutex);
 
@@ -856,12 +1022,15 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	intel_plane->obj = obj;
 
 	if (intel_crtc->active) {
-		/*
-		 * Be sure to re-enable the primary before the sprite is no longer
-		 * covering it fully.
-		 */
-		if (!disable_primary)
-			intel_enable_primary(crtc);
+		bool primary_was_enabled = intel_crtc->primary_enabled;
+
+		intel_crtc->primary_enabled = primary_enabled;
+
+		if (primary_was_enabled != primary_enabled)
+			intel_crtc_wait_for_pending_flips(crtc);
+
+		if (primary_was_enabled && !primary_enabled)
+			intel_pre_disable_primary(crtc);
 
 		if (visible)
 			intel_plane->update_plane(plane, crtc, fb, obj,
@@ -870,8 +1039,8 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		else
 			intel_plane->disable_plane(plane, crtc);
 
-		if (disable_primary)
-			intel_disable_primary(crtc);
+		if (!primary_was_enabled && primary_enabled)
+			intel_post_enable_primary(crtc);
 	}
 
 	/* Unpin old obj after new one is active to avoid ugliness */
@@ -909,8 +1078,14 @@ intel_disable_plane(struct drm_plane *plane)
 	intel_crtc = to_intel_crtc(plane->crtc);
 
 	if (intel_crtc->active) {
-		intel_enable_primary(plane->crtc);
+		bool primary_was_enabled = intel_crtc->primary_enabled;
+
+		intel_crtc->primary_enabled = true;
+
 		intel_plane->disable_plane(plane, plane->crtc);
+
+		if (!primary_was_enabled && intel_crtc->primary_enabled)
+			intel_post_enable_primary(plane->crtc);
 	}
 
 	if (intel_plane->obj) {
diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c
index 22cf0f4ba24..67c6c9a2eb1 100644
--- a/drivers/gpu/drm/i915/intel_tv.c
+++ b/drivers/gpu/drm/i915/intel_tv.c
@@ -934,7 +934,86 @@ intel_tv_compute_config(struct intel_encoder *encoder,
 	return true;
 }
 
-static void intel_tv_mode_set(struct intel_encoder *encoder)
+static void
+set_tv_mode_timings(struct drm_i915_private *dev_priv,
+		    const struct tv_mode *tv_mode,
+		    bool burst_ena)
+{
+	u32 hctl1, hctl2, hctl3;
+	u32 vctl1, vctl2, vctl3, vctl4, vctl5, vctl6, vctl7;
+
+	hctl1 = (tv_mode->hsync_end << TV_HSYNC_END_SHIFT) |
+		(tv_mode->htotal << TV_HTOTAL_SHIFT);
+
+	hctl2 = (tv_mode->hburst_start << 16) |
+		(tv_mode->hburst_len << TV_HBURST_LEN_SHIFT);
+
+	if (burst_ena)
+		hctl2 |= TV_BURST_ENA;
+
+	hctl3 = (tv_mode->hblank_start << TV_HBLANK_START_SHIFT) |
+		(tv_mode->hblank_end << TV_HBLANK_END_SHIFT);
+
+	vctl1 = (tv_mode->nbr_end << TV_NBR_END_SHIFT) |
+		(tv_mode->vi_end_f1 << TV_VI_END_F1_SHIFT) |
+		(tv_mode->vi_end_f2 << TV_VI_END_F2_SHIFT);
+
+	vctl2 = (tv_mode->vsync_len << TV_VSYNC_LEN_SHIFT) |
+		(tv_mode->vsync_start_f1 << TV_VSYNC_START_F1_SHIFT) |
+		(tv_mode->vsync_start_f2 << TV_VSYNC_START_F2_SHIFT);
+
+	vctl3 = (tv_mode->veq_len << TV_VEQ_LEN_SHIFT) |
+		(tv_mode->veq_start_f1 << TV_VEQ_START_F1_SHIFT) |
+		(tv_mode->veq_start_f2 << TV_VEQ_START_F2_SHIFT);
+
+	if (tv_mode->veq_ena)
+		vctl3 |= TV_EQUAL_ENA;
+
+	vctl4 = (tv_mode->vburst_start_f1 << TV_VBURST_START_F1_SHIFT) |
+		(tv_mode->vburst_end_f1 << TV_VBURST_END_F1_SHIFT);
+
+	vctl5 = (tv_mode->vburst_start_f2 << TV_VBURST_START_F2_SHIFT) |
+		(tv_mode->vburst_end_f2 << TV_VBURST_END_F2_SHIFT);
+
+	vctl6 = (tv_mode->vburst_start_f3 << TV_VBURST_START_F3_SHIFT) |
+		(tv_mode->vburst_end_f3 << TV_VBURST_END_F3_SHIFT);
+
+	vctl7 = (tv_mode->vburst_start_f4 << TV_VBURST_START_F4_SHIFT) |
+		(tv_mode->vburst_end_f4 << TV_VBURST_END_F4_SHIFT);
+
+	I915_WRITE(TV_H_CTL_1, hctl1);
+	I915_WRITE(TV_H_CTL_2, hctl2);
+	I915_WRITE(TV_H_CTL_3, hctl3);
+	I915_WRITE(TV_V_CTL_1, vctl1);
+	I915_WRITE(TV_V_CTL_2, vctl2);
+	I915_WRITE(TV_V_CTL_3, vctl3);
+	I915_WRITE(TV_V_CTL_4, vctl4);
+	I915_WRITE(TV_V_CTL_5, vctl5);
+	I915_WRITE(TV_V_CTL_6, vctl6);
+	I915_WRITE(TV_V_CTL_7, vctl7);
+}
+
+static void set_color_conversion(struct drm_i915_private *dev_priv,
+				 const struct color_conversion *color_conversion)
+{
+	if (!color_conversion)
+		return;
+
+	I915_WRITE(TV_CSC_Y, (color_conversion->ry << 16) |
+		   color_conversion->gy);
+	I915_WRITE(TV_CSC_Y2, (color_conversion->by << 16) |
+		   color_conversion->ay);
+	I915_WRITE(TV_CSC_U, (color_conversion->ru << 16) |
+		   color_conversion->gu);
+	I915_WRITE(TV_CSC_U2, (color_conversion->bu << 16) |
+		   color_conversion->au);
+	I915_WRITE(TV_CSC_V, (color_conversion->rv << 16) |
+		   color_conversion->gv);
+	I915_WRITE(TV_CSC_V2, (color_conversion->bv << 16) |
+		   color_conversion->av);
+}
+
+static void intel_tv_pre_enable(struct intel_encoder *encoder)
 {
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -942,14 +1021,13 @@ static void intel_tv_mode_set(struct intel_encoder *encoder)
 	struct intel_tv *intel_tv = enc_to_tv(encoder);
 	const struct tv_mode *tv_mode = intel_tv_mode_find(intel_tv);
 	u32 tv_ctl;
-	u32 hctl1, hctl2, hctl3;
-	u32 vctl1, vctl2, vctl3, vctl4, vctl5, vctl6, vctl7;
 	u32 scctl1, scctl2, scctl3;
 	int i, j;
 	const struct video_levels *video_levels;
 	const struct color_conversion *color_conversion;
 	bool burst_ena;
-	int pipe = intel_crtc->pipe;
+	int xpos = 0x0, ypos = 0x0;
+	unsigned int xsize, ysize;
 
 	if (!tv_mode)
 		return;	/* can't happen (mode_prepare prevents this) */
@@ -982,44 +1060,6 @@ static void intel_tv_mode_set(struct intel_encoder *encoder)
 		burst_ena = tv_mode->burst_ena;
 		break;
 	}
-	hctl1 = (tv_mode->hsync_end << TV_HSYNC_END_SHIFT) |
-		(tv_mode->htotal << TV_HTOTAL_SHIFT);
-
-	hctl2 = (tv_mode->hburst_start << 16) |
-		(tv_mode->hburst_len << TV_HBURST_LEN_SHIFT);
-
-	if (burst_ena)
-		hctl2 |= TV_BURST_ENA;
-
-	hctl3 = (tv_mode->hblank_start << TV_HBLANK_START_SHIFT) |
-		(tv_mode->hblank_end << TV_HBLANK_END_SHIFT);
-
-	vctl1 = (tv_mode->nbr_end << TV_NBR_END_SHIFT) |
-		(tv_mode->vi_end_f1 << TV_VI_END_F1_SHIFT) |
-		(tv_mode->vi_end_f2 << TV_VI_END_F2_SHIFT);
-
-	vctl2 = (tv_mode->vsync_len << TV_VSYNC_LEN_SHIFT) |
-		(tv_mode->vsync_start_f1 << TV_VSYNC_START_F1_SHIFT) |
-		(tv_mode->vsync_start_f2 << TV_VSYNC_START_F2_SHIFT);
-
-	vctl3 = (tv_mode->veq_len << TV_VEQ_LEN_SHIFT) |
-		(tv_mode->veq_start_f1 << TV_VEQ_START_F1_SHIFT) |
-		(tv_mode->veq_start_f2 << TV_VEQ_START_F2_SHIFT);
-
-	if (tv_mode->veq_ena)
-		vctl3 |= TV_EQUAL_ENA;
-
-	vctl4 = (tv_mode->vburst_start_f1 << TV_VBURST_START_F1_SHIFT) |
-		(tv_mode->vburst_end_f1 << TV_VBURST_END_F1_SHIFT);
-
-	vctl5 = (tv_mode->vburst_start_f2 << TV_VBURST_START_F2_SHIFT) |
-		(tv_mode->vburst_end_f2 << TV_VBURST_END_F2_SHIFT);
-
-	vctl6 = (tv_mode->vburst_start_f3 << TV_VBURST_START_F3_SHIFT) |
-		(tv_mode->vburst_end_f3 << TV_VBURST_END_F3_SHIFT);
-
-	vctl7 = (tv_mode->vburst_start_f4 << TV_VBURST_START_F4_SHIFT) |
-		(tv_mode->vburst_end_f4 << TV_VBURST_END_F4_SHIFT);
 
 	if (intel_crtc->pipe == 1)
 		tv_ctl |= TV_ENC_PIPEB_SELECT;
@@ -1051,37 +1091,16 @@ static void intel_tv_mode_set(struct intel_encoder *encoder)
 		tv_mode->dda3_inc << TV_SCDDA3_INC_SHIFT;
 
 	/* Enable two fixes for the chips that need them. */
-	if (dev->pdev->device < 0x2772)
+	if (IS_I915GM(dev))
 		tv_ctl |= TV_ENC_C0_FIX | TV_ENC_SDP_FIX;
 
-	I915_WRITE(TV_H_CTL_1, hctl1);
-	I915_WRITE(TV_H_CTL_2, hctl2);
-	I915_WRITE(TV_H_CTL_3, hctl3);
-	I915_WRITE(TV_V_CTL_1, vctl1);
-	I915_WRITE(TV_V_CTL_2, vctl2);
-	I915_WRITE(TV_V_CTL_3, vctl3);
-	I915_WRITE(TV_V_CTL_4, vctl4);
-	I915_WRITE(TV_V_CTL_5, vctl5);
-	I915_WRITE(TV_V_CTL_6, vctl6);
-	I915_WRITE(TV_V_CTL_7, vctl7);
+	set_tv_mode_timings(dev_priv, tv_mode, burst_ena);
+
 	I915_WRITE(TV_SC_CTL_1, scctl1);
 	I915_WRITE(TV_SC_CTL_2, scctl2);
 	I915_WRITE(TV_SC_CTL_3, scctl3);
 
-	if (color_conversion) {
-		I915_WRITE(TV_CSC_Y, (color_conversion->ry << 16) |
-			   color_conversion->gy);
-		I915_WRITE(TV_CSC_Y2, (color_conversion->by << 16) |
-			   color_conversion->ay);
-		I915_WRITE(TV_CSC_U, (color_conversion->ru << 16) |
-			   color_conversion->gu);
-		I915_WRITE(TV_CSC_U2, (color_conversion->bu << 16) |
-			   color_conversion->au);
-		I915_WRITE(TV_CSC_V, (color_conversion->rv << 16) |
-			   color_conversion->gv);
-		I915_WRITE(TV_CSC_V2, (color_conversion->bv << 16) |
-			   color_conversion->av);
-	}
+	set_color_conversion(dev_priv, color_conversion);
 
 	if (INTEL_INFO(dev)->gen >= 4)
 		I915_WRITE(TV_CLR_KNOBS, 0x00404000);
@@ -1092,46 +1111,25 @@ static void intel_tv_mode_set(struct intel_encoder *encoder)
 		I915_WRITE(TV_CLR_LEVEL,
 			   ((video_levels->black << TV_BLACK_LEVEL_SHIFT) |
 			    (video_levels->blank << TV_BLANK_LEVEL_SHIFT)));
-	{
-		int pipeconf_reg = PIPECONF(pipe);
-		int dspcntr_reg = DSPCNTR(intel_crtc->plane);
-		int pipeconf = I915_READ(pipeconf_reg);
-		int dspcntr = I915_READ(dspcntr_reg);
-		int xpos = 0x0, ypos = 0x0;
-		unsigned int xsize, ysize;
-		/* Pipe must be off here */
-		I915_WRITE(dspcntr_reg, dspcntr & ~DISPLAY_PLANE_ENABLE);
-		intel_flush_primary_plane(dev_priv, intel_crtc->plane);
-
-		/* Wait for vblank for the disable to take effect */
-		if (IS_GEN2(dev))
-			intel_wait_for_vblank(dev, intel_crtc->pipe);
-
-		I915_WRITE(pipeconf_reg, pipeconf & ~PIPECONF_ENABLE);
-		/* Wait for vblank for the disable to take effect. */
-		intel_wait_for_pipe_off(dev, intel_crtc->pipe);
-
-		/* Filter ctl must be set before TV_WIN_SIZE */
-		I915_WRITE(TV_FILTER_CTL_1, TV_AUTO_SCALE);
-		xsize = tv_mode->hblank_start - tv_mode->hblank_end;
-		if (tv_mode->progressive)
-			ysize = tv_mode->nbr_end + 1;
-		else
-			ysize = 2*tv_mode->nbr_end + 1;
-
-		xpos += intel_tv->margin[TV_MARGIN_LEFT];
-		ypos += intel_tv->margin[TV_MARGIN_TOP];
-		xsize -= (intel_tv->margin[TV_MARGIN_LEFT] +
-			  intel_tv->margin[TV_MARGIN_RIGHT]);
-		ysize -= (intel_tv->margin[TV_MARGIN_TOP] +
-			  intel_tv->margin[TV_MARGIN_BOTTOM]);
-		I915_WRITE(TV_WIN_POS, (xpos<<16)|ypos);
-		I915_WRITE(TV_WIN_SIZE, (xsize<<16)|ysize);
-
-		I915_WRITE(pipeconf_reg, pipeconf);
-		I915_WRITE(dspcntr_reg, dspcntr);
-		intel_flush_primary_plane(dev_priv, intel_crtc->plane);
-	}
+
+	assert_pipe_disabled(dev_priv, intel_crtc->pipe);
+
+	/* Filter ctl must be set before TV_WIN_SIZE */
+	I915_WRITE(TV_FILTER_CTL_1, TV_AUTO_SCALE);
+	xsize = tv_mode->hblank_start - tv_mode->hblank_end;
+	if (tv_mode->progressive)
+		ysize = tv_mode->nbr_end + 1;
+	else
+		ysize = 2*tv_mode->nbr_end + 1;
+
+	xpos += intel_tv->margin[TV_MARGIN_LEFT];
+	ypos += intel_tv->margin[TV_MARGIN_TOP];
+	xsize -= (intel_tv->margin[TV_MARGIN_LEFT] +
+		  intel_tv->margin[TV_MARGIN_RIGHT]);
+	ysize -= (intel_tv->margin[TV_MARGIN_TOP] +
+		  intel_tv->margin[TV_MARGIN_BOTTOM]);
+	I915_WRITE(TV_WIN_POS, (xpos<<16)|ypos);
+	I915_WRITE(TV_WIN_SIZE, (xsize<<16)|ysize);
 
 	j = 0;
 	for (i = 0; i < 60; i++)
@@ -1189,8 +1187,8 @@ intel_tv_detect_type(struct intel_tv *intel_tv,
 	if (connector->polled & DRM_CONNECTOR_POLL_HPD) {
 		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 		i915_disable_pipestat(dev_priv, 0,
-				      PIPE_HOTPLUG_INTERRUPT_ENABLE |
-				      PIPE_HOTPLUG_TV_INTERRUPT_ENABLE);
+				      PIPE_HOTPLUG_INTERRUPT_STATUS |
+				      PIPE_HOTPLUG_TV_INTERRUPT_STATUS);
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 	}
 
@@ -1266,8 +1264,8 @@ intel_tv_detect_type(struct intel_tv *intel_tv,
 	if (connector->polled & DRM_CONNECTOR_POLL_HPD) {
 		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 		i915_enable_pipestat(dev_priv, 0,
-				     PIPE_HOTPLUG_INTERRUPT_ENABLE |
-				     PIPE_HOTPLUG_TV_INTERRUPT_ENABLE);
+				     PIPE_HOTPLUG_INTERRUPT_STATUS |
+				     PIPE_HOTPLUG_TV_INTERRUPT_STATUS);
 		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
 	}
 
@@ -1316,17 +1314,18 @@ intel_tv_detect(struct drm_connector *connector, bool force)
 	int type;
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s] force=%d\n",
-		      connector->base.id, drm_get_connector_name(connector),
+		      connector->base.id, connector->name,
 		      force);
 
 	mode = reported_modes[0];
 
 	if (force) {
 		struct intel_load_detect_pipe tmp;
+		struct drm_modeset_acquire_ctx ctx;
 
-		if (intel_get_load_detect_pipe(connector, &mode, &tmp)) {
+		if (intel_get_load_detect_pipe(connector, &mode, &tmp, &ctx)) {
 			type = intel_tv_detect_type(intel_tv, connector);
-			intel_release_load_detect_pipe(connector, &tmp);
+			intel_release_load_detect_pipe(connector, &tmp, &ctx);
 		} else
 			return connector_status_unknown;
 	} else
@@ -1536,9 +1535,14 @@ static int tv_is_present_in_vbt(struct drm_device *dev)
 		/*
 		 * If the device type is not TV, continue.
 		 */
-		if (p_child->old.device_type != DEVICE_TYPE_INT_TV &&
-			p_child->old.device_type != DEVICE_TYPE_TV)
+		switch (p_child->old.device_type) {
+		case DEVICE_TYPE_INT_TV:
+		case DEVICE_TYPE_TV:
+		case DEVICE_TYPE_TV_SVIDEO_COMPOSITE:
+			break;
+		default:
 			continue;
+		}
 		/* Only when the addin_offset is non-zero, it is regarded
 		 * as present.
 		 */
@@ -1629,18 +1633,18 @@ intel_tv_init(struct drm_device *dev)
 
 	intel_encoder->compute_config = intel_tv_compute_config;
 	intel_encoder->get_config = intel_tv_get_config;
-	intel_encoder->mode_set = intel_tv_mode_set;
+	intel_encoder->pre_enable = intel_tv_pre_enable;
 	intel_encoder->enable = intel_enable_tv;
 	intel_encoder->disable = intel_disable_tv;
 	intel_encoder->get_hw_state = intel_tv_get_hw_state;
 	intel_connector->get_hw_state = intel_connector_get_hw_state;
+	intel_connector->unregister = intel_connector_unregister;
 
 	intel_connector_attach_encoder(intel_connector, intel_encoder);
 	intel_encoder->type = INTEL_OUTPUT_TVOUT;
 	intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
-	intel_encoder->cloneable = false;
+	intel_encoder->cloneable = 0;
 	intel_encoder->base.possible_crtcs = ((1 << 0) | (1 << 1));
-	intel_encoder->base.possible_clones = (1 << INTEL_OUTPUT_TVOUT);
 	intel_tv->type = DRM_MODE_CONNECTOR_Unknown;
 
 	/* BIOS margin values */
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index 25cbe073c38..4f6fef7ac06 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -40,6 +40,12 @@
 
 #define __raw_posting_read(dev_priv__, reg__) (void)__raw_i915_read32(dev_priv__, reg__)
 
+static void
+assert_device_not_suspended(struct drm_i915_private *dev_priv)
+{
+	WARN(HAS_RUNTIME_PM(dev_priv->dev) && dev_priv->pm.suspended,
+	     "Device suspended\n");
+}
 
 static void __gen6_gt_wait_for_thread_c0(struct drm_i915_private *dev_priv)
 {
@@ -64,7 +70,8 @@ static void __gen6_gt_force_wake_reset(struct drm_i915_private *dev_priv)
 	__raw_posting_read(dev_priv, ECOBUS);
 }
 
-static void __gen6_gt_force_wake_get(struct drm_i915_private *dev_priv)
+static void __gen6_gt_force_wake_get(struct drm_i915_private *dev_priv,
+							int fw_engine)
 {
 	if (wait_for_atomic((__raw_i915_read32(dev_priv, FORCEWAKE_ACK) & 1) == 0,
 			    FORCEWAKE_ACK_TIMEOUT_MS))
@@ -82,14 +89,15 @@ static void __gen6_gt_force_wake_get(struct drm_i915_private *dev_priv)
 	__gen6_gt_wait_for_thread_c0(dev_priv);
 }
 
-static void __gen6_gt_force_wake_mt_reset(struct drm_i915_private *dev_priv)
+static void __gen7_gt_force_wake_mt_reset(struct drm_i915_private *dev_priv)
 {
 	__raw_i915_write32(dev_priv, FORCEWAKE_MT, _MASKED_BIT_DISABLE(0xffff));
 	/* something from same cacheline, but !FORCEWAKE_MT */
 	__raw_posting_read(dev_priv, ECOBUS);
 }
 
-static void __gen6_gt_force_wake_mt_get(struct drm_i915_private *dev_priv)
+static void __gen7_gt_force_wake_mt_get(struct drm_i915_private *dev_priv,
+							int fw_engine)
 {
 	u32 forcewake_ack;
 
@@ -121,12 +129,12 @@ static void gen6_gt_check_fifodbg(struct drm_i915_private *dev_priv)
 	u32 gtfifodbg;
 
 	gtfifodbg = __raw_i915_read32(dev_priv, GTFIFODBG);
-	if (WARN(gtfifodbg & GT_FIFO_CPU_ERROR_MASK,
-	     "MMIO read or write has been dropped %x\n", gtfifodbg))
-		__raw_i915_write32(dev_priv, GTFIFODBG, GT_FIFO_CPU_ERROR_MASK);
+	if (WARN(gtfifodbg, "GT wake FIFO error 0x%x\n", gtfifodbg))
+		__raw_i915_write32(dev_priv, GTFIFODBG, gtfifodbg);
 }
 
-static void __gen6_gt_force_wake_put(struct drm_i915_private *dev_priv)
+static void __gen6_gt_force_wake_put(struct drm_i915_private *dev_priv,
+							int fw_engine)
 {
 	__raw_i915_write32(dev_priv, FORCEWAKE, 0);
 	/* something from same cacheline, but !FORCEWAKE */
@@ -134,25 +142,35 @@ static void __gen6_gt_force_wake_put(struct drm_i915_private *dev_priv)
 	gen6_gt_check_fifodbg(dev_priv);
 }
 
-static void __gen6_gt_force_wake_mt_put(struct drm_i915_private *dev_priv)
+static void __gen7_gt_force_wake_mt_put(struct drm_i915_private *dev_priv,
+							int fw_engine)
 {
 	__raw_i915_write32(dev_priv, FORCEWAKE_MT,
 			   _MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
 	/* something from same cacheline, but !FORCEWAKE_MT */
 	__raw_posting_read(dev_priv, ECOBUS);
-	gen6_gt_check_fifodbg(dev_priv);
+
+	if (IS_GEN7(dev_priv->dev))
+		gen6_gt_check_fifodbg(dev_priv);
 }
 
 static int __gen6_gt_wait_for_fifo(struct drm_i915_private *dev_priv)
 {
 	int ret = 0;
 
+	/* On VLV, FIFO will be shared by both SW and HW.
+	 * So, we need to read the FREE_ENTRIES everytime */
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		dev_priv->uncore.fifo_count =
+			__raw_i915_read32(dev_priv, GTFIFOCTL) &
+						GT_FIFO_FREE_ENTRIES_MASK;
+
 	if (dev_priv->uncore.fifo_count < GT_FIFO_NUM_RESERVED_ENTRIES) {
 		int loop = 500;
-		u32 fifo = __raw_i915_read32(dev_priv, GT_FIFO_FREE_ENTRIES);
+		u32 fifo = __raw_i915_read32(dev_priv, GTFIFOCTL) & GT_FIFO_FREE_ENTRIES_MASK;
 		while (fifo <= GT_FIFO_NUM_RESERVED_ENTRIES && loop--) {
 			udelay(10);
-			fifo = __raw_i915_read32(dev_priv, GT_FIFO_FREE_ENTRIES);
+			fifo = __raw_i915_read32(dev_priv, GTFIFOCTL) & GT_FIFO_FREE_ENTRIES_MASK;
 		}
 		if (WARN_ON(loop < 0 && fifo <= GT_FIFO_NUM_RESERVED_ENTRIES))
 			++ret;
@@ -167,67 +185,185 @@ static void vlv_force_wake_reset(struct drm_i915_private *dev_priv)
 {
 	__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
 			   _MASKED_BIT_DISABLE(0xffff));
+	__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
+			   _MASKED_BIT_DISABLE(0xffff));
 	/* something from same cacheline, but !FORCEWAKE_VLV */
 	__raw_posting_read(dev_priv, FORCEWAKE_ACK_VLV);
 }
 
-static void vlv_force_wake_get(struct drm_i915_private *dev_priv)
+static void __vlv_force_wake_get(struct drm_i915_private *dev_priv,
+						int fw_engine)
 {
-	if (wait_for_atomic((__raw_i915_read32(dev_priv, FORCEWAKE_ACK_VLV) & FORCEWAKE_KERNEL) == 0,
-			    FORCEWAKE_ACK_TIMEOUT_MS))
-		DRM_ERROR("Timed out waiting for forcewake old ack to clear.\n");
-
-	__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
-			   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
-	__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
-			   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
-
-	if (wait_for_atomic((__raw_i915_read32(dev_priv, FORCEWAKE_ACK_VLV) & FORCEWAKE_KERNEL),
-			    FORCEWAKE_ACK_TIMEOUT_MS))
-		DRM_ERROR("Timed out waiting for GT to ack forcewake request.\n");
+	/* Check for Render Engine */
+	if (FORCEWAKE_RENDER & fw_engine) {
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_VLV) &
+						FORCEWAKE_KERNEL) == 0,
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: Render forcewake old ack to clear.\n");
+
+		__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
+				   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
+
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_VLV) &
+						FORCEWAKE_KERNEL),
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: waiting for Render to ack.\n");
+	}
 
-	if (wait_for_atomic((__raw_i915_read32(dev_priv, FORCEWAKE_ACK_MEDIA_VLV) &
-			     FORCEWAKE_KERNEL),
-			    FORCEWAKE_ACK_TIMEOUT_MS))
-		DRM_ERROR("Timed out waiting for media to ack forcewake request.\n");
+	/* Check for Media Engine */
+	if (FORCEWAKE_MEDIA & fw_engine) {
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_MEDIA_VLV) &
+						FORCEWAKE_KERNEL) == 0,
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: Media forcewake old ack to clear.\n");
+
+		__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
+				   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
+
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_MEDIA_VLV) &
+						FORCEWAKE_KERNEL),
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: waiting for media to ack.\n");
+	}
 
 	/* WaRsForcewakeWaitTC0:vlv */
 	__gen6_gt_wait_for_thread_c0(dev_priv);
+
 }
 
-static void vlv_force_wake_put(struct drm_i915_private *dev_priv)
+static void __vlv_force_wake_put(struct drm_i915_private *dev_priv,
+					int fw_engine)
 {
-	__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
-			   _MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
-	__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
-			   _MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
+
+	/* Check for Render Engine */
+	if (FORCEWAKE_RENDER & fw_engine)
+		__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
+					_MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
+
+
+	/* Check for Media Engine */
+	if (FORCEWAKE_MEDIA & fw_engine)
+		__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
+				_MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
+
 	/* The below doubles as a POSTING_READ */
 	gen6_gt_check_fifodbg(dev_priv);
+
+}
+
+static void vlv_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine)
+{
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+
+	if (fw_engine & FORCEWAKE_RENDER &&
+	    dev_priv->uncore.fw_rendercount++ != 0)
+		fw_engine &= ~FORCEWAKE_RENDER;
+	if (fw_engine & FORCEWAKE_MEDIA &&
+	    dev_priv->uncore.fw_mediacount++ != 0)
+		fw_engine &= ~FORCEWAKE_MEDIA;
+
+	if (fw_engine)
+		dev_priv->uncore.funcs.force_wake_get(dev_priv, fw_engine);
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+}
+
+static void vlv_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine)
+{
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+
+	if (fw_engine & FORCEWAKE_RENDER) {
+		WARN_ON(!dev_priv->uncore.fw_rendercount);
+		if (--dev_priv->uncore.fw_rendercount != 0)
+			fw_engine &= ~FORCEWAKE_RENDER;
+	}
+
+	if (fw_engine & FORCEWAKE_MEDIA) {
+		WARN_ON(!dev_priv->uncore.fw_mediacount);
+		if (--dev_priv->uncore.fw_mediacount != 0)
+			fw_engine &= ~FORCEWAKE_MEDIA;
+	}
+
+	if (fw_engine)
+		dev_priv->uncore.funcs.force_wake_put(dev_priv, fw_engine);
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
-static void gen6_force_wake_work(struct work_struct *work)
+static void gen6_force_wake_timer(unsigned long arg)
 {
-	struct drm_i915_private *dev_priv =
-		container_of(work, typeof(*dev_priv), uncore.force_wake_work.work);
+	struct drm_i915_private *dev_priv = (void *)arg;
 	unsigned long irqflags;
 
+	assert_device_not_suspended(dev_priv);
+
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+	WARN_ON(!dev_priv->uncore.forcewake_count);
+
 	if (--dev_priv->uncore.forcewake_count == 0)
-		dev_priv->uncore.funcs.force_wake_put(dev_priv);
+		dev_priv->uncore.funcs.force_wake_put(dev_priv, FORCEWAKE_ALL);
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+
+	intel_runtime_pm_put(dev_priv);
 }
 
-static void intel_uncore_forcewake_reset(struct drm_device *dev)
+static void intel_uncore_forcewake_reset(struct drm_device *dev, bool restore)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long irqflags;
 
-	if (IS_VALLEYVIEW(dev)) {
+	if (del_timer_sync(&dev_priv->uncore.force_wake_timer))
+		gen6_force_wake_timer((unsigned long)dev_priv);
+
+	/* Hold uncore.lock across reset to prevent any register access
+	 * with forcewake not set correctly
+	 */
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+
+	if (IS_VALLEYVIEW(dev))
 		vlv_force_wake_reset(dev_priv);
-	} else if (INTEL_INFO(dev)->gen >= 6) {
+	else if (IS_GEN6(dev) || IS_GEN7(dev))
 		__gen6_gt_force_wake_reset(dev_priv);
-		if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev))
-			__gen6_gt_force_wake_mt_reset(dev_priv);
+
+	if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev) || IS_GEN8(dev))
+		__gen7_gt_force_wake_mt_reset(dev_priv);
+
+	if (restore) { /* If reset with a user forcewake, try to restore */
+		unsigned fw = 0;
+
+		if (IS_VALLEYVIEW(dev)) {
+			if (dev_priv->uncore.fw_rendercount)
+				fw |= FORCEWAKE_RENDER;
+
+			if (dev_priv->uncore.fw_mediacount)
+				fw |= FORCEWAKE_MEDIA;
+		} else {
+			if (dev_priv->uncore.forcewake_count)
+				fw = FORCEWAKE_ALL;
+		}
+
+		if (fw)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv, fw);
+
+		if (IS_GEN6(dev) || IS_GEN7(dev))
+			dev_priv->uncore.fifo_count =
+				__raw_i915_read32(dev_priv, GTFIFOCTL) &
+				GT_FIFO_FREE_ENTRIES_MASK;
+	} else {
+		dev_priv->uncore.forcewake_count = 0;
+		dev_priv->uncore.fw_rendercount = 0;
+		dev_priv->uncore.fw_mediacount = 0;
 	}
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
 void intel_uncore_early_sanitize(struct drm_device *dev)
@@ -237,7 +373,7 @@ void intel_uncore_early_sanitize(struct drm_device *dev)
 	if (HAS_FPGA_DBG_UNCLAIMED(dev))
 		__raw_i915_write32(dev_priv, FPGA_DBG, FPGA_DBG_RM_NOCLAIM);
 
-	if (IS_HASWELL(dev) &&
+	if ((IS_HASWELL(dev) || IS_BROADWELL(dev)) &&
 	    (__raw_i915_read32(dev_priv, HSW_EDRAM_PRESENT) == 1)) {
 		/* The docs do not explain exactly how the calculation can be
 		 * made. It is somewhat guessable, but for now, it's always
@@ -248,31 +384,18 @@ void intel_uncore_early_sanitize(struct drm_device *dev)
 		DRM_INFO("Found %zuMB of eLLC\n", dev_priv->ellc_size);
 	}
 
-	intel_uncore_forcewake_reset(dev);
+	/* clear out old GT FIFO errors */
+	if (IS_GEN6(dev) || IS_GEN7(dev))
+		__raw_i915_write32(dev_priv, GTFIFODBG,
+				   __raw_i915_read32(dev_priv, GTFIFODBG));
+
+	intel_uncore_forcewake_reset(dev, false);
 }
 
 void intel_uncore_sanitize(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 reg_val;
-
-	intel_uncore_forcewake_reset(dev);
-
 	/* BIOS often leaves RC6 enabled, but disable it for hw init */
 	intel_disable_gt_powersave(dev);
-
-	/* Turn off power gate, require especially for the BIOS less system */
-	if (IS_VALLEYVIEW(dev)) {
-
-		mutex_lock(&dev_priv->rps.hw_lock);
-		reg_val = vlv_punit_read(dev_priv, PUNIT_REG_PWRGT_STATUS);
-
-		if (reg_val & (RENDER_PWRGT | MEDIA_PWRGT | DISP2D_PWRGT))
-			vlv_punit_write(dev_priv, PUNIT_REG_PWRGT_CTRL, 0x0);
-
-		mutex_unlock(&dev_priv->rps.hw_lock);
-
-	}
 }
 
 /*
@@ -281,43 +404,82 @@ void intel_uncore_sanitize(struct drm_device *dev)
  * be called at the beginning of the sequence followed by a call to
  * gen6_gt_force_wake_put() at the end of the sequence.
  */
-void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv)
+void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine)
 {
 	unsigned long irqflags;
 
 	if (!dev_priv->uncore.funcs.force_wake_get)
 		return;
 
+	intel_runtime_pm_get(dev_priv);
+
+	/* Redirect to VLV specific routine */
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		return vlv_force_wake_get(dev_priv, fw_engine);
+
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 	if (dev_priv->uncore.forcewake_count++ == 0)
-		dev_priv->uncore.funcs.force_wake_get(dev_priv);
+		dev_priv->uncore.funcs.force_wake_get(dev_priv, FORCEWAKE_ALL);
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
 /*
  * see gen6_gt_force_wake_get()
  */
-void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv)
+void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine)
 {
 	unsigned long irqflags;
+	bool delayed = false;
 
 	if (!dev_priv->uncore.funcs.force_wake_put)
 		return;
 
+	/* Redirect to VLV specific routine */
+	if (IS_VALLEYVIEW(dev_priv->dev)) {
+		vlv_force_wake_put(dev_priv, fw_engine);
+		goto out;
+	}
+
+
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+	WARN_ON(!dev_priv->uncore.forcewake_count);
+
 	if (--dev_priv->uncore.forcewake_count == 0) {
 		dev_priv->uncore.forcewake_count++;
-		mod_delayed_work(dev_priv->wq,
-				 &dev_priv->uncore.force_wake_work,
-				 1);
+		delayed = true;
+		mod_timer_pinned(&dev_priv->uncore.force_wake_timer,
+				 jiffies + 1);
 	}
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+
+out:
+	if (!delayed)
+		intel_runtime_pm_put(dev_priv);
+}
+
+void assert_force_wake_inactive(struct drm_i915_private *dev_priv)
+{
+	if (!dev_priv->uncore.funcs.force_wake_get)
+		return;
+
+	WARN_ON(dev_priv->uncore.forcewake_count > 0);
 }
 
 /* We give fast paths for the really cool registers */
 #define NEEDS_FORCE_WAKE(dev_priv, reg) \
 	 ((reg) < 0x40000 && (reg) != FORCEWAKE)
 
+#define FORCEWAKE_VLV_RENDER_RANGE_OFFSET(reg) \
+	(((reg) >= 0x2000 && (reg) < 0x4000) ||\
+	((reg) >= 0x5000 && (reg) < 0x8000) ||\
+	((reg) >= 0xB000 && (reg) < 0x12000) ||\
+	((reg) >= 0x2E000 && (reg) < 0x30000))
+
+#define FORCEWAKE_VLV_MEDIA_RANGE_OFFSET(reg)\
+	(((reg) >= 0x12000 && (reg) < 0x14000) ||\
+	((reg) >= 0x22000 && (reg) < 0x24000) ||\
+	((reg) >= 0x30000 && (reg) < 0x40000))
+
 static void
 ilk_dummy_write(struct drm_i915_private *dev_priv)
 {
@@ -349,6 +511,7 @@ hsw_unclaimed_reg_check(struct drm_i915_private *dev_priv, u32 reg)
 #define REG_READ_HEADER(x) \
 	unsigned long irqflags; \
 	u##x val = 0; \
+	assert_device_not_suspended(dev_priv); \
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags)
 
 #define REG_READ_FOOTER \
@@ -377,18 +540,44 @@ gen5_read##x(struct drm_i915_private *dev_priv, off_t reg, bool trace) { \
 static u##x \
 gen6_read##x(struct drm_i915_private *dev_priv, off_t reg, bool trace) { \
 	REG_READ_HEADER(x); \
-	if (NEEDS_FORCE_WAKE((dev_priv), (reg))) { \
-		if (dev_priv->uncore.forcewake_count == 0) \
-			dev_priv->uncore.funcs.force_wake_get(dev_priv); \
+	if (dev_priv->uncore.forcewake_count == 0 && \
+	    NEEDS_FORCE_WAKE((dev_priv), (reg))) { \
+		dev_priv->uncore.funcs.force_wake_get(dev_priv, \
+						      FORCEWAKE_ALL); \
 		val = __raw_i915_read##x(dev_priv, reg); \
-		if (dev_priv->uncore.forcewake_count == 0) \
-			dev_priv->uncore.funcs.force_wake_put(dev_priv); \
+		dev_priv->uncore.funcs.force_wake_put(dev_priv, \
+						      FORCEWAKE_ALL); \
 	} else { \
 		val = __raw_i915_read##x(dev_priv, reg); \
 	} \
 	REG_READ_FOOTER; \
 }
 
+#define __vlv_read(x) \
+static u##x \
+vlv_read##x(struct drm_i915_private *dev_priv, off_t reg, bool trace) { \
+	unsigned fwengine = 0; \
+	REG_READ_HEADER(x); \
+	if (FORCEWAKE_VLV_RENDER_RANGE_OFFSET(reg)) { \
+		if (dev_priv->uncore.fw_rendercount == 0) \
+			fwengine = FORCEWAKE_RENDER; \
+	} else if (FORCEWAKE_VLV_MEDIA_RANGE_OFFSET(reg)) { \
+		if (dev_priv->uncore.fw_mediacount == 0) \
+			fwengine = FORCEWAKE_MEDIA; \
+	}  \
+	if (fwengine) \
+		dev_priv->uncore.funcs.force_wake_get(dev_priv, fwengine); \
+	val = __raw_i915_read##x(dev_priv, reg); \
+	if (fwengine) \
+		dev_priv->uncore.funcs.force_wake_put(dev_priv, fwengine); \
+	REG_READ_FOOTER; \
+}
+
+
+__vlv_read(8)
+__vlv_read(16)
+__vlv_read(32)
+__vlv_read(64)
 __gen6_read(8)
 __gen6_read(16)
 __gen6_read(32)
@@ -402,6 +591,7 @@ __gen4_read(16)
 __gen4_read(32)
 __gen4_read(64)
 
+#undef __vlv_read
 #undef __gen6_read
 #undef __gen5_read
 #undef __gen4_read
@@ -411,14 +601,18 @@ __gen4_read(64)
 #define REG_WRITE_HEADER \
 	unsigned long irqflags; \
 	trace_i915_reg_rw(true, reg, val, sizeof(val), trace); \
+	assert_device_not_suspended(dev_priv); \
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags)
 
+#define REG_WRITE_FOOTER \
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags)
+
 #define __gen4_write(x) \
 static void \
 gen4_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace) { \
 	REG_WRITE_HEADER; \
 	__raw_i915_write##x(dev_priv, reg, val); \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 #define __gen5_write(x) \
@@ -427,7 +621,7 @@ gen5_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace
 	REG_WRITE_HEADER; \
 	ilk_dummy_write(dev_priv); \
 	__raw_i915_write##x(dev_priv, reg, val); \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 #define __gen6_write(x) \
@@ -442,7 +636,7 @@ gen6_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace
 	if (unlikely(__fifo_ret)) { \
 		gen6_gt_check_fifodbg(dev_priv); \
 	} \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 #define __hsw_write(x) \
@@ -459,7 +653,7 @@ hsw_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace)
 		gen6_gt_check_fifodbg(dev_priv); \
 	} \
 	hsw_unclaimed_reg_check(dev_priv, reg); \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 static const u32 gen8_shadowed_regs[] = {
@@ -486,16 +680,19 @@ static bool is_gen8_shadowed(struct drm_i915_private *dev_priv, u32 reg)
 #define __gen8_write(x) \
 static void \
 gen8_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace) { \
-	bool __needs_put = !is_gen8_shadowed(dev_priv, reg); \
 	REG_WRITE_HEADER; \
-	if (__needs_put) { \
-		dev_priv->uncore.funcs.force_wake_get(dev_priv); \
-	} \
-	__raw_i915_write##x(dev_priv, reg, val); \
-	if (__needs_put) { \
-		dev_priv->uncore.funcs.force_wake_put(dev_priv); \
+	if (reg < 0x40000 && !is_gen8_shadowed(dev_priv, reg)) { \
+		if (dev_priv->uncore.forcewake_count == 0) \
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,	\
+							      FORCEWAKE_ALL); \
+		__raw_i915_write##x(dev_priv, reg, val); \
+		if (dev_priv->uncore.forcewake_count == 0) \
+			dev_priv->uncore.funcs.force_wake_put(dev_priv, \
+							      FORCEWAKE_ALL); \
+	} else { \
+		__raw_i915_write##x(dev_priv, reg, val); \
 	} \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 __gen8_write(8)
@@ -524,21 +721,24 @@ __gen4_write(64)
 #undef __gen6_write
 #undef __gen5_write
 #undef __gen4_write
+#undef REG_WRITE_FOOTER
 #undef REG_WRITE_HEADER
 
 void intel_uncore_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	INIT_DELAYED_WORK(&dev_priv->uncore.force_wake_work,
-			  gen6_force_wake_work);
+	setup_timer(&dev_priv->uncore.force_wake_timer,
+		    gen6_force_wake_timer, (unsigned long)dev_priv);
+
+	intel_uncore_early_sanitize(dev);
 
 	if (IS_VALLEYVIEW(dev)) {
-		dev_priv->uncore.funcs.force_wake_get = vlv_force_wake_get;
-		dev_priv->uncore.funcs.force_wake_put = vlv_force_wake_put;
+		dev_priv->uncore.funcs.force_wake_get = __vlv_force_wake_get;
+		dev_priv->uncore.funcs.force_wake_put = __vlv_force_wake_put;
 	} else if (IS_HASWELL(dev) || IS_GEN8(dev)) {
-		dev_priv->uncore.funcs.force_wake_get = __gen6_gt_force_wake_mt_get;
-		dev_priv->uncore.funcs.force_wake_put = __gen6_gt_force_wake_mt_put;
+		dev_priv->uncore.funcs.force_wake_get = __gen7_gt_force_wake_mt_get;
+		dev_priv->uncore.funcs.force_wake_put = __gen7_gt_force_wake_mt_put;
 	} else if (IS_IVYBRIDGE(dev)) {
 		u32 ecobus;
 
@@ -552,16 +752,16 @@ void intel_uncore_init(struct drm_device *dev)
 		 * forcewake being disabled.
 		 */
 		mutex_lock(&dev->struct_mutex);
-		__gen6_gt_force_wake_mt_get(dev_priv);
+		__gen7_gt_force_wake_mt_get(dev_priv, FORCEWAKE_ALL);
 		ecobus = __raw_i915_read32(dev_priv, ECOBUS);
-		__gen6_gt_force_wake_mt_put(dev_priv);
+		__gen7_gt_force_wake_mt_put(dev_priv, FORCEWAKE_ALL);
 		mutex_unlock(&dev->struct_mutex);
 
 		if (ecobus & FORCEWAKE_MT_ENABLE) {
 			dev_priv->uncore.funcs.force_wake_get =
-				__gen6_gt_force_wake_mt_get;
+				__gen7_gt_force_wake_mt_get;
 			dev_priv->uncore.funcs.force_wake_put =
-				__gen6_gt_force_wake_mt_put;
+				__gen7_gt_force_wake_mt_put;
 		} else {
 			DRM_INFO("No MT forcewake available on Ivybridge, this can result in issues\n");
 			DRM_INFO("when using vblank-synced partial screen updates.\n");
@@ -601,10 +801,18 @@ void intel_uncore_init(struct drm_device *dev)
 			dev_priv->uncore.funcs.mmio_writel  = gen6_write32;
 			dev_priv->uncore.funcs.mmio_writeq  = gen6_write64;
 		}
-		dev_priv->uncore.funcs.mmio_readb  = gen6_read8;
-		dev_priv->uncore.funcs.mmio_readw  = gen6_read16;
-		dev_priv->uncore.funcs.mmio_readl  = gen6_read32;
-		dev_priv->uncore.funcs.mmio_readq  = gen6_read64;
+
+		if (IS_VALLEYVIEW(dev)) {
+			dev_priv->uncore.funcs.mmio_readb  = vlv_read8;
+			dev_priv->uncore.funcs.mmio_readw  = vlv_read16;
+			dev_priv->uncore.funcs.mmio_readl  = vlv_read32;
+			dev_priv->uncore.funcs.mmio_readq  = vlv_read64;
+		} else {
+			dev_priv->uncore.funcs.mmio_readb  = gen6_read8;
+			dev_priv->uncore.funcs.mmio_readw  = gen6_read16;
+			dev_priv->uncore.funcs.mmio_readl  = gen6_read32;
+			dev_priv->uncore.funcs.mmio_readq  = gen6_read64;
+		}
 		break;
 	case 5:
 		dev_priv->uncore.funcs.mmio_writeb  = gen5_write8;
@@ -633,20 +841,20 @@ void intel_uncore_init(struct drm_device *dev)
 
 void intel_uncore_fini(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	flush_delayed_work(&dev_priv->uncore.force_wake_work);
-
 	/* Paranoia: make sure we have disabled everything before we exit. */
 	intel_uncore_sanitize(dev);
+	intel_uncore_forcewake_reset(dev, false);
 }
 
+#define GEN_RANGE(l, h) GENMASK(h, l)
+
 static const struct register_whitelist {
 	uint64_t offset;
 	uint32_t size;
-	uint32_t gen_bitmask; /* support gens, 0x10 for 4, 0x30 for 4 and 5, etc. */
+	/* supported gens, 0x10 for 4, 0x30 for 4 and 5, etc. */
+	uint32_t gen_bitmask;
 } whitelist[] = {
-	{ RING_TIMESTAMP(RENDER_RING_BASE), 8, 0xF0 },
+	{ RING_TIMESTAMP(RENDER_RING_BASE), 8, GEN_RANGE(4, 8) },
 };
 
 int i915_reg_read_ioctl(struct drm_device *dev,
@@ -655,7 +863,7 @@ int i915_reg_read_ioctl(struct drm_device *dev,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_reg_read *reg = data;
 	struct register_whitelist const *entry = whitelist;
-	int i;
+	int i, ret = 0;
 
 	for (i = 0; i < ARRAY_SIZE(whitelist); i++, entry++) {
 		if (entry->offset == reg->offset &&
@@ -666,6 +874,8 @@ int i915_reg_read_ioctl(struct drm_device *dev,
 	if (i == ARRAY_SIZE(whitelist))
 		return -EINVAL;
 
+	intel_runtime_pm_get(dev_priv);
+
 	switch (entry->size) {
 	case 8:
 		reg->val = I915_READ64(reg->offset);
@@ -681,8 +891,50 @@ int i915_reg_read_ioctl(struct drm_device *dev,
 		break;
 	default:
 		WARN_ON(1);
+		ret = -EINVAL;
+		goto out;
+	}
+
+out:
+	intel_runtime_pm_put(dev_priv);
+	return ret;
+}
+
+int i915_get_reset_stats_ioctl(struct drm_device *dev,
+			       void *data, struct drm_file *file)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_reset_stats *args = data;
+	struct i915_ctx_hang_stats *hs;
+	struct intel_context *ctx;
+	int ret;
+
+	if (args->flags || args->pad)
 		return -EINVAL;
+
+	if (args->ctx_id == DEFAULT_CONTEXT_ID && !capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	ctx = i915_gem_context_get(file->driver_priv, args->ctx_id);
+	if (IS_ERR(ctx)) {
+		mutex_unlock(&dev->struct_mutex);
+		return PTR_ERR(ctx);
 	}
+	hs = &ctx->hang_stats;
+
+	if (capable(CAP_SYS_ADMIN))
+		args->reset_count = i915_reset_count(&dev_priv->gpu_error);
+	else
+		args->reset_count = 0;
+
+	args->batch_active = hs->batch_active;
+	args->batch_pending = hs->batch_pending;
+
+	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
 }
@@ -698,6 +950,9 @@ static int i965_do_reset(struct drm_device *dev)
 {
 	int ret;
 
+	/* FIXME: i965g/gm need a display save/restore for gpu reset. */
+	return -ENODEV;
+
 	/*
 	 * Set the domains we want to reset (GRDOM/bits 2 and 3) as
 	 * well as the reset bit (GR/bit 0).  Setting the GR bit
@@ -709,7 +964,6 @@ static int i965_do_reset(struct drm_device *dev)
 	if (ret)
 		return ret;
 
-	/* We can't reset render&media without also resetting display ... */
 	pci_write_config_byte(dev->pdev, I965_GDRST,
 			      GRDOM_MEDIA | GRDOM_RESET_ENABLE);
 
@@ -722,38 +976,64 @@ static int i965_do_reset(struct drm_device *dev)
 	return 0;
 }
 
+static int g4x_do_reset(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
+
+	pci_write_config_byte(dev->pdev, I965_GDRST,
+			      GRDOM_RENDER | GRDOM_RESET_ENABLE);
+	ret =  wait_for(i965_reset_complete(dev), 500);
+	if (ret)
+		return ret;
+
+	/* WaVcpClkGateDisableForMediaReset:ctg,elk */
+	I915_WRITE(VDECCLK_GATE_D, I915_READ(VDECCLK_GATE_D) | VCP_UNIT_CLOCK_GATE_DISABLE);
+	POSTING_READ(VDECCLK_GATE_D);
+
+	pci_write_config_byte(dev->pdev, I965_GDRST,
+			      GRDOM_MEDIA | GRDOM_RESET_ENABLE);
+	ret =  wait_for(i965_reset_complete(dev), 500);
+	if (ret)
+		return ret;
+
+	/* WaVcpClkGateDisableForMediaReset:ctg,elk */
+	I915_WRITE(VDECCLK_GATE_D, I915_READ(VDECCLK_GATE_D) & ~VCP_UNIT_CLOCK_GATE_DISABLE);
+	POSTING_READ(VDECCLK_GATE_D);
+
+	pci_write_config_byte(dev->pdev, I965_GDRST, 0);
+
+	return 0;
+}
+
 static int ironlake_do_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 gdrst;
 	int ret;
 
-	gdrst = I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR);
-	gdrst &= ~GRDOM_MASK;
 	I915_WRITE(MCHBAR_MIRROR_BASE + ILK_GDSR,
-		   gdrst | GRDOM_RENDER | GRDOM_RESET_ENABLE);
-	ret = wait_for(I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR) & 0x1, 500);
+		   ILK_GRDOM_RENDER | ILK_GRDOM_RESET_ENABLE);
+	ret = wait_for((I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR) &
+			ILK_GRDOM_RESET_ENABLE) == 0, 500);
 	if (ret)
 		return ret;
 
-	/* We can't reset render&media without also resetting display ... */
-	gdrst = I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR);
-	gdrst &= ~GRDOM_MASK;
 	I915_WRITE(MCHBAR_MIRROR_BASE + ILK_GDSR,
-		   gdrst | GRDOM_MEDIA | GRDOM_RESET_ENABLE);
-	return wait_for(I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR) & 0x1, 500);
+		   ILK_GRDOM_MEDIA | ILK_GRDOM_RESET_ENABLE);
+	ret = wait_for((I915_READ(MCHBAR_MIRROR_BASE + ILK_GDSR) &
+			ILK_GRDOM_RESET_ENABLE) == 0, 500);
+	if (ret)
+		return ret;
+
+	I915_WRITE(MCHBAR_MIRROR_BASE + ILK_GDSR, 0);
+
+	return 0;
 }
 
 static int gen6_do_reset(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int	ret;
-	unsigned long irqflags;
-
-	/* Hold uncore.lock across reset to prevent any register access
-	 * with forcewake not set correctly
-	 */
-	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 
 	/* Reset the chip */
 
@@ -766,18 +1046,8 @@ static int gen6_do_reset(struct drm_device *dev)
 	/* Spin waiting for the device to ack the reset request */
 	ret = wait_for((__raw_i915_read32(dev_priv, GEN6_GDRST) & GEN6_GRDOM_FULL) == 0, 500);
 
-	intel_uncore_forcewake_reset(dev);
+	intel_uncore_forcewake_reset(dev, true);
 
-	/* If reset with a user forcewake, try to restore, otherwise turn it off */
-	if (dev_priv->uncore.forcewake_count)
-		dev_priv->uncore.funcs.force_wake_get(dev_priv);
-	else
-		dev_priv->uncore.funcs.force_wake_put(dev_priv);
-
-	/* Restore fifo count */
-	dev_priv->uncore.fifo_count = __raw_i915_read32(dev_priv, GT_FIFO_FREE_ENTRIES);
-
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 	return ret;
 }
 
@@ -788,20 +1058,15 @@ int intel_gpu_reset(struct drm_device *dev)
 	case 7:
 	case 6: return gen6_do_reset(dev);
 	case 5: return ironlake_do_reset(dev);
-	case 4: return i965_do_reset(dev);
+	case 4:
+		if (IS_G4X(dev))
+			return g4x_do_reset(dev);
+		else
+			return i965_do_reset(dev);
 	default: return -ENODEV;
 	}
 }
 
-void intel_uncore_clear_errors(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	/* XXX needs spinlock around caller's grouping */
-	if (HAS_FPGA_DBG_UNCLAIMED(dev))
-		__raw_i915_write32(dev_priv, FPGA_DBG, FPGA_DBG_RM_NOCLAIM);
-}
-
 void intel_uncore_check_errors(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
diff --git a/drivers/gpu/drm/mga/mga_dma.c b/drivers/gpu/drm/mga/mga_dma.c
index 087db33f6cf..c3bf059ba72 100644
--- a/drivers/gpu/drm/mga/mga_dma.c
+++ b/drivers/gpu/drm/mga/mga_dma.c
@@ -1075,10 +1075,10 @@ static int mga_dma_get_buffers(struct drm_device *dev,
 
 		buf->file_priv = file_priv;
 
-		if (DRM_COPY_TO_USER(&d->request_indices[i],
+		if (copy_to_user(&d->request_indices[i],
 				     &buf->idx, sizeof(buf->idx)))
 			return -EFAULT;
-		if (DRM_COPY_TO_USER(&d->request_sizes[i],
+		if (copy_to_user(&d->request_sizes[i],
 				     &buf->total, sizeof(buf->total)))
 			return -EFAULT;
 
diff --git a/drivers/gpu/drm/mga/mga_drv.h b/drivers/gpu/drm/mga/mga_drv.h
index ca4bc54ea21..fe453213600 100644
--- a/drivers/gpu/drm/mga/mga_drv.h
+++ b/drivers/gpu/drm/mga/mga_drv.h
@@ -186,14 +186,14 @@ extern void mga_disable_vblank(struct drm_device *dev, int crtc);
 extern u32 mga_get_vblank_counter(struct drm_device *dev, int crtc);
 extern int mga_driver_fence_wait(struct drm_device *dev, unsigned int *sequence);
 extern int mga_driver_vblank_wait(struct drm_device *dev, unsigned int *sequence);
-extern irqreturn_t mga_driver_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t mga_driver_irq_handler(int irq, void *arg);
 extern void mga_driver_irq_preinstall(struct drm_device *dev);
 extern int mga_driver_irq_postinstall(struct drm_device *dev);
 extern void mga_driver_irq_uninstall(struct drm_device *dev);
 extern long mga_compat_ioctl(struct file *filp, unsigned int cmd,
 			     unsigned long arg);
 
-#define mga_flush_write_combine()	DRM_WRITEMEMORYBARRIER()
+#define mga_flush_write_combine()	wmb()
 
 #define MGA_READ8(reg)		DRM_READ8(dev_priv->mmio, (reg))
 #define MGA_READ(reg)		DRM_READ32(dev_priv->mmio, (reg))
diff --git a/drivers/gpu/drm/mga/mga_ioc32.c b/drivers/gpu/drm/mga/mga_ioc32.c
index 709e90db8c4..729bfd56b55 100644
--- a/drivers/gpu/drm/mga/mga_ioc32.c
+++ b/drivers/gpu/drm/mga/mga_ioc32.c
@@ -34,6 +34,7 @@
 
 #include <drm/drmP.h>
 #include <drm/mga_drm.h>
+#include "mga_drv.h"
 
 typedef struct drm32_mga_init {
 	int func;
@@ -213,7 +214,7 @@ long mga_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	if (nr < DRM_COMMAND_BASE)
 		return drm_compat_ioctl(filp, cmd, arg);
 
-	if (nr < DRM_COMMAND_BASE + DRM_ARRAY_SIZE(mga_compat_ioctls))
+	if (nr < DRM_COMMAND_BASE + ARRAY_SIZE(mga_compat_ioctls))
 		fn = mga_compat_ioctls[nr - DRM_COMMAND_BASE];
 
 	if (fn != NULL)
diff --git a/drivers/gpu/drm/mga/mga_irq.c b/drivers/gpu/drm/mga/mga_irq.c
index 2b0ceb8dc11..1b071b8ff9d 100644
--- a/drivers/gpu/drm/mga/mga_irq.c
+++ b/drivers/gpu/drm/mga/mga_irq.c
@@ -47,7 +47,7 @@ u32 mga_get_vblank_counter(struct drm_device *dev, int crtc)
 }
 
 
-irqreturn_t mga_driver_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t mga_driver_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	drm_mga_private_t *dev_priv = (drm_mga_private_t *) dev->dev_private;
@@ -79,7 +79,7 @@ irqreturn_t mga_driver_irq_handler(DRM_IRQ_ARGS)
 			MGA_WRITE(MGA_PRIMEND, prim_end);
 
 		atomic_inc(&dev_priv->last_fence_retired);
-		DRM_WAKEUP(&dev_priv->fence_queue);
+		wake_up(&dev_priv->fence_queue);
 		handled = 1;
 	}
 
@@ -128,7 +128,7 @@ int mga_driver_fence_wait(struct drm_device *dev, unsigned int *sequence)
 	 * by about a day rather than she wants to wait for years
 	 * using fences.
 	 */
-	DRM_WAIT_ON(ret, dev_priv->fence_queue, 3 * DRM_HZ,
+	DRM_WAIT_ON(ret, dev_priv->fence_queue, 3 * HZ,
 		    (((cur_fence = atomic_read(&dev_priv->last_fence_retired))
 		      - *sequence) <= (1 << 23)));
 
@@ -151,7 +151,7 @@ int mga_driver_irq_postinstall(struct drm_device *dev)
 {
 	drm_mga_private_t *dev_priv = (drm_mga_private_t *) dev->dev_private;
 
-	DRM_INIT_WAITQUEUE(&dev_priv->fence_queue);
+	init_waitqueue_head(&dev_priv->fence_queue);
 
 	/* Turn on soft trap interrupt.  Vertical blank interrupts are enabled
 	 * in mga_enable_vblank.
diff --git a/drivers/gpu/drm/mga/mga_state.c b/drivers/gpu/drm/mga/mga_state.c
index 37cc2fb4ead..792f924496f 100644
--- a/drivers/gpu/drm/mga/mga_state.c
+++ b/drivers/gpu/drm/mga/mga_state.c
@@ -1020,7 +1020,7 @@ static int mga_getparam(struct drm_device *dev, void *data, struct drm_file *fil
 
 	switch (param->param) {
 	case MGA_PARAM_IRQ_NR:
-		value = drm_dev_to_irq(dev);
+		value = dev->pdev->irq;
 		break;
 	case MGA_PARAM_CARD_TYPE:
 		value = dev_priv->chipset;
@@ -1029,7 +1029,7 @@ static int mga_getparam(struct drm_device *dev, void *data, struct drm_file *fil
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_TO_USER(param->value, &value, sizeof(int))) {
+	if (copy_to_user(param->value, &value, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
@@ -1099,4 +1099,4 @@ const struct drm_ioctl_desc mga_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(MGA_DMA_BOOTSTRAP, mga_dma_bootstrap, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
 };
 
-int mga_max_ioctl = DRM_ARRAY_SIZE(mga_ioctls);
+int mga_max_ioctl = ARRAY_SIZE(mga_ioctls);
diff --git a/drivers/gpu/drm/mgag200/mgag200_cursor.c b/drivers/gpu/drm/mgag200/mgag200_cursor.c
index 801731aeab6..9f9780b7ddf 100644
--- a/drivers/gpu/drm/mgag200/mgag200_cursor.c
+++ b/drivers/gpu/drm/mgag200/mgag200_cursor.c
@@ -22,8 +22,10 @@ static void mga_hide_cursor(struct mga_device *mdev)
 {
 	WREG8(MGA_CURPOSXL, 0);
 	WREG8(MGA_CURPOSXH, 0);
-	mgag200_bo_unpin(mdev->cursor.pixels_1);
-	mgag200_bo_unpin(mdev->cursor.pixels_2);
+	if (mdev->cursor.pixels_1->pin_count)
+		mgag200_bo_unpin(mdev->cursor.pixels_1);
+	if (mdev->cursor.pixels_2->pin_count)
+		mgag200_bo_unpin(mdev->cursor.pixels_2);
 }
 
 int mga_crtc_cursor_set(struct drm_crtc *crtc,
@@ -32,7 +34,7 @@ int mga_crtc_cursor_set(struct drm_crtc *crtc,
 			uint32_t width,
 			uint32_t height)
 {
-	struct drm_device *dev = (struct drm_device *)file_priv->minor->dev;
+	struct drm_device *dev = crtc->dev;
 	struct mga_device *mdev = (struct mga_device *)dev->dev_private;
 	struct mgag200_bo *pixels_1 = mdev->cursor.pixels_1;
 	struct mgag200_bo *pixels_2 = mdev->cursor.pixels_2;
diff --git a/drivers/gpu/drm/mgag200/mgag200_fb.c b/drivers/gpu/drm/mgag200/mgag200_fb.c
index 964f58cee5e..13b7dd83faa 100644
--- a/drivers/gpu/drm/mgag200/mgag200_fb.c
+++ b/drivers/gpu/drm/mgag200/mgag200_fb.c
@@ -41,7 +41,7 @@ static void mga_dirty_update(struct mga_fbdev *mfbdev,
 	 * then the BO is being moved and we should
 	 * store up the damage until later.
 	 */
-	if (!in_interrupt())
+	if (drm_can_sleep())
 		ret = mgag200_bo_reserve(bo, true);
 	if (ret) {
 		if (ret != -EBUSY)
@@ -282,6 +282,11 @@ int mgag200_fbdev_init(struct mga_device *mdev)
 {
 	struct mga_fbdev *mfbdev;
 	int ret;
+	int bpp_sel = 32;
+
+	/* prefer 16bpp on low end gpus with limited VRAM */
+	if (IS_G200_SE(mdev) && mdev->mc.vram_size < (2048*1024))
+		bpp_sel = 16;
 
 	mfbdev = devm_kzalloc(mdev->dev->dev, sizeof(struct mga_fbdev), GFP_KERNEL);
 	if (!mfbdev)
@@ -301,7 +306,7 @@ int mgag200_fbdev_init(struct mga_device *mdev)
 	/* disable all the possible outputs/crtcs before entering KMS mode */
 	drm_helper_disable_unused_functions(mdev->dev);
 
-	drm_fb_helper_initial_config(&mfbdev->helper, 32);
+	drm_fb_helper_initial_config(&mfbdev->helper, bpp_sel);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/mgag200/mgag200_main.c b/drivers/gpu/drm/mgag200/mgag200_main.c
index b1120cb1db6..f6b283b8375 100644
--- a/drivers/gpu/drm/mgag200/mgag200_main.c
+++ b/drivers/gpu/drm/mgag200/mgag200_main.c
@@ -217,7 +217,10 @@ int mgag200_driver_load(struct drm_device *dev, unsigned long flags)
 
 	drm_mode_config_init(dev);
 	dev->mode_config.funcs = (void *)&mga_mode_funcs;
-	dev->mode_config.preferred_depth = 24;
+	if (IS_G200_SE(mdev) && mdev->mc.vram_size < (2048*1024))
+		dev->mode_config.preferred_depth = 16;
+	else
+		dev->mode_config.preferred_depth = 24;
 	dev->mode_config.prefer_shadow = 1;
 
 	r = mgag200_modeset_init(mdev);
@@ -310,7 +313,7 @@ int mgag200_dumb_create(struct drm_file *file,
 	return 0;
 }
 
-void mgag200_bo_unref(struct mgag200_bo **bo)
+static void mgag200_bo_unref(struct mgag200_bo **bo)
 {
 	struct ttm_buffer_object *tbo;
 
@@ -319,17 +322,13 @@ void mgag200_bo_unref(struct mgag200_bo **bo)
 
 	tbo = &((*bo)->bo);
 	ttm_bo_unref(&tbo);
-	if (tbo == NULL)
-		*bo = NULL;
-
+	*bo = NULL;
 }
 
 void mgag200_gem_free_object(struct drm_gem_object *obj)
 {
 	struct mgag200_bo *mgag200_bo = gem_to_mga_bo(obj);
 
-	if (!mgag200_bo)
-		return;
 	mgag200_bo_unref(&mgag200_bo);
 }
 
diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c
index ee6ed633b7b..a034ed40825 100644
--- a/drivers/gpu/drm/mgag200/mgag200_mode.c
+++ b/drivers/gpu/drm/mgag200/mgag200_mode.c
@@ -29,7 +29,7 @@ static void mga_crtc_load_lut(struct drm_crtc *crtc)
 	struct mga_crtc *mga_crtc = to_mga_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct mga_device *mdev = dev->dev_private;
-	struct drm_framebuffer *fb = crtc->fb;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	int i;
 
 	if (!crtc->enabled)
@@ -691,7 +691,7 @@ static void mga_g200wb_commit(struct drm_crtc *crtc)
    CRTCEXT0 has to be programmed last to trigger an update and make the
    new addr variable take effect.
  */
-void mga_set_start_address(struct drm_crtc *crtc, unsigned offset)
+static void mga_set_start_address(struct drm_crtc *crtc, unsigned offset)
 {
 	struct mga_device *mdev = crtc->dev->dev_private;
 	u32 addr;
@@ -742,7 +742,7 @@ static int mga_crtc_do_set_base(struct drm_crtc *crtc,
 		mgag200_bo_unreserve(bo);
 	}
 
-	mga_fb = to_mga_framebuffer(crtc->fb);
+	mga_fb = to_mga_framebuffer(crtc->primary->fb);
 	obj = mga_fb->obj;
 	bo = gem_to_mga_bo(obj);
 
@@ -805,7 +805,7 @@ static int mga_crtc_mode_set(struct drm_crtc *crtc,
 		/* 0x48: */        0,    0,    0,    0,    0,    0,    0,    0
 	};
 
-	bppshift = mdev->bpp_shifts[(crtc->fb->bits_per_pixel >> 3) - 1];
+	bppshift = mdev->bpp_shifts[(crtc->primary->fb->bits_per_pixel >> 3) - 1];
 
 	switch (mdev->type) {
 	case G200_SE_A:
@@ -843,12 +843,12 @@ static int mga_crtc_mode_set(struct drm_crtc *crtc,
 		break;
 	}
 
-	switch (crtc->fb->bits_per_pixel) {
+	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
 		dacvalue[MGA1064_MUL_CTL] = MGA1064_MUL_CTL_8bits;
 		break;
 	case 16:
-		if (crtc->fb->depth == 15)
+		if (crtc->primary->fb->depth == 15)
 			dacvalue[MGA1064_MUL_CTL] = MGA1064_MUL_CTL_15bits;
 		else
 			dacvalue[MGA1064_MUL_CTL] = MGA1064_MUL_CTL_16bits;
@@ -896,8 +896,8 @@ static int mga_crtc_mode_set(struct drm_crtc *crtc,
 	WREG_SEQ(3, 0);
 	WREG_SEQ(4, 0xe);
 
-	pitch = crtc->fb->pitches[0] / (crtc->fb->bits_per_pixel / 8);
-	if (crtc->fb->bits_per_pixel == 24)
+	pitch = crtc->primary->fb->pitches[0] / (crtc->primary->fb->bits_per_pixel / 8);
+	if (crtc->primary->fb->bits_per_pixel == 24)
 		pitch = (pitch * 3) >> (4 - bppshift);
 	else
 		pitch = pitch >> (4 - bppshift);
@@ -974,7 +974,7 @@ static int mga_crtc_mode_set(struct drm_crtc *crtc,
 		((vdisplay & 0xc00) >> 7) |
 		((vsyncstart & 0xc00) >> 5) |
 		((vdisplay & 0x400) >> 3);
-	if (crtc->fb->bits_per_pixel == 24)
+	if (crtc->primary->fb->bits_per_pixel == 24)
 		ext_vga[3] = (((1 << bppshift) * 3) - 1) | 0x80;
 	else
 		ext_vga[3] = ((1 << bppshift) - 1) | 0x80;
@@ -1034,9 +1034,9 @@ static int mga_crtc_mode_set(struct drm_crtc *crtc,
 			u32 bpp;
 			u32 mb;
 
-			if (crtc->fb->bits_per_pixel > 16)
+			if (crtc->primary->fb->bits_per_pixel > 16)
 				bpp = 32;
-			else if (crtc->fb->bits_per_pixel > 8)
+			else if (crtc->primary->fb->bits_per_pixel > 8)
 				bpp = 16;
 			else
 				bpp = 8;
@@ -1277,8 +1277,8 @@ static void mga_crtc_disable(struct drm_crtc *crtc)
 	int ret;
 	DRM_DEBUG_KMS("\n");
 	mga_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
-	if (crtc->fb) {
-		struct mga_framebuffer *mga_fb = to_mga_framebuffer(crtc->fb);
+	if (crtc->primary->fb) {
+		struct mga_framebuffer *mga_fb = to_mga_framebuffer(crtc->primary->fb);
 		struct drm_gem_object *obj = mga_fb->obj;
 		struct mgag200_bo *bo = gem_to_mga_bo(obj);
 		ret = mgag200_bo_reserve(bo, false);
@@ -1287,7 +1287,7 @@ static void mga_crtc_disable(struct drm_crtc *crtc)
 		mgag200_bo_push_sysram(bo);
 		mgag200_bo_unreserve(bo);
 	}
-	crtc->fb = NULL;
+	crtc->primary->fb = NULL;
 }
 
 /* These provide the minimum set of functions required to handle a CRTC */
@@ -1398,7 +1398,7 @@ static void mga_encoder_commit(struct drm_encoder *encoder)
 {
 }
 
-void mga_encoder_destroy(struct drm_encoder *encoder)
+static void mga_encoder_destroy(struct drm_encoder *encoder)
 {
 	struct mga_encoder *mga_encoder = to_mga_encoder(encoder);
 	drm_encoder_cleanup(encoder);
@@ -1519,11 +1519,11 @@ static int mga_vga_mode_valid(struct drm_connector *connector,
 		(mga_vga_calculate_mode_bandwidth(mode, bpp)
 			> (32700 * 1024))) {
 		return MODE_BANDWIDTH;
-	} else if (mode->type == G200_EH &&
+	} else if (mdev->type == G200_EH &&
 		(mga_vga_calculate_mode_bandwidth(mode, bpp)
 			> (37500 * 1024))) {
 		return MODE_BANDWIDTH;
-	} else if (mode->type == G200_ER &&
+	} else if (mdev->type == G200_ER &&
 		(mga_vga_calculate_mode_bandwidth(mode,
 			bpp) > (55000 * 1024))) {
 		return MODE_BANDWIDTH;
@@ -1558,7 +1558,7 @@ static int mga_vga_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
-struct drm_encoder *mga_connector_best_encoder(struct drm_connector
+static struct drm_encoder *mga_connector_best_encoder(struct drm_connector
 						  *connector)
 {
 	int enc_id = connector->encoder_ids[0];
diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c
index 07b192fe15c..5a00e90696d 100644
--- a/drivers/gpu/drm/mgag200/mgag200_ttm.c
+++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c
@@ -80,7 +80,7 @@ static int mgag200_ttm_global_init(struct mga_device *ast)
 	return 0;
 }
 
-void
+static void
 mgag200_ttm_global_release(struct mga_device *ast)
 {
 	if (ast->ttm.mem_global_ref.release == NULL)
@@ -102,7 +102,7 @@ static void mgag200_bo_ttm_destroy(struct ttm_buffer_object *tbo)
 	kfree(bo);
 }
 
-bool mgag200_ttm_bo_is_mgag200_bo(struct ttm_buffer_object *bo)
+static bool mgag200_ttm_bo_is_mgag200_bo(struct ttm_buffer_object *bo)
 {
 	if (bo->destroy == &mgag200_bo_ttm_destroy)
 		return true;
@@ -208,7 +208,7 @@ static struct ttm_backend_func mgag200_tt_backend_func = {
 };
 
 
-struct ttm_tt *mgag200_ttm_tt_create(struct ttm_bo_device *bdev,
+static struct ttm_tt *mgag200_ttm_tt_create(struct ttm_bo_device *bdev,
 				 unsigned long size, uint32_t page_flags,
 				 struct page *dummy_read_page)
 {
@@ -259,7 +259,9 @@ int mgag200_mm_init(struct mga_device *mdev)
 
 	ret = ttm_bo_device_init(&mdev->ttm.bdev,
 				 mdev->ttm.bo_global_ref.ref.object,
-				 &mgag200_bo_driver, DRM_FILE_PAGE_OFFSET,
+				 &mgag200_bo_driver,
+				 dev->anon_inode->i_mapping,
+				 DRM_FILE_PAGE_OFFSET,
 				 true);
 	if (ret) {
 		DRM_ERROR("Error initialising bo driver; %d\n", ret);
@@ -324,7 +326,6 @@ int mgag200_bo_create(struct drm_device *dev, int size, int align,
 	}
 
 	mgabo->bo.bdev = &mdev->ttm.bdev;
-	mgabo->bo.bdev->dev_mapping = dev->dev_mapping;
 
 	mgag200_ttm_placement(mgabo, TTM_PL_FLAG_VRAM | TTM_PL_FLAG_SYSTEM);
 
diff --git a/drivers/gpu/drm/msm/Kconfig b/drivers/gpu/drm/msm/Kconfig
index f39ab7554fc..f1238896785 100644
--- a/drivers/gpu/drm/msm/Kconfig
+++ b/drivers/gpu/drm/msm/Kconfig
@@ -2,8 +2,8 @@
 config DRM_MSM
 	tristate "MSM DRM"
 	depends on DRM
-	depends on ARCH_MSM
-	depends on ARCH_MSM8960
+	depends on MSM_IOMMU
+	depends on ARCH_QCOM || (ARM && COMPILE_TEST)
 	select DRM_KMS_HELPER
 	select SHMEM
 	select TMPFS
diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile
index e5fa12b0d21..93ca49c8df4 100644
--- a/drivers/gpu/drm/msm/Makefile
+++ b/drivers/gpu/drm/msm/Makefile
@@ -7,23 +7,35 @@ msm-y := \
 	adreno/adreno_gpu.o \
 	adreno/a3xx_gpu.o \
 	hdmi/hdmi.o \
+	hdmi/hdmi_audio.o \
 	hdmi/hdmi_bridge.o \
 	hdmi/hdmi_connector.o \
 	hdmi/hdmi_i2c.o \
 	hdmi/hdmi_phy_8960.o \
 	hdmi/hdmi_phy_8x60.o \
-	mdp4/mdp4_crtc.o \
-	mdp4/mdp4_dtv_encoder.o \
-	mdp4/mdp4_format.o \
-	mdp4/mdp4_irq.o \
-	mdp4/mdp4_kms.o \
-	mdp4/mdp4_plane.o \
+	hdmi/hdmi_phy_8x74.o \
+	mdp/mdp_format.o \
+	mdp/mdp_kms.o \
+	mdp/mdp4/mdp4_crtc.o \
+	mdp/mdp4/mdp4_dtv_encoder.o \
+	mdp/mdp4/mdp4_irq.o \
+	mdp/mdp4/mdp4_kms.o \
+	mdp/mdp4/mdp4_plane.o \
+	mdp/mdp5/mdp5_crtc.o \
+	mdp/mdp5/mdp5_encoder.o \
+	mdp/mdp5/mdp5_irq.o \
+	mdp/mdp5/mdp5_kms.o \
+	mdp/mdp5/mdp5_plane.o \
+	mdp/mdp5/mdp5_smp.o \
 	msm_drv.o \
 	msm_fb.o \
 	msm_gem.o \
 	msm_gem_prime.o \
 	msm_gem_submit.o \
 	msm_gpu.o \
+	msm_iommu.o \
+	msm_perf.o \
+	msm_rd.o \
 	msm_ringbuffer.o
 
 msm-$(CONFIG_DRM_MSM_FBDEV) += msm_fbdev.o
diff --git a/drivers/gpu/drm/msm/NOTES b/drivers/gpu/drm/msm/NOTES
index e036f6c1db9..9c4255b9802 100644
--- a/drivers/gpu/drm/msm/NOTES
+++ b/drivers/gpu/drm/msm/NOTES
@@ -4,7 +4,7 @@ In the current snapdragon SoC's, we have (at least) 3 different
 display controller blocks at play:
  + MDP3 - ?? seems to be what is on geeksphone peak device
  + MDP4 - S3 (APQ8060, touchpad), S4-pro (APQ8064, nexus4 & ifc6410)
- + MDSS - snapdragon 800
+ + MDP5 - snapdragon 800
 
 (I don't have a completely clear picture on which display controller
 maps to which part #)
@@ -46,6 +46,24 @@ and treat the MDP4 block's irq as "the" irq.  Even though the connectors
 may have their own irqs which they install themselves.  For this reason
 the display controller is the "master" device.
 
+For MDP5, the mapping is:
+
+  plane   -> PIPE{RGBn,VIGn}             \
+  crtc    -> LM (layer mixer)            |-> MDP "device"
+  encoder -> INTF                        /
+  connector -> HDMI/DSI/eDP/etc          --> other device(s)
+
+Unlike MDP4, it appears we can get by with a single encoder, rather
+than needing a different implementation for DTV, DSI, etc.  (Ie. the
+register interface is same, just different bases.)
+
+Also unlike MDP4, with MDP5 all the IRQs for other blocks (HDMI, DSI,
+etc) are routed through MDP.
+
+And finally, MDP5 has this "Shared Memory Pool" (called "SMP"), from
+which blocks need to be allocated to the active pipes based on fetch
+stride.
+
 Each connector probably ends up being a separate device, just for the
 logistics of finding/mapping io region, irq, etc.  Idealy we would
 have a better way than just stashing the platform device in a global
diff --git a/drivers/gpu/drm/msm/adreno/a2xx.xml.h b/drivers/gpu/drm/msm/adreno/a2xx.xml.h
index 9588098741b..85d615e7d62 100644
--- a/drivers/gpu/drm/msm/adreno/a2xx.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a2xx.xml.h
@@ -8,12 +8,13 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml              (    327 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/a2xx/a2xx.xml           (  31003 bytes, from 2013-09-19 18:50:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_common.xml       (   8983 bytes, from 2013-07-24 01:38:36)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_pm4.xml          (   9759 bytes, from 2013-09-10 00:52:33)
-- /home/robclark/src/freedreno/envytools/rnndb/a3xx/a3xx.xml           (  51983 bytes, from 2013-09-10 00:52:32)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
+- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32814 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   8900 bytes, from 2013-10-22 23:57:49)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  10574 bytes, from 2013-11-13 05:44:45)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  53644 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (   8344 bytes, from 2013-11-30 14:49:47)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -202,6 +203,12 @@ enum a2xx_rb_copy_sample_select {
 	SAMPLE_0123 = 6,
 };
 
+enum adreno_mmu_clnt_beh {
+	BEH_NEVR = 0,
+	BEH_TRAN_RNG = 1,
+	BEH_TRAN_FLT = 2,
+};
+
 enum sq_tex_clamp {
 	SQ_TEX_WRAP = 0,
 	SQ_TEX_MIRROR = 1,
@@ -238,6 +245,92 @@ enum sq_tex_filter {
 
 #define REG_A2XX_CP_PFP_UCODE_DATA				0x000000c1
 
+#define REG_A2XX_MH_MMU_CONFIG					0x00000040
+#define A2XX_MH_MMU_CONFIG_MMU_ENABLE				0x00000001
+#define A2XX_MH_MMU_CONFIG_SPLIT_MODE_ENABLE			0x00000002
+#define A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__MASK		0x00000030
+#define A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__SHIFT		4
+static inline uint32_t A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__MASK		0x000000c0
+#define A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__SHIFT		6
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__MASK		0x00000300
+#define A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__SHIFT		8
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__MASK		0x00000c00
+#define A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__SHIFT		10
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__MASK		0x00003000
+#define A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__SHIFT		12
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__MASK		0x0000c000
+#define A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__SHIFT		14
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__MASK		0x00030000
+#define A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__SHIFT		16
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__MASK		0x000c0000
+#define A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__SHIFT		18
+static inline uint32_t A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__MASK		0x00300000
+#define A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__SHIFT		20
+static inline uint32_t A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__MASK		0x00c00000
+#define A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__SHIFT		22
+static inline uint32_t A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__MASK		0x03000000
+#define A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__SHIFT		24
+static inline uint32_t A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__MASK;
+}
+
+#define REG_A2XX_MH_MMU_VA_RANGE				0x00000041
+
+#define REG_A2XX_MH_MMU_PT_BASE					0x00000042
+
+#define REG_A2XX_MH_MMU_PAGE_FAULT				0x00000043
+
+#define REG_A2XX_MH_MMU_TRAN_ERROR				0x00000044
+
+#define REG_A2XX_MH_MMU_INVALIDATE				0x00000045
+
+#define REG_A2XX_MH_MMU_MPU_BASE				0x00000046
+
+#define REG_A2XX_MH_MMU_MPU_END					0x00000047
+
+#define REG_A2XX_NQWAIT_UNTIL					0x00000394
+
 #define REG_A2XX_RBBM_PERFCOUNTER1_SELECT			0x00000395
 
 #define REG_A2XX_RBBM_PERFCOUNTER1_LO				0x00000397
@@ -276,20 +369,6 @@ enum sq_tex_filter {
 
 #define REG_A2XX_CP_PERFCOUNTER_HI				0x00000447
 
-#define REG_A2XX_CP_ST_BASE					0x0000044d
-
-#define REG_A2XX_CP_ST_BUFSZ					0x0000044e
-
-#define REG_A2XX_CP_IB1_BASE					0x00000458
-
-#define REG_A2XX_CP_IB1_BUFSZ					0x00000459
-
-#define REG_A2XX_CP_IB2_BASE					0x0000045a
-
-#define REG_A2XX_CP_IB2_BUFSZ					0x0000045b
-
-#define REG_A2XX_CP_STAT					0x0000047f
-
 #define REG_A2XX_RBBM_STATUS					0x000005d0
 #define A2XX_RBBM_STATUS_CMDFIFO_AVAIL__MASK			0x0000001f
 #define A2XX_RBBM_STATUS_CMDFIFO_AVAIL__SHIFT			0
@@ -808,6 +887,12 @@ static inline uint32_t A2XX_SQ_CONTEXT_MISC_PARAM_GEN_POS(uint32_t val)
 
 #define REG_A2XX_SQ_VS_PROGRAM					0x000021f7
 
+#define REG_A2XX_VGT_EVENT_INITIATOR				0x000021f9
+
+#define REG_A2XX_VGT_DRAW_INITIATOR				0x000021fc
+
+#define REG_A2XX_VGT_IMMED_DATA					0x000021fd
+
 #define REG_A2XX_RB_DEPTHCONTROL				0x00002200
 #define A2XX_RB_DEPTHCONTROL_STENCIL_ENABLE			0x00000001
 #define A2XX_RB_DEPTHCONTROL_Z_ENABLE				0x00000002
diff --git a/drivers/gpu/drm/msm/adreno/a3xx.xml.h b/drivers/gpu/drm/msm/adreno/a3xx.xml.h
index d4afdf65755..a7be56163d2 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a3xx.xml.h
@@ -8,12 +8,13 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml              (    327 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/a2xx/a2xx.xml           (  31003 bytes, from 2013-09-19 18:50:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_common.xml       (   8983 bytes, from 2013-07-24 01:38:36)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_pm4.xml          (   9759 bytes, from 2013-09-10 00:52:33)
-- /home/robclark/src/freedreno/envytools/rnndb/a3xx/a3xx.xml           (  51983 bytes, from 2013-09-10 00:52:32)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
+- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32814 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   8900 bytes, from 2013-10-22 23:57:49)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  10574 bytes, from 2013-11-13 05:44:45)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  53644 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (   8344 bytes, from 2013-11-30 14:49:47)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -292,6 +293,8 @@ enum a3xx_tex_type {
 #define A3XX_RBBM_STATUS_GPU_BUSY_NOHC				0x40000000
 #define A3XX_RBBM_STATUS_GPU_BUSY				0x80000000
 
+#define REG_A3XX_RBBM_NQWAIT_UNTIL				0x00000040
+
 #define REG_A3XX_RBBM_WAIT_IDLE_CLOCKS_CTL			0x00000033
 
 #define REG_A3XX_RBBM_INTERFACE_HANG_INT_CTL			0x00000050
@@ -304,6 +307,8 @@ enum a3xx_tex_type {
 
 #define REG_A3XX_RBBM_INTERFACE_HANG_MASK_CTL3			0x0000005a
 
+#define REG_A3XX_RBBM_INT_SET_CMD				0x00000060
+
 #define REG_A3XX_RBBM_INT_CLEAR_CMD				0x00000061
 
 #define REG_A3XX_RBBM_INT_0_MASK				0x00000063
@@ -937,13 +942,13 @@ static inline uint32_t A3XX_RB_BLEND_ALPHA_FLOAT(float val)
 	return ((util_float_to_half(val)) << A3XX_RB_BLEND_ALPHA_FLOAT__SHIFT) & A3XX_RB_BLEND_ALPHA_FLOAT__MASK;
 }
 
-#define REG_A3XX_UNKNOWN_20E8					0x000020e8
+#define REG_A3XX_RB_CLEAR_COLOR_DW0				0x000020e8
 
-#define REG_A3XX_UNKNOWN_20E9					0x000020e9
+#define REG_A3XX_RB_CLEAR_COLOR_DW1				0x000020e9
 
-#define REG_A3XX_UNKNOWN_20EA					0x000020ea
+#define REG_A3XX_RB_CLEAR_COLOR_DW2				0x000020ea
 
-#define REG_A3XX_UNKNOWN_20EB					0x000020eb
+#define REG_A3XX_RB_CLEAR_COLOR_DW3				0x000020eb
 
 #define REG_A3XX_RB_COPY_CONTROL				0x000020ec
 #define A3XX_RB_COPY_CONTROL_MSAA_RESOLVE__MASK			0x00000003
@@ -1026,7 +1031,7 @@ static inline uint32_t A3XX_RB_DEPTH_CONTROL_ZFUNC(enum adreno_compare_func val)
 #define A3XX_RB_DEPTH_CONTROL_BF_ENABLE				0x00000080
 #define A3XX_RB_DEPTH_CONTROL_Z_TEST_ENABLE			0x80000000
 
-#define REG_A3XX_UNKNOWN_2101					0x00002101
+#define REG_A3XX_RB_DEPTH_CLEAR					0x00002101
 
 #define REG_A3XX_RB_DEPTH_INFO					0x00002102
 #define A3XX_RB_DEPTH_INFO_DEPTH_FORMAT__MASK			0x00000001
@@ -1103,11 +1108,11 @@ static inline uint32_t A3XX_RB_STENCIL_CONTROL_ZFAIL_BF(enum adreno_stencil_op v
 	return ((val) << A3XX_RB_STENCIL_CONTROL_ZFAIL_BF__SHIFT) & A3XX_RB_STENCIL_CONTROL_ZFAIL_BF__MASK;
 }
 
-#define REG_A3XX_UNKNOWN_2105					0x00002105
+#define REG_A3XX_RB_STENCIL_CLEAR				0x00002105
 
-#define REG_A3XX_UNKNOWN_2106					0x00002106
+#define REG_A3XX_RB_STENCIL_BUF_INFO				0x00002106
 
-#define REG_A3XX_UNKNOWN_2107					0x00002107
+#define REG_A3XX_RB_STENCIL_BUF_PITCH				0x00002107
 
 #define REG_A3XX_RB_STENCILREFMASK				0x00002108
 #define A3XX_RB_STENCILREFMASK_STENCILREF__MASK			0x000000ff
@@ -1149,20 +1154,31 @@ static inline uint32_t A3XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK(uint32_t val)
 	return ((val) << A3XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK__SHIFT) & A3XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK__MASK;
 }
 
-#define REG_A3XX_PA_SC_WINDOW_OFFSET				0x0000210e
-#define A3XX_PA_SC_WINDOW_OFFSET_X__MASK			0x0000ffff
-#define A3XX_PA_SC_WINDOW_OFFSET_X__SHIFT			0
-static inline uint32_t A3XX_PA_SC_WINDOW_OFFSET_X(uint32_t val)
+#define REG_A3XX_RB_LRZ_VSC_CONTROL				0x0000210c
+#define A3XX_RB_LRZ_VSC_CONTROL_BINNING_ENABLE			0x00000002
+
+#define REG_A3XX_RB_WINDOW_OFFSET				0x0000210e
+#define A3XX_RB_WINDOW_OFFSET_X__MASK				0x0000ffff
+#define A3XX_RB_WINDOW_OFFSET_X__SHIFT				0
+static inline uint32_t A3XX_RB_WINDOW_OFFSET_X(uint32_t val)
 {
-	return ((val) << A3XX_PA_SC_WINDOW_OFFSET_X__SHIFT) & A3XX_PA_SC_WINDOW_OFFSET_X__MASK;
+	return ((val) << A3XX_RB_WINDOW_OFFSET_X__SHIFT) & A3XX_RB_WINDOW_OFFSET_X__MASK;
 }
-#define A3XX_PA_SC_WINDOW_OFFSET_Y__MASK			0xffff0000
-#define A3XX_PA_SC_WINDOW_OFFSET_Y__SHIFT			16
-static inline uint32_t A3XX_PA_SC_WINDOW_OFFSET_Y(uint32_t val)
+#define A3XX_RB_WINDOW_OFFSET_Y__MASK				0xffff0000
+#define A3XX_RB_WINDOW_OFFSET_Y__SHIFT				16
+static inline uint32_t A3XX_RB_WINDOW_OFFSET_Y(uint32_t val)
 {
-	return ((val) << A3XX_PA_SC_WINDOW_OFFSET_Y__SHIFT) & A3XX_PA_SC_WINDOW_OFFSET_Y__MASK;
+	return ((val) << A3XX_RB_WINDOW_OFFSET_Y__SHIFT) & A3XX_RB_WINDOW_OFFSET_Y__MASK;
 }
 
+#define REG_A3XX_RB_SAMPLE_COUNT_CONTROL			0x00002110
+
+#define REG_A3XX_RB_SAMPLE_COUNT_ADDR				0x00002111
+
+#define REG_A3XX_RB_Z_CLAMP_MIN					0x00002114
+
+#define REG_A3XX_RB_Z_CLAMP_MAX					0x00002115
+
 #define REG_A3XX_PC_VSTREAM_CONTROL				0x000021e4
 
 #define REG_A3XX_PC_VERTEX_REUSE_BLOCK_CNTL			0x000021ea
@@ -1309,6 +1325,8 @@ static inline uint32_t A3XX_HLSQ_CONST_FSPRESV_RANGE_REG_ENDENTRY(uint32_t val)
 
 #define REG_A3XX_HLSQ_CL_KERNEL_GROUP_X_REG			0x00002215
 
+#define REG_A3XX_HLSQ_CL_KERNEL_GROUP_Y_REG			0x00002216
+
 #define REG_A3XX_HLSQ_CL_KERNEL_GROUP_Z_REG			0x00002217
 
 #define REG_A3XX_HLSQ_CL_WG_OFFSET_REG				0x0000221a
@@ -1491,12 +1509,13 @@ static inline uint32_t REG_A3XX_VPC_VARYING_PS_REPL_MODE(uint32_t i0) { return 0
 
 #define REG_A3XX_SP_SP_CTRL_REG					0x000022c0
 #define A3XX_SP_SP_CTRL_REG_RESOLVE				0x00010000
-#define A3XX_SP_SP_CTRL_REG_CONSTMODE__MASK			0x000c0000
+#define A3XX_SP_SP_CTRL_REG_CONSTMODE__MASK			0x00040000
 #define A3XX_SP_SP_CTRL_REG_CONSTMODE__SHIFT			18
 static inline uint32_t A3XX_SP_SP_CTRL_REG_CONSTMODE(uint32_t val)
 {
 	return ((val) << A3XX_SP_SP_CTRL_REG_CONSTMODE__SHIFT) & A3XX_SP_SP_CTRL_REG_CONSTMODE__MASK;
 }
+#define A3XX_SP_SP_CTRL_REG_BINNING				0x00080000
 #define A3XX_SP_SP_CTRL_REG_SLEEPMODE__MASK			0x00300000
 #define A3XX_SP_SP_CTRL_REG_SLEEPMODE__SHIFT			20
 static inline uint32_t A3XX_SP_SP_CTRL_REG_SLEEPMODE(uint32_t val)
@@ -1669,7 +1688,7 @@ static inline uint32_t A3XX_SP_VS_OBJ_OFFSET_REG_SHADEROBJOFFSET(uint32_t val)
 
 #define REG_A3XX_SP_VS_OBJ_START_REG				0x000022d5
 
-#define REG_A3XX_SP_VS_PVT_MEM_CTRL_REG				0x000022d6
+#define REG_A3XX_SP_VS_PVT_MEM_PARAM_REG			0x000022d6
 
 #define REG_A3XX_SP_VS_PVT_MEM_ADDR_REG				0x000022d7
 
@@ -1772,7 +1791,7 @@ static inline uint32_t A3XX_SP_FS_OBJ_OFFSET_REG_SHADEROBJOFFSET(uint32_t val)
 
 #define REG_A3XX_SP_FS_OBJ_START_REG				0x000022e3
 
-#define REG_A3XX_SP_FS_PVT_MEM_CTRL_REG				0x000022e4
+#define REG_A3XX_SP_FS_PVT_MEM_PARAM_REG			0x000022e4
 
 #define REG_A3XX_SP_FS_PVT_MEM_ADDR_REG				0x000022e5
 
@@ -1943,6 +1962,9 @@ static inline uint32_t REG_A3XX_VSC_PIPE_DATA_ADDRESS(uint32_t i0) { return 0x00
 
 static inline uint32_t REG_A3XX_VSC_PIPE_DATA_LENGTH(uint32_t i0) { return 0x00000c08 + 0x3*i0; }
 
+#define REG_A3XX_VSC_BIN_CONTROL				0x00000c3c
+#define A3XX_VSC_BIN_CONTROL_BINNING_ENABLE			0x00000001
+
 #define REG_A3XX_UNKNOWN_0C3D					0x00000c3d
 
 #define REG_A3XX_PC_PERFCOUNTER0_SELECT				0x00000c48
@@ -1953,7 +1975,7 @@ static inline uint32_t REG_A3XX_VSC_PIPE_DATA_LENGTH(uint32_t i0) { return 0x000
 
 #define REG_A3XX_PC_PERFCOUNTER3_SELECT				0x00000c4b
 
-#define REG_A3XX_UNKNOWN_0C81					0x00000c81
+#define REG_A3XX_GRAS_TSE_DEBUG_ECO				0x00000c81
 
 #define REG_A3XX_GRAS_PERFCOUNTER0_SELECT			0x00000c88
 
@@ -1975,22 +1997,24 @@ static inline uint32_t REG_A3XX_GRAS_CL_USER_PLANE_W(uint32_t i0) { return 0x000
 
 #define REG_A3XX_RB_GMEM_BASE_ADDR				0x00000cc0
 
+#define REG_A3XX_RB_DEBUG_ECO_CONTROLS_ADDR			0x00000cc1
+
 #define REG_A3XX_RB_PERFCOUNTER0_SELECT				0x00000cc6
 
 #define REG_A3XX_RB_PERFCOUNTER1_SELECT				0x00000cc7
 
-#define REG_A3XX_RB_WINDOW_SIZE					0x00000ce0
-#define A3XX_RB_WINDOW_SIZE_WIDTH__MASK				0x00003fff
-#define A3XX_RB_WINDOW_SIZE_WIDTH__SHIFT			0
-static inline uint32_t A3XX_RB_WINDOW_SIZE_WIDTH(uint32_t val)
+#define REG_A3XX_RB_FRAME_BUFFER_DIMENSION			0x00000ce0
+#define A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__MASK		0x00003fff
+#define A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__SHIFT		0
+static inline uint32_t A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH(uint32_t val)
 {
-	return ((val) << A3XX_RB_WINDOW_SIZE_WIDTH__SHIFT) & A3XX_RB_WINDOW_SIZE_WIDTH__MASK;
+	return ((val) << A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__SHIFT) & A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__MASK;
 }
-#define A3XX_RB_WINDOW_SIZE_HEIGHT__MASK			0x0fffc000
-#define A3XX_RB_WINDOW_SIZE_HEIGHT__SHIFT			14
-static inline uint32_t A3XX_RB_WINDOW_SIZE_HEIGHT(uint32_t val)
+#define A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__MASK		0x0fffc000
+#define A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__SHIFT		14
+static inline uint32_t A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT(uint32_t val)
 {
-	return ((val) << A3XX_RB_WINDOW_SIZE_HEIGHT__SHIFT) & A3XX_RB_WINDOW_SIZE_HEIGHT__MASK;
+	return ((val) << A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__SHIFT) & A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__MASK;
 }
 
 #define REG_A3XX_HLSQ_PERFCOUNTER0_SELECT			0x00000e00
@@ -2088,6 +2112,14 @@ static inline uint32_t A3XX_UCHE_CACHE_INVALIDATE1_REG_OPCODE(enum a3xx_cache_op
 
 #define REG_A3XX_TP_PERFCOUNTER5_SELECT				0x00000f09
 
+#define REG_A3XX_VGT_CL_INITIATOR				0x000021f0
+
+#define REG_A3XX_VGT_EVENT_INITIATOR				0x000021f9
+
+#define REG_A3XX_VGT_DRAW_INITIATOR				0x000021fc
+
+#define REG_A3XX_VGT_IMMED_DATA					0x000021fd
+
 #define REG_A3XX_TEX_SAMP_0					0x00000000
 #define A3XX_TEX_SAMP_0_MIPFILTER_LINEAR			0x00000002
 #define A3XX_TEX_SAMP_0_XY_MAG__MASK				0x0000000c
@@ -2123,6 +2155,18 @@ static inline uint32_t A3XX_TEX_SAMP_0_WRAP_R(enum a3xx_tex_clamp val)
 #define A3XX_TEX_SAMP_0_UNNORM_COORDS				0x80000000
 
 #define REG_A3XX_TEX_SAMP_1					0x00000001
+#define A3XX_TEX_SAMP_1_MAX_LOD__MASK				0x003ff000
+#define A3XX_TEX_SAMP_1_MAX_LOD__SHIFT				12
+static inline uint32_t A3XX_TEX_SAMP_1_MAX_LOD(float val)
+{
+	return ((((uint32_t)(val * 12.0))) << A3XX_TEX_SAMP_1_MAX_LOD__SHIFT) & A3XX_TEX_SAMP_1_MAX_LOD__MASK;
+}
+#define A3XX_TEX_SAMP_1_MIN_LOD__MASK				0xffc00000
+#define A3XX_TEX_SAMP_1_MIN_LOD__SHIFT				22
+static inline uint32_t A3XX_TEX_SAMP_1_MIN_LOD(float val)
+{
+	return ((((uint32_t)(val * 12.0))) << A3XX_TEX_SAMP_1_MIN_LOD__SHIFT) & A3XX_TEX_SAMP_1_MIN_LOD__MASK;
+}
 
 #define REG_A3XX_TEX_CONST_0					0x00000000
 #define A3XX_TEX_CONST_0_TILED					0x00000001
diff --git a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
index 035bd13dc8b..942e09d898a 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
@@ -15,6 +15,10 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
+#ifdef CONFIG_MSM_OCMEM
+#  include <mach/ocmem.h>
+#endif
+
 #include "a3xx_gpu.h"
 
 #define A3XX_INT0_MASK \
@@ -31,7 +35,11 @@
 	 A3XX_INT0_CP_AHB_ERROR_HALT |     \
 	 A3XX_INT0_UCHE_OOB_ACCESS)
 
-static struct platform_device *a3xx_pdev;
+
+static bool hang_debug = false;
+MODULE_PARM_DESC(hang_debug, "Dump registers when hang is detected (can be slow!)");
+module_param_named(hang_debug, hang_debug, bool, 0600);
+static void a3xx_dump(struct msm_gpu *gpu);
 
 static void a3xx_me_init(struct msm_gpu *gpu)
 {
@@ -63,6 +71,7 @@ static void a3xx_me_init(struct msm_gpu *gpu)
 static int a3xx_hw_init(struct msm_gpu *gpu)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	struct a3xx_gpu *a3xx_gpu = to_a3xx_gpu(adreno_gpu);
 	uint32_t *ptr, len;
 	int i, ret;
 
@@ -105,6 +114,21 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT, 0x000000ff);
 		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT_CONF, 0x000000a4);
 
+	} else if (adreno_is_a330v2(adreno_gpu)) {
+		/*
+		 * Most of the VBIF registers on 8974v2 have the correct
+		 * values at power on, so we won't modify those if we don't
+		 * need to
+		 */
+		/* Enable 1k sort: */
+		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT, 0x0001003f);
+		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT_CONF, 0x000000a4);
+		/* Enable WR-REQ: */
+		gpu_write(gpu, REG_A3XX_VBIF_GATE_OFF_WRREQ_EN, 0x00003f);
+		gpu_write(gpu, REG_A3XX_VBIF_DDR_OUT_MAX_BURST, 0x0000303);
+		/* Set up VBIF_ROUND_ROBIN_QOS_ARB: */
+		gpu_write(gpu, REG_A3XX_VBIF_ROUND_ROBIN_QOS_ARB, 0x0003);
+
 	} else if (adreno_is_a330(adreno_gpu)) {
 		/* Set up 16 deep read/write request queues: */
 		gpu_write(gpu, REG_A3XX_VBIF_IN_RD_LIM_CONF0, 0x18181818);
@@ -121,10 +145,10 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 		/* Set up VBIF_ROUND_ROBIN_QOS_ARB: */
 		gpu_write(gpu, REG_A3XX_VBIF_ROUND_ROBIN_QOS_ARB, 0x0001);
 		/* Set up AOOO: */
-		gpu_write(gpu, REG_A3XX_VBIF_OUT_AXI_AOOO_EN, 0x0000ffff);
-		gpu_write(gpu, REG_A3XX_VBIF_OUT_AXI_AOOO, 0xffffffff);
+		gpu_write(gpu, REG_A3XX_VBIF_OUT_AXI_AOOO_EN, 0x0000003f);
+		gpu_write(gpu, REG_A3XX_VBIF_OUT_AXI_AOOO, 0x003f003f);
 		/* Enable 1K sort: */
-		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT, 0x0001ffff);
+		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT, 0x0001003f);
 		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT_CONF, 0x000000a4);
 		/* Disable VBIF clock gating. This is to enable AXI running
 		 * higher frequency than GPU:
@@ -162,23 +186,32 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 	gpu_write(gpu, REG_A3XX_UCHE_CACHE_MODE_CONTROL_REG, 0x00000001);
 
 	/* Enable Clock gating: */
-	gpu_write(gpu, REG_A3XX_RBBM_CLOCK_CTL, 0xbfffffff);
-
-	/* Set the OCMEM base address for A330 */
-//TODO:
-//	if (adreno_is_a330(adreno_gpu)) {
-//		gpu_write(gpu, REG_A3XX_RB_GMEM_BASE_ADDR,
-//			(unsigned int)(a3xx_gpu->ocmem_base >> 14));
-//	}
+	if (adreno_is_a320(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_CLOCK_CTL, 0xbfffffff);
+	else if (adreno_is_a330v2(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_CLOCK_CTL, 0xaaaaaaaa);
+	else if (adreno_is_a330(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_CLOCK_CTL, 0xbffcffff);
+
+	if (adreno_is_a330v2(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_GPR0_CTL, 0x05515455);
+	else if (adreno_is_a330(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_GPR0_CTL, 0x00000000);
+
+	/* Set the OCMEM base address for A330, etc */
+	if (a3xx_gpu->ocmem_hdl) {
+		gpu_write(gpu, REG_A3XX_RB_GMEM_BASE_ADDR,
+			(unsigned int)(a3xx_gpu->ocmem_base >> 14));
+	}
 
 	/* Turn on performance counters: */
 	gpu_write(gpu, REG_A3XX_RBBM_PERFCTR_CTL, 0x01);
 
-	/* Set SP perfcounter 7 to count SP_FS_FULL_ALU_INSTRUCTIONS
-	 * we will use this to augment our hang detection:
-	 */
-	gpu_write(gpu, REG_A3XX_SP_PERFCOUNTER7_SELECT,
-			SP_FS_FULL_ALU_INSTRUCTIONS);
+	/* Enable the perfcntrs that we use.. */
+	for (i = 0; i < gpu->num_perfcntrs; i++) {
+		const struct msm_gpu_perfcntr *perfcntr = &gpu->perfcntrs[i];
+		gpu_write(gpu, perfcntr->select_reg, perfcntr->select_val);
+	}
 
 	gpu_write(gpu, REG_A3XX_RBBM_INT_0_MASK, A3XX_INT0_MASK);
 
@@ -219,7 +252,7 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 	/* Load PM4: */
 	ptr = (uint32_t *)(adreno_gpu->pm4->data);
 	len = adreno_gpu->pm4->size / 4;
-	DBG("loading PM4 ucode version: %u", ptr[0]);
+	DBG("loading PM4 ucode version: %x", ptr[1]);
 
 	gpu_write(gpu, REG_AXXX_CP_DEBUG,
 			AXXX_CP_DEBUG_DYNAMIC_CLK_DISABLE |
@@ -231,19 +264,26 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 	/* Load PFP: */
 	ptr = (uint32_t *)(adreno_gpu->pfp->data);
 	len = adreno_gpu->pfp->size / 4;
-	DBG("loading PFP ucode version: %u", ptr[0]);
+	DBG("loading PFP ucode version: %x", ptr[5]);
 
 	gpu_write(gpu, REG_A3XX_CP_PFP_UCODE_ADDR, 0);
 	for (i = 1; i < len; i++)
 		gpu_write(gpu, REG_A3XX_CP_PFP_UCODE_DATA, ptr[i]);
 
 	/* CP ROQ queue sizes (bytes) - RB:16, ST:16, IB1:32, IB2:64 */
-	if (adreno_is_a305(adreno_gpu) || adreno_is_a320(adreno_gpu))
+	if (adreno_is_a305(adreno_gpu) || adreno_is_a320(adreno_gpu)) {
 		gpu_write(gpu, REG_AXXX_CP_QUEUE_THRESHOLDS,
 				AXXX_CP_QUEUE_THRESHOLDS_CSQ_IB1_START(2) |
 				AXXX_CP_QUEUE_THRESHOLDS_CSQ_IB2_START(6) |
 				AXXX_CP_QUEUE_THRESHOLDS_CSQ_ST_START(14));
-
+	} else if (adreno_is_a330(adreno_gpu)) {
+		/* NOTE: this (value take from downstream android driver)
+		 * includes some bits outside of the known bitfields.  But
+		 * A330 has this "MERCIU queue" thing too, which might
+		 * explain a new bitfield or reshuffling:
+		 */
+		gpu_write(gpu, REG_AXXX_CP_QUEUE_THRESHOLDS, 0x003e2008);
+	}
 
 	/* clear ME_HALT to start micro engine */
 	gpu_write(gpu, REG_AXXX_CP_ME_CNTL, 0);
@@ -253,6 +293,17 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 	return 0;
 }
 
+static void a3xx_recover(struct msm_gpu *gpu)
+{
+	/* dump registers before resetting gpu, if enabled: */
+	if (hang_debug)
+		a3xx_dump(gpu);
+	gpu_write(gpu, REG_A3XX_RBBM_SW_RESET_CMD, 1);
+	gpu_read(gpu, REG_A3XX_RBBM_SW_RESET_CMD);
+	gpu_write(gpu, REG_A3XX_RBBM_SW_RESET_CMD, 0);
+	adreno_recover(gpu);
+}
+
 static void a3xx_destroy(struct msm_gpu *gpu)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
@@ -261,27 +312,24 @@ static void a3xx_destroy(struct msm_gpu *gpu)
 	DBG("%s", gpu->name);
 
 	adreno_gpu_cleanup(adreno_gpu);
-	put_device(&a3xx_gpu->pdev->dev);
+
+#ifdef CONFIG_MSM_OCMEM
+	if (a3xx_gpu->ocmem_base)
+		ocmem_free(OCMEM_GRAPHICS, a3xx_gpu->ocmem_hdl);
+#endif
+
 	kfree(a3xx_gpu);
 }
 
 static void a3xx_idle(struct msm_gpu *gpu)
 {
-	unsigned long t;
-
 	/* wait for ringbuffer to drain: */
 	adreno_idle(gpu);
 
-	t = jiffies + ADRENO_IDLE_TIMEOUT;
-
 	/* then wait for GPU to finish: */
-	do {
-		uint32_t rbbm_status = gpu_read(gpu, REG_A3XX_RBBM_STATUS);
-		if (!(rbbm_status & A3XX_RBBM_STATUS_GPU_BUSY))
-			return;
-	} while(time_before(jiffies, t));
-
-	DRM_ERROR("timeout waiting for %s to idle!\n", gpu->name);
+	if (spin_until(!(gpu_read(gpu, REG_A3XX_RBBM_STATUS) &
+			A3XX_RBBM_STATUS_GPU_BUSY)))
+		DRM_ERROR("%s: timeout waiting for GPU to idle!\n", gpu->name);
 
 	/* TODO maybe we need to reset GPU here to recover from hang? */
 }
@@ -302,7 +350,6 @@ static irqreturn_t a3xx_irq(struct msm_gpu *gpu)
 	return IRQ_HANDLED;
 }
 
-#ifdef CONFIG_DEBUG_FS
 static const unsigned int a3xx_registers[] = {
 	0x0000, 0x0002, 0x0010, 0x0012, 0x0018, 0x0018, 0x0020, 0x0027,
 	0x0029, 0x002b, 0x002e, 0x0033, 0x0040, 0x0042, 0x0050, 0x005c,
@@ -342,11 +389,18 @@ static const unsigned int a3xx_registers[] = {
 	0x303c, 0x303c, 0x305e, 0x305f,
 };
 
+#ifdef CONFIG_DEBUG_FS
 static void a3xx_show(struct msm_gpu *gpu, struct seq_file *m)
 {
+	struct drm_device *dev = gpu->dev;
 	int i;
 
 	adreno_show(gpu, m);
+
+	mutex_lock(&dev->struct_mutex);
+
+	gpu->funcs->pm_resume(gpu);
+
 	seq_printf(m, "status:   %08x\n",
 			gpu_read(gpu, REG_A3XX_RBBM_STATUS));
 
@@ -362,16 +416,43 @@ static void a3xx_show(struct msm_gpu *gpu, struct seq_file *m)
 			seq_printf(m, "IO:R %08x %08x\n", addr<<2, val);
 		}
 	}
+
+	gpu->funcs->pm_suspend(gpu);
+
+	mutex_unlock(&dev->struct_mutex);
 }
 #endif
 
+/* would be nice to not have to duplicate the _show() stuff with printk(): */
+static void a3xx_dump(struct msm_gpu *gpu)
+{
+	int i;
+
+	adreno_dump(gpu);
+	printk("status:   %08x\n",
+			gpu_read(gpu, REG_A3XX_RBBM_STATUS));
+
+	/* dump these out in a form that can be parsed by demsm: */
+	printk("IO:region %s 00000000 00020000\n", gpu->name);
+	for (i = 0; i < ARRAY_SIZE(a3xx_registers); i += 2) {
+		uint32_t start = a3xx_registers[i];
+		uint32_t end   = a3xx_registers[i+1];
+		uint32_t addr;
+
+		for (addr = start; addr <= end; addr++) {
+			uint32_t val = gpu_read(gpu, addr);
+			printk("IO:R %08x %08x\n", addr<<2, val);
+		}
+	}
+}
+
 static const struct adreno_gpu_funcs funcs = {
 	.base = {
 		.get_param = adreno_get_param,
 		.hw_init = a3xx_hw_init,
 		.pm_suspend = msm_gpu_pm_suspend,
 		.pm_resume = msm_gpu_pm_resume,
-		.recover = adreno_recover,
+		.recover = a3xx_recover,
 		.last_fence = adreno_last_fence,
 		.submit = adreno_submit,
 		.flush = adreno_flush,
@@ -384,11 +465,20 @@ static const struct adreno_gpu_funcs funcs = {
 	},
 };
 
+static const struct msm_gpu_perfcntr perfcntrs[] = {
+	{ REG_A3XX_SP_PERFCOUNTER6_SELECT, REG_A3XX_RBBM_PERFCTR_SP_6_LO,
+			SP_ALU_ACTIVE_CYCLES, "ALUACTIVE" },
+	{ REG_A3XX_SP_PERFCOUNTER7_SELECT, REG_A3XX_RBBM_PERFCTR_SP_7_LO,
+			SP_FS_FULL_ALU_INSTRUCTIONS, "ALUFULL" },
+};
+
 struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 {
 	struct a3xx_gpu *a3xx_gpu = NULL;
+	struct adreno_gpu *adreno_gpu;
 	struct msm_gpu *gpu;
-	struct platform_device *pdev = a3xx_pdev;
+	struct msm_drm_private *priv = dev->dev_private;
+	struct platform_device *pdev = priv->gpu_pdev;
 	struct adreno_platform_config *config;
 	int ret;
 
@@ -406,24 +496,57 @@ struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 		goto fail;
 	}
 
-	gpu = &a3xx_gpu->base.base;
+	adreno_gpu = &a3xx_gpu->base;
+	gpu = &adreno_gpu->base;
 
-	get_device(&pdev->dev);
 	a3xx_gpu->pdev = pdev;
 
 	gpu->fast_rate = config->fast_rate;
 	gpu->slow_rate = config->slow_rate;
 	gpu->bus_freq  = config->bus_freq;
+#ifdef CONFIG_MSM_BUS_SCALING
+	gpu->bus_scale_table = config->bus_scale_table;
+#endif
 
 	DBG("fast_rate=%u, slow_rate=%u, bus_freq=%u",
 			gpu->fast_rate, gpu->slow_rate, gpu->bus_freq);
 
-	ret = adreno_gpu_init(dev, pdev, &a3xx_gpu->base,
-			&funcs, config->rev);
+	gpu->perfcntrs = perfcntrs;
+	gpu->num_perfcntrs = ARRAY_SIZE(perfcntrs);
+
+	ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs, config->rev);
 	if (ret)
 		goto fail;
 
-	return &a3xx_gpu->base.base;
+	/* if needed, allocate gmem: */
+	if (adreno_is_a330(adreno_gpu)) {
+#ifdef CONFIG_MSM_OCMEM
+		/* TODO this is different/missing upstream: */
+		struct ocmem_buf *ocmem_hdl =
+				ocmem_allocate(OCMEM_GRAPHICS, adreno_gpu->gmem);
+
+		a3xx_gpu->ocmem_hdl = ocmem_hdl;
+		a3xx_gpu->ocmem_base = ocmem_hdl->addr;
+		adreno_gpu->gmem = ocmem_hdl->len;
+		DBG("using %dK of OCMEM at 0x%08x", adreno_gpu->gmem / 1024,
+				a3xx_gpu->ocmem_base);
+#endif
+	}
+
+	if (!gpu->mmu) {
+		/* TODO we think it is possible to configure the GPU to
+		 * restrict access to VRAM carveout.  But the required
+		 * registers are unknown.  For now just bail out and
+		 * limp along with just modesetting.  If it turns out
+		 * to not be possible to restrict access, then we must
+		 * implement a cmdstream validator.
+		 */
+		dev_err(dev->dev, "No memory protection without IOMMU\n");
+		ret = -ENXIO;
+		goto fail;
+	}
+
+	return gpu;
 
 fail:
 	if (a3xx_gpu)
@@ -436,19 +559,66 @@ fail:
  * The a3xx device:
  */
 
-static int a3xx_probe(struct platform_device *pdev)
+#if defined(CONFIG_MSM_BUS_SCALING) && !defined(CONFIG_OF)
+#  include <mach/kgsl.h>
+#endif
+
+static void set_gpu_pdev(struct drm_device *dev,
+		struct platform_device *pdev)
+{
+	struct msm_drm_private *priv = dev->dev_private;
+	priv->gpu_pdev = pdev;
+}
+
+static int a3xx_bind(struct device *dev, struct device *master, void *data)
 {
 	static struct adreno_platform_config config = {};
 #ifdef CONFIG_OF
-	/* TODO */
+	struct device_node *child, *node = dev->of_node;
+	u32 val;
+	int ret;
+
+	ret = of_property_read_u32(node, "qcom,chipid", &val);
+	if (ret) {
+		dev_err(dev, "could not find chipid: %d\n", ret);
+		return ret;
+	}
+
+	config.rev = ADRENO_REV((val >> 24) & 0xff,
+			(val >> 16) & 0xff, (val >> 8) & 0xff, val & 0xff);
+
+	/* find clock rates: */
+	config.fast_rate = 0;
+	config.slow_rate = ~0;
+	for_each_child_of_node(node, child) {
+		if (of_device_is_compatible(child, "qcom,gpu-pwrlevels")) {
+			struct device_node *pwrlvl;
+			for_each_child_of_node(child, pwrlvl) {
+				ret = of_property_read_u32(pwrlvl, "qcom,gpu-freq", &val);
+				if (ret) {
+					dev_err(dev, "could not find gpu-freq: %d\n", ret);
+					return ret;
+				}
+				config.fast_rate = max(config.fast_rate, val);
+				config.slow_rate = min(config.slow_rate, val);
+			}
+		}
+	}
+
+	if (!config.fast_rate) {
+		dev_err(dev, "could not find clk rates\n");
+		return -ENXIO;
+	}
+
 #else
+	struct kgsl_device_platform_data *pdata = dev->platform_data;
 	uint32_t version = socinfo_get_version();
 	if (cpu_is_apq8064ab()) {
 		config.fast_rate = 450000000;
 		config.slow_rate = 27000000;
 		config.bus_freq  = 4;
 		config.rev = ADRENO_REV(3, 2, 1, 0);
-	} else if (cpu_is_apq8064() || cpu_is_msm8960ab()) {
+	} else if (cpu_is_apq8064()) {
 		config.fast_rate = 400000000;
 		config.slow_rate = 27000000;
 		config.bus_freq  = 4;
@@ -461,6 +631,16 @@ static int a3xx_probe(struct platform_device *pdev)
 		else
 			config.rev = ADRENO_REV(3, 2, 0, 0);
 
+	} else if (cpu_is_msm8960ab()) {
+		config.fast_rate = 400000000;
+		config.slow_rate = 320000000;
+		config.bus_freq  = 4;
+
+		if (SOCINFO_VERSION_MINOR(version) == 0)
+			config.rev = ADRENO_REV(3, 2, 1, 0);
+		else
+			config.rev = ADRENO_REV(3, 2, 1, 1);
+
 	} else if (cpu_is_msm8930()) {
 		config.fast_rate = 400000000;
 		config.slow_rate = 27000000;
@@ -473,22 +653,49 @@ static int a3xx_probe(struct platform_device *pdev)
 			config.rev = ADRENO_REV(3, 0, 5, 0);
 
 	}
+#  ifdef CONFIG_MSM_BUS_SCALING
+	config.bus_scale_table = pdata->bus_scale_table;
+#  endif
 #endif
-	pdev->dev.platform_data = &config;
-	a3xx_pdev = pdev;
+	dev->platform_data = &config;
+	set_gpu_pdev(dev_get_drvdata(master), to_platform_device(dev));
 	return 0;
 }
 
+static void a3xx_unbind(struct device *dev, struct device *master,
+		void *data)
+{
+	set_gpu_pdev(dev_get_drvdata(master), NULL);
+}
+
+static const struct component_ops a3xx_ops = {
+		.bind   = a3xx_bind,
+		.unbind = a3xx_unbind,
+};
+
+static int a3xx_probe(struct platform_device *pdev)
+{
+	return component_add(&pdev->dev, &a3xx_ops);
+}
+
 static int a3xx_remove(struct platform_device *pdev)
 {
-	a3xx_pdev = NULL;
+	component_del(&pdev->dev, &a3xx_ops);
 	return 0;
 }
 
+static const struct of_device_id dt_match[] = {
+	{ .compatible = "qcom,kgsl-3d0" },
+	{}
+};
+
 static struct platform_driver a3xx_driver = {
 	.probe = a3xx_probe,
 	.remove = a3xx_remove,
-	.driver.name = "kgsl-3d0",
+	.driver = {
+		.name = "kgsl-3d0",
+		.of_match_table = dt_match,
+	},
 };
 
 void __init a3xx_register(void)
diff --git a/drivers/gpu/drm/msm/adreno/a3xx_gpu.h b/drivers/gpu/drm/msm/adreno/a3xx_gpu.h
index 32c398c2d00..bb9a8ca0507 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/a3xx_gpu.h
@@ -24,6 +24,10 @@
 struct a3xx_gpu {
 	struct adreno_gpu base;
 	struct platform_device *pdev;
+
+	/* if OCMEM is used for GMEM: */
+	uint32_t ocmem_base;
+	void *ocmem_hdl;
 };
 #define to_a3xx_gpu(x) container_of(x, struct a3xx_gpu, base)
 
diff --git a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
index 33dcc606c7c..d6e6ce2d1ab 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
@@ -8,12 +8,13 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml              (    327 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/a2xx/a2xx.xml           (  31003 bytes, from 2013-09-19 18:50:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_common.xml       (   8983 bytes, from 2013-07-24 01:38:36)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_pm4.xml          (   9759 bytes, from 2013-09-10 00:52:33)
-- /home/robclark/src/freedreno/envytools/rnndb/a3xx/a3xx.xml           (  51983 bytes, from 2013-09-10 00:52:32)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
+- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32814 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   8900 bytes, from 2013-10-22 23:57:49)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  10574 bytes, from 2013-11-13 05:44:45)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  53644 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (   8344 bytes, from 2013-11-30 14:49:47)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -115,96 +116,6 @@ enum adreno_rb_depth_format {
 	DEPTHX_24_8 = 1,
 };
 
-enum adreno_mmu_clnt_beh {
-	BEH_NEVR = 0,
-	BEH_TRAN_RNG = 1,
-	BEH_TRAN_FLT = 2,
-};
-
-#define REG_AXXX_MH_MMU_CONFIG					0x00000040
-#define AXXX_MH_MMU_CONFIG_MMU_ENABLE				0x00000001
-#define AXXX_MH_MMU_CONFIG_SPLIT_MODE_ENABLE			0x00000002
-#define AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__MASK		0x00000030
-#define AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__SHIFT		4
-static inline uint32_t AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__MASK		0x000000c0
-#define AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__SHIFT		6
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__MASK		0x00000300
-#define AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__SHIFT		8
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__MASK		0x00000c00
-#define AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__SHIFT		10
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__MASK		0x00003000
-#define AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__SHIFT		12
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__MASK		0x0000c000
-#define AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__SHIFT		14
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__MASK		0x00030000
-#define AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__SHIFT		16
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__MASK		0x000c0000
-#define AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__SHIFT		18
-static inline uint32_t AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__MASK		0x00300000
-#define AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__SHIFT		20
-static inline uint32_t AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__MASK		0x00c00000
-#define AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__SHIFT		22
-static inline uint32_t AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__MASK		0x03000000
-#define AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__SHIFT		24
-static inline uint32_t AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__MASK;
-}
-
-#define REG_AXXX_MH_MMU_VA_RANGE				0x00000041
-
-#define REG_AXXX_MH_MMU_PT_BASE					0x00000042
-
-#define REG_AXXX_MH_MMU_PAGE_FAULT				0x00000043
-
-#define REG_AXXX_MH_MMU_TRAN_ERROR				0x00000044
-
-#define REG_AXXX_MH_MMU_INVALIDATE				0x00000045
-
-#define REG_AXXX_MH_MMU_MPU_BASE				0x00000046
-
-#define REG_AXXX_MH_MMU_MPU_END					0x00000047
-
 #define REG_AXXX_CP_RB_BASE					0x000001c0
 
 #define REG_AXXX_CP_RB_CNTL					0x000001c1
@@ -275,6 +186,18 @@ static inline uint32_t AXXX_CP_QUEUE_THRESHOLDS_CSQ_ST_START(uint32_t val)
 }
 
 #define REG_AXXX_CP_MEQ_THRESHOLDS				0x000001d6
+#define AXXX_CP_MEQ_THRESHOLDS_MEQ_END__MASK			0x001f0000
+#define AXXX_CP_MEQ_THRESHOLDS_MEQ_END__SHIFT			16
+static inline uint32_t AXXX_CP_MEQ_THRESHOLDS_MEQ_END(uint32_t val)
+{
+	return ((val) << AXXX_CP_MEQ_THRESHOLDS_MEQ_END__SHIFT) & AXXX_CP_MEQ_THRESHOLDS_MEQ_END__MASK;
+}
+#define AXXX_CP_MEQ_THRESHOLDS_ROQ_END__MASK			0x1f000000
+#define AXXX_CP_MEQ_THRESHOLDS_ROQ_END__SHIFT			24
+static inline uint32_t AXXX_CP_MEQ_THRESHOLDS_ROQ_END(uint32_t val)
+{
+	return ((val) << AXXX_CP_MEQ_THRESHOLDS_ROQ_END__SHIFT) & AXXX_CP_MEQ_THRESHOLDS_ROQ_END__MASK;
+}
 
 #define REG_AXXX_CP_CSQ_AVAIL					0x000001d7
 #define AXXX_CP_CSQ_AVAIL_RING__MASK				0x0000007f
@@ -402,6 +325,36 @@ static inline uint32_t AXXX_CP_CSQ_IB2_STAT_WPTR(uint32_t val)
 	return ((val) << AXXX_CP_CSQ_IB2_STAT_WPTR__SHIFT) & AXXX_CP_CSQ_IB2_STAT_WPTR__MASK;
 }
 
+#define REG_AXXX_CP_NON_PREFETCH_CNTRS				0x00000440
+
+#define REG_AXXX_CP_STQ_ST_STAT					0x00000443
+
+#define REG_AXXX_CP_ST_BASE					0x0000044d
+
+#define REG_AXXX_CP_ST_BUFSZ					0x0000044e
+
+#define REG_AXXX_CP_MEQ_STAT					0x0000044f
+
+#define REG_AXXX_CP_MIU_TAG_STAT				0x00000452
+
+#define REG_AXXX_CP_BIN_MASK_LO					0x00000454
+
+#define REG_AXXX_CP_BIN_MASK_HI					0x00000455
+
+#define REG_AXXX_CP_BIN_SELECT_LO				0x00000456
+
+#define REG_AXXX_CP_BIN_SELECT_HI				0x00000457
+
+#define REG_AXXX_CP_IB1_BASE					0x00000458
+
+#define REG_AXXX_CP_IB1_BUFSZ					0x00000459
+
+#define REG_AXXX_CP_IB2_BASE					0x0000045a
+
+#define REG_AXXX_CP_IB2_BUFSZ					0x0000045b
+
+#define REG_AXXX_CP_STAT					0x0000047f
+
 #define REG_AXXX_CP_SCRATCH_REG0				0x00000578
 
 #define REG_AXXX_CP_SCRATCH_REG1				0x00000579
@@ -418,6 +371,26 @@ static inline uint32_t AXXX_CP_CSQ_IB2_STAT_WPTR(uint32_t val)
 
 #define REG_AXXX_CP_SCRATCH_REG7				0x0000057f
 
+#define REG_AXXX_CP_ME_VS_EVENT_SRC				0x00000600
+
+#define REG_AXXX_CP_ME_VS_EVENT_ADDR				0x00000601
+
+#define REG_AXXX_CP_ME_VS_EVENT_DATA				0x00000602
+
+#define REG_AXXX_CP_ME_VS_EVENT_ADDR_SWM			0x00000603
+
+#define REG_AXXX_CP_ME_VS_EVENT_DATA_SWM			0x00000604
+
+#define REG_AXXX_CP_ME_PS_EVENT_SRC				0x00000605
+
+#define REG_AXXX_CP_ME_PS_EVENT_ADDR				0x00000606
+
+#define REG_AXXX_CP_ME_PS_EVENT_DATA				0x00000607
+
+#define REG_AXXX_CP_ME_PS_EVENT_ADDR_SWM			0x00000608
+
+#define REG_AXXX_CP_ME_PS_EVENT_DATA_SWM			0x00000609
+
 #define REG_AXXX_CP_ME_CF_EVENT_SRC				0x0000060a
 
 #define REG_AXXX_CP_ME_CF_EVENT_ADDR				0x0000060b
@@ -428,5 +401,11 @@ static inline uint32_t AXXX_CP_CSQ_IB2_STAT_WPTR(uint32_t val)
 
 #define REG_AXXX_CP_ME_NRT_DATA					0x0000060e
 
+#define REG_AXXX_CP_ME_VS_FETCH_DONE_SRC			0x00000612
+
+#define REG_AXXX_CP_ME_VS_FETCH_DONE_ADDR			0x00000613
+
+#define REG_AXXX_CP_ME_VS_FETCH_DONE_DATA			0x00000614
+
 
 #endif /* ADRENO_COMMON_XML */
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
index a0b9d8a95b1..28ca8cd8b09 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
@@ -17,6 +17,7 @@
 
 #include "adreno_gpu.h"
 #include "msm_gem.h"
+#include "msm_mmu.h"
 
 struct adreno_info {
 	struct adreno_rev rev;
@@ -44,7 +45,7 @@ static const struct adreno_info gpulist[] = {
 		.pfpfw = "a300_pfp.fw",
 		.gmem  = SZ_512K,
 	}, {
-		.rev   = ADRENO_REV(3, 3, 0, 0),
+		.rev   = ADRENO_REV(3, 3, 0, ANY_ID),
 		.revn  = 330,
 		.name  = "A330",
 		.pm4fw = "a330_pm4.fw",
@@ -53,6 +54,11 @@ static const struct adreno_info gpulist[] = {
 	},
 };
 
+MODULE_FIRMWARE("a300_pm4.fw");
+MODULE_FIRMWARE("a300_pfp.fw");
+MODULE_FIRMWARE("a330_pm4.fw");
+MODULE_FIRMWARE("a330_pfp.fw");
+
 #define RB_SIZE    SZ_32K
 #define RB_BLKSIZE 16
 
@@ -65,7 +71,13 @@ int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value)
 		*value = adreno_gpu->info->revn;
 		return 0;
 	case MSM_PARAM_GMEM_SIZE:
-		*value = adreno_gpu->info->gmem;
+		*value = adreno_gpu->gmem;
+		return 0;
+	case MSM_PARAM_CHIP_ID:
+		*value = adreno_gpu->rev.patchid |
+				(adreno_gpu->rev.minor << 8) |
+				(adreno_gpu->rev.major << 16) |
+				(adreno_gpu->rev.core << 24);
 		return 0;
 	default:
 		DBG("%s: invalid param: %u", gpu->name, param);
@@ -86,7 +98,7 @@ int adreno_hw_init(struct msm_gpu *gpu)
 	gpu_write(gpu, REG_AXXX_CP_RB_CNTL,
 			/* size is log2(quad-words): */
 			AXXX_CP_RB_CNTL_BUFSZ(ilog2(gpu->rb->size / 8)) |
-			AXXX_CP_RB_CNTL_BLKSZ(RB_BLKSIZE));
+			AXXX_CP_RB_CNTL_BLKSZ(ilog2(RB_BLKSIZE / 8)));
 
 	/* Setup ringbuffer address: */
 	gpu_write(gpu, REG_AXXX_CP_RB_BASE, gpu->rb_iova);
@@ -219,19 +231,11 @@ void adreno_flush(struct msm_gpu *gpu)
 void adreno_idle(struct msm_gpu *gpu)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
-	uint32_t rptr, wptr = get_wptr(gpu->rb);
-	unsigned long t;
-
-	t = jiffies + ADRENO_IDLE_TIMEOUT;
-
-	/* then wait for CP to drain ringbuffer: */
-	do {
-		rptr = adreno_gpu->memptrs->rptr;
-		if (rptr == wptr)
-			return;
-	} while(time_before(jiffies, t));
+	uint32_t wptr = get_wptr(gpu->rb);
 
-	DRM_ERROR("%s: timeout waiting to drain ringbuffer!\n", gpu->name);
+	/* wait for CP to drain ringbuffer: */
+	if (spin_until(adreno_gpu->memptrs->rptr == wptr))
+		DRM_ERROR("%s: timeout waiting to drain ringbuffer!\n", gpu->name);
 
 	/* TODO maybe we need to reset GPU here to recover from hang? */
 }
@@ -254,22 +258,37 @@ void adreno_show(struct msm_gpu *gpu, struct seq_file *m)
 }
 #endif
 
-void adreno_wait_ring(struct msm_gpu *gpu, uint32_t ndwords)
+/* would be nice to not have to duplicate the _show() stuff with printk(): */
+void adreno_dump(struct msm_gpu *gpu)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
-	uint32_t freedwords;
-	unsigned long t = jiffies + ADRENO_IDLE_TIMEOUT;
-	do {
-		uint32_t size = gpu->rb->size / 4;
-		uint32_t wptr = get_wptr(gpu->rb);
-		uint32_t rptr = adreno_gpu->memptrs->rptr;
-		freedwords = (rptr + (size - 1) - wptr) % size;
-
-		if (time_after(jiffies, t)) {
-			DRM_ERROR("%s: timeout waiting for ringbuffer space\n", gpu->name);
-			break;
-		}
-	} while(freedwords < ndwords);
+
+	printk("revision: %d (%d.%d.%d.%d)\n",
+			adreno_gpu->info->revn, adreno_gpu->rev.core,
+			adreno_gpu->rev.major, adreno_gpu->rev.minor,
+			adreno_gpu->rev.patchid);
+
+	printk("fence:    %d/%d\n", adreno_gpu->memptrs->fence,
+			gpu->submitted_fence);
+	printk("rptr:     %d\n", adreno_gpu->memptrs->rptr);
+	printk("wptr:     %d\n", adreno_gpu->memptrs->wptr);
+	printk("rb wptr:  %d\n", get_wptr(gpu->rb));
+
+}
+
+static uint32_t ring_freewords(struct msm_gpu *gpu)
+{
+	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	uint32_t size = gpu->rb->size / 4;
+	uint32_t wptr = get_wptr(gpu->rb);
+	uint32_t rptr = adreno_gpu->memptrs->rptr;
+	return (rptr + (size - 1) - wptr) % size;
+}
+
+void adreno_wait_ring(struct msm_gpu *gpu, uint32_t ndwords)
+{
+	if (spin_until(ring_freewords(gpu) >= ndwords))
+		DRM_ERROR("%s: timeout waiting for ringbuffer space\n", gpu->name);
 }
 
 static const char *iommu_ports[] = {
@@ -286,6 +305,7 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct adreno_gpu *gpu, const struct adreno_gpu_funcs *funcs,
 		struct adreno_rev rev)
 {
+	struct msm_mmu *mmu;
 	int i, ret;
 
 	/* identify gpu: */
@@ -311,6 +331,7 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 			rev.core, rev.major, rev.minor, rev.patchid);
 
 	gpu->funcs = funcs;
+	gpu->gmem = gpu->info->gmem;
 	gpu->rev = rev;
 
 	ret = request_firmware(&gpu->pm4, gpu->info->pm4fw, drm->dev);
@@ -333,10 +354,13 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 	if (ret)
 		return ret;
 
-	ret = msm_iommu_attach(drm, gpu->base.iommu,
-			iommu_ports, ARRAY_SIZE(iommu_ports));
-	if (ret)
-		return ret;
+	mmu = gpu->base.mmu;
+	if (mmu) {
+		ret = mmu->funcs->attach(mmu, iommu_ports,
+				ARRAY_SIZE(iommu_ports));
+		if (ret)
+			return ret;
+	}
 
 	gpu->memptrs_bo = msm_gem_new(drm, sizeof(*gpu->memptrs),
 			MSM_BO_UNCACHED);
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
index f73abfba7c2..63c36ce3302 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
@@ -51,6 +51,7 @@ struct adreno_gpu {
 	struct msm_gpu base;
 	struct adreno_rev rev;
 	const struct adreno_info *info;
+	uint32_t gmem;  /* actual gmem size */
 	uint32_t revn;  /* numeric revision name */
 	const struct adreno_gpu_funcs *funcs;
 
@@ -70,9 +71,25 @@ struct adreno_gpu {
 struct adreno_platform_config {
 	struct adreno_rev rev;
 	uint32_t fast_rate, slow_rate, bus_freq;
+#ifdef CONFIG_MSM_BUS_SCALING
+	struct msm_bus_scale_pdata *bus_scale_table;
+#endif
 };
 
-#define ADRENO_IDLE_TIMEOUT (20 * 1000)
+#define ADRENO_IDLE_TIMEOUT msecs_to_jiffies(1000)
+
+#define spin_until(X) ({                                   \
+	int __ret = -ETIMEDOUT;                            \
+	unsigned long __t = jiffies + ADRENO_IDLE_TIMEOUT; \
+	do {                                               \
+		if (X) {                                   \
+			__ret = 0;                         \
+			break;                             \
+		}                                          \
+	} while (time_before(jiffies, __t));               \
+	__ret;                                             \
+})
+
 
 static inline bool adreno_is_a3xx(struct adreno_gpu *gpu)
 {
@@ -94,6 +111,11 @@ static inline bool adreno_is_a330(struct adreno_gpu *gpu)
 	return gpu->revn == 330;
 }
 
+static inline bool adreno_is_a330v2(struct adreno_gpu *gpu)
+{
+	return adreno_is_a330(gpu) && (gpu->rev.patchid > 0);
+}
+
 int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value);
 int adreno_hw_init(struct msm_gpu *gpu);
 uint32_t adreno_last_fence(struct msm_gpu *gpu);
@@ -105,6 +127,7 @@ void adreno_idle(struct msm_gpu *gpu);
 #ifdef CONFIG_DEBUG_FS
 void adreno_show(struct msm_gpu *gpu, struct seq_file *m);
 #endif
+void adreno_dump(struct msm_gpu *gpu);
 void adreno_wait_ring(struct msm_gpu *gpu, uint32_t ndwords);
 
 int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
diff --git a/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h b/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
index 259ad709b0c..ae992c71703 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
@@ -8,12 +8,13 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml              (    327 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/a2xx/a2xx.xml           (  31003 bytes, from 2013-09-19 18:50:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_common.xml       (   8983 bytes, from 2013-07-24 01:38:36)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_pm4.xml          (   9759 bytes, from 2013-09-10 00:52:33)
-- /home/robclark/src/freedreno/envytools/rnndb/a3xx/a3xx.xml           (  51983 bytes, from 2013-09-10 00:52:32)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
+- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml  (   1453 bytes, from 2013-03-31 16:51:27)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a2xx.xml          (  32814 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   8900 bytes, from 2013-10-22 23:57:49)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  10574 bytes, from 2013-11-13 05:44:45)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  53644 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (   8344 bytes, from 2013-11-30 14:49:47)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -66,13 +67,15 @@ enum vgt_event_type {
 
 enum pc_di_primtype {
 	DI_PT_NONE = 0,
-	DI_PT_POINTLIST = 1,
+	DI_PT_POINTLIST_A2XX = 1,
 	DI_PT_LINELIST = 2,
 	DI_PT_LINESTRIP = 3,
 	DI_PT_TRILIST = 4,
 	DI_PT_TRIFAN = 5,
 	DI_PT_TRISTRIP = 6,
+	DI_PT_LINELOOP = 7,
 	DI_PT_RECTLIST = 8,
+	DI_PT_POINTLIST_A3XX = 9,
 	DI_PT_QUADLIST = 13,
 	DI_PT_QUADSTRIP = 14,
 	DI_PT_POLYGON = 15,
@@ -119,7 +122,7 @@ enum adreno_pm4_type3_packets {
 	CP_WAIT_FOR_IDLE = 38,
 	CP_WAIT_REG_MEM = 60,
 	CP_WAIT_REG_EQ = 82,
-	CP_WAT_REG_GTE = 83,
+	CP_WAIT_REG_GTE = 83,
 	CP_WAIT_UNTIL_READ = 92,
 	CP_WAIT_IB_PFD_COMPLETE = 93,
 	CP_REG_RMW = 33,
@@ -151,7 +154,6 @@ enum adreno_pm4_type3_packets {
 	CP_CONTEXT_UPDATE = 94,
 	CP_INTERRUPT = 64,
 	CP_IM_STORE = 44,
-	CP_SET_BIN_BASE_OFFSET = 75,
 	CP_SET_DRAW_INIT_FLAGS = 75,
 	CP_SET_PROTECTED_MODE = 95,
 	CP_LOAD_STATE = 48,
@@ -159,6 +161,16 @@ enum adreno_pm4_type3_packets {
 	CP_COND_INDIRECT_BUFFER_PFD = 50,
 	CP_INDIRECT_BUFFER_PFE = 63,
 	CP_SET_BIN = 76,
+	CP_TEST_TWO_MEMS = 113,
+	CP_WAIT_FOR_ME = 19,
+	IN_IB_PREFETCH_END = 23,
+	IN_SUBBLK_PREFETCH = 31,
+	IN_INSTR_PREFETCH = 32,
+	IN_INSTR_MATCH = 71,
+	IN_CONST_PREFETCH = 73,
+	IN_INCR_UPDT_STATE = 85,
+	IN_INCR_UPDT_CONST = 86,
+	IN_INCR_UPDT_INSTR = 87,
 };
 
 enum adreno_state_block {
diff --git a/drivers/gpu/drm/msm/dsi/dsi.xml.h b/drivers/gpu/drm/msm/dsi/dsi.xml.h
index 6d4c62bf70d..87be647e382 100644
--- a/drivers/gpu/drm/msm/dsi/dsi.xml.h
+++ b/drivers/gpu/drm/msm/dsi/dsi.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    595 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2013-12-03 20:59:13)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h b/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
index d1df38bf574..747a6ef4211 100644
--- a/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
+++ b/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    595 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2013-12-03 20:59:13)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/dsi/sfpb.xml.h b/drivers/gpu/drm/msm/dsi/sfpb.xml.h
index 0030a111302..48e03acf19b 100644
--- a/drivers/gpu/drm/msm/dsi/sfpb.xml.h
+++ b/drivers/gpu/drm/msm/dsi/sfpb.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    595 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2013-12-03 20:59:13)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.c b/drivers/gpu/drm/msm/hdmi/hdmi.c
index 50d11df35b2..7f7aadef8a8 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.c
@@ -17,8 +17,6 @@
 
 #include "hdmi.h"
 
-static struct platform_device *hdmi_pdev;
-
 void hdmi_set_mode(struct hdmi *hdmi, bool power_on)
 {
 	uint32_t ctrl = 0;
@@ -41,7 +39,7 @@ void hdmi_set_mode(struct hdmi *hdmi, bool power_on)
 			power_on ? "Enable" : "Disable", ctrl);
 }
 
-static irqreturn_t hdmi_irq(int irq, void *dev_id)
+irqreturn_t hdmi_irq(int irq, void *dev_id)
 {
 	struct hdmi *hdmi = dev_id;
 
@@ -67,17 +65,17 @@ void hdmi_destroy(struct kref *kref)
 	if (hdmi->i2c)
 		hdmi_i2c_destroy(hdmi->i2c);
 
-	put_device(&hdmi->pdev->dev);
+	platform_set_drvdata(hdmi->pdev, NULL);
 }
 
 /* initialize connector */
-int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
+struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 {
 	struct hdmi *hdmi = NULL;
 	struct msm_drm_private *priv = dev->dev_private;
-	struct platform_device *pdev = hdmi_pdev;
+	struct platform_device *pdev = priv->hdmi_pdev;
 	struct hdmi_platform_config *config;
-	int ret;
+	int i, ret;
 
 	if (!pdev) {
 		dev_err(dev->dev, "no hdmi device\n");
@@ -95,12 +93,13 @@ int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 
 	kref_init(&hdmi->refcount);
 
-	get_device(&pdev->dev);
-
 	hdmi->dev = dev;
 	hdmi->pdev = pdev;
+	hdmi->config = config;
 	hdmi->encoder = encoder;
 
+	hdmi_audio_infoframe_init(&hdmi->audio.infoframe);
+
 	/* not sure about which phy maps to which msm.. probably I miss some */
 	if (config->phy_init)
 		hdmi->phy = config->phy_init(hdmi);
@@ -114,44 +113,70 @@ int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 		goto fail;
 	}
 
-	hdmi->mmio = msm_ioremap(pdev, "hdmi_msm_hdmi_addr", "HDMI");
+	hdmi->mmio = msm_ioremap(pdev, config->mmio_name, "HDMI");
 	if (IS_ERR(hdmi->mmio)) {
 		ret = PTR_ERR(hdmi->mmio);
 		goto fail;
 	}
 
-	hdmi->mvs = devm_regulator_get(&pdev->dev, "8901_hdmi_mvs");
-	if (IS_ERR(hdmi->mvs))
-		hdmi->mvs = devm_regulator_get(&pdev->dev, "hdmi_mvs");
-	if (IS_ERR(hdmi->mvs)) {
-		ret = PTR_ERR(hdmi->mvs);
-		dev_err(dev->dev, "failed to get mvs regulator: %d\n", ret);
-		goto fail;
+	BUG_ON(config->hpd_reg_cnt > ARRAY_SIZE(hdmi->hpd_regs));
+	for (i = 0; i < config->hpd_reg_cnt; i++) {
+		struct regulator *reg;
+
+		reg = devm_regulator_get(&pdev->dev, config->hpd_reg_names[i]);
+		if (IS_ERR(reg)) {
+			ret = PTR_ERR(reg);
+			dev_err(dev->dev, "failed to get hpd regulator: %s (%d)\n",
+					config->hpd_reg_names[i], ret);
+			goto fail;
+		}
+
+		hdmi->hpd_regs[i] = reg;
 	}
 
-	hdmi->mpp0 = devm_regulator_get(&pdev->dev, "8901_mpp0");
-	if (IS_ERR(hdmi->mpp0))
-		hdmi->mpp0 = NULL;
+	BUG_ON(config->pwr_reg_cnt > ARRAY_SIZE(hdmi->pwr_regs));
+	for (i = 0; i < config->pwr_reg_cnt; i++) {
+		struct regulator *reg;
 
-	hdmi->clk = devm_clk_get(&pdev->dev, "core_clk");
-	if (IS_ERR(hdmi->clk)) {
-		ret = PTR_ERR(hdmi->clk);
-		dev_err(dev->dev, "failed to get 'clk': %d\n", ret);
-		goto fail;
+		reg = devm_regulator_get(&pdev->dev, config->pwr_reg_names[i]);
+		if (IS_ERR(reg)) {
+			ret = PTR_ERR(reg);
+			dev_err(dev->dev, "failed to get pwr regulator: %s (%d)\n",
+					config->pwr_reg_names[i], ret);
+			goto fail;
+		}
+
+		hdmi->pwr_regs[i] = reg;
 	}
 
-	hdmi->m_pclk = devm_clk_get(&pdev->dev, "master_iface_clk");
-	if (IS_ERR(hdmi->m_pclk)) {
-		ret = PTR_ERR(hdmi->m_pclk);
-		dev_err(dev->dev, "failed to get 'm_pclk': %d\n", ret);
-		goto fail;
+	BUG_ON(config->hpd_clk_cnt > ARRAY_SIZE(hdmi->hpd_clks));
+	for (i = 0; i < config->hpd_clk_cnt; i++) {
+		struct clk *clk;
+
+		clk = devm_clk_get(&pdev->dev, config->hpd_clk_names[i]);
+		if (IS_ERR(clk)) {
+			ret = PTR_ERR(clk);
+			dev_err(dev->dev, "failed to get hpd clk: %s (%d)\n",
+					config->hpd_clk_names[i], ret);
+			goto fail;
+		}
+
+		hdmi->hpd_clks[i] = clk;
 	}
 
-	hdmi->s_pclk = devm_clk_get(&pdev->dev, "slave_iface_clk");
-	if (IS_ERR(hdmi->s_pclk)) {
-		ret = PTR_ERR(hdmi->s_pclk);
-		dev_err(dev->dev, "failed to get 's_pclk': %d\n", ret);
-		goto fail;
+	BUG_ON(config->pwr_clk_cnt > ARRAY_SIZE(hdmi->pwr_clks));
+	for (i = 0; i < config->pwr_clk_cnt; i++) {
+		struct clk *clk;
+
+		clk = devm_clk_get(&pdev->dev, config->pwr_clk_names[i]);
+		if (IS_ERR(clk)) {
+			ret = PTR_ERR(clk);
+			dev_err(dev->dev, "failed to get pwr clk: %s (%d)\n",
+					config->pwr_clk_names[i], ret);
+			goto fail;
+		}
+
+		hdmi->pwr_clks[i] = clk;
 	}
 
 	hdmi->i2c = hdmi_i2c_init(hdmi);
@@ -178,20 +203,22 @@ int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 		goto fail;
 	}
 
-	hdmi->irq = platform_get_irq(pdev, 0);
-	if (hdmi->irq < 0) {
-		ret = hdmi->irq;
-		dev_err(dev->dev, "failed to get irq: %d\n", ret);
-		goto fail;
-	}
+	if (!config->shared_irq) {
+		hdmi->irq = platform_get_irq(pdev, 0);
+		if (hdmi->irq < 0) {
+			ret = hdmi->irq;
+			dev_err(dev->dev, "failed to get irq: %d\n", ret);
+			goto fail;
+		}
 
-	ret = devm_request_threaded_irq(&pdev->dev, hdmi->irq,
-			NULL, hdmi_irq, IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
-			"hdmi_isr", hdmi);
-	if (ret < 0) {
-		dev_err(dev->dev, "failed to request IRQ%u: %d\n",
-				hdmi->irq, ret);
-		goto fail;
+		ret = devm_request_threaded_irq(&pdev->dev, hdmi->irq,
+				NULL, hdmi_irq, IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"hdmi_isr", hdmi);
+		if (ret < 0) {
+			dev_err(dev->dev, "failed to request IRQ%u: %d\n",
+					hdmi->irq, ret);
+			goto fail;
+		}
 	}
 
 	encoder->bridge = hdmi->bridge;
@@ -199,7 +226,9 @@ int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 	priv->bridges[priv->num_bridges++]       = hdmi->bridge;
 	priv->connectors[priv->num_connectors++] = hdmi->connector;
 
-	return 0;
+	platform_set_drvdata(pdev, hdmi);
+
+	return hdmi;
 
 fail:
 	if (hdmi) {
@@ -211,54 +240,150 @@ fail:
 		hdmi_destroy(&hdmi->refcount);
 	}
 
-	return ret;
+	return ERR_PTR(ret);
 }
 
 /*
  * The hdmi device:
  */
 
-static int hdmi_dev_probe(struct platform_device *pdev)
+#include <linux/of_gpio.h>
+
+static void set_hdmi_pdev(struct drm_device *dev,
+		struct platform_device *pdev)
+{
+	struct msm_drm_private *priv = dev->dev_private;
+	priv->hdmi_pdev = pdev;
+}
+
+static int hdmi_bind(struct device *dev, struct device *master, void *data)
 {
 	static struct hdmi_platform_config config = {};
 #ifdef CONFIG_OF
-	/* TODO */
+	struct device_node *of_node = dev->of_node;
+
+	int get_gpio(const char *name)
+	{
+		int gpio = of_get_named_gpio(of_node, name, 0);
+		if (gpio < 0) {
+			dev_err(dev, "failed to get gpio: %s (%d)\n",
+					name, gpio);
+			gpio = -1;
+		}
+		return gpio;
+	}
+
+	/* TODO actually use DT.. */
+	static const char *hpd_reg_names[] = {"hpd-gdsc", "hpd-5v"};
+	static const char *pwr_reg_names[] = {"core-vdda", "core-vcc"};
+	static const char *hpd_clk_names[] = {"iface_clk", "core_clk", "mdp_core_clk"};
+	static unsigned long hpd_clk_freq[] = {0, 19200000, 0};
+	static const char *pwr_clk_names[] = {"extp_clk", "alt_iface_clk"};
+
+	config.phy_init      = hdmi_phy_8x74_init;
+	config.mmio_name     = "core_physical";
+	config.hpd_reg_names = hpd_reg_names;
+	config.hpd_reg_cnt   = ARRAY_SIZE(hpd_reg_names);
+	config.pwr_reg_names = pwr_reg_names;
+	config.pwr_reg_cnt   = ARRAY_SIZE(pwr_reg_names);
+	config.hpd_clk_names = hpd_clk_names;
+	config.hpd_freq      = hpd_clk_freq;
+	config.hpd_clk_cnt   = ARRAY_SIZE(hpd_clk_names);
+	config.pwr_clk_names = pwr_clk_names;
+	config.pwr_clk_cnt   = ARRAY_SIZE(pwr_clk_names);
+	config.ddc_clk_gpio  = get_gpio("qcom,hdmi-tx-ddc-clk");
+	config.ddc_data_gpio = get_gpio("qcom,hdmi-tx-ddc-data");
+	config.hpd_gpio      = get_gpio("qcom,hdmi-tx-hpd");
+	config.mux_en_gpio   = get_gpio("qcom,hdmi-tx-mux-en");
+	config.mux_sel_gpio  = get_gpio("qcom,hdmi-tx-mux-sel");
+	config.shared_irq    = true;
+
 #else
+	static const char *hpd_clk_names[] = {
+			"core_clk", "master_iface_clk", "slave_iface_clk",
+	};
 	if (cpu_is_apq8064()) {
+		static const char *hpd_reg_names[] = {"8921_hdmi_mvs"};
 		config.phy_init      = hdmi_phy_8960_init;
+		config.mmio_name     = "hdmi_msm_hdmi_addr";
+		config.hpd_reg_names = hpd_reg_names;
+		config.hpd_reg_cnt   = ARRAY_SIZE(hpd_reg_names);
+		config.hpd_clk_names = hpd_clk_names;
+		config.hpd_clk_cnt   = ARRAY_SIZE(hpd_clk_names);
 		config.ddc_clk_gpio  = 70;
 		config.ddc_data_gpio = 71;
 		config.hpd_gpio      = 72;
-		config.pmic_gpio     = 13 + NR_GPIO_IRQS;
-	} else if (cpu_is_msm8960()) {
+		config.mux_en_gpio   = -1;
+		config.mux_sel_gpio  = -1;
+	} else if (cpu_is_msm8960() || cpu_is_msm8960ab()) {
+		static const char *hpd_reg_names[] = {"8921_hdmi_mvs"};
 		config.phy_init      = hdmi_phy_8960_init;
+		config.mmio_name     = "hdmi_msm_hdmi_addr";
+		config.hpd_reg_names = hpd_reg_names;
+		config.hpd_reg_cnt   = ARRAY_SIZE(hpd_reg_names);
+		config.hpd_clk_names = hpd_clk_names;
+		config.hpd_clk_cnt   = ARRAY_SIZE(hpd_clk_names);
 		config.ddc_clk_gpio  = 100;
 		config.ddc_data_gpio = 101;
 		config.hpd_gpio      = 102;
-		config.pmic_gpio     = -1;
+		config.mux_en_gpio   = -1;
+		config.mux_sel_gpio  = -1;
 	} else if (cpu_is_msm8x60()) {
+		static const char *hpd_reg_names[] = {
+				"8901_hdmi_mvs", "8901_mpp0"
+		};
 		config.phy_init      = hdmi_phy_8x60_init;
+		config.mmio_name     = "hdmi_msm_hdmi_addr";
+		config.hpd_reg_names = hpd_reg_names;
+		config.hpd_reg_cnt   = ARRAY_SIZE(hpd_reg_names);
+		config.hpd_clk_names = hpd_clk_names;
+		config.hpd_clk_cnt   = ARRAY_SIZE(hpd_clk_names);
 		config.ddc_clk_gpio  = 170;
 		config.ddc_data_gpio = 171;
 		config.hpd_gpio      = 172;
-		config.pmic_gpio     = -1;
+		config.mux_en_gpio   = -1;
+		config.mux_sel_gpio  = -1;
 	}
 #endif
-	pdev->dev.platform_data = &config;
-	hdmi_pdev = pdev;
+	dev->platform_data = &config;
+	set_hdmi_pdev(dev_get_drvdata(master), to_platform_device(dev));
 	return 0;
 }
 
+static void hdmi_unbind(struct device *dev, struct device *master,
+		void *data)
+{
+	set_hdmi_pdev(dev_get_drvdata(master), NULL);
+}
+
+static const struct component_ops hdmi_ops = {
+		.bind   = hdmi_bind,
+		.unbind = hdmi_unbind,
+};
+
+static int hdmi_dev_probe(struct platform_device *pdev)
+{
+	return component_add(&pdev->dev, &hdmi_ops);
+}
+
 static int hdmi_dev_remove(struct platform_device *pdev)
 {
-	hdmi_pdev = NULL;
+	component_del(&pdev->dev, &hdmi_ops);
 	return 0;
 }
 
+static const struct of_device_id dt_match[] = {
+	{ .compatible = "qcom,hdmi-tx" },
+	{}
+};
+
 static struct platform_driver hdmi_driver = {
 	.probe = hdmi_dev_probe,
 	.remove = hdmi_dev_remove,
-	.driver.name = "hdmi_msm",
+	.driver = {
+		.name = "hdmi_msm",
+		.of_match_table = dt_match,
+	},
 };
 
 void __init hdmi_register(void)
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.h b/drivers/gpu/drm/msm/hdmi/hdmi.h
index 2c2ec566394..9d7723c6528 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.h
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.h
@@ -22,12 +22,20 @@
 #include <linux/clk.h>
 #include <linux/platform_device.h>
 #include <linux/regulator/consumer.h>
+#include <linux/hdmi.h>
 
 #include "msm_drv.h"
 #include "hdmi.xml.h"
 
 
 struct hdmi_phy;
+struct hdmi_platform_config;
+
+struct hdmi_audio {
+	bool enabled;
+	struct hdmi_audio_infoframe infoframe;
+	int rate;
+};
 
 struct hdmi {
 	struct kref refcount;
@@ -35,14 +43,21 @@ struct hdmi {
 	struct drm_device *dev;
 	struct platform_device *pdev;
 
-	void __iomem *mmio;
+	const struct hdmi_platform_config *config;
 
-	struct regulator *mvs;        /* HDMI_5V */
-	struct regulator *mpp0;       /* External 5V */
+	/* audio state: */
+	struct hdmi_audio audio;
 
-	struct clk *clk;
-	struct clk *m_pclk;
-	struct clk *s_pclk;
+	/* video state: */
+	bool power_on;
+	unsigned long int pixclock;
+
+	void __iomem *mmio;
+
+	struct regulator *hpd_regs[2];
+	struct regulator *pwr_regs[2];
+	struct clk *hpd_clks[3];
+	struct clk *pwr_clks[2];
 
 	struct hdmi_phy *phy;
 	struct i2c_adapter *i2c;
@@ -60,7 +75,30 @@ struct hdmi {
 /* platform config data (ie. from DT, or pdata) */
 struct hdmi_platform_config {
 	struct hdmi_phy *(*phy_init)(struct hdmi *hdmi);
-	int ddc_clk_gpio, ddc_data_gpio, hpd_gpio, pmic_gpio;
+	const char *mmio_name;
+
+	/* regulators that need to be on for hpd: */
+	const char **hpd_reg_names;
+	int hpd_reg_cnt;
+
+	/* regulators that need to be on for screen pwr: */
+	const char **pwr_reg_names;
+	int pwr_reg_cnt;
+
+	/* clks that need to be on for hpd: */
+	const char **hpd_clk_names;
+	const long unsigned *hpd_freq;
+	int hpd_clk_cnt;
+
+	/* clks that need to be on for screen pwr (ie pixel clk): */
+	const char **pwr_clk_names;
+	int pwr_clk_cnt;
+
+	/* gpio's: */
+	int ddc_clk_gpio, ddc_data_gpio, hpd_gpio, mux_en_gpio, mux_sel_gpio;
+
+	/* older devices had their own irq, mdp5+ it is shared w/ mdp: */
+	bool shared_irq;
 };
 
 void hdmi_set_mode(struct hdmi *hdmi, bool power_on);
@@ -106,6 +144,18 @@ struct hdmi_phy {
 
 struct hdmi_phy *hdmi_phy_8960_init(struct hdmi *hdmi);
 struct hdmi_phy *hdmi_phy_8x60_init(struct hdmi *hdmi);
+struct hdmi_phy *hdmi_phy_8x74_init(struct hdmi *hdmi);
+
+/*
+ * audio:
+ */
+
+int hdmi_audio_update(struct hdmi *hdmi);
+int hdmi_audio_info_setup(struct hdmi *hdmi, bool enabled,
+	uint32_t num_of_channels, uint32_t channel_allocation,
+	uint32_t level_shift, bool down_mix);
+void hdmi_audio_set_sample_rate(struct hdmi *hdmi, int rate);
+
 
 /*
  * hdmi bridge:
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.xml.h b/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
index 4e939f82918..e2636582cfd 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    595 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2013-12-03 20:59:13)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -212,6 +214,20 @@ static inline uint32_t HDMI_HDCP_LINK0_STATUS_KEY_STATE(enum hdmi_hdcp_key_state
 #define REG_HDMI_HDCP_RESET					0x00000130
 #define HDMI_HDCP_RESET_LINK0_DEAUTHENTICATE			0x00000001
 
+#define REG_HDMI_VENSPEC_INFO0					0x0000016c
+
+#define REG_HDMI_VENSPEC_INFO1					0x00000170
+
+#define REG_HDMI_VENSPEC_INFO2					0x00000174
+
+#define REG_HDMI_VENSPEC_INFO3					0x00000178
+
+#define REG_HDMI_VENSPEC_INFO4					0x0000017c
+
+#define REG_HDMI_VENSPEC_INFO5					0x00000180
+
+#define REG_HDMI_VENSPEC_INFO6					0x00000184
+
 #define REG_HDMI_AUDIO_CFG					0x000001d0
 #define HDMI_AUDIO_CFG_ENGINE_ENABLE				0x00000001
 #define HDMI_AUDIO_CFG_FIFO_WATERMARK__MASK			0x000000f0
@@ -235,6 +251,9 @@ static inline uint32_t HDMI_DDC_CTRL_TRANSACTION_CNT(uint32_t val)
 	return ((val) << HDMI_DDC_CTRL_TRANSACTION_CNT__SHIFT) & HDMI_DDC_CTRL_TRANSACTION_CNT__MASK;
 }
 
+#define REG_HDMI_DDC_ARBITRATION				0x00000210
+#define HDMI_DDC_ARBITRATION_HW_ARBITRATION			0x00000010
+
 #define REG_HDMI_DDC_INT_CTRL					0x00000214
 #define HDMI_DDC_INT_CTRL_SW_DONE_INT				0x00000001
 #define HDMI_DDC_INT_CTRL_SW_DONE_ACK				0x00000002
@@ -340,6 +359,20 @@ static inline uint32_t HDMI_DDC_REF_REFTIMER(uint32_t val)
 	return ((val) << HDMI_DDC_REF_REFTIMER__SHIFT) & HDMI_DDC_REF_REFTIMER__MASK;
 }
 
+#define REG_HDMI_CEC_STATUS					0x00000298
+
+#define REG_HDMI_CEC_INT					0x0000029c
+
+#define REG_HDMI_CEC_ADDR					0x000002a0
+
+#define REG_HDMI_CEC_TIME					0x000002a4
+
+#define REG_HDMI_CEC_REFTIMER					0x000002a8
+
+#define REG_HDMI_CEC_RD_DATA					0x000002ac
+
+#define REG_HDMI_CEC_RD_FILTER					0x000002b0
+
 #define REG_HDMI_ACTIVE_HSYNC					0x000002b4
 #define HDMI_ACTIVE_HSYNC_START__MASK				0x00000fff
 #define HDMI_ACTIVE_HSYNC_START__SHIFT				0
@@ -410,17 +443,33 @@ static inline uint32_t HDMI_VSYNC_TOTAL_F2_V_TOTAL(uint32_t val)
 #define HDMI_FRAME_CTRL_HSYNC_LOW				0x20000000
 #define HDMI_FRAME_CTRL_INTERLACED_EN				0x80000000
 
+#define REG_HDMI_AUD_INT					0x000002cc
+#define HDMI_AUD_INT_AUD_FIFO_URUN_INT				0x00000001
+#define HDMI_AUD_INT_AUD_FIFO_URAN_MASK				0x00000002
+#define HDMI_AUD_INT_AUD_SAM_DROP_INT				0x00000004
+#define HDMI_AUD_INT_AUD_SAM_DROP_MASK				0x00000008
+
 #define REG_HDMI_PHY_CTRL					0x000002d4
 #define HDMI_PHY_CTRL_SW_RESET_PLL				0x00000001
 #define HDMI_PHY_CTRL_SW_RESET_PLL_LOW				0x00000002
 #define HDMI_PHY_CTRL_SW_RESET					0x00000004
 #define HDMI_PHY_CTRL_SW_RESET_LOW				0x00000008
 
-#define REG_HDMI_AUD_INT					0x000002cc
-#define HDMI_AUD_INT_AUD_FIFO_URUN_INT				0x00000001
-#define HDMI_AUD_INT_AUD_FIFO_URAN_MASK				0x00000002
-#define HDMI_AUD_INT_AUD_SAM_DROP_INT				0x00000004
-#define HDMI_AUD_INT_AUD_SAM_DROP_MASK				0x00000008
+#define REG_HDMI_CEC_WR_RANGE					0x000002dc
+
+#define REG_HDMI_CEC_RD_RANGE					0x000002e0
+
+#define REG_HDMI_VERSION					0x000002e4
+
+#define REG_HDMI_CEC_COMPL_CTL					0x00000360
+
+#define REG_HDMI_CEC_RD_START_RANGE				0x00000364
+
+#define REG_HDMI_CEC_RD_TOTAL_RANGE				0x00000368
+
+#define REG_HDMI_CEC_RD_ERR_RESP_LO				0x0000036c
+
+#define REG_HDMI_CEC_WR_CHECK_CONFIG				0x00000370
 
 #define REG_HDMI_8x60_PHY_REG0					0x00000300
 #define HDMI_8x60_PHY_REG0_DESER_DEL_CTRL__MASK			0x0000001c
@@ -504,5 +553,23 @@ static inline uint32_t HDMI_8x60_PHY_REG1_OUTVOL_SWING_CTRL(uint32_t val)
 
 #define REG_HDMI_8960_PHY_REG12					0x00000430
 
+#define REG_HDMI_8x74_ANA_CFG0					0x00000000
+
+#define REG_HDMI_8x74_ANA_CFG1					0x00000004
+
+#define REG_HDMI_8x74_PD_CTRL0					0x00000010
+
+#define REG_HDMI_8x74_PD_CTRL1					0x00000014
+
+#define REG_HDMI_8x74_BIST_CFG0					0x00000034
+
+#define REG_HDMI_8x74_BIST_PATN0				0x0000003c
+
+#define REG_HDMI_8x74_BIST_PATN1				0x00000040
+
+#define REG_HDMI_8x74_BIST_PATN2				0x00000044
+
+#define REG_HDMI_8x74_BIST_PATN3				0x00000048
+
 
 #endif /* HDMI_XML */
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_audio.c b/drivers/gpu/drm/msm/hdmi/hdmi_audio.c
new file mode 100644
index 00000000000..872485f6013
--- /dev/null
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_audio.c
@@ -0,0 +1,273 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/hdmi.h>
+#include "hdmi.h"
+
+
+/* Supported HDMI Audio channels */
+#define MSM_HDMI_AUDIO_CHANNEL_2		0
+#define MSM_HDMI_AUDIO_CHANNEL_4		1
+#define MSM_HDMI_AUDIO_CHANNEL_6		2
+#define MSM_HDMI_AUDIO_CHANNEL_8		3
+
+/* maps MSM_HDMI_AUDIO_CHANNEL_n consts used by audio driver to # of channels: */
+static int nchannels[] = { 2, 4, 6, 8 };
+
+/* Supported HDMI Audio sample rates */
+#define MSM_HDMI_SAMPLE_RATE_32KHZ		0
+#define MSM_HDMI_SAMPLE_RATE_44_1KHZ		1
+#define MSM_HDMI_SAMPLE_RATE_48KHZ		2
+#define MSM_HDMI_SAMPLE_RATE_88_2KHZ		3
+#define MSM_HDMI_SAMPLE_RATE_96KHZ		4
+#define MSM_HDMI_SAMPLE_RATE_176_4KHZ		5
+#define MSM_HDMI_SAMPLE_RATE_192KHZ		6
+#define MSM_HDMI_SAMPLE_RATE_MAX		7
+
+
+struct hdmi_msm_audio_acr {
+	uint32_t n;	/* N parameter for clock regeneration */
+	uint32_t cts;	/* CTS parameter for clock regeneration */
+};
+
+struct hdmi_msm_audio_arcs {
+	unsigned long int pixclock;
+	struct hdmi_msm_audio_acr lut[MSM_HDMI_SAMPLE_RATE_MAX];
+};
+
+#define HDMI_MSM_AUDIO_ARCS(pclk, ...) { (1000 * (pclk)), __VA_ARGS__ }
+
+/* Audio constants lookup table for hdmi_msm_audio_acr_setup */
+/* Valid Pixel-Clock rates: 25.2MHz, 27MHz, 27.03MHz, 74.25MHz, 148.5MHz */
+static const struct hdmi_msm_audio_arcs acr_lut[] = {
+	/*  25.200MHz  */
+	HDMI_MSM_AUDIO_ARCS(25200, {
+		{4096, 25200}, {6272, 28000}, {6144, 25200}, {12544, 28000},
+		{12288, 25200}, {25088, 28000}, {24576, 25200} }),
+	/*  27.000MHz  */
+	HDMI_MSM_AUDIO_ARCS(27000, {
+		{4096, 27000}, {6272, 30000}, {6144, 27000}, {12544, 30000},
+		{12288, 27000}, {25088, 30000}, {24576, 27000} }),
+	/*  27.027MHz */
+	HDMI_MSM_AUDIO_ARCS(27030, {
+		{4096, 27027}, {6272, 30030}, {6144, 27027}, {12544, 30030},
+		{12288, 27027}, {25088, 30030}, {24576, 27027} }),
+	/*  74.250MHz */
+	HDMI_MSM_AUDIO_ARCS(74250, {
+		{4096, 74250}, {6272, 82500}, {6144, 74250}, {12544, 82500},
+		{12288, 74250}, {25088, 82500}, {24576, 74250} }),
+	/* 148.500MHz */
+	HDMI_MSM_AUDIO_ARCS(148500, {
+		{4096, 148500}, {6272, 165000}, {6144, 148500}, {12544, 165000},
+		{12288, 148500}, {25088, 165000}, {24576, 148500} }),
+};
+
+static const struct hdmi_msm_audio_arcs *get_arcs(unsigned long int pixclock)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(acr_lut); i++) {
+		const struct hdmi_msm_audio_arcs *arcs = &acr_lut[i];
+		if (arcs->pixclock == pixclock)
+			return arcs;
+	}
+
+	return NULL;
+}
+
+int hdmi_audio_update(struct hdmi *hdmi)
+{
+	struct hdmi_audio *audio = &hdmi->audio;
+	struct hdmi_audio_infoframe *info = &audio->infoframe;
+	const struct hdmi_msm_audio_arcs *arcs = NULL;
+	bool enabled = audio->enabled;
+	uint32_t acr_pkt_ctrl, vbi_pkt_ctrl, aud_pkt_ctrl;
+	uint32_t infofrm_ctrl, audio_config;
+
+	DBG("audio: enabled=%d, channels=%d, channel_allocation=0x%x, "
+		"level_shift_value=%d, downmix_inhibit=%d, rate=%d",
+		audio->enabled, info->channels,  info->channel_allocation,
+		info->level_shift_value, info->downmix_inhibit, audio->rate);
+	DBG("video: power_on=%d, pixclock=%lu", hdmi->power_on, hdmi->pixclock);
+
+	if (enabled && !(hdmi->power_on && hdmi->pixclock)) {
+		DBG("disabling audio: no video");
+		enabled = false;
+	}
+
+	if (enabled) {
+		arcs = get_arcs(hdmi->pixclock);
+		if (!arcs) {
+			DBG("disabling audio: unsupported pixclock: %lu",
+					hdmi->pixclock);
+			enabled = false;
+		}
+	}
+
+	/* Read first before writing */
+	acr_pkt_ctrl = hdmi_read(hdmi, REG_HDMI_ACR_PKT_CTRL);
+	vbi_pkt_ctrl = hdmi_read(hdmi, REG_HDMI_VBI_PKT_CTRL);
+	aud_pkt_ctrl = hdmi_read(hdmi, REG_HDMI_AUDIO_PKT_CTRL1);
+	infofrm_ctrl = hdmi_read(hdmi, REG_HDMI_INFOFRAME_CTRL0);
+	audio_config = hdmi_read(hdmi, REG_HDMI_AUDIO_CFG);
+
+	/* Clear N/CTS selection bits */
+	acr_pkt_ctrl &= ~HDMI_ACR_PKT_CTRL_SELECT__MASK;
+
+	if (enabled) {
+		uint32_t n, cts, multiplier;
+		enum hdmi_acr_cts select;
+		uint8_t buf[14];
+
+		n   = arcs->lut[audio->rate].n;
+		cts = arcs->lut[audio->rate].cts;
+
+		if ((MSM_HDMI_SAMPLE_RATE_192KHZ == audio->rate) ||
+				(MSM_HDMI_SAMPLE_RATE_176_4KHZ == audio->rate)) {
+			multiplier = 4;
+			n >>= 2; /* divide N by 4 and use multiplier */
+		} else if ((MSM_HDMI_SAMPLE_RATE_96KHZ == audio->rate) ||
+				(MSM_HDMI_SAMPLE_RATE_88_2KHZ == audio->rate)) {
+			multiplier = 2;
+			n >>= 1; /* divide N by 2 and use multiplier */
+		} else {
+			multiplier = 1;
+		}
+
+		DBG("n=%u, cts=%u, multiplier=%u", n, cts, multiplier);
+
+		acr_pkt_ctrl |= HDMI_ACR_PKT_CTRL_SOURCE;
+		acr_pkt_ctrl |= HDMI_ACR_PKT_CTRL_AUDIO_PRIORITY;
+		acr_pkt_ctrl |= HDMI_ACR_PKT_CTRL_N_MULTIPLIER(multiplier);
+
+		if ((MSM_HDMI_SAMPLE_RATE_48KHZ == audio->rate) ||
+				(MSM_HDMI_SAMPLE_RATE_96KHZ == audio->rate) ||
+				(MSM_HDMI_SAMPLE_RATE_192KHZ == audio->rate))
+			select = ACR_48;
+		else if ((MSM_HDMI_SAMPLE_RATE_44_1KHZ == audio->rate) ||
+				(MSM_HDMI_SAMPLE_RATE_88_2KHZ == audio->rate) ||
+				(MSM_HDMI_SAMPLE_RATE_176_4KHZ == audio->rate))
+			select = ACR_44;
+		else /* default to 32k */
+			select = ACR_32;
+
+		acr_pkt_ctrl |= HDMI_ACR_PKT_CTRL_SELECT(select);
+
+		hdmi_write(hdmi, REG_HDMI_ACR_0(select - 1),
+				HDMI_ACR_0_CTS(cts));
+		hdmi_write(hdmi, REG_HDMI_ACR_1(select - 1),
+				HDMI_ACR_1_N(n));
+
+		hdmi_write(hdmi, REG_HDMI_AUDIO_PKT_CTRL2,
+				COND(info->channels != 2, HDMI_AUDIO_PKT_CTRL2_LAYOUT) |
+				HDMI_AUDIO_PKT_CTRL2_OVERRIDE);
+
+		acr_pkt_ctrl |= HDMI_ACR_PKT_CTRL_CONT;
+		acr_pkt_ctrl |= HDMI_ACR_PKT_CTRL_SEND;
+
+		/* configure infoframe: */
+		hdmi_audio_infoframe_pack(info, buf, sizeof(buf));
+		hdmi_write(hdmi, REG_HDMI_AUDIO_INFO0,
+				(buf[3] <<  0) || (buf[4] <<  8) ||
+				(buf[5] << 16) || (buf[6] << 24));
+		hdmi_write(hdmi, REG_HDMI_AUDIO_INFO1,
+				(buf[7] <<  0) || (buf[8] << 8));
+
+		hdmi_write(hdmi, REG_HDMI_GC, 0);
+
+		vbi_pkt_ctrl |= HDMI_VBI_PKT_CTRL_GC_ENABLE;
+		vbi_pkt_ctrl |= HDMI_VBI_PKT_CTRL_GC_EVERY_FRAME;
+
+		aud_pkt_ctrl |= HDMI_AUDIO_PKT_CTRL1_AUDIO_SAMPLE_SEND;
+
+		infofrm_ctrl |= HDMI_INFOFRAME_CTRL0_AUDIO_INFO_SEND;
+		infofrm_ctrl |= HDMI_INFOFRAME_CTRL0_AUDIO_INFO_CONT;
+		infofrm_ctrl |= HDMI_INFOFRAME_CTRL0_AUDIO_INFO_SOURCE;
+		infofrm_ctrl |= HDMI_INFOFRAME_CTRL0_AUDIO_INFO_UPDATE;
+
+		audio_config &= ~HDMI_AUDIO_CFG_FIFO_WATERMARK__MASK;
+		audio_config |= HDMI_AUDIO_CFG_FIFO_WATERMARK(4);
+		audio_config |= HDMI_AUDIO_CFG_ENGINE_ENABLE;
+	} else {
+		hdmi_write(hdmi, REG_HDMI_GC, HDMI_GC_MUTE);
+		acr_pkt_ctrl &= ~HDMI_ACR_PKT_CTRL_CONT;
+		acr_pkt_ctrl &= ~HDMI_ACR_PKT_CTRL_SEND;
+		vbi_pkt_ctrl &= ~HDMI_VBI_PKT_CTRL_GC_ENABLE;
+		vbi_pkt_ctrl &= ~HDMI_VBI_PKT_CTRL_GC_EVERY_FRAME;
+		aud_pkt_ctrl &= ~HDMI_AUDIO_PKT_CTRL1_AUDIO_SAMPLE_SEND;
+		infofrm_ctrl &= ~HDMI_INFOFRAME_CTRL0_AUDIO_INFO_SEND;
+		infofrm_ctrl &= ~HDMI_INFOFRAME_CTRL0_AUDIO_INFO_CONT;
+		infofrm_ctrl &= ~HDMI_INFOFRAME_CTRL0_AUDIO_INFO_SOURCE;
+		infofrm_ctrl &= ~HDMI_INFOFRAME_CTRL0_AUDIO_INFO_UPDATE;
+		audio_config &= ~HDMI_AUDIO_CFG_ENGINE_ENABLE;
+	}
+
+	hdmi_write(hdmi, REG_HDMI_ACR_PKT_CTRL, acr_pkt_ctrl);
+	hdmi_write(hdmi, REG_HDMI_VBI_PKT_CTRL, vbi_pkt_ctrl);
+	hdmi_write(hdmi, REG_HDMI_AUDIO_PKT_CTRL1, aud_pkt_ctrl);
+	hdmi_write(hdmi, REG_HDMI_INFOFRAME_CTRL0, infofrm_ctrl);
+
+	hdmi_write(hdmi, REG_HDMI_AUD_INT,
+			COND(enabled, HDMI_AUD_INT_AUD_FIFO_URUN_INT) |
+			COND(enabled, HDMI_AUD_INT_AUD_SAM_DROP_INT));
+
+	hdmi_write(hdmi, REG_HDMI_AUDIO_CFG, audio_config);
+
+
+	DBG("audio %sabled", enabled ? "en" : "dis");
+
+	return 0;
+}
+
+int hdmi_audio_info_setup(struct hdmi *hdmi, bool enabled,
+	uint32_t num_of_channels, uint32_t channel_allocation,
+	uint32_t level_shift, bool down_mix)
+{
+	struct hdmi_audio *audio;
+
+	if (!hdmi)
+		return -ENXIO;
+
+	audio = &hdmi->audio;
+
+	if (num_of_channels >= ARRAY_SIZE(nchannels))
+		return -EINVAL;
+
+	audio->enabled = enabled;
+	audio->infoframe.channels = nchannels[num_of_channels];
+	audio->infoframe.channel_allocation = channel_allocation;
+	audio->infoframe.level_shift_value = level_shift;
+	audio->infoframe.downmix_inhibit = down_mix;
+
+	return hdmi_audio_update(hdmi);
+}
+
+void hdmi_audio_set_sample_rate(struct hdmi *hdmi, int rate)
+{
+	struct hdmi_audio *audio;
+
+	if (!hdmi)
+		return;
+
+	audio = &hdmi->audio;
+
+	if ((rate < 0) || (rate >= MSM_HDMI_SAMPLE_RATE_MAX))
+		return;
+
+	audio->rate = rate;
+	hdmi_audio_update(hdmi);
+}
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c b/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c
index 5a8ee3473cf..f6cf745c249 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c
@@ -19,10 +19,7 @@
 
 struct hdmi_bridge {
 	struct drm_bridge base;
-
 	struct hdmi *hdmi;
-
-	unsigned long int pixclock;
 };
 #define to_hdmi_bridge(x) container_of(x, struct hdmi_bridge, base)
 
@@ -34,6 +31,65 @@ static void hdmi_bridge_destroy(struct drm_bridge *bridge)
 	kfree(hdmi_bridge);
 }
 
+static void power_on(struct drm_bridge *bridge)
+{
+	struct drm_device *dev = bridge->dev;
+	struct hdmi_bridge *hdmi_bridge = to_hdmi_bridge(bridge);
+	struct hdmi *hdmi = hdmi_bridge->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
+	int i, ret;
+
+	for (i = 0; i < config->pwr_reg_cnt; i++) {
+		ret = regulator_enable(hdmi->pwr_regs[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to enable pwr regulator: %s (%d)\n",
+					config->pwr_reg_names[i], ret);
+		}
+	}
+
+	if (config->pwr_clk_cnt > 0) {
+		DBG("pixclock: %lu", hdmi->pixclock);
+		ret = clk_set_rate(hdmi->pwr_clks[0], hdmi->pixclock);
+		if (ret) {
+			dev_err(dev->dev, "failed to set pixel clk: %s (%d)\n",
+					config->pwr_clk_names[0], ret);
+		}
+	}
+
+	for (i = 0; i < config->pwr_clk_cnt; i++) {
+		ret = clk_prepare_enable(hdmi->pwr_clks[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to enable pwr clk: %s (%d)\n",
+					config->pwr_clk_names[i], ret);
+		}
+	}
+}
+
+static void power_off(struct drm_bridge *bridge)
+{
+	struct drm_device *dev = bridge->dev;
+	struct hdmi_bridge *hdmi_bridge = to_hdmi_bridge(bridge);
+	struct hdmi *hdmi = hdmi_bridge->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
+	int i, ret;
+
+	/* TODO do we need to wait for final vblank somewhere before
+	 * cutting the clocks?
+	 */
+	mdelay(16 + 4);
+
+	for (i = 0; i < config->pwr_clk_cnt; i++)
+		clk_disable_unprepare(hdmi->pwr_clks[i]);
+
+	for (i = 0; i < config->pwr_reg_cnt; i++) {
+		ret = regulator_disable(hdmi->pwr_regs[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to disable pwr regulator: %s (%d)\n",
+					config->pwr_reg_names[i], ret);
+		}
+	}
+}
+
 static void hdmi_bridge_pre_enable(struct drm_bridge *bridge)
 {
 	struct hdmi_bridge *hdmi_bridge = to_hdmi_bridge(bridge);
@@ -41,7 +97,14 @@ static void hdmi_bridge_pre_enable(struct drm_bridge *bridge)
 	struct hdmi_phy *phy = hdmi->phy;
 
 	DBG("power up");
-	phy->funcs->powerup(phy, hdmi_bridge->pixclock);
+
+	if (!hdmi->power_on) {
+		power_on(bridge);
+		hdmi->power_on = true;
+		hdmi_audio_update(hdmi);
+	}
+
+	phy->funcs->powerup(phy, hdmi->pixclock);
 	hdmi_set_mode(hdmi, true);
 }
 
@@ -62,6 +125,12 @@ static void hdmi_bridge_post_disable(struct drm_bridge *bridge)
 	DBG("power down");
 	hdmi_set_mode(hdmi, false);
 	phy->funcs->powerdown(phy);
+
+	if (hdmi->power_on) {
+		power_off(bridge);
+		hdmi->power_on = false;
+		hdmi_audio_update(hdmi);
+	}
 }
 
 static void hdmi_bridge_mode_set(struct drm_bridge *bridge,
@@ -75,7 +144,7 @@ static void hdmi_bridge_mode_set(struct drm_bridge *bridge,
 
 	mode = adjusted_mode;
 
-	hdmi_bridge->pixclock = mode->clock * 1000;
+	hdmi->pixclock = mode->clock * 1000;
 
 	hdmi->hdmi_mode = drm_match_cea_mode(mode) > 1;
 
@@ -123,9 +192,7 @@ static void hdmi_bridge_mode_set(struct drm_bridge *bridge,
 	DBG("frame_ctrl=%08x", frame_ctrl);
 	hdmi_write(hdmi, REG_HDMI_FRAME_CTRL, frame_ctrl);
 
-	// TODO until we have audio, this might be safest:
-	if (hdmi->hdmi_mode)
-		hdmi_write(hdmi, REG_HDMI_GC, HDMI_GC_MUTE);
+	hdmi_audio_update(hdmi);
 }
 
 static const struct drm_bridge_funcs hdmi_bridge_funcs = {
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_connector.c b/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
index 823eee521a3..28f7e3ec6c2 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
@@ -17,19 +17,20 @@
 
 #include <linux/gpio.h>
 
+#include "msm_kms.h"
 #include "hdmi.h"
 
 struct hdmi_connector {
 	struct drm_connector base;
 	struct hdmi *hdmi;
+	struct work_struct hpd_work;
 };
 #define to_hdmi_connector(x) container_of(x, struct hdmi_connector, base)
 
 static int gpio_config(struct hdmi *hdmi, bool on)
 {
 	struct drm_device *dev = hdmi->dev;
-	struct hdmi_platform_config *config =
-			hdmi->pdev->dev.platform_data;
+	const struct hdmi_platform_config *config = hdmi->config;
 	int ret;
 
 	if (on) {
@@ -39,26 +40,43 @@ static int gpio_config(struct hdmi *hdmi, bool on)
 				"HDMI_DDC_CLK", config->ddc_clk_gpio, ret);
 			goto error1;
 		}
+		gpio_set_value_cansleep(config->ddc_clk_gpio, 1);
+
 		ret = gpio_request(config->ddc_data_gpio, "HDMI_DDC_DATA");
 		if (ret) {
 			dev_err(dev->dev, "'%s'(%d) gpio_request failed: %d\n",
 				"HDMI_DDC_DATA", config->ddc_data_gpio, ret);
 			goto error2;
 		}
+		gpio_set_value_cansleep(config->ddc_data_gpio, 1);
+
 		ret = gpio_request(config->hpd_gpio, "HDMI_HPD");
 		if (ret) {
 			dev_err(dev->dev, "'%s'(%d) gpio_request failed: %d\n",
 				"HDMI_HPD", config->hpd_gpio, ret);
 			goto error3;
 		}
-		if (config->pmic_gpio != -1) {
-			ret = gpio_request(config->pmic_gpio, "PMIC_HDMI_MUX_SEL");
+		gpio_direction_input(config->hpd_gpio);
+		gpio_set_value_cansleep(config->hpd_gpio, 1);
+
+		if (config->mux_en_gpio != -1) {
+			ret = gpio_request(config->mux_en_gpio, "HDMI_MUX_EN");
 			if (ret) {
 				dev_err(dev->dev, "'%s'(%d) gpio_request failed: %d\n",
-					"PMIC_HDMI_MUX_SEL", config->pmic_gpio, ret);
+					"HDMI_MUX_SEL", config->mux_en_gpio, ret);
 				goto error4;
 			}
-			gpio_set_value_cansleep(config->pmic_gpio, 0);
+			gpio_set_value_cansleep(config->mux_en_gpio, 1);
+		}
+
+		if (config->mux_sel_gpio != -1) {
+			ret = gpio_request(config->mux_sel_gpio, "HDMI_MUX_SEL");
+			if (ret) {
+				dev_err(dev->dev, "'%s'(%d) gpio_request failed: %d\n",
+					"HDMI_MUX_SEL", config->mux_sel_gpio, ret);
+				goto error5;
+			}
+			gpio_set_value_cansleep(config->mux_sel_gpio, 0);
 		}
 		DBG("gpio on");
 	} else {
@@ -66,15 +84,23 @@ static int gpio_config(struct hdmi *hdmi, bool on)
 		gpio_free(config->ddc_data_gpio);
 		gpio_free(config->hpd_gpio);
 
-		if (config->pmic_gpio != -1) {
-			gpio_set_value_cansleep(config->pmic_gpio, 1);
-			gpio_free(config->pmic_gpio);
+		if (config->mux_en_gpio != -1) {
+			gpio_set_value_cansleep(config->mux_en_gpio, 0);
+			gpio_free(config->mux_en_gpio);
+		}
+
+		if (config->mux_sel_gpio != -1) {
+			gpio_set_value_cansleep(config->mux_sel_gpio, 1);
+			gpio_free(config->mux_sel_gpio);
 		}
 		DBG("gpio off");
 	}
 
 	return 0;
 
+error5:
+	if (config->mux_en_gpio != -1)
+		gpio_free(config->mux_en_gpio);
 error4:
 	gpio_free(config->hpd_gpio);
 error3:
@@ -88,10 +114,11 @@ error1:
 static int hpd_enable(struct hdmi_connector *hdmi_connector)
 {
 	struct hdmi *hdmi = hdmi_connector->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
 	struct drm_device *dev = hdmi_connector->base.dev;
 	struct hdmi_phy *phy = hdmi->phy;
 	uint32_t hpd_ctrl;
-	int ret;
+	int i, ret;
 
 	ret = gpio_config(hdmi, true);
 	if (ret) {
@@ -99,31 +126,30 @@ static int hpd_enable(struct hdmi_connector *hdmi_connector)
 		goto fail;
 	}
 
-	ret = clk_prepare_enable(hdmi->clk);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable 'clk': %d\n", ret);
-		goto fail;
-	}
-
-	ret = clk_prepare_enable(hdmi->m_pclk);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable 'm_pclk': %d\n", ret);
-		goto fail;
-	}
+	for (i = 0; i < config->hpd_clk_cnt; i++) {
+		if (config->hpd_freq && config->hpd_freq[i]) {
+			ret = clk_set_rate(hdmi->hpd_clks[i],
+					config->hpd_freq[i]);
+			if (ret)
+				dev_warn(dev->dev, "failed to set clk %s (%d)\n",
+						config->hpd_clk_names[i], ret);
+		}
 
-	ret = clk_prepare_enable(hdmi->s_pclk);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable 's_pclk': %d\n", ret);
-		goto fail;
+		ret = clk_prepare_enable(hdmi->hpd_clks[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to enable hpd clk: %s (%d)\n",
+					config->hpd_clk_names[i], ret);
+			goto fail;
+		}
 	}
 
-	if (hdmi->mpp0)
-		ret = regulator_enable(hdmi->mpp0);
-	if (!ret)
-		ret = regulator_enable(hdmi->mvs);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable regulators: %d\n", ret);
-		goto fail;
+	for (i = 0; i < config->hpd_reg_cnt; i++) {
+		ret = regulator_enable(hdmi->hpd_regs[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to enable hpd regulator: %s (%d)\n",
+					config->hpd_reg_names[i], ret);
+			goto fail;
+		}
 	}
 
 	hdmi_set_mode(hdmi, false);
@@ -156,26 +182,26 @@ fail:
 static int hdp_disable(struct hdmi_connector *hdmi_connector)
 {
 	struct hdmi *hdmi = hdmi_connector->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
 	struct drm_device *dev = hdmi_connector->base.dev;
-	int ret = 0;
+	int i, ret = 0;
 
 	/* Disable HPD interrupt */
 	hdmi_write(hdmi, REG_HDMI_HPD_INT_CTRL, 0);
 
 	hdmi_set_mode(hdmi, false);
 
-	if (hdmi->mpp0)
-		ret = regulator_disable(hdmi->mpp0);
-	if (!ret)
-		ret = regulator_disable(hdmi->mvs);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable regulators: %d\n", ret);
-		goto fail;
+	for (i = 0; i < config->hpd_reg_cnt; i++) {
+		ret = regulator_disable(hdmi->hpd_regs[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to disable hpd regulator: %s (%d)\n",
+					config->hpd_reg_names[i], ret);
+			goto fail;
+		}
 	}
 
-	clk_disable_unprepare(hdmi->clk);
-	clk_disable_unprepare(hdmi->m_pclk);
-	clk_disable_unprepare(hdmi->s_pclk);
+	for (i = 0; i < config->hpd_clk_cnt; i++)
+		clk_disable_unprepare(hdmi->hpd_clks[i]);
 
 	ret = gpio_config(hdmi, false);
 	if (ret) {
@@ -189,9 +215,19 @@ fail:
 	return ret;
 }
 
+static void
+hotplug_work(struct work_struct *work)
+{
+	struct hdmi_connector *hdmi_connector =
+		container_of(work, struct hdmi_connector, hpd_work);
+	struct drm_connector *connector = &hdmi_connector->base;
+	drm_helper_hpd_irq_event(connector->dev);
+}
+
 void hdmi_connector_irq(struct drm_connector *connector)
 {
 	struct hdmi_connector *hdmi_connector = to_hdmi_connector(connector);
+	struct msm_drm_private *priv = connector->dev->dev_private;
 	struct hdmi *hdmi = hdmi_connector->hdmi;
 	uint32_t hpd_int_status, hpd_int_ctrl;
 
@@ -209,37 +245,59 @@ void hdmi_connector_irq(struct drm_connector *connector)
 		hdmi_write(hdmi, REG_HDMI_HPD_INT_CTRL,
 				hpd_int_ctrl | HDMI_HPD_INT_CTRL_INT_ACK);
 
-		drm_helper_hpd_irq_event(connector->dev);
-
 		/* detect disconnect if we are connected or visa versa: */
 		hpd_int_ctrl = HDMI_HPD_INT_CTRL_INT_EN;
 		if (!detected)
 			hpd_int_ctrl |= HDMI_HPD_INT_CTRL_INT_CONNECT;
 		hdmi_write(hdmi, REG_HDMI_HPD_INT_CTRL, hpd_int_ctrl);
+
+		queue_work(priv->wq, &hdmi_connector->hpd_work);
 	}
 }
 
+static enum drm_connector_status detect_reg(struct hdmi *hdmi)
+{
+	uint32_t hpd_int_status = hdmi_read(hdmi, REG_HDMI_HPD_INT_STATUS);
+	return (hpd_int_status & HDMI_HPD_INT_STATUS_CABLE_DETECTED) ?
+			connector_status_connected : connector_status_disconnected;
+}
+
+static enum drm_connector_status detect_gpio(struct hdmi *hdmi)
+{
+	const struct hdmi_platform_config *config = hdmi->config;
+	return gpio_get_value(config->hpd_gpio) ?
+			connector_status_connected :
+			connector_status_disconnected;
+}
+
 static enum drm_connector_status hdmi_connector_detect(
 		struct drm_connector *connector, bool force)
 {
 	struct hdmi_connector *hdmi_connector = to_hdmi_connector(connector);
 	struct hdmi *hdmi = hdmi_connector->hdmi;
-	uint32_t hpd_int_status;
+	enum drm_connector_status stat_gpio, stat_reg;
 	int retry = 20;
 
-	hpd_int_status = hdmi_read(hdmi, REG_HDMI_HPD_INT_STATUS);
+	do {
+		stat_gpio = detect_gpio(hdmi);
+		stat_reg  = detect_reg(hdmi);
+
+		if (stat_gpio == stat_reg)
+			break;
 
-	/* sense seems to in some cases be momentarily de-asserted, don't
-	 * let that trick us into thinking the monitor is gone:
-	 */
-	while (retry-- && !(hpd_int_status & HDMI_HPD_INT_STATUS_CABLE_DETECTED)) {
 		mdelay(10);
-		hpd_int_status = hdmi_read(hdmi, REG_HDMI_HPD_INT_STATUS);
-		DBG("status=%08x", hpd_int_status);
+	} while (--retry);
+
+	/* the status we get from reading gpio seems to be more reliable,
+	 * so trust that one the most if we didn't manage to get hdmi and
+	 * gpio status to agree:
+	 */
+	if (stat_gpio != stat_reg) {
+		DBG("HDMI_HPD_INT_STATUS tells us: %d", stat_reg);
+		DBG("hpd gpio tells us: %d", stat_gpio);
 	}
 
-	return (hpd_int_status & HDMI_HPD_INT_STATUS_CABLE_DETECTED) ?
-			connector_status_connected : connector_status_disconnected;
+	return stat_gpio;
 }
 
 static void hdmi_connector_destroy(struct drm_connector *connector)
@@ -285,6 +343,8 @@ static int hdmi_connector_mode_valid(struct drm_connector *connector,
 				 struct drm_display_mode *mode)
 {
 	struct hdmi_connector *hdmi_connector = to_hdmi_connector(connector);
+	struct hdmi *hdmi = hdmi_connector->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
 	struct msm_drm_private *priv = connector->dev->dev_private;
 	struct msm_kms *kms = priv->kms;
 	long actual, requested;
@@ -293,6 +353,13 @@ static int hdmi_connector_mode_valid(struct drm_connector *connector,
 	actual = kms->funcs->round_pixclk(kms,
 			requested, hdmi_connector->hdmi->encoder);
 
+	/* for mdp5/apq8074, we manage our own pixel clk (as opposed to
+	 * mdp4/dtv stuff where pixel clk is assigned to mdp/encoder
+	 * instead):
+	 */
+	if (config->pwr_clk_cnt > 0)
+		actual = clk_round_rate(hdmi->pwr_clks[0], actual);
+
 	DBG("requested=%ld, actual=%ld", requested, actual);
 
 	if (actual != requested)
@@ -335,6 +402,7 @@ struct drm_connector *hdmi_connector_init(struct hdmi *hdmi)
 	}
 
 	hdmi_connector->hdmi = hdmi_reference(hdmi);
+	INIT_WORK(&hdmi_connector->hpd_work, hotplug_work);
 
 	connector = &hdmi_connector->base;
 
@@ -342,7 +410,8 @@ struct drm_connector *hdmi_connector_init(struct hdmi *hdmi)
 			DRM_MODE_CONNECTOR_HDMIA);
 	drm_connector_helper_add(connector, &hdmi_connector_helper_funcs);
 
-	connector->polled = DRM_CONNECTOR_POLL_HPD;
+	connector->polled = DRM_CONNECTOR_POLL_CONNECT |
+			DRM_CONNECTOR_POLL_DISCONNECT;
 
 	connector->interlace_allowed = 1;
 	connector->doublescan_allowed = 0;
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_phy_8x74.c b/drivers/gpu/drm/msm/hdmi/hdmi_phy_8x74.c
new file mode 100644
index 00000000000..59fa6cdacb2
--- /dev/null
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_phy_8x74.c
@@ -0,0 +1,157 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "hdmi.h"
+
+struct hdmi_phy_8x74 {
+	struct hdmi_phy base;
+	struct hdmi *hdmi;
+	void __iomem *mmio;
+};
+#define to_hdmi_phy_8x74(x) container_of(x, struct hdmi_phy_8x74, base)
+
+
+static void phy_write(struct hdmi_phy_8x74 *phy, u32 reg, u32 data)
+{
+	msm_writel(data, phy->mmio + reg);
+}
+
+//static u32 phy_read(struct hdmi_phy_8x74 *phy, u32 reg)
+//{
+//	return msm_readl(phy->mmio + reg);
+//}
+
+static void hdmi_phy_8x74_destroy(struct hdmi_phy *phy)
+{
+	struct hdmi_phy_8x74 *phy_8x74 = to_hdmi_phy_8x74(phy);
+	kfree(phy_8x74);
+}
+
+static void hdmi_phy_8x74_reset(struct hdmi_phy *phy)
+{
+	struct hdmi_phy_8x74 *phy_8x74 = to_hdmi_phy_8x74(phy);
+	struct hdmi *hdmi = phy_8x74->hdmi;
+	unsigned int val;
+
+	/* NOTE that HDMI_PHY_CTL is in core mmio, not phy mmio: */
+
+	val = hdmi_read(hdmi, REG_HDMI_PHY_CTRL);
+
+	if (val & HDMI_PHY_CTRL_SW_RESET_LOW) {
+		/* pull low */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val & ~HDMI_PHY_CTRL_SW_RESET);
+	} else {
+		/* pull high */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val | HDMI_PHY_CTRL_SW_RESET);
+	}
+
+	if (val & HDMI_PHY_CTRL_SW_RESET_PLL_LOW) {
+		/* pull low */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val & ~HDMI_PHY_CTRL_SW_RESET_PLL);
+	} else {
+		/* pull high */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val | HDMI_PHY_CTRL_SW_RESET_PLL);
+	}
+
+	msleep(100);
+
+	if (val & HDMI_PHY_CTRL_SW_RESET_LOW) {
+		/* pull high */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val | HDMI_PHY_CTRL_SW_RESET);
+	} else {
+		/* pull low */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val & ~HDMI_PHY_CTRL_SW_RESET);
+	}
+
+	if (val & HDMI_PHY_CTRL_SW_RESET_PLL_LOW) {
+		/* pull high */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val | HDMI_PHY_CTRL_SW_RESET_PLL);
+	} else {
+		/* pull low */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val & ~HDMI_PHY_CTRL_SW_RESET_PLL);
+	}
+}
+
+static void hdmi_phy_8x74_powerup(struct hdmi_phy *phy,
+		unsigned long int pixclock)
+{
+	struct hdmi_phy_8x74 *phy_8x74 = to_hdmi_phy_8x74(phy);
+
+	phy_write(phy_8x74, REG_HDMI_8x74_ANA_CFG0,   0x1b);
+	phy_write(phy_8x74, REG_HDMI_8x74_ANA_CFG1,   0xf2);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_CFG0,  0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_PATN0, 0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_PATN1, 0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_PATN2, 0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_PATN3, 0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_PD_CTRL1,   0x20);
+}
+
+static void hdmi_phy_8x74_powerdown(struct hdmi_phy *phy)
+{
+	struct hdmi_phy_8x74 *phy_8x74 = to_hdmi_phy_8x74(phy);
+	phy_write(phy_8x74, REG_HDMI_8x74_PD_CTRL0, 0x7f);
+}
+
+static const struct hdmi_phy_funcs hdmi_phy_8x74_funcs = {
+		.destroy = hdmi_phy_8x74_destroy,
+		.reset = hdmi_phy_8x74_reset,
+		.powerup = hdmi_phy_8x74_powerup,
+		.powerdown = hdmi_phy_8x74_powerdown,
+};
+
+struct hdmi_phy *hdmi_phy_8x74_init(struct hdmi *hdmi)
+{
+	struct hdmi_phy_8x74 *phy_8x74;
+	struct hdmi_phy *phy = NULL;
+	int ret;
+
+	phy_8x74 = kzalloc(sizeof(*phy_8x74), GFP_KERNEL);
+	if (!phy_8x74) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	phy = &phy_8x74->base;
+
+	phy->funcs = &hdmi_phy_8x74_funcs;
+
+	phy_8x74->hdmi = hdmi;
+
+	/* for 8x74, the phy mmio is mapped separately: */
+	phy_8x74->mmio = msm_ioremap(hdmi->pdev,
+			"phy_physical", "HDMI_8x74");
+	if (IS_ERR(phy_8x74->mmio)) {
+		ret = PTR_ERR(phy_8x74->mmio);
+		goto fail;
+	}
+
+	return phy;
+
+fail:
+	if (phy)
+		hdmi_phy_8x74_destroy(phy);
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/hdmi/qfprom.xml.h b/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
index dbde4f6339b..d591567173c 100644
--- a/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
+++ b/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    595 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2013-12-03 20:59:13)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4.xml.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
index 9908ffe1c3a..416a26e1e58 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4.xml.h
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 git clone https://github.com/freedreno/envytools.git
 
 The rules-ng-ng source files this header was generated from are:
-- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    595 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
 - /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2013-12-03 20:59:13)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
 - /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
 - /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -42,27 +44,6 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 */
 
 
-enum mdp4_bpc {
-	BPC1 = 0,
-	BPC5 = 1,
-	BPC6 = 2,
-	BPC8 = 3,
-};
-
-enum mdp4_bpc_alpha {
-	BPC1A = 0,
-	BPC4A = 1,
-	BPC6A = 2,
-	BPC8A = 3,
-};
-
-enum mdp4_alpha_type {
-	FG_CONST = 0,
-	BG_CONST = 1,
-	FG_PIXEL = 2,
-	BG_PIXEL = 3,
-};
-
 enum mdp4_pipe {
 	VG1 = 0,
 	VG2 = 1,
@@ -79,15 +60,6 @@ enum mdp4_mixer {
 	MIXER2 = 2,
 };
 
-enum mdp4_mixer_stage_id {
-	STAGE_UNUSED = 0,
-	STAGE_BASE = 1,
-	STAGE0 = 2,
-	STAGE1 = 3,
-	STAGE2 = 4,
-	STAGE3 = 5,
-};
-
 enum mdp4_intf {
 	INTF_LCDC_DTV = 0,
 	INTF_DSI_VIDEO = 1,
@@ -194,56 +166,56 @@ static inline uint32_t MDP4_DISP_INTF_SEL_EXT(enum mdp4_intf val)
 #define REG_MDP4_LAYERMIXER2_IN_CFG				0x000100f0
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE0__MASK			0x00000007
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE0__SHIFT			0
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE0(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE0(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE0__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE0__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE0_MIXER1			0x00000008
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE1__MASK			0x00000070
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE1__SHIFT			4
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE1(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE1(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE1__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE1__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE1_MIXER1			0x00000080
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE2__MASK			0x00000700
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE2__SHIFT			8
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE2(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE2(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE2__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE2__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE2_MIXER1			0x00000800
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE3__MASK			0x00007000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE3__SHIFT			12
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE3(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE3(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE3__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE3__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE3_MIXER1			0x00008000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE4__MASK			0x00070000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE4__SHIFT			16
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE4(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE4(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE4__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE4__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE4_MIXER1			0x00080000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE5__MASK			0x00700000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE5__SHIFT			20
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE5(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE5(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE5__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE5__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE5_MIXER1			0x00800000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE6__MASK			0x07000000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE6__SHIFT			24
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE6(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE6(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE6__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE6__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE6_MIXER1			0x08000000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE7__MASK			0x70000000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE7__SHIFT			28
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE7(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE7(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE7__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE7__MASK;
 }
@@ -254,56 +226,56 @@ static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE7(enum mdp4_mixer_stage_id va
 #define REG_MDP4_LAYERMIXER_IN_CFG				0x00010100
 #define MDP4_LAYERMIXER_IN_CFG_PIPE0__MASK			0x00000007
 #define MDP4_LAYERMIXER_IN_CFG_PIPE0__SHIFT			0
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE0(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE0(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE0__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE0__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE0_MIXER1			0x00000008
 #define MDP4_LAYERMIXER_IN_CFG_PIPE1__MASK			0x00000070
 #define MDP4_LAYERMIXER_IN_CFG_PIPE1__SHIFT			4
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE1(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE1(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE1__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE1__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE1_MIXER1			0x00000080
 #define MDP4_LAYERMIXER_IN_CFG_PIPE2__MASK			0x00000700
 #define MDP4_LAYERMIXER_IN_CFG_PIPE2__SHIFT			8
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE2(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE2(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE2__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE2__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE2_MIXER1			0x00000800
 #define MDP4_LAYERMIXER_IN_CFG_PIPE3__MASK			0x00007000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE3__SHIFT			12
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE3(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE3(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE3__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE3__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE3_MIXER1			0x00008000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE4__MASK			0x00070000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE4__SHIFT			16
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE4(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE4(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE4__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE4__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE4_MIXER1			0x00080000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE5__MASK			0x00700000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE5__SHIFT			20
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE5(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE5(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE5__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE5__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE5_MIXER1			0x00800000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE6__MASK			0x07000000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE6__SHIFT			24
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE6(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE6(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE6__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE6__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE6_MIXER1			0x08000000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE7__MASK			0x70000000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE7__SHIFT			28
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE7(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE7(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE7__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE7__MASK;
 }
@@ -369,7 +341,7 @@ static inline uint32_t REG_MDP4_OVLP_STAGE(uint32_t i0, uint32_t i1) { return 0x
 static inline uint32_t REG_MDP4_OVLP_STAGE_OP(uint32_t i0, uint32_t i1) { return 0x00000000 + __offset_OVLP(i0) + __offset_STAGE(i1); }
 #define MDP4_OVLP_STAGE_OP_FG_ALPHA__MASK			0x00000003
 #define MDP4_OVLP_STAGE_OP_FG_ALPHA__SHIFT			0
-static inline uint32_t MDP4_OVLP_STAGE_OP_FG_ALPHA(enum mdp4_alpha_type val)
+static inline uint32_t MDP4_OVLP_STAGE_OP_FG_ALPHA(enum mdp_alpha_type val)
 {
 	return ((val) << MDP4_OVLP_STAGE_OP_FG_ALPHA__SHIFT) & MDP4_OVLP_STAGE_OP_FG_ALPHA__MASK;
 }
@@ -377,7 +349,7 @@ static inline uint32_t MDP4_OVLP_STAGE_OP_FG_ALPHA(enum mdp4_alpha_type val)
 #define MDP4_OVLP_STAGE_OP_FG_MOD_ALPHA				0x00000008
 #define MDP4_OVLP_STAGE_OP_BG_ALPHA__MASK			0x00000030
 #define MDP4_OVLP_STAGE_OP_BG_ALPHA__SHIFT			4
-static inline uint32_t MDP4_OVLP_STAGE_OP_BG_ALPHA(enum mdp4_alpha_type val)
+static inline uint32_t MDP4_OVLP_STAGE_OP_BG_ALPHA(enum mdp_alpha_type val)
 {
 	return ((val) << MDP4_OVLP_STAGE_OP_BG_ALPHA__SHIFT) & MDP4_OVLP_STAGE_OP_BG_ALPHA__MASK;
 }
@@ -472,19 +444,19 @@ static inline uint32_t REG_MDP4_DMA(enum mdp4_dma i0) { return 0x00000000 + __of
 static inline uint32_t REG_MDP4_DMA_CONFIG(enum mdp4_dma i0) { return 0x00000000 + __offset_DMA(i0); }
 #define MDP4_DMA_CONFIG_G_BPC__MASK				0x00000003
 #define MDP4_DMA_CONFIG_G_BPC__SHIFT				0
-static inline uint32_t MDP4_DMA_CONFIG_G_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_DMA_CONFIG_G_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_DMA_CONFIG_G_BPC__SHIFT) & MDP4_DMA_CONFIG_G_BPC__MASK;
 }
 #define MDP4_DMA_CONFIG_B_BPC__MASK				0x0000000c
 #define MDP4_DMA_CONFIG_B_BPC__SHIFT				2
-static inline uint32_t MDP4_DMA_CONFIG_B_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_DMA_CONFIG_B_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_DMA_CONFIG_B_BPC__SHIFT) & MDP4_DMA_CONFIG_B_BPC__MASK;
 }
 #define MDP4_DMA_CONFIG_R_BPC__MASK				0x00000030
 #define MDP4_DMA_CONFIG_R_BPC__SHIFT				4
-static inline uint32_t MDP4_DMA_CONFIG_R_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_DMA_CONFIG_R_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_DMA_CONFIG_R_BPC__SHIFT) & MDP4_DMA_CONFIG_R_BPC__MASK;
 }
@@ -710,25 +682,25 @@ static inline uint32_t MDP4_PIPE_FRAME_SIZE_WIDTH(uint32_t val)
 static inline uint32_t REG_MDP4_PIPE_SRC_FORMAT(enum mdp4_pipe i0) { return 0x00020050 + 0x10000*i0; }
 #define MDP4_PIPE_SRC_FORMAT_G_BPC__MASK			0x00000003
 #define MDP4_PIPE_SRC_FORMAT_G_BPC__SHIFT			0
-static inline uint32_t MDP4_PIPE_SRC_FORMAT_G_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_PIPE_SRC_FORMAT_G_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_PIPE_SRC_FORMAT_G_BPC__SHIFT) & MDP4_PIPE_SRC_FORMAT_G_BPC__MASK;
 }
 #define MDP4_PIPE_SRC_FORMAT_B_BPC__MASK			0x0000000c
 #define MDP4_PIPE_SRC_FORMAT_B_BPC__SHIFT			2
-static inline uint32_t MDP4_PIPE_SRC_FORMAT_B_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_PIPE_SRC_FORMAT_B_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_PIPE_SRC_FORMAT_B_BPC__SHIFT) & MDP4_PIPE_SRC_FORMAT_B_BPC__MASK;
 }
 #define MDP4_PIPE_SRC_FORMAT_R_BPC__MASK			0x00000030
 #define MDP4_PIPE_SRC_FORMAT_R_BPC__SHIFT			4
-static inline uint32_t MDP4_PIPE_SRC_FORMAT_R_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_PIPE_SRC_FORMAT_R_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_PIPE_SRC_FORMAT_R_BPC__SHIFT) & MDP4_PIPE_SRC_FORMAT_R_BPC__MASK;
 }
 #define MDP4_PIPE_SRC_FORMAT_A_BPC__MASK			0x000000c0
 #define MDP4_PIPE_SRC_FORMAT_A_BPC__SHIFT			6
-static inline uint32_t MDP4_PIPE_SRC_FORMAT_A_BPC(enum mdp4_bpc_alpha val)
+static inline uint32_t MDP4_PIPE_SRC_FORMAT_A_BPC(enum mdp_bpc_alpha val)
 {
 	return ((val) << MDP4_PIPE_SRC_FORMAT_A_BPC__SHIFT) & MDP4_PIPE_SRC_FORMAT_A_BPC__MASK;
 }
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
index 019d530187f..74cebb51e8c 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_crtc.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
@@ -39,6 +39,7 @@ struct mdp4_crtc {
 		spinlock_t lock;
 		bool stale;
 		uint32_t width, height;
+		uint32_t x, y;
 
 		/* next cursor to scan-out: */
 		uint32_t next_iova;
@@ -57,44 +58,100 @@ struct mdp4_crtc {
 #define PENDING_FLIP   0x2
 	atomic_t pending;
 
-	/* the fb that we currently hold a scanout ref to: */
+	/* the fb that we logically (from PoV of KMS API) hold a ref
+	 * to.  Which we may not yet be scanning out (we may still
+	 * be scanning out previous in case of page_flip while waiting
+	 * for gpu rendering to complete:
+	 */
 	struct drm_framebuffer *fb;
 
+	/* the fb that we currently hold a scanout ref to: */
+	struct drm_framebuffer *scanout_fb;
+
 	/* for unref'ing framebuffers after scanout completes: */
 	struct drm_flip_work unref_fb_work;
 
 	/* for unref'ing cursor bo's after scanout completes: */
 	struct drm_flip_work unref_cursor_work;
 
-	struct mdp4_irq vblank;
-	struct mdp4_irq err;
+	struct mdp_irq vblank;
+	struct mdp_irq err;
 };
 #define to_mdp4_crtc(x) container_of(x, struct mdp4_crtc, base)
 
 static struct mdp4_kms *get_kms(struct drm_crtc *crtc)
 {
 	struct msm_drm_private *priv = crtc->dev->dev_private;
-	return to_mdp4_kms(priv->kms);
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
 }
 
-static void update_fb(struct drm_crtc *crtc, bool async,
-		struct drm_framebuffer *new_fb)
+static void request_pending(struct drm_crtc *crtc, uint32_t pending)
 {
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
-	struct drm_framebuffer *old_fb = mdp4_crtc->fb;
 
-	if (old_fb)
-		drm_flip_work_queue(&mdp4_crtc->unref_fb_work, old_fb);
+	atomic_or(pending, &mdp4_crtc->pending);
+	mdp_irq_register(&get_kms(crtc)->base, &mdp4_crtc->vblank);
+}
+
+static void crtc_flush(struct drm_crtc *crtc)
+{
+	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
+	struct mdp4_kms *mdp4_kms = get_kms(crtc);
+	uint32_t i, flush = 0;
+
+	for (i = 0; i < ARRAY_SIZE(mdp4_crtc->planes); i++) {
+		struct drm_plane *plane = mdp4_crtc->planes[i];
+		if (plane) {
+			enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane);
+			flush |= pipe2flush(pipe_id);
+		}
+	}
+	flush |= ovlp2flush(mdp4_crtc->ovlp);
+
+	DBG("%s: flush=%08x", mdp4_crtc->name, flush);
+
+	mdp4_write(mdp4_kms, REG_MDP4_OVERLAY_FLUSH, flush);
+}
+
+static void update_fb(struct drm_crtc *crtc, struct drm_framebuffer *new_fb)
+{
+	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
+	struct drm_framebuffer *old_fb = mdp4_crtc->fb;
 
 	/* grab reference to incoming scanout fb: */
 	drm_framebuffer_reference(new_fb);
-	mdp4_crtc->base.fb = new_fb;
+	mdp4_crtc->base.primary->fb = new_fb;
 	mdp4_crtc->fb = new_fb;
 
-	if (!async) {
-		/* enable vblank to pick up the old_fb */
-		mdp4_irq_register(get_kms(crtc), &mdp4_crtc->vblank);
-	}
+	if (old_fb)
+		drm_flip_work_queue(&mdp4_crtc->unref_fb_work, old_fb);
+}
+
+/* unlike update_fb(), take a ref to the new scanout fb *before* updating
+ * plane, then call this.  Needed to ensure we don't unref the buffer that
+ * is actually still being scanned out.
+ *
+ * Note that this whole thing goes away with atomic.. since we can defer
+ * calling into driver until rendering is done.
+ */
+static void update_scanout(struct drm_crtc *crtc, struct drm_framebuffer *fb)
+{
+	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
+
+	/* flush updates, to make sure hw is updated to new scanout fb,
+	 * so that we can safely queue unref to current fb (ie. next
+	 * vblank we know hw is done w/ previous scanout_fb).
+	 */
+	crtc_flush(crtc);
+
+	if (mdp4_crtc->scanout_fb)
+		drm_flip_work_queue(&mdp4_crtc->unref_fb_work,
+				mdp4_crtc->scanout_fb);
+
+	mdp4_crtc->scanout_fb = fb;
+
+	/* enable vblank to complete flip: */
+	request_pending(crtc, PENDING_FLIP);
 }
 
 /* if file!=NULL, this is preclose potential cancel-flip path */
@@ -120,49 +177,19 @@ static void complete_flip(struct drm_crtc *crtc, struct drm_file *file)
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 }
 
-static void crtc_flush(struct drm_crtc *crtc)
-{
-	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
-	struct mdp4_kms *mdp4_kms = get_kms(crtc);
-	uint32_t i, flush = 0;
-
-	for (i = 0; i < ARRAY_SIZE(mdp4_crtc->planes); i++) {
-		struct drm_plane *plane = mdp4_crtc->planes[i];
-		if (plane) {
-			enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane);
-			flush |= pipe2flush(pipe_id);
-		}
-	}
-	flush |= ovlp2flush(mdp4_crtc->ovlp);
-
-	DBG("%s: flush=%08x", mdp4_crtc->name, flush);
-
-	mdp4_write(mdp4_kms, REG_MDP4_OVERLAY_FLUSH, flush);
-}
-
-static void request_pending(struct drm_crtc *crtc, uint32_t pending)
-{
-	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
-
-	atomic_or(pending, &mdp4_crtc->pending);
-	mdp4_irq_register(get_kms(crtc), &mdp4_crtc->vblank);
-}
-
 static void pageflip_cb(struct msm_fence_cb *cb)
 {
 	struct mdp4_crtc *mdp4_crtc =
 		container_of(cb, struct mdp4_crtc, pageflip_cb);
 	struct drm_crtc *crtc = &mdp4_crtc->base;
-	struct drm_framebuffer *fb = crtc->fb;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 
 	if (!fb)
 		return;
 
+	drm_framebuffer_reference(fb);
 	mdp4_plane_set_scanout(mdp4_crtc->plane, fb);
-	crtc_flush(crtc);
-
-	/* enable vblank to complete flip: */
-	request_pending(crtc, PENDING_FLIP);
+	update_scanout(crtc, fb);
 }
 
 static void unref_fb_worker(struct drm_flip_work *work, void *val)
@@ -190,8 +217,6 @@ static void mdp4_crtc_destroy(struct drm_crtc *crtc)
 {
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
 
-	mdp4_crtc->plane->funcs->destroy(mdp4_crtc->plane);
-
 	drm_crtc_cleanup(crtc);
 	drm_flip_work_cleanup(&mdp4_crtc->unref_fb_work);
 	drm_flip_work_cleanup(&mdp4_crtc->unref_cursor_work);
@@ -210,9 +235,9 @@ static void mdp4_crtc_dpms(struct drm_crtc *crtc, int mode)
 	if (enabled != mdp4_crtc->enabled) {
 		if (enabled) {
 			mdp4_enable(mdp4_kms);
-			mdp4_irq_register(mdp4_kms, &mdp4_crtc->err);
+			mdp_irq_register(&mdp4_kms->base, &mdp4_crtc->err);
 		} else {
-			mdp4_irq_unregister(mdp4_kms, &mdp4_crtc->err);
+			mdp_irq_unregister(&mdp4_kms->base, &mdp4_crtc->err);
 			mdp4_disable(mdp4_kms);
 		}
 		mdp4_crtc->enabled = enabled;
@@ -232,7 +257,7 @@ static void blend_setup(struct drm_crtc *crtc)
 	struct mdp4_kms *mdp4_kms = get_kms(crtc);
 	int i, ovlp = mdp4_crtc->ovlp;
 	uint32_t mixer_cfg = 0;
-	static const enum mdp4_mixer_stage_id stages[] = {
+	static const enum mdp_mixer_stage_id stages[] = {
 			STAGE_BASE, STAGE0, STAGE1, STAGE2, STAGE3,
 	};
 	/* statically (for now) map planes to mixer stage (z-order): */
@@ -262,8 +287,8 @@ static void blend_setup(struct drm_crtc *crtc)
 			enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane);
 			int idx = idxs[pipe_id];
 			if (idx > 0) {
-				const struct mdp4_format *format =
-					to_mdp4_format(msm_framebuffer_format(plane->fb));
+				const struct mdp_format *format =
+					to_mdp_format(msm_framebuffer_format(plane->fb));
 				alpha[idx-1] = format->alpha_enable;
 			}
 			mixer_cfg |= mixercfg(mdp4_crtc->mixer, pipe_id, stages[idx]);
@@ -320,6 +345,20 @@ static int mdp4_crtc_mode_set(struct drm_crtc *crtc,
 			mode->vsync_end, mode->vtotal,
 			mode->type, mode->flags);
 
+	/* grab extra ref for update_scanout() */
+	drm_framebuffer_reference(crtc->primary->fb);
+
+	ret = mdp4_plane_mode_set(mdp4_crtc->plane, crtc, crtc->primary->fb,
+			0, 0, mode->hdisplay, mode->vdisplay,
+			x << 16, y << 16,
+			mode->hdisplay << 16, mode->vdisplay << 16);
+	if (ret) {
+		drm_framebuffer_unreference(crtc->primary->fb);
+		dev_err(crtc->dev->dev, "%s: failed to set mode on plane: %d\n",
+				mdp4_crtc->name, ret);
+		return ret;
+	}
+
 	mdp4_write(mdp4_kms, REG_MDP4_DMA_SRC_SIZE(dma),
 			MDP4_DMA_SRC_SIZE_WIDTH(mode->hdisplay) |
 			MDP4_DMA_SRC_SIZE_HEIGHT(mode->vdisplay));
@@ -327,7 +366,7 @@ static int mdp4_crtc_mode_set(struct drm_crtc *crtc,
 	/* take data from pipe: */
 	mdp4_write(mdp4_kms, REG_MDP4_DMA_SRC_BASE(dma), 0);
 	mdp4_write(mdp4_kms, REG_MDP4_DMA_SRC_STRIDE(dma),
-			crtc->fb->pitches[0]);
+			crtc->primary->fb->pitches[0]);
 	mdp4_write(mdp4_kms, REG_MDP4_DMA_DST_SIZE(dma),
 			MDP4_DMA_DST_SIZE_WIDTH(0) |
 			MDP4_DMA_DST_SIZE_HEIGHT(0));
@@ -337,28 +376,19 @@ static int mdp4_crtc_mode_set(struct drm_crtc *crtc,
 			MDP4_OVLP_SIZE_WIDTH(mode->hdisplay) |
 			MDP4_OVLP_SIZE_HEIGHT(mode->vdisplay));
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_STRIDE(ovlp),
-			crtc->fb->pitches[0]);
+			crtc->primary->fb->pitches[0]);
 
 	mdp4_write(mdp4_kms, REG_MDP4_OVLP_CFG(ovlp), 1);
 
-	update_fb(crtc, false, crtc->fb);
-
-	ret = mdp4_plane_mode_set(mdp4_crtc->plane, crtc, crtc->fb,
-			0, 0, mode->hdisplay, mode->vdisplay,
-			x << 16, y << 16,
-			mode->hdisplay << 16, mode->vdisplay << 16);
-	if (ret) {
-		dev_err(crtc->dev->dev, "%s: failed to set mode on plane: %d\n",
-				mdp4_crtc->name, ret);
-		return ret;
-	}
-
 	if (dma == DMA_E) {
 		mdp4_write(mdp4_kms, REG_MDP4_DMA_E_QUANT(0), 0x00ff0000);
 		mdp4_write(mdp4_kms, REG_MDP4_DMA_E_QUANT(1), 0x00ff0000);
 		mdp4_write(mdp4_kms, REG_MDP4_DMA_E_QUANT(2), 0x00ff0000);
 	}
 
+	update_fb(crtc, crtc->primary->fb);
+	update_scanout(crtc, crtc->primary->fb);
+
 	return 0;
 }
 
@@ -385,13 +415,24 @@ static int mdp4_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
 	struct drm_plane *plane = mdp4_crtc->plane;
 	struct drm_display_mode *mode = &crtc->mode;
+	int ret;
 
-	update_fb(crtc, false, crtc->fb);
+	/* grab extra ref for update_scanout() */
+	drm_framebuffer_reference(crtc->primary->fb);
 
-	return mdp4_plane_mode_set(plane, crtc, crtc->fb,
+	ret = mdp4_plane_mode_set(plane, crtc, crtc->primary->fb,
 			0, 0, mode->hdisplay, mode->vdisplay,
 			x << 16, y << 16,
 			mode->hdisplay << 16, mode->vdisplay << 16);
+	if (ret) {
+		drm_framebuffer_unreference(crtc->primary->fb);
+		return ret;
+	}
+
+	update_fb(crtc, crtc->primary->fb);
+	update_scanout(crtc, crtc->primary->fb);
+
+	return 0;
 }
 
 static void mdp4_crtc_load_lut(struct drm_crtc *crtc)
@@ -419,7 +460,7 @@ static int mdp4_crtc_page_flip(struct drm_crtc *crtc,
 	mdp4_crtc->event = event;
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	update_fb(crtc, true, new_fb);
+	update_fb(crtc, new_fb);
 
 	return msm_gem_queue_inactive_cb(obj, &mdp4_crtc->pageflip_cb);
 }
@@ -442,12 +483,12 @@ static int mdp4_crtc_set_property(struct drm_crtc *crtc,
 static void update_cursor(struct drm_crtc *crtc)
 {
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
+	struct mdp4_kms *mdp4_kms = get_kms(crtc);
 	enum mdp4_dma dma = mdp4_crtc->dma;
 	unsigned long flags;
 
 	spin_lock_irqsave(&mdp4_crtc->cursor.lock, flags);
 	if (mdp4_crtc->cursor.stale) {
-		struct mdp4_kms *mdp4_kms = get_kms(crtc);
 		struct drm_gem_object *next_bo = mdp4_crtc->cursor.next_bo;
 		struct drm_gem_object *prev_bo = mdp4_crtc->cursor.scanout_bo;
 		uint32_t iova = mdp4_crtc->cursor.next_iova;
@@ -467,9 +508,8 @@ static void update_cursor(struct drm_crtc *crtc)
 					MDP4_DMA_CURSOR_BLEND_CONFIG_CURSOR_EN);
 		} else {
 			/* disable cursor: */
-			mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_BASE(dma), 0);
-			mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_BLEND_CONFIG(dma),
-					MDP4_DMA_CURSOR_BLEND_CONFIG_FORMAT(CURSOR_ARGB));
+			mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_BASE(dma),
+					mdp4_kms->blank_cursor_iova);
 		}
 
 		/* and drop the iova ref + obj rev when done scanning out: */
@@ -479,6 +519,11 @@ static void update_cursor(struct drm_crtc *crtc)
 		mdp4_crtc->cursor.scanout_bo = next_bo;
 		mdp4_crtc->cursor.stale = false;
 	}
+
+	mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_POS(dma),
+			MDP4_DMA_CURSOR_POS_X(mdp4_crtc->cursor.x) |
+			MDP4_DMA_CURSOR_POS_Y(mdp4_crtc->cursor.y));
+
 	spin_unlock_irqrestore(&mdp4_crtc->cursor.lock, flags);
 }
 
@@ -526,8 +571,7 @@ static int mdp4_crtc_cursor_set(struct drm_crtc *crtc,
 
 	if (old_bo) {
 		/* drop our previous reference: */
-		msm_gem_put_iova(old_bo, mdp4_kms->id);
-		drm_gem_object_unreference_unlocked(old_bo);
+		drm_flip_work_queue(&mdp4_crtc->unref_cursor_work, old_bo);
 	}
 
 	request_pending(crtc, PENDING_CURSOR);
@@ -542,12 +586,15 @@ fail:
 static int mdp4_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
 {
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
-	struct mdp4_kms *mdp4_kms = get_kms(crtc);
-	enum mdp4_dma dma = mdp4_crtc->dma;
+	unsigned long flags;
 
-	mdp4_write(mdp4_kms, REG_MDP4_DMA_CURSOR_POS(dma),
-			MDP4_DMA_CURSOR_POS_X(x) |
-			MDP4_DMA_CURSOR_POS_Y(y));
+	spin_lock_irqsave(&mdp4_crtc->cursor.lock, flags);
+	mdp4_crtc->cursor.x = x;
+	mdp4_crtc->cursor.y = y;
+	spin_unlock_irqrestore(&mdp4_crtc->cursor.lock, flags);
+
+	crtc_flush(crtc);
+	request_pending(crtc, PENDING_CURSOR);
 
 	return 0;
 }
@@ -571,14 +618,14 @@ static const struct drm_crtc_helper_funcs mdp4_crtc_helper_funcs = {
 	.load_lut = mdp4_crtc_load_lut,
 };
 
-static void mdp4_crtc_vblank_irq(struct mdp4_irq *irq, uint32_t irqstatus)
+static void mdp4_crtc_vblank_irq(struct mdp_irq *irq, uint32_t irqstatus)
 {
 	struct mdp4_crtc *mdp4_crtc = container_of(irq, struct mdp4_crtc, vblank);
 	struct drm_crtc *crtc = &mdp4_crtc->base;
 	struct msm_drm_private *priv = crtc->dev->dev_private;
 	unsigned pending;
 
-	mdp4_irq_unregister(get_kms(crtc), &mdp4_crtc->vblank);
+	mdp_irq_unregister(&get_kms(crtc)->base, &mdp4_crtc->vblank);
 
 	pending = atomic_xchg(&mdp4_crtc->pending, 0);
 
@@ -593,7 +640,7 @@ static void mdp4_crtc_vblank_irq(struct mdp4_irq *irq, uint32_t irqstatus)
 	}
 }
 
-static void mdp4_crtc_err_irq(struct mdp4_irq *irq, uint32_t irqstatus)
+static void mdp4_crtc_err_irq(struct mdp_irq *irq, uint32_t irqstatus)
 {
 	struct mdp4_crtc *mdp4_crtc = container_of(irq, struct mdp4_crtc, err);
 	struct drm_crtc *crtc = &mdp4_crtc->base;
@@ -688,6 +735,9 @@ void mdp4_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane)
 
 void mdp4_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane)
 {
+	/* don't actually detatch our primary plane: */
+	if (to_mdp4_crtc(crtc)->plane == plane)
+		return;
 	set_attach(crtc, mdp4_plane_pipe(plane), NULL);
 }
 
@@ -713,6 +763,7 @@ struct drm_crtc *mdp4_crtc_init(struct drm_device *dev,
 	crtc = &mdp4_crtc->base;
 
 	mdp4_crtc->plane = plane;
+	mdp4_crtc->id = id;
 
 	mdp4_crtc->ovlp = ovlp_id;
 	mdp4_crtc->dma = dma_id;
@@ -738,7 +789,7 @@ struct drm_crtc *mdp4_crtc_init(struct drm_device *dev,
 
 	INIT_FENCE_CB(&mdp4_crtc->pageflip_cb, pageflip_cb);
 
-	drm_crtc_init(dev, crtc, &mdp4_crtc_funcs);
+	drm_crtc_init_with_planes(dev, crtc, plane, NULL, &mdp4_crtc_funcs);
 	drm_crtc_helper_add(crtc, &mdp4_crtc_helper_funcs);
 
 	mdp4_plane_install_properties(mdp4_crtc->plane, &crtc->base);
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_dtv_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
index 5e0dcae70ab..067ed03b35f 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_dtv_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
@@ -15,8 +15,6 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include <mach/clk.h>
-
 #include "mdp4_kms.h"
 
 #include "drm_crtc.h"
@@ -37,7 +35,7 @@ struct mdp4_dtv_encoder {
 static struct mdp4_kms *get_kms(struct drm_encoder *encoder)
 {
 	struct msm_drm_private *priv = encoder->dev->dev_private;
-	return to_mdp4_kms(priv->kms);
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
 }
 
 #ifdef CONFIG_MSM_BUS_SCALING
@@ -139,7 +137,7 @@ static void mdp4_dtv_encoder_dpms(struct drm_encoder *encoder, int mode)
 		 * the settings changes for the new modeset (like new
 		 * scanout buffer) don't latch properly..
 		 */
-		mdp4_irq_wait(mdp4_kms, MDP4_IRQ_EXTERNAL_VSYNC);
+		mdp_irq_wait(&mdp4_kms->base, MDP4_IRQ_EXTERNAL_VSYNC);
 
 		clk_disable_unprepare(mdp4_dtv_encoder->src_clk);
 		clk_disable_unprepare(mdp4_dtv_encoder->hdmi_clk);
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_irq.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_irq.c
new file mode 100644
index 00000000000..8edd531cb62
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_irq.c
@@ -0,0 +1,93 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+
+#include "msm_drv.h"
+#include "mdp4_kms.h"
+
+void mdp4_set_irqmask(struct mdp_kms *mdp_kms, uint32_t irqmask)
+{
+	mdp4_write(to_mdp4_kms(mdp_kms), REG_MDP4_INTR_ENABLE, irqmask);
+}
+
+static void mdp4_irq_error_handler(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	DRM_ERROR("errors: %08x\n", irqstatus);
+}
+
+void mdp4_irq_preinstall(struct msm_kms *kms)
+{
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
+	mdp4_write(mdp4_kms, REG_MDP4_INTR_CLEAR, 0xffffffff);
+}
+
+int mdp4_irq_postinstall(struct msm_kms *kms)
+{
+	struct mdp_kms *mdp_kms = to_mdp_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(mdp_kms);
+	struct mdp_irq *error_handler = &mdp4_kms->error_handler;
+
+	error_handler->irq = mdp4_irq_error_handler;
+	error_handler->irqmask = MDP4_IRQ_PRIMARY_INTF_UDERRUN |
+			MDP4_IRQ_EXTERNAL_INTF_UDERRUN;
+
+	mdp_irq_register(mdp_kms, error_handler);
+
+	return 0;
+}
+
+void mdp4_irq_uninstall(struct msm_kms *kms)
+{
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
+	mdp4_write(mdp4_kms, REG_MDP4_INTR_ENABLE, 0x00000000);
+}
+
+irqreturn_t mdp4_irq(struct msm_kms *kms)
+{
+	struct mdp_kms *mdp_kms = to_mdp_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(mdp_kms);
+	struct drm_device *dev = mdp4_kms->dev;
+	struct msm_drm_private *priv = dev->dev_private;
+	unsigned int id;
+	uint32_t status;
+
+	status = mdp4_read(mdp4_kms, REG_MDP4_INTR_STATUS);
+	mdp4_write(mdp4_kms, REG_MDP4_INTR_CLEAR, status);
+
+	VERB("status=%08x", status);
+
+	mdp_dispatch_irqs(mdp_kms, status);
+
+	for (id = 0; id < priv->num_crtcs; id++)
+		if (status & mdp4_crtc_vblank(priv->crtcs[id]))
+			drm_handle_vblank(dev, id);
+
+	return IRQ_HANDLED;
+}
+
+int mdp4_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
+{
+	mdp_update_vblank_mask(to_mdp_kms(kms),
+			mdp4_crtc_vblank(crtc), true);
+	return 0;
+}
+
+void mdp4_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
+{
+	mdp_update_vblank_mask(to_mdp_kms(kms),
+			mdp4_crtc_vblank(crtc), false);
+}
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_kms.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
index 8972ac35a43..0bb4faa1752 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_kms.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
@@ -17,13 +17,14 @@
 
 
 #include "msm_drv.h"
+#include "msm_mmu.h"
 #include "mdp4_kms.h"
 
 static struct mdp4_platform_config *mdp4_get_config(struct platform_device *dev);
 
 static int mdp4_hw_init(struct msm_kms *kms)
 {
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
 	struct drm_device *dev = mdp4_kms->dev;
 	uint32_t version, major, minor, dmap_cfg, vg_cfg;
 	unsigned long clk;
@@ -31,12 +32,14 @@ static int mdp4_hw_init(struct msm_kms *kms)
 
 	pm_runtime_get_sync(dev->dev);
 
+	mdp4_enable(mdp4_kms);
 	version = mdp4_read(mdp4_kms, REG_MDP4_VERSION);
+	mdp4_disable(mdp4_kms);
 
 	major = FIELD(version, MDP4_VERSION_MAJOR);
 	minor = FIELD(version, MDP4_VERSION_MINOR);
 
-	DBG("found MDP version v%d.%d", major, minor);
+	DBG("found MDP4 version v%d.%d", major, minor);
 
 	if (major != 4) {
 		dev_err(dev->dev, "unexpected MDP version: v%d.%d\n",
@@ -130,7 +133,7 @@ static long mdp4_round_pixclk(struct msm_kms *kms, unsigned long rate,
 
 static void mdp4_preclose(struct msm_kms *kms, struct drm_file *file)
 {
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
 	struct msm_drm_private *priv = mdp4_kms->dev->dev_private;
 	unsigned i;
 
@@ -140,11 +143,16 @@ static void mdp4_preclose(struct msm_kms *kms, struct drm_file *file)
 
 static void mdp4_destroy(struct msm_kms *kms)
 {
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
+	if (mdp4_kms->blank_cursor_iova)
+		msm_gem_put_iova(mdp4_kms->blank_cursor_bo, mdp4_kms->id);
+	if (mdp4_kms->blank_cursor_bo)
+		drm_gem_object_unreference(mdp4_kms->blank_cursor_bo);
 	kfree(mdp4_kms);
 }
 
-static const struct msm_kms_funcs kms_funcs = {
+static const struct mdp_kms_funcs kms_funcs = {
+	.base = {
 		.hw_init         = mdp4_hw_init,
 		.irq_preinstall  = mdp4_irq_preinstall,
 		.irq_postinstall = mdp4_irq_postinstall,
@@ -152,10 +160,12 @@ static const struct msm_kms_funcs kms_funcs = {
 		.irq             = mdp4_irq,
 		.enable_vblank   = mdp4_enable_vblank,
 		.disable_vblank  = mdp4_disable_vblank,
-		.get_format      = mdp4_get_format,
+		.get_format      = mdp_get_format,
 		.round_pixclk    = mdp4_round_pixclk,
 		.preclose        = mdp4_preclose,
 		.destroy         = mdp4_destroy,
+	},
+	.set_irqmask         = mdp4_set_irqmask,
 };
 
 int mdp4_disable(struct mdp4_kms *mdp4_kms)
@@ -189,6 +199,7 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 	struct drm_plane *plane;
 	struct drm_crtc *crtc;
 	struct drm_encoder *encoder;
+	struct hdmi *hdmi;
 	int ret;
 
 	/*
@@ -238,9 +249,10 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 	encoder->possible_crtcs = 0x1;     /* DTV can be hooked to DMA_E */
 	priv->encoders[priv->num_encoders++] = encoder;
 
-	ret = hdmi_init(dev, encoder);
-	if (ret) {
-		dev_err(dev->dev, "failed to initialize HDMI\n");
+	hdmi = hdmi_init(dev, encoder);
+	if (IS_ERR(hdmi)) {
+		ret = PTR_ERR(hdmi);
+		dev_err(dev->dev, "failed to initialize HDMI: %d\n", ret);
 		goto fail;
 	}
 
@@ -260,6 +272,7 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev)
 	struct mdp4_platform_config *config = mdp4_get_config(pdev);
 	struct mdp4_kms *mdp4_kms;
 	struct msm_kms *kms = NULL;
+	struct msm_mmu *mmu;
 	int ret;
 
 	mdp4_kms = kzalloc(sizeof(*mdp4_kms), GFP_KERNEL);
@@ -269,8 +282,9 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev)
 		goto fail;
 	}
 
-	kms = &mdp4_kms->base;
-	kms->funcs = &kms_funcs;
+	mdp_kms_init(&mdp4_kms->base, &kms_funcs);
+
+	kms = &mdp4_kms->base.base;
 
 	mdp4_kms->dev = dev;
 
@@ -322,27 +336,34 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev)
 	clk_set_rate(mdp4_kms->clk, config->max_clk);
 	clk_set_rate(mdp4_kms->lut_clk, config->max_clk);
 
-	if (!config->iommu) {
-		dev_err(dev->dev, "no iommu\n");
-		ret = -ENXIO;
-		goto fail;
-	}
-
 	/* make sure things are off before attaching iommu (bootloader could
 	 * have left things on, in which case we'll start getting faults if
 	 * we don't disable):
 	 */
+	mdp4_enable(mdp4_kms);
 	mdp4_write(mdp4_kms, REG_MDP4_DTV_ENABLE, 0);
 	mdp4_write(mdp4_kms, REG_MDP4_LCDC_ENABLE, 0);
 	mdp4_write(mdp4_kms, REG_MDP4_DSI_ENABLE, 0);
+	mdp4_disable(mdp4_kms);
 	mdelay(16);
 
-	ret = msm_iommu_attach(dev, config->iommu,
-			iommu_ports, ARRAY_SIZE(iommu_ports));
-	if (ret)
-		goto fail;
+	if (config->iommu) {
+		mmu = msm_iommu_new(dev, config->iommu);
+		if (IS_ERR(mmu)) {
+			ret = PTR_ERR(mmu);
+			goto fail;
+		}
+		ret = mmu->funcs->attach(mmu, iommu_ports,
+				ARRAY_SIZE(iommu_ports));
+		if (ret)
+			goto fail;
+	} else {
+		dev_info(dev->dev, "no iommu, fallback to phys "
+				"contig buffers for scanout\n");
+		mmu = NULL;
+	}
 
-	mdp4_kms->id = msm_register_iommu(dev, config->iommu);
+	mdp4_kms->id = msm_register_mmu(dev, mmu);
 	if (mdp4_kms->id < 0) {
 		ret = mdp4_kms->id;
 		dev_err(dev->dev, "failed to register mdp4 iommu: %d\n", ret);
@@ -355,6 +376,23 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev)
 		goto fail;
 	}
 
+	mutex_lock(&dev->struct_mutex);
+	mdp4_kms->blank_cursor_bo = msm_gem_new(dev, SZ_16K, MSM_BO_WC);
+	mutex_unlock(&dev->struct_mutex);
+	if (IS_ERR(mdp4_kms->blank_cursor_bo)) {
+		ret = PTR_ERR(mdp4_kms->blank_cursor_bo);
+		dev_err(dev->dev, "could not allocate blank-cursor bo: %d\n", ret);
+		mdp4_kms->blank_cursor_bo = NULL;
+		goto fail;
+	}
+
+	ret = msm_gem_get_iova(mdp4_kms->blank_cursor_bo, mdp4_kms->id,
+			&mdp4_kms->blank_cursor_iova);
+	if (ret) {
+		dev_err(dev->dev, "could not pin blank-cursor bo: %d\n", ret);
+		goto fail;
+	}
+
 	return kms;
 
 fail:
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_kms.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
index eb015c83408..715520c54cd 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_kms.h
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
@@ -18,29 +18,13 @@
 #ifndef __MDP4_KMS_H__
 #define __MDP4_KMS_H__
 
-#include <linux/clk.h>
-#include <linux/platform_device.h>
-#include <linux/regulator/consumer.h>
-
 #include "msm_drv.h"
+#include "msm_kms.h"
+#include "mdp/mdp_kms.h"
 #include "mdp4.xml.h"
 
-
-/* For transiently registering for different MDP4 irqs that various parts
- * of the KMS code need during setup/configuration.  We these are not
- * necessarily the same as what drm_vblank_get/put() are requesting, and
- * the hysteresis in drm_vblank_put() is not necessarily desirable for
- * internal housekeeping related irq usage.
- */
-struct mdp4_irq {
-	struct list_head node;
-	uint32_t irqmask;
-	bool registered;
-	void (*irq)(struct mdp4_irq *irq, uint32_t irqstatus);
-};
-
 struct mdp4_kms {
-	struct msm_kms base;
+	struct mdp_kms base;
 
 	struct drm_device *dev;
 
@@ -59,11 +43,11 @@ struct mdp4_kms {
 	struct clk *pclk;
 	struct clk *lut_clk;
 
-	/* irq handling: */
-	bool in_irq;
-	struct list_head irq_list;    /* list of mdp4_irq */
-	uint32_t vblank_mask;         /* irq bits set for userspace vblank */
-	struct mdp4_irq error_handler;
+	struct mdp_irq error_handler;
+
+	/* empty/blank cursor bo to use when cursor is "disabled" */
+	struct drm_gem_object *blank_cursor_bo;
+	uint32_t blank_cursor_iova;
 };
 #define to_mdp4_kms(x) container_of(x, struct mdp4_kms, base)
 
@@ -73,16 +57,6 @@ struct mdp4_platform_config {
 	uint32_t max_clk;
 };
 
-struct mdp4_format {
-	struct msm_format base;
-	enum mdp4_bpc bpc_r, bpc_g, bpc_b;
-	enum mdp4_bpc_alpha bpc_a;
-	uint8_t unpack[4];
-	bool alpha_enable, unpack_tight;
-	uint8_t cpp, unpack_count;
-};
-#define to_mdp4_format(x) container_of(x, struct mdp4_format, base)
-
 static inline void mdp4_write(struct mdp4_kms *mdp4_kms, u32 reg, u32 data)
 {
 	msm_writel(data, mdp4_kms->mmio + reg);
@@ -134,7 +108,7 @@ static inline uint32_t dma2err(enum mdp4_dma dma)
 }
 
 static inline uint32_t mixercfg(int mixer, enum mdp4_pipe pipe,
-		enum mdp4_mixer_stage_id stage)
+		enum mdp_mixer_stage_id stage)
 {
 	uint32_t mixer_cfg = 0;
 
@@ -178,19 +152,23 @@ static inline uint32_t mixercfg(int mixer, enum mdp4_pipe pipe,
 int mdp4_disable(struct mdp4_kms *mdp4_kms);
 int mdp4_enable(struct mdp4_kms *mdp4_kms);
 
+void mdp4_set_irqmask(struct mdp_kms *mdp_kms, uint32_t irqmask);
 void mdp4_irq_preinstall(struct msm_kms *kms);
 int mdp4_irq_postinstall(struct msm_kms *kms);
 void mdp4_irq_uninstall(struct msm_kms *kms);
 irqreturn_t mdp4_irq(struct msm_kms *kms);
-void mdp4_irq_wait(struct mdp4_kms *mdp4_kms, uint32_t irqmask);
-void mdp4_irq_register(struct mdp4_kms *mdp4_kms, struct mdp4_irq *irq);
-void mdp4_irq_unregister(struct mdp4_kms *mdp4_kms, struct mdp4_irq *irq);
 int mdp4_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
 void mdp4_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
 
-uint32_t mdp4_get_formats(enum mdp4_pipe pipe_id, uint32_t *formats,
-		uint32_t max_formats);
-const struct msm_format *mdp4_get_format(struct msm_kms *kms, uint32_t format);
+static inline
+uint32_t mdp4_get_formats(enum mdp4_pipe pipe_id, uint32_t *pixel_formats,
+		uint32_t max_formats)
+{
+	/* TODO when we have YUV, we need to filter supported formats
+	 * based on pipe_id..
+	 */
+	return mdp_get_formats(pixel_formats, max_formats);
+}
 
 void mdp4_plane_install_properties(struct drm_plane *plane,
 		struct drm_mode_object *obj);
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_plane.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c
index 0f0af243f6f..66f33dba1eb 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_plane.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c
@@ -34,7 +34,7 @@ struct mdp4_plane {
 static struct mdp4_kms *get_kms(struct drm_plane *plane)
 {
 	struct msm_drm_private *priv = plane->dev->dev_private;
-	return to_mdp4_kms(priv->kms);
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
 }
 
 static int mdp4_plane_update(struct drm_plane *plane,
@@ -132,7 +132,7 @@ int mdp4_plane_mode_set(struct drm_plane *plane,
 	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
 	struct mdp4_kms *mdp4_kms = get_kms(plane);
 	enum mdp4_pipe pipe = mdp4_plane->pipe;
-	const struct mdp4_format *format;
+	const struct mdp_format *format;
 	uint32_t op_mode = 0;
 	uint32_t phasex_step = MDP4_VG_PHASE_STEP_DEFAULT;
 	uint32_t phasey_step = MDP4_VG_PHASE_STEP_DEFAULT;
@@ -170,12 +170,12 @@ int mdp4_plane_mode_set(struct drm_plane *plane,
 			MDP4_PIPE_DST_SIZE_HEIGHT(crtc_h));
 
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_DST_XY(pipe),
-			MDP4_PIPE_SRC_XY_X(crtc_x) |
-			MDP4_PIPE_SRC_XY_Y(crtc_y));
+			MDP4_PIPE_DST_XY_X(crtc_x) |
+			MDP4_PIPE_DST_XY_Y(crtc_y));
 
 	mdp4_plane_set_scanout(plane, fb);
 
-	format = to_mdp4_format(msm_framebuffer_format(fb));
+	format = to_mdp_format(msm_framebuffer_format(fb));
 
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_SRC_FORMAT(pipe),
 			MDP4_PIPE_SRC_FORMAT_A_BPC(format->bpc_a) |
@@ -222,6 +222,7 @@ struct drm_plane *mdp4_plane_init(struct drm_device *dev,
 	struct drm_plane *plane = NULL;
 	struct mdp4_plane *mdp4_plane;
 	int ret;
+	enum drm_plane_type type;
 
 	mdp4_plane = kzalloc(sizeof(*mdp4_plane), GFP_KERNEL);
 	if (!mdp4_plane) {
@@ -237,9 +238,10 @@ struct drm_plane *mdp4_plane_init(struct drm_device *dev,
 	mdp4_plane->nformats = mdp4_get_formats(pipe_id, mdp4_plane->formats,
 			ARRAY_SIZE(mdp4_plane->formats));
 
-	drm_plane_init(dev, plane, 0xff, &mdp4_plane_funcs,
-			mdp4_plane->formats, mdp4_plane->nformats,
-			private_plane);
+	type = private_plane ? DRM_PLANE_TYPE_PRIMARY : DRM_PLANE_TYPE_OVERLAY;
+	drm_universal_plane_init(dev, plane, 0xff, &mdp4_plane_funcs,
+				 mdp4_plane->formats, mdp4_plane->nformats,
+				 type);
 
 	mdp4_plane_install_properties(plane, &plane->base);
 
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h
new file mode 100644
index 00000000000..0aa51517f82
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h
@@ -0,0 +1,1036 @@
+#ifndef MDP5_XML
+#define MDP5_XML
+
+/* Autogenerated file, DO NOT EDIT manually!
+
+This file was generated by the rules-ng-ng headergen tool in this git repository:
+http://github.com/freedreno/envytools/
+git clone https://github.com/freedreno/envytools.git
+
+The rules-ng-ng source files this header was generated from are:
+- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
+- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2013-12-03 20:59:13)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
+
+Copyright (C) 2013 by the following authors:
+- Rob Clark <robdclark@gmail.com> (robclark)
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice (including the
+next paragraph) shall be included in all copies or substantial
+portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+IN NO EVENT SHALL THE COPYRIGHT OWNER(S) AND/OR ITS SUPPLIERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+*/
+
+
+enum mdp5_intf {
+	INTF_DSI = 1,
+	INTF_HDMI = 3,
+	INTF_LCDC = 5,
+	INTF_eDP = 9,
+};
+
+enum mdp5_intfnum {
+	NO_INTF = 0,
+	INTF0 = 1,
+	INTF1 = 2,
+	INTF2 = 3,
+	INTF3 = 4,
+};
+
+enum mdp5_pipe {
+	SSPP_VIG0 = 0,
+	SSPP_VIG1 = 1,
+	SSPP_VIG2 = 2,
+	SSPP_RGB0 = 3,
+	SSPP_RGB1 = 4,
+	SSPP_RGB2 = 5,
+	SSPP_DMA0 = 6,
+	SSPP_DMA1 = 7,
+};
+
+enum mdp5_ctl_mode {
+	MODE_NONE = 0,
+	MODE_ROT0 = 1,
+	MODE_ROT1 = 2,
+	MODE_WB0 = 3,
+	MODE_WB1 = 4,
+	MODE_WFD = 5,
+};
+
+enum mdp5_pack_3d {
+	PACK_3D_FRAME_INT = 0,
+	PACK_3D_H_ROW_INT = 1,
+	PACK_3D_V_ROW_INT = 2,
+	PACK_3D_COL_INT = 3,
+};
+
+enum mdp5_chroma_samp_type {
+	CHROMA_RGB = 0,
+	CHROMA_H2V1 = 1,
+	CHROMA_H1V2 = 2,
+	CHROMA_420 = 3,
+};
+
+enum mdp5_scale_filter {
+	SCALE_FILTER_NEAREST = 0,
+	SCALE_FILTER_BIL = 1,
+	SCALE_FILTER_PCMN = 2,
+	SCALE_FILTER_CA = 3,
+};
+
+enum mdp5_pipe_bwc {
+	BWC_LOSSLESS = 0,
+	BWC_Q_HIGH = 1,
+	BWC_Q_MED = 2,
+};
+
+enum mdp5_client_id {
+	CID_UNUSED = 0,
+	CID_VIG0_Y = 1,
+	CID_VIG0_CR = 2,
+	CID_VIG0_CB = 3,
+	CID_VIG1_Y = 4,
+	CID_VIG1_CR = 5,
+	CID_VIG1_CB = 6,
+	CID_VIG2_Y = 7,
+	CID_VIG2_CR = 8,
+	CID_VIG2_CB = 9,
+	CID_DMA0_Y = 10,
+	CID_DMA0_CR = 11,
+	CID_DMA0_CB = 12,
+	CID_DMA1_Y = 13,
+	CID_DMA1_CR = 14,
+	CID_DMA1_CB = 15,
+	CID_RGB0 = 16,
+	CID_RGB1 = 17,
+	CID_RGB2 = 18,
+	CID_MAX = 19,
+};
+
+enum mdp5_igc_type {
+	IGC_VIG = 0,
+	IGC_RGB = 1,
+	IGC_DMA = 2,
+	IGC_DSPP = 3,
+};
+
+#define MDP5_IRQ_INTF0_WB_ROT_COMP				0x00000001
+#define MDP5_IRQ_INTF1_WB_ROT_COMP				0x00000002
+#define MDP5_IRQ_INTF2_WB_ROT_COMP				0x00000004
+#define MDP5_IRQ_INTF3_WB_ROT_COMP				0x00000008
+#define MDP5_IRQ_INTF0_WB_WFD					0x00000010
+#define MDP5_IRQ_INTF1_WB_WFD					0x00000020
+#define MDP5_IRQ_INTF2_WB_WFD					0x00000040
+#define MDP5_IRQ_INTF3_WB_WFD					0x00000080
+#define MDP5_IRQ_INTF0_PING_PONG_COMP				0x00000100
+#define MDP5_IRQ_INTF1_PING_PONG_COMP				0x00000200
+#define MDP5_IRQ_INTF2_PING_PONG_COMP				0x00000400
+#define MDP5_IRQ_INTF3_PING_PONG_COMP				0x00000800
+#define MDP5_IRQ_INTF0_PING_PONG_RD_PTR				0x00001000
+#define MDP5_IRQ_INTF1_PING_PONG_RD_PTR				0x00002000
+#define MDP5_IRQ_INTF2_PING_PONG_RD_PTR				0x00004000
+#define MDP5_IRQ_INTF3_PING_PONG_RD_PTR				0x00008000
+#define MDP5_IRQ_INTF0_PING_PONG_WR_PTR				0x00010000
+#define MDP5_IRQ_INTF1_PING_PONG_WR_PTR				0x00020000
+#define MDP5_IRQ_INTF2_PING_PONG_WR_PTR				0x00040000
+#define MDP5_IRQ_INTF3_PING_PONG_WR_PTR				0x00080000
+#define MDP5_IRQ_INTF0_PING_PONG_AUTO_REF			0x00100000
+#define MDP5_IRQ_INTF1_PING_PONG_AUTO_REF			0x00200000
+#define MDP5_IRQ_INTF2_PING_PONG_AUTO_REF			0x00400000
+#define MDP5_IRQ_INTF3_PING_PONG_AUTO_REF			0x00800000
+#define MDP5_IRQ_INTF0_UNDER_RUN				0x01000000
+#define MDP5_IRQ_INTF0_VSYNC					0x02000000
+#define MDP5_IRQ_INTF1_UNDER_RUN				0x04000000
+#define MDP5_IRQ_INTF1_VSYNC					0x08000000
+#define MDP5_IRQ_INTF2_UNDER_RUN				0x10000000
+#define MDP5_IRQ_INTF2_VSYNC					0x20000000
+#define MDP5_IRQ_INTF3_UNDER_RUN				0x40000000
+#define MDP5_IRQ_INTF3_VSYNC					0x80000000
+#define REG_MDP5_HW_VERSION					0x00000000
+
+#define REG_MDP5_HW_INTR_STATUS					0x00000010
+#define MDP5_HW_INTR_STATUS_INTR_MDP				0x00000001
+#define MDP5_HW_INTR_STATUS_INTR_DSI0				0x00000010
+#define MDP5_HW_INTR_STATUS_INTR_DSI1				0x00000020
+#define MDP5_HW_INTR_STATUS_INTR_HDMI				0x00000100
+#define MDP5_HW_INTR_STATUS_INTR_EDP				0x00001000
+
+#define REG_MDP5_MDP_VERSION					0x00000100
+#define MDP5_MDP_VERSION_MINOR__MASK				0x00ff0000
+#define MDP5_MDP_VERSION_MINOR__SHIFT				16
+static inline uint32_t MDP5_MDP_VERSION_MINOR(uint32_t val)
+{
+	return ((val) << MDP5_MDP_VERSION_MINOR__SHIFT) & MDP5_MDP_VERSION_MINOR__MASK;
+}
+#define MDP5_MDP_VERSION_MAJOR__MASK				0xf0000000
+#define MDP5_MDP_VERSION_MAJOR__SHIFT				28
+static inline uint32_t MDP5_MDP_VERSION_MAJOR(uint32_t val)
+{
+	return ((val) << MDP5_MDP_VERSION_MAJOR__SHIFT) & MDP5_MDP_VERSION_MAJOR__MASK;
+}
+
+#define REG_MDP5_DISP_INTF_SEL					0x00000104
+#define MDP5_DISP_INTF_SEL_INTF0__MASK				0x000000ff
+#define MDP5_DISP_INTF_SEL_INTF0__SHIFT				0
+static inline uint32_t MDP5_DISP_INTF_SEL_INTF0(enum mdp5_intf val)
+{
+	return ((val) << MDP5_DISP_INTF_SEL_INTF0__SHIFT) & MDP5_DISP_INTF_SEL_INTF0__MASK;
+}
+#define MDP5_DISP_INTF_SEL_INTF1__MASK				0x0000ff00
+#define MDP5_DISP_INTF_SEL_INTF1__SHIFT				8
+static inline uint32_t MDP5_DISP_INTF_SEL_INTF1(enum mdp5_intf val)
+{
+	return ((val) << MDP5_DISP_INTF_SEL_INTF1__SHIFT) & MDP5_DISP_INTF_SEL_INTF1__MASK;
+}
+#define MDP5_DISP_INTF_SEL_INTF2__MASK				0x00ff0000
+#define MDP5_DISP_INTF_SEL_INTF2__SHIFT				16
+static inline uint32_t MDP5_DISP_INTF_SEL_INTF2(enum mdp5_intf val)
+{
+	return ((val) << MDP5_DISP_INTF_SEL_INTF2__SHIFT) & MDP5_DISP_INTF_SEL_INTF2__MASK;
+}
+#define MDP5_DISP_INTF_SEL_INTF3__MASK				0xff000000
+#define MDP5_DISP_INTF_SEL_INTF3__SHIFT				24
+static inline uint32_t MDP5_DISP_INTF_SEL_INTF3(enum mdp5_intf val)
+{
+	return ((val) << MDP5_DISP_INTF_SEL_INTF3__SHIFT) & MDP5_DISP_INTF_SEL_INTF3__MASK;
+}
+
+#define REG_MDP5_INTR_EN					0x00000110
+
+#define REG_MDP5_INTR_STATUS					0x00000114
+
+#define REG_MDP5_INTR_CLEAR					0x00000118
+
+#define REG_MDP5_HIST_INTR_EN					0x0000011c
+
+#define REG_MDP5_HIST_INTR_STATUS				0x00000120
+
+#define REG_MDP5_HIST_INTR_CLEAR				0x00000124
+
+static inline uint32_t REG_MDP5_SMP_ALLOC_W(uint32_t i0) { return 0x00000180 + 0x4*i0; }
+
+static inline uint32_t REG_MDP5_SMP_ALLOC_W_REG(uint32_t i0) { return 0x00000180 + 0x4*i0; }
+#define MDP5_SMP_ALLOC_W_REG_CLIENT0__MASK			0x000000ff
+#define MDP5_SMP_ALLOC_W_REG_CLIENT0__SHIFT			0
+static inline uint32_t MDP5_SMP_ALLOC_W_REG_CLIENT0(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_W_REG_CLIENT0__SHIFT) & MDP5_SMP_ALLOC_W_REG_CLIENT0__MASK;
+}
+#define MDP5_SMP_ALLOC_W_REG_CLIENT1__MASK			0x0000ff00
+#define MDP5_SMP_ALLOC_W_REG_CLIENT1__SHIFT			8
+static inline uint32_t MDP5_SMP_ALLOC_W_REG_CLIENT1(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_W_REG_CLIENT1__SHIFT) & MDP5_SMP_ALLOC_W_REG_CLIENT1__MASK;
+}
+#define MDP5_SMP_ALLOC_W_REG_CLIENT2__MASK			0x00ff0000
+#define MDP5_SMP_ALLOC_W_REG_CLIENT2__SHIFT			16
+static inline uint32_t MDP5_SMP_ALLOC_W_REG_CLIENT2(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_W_REG_CLIENT2__SHIFT) & MDP5_SMP_ALLOC_W_REG_CLIENT2__MASK;
+}
+
+static inline uint32_t REG_MDP5_SMP_ALLOC_R(uint32_t i0) { return 0x00000230 + 0x4*i0; }
+
+static inline uint32_t REG_MDP5_SMP_ALLOC_R_REG(uint32_t i0) { return 0x00000230 + 0x4*i0; }
+#define MDP5_SMP_ALLOC_R_REG_CLIENT0__MASK			0x000000ff
+#define MDP5_SMP_ALLOC_R_REG_CLIENT0__SHIFT			0
+static inline uint32_t MDP5_SMP_ALLOC_R_REG_CLIENT0(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_R_REG_CLIENT0__SHIFT) & MDP5_SMP_ALLOC_R_REG_CLIENT0__MASK;
+}
+#define MDP5_SMP_ALLOC_R_REG_CLIENT1__MASK			0x0000ff00
+#define MDP5_SMP_ALLOC_R_REG_CLIENT1__SHIFT			8
+static inline uint32_t MDP5_SMP_ALLOC_R_REG_CLIENT1(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_R_REG_CLIENT1__SHIFT) & MDP5_SMP_ALLOC_R_REG_CLIENT1__MASK;
+}
+#define MDP5_SMP_ALLOC_R_REG_CLIENT2__MASK			0x00ff0000
+#define MDP5_SMP_ALLOC_R_REG_CLIENT2__SHIFT			16
+static inline uint32_t MDP5_SMP_ALLOC_R_REG_CLIENT2(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_R_REG_CLIENT2__SHIFT) & MDP5_SMP_ALLOC_R_REG_CLIENT2__MASK;
+}
+
+static inline uint32_t __offset_IGC(enum mdp5_igc_type idx)
+{
+	switch (idx) {
+		case IGC_VIG: return 0x00000300;
+		case IGC_RGB: return 0x00000310;
+		case IGC_DMA: return 0x00000320;
+		case IGC_DSPP: return 0x00000400;
+		default: return INVALID_IDX(idx);
+	}
+}
+static inline uint32_t REG_MDP5_IGC(enum mdp5_igc_type i0) { return 0x00000000 + __offset_IGC(i0); }
+
+static inline uint32_t REG_MDP5_IGC_LUT(enum mdp5_igc_type i0, uint32_t i1) { return 0x00000000 + __offset_IGC(i0) + 0x4*i1; }
+
+static inline uint32_t REG_MDP5_IGC_LUT_REG(enum mdp5_igc_type i0, uint32_t i1) { return 0x00000000 + __offset_IGC(i0) + 0x4*i1; }
+#define MDP5_IGC_LUT_REG_VAL__MASK				0x00000fff
+#define MDP5_IGC_LUT_REG_VAL__SHIFT				0
+static inline uint32_t MDP5_IGC_LUT_REG_VAL(uint32_t val)
+{
+	return ((val) << MDP5_IGC_LUT_REG_VAL__SHIFT) & MDP5_IGC_LUT_REG_VAL__MASK;
+}
+#define MDP5_IGC_LUT_REG_INDEX_UPDATE				0x02000000
+#define MDP5_IGC_LUT_REG_DISABLE_PIPE_0				0x10000000
+#define MDP5_IGC_LUT_REG_DISABLE_PIPE_1				0x20000000
+#define MDP5_IGC_LUT_REG_DISABLE_PIPE_2				0x40000000
+
+static inline uint32_t REG_MDP5_CTL(uint32_t i0) { return 0x00000600 + 0x100*i0; }
+
+static inline uint32_t REG_MDP5_CTL_LAYER(uint32_t i0, uint32_t i1) { return 0x00000600 + 0x100*i0 + 0x4*i1; }
+
+static inline uint32_t REG_MDP5_CTL_LAYER_REG(uint32_t i0, uint32_t i1) { return 0x00000600 + 0x100*i0 + 0x4*i1; }
+#define MDP5_CTL_LAYER_REG_VIG0__MASK				0x00000007
+#define MDP5_CTL_LAYER_REG_VIG0__SHIFT				0
+static inline uint32_t MDP5_CTL_LAYER_REG_VIG0(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_VIG0__SHIFT) & MDP5_CTL_LAYER_REG_VIG0__MASK;
+}
+#define MDP5_CTL_LAYER_REG_VIG1__MASK				0x00000038
+#define MDP5_CTL_LAYER_REG_VIG1__SHIFT				3
+static inline uint32_t MDP5_CTL_LAYER_REG_VIG1(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_VIG1__SHIFT) & MDP5_CTL_LAYER_REG_VIG1__MASK;
+}
+#define MDP5_CTL_LAYER_REG_VIG2__MASK				0x000001c0
+#define MDP5_CTL_LAYER_REG_VIG2__SHIFT				6
+static inline uint32_t MDP5_CTL_LAYER_REG_VIG2(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_VIG2__SHIFT) & MDP5_CTL_LAYER_REG_VIG2__MASK;
+}
+#define MDP5_CTL_LAYER_REG_RGB0__MASK				0x00000e00
+#define MDP5_CTL_LAYER_REG_RGB0__SHIFT				9
+static inline uint32_t MDP5_CTL_LAYER_REG_RGB0(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_RGB0__SHIFT) & MDP5_CTL_LAYER_REG_RGB0__MASK;
+}
+#define MDP5_CTL_LAYER_REG_RGB1__MASK				0x00007000
+#define MDP5_CTL_LAYER_REG_RGB1__SHIFT				12
+static inline uint32_t MDP5_CTL_LAYER_REG_RGB1(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_RGB1__SHIFT) & MDP5_CTL_LAYER_REG_RGB1__MASK;
+}
+#define MDP5_CTL_LAYER_REG_RGB2__MASK				0x00038000
+#define MDP5_CTL_LAYER_REG_RGB2__SHIFT				15
+static inline uint32_t MDP5_CTL_LAYER_REG_RGB2(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_RGB2__SHIFT) & MDP5_CTL_LAYER_REG_RGB2__MASK;
+}
+#define MDP5_CTL_LAYER_REG_DMA0__MASK				0x001c0000
+#define MDP5_CTL_LAYER_REG_DMA0__SHIFT				18
+static inline uint32_t MDP5_CTL_LAYER_REG_DMA0(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_DMA0__SHIFT) & MDP5_CTL_LAYER_REG_DMA0__MASK;
+}
+#define MDP5_CTL_LAYER_REG_DMA1__MASK				0x00e00000
+#define MDP5_CTL_LAYER_REG_DMA1__SHIFT				21
+static inline uint32_t MDP5_CTL_LAYER_REG_DMA1(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_DMA1__SHIFT) & MDP5_CTL_LAYER_REG_DMA1__MASK;
+}
+#define MDP5_CTL_LAYER_REG_BORDER_COLOR				0x01000000
+#define MDP5_CTL_LAYER_REG_CURSOR_OUT				0x02000000
+
+static inline uint32_t REG_MDP5_CTL_OP(uint32_t i0) { return 0x00000614 + 0x100*i0; }
+#define MDP5_CTL_OP_MODE__MASK					0x0000000f
+#define MDP5_CTL_OP_MODE__SHIFT					0
+static inline uint32_t MDP5_CTL_OP_MODE(enum mdp5_ctl_mode val)
+{
+	return ((val) << MDP5_CTL_OP_MODE__SHIFT) & MDP5_CTL_OP_MODE__MASK;
+}
+#define MDP5_CTL_OP_INTF_NUM__MASK				0x00000070
+#define MDP5_CTL_OP_INTF_NUM__SHIFT				4
+static inline uint32_t MDP5_CTL_OP_INTF_NUM(enum mdp5_intfnum val)
+{
+	return ((val) << MDP5_CTL_OP_INTF_NUM__SHIFT) & MDP5_CTL_OP_INTF_NUM__MASK;
+}
+#define MDP5_CTL_OP_CMD_MODE					0x00020000
+#define MDP5_CTL_OP_PACK_3D_ENABLE				0x00080000
+#define MDP5_CTL_OP_PACK_3D__MASK				0x00300000
+#define MDP5_CTL_OP_PACK_3D__SHIFT				20
+static inline uint32_t MDP5_CTL_OP_PACK_3D(enum mdp5_pack_3d val)
+{
+	return ((val) << MDP5_CTL_OP_PACK_3D__SHIFT) & MDP5_CTL_OP_PACK_3D__MASK;
+}
+
+static inline uint32_t REG_MDP5_CTL_FLUSH(uint32_t i0) { return 0x00000618 + 0x100*i0; }
+#define MDP5_CTL_FLUSH_VIG0					0x00000001
+#define MDP5_CTL_FLUSH_VIG1					0x00000002
+#define MDP5_CTL_FLUSH_VIG2					0x00000004
+#define MDP5_CTL_FLUSH_RGB0					0x00000008
+#define MDP5_CTL_FLUSH_RGB1					0x00000010
+#define MDP5_CTL_FLUSH_RGB2					0x00000020
+#define MDP5_CTL_FLUSH_LM0					0x00000040
+#define MDP5_CTL_FLUSH_LM1					0x00000080
+#define MDP5_CTL_FLUSH_LM2					0x00000100
+#define MDP5_CTL_FLUSH_DMA0					0x00000800
+#define MDP5_CTL_FLUSH_DMA1					0x00001000
+#define MDP5_CTL_FLUSH_DSPP0					0x00002000
+#define MDP5_CTL_FLUSH_DSPP1					0x00004000
+#define MDP5_CTL_FLUSH_DSPP2					0x00008000
+#define MDP5_CTL_FLUSH_CTL					0x00020000
+
+static inline uint32_t REG_MDP5_CTL_START(uint32_t i0) { return 0x0000061c + 0x100*i0; }
+
+static inline uint32_t REG_MDP5_CTL_PACK_3D(uint32_t i0) { return 0x00000620 + 0x100*i0; }
+
+static inline uint32_t REG_MDP5_PIPE(enum mdp5_pipe i0) { return 0x00001200 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_HIST_CTL_BASE(enum mdp5_pipe i0) { return 0x000014c4 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_HIST_LUT_BASE(enum mdp5_pipe i0) { return 0x000014f0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_HIST_LUT_SWAP(enum mdp5_pipe i0) { return 0x00001500 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC_SIZE(enum mdp5_pipe i0) { return 0x00001200 + 0x400*i0; }
+#define MDP5_PIPE_SRC_SIZE_HEIGHT__MASK				0xffff0000
+#define MDP5_PIPE_SRC_SIZE_HEIGHT__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_SIZE_HEIGHT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_SIZE_HEIGHT__SHIFT) & MDP5_PIPE_SRC_SIZE_HEIGHT__MASK;
+}
+#define MDP5_PIPE_SRC_SIZE_WIDTH__MASK				0x0000ffff
+#define MDP5_PIPE_SRC_SIZE_WIDTH__SHIFT				0
+static inline uint32_t MDP5_PIPE_SRC_SIZE_WIDTH(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_SIZE_WIDTH__SHIFT) & MDP5_PIPE_SRC_SIZE_WIDTH__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_IMG_SIZE(enum mdp5_pipe i0) { return 0x00001204 + 0x400*i0; }
+#define MDP5_PIPE_SRC_IMG_SIZE_HEIGHT__MASK			0xffff0000
+#define MDP5_PIPE_SRC_IMG_SIZE_HEIGHT__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_IMG_SIZE_HEIGHT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_IMG_SIZE_HEIGHT__SHIFT) & MDP5_PIPE_SRC_IMG_SIZE_HEIGHT__MASK;
+}
+#define MDP5_PIPE_SRC_IMG_SIZE_WIDTH__MASK			0x0000ffff
+#define MDP5_PIPE_SRC_IMG_SIZE_WIDTH__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_IMG_SIZE_WIDTH(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_IMG_SIZE_WIDTH__SHIFT) & MDP5_PIPE_SRC_IMG_SIZE_WIDTH__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_XY(enum mdp5_pipe i0) { return 0x00001208 + 0x400*i0; }
+#define MDP5_PIPE_SRC_XY_Y__MASK				0xffff0000
+#define MDP5_PIPE_SRC_XY_Y__SHIFT				16
+static inline uint32_t MDP5_PIPE_SRC_XY_Y(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_XY_Y__SHIFT) & MDP5_PIPE_SRC_XY_Y__MASK;
+}
+#define MDP5_PIPE_SRC_XY_X__MASK				0x0000ffff
+#define MDP5_PIPE_SRC_XY_X__SHIFT				0
+static inline uint32_t MDP5_PIPE_SRC_XY_X(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_XY_X__SHIFT) & MDP5_PIPE_SRC_XY_X__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_OUT_SIZE(enum mdp5_pipe i0) { return 0x0000120c + 0x400*i0; }
+#define MDP5_PIPE_OUT_SIZE_HEIGHT__MASK				0xffff0000
+#define MDP5_PIPE_OUT_SIZE_HEIGHT__SHIFT			16
+static inline uint32_t MDP5_PIPE_OUT_SIZE_HEIGHT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_OUT_SIZE_HEIGHT__SHIFT) & MDP5_PIPE_OUT_SIZE_HEIGHT__MASK;
+}
+#define MDP5_PIPE_OUT_SIZE_WIDTH__MASK				0x0000ffff
+#define MDP5_PIPE_OUT_SIZE_WIDTH__SHIFT				0
+static inline uint32_t MDP5_PIPE_OUT_SIZE_WIDTH(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_OUT_SIZE_WIDTH__SHIFT) & MDP5_PIPE_OUT_SIZE_WIDTH__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_OUT_XY(enum mdp5_pipe i0) { return 0x00001210 + 0x400*i0; }
+#define MDP5_PIPE_OUT_XY_Y__MASK				0xffff0000
+#define MDP5_PIPE_OUT_XY_Y__SHIFT				16
+static inline uint32_t MDP5_PIPE_OUT_XY_Y(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_OUT_XY_Y__SHIFT) & MDP5_PIPE_OUT_XY_Y__MASK;
+}
+#define MDP5_PIPE_OUT_XY_X__MASK				0x0000ffff
+#define MDP5_PIPE_OUT_XY_X__SHIFT				0
+static inline uint32_t MDP5_PIPE_OUT_XY_X(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_OUT_XY_X__SHIFT) & MDP5_PIPE_OUT_XY_X__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC0_ADDR(enum mdp5_pipe i0) { return 0x00001214 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC1_ADDR(enum mdp5_pipe i0) { return 0x00001218 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC2_ADDR(enum mdp5_pipe i0) { return 0x0000121c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC3_ADDR(enum mdp5_pipe i0) { return 0x00001220 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC_STRIDE_A(enum mdp5_pipe i0) { return 0x00001224 + 0x400*i0; }
+#define MDP5_PIPE_SRC_STRIDE_A_P0__MASK				0x0000ffff
+#define MDP5_PIPE_SRC_STRIDE_A_P0__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_STRIDE_A_P0(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_STRIDE_A_P0__SHIFT) & MDP5_PIPE_SRC_STRIDE_A_P0__MASK;
+}
+#define MDP5_PIPE_SRC_STRIDE_A_P1__MASK				0xffff0000
+#define MDP5_PIPE_SRC_STRIDE_A_P1__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_STRIDE_A_P1(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_STRIDE_A_P1__SHIFT) & MDP5_PIPE_SRC_STRIDE_A_P1__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_STRIDE_B(enum mdp5_pipe i0) { return 0x00001228 + 0x400*i0; }
+#define MDP5_PIPE_SRC_STRIDE_B_P2__MASK				0x0000ffff
+#define MDP5_PIPE_SRC_STRIDE_B_P2__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_STRIDE_B_P2(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_STRIDE_B_P2__SHIFT) & MDP5_PIPE_SRC_STRIDE_B_P2__MASK;
+}
+#define MDP5_PIPE_SRC_STRIDE_B_P3__MASK				0xffff0000
+#define MDP5_PIPE_SRC_STRIDE_B_P3__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_STRIDE_B_P3(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_STRIDE_B_P3__SHIFT) & MDP5_PIPE_SRC_STRIDE_B_P3__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_STILE_FRAME_SIZE(enum mdp5_pipe i0) { return 0x0000122c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC_FORMAT(enum mdp5_pipe i0) { return 0x00001230 + 0x400*i0; }
+#define MDP5_PIPE_SRC_FORMAT_G_BPC__MASK			0x00000003
+#define MDP5_PIPE_SRC_FORMAT_G_BPC__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_G_BPC(enum mdp_bpc val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_G_BPC__SHIFT) & MDP5_PIPE_SRC_FORMAT_G_BPC__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_B_BPC__MASK			0x0000000c
+#define MDP5_PIPE_SRC_FORMAT_B_BPC__SHIFT			2
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_B_BPC(enum mdp_bpc val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_B_BPC__SHIFT) & MDP5_PIPE_SRC_FORMAT_B_BPC__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_R_BPC__MASK			0x00000030
+#define MDP5_PIPE_SRC_FORMAT_R_BPC__SHIFT			4
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_R_BPC(enum mdp_bpc val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_R_BPC__SHIFT) & MDP5_PIPE_SRC_FORMAT_R_BPC__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_A_BPC__MASK			0x000000c0
+#define MDP5_PIPE_SRC_FORMAT_A_BPC__SHIFT			6
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_A_BPC(enum mdp_bpc_alpha val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_A_BPC__SHIFT) & MDP5_PIPE_SRC_FORMAT_A_BPC__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_ALPHA_ENABLE			0x00000100
+#define MDP5_PIPE_SRC_FORMAT_CPP__MASK				0x00000600
+#define MDP5_PIPE_SRC_FORMAT_CPP__SHIFT				9
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_CPP(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_CPP__SHIFT) & MDP5_PIPE_SRC_FORMAT_CPP__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_ROT90				0x00000800
+#define MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT__MASK			0x00003000
+#define MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT__SHIFT		12
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT__SHIFT) & MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_UNPACK_TIGHT			0x00020000
+#define MDP5_PIPE_SRC_FORMAT_UNPACK_ALIGN_MSB			0x00040000
+#define MDP5_PIPE_SRC_FORMAT_NUM_PLANES__MASK			0x00780000
+#define MDP5_PIPE_SRC_FORMAT_NUM_PLANES__SHIFT			19
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_NUM_PLANES(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_NUM_PLANES__SHIFT) & MDP5_PIPE_SRC_FORMAT_NUM_PLANES__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP__MASK			0x01800000
+#define MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP__SHIFT			23
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP(enum mdp5_chroma_samp_type val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP__SHIFT) & MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_UNPACK(enum mdp5_pipe i0) { return 0x00001234 + 0x400*i0; }
+#define MDP5_PIPE_SRC_UNPACK_ELEM0__MASK			0x000000ff
+#define MDP5_PIPE_SRC_UNPACK_ELEM0__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_UNPACK_ELEM0(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_UNPACK_ELEM0__SHIFT) & MDP5_PIPE_SRC_UNPACK_ELEM0__MASK;
+}
+#define MDP5_PIPE_SRC_UNPACK_ELEM1__MASK			0x0000ff00
+#define MDP5_PIPE_SRC_UNPACK_ELEM1__SHIFT			8
+static inline uint32_t MDP5_PIPE_SRC_UNPACK_ELEM1(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_UNPACK_ELEM1__SHIFT) & MDP5_PIPE_SRC_UNPACK_ELEM1__MASK;
+}
+#define MDP5_PIPE_SRC_UNPACK_ELEM2__MASK			0x00ff0000
+#define MDP5_PIPE_SRC_UNPACK_ELEM2__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_UNPACK_ELEM2(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_UNPACK_ELEM2__SHIFT) & MDP5_PIPE_SRC_UNPACK_ELEM2__MASK;
+}
+#define MDP5_PIPE_SRC_UNPACK_ELEM3__MASK			0xff000000
+#define MDP5_PIPE_SRC_UNPACK_ELEM3__SHIFT			24
+static inline uint32_t MDP5_PIPE_SRC_UNPACK_ELEM3(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_UNPACK_ELEM3__SHIFT) & MDP5_PIPE_SRC_UNPACK_ELEM3__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_OP_MODE(enum mdp5_pipe i0) { return 0x00001238 + 0x400*i0; }
+#define MDP5_PIPE_SRC_OP_MODE_BWC_EN				0x00000001
+#define MDP5_PIPE_SRC_OP_MODE_BWC__MASK				0x00000006
+#define MDP5_PIPE_SRC_OP_MODE_BWC__SHIFT			1
+static inline uint32_t MDP5_PIPE_SRC_OP_MODE_BWC(enum mdp5_pipe_bwc val)
+{
+	return ((val) << MDP5_PIPE_SRC_OP_MODE_BWC__SHIFT) & MDP5_PIPE_SRC_OP_MODE_BWC__MASK;
+}
+#define MDP5_PIPE_SRC_OP_MODE_FLIP_LR				0x00002000
+#define MDP5_PIPE_SRC_OP_MODE_FLIP_UD				0x00004000
+#define MDP5_PIPE_SRC_OP_MODE_IGC_EN				0x00010000
+#define MDP5_PIPE_SRC_OP_MODE_IGC_ROM_0				0x00020000
+#define MDP5_PIPE_SRC_OP_MODE_IGC_ROM_1				0x00040000
+#define MDP5_PIPE_SRC_OP_MODE_DEINTERLACE			0x00400000
+#define MDP5_PIPE_SRC_OP_MODE_DEINTERLACE_ODD			0x00800000
+
+static inline uint32_t REG_MDP5_PIPE_SRC_CONSTANT_COLOR(enum mdp5_pipe i0) { return 0x0000123c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_FETCH_CONFIG(enum mdp5_pipe i0) { return 0x00001248 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_VC1_RANGE(enum mdp5_pipe i0) { return 0x0000124c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_REQPRIO_FIFO_WM_0(enum mdp5_pipe i0) { return 0x00001250 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_REQPRIO_FIFO_WM_1(enum mdp5_pipe i0) { return 0x00001254 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_REQPRIO_FIFO_WM_2(enum mdp5_pipe i0) { return 0x00001258 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC_ADDR_SW_STATUS(enum mdp5_pipe i0) { return 0x00001270 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_CURRENT_SRC0_ADDR(enum mdp5_pipe i0) { return 0x000012a4 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_CURRENT_SRC1_ADDR(enum mdp5_pipe i0) { return 0x000012a8 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_CURRENT_SRC2_ADDR(enum mdp5_pipe i0) { return 0x000012ac + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_CURRENT_SRC3_ADDR(enum mdp5_pipe i0) { return 0x000012b0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_DECIMATION(enum mdp5_pipe i0) { return 0x000012b4 + 0x400*i0; }
+#define MDP5_PIPE_DECIMATION_VERT__MASK				0x000000ff
+#define MDP5_PIPE_DECIMATION_VERT__SHIFT			0
+static inline uint32_t MDP5_PIPE_DECIMATION_VERT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_DECIMATION_VERT__SHIFT) & MDP5_PIPE_DECIMATION_VERT__MASK;
+}
+#define MDP5_PIPE_DECIMATION_HORZ__MASK				0x0000ff00
+#define MDP5_PIPE_DECIMATION_HORZ__SHIFT			8
+static inline uint32_t MDP5_PIPE_DECIMATION_HORZ(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_DECIMATION_HORZ__SHIFT) & MDP5_PIPE_DECIMATION_HORZ__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_CONFIG(enum mdp5_pipe i0) { return 0x00001404 + 0x400*i0; }
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_EN			0x00000001
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_EN			0x00000002
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER__MASK		0x00000300
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER__SHIFT		8
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER__MASK		0x00000c00
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER__SHIFT		10
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER__MASK		0x00003000
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER__SHIFT		12
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER__MASK		0x0000c000
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER__SHIFT		14
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER__MASK		0x00030000
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER__SHIFT		16
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER__MASK		0x000c0000
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER__SHIFT		18
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_PHASE_STEP_X(enum mdp5_pipe i0) { return 0x00001410 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_PHASE_STEP_Y(enum mdp5_pipe i0) { return 0x00001414 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_INIT_PHASE_X(enum mdp5_pipe i0) { return 0x00001420 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_INIT_PHASE_Y(enum mdp5_pipe i0) { return 0x00001424 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM(uint32_t i0) { return 0x00003200 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_COLOR_OUT(uint32_t i0) { return 0x00003200 + 0x400*i0; }
+#define MDP5_LM_BLEND_COLOR_OUT_STAGE0_FG_ALPHA			0x00000002
+#define MDP5_LM_BLEND_COLOR_OUT_STAGE1_FG_ALPHA			0x00000004
+#define MDP5_LM_BLEND_COLOR_OUT_STAGE2_FG_ALPHA			0x00000008
+#define MDP5_LM_BLEND_COLOR_OUT_STAGE3_FG_ALPHA			0x00000010
+
+static inline uint32_t REG_MDP5_LM_OUT_SIZE(uint32_t i0) { return 0x00003204 + 0x400*i0; }
+#define MDP5_LM_OUT_SIZE_HEIGHT__MASK				0xffff0000
+#define MDP5_LM_OUT_SIZE_HEIGHT__SHIFT				16
+static inline uint32_t MDP5_LM_OUT_SIZE_HEIGHT(uint32_t val)
+{
+	return ((val) << MDP5_LM_OUT_SIZE_HEIGHT__SHIFT) & MDP5_LM_OUT_SIZE_HEIGHT__MASK;
+}
+#define MDP5_LM_OUT_SIZE_WIDTH__MASK				0x0000ffff
+#define MDP5_LM_OUT_SIZE_WIDTH__SHIFT				0
+static inline uint32_t MDP5_LM_OUT_SIZE_WIDTH(uint32_t val)
+{
+	return ((val) << MDP5_LM_OUT_SIZE_WIDTH__SHIFT) & MDP5_LM_OUT_SIZE_WIDTH__MASK;
+}
+
+static inline uint32_t REG_MDP5_LM_BORDER_COLOR_0(uint32_t i0) { return 0x00003208 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_BORDER_COLOR_1(uint32_t i0) { return 0x00003210 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_BLEND(uint32_t i0, uint32_t i1) { return 0x00003220 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_OP_MODE(uint32_t i0, uint32_t i1) { return 0x00003220 + 0x400*i0 + 0x30*i1; }
+#define MDP5_LM_BLEND_OP_MODE_FG_ALPHA__MASK			0x00000003
+#define MDP5_LM_BLEND_OP_MODE_FG_ALPHA__SHIFT			0
+static inline uint32_t MDP5_LM_BLEND_OP_MODE_FG_ALPHA(enum mdp_alpha_type val)
+{
+	return ((val) << MDP5_LM_BLEND_OP_MODE_FG_ALPHA__SHIFT) & MDP5_LM_BLEND_OP_MODE_FG_ALPHA__MASK;
+}
+#define MDP5_LM_BLEND_OP_MODE_FG_INV_ALPHA			0x00000004
+#define MDP5_LM_BLEND_OP_MODE_FG_MOD_ALPHA			0x00000008
+#define MDP5_LM_BLEND_OP_MODE_FG_INV_MOD_ALPHA			0x00000010
+#define MDP5_LM_BLEND_OP_MODE_FG_TRANSP_EN			0x00000020
+#define MDP5_LM_BLEND_OP_MODE_BG_ALPHA__MASK			0x00000300
+#define MDP5_LM_BLEND_OP_MODE_BG_ALPHA__SHIFT			8
+static inline uint32_t MDP5_LM_BLEND_OP_MODE_BG_ALPHA(enum mdp_alpha_type val)
+{
+	return ((val) << MDP5_LM_BLEND_OP_MODE_BG_ALPHA__SHIFT) & MDP5_LM_BLEND_OP_MODE_BG_ALPHA__MASK;
+}
+#define MDP5_LM_BLEND_OP_MODE_BG_INV_ALPHA			0x00000400
+#define MDP5_LM_BLEND_OP_MODE_BG_MOD_ALPHA			0x00000800
+#define MDP5_LM_BLEND_OP_MODE_BG_INV_MOD_ALPHA			0x00001000
+#define MDP5_LM_BLEND_OP_MODE_BG_TRANSP_EN			0x00002000
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_ALPHA(uint32_t i0, uint32_t i1) { return 0x00003224 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_ALPHA(uint32_t i0, uint32_t i1) { return 0x00003228 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_TRANSP_LOW0(uint32_t i0, uint32_t i1) { return 0x0000322c + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_TRANSP_LOW1(uint32_t i0, uint32_t i1) { return 0x00003230 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_TRANSP_HIGH0(uint32_t i0, uint32_t i1) { return 0x00003234 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_TRANSP_HIGH1(uint32_t i0, uint32_t i1) { return 0x00003238 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_TRANSP_LOW0(uint32_t i0, uint32_t i1) { return 0x0000323c + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_TRANSP_LOW1(uint32_t i0, uint32_t i1) { return 0x00003240 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_TRANSP_HIGH0(uint32_t i0, uint32_t i1) { return 0x00003244 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_TRANSP_HIGH1(uint32_t i0, uint32_t i1) { return 0x00003248 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_IMG_SIZE(uint32_t i0) { return 0x000032e0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_SIZE(uint32_t i0) { return 0x000032e4 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_XY(uint32_t i0) { return 0x000032e8 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_STRIDE(uint32_t i0) { return 0x000032dc + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_FORMAT(uint32_t i0) { return 0x000032ec + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BASE_ADDR(uint32_t i0) { return 0x000032f0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_START_XY(uint32_t i0) { return 0x000032f4 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_CONFIG(uint32_t i0) { return 0x000032f8 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_PARAM(uint32_t i0) { return 0x000032fc + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_TRANSP_LOW0(uint32_t i0) { return 0x00003300 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_TRANSP_LOW1(uint32_t i0) { return 0x00003304 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_TRANSP_HIGH0(uint32_t i0) { return 0x00003308 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_TRANSP_HIGH1(uint32_t i0) { return 0x0000330c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_GC_LUT_BASE(uint32_t i0) { return 0x00003310 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP(uint32_t i0) { return 0x00004600 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_OP_MODE(uint32_t i0) { return 0x00004600 + 0x400*i0; }
+#define MDP5_DSPP_OP_MODE_IGC_LUT_EN				0x00000001
+#define MDP5_DSPP_OP_MODE_IGC_TBL_IDX__MASK			0x0000000e
+#define MDP5_DSPP_OP_MODE_IGC_TBL_IDX__SHIFT			1
+static inline uint32_t MDP5_DSPP_OP_MODE_IGC_TBL_IDX(uint32_t val)
+{
+	return ((val) << MDP5_DSPP_OP_MODE_IGC_TBL_IDX__SHIFT) & MDP5_DSPP_OP_MODE_IGC_TBL_IDX__MASK;
+}
+#define MDP5_DSPP_OP_MODE_PCC_EN				0x00000010
+#define MDP5_DSPP_OP_MODE_DITHER_EN				0x00000100
+#define MDP5_DSPP_OP_MODE_HIST_EN				0x00010000
+#define MDP5_DSPP_OP_MODE_AUTO_CLEAR				0x00020000
+#define MDP5_DSPP_OP_MODE_HIST_LUT_EN				0x00080000
+#define MDP5_DSPP_OP_MODE_PA_EN					0x00100000
+#define MDP5_DSPP_OP_MODE_GAMUT_EN				0x00800000
+#define MDP5_DSPP_OP_MODE_GAMUT_ORDER				0x01000000
+
+static inline uint32_t REG_MDP5_DSPP_PCC_BASE(uint32_t i0) { return 0x00004630 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_DITHER_DEPTH(uint32_t i0) { return 0x00004750 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_HIST_CTL_BASE(uint32_t i0) { return 0x00004810 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_HIST_LUT_BASE(uint32_t i0) { return 0x00004830 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_HIST_LUT_SWAP(uint32_t i0) { return 0x00004834 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_PA_BASE(uint32_t i0) { return 0x00004838 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_GAMUT_BASE(uint32_t i0) { return 0x000048dc + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_GC_BASE(uint32_t i0) { return 0x000048b0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_INTF(uint32_t i0) { return 0x00012500 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TIMING_ENGINE_EN(uint32_t i0) { return 0x00012500 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_CONFIG(uint32_t i0) { return 0x00012504 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_HSYNC_CTL(uint32_t i0) { return 0x00012508 + 0x200*i0; }
+#define MDP5_INTF_HSYNC_CTL_PULSEW__MASK			0x0000ffff
+#define MDP5_INTF_HSYNC_CTL_PULSEW__SHIFT			0
+static inline uint32_t MDP5_INTF_HSYNC_CTL_PULSEW(uint32_t val)
+{
+	return ((val) << MDP5_INTF_HSYNC_CTL_PULSEW__SHIFT) & MDP5_INTF_HSYNC_CTL_PULSEW__MASK;
+}
+#define MDP5_INTF_HSYNC_CTL_PERIOD__MASK			0xffff0000
+#define MDP5_INTF_HSYNC_CTL_PERIOD__SHIFT			16
+static inline uint32_t MDP5_INTF_HSYNC_CTL_PERIOD(uint32_t val)
+{
+	return ((val) << MDP5_INTF_HSYNC_CTL_PERIOD__SHIFT) & MDP5_INTF_HSYNC_CTL_PERIOD__MASK;
+}
+
+static inline uint32_t REG_MDP5_INTF_VSYNC_PERIOD_F0(uint32_t i0) { return 0x0001250c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_VSYNC_PERIOD_F1(uint32_t i0) { return 0x00012510 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_VSYNC_LEN_F0(uint32_t i0) { return 0x00012514 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_VSYNC_LEN_F1(uint32_t i0) { return 0x00012518 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_VSTART_F0(uint32_t i0) { return 0x0001251c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_VSTART_F1(uint32_t i0) { return 0x00012520 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_VEND_F0(uint32_t i0) { return 0x00012524 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_VEND_F1(uint32_t i0) { return 0x00012528 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_VSTART_F0(uint32_t i0) { return 0x0001252c + 0x200*i0; }
+#define MDP5_INTF_ACTIVE_VSTART_F0_VAL__MASK			0x7fffffff
+#define MDP5_INTF_ACTIVE_VSTART_F0_VAL__SHIFT			0
+static inline uint32_t MDP5_INTF_ACTIVE_VSTART_F0_VAL(uint32_t val)
+{
+	return ((val) << MDP5_INTF_ACTIVE_VSTART_F0_VAL__SHIFT) & MDP5_INTF_ACTIVE_VSTART_F0_VAL__MASK;
+}
+#define MDP5_INTF_ACTIVE_VSTART_F0_ACTIVE_V_ENABLE		0x80000000
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_VSTART_F1(uint32_t i0) { return 0x00012530 + 0x200*i0; }
+#define MDP5_INTF_ACTIVE_VSTART_F1_VAL__MASK			0x7fffffff
+#define MDP5_INTF_ACTIVE_VSTART_F1_VAL__SHIFT			0
+static inline uint32_t MDP5_INTF_ACTIVE_VSTART_F1_VAL(uint32_t val)
+{
+	return ((val) << MDP5_INTF_ACTIVE_VSTART_F1_VAL__SHIFT) & MDP5_INTF_ACTIVE_VSTART_F1_VAL__MASK;
+}
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_VEND_F0(uint32_t i0) { return 0x00012534 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_VEND_F1(uint32_t i0) { return 0x00012538 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_HCTL(uint32_t i0) { return 0x0001253c + 0x200*i0; }
+#define MDP5_INTF_DISPLAY_HCTL_START__MASK			0x0000ffff
+#define MDP5_INTF_DISPLAY_HCTL_START__SHIFT			0
+static inline uint32_t MDP5_INTF_DISPLAY_HCTL_START(uint32_t val)
+{
+	return ((val) << MDP5_INTF_DISPLAY_HCTL_START__SHIFT) & MDP5_INTF_DISPLAY_HCTL_START__MASK;
+}
+#define MDP5_INTF_DISPLAY_HCTL_END__MASK			0xffff0000
+#define MDP5_INTF_DISPLAY_HCTL_END__SHIFT			16
+static inline uint32_t MDP5_INTF_DISPLAY_HCTL_END(uint32_t val)
+{
+	return ((val) << MDP5_INTF_DISPLAY_HCTL_END__SHIFT) & MDP5_INTF_DISPLAY_HCTL_END__MASK;
+}
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_HCTL(uint32_t i0) { return 0x00012540 + 0x200*i0; }
+#define MDP5_INTF_ACTIVE_HCTL_START__MASK			0x00007fff
+#define MDP5_INTF_ACTIVE_HCTL_START__SHIFT			0
+static inline uint32_t MDP5_INTF_ACTIVE_HCTL_START(uint32_t val)
+{
+	return ((val) << MDP5_INTF_ACTIVE_HCTL_START__SHIFT) & MDP5_INTF_ACTIVE_HCTL_START__MASK;
+}
+#define MDP5_INTF_ACTIVE_HCTL_END__MASK				0x7fff0000
+#define MDP5_INTF_ACTIVE_HCTL_END__SHIFT			16
+static inline uint32_t MDP5_INTF_ACTIVE_HCTL_END(uint32_t val)
+{
+	return ((val) << MDP5_INTF_ACTIVE_HCTL_END__SHIFT) & MDP5_INTF_ACTIVE_HCTL_END__MASK;
+}
+#define MDP5_INTF_ACTIVE_HCTL_ACTIVE_H_ENABLE			0x80000000
+
+static inline uint32_t REG_MDP5_INTF_BORDER_COLOR(uint32_t i0) { return 0x00012544 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_UNDERFLOW_COLOR(uint32_t i0) { return 0x00012548 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_HSYNC_SKEW(uint32_t i0) { return 0x0001254c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_POLARITY_CTL(uint32_t i0) { return 0x00012550 + 0x200*i0; }
+#define MDP5_INTF_POLARITY_CTL_HSYNC_LOW			0x00000001
+#define MDP5_INTF_POLARITY_CTL_VSYNC_LOW			0x00000002
+#define MDP5_INTF_POLARITY_CTL_DATA_EN_LOW			0x00000004
+
+static inline uint32_t REG_MDP5_INTF_TEST_CTL(uint32_t i0) { return 0x00012554 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TP_COLOR0(uint32_t i0) { return 0x00012558 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TP_COLOR1(uint32_t i0) { return 0x0001255c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DSI_CMD_MODE_TRIGGER_EN(uint32_t i0) { return 0x00012584 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_PANEL_FORMAT(uint32_t i0) { return 0x00012590 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_FRAME_LINE_COUNT_EN(uint32_t i0) { return 0x000125a8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_FRAME_COUNT(uint32_t i0) { return 0x000125ac + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_LINE_COUNT(uint32_t i0) { return 0x000125b0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DEFLICKER_CONFIG(uint32_t i0) { return 0x000125f0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DEFLICKER_STRNG_COEFF(uint32_t i0) { return 0x000125f4 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DEFLICKER_WEAK_COEFF(uint32_t i0) { return 0x000125f8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_ENABLE(uint32_t i0) { return 0x00012600 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_MAIN_CONTROL(uint32_t i0) { return 0x00012604 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_VIDEO_CONFIG(uint32_t i0) { return 0x00012608 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_COMPONENT_LIMITS(uint32_t i0) { return 0x0001260c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_RECTANGLE(uint32_t i0) { return 0x00012610 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_INITIAL_VALUE(uint32_t i0) { return 0x00012614 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_BLK_WHITE_PATTERN_FRAME(uint32_t i0) { return 0x00012618 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_RGB_MAPPING(uint32_t i0) { return 0x0001261c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD(uint32_t i0) { return 0x00013100 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BYPASS(uint32_t i0) { return 0x00013100 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CTRL_0(uint32_t i0) { return 0x00013104 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CTRL_1(uint32_t i0) { return 0x00013108 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_FRAME_SIZE(uint32_t i0) { return 0x0001310c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CON_CTRL_0(uint32_t i0) { return 0x00013110 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CON_CTRL_1(uint32_t i0) { return 0x00013114 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_STR_MAN(uint32_t i0) { return 0x00013118 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_VAR(uint32_t i0) { return 0x0001311c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_DITH(uint32_t i0) { return 0x00013120 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_DITH_CTRL(uint32_t i0) { return 0x00013124 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_AMP_LIM(uint32_t i0) { return 0x00013128 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_SLOPE(uint32_t i0) { return 0x0001312c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BW_LVL(uint32_t i0) { return 0x00013130 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_LOGO_POS(uint32_t i0) { return 0x00013134 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_LUT_FI(uint32_t i0) { return 0x00013138 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_LUT_CC(uint32_t i0) { return 0x0001317c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_STR_LIM(uint32_t i0) { return 0x000131c8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CALIB_AB(uint32_t i0) { return 0x000131cc + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CALIB_CD(uint32_t i0) { return 0x000131d0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_MODE_SEL(uint32_t i0) { return 0x000131d4 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_TFILT_CTRL(uint32_t i0) { return 0x000131d8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BL_MINMAX(uint32_t i0) { return 0x000131dc + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BL(uint32_t i0) { return 0x000131e0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BL_MAX(uint32_t i0) { return 0x000131e8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_AL(uint32_t i0) { return 0x000131ec + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_AL_MIN(uint32_t i0) { return 0x000131f0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_AL_FILT(uint32_t i0) { return 0x000131f4 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CFG_BUF(uint32_t i0) { return 0x000131f8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_LUT_AL(uint32_t i0) { return 0x00013200 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_TARG_STR(uint32_t i0) { return 0x00013244 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_START_CALC(uint32_t i0) { return 0x00013248 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_STR_OUT(uint32_t i0) { return 0x0001324c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BL_OUT(uint32_t i0) { return 0x00013254 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CALC_DONE(uint32_t i0) { return 0x00013258 + 0x200*i0; }
+
+
+#endif /* MDP5_XML */
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
new file mode 100644
index 00000000000..ebe2e60f3ab
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
@@ -0,0 +1,575 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "mdp5_kms.h"
+
+#include <drm/drm_mode.h>
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+#include "drm_flip_work.h"
+
+struct mdp5_crtc {
+	struct drm_crtc base;
+	char name[8];
+	struct drm_plane *plane;
+	struct drm_plane *planes[8];
+	int id;
+	bool enabled;
+
+	/* which mixer/encoder we route output to: */
+	int mixer;
+
+	/* if there is a pending flip, these will be non-null: */
+	struct drm_pending_vblank_event *event;
+	struct msm_fence_cb pageflip_cb;
+
+#define PENDING_CURSOR 0x1
+#define PENDING_FLIP   0x2
+	atomic_t pending;
+
+	/* the fb that we logically (from PoV of KMS API) hold a ref
+	 * to.  Which we may not yet be scanning out (we may still
+	 * be scanning out previous in case of page_flip while waiting
+	 * for gpu rendering to complete:
+	 */
+	struct drm_framebuffer *fb;
+
+	/* the fb that we currently hold a scanout ref to: */
+	struct drm_framebuffer *scanout_fb;
+
+	/* for unref'ing framebuffers after scanout completes: */
+	struct drm_flip_work unref_fb_work;
+
+	struct mdp_irq vblank;
+	struct mdp_irq err;
+};
+#define to_mdp5_crtc(x) container_of(x, struct mdp5_crtc, base)
+
+static struct mdp5_kms *get_kms(struct drm_crtc *crtc)
+{
+	struct msm_drm_private *priv = crtc->dev->dev_private;
+	return to_mdp5_kms(to_mdp_kms(priv->kms));
+}
+
+static void request_pending(struct drm_crtc *crtc, uint32_t pending)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+
+	atomic_or(pending, &mdp5_crtc->pending);
+	mdp_irq_register(&get_kms(crtc)->base, &mdp5_crtc->vblank);
+}
+
+static void crtc_flush(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	int id = mdp5_crtc->id;
+	uint32_t i, flush = 0;
+
+	for (i = 0; i < ARRAY_SIZE(mdp5_crtc->planes); i++) {
+		struct drm_plane *plane = mdp5_crtc->planes[i];
+		if (plane) {
+			enum mdp5_pipe pipe = mdp5_plane_pipe(plane);
+			flush |= pipe2flush(pipe);
+		}
+	}
+	flush |= mixer2flush(mdp5_crtc->id);
+	flush |= MDP5_CTL_FLUSH_CTL;
+
+	DBG("%s: flush=%08x", mdp5_crtc->name, flush);
+
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_FLUSH(id), flush);
+}
+
+static void update_fb(struct drm_crtc *crtc, struct drm_framebuffer *new_fb)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct drm_framebuffer *old_fb = mdp5_crtc->fb;
+
+	/* grab reference to incoming scanout fb: */
+	drm_framebuffer_reference(new_fb);
+	mdp5_crtc->base.primary->fb = new_fb;
+	mdp5_crtc->fb = new_fb;
+
+	if (old_fb)
+		drm_flip_work_queue(&mdp5_crtc->unref_fb_work, old_fb);
+}
+
+/* unlike update_fb(), take a ref to the new scanout fb *before* updating
+ * plane, then call this.  Needed to ensure we don't unref the buffer that
+ * is actually still being scanned out.
+ *
+ * Note that this whole thing goes away with atomic.. since we can defer
+ * calling into driver until rendering is done.
+ */
+static void update_scanout(struct drm_crtc *crtc, struct drm_framebuffer *fb)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+
+	/* flush updates, to make sure hw is updated to new scanout fb,
+	 * so that we can safely queue unref to current fb (ie. next
+	 * vblank we know hw is done w/ previous scanout_fb).
+	 */
+	crtc_flush(crtc);
+
+	if (mdp5_crtc->scanout_fb)
+		drm_flip_work_queue(&mdp5_crtc->unref_fb_work,
+				mdp5_crtc->scanout_fb);
+
+	mdp5_crtc->scanout_fb = fb;
+
+	/* enable vblank to complete flip: */
+	request_pending(crtc, PENDING_FLIP);
+}
+
+/* if file!=NULL, this is preclose potential cancel-flip path */
+static void complete_flip(struct drm_crtc *crtc, struct drm_file *file)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct drm_device *dev = crtc->dev;
+	struct drm_pending_vblank_event *event;
+	unsigned long flags, i;
+
+	spin_lock_irqsave(&dev->event_lock, flags);
+	event = mdp5_crtc->event;
+	if (event) {
+		/* if regular vblank case (!file) or if cancel-flip from
+		 * preclose on file that requested flip, then send the
+		 * event:
+		 */
+		if (!file || (event->base.file_priv == file)) {
+			mdp5_crtc->event = NULL;
+			drm_send_vblank_event(dev, mdp5_crtc->id, event);
+		}
+	}
+	spin_unlock_irqrestore(&dev->event_lock, flags);
+
+	for (i = 0; i < ARRAY_SIZE(mdp5_crtc->planes); i++) {
+		struct drm_plane *plane = mdp5_crtc->planes[i];
+		if (plane)
+			mdp5_plane_complete_flip(plane);
+	}
+}
+
+static void pageflip_cb(struct msm_fence_cb *cb)
+{
+	struct mdp5_crtc *mdp5_crtc =
+		container_of(cb, struct mdp5_crtc, pageflip_cb);
+	struct drm_crtc *crtc = &mdp5_crtc->base;
+	struct drm_framebuffer *fb = mdp5_crtc->fb;
+
+	if (!fb)
+		return;
+
+	drm_framebuffer_reference(fb);
+	mdp5_plane_set_scanout(mdp5_crtc->plane, fb);
+	update_scanout(crtc, fb);
+}
+
+static void unref_fb_worker(struct drm_flip_work *work, void *val)
+{
+	struct mdp5_crtc *mdp5_crtc =
+		container_of(work, struct mdp5_crtc, unref_fb_work);
+	struct drm_device *dev = mdp5_crtc->base.dev;
+
+	mutex_lock(&dev->mode_config.mutex);
+	drm_framebuffer_unreference(val);
+	mutex_unlock(&dev->mode_config.mutex);
+}
+
+static void mdp5_crtc_destroy(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+
+	drm_crtc_cleanup(crtc);
+	drm_flip_work_cleanup(&mdp5_crtc->unref_fb_work);
+
+	kfree(mdp5_crtc);
+}
+
+static void mdp5_crtc_dpms(struct drm_crtc *crtc, int mode)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	bool enabled = (mode == DRM_MODE_DPMS_ON);
+
+	DBG("%s: mode=%d", mdp5_crtc->name, mode);
+
+	if (enabled != mdp5_crtc->enabled) {
+		if (enabled) {
+			mdp5_enable(mdp5_kms);
+			mdp_irq_register(&mdp5_kms->base, &mdp5_crtc->err);
+		} else {
+			mdp_irq_unregister(&mdp5_kms->base, &mdp5_crtc->err);
+			mdp5_disable(mdp5_kms);
+		}
+		mdp5_crtc->enabled = enabled;
+	}
+}
+
+static bool mdp5_crtc_mode_fixup(struct drm_crtc *crtc,
+		const struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static void blend_setup(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	int id = mdp5_crtc->id;
+
+	/*
+	 * Hard-coded setup for now until I figure out how the
+	 * layer-mixer works
+	 */
+
+	/* LM[id]: */
+	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_COLOR_OUT(id),
+			MDP5_LM_BLEND_COLOR_OUT_STAGE0_FG_ALPHA);
+	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_OP_MODE(id, 0),
+			MDP5_LM_BLEND_OP_MODE_FG_ALPHA(FG_CONST) |
+			MDP5_LM_BLEND_OP_MODE_BG_ALPHA(FG_PIXEL) |
+			MDP5_LM_BLEND_OP_MODE_BG_INV_ALPHA);
+	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_FG_ALPHA(id, 0), 0xff);
+	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_BG_ALPHA(id, 0), 0x00);
+
+	/* NOTE: seems that LM[n] and CTL[m], we do not need n==m.. but
+	 * we want to be setting CTL[m].LAYER[n].  Not sure what the
+	 * point of having CTL[m].LAYER[o] (for o!=n).. maybe that is
+	 * used when chaining up mixers for high resolution displays?
+	 */
+
+	/* CTL[id]: */
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 0),
+			MDP5_CTL_LAYER_REG_RGB0(STAGE0) |
+			MDP5_CTL_LAYER_REG_BORDER_COLOR);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 1), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 2), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 3), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 4), 0);
+}
+
+static int mdp5_crtc_mode_set(struct drm_crtc *crtc,
+		struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode,
+		int x, int y,
+		struct drm_framebuffer *old_fb)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	int ret;
+
+	mode = adjusted_mode;
+
+	DBG("%s: set mode: %d:\"%s\" %d %d %d %d %d %d %d %d %d %d 0x%x 0x%x",
+			mdp5_crtc->name, mode->base.id, mode->name,
+			mode->vrefresh, mode->clock,
+			mode->hdisplay, mode->hsync_start,
+			mode->hsync_end, mode->htotal,
+			mode->vdisplay, mode->vsync_start,
+			mode->vsync_end, mode->vtotal,
+			mode->type, mode->flags);
+
+	/* grab extra ref for update_scanout() */
+	drm_framebuffer_reference(crtc->primary->fb);
+
+	ret = mdp5_plane_mode_set(mdp5_crtc->plane, crtc, crtc->primary->fb,
+			0, 0, mode->hdisplay, mode->vdisplay,
+			x << 16, y << 16,
+			mode->hdisplay << 16, mode->vdisplay << 16);
+	if (ret) {
+		drm_framebuffer_unreference(crtc->primary->fb);
+		dev_err(crtc->dev->dev, "%s: failed to set mode on plane: %d\n",
+				mdp5_crtc->name, ret);
+		return ret;
+	}
+
+	mdp5_write(mdp5_kms, REG_MDP5_LM_OUT_SIZE(mdp5_crtc->id),
+			MDP5_LM_OUT_SIZE_WIDTH(mode->hdisplay) |
+			MDP5_LM_OUT_SIZE_HEIGHT(mode->vdisplay));
+
+	update_fb(crtc, crtc->primary->fb);
+	update_scanout(crtc, crtc->primary->fb);
+
+	return 0;
+}
+
+static void mdp5_crtc_prepare(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	DBG("%s", mdp5_crtc->name);
+	/* make sure we hold a ref to mdp clks while setting up mode: */
+	mdp5_enable(get_kms(crtc));
+	mdp5_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
+}
+
+static void mdp5_crtc_commit(struct drm_crtc *crtc)
+{
+	mdp5_crtc_dpms(crtc, DRM_MODE_DPMS_ON);
+	crtc_flush(crtc);
+	/* drop the ref to mdp clk's that we got in prepare: */
+	mdp5_disable(get_kms(crtc));
+}
+
+static int mdp5_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
+		struct drm_framebuffer *old_fb)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct drm_plane *plane = mdp5_crtc->plane;
+	struct drm_display_mode *mode = &crtc->mode;
+	int ret;
+
+	/* grab extra ref for update_scanout() */
+	drm_framebuffer_reference(crtc->primary->fb);
+
+	ret = mdp5_plane_mode_set(plane, crtc, crtc->primary->fb,
+			0, 0, mode->hdisplay, mode->vdisplay,
+			x << 16, y << 16,
+			mode->hdisplay << 16, mode->vdisplay << 16);
+	if (ret) {
+		drm_framebuffer_unreference(crtc->primary->fb);
+		return ret;
+	}
+
+	update_fb(crtc, crtc->primary->fb);
+	update_scanout(crtc, crtc->primary->fb);
+
+	return 0;
+}
+
+static void mdp5_crtc_load_lut(struct drm_crtc *crtc)
+{
+}
+
+static int mdp5_crtc_page_flip(struct drm_crtc *crtc,
+		struct drm_framebuffer *new_fb,
+		struct drm_pending_vblank_event *event,
+		uint32_t page_flip_flags)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct drm_device *dev = crtc->dev;
+	struct drm_gem_object *obj;
+	unsigned long flags;
+
+	if (mdp5_crtc->event) {
+		dev_err(dev->dev, "already pending flip!\n");
+		return -EBUSY;
+	}
+
+	obj = msm_framebuffer_bo(new_fb, 0);
+
+	spin_lock_irqsave(&dev->event_lock, flags);
+	mdp5_crtc->event = event;
+	spin_unlock_irqrestore(&dev->event_lock, flags);
+
+	update_fb(crtc, new_fb);
+
+	return msm_gem_queue_inactive_cb(obj, &mdp5_crtc->pageflip_cb);
+}
+
+static int mdp5_crtc_set_property(struct drm_crtc *crtc,
+		struct drm_property *property, uint64_t val)
+{
+	// XXX
+	return -EINVAL;
+}
+
+static const struct drm_crtc_funcs mdp5_crtc_funcs = {
+	.set_config = drm_crtc_helper_set_config,
+	.destroy = mdp5_crtc_destroy,
+	.page_flip = mdp5_crtc_page_flip,
+	.set_property = mdp5_crtc_set_property,
+};
+
+static const struct drm_crtc_helper_funcs mdp5_crtc_helper_funcs = {
+	.dpms = mdp5_crtc_dpms,
+	.mode_fixup = mdp5_crtc_mode_fixup,
+	.mode_set = mdp5_crtc_mode_set,
+	.prepare = mdp5_crtc_prepare,
+	.commit = mdp5_crtc_commit,
+	.mode_set_base = mdp5_crtc_mode_set_base,
+	.load_lut = mdp5_crtc_load_lut,
+};
+
+static void mdp5_crtc_vblank_irq(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	struct mdp5_crtc *mdp5_crtc = container_of(irq, struct mdp5_crtc, vblank);
+	struct drm_crtc *crtc = &mdp5_crtc->base;
+	struct msm_drm_private *priv = crtc->dev->dev_private;
+	unsigned pending;
+
+	mdp_irq_unregister(&get_kms(crtc)->base, &mdp5_crtc->vblank);
+
+	pending = atomic_xchg(&mdp5_crtc->pending, 0);
+
+	if (pending & PENDING_FLIP) {
+		complete_flip(crtc, NULL);
+		drm_flip_work_commit(&mdp5_crtc->unref_fb_work, priv->wq);
+	}
+}
+
+static void mdp5_crtc_err_irq(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	struct mdp5_crtc *mdp5_crtc = container_of(irq, struct mdp5_crtc, err);
+	struct drm_crtc *crtc = &mdp5_crtc->base;
+	DBG("%s: error: %08x", mdp5_crtc->name, irqstatus);
+	crtc_flush(crtc);
+}
+
+uint32_t mdp5_crtc_vblank(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	return mdp5_crtc->vblank.irqmask;
+}
+
+void mdp5_crtc_cancel_pending_flip(struct drm_crtc *crtc, struct drm_file *file)
+{
+	DBG("cancel: %p", file);
+	complete_flip(crtc, file);
+}
+
+/* set interface for routing crtc->encoder: */
+void mdp5_crtc_set_intf(struct drm_crtc *crtc, int intf,
+		enum mdp5_intf intf_id)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	static const enum mdp5_intfnum intfnum[] = {
+			INTF0, INTF1, INTF2, INTF3,
+	};
+	uint32_t intf_sel;
+
+	/* now that we know what irq's we want: */
+	mdp5_crtc->err.irqmask = intf2err(intf);
+	mdp5_crtc->vblank.irqmask = intf2vblank(intf);
+
+	/* when called from modeset_init(), skip the rest until later: */
+	if (!mdp5_kms)
+		return;
+
+	intf_sel = mdp5_read(mdp5_kms, REG_MDP5_DISP_INTF_SEL);
+
+	switch (intf) {
+	case 0:
+		intf_sel &= ~MDP5_DISP_INTF_SEL_INTF0__MASK;
+		intf_sel |= MDP5_DISP_INTF_SEL_INTF0(intf_id);
+		break;
+	case 1:
+		intf_sel &= ~MDP5_DISP_INTF_SEL_INTF1__MASK;
+		intf_sel |= MDP5_DISP_INTF_SEL_INTF1(intf_id);
+		break;
+	case 2:
+		intf_sel &= ~MDP5_DISP_INTF_SEL_INTF2__MASK;
+		intf_sel |= MDP5_DISP_INTF_SEL_INTF2(intf_id);
+		break;
+	case 3:
+		intf_sel &= ~MDP5_DISP_INTF_SEL_INTF3__MASK;
+		intf_sel |= MDP5_DISP_INTF_SEL_INTF3(intf_id);
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	blend_setup(crtc);
+
+	DBG("%s: intf_sel=%08x", mdp5_crtc->name, intf_sel);
+
+	mdp5_write(mdp5_kms, REG_MDP5_DISP_INTF_SEL, intf_sel);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(mdp5_crtc->id),
+			MDP5_CTL_OP_MODE(MODE_NONE) |
+			MDP5_CTL_OP_INTF_NUM(intfnum[intf]));
+
+	crtc_flush(crtc);
+}
+
+static void set_attach(struct drm_crtc *crtc, enum mdp5_pipe pipe_id,
+		struct drm_plane *plane)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+
+	BUG_ON(pipe_id >= ARRAY_SIZE(mdp5_crtc->planes));
+
+	if (mdp5_crtc->planes[pipe_id] == plane)
+		return;
+
+	mdp5_crtc->planes[pipe_id] = plane;
+	blend_setup(crtc);
+	if (mdp5_crtc->enabled && (plane != mdp5_crtc->plane))
+		crtc_flush(crtc);
+}
+
+void mdp5_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane)
+{
+	set_attach(crtc, mdp5_plane_pipe(plane), plane);
+}
+
+void mdp5_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane)
+{
+	/* don't actually detatch our primary plane: */
+	if (to_mdp5_crtc(crtc)->plane == plane)
+		return;
+	set_attach(crtc, mdp5_plane_pipe(plane), NULL);
+}
+
+/* initialize crtc */
+struct drm_crtc *mdp5_crtc_init(struct drm_device *dev,
+		struct drm_plane *plane, int id)
+{
+	struct drm_crtc *crtc = NULL;
+	struct mdp5_crtc *mdp5_crtc;
+	int ret;
+
+	mdp5_crtc = kzalloc(sizeof(*mdp5_crtc), GFP_KERNEL);
+	if (!mdp5_crtc) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	crtc = &mdp5_crtc->base;
+
+	mdp5_crtc->plane = plane;
+	mdp5_crtc->id = id;
+
+	mdp5_crtc->vblank.irq = mdp5_crtc_vblank_irq;
+	mdp5_crtc->err.irq = mdp5_crtc_err_irq;
+
+	snprintf(mdp5_crtc->name, sizeof(mdp5_crtc->name), "%s:%d",
+			pipe2name(mdp5_plane_pipe(plane)), id);
+
+	ret = drm_flip_work_init(&mdp5_crtc->unref_fb_work, 16,
+			"unref fb", unref_fb_worker);
+	if (ret)
+		goto fail;
+
+	INIT_FENCE_CB(&mdp5_crtc->pageflip_cb, pageflip_cb);
+
+	drm_crtc_init_with_planes(dev, crtc, plane, NULL, &mdp5_crtc_funcs);
+	drm_crtc_helper_add(crtc, &mdp5_crtc_helper_funcs);
+
+	mdp5_plane_install_properties(mdp5_crtc->plane, &crtc->base);
+
+	return crtc;
+
+fail:
+	if (crtc)
+		mdp5_crtc_destroy(crtc);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
new file mode 100644
index 00000000000..edec7bfaa95
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
@@ -0,0 +1,258 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "mdp5_kms.h"
+
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+
+struct mdp5_encoder {
+	struct drm_encoder base;
+	int intf;
+	enum mdp5_intf intf_id;
+	bool enabled;
+	uint32_t bsc;
+};
+#define to_mdp5_encoder(x) container_of(x, struct mdp5_encoder, base)
+
+static struct mdp5_kms *get_kms(struct drm_encoder *encoder)
+{
+	struct msm_drm_private *priv = encoder->dev->dev_private;
+	return to_mdp5_kms(to_mdp_kms(priv->kms));
+}
+
+#ifdef CONFIG_MSM_BUS_SCALING
+#include <mach/board.h>
+#include <mach/msm_bus.h>
+#include <mach/msm_bus_board.h>
+#define MDP_BUS_VECTOR_ENTRY(ab_val, ib_val)		\
+	{						\
+		.src = MSM_BUS_MASTER_MDP_PORT0,	\
+		.dst = MSM_BUS_SLAVE_EBI_CH0,		\
+		.ab = (ab_val),				\
+		.ib = (ib_val),				\
+	}
+
+static struct msm_bus_vectors mdp_bus_vectors[] = {
+	MDP_BUS_VECTOR_ENTRY(0, 0),
+	MDP_BUS_VECTOR_ENTRY(2000000000, 2000000000),
+};
+static struct msm_bus_paths mdp_bus_usecases[] = { {
+		.num_paths = 1,
+		.vectors = &mdp_bus_vectors[0],
+}, {
+		.num_paths = 1,
+		.vectors = &mdp_bus_vectors[1],
+} };
+static struct msm_bus_scale_pdata mdp_bus_scale_table = {
+	.usecase = mdp_bus_usecases,
+	.num_usecases = ARRAY_SIZE(mdp_bus_usecases),
+	.name = "mdss_mdp",
+};
+
+static void bs_init(struct mdp5_encoder *mdp5_encoder)
+{
+	mdp5_encoder->bsc = msm_bus_scale_register_client(
+			&mdp_bus_scale_table);
+	DBG("bus scale client: %08x", mdp5_encoder->bsc);
+}
+
+static void bs_fini(struct mdp5_encoder *mdp5_encoder)
+{
+	if (mdp5_encoder->bsc) {
+		msm_bus_scale_unregister_client(mdp5_encoder->bsc);
+		mdp5_encoder->bsc = 0;
+	}
+}
+
+static void bs_set(struct mdp5_encoder *mdp5_encoder, int idx)
+{
+	if (mdp5_encoder->bsc) {
+		DBG("set bus scaling: %d", idx);
+		/* HACK: scaling down, and then immediately back up
+		 * seems to leave things broken (underflow).. so
+		 * never disable:
+		 */
+		idx = 1;
+		msm_bus_scale_client_update_request(mdp5_encoder->bsc, idx);
+	}
+}
+#else
+static void bs_init(struct mdp5_encoder *mdp5_encoder) {}
+static void bs_fini(struct mdp5_encoder *mdp5_encoder) {}
+static void bs_set(struct mdp5_encoder *mdp5_encoder, int idx) {}
+#endif
+
+static void mdp5_encoder_destroy(struct drm_encoder *encoder)
+{
+	struct mdp5_encoder *mdp5_encoder = to_mdp5_encoder(encoder);
+	bs_fini(mdp5_encoder);
+	drm_encoder_cleanup(encoder);
+	kfree(mdp5_encoder);
+}
+
+static const struct drm_encoder_funcs mdp5_encoder_funcs = {
+	.destroy = mdp5_encoder_destroy,
+};
+
+static void mdp5_encoder_dpms(struct drm_encoder *encoder, int mode)
+{
+	struct mdp5_encoder *mdp5_encoder = to_mdp5_encoder(encoder);
+	struct mdp5_kms *mdp5_kms = get_kms(encoder);
+	int intf = mdp5_encoder->intf;
+	bool enabled = (mode == DRM_MODE_DPMS_ON);
+
+	DBG("mode=%d", mode);
+
+	if (enabled == mdp5_encoder->enabled)
+		return;
+
+	if (enabled) {
+		bs_set(mdp5_encoder, 1);
+		mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(intf), 1);
+	} else {
+		mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(intf), 0);
+		bs_set(mdp5_encoder, 0);
+	}
+
+	mdp5_encoder->enabled = enabled;
+}
+
+static bool mdp5_encoder_mode_fixup(struct drm_encoder *encoder,
+		const struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static void mdp5_encoder_mode_set(struct drm_encoder *encoder,
+		struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	struct mdp5_encoder *mdp5_encoder = to_mdp5_encoder(encoder);
+	struct mdp5_kms *mdp5_kms = get_kms(encoder);
+	int intf = mdp5_encoder->intf;
+	uint32_t dtv_hsync_skew, vsync_period, vsync_len, ctrl_pol;
+	uint32_t display_v_start, display_v_end;
+	uint32_t hsync_start_x, hsync_end_x;
+	uint32_t format;
+
+	mode = adjusted_mode;
+
+	DBG("set mode: %d:\"%s\" %d %d %d %d %d %d %d %d %d %d 0x%x 0x%x",
+			mode->base.id, mode->name,
+			mode->vrefresh, mode->clock,
+			mode->hdisplay, mode->hsync_start,
+			mode->hsync_end, mode->htotal,
+			mode->vdisplay, mode->vsync_start,
+			mode->vsync_end, mode->vtotal,
+			mode->type, mode->flags);
+
+	ctrl_pol = 0;
+	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+		ctrl_pol |= MDP5_INTF_POLARITY_CTL_HSYNC_LOW;
+	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+		ctrl_pol |= MDP5_INTF_POLARITY_CTL_VSYNC_LOW;
+	/* probably need to get DATA_EN polarity from panel.. */
+
+	dtv_hsync_skew = 0;  /* get this from panel? */
+	format = 0x213f;     /* get this from panel? */
+
+	hsync_start_x = (mode->htotal - mode->hsync_start);
+	hsync_end_x = mode->htotal - (mode->hsync_start - mode->hdisplay) - 1;
+
+	vsync_period = mode->vtotal * mode->htotal;
+	vsync_len = (mode->vsync_end - mode->vsync_start) * mode->htotal;
+	display_v_start = (mode->vtotal - mode->vsync_start) * mode->htotal + dtv_hsync_skew;
+	display_v_end = vsync_period - ((mode->vsync_start - mode->vdisplay) * mode->htotal) + dtv_hsync_skew - 1;
+
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_HSYNC_CTL(intf),
+			MDP5_INTF_HSYNC_CTL_PULSEW(mode->hsync_end - mode->hsync_start) |
+			MDP5_INTF_HSYNC_CTL_PERIOD(mode->htotal));
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_VSYNC_PERIOD_F0(intf), vsync_period);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_VSYNC_LEN_F0(intf), vsync_len);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_DISPLAY_HCTL(intf),
+			MDP5_INTF_DISPLAY_HCTL_START(hsync_start_x) |
+			MDP5_INTF_DISPLAY_HCTL_END(hsync_end_x));
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_DISPLAY_VSTART_F0(intf), display_v_start);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_DISPLAY_VEND_F0(intf), display_v_end);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_BORDER_COLOR(intf), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_UNDERFLOW_COLOR(intf), 0xff);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_HSYNC_SKEW(intf), dtv_hsync_skew);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_POLARITY_CTL(intf), ctrl_pol);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_ACTIVE_HCTL(intf),
+			MDP5_INTF_ACTIVE_HCTL_START(0) |
+			MDP5_INTF_ACTIVE_HCTL_END(0));
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_ACTIVE_VSTART_F0(intf), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_ACTIVE_VEND_F0(intf), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_PANEL_FORMAT(intf), format);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_FRAME_LINE_COUNT_EN(intf), 0x3);  /* frame+line? */
+}
+
+static void mdp5_encoder_prepare(struct drm_encoder *encoder)
+{
+	mdp5_encoder_dpms(encoder, DRM_MODE_DPMS_OFF);
+}
+
+static void mdp5_encoder_commit(struct drm_encoder *encoder)
+{
+	struct mdp5_encoder *mdp5_encoder = to_mdp5_encoder(encoder);
+	mdp5_crtc_set_intf(encoder->crtc, mdp5_encoder->intf,
+			mdp5_encoder->intf_id);
+	mdp5_encoder_dpms(encoder, DRM_MODE_DPMS_ON);
+}
+
+static const struct drm_encoder_helper_funcs mdp5_encoder_helper_funcs = {
+	.dpms = mdp5_encoder_dpms,
+	.mode_fixup = mdp5_encoder_mode_fixup,
+	.mode_set = mdp5_encoder_mode_set,
+	.prepare = mdp5_encoder_prepare,
+	.commit = mdp5_encoder_commit,
+};
+
+/* initialize encoder */
+struct drm_encoder *mdp5_encoder_init(struct drm_device *dev, int intf,
+		enum mdp5_intf intf_id)
+{
+	struct drm_encoder *encoder = NULL;
+	struct mdp5_encoder *mdp5_encoder;
+	int ret;
+
+	mdp5_encoder = kzalloc(sizeof(*mdp5_encoder), GFP_KERNEL);
+	if (!mdp5_encoder) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	mdp5_encoder->intf = intf;
+	mdp5_encoder->intf_id = intf_id;
+	encoder = &mdp5_encoder->base;
+
+	drm_encoder_init(dev, encoder, &mdp5_encoder_funcs,
+			 DRM_MODE_ENCODER_TMDS);
+	drm_encoder_helper_add(encoder, &mdp5_encoder_helper_funcs);
+
+	bs_init(mdp5_encoder);
+
+	return encoder;
+
+fail:
+	if (encoder)
+		mdp5_encoder_destroy(encoder);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c
new file mode 100644
index 00000000000..f2b985bc2ad
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c
@@ -0,0 +1,111 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+
+#include "msm_drv.h"
+#include "mdp5_kms.h"
+
+void mdp5_set_irqmask(struct mdp_kms *mdp_kms, uint32_t irqmask)
+{
+	mdp5_write(to_mdp5_kms(mdp_kms), REG_MDP5_INTR_EN, irqmask);
+}
+
+static void mdp5_irq_error_handler(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	DRM_ERROR("errors: %08x\n", irqstatus);
+}
+
+void mdp5_irq_preinstall(struct msm_kms *kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	mdp5_write(mdp5_kms, REG_MDP5_INTR_CLEAR, 0xffffffff);
+}
+
+int mdp5_irq_postinstall(struct msm_kms *kms)
+{
+	struct mdp_kms *mdp_kms = to_mdp_kms(kms);
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(mdp_kms);
+	struct mdp_irq *error_handler = &mdp5_kms->error_handler;
+
+	error_handler->irq = mdp5_irq_error_handler;
+	error_handler->irqmask = MDP5_IRQ_INTF0_UNDER_RUN |
+			MDP5_IRQ_INTF1_UNDER_RUN |
+			MDP5_IRQ_INTF2_UNDER_RUN |
+			MDP5_IRQ_INTF3_UNDER_RUN;
+
+	mdp_irq_register(mdp_kms, error_handler);
+
+	return 0;
+}
+
+void mdp5_irq_uninstall(struct msm_kms *kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	mdp5_write(mdp5_kms, REG_MDP5_INTR_EN, 0x00000000);
+}
+
+static void mdp5_irq_mdp(struct mdp_kms *mdp_kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(mdp_kms);
+	struct drm_device *dev = mdp5_kms->dev;
+	struct msm_drm_private *priv = dev->dev_private;
+	unsigned int id;
+	uint32_t status;
+
+	status = mdp5_read(mdp5_kms, REG_MDP5_INTR_STATUS);
+	mdp5_write(mdp5_kms, REG_MDP5_INTR_CLEAR, status);
+
+	VERB("status=%08x", status);
+
+	mdp_dispatch_irqs(mdp_kms, status);
+
+	for (id = 0; id < priv->num_crtcs; id++)
+		if (status & mdp5_crtc_vblank(priv->crtcs[id]))
+			drm_handle_vblank(dev, id);
+}
+
+irqreturn_t mdp5_irq(struct msm_kms *kms)
+{
+	struct mdp_kms *mdp_kms = to_mdp_kms(kms);
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(mdp_kms);
+	uint32_t intr;
+
+	intr = mdp5_read(mdp5_kms, REG_MDP5_HW_INTR_STATUS);
+
+	VERB("intr=%08x", intr);
+
+	if (intr & MDP5_HW_INTR_STATUS_INTR_MDP)
+		mdp5_irq_mdp(mdp_kms);
+
+	if (intr & MDP5_HW_INTR_STATUS_INTR_HDMI)
+		hdmi_irq(0, mdp5_kms->hdmi);
+
+	return IRQ_HANDLED;
+}
+
+int mdp5_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
+{
+	mdp_update_vblank_mask(to_mdp_kms(kms),
+			mdp5_crtc_vblank(crtc), true);
+	return 0;
+}
+
+void mdp5_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
+{
+	mdp_update_vblank_mask(to_mdp_kms(kms),
+			mdp5_crtc_vblank(crtc), false);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
new file mode 100644
index 00000000000..71510ee26e9
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
@@ -0,0 +1,372 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+
+#include "msm_drv.h"
+#include "msm_mmu.h"
+#include "mdp5_kms.h"
+
+static const char *iommu_ports[] = {
+		"mdp_0",
+};
+
+static struct mdp5_platform_config *mdp5_get_config(struct platform_device *dev);
+
+static int mdp5_hw_init(struct msm_kms *kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	struct drm_device *dev = mdp5_kms->dev;
+	uint32_t version, major, minor;
+	int ret = 0;
+
+	pm_runtime_get_sync(dev->dev);
+
+	mdp5_enable(mdp5_kms);
+	version = mdp5_read(mdp5_kms, REG_MDP5_MDP_VERSION);
+	mdp5_disable(mdp5_kms);
+
+	major = FIELD(version, MDP5_MDP_VERSION_MAJOR);
+	minor = FIELD(version, MDP5_MDP_VERSION_MINOR);
+
+	DBG("found MDP5 version v%d.%d", major, minor);
+
+	if ((major != 1) || ((minor != 0) && (minor != 2))) {
+		dev_err(dev->dev, "unexpected MDP version: v%d.%d\n",
+				major, minor);
+		ret = -ENXIO;
+		goto out;
+	}
+
+	mdp5_kms->rev = minor;
+
+	/* Magic unknown register writes:
+	 *
+	 *    W VBIF:0x004 00000001      (mdss_mdp.c:839)
+	 *    W MDP5:0x2e0 0xe9          (mdss_mdp.c:839)
+	 *    W MDP5:0x2e4 0x55          (mdss_mdp.c:839)
+	 *    W MDP5:0x3ac 0xc0000ccc    (mdss_mdp.c:839)
+	 *    W MDP5:0x3b4 0xc0000ccc    (mdss_mdp.c:839)
+	 *    W MDP5:0x3bc 0xcccccc      (mdss_mdp.c:839)
+	 *    W MDP5:0x4a8 0xcccc0c0     (mdss_mdp.c:839)
+	 *    W MDP5:0x4b0 0xccccc0c0    (mdss_mdp.c:839)
+	 *    W MDP5:0x4b8 0xccccc000    (mdss_mdp.c:839)
+	 *
+	 * Downstream fbdev driver gets these register offsets/values
+	 * from DT.. not really sure what these registers are or if
+	 * different values for different boards/SoC's, etc.  I guess
+	 * they are the golden registers.
+	 *
+	 * Not setting these does not seem to cause any problem.  But
+	 * we may be getting lucky with the bootloader initializing
+	 * them for us.  OTOH, if we can always count on the bootloader
+	 * setting the golden registers, then perhaps we don't need to
+	 * care.
+	 */
+
+	mdp5_write(mdp5_kms, REG_MDP5_DISP_INTF_SEL, 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(0), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(1), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(2), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(3), 0);
+
+out:
+	pm_runtime_put_sync(dev->dev);
+
+	return ret;
+}
+
+static long mdp5_round_pixclk(struct msm_kms *kms, unsigned long rate,
+		struct drm_encoder *encoder)
+{
+	return rate;
+}
+
+static void mdp5_preclose(struct msm_kms *kms, struct drm_file *file)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	struct msm_drm_private *priv = mdp5_kms->dev->dev_private;
+	unsigned i;
+
+	for (i = 0; i < priv->num_crtcs; i++)
+		mdp5_crtc_cancel_pending_flip(priv->crtcs[i], file);
+}
+
+static void mdp5_destroy(struct msm_kms *kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	struct msm_mmu *mmu = mdp5_kms->mmu;
+
+	if (mmu) {
+		mmu->funcs->detach(mmu, iommu_ports, ARRAY_SIZE(iommu_ports));
+		mmu->funcs->destroy(mmu);
+	}
+	kfree(mdp5_kms);
+}
+
+static const struct mdp_kms_funcs kms_funcs = {
+	.base = {
+		.hw_init         = mdp5_hw_init,
+		.irq_preinstall  = mdp5_irq_preinstall,
+		.irq_postinstall = mdp5_irq_postinstall,
+		.irq_uninstall   = mdp5_irq_uninstall,
+		.irq             = mdp5_irq,
+		.enable_vblank   = mdp5_enable_vblank,
+		.disable_vblank  = mdp5_disable_vblank,
+		.get_format      = mdp_get_format,
+		.round_pixclk    = mdp5_round_pixclk,
+		.preclose        = mdp5_preclose,
+		.destroy         = mdp5_destroy,
+	},
+	.set_irqmask         = mdp5_set_irqmask,
+};
+
+int mdp5_disable(struct mdp5_kms *mdp5_kms)
+{
+	DBG("");
+
+	clk_disable_unprepare(mdp5_kms->ahb_clk);
+	clk_disable_unprepare(mdp5_kms->axi_clk);
+	clk_disable_unprepare(mdp5_kms->core_clk);
+	clk_disable_unprepare(mdp5_kms->lut_clk);
+
+	return 0;
+}
+
+int mdp5_enable(struct mdp5_kms *mdp5_kms)
+{
+	DBG("");
+
+	clk_prepare_enable(mdp5_kms->ahb_clk);
+	clk_prepare_enable(mdp5_kms->axi_clk);
+	clk_prepare_enable(mdp5_kms->core_clk);
+	clk_prepare_enable(mdp5_kms->lut_clk);
+
+	return 0;
+}
+
+static int modeset_init(struct mdp5_kms *mdp5_kms)
+{
+	static const enum mdp5_pipe crtcs[] = {
+			SSPP_RGB0, SSPP_RGB1, SSPP_RGB2,
+	};
+	struct drm_device *dev = mdp5_kms->dev;
+	struct msm_drm_private *priv = dev->dev_private;
+	struct drm_encoder *encoder;
+	int i, ret;
+
+	/* construct CRTCs: */
+	for (i = 0; i < ARRAY_SIZE(crtcs); i++) {
+		struct drm_plane *plane;
+		struct drm_crtc *crtc;
+
+		plane = mdp5_plane_init(dev, crtcs[i], true);
+		if (IS_ERR(plane)) {
+			ret = PTR_ERR(plane);
+			dev_err(dev->dev, "failed to construct plane for %s (%d)\n",
+					pipe2name(crtcs[i]), ret);
+			goto fail;
+		}
+
+		crtc  = mdp5_crtc_init(dev, plane, i);
+		if (IS_ERR(crtc)) {
+			ret = PTR_ERR(crtc);
+			dev_err(dev->dev, "failed to construct crtc for %s (%d)\n",
+					pipe2name(crtcs[i]), ret);
+			goto fail;
+		}
+		priv->crtcs[priv->num_crtcs++] = crtc;
+	}
+
+	/* Construct encoder for HDMI: */
+	encoder = mdp5_encoder_init(dev, 3, INTF_HDMI);
+	if (IS_ERR(encoder)) {
+		dev_err(dev->dev, "failed to construct encoder\n");
+		ret = PTR_ERR(encoder);
+		goto fail;
+	}
+
+	/* NOTE: the vsync and error irq's are actually associated with
+	 * the INTF/encoder.. the easiest way to deal with this (ie. what
+	 * we do now) is assume a fixed relationship between crtc's and
+	 * encoders.  I'm not sure if there is ever a need to more freely
+	 * assign crtcs to encoders, but if there is then we need to take
+	 * care of error and vblank irq's that the crtc has registered,
+	 * and also update user-requested vblank_mask.
+	 */
+	encoder->possible_crtcs = BIT(0);
+	mdp5_crtc_set_intf(priv->crtcs[0], 3, INTF_HDMI);
+
+	priv->encoders[priv->num_encoders++] = encoder;
+
+	/* Construct bridge/connector for HDMI: */
+	mdp5_kms->hdmi = hdmi_init(dev, encoder);
+	if (IS_ERR(mdp5_kms->hdmi)) {
+		ret = PTR_ERR(mdp5_kms->hdmi);
+		dev_err(dev->dev, "failed to initialize HDMI: %d\n", ret);
+		goto fail;
+	}
+
+	return 0;
+
+fail:
+	return ret;
+}
+
+static int get_clk(struct platform_device *pdev, struct clk **clkp,
+		const char *name)
+{
+	struct device *dev = &pdev->dev;
+	struct clk *clk = devm_clk_get(dev, name);
+	if (IS_ERR(clk)) {
+		dev_err(dev, "failed to get %s (%ld)\n", name, PTR_ERR(clk));
+		return PTR_ERR(clk);
+	}
+	*clkp = clk;
+	return 0;
+}
+
+struct msm_kms *mdp5_kms_init(struct drm_device *dev)
+{
+	struct platform_device *pdev = dev->platformdev;
+	struct mdp5_platform_config *config = mdp5_get_config(pdev);
+	struct mdp5_kms *mdp5_kms;
+	struct msm_kms *kms = NULL;
+	struct msm_mmu *mmu;
+	int ret;
+
+	mdp5_kms = kzalloc(sizeof(*mdp5_kms), GFP_KERNEL);
+	if (!mdp5_kms) {
+		dev_err(dev->dev, "failed to allocate kms\n");
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	mdp_kms_init(&mdp5_kms->base, &kms_funcs);
+
+	kms = &mdp5_kms->base.base;
+
+	mdp5_kms->dev = dev;
+	mdp5_kms->smp_blk_cnt = config->smp_blk_cnt;
+
+	mdp5_kms->mmio = msm_ioremap(pdev, "mdp_phys", "MDP5");
+	if (IS_ERR(mdp5_kms->mmio)) {
+		ret = PTR_ERR(mdp5_kms->mmio);
+		goto fail;
+	}
+
+	mdp5_kms->vbif = msm_ioremap(pdev, "vbif_phys", "VBIF");
+	if (IS_ERR(mdp5_kms->vbif)) {
+		ret = PTR_ERR(mdp5_kms->vbif);
+		goto fail;
+	}
+
+	mdp5_kms->vdd = devm_regulator_get(&pdev->dev, "vdd");
+	if (IS_ERR(mdp5_kms->vdd)) {
+		ret = PTR_ERR(mdp5_kms->vdd);
+		goto fail;
+	}
+
+	ret = regulator_enable(mdp5_kms->vdd);
+	if (ret) {
+		dev_err(dev->dev, "failed to enable regulator vdd: %d\n", ret);
+		goto fail;
+	}
+
+	ret = get_clk(pdev, &mdp5_kms->axi_clk, "bus_clk");
+	if (ret)
+		goto fail;
+	ret = get_clk(pdev, &mdp5_kms->ahb_clk, "iface_clk");
+	if (ret)
+		goto fail;
+	ret = get_clk(pdev, &mdp5_kms->src_clk, "core_clk_src");
+	if (ret)
+		goto fail;
+	ret = get_clk(pdev, &mdp5_kms->core_clk, "core_clk");
+	if (ret)
+		goto fail;
+	ret = get_clk(pdev, &mdp5_kms->lut_clk, "lut_clk");
+	if (ret)
+		goto fail;
+	ret = get_clk(pdev, &mdp5_kms->vsync_clk, "vsync_clk");
+	if (ret)
+		goto fail;
+
+	ret = clk_set_rate(mdp5_kms->src_clk, config->max_clk);
+
+	/* make sure things are off before attaching iommu (bootloader could
+	 * have left things on, in which case we'll start getting faults if
+	 * we don't disable):
+	 */
+	mdp5_enable(mdp5_kms);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(0), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(1), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(2), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(3), 0);
+	mdp5_disable(mdp5_kms);
+	mdelay(16);
+
+	if (config->iommu) {
+		mmu = msm_iommu_new(dev, config->iommu);
+		if (IS_ERR(mmu)) {
+			ret = PTR_ERR(mmu);
+			dev_err(dev->dev, "failed to init iommu: %d\n", ret);
+			goto fail;
+		}
+
+		ret = mmu->funcs->attach(mmu, iommu_ports,
+				ARRAY_SIZE(iommu_ports));
+		if (ret) {
+			dev_err(dev->dev, "failed to attach iommu: %d\n", ret);
+			mmu->funcs->destroy(mmu);
+			goto fail;
+		}
+	} else {
+		dev_info(dev->dev, "no iommu, fallback to phys "
+				"contig buffers for scanout\n");
+		mmu = NULL;
+	}
+	mdp5_kms->mmu = mmu;
+
+	mdp5_kms->id = msm_register_mmu(dev, mmu);
+	if (mdp5_kms->id < 0) {
+		ret = mdp5_kms->id;
+		dev_err(dev->dev, "failed to register mdp5 iommu: %d\n", ret);
+		goto fail;
+	}
+
+	ret = modeset_init(mdp5_kms);
+	if (ret) {
+		dev_err(dev->dev, "modeset_init failed: %d\n", ret);
+		goto fail;
+	}
+
+	return kms;
+
+fail:
+	if (kms)
+		mdp5_destroy(kms);
+	return ERR_PTR(ret);
+}
+
+static struct mdp5_platform_config *mdp5_get_config(struct platform_device *dev)
+{
+	static struct mdp5_platform_config config = {};
+#ifdef CONFIG_OF
+	/* TODO */
+#endif
+	return &config;
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
new file mode 100644
index 00000000000..6e981b692d1
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
@@ -0,0 +1,214 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __MDP5_KMS_H__
+#define __MDP5_KMS_H__
+
+#include "msm_drv.h"
+#include "msm_kms.h"
+#include "mdp/mdp_kms.h"
+#include "mdp5.xml.h"
+#include "mdp5_smp.h"
+
+struct mdp5_kms {
+	struct mdp_kms base;
+
+	struct drm_device *dev;
+
+	int rev;
+
+	/* mapper-id used to request GEM buffer mapped for scanout: */
+	int id;
+	struct msm_mmu *mmu;
+
+	/* for tracking smp allocation amongst pipes: */
+	mdp5_smp_state_t smp_state;
+	struct mdp5_client_smp_state smp_client_state[CID_MAX];
+	int smp_blk_cnt;
+
+	/* io/register spaces: */
+	void __iomem *mmio, *vbif;
+
+	struct regulator *vdd;
+
+	struct clk *axi_clk;
+	struct clk *ahb_clk;
+	struct clk *src_clk;
+	struct clk *core_clk;
+	struct clk *lut_clk;
+	struct clk *vsync_clk;
+
+	struct hdmi *hdmi;
+
+	struct mdp_irq error_handler;
+};
+#define to_mdp5_kms(x) container_of(x, struct mdp5_kms, base)
+
+/* platform config data (ie. from DT, or pdata) */
+struct mdp5_platform_config {
+	struct iommu_domain *iommu;
+	uint32_t max_clk;
+	int smp_blk_cnt;
+};
+
+static inline void mdp5_write(struct mdp5_kms *mdp5_kms, u32 reg, u32 data)
+{
+	msm_writel(data, mdp5_kms->mmio + reg);
+}
+
+static inline u32 mdp5_read(struct mdp5_kms *mdp5_kms, u32 reg)
+{
+	return msm_readl(mdp5_kms->mmio + reg);
+}
+
+static inline const char *pipe2name(enum mdp5_pipe pipe)
+{
+	static const char *names[] = {
+#define NAME(n) [SSPP_ ## n] = #n
+		NAME(VIG0), NAME(VIG1), NAME(VIG2),
+		NAME(RGB0), NAME(RGB1), NAME(RGB2),
+		NAME(DMA0), NAME(DMA1),
+#undef NAME
+	};
+	return names[pipe];
+}
+
+static inline uint32_t pipe2flush(enum mdp5_pipe pipe)
+{
+	switch (pipe) {
+	case SSPP_VIG0: return MDP5_CTL_FLUSH_VIG0;
+	case SSPP_VIG1: return MDP5_CTL_FLUSH_VIG1;
+	case SSPP_VIG2: return MDP5_CTL_FLUSH_VIG2;
+	case SSPP_RGB0: return MDP5_CTL_FLUSH_RGB0;
+	case SSPP_RGB1: return MDP5_CTL_FLUSH_RGB1;
+	case SSPP_RGB2: return MDP5_CTL_FLUSH_RGB2;
+	case SSPP_DMA0: return MDP5_CTL_FLUSH_DMA0;
+	case SSPP_DMA1: return MDP5_CTL_FLUSH_DMA1;
+	default:        return 0;
+	}
+}
+
+static inline int pipe2nclients(enum mdp5_pipe pipe)
+{
+	switch (pipe) {
+	case SSPP_RGB0:
+	case SSPP_RGB1:
+	case SSPP_RGB2:
+		return 1;
+	default:
+		return 3;
+	}
+}
+
+static inline enum mdp5_client_id pipe2client(enum mdp5_pipe pipe, int plane)
+{
+	WARN_ON(plane >= pipe2nclients(pipe));
+	switch (pipe) {
+	case SSPP_VIG0: return CID_VIG0_Y + plane;
+	case SSPP_VIG1: return CID_VIG1_Y + plane;
+	case SSPP_VIG2: return CID_VIG2_Y + plane;
+	case SSPP_RGB0: return CID_RGB0;
+	case SSPP_RGB1: return CID_RGB1;
+	case SSPP_RGB2: return CID_RGB2;
+	case SSPP_DMA0: return CID_DMA0_Y + plane;
+	case SSPP_DMA1: return CID_DMA1_Y + plane;
+	default:        return CID_UNUSED;
+	}
+}
+
+static inline uint32_t mixer2flush(int lm)
+{
+	switch (lm) {
+	case 0:  return MDP5_CTL_FLUSH_LM0;
+	case 1:  return MDP5_CTL_FLUSH_LM1;
+	case 2:  return MDP5_CTL_FLUSH_LM2;
+	default: return 0;
+	}
+}
+
+static inline uint32_t intf2err(int intf)
+{
+	switch (intf) {
+	case 0:  return MDP5_IRQ_INTF0_UNDER_RUN;
+	case 1:  return MDP5_IRQ_INTF1_UNDER_RUN;
+	case 2:  return MDP5_IRQ_INTF2_UNDER_RUN;
+	case 3:  return MDP5_IRQ_INTF3_UNDER_RUN;
+	default: return 0;
+	}
+}
+
+static inline uint32_t intf2vblank(int intf)
+{
+	switch (intf) {
+	case 0:  return MDP5_IRQ_INTF0_VSYNC;
+	case 1:  return MDP5_IRQ_INTF1_VSYNC;
+	case 2:  return MDP5_IRQ_INTF2_VSYNC;
+	case 3:  return MDP5_IRQ_INTF3_VSYNC;
+	default: return 0;
+	}
+}
+
+int mdp5_disable(struct mdp5_kms *mdp5_kms);
+int mdp5_enable(struct mdp5_kms *mdp5_kms);
+
+void mdp5_set_irqmask(struct mdp_kms *mdp_kms, uint32_t irqmask);
+void mdp5_irq_preinstall(struct msm_kms *kms);
+int mdp5_irq_postinstall(struct msm_kms *kms);
+void mdp5_irq_uninstall(struct msm_kms *kms);
+irqreturn_t mdp5_irq(struct msm_kms *kms);
+int mdp5_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
+void mdp5_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
+
+static inline
+uint32_t mdp5_get_formats(enum mdp5_pipe pipe, uint32_t *pixel_formats,
+		uint32_t max_formats)
+{
+	/* TODO when we have YUV, we need to filter supported formats
+	 * based on pipe id..
+	 */
+	return mdp_get_formats(pixel_formats, max_formats);
+}
+
+void mdp5_plane_install_properties(struct drm_plane *plane,
+		struct drm_mode_object *obj);
+void mdp5_plane_set_scanout(struct drm_plane *plane,
+		struct drm_framebuffer *fb);
+int mdp5_plane_mode_set(struct drm_plane *plane,
+		struct drm_crtc *crtc, struct drm_framebuffer *fb,
+		int crtc_x, int crtc_y,
+		unsigned int crtc_w, unsigned int crtc_h,
+		uint32_t src_x, uint32_t src_y,
+		uint32_t src_w, uint32_t src_h);
+void mdp5_plane_complete_flip(struct drm_plane *plane);
+enum mdp5_pipe mdp5_plane_pipe(struct drm_plane *plane);
+struct drm_plane *mdp5_plane_init(struct drm_device *dev,
+		enum mdp5_pipe pipe, bool private_plane);
+
+uint32_t mdp5_crtc_vblank(struct drm_crtc *crtc);
+
+void mdp5_crtc_cancel_pending_flip(struct drm_crtc *crtc, struct drm_file *file);
+void mdp5_crtc_set_intf(struct drm_crtc *crtc, int intf,
+		enum mdp5_intf intf_id);
+void mdp5_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane);
+void mdp5_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane);
+struct drm_crtc *mdp5_crtc_init(struct drm_device *dev,
+		struct drm_plane *plane, int id);
+
+struct drm_encoder *mdp5_encoder_init(struct drm_device *dev, int intf,
+		enum mdp5_intf intf_id);
+
+#endif /* __MDP5_KMS_H__ */
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
new file mode 100644
index 00000000000..f3daec4412a
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
@@ -0,0 +1,394 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "mdp5_kms.h"
+
+
+struct mdp5_plane {
+	struct drm_plane base;
+	const char *name;
+
+	enum mdp5_pipe pipe;
+
+	uint32_t nformats;
+	uint32_t formats[32];
+
+	bool enabled;
+};
+#define to_mdp5_plane(x) container_of(x, struct mdp5_plane, base)
+
+static struct mdp5_kms *get_kms(struct drm_plane *plane)
+{
+	struct msm_drm_private *priv = plane->dev->dev_private;
+	return to_mdp5_kms(to_mdp_kms(priv->kms));
+}
+
+static int mdp5_plane_update(struct drm_plane *plane,
+		struct drm_crtc *crtc, struct drm_framebuffer *fb,
+		int crtc_x, int crtc_y,
+		unsigned int crtc_w, unsigned int crtc_h,
+		uint32_t src_x, uint32_t src_y,
+		uint32_t src_w, uint32_t src_h)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+
+	mdp5_plane->enabled = true;
+
+	if (plane->fb)
+		drm_framebuffer_unreference(plane->fb);
+
+	drm_framebuffer_reference(fb);
+
+	return mdp5_plane_mode_set(plane, crtc, fb,
+			crtc_x, crtc_y, crtc_w, crtc_h,
+			src_x, src_y, src_w, src_h);
+}
+
+static int mdp5_plane_disable(struct drm_plane *plane)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	int i;
+
+	DBG("%s: disable", mdp5_plane->name);
+
+	/* update our SMP request to zero (release all our blks): */
+	for (i = 0; i < pipe2nclients(pipe); i++)
+		mdp5_smp_request(mdp5_kms, pipe2client(pipe, i), 0);
+
+	/* TODO detaching now will cause us not to get the last
+	 * vblank and mdp5_smp_commit().. so other planes will
+	 * still see smp blocks previously allocated to us as
+	 * in-use..
+	 */
+	if (plane->crtc)
+		mdp5_crtc_detach(plane->crtc, plane);
+
+	return 0;
+}
+
+static void mdp5_plane_destroy(struct drm_plane *plane)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct msm_drm_private *priv = plane->dev->dev_private;
+
+	if (priv->kms)
+		mdp5_plane_disable(plane);
+
+	drm_plane_cleanup(plane);
+
+	kfree(mdp5_plane);
+}
+
+/* helper to install properties which are common to planes and crtcs */
+void mdp5_plane_install_properties(struct drm_plane *plane,
+		struct drm_mode_object *obj)
+{
+	// XXX
+}
+
+int mdp5_plane_set_property(struct drm_plane *plane,
+		struct drm_property *property, uint64_t val)
+{
+	// XXX
+	return -EINVAL;
+}
+
+static const struct drm_plane_funcs mdp5_plane_funcs = {
+		.update_plane = mdp5_plane_update,
+		.disable_plane = mdp5_plane_disable,
+		.destroy = mdp5_plane_destroy,
+		.set_property = mdp5_plane_set_property,
+};
+
+void mdp5_plane_set_scanout(struct drm_plane *plane,
+		struct drm_framebuffer *fb)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	uint32_t nplanes = drm_format_num_planes(fb->pixel_format);
+	uint32_t iova[4];
+	int i;
+
+	for (i = 0; i < nplanes; i++) {
+		struct drm_gem_object *bo = msm_framebuffer_bo(fb, i);
+		msm_gem_get_iova(bo, mdp5_kms->id, &iova[i]);
+	}
+	for (; i < 4; i++)
+		iova[i] = 0;
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_STRIDE_A(pipe),
+			MDP5_PIPE_SRC_STRIDE_A_P0(fb->pitches[0]) |
+			MDP5_PIPE_SRC_STRIDE_A_P1(fb->pitches[1]));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_STRIDE_B(pipe),
+			MDP5_PIPE_SRC_STRIDE_B_P2(fb->pitches[2]) |
+			MDP5_PIPE_SRC_STRIDE_B_P3(fb->pitches[3]));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC0_ADDR(pipe), iova[0]);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC1_ADDR(pipe), iova[1]);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC2_ADDR(pipe), iova[2]);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC3_ADDR(pipe), iova[3]);
+
+	plane->fb = fb;
+}
+
+/* NOTE: looks like if horizontal decimation is used (if we supported that)
+ * then the width used to calculate SMP block requirements is the post-
+ * decimated width.  Ie. SMP buffering sits downstream of decimation (which
+ * presumably happens during the dma from scanout buffer).
+ */
+static int request_smp_blocks(struct drm_plane *plane, uint32_t format,
+		uint32_t nplanes, uint32_t width)
+{
+	struct drm_device *dev = plane->dev;
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	int i, hsub, nlines, nblks, ret;
+
+	hsub = drm_format_horz_chroma_subsampling(format);
+
+	/* different if BWC (compressed framebuffer?) enabled: */
+	nlines = 2;
+
+	for (i = 0, nblks = 0; i < nplanes; i++) {
+		int n, fetch_stride, cpp;
+
+		cpp = drm_format_plane_cpp(format, i);
+		fetch_stride = width * cpp / (i ? hsub : 1);
+
+		n = DIV_ROUND_UP(fetch_stride * nlines, SMP_BLK_SIZE);
+
+		/* for hw rev v1.00 */
+		if (mdp5_kms->rev == 0)
+			n = roundup_pow_of_two(n);
+
+		DBG("%s[%d]: request %d SMP blocks", mdp5_plane->name, i, n);
+		ret = mdp5_smp_request(mdp5_kms, pipe2client(pipe, i), n);
+		if (ret) {
+			dev_err(dev->dev, "Could not allocate %d SMP blocks: %d\n",
+					n, ret);
+			return ret;
+		}
+
+		nblks += n;
+	}
+
+	/* in success case, return total # of blocks allocated: */
+	return nblks;
+}
+
+static void set_fifo_thresholds(struct drm_plane *plane, int nblks)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	uint32_t val;
+
+	/* 1/4 of SMP pool that is being fetched */
+	val = (nblks * SMP_ENTRIES_PER_BLK) / 4;
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_0(pipe), val * 1);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_1(pipe), val * 2);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_2(pipe), val * 3);
+
+}
+
+int mdp5_plane_mode_set(struct drm_plane *plane,
+		struct drm_crtc *crtc, struct drm_framebuffer *fb,
+		int crtc_x, int crtc_y,
+		unsigned int crtc_w, unsigned int crtc_h,
+		uint32_t src_x, uint32_t src_y,
+		uint32_t src_w, uint32_t src_h)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	const struct mdp_format *format;
+	uint32_t nplanes, config = 0;
+	uint32_t phasex_step = 0, phasey_step = 0;
+	uint32_t hdecm = 0, vdecm = 0;
+	int i, nblks;
+
+	nplanes = drm_format_num_planes(fb->pixel_format);
+
+	/* bad formats should already be rejected: */
+	if (WARN_ON(nplanes > pipe2nclients(pipe)))
+		return -EINVAL;
+
+	/* src values are in Q16 fixed point, convert to integer: */
+	src_x = src_x >> 16;
+	src_y = src_y >> 16;
+	src_w = src_w >> 16;
+	src_h = src_h >> 16;
+
+	DBG("%s: FB[%u] %u,%u,%u,%u -> CRTC[%u] %d,%d,%u,%u", mdp5_plane->name,
+			fb->base.id, src_x, src_y, src_w, src_h,
+			crtc->base.id, crtc_x, crtc_y, crtc_w, crtc_h);
+
+	/*
+	 * Calculate and request required # of smp blocks:
+	 */
+	nblks = request_smp_blocks(plane, fb->pixel_format, nplanes, src_w);
+	if (nblks < 0)
+		return nblks;
+
+	/*
+	 * Currently we update the hw for allocations/requests immediately,
+	 * but once atomic modeset/pageflip is in place, the allocation
+	 * would move into atomic->check_plane_state(), while updating the
+	 * hw would remain here:
+	 */
+	for (i = 0; i < pipe2nclients(pipe); i++)
+		mdp5_smp_configure(mdp5_kms, pipe2client(pipe, i));
+
+	if (src_w != crtc_w) {
+		config |= MDP5_PIPE_SCALE_CONFIG_SCALEX_EN;
+		/* TODO calc phasex_step, hdecm */
+	}
+
+	if (src_h != crtc_h) {
+		config |= MDP5_PIPE_SCALE_CONFIG_SCALEY_EN;
+		/* TODO calc phasey_step, vdecm */
+	}
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_IMG_SIZE(pipe),
+			MDP5_PIPE_SRC_IMG_SIZE_WIDTH(src_w) |
+			MDP5_PIPE_SRC_IMG_SIZE_HEIGHT(src_h));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_SIZE(pipe),
+			MDP5_PIPE_SRC_SIZE_WIDTH(src_w) |
+			MDP5_PIPE_SRC_SIZE_HEIGHT(src_h));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_XY(pipe),
+			MDP5_PIPE_SRC_XY_X(src_x) |
+			MDP5_PIPE_SRC_XY_Y(src_y));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_OUT_SIZE(pipe),
+			MDP5_PIPE_OUT_SIZE_WIDTH(crtc_w) |
+			MDP5_PIPE_OUT_SIZE_HEIGHT(crtc_h));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_OUT_XY(pipe),
+			MDP5_PIPE_OUT_XY_X(crtc_x) |
+			MDP5_PIPE_OUT_XY_Y(crtc_y));
+
+	mdp5_plane_set_scanout(plane, fb);
+
+	format = to_mdp_format(msm_framebuffer_format(fb));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_FORMAT(pipe),
+			MDP5_PIPE_SRC_FORMAT_A_BPC(format->bpc_a) |
+			MDP5_PIPE_SRC_FORMAT_R_BPC(format->bpc_r) |
+			MDP5_PIPE_SRC_FORMAT_G_BPC(format->bpc_g) |
+			MDP5_PIPE_SRC_FORMAT_B_BPC(format->bpc_b) |
+			COND(format->alpha_enable, MDP5_PIPE_SRC_FORMAT_ALPHA_ENABLE) |
+			MDP5_PIPE_SRC_FORMAT_CPP(format->cpp - 1) |
+			MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT(format->unpack_count - 1) |
+			COND(format->unpack_tight, MDP5_PIPE_SRC_FORMAT_UNPACK_TIGHT) |
+			MDP5_PIPE_SRC_FORMAT_NUM_PLANES(nplanes - 1) |
+			MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP(CHROMA_RGB));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_UNPACK(pipe),
+			MDP5_PIPE_SRC_UNPACK_ELEM0(format->unpack[0]) |
+			MDP5_PIPE_SRC_UNPACK_ELEM1(format->unpack[1]) |
+			MDP5_PIPE_SRC_UNPACK_ELEM2(format->unpack[2]) |
+			MDP5_PIPE_SRC_UNPACK_ELEM3(format->unpack[3]));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_OP_MODE(pipe),
+			MDP5_PIPE_SRC_OP_MODE_BWC(BWC_LOSSLESS));
+
+	/* not using secure mode: */
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_ADDR_SW_STATUS(pipe), 0);
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SCALE_PHASE_STEP_X(pipe), phasex_step);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SCALE_PHASE_STEP_Y(pipe), phasey_step);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_DECIMATION(pipe),
+			MDP5_PIPE_DECIMATION_VERT(vdecm) |
+			MDP5_PIPE_DECIMATION_HORZ(hdecm));
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SCALE_CONFIG(pipe),
+			MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER(SCALE_FILTER_NEAREST));
+
+	set_fifo_thresholds(plane, nblks);
+
+	/* TODO detach from old crtc (if we had more than one) */
+	mdp5_crtc_attach(crtc, plane);
+
+	return 0;
+}
+
+void mdp5_plane_complete_flip(struct drm_plane *plane)
+{
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = to_mdp5_plane(plane)->pipe;
+	int i;
+
+	for (i = 0; i < pipe2nclients(pipe); i++)
+		mdp5_smp_commit(mdp5_kms, pipe2client(pipe, i));
+}
+
+enum mdp5_pipe mdp5_plane_pipe(struct drm_plane *plane)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	return mdp5_plane->pipe;
+}
+
+/* initialize plane */
+struct drm_plane *mdp5_plane_init(struct drm_device *dev,
+		enum mdp5_pipe pipe, bool private_plane)
+{
+	struct drm_plane *plane = NULL;
+	struct mdp5_plane *mdp5_plane;
+	int ret;
+	enum drm_plane_type type;
+
+	mdp5_plane = kzalloc(sizeof(*mdp5_plane), GFP_KERNEL);
+	if (!mdp5_plane) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	plane = &mdp5_plane->base;
+
+	mdp5_plane->pipe = pipe;
+	mdp5_plane->name = pipe2name(pipe);
+
+	mdp5_plane->nformats = mdp5_get_formats(pipe, mdp5_plane->formats,
+			ARRAY_SIZE(mdp5_plane->formats));
+
+	type = private_plane ? DRM_PLANE_TYPE_PRIMARY : DRM_PLANE_TYPE_OVERLAY;
+	drm_universal_plane_init(dev, plane, 0xff, &mdp5_plane_funcs,
+				 mdp5_plane->formats, mdp5_plane->nformats,
+				 type);
+
+	mdp5_plane_install_properties(plane, &plane->base);
+
+	return plane;
+
+fail:
+	if (plane)
+		mdp5_plane_destroy(plane);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c
new file mode 100644
index 00000000000..2d0236b963a
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c
@@ -0,0 +1,173 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+
+#include "mdp5_kms.h"
+#include "mdp5_smp.h"
+
+
+/* SMP - Shared Memory Pool
+ *
+ * These are shared between all the clients, where each plane in a
+ * scanout buffer is a SMP client.  Ie. scanout of 3 plane I420 on
+ * pipe VIG0 => 3 clients: VIG0_Y, VIG0_CB, VIG0_CR.
+ *
+ * Based on the size of the attached scanout buffer, a certain # of
+ * blocks must be allocated to that client out of the shared pool.
+ *
+ * For each block, it can be either free, or pending/in-use by a
+ * client.  The updates happen in three steps:
+ *
+ *  1) mdp5_smp_request():
+ *     When plane scanout is setup, calculate required number of
+ *     blocks needed per client, and request.  Blocks not inuse or
+ *     pending by any other client are added to client's pending
+ *     set.
+ *
+ *  2) mdp5_smp_configure():
+ *     As hw is programmed, before FLUSH, MDP5_SMP_ALLOC registers
+ *     are configured for the union(pending, inuse)
+ *
+ *  3) mdp5_smp_commit():
+ *     After next vblank, copy pending -> inuse.  Optionally update
+ *     MDP5_SMP_ALLOC registers if there are newly unused blocks
+ *
+ * On the next vblank after changes have been committed to hw, the
+ * client's pending blocks become it's in-use blocks (and no-longer
+ * in-use blocks become available to other clients).
+ *
+ * btw, hurray for confusing overloaded acronyms!  :-/
+ *
+ * NOTE: for atomic modeset/pageflip NONBLOCK operations, step #1
+ * should happen at (or before)? atomic->check().  And we'd need
+ * an API to discard previous requests if update is aborted or
+ * (test-only).
+ *
+ * TODO would perhaps be nice to have debugfs to dump out kernel
+ * inuse and pending state of all clients..
+ */
+
+static DEFINE_SPINLOCK(smp_lock);
+
+
+/* step #1: update # of blocks pending for the client: */
+int mdp5_smp_request(struct mdp5_kms *mdp5_kms,
+		enum mdp5_client_id cid, int nblks)
+{
+	struct mdp5_client_smp_state *ps = &mdp5_kms->smp_client_state[cid];
+	int i, ret, avail, cur_nblks, cnt = mdp5_kms->smp_blk_cnt;
+	unsigned long flags;
+
+	spin_lock_irqsave(&smp_lock, flags);
+
+	avail = cnt - bitmap_weight(mdp5_kms->smp_state, cnt);
+	if (nblks > avail) {
+		ret = -ENOSPC;
+		goto fail;
+	}
+
+	cur_nblks = bitmap_weight(ps->pending, cnt);
+	if (nblks > cur_nblks) {
+		/* grow the existing pending reservation: */
+		for (i = cur_nblks; i < nblks; i++) {
+			int blk = find_first_zero_bit(mdp5_kms->smp_state, cnt);
+			set_bit(blk, ps->pending);
+			set_bit(blk, mdp5_kms->smp_state);
+		}
+	} else {
+		/* shrink the existing pending reservation: */
+		for (i = cur_nblks; i > nblks; i--) {
+			int blk = find_first_bit(ps->pending, cnt);
+			clear_bit(blk, ps->pending);
+			/* don't clear in global smp_state until _commit() */
+		}
+	}
+
+fail:
+	spin_unlock_irqrestore(&smp_lock, flags);
+	return 0;
+}
+
+static void update_smp_state(struct mdp5_kms *mdp5_kms,
+		enum mdp5_client_id cid, mdp5_smp_state_t *assigned)
+{
+	int cnt = mdp5_kms->smp_blk_cnt;
+	uint32_t blk, val;
+
+	for_each_set_bit(blk, *assigned, cnt) {
+		int idx = blk / 3;
+		int fld = blk % 3;
+
+		val = mdp5_read(mdp5_kms, REG_MDP5_SMP_ALLOC_W_REG(idx));
+
+		switch (fld) {
+		case 0:
+			val &= ~MDP5_SMP_ALLOC_W_REG_CLIENT0__MASK;
+			val |= MDP5_SMP_ALLOC_W_REG_CLIENT0(cid);
+			break;
+		case 1:
+			val &= ~MDP5_SMP_ALLOC_W_REG_CLIENT1__MASK;
+			val |= MDP5_SMP_ALLOC_W_REG_CLIENT1(cid);
+			break;
+		case 2:
+			val &= ~MDP5_SMP_ALLOC_W_REG_CLIENT2__MASK;
+			val |= MDP5_SMP_ALLOC_W_REG_CLIENT2(cid);
+			break;
+		}
+
+		mdp5_write(mdp5_kms, REG_MDP5_SMP_ALLOC_W_REG(idx), val);
+		mdp5_write(mdp5_kms, REG_MDP5_SMP_ALLOC_R_REG(idx), val);
+	}
+}
+
+/* step #2: configure hw for union(pending, inuse): */
+void mdp5_smp_configure(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid)
+{
+	struct mdp5_client_smp_state *ps = &mdp5_kms->smp_client_state[cid];
+	int cnt = mdp5_kms->smp_blk_cnt;
+	mdp5_smp_state_t assigned;
+
+	bitmap_or(assigned, ps->inuse, ps->pending, cnt);
+	update_smp_state(mdp5_kms, cid, &assigned);
+}
+
+/* step #3: after vblank, copy pending -> inuse: */
+void mdp5_smp_commit(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid)
+{
+	struct mdp5_client_smp_state *ps = &mdp5_kms->smp_client_state[cid];
+	int cnt = mdp5_kms->smp_blk_cnt;
+	mdp5_smp_state_t released;
+
+	/*
+	 * Figure out if there are any blocks we where previously
+	 * using, which can be released and made available to other
+	 * clients:
+	 */
+	if (bitmap_andnot(released, ps->inuse, ps->pending, cnt)) {
+		unsigned long flags;
+
+		spin_lock_irqsave(&smp_lock, flags);
+		/* clear released blocks: */
+		bitmap_andnot(mdp5_kms->smp_state, mdp5_kms->smp_state,
+				released, cnt);
+		spin_unlock_irqrestore(&smp_lock, flags);
+
+		update_smp_state(mdp5_kms, CID_UNUSED, &released);
+	}
+
+	bitmap_copy(ps->inuse, ps->pending, cnt);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h
new file mode 100644
index 00000000000..0ab739e1a1d
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h
@@ -0,0 +1,41 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __MDP5_SMP_H__
+#define __MDP5_SMP_H__
+
+#include "msm_drv.h"
+
+#define MAX_SMP_BLOCKS  22
+#define SMP_BLK_SIZE    4096
+#define SMP_ENTRIES_PER_BLK (SMP_BLK_SIZE / 16)
+
+typedef DECLARE_BITMAP(mdp5_smp_state_t, MAX_SMP_BLOCKS);
+
+struct mdp5_client_smp_state {
+	mdp5_smp_state_t inuse;
+	mdp5_smp_state_t pending;
+};
+
+struct mdp5_kms;
+
+int mdp5_smp_request(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid, int nblks);
+void mdp5_smp_configure(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid);
+void mdp5_smp_commit(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid);
+
+
+#endif /* __MDP5_SMP_H__ */
diff --git a/drivers/gpu/drm/msm/mdp/mdp_common.xml.h b/drivers/gpu/drm/msm/mdp/mdp_common.xml.h
new file mode 100644
index 00000000000..a9629b85b98
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp_common.xml.h
@@ -0,0 +1,78 @@
+#ifndef MDP_COMMON_XML
+#define MDP_COMMON_XML
+
+/* Autogenerated file, DO NOT EDIT manually!
+
+This file was generated by the rules-ng-ng headergen tool in this git repository:
+http://github.com/freedreno/envytools/
+git clone https://github.com/freedreno/envytools.git
+
+The rules-ng-ng source files this header was generated from are:
+- /home/robclark/src/freedreno/envytools/rnndb/msm.xml                 (    647 bytes, from 2013-11-30 14:45:35)
+- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp4.xml            (  17996 bytes, from 2013-12-01 19:10:31)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp_common.xml      (   1615 bytes, from 2013-11-30 15:00:52)
+- /home/robclark/src/freedreno/envytools/rnndb/mdp/mdp5.xml            (  22517 bytes, from 2013-12-03 20:59:13)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/dsi.xml             (  11712 bytes, from 2013-08-17 17:13:43)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/sfpb.xml            (    344 bytes, from 2013-08-11 19:26:32)
+- /home/robclark/src/freedreno/envytools/rnndb/dsi/mmss_cc.xml         (   1544 bytes, from 2013-08-16 19:17:05)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
+
+Copyright (C) 2013 by the following authors:
+- Rob Clark <robdclark@gmail.com> (robclark)
+
+Permission is hereby granted, free of charge, to any person obtaining
+a copy of this software and associated documentation files (the
+"Software"), to deal in the Software without restriction, including
+without limitation the rights to use, copy, modify, merge, publish,
+distribute, sublicense, and/or sell copies of the Software, and to
+permit persons to whom the Software is furnished to do so, subject to
+the following conditions:
+
+The above copyright notice and this permission notice (including the
+next paragraph) shall be included in all copies or substantial
+portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+IN NO EVENT SHALL THE COPYRIGHT OWNER(S) AND/OR ITS SUPPLIERS BE
+LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+*/
+
+
+enum mdp_mixer_stage_id {
+	STAGE_UNUSED = 0,
+	STAGE_BASE = 1,
+	STAGE0 = 2,
+	STAGE1 = 3,
+	STAGE2 = 4,
+	STAGE3 = 5,
+};
+
+enum mdp_alpha_type {
+	FG_CONST = 0,
+	BG_CONST = 1,
+	FG_PIXEL = 2,
+	BG_PIXEL = 3,
+};
+
+enum mdp_bpc {
+	BPC1 = 0,
+	BPC5 = 1,
+	BPC6 = 2,
+	BPC8 = 3,
+};
+
+enum mdp_bpc_alpha {
+	BPC1A = 0,
+	BPC4A = 1,
+	BPC6A = 2,
+	BPC8A = 3,
+};
+
+
+#endif /* MDP_COMMON_XML */
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_format.c b/drivers/gpu/drm/msm/mdp/mdp_format.c
index 17330b0927b..e0a6ffbe6ab 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_format.c
+++ b/drivers/gpu/drm/msm/mdp/mdp_format.c
@@ -17,7 +17,7 @@
 
 
 #include "msm_drv.h"
-#include "mdp4_kms.h"
+#include "mdp_kms.h"
 
 #define FMT(name, a, r, g, b, e0, e1, e2, e3, alpha, tight, c, cnt) { \
 		.base = { .pixel_format = DRM_FORMAT_ ## name }, \
@@ -34,7 +34,7 @@
 
 #define BPC0A 0
 
-static const struct mdp4_format formats[] = {
+static const struct mdp_format formats[] = {
 	/*  name      a  r  g  b   e0 e1 e2 e3  alpha   tight  cpp cnt */
 	FMT(ARGB8888, 8, 8, 8, 8,  1, 0, 2, 3,  true,   true,  4,  4),
 	FMT(XRGB8888, 8, 8, 8, 8,  1, 0, 2, 3,  false,  true,  4,  4),
@@ -44,12 +44,11 @@ static const struct mdp4_format formats[] = {
 	FMT(BGR565,   0, 5, 6, 5,  2, 0, 1, 0,  false,  true,  2,  3),
 };
 
-uint32_t mdp4_get_formats(enum mdp4_pipe pipe_id, uint32_t *pixel_formats,
-		uint32_t max_formats)
+uint32_t mdp_get_formats(uint32_t *pixel_formats, uint32_t max_formats)
 {
 	uint32_t i;
 	for (i = 0; i < ARRAY_SIZE(formats); i++) {
-		const struct mdp4_format *f = &formats[i];
+		const struct mdp_format *f = &formats[i];
 
 		if (i == max_formats)
 			break;
@@ -60,11 +59,11 @@ uint32_t mdp4_get_formats(enum mdp4_pipe pipe_id, uint32_t *pixel_formats,
 	return i;
 }
 
-const struct msm_format *mdp4_get_format(struct msm_kms *kms, uint32_t format)
+const struct msm_format *mdp_get_format(struct msm_kms *kms, uint32_t format)
 {
 	int i;
 	for (i = 0; i < ARRAY_SIZE(formats); i++) {
-		const struct mdp4_format *f = &formats[i];
+		const struct mdp_format *f = &formats[i];
 		if (f->base.pixel_format == format)
 			return &f->base;
 	}
diff --git a/drivers/gpu/drm/msm/mdp/mdp_kms.c b/drivers/gpu/drm/msm/mdp/mdp_kms.c
new file mode 100644
index 00000000000..03455b64a24
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp_kms.c
@@ -0,0 +1,145 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+
+#include "msm_drv.h"
+#include "mdp_kms.h"
+
+
+struct mdp_irq_wait {
+	struct mdp_irq irq;
+	int count;
+};
+
+static DECLARE_WAIT_QUEUE_HEAD(wait_event);
+
+static DEFINE_SPINLOCK(list_lock);
+
+static void update_irq(struct mdp_kms *mdp_kms)
+{
+	struct mdp_irq *irq;
+	uint32_t irqmask = mdp_kms->vblank_mask;
+
+	BUG_ON(!spin_is_locked(&list_lock));
+
+	list_for_each_entry(irq, &mdp_kms->irq_list, node)
+		irqmask |= irq->irqmask;
+
+	mdp_kms->funcs->set_irqmask(mdp_kms, irqmask);
+}
+
+static void update_irq_unlocked(struct mdp_kms *mdp_kms)
+{
+	unsigned long flags;
+	spin_lock_irqsave(&list_lock, flags);
+	update_irq(mdp_kms);
+	spin_unlock_irqrestore(&list_lock, flags);
+}
+
+void mdp_dispatch_irqs(struct mdp_kms *mdp_kms, uint32_t status)
+{
+	struct mdp_irq *handler, *n;
+	unsigned long flags;
+
+	spin_lock_irqsave(&list_lock, flags);
+	mdp_kms->in_irq = true;
+	list_for_each_entry_safe(handler, n, &mdp_kms->irq_list, node) {
+		if (handler->irqmask & status) {
+			spin_unlock_irqrestore(&list_lock, flags);
+			handler->irq(handler, handler->irqmask & status);
+			spin_lock_irqsave(&list_lock, flags);
+		}
+	}
+	mdp_kms->in_irq = false;
+	update_irq(mdp_kms);
+	spin_unlock_irqrestore(&list_lock, flags);
+
+}
+
+void mdp_update_vblank_mask(struct mdp_kms *mdp_kms, uint32_t mask, bool enable)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&list_lock, flags);
+	if (enable)
+		mdp_kms->vblank_mask |= mask;
+	else
+		mdp_kms->vblank_mask &= ~mask;
+	update_irq(mdp_kms);
+	spin_unlock_irqrestore(&list_lock, flags);
+}
+
+static void wait_irq(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	struct mdp_irq_wait *wait =
+			container_of(irq, struct mdp_irq_wait, irq);
+	wait->count--;
+	wake_up_all(&wait_event);
+}
+
+void mdp_irq_wait(struct mdp_kms *mdp_kms, uint32_t irqmask)
+{
+	struct mdp_irq_wait wait = {
+		.irq = {
+			.irq = wait_irq,
+			.irqmask = irqmask,
+		},
+		.count = 1,
+	};
+	mdp_irq_register(mdp_kms, &wait.irq);
+	wait_event_timeout(wait_event, (wait.count <= 0),
+			msecs_to_jiffies(100));
+	mdp_irq_unregister(mdp_kms, &wait.irq);
+}
+
+void mdp_irq_register(struct mdp_kms *mdp_kms, struct mdp_irq *irq)
+{
+	unsigned long flags;
+	bool needs_update = false;
+
+	spin_lock_irqsave(&list_lock, flags);
+
+	if (!irq->registered) {
+		irq->registered = true;
+		list_add(&irq->node, &mdp_kms->irq_list);
+		needs_update = !mdp_kms->in_irq;
+	}
+
+	spin_unlock_irqrestore(&list_lock, flags);
+
+	if (needs_update)
+		update_irq_unlocked(mdp_kms);
+}
+
+void mdp_irq_unregister(struct mdp_kms *mdp_kms, struct mdp_irq *irq)
+{
+	unsigned long flags;
+	bool needs_update = false;
+
+	spin_lock_irqsave(&list_lock, flags);
+
+	if (irq->registered) {
+		irq->registered = false;
+		list_del(&irq->node);
+		needs_update = !mdp_kms->in_irq;
+	}
+
+	spin_unlock_irqrestore(&list_lock, flags);
+
+	if (needs_update)
+		update_irq_unlocked(mdp_kms);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp_kms.h b/drivers/gpu/drm/msm/mdp/mdp_kms.h
new file mode 100644
index 00000000000..99557b5ad4f
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp_kms.h
@@ -0,0 +1,97 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __MDP_KMS_H__
+#define __MDP_KMS_H__
+
+#include <linux/clk.h>
+#include <linux/platform_device.h>
+#include <linux/regulator/consumer.h>
+
+#include "msm_drv.h"
+#include "msm_kms.h"
+#include "mdp_common.xml.h"
+
+struct mdp_kms;
+
+struct mdp_kms_funcs {
+	struct msm_kms_funcs base;
+	void (*set_irqmask)(struct mdp_kms *mdp_kms, uint32_t irqmask);
+};
+
+struct mdp_kms {
+	struct msm_kms base;
+
+	const struct mdp_kms_funcs *funcs;
+
+	/* irq handling: */
+	bool in_irq;
+	struct list_head irq_list;    /* list of mdp4_irq */
+	uint32_t vblank_mask;         /* irq bits set for userspace vblank */
+};
+#define to_mdp_kms(x) container_of(x, struct mdp_kms, base)
+
+static inline void mdp_kms_init(struct mdp_kms *mdp_kms,
+		const struct mdp_kms_funcs *funcs)
+{
+	mdp_kms->funcs = funcs;
+	INIT_LIST_HEAD(&mdp_kms->irq_list);
+	msm_kms_init(&mdp_kms->base, &funcs->base);
+}
+
+/*
+ * irq helpers:
+ */
+
+/* For transiently registering for different MDP irqs that various parts
+ * of the KMS code need during setup/configuration.  These are not
+ * necessarily the same as what drm_vblank_get/put() are requesting, and
+ * the hysteresis in drm_vblank_put() is not necessarily desirable for
+ * internal housekeeping related irq usage.
+ */
+struct mdp_irq {
+	struct list_head node;
+	uint32_t irqmask;
+	bool registered;
+	void (*irq)(struct mdp_irq *irq, uint32_t irqstatus);
+};
+
+void mdp_dispatch_irqs(struct mdp_kms *mdp_kms, uint32_t status);
+void mdp_update_vblank_mask(struct mdp_kms *mdp_kms, uint32_t mask, bool enable);
+void mdp_irq_wait(struct mdp_kms *mdp_kms, uint32_t irqmask);
+void mdp_irq_register(struct mdp_kms *mdp_kms, struct mdp_irq *irq);
+void mdp_irq_unregister(struct mdp_kms *mdp_kms, struct mdp_irq *irq);
+
+
+/*
+ * pixel format helpers:
+ */
+
+struct mdp_format {
+	struct msm_format base;
+	enum mdp_bpc bpc_r, bpc_g, bpc_b;
+	enum mdp_bpc_alpha bpc_a;
+	uint8_t unpack[4];
+	bool alpha_enable, unpack_tight;
+	uint8_t cpp, unpack_count;
+};
+#define to_mdp_format(x) container_of(x, struct mdp_format, base)
+
+uint32_t mdp_get_formats(uint32_t *formats, uint32_t max_formats);
+const struct msm_format *mdp_get_format(struct msm_kms *kms, uint32_t format);
+
+#endif /* __MDP_KMS_H__ */
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_irq.c b/drivers/gpu/drm/msm/mdp4/mdp4_irq.c
deleted file mode 100644
index 5c6b7fca4ed..00000000000
--- a/drivers/gpu/drm/msm/mdp4/mdp4_irq.c
+++ /dev/null
@@ -1,203 +0,0 @@
-/*
- * Copyright (C) 2013 Red Hat
- * Author: Rob Clark <robdclark@gmail.com>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
- */
-
-
-#include "msm_drv.h"
-#include "mdp4_kms.h"
-
-
-struct mdp4_irq_wait {
-	struct mdp4_irq irq;
-	int count;
-};
-
-static DECLARE_WAIT_QUEUE_HEAD(wait_event);
-
-static DEFINE_SPINLOCK(list_lock);
-
-static void update_irq(struct mdp4_kms *mdp4_kms)
-{
-	struct mdp4_irq *irq;
-	uint32_t irqmask = mdp4_kms->vblank_mask;
-
-	BUG_ON(!spin_is_locked(&list_lock));
-
-	list_for_each_entry(irq, &mdp4_kms->irq_list, node)
-		irqmask |= irq->irqmask;
-
-	mdp4_write(mdp4_kms, REG_MDP4_INTR_ENABLE, irqmask);
-}
-
-static void update_irq_unlocked(struct mdp4_kms *mdp4_kms)
-{
-	unsigned long flags;
-	spin_lock_irqsave(&list_lock, flags);
-	update_irq(mdp4_kms);
-	spin_unlock_irqrestore(&list_lock, flags);
-}
-
-static void mdp4_irq_error_handler(struct mdp4_irq *irq, uint32_t irqstatus)
-{
-	DRM_ERROR("errors: %08x\n", irqstatus);
-}
-
-void mdp4_irq_preinstall(struct msm_kms *kms)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	mdp4_write(mdp4_kms, REG_MDP4_INTR_CLEAR, 0xffffffff);
-}
-
-int mdp4_irq_postinstall(struct msm_kms *kms)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	struct mdp4_irq *error_handler = &mdp4_kms->error_handler;
-
-	INIT_LIST_HEAD(&mdp4_kms->irq_list);
-
-	error_handler->irq = mdp4_irq_error_handler;
-	error_handler->irqmask = MDP4_IRQ_PRIMARY_INTF_UDERRUN |
-			MDP4_IRQ_EXTERNAL_INTF_UDERRUN;
-
-	mdp4_irq_register(mdp4_kms, error_handler);
-
-	return 0;
-}
-
-void mdp4_irq_uninstall(struct msm_kms *kms)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	mdp4_write(mdp4_kms, REG_MDP4_INTR_ENABLE, 0x00000000);
-}
-
-irqreturn_t mdp4_irq(struct msm_kms *kms)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	struct drm_device *dev = mdp4_kms->dev;
-	struct msm_drm_private *priv = dev->dev_private;
-	struct mdp4_irq *handler, *n;
-	unsigned long flags;
-	unsigned int id;
-	uint32_t status;
-
-	status = mdp4_read(mdp4_kms, REG_MDP4_INTR_STATUS);
-	mdp4_write(mdp4_kms, REG_MDP4_INTR_CLEAR, status);
-
-	VERB("status=%08x", status);
-
-	for (id = 0; id < priv->num_crtcs; id++)
-		if (status & mdp4_crtc_vblank(priv->crtcs[id]))
-			drm_handle_vblank(dev, id);
-
-	spin_lock_irqsave(&list_lock, flags);
-	mdp4_kms->in_irq = true;
-	list_for_each_entry_safe(handler, n, &mdp4_kms->irq_list, node) {
-		if (handler->irqmask & status) {
-			spin_unlock_irqrestore(&list_lock, flags);
-			handler->irq(handler, handler->irqmask & status);
-			spin_lock_irqsave(&list_lock, flags);
-		}
-	}
-	mdp4_kms->in_irq = false;
-	update_irq(mdp4_kms);
-	spin_unlock_irqrestore(&list_lock, flags);
-
-	return IRQ_HANDLED;
-}
-
-int mdp4_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	unsigned long flags;
-
-	spin_lock_irqsave(&list_lock, flags);
-	mdp4_kms->vblank_mask |= mdp4_crtc_vblank(crtc);
-	update_irq(mdp4_kms);
-	spin_unlock_irqrestore(&list_lock, flags);
-
-	return 0;
-}
-
-void mdp4_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	unsigned long flags;
-
-	spin_lock_irqsave(&list_lock, flags);
-	mdp4_kms->vblank_mask &= ~mdp4_crtc_vblank(crtc);
-	update_irq(mdp4_kms);
-	spin_unlock_irqrestore(&list_lock, flags);
-}
-
-static void wait_irq(struct mdp4_irq *irq, uint32_t irqstatus)
-{
-	struct mdp4_irq_wait *wait =
-			container_of(irq, struct mdp4_irq_wait, irq);
-	wait->count--;
-	wake_up_all(&wait_event);
-}
-
-void mdp4_irq_wait(struct mdp4_kms *mdp4_kms, uint32_t irqmask)
-{
-	struct mdp4_irq_wait wait = {
-		.irq = {
-			.irq = wait_irq,
-			.irqmask = irqmask,
-		},
-		.count = 1,
-	};
-	mdp4_irq_register(mdp4_kms, &wait.irq);
-	wait_event(wait_event, (wait.count <= 0));
-	mdp4_irq_unregister(mdp4_kms, &wait.irq);
-}
-
-void mdp4_irq_register(struct mdp4_kms *mdp4_kms, struct mdp4_irq *irq)
-{
-	unsigned long flags;
-	bool needs_update = false;
-
-	spin_lock_irqsave(&list_lock, flags);
-
-	if (!irq->registered) {
-		irq->registered = true;
-		list_add(&irq->node, &mdp4_kms->irq_list);
-		needs_update = !mdp4_kms->in_irq;
-	}
-
-	spin_unlock_irqrestore(&list_lock, flags);
-
-	if (needs_update)
-		update_irq_unlocked(mdp4_kms);
-}
-
-void mdp4_irq_unregister(struct mdp4_kms *mdp4_kms, struct mdp4_irq *irq)
-{
-	unsigned long flags;
-	bool needs_update = false;
-
-	spin_lock_irqsave(&list_lock, flags);
-
-	if (irq->registered) {
-		irq->registered = false;
-		list_del(&irq->node);
-		needs_update = !mdp4_kms->in_irq;
-	}
-
-	spin_unlock_irqrestore(&list_lock, flags);
-
-	if (needs_update)
-		update_irq_unlocked(mdp4_kms);
-}
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
index 86537692e45..9a5d87db5c2 100644
--- a/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@ -17,6 +17,7 @@
 
 #include "msm_drv.h"
 #include "msm_gpu.h"
+#include "msm_kms.h"
 
 static void msm_fb_output_poll_changed(struct drm_device *dev)
 {
@@ -30,50 +31,19 @@ static const struct drm_mode_config_funcs mode_config_funcs = {
 	.output_poll_changed = msm_fb_output_poll_changed,
 };
 
-static int msm_fault_handler(struct iommu_domain *iommu, struct device *dev,
-		unsigned long iova, int flags, void *arg)
-{
-	DBG("*** fault: iova=%08lx, flags=%d", iova, flags);
-	return 0;
-}
-
-int msm_register_iommu(struct drm_device *dev, struct iommu_domain *iommu)
+int msm_register_mmu(struct drm_device *dev, struct msm_mmu *mmu)
 {
 	struct msm_drm_private *priv = dev->dev_private;
-	int idx = priv->num_iommus++;
+	int idx = priv->num_mmus++;
 
-	if (WARN_ON(idx >= ARRAY_SIZE(priv->iommus)))
+	if (WARN_ON(idx >= ARRAY_SIZE(priv->mmus)))
 		return -EINVAL;
 
-	priv->iommus[idx] = iommu;
-
-	iommu_set_fault_handler(iommu, msm_fault_handler, dev);
-
-	/* need to iommu_attach_device() somewhere??  on resume?? */
+	priv->mmus[idx] = mmu;
 
 	return idx;
 }
 
-int msm_iommu_attach(struct drm_device *dev, struct iommu_domain *iommu,
-		const char **names, int cnt)
-{
-	int i, ret;
-
-	for (i = 0; i < cnt; i++) {
-		/* TODO maybe some day msm iommu won't require this hack: */
-		struct device *msm_iommu_get_ctx(const char *ctx_name);
-		struct device *ctx = msm_iommu_get_ctx(names[i]);
-		if (!ctx)
-			continue;
-		ret = iommu_attach_device(iommu, ctx);
-		if (ret) {
-			dev_warn(dev->dev, "could not attach iommu to %s", names[i]);
-			return ret;
-		}
-	}
-	return 0;
-}
-
 #ifdef CONFIG_DRM_MSM_REGISTER_LOGGING
 static bool reglog = false;
 MODULE_PARM_DESC(reglog, "Enable register read/write logging");
@@ -82,6 +52,14 @@ module_param(reglog, bool, 0600);
 #define reglog 0
 #endif
 
+static char *vram;
+MODULE_PARM_DESC(vram, "Configure VRAM size (for devices without IOMMU/GPUMMU");
+module_param(vram, charp, 0);
+
+/*
+ * Util/helpers:
+ */
+
 void __iomem *msm_ioremap(struct platform_device *pdev, const char *name,
 		const char *dbgname)
 {
@@ -161,6 +139,16 @@ static int msm_unload(struct drm_device *dev)
 		mutex_unlock(&dev->struct_mutex);
 	}
 
+	if (priv->vram.paddr) {
+		DEFINE_DMA_ATTRS(attrs);
+		dma_set_attr(DMA_ATTR_NO_KERNEL_MAPPING, &attrs);
+		drm_mm_takedown(&priv->vram.mm);
+		dma_free_attrs(dev->dev, priv->vram.size, NULL,
+				priv->vram.paddr, &attrs);
+	}
+
+	component_unbind_all(dev->dev, dev);
+
 	dev->dev_private = NULL;
 
 	kfree(priv);
@@ -168,6 +156,24 @@ static int msm_unload(struct drm_device *dev)
 	return 0;
 }
 
+static int get_mdp_ver(struct platform_device *pdev)
+{
+#ifdef CONFIG_OF
+	static const struct of_device_id match_types[] = { {
+		.compatible = "qcom,mdss_mdp",
+		.data	= (void	*)5,
+	}, {
+		/* end node */
+	} };
+	struct device *dev = &pdev->dev;
+	const struct of_device_id *match;
+	match = of_match_node(match_types, dev->of_node);
+	if (match)
+		return (int)match->data;
+#endif
+	return 4;
+}
+
 static int msm_load(struct drm_device *dev, unsigned long flags)
 {
 	struct platform_device *pdev = dev->platformdev;
@@ -175,6 +181,7 @@ static int msm_load(struct drm_device *dev, unsigned long flags)
 	struct msm_kms *kms;
 	int ret;
 
+
 	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
 	if (!priv) {
 		dev_err(dev->dev, "failed to allocate private data\n");
@@ -191,7 +198,60 @@ static int msm_load(struct drm_device *dev, unsigned long flags)
 
 	drm_mode_config_init(dev);
 
-	kms = mdp4_kms_init(dev);
+	/* if we have no IOMMU, then we need to use carveout allocator.
+	 * Grab the entire CMA chunk carved out in early startup in
+	 * mach-msm:
+	 */
+	if (!iommu_present(&platform_bus_type)) {
+		DEFINE_DMA_ATTRS(attrs);
+		unsigned long size;
+		void *p;
+
+		DBG("using %s VRAM carveout", vram);
+		size = memparse(vram, NULL);
+		priv->vram.size = size;
+
+		drm_mm_init(&priv->vram.mm, 0, (size >> PAGE_SHIFT) - 1);
+
+		dma_set_attr(DMA_ATTR_NO_KERNEL_MAPPING, &attrs);
+		dma_set_attr(DMA_ATTR_WRITE_COMBINE, &attrs);
+
+		/* note that for no-kernel-mapping, the vaddr returned
+		 * is bogus, but non-null if allocation succeeded:
+		 */
+		p = dma_alloc_attrs(dev->dev, size,
+				&priv->vram.paddr, GFP_KERNEL, &attrs);
+		if (!p) {
+			dev_err(dev->dev, "failed to allocate VRAM\n");
+			priv->vram.paddr = 0;
+			ret = -ENOMEM;
+			goto fail;
+		}
+
+		dev_info(dev->dev, "VRAM: %08x->%08x\n",
+				(uint32_t)priv->vram.paddr,
+				(uint32_t)(priv->vram.paddr + size));
+	}
+
+	platform_set_drvdata(pdev, dev);
+
+	/* Bind all our sub-components: */
+	ret = component_bind_all(dev->dev, dev);
+	if (ret)
+		return ret;
+
+	switch (get_mdp_ver(pdev)) {
+	case 4:
+		kms = mdp4_kms_init(dev);
+		break;
+	case 5:
+		kms = mdp5_kms_init(dev);
+		break;
+	default:
+		kms = ERR_PTR(-ENODEV);
+		break;
+	}
+
 	if (IS_ERR(kms)) {
 		/*
 		 * NOTE: once we have GPU support, having no kms should not
@@ -228,19 +288,21 @@ static int msm_load(struct drm_device *dev, unsigned long flags)
 	}
 
 	pm_runtime_get_sync(dev->dev);
-	ret = drm_irq_install(dev);
+	ret = drm_irq_install(dev, platform_get_irq(dev->platformdev, 0));
 	pm_runtime_put_sync(dev->dev);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to install IRQ handler\n");
 		goto fail;
 	}
 
-	platform_set_drvdata(pdev, dev);
-
 #ifdef CONFIG_DRM_MSM_FBDEV
 	priv->fbdev = msm_fbdev_init(dev);
 #endif
 
+	ret = msm_debugfs_late_init(dev);
+	if (ret)
+		goto fail;
+
 	drm_kms_helper_poll_init(dev);
 
 	return 0;
@@ -265,7 +327,6 @@ static void load_gpu(struct drm_device *dev)
 		gpu = NULL;
 		/* not fatal */
 	}
-	mutex_unlock(&dev->struct_mutex);
 
 	if (gpu) {
 		int ret;
@@ -275,10 +336,16 @@ static void load_gpu(struct drm_device *dev)
 			dev_err(dev->dev, "gpu hw init failed: %d\n", ret);
 			gpu->funcs->destroy(gpu);
 			gpu = NULL;
+		} else {
+			/* give inactive pm a chance to kick in: */
+			msm_gpu_retire(gpu);
 		}
+
 	}
 
 	priv->gpu = gpu;
+
+	mutex_unlock(&dev->struct_mutex);
 }
 
 static int msm_open(struct drm_device *dev, struct drm_file *file)
@@ -319,14 +386,11 @@ static void msm_preclose(struct drm_device *dev, struct drm_file *file)
 static void msm_lastclose(struct drm_device *dev)
 {
 	struct msm_drm_private *priv = dev->dev_private;
-	if (priv->fbdev) {
-		drm_modeset_lock_all(dev);
-		drm_fb_helper_restore_fbdev_mode(priv->fbdev);
-		drm_modeset_unlock_all(dev);
-	}
+	if (priv->fbdev)
+		drm_fb_helper_restore_fbdev_mode_unlocked(priv->fbdev);
 }
 
-static irqreturn_t msm_irq(DRM_IRQ_ARGS)
+static irqreturn_t msm_irq(int irq, void *arg)
 {
 	struct drm_device *dev = arg;
 	struct msm_drm_private *priv = dev->dev_private;
@@ -415,7 +479,7 @@ static int msm_gem_show(struct drm_device *dev, struct seq_file *m)
 
 static int msm_mm_show(struct drm_device *dev, struct seq_file *m)
 {
-	return drm_mm_dump_table(m, dev->mm_private);
+	return drm_mm_dump_table(m, &dev->vma_offset_manager->vm_addr_space_mm);
 }
 
 static int msm_fb_show(struct drm_device *dev, struct seq_file *m)
@@ -468,6 +532,41 @@ static struct drm_info_list msm_debugfs_list[] = {
 		{ "fb", show_locked, 0, msm_fb_show },
 };
 
+static int late_init_minor(struct drm_minor *minor)
+{
+	int ret;
+
+	if (!minor)
+		return 0;
+
+	ret = msm_rd_debugfs_init(minor);
+	if (ret) {
+		dev_err(minor->dev->dev, "could not install rd debugfs\n");
+		return ret;
+	}
+
+	ret = msm_perf_debugfs_init(minor);
+	if (ret) {
+		dev_err(minor->dev->dev, "could not install perf debugfs\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+int msm_debugfs_late_init(struct drm_device *dev)
+{
+	int ret;
+	ret = late_init_minor(dev->primary);
+	if (ret)
+		return ret;
+	ret = late_init_minor(dev->render);
+	if (ret)
+		return ret;
+	ret = late_init_minor(dev->control);
+	return ret;
+}
+
 static int msm_debugfs_init(struct drm_minor *minor)
 {
 	struct drm_device *dev = minor->dev;
@@ -482,13 +581,17 @@ static int msm_debugfs_init(struct drm_minor *minor)
 		return ret;
 	}
 
-	return ret;
+	return 0;
 }
 
 static void msm_debugfs_cleanup(struct drm_minor *minor)
 {
 	drm_debugfs_remove_files(msm_debugfs_list,
 			ARRAY_SIZE(msm_debugfs_list), minor);
+	if (!minor->dev->dev_private)
+		return;
+	msm_rd_debugfs_cleanup(minor);
+	msm_perf_debugfs_cleanup(minor);
 }
 #endif
 
@@ -601,6 +704,12 @@ static int msm_ioctl_gem_new(struct drm_device *dev, void *data,
 		struct drm_file *file)
 {
 	struct drm_msm_gem_new *args = data;
+
+	if (args->flags & ~MSM_BO_FLAGS) {
+		DRM_ERROR("invalid flags: %08x\n", args->flags);
+		return -EINVAL;
+	}
+
 	return msm_gem_new_handle(dev, file, args->size,
 			args->flags, &args->handle);
 }
@@ -614,6 +723,11 @@ static int msm_ioctl_gem_cpu_prep(struct drm_device *dev, void *data,
 	struct drm_gem_object *obj;
 	int ret;
 
+	if (args->op & ~MSM_PREP_FLAGS) {
+		DRM_ERROR("invalid op: %08x\n", args->op);
+		return -EINVAL;
+	}
+
 	obj = drm_gem_object_lookup(dev, file, args->handle);
 	if (!obj)
 		return -ENOENT;
@@ -668,7 +782,14 @@ static int msm_ioctl_wait_fence(struct drm_device *dev, void *data,
 		struct drm_file *file)
 {
 	struct drm_msm_wait_fence *args = data;
-	return msm_wait_fence_interruptable(dev, args->fence, &TS(args->timeout));
+
+	if (args->pad) {
+		DRM_ERROR("invalid pad: %08x\n", args->pad);
+		return -EINVAL;
+	}
+
+	return msm_wait_fence_interruptable(dev, args->fence,
+			&TS(args->timeout));
 }
 
 static const struct drm_ioctl_desc msm_ioctls[] = {
@@ -773,17 +894,110 @@ static const struct dev_pm_ops msm_pm_ops = {
 };
 
 /*
+ * Componentized driver support:
+ */
+
+#ifdef CONFIG_OF
+/* NOTE: the CONFIG_OF case duplicates the same code as exynos or imx
+ * (or probably any other).. so probably some room for some helpers
+ */
+static int compare_of(struct device *dev, void *data)
+{
+	return dev->of_node == data;
+}
+
+static int msm_drm_add_components(struct device *master, struct master *m)
+{
+	struct device_node *np = master->of_node;
+	unsigned i;
+	int ret;
+
+	for (i = 0; ; i++) {
+		struct device_node *node;
+
+		node = of_parse_phandle(np, "connectors", i);
+		if (!node)
+			break;
+
+		ret = component_master_add_child(m, compare_of, node);
+		of_node_put(node);
+
+		if (ret)
+			return ret;
+	}
+	return 0;
+}
+#else
+static int compare_dev(struct device *dev, void *data)
+{
+	return dev == data;
+}
+
+static int msm_drm_add_components(struct device *master, struct master *m)
+{
+	/* For non-DT case, it kinda sucks.  We don't actually have a way
+	 * to know whether or not we are waiting for certain devices (or if
+	 * they are simply not present).  But for non-DT we only need to
+	 * care about apq8064/apq8060/etc (all mdp4/a3xx):
+	 */
+	static const char *devnames[] = {
+			"hdmi_msm.0", "kgsl-3d0.0",
+	};
+	int i;
+
+	DBG("Adding components..");
+
+	for (i = 0; i < ARRAY_SIZE(devnames); i++) {
+		struct device *dev;
+		int ret;
+
+		dev = bus_find_device_by_name(&platform_bus_type,
+				NULL, devnames[i]);
+		if (!dev) {
+			dev_info(master, "still waiting for %s\n", devnames[i]);
+			return -EPROBE_DEFER;
+		}
+
+		ret = component_master_add_child(m, compare_dev, dev);
+		if (ret) {
+			DBG("could not add child: %d", ret);
+			return ret;
+		}
+	}
+
+	return 0;
+}
+#endif
+
+static int msm_drm_bind(struct device *dev)
+{
+	return drm_platform_init(&msm_driver, to_platform_device(dev));
+}
+
+static void msm_drm_unbind(struct device *dev)
+{
+	drm_put_dev(platform_get_drvdata(to_platform_device(dev)));
+}
+
+static const struct component_master_ops msm_drm_ops = {
+		.add_components = msm_drm_add_components,
+		.bind = msm_drm_bind,
+		.unbind = msm_drm_unbind,
+};
+
+/*
  * Platform driver:
  */
 
 static int msm_pdev_probe(struct platform_device *pdev)
 {
-	return drm_platform_init(&msm_driver, pdev);
+	pdev->dev.coherent_dma_mask = DMA_BIT_MASK(32);
+	return component_master_add(&pdev->dev, &msm_drm_ops);
 }
 
 static int msm_pdev_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&msm_driver, pdev);
+	component_master_del(&pdev->dev, &msm_drm_ops);
 
 	return 0;
 }
@@ -793,12 +1007,19 @@ static const struct platform_device_id msm_id[] = {
 	{ }
 };
 
+static const struct of_device_id dt_match[] = {
+	{ .compatible = "qcom,mdss_mdp" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, dt_match);
+
 static struct platform_driver msm_platform_driver = {
 	.probe      = msm_pdev_probe,
 	.remove     = msm_pdev_remove,
 	.driver     = {
 		.owner  = THIS_MODULE,
 		.name   = "msm",
+		.of_match_table = dt_match,
 		.pm     = &msm_pm_ops,
 	},
 	.id_table   = msm_id,
diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h
index d39f0862b19..8a2c5fd0893 100644
--- a/drivers/gpu/drm/msm/msm_drv.h
+++ b/drivers/gpu/drm/msm/msm_drv.h
@@ -22,6 +22,7 @@
 #include <linux/clk.h>
 #include <linux/cpufreq.h>
 #include <linux/module.h>
+#include <linux/component.h>
 #include <linux/platform_device.h>
 #include <linux/pm.h>
 #include <linux/pm_runtime.h>
@@ -31,6 +32,15 @@
 #include <linux/types.h>
 #include <asm/sizes.h>
 
+
+#if defined(CONFIG_COMPILE_TEST) && !defined(CONFIG_ARCH_QCOM)
+/* stubs we need for compile-test: */
+static inline struct device *msm_iommu_get_ctx(const char *ctx_name)
+{
+	return NULL;
+}
+#endif
+
 #ifndef CONFIG_OF
 #include <mach/board.h>
 #include <mach/socinfo.h>
@@ -44,6 +54,10 @@
 
 struct msm_kms;
 struct msm_gpu;
+struct msm_mmu;
+struct msm_rd_state;
+struct msm_perf_state;
+struct msm_gem_submit;
 
 #define NUM_DOMAINS 2    /* one for KMS, then one per gpu core (?) */
 
@@ -59,6 +73,9 @@ struct msm_drm_private {
 
 	struct msm_kms *kms;
 
+	/* subordinate devices, if present: */
+	struct platform_device *hdmi_pdev, *gpu_pdev;
+
 	/* when we have more than one 'msm_gpu' these need to be an array: */
 	struct msm_gpu *gpu;
 	struct msm_file_private *lastctx;
@@ -68,6 +85,9 @@ struct msm_drm_private {
 	uint32_t next_fence, completed_fence;
 	wait_queue_head_t fence_event;
 
+	struct msm_rd_state *rd;
+	struct msm_perf_state *perf;
+
 	/* list of GEM objects: */
 	struct list_head inactive_list;
 
@@ -76,9 +96,9 @@ struct msm_drm_private {
 	/* callbacks deferred until bo is inactive: */
 	struct list_head fence_cbs;
 
-	/* registered IOMMU domains: */
-	unsigned int num_iommus;
-	struct iommu_domain *iommus[NUM_DOMAINS];
+	/* registered MMUs: */
+	unsigned int num_mmus;
+	struct msm_mmu *mmus[NUM_DOMAINS];
 
 	unsigned int num_planes;
 	struct drm_plane *planes[8];
@@ -94,6 +114,16 @@ struct msm_drm_private {
 
 	unsigned int num_connectors;
 	struct drm_connector *connectors[8];
+
+	/* VRAM carveout, used when no IOMMU: */
+	struct {
+		unsigned long size;
+		dma_addr_t paddr;
+		/* NOTE: mm managed at the page level, size is in # of pages
+		 * and position mm_node->start is in # of pages:
+		 */
+		struct drm_mm mm;
+	} vram;
 };
 
 struct msm_format {
@@ -114,39 +144,7 @@ void __msm_fence_worker(struct work_struct *work);
 		(_cb)->func = _func;                         \
 	} while (0)
 
-/* As there are different display controller blocks depending on the
- * snapdragon version, the kms support is split out and the appropriate
- * implementation is loaded at runtime.  The kms module is responsible
- * for constructing the appropriate planes/crtcs/encoders/connectors.
- */
-struct msm_kms_funcs {
-	/* hw initialization: */
-	int (*hw_init)(struct msm_kms *kms);
-	/* irq handling: */
-	void (*irq_preinstall)(struct msm_kms *kms);
-	int (*irq_postinstall)(struct msm_kms *kms);
-	void (*irq_uninstall)(struct msm_kms *kms);
-	irqreturn_t (*irq)(struct msm_kms *kms);
-	int (*enable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
-	void (*disable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
-	/* misc: */
-	const struct msm_format *(*get_format)(struct msm_kms *kms, uint32_t format);
-	long (*round_pixclk)(struct msm_kms *kms, unsigned long rate,
-			struct drm_encoder *encoder);
-	/* cleanup: */
-	void (*preclose)(struct msm_kms *kms, struct drm_file *file);
-	void (*destroy)(struct msm_kms *kms);
-};
-
-struct msm_kms {
-	const struct msm_kms_funcs *funcs;
-};
-
-struct msm_kms *mdp4_kms_init(struct drm_device *dev);
-
-int msm_register_iommu(struct drm_device *dev, struct iommu_domain *iommu);
-int msm_iommu_attach(struct drm_device *dev, struct iommu_domain *iommu,
-		const char **names, int cnt);
+int msm_register_mmu(struct drm_device *dev, struct msm_mmu *mmu);
 
 int msm_wait_fence_interruptable(struct drm_device *dev, uint32_t fence,
 		struct timespec *timeout);
@@ -202,7 +200,9 @@ struct drm_framebuffer *msm_framebuffer_create(struct drm_device *dev,
 
 struct drm_fb_helper *msm_fbdev_init(struct drm_device *dev);
 
-int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder);
+struct hdmi;
+struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder);
+irqreturn_t hdmi_irq(int irq, void *dev_id);
 void __init hdmi_register(void);
 void __exit hdmi_unregister(void);
 
@@ -210,6 +210,15 @@ void __exit hdmi_unregister(void);
 void msm_gem_describe(struct drm_gem_object *obj, struct seq_file *m);
 void msm_gem_describe_objects(struct list_head *list, struct seq_file *m);
 void msm_framebuffer_describe(struct drm_framebuffer *fb, struct seq_file *m);
+int msm_debugfs_late_init(struct drm_device *dev);
+int msm_rd_debugfs_init(struct drm_minor *minor);
+void msm_rd_debugfs_cleanup(struct drm_minor *minor);
+void msm_rd_dump_submit(struct msm_gem_submit *submit);
+int msm_perf_debugfs_init(struct drm_minor *minor);
+void msm_perf_debugfs_cleanup(struct drm_minor *minor);
+#else
+static inline int msm_debugfs_late_init(struct drm_device *dev) { return 0; }
+static inline void msm_rd_dump_submit(struct msm_gem_submit *submit) {}
 #endif
 
 void __iomem *msm_ioremap(struct platform_device *pdev, const char *name,
diff --git a/drivers/gpu/drm/msm/msm_fb.c b/drivers/gpu/drm/msm/msm_fb.c
index 0286c0eeb10..81bafdf19ab 100644
--- a/drivers/gpu/drm/msm/msm_fb.c
+++ b/drivers/gpu/drm/msm/msm_fb.c
@@ -16,6 +16,7 @@
  */
 
 #include "msm_drv.h"
+#include "msm_kms.h"
 
 #include "drm_crtc.h"
 #include "drm_crtc_helper.h"
diff --git a/drivers/gpu/drm/msm/msm_fbdev.c b/drivers/gpu/drm/msm/msm_fbdev.c
index 6c6d7d4c9b4..5107fc4826b 100644
--- a/drivers/gpu/drm/msm/msm_fbdev.c
+++ b/drivers/gpu/drm/msm/msm_fbdev.c
@@ -59,14 +59,11 @@ static int msm_fbdev_create(struct drm_fb_helper *helper,
 	struct drm_framebuffer *fb = NULL;
 	struct fb_info *fbi = NULL;
 	struct drm_mode_fb_cmd2 mode_cmd = {0};
-	dma_addr_t paddr;
+	uint32_t paddr;
 	int ret, size;
 
-	/* only doing ARGB32 since this is what is needed to alpha-blend
-	 * with video overlays:
-	 */
 	sizes->surface_bpp = 32;
-	sizes->surface_depth = 32;
+	sizes->surface_depth = 24;
 
 	DBG("create fbdev: %dx%d@%d (%dx%d)", sizes->surface_width,
 			sizes->surface_height, sizes->surface_bpp,
diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c
index e587d251c59..690d7e7b6d1 100644
--- a/drivers/gpu/drm/msm/msm_gem.c
+++ b/drivers/gpu/drm/msm/msm_gem.c
@@ -22,7 +22,45 @@
 #include "msm_drv.h"
 #include "msm_gem.h"
 #include "msm_gpu.h"
+#include "msm_mmu.h"
 
+static dma_addr_t physaddr(struct drm_gem_object *obj)
+{
+	struct msm_gem_object *msm_obj = to_msm_bo(obj);
+	struct msm_drm_private *priv = obj->dev->dev_private;
+	return (((dma_addr_t)msm_obj->vram_node->start) << PAGE_SHIFT) +
+			priv->vram.paddr;
+}
+
+/* allocate pages from VRAM carveout, used when no IOMMU: */
+static struct page **get_pages_vram(struct drm_gem_object *obj,
+		int npages)
+{
+	struct msm_gem_object *msm_obj = to_msm_bo(obj);
+	struct msm_drm_private *priv = obj->dev->dev_private;
+	dma_addr_t paddr;
+	struct page **p;
+	int ret, i;
+
+	p = drm_malloc_ab(npages, sizeof(struct page *));
+	if (!p)
+		return ERR_PTR(-ENOMEM);
+
+	ret = drm_mm_insert_node(&priv->vram.mm, msm_obj->vram_node,
+			npages, 0, DRM_MM_SEARCH_DEFAULT);
+	if (ret) {
+		drm_free_large(p);
+		return ERR_PTR(ret);
+	}
+
+	paddr = physaddr(obj);
+	for (i = 0; i < npages; i++) {
+		p[i] = phys_to_page(paddr);
+		paddr += PAGE_SIZE;
+	}
+
+	return p;
+}
 
 /* called with dev->struct_mutex held */
 static struct page **get_pages(struct drm_gem_object *obj)
@@ -31,9 +69,14 @@ static struct page **get_pages(struct drm_gem_object *obj)
 
 	if (!msm_obj->pages) {
 		struct drm_device *dev = obj->dev;
-		struct page **p = drm_gem_get_pages(obj, 0);
+		struct page **p;
 		int npages = obj->size >> PAGE_SHIFT;
 
+		if (iommu_present(&platform_bus_type))
+			p = drm_gem_get_pages(obj, 0);
+		else
+			p = get_pages_vram(obj, npages);
+
 		if (IS_ERR(p)) {
 			dev_err(dev->dev, "could not get pages: %ld\n",
 					PTR_ERR(p));
@@ -73,7 +116,13 @@ static void put_pages(struct drm_gem_object *obj)
 		sg_free_table(msm_obj->sgt);
 		kfree(msm_obj->sgt);
 
-		drm_gem_put_pages(obj, msm_obj->pages, true, false);
+		if (iommu_present(&platform_bus_type))
+			drm_gem_put_pages(obj, msm_obj->pages, true, false);
+		else {
+			drm_mm_remove_node(msm_obj->vram_node);
+			drm_free_large(msm_obj->pages);
+		}
+
 		msm_obj->pages = NULL;
 	}
 }
@@ -138,7 +187,6 @@ int msm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 int msm_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 {
 	struct drm_gem_object *obj = vma->vm_private_data;
-	struct msm_gem_object *msm_obj = to_msm_bo(obj);
 	struct drm_device *dev = obj->dev;
 	struct page **pages;
 	unsigned long pfn;
@@ -163,7 +211,7 @@ int msm_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	pgoff = ((unsigned long)vmf->virtual_address -
 			vma->vm_start) >> PAGE_SHIFT;
 
-	pfn = page_to_pfn(msm_obj->pages[pgoff]);
+	pfn = page_to_pfn(pages[pgoff]);
 
 	VERB("Inserting %p pfn %lx, pa %lx", vmf->virtual_address,
 			pfn, pfn << PAGE_SHIFT);
@@ -219,67 +267,6 @@ uint64_t msm_gem_mmap_offset(struct drm_gem_object *obj)
 	return offset;
 }
 
-/* helpers for dealing w/ iommu: */
-static int map_range(struct iommu_domain *domain, unsigned int iova,
-		struct sg_table *sgt, unsigned int len, int prot)
-{
-	struct scatterlist *sg;
-	unsigned int da = iova;
-	unsigned int i, j;
-	int ret;
-
-	if (!domain || !sgt)
-		return -EINVAL;
-
-	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
-		u32 pa = sg_phys(sg) - sg->offset;
-		size_t bytes = sg->length + sg->offset;
-
-		VERB("map[%d]: %08x %08x(%x)", i, iova, pa, bytes);
-
-		ret = iommu_map(domain, da, pa, bytes, prot);
-		if (ret)
-			goto fail;
-
-		da += bytes;
-	}
-
-	return 0;
-
-fail:
-	da = iova;
-
-	for_each_sg(sgt->sgl, sg, i, j) {
-		size_t bytes = sg->length + sg->offset;
-		iommu_unmap(domain, da, bytes);
-		da += bytes;
-	}
-	return ret;
-}
-
-static void unmap_range(struct iommu_domain *domain, unsigned int iova,
-		struct sg_table *sgt, unsigned int len)
-{
-	struct scatterlist *sg;
-	unsigned int da = iova;
-	int i;
-
-	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
-		size_t bytes = sg->length + sg->offset;
-		size_t unmapped;
-
-		unmapped = iommu_unmap(domain, da, bytes);
-		if (unmapped < bytes)
-			break;
-
-		VERB("unmap[%d]: %08x(%x)", i, iova, bytes);
-
-		BUG_ON(!IS_ALIGNED(bytes, PAGE_SIZE));
-
-		da += bytes;
-	}
-}
-
 /* should be called under struct_mutex.. although it can be called
  * from atomic context without struct_mutex to acquire an extra
  * iova ref if you know one is already held.
@@ -291,19 +278,30 @@ int msm_gem_get_iova_locked(struct drm_gem_object *obj, int id,
 		uint32_t *iova)
 {
 	struct msm_gem_object *msm_obj = to_msm_bo(obj);
+	struct drm_device *dev = obj->dev;
 	int ret = 0;
 
 	if (!msm_obj->domain[id].iova) {
 		struct msm_drm_private *priv = obj->dev->dev_private;
-		uint32_t offset = (uint32_t)mmap_offset(obj);
-		struct page **pages;
-		pages = get_pages(obj);
+		struct msm_mmu *mmu = priv->mmus[id];
+		struct page **pages = get_pages(obj);
+
+		if (!mmu) {
+			dev_err(dev->dev, "null MMU pointer\n");
+			return -EINVAL;
+		}
+
 		if (IS_ERR(pages))
 			return PTR_ERR(pages);
-		// XXX ideally we would not map buffers writable when not needed...
-		ret = map_range(priv->iommus[id], offset, msm_obj->sgt,
-				obj->size, IOMMU_READ | IOMMU_WRITE);
-		msm_obj->domain[id].iova = offset;
+
+		if (iommu_present(&platform_bus_type)) {
+			uint32_t offset = (uint32_t)mmap_offset(obj);
+			ret = mmu->funcs->map(mmu, offset, msm_obj->sgt,
+					obj->size, IOMMU_READ | IOMMU_WRITE);
+			msm_obj->domain[id].iova = offset;
+		} else {
+			msm_obj->domain[id].iova = physaddr(obj);
+		}
 	}
 
 	if (!ret)
@@ -514,6 +512,7 @@ void msm_gem_describe_objects(struct list_head *list, struct seq_file *m)
 void msm_gem_free_object(struct drm_gem_object *obj)
 {
 	struct drm_device *dev = obj->dev;
+	struct msm_drm_private *priv = obj->dev->dev_private;
 	struct msm_gem_object *msm_obj = to_msm_bo(obj);
 	int id;
 
@@ -525,11 +524,10 @@ void msm_gem_free_object(struct drm_gem_object *obj)
 	list_del(&msm_obj->mm_list);
 
 	for (id = 0; id < ARRAY_SIZE(msm_obj->domain); id++) {
-		if (msm_obj->domain[id].iova) {
-			struct msm_drm_private *priv = obj->dev->dev_private;
+		struct msm_mmu *mmu = priv->mmus[id];
+		if (mmu && msm_obj->domain[id].iova) {
 			uint32_t offset = (uint32_t)mmap_offset(obj);
-			unmap_range(priv->iommus[id], offset,
-					msm_obj->sgt, obj->size);
+			mmu->funcs->unmap(mmu, offset, msm_obj->sgt, obj->size);
 		}
 	}
 
@@ -591,6 +589,7 @@ static int msm_gem_new_impl(struct drm_device *dev,
 {
 	struct msm_drm_private *priv = dev->dev_private;
 	struct msm_gem_object *msm_obj;
+	unsigned sz;
 
 	switch (flags & MSM_BO_CACHE_MASK) {
 	case MSM_BO_UNCACHED:
@@ -603,10 +602,17 @@ static int msm_gem_new_impl(struct drm_device *dev,
 		return -EINVAL;
 	}
 
-	msm_obj = kzalloc(sizeof(*msm_obj), GFP_KERNEL);
+	sz = sizeof(*msm_obj);
+	if (!iommu_present(&platform_bus_type))
+		sz += sizeof(struct drm_mm_node);
+
+	msm_obj = kzalloc(sz, GFP_KERNEL);
 	if (!msm_obj)
 		return -ENOMEM;
 
+	if (!iommu_present(&platform_bus_type))
+		msm_obj->vram_node = (void *)&msm_obj[1];
+
 	msm_obj->flags = flags;
 
 	msm_obj->resv = &msm_obj->_resv;
@@ -623,7 +629,7 @@ static int msm_gem_new_impl(struct drm_device *dev,
 struct drm_gem_object *msm_gem_new(struct drm_device *dev,
 		uint32_t size, uint32_t flags)
 {
-	struct drm_gem_object *obj;
+	struct drm_gem_object *obj = NULL;
 	int ret;
 
 	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
@@ -634,15 +640,19 @@ struct drm_gem_object *msm_gem_new(struct drm_device *dev,
 	if (ret)
 		goto fail;
 
-	ret = drm_gem_object_init(dev, obj, size);
-	if (ret)
-		goto fail;
+	if (iommu_present(&platform_bus_type)) {
+		ret = drm_gem_object_init(dev, obj, size);
+		if (ret)
+			goto fail;
+	} else {
+		drm_gem_private_object_init(dev, obj, size);
+	}
 
 	return obj;
 
 fail:
 	if (obj)
-		drm_gem_object_unreference_unlocked(obj);
+		drm_gem_object_unreference(obj);
 
 	return ERR_PTR(ret);
 }
@@ -654,6 +664,12 @@ struct drm_gem_object *msm_gem_import(struct drm_device *dev,
 	struct drm_gem_object *obj;
 	int ret, npages;
 
+	/* if we don't have IOMMU, don't bother pretending we can import: */
+	if (!iommu_present(&platform_bus_type)) {
+		dev_err(dev->dev, "cannot import without IOMMU\n");
+		return ERR_PTR(-EINVAL);
+	}
+
 	size = PAGE_ALIGN(size);
 
 	ret = msm_gem_new_impl(dev, size, MSM_BO_WC, &obj);
diff --git a/drivers/gpu/drm/msm/msm_gem.h b/drivers/gpu/drm/msm/msm_gem.h
index f4f23a578d9..bfb052688f8 100644
--- a/drivers/gpu/drm/msm/msm_gem.h
+++ b/drivers/gpu/drm/msm/msm_gem.h
@@ -57,6 +57,11 @@ struct msm_gem_object {
 	/* normally (resv == &_resv) except for imported bo's */
 	struct reservation_object *resv;
 	struct reservation_object _resv;
+
+	/* For physically contiguous buffers.  Used when we don't have
+	 * an IOMMU.
+	 */
+	struct drm_mm_node *vram_node;
 };
 #define to_msm_bo(x) container_of(x, struct msm_gem_object, base)
 
@@ -85,6 +90,7 @@ struct msm_gem_submit {
 		uint32_t type;
 		uint32_t size;  /* in dwords */
 		uint32_t iova;
+		uint32_t idx;   /* cmdstream buffer idx in bos[] */
 	} cmd[MAX_CMDS];
 	struct {
 		uint32_t flags;
diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c
index 5281d4bc37f..cd0554f6831 100644
--- a/drivers/gpu/drm/msm/msm_gem_submit.c
+++ b/drivers/gpu/drm/msm/msm_gem_submit.c
@@ -23,7 +23,6 @@
  * Cmdstream submission:
  */
 
-#define BO_INVALID_FLAGS ~(MSM_SUBMIT_BO_READ | MSM_SUBMIT_BO_WRITE)
 /* make sure these don't conflict w/ MSM_SUBMIT_BO_x */
 #define BO_VALID    0x8000
 #define BO_LOCKED   0x4000
@@ -77,7 +76,7 @@ static int submit_lookup_objects(struct msm_gem_submit *submit,
 			goto out_unlock;
 		}
 
-		if (submit_bo.flags & BO_INVALID_FLAGS) {
+		if (submit_bo.flags & ~MSM_SUBMIT_BO_FLAGS) {
 			DRM_ERROR("invalid flags: %x\n", submit_bo.flags);
 			ret = -EINVAL;
 			goto out_unlock;
@@ -163,7 +162,7 @@ retry:
 
 
 		/* if locking succeeded, pin bo: */
-		ret = msm_gem_get_iova(&msm_obj->base,
+		ret = msm_gem_get_iova_locked(&msm_obj->base,
 				submit->gpu->id, &iova);
 
 		/* this would break the logic in the fail path.. there is no
@@ -247,7 +246,7 @@ static int submit_reloc(struct msm_gem_submit *submit, struct msm_gem_object *ob
 	/* For now, just map the entire thing.  Eventually we probably
 	 * to do it page-by-page, w/ kmap() if not vmap()d..
 	 */
-	ptr = msm_gem_vaddr(&obj->base);
+	ptr = msm_gem_vaddr_locked(&obj->base);
 
 	if (IS_ERR(ptr)) {
 		ret = PTR_ERR(ptr);
@@ -307,14 +306,12 @@ static void submit_cleanup(struct msm_gem_submit *submit, bool fail)
 {
 	unsigned i;
 
-	mutex_lock(&submit->dev->struct_mutex);
 	for (i = 0; i < submit->nr_bos; i++) {
 		struct msm_gem_object *msm_obj = submit->bos[i].obj;
 		submit_unlock_unpin_bo(submit, i);
 		list_del_init(&msm_obj->submit_entry);
 		drm_gem_object_unreference(&msm_obj->base);
 	}
-	mutex_unlock(&submit->dev->struct_mutex);
 
 	ww_acquire_fini(&submit->ticket);
 	kfree(submit);
@@ -342,6 +339,8 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 	if (args->nr_cmds > MAX_CMDS)
 		return -EINVAL;
 
+	mutex_lock(&dev->struct_mutex);
+
 	submit = submit_create(dev, gpu, args->nr_bos);
 	if (!submit) {
 		ret = -ENOMEM;
@@ -369,6 +368,18 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 			goto out;
 		}
 
+		/* validate input from userspace: */
+		switch (submit_cmd.type) {
+		case MSM_SUBMIT_CMD_BUF:
+		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
+		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
+			break;
+		default:
+			DRM_ERROR("invalid type: %08x\n", submit_cmd.type);
+			ret = -EINVAL;
+			goto out;
+		}
+
 		ret = submit_bo(submit, submit_cmd.submit_idx,
 				&msm_obj, &iova, NULL);
 		if (ret)
@@ -391,6 +402,7 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 		submit->cmd[i].type = submit_cmd.type;
 		submit->cmd[i].size = submit_cmd.size / 4;
 		submit->cmd[i].iova = iova + submit_cmd.submit_offset;
+		submit->cmd[i].idx  = submit_cmd.submit_idx;
 
 		if (submit->valid)
 			continue;
@@ -410,5 +422,6 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 out:
 	if (submit)
 		submit_cleanup(submit, !!ret);
+	mutex_unlock(&dev->struct_mutex);
 	return ret;
 }
diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c
index 4583d61556f..c6322197db8 100644
--- a/drivers/gpu/drm/msm/msm_gpu.c
+++ b/drivers/gpu/drm/msm/msm_gpu.c
@@ -17,6 +17,7 @@
 
 #include "msm_gpu.h"
 #include "msm_gem.h"
+#include "msm_mmu.h"
 
 
 /*
@@ -25,20 +26,10 @@
 
 #ifdef CONFIG_MSM_BUS_SCALING
 #include <mach/board.h>
-#include <mach/kgsl.h>
-static void bs_init(struct msm_gpu *gpu, struct platform_device *pdev)
+static void bs_init(struct msm_gpu *gpu)
 {
-	struct drm_device *dev = gpu->dev;
-	struct kgsl_device_platform_data *pdata;
-
-	if (!pdev) {
-		dev_err(dev->dev, "could not find dtv pdata\n");
-		return;
-	}
-
-	pdata = pdev->dev.platform_data;
-	if (pdata->bus_scale_table) {
-		gpu->bsc = msm_bus_scale_register_client(pdata->bus_scale_table);
+	if (gpu->bus_scale_table) {
+		gpu->bsc = msm_bus_scale_register_client(gpu->bus_scale_table);
 		DBG("bus scale client: %08x", gpu->bsc);
 	}
 }
@@ -59,7 +50,7 @@ static void bs_set(struct msm_gpu *gpu, int idx)
 	}
 }
 #else
-static void bs_init(struct msm_gpu *gpu, struct platform_device *pdev) {}
+static void bs_init(struct msm_gpu *gpu) {}
 static void bs_fini(struct msm_gpu *gpu) {}
 static void bs_set(struct msm_gpu *gpu, int idx) {}
 #endif
@@ -163,9 +154,18 @@ static int disable_axi(struct msm_gpu *gpu)
 
 int msm_gpu_pm_resume(struct msm_gpu *gpu)
 {
+	struct drm_device *dev = gpu->dev;
 	int ret;
 
-	DBG("%s", gpu->name);
+	DBG("%s: active_cnt=%d", gpu->name, gpu->active_cnt);
+
+	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
+
+	if (gpu->active_cnt++ > 0)
+		return 0;
+
+	if (WARN_ON(gpu->active_cnt <= 0))
+		return -EINVAL;
 
 	ret = enable_pwrrail(gpu);
 	if (ret)
@@ -184,9 +184,18 @@ int msm_gpu_pm_resume(struct msm_gpu *gpu)
 
 int msm_gpu_pm_suspend(struct msm_gpu *gpu)
 {
+	struct drm_device *dev = gpu->dev;
 	int ret;
 
-	DBG("%s", gpu->name);
+	DBG("%s: active_cnt=%d", gpu->name, gpu->active_cnt);
+
+	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
+
+	if (--gpu->active_cnt > 0)
+		return 0;
+
+	if (WARN_ON(gpu->active_cnt < 0))
+		return -EINVAL;
 
 	ret = disable_axi(gpu);
 	if (ret)
@@ -204,6 +213,55 @@ int msm_gpu_pm_suspend(struct msm_gpu *gpu)
 }
 
 /*
+ * Inactivity detection (for suspend):
+ */
+
+static void inactive_worker(struct work_struct *work)
+{
+	struct msm_gpu *gpu = container_of(work, struct msm_gpu, inactive_work);
+	struct drm_device *dev = gpu->dev;
+
+	if (gpu->inactive)
+		return;
+
+	DBG("%s: inactive!\n", gpu->name);
+	mutex_lock(&dev->struct_mutex);
+	if (!(msm_gpu_active(gpu) || gpu->inactive)) {
+		disable_axi(gpu);
+		disable_clk(gpu);
+		gpu->inactive = true;
+	}
+	mutex_unlock(&dev->struct_mutex);
+}
+
+static void inactive_handler(unsigned long data)
+{
+	struct msm_gpu *gpu = (struct msm_gpu *)data;
+	struct msm_drm_private *priv = gpu->dev->dev_private;
+
+	queue_work(priv->wq, &gpu->inactive_work);
+}
+
+/* cancel inactive timer and make sure we are awake: */
+static void inactive_cancel(struct msm_gpu *gpu)
+{
+	DBG("%s", gpu->name);
+	del_timer(&gpu->inactive_timer);
+	if (gpu->inactive) {
+		enable_clk(gpu);
+		enable_axi(gpu);
+		gpu->inactive = false;
+	}
+}
+
+static void inactive_start(struct msm_gpu *gpu)
+{
+	DBG("%s", gpu->name);
+	mod_timer(&gpu->inactive_timer,
+			round_jiffies_up(jiffies + DRM_MSM_INACTIVE_JIFFIES));
+}
+
+/*
  * Hangcheck detection for locked gpu:
  */
 
@@ -215,7 +273,10 @@ static void recover_worker(struct work_struct *work)
 	dev_err(dev->dev, "%s: hangcheck recover!\n", gpu->name);
 
 	mutex_lock(&dev->struct_mutex);
-	gpu->funcs->recover(gpu);
+	if (msm_gpu_active(gpu)) {
+		inactive_cancel(gpu);
+		gpu->funcs->recover(gpu);
+	}
 	mutex_unlock(&dev->struct_mutex);
 
 	msm_gpu_retire(gpu);
@@ -259,6 +320,101 @@ static void hangcheck_handler(unsigned long data)
 }
 
 /*
+ * Performance Counters:
+ */
+
+/* called under perf_lock */
+static int update_hw_cntrs(struct msm_gpu *gpu, uint32_t ncntrs, uint32_t *cntrs)
+{
+	uint32_t current_cntrs[ARRAY_SIZE(gpu->last_cntrs)];
+	int i, n = min(ncntrs, gpu->num_perfcntrs);
+
+	/* read current values: */
+	for (i = 0; i < gpu->num_perfcntrs; i++)
+		current_cntrs[i] = gpu_read(gpu, gpu->perfcntrs[i].sample_reg);
+
+	/* update cntrs: */
+	for (i = 0; i < n; i++)
+		cntrs[i] = current_cntrs[i] - gpu->last_cntrs[i];
+
+	/* save current values: */
+	for (i = 0; i < gpu->num_perfcntrs; i++)
+		gpu->last_cntrs[i] = current_cntrs[i];
+
+	return n;
+}
+
+static void update_sw_cntrs(struct msm_gpu *gpu)
+{
+	ktime_t time;
+	uint32_t elapsed;
+	unsigned long flags;
+
+	spin_lock_irqsave(&gpu->perf_lock, flags);
+	if (!gpu->perfcntr_active)
+		goto out;
+
+	time = ktime_get();
+	elapsed = ktime_to_us(ktime_sub(time, gpu->last_sample.time));
+
+	gpu->totaltime += elapsed;
+	if (gpu->last_sample.active)
+		gpu->activetime += elapsed;
+
+	gpu->last_sample.active = msm_gpu_active(gpu);
+	gpu->last_sample.time = time;
+
+out:
+	spin_unlock_irqrestore(&gpu->perf_lock, flags);
+}
+
+void msm_gpu_perfcntr_start(struct msm_gpu *gpu)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&gpu->perf_lock, flags);
+	/* we could dynamically enable/disable perfcntr registers too.. */
+	gpu->last_sample.active = msm_gpu_active(gpu);
+	gpu->last_sample.time = ktime_get();
+	gpu->activetime = gpu->totaltime = 0;
+	gpu->perfcntr_active = true;
+	update_hw_cntrs(gpu, 0, NULL);
+	spin_unlock_irqrestore(&gpu->perf_lock, flags);
+}
+
+void msm_gpu_perfcntr_stop(struct msm_gpu *gpu)
+{
+	gpu->perfcntr_active = false;
+}
+
+/* returns -errno or # of cntrs sampled */
+int msm_gpu_perfcntr_sample(struct msm_gpu *gpu, uint32_t *activetime,
+		uint32_t *totaltime, uint32_t ncntrs, uint32_t *cntrs)
+{
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&gpu->perf_lock, flags);
+
+	if (!gpu->perfcntr_active) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	*activetime = gpu->activetime;
+	*totaltime = gpu->totaltime;
+
+	gpu->activetime = gpu->totaltime = 0;
+
+	ret = update_hw_cntrs(gpu, ncntrs, cntrs);
+
+out:
+	spin_unlock_irqrestore(&gpu->perf_lock, flags);
+
+	return ret;
+}
+
+/*
  * Cmdstream submission/retirement:
  */
 
@@ -290,6 +446,9 @@ static void retire_worker(struct work_struct *work)
 	}
 
 	mutex_unlock(&dev->struct_mutex);
+
+	if (!msm_gpu_active(gpu))
+		inactive_start(gpu);
 }
 
 /* call from irq handler to schedule work to retire bo's */
@@ -297,6 +456,7 @@ void msm_gpu_retire(struct msm_gpu *gpu)
 {
 	struct msm_drm_private *priv = gpu->dev->dev_private;
 	queue_work(priv->wq, &gpu->retire_work);
+	update_sw_cntrs(gpu);
 }
 
 /* add bo's to gpu's ring, and kick gpu: */
@@ -307,12 +467,18 @@ int msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 	struct msm_drm_private *priv = dev->dev_private;
 	int i, ret;
 
-	mutex_lock(&dev->struct_mutex);
-
 	submit->fence = ++priv->next_fence;
 
 	gpu->submitted_fence = submit->fence;
 
+	inactive_cancel(gpu);
+
+	msm_rd_dump_submit(submit);
+
+	gpu->submitted_fence = submit->fence;
+
+	update_sw_cntrs(gpu);
+
 	ret = gpu->funcs->submit(gpu, submit, ctx);
 	priv->lastctx = ctx;
 
@@ -340,7 +506,6 @@ int msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 			msm_gem_move_to_active(&msm_obj->base, gpu, true, submit->fence);
 	}
 	hangcheck_timer_reset(gpu);
-	mutex_unlock(&dev->struct_mutex);
 
 	return ret;
 }
@@ -363,19 +528,29 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
 		const char *name, const char *ioname, const char *irqname, int ringsz)
 {
+	struct iommu_domain *iommu;
 	int i, ret;
 
+	if (WARN_ON(gpu->num_perfcntrs > ARRAY_SIZE(gpu->last_cntrs)))
+		gpu->num_perfcntrs = ARRAY_SIZE(gpu->last_cntrs);
+
 	gpu->dev = drm;
 	gpu->funcs = funcs;
 	gpu->name = name;
+	gpu->inactive = true;
 
 	INIT_LIST_HEAD(&gpu->active_list);
 	INIT_WORK(&gpu->retire_work, retire_worker);
+	INIT_WORK(&gpu->inactive_work, inactive_worker);
 	INIT_WORK(&gpu->recover_work, recover_worker);
 
+	setup_timer(&gpu->inactive_timer, inactive_handler,
+			(unsigned long)gpu);
 	setup_timer(&gpu->hangcheck_timer, hangcheck_handler,
 			(unsigned long)gpu);
 
+	spin_lock_init(&gpu->perf_lock);
+
 	BUG_ON(ARRAY_SIZE(clk_names) != ARRAY_SIZE(gpu->grp_clks));
 
 	/* Map registers: */
@@ -428,13 +603,14 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 	 * and have separate page tables per context.  For now, to keep things
 	 * simple and to get something working, just use a single address space:
 	 */
-	gpu->iommu = iommu_domain_alloc(&platform_bus_type);
-	if (!gpu->iommu) {
-		dev_err(drm->dev, "failed to allocate IOMMU\n");
-		ret = -ENOMEM;
-		goto fail;
+	iommu = iommu_domain_alloc(&platform_bus_type);
+	if (iommu) {
+		dev_info(drm->dev, "%s: using IOMMU\n", name);
+		gpu->mmu = msm_iommu_new(drm, iommu);
+	} else {
+		dev_info(drm->dev, "%s: no IOMMU, fallback to VRAM carveout!\n", name);
 	}
-	gpu->id = msm_register_iommu(drm, gpu->iommu);
+	gpu->id = msm_register_mmu(drm, gpu->mmu);
 
 	/* Create ringbuffer: */
 	gpu->rb = msm_ringbuffer_new(gpu, ringsz);
@@ -452,7 +628,7 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		goto fail;
 	}
 
-	bs_init(gpu, pdev);
+	bs_init(gpu);
 
 	return 0;
 
@@ -474,6 +650,6 @@ void msm_gpu_cleanup(struct msm_gpu *gpu)
 		msm_ringbuffer_destroy(gpu->rb);
 	}
 
-	if (gpu->iommu)
-		iommu_domain_free(gpu->iommu);
+	if (gpu->mmu)
+		gpu->mmu->funcs->destroy(gpu->mmu);
 }
diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h
index 8cd829e520b..9b579b79284 100644
--- a/drivers/gpu/drm/msm/msm_gpu.h
+++ b/drivers/gpu/drm/msm/msm_gpu.h
@@ -25,6 +25,7 @@
 #include "msm_ringbuffer.h"
 
 struct msm_gem_submit;
+struct msm_gpu_perfcntr;
 
 /* So far, with hardware that I've seen to date, we can have:
  *  + zero, one, or two z180 2d cores
@@ -64,6 +65,18 @@ struct msm_gpu {
 	struct drm_device *dev;
 	const struct msm_gpu_funcs *funcs;
 
+	/* performance counters (hw & sw): */
+	spinlock_t perf_lock;
+	bool perfcntr_active;
+	struct {
+		bool active;
+		ktime_t time;
+	} last_sample;
+	uint32_t totaltime, activetime;    /* sw counters */
+	uint32_t last_cntrs[5];            /* hw counters */
+	const struct msm_gpu_perfcntr *perfcntrs;
+	uint32_t num_perfcntrs;
+
 	struct msm_ringbuffer *rb;
 	uint32_t rb_iova;
 
@@ -72,22 +85,35 @@ struct msm_gpu {
 
 	uint32_t submitted_fence;
 
+	/* is gpu powered/active? */
+	int active_cnt;
+	bool inactive;
+
 	/* worker for handling active-list retiring: */
 	struct work_struct retire_work;
 
 	void __iomem *mmio;
 	int irq;
 
-	struct iommu_domain *iommu;
+	struct msm_mmu *mmu;
 	int id;
 
 	/* Power Control: */
 	struct regulator *gpu_reg, *gpu_cx;
 	struct clk *ebi1_clk, *grp_clks[5];
 	uint32_t fast_rate, slow_rate, bus_freq;
+
+#ifdef CONFIG_MSM_BUS_SCALING
+	struct msm_bus_scale_pdata *bus_scale_table;
 	uint32_t bsc;
+#endif
 
-	/* Hang Detction: */
+	/* Hang and Inactivity Detection:
+	 */
+#define DRM_MSM_INACTIVE_PERIOD   66 /* in ms (roughly four frames) */
+#define DRM_MSM_INACTIVE_JIFFIES  msecs_to_jiffies(DRM_MSM_INACTIVE_PERIOD)
+	struct timer_list inactive_timer;
+	struct work_struct inactive_work;
 #define DRM_MSM_HANGCHECK_PERIOD 500 /* in ms */
 #define DRM_MSM_HANGCHECK_JIFFIES msecs_to_jiffies(DRM_MSM_HANGCHECK_PERIOD)
 	struct timer_list hangcheck_timer;
@@ -95,6 +121,24 @@ struct msm_gpu {
 	struct work_struct recover_work;
 };
 
+static inline bool msm_gpu_active(struct msm_gpu *gpu)
+{
+	return gpu->submitted_fence > gpu->funcs->last_fence(gpu);
+}
+
+/* Perf-Counters:
+ * The select_reg and select_val are just there for the benefit of the child
+ * class that actually enables the perf counter..  but msm_gpu base class
+ * will handle sampling/displaying the counters.
+ */
+
+struct msm_gpu_perfcntr {
+	uint32_t select_reg;
+	uint32_t sample_reg;
+	uint32_t select_val;
+	const char *name;
+};
+
 static inline void gpu_write(struct msm_gpu *gpu, u32 reg, u32 data)
 {
 	msm_writel(data, gpu->mmio + (reg << 2));
@@ -108,6 +152,11 @@ static inline u32 gpu_read(struct msm_gpu *gpu, u32 reg)
 int msm_gpu_pm_suspend(struct msm_gpu *gpu);
 int msm_gpu_pm_resume(struct msm_gpu *gpu);
 
+void msm_gpu_perfcntr_start(struct msm_gpu *gpu);
+void msm_gpu_perfcntr_stop(struct msm_gpu *gpu);
+int msm_gpu_perfcntr_sample(struct msm_gpu *gpu, uint32_t *activetime,
+		uint32_t *totaltime, uint32_t ncntrs, uint32_t *cntrs);
+
 void msm_gpu_retire(struct msm_gpu *gpu);
 int msm_gpu_submit(struct msm_gpu *gpu, struct msm_gem_submit *submit,
 		struct msm_file_private *ctx);
diff --git a/drivers/gpu/drm/msm/msm_iommu.c b/drivers/gpu/drm/msm/msm_iommu.c
new file mode 100644
index 00000000000..4b2ad9181ed
--- /dev/null
+++ b/drivers/gpu/drm/msm/msm_iommu.c
@@ -0,0 +1,165 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "msm_drv.h"
+#include "msm_mmu.h"
+
+struct msm_iommu {
+	struct msm_mmu base;
+	struct iommu_domain *domain;
+};
+#define to_msm_iommu(x) container_of(x, struct msm_iommu, base)
+
+static int msm_fault_handler(struct iommu_domain *iommu, struct device *dev,
+		unsigned long iova, int flags, void *arg)
+{
+	DBG("*** fault: iova=%08lx, flags=%d", iova, flags);
+	return -ENOSYS;
+}
+
+static int msm_iommu_attach(struct msm_mmu *mmu, const char **names, int cnt)
+{
+	struct drm_device *dev = mmu->dev;
+	struct msm_iommu *iommu = to_msm_iommu(mmu);
+	int i, ret;
+
+	for (i = 0; i < cnt; i++) {
+		struct device *msm_iommu_get_ctx(const char *ctx_name);
+		struct device *ctx = msm_iommu_get_ctx(names[i]);
+		if (IS_ERR_OR_NULL(ctx)) {
+			dev_warn(dev->dev, "couldn't get %s context", names[i]);
+			continue;
+		}
+		ret = iommu_attach_device(iommu->domain, ctx);
+		if (ret) {
+			dev_warn(dev->dev, "could not attach iommu to %s", names[i]);
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+static void msm_iommu_detach(struct msm_mmu *mmu, const char **names, int cnt)
+{
+	struct msm_iommu *iommu = to_msm_iommu(mmu);
+	int i;
+
+	for (i = 0; i < cnt; i++) {
+		struct device *msm_iommu_get_ctx(const char *ctx_name);
+		struct device *ctx = msm_iommu_get_ctx(names[i]);
+		if (IS_ERR_OR_NULL(ctx))
+			continue;
+		iommu_detach_device(iommu->domain, ctx);
+	}
+}
+
+static int msm_iommu_map(struct msm_mmu *mmu, uint32_t iova,
+		struct sg_table *sgt, unsigned len, int prot)
+{
+	struct msm_iommu *iommu = to_msm_iommu(mmu);
+	struct iommu_domain *domain = iommu->domain;
+	struct scatterlist *sg;
+	unsigned int da = iova;
+	unsigned int i, j;
+	int ret;
+
+	if (!domain || !sgt)
+		return -EINVAL;
+
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
+		u32 pa = sg_phys(sg) - sg->offset;
+		size_t bytes = sg->length + sg->offset;
+
+		VERB("map[%d]: %08x %08x(%x)", i, iova, pa, bytes);
+
+		ret = iommu_map(domain, da, pa, bytes, prot);
+		if (ret)
+			goto fail;
+
+		da += bytes;
+	}
+
+	return 0;
+
+fail:
+	da = iova;
+
+	for_each_sg(sgt->sgl, sg, i, j) {
+		size_t bytes = sg->length + sg->offset;
+		iommu_unmap(domain, da, bytes);
+		da += bytes;
+	}
+	return ret;
+}
+
+static int msm_iommu_unmap(struct msm_mmu *mmu, uint32_t iova,
+		struct sg_table *sgt, unsigned len)
+{
+	struct msm_iommu *iommu = to_msm_iommu(mmu);
+	struct iommu_domain *domain = iommu->domain;
+	struct scatterlist *sg;
+	unsigned int da = iova;
+	int i;
+
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
+		size_t bytes = sg->length + sg->offset;
+		size_t unmapped;
+
+		unmapped = iommu_unmap(domain, da, bytes);
+		if (unmapped < bytes)
+			return unmapped;
+
+		VERB("unmap[%d]: %08x(%x)", i, iova, bytes);
+
+		BUG_ON(!PAGE_ALIGNED(bytes));
+
+		da += bytes;
+	}
+
+	return 0;
+}
+
+static void msm_iommu_destroy(struct msm_mmu *mmu)
+{
+	struct msm_iommu *iommu = to_msm_iommu(mmu);
+	iommu_domain_free(iommu->domain);
+	kfree(iommu);
+}
+
+static const struct msm_mmu_funcs funcs = {
+		.attach = msm_iommu_attach,
+		.detach = msm_iommu_detach,
+		.map = msm_iommu_map,
+		.unmap = msm_iommu_unmap,
+		.destroy = msm_iommu_destroy,
+};
+
+struct msm_mmu *msm_iommu_new(struct drm_device *dev, struct iommu_domain *domain)
+{
+	struct msm_iommu *iommu;
+
+	iommu = kzalloc(sizeof(*iommu), GFP_KERNEL);
+	if (!iommu)
+		return ERR_PTR(-ENOMEM);
+
+	iommu->domain = domain;
+	msm_mmu_init(&iommu->base, dev, &funcs);
+	iommu_set_fault_handler(domain, msm_fault_handler, dev);
+
+	return &iommu->base;
+}
diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
new file mode 100644
index 00000000000..06437745bc2
--- /dev/null
+++ b/drivers/gpu/drm/msm/msm_kms.h
@@ -0,0 +1,68 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __MSM_KMS_H__
+#define __MSM_KMS_H__
+
+#include <linux/clk.h>
+#include <linux/regulator/consumer.h>
+
+#include "msm_drv.h"
+
+/* As there are different display controller blocks depending on the
+ * snapdragon version, the kms support is split out and the appropriate
+ * implementation is loaded at runtime.  The kms module is responsible
+ * for constructing the appropriate planes/crtcs/encoders/connectors.
+ */
+struct msm_kms_funcs {
+	/* hw initialization: */
+	int (*hw_init)(struct msm_kms *kms);
+	/* irq handling: */
+	void (*irq_preinstall)(struct msm_kms *kms);
+	int (*irq_postinstall)(struct msm_kms *kms);
+	void (*irq_uninstall)(struct msm_kms *kms);
+	irqreturn_t (*irq)(struct msm_kms *kms);
+	int (*enable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
+	void (*disable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
+	/* misc: */
+	const struct msm_format *(*get_format)(struct msm_kms *kms, uint32_t format);
+	long (*round_pixclk)(struct msm_kms *kms, unsigned long rate,
+			struct drm_encoder *encoder);
+	/* cleanup: */
+	void (*preclose)(struct msm_kms *kms, struct drm_file *file);
+	void (*destroy)(struct msm_kms *kms);
+};
+
+struct msm_kms {
+	const struct msm_kms_funcs *funcs;
+
+	/* irq handling: */
+	bool in_irq;
+	struct list_head irq_list;    /* list of mdp4_irq */
+	uint32_t vblank_mask;         /* irq bits set for userspace vblank */
+};
+
+static inline void msm_kms_init(struct msm_kms *kms,
+		const struct msm_kms_funcs *funcs)
+{
+	kms->funcs = funcs;
+}
+
+struct msm_kms *mdp4_kms_init(struct drm_device *dev);
+struct msm_kms *mdp5_kms_init(struct drm_device *dev);
+
+#endif /* __MSM_KMS_H__ */
diff --git a/drivers/gpu/drm/msm/msm_mmu.h b/drivers/gpu/drm/msm/msm_mmu.h
new file mode 100644
index 00000000000..21da6d154f7
--- /dev/null
+++ b/drivers/gpu/drm/msm/msm_mmu.h
@@ -0,0 +1,48 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef __MSM_MMU_H__
+#define __MSM_MMU_H__
+
+#include <linux/iommu.h>
+
+struct msm_mmu_funcs {
+	int (*attach)(struct msm_mmu *mmu, const char **names, int cnt);
+	void (*detach)(struct msm_mmu *mmu, const char **names, int cnt);
+	int (*map)(struct msm_mmu *mmu, uint32_t iova, struct sg_table *sgt,
+			unsigned len, int prot);
+	int (*unmap)(struct msm_mmu *mmu, uint32_t iova, struct sg_table *sgt,
+			unsigned len);
+	void (*destroy)(struct msm_mmu *mmu);
+};
+
+struct msm_mmu {
+	const struct msm_mmu_funcs *funcs;
+	struct drm_device *dev;
+};
+
+static inline void msm_mmu_init(struct msm_mmu *mmu, struct drm_device *dev,
+		const struct msm_mmu_funcs *funcs)
+{
+	mmu->dev = dev;
+	mmu->funcs = funcs;
+}
+
+struct msm_mmu *msm_iommu_new(struct drm_device *dev, struct iommu_domain *domain);
+struct msm_mmu *msm_gpummu_new(struct drm_device *dev, struct msm_gpu *gpu);
+
+#endif /* __MSM_MMU_H__ */
diff --git a/drivers/gpu/drm/msm/msm_perf.c b/drivers/gpu/drm/msm/msm_perf.c
new file mode 100644
index 00000000000..830857c47c8
--- /dev/null
+++ b/drivers/gpu/drm/msm/msm_perf.c
@@ -0,0 +1,275 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+/* For profiling, userspace can:
+ *
+ *   tail -f /sys/kernel/debug/dri/<minor>/gpu
+ *
+ * This will enable performance counters/profiling to track the busy time
+ * and any gpu specific performance counters that are supported.
+ */
+
+#ifdef CONFIG_DEBUG_FS
+
+#include <linux/debugfs.h>
+
+#include "msm_drv.h"
+#include "msm_gpu.h"
+
+struct msm_perf_state {
+	struct drm_device *dev;
+
+	bool open;
+	int cnt;
+	struct mutex read_lock;
+
+	char buf[256];
+	int buftot, bufpos;
+
+	unsigned long next_jiffies;
+
+	struct dentry *ent;
+	struct drm_info_node *node;
+};
+
+#define SAMPLE_TIME (HZ/4)
+
+/* wait for next sample time: */
+static int wait_sample(struct msm_perf_state *perf)
+{
+	unsigned long start_jiffies = jiffies;
+
+	if (time_after(perf->next_jiffies, start_jiffies)) {
+		unsigned long remaining_jiffies =
+			perf->next_jiffies - start_jiffies;
+		int ret = schedule_timeout_interruptible(remaining_jiffies);
+		if (ret > 0) {
+			/* interrupted */
+			return -ERESTARTSYS;
+		}
+	}
+	perf->next_jiffies += SAMPLE_TIME;
+	return 0;
+}
+
+static int refill_buf(struct msm_perf_state *perf)
+{
+	struct msm_drm_private *priv = perf->dev->dev_private;
+	struct msm_gpu *gpu = priv->gpu;
+	char *ptr = perf->buf;
+	int rem = sizeof(perf->buf);
+	int i, n;
+
+	if ((perf->cnt++ % 32) == 0) {
+		/* Header line: */
+		n = snprintf(ptr, rem, "%%BUSY");
+		ptr += n;
+		rem -= n;
+
+		for (i = 0; i < gpu->num_perfcntrs; i++) {
+			const struct msm_gpu_perfcntr *perfcntr = &gpu->perfcntrs[i];
+			n = snprintf(ptr, rem, "\t%s", perfcntr->name);
+			ptr += n;
+			rem -= n;
+		}
+	} else {
+		/* Sample line: */
+		uint32_t activetime = 0, totaltime = 0;
+		uint32_t cntrs[5];
+		uint32_t val;
+		int ret;
+
+		/* sleep until next sample time: */
+		ret = wait_sample(perf);
+		if (ret)
+			return ret;
+
+		ret = msm_gpu_perfcntr_sample(gpu, &activetime, &totaltime,
+				ARRAY_SIZE(cntrs), cntrs);
+		if (ret < 0)
+			return ret;
+
+		val = totaltime ? 1000 * activetime / totaltime : 0;
+		n = snprintf(ptr, rem, "%3d.%d%%", val / 10, val % 10);
+		ptr += n;
+		rem -= n;
+
+		for (i = 0; i < ret; i++) {
+			/* cycle counters (I think).. convert to MHz.. */
+			val = cntrs[i] / 10000;
+			n = snprintf(ptr, rem, "\t%5d.%02d",
+					val / 100, val % 100);
+			ptr += n;
+			rem -= n;
+		}
+	}
+
+	n = snprintf(ptr, rem, "\n");
+	ptr += n;
+	rem -= n;
+
+	perf->bufpos = 0;
+	perf->buftot = ptr - perf->buf;
+
+	return 0;
+}
+
+static ssize_t perf_read(struct file *file, char __user *buf,
+		size_t sz, loff_t *ppos)
+{
+	struct msm_perf_state *perf = file->private_data;
+	int n = 0, ret;
+
+	mutex_lock(&perf->read_lock);
+
+	if (perf->bufpos >= perf->buftot) {
+		ret = refill_buf(perf);
+		if (ret)
+			goto out;
+	}
+
+	n = min((int)sz, perf->buftot - perf->bufpos);
+	ret = copy_to_user(buf, &perf->buf[perf->bufpos], n);
+	if (ret)
+		goto out;
+
+	perf->bufpos += n;
+	*ppos += n;
+
+out:
+	mutex_unlock(&perf->read_lock);
+	if (ret)
+		return ret;
+	return n;
+}
+
+static int perf_open(struct inode *inode, struct file *file)
+{
+	struct msm_perf_state *perf = inode->i_private;
+	struct drm_device *dev = perf->dev;
+	struct msm_drm_private *priv = dev->dev_private;
+	struct msm_gpu *gpu = priv->gpu;
+	int ret = 0;
+
+	mutex_lock(&dev->struct_mutex);
+
+	if (perf->open || !gpu) {
+		ret = -EBUSY;
+		goto out;
+	}
+
+	file->private_data = perf;
+	perf->open = true;
+	perf->cnt = 0;
+	perf->buftot = 0;
+	perf->bufpos = 0;
+	msm_gpu_perfcntr_start(gpu);
+	perf->next_jiffies = jiffies + SAMPLE_TIME;
+
+out:
+	mutex_unlock(&dev->struct_mutex);
+	return ret;
+}
+
+static int perf_release(struct inode *inode, struct file *file)
+{
+	struct msm_perf_state *perf = inode->i_private;
+	struct msm_drm_private *priv = perf->dev->dev_private;
+	msm_gpu_perfcntr_stop(priv->gpu);
+	perf->open = false;
+	return 0;
+}
+
+
+static const struct file_operations perf_debugfs_fops = {
+	.owner = THIS_MODULE,
+	.open = perf_open,
+	.read = perf_read,
+	.llseek = no_llseek,
+	.release = perf_release,
+};
+
+int msm_perf_debugfs_init(struct drm_minor *minor)
+{
+	struct msm_drm_private *priv = minor->dev->dev_private;
+	struct msm_perf_state *perf;
+
+	/* only create on first minor: */
+	if (priv->perf)
+		return 0;
+
+	perf = kzalloc(sizeof(*perf), GFP_KERNEL);
+	if (!perf)
+		return -ENOMEM;
+
+	perf->dev = minor->dev;
+
+	mutex_init(&perf->read_lock);
+	priv->perf = perf;
+
+	perf->node = kzalloc(sizeof(*perf->node), GFP_KERNEL);
+	if (!perf->node)
+		goto fail;
+
+	perf->ent = debugfs_create_file("perf", S_IFREG | S_IRUGO,
+			minor->debugfs_root, perf, &perf_debugfs_fops);
+	if (!perf->ent) {
+		DRM_ERROR("Cannot create /sys/kernel/debug/dri/%s/perf\n",
+				minor->debugfs_root->d_name.name);
+		goto fail;
+	}
+
+	perf->node->minor = minor;
+	perf->node->dent  = perf->ent;
+	perf->node->info_ent = NULL;
+
+	mutex_lock(&minor->debugfs_lock);
+	list_add(&perf->node->list, &minor->debugfs_list);
+	mutex_unlock(&minor->debugfs_lock);
+
+	return 0;
+
+fail:
+	msm_perf_debugfs_cleanup(minor);
+	return -1;
+}
+
+void msm_perf_debugfs_cleanup(struct drm_minor *minor)
+{
+	struct msm_drm_private *priv = minor->dev->dev_private;
+	struct msm_perf_state *perf = priv->perf;
+
+	if (!perf)
+		return;
+
+	priv->perf = NULL;
+
+	debugfs_remove(perf->ent);
+
+	if (perf->node) {
+		mutex_lock(&minor->debugfs_lock);
+		list_del(&perf->node->list);
+		mutex_unlock(&minor->debugfs_lock);
+		kfree(perf->node);
+	}
+
+	mutex_destroy(&perf->read_lock);
+
+	kfree(perf);
+}
+
+#endif
diff --git a/drivers/gpu/drm/msm/msm_rd.c b/drivers/gpu/drm/msm/msm_rd.c
new file mode 100644
index 00000000000..9a78c48817c
--- /dev/null
+++ b/drivers/gpu/drm/msm/msm_rd.c
@@ -0,0 +1,337 @@
+/*
+ * Copyright (C) 2013 Red Hat
+ * Author: Rob Clark <robdclark@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+/* For debugging crashes, userspace can:
+ *
+ *   tail -f /sys/kernel/debug/dri/<minor>/rd > logfile.rd
+ *
+ * To log the cmdstream in a format that is understood by freedreno/cffdump
+ * utility.  By comparing the last successfully completed fence #, to the
+ * cmdstream for the next fence, you can narrow down which process and submit
+ * caused the gpu crash/lockup.
+ *
+ * This bypasses drm_debugfs_create_files() mainly because we need to use
+ * our own fops for a bit more control.  In particular, we don't want to
+ * do anything if userspace doesn't have the debugfs file open.
+ */
+
+#ifdef CONFIG_DEBUG_FS
+
+#include <linux/kfifo.h>
+#include <linux/debugfs.h>
+#include <linux/circ_buf.h>
+#include <linux/wait.h>
+
+#include "msm_drv.h"
+#include "msm_gpu.h"
+#include "msm_gem.h"
+
+enum rd_sect_type {
+	RD_NONE,
+	RD_TEST,       /* ascii text */
+	RD_CMD,        /* ascii text */
+	RD_GPUADDR,    /* u32 gpuaddr, u32 size */
+	RD_CONTEXT,    /* raw dump */
+	RD_CMDSTREAM,  /* raw dump */
+	RD_CMDSTREAM_ADDR, /* gpu addr of cmdstream */
+	RD_PARAM,      /* u32 param_type, u32 param_val, u32 bitlen */
+	RD_FLUSH,      /* empty, clear previous params */
+	RD_PROGRAM,    /* shader program, raw dump */
+	RD_VERT_SHADER,
+	RD_FRAG_SHADER,
+	RD_BUFFER_CONTENTS,
+	RD_GPU_ID,
+};
+
+#define BUF_SZ 512  /* should be power of 2 */
+
+/* space used: */
+#define circ_count(circ) \
+	(CIRC_CNT((circ)->head, (circ)->tail, BUF_SZ))
+#define circ_count_to_end(circ) \
+	(CIRC_CNT_TO_END((circ)->head, (circ)->tail, BUF_SZ))
+/* space available: */
+#define circ_space(circ) \
+	(CIRC_SPACE((circ)->head, (circ)->tail, BUF_SZ))
+#define circ_space_to_end(circ) \
+	(CIRC_SPACE_TO_END((circ)->head, (circ)->tail, BUF_SZ))
+
+struct msm_rd_state {
+	struct drm_device *dev;
+
+	bool open;
+
+	struct dentry *ent;
+	struct drm_info_node *node;
+
+	/* current submit to read out: */
+	struct msm_gem_submit *submit;
+
+	/* fifo access is synchronized on the producer side by
+	 * struct_mutex held by submit code (otherwise we could
+	 * end up w/ cmds logged in different order than they
+	 * were executed).  And read_lock synchronizes the reads
+	 */
+	struct mutex read_lock;
+
+	wait_queue_head_t fifo_event;
+	struct circ_buf fifo;
+
+	char buf[BUF_SZ];
+};
+
+static void rd_write(struct msm_rd_state *rd, const void *buf, int sz)
+{
+	struct circ_buf *fifo = &rd->fifo;
+	const char *ptr = buf;
+
+	while (sz > 0) {
+		char *fptr = &fifo->buf[fifo->head];
+		int n;
+
+		wait_event(rd->fifo_event, circ_space(&rd->fifo) > 0);
+
+		n = min(sz, circ_space_to_end(&rd->fifo));
+		memcpy(fptr, ptr, n);
+
+		fifo->head = (fifo->head + n) & (BUF_SZ - 1);
+		sz  -= n;
+		ptr += n;
+
+		wake_up_all(&rd->fifo_event);
+	}
+}
+
+static void rd_write_section(struct msm_rd_state *rd,
+		enum rd_sect_type type, const void *buf, int sz)
+{
+	rd_write(rd, &type, 4);
+	rd_write(rd, &sz, 4);
+	rd_write(rd, buf, sz);
+}
+
+static ssize_t rd_read(struct file *file, char __user *buf,
+		size_t sz, loff_t *ppos)
+{
+	struct msm_rd_state *rd = file->private_data;
+	struct circ_buf *fifo = &rd->fifo;
+	const char *fptr = &fifo->buf[fifo->tail];
+	int n = 0, ret = 0;
+
+	mutex_lock(&rd->read_lock);
+
+	ret = wait_event_interruptible(rd->fifo_event,
+			circ_count(&rd->fifo) > 0);
+	if (ret)
+		goto out;
+
+	n = min_t(int, sz, circ_count_to_end(&rd->fifo));
+	ret = copy_to_user(buf, fptr, n);
+	if (ret)
+		goto out;
+
+	fifo->tail = (fifo->tail + n) & (BUF_SZ - 1);
+	*ppos += n;
+
+	wake_up_all(&rd->fifo_event);
+
+out:
+	mutex_unlock(&rd->read_lock);
+	if (ret)
+		return ret;
+	return n;
+}
+
+static int rd_open(struct inode *inode, struct file *file)
+{
+	struct msm_rd_state *rd = inode->i_private;
+	struct drm_device *dev = rd->dev;
+	struct msm_drm_private *priv = dev->dev_private;
+	struct msm_gpu *gpu = priv->gpu;
+	uint64_t val;
+	uint32_t gpu_id;
+	int ret = 0;
+
+	mutex_lock(&dev->struct_mutex);
+
+	if (rd->open || !gpu) {
+		ret = -EBUSY;
+		goto out;
+	}
+
+	file->private_data = rd;
+	rd->open = true;
+
+	/* the parsing tools need to know gpu-id to know which
+	 * register database to load.
+	 */
+	gpu->funcs->get_param(gpu, MSM_PARAM_GPU_ID, &val);
+	gpu_id = val;
+
+	rd_write_section(rd, RD_GPU_ID, &gpu_id, sizeof(gpu_id));
+
+out:
+	mutex_unlock(&dev->struct_mutex);
+	return ret;
+}
+
+static int rd_release(struct inode *inode, struct file *file)
+{
+	struct msm_rd_state *rd = inode->i_private;
+	rd->open = false;
+	return 0;
+}
+
+
+static const struct file_operations rd_debugfs_fops = {
+	.owner = THIS_MODULE,
+	.open = rd_open,
+	.read = rd_read,
+	.llseek = no_llseek,
+	.release = rd_release,
+};
+
+int msm_rd_debugfs_init(struct drm_minor *minor)
+{
+	struct msm_drm_private *priv = minor->dev->dev_private;
+	struct msm_rd_state *rd;
+
+	/* only create on first minor: */
+	if (priv->rd)
+		return 0;
+
+	rd = kzalloc(sizeof(*rd), GFP_KERNEL);
+	if (!rd)
+		return -ENOMEM;
+
+	rd->dev = minor->dev;
+	rd->fifo.buf = rd->buf;
+
+	mutex_init(&rd->read_lock);
+	priv->rd = rd;
+
+	init_waitqueue_head(&rd->fifo_event);
+
+	rd->node = kzalloc(sizeof(*rd->node), GFP_KERNEL);
+	if (!rd->node)
+		goto fail;
+
+	rd->ent = debugfs_create_file("rd", S_IFREG | S_IRUGO,
+			minor->debugfs_root, rd, &rd_debugfs_fops);
+	if (!rd->ent) {
+		DRM_ERROR("Cannot create /sys/kernel/debug/dri/%s/rd\n",
+				minor->debugfs_root->d_name.name);
+		goto fail;
+	}
+
+	rd->node->minor = minor;
+	rd->node->dent  = rd->ent;
+	rd->node->info_ent = NULL;
+
+	mutex_lock(&minor->debugfs_lock);
+	list_add(&rd->node->list, &minor->debugfs_list);
+	mutex_unlock(&minor->debugfs_lock);
+
+	return 0;
+
+fail:
+	msm_rd_debugfs_cleanup(minor);
+	return -1;
+}
+
+void msm_rd_debugfs_cleanup(struct drm_minor *minor)
+{
+	struct msm_drm_private *priv = minor->dev->dev_private;
+	struct msm_rd_state *rd = priv->rd;
+
+	if (!rd)
+		return;
+
+	priv->rd = NULL;
+
+	debugfs_remove(rd->ent);
+
+	if (rd->node) {
+		mutex_lock(&minor->debugfs_lock);
+		list_del(&rd->node->list);
+		mutex_unlock(&minor->debugfs_lock);
+		kfree(rd->node);
+	}
+
+	mutex_destroy(&rd->read_lock);
+
+	kfree(rd);
+}
+
+/* called under struct_mutex */
+void msm_rd_dump_submit(struct msm_gem_submit *submit)
+{
+	struct drm_device *dev = submit->dev;
+	struct msm_drm_private *priv = dev->dev_private;
+	struct msm_rd_state *rd = priv->rd;
+	char msg[128];
+	int i, n;
+
+	if (!rd->open)
+		return;
+
+	/* writing into fifo is serialized by caller, and
+	 * rd->read_lock is used to serialize the reads
+	 */
+	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
+
+	n = snprintf(msg, sizeof(msg), "%.*s/%d: fence=%u",
+			TASK_COMM_LEN, current->comm, task_pid_nr(current),
+			submit->fence);
+
+	rd_write_section(rd, RD_CMD, msg, ALIGN(n, 4));
+
+	/* could be nice to have an option (module-param?) to snapshot
+	 * all the bo's associated with the submit.  Handy to see vtx
+	 * buffers, etc.  For now just the cmdstream bo's is enough.
+	 */
+
+	for (i = 0; i < submit->nr_cmds; i++) {
+		uint32_t idx  = submit->cmd[i].idx;
+		uint32_t iova = submit->cmd[i].iova;
+		uint32_t szd  = submit->cmd[i].size; /* in dwords */
+		struct msm_gem_object *obj = submit->bos[idx].obj;
+		const char *buf = msm_gem_vaddr_locked(&obj->base);
+
+		buf += iova - submit->bos[idx].iova;
+
+		rd_write_section(rd, RD_GPUADDR,
+				(uint32_t[2]){ iova, szd * 4 }, 8);
+		rd_write_section(rd, RD_BUFFER_CONTENTS,
+				buf, szd * 4);
+
+		switch (submit->cmd[i].type) {
+		case MSM_SUBMIT_CMD_IB_TARGET_BUF:
+			/* ignore IB-targets, we've logged the buffer, the
+			 * parser tool will follow the IB based on the logged
+			 * buffer/gpuaddr, so nothing more to do.
+			 */
+			break;
+		case MSM_SUBMIT_CMD_CTX_RESTORE_BUF:
+		case MSM_SUBMIT_CMD_BUF:
+			rd_write_section(rd, RD_CMDSTREAM_ADDR,
+					(uint32_t[2]){ iova, szd }, 8);
+			break;
+		}
+	}
+}
+#endif
diff --git a/drivers/gpu/drm/nouveau/Kconfig b/drivers/gpu/drm/nouveau/Kconfig
index 7cf787d697b..637c29a3312 100644
--- a/drivers/gpu/drm/nouveau/Kconfig
+++ b/drivers/gpu/drm/nouveau/Kconfig
@@ -11,7 +11,7 @@ config DRM_NOUVEAU
 	select FB
 	select FRAMEBUFFER_CONSOLE if !EXPERT
 	select FB_BACKLIGHT if DRM_NOUVEAU_BACKLIGHT
-	select ACPI_VIDEO if ACPI && X86 && BACKLIGHT_CLASS_DEVICE && VIDEO_OUTPUT_CONTROL && INPUT
+	select ACPI_VIDEO if ACPI && X86 && BACKLIGHT_CLASS_DEVICE && INPUT
 	select X86_PLATFORM_DEVICES if ACPI && X86
 	select ACPI_WMI if ACPI && X86
 	select MXM_WMI if ACPI && X86
@@ -19,7 +19,6 @@ config DRM_NOUVEAU
 	# Similar to i915, we need to select ACPI_VIDEO and it's dependencies
 	select BACKLIGHT_LCD_SUPPORT if ACPI && X86
 	select BACKLIGHT_CLASS_DEVICE if ACPI && X86
-	select VIDEO_OUTPUT_CONTROL if ACPI && X86
 	select INPUT if ACPI && X86
 	select THERMAL if ACPI && X86
 	select ACPI_VIDEO if ACPI && X86
diff --git a/drivers/gpu/drm/nouveau/Makefile b/drivers/gpu/drm/nouveau/Makefile
index b3fa1ba191b..8b307e14363 100644
--- a/drivers/gpu/drm/nouveau/Makefile
+++ b/drivers/gpu/drm/nouveau/Makefile
@@ -41,12 +41,14 @@ nouveau-y += core/subdev/bios/init.o
 nouveau-y += core/subdev/bios/mxm.o
 nouveau-y += core/subdev/bios/perf.o
 nouveau-y += core/subdev/bios/pll.o
+nouveau-y += core/subdev/bios/ramcfg.o
 nouveau-y += core/subdev/bios/rammap.o
 nouveau-y += core/subdev/bios/timing.o
 nouveau-y += core/subdev/bios/therm.o
 nouveau-y += core/subdev/bios/vmap.o
 nouveau-y += core/subdev/bios/volt.o
 nouveau-y += core/subdev/bios/xpio.o
+nouveau-y += core/subdev/bios/P0260.o
 nouveau-y += core/subdev/bus/hwsq.o
 nouveau-y += core/subdev/bus/nv04.o
 nouveau-y += core/subdev/bus/nv31.o
@@ -71,8 +73,12 @@ nouveau-y += core/subdev/devinit/nv10.o
 nouveau-y += core/subdev/devinit/nv1a.o
 nouveau-y += core/subdev/devinit/nv20.o
 nouveau-y += core/subdev/devinit/nv50.o
+nouveau-y += core/subdev/devinit/nv84.o
+nouveau-y += core/subdev/devinit/nv98.o
 nouveau-y += core/subdev/devinit/nva3.o
+nouveau-y += core/subdev/devinit/nvaf.o
 nouveau-y += core/subdev/devinit/nvc0.o
+nouveau-y += core/subdev/devinit/gm107.o
 nouveau-y += core/subdev/fb/base.o
 nouveau-y += core/subdev/fb/nv04.o
 nouveau-y += core/subdev/fb/nv10.o
@@ -96,6 +102,8 @@ nouveau-y += core/subdev/fb/nvaa.o
 nouveau-y += core/subdev/fb/nvaf.o
 nouveau-y += core/subdev/fb/nvc0.o
 nouveau-y += core/subdev/fb/nve0.o
+nouveau-y += core/subdev/fb/gk20a.o
+nouveau-y += core/subdev/fb/gm107.o
 nouveau-y += core/subdev/fb/ramnv04.o
 nouveau-y += core/subdev/fb/ramnv10.o
 nouveau-y += core/subdev/fb/ramnv1a.o
@@ -110,33 +118,44 @@ nouveau-y += core/subdev/fb/ramnva3.o
 nouveau-y += core/subdev/fb/ramnvaa.o
 nouveau-y += core/subdev/fb/ramnvc0.o
 nouveau-y += core/subdev/fb/ramnve0.o
+nouveau-y += core/subdev/fb/ramgk20a.o
+nouveau-y += core/subdev/fb/ramgm107.o
 nouveau-y += core/subdev/fb/sddr3.o
 nouveau-y += core/subdev/fb/gddr5.o
 nouveau-y += core/subdev/gpio/base.o
 nouveau-y += core/subdev/gpio/nv10.o
 nouveau-y += core/subdev/gpio/nv50.o
+nouveau-y += core/subdev/gpio/nv92.o
 nouveau-y += core/subdev/gpio/nvd0.o
 nouveau-y += core/subdev/gpio/nve0.o
 nouveau-y += core/subdev/i2c/base.o
 nouveau-y += core/subdev/i2c/anx9805.o
 nouveau-y += core/subdev/i2c/aux.o
 nouveau-y += core/subdev/i2c/bit.o
+nouveau-y += core/subdev/i2c/pad.o
+nouveau-y += core/subdev/i2c/padnv04.o
+nouveau-y += core/subdev/i2c/padnv94.o
 nouveau-y += core/subdev/i2c/nv04.o
 nouveau-y += core/subdev/i2c/nv4e.o
 nouveau-y += core/subdev/i2c/nv50.o
 nouveau-y += core/subdev/i2c/nv94.o
 nouveau-y += core/subdev/i2c/nvd0.o
+nouveau-y += core/subdev/i2c/gf117.o
+nouveau-y += core/subdev/i2c/nve0.o
 nouveau-y += core/subdev/ibus/nvc0.o
 nouveau-y += core/subdev/ibus/nve0.o
+nouveau-y += core/subdev/ibus/gk20a.o
 nouveau-y += core/subdev/instmem/base.o
 nouveau-y += core/subdev/instmem/nv04.o
 nouveau-y += core/subdev/instmem/nv40.o
 nouveau-y += core/subdev/instmem/nv50.o
-nouveau-y += core/subdev/ltcg/nvc0.o
+nouveau-y += core/subdev/ltcg/gf100.o
+nouveau-y += core/subdev/ltcg/gm107.o
 nouveau-y += core/subdev/mc/base.o
 nouveau-y += core/subdev/mc/nv04.o
 nouveau-y += core/subdev/mc/nv40.o
 nouveau-y += core/subdev/mc/nv44.o
+nouveau-y += core/subdev/mc/nv4c.o
 nouveau-y += core/subdev/mc/nv50.o
 nouveau-y += core/subdev/mc/nv94.o
 nouveau-y += core/subdev/mc/nv98.o
@@ -165,6 +184,7 @@ nouveau-y += core/subdev/therm/nva3.o
 nouveau-y += core/subdev/therm/nvd0.o
 nouveau-y += core/subdev/timer/base.o
 nouveau-y += core/subdev/timer/nv04.o
+nouveau-y += core/subdev/timer/gk20a.o
 nouveau-y += core/subdev/vm/base.o
 nouveau-y += core/subdev/vm/nv04.o
 nouveau-y += core/subdev/vm/nv41.o
@@ -201,7 +221,11 @@ nouveau-y += core/engine/device/nv40.o
 nouveau-y += core/engine/device/nv50.o
 nouveau-y += core/engine/device/nvc0.o
 nouveau-y += core/engine/device/nve0.o
+nouveau-y += core/engine/device/gm100.o
 nouveau-y += core/engine/disp/base.o
+nouveau-y += core/engine/disp/conn.o
+nouveau-y += core/engine/disp/outp.o
+nouveau-y += core/engine/disp/outpdp.o
 nouveau-y += core/engine/disp/nv04.o
 nouveau-y += core/engine/disp/nv50.o
 nouveau-y += core/engine/disp/nv84.o
@@ -211,6 +235,7 @@ nouveau-y += core/engine/disp/nva3.o
 nouveau-y += core/engine/disp/nvd0.o
 nouveau-y += core/engine/disp/nve0.o
 nouveau-y += core/engine/disp/nvf0.o
+nouveau-y += core/engine/disp/gm107.o
 nouveau-y += core/engine/disp/dacnv50.o
 nouveau-y += core/engine/disp/dport.o
 nouveau-y += core/engine/disp/hdanva3.o
@@ -232,16 +257,21 @@ nouveau-y += core/engine/fifo/nv50.o
 nouveau-y += core/engine/fifo/nv84.o
 nouveau-y += core/engine/fifo/nvc0.o
 nouveau-y += core/engine/fifo/nve0.o
+nouveau-y += core/engine/fifo/gk20a.o
+nouveau-y += core/engine/fifo/nv108.o
 nouveau-y += core/engine/graph/ctxnv40.o
 nouveau-y += core/engine/graph/ctxnv50.o
 nouveau-y += core/engine/graph/ctxnvc0.o
 nouveau-y += core/engine/graph/ctxnvc1.o
-nouveau-y += core/engine/graph/ctxnvc3.o
+nouveau-y += core/engine/graph/ctxnvc4.o
 nouveau-y += core/engine/graph/ctxnvc8.o
 nouveau-y += core/engine/graph/ctxnvd7.o
 nouveau-y += core/engine/graph/ctxnvd9.o
 nouveau-y += core/engine/graph/ctxnve4.o
+nouveau-y += core/engine/graph/ctxgk20a.o
 nouveau-y += core/engine/graph/ctxnvf0.o
+nouveau-y += core/engine/graph/ctxnv108.o
+nouveau-y += core/engine/graph/ctxgm107.o
 nouveau-y += core/engine/graph/nv04.o
 nouveau-y += core/engine/graph/nv10.o
 nouveau-y += core/engine/graph/nv20.o
@@ -254,12 +284,15 @@ nouveau-y += core/engine/graph/nv40.o
 nouveau-y += core/engine/graph/nv50.o
 nouveau-y += core/engine/graph/nvc0.o
 nouveau-y += core/engine/graph/nvc1.o
-nouveau-y += core/engine/graph/nvc3.o
+nouveau-y += core/engine/graph/nvc4.o
 nouveau-y += core/engine/graph/nvc8.o
 nouveau-y += core/engine/graph/nvd7.o
 nouveau-y += core/engine/graph/nvd9.o
 nouveau-y += core/engine/graph/nve4.o
+nouveau-y += core/engine/graph/gk20a.o
 nouveau-y += core/engine/graph/nvf0.o
+nouveau-y += core/engine/graph/nv108.o
+nouveau-y += core/engine/graph/gm107.o
 nouveau-y += core/engine/mpeg/nv31.o
 nouveau-y += core/engine/mpeg/nv40.o
 nouveau-y += core/engine/mpeg/nv44.o
diff --git a/drivers/gpu/drm/nouveau/core/core/engine.c b/drivers/gpu/drm/nouveau/core/core/engine.c
index c8bed4a2683..1f6954ae9dd 100644
--- a/drivers/gpu/drm/nouveau/core/core/engine.c
+++ b/drivers/gpu/drm/nouveau/core/core/engine.c
@@ -42,11 +42,24 @@ nouveau_engine_create_(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	if ( parent &&
-	    !nouveau_boolopt(nv_device(parent)->cfgopt, iname, enable)) {
-		if (!enable)
-			nv_warn(engine, "disabled, %s=1 to enable\n", iname);
-		return -ENODEV;
+	if (parent) {
+		struct nouveau_device *device = nv_device(parent);
+		int engidx = nv_engidx(nv_object(engine));
+
+		if (device->disable_mask & (1ULL << engidx)) {
+			if (!nouveau_boolopt(device->cfgopt, iname, false)) {
+				nv_debug(engine, "engine disabled by hw/fw\n");
+				return -ENODEV;
+			}
+
+			nv_warn(engine, "ignoring hw/fw engine disable\n");
+		}
+
+		if (!nouveau_boolopt(device->cfgopt, iname, enable)) {
+			if (!enable)
+				nv_warn(engine, "disabled, %s=1 to enable\n", iname);
+			return -ENODEV;
+		}
 	}
 
 	INIT_LIST_HEAD(&engine->contexts);
diff --git a/drivers/gpu/drm/nouveau/core/core/event.c b/drivers/gpu/drm/nouveau/core/core/event.c
index 3f3c76581a9..ae81d3b5d8b 100644
--- a/drivers/gpu/drm/nouveau/core/core/event.c
+++ b/drivers/gpu/drm/nouveau/core/core/event.c
@@ -28,14 +28,20 @@ nouveau_event_put(struct nouveau_eventh *handler)
 {
 	struct nouveau_event *event = handler->event;
 	unsigned long flags;
-	if (__test_and_clear_bit(NVKM_EVENT_ENABLE, &handler->flags)) {
-		spin_lock_irqsave(&event->refs_lock, flags);
-		if (!--event->index[handler->index].refs) {
+	u32 m, t;
+
+	if (!__test_and_clear_bit(NVKM_EVENT_ENABLE, &handler->flags))
+		return;
+
+	spin_lock_irqsave(&event->refs_lock, flags);
+	for (m = handler->types; t = __ffs(m), m; m &= ~(1 << t)) {
+		if (!--event->refs[handler->index * event->types_nr + t]) {
 			if (event->disable)
-				event->disable(event, handler->index);
+				event->disable(event, 1 << t, handler->index);
 		}
-		spin_unlock_irqrestore(&event->refs_lock, flags);
+
 	}
+	spin_unlock_irqrestore(&event->refs_lock, flags);
 }
 
 void
@@ -43,14 +49,20 @@ nouveau_event_get(struct nouveau_eventh *handler)
 {
 	struct nouveau_event *event = handler->event;
 	unsigned long flags;
-	if (!__test_and_set_bit(NVKM_EVENT_ENABLE, &handler->flags)) {
-		spin_lock_irqsave(&event->refs_lock, flags);
-		if (!event->index[handler->index].refs++) {
+	u32 m, t;
+
+	if (__test_and_set_bit(NVKM_EVENT_ENABLE, &handler->flags))
+		return;
+
+	spin_lock_irqsave(&event->refs_lock, flags);
+	for (m = handler->types; t = __ffs(m), m; m &= ~(1 << t)) {
+		if (!event->refs[handler->index * event->types_nr + t]++) {
 			if (event->enable)
-				event->enable(event, handler->index);
+				event->enable(event, 1 << t, handler->index);
 		}
-		spin_unlock_irqrestore(&event->refs_lock, flags);
+
 	}
+	spin_unlock_irqrestore(&event->refs_lock, flags);
 }
 
 static void
@@ -65,38 +77,47 @@ nouveau_event_fini(struct nouveau_eventh *handler)
 }
 
 static int
-nouveau_event_init(struct nouveau_event *event, int index,
-		   int (*func)(void *, int), void *priv,
+nouveau_event_init(struct nouveau_event *event, u32 types, int index,
+		   int (*func)(void *, u32, int), void *priv,
 		   struct nouveau_eventh *handler)
 {
 	unsigned long flags;
 
+	if (types & ~((1 << event->types_nr) - 1))
+		return -EINVAL;
 	if (index >= event->index_nr)
 		return -EINVAL;
 
 	handler->event = event;
 	handler->flags = 0;
+	handler->types = types;
 	handler->index = index;
 	handler->func = func;
 	handler->priv = priv;
 
 	spin_lock_irqsave(&event->list_lock, flags);
-	list_add_tail(&handler->head, &event->index[index].list);
+	list_add_tail(&handler->head, &event->list[index]);
 	spin_unlock_irqrestore(&event->list_lock, flags);
 	return 0;
 }
 
 int
-nouveau_event_new(struct nouveau_event *event, int index,
-		  int (*func)(void *, int), void *priv,
+nouveau_event_new(struct nouveau_event *event, u32 types, int index,
+		  int (*func)(void *, u32, int), void *priv,
 		  struct nouveau_eventh **phandler)
 {
 	struct nouveau_eventh *handler;
 	int ret = -ENOMEM;
 
+	if (event->check) {
+		ret = event->check(event, types, index);
+		if (ret)
+			return ret;
+	}
+
 	handler = *phandler = kmalloc(sizeof(*handler), GFP_KERNEL);
 	if (handler) {
-		ret = nouveau_event_init(event, index, func, priv, handler);
+		ret = nouveau_event_init(event, types, index, func, priv, handler);
 		if (ret)
 			kfree(handler);
 	}
@@ -116,7 +137,7 @@ nouveau_event_ref(struct nouveau_eventh *handler, struct nouveau_eventh **ref)
 }
 
 void
-nouveau_event_trigger(struct nouveau_event *event, int index)
+nouveau_event_trigger(struct nouveau_event *event, u32 types, int index)
 {
 	struct nouveau_eventh *handler;
 	unsigned long flags;
@@ -125,10 +146,15 @@ nouveau_event_trigger(struct nouveau_event *event, int index)
 		return;
 
 	spin_lock_irqsave(&event->list_lock, flags);
-	list_for_each_entry(handler, &event->index[index].list, head) {
-		if (test_bit(NVKM_EVENT_ENABLE, &handler->flags) &&
-		    handler->func(handler->priv, index) == NVKM_EVENT_DROP)
-			nouveau_event_put(handler);
+	list_for_each_entry(handler, &event->list[index], head) {
+		if (!test_bit(NVKM_EVENT_ENABLE, &handler->flags))
+			continue;
+		if (!(handler->types & types))
+			continue;
+		if (handler->func(handler->priv, handler->types & types, index)
+				!= NVKM_EVENT_DROP)
+			continue;
+		nouveau_event_put(handler);
 	}
 	spin_unlock_irqrestore(&event->list_lock, flags);
 }
@@ -144,20 +170,27 @@ nouveau_event_destroy(struct nouveau_event **pevent)
 }
 
 int
-nouveau_event_create(int index_nr, struct nouveau_event **pevent)
+nouveau_event_create(int types_nr, int index_nr, struct nouveau_event **pevent)
 {
 	struct nouveau_event *event;
 	int i;
 
-	event = *pevent = kzalloc(sizeof(*event) + index_nr *
-				  sizeof(event->index[0]), GFP_KERNEL);
+	event = *pevent = kzalloc(sizeof(*event) + (index_nr * types_nr) *
+				  sizeof(event->refs[0]), GFP_KERNEL);
 	if (!event)
 		return -ENOMEM;
 
+	event->list = kmalloc(sizeof(*event->list) * index_nr, GFP_KERNEL);
+	if (!event->list) {
+		kfree(event);
+		return -ENOMEM;
+	}
+
 	spin_lock_init(&event->list_lock);
 	spin_lock_init(&event->refs_lock);
 	for (i = 0; i < index_nr; i++)
-		INIT_LIST_HEAD(&event->index[i].list);
+		INIT_LIST_HEAD(&event->list[i]);
+	event->types_nr = types_nr;
 	event->index_nr = index_nr;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/core/namedb.c b/drivers/gpu/drm/nouveau/core/core/namedb.c
index 1ce95a8709d..0594a599f6f 100644
--- a/drivers/gpu/drm/nouveau/core/core/namedb.c
+++ b/drivers/gpu/drm/nouveau/core/core/namedb.c
@@ -167,7 +167,7 @@ int
 nouveau_namedb_create_(struct nouveau_object *parent,
 		       struct nouveau_object *engine,
 		       struct nouveau_oclass *oclass, u32 pclass,
-		       struct nouveau_oclass *sclass, u32 engcls,
+		       struct nouveau_oclass *sclass, u64 engcls,
 		       int length, void **pobject)
 {
 	struct nouveau_namedb *namedb;
diff --git a/drivers/gpu/drm/nouveau/core/core/object.c b/drivers/gpu/drm/nouveau/core/core/object.c
index 7f48e288215..12453855590 100644
--- a/drivers/gpu/drm/nouveau/core/core/object.c
+++ b/drivers/gpu/drm/nouveau/core/core/object.c
@@ -156,7 +156,7 @@ nouveau_object_ctor(struct nouveau_object *parent,
 	}
 
 	if (ret == 0) {
-		nv_debug(object, "created\n");
+		nv_trace(object, "created\n");
 		atomic_set(&object->refcount, 1);
 	}
 
@@ -166,7 +166,7 @@ nouveau_object_ctor(struct nouveau_object *parent,
 static void
 nouveau_object_dtor(struct nouveau_object *object)
 {
-	nv_debug(object, "destroying\n");
+	nv_trace(object, "destroying\n");
 	nv_ofuncs(object)->dtor(object);
 }
 
@@ -337,7 +337,7 @@ nouveau_object_inc(struct nouveau_object *object)
 		goto fail_self;
 	}
 
-	nv_debug(object, "initialised\n");
+	nv_trace(object, "initialised\n");
 	return 0;
 
 fail_self:
@@ -375,7 +375,7 @@ nouveau_object_decf(struct nouveau_object *object)
 	if (object->parent)
 		nouveau_object_dec(object->parent, false);
 
-	nv_debug(object, "stopped\n");
+	nv_trace(object, "stopped\n");
 	return 0;
 }
 
@@ -411,7 +411,7 @@ nouveau_object_decs(struct nouveau_object *object)
 		}
 	}
 
-	nv_debug(object, "suspended\n");
+	nv_trace(object, "suspended\n");
 	return 0;
 
 fail_parent:
diff --git a/drivers/gpu/drm/nouveau/core/core/parent.c b/drivers/gpu/drm/nouveau/core/core/parent.c
index 313380ce632..dee5d1235e9 100644
--- a/drivers/gpu/drm/nouveau/core/core/parent.c
+++ b/drivers/gpu/drm/nouveau/core/core/parent.c
@@ -49,7 +49,7 @@ nouveau_parent_sclass(struct nouveau_object *parent, u16 handle,
 
 	mask = nv_parent(parent)->engine;
 	while (mask) {
-		int i = ffsll(mask) - 1;
+		int i = __ffs64(mask);
 
 		if (nv_iclass(parent, NV_CLIENT_CLASS))
 			engine = nv_engine(nv_client(parent)->device);
diff --git a/drivers/gpu/drm/nouveau/core/engine/copy/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/copy/nvc0.c
index 993df09ad64..ac3291f781f 100644
--- a/drivers/gpu/drm/nouveau/core/engine/copy/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/copy/nvc0.c
@@ -105,9 +105,6 @@ nvc0_copy0_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nvc0_copy_priv *priv;
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000100)
-		return -ENODEV;
-
 	ret = nouveau_falcon_create(parent, engine, oclass, 0x104000, true,
 				    "PCE0", "copy0", &priv);
 	*pobject = nv_object(priv);
@@ -133,9 +130,6 @@ nvc0_copy1_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nvc0_copy_priv *priv;
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000200)
-		return -ENODEV;
-
 	ret = nouveau_falcon_create(parent, engine, oclass, 0x105000, true,
 				    "PCE1", "copy1", &priv);
 	*pobject = nv_object(priv);
diff --git a/drivers/gpu/drm/nouveau/core/engine/copy/nve0.c b/drivers/gpu/drm/nouveau/core/engine/copy/nve0.c
index 30f1ef1edcc..748a61eb3c6 100644
--- a/drivers/gpu/drm/nouveau/core/engine/copy/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/copy/nve0.c
@@ -88,9 +88,6 @@ nve0_copy0_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nve0_copy_priv *priv;
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000100)
-		return -ENODEV;
-
 	ret = nouveau_engine_create(parent, engine, oclass, true,
 				    "PCE0", "copy0", &priv);
 	*pobject = nv_object(priv);
@@ -112,9 +109,6 @@ nve0_copy1_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nve0_copy_priv *priv;
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000200)
-		return -ENODEV;
-
 	ret = nouveau_engine_create(parent, engine, oclass, true,
 				    "PCE1", "copy1", &priv);
 	*pobject = nv_object(priv);
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/base.c b/drivers/gpu/drm/nouveau/core/engine/device/base.c
index dd01c6c435d..18c8c7245b7 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/base.c
@@ -131,8 +131,8 @@ nouveau_devobj_ctor(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	mmio_base = pci_resource_start(device->pdev, 0);
-	mmio_size = pci_resource_len(device->pdev, 0);
+	mmio_base = nv_device_resource_start(device, 0);
+	mmio_size = nv_device_resource_len(device, 0);
 
 	/* translate api disable mask into internal mapping */
 	disable = args->debug0;
@@ -185,6 +185,7 @@ nouveau_devobj_ctor(struct nouveau_object *parent,
 			case 0x0e0:
 			case 0x0f0:
 			case 0x100: device->card_type = NV_E0; break;
+			case 0x110: device->card_type = GM100; break;
 			default:
 				break;
 			}
@@ -208,6 +209,7 @@ nouveau_devobj_ctor(struct nouveau_object *parent,
 		case NV_C0:
 		case NV_D0: ret = nvc0_identify(device); break;
 		case NV_E0: ret = nve0_identify(device); break;
+		case GM100: ret = gm100_identify(device); break;
 		default:
 			ret = -EINVAL;
 			break;
@@ -446,6 +448,72 @@ nouveau_device_dtor(struct nouveau_object *object)
 	nouveau_engine_destroy(&device->base);
 }
 
+resource_size_t
+nv_device_resource_start(struct nouveau_device *device, unsigned int bar)
+{
+	if (nv_device_is_pci(device)) {
+		return pci_resource_start(device->pdev, bar);
+	} else {
+		struct resource *res;
+		res = platform_get_resource(device->platformdev,
+					    IORESOURCE_MEM, bar);
+		if (!res)
+			return 0;
+		return res->start;
+	}
+}
+
+resource_size_t
+nv_device_resource_len(struct nouveau_device *device, unsigned int bar)
+{
+	if (nv_device_is_pci(device)) {
+		return pci_resource_len(device->pdev, bar);
+	} else {
+		struct resource *res;
+		res = platform_get_resource(device->platformdev,
+					    IORESOURCE_MEM, bar);
+		if (!res)
+			return 0;
+		return resource_size(res);
+	}
+}
+
+dma_addr_t
+nv_device_map_page(struct nouveau_device *device, struct page *page)
+{
+	dma_addr_t ret;
+
+	if (nv_device_is_pci(device)) {
+		ret = pci_map_page(device->pdev, page, 0, PAGE_SIZE,
+				   PCI_DMA_BIDIRECTIONAL);
+		if (pci_dma_mapping_error(device->pdev, ret))
+			ret = 0;
+	} else {
+		ret = page_to_phys(page);
+	}
+
+	return ret;
+}
+
+void
+nv_device_unmap_page(struct nouveau_device *device, dma_addr_t addr)
+{
+	if (nv_device_is_pci(device))
+		pci_unmap_page(device->pdev, addr, PAGE_SIZE,
+			       PCI_DMA_BIDIRECTIONAL);
+}
+
+int
+nv_device_get_irq(struct nouveau_device *device, bool stall)
+{
+	if (nv_device_is_pci(device)) {
+		return device->pdev->irq;
+	} else {
+		return platform_get_irq_byname(device->platformdev,
+					       stall ? "stall" : "nonstall");
+	}
+}
+
 static struct nouveau_oclass
 nouveau_device_oclass = {
 	.handle = NV_ENGINE(DEVICE, 0x00),
@@ -457,8 +525,8 @@ nouveau_device_oclass = {
 };
 
 int
-nouveau_device_create_(struct pci_dev *pdev, u64 name, const char *sname,
-		       const char *cfg, const char *dbg,
+nouveau_device_create_(void *dev, enum nv_bus_type type, u64 name,
+		       const char *sname, const char *cfg, const char *dbg,
 		       int length, void **pobject)
 {
 	struct nouveau_device *device;
@@ -476,7 +544,14 @@ nouveau_device_create_(struct pci_dev *pdev, u64 name, const char *sname,
 	if (ret)
 		goto done;
 
-	device->pdev = pdev;
+	switch (type) {
+	case NOUVEAU_BUS_PCI:
+		device->pdev = dev;
+		break;
+	case NOUVEAU_BUS_PLATFORM:
+		device->platformdev = dev;
+		break;
+	}
 	device->handle = name;
 	device->cfgopt = cfg;
 	device->dbgopt = dbg;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/gm100.c b/drivers/gpu/drm/nouveau/core/engine/device/gm100.c
new file mode 100644
index 00000000000..a520029e25d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/device/gm100.c
@@ -0,0 +1,106 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bus.h>
+#include <subdev/gpio.h>
+#include <subdev/i2c.h>
+#include <subdev/clock.h>
+#include <subdev/therm.h>
+#include <subdev/mxm.h>
+#include <subdev/devinit.h>
+#include <subdev/mc.h>
+#include <subdev/timer.h>
+#include <subdev/fb.h>
+#include <subdev/ltcg.h>
+#include <subdev/ibus.h>
+#include <subdev/instmem.h>
+#include <subdev/vm.h>
+#include <subdev/bar.h>
+#include <subdev/pwr.h>
+#include <subdev/volt.h>
+
+#include <engine/device.h>
+#include <engine/dmaobj.h>
+#include <engine/fifo.h>
+#include <engine/software.h>
+#include <engine/graph.h>
+#include <engine/disp.h>
+#include <engine/copy.h>
+#include <engine/bsp.h>
+#include <engine/vp.h>
+#include <engine/ppp.h>
+#include <engine/perfmon.h>
+
+int
+gm100_identify(struct nouveau_device *device)
+{
+	switch (device->chipset) {
+	case 0x117:
+		device->cname = "GM107";
+		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
+#if 0
+		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
+#endif
+		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  gm107_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
+		device->oclass[NVDEV_SUBDEV_TIMER  ] = &gk20a_timer_oclass;
+		device->oclass[NVDEV_SUBDEV_FB     ] =  gm107_fb_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gm107_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
+		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
+#if 0
+		device->oclass[NVDEV_SUBDEV_PWR    ] = &nv108_pwr_oclass;
+		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
+#endif
+		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvd0_dmaeng_oclass;
+		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv108_fifo_oclass;
+		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
+		device->oclass[NVDEV_ENGINE_GR     ] =  gm107_graph_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  gm107_disp_oclass;
+		device->oclass[NVDEV_ENGINE_COPY0  ] = &nve0_copy0_oclass;
+#if 0
+		device->oclass[NVDEV_ENGINE_COPY1  ] = &nve0_copy1_oclass;
+#endif
+		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
+#if 0
+		device->oclass[NVDEV_ENGINE_BSP    ] = &nve0_bsp_oclass;
+		device->oclass[NVDEV_ENGINE_VP     ] = &nve0_vp_oclass;
+		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
+#endif
+		break;
+	default:
+		nv_fatal(device, "unknown Maxwell chipset\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv04.c b/drivers/gpu/drm/nouveau/core/engine/device/nv04.c
index dbd2dde7b7e..40b29d0214c 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv04.c
@@ -47,38 +47,38 @@ nv04_identify(struct nouveau_device *device)
 	case 0x04:
 		device->cname = "NV04";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv04_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv04_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv04_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv04_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv04_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv04_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x05:
 		device->cname = "NV05";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv05_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv05_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv04_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv04_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv04_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv04_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	default:
 		nv_fatal(device, "unknown RIVA chipset\n");
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv10.c b/drivers/gpu/drm/nouveau/core/engine/device/nv10.c
index 6e03dd6abee..5f7c25ff523 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv10.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv10.c
@@ -48,152 +48,152 @@ nv10_identify(struct nouveau_device *device)
 	case 0x10:
 		device->cname = "NV10";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv10_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x15:
 		device->cname = "NV15";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv10_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv10_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x16:
 		device->cname = "NV16";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv10_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv10_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x1a:
 		device->cname = "nForce";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv1a_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv10_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv10_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x11:
 		device->cname = "NV11";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv10_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv10_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x17:
 		device->cname = "NV17";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv10_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x1f:
 		device->cname = "nForce2";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv1a_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv10_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x18:
 		device->cname = "NV18";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv10_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	default:
 		nv_fatal(device, "unknown Celsius chipset\n");
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv20.c b/drivers/gpu/drm/nouveau/core/engine/device/nv20.c
index dcde53b9f07..75fed11bba0 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv20.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv20.c
@@ -49,78 +49,78 @@ nv20_identify(struct nouveau_device *device)
 	case 0x20:
 		device->cname = "NV20";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv20_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv20_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x25:
 		device->cname = "NV25";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv25_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv25_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x28:
 		device->cname = "NV28";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv25_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv25_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x2a:
 		device->cname = "NV2A";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv25_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv2a_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	default:
 		nv_fatal(device, "unknown Kelvin chipset\n");
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv30.c b/drivers/gpu/drm/nouveau/core/engine/device/nv30.c
index 7b8662ef4f5..36919d7db7c 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv30.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv30.c
@@ -49,100 +49,100 @@ nv30_identify(struct nouveau_device *device)
 	case 0x30:
 		device->cname = "NV30";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv30_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv30_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x35:
 		device->cname = "NV35";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv35_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv35_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x31:
 		device->cname = "NV31";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv30_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv30_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv31_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x36:
 		device->cname = "NV36";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv36_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv35_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv31_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	case 0x34:
 		device->cname = "NV34";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv34_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv31_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		break;
 	default:
 		nv_fatal(device, "unknown Rankine chipset\n");
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv40.c b/drivers/gpu/drm/nouveau/core/engine/device/nv40.c
index c8c41e93695..1130a62be2c 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv40.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv40.c
@@ -53,16 +53,16 @@ nv40_identify(struct nouveau_device *device)
 	case 0x40:
 		device->cname = "NV40";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv40_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -70,22 +70,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv40_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x41:
 		device->cname = "NV41";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv41_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -93,22 +93,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv40_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x42:
 		device->cname = "NV42";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv41_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -116,22 +116,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv40_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x43:
 		device->cname = "NV43";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv41_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -139,22 +139,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv40_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x45:
 		device->cname = "NV45";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv40_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -162,22 +162,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x47:
 		device->cname = "G70";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv47_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -185,22 +185,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x49:
 		device->cname = "G71";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv49_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -208,22 +208,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x4b:
 		device->cname = "G73";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv49_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -231,22 +231,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x44:
 		device->cname = "NV44";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv44_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -254,22 +254,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x46:
 		device->cname = "G72";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -277,22 +277,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x4a:
 		device->cname = "NV44A";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv44_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -300,22 +300,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x4c:
 		device->cname = "C61";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
-		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_MC     ] =  nv4c_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -323,22 +323,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x4e:
 		device->cname = "C51";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv4e_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv4e_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
-		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_MC     ] =  nv4c_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv4e_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -346,22 +346,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x63:
 		device->cname = "C73";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
-		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_MC     ] =  nv4c_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -369,22 +369,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x67:
 		device->cname = "C67";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
-		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_MC     ] =  nv4c_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -392,22 +392,22 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	case 0x68:
 		device->cname = "C68";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv10_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
-		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_MC     ] =  nv4c_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -415,7 +415,7 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv10_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv40_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv44_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv04_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv04_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv40_perfmon_oclass;
 		break;
 	default:
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv50.c b/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
index db3fc7be856..ef0b0bde1a9 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
@@ -60,17 +60,17 @@ nv50_identify(struct nouveau_device *device)
 	case 0x50:
 		device->cname = "G80";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv50_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv50_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv50_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv50_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv50_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv50_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -79,23 +79,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_SW     ] =  nv50_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv50_graph_oclass;
 		device->oclass[NVDEV_ENGINE_MPEG   ] = &nv50_mpeg_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv50_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv50_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv50_perfmon_oclass;
 		break;
 	case 0x84:
 		device->cname = "G84";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv50_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv50_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -107,23 +107,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_VP     ] = &nv84_vp_oclass;
 		device->oclass[NVDEV_ENGINE_CRYPT  ] = &nv84_crypt_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv84_bsp_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv84_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv84_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv84_perfmon_oclass;
 		break;
 	case 0x86:
 		device->cname = "G86";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv50_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv50_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv50_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -135,23 +135,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_VP     ] = &nv84_vp_oclass;
 		device->oclass[NVDEV_ENGINE_CRYPT  ] = &nv84_crypt_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv84_bsp_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv84_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv84_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv84_perfmon_oclass;
 		break;
 	case 0x92:
 		device->cname = "G92";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv50_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv50_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -163,23 +163,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_VP     ] = &nv84_vp_oclass;
 		device->oclass[NVDEV_ENGINE_CRYPT  ] = &nv84_crypt_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv84_bsp_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv84_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv84_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv84_perfmon_oclass;
 		break;
 	case 0x94:
 		device->cname = "G94";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv94_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -191,23 +191,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_VP     ] = &nv84_vp_oclass;
 		device->oclass[NVDEV_ENGINE_CRYPT  ] = &nv84_crypt_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv84_bsp_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv94_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv94_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv84_perfmon_oclass;
 		break;
 	case 0x96:
 		device->cname = "G96";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv94_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -219,23 +219,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_VP     ] = &nv84_vp_oclass;
 		device->oclass[NVDEV_ENGINE_CRYPT  ] = &nv84_crypt_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv84_bsp_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv94_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv94_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv84_perfmon_oclass;
 		break;
 	case 0x98:
 		device->cname = "G98";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv98_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -247,23 +247,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_CRYPT  ] = &nv98_crypt_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv98_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nv98_ppp_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv94_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv94_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv84_perfmon_oclass;
 		break;
 	case 0xa0:
 		device->cname = "G200";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv50_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv50_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -275,23 +275,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_VP     ] = &nv84_vp_oclass;
 		device->oclass[NVDEV_ENGINE_CRYPT  ] = &nv84_crypt_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv84_bsp_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva0_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva0_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv84_perfmon_oclass;
 		break;
 	case 0xaa:
 		device->cname = "MCP77/MCP78";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nvaa_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv98_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvaa_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -303,23 +303,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_CRYPT  ] = &nv98_crypt_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv98_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nv98_ppp_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv94_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv94_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv84_perfmon_oclass;
 		break;
 	case 0xac:
 		device->cname = "MCP79/MCP7A";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nvaa_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv98_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvaa_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -331,23 +331,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_CRYPT  ] = &nv98_crypt_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv98_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nv98_ppp_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nv94_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nv94_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nv84_perfmon_oclass;
 		break;
 	case 0xa3:
 		device->cname = "GT215";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nva3_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nva3_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nva3_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nva3_pwr_oclass;
@@ -361,23 +361,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv98_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nv98_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nva3_copy_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nva3_perfmon_oclass;
 		break;
 	case 0xa5:
 		device->cname = "GT216";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nva3_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nva3_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nva3_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nva3_pwr_oclass;
@@ -390,23 +390,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv98_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nv98_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nva3_copy_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nva3_perfmon_oclass;
 		break;
 	case 0xa8:
 		device->cname = "GT218";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nva3_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nva3_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nva3_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nva3_pwr_oclass;
@@ -419,23 +419,23 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv98_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nv98_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nva3_copy_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nva3_perfmon_oclass;
 		break;
 	case 0xaf:
 		device->cname = "MCP89";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nva3_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvaf_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvaf_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nva3_pwr_oclass;
@@ -448,7 +448,7 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nv98_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nv98_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nva3_copy_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] =  nva3_perfmon_oclass;
 		break;
 	default:
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
index dbc5e33de94..8d55ed633b1 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
@@ -60,19 +60,19 @@ nvc0_identify(struct nouveau_device *device)
 	case 0xc0:
 		device->cname = "GF100";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc0_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -86,25 +86,25 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nvc0_copy1_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	case 0xc4:
 		device->cname = "GF104";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc0_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -112,31 +112,31 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvc0_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nvc0_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
-		device->oclass[NVDEV_ENGINE_GR     ] =  nvc3_graph_oclass;
+		device->oclass[NVDEV_ENGINE_GR     ] =  nvc4_graph_oclass;
 		device->oclass[NVDEV_ENGINE_VP     ] = &nvc0_vp_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nvc0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nvc0_copy1_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	case 0xc3:
 		device->cname = "GF106";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -144,30 +144,30 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvc0_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nvc0_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
-		device->oclass[NVDEV_ENGINE_GR     ] =  nvc3_graph_oclass;
+		device->oclass[NVDEV_ENGINE_GR     ] =  nvc4_graph_oclass;
 		device->oclass[NVDEV_ENGINE_VP     ] = &nvc0_vp_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nvc0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	case 0xce:
 		device->cname = "GF114";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc0_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -175,31 +175,31 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvc0_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nvc0_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
-		device->oclass[NVDEV_ENGINE_GR     ] =  nvc3_graph_oclass;
+		device->oclass[NVDEV_ENGINE_GR     ] =  nvc4_graph_oclass;
 		device->oclass[NVDEV_ENGINE_VP     ] = &nvc0_vp_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nvc0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nvc0_copy1_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	case 0xcf:
 		device->cname = "GF116";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -207,31 +207,31 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvc0_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nvc0_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
-		device->oclass[NVDEV_ENGINE_GR     ] =  nvc3_graph_oclass;
+		device->oclass[NVDEV_ENGINE_GR     ] =  nvc4_graph_oclass;
 		device->oclass[NVDEV_ENGINE_VP     ] = &nvc0_vp_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nvc0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nvc0_copy1_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	case 0xc1:
 		device->cname = "GF108";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -244,25 +244,25 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nvc0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	case 0xc8:
 		device->cname = "GF110";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv50_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv94_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nv92_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nv94_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc0_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -276,25 +276,25 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nvc0_copy1_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nva3_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nva3_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	case 0xd9:
 		device->cname = "GF119";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nvd0_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nvd0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nvd0_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -307,25 +307,25 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nvc0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nvd0_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nvd0_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	case 0xd7:
 		device->cname = "GF117";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nvd0_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nvd0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  gf117_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvd0_dmaeng_oclass;
@@ -336,7 +336,7 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nvc0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nvc0_copy0_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nvd0_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nvd0_disp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvc0_perfmon_oclass;
 		break;
 	default:
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nve0.c b/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
index 3900104976f..2d1e97d4264 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
@@ -60,19 +60,19 @@ nve0_identify(struct nouveau_device *device)
 	case 0xe4:
 		device->cname = "GK104";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nve0_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -81,7 +81,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nve0_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] =  nve4_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nve0_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nve0_disp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nve0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nve0_copy1_oclass;
 		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
@@ -93,19 +93,19 @@ nve0_identify(struct nouveau_device *device)
 	case 0xe7:
 		device->cname = "GK107";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nve0_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -114,7 +114,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nve0_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] =  nve4_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nve0_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nve0_disp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nve0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nve0_copy1_oclass;
 		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
@@ -126,19 +126,19 @@ nve0_identify(struct nouveau_device *device)
 	case 0xe6:
 		device->cname = "GK106";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nve0_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -147,7 +147,7 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nve0_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] =  nve4_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nve0_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nve0_disp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nve0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nve0_copy1_oclass;
 		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
@@ -156,22 +156,72 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nve0_perfmon_oclass;
 		break;
+	case 0xea:
+		device->cname = "GK20A";
+		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
+		device->oclass[NVDEV_SUBDEV_TIMER  ] = &gk20a_timer_oclass;
+		device->oclass[NVDEV_SUBDEV_FB     ] =  gk20a_fb_oclass;
+		device->oclass[NVDEV_SUBDEV_IBUS   ] = &gk20a_ibus_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] = nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
+		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
+		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvd0_dmaeng_oclass;
+		device->oclass[NVDEV_ENGINE_FIFO   ] =  gk20a_fifo_oclass;
+		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
+		device->oclass[NVDEV_ENGINE_GR     ] =  gk20a_graph_oclass;
+		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
+		device->oclass[NVDEV_ENGINE_PERFMON] = &nve0_perfmon_oclass;
+		break;
 	case 0xf0:
 		device->cname = "GK110";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nve0_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
+		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
+		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
+		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
+		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
+		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
+		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
+		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
+		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
+		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvd0_dmaeng_oclass;
+		device->oclass[NVDEV_ENGINE_FIFO   ] =  nve0_fifo_oclass;
+		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
+		device->oclass[NVDEV_ENGINE_GR     ] =  nvf0_graph_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nvf0_disp_oclass;
+		device->oclass[NVDEV_ENGINE_COPY0  ] = &nve0_copy0_oclass;
+		device->oclass[NVDEV_ENGINE_COPY1  ] = &nve0_copy1_oclass;
+		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
+		device->oclass[NVDEV_ENGINE_BSP    ] = &nve0_bsp_oclass;
+		device->oclass[NVDEV_ENGINE_VP     ] = &nve0_vp_oclass;
+		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
+		device->oclass[NVDEV_ENGINE_PERFMON] = &nvf0_perfmon_oclass;
+		break;
+	case 0xf1:
+		device->cname = "GK110B";
+		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nvd0_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -180,52 +230,46 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nve0_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] =  nvf0_graph_oclass;
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nvf0_disp_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nvf0_disp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nve0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nve0_copy1_oclass;
 		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
-#if 0
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nve0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_VP     ] = &nve0_vp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
-#endif
 		device->oclass[NVDEV_ENGINE_PERFMON] = &nvf0_perfmon_oclass;
 		break;
 	case 0x108:
 		device->cname = "GK208";
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
-		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nve0_gpio_oclass;
-		device->oclass[NVDEV_SUBDEV_I2C    ] = &nvd0_i2c_oclass;
+		device->oclass[NVDEV_SUBDEV_GPIO   ] =  nve0_gpio_oclass;
+		device->oclass[NVDEV_SUBDEV_I2C    ] =  nve0_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
+		device->oclass[NVDEV_SUBDEV_LTCG   ] =  gf100_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nv108_pwr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvd0_dmaeng_oclass;
-#if 0
-		device->oclass[NVDEV_ENGINE_FIFO   ] =  nve0_fifo_oclass;
+		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv108_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
-		device->oclass[NVDEV_ENGINE_GR     ] =  nvf0_graph_oclass;
-#endif
-		device->oclass[NVDEV_ENGINE_DISP   ] = &nvf0_disp_oclass;
-#if 0
+		device->oclass[NVDEV_ENGINE_GR     ] =  nv108_graph_oclass;
+		device->oclass[NVDEV_ENGINE_DISP   ] =  nvf0_disp_oclass;
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nve0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nve0_copy1_oclass;
 		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nve0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_VP     ] = &nve0_vp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
-#endif
 		break;
 	default:
 		nv_fatal(device, "unknown Kepler chipset\n");
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/base.c b/drivers/gpu/drm/nouveau/core/engine/disp/base.c
index 7a5cae42834..9c38c5e4050 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/base.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/base.c
@@ -22,13 +22,89 @@
  * Authors: Ben Skeggs
  */
 
-#include <engine/disp.h>
+#include "priv.h"
+#include "outp.h"
+#include "conn.h"
+
+static int
+nouveau_disp_hpd_check(struct nouveau_event *event, u32 types, int index)
+{
+	struct nouveau_disp *disp = event->priv;
+	struct nvkm_output *outp;
+	list_for_each_entry(outp, &disp->outp, head) {
+		if (outp->conn->index == index) {
+			if (outp->conn->hpd.event)
+				return 0;
+			break;
+		}
+	}
+	return -ENOSYS;
+}
+
+int
+_nouveau_disp_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nouveau_disp *disp = (void *)object;
+	struct nvkm_output *outp;
+	int ret;
+
+	list_for_each_entry(outp, &disp->outp, head) {
+		ret = nv_ofuncs(outp)->fini(nv_object(outp), suspend);
+		if (ret && suspend)
+			goto fail_outp;
+	}
+
+	return nouveau_engine_fini(&disp->base, suspend);
+
+fail_outp:
+	list_for_each_entry_continue_reverse(outp, &disp->outp, head) {
+		nv_ofuncs(outp)->init(nv_object(outp));
+	}
+
+	return ret;
+}
+
+int
+_nouveau_disp_init(struct nouveau_object *object)
+{
+	struct nouveau_disp *disp = (void *)object;
+	struct nvkm_output *outp;
+	int ret;
+
+	ret = nouveau_engine_init(&disp->base);
+	if (ret)
+		return ret;
+
+	list_for_each_entry(outp, &disp->outp, head) {
+		ret = nv_ofuncs(outp)->init(nv_object(outp));
+		if (ret)
+			goto fail_outp;
+	}
+
+	return ret;
+
+fail_outp:
+	list_for_each_entry_continue_reverse(outp, &disp->outp, head) {
+		nv_ofuncs(outp)->fini(nv_object(outp), false);
+	}
+
+	return ret;
+}
 
 void
 _nouveau_disp_dtor(struct nouveau_object *object)
 {
 	struct nouveau_disp *disp = (void *)object;
+	struct nvkm_output *outp, *outt;
+
 	nouveau_event_destroy(&disp->vblank);
+
+	if (disp->outp.next) {
+		list_for_each_entry_safe(outp, outt, &disp->outp, head) {
+			nouveau_object_ref(NULL, (struct nouveau_object **)&outp);
+		}
+	}
+
 	nouveau_engine_destroy(&disp->base);
 }
 
@@ -39,8 +115,15 @@ nouveau_disp_create_(struct nouveau_object *parent,
 		     const char *intname, const char *extname,
 		     int length, void **pobject)
 {
+	struct nouveau_disp_impl *impl = (void *)oclass;
+	struct nouveau_bios *bios = nouveau_bios(parent);
 	struct nouveau_disp *disp;
-	int ret;
+	struct nouveau_oclass **sclass;
+	struct nouveau_object *object;
+	struct dcb_output dcbE;
+	u8  hpd = 0, ver, hdr;
+	u32 data;
+	int ret, i;
 
 	ret = nouveau_engine_create_(parent, engine, oclass, true,
 				     intname, extname, length, pobject);
@@ -48,5 +131,42 @@ nouveau_disp_create_(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	return nouveau_event_create(heads, &disp->vblank);
+	INIT_LIST_HEAD(&disp->outp);
+
+	/* create output objects for each display path in the vbios */
+	i = -1;
+	while ((data = dcb_outp_parse(bios, ++i, &ver, &hdr, &dcbE))) {
+		if (dcbE.type == DCB_OUTPUT_UNUSED)
+			continue;
+		if (dcbE.type == DCB_OUTPUT_EOL)
+			break;
+		data = dcbE.location << 4 | dcbE.type;
+
+		oclass = nvkm_output_oclass;
+		sclass = impl->outp;
+		while (sclass && sclass[0]) {
+			if (sclass[0]->handle == data) {
+				oclass = sclass[0];
+				break;
+			}
+			sclass++;
+		}
+
+		nouveau_object_ctor(*pobject, *pobject, oclass,
+				    &dcbE, i, &object);
+		hpd = max(hpd, (u8)(dcbE.connector + 1));
+	}
+
+	ret = nouveau_event_create(3, hpd, &disp->hpd);
+	if (ret)
+		return ret;
+
+	disp->hpd->priv = disp;
+	disp->hpd->check = nouveau_disp_hpd_check;
+
+	ret = nouveau_event_create(1, heads, &disp->vblank);
+	if (ret)
+		return ret;
+
+	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/conn.c b/drivers/gpu/drm/nouveau/core/engine/disp/conn.c
new file mode 100644
index 00000000000..4ffbc70ecf5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/conn.c
@@ -0,0 +1,172 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/gpio.h>
+
+#include "conn.h"
+#include "outp.h"
+
+static void
+nvkm_connector_hpd_work(struct work_struct *w)
+{
+	struct nvkm_connector *conn = container_of(w, typeof(*conn), hpd.work);
+	struct nouveau_disp *disp = nouveau_disp(conn);
+	struct nouveau_gpio *gpio = nouveau_gpio(conn);
+	u32 send = NVKM_HPD_UNPLUG;
+	if (gpio->get(gpio, 0, DCB_GPIO_UNUSED, conn->hpd.event->index))
+		send = NVKM_HPD_PLUG;
+	nouveau_event_trigger(disp->hpd, send, conn->index);
+	nouveau_event_get(conn->hpd.event);
+}
+
+static int
+nvkm_connector_hpd(void *data, u32 type, int index)
+{
+	struct nvkm_connector *conn = data;
+	DBG("HPD: %d\n", type);
+	schedule_work(&conn->hpd.work);
+	return NVKM_EVENT_DROP;
+}
+
+int
+_nvkm_connector_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nvkm_connector *conn = (void *)object;
+	if (conn->hpd.event)
+		nouveau_event_put(conn->hpd.event);
+	return nouveau_object_fini(&conn->base, suspend);
+}
+
+int
+_nvkm_connector_init(struct nouveau_object *object)
+{
+	struct nvkm_connector *conn = (void *)object;
+	int ret = nouveau_object_init(&conn->base);
+	if (ret == 0) {
+		if (conn->hpd.event)
+			nouveau_event_get(conn->hpd.event);
+	}
+	return ret;
+}
+
+void
+_nvkm_connector_dtor(struct nouveau_object *object)
+{
+	struct nvkm_connector *conn = (void *)object;
+	nouveau_event_ref(NULL, &conn->hpd.event);
+	nouveau_object_destroy(&conn->base);
+}
+
+int
+nvkm_connector_create_(struct nouveau_object *parent,
+		       struct nouveau_object *engine,
+		       struct nouveau_oclass *oclass,
+		       struct nvbios_connE *info, int index,
+		       int length, void **pobject)
+{
+	static const u8 hpd[] = { 0x07, 0x08, 0x51, 0x52, 0x5e, 0x5f, 0x60 };
+	struct nouveau_gpio *gpio = nouveau_gpio(parent);
+	struct nouveau_disp *disp = (void *)engine;
+	struct nvkm_connector *conn;
+	struct nvkm_output *outp;
+	struct dcb_gpio_func func;
+	int ret;
+
+	list_for_each_entry(outp, &disp->outp, head) {
+		if (outp->conn && outp->conn->index == index) {
+			atomic_inc(&nv_object(outp->conn)->refcount);
+			*pobject = outp->conn;
+			return 1;
+		}
+	}
+
+	ret = nouveau_object_create_(parent, engine, oclass, 0, length, pobject);
+	conn = *pobject;
+	if (ret)
+		return ret;
+
+	conn->info = *info;
+	conn->index = index;
+
+	DBG("type %02x loc %d hpd %02x dp %x di %x sr %x lcdid %x\n",
+	    info->type, info->location, info->hpd, info->dp,
+	    info->di, info->sr, info->lcdid);
+
+	if ((info->hpd = ffs(info->hpd))) {
+		if (--info->hpd >= ARRAY_SIZE(hpd)) {
+			ERR("hpd %02x unknown\n", info->hpd);
+			goto done;
+		}
+		info->hpd = hpd[info->hpd];
+
+		ret = gpio->find(gpio, 0, info->hpd, DCB_GPIO_UNUSED, &func);
+		if (ret) {
+			ERR("func %02x lookup failed, %d\n", info->hpd, ret);
+			goto done;
+		}
+
+		ret = nouveau_event_new(gpio->events, NVKM_GPIO_TOGGLED,
+					func.line, nvkm_connector_hpd,
+					conn, &conn->hpd.event);
+		if (ret) {
+			ERR("func %02x failed, %d\n", info->hpd, ret);
+		} else {
+			DBG("func %02x (HPD)\n", info->hpd);
+		}
+	}
+
+done:
+	INIT_WORK(&conn->hpd.work, nvkm_connector_hpd_work);
+	return 0;
+}
+
+int
+_nvkm_connector_ctor(struct nouveau_object *parent,
+		     struct nouveau_object *engine,
+		     struct nouveau_oclass *oclass, void *info, u32 index,
+		     struct nouveau_object **pobject)
+{
+	struct nvkm_connector *conn;
+	int ret;
+
+	ret = nvkm_connector_create(parent, engine, oclass, info, index, &conn);
+	*pobject = nv_object(conn);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+struct nouveau_oclass *
+nvkm_connector_oclass = &(struct nvkm_connector_impl) {
+	.base = {
+		.handle = 0,
+		.ofuncs = &(struct nouveau_ofuncs) {
+			.ctor = _nvkm_connector_ctor,
+			.dtor = _nvkm_connector_dtor,
+			.init = _nvkm_connector_init,
+			.fini = _nvkm_connector_fini,
+		},
+	},
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/conn.h b/drivers/gpu/drm/nouveau/core/engine/disp/conn.h
new file mode 100644
index 00000000000..035ebeacbb1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/conn.h
@@ -0,0 +1,59 @@
+#ifndef __NVKM_DISP_CONN_H__
+#define __NVKM_DISP_CONN_H__
+
+#include "priv.h"
+
+struct nvkm_connector {
+	struct nouveau_object base;
+	struct list_head head;
+
+	struct nvbios_connE info;
+	int index;
+
+	struct {
+		struct nouveau_eventh *event;
+		struct work_struct work;
+	} hpd;
+};
+
+#define nvkm_connector_create(p,e,c,b,i,d)                                     \
+	nvkm_connector_create_((p), (e), (c), (b), (i), sizeof(**d), (void **)d)
+#define nvkm_connector_destroy(d) ({                                           \
+	struct nvkm_connector *disp = (d);                                     \
+	_nvkm_connector_dtor(nv_object(disp));                                 \
+})
+#define nvkm_connector_init(d) ({                                              \
+	struct nvkm_connector *disp = (d);                                     \
+	_nvkm_connector_init(nv_object(disp));                                 \
+})
+#define nvkm_connector_fini(d,s) ({                                            \
+	struct nvkm_connector *disp = (d);                                     \
+	_nvkm_connector_fini(nv_object(disp), (s));                            \
+})
+
+int nvkm_connector_create_(struct nouveau_object *, struct nouveau_object *,
+			   struct nouveau_oclass *, struct nvbios_connE *,
+			   int, int, void **);
+
+int  _nvkm_connector_ctor(struct nouveau_object *, struct nouveau_object *,
+			  struct nouveau_oclass *, void *, u32,
+			  struct nouveau_object **);
+void _nvkm_connector_dtor(struct nouveau_object *);
+int  _nvkm_connector_init(struct nouveau_object *);
+int  _nvkm_connector_fini(struct nouveau_object *, bool);
+
+struct nvkm_connector_impl {
+	struct nouveau_oclass base;
+};
+
+#ifndef MSG
+#define MSG(l,f,a...) do {                                                     \
+	struct nvkm_connector *_conn = (void *)conn;                           \
+	nv_##l(nv_object(conn)->engine, "%02x:%02x%02x: "f, _conn->index,      \
+	       _conn->info.location, _conn->info.type, ##a);                   \
+} while(0)
+#define DBG(f,a...) MSG(debug, f, ##a)
+#define ERR(f,a...) MSG(error, f, ##a)
+#endif
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/dport.c b/drivers/gpu/drm/nouveau/core/engine/disp/dport.c
index 1bd4c63369c..5a5b59b2113 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/dport.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/dport.c
@@ -30,42 +30,38 @@
 
 #include <engine/disp.h>
 
-#include "dport.h"
+#include <core/class.h>
 
-#define DBG(fmt, args...) nv_debug(dp->disp, "DP:%04x:%04x: " fmt,             \
-				   dp->outp->hasht, dp->outp->hashm, ##args)
-#define ERR(fmt, args...) nv_error(dp->disp, "DP:%04x:%04x: " fmt,             \
-				   dp->outp->hasht, dp->outp->hashm, ##args)
+#include "dport.h"
+#include "outpdp.h"
 
 /******************************************************************************
  * link training
  *****************************************************************************/
 struct dp_state {
-	const struct nouveau_dp_func *func;
-	struct nouveau_disp *disp;
-	struct dcb_output *outp;
-	struct nvbios_dpout info;
-	u8 version;
-	struct nouveau_i2c_port *aux;
-	int head;
-	u8  dpcd[4];
+	struct nvkm_output_dp *outp;
 	int link_nr;
 	u32 link_bw;
 	u8  stat[6];
 	u8  conf[4];
+	bool pc2;
+	u8  pc2stat;
+	u8  pc2conf[2];
 };
 
 static int
 dp_set_link_config(struct dp_state *dp)
 {
-	struct nouveau_disp *disp = dp->disp;
+	struct nvkm_output_dp_impl *impl = (void *)nv_oclass(dp->outp);
+	struct nvkm_output_dp *outp = dp->outp;
+	struct nouveau_disp *disp = nouveau_disp(outp);
 	struct nouveau_bios *bios = nouveau_bios(disp);
 	struct nvbios_init init = {
-		.subdev = nv_subdev(dp->disp),
+		.subdev = nv_subdev(disp),
 		.bios = bios,
 		.offset = 0x0000,
-		.outp = dp->outp,
-		.crtc = dp->head,
+		.outp = &outp->base.info,
+		.crtc = -1,
 		.execute = 1,
 	};
 	u32 lnkcmp;
@@ -75,8 +71,8 @@ dp_set_link_config(struct dp_state *dp)
 	DBG("%d lanes at %d KB/s\n", dp->link_nr, dp->link_bw);
 
 	/* set desired link configuration on the source */
-	if ((lnkcmp = dp->info.lnkcmp)) {
-		if (dp->version < 0x30) {
+	if ((lnkcmp = dp->outp->info.lnkcmp)) {
+		if (outp->version < 0x30) {
 			while ((dp->link_bw / 10) < nv_ro16(bios, lnkcmp))
 				lnkcmp += 4;
 			init.offset = nv_ro16(bios, lnkcmp + 2);
@@ -89,73 +85,112 @@ dp_set_link_config(struct dp_state *dp)
 		nvbios_exec(&init);
 	}
 
-	ret = dp->func->lnk_ctl(dp->disp, dp->outp, dp->head,
-				dp->link_nr, dp->link_bw / 27000,
-				dp->dpcd[DPCD_RC02] &
-					 DPCD_RC02_ENHANCED_FRAME_CAP);
+	ret = impl->lnk_ctl(outp, dp->link_nr, dp->link_bw / 27000,
+			    outp->dpcd[DPCD_RC02] &
+				       DPCD_RC02_ENHANCED_FRAME_CAP);
 	if (ret) {
-		ERR("lnk_ctl failed with %d\n", ret);
+		if (ret < 0)
+			ERR("lnk_ctl failed with %d\n", ret);
 		return ret;
 	}
 
+	impl->lnk_pwr(outp, dp->link_nr);
+
 	/* set desired link configuration on the sink */
 	sink[0] = dp->link_bw / 27000;
 	sink[1] = dp->link_nr;
-	if (dp->dpcd[DPCD_RC02] & DPCD_RC02_ENHANCED_FRAME_CAP)
+	if (outp->dpcd[DPCD_RC02] & DPCD_RC02_ENHANCED_FRAME_CAP)
 		sink[1] |= DPCD_LC01_ENHANCED_FRAME_EN;
 
-	return nv_wraux(dp->aux, DPCD_LC00, sink, 2);
+	return nv_wraux(outp->base.edid, DPCD_LC00_LINK_BW_SET, sink, 2);
 }
 
 static void
 dp_set_training_pattern(struct dp_state *dp, u8 pattern)
 {
+	struct nvkm_output_dp_impl *impl = (void *)nv_oclass(dp->outp);
+	struct nvkm_output_dp *outp = dp->outp;
 	u8 sink_tp;
 
 	DBG("training pattern %d\n", pattern);
-	dp->func->pattern(dp->disp, dp->outp, dp->head, pattern);
+	impl->pattern(outp, pattern);
 
-	nv_rdaux(dp->aux, DPCD_LC02, &sink_tp, 1);
+	nv_rdaux(outp->base.edid, DPCD_LC02, &sink_tp, 1);
 	sink_tp &= ~DPCD_LC02_TRAINING_PATTERN_SET;
 	sink_tp |= pattern;
-	nv_wraux(dp->aux, DPCD_LC02, &sink_tp, 1);
+	nv_wraux(outp->base.edid, DPCD_LC02, &sink_tp, 1);
 }
 
 static int
-dp_link_train_commit(struct dp_state *dp)
+dp_link_train_commit(struct dp_state *dp, bool pc)
 {
-	int i;
+	struct nvkm_output_dp_impl *impl = (void *)nv_oclass(dp->outp);
+	struct nvkm_output_dp *outp = dp->outp;
+	int ret, i;
 
 	for (i = 0; i < dp->link_nr; i++) {
 		u8 lane = (dp->stat[4 + (i >> 1)] >> ((i & 1) * 4)) & 0xf;
+		u8 lpc2 = (dp->pc2stat >> (i * 2)) & 0x3;
 		u8 lpre = (lane & 0x0c) >> 2;
 		u8 lvsw = (lane & 0x03) >> 0;
+		u8 hivs = 3 - lpre;
+		u8 hipe = 3;
+		u8 hipc = 3;
+
+		if (lpc2 >= hipc)
+			lpc2 = hipc | DPCD_LC0F_LANE0_MAX_POST_CURSOR2_REACHED;
+		if (lpre >= hipe) {
+			lpre = hipe | DPCD_LC03_MAX_SWING_REACHED; /* yes. */
+			lvsw = hivs = 3 - (lpre & 3);
+		} else
+		if (lvsw >= hivs) {
+			lvsw = hivs | DPCD_LC03_MAX_SWING_REACHED;
+		}
 
 		dp->conf[i] = (lpre << 3) | lvsw;
-		if (lvsw == 3)
-			dp->conf[i] |= DPCD_LC03_MAX_SWING_REACHED;
-		if (lpre == 3)
-			dp->conf[i] |= DPCD_LC03_MAX_PRE_EMPHASIS_REACHED;
+		dp->pc2conf[i >> 1] |= lpc2 << ((i & 1) * 4);
 
-		DBG("config lane %d %02x\n", i, dp->conf[i]);
-		dp->func->drv_ctl(dp->disp, dp->outp, dp->head, i, lvsw, lpre);
+		DBG("config lane %d %02x %02x\n", i, dp->conf[i], lpc2);
+		impl->drv_ctl(outp, i, lvsw & 3, lpre & 3, lpc2 & 3);
 	}
 
-	return nv_wraux(dp->aux, DPCD_LC03(0), dp->conf, 4);
+	ret = nv_wraux(outp->base.edid, DPCD_LC03(0), dp->conf, 4);
+	if (ret)
+		return ret;
+
+	if (pc) {
+		ret = nv_wraux(outp->base.edid, DPCD_LC0F, dp->pc2conf, 2);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
 }
 
 static int
-dp_link_train_update(struct dp_state *dp, u32 delay)
+dp_link_train_update(struct dp_state *dp, bool pc, u32 delay)
 {
+	struct nvkm_output_dp *outp = dp->outp;
 	int ret;
 
-	udelay(delay);
+	if (outp->dpcd[DPCD_RC0E_AUX_RD_INTERVAL])
+		mdelay(outp->dpcd[DPCD_RC0E_AUX_RD_INTERVAL] * 4);
+	else
+		udelay(delay);
 
-	ret = nv_rdaux(dp->aux, DPCD_LS02, dp->stat, 6);
+	ret = nv_rdaux(outp->base.edid, DPCD_LS02, dp->stat, 6);
 	if (ret)
 		return ret;
 
-	DBG("status %6ph\n", dp->stat);
+	if (pc) {
+		ret = nv_rdaux(outp->base.edid, DPCD_LS0C, &dp->pc2stat, 1);
+		if (ret)
+			dp->pc2stat = 0x00;
+		DBG("status %6ph pc2 %02x\n", dp->stat, dp->pc2stat);
+	} else {
+		DBG("status %6ph\n", dp->stat);
+	}
+
 	return 0;
 }
 
@@ -169,8 +204,8 @@ dp_link_train_cr(struct dp_state *dp)
 	dp_set_training_pattern(dp, 1);
 
 	do {
-		if (dp_link_train_commit(dp) ||
-		    dp_link_train_update(dp, 100))
+		if (dp_link_train_commit(dp, false) ||
+		    dp_link_train_update(dp, false, 100))
 			break;
 
 		cr_done = true;
@@ -196,13 +231,19 @@ dp_link_train_cr(struct dp_state *dp)
 static int
 dp_link_train_eq(struct dp_state *dp)
 {
+	struct nvkm_output_dp *outp = dp->outp;
 	bool eq_done = false, cr_done = true;
 	int tries = 0, i;
 
-	dp_set_training_pattern(dp, 2);
+	if (outp->dpcd[2] & DPCD_RC02_TPS3_SUPPORTED)
+		dp_set_training_pattern(dp, 3);
+	else
+		dp_set_training_pattern(dp, 2);
 
 	do {
-		if (dp_link_train_update(dp, 400))
+		if ((tries &&
+		    dp_link_train_commit(dp, dp->pc2)) ||
+		    dp_link_train_update(dp, dp->pc2, 400))
 			break;
 
 		eq_done = !!(dp->stat[2] & DPCD_LS04_INTERLANE_ALIGN_DONE);
@@ -214,9 +255,6 @@ dp_link_train_eq(struct dp_state *dp)
 			    !(lane & DPCD_LS02_LANE0_SYMBOL_LOCKED))
 				eq_done = false;
 		}
-
-		if (dp_link_train_commit(dp))
-			break;
 	} while (!eq_done && cr_done && ++tries <= 5);
 
 	return eq_done ? 0 : -1;
@@ -225,113 +263,109 @@ dp_link_train_eq(struct dp_state *dp)
 static void
 dp_link_train_init(struct dp_state *dp, bool spread)
 {
+	struct nvkm_output_dp *outp = dp->outp;
+	struct nouveau_disp *disp = nouveau_disp(outp);
+	struct nouveau_bios *bios = nouveau_bios(disp);
 	struct nvbios_init init = {
-		.subdev = nv_subdev(dp->disp),
-		.bios = nouveau_bios(dp->disp),
-		.outp = dp->outp,
-		.crtc = dp->head,
+		.subdev = nv_subdev(disp),
+		.bios = bios,
+		.outp = &outp->base.info,
+		.crtc = -1,
 		.execute = 1,
 	};
 
 	/* set desired spread */
 	if (spread)
-		init.offset = dp->info.script[2];
+		init.offset = outp->info.script[2];
 	else
-		init.offset = dp->info.script[3];
+		init.offset = outp->info.script[3];
 	nvbios_exec(&init);
 
 	/* pre-train script */
-	init.offset = dp->info.script[0];
+	init.offset = outp->info.script[0];
 	nvbios_exec(&init);
 }
 
 static void
 dp_link_train_fini(struct dp_state *dp)
 {
+	struct nvkm_output_dp *outp = dp->outp;
+	struct nouveau_disp *disp = nouveau_disp(outp);
+	struct nouveau_bios *bios = nouveau_bios(disp);
 	struct nvbios_init init = {
-		.subdev = nv_subdev(dp->disp),
-		.bios = nouveau_bios(dp->disp),
-		.outp = dp->outp,
-		.crtc = dp->head,
+		.subdev = nv_subdev(disp),
+		.bios = bios,
+		.outp = &outp->base.info,
+		.crtc = -1,
 		.execute = 1,
 	};
 
 	/* post-train script */
-	init.offset = dp->info.script[1],
+	init.offset = outp->info.script[1],
 	nvbios_exec(&init);
 }
 
-int
-nouveau_dp_train(struct nouveau_disp *disp, const struct nouveau_dp_func *func,
-		 struct dcb_output *outp, int head, u32 datarate)
+static const struct dp_rates {
+	u32 rate;
+	u8  bw;
+	u8  nr;
+} nouveau_dp_rates[] = {
+	{ 2160000, 0x14, 4 },
+	{ 1080000, 0x0a, 4 },
+	{ 1080000, 0x14, 2 },
+	{  648000, 0x06, 4 },
+	{  540000, 0x0a, 2 },
+	{  540000, 0x14, 1 },
+	{  324000, 0x06, 2 },
+	{  270000, 0x0a, 1 },
+	{  162000, 0x06, 1 },
+	{}
+};
+
+void
+nouveau_dp_train(struct work_struct *w)
 {
-	struct nouveau_bios *bios = nouveau_bios(disp);
-	struct nouveau_i2c *i2c = nouveau_i2c(disp);
+	struct nvkm_output_dp *outp = container_of(w, typeof(*outp), lt.work);
+	struct nouveau_disp *disp = nouveau_disp(outp);
+	const struct dp_rates *cfg = nouveau_dp_rates;
 	struct dp_state _dp = {
-		.disp = disp,
-		.func = func,
 		.outp = outp,
-		.head = head,
 	}, *dp = &_dp;
-	const u32 bw_list[] = { 270000, 162000, 0 };
-	const u32 *link_bw = bw_list;
-	u8  hdr, cnt, len;
-	u32 data;
+	u32 datarate = 0;
 	int ret;
 
-	/* find the bios displayport data relevant to this output */
-	data = nvbios_dpout_match(bios, outp->hasht, outp->hashm, &dp->version,
-				 &hdr, &cnt, &len, &dp->info);
-	if (!data) {
-		ERR("bios data not found\n");
-		return -EINVAL;
-	}
-
-	/* acquire the aux channel and fetch some info about the display */
-	if (outp->location)
-		dp->aux = i2c->find_type(i2c, NV_I2C_TYPE_EXTAUX(outp->extdev));
-	else
-		dp->aux = i2c->find(i2c, NV_I2C_TYPE_DCBI2C(outp->i2c_index));
-	if (!dp->aux) {
-		ERR("no aux channel?!\n");
-		return -ENODEV;
+	/* bring capabilities within encoder limits */
+	if (nv_mclass(disp) < NVD0_DISP_CLASS)
+		outp->dpcd[2] &= ~DPCD_RC02_TPS3_SUPPORTED;
+	if ((outp->dpcd[2] & 0x1f) > outp->base.info.dpconf.link_nr) {
+		outp->dpcd[2] &= ~DPCD_RC02_MAX_LANE_COUNT;
+		outp->dpcd[2] |= outp->base.info.dpconf.link_nr;
 	}
-
-	ret = nv_rdaux(dp->aux, 0x00000, dp->dpcd, sizeof(dp->dpcd));
-	if (ret) {
-		/* it's possible the display has been unplugged before we
-		 * get here.  we still need to execute the full set of
-		 * vbios scripts, and program the OR at a high enough
-		 * frequency to satisfy the target mode.  failure to do
-		 * so results at best in an UPDATE hanging, and at worst
-		 * with PDISP running away to join the circus.
-		 */
-		dp->dpcd[1] = link_bw[0] / 27000;
-		dp->dpcd[2] = 4;
-		dp->dpcd[3] = 0x00;
-		ERR("failed to read DPCD\n");
+	if (outp->dpcd[1] > outp->base.info.dpconf.link_bw)
+		outp->dpcd[1] = outp->base.info.dpconf.link_bw;
+	dp->pc2 = outp->dpcd[2] & DPCD_RC02_TPS3_SUPPORTED;
+
+	/* restrict link config to the lowest required rate, if requested */
+	if (datarate) {
+		datarate = (datarate / 8) * 10; /* 8B/10B coding overhead */
+		while (cfg[1].rate >= datarate)
+			cfg++;
 	}
+	cfg--;
 
-	/* adjust required bandwidth for 8B/10B coding overhead */
-	datarate = (datarate / 8) * 10;
+	/* disable link interrupt handling during link training */
+	nouveau_event_put(outp->irq);
 
 	/* enable down-spreading and execute pre-train script from vbios */
-	dp_link_train_init(dp, dp->dpcd[3] & 0x01);
+	dp_link_train_init(dp, outp->dpcd[3] & 0x01);
 
-	/* start off at highest link rate supported by encoder and display */
-	while (*link_bw > (dp->dpcd[1] * 27000))
-		link_bw++;
-
-	while ((ret = -EIO) && link_bw[0]) {
-		/* find minimum required lane count at this link rate */
-		dp->link_nr = dp->dpcd[2] & DPCD_RC02_MAX_LANE_COUNT;
-		while ((dp->link_nr >> 1) * link_bw[0] > datarate)
-			dp->link_nr >>= 1;
-
-		/* drop link rate to minimum with this lane count */
-		while ((link_bw[1] * dp->link_nr) > datarate)
-			link_bw++;
-		dp->link_bw = link_bw[0];
+	while (ret = -EIO, (++cfg)->rate) {
+		/* select next configuration supported by encoder and sink */
+		while (cfg->nr > (outp->dpcd[2] & DPCD_RC02_MAX_LANE_COUNT) ||
+		       cfg->bw > (outp->dpcd[DPCD_RC01_MAX_LINK_RATE]))
+			cfg++;
+		dp->link_bw = cfg->bw * 27000;
+		dp->link_nr = cfg->nr;
 
 		/* program selected link configuration */
 		ret = dp_set_link_config(dp);
@@ -348,17 +382,18 @@ nouveau_dp_train(struct nouveau_disp *disp, const struct nouveau_dp_func *func,
 			 */
 			break;
 		}
-
-		/* retry at lower rate */
-		link_bw++;
 	}
 
-	/* finish link training */
+	/* finish link training and execute post-train script from vbios */
 	dp_set_training_pattern(dp, 0);
 	if (ret < 0)
 		ERR("link training failed\n");
 
-	/* execute post-train script from vbios */
 	dp_link_train_fini(dp);
-	return (ret < 0) ? false : true;
+
+	/* signal completion and enable link interrupt handling */
+	DBG("training complete\n");
+	atomic_set(&outp->lt.done, 1);
+	wake_up(&outp->lt.wait);
+	nouveau_event_get(outp->irq);
 }
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/dport.h b/drivers/gpu/drm/nouveau/core/engine/disp/dport.h
index 0e1bbd18ff6..5628d2d5ec7 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/dport.h
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/dport.h
@@ -2,19 +2,18 @@
 #define __NVKM_DISP_DPORT_H__
 
 /* DPCD Receiver Capabilities */
-#define DPCD_RC00                                                       0x00000
-#define DPCD_RC00_DPCD_REV                                                 0xff
-#define DPCD_RC01                                                       0x00001
-#define DPCD_RC01_MAX_LINK_RATE                                            0xff
+#define DPCD_RC00_DPCD_REV                                              0x00000
+#define DPCD_RC01_MAX_LINK_RATE                                         0x00001
 #define DPCD_RC02                                                       0x00002
 #define DPCD_RC02_ENHANCED_FRAME_CAP                                       0x80
+#define DPCD_RC02_TPS3_SUPPORTED                                           0x40
 #define DPCD_RC02_MAX_LANE_COUNT                                           0x1f
 #define DPCD_RC03                                                       0x00003
 #define DPCD_RC03_MAX_DOWNSPREAD                                           0x01
+#define DPCD_RC0E_AUX_RD_INTERVAL                                       0x0000e
 
 /* DPCD Link Configuration */
-#define DPCD_LC00                                                       0x00100
-#define DPCD_LC00_LINK_BW_SET                                              0xff
+#define DPCD_LC00_LINK_BW_SET                                           0x00100
 #define DPCD_LC01                                                       0x00101
 #define DPCD_LC01_ENHANCED_FRAME_EN                                        0x80
 #define DPCD_LC01_LANE_COUNT_SET                                           0x1f
@@ -25,6 +24,16 @@
 #define DPCD_LC03_PRE_EMPHASIS_SET                                         0x18
 #define DPCD_LC03_MAX_SWING_REACHED                                        0x04
 #define DPCD_LC03_VOLTAGE_SWING_SET                                        0x03
+#define DPCD_LC0F                                                       0x0010f
+#define DPCD_LC0F_LANE1_MAX_POST_CURSOR2_REACHED                           0x40
+#define DPCD_LC0F_LANE1_POST_CURSOR2_SET                                   0x30
+#define DPCD_LC0F_LANE0_MAX_POST_CURSOR2_REACHED                           0x04
+#define DPCD_LC0F_LANE0_POST_CURSOR2_SET                                   0x03
+#define DPCD_LC10                                                       0x00110
+#define DPCD_LC10_LANE3_MAX_POST_CURSOR2_REACHED                           0x40
+#define DPCD_LC10_LANE3_POST_CURSOR2_SET                                   0x30
+#define DPCD_LC10_LANE2_MAX_POST_CURSOR2_REACHED                           0x04
+#define DPCD_LC10_LANE2_POST_CURSOR2_SET                                   0x03
 
 /* DPCD Link/Sink Status */
 #define DPCD_LS02                                                       0x00202
@@ -55,24 +64,12 @@
 #define DPCD_LS07_LANE3_VOLTAGE_SWING                                      0x30
 #define DPCD_LS07_LANE2_PRE_EMPHASIS                                       0x0c
 #define DPCD_LS07_LANE2_VOLTAGE_SWING                                      0x03
+#define DPCD_LS0C                                                       0x0020c
+#define DPCD_LS0C_LANE3_POST_CURSOR2                                       0xc0
+#define DPCD_LS0C_LANE2_POST_CURSOR2                                       0x30
+#define DPCD_LS0C_LANE1_POST_CURSOR2                                       0x0c
+#define DPCD_LS0C_LANE0_POST_CURSOR2                                       0x03
 
-struct nouveau_disp;
-struct dcb_output;
-
-struct nouveau_dp_func {
-	int (*pattern)(struct nouveau_disp *, struct dcb_output *,
-		       int head, int pattern);
-	int (*lnk_ctl)(struct nouveau_disp *, struct dcb_output *, int head,
-		       int link_nr, int link_bw, bool enh_frame);
-	int (*drv_ctl)(struct nouveau_disp *, struct dcb_output *, int head,
-		       int lane, int swing, int preem);
-};
-
-extern const struct nouveau_dp_func nv94_sor_dp_func;
-extern const struct nouveau_dp_func nvd0_sor_dp_func;
-extern const struct nouveau_dp_func nv50_pior_dp_func;
-
-int nouveau_dp_train(struct nouveau_disp *, const struct nouveau_dp_func *,
-		     struct dcb_output *, int, u32);
+void nouveau_dp_train(struct work_struct *);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c b/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c
new file mode 100644
index 00000000000..9fc7447fec9
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/gm107.c
@@ -0,0 +1,101 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <engine/software.h>
+#include <engine/disp.h>
+
+#include <core/class.h>
+
+#include "nv50.h"
+
+/*******************************************************************************
+ * Base display object
+ ******************************************************************************/
+
+static struct nouveau_oclass
+gm107_disp_sclass[] = {
+	{ GM107_DISP_MAST_CLASS, &nvd0_disp_mast_ofuncs },
+	{ GM107_DISP_SYNC_CLASS, &nvd0_disp_sync_ofuncs },
+	{ GM107_DISP_OVLY_CLASS, &nvd0_disp_ovly_ofuncs },
+	{ GM107_DISP_OIMM_CLASS, &nvd0_disp_oimm_ofuncs },
+	{ GM107_DISP_CURS_CLASS, &nvd0_disp_curs_ofuncs },
+	{}
+};
+
+static struct nouveau_oclass
+gm107_disp_base_oclass[] = {
+	{ GM107_DISP_CLASS, &nvd0_disp_base_ofuncs, nvd0_disp_base_omthds },
+	{}
+};
+
+/*******************************************************************************
+ * Display engine implementation
+ ******************************************************************************/
+
+static int
+gm107_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct nv50_disp_priv *priv;
+	int heads = nv_rd32(parent, 0x022448);
+	int ret;
+
+	ret = nouveau_disp_create(parent, engine, oclass, heads,
+				  "PDISP", "display", &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	nv_engine(priv)->sclass = gm107_disp_base_oclass;
+	nv_engine(priv)->cclass = &nv50_disp_cclass;
+	nv_subdev(priv)->intr = nvd0_disp_intr;
+	INIT_WORK(&priv->supervisor, nvd0_disp_intr_supervisor);
+	priv->sclass = gm107_disp_sclass;
+	priv->head.nr = heads;
+	priv->dac.nr = 3;
+	priv->sor.nr = 4;
+	priv->dac.power = nv50_dac_power;
+	priv->dac.sense = nv50_dac_sense;
+	priv->sor.power = nv50_sor_power;
+	priv->sor.hda_eld = nvd0_hda_eld;
+	priv->sor.hdmi = nvd0_hdmi_ctrl;
+	return 0;
+}
+
+struct nouveau_oclass *
+gm107_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x07),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gm107_disp_ctor,
+		.dtor = _nouveau_disp_dtor,
+		.init = _nouveau_disp_init,
+		.fini = _nouveau_disp_fini,
+	},
+	.base.outp =  nvd0_disp_outp_sclass,
+	.mthd.core = &nve0_disp_mast_mthd_chan,
+	.mthd.base = &nvd0_disp_sync_mthd_chan,
+	.mthd.ovly = &nve0_disp_ovly_mthd_chan,
+	.mthd.prev = -0x020000,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv04.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv04.c
index a0bc8a89b69..a32666ed0c4 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv04.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 
-#include <engine/disp.h>
+#include "priv.h"
 
 #include <core/event.h>
 #include <core/class.h>
@@ -31,9 +31,53 @@ struct nv04_disp_priv {
 	struct nouveau_disp base;
 };
 
+static int
+nv04_disp_scanoutpos(struct nouveau_object *object, u32 mthd,
+		     void *data, u32 size)
+{
+	struct nv04_disp_priv *priv = (void *)object->engine;
+	struct nv04_display_scanoutpos *args = data;
+	const int head = (mthd & NV04_DISP_MTHD_HEAD);
+	u32 line;
+
+	if (size < sizeof(*args))
+		return -EINVAL;
+
+	args->vblanks = nv_rd32(priv, 0x680800 + (head * 0x2000)) & 0xffff;
+	args->vtotal  = nv_rd32(priv, 0x680804 + (head * 0x2000)) & 0xffff;
+	args->vblanke = args->vtotal - 1;
+
+	args->hblanks = nv_rd32(priv, 0x680820 + (head * 0x2000)) & 0xffff;
+	args->htotal  = nv_rd32(priv, 0x680824 + (head * 0x2000)) & 0xffff;
+	args->hblanke = args->htotal - 1;
+
+	/*
+	 * If output is vga instead of digital then vtotal/htotal is invalid
+	 * so we have to give up and trigger the timestamping fallback in the
+	 * drm core.
+	 */
+	if (!args->vtotal || !args->htotal)
+		return -ENOTSUPP;
+
+	args->time[0] = ktime_to_ns(ktime_get());
+	line = nv_rd32(priv, 0x600868 + (head * 0x2000));
+	args->time[1] = ktime_to_ns(ktime_get());
+	args->hline = (line & 0xffff0000) >> 16;
+	args->vline = (line & 0x0000ffff);
+	return 0;
+}
+
+#define HEAD_MTHD(n) (n), (n) + 0x01
+
+static struct nouveau_omthds
+nv04_disp_omthds[] = {
+	{ HEAD_MTHD(NV04_DISP_SCANOUTPOS), nv04_disp_scanoutpos },
+	{}
+};
+
 static struct nouveau_oclass
 nv04_disp_sclass[] = {
-	{ NV04_DISP_CLASS, &nouveau_object_ofuncs },
+	{ NV04_DISP_CLASS, &nouveau_object_ofuncs, nv04_disp_omthds },
 	{},
 };
 
@@ -42,13 +86,13 @@ nv04_disp_sclass[] = {
  ******************************************************************************/
 
 static void
-nv04_disp_vblank_enable(struct nouveau_event *event, int head)
+nv04_disp_vblank_enable(struct nouveau_event *event, int type, int head)
 {
 	nv_wr32(event->priv, 0x600140 + (head * 0x2000) , 0x00000001);
 }
 
 static void
-nv04_disp_vblank_disable(struct nouveau_event *event, int head)
+nv04_disp_vblank_disable(struct nouveau_event *event, int type, int head)
 {
 	nv_wr32(event->priv, 0x600140 + (head * 0x2000) , 0x00000000);
 }
@@ -62,12 +106,12 @@ nv04_disp_intr(struct nouveau_subdev *subdev)
 	u32 pvideo;
 
 	if (crtc0 & 0x00000001) {
-		nouveau_event_trigger(priv->base.vblank, 0);
+		nouveau_event_trigger(priv->base.vblank, 1, 0);
 		nv_wr32(priv, 0x600100, 0x00000001);
 	}
 
 	if (crtc1 & 0x00000001) {
-		nouveau_event_trigger(priv->base.vblank, 1);
+		nouveau_event_trigger(priv->base.vblank, 1, 1);
 		nv_wr32(priv, 0x602100, 0x00000001);
 	}
 
@@ -102,13 +146,13 @@ nv04_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	return 0;
 }
 
-struct nouveau_oclass
-nv04_disp_oclass = {
-	.handle = NV_ENGINE(DISP, 0x04),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv04_disp_oclass = &(struct nouveau_disp_impl) {
+	.base.handle = NV_ENGINE(DISP, 0x04),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv04_disp_ctor,
 		.dtor = _nouveau_disp_dtor,
 		.init = _nouveau_disp_init,
 		.fini = _nouveau_disp_fini,
 	},
-};
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
index c168ae3eaa9..2283c442a10 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
@@ -26,8 +26,7 @@
 #include <core/parent.h>
 #include <core/handle.h>
 #include <core/class.h>
-
-#include <engine/disp.h>
+#include <core/enum.h>
 
 #include <subdev/bios.h>
 #include <subdev/bios/dcb.h>
@@ -227,6 +226,177 @@ nv50_disp_dmac_fini(struct nouveau_object *object, bool suspend)
  * EVO master channel object
  ******************************************************************************/
 
+static void
+nv50_disp_mthd_list(struct nv50_disp_priv *priv, int debug, u32 base, int c,
+		    const struct nv50_disp_mthd_list *list, int inst)
+{
+	struct nouveau_object *disp = nv_object(priv);
+	int i;
+
+	for (i = 0; list->data[i].mthd; i++) {
+		if (list->data[i].addr) {
+			u32 next = nv_rd32(priv, list->data[i].addr + base + 0);
+			u32 prev = nv_rd32(priv, list->data[i].addr + base + c);
+			u32 mthd = list->data[i].mthd + (list->mthd * inst);
+			const char *name = list->data[i].name;
+			char mods[16];
+
+			if (prev != next)
+				snprintf(mods, sizeof(mods), "-> 0x%08x", next);
+			else
+				snprintf(mods, sizeof(mods), "%13c", ' ');
+
+			nv_printk_(disp, debug, "\t0x%04x: 0x%08x %s%s%s\n",
+				   mthd, prev, mods, name ? " // " : "",
+				   name ? name : "");
+		}
+	}
+}
+
+void
+nv50_disp_mthd_chan(struct nv50_disp_priv *priv, int debug, int head,
+		    const struct nv50_disp_mthd_chan *chan)
+{
+	struct nouveau_object *disp = nv_object(priv);
+	const struct nv50_disp_impl *impl = (void *)disp->oclass;
+	const struct nv50_disp_mthd_list *list;
+	int i, j;
+
+	if (debug > nv_subdev(priv)->debug)
+		return;
+
+	for (i = 0; (list = chan->data[i].mthd) != NULL; i++) {
+		u32 base = head * chan->addr;
+		for (j = 0; j < chan->data[i].nr; j++, base += list->addr) {
+			const char *cname = chan->name;
+			const char *sname = "";
+			char cname_[16], sname_[16];
+
+			if (chan->addr) {
+				snprintf(cname_, sizeof(cname_), "%s %d",
+					 chan->name, head);
+				cname = cname_;
+			}
+
+			if (chan->data[i].nr > 1) {
+				snprintf(sname_, sizeof(sname_), " - %s %d",
+					 chan->data[i].name, j);
+				sname = sname_;
+			}
+
+			nv_printk_(disp, debug, "%s%s:\n", cname, sname);
+			nv50_disp_mthd_list(priv, debug, base, impl->mthd.prev,
+					    list, j);
+		}
+	}
+}
+
+const struct nv50_disp_mthd_list
+nv50_disp_mast_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0080, 0x000000 },
+		{ 0x0084, 0x610bb8 },
+		{ 0x0088, 0x610b9c },
+		{ 0x008c, 0x000000 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_list
+nv50_disp_mast_mthd_dac = {
+	.mthd = 0x0080,
+	.addr = 0x000008,
+	.data = {
+		{ 0x0400, 0x610b58 },
+		{ 0x0404, 0x610bdc },
+		{ 0x0420, 0x610828 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_list
+nv50_disp_mast_mthd_sor = {
+	.mthd = 0x0040,
+	.addr = 0x000008,
+	.data = {
+		{ 0x0600, 0x610b70 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_list
+nv50_disp_mast_mthd_pior = {
+	.mthd = 0x0040,
+	.addr = 0x000008,
+	.data = {
+		{ 0x0700, 0x610b80 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_list
+nv50_disp_mast_mthd_head = {
+	.mthd = 0x0400,
+	.addr = 0x000540,
+	.data = {
+		{ 0x0800, 0x610ad8 },
+		{ 0x0804, 0x610ad0 },
+		{ 0x0808, 0x610a48 },
+		{ 0x080c, 0x610a78 },
+		{ 0x0810, 0x610ac0 },
+		{ 0x0814, 0x610af8 },
+		{ 0x0818, 0x610b00 },
+		{ 0x081c, 0x610ae8 },
+		{ 0x0820, 0x610af0 },
+		{ 0x0824, 0x610b08 },
+		{ 0x0828, 0x610b10 },
+		{ 0x082c, 0x610a68 },
+		{ 0x0830, 0x610a60 },
+		{ 0x0834, 0x000000 },
+		{ 0x0838, 0x610a40 },
+		{ 0x0840, 0x610a24 },
+		{ 0x0844, 0x610a2c },
+		{ 0x0848, 0x610aa8 },
+		{ 0x084c, 0x610ab0 },
+		{ 0x0860, 0x610a84 },
+		{ 0x0864, 0x610a90 },
+		{ 0x0868, 0x610b18 },
+		{ 0x086c, 0x610b20 },
+		{ 0x0870, 0x610ac8 },
+		{ 0x0874, 0x610a38 },
+		{ 0x0880, 0x610a58 },
+		{ 0x0884, 0x610a9c },
+		{ 0x08a0, 0x610a70 },
+		{ 0x08a4, 0x610a50 },
+		{ 0x08a8, 0x610ae0 },
+		{ 0x08c0, 0x610b28 },
+		{ 0x08c4, 0x610b30 },
+		{ 0x08c8, 0x610b40 },
+		{ 0x08d4, 0x610b38 },
+		{ 0x08d8, 0x610b48 },
+		{ 0x08dc, 0x610b50 },
+		{ 0x0900, 0x610a18 },
+		{ 0x0904, 0x610ab8 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_chan
+nv50_disp_mast_mthd_chan = {
+	.name = "Core",
+	.addr = 0x000000,
+	.data = {
+		{ "Global", 1, &nv50_disp_mast_mthd_base },
+		{    "DAC", 3, &nv50_disp_mast_mthd_dac  },
+		{    "SOR", 2, &nv50_disp_mast_mthd_sor  },
+		{   "PIOR", 3, &nv50_disp_mast_mthd_pior },
+		{   "HEAD", 2, &nv50_disp_mast_mthd_head },
+		{}
+	}
+};
+
 static int
 nv50_disp_mast_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
@@ -323,6 +493,56 @@ nv50_disp_mast_ofuncs = {
  * EVO sync channel objects
  ******************************************************************************/
 
+static const struct nv50_disp_mthd_list
+nv50_disp_sync_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0080, 0x000000 },
+		{ 0x0084, 0x0008c4 },
+		{ 0x0088, 0x0008d0 },
+		{ 0x008c, 0x0008dc },
+		{ 0x0090, 0x0008e4 },
+		{ 0x0094, 0x610884 },
+		{ 0x00a0, 0x6108a0 },
+		{ 0x00a4, 0x610878 },
+		{ 0x00c0, 0x61086c },
+		{ 0x00e0, 0x610858 },
+		{ 0x00e4, 0x610860 },
+		{ 0x00e8, 0x6108ac },
+		{ 0x00ec, 0x6108b4 },
+		{ 0x0100, 0x610894 },
+		{ 0x0110, 0x6108bc },
+		{ 0x0114, 0x61088c },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_list
+nv50_disp_sync_mthd_image = {
+	.mthd = 0x0400,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0800, 0x6108f0 },
+		{ 0x0804, 0x6108fc },
+		{ 0x0808, 0x61090c },
+		{ 0x080c, 0x610914 },
+		{ 0x0810, 0x610904 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_chan
+nv50_disp_sync_mthd_chan = {
+	.name = "Base",
+	.addr = 0x000540,
+	.data = {
+		{ "Global", 1, &nv50_disp_sync_mthd_base },
+		{  "Image", 2, &nv50_disp_sync_mthd_image },
+		{}
+	}
+};
+
 static int
 nv50_disp_sync_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
@@ -362,6 +582,44 @@ nv50_disp_sync_ofuncs = {
  * EVO overlay channel objects
  ******************************************************************************/
 
+const struct nv50_disp_mthd_list
+nv50_disp_ovly_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0080, 0x000000 },
+		{ 0x0084, 0x0009a0 },
+		{ 0x0088, 0x0009c0 },
+		{ 0x008c, 0x0009c8 },
+		{ 0x0090, 0x6109b4 },
+		{ 0x0094, 0x610970 },
+		{ 0x00a0, 0x610998 },
+		{ 0x00a4, 0x610964 },
+		{ 0x00c0, 0x610958 },
+		{ 0x00e0, 0x6109a8 },
+		{ 0x00e4, 0x6109d0 },
+		{ 0x00e8, 0x6109d8 },
+		{ 0x0100, 0x61094c },
+		{ 0x0104, 0x610984 },
+		{ 0x0108, 0x61098c },
+		{ 0x0800, 0x6109f8 },
+		{ 0x0808, 0x610a08 },
+		{ 0x080c, 0x610a10 },
+		{ 0x0810, 0x610a00 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_chan
+nv50_disp_ovly_mthd_chan = {
+	.name = "Overlay",
+	.addr = 0x000540,
+	.data = {
+		{ "Global", 1, &nv50_disp_ovly_mthd_base },
+		{}
+	}
+};
+
 static int
 nv50_disp_ovly_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
@@ -541,14 +799,43 @@ nv50_disp_curs_ofuncs = {
  * Base display object
  ******************************************************************************/
 
+int
+nv50_disp_base_scanoutpos(struct nouveau_object *object, u32 mthd,
+			  void *data, u32 size)
+{
+	struct nv50_disp_priv *priv = (void *)object->engine;
+	struct nv04_display_scanoutpos *args = data;
+	const int head = (mthd & NV50_DISP_MTHD_HEAD);
+	u32 blanke, blanks, total;
+
+	if (size < sizeof(*args) || head >= priv->head.nr)
+		return -EINVAL;
+	blanke = nv_rd32(priv, 0x610aec + (head * 0x540));
+	blanks = nv_rd32(priv, 0x610af4 + (head * 0x540));
+	total  = nv_rd32(priv, 0x610afc + (head * 0x540));
+
+	args->vblanke = (blanke & 0xffff0000) >> 16;
+	args->hblanke = (blanke & 0x0000ffff);
+	args->vblanks = (blanks & 0xffff0000) >> 16;
+	args->hblanks = (blanks & 0x0000ffff);
+	args->vtotal  = ( total & 0xffff0000) >> 16;
+	args->htotal  = ( total & 0x0000ffff);
+
+	args->time[0] = ktime_to_ns(ktime_get());
+	args->vline   = nv_rd32(priv, 0x616340 + (head * 0x800)) & 0xffff;
+	args->time[1] = ktime_to_ns(ktime_get()); /* vline read locks hline */
+	args->hline   = nv_rd32(priv, 0x616344 + (head * 0x800)) & 0xffff;
+	return 0;
+}
+
 static void
-nv50_disp_base_vblank_enable(struct nouveau_event *event, int head)
+nv50_disp_base_vblank_enable(struct nouveau_event *event, int type, int head)
 {
 	nv_mask(event->priv, 0x61002c, (4 << head), (4 << head));
 }
 
 static void
-nv50_disp_base_vblank_disable(struct nouveau_event *event, int head)
+nv50_disp_base_vblank_disable(struct nouveau_event *event, int type, int head)
 {
 	nv_mask(event->priv, 0x61002c, (4 << head), 0);
 }
@@ -675,6 +962,7 @@ nv50_disp_base_ofuncs = {
 
 static struct nouveau_omthds
 nv50_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nv50_disp_base_scanoutpos },
 	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
 	{ SOR_MTHD(NV50_DISP_SOR_LVDS_SCRIPT) , nv50_sor_mthd },
 	{ DAC_MTHD(NV50_DISP_DAC_PWR)         , nv50_dac_mthd },
@@ -752,40 +1040,94 @@ nv50_disp_cclass = {
  * Display engine implementation
  ******************************************************************************/
 
-static void
-nv50_disp_intr_error(struct nv50_disp_priv *priv)
-{
-	u32 channels = (nv_rd32(priv, 0x610020) & 0x001f0000) >> 16;
-	u32 addr, data;
-	int chid;
-
-	for (chid = 0; chid < 5; chid++) {
-		if (!(channels & (1 << chid)))
-			continue;
+static const struct nouveau_enum
+nv50_disp_intr_error_type[] = {
+	{ 3, "ILLEGAL_MTHD" },
+	{ 4, "INVALID_VALUE" },
+	{ 5, "INVALID_STATE" },
+	{ 7, "INVALID_HANDLE" },
+	{}
+};
 
-		nv_wr32(priv, 0x610020, 0x00010000 << chid);
-		addr = nv_rd32(priv, 0x610080 + (chid * 0x08));
-		data = nv_rd32(priv, 0x610084 + (chid * 0x08));
-		nv_wr32(priv, 0x610080 + (chid * 0x08), 0x90000000);
+static const struct nouveau_enum
+nv50_disp_intr_error_code[] = {
+	{ 0x00, "" },
+	{}
+};
 
-		nv_error(priv, "chid %d mthd 0x%04x data 0x%08x 0x%08x\n",
-			 chid, addr & 0xffc, data, addr);
+static void
+nv50_disp_intr_error(struct nv50_disp_priv *priv, int chid)
+{
+	struct nv50_disp_impl *impl = (void *)nv_object(priv)->oclass;
+	u32 data = nv_rd32(priv, 0x610084 + (chid * 0x08));
+	u32 addr = nv_rd32(priv, 0x610080 + (chid * 0x08));
+	u32 code = (addr & 0x00ff0000) >> 16;
+	u32 type = (addr & 0x00007000) >> 12;
+	u32 mthd = (addr & 0x00000ffc);
+	const struct nouveau_enum *ec, *et;
+	char ecunk[6], etunk[6];
+
+	et = nouveau_enum_find(nv50_disp_intr_error_type, type);
+	if (!et)
+		snprintf(etunk, sizeof(etunk), "UNK%02X", type);
+
+	ec = nouveau_enum_find(nv50_disp_intr_error_code, code);
+	if (!ec)
+		snprintf(ecunk, sizeof(ecunk), "UNK%02X", code);
+
+	nv_error(priv, "%s [%s] chid %d mthd 0x%04x data 0x%08x\n",
+		 et ? et->name : etunk, ec ? ec->name : ecunk,
+		 chid, mthd, data);
+
+	if (chid == 0) {
+		switch (mthd) {
+		case 0x0080:
+			nv50_disp_mthd_chan(priv, NV_DBG_ERROR, chid - 0,
+					    impl->mthd.core);
+			break;
+		default:
+			break;
+		}
+	} else
+	if (chid <= 2) {
+		switch (mthd) {
+		case 0x0080:
+			nv50_disp_mthd_chan(priv, NV_DBG_ERROR, chid - 1,
+					    impl->mthd.base);
+			break;
+		default:
+			break;
+		}
+	} else
+	if (chid <= 4) {
+		switch (mthd) {
+		case 0x0080:
+			nv50_disp_mthd_chan(priv, NV_DBG_ERROR, chid - 3,
+					    impl->mthd.ovly);
+			break;
+		default:
+			break;
+		}
 	}
+
+	nv_wr32(priv, 0x610020, 0x00010000 << chid);
+	nv_wr32(priv, 0x610080 + (chid * 0x08), 0x90000000);
 }
 
-static u16
-exec_lookup(struct nv50_disp_priv *priv, int head, int outp, u32 ctrl,
-	    struct dcb_output *dcb, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+static struct nvkm_output *
+exec_lookup(struct nv50_disp_priv *priv, int head, int or, u32 ctrl,
+	    u32 *data, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
 	    struct nvbios_outp *info)
 {
 	struct nouveau_bios *bios = nouveau_bios(priv);
-	u16 mask, type, data;
+	struct nvkm_output *outp;
+	u16 mask, type;
 
-	if (outp < 4) {
+	if (or < 4) {
 		type = DCB_OUTPUT_ANALOG;
 		mask = 0;
 	} else
-	if (outp < 8) {
+	if (or < 8) {
 		switch (ctrl & 0x00000f00) {
 		case 0x00000000: type = DCB_OUTPUT_LVDS; mask = 1; break;
 		case 0x00000100: type = DCB_OUTPUT_TMDS; mask = 1; break;
@@ -795,45 +1137,48 @@ exec_lookup(struct nv50_disp_priv *priv, int head, int outp, u32 ctrl,
 		case 0x00000900: type = DCB_OUTPUT_DP; mask = 2; break;
 		default:
 			nv_error(priv, "unknown SOR mc 0x%08x\n", ctrl);
-			return 0x0000;
+			return NULL;
 		}
-		outp -= 4;
+		or  -= 4;
 	} else {
-		outp = outp - 8;
+		or   = or - 8;
 		type = 0x0010;
 		mask = 0;
 		switch (ctrl & 0x00000f00) {
-		case 0x00000000: type |= priv->pior.type[outp]; break;
+		case 0x00000000: type |= priv->pior.type[or]; break;
 		default:
 			nv_error(priv, "unknown PIOR mc 0x%08x\n", ctrl);
-			return 0x0000;
+			return NULL;
 		}
 	}
 
 	mask  = 0x00c0 & (mask << 6);
-	mask |= 0x0001 << outp;
+	mask |= 0x0001 << or;
 	mask |= 0x0100 << head;
 
-	data = dcb_outp_match(bios, type, mask, ver, hdr, dcb);
-	if (!data)
-		return 0x0000;
-
-	/* off-chip encoders require matching the exact encoder type */
-	if (dcb->location != 0)
-		type |= dcb->extdev << 8;
+	list_for_each_entry(outp, &priv->base.outp, head) {
+		if ((outp->info.hasht & 0xff) == type &&
+		    (outp->info.hashm & mask) == mask) {
+			*data = nvbios_outp_match(bios, outp->info.hasht,
+							outp->info.hashm,
+						  ver, hdr, cnt, len, info);
+			if (!*data)
+				return NULL;
+			return outp;
+		}
+	}
 
-	return nvbios_outp_match(bios, type, mask, ver, hdr, cnt, len, info);
+	return NULL;
 }
 
-static bool
+static struct nvkm_output *
 exec_script(struct nv50_disp_priv *priv, int head, int id)
 {
 	struct nouveau_bios *bios = nouveau_bios(priv);
+	struct nvkm_output *outp;
 	struct nvbios_outp info;
-	struct dcb_output dcb;
 	u8  ver, hdr, cnt, len;
-	u16 data;
-	u32 ctrl = 0x00000000;
+	u32 data, ctrl = 0;
 	u32 reg;
 	int i;
 
@@ -863,36 +1208,35 @@ exec_script(struct nv50_disp_priv *priv, int head, int id)
 	}
 
 	if (!(ctrl & (1 << head)))
-		return false;
+		return NULL;
 	i--;
 
-	data = exec_lookup(priv, head, i, ctrl, &dcb, &ver, &hdr, &cnt, &len, &info);
-	if (data) {
+	outp = exec_lookup(priv, head, i, ctrl, &data, &ver, &hdr, &cnt, &len, &info);
+	if (outp) {
 		struct nvbios_init init = {
 			.subdev = nv_subdev(priv),
 			.bios = bios,
 			.offset = info.script[id],
-			.outp = &dcb,
+			.outp = &outp->info,
 			.crtc = head,
 			.execute = 1,
 		};
 
-		return nvbios_exec(&init) == 0;
+		nvbios_exec(&init);
 	}
 
-	return false;
+	return outp;
 }
 
-static u32
-exec_clkcmp(struct nv50_disp_priv *priv, int head, int id, u32 pclk,
-	    struct dcb_output *outp)
+static struct nvkm_output *
+exec_clkcmp(struct nv50_disp_priv *priv, int head, int id, u32 pclk, u32 *conf)
 {
 	struct nouveau_bios *bios = nouveau_bios(priv);
+	struct nvkm_output *outp;
 	struct nvbios_outp info1;
 	struct nvbios_ocfg info2;
 	u8  ver, hdr, cnt, len;
-	u32 ctrl = 0x00000000;
-	u32 data, conf = ~0;
+	u32 data, ctrl = 0;
 	u32 reg;
 	int i;
 
@@ -922,37 +1266,37 @@ exec_clkcmp(struct nv50_disp_priv *priv, int head, int id, u32 pclk,
 	}
 
 	if (!(ctrl & (1 << head)))
-		return conf;
+		return NULL;
 	i--;
 
-	data = exec_lookup(priv, head, i, ctrl, outp, &ver, &hdr, &cnt, &len, &info1);
-	if (!data)
-		return conf;
+	outp = exec_lookup(priv, head, i, ctrl, &data, &ver, &hdr, &cnt, &len, &info1);
+	if (!outp)
+		return NULL;
 
-	if (outp->location == 0) {
-		switch (outp->type) {
+	if (outp->info.location == 0) {
+		switch (outp->info.type) {
 		case DCB_OUTPUT_TMDS:
-			conf = (ctrl & 0x00000f00) >> 8;
+			*conf = (ctrl & 0x00000f00) >> 8;
 			if (pclk >= 165000)
-				conf |= 0x0100;
+				*conf |= 0x0100;
 			break;
 		case DCB_OUTPUT_LVDS:
-			conf = priv->sor.lvdsconf;
+			*conf = priv->sor.lvdsconf;
 			break;
 		case DCB_OUTPUT_DP:
-			conf = (ctrl & 0x00000f00) >> 8;
+			*conf = (ctrl & 0x00000f00) >> 8;
 			break;
 		case DCB_OUTPUT_ANALOG:
 		default:
-			conf = 0x00ff;
+			*conf = 0x00ff;
 			break;
 		}
 	} else {
-		conf = (ctrl & 0x00000f00) >> 8;
+		*conf = (ctrl & 0x00000f00) >> 8;
 		pclk = pclk / 2;
 	}
 
-	data = nvbios_ocfg_match(bios, data, conf, &ver, &hdr, &cnt, &len, &info2);
+	data = nvbios_ocfg_match(bios, data, *conf, &ver, &hdr, &cnt, &len, &info2);
 	if (data && id < 0xff) {
 		data = nvbios_oclk_match(bios, info2.clkcmp[id], pclk);
 		if (data) {
@@ -960,7 +1304,7 @@ exec_clkcmp(struct nv50_disp_priv *priv, int head, int id, u32 pclk,
 				.subdev = nv_subdev(priv),
 				.bios = bios,
 				.offset = data,
-				.outp = outp,
+				.outp = &outp->info,
 				.crtc = head,
 				.execute = 1,
 			};
@@ -969,7 +1313,7 @@ exec_clkcmp(struct nv50_disp_priv *priv, int head, int id, u32 pclk,
 		}
 	}
 
-	return conf;
+	return outp;
 }
 
 static void
@@ -981,7 +1325,35 @@ nv50_disp_intr_unk10_0(struct nv50_disp_priv *priv, int head)
 static void
 nv50_disp_intr_unk20_0(struct nv50_disp_priv *priv, int head)
 {
-	exec_script(priv, head, 2);
+	struct nvkm_output *outp = exec_script(priv, head, 2);
+
+	/* the binary driver does this outside of the supervisor handling
+	 * (after the third supervisor from a detach).  we (currently?)
+	 * allow both detach/attach to happen in the same set of
+	 * supervisor interrupts, so it would make sense to execute this
+	 * (full power down?) script after all the detach phases of the
+	 * supervisor handling.  like with training if needed from the
+	 * second supervisor, nvidia doesn't do this, so who knows if it's
+	 * entirely safe, but it does appear to work..
+	 *
+	 * without this script being run, on some configurations i've
+	 * seen, switching from DP to TMDS on a DP connector may result
+	 * in a blank screen (SOR_PWR off/on can restore it)
+	 */
+	if (outp && outp->info.type == DCB_OUTPUT_DP) {
+		struct nvkm_output_dp *outpdp = (void *)outp;
+		struct nvbios_init init = {
+			.subdev = nv_subdev(priv),
+			.bios = nouveau_bios(priv),
+			.outp = &outp->info,
+			.crtc = head,
+			.offset = outpdp->info.script[4],
+			.execute = 1,
+		};
+
+		nvbios_exec(&init);
+		atomic_set(&outpdp->lt.done, 0);
+	}
 }
 
 static void
@@ -1103,56 +1475,83 @@ nv50_disp_intr_unk20_2_dp(struct nv50_disp_priv *priv,
 static void
 nv50_disp_intr_unk20_2(struct nv50_disp_priv *priv, int head)
 {
-	struct dcb_output outp;
+	struct nvkm_output *outp;
 	u32 pclk = nv_rd32(priv, 0x610ad0 + (head * 0x540)) & 0x3fffff;
 	u32 hval, hreg = 0x614200 + (head * 0x800);
 	u32 oval, oreg;
-	u32 mask;
-	u32 conf = exec_clkcmp(priv, head, 0xff, pclk, &outp);
-	if (conf != ~0) {
-		if (outp.location == 0 && outp.type == DCB_OUTPUT_DP) {
-			u32 soff = (ffs(outp.or) - 1) * 0x08;
-			u32 ctrl = nv_rd32(priv, 0x610798 + soff);
-			u32 datarate;
-
-			switch ((ctrl & 0x000f0000) >> 16) {
-			case 6: datarate = pclk * 30 / 8; break;
-			case 5: datarate = pclk * 24 / 8; break;
-			case 2:
-			default:
-				datarate = pclk * 18 / 8;
-				break;
-			}
+	u32 mask, conf;
 
-			nouveau_dp_train(&priv->base, priv->sor.dp,
-					 &outp, head, datarate);
-		}
+	outp = exec_clkcmp(priv, head, 0xff, pclk, &conf);
+	if (!outp)
+		return;
 
-		exec_clkcmp(priv, head, 0, pclk, &outp);
-
-		if (!outp.location && outp.type == DCB_OUTPUT_ANALOG) {
-			oreg = 0x614280 + (ffs(outp.or) - 1) * 0x800;
-			oval = 0x00000000;
-			hval = 0x00000000;
-			mask = 0xffffffff;
-		} else
-		if (!outp.location) {
-			if (outp.type == DCB_OUTPUT_DP)
-				nv50_disp_intr_unk20_2_dp(priv, &outp, pclk);
-			oreg = 0x614300 + (ffs(outp.or) - 1) * 0x800;
-			oval = (conf & 0x0100) ? 0x00000101 : 0x00000000;
-			hval = 0x00000000;
-			mask = 0x00000707;
+	/* we allow both encoder attach and detach operations to occur
+	 * within a single supervisor (ie. modeset) sequence.  the
+	 * encoder detach scripts quite often switch off power to the
+	 * lanes, which requires the link to be re-trained.
+	 *
+	 * this is not generally an issue as the sink "must" (heh)
+	 * signal an irq when it's lost sync so the driver can
+	 * re-train.
+	 *
+	 * however, on some boards, if one does not configure at least
+	 * the gpu side of the link *before* attaching, then various
+	 * things can go horribly wrong (PDISP disappearing from mmio,
+	 * third supervisor never happens, etc).
+	 *
+	 * the solution is simply to retrain here, if necessary.  last
+	 * i checked, the binary driver userspace does not appear to
+	 * trigger this situation (it forces an UPDATE between steps).
+	 */
+	if (outp->info.type == DCB_OUTPUT_DP) {
+		u32 soff = (ffs(outp->info.or) - 1) * 0x08;
+		u32 ctrl, datarate;
+
+		if (outp->info.location == 0) {
+			ctrl = nv_rd32(priv, 0x610794 + soff);
+			soff = 1;
 		} else {
-			oreg = 0x614380 + (ffs(outp.or) - 1) * 0x800;
-			oval = 0x00000001;
-			hval = 0x00000001;
-			mask = 0x00000707;
+			ctrl = nv_rd32(priv, 0x610b80 + soff);
+			soff = 2;
+		}
+
+		switch ((ctrl & 0x000f0000) >> 16) {
+		case 6: datarate = pclk * 30; break;
+		case 5: datarate = pclk * 24; break;
+		case 2:
+		default:
+			datarate = pclk * 18;
+			break;
 		}
 
-		nv_mask(priv, hreg, 0x0000000f, hval);
-		nv_mask(priv, oreg, mask, oval);
+		if (nvkm_output_dp_train(outp, datarate / soff, true))
+			ERR("link not trained before attach\n");
 	}
+
+	exec_clkcmp(priv, head, 0, pclk, &conf);
+
+	if (!outp->info.location && outp->info.type == DCB_OUTPUT_ANALOG) {
+		oreg = 0x614280 + (ffs(outp->info.or) - 1) * 0x800;
+		oval = 0x00000000;
+		hval = 0x00000000;
+		mask = 0xffffffff;
+	} else
+	if (!outp->info.location) {
+		if (outp->info.type == DCB_OUTPUT_DP)
+			nv50_disp_intr_unk20_2_dp(priv, &outp->info, pclk);
+		oreg = 0x614300 + (ffs(outp->info.or) - 1) * 0x800;
+		oval = (conf & 0x0100) ? 0x00000101 : 0x00000000;
+		hval = 0x00000000;
+		mask = 0x00000707;
+	} else {
+		oreg = 0x614380 + (ffs(outp->info.or) - 1) * 0x800;
+		oval = 0x00000001;
+		hval = 0x00000001;
+		mask = 0x00000707;
+	}
+
+	nv_mask(priv, hreg, 0x0000000f, hval);
+	nv_mask(priv, oreg, mask, oval);
 }
 
 /* If programming a TMDS output on a SOR that can also be configured for
@@ -1180,30 +1579,16 @@ nv50_disp_intr_unk40_0_tmds(struct nv50_disp_priv *priv, struct dcb_output *outp
 static void
 nv50_disp_intr_unk40_0(struct nv50_disp_priv *priv, int head)
 {
-	struct dcb_output outp;
+	struct nvkm_output *outp;
 	u32 pclk = nv_rd32(priv, 0x610ad0 + (head * 0x540)) & 0x3fffff;
-	if (exec_clkcmp(priv, head, 1, pclk, &outp) != ~0) {
-		if (outp.location == 0 && outp.type == DCB_OUTPUT_TMDS)
-			nv50_disp_intr_unk40_0_tmds(priv, &outp);
-		else
-		if (outp.location == 1 && outp.type == DCB_OUTPUT_DP) {
-			u32 soff = (ffs(outp.or) - 1) * 0x08;
-			u32 ctrl = nv_rd32(priv, 0x610b84 + soff);
-			u32 datarate;
-
-			switch ((ctrl & 0x000f0000) >> 16) {
-			case 6: datarate = pclk * 30 / 8; break;
-			case 5: datarate = pclk * 24 / 8; break;
-			case 2:
-			default:
-				datarate = pclk * 18 / 8;
-				break;
-			}
+	u32 conf;
 
-			nouveau_dp_train(&priv->base, priv->pior.dp,
-					 &outp, head, datarate);
-		}
-	}
+	outp = exec_clkcmp(priv, head, 1, pclk, &conf);
+	if (!outp)
+		return;
+
+	if (outp->info.location == 0 && outp->info.type == DCB_OUTPUT_TMDS)
+		nv50_disp_intr_unk40_0_tmds(priv, &outp->info);
 }
 
 void
@@ -1211,12 +1596,14 @@ nv50_disp_intr_supervisor(struct work_struct *work)
 {
 	struct nv50_disp_priv *priv =
 		container_of(work, struct nv50_disp_priv, supervisor);
+	struct nv50_disp_impl *impl = (void *)nv_object(priv)->oclass;
 	u32 super = nv_rd32(priv, 0x610030);
 	int head;
 
 	nv_debug(priv, "supervisor 0x%08x 0x%08x\n", priv->super, super);
 
 	if (priv->super & 0x00000010) {
+		nv50_disp_mthd_chan(priv, NV_DBG_DEBUG, 0, impl->mthd.core);
 		for (head = 0; head < priv->head.nr; head++) {
 			if (!(super & (0x00000020 << head)))
 				continue;
@@ -1260,19 +1647,20 @@ nv50_disp_intr(struct nouveau_subdev *subdev)
 	u32 intr0 = nv_rd32(priv, 0x610020);
 	u32 intr1 = nv_rd32(priv, 0x610024);
 
-	if (intr0 & 0x001f0000) {
-		nv50_disp_intr_error(priv);
-		intr0 &= ~0x001f0000;
+	while (intr0 & 0x001f0000) {
+		u32 chid = __ffs(intr0 & 0x001f0000) - 16;
+		nv50_disp_intr_error(priv, chid);
+		intr0 &= ~(0x00010000 << chid);
 	}
 
 	if (intr1 & 0x00000004) {
-		nouveau_event_trigger(priv->base.vblank, 0);
+		nouveau_event_trigger(priv->base.vblank, 1, 0);
 		nv_wr32(priv, 0x610024, 0x00000004);
 		intr1 &= ~0x00000004;
 	}
 
 	if (intr1 & 0x00000008) {
-		nouveau_event_trigger(priv->base.vblank, 1);
+		nouveau_event_trigger(priv->base.vblank, 1, 1);
 		nv_wr32(priv, 0x610024, 0x00000008);
 		intr1 &= ~0x00000008;
 	}
@@ -1312,17 +1700,27 @@ nv50_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->dac.sense = nv50_dac_sense;
 	priv->sor.power = nv50_sor_power;
 	priv->pior.power = nv50_pior_power;
-	priv->pior.dp = &nv50_pior_dp_func;
 	return 0;
 }
 
-struct nouveau_oclass
-nv50_disp_oclass = {
-	.handle = NV_ENGINE(DISP, 0x50),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv50_disp_outp_sclass[] = {
+	&nv50_pior_dp_impl.base.base,
+	NULL
+};
+
+struct nouveau_oclass *
+nv50_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x50),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv50_disp_ctor,
 		.dtor = _nouveau_disp_dtor,
 		.init = _nouveau_disp_init,
 		.fini = _nouveau_disp_fini,
 	},
-};
+	.base.outp =  nv50_disp_outp_sclass,
+	.mthd.core = &nv50_disp_mast_mthd_chan,
+	.mthd.base = &nv50_disp_sync_mthd_chan,
+	.mthd.ovly = &nv50_disp_ovly_mthd_chan,
+	.mthd.prev = 0x000004,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
index 1ae6ceb5670..1a886472b6f 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
@@ -8,9 +8,21 @@
 #include <core/event.h>
 
 #include <engine/dmaobj.h>
-#include <engine/disp.h>
 
 #include "dport.h"
+#include "priv.h"
+#include "outp.h"
+#include "outpdp.h"
+
+struct nv50_disp_impl {
+	struct nouveau_disp_impl base;
+	struct {
+		const struct nv50_disp_mthd_chan *core;
+		const struct nv50_disp_mthd_chan *base;
+		const struct nv50_disp_mthd_chan *ovly;
+		int prev;
+	} mthd;
+};
 
 struct nv50_disp_priv {
 	struct nouveau_disp base;
@@ -33,16 +45,18 @@ struct nv50_disp_priv {
 		int (*hda_eld)(struct nv50_disp_priv *, int sor, u8 *, u32);
 		int (*hdmi)(struct nv50_disp_priv *, int head, int sor, u32);
 		u32 lvdsconf;
-		const struct nouveau_dp_func *dp;
 	} sor;
 	struct {
 		int nr;
 		int (*power)(struct nv50_disp_priv *, int ext, u32 data);
 		u8 type[3];
-		const struct nouveau_dp_func *dp;
 	} pior;
 };
 
+#define HEAD_MTHD(n) (n), (n) + 0x03
+
+int nv50_disp_base_scanoutpos(struct nouveau_object *, u32, void *, u32);
+
 #define DAC_MTHD(n) (n), (n) + 0x03
 
 int nv50_dac_mthd(struct nouveau_object *, u32, void *, u32);
@@ -120,28 +134,79 @@ struct nv50_disp_pioc {
 	struct nv50_disp_chan base;
 };
 
+struct nv50_disp_mthd_list {
+	u32 mthd;
+	u32 addr;
+	struct {
+		u32 mthd;
+		u32 addr;
+		const char *name;
+	} data[];
+};
+
+struct nv50_disp_mthd_chan {
+	const char *name;
+	u32 addr;
+	struct {
+		const char *name;
+		int nr;
+		const struct nv50_disp_mthd_list *mthd;
+	} data[];
+};
+
 extern struct nouveau_ofuncs nv50_disp_mast_ofuncs;
+extern const struct nv50_disp_mthd_list nv50_disp_mast_mthd_base;
+extern const struct nv50_disp_mthd_list nv50_disp_mast_mthd_sor;
+extern const struct nv50_disp_mthd_list nv50_disp_mast_mthd_pior;
 extern struct nouveau_ofuncs nv50_disp_sync_ofuncs;
+extern const struct nv50_disp_mthd_list nv50_disp_sync_mthd_image;
 extern struct nouveau_ofuncs nv50_disp_ovly_ofuncs;
+extern const struct nv50_disp_mthd_list nv50_disp_ovly_mthd_base;
 extern struct nouveau_ofuncs nv50_disp_oimm_ofuncs;
 extern struct nouveau_ofuncs nv50_disp_curs_ofuncs;
 extern struct nouveau_ofuncs nv50_disp_base_ofuncs;
 extern struct nouveau_oclass nv50_disp_cclass;
+void nv50_disp_mthd_chan(struct nv50_disp_priv *, int debug, int head,
+			 const struct nv50_disp_mthd_chan *);
 void nv50_disp_intr_supervisor(struct work_struct *);
 void nv50_disp_intr(struct nouveau_subdev *);
 
+extern const struct nv50_disp_mthd_chan nv84_disp_mast_mthd_chan;
+extern const struct nv50_disp_mthd_list nv84_disp_mast_mthd_dac;
+extern const struct nv50_disp_mthd_list nv84_disp_mast_mthd_head;
+extern const struct nv50_disp_mthd_chan nv84_disp_sync_mthd_chan;
+extern const struct nv50_disp_mthd_chan nv84_disp_ovly_mthd_chan;
 extern struct nouveau_omthds nv84_disp_base_omthds[];
 
-extern struct nouveau_omthds nva3_disp_base_omthds[];
+extern const struct nv50_disp_mthd_chan nv94_disp_mast_mthd_chan;
 
 extern struct nouveau_ofuncs nvd0_disp_mast_ofuncs;
+extern const struct nv50_disp_mthd_list nvd0_disp_mast_mthd_base;
+extern const struct nv50_disp_mthd_list nvd0_disp_mast_mthd_dac;
+extern const struct nv50_disp_mthd_list nvd0_disp_mast_mthd_sor;
+extern const struct nv50_disp_mthd_list nvd0_disp_mast_mthd_pior;
 extern struct nouveau_ofuncs nvd0_disp_sync_ofuncs;
 extern struct nouveau_ofuncs nvd0_disp_ovly_ofuncs;
+extern const struct nv50_disp_mthd_chan nvd0_disp_sync_mthd_chan;
 extern struct nouveau_ofuncs nvd0_disp_oimm_ofuncs;
 extern struct nouveau_ofuncs nvd0_disp_curs_ofuncs;
+extern struct nouveau_omthds nvd0_disp_base_omthds[];
 extern struct nouveau_ofuncs nvd0_disp_base_ofuncs;
 extern struct nouveau_oclass nvd0_disp_cclass;
 void nvd0_disp_intr_supervisor(struct work_struct *);
 void nvd0_disp_intr(struct nouveau_subdev *);
 
+extern const struct nv50_disp_mthd_chan nve0_disp_mast_mthd_chan;
+extern const struct nv50_disp_mthd_chan nve0_disp_ovly_mthd_chan;
+
+extern struct nvkm_output_dp_impl nv50_pior_dp_impl;
+extern struct nouveau_oclass *nv50_disp_outp_sclass[];
+
+extern struct nvkm_output_dp_impl nv94_sor_dp_impl;
+int nv94_sor_dp_lnk_pwr(struct nvkm_output_dp *, int);
+extern struct nouveau_oclass *nv94_disp_outp_sclass[];
+
+extern struct nvkm_output_dp_impl nvd0_sor_dp_impl;
+extern struct nouveau_oclass *nvd0_disp_outp_sclass[];
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
index d8c74c0883a..1cc62e43468 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
@@ -29,6 +29,179 @@
 
 #include "nv50.h"
 
+/*******************************************************************************
+ * EVO master channel object
+ ******************************************************************************/
+
+const struct nv50_disp_mthd_list
+nv84_disp_mast_mthd_dac = {
+	.mthd = 0x0080,
+	.addr = 0x000008,
+	.data = {
+		{ 0x0400, 0x610b58 },
+		{ 0x0404, 0x610bdc },
+		{ 0x0420, 0x610bc4 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_list
+nv84_disp_mast_mthd_head = {
+	.mthd = 0x0400,
+	.addr = 0x000540,
+	.data = {
+		{ 0x0800, 0x610ad8 },
+		{ 0x0804, 0x610ad0 },
+		{ 0x0808, 0x610a48 },
+		{ 0x080c, 0x610a78 },
+		{ 0x0810, 0x610ac0 },
+		{ 0x0814, 0x610af8 },
+		{ 0x0818, 0x610b00 },
+		{ 0x081c, 0x610ae8 },
+		{ 0x0820, 0x610af0 },
+		{ 0x0824, 0x610b08 },
+		{ 0x0828, 0x610b10 },
+		{ 0x082c, 0x610a68 },
+		{ 0x0830, 0x610a60 },
+		{ 0x0834, 0x000000 },
+		{ 0x0838, 0x610a40 },
+		{ 0x0840, 0x610a24 },
+		{ 0x0844, 0x610a2c },
+		{ 0x0848, 0x610aa8 },
+		{ 0x084c, 0x610ab0 },
+		{ 0x085c, 0x610c5c },
+		{ 0x0860, 0x610a84 },
+		{ 0x0864, 0x610a90 },
+		{ 0x0868, 0x610b18 },
+		{ 0x086c, 0x610b20 },
+		{ 0x0870, 0x610ac8 },
+		{ 0x0874, 0x610a38 },
+		{ 0x0878, 0x610c50 },
+		{ 0x0880, 0x610a58 },
+		{ 0x0884, 0x610a9c },
+		{ 0x089c, 0x610c68 },
+		{ 0x08a0, 0x610a70 },
+		{ 0x08a4, 0x610a50 },
+		{ 0x08a8, 0x610ae0 },
+		{ 0x08c0, 0x610b28 },
+		{ 0x08c4, 0x610b30 },
+		{ 0x08c8, 0x610b40 },
+		{ 0x08d4, 0x610b38 },
+		{ 0x08d8, 0x610b48 },
+		{ 0x08dc, 0x610b50 },
+		{ 0x0900, 0x610a18 },
+		{ 0x0904, 0x610ab8 },
+		{ 0x0910, 0x610c70 },
+		{ 0x0914, 0x610c78 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_chan
+nv84_disp_mast_mthd_chan = {
+	.name = "Core",
+	.addr = 0x000000,
+	.data = {
+		{ "Global", 1, &nv50_disp_mast_mthd_base },
+		{    "DAC", 3, &nv84_disp_mast_mthd_dac  },
+		{    "SOR", 2, &nv50_disp_mast_mthd_sor  },
+		{   "PIOR", 3, &nv50_disp_mast_mthd_pior },
+		{   "HEAD", 2, &nv84_disp_mast_mthd_head },
+		{}
+	}
+};
+
+/*******************************************************************************
+ * EVO sync channel objects
+ ******************************************************************************/
+
+static const struct nv50_disp_mthd_list
+nv84_disp_sync_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0080, 0x000000 },
+		{ 0x0084, 0x0008c4 },
+		{ 0x0088, 0x0008d0 },
+		{ 0x008c, 0x0008dc },
+		{ 0x0090, 0x0008e4 },
+		{ 0x0094, 0x610884 },
+		{ 0x00a0, 0x6108a0 },
+		{ 0x00a4, 0x610878 },
+		{ 0x00c0, 0x61086c },
+		{ 0x00c4, 0x610800 },
+		{ 0x00c8, 0x61080c },
+		{ 0x00cc, 0x610818 },
+		{ 0x00e0, 0x610858 },
+		{ 0x00e4, 0x610860 },
+		{ 0x00e8, 0x6108ac },
+		{ 0x00ec, 0x6108b4 },
+		{ 0x00fc, 0x610824 },
+		{ 0x0100, 0x610894 },
+		{ 0x0104, 0x61082c },
+		{ 0x0110, 0x6108bc },
+		{ 0x0114, 0x61088c },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_chan
+nv84_disp_sync_mthd_chan = {
+	.name = "Base",
+	.addr = 0x000540,
+	.data = {
+		{ "Global", 1, &nv84_disp_sync_mthd_base },
+		{  "Image", 2, &nv50_disp_sync_mthd_image },
+		{}
+	}
+};
+
+/*******************************************************************************
+ * EVO overlay channel objects
+ ******************************************************************************/
+
+static const struct nv50_disp_mthd_list
+nv84_disp_ovly_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0080, 0x000000 },
+		{ 0x0084, 0x6109a0 },
+		{ 0x0088, 0x6109c0 },
+		{ 0x008c, 0x6109c8 },
+		{ 0x0090, 0x6109b4 },
+		{ 0x0094, 0x610970 },
+		{ 0x00a0, 0x610998 },
+		{ 0x00a4, 0x610964 },
+		{ 0x00c0, 0x610958 },
+		{ 0x00e0, 0x6109a8 },
+		{ 0x00e4, 0x6109d0 },
+		{ 0x00e8, 0x6109d8 },
+		{ 0x0100, 0x61094c },
+		{ 0x0104, 0x610984 },
+		{ 0x0108, 0x61098c },
+		{ 0x0800, 0x6109f8 },
+		{ 0x0808, 0x610a08 },
+		{ 0x080c, 0x610a10 },
+		{ 0x0810, 0x610a00 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_chan
+nv84_disp_ovly_mthd_chan = {
+	.name = "Overlay",
+	.addr = 0x000540,
+	.data = {
+		{ "Global", 1, &nv84_disp_ovly_mthd_base },
+		{}
+	}
+};
+
+/*******************************************************************************
+ * Base display object
+ ******************************************************************************/
+
 static struct nouveau_oclass
 nv84_disp_sclass[] = {
 	{ NV84_DISP_MAST_CLASS, &nv50_disp_mast_ofuncs },
@@ -41,6 +214,7 @@ nv84_disp_sclass[] = {
 
 struct nouveau_omthds
 nv84_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nv50_disp_base_scanoutpos },
 	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
 	{ SOR_MTHD(NV84_DISP_SOR_HDMI_PWR)    , nv50_sor_mthd },
 	{ SOR_MTHD(NV50_DISP_SOR_LVDS_SCRIPT) , nv50_sor_mthd },
@@ -58,6 +232,10 @@ nv84_disp_base_oclass[] = {
 	{}
 };
 
+/*******************************************************************************
+ * Display engine implementation
+ ******************************************************************************/
+
 static int
 nv84_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	       struct nouveau_oclass *oclass, void *data, u32 size,
@@ -86,17 +264,21 @@ nv84_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hdmi = nv84_hdmi_ctrl;
 	priv->pior.power = nv50_pior_power;
-	priv->pior.dp = &nv50_pior_dp_func;
 	return 0;
 }
 
-struct nouveau_oclass
-nv84_disp_oclass = {
-	.handle = NV_ENGINE(DISP, 0x82),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv84_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x82),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv84_disp_ctor,
 		.dtor = _nouveau_disp_dtor,
 		.init = _nouveau_disp_init,
 		.fini = _nouveau_disp_fini,
 	},
-};
+	.base.outp =  nv50_disp_outp_sclass,
+	.mthd.core = &nv84_disp_mast_mthd_chan,
+	.mthd.base = &nv84_disp_sync_mthd_chan,
+	.mthd.ovly = &nv84_disp_ovly_mthd_chan,
+	.mthd.prev = 0x000004,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
index a66f949c1f8..4f718a9f5ae 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
@@ -29,6 +29,38 @@
 
 #include "nv50.h"
 
+/*******************************************************************************
+ * EVO master channel object
+ ******************************************************************************/
+
+const struct nv50_disp_mthd_list
+nv94_disp_mast_mthd_sor = {
+	.mthd = 0x0040,
+	.addr = 0x000008,
+	.data = {
+		{ 0x0600, 0x610794 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_chan
+nv94_disp_mast_mthd_chan = {
+	.name = "Core",
+	.addr = 0x000000,
+	.data = {
+		{ "Global", 1, &nv50_disp_mast_mthd_base },
+		{    "DAC", 3, &nv84_disp_mast_mthd_dac  },
+		{    "SOR", 4, &nv94_disp_mast_mthd_sor  },
+		{   "PIOR", 3, &nv50_disp_mast_mthd_pior },
+		{   "HEAD", 2, &nv84_disp_mast_mthd_head },
+		{}
+	}
+};
+
+/*******************************************************************************
+ * Base display object
+ ******************************************************************************/
+
 static struct nouveau_oclass
 nv94_disp_sclass[] = {
 	{ NV94_DISP_MAST_CLASS, &nv50_disp_mast_ofuncs },
@@ -41,9 +73,11 @@ nv94_disp_sclass[] = {
 
 static struct nouveau_omthds
 nv94_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nv50_disp_base_scanoutpos },
 	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
 	{ SOR_MTHD(NV84_DISP_SOR_HDMI_PWR)    , nv50_sor_mthd },
 	{ SOR_MTHD(NV50_DISP_SOR_LVDS_SCRIPT) , nv50_sor_mthd },
+	{ SOR_MTHD(NV94_DISP_SOR_DP_PWR)      , nv50_sor_mthd },
 	{ DAC_MTHD(NV50_DISP_DAC_PWR)         , nv50_dac_mthd },
 	{ DAC_MTHD(NV50_DISP_DAC_LOAD)        , nv50_dac_mthd },
 	{ PIOR_MTHD(NV50_DISP_PIOR_PWR)       , nv50_pior_mthd },
@@ -58,6 +92,10 @@ nv94_disp_base_oclass[] = {
 	{}
 };
 
+/*******************************************************************************
+ * Display engine implementation
+ ******************************************************************************/
+
 static int
 nv94_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	       struct nouveau_oclass *oclass, void *data, u32 size,
@@ -85,19 +123,29 @@ nv94_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->dac.sense = nv50_dac_sense;
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hdmi = nv84_hdmi_ctrl;
-	priv->sor.dp = &nv94_sor_dp_func;
 	priv->pior.power = nv50_pior_power;
-	priv->pior.dp = &nv50_pior_dp_func;
 	return 0;
 }
 
-struct nouveau_oclass
-nv94_disp_oclass = {
-	.handle = NV_ENGINE(DISP, 0x88),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv94_disp_outp_sclass[] = {
+	&nv50_pior_dp_impl.base.base,
+	&nv94_sor_dp_impl.base.base,
+	NULL
+};
+
+struct nouveau_oclass *
+nv94_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x88),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv94_disp_ctor,
 		.dtor = _nouveau_disp_dtor,
 		.init = _nouveau_disp_init,
 		.fini = _nouveau_disp_fini,
 	},
-};
+	.base.outp =  nv94_disp_outp_sclass,
+	.mthd.core = &nv94_disp_mast_mthd_chan,
+	.mthd.base = &nv84_disp_sync_mthd_chan,
+	.mthd.ovly = &nv84_disp_ovly_mthd_chan,
+	.mthd.prev = 0x000004,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c
index 6cf8eefac36..6237a9a36f7 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nva0.c
@@ -29,6 +29,55 @@
 
 #include "nv50.h"
 
+/*******************************************************************************
+ * EVO overlay channel objects
+ ******************************************************************************/
+
+static const struct nv50_disp_mthd_list
+nva0_disp_ovly_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0080, 0x000000 },
+		{ 0x0084, 0x6109a0 },
+		{ 0x0088, 0x6109c0 },
+		{ 0x008c, 0x6109c8 },
+		{ 0x0090, 0x6109b4 },
+		{ 0x0094, 0x610970 },
+		{ 0x00a0, 0x610998 },
+		{ 0x00a4, 0x610964 },
+		{ 0x00b0, 0x610c98 },
+		{ 0x00b4, 0x610ca4 },
+		{ 0x00b8, 0x610cac },
+		{ 0x00c0, 0x610958 },
+		{ 0x00e0, 0x6109a8 },
+		{ 0x00e4, 0x6109d0 },
+		{ 0x00e8, 0x6109d8 },
+		{ 0x0100, 0x61094c },
+		{ 0x0104, 0x610984 },
+		{ 0x0108, 0x61098c },
+		{ 0x0800, 0x6109f8 },
+		{ 0x0808, 0x610a08 },
+		{ 0x080c, 0x610a10 },
+		{ 0x0810, 0x610a00 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_chan
+nva0_disp_ovly_mthd_chan = {
+	.name = "Overlay",
+	.addr = 0x000540,
+	.data = {
+		{ "Global", 1, &nva0_disp_ovly_mthd_base },
+		{}
+	}
+};
+
+/*******************************************************************************
+ * Base display object
+ ******************************************************************************/
+
 static struct nouveau_oclass
 nva0_disp_sclass[] = {
 	{ NVA0_DISP_MAST_CLASS, &nv50_disp_mast_ofuncs },
@@ -45,6 +94,10 @@ nva0_disp_base_oclass[] = {
 	{}
 };
 
+/*******************************************************************************
+ * Display engine implementation
+ ******************************************************************************/
+
 static int
 nva0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	       struct nouveau_oclass *oclass, void *data, u32 size,
@@ -73,17 +126,21 @@ nva0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hdmi = nv84_hdmi_ctrl;
 	priv->pior.power = nv50_pior_power;
-	priv->pior.dp = &nv50_pior_dp_func;
 	return 0;
 }
 
-struct nouveau_oclass
-nva0_disp_oclass = {
-	.handle = NV_ENGINE(DISP, 0x83),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nva0_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x83),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nva0_disp_ctor,
 		.dtor = _nouveau_disp_dtor,
 		.init = _nouveau_disp_init,
 		.fini = _nouveau_disp_fini,
 	},
-};
+	.base.outp =  nv50_disp_outp_sclass,
+	.mthd.core = &nv84_disp_mast_mthd_chan,
+	.mthd.base = &nv84_disp_sync_mthd_chan,
+	.mthd.ovly = &nva0_disp_ovly_mthd_chan,
+	.mthd.prev = 0x000004,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c b/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
index b75413169ea..019124d4782 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
@@ -29,6 +29,10 @@
 
 #include "nv50.h"
 
+/*******************************************************************************
+ * Base display object
+ ******************************************************************************/
+
 static struct nouveau_oclass
 nva3_disp_sclass[] = {
 	{ NVA3_DISP_MAST_CLASS, &nv50_disp_mast_ofuncs },
@@ -39,12 +43,14 @@ nva3_disp_sclass[] = {
 	{}
 };
 
-struct nouveau_omthds
+static struct nouveau_omthds
 nva3_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nv50_disp_base_scanoutpos },
 	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
 	{ SOR_MTHD(NVA3_DISP_SOR_HDA_ELD)     , nv50_sor_mthd },
 	{ SOR_MTHD(NV84_DISP_SOR_HDMI_PWR)    , nv50_sor_mthd },
 	{ SOR_MTHD(NV50_DISP_SOR_LVDS_SCRIPT) , nv50_sor_mthd },
+	{ SOR_MTHD(NV94_DISP_SOR_DP_PWR)      , nv50_sor_mthd },
 	{ DAC_MTHD(NV50_DISP_DAC_PWR)         , nv50_dac_mthd },
 	{ DAC_MTHD(NV50_DISP_DAC_LOAD)        , nv50_dac_mthd },
 	{ PIOR_MTHD(NV50_DISP_PIOR_PWR)       , nv50_pior_mthd },
@@ -59,6 +65,10 @@ nva3_disp_base_oclass[] = {
 	{}
 };
 
+/*******************************************************************************
+ * Display engine implementation
+ ******************************************************************************/
+
 static int
 nva3_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	       struct nouveau_oclass *oclass, void *data, u32 size,
@@ -87,19 +97,22 @@ nva3_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nva3_hda_eld;
 	priv->sor.hdmi = nva3_hdmi_ctrl;
-	priv->sor.dp = &nv94_sor_dp_func;
 	priv->pior.power = nv50_pior_power;
-	priv->pior.dp = &nv50_pior_dp_func;
 	return 0;
 }
 
-struct nouveau_oclass
-nva3_disp_oclass = {
-	.handle = NV_ENGINE(DISP, 0x85),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nva3_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x85),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nva3_disp_ctor,
 		.dtor = _nouveau_disp_dtor,
 		.init = _nouveau_disp_init,
 		.fini = _nouveau_disp_fini,
 	},
-};
+	.base.outp =  nv94_disp_outp_sclass,
+	.mthd.core = &nv94_disp_mast_mthd_chan,
+	.mthd.base = &nv84_disp_sync_mthd_chan,
+	.mthd.ovly = &nv84_disp_ovly_mthd_chan,
+	.mthd.prev = 0x000004,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
index 378a015091d..fa30d8196f3 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
@@ -124,6 +124,146 @@ nvd0_disp_dmac_fini(struct nouveau_object *object, bool suspend)
  * EVO master channel object
  ******************************************************************************/
 
+const struct nv50_disp_mthd_list
+nvd0_disp_mast_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0080, 0x660080 },
+		{ 0x0084, 0x660084 },
+		{ 0x0088, 0x660088 },
+		{ 0x008c, 0x000000 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_list
+nvd0_disp_mast_mthd_dac = {
+	.mthd = 0x0020,
+	.addr = 0x000020,
+	.data = {
+		{ 0x0180, 0x660180 },
+		{ 0x0184, 0x660184 },
+		{ 0x0188, 0x660188 },
+		{ 0x0190, 0x660190 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_list
+nvd0_disp_mast_mthd_sor = {
+	.mthd = 0x0020,
+	.addr = 0x000020,
+	.data = {
+		{ 0x0200, 0x660200 },
+		{ 0x0204, 0x660204 },
+		{ 0x0208, 0x660208 },
+		{ 0x0210, 0x660210 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_list
+nvd0_disp_mast_mthd_pior = {
+	.mthd = 0x0020,
+	.addr = 0x000020,
+	.data = {
+		{ 0x0300, 0x660300 },
+		{ 0x0304, 0x660304 },
+		{ 0x0308, 0x660308 },
+		{ 0x0310, 0x660310 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_list
+nvd0_disp_mast_mthd_head = {
+	.mthd = 0x0300,
+	.addr = 0x000300,
+	.data = {
+		{ 0x0400, 0x660400 },
+		{ 0x0404, 0x660404 },
+		{ 0x0408, 0x660408 },
+		{ 0x040c, 0x66040c },
+		{ 0x0410, 0x660410 },
+		{ 0x0414, 0x660414 },
+		{ 0x0418, 0x660418 },
+		{ 0x041c, 0x66041c },
+		{ 0x0420, 0x660420 },
+		{ 0x0424, 0x660424 },
+		{ 0x0428, 0x660428 },
+		{ 0x042c, 0x66042c },
+		{ 0x0430, 0x660430 },
+		{ 0x0434, 0x660434 },
+		{ 0x0438, 0x660438 },
+		{ 0x0440, 0x660440 },
+		{ 0x0444, 0x660444 },
+		{ 0x0448, 0x660448 },
+		{ 0x044c, 0x66044c },
+		{ 0x0450, 0x660450 },
+		{ 0x0454, 0x660454 },
+		{ 0x0458, 0x660458 },
+		{ 0x045c, 0x66045c },
+		{ 0x0460, 0x660460 },
+		{ 0x0468, 0x660468 },
+		{ 0x046c, 0x66046c },
+		{ 0x0470, 0x660470 },
+		{ 0x0474, 0x660474 },
+		{ 0x0480, 0x660480 },
+		{ 0x0484, 0x660484 },
+		{ 0x048c, 0x66048c },
+		{ 0x0490, 0x660490 },
+		{ 0x0494, 0x660494 },
+		{ 0x0498, 0x660498 },
+		{ 0x04b0, 0x6604b0 },
+		{ 0x04b8, 0x6604b8 },
+		{ 0x04bc, 0x6604bc },
+		{ 0x04c0, 0x6604c0 },
+		{ 0x04c4, 0x6604c4 },
+		{ 0x04c8, 0x6604c8 },
+		{ 0x04d0, 0x6604d0 },
+		{ 0x04d4, 0x6604d4 },
+		{ 0x04e0, 0x6604e0 },
+		{ 0x04e4, 0x6604e4 },
+		{ 0x04e8, 0x6604e8 },
+		{ 0x04ec, 0x6604ec },
+		{ 0x04f0, 0x6604f0 },
+		{ 0x04f4, 0x6604f4 },
+		{ 0x04f8, 0x6604f8 },
+		{ 0x04fc, 0x6604fc },
+		{ 0x0500, 0x660500 },
+		{ 0x0504, 0x660504 },
+		{ 0x0508, 0x660508 },
+		{ 0x050c, 0x66050c },
+		{ 0x0510, 0x660510 },
+		{ 0x0514, 0x660514 },
+		{ 0x0518, 0x660518 },
+		{ 0x051c, 0x66051c },
+		{ 0x052c, 0x66052c },
+		{ 0x0530, 0x660530 },
+		{ 0x054c, 0x66054c },
+		{ 0x0550, 0x660550 },
+		{ 0x0554, 0x660554 },
+		{ 0x0558, 0x660558 },
+		{ 0x055c, 0x66055c },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_chan
+nvd0_disp_mast_mthd_chan = {
+	.name = "Core",
+	.addr = 0x000000,
+	.data = {
+		{ "Global", 1, &nvd0_disp_mast_mthd_base },
+		{    "DAC", 3, &nvd0_disp_mast_mthd_dac  },
+		{    "SOR", 8, &nvd0_disp_mast_mthd_sor  },
+		{   "PIOR", 4, &nvd0_disp_mast_mthd_pior },
+		{   "HEAD", 4, &nvd0_disp_mast_mthd_head },
+		{}
+	}
+};
+
 static int
 nvd0_disp_mast_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
@@ -216,6 +356,81 @@ nvd0_disp_mast_ofuncs = {
  * EVO sync channel objects
  ******************************************************************************/
 
+static const struct nv50_disp_mthd_list
+nvd0_disp_sync_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0080, 0x661080 },
+		{ 0x0084, 0x661084 },
+		{ 0x0088, 0x661088 },
+		{ 0x008c, 0x66108c },
+		{ 0x0090, 0x661090 },
+		{ 0x0094, 0x661094 },
+		{ 0x00a0, 0x6610a0 },
+		{ 0x00a4, 0x6610a4 },
+		{ 0x00c0, 0x6610c0 },
+		{ 0x00c4, 0x6610c4 },
+		{ 0x00c8, 0x6610c8 },
+		{ 0x00cc, 0x6610cc },
+		{ 0x00e0, 0x6610e0 },
+		{ 0x00e4, 0x6610e4 },
+		{ 0x00e8, 0x6610e8 },
+		{ 0x00ec, 0x6610ec },
+		{ 0x00fc, 0x6610fc },
+		{ 0x0100, 0x661100 },
+		{ 0x0104, 0x661104 },
+		{ 0x0108, 0x661108 },
+		{ 0x010c, 0x66110c },
+		{ 0x0110, 0x661110 },
+		{ 0x0114, 0x661114 },
+		{ 0x0118, 0x661118 },
+		{ 0x011c, 0x66111c },
+		{ 0x0130, 0x661130 },
+		{ 0x0134, 0x661134 },
+		{ 0x0138, 0x661138 },
+		{ 0x013c, 0x66113c },
+		{ 0x0140, 0x661140 },
+		{ 0x0144, 0x661144 },
+		{ 0x0148, 0x661148 },
+		{ 0x014c, 0x66114c },
+		{ 0x0150, 0x661150 },
+		{ 0x0154, 0x661154 },
+		{ 0x0158, 0x661158 },
+		{ 0x015c, 0x66115c },
+		{ 0x0160, 0x661160 },
+		{ 0x0164, 0x661164 },
+		{ 0x0168, 0x661168 },
+		{ 0x016c, 0x66116c },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_list
+nvd0_disp_sync_mthd_image = {
+	.mthd = 0x0400,
+	.addr = 0x000400,
+	.data = {
+		{ 0x0400, 0x661400 },
+		{ 0x0404, 0x661404 },
+		{ 0x0408, 0x661408 },
+		{ 0x040c, 0x66140c },
+		{ 0x0410, 0x661410 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_chan
+nvd0_disp_sync_mthd_chan = {
+	.name = "Base",
+	.addr = 0x001000,
+	.data = {
+		{ "Global", 1, &nvd0_disp_sync_mthd_base },
+		{  "Image", 2, &nvd0_disp_sync_mthd_image },
+		{}
+	}
+};
+
 static int
 nvd0_disp_sync_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
@@ -256,6 +471,68 @@ nvd0_disp_sync_ofuncs = {
  * EVO overlay channel objects
  ******************************************************************************/
 
+static const struct nv50_disp_mthd_list
+nvd0_disp_ovly_mthd_base = {
+	.mthd = 0x0000,
+	.data = {
+		{ 0x0080, 0x665080 },
+		{ 0x0084, 0x665084 },
+		{ 0x0088, 0x665088 },
+		{ 0x008c, 0x66508c },
+		{ 0x0090, 0x665090 },
+		{ 0x0094, 0x665094 },
+		{ 0x00a0, 0x6650a0 },
+		{ 0x00a4, 0x6650a4 },
+		{ 0x00b0, 0x6650b0 },
+		{ 0x00b4, 0x6650b4 },
+		{ 0x00b8, 0x6650b8 },
+		{ 0x00c0, 0x6650c0 },
+		{ 0x00e0, 0x6650e0 },
+		{ 0x00e4, 0x6650e4 },
+		{ 0x00e8, 0x6650e8 },
+		{ 0x0100, 0x665100 },
+		{ 0x0104, 0x665104 },
+		{ 0x0108, 0x665108 },
+		{ 0x010c, 0x66510c },
+		{ 0x0110, 0x665110 },
+		{ 0x0118, 0x665118 },
+		{ 0x011c, 0x66511c },
+		{ 0x0120, 0x665120 },
+		{ 0x0124, 0x665124 },
+		{ 0x0130, 0x665130 },
+		{ 0x0134, 0x665134 },
+		{ 0x0138, 0x665138 },
+		{ 0x013c, 0x66513c },
+		{ 0x0140, 0x665140 },
+		{ 0x0144, 0x665144 },
+		{ 0x0148, 0x665148 },
+		{ 0x014c, 0x66514c },
+		{ 0x0150, 0x665150 },
+		{ 0x0154, 0x665154 },
+		{ 0x0158, 0x665158 },
+		{ 0x015c, 0x66515c },
+		{ 0x0160, 0x665160 },
+		{ 0x0164, 0x665164 },
+		{ 0x0168, 0x665168 },
+		{ 0x016c, 0x66516c },
+		{ 0x0400, 0x665400 },
+		{ 0x0408, 0x665408 },
+		{ 0x040c, 0x66540c },
+		{ 0x0410, 0x665410 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_chan
+nvd0_disp_ovly_mthd_chan = {
+	.name = "Overlay",
+	.addr = 0x001000,
+	.data = {
+		{ "Global", 1, &nvd0_disp_ovly_mthd_base },
+		{}
+	}
+};
+
 static int
 nvd0_disp_ovly_ctor(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
@@ -440,14 +717,44 @@ nvd0_disp_curs_ofuncs = {
  * Base display object
  ******************************************************************************/
 
+static int
+nvd0_disp_base_scanoutpos(struct nouveau_object *object, u32 mthd,
+			  void *data, u32 size)
+{
+	struct nv50_disp_priv *priv = (void *)object->engine;
+	struct nv04_display_scanoutpos *args = data;
+	const int head = (mthd & NV50_DISP_MTHD_HEAD);
+	u32 blanke, blanks, total;
+
+	if (size < sizeof(*args) || head >= priv->head.nr)
+		return -EINVAL;
+
+	total  = nv_rd32(priv, 0x640414 + (head * 0x300));
+	blanke = nv_rd32(priv, 0x64041c + (head * 0x300));
+	blanks = nv_rd32(priv, 0x640420 + (head * 0x300));
+
+	args->vblanke = (blanke & 0xffff0000) >> 16;
+	args->hblanke = (blanke & 0x0000ffff);
+	args->vblanks = (blanks & 0xffff0000) >> 16;
+	args->hblanks = (blanks & 0x0000ffff);
+	args->vtotal  = ( total & 0xffff0000) >> 16;
+	args->htotal  = ( total & 0x0000ffff);
+
+	args->time[0] = ktime_to_ns(ktime_get());
+	args->vline   = nv_rd32(priv, 0x616340 + (head * 0x800)) & 0xffff;
+	args->time[1] = ktime_to_ns(ktime_get()); /* vline read locks hline */
+	args->hline   = nv_rd32(priv, 0x616344 + (head * 0x800)) & 0xffff;
+	return 0;
+}
+
 static void
-nvd0_disp_base_vblank_enable(struct nouveau_event *event, int head)
+nvd0_disp_base_vblank_enable(struct nouveau_event *event, int type, int head)
 {
 	nv_mask(event->priv, 0x6100c0 + (head * 0x800), 0x00000001, 0x00000001);
 }
 
 static void
-nvd0_disp_base_vblank_disable(struct nouveau_event *event, int head)
+nvd0_disp_base_vblank_disable(struct nouveau_event *event, int type, int head)
 {
 	nv_mask(event->priv, 0x6100c0 + (head * 0x800), 0x00000001, 0x00000000);
 }
@@ -573,9 +880,25 @@ nvd0_disp_base_ofuncs = {
 	.fini = nvd0_disp_base_fini,
 };
 
+struct nouveau_omthds
+nvd0_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nvd0_disp_base_scanoutpos },
+	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
+	{ SOR_MTHD(NVA3_DISP_SOR_HDA_ELD)     , nv50_sor_mthd },
+	{ SOR_MTHD(NV84_DISP_SOR_HDMI_PWR)    , nv50_sor_mthd },
+	{ SOR_MTHD(NV50_DISP_SOR_LVDS_SCRIPT) , nv50_sor_mthd },
+	{ SOR_MTHD(NV94_DISP_SOR_DP_PWR)      , nv50_sor_mthd },
+	{ DAC_MTHD(NV50_DISP_DAC_PWR)         , nv50_dac_mthd },
+	{ DAC_MTHD(NV50_DISP_DAC_LOAD)        , nv50_dac_mthd },
+	{ PIOR_MTHD(NV50_DISP_PIOR_PWR)       , nv50_pior_mthd },
+	{ PIOR_MTHD(NV50_DISP_PIOR_TMDS_PWR)  , nv50_pior_mthd },
+	{ PIOR_MTHD(NV50_DISP_PIOR_DP_PWR)    , nv50_pior_mthd },
+	{},
+};
+
 static struct nouveau_oclass
 nvd0_disp_base_oclass[] = {
-	{ NVD0_DISP_CLASS, &nvd0_disp_base_ofuncs, nva3_disp_base_omthds },
+	{ NVD0_DISP_CLASS, &nvd0_disp_base_ofuncs, nvd0_disp_base_omthds },
 	{}
 };
 
@@ -593,19 +916,20 @@ nvd0_disp_sclass[] = {
  * Display engine implementation
  ******************************************************************************/
 
-static u16
-exec_lookup(struct nv50_disp_priv *priv, int head, int outp, u32 ctrl,
-	    struct dcb_output *dcb, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+static struct nvkm_output *
+exec_lookup(struct nv50_disp_priv *priv, int head, int or, u32 ctrl,
+	    u32 *data, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
 	    struct nvbios_outp *info)
 {
 	struct nouveau_bios *bios = nouveau_bios(priv);
-	u16 mask, type, data;
+	struct nvkm_output *outp;
+	u16 mask, type;
 
-	if (outp < 4) {
+	if (or < 4) {
 		type = DCB_OUTPUT_ANALOG;
 		mask = 0;
 	} else {
-		outp -= 4;
+		or -= 4;
 		switch (ctrl & 0x00000f00) {
 		case 0x00000000: type = DCB_OUTPUT_LVDS; mask = 1; break;
 		case 0x00000100: type = DCB_OUTPUT_TMDS; mask = 1; break;
@@ -617,101 +941,106 @@ exec_lookup(struct nv50_disp_priv *priv, int head, int outp, u32 ctrl,
 			nv_error(priv, "unknown SOR mc 0x%08x\n", ctrl);
 			return 0x0000;
 		}
-		dcb->sorconf.link = mask;
 	}
 
 	mask  = 0x00c0 & (mask << 6);
-	mask |= 0x0001 << outp;
+	mask |= 0x0001 << or;
 	mask |= 0x0100 << head;
 
-	data = dcb_outp_match(bios, type, mask, ver, hdr, dcb);
-	if (!data)
-		return 0x0000;
+	list_for_each_entry(outp, &priv->base.outp, head) {
+		if ((outp->info.hasht & 0xff) == type &&
+		    (outp->info.hashm & mask) == mask) {
+			*data = nvbios_outp_match(bios, outp->info.hasht,
+							outp->info.hashm,
+						  ver, hdr, cnt, len, info);
+			if (!*data)
+				return NULL;
+			return outp;
+		}
+	}
 
-	return nvbios_outp_match(bios, type, mask, ver, hdr, cnt, len, info);
+	return NULL;
 }
 
-static bool
+static struct nvkm_output *
 exec_script(struct nv50_disp_priv *priv, int head, int id)
 {
 	struct nouveau_bios *bios = nouveau_bios(priv);
+	struct nvkm_output *outp;
 	struct nvbios_outp info;
-	struct dcb_output dcb;
 	u8  ver, hdr, cnt, len;
-	u32 ctrl = 0x00000000;
-	u16 data;
-	int outp;
+	u32 data, ctrl = 0;
+	int or;
 
-	for (outp = 0; !(ctrl & (1 << head)) && outp < 8; outp++) {
-		ctrl = nv_rd32(priv, 0x640180 + (outp * 0x20));
+	for (or = 0; !(ctrl & (1 << head)) && or < 8; or++) {
+		ctrl = nv_rd32(priv, 0x640180 + (or * 0x20));
 		if (ctrl & (1 << head))
 			break;
 	}
 
-	if (outp == 8)
-		return false;
+	if (or == 8)
+		return NULL;
 
-	data = exec_lookup(priv, head, outp, ctrl, &dcb, &ver, &hdr, &cnt, &len, &info);
-	if (data) {
+	outp = exec_lookup(priv, head, or, ctrl, &data, &ver, &hdr, &cnt, &len, &info);
+	if (outp) {
 		struct nvbios_init init = {
 			.subdev = nv_subdev(priv),
 			.bios = bios,
 			.offset = info.script[id],
-			.outp = &dcb,
+			.outp = &outp->info,
 			.crtc = head,
 			.execute = 1,
 		};
 
-		return nvbios_exec(&init) == 0;
+		nvbios_exec(&init);
 	}
 
-	return false;
+	return outp;
 }
 
-static u32
-exec_clkcmp(struct nv50_disp_priv *priv, int head, int id,
-	    u32 pclk, struct dcb_output *dcb)
+static struct nvkm_output *
+exec_clkcmp(struct nv50_disp_priv *priv, int head, int id, u32 pclk, u32 *conf)
 {
 	struct nouveau_bios *bios = nouveau_bios(priv);
+	struct nvkm_output *outp;
 	struct nvbios_outp info1;
 	struct nvbios_ocfg info2;
 	u8  ver, hdr, cnt, len;
-	u32 ctrl = 0x00000000;
-	u32 data, conf = ~0;
-	int outp;
+	u32 data, ctrl = 0;
+	int or;
 
-	for (outp = 0; !(ctrl & (1 << head)) && outp < 8; outp++) {
-		ctrl = nv_rd32(priv, 0x660180 + (outp * 0x20));
+	for (or = 0; !(ctrl & (1 << head)) && or < 8; or++) {
+		ctrl = nv_rd32(priv, 0x660180 + (or * 0x20));
 		if (ctrl & (1 << head))
 			break;
 	}
 
-	if (outp == 8)
-		return false;
+	if (or == 8)
+		return NULL;
 
-	data = exec_lookup(priv, head, outp, ctrl, dcb, &ver, &hdr, &cnt, &len, &info1);
-	if (data == 0x0000)
-		return conf;
+	outp = exec_lookup(priv, head, or, ctrl, &data, &ver, &hdr, &cnt, &len, &info1);
+	if (!outp)
+		return NULL;
 
-	switch (dcb->type) {
+	switch (outp->info.type) {
 	case DCB_OUTPUT_TMDS:
-		conf = (ctrl & 0x00000f00) >> 8;
+		*conf = (ctrl & 0x00000f00) >> 8;
 		if (pclk >= 165000)
-			conf |= 0x0100;
+			*conf |= 0x0100;
 		break;
 	case DCB_OUTPUT_LVDS:
-		conf = priv->sor.lvdsconf;
+		*conf = priv->sor.lvdsconf;
 		break;
 	case DCB_OUTPUT_DP:
-		conf = (ctrl & 0x00000f00) >> 8;
+		*conf = (ctrl & 0x00000f00) >> 8;
 		break;
 	case DCB_OUTPUT_ANALOG:
 	default:
-		conf = 0x00ff;
+		*conf = 0x00ff;
 		break;
 	}
 
-	data = nvbios_ocfg_match(bios, data, conf, &ver, &hdr, &cnt, &len, &info2);
+	data = nvbios_ocfg_match(bios, data, *conf, &ver, &hdr, &cnt, &len, &info2);
 	if (data && id < 0xff) {
 		data = nvbios_oclk_match(bios, info2.clkcmp[id], pclk);
 		if (data) {
@@ -719,7 +1048,7 @@ exec_clkcmp(struct nv50_disp_priv *priv, int head, int id,
 				.subdev = nv_subdev(priv),
 				.bios = bios,
 				.offset = data,
-				.outp = dcb,
+				.outp = &outp->info,
 				.crtc = head,
 				.execute = 1,
 			};
@@ -728,7 +1057,7 @@ exec_clkcmp(struct nv50_disp_priv *priv, int head, int id,
 		}
 	}
 
-	return conf;
+	return outp;
 }
 
 static void
@@ -740,7 +1069,23 @@ nvd0_disp_intr_unk1_0(struct nv50_disp_priv *priv, int head)
 static void
 nvd0_disp_intr_unk2_0(struct nv50_disp_priv *priv, int head)
 {
-	exec_script(priv, head, 2);
+	struct nvkm_output *outp = exec_script(priv, head, 2);
+
+	/* see note in nv50_disp_intr_unk20_0() */
+	if (outp && outp->info.type == DCB_OUTPUT_DP) {
+		struct nvkm_output_dp *outpdp = (void *)outp;
+		struct nvbios_init init = {
+			.subdev = nv_subdev(priv),
+			.bios = nouveau_bios(priv),
+			.outp = &outp->info,
+			.crtc = head,
+			.offset = outpdp->info.script[4],
+			.execute = 1,
+		};
+
+		nvbios_exec(&init);
+		atomic_set(&outpdp->lt.done, 0);
+	}
 }
 
 static void
@@ -802,49 +1147,52 @@ nvd0_disp_intr_unk2_2_tu(struct nv50_disp_priv *priv, int head,
 static void
 nvd0_disp_intr_unk2_2(struct nv50_disp_priv *priv, int head)
 {
-	struct dcb_output outp;
+	struct nvkm_output *outp;
 	u32 pclk = nv_rd32(priv, 0x660450 + (head * 0x300)) / 1000;
-	u32 conf = exec_clkcmp(priv, head, 0xff, pclk, &outp);
-	if (conf != ~0) {
-		u32 addr, data;
-
-		if (outp.type == DCB_OUTPUT_DP) {
-			u32 sync = nv_rd32(priv, 0x660404 + (head * 0x300));
-			switch ((sync & 0x000003c0) >> 6) {
-			case 6: pclk = pclk * 30 / 8; break;
-			case 5: pclk = pclk * 24 / 8; break;
-			case 2:
-			default:
-				pclk = pclk * 18 / 8;
-				break;
-			}
-
-			nouveau_dp_train(&priv->base, priv->sor.dp,
-					 &outp, head, pclk);
+	u32 conf, addr, data;
+
+	outp = exec_clkcmp(priv, head, 0xff, pclk, &conf);
+	if (!outp)
+		return;
+
+	/* see note in nv50_disp_intr_unk20_2() */
+	if (outp->info.type == DCB_OUTPUT_DP) {
+		u32 sync = nv_rd32(priv, 0x660404 + (head * 0x300));
+		switch ((sync & 0x000003c0) >> 6) {
+		case 6: pclk = pclk * 30; break;
+		case 5: pclk = pclk * 24; break;
+		case 2:
+		default:
+			pclk = pclk * 18;
+			break;
 		}
 
-		exec_clkcmp(priv, head, 0, pclk, &outp);
+		if (nvkm_output_dp_train(outp, pclk, true))
+			ERR("link not trained before attach\n");
+	}
 
-		if (outp.type == DCB_OUTPUT_ANALOG) {
-			addr = 0x612280 + (ffs(outp.or) - 1) * 0x800;
-			data = 0x00000000;
-		} else {
-			if (outp.type == DCB_OUTPUT_DP)
-				nvd0_disp_intr_unk2_2_tu(priv, head, &outp);
-			addr = 0x612300 + (ffs(outp.or) - 1) * 0x800;
-			data = (conf & 0x0100) ? 0x00000101 : 0x00000000;
-		}
+	exec_clkcmp(priv, head, 0, pclk, &conf);
 
-		nv_mask(priv, addr, 0x00000707, data);
+	if (outp->info.type == DCB_OUTPUT_ANALOG) {
+		addr = 0x612280 + (ffs(outp->info.or) - 1) * 0x800;
+		data = 0x00000000;
+	} else {
+		if (outp->info.type == DCB_OUTPUT_DP)
+			nvd0_disp_intr_unk2_2_tu(priv, head, &outp->info);
+		addr = 0x612300 + (ffs(outp->info.or) - 1) * 0x800;
+		data = (conf & 0x0100) ? 0x00000101 : 0x00000000;
 	}
+
+	nv_mask(priv, addr, 0x00000707, data);
 }
 
 static void
 nvd0_disp_intr_unk4_0(struct nv50_disp_priv *priv, int head)
 {
-	struct dcb_output outp;
 	u32 pclk = nv_rd32(priv, 0x660450 + (head * 0x300)) / 1000;
-	exec_clkcmp(priv, head, 1, pclk, &outp);
+	u32 conf;
+
+	exec_clkcmp(priv, head, 1, pclk, &conf);
 }
 
 void
@@ -852,19 +1200,22 @@ nvd0_disp_intr_supervisor(struct work_struct *work)
 {
 	struct nv50_disp_priv *priv =
 		container_of(work, struct nv50_disp_priv, supervisor);
+	struct nv50_disp_impl *impl = (void *)nv_object(priv)->oclass;
 	u32 mask[4];
 	int head;
 
-	nv_debug(priv, "supervisor %08x\n", priv->super);
+	nv_debug(priv, "supervisor %d\n", ffs(priv->super));
 	for (head = 0; head < priv->head.nr; head++) {
 		mask[head] = nv_rd32(priv, 0x6101d4 + (head * 0x800));
 		nv_debug(priv, "head %d: 0x%08x\n", head, mask[head]);
 	}
 
 	if (priv->super & 0x00000001) {
+		nv50_disp_mthd_chan(priv, NV_DBG_DEBUG, 0, impl->mthd.core);
 		for (head = 0; head < priv->head.nr; head++) {
 			if (!(mask[head] & 0x00001000))
 				continue;
+			nv_debug(priv, "supervisor 1.0 - head %d\n", head);
 			nvd0_disp_intr_unk1_0(priv, head);
 		}
 	} else
@@ -872,16 +1223,19 @@ nvd0_disp_intr_supervisor(struct work_struct *work)
 		for (head = 0; head < priv->head.nr; head++) {
 			if (!(mask[head] & 0x00001000))
 				continue;
+			nv_debug(priv, "supervisor 2.0 - head %d\n", head);
 			nvd0_disp_intr_unk2_0(priv, head);
 		}
 		for (head = 0; head < priv->head.nr; head++) {
 			if (!(mask[head] & 0x00010000))
 				continue;
+			nv_debug(priv, "supervisor 2.1 - head %d\n", head);
 			nvd0_disp_intr_unk2_1(priv, head);
 		}
 		for (head = 0; head < priv->head.nr; head++) {
 			if (!(mask[head] & 0x00001000))
 				continue;
+			nv_debug(priv, "supervisor 2.2 - head %d\n", head);
 			nvd0_disp_intr_unk2_2(priv, head);
 		}
 	} else
@@ -889,6 +1243,7 @@ nvd0_disp_intr_supervisor(struct work_struct *work)
 		for (head = 0; head < priv->head.nr; head++) {
 			if (!(mask[head] & 0x00001000))
 				continue;
+			nv_debug(priv, "supervisor 3.0 - head %d\n", head);
 			nvd0_disp_intr_unk4_0(priv, head);
 		}
 	}
@@ -898,6 +1253,53 @@ nvd0_disp_intr_supervisor(struct work_struct *work)
 	nv_wr32(priv, 0x6101d0, 0x80000000);
 }
 
+static void
+nvd0_disp_intr_error(struct nv50_disp_priv *priv, int chid)
+{
+	const struct nv50_disp_impl *impl = (void *)nv_object(priv)->oclass;
+	u32 mthd = nv_rd32(priv, 0x6101f0 + (chid * 12));
+	u32 data = nv_rd32(priv, 0x6101f4 + (chid * 12));
+	u32 unkn = nv_rd32(priv, 0x6101f8 + (chid * 12));
+
+	nv_error(priv, "chid %d mthd 0x%04x data 0x%08x "
+		       "0x%08x 0x%08x\n",
+		 chid, (mthd & 0x0000ffc), data, mthd, unkn);
+
+	if (chid == 0) {
+		switch (mthd & 0xffc) {
+		case 0x0080:
+			nv50_disp_mthd_chan(priv, NV_DBG_ERROR, chid - 0,
+					    impl->mthd.core);
+			break;
+		default:
+			break;
+		}
+	} else
+	if (chid <= 4) {
+		switch (mthd & 0xffc) {
+		case 0x0080:
+			nv50_disp_mthd_chan(priv, NV_DBG_ERROR, chid - 1,
+					    impl->mthd.base);
+			break;
+		default:
+			break;
+		}
+	} else
+	if (chid <= 8) {
+		switch (mthd & 0xffc) {
+		case 0x0080:
+			nv50_disp_mthd_chan(priv, NV_DBG_ERROR, chid - 5,
+					    impl->mthd.ovly);
+			break;
+		default:
+			break;
+		}
+	}
+
+	nv_wr32(priv, 0x61009c, (1 << chid));
+	nv_wr32(priv, 0x6101f0 + (chid * 12), 0x90000000);
+}
+
 void
 nvd0_disp_intr(struct nouveau_subdev *subdev)
 {
@@ -914,18 +1316,8 @@ nvd0_disp_intr(struct nouveau_subdev *subdev)
 	if (intr & 0x00000002) {
 		u32 stat = nv_rd32(priv, 0x61009c);
 		int chid = ffs(stat) - 1;
-		if (chid >= 0) {
-			u32 mthd = nv_rd32(priv, 0x6101f0 + (chid * 12));
-			u32 data = nv_rd32(priv, 0x6101f4 + (chid * 12));
-			u32 unkn = nv_rd32(priv, 0x6101f8 + (chid * 12));
-
-			nv_error(priv, "chid %d mthd 0x%04x data 0x%08x "
-				       "0x%08x 0x%08x\n",
-				 chid, (mthd & 0x0000ffc), data, mthd, unkn);
-			nv_wr32(priv, 0x61009c, (1 << chid));
-			nv_wr32(priv, 0x6101f0 + (chid * 12), 0x90000000);
-		}
-
+		if (chid >= 0)
+			nvd0_disp_intr_error(priv, chid);
 		intr &= ~0x00000002;
 	}
 
@@ -951,7 +1343,7 @@ nvd0_disp_intr(struct nouveau_subdev *subdev)
 		if (mask & intr) {
 			u32 stat = nv_rd32(priv, 0x6100bc + (i * 0x800));
 			if (stat & 0x00000001)
-				nouveau_event_trigger(priv->base.vblank, i);
+				nouveau_event_trigger(priv->base.vblank, 1, i);
 			nv_mask(priv, 0x6100bc + (i * 0x800), 0, 0);
 			nv_rd32(priv, 0x6100c0 + (i * 0x800));
 		}
@@ -967,9 +1359,6 @@ nvd0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	int heads = nv_rd32(parent, 0x022448);
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000001)
-		return -ENODEV;
-
 	ret = nouveau_disp_create(parent, engine, oclass, heads,
 				  "PDISP", "display", &priv);
 	*pobject = nv_object(priv);
@@ -989,17 +1378,27 @@ nvd0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nvd0_hda_eld;
 	priv->sor.hdmi = nvd0_hdmi_ctrl;
-	priv->sor.dp = &nvd0_sor_dp_func;
 	return 0;
 }
 
-struct nouveau_oclass
-nvd0_disp_oclass = {
-	.handle = NV_ENGINE(DISP, 0x90),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nvd0_disp_outp_sclass[] = {
+	&nvd0_sor_dp_impl.base.base,
+	NULL
+};
+
+struct nouveau_oclass *
+nvd0_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x90),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nvd0_disp_ctor,
 		.dtor = _nouveau_disp_dtor,
 		.init = _nouveau_disp_init,
 		.fini = _nouveau_disp_fini,
 	},
-};
+	.base.outp =  nvd0_disp_outp_sclass,
+	.mthd.core = &nvd0_disp_mast_mthd_chan,
+	.mthd.base = &nvd0_disp_sync_mthd_chan,
+	.mthd.ovly = &nvd0_disp_ovly_mthd_chan,
+	.mthd.prev = -0x020000,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
index fb1fe6ae5e7..11328e3f5df 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
@@ -29,6 +29,175 @@
 
 #include "nv50.h"
 
+/*******************************************************************************
+ * EVO master channel object
+ ******************************************************************************/
+
+static const struct nv50_disp_mthd_list
+nve0_disp_mast_mthd_head = {
+	.mthd = 0x0300,
+	.addr = 0x000300,
+	.data = {
+		{ 0x0400, 0x660400 },
+		{ 0x0404, 0x660404 },
+		{ 0x0408, 0x660408 },
+		{ 0x040c, 0x66040c },
+		{ 0x0410, 0x660410 },
+		{ 0x0414, 0x660414 },
+		{ 0x0418, 0x660418 },
+		{ 0x041c, 0x66041c },
+		{ 0x0420, 0x660420 },
+		{ 0x0424, 0x660424 },
+		{ 0x0428, 0x660428 },
+		{ 0x042c, 0x66042c },
+		{ 0x0430, 0x660430 },
+		{ 0x0434, 0x660434 },
+		{ 0x0438, 0x660438 },
+		{ 0x0440, 0x660440 },
+		{ 0x0444, 0x660444 },
+		{ 0x0448, 0x660448 },
+		{ 0x044c, 0x66044c },
+		{ 0x0450, 0x660450 },
+		{ 0x0454, 0x660454 },
+		{ 0x0458, 0x660458 },
+		{ 0x045c, 0x66045c },
+		{ 0x0460, 0x660460 },
+		{ 0x0468, 0x660468 },
+		{ 0x046c, 0x66046c },
+		{ 0x0470, 0x660470 },
+		{ 0x0474, 0x660474 },
+		{ 0x047c, 0x66047c },
+		{ 0x0480, 0x660480 },
+		{ 0x0484, 0x660484 },
+		{ 0x0488, 0x660488 },
+		{ 0x048c, 0x66048c },
+		{ 0x0490, 0x660490 },
+		{ 0x0494, 0x660494 },
+		{ 0x0498, 0x660498 },
+		{ 0x04a0, 0x6604a0 },
+		{ 0x04b0, 0x6604b0 },
+		{ 0x04b8, 0x6604b8 },
+		{ 0x04bc, 0x6604bc },
+		{ 0x04c0, 0x6604c0 },
+		{ 0x04c4, 0x6604c4 },
+		{ 0x04c8, 0x6604c8 },
+		{ 0x04d0, 0x6604d0 },
+		{ 0x04d4, 0x6604d4 },
+		{ 0x04e0, 0x6604e0 },
+		{ 0x04e4, 0x6604e4 },
+		{ 0x04e8, 0x6604e8 },
+		{ 0x04ec, 0x6604ec },
+		{ 0x04f0, 0x6604f0 },
+		{ 0x04f4, 0x6604f4 },
+		{ 0x04f8, 0x6604f8 },
+		{ 0x04fc, 0x6604fc },
+		{ 0x0500, 0x660500 },
+		{ 0x0504, 0x660504 },
+		{ 0x0508, 0x660508 },
+		{ 0x050c, 0x66050c },
+		{ 0x0510, 0x660510 },
+		{ 0x0514, 0x660514 },
+		{ 0x0518, 0x660518 },
+		{ 0x051c, 0x66051c },
+		{ 0x0520, 0x660520 },
+		{ 0x0524, 0x660524 },
+		{ 0x052c, 0x66052c },
+		{ 0x0530, 0x660530 },
+		{ 0x054c, 0x66054c },
+		{ 0x0550, 0x660550 },
+		{ 0x0554, 0x660554 },
+		{ 0x0558, 0x660558 },
+		{ 0x055c, 0x66055c },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_chan
+nve0_disp_mast_mthd_chan = {
+	.name = "Core",
+	.addr = 0x000000,
+	.data = {
+		{ "Global", 1, &nvd0_disp_mast_mthd_base },
+		{    "DAC", 3, &nvd0_disp_mast_mthd_dac  },
+		{    "SOR", 8, &nvd0_disp_mast_mthd_sor  },
+		{   "PIOR", 4, &nvd0_disp_mast_mthd_pior },
+		{   "HEAD", 4, &nve0_disp_mast_mthd_head },
+		{}
+	}
+};
+
+/*******************************************************************************
+ * EVO overlay channel objects
+ ******************************************************************************/
+
+static const struct nv50_disp_mthd_list
+nve0_disp_ovly_mthd_base = {
+	.mthd = 0x0000,
+	.data = {
+		{ 0x0080, 0x665080 },
+		{ 0x0084, 0x665084 },
+		{ 0x0088, 0x665088 },
+		{ 0x008c, 0x66508c },
+		{ 0x0090, 0x665090 },
+		{ 0x0094, 0x665094 },
+		{ 0x00a0, 0x6650a0 },
+		{ 0x00a4, 0x6650a4 },
+		{ 0x00b0, 0x6650b0 },
+		{ 0x00b4, 0x6650b4 },
+		{ 0x00b8, 0x6650b8 },
+		{ 0x00c0, 0x6650c0 },
+		{ 0x00c4, 0x6650c4 },
+		{ 0x00e0, 0x6650e0 },
+		{ 0x00e4, 0x6650e4 },
+		{ 0x00e8, 0x6650e8 },
+		{ 0x0100, 0x665100 },
+		{ 0x0104, 0x665104 },
+		{ 0x0108, 0x665108 },
+		{ 0x010c, 0x66510c },
+		{ 0x0110, 0x665110 },
+		{ 0x0118, 0x665118 },
+		{ 0x011c, 0x66511c },
+		{ 0x0120, 0x665120 },
+		{ 0x0124, 0x665124 },
+		{ 0x0130, 0x665130 },
+		{ 0x0134, 0x665134 },
+		{ 0x0138, 0x665138 },
+		{ 0x013c, 0x66513c },
+		{ 0x0140, 0x665140 },
+		{ 0x0144, 0x665144 },
+		{ 0x0148, 0x665148 },
+		{ 0x014c, 0x66514c },
+		{ 0x0150, 0x665150 },
+		{ 0x0154, 0x665154 },
+		{ 0x0158, 0x665158 },
+		{ 0x015c, 0x66515c },
+		{ 0x0160, 0x665160 },
+		{ 0x0164, 0x665164 },
+		{ 0x0168, 0x665168 },
+		{ 0x016c, 0x66516c },
+		{ 0x0400, 0x665400 },
+		{ 0x0404, 0x665404 },
+		{ 0x0408, 0x665408 },
+		{ 0x040c, 0x66540c },
+		{ 0x0410, 0x665410 },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_chan
+nve0_disp_ovly_mthd_chan = {
+	.name = "Overlay",
+	.addr = 0x001000,
+	.data = {
+		{ "Global", 1, &nve0_disp_ovly_mthd_base },
+		{}
+	}
+};
+
+/*******************************************************************************
+ * Base display object
+ ******************************************************************************/
+
 static struct nouveau_oclass
 nve0_disp_sclass[] = {
 	{ NVE0_DISP_MAST_CLASS, &nvd0_disp_mast_ofuncs },
@@ -41,10 +210,14 @@ nve0_disp_sclass[] = {
 
 static struct nouveau_oclass
 nve0_disp_base_oclass[] = {
-	{ NVE0_DISP_CLASS, &nvd0_disp_base_ofuncs, nva3_disp_base_omthds },
+	{ NVE0_DISP_CLASS, &nvd0_disp_base_ofuncs, nvd0_disp_base_omthds },
 	{}
 };
 
+/*******************************************************************************
+ * Display engine implementation
+ ******************************************************************************/
+
 static int
 nve0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	       struct nouveau_oclass *oclass, void *data, u32 size,
@@ -54,9 +227,6 @@ nve0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	int heads = nv_rd32(parent, 0x022448);
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000001)
-		return -ENODEV;
-
 	ret = nouveau_disp_create(parent, engine, oclass, heads,
 				  "PDISP", "display", &priv);
 	*pobject = nv_object(priv);
@@ -76,17 +246,21 @@ nve0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nvd0_hda_eld;
 	priv->sor.hdmi = nvd0_hdmi_ctrl;
-	priv->sor.dp = &nvd0_sor_dp_func;
 	return 0;
 }
 
-struct nouveau_oclass
-nve0_disp_oclass = {
-	.handle = NV_ENGINE(DISP, 0x91),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nve0_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x91),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nve0_disp_ctor,
 		.dtor = _nouveau_disp_dtor,
 		.init = _nouveau_disp_init,
 		.fini = _nouveau_disp_fini,
 	},
-};
+	.base.outp =  nvd0_disp_outp_sclass,
+	.mthd.core = &nve0_disp_mast_mthd_chan,
+	.mthd.base = &nvd0_disp_sync_mthd_chan,
+	.mthd.ovly = &nve0_disp_ovly_mthd_chan,
+	.mthd.prev = -0x020000,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
index 42aa6b97dbe..104388081d7 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
@@ -29,6 +29,10 @@
 
 #include "nv50.h"
 
+/*******************************************************************************
+ * Base display object
+ ******************************************************************************/
+
 static struct nouveau_oclass
 nvf0_disp_sclass[] = {
 	{ NVF0_DISP_MAST_CLASS, &nvd0_disp_mast_ofuncs },
@@ -41,10 +45,14 @@ nvf0_disp_sclass[] = {
 
 static struct nouveau_oclass
 nvf0_disp_base_oclass[] = {
-	{ NVF0_DISP_CLASS, &nvd0_disp_base_ofuncs, nva3_disp_base_omthds },
+	{ NVF0_DISP_CLASS, &nvd0_disp_base_ofuncs, nvd0_disp_base_omthds },
 	{}
 };
 
+/*******************************************************************************
+ * Display engine implementation
+ ******************************************************************************/
+
 static int
 nvf0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	       struct nouveau_oclass *oclass, void *data, u32 size,
@@ -54,9 +62,6 @@ nvf0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	int heads = nv_rd32(parent, 0x022448);
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000001)
-		return -ENODEV;
-
 	ret = nouveau_disp_create(parent, engine, oclass, heads,
 				  "PDISP", "display", &priv);
 	*pobject = nv_object(priv);
@@ -76,17 +81,21 @@ nvf0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->sor.power = nv50_sor_power;
 	priv->sor.hda_eld = nvd0_hda_eld;
 	priv->sor.hdmi = nvd0_hdmi_ctrl;
-	priv->sor.dp = &nvd0_sor_dp_func;
 	return 0;
 }
 
-struct nouveau_oclass
-nvf0_disp_oclass = {
-	.handle = NV_ENGINE(DISP, 0x92),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nvf0_disp_oclass = &(struct nv50_disp_impl) {
+	.base.base.handle = NV_ENGINE(DISP, 0x92),
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nvf0_disp_ctor,
 		.dtor = _nouveau_disp_dtor,
 		.init = _nouveau_disp_init,
 		.fini = _nouveau_disp_fini,
 	},
-};
+	.base.outp =  nvd0_disp_outp_sclass,
+	.mthd.core = &nve0_disp_mast_mthd_chan,
+	.mthd.base = &nvd0_disp_sync_mthd_chan,
+	.mthd.ovly = &nve0_disp_ovly_mthd_chan,
+	.mthd.prev = -0x020000,
+}.base.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/outp.c b/drivers/gpu/drm/nouveau/core/engine/disp/outp.c
new file mode 100644
index 00000000000..ad9ba7ccec7
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/outp.c
@@ -0,0 +1,137 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/i2c.h>
+#include <subdev/bios.h>
+#include <subdev/bios/conn.h>
+
+#include "outp.h"
+
+int
+_nvkm_output_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nvkm_output *outp = (void *)object;
+	nv_ofuncs(outp->conn)->fini(nv_object(outp->conn), suspend);
+	return nouveau_object_fini(&outp->base, suspend);
+}
+
+int
+_nvkm_output_init(struct nouveau_object *object)
+{
+	struct nvkm_output *outp = (void *)object;
+	int ret = nouveau_object_init(&outp->base);
+	if (ret == 0)
+		nv_ofuncs(outp->conn)->init(nv_object(outp->conn));
+	return 0;
+}
+
+void
+_nvkm_output_dtor(struct nouveau_object *object)
+{
+	struct nvkm_output *outp = (void *)object;
+	list_del(&outp->head);
+	nouveau_object_ref(NULL, (void *)&outp->conn);
+	nouveau_object_destroy(&outp->base);
+}
+
+int
+nvkm_output_create_(struct nouveau_object *parent,
+		    struct nouveau_object *engine,
+		    struct nouveau_oclass *oclass,
+		    struct dcb_output *dcbE, int index,
+		    int length, void **pobject)
+{
+	struct nouveau_bios *bios = nouveau_bios(engine);
+	struct nouveau_i2c *i2c = nouveau_i2c(parent);
+	struct nouveau_disp *disp = (void *)engine;
+	struct nvbios_connE connE;
+	struct nvkm_output *outp;
+	u8  ver, hdr;
+	u32 data;
+	int ret;
+
+	ret = nouveau_object_create_(parent, engine, oclass, 0, length, pobject);
+	outp = *pobject;
+	if (ret)
+		return ret;
+
+	outp->info = *dcbE;
+	outp->index = index;
+
+	DBG("type %02x loc %d or %d link %d con %x edid %x bus %d head %x\n",
+	    dcbE->type, dcbE->location, dcbE->or, dcbE->type >= 2 ?
+	    dcbE->sorconf.link : 0, dcbE->connector, dcbE->i2c_index,
+	    dcbE->bus, dcbE->heads);
+
+	outp->port = i2c->find(i2c, outp->info.i2c_index);
+	outp->edid = outp->port;
+
+	data = nvbios_connEp(bios, outp->info.connector, &ver, &hdr, &connE);
+	if (!data) {
+		DBG("vbios connector data not found\n");
+		memset(&connE, 0x00, sizeof(connE));
+		connE.type = DCB_CONNECTOR_NONE;
+	}
+
+	ret = nouveau_object_ctor(parent, engine, nvkm_connector_oclass,
+				 &connE, outp->info.connector,
+				 (struct nouveau_object **)&outp->conn);
+	if (ret < 0) {
+		ERR("error %d creating connector, disabling\n", ret);
+		return ret;
+	}
+
+	list_add_tail(&outp->head, &disp->outp);
+	return 0;
+}
+
+int
+_nvkm_output_ctor(struct nouveau_object *parent,
+		  struct nouveau_object *engine,
+		  struct nouveau_oclass *oclass, void *dcbE, u32 index,
+		  struct nouveau_object **pobject)
+{
+	struct nvkm_output *outp;
+	int ret;
+
+	ret = nvkm_output_create(parent, engine, oclass, dcbE, index, &outp);
+	*pobject = nv_object(outp);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+struct nouveau_oclass *
+nvkm_output_oclass = &(struct nvkm_output_impl) {
+	.base = {
+		.handle = 0,
+		.ofuncs = &(struct nouveau_ofuncs) {
+			.ctor = _nvkm_output_ctor,
+			.dtor = _nvkm_output_dtor,
+			.init = _nvkm_output_init,
+			.fini = _nvkm_output_fini,
+		},
+	},
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/outp.h b/drivers/gpu/drm/nouveau/core/engine/disp/outp.h
new file mode 100644
index 00000000000..bc76fbf8571
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/outp.h
@@ -0,0 +1,59 @@
+#ifndef __NVKM_DISP_OUTP_H__
+#define __NVKM_DISP_OUTP_H__
+
+#include "priv.h"
+
+struct nvkm_output {
+	struct nouveau_object base;
+	struct list_head head;
+
+	struct dcb_output info;
+	int index;
+
+	struct nouveau_i2c_port *port;
+	struct nouveau_i2c_port *edid;
+
+	struct nvkm_connector *conn;
+};
+
+#define nvkm_output_create(p,e,c,b,i,d)                                        \
+	nvkm_output_create_((p), (e), (c), (b), (i), sizeof(**d), (void **)d)
+#define nvkm_output_destroy(d) ({                                              \
+	struct nvkm_output *_outp = (d);                                       \
+	_nvkm_output_dtor(nv_object(_outp));                                   \
+})
+#define nvkm_output_init(d) ({                                                 \
+	struct nvkm_output *_outp = (d);                                       \
+	_nvkm_output_init(nv_object(_outp));                                   \
+})
+#define nvkm_output_fini(d,s) ({                                               \
+	struct nvkm_output *_outp = (d);                                       \
+	_nvkm_output_fini(nv_object(_outp), (s));                              \
+})
+
+int nvkm_output_create_(struct nouveau_object *, struct nouveau_object *,
+			struct nouveau_oclass *, struct dcb_output *,
+			int, int, void **);
+
+int  _nvkm_output_ctor(struct nouveau_object *, struct nouveau_object *,
+		       struct nouveau_oclass *, void *, u32,
+		       struct nouveau_object **);
+void _nvkm_output_dtor(struct nouveau_object *);
+int  _nvkm_output_init(struct nouveau_object *);
+int  _nvkm_output_fini(struct nouveau_object *, bool);
+
+struct nvkm_output_impl {
+	struct nouveau_oclass base;
+};
+
+#ifndef MSG
+#define MSG(l,f,a...) do {                                                     \
+	struct nvkm_output *_outp = (void *)outp;                              \
+	nv_##l(nv_object(outp)->engine, "%02x:%04x:%04x: "f, _outp->index,     \
+	       _outp->info.hasht, _outp->info.hashm, ##a);                     \
+} while(0)
+#define DBG(f,a...) MSG(debug, f, ##a)
+#define ERR(f,a...) MSG(error, f, ##a)
+#endif
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c b/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c
new file mode 100644
index 00000000000..eb2d7789555
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.c
@@ -0,0 +1,278 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/i2c.h>
+
+#include "outpdp.h"
+#include "conn.h"
+#include "dport.h"
+
+int
+nvkm_output_dp_train(struct nvkm_output *base, u32 datarate, bool wait)
+{
+	struct nvkm_output_dp *outp = (void *)base;
+	bool retrain = true;
+	u8 link[2], stat[3];
+	u32 linkrate;
+	int ret, i;
+
+	/* check that the link is trained at a high enough rate */
+	ret = nv_rdaux(outp->base.edid, DPCD_LC00_LINK_BW_SET, link, 2);
+	if (ret) {
+		DBG("failed to read link config, assuming no sink\n");
+		goto done;
+	}
+
+	linkrate = link[0] * 27000 * (link[1] & DPCD_LC01_LANE_COUNT_SET);
+	linkrate = (linkrate * 8) / 10; /* 8B/10B coding overhead */
+	datarate = (datarate + 9) / 10; /* -> decakilobits */
+	if (linkrate < datarate) {
+		DBG("link not trained at sufficient rate\n");
+		goto done;
+	}
+
+	/* check that link is still trained */
+	ret = nv_rdaux(outp->base.edid, DPCD_LS02, stat, 3);
+	if (ret) {
+		DBG("failed to read link status, assuming no sink\n");
+		goto done;
+	}
+
+	if (stat[2] & DPCD_LS04_INTERLANE_ALIGN_DONE) {
+		for (i = 0; i < (link[1] & DPCD_LC01_LANE_COUNT_SET); i++) {
+			u8 lane = (stat[i >> 1] >> ((i & 1) * 4)) & 0x0f;
+			if (!(lane & DPCD_LS02_LANE0_CR_DONE) ||
+			    !(lane & DPCD_LS02_LANE0_CHANNEL_EQ_DONE) ||
+			    !(lane & DPCD_LS02_LANE0_SYMBOL_LOCKED)) {
+				DBG("lane %d not equalised\n", lane);
+				goto done;
+			}
+		}
+		retrain = false;
+	} else {
+		DBG("no inter-lane alignment\n");
+	}
+
+done:
+	if (retrain || !atomic_read(&outp->lt.done)) {
+		/* no sink, but still need to configure source */
+		if (outp->dpcd[DPCD_RC00_DPCD_REV] == 0x00) {
+			outp->dpcd[DPCD_RC01_MAX_LINK_RATE] =
+				outp->base.info.dpconf.link_bw;
+			outp->dpcd[DPCD_RC02] =
+				outp->base.info.dpconf.link_nr;
+		}
+		atomic_set(&outp->lt.done, 0);
+		schedule_work(&outp->lt.work);
+	} else {
+		nouveau_event_get(outp->irq);
+	}
+
+	if (wait) {
+		if (!wait_event_timeout(outp->lt.wait,
+					atomic_read(&outp->lt.done),
+					msecs_to_jiffies(2000)))
+			ret = -ETIMEDOUT;
+	}
+
+	return ret;
+}
+
+static void
+nvkm_output_dp_enable(struct nvkm_output_dp *outp, bool present)
+{
+	struct nouveau_i2c_port *port = outp->base.edid;
+	if (present) {
+		if (!outp->present) {
+			nouveau_i2c(port)->acquire_pad(port, 0);
+			DBG("aux power -> always\n");
+			outp->present = true;
+		}
+		nvkm_output_dp_train(&outp->base, 0, true);
+	} else {
+		if (outp->present) {
+			nouveau_i2c(port)->release_pad(port);
+			DBG("aux power -> demand\n");
+			outp->present = false;
+		}
+		atomic_set(&outp->lt.done, 0);
+	}
+}
+
+static void
+nvkm_output_dp_detect(struct nvkm_output_dp *outp)
+{
+	struct nouveau_i2c_port *port = outp->base.edid;
+	int ret = nouveau_i2c(port)->acquire_pad(port, 0);
+	if (ret == 0) {
+		ret = nv_rdaux(outp->base.edid, DPCD_RC00_DPCD_REV,
+			       outp->dpcd, sizeof(outp->dpcd));
+		nvkm_output_dp_enable(outp, ret == 0);
+		nouveau_i2c(port)->release_pad(port);
+	}
+}
+
+static void
+nvkm_output_dp_service_work(struct work_struct *work)
+{
+	struct nvkm_output_dp *outp = container_of(work, typeof(*outp), work);
+	struct nouveau_disp *disp = nouveau_disp(outp);
+	int type = atomic_xchg(&outp->pending, 0);
+	u32 send = 0;
+
+	if (type & (NVKM_I2C_PLUG | NVKM_I2C_UNPLUG)) {
+		nvkm_output_dp_detect(outp);
+		if (type & NVKM_I2C_UNPLUG)
+			send |= NVKM_HPD_UNPLUG;
+		if (type & NVKM_I2C_PLUG)
+			send |= NVKM_HPD_PLUG;
+		nouveau_event_get(outp->base.conn->hpd.event);
+	}
+
+	if (type & NVKM_I2C_IRQ) {
+		nvkm_output_dp_train(&outp->base, 0, true);
+		send |= NVKM_HPD_IRQ;
+	}
+
+	nouveau_event_trigger(disp->hpd, send, outp->base.info.connector);
+}
+
+static int
+nvkm_output_dp_service(void *data, u32 type, int index)
+{
+	struct nvkm_output_dp *outp = data;
+	DBG("HPD: %d\n", type);
+	atomic_or(type, &outp->pending);
+	schedule_work(&outp->work);
+	return NVKM_EVENT_DROP;
+}
+
+int
+_nvkm_output_dp_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nvkm_output_dp *outp = (void *)object;
+	nouveau_event_put(outp->irq);
+	nvkm_output_dp_enable(outp, false);
+	return nvkm_output_fini(&outp->base, suspend);
+}
+
+int
+_nvkm_output_dp_init(struct nouveau_object *object)
+{
+	struct nvkm_output_dp *outp = (void *)object;
+	nvkm_output_dp_detect(outp);
+	return nvkm_output_init(&outp->base);
+}
+
+void
+_nvkm_output_dp_dtor(struct nouveau_object *object)
+{
+	struct nvkm_output_dp *outp = (void *)object;
+	nouveau_event_ref(NULL, &outp->irq);
+	nvkm_output_destroy(&outp->base);
+}
+
+int
+nvkm_output_dp_create_(struct nouveau_object *parent,
+		       struct nouveau_object *engine,
+		       struct nouveau_oclass *oclass,
+		       struct dcb_output *info, int index,
+		       int length, void **pobject)
+{
+	struct nouveau_bios *bios = nouveau_bios(parent);
+	struct nouveau_i2c *i2c = nouveau_i2c(parent);
+	struct nvkm_output_dp *outp;
+	u8  hdr, cnt, len;
+	u32 data;
+	int ret;
+
+	ret = nvkm_output_create_(parent, engine, oclass, info, index,
+				  length, pobject);
+	outp = *pobject;
+	if (ret)
+		return ret;
+
+	nouveau_event_ref(NULL, &outp->base.conn->hpd.event);
+
+	/* access to the aux channel is not optional... */
+	if (!outp->base.edid) {
+		ERR("aux channel not found\n");
+		return -ENODEV;
+	}
+
+	/* nor is the bios data for this output... */
+	data = nvbios_dpout_match(bios, outp->base.info.hasht,
+				  outp->base.info.hashm, &outp->version,
+				  &hdr, &cnt, &len, &outp->info);
+	if (!data) {
+		ERR("no bios dp data\n");
+		return -ENODEV;
+	}
+
+	DBG("bios dp %02x %02x %02x %02x\n", outp->version, hdr, cnt, len);
+
+	/* link training */
+	INIT_WORK(&outp->lt.work, nouveau_dp_train);
+	init_waitqueue_head(&outp->lt.wait);
+	atomic_set(&outp->lt.done, 0);
+
+	/* link maintenance */
+	ret = nouveau_event_new(i2c->ntfy, NVKM_I2C_IRQ, outp->base.edid->index,
+				nvkm_output_dp_service, outp, &outp->irq);
+	if (ret) {
+		ERR("error monitoring aux irq event: %d\n", ret);
+		return ret;
+	}
+
+	INIT_WORK(&outp->work, nvkm_output_dp_service_work);
+
+	/* hotplug detect, replaces gpio-based mechanism with aux events */
+	ret = nouveau_event_new(i2c->ntfy, NVKM_I2C_PLUG | NVKM_I2C_UNPLUG,
+				outp->base.edid->index,
+				nvkm_output_dp_service, outp,
+			       &outp->base.conn->hpd.event);
+	if (ret) {
+		ERR("error monitoring aux hpd events: %d\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+int
+_nvkm_output_dp_ctor(struct nouveau_object *parent,
+		     struct nouveau_object *engine,
+		     struct nouveau_oclass *oclass, void *info, u32 index,
+		     struct nouveau_object **pobject)
+{
+	struct nvkm_output_dp *outp;
+	int ret;
+
+	ret = nvkm_output_dp_create(parent, engine, oclass, info, index, &outp);
+	*pobject = nv_object(outp);
+	if (ret)
+		return ret;
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.h b/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.h
new file mode 100644
index 00000000000..ff33ba12cb6
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/outpdp.h
@@ -0,0 +1,65 @@
+#ifndef __NVKM_DISP_OUTP_DP_H__
+#define __NVKM_DISP_OUTP_DP_H__
+
+#include <subdev/bios.h>
+#include <subdev/bios/dp.h>
+
+#include "outp.h"
+
+struct nvkm_output_dp {
+	struct nvkm_output base;
+
+	struct nvbios_dpout info;
+	u8 version;
+
+	struct nouveau_eventh *irq;
+	struct nouveau_eventh *hpd;
+	struct work_struct work;
+	atomic_t pending;
+	bool present;
+	u8 dpcd[16];
+
+	struct {
+		struct work_struct work;
+		wait_queue_head_t wait;
+		atomic_t done;
+	} lt;
+};
+
+#define nvkm_output_dp_create(p,e,c,b,i,d)                                     \
+	nvkm_output_dp_create_((p), (e), (c), (b), (i), sizeof(**d), (void **)d)
+#define nvkm_output_dp_destroy(d) ({                                           \
+	struct nvkm_output_dp *_outp = (d);                                    \
+	_nvkm_output_dp_dtor(nv_object(_outp));                                \
+})
+#define nvkm_output_dp_init(d) ({                                              \
+	struct nvkm_output_dp *_outp = (d);                                    \
+	_nvkm_output_dp_init(nv_object(_outp));                                \
+})
+#define nvkm_output_dp_fini(d,s) ({                                            \
+	struct nvkm_output_dp *_outp = (d);                                    \
+	_nvkm_output_dp_fini(nv_object(_outp), (s));                           \
+})
+
+int nvkm_output_dp_create_(struct nouveau_object *, struct nouveau_object *,
+			   struct nouveau_oclass *, struct dcb_output *,
+			   int, int, void **);
+
+int  _nvkm_output_dp_ctor(struct nouveau_object *, struct nouveau_object *,
+			  struct nouveau_oclass *, void *, u32,
+			  struct nouveau_object **);
+void _nvkm_output_dp_dtor(struct nouveau_object *);
+int  _nvkm_output_dp_init(struct nouveau_object *);
+int  _nvkm_output_dp_fini(struct nouveau_object *, bool);
+
+struct nvkm_output_dp_impl {
+	struct nvkm_output_impl base;
+	int (*pattern)(struct nvkm_output_dp *, int);
+	int (*lnk_pwr)(struct nvkm_output_dp *, int nr);
+	int (*lnk_ctl)(struct nvkm_output_dp *, int nr, int bw, bool ef);
+	int (*drv_ctl)(struct nvkm_output_dp *, int ln, int vs, int pe, int pc);
+};
+
+int nvkm_output_dp_train(struct nvkm_output *, u32 rate, bool wait);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/piornv50.c b/drivers/gpu/drm/nouveau/core/engine/disp/piornv50.c
index 2c8ce351b52..fe0f256f11b 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/piornv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/piornv50.c
@@ -33,68 +33,107 @@
 #include "nv50.h"
 
 /******************************************************************************
- * DisplayPort
+ * TMDS
  *****************************************************************************/
-static struct nouveau_i2c_port *
-nv50_pior_dp_find(struct nouveau_disp *disp, struct dcb_output *outp)
+
+static int
+nv50_pior_tmds_ctor(struct nouveau_object *parent,
+		    struct nouveau_object *engine,
+		    struct nouveau_oclass *oclass, void *info, u32 index,
+		    struct nouveau_object **pobject)
 {
-	struct nouveau_i2c *i2c = nouveau_i2c(disp);
-	return i2c->find_type(i2c, NV_I2C_TYPE_EXTAUX(outp->extdev));
+	struct nouveau_i2c *i2c = nouveau_i2c(parent);
+	struct nvkm_output *outp;
+	int ret;
+
+	ret = nvkm_output_create(parent, engine, oclass, info, index, &outp);
+	*pobject = nv_object(outp);
+	if (ret)
+		return ret;
+
+	outp->edid = i2c->find_type(i2c, NV_I2C_TYPE_EXTDDC(outp->info.extdev));
+	return 0;
 }
 
+struct nvkm_output_impl
+nv50_pior_tmds_impl = {
+	.base.handle = DCB_OUTPUT_TMDS | 0x0100,
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_pior_tmds_ctor,
+		.dtor = _nvkm_output_dtor,
+		.init = _nvkm_output_init,
+		.fini = _nvkm_output_fini,
+	},
+};
+
+/******************************************************************************
+ * DisplayPort
+ *****************************************************************************/
+
 static int
-nv50_pior_dp_pattern(struct nouveau_disp *disp, struct dcb_output *outp,
-		     int head, int pattern)
+nv50_pior_dp_pattern(struct nvkm_output_dp *outp, int pattern)
 {
-	struct nouveau_i2c_port *port;
-	int ret = -EINVAL;
-
-	port = nv50_pior_dp_find(disp, outp);
-	if (port) {
-		if (port->func->pattern)
-			ret = port->func->pattern(port, pattern);
-		else
-			ret = 0;
-	}
-
-	return ret;
+	struct nouveau_i2c_port *port = outp->base.edid;
+	if (port && port->func->pattern)
+		return port->func->pattern(port, pattern);
+	return port ? 0 : -ENODEV;
 }
 
 static int
-nv50_pior_dp_lnk_ctl(struct nouveau_disp *disp, struct dcb_output *outp,
-		     int head, int lane_nr, int link_bw, bool enh)
+nv50_pior_dp_lnk_pwr(struct nvkm_output_dp *outp, int nr)
 {
-	struct nouveau_i2c_port *port;
-	int ret = -EINVAL;
+	return 0;
+}
 
-	port = nv50_pior_dp_find(disp, outp);
+static int
+nv50_pior_dp_lnk_ctl(struct nvkm_output_dp *outp, int nr, int bw, bool ef)
+{
+	struct nouveau_i2c_port *port = outp->base.edid;
 	if (port && port->func->lnk_ctl)
-		ret = port->func->lnk_ctl(port, lane_nr, link_bw, enh);
+		return port->func->lnk_ctl(port, nr, bw, ef);
+	return port ? 0 : -ENODEV;
+}
 
-	return ret;
+static int
+nv50_pior_dp_drv_ctl(struct nvkm_output_dp *outp, int ln, int vs, int pe, int pc)
+{
+	struct nouveau_i2c_port *port = outp->base.edid;
+	if (port && port->func->drv_ctl)
+		return port->func->drv_ctl(port, ln, vs, pe);
+	return port ? 0 : -ENODEV;
 }
 
 static int
-nv50_pior_dp_drv_ctl(struct nouveau_disp *disp, struct dcb_output *outp,
-		     int head, int lane, int vsw, int pre)
+nv50_pior_dp_ctor(struct nouveau_object *parent,
+		  struct nouveau_object *engine,
+		  struct nouveau_oclass *oclass, void *info, u32 index,
+		  struct nouveau_object **pobject)
 {
-	struct nouveau_i2c_port *port;
-	int ret = -EINVAL;
-
-	port = nv50_pior_dp_find(disp, outp);
-	if (port) {
-		if (port->func->drv_ctl)
-			ret = port->func->drv_ctl(port, lane, vsw, pre);
-		else
-			ret = 0;
-	}
+	struct nouveau_i2c *i2c = nouveau_i2c(parent);
+	struct nvkm_output_dp *outp;
+	int ret;
 
-	return ret;
+	ret = nvkm_output_dp_create(parent, engine, oclass, info, index, &outp);
+	*pobject = nv_object(outp);
+	if (ret)
+		return ret;
+
+	outp->base.edid = i2c->find_type(i2c, NV_I2C_TYPE_EXTAUX(
+					 outp->base.info.extdev));
+	return 0;
 }
 
-const struct nouveau_dp_func
-nv50_pior_dp_func = {
+struct nvkm_output_dp_impl
+nv50_pior_dp_impl = {
+	.base.base.handle = DCB_OUTPUT_DP | 0x0010,
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_pior_dp_ctor,
+		.dtor = _nvkm_output_dp_dtor,
+		.init = _nvkm_output_dp_init,
+		.fini = _nvkm_output_dp_fini,
+	},
 	.pattern = nv50_pior_dp_pattern,
+	.lnk_pwr = nv50_pior_dp_lnk_pwr,
 	.lnk_ctl = nv50_pior_dp_lnk_ctl,
 	.drv_ctl = nv50_pior_dp_drv_ctl,
 };
@@ -102,6 +141,7 @@ nv50_pior_dp_func = {
 /******************************************************************************
  * General PIOR handling
  *****************************************************************************/
+
 int
 nv50_pior_power(struct nv50_disp_priv *priv, int or, u32 data)
 {
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/priv.h b/drivers/gpu/drm/nouveau/core/engine/disp/priv.h
new file mode 100644
index 00000000000..26e9a42569c
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/priv.h
@@ -0,0 +1,42 @@
+#ifndef __NVKM_DISP_PRIV_H__
+#define __NVKM_DISP_PRIV_H__
+
+#include <subdev/bios.h>
+#include <subdev/bios/dcb.h>
+#include <subdev/bios/conn.h>
+
+#include <engine/disp.h>
+
+struct nouveau_disp_impl {
+	struct nouveau_oclass base;
+	struct nouveau_oclass **outp;
+	struct nouveau_oclass **conn;
+};
+
+#define nouveau_disp_create(p,e,c,h,i,x,d)                                     \
+	nouveau_disp_create_((p), (e), (c), (h), (i), (x),                     \
+			     sizeof(**d), (void **)d)
+#define nouveau_disp_destroy(d) ({                                             \
+	struct nouveau_disp *disp = (d);                                       \
+	_nouveau_disp_dtor(nv_object(disp));                                   \
+})
+#define nouveau_disp_init(d) ({                                                \
+	struct nouveau_disp *disp = (d);                                       \
+	_nouveau_disp_init(nv_object(disp));                                   \
+})
+#define nouveau_disp_fini(d,s) ({                                              \
+	struct nouveau_disp *disp = (d);                                       \
+	_nouveau_disp_fini(nv_object(disp), (s));                              \
+})
+
+int  nouveau_disp_create_(struct nouveau_object *, struct nouveau_object *,
+			  struct nouveau_oclass *, int heads,
+			  const char *, const char *, int, void **);
+void _nouveau_disp_dtor(struct nouveau_object *);
+int  _nouveau_disp_init(struct nouveau_object *);
+int  _nouveau_disp_fini(struct nouveau_object *, bool);
+
+extern struct nouveau_oclass *nvkm_output_oclass;
+extern struct nouveau_oclass *nvkm_connector_oclass;
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/sornv50.c b/drivers/gpu/drm/nouveau/core/engine/disp/sornv50.c
index 526b7524289..7a1ebdfa9e1 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/sornv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/sornv50.c
@@ -47,8 +47,12 @@ int
 nv50_sor_mthd(struct nouveau_object *object, u32 mthd, void *args, u32 size)
 {
 	struct nv50_disp_priv *priv = (void *)object->engine;
+	const u8  type = (mthd & NV50_DISP_SOR_MTHD_TYPE) >> 12;
 	const u8  head = (mthd & NV50_DISP_SOR_MTHD_HEAD) >> 3;
+	const u8  link = (mthd & NV50_DISP_SOR_MTHD_LINK) >> 2;
 	const u8    or = (mthd & NV50_DISP_SOR_MTHD_OR);
+	const u16 mask = (0x0100 << head) | (0x0040 << link) | (0x0001 << or);
+	struct nvkm_output *outp = NULL, *temp;
 	u32 data;
 	int ret = -EINVAL;
 
@@ -56,6 +60,13 @@ nv50_sor_mthd(struct nouveau_object *object, u32 mthd, void *args, u32 size)
 		return -EINVAL;
 	data = *(u32 *)args;
 
+	list_for_each_entry(temp, &priv->base.outp, head) {
+		if ((temp->info.hasht & 0xff) == type &&
+		    (temp->info.hashm & mask) == mask) {
+			outp = temp;
+			break;
+		}
+	}
 
 	switch (mthd & ~0x3f) {
 	case NV50_DISP_SOR_PWR:
@@ -71,6 +82,24 @@ nv50_sor_mthd(struct nouveau_object *object, u32 mthd, void *args, u32 size)
 		priv->sor.lvdsconf = data & NV50_DISP_SOR_LVDS_SCRIPT_ID;
 		ret = 0;
 		break;
+	case NV94_DISP_SOR_DP_PWR:
+		if (outp) {
+			struct nvkm_output_dp *outpdp = (void *)outp;
+			switch (data) {
+			case NV94_DISP_SOR_DP_PWR_STATE_OFF:
+				nouveau_event_put(outpdp->irq);
+				((struct nvkm_output_dp_impl *)nv_oclass(outp))
+					->lnk_pwr(outpdp, 0);
+				atomic_set(&outpdp->lt.done, 0);
+				break;
+			case NV94_DISP_SOR_DP_PWR_STATE_ON:
+				nvkm_output_dp_train(&outpdp->base, 0, true);
+				break;
+			default:
+				return -EINVAL;
+			}
+		}
+		break;
 	default:
 		BUG_ON(1);
 	}
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/sornv94.c b/drivers/gpu/drm/nouveau/core/engine/disp/sornv94.c
index eea3ef59693..05487cda84a 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/sornv94.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/sornv94.c
@@ -29,19 +29,21 @@
 #include <subdev/bios/dcb.h>
 #include <subdev/bios/dp.h>
 #include <subdev/bios/init.h>
+#include <subdev/timer.h>
 
 #include "nv50.h"
+#include "outpdp.h"
 
 static inline u32
-nv94_sor_soff(struct dcb_output *outp)
+nv94_sor_soff(struct nvkm_output_dp *outp)
 {
-	return (ffs(outp->or) - 1) * 0x800;
+	return (ffs(outp->base.info.or) - 1) * 0x800;
 }
 
 static inline u32
-nv94_sor_loff(struct dcb_output *outp)
+nv94_sor_loff(struct nvkm_output_dp *outp)
 {
-	return nv94_sor_soff(outp) + !(outp->sorconf.link & 1) * 0x80;
+	return nv94_sor_soff(outp) + !(outp->base.info.sorconf.link & 1) * 0x80;
 }
 
 static inline u32
@@ -55,77 +57,96 @@ nv94_sor_dp_lane_map(struct nv50_disp_priv *priv, u8 lane)
 }
 
 static int
-nv94_sor_dp_pattern(struct nouveau_disp *disp, struct dcb_output *outp,
-		    int head, int pattern)
+nv94_sor_dp_pattern(struct nvkm_output_dp *outp, int pattern)
 {
-	struct nv50_disp_priv *priv = (void *)disp;
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
 	const u32 loff = nv94_sor_loff(outp);
 	nv_mask(priv, 0x61c10c + loff, 0x0f000000, pattern << 24);
 	return 0;
 }
 
+int
+nv94_sor_dp_lnk_pwr(struct nvkm_output_dp *outp, int nr)
+{
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
+	const u32 soff = nv94_sor_soff(outp);
+	const u32 loff = nv94_sor_loff(outp);
+	u32 mask = 0, i;
+
+	for (i = 0; i < nr; i++)
+		mask |= 1 << (nv94_sor_dp_lane_map(priv, i) >> 3);
+
+	nv_mask(priv, 0x61c130 + loff, 0x0000000f, mask);
+	nv_mask(priv, 0x61c034 + soff, 0x80000000, 0x80000000);
+	nv_wait(priv, 0x61c034 + soff, 0x80000000, 0x00000000);
+	return 0;
+}
+
 static int
-nv94_sor_dp_lnk_ctl(struct nouveau_disp *disp, struct dcb_output *outp,
-		    int head, int link_nr, int link_bw, bool enh_frame)
+nv94_sor_dp_lnk_ctl(struct nvkm_output_dp *outp, int nr, int bw, bool ef)
 {
-	struct nv50_disp_priv *priv = (void *)disp;
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
 	const u32 soff = nv94_sor_soff(outp);
 	const u32 loff = nv94_sor_loff(outp);
 	u32 dpctrl = 0x00000000;
 	u32 clksor = 0x00000000;
-	u32 lane = 0;
-	int i;
 
-	dpctrl |= ((1 << link_nr) - 1) << 16;
-	if (enh_frame)
+	dpctrl |= ((1 << nr) - 1) << 16;
+	if (ef)
 		dpctrl |= 0x00004000;
-	if (link_bw > 0x06)
+	if (bw > 0x06)
 		clksor |= 0x00040000;
 
-	for (i = 0; i < link_nr; i++)
-		lane |= 1 << (nv94_sor_dp_lane_map(priv, i) >> 3);
-
 	nv_mask(priv, 0x614300 + soff, 0x000c0000, clksor);
 	nv_mask(priv, 0x61c10c + loff, 0x001f4000, dpctrl);
-	nv_mask(priv, 0x61c130 + loff, 0x0000000f, lane);
 	return 0;
 }
 
 static int
-nv94_sor_dp_drv_ctl(struct nouveau_disp *disp, struct dcb_output *outp,
-		    int head, int lane, int swing, int preem)
+nv94_sor_dp_drv_ctl(struct nvkm_output_dp *outp, int ln, int vs, int pe, int pc)
 {
-	struct nouveau_bios *bios = nouveau_bios(disp);
-	struct nv50_disp_priv *priv = (void *)disp;
-	const u32 shift = nv94_sor_dp_lane_map(priv, lane);
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
+	struct nouveau_bios *bios = nouveau_bios(priv);
+	const u32 shift = nv94_sor_dp_lane_map(priv, ln);
 	const u32 loff = nv94_sor_loff(outp);
 	u32 addr, data[3];
 	u8  ver, hdr, cnt, len;
 	struct nvbios_dpout info;
 	struct nvbios_dpcfg ocfg;
 
-	addr = nvbios_dpout_match(bios, outp->hasht, outp->hashm,
+	addr = nvbios_dpout_match(bios, outp->base.info.hasht,
+					outp->base.info.hashm,
 				 &ver, &hdr, &cnt, &len, &info);
 	if (!addr)
 		return -ENODEV;
 
-	addr = nvbios_dpcfg_match(bios, addr, 0, swing, preem,
+	addr = nvbios_dpcfg_match(bios, addr, 0, vs, pe,
 				 &ver, &hdr, &cnt, &len, &ocfg);
 	if (!addr)
 		return -EINVAL;
 
 	data[0] = nv_rd32(priv, 0x61c118 + loff) & ~(0x000000ff << shift);
 	data[1] = nv_rd32(priv, 0x61c120 + loff) & ~(0x000000ff << shift);
-	data[2] = nv_rd32(priv, 0x61c130 + loff) & ~(0x0000ff00);
-	nv_wr32(priv, 0x61c118 + loff, data[0] | (ocfg.drv << shift));
-	nv_wr32(priv, 0x61c120 + loff, data[1] | (ocfg.pre << shift));
-	nv_wr32(priv, 0x61c130 + loff, data[2] | (ocfg.unk << 8));
+	data[2] = nv_rd32(priv, 0x61c130 + loff);
+	if ((data[2] & 0x0000ff00) < (ocfg.tx_pu << 8) || ln == 0)
+		data[2] = (data[2] & ~0x0000ff00) | (ocfg.tx_pu << 8);
+	nv_wr32(priv, 0x61c118 + loff, data[0] | (ocfg.dc << shift));
+	nv_wr32(priv, 0x61c120 + loff, data[1] | (ocfg.pe << shift));
+	nv_wr32(priv, 0x61c130 + loff, data[2] | (ocfg.tx_pu << 8));
 	return 0;
 }
 
-const struct nouveau_dp_func
-nv94_sor_dp_func = {
+struct nvkm_output_dp_impl
+nv94_sor_dp_impl = {
+	.base.base.handle = DCB_OUTPUT_DP,
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nvkm_output_dp_ctor,
+		.dtor = _nvkm_output_dp_dtor,
+		.init = _nvkm_output_dp_init,
+		.fini = _nvkm_output_dp_fini,
+	},
 	.pattern = nv94_sor_dp_pattern,
+	.lnk_pwr = nv94_sor_dp_lnk_pwr,
 	.lnk_ctl = nv94_sor_dp_lnk_ctl,
 	.drv_ctl = nv94_sor_dp_drv_ctl,
 };
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/sornvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/sornvd0.c
index d2df572f16a..97f0e9cd3d4 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/sornvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/sornvd0.c
@@ -29,19 +29,20 @@
 #include <subdev/bios/dcb.h>
 #include <subdev/bios/dp.h>
 #include <subdev/bios/init.h>
+#include <subdev/timer.h>
 
 #include "nv50.h"
 
 static inline u32
-nvd0_sor_soff(struct dcb_output *outp)
+nvd0_sor_soff(struct nvkm_output_dp *outp)
 {
-	return (ffs(outp->or) - 1) * 0x800;
+	return (ffs(outp->base.info.or) - 1) * 0x800;
 }
 
 static inline u32
-nvd0_sor_loff(struct dcb_output *outp)
+nvd0_sor_loff(struct nvkm_output_dp *outp)
 {
-	return nvd0_sor_soff(outp) + !(outp->sorconf.link & 1) * 0x80;
+	return nvd0_sor_soff(outp) + !(outp->base.info.sorconf.link & 1) * 0x80;
 }
 
 static inline u32
@@ -52,77 +53,80 @@ nvd0_sor_dp_lane_map(struct nv50_disp_priv *priv, u8 lane)
 }
 
 static int
-nvd0_sor_dp_pattern(struct nouveau_disp *disp, struct dcb_output *outp,
-		    int head, int pattern)
+nvd0_sor_dp_pattern(struct nvkm_output_dp *outp, int pattern)
 {
-	struct nv50_disp_priv *priv = (void *)disp;
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
 	const u32 loff = nvd0_sor_loff(outp);
 	nv_mask(priv, 0x61c110 + loff, 0x0f0f0f0f, 0x01010101 * pattern);
 	return 0;
 }
 
 static int
-nvd0_sor_dp_lnk_ctl(struct nouveau_disp *disp, struct dcb_output *outp,
-		    int head, int link_nr, int link_bw, bool enh_frame)
+nvd0_sor_dp_lnk_ctl(struct nvkm_output_dp *outp, int nr, int bw, bool ef)
 {
-	struct nv50_disp_priv *priv = (void *)disp;
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
 	const u32 soff = nvd0_sor_soff(outp);
 	const u32 loff = nvd0_sor_loff(outp);
 	u32 dpctrl = 0x00000000;
 	u32 clksor = 0x00000000;
-	u32 lane = 0;
-	int i;
 
-	clksor |= link_bw << 18;
-	dpctrl |= ((1 << link_nr) - 1) << 16;
-	if (enh_frame)
+	clksor |= bw << 18;
+	dpctrl |= ((1 << nr) - 1) << 16;
+	if (ef)
 		dpctrl |= 0x00004000;
 
-	for (i = 0; i < link_nr; i++)
-		lane |= 1 << (nvd0_sor_dp_lane_map(priv, i) >> 3);
-
 	nv_mask(priv, 0x612300 + soff, 0x007c0000, clksor);
 	nv_mask(priv, 0x61c10c + loff, 0x001f4000, dpctrl);
-	nv_mask(priv, 0x61c130 + loff, 0x0000000f, lane);
 	return 0;
 }
 
 static int
-nvd0_sor_dp_drv_ctl(struct nouveau_disp *disp, struct dcb_output *outp,
-		    int head, int lane, int swing, int preem)
+nvd0_sor_dp_drv_ctl(struct nvkm_output_dp *outp, int ln, int vs, int pe, int pc)
 {
-	struct nouveau_bios *bios = nouveau_bios(disp);
-	struct nv50_disp_priv *priv = (void *)disp;
-	const u32 shift = nvd0_sor_dp_lane_map(priv, lane);
+	struct nv50_disp_priv *priv = (void *)nouveau_disp(outp);
+	struct nouveau_bios *bios = nouveau_bios(priv);
+	const u32 shift = nvd0_sor_dp_lane_map(priv, ln);
 	const u32 loff = nvd0_sor_loff(outp);
-	u32 addr, data[3];
+	u32 addr, data[4];
 	u8  ver, hdr, cnt, len;
 	struct nvbios_dpout info;
 	struct nvbios_dpcfg ocfg;
 
-	addr = nvbios_dpout_match(bios, outp->hasht, outp->hashm,
+	addr = nvbios_dpout_match(bios, outp->base.info.hasht,
+					outp->base.info.hashm,
 				 &ver, &hdr, &cnt, &len, &info);
 	if (!addr)
 		return -ENODEV;
 
-	addr = nvbios_dpcfg_match(bios, addr, 0, swing, preem,
+	addr = nvbios_dpcfg_match(bios, addr, pc, vs, pe,
 				 &ver, &hdr, &cnt, &len, &ocfg);
 	if (!addr)
 		return -EINVAL;
 
 	data[0] = nv_rd32(priv, 0x61c118 + loff) & ~(0x000000ff << shift);
 	data[1] = nv_rd32(priv, 0x61c120 + loff) & ~(0x000000ff << shift);
-	data[2] = nv_rd32(priv, 0x61c130 + loff) & ~(0x0000ff00);
-	nv_wr32(priv, 0x61c118 + loff, data[0] | (ocfg.drv << shift));
-	nv_wr32(priv, 0x61c120 + loff, data[1] | (ocfg.pre << shift));
-	nv_wr32(priv, 0x61c130 + loff, data[2] | (ocfg.unk << 8));
-	nv_mask(priv, 0x61c13c + loff, 0x00000000, 0x00000000);
+	data[2] = nv_rd32(priv, 0x61c130 + loff);
+	if ((data[2] & 0x0000ff00) < (ocfg.tx_pu << 8) || ln == 0)
+		data[2] = (data[2] & ~0x0000ff00) | (ocfg.tx_pu << 8);
+	nv_wr32(priv, 0x61c118 + loff, data[0] | (ocfg.dc << shift));
+	nv_wr32(priv, 0x61c120 + loff, data[1] | (ocfg.pe << shift));
+	nv_wr32(priv, 0x61c130 + loff, data[2] | (ocfg.tx_pu << 8));
+	data[3] = nv_rd32(priv, 0x61c13c + loff) & ~(0x000000ff << shift);
+	nv_wr32(priv, 0x61c13c + loff, data[3] | (ocfg.pc << shift));
 	return 0;
 }
 
-const struct nouveau_dp_func
-nvd0_sor_dp_func = {
+struct nvkm_output_dp_impl
+nvd0_sor_dp_impl = {
+	.base.base.handle = DCB_OUTPUT_DP,
+	.base.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nvkm_output_dp_ctor,
+		.dtor = _nvkm_output_dp_dtor,
+		.init = _nvkm_output_dp_init,
+		.fini = _nvkm_output_dp_fini,
+	},
 	.pattern = nvd0_sor_dp_pattern,
+	.lnk_pwr = nv94_sor_dp_lnk_pwr,
 	.lnk_ctl = nvd0_sor_dp_lnk_ctl,
 	.drv_ctl = nvd0_sor_dp_drv_ctl,
 };
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/vga.c b/drivers/gpu/drm/nouveau/core/engine/disp/vga.c
index 5a1c6847459..8836c3cb99c 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/vga.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/vga.c
@@ -138,10 +138,15 @@ nv_wrvgai(void *obj, int head, u16 port, u8 index, u8 value)
 bool
 nv_lockvgac(void *obj, bool lock)
 {
+	struct nouveau_device *dev = nv_device(obj);
+
 	bool locked = !nv_rdvgac(obj, 0, 0x1f);
 	u8 data = lock ? 0x99 : 0x57;
-	nv_wrvgac(obj, 0, 0x1f, data);
-	if (nv_device(obj)->chipset == 0x11) {
+	if (dev->card_type < NV_50)
+		nv_wrvgac(obj, 0, 0x1f, data);
+	else
+		nv_wrvgac(obj, 0, 0x3f, data);
+	if (dev->chipset == 0x11) {
 		if (!(nv_rd32(obj, 0x001084) & 0x10000000))
 			nv_wrvgac(obj, 1, 0x1f, data);
 	}
diff --git a/drivers/gpu/drm/nouveau/core/engine/dmaobj/nvd0.c b/drivers/gpu/drm/nouveau/core/engine/dmaobj/nvd0.c
index 944e73ac485..1cfb3bb9013 100644
--- a/drivers/gpu/drm/nouveau/core/engine/dmaobj/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/dmaobj/nvd0.c
@@ -53,6 +53,9 @@ nvd0_dmaobj_bind(struct nouveau_dmaeng *dmaeng,
 		case NVF0_DISP_MAST_CLASS:
 		case NVF0_DISP_SYNC_CLASS:
 		case NVF0_DISP_OVLY_CLASS:
+		case GM107_DISP_MAST_CLASS:
+		case GM107_DISP_SYNC_CLASS:
+		case GM107_DISP_OVLY_CLASS:
 			break;
 		default:
 			return -EINVAL;
diff --git a/drivers/gpu/drm/nouveau/core/engine/falcon.c b/drivers/gpu/drm/nouveau/core/engine/falcon.c
index e03fc8e4dc1..2914646c870 100644
--- a/drivers/gpu/drm/nouveau/core/engine/falcon.c
+++ b/drivers/gpu/drm/nouveau/core/engine/falcon.c
@@ -56,6 +56,16 @@ _nouveau_falcon_wr32(struct nouveau_object *object, u64 addr, u32 data)
 	nv_wr32(falcon, falcon->addr + addr, data);
 }
 
+static void *
+vmemdup(const void *src, size_t len)
+{
+	void *p = vmalloc(len);
+
+	if (p)
+		memcpy(p, src, len);
+	return p;
+}
+
 int
 _nouveau_falcon_init(struct nouveau_object *object)
 {
@@ -109,9 +119,9 @@ _nouveau_falcon_init(struct nouveau_object *object)
 		snprintf(name, sizeof(name), "nouveau/nv%02x_fuc%03x",
 			 device->chipset, falcon->addr >> 12);
 
-		ret = request_firmware(&fw, name, &device->pdev->dev);
+		ret = request_firmware(&fw, name, nv_device_base(device));
 		if (ret == 0) {
-			falcon->code.data = kmemdup(fw->data, fw->size, GFP_KERNEL);
+			falcon->code.data = vmemdup(fw->data, fw->size);
 			falcon->code.size = fw->size;
 			falcon->data.data = NULL;
 			falcon->data.size = 0;
@@ -128,13 +138,13 @@ _nouveau_falcon_init(struct nouveau_object *object)
 		snprintf(name, sizeof(name), "nouveau/nv%02x_fuc%03xd",
 			 device->chipset, falcon->addr >> 12);
 
-		ret = request_firmware(&fw, name, &device->pdev->dev);
+		ret = request_firmware(&fw, name, nv_device_base(device));
 		if (ret) {
 			nv_error(falcon, "unable to load firmware data\n");
 			return ret;
 		}
 
-		falcon->data.data = kmemdup(fw->data, fw->size, GFP_KERNEL);
+		falcon->data.data = vmemdup(fw->data, fw->size);
 		falcon->data.size = fw->size;
 		release_firmware(fw);
 		if (!falcon->data.data)
@@ -143,13 +153,13 @@ _nouveau_falcon_init(struct nouveau_object *object)
 		snprintf(name, sizeof(name), "nouveau/nv%02x_fuc%03xc",
 			 device->chipset, falcon->addr >> 12);
 
-		ret = request_firmware(&fw, name, &device->pdev->dev);
+		ret = request_firmware(&fw, name, nv_device_base(device));
 		if (ret) {
 			nv_error(falcon, "unable to load firmware code\n");
 			return ret;
 		}
 
-		falcon->code.data = kmemdup(fw->data, fw->size, GFP_KERNEL);
+		falcon->code.data = vmemdup(fw->data, fw->size);
 		falcon->code.size = fw->size;
 		release_firmware(fw);
 		if (!falcon->code.data)
@@ -235,8 +245,8 @@ _nouveau_falcon_fini(struct nouveau_object *object, bool suspend)
 	if (!suspend) {
 		nouveau_gpuobj_ref(NULL, &falcon->core);
 		if (falcon->external) {
-			kfree(falcon->data.data);
-			kfree(falcon->code.data);
+			vfree(falcon->data.data);
+			vfree(falcon->code.data);
 			falcon->code.data = NULL;
 		}
 	}
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/base.c b/drivers/gpu/drm/nouveau/core/engine/fifo/base.c
index d3ec436d9cb..56ed3d73bf8 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/base.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/base.c
@@ -86,12 +86,12 @@ nouveau_fifo_channel_create_(struct nouveau_object *parent,
 	}
 
 	/* map fifo control registers */
-	chan->user = ioremap(pci_resource_start(device->pdev, bar) + addr +
+	chan->user = ioremap(nv_device_resource_start(device, bar) + addr +
 			     (chan->chid * size), size);
 	if (!chan->user)
 		return -EFAULT;
 
-	nouveau_event_trigger(priv->cevent, 0);
+	nouveau_event_trigger(priv->cevent, 1, 0);
 
 	chan->size = size;
 	return 0;
@@ -194,11 +194,11 @@ nouveau_fifo_create_(struct nouveau_object *parent,
 	if (!priv->channel)
 		return -ENOMEM;
 
-	ret = nouveau_event_create(1, &priv->cevent);
+	ret = nouveau_event_create(1, 1, &priv->cevent);
 	if (ret)
 		return ret;
 
-	ret = nouveau_event_create(1, &priv->uevent);
+	ret = nouveau_event_create(1, 1, &priv->uevent);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/gk20a.c b/drivers/gpu/drm/nouveau/core/engine/fifo/gk20a.c
new file mode 100644
index 00000000000..327456eae96
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/gk20a.c
@@ -0,0 +1,35 @@
+/*
+ * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include "nve0.h"
+
+struct nouveau_oclass *
+gk20a_fifo_oclass = &(struct nve0_fifo_impl) {
+	.base.handle = NV_ENGINE(FIFO, 0xea),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nve0_fifo_ctor,
+		.dtor = nve0_fifo_dtor,
+		.init = nve0_fifo_init,
+		.fini = nve0_fifo_fini,
+	},
+	.channels = 128,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nv04.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nv04.c
index 54f26cc801c..c61b16a6388 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nv04.c
@@ -539,7 +539,7 @@ nv04_fifo_intr(struct nouveau_subdev *subdev)
 			}
 
 			if (status & 0x40000000) {
-				nouveau_event_trigger(priv->base.uevent, 0);
+				nouveau_event_trigger(priv->base.uevent, 1, 0);
 				nv_wr32(priv, 0x002100, 0x40000000);
 				status &= ~0x40000000;
 			}
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nv108.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nv108.c
new file mode 100644
index 00000000000..09362a51ba5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nv108.c
@@ -0,0 +1,37 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nve0.h"
+
+struct nouveau_oclass *
+nv108_fifo_oclass = &(struct nve0_fifo_impl) {
+	.base.handle = NV_ENGINE(FIFO, 0x08),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nve0_fifo_ctor,
+		.dtor = nve0_fifo_dtor,
+		.init = nve0_fifo_init,
+		.fini = _nouveau_fifo_fini,
+	},
+	.channels = 1024,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nv84.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nv84.c
index fe0f41e65d9..6e5ac16e546 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/nv84.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nv84.c
@@ -389,14 +389,14 @@ nv84_fifo_cclass = {
  ******************************************************************************/
 
 static void
-nv84_fifo_uevent_enable(struct nouveau_event *event, int index)
+nv84_fifo_uevent_enable(struct nouveau_event *event, int type, int index)
 {
 	struct nv84_fifo_priv *priv = event->priv;
 	nv_mask(priv, 0x002140, 0x40000000, 0x40000000);
 }
 
 static void
-nv84_fifo_uevent_disable(struct nouveau_event *event, int index)
+nv84_fifo_uevent_disable(struct nouveau_event *event, int type, int index)
 {
 	struct nv84_fifo_priv *priv = event->priv;
 	nv_mask(priv, 0x002140, 0x40000000, 0x00000000);
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c
index 9ac94d4e564..ae4a4dc5642 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c
@@ -33,6 +33,7 @@
 
 #include <subdev/timer.h>
 #include <subdev/bar.h>
+#include <subdev/fb.h>
 #include <subdev/vm.h>
 
 #include <engine/dmaobj.h>
@@ -40,8 +41,16 @@
 
 struct nvc0_fifo_priv {
 	struct nouveau_fifo base;
-	struct nouveau_gpuobj *playlist[2];
-	int cur_playlist;
+
+	struct work_struct fault;
+	u64 mask;
+
+	struct {
+		struct nouveau_gpuobj *mem[2];
+		int active;
+		wait_queue_head_t wait;
+	} runlist;
+
 	struct {
 		struct nouveau_gpuobj *mem;
 		struct nouveau_vma bar;
@@ -57,6 +66,11 @@ struct nvc0_fifo_base {
 
 struct nvc0_fifo_chan {
 	struct nouveau_fifo_chan base;
+	enum {
+		STOPPED,
+		RUNNING,
+		KILLED
+	} state;
 };
 
 /*******************************************************************************
@@ -64,29 +78,33 @@ struct nvc0_fifo_chan {
  ******************************************************************************/
 
 static void
-nvc0_fifo_playlist_update(struct nvc0_fifo_priv *priv)
+nvc0_fifo_runlist_update(struct nvc0_fifo_priv *priv)
 {
 	struct nouveau_bar *bar = nouveau_bar(priv);
 	struct nouveau_gpuobj *cur;
 	int i, p;
 
 	mutex_lock(&nv_subdev(priv)->mutex);
-	cur = priv->playlist[priv->cur_playlist];
-	priv->cur_playlist = !priv->cur_playlist;
+	cur = priv->runlist.mem[priv->runlist.active];
+	priv->runlist.active = !priv->runlist.active;
 
 	for (i = 0, p = 0; i < 128; i++) {
-		if (!(nv_rd32(priv, 0x003004 + (i * 8)) & 1))
-			continue;
-		nv_wo32(cur, p + 0, i);
-		nv_wo32(cur, p + 4, 0x00000004);
-		p += 8;
+		struct nvc0_fifo_chan *chan = (void *)priv->base.channel[i];
+		if (chan && chan->state == RUNNING) {
+			nv_wo32(cur, p + 0, i);
+			nv_wo32(cur, p + 4, 0x00000004);
+			p += 8;
+		}
 	}
 	bar->flush(bar);
 
 	nv_wr32(priv, 0x002270, cur->addr >> 12);
 	nv_wr32(priv, 0x002274, 0x01f00000 | (p >> 3));
-	if (!nv_wait(priv, 0x00227c, 0x00100000, 0x00000000))
-		nv_error(priv, "playlist update failed\n");
+
+	if (wait_event_timeout(priv->runlist.wait,
+			       !(nv_rd32(priv, 0x00227c) & 0x00100000),
+			       msecs_to_jiffies(2000)) == 0)
+		nv_error(priv, "runlist update timeout\n");
 	mutex_unlock(&nv_subdev(priv)->mutex);
 }
 
@@ -238,30 +256,32 @@ nvc0_fifo_chan_init(struct nouveau_object *object)
 		return ret;
 
 	nv_wr32(priv, 0x003000 + (chid * 8), 0xc0000000 | base->addr >> 12);
-	nv_wr32(priv, 0x003004 + (chid * 8), 0x001f0001);
-	nvc0_fifo_playlist_update(priv);
+
+	if (chan->state == STOPPED && (chan->state = RUNNING) == RUNNING) {
+		nv_wr32(priv, 0x003004 + (chid * 8), 0x001f0001);
+		nvc0_fifo_runlist_update(priv);
+	}
+
 	return 0;
 }
 
+static void nvc0_fifo_intr_engine(struct nvc0_fifo_priv *priv);
+
 static int
 nvc0_fifo_chan_fini(struct nouveau_object *object, bool suspend)
 {
 	struct nvc0_fifo_priv *priv = (void *)object->engine;
 	struct nvc0_fifo_chan *chan = (void *)object;
 	u32 chid = chan->base.chid;
-	u32 mask, engine;
 
-	nv_mask(priv, 0x003004 + (chid * 8), 0x00000001, 0x00000000);
-	nvc0_fifo_playlist_update(priv);
-	mask = nv_rd32(priv, 0x0025a4);
-	for (engine = 0; mask && engine < 16; engine++) {
-		if (!(mask & (1 << engine)))
-			continue;
-		nv_mask(priv, 0x0025a8 + (engine * 4), 0x00000000, 0x00000000);
-		mask &= ~(1 << engine);
+	if (chan->state == RUNNING && (chan->state = STOPPED) == STOPPED) {
+		nv_mask(priv, 0x003004 + (chid * 8), 0x00000001, 0x00000000);
+		nvc0_fifo_runlist_update(priv);
 	}
-	nv_wr32(priv, 0x003000 + (chid * 8), 0x00000000);
 
+	nvc0_fifo_intr_engine(priv);
+
+	nv_wr32(priv, 0x003000 + (chid * 8), 0x00000000);
 	return nouveau_fifo_channel_fini(&chan->base, suspend);
 }
 
@@ -344,11 +364,177 @@ nvc0_fifo_cclass = {
  * PFIFO engine
  ******************************************************************************/
 
-static const struct nouveau_enum nvc0_fifo_fault_unit[] = {
+static inline int
+nvc0_fifo_engidx(struct nvc0_fifo_priv *priv, u32 engn)
+{
+	switch (engn) {
+	case NVDEV_ENGINE_GR   : engn = 0; break;
+	case NVDEV_ENGINE_BSP  : engn = 1; break;
+	case NVDEV_ENGINE_PPP  : engn = 2; break;
+	case NVDEV_ENGINE_VP   : engn = 3; break;
+	case NVDEV_ENGINE_COPY0: engn = 4; break;
+	case NVDEV_ENGINE_COPY1: engn = 5; break;
+	default:
+		return -1;
+	}
+
+	return engn;
+}
+
+static inline struct nouveau_engine *
+nvc0_fifo_engine(struct nvc0_fifo_priv *priv, u32 engn)
+{
+	switch (engn) {
+	case 0: engn = NVDEV_ENGINE_GR; break;
+	case 1: engn = NVDEV_ENGINE_BSP; break;
+	case 2: engn = NVDEV_ENGINE_PPP; break;
+	case 3: engn = NVDEV_ENGINE_VP; break;
+	case 4: engn = NVDEV_ENGINE_COPY0; break;
+	case 5: engn = NVDEV_ENGINE_COPY1; break;
+	default:
+		return NULL;
+	}
+
+	return nouveau_engine(priv, engn);
+}
+
+static void
+nvc0_fifo_recover_work(struct work_struct *work)
+{
+	struct nvc0_fifo_priv *priv = container_of(work, typeof(*priv), fault);
+	struct nouveau_object *engine;
+	unsigned long flags;
+	u32 engn, engm = 0;
+	u64 mask, todo;
+
+	spin_lock_irqsave(&priv->base.lock, flags);
+	mask = priv->mask;
+	priv->mask = 0ULL;
+	spin_unlock_irqrestore(&priv->base.lock, flags);
+
+	for (todo = mask; engn = __ffs64(todo), todo; todo &= ~(1 << engn))
+		engm |= 1 << nvc0_fifo_engidx(priv, engn);
+	nv_mask(priv, 0x002630, engm, engm);
+
+	for (todo = mask; engn = __ffs64(todo), todo; todo &= ~(1 << engn)) {
+		if ((engine = (void *)nouveau_engine(priv, engn))) {
+			nv_ofuncs(engine)->fini(engine, false);
+			WARN_ON(nv_ofuncs(engine)->init(engine));
+		}
+	}
+
+	nvc0_fifo_runlist_update(priv);
+	nv_wr32(priv, 0x00262c, engm);
+	nv_mask(priv, 0x002630, engm, 0x00000000);
+}
+
+static void
+nvc0_fifo_recover(struct nvc0_fifo_priv *priv, struct nouveau_engine *engine,
+		  struct nvc0_fifo_chan *chan)
+{
+	struct nouveau_object *engobj = nv_object(engine);
+	u32 chid = chan->base.chid;
+	unsigned long flags;
+
+	nv_error(priv, "%s engine fault on channel %d, recovering...\n",
+		       nv_subdev(engine)->name, chid);
+
+	nv_mask(priv, 0x003004 + (chid * 0x08), 0x00000001, 0x00000000);
+	chan->state = KILLED;
+
+	spin_lock_irqsave(&priv->base.lock, flags);
+	priv->mask |= 1ULL << nv_engidx(engobj);
+	spin_unlock_irqrestore(&priv->base.lock, flags);
+	schedule_work(&priv->fault);
+}
+
+static int
+nvc0_fifo_swmthd(struct nvc0_fifo_priv *priv, u32 chid, u32 mthd, u32 data)
+{
+	struct nvc0_fifo_chan *chan = NULL;
+	struct nouveau_handle *bind;
+	unsigned long flags;
+	int ret = -EINVAL;
+
+	spin_lock_irqsave(&priv->base.lock, flags);
+	if (likely(chid >= priv->base.min && chid <= priv->base.max))
+		chan = (void *)priv->base.channel[chid];
+	if (unlikely(!chan))
+		goto out;
+
+	bind = nouveau_namedb_get_class(nv_namedb(chan), 0x906e);
+	if (likely(bind)) {
+		if (!mthd || !nv_call(bind->object, mthd, data))
+			ret = 0;
+		nouveau_namedb_put(bind);
+	}
+
+out:
+	spin_unlock_irqrestore(&priv->base.lock, flags);
+	return ret;
+}
+
+static const struct nouveau_enum
+nvc0_fifo_sched_reason[] = {
+	{ 0x0a, "CTXSW_TIMEOUT" },
+	{}
+};
+
+static void
+nvc0_fifo_intr_sched_ctxsw(struct nvc0_fifo_priv *priv)
+{
+	struct nouveau_engine *engine;
+	struct nvc0_fifo_chan *chan;
+	u32 engn;
+
+	for (engn = 0; engn < 6; engn++) {
+		u32 stat = nv_rd32(priv, 0x002640 + (engn * 0x04));
+		u32 busy = (stat & 0x80000000);
+		u32 save = (stat & 0x00100000); /* maybe? */
+		u32 unk0 = (stat & 0x00040000);
+		u32 unk1 = (stat & 0x00001000);
+		u32 chid = (stat & 0x0000007f);
+		(void)save;
+
+		if (busy && unk0 && unk1) {
+			if (!(chan = (void *)priv->base.channel[chid]))
+				continue;
+			if (!(engine = nvc0_fifo_engine(priv, engn)))
+				continue;
+			nvc0_fifo_recover(priv, engine, chan);
+		}
+	}
+}
+
+static void
+nvc0_fifo_intr_sched(struct nvc0_fifo_priv *priv)
+{
+	u32 intr = nv_rd32(priv, 0x00254c);
+	u32 code = intr & 0x000000ff;
+	const struct nouveau_enum *en;
+	char enunk[6] = "";
+
+	en = nouveau_enum_find(nvc0_fifo_sched_reason, code);
+	if (!en)
+		snprintf(enunk, sizeof(enunk), "UNK%02x", code);
+
+	nv_error(priv, "SCHED_ERROR [ %s ]\n", en ? en->name : enunk);
+
+	switch (code) {
+	case 0x0a:
+		nvc0_fifo_intr_sched_ctxsw(priv);
+		break;
+	default:
+		break;
+	}
+}
+
+static const struct nouveau_enum
+nvc0_fifo_fault_engine[] = {
 	{ 0x00, "PGRAPH", NULL, NVDEV_ENGINE_GR },
-	{ 0x03, "PEEPHOLE" },
-	{ 0x04, "BAR1" },
-	{ 0x05, "BAR3" },
+	{ 0x03, "PEEPHOLE", NULL, NVDEV_ENGINE_IFB },
+	{ 0x04, "BAR1", NULL, NVDEV_SUBDEV_BAR },
+	{ 0x05, "BAR3", NULL, NVDEV_SUBDEV_INSTMEM },
 	{ 0x07, "PFIFO", NULL, NVDEV_ENGINE_FIFO },
 	{ 0x10, "PBSP", NULL, NVDEV_ENGINE_BSP },
 	{ 0x11, "PPPP", NULL, NVDEV_ENGINE_PPP },
@@ -360,7 +546,8 @@ static const struct nouveau_enum nvc0_fifo_fault_unit[] = {
 	{}
 };
 
-static const struct nouveau_enum nvc0_fifo_fault_reason[] = {
+static const struct nouveau_enum
+nvc0_fifo_fault_reason[] = {
 	{ 0x00, "PT_NOT_PRESENT" },
 	{ 0x01, "PT_TOO_SHORT" },
 	{ 0x02, "PAGE_NOT_PRESENT" },
@@ -373,7 +560,8 @@ static const struct nouveau_enum nvc0_fifo_fault_reason[] = {
 	{}
 };
 
-static const struct nouveau_enum nvc0_fifo_fault_hubclient[] = {
+static const struct nouveau_enum
+nvc0_fifo_fault_hubclient[] = {
 	{ 0x01, "PCOPY0" },
 	{ 0x02, "PCOPY1" },
 	{ 0x04, "DISPATCH" },
@@ -391,7 +579,8 @@ static const struct nouveau_enum nvc0_fifo_fault_hubclient[] = {
 	{}
 };
 
-static const struct nouveau_enum nvc0_fifo_fault_gpcclient[] = {
+static const struct nouveau_enum
+nvc0_fifo_fault_gpcclient[] = {
 	{ 0x01, "TEX" },
 	{ 0x0c, "ESETUP" },
 	{ 0x0e, "CTXCTL" },
@@ -399,92 +588,92 @@ static const struct nouveau_enum nvc0_fifo_fault_gpcclient[] = {
 	{}
 };
 
-static const struct nouveau_bitfield nvc0_fifo_subfifo_intr[] = {
-/*	{ 0x00008000, "" }	seen with null ib push */
-	{ 0x00200000, "ILLEGAL_MTHD" },
-	{ 0x00800000, "EMPTY_SUBC" },
-	{}
-};
-
 static void
-nvc0_fifo_isr_vm_fault(struct nvc0_fifo_priv *priv, int unit)
+nvc0_fifo_intr_fault(struct nvc0_fifo_priv *priv, int unit)
 {
 	u32 inst = nv_rd32(priv, 0x002800 + (unit * 0x10));
 	u32 valo = nv_rd32(priv, 0x002804 + (unit * 0x10));
 	u32 vahi = nv_rd32(priv, 0x002808 + (unit * 0x10));
 	u32 stat = nv_rd32(priv, 0x00280c + (unit * 0x10));
+	u32 gpc    = (stat & 0x1f000000) >> 24;
 	u32 client = (stat & 0x00001f00) >> 8;
-	const struct nouveau_enum *en;
-	struct nouveau_engine *engine;
-	struct nouveau_object *engctx = NULL;
-
-	switch (unit) {
-	case 3: /* PEEPHOLE */
-		nv_mask(priv, 0x001718, 0x00000000, 0x00000000);
-		break;
-	case 4: /* BAR1 */
-		nv_mask(priv, 0x001704, 0x00000000, 0x00000000);
-		break;
-	case 5: /* BAR3 */
-		nv_mask(priv, 0x001714, 0x00000000, 0x00000000);
-		break;
-	default:
-		break;
+	u32 write  = (stat & 0x00000080);
+	u32 hub    = (stat & 0x00000040);
+	u32 reason = (stat & 0x0000000f);
+	struct nouveau_object *engctx = NULL, *object;
+	struct nouveau_engine *engine = NULL;
+	const struct nouveau_enum *er, *eu, *ec;
+	char erunk[6] = "";
+	char euunk[6] = "";
+	char ecunk[6] = "";
+	char gpcid[3] = "";
+
+	er = nouveau_enum_find(nvc0_fifo_fault_reason, reason);
+	if (!er)
+		snprintf(erunk, sizeof(erunk), "UNK%02X", reason);
+
+	eu = nouveau_enum_find(nvc0_fifo_fault_engine, unit);
+	if (eu) {
+		switch (eu->data2) {
+		case NVDEV_SUBDEV_BAR:
+			nv_mask(priv, 0x001704, 0x00000000, 0x00000000);
+			break;
+		case NVDEV_SUBDEV_INSTMEM:
+			nv_mask(priv, 0x001714, 0x00000000, 0x00000000);
+			break;
+		case NVDEV_ENGINE_IFB:
+			nv_mask(priv, 0x001718, 0x00000000, 0x00000000);
+			break;
+		default:
+			engine = nouveau_engine(priv, eu->data2);
+			if (engine)
+				engctx = nouveau_engctx_get(engine, inst);
+			break;
+		}
+	} else {
+		snprintf(euunk, sizeof(euunk), "UNK%02x", unit);
 	}
 
-	nv_error(priv, "%s fault at 0x%010llx [", (stat & 0x00000080) ?
-		 "write" : "read", (u64)vahi << 32 | valo);
-	nouveau_enum_print(nvc0_fifo_fault_reason, stat & 0x0000000f);
-	pr_cont("] from ");
-	en = nouveau_enum_print(nvc0_fifo_fault_unit, unit);
-	if (stat & 0x00000040) {
-		pr_cont("/");
-		nouveau_enum_print(nvc0_fifo_fault_hubclient, client);
+	if (hub) {
+		ec = nouveau_enum_find(nvc0_fifo_fault_hubclient, client);
 	} else {
-		pr_cont("/GPC%d/", (stat & 0x1f000000) >> 24);
-		nouveau_enum_print(nvc0_fifo_fault_gpcclient, client);
+		ec = nouveau_enum_find(nvc0_fifo_fault_gpcclient, client);
+		snprintf(gpcid, sizeof(gpcid), "%d", gpc);
 	}
 
-	if (en && en->data2) {
-		engine = nouveau_engine(priv, en->data2);
-		if (engine)
-			engctx = nouveau_engctx_get(engine, inst);
-
+	if (!ec)
+		snprintf(ecunk, sizeof(ecunk), "UNK%02x", client);
+
+	nv_error(priv, "%s fault at 0x%010llx [%s] from %s/%s%s%s%s on "
+		       "channel 0x%010llx [%s]\n", write ? "write" : "read",
+		 (u64)vahi << 32 | valo, er ? er->name : erunk,
+		 eu ? eu->name : euunk, hub ? "" : "GPC", gpcid, hub ? "" : "/",
+		 ec ? ec->name : ecunk, (u64)inst << 12,
+		 nouveau_client_name(engctx));
+
+	object = engctx;
+	while (object) {
+		switch (nv_mclass(object)) {
+		case NVC0_CHANNEL_IND_CLASS:
+			nvc0_fifo_recover(priv, engine, (void *)object);
+			break;
+		}
+		object = object->parent;
 	}
-	pr_cont(" on channel 0x%010llx [%s]\n", (u64)inst << 12,
-			nouveau_client_name(engctx));
 
 	nouveau_engctx_put(engctx);
 }
 
-static int
-nvc0_fifo_swmthd(struct nvc0_fifo_priv *priv, u32 chid, u32 mthd, u32 data)
-{
-	struct nvc0_fifo_chan *chan = NULL;
-	struct nouveau_handle *bind;
-	unsigned long flags;
-	int ret = -EINVAL;
-
-	spin_lock_irqsave(&priv->base.lock, flags);
-	if (likely(chid >= priv->base.min && chid <= priv->base.max))
-		chan = (void *)priv->base.channel[chid];
-	if (unlikely(!chan))
-		goto out;
-
-	bind = nouveau_namedb_get_class(nv_namedb(chan), 0x906e);
-	if (likely(bind)) {
-		if (!mthd || !nv_call(bind->object, mthd, data))
-			ret = 0;
-		nouveau_namedb_put(bind);
-	}
-
-out:
-	spin_unlock_irqrestore(&priv->base.lock, flags);
-	return ret;
-}
+static const struct nouveau_bitfield
+nvc0_fifo_pbdma_intr[] = {
+/*	{ 0x00008000, "" }	seen with null ib push */
+	{ 0x00200000, "ILLEGAL_MTHD" },
+	{ 0x00800000, "EMPTY_SUBC" },
+	{}
+};
 
 static void
-nvc0_fifo_isr_subfifo_intr(struct nvc0_fifo_priv *priv, int unit)
+nvc0_fifo_intr_pbdma(struct nvc0_fifo_priv *priv, int unit)
 {
 	u32 stat = nv_rd32(priv, 0x040108 + (unit * 0x2000));
 	u32 addr = nv_rd32(priv, 0x0400c0 + (unit * 0x2000));
@@ -500,11 +689,11 @@ nvc0_fifo_isr_subfifo_intr(struct nvc0_fifo_priv *priv, int unit)
 	}
 
 	if (show) {
-		nv_error(priv, "SUBFIFO%d:", unit);
-		nouveau_bitfield_print(nvc0_fifo_subfifo_intr, show);
+		nv_error(priv, "PBDMA%d:", unit);
+		nouveau_bitfield_print(nvc0_fifo_pbdma_intr, show);
 		pr_cont("\n");
 		nv_error(priv,
-			 "SUBFIFO%d: ch %d [%s] subc %d mthd 0x%04x data 0x%08x\n",
+			 "PBDMA%d: ch %d [%s] subc %d mthd 0x%04x data 0x%08x\n",
 			 unit, chid,
 			 nouveau_client_name_for_fifo_chid(&priv->base, chid),
 			 subc, mthd, data);
@@ -515,6 +704,56 @@ nvc0_fifo_isr_subfifo_intr(struct nvc0_fifo_priv *priv, int unit)
 }
 
 static void
+nvc0_fifo_intr_runlist(struct nvc0_fifo_priv *priv)
+{
+	u32 intr = nv_rd32(priv, 0x002a00);
+
+	if (intr & 0x10000000) {
+		wake_up(&priv->runlist.wait);
+		nv_wr32(priv, 0x002a00, 0x10000000);
+		intr &= ~0x10000000;
+	}
+
+	if (intr) {
+		nv_error(priv, "RUNLIST 0x%08x\n", intr);
+		nv_wr32(priv, 0x002a00, intr);
+	}
+}
+
+static void
+nvc0_fifo_intr_engine_unit(struct nvc0_fifo_priv *priv, int engn)
+{
+	u32 intr = nv_rd32(priv, 0x0025a8 + (engn * 0x04));
+	u32 inte = nv_rd32(priv, 0x002628);
+	u32 unkn;
+
+	for (unkn = 0; unkn < 8; unkn++) {
+		u32 ints = (intr >> (unkn * 0x04)) & inte;
+		if (ints & 0x1) {
+			nouveau_event_trigger(priv->base.uevent, 1, 0);
+			ints &= ~1;
+		}
+		if (ints) {
+			nv_error(priv, "ENGINE %d %d %01x", engn, unkn, ints);
+			nv_mask(priv, 0x002628, ints, 0);
+		}
+	}
+
+	nv_wr32(priv, 0x0025a8 + (engn * 0x04), intr);
+}
+
+static void
+nvc0_fifo_intr_engine(struct nvc0_fifo_priv *priv)
+{
+	u32 mask = nv_rd32(priv, 0x0025a4);
+	while (mask) {
+		u32 unit = __ffs(mask);
+		nvc0_fifo_intr_engine_unit(priv, unit);
+		mask &= ~(1 << unit);
+	}
+}
+
+static void
 nvc0_fifo_intr(struct nouveau_subdev *subdev)
 {
 	struct nvc0_fifo_priv *priv = (void *)subdev;
@@ -529,8 +768,7 @@ nvc0_fifo_intr(struct nouveau_subdev *subdev)
 	}
 
 	if (stat & 0x00000100) {
-		u32 intr = nv_rd32(priv, 0x00254c);
-		nv_warn(priv, "INTR 0x00000100: 0x%08x\n", intr);
+		nvc0_fifo_intr_sched(priv);
 		nv_wr32(priv, 0x002100, 0x00000100);
 		stat &= ~0x00000100;
 	}
@@ -550,64 +788,53 @@ nvc0_fifo_intr(struct nouveau_subdev *subdev)
 	}
 
 	if (stat & 0x10000000) {
-		u32 units = nv_rd32(priv, 0x00259c);
-		u32 u = units;
-
-		while (u) {
-			int i = ffs(u) - 1;
-			nvc0_fifo_isr_vm_fault(priv, i);
-			u &= ~(1 << i);
+		u32 mask = nv_rd32(priv, 0x00259c);
+		while (mask) {
+			u32 unit = __ffs(mask);
+			nvc0_fifo_intr_fault(priv, unit);
+			nv_wr32(priv, 0x00259c, (1 << unit));
+			mask &= ~(1 << unit);
 		}
-
-		nv_wr32(priv, 0x00259c, units);
 		stat &= ~0x10000000;
 	}
 
 	if (stat & 0x20000000) {
-		u32 units = nv_rd32(priv, 0x0025a0);
-		u32 u = units;
-
-		while (u) {
-			int i = ffs(u) - 1;
-			nvc0_fifo_isr_subfifo_intr(priv, i);
-			u &= ~(1 << i);
+		u32 mask = nv_rd32(priv, 0x0025a0);
+		while (mask) {
+			u32 unit = __ffs(mask);
+			nvc0_fifo_intr_pbdma(priv, unit);
+			nv_wr32(priv, 0x0025a0, (1 << unit));
+			mask &= ~(1 << unit);
 		}
-
-		nv_wr32(priv, 0x0025a0, units);
 		stat &= ~0x20000000;
 	}
 
 	if (stat & 0x40000000) {
-		u32 intr0 = nv_rd32(priv, 0x0025a4);
-		u32 intr1 = nv_mask(priv, 0x002a00, 0x00000000, 0x00000);
-		nv_debug(priv, "INTR 0x40000000: 0x%08x 0x%08x\n",
-			       intr0, intr1);
+		nvc0_fifo_intr_runlist(priv);
 		stat &= ~0x40000000;
 	}
 
 	if (stat & 0x80000000) {
-		u32 intr = nv_mask(priv, 0x0025a8, 0x00000000, 0x00000000);
-		nouveau_event_trigger(priv->base.uevent, 0);
-		nv_debug(priv, "INTR 0x80000000: 0x%08x\n", intr);
+		nvc0_fifo_intr_engine(priv);
 		stat &= ~0x80000000;
 	}
 
 	if (stat) {
-		nv_fatal(priv, "unhandled status 0x%08x\n", stat);
+		nv_error(priv, "INTR 0x%08x\n", stat);
+		nv_mask(priv, 0x002140, stat, 0x00000000);
 		nv_wr32(priv, 0x002100, stat);
-		nv_wr32(priv, 0x002140, 0);
 	}
 }
 
 static void
-nvc0_fifo_uevent_enable(struct nouveau_event *event, int index)
+nvc0_fifo_uevent_enable(struct nouveau_event *event, int type, int index)
 {
 	struct nvc0_fifo_priv *priv = event->priv;
 	nv_mask(priv, 0x002140, 0x80000000, 0x80000000);
 }
 
 static void
-nvc0_fifo_uevent_disable(struct nouveau_event *event, int index)
+nvc0_fifo_uevent_disable(struct nouveau_event *event, int type, int index)
 {
 	struct nvc0_fifo_priv *priv = event->priv;
 	nv_mask(priv, 0x002140, 0x80000000, 0x00000000);
@@ -626,16 +853,20 @@ nvc0_fifo_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	INIT_WORK(&priv->fault, nvc0_fifo_recover_work);
+
 	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x1000, 0x1000, 0,
-				&priv->playlist[0]);
+				&priv->runlist.mem[0]);
 	if (ret)
 		return ret;
 
 	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x1000, 0x1000, 0,
-				&priv->playlist[1]);
+				&priv->runlist.mem[1]);
 	if (ret)
 		return ret;
 
+	init_waitqueue_head(&priv->runlist.wait);
+
 	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 128 * 0x1000, 0x1000, 0,
 				&priv->user.mem);
 	if (ret)
@@ -664,8 +895,8 @@ nvc0_fifo_dtor(struct nouveau_object *object)
 
 	nouveau_gpuobj_unmap(&priv->user.bar);
 	nouveau_gpuobj_ref(NULL, &priv->user.mem);
-	nouveau_gpuobj_ref(NULL, &priv->playlist[1]);
-	nouveau_gpuobj_ref(NULL, &priv->playlist[0]);
+	nouveau_gpuobj_ref(NULL, &priv->runlist.mem[0]);
+	nouveau_gpuobj_ref(NULL, &priv->runlist.mem[1]);
 
 	nouveau_fifo_destroy(&priv->base);
 }
@@ -684,9 +915,9 @@ nvc0_fifo_init(struct nouveau_object *object)
 	nv_wr32(priv, 0x002204, 0xffffffff);
 
 	priv->spoon_nr = hweight32(nv_rd32(priv, 0x002204));
-	nv_debug(priv, "%d subfifo(s)\n", priv->spoon_nr);
+	nv_debug(priv, "%d PBDMA unit(s)\n", priv->spoon_nr);
 
-	/* assign engines to subfifos */
+	/* assign engines to PBDMAs */
 	if (priv->spoon_nr >= 3) {
 		nv_wr32(priv, 0x002208, ~(1 << 0)); /* PGRAPH */
 		nv_wr32(priv, 0x00220c, ~(1 << 1)); /* PVP */
@@ -696,7 +927,7 @@ nvc0_fifo_init(struct nouveau_object *object)
 		nv_wr32(priv, 0x00221c, ~(1 << 1)); /* PCE1 */
 	}
 
-	/* PSUBFIFO[n] */
+	/* PBDMA[n] */
 	for (i = 0; i < priv->spoon_nr; i++) {
 		nv_mask(priv, 0x04013c + (i * 0x2000), 0x10000100, 0x00000000);
 		nv_wr32(priv, 0x040108 + (i * 0x2000), 0xffffffff); /* INTR */
@@ -706,10 +937,9 @@ nvc0_fifo_init(struct nouveau_object *object)
 	nv_mask(priv, 0x002200, 0x00000001, 0x00000001);
 	nv_wr32(priv, 0x002254, 0x10000000 | priv->user.bar.offset >> 12);
 
-	nv_wr32(priv, 0x002a00, 0xffffffff); /* clears PFIFO.INTR bit 30 */
 	nv_wr32(priv, 0x002100, 0xffffffff);
-	nv_wr32(priv, 0x002140, 0x3fffffff);
-	nv_wr32(priv, 0x002628, 0x00000001); /* makes mthd 0x20 work */
+	nv_wr32(priv, 0x002140, 0x7fffffff);
+	nv_wr32(priv, 0x002628, 0x00000001); /* ENGINE_INTR_EN */
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c
index 04f412922d2..298063edb92 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c
@@ -33,10 +33,12 @@
 
 #include <subdev/timer.h>
 #include <subdev/bar.h>
+#include <subdev/fb.h>
 #include <subdev/vm.h>
 
 #include <engine/dmaobj.h>
-#include <engine/fifo.h>
+
+#include "nve0.h"
 
 #define _(a,b) { (a), ((1ULL << (a)) | (b)) }
 static const struct {
@@ -56,12 +58,17 @@ static const struct {
 #define FIFO_ENGINE_NR ARRAY_SIZE(fifo_engine)
 
 struct nve0_fifo_engn {
-	struct nouveau_gpuobj *playlist[2];
-	int cur_playlist;
+	struct nouveau_gpuobj *runlist[2];
+	int cur_runlist;
+	wait_queue_head_t wait;
 };
 
 struct nve0_fifo_priv {
 	struct nouveau_fifo base;
+
+	struct work_struct fault;
+	u64 mask;
+
 	struct nve0_fifo_engn engine[FIFO_ENGINE_NR];
 	struct {
 		struct nouveau_gpuobj *mem;
@@ -79,6 +86,11 @@ struct nve0_fifo_base {
 struct nve0_fifo_chan {
 	struct nouveau_fifo_chan base;
 	u32 engine;
+	enum {
+		STOPPED,
+		RUNNING,
+		KILLED
+	} state;
 };
 
 /*******************************************************************************
@@ -86,32 +98,34 @@ struct nve0_fifo_chan {
  ******************************************************************************/
 
 static void
-nve0_fifo_playlist_update(struct nve0_fifo_priv *priv, u32 engine)
+nve0_fifo_runlist_update(struct nve0_fifo_priv *priv, u32 engine)
 {
 	struct nouveau_bar *bar = nouveau_bar(priv);
 	struct nve0_fifo_engn *engn = &priv->engine[engine];
 	struct nouveau_gpuobj *cur;
-	u32 match = (engine << 16) | 0x00000001;
 	int i, p;
 
 	mutex_lock(&nv_subdev(priv)->mutex);
-	cur = engn->playlist[engn->cur_playlist];
-	engn->cur_playlist = !engn->cur_playlist;
+	cur = engn->runlist[engn->cur_runlist];
+	engn->cur_runlist = !engn->cur_runlist;
 
 	for (i = 0, p = 0; i < priv->base.max; i++) {
-		u32 ctrl = nv_rd32(priv, 0x800004 + (i * 8)) & 0x001f0001;
-		if (ctrl != match)
-			continue;
-		nv_wo32(cur, p + 0, i);
-		nv_wo32(cur, p + 4, 0x00000000);
-		p += 8;
+		struct nve0_fifo_chan *chan = (void *)priv->base.channel[i];
+		if (chan && chan->state == RUNNING && chan->engine == engine) {
+			nv_wo32(cur, p + 0, i);
+			nv_wo32(cur, p + 4, 0x00000000);
+			p += 8;
+		}
 	}
 	bar->flush(bar);
 
 	nv_wr32(priv, 0x002270, cur->addr >> 12);
 	nv_wr32(priv, 0x002274, (engine << 20) | (p >> 3));
-	if (!nv_wait(priv, 0x002284 + (engine * 4), 0x00100000, 0x00000000))
-		nv_error(priv, "playlist %d update timeout\n", engine);
+
+	if (wait_event_timeout(engn->wait, !(nv_rd32(priv, 0x002284 +
+			       (engine * 0x08)) & 0x00100000),
+				msecs_to_jiffies(2000)) == 0)
+		nv_error(priv, "runlist %d update timeout\n", engine);
 	mutex_unlock(&nv_subdev(priv)->mutex);
 }
 
@@ -127,9 +141,11 @@ nve0_fifo_context_attach(struct nouveau_object *parent,
 
 	switch (nv_engidx(object->engine)) {
 	case NVDEV_ENGINE_SW   :
+		return 0;
 	case NVDEV_ENGINE_COPY0:
 	case NVDEV_ENGINE_COPY1:
 	case NVDEV_ENGINE_COPY2:
+		nv_engctx(ectx)->addr = nv_gpuobj(base)->addr >> 12;
 		return 0;
 	case NVDEV_ENGINE_GR   : addr = 0x0210; break;
 	case NVDEV_ENGINE_BSP  : addr = 0x0270; break;
@@ -277,9 +293,13 @@ nve0_fifo_chan_init(struct nouveau_object *object)
 
 	nv_mask(priv, 0x800004 + (chid * 8), 0x000f0000, chan->engine << 16);
 	nv_wr32(priv, 0x800000 + (chid * 8), 0x80000000 | base->addr >> 12);
-	nv_mask(priv, 0x800004 + (chid * 8), 0x00000400, 0x00000400);
-	nve0_fifo_playlist_update(priv, chan->engine);
-	nv_mask(priv, 0x800004 + (chid * 8), 0x00000400, 0x00000400);
+
+	if (chan->state == STOPPED && (chan->state = RUNNING) == RUNNING) {
+		nv_mask(priv, 0x800004 + (chid * 8), 0x00000400, 0x00000400);
+		nve0_fifo_runlist_update(priv, chan->engine);
+		nv_mask(priv, 0x800004 + (chid * 8), 0x00000400, 0x00000400);
+	}
+
 	return 0;
 }
 
@@ -290,10 +310,12 @@ nve0_fifo_chan_fini(struct nouveau_object *object, bool suspend)
 	struct nve0_fifo_chan *chan = (void *)object;
 	u32 chid = chan->base.chid;
 
-	nv_mask(priv, 0x800004 + (chid * 8), 0x00000800, 0x00000800);
-	nve0_fifo_playlist_update(priv, chan->engine);
-	nv_wr32(priv, 0x800000 + (chid * 8), 0x00000000);
+	if (chan->state == RUNNING && (chan->state = STOPPED) == STOPPED) {
+		nv_mask(priv, 0x800004 + (chid * 8), 0x00000800, 0x00000800);
+		nve0_fifo_runlist_update(priv, chan->engine);
+	}
 
+	nv_wr32(priv, 0x800000 + (chid * 8), 0x00000000);
 	return nouveau_fifo_channel_fini(&chan->base, suspend);
 }
 
@@ -375,73 +397,81 @@ nve0_fifo_cclass = {
  * PFIFO engine
  ******************************************************************************/
 
-static const struct nouveau_enum nve0_fifo_fault_unit[] = {
-	{}
-};
-
-static const struct nouveau_enum nve0_fifo_fault_reason[] = {
-	{ 0x00, "PT_NOT_PRESENT" },
-	{ 0x01, "PT_TOO_SHORT" },
-	{ 0x02, "PAGE_NOT_PRESENT" },
-	{ 0x03, "VM_LIMIT_EXCEEDED" },
-	{ 0x04, "NO_CHANNEL" },
-	{ 0x05, "PAGE_SYSTEM_ONLY" },
-	{ 0x06, "PAGE_READ_ONLY" },
-	{ 0x0a, "COMPRESSED_SYSRAM" },
-	{ 0x0c, "INVALID_STORAGE_TYPE" },
-	{}
-};
-
-static const struct nouveau_enum nve0_fifo_fault_hubclient[] = {
-	{}
-};
+static inline int
+nve0_fifo_engidx(struct nve0_fifo_priv *priv, u32 engn)
+{
+	switch (engn) {
+	case NVDEV_ENGINE_GR   :
+	case NVDEV_ENGINE_COPY2: engn = 0; break;
+	case NVDEV_ENGINE_BSP  : engn = 1; break;
+	case NVDEV_ENGINE_PPP  : engn = 2; break;
+	case NVDEV_ENGINE_VP   : engn = 3; break;
+	case NVDEV_ENGINE_COPY0: engn = 4; break;
+	case NVDEV_ENGINE_COPY1: engn = 5; break;
+	case NVDEV_ENGINE_VENC : engn = 6; break;
+	default:
+		return -1;
+	}
 
-static const struct nouveau_enum nve0_fifo_fault_gpcclient[] = {
-	{}
-};
+	return engn;
+}
 
-static const struct nouveau_bitfield nve0_fifo_subfifo_intr[] = {
-	{ 0x00200000, "ILLEGAL_MTHD" },
-	{ 0x00800000, "EMPTY_SUBC" },
-	{}
-};
+static inline struct nouveau_engine *
+nve0_fifo_engine(struct nve0_fifo_priv *priv, u32 engn)
+{
+	if (engn >= ARRAY_SIZE(fifo_engine))
+		return NULL;
+	return nouveau_engine(priv, fifo_engine[engn].subdev);
+}
 
 static void
-nve0_fifo_isr_vm_fault(struct nve0_fifo_priv *priv, int unit)
+nve0_fifo_recover_work(struct work_struct *work)
 {
-	u32 inst = nv_rd32(priv, 0x2800 + (unit * 0x10));
-	u32 valo = nv_rd32(priv, 0x2804 + (unit * 0x10));
-	u32 vahi = nv_rd32(priv, 0x2808 + (unit * 0x10));
-	u32 stat = nv_rd32(priv, 0x280c + (unit * 0x10));
-	u32 client = (stat & 0x00001f00) >> 8;
-	const struct nouveau_enum *en;
-	struct nouveau_engine *engine;
-	struct nouveau_object *engctx = NULL;
-
-	nv_error(priv, "PFIFO: %s fault at 0x%010llx [", (stat & 0x00000080) ?
-		       "write" : "read", (u64)vahi << 32 | valo);
-	nouveau_enum_print(nve0_fifo_fault_reason, stat & 0x0000000f);
-	pr_cont("] from ");
-	en = nouveau_enum_print(nve0_fifo_fault_unit, unit);
-	if (stat & 0x00000040) {
-		pr_cont("/");
-		nouveau_enum_print(nve0_fifo_fault_hubclient, client);
-	} else {
-		pr_cont("/GPC%d/", (stat & 0x1f000000) >> 24);
-		nouveau_enum_print(nve0_fifo_fault_gpcclient, client);
-	}
+	struct nve0_fifo_priv *priv = container_of(work, typeof(*priv), fault);
+	struct nouveau_object *engine;
+	unsigned long flags;
+	u32 engn, engm = 0;
+	u64 mask, todo;
 
-	if (en && en->data2) {
-		engine = nouveau_engine(priv, en->data2);
-		if (engine)
-			engctx = nouveau_engctx_get(engine, inst);
+	spin_lock_irqsave(&priv->base.lock, flags);
+	mask = priv->mask;
+	priv->mask = 0ULL;
+	spin_unlock_irqrestore(&priv->base.lock, flags);
 
+	for (todo = mask; engn = __ffs64(todo), todo; todo &= ~(1 << engn))
+		engm |= 1 << nve0_fifo_engidx(priv, engn);
+	nv_mask(priv, 0x002630, engm, engm);
+
+	for (todo = mask; engn = __ffs64(todo), todo; todo &= ~(1 << engn)) {
+		if ((engine = (void *)nouveau_engine(priv, engn))) {
+			nv_ofuncs(engine)->fini(engine, false);
+			WARN_ON(nv_ofuncs(engine)->init(engine));
+		}
+		nve0_fifo_runlist_update(priv, nve0_fifo_engidx(priv, engn));
 	}
 
-	pr_cont(" on channel 0x%010llx [%s]\n", (u64)inst << 12,
-			nouveau_client_name(engctx));
+	nv_wr32(priv, 0x00262c, engm);
+	nv_mask(priv, 0x002630, engm, 0x00000000);
+}
 
-	nouveau_engctx_put(engctx);
+static void
+nve0_fifo_recover(struct nve0_fifo_priv *priv, struct nouveau_engine *engine,
+		  struct nve0_fifo_chan *chan)
+{
+	struct nouveau_object *engobj = nv_object(engine);
+	u32 chid = chan->base.chid;
+	unsigned long flags;
+
+	nv_error(priv, "%s engine fault on channel %d, recovering...\n",
+		       nv_subdev(engine)->name, chid);
+
+	nv_mask(priv, 0x800004 + (chid * 0x08), 0x00000800, 0x00000800);
+	chan->state = KILLED;
+
+	spin_lock_irqsave(&priv->base.lock, flags);
+	priv->mask |= 1ULL << nv_engidx(engobj);
+	spin_unlock_irqrestore(&priv->base.lock, flags);
+	schedule_work(&priv->fault);
 }
 
 static int
@@ -470,8 +500,321 @@ out:
 	return ret;
 }
 
+static const struct nouveau_enum
+nve0_fifo_bind_reason[] = {
+	{ 0x01, "BIND_NOT_UNBOUND" },
+	{ 0x02, "SNOOP_WITHOUT_BAR1" },
+	{ 0x03, "UNBIND_WHILE_RUNNING" },
+	{ 0x05, "INVALID_RUNLIST" },
+	{ 0x06, "INVALID_CTX_TGT" },
+	{ 0x0b, "UNBIND_WHILE_PARKED" },
+	{}
+};
+
+static void
+nve0_fifo_intr_bind(struct nve0_fifo_priv *priv)
+{
+	u32 intr = nv_rd32(priv, 0x00252c);
+	u32 code = intr & 0x000000ff;
+	const struct nouveau_enum *en;
+	char enunk[6] = "";
+
+	en = nouveau_enum_find(nve0_fifo_bind_reason, code);
+	if (!en)
+		snprintf(enunk, sizeof(enunk), "UNK%02x", code);
+
+	nv_error(priv, "BIND_ERROR [ %s ]\n", en ? en->name : enunk);
+}
+
+static const struct nouveau_enum
+nve0_fifo_sched_reason[] = {
+	{ 0x0a, "CTXSW_TIMEOUT" },
+	{}
+};
+
+static void
+nve0_fifo_intr_sched_ctxsw(struct nve0_fifo_priv *priv)
+{
+	struct nouveau_engine *engine;
+	struct nve0_fifo_chan *chan;
+	u32 engn;
+
+	for (engn = 0; engn < ARRAY_SIZE(fifo_engine); engn++) {
+		u32 stat = nv_rd32(priv, 0x002640 + (engn * 0x04));
+		u32 busy = (stat & 0x80000000);
+		u32 next = (stat & 0x07ff0000) >> 16;
+		u32 chsw = (stat & 0x00008000);
+		u32 save = (stat & 0x00004000);
+		u32 load = (stat & 0x00002000);
+		u32 prev = (stat & 0x000007ff);
+		u32 chid = load ? next : prev;
+		(void)save;
+
+		if (busy && chsw) {
+			if (!(chan = (void *)priv->base.channel[chid]))
+				continue;
+			if (!(engine = nve0_fifo_engine(priv, engn)))
+				continue;
+			nve0_fifo_recover(priv, engine, chan);
+		}
+	}
+}
+
+static void
+nve0_fifo_intr_sched(struct nve0_fifo_priv *priv)
+{
+	u32 intr = nv_rd32(priv, 0x00254c);
+	u32 code = intr & 0x000000ff;
+	const struct nouveau_enum *en;
+	char enunk[6] = "";
+
+	en = nouveau_enum_find(nve0_fifo_sched_reason, code);
+	if (!en)
+		snprintf(enunk, sizeof(enunk), "UNK%02x", code);
+
+	nv_error(priv, "SCHED_ERROR [ %s ]\n", en ? en->name : enunk);
+
+	switch (code) {
+	case 0x0a:
+		nve0_fifo_intr_sched_ctxsw(priv);
+		break;
+	default:
+		break;
+	}
+}
+
+static void
+nve0_fifo_intr_chsw(struct nve0_fifo_priv *priv)
+{
+	u32 stat = nv_rd32(priv, 0x00256c);
+	nv_error(priv, "CHSW_ERROR 0x%08x\n", stat);
+	nv_wr32(priv, 0x00256c, stat);
+}
+
+static void
+nve0_fifo_intr_dropped_fault(struct nve0_fifo_priv *priv)
+{
+	u32 stat = nv_rd32(priv, 0x00259c);
+	nv_error(priv, "DROPPED_MMU_FAULT 0x%08x\n", stat);
+}
+
+static const struct nouveau_enum
+nve0_fifo_fault_engine[] = {
+	{ 0x00, "GR", NULL, NVDEV_ENGINE_GR },
+	{ 0x03, "IFB", NULL, NVDEV_ENGINE_IFB },
+	{ 0x04, "BAR1", NULL, NVDEV_SUBDEV_BAR },
+	{ 0x05, "BAR3", NULL, NVDEV_SUBDEV_INSTMEM },
+	{ 0x07, "PBDMA0", NULL, NVDEV_ENGINE_FIFO },
+	{ 0x08, "PBDMA1", NULL, NVDEV_ENGINE_FIFO },
+	{ 0x09, "PBDMA2", NULL, NVDEV_ENGINE_FIFO },
+	{ 0x10, "MSVLD", NULL, NVDEV_ENGINE_BSP },
+	{ 0x11, "MSPPP", NULL, NVDEV_ENGINE_PPP },
+	{ 0x13, "PERF" },
+	{ 0x14, "MSPDEC", NULL, NVDEV_ENGINE_VP },
+	{ 0x15, "CE0", NULL, NVDEV_ENGINE_COPY0 },
+	{ 0x16, "CE1", NULL, NVDEV_ENGINE_COPY1 },
+	{ 0x17, "PMU" },
+	{ 0x19, "MSENC", NULL, NVDEV_ENGINE_VENC },
+	{ 0x1b, "CE2", NULL, NVDEV_ENGINE_COPY2 },
+	{}
+};
+
+static const struct nouveau_enum
+nve0_fifo_fault_reason[] = {
+	{ 0x00, "PDE" },
+	{ 0x01, "PDE_SIZE" },
+	{ 0x02, "PTE" },
+	{ 0x03, "VA_LIMIT_VIOLATION" },
+	{ 0x04, "UNBOUND_INST_BLOCK" },
+	{ 0x05, "PRIV_VIOLATION" },
+	{ 0x06, "RO_VIOLATION" },
+	{ 0x07, "WO_VIOLATION" },
+	{ 0x08, "PITCH_MASK_VIOLATION" },
+	{ 0x09, "WORK_CREATION" },
+	{ 0x0a, "UNSUPPORTED_APERTURE" },
+	{ 0x0b, "COMPRESSION_FAILURE" },
+	{ 0x0c, "UNSUPPORTED_KIND" },
+	{ 0x0d, "REGION_VIOLATION" },
+	{ 0x0e, "BOTH_PTES_VALID" },
+	{ 0x0f, "INFO_TYPE_POISONED" },
+	{}
+};
+
+static const struct nouveau_enum
+nve0_fifo_fault_hubclient[] = {
+	{ 0x00, "VIP" },
+	{ 0x01, "CE0" },
+	{ 0x02, "CE1" },
+	{ 0x03, "DNISO" },
+	{ 0x04, "FE" },
+	{ 0x05, "FECS" },
+	{ 0x06, "HOST" },
+	{ 0x07, "HOST_CPU" },
+	{ 0x08, "HOST_CPU_NB" },
+	{ 0x09, "ISO" },
+	{ 0x0a, "MMU" },
+	{ 0x0b, "MSPDEC" },
+	{ 0x0c, "MSPPP" },
+	{ 0x0d, "MSVLD" },
+	{ 0x0e, "NISO" },
+	{ 0x0f, "P2P" },
+	{ 0x10, "PD" },
+	{ 0x11, "PERF" },
+	{ 0x12, "PMU" },
+	{ 0x13, "RASTERTWOD" },
+	{ 0x14, "SCC" },
+	{ 0x15, "SCC_NB" },
+	{ 0x16, "SEC" },
+	{ 0x17, "SSYNC" },
+	{ 0x18, "GR_COPY" },
+	{ 0x19, "CE2" },
+	{ 0x1a, "XV" },
+	{ 0x1b, "MMU_NB" },
+	{ 0x1c, "MSENC" },
+	{ 0x1d, "DFALCON" },
+	{ 0x1e, "SKED" },
+	{ 0x1f, "AFALCON" },
+	{}
+};
+
+static const struct nouveau_enum
+nve0_fifo_fault_gpcclient[] = {
+	{ 0x00, "L1_0" }, { 0x01, "T1_0" }, { 0x02, "PE_0" },
+	{ 0x03, "L1_1" }, { 0x04, "T1_1" }, { 0x05, "PE_1" },
+	{ 0x06, "L1_2" }, { 0x07, "T1_2" }, { 0x08, "PE_2" },
+	{ 0x09, "L1_3" }, { 0x0a, "T1_3" }, { 0x0b, "PE_3" },
+	{ 0x0c, "RAST" },
+	{ 0x0d, "GCC" },
+	{ 0x0e, "GPCCS" },
+	{ 0x0f, "PROP_0" },
+	{ 0x10, "PROP_1" },
+	{ 0x11, "PROP_2" },
+	{ 0x12, "PROP_3" },
+	{ 0x13, "L1_4" }, { 0x14, "T1_4" }, { 0x15, "PE_4" },
+	{ 0x16, "L1_5" }, { 0x17, "T1_5" }, { 0x18, "PE_5" },
+	{ 0x19, "L1_6" }, { 0x1a, "T1_6" }, { 0x1b, "PE_6" },
+	{ 0x1c, "L1_7" }, { 0x1d, "T1_7" }, { 0x1e, "PE_7" },
+	{ 0x1f, "GPM" },
+	{ 0x20, "LTP_UTLB_0" },
+	{ 0x21, "LTP_UTLB_1" },
+	{ 0x22, "LTP_UTLB_2" },
+	{ 0x23, "LTP_UTLB_3" },
+	{ 0x24, "GPC_RGG_UTLB" },
+	{}
+};
+
+static void
+nve0_fifo_intr_fault(struct nve0_fifo_priv *priv, int unit)
+{
+	u32 inst = nv_rd32(priv, 0x002800 + (unit * 0x10));
+	u32 valo = nv_rd32(priv, 0x002804 + (unit * 0x10));
+	u32 vahi = nv_rd32(priv, 0x002808 + (unit * 0x10));
+	u32 stat = nv_rd32(priv, 0x00280c + (unit * 0x10));
+	u32 gpc    = (stat & 0x1f000000) >> 24;
+	u32 client = (stat & 0x00001f00) >> 8;
+	u32 write  = (stat & 0x00000080);
+	u32 hub    = (stat & 0x00000040);
+	u32 reason = (stat & 0x0000000f);
+	struct nouveau_object *engctx = NULL, *object;
+	struct nouveau_engine *engine = NULL;
+	const struct nouveau_enum *er, *eu, *ec;
+	char erunk[6] = "";
+	char euunk[6] = "";
+	char ecunk[6] = "";
+	char gpcid[3] = "";
+
+	er = nouveau_enum_find(nve0_fifo_fault_reason, reason);
+	if (!er)
+		snprintf(erunk, sizeof(erunk), "UNK%02X", reason);
+
+	eu = nouveau_enum_find(nve0_fifo_fault_engine, unit);
+	if (eu) {
+		switch (eu->data2) {
+		case NVDEV_SUBDEV_BAR:
+			nv_mask(priv, 0x001704, 0x00000000, 0x00000000);
+			break;
+		case NVDEV_SUBDEV_INSTMEM:
+			nv_mask(priv, 0x001714, 0x00000000, 0x00000000);
+			break;
+		case NVDEV_ENGINE_IFB:
+			nv_mask(priv, 0x001718, 0x00000000, 0x00000000);
+			break;
+		default:
+			engine = nouveau_engine(priv, eu->data2);
+			if (engine)
+				engctx = nouveau_engctx_get(engine, inst);
+			break;
+		}
+	} else {
+		snprintf(euunk, sizeof(euunk), "UNK%02x", unit);
+	}
+
+	if (hub) {
+		ec = nouveau_enum_find(nve0_fifo_fault_hubclient, client);
+	} else {
+		ec = nouveau_enum_find(nve0_fifo_fault_gpcclient, client);
+		snprintf(gpcid, sizeof(gpcid), "%d", gpc);
+	}
+
+	if (!ec)
+		snprintf(ecunk, sizeof(ecunk), "UNK%02x", client);
+
+	nv_error(priv, "%s fault at 0x%010llx [%s] from %s/%s%s%s%s on "
+		       "channel 0x%010llx [%s]\n", write ? "write" : "read",
+		 (u64)vahi << 32 | valo, er ? er->name : erunk,
+		 eu ? eu->name : euunk, hub ? "" : "GPC", gpcid, hub ? "" : "/",
+		 ec ? ec->name : ecunk, (u64)inst << 12,
+		 nouveau_client_name(engctx));
+
+	object = engctx;
+	while (object) {
+		switch (nv_mclass(object)) {
+		case NVE0_CHANNEL_IND_CLASS:
+			nve0_fifo_recover(priv, engine, (void *)object);
+			break;
+		}
+		object = object->parent;
+	}
+
+	nouveau_engctx_put(engctx);
+}
+
+static const struct nouveau_bitfield nve0_fifo_pbdma_intr[] = {
+	{ 0x00000001, "MEMREQ" },
+	{ 0x00000002, "MEMACK_TIMEOUT" },
+	{ 0x00000004, "MEMACK_EXTRA" },
+	{ 0x00000008, "MEMDAT_TIMEOUT" },
+	{ 0x00000010, "MEMDAT_EXTRA" },
+	{ 0x00000020, "MEMFLUSH" },
+	{ 0x00000040, "MEMOP" },
+	{ 0x00000080, "LBCONNECT" },
+	{ 0x00000100, "LBREQ" },
+	{ 0x00000200, "LBACK_TIMEOUT" },
+	{ 0x00000400, "LBACK_EXTRA" },
+	{ 0x00000800, "LBDAT_TIMEOUT" },
+	{ 0x00001000, "LBDAT_EXTRA" },
+	{ 0x00002000, "GPFIFO" },
+	{ 0x00004000, "GPPTR" },
+	{ 0x00008000, "GPENTRY" },
+	{ 0x00010000, "GPCRC" },
+	{ 0x00020000, "PBPTR" },
+	{ 0x00040000, "PBENTRY" },
+	{ 0x00080000, "PBCRC" },
+	{ 0x00100000, "XBARCONNECT" },
+	{ 0x00200000, "METHOD" },
+	{ 0x00400000, "METHODCRC" },
+	{ 0x00800000, "DEVICE" },
+	{ 0x02000000, "SEMAPHORE" },
+	{ 0x04000000, "ACQUIRE" },
+	{ 0x08000000, "PRI" },
+	{ 0x20000000, "NO_CTXSW_SEG" },
+	{ 0x40000000, "PBSEG" },
+	{ 0x80000000, "SIGNATURE" },
+	{}
+};
+
 static void
-nve0_fifo_isr_subfifo_intr(struct nve0_fifo_priv *priv, int unit)
+nve0_fifo_intr_pbdma(struct nve0_fifo_priv *priv, int unit)
 {
 	u32 stat = nv_rd32(priv, 0x040108 + (unit * 0x2000));
 	u32 addr = nv_rd32(priv, 0x0400c0 + (unit * 0x2000));
@@ -487,11 +830,11 @@ nve0_fifo_isr_subfifo_intr(struct nve0_fifo_priv *priv, int unit)
 	}
 
 	if (show) {
-		nv_error(priv, "SUBFIFO%d:", unit);
-		nouveau_bitfield_print(nve0_fifo_subfifo_intr, show);
+		nv_error(priv, "PBDMA%d:", unit);
+		nouveau_bitfield_print(nve0_fifo_pbdma_intr, show);
 		pr_cont("\n");
 		nv_error(priv,
-			 "SUBFIFO%d: ch %d [%s] subc %d mthd 0x%04x data 0x%08x\n",
+			 "PBDMA%d: ch %d [%s] subc %d mthd 0x%04x data 0x%08x\n",
 			 unit, chid,
 			 nouveau_client_name_for_fifo_chid(&priv->base, chid),
 			 subc, mthd, data);
@@ -502,126 +845,171 @@ nve0_fifo_isr_subfifo_intr(struct nve0_fifo_priv *priv, int unit)
 }
 
 static void
+nve0_fifo_intr_runlist(struct nve0_fifo_priv *priv)
+{
+	u32 mask = nv_rd32(priv, 0x002a00);
+	while (mask) {
+		u32 engn = __ffs(mask);
+		wake_up(&priv->engine[engn].wait);
+		nv_wr32(priv, 0x002a00, 1 << engn);
+		mask &= ~(1 << engn);
+	}
+}
+
+static void
+nve0_fifo_intr_engine(struct nve0_fifo_priv *priv)
+{
+	nouveau_event_trigger(priv->base.uevent, 1, 0);
+}
+
+static void
 nve0_fifo_intr(struct nouveau_subdev *subdev)
 {
 	struct nve0_fifo_priv *priv = (void *)subdev;
 	u32 mask = nv_rd32(priv, 0x002140);
 	u32 stat = nv_rd32(priv, 0x002100) & mask;
 
+	if (stat & 0x00000001) {
+		nve0_fifo_intr_bind(priv);
+		nv_wr32(priv, 0x002100, 0x00000001);
+		stat &= ~0x00000001;
+	}
+
+	if (stat & 0x00000010) {
+		nv_error(priv, "PIO_ERROR\n");
+		nv_wr32(priv, 0x002100, 0x00000010);
+		stat &= ~0x00000010;
+	}
+
 	if (stat & 0x00000100) {
-		nv_warn(priv, "unknown status 0x00000100\n");
+		nve0_fifo_intr_sched(priv);
 		nv_wr32(priv, 0x002100, 0x00000100);
 		stat &= ~0x00000100;
 	}
 
-	if (stat & 0x10000000) {
-		u32 units = nv_rd32(priv, 0x00259c);
-		u32 u = units;
+	if (stat & 0x00010000) {
+		nve0_fifo_intr_chsw(priv);
+		nv_wr32(priv, 0x002100, 0x00010000);
+		stat &= ~0x00010000;
+	}
 
-		while (u) {
-			int i = ffs(u) - 1;
-			nve0_fifo_isr_vm_fault(priv, i);
-			u &= ~(1 << i);
-		}
+	if (stat & 0x00800000) {
+		nv_error(priv, "FB_FLUSH_TIMEOUT\n");
+		nv_wr32(priv, 0x002100, 0x00800000);
+		stat &= ~0x00800000;
+	}
 
-		nv_wr32(priv, 0x00259c, units);
-		stat &= ~0x10000000;
+	if (stat & 0x01000000) {
+		nv_error(priv, "LB_ERROR\n");
+		nv_wr32(priv, 0x002100, 0x01000000);
+		stat &= ~0x01000000;
 	}
 
-	if (stat & 0x20000000) {
-		u32 units = nv_rd32(priv, 0x0025a0);
-		u32 u = units;
+	if (stat & 0x08000000) {
+		nve0_fifo_intr_dropped_fault(priv);
+		nv_wr32(priv, 0x002100, 0x08000000);
+		stat &= ~0x08000000;
+	}
 
-		while (u) {
-			int i = ffs(u) - 1;
-			nve0_fifo_isr_subfifo_intr(priv, i);
-			u &= ~(1 << i);
+	if (stat & 0x10000000) {
+		u32 mask = nv_rd32(priv, 0x00259c);
+		while (mask) {
+			u32 unit = __ffs(mask);
+			nve0_fifo_intr_fault(priv, unit);
+			nv_wr32(priv, 0x00259c, (1 << unit));
+			mask &= ~(1 << unit);
 		}
+		stat &= ~0x10000000;
+	}
 
-		nv_wr32(priv, 0x0025a0, units);
+	if (stat & 0x20000000) {
+		u32 mask = nv_rd32(priv, 0x0025a0);
+		while (mask) {
+			u32 unit = __ffs(mask);
+			nve0_fifo_intr_pbdma(priv, unit);
+			nv_wr32(priv, 0x0025a0, (1 << unit));
+			mask &= ~(1 << unit);
+		}
 		stat &= ~0x20000000;
 	}
 
 	if (stat & 0x40000000) {
-		nv_warn(priv, "unknown status 0x40000000\n");
-		nv_mask(priv, 0x002a00, 0x00000000, 0x00000000);
+		nve0_fifo_intr_runlist(priv);
 		stat &= ~0x40000000;
 	}
 
 	if (stat & 0x80000000) {
-		nouveau_event_trigger(priv->base.uevent, 0);
+		nve0_fifo_intr_engine(priv);
 		nv_wr32(priv, 0x002100, 0x80000000);
 		stat &= ~0x80000000;
 	}
 
 	if (stat) {
-		nv_fatal(priv, "unhandled status 0x%08x\n", stat);
+		nv_error(priv, "INTR 0x%08x\n", stat);
+		nv_mask(priv, 0x002140, stat, 0x00000000);
 		nv_wr32(priv, 0x002100, stat);
-		nv_wr32(priv, 0x002140, 0);
 	}
 }
 
 static void
-nve0_fifo_uevent_enable(struct nouveau_event *event, int index)
+nve0_fifo_uevent_enable(struct nouveau_event *event, int type, int index)
 {
 	struct nve0_fifo_priv *priv = event->priv;
 	nv_mask(priv, 0x002140, 0x80000000, 0x80000000);
 }
 
 static void
-nve0_fifo_uevent_disable(struct nouveau_event *event, int index)
+nve0_fifo_uevent_disable(struct nouveau_event *event, int type, int index)
 {
 	struct nve0_fifo_priv *priv = event->priv;
 	nv_mask(priv, 0x002140, 0x80000000, 0x00000000);
 }
 
-static int
-nve0_fifo_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	       struct nouveau_oclass *oclass, void *data, u32 size,
-	       struct nouveau_object **pobject)
+int
+nve0_fifo_fini(struct nouveau_object *object, bool suspend)
 {
-	struct nve0_fifo_priv *priv;
-	int ret, i;
+	struct nve0_fifo_priv *priv = (void *)object;
+	int ret;
 
-	ret = nouveau_fifo_create(parent, engine, oclass, 0, 4095, &priv);
-	*pobject = nv_object(priv);
+	ret = nouveau_fifo_fini(&priv->base, suspend);
 	if (ret)
 		return ret;
 
-	for (i = 0; i < FIFO_ENGINE_NR; i++) {
-		ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0x1000,
-					 0, &priv->engine[i].playlist[0]);
-		if (ret)
-			return ret;
+	/* allow mmu fault interrupts, even when we're not using fifo */
+	nv_mask(priv, 0x002140, 0x10000000, 0x10000000);
+	return 0;
+}
 
-		ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0x1000,
-					 0, &priv->engine[i].playlist[1]);
-		if (ret)
-			return ret;
-	}
+int
+nve0_fifo_init(struct nouveau_object *object)
+{
+	struct nve0_fifo_priv *priv = (void *)object;
+	int ret, i;
 
-	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 4096 * 0x200, 0x1000,
-				 NVOBJ_FLAG_ZERO_ALLOC, &priv->user.mem);
+	ret = nouveau_fifo_init(&priv->base);
 	if (ret)
 		return ret;
 
-	ret = nouveau_gpuobj_map(priv->user.mem, NV_MEM_ACCESS_RW,
-				&priv->user.bar);
-	if (ret)
-		return ret;
+	/* enable all available PBDMA units */
+	nv_wr32(priv, 0x000204, 0xffffffff);
+	priv->spoon_nr = hweight32(nv_rd32(priv, 0x000204));
+	nv_debug(priv, "%d PBDMA unit(s)\n", priv->spoon_nr);
 
-	priv->base.uevent->enable = nve0_fifo_uevent_enable;
-	priv->base.uevent->disable = nve0_fifo_uevent_disable;
-	priv->base.uevent->priv = priv;
+	/* PBDMA[n] */
+	for (i = 0; i < priv->spoon_nr; i++) {
+		nv_mask(priv, 0x04013c + (i * 0x2000), 0x10000100, 0x00000000);
+		nv_wr32(priv, 0x040108 + (i * 0x2000), 0xffffffff); /* INTR */
+		nv_wr32(priv, 0x04010c + (i * 0x2000), 0xfffffeff); /* INTREN */
+	}
 
-	nv_subdev(priv)->unit = 0x00000100;
-	nv_subdev(priv)->intr = nve0_fifo_intr;
-	nv_engine(priv)->cclass = &nve0_fifo_cclass;
-	nv_engine(priv)->sclass = nve0_fifo_sclass;
+	nv_wr32(priv, 0x002254, 0x10000000 | priv->user.bar.offset >> 12);
+
+	nv_wr32(priv, 0x002100, 0xffffffff);
+	nv_wr32(priv, 0x002140, 0x7fffffff);
 	return 0;
 }
 
-static void
+void
 nve0_fifo_dtor(struct nouveau_object *object)
 {
 	struct nve0_fifo_priv *priv = (void *)object;
@@ -631,50 +1019,73 @@ nve0_fifo_dtor(struct nouveau_object *object)
 	nouveau_gpuobj_ref(NULL, &priv->user.mem);
 
 	for (i = 0; i < FIFO_ENGINE_NR; i++) {
-		nouveau_gpuobj_ref(NULL, &priv->engine[i].playlist[1]);
-		nouveau_gpuobj_ref(NULL, &priv->engine[i].playlist[0]);
+		nouveau_gpuobj_ref(NULL, &priv->engine[i].runlist[1]);
+		nouveau_gpuobj_ref(NULL, &priv->engine[i].runlist[0]);
 	}
 
 	nouveau_fifo_destroy(&priv->base);
 }
 
-static int
-nve0_fifo_init(struct nouveau_object *object)
+int
+nve0_fifo_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
 {
-	struct nve0_fifo_priv *priv = (void *)object;
+	struct nve0_fifo_impl *impl = (void *)oclass;
+	struct nve0_fifo_priv *priv;
 	int ret, i;
 
-	ret = nouveau_fifo_init(&priv->base);
+	ret = nouveau_fifo_create(parent, engine, oclass, 0,
+				  impl->channels - 1, &priv);
+	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
 
-	/* enable all available PSUBFIFOs */
-	nv_wr32(priv, 0x000204, 0xffffffff);
-	priv->spoon_nr = hweight32(nv_rd32(priv, 0x000204));
-	nv_debug(priv, "%d subfifo(s)\n", priv->spoon_nr);
+	INIT_WORK(&priv->fault, nve0_fifo_recover_work);
 
-	/* PSUBFIFO[n] */
-	for (i = 0; i < priv->spoon_nr; i++) {
-		nv_mask(priv, 0x04013c + (i * 0x2000), 0x10000100, 0x00000000);
-		nv_wr32(priv, 0x040108 + (i * 0x2000), 0xffffffff); /* INTR */
-		nv_wr32(priv, 0x04010c + (i * 0x2000), 0xfffffeff); /* INTREN */
+	for (i = 0; i < FIFO_ENGINE_NR; i++) {
+		ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0x1000,
+					 0, &priv->engine[i].runlist[0]);
+		if (ret)
+			return ret;
+
+		ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0x1000,
+					 0, &priv->engine[i].runlist[1]);
+		if (ret)
+			return ret;
+
+		init_waitqueue_head(&priv->engine[i].wait);
 	}
 
-	nv_wr32(priv, 0x002254, 0x10000000 | priv->user.bar.offset >> 12);
+	ret = nouveau_gpuobj_new(nv_object(priv), NULL, impl->channels * 0x200,
+				0x1000, NVOBJ_FLAG_ZERO_ALLOC, &priv->user.mem);
+	if (ret)
+		return ret;
 
-	nv_wr32(priv, 0x002a00, 0xffffffff);
-	nv_wr32(priv, 0x002100, 0xffffffff);
-	nv_wr32(priv, 0x002140, 0x3fffffff);
+	ret = nouveau_gpuobj_map(priv->user.mem, NV_MEM_ACCESS_RW,
+				&priv->user.bar);
+	if (ret)
+		return ret;
+
+	priv->base.uevent->enable = nve0_fifo_uevent_enable;
+	priv->base.uevent->disable = nve0_fifo_uevent_disable;
+	priv->base.uevent->priv = priv;
+
+	nv_subdev(priv)->unit = 0x00000100;
+	nv_subdev(priv)->intr = nve0_fifo_intr;
+	nv_engine(priv)->cclass = &nve0_fifo_cclass;
+	nv_engine(priv)->sclass = nve0_fifo_sclass;
 	return 0;
 }
 
 struct nouveau_oclass *
-nve0_fifo_oclass = &(struct nouveau_oclass) {
-	.handle = NV_ENGINE(FIFO, 0xe0),
-	.ofuncs = &(struct nouveau_ofuncs) {
+nve0_fifo_oclass = &(struct nve0_fifo_impl) {
+	.base.handle = NV_ENGINE(FIFO, 0xe0),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nve0_fifo_ctor,
 		.dtor = nve0_fifo_dtor,
 		.init = nve0_fifo_init,
-		.fini = _nouveau_fifo_fini,
+		.fini = nve0_fifo_fini,
 	},
-};
+	.channels = 4096,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.h b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.h
new file mode 100644
index 00000000000..e96b32bb1bb
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.h
@@ -0,0 +1,18 @@
+#ifndef __NVKM_FIFO_NVE0_H__
+#define __NVKM_FIFO_NVE0_H__
+
+#include <engine/fifo.h>
+
+int  nve0_fifo_ctor(struct nouveau_object *, struct nouveau_object *,
+		    struct nouveau_oclass *, void *, u32,
+		    struct nouveau_object **);
+void nve0_fifo_dtor(struct nouveau_object *);
+int  nve0_fifo_init(struct nouveau_object *);
+int  nve0_fifo_fini(struct nouveau_object *, bool);
+
+struct nve0_fifo_impl {
+	struct nouveau_oclass base;
+	u32 channels;
+};
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxgk20a.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxgk20a.c
new file mode 100644
index 00000000000..224ee0287ab
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxgk20a.c
@@ -0,0 +1,53 @@
+/*
+ * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include "ctxnvc0.h"
+
+static const struct nvc0_graph_pack
+gk20a_grctx_pack_mthd[] = {
+	{ nve4_grctx_init_a097_0, 0xa297 },
+	{ nvc0_grctx_init_902d_0, 0x902d },
+	{}
+};
+
+struct nouveau_oclass *
+gk20a_grctx_oclass = &(struct nvc0_grctx_oclass) {
+	.base.handle = NV_ENGCTX(GR, 0xea),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nvc0_graph_context_ctor,
+		.dtor = nvc0_graph_context_dtor,
+		.init = _nouveau_graph_context_init,
+		.fini = _nouveau_graph_context_fini,
+		.rd32 = _nouveau_graph_context_rd32,
+		.wr32 = _nouveau_graph_context_wr32,
+	},
+	.main  = nve4_grctx_generate_main,
+	.mods  = nve4_grctx_generate_mods,
+	.unkn  = nve4_grctx_generate_unkn,
+	.hub   = nve4_grctx_pack_hub,
+	.gpc   = nve4_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nve4_grctx_pack_tpc,
+	.ppc   = nve4_grctx_pack_ppc,
+	.icmd  = nve4_grctx_pack_icmd,
+	.mthd  = gk20a_grctx_pack_mthd,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxgm107.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxgm107.c
new file mode 100644
index 00000000000..b0d0fb2f4d0
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxgm107.c
@@ -0,0 +1,991 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include "ctxnvc0.h"
+
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+gm107_grctx_init_icmd_0[] = {
+	{ 0x001000,   1, 0x01, 0x00000004 },
+	{ 0x000039,   3, 0x01, 0x00000000 },
+	{ 0x0000a9,   1, 0x01, 0x0000ffff },
+	{ 0x000038,   1, 0x01, 0x0fac6881 },
+	{ 0x00003d,   1, 0x01, 0x00000001 },
+	{ 0x0000e8,   8, 0x01, 0x00000400 },
+	{ 0x000078,   8, 0x01, 0x00000300 },
+	{ 0x000050,   1, 0x01, 0x00000011 },
+	{ 0x000058,   8, 0x01, 0x00000008 },
+	{ 0x000208,   8, 0x01, 0x00000001 },
+	{ 0x000081,   1, 0x01, 0x00000001 },
+	{ 0x000085,   1, 0x01, 0x00000004 },
+	{ 0x000088,   1, 0x01, 0x00000400 },
+	{ 0x000090,   1, 0x01, 0x00000300 },
+	{ 0x000098,   1, 0x01, 0x00001001 },
+	{ 0x0000e3,   1, 0x01, 0x00000001 },
+	{ 0x0000da,   1, 0x01, 0x00000001 },
+	{ 0x0000f8,   1, 0x01, 0x00000003 },
+	{ 0x0000fa,   1, 0x01, 0x00000001 },
+	{ 0x0000b1,   2, 0x01, 0x00000001 },
+	{ 0x00009f,   4, 0x01, 0x0000ffff },
+	{ 0x0000a8,   1, 0x01, 0x0000ffff },
+	{ 0x0000ad,   1, 0x01, 0x0000013e },
+	{ 0x0000e1,   1, 0x01, 0x00000010 },
+	{ 0x000290,  16, 0x01, 0x00000000 },
+	{ 0x0003b0,  16, 0x01, 0x00000000 },
+	{ 0x0002a0,  16, 0x01, 0x00000000 },
+	{ 0x000420,  16, 0x01, 0x00000000 },
+	{ 0x0002b0,  16, 0x01, 0x00000000 },
+	{ 0x000430,  16, 0x01, 0x00000000 },
+	{ 0x0002c0,  16, 0x01, 0x00000000 },
+	{ 0x0004d0,  16, 0x01, 0x00000000 },
+	{ 0x000720,  16, 0x01, 0x00000000 },
+	{ 0x0008c0,  16, 0x01, 0x00000000 },
+	{ 0x000890,  16, 0x01, 0x00000000 },
+	{ 0x0008e0,  16, 0x01, 0x00000000 },
+	{ 0x0008a0,  16, 0x01, 0x00000000 },
+	{ 0x0008f0,  16, 0x01, 0x00000000 },
+	{ 0x00094c,   1, 0x01, 0x000000ff },
+	{ 0x00094d,   1, 0x01, 0xffffffff },
+	{ 0x00094e,   1, 0x01, 0x00000002 },
+	{ 0x0002f2,   2, 0x01, 0x00000001 },
+	{ 0x0002f5,   1, 0x01, 0x00000001 },
+	{ 0x0002f7,   1, 0x01, 0x00000001 },
+	{ 0x000303,   1, 0x01, 0x00000001 },
+	{ 0x0002e6,   1, 0x01, 0x00000001 },
+	{ 0x000466,   1, 0x01, 0x00000052 },
+	{ 0x000301,   1, 0x01, 0x3f800000 },
+	{ 0x000304,   1, 0x01, 0x30201000 },
+	{ 0x000305,   1, 0x01, 0x70605040 },
+	{ 0x000306,   1, 0x01, 0xb8a89888 },
+	{ 0x000307,   1, 0x01, 0xf8e8d8c8 },
+	{ 0x00030a,   1, 0x01, 0x00ffff00 },
+	{ 0x0000de,   1, 0x01, 0x00000001 },
+	{ 0x00030b,   1, 0x01, 0x0000001a },
+	{ 0x00030c,   1, 0x01, 0x00000001 },
+	{ 0x000318,   1, 0x01, 0x00000001 },
+	{ 0x000340,   1, 0x01, 0x00000000 },
+	{ 0x00037d,   1, 0x01, 0x00000006 },
+	{ 0x0003a0,   1, 0x01, 0x00000002 },
+	{ 0x0003aa,   1, 0x01, 0x00000001 },
+	{ 0x0003a9,   1, 0x01, 0x00000001 },
+	{ 0x000380,   1, 0x01, 0x00000001 },
+	{ 0x000383,   1, 0x01, 0x00000011 },
+	{ 0x000360,   1, 0x01, 0x00000040 },
+	{ 0x000366,   2, 0x01, 0x00000000 },
+	{ 0x000368,   1, 0x01, 0x00000fff },
+	{ 0x000370,   2, 0x01, 0x00000000 },
+	{ 0x000372,   1, 0x01, 0x000fffff },
+	{ 0x00037a,   1, 0x01, 0x00000012 },
+	{ 0x000619,   1, 0x01, 0x00000003 },
+	{ 0x000811,   1, 0x01, 0x00000003 },
+	{ 0x000812,   1, 0x01, 0x00000004 },
+	{ 0x000813,   1, 0x01, 0x00000006 },
+	{ 0x000814,   1, 0x01, 0x00000008 },
+	{ 0x000815,   1, 0x01, 0x0000000b },
+	{ 0x000800,   6, 0x01, 0x00000001 },
+	{ 0x000632,   1, 0x01, 0x00000001 },
+	{ 0x000633,   1, 0x01, 0x00000002 },
+	{ 0x000634,   1, 0x01, 0x00000003 },
+	{ 0x000635,   1, 0x01, 0x00000004 },
+	{ 0x000654,   1, 0x01, 0x3f800000 },
+	{ 0x000657,   1, 0x01, 0x3f800000 },
+	{ 0x000655,   2, 0x01, 0x3f800000 },
+	{ 0x0006cd,   1, 0x01, 0x3f800000 },
+	{ 0x0007f5,   1, 0x01, 0x3f800000 },
+	{ 0x0007dc,   1, 0x01, 0x39291909 },
+	{ 0x0007dd,   1, 0x01, 0x79695949 },
+	{ 0x0007de,   1, 0x01, 0xb9a99989 },
+	{ 0x0007df,   1, 0x01, 0xf9e9d9c9 },
+	{ 0x0007e8,   1, 0x01, 0x00003210 },
+	{ 0x0007e9,   1, 0x01, 0x00007654 },
+	{ 0x0007ea,   1, 0x01, 0x00000098 },
+	{ 0x0007ec,   1, 0x01, 0x39291909 },
+	{ 0x0007ed,   1, 0x01, 0x79695949 },
+	{ 0x0007ee,   1, 0x01, 0xb9a99989 },
+	{ 0x0007ef,   1, 0x01, 0xf9e9d9c9 },
+	{ 0x0007f0,   1, 0x01, 0x00003210 },
+	{ 0x0007f1,   1, 0x01, 0x00007654 },
+	{ 0x0007f2,   1, 0x01, 0x00000098 },
+	{ 0x0005a5,   1, 0x01, 0x00000001 },
+	{ 0x0005d0,   1, 0x01, 0x20181008 },
+	{ 0x0005d1,   1, 0x01, 0x40383028 },
+	{ 0x0005d2,   1, 0x01, 0x60585048 },
+	{ 0x0005d3,   1, 0x01, 0x80787068 },
+	{ 0x000980, 128, 0x01, 0x00000000 },
+	{ 0x000468,   1, 0x01, 0x00000004 },
+	{ 0x00046c,   1, 0x01, 0x00000001 },
+	{ 0x000470,  96, 0x01, 0x00000000 },
+	{ 0x000510,  16, 0x01, 0x3f800000 },
+	{ 0x000520,   1, 0x01, 0x000002b6 },
+	{ 0x000529,   1, 0x01, 0x00000001 },
+	{ 0x000530,  16, 0x01, 0xffff0000 },
+	{ 0x000550,  32, 0x01, 0xffff0000 },
+	{ 0x000585,   1, 0x01, 0x0000003f },
+	{ 0x000576,   1, 0x01, 0x00000003 },
+	{ 0x00057b,   1, 0x01, 0x00000059 },
+	{ 0x000586,   1, 0x01, 0x00000040 },
+	{ 0x000582,   2, 0x01, 0x00000080 },
+	{ 0x000595,   1, 0x01, 0x00400040 },
+	{ 0x000596,   1, 0x01, 0x00000492 },
+	{ 0x000597,   1, 0x01, 0x08080203 },
+	{ 0x0005ad,   1, 0x01, 0x00000008 },
+	{ 0x000598,   1, 0x01, 0x00020001 },
+	{ 0x0005c2,   1, 0x01, 0x00000001 },
+	{ 0x000638,   2, 0x01, 0x00000001 },
+	{ 0x00063a,   1, 0x01, 0x00000002 },
+	{ 0x00063b,   2, 0x01, 0x00000001 },
+	{ 0x00063d,   1, 0x01, 0x00000002 },
+	{ 0x00063e,   1, 0x01, 0x00000001 },
+	{ 0x0008b8,   8, 0x01, 0x00000001 },
+	{ 0x000900,   8, 0x01, 0x00000001 },
+	{ 0x000908,   8, 0x01, 0x00000002 },
+	{ 0x000910,  16, 0x01, 0x00000001 },
+	{ 0x000920,   8, 0x01, 0x00000002 },
+	{ 0x000928,   8, 0x01, 0x00000001 },
+	{ 0x000662,   1, 0x01, 0x00000001 },
+	{ 0x000648,   9, 0x01, 0x00000001 },
+	{ 0x000658,   1, 0x01, 0x0000000f },
+	{ 0x0007ff,   1, 0x01, 0x0000000a },
+	{ 0x00066a,   1, 0x01, 0x40000000 },
+	{ 0x00066b,   1, 0x01, 0x10000000 },
+	{ 0x00066c,   2, 0x01, 0xffff0000 },
+	{ 0x0007af,   2, 0x01, 0x00000008 },
+	{ 0x0007f6,   1, 0x01, 0x00000001 },
+	{ 0x0006b2,   1, 0x01, 0x00000055 },
+	{ 0x0007ad,   1, 0x01, 0x00000003 },
+	{ 0x000971,   1, 0x01, 0x00000008 },
+	{ 0x000972,   1, 0x01, 0x00000040 },
+	{ 0x000973,   1, 0x01, 0x0000012c },
+	{ 0x00097c,   1, 0x01, 0x00000040 },
+	{ 0x000975,   1, 0x01, 0x00000020 },
+	{ 0x000976,   1, 0x01, 0x00000001 },
+	{ 0x000977,   1, 0x01, 0x00000020 },
+	{ 0x000978,   1, 0x01, 0x00000001 },
+	{ 0x000957,   1, 0x01, 0x00000003 },
+	{ 0x00095e,   1, 0x01, 0x20164010 },
+	{ 0x00095f,   1, 0x01, 0x00000020 },
+	{ 0x000a0d,   1, 0x01, 0x00000006 },
+	{ 0x00097d,   1, 0x01, 0x0000000c },
+	{ 0x000683,   1, 0x01, 0x00000006 },
+	{ 0x000687,   1, 0x01, 0x003fffff },
+	{ 0x0006a0,   1, 0x01, 0x00000005 },
+	{ 0x000840,   1, 0x01, 0x00400008 },
+	{ 0x000841,   1, 0x01, 0x08000080 },
+	{ 0x000842,   1, 0x01, 0x00400008 },
+	{ 0x000843,   1, 0x01, 0x08000080 },
+	{ 0x000818,   8, 0x01, 0x00000000 },
+	{ 0x000848,  16, 0x01, 0x00000000 },
+	{ 0x000738,   1, 0x01, 0x00000000 },
+	{ 0x0006aa,   1, 0x01, 0x00000001 },
+	{ 0x0006ab,   1, 0x01, 0x00000002 },
+	{ 0x0006ac,   1, 0x01, 0x00000080 },
+	{ 0x0006ad,   2, 0x01, 0x00000100 },
+	{ 0x0006b1,   1, 0x01, 0x00000011 },
+	{ 0x0006bb,   1, 0x01, 0x000000cf },
+	{ 0x0006ce,   1, 0x01, 0x2a712488 },
+	{ 0x000739,   1, 0x01, 0x4085c000 },
+	{ 0x00073a,   1, 0x01, 0x00000080 },
+	{ 0x000786,   1, 0x01, 0x80000100 },
+	{ 0x00073c,   1, 0x01, 0x00010100 },
+	{ 0x00073d,   1, 0x01, 0x02800000 },
+	{ 0x000787,   1, 0x01, 0x000000cf },
+	{ 0x00078c,   1, 0x01, 0x00000008 },
+	{ 0x000792,   1, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
+	{ 0x000797,   1, 0x01, 0x000000cf },
+	{ 0x000836,   1, 0x01, 0x00000001 },
+	{ 0x00079a,   1, 0x01, 0x00000002 },
+	{ 0x000833,   1, 0x01, 0x04444480 },
+	{ 0x0007a1,   1, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
+	{ 0x000831,   1, 0x01, 0x00000004 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x000a04,   1, 0x01, 0x000000ff },
+	{ 0x000a0b,   1, 0x01, 0x00000040 },
+	{ 0x00097f,   1, 0x01, 0x00000100 },
+	{ 0x000a02,   1, 0x01, 0x00000001 },
+	{ 0x000809,   1, 0x01, 0x00000007 },
+	{ 0x00c221,   1, 0x01, 0x00000040 },
+	{ 0x00c1b0,   8, 0x01, 0x0000000f },
+	{ 0x00c1b8,   1, 0x01, 0x0fac6881 },
+	{ 0x00c1b9,   1, 0x01, 0x00fac688 },
+	{ 0x00c401,   1, 0x01, 0x00000001 },
+	{ 0x00c402,   1, 0x01, 0x00010001 },
+	{ 0x00c403,   2, 0x01, 0x00000001 },
+	{ 0x00c40e,   1, 0x01, 0x00000020 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000002 },
+	{ 0x0006aa,   1, 0x01, 0x00000001 },
+	{ 0x0006ad,   2, 0x01, 0x00000100 },
+	{ 0x0006b1,   1, 0x01, 0x00000011 },
+	{ 0x00078c,   1, 0x01, 0x00000008 },
+	{ 0x000792,   1, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
+	{ 0x000797,   1, 0x01, 0x000000cf },
+	{ 0x00079a,   1, 0x01, 0x00000002 },
+	{ 0x0007a1,   1, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
+	{ 0x000831,   1, 0x01, 0x00000004 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000008 },
+	{ 0x000039,   3, 0x01, 0x00000000 },
+	{ 0x000380,   1, 0x01, 0x00000001 },
+	{ 0x000366,   2, 0x01, 0x00000000 },
+	{ 0x000368,   1, 0x01, 0x00000fff },
+	{ 0x000370,   2, 0x01, 0x00000000 },
+	{ 0x000372,   1, 0x01, 0x000fffff },
+	{ 0x000813,   1, 0x01, 0x00000006 },
+	{ 0x000814,   1, 0x01, 0x00000008 },
+	{ 0x000818,   8, 0x01, 0x00000000 },
+	{ 0x000848,  16, 0x01, 0x00000000 },
+	{ 0x000738,   1, 0x01, 0x00000000 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x000a04,   1, 0x01, 0x000000ff },
+	{ 0x000a0b,   1, 0x01, 0x00000040 },
+	{ 0x00097f,   1, 0x01, 0x00000100 },
+	{ 0x000a02,   1, 0x01, 0x00000001 },
+	{ 0x000809,   1, 0x01, 0x00000007 },
+	{ 0x00c221,   1, 0x01, 0x00000040 },
+	{ 0x00c401,   1, 0x01, 0x00000001 },
+	{ 0x00c402,   1, 0x01, 0x00010001 },
+	{ 0x00c403,   2, 0x01, 0x00000001 },
+	{ 0x00c40e,   1, 0x01, 0x00000020 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000001 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+gm107_grctx_pack_icmd[] = {
+	{ gm107_grctx_init_icmd_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_b097_0[] = {
+	{ 0x000800,   8, 0x40, 0x00000000 },
+	{ 0x000804,   8, 0x40, 0x00000000 },
+	{ 0x000808,   8, 0x40, 0x00000400 },
+	{ 0x00080c,   8, 0x40, 0x00000300 },
+	{ 0x000810,   1, 0x04, 0x000000cf },
+	{ 0x000850,   7, 0x40, 0x00000000 },
+	{ 0x000814,   8, 0x40, 0x00000040 },
+	{ 0x000818,   8, 0x40, 0x00000001 },
+	{ 0x00081c,   8, 0x40, 0x00000000 },
+	{ 0x000820,   8, 0x40, 0x00000000 },
+	{ 0x001c00,  16, 0x10, 0x00000000 },
+	{ 0x001c04,  16, 0x10, 0x00000000 },
+	{ 0x001c08,  16, 0x10, 0x00000000 },
+	{ 0x001c0c,  16, 0x10, 0x00000000 },
+	{ 0x001d00,  16, 0x10, 0x00000000 },
+	{ 0x001d04,  16, 0x10, 0x00000000 },
+	{ 0x001d08,  16, 0x10, 0x00000000 },
+	{ 0x001d0c,  16, 0x10, 0x00000000 },
+	{ 0x001f00,  16, 0x08, 0x00000000 },
+	{ 0x001f04,  16, 0x08, 0x00000000 },
+	{ 0x001f80,  16, 0x08, 0x00000000 },
+	{ 0x001f84,  16, 0x08, 0x00000000 },
+	{ 0x002000,   1, 0x04, 0x00000000 },
+	{ 0x002040,   1, 0x04, 0x00000011 },
+	{ 0x002080,   1, 0x04, 0x00000020 },
+	{ 0x0020c0,   1, 0x04, 0x00000030 },
+	{ 0x002100,   1, 0x04, 0x00000040 },
+	{ 0x002140,   1, 0x04, 0x00000051 },
+	{ 0x00200c,   6, 0x40, 0x00000001 },
+	{ 0x002010,   1, 0x04, 0x00000000 },
+	{ 0x002050,   1, 0x04, 0x00000000 },
+	{ 0x002090,   1, 0x04, 0x00000001 },
+	{ 0x0020d0,   1, 0x04, 0x00000002 },
+	{ 0x002110,   1, 0x04, 0x00000003 },
+	{ 0x002150,   1, 0x04, 0x00000004 },
+	{ 0x000380,   4, 0x20, 0x00000000 },
+	{ 0x000384,   4, 0x20, 0x00000000 },
+	{ 0x000388,   4, 0x20, 0x00000000 },
+	{ 0x00038c,   4, 0x20, 0x00000000 },
+	{ 0x000700,   4, 0x10, 0x00000000 },
+	{ 0x000704,   4, 0x10, 0x00000000 },
+	{ 0x000708,   4, 0x10, 0x00000000 },
+	{ 0x002800, 128, 0x04, 0x00000000 },
+	{ 0x000a00,  16, 0x20, 0x00000000 },
+	{ 0x000a04,  16, 0x20, 0x00000000 },
+	{ 0x000a08,  16, 0x20, 0x00000000 },
+	{ 0x000a0c,  16, 0x20, 0x00000000 },
+	{ 0x000a10,  16, 0x20, 0x00000000 },
+	{ 0x000a14,  16, 0x20, 0x00000000 },
+	{ 0x000c00,  16, 0x10, 0x00000000 },
+	{ 0x000c04,  16, 0x10, 0x00000000 },
+	{ 0x000c08,  16, 0x10, 0x00000000 },
+	{ 0x000c0c,  16, 0x10, 0x3f800000 },
+	{ 0x000d00,   8, 0x08, 0xffff0000 },
+	{ 0x000d04,   8, 0x08, 0xffff0000 },
+	{ 0x000e00,  16, 0x10, 0x00000000 },
+	{ 0x000e04,  16, 0x10, 0xffff0000 },
+	{ 0x000e08,  16, 0x10, 0xffff0000 },
+	{ 0x000d40,   4, 0x08, 0x00000000 },
+	{ 0x000d44,   4, 0x08, 0x00000000 },
+	{ 0x001e00,   8, 0x20, 0x00000001 },
+	{ 0x001e04,   8, 0x20, 0x00000001 },
+	{ 0x001e08,   8, 0x20, 0x00000002 },
+	{ 0x001e0c,   8, 0x20, 0x00000001 },
+	{ 0x001e10,   8, 0x20, 0x00000001 },
+	{ 0x001e14,   8, 0x20, 0x00000002 },
+	{ 0x001e18,   8, 0x20, 0x00000001 },
+	{ 0x001480,   8, 0x10, 0x00000000 },
+	{ 0x001484,   8, 0x10, 0x00000000 },
+	{ 0x001488,   8, 0x10, 0x00000000 },
+	{ 0x003400, 128, 0x04, 0x00000000 },
+	{ 0x00030c,   1, 0x04, 0x00000001 },
+	{ 0x001944,   1, 0x04, 0x00000000 },
+	{ 0x001514,   1, 0x04, 0x00000000 },
+	{ 0x000d68,   1, 0x04, 0x0000ffff },
+	{ 0x00121c,   1, 0x04, 0x0fac6881 },
+	{ 0x000fac,   1, 0x04, 0x00000001 },
+	{ 0x001538,   1, 0x04, 0x00000001 },
+	{ 0x000fe0,   2, 0x04, 0x00000000 },
+	{ 0x000fe8,   1, 0x04, 0x00000014 },
+	{ 0x000fec,   1, 0x04, 0x00000040 },
+	{ 0x000ff0,   1, 0x04, 0x00000000 },
+	{ 0x00179c,   1, 0x04, 0x00000000 },
+	{ 0x001228,   1, 0x04, 0x00000400 },
+	{ 0x00122c,   1, 0x04, 0x00000300 },
+	{ 0x001230,   1, 0x04, 0x00010001 },
+	{ 0x0007f8,   1, 0x04, 0x00000000 },
+	{ 0x0015b4,   1, 0x04, 0x00000001 },
+	{ 0x0015cc,   1, 0x04, 0x00000000 },
+	{ 0x001534,   1, 0x04, 0x00000000 },
+	{ 0x000754,   1, 0x04, 0x00000001 },
+	{ 0x000fb0,   1, 0x04, 0x00000000 },
+	{ 0x0015d0,   1, 0x04, 0x00000000 },
+	{ 0x00153c,   1, 0x04, 0x00000000 },
+	{ 0x0016b4,   1, 0x04, 0x00000003 },
+	{ 0x000fbc,   4, 0x04, 0x0000ffff },
+	{ 0x000df8,   2, 0x04, 0x00000000 },
+	{ 0x001948,   1, 0x04, 0x00000000 },
+	{ 0x001970,   1, 0x04, 0x00000001 },
+	{ 0x00161c,   1, 0x04, 0x000009f0 },
+	{ 0x000dcc,   1, 0x04, 0x00000010 },
+	{ 0x0015e4,   1, 0x04, 0x00000000 },
+	{ 0x001160,  32, 0x04, 0x25e00040 },
+	{ 0x001880,  32, 0x04, 0x00000000 },
+	{ 0x000f84,   2, 0x04, 0x00000000 },
+	{ 0x0017c8,   2, 0x04, 0x00000000 },
+	{ 0x0017d0,   1, 0x04, 0x000000ff },
+	{ 0x0017d4,   1, 0x04, 0xffffffff },
+	{ 0x0017d8,   1, 0x04, 0x00000002 },
+	{ 0x0017dc,   1, 0x04, 0x00000000 },
+	{ 0x0015f4,   2, 0x04, 0x00000000 },
+	{ 0x001434,   2, 0x04, 0x00000000 },
+	{ 0x000d74,   1, 0x04, 0x00000000 },
+	{ 0x0013a4,   1, 0x04, 0x00000000 },
+	{ 0x001318,   1, 0x04, 0x00000001 },
+	{ 0x001080,   2, 0x04, 0x00000000 },
+	{ 0x001088,   2, 0x04, 0x00000001 },
+	{ 0x001090,   1, 0x04, 0x00000000 },
+	{ 0x001094,   1, 0x04, 0x00000001 },
+	{ 0x001098,   1, 0x04, 0x00000000 },
+	{ 0x00109c,   1, 0x04, 0x00000001 },
+	{ 0x0010a0,   2, 0x04, 0x00000000 },
+	{ 0x001644,   1, 0x04, 0x00000000 },
+	{ 0x000748,   1, 0x04, 0x00000000 },
+	{ 0x000de8,   1, 0x04, 0x00000000 },
+	{ 0x001648,   1, 0x04, 0x00000000 },
+	{ 0x0012a4,   1, 0x04, 0x00000000 },
+	{ 0x001120,   4, 0x04, 0x00000000 },
+	{ 0x001118,   1, 0x04, 0x00000000 },
+	{ 0x00164c,   1, 0x04, 0x00000000 },
+	{ 0x001658,   1, 0x04, 0x00000000 },
+	{ 0x001910,   1, 0x04, 0x00000290 },
+	{ 0x001518,   1, 0x04, 0x00000000 },
+	{ 0x00165c,   1, 0x04, 0x00000001 },
+	{ 0x001520,   1, 0x04, 0x00000000 },
+	{ 0x001604,   1, 0x04, 0x00000000 },
+	{ 0x001570,   1, 0x04, 0x00000000 },
+	{ 0x0013b0,   2, 0x04, 0x3f800000 },
+	{ 0x00020c,   1, 0x04, 0x00000000 },
+	{ 0x001670,   1, 0x04, 0x30201000 },
+	{ 0x001674,   1, 0x04, 0x70605040 },
+	{ 0x001678,   1, 0x04, 0xb8a89888 },
+	{ 0x00167c,   1, 0x04, 0xf8e8d8c8 },
+	{ 0x00166c,   1, 0x04, 0x00000000 },
+	{ 0x001680,   1, 0x04, 0x00ffff00 },
+	{ 0x0012d0,   1, 0x04, 0x00000003 },
+	{ 0x0012d4,   1, 0x04, 0x00000002 },
+	{ 0x001684,   2, 0x04, 0x00000000 },
+	{ 0x000dac,   2, 0x04, 0x00001b02 },
+	{ 0x000db4,   1, 0x04, 0x00000000 },
+	{ 0x00168c,   1, 0x04, 0x00000000 },
+	{ 0x0015bc,   1, 0x04, 0x00000000 },
+	{ 0x00156c,   1, 0x04, 0x00000000 },
+	{ 0x00187c,   1, 0x04, 0x00000000 },
+	{ 0x001110,   1, 0x04, 0x00000001 },
+	{ 0x000dc0,   3, 0x04, 0x00000000 },
+	{ 0x000f40,   5, 0x04, 0x00000000 },
+	{ 0x001234,   1, 0x04, 0x00000000 },
+	{ 0x001690,   1, 0x04, 0x00000000 },
+	{ 0x000790,   5, 0x04, 0x00000000 },
+	{ 0x00077c,   1, 0x04, 0x00000000 },
+	{ 0x001000,   1, 0x04, 0x00000010 },
+	{ 0x0010fc,   1, 0x04, 0x00000000 },
+	{ 0x001290,   1, 0x04, 0x00000000 },
+	{ 0x000218,   1, 0x04, 0x00000010 },
+	{ 0x0012d8,   1, 0x04, 0x00000000 },
+	{ 0x0012dc,   1, 0x04, 0x00000010 },
+	{ 0x000d94,   1, 0x04, 0x00000001 },
+	{ 0x00155c,   2, 0x04, 0x00000000 },
+	{ 0x001564,   1, 0x04, 0x00000fff },
+	{ 0x001574,   2, 0x04, 0x00000000 },
+	{ 0x00157c,   1, 0x04, 0x000fffff },
+	{ 0x001354,   1, 0x04, 0x00000000 },
+	{ 0x001610,   1, 0x04, 0x00000012 },
+	{ 0x001608,   2, 0x04, 0x00000000 },
+	{ 0x00260c,   1, 0x04, 0x00000000 },
+	{ 0x0007ac,   1, 0x04, 0x00000000 },
+	{ 0x00162c,   1, 0x04, 0x00000003 },
+	{ 0x000210,   1, 0x04, 0x00000000 },
+	{ 0x000320,   1, 0x04, 0x00000000 },
+	{ 0x000324,   6, 0x04, 0x3f800000 },
+	{ 0x000750,   1, 0x04, 0x00000000 },
+	{ 0x000760,   1, 0x04, 0x39291909 },
+	{ 0x000764,   1, 0x04, 0x79695949 },
+	{ 0x000768,   1, 0x04, 0xb9a99989 },
+	{ 0x00076c,   1, 0x04, 0xf9e9d9c9 },
+	{ 0x000770,   1, 0x04, 0x30201000 },
+	{ 0x000774,   1, 0x04, 0x70605040 },
+	{ 0x000778,   1, 0x04, 0x00009080 },
+	{ 0x000780,   1, 0x04, 0x39291909 },
+	{ 0x000784,   1, 0x04, 0x79695949 },
+	{ 0x000788,   1, 0x04, 0xb9a99989 },
+	{ 0x00078c,   1, 0x04, 0xf9e9d9c9 },
+	{ 0x0007d0,   1, 0x04, 0x30201000 },
+	{ 0x0007d4,   1, 0x04, 0x70605040 },
+	{ 0x0007d8,   1, 0x04, 0x00009080 },
+	{ 0x00037c,   1, 0x04, 0x00000001 },
+	{ 0x000740,   2, 0x04, 0x00000000 },
+	{ 0x002600,   1, 0x04, 0x00000000 },
+	{ 0x001918,   1, 0x04, 0x00000000 },
+	{ 0x00191c,   1, 0x04, 0x00000900 },
+	{ 0x001920,   1, 0x04, 0x00000405 },
+	{ 0x001308,   1, 0x04, 0x00000001 },
+	{ 0x001924,   1, 0x04, 0x00000000 },
+	{ 0x0013ac,   1, 0x04, 0x00000000 },
+	{ 0x00192c,   1, 0x04, 0x00000001 },
+	{ 0x00193c,   1, 0x04, 0x00002c1c },
+	{ 0x000d7c,   1, 0x04, 0x00000000 },
+	{ 0x000f8c,   1, 0x04, 0x00000000 },
+	{ 0x0002c0,   1, 0x04, 0x00000001 },
+	{ 0x001510,   1, 0x04, 0x00000000 },
+	{ 0x001940,   1, 0x04, 0x00000000 },
+	{ 0x000ff4,   2, 0x04, 0x00000000 },
+	{ 0x00194c,   2, 0x04, 0x00000000 },
+	{ 0x001968,   1, 0x04, 0x00000000 },
+	{ 0x001590,   1, 0x04, 0x0000003f },
+	{ 0x0007e8,   4, 0x04, 0x00000000 },
+	{ 0x00196c,   1, 0x04, 0x00000011 },
+	{ 0x0002e4,   1, 0x04, 0x0000b001 },
+	{ 0x00036c,   2, 0x04, 0x00000000 },
+	{ 0x00197c,   1, 0x04, 0x00000000 },
+	{ 0x000fcc,   2, 0x04, 0x00000000 },
+	{ 0x0002d8,   1, 0x04, 0x00000040 },
+	{ 0x001980,   1, 0x04, 0x00000080 },
+	{ 0x001504,   1, 0x04, 0x00000080 },
+	{ 0x001984,   1, 0x04, 0x00000000 },
+	{ 0x000f60,   1, 0x04, 0x00000000 },
+	{ 0x000f64,   1, 0x04, 0x00400040 },
+	{ 0x000f68,   1, 0x04, 0x00002212 },
+	{ 0x000f6c,   1, 0x04, 0x08080203 },
+	{ 0x001108,   1, 0x04, 0x00000008 },
+	{ 0x000f70,   1, 0x04, 0x00080001 },
+	{ 0x000ffc,   1, 0x04, 0x00000000 },
+	{ 0x000300,   1, 0x04, 0x00000001 },
+	{ 0x0013a8,   1, 0x04, 0x00000000 },
+	{ 0x0012ec,   1, 0x04, 0x00000000 },
+	{ 0x001310,   1, 0x04, 0x00000000 },
+	{ 0x001314,   1, 0x04, 0x00000001 },
+	{ 0x001380,   1, 0x04, 0x00000000 },
+	{ 0x001384,   4, 0x04, 0x00000001 },
+	{ 0x001394,   1, 0x04, 0x00000000 },
+	{ 0x00139c,   1, 0x04, 0x00000000 },
+	{ 0x001398,   1, 0x04, 0x00000000 },
+	{ 0x001594,   1, 0x04, 0x00000000 },
+	{ 0x001598,   4, 0x04, 0x00000001 },
+	{ 0x000f54,   3, 0x04, 0x00000000 },
+	{ 0x0019bc,   1, 0x04, 0x00000000 },
+	{ 0x000f9c,   2, 0x04, 0x00000000 },
+	{ 0x0012cc,   1, 0x04, 0x00000000 },
+	{ 0x0012e8,   1, 0x04, 0x00000000 },
+	{ 0x00130c,   1, 0x04, 0x00000001 },
+	{ 0x001360,   8, 0x04, 0x00000000 },
+	{ 0x00133c,   2, 0x04, 0x00000001 },
+	{ 0x001344,   1, 0x04, 0x00000002 },
+	{ 0x001348,   2, 0x04, 0x00000001 },
+	{ 0x001350,   1, 0x04, 0x00000002 },
+	{ 0x001358,   1, 0x04, 0x00000001 },
+	{ 0x0012e4,   1, 0x04, 0x00000000 },
+	{ 0x00131c,   4, 0x04, 0x00000000 },
+	{ 0x0019c0,   1, 0x04, 0x00000000 },
+	{ 0x001140,   1, 0x04, 0x00000000 },
+	{ 0x000dd0,   1, 0x04, 0x00000000 },
+	{ 0x000dd4,   1, 0x04, 0x00000001 },
+	{ 0x0002f4,   1, 0x04, 0x00000000 },
+	{ 0x0019c4,   1, 0x04, 0x00000000 },
+	{ 0x0019c8,   1, 0x04, 0x00001500 },
+	{ 0x00135c,   1, 0x04, 0x00000000 },
+	{ 0x000f90,   1, 0x04, 0x00000000 },
+	{ 0x0019e0,   8, 0x04, 0x00000001 },
+	{ 0x0019cc,   1, 0x04, 0x00000001 },
+	{ 0x0015b8,   1, 0x04, 0x00000000 },
+	{ 0x001a00,   1, 0x04, 0x00001111 },
+	{ 0x001a04,   7, 0x04, 0x00000000 },
+	{ 0x000d6c,   2, 0x04, 0xffff0000 },
+	{ 0x0010f8,   1, 0x04, 0x00001010 },
+	{ 0x000d80,   5, 0x04, 0x00000000 },
+	{ 0x000da0,   1, 0x04, 0x00000000 },
+	{ 0x0007a4,   2, 0x04, 0x00000000 },
+	{ 0x001508,   1, 0x04, 0x80000000 },
+	{ 0x00150c,   1, 0x04, 0x40000000 },
+	{ 0x001668,   1, 0x04, 0x00000000 },
+	{ 0x000318,   2, 0x04, 0x00000008 },
+	{ 0x000d9c,   1, 0x04, 0x00000001 },
+	{ 0x000f14,   1, 0x04, 0x00000000 },
+	{ 0x000374,   1, 0x04, 0x00000000 },
+	{ 0x000378,   1, 0x04, 0x0000000c },
+	{ 0x0007dc,   1, 0x04, 0x00000000 },
+	{ 0x00074c,   1, 0x04, 0x00000055 },
+	{ 0x001420,   1, 0x04, 0x00000003 },
+	{ 0x001008,   1, 0x04, 0x00000008 },
+	{ 0x00100c,   1, 0x04, 0x00000040 },
+	{ 0x001010,   1, 0x04, 0x0000012c },
+	{ 0x000d60,   1, 0x04, 0x00000040 },
+	{ 0x001018,   1, 0x04, 0x00000020 },
+	{ 0x00101c,   1, 0x04, 0x00000001 },
+	{ 0x001020,   1, 0x04, 0x00000020 },
+	{ 0x001024,   1, 0x04, 0x00000001 },
+	{ 0x001444,   3, 0x04, 0x00000000 },
+	{ 0x000360,   1, 0x04, 0x20164010 },
+	{ 0x000364,   1, 0x04, 0x00000020 },
+	{ 0x000368,   1, 0x04, 0x00000000 },
+	{ 0x000da8,   1, 0x04, 0x00000030 },
+	{ 0x000de4,   1, 0x04, 0x00000000 },
+	{ 0x000204,   1, 0x04, 0x00000006 },
+	{ 0x0002d0,   1, 0x04, 0x003fffff },
+	{ 0x001220,   1, 0x04, 0x00000005 },
+	{ 0x000fdc,   1, 0x04, 0x00000000 },
+	{ 0x000f98,   1, 0x04, 0x00400008 },
+	{ 0x001284,   1, 0x04, 0x08000080 },
+	{ 0x001450,   1, 0x04, 0x00400008 },
+	{ 0x001454,   1, 0x04, 0x08000080 },
+	{ 0x000214,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+gm107_grctx_pack_mthd[] = {
+	{ gm107_grctx_init_b097_0, 0xb097 },
+	{ nvc0_grctx_init_902d_0, 0x902d },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_fe_0[] = {
+	{ 0x404004,   8, 0x04, 0x00000000 },
+	{ 0x404024,   1, 0x04, 0x0000e000 },
+	{ 0x404028,   8, 0x04, 0x00000000 },
+	{ 0x4040a8,   8, 0x04, 0x00000000 },
+	{ 0x4040c8,   1, 0x04, 0xf800008f },
+	{ 0x4040d0,   6, 0x04, 0x00000000 },
+	{ 0x4040f8,   1, 0x04, 0x00000000 },
+	{ 0x404100,  10, 0x04, 0x00000000 },
+	{ 0x404130,   2, 0x04, 0x00000000 },
+	{ 0x404150,   1, 0x04, 0x0000002e },
+	{ 0x404154,   1, 0x04, 0x00000400 },
+	{ 0x404158,   1, 0x04, 0x00000200 },
+	{ 0x404164,   1, 0x04, 0x00000045 },
+	{ 0x40417c,   2, 0x04, 0x00000000 },
+	{ 0x404194,   1, 0x04, 0x01000700 },
+	{ 0x4041a0,   4, 0x04, 0x00000000 },
+	{ 0x404200,   4, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_ds_0[] = {
+	{ 0x405800,   1, 0x04, 0x0f8001bf },
+	{ 0x405830,   1, 0x04, 0x0aa01000 },
+	{ 0x405834,   1, 0x04, 0x08000000 },
+	{ 0x405838,   1, 0x04, 0x00000000 },
+	{ 0x405854,   1, 0x04, 0x00000000 },
+	{ 0x405870,   4, 0x04, 0x00000001 },
+	{ 0x405a00,   2, 0x04, 0x00000000 },
+	{ 0x405a18,   1, 0x04, 0x00000000 },
+	{ 0x405a1c,   1, 0x04, 0x000000ff },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_pd_0[] = {
+	{ 0x406020,   1, 0x04, 0x07410001 },
+	{ 0x406028,   4, 0x04, 0x00000001 },
+	{ 0x4064a8,   1, 0x04, 0x00000000 },
+	{ 0x4064ac,   1, 0x04, 0x00003fff },
+	{ 0x4064b0,   3, 0x04, 0x00000000 },
+	{ 0x4064c0,   1, 0x04, 0x80400280 },
+	{ 0x4064c4,   1, 0x04, 0x0400ffff },
+	{ 0x4064c8,   1, 0x04, 0x018001ff },
+	{ 0x4064cc,   9, 0x04, 0x00000000 },
+	{ 0x4064fc,   1, 0x04, 0x0000022a },
+	{ 0x406500,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_be_0[] = {
+	{ 0x408800,   1, 0x04, 0x32802a3c },
+	{ 0x408804,   1, 0x04, 0x00000040 },
+	{ 0x408808,   1, 0x04, 0x1003e005 },
+	{ 0x408840,   1, 0x04, 0x0000000b },
+	{ 0x408900,   1, 0x04, 0xb080b801 },
+	{ 0x408904,   1, 0x04, 0x63038001 },
+	{ 0x408908,   1, 0x04, 0x02c8102f },
+	{ 0x408980,   1, 0x04, 0x0000011d },
+	{}
+};
+
+static const struct nvc0_graph_pack
+gm107_grctx_pack_hub[] = {
+	{ nvc0_grctx_init_main_0 },
+	{ gm107_grctx_init_fe_0 },
+	{ nvf0_grctx_init_pri_0 },
+	{ nve4_grctx_init_memfmt_0 },
+	{ gm107_grctx_init_ds_0 },
+	{ nvf0_grctx_init_cwd_0 },
+	{ gm107_grctx_init_pd_0 },
+	{ nv108_grctx_init_rstr2d_0 },
+	{ nve4_grctx_init_scc_0 },
+	{ gm107_grctx_init_be_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_gpc_unk_0[] = {
+	{ 0x418380,   1, 0x04, 0x00000056 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_gpc_unk_1[] = {
+	{ 0x418600,   1, 0x04, 0x0000007f },
+	{ 0x418684,   1, 0x04, 0x0000001f },
+	{ 0x418700,   1, 0x04, 0x00000002 },
+	{ 0x418704,   1, 0x04, 0x00000080 },
+	{ 0x418708,   1, 0x04, 0x40000000 },
+	{ 0x41870c,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_setup_0[] = {
+	{ 0x418800,   1, 0x04, 0x7006863a },
+	{ 0x418810,   1, 0x04, 0x00000000 },
+	{ 0x418828,   1, 0x04, 0x00000044 },
+	{ 0x418830,   1, 0x04, 0x10000001 },
+	{ 0x4188d8,   1, 0x04, 0x00000008 },
+	{ 0x4188e0,   1, 0x04, 0x01000000 },
+	{ 0x4188e8,   5, 0x04, 0x00000000 },
+	{ 0x4188fc,   1, 0x04, 0x20100058 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_gpc_unk_2[] = {
+	{ 0x418d24,   1, 0x04, 0x00000000 },
+	{ 0x418e00,   1, 0x04, 0x90000000 },
+	{ 0x418e24,   1, 0x04, 0x00000000 },
+	{ 0x418e28,   1, 0x04, 0x00000030 },
+	{ 0x418e30,   1, 0x04, 0x00000000 },
+	{ 0x418e34,   1, 0x04, 0x00010000 },
+	{ 0x418e38,   1, 0x04, 0x00000000 },
+	{ 0x418e40,  22, 0x04, 0x00000000 },
+	{ 0x418ea0,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+gm107_grctx_pack_gpc[] = {
+	{ gm107_grctx_init_gpc_unk_0 },
+	{ nv108_grctx_init_prop_0 },
+	{ gm107_grctx_init_gpc_unk_1 },
+	{ gm107_grctx_init_setup_0 },
+	{ nvc0_grctx_init_zcull_0 },
+	{ nv108_grctx_init_crstr_0 },
+	{ nve4_grctx_init_gpm_0 },
+	{ gm107_grctx_init_gpc_unk_2 },
+	{ nvc0_grctx_init_gcc_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_tex_0[] = {
+	{ 0x419a00,   1, 0x04, 0x000300f0 },
+	{ 0x419a04,   1, 0x04, 0x00000005 },
+	{ 0x419a08,   1, 0x04, 0x00000421 },
+	{ 0x419a0c,   1, 0x04, 0x00120000 },
+	{ 0x419a10,   1, 0x04, 0x00000000 },
+	{ 0x419a14,   1, 0x04, 0x00002200 },
+	{ 0x419a1c,   1, 0x04, 0x0000c000 },
+	{ 0x419a20,   1, 0x04, 0x20008a00 },
+	{ 0x419a30,   1, 0x04, 0x00000001 },
+	{ 0x419a3c,   1, 0x04, 0x00000002 },
+	{ 0x419ac4,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_mpc_0[] = {
+	{ 0x419c00,   1, 0x04, 0x0000001a },
+	{ 0x419c04,   1, 0x04, 0x80000006 },
+	{ 0x419c08,   1, 0x04, 0x00000002 },
+	{ 0x419c20,   1, 0x04, 0x00000000 },
+	{ 0x419c24,   1, 0x04, 0x00084210 },
+	{ 0x419c28,   1, 0x04, 0x3efbefbe },
+	{ 0x419c2c,   1, 0x04, 0x00000000 },
+	{ 0x419c34,   1, 0x04, 0x01ff1ff3 },
+	{ 0x419c3c,   1, 0x04, 0x00001919 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_l1c_0[] = {
+	{ 0x419c84,   1, 0x04, 0x00000020 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_sm_0[] = {
+	{ 0x419e04,   3, 0x04, 0x00000000 },
+	{ 0x419e10,   1, 0x04, 0x00001c02 },
+	{ 0x419e44,   1, 0x04, 0x00d3eff2 },
+	{ 0x419e48,   1, 0x04, 0x00000000 },
+	{ 0x419e4c,   1, 0x04, 0x0000007f },
+	{ 0x419e50,   1, 0x04, 0x00000000 },
+	{ 0x419e60,   4, 0x04, 0x00000000 },
+	{ 0x419e74,  10, 0x04, 0x00000000 },
+	{ 0x419eac,   1, 0x04, 0x0001cf8b },
+	{ 0x419eb0,   1, 0x04, 0x00030300 },
+	{ 0x419eb8,   1, 0x04, 0x00000000 },
+	{ 0x419ef0,  24, 0x04, 0x00000000 },
+	{ 0x419f68,   2, 0x04, 0x00000000 },
+	{ 0x419f70,   1, 0x04, 0x00000020 },
+	{ 0x419f78,   1, 0x04, 0x000003eb },
+	{ 0x419f7c,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+gm107_grctx_pack_tpc[] = {
+	{ nvd7_grctx_init_pe_0 },
+	{ gm107_grctx_init_tex_0 },
+	{ gm107_grctx_init_mpc_0 },
+	{ gm107_grctx_init_l1c_0 },
+	{ gm107_grctx_init_sm_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_cbm_0[] = {
+	{ 0x41bec0,   1, 0x04, 0x00000000 },
+	{ 0x41bec4,   1, 0x04, 0x01050000 },
+	{ 0x41bee4,   1, 0x04, 0x00000000 },
+	{ 0x41bef0,   1, 0x04, 0x000003ff },
+	{ 0x41bef4,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_grctx_init_wwdx_0[] = {
+	{ 0x41bf00,   1, 0x04, 0x0a418820 },
+	{ 0x41bf04,   1, 0x04, 0x062080e6 },
+	{ 0x41bf08,   1, 0x04, 0x020398a4 },
+	{ 0x41bf0c,   1, 0x04, 0x0e629062 },
+	{ 0x41bf10,   1, 0x04, 0x0a418820 },
+	{ 0x41bf14,   1, 0x04, 0x000000e6 },
+	{ 0x41bfd0,   1, 0x04, 0x00900103 },
+	{ 0x41bfe0,   1, 0x04, 0x80000000 },
+	{ 0x41bfe4,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+gm107_grctx_pack_ppc[] = {
+	{ nve4_grctx_init_pes_0 },
+	{ gm107_grctx_init_cbm_0 },
+	{ gm107_grctx_init_wwdx_0 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
+static void
+gm107_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
+{
+	mmio_data(0x003000, 0x0100, NV_MEM_ACCESS_RW | NV_MEM_ACCESS_SYS);
+	mmio_data(0x008000, 0x0100, NV_MEM_ACCESS_RW | NV_MEM_ACCESS_SYS);
+	mmio_data(0x200000, 0x1000, NV_MEM_ACCESS_RW);
+
+	mmio_list(0x40800c, 0x00000000,  8, 1);
+	mmio_list(0x408010, 0x80000000,  0, 0);
+	mmio_list(0x419004, 0x00000000,  8, 1);
+	mmio_list(0x419008, 0x00000000,  0, 0);
+	mmio_list(0x4064cc, 0x80000000,  0, 0);
+	mmio_list(0x418e30, 0x80000000,  0, 0);
+
+	mmio_list(0x408004, 0x00000000,  8, 0);
+	mmio_list(0x408008, 0x80000030,  0, 0);
+	mmio_list(0x418e24, 0x00000000,  8, 0);
+	mmio_list(0x418e28, 0x80000030,  0, 0);
+
+	mmio_list(0x4064c8, 0x018002c0,  0, 0);
+
+	mmio_list(0x418810, 0x80000000, 12, 2);
+	mmio_list(0x419848, 0x10000000, 12, 2);
+	mmio_list(0x419c2c, 0x10000000, 12, 2);
+
+	mmio_list(0x405830, 0x0aa01000,  0, 0);
+	mmio_list(0x4064c4, 0x0400ffff,  0, 0);
+
+	/*XXX*/
+	mmio_list(0x5030c0, 0x00001540,  0, 0);
+	mmio_list(0x5030f4, 0x00000000,  0, 0);
+	mmio_list(0x5030e4, 0x00002000,  0, 0);
+	mmio_list(0x5030f8, 0x00003fc0,  0, 0);
+	mmio_list(0x418ea0, 0x07151540,  0, 0);
+
+	mmio_list(0x5032c0, 0x00001540,  0, 0);
+	mmio_list(0x5032f4, 0x00001fe0,  0, 0);
+	mmio_list(0x5032e4, 0x00002000,  0, 0);
+	mmio_list(0x5032f8, 0x00006fc0,  0, 0);
+	mmio_list(0x418ea4, 0x07151540,  0, 0);
+}
+
+static void
+gm107_grctx_generate_tpcid(struct nvc0_graph_priv *priv)
+{
+	int gpc, tpc, id;
+
+	for (tpc = 0, id = 0; tpc < 4; tpc++) {
+		for (gpc = 0; gpc < priv->gpc_nr; gpc++) {
+			if (tpc < priv->tpc_nr[gpc]) {
+				nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x698), id);
+				nv_wr32(priv, GPC_UNIT(gpc, 0x0c10 + tpc * 4), id);
+				nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x088), id);
+				id++;
+			}
+
+			nv_wr32(priv, GPC_UNIT(gpc, 0x0c08), priv->tpc_nr[gpc]);
+			nv_wr32(priv, GPC_UNIT(gpc, 0x0c8c), priv->tpc_nr[gpc]);
+		}
+	}
+}
+
+static void
+gm107_grctx_generate_main(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
+{
+	struct nvc0_grctx_oclass *oclass = (void *)nv_engine(priv)->cclass;
+	int i;
+
+	nvc0_graph_mmio(priv, oclass->hub);
+	nvc0_graph_mmio(priv, oclass->gpc);
+	nvc0_graph_mmio(priv, oclass->zcull);
+	nvc0_graph_mmio(priv, oclass->tpc);
+	nvc0_graph_mmio(priv, oclass->ppc);
+
+	nv_wr32(priv, 0x404154, 0x00000000);
+
+	oclass->mods(priv, info);
+	oclass->unkn(priv);
+
+	gm107_grctx_generate_tpcid(priv);
+	nvc0_grctx_generate_r406028(priv);
+	nve4_grctx_generate_r418bb8(priv);
+	nvc0_grctx_generate_r406800(priv);
+
+	nv_wr32(priv, 0x4064d0, 0x00000001);
+	for (i = 1; i < 8; i++)
+		nv_wr32(priv, 0x4064d0 + (i * 0x04), 0x00000000);
+	nv_wr32(priv, 0x406500, 0x00000001);
+
+	nv_wr32(priv, 0x405b00, (priv->tpc_total << 8) | priv->gpc_nr);
+
+	if (priv->gpc_nr == 1) {
+		nv_mask(priv, 0x408850, 0x0000000f, priv->tpc_nr[0]);
+		nv_mask(priv, 0x408958, 0x0000000f, priv->tpc_nr[0]);
+	} else {
+		nv_mask(priv, 0x408850, 0x0000000f, priv->gpc_nr);
+		nv_mask(priv, 0x408958, 0x0000000f, priv->gpc_nr);
+	}
+
+	nvc0_graph_icmd(priv, oclass->icmd);
+	nv_wr32(priv, 0x404154, 0x00000400);
+	nvc0_graph_mthd(priv, oclass->mthd);
+
+	nv_mask(priv, 0x419e00, 0x00808080, 0x00808080);
+	nv_mask(priv, 0x419ccc, 0x80000000, 0x80000000);
+	nv_mask(priv, 0x419f80, 0x80000000, 0x80000000);
+	nv_mask(priv, 0x419f88, 0x80000000, 0x80000000);
+}
+
+struct nouveau_oclass *
+gm107_grctx_oclass = &(struct nvc0_grctx_oclass) {
+	.base.handle = NV_ENGCTX(GR, 0x08),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nvc0_graph_context_ctor,
+		.dtor = nvc0_graph_context_dtor,
+		.init = _nouveau_graph_context_init,
+		.fini = _nouveau_graph_context_fini,
+		.rd32 = _nouveau_graph_context_rd32,
+		.wr32 = _nouveau_graph_context_wr32,
+	},
+	.main  = gm107_grctx_generate_main,
+	.mods  = gm107_grctx_generate_mods,
+	.unkn  = nve4_grctx_generate_unkn,
+	.hub   = gm107_grctx_pack_hub,
+	.gpc   = gm107_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = gm107_grctx_pack_tpc,
+	.ppc   = gm107_grctx_pack_ppc,
+	.icmd  = gm107_grctx_pack_icmd,
+	.mthd  = gm107_grctx_pack_mthd,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv108.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv108.c
new file mode 100644
index 00000000000..8de4a429154
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv108.c
@@ -0,0 +1,599 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include "ctxnvc0.h"
+
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+nv108_grctx_init_icmd_0[] = {
+	{ 0x001000,   1, 0x01, 0x00000004 },
+	{ 0x000039,   3, 0x01, 0x00000000 },
+	{ 0x0000a9,   1, 0x01, 0x0000ffff },
+	{ 0x000038,   1, 0x01, 0x0fac6881 },
+	{ 0x00003d,   1, 0x01, 0x00000001 },
+	{ 0x0000e8,   8, 0x01, 0x00000400 },
+	{ 0x000078,   8, 0x01, 0x00000300 },
+	{ 0x000050,   1, 0x01, 0x00000011 },
+	{ 0x000058,   8, 0x01, 0x00000008 },
+	{ 0x000208,   8, 0x01, 0x00000001 },
+	{ 0x000081,   1, 0x01, 0x00000001 },
+	{ 0x000085,   1, 0x01, 0x00000004 },
+	{ 0x000088,   1, 0x01, 0x00000400 },
+	{ 0x000090,   1, 0x01, 0x00000300 },
+	{ 0x000098,   1, 0x01, 0x00001001 },
+	{ 0x0000e3,   1, 0x01, 0x00000001 },
+	{ 0x0000da,   1, 0x01, 0x00000001 },
+	{ 0x0000f8,   1, 0x01, 0x00000003 },
+	{ 0x0000fa,   1, 0x01, 0x00000001 },
+	{ 0x00009f,   4, 0x01, 0x0000ffff },
+	{ 0x0000b1,   1, 0x01, 0x00000001 },
+	{ 0x0000ad,   1, 0x01, 0x0000013e },
+	{ 0x0000e1,   1, 0x01, 0x00000010 },
+	{ 0x000290,  16, 0x01, 0x00000000 },
+	{ 0x0003b0,  16, 0x01, 0x00000000 },
+	{ 0x0002a0,  16, 0x01, 0x00000000 },
+	{ 0x000420,  16, 0x01, 0x00000000 },
+	{ 0x0002b0,  16, 0x01, 0x00000000 },
+	{ 0x000430,  16, 0x01, 0x00000000 },
+	{ 0x0002c0,  16, 0x01, 0x00000000 },
+	{ 0x0004d0,  16, 0x01, 0x00000000 },
+	{ 0x000720,  16, 0x01, 0x00000000 },
+	{ 0x0008c0,  16, 0x01, 0x00000000 },
+	{ 0x000890,  16, 0x01, 0x00000000 },
+	{ 0x0008e0,  16, 0x01, 0x00000000 },
+	{ 0x0008a0,  16, 0x01, 0x00000000 },
+	{ 0x0008f0,  16, 0x01, 0x00000000 },
+	{ 0x00094c,   1, 0x01, 0x000000ff },
+	{ 0x00094d,   1, 0x01, 0xffffffff },
+	{ 0x00094e,   1, 0x01, 0x00000002 },
+	{ 0x0002ec,   1, 0x01, 0x00000001 },
+	{ 0x0002f2,   2, 0x01, 0x00000001 },
+	{ 0x0002f5,   1, 0x01, 0x00000001 },
+	{ 0x0002f7,   1, 0x01, 0x00000001 },
+	{ 0x000303,   1, 0x01, 0x00000001 },
+	{ 0x0002e6,   1, 0x01, 0x00000001 },
+	{ 0x000466,   1, 0x01, 0x00000052 },
+	{ 0x000301,   1, 0x01, 0x3f800000 },
+	{ 0x000304,   1, 0x01, 0x30201000 },
+	{ 0x000305,   1, 0x01, 0x70605040 },
+	{ 0x000306,   1, 0x01, 0xb8a89888 },
+	{ 0x000307,   1, 0x01, 0xf8e8d8c8 },
+	{ 0x00030a,   1, 0x01, 0x00ffff00 },
+	{ 0x00030b,   1, 0x01, 0x0000001a },
+	{ 0x00030c,   1, 0x01, 0x00000001 },
+	{ 0x000318,   1, 0x01, 0x00000001 },
+	{ 0x000340,   1, 0x01, 0x00000000 },
+	{ 0x000375,   1, 0x01, 0x00000001 },
+	{ 0x00037d,   1, 0x01, 0x00000006 },
+	{ 0x0003a0,   1, 0x01, 0x00000002 },
+	{ 0x0003aa,   1, 0x01, 0x00000001 },
+	{ 0x0003a9,   1, 0x01, 0x00000001 },
+	{ 0x000380,   1, 0x01, 0x00000001 },
+	{ 0x000383,   1, 0x01, 0x00000011 },
+	{ 0x000360,   1, 0x01, 0x00000040 },
+	{ 0x000366,   2, 0x01, 0x00000000 },
+	{ 0x000368,   1, 0x01, 0x00000fff },
+	{ 0x000370,   2, 0x01, 0x00000000 },
+	{ 0x000372,   1, 0x01, 0x000fffff },
+	{ 0x00037a,   1, 0x01, 0x00000012 },
+	{ 0x000619,   1, 0x01, 0x00000003 },
+	{ 0x000811,   1, 0x01, 0x00000003 },
+	{ 0x000812,   1, 0x01, 0x00000004 },
+	{ 0x000813,   1, 0x01, 0x00000006 },
+	{ 0x000814,   1, 0x01, 0x00000008 },
+	{ 0x000815,   1, 0x01, 0x0000000b },
+	{ 0x000800,   6, 0x01, 0x00000001 },
+	{ 0x000632,   1, 0x01, 0x00000001 },
+	{ 0x000633,   1, 0x01, 0x00000002 },
+	{ 0x000634,   1, 0x01, 0x00000003 },
+	{ 0x000635,   1, 0x01, 0x00000004 },
+	{ 0x000654,   1, 0x01, 0x3f800000 },
+	{ 0x000657,   1, 0x01, 0x3f800000 },
+	{ 0x000655,   2, 0x01, 0x3f800000 },
+	{ 0x0006cd,   1, 0x01, 0x3f800000 },
+	{ 0x0007f5,   1, 0x01, 0x3f800000 },
+	{ 0x0007dc,   1, 0x01, 0x39291909 },
+	{ 0x0007dd,   1, 0x01, 0x79695949 },
+	{ 0x0007de,   1, 0x01, 0xb9a99989 },
+	{ 0x0007df,   1, 0x01, 0xf9e9d9c9 },
+	{ 0x0007e8,   1, 0x01, 0x00003210 },
+	{ 0x0007e9,   1, 0x01, 0x00007654 },
+	{ 0x0007ea,   1, 0x01, 0x00000098 },
+	{ 0x0007ec,   1, 0x01, 0x39291909 },
+	{ 0x0007ed,   1, 0x01, 0x79695949 },
+	{ 0x0007ee,   1, 0x01, 0xb9a99989 },
+	{ 0x0007ef,   1, 0x01, 0xf9e9d9c9 },
+	{ 0x0007f0,   1, 0x01, 0x00003210 },
+	{ 0x0007f1,   1, 0x01, 0x00007654 },
+	{ 0x0007f2,   1, 0x01, 0x00000098 },
+	{ 0x0005a5,   1, 0x01, 0x00000001 },
+	{ 0x000980, 128, 0x01, 0x00000000 },
+	{ 0x000468,   1, 0x01, 0x00000004 },
+	{ 0x00046c,   1, 0x01, 0x00000001 },
+	{ 0x000470,  96, 0x01, 0x00000000 },
+	{ 0x000510,  16, 0x01, 0x3f800000 },
+	{ 0x000520,   1, 0x01, 0x000002b6 },
+	{ 0x000529,   1, 0x01, 0x00000001 },
+	{ 0x000530,  16, 0x01, 0xffff0000 },
+	{ 0x000585,   1, 0x01, 0x0000003f },
+	{ 0x000576,   1, 0x01, 0x00000003 },
+	{ 0x00057b,   1, 0x01, 0x00000059 },
+	{ 0x000586,   1, 0x01, 0x00000040 },
+	{ 0x000582,   2, 0x01, 0x00000080 },
+	{ 0x0005c2,   1, 0x01, 0x00000001 },
+	{ 0x000638,   2, 0x01, 0x00000001 },
+	{ 0x00063a,   1, 0x01, 0x00000002 },
+	{ 0x00063b,   2, 0x01, 0x00000001 },
+	{ 0x00063d,   1, 0x01, 0x00000002 },
+	{ 0x00063e,   1, 0x01, 0x00000001 },
+	{ 0x0008b8,   8, 0x01, 0x00000001 },
+	{ 0x000900,   8, 0x01, 0x00000001 },
+	{ 0x000908,   8, 0x01, 0x00000002 },
+	{ 0x000910,  16, 0x01, 0x00000001 },
+	{ 0x000920,   8, 0x01, 0x00000002 },
+	{ 0x000928,   8, 0x01, 0x00000001 },
+	{ 0x000662,   1, 0x01, 0x00000001 },
+	{ 0x000648,   9, 0x01, 0x00000001 },
+	{ 0x000658,   1, 0x01, 0x0000000f },
+	{ 0x0007ff,   1, 0x01, 0x0000000a },
+	{ 0x00066a,   1, 0x01, 0x40000000 },
+	{ 0x00066b,   1, 0x01, 0x10000000 },
+	{ 0x00066c,   2, 0x01, 0xffff0000 },
+	{ 0x0007af,   2, 0x01, 0x00000008 },
+	{ 0x0007f6,   1, 0x01, 0x00000001 },
+	{ 0x00080b,   1, 0x01, 0x00000002 },
+	{ 0x0006b2,   1, 0x01, 0x00000055 },
+	{ 0x0007ad,   1, 0x01, 0x00000003 },
+	{ 0x000937,   1, 0x01, 0x00000001 },
+	{ 0x000971,   1, 0x01, 0x00000008 },
+	{ 0x000972,   1, 0x01, 0x00000040 },
+	{ 0x000973,   1, 0x01, 0x0000012c },
+	{ 0x00097c,   1, 0x01, 0x00000040 },
+	{ 0x000979,   1, 0x01, 0x00000003 },
+	{ 0x000975,   1, 0x01, 0x00000020 },
+	{ 0x000976,   1, 0x01, 0x00000001 },
+	{ 0x000977,   1, 0x01, 0x00000020 },
+	{ 0x000978,   1, 0x01, 0x00000001 },
+	{ 0x000957,   1, 0x01, 0x00000003 },
+	{ 0x00095e,   1, 0x01, 0x20164010 },
+	{ 0x00095f,   1, 0x01, 0x00000020 },
+	{ 0x000a0d,   1, 0x01, 0x00000006 },
+	{ 0x00097d,   1, 0x01, 0x00000020 },
+	{ 0x000683,   1, 0x01, 0x00000006 },
+	{ 0x000685,   1, 0x01, 0x003fffff },
+	{ 0x000687,   1, 0x01, 0x003fffff },
+	{ 0x0006a0,   1, 0x01, 0x00000005 },
+	{ 0x000840,   1, 0x01, 0x00400008 },
+	{ 0x000841,   1, 0x01, 0x08000080 },
+	{ 0x000842,   1, 0x01, 0x00400008 },
+	{ 0x000843,   1, 0x01, 0x08000080 },
+	{ 0x0006aa,   1, 0x01, 0x00000001 },
+	{ 0x0006ab,   1, 0x01, 0x00000002 },
+	{ 0x0006ac,   1, 0x01, 0x00000080 },
+	{ 0x0006ad,   2, 0x01, 0x00000100 },
+	{ 0x0006b1,   1, 0x01, 0x00000011 },
+	{ 0x0006bb,   1, 0x01, 0x000000cf },
+	{ 0x0006ce,   1, 0x01, 0x2a712488 },
+	{ 0x000739,   1, 0x01, 0x4085c000 },
+	{ 0x00073a,   1, 0x01, 0x00000080 },
+	{ 0x000786,   1, 0x01, 0x80000100 },
+	{ 0x00073c,   1, 0x01, 0x00010100 },
+	{ 0x00073d,   1, 0x01, 0x02800000 },
+	{ 0x000787,   1, 0x01, 0x000000cf },
+	{ 0x00078c,   1, 0x01, 0x00000008 },
+	{ 0x000792,   1, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
+	{ 0x000797,   1, 0x01, 0x000000cf },
+	{ 0x000836,   1, 0x01, 0x00000001 },
+	{ 0x00079a,   1, 0x01, 0x00000002 },
+	{ 0x000833,   1, 0x01, 0x04444480 },
+	{ 0x0007a1,   1, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
+	{ 0x000831,   1, 0x01, 0x00000004 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x000a04,   1, 0x01, 0x000000ff },
+	{ 0x000a0b,   1, 0x01, 0x00000040 },
+	{ 0x00097f,   1, 0x01, 0x00000100 },
+	{ 0x000a02,   1, 0x01, 0x00000001 },
+	{ 0x000809,   1, 0x01, 0x00000007 },
+	{ 0x00c221,   1, 0x01, 0x00000040 },
+	{ 0x00c1b0,   8, 0x01, 0x0000000f },
+	{ 0x00c1b8,   1, 0x01, 0x0fac6881 },
+	{ 0x00c1b9,   1, 0x01, 0x00fac688 },
+	{ 0x00c401,   1, 0x01, 0x00000001 },
+	{ 0x00c402,   1, 0x01, 0x00010001 },
+	{ 0x00c403,   2, 0x01, 0x00000001 },
+	{ 0x00c40e,   1, 0x01, 0x00000020 },
+	{ 0x00c500,   1, 0x01, 0x00000003 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000002 },
+	{ 0x0006aa,   1, 0x01, 0x00000001 },
+	{ 0x0006ad,   2, 0x01, 0x00000100 },
+	{ 0x0006b1,   1, 0x01, 0x00000011 },
+	{ 0x00078c,   1, 0x01, 0x00000008 },
+	{ 0x000792,   1, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
+	{ 0x000797,   1, 0x01, 0x000000cf },
+	{ 0x00079a,   1, 0x01, 0x00000002 },
+	{ 0x0007a1,   1, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
+	{ 0x000831,   1, 0x01, 0x00000004 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000008 },
+	{ 0x000039,   3, 0x01, 0x00000000 },
+	{ 0x000380,   1, 0x01, 0x00000001 },
+	{ 0x000366,   2, 0x01, 0x00000000 },
+	{ 0x000368,   1, 0x01, 0x00000fff },
+	{ 0x000370,   2, 0x01, 0x00000000 },
+	{ 0x000372,   1, 0x01, 0x000fffff },
+	{ 0x000813,   1, 0x01, 0x00000006 },
+	{ 0x000814,   1, 0x01, 0x00000008 },
+	{ 0x000957,   1, 0x01, 0x00000003 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x000a04,   1, 0x01, 0x000000ff },
+	{ 0x000a0b,   1, 0x01, 0x00000040 },
+	{ 0x00097f,   1, 0x01, 0x00000100 },
+	{ 0x000a02,   1, 0x01, 0x00000001 },
+	{ 0x000809,   1, 0x01, 0x00000007 },
+	{ 0x00c221,   1, 0x01, 0x00000040 },
+	{ 0x00c401,   1, 0x01, 0x00000001 },
+	{ 0x00c402,   1, 0x01, 0x00010001 },
+	{ 0x00c403,   2, 0x01, 0x00000001 },
+	{ 0x00c40e,   1, 0x01, 0x00000020 },
+	{ 0x00c500,   1, 0x01, 0x00000003 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000001 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+nv108_grctx_pack_icmd[] = {
+	{ nv108_grctx_init_icmd_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_fe_0[] = {
+	{ 0x404004,   8, 0x04, 0x00000000 },
+	{ 0x404024,   1, 0x04, 0x0000e000 },
+	{ 0x404028,   8, 0x04, 0x00000000 },
+	{ 0x4040a8,   8, 0x04, 0x00000000 },
+	{ 0x4040c8,   1, 0x04, 0xf800008f },
+	{ 0x4040d0,   6, 0x04, 0x00000000 },
+	{ 0x4040e8,   1, 0x04, 0x00001000 },
+	{ 0x4040f8,   1, 0x04, 0x00000000 },
+	{ 0x404100,  10, 0x04, 0x00000000 },
+	{ 0x404130,   2, 0x04, 0x00000000 },
+	{ 0x404138,   1, 0x04, 0x20000040 },
+	{ 0x404150,   1, 0x04, 0x0000002e },
+	{ 0x404154,   1, 0x04, 0x00000400 },
+	{ 0x404158,   1, 0x04, 0x00000200 },
+	{ 0x404164,   1, 0x04, 0x00000055 },
+	{ 0x40417c,   2, 0x04, 0x00000000 },
+	{ 0x404194,   1, 0x04, 0x01000700 },
+	{ 0x4041a0,   4, 0x04, 0x00000000 },
+	{ 0x404200,   1, 0x04, 0x0000a197 },
+	{ 0x404204,   1, 0x04, 0x0000a1c0 },
+	{ 0x404208,   1, 0x04, 0x0000a140 },
+	{ 0x40420c,   1, 0x04, 0x0000902d },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_ds_0[] = {
+	{ 0x405800,   1, 0x04, 0x0f8000bf },
+	{ 0x405830,   1, 0x04, 0x02180648 },
+	{ 0x405834,   1, 0x04, 0x08000000 },
+	{ 0x405838,   1, 0x04, 0x00000000 },
+	{ 0x405854,   1, 0x04, 0x00000000 },
+	{ 0x405870,   4, 0x04, 0x00000001 },
+	{ 0x405a00,   2, 0x04, 0x00000000 },
+	{ 0x405a18,   1, 0x04, 0x00000000 },
+	{ 0x405a1c,   1, 0x04, 0x000000ff },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_pd_0[] = {
+	{ 0x406020,   1, 0x04, 0x034103c1 },
+	{ 0x406028,   4, 0x04, 0x00000001 },
+	{ 0x4064a8,   1, 0x04, 0x00000000 },
+	{ 0x4064ac,   1, 0x04, 0x00003fff },
+	{ 0x4064b0,   3, 0x04, 0x00000000 },
+	{ 0x4064c0,   1, 0x04, 0x802000f0 },
+	{ 0x4064c4,   1, 0x04, 0x0192ffff },
+	{ 0x4064c8,   1, 0x04, 0x00c20200 },
+	{ 0x4064cc,   9, 0x04, 0x00000000 },
+	{ 0x4064fc,   1, 0x04, 0x0000022a },
+	{}
+};
+
+const struct nvc0_graph_init
+nv108_grctx_init_rstr2d_0[] = {
+	{ 0x407804,   1, 0x04, 0x00000063 },
+	{ 0x40780c,   1, 0x04, 0x0a418820 },
+	{ 0x407810,   1, 0x04, 0x062080e6 },
+	{ 0x407814,   1, 0x04, 0x020398a4 },
+	{ 0x407818,   1, 0x04, 0x0e629062 },
+	{ 0x40781c,   1, 0x04, 0x0a418820 },
+	{ 0x407820,   1, 0x04, 0x000000e6 },
+	{ 0x4078bc,   1, 0x04, 0x00000103 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_be_0[] = {
+	{ 0x408800,   1, 0x04, 0x32802a3c },
+	{ 0x408804,   1, 0x04, 0x00000040 },
+	{ 0x408808,   1, 0x04, 0x1003e005 },
+	{ 0x408840,   1, 0x04, 0x0000000b },
+	{ 0x408900,   1, 0x04, 0xb080b801 },
+	{ 0x408904,   1, 0x04, 0x62000001 },
+	{ 0x408908,   1, 0x04, 0x02c8102f },
+	{ 0x408980,   1, 0x04, 0x0000011d },
+	{}
+};
+
+static const struct nvc0_graph_pack
+nv108_grctx_pack_hub[] = {
+	{ nvc0_grctx_init_main_0 },
+	{ nv108_grctx_init_fe_0 },
+	{ nvf0_grctx_init_pri_0 },
+	{ nve4_grctx_init_memfmt_0 },
+	{ nv108_grctx_init_ds_0 },
+	{ nvf0_grctx_init_cwd_0 },
+	{ nv108_grctx_init_pd_0 },
+	{ nv108_grctx_init_rstr2d_0 },
+	{ nve4_grctx_init_scc_0 },
+	{ nv108_grctx_init_be_0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nv108_grctx_init_prop_0[] = {
+	{ 0x418400,   1, 0x04, 0x38005e00 },
+	{ 0x418404,   1, 0x04, 0x71e0ffff },
+	{ 0x41840c,   1, 0x04, 0x00001008 },
+	{ 0x418410,   1, 0x04, 0x0fff0fff },
+	{ 0x418414,   1, 0x04, 0x02200fff },
+	{ 0x418450,   6, 0x04, 0x00000000 },
+	{ 0x418468,   1, 0x04, 0x00000001 },
+	{ 0x41846c,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_gpc_unk_1[] = {
+	{ 0x418600,   1, 0x04, 0x0000007f },
+	{ 0x418684,   1, 0x04, 0x0000001f },
+	{ 0x418700,   1, 0x04, 0x00000002 },
+	{ 0x418704,   2, 0x04, 0x00000080 },
+	{ 0x41870c,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_setup_0[] = {
+	{ 0x418800,   1, 0x04, 0x7006863a },
+	{ 0x418808,   1, 0x04, 0x00000000 },
+	{ 0x41880c,   1, 0x04, 0x00000030 },
+	{ 0x418810,   1, 0x04, 0x00000000 },
+	{ 0x418828,   1, 0x04, 0x00000044 },
+	{ 0x418830,   1, 0x04, 0x10000001 },
+	{ 0x4188d8,   1, 0x04, 0x00000008 },
+	{ 0x4188e0,   1, 0x04, 0x01000000 },
+	{ 0x4188e8,   5, 0x04, 0x00000000 },
+	{ 0x4188fc,   1, 0x04, 0x20100058 },
+	{}
+};
+
+const struct nvc0_graph_init
+nv108_grctx_init_crstr_0[] = {
+	{ 0x418b00,   1, 0x04, 0x0000001e },
+	{ 0x418b08,   1, 0x04, 0x0a418820 },
+	{ 0x418b0c,   1, 0x04, 0x062080e6 },
+	{ 0x418b10,   1, 0x04, 0x020398a4 },
+	{ 0x418b14,   1, 0x04, 0x0e629062 },
+	{ 0x418b18,   1, 0x04, 0x0a418820 },
+	{ 0x418b1c,   1, 0x04, 0x000000e6 },
+	{ 0x418bb8,   1, 0x04, 0x00000103 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_gpm_0[] = {
+	{ 0x418c08,   1, 0x04, 0x00000001 },
+	{ 0x418c10,   8, 0x04, 0x00000000 },
+	{ 0x418c40,   1, 0x04, 0xffffffff },
+	{ 0x418c6c,   1, 0x04, 0x00000001 },
+	{ 0x418c80,   1, 0x04, 0x2020000c },
+	{ 0x418c8c,   1, 0x04, 0x00000001 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+nv108_grctx_pack_gpc[] = {
+	{ nvc0_grctx_init_gpc_unk_0 },
+	{ nv108_grctx_init_prop_0 },
+	{ nv108_grctx_init_gpc_unk_1 },
+	{ nv108_grctx_init_setup_0 },
+	{ nvc0_grctx_init_zcull_0 },
+	{ nv108_grctx_init_crstr_0 },
+	{ nv108_grctx_init_gpm_0 },
+	{ nvf0_grctx_init_gpc_unk_2 },
+	{ nvc0_grctx_init_gcc_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_tex_0[] = {
+	{ 0x419a00,   1, 0x04, 0x000100f0 },
+	{ 0x419a04,   1, 0x04, 0x00000001 },
+	{ 0x419a08,   1, 0x04, 0x00000421 },
+	{ 0x419a0c,   1, 0x04, 0x00120000 },
+	{ 0x419a10,   1, 0x04, 0x00000000 },
+	{ 0x419a14,   1, 0x04, 0x00000200 },
+	{ 0x419a1c,   1, 0x04, 0x0000c000 },
+	{ 0x419a20,   1, 0x04, 0x00000800 },
+	{ 0x419a30,   1, 0x04, 0x00000001 },
+	{ 0x419ac4,   1, 0x04, 0x0037f440 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_sm_0[] = {
+	{ 0x419e04,   1, 0x04, 0x00000000 },
+	{ 0x419e08,   1, 0x04, 0x0000001d },
+	{ 0x419e0c,   1, 0x04, 0x00000000 },
+	{ 0x419e10,   1, 0x04, 0x00001c02 },
+	{ 0x419e44,   1, 0x04, 0x0013eff2 },
+	{ 0x419e48,   1, 0x04, 0x00000000 },
+	{ 0x419e4c,   1, 0x04, 0x0000007f },
+	{ 0x419e50,   2, 0x04, 0x00000000 },
+	{ 0x419e58,   1, 0x04, 0x00000001 },
+	{ 0x419e5c,   3, 0x04, 0x00000000 },
+	{ 0x419e68,   1, 0x04, 0x00000002 },
+	{ 0x419e6c,  12, 0x04, 0x00000000 },
+	{ 0x419eac,   1, 0x04, 0x00001f8f },
+	{ 0x419eb0,   1, 0x04, 0x0db00d2f },
+	{ 0x419eb8,   1, 0x04, 0x00000000 },
+	{ 0x419ec8,   1, 0x04, 0x0001304f },
+	{ 0x419f30,   4, 0x04, 0x00000000 },
+	{ 0x419f40,   1, 0x04, 0x00000018 },
+	{ 0x419f44,   3, 0x04, 0x00000000 },
+	{ 0x419f58,   1, 0x04, 0x00000020 },
+	{ 0x419f70,   1, 0x04, 0x00000000 },
+	{ 0x419f78,   1, 0x04, 0x000001eb },
+	{ 0x419f7c,   1, 0x04, 0x00000404 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+nv108_grctx_pack_tpc[] = {
+	{ nvd7_grctx_init_pe_0 },
+	{ nv108_grctx_init_tex_0 },
+	{ nvf0_grctx_init_mpc_0 },
+	{ nvf0_grctx_init_l1c_0 },
+	{ nv108_grctx_init_sm_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_grctx_init_cbm_0[] = {
+	{ 0x41bec0,   1, 0x04, 0x10000000 },
+	{ 0x41bec4,   1, 0x04, 0x00037f7f },
+	{ 0x41bee4,   1, 0x04, 0x00000000 },
+	{ 0x41bef0,   1, 0x04, 0x000003ff },
+	{}
+};
+
+static const struct nvc0_graph_pack
+nv108_grctx_pack_ppc[] = {
+	{ nve4_grctx_init_pes_0 },
+	{ nv108_grctx_init_cbm_0 },
+	{ nvd7_grctx_init_wwdx_0 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
+static void
+nv108_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
+{
+	u32 magic[GPC_MAX][2];
+	u32 offset;
+	int gpc;
+
+	mmio_data(0x003000, 0x0100, NV_MEM_ACCESS_RW | NV_MEM_ACCESS_SYS);
+	mmio_data(0x008000, 0x0100, NV_MEM_ACCESS_RW | NV_MEM_ACCESS_SYS);
+	mmio_data(0x060000, 0x1000, NV_MEM_ACCESS_RW);
+	mmio_list(0x40800c, 0x00000000,  8, 1);
+	mmio_list(0x408010, 0x80000000,  0, 0);
+	mmio_list(0x419004, 0x00000000,  8, 1);
+	mmio_list(0x419008, 0x00000000,  0, 0);
+	mmio_list(0x4064cc, 0x80000000,  0, 0);
+	mmio_list(0x408004, 0x00000000,  8, 0);
+	mmio_list(0x408008, 0x80000030,  0, 0);
+	mmio_list(0x418808, 0x00000000,  8, 0);
+	mmio_list(0x41880c, 0x80000030,  0, 0);
+	mmio_list(0x4064c8, 0x00c20200,  0, 0);
+	mmio_list(0x418810, 0x80000000, 12, 2);
+	mmio_list(0x419848, 0x10000000, 12, 2);
+
+	mmio_list(0x405830, 0x02180648,  0, 0);
+	mmio_list(0x4064c4, 0x0192ffff,  0, 0);
+
+	for (gpc = 0, offset = 0; gpc < priv->gpc_nr; gpc++) {
+		u16 magic0 = 0x0218 * priv->tpc_nr[gpc];
+		u16 magic1 = 0x0648 * priv->tpc_nr[gpc];
+		magic[gpc][0]  = 0x10000000 | (magic0 << 16) | offset;
+		magic[gpc][1]  = 0x00000000 | (magic1 << 16);
+		offset += 0x0324 * priv->tpc_nr[gpc];
+	}
+
+	for (gpc = 0; gpc < priv->gpc_nr; gpc++) {
+		mmio_list(GPC_UNIT(gpc, 0x30c0), magic[gpc][0], 0, 0);
+		mmio_list(GPC_UNIT(gpc, 0x30e4), magic[gpc][1] | offset, 0, 0);
+		offset += 0x07ff * priv->tpc_nr[gpc];
+	}
+
+	mmio_list(0x17e91c, 0x0b040a0b, 0, 0);
+	mmio_list(0x17e920, 0x00090d08, 0, 0);
+}
+
+struct nouveau_oclass *
+nv108_grctx_oclass = &(struct nvc0_grctx_oclass) {
+	.base.handle = NV_ENGCTX(GR, 0x08),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nvc0_graph_context_ctor,
+		.dtor = nvc0_graph_context_dtor,
+		.init = _nouveau_graph_context_init,
+		.fini = _nouveau_graph_context_fini,
+		.rd32 = _nouveau_graph_context_rd32,
+		.wr32 = _nouveau_graph_context_wr32,
+	},
+	.main  = nve4_grctx_generate_main,
+	.mods  = nv108_grctx_generate_mods,
+	.unkn  = nve4_grctx_generate_unkn,
+	.hub   = nv108_grctx_pack_hub,
+	.gpc   = nv108_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nv108_grctx_pack_tpc,
+	.ppc   = nv108_grctx_pack_ppc,
+	.icmd  = nv108_grctx_pack_icmd,
+	.mthd  = nvf0_grctx_pack_mthd,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc0.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc0.c
index fe67415c3e1..833a96508c4 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc0.c
@@ -22,10 +22,14 @@
  * Authors: Ben Skeggs
  */
 
-#include "nvc0.h"
+#include "ctxnvc0.h"
 
-struct nvc0_graph_init
-nvc0_grctx_init_icmd[] = {
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+nvc0_grctx_init_icmd_0[] = {
 	{ 0x001000,   1, 0x01, 0x00000004 },
 	{ 0x0000a9,   1, 0x01, 0x0000ffff },
 	{ 0x000038,   1, 0x01, 0x0fac6881 },
@@ -140,8 +144,7 @@ nvc0_grctx_init_icmd[] = {
 	{ 0x000586,   1, 0x01, 0x00000040 },
 	{ 0x000582,   2, 0x01, 0x00000080 },
 	{ 0x0005c2,   1, 0x01, 0x00000001 },
-	{ 0x000638,   1, 0x01, 0x00000001 },
-	{ 0x000639,   1, 0x01, 0x00000001 },
+	{ 0x000638,   2, 0x01, 0x00000001 },
 	{ 0x00063a,   1, 0x01, 0x00000002 },
 	{ 0x00063b,   2, 0x01, 0x00000001 },
 	{ 0x00063d,   1, 0x01, 0x00000002 },
@@ -201,15 +204,13 @@ nvc0_grctx_init_icmd[] = {
 	{ 0x000787,   1, 0x01, 0x000000cf },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x000836,   1, 0x01, 0x00000001 },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x00080c,   1, 0x01, 0x00000002 },
 	{ 0x00080d,   2, 0x01, 0x00000100 },
@@ -235,14 +236,12 @@ nvc0_grctx_init_icmd[] = {
 	{ 0x0006b1,   1, 0x01, 0x00000011 },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x01e100,   1, 0x01, 0x00000001 },
 	{ 0x001000,   1, 0x01, 0x00000014 },
@@ -267,8 +266,14 @@ nvc0_grctx_init_icmd[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_9097[] = {
+const struct nvc0_graph_pack
+nvc0_grctx_pack_icmd[] = {
+	{ nvc0_grctx_init_icmd_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc0_grctx_init_9097_0[] = {
 	{ 0x000800,   8, 0x40, 0x00000000 },
 	{ 0x000804,   8, 0x40, 0x00000000 },
 	{ 0x000808,   8, 0x40, 0x00000400 },
@@ -516,8 +521,7 @@ nvc0_grctx_init_9097[] = {
 	{ 0x001350,   1, 0x04, 0x00000002 },
 	{ 0x001358,   1, 0x04, 0x00000001 },
 	{ 0x0012e4,   1, 0x04, 0x00000000 },
-	{ 0x00131c,   1, 0x04, 0x00000000 },
-	{ 0x001320,   3, 0x04, 0x00000000 },
+	{ 0x00131c,   4, 0x04, 0x00000000 },
 	{ 0x0019c0,   1, 0x04, 0x00000000 },
 	{ 0x001140,   1, 0x04, 0x00000000 },
 	{ 0x0019c4,   1, 0x04, 0x00000000 },
@@ -571,8 +575,8 @@ nvc0_grctx_init_9097[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_902d[] = {
+const struct nvc0_graph_init
+nvc0_grctx_init_902d_0[] = {
 	{ 0x000200,   1, 0x04, 0x000000cf },
 	{ 0x000204,   1, 0x04, 0x00000001 },
 	{ 0x000208,   1, 0x04, 0x00000020 },
@@ -590,8 +594,8 @@ nvc0_grctx_init_902d[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_9039[] = {
+const struct nvc0_graph_init
+nvc0_grctx_init_9039_0[] = {
 	{ 0x00030c,   3, 0x04, 0x00000000 },
 	{ 0x000320,   1, 0x04, 0x00000000 },
 	{ 0x000238,   2, 0x04, 0x00000000 },
@@ -599,8 +603,8 @@ nvc0_grctx_init_9039[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_90c0[] = {
+const struct nvc0_graph_init
+nvc0_grctx_init_90c0_0[] = {
 	{ 0x00270c,   8, 0x20, 0x00000000 },
 	{ 0x00030c,   1, 0x04, 0x00000001 },
 	{ 0x001944,   1, 0x04, 0x00000000 },
@@ -617,38 +621,44 @@ nvc0_grctx_init_90c0[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_base[] = {
+const struct nvc0_graph_pack
+nvc0_grctx_pack_mthd[] = {
+	{ nvc0_grctx_init_9097_0, 0x9097 },
+	{ nvc0_grctx_init_902d_0, 0x902d },
+	{ nvc0_grctx_init_9039_0, 0x9039 },
+	{ nvc0_grctx_init_90c0_0, 0x90c0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_main_0[] = {
 	{ 0x400204,   2, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_unk40xx[] = {
-	{ 0x404004,  10, 0x04, 0x00000000 },
+const struct nvc0_graph_init
+nvc0_grctx_init_fe_0[] = {
+	{ 0x404004,  11, 0x04, 0x00000000 },
 	{ 0x404044,   1, 0x04, 0x00000000 },
-	{ 0x404094,   1, 0x04, 0x00000000 },
-	{ 0x404098,  12, 0x04, 0x00000000 },
+	{ 0x404094,  13, 0x04, 0x00000000 },
 	{ 0x4040c8,   1, 0x04, 0xf0000087 },
 	{ 0x4040d0,   6, 0x04, 0x00000000 },
 	{ 0x4040e8,   1, 0x04, 0x00001000 },
 	{ 0x4040f8,   1, 0x04, 0x00000000 },
-	{ 0x404130,   1, 0x04, 0x00000000 },
-	{ 0x404134,   1, 0x04, 0x00000000 },
+	{ 0x404130,   2, 0x04, 0x00000000 },
 	{ 0x404138,   1, 0x04, 0x20000040 },
 	{ 0x404150,   1, 0x04, 0x0000002e },
 	{ 0x404154,   1, 0x04, 0x00000400 },
 	{ 0x404158,   1, 0x04, 0x00000200 },
 	{ 0x404164,   1, 0x04, 0x00000055 },
 	{ 0x404168,   1, 0x04, 0x00000000 },
-	{ 0x404174,   1, 0x04, 0x00000000 },
-	{ 0x404178,   2, 0x04, 0x00000000 },
+	{ 0x404174,   3, 0x04, 0x00000000 },
 	{ 0x404200,   8, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_unk44xx[] = {
+const struct nvc0_graph_init
+nvc0_grctx_init_pri_0[] = {
 	{ 0x404404,  14, 0x04, 0x00000000 },
 	{ 0x404460,   2, 0x04, 0x00000000 },
 	{ 0x404468,   1, 0x04, 0x00ffffff },
@@ -658,8 +668,8 @@ nvc0_grctx_init_unk44xx[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_unk46xx[] = {
+const struct nvc0_graph_init
+nvc0_grctx_init_memfmt_0[] = {
 	{ 0x404604,   1, 0x04, 0x00000015 },
 	{ 0x404608,   1, 0x04, 0x00000000 },
 	{ 0x40460c,   1, 0x04, 0x00002e00 },
@@ -674,19 +684,14 @@ nvc0_grctx_init_unk46xx[] = {
 	{ 0x4046a0,   1, 0x04, 0x007f0080 },
 	{ 0x4046a4,  18, 0x04, 0x00000000 },
 	{ 0x4046f0,   2, 0x04, 0x00000000 },
-	{}
-};
-
-struct nvc0_graph_init
-nvc0_grctx_init_unk47xx[] = {
 	{ 0x404700,  13, 0x04, 0x00000000 },
 	{ 0x404734,   1, 0x04, 0x00000100 },
 	{ 0x404738,   8, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_unk58xx[] = {
+static const struct nvc0_graph_init
+nvc0_grctx_init_ds_0[] = {
 	{ 0x405800,   1, 0x04, 0x078000bf },
 	{ 0x405830,   1, 0x04, 0x02180000 },
 	{ 0x405834,   2, 0x04, 0x00000000 },
@@ -697,23 +702,18 @@ nvc0_grctx_init_unk58xx[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_unk60xx[] = {
+static const struct nvc0_graph_init
+nvc0_grctx_init_pd_0[] = {
 	{ 0x406020,   1, 0x04, 0x000103c1 },
 	{ 0x406028,   4, 0x04, 0x00000001 },
-	{}
-};
-
-struct nvc0_graph_init
-nvc0_grctx_init_unk64xx[] = {
 	{ 0x4064a8,   1, 0x04, 0x00000000 },
 	{ 0x4064ac,   1, 0x04, 0x00003fff },
 	{ 0x4064b4,   2, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_unk78xx[] = {
+const struct nvc0_graph_init
+nvc0_grctx_init_rstr2d_0[] = {
 	{ 0x407804,   1, 0x04, 0x00000023 },
 	{ 0x40780c,   1, 0x04, 0x0a418820 },
 	{ 0x407810,   1, 0x04, 0x062080e6 },
@@ -725,8 +725,8 @@ nvc0_grctx_init_unk78xx[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_unk80xx[] = {
+const struct nvc0_graph_init
+nvc0_grctx_init_scc_0[] = {
 	{ 0x408000,   2, 0x04, 0x00000000 },
 	{ 0x408008,   1, 0x04, 0x00000018 },
 	{ 0x40800c,   2, 0x04, 0x00000000 },
@@ -736,8 +736,8 @@ nvc0_grctx_init_unk80xx[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_rop[] = {
+static const struct nvc0_graph_init
+nvc0_grctx_init_be_0[] = {
 	{ 0x408800,   1, 0x04, 0x02802a3c },
 	{ 0x408804,   1, 0x04, 0x00000040 },
 	{ 0x408808,   1, 0x04, 0x0003e00d },
@@ -748,9 +748,28 @@ nvc0_grctx_init_rop[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_gpc_0[] = {
+const struct nvc0_graph_pack
+nvc0_grctx_pack_hub[] = {
+	{ nvc0_grctx_init_main_0 },
+	{ nvc0_grctx_init_fe_0 },
+	{ nvc0_grctx_init_pri_0 },
+	{ nvc0_grctx_init_memfmt_0 },
+	{ nvc0_grctx_init_ds_0 },
+	{ nvc0_grctx_init_pd_0 },
+	{ nvc0_grctx_init_rstr2d_0 },
+	{ nvc0_grctx_init_scc_0 },
+	{ nvc0_grctx_init_be_0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_gpc_unk_0[] = {
 	{ 0x418380,   1, 0x04, 0x00000016 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_prop_0[] = {
 	{ 0x418400,   1, 0x04, 0x38004e00 },
 	{ 0x418404,   1, 0x04, 0x71e0ffff },
 	{ 0x418408,   1, 0x04, 0x00000000 },
@@ -760,6 +779,11 @@ nvc0_grctx_init_gpc_0[] = {
 	{ 0x418450,   6, 0x04, 0x00000000 },
 	{ 0x418468,   1, 0x04, 0x00000001 },
 	{ 0x41846c,   2, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_gpc_unk_1[] = {
 	{ 0x418600,   1, 0x04, 0x0000001f },
 	{ 0x418684,   1, 0x04, 0x0000000f },
 	{ 0x418700,   1, 0x04, 0x00000002 },
@@ -767,6 +791,11 @@ nvc0_grctx_init_gpc_0[] = {
 	{ 0x418708,   1, 0x04, 0x00000000 },
 	{ 0x41870c,   1, 0x04, 0x07c80000 },
 	{ 0x418710,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc0_grctx_init_setup_0[] = {
 	{ 0x418800,   1, 0x04, 0x0006860a },
 	{ 0x418808,   3, 0x04, 0x00000000 },
 	{ 0x418828,   1, 0x04, 0x00008442 },
@@ -775,10 +804,20 @@ nvc0_grctx_init_gpc_0[] = {
 	{ 0x4188e0,   1, 0x04, 0x01000000 },
 	{ 0x4188e8,   5, 0x04, 0x00000000 },
 	{ 0x4188fc,   1, 0x04, 0x00100000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_zcull_0[] = {
 	{ 0x41891c,   1, 0x04, 0x00ff00ff },
 	{ 0x418924,   1, 0x04, 0x00000000 },
 	{ 0x418928,   1, 0x04, 0x00ffff00 },
 	{ 0x41892c,   1, 0x04, 0x0000ff00 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_crstr_0[] = {
 	{ 0x418b00,   1, 0x04, 0x00000000 },
 	{ 0x418b08,   1, 0x04, 0x0a418820 },
 	{ 0x418b0c,   1, 0x04, 0x062080e6 },
@@ -787,18 +826,41 @@ nvc0_grctx_init_gpc_0[] = {
 	{ 0x418b18,   1, 0x04, 0x0a418820 },
 	{ 0x418b1c,   1, 0x04, 0x000000e6 },
 	{ 0x418bb8,   1, 0x04, 0x00000103 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_gpm_0[] = {
 	{ 0x418c08,   1, 0x04, 0x00000001 },
 	{ 0x418c10,   8, 0x04, 0x00000000 },
 	{ 0x418c80,   1, 0x04, 0x20200004 },
 	{ 0x418c8c,   1, 0x04, 0x00000001 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_gcc_0[] = {
 	{ 0x419000,   1, 0x04, 0x00000780 },
 	{ 0x419004,   2, 0x04, 0x00000000 },
 	{ 0x419014,   1, 0x04, 0x00000004 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_gpc_1[] = {
+const struct nvc0_graph_pack
+nvc0_grctx_pack_gpc[] = {
+	{ nvc0_grctx_init_gpc_unk_0 },
+	{ nvc0_grctx_init_prop_0 },
+	{ nvc0_grctx_init_gpc_unk_1 },
+	{ nvc0_grctx_init_setup_0 },
+	{ nvc0_grctx_init_zcull_0 },
+	{ nvc0_grctx_init_crstr_0 },
+	{ nvc0_grctx_init_gpm_0 },
+	{ nvc0_grctx_init_gcc_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc0_grctx_init_zcullr_0[] = {
 	{ 0x418a00,   3, 0x04, 0x00000000 },
 	{ 0x418a0c,   1, 0x04, 0x00010000 },
 	{ 0x418a10,   3, 0x04, 0x00000000 },
@@ -826,19 +888,35 @@ nvc0_grctx_init_gpc_1[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_grctx_init_tpc[] = {
+const struct nvc0_graph_pack
+nvc0_grctx_pack_zcull[] = {
+	{ nvc0_grctx_init_zcullr_0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_pe_0[] = {
 	{ 0x419818,   1, 0x04, 0x00000000 },
 	{ 0x41983c,   1, 0x04, 0x00038bc7 },
 	{ 0x419848,   1, 0x04, 0x00000000 },
 	{ 0x419864,   1, 0x04, 0x0000012a },
 	{ 0x419888,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc0_grctx_init_tex_0[] = {
 	{ 0x419a00,   1, 0x04, 0x000001f0 },
 	{ 0x419a04,   1, 0x04, 0x00000001 },
 	{ 0x419a08,   1, 0x04, 0x00000023 },
 	{ 0x419a0c,   1, 0x04, 0x00020000 },
 	{ 0x419a10,   1, 0x04, 0x00000000 },
 	{ 0x419a14,   1, 0x04, 0x00000200 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_wwdx_0[] = {
 	{ 0x419b00,   1, 0x04, 0x0a418820 },
 	{ 0x419b04,   1, 0x04, 0x062080e6 },
 	{ 0x419b08,   1, 0x04, 0x020398a4 },
@@ -848,15 +926,35 @@ nvc0_grctx_init_tpc[] = {
 	{ 0x419bd0,   1, 0x04, 0x00900103 },
 	{ 0x419be0,   1, 0x04, 0x00000001 },
 	{ 0x419be4,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_mpc_0[] = {
 	{ 0x419c00,   1, 0x04, 0x00000002 },
 	{ 0x419c04,   1, 0x04, 0x00000006 },
 	{ 0x419c08,   1, 0x04, 0x00000002 },
 	{ 0x419c20,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc0_grctx_init_l1c_0[] = {
 	{ 0x419cb0,   1, 0x04, 0x00060048 },
 	{ 0x419ce8,   1, 0x04, 0x00000000 },
 	{ 0x419cf4,   1, 0x04, 0x00000183 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_grctx_init_tpccs_0[] = {
 	{ 0x419d20,   1, 0x04, 0x02180000 },
 	{ 0x419d24,   1, 0x04, 0x00001fff },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc0_grctx_init_sm_0[] = {
 	{ 0x419e04,   3, 0x04, 0x00000000 },
 	{ 0x419e10,   1, 0x04, 0x00000002 },
 	{ 0x419e44,   1, 0x04, 0x001beff2 },
@@ -868,6 +966,22 @@ nvc0_grctx_init_tpc[] = {
 	{}
 };
 
+const struct nvc0_graph_pack
+nvc0_grctx_pack_tpc[] = {
+	{ nvc0_grctx_init_pe_0 },
+	{ nvc0_grctx_init_tex_0 },
+	{ nvc0_grctx_init_wwdx_0 },
+	{ nvc0_grctx_init_mpc_0 },
+	{ nvc0_grctx_init_l1c_0 },
+	{ nvc0_grctx_init_tpccs_0 },
+	{ nvc0_grctx_init_sm_0 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
 void
 nvc0_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 {
@@ -1055,14 +1169,14 @@ void
 nvc0_grctx_generate_main(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 {
 	struct nvc0_grctx_oclass *oclass = (void *)nv_engine(priv)->cclass;
-	int i;
 
 	nv_mask(priv, 0x000260, 0x00000001, 0x00000000);
 
-	for (i = 0; oclass->hub[i]; i++)
-		nvc0_graph_mmio(priv, oclass->hub[i]);
-	for (i = 0; oclass->gpc[i]; i++)
-		nvc0_graph_mmio(priv, oclass->gpc[i]);
+	nvc0_graph_mmio(priv, oclass->hub);
+	nvc0_graph_mmio(priv, oclass->gpc);
+	nvc0_graph_mmio(priv, oclass->zcull);
+	nvc0_graph_mmio(priv, oclass->tpc);
+	nvc0_graph_mmio(priv, oclass->ppc);
 
 	nv_wr32(priv, 0x404154, 0x00000000);
 
@@ -1182,46 +1296,6 @@ done:
 	return ret;
 }
 
-struct nvc0_graph_init *
-nvc0_grctx_init_hub[] = {
-	nvc0_grctx_init_base,
-	nvc0_grctx_init_unk40xx,
-	nvc0_grctx_init_unk44xx,
-	nvc0_grctx_init_unk46xx,
-	nvc0_grctx_init_unk47xx,
-	nvc0_grctx_init_unk58xx,
-	nvc0_grctx_init_unk60xx,
-	nvc0_grctx_init_unk64xx,
-	nvc0_grctx_init_unk78xx,
-	nvc0_grctx_init_unk80xx,
-	nvc0_grctx_init_rop,
-	NULL
-};
-
-static struct nvc0_graph_init *
-nvc0_grctx_init_gpc[] = {
-	nvc0_grctx_init_gpc_0,
-	nvc0_grctx_init_gpc_1,
-	nvc0_grctx_init_tpc,
-	NULL
-};
-
-struct nvc0_graph_init
-nvc0_grctx_init_mthd_magic[] = {
-	{ 0x3410, 1, 0x04, 0x00000000 },
-	{}
-};
-
-struct nvc0_graph_mthd
-nvc0_grctx_init_mthd[] = {
-	{ 0x9097, nvc0_grctx_init_9097, },
-	{ 0x902d, nvc0_grctx_init_902d, },
-	{ 0x9039, nvc0_grctx_init_9039, },
-	{ 0x90c0, nvc0_grctx_init_90c0, },
-	{ 0x902d, nvc0_grctx_init_mthd_magic, },
-	{}
-};
-
 struct nouveau_oclass *
 nvc0_grctx_oclass = &(struct nvc0_grctx_oclass) {
 	.base.handle = NV_ENGCTX(GR, 0xc0),
@@ -1233,11 +1307,13 @@ nvc0_grctx_oclass = &(struct nvc0_grctx_oclass) {
 		.rd32 = _nouveau_graph_context_rd32,
 		.wr32 = _nouveau_graph_context_wr32,
 	},
-	.main = nvc0_grctx_generate_main,
-	.mods = nvc0_grctx_generate_mods,
-	.unkn = nvc0_grctx_generate_unkn,
-	.hub  = nvc0_grctx_init_hub,
-	.gpc  = nvc0_grctx_init_gpc,
-	.icmd = nvc0_grctx_init_icmd,
-	.mthd = nvc0_grctx_init_mthd,
+	.main  = nvc0_grctx_generate_main,
+	.mods  = nvc0_grctx_generate_mods,
+	.unkn  = nvc0_grctx_generate_unkn,
+	.hub   = nvc0_grctx_pack_hub,
+	.gpc   = nvc0_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nvc0_grctx_pack_tpc,
+	.icmd  = nvc0_grctx_pack_icmd,
+	.mthd  = nvc0_grctx_pack_mthd,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc0.h b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc0.h
new file mode 100644
index 00000000000..8da8b627b9d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc0.h
@@ -0,0 +1,179 @@
+#ifndef __NVKM_GRCTX_NVC0_H__
+#define __NVKM_GRCTX_NVC0_H__
+
+#include "nvc0.h"
+
+struct nvc0_grctx {
+	struct nvc0_graph_priv *priv;
+	struct nvc0_graph_data *data;
+	struct nvc0_graph_mmio *mmio;
+	int buffer_nr;
+	u64 buffer[4];
+	u64 addr;
+};
+
+struct nvc0_grctx_oclass {
+	struct nouveau_oclass base;
+	/* main context generation function */
+	void  (*main)(struct nvc0_graph_priv *, struct nvc0_grctx *);
+	/* context-specific modify-on-first-load list generation function */
+	void  (*mods)(struct nvc0_graph_priv *, struct nvc0_grctx *);
+	void  (*unkn)(struct nvc0_graph_priv *);
+	/* mmio context data */
+	const struct nvc0_graph_pack *hub;
+	const struct nvc0_graph_pack *gpc;
+	const struct nvc0_graph_pack *zcull;
+	const struct nvc0_graph_pack *tpc;
+	const struct nvc0_graph_pack *ppc;
+	/* indirect context data, generated with icmds/mthds */
+	const struct nvc0_graph_pack *icmd;
+	const struct nvc0_graph_pack *mthd;
+};
+
+#define mmio_data(s,a,p) do {                                                  \
+	info->buffer[info->buffer_nr] = round_up(info->addr, (a));             \
+	info->addr = info->buffer[info->buffer_nr++] + (s);                    \
+	info->data->size = (s);                                                \
+	info->data->align = (a);                                               \
+	info->data->access = (p);                                              \
+	info->data++;                                                          \
+} while(0)
+
+#define mmio_list(r,d,s,b) do {                                                \
+	info->mmio->addr = (r);                                                \
+	info->mmio->data = (d);                                                \
+	info->mmio->shift = (s);                                               \
+	info->mmio->buffer = (b);                                              \
+	info->mmio++;                                                          \
+	nv_wr32(priv, (r), (d) | ((s) ? (info->buffer[(b)] >> (s)) : 0));      \
+} while(0)
+
+extern struct nouveau_oclass *nvc0_grctx_oclass;
+int  nvc0_grctx_generate(struct nvc0_graph_priv *);
+void nvc0_grctx_generate_main(struct nvc0_graph_priv *, struct nvc0_grctx *);
+void nvc0_grctx_generate_mods(struct nvc0_graph_priv *, struct nvc0_grctx *);
+void nvc0_grctx_generate_unkn(struct nvc0_graph_priv *);
+void nvc0_grctx_generate_tpcid(struct nvc0_graph_priv *);
+void nvc0_grctx_generate_r406028(struct nvc0_graph_priv *);
+void nvc0_grctx_generate_r4060a8(struct nvc0_graph_priv *);
+void nvc0_grctx_generate_r418bb8(struct nvc0_graph_priv *);
+void nvc0_grctx_generate_r406800(struct nvc0_graph_priv *);
+
+extern struct nouveau_oclass *nvc1_grctx_oclass;
+void nvc1_grctx_generate_mods(struct nvc0_graph_priv *, struct nvc0_grctx *);
+void nvc1_grctx_generate_unkn(struct nvc0_graph_priv *);
+
+extern struct nouveau_oclass *nvc4_grctx_oclass;
+extern struct nouveau_oclass *nvc8_grctx_oclass;
+extern struct nouveau_oclass *nvd7_grctx_oclass;
+extern struct nouveau_oclass *nvd9_grctx_oclass;
+
+extern struct nouveau_oclass *nve4_grctx_oclass;
+extern struct nouveau_oclass *gk20a_grctx_oclass;
+void nve4_grctx_generate_main(struct nvc0_graph_priv *, struct nvc0_grctx *);
+void nve4_grctx_generate_mods(struct nvc0_graph_priv *, struct nvc0_grctx *);
+void nve4_grctx_generate_unkn(struct nvc0_graph_priv *);
+void nve4_grctx_generate_r418bb8(struct nvc0_graph_priv *);
+
+extern struct nouveau_oclass *nvf0_grctx_oclass;
+extern struct nouveau_oclass *nv108_grctx_oclass;
+extern struct nouveau_oclass *gm107_grctx_oclass;
+
+/* context init value lists */
+
+extern const struct nvc0_graph_pack nvc0_grctx_pack_icmd[];
+
+extern const struct nvc0_graph_pack nvc0_grctx_pack_mthd[];
+extern const struct nvc0_graph_init nvc0_grctx_init_902d_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_9039_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_90c0_0[];
+
+extern const struct nvc0_graph_pack nvc0_grctx_pack_hub[];
+extern const struct nvc0_graph_init nvc0_grctx_init_main_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_fe_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_pri_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_memfmt_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_rstr2d_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_scc_0[];
+
+extern const struct nvc0_graph_pack nvc0_grctx_pack_gpc[];
+extern const struct nvc0_graph_init nvc0_grctx_init_gpc_unk_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_prop_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_gpc_unk_1[];
+extern const struct nvc0_graph_init nvc0_grctx_init_zcull_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_crstr_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_gpm_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_gcc_0[];
+
+extern const struct nvc0_graph_pack nvc0_grctx_pack_zcull[];
+
+extern const struct nvc0_graph_pack nvc0_grctx_pack_tpc[];
+extern const struct nvc0_graph_init nvc0_grctx_init_pe_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_wwdx_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_mpc_0[];
+extern const struct nvc0_graph_init nvc0_grctx_init_tpccs_0[];
+
+extern const struct nvc0_graph_init nvc4_grctx_init_tex_0[];
+extern const struct nvc0_graph_init nvc4_grctx_init_l1c_0[];
+extern const struct nvc0_graph_init nvc4_grctx_init_sm_0[];
+
+extern const struct nvc0_graph_init nvc1_grctx_init_9097_0[];
+
+extern const struct nvc0_graph_init nvc1_grctx_init_gpm_0[];
+
+extern const struct nvc0_graph_init nvc1_grctx_init_pe_0[];
+extern const struct nvc0_graph_init nvc1_grctx_init_wwdx_0[];
+extern const struct nvc0_graph_init nvc1_grctx_init_tpccs_0[];
+
+extern const struct nvc0_graph_init nvc8_grctx_init_9197_0[];
+extern const struct nvc0_graph_init nvc8_grctx_init_9297_0[];
+
+extern const struct nvc0_graph_pack nvd9_grctx_pack_icmd[];
+
+extern const struct nvc0_graph_pack nvd9_grctx_pack_mthd[];
+
+extern const struct nvc0_graph_init nvd9_grctx_init_fe_0[];
+extern const struct nvc0_graph_init nvd9_grctx_init_be_0[];
+
+extern const struct nvc0_graph_init nvd9_grctx_init_prop_0[];
+extern const struct nvc0_graph_init nvd9_grctx_init_gpc_unk_1[];
+extern const struct nvc0_graph_init nvd9_grctx_init_crstr_0[];
+
+extern const struct nvc0_graph_init nvd9_grctx_init_sm_0[];
+
+extern const struct nvc0_graph_init nvd7_grctx_init_pe_0[];
+
+extern const struct nvc0_graph_init nvd7_grctx_init_wwdx_0[];
+
+extern const struct nvc0_graph_init nve4_grctx_init_memfmt_0[];
+extern const struct nvc0_graph_init nve4_grctx_init_ds_0[];
+extern const struct nvc0_graph_init nve4_grctx_init_scc_0[];
+
+extern const struct nvc0_graph_init nve4_grctx_init_gpm_0[];
+
+extern const struct nvc0_graph_init nve4_grctx_init_pes_0[];
+
+extern const struct nvc0_graph_pack nve4_grctx_pack_hub[];
+extern const struct nvc0_graph_pack nve4_grctx_pack_gpc[];
+extern const struct nvc0_graph_pack nve4_grctx_pack_tpc[];
+extern const struct nvc0_graph_pack nve4_grctx_pack_ppc[];
+extern const struct nvc0_graph_pack nve4_grctx_pack_icmd[];
+extern const struct nvc0_graph_init nve4_grctx_init_a097_0[];
+
+extern const struct nvc0_graph_pack nvf0_grctx_pack_mthd[];
+
+extern const struct nvc0_graph_init nvf0_grctx_init_pri_0[];
+extern const struct nvc0_graph_init nvf0_grctx_init_cwd_0[];
+
+extern const struct nvc0_graph_init nvf0_grctx_init_gpc_unk_2[];
+
+extern const struct nvc0_graph_init nvf0_grctx_init_mpc_0[];
+extern const struct nvc0_graph_init nvf0_grctx_init_l1c_0[];
+
+extern const struct nvc0_graph_init nv108_grctx_init_rstr2d_0[];
+
+extern const struct nvc0_graph_init nv108_grctx_init_prop_0[];
+extern const struct nvc0_graph_init nv108_grctx_init_crstr_0[];
+
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc1.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc1.c
index 71b4283f7fa..24a92c569c0 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc1.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc1.c
@@ -22,10 +22,14 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 
-#include "nvc0.h"
+#include "ctxnvc0.h"
 
-static struct nvc0_graph_init
-nvc1_grctx_init_icmd[] = {
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+nvc1_grctx_init_icmd_0[] = {
 	{ 0x001000,   1, 0x01, 0x00000004 },
 	{ 0x0000a9,   1, 0x01, 0x0000ffff },
 	{ 0x000038,   1, 0x01, 0x0fac6881 },
@@ -141,8 +145,7 @@ nvc1_grctx_init_icmd[] = {
 	{ 0x000586,   1, 0x01, 0x00000040 },
 	{ 0x000582,   2, 0x01, 0x00000080 },
 	{ 0x0005c2,   1, 0x01, 0x00000001 },
-	{ 0x000638,   1, 0x01, 0x00000001 },
-	{ 0x000639,   1, 0x01, 0x00000001 },
+	{ 0x000638,   2, 0x01, 0x00000001 },
 	{ 0x00063a,   1, 0x01, 0x00000002 },
 	{ 0x00063b,   2, 0x01, 0x00000001 },
 	{ 0x00063d,   1, 0x01, 0x00000002 },
@@ -202,15 +205,13 @@ nvc1_grctx_init_icmd[] = {
 	{ 0x000787,   1, 0x01, 0x000000cf },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x000836,   1, 0x01, 0x00000001 },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x00080c,   1, 0x01, 0x00000002 },
 	{ 0x00080d,   2, 0x01, 0x00000100 },
@@ -236,14 +237,12 @@ nvc1_grctx_init_icmd[] = {
 	{ 0x0006b1,   1, 0x01, 0x00000011 },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x01e100,   1, 0x01, 0x00000001 },
 	{ 0x001000,   1, 0x01, 0x00000014 },
@@ -268,8 +267,14 @@ nvc1_grctx_init_icmd[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc1_grctx_init_9097[] = {
+static const struct nvc0_graph_pack
+nvc1_grctx_pack_icmd[] = {
+	{ nvc1_grctx_init_icmd_0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc1_grctx_init_9097_0[] = {
 	{ 0x000800,   8, 0x40, 0x00000000 },
 	{ 0x000804,   8, 0x40, 0x00000000 },
 	{ 0x000808,   8, 0x40, 0x00000400 },
@@ -516,8 +521,7 @@ nvc1_grctx_init_9097[] = {
 	{ 0x001350,   1, 0x04, 0x00000002 },
 	{ 0x001358,   1, 0x04, 0x00000001 },
 	{ 0x0012e4,   1, 0x04, 0x00000000 },
-	{ 0x00131c,   1, 0x04, 0x00000000 },
-	{ 0x001320,   3, 0x04, 0x00000000 },
+	{ 0x00131c,   4, 0x04, 0x00000000 },
 	{ 0x0019c0,   1, 0x04, 0x00000000 },
 	{ 0x001140,   1, 0x04, 0x00000000 },
 	{ 0x0019c4,   1, 0x04, 0x00000000 },
@@ -571,15 +575,25 @@ nvc1_grctx_init_9097[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvc1_grctx_init_9197[] = {
+static const struct nvc0_graph_init
+nvc1_grctx_init_9197_0[] = {
 	{ 0x003400, 128, 0x04, 0x00000000 },
 	{ 0x0002e4,   1, 0x04, 0x0000b001 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvc1_grctx_init_unk58xx[] = {
+static const struct nvc0_graph_pack
+nvc1_grctx_pack_mthd[] = {
+	{ nvc1_grctx_init_9097_0, 0x9097 },
+	{ nvc1_grctx_init_9197_0, 0x9197 },
+	{ nvc0_grctx_init_902d_0, 0x902d },
+	{ nvc0_grctx_init_9039_0, 0x9039 },
+	{ nvc0_grctx_init_90c0_0, 0x90c0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc1_grctx_init_ds_0[] = {
 	{ 0x405800,   1, 0x04, 0x0f8000bf },
 	{ 0x405830,   1, 0x04, 0x02180218 },
 	{ 0x405834,   2, 0x04, 0x00000000 },
@@ -590,8 +604,20 @@ nvc1_grctx_init_unk58xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvc1_grctx_init_rop[] = {
+static const struct nvc0_graph_init
+nvc1_grctx_init_pd_0[] = {
+	{ 0x406020,   1, 0x04, 0x000103c1 },
+	{ 0x406028,   4, 0x04, 0x00000001 },
+	{ 0x4064a8,   1, 0x04, 0x00000000 },
+	{ 0x4064ac,   1, 0x04, 0x00003fff },
+	{ 0x4064b4,   2, 0x04, 0x00000000 },
+	{ 0x4064c0,   1, 0x04, 0x80140078 },
+	{ 0x4064c4,   1, 0x04, 0x0086ffff },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc1_grctx_init_be_0[] = {
 	{ 0x408800,   1, 0x04, 0x02802a3c },
 	{ 0x408804,   1, 0x04, 0x00000040 },
 	{ 0x408808,   1, 0x04, 0x1003e005 },
@@ -602,25 +628,22 @@ nvc1_grctx_init_rop[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvc1_grctx_init_gpc_0[] = {
-	{ 0x418380,   1, 0x04, 0x00000016 },
-	{ 0x418400,   1, 0x04, 0x38004e00 },
-	{ 0x418404,   1, 0x04, 0x71e0ffff },
-	{ 0x418408,   1, 0x04, 0x00000000 },
-	{ 0x41840c,   1, 0x04, 0x00001008 },
-	{ 0x418410,   1, 0x04, 0x0fff0fff },
-	{ 0x418414,   1, 0x04, 0x00200fff },
-	{ 0x418450,   6, 0x04, 0x00000000 },
-	{ 0x418468,   1, 0x04, 0x00000001 },
-	{ 0x41846c,   2, 0x04, 0x00000000 },
-	{ 0x418600,   1, 0x04, 0x0000001f },
-	{ 0x418684,   1, 0x04, 0x0000000f },
-	{ 0x418700,   1, 0x04, 0x00000002 },
-	{ 0x418704,   1, 0x04, 0x00000080 },
-	{ 0x418708,   1, 0x04, 0x00000000 },
-	{ 0x41870c,   1, 0x04, 0x07c80000 },
-	{ 0x418710,   1, 0x04, 0x00000000 },
+static const struct nvc0_graph_pack
+nvc1_grctx_pack_hub[] = {
+	{ nvc0_grctx_init_main_0 },
+	{ nvc0_grctx_init_fe_0 },
+	{ nvc0_grctx_init_pri_0 },
+	{ nvc0_grctx_init_memfmt_0 },
+	{ nvc1_grctx_init_ds_0 },
+	{ nvc1_grctx_init_pd_0 },
+	{ nvc0_grctx_init_rstr2d_0 },
+	{ nvc0_grctx_init_scc_0 },
+	{ nvc1_grctx_init_be_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc1_grctx_init_setup_0[] = {
 	{ 0x418800,   1, 0x04, 0x0006860a },
 	{ 0x418808,   3, 0x04, 0x00000000 },
 	{ 0x418828,   1, 0x04, 0x00008442 },
@@ -629,69 +652,44 @@ nvc1_grctx_init_gpc_0[] = {
 	{ 0x4188e0,   1, 0x04, 0x01000000 },
 	{ 0x4188e8,   5, 0x04, 0x00000000 },
 	{ 0x4188fc,   1, 0x04, 0x00100018 },
-	{ 0x41891c,   1, 0x04, 0x00ff00ff },
-	{ 0x418924,   1, 0x04, 0x00000000 },
-	{ 0x418928,   1, 0x04, 0x00ffff00 },
-	{ 0x41892c,   1, 0x04, 0x0000ff00 },
-	{ 0x418a00,   3, 0x04, 0x00000000 },
-	{ 0x418a0c,   1, 0x04, 0x00010000 },
-	{ 0x418a10,   3, 0x04, 0x00000000 },
-	{ 0x418a20,   3, 0x04, 0x00000000 },
-	{ 0x418a2c,   1, 0x04, 0x00010000 },
-	{ 0x418a30,   3, 0x04, 0x00000000 },
-	{ 0x418a40,   3, 0x04, 0x00000000 },
-	{ 0x418a4c,   1, 0x04, 0x00010000 },
-	{ 0x418a50,   3, 0x04, 0x00000000 },
-	{ 0x418a60,   3, 0x04, 0x00000000 },
-	{ 0x418a6c,   1, 0x04, 0x00010000 },
-	{ 0x418a70,   3, 0x04, 0x00000000 },
-	{ 0x418a80,   3, 0x04, 0x00000000 },
-	{ 0x418a8c,   1, 0x04, 0x00010000 },
-	{ 0x418a90,   3, 0x04, 0x00000000 },
-	{ 0x418aa0,   3, 0x04, 0x00000000 },
-	{ 0x418aac,   1, 0x04, 0x00010000 },
-	{ 0x418ab0,   3, 0x04, 0x00000000 },
-	{ 0x418ac0,   3, 0x04, 0x00000000 },
-	{ 0x418acc,   1, 0x04, 0x00010000 },
-	{ 0x418ad0,   3, 0x04, 0x00000000 },
-	{ 0x418ae0,   3, 0x04, 0x00000000 },
-	{ 0x418aec,   1, 0x04, 0x00010000 },
-	{ 0x418af0,   3, 0x04, 0x00000000 },
-	{ 0x418b00,   1, 0x04, 0x00000000 },
-	{ 0x418b08,   1, 0x04, 0x0a418820 },
-	{ 0x418b0c,   1, 0x04, 0x062080e6 },
-	{ 0x418b10,   1, 0x04, 0x020398a4 },
-	{ 0x418b14,   1, 0x04, 0x0e629062 },
-	{ 0x418b18,   1, 0x04, 0x0a418820 },
-	{ 0x418b1c,   1, 0x04, 0x000000e6 },
-	{ 0x418bb8,   1, 0x04, 0x00000103 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc1_grctx_init_gpm_0[] = {
 	{ 0x418c08,   1, 0x04, 0x00000001 },
 	{ 0x418c10,   8, 0x04, 0x00000000 },
 	{ 0x418c6c,   1, 0x04, 0x00000001 },
 	{ 0x418c80,   1, 0x04, 0x20200004 },
 	{ 0x418c8c,   1, 0x04, 0x00000001 },
-	{ 0x419000,   1, 0x04, 0x00000780 },
-	{ 0x419004,   2, 0x04, 0x00000000 },
-	{ 0x419014,   1, 0x04, 0x00000004 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvc1_grctx_init_tpc[] = {
+static const struct nvc0_graph_pack
+nvc1_grctx_pack_gpc[] = {
+	{ nvc0_grctx_init_gpc_unk_0 },
+	{ nvc0_grctx_init_prop_0 },
+	{ nvc0_grctx_init_gpc_unk_1 },
+	{ nvc1_grctx_init_setup_0 },
+	{ nvc0_grctx_init_zcull_0 },
+	{ nvc0_grctx_init_crstr_0 },
+	{ nvc1_grctx_init_gpm_0 },
+	{ nvc0_grctx_init_gcc_0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc1_grctx_init_pe_0[] = {
 	{ 0x419818,   1, 0x04, 0x00000000 },
 	{ 0x41983c,   1, 0x04, 0x00038bc7 },
 	{ 0x419848,   1, 0x04, 0x00000000 },
 	{ 0x419864,   1, 0x04, 0x00000129 },
 	{ 0x419888,   1, 0x04, 0x00000000 },
-	{ 0x419a00,   1, 0x04, 0x000001f0 },
-	{ 0x419a04,   1, 0x04, 0x00000001 },
-	{ 0x419a08,   1, 0x04, 0x00000023 },
-	{ 0x419a0c,   1, 0x04, 0x00020000 },
-	{ 0x419a10,   1, 0x04, 0x00000000 },
-	{ 0x419a14,   1, 0x04, 0x00000200 },
-	{ 0x419a1c,   1, 0x04, 0x00000000 },
-	{ 0x419a20,   1, 0x04, 0x00000800 },
-	{ 0x419ac4,   1, 0x04, 0x0007f440 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc1_grctx_init_wwdx_0[] = {
 	{ 0x419b00,   1, 0x04, 0x0a418820 },
 	{ 0x419b04,   1, 0x04, 0x062080e6 },
 	{ 0x419b08,   1, 0x04, 0x020398a4 },
@@ -701,28 +699,33 @@ nvc1_grctx_init_tpc[] = {
 	{ 0x419bd0,   1, 0x04, 0x00900103 },
 	{ 0x419be0,   1, 0x04, 0x00400001 },
 	{ 0x419be4,   1, 0x04, 0x00000000 },
-	{ 0x419c00,   1, 0x04, 0x00000002 },
-	{ 0x419c04,   1, 0x04, 0x00000006 },
-	{ 0x419c08,   1, 0x04, 0x00000002 },
-	{ 0x419c20,   1, 0x04, 0x00000000 },
-	{ 0x419cb0,   1, 0x04, 0x00020048 },
-	{ 0x419ce8,   1, 0x04, 0x00000000 },
-	{ 0x419cf4,   1, 0x04, 0x00000183 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc1_grctx_init_tpccs_0[] = {
 	{ 0x419d20,   1, 0x04, 0x12180000 },
 	{ 0x419d24,   1, 0x04, 0x00001fff },
 	{ 0x419d44,   1, 0x04, 0x02180218 },
-	{ 0x419e04,   3, 0x04, 0x00000000 },
-	{ 0x419e10,   1, 0x04, 0x00000002 },
-	{ 0x419e44,   1, 0x04, 0x001beff2 },
-	{ 0x419e48,   1, 0x04, 0x00000000 },
-	{ 0x419e4c,   1, 0x04, 0x0000000f },
-	{ 0x419e50,  17, 0x04, 0x00000000 },
-	{ 0x419e98,   1, 0x04, 0x00000000 },
-	{ 0x419ee0,   1, 0x04, 0x00011110 },
-	{ 0x419f30,  11, 0x04, 0x00000000 },
 	{}
 };
 
+static const struct nvc0_graph_pack
+nvc1_grctx_pack_tpc[] = {
+	{ nvc1_grctx_init_pe_0 },
+	{ nvc4_grctx_init_tex_0 },
+	{ nvc1_grctx_init_wwdx_0 },
+	{ nvc0_grctx_init_mpc_0 },
+	{ nvc4_grctx_init_l1c_0 },
+	{ nvc1_grctx_init_tpccs_0 },
+	{ nvc4_grctx_init_sm_0 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
 void
 nvc1_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 {
@@ -771,41 +774,6 @@ nvc1_grctx_generate_unkn(struct nvc0_graph_priv *priv)
 	nv_mask(priv, 0x419c00, 0x00000008, 0x00000008);
 }
 
-static struct nvc0_graph_init *
-nvc1_grctx_init_hub[] = {
-	nvc0_grctx_init_base,
-	nvc0_grctx_init_unk40xx,
-	nvc0_grctx_init_unk44xx,
-	nvc0_grctx_init_unk46xx,
-	nvc0_grctx_init_unk47xx,
-	nvc1_grctx_init_unk58xx,
-	nvc0_grctx_init_unk60xx,
-	nvc0_grctx_init_unk64xx,
-	nvc0_grctx_init_unk78xx,
-	nvc0_grctx_init_unk80xx,
-	nvc1_grctx_init_rop,
-	NULL
-};
-
-struct nvc0_graph_init *
-nvc1_grctx_init_gpc[] = {
-	nvc1_grctx_init_gpc_0,
-	nvc0_grctx_init_gpc_1,
-	nvc1_grctx_init_tpc,
-	NULL
-};
-
-static struct nvc0_graph_mthd
-nvc1_grctx_init_mthd[] = {
-	{ 0x9097, nvc1_grctx_init_9097, },
-	{ 0x9197, nvc1_grctx_init_9197, },
-	{ 0x902d, nvc0_grctx_init_902d, },
-	{ 0x9039, nvc0_grctx_init_9039, },
-	{ 0x90c0, nvc0_grctx_init_90c0, },
-	{ 0x902d, nvc0_grctx_init_mthd_magic, },
-	{}
-};
-
 struct nouveau_oclass *
 nvc1_grctx_oclass = &(struct nvc0_grctx_oclass) {
 	.base.handle = NV_ENGCTX(GR, 0xc1),
@@ -817,11 +785,13 @@ nvc1_grctx_oclass = &(struct nvc0_grctx_oclass) {
 		.rd32 = _nouveau_graph_context_rd32,
 		.wr32 = _nouveau_graph_context_wr32,
 	},
-	.main = nvc0_grctx_generate_main,
-	.mods = nvc1_grctx_generate_mods,
-	.unkn = nvc1_grctx_generate_unkn,
-	.hub  = nvc1_grctx_init_hub,
-	.gpc  = nvc1_grctx_init_gpc,
-	.icmd = nvc1_grctx_init_icmd,
-	.mthd = nvc1_grctx_init_mthd,
+	.main  = nvc0_grctx_generate_main,
+	.mods  = nvc1_grctx_generate_mods,
+	.unkn  = nvc1_grctx_generate_unkn,
+	.hub   = nvc1_grctx_pack_hub,
+	.gpc   = nvc1_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nvc1_grctx_pack_tpc,
+	.icmd  = nvc1_grctx_pack_icmd,
+	.mthd  = nvc1_grctx_pack_mthd,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc3.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc4.c
index 8f237b3bd8c..e11ed553819 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc3.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc4.c
@@ -22,15 +22,14 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 
-#include "nvc0.h"
+#include "ctxnvc0.h"
 
-static struct nvc0_graph_init
-nvc3_grctx_init_tpc[] = {
-	{ 0x419818,   1, 0x04, 0x00000000 },
-	{ 0x41983c,   1, 0x04, 0x00038bc7 },
-	{ 0x419848,   1, 0x04, 0x00000000 },
-	{ 0x419864,   1, 0x04, 0x0000012a },
-	{ 0x419888,   1, 0x04, 0x00000000 },
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
+
+const struct nvc0_graph_init
+nvc4_grctx_init_tex_0[] = {
 	{ 0x419a00,   1, 0x04, 0x000001f0 },
 	{ 0x419a04,   1, 0x04, 0x00000001 },
 	{ 0x419a08,   1, 0x04, 0x00000023 },
@@ -40,24 +39,19 @@ nvc3_grctx_init_tpc[] = {
 	{ 0x419a1c,   1, 0x04, 0x00000000 },
 	{ 0x419a20,   1, 0x04, 0x00000800 },
 	{ 0x419ac4,   1, 0x04, 0x0007f440 },
-	{ 0x419b00,   1, 0x04, 0x0a418820 },
-	{ 0x419b04,   1, 0x04, 0x062080e6 },
-	{ 0x419b08,   1, 0x04, 0x020398a4 },
-	{ 0x419b0c,   1, 0x04, 0x0e629062 },
-	{ 0x419b10,   1, 0x04, 0x0a418820 },
-	{ 0x419b14,   1, 0x04, 0x000000e6 },
-	{ 0x419bd0,   1, 0x04, 0x00900103 },
-	{ 0x419be0,   1, 0x04, 0x00000001 },
-	{ 0x419be4,   1, 0x04, 0x00000000 },
-	{ 0x419c00,   1, 0x04, 0x00000002 },
-	{ 0x419c04,   1, 0x04, 0x00000006 },
-	{ 0x419c08,   1, 0x04, 0x00000002 },
-	{ 0x419c20,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc4_grctx_init_l1c_0[] = {
 	{ 0x419cb0,   1, 0x04, 0x00020048 },
 	{ 0x419ce8,   1, 0x04, 0x00000000 },
 	{ 0x419cf4,   1, 0x04, 0x00000183 },
-	{ 0x419d20,   1, 0x04, 0x02180000 },
-	{ 0x419d24,   1, 0x04, 0x00001fff },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc4_grctx_init_sm_0[] = {
 	{ 0x419e04,   3, 0x04, 0x00000000 },
 	{ 0x419e10,   1, 0x04, 0x00000002 },
 	{ 0x419e44,   1, 0x04, 0x001beff2 },
@@ -70,16 +64,24 @@ nvc3_grctx_init_tpc[] = {
 	{}
 };
 
-struct nvc0_graph_init *
-nvc3_grctx_init_gpc[] = {
-	nvc0_grctx_init_gpc_0,
-	nvc0_grctx_init_gpc_1,
-	nvc3_grctx_init_tpc,
-	NULL
+static const struct nvc0_graph_pack
+nvc4_grctx_pack_tpc[] = {
+	{ nvc0_grctx_init_pe_0 },
+	{ nvc4_grctx_init_tex_0 },
+	{ nvc0_grctx_init_wwdx_0 },
+	{ nvc0_grctx_init_mpc_0 },
+	{ nvc4_grctx_init_l1c_0 },
+	{ nvc0_grctx_init_tpccs_0 },
+	{ nvc4_grctx_init_sm_0 },
+	{}
 };
 
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
 struct nouveau_oclass *
-nvc3_grctx_oclass = &(struct nvc0_grctx_oclass) {
+nvc4_grctx_oclass = &(struct nvc0_grctx_oclass) {
 	.base.handle = NV_ENGCTX(GR, 0xc3),
 	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nvc0_graph_context_ctor,
@@ -89,11 +91,13 @@ nvc3_grctx_oclass = &(struct nvc0_grctx_oclass) {
 		.rd32 = _nouveau_graph_context_rd32,
 		.wr32 = _nouveau_graph_context_wr32,
 	},
-	.main = nvc0_grctx_generate_main,
-	.mods = nvc0_grctx_generate_mods,
-	.unkn = nvc0_grctx_generate_unkn,
-	.hub  = nvc0_grctx_init_hub,
-	.gpc  = nvc3_grctx_init_gpc,
-	.icmd = nvc0_grctx_init_icmd,
-	.mthd = nvc0_grctx_init_mthd,
+	.main  = nvc0_grctx_generate_main,
+	.mods  = nvc0_grctx_generate_mods,
+	.unkn  = nvc0_grctx_generate_unkn,
+	.hub   = nvc0_grctx_pack_hub,
+	.gpc   = nvc0_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nvc4_grctx_pack_tpc,
+	.icmd  = nvc0_grctx_pack_icmd,
+	.mthd  = nvc0_grctx_pack_mthd,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc8.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc8.c
index d0d4ce3c489..feebd58dfe8 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc8.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvc8.c
@@ -22,10 +22,14 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 
-#include "nvc0.h"
+#include "ctxnvc0.h"
 
-static struct nvc0_graph_init
-nvc8_grctx_init_icmd[] = {
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+nvc8_grctx_init_icmd_0[] = {
 	{ 0x001000,   1, 0x01, 0x00000004 },
 	{ 0x0000a9,   1, 0x01, 0x0000ffff },
 	{ 0x000038,   1, 0x01, 0x0fac6881 },
@@ -141,8 +145,7 @@ nvc8_grctx_init_icmd[] = {
 	{ 0x000586,   1, 0x01, 0x00000040 },
 	{ 0x000582,   2, 0x01, 0x00000080 },
 	{ 0x0005c2,   1, 0x01, 0x00000001 },
-	{ 0x000638,   1, 0x01, 0x00000001 },
-	{ 0x000639,   1, 0x01, 0x00000001 },
+	{ 0x000638,   2, 0x01, 0x00000001 },
 	{ 0x00063a,   1, 0x01, 0x00000002 },
 	{ 0x00063b,   2, 0x01, 0x00000001 },
 	{ 0x00063d,   1, 0x01, 0x00000002 },
@@ -203,15 +206,13 @@ nvc8_grctx_init_icmd[] = {
 	{ 0x000787,   1, 0x01, 0x000000cf },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x000836,   1, 0x01, 0x00000001 },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x00080c,   1, 0x01, 0x00000002 },
 	{ 0x00080d,   2, 0x01, 0x00000100 },
@@ -237,14 +238,12 @@ nvc8_grctx_init_icmd[] = {
 	{ 0x0006b1,   1, 0x01, 0x00000011 },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x01e100,   1, 0x01, 0x00000001 },
 	{ 0x001000,   1, 0x01, 0x00000014 },
@@ -269,58 +268,20 @@ nvc8_grctx_init_icmd[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvc8_grctx_init_tpc[] = {
-	{ 0x419818,   1, 0x04, 0x00000000 },
-	{ 0x41983c,   1, 0x04, 0x00038bc7 },
-	{ 0x419848,   1, 0x04, 0x00000000 },
-	{ 0x419864,   1, 0x04, 0x0000012a },
-	{ 0x419888,   1, 0x04, 0x00000000 },
-	{ 0x419a00,   1, 0x04, 0x000001f0 },
-	{ 0x419a04,   1, 0x04, 0x00000001 },
-	{ 0x419a08,   1, 0x04, 0x00000023 },
-	{ 0x419a0c,   1, 0x04, 0x00020000 },
-	{ 0x419a10,   1, 0x04, 0x00000000 },
-	{ 0x419a14,   1, 0x04, 0x00000200 },
-	{ 0x419a1c,   1, 0x04, 0x00000000 },
-	{ 0x419a20,   1, 0x04, 0x00000800 },
-	{ 0x419b00,   1, 0x04, 0x0a418820 },
-	{ 0x419b04,   1, 0x04, 0x062080e6 },
-	{ 0x419b08,   1, 0x04, 0x020398a4 },
-	{ 0x419b0c,   1, 0x04, 0x0e629062 },
-	{ 0x419b10,   1, 0x04, 0x0a418820 },
-	{ 0x419b14,   1, 0x04, 0x000000e6 },
-	{ 0x419bd0,   1, 0x04, 0x00900103 },
-	{ 0x419be0,   1, 0x04, 0x00000001 },
-	{ 0x419be4,   1, 0x04, 0x00000000 },
-	{ 0x419c00,   1, 0x04, 0x00000002 },
-	{ 0x419c04,   1, 0x04, 0x00000006 },
-	{ 0x419c08,   1, 0x04, 0x00000002 },
-	{ 0x419c20,   1, 0x04, 0x00000000 },
-	{ 0x419cb0,   1, 0x04, 0x00060048 },
-	{ 0x419ce8,   1, 0x04, 0x00000000 },
-	{ 0x419cf4,   1, 0x04, 0x00000183 },
-	{ 0x419d20,   1, 0x04, 0x02180000 },
-	{ 0x419d24,   1, 0x04, 0x00001fff },
-	{ 0x419e04,   3, 0x04, 0x00000000 },
-	{ 0x419e10,   1, 0x04, 0x00000002 },
-	{ 0x419e44,   1, 0x04, 0x001beff2 },
-	{ 0x419e48,   1, 0x04, 0x00000000 },
-	{ 0x419e4c,   1, 0x04, 0x0000000f },
-	{ 0x419e50,  17, 0x04, 0x00000000 },
-	{ 0x419e98,   1, 0x04, 0x00000000 },
-	{ 0x419f50,   2, 0x04, 0x00000000 },
+static const struct nvc0_graph_pack
+nvc8_grctx_pack_icmd[] = {
+	{ nvc8_grctx_init_icmd_0 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc8_grctx_init_9197[] = {
+const struct nvc0_graph_init
+nvc8_grctx_init_9197_0[] = {
 	{ 0x0002e4,   1, 0x04, 0x0000b001 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc8_grctx_init_9297[] = {
+const struct nvc0_graph_init
+nvc8_grctx_init_9297_0[] = {
 	{ 0x003400, 128, 0x04, 0x00000000 },
 	{ 0x00036c,   2, 0x04, 0x00000000 },
 	{ 0x0007a4,   2, 0x04, 0x00000000 },
@@ -329,26 +290,47 @@ nvc8_grctx_init_9297[] = {
 	{}
 };
 
-static struct nvc0_graph_mthd
-nvc8_grctx_init_mthd[] = {
-	{ 0x9097, nvc1_grctx_init_9097, },
-	{ 0x9197, nvc8_grctx_init_9197, },
-	{ 0x9297, nvc8_grctx_init_9297, },
-	{ 0x902d, nvc0_grctx_init_902d, },
-	{ 0x9039, nvc0_grctx_init_9039, },
-	{ 0x90c0, nvc0_grctx_init_90c0, },
-	{ 0x902d, nvc0_grctx_init_mthd_magic, },
+static const struct nvc0_graph_pack
+nvc8_grctx_pack_mthd[] = {
+	{ nvc1_grctx_init_9097_0, 0x9097 },
+	{ nvc8_grctx_init_9197_0, 0x9197 },
+	{ nvc8_grctx_init_9297_0, 0x9297 },
+	{ nvc0_grctx_init_902d_0, 0x902d },
+	{ nvc0_grctx_init_9039_0, 0x9039 },
+	{ nvc0_grctx_init_90c0_0, 0x90c0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc8_grctx_init_setup_0[] = {
+	{ 0x418800,   1, 0x04, 0x0006860a },
+	{ 0x418808,   3, 0x04, 0x00000000 },
+	{ 0x418828,   1, 0x04, 0x00008442 },
+	{ 0x418830,   1, 0x04, 0x00000001 },
+	{ 0x4188d8,   1, 0x04, 0x00000008 },
+	{ 0x4188e0,   1, 0x04, 0x01000000 },
+	{ 0x4188e8,   5, 0x04, 0x00000000 },
+	{ 0x4188fc,   1, 0x04, 0x20100000 },
 	{}
 };
 
-static struct nvc0_graph_init *
-nvc8_grctx_init_gpc[] = {
-	nvc0_grctx_init_gpc_0,
-	nvc0_grctx_init_gpc_1,
-	nvc8_grctx_init_tpc,
-	NULL
+static const struct nvc0_graph_pack
+nvc8_grctx_pack_gpc[] = {
+	{ nvc0_grctx_init_gpc_unk_0 },
+	{ nvc0_grctx_init_prop_0 },
+	{ nvc0_grctx_init_gpc_unk_1 },
+	{ nvc8_grctx_init_setup_0 },
+	{ nvc0_grctx_init_zcull_0 },
+	{ nvc0_grctx_init_crstr_0 },
+	{ nvc0_grctx_init_gpm_0 },
+	{ nvc0_grctx_init_gcc_0 },
+	{}
 };
 
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
 struct nouveau_oclass *
 nvc8_grctx_oclass = &(struct nvc0_grctx_oclass) {
 	.base.handle = NV_ENGCTX(GR, 0xc8),
@@ -360,11 +342,13 @@ nvc8_grctx_oclass = &(struct nvc0_grctx_oclass) {
 		.rd32 = _nouveau_graph_context_rd32,
 		.wr32 = _nouveau_graph_context_wr32,
 	},
-	.main = nvc0_grctx_generate_main,
-	.mods = nvc0_grctx_generate_mods,
-	.unkn = nvc0_grctx_generate_unkn,
-	.hub  = nvc0_grctx_init_hub,
-	.gpc  = nvc8_grctx_init_gpc,
-	.icmd = nvc8_grctx_init_icmd,
-	.mthd = nvc8_grctx_init_mthd,
+	.main  = nvc0_grctx_generate_main,
+	.mods  = nvc0_grctx_generate_mods,
+	.unkn  = nvc0_grctx_generate_unkn,
+	.hub   = nvc0_grctx_pack_hub,
+	.gpc   = nvc8_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nvc0_grctx_pack_tpc,
+	.icmd  = nvc8_grctx_pack_icmd,
+	.mthd  = nvc8_grctx_pack_mthd,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvd7.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvd7.c
index c4740d52853..1dbc8d7f2e8 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvd7.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvd7.c
@@ -22,33 +22,14 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 
-#include "nvc0.h"
+#include "ctxnvc0.h"
 
-struct nvc0_graph_init
-nvd7_grctx_init_unk40xx[] = {
-	{ 0x404004,  10, 0x04, 0x00000000 },
-	{ 0x404044,   1, 0x04, 0x00000000 },
-	{ 0x404094,   1, 0x04, 0x00000000 },
-	{ 0x404098,  12, 0x04, 0x00000000 },
-	{ 0x4040c8,   1, 0x04, 0xf0000087 },
-	{ 0x4040d0,   6, 0x04, 0x00000000 },
-	{ 0x4040e8,   1, 0x04, 0x00001000 },
-	{ 0x4040f8,   1, 0x04, 0x00000000 },
-	{ 0x404130,   1, 0x04, 0x00000000 },
-	{ 0x404134,   1, 0x04, 0x00000000 },
-	{ 0x404138,   1, 0x04, 0x20000040 },
-	{ 0x404150,   1, 0x04, 0x0000002e },
-	{ 0x404154,   1, 0x04, 0x00000400 },
-	{ 0x404158,   1, 0x04, 0x00000200 },
-	{ 0x404164,   1, 0x04, 0x00000055 },
-	{ 0x404168,   1, 0x04, 0x00000000 },
-	{ 0x404178,   2, 0x04, 0x00000000 },
-	{ 0x404200,   8, 0x04, 0x00000000 },
-	{}
-};
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
 
-static struct nvc0_graph_init
-nvd7_grctx_init_unk58xx[] = {
+static const struct nvc0_graph_init
+nvd7_grctx_init_ds_0[] = {
 	{ 0x405800,   1, 0x04, 0x0f8000bf },
 	{ 0x405830,   1, 0x04, 0x02180324 },
 	{ 0x405834,   1, 0x04, 0x08000000 },
@@ -60,8 +41,10 @@ nvd7_grctx_init_unk58xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvd7_grctx_init_unk64xx[] = {
+static const struct nvc0_graph_init
+nvd7_grctx_init_pd_0[] = {
+	{ 0x406020,   1, 0x04, 0x000103c1 },
+	{ 0x406028,   4, 0x04, 0x00000001 },
 	{ 0x4064a8,   1, 0x04, 0x00000000 },
 	{ 0x4064ac,   1, 0x04, 0x00003fff },
 	{ 0x4064b4,   3, 0x04, 0x00000000 },
@@ -71,22 +54,22 @@ nvd7_grctx_init_unk64xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvd7_grctx_init_gpc_0[] = {
-	{ 0x418380,   1, 0x04, 0x00000016 },
-	{ 0x418400,   1, 0x04, 0x38004e00 },
-	{ 0x418404,   1, 0x04, 0x71e0ffff },
-	{ 0x41840c,   1, 0x04, 0x00001008 },
-	{ 0x418410,   1, 0x04, 0x0fff0fff },
-	{ 0x418414,   1, 0x04, 0x02200fff },
-	{ 0x418450,   6, 0x04, 0x00000000 },
-	{ 0x418468,   1, 0x04, 0x00000001 },
-	{ 0x41846c,   2, 0x04, 0x00000000 },
-	{ 0x418600,   1, 0x04, 0x0000001f },
-	{ 0x418684,   1, 0x04, 0x0000000f },
-	{ 0x418700,   1, 0x04, 0x00000002 },
-	{ 0x418704,   1, 0x04, 0x00000080 },
-	{ 0x418708,   3, 0x04, 0x00000000 },
+static const struct nvc0_graph_pack
+nvd7_grctx_pack_hub[] = {
+	{ nvc0_grctx_init_main_0 },
+	{ nvd9_grctx_init_fe_0 },
+	{ nvc0_grctx_init_pri_0 },
+	{ nvc0_grctx_init_memfmt_0 },
+	{ nvd7_grctx_init_ds_0 },
+	{ nvd7_grctx_init_pd_0 },
+	{ nvc0_grctx_init_rstr2d_0 },
+	{ nvc0_grctx_init_scc_0 },
+	{ nvd9_grctx_init_be_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd7_grctx_init_setup_0[] = {
 	{ 0x418800,   1, 0x04, 0x7006860a },
 	{ 0x418808,   3, 0x04, 0x00000000 },
 	{ 0x418828,   1, 0x04, 0x00008442 },
@@ -95,34 +78,32 @@ nvd7_grctx_init_gpc_0[] = {
 	{ 0x4188e0,   1, 0x04, 0x01000000 },
 	{ 0x4188e8,   5, 0x04, 0x00000000 },
 	{ 0x4188fc,   1, 0x04, 0x20100018 },
-	{ 0x41891c,   1, 0x04, 0x00ff00ff },
-	{ 0x418924,   1, 0x04, 0x00000000 },
-	{ 0x418928,   1, 0x04, 0x00ffff00 },
-	{ 0x41892c,   1, 0x04, 0x0000ff00 },
-	{ 0x418b00,   1, 0x04, 0x00000006 },
-	{ 0x418b08,   1, 0x04, 0x0a418820 },
-	{ 0x418b0c,   1, 0x04, 0x062080e6 },
-	{ 0x418b10,   1, 0x04, 0x020398a4 },
-	{ 0x418b14,   1, 0x04, 0x0e629062 },
-	{ 0x418b18,   1, 0x04, 0x0a418820 },
-	{ 0x418b1c,   1, 0x04, 0x000000e6 },
-	{ 0x418bb8,   1, 0x04, 0x00000103 },
-	{ 0x418c08,   1, 0x04, 0x00000001 },
-	{ 0x418c10,   8, 0x04, 0x00000000 },
-	{ 0x418c6c,   1, 0x04, 0x00000001 },
-	{ 0x418c80,   1, 0x04, 0x20200004 },
-	{ 0x418c8c,   1, 0x04, 0x00000001 },
-	{ 0x419000,   1, 0x04, 0x00000780 },
-	{ 0x419004,   2, 0x04, 0x00000000 },
-	{ 0x419014,   1, 0x04, 0x00000004 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvd7_grctx_init_tpc[] = {
+static const struct nvc0_graph_pack
+nvd7_grctx_pack_gpc[] = {
+	{ nvc0_grctx_init_gpc_unk_0 },
+	{ nvd9_grctx_init_prop_0 },
+	{ nvd9_grctx_init_gpc_unk_1 },
+	{ nvd7_grctx_init_setup_0 },
+	{ nvc0_grctx_init_zcull_0 },
+	{ nvd9_grctx_init_crstr_0 },
+	{ nvc1_grctx_init_gpm_0 },
+	{ nvc0_grctx_init_gcc_0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd7_grctx_init_pe_0[] = {
 	{ 0x419848,   1, 0x04, 0x00000000 },
 	{ 0x419864,   1, 0x04, 0x00000129 },
 	{ 0x419888,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd7_grctx_init_tex_0[] = {
 	{ 0x419a00,   1, 0x04, 0x000001f0 },
 	{ 0x419a04,   1, 0x04, 0x00000001 },
 	{ 0x419a08,   1, 0x04, 0x00000023 },
@@ -132,33 +113,46 @@ nvd7_grctx_init_tpc[] = {
 	{ 0x419a1c,   1, 0x04, 0x00008000 },
 	{ 0x419a20,   1, 0x04, 0x00000800 },
 	{ 0x419ac4,   1, 0x04, 0x0017f440 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd7_grctx_init_mpc_0[] = {
 	{ 0x419c00,   1, 0x04, 0x0000000a },
 	{ 0x419c04,   1, 0x04, 0x00000006 },
 	{ 0x419c08,   1, 0x04, 0x00000002 },
 	{ 0x419c20,   1, 0x04, 0x00000000 },
 	{ 0x419c24,   1, 0x04, 0x00084210 },
 	{ 0x419c28,   1, 0x04, 0x3efbefbe },
-	{ 0x419cb0,   1, 0x04, 0x00020048 },
-	{ 0x419ce8,   1, 0x04, 0x00000000 },
-	{ 0x419cf4,   1, 0x04, 0x00000183 },
-	{ 0x419e04,   3, 0x04, 0x00000000 },
-	{ 0x419e10,   1, 0x04, 0x00000002 },
-	{ 0x419e44,   1, 0x04, 0x001beff2 },
-	{ 0x419e48,   1, 0x04, 0x00000000 },
-	{ 0x419e4c,   1, 0x04, 0x0000000f },
-	{ 0x419e50,  17, 0x04, 0x00000000 },
-	{ 0x419e98,   1, 0x04, 0x00000000 },
-	{ 0x419ee0,   1, 0x04, 0x00010110 },
-	{ 0x419f30,  11, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvd7_grctx_init_unk[] = {
+static const struct nvc0_graph_pack
+nvd7_grctx_pack_tpc[] = {
+	{ nvd7_grctx_init_pe_0 },
+	{ nvd7_grctx_init_tex_0 },
+	{ nvd7_grctx_init_mpc_0 },
+	{ nvc4_grctx_init_l1c_0 },
+	{ nvd9_grctx_init_sm_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd7_grctx_init_pes_0[] = {
 	{ 0x41be24,   1, 0x04, 0x00000002 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd7_grctx_init_cbm_0[] = {
 	{ 0x41bec0,   1, 0x04, 0x12180000 },
 	{ 0x41bec4,   1, 0x04, 0x00003fff },
 	{ 0x41bee4,   1, 0x04, 0x03240218 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd7_grctx_init_wwdx_0[] = {
 	{ 0x41bf00,   1, 0x04, 0x0a418820 },
 	{ 0x41bf04,   1, 0x04, 0x062080e6 },
 	{ 0x41bf08,   1, 0x04, 0x020398a4 },
@@ -171,6 +165,18 @@ nvd7_grctx_init_unk[] = {
 	{}
 };
 
+static const struct nvc0_graph_pack
+nvd7_grctx_pack_ppc[] = {
+	{ nvd7_grctx_init_pes_0 },
+	{ nvd7_grctx_init_cbm_0 },
+	{ nvd7_grctx_init_wwdx_0 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
 static void
 nvd7_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 {
@@ -219,10 +225,11 @@ nvd7_grctx_generate_main(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 
 	nv_mask(priv, 0x000260, 0x00000001, 0x00000000);
 
-	for (i = 0; oclass->hub[i]; i++)
-		nvc0_graph_mmio(priv, oclass->hub[i]);
-	for (i = 0; oclass->gpc[i]; i++)
-		nvc0_graph_mmio(priv, oclass->gpc[i]);
+	nvc0_graph_mmio(priv, oclass->hub);
+	nvc0_graph_mmio(priv, oclass->gpc);
+	nvc0_graph_mmio(priv, oclass->zcull);
+	nvc0_graph_mmio(priv, oclass->tpc);
+	nvc0_graph_mmio(priv, oclass->ppc);
 
 	nv_wr32(priv, 0x404154, 0x00000000);
 
@@ -244,32 +251,6 @@ nvd7_grctx_generate_main(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 	nv_mask(priv, 0x000260, 0x00000001, 0x00000001);
 }
 
-
-static struct nvc0_graph_init *
-nvd7_grctx_init_hub[] = {
-	nvc0_grctx_init_base,
-	nvd7_grctx_init_unk40xx,
-	nvc0_grctx_init_unk44xx,
-	nvc0_grctx_init_unk46xx,
-	nvc0_grctx_init_unk47xx,
-	nvd7_grctx_init_unk58xx,
-	nvc0_grctx_init_unk60xx,
-	nvd7_grctx_init_unk64xx,
-	nvc0_grctx_init_unk78xx,
-	nvc0_grctx_init_unk80xx,
-	nvd9_grctx_init_rop,
-	NULL
-};
-
-struct nvc0_graph_init *
-nvd7_grctx_init_gpc[] = {
-	nvd7_grctx_init_gpc_0,
-	nvc0_grctx_init_gpc_1,
-	nvd7_grctx_init_tpc,
-	nvd7_grctx_init_unk,
-	NULL
-};
-
 struct nouveau_oclass *
 nvd7_grctx_oclass = &(struct nvc0_grctx_oclass) {
 	.base.handle = NV_ENGCTX(GR, 0xd7),
@@ -281,11 +262,14 @@ nvd7_grctx_oclass = &(struct nvc0_grctx_oclass) {
 		.rd32 = _nouveau_graph_context_rd32,
 		.wr32 = _nouveau_graph_context_wr32,
 	},
-	.main = nvd7_grctx_generate_main,
-	.mods = nvd7_grctx_generate_mods,
-	.unkn = nve4_grctx_generate_unkn,
-	.hub  = nvd7_grctx_init_hub,
-	.gpc  = nvd7_grctx_init_gpc,
-	.icmd = nvd9_grctx_init_icmd,
-	.mthd = nvd9_grctx_init_mthd,
+	.main  = nvd7_grctx_generate_main,
+	.mods  = nvd7_grctx_generate_mods,
+	.unkn  = nve4_grctx_generate_unkn,
+	.hub   = nvd7_grctx_pack_hub,
+	.gpc   = nvd7_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nvd7_grctx_pack_tpc,
+	.ppc   = nvd7_grctx_pack_ppc,
+	.icmd  = nvd9_grctx_pack_icmd,
+	.mthd  = nvd9_grctx_pack_mthd,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvd9.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvd9.c
index a1102cbf2fd..c665fb7e466 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvd9.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvd9.c
@@ -22,38 +22,14 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 
-#include "nvc0.h"
+#include "ctxnvc0.h"
 
-struct nvc0_graph_init
-nvd9_grctx_init_90c0[] = {
-	{ 0x002700,   4, 0x40, 0x00000000 },
-	{ 0x002720,   4, 0x40, 0x00000000 },
-	{ 0x002704,   4, 0x40, 0x00000000 },
-	{ 0x002724,   4, 0x40, 0x00000000 },
-	{ 0x002708,   4, 0x40, 0x00000000 },
-	{ 0x002728,   4, 0x40, 0x00000000 },
-	{ 0x00270c,   8, 0x20, 0x00000000 },
-	{ 0x002710,   4, 0x40, 0x00014000 },
-	{ 0x002730,   4, 0x40, 0x00014000 },
-	{ 0x002714,   4, 0x40, 0x00000040 },
-	{ 0x002734,   4, 0x40, 0x00000040 },
-	{ 0x00030c,   1, 0x04, 0x00000001 },
-	{ 0x001944,   1, 0x04, 0x00000000 },
-	{ 0x000758,   1, 0x04, 0x00000100 },
-	{ 0x0002c4,   1, 0x04, 0x00000000 },
-	{ 0x000790,   5, 0x04, 0x00000000 },
-	{ 0x00077c,   1, 0x04, 0x00000000 },
-	{ 0x000204,   3, 0x04, 0x00000000 },
-	{ 0x000214,   1, 0x04, 0x00000000 },
-	{ 0x00024c,   1, 0x04, 0x00000000 },
-	{ 0x000d94,   1, 0x04, 0x00000001 },
-	{ 0x001608,   2, 0x04, 0x00000000 },
-	{ 0x001664,   1, 0x04, 0x00000000 },
-	{}
-};
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
 
-struct nvc0_graph_init
-nvd9_grctx_init_icmd[] = {
+static const struct nvc0_graph_init
+nvd9_grctx_init_icmd_0[] = {
 	{ 0x001000,   1, 0x01, 0x00000004 },
 	{ 0x0000a9,   1, 0x01, 0x0000ffff },
 	{ 0x000038,   1, 0x01, 0x0fac6881 },
@@ -171,8 +147,7 @@ nvd9_grctx_init_icmd[] = {
 	{ 0x000586,   1, 0x01, 0x00000040 },
 	{ 0x000582,   2, 0x01, 0x00000080 },
 	{ 0x0005c2,   1, 0x01, 0x00000001 },
-	{ 0x000638,   1, 0x01, 0x00000001 },
-	{ 0x000639,   1, 0x01, 0x00000001 },
+	{ 0x000638,   2, 0x01, 0x00000001 },
 	{ 0x00063a,   1, 0x01, 0x00000002 },
 	{ 0x00063b,   2, 0x01, 0x00000001 },
 	{ 0x00063d,   1, 0x01, 0x00000002 },
@@ -233,15 +208,13 @@ nvd9_grctx_init_icmd[] = {
 	{ 0x000787,   1, 0x01, 0x000000cf },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x000836,   1, 0x01, 0x00000001 },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x00080c,   1, 0x01, 0x00000002 },
 	{ 0x00080d,   2, 0x01, 0x00000100 },
@@ -267,14 +240,12 @@ nvd9_grctx_init_icmd[] = {
 	{ 0x0006b1,   1, 0x01, 0x00000011 },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x01e100,   1, 0x01, 0x00000001 },
 	{ 0x001000,   1, 0x01, 0x00000014 },
@@ -299,18 +270,56 @@ nvd9_grctx_init_icmd[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvd9_grctx_init_unk40xx[] = {
-	{ 0x404004,  11, 0x04, 0x00000000 },
+const struct nvc0_graph_pack
+nvd9_grctx_pack_icmd[] = {
+	{ nvd9_grctx_init_icmd_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd9_grctx_init_90c0_0[] = {
+	{ 0x002700,   8, 0x20, 0x00000000 },
+	{ 0x002704,   8, 0x20, 0x00000000 },
+	{ 0x002708,   8, 0x20, 0x00000000 },
+	{ 0x00270c,   8, 0x20, 0x00000000 },
+	{ 0x002710,   8, 0x20, 0x00014000 },
+	{ 0x002714,   8, 0x20, 0x00000040 },
+	{ 0x00030c,   1, 0x04, 0x00000001 },
+	{ 0x001944,   1, 0x04, 0x00000000 },
+	{ 0x000758,   1, 0x04, 0x00000100 },
+	{ 0x0002c4,   1, 0x04, 0x00000000 },
+	{ 0x000790,   5, 0x04, 0x00000000 },
+	{ 0x00077c,   1, 0x04, 0x00000000 },
+	{ 0x000204,   3, 0x04, 0x00000000 },
+	{ 0x000214,   1, 0x04, 0x00000000 },
+	{ 0x00024c,   1, 0x04, 0x00000000 },
+	{ 0x000d94,   1, 0x04, 0x00000001 },
+	{ 0x001608,   2, 0x04, 0x00000000 },
+	{ 0x001664,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_pack
+nvd9_grctx_pack_mthd[] = {
+	{ nvc1_grctx_init_9097_0, 0x9097 },
+	{ nvc8_grctx_init_9197_0, 0x9197 },
+	{ nvc8_grctx_init_9297_0, 0x9297 },
+	{ nvc0_grctx_init_902d_0, 0x902d },
+	{ nvc0_grctx_init_9039_0, 0x9039 },
+	{ nvd9_grctx_init_90c0_0, 0x90c0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd9_grctx_init_fe_0[] = {
+	{ 0x404004,  10, 0x04, 0x00000000 },
 	{ 0x404044,   1, 0x04, 0x00000000 },
-	{ 0x404094,   1, 0x04, 0x00000000 },
-	{ 0x404098,  12, 0x04, 0x00000000 },
+	{ 0x404094,  13, 0x04, 0x00000000 },
 	{ 0x4040c8,   1, 0x04, 0xf0000087 },
 	{ 0x4040d0,   6, 0x04, 0x00000000 },
 	{ 0x4040e8,   1, 0x04, 0x00001000 },
 	{ 0x4040f8,   1, 0x04, 0x00000000 },
-	{ 0x404130,   1, 0x04, 0x00000000 },
-	{ 0x404134,   1, 0x04, 0x00000000 },
+	{ 0x404130,   2, 0x04, 0x00000000 },
 	{ 0x404138,   1, 0x04, 0x20000040 },
 	{ 0x404150,   1, 0x04, 0x0000002e },
 	{ 0x404154,   1, 0x04, 0x00000400 },
@@ -322,8 +331,8 @@ nvd9_grctx_init_unk40xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvd9_grctx_init_unk58xx[] = {
+static const struct nvc0_graph_init
+nvd9_grctx_init_ds_0[] = {
 	{ 0x405800,   1, 0x04, 0x0f8000bf },
 	{ 0x405830,   1, 0x04, 0x02180218 },
 	{ 0x405834,   1, 0x04, 0x08000000 },
@@ -335,8 +344,10 @@ nvd9_grctx_init_unk58xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvd9_grctx_init_unk64xx[] = {
+static const struct nvc0_graph_init
+nvd9_grctx_init_pd_0[] = {
+	{ 0x406020,   1, 0x04, 0x000103c1 },
+	{ 0x406028,   4, 0x04, 0x00000001 },
 	{ 0x4064a8,   1, 0x04, 0x00000000 },
 	{ 0x4064ac,   1, 0x04, 0x00003fff },
 	{ 0x4064b4,   3, 0x04, 0x00000000 },
@@ -345,21 +356,34 @@ nvd9_grctx_init_unk64xx[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvd9_grctx_init_rop[] = {
+const struct nvc0_graph_init
+nvd9_grctx_init_be_0[] = {
 	{ 0x408800,   1, 0x04, 0x02802a3c },
 	{ 0x408804,   1, 0x04, 0x00000040 },
 	{ 0x408808,   1, 0x04, 0x1043e005 },
 	{ 0x408900,   1, 0x04, 0x3080b801 },
-	{ 0x408904,   1, 0x04, 0x1043e005 },
+	{ 0x408904,   1, 0x04, 0x62000001 },
 	{ 0x408908,   1, 0x04, 0x00c8102f },
 	{ 0x408980,   1, 0x04, 0x0000011d },
 	{}
 };
 
-static struct nvc0_graph_init
-nvd9_grctx_init_gpc_0[] = {
-	{ 0x418380,   1, 0x04, 0x00000016 },
+static const struct nvc0_graph_pack
+nvd9_grctx_pack_hub[] = {
+	{ nvc0_grctx_init_main_0 },
+	{ nvd9_grctx_init_fe_0 },
+	{ nvc0_grctx_init_pri_0 },
+	{ nvc0_grctx_init_memfmt_0 },
+	{ nvd9_grctx_init_ds_0 },
+	{ nvd9_grctx_init_pd_0 },
+	{ nvc0_grctx_init_rstr2d_0 },
+	{ nvc0_grctx_init_scc_0 },
+	{ nvd9_grctx_init_be_0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd9_grctx_init_prop_0[] = {
 	{ 0x418400,   1, 0x04, 0x38004e00 },
 	{ 0x418404,   1, 0x04, 0x71e0ffff },
 	{ 0x41840c,   1, 0x04, 0x00001008 },
@@ -368,11 +392,21 @@ nvd9_grctx_init_gpc_0[] = {
 	{ 0x418450,   6, 0x04, 0x00000000 },
 	{ 0x418468,   1, 0x04, 0x00000001 },
 	{ 0x41846c,   2, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd9_grctx_init_gpc_unk_1[] = {
 	{ 0x418600,   1, 0x04, 0x0000001f },
 	{ 0x418684,   1, 0x04, 0x0000000f },
 	{ 0x418700,   1, 0x04, 0x00000002 },
 	{ 0x418704,   1, 0x04, 0x00000080 },
 	{ 0x418708,   3, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd9_grctx_init_setup_0[] = {
 	{ 0x418800,   1, 0x04, 0x7006860a },
 	{ 0x418808,   3, 0x04, 0x00000000 },
 	{ 0x418828,   1, 0x04, 0x00008442 },
@@ -381,10 +415,11 @@ nvd9_grctx_init_gpc_0[] = {
 	{ 0x4188e0,   1, 0x04, 0x01000000 },
 	{ 0x4188e8,   5, 0x04, 0x00000000 },
 	{ 0x4188fc,   1, 0x04, 0x20100008 },
-	{ 0x41891c,   1, 0x04, 0x00ff00ff },
-	{ 0x418924,   1, 0x04, 0x00000000 },
-	{ 0x418928,   1, 0x04, 0x00ffff00 },
-	{ 0x41892c,   1, 0x04, 0x0000ff00 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd9_grctx_init_crstr_0[] = {
 	{ 0x418b00,   1, 0x04, 0x00000006 },
 	{ 0x418b08,   1, 0x04, 0x0a418820 },
 	{ 0x418b0c,   1, 0x04, 0x062080e6 },
@@ -393,24 +428,24 @@ nvd9_grctx_init_gpc_0[] = {
 	{ 0x418b18,   1, 0x04, 0x0a418820 },
 	{ 0x418b1c,   1, 0x04, 0x000000e6 },
 	{ 0x418bb8,   1, 0x04, 0x00000103 },
-	{ 0x418c08,   1, 0x04, 0x00000001 },
-	{ 0x418c10,   8, 0x04, 0x00000000 },
-	{ 0x418c6c,   1, 0x04, 0x00000001 },
-	{ 0x418c80,   1, 0x04, 0x20200004 },
-	{ 0x418c8c,   1, 0x04, 0x00000001 },
-	{ 0x419000,   1, 0x04, 0x00000780 },
-	{ 0x419004,   2, 0x04, 0x00000000 },
-	{ 0x419014,   1, 0x04, 0x00000004 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvd9_grctx_init_tpc[] = {
-	{ 0x419818,   1, 0x04, 0x00000000 },
-	{ 0x41983c,   1, 0x04, 0x00038bc7 },
-	{ 0x419848,   1, 0x04, 0x00000000 },
-	{ 0x419864,   1, 0x04, 0x00000129 },
-	{ 0x419888,   1, 0x04, 0x00000000 },
+static const struct nvc0_graph_pack
+nvd9_grctx_pack_gpc[] = {
+	{ nvc0_grctx_init_gpc_unk_0 },
+	{ nvd9_grctx_init_prop_0 },
+	{ nvd9_grctx_init_gpc_unk_1 },
+	{ nvd9_grctx_init_setup_0 },
+	{ nvc0_grctx_init_zcull_0 },
+	{ nvd9_grctx_init_crstr_0 },
+	{ nvc1_grctx_init_gpm_0 },
+	{ nvc0_grctx_init_gcc_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd9_grctx_init_tex_0[] = {
 	{ 0x419a00,   1, 0x04, 0x000001f0 },
 	{ 0x419a04,   1, 0x04, 0x00000001 },
 	{ 0x419a08,   1, 0x04, 0x00000023 },
@@ -420,27 +455,22 @@ nvd9_grctx_init_tpc[] = {
 	{ 0x419a1c,   1, 0x04, 0x00000000 },
 	{ 0x419a20,   1, 0x04, 0x00000800 },
 	{ 0x419ac4,   1, 0x04, 0x0017f440 },
-	{ 0x419b00,   1, 0x04, 0x0a418820 },
-	{ 0x419b04,   1, 0x04, 0x062080e6 },
-	{ 0x419b08,   1, 0x04, 0x020398a4 },
-	{ 0x419b0c,   1, 0x04, 0x0e629062 },
-	{ 0x419b10,   1, 0x04, 0x0a418820 },
-	{ 0x419b14,   1, 0x04, 0x000000e6 },
-	{ 0x419bd0,   1, 0x04, 0x00900103 },
-	{ 0x419be0,   1, 0x04, 0x00400001 },
-	{ 0x419be4,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd9_grctx_init_mpc_0[] = {
 	{ 0x419c00,   1, 0x04, 0x0000000a },
 	{ 0x419c04,   1, 0x04, 0x00000006 },
 	{ 0x419c08,   1, 0x04, 0x00000002 },
 	{ 0x419c20,   1, 0x04, 0x00000000 },
 	{ 0x419c24,   1, 0x04, 0x00084210 },
 	{ 0x419c28,   1, 0x04, 0x3cf3cf3c },
-	{ 0x419cb0,   1, 0x04, 0x00020048 },
-	{ 0x419ce8,   1, 0x04, 0x00000000 },
-	{ 0x419cf4,   1, 0x04, 0x00000183 },
-	{ 0x419d20,   1, 0x04, 0x12180000 },
-	{ 0x419d24,   1, 0x04, 0x00001fff },
-	{ 0x419d44,   1, 0x04, 0x02180218 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd9_grctx_init_sm_0[] = {
 	{ 0x419e04,   3, 0x04, 0x00000000 },
 	{ 0x419e10,   1, 0x04, 0x00000002 },
 	{ 0x419e44,   1, 0x04, 0x001beff2 },
@@ -453,47 +483,21 @@ nvd9_grctx_init_tpc[] = {
 	{}
 };
 
-static struct nvc0_graph_init *
-nvd9_grctx_init_hub[] = {
-	nvc0_grctx_init_base,
-	nvd9_grctx_init_unk40xx,
-	nvc0_grctx_init_unk44xx,
-	nvc0_grctx_init_unk46xx,
-	nvc0_grctx_init_unk47xx,
-	nvd9_grctx_init_unk58xx,
-	nvc0_grctx_init_unk60xx,
-	nvd9_grctx_init_unk64xx,
-	nvc0_grctx_init_unk78xx,
-	nvc0_grctx_init_unk80xx,
-	nvd9_grctx_init_rop,
-	NULL
-};
-
-struct nvc0_graph_init *
-nvd9_grctx_init_gpc[] = {
-	nvd9_grctx_init_gpc_0,
-	nvc0_grctx_init_gpc_1,
-	nvd9_grctx_init_tpc,
-	NULL
-};
-
-struct nvc0_graph_init
-nvd9_grctx_init_mthd_magic[] = {
-	{ 0x3410, 1, 0x04, 0x80002006 },
+static const struct nvc0_graph_pack
+nvd9_grctx_pack_tpc[] = {
+	{ nvc1_grctx_init_pe_0 },
+	{ nvd9_grctx_init_tex_0 },
+	{ nvc1_grctx_init_wwdx_0 },
+	{ nvd9_grctx_init_mpc_0 },
+	{ nvc4_grctx_init_l1c_0 },
+	{ nvc1_grctx_init_tpccs_0 },
+	{ nvd9_grctx_init_sm_0 },
 	{}
 };
 
-struct nvc0_graph_mthd
-nvd9_grctx_init_mthd[] = {
-	{ 0x9097, nvc1_grctx_init_9097, },
-	{ 0x9197, nvc8_grctx_init_9197, },
-	{ 0x9297, nvc8_grctx_init_9297, },
-	{ 0x902d, nvc0_grctx_init_902d, },
-	{ 0x9039, nvc0_grctx_init_9039, },
-	{ 0x90c0, nvd9_grctx_init_90c0, },
-	{ 0x902d, nvd9_grctx_init_mthd_magic, },
-	{}
-};
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
 
 struct nouveau_oclass *
 nvd9_grctx_oclass = &(struct nvc0_grctx_oclass) {
@@ -506,11 +510,13 @@ nvd9_grctx_oclass = &(struct nvc0_grctx_oclass) {
 		.rd32 = _nouveau_graph_context_rd32,
 		.wr32 = _nouveau_graph_context_wr32,
 	},
-	.main = nvc0_grctx_generate_main,
-	.mods = nvc1_grctx_generate_mods,
-	.unkn = nvc1_grctx_generate_unkn,
-	.hub  = nvd9_grctx_init_hub,
-	.gpc  = nvd9_grctx_init_gpc,
-	.icmd = nvd9_grctx_init_icmd,
-	.mthd = nvd9_grctx_init_mthd,
+	.main  = nvc0_grctx_generate_main,
+	.mods  = nvc1_grctx_generate_mods,
+	.unkn  = nvc1_grctx_generate_unkn,
+	.hub   = nvd9_grctx_pack_hub,
+	.gpc   = nvd9_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nvd9_grctx_pack_tpc,
+	.icmd  = nvd9_grctx_pack_icmd,
+	.mthd  = nvd9_grctx_pack_mthd,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnve4.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnve4.c
index e2de73ee5ee..c5b24923858 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnve4.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnve4.c
@@ -22,10 +22,14 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 
-#include "nvc0.h"
+#include "ctxnvc0.h"
 
-struct nvc0_graph_init
-nve4_grctx_init_icmd[] = {
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+nve4_grctx_init_icmd_0[] = {
 	{ 0x001000,   1, 0x01, 0x00000004 },
 	{ 0x000039,   3, 0x01, 0x00000000 },
 	{ 0x0000a9,   1, 0x01, 0x0000ffff },
@@ -138,8 +142,7 @@ nve4_grctx_init_icmd[] = {
 	{ 0x000586,   1, 0x01, 0x00000040 },
 	{ 0x000582,   2, 0x01, 0x00000080 },
 	{ 0x0005c2,   1, 0x01, 0x00000001 },
-	{ 0x000638,   1, 0x01, 0x00000001 },
-	{ 0x000639,   1, 0x01, 0x00000001 },
+	{ 0x000638,   2, 0x01, 0x00000001 },
 	{ 0x00063a,   1, 0x01, 0x00000002 },
 	{ 0x00063b,   2, 0x01, 0x00000001 },
 	{ 0x00063d,   1, 0x01, 0x00000002 },
@@ -197,15 +200,13 @@ nve4_grctx_init_icmd[] = {
 	{ 0x000787,   1, 0x01, 0x000000cf },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x000836,   1, 0x01, 0x00000001 },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x000b07,   1, 0x01, 0x00000002 },
 	{ 0x000b08,   2, 0x01, 0x00000100 },
@@ -231,14 +232,12 @@ nve4_grctx_init_icmd[] = {
 	{ 0x0006b1,   1, 0x01, 0x00000011 },
 	{ 0x00078c,   1, 0x01, 0x00000008 },
 	{ 0x000792,   1, 0x01, 0x00000001 },
-	{ 0x000794,   1, 0x01, 0x00000001 },
-	{ 0x000795,   2, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
 	{ 0x000797,   1, 0x01, 0x000000cf },
 	{ 0x00079a,   1, 0x01, 0x00000002 },
 	{ 0x000833,   1, 0x01, 0x04444480 },
 	{ 0x0007a1,   1, 0x01, 0x00000001 },
-	{ 0x0007a3,   1, 0x01, 0x00000001 },
-	{ 0x0007a4,   2, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
 	{ 0x000831,   1, 0x01, 0x00000004 },
 	{ 0x01e100,   1, 0x01, 0x00000001 },
 	{ 0x001000,   1, 0x01, 0x00000008 },
@@ -273,8 +272,14 @@ nve4_grctx_init_icmd[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nve4_grctx_init_a097[] = {
+const struct nvc0_graph_pack
+nve4_grctx_pack_icmd[] = {
+	{ nve4_grctx_init_icmd_0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nve4_grctx_init_a097_0[] = {
 	{ 0x000800,   8, 0x40, 0x00000000 },
 	{ 0x000804,   8, 0x40, 0x00000000 },
 	{ 0x000808,   8, 0x40, 0x00000400 },
@@ -517,8 +522,7 @@ nve4_grctx_init_a097[] = {
 	{ 0x001350,   1, 0x04, 0x00000002 },
 	{ 0x001358,   1, 0x04, 0x00000001 },
 	{ 0x0012e4,   1, 0x04, 0x00000000 },
-	{ 0x00131c,   1, 0x04, 0x00000000 },
-	{ 0x001320,   3, 0x04, 0x00000000 },
+	{ 0x00131c,   4, 0x04, 0x00000000 },
 	{ 0x0019c0,   1, 0x04, 0x00000000 },
 	{ 0x001140,   1, 0x04, 0x00000000 },
 	{ 0x0019c4,   1, 0x04, 0x00000000 },
@@ -574,19 +578,24 @@ nve4_grctx_init_a097[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_grctx_init_unk40xx[] = {
+static const struct nvc0_graph_pack
+nve4_grctx_pack_mthd[] = {
+	{ nve4_grctx_init_a097_0, 0xa097 },
+	{ nvc0_grctx_init_902d_0, 0x902d },
+	{}
+};
+
+static const struct nvc0_graph_init
+nve4_grctx_init_fe_0[] = {
 	{ 0x404010,   5, 0x04, 0x00000000 },
 	{ 0x404024,   1, 0x04, 0x0000e000 },
 	{ 0x404028,   1, 0x04, 0x00000000 },
-	{ 0x4040a8,   1, 0x04, 0x00000000 },
-	{ 0x4040ac,   7, 0x04, 0x00000000 },
+	{ 0x4040a8,   8, 0x04, 0x00000000 },
 	{ 0x4040c8,   1, 0x04, 0xf800008f },
 	{ 0x4040d0,   6, 0x04, 0x00000000 },
 	{ 0x4040e8,   1, 0x04, 0x00001000 },
 	{ 0x4040f8,   1, 0x04, 0x00000000 },
-	{ 0x404130,   1, 0x04, 0x00000000 },
-	{ 0x404134,   1, 0x04, 0x00000000 },
+	{ 0x404130,   2, 0x04, 0x00000000 },
 	{ 0x404138,   1, 0x04, 0x20000040 },
 	{ 0x404150,   1, 0x04, 0x0000002e },
 	{ 0x404154,   1, 0x04, 0x00000400 },
@@ -597,8 +606,8 @@ nve4_grctx_init_unk40xx[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nve4_grctx_init_unk46xx[] = {
+const struct nvc0_graph_init
+nve4_grctx_init_memfmt_0[] = {
 	{ 0x404604,   1, 0x04, 0x00000014 },
 	{ 0x404608,   1, 0x04, 0x00000000 },
 	{ 0x40460c,   1, 0x04, 0x00003fff },
@@ -614,11 +623,6 @@ nve4_grctx_init_unk46xx[] = {
 	{ 0x4046a0,   1, 0x04, 0x007f0080 },
 	{ 0x4046a4,   8, 0x04, 0x00000000 },
 	{ 0x4046c8,   3, 0x04, 0x00000000 },
-	{}
-};
-
-struct nvc0_graph_init
-nve4_grctx_init_unk47xx[] = {
 	{ 0x404700,   3, 0x04, 0x00000000 },
 	{ 0x404718,   7, 0x04, 0x00000000 },
 	{ 0x404734,   1, 0x04, 0x00000100 },
@@ -628,8 +632,8 @@ nve4_grctx_init_unk47xx[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nve4_grctx_init_unk58xx[] = {
+const struct nvc0_graph_init
+nve4_grctx_init_ds_0[] = {
 	{ 0x405800,   1, 0x04, 0x0f8000bf },
 	{ 0x405830,   1, 0x04, 0x02180648 },
 	{ 0x405834,   1, 0x04, 0x08000000 },
@@ -641,22 +645,17 @@ nve4_grctx_init_unk58xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_grctx_init_unk5bxx[] = {
+static const struct nvc0_graph_init
+nve4_grctx_init_cwd_0[] = {
 	{ 0x405b00,   1, 0x04, 0x00000000 },
 	{ 0x405b10,   1, 0x04, 0x00001000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_grctx_init_unk60xx[] = {
+static const struct nvc0_graph_init
+nve4_grctx_init_pd_0[] = {
 	{ 0x406020,   1, 0x04, 0x004103c1 },
 	{ 0x406028,   4, 0x04, 0x00000001 },
-	{}
-};
-
-static struct nvc0_graph_init
-nve4_grctx_init_unk64xx[] = {
 	{ 0x4064a8,   1, 0x04, 0x00000000 },
 	{ 0x4064ac,   1, 0x04, 0x00003fff },
 	{ 0x4064b4,   2, 0x04, 0x00000000 },
@@ -668,14 +667,14 @@ nve4_grctx_init_unk64xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_grctx_init_unk70xx[] = {
+static const struct nvc0_graph_init
+nve4_grctx_init_sked_0[] = {
 	{ 0x407040,   1, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nve4_grctx_init_unk80xx[] = {
+const struct nvc0_graph_init
+nve4_grctx_init_scc_0[] = {
 	{ 0x408000,   2, 0x04, 0x00000000 },
 	{ 0x408008,   1, 0x04, 0x00000030 },
 	{ 0x40800c,   2, 0x04, 0x00000000 },
@@ -685,8 +684,8 @@ nve4_grctx_init_unk80xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_grctx_init_rop[] = {
+static const struct nvc0_graph_init
+nve4_grctx_init_be_0[] = {
 	{ 0x408800,   1, 0x04, 0x02802a3c },
 	{ 0x408804,   1, 0x04, 0x00000040 },
 	{ 0x408808,   1, 0x04, 0x1043e005 },
@@ -698,22 +697,24 @@ nve4_grctx_init_rop[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_grctx_init_gpc_0[] = {
-	{ 0x418380,   1, 0x04, 0x00000016 },
-	{ 0x418400,   1, 0x04, 0x38004e00 },
-	{ 0x418404,   1, 0x04, 0x71e0ffff },
-	{ 0x41840c,   1, 0x04, 0x00001008 },
-	{ 0x418410,   1, 0x04, 0x0fff0fff },
-	{ 0x418414,   1, 0x04, 0x02200fff },
-	{ 0x418450,   6, 0x04, 0x00000000 },
-	{ 0x418468,   1, 0x04, 0x00000001 },
-	{ 0x41846c,   2, 0x04, 0x00000000 },
-	{ 0x418600,   1, 0x04, 0x0000001f },
-	{ 0x418684,   1, 0x04, 0x0000000f },
-	{ 0x418700,   1, 0x04, 0x00000002 },
-	{ 0x418704,   1, 0x04, 0x00000080 },
-	{ 0x418708,   3, 0x04, 0x00000000 },
+const struct nvc0_graph_pack
+nve4_grctx_pack_hub[] = {
+	{ nvc0_grctx_init_main_0 },
+	{ nve4_grctx_init_fe_0 },
+	{ nvc0_grctx_init_pri_0 },
+	{ nve4_grctx_init_memfmt_0 },
+	{ nve4_grctx_init_ds_0 },
+	{ nve4_grctx_init_cwd_0 },
+	{ nve4_grctx_init_pd_0 },
+	{ nve4_grctx_init_sked_0 },
+	{ nvc0_grctx_init_rstr2d_0 },
+	{ nve4_grctx_init_scc_0 },
+	{ nve4_grctx_init_be_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nve4_grctx_init_setup_0[] = {
 	{ 0x418800,   1, 0x04, 0x7006860a },
 	{ 0x418808,   3, 0x04, 0x00000000 },
 	{ 0x418828,   1, 0x04, 0x00000044 },
@@ -722,35 +723,35 @@ nve4_grctx_init_gpc_0[] = {
 	{ 0x4188e0,   1, 0x04, 0x01000000 },
 	{ 0x4188e8,   5, 0x04, 0x00000000 },
 	{ 0x4188fc,   1, 0x04, 0x20100018 },
-	{ 0x41891c,   1, 0x04, 0x00ff00ff },
-	{ 0x418924,   1, 0x04, 0x00000000 },
-	{ 0x418928,   1, 0x04, 0x00ffff00 },
-	{ 0x41892c,   1, 0x04, 0x0000ff00 },
-	{ 0x418b00,   1, 0x04, 0x00000006 },
-	{ 0x418b08,   1, 0x04, 0x0a418820 },
-	{ 0x418b0c,   1, 0x04, 0x062080e6 },
-	{ 0x418b10,   1, 0x04, 0x020398a4 },
-	{ 0x418b14,   1, 0x04, 0x0e629062 },
-	{ 0x418b18,   1, 0x04, 0x0a418820 },
-	{ 0x418b1c,   1, 0x04, 0x000000e6 },
-	{ 0x418bb8,   1, 0x04, 0x00000103 },
+	{}
+};
+
+const struct nvc0_graph_init
+nve4_grctx_init_gpm_0[] = {
 	{ 0x418c08,   1, 0x04, 0x00000001 },
 	{ 0x418c10,   8, 0x04, 0x00000000 },
 	{ 0x418c40,   1, 0x04, 0xffffffff },
 	{ 0x418c6c,   1, 0x04, 0x00000001 },
 	{ 0x418c80,   1, 0x04, 0x20200004 },
 	{ 0x418c8c,   1, 0x04, 0x00000001 },
-	{ 0x419000,   1, 0x04, 0x00000780 },
-	{ 0x419004,   2, 0x04, 0x00000000 },
-	{ 0x419014,   1, 0x04, 0x00000004 },
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_grctx_init_tpc[] = {
-	{ 0x419848,   1, 0x04, 0x00000000 },
-	{ 0x419864,   1, 0x04, 0x00000129 },
-	{ 0x419888,   1, 0x04, 0x00000000 },
+const struct nvc0_graph_pack
+nve4_grctx_pack_gpc[] = {
+	{ nvc0_grctx_init_gpc_unk_0 },
+	{ nvd9_grctx_init_prop_0 },
+	{ nvd9_grctx_init_gpc_unk_1 },
+	{ nve4_grctx_init_setup_0 },
+	{ nvc0_grctx_init_zcull_0 },
+	{ nvd9_grctx_init_crstr_0 },
+	{ nve4_grctx_init_gpm_0 },
+	{ nvc0_grctx_init_gcc_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nve4_grctx_init_tex_0[] = {
 	{ 0x419a00,   1, 0x04, 0x000000f0 },
 	{ 0x419a04,   1, 0x04, 0x00000001 },
 	{ 0x419a08,   1, 0x04, 0x00000021 },
@@ -761,14 +762,29 @@ nve4_grctx_init_tpc[] = {
 	{ 0x419a20,   1, 0x04, 0x00000800 },
 	{ 0x419a30,   1, 0x04, 0x00000001 },
 	{ 0x419ac4,   1, 0x04, 0x0037f440 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nve4_grctx_init_mpc_0[] = {
 	{ 0x419c00,   1, 0x04, 0x0000000a },
 	{ 0x419c04,   1, 0x04, 0x80000006 },
 	{ 0x419c08,   1, 0x04, 0x00000002 },
 	{ 0x419c20,   1, 0x04, 0x00000000 },
 	{ 0x419c24,   1, 0x04, 0x00084210 },
 	{ 0x419c28,   1, 0x04, 0x3efbefbe },
+	{}
+};
+
+static const struct nvc0_graph_init
+nve4_grctx_init_l1c_0[] = {
 	{ 0x419ce8,   1, 0x04, 0x00000000 },
 	{ 0x419cf4,   1, 0x04, 0x00003203 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nve4_grctx_init_sm_0[] = {
 	{ 0x419e04,   3, 0x04, 0x00000000 },
 	{ 0x419e10,   1, 0x04, 0x00000402 },
 	{ 0x419e44,   1, 0x04, 0x0013eff2 },
@@ -782,29 +798,47 @@ nve4_grctx_init_tpc[] = {
 	{ 0x419f58,   1, 0x04, 0x00000000 },
 	{ 0x419f70,   1, 0x04, 0x00000000 },
 	{ 0x419f78,   1, 0x04, 0x0000000b },
-	{ 0x419f7c,   1, 0x04, 0x0000027a },
+	{ 0x419f7c,   1, 0x04, 0x0000027c },
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_grctx_init_unk[] = {
+const struct nvc0_graph_pack
+nve4_grctx_pack_tpc[] = {
+	{ nvd7_grctx_init_pe_0 },
+	{ nve4_grctx_init_tex_0 },
+	{ nve4_grctx_init_mpc_0 },
+	{ nve4_grctx_init_l1c_0 },
+	{ nve4_grctx_init_sm_0 },
+	{}
+};
+
+const struct nvc0_graph_init
+nve4_grctx_init_pes_0[] = {
 	{ 0x41be24,   1, 0x04, 0x00000006 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nve4_grctx_init_cbm_0[] = {
 	{ 0x41bec0,   1, 0x04, 0x12180000 },
 	{ 0x41bec4,   1, 0x04, 0x00037f7f },
 	{ 0x41bee4,   1, 0x04, 0x06480430 },
-	{ 0x41bf00,   1, 0x04, 0x0a418820 },
-	{ 0x41bf04,   1, 0x04, 0x062080e6 },
-	{ 0x41bf08,   1, 0x04, 0x020398a4 },
-	{ 0x41bf0c,   1, 0x04, 0x0e629062 },
-	{ 0x41bf10,   1, 0x04, 0x0a418820 },
-	{ 0x41bf14,   1, 0x04, 0x000000e6 },
-	{ 0x41bfd0,   1, 0x04, 0x00900103 },
-	{ 0x41bfe0,   1, 0x04, 0x00400001 },
-	{ 0x41bfe4,   1, 0x04, 0x00000000 },
 	{}
 };
 
-static void
+const struct nvc0_graph_pack
+nve4_grctx_pack_ppc[] = {
+	{ nve4_grctx_init_pes_0 },
+	{ nve4_grctx_init_cbm_0 },
+	{ nvd7_grctx_init_wwdx_0 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
+void
 nve4_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 {
 	u32 magic[GPC_MAX][2];
@@ -925,10 +959,11 @@ nve4_grctx_generate_main(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 
 	nv_mask(priv, 0x000260, 0x00000001, 0x00000000);
 
-	for (i = 0; oclass->hub[i]; i++)
-		nvc0_graph_mmio(priv, oclass->hub[i]);
-	for (i = 0; oclass->gpc[i]; i++)
-		nvc0_graph_mmio(priv, oclass->gpc[i]);
+	nvc0_graph_mmio(priv, oclass->hub);
+	nvc0_graph_mmio(priv, oclass->gpc);
+	nvc0_graph_mmio(priv, oclass->zcull);
+	nvc0_graph_mmio(priv, oclass->tpc);
+	nvc0_graph_mmio(priv, oclass->ppc);
 
 	nv_wr32(priv, 0x404154, 0x00000000);
 
@@ -962,41 +997,6 @@ nve4_grctx_generate_main(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 	nv_mask(priv, 0x41be10, 0x00800000, 0x00800000);
 }
 
-static struct nvc0_graph_init *
-nve4_grctx_init_hub[] = {
-	nvc0_grctx_init_base,
-	nve4_grctx_init_unk40xx,
-	nvc0_grctx_init_unk44xx,
-	nve4_grctx_init_unk46xx,
-	nve4_grctx_init_unk47xx,
-	nve4_grctx_init_unk58xx,
-	nve4_grctx_init_unk5bxx,
-	nve4_grctx_init_unk60xx,
-	nve4_grctx_init_unk64xx,
-	nve4_grctx_init_unk70xx,
-	nvc0_grctx_init_unk78xx,
-	nve4_grctx_init_unk80xx,
-	nve4_grctx_init_rop,
-	NULL
-};
-
-struct nvc0_graph_init *
-nve4_grctx_init_gpc[] = {
-	nve4_grctx_init_gpc_0,
-	nvc0_grctx_init_gpc_1,
-	nve4_grctx_init_tpc,
-	nve4_grctx_init_unk,
-	NULL
-};
-
-static struct nvc0_graph_mthd
-nve4_grctx_init_mthd[] = {
-	{ 0xa097, nve4_grctx_init_a097, },
-	{ 0x902d, nvc0_grctx_init_902d, },
-	{ 0x902d, nvc0_grctx_init_mthd_magic, },
-	{}
-};
-
 struct nouveau_oclass *
 nve4_grctx_oclass = &(struct nvc0_grctx_oclass) {
 	.base.handle = NV_ENGCTX(GR, 0xe4),
@@ -1008,11 +1008,14 @@ nve4_grctx_oclass = &(struct nvc0_grctx_oclass) {
 		.rd32 = _nouveau_graph_context_rd32,
 		.wr32 = _nouveau_graph_context_wr32,
 	},
-	.main = nve4_grctx_generate_main,
-	.mods = nve4_grctx_generate_mods,
-	.unkn = nve4_grctx_generate_unkn,
-	.hub  = nve4_grctx_init_hub,
-	.gpc  = nve4_grctx_init_gpc,
-	.icmd = nve4_grctx_init_icmd,
-	.mthd = nve4_grctx_init_mthd,
+	.main  = nve4_grctx_generate_main,
+	.mods  = nve4_grctx_generate_mods,
+	.unkn  = nve4_grctx_generate_unkn,
+	.hub   = nve4_grctx_pack_hub,
+	.gpc   = nve4_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nve4_grctx_pack_tpc,
+	.ppc   = nve4_grctx_pack_ppc,
+	.icmd  = nve4_grctx_pack_icmd,
+	.mthd  = nve4_grctx_pack_mthd,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvf0.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvf0.c
index dcb2ebb8c29..dec03f04114 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvf0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvf0.c
@@ -22,10 +22,580 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 
-#include "nvc0.h"
+#include "ctxnvc0.h"
 
-static struct nvc0_graph_init
-nvf0_grctx_init_unk40xx[] = {
+/*******************************************************************************
+ * PGRAPH context register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+nvf0_grctx_init_icmd_0[] = {
+	{ 0x001000,   1, 0x01, 0x00000004 },
+	{ 0x000039,   3, 0x01, 0x00000000 },
+	{ 0x0000a9,   1, 0x01, 0x0000ffff },
+	{ 0x000038,   1, 0x01, 0x0fac6881 },
+	{ 0x00003d,   1, 0x01, 0x00000001 },
+	{ 0x0000e8,   8, 0x01, 0x00000400 },
+	{ 0x000078,   8, 0x01, 0x00000300 },
+	{ 0x000050,   1, 0x01, 0x00000011 },
+	{ 0x000058,   8, 0x01, 0x00000008 },
+	{ 0x000208,   8, 0x01, 0x00000001 },
+	{ 0x000081,   1, 0x01, 0x00000001 },
+	{ 0x000085,   1, 0x01, 0x00000004 },
+	{ 0x000088,   1, 0x01, 0x00000400 },
+	{ 0x000090,   1, 0x01, 0x00000300 },
+	{ 0x000098,   1, 0x01, 0x00001001 },
+	{ 0x0000e3,   1, 0x01, 0x00000001 },
+	{ 0x0000da,   1, 0x01, 0x00000001 },
+	{ 0x0000f8,   1, 0x01, 0x00000003 },
+	{ 0x0000fa,   1, 0x01, 0x00000001 },
+	{ 0x00009f,   4, 0x01, 0x0000ffff },
+	{ 0x0000b1,   1, 0x01, 0x00000001 },
+	{ 0x0000ad,   1, 0x01, 0x0000013e },
+	{ 0x0000e1,   1, 0x01, 0x00000010 },
+	{ 0x000290,  16, 0x01, 0x00000000 },
+	{ 0x0003b0,  16, 0x01, 0x00000000 },
+	{ 0x0002a0,  16, 0x01, 0x00000000 },
+	{ 0x000420,  16, 0x01, 0x00000000 },
+	{ 0x0002b0,  16, 0x01, 0x00000000 },
+	{ 0x000430,  16, 0x01, 0x00000000 },
+	{ 0x0002c0,  16, 0x01, 0x00000000 },
+	{ 0x0004d0,  16, 0x01, 0x00000000 },
+	{ 0x000720,  16, 0x01, 0x00000000 },
+	{ 0x0008c0,  16, 0x01, 0x00000000 },
+	{ 0x000890,  16, 0x01, 0x00000000 },
+	{ 0x0008e0,  16, 0x01, 0x00000000 },
+	{ 0x0008a0,  16, 0x01, 0x00000000 },
+	{ 0x0008f0,  16, 0x01, 0x00000000 },
+	{ 0x00094c,   1, 0x01, 0x000000ff },
+	{ 0x00094d,   1, 0x01, 0xffffffff },
+	{ 0x00094e,   1, 0x01, 0x00000002 },
+	{ 0x0002ec,   1, 0x01, 0x00000001 },
+	{ 0x0002f2,   2, 0x01, 0x00000001 },
+	{ 0x0002f5,   1, 0x01, 0x00000001 },
+	{ 0x0002f7,   1, 0x01, 0x00000001 },
+	{ 0x000303,   1, 0x01, 0x00000001 },
+	{ 0x0002e6,   1, 0x01, 0x00000001 },
+	{ 0x000466,   1, 0x01, 0x00000052 },
+	{ 0x000301,   1, 0x01, 0x3f800000 },
+	{ 0x000304,   1, 0x01, 0x30201000 },
+	{ 0x000305,   1, 0x01, 0x70605040 },
+	{ 0x000306,   1, 0x01, 0xb8a89888 },
+	{ 0x000307,   1, 0x01, 0xf8e8d8c8 },
+	{ 0x00030a,   1, 0x01, 0x00ffff00 },
+	{ 0x00030b,   1, 0x01, 0x0000001a },
+	{ 0x00030c,   1, 0x01, 0x00000001 },
+	{ 0x000318,   1, 0x01, 0x00000001 },
+	{ 0x000340,   1, 0x01, 0x00000000 },
+	{ 0x000375,   1, 0x01, 0x00000001 },
+	{ 0x00037d,   1, 0x01, 0x00000006 },
+	{ 0x0003a0,   1, 0x01, 0x00000002 },
+	{ 0x0003aa,   1, 0x01, 0x00000001 },
+	{ 0x0003a9,   1, 0x01, 0x00000001 },
+	{ 0x000380,   1, 0x01, 0x00000001 },
+	{ 0x000383,   1, 0x01, 0x00000011 },
+	{ 0x000360,   1, 0x01, 0x00000040 },
+	{ 0x000366,   2, 0x01, 0x00000000 },
+	{ 0x000368,   1, 0x01, 0x00000fff },
+	{ 0x000370,   2, 0x01, 0x00000000 },
+	{ 0x000372,   1, 0x01, 0x000fffff },
+	{ 0x00037a,   1, 0x01, 0x00000012 },
+	{ 0x000619,   1, 0x01, 0x00000003 },
+	{ 0x000811,   1, 0x01, 0x00000003 },
+	{ 0x000812,   1, 0x01, 0x00000004 },
+	{ 0x000813,   1, 0x01, 0x00000006 },
+	{ 0x000814,   1, 0x01, 0x00000008 },
+	{ 0x000815,   1, 0x01, 0x0000000b },
+	{ 0x000800,   6, 0x01, 0x00000001 },
+	{ 0x000632,   1, 0x01, 0x00000001 },
+	{ 0x000633,   1, 0x01, 0x00000002 },
+	{ 0x000634,   1, 0x01, 0x00000003 },
+	{ 0x000635,   1, 0x01, 0x00000004 },
+	{ 0x000654,   1, 0x01, 0x3f800000 },
+	{ 0x000657,   1, 0x01, 0x3f800000 },
+	{ 0x000655,   2, 0x01, 0x3f800000 },
+	{ 0x0006cd,   1, 0x01, 0x3f800000 },
+	{ 0x0007f5,   1, 0x01, 0x3f800000 },
+	{ 0x0007dc,   1, 0x01, 0x39291909 },
+	{ 0x0007dd,   1, 0x01, 0x79695949 },
+	{ 0x0007de,   1, 0x01, 0xb9a99989 },
+	{ 0x0007df,   1, 0x01, 0xf9e9d9c9 },
+	{ 0x0007e8,   1, 0x01, 0x00003210 },
+	{ 0x0007e9,   1, 0x01, 0x00007654 },
+	{ 0x0007ea,   1, 0x01, 0x00000098 },
+	{ 0x0007ec,   1, 0x01, 0x39291909 },
+	{ 0x0007ed,   1, 0x01, 0x79695949 },
+	{ 0x0007ee,   1, 0x01, 0xb9a99989 },
+	{ 0x0007ef,   1, 0x01, 0xf9e9d9c9 },
+	{ 0x0007f0,   1, 0x01, 0x00003210 },
+	{ 0x0007f1,   1, 0x01, 0x00007654 },
+	{ 0x0007f2,   1, 0x01, 0x00000098 },
+	{ 0x0005a5,   1, 0x01, 0x00000001 },
+	{ 0x000980, 128, 0x01, 0x00000000 },
+	{ 0x000468,   1, 0x01, 0x00000004 },
+	{ 0x00046c,   1, 0x01, 0x00000001 },
+	{ 0x000470,  96, 0x01, 0x00000000 },
+	{ 0x000510,  16, 0x01, 0x3f800000 },
+	{ 0x000520,   1, 0x01, 0x000002b6 },
+	{ 0x000529,   1, 0x01, 0x00000001 },
+	{ 0x000530,  16, 0x01, 0xffff0000 },
+	{ 0x000585,   1, 0x01, 0x0000003f },
+	{ 0x000576,   1, 0x01, 0x00000003 },
+	{ 0x00057b,   1, 0x01, 0x00000059 },
+	{ 0x000586,   1, 0x01, 0x00000040 },
+	{ 0x000582,   2, 0x01, 0x00000080 },
+	{ 0x0005c2,   1, 0x01, 0x00000001 },
+	{ 0x000638,   2, 0x01, 0x00000001 },
+	{ 0x00063a,   1, 0x01, 0x00000002 },
+	{ 0x00063b,   2, 0x01, 0x00000001 },
+	{ 0x00063d,   1, 0x01, 0x00000002 },
+	{ 0x00063e,   1, 0x01, 0x00000001 },
+	{ 0x0008b8,   8, 0x01, 0x00000001 },
+	{ 0x000900,   8, 0x01, 0x00000001 },
+	{ 0x000908,   8, 0x01, 0x00000002 },
+	{ 0x000910,  16, 0x01, 0x00000001 },
+	{ 0x000920,   8, 0x01, 0x00000002 },
+	{ 0x000928,   8, 0x01, 0x00000001 },
+	{ 0x000662,   1, 0x01, 0x00000001 },
+	{ 0x000648,   9, 0x01, 0x00000001 },
+	{ 0x000658,   1, 0x01, 0x0000000f },
+	{ 0x0007ff,   1, 0x01, 0x0000000a },
+	{ 0x00066a,   1, 0x01, 0x40000000 },
+	{ 0x00066b,   1, 0x01, 0x10000000 },
+	{ 0x00066c,   2, 0x01, 0xffff0000 },
+	{ 0x0007af,   2, 0x01, 0x00000008 },
+	{ 0x0007f6,   1, 0x01, 0x00000001 },
+	{ 0x00080b,   1, 0x01, 0x00000002 },
+	{ 0x0006b2,   1, 0x01, 0x00000055 },
+	{ 0x0007ad,   1, 0x01, 0x00000003 },
+	{ 0x000937,   1, 0x01, 0x00000001 },
+	{ 0x000971,   1, 0x01, 0x00000008 },
+	{ 0x000972,   1, 0x01, 0x00000040 },
+	{ 0x000973,   1, 0x01, 0x0000012c },
+	{ 0x00097c,   1, 0x01, 0x00000040 },
+	{ 0x000979,   1, 0x01, 0x00000003 },
+	{ 0x000975,   1, 0x01, 0x00000020 },
+	{ 0x000976,   1, 0x01, 0x00000001 },
+	{ 0x000977,   1, 0x01, 0x00000020 },
+	{ 0x000978,   1, 0x01, 0x00000001 },
+	{ 0x000957,   1, 0x01, 0x00000003 },
+	{ 0x00095e,   1, 0x01, 0x20164010 },
+	{ 0x00095f,   1, 0x01, 0x00000020 },
+	{ 0x000a0d,   1, 0x01, 0x00000006 },
+	{ 0x00097d,   1, 0x01, 0x00000020 },
+	{ 0x000683,   1, 0x01, 0x00000006 },
+	{ 0x000685,   1, 0x01, 0x003fffff },
+	{ 0x000687,   1, 0x01, 0x003fffff },
+	{ 0x0006a0,   1, 0x01, 0x00000005 },
+	{ 0x000840,   1, 0x01, 0x00400008 },
+	{ 0x000841,   1, 0x01, 0x08000080 },
+	{ 0x000842,   1, 0x01, 0x00400008 },
+	{ 0x000843,   1, 0x01, 0x08000080 },
+	{ 0x0006aa,   1, 0x01, 0x00000001 },
+	{ 0x0006ab,   1, 0x01, 0x00000002 },
+	{ 0x0006ac,   1, 0x01, 0x00000080 },
+	{ 0x0006ad,   2, 0x01, 0x00000100 },
+	{ 0x0006b1,   1, 0x01, 0x00000011 },
+	{ 0x0006bb,   1, 0x01, 0x000000cf },
+	{ 0x0006ce,   1, 0x01, 0x2a712488 },
+	{ 0x000739,   1, 0x01, 0x4085c000 },
+	{ 0x00073a,   1, 0x01, 0x00000080 },
+	{ 0x000786,   1, 0x01, 0x80000100 },
+	{ 0x00073c,   1, 0x01, 0x00010100 },
+	{ 0x00073d,   1, 0x01, 0x02800000 },
+	{ 0x000787,   1, 0x01, 0x000000cf },
+	{ 0x00078c,   1, 0x01, 0x00000008 },
+	{ 0x000792,   1, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
+	{ 0x000797,   1, 0x01, 0x000000cf },
+	{ 0x000836,   1, 0x01, 0x00000001 },
+	{ 0x00079a,   1, 0x01, 0x00000002 },
+	{ 0x000833,   1, 0x01, 0x04444480 },
+	{ 0x0007a1,   1, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
+	{ 0x000831,   1, 0x01, 0x00000004 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x000a04,   1, 0x01, 0x000000ff },
+	{ 0x000a0b,   1, 0x01, 0x00000040 },
+	{ 0x00097f,   1, 0x01, 0x00000100 },
+	{ 0x000a02,   1, 0x01, 0x00000001 },
+	{ 0x000809,   1, 0x01, 0x00000007 },
+	{ 0x00c221,   1, 0x01, 0x00000040 },
+	{ 0x00c1b0,   8, 0x01, 0x0000000f },
+	{ 0x00c1b8,   1, 0x01, 0x0fac6881 },
+	{ 0x00c1b9,   1, 0x01, 0x00fac688 },
+	{ 0x00c401,   1, 0x01, 0x00000001 },
+	{ 0x00c402,   1, 0x01, 0x00010001 },
+	{ 0x00c403,   2, 0x01, 0x00000001 },
+	{ 0x00c40e,   1, 0x01, 0x00000020 },
+	{ 0x00c500,   1, 0x01, 0x00000003 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000002 },
+	{ 0x0006aa,   1, 0x01, 0x00000001 },
+	{ 0x0006ad,   2, 0x01, 0x00000100 },
+	{ 0x0006b1,   1, 0x01, 0x00000011 },
+	{ 0x00078c,   1, 0x01, 0x00000008 },
+	{ 0x000792,   1, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
+	{ 0x000797,   1, 0x01, 0x000000cf },
+	{ 0x00079a,   1, 0x01, 0x00000002 },
+	{ 0x000833,   1, 0x01, 0x04444480 },
+	{ 0x0007a1,   1, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
+	{ 0x000831,   1, 0x01, 0x00000004 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000008 },
+	{ 0x000039,   3, 0x01, 0x00000000 },
+	{ 0x000380,   1, 0x01, 0x00000001 },
+	{ 0x000366,   2, 0x01, 0x00000000 },
+	{ 0x000368,   1, 0x01, 0x00000fff },
+	{ 0x000370,   2, 0x01, 0x00000000 },
+	{ 0x000372,   1, 0x01, 0x000fffff },
+	{ 0x000813,   1, 0x01, 0x00000006 },
+	{ 0x000814,   1, 0x01, 0x00000008 },
+	{ 0x000957,   1, 0x01, 0x00000003 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x000a04,   1, 0x01, 0x000000ff },
+	{ 0x000a0b,   1, 0x01, 0x00000040 },
+	{ 0x00097f,   1, 0x01, 0x00000100 },
+	{ 0x000a02,   1, 0x01, 0x00000001 },
+	{ 0x000809,   1, 0x01, 0x00000007 },
+	{ 0x00c221,   1, 0x01, 0x00000040 },
+	{ 0x00c401,   1, 0x01, 0x00000001 },
+	{ 0x00c402,   1, 0x01, 0x00010001 },
+	{ 0x00c403,   2, 0x01, 0x00000001 },
+	{ 0x00c40e,   1, 0x01, 0x00000020 },
+	{ 0x00c500,   1, 0x01, 0x00000003 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000001 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+nvf0_grctx_pack_icmd[] = {
+	{ nvf0_grctx_init_icmd_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvf0_grctx_init_a197_0[] = {
+	{ 0x000800,   8, 0x40, 0x00000000 },
+	{ 0x000804,   8, 0x40, 0x00000000 },
+	{ 0x000808,   8, 0x40, 0x00000400 },
+	{ 0x00080c,   8, 0x40, 0x00000300 },
+	{ 0x000810,   1, 0x04, 0x000000cf },
+	{ 0x000850,   7, 0x40, 0x00000000 },
+	{ 0x000814,   8, 0x40, 0x00000040 },
+	{ 0x000818,   8, 0x40, 0x00000001 },
+	{ 0x00081c,   8, 0x40, 0x00000000 },
+	{ 0x000820,   8, 0x40, 0x00000000 },
+	{ 0x001c00,  16, 0x10, 0x00000000 },
+	{ 0x001c04,  16, 0x10, 0x00000000 },
+	{ 0x001c08,  16, 0x10, 0x00000000 },
+	{ 0x001c0c,  16, 0x10, 0x00000000 },
+	{ 0x001d00,  16, 0x10, 0x00000000 },
+	{ 0x001d04,  16, 0x10, 0x00000000 },
+	{ 0x001d08,  16, 0x10, 0x00000000 },
+	{ 0x001d0c,  16, 0x10, 0x00000000 },
+	{ 0x001f00,  16, 0x08, 0x00000000 },
+	{ 0x001f04,  16, 0x08, 0x00000000 },
+	{ 0x001f80,  16, 0x08, 0x00000000 },
+	{ 0x001f84,  16, 0x08, 0x00000000 },
+	{ 0x002000,   1, 0x04, 0x00000000 },
+	{ 0x002040,   1, 0x04, 0x00000011 },
+	{ 0x002080,   1, 0x04, 0x00000020 },
+	{ 0x0020c0,   1, 0x04, 0x00000030 },
+	{ 0x002100,   1, 0x04, 0x00000040 },
+	{ 0x002140,   1, 0x04, 0x00000051 },
+	{ 0x00200c,   6, 0x40, 0x00000001 },
+	{ 0x002010,   1, 0x04, 0x00000000 },
+	{ 0x002050,   1, 0x04, 0x00000000 },
+	{ 0x002090,   1, 0x04, 0x00000001 },
+	{ 0x0020d0,   1, 0x04, 0x00000002 },
+	{ 0x002110,   1, 0x04, 0x00000003 },
+	{ 0x002150,   1, 0x04, 0x00000004 },
+	{ 0x000380,   4, 0x20, 0x00000000 },
+	{ 0x000384,   4, 0x20, 0x00000000 },
+	{ 0x000388,   4, 0x20, 0x00000000 },
+	{ 0x00038c,   4, 0x20, 0x00000000 },
+	{ 0x000700,   4, 0x10, 0x00000000 },
+	{ 0x000704,   4, 0x10, 0x00000000 },
+	{ 0x000708,   4, 0x10, 0x00000000 },
+	{ 0x002800, 128, 0x04, 0x00000000 },
+	{ 0x000a00,  16, 0x20, 0x00000000 },
+	{ 0x000a04,  16, 0x20, 0x00000000 },
+	{ 0x000a08,  16, 0x20, 0x00000000 },
+	{ 0x000a0c,  16, 0x20, 0x00000000 },
+	{ 0x000a10,  16, 0x20, 0x00000000 },
+	{ 0x000a14,  16, 0x20, 0x00000000 },
+	{ 0x000c00,  16, 0x10, 0x00000000 },
+	{ 0x000c04,  16, 0x10, 0x00000000 },
+	{ 0x000c08,  16, 0x10, 0x00000000 },
+	{ 0x000c0c,  16, 0x10, 0x3f800000 },
+	{ 0x000d00,   8, 0x08, 0xffff0000 },
+	{ 0x000d04,   8, 0x08, 0xffff0000 },
+	{ 0x000e00,  16, 0x10, 0x00000000 },
+	{ 0x000e04,  16, 0x10, 0xffff0000 },
+	{ 0x000e08,  16, 0x10, 0xffff0000 },
+	{ 0x000d40,   4, 0x08, 0x00000000 },
+	{ 0x000d44,   4, 0x08, 0x00000000 },
+	{ 0x001e00,   8, 0x20, 0x00000001 },
+	{ 0x001e04,   8, 0x20, 0x00000001 },
+	{ 0x001e08,   8, 0x20, 0x00000002 },
+	{ 0x001e0c,   8, 0x20, 0x00000001 },
+	{ 0x001e10,   8, 0x20, 0x00000001 },
+	{ 0x001e14,   8, 0x20, 0x00000002 },
+	{ 0x001e18,   8, 0x20, 0x00000001 },
+	{ 0x003400, 128, 0x04, 0x00000000 },
+	{ 0x00030c,   1, 0x04, 0x00000001 },
+	{ 0x001944,   1, 0x04, 0x00000000 },
+	{ 0x001514,   1, 0x04, 0x00000000 },
+	{ 0x000d68,   1, 0x04, 0x0000ffff },
+	{ 0x00121c,   1, 0x04, 0x0fac6881 },
+	{ 0x000fac,   1, 0x04, 0x00000001 },
+	{ 0x001538,   1, 0x04, 0x00000001 },
+	{ 0x000fe0,   2, 0x04, 0x00000000 },
+	{ 0x000fe8,   1, 0x04, 0x00000014 },
+	{ 0x000fec,   1, 0x04, 0x00000040 },
+	{ 0x000ff0,   1, 0x04, 0x00000000 },
+	{ 0x00179c,   1, 0x04, 0x00000000 },
+	{ 0x001228,   1, 0x04, 0x00000400 },
+	{ 0x00122c,   1, 0x04, 0x00000300 },
+	{ 0x001230,   1, 0x04, 0x00010001 },
+	{ 0x0007f8,   1, 0x04, 0x00000000 },
+	{ 0x0015b4,   1, 0x04, 0x00000001 },
+	{ 0x0015cc,   1, 0x04, 0x00000000 },
+	{ 0x001534,   1, 0x04, 0x00000000 },
+	{ 0x000fb0,   1, 0x04, 0x00000000 },
+	{ 0x0015d0,   1, 0x04, 0x00000000 },
+	{ 0x00153c,   1, 0x04, 0x00000000 },
+	{ 0x0016b4,   1, 0x04, 0x00000003 },
+	{ 0x000fbc,   4, 0x04, 0x0000ffff },
+	{ 0x000df8,   2, 0x04, 0x00000000 },
+	{ 0x001948,   1, 0x04, 0x00000000 },
+	{ 0x001970,   1, 0x04, 0x00000001 },
+	{ 0x00161c,   1, 0x04, 0x000009f0 },
+	{ 0x000dcc,   1, 0x04, 0x00000010 },
+	{ 0x00163c,   1, 0x04, 0x00000000 },
+	{ 0x0015e4,   1, 0x04, 0x00000000 },
+	{ 0x001160,  32, 0x04, 0x25e00040 },
+	{ 0x001880,  32, 0x04, 0x00000000 },
+	{ 0x000f84,   2, 0x04, 0x00000000 },
+	{ 0x0017c8,   2, 0x04, 0x00000000 },
+	{ 0x0017d0,   1, 0x04, 0x000000ff },
+	{ 0x0017d4,   1, 0x04, 0xffffffff },
+	{ 0x0017d8,   1, 0x04, 0x00000002 },
+	{ 0x0017dc,   1, 0x04, 0x00000000 },
+	{ 0x0015f4,   2, 0x04, 0x00000000 },
+	{ 0x001434,   2, 0x04, 0x00000000 },
+	{ 0x000d74,   1, 0x04, 0x00000000 },
+	{ 0x000dec,   1, 0x04, 0x00000001 },
+	{ 0x0013a4,   1, 0x04, 0x00000000 },
+	{ 0x001318,   1, 0x04, 0x00000001 },
+	{ 0x001644,   1, 0x04, 0x00000000 },
+	{ 0x000748,   1, 0x04, 0x00000000 },
+	{ 0x000de8,   1, 0x04, 0x00000000 },
+	{ 0x001648,   1, 0x04, 0x00000000 },
+	{ 0x0012a4,   1, 0x04, 0x00000000 },
+	{ 0x001120,   4, 0x04, 0x00000000 },
+	{ 0x001118,   1, 0x04, 0x00000000 },
+	{ 0x00164c,   1, 0x04, 0x00000000 },
+	{ 0x001658,   1, 0x04, 0x00000000 },
+	{ 0x001910,   1, 0x04, 0x00000290 },
+	{ 0x001518,   1, 0x04, 0x00000000 },
+	{ 0x00165c,   1, 0x04, 0x00000001 },
+	{ 0x001520,   1, 0x04, 0x00000000 },
+	{ 0x001604,   1, 0x04, 0x00000000 },
+	{ 0x001570,   1, 0x04, 0x00000000 },
+	{ 0x0013b0,   2, 0x04, 0x3f800000 },
+	{ 0x00020c,   1, 0x04, 0x00000000 },
+	{ 0x001670,   1, 0x04, 0x30201000 },
+	{ 0x001674,   1, 0x04, 0x70605040 },
+	{ 0x001678,   1, 0x04, 0xb8a89888 },
+	{ 0x00167c,   1, 0x04, 0xf8e8d8c8 },
+	{ 0x00166c,   1, 0x04, 0x00000000 },
+	{ 0x001680,   1, 0x04, 0x00ffff00 },
+	{ 0x0012d0,   1, 0x04, 0x00000003 },
+	{ 0x0012d4,   1, 0x04, 0x00000002 },
+	{ 0x001684,   2, 0x04, 0x00000000 },
+	{ 0x000dac,   2, 0x04, 0x00001b02 },
+	{ 0x000db4,   1, 0x04, 0x00000000 },
+	{ 0x00168c,   1, 0x04, 0x00000000 },
+	{ 0x0015bc,   1, 0x04, 0x00000000 },
+	{ 0x00156c,   1, 0x04, 0x00000000 },
+	{ 0x00187c,   1, 0x04, 0x00000000 },
+	{ 0x001110,   1, 0x04, 0x00000001 },
+	{ 0x000dc0,   3, 0x04, 0x00000000 },
+	{ 0x001234,   1, 0x04, 0x00000000 },
+	{ 0x001690,   1, 0x04, 0x00000000 },
+	{ 0x0012ac,   1, 0x04, 0x00000001 },
+	{ 0x0002c4,   1, 0x04, 0x00000000 },
+	{ 0x000790,   5, 0x04, 0x00000000 },
+	{ 0x00077c,   1, 0x04, 0x00000000 },
+	{ 0x001000,   1, 0x04, 0x00000010 },
+	{ 0x0010fc,   1, 0x04, 0x00000000 },
+	{ 0x001290,   1, 0x04, 0x00000000 },
+	{ 0x000218,   1, 0x04, 0x00000010 },
+	{ 0x0012d8,   1, 0x04, 0x00000000 },
+	{ 0x0012dc,   1, 0x04, 0x00000010 },
+	{ 0x000d94,   1, 0x04, 0x00000001 },
+	{ 0x00155c,   2, 0x04, 0x00000000 },
+	{ 0x001564,   1, 0x04, 0x00000fff },
+	{ 0x001574,   2, 0x04, 0x00000000 },
+	{ 0x00157c,   1, 0x04, 0x000fffff },
+	{ 0x001354,   1, 0x04, 0x00000000 },
+	{ 0x001610,   1, 0x04, 0x00000012 },
+	{ 0x001608,   2, 0x04, 0x00000000 },
+	{ 0x00260c,   1, 0x04, 0x00000000 },
+	{ 0x0007ac,   1, 0x04, 0x00000000 },
+	{ 0x00162c,   1, 0x04, 0x00000003 },
+	{ 0x000210,   1, 0x04, 0x00000000 },
+	{ 0x000320,   1, 0x04, 0x00000000 },
+	{ 0x000324,   6, 0x04, 0x3f800000 },
+	{ 0x000750,   1, 0x04, 0x00000000 },
+	{ 0x000760,   1, 0x04, 0x39291909 },
+	{ 0x000764,   1, 0x04, 0x79695949 },
+	{ 0x000768,   1, 0x04, 0xb9a99989 },
+	{ 0x00076c,   1, 0x04, 0xf9e9d9c9 },
+	{ 0x000770,   1, 0x04, 0x30201000 },
+	{ 0x000774,   1, 0x04, 0x70605040 },
+	{ 0x000778,   1, 0x04, 0x00009080 },
+	{ 0x000780,   1, 0x04, 0x39291909 },
+	{ 0x000784,   1, 0x04, 0x79695949 },
+	{ 0x000788,   1, 0x04, 0xb9a99989 },
+	{ 0x00078c,   1, 0x04, 0xf9e9d9c9 },
+	{ 0x0007d0,   1, 0x04, 0x30201000 },
+	{ 0x0007d4,   1, 0x04, 0x70605040 },
+	{ 0x0007d8,   1, 0x04, 0x00009080 },
+	{ 0x00037c,   1, 0x04, 0x00000001 },
+	{ 0x000740,   2, 0x04, 0x00000000 },
+	{ 0x002600,   1, 0x04, 0x00000000 },
+	{ 0x001918,   1, 0x04, 0x00000000 },
+	{ 0x00191c,   1, 0x04, 0x00000900 },
+	{ 0x001920,   1, 0x04, 0x00000405 },
+	{ 0x001308,   1, 0x04, 0x00000001 },
+	{ 0x001924,   1, 0x04, 0x00000000 },
+	{ 0x0013ac,   1, 0x04, 0x00000000 },
+	{ 0x00192c,   1, 0x04, 0x00000001 },
+	{ 0x00193c,   1, 0x04, 0x00002c1c },
+	{ 0x000d7c,   1, 0x04, 0x00000000 },
+	{ 0x000f8c,   1, 0x04, 0x00000000 },
+	{ 0x0002c0,   1, 0x04, 0x00000001 },
+	{ 0x001510,   1, 0x04, 0x00000000 },
+	{ 0x001940,   1, 0x04, 0x00000000 },
+	{ 0x000ff4,   2, 0x04, 0x00000000 },
+	{ 0x00194c,   2, 0x04, 0x00000000 },
+	{ 0x001968,   1, 0x04, 0x00000000 },
+	{ 0x001590,   1, 0x04, 0x0000003f },
+	{ 0x0007e8,   4, 0x04, 0x00000000 },
+	{ 0x00196c,   1, 0x04, 0x00000011 },
+	{ 0x0002e4,   1, 0x04, 0x0000b001 },
+	{ 0x00036c,   2, 0x04, 0x00000000 },
+	{ 0x00197c,   1, 0x04, 0x00000000 },
+	{ 0x000fcc,   2, 0x04, 0x00000000 },
+	{ 0x0002d8,   1, 0x04, 0x00000040 },
+	{ 0x001980,   1, 0x04, 0x00000080 },
+	{ 0x001504,   1, 0x04, 0x00000080 },
+	{ 0x001984,   1, 0x04, 0x00000000 },
+	{ 0x000300,   1, 0x04, 0x00000001 },
+	{ 0x0013a8,   1, 0x04, 0x00000000 },
+	{ 0x0012ec,   1, 0x04, 0x00000000 },
+	{ 0x001310,   1, 0x04, 0x00000000 },
+	{ 0x001314,   1, 0x04, 0x00000001 },
+	{ 0x001380,   1, 0x04, 0x00000000 },
+	{ 0x001384,   4, 0x04, 0x00000001 },
+	{ 0x001394,   1, 0x04, 0x00000000 },
+	{ 0x00139c,   1, 0x04, 0x00000000 },
+	{ 0x001398,   1, 0x04, 0x00000000 },
+	{ 0x001594,   1, 0x04, 0x00000000 },
+	{ 0x001598,   4, 0x04, 0x00000001 },
+	{ 0x000f54,   3, 0x04, 0x00000000 },
+	{ 0x0019bc,   1, 0x04, 0x00000000 },
+	{ 0x000f9c,   2, 0x04, 0x00000000 },
+	{ 0x0012cc,   1, 0x04, 0x00000000 },
+	{ 0x0012e8,   1, 0x04, 0x00000000 },
+	{ 0x00130c,   1, 0x04, 0x00000001 },
+	{ 0x001360,   8, 0x04, 0x00000000 },
+	{ 0x00133c,   2, 0x04, 0x00000001 },
+	{ 0x001344,   1, 0x04, 0x00000002 },
+	{ 0x001348,   2, 0x04, 0x00000001 },
+	{ 0x001350,   1, 0x04, 0x00000002 },
+	{ 0x001358,   1, 0x04, 0x00000001 },
+	{ 0x0012e4,   1, 0x04, 0x00000000 },
+	{ 0x00131c,   4, 0x04, 0x00000000 },
+	{ 0x0019c0,   1, 0x04, 0x00000000 },
+	{ 0x001140,   1, 0x04, 0x00000000 },
+	{ 0x0019c4,   1, 0x04, 0x00000000 },
+	{ 0x0019c8,   1, 0x04, 0x00001500 },
+	{ 0x00135c,   1, 0x04, 0x00000000 },
+	{ 0x000f90,   1, 0x04, 0x00000000 },
+	{ 0x0019e0,   8, 0x04, 0x00000001 },
+	{ 0x0019cc,   1, 0x04, 0x00000001 },
+	{ 0x0015b8,   1, 0x04, 0x00000000 },
+	{ 0x001a00,   1, 0x04, 0x00001111 },
+	{ 0x001a04,   7, 0x04, 0x00000000 },
+	{ 0x000d6c,   2, 0x04, 0xffff0000 },
+	{ 0x0010f8,   1, 0x04, 0x00001010 },
+	{ 0x000d80,   5, 0x04, 0x00000000 },
+	{ 0x000da0,   1, 0x04, 0x00000000 },
+	{ 0x0007a4,   2, 0x04, 0x00000000 },
+	{ 0x001508,   1, 0x04, 0x80000000 },
+	{ 0x00150c,   1, 0x04, 0x40000000 },
+	{ 0x001668,   1, 0x04, 0x00000000 },
+	{ 0x000318,   2, 0x04, 0x00000008 },
+	{ 0x000d9c,   1, 0x04, 0x00000001 },
+	{ 0x000ddc,   1, 0x04, 0x00000002 },
+	{ 0x000374,   1, 0x04, 0x00000000 },
+	{ 0x000378,   1, 0x04, 0x00000020 },
+	{ 0x0007dc,   1, 0x04, 0x00000000 },
+	{ 0x00074c,   1, 0x04, 0x00000055 },
+	{ 0x001420,   1, 0x04, 0x00000003 },
+	{ 0x0017bc,   2, 0x04, 0x00000000 },
+	{ 0x0017c4,   1, 0x04, 0x00000001 },
+	{ 0x001008,   1, 0x04, 0x00000008 },
+	{ 0x00100c,   1, 0x04, 0x00000040 },
+	{ 0x001010,   1, 0x04, 0x0000012c },
+	{ 0x000d60,   1, 0x04, 0x00000040 },
+	{ 0x00075c,   1, 0x04, 0x00000003 },
+	{ 0x001018,   1, 0x04, 0x00000020 },
+	{ 0x00101c,   1, 0x04, 0x00000001 },
+	{ 0x001020,   1, 0x04, 0x00000020 },
+	{ 0x001024,   1, 0x04, 0x00000001 },
+	{ 0x001444,   3, 0x04, 0x00000000 },
+	{ 0x000360,   1, 0x04, 0x20164010 },
+	{ 0x000364,   1, 0x04, 0x00000020 },
+	{ 0x000368,   1, 0x04, 0x00000000 },
+	{ 0x000de4,   1, 0x04, 0x00000000 },
+	{ 0x000204,   1, 0x04, 0x00000006 },
+	{ 0x000208,   1, 0x04, 0x00000000 },
+	{ 0x0002cc,   2, 0x04, 0x003fffff },
+	{ 0x001220,   1, 0x04, 0x00000005 },
+	{ 0x000fdc,   1, 0x04, 0x00000000 },
+	{ 0x000f98,   1, 0x04, 0x00400008 },
+	{ 0x001284,   1, 0x04, 0x08000080 },
+	{ 0x001450,   1, 0x04, 0x00400008 },
+	{ 0x001454,   1, 0x04, 0x08000080 },
+	{ 0x000214,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_pack
+nvf0_grctx_pack_mthd[] = {
+	{ nvf0_grctx_init_a197_0, 0xa197 },
+	{ nvc0_grctx_init_902d_0, 0x902d },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvf0_grctx_init_fe_0[] = {
 	{ 0x404004,   8, 0x04, 0x00000000 },
 	{ 0x404024,   1, 0x04, 0x0000e000 },
 	{ 0x404028,   8, 0x04, 0x00000000 },
@@ -50,8 +620,8 @@ nvf0_grctx_init_unk40xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_grctx_init_unk44xx[] = {
+const struct nvc0_graph_init
+nvf0_grctx_init_pri_0[] = {
 	{ 0x404404,  12, 0x04, 0x00000000 },
 	{ 0x404438,   1, 0x04, 0x00000000 },
 	{ 0x404460,   2, 0x04, 0x00000000 },
@@ -62,23 +632,18 @@ nvf0_grctx_init_unk44xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_grctx_init_unk5bxx[] = {
+const struct nvc0_graph_init
+nvf0_grctx_init_cwd_0[] = {
 	{ 0x405b00,   1, 0x04, 0x00000000 },
 	{ 0x405b10,   1, 0x04, 0x00001000 },
 	{ 0x405b20,   1, 0x04, 0x04000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_grctx_init_unk60xx[] = {
+static const struct nvc0_graph_init
+nvf0_grctx_init_pd_0[] = {
 	{ 0x406020,   1, 0x04, 0x034103c1 },
 	{ 0x406028,   4, 0x04, 0x00000001 },
-	{}
-};
-
-static struct nvc0_graph_init
-nvf0_grctx_init_unk64xx[] = {
 	{ 0x4064a8,   1, 0x04, 0x00000000 },
 	{ 0x4064ac,   1, 0x04, 0x00003fff },
 	{ 0x4064b0,   3, 0x04, 0x00000000 },
@@ -90,8 +655,8 @@ nvf0_grctx_init_unk64xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_grctx_init_unk88xx[] = {
+static const struct nvc0_graph_init
+nvf0_grctx_init_be_0[] = {
 	{ 0x408800,   1, 0x04, 0x12802a3c },
 	{ 0x408804,   1, 0x04, 0x00000040 },
 	{ 0x408808,   1, 0x04, 0x1003e005 },
@@ -103,22 +668,23 @@ nvf0_grctx_init_unk88xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_grctx_init_gpc_0[] = {
-	{ 0x418380,   1, 0x04, 0x00000016 },
-	{ 0x418400,   1, 0x04, 0x38004e00 },
-	{ 0x418404,   1, 0x04, 0x71e0ffff },
-	{ 0x41840c,   1, 0x04, 0x00001008 },
-	{ 0x418410,   1, 0x04, 0x0fff0fff },
-	{ 0x418414,   1, 0x04, 0x02200fff },
-	{ 0x418450,   6, 0x04, 0x00000000 },
-	{ 0x418468,   1, 0x04, 0x00000001 },
-	{ 0x41846c,   2, 0x04, 0x00000000 },
-	{ 0x418600,   1, 0x04, 0x0000001f },
-	{ 0x418684,   1, 0x04, 0x0000000f },
-	{ 0x418700,   1, 0x04, 0x00000002 },
-	{ 0x418704,   1, 0x04, 0x00000080 },
-	{ 0x418708,   3, 0x04, 0x00000000 },
+static const struct nvc0_graph_pack
+nvf0_grctx_pack_hub[] = {
+	{ nvc0_grctx_init_main_0 },
+	{ nvf0_grctx_init_fe_0 },
+	{ nvf0_grctx_init_pri_0 },
+	{ nve4_grctx_init_memfmt_0 },
+	{ nve4_grctx_init_ds_0 },
+	{ nvf0_grctx_init_cwd_0 },
+	{ nvf0_grctx_init_pd_0 },
+	{ nvc0_grctx_init_rstr2d_0 },
+	{ nve4_grctx_init_scc_0 },
+	{ nvf0_grctx_init_be_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvf0_grctx_init_setup_0[] = {
 	{ 0x418800,   1, 0x04, 0x7006860a },
 	{ 0x418808,   1, 0x04, 0x00000000 },
 	{ 0x41880c,   1, 0x04, 0x00000030 },
@@ -129,36 +695,31 @@ nvf0_grctx_init_gpc_0[] = {
 	{ 0x4188e0,   1, 0x04, 0x01000000 },
 	{ 0x4188e8,   5, 0x04, 0x00000000 },
 	{ 0x4188fc,   1, 0x04, 0x20100018 },
-	{ 0x41891c,   1, 0x04, 0x00ff00ff },
-	{ 0x418924,   1, 0x04, 0x00000000 },
-	{ 0x418928,   1, 0x04, 0x00ffff00 },
-	{ 0x41892c,   1, 0x04, 0x0000ff00 },
-	{ 0x418b00,   1, 0x04, 0x00000006 },
-	{ 0x418b08,   1, 0x04, 0x0a418820 },
-	{ 0x418b0c,   1, 0x04, 0x062080e6 },
-	{ 0x418b10,   1, 0x04, 0x020398a4 },
-	{ 0x418b14,   1, 0x04, 0x0e629062 },
-	{ 0x418b18,   1, 0x04, 0x0a418820 },
-	{ 0x418b1c,   1, 0x04, 0x000000e6 },
-	{ 0x418bb8,   1, 0x04, 0x00000103 },
-	{ 0x418c08,   1, 0x04, 0x00000001 },
-	{ 0x418c10,   8, 0x04, 0x00000000 },
-	{ 0x418c40,   1, 0x04, 0xffffffff },
-	{ 0x418c6c,   1, 0x04, 0x00000001 },
-	{ 0x418c80,   1, 0x04, 0x20200004 },
-	{ 0x418c8c,   1, 0x04, 0x00000001 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvf0_grctx_init_gpc_unk_2[] = {
 	{ 0x418d24,   1, 0x04, 0x00000000 },
-	{ 0x419000,   1, 0x04, 0x00000780 },
-	{ 0x419004,   2, 0x04, 0x00000000 },
-	{ 0x419014,   1, 0x04, 0x00000004 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_grctx_init_tpc[] = {
-	{ 0x419848,   1, 0x04, 0x00000000 },
-	{ 0x419864,   1, 0x04, 0x00000129 },
-	{ 0x419888,   1, 0x04, 0x00000000 },
+static const struct nvc0_graph_pack
+nvf0_grctx_pack_gpc[] = {
+	{ nvc0_grctx_init_gpc_unk_0 },
+	{ nvd9_grctx_init_prop_0 },
+	{ nvd9_grctx_init_gpc_unk_1 },
+	{ nvf0_grctx_init_setup_0 },
+	{ nvc0_grctx_init_zcull_0 },
+	{ nvd9_grctx_init_crstr_0 },
+	{ nve4_grctx_init_gpm_0 },
+	{ nvf0_grctx_init_gpc_unk_2 },
+	{ nvc0_grctx_init_gcc_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvf0_grctx_init_tex_0[] = {
 	{ 0x419a00,   1, 0x04, 0x000000f0 },
 	{ 0x419a04,   1, 0x04, 0x00000001 },
 	{ 0x419a08,   1, 0x04, 0x00000021 },
@@ -169,14 +730,29 @@ nvf0_grctx_init_tpc[] = {
 	{ 0x419a20,   1, 0x04, 0x00020800 },
 	{ 0x419a30,   1, 0x04, 0x00000001 },
 	{ 0x419ac4,   1, 0x04, 0x0037f440 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvf0_grctx_init_mpc_0[] = {
 	{ 0x419c00,   1, 0x04, 0x0000001a },
 	{ 0x419c04,   1, 0x04, 0x80000006 },
 	{ 0x419c08,   1, 0x04, 0x00000002 },
 	{ 0x419c20,   1, 0x04, 0x00000000 },
 	{ 0x419c24,   1, 0x04, 0x00084210 },
 	{ 0x419c28,   1, 0x04, 0x3efbefbe },
+	{}
+};
+
+const struct nvc0_graph_init
+nvf0_grctx_init_l1c_0[] = {
 	{ 0x419ce8,   1, 0x04, 0x00000000 },
 	{ 0x419cf4,   1, 0x04, 0x00000203 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvf0_grctx_init_sm_0[] = {
 	{ 0x419e04,   1, 0x04, 0x00000000 },
 	{ 0x419e08,   1, 0x04, 0x0000001d },
 	{ 0x419e0c,   1, 0x04, 0x00000000 },
@@ -189,8 +765,8 @@ nvf0_grctx_init_tpc[] = {
 	{ 0x419e5c,   3, 0x04, 0x00000000 },
 	{ 0x419e68,   1, 0x04, 0x00000002 },
 	{ 0x419e6c,  12, 0x04, 0x00000000 },
-	{ 0x419eac,   1, 0x04, 0x00001fcf },
-	{ 0x419eb0,   1, 0x04, 0x0db00da0 },
+	{ 0x419eac,   1, 0x04, 0x00001f8f },
+	{ 0x419eb0,   1, 0x04, 0x0db00d2f },
 	{ 0x419eb8,   1, 0x04, 0x00000000 },
 	{ 0x419ec8,   1, 0x04, 0x0001304f },
 	{ 0x419f30,   4, 0x04, 0x00000000 },
@@ -203,24 +779,36 @@ nvf0_grctx_init_tpc[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_grctx_init_unk[] = {
-	{ 0x41be24,   1, 0x04, 0x00000006 },
+static const struct nvc0_graph_pack
+nvf0_grctx_pack_tpc[] = {
+	{ nvd7_grctx_init_pe_0 },
+	{ nvf0_grctx_init_tex_0 },
+	{ nvf0_grctx_init_mpc_0 },
+	{ nvf0_grctx_init_l1c_0 },
+	{ nvf0_grctx_init_sm_0 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvf0_grctx_init_cbm_0[] = {
 	{ 0x41bec0,   1, 0x04, 0x10000000 },
 	{ 0x41bec4,   1, 0x04, 0x00037f7f },
 	{ 0x41bee4,   1, 0x04, 0x00000000 },
-	{ 0x41bf00,   1, 0x04, 0x0a418820 },
-	{ 0x41bf04,   1, 0x04, 0x062080e6 },
-	{ 0x41bf08,   1, 0x04, 0x020398a4 },
-	{ 0x41bf0c,   1, 0x04, 0x0e629062 },
-	{ 0x41bf10,   1, 0x04, 0x0a418820 },
-	{ 0x41bf14,   1, 0x04, 0x000000e6 },
-	{ 0x41bfd0,   1, 0x04, 0x00900103 },
-	{ 0x41bfe0,   1, 0x04, 0x00400001 },
-	{ 0x41bfe4,   1, 0x04, 0x00000000 },
 	{}
 };
 
+static const struct nvc0_graph_pack
+nvf0_grctx_pack_ppc[] = {
+	{ nve4_grctx_init_pes_0 },
+	{ nvf0_grctx_init_cbm_0 },
+	{ nvd7_grctx_init_wwdx_0 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
 static void
 nvf0_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 {
@@ -254,7 +842,7 @@ nvf0_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 		u16 magic3 = 0x0648;
 		magic[gpc][0]  = 0x10000000 | (magic0 << 16) | offset;
 		magic[gpc][1]  = 0x00000000 | (magic1 << 16);
-		offset += 0x0324 * (priv->tpc_nr[gpc] - 1);;
+		offset += 0x0324 * (priv->tpc_nr[gpc] - 1);
 		magic[gpc][2]  = 0x10000000 | (magic2 << 16) | offset;
 		magic[gpc][3]  = 0x00000000 | (magic3 << 16);
 		offset += 0x0324;
@@ -273,40 +861,6 @@ nvf0_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 	mmio_list(0x17e920, 0x00090a05, 0, 0);
 }
 
-static struct nvc0_graph_init *
-nvf0_grctx_init_hub[] = {
-	nvc0_grctx_init_base,
-	nvf0_grctx_init_unk40xx,
-	nvf0_grctx_init_unk44xx,
-	nve4_grctx_init_unk46xx,
-	nve4_grctx_init_unk47xx,
-	nve4_grctx_init_unk58xx,
-	nvf0_grctx_init_unk5bxx,
-	nvf0_grctx_init_unk60xx,
-	nvf0_grctx_init_unk64xx,
-	nve4_grctx_init_unk80xx,
-	nvf0_grctx_init_unk88xx,
-	nvd9_grctx_init_rop,
-	NULL
-};
-
-struct nvc0_graph_init *
-nvf0_grctx_init_gpc[] = {
-	nvf0_grctx_init_gpc_0,
-	nvc0_grctx_init_gpc_1,
-	nvf0_grctx_init_tpc,
-	nvf0_grctx_init_unk,
-	NULL
-};
-
-static struct nvc0_graph_mthd
-nvf0_grctx_init_mthd[] = {
-	{ 0xa197, nvc1_grctx_init_9097, },
-	{ 0x902d, nvc0_grctx_init_902d, },
-	{ 0x902d, nvc0_grctx_init_mthd_magic, },
-	{}
-};
-
 struct nouveau_oclass *
 nvf0_grctx_oclass = &(struct nvc0_grctx_oclass) {
 	.base.handle = NV_ENGCTX(GR, 0xf0),
@@ -318,11 +872,14 @@ nvf0_grctx_oclass = &(struct nvc0_grctx_oclass) {
 		.rd32 = _nouveau_graph_context_rd32,
 		.wr32 = _nouveau_graph_context_wr32,
 	},
-	.main = nve4_grctx_generate_main,
-	.mods = nvf0_grctx_generate_mods,
-	.unkn = nve4_grctx_generate_unkn,
-	.hub  = nvf0_grctx_init_hub,
-	.gpc  = nvf0_grctx_init_gpc,
-	.icmd = nvc0_grctx_init_icmd,
-	.mthd = nvf0_grctx_init_mthd,
+	.main  = nve4_grctx_generate_main,
+	.mods  = nvf0_grctx_generate_mods,
+	.unkn  = nve4_grctx_generate_unkn,
+	.hub   = nvf0_grctx_pack_hub,
+	.gpc   = nvf0_grctx_pack_gpc,
+	.zcull = nvc0_grctx_pack_zcull,
+	.tpc   = nvf0_grctx_pack_tpc,
+	.ppc   = nvf0_grctx_pack_ppc,
+	.icmd  = nvf0_grctx_pack_icmd,
+	.mthd  = nvf0_grctx_pack_mthd,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/com.fuc b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/com.fuc
index 5d24b6de16c..e37d8106ae1 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/com.fuc
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/com.fuc
@@ -38,7 +38,7 @@ queue_put:
 	cmpu b32 $r8 $r9
 	bra ne #queue_put_next
 		mov $r15 E_CMD_OVERFLOW
-		call #error
+		call(error)
 		ret
 
 	// store cmd/data on queue
@@ -92,18 +92,16 @@ queue_get_done:
 // Out: $r15 value
 //
 nv_rd32:
-	mov $r11 0x728
-	shl b32 $r11 6
 	mov b32 $r12 $r14
 	bset $r12 31			// MMIO_CTRL_PENDING
-	iowr I[$r11 + 0x000] $r12	// MMIO_CTRL
+	nv_iowr(NV_PGRAPH_FECS_MMIO_CTRL, 0, $r12)
 	nv_rd32_wait:
-		iord $r12 I[$r11 + 0x000]
+		nv_iord($r12, NV_PGRAPH_FECS_MMIO_CTRL, 0)
 		xbit $r12 $r12 31
 		bra ne #nv_rd32_wait
 	mov $r10 6			// DONE_MMIO_RD
-	call #wait_doneo
-	iord $r15 I[$r11 + 0x100]	// MMIO_RDVAL
+	call(wait_doneo)
+	nv_iord($r15, NV_PGRAPH_FECS_MMIO_RDVAL, 0)
 	ret
 
 // nv_wr32 - write 32-bit value to nv register
@@ -112,37 +110,17 @@ nv_rd32:
 //      $r15 value
 //
 nv_wr32:
-	mov $r11 0x728
-	shl b32 $r11 6
-	iowr I[$r11 + 0x200] $r15	// MMIO_WRVAL
+	nv_iowr(NV_PGRAPH_FECS_MMIO_WRVAL, 0, $r15)
 	mov b32 $r12 $r14
 	bset $r12 31			// MMIO_CTRL_PENDING
 	bset $r12 30			// MMIO_CTRL_WRITE
-	iowr I[$r11 + 0x000] $r12	// MMIO_CTRL
+	nv_iowr(NV_PGRAPH_FECS_MMIO_CTRL, 0, $r12)
 	nv_wr32_wait:
-		iord $r12 I[$r11 + 0x000]
+		nv_iord($r12, NV_PGRAPH_FECS_MMIO_CTRL, 0)
 		xbit $r12 $r12 31
 		bra ne #nv_wr32_wait
 	ret
 
-// (re)set watchdog timer
-//
-// In : $r15 timeout
-//
-watchdog_reset:
-	mov $r8 0x430
-	shl b32 $r8 6
-	bset $r15 31
-	iowr I[$r8 + 0x000] $r15
-	ret
-
-// clear watchdog timer
-watchdog_clear:
-	mov $r8 0x430
-	shl b32 $r8 6
-	iowr I[$r8 + 0x000] $r0
-	ret
-
 // wait_donez - wait on FUC_DONE bit to become clear
 //
 // In : $r10 bit to wait on
@@ -163,13 +141,9 @@ wait_donez:
 //
 wait_doneo:
 	trace_set(T_WAIT);
-	mov $r8 0x818
-	shl b32 $r8 6
-	iowr I[$r8 + 0x000] $r10
+	nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_VAL(6), 0, $r10)
 	wait_doneo_e:
-		mov $r8 0x400
-		shl b32 $r8 6
-		iord $r8 I[$r8 + 0x000]
+		nv_iord($r8, NV_PGRAPH_FECS_SIGNAL, 0)
 		xbit $r8 $r8 $r10
 		bra e #wait_doneo_e
 	trace_clr(T_WAIT)
@@ -209,21 +183,18 @@ mmctx_size:
 //
 mmctx_xfer:
 	trace_set(T_MMCTX)
-	mov $r8 0x710
-	shl b32 $r8 6
 	clear b32 $r9
 	or $r11 $r11
 	bra e #mmctx_base_disabled
-		iowr I[$r8 + 0x000] $r11	// MMCTX_BASE
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_BASE, 0, $r11)
 		bset $r9 0			// BASE_EN
 	mmctx_base_disabled:
 	or $r14 $r14
 	bra e #mmctx_multi_disabled
-		iowr I[$r8 + 0x200] $r14 	// MMCTX_MULTI_STRIDE
-		iowr I[$r8 + 0x300] $r15 	// MMCTX_MULTI_MASK
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_MULTI_STRIDE, 0, $r14)
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_MULTI_MASK, 0, $r15)
 		bset $r9 1			// MULTI_EN
 	mmctx_multi_disabled:
-	add b32 $r8 0x100
 
 	xbit $r11 $r10 0
 	shl b32 $r11 16			// DIR
@@ -231,20 +202,20 @@ mmctx_xfer:
 	xbit $r14 $r10 1
 	shl b32 $r14 17
 	or $r11 $r14			// START_TRIGGER
-	iowr I[$r8 + 0x000] $r11	// MMCTX_CTRL
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_CTRL, 0, $r11)
 
 	// loop over the mmio list, and send requests to the hw
 	mmctx_exec_loop:
 		// wait for space in mmctx queue
 		mmctx_wait_free:
-			iord $r14 I[$r8 + 0x000] // MMCTX_CTRL
+			nv_iord($r14, NV_PGRAPH_FECS_MMCTX_CTRL, 0)
 			and $r14 0x1f
 			bra e #mmctx_wait_free
 
 		// queue up an entry
 		ld b32 $r14 D[$r12]
 		or $r14 $r9
-		iowr I[$r8 + 0x300] $r14
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_QUEUE, 0, $r14)
 		add b32 $r12 4
 		cmpu b32 $r12 $r13
 		bra ne #mmctx_exec_loop
@@ -253,22 +224,22 @@ mmctx_xfer:
 	bra ne #mmctx_stop
 		// wait for queue to empty
 		mmctx_fini_wait:
-			iord $r11 I[$r8 + 0x000]	// MMCTX_CTRL
+			nv_iord($r11, NV_PGRAPH_FECS_MMCTX_CTRL, 0)
 			and $r11 0x1f
 			cmpu b32 $r11 0x10
 			bra ne #mmctx_fini_wait
-		mov $r10 2				// DONE_MMCTX
-		call #wait_donez
+		mov $r10 5			// DONE_MMCTX
+		call(wait_donez)
 		bra #mmctx_done
 	mmctx_stop:
 		xbit $r11 $r10 0
 		shl b32 $r11 16			// DIR
 		bset $r11 12			// QLIMIT = 0x10
 		bset $r11 18			// STOP_TRIGGER
-		iowr I[$r8 + 0x000] $r11	// MMCTX_CTRL
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_CTRL, 0, $r11)
 		mmctx_stop_wait:
 			// wait for STOP_TRIGGER to clear
-			iord $r11 I[$r8 + 0x000] // MMCTX_CTRL
+			nv_iord($r11, NV_PGRAPH_FECS_MMCTX_CTRL, 0)
 			xbit $r11 $r11 18
 			bra ne #mmctx_stop_wait
 	mmctx_done:
@@ -280,28 +251,24 @@ mmctx_xfer:
 strand_wait:
 	push $r10
 	mov $r10 2
-	call #wait_donez
+	call(wait_donez)
 	pop $r10
 	ret
 
 // unknown - call before issuing strand commands
 //
 strand_pre:
-	mov $r8 0x4afc
-	sethi $r8 0x20000
-	mov $r9 0xc
-	iowr I[$r8] $r9
-	call #strand_wait
+	mov $r9 NV_PGRAPH_FECS_STRAND_CMD_ENABLE
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r9)
+	call(strand_wait)
 	ret
 
 // unknown - call after issuing strand commands
 //
 strand_post:
-	mov $r8 0x4afc
-	sethi $r8 0x20000
-	mov $r9 0xd
-	iowr I[$r8] $r9
-	call #strand_wait
+	mov $r9 NV_PGRAPH_FECS_STRAND_CMD_DISABLE
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r9)
+	call(strand_wait)
 	ret
 
 // Selects strand set?!
@@ -309,18 +276,14 @@ strand_post:
 // In: $r14 id
 //
 strand_set:
-	mov $r10 0x4ffc
-	sethi $r10 0x20000
-	sub b32 $r11 $r10 0x500
 	mov $r12 0xf
-	iowr I[$r10 + 0x000] $r12		// 0x93c = 0xf
-	mov $r12 0xb
-	iowr I[$r11 + 0x000] $r12		// 0x928 = 0xb
-	call #strand_wait
-	iowr I[$r10 + 0x000] $r14		// 0x93c = <id>
-	mov $r12 0xa
-	iowr I[$r11 + 0x000] $r12		// 0x928 = 0xa
-	call #strand_wait
+	nv_iowr(NV_PGRAPH_FECS_STRAND_FILTER, 0x3f, $r12)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_DEACTIVATE_FILTER
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	nv_iowr(NV_PGRAPH_FECS_STRAND_FILTER, 0x3f, $r14)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_ACTIVATE_FILTER
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	call(strand_wait)
 	ret
 
 // Initialise strand context data
@@ -332,30 +295,27 @@ strand_set:
 //
 strand_ctx_init:
 	trace_set(T_STRINIT)
-	call #strand_pre
+	call(strand_pre)
 	mov $r14 3
-	call #strand_set
-	mov $r10 0x46fc
-	sethi $r10 0x20000
-	add b32 $r11 $r10 0x400
-	iowr I[$r10 + 0x100] $r0	// STRAND_FIRST_GENE = 0
-	mov $r12 1
-	iowr I[$r11 + 0x000] $r12	// STRAND_CMD = LATCH_FIRST_GENE
-	call #strand_wait
+	call(strand_set)
+
+	clear b32 $r12
+	nv_iowr(NV_PGRAPH_FECS_STRAND_SELECT, 0x3f, $r12)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_SEEK
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	call(strand_wait)
 	sub b32 $r12 $r0 1
-	iowr I[$r10 + 0x000] $r12	// STRAND_GENE_CNT = 0xffffffff
-	mov $r12 2
-	iowr I[$r11 + 0x000] $r12	// STRAND_CMD = LATCH_GENE_CNT
-	call #strand_wait
-	call #strand_post
+	nv_iowr(NV_PGRAPH_FECS_STRAND_DATA, 0x3f, $r12)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_GET_INFO
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	call(strand_wait)
+	call(strand_post)
 
 	// read the size of each strand, poke the context offset of
 	// each into STRAND_{SAVE,LOAD}_SWBASE now, no need to worry
 	// about it later then.
-	mov $r8 0x880
-	shl b32 $r8 6
-	iord $r9 I[$r8 + 0x000]		// STRANDS
-	add b32 $r8 0x2200
+	nv_mkio($r8, NV_PGRAPH_FECS_STRAND_SAVE_SWBASE, 0x00)
+	nv_iord($r9, NV_PGRAPH_FECS_STRANDS_CNT, 0x00)
 	shr b32 $r14 $r15 8
 	ctx_init_strand_loop:
 		iowr I[$r8 + 0x000] $r14	// STRAND_SAVE_SWBASE
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpc.fuc b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpc.fuc
index 5547c1b3f4f..7445f12b1d9 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpc.fuc
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpc.fuc
@@ -54,16 +54,13 @@ mmio_list_base:
 #ifdef INCLUDE_CODE
 // reports an exception to the host
 //
-// In: $r15 error code (see nvc0.fuc)
+// In: $r15 error code (see os.h)
 //
 error:
 	push $r14
-	mov $r14 -0x67ec 	// 0x9814
-	sethi $r14 0x400000
-	call #nv_wr32		// HUB_CTXCTL_CC_SCRATCH[5] = error code
-	add b32 $r14 0x41c
+	nv_wr32(NV_PGRAPH_FECS_CC_SCRATCH_VAL(5), $r15)
 	mov $r15 1
-	call #nv_wr32		// HUB_CTXCTL_INTR_UP_SET
+	nv_wr32(NV_PGRAPH_FECS_INTR_UP_SET, $r15)
 	pop $r14
 	ret
 
@@ -81,49 +78,48 @@ error:
 //
 init:
 	clear b32 $r0
-	mov $sp $r0
+
+	// setup stack
+	nv_iord($r1, NV_PGRAPH_GPCX_GPCCS_CAPS, 0)
+	extr $r1 $r1 9:17
+	shl b32 $r1 8
+	mov $sp $r1
 
 	// enable fifo access
-	mov $r1 0x1200
-	mov $r2 2
-	iowr I[$r1 + 0x000] $r2		// FIFO_ENABLE
+	mov $r2 NV_PGRAPH_GPCX_GPCCS_ACCESS_FIFO
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_ACCESS, 0, $r2)
 
 	// setup i0 handler, and route all interrupts to it
 	mov $r1 #ih
 	mov $iv0 $r1
-	mov $r1 0x400
-	iowr I[$r1 + 0x300] $r0		// INTR_DISPATCH
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_INTR_ROUTE, 0, $r0)
 
 	// enable fifo interrupt
-	mov $r2 4
-	iowr I[$r1 + 0x000] $r2		// INTR_EN_SET
+	mov $r2 NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET_FIFO
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET, 0, $r2)
 
 	// enable interrupts
 	bset $flags ie0
 
 	// figure out which GPC we are, and how many TPCs we have
-	mov $r1 0x608
-	shl b32 $r1 6
-	iord $r2 I[$r1 + 0x000]		// UNITS
+	nv_iord($r2, NV_PGRAPH_GPCX_GPCCS_UNITS, 0)
 	mov $r3 1
 	and $r2 0x1f
 	shl b32 $r3 $r2
 	sub b32 $r3 1
 	st b32 D[$r0 + #tpc_count] $r2
 	st b32 D[$r0 + #tpc_mask] $r3
-	add b32 $r1 0x400
-	iord $r2 I[$r1 + 0x000]		// MYINDEX
+	nv_iord($r2, NV_PGRAPH_GPCX_GPCCS_MYINDEX, 0)
 	st b32 D[$r0 + #gpc_id] $r2
 
 #if NV_PGRAPH_GPCX_UNK__SIZE > 0
 	// figure out which, and how many, UNKs are actually present
-	mov $r14 0x0c30
-	sethi $r14 0x500000
+	imm32($r14, 0x500c30)
 	clear b32 $r2
 	clear b32 $r3
 	clear b32 $r4
 	init_unk_loop:
-		call #nv_rd32
+		call(nv_rd32)
 		cmp b32 $r15 0
 		bra z #init_unk_next
 			mov $r15 1
@@ -146,23 +142,21 @@ init:
 
 	// set mmctx base addresses now so we don't have to do it later,
 	// they don't currently ever change
-	mov $r4 0x700
-	shl b32 $r4 6
 	shr b32 $r5 $r2 8
-	iowr I[$r4 + 0x000] $r5		// MMCTX_SAVE_SWBASE
-	iowr I[$r4 + 0x100] $r5		// MMCTX_LOAD_SWBASE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MMCTX_SAVE_SWBASE, 0, $r5)
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_SWBASE, 0, $r5)
 
 	// calculate GPC mmio context size
 	ld b32 $r14 D[$r0 + #gpc_mmio_list_head]
 	ld b32 $r15 D[$r0 + #gpc_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 	add b32 $r2 $r15
 	add b32 $r3 $r15
 
 	// calculate per-TPC mmio context size
 	ld b32 $r14 D[$r0 + #tpc_mmio_list_head]
 	ld b32 $r15 D[$r0 + #tpc_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 	ld b32 $r14 D[$r0 + #tpc_count]
 	mulu $r14 $r15
 	add b32 $r2 $r14
@@ -172,7 +166,7 @@ init:
 	// calculate per-UNK mmio context size
 	ld b32 $r14 D[$r0 + #unk_mmio_list_head]
 	ld b32 $r15 D[$r0 + #unk_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 	ld b32 $r14 D[$r0 + #unk_count]
 	mulu $r14 $r15
 	add b32 $r2 $r14
@@ -180,9 +174,8 @@ init:
 #endif
 
 	// round up base/size to 256 byte boundary (for strand SWBASE)
-	add b32 $r4 0x1300
 	shr b32 $r3 2
-	iowr I[$r4 + 0x000] $r3		// MMCTX_LOAD_COUNT, wtf for?!?
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_COUNT, 0, $r3) // wtf for?!
 	shr b32 $r2 8
 	shr b32 $r3 6
 	add b32 $r2 1
@@ -192,7 +185,7 @@ init:
 
 	// calculate size of strand context data
 	mov b32 $r15 $r2
-	call #strand_ctx_init
+	call(strand_ctx_init)
 	add b32 $r3 $r15
 
 	// save context size, and tell HUB we're done
@@ -208,7 +201,7 @@ main:
 	bset $flags $p0
 	sleep $p0
 	mov $r13 #cmd_queue
-	call #queue_get
+	call(queue_get)
 	bra $p1 #main
 
 	// 0x0000-0x0003 are all context transfers
@@ -224,13 +217,13 @@ main:
 		or $r1 $r14
 		mov $flags $r1
 		// transfer context data
-		call #ctx_xfer
+		call(ctx_xfer)
 		bra #main
 
 	main_not_ctx_xfer:
 	shl b32 $r15 $r14 16
 	or $r15 E_BAD_COMMAND
-	call #error
+	call(error)
 	bra #main
 
 // interrupt handler
@@ -247,22 +240,20 @@ ih:
 	clear b32 $r0
 
 	// incoming fifo command?
-	iord $r10 I[$r0 + 0x200]	// INTR
-	and $r11 $r10 0x00000004
+	nv_iord($r10, NV_PGRAPH_GPCX_GPCCS_INTR, 0)
+	and $r11 $r10 NV_PGRAPH_GPCX_GPCCS_INTR_FIFO
 	bra e #ih_no_fifo
 		// queue incoming fifo command for later processing
-		mov $r11 0x1900
 		mov $r13 #cmd_queue
-		iord $r14 I[$r11 + 0x100]	// FIFO_CMD
-		iord $r15 I[$r11 + 0x000]	// FIFO_DATA
-		call #queue_put
-		add b32 $r11 0x400
+		nv_iord($r14, NV_PGRAPH_GPCX_GPCCS_FIFO_CMD, 0)
+		nv_iord($r15, NV_PGRAPH_GPCX_GPCCS_FIFO_DATA, 0)
+		call(queue_put)
 		mov $r14 1
-		iowr I[$r11 + 0x000] $r14	// FIFO_ACK
+		nv_iowr(NV_PGRAPH_GPCX_GPCCS_FIFO_ACK, 0, $r14)
 
 	// ack, and wake up main()
 	ih_no_fifo:
-	iowr I[$r0 + 0x100] $r10	// INTR_ACK
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_INTR_ACK, 0, $r10)
 
 	pop $r15
 	pop $r14
@@ -283,9 +274,7 @@ hub_barrier_done:
 	mov $r15 1
 	ld b32 $r14 D[$r0 + #gpc_id]
 	shl b32 $r15 $r14
-	mov $r14 -0x6be8 	// 0x409418 - HUB_BAR_SET
-	sethi $r14 0x400000
-	call #nv_wr32
+	nv_wr32(0x409418, $r15)	// 0x409418 - HUB_BAR_SET
 	ret
 
 // Disables various things, waits a bit, and re-enables them..
@@ -295,16 +284,15 @@ hub_barrier_done:
 // funny things happen.
 //
 ctx_redswitch:
-	mov $r14 0x614
-	shl b32 $r14 6
-	mov $r15 0x020
-	iowr I[$r14] $r15	// GPC_RED_SWITCH = POWER
-	mov $r15 8
+	mov $r15 NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_POWER
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_RED_SWITCH, 0, $r15)
+	mov $r14 8
 	ctx_redswitch_delay:
-		sub b32 $r15 1
+		sub b32 $r14 1
 		bra ne #ctx_redswitch_delay
-	mov $r15 0xa20
-	iowr I[$r14] $r15	// GPC_RED_SWITCH = UNK11, ENABLE, POWER
+	or $r15 NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_UNK11
+	or $r15 NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_ENABLE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_RED_SWITCH, 0, $r15)
 	ret
 
 // Transfer GPC context data between GPU and storage area
@@ -317,46 +305,37 @@ ctx_redswitch:
 //
 ctx_xfer:
 	// set context base address
-	mov $r1 0xa04
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r15// MEM_BASE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MEM_BASE, 0, $r15)
 	bra not $p1 #ctx_xfer_not_load
-		call #ctx_redswitch
+		call(ctx_redswitch)
 	ctx_xfer_not_load:
 
 	// strands
-	mov $r1 0x4afc
-	sethi $r1 0x20000
-	mov $r2 0xc
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x0c
-	call #strand_wait
-	mov $r2 0x47fc
-	sethi $r2 0x20000
-	iowr I[$r2] $r0		// STRAND_FIRST_GENE(0x3f) = 0x00
-	xbit $r2 $flags $p1
-	add b32 $r2 3
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x03/0x04 (SAVE/LOAD)
+	call(strand_pre)
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_STRAND_SELECT, 0x3f, $r2)
+	xbit $r2 $flags $p1	// SAVE/LOAD
+	add b32 $r2 NV_PGRAPH_GPCX_GPCCS_STRAND_CMD_SAVE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_STRAND_CMD, 0x3f, $r2)
 
 	// mmio context
 	xbit $r10 $flags $p1	// direction
 	or $r10 2		// first
-	mov $r11 0x0000
-	sethi $r11 0x500000
+	imm32($r11,0x500000)
 	ld b32 $r12 D[$r0 + #gpc_id]
 	shl b32 $r12 15
 	add b32 $r11 $r12	// base = NV_PGRAPH_GPCn
 	ld b32 $r12 D[$r0 + #gpc_mmio_list_head]
 	ld b32 $r13 D[$r0 + #gpc_mmio_list_tail]
 	mov $r14 0		// not multi
-	call #mmctx_xfer
+	call(mmctx_xfer)
 
 	// per-TPC mmio context
 	xbit $r10 $flags $p1	// direction
 #if !NV_PGRAPH_GPCX_UNK__SIZE
 	or $r10 4		// last
 #endif
-	mov $r11 0x4000
-	sethi $r11 0x500000	// base = NV_PGRAPH_GPC0_TPC0
+	imm32($r11, 0x504000)
 	ld b32 $r12 D[$r0 + #gpc_id]
 	shl b32 $r12 15
 	add b32 $r11 $r12	// base = NV_PGRAPH_GPCn_TPC0
@@ -364,14 +343,13 @@ ctx_xfer:
 	ld b32 $r13 D[$r0 + #tpc_mmio_list_tail]
 	ld b32 $r15 D[$r0 + #tpc_mask]
 	mov $r14 0x800		// stride = 0x800
-	call #mmctx_xfer
+	call(mmctx_xfer)
 
 #if NV_PGRAPH_GPCX_UNK__SIZE > 0
 	// per-UNK mmio context
 	xbit $r10 $flags $p1	// direction
 	or $r10 4		// last
-	mov $r11 0x3000
-	sethi $r11 0x500000	// base = NV_PGRAPH_GPC0_UNK0
+	imm32($r11, 0x503000)
 	ld b32 $r12 D[$r0 + #gpc_id]
 	shl b32 $r12 15
 	add b32 $r11 $r12	// base = NV_PGRAPH_GPCn_UNK0
@@ -379,11 +357,11 @@ ctx_xfer:
 	ld b32 $r13 D[$r0 + #unk_mmio_list_tail]
 	ld b32 $r15 D[$r0 + #unk_mask]
 	mov $r14 0x200		// stride = 0x200
-	call #mmctx_xfer
+	call(mmctx_xfer)
 #endif
 
 	// wait for strands to finish
-	call #strand_wait
+	call(strand_wait)
 
 	// if load, or a save without a load following, do some
 	// unknown stuff that's done after finishing a block of
@@ -391,14 +369,10 @@ ctx_xfer:
 	bra $p1 #ctx_xfer_post
 	bra not $p2 #ctx_xfer_done
 	ctx_xfer_post:
-		mov $r1 0x4afc
-		sethi $r1 0x20000
-		mov $r2 0xd
-		iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x0d
-		call #strand_wait
+		call(strand_post)
 
 	// mark completion in HUB's barrier
 	ctx_xfer_done:
-	call #hub_barrier_done
+	call(hub_barrier_done)
 	ret
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcgm107.fuc5 b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcgm107.fuc5
new file mode 100644
index 00000000000..e730603891d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcgm107.fuc5
@@ -0,0 +1,42 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#define NV_PGRAPH_GPCX_UNK__SIZE                                     0x00000002
+
+#define CHIPSET GK208
+#include "macros.fuc"
+
+.section #gm107_grgpc_data
+#define INCLUDE_DATA
+#include "com.fuc"
+#include "gpc.fuc"
+#undef INCLUDE_DATA
+
+.section #gm107_grgpc_code
+#define INCLUDE_CODE
+bra #init
+#include "com.fuc"
+#include "gpc.fuc"
+.align 256
+#undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcgm107.fuc5.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcgm107.fuc5.h
new file mode 100644
index 00000000000..6d53b67dd3c
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcgm107.fuc5.h
@@ -0,0 +1,473 @@
+uint32_t gm107_grgpc_data[] = {
+/* 0x0000: gpc_mmio_list_head */
+	0x0000006c,
+/* 0x0004: gpc_mmio_list_tail */
+/* 0x0004: tpc_mmio_list_head */
+	0x0000006c,
+/* 0x0008: tpc_mmio_list_tail */
+/* 0x0008: unk_mmio_list_head */
+	0x0000006c,
+/* 0x000c: unk_mmio_list_tail */
+	0x0000006c,
+/* 0x0010: gpc_id */
+	0x00000000,
+/* 0x0014: tpc_count */
+	0x00000000,
+/* 0x0018: tpc_mask */
+	0x00000000,
+/* 0x001c: unk_count */
+	0x00000000,
+/* 0x0020: unk_mask */
+	0x00000000,
+/* 0x0024: cmd_queue */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+};
+
+uint32_t gm107_grgpc_code[] = {
+	0x03140ef5,
+/* 0x0004: queue_put */
+	0x9800d898,
+	0x86f001d9,
+	0xf489a408,
+	0x020f0b1b,
+	0x0002f87e,
+/* 0x001a: queue_put_next */
+	0x98c400f8,
+	0x0384b607,
+	0xb6008dbb,
+	0x8eb50880,
+	0x018fb500,
+	0xf00190b6,
+	0xd9b50f94,
+/* 0x0037: queue_get */
+	0xf400f801,
+	0xd8980131,
+	0x01d99800,
+	0x0bf489a4,
+	0x0789c421,
+	0xbb0394b6,
+	0x90b6009d,
+	0x009e9808,
+	0xb6019f98,
+	0x84f00180,
+	0x00d8b50f,
+/* 0x0063: queue_get_done */
+	0xf80132f4,
+/* 0x0065: nv_rd32 */
+	0xf0ecb200,
+	0x00801fc9,
+	0x0cf601ca,
+/* 0x0073: nv_rd32_wait */
+	0x8c04bd00,
+	0xcf01ca00,
+	0xccc800cc,
+	0xf61bf41f,
+	0xec7e060a,
+	0x008f0000,
+	0xffcf01cb,
+/* 0x008f: nv_wr32 */
+	0x8000f800,
+	0xf601cc00,
+	0x04bd000f,
+	0xc9f0ecb2,
+	0x1ec9f01f,
+	0x01ca0080,
+	0xbd000cf6,
+/* 0x00a9: nv_wr32_wait */
+	0xca008c04,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f61b,
+/* 0x00b8: wait_donez */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x00cf: wait_donez_ne */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf61bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x00ec: wait_doneo */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x0103: wait_doneo_e */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf60bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x0120: mmctx_size */
+/* 0x0122: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0x1bf4efa4,
+	0xf89fb2ec,
+/* 0x013d: mmctx_xfer */
+	0xf094bd00,
+	0x00800199,
+	0x09f60237,
+	0xbd04bd00,
+	0x05bbfd94,
+	0x800f0bf4,
+	0xf601c400,
+	0x04bd000b,
+/* 0x015f: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0xc6008018,
+	0x000ef601,
+	0x008004bd,
+	0x0ff601c7,
+	0xf004bd00,
+/* 0x017a: mmctx_multi_disabled */
+	0xabc80199,
+	0x10b4b600,
+	0xc80cb9f0,
+	0xe4b601ae,
+	0x05befd11,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x0195: mmctx_exec_loop */
+/* 0x0195: mmctx_wait_free */
+	0xc5008e04,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f60b,
+	0x05e9fd00,
+	0x01c80080,
+	0xbd000ef6,
+	0x04c0b604,
+	0x1bf4cda4,
+	0x02abc8df,
+/* 0x01bf: mmctx_fini_wait */
+	0x8b1c1bf4,
+	0xcf01c500,
+	0xb4f000bb,
+	0x10b4b01f,
+	0x0af31bf4,
+	0x00b87e05,
+	0x250ef400,
+/* 0x01d8: mmctx_stop */
+	0xb600abc8,
+	0xb9f010b4,
+	0x12b9f00c,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x01ed: mmctx_stop_wait */
+	0xc5008b04,
+	0x00bbcf01,
+	0xf412bbc8,
+/* 0x01fa: mmctx_done */
+	0x94bdf61b,
+	0x800199f0,
+	0xf6021700,
+	0x04bd0009,
+/* 0x020a: strand_wait */
+	0xa0f900f8,
+	0xb87e020a,
+	0xa0fc0000,
+/* 0x0216: strand_pre */
+	0x0c0900f8,
+	0x024afc80,
+	0xbd0009f6,
+	0x020a7e04,
+/* 0x0227: strand_post */
+	0x0900f800,
+	0x4afc800d,
+	0x0009f602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0238: strand_set */
+	0xfc800f0c,
+	0x0cf6024f,
+	0x0c04bd00,
+	0x4afc800b,
+	0x000cf602,
+	0xfc8004bd,
+	0x0ef6024f,
+	0x0c04bd00,
+	0x4afc800a,
+	0x000cf602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0268: strand_ctx_init */
+	0x99f094bd,
+	0x37008003,
+	0x0009f602,
+	0x167e04bd,
+	0x030e0002,
+	0x0002387e,
+	0xfc80c4bd,
+	0x0cf60247,
+	0x0c04bd00,
+	0x4afc8001,
+	0x000cf602,
+	0x0a7e04bd,
+	0x0c920002,
+	0x46fc8001,
+	0x000cf602,
+	0x020c04bd,
+	0x024afc80,
+	0xbd000cf6,
+	0x020a7e04,
+	0x02277e00,
+	0x42008800,
+	0x20008902,
+	0x0099cf02,
+/* 0x02c7: ctx_init_strand_loop */
+	0xf608fe95,
+	0x8ef6008e,
+	0x808acf40,
+	0xb606a5b6,
+	0xeabb01a0,
+	0x0480b600,
+	0xf40192b6,
+	0xe4b6e81b,
+	0xf2efbc08,
+	0x99f094bd,
+	0x17008003,
+	0x0009f602,
+	0x00f804bd,
+/* 0x02f8: error */
+	0xffb2e0f9,
+	0x4098148e,
+	0x00008f7e,
+	0xffb2010f,
+	0x409c1c8e,
+	0x00008f7e,
+	0x00f8e0fc,
+/* 0x0314: init */
+	0x004104bd,
+	0x0011cf42,
+	0x010911e7,
+	0xfe0814b6,
+	0x02020014,
+	0xf6120040,
+	0x04bd0002,
+	0xfe047241,
+	0x00400010,
+	0x0000f607,
+	0x040204bd,
+	0xf6040040,
+	0x04bd0002,
+	0x821031f4,
+	0xcf018200,
+	0x01030022,
+	0xbb1f24f0,
+	0x32b60432,
+	0x0502b501,
+	0x820603b5,
+	0xcf018600,
+	0x02b50022,
+	0x0c308e04,
+	0xbd24bd50,
+/* 0x0377: init_unk_loop */
+	0x7e44bd34,
+	0xb0000065,
+	0x0bf400f6,
+	0xbb010f0e,
+	0x4ffd04f2,
+	0x0130b605,
+/* 0x038c: init_unk_next */
+	0xb60120b6,
+	0x26b004e0,
+	0xe21bf402,
+/* 0x0398: init_unk_done */
+	0xb50703b5,
+	0x00820804,
+	0x22cf0201,
+	0x9534bd00,
+	0x00800825,
+	0x05f601c0,
+	0x8004bd00,
+	0xf601c100,
+	0x04bd0005,
+	0x98000e98,
+	0x207e010f,
+	0x2fbb0001,
+	0x003fbb00,
+	0x98010e98,
+	0x207e020f,
+	0x0e980001,
+	0x00effd05,
+	0xbb002ebb,
+	0x0e98003e,
+	0x030f9802,
+	0x0001207e,
+	0xfd070e98,
+	0x2ebb00ef,
+	0x003ebb00,
+	0x800235b6,
+	0xf601d300,
+	0x04bd0003,
+	0xb60825b6,
+	0x20b60635,
+	0x0130b601,
+	0xb60824b6,
+	0x2fb20834,
+	0x0002687e,
+	0x80003fbb,
+	0xf6020100,
+	0x04bd0003,
+	0x29f024bd,
+	0x3000801f,
+	0x0002f602,
+/* 0x0436: main */
+	0x31f404bd,
+	0x0028f400,
+	0x377e240d,
+	0x01f40000,
+	0x04e4b0f4,
+	0xfe1d18f4,
+	0x06020181,
+	0x12fd20bd,
+	0x01e4b604,
+	0xfe051efd,
+	0x097e0018,
+	0x0ef40005,
+/* 0x0465: main_not_ctx_xfer */
+	0x10ef94d4,
+	0x7e01f5f0,
+	0xf40002f8,
+/* 0x0472: ih */
+	0x80f9c70e,
+	0xf90188fe,
+	0xf990f980,
+	0xf9b0f9a0,
+	0xf9e0f9d0,
+	0x4a04bdf0,
+	0xaacf0200,
+	0x04abc400,
+	0x0d1f0bf4,
+	0x1a004e24,
+	0x4f00eecf,
+	0xffcf1900,
+	0x00047e00,
+	0x40010e00,
+	0x0ef61d00,
+/* 0x04af: ih_no_fifo */
+	0x4004bd00,
+	0x0af60100,
+	0xfc04bd00,
+	0xfce0fcf0,
+	0xfcb0fcd0,
+	0xfc90fca0,
+	0x0088fe80,
+	0x32f480fc,
+/* 0x04cf: hub_barrier_done */
+	0x0f01f800,
+	0x040e9801,
+	0xb204febb,
+	0x94188eff,
+	0x008f7e40,
+/* 0x04e3: ctx_redswitch */
+	0x0f00f800,
+	0x85008020,
+	0x000ff601,
+	0x080e04bd,
+/* 0x04f0: ctx_redswitch_delay */
+	0xf401e2b6,
+	0xf5f1fd1b,
+	0xf5f10800,
+	0x00800200,
+	0x0ff60185,
+	0xf804bd00,
+/* 0x0509: ctx_xfer */
+	0x81008000,
+	0x000ff602,
+	0x11f404bd,
+	0x04e37e07,
+/* 0x0519: ctx_xfer_not_load */
+	0x02167e00,
+	0x8024bd00,
+	0xf60247fc,
+	0x04bd0002,
+	0xb6012cf0,
+	0xfc800320,
+	0x02f6024a,
+	0xf004bd00,
+	0xa5f001ac,
+	0x00008b02,
+	0x040c9850,
+	0xbb0fc4b6,
+	0x0c9800bc,
+	0x010d9800,
+	0x3d7e000e,
+	0xacf00001,
+	0x40008b01,
+	0x040c9850,
+	0xbb0fc4b6,
+	0x0c9800bc,
+	0x020d9801,
+	0x4e060f98,
+	0x3d7e0800,
+	0xacf00001,
+	0x04a5f001,
+	0x5030008b,
+	0xb6040c98,
+	0xbcbb0fc4,
+	0x020c9800,
+	0x98030d98,
+	0x004e080f,
+	0x013d7e02,
+	0x020a7e00,
+	0x0601f400,
+/* 0x05a3: ctx_xfer_post */
+	0x7e0712f4,
+/* 0x05a7: ctx_xfer_done */
+	0x7e000227,
+	0xf80004cf,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+};
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5 b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5
new file mode 100644
index 00000000000..bd30262d635
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5
@@ -0,0 +1,42 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#define NV_PGRAPH_GPCX_UNK__SIZE                                     0x00000001
+
+#define CHIPSET GK208
+#include "macros.fuc"
+
+.section #nv108_grgpc_data
+#define INCLUDE_DATA
+#include "com.fuc"
+#include "gpc.fuc"
+#undef INCLUDE_DATA
+
+.section #nv108_grgpc_code
+#define INCLUDE_CODE
+bra #init
+#include "com.fuc"
+#include "gpc.fuc"
+.align 256
+#undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5.h
new file mode 100644
index 00000000000..31922707794
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5.h
@@ -0,0 +1,473 @@
+uint32_t nv108_grgpc_data[] = {
+/* 0x0000: gpc_mmio_list_head */
+	0x0000006c,
+/* 0x0004: gpc_mmio_list_tail */
+/* 0x0004: tpc_mmio_list_head */
+	0x0000006c,
+/* 0x0008: tpc_mmio_list_tail */
+/* 0x0008: unk_mmio_list_head */
+	0x0000006c,
+/* 0x000c: unk_mmio_list_tail */
+	0x0000006c,
+/* 0x0010: gpc_id */
+	0x00000000,
+/* 0x0014: tpc_count */
+	0x00000000,
+/* 0x0018: tpc_mask */
+	0x00000000,
+/* 0x001c: unk_count */
+	0x00000000,
+/* 0x0020: unk_mask */
+	0x00000000,
+/* 0x0024: cmd_queue */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+};
+
+uint32_t nv108_grgpc_code[] = {
+	0x03140ef5,
+/* 0x0004: queue_put */
+	0x9800d898,
+	0x86f001d9,
+	0xf489a408,
+	0x020f0b1b,
+	0x0002f87e,
+/* 0x001a: queue_put_next */
+	0x98c400f8,
+	0x0384b607,
+	0xb6008dbb,
+	0x8eb50880,
+	0x018fb500,
+	0xf00190b6,
+	0xd9b50f94,
+/* 0x0037: queue_get */
+	0xf400f801,
+	0xd8980131,
+	0x01d99800,
+	0x0bf489a4,
+	0x0789c421,
+	0xbb0394b6,
+	0x90b6009d,
+	0x009e9808,
+	0xb6019f98,
+	0x84f00180,
+	0x00d8b50f,
+/* 0x0063: queue_get_done */
+	0xf80132f4,
+/* 0x0065: nv_rd32 */
+	0xf0ecb200,
+	0x00801fc9,
+	0x0cf601ca,
+/* 0x0073: nv_rd32_wait */
+	0x8c04bd00,
+	0xcf01ca00,
+	0xccc800cc,
+	0xf61bf41f,
+	0xec7e060a,
+	0x008f0000,
+	0xffcf01cb,
+/* 0x008f: nv_wr32 */
+	0x8000f800,
+	0xf601cc00,
+	0x04bd000f,
+	0xc9f0ecb2,
+	0x1ec9f01f,
+	0x01ca0080,
+	0xbd000cf6,
+/* 0x00a9: nv_wr32_wait */
+	0xca008c04,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f61b,
+/* 0x00b8: wait_donez */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x00cf: wait_donez_ne */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf61bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x00ec: wait_doneo */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x0103: wait_doneo_e */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf60bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x0120: mmctx_size */
+/* 0x0122: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0x1bf4efa4,
+	0xf89fb2ec,
+/* 0x013d: mmctx_xfer */
+	0xf094bd00,
+	0x00800199,
+	0x09f60237,
+	0xbd04bd00,
+	0x05bbfd94,
+	0x800f0bf4,
+	0xf601c400,
+	0x04bd000b,
+/* 0x015f: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0xc6008018,
+	0x000ef601,
+	0x008004bd,
+	0x0ff601c7,
+	0xf004bd00,
+/* 0x017a: mmctx_multi_disabled */
+	0xabc80199,
+	0x10b4b600,
+	0xc80cb9f0,
+	0xe4b601ae,
+	0x05befd11,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x0195: mmctx_exec_loop */
+/* 0x0195: mmctx_wait_free */
+	0xc5008e04,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f60b,
+	0x05e9fd00,
+	0x01c80080,
+	0xbd000ef6,
+	0x04c0b604,
+	0x1bf4cda4,
+	0x02abc8df,
+/* 0x01bf: mmctx_fini_wait */
+	0x8b1c1bf4,
+	0xcf01c500,
+	0xb4f000bb,
+	0x10b4b01f,
+	0x0af31bf4,
+	0x00b87e05,
+	0x250ef400,
+/* 0x01d8: mmctx_stop */
+	0xb600abc8,
+	0xb9f010b4,
+	0x12b9f00c,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x01ed: mmctx_stop_wait */
+	0xc5008b04,
+	0x00bbcf01,
+	0xf412bbc8,
+/* 0x01fa: mmctx_done */
+	0x94bdf61b,
+	0x800199f0,
+	0xf6021700,
+	0x04bd0009,
+/* 0x020a: strand_wait */
+	0xa0f900f8,
+	0xb87e020a,
+	0xa0fc0000,
+/* 0x0216: strand_pre */
+	0x0c0900f8,
+	0x024afc80,
+	0xbd0009f6,
+	0x020a7e04,
+/* 0x0227: strand_post */
+	0x0900f800,
+	0x4afc800d,
+	0x0009f602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0238: strand_set */
+	0xfc800f0c,
+	0x0cf6024f,
+	0x0c04bd00,
+	0x4afc800b,
+	0x000cf602,
+	0xfc8004bd,
+	0x0ef6024f,
+	0x0c04bd00,
+	0x4afc800a,
+	0x000cf602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0268: strand_ctx_init */
+	0x99f094bd,
+	0x37008003,
+	0x0009f602,
+	0x167e04bd,
+	0x030e0002,
+	0x0002387e,
+	0xfc80c4bd,
+	0x0cf60247,
+	0x0c04bd00,
+	0x4afc8001,
+	0x000cf602,
+	0x0a7e04bd,
+	0x0c920002,
+	0x46fc8001,
+	0x000cf602,
+	0x020c04bd,
+	0x024afc80,
+	0xbd000cf6,
+	0x020a7e04,
+	0x02277e00,
+	0x42008800,
+	0x20008902,
+	0x0099cf02,
+/* 0x02c7: ctx_init_strand_loop */
+	0xf608fe95,
+	0x8ef6008e,
+	0x808acf40,
+	0xb606a5b6,
+	0xeabb01a0,
+	0x0480b600,
+	0xf40192b6,
+	0xe4b6e81b,
+	0xf2efbc08,
+	0x99f094bd,
+	0x17008003,
+	0x0009f602,
+	0x00f804bd,
+/* 0x02f8: error */
+	0xffb2e0f9,
+	0x4098148e,
+	0x00008f7e,
+	0xffb2010f,
+	0x409c1c8e,
+	0x00008f7e,
+	0x00f8e0fc,
+/* 0x0314: init */
+	0x004104bd,
+	0x0011cf42,
+	0x010911e7,
+	0xfe0814b6,
+	0x02020014,
+	0xf6120040,
+	0x04bd0002,
+	0xfe047241,
+	0x00400010,
+	0x0000f607,
+	0x040204bd,
+	0xf6040040,
+	0x04bd0002,
+	0x821031f4,
+	0xcf018200,
+	0x01030022,
+	0xbb1f24f0,
+	0x32b60432,
+	0x0502b501,
+	0x820603b5,
+	0xcf018600,
+	0x02b50022,
+	0x0c308e04,
+	0xbd24bd50,
+/* 0x0377: init_unk_loop */
+	0x7e44bd34,
+	0xb0000065,
+	0x0bf400f6,
+	0xbb010f0e,
+	0x4ffd04f2,
+	0x0130b605,
+/* 0x038c: init_unk_next */
+	0xb60120b6,
+	0x26b004e0,
+	0xe21bf401,
+/* 0x0398: init_unk_done */
+	0xb50703b5,
+	0x00820804,
+	0x22cf0201,
+	0x9534bd00,
+	0x00800825,
+	0x05f601c0,
+	0x8004bd00,
+	0xf601c100,
+	0x04bd0005,
+	0x98000e98,
+	0x207e010f,
+	0x2fbb0001,
+	0x003fbb00,
+	0x98010e98,
+	0x207e020f,
+	0x0e980001,
+	0x00effd05,
+	0xbb002ebb,
+	0x0e98003e,
+	0x030f9802,
+	0x0001207e,
+	0xfd070e98,
+	0x2ebb00ef,
+	0x003ebb00,
+	0x800235b6,
+	0xf601d300,
+	0x04bd0003,
+	0xb60825b6,
+	0x20b60635,
+	0x0130b601,
+	0xb60824b6,
+	0x2fb20834,
+	0x0002687e,
+	0x80003fbb,
+	0xf6020100,
+	0x04bd0003,
+	0x29f024bd,
+	0x3000801f,
+	0x0002f602,
+/* 0x0436: main */
+	0x31f404bd,
+	0x0028f400,
+	0x377e240d,
+	0x01f40000,
+	0x04e4b0f4,
+	0xfe1d18f4,
+	0x06020181,
+	0x12fd20bd,
+	0x01e4b604,
+	0xfe051efd,
+	0x097e0018,
+	0x0ef40005,
+/* 0x0465: main_not_ctx_xfer */
+	0x10ef94d4,
+	0x7e01f5f0,
+	0xf40002f8,
+/* 0x0472: ih */
+	0x80f9c70e,
+	0xf90188fe,
+	0xf990f980,
+	0xf9b0f9a0,
+	0xf9e0f9d0,
+	0x4a04bdf0,
+	0xaacf0200,
+	0x04abc400,
+	0x0d1f0bf4,
+	0x1a004e24,
+	0x4f00eecf,
+	0xffcf1900,
+	0x00047e00,
+	0x40010e00,
+	0x0ef61d00,
+/* 0x04af: ih_no_fifo */
+	0x4004bd00,
+	0x0af60100,
+	0xfc04bd00,
+	0xfce0fcf0,
+	0xfcb0fcd0,
+	0xfc90fca0,
+	0x0088fe80,
+	0x32f480fc,
+/* 0x04cf: hub_barrier_done */
+	0x0f01f800,
+	0x040e9801,
+	0xb204febb,
+	0x94188eff,
+	0x008f7e40,
+/* 0x04e3: ctx_redswitch */
+	0x0f00f800,
+	0x85008020,
+	0x000ff601,
+	0x080e04bd,
+/* 0x04f0: ctx_redswitch_delay */
+	0xf401e2b6,
+	0xf5f1fd1b,
+	0xf5f10800,
+	0x00800200,
+	0x0ff60185,
+	0xf804bd00,
+/* 0x0509: ctx_xfer */
+	0x81008000,
+	0x000ff602,
+	0x11f404bd,
+	0x04e37e07,
+/* 0x0519: ctx_xfer_not_load */
+	0x02167e00,
+	0x8024bd00,
+	0xf60247fc,
+	0x04bd0002,
+	0xb6012cf0,
+	0xfc800320,
+	0x02f6024a,
+	0xf004bd00,
+	0xa5f001ac,
+	0x00008b02,
+	0x040c9850,
+	0xbb0fc4b6,
+	0x0c9800bc,
+	0x010d9800,
+	0x3d7e000e,
+	0xacf00001,
+	0x40008b01,
+	0x040c9850,
+	0xbb0fc4b6,
+	0x0c9800bc,
+	0x020d9801,
+	0x4e060f98,
+	0x3d7e0800,
+	0xacf00001,
+	0x04a5f001,
+	0x5030008b,
+	0xb6040c98,
+	0xbcbb0fc4,
+	0x020c9800,
+	0x98030d98,
+	0x004e080f,
+	0x013d7e02,
+	0x020a7e00,
+	0x0601f400,
+/* 0x05a3: ctx_xfer_post */
+	0x7e0712f4,
+/* 0x05a7: ctx_xfer_done */
+	0x7e000227,
+	0xf80004cf,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+};
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvc0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvc0.fuc.h
index f2b0dea8011..325cc7b7b2f 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvc0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvc0.fuc.h
@@ -37,14 +37,14 @@ uint32_t nvc0_grgpc_data[] = {
 };
 
 uint32_t nvc0_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -68,184 +68,214 @@ uint32_t nvc0_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f405,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -259,59 +289,75 @@ uint32_t nvc0_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe042617,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
-	0xf00137f0,
-	0x32bb1f24,
-	0x0132b604,
-	0x80050280,
-	0x10b70603,
-	0x12cf0400,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xf104bd00,
+	0xf0420017,
+	0x11cf0013,
+	0x0911e700,
+	0x0814b601,
+	0xf00014fe,
+	0x07f10227,
+	0x03f01200,
+	0x0002d000,
+	0x17f104bd,
+	0x10fe04e6,
+	0x0007f100,
+	0x0003f007,
+	0xbd0000d0,
+	0x0427f004,
+	0x040007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0xf11031f4,
+	0xf0820027,
+	0x22cf0123,
+	0x0137f000,
+	0xbb1f24f0,
+	0x32b60432,
+	0x05028001,
+	0xf1060380,
+	0xf0860027,
+	0x22cf0123,
 	0x04028000,
 	0x010027f1,
 	0xcf0223f0,
 	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x40b7003e,
-	0x35b61300,
-	0x0043d002,
+	0xf1082595,
+	0xf0c00007,
+	0x05d00103,
+	0xf104bd00,
+	0xf0c10007,
+	0x05d00103,
+	0x9804bd00,
+	0x0f98000e,
+	0x5021f501,
+	0x002fbb01,
+	0x98003fbb,
+	0x0f98010e,
+	0x5021f502,
+	0x050e9801,
+	0xbb00effd,
+	0x3ebb002e,
+	0x0235b600,
+	0xd30007f1,
+	0xd00103f0,
+	0x04bd0003,
 	0xb60825b6,
 	0x20b60635,
 	0x0130b601,
 	0xb60824b6,
 	0x2fb90834,
-	0x7121f502,
+	0xd321f502,
 	0x003fbb02,
 	0x010007f1,
 	0xd00203f0,
@@ -320,7 +366,7 @@ uint32_t nvc0_grgpc_code[] = {
 	0x0007f11f,
 	0x0203f008,
 	0xbd0002d0,
-/* 0x03e9: main */
+/* 0x04a9: main */
 	0x0031f404,
 	0xf00028f4,
 	0x21f41cd7,
@@ -332,74 +378,87 @@ uint32_t nvc0_grgpc_code[] = {
 	0x01e4b604,
 	0xfe051efd,
 	0x21f50018,
-	0x0ef404ad,
-/* 0x0419: main_not_ctx_xfer */
+	0x0ef4059e,
+/* 0x04d9: main_not_ctx_xfer */
 	0x10ef94d3,
 	0xf501f5f0,
-	0xf402fe21,
-/* 0x0426: ih */
+	0xf4037e21,
+/* 0x04e6: ih */
 	0x80f9c60e,
 	0xf90188fe,
 	0xf990f980,
 	0xf9b0f9a0,
 	0xf9e0f9d0,
-	0xcf04bdf0,
-	0xabc4800a,
-	0x1d0bf404,
-	0x1900b7f1,
-	0xcf1cd7f0,
-	0xbfcf40be,
-	0x0421f400,
-	0x0400b0b7,
-	0xd001e7f0,
-/* 0x045e: ih_no_fifo */
-	0x0ad000be,
-	0xfcf0fc40,
-	0xfcd0fce0,
-	0xfca0fcb0,
-	0xfe80fc90,
-	0x80fc0088,
-	0xf80032f4,
-/* 0x0479: hub_barrier_done */
-	0x01f7f001,
-	0xbb040e98,
-	0xe7f104fe,
-	0xe3f09418,
-	0x8d21f440,
-/* 0x048e: ctx_redswitch */
-	0xe7f100f8,
-	0xe4b60614,
-	0x20f7f006,
-	0xf000efd0,
-/* 0x049e: ctx_redswitch_delay */
-	0xf2b608f7,
+	0xf104bdf0,
+	0xf00200a7,
+	0xaacf00a3,
+	0x04abc400,
+	0xf02c0bf4,
+	0xe7f11cd7,
+	0xe3f01a00,
+	0x00eecf00,
+	0x1900f7f1,
+	0xcf00f3f0,
+	0x21f400ff,
+	0x01e7f004,
+	0x1d0007f1,
+	0xd00003f0,
+	0x04bd000e,
+/* 0x0534: ih_no_fifo */
+	0x010007f1,
+	0xd00003f0,
+	0x04bd000a,
+	0xe0fcf0fc,
+	0xb0fcd0fc,
+	0x90fca0fc,
+	0x88fe80fc,
+	0xf480fc00,
+	0x01f80032,
+/* 0x0558: hub_barrier_done */
+	0x9801f7f0,
+	0xfebb040e,
+	0x02ffb904,
+	0x9418e7f1,
+	0xf440e3f0,
+	0x00f89d21,
+/* 0x0570: ctx_redswitch */
+	0xf120f7f0,
+	0xf0850007,
+	0x0fd00103,
+	0xf004bd00,
+/* 0x0582: ctx_redswitch_delay */
+	0xe2b608e7,
 	0xfd1bf401,
-	0x0a20f7f1,
-	0xf800efd0,
-/* 0x04ad: ctx_xfer */
-	0x0417f100,
-	0x0614b60a,
-	0xf4001fd0,
-	0x21f50711,
-/* 0x04be: ctx_xfer_not_load */
-	0x17f1048e,
-	0x13f04afc,
-	0x0c27f002,
-	0xf50012d0,
-	0xf1021521,
-	0xf047fc27,
-	0x20d00223,
-	0x012cf000,
-	0xd00320b6,
-	0xacf00012,
-	0x02a5f001,
-	0xf000b7f0,
+	0x0800f5f1,
+	0x0200f5f1,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000f,
+/* 0x059e: ctx_xfer */
+	0x07f100f8,
+	0x03f08100,
+	0x000fd002,
+	0x11f404bd,
+	0x7021f507,
+/* 0x05b1: ctx_xfer_not_load */
+	0x6a21f505,
+	0xf124bd02,
+	0xf047fc07,
+	0x02d00203,
+	0xf004bd00,
+	0x20b6012c,
+	0xfc07f103,
+	0x0203f04a,
+	0xbd0002d0,
+	0x01acf004,
+	0xf102a5f0,
+	0xf00000b7,
 	0x0c9850b3,
 	0x0fc4b604,
 	0x9800bcbb,
 	0x0d98000c,
 	0x00e7f001,
-	0x016621f5,
+	0x016f21f5,
 	0xf001acf0,
 	0xb7f104a5,
 	0xb3f04000,
@@ -409,17 +468,20 @@ uint32_t nvc0_grgpc_code[] = {
 	0x020d9801,
 	0xf1060f98,
 	0xf50800e7,
-	0xf5016621,
-	0xf4021521,
+	0xf5016f21,
+	0xf4025e21,
 	0x12f40601,
-/* 0x0535: ctx_xfer_post */
-	0xfc17f114,
-	0x0213f04a,
-	0xd00d27f0,
-	0x21f50012,
-/* 0x0546: ctx_xfer_done */
-	0x21f50215,
-	0x00f80479,
+/* 0x0629: ctx_xfer_post */
+	0x7f21f507,
+/* 0x062d: ctx_xfer_done */
+	0x5821f502,
+	0x0000f805,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvd7.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvd7.fuc.h
index dd346c2a162..d1504a4059c 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvd7.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvd7.fuc.h
@@ -41,14 +41,14 @@ uint32_t nvd7_grgpc_data[] = {
 };
 
 uint32_t nvd7_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -72,184 +72,214 @@ uint32_t nvd7_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f405,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -263,81 +293,97 @@ uint32_t nvd7_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe047017,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
-	0xf00137f0,
-	0x32bb1f24,
-	0x0132b604,
-	0x80050280,
-	0x10b70603,
-	0x12cf0400,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xf104bd00,
+	0xf0420017,
+	0x11cf0013,
+	0x0911e700,
+	0x0814b601,
+	0xf00014fe,
+	0x07f10227,
+	0x03f01200,
+	0x0002d000,
+	0x17f104bd,
+	0x10fe0530,
+	0x0007f100,
+	0x0003f007,
+	0xbd0000d0,
+	0x0427f004,
+	0x040007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0xf11031f4,
+	0xf0820027,
+	0x22cf0123,
+	0x0137f000,
+	0xbb1f24f0,
+	0x32b60432,
+	0x05028001,
+	0xf1060380,
+	0xf0860027,
+	0x22cf0123,
 	0x04028000,
 	0x0c30e7f1,
 	0xbd50e3f0,
 	0xbd34bd24,
-/* 0x0371: init_unk_loop */
+/* 0x0421: init_unk_loop */
 	0x6821f444,
 	0xf400f6b0,
 	0xf7f00f0b,
 	0x04f2bb01,
 	0xb6054ffd,
-/* 0x0386: init_unk_next */
+/* 0x0436: init_unk_next */
 	0x20b60130,
 	0x04e0b601,
 	0xf40126b0,
-/* 0x0392: init_unk_done */
+/* 0x0442: init_unk_done */
 	0x0380e21b,
 	0x08048007,
 	0x010027f1,
 	0xcf0223f0,
 	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x0e98003e,
-	0x030f9802,
-	0x014721f5,
-	0xfd070e98,
-	0x2ebb00ef,
-	0x003ebb00,
-	0x130040b7,
-	0xd00235b6,
-	0x25b60043,
+	0xf1082595,
+	0xf0c00007,
+	0x05d00103,
+	0xf104bd00,
+	0xf0c10007,
+	0x05d00103,
+	0x9804bd00,
+	0x0f98000e,
+	0x5021f501,
+	0x002fbb01,
+	0x98003fbb,
+	0x0f98010e,
+	0x5021f502,
+	0x050e9801,
+	0xbb00effd,
+	0x3ebb002e,
+	0x020e9800,
+	0xf5030f98,
+	0x98015021,
+	0xeffd070e,
+	0x002ebb00,
+	0xb6003ebb,
+	0x07f10235,
+	0x03f0d300,
+	0x0003d001,
+	0x25b604bd,
 	0x0635b608,
 	0xb60120b6,
 	0x24b60130,
 	0x0834b608,
 	0xf5022fb9,
-	0xbb027121,
+	0xbb02d321,
 	0x07f1003f,
 	0x03f00100,
 	0x0003d002,
@@ -345,7 +391,7 @@ uint32_t nvd7_grgpc_code[] = {
 	0xf11f29f0,
 	0xf0080007,
 	0x02d00203,
-/* 0x0433: main */
+/* 0x04f3: main */
 	0xf404bd00,
 	0x28f40031,
 	0x24d7f000,
@@ -357,75 +403,88 @@ uint32_t nvd7_grgpc_code[] = {
 	0xb60412fd,
 	0x1efd01e4,
 	0x0018fe05,
-	0x04f721f5,
-/* 0x0463: main_not_ctx_xfer */
+	0x05e821f5,
+/* 0x0523: main_not_ctx_xfer */
 	0x94d30ef4,
 	0xf5f010ef,
-	0xfe21f501,
-	0xc60ef402,
-/* 0x0470: ih */
+	0x7e21f501,
+	0xc60ef403,
+/* 0x0530: ih */
 	0x88fe80f9,
 	0xf980f901,
 	0xf9a0f990,
 	0xf9d0f9b0,
 	0xbdf0f9e0,
-	0x800acf04,
-	0xf404abc4,
-	0xb7f11d0b,
-	0xd7f01900,
-	0x40becf24,
-	0xf400bfcf,
-	0xb0b70421,
-	0xe7f00400,
-	0x00bed001,
-/* 0x04a8: ih_no_fifo */
-	0xfc400ad0,
-	0xfce0fcf0,
-	0xfcb0fcd0,
-	0xfc90fca0,
-	0x0088fe80,
-	0x32f480fc,
-/* 0x04c3: hub_barrier_done */
-	0xf001f800,
-	0x0e9801f7,
-	0x04febb04,
-	0x9418e7f1,
-	0xf440e3f0,
-	0x00f88d21,
-/* 0x04d8: ctx_redswitch */
-	0x0614e7f1,
-	0xf006e4b6,
-	0xefd020f7,
-	0x08f7f000,
-/* 0x04e8: ctx_redswitch_delay */
-	0xf401f2b6,
-	0xf7f1fd1b,
-	0xefd00a20,
-/* 0x04f7: ctx_xfer */
-	0xf100f800,
-	0xb60a0417,
-	0x1fd00614,
-	0x0711f400,
-	0x04d821f5,
-/* 0x0508: ctx_xfer_not_load */
-	0x4afc17f1,
-	0xf00213f0,
-	0x12d00c27,
-	0x1521f500,
-	0xfc27f102,
-	0x0223f047,
-	0xf00020d0,
-	0x20b6012c,
-	0x0012d003,
-	0xf001acf0,
-	0xb7f002a5,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x24d7f02c,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
+	0xf00421f4,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x057e: ih_no_fifo */
+	0x07f104bd,
+	0x03f00100,
+	0x000ad000,
+	0xf0fc04bd,
+	0xd0fce0fc,
+	0xa0fcb0fc,
+	0x80fc90fc,
+	0xfc0088fe,
+	0x0032f480,
+/* 0x05a2: hub_barrier_done */
+	0xf7f001f8,
+	0x040e9801,
+	0xb904febb,
+	0xe7f102ff,
+	0xe3f09418,
+	0x9d21f440,
+/* 0x05ba: ctx_redswitch */
+	0xf7f000f8,
+	0x0007f120,
+	0x0103f085,
+	0xbd000fd0,
+	0x08e7f004,
+/* 0x05cc: ctx_redswitch_delay */
+	0xf401e2b6,
+	0xf5f1fd1b,
+	0xf5f10800,
+	0x07f10200,
+	0x03f08500,
+	0x000fd001,
+	0x00f804bd,
+/* 0x05e8: ctx_xfer */
+	0x810007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0xf50711f4,
+/* 0x05fb: ctx_xfer_not_load */
+	0xf505ba21,
+	0xbd026a21,
+	0xfc07f124,
+	0x0203f047,
+	0xbd0002d0,
+	0x012cf004,
+	0xf10320b6,
+	0xf04afc07,
+	0x02d00203,
+	0xf004bd00,
+	0xa5f001ac,
+	0x00b7f102,
 	0x50b3f000,
 	0xb6040c98,
 	0xbcbb0fc4,
 	0x000c9800,
 	0xf0010d98,
 	0x21f500e7,
-	0xacf00166,
+	0xacf0016f,
 	0x00b7f101,
 	0x50b3f040,
 	0xb6040c98,
@@ -434,7 +493,7 @@ uint32_t nvd7_grgpc_code[] = {
 	0x98020d98,
 	0xe7f1060f,
 	0x21f50800,
-	0xacf00166,
+	0xacf0016f,
 	0x04a5f001,
 	0x3000b7f1,
 	0x9850b3f0,
@@ -443,18 +502,21 @@ uint32_t nvd7_grgpc_code[] = {
 	0x98020c98,
 	0x0f98030d,
 	0x00e7f108,
-	0x6621f502,
-	0x1521f501,
+	0x6f21f502,
+	0x5e21f501,
 	0x0601f402,
-/* 0x05a3: ctx_xfer_post */
-	0xf11412f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00d,
-	0x021521f5,
-/* 0x05b4: ctx_xfer_done */
-	0x04c321f5,
-	0x000000f8,
+/* 0x0697: ctx_xfer_post */
+	0xf50712f4,
+/* 0x069b: ctx_xfer_done */
+	0xf5027f21,
+	0xf805a221,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnve0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnve0.fuc.h
index 7ff5ef6b080..855b220378f 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnve0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnve0.fuc.h
@@ -41,14 +41,14 @@ uint32_t nve0_grgpc_data[] = {
 };
 
 uint32_t nve0_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -72,184 +72,214 @@ uint32_t nve0_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f405,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -263,81 +293,97 @@ uint32_t nve0_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe047017,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
-	0xf00137f0,
-	0x32bb1f24,
-	0x0132b604,
-	0x80050280,
-	0x10b70603,
-	0x12cf0400,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xf104bd00,
+	0xf0420017,
+	0x11cf0013,
+	0x0911e700,
+	0x0814b601,
+	0xf00014fe,
+	0x07f10227,
+	0x03f01200,
+	0x0002d000,
+	0x17f104bd,
+	0x10fe0530,
+	0x0007f100,
+	0x0003f007,
+	0xbd0000d0,
+	0x0427f004,
+	0x040007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0xf11031f4,
+	0xf0820027,
+	0x22cf0123,
+	0x0137f000,
+	0xbb1f24f0,
+	0x32b60432,
+	0x05028001,
+	0xf1060380,
+	0xf0860027,
+	0x22cf0123,
 	0x04028000,
 	0x0c30e7f1,
 	0xbd50e3f0,
 	0xbd34bd24,
-/* 0x0371: init_unk_loop */
+/* 0x0421: init_unk_loop */
 	0x6821f444,
 	0xf400f6b0,
 	0xf7f00f0b,
 	0x04f2bb01,
 	0xb6054ffd,
-/* 0x0386: init_unk_next */
+/* 0x0436: init_unk_next */
 	0x20b60130,
 	0x04e0b601,
 	0xf40126b0,
-/* 0x0392: init_unk_done */
+/* 0x0442: init_unk_done */
 	0x0380e21b,
 	0x08048007,
 	0x010027f1,
 	0xcf0223f0,
 	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x0e98003e,
-	0x030f9802,
-	0x014721f5,
-	0xfd070e98,
-	0x2ebb00ef,
-	0x003ebb00,
-	0x130040b7,
-	0xd00235b6,
-	0x25b60043,
+	0xf1082595,
+	0xf0c00007,
+	0x05d00103,
+	0xf104bd00,
+	0xf0c10007,
+	0x05d00103,
+	0x9804bd00,
+	0x0f98000e,
+	0x5021f501,
+	0x002fbb01,
+	0x98003fbb,
+	0x0f98010e,
+	0x5021f502,
+	0x050e9801,
+	0xbb00effd,
+	0x3ebb002e,
+	0x020e9800,
+	0xf5030f98,
+	0x98015021,
+	0xeffd070e,
+	0x002ebb00,
+	0xb6003ebb,
+	0x07f10235,
+	0x03f0d300,
+	0x0003d001,
+	0x25b604bd,
 	0x0635b608,
 	0xb60120b6,
 	0x24b60130,
 	0x0834b608,
 	0xf5022fb9,
-	0xbb027121,
+	0xbb02d321,
 	0x07f1003f,
 	0x03f00100,
 	0x0003d002,
@@ -345,7 +391,7 @@ uint32_t nve0_grgpc_code[] = {
 	0xf11f29f0,
 	0xf0080007,
 	0x02d00203,
-/* 0x0433: main */
+/* 0x04f3: main */
 	0xf404bd00,
 	0x28f40031,
 	0x24d7f000,
@@ -357,75 +403,88 @@ uint32_t nve0_grgpc_code[] = {
 	0xb60412fd,
 	0x1efd01e4,
 	0x0018fe05,
-	0x04f721f5,
-/* 0x0463: main_not_ctx_xfer */
+	0x05e821f5,
+/* 0x0523: main_not_ctx_xfer */
 	0x94d30ef4,
 	0xf5f010ef,
-	0xfe21f501,
-	0xc60ef402,
-/* 0x0470: ih */
+	0x7e21f501,
+	0xc60ef403,
+/* 0x0530: ih */
 	0x88fe80f9,
 	0xf980f901,
 	0xf9a0f990,
 	0xf9d0f9b0,
 	0xbdf0f9e0,
-	0x800acf04,
-	0xf404abc4,
-	0xb7f11d0b,
-	0xd7f01900,
-	0x40becf24,
-	0xf400bfcf,
-	0xb0b70421,
-	0xe7f00400,
-	0x00bed001,
-/* 0x04a8: ih_no_fifo */
-	0xfc400ad0,
-	0xfce0fcf0,
-	0xfcb0fcd0,
-	0xfc90fca0,
-	0x0088fe80,
-	0x32f480fc,
-/* 0x04c3: hub_barrier_done */
-	0xf001f800,
-	0x0e9801f7,
-	0x04febb04,
-	0x9418e7f1,
-	0xf440e3f0,
-	0x00f88d21,
-/* 0x04d8: ctx_redswitch */
-	0x0614e7f1,
-	0xf006e4b6,
-	0xefd020f7,
-	0x08f7f000,
-/* 0x04e8: ctx_redswitch_delay */
-	0xf401f2b6,
-	0xf7f1fd1b,
-	0xefd00a20,
-/* 0x04f7: ctx_xfer */
-	0xf100f800,
-	0xb60a0417,
-	0x1fd00614,
-	0x0711f400,
-	0x04d821f5,
-/* 0x0508: ctx_xfer_not_load */
-	0x4afc17f1,
-	0xf00213f0,
-	0x12d00c27,
-	0x1521f500,
-	0xfc27f102,
-	0x0223f047,
-	0xf00020d0,
-	0x20b6012c,
-	0x0012d003,
-	0xf001acf0,
-	0xb7f002a5,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x24d7f02c,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
+	0xf00421f4,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x057e: ih_no_fifo */
+	0x07f104bd,
+	0x03f00100,
+	0x000ad000,
+	0xf0fc04bd,
+	0xd0fce0fc,
+	0xa0fcb0fc,
+	0x80fc90fc,
+	0xfc0088fe,
+	0x0032f480,
+/* 0x05a2: hub_barrier_done */
+	0xf7f001f8,
+	0x040e9801,
+	0xb904febb,
+	0xe7f102ff,
+	0xe3f09418,
+	0x9d21f440,
+/* 0x05ba: ctx_redswitch */
+	0xf7f000f8,
+	0x0007f120,
+	0x0103f085,
+	0xbd000fd0,
+	0x08e7f004,
+/* 0x05cc: ctx_redswitch_delay */
+	0xf401e2b6,
+	0xf5f1fd1b,
+	0xf5f10800,
+	0x07f10200,
+	0x03f08500,
+	0x000fd001,
+	0x00f804bd,
+/* 0x05e8: ctx_xfer */
+	0x810007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0xf50711f4,
+/* 0x05fb: ctx_xfer_not_load */
+	0xf505ba21,
+	0xbd026a21,
+	0xfc07f124,
+	0x0203f047,
+	0xbd0002d0,
+	0x012cf004,
+	0xf10320b6,
+	0xf04afc07,
+	0x02d00203,
+	0xf004bd00,
+	0xa5f001ac,
+	0x00b7f102,
 	0x50b3f000,
 	0xb6040c98,
 	0xbcbb0fc4,
 	0x000c9800,
 	0xf0010d98,
 	0x21f500e7,
-	0xacf00166,
+	0xacf0016f,
 	0x00b7f101,
 	0x50b3f040,
 	0xb6040c98,
@@ -434,7 +493,7 @@ uint32_t nve0_grgpc_code[] = {
 	0x98020d98,
 	0xe7f1060f,
 	0x21f50800,
-	0xacf00166,
+	0xacf0016f,
 	0x04a5f001,
 	0x3000b7f1,
 	0x9850b3f0,
@@ -443,18 +502,21 @@ uint32_t nve0_grgpc_code[] = {
 	0x98020c98,
 	0x0f98030d,
 	0x00e7f108,
-	0x6621f502,
-	0x1521f501,
+	0x6f21f502,
+	0x5e21f501,
 	0x0601f402,
-/* 0x05a3: ctx_xfer_post */
-	0xf11412f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00d,
-	0x021521f5,
-/* 0x05b4: ctx_xfer_done */
-	0x04c321f5,
-	0x000000f8,
+/* 0x0697: ctx_xfer_post */
+	0xf50712f4,
+/* 0x069b: ctx_xfer_done */
+	0xf5027f21,
+	0xf805a221,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvf0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvf0.fuc.h
index f870507be88..1b803197d28 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvf0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvf0.fuc.h
@@ -41,14 +41,14 @@ uint32_t nvf0_grgpc_data[] = {
 };
 
 uint32_t nvf0_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -72,184 +72,214 @@ uint32_t nvf0_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f03700,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f037,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f03700,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f037,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf0370007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f405,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f03700,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -263,81 +293,97 @@ uint32_t nvf0_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe047017,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
-	0xf00137f0,
-	0x32bb1f24,
-	0x0132b604,
-	0x80050280,
-	0x10b70603,
-	0x12cf0400,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xf104bd00,
+	0xf0420017,
+	0x11cf0013,
+	0x0911e700,
+	0x0814b601,
+	0xf00014fe,
+	0x07f10227,
+	0x03f01200,
+	0x0002d000,
+	0x17f104bd,
+	0x10fe0530,
+	0x0007f100,
+	0x0003f007,
+	0xbd0000d0,
+	0x0427f004,
+	0x040007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0xf11031f4,
+	0xf0820027,
+	0x22cf0123,
+	0x0137f000,
+	0xbb1f24f0,
+	0x32b60432,
+	0x05028001,
+	0xf1060380,
+	0xf0860027,
+	0x22cf0123,
 	0x04028000,
 	0x0c30e7f1,
 	0xbd50e3f0,
 	0xbd34bd24,
-/* 0x0371: init_unk_loop */
+/* 0x0421: init_unk_loop */
 	0x6821f444,
 	0xf400f6b0,
 	0xf7f00f0b,
 	0x04f2bb01,
 	0xb6054ffd,
-/* 0x0386: init_unk_next */
+/* 0x0436: init_unk_next */
 	0x20b60130,
 	0x04e0b601,
 	0xf40226b0,
-/* 0x0392: init_unk_done */
+/* 0x0442: init_unk_done */
 	0x0380e21b,
 	0x08048007,
 	0x010027f1,
 	0xcf0223f0,
 	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x0e98003e,
-	0x030f9802,
-	0x014721f5,
-	0xfd070e98,
-	0x2ebb00ef,
-	0x003ebb00,
-	0x130040b7,
-	0xd00235b6,
-	0x25b60043,
+	0xf1082595,
+	0xf0c00007,
+	0x05d00103,
+	0xf104bd00,
+	0xf0c10007,
+	0x05d00103,
+	0x9804bd00,
+	0x0f98000e,
+	0x5021f501,
+	0x002fbb01,
+	0x98003fbb,
+	0x0f98010e,
+	0x5021f502,
+	0x050e9801,
+	0xbb00effd,
+	0x3ebb002e,
+	0x020e9800,
+	0xf5030f98,
+	0x98015021,
+	0xeffd070e,
+	0x002ebb00,
+	0xb6003ebb,
+	0x07f10235,
+	0x03f0d300,
+	0x0003d001,
+	0x25b604bd,
 	0x0635b608,
 	0xb60120b6,
 	0x24b60130,
 	0x0834b608,
 	0xf5022fb9,
-	0xbb027121,
+	0xbb02d321,
 	0x07f1003f,
 	0x03f00100,
 	0x0003d002,
@@ -345,7 +391,7 @@ uint32_t nvf0_grgpc_code[] = {
 	0xf11f29f0,
 	0xf0300007,
 	0x02d00203,
-/* 0x0433: main */
+/* 0x04f3: main */
 	0xf404bd00,
 	0x28f40031,
 	0x24d7f000,
@@ -357,75 +403,88 @@ uint32_t nvf0_grgpc_code[] = {
 	0xb60412fd,
 	0x1efd01e4,
 	0x0018fe05,
-	0x04f721f5,
-/* 0x0463: main_not_ctx_xfer */
+	0x05e821f5,
+/* 0x0523: main_not_ctx_xfer */
 	0x94d30ef4,
 	0xf5f010ef,
-	0xfe21f501,
-	0xc60ef402,
-/* 0x0470: ih */
+	0x7e21f501,
+	0xc60ef403,
+/* 0x0530: ih */
 	0x88fe80f9,
 	0xf980f901,
 	0xf9a0f990,
 	0xf9d0f9b0,
 	0xbdf0f9e0,
-	0x800acf04,
-	0xf404abc4,
-	0xb7f11d0b,
-	0xd7f01900,
-	0x40becf24,
-	0xf400bfcf,
-	0xb0b70421,
-	0xe7f00400,
-	0x00bed001,
-/* 0x04a8: ih_no_fifo */
-	0xfc400ad0,
-	0xfce0fcf0,
-	0xfcb0fcd0,
-	0xfc90fca0,
-	0x0088fe80,
-	0x32f480fc,
-/* 0x04c3: hub_barrier_done */
-	0xf001f800,
-	0x0e9801f7,
-	0x04febb04,
-	0x9418e7f1,
-	0xf440e3f0,
-	0x00f88d21,
-/* 0x04d8: ctx_redswitch */
-	0x0614e7f1,
-	0xf006e4b6,
-	0xefd020f7,
-	0x08f7f000,
-/* 0x04e8: ctx_redswitch_delay */
-	0xf401f2b6,
-	0xf7f1fd1b,
-	0xefd00a20,
-/* 0x04f7: ctx_xfer */
-	0xf100f800,
-	0xb60a0417,
-	0x1fd00614,
-	0x0711f400,
-	0x04d821f5,
-/* 0x0508: ctx_xfer_not_load */
-	0x4afc17f1,
-	0xf00213f0,
-	0x12d00c27,
-	0x1521f500,
-	0xfc27f102,
-	0x0223f047,
-	0xf00020d0,
-	0x20b6012c,
-	0x0012d003,
-	0xf001acf0,
-	0xb7f002a5,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x24d7f02c,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
+	0xf00421f4,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x057e: ih_no_fifo */
+	0x07f104bd,
+	0x03f00100,
+	0x000ad000,
+	0xf0fc04bd,
+	0xd0fce0fc,
+	0xa0fcb0fc,
+	0x80fc90fc,
+	0xfc0088fe,
+	0x0032f480,
+/* 0x05a2: hub_barrier_done */
+	0xf7f001f8,
+	0x040e9801,
+	0xb904febb,
+	0xe7f102ff,
+	0xe3f09418,
+	0x9d21f440,
+/* 0x05ba: ctx_redswitch */
+	0xf7f000f8,
+	0x0007f120,
+	0x0103f085,
+	0xbd000fd0,
+	0x08e7f004,
+/* 0x05cc: ctx_redswitch_delay */
+	0xf401e2b6,
+	0xf5f1fd1b,
+	0xf5f10800,
+	0x07f10200,
+	0x03f08500,
+	0x000fd001,
+	0x00f804bd,
+/* 0x05e8: ctx_xfer */
+	0x810007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0xf50711f4,
+/* 0x05fb: ctx_xfer_not_load */
+	0xf505ba21,
+	0xbd026a21,
+	0xfc07f124,
+	0x0203f047,
+	0xbd0002d0,
+	0x012cf004,
+	0xf10320b6,
+	0xf04afc07,
+	0x02d00203,
+	0xf004bd00,
+	0xa5f001ac,
+	0x00b7f102,
 	0x50b3f000,
 	0xb6040c98,
 	0xbcbb0fc4,
 	0x000c9800,
 	0xf0010d98,
 	0x21f500e7,
-	0xacf00166,
+	0xacf0016f,
 	0x00b7f101,
 	0x50b3f040,
 	0xb6040c98,
@@ -434,7 +493,7 @@ uint32_t nvf0_grgpc_code[] = {
 	0x98020d98,
 	0xe7f1060f,
 	0x21f50800,
-	0xacf00166,
+	0xacf0016f,
 	0x04a5f001,
 	0x3000b7f1,
 	0x9850b3f0,
@@ -443,18 +502,21 @@ uint32_t nvf0_grgpc_code[] = {
 	0x98020c98,
 	0x0f98030d,
 	0x00e7f108,
-	0x6621f502,
-	0x1521f501,
+	0x6f21f502,
+	0x5e21f501,
 	0x0601f402,
-/* 0x05a3: ctx_xfer_post */
-	0xf11412f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00d,
-	0x021521f5,
-/* 0x05b4: ctx_xfer_done */
-	0x04c321f5,
-	0x000000f8,
+/* 0x0697: ctx_xfer_post */
+	0xf50712f4,
+/* 0x069b: ctx_xfer_done */
+	0xf5027f21,
+	0xf805a221,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hub.fuc b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hub.fuc
index b82d2ae8991..b4ad18bf5a2 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hub.fuc
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hub.fuc
@@ -49,7 +49,7 @@ hub_mmio_list_next:
 #ifdef INCLUDE_CODE
 // reports an exception to the host
 //
-// In: $r15 error code (see nvc0.fuc)
+// In: $r15 error code (see os.h)
 //
 error:
 	nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_VAL(5), 0, $r15)
@@ -68,60 +68,57 @@ error:
 //
 init:
 	clear b32 $r0
-	mov $sp $r0
 	mov $xdbase $r0
 
+	// setup stack
+	nv_iord($r1, NV_PGRAPH_FECS_CAPS, 0)
+	extr $r1 $r1 9:17
+	shl b32 $r1 8
+	mov $sp $r1
+
 	// enable fifo access
-	mov $r1 0x1200
-	mov $r2 2
-	iowr I[$r1 + 0x000] $r2	// FIFO_ENABLE
+	mov $r2 NV_PGRAPH_FECS_ACCESS_FIFO
+	nv_iowr(NV_PGRAPH_FECS_ACCESS, 0, $r2)
 
 	// setup i0 handler, and route all interrupts to it
 	mov $r1 #ih
 	mov $iv0 $r1
-	mov $r1 0x400
-	iowr I[$r1 + 0x300] $r0	// INTR_DISPATCH
 
-	// route HUB_CHANNEL_SWITCH to fuc interrupt 8
-	mov $r3 0x404
-	shl b32 $r3 6
-	mov $r2 0x2003		// { HUB_CHANNEL_SWITCH, ZERO } -> intr 8
-	iowr I[$r3 + 0x000] $r2
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_FECS_INTR_ROUTE, 0, $r2)
+
+	// route HUB_CHSW_PULSE to fuc interrupt 8
+	mov $r2 0x2003		// { HUB_CHSW_PULSE, ZERO } -> intr 8
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 0, $r2)
 
 	// not sure what these are, route them because NVIDIA does, and
 	// the IRQ handler will signal the host if we ever get one.. we
 	// may find out if/why we need to handle these if so..
 	//
-	mov $r2 0x2004
-	iowr I[$r3 + 0x004] $r2 // { 0x04, ZERO } -> intr 9
-	mov $r2 0x200b
-	iowr I[$r3 + 0x008] $r2 // { 0x0b, ZERO } -> intr 10
-	mov $r2 0x200c
-	iowr I[$r3 + 0x01c] $r2 // { 0x0c, ZERO } -> intr 15
+	mov $r2 0x2004		// { 0x04, ZERO } -> intr 9
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 1, $r2)
+	mov $r2 0x200b		// { HUB_FIRMWARE_MTHD, ZERO } -> intr 10
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 2, $r2)
+	mov $r2 0x200c		// { 0x0c, ZERO } -> intr 15
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 7, $r2)
 
 	// enable all INTR_UP interrupts
-	mov $r2 0xc24
-	shl b32 $r2 6
-	not b32 $r3 $r0
-	iowr I[$r2] $r3
+	sub b32 $r3 $r0 1
+	nv_iowr(NV_PGRAPH_FECS_INTR_UP_EN, 0, $r3)
 
-	// enable fifo, ctxsw, 9, 10, 15 interrupts
-	mov $r2 -0x78fc		// 0x8704
-	sethi $r2 0
-	iowr I[$r1 + 0x000] $r2	// INTR_EN_SET
+	// enable fifo, ctxsw, 9, fwmthd, 15 interrupts
+	imm32($r2, 0x8704)
+	nv_iowr(NV_PGRAPH_FECS_INTR_EN_SET, 0, $r2)
 
 	// fifo level triggered, rest edge
-	sub b32 $r1 0x100
-	mov $r2 4
-	iowr I[$r1] $r2
+	mov $r2 NV_PGRAPH_FECS_INTR_MODE_FIFO_LEVEL
+	nv_iowr(NV_PGRAPH_FECS_INTR_MODE, 0, $r2)
 
 	// enable interrupts
 	bset $flags ie0
 
 	// fetch enabled GPC/ROP counts
-	mov $r14 -0x69fc	// 0x409604
-	sethi $r14 0x400000
-	call #nv_rd32
+	nv_rd32($r14, 0x409604)
 	extr $r1 $r15 16:20
 	st b32 D[$r0 + #rop_count] $r1
 	and $r15 0x1f
@@ -131,37 +128,40 @@ init:
 	mov $r1 1
 	shl b32 $r1 $r15
 	sub b32 $r1 1
-	mov $r2 0x40c
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r1
-	iowr I[$r2 + 0x100] $r1
+	nv_iowr(NV_PGRAPH_FECS_BAR_MASK0, 0, $r1)
+	nv_iowr(NV_PGRAPH_FECS_BAR_MASK1, 0, $r1)
 
 	// context size calculation, reserve first 256 bytes for use by fuc
 	mov $r1 256
 
+	//
+	mov $r15 2
+	call(ctx_4170s)
+	call(ctx_4170w)
+	mov $r15 0x10
+	call(ctx_86c)
+
 	// calculate size of mmio context data
 	ld b32 $r14 D[$r0 + #hub_mmio_list_head]
 	ld b32 $r15 D[$r0 + #hub_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 
 	// set mmctx base addresses now so we don't have to do it later,
 	// they don't (currently) ever change
-	mov $r3 0x700
-	shl b32 $r3 6
 	shr b32 $r4 $r1 8
-	iowr I[$r3 + 0x000] $r4		// MMCTX_SAVE_SWBASE
-	iowr I[$r3 + 0x100] $r4		// MMCTX_LOAD_SWBASE
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_SAVE_SWBASE, 0, $r4)
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_LOAD_SWBASE, 0, $r4)
 	add b32 $r3 0x1300
 	add b32 $r1 $r15
 	shr b32 $r15 2
-	iowr I[$r3 + 0x000] $r15	// MMCTX_LOAD_COUNT, wtf for?!?
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_LOAD_COUNT, 0, $r15) // wtf??
 
 	// strands, base offset needs to be aligned to 256 bytes
 	shr b32 $r1 8
 	add b32 $r1 1
 	shl b32 $r1 8
 	mov b32 $r15 $r1
-	call #strand_ctx_init
+	call(strand_ctx_init)
 	add b32 $r1 $r15
 
 	// initialise each GPC in sequence by passing in the offset of its
@@ -173,30 +173,29 @@ init:
 	// in GPCn_CC_SCRATCH[1]
 	//
 	ld b32 $r3 D[$r0 + #gpc_count]
-	mov $r4 0x2000
-	sethi $r4 0x500000
+	imm32($r4, 0x502000)
 	init_gpc:
 		// setup, and start GPC ucode running
 		add b32 $r14 $r4 0x804
 		mov b32 $r15 $r1
-		call #nv_wr32			// CC_SCRATCH[1] = ctx offset
+		call(nv_wr32)			// CC_SCRATCH[1] = ctx offset
 		add b32 $r14 $r4 0x10c
 		clear b32 $r15
-		call #nv_wr32
+		call(nv_wr32)
 		add b32 $r14 $r4 0x104
-		call #nv_wr32			// ENTRY
+		call(nv_wr32)			// ENTRY
 		add b32 $r14 $r4 0x100
 		mov $r15 2			// CTRL_START_TRIGGER
-		call #nv_wr32			// CTRL
+		call(nv_wr32)			// CTRL
 
 		// wait for it to complete, and adjust context size
 		add b32 $r14 $r4 0x800
 		init_gpc_wait:
-			call #nv_rd32
+			call(nv_rd32)
 			xbit $r15 $r15 31
 			bra e #init_gpc_wait
 		add b32 $r14 $r4 0x804
-		call #nv_rd32
+		call(nv_rd32)
 		add b32 $r1 $r15
 
 		// next!
@@ -204,6 +203,12 @@ init:
 		sub b32 $r3 1
 		bra ne #init_gpc
 
+	//
+	mov $r15 0
+	call(ctx_86c)
+	mov $r15 0
+	call(ctx_4170s)
+
 	// save context size, and tell host we're ready
 	nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_VAL(1), 0, $r1)
 	clear b32 $r1
@@ -218,17 +223,15 @@ main:
 	bset $flags $p0
 	sleep $p0
 	mov $r13 #cmd_queue
-	call #queue_get
+	call(queue_get)
 	bra $p1 #main
 
 	// context switch, requested by GPU?
 	cmpu b32 $r14 0x4001
 	bra ne #main_not_ctx_switch
 		trace_set(T_AUTO)
-		mov $r1 0xb00
-		shl b32 $r1 6
-		iord $r2 I[$r1 + 0x100]		// CHAN_NEXT
-		iord $r1 I[$r1 + 0x000]		// CHAN_CUR
+		nv_iord($r1, NV_PGRAPH_FECS_CHAN_ADDR, 0)
+		nv_iord($r2, NV_PGRAPH_FECS_CHAN_NEXT, 0)
 
 		xbit $r3 $r1 31
 		bra e #chsw_no_prev
@@ -239,12 +242,12 @@ main:
 				trace_set(T_SAVE)
 				bclr $flags $p1
 				bset $flags $p2
-				call #ctx_xfer
+				call(ctx_xfer)
 				trace_clr(T_SAVE);
 				pop $r2
 				trace_set(T_LOAD);
 				bset $flags $p1
-				call #ctx_xfer
+				call(ctx_xfer)
 				trace_clr(T_LOAD);
 				bra #chsw_done
 			chsw_prev_no_next:
@@ -252,25 +255,21 @@ main:
 				mov b32 $r2 $r1
 				bclr $flags $p1
 				bclr $flags $p2
-				call #ctx_xfer
+				call(ctx_xfer)
 				pop $r2
-				mov $r1 0xb00
-				shl b32 $r1 6
-				iowr I[$r1] $r2
+				nv_iowr(NV_PGRAPH_FECS_CHAN_ADDR, 0, $r2)
 				bra #chsw_done
 		chsw_no_prev:
 			xbit $r3 $r2 31
 			bra e #chsw_done
 				bset $flags $p1
 				bclr $flags $p2
-				call #ctx_xfer
+				call(ctx_xfer)
 
 		// ack the context switch request
 		chsw_done:
-		mov $r1 0xb0c
-		shl b32 $r1 6
-		mov $r2 1
-		iowr I[$r1 + 0x000] $r2		// 0x409b0c
+		mov $r2 NV_PGRAPH_FECS_CHSW_ACK
+		nv_iowr(NV_PGRAPH_FECS_CHSW, 0, $r2)
 		trace_clr(T_AUTO)
 		bra #main
 
@@ -279,7 +278,7 @@ main:
 	cmpu b32 $r14 0x0001
 	bra ne #main_not_ctx_chan
 		mov b32 $r2 $r15
-		call #ctx_chan
+		call(ctx_chan)
 		bra #main_done
 
 	// request to store current channel context?
@@ -289,14 +288,14 @@ main:
 		trace_set(T_SAVE)
 		bclr $flags $p1
 		bclr $flags $p2
-		call #ctx_xfer
+		call(ctx_xfer)
 		trace_clr(T_SAVE)
 		bra #main_done
 
 	main_not_ctx_save:
 		shl b32 $r15 $r14 16
 		or $r15 E_BAD_COMMAND
-		call #error
+		call(error)
 		bra #main
 
 	main_done:
@@ -319,41 +318,58 @@ ih:
 	clear b32 $r0
 
 	// incoming fifo command?
-	iord $r10 I[$r0 + 0x200]	// INTR
-	and $r11 $r10 0x00000004
+	nv_iord($r10, NV_PGRAPH_FECS_INTR, 0)
+	and $r11 $r10 NV_PGRAPH_FECS_INTR_FIFO
 	bra e #ih_no_fifo
 		// queue incoming fifo command for later processing
-		mov $r11 0x1900
 		mov $r13 #cmd_queue
-		iord $r14 I[$r11 + 0x100]	// FIFO_CMD
-		iord $r15 I[$r11 + 0x000]	// FIFO_DATA
-		call #queue_put
+		nv_iord($r14, NV_PGRAPH_FECS_FIFO_CMD, 0)
+		nv_iord($r15, NV_PGRAPH_FECS_FIFO_DATA, 0)
+		call(queue_put)
 		add b32 $r11 0x400
 		mov $r14 1
-		iowr I[$r11 + 0x000] $r14	// FIFO_ACK
+		nv_iowr(NV_PGRAPH_FECS_FIFO_ACK, 0, $r14)
 
 	// context switch request?
 	ih_no_fifo:
-	and $r11 $r10 0x00000100
+	and $r11 $r10 NV_PGRAPH_FECS_INTR_CHSW
 	bra e #ih_no_ctxsw
 		// enqueue a context switch for later processing
 		mov $r13 #cmd_queue
 		mov $r14 0x4001
-		call #queue_put
+		call(queue_put)
 
-	// anything we didn't handle, bring it to the host's attention
+	// firmware method?
 	ih_no_ctxsw:
-	mov $r11 0x104
+	and $r11 $r10 NV_PGRAPH_FECS_INTR_FWMTHD
+	bra e #ih_no_fwmthd
+		// none we handle; report to host and ack
+		nv_rd32($r15, NV_PGRAPH_TRAPPED_DATA_LO)
+		nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_VAL(4), 0, $r15)
+		nv_rd32($r15, NV_PGRAPH_TRAPPED_ADDR)
+		nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_VAL(3), 0, $r15)
+		extr $r14 $r15 16:18
+		shl b32 $r14 $r14 2
+		imm32($r15, NV_PGRAPH_FE_OBJECT_TABLE(0))
+		add b32 $r14 $r15
+		call(nv_rd32)
+		nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_VAL(2), 0, $r15)
+		mov $r15 E_BAD_FWMTHD
+		call(error)
+		mov $r11 0x100
+		nv_wr32(0x400144, $r11)
+
+	// anything we didn't handle, bring it to the host's attention
+	ih_no_fwmthd:
+	mov $r11 0x504 // FIFO | CHSW | FWMTHD
 	not b32 $r11
 	and $r11 $r10 $r11
 	bra e #ih_no_other
-		mov $r10 0xc1c
-		shl b32 $r10 6
-		iowr I[$r10] $r11	// INTR_UP_SET
+		nv_iowr(NV_PGRAPH_FECS_INTR_UP_SET, 0, $r11)
 
 	// ack, and wake up main()
 	ih_no_other:
-	iowr I[$r0 + 0x100] $r10	// INTR_ACK
+	nv_iowr(NV_PGRAPH_FECS_INTR_ACK, 0, $r10)
 
 	pop $r15
 	pop $r14
@@ -370,12 +386,10 @@ ih:
 #if CHIPSET < GK100
 // Not real sure, but, MEM_CMD 7 will hang forever if this isn't done
 ctx_4160s:
-	mov $r14 0x4160
-	sethi $r14 0x400000
 	mov $r15 1
-	call #nv_wr32
+	nv_wr32(0x404160, $r15)
 	ctx_4160s_wait:
-		call #nv_rd32
+		nv_rd32($r15, 0x404160)
 		xbit $r15 $r15 4
 		bra e #ctx_4160s_wait
 	ret
@@ -384,10 +398,8 @@ ctx_4160s:
 // to hang with STATUS=0x00000007 until it's cleared.. fbcon can
 // still function with it set however...
 ctx_4160c:
-	mov $r14 0x4160
-	sethi $r14 0x400000
 	clear b32 $r15
-	call #nv_wr32
+	nv_wr32(0x404160, $r15)
 	ret
 #endif
 
@@ -396,18 +408,14 @@ ctx_4160c:
 // In: $r15 value to set 0x404170 to
 //
 ctx_4170s:
-	mov $r14 0x4170
-	sethi $r14 0x400000
 	or $r15 0x10
-	call #nv_wr32
+	nv_wr32(0x404170, $r15)
 	ret
 
 // Waits for a ctx_4170s() call to complete
 //
 ctx_4170w:
-	mov $r14 0x4170
-	sethi $r14 0x400000
-	call #nv_rd32
+	nv_rd32($r15, 0x404170)
 	and $r15 0x10
 	bra ne #ctx_4170w
 	ret
@@ -419,16 +427,18 @@ ctx_4170w:
 // funny things happen.
 //
 ctx_redswitch:
-	mov $r14 0x614
-	shl b32 $r14 6
-	mov $r15 0x270
-	iowr I[$r14] $r15	// HUB_RED_SWITCH = ENABLE_GPC, POWER_ALL
+	mov $r14 NV_PGRAPH_FECS_RED_SWITCH_ENABLE_GPC
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_POWER_ROP
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_POWER_GPC
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_POWER_MAIN
+	nv_iowr(NV_PGRAPH_FECS_RED_SWITCH, 0, $r14)
 	mov $r15 8
 	ctx_redswitch_delay:
 		sub b32 $r15 1
 		bra ne #ctx_redswitch_delay
-	mov $r15 0x770
-	iowr I[$r14] $r15	// HUB_RED_SWITCH = ENABLE_ALL, POWER_ALL
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_ENABLE_ROP
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_ENABLE_MAIN
+	nv_iowr(NV_PGRAPH_FECS_RED_SWITCH, 0, $r14)
 	ret
 
 // Not a clue what this is for, except that unless the value is 0x10, the
@@ -437,15 +447,18 @@ ctx_redswitch:
 // In: $r15 value to set to (0x00/0x10 are used)
 //
 ctx_86c:
-	mov $r14 0x86c
-	shl b32 $r14 6
-	iowr I[$r14] $r15	// HUB(0x86c) = val
-	mov $r14 -0x75ec
-	sethi $r14 0x400000
-	call #nv_wr32		// ROP(0xa14) = val
-	mov $r14 -0x5794
-	sethi $r14 0x410000
-	call #nv_wr32		// GPC(0x86c) = val
+	nv_iowr(NV_PGRAPH_FECS_UNK86C, 0, $r15)
+	nv_wr32(0x408a14, $r15)
+	nv_wr32(NV_PGRAPH_GPCX_GPCCS_UNK86C, $r15)
+	ret
+
+// In: $r15 NV_PGRAPH_FECS_MEM_CMD_*
+ctx_mem:
+	nv_iowr(NV_PGRAPH_FECS_MEM_CMD, 0, $r15)
+	ctx_mem_wait:
+		nv_iord($r15, NV_PGRAPH_FECS_MEM_CMD, 0)
+		or $r15 $r15
+		bra ne #ctx_mem_wait
 	ret
 
 // ctx_load - load's a channel's ctxctl data, and selects its vm
@@ -457,23 +470,14 @@ ctx_load:
 
 	// switch to channel, somewhat magic in parts..
 	mov $r10 12		// DONE_UNK12
-	call #wait_donez
-	mov $r1 0xa24
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r0	// 0x409a24
-	mov $r3 0xb00
-	shl b32 $r3 6
-	iowr I[$r3 + 0x100] $r2	// CHAN_NEXT
-	mov $r1 0xa0c
-	shl b32 $r1 6
-	mov $r4 7
-	iowr I[$r1 + 0x000] $r2 // MEM_CHAN
-	iowr I[$r1 + 0x100] $r4	// MEM_CMD
-	ctx_chan_wait_0:
-		iord $r4 I[$r1 + 0x100]
-		and $r4 0x1f
-		bra ne #ctx_chan_wait_0
-	iowr I[$r3 + 0x000] $r2	// CHAN_CUR
+	call(wait_donez)
+	clear b32 $r15
+	nv_iowr(0x409a24, 0, $r15)
+	nv_iowr(NV_PGRAPH_FECS_CHAN_NEXT, 0, $r2)
+	nv_iowr(NV_PGRAPH_FECS_MEM_CHAN, 0, $r2)
+	mov $r15 NV_PGRAPH_FECS_MEM_CMD_LOAD_CHAN
+	call(ctx_mem)
+	nv_iowr(NV_PGRAPH_FECS_CHAN_ADDR, 0, $r2)
 
 	// load channel header, fetch PGRAPH context pointer
 	mov $xtargets $r0
@@ -482,14 +486,10 @@ ctx_load:
 	add b32 $r2 2
 
 	trace_set(T_LCHAN)
-	mov $r1 0xa04
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r2		// MEM_BASE
-	mov $r1 0xa20
-	shl b32 $r1 6
-	mov $r2 0x0002
-	sethi $r2 0x80000000
-	iowr I[$r1 + 0x000] $r2		// MEM_TARGET = vram
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r2)
+	imm32($r2, NV_PGRAPH_FECS_MEM_TARGET_UNK31)
+	or  $r2 NV_PGRAPH_FECS_MEM_TARGET_AS_VRAM
+	nv_iowr(NV_PGRAPH_FECS_MEM_TARGET, 0, $r2)
 	mov $r1 0x10			// chan + 0x0210
 	mov $r2 #xfer_data
 	sethi $r2 0x00020000		// 16 bytes
@@ -507,13 +507,9 @@ ctx_load:
 
 	// set transfer base to start of context, and fetch context header
 	trace_set(T_LCTXH)
-	mov $r2 0xa04
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r1		// MEM_BASE
-	mov $r2 1
-	mov $r1 0xa20
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r2		// MEM_TARGET = vm
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r1)
+	mov $r2 NV_PGRAPH_FECS_MEM_TARGET_AS_VM
+	nv_iowr(NV_PGRAPH_FECS_MEM_TARGET, 0, $r2)
 	mov $r1 #chan_data
 	sethi $r1 0x00060000		// 256 bytes
 	xdld $r0 $r1
@@ -532,21 +528,15 @@ ctx_load:
 //
 ctx_chan:
 #if CHIPSET < GK100
-	call #ctx_4160s
+	call(ctx_4160s)
 #endif
-	call #ctx_load
+	call(ctx_load)
 	mov $r10 12			// DONE_UNK12
-	call #wait_donez
-	mov $r1 0xa10
-	shl b32 $r1 6
-	mov $r2 5
-	iowr I[$r1 + 0x000] $r2		// MEM_CMD = 5 (???)
-	ctx_chan_wait:
-		iord $r2 I[$r1 + 0x000]
-		or $r2 $r2
-		bra ne #ctx_chan_wait
+	call(wait_donez)
+	mov $r15 5 // MEM_CMD 5 ???
+	call(ctx_mem)
 #if CHIPSET < GK100
-	call #ctx_4160c
+	call(ctx_4160c)
 #endif
 	ret
 
@@ -562,9 +552,7 @@ ctx_chan:
 ctx_mmio_exec:
 	// set transfer base to be the mmio list
 	ld b32 $r3 D[$r0 + #chan_mmio_address]
-	mov $r2 0xa04
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r3		// MEM_BASE
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r3)
 
 	clear b32 $r3
 	ctx_mmio_loop:
@@ -580,7 +568,7 @@ ctx_mmio_exec:
 		ctx_mmio_pull:
 		ld b32 $r14 D[$r4 + #xfer_data + 0x00]
 		ld b32 $r15 D[$r4 + #xfer_data + 0x04]
-		call #nv_wr32
+		call(nv_wr32)
 
 		// next!
 		add b32 $r3 8
@@ -590,7 +578,7 @@ ctx_mmio_exec:
 	// set transfer base back to the current context
 	ctx_mmio_done:
 	ld b32 $r3 D[$r0 + #ctx_current]
-	iowr I[$r2 + 0x000] $r3		// MEM_BASE
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r3)
 
 	// disable the mmio list now, we don't need/want to execute it again
 	st b32 D[$r0 + #chan_mmio_count] $r0
@@ -610,12 +598,10 @@ ctx_mmio_exec:
 //
 ctx_xfer:
 	// according to mwk, some kind of wait for idle
-	mov $r15 0xc00
-	shl b32 $r15 6
 	mov $r14 4
-	iowr I[$r15 + 0x200] $r14
+	nv_iowr(0x409c08, 0, $r14)
 	ctx_xfer_idle:
-		iord $r14 I[$r15 + 0x000]
+		nv_iord($r14, 0x409c00, 0)
 		and $r14 0x2000
 		bra ne #ctx_xfer_idle
 
@@ -623,50 +609,42 @@ ctx_xfer:
 	bra $p2 #ctx_xfer_pre_load
 	ctx_xfer_pre:
 		mov $r15 0x10
-		call #ctx_86c
+		call(ctx_86c)
 #if CHIPSET < GK100
-		call #ctx_4160s
+		call(ctx_4160s)
 #endif
 		bra not $p1 #ctx_xfer_exec
 
 	ctx_xfer_pre_load:
 		mov $r15 2
-		call #ctx_4170s
-		call #ctx_4170w
-		call #ctx_redswitch
+		call(ctx_4170s)
+		call(ctx_4170w)
+		call(ctx_redswitch)
 		clear b32 $r15
-		call #ctx_4170s
-		call #ctx_load
+		call(ctx_4170s)
+		call(ctx_load)
 
 	// fetch context pointer, and initiate xfer on all GPCs
 	ctx_xfer_exec:
 	ld b32 $r1 D[$r0 + #ctx_current]
-	mov $r2 0x414
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r0	// BAR_STATUS = reset
-	mov $r14 -0x5b00
-	sethi $r14 0x410000
-	mov b32 $r15 $r1
-	call #nv_wr32		// GPC_BCAST_WRCMD_DATA = ctx pointer
-	add b32 $r14 4
+
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_FECS_BAR, 0, $r2)
+
+	nv_wr32(0x41a500, $r1)	// GPC_BCAST_WRCMD_DATA = ctx pointer
 	xbit $r15 $flags $p1
 	xbit $r2 $flags $p2
 	shl b32 $r2 1
 	or $r15 $r2
-	call #nv_wr32		// GPC_BCAST_WRCMD_CMD = GPC_XFER(type)
+	nv_wr32(0x41a504, $r15)	// GPC_BCAST_WRCMD_CMD = GPC_XFER(type)
 
 	// strands
-	mov $r1 0x4afc
-	sethi $r1 0x20000
-	mov $r2 0xc
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x0c
-	call #strand_wait
-	mov $r2 0x47fc
-	sethi $r2 0x20000
-	iowr I[$r2] $r0		// STRAND_FIRST_GENE(0x3f) = 0x00
-	xbit $r2 $flags $p1
-	add b32 $r2 3
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x03/0x04 (SAVE/LOAD)
+	call(strand_pre)
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_FECS_STRAND_SELECT, 0x3f, $r2)
+	xbit $r2 $flags $p1	// SAVE/LOAD
+	add b32 $r2 NV_PGRAPH_FECS_STRAND_CMD_SAVE
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r2)
 
 	// mmio context
 	xbit $r10 $flags $p1	// direction
@@ -675,48 +653,42 @@ ctx_xfer:
 	ld b32 $r12 D[$r0 + #hub_mmio_list_head]
 	ld b32 $r13 D[$r0 + #hub_mmio_list_tail]
 	mov $r14 0		// not multi
-	call #mmctx_xfer
+	call(mmctx_xfer)
 
 	// wait for GPCs to all complete
 	mov $r10 8		// DONE_BAR
-	call #wait_doneo
+	call(wait_doneo)
 
 	// wait for strand xfer to complete
-	call #strand_wait
+	call(strand_wait)
 
 	// post-op
 	bra $p1 #ctx_xfer_post
 		mov $r10 12		// DONE_UNK12
-		call #wait_donez
-		mov $r1 0xa10
-		shl b32 $r1 6
-		mov $r2 5
-		iowr I[$r1] $r2		// MEM_CMD
-		ctx_xfer_post_save_wait:
-			iord $r2 I[$r1]
-			or $r2 $r2
-			bra ne #ctx_xfer_post_save_wait
+		call(wait_donez)
+		mov $r15 5 // MEM_CMD 5 ???
+		call(ctx_mem)
 
 	bra $p2 #ctx_xfer_done
 	ctx_xfer_post:
 		mov $r15 2
-		call #ctx_4170s
+		call(ctx_4170s)
 		clear b32 $r15
-		call #ctx_86c
-		call #strand_post
-		call #ctx_4170w
+		call(ctx_86c)
+		call(strand_post)
+		call(ctx_4170w)
 		clear b32 $r15
-		call #ctx_4170s
+		call(ctx_4170s)
 
 		bra not $p1 #ctx_xfer_no_post_mmio
 		ld b32 $r1 D[$r0 + #chan_mmio_count]
 		or $r1 $r1
 		bra e #ctx_xfer_no_post_mmio
-			call #ctx_mmio_exec
+			call(ctx_mmio_exec)
 
 		ctx_xfer_no_post_mmio:
 #if CHIPSET < GK100
-		call #ctx_4160c
+		call(ctx_4160c)
 #endif
 
 	ctx_xfer_done:
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubgm107.fuc5 b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubgm107.fuc5
new file mode 100644
index 00000000000..27591b3086a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubgm107.fuc5
@@ -0,0 +1,40 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#define CHIPSET GK208
+#include "macros.fuc"
+
+.section #gm107_grhub_data
+#define INCLUDE_DATA
+#include "com.fuc"
+#include "hub.fuc"
+#undef INCLUDE_DATA
+
+.section #gm107_grhub_code
+#define INCLUDE_CODE
+bra #init
+#include "com.fuc"
+#include "hub.fuc"
+.align 256
+#undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubgm107.fuc5.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubgm107.fuc5.h
new file mode 100644
index 00000000000..5f953c5c20b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubgm107.fuc5.h
@@ -0,0 +1,916 @@
+uint32_t gm107_grhub_data[] = {
+/* 0x0000: hub_mmio_list_head */
+	0x00000300,
+/* 0x0004: hub_mmio_list_tail */
+	0x00000304,
+/* 0x0008: gpc_count */
+	0x00000000,
+/* 0x000c: rop_count */
+	0x00000000,
+/* 0x0010: cmd_queue */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0058: ctx_current */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0100: chan_data */
+/* 0x0100: chan_mmio_count */
+	0x00000000,
+/* 0x0104: chan_mmio_address */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0200: xfer_data */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0300: hub_mmio_list_base */
+	0x0417e91c,
+};
+
+uint32_t gm107_grhub_code[] = {
+	0x030e0ef5,
+/* 0x0004: queue_put */
+	0x9800d898,
+	0x86f001d9,
+	0xf489a408,
+	0x020f0b1b,
+	0x0002f87e,
+/* 0x001a: queue_put_next */
+	0x98c400f8,
+	0x0384b607,
+	0xb6008dbb,
+	0x8eb50880,
+	0x018fb500,
+	0xf00190b6,
+	0xd9b50f94,
+/* 0x0037: queue_get */
+	0xf400f801,
+	0xd8980131,
+	0x01d99800,
+	0x0bf489a4,
+	0x0789c421,
+	0xbb0394b6,
+	0x90b6009d,
+	0x009e9808,
+	0xb6019f98,
+	0x84f00180,
+	0x00d8b50f,
+/* 0x0063: queue_get_done */
+	0xf80132f4,
+/* 0x0065: nv_rd32 */
+	0xf0ecb200,
+	0x00801fc9,
+	0x0cf601ca,
+/* 0x0073: nv_rd32_wait */
+	0x8c04bd00,
+	0xcf01ca00,
+	0xccc800cc,
+	0xf61bf41f,
+	0xec7e060a,
+	0x008f0000,
+	0xffcf01cb,
+/* 0x008f: nv_wr32 */
+	0x8000f800,
+	0xf601cc00,
+	0x04bd000f,
+	0xc9f0ecb2,
+	0x1ec9f01f,
+	0x01ca0080,
+	0xbd000cf6,
+/* 0x00a9: nv_wr32_wait */
+	0xca008c04,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f61b,
+/* 0x00b8: wait_donez */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x00cf: wait_donez_ne */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf61bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x00ec: wait_doneo */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x0103: wait_doneo_e */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf60bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x0120: mmctx_size */
+/* 0x0122: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0x1bf4efa4,
+	0xf89fb2ec,
+/* 0x013d: mmctx_xfer */
+	0xf094bd00,
+	0x00800199,
+	0x09f60237,
+	0xbd04bd00,
+	0x05bbfd94,
+	0x800f0bf4,
+	0xf601c400,
+	0x04bd000b,
+/* 0x015f: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0xc6008018,
+	0x000ef601,
+	0x008004bd,
+	0x0ff601c7,
+	0xf004bd00,
+/* 0x017a: mmctx_multi_disabled */
+	0xabc80199,
+	0x10b4b600,
+	0xc80cb9f0,
+	0xe4b601ae,
+	0x05befd11,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x0195: mmctx_exec_loop */
+/* 0x0195: mmctx_wait_free */
+	0xc5008e04,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f60b,
+	0x05e9fd00,
+	0x01c80080,
+	0xbd000ef6,
+	0x04c0b604,
+	0x1bf4cda4,
+	0x02abc8df,
+/* 0x01bf: mmctx_fini_wait */
+	0x8b1c1bf4,
+	0xcf01c500,
+	0xb4f000bb,
+	0x10b4b01f,
+	0x0af31bf4,
+	0x00b87e05,
+	0x250ef400,
+/* 0x01d8: mmctx_stop */
+	0xb600abc8,
+	0xb9f010b4,
+	0x12b9f00c,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x01ed: mmctx_stop_wait */
+	0xc5008b04,
+	0x00bbcf01,
+	0xf412bbc8,
+/* 0x01fa: mmctx_done */
+	0x94bdf61b,
+	0x800199f0,
+	0xf6021700,
+	0x04bd0009,
+/* 0x020a: strand_wait */
+	0xa0f900f8,
+	0xb87e020a,
+	0xa0fc0000,
+/* 0x0216: strand_pre */
+	0x0c0900f8,
+	0x024afc80,
+	0xbd0009f6,
+	0x020a7e04,
+/* 0x0227: strand_post */
+	0x0900f800,
+	0x4afc800d,
+	0x0009f602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0238: strand_set */
+	0xfc800f0c,
+	0x0cf6024f,
+	0x0c04bd00,
+	0x4afc800b,
+	0x000cf602,
+	0xfc8004bd,
+	0x0ef6024f,
+	0x0c04bd00,
+	0x4afc800a,
+	0x000cf602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0268: strand_ctx_init */
+	0x99f094bd,
+	0x37008003,
+	0x0009f602,
+	0x167e04bd,
+	0x030e0002,
+	0x0002387e,
+	0xfc80c4bd,
+	0x0cf60247,
+	0x0c04bd00,
+	0x4afc8001,
+	0x000cf602,
+	0x0a7e04bd,
+	0x0c920002,
+	0x46fc8001,
+	0x000cf602,
+	0x020c04bd,
+	0x024afc80,
+	0xbd000cf6,
+	0x020a7e04,
+	0x02277e00,
+	0x42008800,
+	0x20008902,
+	0x0099cf02,
+/* 0x02c7: ctx_init_strand_loop */
+	0xf608fe95,
+	0x8ef6008e,
+	0x808acf40,
+	0xb606a5b6,
+	0xeabb01a0,
+	0x0480b600,
+	0xf40192b6,
+	0xe4b6e81b,
+	0xf2efbc08,
+	0x99f094bd,
+	0x17008003,
+	0x0009f602,
+	0x00f804bd,
+/* 0x02f8: error */
+	0x02050080,
+	0xbd000ff6,
+	0x80010f04,
+	0xf6030700,
+	0x04bd000f,
+/* 0x030e: init */
+	0x04bd00f8,
+	0x410007fe,
+	0x11cf4200,
+	0x0911e700,
+	0x0814b601,
+	0x020014fe,
+	0x12004002,
+	0xbd0002f6,
+	0x05c94104,
+	0xbd0010fe,
+	0x07004024,
+	0xbd0002f6,
+	0x20034204,
+	0x01010080,
+	0xbd0002f6,
+	0x20044204,
+	0x01010480,
+	0xbd0002f6,
+	0x200b4204,
+	0x01010880,
+	0xbd0002f6,
+	0x200c4204,
+	0x01011c80,
+	0xbd0002f6,
+	0x01039204,
+	0x03090080,
+	0xbd0003f6,
+	0x87044204,
+	0xf6040040,
+	0x04bd0002,
+	0x00400402,
+	0x0002f603,
+	0x31f404bd,
+	0x96048e10,
+	0x00657e40,
+	0xc7feb200,
+	0x01b590f1,
+	0x1ff4f003,
+	0x01020fb5,
+	0x041fbb01,
+	0x800112b6,
+	0xf6010300,
+	0x04bd0001,
+	0x01040080,
+	0xbd0001f6,
+	0x01004104,
+	0xa87e020f,
+	0xb77e0006,
+	0x100f0006,
+	0x0006f97e,
+	0x98000e98,
+	0x207e010f,
+	0x14950001,
+	0xc0008008,
+	0x0004f601,
+	0x008004bd,
+	0x04f601c1,
+	0xb704bd00,
+	0xbb130030,
+	0xf5b6001f,
+	0xd3008002,
+	0x000ff601,
+	0x15b604bd,
+	0x0110b608,
+	0xb20814b6,
+	0x02687e1f,
+	0x001fbb00,
+	0x84020398,
+/* 0x041f: init_gpc */
+	0xb8502000,
+	0x0008044e,
+	0x8f7e1fb2,
+	0x4eb80000,
+	0xbd00010c,
+	0x008f7ef4,
+	0x044eb800,
+	0x8f7e0001,
+	0x4eb80000,
+	0x0f000100,
+	0x008f7e02,
+	0x004eb800,
+/* 0x044e: init_gpc_wait */
+	0x657e0008,
+	0xffc80000,
+	0xf90bf41f,
+	0x08044eb8,
+	0x00657e00,
+	0x001fbb00,
+	0x800040b7,
+	0xf40132b6,
+	0x000fb41b,
+	0x0006f97e,
+	0xa87e000f,
+	0x00800006,
+	0x01f60201,
+	0xbd04bd00,
+	0x1f19f014,
+	0x02300080,
+	0xbd0001f6,
+/* 0x0491: main */
+	0x0031f404,
+	0x0d0028f4,
+	0x00377e10,
+	0xf401f400,
+	0x4001e4b1,
+	0x00c71bf5,
+	0x99f094bd,
+	0x37008004,
+	0x0009f602,
+	0x008104bd,
+	0x11cf02c0,
+	0xc1008200,
+	0x0022cf02,
+	0xf41f13c8,
+	0x23c8770b,
+	0x550bf41f,
+	0x12b220f9,
+	0x99f094bd,
+	0x37008007,
+	0x0009f602,
+	0x32f404bd,
+	0x0231f401,
+	0x00087c7e,
+	0x99f094bd,
+	0x17008007,
+	0x0009f602,
+	0x20fc04bd,
+	0x99f094bd,
+	0x37008006,
+	0x0009f602,
+	0x31f404bd,
+	0x087c7e01,
+	0xf094bd00,
+	0x00800699,
+	0x09f60217,
+	0xf404bd00,
+/* 0x0522: chsw_prev_no_next */
+	0x20f92f0e,
+	0x32f412b2,
+	0x0232f401,
+	0x00087c7e,
+	0x008020fc,
+	0x02f602c0,
+	0xf404bd00,
+/* 0x053e: chsw_no_prev */
+	0x23c8130e,
+	0x0d0bf41f,
+	0xf40131f4,
+	0x7c7e0232,
+/* 0x054e: chsw_done */
+	0x01020008,
+	0x02c30080,
+	0xbd0002f6,
+	0xf094bd04,
+	0x00800499,
+	0x09f60217,
+	0xf504bd00,
+/* 0x056b: main_not_ctx_switch */
+	0xb0ff2a0e,
+	0x1bf401e4,
+	0x7ef2b20c,
+	0xf400081c,
+/* 0x057a: main_not_ctx_chan */
+	0xe4b0400e,
+	0x2c1bf402,
+	0x99f094bd,
+	0x37008007,
+	0x0009f602,
+	0x32f404bd,
+	0x0232f401,
+	0x00087c7e,
+	0x99f094bd,
+	0x17008007,
+	0x0009f602,
+	0x0ef404bd,
+/* 0x05a9: main_not_ctx_save */
+	0x10ef9411,
+	0x7e01f5f0,
+	0xf50002f8,
+/* 0x05b7: main_done */
+	0xbdfede0e,
+	0x1f29f024,
+	0x02300080,
+	0xbd0002f6,
+	0xcc0ef504,
+/* 0x05c9: ih */
+	0xfe80f9fe,
+	0x80f90188,
+	0xa0f990f9,
+	0xd0f9b0f9,
+	0xf0f9e0f9,
+	0x004a04bd,
+	0x00aacf02,
+	0xf404abc4,
+	0x100d230b,
+	0xcf1a004e,
+	0x004f00ee,
+	0x00ffcf19,
+	0x0000047e,
+	0x0400b0b7,
+	0x0040010e,
+	0x000ef61d,
+/* 0x060a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x4e100d0c,
+	0x047e4001,
+/* 0x061a: ih_no_ctxsw */
+	0xabe40000,
+	0x0bf40400,
+	0x07088e56,
+	0x00657e40,
+	0x80ffb200,
+	0xf6020400,
+	0x04bd000f,
+	0x4007048e,
+	0x0000657e,
+	0x0080ffb2,
+	0x0ff60203,
+	0xc704bd00,
+	0xee9450fe,
+	0x07008f02,
+	0x00efbb40,
+	0x0000657e,
+	0x02020080,
+	0xbd000ff6,
+	0x7e030f04,
+	0x4b0002f8,
+	0xbfb20100,
+	0x4001448e,
+	0x00008f7e,
+/* 0x0674: ih_no_fwmthd */
+	0xbd05044b,
+	0xb4abffb0,
+	0x800c0bf4,
+	0xf6030700,
+	0x04bd000b,
+/* 0x0688: ih_no_other */
+	0xf6010040,
+	0x04bd000a,
+	0xe0fcf0fc,
+	0xb0fcd0fc,
+	0x90fca0fc,
+	0x88fe80fc,
+	0xf480fc00,
+	0x01f80032,
+/* 0x06a8: ctx_4170s */
+	0xb210f5f0,
+	0x41708eff,
+	0x008f7e40,
+/* 0x06b7: ctx_4170w */
+	0x8e00f800,
+	0x7e404170,
+	0xb2000065,
+	0x10f4f0ff,
+	0xf8f31bf4,
+/* 0x06c9: ctx_redswitch */
+	0x02004e00,
+	0xf040e5f0,
+	0xe5f020e5,
+	0x85008010,
+	0x000ef601,
+	0x080f04bd,
+/* 0x06e0: ctx_redswitch_delay */
+	0xf401f2b6,
+	0xe5f1fd1b,
+	0xe5f10400,
+	0x00800100,
+	0x0ef60185,
+	0xf804bd00,
+/* 0x06f9: ctx_86c */
+	0x23008000,
+	0x000ff602,
+	0xffb204bd,
+	0x408a148e,
+	0x00008f7e,
+	0x8c8effb2,
+	0x8f7e41a8,
+	0x00f80000,
+/* 0x0718: ctx_mem */
+	0x02840080,
+	0xbd000ff6,
+/* 0x0721: ctx_mem_wait */
+	0x84008f04,
+	0x00ffcf02,
+	0xf405fffd,
+	0x00f8f61b,
+/* 0x0730: ctx_load */
+	0x99f094bd,
+	0x37008005,
+	0x0009f602,
+	0x0c0a04bd,
+	0x0000b87e,
+	0x0080f4bd,
+	0x0ff60289,
+	0x8004bd00,
+	0xf602c100,
+	0x04bd0002,
+	0x02830080,
+	0xbd0002f6,
+	0x7e070f04,
+	0x80000718,
+	0xf602c000,
+	0x04bd0002,
+	0xf0000bfe,
+	0x24b61f2a,
+	0x0220b604,
+	0x99f094bd,
+	0x37008008,
+	0x0009f602,
+	0x008004bd,
+	0x02f60281,
+	0xd204bd00,
+	0x80000000,
+	0x800225f0,
+	0xf6028800,
+	0x04bd0002,
+	0x00421001,
+	0x0223f002,
+	0xf80512fa,
+	0xf094bd03,
+	0x00800899,
+	0x09f60217,
+	0x9804bd00,
+	0x14b68101,
+	0x80029818,
+	0xfd0825b6,
+	0x01b50512,
+	0xf094bd16,
+	0x00800999,
+	0x09f60237,
+	0x8004bd00,
+	0xf6028100,
+	0x04bd0001,
+	0x00800102,
+	0x02f60288,
+	0x4104bd00,
+	0x13f00100,
+	0x0501fa06,
+	0x94bd03f8,
+	0x800999f0,
+	0xf6021700,
+	0x04bd0009,
+	0x99f094bd,
+	0x17008005,
+	0x0009f602,
+	0x00f804bd,
+/* 0x081c: ctx_chan */
+	0x0007307e,
+	0xb87e0c0a,
+	0x050f0000,
+	0x0007187e,
+/* 0x082e: ctx_mmio_exec */
+	0x039800f8,
+	0x81008041,
+	0x0003f602,
+	0x34bd04bd,
+/* 0x083c: ctx_mmio_loop */
+	0xf4ff34c4,
+	0x00450e1b,
+	0x0653f002,
+	0xf80535fa,
+/* 0x084d: ctx_mmio_pull */
+	0x804e9803,
+	0x7e814f98,
+	0xb600008f,
+	0x12b60830,
+	0xdf1bf401,
+/* 0x0860: ctx_mmio_done */
+	0x80160398,
+	0xf6028100,
+	0x04bd0003,
+	0x414000b5,
+	0x13f00100,
+	0x0601fa06,
+	0x00f803f8,
+/* 0x087c: ctx_xfer */
+	0x0080040e,
+	0x0ef60302,
+/* 0x0887: ctx_xfer_idle */
+	0x8e04bd00,
+	0xcf030000,
+	0xe4f100ee,
+	0x1bf42000,
+	0x0611f4f5,
+/* 0x089b: ctx_xfer_pre */
+	0x0f0c02f4,
+	0x06f97e10,
+	0x1b11f400,
+/* 0x08a4: ctx_xfer_pre_load */
+	0xa87e020f,
+	0xb77e0006,
+	0xc97e0006,
+	0xf4bd0006,
+	0x0006a87e,
+	0x0007307e,
+/* 0x08bc: ctx_xfer_exec */
+	0xbd160198,
+	0x05008024,
+	0x0002f601,
+	0x1fb204bd,
+	0x41a5008e,
+	0x00008f7e,
+	0xf001fcf0,
+	0x24b6022c,
+	0x05f2fd01,
+	0x048effb2,
+	0x8f7e41a5,
+	0x167e0000,
+	0x24bd0002,
+	0x0247fc80,
+	0xbd0002f6,
+	0x012cf004,
+	0x800320b6,
+	0xf6024afc,
+	0x04bd0002,
+	0xf001acf0,
+	0x000b06a5,
+	0x98000c98,
+	0x000e010d,
+	0x00013d7e,
+	0xec7e080a,
+	0x0a7e0000,
+	0x01f40002,
+	0x7e0c0a12,
+	0x0f0000b8,
+	0x07187e05,
+	0x2d02f400,
+/* 0x0938: ctx_xfer_post */
+	0xa87e020f,
+	0xf4bd0006,
+	0x0006f97e,
+	0x0002277e,
+	0x0006b77e,
+	0xa87ef4bd,
+	0x11f40006,
+	0x40019810,
+	0xf40511fd,
+	0x2e7e070b,
+/* 0x0962: ctx_xfer_no_post_mmio */
+/* 0x0962: ctx_xfer_done */
+	0x00f80008,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+};
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5 b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5
new file mode 100644
index 00000000000..7c5d25630fa
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5
@@ -0,0 +1,40 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#define CHIPSET GK208
+#include "macros.fuc"
+
+.section #nv108_grhub_data
+#define INCLUDE_DATA
+#include "com.fuc"
+#include "hub.fuc"
+#undef INCLUDE_DATA
+
+.section #nv108_grhub_code
+#define INCLUDE_CODE
+bra #init
+#include "com.fuc"
+#include "hub.fuc"
+.align 256
+#undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5.h
new file mode 100644
index 00000000000..e49b5a877ae
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5.h
@@ -0,0 +1,916 @@
+uint32_t nv108_grhub_data[] = {
+/* 0x0000: hub_mmio_list_head */
+	0x00000300,
+/* 0x0004: hub_mmio_list_tail */
+	0x00000304,
+/* 0x0008: gpc_count */
+	0x00000000,
+/* 0x000c: rop_count */
+	0x00000000,
+/* 0x0010: cmd_queue */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0058: ctx_current */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0100: chan_data */
+/* 0x0100: chan_mmio_count */
+	0x00000000,
+/* 0x0104: chan_mmio_address */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0200: xfer_data */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0300: hub_mmio_list_base */
+	0x0417e91c,
+};
+
+uint32_t nv108_grhub_code[] = {
+	0x030e0ef5,
+/* 0x0004: queue_put */
+	0x9800d898,
+	0x86f001d9,
+	0xf489a408,
+	0x020f0b1b,
+	0x0002f87e,
+/* 0x001a: queue_put_next */
+	0x98c400f8,
+	0x0384b607,
+	0xb6008dbb,
+	0x8eb50880,
+	0x018fb500,
+	0xf00190b6,
+	0xd9b50f94,
+/* 0x0037: queue_get */
+	0xf400f801,
+	0xd8980131,
+	0x01d99800,
+	0x0bf489a4,
+	0x0789c421,
+	0xbb0394b6,
+	0x90b6009d,
+	0x009e9808,
+	0xb6019f98,
+	0x84f00180,
+	0x00d8b50f,
+/* 0x0063: queue_get_done */
+	0xf80132f4,
+/* 0x0065: nv_rd32 */
+	0xf0ecb200,
+	0x00801fc9,
+	0x0cf601ca,
+/* 0x0073: nv_rd32_wait */
+	0x8c04bd00,
+	0xcf01ca00,
+	0xccc800cc,
+	0xf61bf41f,
+	0xec7e060a,
+	0x008f0000,
+	0xffcf01cb,
+/* 0x008f: nv_wr32 */
+	0x8000f800,
+	0xf601cc00,
+	0x04bd000f,
+	0xc9f0ecb2,
+	0x1ec9f01f,
+	0x01ca0080,
+	0xbd000cf6,
+/* 0x00a9: nv_wr32_wait */
+	0xca008c04,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f61b,
+/* 0x00b8: wait_donez */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x00cf: wait_donez_ne */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf61bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x00ec: wait_doneo */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x0103: wait_doneo_e */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf60bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x0120: mmctx_size */
+/* 0x0122: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0x1bf4efa4,
+	0xf89fb2ec,
+/* 0x013d: mmctx_xfer */
+	0xf094bd00,
+	0x00800199,
+	0x09f60237,
+	0xbd04bd00,
+	0x05bbfd94,
+	0x800f0bf4,
+	0xf601c400,
+	0x04bd000b,
+/* 0x015f: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0xc6008018,
+	0x000ef601,
+	0x008004bd,
+	0x0ff601c7,
+	0xf004bd00,
+/* 0x017a: mmctx_multi_disabled */
+	0xabc80199,
+	0x10b4b600,
+	0xc80cb9f0,
+	0xe4b601ae,
+	0x05befd11,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x0195: mmctx_exec_loop */
+/* 0x0195: mmctx_wait_free */
+	0xc5008e04,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f60b,
+	0x05e9fd00,
+	0x01c80080,
+	0xbd000ef6,
+	0x04c0b604,
+	0x1bf4cda4,
+	0x02abc8df,
+/* 0x01bf: mmctx_fini_wait */
+	0x8b1c1bf4,
+	0xcf01c500,
+	0xb4f000bb,
+	0x10b4b01f,
+	0x0af31bf4,
+	0x00b87e05,
+	0x250ef400,
+/* 0x01d8: mmctx_stop */
+	0xb600abc8,
+	0xb9f010b4,
+	0x12b9f00c,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x01ed: mmctx_stop_wait */
+	0xc5008b04,
+	0x00bbcf01,
+	0xf412bbc8,
+/* 0x01fa: mmctx_done */
+	0x94bdf61b,
+	0x800199f0,
+	0xf6021700,
+	0x04bd0009,
+/* 0x020a: strand_wait */
+	0xa0f900f8,
+	0xb87e020a,
+	0xa0fc0000,
+/* 0x0216: strand_pre */
+	0x0c0900f8,
+	0x024afc80,
+	0xbd0009f6,
+	0x020a7e04,
+/* 0x0227: strand_post */
+	0x0900f800,
+	0x4afc800d,
+	0x0009f602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0238: strand_set */
+	0xfc800f0c,
+	0x0cf6024f,
+	0x0c04bd00,
+	0x4afc800b,
+	0x000cf602,
+	0xfc8004bd,
+	0x0ef6024f,
+	0x0c04bd00,
+	0x4afc800a,
+	0x000cf602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0268: strand_ctx_init */
+	0x99f094bd,
+	0x37008003,
+	0x0009f602,
+	0x167e04bd,
+	0x030e0002,
+	0x0002387e,
+	0xfc80c4bd,
+	0x0cf60247,
+	0x0c04bd00,
+	0x4afc8001,
+	0x000cf602,
+	0x0a7e04bd,
+	0x0c920002,
+	0x46fc8001,
+	0x000cf602,
+	0x020c04bd,
+	0x024afc80,
+	0xbd000cf6,
+	0x020a7e04,
+	0x02277e00,
+	0x42008800,
+	0x20008902,
+	0x0099cf02,
+/* 0x02c7: ctx_init_strand_loop */
+	0xf608fe95,
+	0x8ef6008e,
+	0x808acf40,
+	0xb606a5b6,
+	0xeabb01a0,
+	0x0480b600,
+	0xf40192b6,
+	0xe4b6e81b,
+	0xf2efbc08,
+	0x99f094bd,
+	0x17008003,
+	0x0009f602,
+	0x00f804bd,
+/* 0x02f8: error */
+	0x02050080,
+	0xbd000ff6,
+	0x80010f04,
+	0xf6030700,
+	0x04bd000f,
+/* 0x030e: init */
+	0x04bd00f8,
+	0x410007fe,
+	0x11cf4200,
+	0x0911e700,
+	0x0814b601,
+	0x020014fe,
+	0x12004002,
+	0xbd0002f6,
+	0x05c94104,
+	0xbd0010fe,
+	0x07004024,
+	0xbd0002f6,
+	0x20034204,
+	0x01010080,
+	0xbd0002f6,
+	0x20044204,
+	0x01010480,
+	0xbd0002f6,
+	0x200b4204,
+	0x01010880,
+	0xbd0002f6,
+	0x200c4204,
+	0x01011c80,
+	0xbd0002f6,
+	0x01039204,
+	0x03090080,
+	0xbd0003f6,
+	0x87044204,
+	0xf6040040,
+	0x04bd0002,
+	0x00400402,
+	0x0002f603,
+	0x31f404bd,
+	0x96048e10,
+	0x00657e40,
+	0xc7feb200,
+	0x01b590f1,
+	0x1ff4f003,
+	0x01020fb5,
+	0x041fbb01,
+	0x800112b6,
+	0xf6010300,
+	0x04bd0001,
+	0x01040080,
+	0xbd0001f6,
+	0x01004104,
+	0xa87e020f,
+	0xb77e0006,
+	0x100f0006,
+	0x0006f97e,
+	0x98000e98,
+	0x207e010f,
+	0x14950001,
+	0xc0008008,
+	0x0004f601,
+	0x008004bd,
+	0x04f601c1,
+	0xb704bd00,
+	0xbb130030,
+	0xf5b6001f,
+	0xd3008002,
+	0x000ff601,
+	0x15b604bd,
+	0x0110b608,
+	0xb20814b6,
+	0x02687e1f,
+	0x001fbb00,
+	0x84020398,
+/* 0x041f: init_gpc */
+	0xb8502000,
+	0x0008044e,
+	0x8f7e1fb2,
+	0x4eb80000,
+	0xbd00010c,
+	0x008f7ef4,
+	0x044eb800,
+	0x8f7e0001,
+	0x4eb80000,
+	0x0f000100,
+	0x008f7e02,
+	0x004eb800,
+/* 0x044e: init_gpc_wait */
+	0x657e0008,
+	0xffc80000,
+	0xf90bf41f,
+	0x08044eb8,
+	0x00657e00,
+	0x001fbb00,
+	0x800040b7,
+	0xf40132b6,
+	0x000fb41b,
+	0x0006f97e,
+	0xa87e000f,
+	0x00800006,
+	0x01f60201,
+	0xbd04bd00,
+	0x1f19f014,
+	0x02300080,
+	0xbd0001f6,
+/* 0x0491: main */
+	0x0031f404,
+	0x0d0028f4,
+	0x00377e10,
+	0xf401f400,
+	0x4001e4b1,
+	0x00c71bf5,
+	0x99f094bd,
+	0x37008004,
+	0x0009f602,
+	0x008104bd,
+	0x11cf02c0,
+	0xc1008200,
+	0x0022cf02,
+	0xf41f13c8,
+	0x23c8770b,
+	0x550bf41f,
+	0x12b220f9,
+	0x99f094bd,
+	0x37008007,
+	0x0009f602,
+	0x32f404bd,
+	0x0231f401,
+	0x00087c7e,
+	0x99f094bd,
+	0x17008007,
+	0x0009f602,
+	0x20fc04bd,
+	0x99f094bd,
+	0x37008006,
+	0x0009f602,
+	0x31f404bd,
+	0x087c7e01,
+	0xf094bd00,
+	0x00800699,
+	0x09f60217,
+	0xf404bd00,
+/* 0x0522: chsw_prev_no_next */
+	0x20f92f0e,
+	0x32f412b2,
+	0x0232f401,
+	0x00087c7e,
+	0x008020fc,
+	0x02f602c0,
+	0xf404bd00,
+/* 0x053e: chsw_no_prev */
+	0x23c8130e,
+	0x0d0bf41f,
+	0xf40131f4,
+	0x7c7e0232,
+/* 0x054e: chsw_done */
+	0x01020008,
+	0x02c30080,
+	0xbd0002f6,
+	0xf094bd04,
+	0x00800499,
+	0x09f60217,
+	0xf504bd00,
+/* 0x056b: main_not_ctx_switch */
+	0xb0ff2a0e,
+	0x1bf401e4,
+	0x7ef2b20c,
+	0xf400081c,
+/* 0x057a: main_not_ctx_chan */
+	0xe4b0400e,
+	0x2c1bf402,
+	0x99f094bd,
+	0x37008007,
+	0x0009f602,
+	0x32f404bd,
+	0x0232f401,
+	0x00087c7e,
+	0x99f094bd,
+	0x17008007,
+	0x0009f602,
+	0x0ef404bd,
+/* 0x05a9: main_not_ctx_save */
+	0x10ef9411,
+	0x7e01f5f0,
+	0xf50002f8,
+/* 0x05b7: main_done */
+	0xbdfede0e,
+	0x1f29f024,
+	0x02300080,
+	0xbd0002f6,
+	0xcc0ef504,
+/* 0x05c9: ih */
+	0xfe80f9fe,
+	0x80f90188,
+	0xa0f990f9,
+	0xd0f9b0f9,
+	0xf0f9e0f9,
+	0x004a04bd,
+	0x00aacf02,
+	0xf404abc4,
+	0x100d230b,
+	0xcf1a004e,
+	0x004f00ee,
+	0x00ffcf19,
+	0x0000047e,
+	0x0400b0b7,
+	0x0040010e,
+	0x000ef61d,
+/* 0x060a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x4e100d0c,
+	0x047e4001,
+/* 0x061a: ih_no_ctxsw */
+	0xabe40000,
+	0x0bf40400,
+	0x07088e56,
+	0x00657e40,
+	0x80ffb200,
+	0xf6020400,
+	0x04bd000f,
+	0x4007048e,
+	0x0000657e,
+	0x0080ffb2,
+	0x0ff60203,
+	0xc704bd00,
+	0xee9450fe,
+	0x07008f02,
+	0x00efbb40,
+	0x0000657e,
+	0x02020080,
+	0xbd000ff6,
+	0x7e030f04,
+	0x4b0002f8,
+	0xbfb20100,
+	0x4001448e,
+	0x00008f7e,
+/* 0x0674: ih_no_fwmthd */
+	0xbd05044b,
+	0xb4abffb0,
+	0x800c0bf4,
+	0xf6030700,
+	0x04bd000b,
+/* 0x0688: ih_no_other */
+	0xf6010040,
+	0x04bd000a,
+	0xe0fcf0fc,
+	0xb0fcd0fc,
+	0x90fca0fc,
+	0x88fe80fc,
+	0xf480fc00,
+	0x01f80032,
+/* 0x06a8: ctx_4170s */
+	0xb210f5f0,
+	0x41708eff,
+	0x008f7e40,
+/* 0x06b7: ctx_4170w */
+	0x8e00f800,
+	0x7e404170,
+	0xb2000065,
+	0x10f4f0ff,
+	0xf8f31bf4,
+/* 0x06c9: ctx_redswitch */
+	0x02004e00,
+	0xf040e5f0,
+	0xe5f020e5,
+	0x85008010,
+	0x000ef601,
+	0x080f04bd,
+/* 0x06e0: ctx_redswitch_delay */
+	0xf401f2b6,
+	0xe5f1fd1b,
+	0xe5f10400,
+	0x00800100,
+	0x0ef60185,
+	0xf804bd00,
+/* 0x06f9: ctx_86c */
+	0x23008000,
+	0x000ff602,
+	0xffb204bd,
+	0x408a148e,
+	0x00008f7e,
+	0x8c8effb2,
+	0x8f7e41a8,
+	0x00f80000,
+/* 0x0718: ctx_mem */
+	0x02840080,
+	0xbd000ff6,
+/* 0x0721: ctx_mem_wait */
+	0x84008f04,
+	0x00ffcf02,
+	0xf405fffd,
+	0x00f8f61b,
+/* 0x0730: ctx_load */
+	0x99f094bd,
+	0x37008005,
+	0x0009f602,
+	0x0c0a04bd,
+	0x0000b87e,
+	0x0080f4bd,
+	0x0ff60289,
+	0x8004bd00,
+	0xf602c100,
+	0x04bd0002,
+	0x02830080,
+	0xbd0002f6,
+	0x7e070f04,
+	0x80000718,
+	0xf602c000,
+	0x04bd0002,
+	0xf0000bfe,
+	0x24b61f2a,
+	0x0220b604,
+	0x99f094bd,
+	0x37008008,
+	0x0009f602,
+	0x008004bd,
+	0x02f60281,
+	0xd204bd00,
+	0x80000000,
+	0x800225f0,
+	0xf6028800,
+	0x04bd0002,
+	0x00421001,
+	0x0223f002,
+	0xf80512fa,
+	0xf094bd03,
+	0x00800899,
+	0x09f60217,
+	0x9804bd00,
+	0x14b68101,
+	0x80029818,
+	0xfd0825b6,
+	0x01b50512,
+	0xf094bd16,
+	0x00800999,
+	0x09f60237,
+	0x8004bd00,
+	0xf6028100,
+	0x04bd0001,
+	0x00800102,
+	0x02f60288,
+	0x4104bd00,
+	0x13f00100,
+	0x0501fa06,
+	0x94bd03f8,
+	0x800999f0,
+	0xf6021700,
+	0x04bd0009,
+	0x99f094bd,
+	0x17008005,
+	0x0009f602,
+	0x00f804bd,
+/* 0x081c: ctx_chan */
+	0x0007307e,
+	0xb87e0c0a,
+	0x050f0000,
+	0x0007187e,
+/* 0x082e: ctx_mmio_exec */
+	0x039800f8,
+	0x81008041,
+	0x0003f602,
+	0x34bd04bd,
+/* 0x083c: ctx_mmio_loop */
+	0xf4ff34c4,
+	0x00450e1b,
+	0x0653f002,
+	0xf80535fa,
+/* 0x084d: ctx_mmio_pull */
+	0x804e9803,
+	0x7e814f98,
+	0xb600008f,
+	0x12b60830,
+	0xdf1bf401,
+/* 0x0860: ctx_mmio_done */
+	0x80160398,
+	0xf6028100,
+	0x04bd0003,
+	0x414000b5,
+	0x13f00100,
+	0x0601fa06,
+	0x00f803f8,
+/* 0x087c: ctx_xfer */
+	0x0080040e,
+	0x0ef60302,
+/* 0x0887: ctx_xfer_idle */
+	0x8e04bd00,
+	0xcf030000,
+	0xe4f100ee,
+	0x1bf42000,
+	0x0611f4f5,
+/* 0x089b: ctx_xfer_pre */
+	0x0f0c02f4,
+	0x06f97e10,
+	0x1b11f400,
+/* 0x08a4: ctx_xfer_pre_load */
+	0xa87e020f,
+	0xb77e0006,
+	0xc97e0006,
+	0xf4bd0006,
+	0x0006a87e,
+	0x0007307e,
+/* 0x08bc: ctx_xfer_exec */
+	0xbd160198,
+	0x05008024,
+	0x0002f601,
+	0x1fb204bd,
+	0x41a5008e,
+	0x00008f7e,
+	0xf001fcf0,
+	0x24b6022c,
+	0x05f2fd01,
+	0x048effb2,
+	0x8f7e41a5,
+	0x167e0000,
+	0x24bd0002,
+	0x0247fc80,
+	0xbd0002f6,
+	0x012cf004,
+	0x800320b6,
+	0xf6024afc,
+	0x04bd0002,
+	0xf001acf0,
+	0x000b06a5,
+	0x98000c98,
+	0x000e010d,
+	0x00013d7e,
+	0xec7e080a,
+	0x0a7e0000,
+	0x01f40002,
+	0x7e0c0a12,
+	0x0f0000b8,
+	0x07187e05,
+	0x2d02f400,
+/* 0x0938: ctx_xfer_post */
+	0xa87e020f,
+	0xf4bd0006,
+	0x0006f97e,
+	0x0002277e,
+	0x0006b77e,
+	0xa87ef4bd,
+	0x11f40006,
+	0x40019810,
+	0xf40511fd,
+	0x2e7e070b,
+/* 0x0962: ctx_xfer_no_post_mmio */
+/* 0x0962: ctx_xfer_done */
+	0x00f80008,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+};
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvc0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvc0.fuc.h
index b59f694c042..92dfe6a4ac8 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvc0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvc0.fuc.h
@@ -206,14 +206,14 @@ uint32_t nvc0_grhub_data[] = {
 };
 
 uint32_t nvc0_grhub_code[] = {
-	0x031b0ef5,
+	0x039b0ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -237,184 +237,214 @@ uint32_t nvc0_grhub_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f405,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -428,7 +458,7 @@ uint32_t nvc0_grhub_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0x07f100f8,
 	0x03f00500,
 	0x000fd002,
@@ -436,82 +466,117 @@ uint32_t nvc0_grhub_code[] = {
 	0x0007f101,
 	0x0303f007,
 	0xbd000fd0,
-/* 0x031b: init */
+/* 0x039b: init */
 	0xbd00f804,
-	0x0004fe04,
-	0xf10007fe,
-	0xf0120017,
-	0x12d00227,
-	0xb117f100,
-	0x0010fe05,
-	0x040017f1,
-	0xf1c010d0,
-	0xb6040437,
-	0x27f10634,
-	0x32d02003,
-	0x0427f100,
-	0x0132d020,
+	0x0007fe04,
+	0x420017f1,
+	0xcf0013f0,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf10227f0,
+	0xf0120007,
+	0x02d00003,
+	0xf104bd00,
+	0xfe06c817,
+	0x24bd0010,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0x200327f1,
+	0x010007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200427f1,
+	0x010407f1,
+	0xd00103f0,
+	0x04bd0002,
 	0x200b27f1,
-	0xf10232d0,
-	0xd0200c27,
-	0x27f10732,
-	0x24b60c24,
-	0x0003b906,
-	0xf10023d0,
+	0x010807f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200c27f1,
+	0x011c07f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1010392,
+	0xf0090007,
+	0x03d00303,
+	0xf104bd00,
 	0xf0870427,
-	0x12d00023,
-	0x0012b700,
-	0x0427f001,
-	0xf40012d0,
-	0xe7f11031,
-	0xe3f09604,
-	0x6821f440,
-	0x8090f1c7,
-	0xf4f00301,
-	0x020f801f,
-	0xbb0117f0,
-	0x12b6041f,
-	0x0c27f101,
-	0x0624b604,
-	0xd00021d0,
-	0x17f14021,
-	0x0e980100,
-	0x010f9800,
-	0x014721f5,
-	0x070037f1,
-	0x950634b6,
-	0x34d00814,
-	0x4034d000,
-	0x130030b7,
-	0xb6001fbb,
-	0x3fd002f5,
-	0x0815b600,
-	0xb60110b6,
-	0x1fb90814,
-	0x7121f502,
-	0x001fbb02,
-	0xf1020398,
-	0xf0200047,
-/* 0x03f6: init_gpc */
-	0x4ea05043,
-	0x1fb90804,
-	0x8d21f402,
-	0x010c4ea0,
-	0x21f4f4bd,
-	0x044ea08d,
-	0x8d21f401,
-	0x01004ea0,
-	0xf402f7f0,
-	0x4ea08d21,
-/* 0x041e: init_gpc_wait */
-	0x21f40800,
-	0x1fffc868,
-	0xa0fa0bf4,
-	0xf408044e,
-	0x1fbb6821,
-	0x0040b700,
-	0x0132b680,
-	0xf1be1bf4,
+	0x07f10023,
+	0x03f00400,
+	0x0002d000,
+	0x27f004bd,
+	0x0007f104,
+	0x0003f003,
+	0xbd0002d0,
+	0x1031f404,
+	0x9604e7f1,
+	0xf440e3f0,
+	0xfeb96821,
+	0x90f1c702,
+	0xf0030180,
+	0x0f801ff4,
+	0x0117f002,
+	0xb6041fbb,
+	0x07f10112,
+	0x03f00300,
+	0x0001d001,
+	0x07f104bd,
+	0x03f00400,
+	0x0001d001,
+	0x17f104bd,
+	0xf7f00100,
+	0x0d21f502,
+	0x1f21f508,
+	0x10f7f008,
+	0x086c21f5,
+	0x98000e98,
+	0x21f5010f,
+	0x14950150,
+	0x0007f108,
+	0x0103f0c0,
+	0xbd0004d0,
+	0x0007f104,
+	0x0103f0c1,
+	0xbd0004d0,
+	0x0030b704,
+	0x001fbb13,
+	0xf102f5b6,
+	0xf0d30007,
+	0x0fd00103,
+	0xb604bd00,
+	0x10b60815,
+	0x0814b601,
+	0xf5021fb9,
+	0xbb02d321,
+	0x0398001f,
+	0x0047f102,
+	0x5043f020,
+/* 0x04f4: init_gpc */
+	0x08044ea0,
+	0xf4021fb9,
+	0x4ea09d21,
+	0xf4bd010c,
+	0xa09d21f4,
+	0xf401044e,
+	0x4ea09d21,
+	0xf7f00100,
+	0x9d21f402,
+	0x08004ea0,
+/* 0x051c: init_gpc_wait */
+	0xc86821f4,
+	0x0bf41fff,
+	0x044ea0fa,
+	0x6821f408,
+	0xb7001fbb,
+	0xb6800040,
+	0x1bf40132,
+	0x00f7f0be,
+	0x086c21f5,
+	0xf500f7f0,
+	0xf1080d21,
 	0xf0010007,
 	0x01d00203,
 	0xbd04bd00,
@@ -519,360 +584,421 @@ uint32_t nvc0_grhub_code[] = {
 	0x080007f1,
 	0xd00203f0,
 	0x04bd0001,
-/* 0x0458: main */
+/* 0x0564: main */
 	0xf40031f4,
 	0xd7f00028,
 	0x3921f410,
 	0xb1f401f4,
 	0xf54001e4,
-	0xbd00de1b,
+	0xbd00e91b,
 	0x0499f094,
 	0x0f0007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0b0017f1,
-	0xcf0614b6,
-	0x11cf4012,
-	0x1f13c800,
-	0x00870bf5,
-	0xf41f23c8,
-	0x20f9620b,
-	0xbd0212b9,
-	0x0799f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0xf40132f4,
-	0x21f50231,
-	0x94bd082f,
+	0xc00017f1,
+	0xcf0213f0,
+	0x27f10011,
+	0x23f0c100,
+	0x0022cf02,
+	0xf51f13c8,
+	0xc800890b,
+	0x0bf41f23,
+	0xb920f962,
+	0x94bd0212,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
-	0xfc04bd00,
-	0xf094bd20,
-	0x07f10699,
-	0x03f00f00,
-	0x0009d002,
-	0x31f404bd,
-	0x2f21f501,
-	0xf094bd08,
-	0x07f10699,
+	0xf404bd00,
+	0x31f40132,
+	0x4021f502,
+	0xf094bd0a,
+	0x07f10799,
 	0x03f01700,
 	0x0009d002,
-	0x0ef404bd,
-/* 0x04f9: chsw_prev_no_next */
-	0xb920f931,
-	0x32f40212,
-	0x0232f401,
-	0x082f21f5,
-	0x17f120fc,
-	0x14b60b00,
-	0x0012d006,
-/* 0x0517: chsw_no_prev */
-	0xc8130ef4,
-	0x0bf41f23,
-	0x0131f40d,
-	0xf50232f4,
-/* 0x0527: chsw_done */
-	0xf1082f21,
-	0xb60b0c17,
-	0x27f00614,
-	0x0012d001,
+	0x20fc04bd,
 	0x99f094bd,
-	0x0007f104,
+	0x0007f106,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0131f404,
+	0x0a4021f5,
+	0x99f094bd,
+	0x0007f106,
 	0x0203f017,
 	0xbd0009d0,
-	0x130ef504,
-/* 0x0549: main_not_ctx_switch */
-	0x01e4b0ff,
-	0xb90d1bf4,
-	0x21f502f2,
-	0x0ef407bb,
-/* 0x0559: main_not_ctx_chan */
-	0x02e4b046,
-	0xbd321bf4,
-	0x0799f094,
-	0x0f0007f1,
+	0x330ef404,
+/* 0x060c: chsw_prev_no_next */
+	0x12b920f9,
+	0x0132f402,
+	0xf50232f4,
+	0xfc0a4021,
+	0x0007f120,
+	0x0203f0c0,
+	0xbd0002d0,
+	0x130ef404,
+/* 0x062c: chsw_no_prev */
+	0xf41f23c8,
+	0x31f40d0b,
+	0x0232f401,
+	0x0a4021f5,
+/* 0x063c: chsw_done */
+	0xf10127f0,
+	0xf0c30007,
+	0x02d00203,
+	0xbd04bd00,
+	0x0499f094,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0xf40132f4,
-	0x21f50232,
-	0x94bd082f,
+	0xff080ef5,
+/* 0x0660: main_not_ctx_switch */
+	0xf401e4b0,
+	0xf2b90d1b,
+	0xd021f502,
+	0x460ef409,
+/* 0x0670: main_not_ctx_chan */
+	0xf402e4b0,
+	0x94bd321b,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
 	0xf404bd00,
-/* 0x058e: main_not_ctx_save */
-	0xef94110e,
-	0x01f5f010,
-	0x02fe21f5,
-	0xfec00ef5,
-/* 0x059c: main_done */
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f008,
-	0xbd0002d0,
-	0xab0ef504,
-/* 0x05b1: ih */
-	0xfe80f9fe,
-	0x80f90188,
-	0xa0f990f9,
-	0xd0f9b0f9,
-	0xf0f9e0f9,
-	0x0acf04bd,
-	0x04abc480,
-	0xf11d0bf4,
-	0xf01900b7,
-	0xbecf10d7,
-	0x00bfcf40,
+	0x32f40132,
+	0x4021f502,
+	0xf094bd0a,
+	0x07f10799,
+	0x03f01700,
+	0x0009d002,
+	0x0ef404bd,
+/* 0x06a5: main_not_ctx_save */
+	0x10ef9411,
+	0xf501f5f0,
+	0xf5037e21,
+/* 0x06b3: main_done */
+	0xbdfeb50e,
+	0x1f29f024,
+	0x080007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xfea00ef5,
+/* 0x06c8: ih */
+	0x88fe80f9,
+	0xf980f901,
+	0xf9a0f990,
+	0xf9d0f9b0,
+	0xbdf0f9e0,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x10d7f030,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
 	0xb70421f4,
 	0xf00400b0,
-	0xbed001e7,
-/* 0x05e9: ih_no_fifo */
-	0x00abe400,
-	0x0d0bf401,
-	0xf110d7f0,
-	0xf44001e7,
-/* 0x05fa: ih_no_ctxsw */
-	0xb7f10421,
-	0xb0bd0104,
-	0xf4b4abff,
-	0xa7f10d0b,
-	0xa4b60c1c,
-	0x00abd006,
-/* 0x0610: ih_no_other */
-	0xfc400ad0,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x071a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x10d7f00d,
+	0x4001e7f1,
+/* 0x072b: ih_no_ctxsw */
+	0xe40421f4,
+	0xf40400ab,
+	0xe7f16c0b,
+	0xe3f00708,
+	0x6821f440,
+	0xf102ffb9,
+	0xf0040007,
+	0x0fd00203,
+	0xf104bd00,
+	0xf00704e7,
+	0x21f440e3,
+	0x02ffb968,
+	0x030007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0x9450fec7,
+	0xf7f102ee,
+	0xf3f00700,
+	0x00efbb40,
+	0xf16821f4,
+	0xf0020007,
+	0x0fd00203,
+	0xf004bd00,
+	0x21f503f7,
+	0xb7f1037e,
+	0xbfb90100,
+	0x44e7f102,
+	0x40e3f001,
+/* 0x079b: ih_no_fwmthd */
+	0xf19d21f4,
+	0xbd0504b7,
+	0xb4abffb0,
+	0xf10f0bf4,
+	0xf0070007,
+	0x0bd00303,
+/* 0x07b3: ih_no_other */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
 	0xfce0fcf0,
 	0xfcb0fcd0,
 	0xfc90fca0,
 	0x0088fe80,
 	0x32f480fc,
-/* 0x062b: ctx_4160s */
-	0xf101f800,
-	0xf04160e7,
-	0xf7f040e3,
-	0x8d21f401,
-/* 0x0638: ctx_4160s_wait */
-	0xc86821f4,
-	0x0bf404ff,
-/* 0x0643: ctx_4160c */
-	0xf100f8fa,
+/* 0x07d7: ctx_4160s */
+	0xf001f800,
+	0xffb901f7,
+	0x60e7f102,
+	0x40e3f041,
+/* 0x07e7: ctx_4160s_wait */
+	0xf19d21f4,
 	0xf04160e7,
-	0xf4bd40e3,
-	0xf88d21f4,
-/* 0x0651: ctx_4170s */
-	0x70e7f100,
+	0x21f440e3,
+	0x02ffb968,
+	0xf404ffc8,
+	0x00f8f00b,
+/* 0x07fc: ctx_4160c */
+	0xffb9f4bd,
+	0x60e7f102,
 	0x40e3f041,
-	0xf410f5f0,
-	0x00f88d21,
-/* 0x0660: ctx_4170w */
-	0x4170e7f1,
-	0xf440e3f0,
-	0xf4f06821,
-	0xf31bf410,
-/* 0x0672: ctx_redswitch */
-	0xe7f100f8,
-	0xe4b60614,
-	0x70f7f106,
-	0x00efd002,
-/* 0x0683: ctx_redswitch_delay */
-	0xb608f7f0,
-	0x1bf401f2,
-	0x70f7f1fd,
-	0x00efd007,
-/* 0x0692: ctx_86c */
-	0xe7f100f8,
-	0xe4b6086c,
-	0x00efd006,
-	0x8a14e7f1,
-	0xf440e3f0,
-	0xe7f18d21,
-	0xe3f0a86c,
-	0x8d21f441,
-/* 0x06b2: ctx_load */
+	0xf89d21f4,
+/* 0x080d: ctx_4170s */
+	0x10f5f000,
+	0xf102ffb9,
+	0xf04170e7,
+	0x21f440e3,
+/* 0x081f: ctx_4170w */
+	0xf100f89d,
+	0xf04170e7,
+	0x21f440e3,
+	0x02ffb968,
+	0xf410f4f0,
+	0x00f8f01b,
+/* 0x0834: ctx_redswitch */
+	0x0200e7f1,
+	0xf040e5f0,
+	0xe5f020e5,
+	0x0007f110,
+	0x0103f085,
+	0xbd000ed0,
+	0x08f7f004,
+/* 0x0850: ctx_redswitch_delay */
+	0xf401f2b6,
+	0xe5f1fd1b,
+	0xe5f10400,
+	0x07f10100,
+	0x03f08500,
+	0x000ed001,
+	0x00f804bd,
+/* 0x086c: ctx_86c */
+	0x1b0007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0xf102ffb9,
+	0xf08a14e7,
+	0x21f440e3,
+	0x02ffb99d,
+	0xa86ce7f1,
+	0xf441e3f0,
+	0x00f89d21,
+/* 0x0894: ctx_mem */
+	0x840007f1,
+	0xd00203f0,
+	0x04bd000f,
+/* 0x08a0: ctx_mem_wait */
+	0x8400f7f1,
+	0xcf02f3f0,
+	0xfffd00ff,
+	0xf31bf405,
+/* 0x08b2: ctx_load */
 	0x94bd00f8,
 	0xf10599f0,
 	0xf00f0007,
 	0x09d00203,
 	0xf004bd00,
 	0x21f40ca7,
-	0x2417f1c9,
-	0x0614b60a,
-	0xf10010d0,
-	0xb60b0037,
-	0x32d00634,
-	0x0c17f140,
-	0x0614b60a,
-	0xd00747f0,
-	0x14d00012,
-/* 0x06ed: ctx_chan_wait_0 */
-	0x4014cf40,
-	0xf41f44f0,
-	0x32d0fa1b,
-	0x000bfe00,
-	0xb61f2af0,
-	0x20b60424,
-	0xf094bd02,
+	0xf1f4bdd0,
+	0xf0890007,
+	0x0fd00203,
+	0xf104bd00,
+	0xf0c10007,
+	0x02d00203,
+	0xf104bd00,
+	0xf0830007,
+	0x02d00203,
+	0xf004bd00,
+	0x21f507f7,
+	0x07f10894,
+	0x03f0c000,
+	0x0002d002,
+	0x0bfe04bd,
+	0x1f2af000,
+	0xb60424b6,
+	0x94bd0220,
+	0xf10899f0,
+	0xf00f0007,
+	0x09d00203,
+	0xf104bd00,
+	0xf0810007,
+	0x02d00203,
+	0xf104bd00,
+	0xf1000027,
+	0xf0800023,
+	0x07f10225,
+	0x03f08800,
+	0x0002d002,
+	0x17f004bd,
+	0x0027f110,
+	0x0223f002,
+	0xf80512fa,
+	0xf094bd03,
 	0x07f10899,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x17f104bd,
-	0x14b60a04,
-	0x0012d006,
-	0x0a2017f1,
-	0xf00614b6,
-	0x23f10227,
-	0x12d08000,
-	0x1017f000,
-	0x020027f1,
-	0xfa0223f0,
-	0x03f80512,
+	0x019804bd,
+	0x1814b681,
+	0xb6800298,
+	0x12fd0825,
+	0x16018005,
 	0x99f094bd,
-	0x0007f108,
-	0x0203f017,
+	0x0007f109,
+	0x0203f00f,
 	0xbd0009d0,
-	0x81019804,
-	0x981814b6,
-	0x25b68002,
-	0x0512fd08,
-	0xbd160180,
-	0x0999f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0x0a0427f1,
-	0xd00624b6,
-	0x27f00021,
-	0x2017f101,
-	0x0614b60a,
-	0xf10012d0,
-	0xf0010017,
-	0x01fa0613,
-	0xbd03f805,
-	0x0999f094,
-	0x170007f1,
+	0x0007f104,
+	0x0203f081,
+	0xbd0001d0,
+	0x0127f004,
+	0x880007f1,
 	0xd00203f0,
-	0x04bd0009,
+	0x04bd0002,
+	0x010017f1,
+	0xfa0613f0,
+	0x03f80501,
 	0x99f094bd,
-	0x0007f105,
+	0x0007f109,
 	0x0203f017,
 	0xbd0009d0,
-/* 0x07bb: ctx_chan */
-	0xf500f804,
-	0xf5062b21,
-	0xf006b221,
-	0x21f40ca7,
-	0x1017f1c9,
-	0x0614b60a,
-	0xd00527f0,
-/* 0x07d6: ctx_chan_wait */
-	0x12cf0012,
-	0x0522fd00,
-	0xf5fa1bf4,
-	0xf8064321,
-/* 0x07e5: ctx_mmio_exec */
-	0x41039800,
-	0x0a0427f1,
-	0xd00624b6,
-	0x34bd0023,
-/* 0x07f4: ctx_mmio_loop */
+	0xf094bd04,
+	0x07f10599,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x09d0: ctx_chan */
+	0x07d721f5,
+	0x08b221f5,
+	0xf40ca7f0,
+	0xf7f0d021,
+	0x9421f505,
+	0xfc21f508,
+/* 0x09eb: ctx_mmio_exec */
+	0x9800f807,
+	0x07f14103,
+	0x03f08100,
+	0x0003d002,
+	0x34bd04bd,
+/* 0x09fc: ctx_mmio_loop */
 	0xf4ff34c4,
 	0x57f10f1b,
 	0x53f00200,
 	0x0535fa06,
-/* 0x0806: ctx_mmio_pull */
+/* 0x0a0e: ctx_mmio_pull */
 	0x4e9803f8,
 	0x814f9880,
-	0xb68d21f4,
+	0xb69d21f4,
 	0x12b60830,
 	0xdf1bf401,
-/* 0x0818: ctx_mmio_done */
-	0xd0160398,
-	0x00800023,
-	0x0017f140,
-	0x0613f001,
-	0xf80601fa,
-/* 0x082f: ctx_xfer */
-	0xf100f803,
-	0xb60c00f7,
-	0xe7f006f4,
-	0x80fed004,
-/* 0x083c: ctx_xfer_idle */
-	0xf100fecf,
-	0xf42000e4,
-	0x11f4f91b,
-	0x1102f406,
-/* 0x084c: ctx_xfer_pre */
-	0xf510f7f0,
-	0xf5069221,
-	0xf4062b21,
-/* 0x085a: ctx_xfer_pre_load */
-	0xf7f01c11,
-	0x5121f502,
-	0x6021f506,
-	0x7221f506,
-	0xf5f4bd06,
-	0xf5065121,
-/* 0x0873: ctx_xfer_exec */
-	0x9806b221,
-	0x27f11601,
-	0x24b60414,
-	0x0020d006,
-	0xa500e7f1,
-	0xb941e3f0,
-	0x21f4021f,
-	0x04e0b68d,
-	0xf001fcf0,
-	0x24b6022c,
-	0x05f2fd01,
-	0xf18d21f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00c,
-	0x021521f5,
-	0x47fc27f1,
-	0xd00223f0,
-	0x2cf00020,
+/* 0x0a20: ctx_mmio_done */
+	0xf1160398,
+	0xf0810007,
+	0x03d00203,
+	0x8004bd00,
+	0x17f14000,
+	0x13f00100,
+	0x0601fa06,
+	0x00f803f8,
+/* 0x0a40: ctx_xfer */
+	0xf104e7f0,
+	0xf0020007,
+	0x0ed00303,
+/* 0x0a4f: ctx_xfer_idle */
+	0xf104bd00,
+	0xf00000e7,
+	0xeecf03e3,
+	0x00e4f100,
+	0xf21bf420,
+	0xf40611f4,
+/* 0x0a66: ctx_xfer_pre */
+	0xf7f01102,
+	0x6c21f510,
+	0xd721f508,
+	0x1c11f407,
+/* 0x0a74: ctx_xfer_pre_load */
+	0xf502f7f0,
+	0xf5080d21,
+	0xf5081f21,
+	0xbd083421,
+	0x0d21f5f4,
+	0xb221f508,
+/* 0x0a8d: ctx_xfer_exec */
+	0x16019808,
+	0x07f124bd,
+	0x03f00500,
+	0x0002d001,
+	0x1fb904bd,
+	0x00e7f102,
+	0x41e3f0a5,
+	0xf09d21f4,
+	0x2cf001fc,
+	0x0124b602,
+	0xb905f2fd,
+	0xe7f102ff,
+	0xe3f0a504,
+	0x9d21f441,
+	0x026a21f5,
+	0x07f124bd,
+	0x03f047fc,
+	0x0002d002,
+	0x2cf004bd,
 	0x0320b601,
-	0xf00012d0,
-	0xa5f001ac,
-	0x00b7f006,
-	0x98000c98,
-	0xe7f0010d,
-	0x6621f500,
-	0x08a7f001,
-	0x010921f5,
-	0x021521f5,
-	0xf02201f4,
-	0x21f40ca7,
-	0x1017f1c9,
-	0x0614b60a,
-	0xd00527f0,
-/* 0x08fa: ctx_xfer_post_save_wait */
-	0x12cf0012,
-	0x0522fd00,
-	0xf4fa1bf4,
-/* 0x0906: ctx_xfer_post */
-	0xf7f03202,
-	0x5121f502,
-	0xf5f4bd06,
-	0xf5069221,
-	0xf5023421,
-	0xbd066021,
-	0x5121f5f4,
-	0x1011f406,
-	0xfd400198,
-	0x0bf40511,
-	0xe521f507,
-/* 0x0931: ctx_xfer_no_post_mmio */
-	0x4321f507,
-/* 0x0935: ctx_xfer_done */
-	0x0000f806,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf001acf0,
+	0xb7f006a5,
+	0x000c9800,
+	0xf0010d98,
+	0x21f500e7,
+	0xa7f0016f,
+	0x1021f508,
+	0x5e21f501,
+	0x1301f402,
+	0xf40ca7f0,
+	0xf7f0d021,
+	0x9421f505,
+	0x3202f408,
+/* 0x0b1c: ctx_xfer_post */
+	0xf502f7f0,
+	0xbd080d21,
+	0x6c21f5f4,
+	0x7f21f508,
+	0x1f21f502,
+	0xf5f4bd08,
+	0xf4080d21,
+	0x01981011,
+	0x0511fd40,
+	0xf5070bf4,
+/* 0x0b47: ctx_xfer_no_post_mmio */
+	0xf509eb21,
+/* 0x0b4b: ctx_xfer_done */
+	0xf807fc21,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvd7.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvd7.fuc.h
index a1b9f763996..62b0c7601d8 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvd7.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvd7.fuc.h
@@ -206,14 +206,14 @@ uint32_t nvd7_grhub_data[] = {
 };
 
 uint32_t nvd7_grhub_code[] = {
-	0x031b0ef5,
+	0x039b0ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -237,184 +237,214 @@ uint32_t nvd7_grhub_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f405,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -428,7 +458,7 @@ uint32_t nvd7_grhub_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0x07f100f8,
 	0x03f00500,
 	0x000fd002,
@@ -436,82 +466,117 @@ uint32_t nvd7_grhub_code[] = {
 	0x0007f101,
 	0x0303f007,
 	0xbd000fd0,
-/* 0x031b: init */
+/* 0x039b: init */
 	0xbd00f804,
-	0x0004fe04,
-	0xf10007fe,
-	0xf0120017,
-	0x12d00227,
-	0xb117f100,
-	0x0010fe05,
-	0x040017f1,
-	0xf1c010d0,
-	0xb6040437,
-	0x27f10634,
-	0x32d02003,
-	0x0427f100,
-	0x0132d020,
+	0x0007fe04,
+	0x420017f1,
+	0xcf0013f0,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf10227f0,
+	0xf0120007,
+	0x02d00003,
+	0xf104bd00,
+	0xfe06c817,
+	0x24bd0010,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0x200327f1,
+	0x010007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200427f1,
+	0x010407f1,
+	0xd00103f0,
+	0x04bd0002,
 	0x200b27f1,
-	0xf10232d0,
-	0xd0200c27,
-	0x27f10732,
-	0x24b60c24,
-	0x0003b906,
-	0xf10023d0,
+	0x010807f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200c27f1,
+	0x011c07f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1010392,
+	0xf0090007,
+	0x03d00303,
+	0xf104bd00,
 	0xf0870427,
-	0x12d00023,
-	0x0012b700,
-	0x0427f001,
-	0xf40012d0,
-	0xe7f11031,
-	0xe3f09604,
-	0x6821f440,
-	0x8090f1c7,
-	0xf4f00301,
-	0x020f801f,
-	0xbb0117f0,
-	0x12b6041f,
-	0x0c27f101,
-	0x0624b604,
-	0xd00021d0,
-	0x17f14021,
-	0x0e980100,
-	0x010f9800,
-	0x014721f5,
-	0x070037f1,
-	0x950634b6,
-	0x34d00814,
-	0x4034d000,
-	0x130030b7,
-	0xb6001fbb,
-	0x3fd002f5,
-	0x0815b600,
-	0xb60110b6,
-	0x1fb90814,
-	0x7121f502,
-	0x001fbb02,
-	0xf1020398,
-	0xf0200047,
-/* 0x03f6: init_gpc */
-	0x4ea05043,
-	0x1fb90804,
-	0x8d21f402,
-	0x010c4ea0,
-	0x21f4f4bd,
-	0x044ea08d,
-	0x8d21f401,
-	0x01004ea0,
-	0xf402f7f0,
-	0x4ea08d21,
-/* 0x041e: init_gpc_wait */
-	0x21f40800,
-	0x1fffc868,
-	0xa0fa0bf4,
-	0xf408044e,
-	0x1fbb6821,
-	0x0040b700,
-	0x0132b680,
-	0xf1be1bf4,
+	0x07f10023,
+	0x03f00400,
+	0x0002d000,
+	0x27f004bd,
+	0x0007f104,
+	0x0003f003,
+	0xbd0002d0,
+	0x1031f404,
+	0x9604e7f1,
+	0xf440e3f0,
+	0xfeb96821,
+	0x90f1c702,
+	0xf0030180,
+	0x0f801ff4,
+	0x0117f002,
+	0xb6041fbb,
+	0x07f10112,
+	0x03f00300,
+	0x0001d001,
+	0x07f104bd,
+	0x03f00400,
+	0x0001d001,
+	0x17f104bd,
+	0xf7f00100,
+	0x0d21f502,
+	0x1f21f508,
+	0x10f7f008,
+	0x086c21f5,
+	0x98000e98,
+	0x21f5010f,
+	0x14950150,
+	0x0007f108,
+	0x0103f0c0,
+	0xbd0004d0,
+	0x0007f104,
+	0x0103f0c1,
+	0xbd0004d0,
+	0x0030b704,
+	0x001fbb13,
+	0xf102f5b6,
+	0xf0d30007,
+	0x0fd00103,
+	0xb604bd00,
+	0x10b60815,
+	0x0814b601,
+	0xf5021fb9,
+	0xbb02d321,
+	0x0398001f,
+	0x0047f102,
+	0x5043f020,
+/* 0x04f4: init_gpc */
+	0x08044ea0,
+	0xf4021fb9,
+	0x4ea09d21,
+	0xf4bd010c,
+	0xa09d21f4,
+	0xf401044e,
+	0x4ea09d21,
+	0xf7f00100,
+	0x9d21f402,
+	0x08004ea0,
+/* 0x051c: init_gpc_wait */
+	0xc86821f4,
+	0x0bf41fff,
+	0x044ea0fa,
+	0x6821f408,
+	0xb7001fbb,
+	0xb6800040,
+	0x1bf40132,
+	0x00f7f0be,
+	0x086c21f5,
+	0xf500f7f0,
+	0xf1080d21,
 	0xf0010007,
 	0x01d00203,
 	0xbd04bd00,
@@ -519,360 +584,421 @@ uint32_t nvd7_grhub_code[] = {
 	0x080007f1,
 	0xd00203f0,
 	0x04bd0001,
-/* 0x0458: main */
+/* 0x0564: main */
 	0xf40031f4,
 	0xd7f00028,
 	0x3921f410,
 	0xb1f401f4,
 	0xf54001e4,
-	0xbd00de1b,
+	0xbd00e91b,
 	0x0499f094,
 	0x0f0007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0b0017f1,
-	0xcf0614b6,
-	0x11cf4012,
-	0x1f13c800,
-	0x00870bf5,
-	0xf41f23c8,
-	0x20f9620b,
-	0xbd0212b9,
-	0x0799f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0xf40132f4,
-	0x21f50231,
-	0x94bd082f,
+	0xc00017f1,
+	0xcf0213f0,
+	0x27f10011,
+	0x23f0c100,
+	0x0022cf02,
+	0xf51f13c8,
+	0xc800890b,
+	0x0bf41f23,
+	0xb920f962,
+	0x94bd0212,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
-	0xfc04bd00,
-	0xf094bd20,
-	0x07f10699,
-	0x03f00f00,
-	0x0009d002,
-	0x31f404bd,
-	0x2f21f501,
-	0xf094bd08,
-	0x07f10699,
+	0xf404bd00,
+	0x31f40132,
+	0x4021f502,
+	0xf094bd0a,
+	0x07f10799,
 	0x03f01700,
 	0x0009d002,
-	0x0ef404bd,
-/* 0x04f9: chsw_prev_no_next */
-	0xb920f931,
-	0x32f40212,
-	0x0232f401,
-	0x082f21f5,
-	0x17f120fc,
-	0x14b60b00,
-	0x0012d006,
-/* 0x0517: chsw_no_prev */
-	0xc8130ef4,
-	0x0bf41f23,
-	0x0131f40d,
-	0xf50232f4,
-/* 0x0527: chsw_done */
-	0xf1082f21,
-	0xb60b0c17,
-	0x27f00614,
-	0x0012d001,
+	0x20fc04bd,
 	0x99f094bd,
-	0x0007f104,
+	0x0007f106,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0131f404,
+	0x0a4021f5,
+	0x99f094bd,
+	0x0007f106,
 	0x0203f017,
 	0xbd0009d0,
-	0x130ef504,
-/* 0x0549: main_not_ctx_switch */
-	0x01e4b0ff,
-	0xb90d1bf4,
-	0x21f502f2,
-	0x0ef407bb,
-/* 0x0559: main_not_ctx_chan */
-	0x02e4b046,
-	0xbd321bf4,
-	0x0799f094,
-	0x0f0007f1,
+	0x330ef404,
+/* 0x060c: chsw_prev_no_next */
+	0x12b920f9,
+	0x0132f402,
+	0xf50232f4,
+	0xfc0a4021,
+	0x0007f120,
+	0x0203f0c0,
+	0xbd0002d0,
+	0x130ef404,
+/* 0x062c: chsw_no_prev */
+	0xf41f23c8,
+	0x31f40d0b,
+	0x0232f401,
+	0x0a4021f5,
+/* 0x063c: chsw_done */
+	0xf10127f0,
+	0xf0c30007,
+	0x02d00203,
+	0xbd04bd00,
+	0x0499f094,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0xf40132f4,
-	0x21f50232,
-	0x94bd082f,
+	0xff080ef5,
+/* 0x0660: main_not_ctx_switch */
+	0xf401e4b0,
+	0xf2b90d1b,
+	0xd021f502,
+	0x460ef409,
+/* 0x0670: main_not_ctx_chan */
+	0xf402e4b0,
+	0x94bd321b,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
 	0xf404bd00,
-/* 0x058e: main_not_ctx_save */
-	0xef94110e,
-	0x01f5f010,
-	0x02fe21f5,
-	0xfec00ef5,
-/* 0x059c: main_done */
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f008,
-	0xbd0002d0,
-	0xab0ef504,
-/* 0x05b1: ih */
-	0xfe80f9fe,
-	0x80f90188,
-	0xa0f990f9,
-	0xd0f9b0f9,
-	0xf0f9e0f9,
-	0x0acf04bd,
-	0x04abc480,
-	0xf11d0bf4,
-	0xf01900b7,
-	0xbecf10d7,
-	0x00bfcf40,
+	0x32f40132,
+	0x4021f502,
+	0xf094bd0a,
+	0x07f10799,
+	0x03f01700,
+	0x0009d002,
+	0x0ef404bd,
+/* 0x06a5: main_not_ctx_save */
+	0x10ef9411,
+	0xf501f5f0,
+	0xf5037e21,
+/* 0x06b3: main_done */
+	0xbdfeb50e,
+	0x1f29f024,
+	0x080007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xfea00ef5,
+/* 0x06c8: ih */
+	0x88fe80f9,
+	0xf980f901,
+	0xf9a0f990,
+	0xf9d0f9b0,
+	0xbdf0f9e0,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x10d7f030,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
 	0xb70421f4,
 	0xf00400b0,
-	0xbed001e7,
-/* 0x05e9: ih_no_fifo */
-	0x00abe400,
-	0x0d0bf401,
-	0xf110d7f0,
-	0xf44001e7,
-/* 0x05fa: ih_no_ctxsw */
-	0xb7f10421,
-	0xb0bd0104,
-	0xf4b4abff,
-	0xa7f10d0b,
-	0xa4b60c1c,
-	0x00abd006,
-/* 0x0610: ih_no_other */
-	0xfc400ad0,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x071a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x10d7f00d,
+	0x4001e7f1,
+/* 0x072b: ih_no_ctxsw */
+	0xe40421f4,
+	0xf40400ab,
+	0xe7f16c0b,
+	0xe3f00708,
+	0x6821f440,
+	0xf102ffb9,
+	0xf0040007,
+	0x0fd00203,
+	0xf104bd00,
+	0xf00704e7,
+	0x21f440e3,
+	0x02ffb968,
+	0x030007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0x9450fec7,
+	0xf7f102ee,
+	0xf3f00700,
+	0x00efbb40,
+	0xf16821f4,
+	0xf0020007,
+	0x0fd00203,
+	0xf004bd00,
+	0x21f503f7,
+	0xb7f1037e,
+	0xbfb90100,
+	0x44e7f102,
+	0x40e3f001,
+/* 0x079b: ih_no_fwmthd */
+	0xf19d21f4,
+	0xbd0504b7,
+	0xb4abffb0,
+	0xf10f0bf4,
+	0xf0070007,
+	0x0bd00303,
+/* 0x07b3: ih_no_other */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
 	0xfce0fcf0,
 	0xfcb0fcd0,
 	0xfc90fca0,
 	0x0088fe80,
 	0x32f480fc,
-/* 0x062b: ctx_4160s */
-	0xf101f800,
-	0xf04160e7,
-	0xf7f040e3,
-	0x8d21f401,
-/* 0x0638: ctx_4160s_wait */
-	0xc86821f4,
-	0x0bf404ff,
-/* 0x0643: ctx_4160c */
-	0xf100f8fa,
+/* 0x07d7: ctx_4160s */
+	0xf001f800,
+	0xffb901f7,
+	0x60e7f102,
+	0x40e3f041,
+/* 0x07e7: ctx_4160s_wait */
+	0xf19d21f4,
 	0xf04160e7,
-	0xf4bd40e3,
-	0xf88d21f4,
-/* 0x0651: ctx_4170s */
-	0x70e7f100,
+	0x21f440e3,
+	0x02ffb968,
+	0xf404ffc8,
+	0x00f8f00b,
+/* 0x07fc: ctx_4160c */
+	0xffb9f4bd,
+	0x60e7f102,
 	0x40e3f041,
-	0xf410f5f0,
-	0x00f88d21,
-/* 0x0660: ctx_4170w */
-	0x4170e7f1,
-	0xf440e3f0,
-	0xf4f06821,
-	0xf31bf410,
-/* 0x0672: ctx_redswitch */
-	0xe7f100f8,
-	0xe4b60614,
-	0x70f7f106,
-	0x00efd002,
-/* 0x0683: ctx_redswitch_delay */
-	0xb608f7f0,
-	0x1bf401f2,
-	0x70f7f1fd,
-	0x00efd007,
-/* 0x0692: ctx_86c */
-	0xe7f100f8,
-	0xe4b6086c,
-	0x00efd006,
-	0x8a14e7f1,
-	0xf440e3f0,
-	0xe7f18d21,
-	0xe3f0a86c,
-	0x8d21f441,
-/* 0x06b2: ctx_load */
+	0xf89d21f4,
+/* 0x080d: ctx_4170s */
+	0x10f5f000,
+	0xf102ffb9,
+	0xf04170e7,
+	0x21f440e3,
+/* 0x081f: ctx_4170w */
+	0xf100f89d,
+	0xf04170e7,
+	0x21f440e3,
+	0x02ffb968,
+	0xf410f4f0,
+	0x00f8f01b,
+/* 0x0834: ctx_redswitch */
+	0x0200e7f1,
+	0xf040e5f0,
+	0xe5f020e5,
+	0x0007f110,
+	0x0103f085,
+	0xbd000ed0,
+	0x08f7f004,
+/* 0x0850: ctx_redswitch_delay */
+	0xf401f2b6,
+	0xe5f1fd1b,
+	0xe5f10400,
+	0x07f10100,
+	0x03f08500,
+	0x000ed001,
+	0x00f804bd,
+/* 0x086c: ctx_86c */
+	0x1b0007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0xf102ffb9,
+	0xf08a14e7,
+	0x21f440e3,
+	0x02ffb99d,
+	0xa86ce7f1,
+	0xf441e3f0,
+	0x00f89d21,
+/* 0x0894: ctx_mem */
+	0x840007f1,
+	0xd00203f0,
+	0x04bd000f,
+/* 0x08a0: ctx_mem_wait */
+	0x8400f7f1,
+	0xcf02f3f0,
+	0xfffd00ff,
+	0xf31bf405,
+/* 0x08b2: ctx_load */
 	0x94bd00f8,
 	0xf10599f0,
 	0xf00f0007,
 	0x09d00203,
 	0xf004bd00,
 	0x21f40ca7,
-	0x2417f1c9,
-	0x0614b60a,
-	0xf10010d0,
-	0xb60b0037,
-	0x32d00634,
-	0x0c17f140,
-	0x0614b60a,
-	0xd00747f0,
-	0x14d00012,
-/* 0x06ed: ctx_chan_wait_0 */
-	0x4014cf40,
-	0xf41f44f0,
-	0x32d0fa1b,
-	0x000bfe00,
-	0xb61f2af0,
-	0x20b60424,
-	0xf094bd02,
+	0xf1f4bdd0,
+	0xf0890007,
+	0x0fd00203,
+	0xf104bd00,
+	0xf0c10007,
+	0x02d00203,
+	0xf104bd00,
+	0xf0830007,
+	0x02d00203,
+	0xf004bd00,
+	0x21f507f7,
+	0x07f10894,
+	0x03f0c000,
+	0x0002d002,
+	0x0bfe04bd,
+	0x1f2af000,
+	0xb60424b6,
+	0x94bd0220,
+	0xf10899f0,
+	0xf00f0007,
+	0x09d00203,
+	0xf104bd00,
+	0xf0810007,
+	0x02d00203,
+	0xf104bd00,
+	0xf1000027,
+	0xf0800023,
+	0x07f10225,
+	0x03f08800,
+	0x0002d002,
+	0x17f004bd,
+	0x0027f110,
+	0x0223f002,
+	0xf80512fa,
+	0xf094bd03,
 	0x07f10899,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x17f104bd,
-	0x14b60a04,
-	0x0012d006,
-	0x0a2017f1,
-	0xf00614b6,
-	0x23f10227,
-	0x12d08000,
-	0x1017f000,
-	0x020027f1,
-	0xfa0223f0,
-	0x03f80512,
+	0x019804bd,
+	0x1814b681,
+	0xb6800298,
+	0x12fd0825,
+	0x16018005,
 	0x99f094bd,
-	0x0007f108,
-	0x0203f017,
+	0x0007f109,
+	0x0203f00f,
 	0xbd0009d0,
-	0x81019804,
-	0x981814b6,
-	0x25b68002,
-	0x0512fd08,
-	0xbd160180,
-	0x0999f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0x0a0427f1,
-	0xd00624b6,
-	0x27f00021,
-	0x2017f101,
-	0x0614b60a,
-	0xf10012d0,
-	0xf0010017,
-	0x01fa0613,
-	0xbd03f805,
-	0x0999f094,
-	0x170007f1,
+	0x0007f104,
+	0x0203f081,
+	0xbd0001d0,
+	0x0127f004,
+	0x880007f1,
 	0xd00203f0,
-	0x04bd0009,
+	0x04bd0002,
+	0x010017f1,
+	0xfa0613f0,
+	0x03f80501,
 	0x99f094bd,
-	0x0007f105,
+	0x0007f109,
 	0x0203f017,
 	0xbd0009d0,
-/* 0x07bb: ctx_chan */
-	0xf500f804,
-	0xf5062b21,
-	0xf006b221,
-	0x21f40ca7,
-	0x1017f1c9,
-	0x0614b60a,
-	0xd00527f0,
-/* 0x07d6: ctx_chan_wait */
-	0x12cf0012,
-	0x0522fd00,
-	0xf5fa1bf4,
-	0xf8064321,
-/* 0x07e5: ctx_mmio_exec */
-	0x41039800,
-	0x0a0427f1,
-	0xd00624b6,
-	0x34bd0023,
-/* 0x07f4: ctx_mmio_loop */
+	0xf094bd04,
+	0x07f10599,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x09d0: ctx_chan */
+	0x07d721f5,
+	0x08b221f5,
+	0xf40ca7f0,
+	0xf7f0d021,
+	0x9421f505,
+	0xfc21f508,
+/* 0x09eb: ctx_mmio_exec */
+	0x9800f807,
+	0x07f14103,
+	0x03f08100,
+	0x0003d002,
+	0x34bd04bd,
+/* 0x09fc: ctx_mmio_loop */
 	0xf4ff34c4,
 	0x57f10f1b,
 	0x53f00200,
 	0x0535fa06,
-/* 0x0806: ctx_mmio_pull */
+/* 0x0a0e: ctx_mmio_pull */
 	0x4e9803f8,
 	0x814f9880,
-	0xb68d21f4,
+	0xb69d21f4,
 	0x12b60830,
 	0xdf1bf401,
-/* 0x0818: ctx_mmio_done */
-	0xd0160398,
-	0x00800023,
-	0x0017f140,
-	0x0613f001,
-	0xf80601fa,
-/* 0x082f: ctx_xfer */
-	0xf100f803,
-	0xb60c00f7,
-	0xe7f006f4,
-	0x80fed004,
-/* 0x083c: ctx_xfer_idle */
-	0xf100fecf,
-	0xf42000e4,
-	0x11f4f91b,
-	0x1102f406,
-/* 0x084c: ctx_xfer_pre */
-	0xf510f7f0,
-	0xf5069221,
-	0xf4062b21,
-/* 0x085a: ctx_xfer_pre_load */
-	0xf7f01c11,
-	0x5121f502,
-	0x6021f506,
-	0x7221f506,
-	0xf5f4bd06,
-	0xf5065121,
-/* 0x0873: ctx_xfer_exec */
-	0x9806b221,
-	0x27f11601,
-	0x24b60414,
-	0x0020d006,
-	0xa500e7f1,
-	0xb941e3f0,
-	0x21f4021f,
-	0x04e0b68d,
-	0xf001fcf0,
-	0x24b6022c,
-	0x05f2fd01,
-	0xf18d21f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00c,
-	0x021521f5,
-	0x47fc27f1,
-	0xd00223f0,
-	0x2cf00020,
+/* 0x0a20: ctx_mmio_done */
+	0xf1160398,
+	0xf0810007,
+	0x03d00203,
+	0x8004bd00,
+	0x17f14000,
+	0x13f00100,
+	0x0601fa06,
+	0x00f803f8,
+/* 0x0a40: ctx_xfer */
+	0xf104e7f0,
+	0xf0020007,
+	0x0ed00303,
+/* 0x0a4f: ctx_xfer_idle */
+	0xf104bd00,
+	0xf00000e7,
+	0xeecf03e3,
+	0x00e4f100,
+	0xf21bf420,
+	0xf40611f4,
+/* 0x0a66: ctx_xfer_pre */
+	0xf7f01102,
+	0x6c21f510,
+	0xd721f508,
+	0x1c11f407,
+/* 0x0a74: ctx_xfer_pre_load */
+	0xf502f7f0,
+	0xf5080d21,
+	0xf5081f21,
+	0xbd083421,
+	0x0d21f5f4,
+	0xb221f508,
+/* 0x0a8d: ctx_xfer_exec */
+	0x16019808,
+	0x07f124bd,
+	0x03f00500,
+	0x0002d001,
+	0x1fb904bd,
+	0x00e7f102,
+	0x41e3f0a5,
+	0xf09d21f4,
+	0x2cf001fc,
+	0x0124b602,
+	0xb905f2fd,
+	0xe7f102ff,
+	0xe3f0a504,
+	0x9d21f441,
+	0x026a21f5,
+	0x07f124bd,
+	0x03f047fc,
+	0x0002d002,
+	0x2cf004bd,
 	0x0320b601,
-	0xf00012d0,
-	0xa5f001ac,
-	0x00b7f006,
-	0x98000c98,
-	0xe7f0010d,
-	0x6621f500,
-	0x08a7f001,
-	0x010921f5,
-	0x021521f5,
-	0xf02201f4,
-	0x21f40ca7,
-	0x1017f1c9,
-	0x0614b60a,
-	0xd00527f0,
-/* 0x08fa: ctx_xfer_post_save_wait */
-	0x12cf0012,
-	0x0522fd00,
-	0xf4fa1bf4,
-/* 0x0906: ctx_xfer_post */
-	0xf7f03202,
-	0x5121f502,
-	0xf5f4bd06,
-	0xf5069221,
-	0xf5023421,
-	0xbd066021,
-	0x5121f5f4,
-	0x1011f406,
-	0xfd400198,
-	0x0bf40511,
-	0xe521f507,
-/* 0x0931: ctx_xfer_no_post_mmio */
-	0x4321f507,
-/* 0x0935: ctx_xfer_done */
-	0x0000f806,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf001acf0,
+	0xb7f006a5,
+	0x000c9800,
+	0xf0010d98,
+	0x21f500e7,
+	0xa7f0016f,
+	0x1021f508,
+	0x5e21f501,
+	0x1301f402,
+	0xf40ca7f0,
+	0xf7f0d021,
+	0x9421f505,
+	0x3202f408,
+/* 0x0b1c: ctx_xfer_post */
+	0xf502f7f0,
+	0xbd080d21,
+	0x6c21f5f4,
+	0x7f21f508,
+	0x1f21f502,
+	0xf5f4bd08,
+	0xf4080d21,
+	0x01981011,
+	0x0511fd40,
+	0xf5070bf4,
+/* 0x0b47: ctx_xfer_no_post_mmio */
+	0xf509eb21,
+/* 0x0b4b: ctx_xfer_done */
+	0xf807fc21,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnve0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnve0.fuc.h
index eb7bc0e9576..51c3797d853 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnve0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnve0.fuc.h
@@ -206,14 +206,14 @@ uint32_t nve0_grhub_data[] = {
 };
 
 uint32_t nve0_grhub_code[] = {
-	0x031b0ef5,
+	0x039b0ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -237,184 +237,214 @@ uint32_t nve0_grhub_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f405,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -428,7 +458,7 @@ uint32_t nve0_grhub_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0x07f100f8,
 	0x03f00500,
 	0x000fd002,
@@ -436,82 +466,117 @@ uint32_t nve0_grhub_code[] = {
 	0x0007f101,
 	0x0303f007,
 	0xbd000fd0,
-/* 0x031b: init */
+/* 0x039b: init */
 	0xbd00f804,
-	0x0004fe04,
-	0xf10007fe,
-	0xf0120017,
-	0x12d00227,
-	0xb117f100,
-	0x0010fe05,
-	0x040017f1,
-	0xf1c010d0,
-	0xb6040437,
-	0x27f10634,
-	0x32d02003,
-	0x0427f100,
-	0x0132d020,
+	0x0007fe04,
+	0x420017f1,
+	0xcf0013f0,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf10227f0,
+	0xf0120007,
+	0x02d00003,
+	0xf104bd00,
+	0xfe06c817,
+	0x24bd0010,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0x200327f1,
+	0x010007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200427f1,
+	0x010407f1,
+	0xd00103f0,
+	0x04bd0002,
 	0x200b27f1,
-	0xf10232d0,
-	0xd0200c27,
-	0x27f10732,
-	0x24b60c24,
-	0x0003b906,
-	0xf10023d0,
+	0x010807f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200c27f1,
+	0x011c07f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1010392,
+	0xf0090007,
+	0x03d00303,
+	0xf104bd00,
 	0xf0870427,
-	0x12d00023,
-	0x0012b700,
-	0x0427f001,
-	0xf40012d0,
-	0xe7f11031,
-	0xe3f09604,
-	0x6821f440,
-	0x8090f1c7,
-	0xf4f00301,
-	0x020f801f,
-	0xbb0117f0,
-	0x12b6041f,
-	0x0c27f101,
-	0x0624b604,
-	0xd00021d0,
-	0x17f14021,
-	0x0e980100,
-	0x010f9800,
-	0x014721f5,
-	0x070037f1,
-	0x950634b6,
-	0x34d00814,
-	0x4034d000,
-	0x130030b7,
-	0xb6001fbb,
-	0x3fd002f5,
-	0x0815b600,
-	0xb60110b6,
-	0x1fb90814,
-	0x7121f502,
-	0x001fbb02,
-	0xf1020398,
-	0xf0200047,
-/* 0x03f6: init_gpc */
-	0x4ea05043,
-	0x1fb90804,
-	0x8d21f402,
-	0x010c4ea0,
-	0x21f4f4bd,
-	0x044ea08d,
-	0x8d21f401,
-	0x01004ea0,
-	0xf402f7f0,
-	0x4ea08d21,
-/* 0x041e: init_gpc_wait */
-	0x21f40800,
-	0x1fffc868,
-	0xa0fa0bf4,
-	0xf408044e,
-	0x1fbb6821,
-	0x0040b700,
-	0x0132b680,
-	0xf1be1bf4,
+	0x07f10023,
+	0x03f00400,
+	0x0002d000,
+	0x27f004bd,
+	0x0007f104,
+	0x0003f003,
+	0xbd0002d0,
+	0x1031f404,
+	0x9604e7f1,
+	0xf440e3f0,
+	0xfeb96821,
+	0x90f1c702,
+	0xf0030180,
+	0x0f801ff4,
+	0x0117f002,
+	0xb6041fbb,
+	0x07f10112,
+	0x03f00300,
+	0x0001d001,
+	0x07f104bd,
+	0x03f00400,
+	0x0001d001,
+	0x17f104bd,
+	0xf7f00100,
+	0xd721f502,
+	0xe921f507,
+	0x10f7f007,
+	0x083621f5,
+	0x98000e98,
+	0x21f5010f,
+	0x14950150,
+	0x0007f108,
+	0x0103f0c0,
+	0xbd0004d0,
+	0x0007f104,
+	0x0103f0c1,
+	0xbd0004d0,
+	0x0030b704,
+	0x001fbb13,
+	0xf102f5b6,
+	0xf0d30007,
+	0x0fd00103,
+	0xb604bd00,
+	0x10b60815,
+	0x0814b601,
+	0xf5021fb9,
+	0xbb02d321,
+	0x0398001f,
+	0x0047f102,
+	0x5043f020,
+/* 0x04f4: init_gpc */
+	0x08044ea0,
+	0xf4021fb9,
+	0x4ea09d21,
+	0xf4bd010c,
+	0xa09d21f4,
+	0xf401044e,
+	0x4ea09d21,
+	0xf7f00100,
+	0x9d21f402,
+	0x08004ea0,
+/* 0x051c: init_gpc_wait */
+	0xc86821f4,
+	0x0bf41fff,
+	0x044ea0fa,
+	0x6821f408,
+	0xb7001fbb,
+	0xb6800040,
+	0x1bf40132,
+	0x00f7f0be,
+	0x083621f5,
+	0xf500f7f0,
+	0xf107d721,
 	0xf0010007,
 	0x01d00203,
 	0xbd04bd00,
@@ -519,340 +584,401 @@ uint32_t nve0_grhub_code[] = {
 	0x080007f1,
 	0xd00203f0,
 	0x04bd0001,
-/* 0x0458: main */
+/* 0x0564: main */
 	0xf40031f4,
 	0xd7f00028,
 	0x3921f410,
 	0xb1f401f4,
 	0xf54001e4,
-	0xbd00de1b,
+	0xbd00e91b,
 	0x0499f094,
 	0x0f0007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0b0017f1,
-	0xcf0614b6,
-	0x11cf4012,
-	0x1f13c800,
-	0x00870bf5,
-	0xf41f23c8,
-	0x20f9620b,
-	0xbd0212b9,
-	0x0799f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0xf40132f4,
-	0x21f50231,
-	0x94bd0801,
+	0xc00017f1,
+	0xcf0213f0,
+	0x27f10011,
+	0x23f0c100,
+	0x0022cf02,
+	0xf51f13c8,
+	0xc800890b,
+	0x0bf41f23,
+	0xb920f962,
+	0x94bd0212,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
-	0xfc04bd00,
-	0xf094bd20,
-	0x07f10699,
-	0x03f00f00,
-	0x0009d002,
-	0x31f404bd,
-	0x0121f501,
-	0xf094bd08,
-	0x07f10699,
+	0xf404bd00,
+	0x31f40132,
+	0x0221f502,
+	0xf094bd0a,
+	0x07f10799,
 	0x03f01700,
 	0x0009d002,
-	0x0ef404bd,
-/* 0x04f9: chsw_prev_no_next */
-	0xb920f931,
-	0x32f40212,
-	0x0232f401,
-	0x080121f5,
-	0x17f120fc,
-	0x14b60b00,
-	0x0012d006,
-/* 0x0517: chsw_no_prev */
-	0xc8130ef4,
-	0x0bf41f23,
-	0x0131f40d,
-	0xf50232f4,
-/* 0x0527: chsw_done */
-	0xf1080121,
-	0xb60b0c17,
-	0x27f00614,
-	0x0012d001,
+	0x20fc04bd,
 	0x99f094bd,
-	0x0007f104,
+	0x0007f106,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0131f404,
+	0x0a0221f5,
+	0x99f094bd,
+	0x0007f106,
 	0x0203f017,
 	0xbd0009d0,
-	0x130ef504,
-/* 0x0549: main_not_ctx_switch */
-	0x01e4b0ff,
-	0xb90d1bf4,
-	0x21f502f2,
-	0x0ef40795,
-/* 0x0559: main_not_ctx_chan */
-	0x02e4b046,
-	0xbd321bf4,
-	0x0799f094,
-	0x0f0007f1,
+	0x330ef404,
+/* 0x060c: chsw_prev_no_next */
+	0x12b920f9,
+	0x0132f402,
+	0xf50232f4,
+	0xfc0a0221,
+	0x0007f120,
+	0x0203f0c0,
+	0xbd0002d0,
+	0x130ef404,
+/* 0x062c: chsw_no_prev */
+	0xf41f23c8,
+	0x31f40d0b,
+	0x0232f401,
+	0x0a0221f5,
+/* 0x063c: chsw_done */
+	0xf10127f0,
+	0xf0c30007,
+	0x02d00203,
+	0xbd04bd00,
+	0x0499f094,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0xf40132f4,
-	0x21f50232,
-	0x94bd0801,
+	0xff080ef5,
+/* 0x0660: main_not_ctx_switch */
+	0xf401e4b0,
+	0xf2b90d1b,
+	0x9a21f502,
+	0x460ef409,
+/* 0x0670: main_not_ctx_chan */
+	0xf402e4b0,
+	0x94bd321b,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
 	0xf404bd00,
-/* 0x058e: main_not_ctx_save */
-	0xef94110e,
-	0x01f5f010,
-	0x02fe21f5,
-	0xfec00ef5,
-/* 0x059c: main_done */
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f008,
-	0xbd0002d0,
-	0xab0ef504,
-/* 0x05b1: ih */
-	0xfe80f9fe,
-	0x80f90188,
-	0xa0f990f9,
-	0xd0f9b0f9,
-	0xf0f9e0f9,
-	0x0acf04bd,
-	0x04abc480,
-	0xf11d0bf4,
-	0xf01900b7,
-	0xbecf10d7,
-	0x00bfcf40,
+	0x32f40132,
+	0x0221f502,
+	0xf094bd0a,
+	0x07f10799,
+	0x03f01700,
+	0x0009d002,
+	0x0ef404bd,
+/* 0x06a5: main_not_ctx_save */
+	0x10ef9411,
+	0xf501f5f0,
+	0xf5037e21,
+/* 0x06b3: main_done */
+	0xbdfeb50e,
+	0x1f29f024,
+	0x080007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xfea00ef5,
+/* 0x06c8: ih */
+	0x88fe80f9,
+	0xf980f901,
+	0xf9a0f990,
+	0xf9d0f9b0,
+	0xbdf0f9e0,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x10d7f030,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
 	0xb70421f4,
 	0xf00400b0,
-	0xbed001e7,
-/* 0x05e9: ih_no_fifo */
-	0x00abe400,
-	0x0d0bf401,
-	0xf110d7f0,
-	0xf44001e7,
-/* 0x05fa: ih_no_ctxsw */
-	0xb7f10421,
-	0xb0bd0104,
-	0xf4b4abff,
-	0xa7f10d0b,
-	0xa4b60c1c,
-	0x00abd006,
-/* 0x0610: ih_no_other */
-	0xfc400ad0,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x071a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x10d7f00d,
+	0x4001e7f1,
+/* 0x072b: ih_no_ctxsw */
+	0xe40421f4,
+	0xf40400ab,
+	0xe7f16c0b,
+	0xe3f00708,
+	0x6821f440,
+	0xf102ffb9,
+	0xf0040007,
+	0x0fd00203,
+	0xf104bd00,
+	0xf00704e7,
+	0x21f440e3,
+	0x02ffb968,
+	0x030007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0x9450fec7,
+	0xf7f102ee,
+	0xf3f00700,
+	0x00efbb40,
+	0xf16821f4,
+	0xf0020007,
+	0x0fd00203,
+	0xf004bd00,
+	0x21f503f7,
+	0xb7f1037e,
+	0xbfb90100,
+	0x44e7f102,
+	0x40e3f001,
+/* 0x079b: ih_no_fwmthd */
+	0xf19d21f4,
+	0xbd0504b7,
+	0xb4abffb0,
+	0xf10f0bf4,
+	0xf0070007,
+	0x0bd00303,
+/* 0x07b3: ih_no_other */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
 	0xfce0fcf0,
 	0xfcb0fcd0,
 	0xfc90fca0,
 	0x0088fe80,
 	0x32f480fc,
-/* 0x062b: ctx_4170s */
-	0xf101f800,
-	0xf04170e7,
-	0xf5f040e3,
-	0x8d21f410,
-/* 0x063a: ctx_4170w */
+/* 0x07d7: ctx_4170s */
+	0xf001f800,
+	0xffb910f5,
+	0x70e7f102,
+	0x40e3f041,
+	0xf89d21f4,
+/* 0x07e9: ctx_4170w */
+	0x70e7f100,
+	0x40e3f041,
+	0xb96821f4,
+	0xf4f002ff,
+	0xf01bf410,
+/* 0x07fe: ctx_redswitch */
 	0xe7f100f8,
-	0xe3f04170,
-	0x6821f440,
-	0xf410f4f0,
+	0xe5f00200,
+	0x20e5f040,
+	0xf110e5f0,
+	0xf0850007,
+	0x0ed00103,
+	0xf004bd00,
+/* 0x081a: ctx_redswitch_delay */
+	0xf2b608f7,
+	0xfd1bf401,
+	0x0400e5f1,
+	0x0100e5f1,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000e,
+/* 0x0836: ctx_86c */
+	0x07f100f8,
+	0x03f01b00,
+	0x000fd002,
+	0xffb904bd,
+	0x14e7f102,
+	0x40e3f08a,
+	0xb99d21f4,
+	0xe7f102ff,
+	0xe3f0a86c,
+	0x9d21f441,
+/* 0x085e: ctx_mem */
+	0x07f100f8,
+	0x03f08400,
+	0x000fd002,
+/* 0x086a: ctx_mem_wait */
+	0xf7f104bd,
+	0xf3f08400,
+	0x00ffcf02,
+	0xf405fffd,
 	0x00f8f31b,
-/* 0x064c: ctx_redswitch */
-	0x0614e7f1,
-	0xf106e4b6,
-	0xd00270f7,
-	0xf7f000ef,
-/* 0x065d: ctx_redswitch_delay */
-	0x01f2b608,
-	0xf1fd1bf4,
-	0xd00770f7,
-	0x00f800ef,
-/* 0x066c: ctx_86c */
-	0x086ce7f1,
-	0xd006e4b6,
-	0xe7f100ef,
-	0xe3f08a14,
-	0x8d21f440,
-	0xa86ce7f1,
-	0xf441e3f0,
-	0x00f88d21,
-/* 0x068c: ctx_load */
+/* 0x087c: ctx_load */
 	0x99f094bd,
 	0x0007f105,
 	0x0203f00f,
 	0xbd0009d0,
 	0x0ca7f004,
-	0xf1c921f4,
-	0xb60a2417,
-	0x10d00614,
-	0x0037f100,
-	0x0634b60b,
-	0xf14032d0,
-	0xb60a0c17,
-	0x47f00614,
-	0x0012d007,
-/* 0x06c7: ctx_chan_wait_0 */
-	0xcf4014d0,
-	0x44f04014,
-	0xfa1bf41f,
-	0xfe0032d0,
-	0x2af0000b,
-	0x0424b61f,
-	0xbd0220b6,
+	0xbdd021f4,
+	0x0007f1f4,
+	0x0203f089,
+	0xbd000fd0,
+	0x0007f104,
+	0x0203f0c1,
+	0xbd0002d0,
+	0x0007f104,
+	0x0203f083,
+	0xbd0002d0,
+	0x07f7f004,
+	0x085e21f5,
+	0xc00007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf0000bfe,
+	0x24b61f2a,
+	0x0220b604,
+	0x99f094bd,
+	0x0007f108,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f081,
+	0xbd0002d0,
+	0x0027f104,
+	0x0023f100,
+	0x0225f080,
+	0x880007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf11017f0,
+	0xf0020027,
+	0x12fa0223,
+	0xbd03f805,
 	0x0899f094,
-	0x0f0007f1,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0a0417f1,
-	0xd00614b6,
-	0x17f10012,
-	0x14b60a20,
-	0x0227f006,
-	0x800023f1,
-	0xf00012d0,
-	0x27f11017,
-	0x23f00200,
-	0x0512fa02,
-	0x94bd03f8,
-	0xf10899f0,
-	0xf0170007,
+	0xb6810198,
+	0x02981814,
+	0x0825b680,
+	0x800512fd,
+	0x94bd1601,
+	0xf10999f0,
+	0xf00f0007,
 	0x09d00203,
-	0x9804bd00,
-	0x14b68101,
-	0x80029818,
-	0xfd0825b6,
-	0x01800512,
-	0xf094bd16,
-	0x07f10999,
-	0x03f00f00,
-	0x0009d002,
-	0x27f104bd,
-	0x24b60a04,
-	0x0021d006,
-	0xf10127f0,
-	0xb60a2017,
-	0x12d00614,
-	0x0017f100,
-	0x0613f001,
-	0xf80501fa,
-	0xf094bd03,
-	0x07f10999,
-	0x03f01700,
-	0x0009d002,
-	0x94bd04bd,
-	0xf10599f0,
+	0xf104bd00,
+	0xf0810007,
+	0x01d00203,
+	0xf004bd00,
+	0x07f10127,
+	0x03f08800,
+	0x0002d002,
+	0x17f104bd,
+	0x13f00100,
+	0x0501fa06,
+	0x94bd03f8,
+	0xf10999f0,
 	0xf0170007,
 	0x09d00203,
-	0xf804bd00,
-/* 0x0795: ctx_chan */
-	0x8c21f500,
-	0x0ca7f006,
-	0xf1c921f4,
-	0xb60a1017,
-	0x27f00614,
-	0x0012d005,
-/* 0x07ac: ctx_chan_wait */
-	0xfd0012cf,
-	0x1bf40522,
-/* 0x07b7: ctx_mmio_exec */
-	0x9800f8fa,
-	0x27f14103,
-	0x24b60a04,
-	0x0023d006,
-/* 0x07c6: ctx_mmio_loop */
+	0xbd04bd00,
+	0x0599f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x099a: ctx_chan */
+	0x21f500f8,
+	0xa7f0087c,
+	0xd021f40c,
+	0xf505f7f0,
+	0xf8085e21,
+/* 0x09ad: ctx_mmio_exec */
+	0x41039800,
+	0x810007f1,
+	0xd00203f0,
+	0x04bd0003,
+/* 0x09be: ctx_mmio_loop */
 	0x34c434bd,
 	0x0f1bf4ff,
 	0x020057f1,
 	0xfa0653f0,
 	0x03f80535,
-/* 0x07d8: ctx_mmio_pull */
+/* 0x09d0: ctx_mmio_pull */
 	0x98804e98,
 	0x21f4814f,
-	0x0830b68d,
+	0x0830b69d,
 	0xf40112b6,
-/* 0x07ea: ctx_mmio_done */
+/* 0x09e2: ctx_mmio_done */
 	0x0398df1b,
-	0x0023d016,
-	0xf1400080,
-	0xf0010017,
-	0x01fa0613,
-	0xf803f806,
-/* 0x0801: ctx_xfer */
-	0x00f7f100,
-	0x06f4b60c,
-	0xd004e7f0,
-/* 0x080e: ctx_xfer_idle */
-	0xfecf80fe,
-	0x00e4f100,
-	0xf91bf420,
-	0xf40611f4,
-/* 0x081e: ctx_xfer_pre */
-	0xf7f00d02,
-	0x6c21f510,
-	0x1c11f406,
-/* 0x0828: ctx_xfer_pre_load */
-	0xf502f7f0,
-	0xf5062b21,
-	0xf5063a21,
-	0xbd064c21,
-	0x2b21f5f4,
-	0x8c21f506,
-/* 0x0841: ctx_xfer_exec */
-	0x16019806,
-	0x041427f1,
-	0xd00624b6,
-	0xe7f10020,
-	0xe3f0a500,
-	0x021fb941,
-	0xb68d21f4,
-	0xfcf004e0,
-	0x022cf001,
-	0xfd0124b6,
-	0x21f405f2,
-	0xfc17f18d,
-	0x0213f04a,
-	0xd00c27f0,
-	0x21f50012,
-	0x27f10215,
-	0x23f047fc,
-	0x0020d002,
+	0x0007f116,
+	0x0203f081,
+	0xbd0003d0,
+	0x40008004,
+	0x010017f1,
+	0xfa0613f0,
+	0x03f80601,
+/* 0x0a02: ctx_xfer */
+	0xe7f000f8,
+	0x0007f104,
+	0x0303f002,
+	0xbd000ed0,
+/* 0x0a11: ctx_xfer_idle */
+	0x00e7f104,
+	0x03e3f000,
+	0xf100eecf,
+	0xf42000e4,
+	0x11f4f21b,
+	0x0d02f406,
+/* 0x0a28: ctx_xfer_pre */
+	0xf510f7f0,
+	0xf4083621,
+/* 0x0a32: ctx_xfer_pre_load */
+	0xf7f01c11,
+	0xd721f502,
+	0xe921f507,
+	0xfe21f507,
+	0xf5f4bd07,
+	0xf507d721,
+/* 0x0a4b: ctx_xfer_exec */
+	0x98087c21,
+	0x24bd1601,
+	0x050007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1021fb9,
+	0xf0a500e7,
+	0x21f441e3,
+	0x01fcf09d,
+	0xb6022cf0,
+	0xf2fd0124,
+	0x02ffb905,
+	0xa504e7f1,
+	0xf441e3f0,
+	0x21f59d21,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
 	0xb6012cf0,
-	0x12d00320,
-	0x01acf000,
-	0xf006a5f0,
-	0x0c9800b7,
-	0x010d9800,
-	0xf500e7f0,
-	0xf0016621,
-	0x21f508a7,
-	0x21f50109,
-	0x01f40215,
-	0x0ca7f022,
-	0xf1c921f4,
-	0xb60a1017,
-	0x27f00614,
-	0x0012d005,
-/* 0x08c8: ctx_xfer_post_save_wait */
-	0xfd0012cf,
-	0x1bf40522,
-	0x2e02f4fa,
-/* 0x08d4: ctx_xfer_post */
-	0xf502f7f0,
-	0xbd062b21,
-	0x6c21f5f4,
-	0x3421f506,
-	0x3a21f502,
-	0xf5f4bd06,
-	0xf4062b21,
-	0x01981011,
-	0x0511fd40,
-	0xf5070bf4,
-/* 0x08ff: ctx_xfer_no_post_mmio */
-/* 0x08ff: ctx_xfer_done */
-	0xf807b721,
-	0x00000000,
-	0x00000000,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x06a5f001,
+	0x9800b7f0,
+	0x0d98000c,
+	0x00e7f001,
+	0x016f21f5,
+	0xf508a7f0,
+	0xf5011021,
+	0xf4025e21,
+	0xa7f01301,
+	0xd021f40c,
+	0xf505f7f0,
+	0xf4085e21,
+/* 0x0ada: ctx_xfer_post */
+	0xf7f02e02,
+	0xd721f502,
+	0xf5f4bd07,
+	0xf5083621,
+	0xf5027f21,
+	0xbd07e921,
+	0xd721f5f4,
+	0x1011f407,
+	0xfd400198,
+	0x0bf40511,
+	0xad21f507,
+/* 0x0b05: ctx_xfer_no_post_mmio */
+/* 0x0b05: ctx_xfer_done */
+	0x0000f809,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvf0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvf0.fuc.h
index 438506d1474..a0af4b703a8 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvf0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvf0.fuc.h
@@ -206,14 +206,14 @@ uint32_t nvf0_grhub_data[] = {
 };
 
 uint32_t nvf0_grhub_code[] = {
-	0x031b0ef5,
+	0x039b0ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -237,184 +237,214 @@ uint32_t nvf0_grhub_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f03700,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f037,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f03700,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f037,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf0370007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f405,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f03700,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -428,7 +458,7 @@ uint32_t nvf0_grhub_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0x07f100f8,
 	0x03f00500,
 	0x000fd002,
@@ -436,82 +466,117 @@ uint32_t nvf0_grhub_code[] = {
 	0x0007f101,
 	0x0303f007,
 	0xbd000fd0,
-/* 0x031b: init */
+/* 0x039b: init */
 	0xbd00f804,
-	0x0004fe04,
-	0xf10007fe,
-	0xf0120017,
-	0x12d00227,
-	0xb117f100,
-	0x0010fe05,
-	0x040017f1,
-	0xf1c010d0,
-	0xb6040437,
-	0x27f10634,
-	0x32d02003,
-	0x0427f100,
-	0x0132d020,
+	0x0007fe04,
+	0x420017f1,
+	0xcf0013f0,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf10227f0,
+	0xf0120007,
+	0x02d00003,
+	0xf104bd00,
+	0xfe06c817,
+	0x24bd0010,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0x200327f1,
+	0x010007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200427f1,
+	0x010407f1,
+	0xd00103f0,
+	0x04bd0002,
 	0x200b27f1,
-	0xf10232d0,
-	0xd0200c27,
-	0x27f10732,
-	0x24b60c24,
-	0x0003b906,
-	0xf10023d0,
+	0x010807f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200c27f1,
+	0x011c07f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1010392,
+	0xf0090007,
+	0x03d00303,
+	0xf104bd00,
 	0xf0870427,
-	0x12d00023,
-	0x0012b700,
-	0x0427f001,
-	0xf40012d0,
-	0xe7f11031,
-	0xe3f09604,
-	0x6821f440,
-	0x8090f1c7,
-	0xf4f00301,
-	0x020f801f,
-	0xbb0117f0,
-	0x12b6041f,
-	0x0c27f101,
-	0x0624b604,
-	0xd00021d0,
-	0x17f14021,
-	0x0e980100,
-	0x010f9800,
-	0x014721f5,
-	0x070037f1,
-	0x950634b6,
-	0x34d00814,
-	0x4034d000,
-	0x130030b7,
-	0xb6001fbb,
-	0x3fd002f5,
-	0x0815b600,
-	0xb60110b6,
-	0x1fb90814,
-	0x7121f502,
-	0x001fbb02,
-	0xf1020398,
-	0xf0200047,
-/* 0x03f6: init_gpc */
-	0x4ea05043,
-	0x1fb90804,
-	0x8d21f402,
-	0x010c4ea0,
-	0x21f4f4bd,
-	0x044ea08d,
-	0x8d21f401,
-	0x01004ea0,
-	0xf402f7f0,
-	0x4ea08d21,
-/* 0x041e: init_gpc_wait */
-	0x21f40800,
-	0x1fffc868,
-	0xa0fa0bf4,
-	0xf408044e,
-	0x1fbb6821,
-	0x0040b700,
-	0x0132b680,
-	0xf1be1bf4,
+	0x07f10023,
+	0x03f00400,
+	0x0002d000,
+	0x27f004bd,
+	0x0007f104,
+	0x0003f003,
+	0xbd0002d0,
+	0x1031f404,
+	0x9604e7f1,
+	0xf440e3f0,
+	0xfeb96821,
+	0x90f1c702,
+	0xf0030180,
+	0x0f801ff4,
+	0x0117f002,
+	0xb6041fbb,
+	0x07f10112,
+	0x03f00300,
+	0x0001d001,
+	0x07f104bd,
+	0x03f00400,
+	0x0001d001,
+	0x17f104bd,
+	0xf7f00100,
+	0xd721f502,
+	0xe921f507,
+	0x10f7f007,
+	0x083621f5,
+	0x98000e98,
+	0x21f5010f,
+	0x14950150,
+	0x0007f108,
+	0x0103f0c0,
+	0xbd0004d0,
+	0x0007f104,
+	0x0103f0c1,
+	0xbd0004d0,
+	0x0030b704,
+	0x001fbb13,
+	0xf102f5b6,
+	0xf0d30007,
+	0x0fd00103,
+	0xb604bd00,
+	0x10b60815,
+	0x0814b601,
+	0xf5021fb9,
+	0xbb02d321,
+	0x0398001f,
+	0x0047f102,
+	0x5043f020,
+/* 0x04f4: init_gpc */
+	0x08044ea0,
+	0xf4021fb9,
+	0x4ea09d21,
+	0xf4bd010c,
+	0xa09d21f4,
+	0xf401044e,
+	0x4ea09d21,
+	0xf7f00100,
+	0x9d21f402,
+	0x08004ea0,
+/* 0x051c: init_gpc_wait */
+	0xc86821f4,
+	0x0bf41fff,
+	0x044ea0fa,
+	0x6821f408,
+	0xb7001fbb,
+	0xb6800040,
+	0x1bf40132,
+	0x00f7f0be,
+	0x083621f5,
+	0xf500f7f0,
+	0xf107d721,
 	0xf0010007,
 	0x01d00203,
 	0xbd04bd00,
@@ -519,340 +584,401 @@ uint32_t nvf0_grhub_code[] = {
 	0x300007f1,
 	0xd00203f0,
 	0x04bd0001,
-/* 0x0458: main */
+/* 0x0564: main */
 	0xf40031f4,
 	0xd7f00028,
 	0x3921f410,
 	0xb1f401f4,
 	0xf54001e4,
-	0xbd00de1b,
+	0xbd00e91b,
 	0x0499f094,
 	0x370007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0b0017f1,
-	0xcf0614b6,
-	0x11cf4012,
-	0x1f13c800,
-	0x00870bf5,
-	0xf41f23c8,
-	0x20f9620b,
-	0xbd0212b9,
-	0x0799f094,
-	0x370007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0xf40132f4,
-	0x21f50231,
-	0x94bd0801,
+	0xc00017f1,
+	0xcf0213f0,
+	0x27f10011,
+	0x23f0c100,
+	0x0022cf02,
+	0xf51f13c8,
+	0xc800890b,
+	0x0bf41f23,
+	0xb920f962,
+	0x94bd0212,
 	0xf10799f0,
-	0xf0170007,
+	0xf0370007,
 	0x09d00203,
-	0xfc04bd00,
-	0xf094bd20,
-	0x07f10699,
-	0x03f03700,
-	0x0009d002,
-	0x31f404bd,
-	0x0121f501,
-	0xf094bd08,
-	0x07f10699,
+	0xf404bd00,
+	0x31f40132,
+	0x0221f502,
+	0xf094bd0a,
+	0x07f10799,
 	0x03f01700,
 	0x0009d002,
-	0x0ef404bd,
-/* 0x04f9: chsw_prev_no_next */
-	0xb920f931,
-	0x32f40212,
-	0x0232f401,
-	0x080121f5,
-	0x17f120fc,
-	0x14b60b00,
-	0x0012d006,
-/* 0x0517: chsw_no_prev */
-	0xc8130ef4,
-	0x0bf41f23,
-	0x0131f40d,
-	0xf50232f4,
-/* 0x0527: chsw_done */
-	0xf1080121,
-	0xb60b0c17,
-	0x27f00614,
-	0x0012d001,
+	0x20fc04bd,
 	0x99f094bd,
-	0x0007f104,
+	0x0007f106,
+	0x0203f037,
+	0xbd0009d0,
+	0x0131f404,
+	0x0a0221f5,
+	0x99f094bd,
+	0x0007f106,
 	0x0203f017,
 	0xbd0009d0,
-	0x130ef504,
-/* 0x0549: main_not_ctx_switch */
-	0x01e4b0ff,
-	0xb90d1bf4,
-	0x21f502f2,
-	0x0ef40795,
-/* 0x0559: main_not_ctx_chan */
-	0x02e4b046,
-	0xbd321bf4,
-	0x0799f094,
-	0x370007f1,
+	0x330ef404,
+/* 0x060c: chsw_prev_no_next */
+	0x12b920f9,
+	0x0132f402,
+	0xf50232f4,
+	0xfc0a0221,
+	0x0007f120,
+	0x0203f0c0,
+	0xbd0002d0,
+	0x130ef404,
+/* 0x062c: chsw_no_prev */
+	0xf41f23c8,
+	0x31f40d0b,
+	0x0232f401,
+	0x0a0221f5,
+/* 0x063c: chsw_done */
+	0xf10127f0,
+	0xf0c30007,
+	0x02d00203,
+	0xbd04bd00,
+	0x0499f094,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0xf40132f4,
-	0x21f50232,
-	0x94bd0801,
+	0xff080ef5,
+/* 0x0660: main_not_ctx_switch */
+	0xf401e4b0,
+	0xf2b90d1b,
+	0x9a21f502,
+	0x460ef409,
+/* 0x0670: main_not_ctx_chan */
+	0xf402e4b0,
+	0x94bd321b,
 	0xf10799f0,
-	0xf0170007,
+	0xf0370007,
 	0x09d00203,
 	0xf404bd00,
-/* 0x058e: main_not_ctx_save */
-	0xef94110e,
-	0x01f5f010,
-	0x02fe21f5,
-	0xfec00ef5,
-/* 0x059c: main_done */
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f030,
-	0xbd0002d0,
-	0xab0ef504,
-/* 0x05b1: ih */
-	0xfe80f9fe,
-	0x80f90188,
-	0xa0f990f9,
-	0xd0f9b0f9,
-	0xf0f9e0f9,
-	0x0acf04bd,
-	0x04abc480,
-	0xf11d0bf4,
-	0xf01900b7,
-	0xbecf10d7,
-	0x00bfcf40,
+	0x32f40132,
+	0x0221f502,
+	0xf094bd0a,
+	0x07f10799,
+	0x03f01700,
+	0x0009d002,
+	0x0ef404bd,
+/* 0x06a5: main_not_ctx_save */
+	0x10ef9411,
+	0xf501f5f0,
+	0xf5037e21,
+/* 0x06b3: main_done */
+	0xbdfeb50e,
+	0x1f29f024,
+	0x300007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xfea00ef5,
+/* 0x06c8: ih */
+	0x88fe80f9,
+	0xf980f901,
+	0xf9a0f990,
+	0xf9d0f9b0,
+	0xbdf0f9e0,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x10d7f030,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
 	0xb70421f4,
 	0xf00400b0,
-	0xbed001e7,
-/* 0x05e9: ih_no_fifo */
-	0x00abe400,
-	0x0d0bf401,
-	0xf110d7f0,
-	0xf44001e7,
-/* 0x05fa: ih_no_ctxsw */
-	0xb7f10421,
-	0xb0bd0104,
-	0xf4b4abff,
-	0xa7f10d0b,
-	0xa4b60c1c,
-	0x00abd006,
-/* 0x0610: ih_no_other */
-	0xfc400ad0,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x071a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x10d7f00d,
+	0x4001e7f1,
+/* 0x072b: ih_no_ctxsw */
+	0xe40421f4,
+	0xf40400ab,
+	0xe7f16c0b,
+	0xe3f00708,
+	0x6821f440,
+	0xf102ffb9,
+	0xf0040007,
+	0x0fd00203,
+	0xf104bd00,
+	0xf00704e7,
+	0x21f440e3,
+	0x02ffb968,
+	0x030007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0x9450fec7,
+	0xf7f102ee,
+	0xf3f00700,
+	0x00efbb40,
+	0xf16821f4,
+	0xf0020007,
+	0x0fd00203,
+	0xf004bd00,
+	0x21f503f7,
+	0xb7f1037e,
+	0xbfb90100,
+	0x44e7f102,
+	0x40e3f001,
+/* 0x079b: ih_no_fwmthd */
+	0xf19d21f4,
+	0xbd0504b7,
+	0xb4abffb0,
+	0xf10f0bf4,
+	0xf0070007,
+	0x0bd00303,
+/* 0x07b3: ih_no_other */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
 	0xfce0fcf0,
 	0xfcb0fcd0,
 	0xfc90fca0,
 	0x0088fe80,
 	0x32f480fc,
-/* 0x062b: ctx_4170s */
-	0xf101f800,
-	0xf04170e7,
-	0xf5f040e3,
-	0x8d21f410,
-/* 0x063a: ctx_4170w */
+/* 0x07d7: ctx_4170s */
+	0xf001f800,
+	0xffb910f5,
+	0x70e7f102,
+	0x40e3f041,
+	0xf89d21f4,
+/* 0x07e9: ctx_4170w */
+	0x70e7f100,
+	0x40e3f041,
+	0xb96821f4,
+	0xf4f002ff,
+	0xf01bf410,
+/* 0x07fe: ctx_redswitch */
 	0xe7f100f8,
-	0xe3f04170,
-	0x6821f440,
-	0xf410f4f0,
+	0xe5f00200,
+	0x20e5f040,
+	0xf110e5f0,
+	0xf0850007,
+	0x0ed00103,
+	0xf004bd00,
+/* 0x081a: ctx_redswitch_delay */
+	0xf2b608f7,
+	0xfd1bf401,
+	0x0400e5f1,
+	0x0100e5f1,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000e,
+/* 0x0836: ctx_86c */
+	0x07f100f8,
+	0x03f02300,
+	0x000fd002,
+	0xffb904bd,
+	0x14e7f102,
+	0x40e3f08a,
+	0xb99d21f4,
+	0xe7f102ff,
+	0xe3f0a88c,
+	0x9d21f441,
+/* 0x085e: ctx_mem */
+	0x07f100f8,
+	0x03f08400,
+	0x000fd002,
+/* 0x086a: ctx_mem_wait */
+	0xf7f104bd,
+	0xf3f08400,
+	0x00ffcf02,
+	0xf405fffd,
 	0x00f8f31b,
-/* 0x064c: ctx_redswitch */
-	0x0614e7f1,
-	0xf106e4b6,
-	0xd00270f7,
-	0xf7f000ef,
-/* 0x065d: ctx_redswitch_delay */
-	0x01f2b608,
-	0xf1fd1bf4,
-	0xd00770f7,
-	0x00f800ef,
-/* 0x066c: ctx_86c */
-	0x086ce7f1,
-	0xd006e4b6,
-	0xe7f100ef,
-	0xe3f08a14,
-	0x8d21f440,
-	0xa86ce7f1,
-	0xf441e3f0,
-	0x00f88d21,
-/* 0x068c: ctx_load */
+/* 0x087c: ctx_load */
 	0x99f094bd,
 	0x0007f105,
 	0x0203f037,
 	0xbd0009d0,
 	0x0ca7f004,
-	0xf1c921f4,
-	0xb60a2417,
-	0x10d00614,
-	0x0037f100,
-	0x0634b60b,
-	0xf14032d0,
-	0xb60a0c17,
-	0x47f00614,
-	0x0012d007,
-/* 0x06c7: ctx_chan_wait_0 */
-	0xcf4014d0,
-	0x44f04014,
-	0xfa1bf41f,
-	0xfe0032d0,
-	0x2af0000b,
-	0x0424b61f,
-	0xbd0220b6,
+	0xbdd021f4,
+	0x0007f1f4,
+	0x0203f089,
+	0xbd000fd0,
+	0x0007f104,
+	0x0203f0c1,
+	0xbd0002d0,
+	0x0007f104,
+	0x0203f083,
+	0xbd0002d0,
+	0x07f7f004,
+	0x085e21f5,
+	0xc00007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf0000bfe,
+	0x24b61f2a,
+	0x0220b604,
+	0x99f094bd,
+	0x0007f108,
+	0x0203f037,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f081,
+	0xbd0002d0,
+	0x0027f104,
+	0x0023f100,
+	0x0225f080,
+	0x880007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf11017f0,
+	0xf0020027,
+	0x12fa0223,
+	0xbd03f805,
 	0x0899f094,
-	0x370007f1,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0a0417f1,
-	0xd00614b6,
-	0x17f10012,
-	0x14b60a20,
-	0x0227f006,
-	0x800023f1,
-	0xf00012d0,
-	0x27f11017,
-	0x23f00200,
-	0x0512fa02,
-	0x94bd03f8,
-	0xf10899f0,
-	0xf0170007,
+	0xb6810198,
+	0x02981814,
+	0x0825b680,
+	0x800512fd,
+	0x94bd1601,
+	0xf10999f0,
+	0xf0370007,
 	0x09d00203,
-	0x9804bd00,
-	0x14b68101,
-	0x80029818,
-	0xfd0825b6,
-	0x01800512,
-	0xf094bd16,
-	0x07f10999,
-	0x03f03700,
-	0x0009d002,
-	0x27f104bd,
-	0x24b60a04,
-	0x0021d006,
-	0xf10127f0,
-	0xb60a2017,
-	0x12d00614,
-	0x0017f100,
-	0x0613f001,
-	0xf80501fa,
-	0xf094bd03,
-	0x07f10999,
-	0x03f01700,
-	0x0009d002,
-	0x94bd04bd,
-	0xf10599f0,
+	0xf104bd00,
+	0xf0810007,
+	0x01d00203,
+	0xf004bd00,
+	0x07f10127,
+	0x03f08800,
+	0x0002d002,
+	0x17f104bd,
+	0x13f00100,
+	0x0501fa06,
+	0x94bd03f8,
+	0xf10999f0,
 	0xf0170007,
 	0x09d00203,
-	0xf804bd00,
-/* 0x0795: ctx_chan */
-	0x8c21f500,
-	0x0ca7f006,
-	0xf1c921f4,
-	0xb60a1017,
-	0x27f00614,
-	0x0012d005,
-/* 0x07ac: ctx_chan_wait */
-	0xfd0012cf,
-	0x1bf40522,
-/* 0x07b7: ctx_mmio_exec */
-	0x9800f8fa,
-	0x27f14103,
-	0x24b60a04,
-	0x0023d006,
-/* 0x07c6: ctx_mmio_loop */
+	0xbd04bd00,
+	0x0599f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x099a: ctx_chan */
+	0x21f500f8,
+	0xa7f0087c,
+	0xd021f40c,
+	0xf505f7f0,
+	0xf8085e21,
+/* 0x09ad: ctx_mmio_exec */
+	0x41039800,
+	0x810007f1,
+	0xd00203f0,
+	0x04bd0003,
+/* 0x09be: ctx_mmio_loop */
 	0x34c434bd,
 	0x0f1bf4ff,
 	0x020057f1,
 	0xfa0653f0,
 	0x03f80535,
-/* 0x07d8: ctx_mmio_pull */
+/* 0x09d0: ctx_mmio_pull */
 	0x98804e98,
 	0x21f4814f,
-	0x0830b68d,
+	0x0830b69d,
 	0xf40112b6,
-/* 0x07ea: ctx_mmio_done */
+/* 0x09e2: ctx_mmio_done */
 	0x0398df1b,
-	0x0023d016,
-	0xf1400080,
-	0xf0010017,
-	0x01fa0613,
-	0xf803f806,
-/* 0x0801: ctx_xfer */
-	0x00f7f100,
-	0x06f4b60c,
-	0xd004e7f0,
-/* 0x080e: ctx_xfer_idle */
-	0xfecf80fe,
-	0x00e4f100,
-	0xf91bf420,
-	0xf40611f4,
-/* 0x081e: ctx_xfer_pre */
-	0xf7f00d02,
-	0x6c21f510,
-	0x1c11f406,
-/* 0x0828: ctx_xfer_pre_load */
-	0xf502f7f0,
-	0xf5062b21,
-	0xf5063a21,
-	0xbd064c21,
-	0x2b21f5f4,
-	0x8c21f506,
-/* 0x0841: ctx_xfer_exec */
-	0x16019806,
-	0x041427f1,
-	0xd00624b6,
-	0xe7f10020,
-	0xe3f0a500,
-	0x021fb941,
-	0xb68d21f4,
-	0xfcf004e0,
-	0x022cf001,
-	0xfd0124b6,
-	0x21f405f2,
-	0xfc17f18d,
-	0x0213f04a,
-	0xd00c27f0,
-	0x21f50012,
-	0x27f10215,
-	0x23f047fc,
-	0x0020d002,
+	0x0007f116,
+	0x0203f081,
+	0xbd0003d0,
+	0x40008004,
+	0x010017f1,
+	0xfa0613f0,
+	0x03f80601,
+/* 0x0a02: ctx_xfer */
+	0xe7f000f8,
+	0x0007f104,
+	0x0303f002,
+	0xbd000ed0,
+/* 0x0a11: ctx_xfer_idle */
+	0x00e7f104,
+	0x03e3f000,
+	0xf100eecf,
+	0xf42000e4,
+	0x11f4f21b,
+	0x0d02f406,
+/* 0x0a28: ctx_xfer_pre */
+	0xf510f7f0,
+	0xf4083621,
+/* 0x0a32: ctx_xfer_pre_load */
+	0xf7f01c11,
+	0xd721f502,
+	0xe921f507,
+	0xfe21f507,
+	0xf5f4bd07,
+	0xf507d721,
+/* 0x0a4b: ctx_xfer_exec */
+	0x98087c21,
+	0x24bd1601,
+	0x050007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1021fb9,
+	0xf0a500e7,
+	0x21f441e3,
+	0x01fcf09d,
+	0xb6022cf0,
+	0xf2fd0124,
+	0x02ffb905,
+	0xa504e7f1,
+	0xf441e3f0,
+	0x21f59d21,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
 	0xb6012cf0,
-	0x12d00320,
-	0x01acf000,
-	0xf006a5f0,
-	0x0c9800b7,
-	0x010d9800,
-	0xf500e7f0,
-	0xf0016621,
-	0x21f508a7,
-	0x21f50109,
-	0x01f40215,
-	0x0ca7f022,
-	0xf1c921f4,
-	0xb60a1017,
-	0x27f00614,
-	0x0012d005,
-/* 0x08c8: ctx_xfer_post_save_wait */
-	0xfd0012cf,
-	0x1bf40522,
-	0x2e02f4fa,
-/* 0x08d4: ctx_xfer_post */
-	0xf502f7f0,
-	0xbd062b21,
-	0x6c21f5f4,
-	0x3421f506,
-	0x3a21f502,
-	0xf5f4bd06,
-	0xf4062b21,
-	0x01981011,
-	0x0511fd40,
-	0xf5070bf4,
-/* 0x08ff: ctx_xfer_no_post_mmio */
-/* 0x08ff: ctx_xfer_done */
-	0xf807b721,
-	0x00000000,
-	0x00000000,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x06a5f001,
+	0x9800b7f0,
+	0x0d98000c,
+	0x00e7f001,
+	0x016f21f5,
+	0xf508a7f0,
+	0xf5011021,
+	0xf4025e21,
+	0xa7f01301,
+	0xd021f40c,
+	0xf505f7f0,
+	0xf4085e21,
+/* 0x0ada: ctx_xfer_post */
+	0xf7f02e02,
+	0xd721f502,
+	0xf5f4bd07,
+	0xf5083621,
+	0xf5027f21,
+	0xbd07e921,
+	0xd721f5f4,
+	0x1011f407,
+	0xfd400198,
+	0x0bf40511,
+	0xad21f507,
+/* 0x0b05: ctx_xfer_no_post_mmio */
+/* 0x0b05: ctx_xfer_done */
+	0x0000f809,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/macros.fuc b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/macros.fuc
index 33a5a82eccb..2a0b0f84429 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/macros.fuc
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/macros.fuc
@@ -28,28 +28,142 @@
 #define GF117 0xd7
 #define GK100 0xe0
 #define GK110 0xf0
+#define GK208 0x108
 
+#define NV_PGRAPH_TRAPPED_ADDR                                         0x400704
+#define NV_PGRAPH_TRAPPED_DATA_LO                                      0x400708
+#define NV_PGRAPH_TRAPPED_DATA_HI                                      0x40070c
+
+#define NV_PGRAPH_FE_OBJECT_TABLE(n)                        ((n) * 4 + 0x400700)
+
+#define NV_PGRAPH_FECS_INTR_ACK                                        0x409004
+#define NV_PGRAPH_FECS_INTR                                            0x409008
+#define NV_PGRAPH_FECS_INTR_FWMTHD                                   0x00000400
+#define NV_PGRAPH_FECS_INTR_CHSW                                     0x00000100
+#define NV_PGRAPH_FECS_INTR_FIFO                                     0x00000004
+#define NV_PGRAPH_FECS_INTR_MODE                                       0x40900c
+#define NV_PGRAPH_FECS_INTR_MODE_FIFO                                0x00000004
+#define NV_PGRAPH_FECS_INTR_MODE_FIFO_LEVEL                          0x00000004
+#define NV_PGRAPH_FECS_INTR_MODE_FIFO_EDGE                           0x00000000
+#define NV_PGRAPH_FECS_INTR_EN_SET                                     0x409010
+#define NV_PGRAPH_FECS_INTR_EN_SET_FIFO                              0x00000004
+#define NV_PGRAPH_FECS_INTR_ROUTE                                      0x40901c
+#define NV_PGRAPH_FECS_ACCESS                                          0x409048
+#define NV_PGRAPH_FECS_ACCESS_FIFO                                   0x00000002
+#define NV_PGRAPH_FECS_FIFO_DATA                                       0x409064
+#define NV_PGRAPH_FECS_FIFO_CMD                                        0x409068
+#define NV_PGRAPH_FECS_FIFO_ACK                                        0x409074
+#define NV_PGRAPH_FECS_CAPS                                            0x409108
 #define NV_PGRAPH_FECS_SIGNAL                                          0x409400
+#define NV_PGRAPH_FECS_IROUTE                                          0x409404
+#define NV_PGRAPH_FECS_BAR_MASK0                                       0x40940c
+#define NV_PGRAPH_FECS_BAR_MASK1                                       0x409410
+#define NV_PGRAPH_FECS_BAR                                             0x409414
+#define NV_PGRAPH_FECS_BAR_SET                                         0x409418
+#define NV_PGRAPH_FECS_RED_SWITCH                                      0x409614
+#define NV_PGRAPH_FECS_RED_SWITCH_ENABLE_ROP                         0x00000400
+#define NV_PGRAPH_FECS_RED_SWITCH_ENABLE_GPC                         0x00000200
+#define NV_PGRAPH_FECS_RED_SWITCH_ENABLE_MAIN                        0x00000100
+#define NV_PGRAPH_FECS_RED_SWITCH_POWER_ROP                          0x00000040
+#define NV_PGRAPH_FECS_RED_SWITCH_POWER_GPC                          0x00000020
+#define NV_PGRAPH_FECS_RED_SWITCH_POWER_MAIN                         0x00000010
+#define NV_PGRAPH_FECS_RED_SWITCH_PAUSE_GPC                          0x00000002
+#define NV_PGRAPH_FECS_RED_SWITCH_PAUSE_MAIN                         0x00000001
+#define NV_PGRAPH_FECS_MMCTX_SAVE_SWBASE                               0x409700
+#define NV_PGRAPH_FECS_MMCTX_LOAD_SWBASE                               0x409704
+#define NV_PGRAPH_FECS_MMCTX_LOAD_COUNT                                0x40974c
+#define NV_PGRAPH_FECS_MMCTX_SAVE_SWBASE                               0x409700
+#define NV_PGRAPH_FECS_MMCTX_LOAD_SWBASE                               0x409704
+#define NV_PGRAPH_FECS_MMCTX_BASE                                      0x409710
+#define NV_PGRAPH_FECS_MMCTX_CTRL                                      0x409714
+#define NV_PGRAPH_FECS_MMCTX_MULTI_STRIDE                              0x409718
+#define NV_PGRAPH_FECS_MMCTX_MULTI_MASK                                0x40971c
+#define NV_PGRAPH_FECS_MMCTX_QUEUE                                     0x409720
+#define NV_PGRAPH_FECS_MMIO_CTRL                                       0x409728
+#define NV_PGRAPH_FECS_MMIO_RDVAL                                      0x40972c
+#define NV_PGRAPH_FECS_MMIO_WRVAL                                      0x409730
+#define NV_PGRAPH_FECS_MMCTX_LOAD_COUNT                                0x40974c
 #if CHIPSET < GK110
 #define NV_PGRAPH_FECS_CC_SCRATCH_VAL(n)                    ((n) * 4 + 0x409800)
 #define NV_PGRAPH_FECS_CC_SCRATCH_SET(n)                    ((n) * 4 + 0x409820)
 #define NV_PGRAPH_FECS_CC_SCRATCH_CLR(n)                    ((n) * 4 + 0x409840)
+#define NV_PGRAPH_FECS_UNK86C                                          0x40986c
 #else
 #define NV_PGRAPH_FECS_CC_SCRATCH_VAL(n)                    ((n) * 4 + 0x409800)
 #define NV_PGRAPH_FECS_CC_SCRATCH_CLR(n)                    ((n) * 4 + 0x409840)
+#define NV_PGRAPH_FECS_UNK86C                                          0x40988c
 #define NV_PGRAPH_FECS_CC_SCRATCH_SET(n)                    ((n) * 4 + 0x4098c0)
 #endif
+#define NV_PGRAPH_FECS_STRANDS_CNT                                     0x409880
+#define NV_PGRAPH_FECS_STRAND_SAVE_SWBASE                              0x409908
+#define NV_PGRAPH_FECS_STRAND_LOAD_SWBASE                              0x40990c
+#define NV_PGRAPH_FECS_STRAND_WORDS                                    0x409910
+#define NV_PGRAPH_FECS_STRAND_DATA                                     0x409918
+#define NV_PGRAPH_FECS_STRAND_SELECT                                   0x40991c
+#define NV_PGRAPH_FECS_STRAND_CMD                                      0x409928
+#define NV_PGRAPH_FECS_STRAND_CMD_SEEK                               0x00000001
+#define NV_PGRAPH_FECS_STRAND_CMD_GET_INFO                           0x00000002
+#define NV_PGRAPH_FECS_STRAND_CMD_SAVE                               0x00000003
+#define NV_PGRAPH_FECS_STRAND_CMD_LOAD                               0x00000004
+#define NV_PGRAPH_FECS_STRAND_CMD_ACTIVATE_FILTER                    0x0000000a
+#define NV_PGRAPH_FECS_STRAND_CMD_DEACTIVATE_FILTER                  0x0000000b
+#define NV_PGRAPH_FECS_STRAND_CMD_ENABLE                             0x0000000c
+#define NV_PGRAPH_FECS_STRAND_CMD_DISABLE                            0x0000000d
+#define NV_PGRAPH_FECS_STRAND_FILTER                                   0x40993c
+#define NV_PGRAPH_FECS_MEM_BASE                                        0x409a04
+#define NV_PGRAPH_FECS_MEM_CHAN                                        0x409a0c
+#define NV_PGRAPH_FECS_MEM_CMD                                         0x409a10
+#define NV_PGRAPH_FECS_MEM_CMD_LOAD_CHAN                             0x00000007
+#define NV_PGRAPH_FECS_MEM_TARGET                                      0x409a20
+#define NV_PGRAPH_FECS_MEM_TARGET_UNK31                              0x80000000
+#define NV_PGRAPH_FECS_MEM_TARGET_AS                                 0x0000001f
+#define NV_PGRAPH_FECS_MEM_TARGET_AS_VM                              0x00000001
+#define NV_PGRAPH_FECS_MEM_TARGET_AS_VRAM                            0x00000002
+#define NV_PGRAPH_FECS_CHAN_ADDR                                       0x409b00
+#define NV_PGRAPH_FECS_CHAN_NEXT                                       0x409b04
+#define NV_PGRAPH_FECS_CHSW                                            0x409b0c
+#define NV_PGRAPH_FECS_CHSW_ACK                                      0x00000001
 #define NV_PGRAPH_FECS_INTR_UP_SET                                     0x409c1c
+#define NV_PGRAPH_FECS_INTR_UP_EN                                      0x409c24
 
+#define NV_PGRAPH_GPCX_GPCCS_INTR_ACK                                  0x41a004
+#define NV_PGRAPH_GPCX_GPCCS_INTR                                      0x41a008
+#define NV_PGRAPH_GPCX_GPCCS_INTR_FIFO                               0x00000004
+#define NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET                               0x41a010
+#define NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET_FIFO                        0x00000004
+#define NV_PGRAPH_GPCX_GPCCS_INTR_ROUTE                                0x41a01c
+#define NV_PGRAPH_GPCX_GPCCS_ACCESS                                    0x41a048
+#define NV_PGRAPH_GPCX_GPCCS_ACCESS_FIFO                             0x00000002
+#define NV_PGRAPH_GPCX_GPCCS_FIFO_DATA                                 0x41a064
+#define NV_PGRAPH_GPCX_GPCCS_FIFO_CMD                                  0x41a068
+#define NV_PGRAPH_GPCX_GPCCS_FIFO_ACK                                  0x41a074
+#define NV_PGRAPH_GPCX_GPCCS_UNITS                                     0x41a608
+#define NV_PGRAPH_GPCX_GPCCS_CAPS                                      0x41a108
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH                                0x41a614
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_UNK11                        0x00000800
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_ENABLE                       0x00000200
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_POWER                        0x00000020
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_PAUSE                        0x00000002
+#define NV_PGRAPH_GPCX_GPCCS_MYINDEX                                   0x41a618
+#define NV_PGRAPH_GPCX_GPCCS_MMCTX_SAVE_SWBASE                         0x41a700
+#define NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_SWBASE                         0x41a704
+#define NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_COUNT                          0x41a74c
 #if CHIPSET < GK110
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_VAL(n)              ((n) * 4 + 0x41a800)
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_SET(n)              ((n) * 4 + 0x41a820)
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_CLR(n)              ((n) * 4 + 0x41a840)
+#define NV_PGRAPH_GPCX_GPCCS_UNK86C                                    0x41a86c
 #else
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_VAL(n)              ((n) * 4 + 0x41a800)
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_CLR(n)              ((n) * 4 + 0x41a840)
+#define NV_PGRAPH_GPCX_GPCCS_UNK86C                                    0x41a88c
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_SET(n)              ((n) * 4 + 0x41a8c0)
 #endif
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_SELECT                             0x41a91c
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_CMD                                0x41a928
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_CMD_SAVE                         0x00000003
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_CMD_LOAD                         0x00000004
+#define NV_PGRAPH_GPCX_GPCCS_MEM_BASE                                  0x41aa04
 
 #define mmctx_data(r,c) .b32 (((c - 1) << 26) | r)
 #define queue_init      .skip 72 // (2 * 4) + ((8 * 4) * 2)
@@ -65,24 +179,50 @@
 #define T_LCHAN   8
 #define T_LCTXH   9
 
-#define nv_mkmm(rv,r) /*
-*/	movw rv  ((r) & 0x0000fffc) /*
-*/	sethi rv ((r) & 0x00ff0000)
+#if CHIPSET < GK208
+#define imm32(reg,val) /*
+*/	movw reg  ((val) & 0x0000ffff) /*
+*/	sethi reg ((val) & 0xffff0000)
+#else
+#define imm32(reg,val) /*
+*/	mov reg (val)
+#endif
+
 #define nv_mkio(rv,r,i) /*
-*/	nv_mkmm(rv, (((r) & 0xffc) << 6) | ((i) << 2))
+*/	imm32(rv, (((r) & 0xffc) << 6) | ((i) << 2))
+
+#define hash #
+#define fn(a) a
+#if CHIPSET < GK208
+#define call(a) call fn(hash)a
+#else
+#define call(a) lcall fn(hash)a
+#endif
 
 #define nv_iord(rv,r,i) /*
 */	nv_mkio(rv,r,i) /*
 */	iord rv I[rv]
+
 #define nv_iowr(r,i,rv) /*
 */	nv_mkio($r0,r,i) /*
 */	iowr I[$r0] rv /*
 */	clear b32 $r0
 
+#define nv_rd32(reg,addr) /*
+*/	imm32($r14, addr) /*
+*/	call(nv_rd32) /*
+*/	mov b32 reg $r15
+
+#define nv_wr32(addr,reg) /*
+*/	mov b32 $r15 reg /*
+*/	imm32($r14, addr) /*
+*/	call(nv_wr32)
+
 #define trace_set(bit) /*
 */	clear b32 $r9 /*
 */	bset $r9 bit /*
 */	nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_SET(7), 0, $r9)
+
 #define trace_clr(bit) /*
 */	clear b32 $r9 /*
 */	bset $r9 bit /*
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/os.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/os.h
index fd1d380de09..1718ae4e822 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/os.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/os.h
@@ -3,5 +3,6 @@
 
 #define E_BAD_COMMAND  0x00000001
 #define E_CMD_OVERFLOW 0x00000002
+#define E_BAD_FWMTHD   0x00000003
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/gk20a.c b/drivers/gpu/drm/nouveau/core/engine/graph/gk20a.c
new file mode 100644
index 00000000000..83048a56430
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/gk20a.c
@@ -0,0 +1,47 @@
+/*
+ * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include "nvc0.h"
+#include "ctxnvc0.h"
+
+static struct nouveau_oclass
+gk20a_graph_sclass[] = {
+	{ 0x902d, &nouveau_object_ofuncs },
+	{ 0xa040, &nouveau_object_ofuncs },
+	{ 0xa297, &nouveau_object_ofuncs },
+	{ 0xa0c0, &nouveau_object_ofuncs },
+	{}
+};
+
+struct nouveau_oclass *
+gk20a_graph_oclass = &(struct nvc0_graph_oclass) {
+	.base.handle = NV_ENGINE(GR, 0xea),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nvc0_graph_ctor,
+		.dtor = nvc0_graph_dtor,
+		.init = nve4_graph_init,
+		.fini = nve4_graph_fini,
+	},
+	.cclass = &gk20a_grctx_oclass,
+	.sclass = gk20a_graph_sclass,
+	.mmio = nve4_graph_pack_mmio,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/gm107.c b/drivers/gpu/drm/nouveau/core/engine/graph/gm107.c
new file mode 100644
index 00000000000..21c5f31d607
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/gm107.c
@@ -0,0 +1,465 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/P0260.h>
+
+#include "nvc0.h"
+#include "ctxnvc0.h"
+
+/*******************************************************************************
+ * Graphics object classes
+ ******************************************************************************/
+
+static struct nouveau_oclass
+gm107_graph_sclass[] = {
+	{ 0x902d, &nouveau_object_ofuncs },
+	{ 0xa140, &nouveau_object_ofuncs },
+	{ 0xb097, &nouveau_object_ofuncs },
+	{ 0xb0c0, &nouveau_object_ofuncs },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+gm107_graph_init_main_0[] = {
+	{ 0x400080,   1, 0x04, 0x003003c2 },
+	{ 0x400088,   1, 0x04, 0x0001bfe7 },
+	{ 0x40008c,   1, 0x04, 0x00060000 },
+	{ 0x400090,   1, 0x04, 0x00000030 },
+	{ 0x40013c,   1, 0x04, 0x003901f3 },
+	{ 0x400140,   1, 0x04, 0x00000100 },
+	{ 0x400144,   1, 0x04, 0x00000000 },
+	{ 0x400148,   1, 0x04, 0x00000110 },
+	{ 0x400138,   1, 0x04, 0x00000000 },
+	{ 0x400130,   2, 0x04, 0x00000000 },
+	{ 0x400124,   1, 0x04, 0x00000002 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_ds_0[] = {
+	{ 0x405844,   1, 0x04, 0x00ffffff },
+	{ 0x405850,   1, 0x04, 0x00000000 },
+	{ 0x405900,   1, 0x04, 0x00000000 },
+	{ 0x405908,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_scc_0[] = {
+	{ 0x40803c,   1, 0x04, 0x00000010 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_sked_0[] = {
+	{ 0x407010,   1, 0x04, 0x00000000 },
+	{ 0x407040,   1, 0x04, 0x40440424 },
+	{ 0x407048,   1, 0x04, 0x0000000a },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_prop_0[] = {
+	{ 0x418408,   1, 0x04, 0x00000000 },
+	{ 0x4184a0,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_setup_1[] = {
+	{ 0x4188c8,   2, 0x04, 0x00000000 },
+	{ 0x4188d0,   1, 0x04, 0x00010000 },
+	{ 0x4188d4,   1, 0x04, 0x00010201 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_zcull_0[] = {
+	{ 0x418910,   1, 0x04, 0x00010001 },
+	{ 0x418914,   1, 0x04, 0x00000301 },
+	{ 0x418918,   1, 0x04, 0x00800000 },
+	{ 0x418930,   2, 0x04, 0x00000000 },
+	{ 0x418980,   1, 0x04, 0x77777770 },
+	{ 0x418984,   3, 0x04, 0x77777777 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_gpc_unk_1[] = {
+	{ 0x418d00,   1, 0x04, 0x00000000 },
+	{ 0x418f00,   1, 0x04, 0x00000400 },
+	{ 0x418f08,   1, 0x04, 0x00000000 },
+	{ 0x418e08,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_tpccs_0[] = {
+	{ 0x419dc4,   1, 0x04, 0x00000000 },
+	{ 0x419dc8,   1, 0x04, 0x00000501 },
+	{ 0x419dd0,   1, 0x04, 0x00000000 },
+	{ 0x419dd4,   1, 0x04, 0x00000100 },
+	{ 0x419dd8,   1, 0x04, 0x00000001 },
+	{ 0x419ddc,   1, 0x04, 0x00000002 },
+	{ 0x419de0,   1, 0x04, 0x00000001 },
+	{ 0x419d0c,   1, 0x04, 0x00000000 },
+	{ 0x419d10,   1, 0x04, 0x00000014 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_tex_0[] = {
+	{ 0x419ab0,   1, 0x04, 0x00000000 },
+	{ 0x419ab8,   1, 0x04, 0x000000e7 },
+	{ 0x419abc,   1, 0x04, 0x00000000 },
+	{ 0x419acc,   1, 0x04, 0x000000ff },
+	{ 0x419ac0,   1, 0x04, 0x00000000 },
+	{ 0x419aa8,   2, 0x04, 0x00000000 },
+	{ 0x419ad0,   2, 0x04, 0x00000000 },
+	{ 0x419ae0,   2, 0x04, 0x00000000 },
+	{ 0x419af0,   4, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_pe_0[] = {
+	{ 0x419900,   1, 0x04, 0x000000ff },
+	{ 0x41980c,   1, 0x04, 0x00000010 },
+	{ 0x419844,   1, 0x04, 0x00000000 },
+	{ 0x419838,   1, 0x04, 0x000000ff },
+	{ 0x419850,   1, 0x04, 0x00000004 },
+	{ 0x419854,   2, 0x04, 0x00000000 },
+	{ 0x419894,   3, 0x04, 0x00100401 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_l1c_0[] = {
+	{ 0x419c98,   1, 0x04, 0x00000000 },
+	{ 0x419cc0,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_sm_0[] = {
+	{ 0x419e30,   1, 0x04, 0x000000ff },
+	{ 0x419e00,   1, 0x04, 0x00000000 },
+	{ 0x419ea0,   1, 0x04, 0x00000000 },
+	{ 0x419ee4,   1, 0x04, 0x00000000 },
+	{ 0x419ea4,   1, 0x04, 0x00000100 },
+	{ 0x419ea8,   1, 0x04, 0x01000000 },
+	{ 0x419ee8,   1, 0x04, 0x00000091 },
+	{ 0x419eb4,   1, 0x04, 0x00000000 },
+	{ 0x419ebc,   2, 0x04, 0x00000000 },
+	{ 0x419edc,   1, 0x04, 0x000c1810 },
+	{ 0x419ed8,   1, 0x04, 0x00000000 },
+	{ 0x419ee0,   1, 0x04, 0x00000000 },
+	{ 0x419f74,   1, 0x04, 0x00005155 },
+	{ 0x419f80,   4, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_l1c_1[] = {
+	{ 0x419ccc,   2, 0x04, 0x00000000 },
+	{ 0x419c80,   1, 0x04, 0x3f006022 },
+	{ 0x419c88,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_pes_0[] = {
+	{ 0x41be50,   1, 0x04, 0x000000ff },
+	{ 0x41be04,   1, 0x04, 0x00000000 },
+	{ 0x41be08,   1, 0x04, 0x00000004 },
+	{ 0x41be0c,   1, 0x04, 0x00000008 },
+	{ 0x41be10,   1, 0x04, 0x0e3b8bc7 },
+	{ 0x41be14,   2, 0x04, 0x00000000 },
+	{ 0x41be3c,   5, 0x04, 0x00100401 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_wwdx_0[] = {
+	{ 0x41bfd4,   1, 0x04, 0x00800000 },
+	{ 0x41bfdc,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_cbm_0[] = {
+	{ 0x41becc,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_be_0[] = {
+	{ 0x408890,   1, 0x04, 0x000000ff },
+	{ 0x40880c,   1, 0x04, 0x00000000 },
+	{ 0x408850,   1, 0x04, 0x00000004 },
+	{ 0x408878,   1, 0x04, 0x00c81603 },
+	{ 0x40887c,   1, 0x04, 0x80543432 },
+	{ 0x408880,   1, 0x04, 0x0010581e },
+	{ 0x408884,   1, 0x04, 0x00001205 },
+	{ 0x408974,   1, 0x04, 0x000000ff },
+	{ 0x408910,   9, 0x04, 0x00000000 },
+	{ 0x408950,   1, 0x04, 0x00000000 },
+	{ 0x408954,   1, 0x04, 0x0000ffff },
+	{ 0x408958,   1, 0x04, 0x00000034 },
+	{ 0x40895c,   1, 0x04, 0x8531a003 },
+	{ 0x408960,   1, 0x04, 0x0561985a },
+	{ 0x408964,   1, 0x04, 0x04e15c4f },
+	{ 0x408968,   1, 0x04, 0x02808833 },
+	{ 0x40896c,   1, 0x04, 0x01f02438 },
+	{ 0x408970,   1, 0x04, 0x00012c00 },
+	{ 0x408984,   1, 0x04, 0x00000000 },
+	{ 0x408988,   1, 0x04, 0x08040201 },
+	{ 0x40898c,   1, 0x04, 0x80402010 },
+	{}
+};
+
+static const struct nvc0_graph_init
+gm107_graph_init_sm_1[] = {
+	{ 0x419e5c,   1, 0x04, 0x00000000 },
+	{ 0x419e58,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+gm107_graph_pack_mmio[] = {
+	{ gm107_graph_init_main_0 },
+	{ nvf0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvc0_graph_init_pd_0 },
+	{ gm107_graph_init_ds_0 },
+	{ gm107_graph_init_scc_0 },
+	{ gm107_graph_init_sked_0 },
+	{ nvf0_graph_init_cwd_0 },
+	{ gm107_graph_init_prop_0 },
+	{ nv108_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ gm107_graph_init_setup_1 },
+	{ gm107_graph_init_zcull_0 },
+	{ nvc0_graph_init_gpm_0 },
+	{ gm107_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ gm107_graph_init_tpccs_0 },
+	{ gm107_graph_init_tex_0 },
+	{ gm107_graph_init_pe_0 },
+	{ gm107_graph_init_l1c_0 },
+	{ nvc0_graph_init_mpc_0 },
+	{ gm107_graph_init_sm_0 },
+	{ gm107_graph_init_l1c_1 },
+	{ gm107_graph_init_pes_0 },
+	{ gm107_graph_init_wwdx_0 },
+	{ gm107_graph_init_cbm_0 },
+	{ gm107_graph_init_be_0 },
+	{ gm107_graph_init_sm_1 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
+static void
+gm107_graph_init_bios(struct nvc0_graph_priv *priv)
+{
+	static const struct {
+		u32 ctrl;
+		u32 data;
+	} regs[] = {
+		{ 0x419ed8, 0x419ee0 },
+		{ 0x419ad0, 0x419ad4 },
+		{ 0x419ae0, 0x419ae4 },
+		{ 0x419af0, 0x419af4 },
+		{ 0x419af8, 0x419afc },
+	};
+	struct nouveau_bios *bios = nouveau_bios(priv);
+	struct nvbios_P0260E infoE;
+	struct nvbios_P0260X infoX;
+	int E = -1, X;
+	u8 ver, hdr;
+
+	while (nvbios_P0260Ep(bios, ++E, &ver, &hdr, &infoE)) {
+		if (X = -1, E < ARRAY_SIZE(regs)) {
+			nv_wr32(priv, regs[E].ctrl, infoE.data);
+			while (nvbios_P0260Xp(bios, ++X, &ver, &hdr, &infoX))
+				nv_wr32(priv, regs[E].data, infoX.data);
+		}
+	}
+}
+
+int
+gm107_graph_init(struct nouveau_object *object)
+{
+	struct nvc0_graph_oclass *oclass = (void *)object->oclass;
+	struct nvc0_graph_priv *priv = (void *)object;
+	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, priv->tpc_total);
+	u32 data[TPC_MAX / 8] = {};
+	u8  tpcnr[GPC_MAX];
+	int gpc, tpc, ppc, rop;
+	int ret, i;
+
+	ret = nouveau_graph_init(&priv->base);
+	if (ret)
+		return ret;
+
+	nv_wr32(priv, GPC_BCAST(0x0880), 0x00000000);
+	nv_wr32(priv, GPC_BCAST(0x0890), 0x00000000);
+	nv_wr32(priv, GPC_BCAST(0x0894), 0x00000000);
+	nv_wr32(priv, GPC_BCAST(0x08b4), priv->unk4188b4->addr >> 8);
+	nv_wr32(priv, GPC_BCAST(0x08b8), priv->unk4188b8->addr >> 8);
+
+	nvc0_graph_mmio(priv, oclass->mmio);
+
+	gm107_graph_init_bios(priv);
+
+	nv_wr32(priv, GPC_UNIT(0, 0x3018), 0x00000001);
+
+	memset(data, 0x00, sizeof(data));
+	memcpy(tpcnr, priv->tpc_nr, sizeof(priv->tpc_nr));
+	for (i = 0, gpc = -1; i < priv->tpc_total; i++) {
+		do {
+			gpc = (gpc + 1) % priv->gpc_nr;
+		} while (!tpcnr[gpc]);
+		tpc = priv->tpc_nr[gpc] - tpcnr[gpc]--;
+
+		data[i / 8] |= tpc << ((i % 8) * 4);
+	}
+
+	nv_wr32(priv, GPC_BCAST(0x0980), data[0]);
+	nv_wr32(priv, GPC_BCAST(0x0984), data[1]);
+	nv_wr32(priv, GPC_BCAST(0x0988), data[2]);
+	nv_wr32(priv, GPC_BCAST(0x098c), data[3]);
+
+	for (gpc = 0; gpc < priv->gpc_nr; gpc++) {
+		nv_wr32(priv, GPC_UNIT(gpc, 0x0914),
+			priv->magic_not_rop_nr << 8 | priv->tpc_nr[gpc]);
+		nv_wr32(priv, GPC_UNIT(gpc, 0x0910), 0x00040000 |
+			priv->tpc_total);
+		nv_wr32(priv, GPC_UNIT(gpc, 0x0918), magicgpc918);
+	}
+
+	nv_wr32(priv, GPC_BCAST(0x3fd4), magicgpc918);
+	nv_wr32(priv, GPC_BCAST(0x08ac), nv_rd32(priv, 0x100800));
+
+	nv_wr32(priv, 0x400500, 0x00010001);
+
+	nv_wr32(priv, 0x400100, 0xffffffff);
+	nv_wr32(priv, 0x40013c, 0xffffffff);
+	nv_wr32(priv, 0x400124, 0x00000002);
+	nv_wr32(priv, 0x409c24, 0x000e0000);
+
+	nv_wr32(priv, 0x404000, 0xc0000000);
+	nv_wr32(priv, 0x404600, 0xc0000000);
+	nv_wr32(priv, 0x408030, 0xc0000000);
+	nv_wr32(priv, 0x404490, 0xc0000000);
+	nv_wr32(priv, 0x406018, 0xc0000000);
+	nv_wr32(priv, 0x407020, 0x40000000);
+	nv_wr32(priv, 0x405840, 0xc0000000);
+	nv_wr32(priv, 0x405844, 0x00ffffff);
+	nv_mask(priv, 0x419cc0, 0x00000008, 0x00000008);
+
+	for (gpc = 0; gpc < priv->gpc_nr; gpc++) {
+		for (ppc = 0; ppc < 2 /* priv->ppc_nr[gpc] */; ppc++)
+			nv_wr32(priv, PPC_UNIT(gpc, ppc, 0x038), 0xc0000000);
+		nv_wr32(priv, GPC_UNIT(gpc, 0x0420), 0xc0000000);
+		nv_wr32(priv, GPC_UNIT(gpc, 0x0900), 0xc0000000);
+		nv_wr32(priv, GPC_UNIT(gpc, 0x1028), 0xc0000000);
+		nv_wr32(priv, GPC_UNIT(gpc, 0x0824), 0xc0000000);
+		for (tpc = 0; tpc < priv->tpc_nr[gpc]; tpc++) {
+			nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
+			nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
+			nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
+			nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
+			nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
+			nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
+			nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
+			nv_wr32(priv, TPC_UNIT(gpc, tpc, 0x64c), 0x00000005);
+		}
+		nv_wr32(priv, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
+		nv_wr32(priv, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
+	}
+
+	for (rop = 0; rop < priv->rop_nr; rop++) {
+		nv_wr32(priv, ROP_UNIT(rop, 0x144), 0x40000000);
+		nv_wr32(priv, ROP_UNIT(rop, 0x070), 0x40000000);
+		nv_wr32(priv, ROP_UNIT(rop, 0x204), 0xffffffff);
+		nv_wr32(priv, ROP_UNIT(rop, 0x208), 0xffffffff);
+	}
+
+	nv_wr32(priv, 0x400108, 0xffffffff);
+	nv_wr32(priv, 0x400138, 0xffffffff);
+	nv_wr32(priv, 0x400118, 0xffffffff);
+	nv_wr32(priv, 0x400130, 0xffffffff);
+	nv_wr32(priv, 0x40011c, 0xffffffff);
+	nv_wr32(priv, 0x400134, 0xffffffff);
+
+	nv_wr32(priv, 0x400054, 0x2c350f63);
+	return nvc0_graph_init_ctxctl(priv);
+}
+
+#include "fuc/hubgm107.fuc5.h"
+
+static struct nvc0_graph_ucode
+gm107_graph_fecs_ucode = {
+	.code.data = gm107_grhub_code,
+	.code.size = sizeof(gm107_grhub_code),
+	.data.data = gm107_grhub_data,
+	.data.size = sizeof(gm107_grhub_data),
+};
+
+#include "fuc/gpcgm107.fuc5.h"
+
+static struct nvc0_graph_ucode
+gm107_graph_gpccs_ucode = {
+	.code.data = gm107_grgpc_code,
+	.code.size = sizeof(gm107_grgpc_code),
+	.data.data = gm107_grgpc_data,
+	.data.size = sizeof(gm107_grgpc_data),
+};
+
+struct nouveau_oclass *
+gm107_graph_oclass = &(struct nvc0_graph_oclass) {
+	.base.handle = NV_ENGINE(GR, 0x07),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nvc0_graph_ctor,
+		.dtor = nvc0_graph_dtor,
+		.init = gm107_graph_init,
+		.fini = _nouveau_graph_fini,
+	},
+	.cclass = &gm107_grctx_oclass,
+	.sclass =  gm107_graph_sclass,
+	.mmio = gm107_graph_pack_mmio,
+	.fecs.ucode = 0 ? &gm107_graph_fecs_ucode : NULL,
+	.gpccs.ucode = &gm107_graph_gpccs_ucode,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nv108.c b/drivers/gpu/drm/nouveau/core/engine/graph/nv108.c
new file mode 100644
index 00000000000..00ea1a08982
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nv108.c
@@ -0,0 +1,223 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include "nvc0.h"
+#include "ctxnvc0.h"
+
+/*******************************************************************************
+ * Graphics object classes
+ ******************************************************************************/
+
+static struct nouveau_oclass
+nv108_graph_sclass[] = {
+	{ 0x902d, &nouveau_object_ofuncs },
+	{ 0xa140, &nouveau_object_ofuncs },
+	{ 0xa197, &nouveau_object_ofuncs },
+	{ 0xa1c0, &nouveau_object_ofuncs },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+nv108_graph_init_main_0[] = {
+	{ 0x400080,   1, 0x04, 0x003083c2 },
+	{ 0x400088,   1, 0x04, 0x0001bfe7 },
+	{ 0x40008c,   1, 0x04, 0x00000000 },
+	{ 0x400090,   1, 0x04, 0x00000030 },
+	{ 0x40013c,   1, 0x04, 0x003901f7 },
+	{ 0x400140,   1, 0x04, 0x00000100 },
+	{ 0x400144,   1, 0x04, 0x00000000 },
+	{ 0x400148,   1, 0x04, 0x00000110 },
+	{ 0x400138,   1, 0x04, 0x00000000 },
+	{ 0x400130,   2, 0x04, 0x00000000 },
+	{ 0x400124,   1, 0x04, 0x00000002 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_graph_init_ds_0[] = {
+	{ 0x405844,   1, 0x04, 0x00ffffff },
+	{ 0x405850,   1, 0x04, 0x00000000 },
+	{ 0x405900,   1, 0x04, 0x00000000 },
+	{ 0x405908,   1, 0x04, 0x00000000 },
+	{ 0x405928,   2, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nv108_graph_init_gpc_unk_0[] = {
+	{ 0x418604,   1, 0x04, 0x00000000 },
+	{ 0x418680,   1, 0x04, 0x00000000 },
+	{ 0x418714,   1, 0x04, 0x00000000 },
+	{ 0x418384,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_graph_init_setup_1[] = {
+	{ 0x4188c8,   2, 0x04, 0x00000000 },
+	{ 0x4188d0,   1, 0x04, 0x00010000 },
+	{ 0x4188d4,   1, 0x04, 0x00000201 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_graph_init_tex_0[] = {
+	{ 0x419ab0,   1, 0x04, 0x00000000 },
+	{ 0x419ac8,   1, 0x04, 0x00000000 },
+	{ 0x419ab8,   1, 0x04, 0x000000e7 },
+	{ 0x419abc,   2, 0x04, 0x00000000 },
+	{ 0x419ab4,   1, 0x04, 0x00000000 },
+	{ 0x419aa8,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nv108_graph_init_l1c_0[] = {
+	{ 0x419c98,   1, 0x04, 0x00000000 },
+	{ 0x419ca8,   1, 0x04, 0x00000000 },
+	{ 0x419cb0,   1, 0x04, 0x01000000 },
+	{ 0x419cb4,   1, 0x04, 0x00000000 },
+	{ 0x419cb8,   1, 0x04, 0x00b08bea },
+	{ 0x419c84,   1, 0x04, 0x00010384 },
+	{ 0x419cbc,   1, 0x04, 0x281b3646 },
+	{ 0x419cc0,   2, 0x04, 0x00000000 },
+	{ 0x419c80,   1, 0x04, 0x00000230 },
+	{ 0x419ccc,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+nv108_graph_pack_mmio[] = {
+	{ nv108_graph_init_main_0 },
+	{ nvf0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvd9_graph_init_pd_0 },
+	{ nv108_graph_init_ds_0 },
+	{ nvc0_graph_init_scc_0 },
+	{ nvf0_graph_init_sked_0 },
+	{ nvf0_graph_init_cwd_0 },
+	{ nvd9_graph_init_prop_0 },
+	{ nv108_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ nv108_graph_init_setup_1 },
+	{ nvc0_graph_init_zcull_0 },
+	{ nvd9_graph_init_gpm_0 },
+	{ nvf0_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ nve4_graph_init_tpccs_0 },
+	{ nv108_graph_init_tex_0 },
+	{ nve4_graph_init_pe_0 },
+	{ nv108_graph_init_l1c_0 },
+	{ nvc0_graph_init_mpc_0 },
+	{ nvf0_graph_init_sm_0 },
+	{ nvd7_graph_init_pes_0 },
+	{ nvd7_graph_init_wwdx_0 },
+	{ nvd7_graph_init_cbm_0 },
+	{ nve4_graph_init_be_0 },
+	{ nvc0_graph_init_fe_1 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
+static int
+nv108_graph_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nvc0_graph_priv *priv = (void *)object;
+	static const struct {
+		u32 addr;
+		u32 data;
+	} magic[] = {
+		{ 0x020520, 0xfffffffc },
+		{ 0x020524, 0xfffffffe },
+		{ 0x020524, 0xfffffffc },
+		{ 0x020524, 0xfffffff8 },
+		{ 0x020524, 0xffffffe0 },
+		{ 0x020530, 0xfffffffe },
+		{ 0x02052c, 0xfffffffa },
+		{ 0x02052c, 0xfffffff0 },
+		{ 0x02052c, 0xffffffc0 },
+		{ 0x02052c, 0xffffff00 },
+		{ 0x02052c, 0xfffffc00 },
+		{ 0x02052c, 0xfffcfc00 },
+		{ 0x02052c, 0xfff0fc00 },
+		{ 0x02052c, 0xff80fc00 },
+		{ 0x020528, 0xfffffffe },
+		{ 0x020528, 0xfffffffc },
+	};
+	int i;
+
+	nv_mask(priv, 0x000200, 0x08001000, 0x00000000);
+	nv_mask(priv, 0x0206b4, 0x00000000, 0x00000000);
+	for (i = 0; i < ARRAY_SIZE(magic); i++) {
+		nv_wr32(priv, magic[i].addr, magic[i].data);
+		nv_wait(priv, magic[i].addr, 0x80000000, 0x00000000);
+	}
+
+	return nouveau_graph_fini(&priv->base, suspend);
+}
+
+#include "fuc/hubnv108.fuc5.h"
+
+static struct nvc0_graph_ucode
+nv108_graph_fecs_ucode = {
+	.code.data = nv108_grhub_code,
+	.code.size = sizeof(nv108_grhub_code),
+	.data.data = nv108_grhub_data,
+	.data.size = sizeof(nv108_grhub_data),
+};
+
+#include "fuc/gpcnv108.fuc5.h"
+
+static struct nvc0_graph_ucode
+nv108_graph_gpccs_ucode = {
+	.code.data = nv108_grgpc_code,
+	.code.size = sizeof(nv108_grgpc_code),
+	.data.data = nv108_grgpc_data,
+	.data.size = sizeof(nv108_grgpc_data),
+};
+
+struct nouveau_oclass *
+nv108_graph_oclass = &(struct nvc0_graph_oclass) {
+	.base.handle = NV_ENGINE(GR, 0x08),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nvc0_graph_ctor,
+		.dtor = nvc0_graph_dtor,
+		.init = nve4_graph_init,
+		.fini = nv108_graph_fini,
+	},
+	.cclass = &nv108_grctx_oclass,
+	.sclass =  nv108_graph_sclass,
+	.mmio = nv108_graph_pack_mmio,
+	.fecs.ucode = &nv108_graph_fecs_ucode,
+	.gpccs.ucode = &nv108_graph_gpccs_ucode,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nv20.c b/drivers/gpu/drm/nouveau/core/engine/graph/nv20.c
index b2455931590..d145e080899 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nv20.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nv20.c
@@ -349,7 +349,7 @@ nv20_graph_init(struct nouveau_object *object)
 	nv_wr32(priv, NV10_PGRAPH_SURFACE, tmp);
 
 	/* begin RAM config */
-	vramsz = pci_resource_len(nv_device(priv)->pdev, 0) - 1;
+	vramsz = nv_device_resource_len(nv_device(priv), 0) - 1;
 	nv_wr32(priv, 0x4009A4, nv_rd32(priv, 0x100200));
 	nv_wr32(priv, 0x4009A8, nv_rd32(priv, 0x100204));
 	nv_wr32(priv, NV10_PGRAPH_RDI_INDEX, 0x00EA0000);
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nv40.c b/drivers/gpu/drm/nouveau/core/engine/graph/nv40.c
index 193a5de1b48..6477fbf6a55 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nv40.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nv40.c
@@ -484,7 +484,7 @@ nv40_graph_init(struct nouveau_object *object)
 		engine->tile_prog(engine, i);
 
 	/* begin RAM config */
-	vramsz = pci_resource_len(nv_device(priv)->pdev, 0) - 1;
+	vramsz = nv_device_resource_len(nv_device(priv), 0) - 1;
 	switch (nv_device(priv)->chipset) {
 	case 0x40:
 		nv_wr32(priv, 0x4009A4, nv_rd32(priv, 0x100200));
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c b/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c
index 03de5175dd9..20665c21d80 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c
@@ -197,34 +197,35 @@ static const struct nouveau_bitfield nv50_pgraph_status[] = {
 	{ 0x00000080, "UNK7" },
 	{ 0x00000100, "CTXPROG" },
 	{ 0x00000200, "VFETCH" },
-	{ 0x00000400, "CCACHE_UNK4" },
-	{ 0x00000800, "STRMOUT_GSCHED_UNK5" },
-	{ 0x00001000, "UNK14XX" },
-	{ 0x00002000, "UNK24XX_CSCHED" },
-	{ 0x00004000, "UNK1CXX" },
+	{ 0x00000400, "CCACHE_PREGEOM" },
+	{ 0x00000800, "STRMOUT_VATTR_POSTGEOM" },
+	{ 0x00001000, "VCLIP" },
+	{ 0x00002000, "RATTR_APLANE" },
+	{ 0x00004000, "TRAST" },
 	{ 0x00008000, "CLIPID" },
 	{ 0x00010000, "ZCULL" },
 	{ 0x00020000, "ENG2D" },
-	{ 0x00040000, "UNK34XX" },
-	{ 0x00080000, "TPRAST" },
-	{ 0x00100000, "TPROP" },
-	{ 0x00200000, "TEX" },
-	{ 0x00400000, "TPVP" },
-	{ 0x00800000, "MP" },
+	{ 0x00040000, "RMASK" },
+	{ 0x00080000, "TPC_RAST" },
+	{ 0x00100000, "TPC_PROP" },
+	{ 0x00200000, "TPC_TEX" },
+	{ 0x00400000, "TPC_GEOM" },
+	{ 0x00800000, "TPC_MP" },
 	{ 0x01000000, "ROP" },
 	{}
 };
 
 static const char *const nv50_pgraph_vstatus_0[] = {
-	"VFETCH", "CCACHE", "UNK4", "UNK5", "GSCHED", "STRMOUT", "UNK14XX", NULL
+	"VFETCH", "CCACHE", "PREGEOM", "POSTGEOM", "VATTR", "STRMOUT", "VCLIP",
+	NULL
 };
 
 static const char *const nv50_pgraph_vstatus_1[] = {
-	"TPRAST", "TPROP", "TEXTURE", "TPVP", "MP", NULL
+	"TPC_RAST", "TPC_PROP", "TPC_TEX", "TPC_GEOM", "TPC_MP", NULL
 };
 
 static const char *const nv50_pgraph_vstatus_2[] = {
-	"UNK24XX", "CSCHED", "UNK1CXX", "CLIPID", "ZCULL", "ENG2D", "UNK34XX",
+	"RATTR", "APLANE", "TRAST", "CLIPID", "ZCULL", "ENG2D", "RMASK",
 	"ROP", NULL
 };
 
@@ -304,12 +305,37 @@ nv84_graph_tlb_flush(struct nouveau_engine *engine)
 	return timeout ? -EBUSY : 0;
 }
 
-static const struct nouveau_enum nv50_mp_exec_error_names[] = {
-	{ 3, "STACK_UNDERFLOW", NULL },
-	{ 4, "QUADON_ACTIVE", NULL },
-	{ 8, "TIMEOUT", NULL },
-	{ 0x10, "INVALID_OPCODE", NULL },
-	{ 0x40, "BREAKPOINT", NULL },
+static const struct nouveau_bitfield nv50_mp_exec_errors[] = {
+	{ 0x01, "STACK_UNDERFLOW" },
+	{ 0x02, "STACK_MISMATCH" },
+	{ 0x04, "QUADON_ACTIVE" },
+	{ 0x08, "TIMEOUT" },
+	{ 0x10, "INVALID_OPCODE" },
+	{ 0x20, "PM_OVERFLOW" },
+	{ 0x40, "BREAKPOINT" },
+	{}
+};
+
+static const struct nouveau_bitfield nv50_mpc_traps[] = {
+	{ 0x0000001, "LOCAL_LIMIT_READ" },
+	{ 0x0000010, "LOCAL_LIMIT_WRITE" },
+	{ 0x0000040, "STACK_LIMIT" },
+	{ 0x0000100, "GLOBAL_LIMIT_READ" },
+	{ 0x0001000, "GLOBAL_LIMIT_WRITE" },
+	{ 0x0010000, "MP0" },
+	{ 0x0020000, "MP1" },
+	{ 0x0040000, "GLOBAL_LIMIT_RED" },
+	{ 0x0400000, "GLOBAL_LIMIT_ATOM" },
+	{ 0x4000000, "MP2" },
+	{}
+};
+
+static const struct nouveau_bitfield nv50_tex_traps[] = {
+	{ 0x00000001, "" }, /* any bit set? */
+	{ 0x00000002, "FAULT" },
+	{ 0x00000004, "STORAGE_TYPE_MISMATCH" },
+	{ 0x00000008, "LINEAR_MISMATCH" },
+	{ 0x00000020, "WRONG_MEMTYPE" },
 	{}
 };
 
@@ -396,6 +422,60 @@ static const struct nouveau_bitfield nv50_graph_intr_name[] = {
 	{}
 };
 
+static const struct nouveau_bitfield nv50_graph_trap_prop[] = {
+	{ 0x00000004, "SURF_WIDTH_OVERRUN" },
+	{ 0x00000008, "SURF_HEIGHT_OVERRUN" },
+	{ 0x00000010, "DST2D_FAULT" },
+	{ 0x00000020, "ZETA_FAULT" },
+	{ 0x00000040, "RT_FAULT" },
+	{ 0x00000080, "CUDA_FAULT" },
+	{ 0x00000100, "DST2D_STORAGE_TYPE_MISMATCH" },
+	{ 0x00000200, "ZETA_STORAGE_TYPE_MISMATCH" },
+	{ 0x00000400, "RT_STORAGE_TYPE_MISMATCH" },
+	{ 0x00000800, "DST2D_LINEAR_MISMATCH" },
+	{ 0x00001000, "RT_LINEAR_MISMATCH" },
+	{}
+};
+
+static void
+nv50_priv_prop_trap(struct nv50_graph_priv *priv,
+		    u32 ustatus_addr, u32 ustatus, u32 tp)
+{
+	u32 e0c = nv_rd32(priv, ustatus_addr + 0x04);
+	u32 e10 = nv_rd32(priv, ustatus_addr + 0x08);
+	u32 e14 = nv_rd32(priv, ustatus_addr + 0x0c);
+	u32 e18 = nv_rd32(priv, ustatus_addr + 0x10);
+	u32 e1c = nv_rd32(priv, ustatus_addr + 0x14);
+	u32 e20 = nv_rd32(priv, ustatus_addr + 0x18);
+	u32 e24 = nv_rd32(priv, ustatus_addr + 0x1c);
+
+	/* CUDA memory: l[], g[] or stack. */
+	if (ustatus & 0x00000080) {
+		if (e18 & 0x80000000) {
+			/* g[] read fault? */
+			nv_error(priv, "TRAP_PROP - TP %d - CUDA_FAULT - Global read fault at address %02x%08x\n",
+					 tp, e14, e10 | ((e18 >> 24) & 0x1f));
+			e18 &= ~0x1f000000;
+		} else if (e18 & 0xc) {
+			/* g[] write fault? */
+			nv_error(priv, "TRAP_PROP - TP %d - CUDA_FAULT - Global write fault at address %02x%08x\n",
+				 tp, e14, e10 | ((e18 >> 7) & 0x1f));
+			e18 &= ~0x00000f80;
+		} else {
+			nv_error(priv, "TRAP_PROP - TP %d - Unknown CUDA fault at address %02x%08x\n",
+				 tp, e14, e10);
+		}
+		ustatus &= ~0x00000080;
+	}
+	if (ustatus) {
+		nv_error(priv, "TRAP_PROP - TP %d -", tp);
+		nouveau_bitfield_print(nv50_graph_trap_prop, ustatus);
+		pr_cont(" - Address %02x%08x\n", e14, e10);
+	}
+	nv_error(priv, "TRAP_PROP - TP %d - e0c: %08x, e18: %08x, e1c: %08x, e20: %08x, e24: %08x\n",
+		 tp, e0c, e18, e1c, e20, e24);
+}
+
 static void
 nv50_priv_mp_trap(struct nv50_graph_priv *priv, int tpid, int display)
 {
@@ -420,8 +500,8 @@ nv50_priv_mp_trap(struct nv50_graph_priv *priv, int tpid, int display)
 			oplow = nv_rd32(priv, addr + 0x70);
 			ophigh = nv_rd32(priv, addr + 0x74);
 			nv_error(priv, "TRAP_MP_EXEC - "
-					"TP %d MP %d: ", tpid, i);
-			nouveau_enum_print(nv50_mp_exec_error_names, status);
+					"TP %d MP %d:", tpid, i);
+			nouveau_bitfield_print(nv50_mp_exec_errors, status);
 			pr_cont(" at %06x warp %d, opcode %08x %08x\n",
 					pc&0xffffff, pc >> 24,
 					oplow, ophigh);
@@ -461,6 +541,13 @@ nv50_priv_tp_trap(struct nv50_graph_priv *priv, int type, u32 ustatus_old,
 				for (r = ustatus_addr + 4; r <= ustatus_addr + 0x10; r += 4)
 					nv_error(priv, "\t0x%08x: 0x%08x\n", r,
 						nv_rd32(priv, r));
+				if (ustatus) {
+					nv_error(priv, "%s - TP%d:", name, i);
+					nouveau_bitfield_print(nv50_tex_traps,
+							       ustatus);
+					pr_cont("\n");
+					ustatus = 0;
+				}
 			}
 			break;
 		case 7: /* MP error */
@@ -468,60 +555,19 @@ nv50_priv_tp_trap(struct nv50_graph_priv *priv, int type, u32 ustatus_old,
 				nv50_priv_mp_trap(priv, i, display);
 				ustatus &= ~0x04030000;
 			}
-			break;
-		case 8: /* TPDMA error */
-			{
-			u32 e0c = nv_rd32(priv, ustatus_addr + 4);
-			u32 e10 = nv_rd32(priv, ustatus_addr + 8);
-			u32 e14 = nv_rd32(priv, ustatus_addr + 0xc);
-			u32 e18 = nv_rd32(priv, ustatus_addr + 0x10);
-			u32 e1c = nv_rd32(priv, ustatus_addr + 0x14);
-			u32 e20 = nv_rd32(priv, ustatus_addr + 0x18);
-			u32 e24 = nv_rd32(priv, ustatus_addr + 0x1c);
-			/* 2d engine destination */
-			if (ustatus & 0x00000010) {
-				if (display) {
-					nv_error(priv, "TRAP_TPDMA_2D - TP %d - Unknown fault at address %02x%08x\n",
-							i, e14, e10);
-					nv_error(priv, "TRAP_TPDMA_2D - TP %d - e0c: %08x, e18: %08x, e1c: %08x, e20: %08x, e24: %08x\n",
-							i, e0c, e18, e1c, e20, e24);
-				}
-				ustatus &= ~0x00000010;
-			}
-			/* Render target */
-			if (ustatus & 0x00000040) {
-				if (display) {
-					nv_error(priv, "TRAP_TPDMA_RT - TP %d - Unknown fault at address %02x%08x\n",
-							i, e14, e10);
-					nv_error(priv, "TRAP_TPDMA_RT - TP %d - e0c: %08x, e18: %08x, e1c: %08x, e20: %08x, e24: %08x\n",
-							i, e0c, e18, e1c, e20, e24);
-				}
-				ustatus &= ~0x00000040;
-			}
-			/* CUDA memory: l[], g[] or stack. */
-			if (ustatus & 0x00000080) {
-				if (display) {
-					if (e18 & 0x80000000) {
-						/* g[] read fault? */
-						nv_error(priv, "TRAP_TPDMA - TP %d - Global read fault at address %02x%08x\n",
-								i, e14, e10 | ((e18 >> 24) & 0x1f));
-						e18 &= ~0x1f000000;
-					} else if (e18 & 0xc) {
-						/* g[] write fault? */
-						nv_error(priv, "TRAP_TPDMA - TP %d - Global write fault at address %02x%08x\n",
-								i, e14, e10 | ((e18 >> 7) & 0x1f));
-						e18 &= ~0x00000f80;
-					} else {
-						nv_error(priv, "TRAP_TPDMA - TP %d - Unknown CUDA fault at address %02x%08x\n",
-								i, e14, e10);
-					}
-					nv_error(priv, "TRAP_TPDMA - TP %d - e0c: %08x, e18: %08x, e1c: %08x, e20: %08x, e24: %08x\n",
-							i, e0c, e18, e1c, e20, e24);
-				}
-				ustatus &= ~0x00000080;
-			}
+			if (ustatus && display) {
+				nv_error(priv, "%s - TP%d:", name, i);
+				nouveau_bitfield_print(nv50_mpc_traps, ustatus);
+				pr_cont("\n");
+				ustatus = 0;
 			}
 			break;
+		case 8: /* PROP error */
+			if (display)
+				nv50_priv_prop_trap(
+						priv, ustatus_addr, ustatus, i);
+			ustatus = 0;
+			break;
 		}
 		if (ustatus) {
 			if (display)
@@ -727,11 +773,11 @@ nv50_graph_trap_handler(struct nv50_graph_priv *priv, u32 display,
 		status &= ~0x080;
 	}
 
-	/* TPDMA:  Handles TP-initiated uncached memory accesses:
+	/* PROP:  Handles TP-initiated uncached memory accesses:
 	 * l[], g[], stack, 2d surfaces, render targets. */
 	if (status & 0x100) {
 		nv50_priv_tp_trap(priv, 8, 0x408e08, 0x408708, display,
-				    "TRAP_TPDMA");
+				    "TRAP_PROP");
 		nv_wr32(priv, 0x400108, 0x100);
 		status &= ~0x100;
 	}
@@ -760,7 +806,7 @@ nv50_graph_intr(struct nouveau_subdev *subdev)
 	u32 mthd = (addr & 0x00001ffc);
 	u32 data = nv_rd32(priv, 0x400708);
 	u32 class = nv_rd32(priv, 0x400814);
-	u32 show = stat;
+	u32 show = stat, show_bitfield = stat;
 	int chid;
 
 	engctx = nouveau_engctx_get(engine, inst);
@@ -778,21 +824,26 @@ nv50_graph_intr(struct nouveau_subdev *subdev)
 		nv_error(priv, "DATA_ERROR ");
 		nouveau_enum_print(nv50_data_error_names, ecode);
 		pr_cont("\n");
+		show_bitfield &= ~0x00100000;
 	}
 
 	if (stat & 0x00200000) {
 		if (!nv50_graph_trap_handler(priv, show, chid, (u64)inst << 12,
 				engctx))
 			show &= ~0x00200000;
+		show_bitfield &= ~0x00200000;
 	}
 
 	nv_wr32(priv, 0x400100, stat);
 	nv_wr32(priv, 0x400500, 0x00010001);
 
 	if (show) {
-		nv_error(priv, "%s", "");
-		nouveau_bitfield_print(nv50_graph_intr_name, show);
-		pr_cont("\n");
+		show &= show_bitfield;
+		if (show) {
+			nv_error(priv, "%s", "");
+			nouveau_bitfield_print(nv50_graph_intr_name, show);
+			pr_cont("\n");
+		}
 		nv_error(priv,
 			 "ch %d [0x%010llx %s] subc %d class 0x%04x mthd 0x%04x data 0x%08x\n",
 			 chid, (u64)inst << 12, nouveau_client_name(engctx),
@@ -850,7 +901,7 @@ nv50_graph_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		nv_engine(priv)->sclass = nvaf_graph_sclass;
 		break;
 
-	};
+	}
 
 	/* unfortunate hw bug workaround... */
 	if (nv_device(priv)->chipset != 0x50 &&
@@ -925,7 +976,6 @@ nv50_graph_init(struct nouveau_object *object)
 		break;
 	case 0xa0:
 	default:
-		nv_wr32(priv, 0x402cc0, 0x00000000);
 		if (nv_device(priv)->chipset == 0xa0 ||
 		    nv_device(priv)->chipset == 0xaa ||
 		    nv_device(priv)->chipset == 0xac) {
@@ -940,10 +990,10 @@ nv50_graph_init(struct nouveau_object *object)
 
 	/* zero out zcull regions */
 	for (i = 0; i < 8; i++) {
-		nv_wr32(priv, 0x402c20 + (i * 8), 0x00000000);
-		nv_wr32(priv, 0x402c24 + (i * 8), 0x00000000);
-		nv_wr32(priv, 0x402c28 + (i * 8), 0x00000000);
-		nv_wr32(priv, 0x402c2c + (i * 8), 0x00000000);
+		nv_wr32(priv, 0x402c20 + (i * 0x10), 0x00000000);
+		nv_wr32(priv, 0x402c24 + (i * 0x10), 0x00000000);
+		nv_wr32(priv, 0x402c28 + (i * 0x10), 0x00000000);
+		nv_wr32(priv, 0x402c2c + (i * 0x10), 0x00000000);
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c
index 5c8a63dc506..aa083891635 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c
@@ -23,6 +23,7 @@
  */
 
 #include "nvc0.h"
+#include "ctxnvc0.h"
 
 /*******************************************************************************
  * Graphics object classes
@@ -146,11 +147,11 @@ nvc0_graph_context_dtor(struct nouveau_object *object)
 }
 
 /*******************************************************************************
- * PGRAPH engine/subdev functions
+ * PGRAPH register lists
  ******************************************************************************/
 
-struct nvc0_graph_init
-nvc0_graph_init_regs[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_main_0[] = {
 	{ 0x400080,   1, 0x04, 0x003083c2 },
 	{ 0x400088,   1, 0x04, 0x00006fe7 },
 	{ 0x40008c,   1, 0x04, 0x00000000 },
@@ -165,95 +166,170 @@ nvc0_graph_init_regs[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_graph_init_unk40xx[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_fe_0[] = {
 	{ 0x40415c,   1, 0x04, 0x00000000 },
 	{ 0x404170,   1, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_graph_init_unk44xx[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_pri_0[] = {
 	{ 0x404488,   2, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_graph_init_unk78xx[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_rstr2d_0[] = {
 	{ 0x407808,   1, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_graph_init_unk60xx[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_pd_0[] = {
 	{ 0x406024,   1, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_graph_init_unk58xx[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_ds_0[] = {
 	{ 0x405844,   1, 0x04, 0x00ffffff },
 	{ 0x405850,   1, 0x04, 0x00000000 },
 	{ 0x405908,   1, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_graph_init_unk80xx[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_scc_0[] = {
 	{ 0x40803c,   1, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_graph_init_gpc[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_prop_0[] = {
 	{ 0x4184a0,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_gpc_unk_0[] = {
 	{ 0x418604,   1, 0x04, 0x00000000 },
 	{ 0x418680,   1, 0x04, 0x00000000 },
 	{ 0x418714,   1, 0x04, 0x80000000 },
 	{ 0x418384,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_setup_0[] = {
 	{ 0x418814,   3, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_crstr_0[] = {
 	{ 0x418b04,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_setup_1[] = {
 	{ 0x4188c8,   1, 0x04, 0x80000000 },
 	{ 0x4188cc,   1, 0x04, 0x00000000 },
 	{ 0x4188d0,   1, 0x04, 0x00010000 },
 	{ 0x4188d4,   1, 0x04, 0x00000001 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_zcull_0[] = {
 	{ 0x418910,   1, 0x04, 0x00010001 },
 	{ 0x418914,   1, 0x04, 0x00000301 },
 	{ 0x418918,   1, 0x04, 0x00800000 },
 	{ 0x418980,   1, 0x04, 0x77777770 },
 	{ 0x418984,   3, 0x04, 0x77777777 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_gpm_0[] = {
 	{ 0x418c04,   1, 0x04, 0x00000000 },
 	{ 0x418c88,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_gpc_unk_1[] = {
 	{ 0x418d00,   1, 0x04, 0x00000000 },
 	{ 0x418f08,   1, 0x04, 0x00000000 },
 	{ 0x418e00,   1, 0x04, 0x00000050 },
 	{ 0x418e08,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_gcc_0[] = {
 	{ 0x41900c,   1, 0x04, 0x00000000 },
 	{ 0x419018,   1, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvc0_graph_init_tpc[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_tpccs_0[] = {
 	{ 0x419d08,   2, 0x04, 0x00000000 },
 	{ 0x419d10,   1, 0x04, 0x00000014 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_tex_0[] = {
 	{ 0x419ab0,   1, 0x04, 0x00000000 },
 	{ 0x419ab8,   1, 0x04, 0x000000e7 },
 	{ 0x419abc,   2, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_pe_0[] = {
 	{ 0x41980c,   3, 0x04, 0x00000000 },
 	{ 0x419844,   1, 0x04, 0x00000000 },
 	{ 0x41984c,   1, 0x04, 0x00005bc5 },
 	{ 0x419850,   4, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_l1c_0[] = {
 	{ 0x419c98,   1, 0x04, 0x00000000 },
 	{ 0x419ca8,   1, 0x04, 0x80000000 },
 	{ 0x419cb4,   1, 0x04, 0x00000000 },
 	{ 0x419cb8,   1, 0x04, 0x00008bf4 },
 	{ 0x419cbc,   1, 0x04, 0x28137606 },
 	{ 0x419cc0,   2, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_wwdx_0[] = {
 	{ 0x419bd4,   1, 0x04, 0x00800000 },
 	{ 0x419bdc,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_tpccs_1[] = {
 	{ 0x419d2c,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_mpc_0[] = {
 	{ 0x419c0c,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc0_graph_init_sm_0[] = {
 	{ 0x419e00,   1, 0x04, 0x00000000 },
 	{ 0x419ea0,   1, 0x04, 0x00000000 },
 	{ 0x419ea4,   1, 0x04, 0x00000100 },
@@ -270,8 +346,8 @@ nvc0_graph_init_tpc[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_graph_init_unk88xx[] = {
+const struct nvc0_graph_init
+nvc0_graph_init_be_0[] = {
 	{ 0x40880c,   1, 0x04, 0x00000000 },
 	{ 0x408910,   9, 0x04, 0x00000000 },
 	{ 0x408950,   1, 0x04, 0x00000000 },
@@ -282,18 +358,64 @@ nvc0_graph_init_unk88xx[] = {
 	{}
 };
 
-struct nvc0_graph_init
-nvc0_graph_tpc_0[] = {
-	{ 0x50405c,   1, 0x04, 0x00000001 },
+const struct nvc0_graph_init
+nvc0_graph_init_fe_1[] = {
+	{ 0x4040f0,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc0_graph_init_pe_1[] = {
+	{ 0x419880,   1, 0x04, 0x00000002 },
 	{}
 };
 
+static const struct nvc0_graph_pack
+nvc0_graph_pack_mmio[] = {
+	{ nvc0_graph_init_main_0 },
+	{ nvc0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvc0_graph_init_pd_0 },
+	{ nvc0_graph_init_ds_0 },
+	{ nvc0_graph_init_scc_0 },
+	{ nvc0_graph_init_prop_0 },
+	{ nvc0_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ nvc0_graph_init_setup_1 },
+	{ nvc0_graph_init_zcull_0 },
+	{ nvc0_graph_init_gpm_0 },
+	{ nvc0_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ nvc0_graph_init_tpccs_0 },
+	{ nvc0_graph_init_tex_0 },
+	{ nvc0_graph_init_pe_0 },
+	{ nvc0_graph_init_l1c_0 },
+	{ nvc0_graph_init_wwdx_0 },
+	{ nvc0_graph_init_tpccs_1 },
+	{ nvc0_graph_init_mpc_0 },
+	{ nvc0_graph_init_sm_0 },
+	{ nvc0_graph_init_be_0 },
+	{ nvc0_graph_init_fe_1 },
+	{ nvc0_graph_init_pe_1 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
 void
-nvc0_graph_mmio(struct nvc0_graph_priv *priv, struct nvc0_graph_init *init)
+nvc0_graph_mmio(struct nvc0_graph_priv *priv, const struct nvc0_graph_pack *p)
 {
-	for (; init && init->count; init++) {
-		u32 addr = init->addr, i;
-		for (i = 0; i < init->count; i++) {
+	const struct nvc0_graph_pack *pack;
+	const struct nvc0_graph_init *init;
+
+	pack_for_each_init(init, pack, p) {
+		u32 next = init->addr + init->count * init->pitch;
+		u32 addr = init->addr;
+		while (addr < next) {
 			nv_wr32(priv, addr, init->data);
 			addr += init->pitch;
 		}
@@ -301,49 +423,53 @@ nvc0_graph_mmio(struct nvc0_graph_priv *priv, struct nvc0_graph_init *init)
 }
 
 void
-nvc0_graph_icmd(struct nvc0_graph_priv *priv, struct nvc0_graph_init *init)
+nvc0_graph_icmd(struct nvc0_graph_priv *priv, const struct nvc0_graph_pack *p)
 {
-	u32 addr, data;
-	int i, j;
+	const struct nvc0_graph_pack *pack;
+	const struct nvc0_graph_init *init;
+	u32 data = 0;
 
 	nv_wr32(priv, 0x400208, 0x80000000);
-	for (i = 0; init->count; init++, i++) {
-		if (!i || data != init->data) {
+
+	pack_for_each_init(init, pack, p) {
+		u32 next = init->addr + init->count * init->pitch;
+		u32 addr = init->addr;
+
+		if ((pack == p && init == p->init) || data != init->data) {
 			nv_wr32(priv, 0x400204, init->data);
 			data = init->data;
 		}
 
-		addr = init->addr;
-		for (j = 0; j < init->count; j++) {
+		while (addr < next) {
 			nv_wr32(priv, 0x400200, addr);
+			nv_wait(priv, 0x400700, 0x00000002, 0x00000000);
 			addr += init->pitch;
-			while (nv_rd32(priv, 0x400700) & 0x00000002) {}
 		}
 	}
+
 	nv_wr32(priv, 0x400208, 0x00000000);
 }
 
 void
-nvc0_graph_mthd(struct nvc0_graph_priv *priv, struct nvc0_graph_mthd *mthds)
+nvc0_graph_mthd(struct nvc0_graph_priv *priv, const struct nvc0_graph_pack *p)
 {
-	struct nvc0_graph_mthd *mthd;
-	struct nvc0_graph_init *init;
-	int i = 0, j;
-	u32 data;
-
-	while ((mthd = &mthds[i++]) && (init = mthd->init)) {
-		u32  addr = 0x80000000 | mthd->oclass;
-		for (data = 0; init->count; init++) {
-			if (init == mthd->init || data != init->data) {
-				nv_wr32(priv, 0x40448c, init->data);
-				data = init->data;
-			}
+	const struct nvc0_graph_pack *pack;
+	const struct nvc0_graph_init *init;
+	u32 data = 0;
 
-			addr = (addr & 0x8000ffff) | (init->addr << 14);
-			for (j = 0; j < init->count; j++) {
-				nv_wr32(priv, 0x404488, addr);
-				addr += init->pitch << 14;
-			}
+	pack_for_each_init(init, pack, p) {
+		u32 ctrl = 0x80000000 | pack->type;
+		u32 next = init->addr + init->count * init->pitch;
+		u32 addr = init->addr;
+
+		if ((pack == p && init == p->init) || data != init->data) {
+			nv_wr32(priv, 0x40448c, init->data);
+			data = init->data;
+		}
+
+		while (addr < next) {
+			nv_wr32(priv, 0x404488, ctrl | (addr << 14));
+			addr += init->pitch;
 		}
 	}
 }
@@ -663,17 +789,40 @@ nvc0_graph_ctxctl_debug(struct nvc0_graph_priv *priv)
 static void
 nvc0_graph_ctxctl_isr(struct nvc0_graph_priv *priv)
 {
-	u32 ustat = nv_rd32(priv, 0x409c18);
+	u32 stat = nv_rd32(priv, 0x409c18);
+
+	if (stat & 0x00000001) {
+		u32 code = nv_rd32(priv, 0x409814);
+		if (code == E_BAD_FWMTHD) {
+			u32 class = nv_rd32(priv, 0x409808);
+			u32  addr = nv_rd32(priv, 0x40980c);
+			u32  subc = (addr & 0x00070000) >> 16;
+			u32  mthd = (addr & 0x00003ffc);
+			u32  data = nv_rd32(priv, 0x409810);
+
+			nv_error(priv, "FECS MTHD subc %d class 0x%04x "
+				       "mthd 0x%04x data 0x%08x\n",
+				 subc, class, mthd, data);
 
-	if (ustat & 0x00000001)
-		nv_error(priv, "CTXCTL ucode error\n");
-	if (ustat & 0x00080000)
-		nv_error(priv, "CTXCTL watchdog timeout\n");
-	if (ustat & ~0x00080001)
-		nv_error(priv, "CTXCTL 0x%08x\n", ustat);
+			nv_wr32(priv, 0x409c20, 0x00000001);
+			stat &= ~0x00000001;
+		} else {
+			nv_error(priv, "FECS ucode error %d\n", code);
+		}
+	}
+
+	if (stat & 0x00080000) {
+		nv_error(priv, "FECS watchdog timeout\n");
+		nvc0_graph_ctxctl_debug(priv);
+		nv_wr32(priv, 0x409c20, 0x00080000);
+		stat &= ~0x00080000;
+	}
 
-	nvc0_graph_ctxctl_debug(priv);
-	nv_wr32(priv, 0x409c20, ustat);
+	if (stat) {
+		nv_error(priv, "FECS 0x%08x\n", stat);
+		nvc0_graph_ctxctl_debug(priv);
+		nv_wr32(priv, 0x409c20, stat);
+	}
 }
 
 static void
@@ -768,15 +917,20 @@ nvc0_graph_init_fw(struct nvc0_graph_priv *priv, u32 fuc_base,
 			nv_wr32(priv, fuc_base + 0x0188, i >> 6);
 		nv_wr32(priv, fuc_base + 0x0184, code->data[i]);
 	}
+
+	/* code must be padded to 0x40 words */
+	for (; i & 0x3f; i++)
+		nv_wr32(priv, fuc_base + 0x0184, 0);
 }
 
 static void
 nvc0_graph_init_csdata(struct nvc0_graph_priv *priv,
-		       struct nvc0_graph_init *init,
+		       const struct nvc0_graph_pack *pack,
 		       u32 falcon, u32 starstar, u32 base)
 {
-	u32 addr = init->addr;
-	u32 next = addr;
+	const struct nvc0_graph_pack *iter;
+	const struct nvc0_graph_init *init;
+	u32 addr = ~0, prev = ~0, xfer = 0;
 	u32 star, temp;
 
 	nv_wr32(priv, falcon + 0x01c0, 0x02000000 + starstar);
@@ -786,22 +940,28 @@ nvc0_graph_init_csdata(struct nvc0_graph_priv *priv,
 		star = temp;
 	nv_wr32(priv, falcon + 0x01c0, 0x01000000 + star);
 
-	do {
-		if (init->addr != next) {
-			while (addr < next) {
-				u32 nr = min((int)(next - addr) / 4, 32);
-				nv_wr32(priv, falcon + 0x01c4,
-					((nr - 1) << 26) | (addr - base));
-				addr += nr * 4;
-				star += 4;
+	pack_for_each_init(init, iter, pack) {
+		u32 head = init->addr - base;
+		u32 tail = head + init->count * init->pitch;
+		while (head < tail) {
+			if (head != prev + 4 || xfer >= 32) {
+				if (xfer) {
+					u32 data = ((--xfer << 26) | addr);
+					nv_wr32(priv, falcon + 0x01c4, data);
+					star += 4;
+				}
+				addr = head;
+				xfer = 0;
 			}
-			addr = next = init->addr;
+			prev = head;
+			xfer = xfer + 1;
+			head = head + init->pitch;
 		}
-		next += init->count * 4;
-	} while ((init++)->count);
+	}
 
+	nv_wr32(priv, falcon + 0x01c4, (--xfer << 26) | addr);
 	nv_wr32(priv, falcon + 0x01c0, 0x01000004 + starstar);
-	nv_wr32(priv, falcon + 0x01c4, star);
+	nv_wr32(priv, falcon + 0x01c4, star + 4);
 }
 
 int
@@ -809,7 +969,6 @@ nvc0_graph_init_ctxctl(struct nvc0_graph_priv *priv)
 {
 	struct nvc0_graph_oclass *oclass = (void *)nv_object(priv)->oclass;
 	struct nvc0_grctx_oclass *cclass = (void *)nv_engine(priv)->cclass;
-	struct nvc0_graph_init *init;
 	u32 r000260;
 	int i;
 
@@ -901,6 +1060,9 @@ nvc0_graph_init_ctxctl(struct nvc0_graph_priv *priv)
 		}
 
 		return 0;
+	} else
+	if (!oclass->fecs.ucode) {
+		return -ENOSYS;
 	}
 
 	/* load HUB microcode */
@@ -916,10 +1078,6 @@ nvc0_graph_init_ctxctl(struct nvc0_graph_priv *priv)
 		nv_wr32(priv, 0x409184, oclass->fecs.ucode->code.data[i]);
 	}
 
-	for (i = 0; (init = cclass->hub[i]); i++) {
-		nvc0_graph_init_csdata(priv, init, 0x409000, 0x000, 0x000000);
-	}
-
 	/* load GPC microcode */
 	nv_wr32(priv, 0x41a1c0, 0x01000000);
 	for (i = 0; i < oclass->gpccs.ucode->data.size / 4; i++)
@@ -933,12 +1091,11 @@ nvc0_graph_init_ctxctl(struct nvc0_graph_priv *priv)
 	}
 	nv_wr32(priv, 0x000260, r000260);
 
-	if ((init = cclass->gpc[0]))
-		nvc0_graph_init_csdata(priv, init, 0x41a000, 0x000, 0x418000);
-	if ((init = cclass->gpc[2]))
-		nvc0_graph_init_csdata(priv, init, 0x41a000, 0x004, 0x419800);
-	if ((init = cclass->gpc[3]))
-		nvc0_graph_init_csdata(priv, init, 0x41a000, 0x008, 0x41be00);
+	/* load register lists */
+	nvc0_graph_init_csdata(priv, cclass->hub, 0x409000, 0x000, 0x000000);
+	nvc0_graph_init_csdata(priv, cclass->gpc, 0x41a000, 0x000, 0x418000);
+	nvc0_graph_init_csdata(priv, cclass->tpc, 0x41a000, 0x004, 0x419800);
+	nvc0_graph_init_csdata(priv, cclass->ppc, 0x41a000, 0x008, 0x41be00);
 
 	/* start HUB ucode running, it'll init the GPCs */
 	nv_wr32(priv, 0x40910c, 0x00000000);
@@ -985,8 +1142,7 @@ nvc0_graph_init(struct nouveau_object *object)
 	nv_wr32(priv, GPC_BCAST(0x08b4), priv->unk4188b4->addr >> 8);
 	nv_wr32(priv, GPC_BCAST(0x08b8), priv->unk4188b8->addr >> 8);
 
-	for (i = 0; oclass->mmio[i]; i++)
-		nvc0_graph_mmio(priv, oclass->mmio[i]);
+	nvc0_graph_mmio(priv, oclass->mmio);
 
 	memcpy(tpcnr, priv->tpc_nr, sizeof(priv->tpc_nr));
 	for (i = 0, gpc = -1; i < priv->tpc_total; i++) {
@@ -1088,10 +1244,10 @@ nvc0_graph_ctor_fw(struct nvc0_graph_priv *priv, const char *fwname,
 	int ret;
 
 	snprintf(f, sizeof(f), "nouveau/nv%02x_%s", device->chipset, fwname);
-	ret = request_firmware(&fw, f, &device->pdev->dev);
+	ret = request_firmware(&fw, f, nv_device_base(device));
 	if (ret) {
 		snprintf(f, sizeof(f), "nouveau/%s", fwname);
-		ret = request_firmware(&fw, f, &device->pdev->dev);
+		ret = request_firmware(&fw, f, nv_device_base(device));
 		if (ret) {
 			nv_error(priv, "failed to load %s\n", fwname);
 			return ret;
@@ -1130,10 +1286,14 @@ nvc0_graph_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nvc0_graph_oclass *oclass = (void *)bclass;
 	struct nouveau_device *device = nv_device(parent);
 	struct nvc0_graph_priv *priv;
+	bool use_ext_fw, enable;
 	int ret, i;
 
-	ret = nouveau_graph_create(parent, engine, bclass,
-				   (oclass->fecs.ucode != NULL), &priv);
+	use_ext_fw = nouveau_boolopt(device->cfgopt, "NvGrUseFW",
+				     oclass->fecs.ucode == NULL);
+	enable = use_ext_fw || oclass->fecs.ucode != NULL;
+
+	ret = nouveau_graph_create(parent, engine, bclass, enable, &priv);
 	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
@@ -1143,7 +1303,7 @@ nvc0_graph_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 	priv->base.units = nvc0_graph_units;
 
-	if (nouveau_boolopt(device->cfgopt, "NvGrUseFW", false)) {
+	if (use_ext_fw) {
 		nv_info(priv, "using external firmware\n");
 		if (nvc0_graph_ctor_fw(priv, "fuc409c", &priv->fuc409c) ||
 		    nvc0_graph_ctor_fw(priv, "fuc409d", &priv->fuc409d) ||
@@ -1217,22 +1377,6 @@ nvc0_graph_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	return 0;
 }
 
-struct nvc0_graph_init *
-nvc0_graph_init_mmio[] = {
-	nvc0_graph_init_regs,
-	nvc0_graph_init_unk40xx,
-	nvc0_graph_init_unk44xx,
-	nvc0_graph_init_unk78xx,
-	nvc0_graph_init_unk60xx,
-	nvc0_graph_init_unk58xx,
-	nvc0_graph_init_unk80xx,
-	nvc0_graph_init_gpc,
-	nvc0_graph_init_tpc,
-	nvc0_graph_init_unk88xx,
-	nvc0_graph_tpc_0,
-	NULL
-};
-
 #include "fuc/hubnvc0.fuc.h"
 
 struct nvc0_graph_ucode
@@ -1264,7 +1408,7 @@ nvc0_graph_oclass = &(struct nvc0_graph_oclass) {
 	},
 	.cclass = &nvc0_grctx_oclass,
 	.sclass =  nvc0_graph_sclass,
-	.mmio = nvc0_graph_init_mmio,
+	.mmio = nvc0_graph_pack_mmio,
 	.fecs.ucode = &nvc0_graph_fecs_ucode,
 	.gpccs.ucode = &nvc0_graph_gpccs_ucode,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.h b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.h
index ea17a80ad7f..ffc289198dd 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.h
@@ -38,6 +38,8 @@
 #include <engine/fifo.h>
 #include <engine/graph.h>
 
+#include "fuc/os.h"
+
 #define GPC_MAX 32
 #define TPC_MAX (GPC_MAX * 8)
 
@@ -45,6 +47,7 @@
 #define ROP_UNIT(u, r)    (0x410000 + (u) * 0x400 + (r))
 #define GPC_BCAST(r)      (0x418000 + (r))
 #define GPC_UNIT(t, r)    (0x500000 + (t) * 0x8000 + (r))
+#define PPC_UNIT(t, m, r) (0x503000 + (t) * 0x8000 + (m) * 0x200 + (r))
 #define TPC_UNIT(t, m, r) (0x504000 + (t) * 0x8000 + (m) * 0x800 + (r))
 
 struct nvc0_graph_data {
@@ -102,8 +105,6 @@ struct nvc0_graph_chan {
 	} data[4];
 };
 
-int  nvc0_grctx_generate(struct nvc0_graph_priv *);
-
 int  nvc0_graph_context_ctor(struct nouveau_object *, struct nouveau_object *,
 			     struct nouveau_oclass *, void *, u32,
 			     struct nouveau_object **);
@@ -117,6 +118,7 @@ int  nvc0_graph_ctor(struct nouveau_object *, struct nouveau_object *,
 		     struct nouveau_object **);
 void nvc0_graph_dtor(struct nouveau_object *);
 int  nvc0_graph_init(struct nouveau_object *);
+int  nve4_graph_fini(struct nouveau_object *, bool);
 int  nve4_graph_init(struct nouveau_object *);
 
 extern struct nouveau_oclass nvc0_graph_sclass[];
@@ -130,34 +132,14 @@ struct nvc0_graph_init {
 	u32 data;
 };
 
-struct nvc0_graph_mthd {
-	u16 oclass;
-	struct nvc0_graph_init *init;
+struct nvc0_graph_pack {
+	const struct nvc0_graph_init *init;
+	u32 type;
 };
 
-struct nvc0_grctx {
-	struct nvc0_graph_priv *priv;
-	struct nvc0_graph_data *data;
-	struct nvc0_graph_mmio *mmio;
-	int buffer_nr;
-	u64 buffer[4];
-	u64 addr;
-};
-
-struct nvc0_grctx_oclass {
-	struct nouveau_oclass base;
-	/* main context generation function */
-	void  (*main)(struct nvc0_graph_priv *, struct nvc0_grctx *);
-	/* context-specific modify-on-first-load list generation function */
-	void  (*mods)(struct nvc0_graph_priv *, struct nvc0_grctx *);
-	void  (*unkn)(struct nvc0_graph_priv *);
-	/* mmio context data */
-	struct nvc0_graph_init **hub;
-	struct nvc0_graph_init **gpc;
-	/* indirect context data, generated with icmds/mthds */
-	struct nvc0_graph_init *icmd;
-	struct nvc0_graph_mthd *mthd;
-};
+#define pack_for_each_init(init, pack, head)                                   \
+	for (pack = head; pack && pack->init; pack++)                          \
+		  for (init = pack->init; init && init->count; init++)
 
 struct nvc0_graph_ucode {
 	struct nvc0_graph_fuc code;
@@ -171,7 +153,7 @@ struct nvc0_graph_oclass {
 	struct nouveau_oclass base;
 	struct nouveau_oclass **cclass;
 	struct nouveau_oclass *sclass;
-	struct nvc0_graph_init **mmio;
+	const struct nvc0_graph_pack *mmio;
 	struct {
 		struct nvc0_graph_ucode *ucode;
 	} fecs;
@@ -180,109 +162,73 @@ struct nvc0_graph_oclass {
 	} gpccs;
 };
 
-void nvc0_graph_mmio(struct nvc0_graph_priv *, struct nvc0_graph_init *);
-void nvc0_graph_icmd(struct nvc0_graph_priv *, struct nvc0_graph_init *);
-void nvc0_graph_mthd(struct nvc0_graph_priv *, struct nvc0_graph_mthd *);
+void nvc0_graph_mmio(struct nvc0_graph_priv *, const struct nvc0_graph_pack *);
+void nvc0_graph_icmd(struct nvc0_graph_priv *, const struct nvc0_graph_pack *);
+void nvc0_graph_mthd(struct nvc0_graph_priv *, const struct nvc0_graph_pack *);
 int  nvc0_graph_init_ctxctl(struct nvc0_graph_priv *);
 
-extern struct nvc0_graph_init nvc0_graph_init_regs[];
-extern struct nvc0_graph_init nvc0_graph_init_unk40xx[];
-extern struct nvc0_graph_init nvc0_graph_init_unk44xx[];
-extern struct nvc0_graph_init nvc0_graph_init_unk78xx[];
-extern struct nvc0_graph_init nvc0_graph_init_unk60xx[];
-extern struct nvc0_graph_init nvc0_graph_init_unk58xx[];
-extern struct nvc0_graph_init nvc0_graph_init_unk80xx[];
-extern struct nvc0_graph_init nvc0_graph_init_gpc[];
-extern struct nvc0_graph_init nvc0_graph_init_unk88xx[];
-extern struct nvc0_graph_init nvc0_graph_tpc_0[];
-
-extern struct nvc0_graph_init nvc3_graph_init_unk58xx[];
-
-extern struct nvc0_graph_init nvd9_graph_init_unk58xx[];
-extern struct nvc0_graph_init nvd9_graph_init_unk64xx[];
-
-extern struct nvc0_graph_init nve4_graph_init_regs[];
-extern struct nvc0_graph_init nve4_graph_init_unk[];
-extern struct nvc0_graph_init nve4_graph_init_unk88xx[];
-
-int  nvc0_grctx_generate(struct nvc0_graph_priv *);
-void nvc0_grctx_generate_main(struct nvc0_graph_priv *, struct nvc0_grctx *);
-void nvc0_grctx_generate_mods(struct nvc0_graph_priv *, struct nvc0_grctx *);
-void nvc0_grctx_generate_unkn(struct nvc0_graph_priv *);
-void nvc0_grctx_generate_tpcid(struct nvc0_graph_priv *);
-void nvc0_grctx_generate_r406028(struct nvc0_graph_priv *);
-void nvc0_grctx_generate_r4060a8(struct nvc0_graph_priv *);
-void nvc0_grctx_generate_r418bb8(struct nvc0_graph_priv *);
-void nve4_grctx_generate_r418bb8(struct nvc0_graph_priv *);
-void nvc0_grctx_generate_r406800(struct nvc0_graph_priv *);
-
-extern struct nouveau_oclass *nvc0_grctx_oclass;
-extern struct nvc0_graph_init *nvc0_grctx_init_hub[];
-extern struct nvc0_graph_init nvc0_grctx_init_base[];
-extern struct nvc0_graph_init nvc0_grctx_init_unk40xx[];
-extern struct nvc0_graph_init nvc0_grctx_init_unk44xx[];
-extern struct nvc0_graph_init nvc0_grctx_init_unk46xx[];
-extern struct nvc0_graph_init nvc0_grctx_init_unk47xx[];
-extern struct nvc0_graph_init nvc0_grctx_init_unk60xx[];
-extern struct nvc0_graph_init nvc0_grctx_init_unk64xx[];
-extern struct nvc0_graph_init nvc0_grctx_init_unk78xx[];
-extern struct nvc0_graph_init nvc0_grctx_init_unk80xx[];
-extern struct nvc0_graph_init nvc0_grctx_init_gpc_0[];
-extern struct nvc0_graph_init nvc0_grctx_init_gpc_1[];
-extern struct nvc0_graph_init nvc0_grctx_init_tpc[];
-extern struct nvc0_graph_init nvc0_grctx_init_icmd[];
-extern struct nvc0_graph_init nvd9_grctx_init_icmd[]; //
-
-extern struct nvc0_graph_mthd nvc0_grctx_init_mthd[];
-extern struct nvc0_graph_init nvc0_grctx_init_902d[];
-extern struct nvc0_graph_init nvc0_grctx_init_9039[];
-extern struct nvc0_graph_init nvc0_grctx_init_90c0[];
-extern struct nvc0_graph_init nvc0_grctx_init_mthd_magic[];
-
-void nvc1_grctx_generate_mods(struct nvc0_graph_priv *, struct nvc0_grctx *);
-void nvc1_grctx_generate_unkn(struct nvc0_graph_priv *);
-extern struct nouveau_oclass *nvc1_grctx_oclass;
-extern struct nvc0_graph_init nvc1_grctx_init_9097[];
-
-extern struct nouveau_oclass *nvc3_grctx_oclass;
-
-extern struct nouveau_oclass *nvc8_grctx_oclass;
-extern struct nvc0_graph_init nvc8_grctx_init_9197[];
-extern struct nvc0_graph_init nvc8_grctx_init_9297[];
-
-extern struct nouveau_oclass *nvd7_grctx_oclass;
-
-extern struct nouveau_oclass *nvd9_grctx_oclass;
-extern struct nvc0_graph_init nvd9_grctx_init_rop[];
-extern struct nvc0_graph_mthd nvd9_grctx_init_mthd[];
-
-void nve4_grctx_generate_main(struct nvc0_graph_priv *, struct nvc0_grctx *);
-void nve4_grctx_generate_unkn(struct nvc0_graph_priv *);
-extern struct nouveau_oclass *nve4_grctx_oclass;
-extern struct nvc0_graph_init nve4_grctx_init_unk46xx[];
-extern struct nvc0_graph_init nve4_grctx_init_unk47xx[];
-extern struct nvc0_graph_init nve4_grctx_init_unk58xx[];
-extern struct nvc0_graph_init nve4_grctx_init_unk80xx[];
-extern struct nvc0_graph_init nve4_grctx_init_unk90xx[];
-
-extern struct nouveau_oclass *nvf0_grctx_oclass;
-
-#define mmio_data(s,a,p) do {                                                  \
-	info->buffer[info->buffer_nr] = round_up(info->addr, (a));             \
-	info->addr = info->buffer[info->buffer_nr++] + (s);                    \
-	info->data->size = (s);                                                \
-	info->data->align = (a);                                               \
-	info->data->access = (p);                                              \
-	info->data++;                                                          \
-} while(0)
-
-#define mmio_list(r,d,s,b) do {                                                \
-	info->mmio->addr = (r);                                                \
-	info->mmio->data = (d);                                                \
-	info->mmio->shift = (s);                                               \
-	info->mmio->buffer = (b);                                              \
-	info->mmio++;                                                          \
-	nv_wr32(priv, (r), (d) | ((s) ? (info->buffer[(b)] >> (s)) : 0));      \
-} while(0)
+/* register init value lists */
+
+extern const struct nvc0_graph_init nvc0_graph_init_main_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_fe_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_pri_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_rstr2d_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_pd_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_ds_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_scc_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_prop_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_gpc_unk_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_setup_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_crstr_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_setup_1[];
+extern const struct nvc0_graph_init nvc0_graph_init_zcull_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_gpm_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_gpc_unk_1[];
+extern const struct nvc0_graph_init nvc0_graph_init_gcc_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_tpccs_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_tex_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_pe_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_l1c_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_wwdx_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_tpccs_1[];
+extern const struct nvc0_graph_init nvc0_graph_init_mpc_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_be_0[];
+extern const struct nvc0_graph_init nvc0_graph_init_fe_1[];
+extern const struct nvc0_graph_init nvc0_graph_init_pe_1[];
+
+extern const struct nvc0_graph_init nvc4_graph_init_ds_0[];
+extern const struct nvc0_graph_init nvc4_graph_init_tex_0[];
+extern const struct nvc0_graph_init nvc4_graph_init_sm_0[];
+
+extern const struct nvc0_graph_init nvc1_graph_init_gpc_unk_0[];
+extern const struct nvc0_graph_init nvc1_graph_init_setup_1[];
+
+extern const struct nvc0_graph_init nvd9_graph_init_pd_0[];
+extern const struct nvc0_graph_init nvd9_graph_init_ds_0[];
+extern const struct nvc0_graph_init nvd9_graph_init_prop_0[];
+extern const struct nvc0_graph_init nvd9_graph_init_gpm_0[];
+extern const struct nvc0_graph_init nvd9_graph_init_gpc_unk_1[];
+extern const struct nvc0_graph_init nvd9_graph_init_tex_0[];
+extern const struct nvc0_graph_init nvd9_graph_init_sm_0[];
+extern const struct nvc0_graph_init nvd9_graph_init_fe_1[];
+
+extern const struct nvc0_graph_init nvd7_graph_init_pes_0[];
+extern const struct nvc0_graph_init nvd7_graph_init_wwdx_0[];
+extern const struct nvc0_graph_init nvd7_graph_init_cbm_0[];
+
+extern const struct nvc0_graph_init nve4_graph_init_main_0[];
+extern const struct nvc0_graph_init nve4_graph_init_tpccs_0[];
+extern const struct nvc0_graph_init nve4_graph_init_pe_0[];
+extern const struct nvc0_graph_init nve4_graph_init_be_0[];
+extern const struct nvc0_graph_pack nve4_graph_pack_mmio[];
+
+extern const struct nvc0_graph_init nvf0_graph_init_fe_0[];
+extern const struct nvc0_graph_init nvf0_graph_init_sked_0[];
+extern const struct nvc0_graph_init nvf0_graph_init_cwd_0[];
+extern const struct nvc0_graph_init nvf0_graph_init_gpc_unk_1[];
+extern const struct nvc0_graph_init nvf0_graph_init_sm_0[];
+
+extern const struct nvc0_graph_init nv108_graph_init_gpc_unk_0[];
+
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvc1.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvc1.c
index bc4a469b86c..30cab0b2eba 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvc1.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvc1.c
@@ -23,6 +23,7 @@
  */
 
 #include "nvc0.h"
+#include "ctxnvc0.h"
 
 /*******************************************************************************
  * Graphics object classes
@@ -39,94 +40,82 @@ nvc1_graph_sclass[] = {
 };
 
 /*******************************************************************************
- * PGRAPH engine/subdev functions
+ * PGRAPH register lists
  ******************************************************************************/
 
-static struct nvc0_graph_init
-nvc1_graph_init_gpc[] = {
-	{ 0x4184a0,   1, 0x04, 0x00000000 },
+const struct nvc0_graph_init
+nvc1_graph_init_gpc_unk_0[] = {
 	{ 0x418604,   1, 0x04, 0x00000000 },
 	{ 0x418680,   1, 0x04, 0x00000000 },
 	{ 0x418714,   1, 0x04, 0x00000000 },
 	{ 0x418384,   1, 0x04, 0x00000000 },
-	{ 0x418814,   3, 0x04, 0x00000000 },
-	{ 0x418b04,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc1_graph_init_setup_1[] = {
 	{ 0x4188c8,   2, 0x04, 0x00000000 },
 	{ 0x4188d0,   1, 0x04, 0x00010000 },
 	{ 0x4188d4,   1, 0x04, 0x00000001 },
-	{ 0x418910,   1, 0x04, 0x00010001 },
-	{ 0x418914,   1, 0x04, 0x00000301 },
-	{ 0x418918,   1, 0x04, 0x00800000 },
-	{ 0x418980,   1, 0x04, 0x77777770 },
-	{ 0x418984,   3, 0x04, 0x77777777 },
-	{ 0x418c04,   1, 0x04, 0x00000000 },
-	{ 0x418c88,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc1_graph_init_gpc_unk_1[] = {
 	{ 0x418d00,   1, 0x04, 0x00000000 },
 	{ 0x418f08,   1, 0x04, 0x00000000 },
 	{ 0x418e00,   1, 0x04, 0x00000003 },
 	{ 0x418e08,   1, 0x04, 0x00000000 },
-	{ 0x41900c,   1, 0x04, 0x00000000 },
-	{ 0x419018,   1, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvc1_graph_init_tpc[] = {
-	{ 0x419d08,   2, 0x04, 0x00000000 },
-	{ 0x419d10,   1, 0x04, 0x00000014 },
-	{ 0x419ab0,   1, 0x04, 0x00000000 },
-	{ 0x419ac8,   1, 0x04, 0x00000000 },
-	{ 0x419ab8,   1, 0x04, 0x000000e7 },
-	{ 0x419abc,   2, 0x04, 0x00000000 },
-	{ 0x41980c,   2, 0x04, 0x00000000 },
+static const struct nvc0_graph_init
+nvc1_graph_init_pe_0[] = {
+	{ 0x41980c,   1, 0x04, 0x00000010 },
+	{ 0x419810,   1, 0x04, 0x00000000 },
 	{ 0x419814,   1, 0x04, 0x00000004 },
 	{ 0x419844,   1, 0x04, 0x00000000 },
 	{ 0x41984c,   1, 0x04, 0x00005bc5 },
 	{ 0x419850,   4, 0x04, 0x00000000 },
 	{ 0x419880,   1, 0x04, 0x00000002 },
-	{ 0x419c98,   1, 0x04, 0x00000000 },
-	{ 0x419ca8,   1, 0x04, 0x80000000 },
-	{ 0x419cb4,   1, 0x04, 0x00000000 },
-	{ 0x419cb8,   1, 0x04, 0x00008bf4 },
-	{ 0x419cbc,   1, 0x04, 0x28137606 },
-	{ 0x419cc0,   2, 0x04, 0x00000000 },
-	{ 0x419bd4,   1, 0x04, 0x00800000 },
-	{ 0x419bdc,   1, 0x04, 0x00000000 },
-	{ 0x419d2c,   1, 0x04, 0x00000000 },
-	{ 0x419c0c,   1, 0x04, 0x00000000 },
-	{ 0x419e00,   1, 0x04, 0x00000000 },
-	{ 0x419ea0,   1, 0x04, 0x00000000 },
-	{ 0x419ea4,   1, 0x04, 0x00000100 },
-	{ 0x419ea8,   1, 0x04, 0x00001100 },
-	{ 0x419eac,   1, 0x04, 0x11100702 },
-	{ 0x419eb0,   1, 0x04, 0x00000003 },
-	{ 0x419eb4,   4, 0x04, 0x00000000 },
-	{ 0x419ec8,   1, 0x04, 0x0e063818 },
-	{ 0x419ecc,   1, 0x04, 0x0e060e06 },
-	{ 0x419ed0,   1, 0x04, 0x00003818 },
-	{ 0x419ed4,   1, 0x04, 0x011104f1 },
-	{ 0x419edc,   1, 0x04, 0x00000000 },
-	{ 0x419f00,   1, 0x04, 0x00000000 },
-	{ 0x419f2c,   1, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init *
-nvc1_graph_init_mmio[] = {
-	nvc0_graph_init_regs,
-	nvc0_graph_init_unk40xx,
-	nvc0_graph_init_unk44xx,
-	nvc0_graph_init_unk78xx,
-	nvc0_graph_init_unk60xx,
-	nvc3_graph_init_unk58xx,
-	nvc0_graph_init_unk80xx,
-	nvc1_graph_init_gpc,
-	nvc1_graph_init_tpc,
-	nvc0_graph_init_unk88xx,
-	nvc0_graph_tpc_0,
-	NULL
+static const struct nvc0_graph_pack
+nvc1_graph_pack_mmio[] = {
+	{ nvc0_graph_init_main_0 },
+	{ nvc0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvc0_graph_init_pd_0 },
+	{ nvc4_graph_init_ds_0 },
+	{ nvc0_graph_init_scc_0 },
+	{ nvc0_graph_init_prop_0 },
+	{ nvc1_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ nvc1_graph_init_setup_1 },
+	{ nvc0_graph_init_zcull_0 },
+	{ nvc0_graph_init_gpm_0 },
+	{ nvc1_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ nvc0_graph_init_tpccs_0 },
+	{ nvc4_graph_init_tex_0 },
+	{ nvc1_graph_init_pe_0 },
+	{ nvc0_graph_init_l1c_0 },
+	{ nvc0_graph_init_wwdx_0 },
+	{ nvc0_graph_init_tpccs_1 },
+	{ nvc0_graph_init_mpc_0 },
+	{ nvc4_graph_init_sm_0 },
+	{ nvc0_graph_init_be_0 },
+	{ nvc0_graph_init_fe_1 },
+	{}
 };
 
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
 struct nouveau_oclass *
 nvc1_graph_oclass = &(struct nvc0_graph_oclass) {
 	.base.handle = NV_ENGINE(GR, 0xc1),
@@ -138,7 +127,7 @@ nvc1_graph_oclass = &(struct nvc0_graph_oclass) {
 	},
 	.cclass = &nvc1_grctx_oclass,
 	.sclass = nvc1_graph_sclass,
-	.mmio = nvc1_graph_init_mmio,
+	.mmio = nvc1_graph_pack_mmio,
 	.fecs.ucode = &nvc0_graph_fecs_ucode,
 	.gpccs.ucode = &nvc0_graph_gpccs_ucode,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvc3.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvc4.c
index d44b3b3ee80..e82e70c5313 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvc3.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvc4.c
@@ -23,13 +23,14 @@
  */
 
 #include "nvc0.h"
+#include "ctxnvc0.h"
 
 /*******************************************************************************
- * PGRAPH engine/subdev functions
+ * PGRAPH register lists
  ******************************************************************************/
 
-struct nvc0_graph_init
-nvc3_graph_init_unk58xx[] = {
+const struct nvc0_graph_init
+nvc4_graph_init_ds_0[] = {
 	{ 0x405844,   1, 0x04, 0x00ffffff },
 	{ 0x405850,   1, 0x04, 0x00000000 },
 	{ 0x405900,   1, 0x04, 0x00002834 },
@@ -37,29 +38,27 @@ nvc3_graph_init_unk58xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nvc3_graph_init_tpc[] = {
-	{ 0x419d08,   2, 0x04, 0x00000000 },
-	{ 0x419d10,   1, 0x04, 0x00000014 },
+const struct nvc0_graph_init
+nvc4_graph_init_tex_0[] = {
 	{ 0x419ab0,   1, 0x04, 0x00000000 },
 	{ 0x419ac8,   1, 0x04, 0x00000000 },
 	{ 0x419ab8,   1, 0x04, 0x000000e7 },
 	{ 0x419abc,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvc4_graph_init_pe_0[] = {
 	{ 0x41980c,   3, 0x04, 0x00000000 },
 	{ 0x419844,   1, 0x04, 0x00000000 },
 	{ 0x41984c,   1, 0x04, 0x00005bc5 },
 	{ 0x419850,   4, 0x04, 0x00000000 },
 	{ 0x419880,   1, 0x04, 0x00000002 },
-	{ 0x419c98,   1, 0x04, 0x00000000 },
-	{ 0x419ca8,   1, 0x04, 0x80000000 },
-	{ 0x419cb4,   1, 0x04, 0x00000000 },
-	{ 0x419cb8,   1, 0x04, 0x00008bf4 },
-	{ 0x419cbc,   1, 0x04, 0x28137606 },
-	{ 0x419cc0,   2, 0x04, 0x00000000 },
-	{ 0x419bd4,   1, 0x04, 0x00800000 },
-	{ 0x419bdc,   1, 0x04, 0x00000000 },
-	{ 0x419d2c,   1, 0x04, 0x00000000 },
-	{ 0x419c0c,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvc4_graph_init_sm_0[] = {
 	{ 0x419e00,   1, 0x04, 0x00000000 },
 	{ 0x419ea0,   1, 0x04, 0x00000000 },
 	{ 0x419ea4,   1, 0x04, 0x00000100 },
@@ -77,24 +76,43 @@ nvc3_graph_init_tpc[] = {
 	{}
 };
 
-static struct nvc0_graph_init *
-nvc3_graph_init_mmio[] = {
-	nvc0_graph_init_regs,
-	nvc0_graph_init_unk40xx,
-	nvc0_graph_init_unk44xx,
-	nvc0_graph_init_unk78xx,
-	nvc0_graph_init_unk60xx,
-	nvc3_graph_init_unk58xx,
-	nvc0_graph_init_unk80xx,
-	nvc0_graph_init_gpc,
-	nvc3_graph_init_tpc,
-	nvc0_graph_init_unk88xx,
-	nvc0_graph_tpc_0,
-	NULL
+static const struct nvc0_graph_pack
+nvc4_graph_pack_mmio[] = {
+	{ nvc0_graph_init_main_0 },
+	{ nvc0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvc0_graph_init_pd_0 },
+	{ nvc4_graph_init_ds_0 },
+	{ nvc0_graph_init_scc_0 },
+	{ nvc0_graph_init_prop_0 },
+	{ nvc0_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ nvc0_graph_init_setup_1 },
+	{ nvc0_graph_init_zcull_0 },
+	{ nvc0_graph_init_gpm_0 },
+	{ nvc0_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ nvc0_graph_init_tpccs_0 },
+	{ nvc4_graph_init_tex_0 },
+	{ nvc4_graph_init_pe_0 },
+	{ nvc0_graph_init_l1c_0 },
+	{ nvc0_graph_init_wwdx_0 },
+	{ nvc0_graph_init_tpccs_1 },
+	{ nvc0_graph_init_mpc_0 },
+	{ nvc4_graph_init_sm_0 },
+	{ nvc0_graph_init_be_0 },
+	{ nvc0_graph_init_fe_1 },
+	{}
 };
 
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
 struct nouveau_oclass *
-nvc3_graph_oclass = &(struct nvc0_graph_oclass) {
+nvc4_graph_oclass = &(struct nvc0_graph_oclass) {
 	.base.handle = NV_ENGINE(GR, 0xc3),
 	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nvc0_graph_ctor,
@@ -102,9 +120,9 @@ nvc3_graph_oclass = &(struct nvc0_graph_oclass) {
 		.init = nvc0_graph_init,
 		.fini = _nouveau_graph_fini,
 	},
-	.cclass = &nvc3_grctx_oclass,
+	.cclass = &nvc4_grctx_oclass,
 	.sclass = nvc0_graph_sclass,
-	.mmio = nvc3_graph_init_mmio,
+	.mmio = nvc4_graph_pack_mmio,
 	.fecs.ucode = &nvc0_graph_fecs_ucode,
 	.gpccs.ucode = &nvc0_graph_gpccs_ucode,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvc8.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvc8.c
index 02845e56731..a6bf783e125 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvc8.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvc8.c
@@ -23,6 +23,7 @@
  */
 
 #include "nvc0.h"
+#include "ctxnvc0.h"
 
 /*******************************************************************************
  * Graphics object classes
@@ -40,58 +41,11 @@ nvc8_graph_sclass[] = {
 };
 
 /*******************************************************************************
- * PGRAPH engine/subdev functions
+ * PGRAPH register lists
  ******************************************************************************/
 
-static struct nvc0_graph_init
-nvc8_graph_init_gpc[] = {
-	{ 0x4184a0,   1, 0x04, 0x00000000 },
-	{ 0x418604,   1, 0x04, 0x00000000 },
-	{ 0x418680,   1, 0x04, 0x00000000 },
-	{ 0x418714,   1, 0x04, 0x80000000 },
-	{ 0x418384,   1, 0x04, 0x00000000 },
-	{ 0x418814,   3, 0x04, 0x00000000 },
-	{ 0x418b04,   1, 0x04, 0x00000000 },
-	{ 0x4188c8,   2, 0x04, 0x00000000 },
-	{ 0x4188d0,   1, 0x04, 0x00010000 },
-	{ 0x4188d4,   1, 0x04, 0x00000001 },
-	{ 0x418910,   1, 0x04, 0x00010001 },
-	{ 0x418914,   1, 0x04, 0x00000301 },
-	{ 0x418918,   1, 0x04, 0x00800000 },
-	{ 0x418980,   1, 0x04, 0x77777770 },
-	{ 0x418984,   3, 0x04, 0x77777777 },
-	{ 0x418c04,   1, 0x04, 0x00000000 },
-	{ 0x418c88,   1, 0x04, 0x00000000 },
-	{ 0x418d00,   1, 0x04, 0x00000000 },
-	{ 0x418f08,   1, 0x04, 0x00000000 },
-	{ 0x418e00,   1, 0x04, 0x00000050 },
-	{ 0x418e08,   1, 0x04, 0x00000000 },
-	{ 0x41900c,   1, 0x04, 0x00000000 },
-	{ 0x419018,   1, 0x04, 0x00000000 },
-	{}
-};
-
-static struct nvc0_graph_init
-nvc8_graph_init_tpc[] = {
-	{ 0x419d08,   2, 0x04, 0x00000000 },
-	{ 0x419d10,   1, 0x04, 0x00000014 },
-	{ 0x419ab0,   1, 0x04, 0x00000000 },
-	{ 0x419ab8,   1, 0x04, 0x000000e7 },
-	{ 0x419abc,   2, 0x04, 0x00000000 },
-	{ 0x41980c,   3, 0x04, 0x00000000 },
-	{ 0x419844,   1, 0x04, 0x00000000 },
-	{ 0x41984c,   1, 0x04, 0x00005bc5 },
-	{ 0x419850,   4, 0x04, 0x00000000 },
-	{ 0x419c98,   1, 0x04, 0x00000000 },
-	{ 0x419ca8,   1, 0x04, 0x80000000 },
-	{ 0x419cb4,   1, 0x04, 0x00000000 },
-	{ 0x419cb8,   1, 0x04, 0x00008bf4 },
-	{ 0x419cbc,   1, 0x04, 0x28137606 },
-	{ 0x419cc0,   2, 0x04, 0x00000000 },
-	{ 0x419bd4,   1, 0x04, 0x00800000 },
-	{ 0x419bdc,   1, 0x04, 0x00000000 },
-	{ 0x419d2c,   1, 0x04, 0x00000000 },
-	{ 0x419c0c,   1, 0x04, 0x00000000 },
+static const struct nvc0_graph_init
+nvc8_graph_init_sm_0[] = {
 	{ 0x419e00,   1, 0x04, 0x00000000 },
 	{ 0x419ea0,   1, 0x04, 0x00000000 },
 	{ 0x419ea4,   1, 0x04, 0x00000100 },
@@ -108,22 +62,42 @@ nvc8_graph_init_tpc[] = {
 	{}
 };
 
-static struct nvc0_graph_init *
-nvc8_graph_init_mmio[] = {
-	nvc0_graph_init_regs,
-	nvc0_graph_init_unk40xx,
-	nvc0_graph_init_unk44xx,
-	nvc0_graph_init_unk78xx,
-	nvc0_graph_init_unk60xx,
-	nvc0_graph_init_unk58xx,
-	nvc0_graph_init_unk80xx,
-	nvc8_graph_init_gpc,
-	nvc8_graph_init_tpc,
-	nvc0_graph_init_unk88xx,
-	nvc0_graph_tpc_0,
-	NULL
+static const struct nvc0_graph_pack
+nvc8_graph_pack_mmio[] = {
+	{ nvc0_graph_init_main_0 },
+	{ nvc0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvc0_graph_init_pd_0 },
+	{ nvc0_graph_init_ds_0 },
+	{ nvc0_graph_init_scc_0 },
+	{ nvc0_graph_init_prop_0 },
+	{ nvc0_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ nvc1_graph_init_setup_1 },
+	{ nvc0_graph_init_zcull_0 },
+	{ nvc0_graph_init_gpm_0 },
+	{ nvc0_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ nvc0_graph_init_tpccs_0 },
+	{ nvc0_graph_init_tex_0 },
+	{ nvc0_graph_init_pe_0 },
+	{ nvc0_graph_init_l1c_0 },
+	{ nvc0_graph_init_wwdx_0 },
+	{ nvc0_graph_init_tpccs_1 },
+	{ nvc0_graph_init_mpc_0 },
+	{ nvc8_graph_init_sm_0 },
+	{ nvc0_graph_init_be_0 },
+	{ nvc0_graph_init_fe_1 },
+	{ nvc0_graph_init_pe_1 },
+	{}
 };
 
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
 struct nouveau_oclass *
 nvc8_graph_oclass = &(struct nvc0_graph_oclass) {
 	.base.handle = NV_ENGINE(GR, 0xc8),
@@ -135,7 +109,7 @@ nvc8_graph_oclass = &(struct nvc0_graph_oclass) {
 	},
 	.cclass = &nvc8_grctx_oclass,
 	.sclass = nvc8_graph_sclass,
-	.mmio = nvc8_graph_init_mmio,
+	.mmio = nvc8_graph_pack_mmio,
 	.fecs.ucode = &nvc0_graph_fecs_ucode,
 	.gpccs.ucode = &nvc0_graph_gpccs_ucode,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvd7.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvd7.c
index 5052d7ab4d7..2a6a94e2a04 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvd7.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvd7.c
@@ -23,6 +23,77 @@
  */
 
 #include "nvc0.h"
+#include "ctxnvc0.h"
+
+/*******************************************************************************
+ * PGRAPH register lists
+ ******************************************************************************/
+
+static const struct nvc0_graph_init
+nvd7_graph_init_pe_0[] = {
+	{ 0x41980c,   1, 0x04, 0x00000010 },
+	{ 0x419844,   1, 0x04, 0x00000000 },
+	{ 0x41984c,   1, 0x04, 0x00005bc8 },
+	{ 0x419850,   3, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd7_graph_init_pes_0[] = {
+	{ 0x41be04,   1, 0x04, 0x00000000 },
+	{ 0x41be08,   1, 0x04, 0x00000004 },
+	{ 0x41be0c,   1, 0x04, 0x00000000 },
+	{ 0x41be10,   1, 0x04, 0x003b8bc7 },
+	{ 0x41be14,   2, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd7_graph_init_wwdx_0[] = {
+	{ 0x41bfd4,   1, 0x04, 0x00800000 },
+	{ 0x41bfdc,   1, 0x04, 0x00000000 },
+	{ 0x41bff8,   2, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd7_graph_init_cbm_0[] = {
+	{ 0x41becc,   1, 0x04, 0x00000000 },
+	{ 0x41bee8,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_pack
+nvd7_graph_pack_mmio[] = {
+	{ nvc0_graph_init_main_0 },
+	{ nvc0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvd9_graph_init_pd_0 },
+	{ nvd9_graph_init_ds_0 },
+	{ nvc0_graph_init_scc_0 },
+	{ nvd9_graph_init_prop_0 },
+	{ nvc1_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ nvc1_graph_init_setup_1 },
+	{ nvc0_graph_init_zcull_0 },
+	{ nvd9_graph_init_gpm_0 },
+	{ nvd9_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ nvc0_graph_init_tpccs_0 },
+	{ nvd9_graph_init_tex_0 },
+	{ nvd7_graph_init_pe_0 },
+	{ nvc0_graph_init_l1c_0 },
+	{ nvc0_graph_init_mpc_0 },
+	{ nvd9_graph_init_sm_0 },
+	{ nvd7_graph_init_pes_0 },
+	{ nvd7_graph_init_wwdx_0 },
+	{ nvd7_graph_init_cbm_0 },
+	{ nvc0_graph_init_be_0 },
+	{ nvd9_graph_init_fe_1 },
+	{}
+};
 
 /*******************************************************************************
  * PGRAPH engine/subdev functions
@@ -48,108 +119,6 @@ nvd7_graph_gpccs_ucode = {
 	.data.size = sizeof(nvd7_grgpc_data),
 };
 
-static struct nvc0_graph_init
-nvd7_graph_init_gpc[] = {
-	{ 0x418408,   1, 0x04, 0x00000000 },
-	{ 0x4184a0,   1, 0x04, 0x00000000 },
-	{ 0x4184a4,   2, 0x04, 0x00000000 },
-	{ 0x418604,   1, 0x04, 0x00000000 },
-	{ 0x418680,   1, 0x04, 0x00000000 },
-	{ 0x418714,   1, 0x04, 0x00000000 },
-	{ 0x418384,   1, 0x04, 0x00000000 },
-	{ 0x418814,   3, 0x04, 0x00000000 },
-	{ 0x418b04,   1, 0x04, 0x00000000 },
-	{ 0x4188c8,   2, 0x04, 0x00000000 },
-	{ 0x4188d0,   1, 0x04, 0x00010000 },
-	{ 0x4188d4,   1, 0x04, 0x00000001 },
-	{ 0x418910,   1, 0x04, 0x00010001 },
-	{ 0x418914,   1, 0x04, 0x00000301 },
-	{ 0x418918,   1, 0x04, 0x00800000 },
-	{ 0x418980,   1, 0x04, 0x77777770 },
-	{ 0x418984,   3, 0x04, 0x77777777 },
-	{ 0x418c04,   1, 0x04, 0x00000000 },
-	{ 0x418c64,   1, 0x04, 0x00000000 },
-	{ 0x418c68,   1, 0x04, 0x00000000 },
-	{ 0x418c88,   1, 0x04, 0x00000000 },
-	{ 0x418cb4,   2, 0x04, 0x00000000 },
-	{ 0x418d00,   1, 0x04, 0x00000000 },
-	{ 0x418d28,   1, 0x04, 0x00000000 },
-	{ 0x418f00,   1, 0x04, 0x00000000 },
-	{ 0x418f08,   1, 0x04, 0x00000000 },
-	{ 0x418f20,   2, 0x04, 0x00000000 },
-	{ 0x418e00,   1, 0x04, 0x00000003 },
-	{ 0x418e08,   1, 0x04, 0x00000000 },
-	{ 0x418e1c,   1, 0x04, 0x00000000 },
-	{ 0x418e20,   1, 0x04, 0x00000000 },
-	{ 0x41900c,   1, 0x04, 0x00000000 },
-	{ 0x419018,   1, 0x04, 0x00000000 },
-	{}
-};
-
-static struct nvc0_graph_init
-nvd7_graph_init_tpc[] = {
-	{ 0x419d08,   2, 0x04, 0x00000000 },
-	{ 0x419d10,   1, 0x04, 0x00000014 },
-	{ 0x419ab0,   1, 0x04, 0x00000000 },
-	{ 0x419ac8,   1, 0x04, 0x00000000 },
-	{ 0x419ab8,   1, 0x04, 0x000000e7 },
-	{ 0x419abc,   2, 0x04, 0x00000000 },
-	{ 0x419ab4,   1, 0x04, 0x00000000 },
-	{ 0x41980c,   1, 0x04, 0x00000010 },
-	{ 0x419844,   1, 0x04, 0x00000000 },
-	{ 0x41984c,   1, 0x04, 0x00005bc8 },
-	{ 0x419850,   2, 0x04, 0x00000000 },
-	{ 0x419c98,   1, 0x04, 0x00000000 },
-	{ 0x419ca8,   1, 0x04, 0x80000000 },
-	{ 0x419cb4,   1, 0x04, 0x00000000 },
-	{ 0x419cb8,   1, 0x04, 0x00008bf4 },
-	{ 0x419cbc,   1, 0x04, 0x28137606 },
-	{ 0x419cc0,   2, 0x04, 0x00000000 },
-	{ 0x419c0c,   1, 0x04, 0x00000000 },
-	{ 0x419e00,   1, 0x04, 0x00000000 },
-	{ 0x419ea0,   1, 0x04, 0x00000000 },
-	{ 0x419ea4,   1, 0x04, 0x00000100 },
-	{ 0x419ea8,   1, 0x04, 0x02001100 },
-	{ 0x419eac,   1, 0x04, 0x11100702 },
-	{ 0x419eb0,   1, 0x04, 0x00000003 },
-	{ 0x419eb4,   4, 0x04, 0x00000000 },
-	{ 0x419ec8,   1, 0x04, 0x0e063818 },
-	{ 0x419ecc,   1, 0x04, 0x0e060e06 },
-	{ 0x419ed0,   1, 0x04, 0x00003818 },
-	{ 0x419ed4,   1, 0x04, 0x011104f1 },
-	{ 0x419edc,   1, 0x04, 0x00000000 },
-	{ 0x419f00,   1, 0x04, 0x00000000 },
-	{ 0x419f2c,   1, 0x04, 0x00000000 },
-	{}
-};
-
-static struct nvc0_graph_init
-nvd7_graph_init_tpc_0[] = {
-	{ 0x40402c,   1, 0x04, 0x00000000 },
-	{ 0x4040f0,   1, 0x04, 0x00000000 },
-	{ 0x404174,   1, 0x04, 0x00000000 },
-	{ 0x503018,   1, 0x04, 0x00000001 },
-	{}
-};
-
-static struct nvc0_graph_init *
-nvd7_graph_init_mmio[] = {
-	nvc0_graph_init_regs,
-	nvc0_graph_init_unk40xx,
-	nvc0_graph_init_unk44xx,
-	nvc0_graph_init_unk78xx,
-	nvc0_graph_init_unk60xx,
-	nvd9_graph_init_unk64xx,
-	nvd9_graph_init_unk58xx,
-	nvc0_graph_init_unk80xx,
-	nvd7_graph_init_gpc,
-	nvd7_graph_init_tpc,
-	nve4_graph_init_unk,
-	nvc0_graph_init_unk88xx,
-	nvd7_graph_init_tpc_0,
-	NULL
-};
-
 struct nouveau_oclass *
 nvd7_graph_oclass = &(struct nvc0_graph_oclass) {
 	.base.handle = NV_ENGINE(GR, 0xd7),
@@ -161,7 +130,7 @@ nvd7_graph_oclass = &(struct nvc0_graph_oclass) {
 	},
 	.cclass = &nvd7_grctx_oclass,
 	.sclass = nvc8_graph_sclass,
-	.mmio = nvd7_graph_init_mmio,
+	.mmio = nvd7_graph_pack_mmio,
 	.fecs.ucode = &nvd7_graph_fecs_ucode,
 	.gpccs.ucode = &nvd7_graph_gpccs_ucode,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvd9.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvd9.c
index 652098e0df3..00fdf202fb9 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvd9.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvd9.c
@@ -23,76 +23,70 @@
  */
 
 #include "nvc0.h"
+#include "ctxnvc0.h"
 
 /*******************************************************************************
- * PGRAPH engine/subdev functions
+ * PGRAPH register lists
  ******************************************************************************/
 
-struct nvc0_graph_init
-nvd9_graph_init_unk64xx[] = {
+const struct nvc0_graph_init
+nvd9_graph_init_pd_0[] = {
+	{ 0x406024,   1, 0x04, 0x00000000 },
 	{ 0x4064f0,   3, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nvd9_graph_init_unk58xx[] = {
+const struct nvc0_graph_init
+nvd9_graph_init_ds_0[] = {
 	{ 0x405844,   1, 0x04, 0x00ffffff },
 	{ 0x405850,   1, 0x04, 0x00000000 },
 	{ 0x405900,   1, 0x04, 0x00002834 },
 	{ 0x405908,   1, 0x04, 0x00000000 },
-	{ 0x405928,   1, 0x04, 0x00000000 },
-	{ 0x40592c,   1, 0x04, 0x00000000 },
+	{ 0x405928,   2, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvd9_graph_init_gpc[] = {
+const struct nvc0_graph_init
+nvd9_graph_init_prop_0[] = {
 	{ 0x418408,   1, 0x04, 0x00000000 },
-	{ 0x4184a0,   1, 0x04, 0x00000000 },
-	{ 0x4184a4,   2, 0x04, 0x00000000 },
-	{ 0x418604,   1, 0x04, 0x00000000 },
-	{ 0x418680,   1, 0x04, 0x00000000 },
-	{ 0x418714,   1, 0x04, 0x00000000 },
-	{ 0x418384,   1, 0x04, 0x00000000 },
-	{ 0x418814,   3, 0x04, 0x00000000 },
-	{ 0x418b04,   1, 0x04, 0x00000000 },
-	{ 0x4188c8,   2, 0x04, 0x00000000 },
-	{ 0x4188d0,   1, 0x04, 0x00010000 },
-	{ 0x4188d4,   1, 0x04, 0x00000001 },
-	{ 0x418910,   1, 0x04, 0x00010001 },
-	{ 0x418914,   1, 0x04, 0x00000301 },
-	{ 0x418918,   1, 0x04, 0x00800000 },
-	{ 0x418980,   1, 0x04, 0x77777770 },
-	{ 0x418984,   3, 0x04, 0x77777777 },
+	{ 0x4184a0,   3, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd9_graph_init_gpm_0[] = {
 	{ 0x418c04,   1, 0x04, 0x00000000 },
-	{ 0x418c64,   1, 0x04, 0x00000000 },
-	{ 0x418c68,   1, 0x04, 0x00000000 },
+	{ 0x418c64,   2, 0x04, 0x00000000 },
 	{ 0x418c88,   1, 0x04, 0x00000000 },
 	{ 0x418cb4,   2, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd9_graph_init_gpc_unk_1[] = {
 	{ 0x418d00,   1, 0x04, 0x00000000 },
-	{ 0x418d28,   1, 0x04, 0x00000000 },
-	{ 0x418d2c,   1, 0x04, 0x00000000 },
+	{ 0x418d28,   2, 0x04, 0x00000000 },
 	{ 0x418f00,   1, 0x04, 0x00000000 },
 	{ 0x418f08,   1, 0x04, 0x00000000 },
 	{ 0x418f20,   2, 0x04, 0x00000000 },
 	{ 0x418e00,   1, 0x04, 0x00000003 },
 	{ 0x418e08,   1, 0x04, 0x00000000 },
-	{ 0x418e1c,   1, 0x04, 0x00000000 },
-	{ 0x418e20,   1, 0x04, 0x00000000 },
-	{ 0x41900c,   1, 0x04, 0x00000000 },
-	{ 0x419018,   1, 0x04, 0x00000000 },
+	{ 0x418e1c,   2, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvd9_graph_init_tpc[] = {
-	{ 0x419d08,   2, 0x04, 0x00000000 },
-	{ 0x419d10,   1, 0x04, 0x00000014 },
+const struct nvc0_graph_init
+nvd9_graph_init_tex_0[] = {
 	{ 0x419ab0,   1, 0x04, 0x00000000 },
 	{ 0x419ac8,   1, 0x04, 0x00000000 },
 	{ 0x419ab8,   1, 0x04, 0x000000e7 },
 	{ 0x419abc,   2, 0x04, 0x00000000 },
 	{ 0x419ab4,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd9_graph_init_pe_0[] = {
 	{ 0x41980c,   1, 0x04, 0x00000010 },
 	{ 0x419810,   1, 0x04, 0x00000000 },
 	{ 0x419814,   1, 0x04, 0x00000004 },
@@ -100,20 +94,26 @@ nvd9_graph_init_tpc[] = {
 	{ 0x41984c,   1, 0x04, 0x0000a918 },
 	{ 0x419850,   4, 0x04, 0x00000000 },
 	{ 0x419880,   1, 0x04, 0x00000002 },
-	{ 0x419c98,   1, 0x04, 0x00000000 },
-	{ 0x419ca8,   1, 0x04, 0x80000000 },
-	{ 0x419cb4,   1, 0x04, 0x00000000 },
-	{ 0x419cb8,   1, 0x04, 0x00008bf4 },
-	{ 0x419cbc,   1, 0x04, 0x28137606 },
-	{ 0x419cc0,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd9_graph_init_wwdx_0[] = {
 	{ 0x419bd4,   1, 0x04, 0x00800000 },
 	{ 0x419bdc,   1, 0x04, 0x00000000 },
-	{ 0x419bf8,   1, 0x04, 0x00000000 },
-	{ 0x419bfc,   1, 0x04, 0x00000000 },
+	{ 0x419bf8,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvd9_graph_init_tpccs_1[] = {
 	{ 0x419d2c,   1, 0x04, 0x00000000 },
-	{ 0x419d48,   1, 0x04, 0x00000000 },
-	{ 0x419d4c,   1, 0x04, 0x00000000 },
-	{ 0x419c0c,   1, 0x04, 0x00000000 },
+	{ 0x419d48,   2, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvd9_graph_init_sm_0[] = {
 	{ 0x419e00,   1, 0x04, 0x00000000 },
 	{ 0x419ea0,   1, 0x04, 0x00000000 },
 	{ 0x419ea4,   1, 0x04, 0x00000100 },
@@ -131,23 +131,49 @@ nvd9_graph_init_tpc[] = {
 	{}
 };
 
-static struct nvc0_graph_init *
-nvd9_graph_init_mmio[] = {
-	nvc0_graph_init_regs,
-	nvc0_graph_init_unk40xx,
-	nvc0_graph_init_unk44xx,
-	nvc0_graph_init_unk78xx,
-	nvc0_graph_init_unk60xx,
-	nvd9_graph_init_unk64xx,
-	nvd9_graph_init_unk58xx,
-	nvc0_graph_init_unk80xx,
-	nvd9_graph_init_gpc,
-	nvd9_graph_init_tpc,
-	nvc0_graph_init_unk88xx,
-	nvc0_graph_tpc_0,
-	NULL
+const struct nvc0_graph_init
+nvd9_graph_init_fe_1[] = {
+	{ 0x40402c,   1, 0x04, 0x00000000 },
+	{ 0x4040f0,   1, 0x04, 0x00000000 },
+	{ 0x404174,   1, 0x04, 0x00000000 },
+	{}
 };
 
+static const struct nvc0_graph_pack
+nvd9_graph_pack_mmio[] = {
+	{ nvc0_graph_init_main_0 },
+	{ nvc0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvd9_graph_init_pd_0 },
+	{ nvd9_graph_init_ds_0 },
+	{ nvc0_graph_init_scc_0 },
+	{ nvd9_graph_init_prop_0 },
+	{ nvc1_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ nvc1_graph_init_setup_1 },
+	{ nvc0_graph_init_zcull_0 },
+	{ nvd9_graph_init_gpm_0 },
+	{ nvd9_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ nvc0_graph_init_tpccs_0 },
+	{ nvd9_graph_init_tex_0 },
+	{ nvd9_graph_init_pe_0 },
+	{ nvc0_graph_init_l1c_0 },
+	{ nvd9_graph_init_wwdx_0 },
+	{ nvd9_graph_init_tpccs_1 },
+	{ nvc0_graph_init_mpc_0 },
+	{ nvd9_graph_init_sm_0 },
+	{ nvc0_graph_init_be_0 },
+	{ nvd9_graph_init_fe_1 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
 struct nouveau_oclass *
 nvd9_graph_oclass = &(struct nvc0_graph_oclass) {
 	.base.handle = NV_ENGINE(GR, 0xd9),
@@ -159,7 +185,7 @@ nvd9_graph_oclass = &(struct nvc0_graph_oclass) {
 	},
 	.cclass = &nvd9_grctx_oclass,
 	.sclass = nvc8_graph_sclass,
-	.mmio = nvd9_graph_init_mmio,
+	.mmio = nvd9_graph_pack_mmio,
 	.fecs.ucode = &nvc0_graph_fecs_ucode,
 	.gpccs.ucode = &nvc0_graph_gpccs_ucode,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nve4.c b/drivers/gpu/drm/nouveau/core/engine/graph/nve4.c
index 05ec09c8851..51e0c075ad3 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nve4.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nve4.c
@@ -23,6 +23,7 @@
  */
 
 #include "nvc0.h"
+#include "ctxnvc0.h"
 
 /*******************************************************************************
  * Graphics object classes
@@ -38,11 +39,11 @@ nve4_graph_sclass[] = {
 };
 
 /*******************************************************************************
- * PGRAPH engine/subdev functions
+ * PGRAPH register lists
  ******************************************************************************/
 
-struct nvc0_graph_init
-nve4_graph_init_regs[] = {
+const struct nvc0_graph_init
+nve4_graph_init_main_0[] = {
 	{ 0x400080,   1, 0x04, 0x003083c2 },
 	{ 0x400088,   1, 0x04, 0x0001ffe7 },
 	{ 0x40008c,   1, 0x04, 0x00000000 },
@@ -57,81 +58,59 @@ nve4_graph_init_regs[] = {
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_graph_init_unk58xx[] = {
+static const struct nvc0_graph_init
+nve4_graph_init_ds_0[] = {
 	{ 0x405844,   1, 0x04, 0x00ffffff },
 	{ 0x405850,   1, 0x04, 0x00000000 },
 	{ 0x405900,   1, 0x04, 0x0000ff34 },
 	{ 0x405908,   1, 0x04, 0x00000000 },
-	{ 0x405928,   1, 0x04, 0x00000000 },
-	{ 0x40592c,   1, 0x04, 0x00000000 },
+	{ 0x405928,   2, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_graph_init_unk70xx[] = {
+static const struct nvc0_graph_init
+nve4_graph_init_sked_0[] = {
 	{ 0x407010,   1, 0x04, 0x00000000 },
 	{}
 };
 
-struct nvc0_graph_init
-nve4_graph_init_unk5bxx[] = {
+static const struct nvc0_graph_init
+nve4_graph_init_cwd_0[] = {
 	{ 0x405b50,   1, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_graph_init_gpc[] = {
-	{ 0x418408,   1, 0x04, 0x00000000 },
-	{ 0x4184a0,   1, 0x04, 0x00000000 },
-	{ 0x4184a4,   2, 0x04, 0x00000000 },
-	{ 0x418604,   1, 0x04, 0x00000000 },
-	{ 0x418680,   1, 0x04, 0x00000000 },
-	{ 0x418714,   1, 0x04, 0x00000000 },
-	{ 0x418384,   1, 0x04, 0x00000000 },
-	{ 0x418814,   3, 0x04, 0x00000000 },
-	{ 0x418b04,   1, 0x04, 0x00000000 },
-	{ 0x4188c8,   2, 0x04, 0x00000000 },
-	{ 0x4188d0,   1, 0x04, 0x00010000 },
-	{ 0x4188d4,   1, 0x04, 0x00000001 },
-	{ 0x418910,   1, 0x04, 0x00010001 },
-	{ 0x418914,   1, 0x04, 0x00000301 },
-	{ 0x418918,   1, 0x04, 0x00800000 },
-	{ 0x418980,   1, 0x04, 0x77777770 },
-	{ 0x418984,   3, 0x04, 0x77777777 },
-	{ 0x418c04,   1, 0x04, 0x00000000 },
-	{ 0x418c64,   1, 0x04, 0x00000000 },
-	{ 0x418c68,   1, 0x04, 0x00000000 },
-	{ 0x418c88,   1, 0x04, 0x00000000 },
-	{ 0x418cb4,   2, 0x04, 0x00000000 },
+static const struct nvc0_graph_init
+nve4_graph_init_gpc_unk_1[] = {
 	{ 0x418d00,   1, 0x04, 0x00000000 },
-	{ 0x418d28,   1, 0x04, 0x00000000 },
-	{ 0x418d2c,   1, 0x04, 0x00000000 },
+	{ 0x418d28,   2, 0x04, 0x00000000 },
 	{ 0x418f00,   1, 0x04, 0x00000000 },
 	{ 0x418f08,   1, 0x04, 0x00000000 },
 	{ 0x418f20,   2, 0x04, 0x00000000 },
 	{ 0x418e00,   1, 0x04, 0x00000060 },
 	{ 0x418e08,   1, 0x04, 0x00000000 },
-	{ 0x418e1c,   1, 0x04, 0x00000000 },
-	{ 0x418e20,   1, 0x04, 0x00000000 },
-	{ 0x41900c,   1, 0x04, 0x00000000 },
-	{ 0x419018,   1, 0x04, 0x00000000 },
+	{ 0x418e1c,   2, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nve4_graph_init_tpc[] = {
+const struct nvc0_graph_init
+nve4_graph_init_tpccs_0[] = {
 	{ 0x419d0c,   1, 0x04, 0x00000000 },
 	{ 0x419d10,   1, 0x04, 0x00000014 },
-	{ 0x419ab0,   1, 0x04, 0x00000000 },
-	{ 0x419ac8,   1, 0x04, 0x00000000 },
-	{ 0x419ab8,   1, 0x04, 0x000000e7 },
-	{ 0x419abc,   2, 0x04, 0x00000000 },
-	{ 0x419ab4,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nve4_graph_init_pe_0[] = {
 	{ 0x41980c,   1, 0x04, 0x00000010 },
 	{ 0x419844,   1, 0x04, 0x00000000 },
 	{ 0x419850,   1, 0x04, 0x00000004 },
 	{ 0x419854,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nve4_graph_init_l1c_0[] = {
 	{ 0x419c98,   1, 0x04, 0x00000000 },
 	{ 0x419ca8,   1, 0x04, 0x00000000 },
 	{ 0x419cb0,   1, 0x04, 0x01000000 },
@@ -141,39 +120,25 @@ nve4_graph_init_tpc[] = {
 	{ 0x419cbc,   1, 0x04, 0x28137646 },
 	{ 0x419cc0,   2, 0x04, 0x00000000 },
 	{ 0x419c80,   1, 0x04, 0x00020232 },
-	{ 0x419c0c,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nve4_graph_init_sm_0[] = {
 	{ 0x419e00,   1, 0x04, 0x00000000 },
 	{ 0x419ea0,   1, 0x04, 0x00000000 },
 	{ 0x419ee4,   1, 0x04, 0x00000000 },
 	{ 0x419ea4,   1, 0x04, 0x00000100 },
 	{ 0x419ea8,   1, 0x04, 0x00000000 },
-	{ 0x419eb4,   1, 0x04, 0x00000000 },
-	{ 0x419eb8,   3, 0x04, 0x00000000 },
+	{ 0x419eb4,   4, 0x04, 0x00000000 },
 	{ 0x419edc,   1, 0x04, 0x00000000 },
 	{ 0x419f00,   1, 0x04, 0x00000000 },
 	{ 0x419f74,   1, 0x04, 0x00000555 },
 	{}
 };
 
-struct nvc0_graph_init
-nve4_graph_init_unk[] = {
-	{ 0x41be04,   1, 0x04, 0x00000000 },
-	{ 0x41be08,   1, 0x04, 0x00000004 },
-	{ 0x41be0c,   1, 0x04, 0x00000000 },
-	{ 0x41be10,   1, 0x04, 0x003b8bc7 },
-	{ 0x41be14,   2, 0x04, 0x00000000 },
-	{ 0x41bfd4,   1, 0x04, 0x00800000 },
-	{ 0x41bfdc,   1, 0x04, 0x00000000 },
-	{ 0x41bff8,   1, 0x04, 0x00000000 },
-	{ 0x41bffc,   1, 0x04, 0x00000000 },
-	{ 0x41becc,   1, 0x04, 0x00000000 },
-	{ 0x41bee8,   1, 0x04, 0x00000000 },
-	{ 0x41beec,   1, 0x04, 0x00000000 },
-	{}
-};
-
-struct nvc0_graph_init
-nve4_graph_init_unk88xx[] = {
+const struct nvc0_graph_init
+nve4_graph_init_be_0[] = {
 	{ 0x40880c,   1, 0x04, 0x00000000 },
 	{ 0x408850,   1, 0x04, 0x00000004 },
 	{ 0x408910,   9, 0x04, 0x00000000 },
@@ -186,6 +151,67 @@ nve4_graph_init_unk88xx[] = {
 	{}
 };
 
+const struct nvc0_graph_pack
+nve4_graph_pack_mmio[] = {
+	{ nve4_graph_init_main_0 },
+	{ nvc0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvd9_graph_init_pd_0 },
+	{ nve4_graph_init_ds_0 },
+	{ nvc0_graph_init_scc_0 },
+	{ nve4_graph_init_sked_0 },
+	{ nve4_graph_init_cwd_0 },
+	{ nvd9_graph_init_prop_0 },
+	{ nvc1_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ nvc1_graph_init_setup_1 },
+	{ nvc0_graph_init_zcull_0 },
+	{ nvd9_graph_init_gpm_0 },
+	{ nve4_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ nve4_graph_init_tpccs_0 },
+	{ nvd9_graph_init_tex_0 },
+	{ nve4_graph_init_pe_0 },
+	{ nve4_graph_init_l1c_0 },
+	{ nvc0_graph_init_mpc_0 },
+	{ nve4_graph_init_sm_0 },
+	{ nvd7_graph_init_pes_0 },
+	{ nvd7_graph_init_wwdx_0 },
+	{ nvd7_graph_init_cbm_0 },
+	{ nve4_graph_init_be_0 },
+	{ nvc0_graph_init_fe_1 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
+int
+nve4_graph_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nvc0_graph_priv *priv = (void *)object;
+
+	/*XXX: this is a nasty hack to power on gr on certain boards
+	 *     where it's disabled by therm, somehow.  ideally it'd
+	 *     be nice to know when we should be doing this, and why,
+	 *     but, it's yet to be determined.  for now we test for
+	 *     the particular mmio error that occurs in the situation,
+	 *     and then bash therm in the way nvidia do.
+	 */
+	nv_mask(priv, 0x000200, 0x08001000, 0x08001000);
+	nv_rd32(priv, 0x000200);
+	if (nv_rd32(priv, 0x400700) == 0xbadf1000) {
+		nv_mask(priv, 0x000200, 0x08001000, 0x00000000);
+		nv_rd32(priv, 0x000200);
+		nv_mask(priv, 0x020004, 0xc0000000, 0x40000000);
+	}
+
+	return nouveau_graph_fini(&priv->base, suspend);
+}
+
 int
 nve4_graph_init(struct nouveau_object *object)
 {
@@ -210,8 +236,7 @@ nve4_graph_init(struct nouveau_object *object)
 	nv_wr32(priv, GPC_BCAST(0x08b4), priv->unk4188b4->addr >> 8);
 	nv_wr32(priv, GPC_BCAST(0x08b8), priv->unk4188b8->addr >> 8);
 
-	for (i = 0; oclass->mmio[i]; i++)
-		nvc0_graph_mmio(priv, oclass->mmio[i]);
+	nvc0_graph_mmio(priv, oclass->mmio);
 
 	nv_wr32(priv, GPC_UNIT(0, 0x3018), 0x00000001);
 
@@ -298,25 +323,6 @@ nve4_graph_init(struct nouveau_object *object)
 	return nvc0_graph_init_ctxctl(priv);
 }
 
-static struct nvc0_graph_init *
-nve4_graph_init_mmio[] = {
-	nve4_graph_init_regs,
-	nvc0_graph_init_unk40xx,
-	nvc0_graph_init_unk44xx,
-	nvc0_graph_init_unk78xx,
-	nvc0_graph_init_unk60xx,
-	nvd9_graph_init_unk64xx,
-	nve4_graph_init_unk58xx,
-	nvc0_graph_init_unk80xx,
-	nve4_graph_init_unk70xx,
-	nve4_graph_init_unk5bxx,
-	nve4_graph_init_gpc,
-	nve4_graph_init_tpc,
-	nve4_graph_init_unk,
-	nve4_graph_init_unk88xx,
-	NULL
-};
-
 #include "fuc/hubnve0.fuc.h"
 
 static struct nvc0_graph_ucode
@@ -344,11 +350,11 @@ nve4_graph_oclass = &(struct nvc0_graph_oclass) {
 		.ctor = nvc0_graph_ctor,
 		.dtor = nvc0_graph_dtor,
 		.init = nve4_graph_init,
-		.fini = _nouveau_graph_fini,
+		.fini = nve4_graph_fini,
 	},
 	.cclass = &nve4_grctx_oclass,
 	.sclass = nve4_graph_sclass,
-	.mmio = nve4_graph_init_mmio,
+	.mmio = nve4_graph_pack_mmio,
 	.fecs.ucode = &nve4_graph_fecs_ucode,
 	.gpccs.ucode = &nve4_graph_gpccs_ucode,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvf0.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvf0.c
index 2f0ac783223..c96762122b9 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvf0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvf0.c
@@ -23,6 +23,7 @@
  */
 
 #include "nvc0.h"
+#include "ctxnvc0.h"
 
 /*******************************************************************************
  * Graphics object classes
@@ -38,86 +39,57 @@ nvf0_graph_sclass[] = {
 };
 
 /*******************************************************************************
- * PGRAPH engine/subdev functions
+ * PGRAPH register lists
  ******************************************************************************/
 
-static struct nvc0_graph_init
-nvf0_graph_init_unk40xx[] = {
+const struct nvc0_graph_init
+nvf0_graph_init_fe_0[] = {
 	{ 0x40415c,   1, 0x04, 0x00000000 },
 	{ 0x404170,   1, 0x04, 0x00000000 },
 	{ 0x4041b4,   1, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_graph_init_unk58xx[] = {
+static const struct nvc0_graph_init
+nvf0_graph_init_ds_0[] = {
 	{ 0x405844,   1, 0x04, 0x00ffffff },
 	{ 0x405850,   1, 0x04, 0x00000000 },
 	{ 0x405900,   1, 0x04, 0x0000ff00 },
 	{ 0x405908,   1, 0x04, 0x00000000 },
-	{ 0x405928,   1, 0x04, 0x00000000 },
-	{ 0x40592c,   1, 0x04, 0x00000000 },
+	{ 0x405928,   2, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_graph_init_unk70xx[] = {
+const struct nvc0_graph_init
+nvf0_graph_init_sked_0[] = {
 	{ 0x407010,   1, 0x04, 0x00000000 },
 	{ 0x407040,   1, 0x04, 0x80440424 },
 	{ 0x407048,   1, 0x04, 0x0000000a },
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_graph_init_unk5bxx[] = {
+const struct nvc0_graph_init
+nvf0_graph_init_cwd_0[] = {
 	{ 0x405b44,   1, 0x04, 0x00000000 },
 	{ 0x405b50,   1, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_graph_init_gpc[] = {
-	{ 0x418408,   1, 0x04, 0x00000000 },
-	{ 0x4184a0,   1, 0x04, 0x00000000 },
-	{ 0x4184a4,   2, 0x04, 0x00000000 },
-	{ 0x418604,   1, 0x04, 0x00000000 },
-	{ 0x418680,   1, 0x04, 0x00000000 },
-	{ 0x418714,   1, 0x04, 0x00000000 },
-	{ 0x418384,   1, 0x04, 0x00000000 },
-	{ 0x418814,   3, 0x04, 0x00000000 },
-	{ 0x418b04,   1, 0x04, 0x00000000 },
-	{ 0x4188c8,   2, 0x04, 0x00000000 },
-	{ 0x4188d0,   1, 0x04, 0x00010000 },
-	{ 0x4188d4,   1, 0x04, 0x00000001 },
-	{ 0x418910,   1, 0x04, 0x00010001 },
-	{ 0x418914,   1, 0x04, 0x00000301 },
-	{ 0x418918,   1, 0x04, 0x00800000 },
-	{ 0x418980,   1, 0x04, 0x77777770 },
-	{ 0x418984,   3, 0x04, 0x77777777 },
-	{ 0x418c04,   1, 0x04, 0x00000000 },
-	{ 0x418c64,   1, 0x04, 0x00000000 },
-	{ 0x418c68,   1, 0x04, 0x00000000 },
-	{ 0x418c88,   1, 0x04, 0x00000000 },
-	{ 0x418cb4,   2, 0x04, 0x00000000 },
+const struct nvc0_graph_init
+nvf0_graph_init_gpc_unk_1[] = {
 	{ 0x418d00,   1, 0x04, 0x00000000 },
-	{ 0x418d28,   1, 0x04, 0x00000000 },
-	{ 0x418d2c,   1, 0x04, 0x00000000 },
+	{ 0x418d28,   2, 0x04, 0x00000000 },
 	{ 0x418f00,   1, 0x04, 0x00000400 },
 	{ 0x418f08,   1, 0x04, 0x00000000 },
-	{ 0x418f20,   1, 0x04, 0x00000000 },
-	{ 0x418f24,   1, 0x04, 0x00000000 },
+	{ 0x418f20,   2, 0x04, 0x00000000 },
 	{ 0x418e00,   1, 0x04, 0x00000000 },
 	{ 0x418e08,   1, 0x04, 0x00000000 },
 	{ 0x418e1c,   2, 0x04, 0x00000000 },
-	{ 0x41900c,   1, 0x04, 0x00000000 },
-	{ 0x419018,   1, 0x04, 0x00000000 },
 	{}
 };
 
-static struct nvc0_graph_init
-nvf0_graph_init_tpc[] = {
-	{ 0x419d0c,   1, 0x04, 0x00000000 },
-	{ 0x419d10,   1, 0x04, 0x00000014 },
+static const struct nvc0_graph_init
+nvf0_graph_init_tex_0[] = {
 	{ 0x419ab0,   1, 0x04, 0x00000000 },
 	{ 0x419ac8,   1, 0x04, 0x00000000 },
 	{ 0x419ab8,   1, 0x04, 0x000000e7 },
@@ -125,10 +97,11 @@ nvf0_graph_init_tpc[] = {
 	{ 0x419abc,   2, 0x04, 0x00000000 },
 	{ 0x419ab4,   1, 0x04, 0x00000000 },
 	{ 0x419aa8,   2, 0x04, 0x00000000 },
-	{ 0x41980c,   1, 0x04, 0x00000010 },
-	{ 0x419844,   1, 0x04, 0x00000000 },
-	{ 0x419850,   1, 0x04, 0x00000004 },
-	{ 0x419854,   2, 0x04, 0x00000000 },
+	{}
+};
+
+static const struct nvc0_graph_init
+nvf0_graph_init_l1c_0[] = {
 	{ 0x419c98,   1, 0x04, 0x00000000 },
 	{ 0x419ca8,   1, 0x04, 0x00000000 },
 	{ 0x419cb0,   1, 0x04, 0x01000000 },
@@ -139,7 +112,11 @@ nvf0_graph_init_tpc[] = {
 	{ 0x419cc0,   2, 0x04, 0x00000000 },
 	{ 0x419c80,   1, 0x04, 0x00020230 },
 	{ 0x419ccc,   2, 0x04, 0x00000000 },
-	{ 0x419c0c,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct nvc0_graph_init
+nvf0_graph_init_sm_0[] = {
 	{ 0x419e00,   1, 0x04, 0x00000080 },
 	{ 0x419ea0,   1, 0x04, 0x00000000 },
 	{ 0x419ee4,   1, 0x04, 0x00000000 },
@@ -155,6 +132,44 @@ nvf0_graph_init_tpc[] = {
 	{}
 };
 
+static const struct nvc0_graph_pack
+nvf0_graph_pack_mmio[] = {
+	{ nve4_graph_init_main_0 },
+	{ nvf0_graph_init_fe_0 },
+	{ nvc0_graph_init_pri_0 },
+	{ nvc0_graph_init_rstr2d_0 },
+	{ nvd9_graph_init_pd_0 },
+	{ nvf0_graph_init_ds_0 },
+	{ nvc0_graph_init_scc_0 },
+	{ nvf0_graph_init_sked_0 },
+	{ nvf0_graph_init_cwd_0 },
+	{ nvd9_graph_init_prop_0 },
+	{ nvc1_graph_init_gpc_unk_0 },
+	{ nvc0_graph_init_setup_0 },
+	{ nvc0_graph_init_crstr_0 },
+	{ nvc1_graph_init_setup_1 },
+	{ nvc0_graph_init_zcull_0 },
+	{ nvd9_graph_init_gpm_0 },
+	{ nvf0_graph_init_gpc_unk_1 },
+	{ nvc0_graph_init_gcc_0 },
+	{ nve4_graph_init_tpccs_0 },
+	{ nvf0_graph_init_tex_0 },
+	{ nve4_graph_init_pe_0 },
+	{ nvf0_graph_init_l1c_0 },
+	{ nvc0_graph_init_mpc_0 },
+	{ nvf0_graph_init_sm_0 },
+	{ nvd7_graph_init_pes_0 },
+	{ nvd7_graph_init_wwdx_0 },
+	{ nvd7_graph_init_cbm_0 },
+	{ nve4_graph_init_be_0 },
+	{ nvc0_graph_init_fe_1 },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
 static int
 nvf0_graph_fini(struct nouveau_object *object, bool suspend)
 {
@@ -192,25 +207,6 @@ nvf0_graph_fini(struct nouveau_object *object, bool suspend)
 	return nouveau_graph_fini(&priv->base, suspend);
 }
 
-static struct nvc0_graph_init *
-nvf0_graph_init_mmio[] = {
-	nve4_graph_init_regs,
-	nvf0_graph_init_unk40xx,
-	nvc0_graph_init_unk44xx,
-	nvc0_graph_init_unk78xx,
-	nvc0_graph_init_unk60xx,
-	nvd9_graph_init_unk64xx,
-	nvf0_graph_init_unk58xx,
-	nvc0_graph_init_unk80xx,
-	nvf0_graph_init_unk70xx,
-	nvf0_graph_init_unk5bxx,
-	nvf0_graph_init_gpc,
-	nvf0_graph_init_tpc,
-	nve4_graph_init_unk,
-	nve4_graph_init_unk88xx,
-	NULL
-};
-
 #include "fuc/hubnvf0.fuc.h"
 
 static struct nvc0_graph_ucode
@@ -242,7 +238,7 @@ nvf0_graph_oclass = &(struct nvc0_graph_oclass) {
 	},
 	.cclass = &nvf0_grctx_oclass,
 	.sclass =  nvf0_graph_sclass,
-	.mmio = nvf0_graph_init_mmio,
-	.fecs.ucode = 0 ? &nvf0_graph_fecs_ucode : NULL,
+	.mmio = nvf0_graph_pack_mmio,
+	.fecs.ucode = &nvf0_graph_fecs_ucode,
 	.gpccs.ucode = &nvf0_graph_gpccs_ucode,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/software/nv50.c b/drivers/gpu/drm/nouveau/core/engine/software/nv50.c
index 5ce686ee729..f3b4d9dbf23 100644
--- a/drivers/gpu/drm/nouveau/core/engine/software/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/software/nv50.c
@@ -124,7 +124,7 @@ nv50_software_sclass[] = {
  ******************************************************************************/
 
 static int
-nv50_software_vblsem_release(void *data, int head)
+nv50_software_vblsem_release(void *data, u32 type, int head)
 {
 	struct nv50_software_chan *chan = data;
 	struct nv50_software_priv *priv = (void *)nv_object(chan)->engine;
@@ -183,7 +183,7 @@ nv50_software_context_ctor(struct nouveau_object *parent,
 		return -ENOMEM;
 
 	for (i = 0; i < chan->vblank.nr_event; i++) {
-		ret = nouveau_event_new(pdisp->vblank, i, pclass->vblank,
+		ret = nouveau_event_new(pdisp->vblank, 1, i, pclass->vblank,
 					chan, &chan->vblank.event[i]);
 		if (ret)
 			return ret;
diff --git a/drivers/gpu/drm/nouveau/core/engine/software/nv50.h b/drivers/gpu/drm/nouveau/core/engine/software/nv50.h
index 2de370c2127..bb49a7a2085 100644
--- a/drivers/gpu/drm/nouveau/core/engine/software/nv50.h
+++ b/drivers/gpu/drm/nouveau/core/engine/software/nv50.h
@@ -19,7 +19,7 @@ int  nv50_software_ctor(struct nouveau_object *, struct nouveau_object *,
 
 struct nv50_software_cclass {
 	struct nouveau_oclass base;
-	int (*vblank)(void *, int);
+	int (*vblank)(void *, u32, int);
 };
 
 struct nv50_software_chan {
diff --git a/drivers/gpu/drm/nouveau/core/engine/software/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/software/nvc0.c
index f9430c1bf3e..135c20f3835 100644
--- a/drivers/gpu/drm/nouveau/core/engine/software/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/software/nvc0.c
@@ -104,7 +104,7 @@ nvc0_software_sclass[] = {
  ******************************************************************************/
 
 static int
-nvc0_software_vblsem_release(void *data, int head)
+nvc0_software_vblsem_release(void *data, u32 type, int head)
 {
 	struct nv50_software_chan *chan = data;
 	struct nv50_software_priv *priv = (void *)nv_object(chan)->engine;
diff --git a/drivers/gpu/drm/nouveau/core/engine/xtensa.c b/drivers/gpu/drm/nouveau/core/engine/xtensa.c
index 5f6ede7c489..92384759d2f 100644
--- a/drivers/gpu/drm/nouveau/core/engine/xtensa.c
+++ b/drivers/gpu/drm/nouveau/core/engine/xtensa.c
@@ -112,7 +112,7 @@ _nouveau_xtensa_init(struct nouveau_object *object)
 		snprintf(name, sizeof(name), "nouveau/nv84_xuc%03x",
 			 xtensa->addr >> 12);
 
-		ret = request_firmware(&fw, name, &device->pdev->dev);
+		ret = request_firmware(&fw, name, nv_device_base(device));
 		if (ret) {
 			nv_warn(xtensa, "unable to load firmware %s\n", name);
 			return ret;
diff --git a/drivers/gpu/drm/nouveau/core/include/core/class.h b/drivers/gpu/drm/nouveau/core/include/core/class.h
index 560c3593dae..e0c812bc884 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/class.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/class.h
@@ -230,9 +230,26 @@ struct nve0_channel_ind_class {
 
 #define NV04_DISP_CLASS                                              0x00000046
 
+#define NV04_DISP_MTHD                                               0x00000000
+#define NV04_DISP_MTHD_HEAD                                          0x00000001
+
+#define NV04_DISP_SCANOUTPOS                                         0x00000000
+
 struct nv04_display_class {
 };
 
+struct nv04_display_scanoutpos {
+	s64 time[2];
+	u32 vblanks;
+	u32 vblanke;
+	u32 vtotal;
+	u32 vline;
+	u32 hblanks;
+	u32 hblanke;
+	u32 htotal;
+	u32 hline;
+};
+
 /* 5070: NV50_DISP
  * 8270: NV84_DISP
  * 8370: NVA0_DISP
@@ -241,6 +258,7 @@ struct nv04_display_class {
  * 9070: NVD0_DISP
  * 9170: NVE0_DISP
  * 9270: NVF0_DISP
+ * 9470: GM107_DISP
  */
 
 #define NV50_DISP_CLASS                                              0x00005070
@@ -251,6 +269,12 @@ struct nv04_display_class {
 #define NVD0_DISP_CLASS                                              0x00009070
 #define NVE0_DISP_CLASS                                              0x00009170
 #define NVF0_DISP_CLASS                                              0x00009270
+#define GM107_DISP_CLASS                                             0x00009470
+
+#define NV50_DISP_MTHD                                               0x00000000
+#define NV50_DISP_MTHD_HEAD                                          0x00000003
+
+#define NV50_DISP_SCANOUTPOS                                         0x00000000
 
 #define NV50_DISP_SOR_MTHD                                           0x00010000
 #define NV50_DISP_SOR_MTHD_TYPE                                      0x0000f000
@@ -271,6 +295,10 @@ struct nv04_display_class {
 #define NV84_DISP_SOR_HDMI_PWR_REKEY                                 0x0000007f
 #define NV50_DISP_SOR_LVDS_SCRIPT                                    0x00013000
 #define NV50_DISP_SOR_LVDS_SCRIPT_ID                                 0x0000ffff
+#define NV94_DISP_SOR_DP_PWR                                         0x00016000
+#define NV94_DISP_SOR_DP_PWR_STATE                                   0x00000001
+#define NV94_DISP_SOR_DP_PWR_STATE_OFF                               0x00000000
+#define NV94_DISP_SOR_DP_PWR_STATE_ON                                0x00000001
 
 #define NV50_DISP_DAC_MTHD                                           0x00020000
 #define NV50_DISP_DAC_MTHD_TYPE                                      0x0000f000
@@ -320,6 +348,7 @@ struct nv50_display_class {
  * 907a: NVD0_DISP_CURS
  * 917a: NVE0_DISP_CURS
  * 927a: NVF0_DISP_CURS
+ * 947a: GM107_DISP_CURS
  */
 
 #define NV50_DISP_CURS_CLASS                                         0x0000507a
@@ -330,6 +359,7 @@ struct nv50_display_class {
 #define NVD0_DISP_CURS_CLASS                                         0x0000907a
 #define NVE0_DISP_CURS_CLASS                                         0x0000917a
 #define NVF0_DISP_CURS_CLASS                                         0x0000927a
+#define GM107_DISP_CURS_CLASS                                        0x0000947a
 
 struct nv50_display_curs_class {
 	u32 head;
@@ -343,6 +373,7 @@ struct nv50_display_curs_class {
  * 907b: NVD0_DISP_OIMM
  * 917b: NVE0_DISP_OIMM
  * 927b: NVE0_DISP_OIMM
+ * 947b: GM107_DISP_OIMM
  */
 
 #define NV50_DISP_OIMM_CLASS                                         0x0000507b
@@ -353,6 +384,7 @@ struct nv50_display_curs_class {
 #define NVD0_DISP_OIMM_CLASS                                         0x0000907b
 #define NVE0_DISP_OIMM_CLASS                                         0x0000917b
 #define NVF0_DISP_OIMM_CLASS                                         0x0000927b
+#define GM107_DISP_OIMM_CLASS                                        0x0000947b
 
 struct nv50_display_oimm_class {
 	u32 head;
@@ -366,6 +398,7 @@ struct nv50_display_oimm_class {
  * 907c: NVD0_DISP_SYNC
  * 917c: NVE0_DISP_SYNC
  * 927c: NVF0_DISP_SYNC
+ * 947c: GM107_DISP_SYNC
  */
 
 #define NV50_DISP_SYNC_CLASS                                         0x0000507c
@@ -376,6 +409,7 @@ struct nv50_display_oimm_class {
 #define NVD0_DISP_SYNC_CLASS                                         0x0000907c
 #define NVE0_DISP_SYNC_CLASS                                         0x0000917c
 #define NVF0_DISP_SYNC_CLASS                                         0x0000927c
+#define GM107_DISP_SYNC_CLASS                                        0x0000947c
 
 struct nv50_display_sync_class {
 	u32 pushbuf;
@@ -390,6 +424,7 @@ struct nv50_display_sync_class {
  * 907d: NVD0_DISP_MAST
  * 917d: NVE0_DISP_MAST
  * 927d: NVF0_DISP_MAST
+ * 947d: GM107_DISP_MAST
  */
 
 #define NV50_DISP_MAST_CLASS                                         0x0000507d
@@ -400,6 +435,7 @@ struct nv50_display_sync_class {
 #define NVD0_DISP_MAST_CLASS                                         0x0000907d
 #define NVE0_DISP_MAST_CLASS                                         0x0000917d
 #define NVF0_DISP_MAST_CLASS                                         0x0000927d
+#define GM107_DISP_MAST_CLASS                                        0x0000947d
 
 struct nv50_display_mast_class {
 	u32 pushbuf;
@@ -413,6 +449,7 @@ struct nv50_display_mast_class {
  * 907e: NVD0_DISP_OVLY
  * 917e: NVE0_DISP_OVLY
  * 927e: NVF0_DISP_OVLY
+ * 947e: GM107_DISP_OVLY
  */
 
 #define NV50_DISP_OVLY_CLASS                                         0x0000507e
@@ -423,6 +460,7 @@ struct nv50_display_mast_class {
 #define NVD0_DISP_OVLY_CLASS                                         0x0000907e
 #define NVE0_DISP_OVLY_CLASS                                         0x0000917e
 #define NVF0_DISP_OVLY_CLASS                                         0x0000927e
+#define GM107_DISP_OVLY_CLASS                                        0x0000947e
 
 struct nv50_display_ovly_class {
 	u32 pushbuf;
diff --git a/drivers/gpu/drm/nouveau/core/include/core/device.h b/drivers/gpu/drm/nouveau/core/include/core/device.h
index ac2881d1776..a8a9a9cf16c 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/device.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/device.h
@@ -38,7 +38,9 @@ enum nv_subdev_type {
 	NVDEV_SUBDEV_THERM,
 	NVDEV_SUBDEV_CLOCK,
 
-	NVDEV_ENGINE_DMAOBJ,
+	NVDEV_ENGINE_FIRST,
+	NVDEV_ENGINE_DMAOBJ = NVDEV_ENGINE_FIRST,
+	NVDEV_ENGINE_IFB,
 	NVDEV_ENGINE_FIFO,
 	NVDEV_ENGINE_SW,
 	NVDEV_ENGINE_GR,
@@ -64,12 +66,14 @@ struct nouveau_device {
 	struct list_head head;
 
 	struct pci_dev *pdev;
+	struct platform_device *platformdev;
 	u64 handle;
 
 	const char *cfgopt;
 	const char *dbgopt;
 	const char *name;
 	const char *cname;
+	u64 disable_mask;
 
 	enum {
 		NV_04    = 0x04,
@@ -82,6 +86,7 @@ struct nouveau_device {
 		NV_C0    = 0xc0,
 		NV_D0    = 0xd0,
 		NV_E0    = 0xe0,
+		GM100    = 0x110,
 	} card_type;
 	u32 chipset;
 	u32 crystal;
@@ -138,4 +143,32 @@ nv_device_match(struct nouveau_object *object, u16 dev, u16 ven, u16 sub)
 	       device->pdev->subsystem_device == sub;
 }
 
+static inline bool
+nv_device_is_pci(struct nouveau_device *device)
+{
+	return device->pdev != NULL;
+}
+
+static inline struct device *
+nv_device_base(struct nouveau_device *device)
+{
+	return nv_device_is_pci(device) ? &device->pdev->dev :
+					  &device->platformdev->dev;
+}
+
+resource_size_t
+nv_device_resource_start(struct nouveau_device *device, unsigned int bar);
+
+resource_size_t
+nv_device_resource_len(struct nouveau_device *device, unsigned int bar);
+
+dma_addr_t
+nv_device_map_page(struct nouveau_device *device, struct page *page);
+
+void
+nv_device_unmap_page(struct nouveau_device *device, dma_addr_t addr);
+
+int
+nv_device_get_irq(struct nouveau_device *device, bool stall);
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/core/event.h b/drivers/gpu/drm/nouveau/core/include/core/event.h
index 5d539ebff3e..ba3f1a76a81 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/event.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/event.h
@@ -12,32 +12,33 @@ struct nouveau_eventh {
 	struct nouveau_event *event;
 	struct list_head head;
 	unsigned long flags;
+	u32 types;
 	int index;
-	int (*func)(void *, int);
+	int (*func)(void *, u32, int);
 	void *priv;
 };
 
 struct nouveau_event {
-	spinlock_t list_lock;
-	spinlock_t refs_lock;
-
 	void *priv;
-	void (*enable)(struct nouveau_event *, int index);
-	void (*disable)(struct nouveau_event *, int index);
+	int (*check)(struct nouveau_event *, u32 type, int index);
+	void (*enable)(struct nouveau_event *, int type, int index);
+	void (*disable)(struct nouveau_event *, int type, int index);
 
+	int types_nr;
 	int index_nr;
-	struct {
-		struct list_head list;
-		int refs;
-	} index[];
+
+	spinlock_t list_lock;
+	struct list_head *list;
+	spinlock_t refs_lock;
+	int refs[];
 };
 
-int  nouveau_event_create(int index_nr, struct nouveau_event **);
+int  nouveau_event_create(int types_nr, int index_nr, struct nouveau_event **);
 void nouveau_event_destroy(struct nouveau_event **);
-void nouveau_event_trigger(struct nouveau_event *, int index);
+void nouveau_event_trigger(struct nouveau_event *, u32 types, int index);
 
-int  nouveau_event_new(struct nouveau_event *, int index,
-		       int (*func)(void *, int), void *,
+int  nouveau_event_new(struct nouveau_event *, u32 types, int index,
+		       int (*func)(void *, u32, int), void *,
 		       struct nouveau_eventh **);
 void nouveau_event_ref(struct nouveau_eventh *, struct nouveau_eventh **);
 void nouveau_event_get(struct nouveau_eventh *);
diff --git a/drivers/gpu/drm/nouveau/core/include/core/namedb.h b/drivers/gpu/drm/nouveau/core/include/core/namedb.h
index 8897e088608..f5b5fd8e1fc 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/namedb.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/namedb.h
@@ -33,7 +33,7 @@ nv_namedb(void *obj)
 
 int  nouveau_namedb_create_(struct nouveau_object *, struct nouveau_object *,
 			    struct nouveau_oclass *, u32 pclass,
-			    struct nouveau_oclass *, u32 engcls,
+			    struct nouveau_oclass *, u64 engcls,
 			    int size, void **);
 
 int  _nouveau_namedb_ctor(struct nouveau_object *, struct nouveau_object *,
diff --git a/drivers/gpu/drm/nouveau/core/include/engine/device.h b/drivers/gpu/drm/nouveau/core/include/engine/device.h
index b3dd2c4c2f1..672d3c8f414 100644
--- a/drivers/gpu/drm/nouveau/core/include/engine/device.h
+++ b/drivers/gpu/drm/nouveau/core/include/engine/device.h
@@ -3,11 +3,20 @@
 
 #include <core/device.h>
 
-#define nouveau_device_create(p,n,s,c,d,u)                                     \
-	nouveau_device_create_((p), (n), (s), (c), (d), sizeof(**u), (void **)u)
+struct platform_device;
 
-int  nouveau_device_create_(struct pci_dev *, u64 name, const char *sname,
-			    const char *cfg, const char *dbg, int, void **);
+enum nv_bus_type {
+	NOUVEAU_BUS_PCI,
+	NOUVEAU_BUS_PLATFORM,
+};
+
+#define nouveau_device_create(p,t,n,s,c,d,u)                                   \
+	nouveau_device_create_((void *)(p), (t), (n), (s), (c), (d),           \
+			       sizeof(**u), (void **)u)
+
+int  nouveau_device_create_(void *, enum nv_bus_type type, u64 name,
+			    const char *sname, const char *cfg, const char *dbg,
+			    int, void **);
 
 int nv04_identify(struct nouveau_device *);
 int nv10_identify(struct nouveau_device *);
@@ -17,6 +26,7 @@ int nv40_identify(struct nouveau_device *);
 int nv50_identify(struct nouveau_device *);
 int nvc0_identify(struct nouveau_device *);
 int nve0_identify(struct nouveau_device *);
+int gm100_identify(struct nouveau_device *);
 
 struct nouveau_device *nouveau_device_find(u64 name);
 
diff --git a/drivers/gpu/drm/nouveau/core/include/engine/disp.h b/drivers/gpu/drm/nouveau/core/include/engine/disp.h
index 4b21fabfbdd..fde84289680 100644
--- a/drivers/gpu/drm/nouveau/core/include/engine/disp.h
+++ b/drivers/gpu/drm/nouveau/core/include/engine/disp.h
@@ -6,8 +6,19 @@
 #include <core/device.h>
 #include <core/event.h>
 
+enum nvkm_hpd_event {
+	NVKM_HPD_PLUG = 1,
+	NVKM_HPD_UNPLUG = 2,
+	NVKM_HPD_IRQ = 4,
+	NVKM_HPD = (NVKM_HPD_PLUG | NVKM_HPD_UNPLUG | NVKM_HPD_IRQ)
+};
+
 struct nouveau_disp {
 	struct nouveau_engine base;
+
+	struct list_head outp;
+	struct nouveau_event *hpd;
+
 	struct nouveau_event *vblank;
 };
 
@@ -17,33 +28,15 @@ nouveau_disp(void *obj)
 	return (void *)nv_device(obj)->subdev[NVDEV_ENGINE_DISP];
 }
 
-#define nouveau_disp_create(p,e,c,h,i,x,d)                                     \
-	nouveau_disp_create_((p), (e), (c), (h), (i), (x),                     \
-			     sizeof(**d), (void **)d)
-#define nouveau_disp_destroy(d) ({                                             \
-	struct nouveau_disp *disp = (d);                                       \
-	_nouveau_disp_dtor(nv_object(disp));                                   \
-})
-#define nouveau_disp_init(d)                                                   \
-	nouveau_engine_init(&(d)->base)
-#define nouveau_disp_fini(d,s)                                                 \
-	nouveau_engine_fini(&(d)->base, (s))
-
-int  nouveau_disp_create_(struct nouveau_object *, struct nouveau_object *,
-			  struct nouveau_oclass *, int heads,
-			  const char *, const char *, int, void **);
-void _nouveau_disp_dtor(struct nouveau_object *);
-#define _nouveau_disp_init _nouveau_engine_init
-#define _nouveau_disp_fini _nouveau_engine_fini
-
-extern struct nouveau_oclass nv04_disp_oclass;
-extern struct nouveau_oclass nv50_disp_oclass;
-extern struct nouveau_oclass nv84_disp_oclass;
-extern struct nouveau_oclass nva0_disp_oclass;
-extern struct nouveau_oclass nv94_disp_oclass;
-extern struct nouveau_oclass nva3_disp_oclass;
-extern struct nouveau_oclass nvd0_disp_oclass;
-extern struct nouveau_oclass nve0_disp_oclass;
-extern struct nouveau_oclass nvf0_disp_oclass;
+extern struct nouveau_oclass *nv04_disp_oclass;
+extern struct nouveau_oclass *nv50_disp_oclass;
+extern struct nouveau_oclass *nv84_disp_oclass;
+extern struct nouveau_oclass *nva0_disp_oclass;
+extern struct nouveau_oclass *nv94_disp_oclass;
+extern struct nouveau_oclass *nva3_disp_oclass;
+extern struct nouveau_oclass *nvd0_disp_oclass;
+extern struct nouveau_oclass *nve0_disp_oclass;
+extern struct nouveau_oclass *nvf0_disp_oclass;
+extern struct nouveau_oclass *gm107_disp_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/engine/fifo.h b/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
index 8c32cf4d83c..b639eb2c74f 100644
--- a/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
+++ b/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
@@ -109,6 +109,8 @@ extern struct nouveau_oclass *nv50_fifo_oclass;
 extern struct nouveau_oclass *nv84_fifo_oclass;
 extern struct nouveau_oclass *nvc0_fifo_oclass;
 extern struct nouveau_oclass *nve0_fifo_oclass;
+extern struct nouveau_oclass *gk20a_fifo_oclass;
+extern struct nouveau_oclass *nv108_fifo_oclass;
 
 void nv04_fifo_intr(struct nouveau_subdev *);
 int  nv04_fifo_context_attach(struct nouveau_object *, struct nouveau_object *);
diff --git a/drivers/gpu/drm/nouveau/core/include/engine/graph.h b/drivers/gpu/drm/nouveau/core/include/engine/graph.h
index 8e1b52312dd..8c1d4772da0 100644
--- a/drivers/gpu/drm/nouveau/core/include/engine/graph.h
+++ b/drivers/gpu/drm/nouveau/core/include/engine/graph.h
@@ -63,12 +63,15 @@ extern struct nouveau_oclass nv40_graph_oclass;
 extern struct nouveau_oclass nv50_graph_oclass;
 extern struct nouveau_oclass *nvc0_graph_oclass;
 extern struct nouveau_oclass *nvc1_graph_oclass;
-extern struct nouveau_oclass *nvc3_graph_oclass;
+extern struct nouveau_oclass *nvc4_graph_oclass;
 extern struct nouveau_oclass *nvc8_graph_oclass;
 extern struct nouveau_oclass *nvd7_graph_oclass;
 extern struct nouveau_oclass *nvd9_graph_oclass;
 extern struct nouveau_oclass *nve4_graph_oclass;
+extern struct nouveau_oclass *gk20a_graph_oclass;
 extern struct nouveau_oclass *nvf0_graph_oclass;
+extern struct nouveau_oclass *nv108_graph_oclass;
+extern struct nouveau_oclass *gm107_graph_oclass;
 
 extern const struct nouveau_bitfield nv04_graph_nsource[];
 extern struct nouveau_ofuncs nv04_graph_ofuncs;
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bar.h b/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
index 4f4ff4502c3..9faa98e67ad 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
@@ -4,8 +4,7 @@
 #include <core/subdev.h>
 #include <core/device.h>
 
-#include <subdev/fb.h>
-
+struct nouveau_mem;
 struct nouveau_vma;
 
 struct nouveau_bar {
@@ -29,27 +28,7 @@ nouveau_bar(void *obj)
 	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_BAR];
 }
 
-#define nouveau_bar_create(p,e,o,d)                                            \
-	nouveau_bar_create_((p), (e), (o), sizeof(**d), (void **)d)
-#define nouveau_bar_init(p)                                                    \
-	nouveau_subdev_init(&(p)->base)
-#define nouveau_bar_fini(p,s)                                                  \
-	nouveau_subdev_fini(&(p)->base, (s))
-
-int nouveau_bar_create_(struct nouveau_object *, struct nouveau_object *,
-			struct nouveau_oclass *, int, void **);
-void nouveau_bar_destroy(struct nouveau_bar *);
-
-void _nouveau_bar_dtor(struct nouveau_object *);
-#define _nouveau_bar_init _nouveau_subdev_init
-#define _nouveau_bar_fini _nouveau_subdev_fini
-
 extern struct nouveau_oclass nv50_bar_oclass;
 extern struct nouveau_oclass nvc0_bar_oclass;
 
-int nouveau_bar_alloc(struct nouveau_bar *, struct nouveau_object *,
-		      struct nouveau_mem *, struct nouveau_object **);
-
-void nv84_bar_flush(struct nouveau_bar *);
-
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/P0260.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/P0260.h
new file mode 100644
index 00000000000..bba01ab1e04
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/P0260.h
@@ -0,0 +1,23 @@
+#ifndef __NVBIOS_P0260_H__
+#define __NVBIOS_P0260_H__
+
+u32 nvbios_P0260Te(struct nouveau_bios *,
+		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *xnr, u8 *xsz);
+
+struct nvbios_P0260E {
+	u32 data;
+};
+
+u32 nvbios_P0260Ee(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr);
+u32 nvbios_P0260Ep(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr,
+		   struct nvbios_P0260E *);
+
+struct nvbios_P0260X {
+	u32 data;
+};
+
+u32 nvbios_P0260Xe(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr);
+u32 nvbios_P0260Xp(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr,
+		   struct nvbios_P0260X *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/conn.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/conn.h
index c1270548fd0..f3930c27cb7 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/conn.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/conn.h
@@ -16,12 +16,31 @@ enum dcb_connector_type {
 	DCB_CONNECTOR_eDP = 0x47,
 	DCB_CONNECTOR_HDMI_0 = 0x60,
 	DCB_CONNECTOR_HDMI_1 = 0x61,
+	DCB_CONNECTOR_HDMI_C = 0x63,
 	DCB_CONNECTOR_DMS59_DP0 = 0x64,
 	DCB_CONNECTOR_DMS59_DP1 = 0x65,
 	DCB_CONNECTOR_NONE = 0xff
 };
 
-u16 dcb_conntab(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
-u16 dcb_conn(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *len);
+struct nvbios_connT {
+};
+
+u32 nvbios_connTe(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_connTp(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		  struct nvbios_connT *info);
+
+struct nvbios_connE {
+	u8 type;
+	u8 location;
+	u8 hpd;
+	u8 dp;
+	u8 di;
+	u8 sr;
+	u8 lcdid;
+};
+
+u32 nvbios_connEe(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *hdr);
+u32 nvbios_connEp(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *hdr,
+		  struct nvbios_connE *info);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/dp.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/dp.h
index 6e54218b55f..728206e2177 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/dp.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/dp.h
@@ -17,9 +17,10 @@ u16 nvbios_dpout_match(struct nouveau_bios *, u16 type, u16 mask,
 		       struct nvbios_dpout *);
 
 struct nvbios_dpcfg {
-	u8 drv;
-	u8 pre;
-	u8 unk;
+	u8 pc;
+	u8 dc;
+	u8 pe;
+	u8 tx_pu;
 };
 
 u16
@@ -27,7 +28,7 @@ nvbios_dpcfg_parse(struct nouveau_bios *, u16 outp, u8 idx,
 		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
 		   struct nvbios_dpcfg *);
 u16
-nvbios_dpcfg_match(struct nouveau_bios *, u16 outp, u8 un, u8 vs, u8 pe,
+nvbios_dpcfg_match(struct nouveau_bios *, u16 outp, u8 pc, u8 vs, u8 pe,
 		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
 		   struct nvbios_dpcfg *);
 
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h
new file mode 100644
index 00000000000..c086ac6d677
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h
@@ -0,0 +1,66 @@
+#ifndef __NVBIOS_RAMCFG_H__
+#define __NVBIOS_RAMCFG_H__
+
+struct nouveau_bios;
+
+struct nvbios_ramcfg {
+	unsigned rammap_11_08_01:1;
+	unsigned rammap_11_08_0c:2;
+	unsigned rammap_11_08_10:1;
+	unsigned rammap_11_11_0c:2;
+
+	unsigned ramcfg_11_01_01:1;
+	unsigned ramcfg_11_01_02:1;
+	unsigned ramcfg_11_01_04:1;
+	unsigned ramcfg_11_01_08:1;
+	unsigned ramcfg_11_01_10:1;
+	unsigned ramcfg_11_01_20:1;
+	unsigned ramcfg_11_01_40:1;
+	unsigned ramcfg_11_01_80:1;
+	unsigned ramcfg_11_02_03:2;
+	unsigned ramcfg_11_02_04:1;
+	unsigned ramcfg_11_02_08:1;
+	unsigned ramcfg_11_02_10:1;
+	unsigned ramcfg_11_02_40:1;
+	unsigned ramcfg_11_02_80:1;
+	unsigned ramcfg_11_03_0f:4;
+	unsigned ramcfg_11_03_30:2;
+	unsigned ramcfg_11_03_c0:2;
+	unsigned ramcfg_11_03_f0:4;
+	unsigned ramcfg_11_04:8;
+	unsigned ramcfg_11_06:8;
+	unsigned ramcfg_11_07_02:1;
+	unsigned ramcfg_11_07_04:1;
+	unsigned ramcfg_11_07_08:1;
+	unsigned ramcfg_11_07_10:1;
+	unsigned ramcfg_11_07_40:1;
+	unsigned ramcfg_11_07_80:1;
+	unsigned ramcfg_11_08_01:1;
+	unsigned ramcfg_11_08_02:1;
+	unsigned ramcfg_11_08_04:1;
+	unsigned ramcfg_11_08_08:1;
+	unsigned ramcfg_11_08_10:1;
+	unsigned ramcfg_11_08_20:1;
+	unsigned ramcfg_11_09:8;
+
+	unsigned timing[11];
+	unsigned timing_20_2e_03:2;
+	unsigned timing_20_2e_30:2;
+	unsigned timing_20_2e_c0:2;
+	unsigned timing_20_2f_03:2;
+	unsigned timing_20_2c_003f:6;
+	unsigned timing_20_2c_1fc0:7;
+	unsigned timing_20_30_f8:5;
+	unsigned timing_20_30_07:3;
+	unsigned timing_20_31_0007:3;
+	unsigned timing_20_31_0078:4;
+	unsigned timing_20_31_0780:4;
+	unsigned timing_20_31_0800:1;
+	unsigned timing_20_31_7000:3;
+	unsigned timing_20_31_8000:1;
+};
+
+u8 nvbios_ramcfg_count(struct nouveau_bios *);
+u8 nvbios_ramcfg_index(struct nouveau_subdev *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
index bc15e032087..5bdf8e4db40 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
@@ -1,11 +1,25 @@
 #ifndef __NVBIOS_RAMMAP_H__
 #define __NVBIOS_RAMMAP_H__
 
-u16 nvbios_rammap_table(struct nouveau_bios *, u8 *ver, u8 *hdr,
-			u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
-u16 nvbios_rammap_entry(struct nouveau_bios *, int idx,
-			u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
-u16 nvbios_rammap_match(struct nouveau_bios *, u16 khz,
-			u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+struct nvbios_ramcfg;
+
+u32 nvbios_rammapTe(struct nouveau_bios *, u8 *ver, u8 *hdr,
+		    u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
+
+u32 nvbios_rammapEe(struct nouveau_bios *, int idx,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_rammapEm(struct nouveau_bios *, u16 mhz,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u32 nvbios_rammapEp(struct nouveau_bios *, u16 mhz,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		    struct nvbios_ramcfg *);
+
+u32 nvbios_rammapSe(struct nouveau_bios *, u32 data,
+		    u8 ever, u8 ehdr, u8 ecnt, u8 elen, int idx,
+		    u8 *ver, u8 *hdr);
+u32 nvbios_rammapSp(struct nouveau_bios *, u32 data,
+		    u8 ever, u8 ehdr, u8 ecnt, u8 elen, int idx,
+		    u8 *ver, u8 *hdr,
+		    struct nvbios_ramcfg *);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h
index 083541dbe9c..8dc5051df55 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/therm.h
@@ -31,6 +31,12 @@ struct nouveau_therm_trip_point {
 	int hysteresis;
 };
 
+enum nvbios_therm_fan_mode {
+	NVBIOS_THERM_FAN_TRIP = 0,
+	NVBIOS_THERM_FAN_LINEAR = 1,
+	NVBIOS_THERM_FAN_OTHER = 2,
+};
+
 struct nvbios_therm_fan {
 	u16 pwm_freq;
 
@@ -40,6 +46,7 @@ struct nvbios_therm_fan {
 	u16 bump_period;
 	u16 slow_down_period;
 
+	enum nvbios_therm_fan_mode fan_mode;
 	struct nouveau_therm_trip_point trip[NOUVEAU_TEMP_FAN_TRIP_MAX];
 	u8 nr_fan_trip;
 	u8 linear_min_temp;
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/timing.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/timing.h
index 963694b5422..76d914b67ab 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/timing.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/timing.h
@@ -1,8 +1,14 @@
 #ifndef __NVBIOS_TIMING_H__
 #define __NVBIOS_TIMING_H__
 
-u16 nvbios_timing_table(struct nouveau_bios *,
-			u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
-u16 nvbios_timing_entry(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr);
+struct nvbios_ramcfg;
+
+u16 nvbios_timingTe(struct nouveau_bios *,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
+u16 nvbios_timingEe(struct nouveau_bios *, int idx,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u16 nvbios_timingEp(struct nouveau_bios *, int idx,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		    struct nvbios_ramcfg *);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/clock.h b/drivers/gpu/drm/nouveau/core/include/subdev/clock.h
index 8f4ced75444..c01e29c9f89 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/clock.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/clock.h
@@ -77,6 +77,8 @@ struct nouveau_clock {
 	int tstate; /* thermal adjustment (max-) */
 	int dstate; /* display adjustment (min+) */
 
+	bool allow_reclock;
+
 	int  (*read)(struct nouveau_clock *, enum nv_clk_src);
 	int  (*calc)(struct nouveau_clock *, struct nouveau_cstate *);
 	int  (*prog)(struct nouveau_clock *);
@@ -106,8 +108,8 @@ struct nouveau_clocks {
 	int mdiv;
 };
 
-#define nouveau_clock_create(p,e,o,i,d)                                        \
-	nouveau_clock_create_((p), (e), (o), (i), sizeof(**d), (void **)d)
+#define nouveau_clock_create(p,e,o,i,r,d)                                      \
+	nouveau_clock_create_((p), (e), (o), (i), (r), sizeof(**d), (void **)d)
 #define nouveau_clock_destroy(p) ({                                            \
 	struct nouveau_clock *clk = (p);                                       \
 	_nouveau_clock_dtor(nv_object(clk));                                   \
@@ -121,7 +123,7 @@ struct nouveau_clocks {
 
 int  nouveau_clock_create_(struct nouveau_object *, struct nouveau_object *,
 			   struct nouveau_oclass *,
-			   struct nouveau_clocks *, int, void **);
+			   struct nouveau_clocks *, bool, int, void **);
 void _nouveau_clock_dtor(struct nouveau_object *);
 int _nouveau_clock_init(struct nouveau_object *);
 #define _nouveau_clock_fini _nouveau_subdev_fini
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h b/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h
index 685c9b12ee4..e292271a84e 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h
@@ -9,7 +9,7 @@ struct nouveau_devinit {
 	bool post;
 	void (*meminit)(struct nouveau_devinit *);
 	int  (*pll_set)(struct nouveau_devinit *, u32 type, u32 freq);
-
+	u32  (*mmio)(struct nouveau_devinit *, u32 addr);
 };
 
 static inline struct nouveau_devinit *
@@ -18,32 +18,17 @@ nouveau_devinit(void *obj)
 	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_DEVINIT];
 }
 
-#define nouveau_devinit_create(p,e,o,d)                                        \
-	nouveau_devinit_create_((p), (e), (o), sizeof(**d), (void **)d)
-#define nouveau_devinit_destroy(p)                                             \
-	nouveau_subdev_destroy(&(p)->base)
-#define nouveau_devinit_init(p) ({                                             \
-	struct nouveau_devinit *d = (p);                                       \
-	_nouveau_devinit_init(nv_object(d));                                   \
-})
-#define nouveau_devinit_fini(p,s) ({                                           \
-	struct nouveau_devinit *d = (p);                                       \
-	_nouveau_devinit_fini(nv_object(d), (s));                              \
-})
-
-int nouveau_devinit_create_(struct nouveau_object *, struct nouveau_object *,
-			    struct nouveau_oclass *, int, void **);
-#define _nouveau_devinit_dtor _nouveau_subdev_dtor
-int _nouveau_devinit_init(struct nouveau_object *);
-int _nouveau_devinit_fini(struct nouveau_object *, bool suspend);
-
-extern struct nouveau_oclass nv04_devinit_oclass;
-extern struct nouveau_oclass nv05_devinit_oclass;
-extern struct nouveau_oclass nv10_devinit_oclass;
-extern struct nouveau_oclass nv1a_devinit_oclass;
-extern struct nouveau_oclass nv20_devinit_oclass;
-extern struct nouveau_oclass nv50_devinit_oclass;
-extern struct nouveau_oclass nva3_devinit_oclass;
-extern struct nouveau_oclass nvc0_devinit_oclass;
+extern struct nouveau_oclass *nv04_devinit_oclass;
+extern struct nouveau_oclass *nv05_devinit_oclass;
+extern struct nouveau_oclass *nv10_devinit_oclass;
+extern struct nouveau_oclass *nv1a_devinit_oclass;
+extern struct nouveau_oclass *nv20_devinit_oclass;
+extern struct nouveau_oclass *nv50_devinit_oclass;
+extern struct nouveau_oclass *nv84_devinit_oclass;
+extern struct nouveau_oclass *nv98_devinit_oclass;
+extern struct nouveau_oclass *nva3_devinit_oclass;
+extern struct nouveau_oclass *nvaf_devinit_oclass;
+extern struct nouveau_oclass *nvc0_devinit_oclass;
+extern struct nouveau_oclass *gm107_devinit_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/fb.h b/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
index d89dbdf39b0..871e73914b2 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
@@ -105,6 +105,15 @@ extern struct nouveau_oclass *nvaa_fb_oclass;
 extern struct nouveau_oclass *nvaf_fb_oclass;
 extern struct nouveau_oclass *nvc0_fb_oclass;
 extern struct nouveau_oclass *nve0_fb_oclass;
+extern struct nouveau_oclass *gk20a_fb_oclass;
+extern struct nouveau_oclass *gm107_fb_oclass;
+
+#include <subdev/bios/ramcfg.h>
+
+struct nouveau_ram_data {
+	struct nvbios_ramcfg bios;
+	u32 freq;
+};
 
 struct nouveau_ram {
 	struct nouveau_object base;
@@ -142,6 +151,12 @@ struct nouveau_ram {
 	} rammap, ramcfg, timing;
 	u32 freq;
 	u32 mr[16];
+	u32 mr1_nuts;
+
+	struct nouveau_ram_data *next;
+	struct nouveau_ram_data former;
+	struct nouveau_ram_data xition;
+	struct nouveau_ram_data target;
 };
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h b/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h
index c85b9f1579a..612d82ab683 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/gpio.h
@@ -8,17 +8,18 @@
 #include <subdev/bios.h>
 #include <subdev/bios/gpio.h>
 
+enum nvkm_gpio_event {
+	NVKM_GPIO_HI = 1,
+	NVKM_GPIO_LO = 2,
+	NVKM_GPIO_TOGGLED = (NVKM_GPIO_HI | NVKM_GPIO_LO),
+};
+
 struct nouveau_gpio {
 	struct nouveau_subdev base;
 
 	struct nouveau_event *events;
 
-	/* hardware interfaces */
 	void (*reset)(struct nouveau_gpio *, u8 func);
-	int  (*drive)(struct nouveau_gpio *, int line, int dir, int out);
-	int  (*sense)(struct nouveau_gpio *, int line);
-
-	/* software interfaces */
 	int  (*find)(struct nouveau_gpio *, int idx, u8 tag, u8 line,
 		     struct dcb_gpio_func *);
 	int  (*set)(struct nouveau_gpio *, int idx, u8 tag, u8 line, int state);
@@ -31,23 +32,10 @@ nouveau_gpio(void *obj)
 	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_GPIO];
 }
 
-#define nouveau_gpio_create(p,e,o,l,d)                                         \
-	nouveau_gpio_create_((p), (e), (o), (l), sizeof(**d), (void **)d)
-#define nouveau_gpio_destroy(p) ({                                             \
-	struct nouveau_gpio *gpio = (p);                                       \
-	_nouveau_gpio_dtor(nv_object(gpio));                                   \
-})
-#define nouveau_gpio_fini(p,s)                                                 \
-	nouveau_subdev_fini(&(p)->base, (s))
-
-int  nouveau_gpio_create_(struct nouveau_object *, struct nouveau_object *,
-			  struct nouveau_oclass *, int, int, void **);
-void _nouveau_gpio_dtor(struct nouveau_object *);
-int  nouveau_gpio_init(struct nouveau_gpio *);
-
-extern struct nouveau_oclass nv10_gpio_oclass;
-extern struct nouveau_oclass nv50_gpio_oclass;
-extern struct nouveau_oclass nvd0_gpio_oclass;
-extern struct nouveau_oclass nve0_gpio_oclass;
+extern struct nouveau_oclass *nv10_gpio_oclass;
+extern struct nouveau_oclass *nv50_gpio_oclass;
+extern struct nouveau_oclass *nv92_gpio_oclass;
+extern struct nouveau_oclass *nvd0_gpio_oclass;
+extern struct nouveau_oclass *nve0_gpio_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/i2c.h b/drivers/gpu/drm/nouveau/core/include/subdev/i2c.h
index 7f50a858b16..825f7bb46b6 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/i2c.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/i2c.h
@@ -14,52 +14,41 @@
 #define NV_I2C_TYPE_EXTDDC(e) (0x0005 | (e) << 8)
 #define NV_I2C_TYPE_EXTAUX(e) (0x0006 | (e) << 8)
 
+enum nvkm_i2c_event {
+	NVKM_I2C_PLUG = 1,
+	NVKM_I2C_UNPLUG = 2,
+	NVKM_I2C_IRQ = 4,
+	NVKM_I2C_DONE = 8,
+	NVKM_I2C_ANY = (NVKM_I2C_PLUG |
+			NVKM_I2C_UNPLUG |
+			NVKM_I2C_IRQ |
+			NVKM_I2C_DONE),
+};
+
 struct nouveau_i2c_port {
 	struct nouveau_object base;
 	struct i2c_adapter adapter;
+	struct mutex mutex;
 
 	struct list_head head;
 	u8  index;
+	int aux;
 
 	const struct nouveau_i2c_func *func;
 };
 
 struct nouveau_i2c_func {
-	void (*acquire)(struct nouveau_i2c_port *);
-	void (*release)(struct nouveau_i2c_port *);
-
 	void (*drive_scl)(struct nouveau_i2c_port *, int);
 	void (*drive_sda)(struct nouveau_i2c_port *, int);
 	int  (*sense_scl)(struct nouveau_i2c_port *);
 	int  (*sense_sda)(struct nouveau_i2c_port *);
 
-	int  (*aux)(struct nouveau_i2c_port *, u8, u32, u8 *, u8);
+	int  (*aux)(struct nouveau_i2c_port *, bool, u8, u32, u8 *, u8);
 	int  (*pattern)(struct nouveau_i2c_port *, int pattern);
 	int  (*lnk_ctl)(struct nouveau_i2c_port *, int nr, int bw, bool enh);
 	int  (*drv_ctl)(struct nouveau_i2c_port *, int lane, int sw, int pe);
 };
 
-#define nouveau_i2c_port_create(p,e,o,i,a,f,d)                                 \
-	nouveau_i2c_port_create_((p), (e), (o), (i), (a), (f),                 \
-				 sizeof(**d), (void **)d)
-#define nouveau_i2c_port_destroy(p) ({                                         \
-	struct nouveau_i2c_port *port = (p);                                   \
-	_nouveau_i2c_port_dtor(nv_object(i2c));                                \
-})
-#define nouveau_i2c_port_init(p)                                               \
-	nouveau_object_init(&(p)->base)
-#define nouveau_i2c_port_fini(p,s)                                             \
-	nouveau_object_fini(&(p)->base, (s))
-
-int nouveau_i2c_port_create_(struct nouveau_object *, struct nouveau_object *,
-			     struct nouveau_oclass *, u8,
-			     const struct i2c_algorithm *,
-			     const struct nouveau_i2c_func *,
-			     int, void **);
-void _nouveau_i2c_port_dtor(struct nouveau_object *);
-#define _nouveau_i2c_port_init nouveau_object_init
-#define _nouveau_i2c_port_fini nouveau_object_fini
-
 struct nouveau_i2c_board_info {
 	struct i2c_board_info dev;
 	u8 udelay; /* set to 0 to use the standard delay */
@@ -67,13 +56,20 @@ struct nouveau_i2c_board_info {
 
 struct nouveau_i2c {
 	struct nouveau_subdev base;
+	struct nouveau_event *ntfy;
 
 	struct nouveau_i2c_port *(*find)(struct nouveau_i2c *, u8 index);
 	struct nouveau_i2c_port *(*find_type)(struct nouveau_i2c *, u16 type);
+	int  (*acquire_pad)(struct nouveau_i2c_port *, unsigned long timeout);
+	void (*release_pad)(struct nouveau_i2c_port *);
+	int  (*acquire)(struct nouveau_i2c_port *, unsigned long timeout);
+	void (*release)(struct nouveau_i2c_port *);
 	int (*identify)(struct nouveau_i2c *, int index,
 			const char *what, struct nouveau_i2c_board_info *,
 			bool (*match)(struct nouveau_i2c_port *,
 				      struct i2c_board_info *, void *), void *);
+
+	wait_queue_head_t wait;
 	struct list_head ports;
 };
 
@@ -83,37 +79,13 @@ nouveau_i2c(void *obj)
 	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_I2C];
 }
 
-#define nouveau_i2c_create(p,e,o,s,d)                                          \
-	nouveau_i2c_create_((p), (e), (o), (s), sizeof(**d), (void **)d)
-#define nouveau_i2c_destroy(p) ({                                              \
-	struct nouveau_i2c *i2c = (p);                                         \
-	_nouveau_i2c_dtor(nv_object(i2c));                                     \
-})
-#define nouveau_i2c_init(p) ({                                                 \
-	struct nouveau_i2c *i2c = (p);                                         \
-	_nouveau_i2c_init(nv_object(i2c));                                     \
-})
-#define nouveau_i2c_fini(p,s) ({                                               \
-	struct nouveau_i2c *i2c = (p);                                         \
-	_nouveau_i2c_fini(nv_object(i2c), (s));                                \
-})
-
-int nouveau_i2c_create_(struct nouveau_object *, struct nouveau_object *,
-			struct nouveau_oclass *, struct nouveau_oclass *,
-			int, void **);
-void _nouveau_i2c_dtor(struct nouveau_object *);
-int  _nouveau_i2c_init(struct nouveau_object *);
-int  _nouveau_i2c_fini(struct nouveau_object *, bool);
-
-extern struct nouveau_oclass nv04_i2c_oclass;
-extern struct nouveau_oclass nv4e_i2c_oclass;
-extern struct nouveau_oclass nv50_i2c_oclass;
-extern struct nouveau_oclass nv94_i2c_oclass;
-extern struct nouveau_oclass nvd0_i2c_oclass;
-extern struct nouveau_oclass nouveau_anx9805_sclass[];
-
-extern const struct i2c_algorithm nouveau_i2c_bit_algo;
-extern const struct i2c_algorithm nouveau_i2c_aux_algo;
+extern struct nouveau_oclass *nv04_i2c_oclass;
+extern struct nouveau_oclass *nv4e_i2c_oclass;
+extern struct nouveau_oclass *nv50_i2c_oclass;
+extern struct nouveau_oclass *nv94_i2c_oclass;
+extern struct nouveau_oclass *nvd0_i2c_oclass;
+extern struct nouveau_oclass *gf117_i2c_oclass;
+extern struct nouveau_oclass *nve0_i2c_oclass;
 
 static inline int
 nv_rdi2cr(struct nouveau_i2c_port *port, u8 addr, u8 reg)
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/ibus.h b/drivers/gpu/drm/nouveau/core/include/subdev/ibus.h
index 88814f159d8..31df634c0fd 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/ibus.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/ibus.h
@@ -30,5 +30,6 @@ nouveau_ibus(void *obj)
 
 extern struct nouveau_oclass nvc0_ibus_oclass;
 extern struct nouveau_oclass nve0_ibus_oclass;
+extern struct nouveau_oclass gk20a_ibus_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/instmem.h b/drivers/gpu/drm/nouveau/core/include/subdev/instmem.h
index 4aca33887aa..c1df26f3230 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/instmem.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/instmem.h
@@ -23,21 +23,6 @@ nv_memobj(void *obj)
 	return obj;
 }
 
-#define nouveau_instobj_create(p,e,o,d)                                        \
-	nouveau_instobj_create_((p), (e), (o), sizeof(**d), (void **)d)
-#define nouveau_instobj_init(p)                                                \
-	nouveau_object_init(&(p)->base)
-#define nouveau_instobj_fini(p,s)                                              \
-	nouveau_object_fini(&(p)->base, (s))
-
-int  nouveau_instobj_create_(struct nouveau_object *, struct nouveau_object *,
-			     struct nouveau_oclass *, int, void **);
-void nouveau_instobj_destroy(struct nouveau_instobj *);
-
-void _nouveau_instobj_dtor(struct nouveau_object *);
-#define _nouveau_instobj_init nouveau_object_init
-#define _nouveau_instobj_fini nouveau_object_fini
-
 struct nouveau_instmem {
 	struct nouveau_subdev base;
 	struct list_head list;
@@ -60,21 +45,8 @@ nouveau_instmem(void *obj)
 	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_INSTMEM];
 }
 
-#define nouveau_instmem_create(p,e,o,d)                                        \
-	nouveau_instmem_create_((p), (e), (o), sizeof(**d), (void **)d)
-#define nouveau_instmem_destroy(p)                                             \
-	nouveau_subdev_destroy(&(p)->base)
-int nouveau_instmem_create_(struct nouveau_object *, struct nouveau_object *,
-			    struct nouveau_oclass *, int, void **);
-int nouveau_instmem_init(struct nouveau_instmem *);
-int nouveau_instmem_fini(struct nouveau_instmem *, bool);
-
-#define _nouveau_instmem_dtor _nouveau_subdev_dtor
-int _nouveau_instmem_init(struct nouveau_object *);
-int _nouveau_instmem_fini(struct nouveau_object *, bool);
-
-extern struct nouveau_oclass nv04_instmem_oclass;
-extern struct nouveau_oclass nv40_instmem_oclass;
-extern struct nouveau_oclass nv50_instmem_oclass;
+extern struct nouveau_oclass *nv04_instmem_oclass;
+extern struct nouveau_oclass *nv40_instmem_oclass;
+extern struct nouveau_oclass *nv50_instmem_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/ltcg.h b/drivers/gpu/drm/nouveau/core/include/subdev/ltcg.h
index a1985ed3d58..c9c1950b774 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/ltcg.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/ltcg.h
@@ -35,6 +35,7 @@ nouveau_ltcg(void *obj)
 #define _nouveau_ltcg_init _nouveau_subdev_init
 #define _nouveau_ltcg_fini _nouveau_subdev_fini
 
-extern struct nouveau_oclass nvc0_ltcg_oclass;
+extern struct nouveau_oclass *gf100_ltcg_oclass;
+extern struct nouveau_oclass *gm107_ltcg_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/mc.h b/drivers/gpu/drm/nouveau/core/include/subdev/mc.h
index adc88b73d91..72b176831be 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/mc.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/mc.h
@@ -12,6 +12,7 @@ struct nouveau_mc_intr {
 struct nouveau_mc {
 	struct nouveau_subdev base;
 	bool use_msi;
+	unsigned int irq;
 };
 
 static inline struct nouveau_mc *
@@ -47,6 +48,7 @@ struct nouveau_mc_oclass {
 extern struct nouveau_oclass *nv04_mc_oclass;
 extern struct nouveau_oclass *nv40_mc_oclass;
 extern struct nouveau_oclass *nv44_mc_oclass;
+extern struct nouveau_oclass *nv4c_mc_oclass;
 extern struct nouveau_oclass *nv50_mc_oclass;
 extern struct nouveau_oclass *nv94_mc_oclass;
 extern struct nouveau_oclass *nv98_mc_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/therm.h b/drivers/gpu/drm/nouveau/core/include/subdev/therm.h
index 69891d4a3fe..d4a68179e58 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/therm.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/therm.h
@@ -31,7 +31,7 @@ struct nouveau_therm {
 	int (*pwm_ctrl)(struct nouveau_therm *, int line, bool);
 	int (*pwm_get)(struct nouveau_therm *, int line, u32 *, u32 *);
 	int (*pwm_set)(struct nouveau_therm *, int line, u32, u32);
-	int (*pwm_clock)(struct nouveau_therm *);
+	int (*pwm_clock)(struct nouveau_therm *, int line);
 
 	int (*fan_get)(struct nouveau_therm *);
 	int (*fan_set)(struct nouveau_therm *, int);
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/timer.h b/drivers/gpu/drm/nouveau/core/include/subdev/timer.h
index 9ab70dfe5b0..db9be803a87 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/timer.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/timer.h
@@ -59,5 +59,6 @@ int nouveau_timer_create_(struct nouveau_object *, struct nouveau_engine *,
 			  struct nouveau_oclass *, int size, void **);
 
 extern struct nouveau_oclass nv04_timer_oclass;
+extern struct nouveau_oclass gk20a_timer_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/vm.h b/drivers/gpu/drm/nouveau/core/include/subdev/vm.h
index fcf57fa309b..c9509039f94 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/vm.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/vm.h
@@ -131,9 +131,5 @@ void nouveau_vm_map(struct nouveau_vma *, struct nouveau_mem *);
 void nouveau_vm_map_at(struct nouveau_vma *, u64 offset, struct nouveau_mem *);
 void nouveau_vm_unmap(struct nouveau_vma *);
 void nouveau_vm_unmap_at(struct nouveau_vma *, u64 offset, u64 length);
-void nouveau_vm_map_sg(struct nouveau_vma *, u64 offset, u64 length,
-		       struct nouveau_mem *);
-void nouveau_vm_map_sg_table(struct nouveau_vma *vma, u64 delta, u64 length,
-		     struct nouveau_mem *mem);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/os.h b/drivers/gpu/drm/nouveau/core/os.h
index 191e739f30d..d0ced94ca54 100644
--- a/drivers/gpu/drm/nouveau/core/os.h
+++ b/drivers/gpu/drm/nouveau/core/os.h
@@ -5,6 +5,7 @@
 #include <linux/slab.h>
 #include <linux/mutex.h>
 #include <linux/pci.h>
+#include <linux/platform_device.h>
 #include <linux/printk.h>
 #include <linux/bitops.h>
 #include <linux/firmware.h>
@@ -23,17 +24,6 @@
 
 #include <asm/unaligned.h>
 
-static inline int
-ffsll(u64 mask)
-{
-	int i;
-	for (i = 0; i < 64; i++) {
-		if (mask & (1ULL << i))
-			return i + 1;
-	}
-	return 0;
-}
-
 #ifndef ioread32_native
 #ifdef __BIG_ENDIAN
 #define ioread16_native ioread16be
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/base.c b/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
index d70ba342aa2..73b1ed20c8d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
@@ -23,7 +23,11 @@
  */
 
 #include <core/object.h>
-#include <subdev/bar.h>
+
+#include <subdev/fb.h>
+#include <subdev/vm.h>
+
+#include "priv.h"
 
 struct nouveau_barobj {
 	struct nouveau_object base;
@@ -114,8 +118,10 @@ nouveau_bar_create_(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	bar->iomem = ioremap(pci_resource_start(device->pdev, 3),
-			     pci_resource_len(device->pdev, 3));
+	if (nv_device_resource_len(device, 3) != 0)
+		bar->iomem = ioremap(nv_device_resource_start(device, 3),
+				     nv_device_resource_len(device, 3));
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/bar/nv50.c
index 160d27f3c7b..f748ba49dfc 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bar/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/nv50.c
@@ -25,10 +25,11 @@
 #include <core/gpuobj.h>
 
 #include <subdev/timer.h>
-#include <subdev/bar.h>
 #include <subdev/fb.h>
 #include <subdev/vm.h>
 
+#include "priv.h"
+
 struct nv50_bar_priv {
 	struct nouveau_bar base;
 	spinlock_t lock;
@@ -138,7 +139,7 @@ nv50_bar_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 	/* BAR3 */
 	start = 0x0100000000ULL;
-	limit = start + pci_resource_len(device->pdev, 3);
+	limit = start + nv_device_resource_len(device, 3);
 
 	ret = nouveau_vm_new(device, start, limit, start, &vm);
 	if (ret)
@@ -172,7 +173,7 @@ nv50_bar_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 	/* BAR1 */
 	start = 0x0000000000ULL;
-	limit = start + pci_resource_len(device->pdev, 1);
+	limit = start + nv_device_resource_len(device, 1);
 
 	ret = nouveau_vm_new(device, start, limit--, start, &vm);
 	if (ret)
@@ -230,7 +231,7 @@ static int
 nv50_bar_init(struct nouveau_object *object)
 {
 	struct nv50_bar_priv *priv = (void *)object;
-	int ret;
+	int ret, i;
 
 	ret = nouveau_bar_init(&priv->base);
 	if (ret)
@@ -248,6 +249,8 @@ nv50_bar_init(struct nouveau_object *object)
 	nv_wr32(priv, 0x001704, 0x40000000 | priv->mem->addr >> 12);
 	nv_wr32(priv, 0x001708, 0x80000000 | priv->bar1->node->offset >> 4);
 	nv_wr32(priv, 0x00170c, 0x80000000 | priv->bar3->node->offset >> 4);
+	for (i = 0; i < 8; i++)
+		nv_wr32(priv, 0x001900 + (i * 4), 0x00000000);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/bar/nvc0.c
index b2ec7411eb2..ca8139b9ab2 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bar/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/nvc0.c
@@ -25,18 +25,21 @@
 #include <core/gpuobj.h>
 
 #include <subdev/timer.h>
-#include <subdev/bar.h>
 #include <subdev/fb.h>
 #include <subdev/vm.h>
 
+#include "priv.h"
+
+struct nvc0_bar_priv_vm {
+	struct nouveau_gpuobj *mem;
+	struct nouveau_gpuobj *pgd;
+	struct nouveau_vm *vm;
+};
+
 struct nvc0_bar_priv {
 	struct nouveau_bar base;
 	spinlock_t lock;
-	struct {
-		struct nouveau_gpuobj *mem;
-		struct nouveau_gpuobj *pgd;
-		struct nouveau_vm *vm;
-	} bar[2];
+	struct nvc0_bar_priv_vm bar[2];
 };
 
 static int
@@ -78,88 +81,87 @@ nvc0_bar_unmap(struct nouveau_bar *bar, struct nouveau_vma *vma)
 }
 
 static int
-nvc0_bar_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	      struct nouveau_oclass *oclass, void *data, u32 size,
-	      struct nouveau_object **pobject)
+nvc0_bar_init_vm(struct nvc0_bar_priv *priv, struct nvc0_bar_priv_vm *bar_vm,
+		 int bar_nr)
 {
-	struct nouveau_device *device = nv_device(parent);
-	struct pci_dev *pdev = device->pdev;
-	struct nvc0_bar_priv *priv;
-	struct nouveau_gpuobj *mem;
+	struct nouveau_device *device = nv_device(&priv->base);
 	struct nouveau_vm *vm;
+	resource_size_t bar_len;
 	int ret;
 
-	ret = nouveau_bar_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	/* BAR3 */
 	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x1000, 0, 0,
-				&priv->bar[0].mem);
-	mem = priv->bar[0].mem;
+				&bar_vm->mem);
 	if (ret)
 		return ret;
 
 	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0, 0,
-				&priv->bar[0].pgd);
+				&bar_vm->pgd);
 	if (ret)
 		return ret;
 
-	ret = nouveau_vm_new(device, 0, pci_resource_len(pdev, 3), 0, &vm);
+	bar_len = nv_device_resource_len(device, bar_nr);
+
+	ret = nouveau_vm_new(device, 0, bar_len, 0, &vm);
 	if (ret)
 		return ret;
 
 	atomic_inc(&vm->engref[NVDEV_SUBDEV_BAR]);
 
-	ret = nouveau_gpuobj_new(nv_object(priv), NULL,
-				 (pci_resource_len(pdev, 3) >> 12) * 8,
-				 0x1000, NVOBJ_FLAG_ZERO_ALLOC,
-				 &vm->pgt[0].obj[0]);
-	vm->pgt[0].refcount[0] = 1;
-	if (ret)
-		return ret;
+	/*
+	 * Bootstrap page table lookup.
+	 */
+	if (bar_nr == 3) {
+		ret = nouveau_gpuobj_new(nv_object(priv), NULL,
+					 (bar_len >> 12) * 8, 0x1000,
+					 NVOBJ_FLAG_ZERO_ALLOC,
+					&vm->pgt[0].obj[0]);
+		vm->pgt[0].refcount[0] = 1;
+		if (ret)
+			return ret;
+	}
 
-	ret = nouveau_vm_ref(vm, &priv->bar[0].vm, priv->bar[0].pgd);
+	ret = nouveau_vm_ref(vm, &bar_vm->vm, bar_vm->pgd);
 	nouveau_vm_ref(NULL, &vm, NULL);
 	if (ret)
 		return ret;
 
-	nv_wo32(mem, 0x0200, lower_32_bits(priv->bar[0].pgd->addr));
-	nv_wo32(mem, 0x0204, upper_32_bits(priv->bar[0].pgd->addr));
-	nv_wo32(mem, 0x0208, lower_32_bits(pci_resource_len(pdev, 3) - 1));
-	nv_wo32(mem, 0x020c, upper_32_bits(pci_resource_len(pdev, 3) - 1));
+	nv_wo32(bar_vm->mem, 0x0200, lower_32_bits(bar_vm->pgd->addr));
+	nv_wo32(bar_vm->mem, 0x0204, upper_32_bits(bar_vm->pgd->addr));
+	nv_wo32(bar_vm->mem, 0x0208, lower_32_bits(bar_len - 1));
+	nv_wo32(bar_vm->mem, 0x020c, upper_32_bits(bar_len - 1));
 
-	/* BAR1 */
-	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x1000, 0, 0,
-				&priv->bar[1].mem);
-	mem = priv->bar[1].mem;
-	if (ret)
-		return ret;
+	return 0;
+}
 
-	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0, 0,
-				&priv->bar[1].pgd);
-	if (ret)
-		return ret;
+static int
+nvc0_bar_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	      struct nouveau_oclass *oclass, void *data, u32 size,
+	      struct nouveau_object **pobject)
+{
+	struct nouveau_device *device = nv_device(parent);
+	struct nvc0_bar_priv *priv;
+	bool has_bar3 = nv_device_resource_len(device, 3) != 0;
+	int ret;
 
-	ret = nouveau_vm_new(device, 0, pci_resource_len(pdev, 1), 0, &vm);
+	ret = nouveau_bar_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
 
-	atomic_inc(&vm->engref[NVDEV_SUBDEV_BAR]);
+	/* BAR3 */
+	if (has_bar3) {
+		ret = nvc0_bar_init_vm(priv, &priv->bar[0], 3);
+		if (ret)
+			return ret;
+		priv->base.alloc = nouveau_bar_alloc;
+		priv->base.kmap = nvc0_bar_kmap;
+	}
 
-	ret = nouveau_vm_ref(vm, &priv->bar[1].vm, priv->bar[1].pgd);
-	nouveau_vm_ref(NULL, &vm, NULL);
+	/* BAR1 */
+	ret = nvc0_bar_init_vm(priv, &priv->bar[1], 1);
 	if (ret)
 		return ret;
 
-	nv_wo32(mem, 0x0200, lower_32_bits(priv->bar[1].pgd->addr));
-	nv_wo32(mem, 0x0204, upper_32_bits(priv->bar[1].pgd->addr));
-	nv_wo32(mem, 0x0208, lower_32_bits(pci_resource_len(pdev, 1) - 1));
-	nv_wo32(mem, 0x020c, upper_32_bits(pci_resource_len(pdev, 1) - 1));
-
-	priv->base.alloc = nouveau_bar_alloc;
-	priv->base.kmap = nvc0_bar_kmap;
 	priv->base.umap = nvc0_bar_umap;
 	priv->base.unmap = nvc0_bar_unmap;
 	priv->base.flush = nv84_bar_flush;
@@ -201,7 +203,9 @@ nvc0_bar_init(struct nouveau_object *object)
 	nv_mask(priv, 0x100c80, 0x00000001, 0x00000000);
 
 	nv_wr32(priv, 0x001704, 0x80000000 | priv->bar[1].mem->addr >> 12);
-	nv_wr32(priv, 0x001714, 0xc0000000 | priv->bar[0].mem->addr >> 12);
+	if (priv->bar[0].mem)
+		nv_wr32(priv, 0x001714,
+			0xc0000000 | priv->bar[0].mem->addr >> 12);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/priv.h b/drivers/gpu/drm/nouveau/core/subdev/bar/priv.h
new file mode 100644
index 00000000000..ffad8f337ea
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/priv.h
@@ -0,0 +1,26 @@
+#ifndef __NVKM_BAR_PRIV_H__
+#define __NVKM_BAR_PRIV_H__
+
+#include <subdev/bar.h>
+
+#define nouveau_bar_create(p,e,o,d)                                            \
+	nouveau_bar_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_bar_init(p)                                                    \
+	nouveau_subdev_init(&(p)->base)
+#define nouveau_bar_fini(p,s)                                                  \
+	nouveau_subdev_fini(&(p)->base, (s))
+
+int nouveau_bar_create_(struct nouveau_object *, struct nouveau_object *,
+			struct nouveau_oclass *, int, void **);
+void nouveau_bar_destroy(struct nouveau_bar *);
+
+void _nouveau_bar_dtor(struct nouveau_object *);
+#define _nouveau_bar_init _nouveau_subdev_init
+#define _nouveau_bar_fini _nouveau_subdev_fini
+
+int  nouveau_bar_alloc(struct nouveau_bar *, struct nouveau_object *,
+		       struct nouveau_mem *, struct nouveau_object **);
+
+void nv84_bar_flush(struct nouveau_bar *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/P0260.c b/drivers/gpu/drm/nouveau/core/subdev/bios/P0260.c
new file mode 100644
index 00000000000..199f4e5f748
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/P0260.c
@@ -0,0 +1,109 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/ramcfg.h>
+#include <subdev/bios/P0260.h>
+
+u32
+nvbios_P0260Te(struct nouveau_bios *bios,
+	       u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *xnr, u8 *xsz)
+{
+	struct bit_entry bit_P;
+	u32 data = 0x00000000;
+
+	if (!bit_entry(bios, 'P', &bit_P)) {
+		if (bit_P.version == 2 && bit_P.length > 0x63)
+			data = nv_ro32(bios, bit_P.offset + 0x60);
+		if (data) {
+			*ver = nv_ro08(bios, data + 0);
+			switch (*ver) {
+			case 0x10:
+				*hdr = nv_ro08(bios, data + 1);
+				*cnt = nv_ro08(bios, data + 2);
+				*len = 4;
+				*xnr = nv_ro08(bios, data + 3);
+				*xsz = 4;
+				return data;
+			default:
+				break;
+			}
+		}
+	}
+
+	return 0x00000000;
+}
+
+u32
+nvbios_P0260Ee(struct nouveau_bios *bios, int idx, u8 *ver, u8 *len)
+{
+	u8  hdr, cnt, xnr, xsz;
+	u32 data = nvbios_P0260Te(bios, ver, &hdr, &cnt, len, &xnr, &xsz);
+	if (data && idx < cnt)
+		return data + hdr + (idx * *len);
+	return 0x00000000;
+}
+
+u32
+nvbios_P0260Ep(struct nouveau_bios *bios, int idx, u8 *ver, u8 *len,
+	       struct nvbios_P0260E *info)
+{
+	u32 data = nvbios_P0260Ee(bios, idx, ver, len);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->data = nv_ro32(bios, data);
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_P0260Xe(struct nouveau_bios *bios, int idx, u8 *ver, u8 *xsz)
+{
+	u8  hdr, cnt, len, xnr;
+	u32 data = nvbios_P0260Te(bios, ver, &hdr, &cnt, &len, &xnr, xsz);
+	if (data && idx < xnr)
+		return data + hdr + (cnt * len) + (idx * *xsz);
+	return 0x00000000;
+}
+
+u32
+nvbios_P0260Xp(struct nouveau_bios *bios, int idx, u8 *ver, u8 *hdr,
+	       struct nvbios_P0260X *info)
+{
+	u32 data = nvbios_P0260Xe(bios, idx, ver, hdr);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x10:
+		info->data = nv_ro32(bios, data);
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
index aa0fbbec7f0..d45704a2c2d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
@@ -90,10 +90,26 @@ nouveau_bios_shadow_pramin(struct nouveau_bios *bios)
 	int i;
 
 	if (device->card_type >= NV_50) {
-		if (  device->card_type < NV_C0 ||
-		    !(nv_rd32(bios, 0x022500) & 0x00000001))
-			addr = (u64)(nv_rd32(bios, 0x619f04) & 0xffffff00) << 8;
+		if (device->card_type >= NV_C0 && device->card_type < GM100) {
+			if (nv_rd32(bios, 0x022500) & 0x00000001)
+				return;
+		} else
+		if (device->card_type >= GM100) {
+			if (nv_rd32(bios, 0x021c04) & 0x00000001)
+				return;
+		}
+
+		addr = nv_rd32(bios, 0x619f04);
+		if (!(addr & 0x00000008)) {
+			nv_debug(bios, "... not enabled\n");
+			return;
+		}
+		if ( (addr & 0x00000003) != 1) {
+			nv_debug(bios, "... not in vram\n");
+			return;
+		}
 
+		addr = (addr & 0xffffff00) << 8;
 		if (!addr) {
 			addr  = (u64)nv_rd32(bios, 0x001700) << 16;
 			addr += 0xf0000;
@@ -130,6 +146,10 @@ nouveau_bios_shadow_prom(struct nouveau_bios *bios)
 	u16 pcir;
 	int i;
 
+	/* there is no prom on nv4x IGP's */
+	if (device->card_type == NV_40 && device->chipset >= 0x4c)
+		return;
+
 	/* enable access to rom */
 	if (device->card_type >= NV_50)
 		pcireg = 0x088050;
@@ -137,6 +157,10 @@ nouveau_bios_shadow_prom(struct nouveau_bios *bios)
 		pcireg = 0x001850;
 	access = nv_mask(bios, pcireg, 0x00000001, 0x00000000);
 
+	/* WARNING: PROM accesses should always be 32-bits aligned. Other
+	 * accesses work on most chipset but do not on Kepler chipsets
+	 */
+
 	/* bail if no rom signature, with a workaround for a PROM reading
 	 * issue on some chipsets.  the first read after a period of
 	 * inactivity returns the wrong result, so retry the first header
@@ -144,31 +168,35 @@ nouveau_bios_shadow_prom(struct nouveau_bios *bios)
 	 */
 	i = 16;
 	do {
-		if (nv_rd08(bios, 0x300000) == 0x55)
+		u32 data = le32_to_cpu(nv_rd32(bios, 0x300000)) & 0xffff;
+		if (data == 0xaa55)
 			break;
 	} while (i--);
 
-	if (!i || nv_rd08(bios, 0x300001) != 0xaa)
-		goto out;
-
-	/* additional check (see note below) - read PCI record header */
-	pcir = nv_rd08(bios, 0x300018) |
-	       nv_rd08(bios, 0x300019) << 8;
-	if (nv_rd08(bios, 0x300000 + pcir) != 'P' ||
-	    nv_rd08(bios, 0x300001 + pcir) != 'C' ||
-	    nv_rd08(bios, 0x300002 + pcir) != 'I' ||
-	    nv_rd08(bios, 0x300003 + pcir) != 'R')
+	if (!i)
 		goto out;
 
 	/* read entire bios image to system memory */
-	bios->size = nv_rd08(bios, 0x300002) * 512;
+	bios->size = (le32_to_cpu(nv_rd32(bios, 0x300000)) >> 16) & 0xff;
+	bios->size = bios->size * 512;
 	if (!bios->size)
 		goto out;
 
 	bios->data = kmalloc(bios->size, GFP_KERNEL);
-	if (bios->data) {
-		for (i = 0; i < bios->size; i++)
-			nv_wo08(bios, i, nv_rd08(bios, 0x300000 + i));
+	if (!bios->data)
+		goto out;
+
+	for (i = 0; i < bios->size; i += 4)
+		((u32 *)bios->data)[i/4] = nv_rd32(bios, 0x300000 + i);
+
+	/* check the PCI record header */
+	pcir = nv_ro16(bios, 0x0018);
+	if (bios->data[pcir + 0] != 'P' ||
+	    bios->data[pcir + 1] != 'C' ||
+	    bios->data[pcir + 2] != 'I' ||
+	    bios->data[pcir + 3] != 'R') {
+		bios->size = 0;
+		kfree(bios->data);
 	}
 
 out:
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/conn.c b/drivers/gpu/drm/nouveau/core/subdev/bios/conn.c
index 5ac010efd95..2ede3bcd96a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/conn.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/conn.c
@@ -28,12 +28,12 @@
 #include <subdev/bios/dcb.h>
 #include <subdev/bios/conn.h>
 
-u16
-dcb_conntab(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+u32
+nvbios_connTe(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 {
-	u16 dcb = dcb_table(bios, ver, hdr, cnt, len);
+	u32 dcb = dcb_table(bios, ver, hdr, cnt, len);
 	if (dcb && *ver >= 0x30 && *hdr >= 0x16) {
-		u16 data = nv_ro16(bios, dcb + 0x14);
+		u32 data = nv_ro16(bios, dcb + 0x14);
 		if (data) {
 			*ver = nv_ro08(bios, data + 0);
 			*hdr = nv_ro08(bios, data + 1);
@@ -42,15 +42,59 @@ dcb_conntab(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 			return data;
 		}
 	}
-	return 0x0000;
+	return 0x00000000;
 }
 
-u16
-dcb_conn(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *len)
+u32
+nvbios_connTp(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+	      struct nvbios_connT *info)
+{
+	u32 data = nvbios_connTe(bios, ver, hdr, cnt, len);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x30:
+	case 0x40:
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
+}
+
+u32
+nvbios_connEe(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *len)
 {
 	u8  hdr, cnt;
-	u16 data = dcb_conntab(bios, ver, &hdr, &cnt, len);
+	u32 data = nvbios_connTe(bios, ver, &hdr, &cnt, len);
 	if (data && idx < cnt)
 		return data + hdr + (idx * *len);
-	return 0x0000;
+	return 0x00000000;
+}
+
+u32
+nvbios_connEp(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *len,
+	      struct nvbios_connE *info)
+{
+	u32 data = nvbios_connEe(bios, idx, ver, len);
+	memset(info, 0x00, sizeof(*info));
+	switch (!!data * *ver) {
+	case 0x30:
+	case 0x40:
+		info->type     =  nv_ro08(bios, data + 0x00);
+		info->location =  nv_ro08(bios, data + 0x01) & 0x0f;
+		info->hpd      = (nv_ro08(bios, data + 0x01) & 0x30) >> 4;
+		info->dp       = (nv_ro08(bios, data + 0x01) & 0xc0) >> 6;
+		if (*len < 4)
+			return data;
+		info->hpd     |= (nv_ro08(bios, data + 0x02) & 0x03) << 2;
+		info->dp      |=  nv_ro08(bios, data + 0x02) & 0x0c;
+		info->di       = (nv_ro08(bios, data + 0x02) & 0xf0) >> 4;
+		info->hpd     |= (nv_ro08(bios, data + 0x03) & 0x07) << 4;
+		info->sr       = (nv_ro08(bios, data + 0x03) & 0x08) >> 3;
+		info->lcdid    = (nv_ro08(bios, data + 0x03) & 0x70) >> 4;
+		return data;
+	default:
+		break;
+	}
+	return 0x00000000;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c b/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
index 2d9b9d7a799..88606bfaf84 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/dcb.c
@@ -142,9 +142,36 @@ dcb_outp_parse(struct nouveau_bios *bios, u8 idx, u8 *ver, u8 *len,
 		if (*ver >= 0x40) {
 			u32 conf = nv_ro32(bios, dcb + 0x04);
 			switch (outp->type) {
+			case DCB_OUTPUT_DP:
+				switch (conf & 0x00e00000) {
+				case 0x00000000:
+					outp->dpconf.link_bw = 0x06;
+					break;
+				case 0x00200000:
+					outp->dpconf.link_bw = 0x0a;
+					break;
+				case 0x00400000:
+				default:
+					outp->dpconf.link_bw = 0x14;
+					break;
+				}
+
+				switch (conf & 0x0f000000) {
+				case 0x0f000000:
+					outp->dpconf.link_nr = 4;
+					break;
+				case 0x03000000:
+					outp->dpconf.link_nr = 2;
+					break;
+				case 0x01000000:
+				default:
+					outp->dpconf.link_nr = 1;
+					break;
+				}
+
+				/* fall-through... */
 			case DCB_OUTPUT_TMDS:
 			case DCB_OUTPUT_LVDS:
-			case DCB_OUTPUT_DP:
 				outp->link = (conf & 0x00000030) >> 4;
 				outp->sorconf.link = outp->link; /*XXX*/
 				outp->extdev = 0x00;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/dp.c b/drivers/gpu/drm/nouveau/core/subdev/bios/dp.c
index 7628fe75922..f309dd65725 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/dp.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/dp.c
@@ -162,18 +162,20 @@ nvbios_dpcfg_parse(struct nouveau_bios *bios, u16 outp, u8 idx,
 		   struct nvbios_dpcfg *info)
 {
 	u16 data = nvbios_dpcfg_entry(bios, outp, idx, ver, hdr, cnt, len);
+	memset(info, 0x00, sizeof(*info));
 	if (data) {
 		switch (*ver) {
 		case 0x21:
-			info->drv = nv_ro08(bios, data + 0x02);
-			info->pre = nv_ro08(bios, data + 0x03);
-			info->unk = nv_ro08(bios, data + 0x04);
+			info->dc    = nv_ro08(bios, data + 0x02);
+			info->pe    = nv_ro08(bios, data + 0x03);
+			info->tx_pu = nv_ro08(bios, data + 0x04);
 			break;
 		case 0x30:
 		case 0x40:
-			info->drv = nv_ro08(bios, data + 0x01);
-			info->pre = nv_ro08(bios, data + 0x02);
-			info->unk = nv_ro08(bios, data + 0x03);
+			info->pc    = nv_ro08(bios, data + 0x00);
+			info->dc    = nv_ro08(bios, data + 0x01);
+			info->pe    = nv_ro08(bios, data + 0x02);
+			info->tx_pu = nv_ro08(bios, data + 0x03);
 			break;
 		default:
 			data = 0x0000;
@@ -184,7 +186,7 @@ nvbios_dpcfg_parse(struct nouveau_bios *bios, u16 outp, u8 idx,
 }
 
 u16
-nvbios_dpcfg_match(struct nouveau_bios *bios, u16 outp, u8 un, u8 vs, u8 pe,
+nvbios_dpcfg_match(struct nouveau_bios *bios, u16 outp, u8 pc, u8 vs, u8 pe,
 		   u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
 		   struct nvbios_dpcfg *info)
 {
@@ -193,16 +195,15 @@ nvbios_dpcfg_match(struct nouveau_bios *bios, u16 outp, u8 un, u8 vs, u8 pe,
 
 	if (*ver >= 0x30) {
 		const u8 vsoff[] = { 0, 4, 7, 9 };
-		idx = (un * 10) + vsoff[vs] + pe;
+		idx = (pc * 10) + vsoff[vs] + pe;
 	} else {
-		while ((data = nvbios_dpcfg_entry(bios, outp, idx,
+		while ((data = nvbios_dpcfg_entry(bios, outp, ++idx,
 						  ver, hdr, cnt, len))) {
 			if (nv_ro08(bios, data + 0x00) == vs &&
 			    nv_ro08(bios, data + 0x01) == pe)
 				break;
-			idx++;
 		}
 	}
 
-	return nvbios_dpcfg_parse(bios, outp, pe, ver, hdr, cnt, len, info);
+	return nvbios_dpcfg_parse(bios, outp, idx, ver, hdr, cnt, len, info);
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/init.c b/drivers/gpu/drm/nouveau/core/subdev/bios/init.c
index df1b1b42309..626380f9e4c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/init.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/init.c
@@ -9,6 +9,7 @@
 #include <subdev/bios/dp.h>
 #include <subdev/bios/gpio.h>
 #include <subdev/bios/init.h>
+#include <subdev/bios/ramcfg.h>
 #include <subdev/devinit.h>
 #include <subdev/i2c.h>
 #include <subdev/vga.h>
@@ -97,15 +98,16 @@ static u8
 init_conn(struct nvbios_init *init)
 {
 	struct nouveau_bios *bios = init->bios;
-	u8  ver, len;
-	u16 conn;
+	struct nvbios_connE connE;
+	u8  ver, hdr;
+	u32 conn;
 
 	if (init_exec(init)) {
 		if (init->outp) {
 			conn = init->outp->connector;
-			conn = dcb_conn(bios, conn, &ver, &len);
+			conn = nvbios_connEp(bios, conn, &ver, &hdr, &connE);
 			if (conn)
-				return nv_ro08(bios, conn);
+				return connE.type;
 		}
 
 		error("script needs connector type\n");
@@ -117,6 +119,8 @@ init_conn(struct nvbios_init *init)
 static inline u32
 init_nvreg(struct nvbios_init *init, u32 reg)
 {
+	struct nouveau_devinit *devinit = nouveau_devinit(init->bios);
+
 	/* C51 (at least) sometimes has the lower bits set which the VBIOS
 	 * interprets to mean that access needs to go through certain IO
 	 * ports instead.  The NVIDIA binary driver has been seen to access
@@ -146,6 +150,9 @@ init_nvreg(struct nvbios_init *init, u32 reg)
 
 	if (reg & ~0x00fffffc)
 		warn("unknown bits in register 0x%08x\n", reg);
+
+	if (devinit->mmio)
+		reg = devinit->mmio(devinit, reg);
 	return reg;
 }
 
@@ -153,7 +160,7 @@ static u32
 init_rd32(struct nvbios_init *init, u32 reg)
 {
 	reg = init_nvreg(init, reg);
-	if (init_exec(init))
+	if (reg != ~0 && init_exec(init))
 		return nv_rd32(init->subdev, reg);
 	return 0x00000000;
 }
@@ -162,7 +169,7 @@ static void
 init_wr32(struct nvbios_init *init, u32 reg, u32 val)
 {
 	reg = init_nvreg(init, reg);
-	if (init_exec(init))
+	if (reg != ~0 && init_exec(init))
 		nv_wr32(init->subdev, reg, val);
 }
 
@@ -170,7 +177,7 @@ static u32
 init_mask(struct nvbios_init *init, u32 reg, u32 mask, u32 val)
 {
 	reg = init_nvreg(init, reg);
-	if (init_exec(init)) {
+	if (reg != ~0 && init_exec(init)) {
 		u32 tmp = nv_rd32(init->subdev, reg);
 		nv_wr32(init->subdev, reg, (tmp & ~mask) | val);
 		return tmp;
@@ -391,43 +398,14 @@ init_unknown_script(struct nouveau_bios *bios)
 	return 0x0000;
 }
 
-static u16
-init_ram_restrict_table(struct nvbios_init *init)
-{
-	struct nouveau_bios *bios = init->bios;
-	struct bit_entry bit_M;
-	u16 data = 0x0000;
-
-	if (!bit_entry(bios, 'M', &bit_M)) {
-		if (bit_M.version == 1 && bit_M.length >= 5)
-			data = nv_ro16(bios, bit_M.offset + 3);
-		if (bit_M.version == 2 && bit_M.length >= 3)
-			data = nv_ro16(bios, bit_M.offset + 1);
-	}
-
-	if (data == 0x0000)
-		warn("ram restrict table not found\n");
-	return data;
-}
-
 static u8
 init_ram_restrict_group_count(struct nvbios_init *init)
 {
-	struct nouveau_bios *bios = init->bios;
-	struct bit_entry bit_M;
-
-	if (!bit_entry(bios, 'M', &bit_M)) {
-		if (bit_M.version == 1 && bit_M.length >= 5)
-			return nv_ro08(bios, bit_M.offset + 2);
-		if (bit_M.version == 2 && bit_M.length >= 3)
-			return nv_ro08(bios, bit_M.offset + 0);
-	}
-
-	return 0x00;
+	return nvbios_ramcfg_count(init->bios);
 }
 
 static u8
-init_ram_restrict_strap(struct nvbios_init *init)
+init_ram_restrict(struct nvbios_init *init)
 {
 	/* This appears to be the behaviour of the VBIOS parser, and *is*
 	 * important to cache the NV_PEXTDEV_BOOT0 on later chipsets to
@@ -438,18 +416,8 @@ init_ram_restrict_strap(struct nvbios_init *init)
 	 * in case *not* re-reading the strap causes similar breakage.
 	 */
 	if (!init->ramcfg || init->bios->version.major < 0x70)
-		init->ramcfg = init_rd32(init, 0x101000);
-	return (init->ramcfg & 0x00000003c) >> 2;
-}
-
-static u8
-init_ram_restrict(struct nvbios_init *init)
-{
-	u8  strap = init_ram_restrict_strap(init);
-	u16 table = init_ram_restrict_table(init);
-	if (table)
-		return nv_ro08(init->bios, table + strap);
-	return 0x00;
+		init->ramcfg = 0x80000000 | nvbios_ramcfg_index(init->subdev);
+	return (init->ramcfg & 0x7fffffff);
 }
 
 static u8
@@ -883,9 +851,8 @@ init_idx_addr_latched(struct nvbios_init *init)
 	u32 data = nv_ro32(bios, init->offset + 13);
 	u8 count = nv_ro08(bios, init->offset + 17);
 
-	trace("INDEX_ADDRESS_LATCHED\t"
-	      "R[0x%06x] : R[0x%06x]\n\tCTRL &= 0x%08x |= 0x%08x\n",
-	      creg, dreg, mask, data);
+	trace("INDEX_ADDRESS_LATCHED\tR[0x%06x] : R[0x%06x]\n", creg, dreg);
+	trace("\tCTRL &= 0x%08x |= 0x%08x\n", mask, data);
 	init->offset += 18;
 
 	while (count--) {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c b/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c
new file mode 100644
index 00000000000..6c401f70ab9
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c
@@ -0,0 +1,68 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs <bskeggs@redhat.com>
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/ramcfg.h>
+
+static u8
+nvbios_ramcfg_strap(struct nouveau_subdev *subdev)
+{
+	return (nv_rd32(subdev, 0x101000) & 0x0000003c) >> 2;
+}
+
+u8
+nvbios_ramcfg_count(struct nouveau_bios *bios)
+{
+	struct bit_entry bit_M;
+
+	if (!bit_entry(bios, 'M', &bit_M)) {
+		if (bit_M.version == 1 && bit_M.length >= 5)
+			return nv_ro08(bios, bit_M.offset + 2);
+		if (bit_M.version == 2 && bit_M.length >= 3)
+			return nv_ro08(bios, bit_M.offset + 0);
+	}
+
+	return 0x00;
+}
+
+u8
+nvbios_ramcfg_index(struct nouveau_subdev *subdev)
+{
+	struct nouveau_bios *bios = nouveau_bios(subdev);
+	u8 strap = nvbios_ramcfg_strap(subdev);
+	u32 xlat = 0x00000000;
+	struct bit_entry bit_M;
+
+	if (!bit_entry(bios, 'M', &bit_M)) {
+		if (bit_M.version == 1 && bit_M.length >= 5)
+			xlat = nv_ro16(bios, bit_M.offset + 3);
+		if (bit_M.version == 2 && bit_M.length >= 3)
+			xlat = nv_ro16(bios, bit_M.offset + 1);
+	}
+
+	if (xlat)
+		strap = nv_ro08(bios, xlat + strap);
+	return strap;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c b/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
index 916fa9d302b..1811b2cb047 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
@@ -24,11 +24,12 @@
 
 #include <subdev/bios.h>
 #include <subdev/bios/bit.h>
+#include <subdev/bios/ramcfg.h>
 #include <subdev/bios/rammap.h>
 
-u16
-nvbios_rammap_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr,
-		    u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
+u32
+nvbios_rammapTe(struct nouveau_bios *bios, u8 *ver, u8 *hdr,
+		u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
 {
 	struct bit_entry bit_P;
 	u16 rammap = 0x0000;
@@ -57,12 +58,12 @@ nvbios_rammap_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr,
 	return 0x0000;
 }
 
-u16
-nvbios_rammap_entry(struct nouveau_bios *bios, int idx,
-		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+u32
+nvbios_rammapEe(struct nouveau_bios *bios, int idx,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 {
 	u8  snr, ssz;
-	u16 rammap = nvbios_rammap_table(bios, ver, hdr, cnt, len, &snr, &ssz);
+	u16 rammap = nvbios_rammapTe(bios, ver, hdr, cnt, len, &snr, &ssz);
 	if (rammap && idx < *cnt) {
 		rammap = rammap + *hdr + (idx * (*len + (snr * ssz)));
 		*hdr = *len;
@@ -73,16 +74,100 @@ nvbios_rammap_entry(struct nouveau_bios *bios, int idx,
 	return 0x0000;
 }
 
-u16
-nvbios_rammap_match(struct nouveau_bios *bios, u16 khz,
-		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+u32
+nvbios_rammapEm(struct nouveau_bios *bios, u16 khz,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 {
 	int idx = 0;
 	u32 data;
-	while ((data = nvbios_rammap_entry(bios, idx++, ver, hdr, cnt, len))) {
+	while ((data = nvbios_rammapEe(bios, idx++, ver, hdr, cnt, len))) {
 		if (khz >= nv_ro16(bios, data + 0x00) &&
 		    khz <= nv_ro16(bios, data + 0x02))
 			break;
 	}
 	return data;
 }
+
+u32
+nvbios_rammapEp(struct nouveau_bios *bios, u16 khz,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		struct nvbios_ramcfg *p)
+{
+	u32 data = nvbios_rammapEm(bios, khz, ver, hdr, cnt, len);
+	memset(p, 0x00, sizeof(*p));
+	switch (!!data * *ver) {
+	case 0x11:
+		p->rammap_11_08_01 = (nv_ro08(bios, data + 0x08) & 0x01) >> 0;
+		p->rammap_11_08_0c = (nv_ro08(bios, data + 0x08) & 0x0c) >> 2;
+		p->rammap_11_08_10 = (nv_ro08(bios, data + 0x08) & 0x10) >> 4;
+		p->rammap_11_11_0c = (nv_ro08(bios, data + 0x11) & 0x0c) >> 2;
+		break;
+	default:
+		data = 0;
+		break;
+	}
+	return data;
+}
+
+u32
+nvbios_rammapSe(struct nouveau_bios *bios, u32 data,
+		u8 ever, u8 ehdr, u8 ecnt, u8 elen, int idx,
+		u8 *ver, u8 *hdr)
+{
+	if (idx < ecnt) {
+		data = data + ehdr + (idx * elen);
+		*ver = ever;
+		*hdr = elen;
+		return data;
+	}
+	return 0;
+}
+
+u32
+nvbios_rammapSp(struct nouveau_bios *bios, u32 data,
+		u8 ever, u8 ehdr, u8 ecnt, u8 elen, int idx,
+		u8 *ver, u8 *hdr, struct nvbios_ramcfg *p)
+{
+	data = nvbios_rammapSe(bios, data, ever, ehdr, ecnt, elen, idx, ver, hdr);
+	switch (!!data * *ver) {
+	case 0x11:
+		p->ramcfg_11_01_01 = (nv_ro08(bios, data + 0x01) & 0x01) >> 0;
+		p->ramcfg_11_01_02 = (nv_ro08(bios, data + 0x01) & 0x02) >> 1;
+		p->ramcfg_11_01_04 = (nv_ro08(bios, data + 0x01) & 0x04) >> 2;
+		p->ramcfg_11_01_08 = (nv_ro08(bios, data + 0x01) & 0x08) >> 3;
+		p->ramcfg_11_01_10 = (nv_ro08(bios, data + 0x01) & 0x10) >> 4;
+		p->ramcfg_11_01_20 = (nv_ro08(bios, data + 0x01) & 0x20) >> 5;
+		p->ramcfg_11_01_40 = (nv_ro08(bios, data + 0x01) & 0x40) >> 6;
+		p->ramcfg_11_01_80 = (nv_ro08(bios, data + 0x01) & 0x80) >> 7;
+		p->ramcfg_11_02_03 = (nv_ro08(bios, data + 0x02) & 0x03) >> 0;
+		p->ramcfg_11_02_04 = (nv_ro08(bios, data + 0x02) & 0x04) >> 2;
+		p->ramcfg_11_02_08 = (nv_ro08(bios, data + 0x02) & 0x08) >> 3;
+		p->ramcfg_11_02_10 = (nv_ro08(bios, data + 0x02) & 0x10) >> 4;
+		p->ramcfg_11_02_40 = (nv_ro08(bios, data + 0x02) & 0x40) >> 6;
+		p->ramcfg_11_02_80 = (nv_ro08(bios, data + 0x02) & 0x80) >> 7;
+		p->ramcfg_11_03_0f = (nv_ro08(bios, data + 0x03) & 0x0f) >> 0;
+		p->ramcfg_11_03_30 = (nv_ro08(bios, data + 0x03) & 0x30) >> 4;
+		p->ramcfg_11_03_c0 = (nv_ro08(bios, data + 0x03) & 0xc0) >> 6;
+		p->ramcfg_11_03_f0 = (nv_ro08(bios, data + 0x03) & 0xf0) >> 4;
+		p->ramcfg_11_04    = (nv_ro08(bios, data + 0x04) & 0xff) >> 0;
+		p->ramcfg_11_06    = (nv_ro08(bios, data + 0x06) & 0xff) >> 0;
+		p->ramcfg_11_07_02 = (nv_ro08(bios, data + 0x07) & 0x02) >> 1;
+		p->ramcfg_11_07_04 = (nv_ro08(bios, data + 0x07) & 0x04) >> 2;
+		p->ramcfg_11_07_08 = (nv_ro08(bios, data + 0x07) & 0x08) >> 3;
+		p->ramcfg_11_07_10 = (nv_ro08(bios, data + 0x07) & 0x10) >> 4;
+		p->ramcfg_11_07_40 = (nv_ro08(bios, data + 0x07) & 0x40) >> 6;
+		p->ramcfg_11_07_80 = (nv_ro08(bios, data + 0x07) & 0x80) >> 7;
+		p->ramcfg_11_08_01 = (nv_ro08(bios, data + 0x08) & 0x01) >> 0;
+		p->ramcfg_11_08_02 = (nv_ro08(bios, data + 0x08) & 0x02) >> 1;
+		p->ramcfg_11_08_04 = (nv_ro08(bios, data + 0x08) & 0x04) >> 2;
+		p->ramcfg_11_08_08 = (nv_ro08(bios, data + 0x08) & 0x08) >> 3;
+		p->ramcfg_11_08_10 = (nv_ro08(bios, data + 0x08) & 0x10) >> 4;
+		p->ramcfg_11_08_20 = (nv_ro08(bios, data + 0x08) & 0x20) >> 5;
+		p->ramcfg_11_09    = (nv_ro08(bios, data + 0x09) & 0xff) >> 0;
+		break;
+	default:
+		data = 0;
+		break;
+	}
+	return data;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/therm.c b/drivers/gpu/drm/nouveau/core/subdev/bios/therm.c
index 22ac6dbd6c8..d1585409407 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/therm.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/therm.c
@@ -164,6 +164,7 @@ nvbios_therm_fan_parse(struct nouveau_bios *bios,
 
 	i = 0;
 	fan->nr_fan_trip = 0;
+	fan->fan_mode = NVBIOS_THERM_FAN_OTHER;
 	while ((entry = nvbios_therm_entry(bios, i++, &ver, &len))) {
 		s16 value = nv_ro16(bios, entry + 1);
 
@@ -174,6 +175,8 @@ nvbios_therm_fan_parse(struct nouveau_bios *bios,
 			break;
 		case 0x24:
 			fan->nr_fan_trip++;
+			if (fan->fan_mode > NVBIOS_THERM_FAN_TRIP)
+				fan->fan_mode = NVBIOS_THERM_FAN_TRIP;
 			cur_trip = &fan->trip[fan->nr_fan_trip - 1];
 			cur_trip->hysteresis = value & 0xf;
 			cur_trip->temp = (value & 0xff0) >> 4;
@@ -194,11 +197,19 @@ nvbios_therm_fan_parse(struct nouveau_bios *bios,
 			fan->slow_down_period = value;
 			break;
 		case 0x46:
+			if (fan->fan_mode > NVBIOS_THERM_FAN_LINEAR)
+				fan->fan_mode = NVBIOS_THERM_FAN_LINEAR;
 			fan->linear_min_temp = nv_ro08(bios, entry + 1);
 			fan->linear_max_temp = nv_ro08(bios, entry + 2);
 			break;
 		}
 	}
 
+	/* starting from fermi, fan management is always linear */
+	if (nv_device(bios)->card_type >= NV_C0 &&
+		fan->fan_mode == NVBIOS_THERM_FAN_OTHER) {
+		fan->fan_mode = NVBIOS_THERM_FAN_LINEAR;
+	}
+
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c b/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
index 151c2d6aaee..350d44ab2ba 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
@@ -24,11 +24,12 @@
 
 #include <subdev/bios.h>
 #include <subdev/bios/bit.h>
+#include <subdev/bios/ramcfg.h>
 #include <subdev/bios/timing.h>
 
 u16
-nvbios_timing_table(struct nouveau_bios *bios,
-		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+nvbios_timingTe(struct nouveau_bios *bios,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
 {
 	struct bit_entry bit_P;
 	u16 timing = 0x0000;
@@ -47,11 +48,15 @@ nvbios_timing_table(struct nouveau_bios *bios,
 				*hdr = nv_ro08(bios, timing + 1);
 				*cnt = nv_ro08(bios, timing + 2);
 				*len = nv_ro08(bios, timing + 3);
+				*snr = 0;
+				*ssz = 0;
 				return timing;
 			case 0x20:
 				*hdr = nv_ro08(bios, timing + 1);
-				*cnt = nv_ro08(bios, timing + 3);
+				*cnt = nv_ro08(bios, timing + 5);
 				*len = nv_ro08(bios, timing + 2);
+				*snr = nv_ro08(bios, timing + 4);
+				*ssz = nv_ro08(bios, timing + 3);
 				return timing;
 			default:
 				break;
@@ -63,11 +68,60 @@ nvbios_timing_table(struct nouveau_bios *bios,
 }
 
 u16
-nvbios_timing_entry(struct nouveau_bios *bios, int idx, u8 *ver, u8 *len)
+nvbios_timingEe(struct nouveau_bios *bios, int idx,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 {
-	u8  hdr, cnt;
-	u16 timing = nvbios_timing_table(bios, ver, &hdr, &cnt, len);
-	if (timing && idx < cnt)
-		return timing + hdr + (idx * *len);
+	u8  snr, ssz;
+	u16 timing = nvbios_timingTe(bios, ver, hdr, cnt, len, &snr, &ssz);
+	if (timing && idx < *cnt) {
+		timing += *hdr + idx * (*len + (snr * ssz));
+		*hdr = *len;
+		*cnt = snr;
+		*len = ssz;
+		return timing;
+	}
 	return 0x0000;
 }
+
+u16
+nvbios_timingEp(struct nouveau_bios *bios, int idx,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		struct nvbios_ramcfg *p)
+{
+	u16 data = nvbios_timingEe(bios, idx, ver, hdr, cnt, len), temp;
+	switch (!!data * *ver) {
+	case 0x20:
+		p->timing[0] = nv_ro32(bios, data + 0x00);
+		p->timing[1] = nv_ro32(bios, data + 0x04);
+		p->timing[2] = nv_ro32(bios, data + 0x08);
+		p->timing[3] = nv_ro32(bios, data + 0x0c);
+		p->timing[4] = nv_ro32(bios, data + 0x10);
+		p->timing[5] = nv_ro32(bios, data + 0x14);
+		p->timing[6] = nv_ro32(bios, data + 0x18);
+		p->timing[7] = nv_ro32(bios, data + 0x1c);
+		p->timing[8] = nv_ro32(bios, data + 0x20);
+		p->timing[9] = nv_ro32(bios, data + 0x24);
+		p->timing[10] = nv_ro32(bios, data + 0x28);
+		p->timing_20_2e_03 = (nv_ro08(bios, data + 0x2e) & 0x03) >> 0;
+		p->timing_20_2e_30 = (nv_ro08(bios, data + 0x2e) & 0x30) >> 4;
+		p->timing_20_2e_c0 = (nv_ro08(bios, data + 0x2e) & 0xc0) >> 6;
+		p->timing_20_2f_03 = (nv_ro08(bios, data + 0x2f) & 0x03) >> 0;
+		temp = nv_ro16(bios, data + 0x2c);
+		p->timing_20_2c_003f = (temp & 0x003f) >> 0;
+		p->timing_20_2c_1fc0 = (temp & 0x1fc0) >> 6;
+		p->timing_20_30_07 = (nv_ro08(bios, data + 0x30) & 0x07) >> 0;
+		p->timing_20_30_f8 = (nv_ro08(bios, data + 0x30) & 0xf8) >> 3;
+		temp = nv_ro16(bios, data + 0x31);
+		p->timing_20_31_0007 = (temp & 0x0007) >> 0;
+		p->timing_20_31_0078 = (temp & 0x0078) >> 3;
+		p->timing_20_31_0780 = (temp & 0x0780) >> 7;
+		p->timing_20_31_0800 = (temp & 0x0800) >> 11;
+		p->timing_20_31_7000 = (temp & 0x7000) >> 12;
+		p->timing_20_31_8000 = (temp & 0x8000) >> 15;
+		break;
+	default:
+		data = 0;
+		break;
+	}
+	return data;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/base.c b/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
index e2938a21b06..22351f594d2 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
@@ -182,9 +182,12 @@ nouveau_pstate_prog(struct nouveau_clock *clk, int pstatei)
 	clk->pstate = pstatei;
 
 	if (pfb->ram->calc) {
-		ret = pfb->ram->calc(pfb, pstate->base.domain[nv_clk_src_mem]);
-		if (ret == 0)
-			ret = pfb->ram->prog(pfb);
+		int khz = pstate->base.domain[nv_clk_src_mem];
+		do {
+			ret = pfb->ram->calc(pfb, khz);
+			if (ret == 0)
+				ret = pfb->ram->prog(pfb);
+		} while (ret > 0);
 		pfb->ram->tidy(pfb);
 	}
 
@@ -343,8 +346,8 @@ nouveau_clock_ustate_update(struct nouveau_clock *clk, int req)
 	struct nouveau_pstate *pstate;
 	int i = 0;
 
-	/* YKW repellant */
-	return -ENOSYS;
+	if (!clk->allow_reclock)
+		return -ENOSYS;
 
 	if (req != -1 && req != -2) {
 		list_for_each_entry(pstate, &clk->states, head) {
@@ -453,6 +456,7 @@ nouveau_clock_create_(struct nouveau_object *parent,
 		      struct nouveau_object *engine,
 		      struct nouveau_oclass *oclass,
 		      struct nouveau_clocks *clocks,
+		      bool allow_reclock,
 		      int length, void **object)
 {
 	struct nouveau_device *device = nv_device(parent);
@@ -475,6 +479,8 @@ nouveau_clock_create_(struct nouveau_object *parent,
 		ret = nouveau_pstate_new(clk, idx++);
 	} while (ret == 0);
 
+	clk->allow_reclock = allow_reclock;
+
 	mode = nouveau_stropt(device->cfgopt, "NvClkMode", &arglen);
 	if (mode) {
 		if (!strncasecmpz(mode, "disabled", arglen)) {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nv04.c
index 30c1f3a4158..eb2d4425a49 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nv04.c
@@ -25,7 +25,7 @@
 #include <subdev/bios.h>
 #include <subdev/bios/pll.h>
 #include <subdev/clock.h>
-#include <subdev/devinit/priv.h>
+#include <subdev/devinit/nv04.h>
 
 #include "pll.h"
 
@@ -82,7 +82,8 @@ nv04_clock_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nv04_clock_priv *priv;
 	int ret;
 
-	ret = nouveau_clock_create(parent, engine, oclass, nv04_domain, &priv);
+	ret = nouveau_clock_create(parent, engine, oclass, nv04_domain, false,
+				   &priv);
 	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nv40.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nv40.c
index db7346f7908..8a9e1683979 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nv40.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nv40.c
@@ -213,7 +213,8 @@ nv40_clock_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nv40_clock_priv *priv;
 	int ret;
 
-	ret = nouveau_clock_create(parent, engine, oclass, nv40_domain, &priv);
+	ret = nouveau_clock_create(parent, engine, oclass, nv40_domain, true,
+				   &priv);
 	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nv50.c
index 250a6d96016..8c132772ba9 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nv50.c
@@ -507,7 +507,7 @@ nv50_clock_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	int ret;
 
 	ret = nouveau_clock_create(parent, engine, oclass, pclass->domains,
-				  &priv);
+				   false, &priv);
 	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c
index 4f5a1373f00..9fb58354a80 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nva3.c
@@ -302,7 +302,8 @@ nva3_clock_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nva3_clock_priv *priv;
 	int ret;
 
-	ret = nouveau_clock_create(parent, engine, oclass, nva3_domain, &priv);
+	ret = nouveau_clock_create(parent, engine, oclass, nva3_domain, false,
+				   &priv);
 	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c
index 7a723b4f564..6a65fc9e966 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nvaa.c
@@ -421,7 +421,8 @@ nvaa_clock_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nvaa_clock_priv *priv;
 	int ret;
 
-	ret = nouveau_clock_create(parent, engine, oclass, nvaa_domains, &priv);
+	ret = nouveau_clock_create(parent, engine, oclass, nvaa_domains, true,
+				   &priv);
 	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nvc0.c
index c3105720ed2..dbf8517f54d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nvc0.c
@@ -437,7 +437,8 @@ nvc0_clock_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nvc0_clock_priv *priv;
 	int ret;
 
-	ret = nouveau_clock_create(parent, engine, oclass, nvc0_domain, &priv);
+	ret = nouveau_clock_create(parent, engine, oclass, nvc0_domain, false,
+				   &priv);
 	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nve0.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nve0.c
index 4c62e84b96f..0e62a324014 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nve0.c
@@ -307,7 +307,6 @@ calc_clk(struct nve0_clock_priv *priv,
 		info->dsrc = src0;
 		if (div0) {
 			info->ddiv |= 0x80000000;
-			info->ddiv |= div0 << 8;
 			info->ddiv |= div0;
 		}
 		if (div1D) {
@@ -352,7 +351,7 @@ nve0_clock_prog_0(struct nve0_clock_priv *priv, int clk)
 {
 	struct nve0_clock_info *info = &priv->eng[clk];
 	if (!info->ssel) {
-		nv_mask(priv, 0x1371d0 + (clk * 0x04), 0x80003f3f, info->ddiv);
+		nv_mask(priv, 0x1371d0 + (clk * 0x04), 0x8000003f, info->ddiv);
 		nv_wr32(priv, 0x137160 + (clk * 0x04), info->dsrc);
 	}
 }
@@ -389,7 +388,10 @@ static void
 nve0_clock_prog_3(struct nve0_clock_priv *priv, int clk)
 {
 	struct nve0_clock_info *info = &priv->eng[clk];
-	nv_mask(priv, 0x137250 + (clk * 0x04), 0x00003f3f, info->mdiv);
+	if (info->ssel)
+		nv_mask(priv, 0x137250 + (clk * 0x04), 0x00003f00, info->mdiv);
+	else
+		nv_mask(priv, 0x137250 + (clk * 0x04), 0x0000003f, info->mdiv);
 }
 
 static void
@@ -457,7 +459,7 @@ nve0_domain[] = {
 	{ nv_clk_src_gpc    , 0x00, NVKM_CLK_DOM_FLAG_CORE, "core", 2000 },
 	{ nv_clk_src_hubk07 , 0x01, NVKM_CLK_DOM_FLAG_CORE },
 	{ nv_clk_src_rop    , 0x02, NVKM_CLK_DOM_FLAG_CORE },
-	{ nv_clk_src_mem    , 0x03, 0, "memory", 1000 },
+	{ nv_clk_src_mem    , 0x03, 0, "memory", 500 },
 	{ nv_clk_src_hubk06 , 0x04, NVKM_CLK_DOM_FLAG_CORE },
 	{ nv_clk_src_hubk01 , 0x05 },
 	{ nv_clk_src_vdec   , 0x06 },
@@ -473,7 +475,8 @@ nve0_clock_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nve0_clock_priv *priv;
 	int ret;
 
-	ret = nouveau_clock_create(parent, engine, oclass, nve0_domain, &priv);
+	ret = nouveau_clock_create(parent, engine, oclass, nve0_domain, true,
+				   &priv);
 	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c
index 79c81d3d9ba..239acfe876c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c
@@ -24,9 +24,11 @@
 
 #include <core/option.h>
 
-#include <subdev/devinit.h>
 #include <subdev/bios.h>
 #include <subdev/bios/init.h>
+#include <subdev/vga.h>
+
+#include "priv.h"
 
 int
 _nouveau_devinit_fini(struct nouveau_object *object, bool suspend)
@@ -37,18 +39,41 @@ _nouveau_devinit_fini(struct nouveau_object *object, bool suspend)
 	if (suspend)
 		devinit->post = true;
 
+	/* unlock the extended vga crtc regs */
+	nv_lockvgac(devinit, false);
+
 	return nouveau_subdev_fini(&devinit->base, suspend);
 }
 
 int
 _nouveau_devinit_init(struct nouveau_object *object)
 {
+	struct nouveau_devinit_impl *impl = (void *)object->oclass;
 	struct nouveau_devinit *devinit = (void *)object;
-	int ret = nouveau_subdev_init(&devinit->base);
+	int ret;
+
+	ret = nouveau_subdev_init(&devinit->base);
+	if (ret)
+		return ret;
+
+	ret = nvbios_init(&devinit->base, devinit->post);
 	if (ret)
 		return ret;
 
-	return nvbios_init(&devinit->base, devinit->post);
+	if (impl->disable)
+		nv_device(devinit)->disable_mask |= impl->disable(devinit);
+	return 0;
+}
+
+void
+_nouveau_devinit_dtor(struct nouveau_object *object)
+{
+	struct nouveau_devinit *devinit = (void *)object;
+
+	/* lock crtc regs */
+	nv_lockvgac(devinit, true);
+
+	nouveau_subdev_destroy(&devinit->base);
 }
 
 int
@@ -57,6 +82,7 @@ nouveau_devinit_create_(struct nouveau_object *parent,
 			struct nouveau_oclass *oclass,
 			int size, void **pobject)
 {
+	struct nouveau_devinit_impl *impl = (void *)oclass;
 	struct nouveau_device *device = nv_device(parent);
 	struct nouveau_devinit *devinit;
 	int ret;
@@ -68,5 +94,8 @@ nouveau_devinit_create_(struct nouveau_object *parent,
 		return ret;
 
 	devinit->post = nouveau_boolopt(device->cfgopt, "NvForcePost", false);
+	devinit->meminit = impl->meminit;
+	devinit->pll_set = impl->pll_set;
+	devinit->mmio    = impl->mmio;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h
index 6b56a0f4cb4..4fe49cf4c99 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/fbmem.h
@@ -24,6 +24,8 @@
  *
  */
 
+#include <core/device.h>
+
 #define NV04_PFB_BOOT_0						0x00100000
 #	define NV04_PFB_BOOT_0_RAM_AMOUNT			0x00000003
 #	define NV04_PFB_BOOT_0_RAM_AMOUNT_32MB			0x00000000
@@ -60,10 +62,10 @@
 #	define NV10_PFB_REFCTRL_VALID_1				(1 << 31)
 
 static inline struct io_mapping *
-fbmem_init(struct pci_dev *pdev)
+fbmem_init(struct nouveau_device *dev)
 {
-	return io_mapping_create_wc(pci_resource_start(pdev, 1),
-				    pci_resource_len(pdev, 1));
+	return io_mapping_create_wc(nv_device_resource_start(dev, 1),
+				    nv_device_resource_len(dev, 1));
 }
 
 static inline void
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/gm107.c
new file mode 100644
index 00000000000..c69bc7f54e3
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/gm107.c
@@ -0,0 +1,56 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nv50.h"
+
+static u64
+gm107_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r021c00 = nv_rd32(priv, 0x021c00);
+	u32 r021c04 = nv_rd32(priv, 0x021c04);
+	u64 disable = 0ULL;
+
+	if (r021c00 & 0x00000001)
+		disable |= (1ULL << NVDEV_ENGINE_COPY0);
+	if (r021c00 & 0x00000004)
+		disable |= (1ULL << NVDEV_ENGINE_COPY2);
+	if (r021c04 & 0x00000001)
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+
+	return disable;
+}
+
+struct nouveau_oclass *
+gm107_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x07),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
+		.dtor = _nouveau_devinit_dtor,
+		.init = nv50_devinit_init,
+		.fini = _nouveau_devinit_fini,
+	},
+	.pll_set = nvc0_devinit_pll_set,
+	.disable = gm107_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c
index 27c8235f1a8..052ad690b46 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c
@@ -27,12 +27,7 @@
 #include <subdev/vga.h>
 
 #include "fbmem.h"
-#include "priv.h"
-
-struct nv04_devinit_priv {
-	struct nouveau_devinit base;
-	int owner;
-};
+#include "nv04.h"
 
 static void
 nv04_devinit_meminit(struct nouveau_devinit *devinit)
@@ -43,7 +38,7 @@ nv04_devinit_meminit(struct nouveau_devinit *devinit)
 	int i;
 
 	/* Map the framebuffer aperture */
-	fb = fbmem_init(nv_device(priv)->pdev);
+	fb = fbmem_init(nv_device(priv));
 	if (!fb) {
 		nv_error(priv, "failed to map fb\n");
 		return;
@@ -393,17 +388,21 @@ int
 nv04_devinit_fini(struct nouveau_object *object, bool suspend)
 {
 	struct nv04_devinit_priv *priv = (void *)object;
+	int ret;
 
 	/* make i2c busses accessible */
 	nv_mask(priv, 0x000200, 0x00000001, 0x00000001);
 
-	/* unlock extended vga crtc regs, and unslave crtcs */
-	nv_lockvgac(priv, false);
+	ret = nouveau_devinit_fini(&priv->base, suspend);
+	if (ret)
+		return ret;
+
+	/* unslave crtcs */
 	if (priv->owner < 0)
 		priv->owner = nv_rdvgaowner(priv);
 	nv_wrvgaowner(priv, 0);
 
-	return nouveau_devinit_fini(&priv->base, suspend);
+	return 0;
 }
 
 int
@@ -431,14 +430,13 @@ nv04_devinit_dtor(struct nouveau_object *object)
 {
 	struct nv04_devinit_priv *priv = (void *)object;
 
-	/* restore vga owner saved at first init, and lock crtc regs  */
+	/* restore vga owner saved at first init */
 	nv_wrvgaowner(priv, priv->owner);
-	nv_lockvgac(priv, true);
 
 	nouveau_devinit_destroy(&priv->base);
 }
 
-static int
+int
 nv04_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
 		  struct nouveau_object **pobject)
@@ -451,19 +449,19 @@ nv04_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	priv->base.meminit = nv04_devinit_meminit;
-	priv->base.pll_set = nv04_devinit_pll_set;
 	priv->owner = -1;
 	return 0;
 }
 
-struct nouveau_oclass
-nv04_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x04),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv04_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x04),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.meminit = nv04_devinit_meminit,
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.h
new file mode 100644
index 00000000000..23470a57510
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.h
@@ -0,0 +1,23 @@
+#ifndef __NVKM_DEVINIT_NV04_H__
+#define __NVKM_DEVINIT_NV04_H__
+
+#include "priv.h"
+
+struct nv04_devinit_priv {
+	struct nouveau_devinit base;
+	u8 owner;
+};
+
+int  nv04_devinit_ctor(struct nouveau_object *, struct nouveau_object *,
+		       struct nouveau_oclass *, void *, u32,
+		       struct nouveau_object **);
+void nv04_devinit_dtor(struct nouveau_object *);
+int  nv04_devinit_init(struct nouveau_object *);
+int  nv04_devinit_fini(struct nouveau_object *, bool);
+int  nv04_devinit_pll_set(struct nouveau_devinit *, u32, u32);
+
+void setPLL_single(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
+void setPLL_double_highregs(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
+void setPLL_double_lowregs(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c
index b1912a8a894..4a19c10e517 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c
@@ -29,12 +29,7 @@
 #include <subdev/vga.h>
 
 #include "fbmem.h"
-#include "priv.h"
-
-struct nv05_devinit_priv {
-	struct nouveau_devinit base;
-	u8 owner;
-};
+#include "nv04.h"
 
 static void
 nv05_devinit_meminit(struct nouveau_devinit *devinit)
@@ -49,7 +44,7 @@ nv05_devinit_meminit(struct nouveau_devinit *devinit)
 		{ 0x06, 0x00 },
 		{ 0x00, 0x00 }
 	};
-	struct nv05_devinit_priv *priv = (void *)devinit;
+	struct nv04_devinit_priv *priv = (void *)devinit;
 	struct nouveau_bios *bios = nouveau_bios(priv);
 	struct io_mapping *fb;
 	u32 patt = 0xdeadbeef;
@@ -58,7 +53,7 @@ nv05_devinit_meminit(struct nouveau_devinit *devinit)
 	int i, v;
 
 	/* Map the framebuffer aperture */
-	fb = fbmem_init(nv_device(priv)->pdev);
+	fb = fbmem_init(nv_device(priv));
 	if (!fb) {
 		nv_error(priv, "failed to map fb\n");
 		return;
@@ -130,31 +125,15 @@ out:
 	fbmem_fini(fb);
 }
 
-static int
-nv05_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nv05_devinit_priv *priv;
-	int ret;
-
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.meminit = nv05_devinit_meminit;
-	priv->base.pll_set = nv04_devinit_pll_set;
-	return 0;
-}
-
-struct nouveau_oclass
-nv05_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x05),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv05_devinit_ctor,
+struct nouveau_oclass *
+nv05_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x05),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.meminit = nv05_devinit_meminit,
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c
index 8d274dba1ef..3b8d657da27 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c
@@ -27,17 +27,12 @@
 #include <subdev/vga.h>
 
 #include "fbmem.h"
-#include "priv.h"
-
-struct nv10_devinit_priv {
-	struct nouveau_devinit base;
-	u8 owner;
-};
+#include "nv04.h"
 
 static void
 nv10_devinit_meminit(struct nouveau_devinit *devinit)
 {
-	struct nv10_devinit_priv *priv = (void *)devinit;
+	struct nv04_devinit_priv *priv = (void *)devinit;
 	static const int mem_width[] = { 0x10, 0x00, 0x20 };
 	int mem_width_count;
 	uint32_t patt = 0xdeadbeef;
@@ -51,7 +46,7 @@ nv10_devinit_meminit(struct nouveau_devinit *devinit)
 		mem_width_count = 2;
 
 	/* Map the framebuffer aperture */
-	fb = fbmem_init(nv_device(priv)->pdev);
+	fb = fbmem_init(nv_device(priv));
 	if (!fb) {
 		nv_error(priv, "failed to map fb\n");
 		return;
@@ -101,31 +96,15 @@ amount_found:
 	fbmem_fini(fb);
 }
 
-static int
-nv10_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nv10_devinit_priv *priv;
-	int ret;
-
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.meminit = nv10_devinit_meminit;
-	priv->base.pll_set = nv04_devinit_pll_set;
-	return 0;
-}
-
-struct nouveau_oclass
-nv10_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x10),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv10_devinit_ctor,
+struct nouveau_oclass *
+nv10_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x10),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.meminit = nv10_devinit_meminit,
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c
index e9743cdabe7..526d0c6faac 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c
@@ -22,37 +22,16 @@
  * Authors: Ben Skeggs
  */
 
-#include "priv.h"
+#include "nv04.h"
 
-struct nv1a_devinit_priv {
-	struct nouveau_devinit base;
-	u8 owner;
-};
-
-static int
-nv1a_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nv1a_devinit_priv *priv;
-	int ret;
-
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.pll_set = nv04_devinit_pll_set;
-	return 0;
-}
-
-struct nouveau_oclass
-nv1a_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x1a),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv1a_devinit_ctor,
+struct nouveau_oclass *
+nv1a_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x1a),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c
index 6cc6080d3bc..04bc9732644 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c
@@ -24,25 +24,20 @@
  *
  */
 
-#include "priv.h"
+#include "nv04.h"
 #include "fbmem.h"
 
-struct nv20_devinit_priv {
-	struct nouveau_devinit base;
-	u8 owner;
-};
-
 static void
 nv20_devinit_meminit(struct nouveau_devinit *devinit)
 {
-	struct nv20_devinit_priv *priv = (void *)devinit;
+	struct nv04_devinit_priv *priv = (void *)devinit;
 	struct nouveau_device *device = nv_device(priv);
 	uint32_t mask = (device->chipset >= 0x25 ? 0x300 : 0x900);
 	uint32_t amount, off;
 	struct io_mapping *fb;
 
 	/* Map the framebuffer aperture */
-	fb = fbmem_init(nv_device(priv)->pdev);
+	fb = fbmem_init(nv_device(priv));
 	if (!fb) {
 		nv_error(priv, "failed to map fb\n");
 		return;
@@ -65,31 +60,15 @@ nv20_devinit_meminit(struct nouveau_devinit *devinit)
 	fbmem_fini(fb);
 }
 
-static int
-nv20_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nv20_devinit_priv *priv;
-	int ret;
-
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.meminit = nv20_devinit_meminit;
-	priv->base.pll_set = nv04_devinit_pll_set;
-	return 0;
-}
-
-struct nouveau_oclass
-nv20_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x20),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv20_devinit_ctor,
+struct nouveau_oclass *
+nv20_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x20),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.meminit = nv20_devinit_meminit,
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c
index 6df72247c47..b46c62a1d5d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c
@@ -28,9 +28,9 @@
 #include <subdev/bios/init.h>
 #include <subdev/vga.h>
 
-#include "priv.h"
+#include "nv50.h"
 
-static int
+int
 nv50_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 {
 	struct nv50_devinit_priv *priv = (void *)devinit;
@@ -74,6 +74,19 @@ nv50_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 	return 0;
 }
 
+static u64
+nv50_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u64 disable = 0ULL;
+
+	if (!(r001540 & 0x40000000))
+		disable |= (1ULL << NVDEV_ENGINE_MPEG);
+
+	return disable;
+}
+
 int
 nv50_devinit_init(struct nouveau_object *object)
 {
@@ -120,7 +133,7 @@ nv50_devinit_init(struct nouveau_object *object)
 	return 0;
 }
 
-static int
+int
 nv50_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
 		  struct nouveau_object **pobject)
@@ -133,17 +146,18 @@ nv50_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	priv->base.pll_set = nv50_devinit_pll_set;
 	return 0;
 }
 
-struct nouveau_oclass
-nv50_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x50),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv50_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x50),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv50_devinit_ctor,
 		.dtor = _nouveau_devinit_dtor,
 		.init = nv50_devinit_init,
 		.fini = _nouveau_devinit_fini,
 	},
-};
+	.pll_set = nv50_devinit_pll_set,
+	.disable = nv50_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h
new file mode 100644
index 00000000000..51d5076333e
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h
@@ -0,0 +1,21 @@
+#ifndef __NVKM_DEVINIT_NV50_H__
+#define __NVKM_DEVINIT_NV50_H__
+
+#include "priv.h"
+
+struct nv50_devinit_priv {
+	struct nouveau_devinit base;
+	u32 r001540;
+};
+
+int  nv50_devinit_ctor(struct nouveau_object *, struct nouveau_object *,
+		       struct nouveau_oclass *, void *, u32,
+		       struct nouveau_object **);
+int  nv50_devinit_init(struct nouveau_object *);
+int  nv50_devinit_pll_set(struct nouveau_devinit *, u32, u32);
+
+int  nva3_devinit_pll_set(struct nouveau_devinit *, u32, u32);
+
+int  nvc0_devinit_pll_set(struct nouveau_devinit *, u32, u32);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c
new file mode 100644
index 00000000000..787422505d8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c
@@ -0,0 +1,63 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nv50.h"
+
+static u64
+nv84_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u32 r00154c = nv_rd32(priv, 0x00154c);
+	u64 disable = 0ULL;
+
+	if (!(r001540 & 0x40000000)) {
+		disable |= (1ULL << NVDEV_ENGINE_MPEG);
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+		disable |= (1ULL << NVDEV_ENGINE_CRYPT);
+	}
+
+	if (!(r00154c & 0x00000004))
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+	if (!(r00154c & 0x00000020))
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (!(r00154c & 0x00000040))
+		disable |= (1ULL << NVDEV_ENGINE_CRYPT);
+
+	return disable;
+}
+
+struct nouveau_oclass *
+nv84_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x84),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
+		.dtor = _nouveau_devinit_dtor,
+		.init = nv50_devinit_init,
+		.fini = _nouveau_devinit_fini,
+	},
+	.pll_set = nv50_devinit_pll_set,
+	.disable = nv84_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c
new file mode 100644
index 00000000000..2b0e963fc6f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c
@@ -0,0 +1,62 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nv50.h"
+
+static u64
+nv98_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u32 r00154c = nv_rd32(priv, 0x00154c);
+	u64 disable = 0ULL;
+
+	if (!(r001540 & 0x40000000)) {
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+		disable |= (1ULL << NVDEV_ENGINE_PPP);
+	}
+
+	if (!(r00154c & 0x00000004))
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+	if (!(r00154c & 0x00000020))
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (!(r00154c & 0x00000040))
+		disable |= (1ULL << NVDEV_ENGINE_CRYPT);
+
+	return disable;
+}
+
+struct nouveau_oclass *
+nv98_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x98),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
+		.dtor = _nouveau_devinit_dtor,
+		.init = nv50_devinit_init,
+		.fini = _nouveau_devinit_fini,
+	},
+	.pll_set = nv50_devinit_pll_set,
+	.disable = nv98_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c
index 76a68b29014..006cf348bda 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c
@@ -22,12 +22,12 @@
  * Authors: Ben Skeggs
  */
 
-#include "priv.h"
+#include "nv50.h"
 
-static int
+int
 nva3_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 {
-	struct nva3_devinit_priv *priv = (void *)devinit;
+	struct nv50_devinit_priv *priv = (void *)devinit;
 	struct nouveau_bios *bios = nouveau_bios(priv);
 	struct nvbios_pll info;
 	int N, fN, M, P;
@@ -58,30 +58,88 @@ nva3_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 	return ret;
 }
 
-static int
-nva3_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
+static u64
+nva3_devinit_disable(struct nouveau_devinit *devinit)
 {
-	struct nv50_devinit_priv *priv;
-	int ret;
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u32 r00154c = nv_rd32(priv, 0x00154c);
+	u64 disable = 0ULL;
 
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
+	if (!(r001540 & 0x40000000)) {
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_PPP);
+	}
+
+	if (!(r00154c & 0x00000004))
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+	if (!(r00154c & 0x00000020))
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (!(r00154c & 0x00000200))
+		disable |= (1ULL << NVDEV_ENGINE_COPY0);
+
+	return disable;
+}
 
-	priv->base.pll_set = nva3_devinit_pll_set;
-	return 0;
+static u32
+nva3_devinit_mmio_part[] = {
+	0x100720, 0x1008bc, 4,
+	0x100a20, 0x100adc, 4,
+	0x100d80, 0x100ddc, 4,
+	0x110000, 0x110f9c, 4,
+	0x111000, 0x11103c, 8,
+	0x111080, 0x1110fc, 4,
+	0x111120, 0x1111fc, 4,
+	0x111300, 0x1114bc, 4,
+	0,
+};
+
+static u32
+nva3_devinit_mmio(struct nouveau_devinit *devinit, u32 addr)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 *mmio = nva3_devinit_mmio_part;
+
+	/* the init tables on some boards have INIT_RAM_RESTRICT_ZM_REG_GROUP
+	 * instructions which touch registers that may not even exist on
+	 * some configurations (Quadro 400), which causes the register
+	 * interface to screw up for some amount of time after attempting to
+	 * write to one of these, and results in all sorts of things going
+	 * horribly wrong.
+	 *
+	 * the binary driver avoids touching these registers at all, however,
+	 * the video bios doesn't care and does what the scripts say.  it's
+	 * presumed that the io-port access to priv registers isn't effected
+	 * by the screw-up bug mentioned above.
+	 *
+	 * really, a new opcode should've been invented to handle these
+	 * requirements, but whatever, it's too late for that now.
+	 */
+	while (mmio[0]) {
+		if (addr >= mmio[0] && addr <= mmio[1]) {
+			u32 part = (addr / mmio[2]) & 7;
+			if (!priv->r001540)
+				priv->r001540 = nv_rd32(priv, 0x001540);
+			if (part >= hweight8((priv->r001540 >> 16) & 0xff))
+				return ~0;
+			return addr;
+		}
+		mmio += 3;
+	}
+
+	return addr;
 }
 
-struct nouveau_oclass
-nva3_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0xa3),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nva3_devinit_ctor,
+struct nouveau_oclass *
+nva3_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0xa3),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
 		.dtor = _nouveau_devinit_dtor,
 		.init = nv50_devinit_init,
 		.fini = _nouveau_devinit_fini,
 	},
-};
+	.pll_set = nva3_devinit_pll_set,
+	.disable = nva3_devinit_disable,
+	.mmio    = nva3_devinit_mmio,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c
new file mode 100644
index 00000000000..4fc68d27eff
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c
@@ -0,0 +1,63 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nv50.h"
+
+static u64
+nvaf_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u32 r00154c = nv_rd32(priv, 0x00154c);
+	u64 disable = 0;
+
+	if (!(r001540 & 0x40000000)) {
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_PPP);
+	}
+
+	if (!(r00154c & 0x00000004))
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+	if (!(r00154c & 0x00000020))
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (!(r00154c & 0x00000040))
+		disable |= (1ULL << NVDEV_ENGINE_VIC);
+	if (!(r00154c & 0x00000200))
+		disable |= (1ULL << NVDEV_ENGINE_COPY0);
+
+	return disable;
+}
+
+struct nouveau_oclass *
+nvaf_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0xaf),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
+		.dtor = _nouveau_devinit_dtor,
+		.init = nv50_devinit_init,
+		.fini = _nouveau_devinit_fini,
+	},
+	.pll_set = nva3_devinit_pll_set,
+	.disable = nvaf_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c
index 19e265bf457..30c765747ee 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c
@@ -22,12 +22,12 @@
  * Authors: Ben Skeggs
  */
 
-#include "priv.h"
+#include "nv50.h"
 
-static int
+int
 nvc0_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 {
-	struct nvc0_devinit_priv *priv = (void *)devinit;
+	struct nv50_devinit_priv *priv = (void *)devinit;
 	struct nouveau_bios *bios = nouveau_bios(priv);
 	struct nvbios_pll info;
 	int N, fN, M, P;
@@ -59,6 +59,33 @@ nvc0_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 	return ret;
 }
 
+static u64
+nvc0_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r022500 = nv_rd32(priv, 0x022500);
+	u64 disable = 0ULL;
+
+	if (r022500 & 0x00000001)
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+
+	if (r022500 & 0x00000002) {
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_PPP);
+	}
+
+	if (r022500 & 0x00000004)
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (r022500 & 0x00000008)
+		disable |= (1ULL << NVDEV_ENGINE_VENC);
+	if (r022500 & 0x00000100)
+		disable |= (1ULL << NVDEV_ENGINE_COPY0);
+	if (r022500 & 0x00000200)
+		disable |= (1ULL << NVDEV_ENGINE_COPY1);
+
+	return disable;
+}
+
 static int
 nvc0_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
@@ -72,19 +99,20 @@ nvc0_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	priv->base.pll_set = nvc0_devinit_pll_set;
 	if (nv_rd32(priv, 0x022500) & 0x00000001)
 		priv->base.post = true;
 	return 0;
 }
 
-struct nouveau_oclass
-nvc0_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0xc0),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nvc0_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0xc0),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nvc0_devinit_ctor,
 		.dtor = _nouveau_devinit_dtor,
 		.init = nv50_devinit_init,
 		.fini = _nouveau_devinit_fini,
 	},
-};
+	.pll_set = nvc0_devinit_pll_set,
+	.disable = nvc0_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h
index 7d622e2b017..f0e8683ad84 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h
@@ -6,20 +6,33 @@
 #include <subdev/clock/pll.h>
 #include <subdev/devinit.h>
 
-void nv04_devinit_dtor(struct nouveau_object *);
-int  nv04_devinit_init(struct nouveau_object *);
-int  nv04_devinit_fini(struct nouveau_object *, bool);
-int  nv04_devinit_pll_set(struct nouveau_devinit *, u32, u32);
-
-void setPLL_single(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
-void setPLL_double_highregs(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
-void setPLL_double_lowregs(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
-
-
-struct nv50_devinit_priv {
-	struct nouveau_devinit base;
+struct nouveau_devinit_impl {
+	struct nouveau_oclass base;
+	void (*meminit)(struct nouveau_devinit *);
+	int  (*pll_set)(struct nouveau_devinit *, u32 type, u32 freq);
+	u64  (*disable)(struct nouveau_devinit *);
+	u32  (*mmio)(struct nouveau_devinit *, u32);
 };
 
-int  nv50_devinit_init(struct nouveau_object *);
+#define nouveau_devinit_create(p,e,o,d)                                        \
+	nouveau_devinit_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_devinit_destroy(p) ({                                          \
+	struct nouveau_devinit *d = (p);                                       \
+	_nouveau_devinit_dtor(nv_object(d));                                   \
+})
+#define nouveau_devinit_init(p) ({                                             \
+	struct nouveau_devinit *d = (p);                                       \
+	_nouveau_devinit_init(nv_object(d));                                   \
+})
+#define nouveau_devinit_fini(p,s) ({                                           \
+	struct nouveau_devinit *d = (p);                                       \
+	_nouveau_devinit_fini(nv_object(d), (s));                              \
+})
+
+int nouveau_devinit_create_(struct nouveau_object *, struct nouveau_object *,
+			    struct nouveau_oclass *, int, void **);
+void _nouveau_devinit_dtor(struct nouveau_object *);
+int _nouveau_devinit_init(struct nouveau_object *);
+int _nouveau_devinit_fini(struct nouveau_object *, bool suspend);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
index 34f9605ffee..66fe959b4f7 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
@@ -25,35 +25,44 @@
 #include <subdev/bios.h>
 #include "priv.h"
 
+/* binary driver only executes this path if the condition (a) is true
+ * for any configuration (combination of rammap+ramcfg+timing) that
+ * can be reached on a given card.  for now, we will execute the branch
+ * unconditionally in the hope that a "false everywhere" in the bios
+ * tables doesn't actually mean "don't touch this".
+ */
+#define NOTE00(a) 1
+
 int
-nouveau_gddr5_calc(struct nouveau_ram *ram)
+nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts)
 {
-	struct nouveau_bios *bios = nouveau_bios(ram);
-	int pd, lf, xd, vh, vr, vo;
-	int WL, CL, WR, at, dt, ds;
+	int pd, lf, xd, vh, vr, vo, l3;
+	int WL, CL, WR, at[2], dt, ds;
 	int rq = ram->freq < 1000000; /* XXX */
 
-	switch (!!ram->ramcfg.data * ram->ramcfg.version) {
+	switch (ram->ramcfg.version) {
 	case 0x11:
-		pd =  (nv_ro08(bios, ram->ramcfg.data + 0x01) & 0x80) >> 7;
-		lf =  (nv_ro08(bios, ram->ramcfg.data + 0x01) & 0x40) >> 6;
-		xd = !(nv_ro08(bios, ram->ramcfg.data + 0x01) & 0x20);
-		vh =  (nv_ro08(bios, ram->ramcfg.data + 0x02) & 0x10) >> 4;
-		vr =  (nv_ro08(bios, ram->ramcfg.data + 0x02) & 0x04) >> 2;
-		vo =   nv_ro08(bios, ram->ramcfg.data + 0x06) & 0xff;
+		pd =  ram->next->bios.ramcfg_11_01_80;
+		lf =  ram->next->bios.ramcfg_11_01_40;
+		xd = !ram->next->bios.ramcfg_11_01_20;
+		vh =  ram->next->bios.ramcfg_11_02_10;
+		vr =  ram->next->bios.ramcfg_11_02_04;
+		vo =  ram->next->bios.ramcfg_11_06;
+		l3 = !ram->next->bios.ramcfg_11_07_02;
 		break;
 	default:
 		return -ENOSYS;
 	}
 
-	switch (!!ram->timing.data * ram->timing.version) {
+	switch (ram->timing.version) {
 	case 0x20:
-		WL = (nv_ro16(bios, ram->timing.data + 0x04) & 0x0f80) >> 7;
-		CL =  nv_ro08(bios, ram->timing.data + 0x04) & 0x1f;
-		WR =  nv_ro08(bios, ram->timing.data + 0x0a) & 0x7f;
-		at = (nv_ro08(bios, ram->timing.data + 0x2e) & 0xc0) >> 6;
-		dt =  nv_ro08(bios, ram->timing.data + 0x2e) & 0x03;
-		ds =  nv_ro08(bios, ram->timing.data + 0x2f) & 0x03;
+		WL = (ram->next->bios.timing[1] & 0x00000f80) >> 7;
+		CL = (ram->next->bios.timing[1] & 0x0000001f);
+		WR = (ram->next->bios.timing[2] & 0x007f0000) >> 16;
+		at[0] = ram->next->bios.timing_20_2e_c0;
+		at[1] = ram->next->bios.timing_20_2e_30;
+		dt =  ram->next->bios.timing_20_2e_03;
+		ds =  ram->next->bios.timing_20_2f_03;
 		break;
 	default:
 		return -ENOSYS;
@@ -71,13 +80,25 @@ nouveau_gddr5_calc(struct nouveau_ram *ram)
 
 	ram->mr[1] &= ~0x0bf;
 	ram->mr[1] |= (xd & 0x01) << 7;
-	ram->mr[1] |= (at & 0x03) << 4;
+	ram->mr[1] |= (at[0] & 0x03) << 4;
 	ram->mr[1] |= (dt & 0x03) << 2;
 	ram->mr[1] |= (ds & 0x03) << 0;
 
+	/* this seems wrong, alternate field used for the broadcast
+	 * on nuts vs non-nuts configs..  meh, it matches for now.
+	 */
+	ram->mr1_nuts = ram->mr[1];
+	if (nuts) {
+		ram->mr[1] &= ~0x030;
+		ram->mr[1] |= (at[1] & 0x03) << 4;
+	}
+
 	ram->mr[3] &= ~0x020;
 	ram->mr[3] |= (rq & 0x01) << 5;
 
+	ram->mr[5] &= ~0x004;
+	ram->mr[5] |= (l3 << 2);
+
 	if (!vo)
 		vo = (ram->mr[6] & 0xff0) >> 4;
 	if (ram->mr[6] & 0x001)
@@ -86,11 +107,16 @@ nouveau_gddr5_calc(struct nouveau_ram *ram)
 	ram->mr[6] |= (vo & 0xff) << 4;
 	ram->mr[6] |= (pd & 0x01) << 0;
 
-	if (!(ram->mr[7] & 0x100))
-		vr = 0; /* binary driver does this.. bug? */
-	ram->mr[7] &= ~0x188;
-	ram->mr[7] |= (vr & 0x01) << 8;
+	if (NOTE00(vr)) {
+		ram->mr[7] &= ~0x300;
+		ram->mr[7] |= (vr & 0x03) << 8;
+	}
+	ram->mr[7] &= ~0x088;
 	ram->mr[7] |= (vh & 0x01) << 7;
 	ram->mr[7] |= (lf & 0x01) << 3;
+
+	ram->mr[8] &= ~0x003;
+	ram->mr[8] |= (WR & 0x10) >> 3;
+	ram->mr[8] |= (CL & 0x10) >> 4;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/gk20a.c b/drivers/gpu/drm/nouveau/core/subdev/fb/gk20a.c
new file mode 100644
index 00000000000..a16024a7477
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/gk20a.c
@@ -0,0 +1,56 @@
+/*
+ * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include "nvc0.h"
+
+struct gk20a_fb_priv {
+	struct nouveau_fb base;
+};
+
+static int
+gk20a_fb_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	     struct nouveau_oclass *oclass, void *data, u32 size,
+	     struct nouveau_object **pobject)
+{
+	struct gk20a_fb_priv *priv;
+	int ret;
+
+	ret = nouveau_fb_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+struct nouveau_oclass *
+gk20a_fb_oclass = &(struct nouveau_fb_impl) {
+	.base.handle = NV_SUBDEV(FB, 0xea),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gk20a_fb_ctor,
+		.dtor = _nouveau_fb_dtor,
+		.init = _nouveau_fb_init,
+		.fini = _nouveau_fb_fini,
+	},
+	.memtype = nvc0_fb_memtype_valid,
+	.ram = &gk20a_ram_oclass,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/fb/gm107.c
new file mode 100644
index 00000000000..c4840aedc2d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/gm107.c
@@ -0,0 +1,38 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nvc0.h"
+
+struct nouveau_oclass *
+gm107_fb_oclass = &(struct nouveau_fb_impl) {
+	.base.handle = NV_SUBDEV(FB, 0x07),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nvc0_fb_ctor,
+		.dtor = nvc0_fb_dtor,
+		.init = nvc0_fb_init,
+		.fini = _nouveau_fb_fini,
+	},
+	.memtype = nvc0_fb_memtype_valid,
+	.ram = &gm107_ram_oclass,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv1a.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv1a.c
index 9159a5ccee9..265d1253624 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv1a.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv1a.c
@@ -36,7 +36,7 @@ nv1a_fb_oclass = &(struct nv04_fb_impl) {
 		.fini = _nouveau_fb_fini,
 	},
 	.base.memtype = nv04_fb_memtype_valid,
-	.base.ram = &nv10_ram_oclass,
+	.base.ram = &nv1a_ram_oclass,
 	.tile.regions = 8,
 	.tile.init = nv10_fb_tile_init,
 	.tile.fini = nv10_fb_tile_fini,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nv50.c
index cbc7f00c127..1fc55c1e91a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nv50.c
@@ -250,10 +250,8 @@ nv50_fb_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 	priv->r100c08_page = alloc_page(GFP_KERNEL | __GFP_ZERO);
 	if (priv->r100c08_page) {
-		priv->r100c08 = pci_map_page(device->pdev, priv->r100c08_page,
-					     0, PAGE_SIZE,
-					     PCI_DMA_BIDIRECTIONAL);
-		if (pci_dma_mapping_error(device->pdev, priv->r100c08))
+		priv->r100c08 = nv_device_map_page(device, priv->r100c08_page);
+		if (!priv->r100c08)
 			nv_warn(priv, "failed 0x100c08 page map\n");
 	} else {
 		nv_warn(priv, "failed 0x100c08 page alloc\n");
@@ -270,8 +268,7 @@ nv50_fb_dtor(struct nouveau_object *object)
 	struct nv50_fb_priv *priv = (void *)object;
 
 	if (priv->r100c08_page) {
-		pci_unmap_page(device->pdev, priv->r100c08, PAGE_SIZE,
-			       PCI_DMA_BIDIRECTIONAL);
+		nv_device_unmap_page(device, priv->r100c08);
 		__free_page(priv->r100c08_page);
 	}
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.c
index e5fc37c4caa..0670ae33ee4 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.c
@@ -33,6 +33,21 @@ nvc0_fb_memtype_valid(struct nouveau_fb *pfb, u32 tile_flags)
 	return likely((nvc0_pte_storage_type_map[memtype] != 0xff));
 }
 
+static void
+nvc0_fb_intr(struct nouveau_subdev *subdev)
+{
+	struct nvc0_fb_priv *priv = (void *)subdev;
+	u32 intr = nv_rd32(priv, 0x000100);
+	if (intr & 0x08000000) {
+		nv_debug(priv, "PFFB intr\n");
+		intr &= ~0x08000000;
+	}
+	if (intr & 0x00002000) {
+		nv_debug(priv, "PBFB intr\n");
+		intr &= ~0x00002000;
+	}
+}
+
 int
 nvc0_fb_init(struct nouveau_object *object)
 {
@@ -55,8 +70,7 @@ nvc0_fb_dtor(struct nouveau_object *object)
 	struct nvc0_fb_priv *priv = (void *)object;
 
 	if (priv->r100c10_page) {
-		pci_unmap_page(device->pdev, priv->r100c10, PAGE_SIZE,
-			       PCI_DMA_BIDIRECTIONAL);
+		nv_device_unmap_page(device, priv->r100c10);
 		__free_page(priv->r100c10_page);
 	}
 
@@ -79,13 +93,12 @@ nvc0_fb_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 	priv->r100c10_page = alloc_page(GFP_KERNEL | __GFP_ZERO);
 	if (priv->r100c10_page) {
-		priv->r100c10 = pci_map_page(device->pdev, priv->r100c10_page,
-					     0, PAGE_SIZE,
-					     PCI_DMA_BIDIRECTIONAL);
-		if (pci_dma_mapping_error(device->pdev, priv->r100c10))
+		priv->r100c10 = nv_device_map_page(device, priv->r100c10_page);
+		if (!priv->r100c10)
 			return -EFAULT;
 	}
 
+	nv_subdev(priv)->intr = nvc0_fb_intr;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.h b/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.h
index 9e1931eb746..705a06d755a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.h
@@ -18,12 +18,14 @@ int  nvc0_fb_init(struct nouveau_object *);
 bool nvc0_fb_memtype_valid(struct nouveau_fb *, u32);
 
 
-#define nvc0_ram_create(p,e,o,d)                                               \
-	nvc0_ram_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nvc0_ram_create(p,e,o,m,d)                                             \
+	nvc0_ram_create_((p), (e), (o), (m), sizeof(**d), (void **)d)
 int  nvc0_ram_create_(struct nouveau_object *, struct nouveau_object *,
-		      struct nouveau_oclass *, int, void **);
+		      struct nouveau_oclass *, u32, int, void **);
 int  nvc0_ram_get(struct nouveau_fb *, u64, u32, u32, u32,
 		  struct nouveau_mem **);
 void nvc0_ram_put(struct nouveau_fb *, struct nouveau_mem **);
 
+int  nve0_ram_init(struct nouveau_object*);
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h b/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
index 493125214e8..82273f832e4 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
@@ -32,9 +32,11 @@ extern struct nouveau_oclass nva3_ram_oclass;
 extern struct nouveau_oclass nvaa_ram_oclass;
 extern struct nouveau_oclass nvc0_ram_oclass;
 extern struct nouveau_oclass nve0_ram_oclass;
+extern struct nouveau_oclass gk20a_ram_oclass;
+extern struct nouveau_oclass gm107_ram_oclass;
 
 int nouveau_sddr3_calc(struct nouveau_ram *ram);
-int nouveau_gddr5_calc(struct nouveau_ram *ram);
+int nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts);
 
 #define nouveau_fb_create(p,e,c,d)                                             \
 	nouveau_fb_create_((p), (e), (c), sizeof(**d), (void **)d)
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h b/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h
index 0f57fcfe0bb..2af9cfd2c60 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramfuc.h
@@ -26,7 +26,7 @@ ramfuc_reg2(u32 addr1, u32 addr2)
 	};
 }
 
-static inline struct ramfuc_reg
+static noinline struct ramfuc_reg
 ramfuc_reg(u32 addr)
 {
 	return ramfuc_reg2(addr, addr);
@@ -107,7 +107,7 @@ ramfuc_nsec(struct ramfuc *ram, u32 nsec)
 
 #define ram_init(s,p)       ramfuc_init(&(s)->base, (p))
 #define ram_exec(s,e)       ramfuc_exec(&(s)->base, (e))
-#define ram_have(s,r)       ((s)->r_##r.addr != 0x000000)
+#define ram_have(s,r)       ((s)->r_##r.addr[0] != 0x000000)
 #define ram_rd32(s,r)       ramfuc_rd32(&(s)->base, &(s)->r_##r)
 #define ram_wr32(s,r,d)     ramfuc_wr32(&(s)->base, &(s)->r_##r, (d))
 #define ram_nuke(s,r)       ramfuc_nuke(&(s)->base, &(s)->r_##r)
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramgk20a.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramgk20a.c
new file mode 100644
index 00000000000..4d77d75e467
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramgk20a.c
@@ -0,0 +1,152 @@
+/*
+ * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include "priv.h"
+
+#include <subdev/fb.h>
+
+struct gk20a_mem {
+	struct nouveau_mem base;
+	void *cpuaddr;
+	dma_addr_t handle;
+};
+#define to_gk20a_mem(m) container_of(m, struct gk20a_mem, base)
+
+static void
+gk20a_ram_put(struct nouveau_fb *pfb, struct nouveau_mem **pmem)
+{
+	struct device *dev = nv_device_base(nv_device(pfb));
+	struct gk20a_mem *mem = to_gk20a_mem(*pmem);
+
+	*pmem = NULL;
+	if (unlikely(mem == NULL))
+		return;
+
+	if (likely(mem->cpuaddr))
+		dma_free_coherent(dev, mem->base.size << PAGE_SHIFT,
+				  mem->cpuaddr, mem->handle);
+
+	kfree(mem->base.pages);
+	kfree(mem);
+}
+
+static int
+gk20a_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
+	     u32 memtype, struct nouveau_mem **pmem)
+{
+	struct device *dev = nv_device_base(nv_device(pfb));
+	struct gk20a_mem *mem;
+	u32 type = memtype & 0xff;
+	u32 npages, order;
+	int i;
+
+	nv_debug(pfb, "%s: size: %llx align: %x, ncmin: %x\n", __func__, size,
+		 align, ncmin);
+
+	npages = size >> PAGE_SHIFT;
+	if (npages == 0)
+		npages = 1;
+
+	if (align == 0)
+		align = PAGE_SIZE;
+	align >>= PAGE_SHIFT;
+
+	/* round alignment to the next power of 2, if needed */
+	order = fls(align);
+	if ((align & (align - 1)) == 0)
+		order--;
+	align = BIT(order);
+
+	/* ensure returned address is correctly aligned */
+	npages = max(align, npages);
+
+	mem = kzalloc(sizeof(*mem), GFP_KERNEL);
+	if (!mem)
+		return -ENOMEM;
+
+	mem->base.size = npages;
+	mem->base.memtype = type;
+
+	mem->base.pages = kzalloc(sizeof(dma_addr_t) * npages, GFP_KERNEL);
+	if (!mem->base.pages) {
+		kfree(mem);
+		return -ENOMEM;
+	}
+
+	*pmem = &mem->base;
+
+	mem->cpuaddr = dma_alloc_coherent(dev, npages << PAGE_SHIFT,
+					  &mem->handle, GFP_KERNEL);
+	if (!mem->cpuaddr) {
+		nv_error(pfb, "%s: cannot allocate memory!\n", __func__);
+		gk20a_ram_put(pfb, pmem);
+		return -ENOMEM;
+	}
+
+	align <<= PAGE_SHIFT;
+
+	/* alignment check */
+	if (unlikely(mem->handle & (align - 1)))
+		nv_warn(pfb, "memory not aligned as requested: %pad (0x%x)\n",
+			&mem->handle, align);
+
+	nv_debug(pfb, "alloc size: 0x%x, align: 0x%x, paddr: %pad, vaddr: %p\n",
+		 npages << PAGE_SHIFT, align, &mem->handle, mem->cpuaddr);
+
+	for (i = 0; i < npages; i++)
+		mem->base.pages[i] = mem->handle + (PAGE_SIZE * i);
+
+	mem->base.offset = (u64)mem->base.pages[0];
+
+	return 0;
+}
+
+static int
+gk20a_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	      struct nouveau_oclass *oclass, void *data, u32 datasize,
+	      struct nouveau_object **pobject)
+{
+	struct nouveau_ram *ram;
+	int ret;
+
+	ret = nouveau_ram_create(parent, engine, oclass, &ram);
+	*pobject = nv_object(ram);
+	if (ret)
+		return ret;
+	ram->type = NV_MEM_TYPE_STOLEN;
+	ram->size = get_num_physpages() << PAGE_SHIFT;
+
+	ram->get = gk20a_ram_get;
+	ram->put = gk20a_ram_put;
+
+	return 0;
+}
+
+struct nouveau_oclass
+gk20a_ram_oclass = {
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gk20a_ram_ctor,
+		.dtor = _nouveau_ram_dtor,
+		.init = _nouveau_ram_init,
+		.fini = _nouveau_ram_fini,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramgm107.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramgm107.c
new file mode 100644
index 00000000000..4c6363595c7
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramgm107.c
@@ -0,0 +1,56 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nvc0.h"
+
+struct gm107_ram {
+	struct nouveau_ram base;
+};
+
+static int
+gm107_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct gm107_ram *ram;
+	int ret;
+
+	ret = nvc0_ram_create(parent, engine, oclass, 0x021c14, &ram);
+	*pobject = nv_object(ram);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+struct nouveau_oclass
+gm107_ram_oclass = {
+	.handle = 0,
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gm107_ram_ctor,
+		.dtor = _nouveau_ram_dtor,
+		.init = nve0_ram_init,
+		.fini = _nouveau_ram_fini,
+	}
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
index 76762a17d89..e5d12c24cc4 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
@@ -70,13 +70,11 @@ nv50_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	struct nv50_ramseq *hwsq = &ram->hwsq;
 	struct nvbios_perfE perfE;
 	struct nvbios_pll mpll;
-	struct bit_entry M;
 	struct {
 		u32 data;
 		u8  size;
 	} ramcfg, timing;
-	u8  ver, hdr, cnt, strap;
-	u32 data;
+	u8  ver, hdr, cnt, len, strap;
 	int N1, M1, N2, M2, P;
 	int ret, i;
 
@@ -93,16 +91,7 @@ nv50_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	} while (perfE.memory < freq);
 
 	/* locate specific data set for the attached memory */
-	if (bit_entry(bios, 'M', &M) || M.version != 1 || M.length < 5) {
-		nv_error(pfb, "invalid/missing memory table\n");
-		return -EINVAL;
-	}
-
-	strap = (nv_rd32(pfb, 0x101000) & 0x0000003c) >> 2;
-	data = nv_ro16(bios, M.offset + 3);
-	if (data)
-		strap = nv_ro08(bios, data + strap);
-
+	strap = nvbios_ramcfg_index(nv_subdev(pfb));
 	if (strap >= cnt) {
 		nv_error(pfb, "invalid ramcfg strap\n");
 		return -EINVAL;
@@ -113,7 +102,8 @@ nv50_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	/* lookup memory timings, if bios says they're present */
 	strap = nv_ro08(bios, ramcfg.data + 0x01);
 	if (strap != 0xff) {
-		timing.data = nvbios_timing_entry(bios, strap, &ver, &hdr);
+		timing.data = nvbios_timingEe(bios, strap, &ver, &hdr,
+					     &cnt, &len);
 		if (!timing.data || ver != 0x10 || hdr < 0x12) {
 			nv_error(pfb, "invalid/missing timing entry "
 				 "%02x %04x %02x %02x\n",
@@ -221,7 +211,7 @@ nv50_ram_prog(struct nouveau_fb *pfb)
 	struct nv50_ram *ram = (void *)pfb->ram;
 	struct nv50_ramseq *hwsq = &ram->hwsq;
 
-	ram_exec(hwsq, nouveau_boolopt(device->cfgopt, "NvMemExec", false));
+	ram_exec(hwsq, nouveau_boolopt(device->cfgopt, "NvMemExec", true));
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
index f6292cd9207..8076fb195dd 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
@@ -79,8 +79,7 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	struct nva3_ram *ram = (void *)pfb->ram;
 	struct nva3_ramfuc *fuc = &ram->fuc;
 	struct nva3_clock_info mclk;
-	struct bit_entry M;
-	u8  ver, cnt, strap;
+	u8  ver, cnt, len, strap;
 	u32 data;
 	struct {
 		u32 data;
@@ -91,24 +90,15 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	int ret;
 
 	/* lookup memory config data relevant to the target frequency */
-	rammap.data = nvbios_rammap_match(bios, freq / 1000, &ver, &rammap.size,
-					 &cnt, &ramcfg.size);
+	rammap.data = nvbios_rammapEm(bios, freq / 1000, &ver, &rammap.size,
+				     &cnt, &ramcfg.size);
 	if (!rammap.data || ver != 0x10 || rammap.size < 0x0e) {
 		nv_error(pfb, "invalid/missing rammap entry\n");
 		return -EINVAL;
 	}
 
 	/* locate specific data set for the attached memory */
-	if (bit_entry(bios, 'M', &M) || M.version != 2 || M.length < 3) {
-		nv_error(pfb, "invalid/missing memory table\n");
-		return -EINVAL;
-	}
-
-	strap = (nv_rd32(pfb, 0x101000) & 0x0000003c) >> 2;
-	data = nv_ro16(bios, M.offset + 1);
-	if (data)
-		strap = nv_ro08(bios, data + strap);
-
+	strap = nvbios_ramcfg_index(nv_subdev(pfb));
 	if (strap >= cnt) {
 		nv_error(pfb, "invalid ramcfg strap\n");
 		return -EINVAL;
@@ -123,8 +113,8 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	/* lookup memory timings, if bios says they're present */
 	strap = nv_ro08(bios, ramcfg.data + 0x01);
 	if (strap != 0xff) {
-		timing.data = nvbios_timing_entry(bios, strap, &ver,
-						 &timing.size);
+		timing.data = nvbios_timingEe(bios, strap, &ver, &timing.size,
+					     &cnt, &len);
 		if (!timing.data || ver != 0x10 || timing.size < 0x19) {
 			nv_error(pfb, "invalid/missing timing entry\n");
 			return -EINVAL;
@@ -319,7 +309,7 @@ nva3_ram_prog(struct nouveau_fb *pfb)
 	struct nouveau_device *device = nv_device(pfb);
 	struct nva3_ram *ram = (void *)pfb->ram;
 	struct nva3_ramfuc *fuc = &ram->fuc;
-	ram_exec(fuc, nouveau_boolopt(device->cfgopt, "NvMemExec", false));
+	ram_exec(fuc, nouveau_boolopt(device->cfgopt, "NvMemExec", true));
 	return 0;
 }
 
@@ -345,21 +335,23 @@ nva3_ram_init(struct nouveau_object *object)
 	/* prepare for ddr link training, and load training patterns */
 	switch (ram->base.type) {
 	case NV_MEM_TYPE_DDR3: {
-		static const u32 pattern[16] = {
-			0xaaaaaaaa, 0xcccccccc, 0xdddddddd, 0xeeeeeeee,
-			0x00000000, 0x11111111, 0x44444444, 0xdddddddd,
-			0x33333333, 0x55555555, 0x77777777, 0x66666666,
-			0x99999999, 0x88888888, 0xeeeeeeee, 0xbbbbbbbb,
-		};
-
-		nv_wr32(pfb, 0x100538, 0x10001ff6); /*XXX*/
-		nv_wr32(pfb, 0x1005a8, 0x0000ffff);
-		nv_mask(pfb, 0x10f800, 0x00000001, 0x00000001);
-		for (i = 0; i < 0x30; i++) {
-			nv_wr32(pfb, 0x10f8c0, (i << 8) | i);
-			nv_wr32(pfb, 0x10f8e0, (i << 8) | i);
-			nv_wr32(pfb, 0x10f900, pattern[i % 16]);
-			nv_wr32(pfb, 0x10f920, pattern[i % 16]);
+		if (nv_device(pfb)->chipset == 0xa8) {
+			static const u32 pattern[16] = {
+				0xaaaaaaaa, 0xcccccccc, 0xdddddddd, 0xeeeeeeee,
+				0x00000000, 0x11111111, 0x44444444, 0xdddddddd,
+				0x33333333, 0x55555555, 0x77777777, 0x66666666,
+				0x99999999, 0x88888888, 0xeeeeeeee, 0xbbbbbbbb,
+			};
+
+			nv_wr32(pfb, 0x100538, 0x10001ff6); /*XXX*/
+			nv_wr32(pfb, 0x1005a8, 0x0000ffff);
+			nv_mask(pfb, 0x10f800, 0x00000001, 0x00000001);
+			for (i = 0; i < 0x30; i++) {
+				nv_wr32(pfb, 0x10f8c0, (i << 8) | i);
+				nv_wr32(pfb, 0x10f8e0, (i << 8) | i);
+				nv_wr32(pfb, 0x10f900, pattern[i % 16]);
+				nv_wr32(pfb, 0x10f920, pattern[i % 16]);
+			}
 		}
 	}
 		break;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
index f464547c6ba..5a6a5027f74 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
@@ -23,7 +23,6 @@
  */
 
 #include <subdev/bios.h>
-#include <subdev/bios/bit.h>
 #include <subdev/bios/pll.h>
 #include <subdev/bios/rammap.h>
 #include <subdev/bios/timing.h>
@@ -134,9 +133,7 @@ nvc0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nvc0_ram *ram = (void *)pfb->ram;
 	struct nvc0_ramfuc *fuc = &ram->fuc;
-	struct bit_entry M;
-	u8  ver, cnt, strap;
-	u32 data;
+	u8  ver, cnt, len, strap;
 	struct {
 		u32 data;
 		u8  size;
@@ -147,24 +144,15 @@ nvc0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	int ret;
 
 	/* lookup memory config data relevant to the target frequency */
-	rammap.data = nvbios_rammap_match(bios, freq / 1000, &ver, &rammap.size,
-					 &cnt, &ramcfg.size);
+	rammap.data = nvbios_rammapEm(bios, freq / 1000, &ver, &rammap.size,
+				     &cnt, &ramcfg.size);
 	if (!rammap.data || ver != 0x10 || rammap.size < 0x0e) {
 		nv_error(pfb, "invalid/missing rammap entry\n");
 		return -EINVAL;
 	}
 
 	/* locate specific data set for the attached memory */
-	if (bit_entry(bios, 'M', &M) || M.version != 2 || M.length < 3) {
-		nv_error(pfb, "invalid/missing memory table\n");
-		return -EINVAL;
-	}
-
-	strap = (nv_rd32(pfb, 0x101000) & 0x0000003c) >> 2;
-	data = nv_ro16(bios, M.offset + 1);
-	if (data)
-		strap = nv_ro08(bios, data + strap);
-
+	strap = nvbios_ramcfg_index(nv_subdev(pfb));
 	if (strap >= cnt) {
 		nv_error(pfb, "invalid ramcfg strap\n");
 		return -EINVAL;
@@ -179,8 +167,8 @@ nvc0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	/* lookup memory timings, if bios says they're present */
 	strap = nv_ro08(bios, ramcfg.data + 0x01);
 	if (strap != 0xff) {
-		timing.data = nvbios_timing_entry(bios, strap, &ver,
-						 &timing.size);
+		timing.data = nvbios_timingEe(bios, strap, &ver, &timing.size,
+					     &cnt, &len);
 		if (!timing.data || ver != 0x10 || timing.size < 0x19) {
 			nv_error(pfb, "invalid/missing timing entry\n");
 			return -EINVAL;
@@ -420,7 +408,7 @@ nvc0_ram_prog(struct nouveau_fb *pfb)
 	struct nouveau_device *device = nv_device(pfb);
 	struct nvc0_ram *ram = (void *)pfb->ram;
 	struct nvc0_ramfuc *fuc = &ram->fuc;
-	ram_exec(fuc, nouveau_boolopt(device->cfgopt, "NvMemExec", false));
+	ram_exec(fuc, nouveau_boolopt(device->cfgopt, "NvMemExec", true));
 	return 0;
 }
 
@@ -517,7 +505,8 @@ nvc0_ram_get(struct nouveau_fb *pfb, u64 size, u32 align, u32 ncmin,
 
 int
 nvc0_ram_create_(struct nouveau_object *parent, struct nouveau_object *engine,
-		 struct nouveau_oclass *oclass, int size, void **pobject)
+		 struct nouveau_oclass *oclass, u32 maskaddr, int size,
+		 void **pobject)
 {
 	struct nouveau_fb *pfb = nouveau_fb(parent);
 	struct nouveau_bios *bios = nouveau_bios(pfb);
@@ -525,7 +514,7 @@ nvc0_ram_create_(struct nouveau_object *parent, struct nouveau_object *engine,
 	const u32 rsvd_head = ( 256 * 1024) >> 12; /* vga memory */
 	const u32 rsvd_tail = (1024 * 1024) >> 12; /* vbios etc */
 	u32 parts = nv_rd32(pfb, 0x022438);
-	u32 pmask = nv_rd32(pfb, 0x022554);
+	u32 pmask = nv_rd32(pfb, maskaddr);
 	u32 bsize = nv_rd32(pfb, 0x10f20c);
 	u32 offset, length;
 	bool uniform = true;
@@ -642,7 +631,7 @@ nvc0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nvc0_ram *ram;
 	int ret;
 
-	ret = nvc0_ram_create(parent, engine, oclass, &ram);
+	ret = nvc0_ram_create(parent, engine, oclass, 0x022554, &ram);
 	*pobject = nv_object(ram);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
index bc86cfd084f..c5b46e30231 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
@@ -25,7 +25,6 @@
 #include <subdev/gpio.h>
 
 #include <subdev/bios.h>
-#include <subdev/bios/bit.h>
 #include <subdev/bios/pll.h>
 #include <subdev/bios/init.h>
 #include <subdev/bios/rammap.h>
@@ -42,6 +41,14 @@
 
 #include "ramfuc.h"
 
+/* binary driver only executes this path if the condition (a) is true
+ * for any configuration (combination of rammap+ramcfg+timing) that
+ * can be reached on a given card.  for now, we will execute the branch
+ * unconditionally in the hope that a "false everywhere" in the bios
+ * tables doesn't actually mean "don't touch this".
+ */
+#define NOTE00(a) 1
+
 struct nve0_ramfuc {
 	struct ramfuc base;
 
@@ -104,7 +111,9 @@ struct nve0_ramfuc {
 	struct ramfuc_reg r_mr[16]; /* MR0 - MR8, MR15 */
 
 	struct ramfuc_reg r_0x62c000;
+
 	struct ramfuc_reg r_0x10f200;
+
 	struct ramfuc_reg r_0x10f210;
 	struct ramfuc_reg r_0x10f310;
 	struct ramfuc_reg r_0x10f314;
@@ -118,12 +127,17 @@ struct nve0_ramfuc {
 	struct ramfuc_reg r_0x10f65c;
 	struct ramfuc_reg r_0x10f6bc;
 	struct ramfuc_reg r_0x100710;
-	struct ramfuc_reg r_0x10f750;
+	struct ramfuc_reg r_0x100750;
 };
 
 struct nve0_ram {
 	struct nouveau_ram base;
 	struct nve0_ramfuc fuc;
+
+	u32 parts;
+	u32 pmask;
+	u32 pnuts;
+
 	int from;
 	int mode;
 	int N1, fN1, M1, P1;
@@ -134,17 +148,17 @@ struct nve0_ram {
  * GDDR5
  ******************************************************************************/
 static void
-train(struct nve0_ramfuc *fuc, u32 magic)
+nve0_ram_train(struct nve0_ramfuc *fuc, u32 mask, u32 data)
 {
 	struct nve0_ram *ram = container_of(fuc, typeof(*ram), fuc);
-	struct nouveau_fb *pfb = nouveau_fb(ram);
-	const int mc = nv_rd32(pfb, 0x02243c);
-	int i;
-
-	ram_mask(fuc, 0x10f910, 0xbc0e0000, magic);
-	ram_mask(fuc, 0x10f914, 0xbc0e0000, magic);
-	for (i = 0; i < mc; i++) {
-		const u32 addr = 0x110974 + (i * 0x1000);
+	u32 addr = 0x110974, i;
+
+	ram_mask(fuc, 0x10f910, mask, data);
+	ram_mask(fuc, 0x10f914, mask, data);
+
+	for (i = 0; (data & 0x80000000) && i < ram->parts; addr += 0x1000, i++) {
+		if (ram->pmask & (1 << i))
+			continue;
 		ram_wait(fuc, addr, 0x0000000f, 0x00000000, 500000);
 	}
 }
@@ -186,6 +200,7 @@ r1373f4_init(struct nve0_ramfuc *fuc)
 	/* (re)program mempll, if required */
 	if (ram->mode == 2) {
 		ram_mask(fuc, 0x1373f4, 0x00010000, 0x00000000);
+		ram_mask(fuc, 0x132000, 0x80000000, 0x80000000);
 		ram_mask(fuc, 0x132000, 0x00000001, 0x00000000);
 		ram_mask(fuc, 0x132004, 0x103fffff, mcoef);
 		ram_mask(fuc, 0x132000, 0x00000001, 0x00000001);
@@ -199,12 +214,12 @@ r1373f4_init(struct nve0_ramfuc *fuc)
 }
 
 static void
-r1373f4_fini(struct nve0_ramfuc *fuc, u32 ramcfg)
+r1373f4_fini(struct nve0_ramfuc *fuc)
 {
 	struct nve0_ram *ram = container_of(fuc, typeof(*ram), fuc);
-	struct nouveau_bios *bios = nouveau_bios(ram);
-	u8 v0 = (nv_ro08(bios, ramcfg + 0x03) & 0xc0) >> 6;
-	u8 v1 = (nv_ro08(bios, ramcfg + 0x03) & 0x30) >> 4;
+	struct nouveau_ram_data *next = ram->base.next;
+	u8 v0 = next->bios.ramcfg_11_03_c0;
+	u8 v1 = next->bios.ramcfg_11_03_30;
 	u32 tmp;
 
 	tmp = ram_rd32(fuc, 0x1373ec) & ~0x00030000;
@@ -220,25 +235,46 @@ r1373f4_fini(struct nve0_ramfuc *fuc, u32 ramcfg)
 	ram_mask(fuc, 0x10f800, 0x00000030, (v0 ^ v1) << 4);
 }
 
+static void
+nve0_ram_nuts(struct nve0_ram *ram, struct ramfuc_reg *reg,
+	      u32 _mask, u32 _data, u32 _copy)
+{
+	struct nve0_fb_priv *priv = (void *)nouveau_fb(ram);
+	struct ramfuc *fuc = &ram->fuc.base;
+	u32 addr = 0x110000 + (reg->addr[0] & 0xfff);
+	u32 mask = _mask | _copy;
+	u32 data = (_data & _mask) | (reg->data & _copy);
+	u32 i;
+
+	for (i = 0; i < 16; i++, addr += 0x1000) {
+		if (ram->pnuts & (1 << i)) {
+			u32 prev = nv_rd32(priv, addr);
+			u32 next = (prev & ~mask) | data;
+			nouveau_memx_wr32(fuc->memx, addr, next);
+		}
+	}
+}
+#define ram_nuts(s,r,m,d,c)                                                    \
+	nve0_ram_nuts((s), &(s)->fuc.r_##r, (m), (d), (c))
+
 static int
 nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 {
-	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
 	struct nve0_ramfuc *fuc = &ram->fuc;
-	const u32 rammap = ram->base.rammap.data;
-	const u32 ramcfg = ram->base.ramcfg.data;
-	const u32 timing = ram->base.timing.data;
-	int vc = !(nv_ro08(bios, ramcfg + 0x02) & 0x08);
-	int mv = 1; /*XXX*/
+	struct nouveau_ram_data *next = ram->base.next;
+	int vc = !next->bios.ramcfg_11_02_08;
+	int mv = !next->bios.ramcfg_11_02_04;
 	u32 mask, data;
 
 	ram_mask(fuc, 0x10f808, 0x40000000, 0x40000000);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0000);
 
 	/* MR1: turn termination on early, for some reason.. */
-	if ((ram->base.mr[1] & 0x03c) != 0x030)
+	if ((ram->base.mr[1] & 0x03c) != 0x030) {
 		ram_mask(fuc, mr[1], 0x03c, ram->base.mr[1] & 0x03c);
+		ram_nuts(ram, mr[1], 0x03c, ram->base.mr1_nuts & 0x03c, 0x000);
+	}
 
 	if (vc == 1 && ram_have(fuc, gpio2E)) {
 		u32 temp  = ram_mask(fuc, gpio2E, 0x3000, fuc->r_func2E[1]);
@@ -250,8 +286,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	ram_mask(fuc, 0x10f200, 0x00000800, 0x00000000);
 
-	ram_mask(fuc, 0x10f914, 0x01020000, 0x000c0000);
-	ram_mask(fuc, 0x10f910, 0x01020000, 0x000c0000);
+	nve0_ram_train(fuc, 0x01020000, 0x000c0000);
 
 	ram_wr32(fuc, 0x10f210, 0x00000000); /* REFRESH_AUTO = 0 */
 	ram_nsec(fuc, 1000);
@@ -280,28 +315,28 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	if (1) {
 		data |= 0x800807e0;
-		switch (nv_ro08(bios, ramcfg + 0x03) & 0xc0) {
-		case 0xc0: data &= ~0x00000040; break;
-		case 0x80: data &= ~0x00000100; break;
-		case 0x40: data &= ~0x80000000; break;
-		case 0x00: data &= ~0x00000400; break;
+		switch (next->bios.ramcfg_11_03_c0) {
+		case 3: data &= ~0x00000040; break;
+		case 2: data &= ~0x00000100; break;
+		case 1: data &= ~0x80000000; break;
+		case 0: data &= ~0x00000400; break;
 		}
 
-		switch (nv_ro08(bios, ramcfg + 0x03) & 0x30) {
-		case 0x30: data &= ~0x00000020; break;
-		case 0x20: data &= ~0x00000080; break;
-		case 0x10: data &= ~0x00080000; break;
-		case 0x00: data &= ~0x00000200; break;
+		switch (next->bios.ramcfg_11_03_30) {
+		case 3: data &= ~0x00000020; break;
+		case 2: data &= ~0x00000080; break;
+		case 1: data &= ~0x00080000; break;
+		case 0: data &= ~0x00000200; break;
 		}
 	}
 
-	if (nv_ro08(bios, ramcfg + 0x02) & 0x80)
+	if (next->bios.ramcfg_11_02_80)
 		mask |= 0x03000000;
-	if (nv_ro08(bios, ramcfg + 0x02) & 0x40)
+	if (next->bios.ramcfg_11_02_40)
 		mask |= 0x00002000;
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x10)
+	if (next->bios.ramcfg_11_07_10)
 		mask |= 0x00004000;
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x08)
+	if (next->bios.ramcfg_11_07_08)
 		mask |= 0x00000003;
 	else {
 		mask |= 0x34000000;
@@ -314,18 +349,18 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	if (ram->from == 2 && ram->mode != 2) {
 		ram_mask(fuc, 0x10f808, 0x00080000, 0x00000000);
-		ram_mask(fuc, 0x10f200, 0x00008000, 0x00008000);
+		ram_mask(fuc, 0x10f200, 0x18008000, 0x00008000);
 		ram_mask(fuc, 0x10f800, 0x00000000, 0x00000004);
 		ram_mask(fuc, 0x10f830, 0x00008000, 0x01040010);
 		ram_mask(fuc, 0x10f830, 0x01000000, 0x00000000);
 		r1373f4_init(fuc);
 		ram_mask(fuc, 0x1373f0, 0x00000002, 0x00000001);
-		r1373f4_fini(fuc, ramcfg);
+		r1373f4_fini(fuc);
 		ram_mask(fuc, 0x10f830, 0x00c00000, 0x00240001);
 	} else
 	if (ram->from != 2 && ram->mode != 2) {
 		r1373f4_init(fuc);
-		r1373f4_fini(fuc, ramcfg);
+		r1373f4_fini(fuc);
 	}
 
 	if (ram_have(fuc, gpioMV)) {
@@ -336,49 +371,54 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 		}
 	}
 
-	if ( (nv_ro08(bios, ramcfg + 0x02) & 0x40) ||
-	     (nv_ro08(bios, ramcfg + 0x07) & 0x10)) {
+	if (next->bios.ramcfg_11_02_40 ||
+	    next->bios.ramcfg_11_07_10) {
 		ram_mask(fuc, 0x132040, 0x00010000, 0x00010000);
 		ram_nsec(fuc, 20000);
 	}
 
 	if (ram->from != 2 && ram->mode == 2) {
+		if (0 /*XXX: Titan */)
+			ram_mask(fuc, 0x10f200, 0x18000000, 0x18000000);
 		ram_mask(fuc, 0x10f800, 0x00000004, 0x00000000);
 		ram_mask(fuc, 0x1373f0, 0x00000000, 0x00000002);
 		ram_mask(fuc, 0x10f830, 0x00800001, 0x00408010);
 		r1373f4_init(fuc);
-		r1373f4_fini(fuc, ramcfg);
+		r1373f4_fini(fuc);
 		ram_mask(fuc, 0x10f808, 0x00000000, 0x00080000);
 		ram_mask(fuc, 0x10f200, 0x00808000, 0x00800000);
 	} else
 	if (ram->from == 2 && ram->mode == 2) {
 		ram_mask(fuc, 0x10f800, 0x00000004, 0x00000000);
 		r1373f4_init(fuc);
-		r1373f4_fini(fuc, ramcfg);
+		r1373f4_fini(fuc);
 	}
 
 	if (ram->mode != 2) /*XXX*/ {
-		if (nv_ro08(bios, ramcfg + 0x07) & 0x40)
+		if (next->bios.ramcfg_11_07_40)
 			ram_mask(fuc, 0x10f670, 0x80000000, 0x80000000);
 	}
 
-	data = (nv_ro08(bios, rammap + 0x11) & 0x0c) >> 2;
-	ram_wr32(fuc, 0x10f65c, 0x00000011 * data);
-	ram_wr32(fuc, 0x10f6b8, 0x01010101 * nv_ro08(bios, ramcfg + 0x09));
-	ram_wr32(fuc, 0x10f6bc, 0x01010101 * nv_ro08(bios, ramcfg + 0x09));
+	ram_wr32(fuc, 0x10f65c, 0x00000011 * next->bios.rammap_11_11_0c);
+	ram_wr32(fuc, 0x10f6b8, 0x01010101 * next->bios.ramcfg_11_09);
+	ram_wr32(fuc, 0x10f6bc, 0x01010101 * next->bios.ramcfg_11_09);
 
-	data = nv_ro08(bios, ramcfg + 0x04);
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08)) {
-		ram_wr32(fuc, 0x10f698, 0x01010101 * data);
-		ram_wr32(fuc, 0x10f69c, 0x01010101 * data);
+	if (!next->bios.ramcfg_11_07_08 && !next->bios.ramcfg_11_07_04) {
+		ram_wr32(fuc, 0x10f698, 0x01010101 * next->bios.ramcfg_11_04);
+		ram_wr32(fuc, 0x10f69c, 0x01010101 * next->bios.ramcfg_11_04);
+	} else
+	if (!next->bios.ramcfg_11_07_08) {
+		ram_wr32(fuc, 0x10f698, 0x00000000);
+		ram_wr32(fuc, 0x10f69c, 0x00000000);
 	}
 
 	if (ram->mode != 2) {
-		u32 temp = ram_rd32(fuc, 0x10f694) & ~0xff00ff00;
-		ram_wr32(fuc, 0x10f694, temp | (0x01000100 * data));
+		u32 data = 0x01000100 * next->bios.ramcfg_11_04;
+		ram_nuke(fuc, 0x10f694);
+		ram_mask(fuc, 0x10f694, 0xff00ff00, data);
 	}
 
-	if (ram->mode == 2 && (nv_ro08(bios, ramcfg + 0x08) & 0x10))
+	if (ram->mode == 2 && next->bios.ramcfg_11_08_10)
 		data = 0x00000080;
 	else
 		data = 0x00000000;
@@ -386,19 +426,19 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	mask = 0x00070000;
 	data = 0x00000000;
-	if (!(nv_ro08(bios, ramcfg + 0x02) & 0x80))
+	if (!next->bios.ramcfg_11_02_80)
 		data |= 0x03000000;
-	if (!(nv_ro08(bios, ramcfg + 0x02) & 0x40))
+	if (!next->bios.ramcfg_11_02_40)
 		data |= 0x00002000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x10))
+	if (!next->bios.ramcfg_11_07_10)
 		data |= 0x00004000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08))
+	if (!next->bios.ramcfg_11_07_08)
 		data |= 0x00000003;
 	else
 		data |= 0x74000000;
 	ram_mask(fuc, 0x10f824, mask, data);
 
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x08)
+	if (next->bios.ramcfg_11_01_08)
 		data = 0x00000000;
 	else
 		data = 0x00001000;
@@ -409,61 +449,90 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 		ram_mask(fuc, 0x10f670, 0x80000000, 0x00000000);
 	}
 
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x01)
+	if (next->bios.ramcfg_11_08_01)
 		data = 0x00100000;
 	else
 		data = 0x00000000;
 	ram_mask(fuc, 0x10f82c, 0x00100000, data);
 
 	data = 0x00000000;
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x08)
+	if (next->bios.ramcfg_11_08_08)
 		data |= 0x00002000;
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x04)
+	if (next->bios.ramcfg_11_08_04)
 		data |= 0x00001000;
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x02)
+	if (next->bios.ramcfg_11_08_02)
 		data |= 0x00004000;
 	ram_mask(fuc, 0x10f830, 0x00007000, data);
 
 	/* PFB timing */
-	ram_mask(fuc, 0x10f248, 0xffffffff, nv_ro32(bios, timing + 0x28));
-	ram_mask(fuc, 0x10f290, 0xffffffff, nv_ro32(bios, timing + 0x00));
-	ram_mask(fuc, 0x10f294, 0xffffffff, nv_ro32(bios, timing + 0x04));
-	ram_mask(fuc, 0x10f298, 0xffffffff, nv_ro32(bios, timing + 0x08));
-	ram_mask(fuc, 0x10f29c, 0xffffffff, nv_ro32(bios, timing + 0x0c));
-	ram_mask(fuc, 0x10f2a0, 0xffffffff, nv_ro32(bios, timing + 0x10));
-	ram_mask(fuc, 0x10f2a4, 0xffffffff, nv_ro32(bios, timing + 0x14));
-	ram_mask(fuc, 0x10f2a8, 0xffffffff, nv_ro32(bios, timing + 0x18));
-	ram_mask(fuc, 0x10f2ac, 0xffffffff, nv_ro32(bios, timing + 0x1c));
-	ram_mask(fuc, 0x10f2cc, 0xffffffff, nv_ro32(bios, timing + 0x20));
-	ram_mask(fuc, 0x10f2e8, 0xffffffff, nv_ro32(bios, timing + 0x24));
-
-	data = (nv_ro08(bios, ramcfg + 0x02) & 0x03) << 8;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x10)
-		data |= 0x70000000;
-	ram_mask(fuc, 0x10f604, 0x70000300, data);
-
-	data = (nv_ro08(bios, timing + 0x30) & 0x07) << 28;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x01)
-		data |= 0x00000100;
-	ram_mask(fuc, 0x10f614, 0x70000000, data);
-
-	data = (nv_ro08(bios, timing + 0x30) & 0x07) << 28;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x02)
-		data |= 0x00000100;
-	ram_mask(fuc, 0x10f610, 0x70000000, data);
+	ram_mask(fuc, 0x10f248, 0xffffffff, next->bios.timing[10]);
+	ram_mask(fuc, 0x10f290, 0xffffffff, next->bios.timing[0]);
+	ram_mask(fuc, 0x10f294, 0xffffffff, next->bios.timing[1]);
+	ram_mask(fuc, 0x10f298, 0xffffffff, next->bios.timing[2]);
+	ram_mask(fuc, 0x10f29c, 0xffffffff, next->bios.timing[3]);
+	ram_mask(fuc, 0x10f2a0, 0xffffffff, next->bios.timing[4]);
+	ram_mask(fuc, 0x10f2a4, 0xffffffff, next->bios.timing[5]);
+	ram_mask(fuc, 0x10f2a8, 0xffffffff, next->bios.timing[6]);
+	ram_mask(fuc, 0x10f2ac, 0xffffffff, next->bios.timing[7]);
+	ram_mask(fuc, 0x10f2cc, 0xffffffff, next->bios.timing[8]);
+	ram_mask(fuc, 0x10f2e8, 0xffffffff, next->bios.timing[9]);
+
+	data = mask = 0x00000000;
+	if (NOTE00(ramcfg_08_20)) {
+		if (next->bios.ramcfg_11_08_20)
+			data |= 0x01000000;
+		mask |= 0x01000000;
+	}
+	ram_mask(fuc, 0x10f200, mask, data);
+
+	data = mask = 0x00000000;
+	if (NOTE00(ramcfg_02_03 != 0)) {
+		data |= next->bios.ramcfg_11_02_03 << 8;
+		mask |= 0x00000300;
+	}
+	if (NOTE00(ramcfg_01_10)) {
+		if (next->bios.ramcfg_11_01_10)
+			data |= 0x70000000;
+		mask |= 0x70000000;
+	}
+	ram_mask(fuc, 0x10f604, mask, data);
+
+	data = mask = 0x00000000;
+	if (NOTE00(timing_30_07 != 0)) {
+		data |= next->bios.timing_20_30_07 << 28;
+		mask |= 0x70000000;
+	}
+	if (NOTE00(ramcfg_01_01)) {
+		if (next->bios.ramcfg_11_01_01)
+			data |= 0x00000100;
+		mask |= 0x00000100;
+	}
+	ram_mask(fuc, 0x10f614, mask, data);
+
+	data = mask = 0x00000000;
+	if (NOTE00(timing_30_07 != 0)) {
+		data |= next->bios.timing_20_30_07 << 28;
+		mask |= 0x70000000;
+	}
+	if (NOTE00(ramcfg_01_02)) {
+		if (next->bios.ramcfg_11_01_02)
+			data |= 0x00000100;
+		mask |= 0x00000100;
+	}
+	ram_mask(fuc, 0x10f610, mask, data);
 
 	mask = 0x33f00000;
 	data = 0x00000000;
-	if (!(nv_ro08(bios, ramcfg + 0x01) & 0x04))
+	if (!next->bios.ramcfg_11_01_04)
 		data |= 0x20200000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x80))
+	if (!next->bios.ramcfg_11_07_80)
 		data |= 0x12800000;
 	/*XXX: see note above about there probably being some condition
 	 *     for the 10f824 stuff that uses ramcfg 3...
 	 */
-	if ( (nv_ro08(bios, ramcfg + 0x03) & 0xf0)) {
-		if (nv_ro08(bios, rammap + 0x08) & 0x0c) {
-			if (!(nv_ro08(bios, ramcfg + 0x07) & 0x80))
+	if (next->bios.ramcfg_11_03_f0) {
+		if (next->bios.rammap_11_08_0c) {
+			if (!next->bios.ramcfg_11_07_80)
 				mask |= 0x00000020;
 			else
 				data |= 0x00000020;
@@ -476,49 +545,53 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	ram_mask(fuc, 0x10f808, mask, data);
 
-	data = nv_ro08(bios, ramcfg + 0x03) & 0x0f;
-	ram_wr32(fuc, 0x10f870, 0x11111111 * data);
+	ram_wr32(fuc, 0x10f870, 0x11111111 * next->bios.ramcfg_11_03_0f);
 
-	data = nv_ro08(bios, ramcfg + 0x02) & 0x03;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x10)
-		data |= 0x00000004;
-	if ((nv_rd32(bios, 0x100770) & 0x00000004) != (data & 0x00000004)) {
-		ram_wr32(fuc, 0x10f750, 0x04000009);
+	data = mask = 0x00000000;
+	if (NOTE00(ramcfg_02_03 != 0)) {
+		data |= next->bios.ramcfg_11_02_03;
+		mask |= 0x00000003;
+	}
+	if (NOTE00(ramcfg_01_10)) {
+		if (next->bios.ramcfg_11_01_10)
+			data |= 0x00000004;
+		mask |= 0x00000004;
+	}
+
+	if ((ram_mask(fuc, 0x100770, mask, data) & mask & 4) != (data & 4)) {
+		ram_mask(fuc, 0x100750, 0x00000008, 0x00000008);
 		ram_wr32(fuc, 0x100710, 0x00000000);
 		ram_wait(fuc, 0x100710, 0x80000000, 0x80000000, 200000);
 	}
-	ram_mask(fuc, 0x100770, 0x00000007, data);
 
-	data = (nv_ro08(bios, timing + 0x30) & 0x07) << 8;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x01)
+	data = next->bios.timing_20_30_07 << 8;
+	if (next->bios.ramcfg_11_01_01)
 		data |= 0x80000000;
 	ram_mask(fuc, 0x100778, 0x00000700, data);
 
-	data = nv_ro16(bios, timing + 0x2c);
-	ram_mask(fuc, 0x10f250, 0x000003f0, (data & 0x003f) <<  4);
-	ram_mask(fuc, 0x10f24c, 0x7f000000, (data & 0x1fc0) << 18);
-
-	data = nv_ro08(bios, timing + 0x30);
-	ram_mask(fuc, 0x10f224, 0x001f0000, (data & 0xf8) << 13);
+	ram_mask(fuc, 0x10f250, 0x000003f0, next->bios.timing_20_2c_003f << 4);
+	data = (next->bios.timing[10] & 0x7f000000) >> 24;
+	if (data < next->bios.timing_20_2c_1fc0)
+		data = next->bios.timing_20_2c_1fc0;
+	ram_mask(fuc, 0x10f24c, 0x7f000000, data << 24);
+	ram_mask(fuc, 0x10f224, 0x001f0000, next->bios.timing_20_30_f8 << 16);
 
-	data = nv_ro16(bios, timing + 0x31);
-	ram_mask(fuc, 0x10fec4, 0x041e0f07, (data & 0x0800) << 15 |
-					    (data & 0x0780) << 10 |
-					    (data & 0x0078) <<  5 |
-					    (data & 0x0007));
-	ram_mask(fuc, 0x10fec8, 0x00000027, (data & 0x8000) >> 10 |
-					    (data & 0x7000) >> 12);
+	ram_mask(fuc, 0x10fec4, 0x041e0f07, next->bios.timing_20_31_0800 << 26 |
+					    next->bios.timing_20_31_0780 << 17 |
+					    next->bios.timing_20_31_0078 << 8 |
+					    next->bios.timing_20_31_0007);
+	ram_mask(fuc, 0x10fec8, 0x00000027, next->bios.timing_20_31_8000 << 5 |
+					    next->bios.timing_20_31_7000);
 
 	ram_wr32(fuc, 0x10f090, 0x4000007e);
-	ram_nsec(fuc, 1000);
+	ram_nsec(fuc, 2000);
 	ram_wr32(fuc, 0x10f314, 0x00000001); /* PRECHARGE */
 	ram_wr32(fuc, 0x10f310, 0x00000001); /* REFRESH */
-	ram_nsec(fuc, 2000);
 	ram_wr32(fuc, 0x10f210, 0x80000000); /* REFRESH_AUTO = 1 */
 
-	if ((nv_ro08(bios, ramcfg + 0x08) & 0x10) && (ram->mode == 2) /*XXX*/) {
+	if (next->bios.ramcfg_11_08_10 && (ram->mode == 2) /*XXX*/) {
 		u32 temp = ram_mask(fuc, 0x10f294, 0xff000000, 0x24000000);
-		train(fuc, 0xa4010000); /*XXX*/
+		nve0_ram_train(fuc, 0xbc0e0000, 0xa4010000); /*XXX*/
 		ram_nsec(fuc, 1000);
 		ram_wr32(fuc, 0x10f294, temp);
 	}
@@ -528,7 +601,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, mr[8], 0xfff, ram->base.mr[8]);
 	ram_nsec(fuc, 1000);
 	ram_mask(fuc, mr[1], 0xfff, ram->base.mr[1]);
-	ram_mask(fuc, mr[5], 0xfff, ram->base.mr[5]);
+	ram_mask(fuc, mr[5], 0xfff, ram->base.mr[5] & ~0x004); /* LP3 later */
 	ram_mask(fuc, mr[6], 0xfff, ram->base.mr[6]);
 	ram_mask(fuc, mr[7], 0xfff, ram->base.mr[7]);
 
@@ -544,12 +617,13 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_wr32(fuc, 0x10f318, 0x00000001); /* NOP? */
 	ram_mask(fuc, 0x10f200, 0x80000000, 0x00000000);
 	ram_nsec(fuc, 1000);
+	ram_nuts(ram, 0x10f200, 0x18808800, 0x00000000, 0x18808800);
 
 	data  = ram_rd32(fuc, 0x10f978);
 	data &= ~0x00046144;
 	data |=  0x0000000b;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08)) {
-		if (!(nv_ro08(bios, ramcfg + 0x07) & 0x04))
+	if (!next->bios.ramcfg_11_07_08) {
+		if (!next->bios.ramcfg_11_07_04)
 			data |= 0x0000200c;
 		else
 			data |= 0x00000000;
@@ -563,44 +637,43 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 		ram_wr32(fuc, 0x10f830, data);
 	}
 
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08)) {
+	if (!next->bios.ramcfg_11_07_08) {
 		data = 0x88020000;
-		if ( (nv_ro08(bios, ramcfg + 0x07) & 0x04))
+		if ( next->bios.ramcfg_11_07_04)
 			data |= 0x10000000;
-		if (!(nv_ro08(bios, rammap + 0x08) & 0x10))
+		if (!next->bios.rammap_11_08_10)
 			data |= 0x00080000;
 	} else {
 		data = 0xa40e0000;
 	}
-	train(fuc, data);
-	ram_nsec(fuc, 1000);
+	nve0_ram_train(fuc, 0xbc0f0000, data);
+	if (1) /* XXX: not always? */
+		ram_nsec(fuc, 1000);
 
 	if (ram->mode == 2) { /*XXX*/
 		ram_mask(fuc, 0x10f800, 0x00000004, 0x00000004);
 	}
 
-	/* MR5: (re)enable LP3 if necessary
-	 * XXX: need to find the switch, keeping off for now
-	 */
-	ram_mask(fuc, mr[5], 0x00000004, 0x00000000);
+	/* LP3 */
+	if (ram_mask(fuc, mr[5], 0x004, ram->base.mr[5]) != ram->base.mr[5])
+		ram_nsec(fuc, 1000);
 
 	if (ram->mode != 2) {
 		ram_mask(fuc, 0x10f830, 0x01000000, 0x01000000);
 		ram_mask(fuc, 0x10f830, 0x01000000, 0x00000000);
 	}
 
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x02) {
-		ram_mask(fuc, 0x10f910, 0x80020000, 0x01000000);
-		ram_mask(fuc, 0x10f914, 0x80020000, 0x01000000);
-	}
+	if (next->bios.ramcfg_11_07_02)
+		nve0_ram_train(fuc, 0x80020000, 0x01000000);
 
 	ram_wr32(fuc, 0x62c000, 0x0f0f0f00);
 
-	if (nv_ro08(bios, rammap + 0x08) & 0x01)
+	if (next->bios.rammap_11_08_01)
 		data = 0x00000800;
 	else
 		data = 0x00000000;
 	ram_mask(fuc, 0x10f200, 0x00000800, data);
+	ram_nuts(ram, 0x10f200, 0x18808800, data, 0x18808800);
 	return 0;
 }
 
@@ -611,17 +684,14 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 static int
 nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 {
-	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
 	struct nve0_ramfuc *fuc = &ram->fuc;
 	const u32 rcoef = ((  ram->P1 << 16) | (ram->N1 << 8) | ram->M1);
 	const u32 runk0 = ram->fN1 << 16;
 	const u32 runk1 = ram->fN1;
-	const u32 rammap = ram->base.rammap.data;
-	const u32 ramcfg = ram->base.ramcfg.data;
-	const u32 timing = ram->base.timing.data;
-	int vc = !(nv_ro08(bios, ramcfg + 0x02) & 0x08);
-	int mv = 1; /*XXX*/
+	struct nouveau_ram_data *next = ram->base.next;
+	int vc = !next->bios.ramcfg_11_02_08;
+	int mv = !next->bios.ramcfg_11_02_04;
 	u32 mask, data;
 
 	ram_mask(fuc, 0x10f808, 0x40000000, 0x40000000);
@@ -636,7 +706,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 	}
 
 	ram_mask(fuc, 0x10f200, 0x00000800, 0x00000000);
-	if ((nv_ro08(bios, ramcfg + 0x03) & 0xf0))
+	if (next->bios.ramcfg_11_03_f0)
 		ram_mask(fuc, 0x10f808, 0x04000000, 0x04000000);
 
 	ram_wr32(fuc, 0x10f314, 0x00000001); /* PRECHARGE */
@@ -661,28 +731,28 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 	if (1) {
 		mask |= 0x800807e0;
 		data |= 0x800807e0;
-		switch (nv_ro08(bios, ramcfg + 0x03) & 0xc0) {
-		case 0xc0: data &= ~0x00000040; break;
-		case 0x80: data &= ~0x00000100; break;
-		case 0x40: data &= ~0x80000000; break;
-		case 0x00: data &= ~0x00000400; break;
+		switch (next->bios.ramcfg_11_03_c0) {
+		case 3: data &= ~0x00000040; break;
+		case 2: data &= ~0x00000100; break;
+		case 1: data &= ~0x80000000; break;
+		case 0: data &= ~0x00000400; break;
 		}
 
-		switch (nv_ro08(bios, ramcfg + 0x03) & 0x30) {
-		case 0x30: data &= ~0x00000020; break;
-		case 0x20: data &= ~0x00000080; break;
-		case 0x10: data &= ~0x00080000; break;
-		case 0x00: data &= ~0x00000200; break;
+		switch (next->bios.ramcfg_11_03_30) {
+		case 3: data &= ~0x00000020; break;
+		case 2: data &= ~0x00000080; break;
+		case 1: data &= ~0x00080000; break;
+		case 0: data &= ~0x00000200; break;
 		}
 	}
 
-	if (nv_ro08(bios, ramcfg + 0x02) & 0x80)
+	if (next->bios.ramcfg_11_02_80)
 		mask |= 0x03000000;
-	if (nv_ro08(bios, ramcfg + 0x02) & 0x40)
+	if (next->bios.ramcfg_11_02_40)
 		mask |= 0x00002000;
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x10)
+	if (next->bios.ramcfg_11_07_10)
 		mask |= 0x00004000;
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x08)
+	if (next->bios.ramcfg_11_07_08)
 		mask |= 0x00000003;
 	else
 		mask |= 0x14000000;
@@ -692,7 +762,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 
 	ram_mask(fuc, 0x1373f4, 0x00000000, 0x00010010);
 	data  = ram_rd32(fuc, 0x1373ec) & ~0x00030000;
-	data |= (nv_ro08(bios, ramcfg + 0x03) & 0x30) << 12;
+	data |= next->bios.ramcfg_11_03_30 << 16;
 	ram_wr32(fuc, 0x1373ec, data);
 	ram_mask(fuc, 0x1373f4, 0x00000003, 0x00000000);
 	ram_mask(fuc, 0x1373f4, 0x00000010, 0x00000000);
@@ -724,68 +794,67 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 		}
 	}
 
-	if ( (nv_ro08(bios, ramcfg + 0x02) & 0x40) ||
-	     (nv_ro08(bios, ramcfg + 0x07) & 0x10)) {
+	if (next->bios.ramcfg_11_02_40 ||
+	    next->bios.ramcfg_11_07_10) {
 		ram_mask(fuc, 0x132040, 0x00010000, 0x00010000);
 		ram_nsec(fuc, 20000);
 	}
 
 	if (ram->mode != 2) /*XXX*/ {
-		if (nv_ro08(bios, ramcfg + 0x07) & 0x40)
+		if (next->bios.ramcfg_11_07_40)
 			ram_mask(fuc, 0x10f670, 0x80000000, 0x80000000);
 	}
 
-	data = (nv_ro08(bios, rammap + 0x11) & 0x0c) >> 2;
-	ram_wr32(fuc, 0x10f65c, 0x00000011 * data);
-	ram_wr32(fuc, 0x10f6b8, 0x01010101 * nv_ro08(bios, ramcfg + 0x09));
-	ram_wr32(fuc, 0x10f6bc, 0x01010101 * nv_ro08(bios, ramcfg + 0x09));
+	ram_wr32(fuc, 0x10f65c, 0x00000011 * next->bios.rammap_11_11_0c);
+	ram_wr32(fuc, 0x10f6b8, 0x01010101 * next->bios.ramcfg_11_09);
+	ram_wr32(fuc, 0x10f6bc, 0x01010101 * next->bios.ramcfg_11_09);
 
 	mask = 0x00010000;
 	data = 0x00000000;
-	if (!(nv_ro08(bios, ramcfg + 0x02) & 0x80))
+	if (!next->bios.ramcfg_11_02_80)
 		data |= 0x03000000;
-	if (!(nv_ro08(bios, ramcfg + 0x02) & 0x40))
+	if (!next->bios.ramcfg_11_02_40)
 		data |= 0x00002000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x10))
+	if (!next->bios.ramcfg_11_07_10)
 		data |= 0x00004000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08))
+	if (!next->bios.ramcfg_11_07_08)
 		data |= 0x00000003;
 	else
 		data |= 0x14000000;
 	ram_mask(fuc, 0x10f824, mask, data);
 	ram_nsec(fuc, 1000);
 
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x01)
+	if (next->bios.ramcfg_11_08_01)
 		data = 0x00100000;
 	else
 		data = 0x00000000;
 	ram_mask(fuc, 0x10f82c, 0x00100000, data);
 
 	/* PFB timing */
-	ram_mask(fuc, 0x10f248, 0xffffffff, nv_ro32(bios, timing + 0x28));
-	ram_mask(fuc, 0x10f290, 0xffffffff, nv_ro32(bios, timing + 0x00));
-	ram_mask(fuc, 0x10f294, 0xffffffff, nv_ro32(bios, timing + 0x04));
-	ram_mask(fuc, 0x10f298, 0xffffffff, nv_ro32(bios, timing + 0x08));
-	ram_mask(fuc, 0x10f29c, 0xffffffff, nv_ro32(bios, timing + 0x0c));
-	ram_mask(fuc, 0x10f2a0, 0xffffffff, nv_ro32(bios, timing + 0x10));
-	ram_mask(fuc, 0x10f2a4, 0xffffffff, nv_ro32(bios, timing + 0x14));
-	ram_mask(fuc, 0x10f2a8, 0xffffffff, nv_ro32(bios, timing + 0x18));
-	ram_mask(fuc, 0x10f2ac, 0xffffffff, nv_ro32(bios, timing + 0x1c));
-	ram_mask(fuc, 0x10f2cc, 0xffffffff, nv_ro32(bios, timing + 0x20));
-	ram_mask(fuc, 0x10f2e8, 0xffffffff, nv_ro32(bios, timing + 0x24));
+	ram_mask(fuc, 0x10f248, 0xffffffff, next->bios.timing[10]);
+	ram_mask(fuc, 0x10f290, 0xffffffff, next->bios.timing[0]);
+	ram_mask(fuc, 0x10f294, 0xffffffff, next->bios.timing[1]);
+	ram_mask(fuc, 0x10f298, 0xffffffff, next->bios.timing[2]);
+	ram_mask(fuc, 0x10f29c, 0xffffffff, next->bios.timing[3]);
+	ram_mask(fuc, 0x10f2a0, 0xffffffff, next->bios.timing[4]);
+	ram_mask(fuc, 0x10f2a4, 0xffffffff, next->bios.timing[5]);
+	ram_mask(fuc, 0x10f2a8, 0xffffffff, next->bios.timing[6]);
+	ram_mask(fuc, 0x10f2ac, 0xffffffff, next->bios.timing[7]);
+	ram_mask(fuc, 0x10f2cc, 0xffffffff, next->bios.timing[8]);
+	ram_mask(fuc, 0x10f2e8, 0xffffffff, next->bios.timing[9]);
 
 	mask = 0x33f00000;
 	data = 0x00000000;
-	if (!(nv_ro08(bios, ramcfg + 0x01) & 0x04))
+	if (!next->bios.ramcfg_11_01_04)
 		data |= 0x20200000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x80))
+	if (!next->bios.ramcfg_11_07_80)
 		data |= 0x12800000;
 	/*XXX: see note above about there probably being some condition
 	 *     for the 10f824 stuff that uses ramcfg 3...
 	 */
-	if ( (nv_ro08(bios, ramcfg + 0x03) & 0xf0)) {
-		if (nv_ro08(bios, rammap + 0x08) & 0x0c) {
-			if (!(nv_ro08(bios, ramcfg + 0x07) & 0x80))
+	if (next->bios.ramcfg_11_03_f0) {
+		if (next->bios.rammap_11_08_0c) {
+			if (!next->bios.ramcfg_11_07_80)
 				mask |= 0x00000020;
 			else
 				data |= 0x00000020;
@@ -799,21 +868,16 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 
 	ram_mask(fuc, 0x10f808, mask, data);
 
-	data = nv_ro08(bios, ramcfg + 0x03) & 0x0f;
-	ram_wr32(fuc, 0x10f870, 0x11111111 * data);
+	ram_wr32(fuc, 0x10f870, 0x11111111 * next->bios.ramcfg_11_03_0f);
 
-	data = nv_ro16(bios, timing + 0x2c);
-	ram_mask(fuc, 0x10f250, 0x000003f0, (data & 0x003f) <<  4);
+	ram_mask(fuc, 0x10f250, 0x000003f0, next->bios.timing_20_2c_003f << 4);
 
-	if (((nv_ro32(bios, timing + 0x2c) & 0x00001fc0) >>  6) >
-	    ((nv_ro32(bios, timing + 0x28) & 0x7f000000) >> 24))
-		data = (nv_ro32(bios, timing + 0x2c) & 0x00001fc0) >>  6;
-	else
-		data = (nv_ro32(bios, timing + 0x28) & 0x1f000000) >> 24;
+	data = (next->bios.timing[10] & 0x7f000000) >> 24;
+	if (data < next->bios.timing_20_2c_1fc0)
+		data = next->bios.timing_20_2c_1fc0;
 	ram_mask(fuc, 0x10f24c, 0x7f000000, data << 24);
 
-	data = nv_ro08(bios, timing + 0x30);
-	ram_mask(fuc, 0x10f224, 0x001f0000, (data & 0xf8) << 13);
+	ram_mask(fuc, 0x10f224, 0x001f0000, next->bios.timing_20_30_f8 << 16);
 
 	ram_wr32(fuc, 0x10f090, 0x4000007f);
 	ram_nsec(fuc, 1000);
@@ -855,7 +919,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 
 	ram_wr32(fuc, 0x62c000, 0x0f0f0f00);
 
-	if (nv_ro08(bios, rammap + 0x08) & 0x01)
+	if (next->bios.rammap_11_08_01)
 		data = 0x00000800;
 	else
 		data = 0x00000000;
@@ -868,21 +932,18 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
  ******************************************************************************/
 
 static int
-nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
+nve0_ram_calc_data(struct nouveau_fb *pfb, u32 freq,
+		   struct nouveau_ram_data *data)
 {
 	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
-	struct nve0_ramfuc *fuc = &ram->fuc;
-	struct bit_entry M;
-	int ret, refclk, strap, i;
-	u32 data;
-	u8  cnt;
+	u8 strap, cnt, len;
 
 	/* lookup memory config data relevant to the target frequency */
-	ram->base.rammap.data = nvbios_rammap_match(bios, freq / 1000,
-						   &ram->base.rammap.version,
-						   &ram->base.rammap.size, &cnt,
-						   &ram->base.ramcfg.size);
+	ram->base.rammap.data = nvbios_rammapEp(bios, freq / 1000,
+					       &ram->base.rammap.version,
+					       &ram->base.rammap.size,
+					       &cnt, &len, &data->bios);
 	if (!ram->base.rammap.data || ram->base.rammap.version != 0x11 ||
 	     ram->base.rammap.size < 0x09) {
 		nv_error(pfb, "invalid/missing rammap entry\n");
@@ -890,24 +951,14 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	}
 
 	/* locate specific data set for the attached memory */
-	if (bit_entry(bios, 'M', &M) || M.version != 2 || M.length < 3) {
-		nv_error(pfb, "invalid/missing memory table\n");
-		return -EINVAL;
-	}
-
-	strap = (nv_rd32(pfb, 0x101000) & 0x0000003c) >> 2;
-	data = nv_ro16(bios, M.offset + 1);
-	if (data)
-		strap = nv_ro08(bios, data + strap);
-
-	if (strap >= cnt) {
-		nv_error(pfb, "invalid ramcfg strap\n");
-		return -EINVAL;
-	}
-
-	ram->base.ramcfg.version = ram->base.rammap.version;
-	ram->base.ramcfg.data = ram->base.rammap.data + ram->base.rammap.size +
-			       (ram->base.ramcfg.size * strap);
+	strap = nvbios_ramcfg_index(nv_subdev(pfb));
+	ram->base.ramcfg.data = nvbios_rammapSp(bios, ram->base.rammap.data,
+						ram->base.rammap.version,
+						ram->base.rammap.size,
+						cnt, len, strap,
+						&ram->base.ramcfg.version,
+						&ram->base.ramcfg.size,
+						&data->bios);
 	if (!ram->base.ramcfg.data || ram->base.ramcfg.version != 0x11 ||
 	     ram->base.ramcfg.size < 0x08) {
 		nv_error(pfb, "invalid/missing ramcfg entry\n");
@@ -918,9 +969,9 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	strap = nv_ro08(bios, ram->base.ramcfg.data + 0x00);
 	if (strap != 0xff) {
 		ram->base.timing.data =
-			nvbios_timing_entry(bios, strap,
-					   &ram->base.timing.version,
-					   &ram->base.timing.size);
+			nvbios_timingEp(bios, strap, &ram->base.timing.version,
+				       &ram->base.timing.size, &cnt, &len,
+				       &data->bios);
 		if (!ram->base.timing.data ||
 		     ram->base.timing.version != 0x20 ||
 		     ram->base.timing.size < 0x33) {
@@ -931,11 +982,23 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		ram->base.timing.data = 0;
 	}
 
+	data->freq = freq;
+	return 0;
+}
+
+static int
+nve0_ram_calc_xits(struct nouveau_fb *pfb, struct nouveau_ram_data *next)
+{
+	struct nve0_ram *ram = (void *)pfb->ram;
+	struct nve0_ramfuc *fuc = &ram->fuc;
+	int refclk, i;
+	int ret;
+
 	ret = ram_init(fuc, pfb);
 	if (ret)
 		return ret;
 
-	ram->mode = (freq > fuc->refpll.vco1.max_freq) ? 2 : 1;
+	ram->mode = (next->freq > fuc->refpll.vco1.max_freq) ? 2 : 1;
 	ram->from = ram_rd32(fuc, 0x1373f4) & 0x0000000f;
 
 	/* XXX: this is *not* what nvidia do.  on fermi nvidia generally
@@ -946,7 +1009,7 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	 * so far, i've seen very weird values being chosen by nvidia on
 	 * kepler boards, no idea how/why they're chosen.
 	 */
-	refclk = freq;
+	refclk = next->freq;
 	if (ram->mode == 2)
 		refclk = fuc->mempll.refclk;
 
@@ -968,7 +1031,7 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		fuc->mempll.min_p = 1;
 		fuc->mempll.max_p = 2;
 
-		ret = nva3_pll_calc(nv_subdev(pfb), &fuc->mempll, freq,
+		ret = nva3_pll_calc(nv_subdev(pfb), &fuc->mempll, next->freq,
 				   &ram->N2, NULL, &ram->M2, &ram->P2);
 		if (ret <= 0) {
 			nv_error(pfb, "unable to calc mempll\n");
@@ -980,17 +1043,18 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		if (ram_have(fuc, mr[i]))
 			ram->base.mr[i] = ram_rd32(fuc, mr[i]);
 	}
+	ram->base.freq = next->freq;
 
 	switch (ram->base.type) {
 	case NV_MEM_TYPE_DDR3:
 		ret = nouveau_sddr3_calc(&ram->base);
 		if (ret == 0)
-			ret = nve0_ram_calc_sddr3(pfb, freq);
+			ret = nve0_ram_calc_sddr3(pfb, next->freq);
 		break;
 	case NV_MEM_TYPE_GDDR5:
-		ret = nouveau_gddr5_calc(&ram->base);
+		ret = nouveau_gddr5_calc(&ram->base, ram->pnuts != 0);
 		if (ret == 0)
-			ret = nve0_ram_calc_gddr5(pfb, freq);
+			ret = nve0_ram_calc_gddr5(pfb, next->freq);
 		break;
 	default:
 		ret = -ENOSYS;
@@ -1001,13 +1065,55 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 }
 
 static int
+nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
+{
+	struct nouveau_clock *clk = nouveau_clock(pfb);
+	struct nve0_ram *ram = (void *)pfb->ram;
+	struct nouveau_ram_data *xits = &ram->base.xition;
+	struct nouveau_ram_data *copy;
+	int ret;
+
+	if (ram->base.next == NULL) {
+		ret = nve0_ram_calc_data(pfb, clk->read(clk, nv_clk_src_mem),
+					&ram->base.former);
+		if (ret)
+			return ret;
+
+		ret = nve0_ram_calc_data(pfb, freq, &ram->base.target);
+		if (ret)
+			return ret;
+
+		if (ram->base.target.freq < ram->base.former.freq) {
+			*xits = ram->base.target;
+			copy = &ram->base.former;
+		} else {
+			*xits = ram->base.former;
+			copy = &ram->base.target;
+		}
+
+		xits->bios.ramcfg_11_02_04 = copy->bios.ramcfg_11_02_04;
+		xits->bios.ramcfg_11_02_03 = copy->bios.ramcfg_11_02_03;
+		xits->bios.timing_20_30_07 = copy->bios.timing_20_30_07;
+
+		ram->base.next = &ram->base.target;
+		if (memcmp(xits, &ram->base.former, sizeof(xits->bios)))
+			ram->base.next = &ram->base.xition;
+	} else {
+		BUG_ON(ram->base.next != &ram->base.xition);
+		ram->base.next = &ram->base.target;
+	}
+
+	return nve0_ram_calc_xits(pfb, ram->base.next);
+}
+
+static int
 nve0_ram_prog(struct nouveau_fb *pfb)
 {
 	struct nouveau_device *device = nv_device(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
 	struct nve0_ramfuc *fuc = &ram->fuc;
-	ram_exec(fuc, nouveau_boolopt(device->cfgopt, "NvMemExec", false));
-	return 0;
+	ram_exec(fuc, nouveau_boolopt(device->cfgopt, "NvMemExec", true));
+	return (ram->base.next == &ram->base.xition);
 }
 
 static void
@@ -1015,10 +1121,11 @@ nve0_ram_tidy(struct nouveau_fb *pfb)
 {
 	struct nve0_ram *ram = (void *)pfb->ram;
 	struct nve0_ramfuc *fuc = &ram->fuc;
+	ram->base.next = NULL;
 	ram_exec(fuc, false);
 }
 
-static int
+int
 nve0_ram_init(struct nouveau_object *object)
 {
 	struct nouveau_fb *pfb = (void *)object->parent;
@@ -1055,7 +1162,7 @@ nve0_ram_init(struct nouveau_object *object)
 	 * binary driver skips the one that's already been setup by
 	 * the init tables.
 	 */
-	data = nvbios_rammap_table(bios, &ver, &hdr, &cnt, &len, &snr, &ssz);
+	data = nvbios_rammapTe(bios, &ver, &hdr, &cnt, &len, &snr, &ssz);
 	if (!data || hdr < 0x15)
 		return -EINVAL;
 
@@ -1073,6 +1180,7 @@ nve0_ram_init(struct nouveau_object *object)
 		data += 4;
 	}
 	nv_wr32(pfb, 0x10f65c, save);
+	nv_mask(pfb, 0x10f584, 0x11000000, 0x00000000);
 
 	switch (ram->base.type) {
 	case NV_MEM_TYPE_GDDR5:
@@ -1117,9 +1225,10 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nouveau_gpio *gpio = nouveau_gpio(pfb);
 	struct dcb_gpio_func func;
 	struct nve0_ram *ram;
-	int ret;
+	int ret, i;
+	u32 tmp;
 
-	ret = nvc0_ram_create(parent, engine, oclass, &ram);
+	ret = nvc0_ram_create(parent, engine, oclass, 0x022554, &ram);
 	*pobject = nv_object(ram);
 	if (ret)
 		return ret;
@@ -1136,6 +1245,25 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		break;
 	}
 
+	/* calculate a mask of differently configured memory partitions,
+	 * because, of course reclocking wasn't complicated enough
+	 * already without having to treat some of them differently to
+	 * the others....
+	 */
+	ram->parts = nv_rd32(pfb, 0x022438);
+	ram->pmask = nv_rd32(pfb, 0x022554);
+	ram->pnuts = 0;
+	for (i = 0, tmp = 0; i < ram->parts; i++) {
+		if (!(ram->pmask & (1 << i))) {
+			u32 cfg1 = nv_rd32(pfb, 0x110204 + (i * 0x1000));
+			if (tmp && tmp != cfg1) {
+				ram->pnuts |= (1 << i);
+				continue;
+			}
+			tmp = cfg1;
+		}
+	}
+
 	// parse bios data for both pll's
 	ret = nvbios_pll_parse(bios, 0x0c, &ram->fuc.refpll);
 	if (ret) {
@@ -1248,7 +1376,7 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	ram->fuc.r_0x10f65c = ramfuc_reg(0x10f65c);
 	ram->fuc.r_0x10f6bc = ramfuc_reg(0x10f6bc);
 	ram->fuc.r_0x100710 = ramfuc_reg(0x100710);
-	ram->fuc.r_0x10f750 = ramfuc_reg(0x10f750);
+	ram->fuc.r_0x100750 = ramfuc_reg(0x100750);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c
index f572c2804c3..45e0202f315 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/base.c
@@ -22,21 +22,24 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/gpio.h>
 #include <subdev/bios.h>
 #include <subdev/bios/gpio.h>
 
+#include "priv.h"
+
 static int
 nouveau_gpio_drive(struct nouveau_gpio *gpio,
 		   int idx, int line, int dir, int out)
 {
-	return gpio->drive ? gpio->drive(gpio, line, dir, out) : -ENODEV;
+	const struct nouveau_gpio_impl *impl = (void *)nv_object(gpio)->oclass;
+	return impl->drive ? impl->drive(gpio, line, dir, out) : -ENODEV;
 }
 
 static int
 nouveau_gpio_sense(struct nouveau_gpio *gpio, int idx, int line)
 {
-	return gpio->sense ? gpio->sense(gpio, line) : -ENODEV;
+	const struct nouveau_gpio_impl *impl = (void *)nv_object(gpio)->oclass;
+	return impl->sense ? impl->sense(gpio, line) : -ENODEV;
 }
 
 static int
@@ -102,6 +105,80 @@ nouveau_gpio_get(struct nouveau_gpio *gpio, int idx, u8 tag, u8 line)
 	return ret;
 }
 
+static void
+nouveau_gpio_intr_disable(struct nouveau_event *event, int type, int index)
+{
+	struct nouveau_gpio *gpio = nouveau_gpio(event->priv);
+	const struct nouveau_gpio_impl *impl = (void *)nv_object(gpio)->oclass;
+	impl->intr_mask(gpio, type, 1 << index, 0);
+}
+
+static void
+nouveau_gpio_intr_enable(struct nouveau_event *event, int type, int index)
+{
+	struct nouveau_gpio *gpio = nouveau_gpio(event->priv);
+	const struct nouveau_gpio_impl *impl = (void *)nv_object(gpio)->oclass;
+	impl->intr_mask(gpio, type, 1 << index, 1 << index);
+}
+
+static void
+nouveau_gpio_intr(struct nouveau_subdev *subdev)
+{
+	struct nouveau_gpio *gpio = nouveau_gpio(subdev);
+	const struct nouveau_gpio_impl *impl = (void *)nv_object(gpio)->oclass;
+	u32 hi, lo, e, i;
+
+	impl->intr_stat(gpio, &hi, &lo);
+
+	for (i = 0; e = 0, (hi | lo) && i < impl->lines; i++) {
+		if (hi & (1 << i))
+			e |= NVKM_GPIO_HI;
+		if (lo & (1 << i))
+			e |= NVKM_GPIO_LO;
+		nouveau_event_trigger(gpio->events, e, i);
+	}
+}
+
+int
+_nouveau_gpio_fini(struct nouveau_object *object, bool suspend)
+{
+	const struct nouveau_gpio_impl *impl = (void *)object->oclass;
+	struct nouveau_gpio *gpio = nouveau_gpio(object);
+	u32 mask = (1 << impl->lines) - 1;
+
+	impl->intr_mask(gpio, NVKM_GPIO_TOGGLED, mask, 0);
+	impl->intr_stat(gpio, &mask, &mask);
+
+	return nouveau_subdev_fini(&gpio->base, suspend);
+}
+
+static struct dmi_system_id gpio_reset_ids[] = {
+	{
+		.ident = "Apple Macbook 10,1",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "MacBookPro10,1"),
+		}
+	},
+	{ }
+};
+
+int
+_nouveau_gpio_init(struct nouveau_object *object)
+{
+	struct nouveau_gpio *gpio = nouveau_gpio(object);
+	int ret;
+
+	ret = nouveau_subdev_init(&gpio->base);
+	if (ret)
+		return ret;
+
+	if (gpio->reset && dmi_check_system(gpio_reset_ids))
+		gpio->reset(gpio, DCB_GPIO_UNUSED);
+
+	return ret;
+}
+
 void
 _nouveau_gpio_dtor(struct nouveau_object *object)
 {
@@ -113,9 +190,10 @@ _nouveau_gpio_dtor(struct nouveau_object *object)
 int
 nouveau_gpio_create_(struct nouveau_object *parent,
 		     struct nouveau_object *engine,
-		     struct nouveau_oclass *oclass, int lines,
+		     struct nouveau_oclass *oclass,
 		     int length, void **pobject)
 {
+	const struct nouveau_gpio_impl *impl = (void *)oclass;
 	struct nouveau_gpio *gpio;
 	int ret;
 
@@ -125,34 +203,34 @@ nouveau_gpio_create_(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	ret = nouveau_event_create(lines, &gpio->events);
-	if (ret)
-		return ret;
-
 	gpio->find = nouveau_gpio_find;
 	gpio->set  = nouveau_gpio_set;
 	gpio->get  = nouveau_gpio_get;
+	gpio->reset = impl->reset;
+
+	ret = nouveau_event_create(2, impl->lines, &gpio->events);
+	if (ret)
+		return ret;
+
+	gpio->events->priv = gpio;
+	gpio->events->enable = nouveau_gpio_intr_enable;
+	gpio->events->disable = nouveau_gpio_intr_disable;
+	nv_subdev(gpio)->intr = nouveau_gpio_intr;
 	return 0;
 }
 
-static struct dmi_system_id gpio_reset_ids[] = {
-	{
-		.ident = "Apple Macbook 10,1",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "MacBookPro10,1"),
-		}
-	},
-	{ }
-};
-
 int
-nouveau_gpio_init(struct nouveau_gpio *gpio)
+_nouveau_gpio_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+		   struct nouveau_oclass *oclass, void *data, u32 size,
+		   struct nouveau_object **pobject)
 {
-	int ret = nouveau_subdev_init(&gpio->base);
-	if (ret == 0 && gpio->reset) {
-		if (dmi_check_system(gpio_reset_ids))
-			gpio->reset(gpio, DCB_GPIO_UNUSED);
-	}
-	return ret;
+	struct nouveau_gpio *gpio;
+	int ret;
+
+	ret = nouveau_gpio_create(parent, engine, oclass, &gpio);
+	*pobject = nv_object(gpio);
+	if (ret)
+		return ret;
+
+	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv10.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv10.c
index 76d5d5465dd..27ad23eaf18 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv10.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv10.c
@@ -26,10 +26,6 @@
 
 #include "priv.h"
 
-struct nv10_gpio_priv {
-	struct nouveau_gpio base;
-};
-
 static int
 nv10_gpio_sense(struct nouveau_gpio *gpio, int line)
 {
@@ -83,95 +79,38 @@ nv10_gpio_drive(struct nouveau_gpio *gpio, int line, int dir, int out)
 }
 
 static void
-nv10_gpio_intr(struct nouveau_subdev *subdev)
-{
-	struct nv10_gpio_priv *priv = (void *)subdev;
-	u32 intr = nv_rd32(priv, 0x001104);
-	u32 hi = (intr & 0x0000ffff) >> 0;
-	u32 lo = (intr & 0xffff0000) >> 16;
-	int i;
-
-	for (i = 0; (hi | lo) && i < 32; i++) {
-		if ((hi | lo) & (1 << i))
-			nouveau_event_trigger(priv->base.events, i);
-	}
-
-	nv_wr32(priv, 0x001104, intr);
-}
-
-static void
-nv10_gpio_intr_enable(struct nouveau_event *event, int line)
-{
-	nv_wr32(event->priv, 0x001104, 0x00010001 << line);
-	nv_mask(event->priv, 0x001144, 0x00010001 << line, 0x00010001 << line);
-}
-
-static void
-nv10_gpio_intr_disable(struct nouveau_event *event, int line)
-{
-	nv_wr32(event->priv, 0x001104, 0x00010001 << line);
-	nv_mask(event->priv, 0x001144, 0x00010001 << line, 0x00000000);
-}
-
-static int
-nv10_gpio_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	       struct nouveau_oclass *oclass, void *data, u32 size,
-	       struct nouveau_object **pobject)
+nv10_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
 {
-	struct nv10_gpio_priv *priv;
-	int ret;
-
-	ret = nouveau_gpio_create(parent, engine, oclass, 16, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.drive = nv10_gpio_drive;
-	priv->base.sense = nv10_gpio_sense;
-	priv->base.events->priv = priv;
-	priv->base.events->enable = nv10_gpio_intr_enable;
-	priv->base.events->disable = nv10_gpio_intr_disable;
-	nv_subdev(priv)->intr = nv10_gpio_intr;
-	return 0;
+	u32 intr = nv_rd32(gpio, 0x001104);
+	u32 stat = nv_rd32(gpio, 0x001144) & intr;
+	*lo = (stat & 0xffff0000) >> 16;
+	*hi = (stat & 0x0000ffff);
+	nv_wr32(gpio, 0x001104, intr);
 }
 
 static void
-nv10_gpio_dtor(struct nouveau_object *object)
-{
-	struct nv10_gpio_priv *priv = (void *)object;
-	nouveau_gpio_destroy(&priv->base);
-}
-
-static int
-nv10_gpio_init(struct nouveau_object *object)
-{
-	struct nv10_gpio_priv *priv = (void *)object;
-	int ret;
-
-	ret = nouveau_gpio_init(&priv->base);
-	if (ret)
-		return ret;
-
-	nv_wr32(priv, 0x001144, 0x00000000);
-	nv_wr32(priv, 0x001104, 0xffffffff);
-	return 0;
-}
-
-static int
-nv10_gpio_fini(struct nouveau_object *object, bool suspend)
+nv10_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
 {
-	struct nv10_gpio_priv *priv = (void *)object;
-	nv_wr32(priv, 0x001144, 0x00000000);
-	return nouveau_gpio_fini(&priv->base, suspend);
+	u32 inte = nv_rd32(gpio, 0x001144);
+	if (type & NVKM_GPIO_LO)
+		inte = (inte & ~(mask << 16)) | (data << 16);
+	if (type & NVKM_GPIO_HI)
+		inte = (inte & ~mask) | data;
+	nv_wr32(gpio, 0x001144, inte);
 }
 
-struct nouveau_oclass
-nv10_gpio_oclass = {
-	.handle = NV_SUBDEV(GPIO, 0x10),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv10_gpio_ctor,
-		.dtor = nv10_gpio_dtor,
-		.init = nv10_gpio_init,
-		.fini = nv10_gpio_fini,
+struct nouveau_oclass *
+nv10_gpio_oclass = &(struct nouveau_gpio_impl) {
+	.base.handle = NV_SUBDEV(GPIO, 0x10),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_gpio_ctor,
+		.dtor = _nouveau_gpio_dtor,
+		.init = _nouveau_gpio_init,
+		.fini = _nouveau_gpio_fini,
 	},
-};
+	.lines = 16,
+	.intr_stat = nv10_gpio_intr_stat,
+	.intr_mask = nv10_gpio_intr_mask,
+	.drive = nv10_gpio_drive,
+	.sense = nv10_gpio_sense,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv50.c
index c4c1d415e7f..1864fa98e6b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv50.c
@@ -24,15 +24,10 @@
 
 #include "priv.h"
 
-struct nv50_gpio_priv {
-	struct nouveau_gpio base;
-};
-
-static void
+void
 nv50_gpio_reset(struct nouveau_gpio *gpio, u8 match)
 {
 	struct nouveau_bios *bios = nouveau_bios(gpio);
-	struct nv50_gpio_priv *priv = (void *)gpio;
 	u8 ver, len;
 	u16 entry;
 	int ent = -1;
@@ -46,7 +41,8 @@ nv50_gpio_reset(struct nouveau_gpio *gpio, u8 match)
 		u8  unk0 = !!(data & 0x02000000);
 		u8  unk1 = !!(data & 0x04000000);
 		u32 val = (unk1 << 16) | unk0;
-		u32 reg = regs[line >> 4]; line &= 0x0f;
+		u32 reg = regs[line >> 4];
+		u32 lsh = line & 0x0f;
 
 		if ( func  == DCB_GPIO_UNUSED ||
 		    (match != DCB_GPIO_UNUSED && match != func))
@@ -54,7 +50,7 @@ nv50_gpio_reset(struct nouveau_gpio *gpio, u8 match)
 
 		gpio->set(gpio, 0, func, line, defs);
 
-		nv_mask(priv, reg, 0x00010001 << line, val << line);
+		nv_mask(gpio, reg, 0x00010001 << lsh, val << lsh);
 	}
 }
 
@@ -71,7 +67,7 @@ nv50_gpio_location(int line, u32 *reg, u32 *shift)
 	return 0;
 }
 
-static int
+int
 nv50_gpio_drive(struct nouveau_gpio *gpio, int line, int dir, int out)
 {
 	u32 reg, shift;
@@ -79,11 +75,11 @@ nv50_gpio_drive(struct nouveau_gpio *gpio, int line, int dir, int out)
 	if (nv50_gpio_location(line, &reg, &shift))
 		return -EINVAL;
 
-	nv_mask(gpio, reg, 7 << shift, (((dir ^ 1) << 1) | out) << shift);
+	nv_mask(gpio, reg, 3 << shift, (((dir ^ 1) << 1) | out) << shift);
 	return 0;
 }
 
-static int
+int
 nv50_gpio_sense(struct nouveau_gpio *gpio, int line)
 {
 	u32 reg, shift;
@@ -94,119 +90,40 @@ nv50_gpio_sense(struct nouveau_gpio *gpio, int line)
 	return !!(nv_rd32(gpio, reg) & (4 << shift));
 }
 
-void
-nv50_gpio_intr(struct nouveau_subdev *subdev)
-{
-	struct nv50_gpio_priv *priv = (void *)subdev;
-	u32 intr0, intr1 = 0;
-	u32 hi, lo;
-	int i;
-
-	intr0 = nv_rd32(priv, 0xe054) & nv_rd32(priv, 0xe050);
-	if (nv_device(priv)->chipset > 0x92)
-		intr1 = nv_rd32(priv, 0xe074) & nv_rd32(priv, 0xe070);
-
-	hi = (intr0 & 0x0000ffff) | (intr1 << 16);
-	lo = (intr0 >> 16) | (intr1 & 0xffff0000);
-
-	for (i = 0; (hi | lo) && i < 32; i++) {
-		if ((hi | lo) & (1 << i))
-			nouveau_event_trigger(priv->base.events, i);
-	}
-
-	nv_wr32(priv, 0xe054, intr0);
-	if (nv_device(priv)->chipset > 0x92)
-		nv_wr32(priv, 0xe074, intr1);
-}
-
-void
-nv50_gpio_intr_enable(struct nouveau_event *event, int line)
-{
-	const u32 addr = line < 16 ? 0xe050 : 0xe070;
-	const u32 mask = 0x00010001 << (line & 0xf);
-	nv_wr32(event->priv, addr + 0x04, mask);
-	nv_mask(event->priv, addr + 0x00, mask, mask);
-}
-
-void
-nv50_gpio_intr_disable(struct nouveau_event *event, int line)
-{
-	const u32 addr = line < 16 ? 0xe050 : 0xe070;
-	const u32 mask = 0x00010001 << (line & 0xf);
-	nv_wr32(event->priv, addr + 0x04, mask);
-	nv_mask(event->priv, addr + 0x00, mask, 0x00000000);
-}
-
-static int
-nv50_gpio_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	       struct nouveau_oclass *oclass, void *data, u32 size,
-	       struct nouveau_object **pobject)
-{
-	struct nv50_gpio_priv *priv;
-	int ret;
-
-	ret = nouveau_gpio_create(parent, engine, oclass,
-				  nv_device(parent)->chipset > 0x92 ? 32 : 16,
-				  &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.reset = nv50_gpio_reset;
-	priv->base.drive = nv50_gpio_drive;
-	priv->base.sense = nv50_gpio_sense;
-	priv->base.events->priv = priv;
-	priv->base.events->enable = nv50_gpio_intr_enable;
-	priv->base.events->disable = nv50_gpio_intr_disable;
-	nv_subdev(priv)->intr = nv50_gpio_intr;
-	return 0;
-}
-
-void
-nv50_gpio_dtor(struct nouveau_object *object)
-{
-	struct nv50_gpio_priv *priv = (void *)object;
-	nouveau_gpio_destroy(&priv->base);
-}
-
-int
-nv50_gpio_init(struct nouveau_object *object)
+static void
+nv50_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
 {
-	struct nv50_gpio_priv *priv = (void *)object;
-	int ret;
-
-	ret = nouveau_gpio_init(&priv->base);
-	if (ret)
-		return ret;
-
-	/* disable, and ack any pending gpio interrupts */
-	nv_wr32(priv, 0xe050, 0x00000000);
-	nv_wr32(priv, 0xe054, 0xffffffff);
-	if (nv_device(priv)->chipset > 0x92) {
-		nv_wr32(priv, 0xe070, 0x00000000);
-		nv_wr32(priv, 0xe074, 0xffffffff);
-	}
-
-	return 0;
+	u32 intr = nv_rd32(gpio, 0x00e054);
+	u32 stat = nv_rd32(gpio, 0x00e050) & intr;
+	*lo = (stat & 0xffff0000) >> 16;
+	*hi = (stat & 0x0000ffff);
+	nv_wr32(gpio, 0x00e054, intr);
 }
 
-int
-nv50_gpio_fini(struct nouveau_object *object, bool suspend)
+static void
+nv50_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
 {
-	struct nv50_gpio_priv *priv = (void *)object;
-	nv_wr32(priv, 0xe050, 0x00000000);
-	if (nv_device(priv)->chipset > 0x92)
-		nv_wr32(priv, 0xe070, 0x00000000);
-	return nouveau_gpio_fini(&priv->base, suspend);
+	u32 inte = nv_rd32(gpio, 0x00e050);
+	if (type & NVKM_GPIO_LO)
+		inte = (inte & ~(mask << 16)) | (data << 16);
+	if (type & NVKM_GPIO_HI)
+		inte = (inte & ~mask) | data;
+	nv_wr32(gpio, 0x00e050, inte);
 }
 
-struct nouveau_oclass
-nv50_gpio_oclass = {
-	.handle = NV_SUBDEV(GPIO, 0x50),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv50_gpio_ctor,
-		.dtor = nv50_gpio_dtor,
-		.init = nv50_gpio_init,
-		.fini = nv50_gpio_fini,
+struct nouveau_oclass *
+nv50_gpio_oclass = &(struct nouveau_gpio_impl) {
+	.base.handle = NV_SUBDEV(GPIO, 0x50),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_gpio_ctor,
+		.dtor = _nouveau_gpio_dtor,
+		.init = _nouveau_gpio_init,
+		.fini = _nouveau_gpio_fini,
 	},
-};
+	.lines = 16,
+	.intr_stat = nv50_gpio_intr_stat,
+	.intr_mask = nv50_gpio_intr_mask,
+	.drive = nv50_gpio_drive,
+	.sense = nv50_gpio_sense,
+	.reset = nv50_gpio_reset,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nv92.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv92.c
new file mode 100644
index 00000000000..252083d376f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nv92.c
@@ -0,0 +1,74 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "priv.h"
+
+void
+nv92_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
+{
+	u32 intr0 = nv_rd32(gpio, 0x00e054);
+	u32 intr1 = nv_rd32(gpio, 0x00e074);
+	u32 stat0 = nv_rd32(gpio, 0x00e050) & intr0;
+	u32 stat1 = nv_rd32(gpio, 0x00e070) & intr1;
+	*lo = (stat1 & 0xffff0000) | (stat0 >> 16);
+	*hi = (stat1 << 16) | (stat0 & 0x0000ffff);
+	nv_wr32(gpio, 0x00e054, intr0);
+	nv_wr32(gpio, 0x00e074, intr1);
+}
+
+void
+nv92_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
+{
+	u32 inte0 = nv_rd32(gpio, 0x00e050);
+	u32 inte1 = nv_rd32(gpio, 0x00e070);
+	if (type & NVKM_GPIO_LO)
+		inte0 = (inte0 & ~(mask << 16)) | (data << 16);
+	if (type & NVKM_GPIO_HI)
+		inte0 = (inte0 & ~(mask & 0xffff)) | (data & 0xffff);
+	mask >>= 16;
+	data >>= 16;
+	if (type & NVKM_GPIO_LO)
+		inte1 = (inte1 & ~(mask << 16)) | (data << 16);
+	if (type & NVKM_GPIO_HI)
+		inte1 = (inte1 & ~mask) | data;
+	nv_wr32(gpio, 0x00e050, inte0);
+	nv_wr32(gpio, 0x00e070, inte1);
+}
+
+struct nouveau_oclass *
+nv92_gpio_oclass = &(struct nouveau_gpio_impl) {
+	.base.handle = NV_SUBDEV(GPIO, 0x92),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_gpio_ctor,
+		.dtor = _nouveau_gpio_dtor,
+		.init = _nouveau_gpio_init,
+		.fini = _nouveau_gpio_fini,
+	},
+	.lines = 32,
+	.intr_stat = nv92_gpio_intr_stat,
+	.intr_mask = nv92_gpio_intr_mask,
+	.drive = nv50_gpio_drive,
+	.sense = nv50_gpio_sense,
+	.reset = nv50_gpio_reset,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c
index 010431e3ace..a4682b0956a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nvd0.c
@@ -24,15 +24,10 @@
 
 #include "priv.h"
 
-struct nvd0_gpio_priv {
-	struct nouveau_gpio base;
-};
-
 void
 nvd0_gpio_reset(struct nouveau_gpio *gpio, u8 match)
 {
 	struct nouveau_bios *bios = nouveau_bios(gpio);
-	struct nvd0_gpio_priv *priv = (void *)gpio;
 	u8 ver, len;
 	u16 entry;
 	int ent = -1;
@@ -51,9 +46,9 @@ nvd0_gpio_reset(struct nouveau_gpio *gpio, u8 match)
 
 		gpio->set(gpio, 0, func, line, defs);
 
-		nv_mask(priv, 0x00d610 + (line * 4), 0xff, unk0);
+		nv_mask(gpio, 0x00d610 + (line * 4), 0xff, unk0);
 		if (unk1--)
-			nv_mask(priv, 0x00d740 + (unk1 * 4), 0xff, line);
+			nv_mask(gpio, 0x00d740 + (unk1 * 4), 0xff, line);
 	}
 }
 
@@ -72,36 +67,19 @@ nvd0_gpio_sense(struct nouveau_gpio *gpio, int line)
 	return !!(nv_rd32(gpio, 0x00d610 + (line * 4)) & 0x00004000);
 }
 
-static int
-nvd0_gpio_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	       struct nouveau_oclass *oclass, void *data, u32 size,
-	       struct nouveau_object **pobject)
-{
-	struct nvd0_gpio_priv *priv;
-	int ret;
-
-	ret = nouveau_gpio_create(parent, engine, oclass, 32, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.reset = nvd0_gpio_reset;
-	priv->base.drive = nvd0_gpio_drive;
-	priv->base.sense = nvd0_gpio_sense;
-	priv->base.events->priv = priv;
-	priv->base.events->enable = nv50_gpio_intr_enable;
-	priv->base.events->disable = nv50_gpio_intr_disable;
-	nv_subdev(priv)->intr = nv50_gpio_intr;
-	return 0;
-}
-
-struct nouveau_oclass
-nvd0_gpio_oclass = {
-	.handle = NV_SUBDEV(GPIO, 0xd0),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nvd0_gpio_ctor,
-		.dtor = nv50_gpio_dtor,
-		.init = nv50_gpio_init,
-		.fini = nv50_gpio_fini,
+struct nouveau_oclass *
+nvd0_gpio_oclass = &(struct nouveau_gpio_impl) {
+	.base.handle = NV_SUBDEV(GPIO, 0xd0),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_gpio_ctor,
+		.dtor = _nouveau_gpio_dtor,
+		.init = _nouveau_gpio_init,
+		.fini = _nouveau_gpio_fini,
 	},
-};
+	.lines = 32,
+	.intr_stat = nv92_gpio_intr_stat,
+	.intr_mask = nv92_gpio_intr_mask,
+	.drive = nvd0_gpio_drive,
+	.sense = nvd0_gpio_sense,
+	.reset = nvd0_gpio_reset,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/nve0.c b/drivers/gpu/drm/nouveau/core/subdev/gpio/nve0.c
index 16b8c5bf5ef..e1145b48c76 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/nve0.c
@@ -24,108 +24,51 @@
 
 #include "priv.h"
 
-struct nve0_gpio_priv {
-	struct nouveau_gpio base;
-};
-
-void
-nve0_gpio_intr(struct nouveau_subdev *subdev)
+static void
+nve0_gpio_intr_stat(struct nouveau_gpio *gpio, u32 *hi, u32 *lo)
 {
-	struct nve0_gpio_priv *priv = (void *)subdev;
-	u32 intr0 = nv_rd32(priv, 0xdc00) & nv_rd32(priv, 0xdc08);
-	u32 intr1 = nv_rd32(priv, 0xdc80) & nv_rd32(priv, 0xdc88);
-	u32 hi = (intr0 & 0x0000ffff) | (intr1 << 16);
-	u32 lo = (intr0 >> 16) | (intr1 & 0xffff0000);
-	int i;
-
-	for (i = 0; (hi | lo) && i < 32; i++) {
-		if ((hi | lo) & (1 << i))
-			nouveau_event_trigger(priv->base.events, i);
-	}
-
-	nv_wr32(priv, 0xdc00, intr0);
-	nv_wr32(priv, 0xdc88, intr1);
+	u32 intr0 = nv_rd32(gpio, 0x00dc00);
+	u32 intr1 = nv_rd32(gpio, 0x00dc80);
+	u32 stat0 = nv_rd32(gpio, 0x00dc08) & intr0;
+	u32 stat1 = nv_rd32(gpio, 0x00dc88) & intr1;
+	*lo = (stat1 & 0xffff0000) | (stat0 >> 16);
+	*hi = (stat1 << 16) | (stat0 & 0x0000ffff);
+	nv_wr32(gpio, 0x00dc00, intr0);
+	nv_wr32(gpio, 0x00dc80, intr1);
 }
 
 void
-nve0_gpio_intr_enable(struct nouveau_event *event, int line)
+nve0_gpio_intr_mask(struct nouveau_gpio *gpio, u32 type, u32 mask, u32 data)
 {
-	const u32 addr = line < 16 ? 0xdc00 : 0xdc80;
-	const u32 mask = 0x00010001 << (line & 0xf);
-	nv_wr32(event->priv, addr + 0x08, mask);
-	nv_mask(event->priv, addr + 0x00, mask, mask);
-}
-
-void
-nve0_gpio_intr_disable(struct nouveau_event *event, int line)
-{
-	const u32 addr = line < 16 ? 0xdc00 : 0xdc80;
-	const u32 mask = 0x00010001 << (line & 0xf);
-	nv_wr32(event->priv, addr + 0x08, mask);
-	nv_mask(event->priv, addr + 0x00, mask, 0x00000000);
-}
-
-int
-nve0_gpio_fini(struct nouveau_object *object, bool suspend)
-{
-	struct nve0_gpio_priv *priv = (void *)object;
-	nv_wr32(priv, 0xdc08, 0x00000000);
-	nv_wr32(priv, 0xdc88, 0x00000000);
-	return nouveau_gpio_fini(&priv->base, suspend);
-}
-
-int
-nve0_gpio_init(struct nouveau_object *object)
-{
-	struct nve0_gpio_priv *priv = (void *)object;
-	int ret;
-
-	ret = nouveau_gpio_init(&priv->base);
-	if (ret)
-		return ret;
-
-	nv_wr32(priv, 0xdc00, 0xffffffff);
-	nv_wr32(priv, 0xdc80, 0xffffffff);
-	return 0;
-}
-
-void
-nve0_gpio_dtor(struct nouveau_object *object)
-{
-	struct nve0_gpio_priv *priv = (void *)object;
-	nouveau_gpio_destroy(&priv->base);
-}
-
-static int
-nve0_gpio_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	       struct nouveau_oclass *oclass, void *data, u32 size,
-	       struct nouveau_object **pobject)
-{
-	struct nve0_gpio_priv *priv;
-	int ret;
-
-	ret = nouveau_gpio_create(parent, engine, oclass, 32, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.reset = nvd0_gpio_reset;
-	priv->base.drive = nvd0_gpio_drive;
-	priv->base.sense = nvd0_gpio_sense;
-	priv->base.events->priv = priv;
-	priv->base.events->enable = nve0_gpio_intr_enable;
-	priv->base.events->disable = nve0_gpio_intr_disable;
-	nv_subdev(priv)->intr = nve0_gpio_intr;
-	return 0;
+	u32 inte0 = nv_rd32(gpio, 0x00dc08);
+	u32 inte1 = nv_rd32(gpio, 0x00dc88);
+	if (type & NVKM_GPIO_LO)
+		inte0 = (inte0 & ~(mask << 16)) | (data << 16);
+	if (type & NVKM_GPIO_HI)
+		inte0 = (inte0 & ~(mask & 0xffff)) | (data & 0xffff);
+	mask >>= 16;
+	data >>= 16;
+	if (type & NVKM_GPIO_LO)
+		inte1 = (inte1 & ~(mask << 16)) | (data << 16);
+	if (type & NVKM_GPIO_HI)
+		inte1 = (inte1 & ~mask) | data;
+	nv_wr32(gpio, 0x00dc08, inte0);
+	nv_wr32(gpio, 0x00dc88, inte1);
 }
 
-struct nouveau_oclass
-nve0_gpio_oclass = {
-	.handle = NV_SUBDEV(GPIO, 0xe0),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nve0_gpio_ctor,
-		.dtor = nv50_gpio_dtor,
-		.init = nve0_gpio_init,
-		.fini = nve0_gpio_fini,
+struct nouveau_oclass *
+nve0_gpio_oclass = &(struct nouveau_gpio_impl) {
+	.base.handle = NV_SUBDEV(GPIO, 0xe0),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_gpio_ctor,
+		.dtor = _nouveau_gpio_dtor,
+		.init = _nouveau_gpio_init,
+		.fini = _nouveau_gpio_fini,
 	},
-};
+	.lines = 32,
+	.intr_stat = nve0_gpio_intr_stat,
+	.intr_mask = nve0_gpio_intr_mask,
+	.drive = nvd0_gpio_drive,
+	.sense = nvd0_gpio_sense,
+	.reset = nvd0_gpio_reset,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h b/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h
index 2ee1c895c78..e1724dfc86a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/gpio/priv.h
@@ -3,15 +3,65 @@
 
 #include <subdev/gpio.h>
 
-void nv50_gpio_dtor(struct nouveau_object *);
-int  nv50_gpio_init(struct nouveau_object *);
-int  nv50_gpio_fini(struct nouveau_object *, bool);
-void nv50_gpio_intr(struct nouveau_subdev *);
-void nv50_gpio_intr_enable(struct nouveau_event *, int line);
-void nv50_gpio_intr_disable(struct nouveau_event *, int line);
+#define nouveau_gpio_create(p,e,o,d)                                           \
+	nouveau_gpio_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_gpio_destroy(p) ({                                             \
+	struct nouveau_gpio *gpio = (p);                                       \
+	_nouveau_gpio_dtor(nv_object(gpio));                                   \
+})
+#define nouveau_gpio_init(p) ({                                                \
+	struct nouveau_gpio *gpio = (p);                                       \
+	_nouveau_gpio_init(nv_object(gpio));                                   \
+})
+#define nouveau_gpio_fini(p,s) ({                                              \
+	struct nouveau_gpio *gpio = (p);                                       \
+	_nouveau_gpio_fini(nv_object(gpio), (s));                              \
+})
+
+int  nouveau_gpio_create_(struct nouveau_object *, struct nouveau_object *,
+			  struct nouveau_oclass *, int, void **);
+int  _nouveau_gpio_ctor(struct nouveau_object *, struct nouveau_object *,
+			struct nouveau_oclass *, void *, u32,
+			struct nouveau_object **);
+void _nouveau_gpio_dtor(struct nouveau_object *);
+int  _nouveau_gpio_init(struct nouveau_object *);
+int  _nouveau_gpio_fini(struct nouveau_object *, bool);
+
+struct nouveau_gpio_impl {
+	struct nouveau_oclass base;
+	int lines;
+
+	/* read and ack pending interrupts, returning only data
+	 * for lines that have not been masked off, while still
+	 * performing the ack for anything that was pending.
+	 */
+	void (*intr_stat)(struct nouveau_gpio *, u32 *, u32 *);
+
+	/* mask on/off interrupts for hi/lo transitions on a
+	 * given set of gpio lines
+	 */
+	void (*intr_mask)(struct nouveau_gpio *, u32, u32, u32);
+
+	/* configure gpio direction and output value */
+	int  (*drive)(struct nouveau_gpio *, int line, int dir, int out);
+
+	/* sense current state of given gpio line */
+	int  (*sense)(struct nouveau_gpio *, int line);
+
+	/*XXX*/
+	void (*reset)(struct nouveau_gpio *, u8);
+};
+
+void nv50_gpio_reset(struct nouveau_gpio *, u8);
+int  nv50_gpio_drive(struct nouveau_gpio *, int, int, int);
+int  nv50_gpio_sense(struct nouveau_gpio *, int);
+
+void nv92_gpio_intr_stat(struct nouveau_gpio *, u32 *, u32 *);
+void nv92_gpio_intr_mask(struct nouveau_gpio *, u32, u32, u32);
 
 void nvd0_gpio_reset(struct nouveau_gpio *, u8);
 int  nvd0_gpio_drive(struct nouveau_gpio *, int, int, int);
 int  nvd0_gpio_sense(struct nouveau_gpio *, int);
 
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/anx9805.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/anx9805.c
index 4b195ac4da6..2c2731a6cf9 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/anx9805.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/anx9805.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 
-#include <subdev/i2c.h>
+#include "port.h"
 
 struct anx9805_i2c_port {
 	struct nouveau_i2c_port base;
@@ -37,6 +37,8 @@ anx9805_train(struct nouveau_i2c_port *port, int link_nr, int link_bw, bool enh)
 	struct nouveau_i2c_port *mast = (void *)nv_object(chan)->parent;
 	u8 tmp, i;
 
+	DBG("ANX9805 train %d 0x%02x %d\n", link_nr, link_bw, enh);
+
 	nv_wri2cr(mast, chan->addr, 0xa0, link_bw);
 	nv_wri2cr(mast, chan->addr, 0xa1, link_nr | (enh ? 0x80 : 0x00));
 	nv_wri2cr(mast, chan->addr, 0xa2, 0x01);
@@ -60,21 +62,29 @@ anx9805_train(struct nouveau_i2c_port *port, int link_nr, int link_bw, bool enh)
 }
 
 static int
-anx9805_aux(struct nouveau_i2c_port *port, u8 type, u32 addr, u8 *data, u8 size)
+anx9805_aux(struct nouveau_i2c_port *port, bool retry,
+	    u8 type, u32 addr, u8 *data, u8 size)
 {
 	struct anx9805_i2c_port *chan = (void *)port;
 	struct nouveau_i2c_port *mast = (void *)nv_object(chan)->parent;
 	int i, ret = -ETIMEDOUT;
+	u8 buf[16] = {};
 	u8 tmp;
 
+	DBG("%02x %05x %d\n", type, addr, size);
+
 	tmp = nv_rdi2cr(mast, chan->ctrl, 0x07) & ~0x04;
 	nv_wri2cr(mast, chan->ctrl, 0x07, tmp | 0x04);
 	nv_wri2cr(mast, chan->ctrl, 0x07, tmp);
 	nv_wri2cr(mast, chan->ctrl, 0xf7, 0x01);
 
 	nv_wri2cr(mast, chan->addr, 0xe4, 0x80);
-	for (i = 0; !(type & 1) && i < size; i++)
-		nv_wri2cr(mast, chan->addr, 0xf0 + i, data[i]);
+	if (!(type & 1)) {
+		memcpy(buf, data, size);
+		DBG("%16ph", buf);
+		for (i = 0; i < size; i++)
+			nv_wri2cr(mast, chan->addr, 0xf0 + i, buf[i]);
+	}
 	nv_wri2cr(mast, chan->addr, 0xe5, ((size - 1) << 4) | type);
 	nv_wri2cr(mast, chan->addr, 0xe6, (addr & 0x000ff) >>  0);
 	nv_wri2cr(mast, chan->addr, 0xe7, (addr & 0x0ff00) >>  8);
@@ -93,8 +103,13 @@ anx9805_aux(struct nouveau_i2c_port *port, u8 type, u32 addr, u8 *data, u8 size)
 		goto done;
 	}
 
-	for (i = 0; (type & 1) && i < size; i++)
-		data[i] = nv_rdi2cr(mast, chan->addr, 0xf0 + i);
+	if (type & 1) {
+		for (i = 0; i < size; i++)
+			buf[i] = nv_rdi2cr(mast, chan->addr, 0xf0 + i);
+		DBG("%16ph", buf);
+		memcpy(data, buf, size);
+	}
+
 	ret = 0;
 done:
 	nv_wri2cr(mast, chan->ctrl, 0xf7, 0x01);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/aux.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/aux.c
index 5de074ad170..02eb42be2e9 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/aux.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/aux.c
@@ -22,15 +22,19 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/i2c.h>
+#include "priv.h"
 
 int
 nv_rdaux(struct nouveau_i2c_port *port, u32 addr, u8 *data, u8 size)
 {
+	struct nouveau_i2c *i2c = nouveau_i2c(port);
 	if (port->func->aux) {
-		if (port->func->acquire)
-			port->func->acquire(port);
-		return port->func->aux(port, 9, addr, data, size);
+		int ret = i2c->acquire(port, 0);
+		if (ret == 0) {
+			ret = port->func->aux(port, true, 9, addr, data, size);
+			i2c->release(port);
+		}
+		return ret;
 	}
 	return -ENODEV;
 }
@@ -38,10 +42,14 @@ nv_rdaux(struct nouveau_i2c_port *port, u32 addr, u8 *data, u8 size)
 int
 nv_wraux(struct nouveau_i2c_port *port, u32 addr, u8 *data, u8 size)
 {
+	struct nouveau_i2c *i2c = nouveau_i2c(port);
 	if (port->func->aux) {
-		if (port->func->acquire)
-			port->func->acquire(port);
-		return port->func->aux(port, 8, addr, data, size);
+		int ret = i2c->acquire(port, 0);
+		if (ret == 0) {
+			ret = port->func->aux(port, true, 8, addr, data, size);
+			i2c->release(port);
+		}
+		return ret;
 	}
 	return -ENODEV;
 }
@@ -50,13 +58,16 @@ static int
 aux_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs, int num)
 {
 	struct nouveau_i2c_port *port = adap->algo_data;
+	struct nouveau_i2c *i2c = nouveau_i2c(port);
 	struct i2c_msg *msg = msgs;
 	int ret, mcnt = num;
 
 	if (!port->func->aux)
 		return -ENODEV;
-	if ( port->func->acquire)
-		port->func->acquire(port);
+
+	ret = i2c->acquire(port, 0);
+	if (ret)
+		return ret;
 
 	while (mcnt--) {
 		u8 remaining = msg->len;
@@ -74,9 +85,11 @@ aux_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs, int num)
 			if (mcnt || remaining > 16)
 				cmd |= 4; /* MOT */
 
-			ret = port->func->aux(port, cmd, msg->addr, ptr, cnt);
-			if (ret < 0)
+			ret = port->func->aux(port, true, cmd, msg->addr, ptr, cnt);
+			if (ret < 0) {
+				i2c->release(port);
 				return ret;
+			}
 
 			ptr += cnt;
 			remaining -= cnt;
@@ -85,6 +98,7 @@ aux_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs, int num)
 		msg++;
 	}
 
+	i2c->release(port);
 	return num;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
index c33c03d2f4a..09ba2cc851c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/base.c
@@ -23,13 +23,16 @@
  */
 
 #include <core/option.h>
+#include <core/event.h>
 
 #include <subdev/bios.h>
 #include <subdev/bios/dcb.h>
 #include <subdev/bios/i2c.h>
-#include <subdev/i2c.h>
 #include <subdev/vga.h>
 
+#include "priv.h"
+#include "pad.h"
+
 /******************************************************************************
  * interface to linux i2c bit-banging algorithm
  *****************************************************************************/
@@ -45,9 +48,15 @@ nouveau_i2c_pre_xfer(struct i2c_adapter *adap)
 {
 	struct i2c_algo_bit_data *bit = adap->algo_data;
 	struct nouveau_i2c_port *port = bit->data;
-	if (port->func->acquire)
-		port->func->acquire(port);
-	return 0;
+	return nouveau_i2c(port)->acquire(port, bit->timeout);
+}
+
+static void
+nouveau_i2c_post_xfer(struct i2c_adapter *adap)
+{
+	struct i2c_algo_bit_data *bit = adap->algo_data;
+	struct nouveau_i2c_port *port = bit->data;
+	return nouveau_i2c(port)->release(port);
 }
 
 static void
@@ -82,6 +91,15 @@ nouveau_i2c_getsda(void *data)
  * base i2c "port" class implementation
  *****************************************************************************/
 
+int
+_nouveau_i2c_port_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nouveau_i2c_port *port = (void *)object;
+	struct nvkm_i2c_pad *pad = nvkm_i2c_pad(port);
+	nv_ofuncs(pad)->fini(nv_object(pad), suspend);
+	return nouveau_object_fini(&port->base, suspend);
+}
+
 void
 _nouveau_i2c_port_dtor(struct nouveau_object *object)
 {
@@ -98,7 +116,7 @@ nouveau_i2c_port_create_(struct nouveau_object *parent,
 			 const struct nouveau_i2c_func *func,
 			 int size, void **pobject)
 {
-	struct nouveau_device *device = nv_device(parent);
+	struct nouveau_device *device = nv_device(engine);
 	struct nouveau_i2c *i2c = (void *)engine;
 	struct nouveau_i2c_port *port;
 	int ret;
@@ -111,10 +129,11 @@ nouveau_i2c_port_create_(struct nouveau_object *parent,
 	snprintf(port->adapter.name, sizeof(port->adapter.name),
 		 "nouveau-%s-%d", device->name, index);
 	port->adapter.owner = THIS_MODULE;
-	port->adapter.dev.parent = &device->pdev->dev;
+	port->adapter.dev.parent = nv_device_base(device);
 	port->index = index;
+	port->aux = -1;
 	port->func = func;
-	i2c_set_adapdata(&port->adapter, i2c);
+	mutex_init(&port->mutex);
 
 	if ( algo == &nouveau_i2c_bit_algo &&
 	    !nouveau_boolopt(device->cfgopt, "NvI2C", CSTMSEL)) {
@@ -128,6 +147,7 @@ nouveau_i2c_port_create_(struct nouveau_object *parent,
 		bit->timeout = usecs_to_jiffies(2200);
 		bit->data = port;
 		bit->pre_xfer = nouveau_i2c_pre_xfer;
+		bit->post_xfer = nouveau_i2c_post_xfer;
 		bit->setsda = nouveau_i2c_setsda;
 		bit->setscl = nouveau_i2c_setscl;
 		bit->getsda = nouveau_i2c_getsda;
@@ -141,7 +161,6 @@ nouveau_i2c_port_create_(struct nouveau_object *parent,
 		ret = i2c_add_adapter(&port->adapter);
 	}
 
-	/* drop port's i2c subdev refcount, i2c handles this itself */
 	if (ret == 0)
 		list_add_tail(&port->head, &i2c->ports);
 	return ret;
@@ -193,6 +212,75 @@ nouveau_i2c_find_type(struct nouveau_i2c *i2c, u16 type)
 	return NULL;
 }
 
+static void
+nouveau_i2c_release_pad(struct nouveau_i2c_port *port)
+{
+	struct nvkm_i2c_pad *pad = nvkm_i2c_pad(port);
+	struct nouveau_i2c *i2c = nouveau_i2c(port);
+
+	if (atomic_dec_and_test(&nv_object(pad)->usecount)) {
+		nv_ofuncs(pad)->fini(nv_object(pad), false);
+		wake_up_all(&i2c->wait);
+	}
+}
+
+static int
+nouveau_i2c_try_acquire_pad(struct nouveau_i2c_port *port)
+{
+	struct nvkm_i2c_pad *pad = nvkm_i2c_pad(port);
+
+	if (atomic_add_return(1, &nv_object(pad)->usecount) != 1) {
+		struct nouveau_object *owner = (void *)pad->port;
+		do {
+			if (owner == (void *)port)
+				return 0;
+			owner = owner->parent;
+		} while(owner);
+		nouveau_i2c_release_pad(port);
+		return -EBUSY;
+	}
+
+	pad->next = port;
+	nv_ofuncs(pad)->init(nv_object(pad));
+	return 0;
+}
+
+static int
+nouveau_i2c_acquire_pad(struct nouveau_i2c_port *port, unsigned long timeout)
+{
+	struct nouveau_i2c *i2c = nouveau_i2c(port);
+
+	if (timeout) {
+		if (wait_event_timeout(i2c->wait,
+				       nouveau_i2c_try_acquire_pad(port) == 0,
+				       timeout) == 0)
+			return -EBUSY;
+	} else {
+		wait_event(i2c->wait, nouveau_i2c_try_acquire_pad(port) == 0);
+	}
+
+	return 0;
+}
+
+static void
+nouveau_i2c_release(struct nouveau_i2c_port *port)
+__releases(pad->mutex)
+{
+	nouveau_i2c(port)->release_pad(port);
+	mutex_unlock(&port->mutex);
+}
+
+static int
+nouveau_i2c_acquire(struct nouveau_i2c_port *port, unsigned long timeout)
+__acquires(pad->mutex)
+{
+	int ret;
+	mutex_lock(&port->mutex);
+	if ((ret = nouveau_i2c(port)->acquire_pad(port, timeout)))
+		mutex_unlock(&port->mutex);
+	return ret;
+}
+
 static int
 nouveau_i2c_identify(struct nouveau_i2c *i2c, int index, const char *what,
 		     struct nouveau_i2c_board_info *info,
@@ -237,11 +325,59 @@ nouveau_i2c_identify(struct nouveau_i2c *i2c, int index, const char *what,
 	return -ENODEV;
 }
 
+static void
+nouveau_i2c_intr_disable(struct nouveau_event *event, int type, int index)
+{
+	struct nouveau_i2c *i2c = nouveau_i2c(event->priv);
+	struct nouveau_i2c_port *port = i2c->find(i2c, index);
+	const struct nouveau_i2c_impl *impl = (void *)nv_object(i2c)->oclass;
+	if (port && port->aux >= 0)
+		impl->aux_mask(i2c, type, 1 << port->aux, 0);
+}
+
+static void
+nouveau_i2c_intr_enable(struct nouveau_event *event, int type, int index)
+{
+	struct nouveau_i2c *i2c = nouveau_i2c(event->priv);
+	struct nouveau_i2c_port *port = i2c->find(i2c, index);
+	const struct nouveau_i2c_impl *impl = (void *)nv_object(i2c)->oclass;
+	if (port && port->aux >= 0)
+		impl->aux_mask(i2c, type, 1 << port->aux, 1 << port->aux);
+}
+
+static void
+nouveau_i2c_intr(struct nouveau_subdev *subdev)
+{
+	struct nouveau_i2c_impl *impl = (void *)nv_oclass(subdev);
+	struct nouveau_i2c *i2c = nouveau_i2c(subdev);
+	struct nouveau_i2c_port *port;
+	u32 hi, lo, rq, tx, e;
+
+	if (impl->aux_stat) {
+		impl->aux_stat(i2c, &hi, &lo, &rq, &tx);
+		if (hi || lo || rq || tx) {
+			list_for_each_entry(port, &i2c->ports, head) {
+				if (e = 0, port->aux < 0)
+					continue;
+
+				if (hi & (1 << port->aux)) e |= NVKM_I2C_PLUG;
+				if (lo & (1 << port->aux)) e |= NVKM_I2C_UNPLUG;
+				if (rq & (1 << port->aux)) e |= NVKM_I2C_IRQ;
+				if (tx & (1 << port->aux)) e |= NVKM_I2C_DONE;
+
+				nouveau_event_trigger(i2c->ntfy, e, port->index);
+			}
+		}
+	}
+}
+
 int
 _nouveau_i2c_fini(struct nouveau_object *object, bool suspend)
 {
+	struct nouveau_i2c_impl *impl = (void *)nv_oclass(object);
 	struct nouveau_i2c *i2c = (void *)object;
 	struct nouveau_i2c_port *port;
+	u32 mask;
 	int ret;
 
 	list_for_each_entry(port, &i2c->ports, head) {
@@ -250,6 +386,11 @@ _nouveau_i2c_fini(struct nouveau_object *object, bool suspend)
 			goto fail;
 	}
 
+	if ((mask = (1 << impl->aux) - 1), impl->aux_stat) {
+		impl->aux_mask(i2c, NVKM_I2C_ANY, mask, 0);
+		impl->aux_stat(i2c, &mask, &mask, &mask, &mask);
+	}
+
 	return nouveau_subdev_fini(&i2c->base, suspend);
 fail:
 	list_for_each_entry_continue_reverse(port, &i2c->ports, head) {
@@ -290,6 +431,8 @@ _nouveau_i2c_dtor(struct nouveau_object *object)
 	struct nouveau_i2c *i2c = (void *)object;
 	struct nouveau_i2c_port *port, *temp;
 
+	nouveau_event_destroy(&i2c->ntfy);
+
 	list_for_each_entry_safe(port, temp, &i2c->ports, head) {
 		nouveau_object_ref(NULL, (struct nouveau_object **)&port);
 	}
@@ -306,14 +449,14 @@ int
 nouveau_i2c_create_(struct nouveau_object *parent,
 		    struct nouveau_object *engine,
 		    struct nouveau_oclass *oclass,
-		    struct nouveau_oclass *sclass,
 		    int length, void **pobject)
 {
+	const struct nouveau_i2c_impl *impl = (void *)oclass;
 	struct nouveau_bios *bios = nouveau_bios(parent);
 	struct nouveau_i2c *i2c;
 	struct nouveau_object *object;
 	struct dcb_i2c_entry info;
-	int ret, i, j, index = -1;
+	int ret, i, j, index = -1, pad;
 	struct dcb_output outp;
 	u8  ver, hdr;
 	u32 data;
@@ -324,24 +467,48 @@ nouveau_i2c_create_(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
+	nv_subdev(i2c)->intr = nouveau_i2c_intr;
 	i2c->find = nouveau_i2c_find;
 	i2c->find_type = nouveau_i2c_find_type;
+	i2c->acquire_pad = nouveau_i2c_acquire_pad;
+	i2c->release_pad = nouveau_i2c_release_pad;
+	i2c->acquire = nouveau_i2c_acquire;
+	i2c->release = nouveau_i2c_release;
 	i2c->identify = nouveau_i2c_identify;
+	init_waitqueue_head(&i2c->wait);
 	INIT_LIST_HEAD(&i2c->ports);
 
 	while (!dcb_i2c_parse(bios, ++index, &info)) {
 		if (info.type == DCB_I2C_UNUSED)
 			continue;
 
-		oclass = sclass;
+		if (info.share != DCB_I2C_UNUSED) {
+			if (info.type == DCB_I2C_NVIO_AUX)
+				pad = info.drive;
+			else
+				pad = info.share;
+			oclass = impl->pad_s;
+		} else {
+			pad = 0x100 + info.drive;
+			oclass = impl->pad_x;
+		}
+
+		ret = nouveau_object_ctor(NULL, *pobject, oclass,
+					  NULL, pad, &parent);
+		if (ret < 0)
+			continue;
+
+		oclass = impl->sclass;
 		do {
 			ret = -EINVAL;
 			if (oclass->handle == info.type) {
-				ret = nouveau_object_ctor(*pobject, *pobject,
+				ret = nouveau_object_ctor(parent, *pobject,
 							  oclass, &info,
 							  index, &object);
 			}
 		} while (ret && (++oclass)->handle);
+
+		nouveau_object_ref(NULL, &parent);
 	}
 
 	/* in addition to the busses specified in the i2c table, there
@@ -380,5 +547,28 @@ nouveau_i2c_create_(struct nouveau_object *parent,
 		}
 	}
 
+	ret = nouveau_event_create(4, index, &i2c->ntfy);
+	if (ret)
+		return ret;
+
+	i2c->ntfy->priv = i2c;
+	i2c->ntfy->enable = nouveau_i2c_intr_enable;
+	i2c->ntfy->disable = nouveau_i2c_intr_disable;
+	return 0;
+}
+
+int
+_nouveau_i2c_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+		  struct nouveau_oclass *oclass, void *data, u32 size,
+		  struct nouveau_object **pobject)
+{
+	struct nouveau_i2c *i2c;
+	int ret;
+
+	ret = nouveau_i2c_create(parent, engine, oclass, &i2c);
+	*pobject = nv_object(i2c);
+	if (ret)
+		return ret;
+
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/bit.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/bit.c
index a6e72d3b06b..813ffc96e86 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/bit.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/bit.c
@@ -22,7 +22,7 @@
  * Authors: Ben Skeggs
  */
 
-#include "subdev/i2c.h"
+#include "priv.h"
 
 #ifdef CONFIG_NOUVEAU_I2C_INTERNAL
 #define T_TIMEOUT  2200000
@@ -187,8 +187,9 @@ i2c_bit_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs, int num)
 	struct i2c_msg *msg = msgs;
 	int ret = 0, mcnt = num;
 
-	if (port->func->acquire)
-		port->func->acquire(port);
+	ret = nouveau_i2c(port)->acquire(port, nsecs_to_jiffies(T_TIMEOUT));
+	if (ret)
+		return ret;
 
 	while (!ret && mcnt--) {
 		u8 remaining = msg->len;
@@ -210,6 +211,7 @@ i2c_bit_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs, int num)
 	}
 
 	i2c_stop(port);
+	nouveau_i2c(port)->release(port);
 	return (ret < 0) ? ret : num;
 }
 #else
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/gf117.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/gf117.c
new file mode 100644
index 00000000000..fa891c39866
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/gf117.c
@@ -0,0 +1,39 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nv50.h"
+
+struct nouveau_oclass *
+gf117_i2c_oclass = &(struct nouveau_i2c_impl) {
+	.base.handle = NV_SUBDEV(I2C, 0xd7),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_i2c_ctor,
+		.dtor = _nouveau_i2c_dtor,
+		.init = _nouveau_i2c_init,
+		.fini = _nouveau_i2c_fini,
+	},
+	.sclass = nvd0_i2c_sclass,
+	.pad_x = &nv04_i2c_pad_oclass,
+	.pad_s = &nv04_i2c_pad_oclass,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv04.c
index 860d5d2365d..b1725bdea96 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv04.c
@@ -22,9 +22,10 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/i2c.h>
 #include <subdev/vga.h>
 
+#include "priv.h"
+
 struct nv04_i2c_priv {
 	struct nouveau_i2c base;
 };
@@ -115,29 +116,15 @@ nv04_i2c_sclass[] = {
 	{}
 };
 
-static int
-nv04_i2c_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	      struct nouveau_oclass *oclass, void *data, u32 size,
-	      struct nouveau_object **pobject)
-{
-	struct nv04_i2c_priv *priv;
-	int ret;
-
-	ret = nouveau_i2c_create(parent, engine, oclass, nv04_i2c_sclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	return 0;
-}
-
-struct nouveau_oclass
-nv04_i2c_oclass = {
-	.handle = NV_SUBDEV(I2C, 0x04),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv04_i2c_ctor,
+struct nouveau_oclass *
+nv04_i2c_oclass = &(struct nouveau_i2c_impl) {
+	.base.handle = NV_SUBDEV(I2C, 0x04),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_i2c_ctor,
 		.dtor = _nouveau_i2c_dtor,
 		.init = _nouveau_i2c_init,
 		.fini = _nouveau_i2c_fini,
 	},
-};
+	.sclass = nv04_i2c_sclass,
+	.pad_x = &nv04_i2c_pad_oclass,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv4e.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv4e.c
index 0c2655a03bb..f16c87ce5ba 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv4e.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv4e.c
@@ -22,9 +22,10 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/i2c.h>
 #include <subdev/vga.h>
 
+#include "priv.h"
+
 struct nv4e_i2c_priv {
 	struct nouveau_i2c base;
 };
@@ -107,29 +108,15 @@ nv4e_i2c_sclass[] = {
 	{}
 };
 
-static int
-nv4e_i2c_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	      struct nouveau_oclass *oclass, void *data, u32 size,
-	      struct nouveau_object **pobject)
-{
-	struct nv4e_i2c_priv *priv;
-	int ret;
-
-	ret = nouveau_i2c_create(parent, engine, oclass, nv4e_i2c_sclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	return 0;
-}
-
-struct nouveau_oclass
-nv4e_i2c_oclass = {
-	.handle = NV_SUBDEV(I2C, 0x4e),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv4e_i2c_ctor,
+struct nouveau_oclass *
+nv4e_i2c_oclass = &(struct nouveau_i2c_impl) {
+	.base.handle = NV_SUBDEV(I2C, 0x4e),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_i2c_ctor,
 		.dtor = _nouveau_i2c_dtor,
 		.init = _nouveau_i2c_init,
 		.fini = _nouveau_i2c_fini,
 	},
-};
+	.sclass = nv4e_i2c_sclass,
+	.pad_x = &nv04_i2c_pad_oclass,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.c
index a8d67a28770..7b8756d4df0 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.c
@@ -121,29 +121,15 @@ nv50_i2c_sclass[] = {
 	{}
 };
 
-static int
-nv50_i2c_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	      struct nouveau_oclass *oclass, void *data, u32 size,
-	      struct nouveau_object **pobject)
-{
-	struct nv50_i2c_priv *priv;
-	int ret;
-
-	ret = nouveau_i2c_create(parent, engine, oclass, nv50_i2c_sclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	return 0;
-}
-
-struct nouveau_oclass
-nv50_i2c_oclass = {
-	.handle = NV_SUBDEV(I2C, 0x50),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv50_i2c_ctor,
+struct nouveau_oclass *
+nv50_i2c_oclass = &(struct nouveau_i2c_impl) {
+	.base.handle = NV_SUBDEV(I2C, 0x50),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_i2c_ctor,
 		.dtor = _nouveau_i2c_dtor,
 		.init = _nouveau_i2c_init,
 		.fini = _nouveau_i2c_fini,
 	},
-};
+	.sclass = nv50_i2c_sclass,
+	.pad_x = &nv04_i2c_pad_oclass,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.h b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.h
index 4e5ba48ebf5..5d2a77421c7 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv50.h
@@ -1,7 +1,7 @@
 #ifndef __NV50_I2C_H__
 #define __NV50_I2C_H__
 
-#include <subdev/i2c.h>
+#include "priv.h"
 
 struct nv50_i2c_priv {
 	struct nouveau_i2c base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv94.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv94.c
index df6d3e4b68b..f59c3a25546 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nv94.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nv94.c
@@ -24,6 +24,36 @@
 
 #include "nv50.h"
 
+void
+nv94_aux_stat(struct nouveau_i2c *i2c, u32 *hi, u32 *lo, u32 *rq, u32 *tx)
+{
+	u32 intr = nv_rd32(i2c, 0x00e06c);
+	u32 stat = nv_rd32(i2c, 0x00e068) & intr, i;
+	for (i = 0, *hi = *lo = *rq = *tx = 0; i < 8; i++) {
+		if ((stat & (1 << (i * 4)))) *hi |= 1 << i;
+		if ((stat & (2 << (i * 4)))) *lo |= 1 << i;
+		if ((stat & (4 << (i * 4)))) *rq |= 1 << i;
+		if ((stat & (8 << (i * 4)))) *tx |= 1 << i;
+	}
+	nv_wr32(i2c, 0x00e06c, intr);
+}
+
+void
+nv94_aux_mask(struct nouveau_i2c *i2c, u32 type, u32 mask, u32 data)
+{
+	u32 temp = nv_rd32(i2c, 0x00e068), i;
+	for (i = 0; i < 8; i++) {
+		if (mask & (1 << i)) {
+			if (!(data & (1 << i))) {
+				temp &= ~(type << (i * 4));
+				continue;
+			}
+			temp |= type << (i * 4);
+		}
+	}
+	nv_wr32(i2c, 0x00e068, temp);
+}
+
 #define AUX_DBG(fmt, args...) nv_debug(aux, "AUXCH(%d): " fmt, ch, ##args)
 #define AUX_ERR(fmt, args...) nv_error(aux, "AUXCH(%d): " fmt, ch, ##args)
 
@@ -69,7 +99,8 @@ auxch_init(struct nouveau_i2c *aux, int ch)
 }
 
 int
-nv94_aux(struct nouveau_i2c_port *base, u8 type, u32 addr, u8 *data, u8 size)
+nv94_aux(struct nouveau_i2c_port *base, bool retry,
+	 u8 type, u32 addr, u8 *data, u8 size)
 {
 	struct nouveau_i2c *aux = nouveau_i2c(base);
 	struct nv50_i2c_port *port = (void *)base;
@@ -105,9 +136,8 @@ nv94_aux(struct nouveau_i2c_port *base, u8 type, u32 addr, u8 *data, u8 size)
 	ctrl |= size - 1;
 	nv_wr32(aux, 0x00e4e0 + (ch * 0x50), addr);
 
-	/* retry transaction a number of times on failure... */
-	ret = -EREMOTEIO;
-	for (retries = 0; retries < 32; retries++) {
+	/* (maybe) retry transaction a number of times on failure... */
+	for (retries = 0; !ret && retries < 32; retries++) {
 		/* reset, and delay a while if this is a retry */
 		nv_wr32(aux, 0x00e4e4 + (ch * 0x50), 0x80000000 | ctrl);
 		nv_wr32(aux, 0x00e4e4 + (ch * 0x50), 0x00000000 | ctrl);
@@ -123,16 +153,21 @@ nv94_aux(struct nouveau_i2c_port *base, u8 type, u32 addr, u8 *data, u8 size)
 			udelay(1);
 			if (!timeout--) {
 				AUX_ERR("tx req timeout 0x%08x\n", ctrl);
+				ret = -EIO;
 				goto out;
 			}
 		} while (ctrl & 0x00010000);
+		ret = 1;
 
 		/* read status, and check if transaction completed ok */
 		stat = nv_mask(aux, 0x00e4e8 + (ch * 0x50), 0, 0);
-		if (!(stat & 0x000f0f00)) {
-			ret = 0;
-			break;
-		}
+		if ((stat & 0x000f0000) == 0x00080000 ||
+		    (stat & 0x000f0000) == 0x00020000)
+			ret = retry ? 0 : 1;
+		if ((stat & 0x00000100))
+			ret = -ETIMEDOUT;
+		if ((stat & 0x00000e00))
+			ret = -EIO;
 
 		AUX_DBG("%02d 0x%08x 0x%08x\n", retries, ctrl, stat);
 	}
@@ -147,29 +182,11 @@ nv94_aux(struct nouveau_i2c_port *base, u8 type, u32 addr, u8 *data, u8 size)
 
 out:
 	auxch_fini(aux, ch);
-	return ret;
-}
-
-void
-nv94_i2c_acquire(struct nouveau_i2c_port *base)
-{
-	struct nv50_i2c_priv *priv = (void *)nv_object(base)->engine;
-	struct nv50_i2c_port *port = (void *)base;
-	if (port->ctrl) {
-		nv_mask(priv, port->ctrl + 0x0c, 0x00000001, 0x00000000);
-		nv_mask(priv, port->ctrl + 0x00, 0x0000f003, port->data);
-	}
-}
-
-void
-nv94_i2c_release(struct nouveau_i2c_port *base)
-{
+	return ret < 0 ? ret : (stat & 0x000f0000) >> 16;
 }
 
 static const struct nouveau_i2c_func
 nv94_i2c_func = {
-	.acquire   = nv94_i2c_acquire,
-	.release   = nv94_i2c_release,
 	.drive_scl = nv50_i2c_drive_scl,
 	.drive_sda = nv50_i2c_drive_sda,
 	.sense_scl = nv50_i2c_sense_scl,
@@ -206,8 +223,6 @@ nv94_i2c_port_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 static const struct nouveau_i2c_func
 nv94_aux_func = {
-	.acquire   = nv94_i2c_acquire,
-	.release   = nv94_i2c_release,
 	.aux       = nv94_aux,
 };
 
@@ -227,6 +242,7 @@ nv94_aux_port_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
+	port->base.aux = info->drive;
 	port->addr = info->drive;
 	if (info->share != DCB_I2C_UNUSED) {
 		port->ctrl = 0x00e500 + (info->drive * 0x50);
@@ -257,29 +273,19 @@ nv94_i2c_sclass[] = {
 	{}
 };
 
-static int
-nv94_i2c_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	      struct nouveau_oclass *oclass, void *data, u32 size,
-	      struct nouveau_object **pobject)
-{
-	struct nv50_i2c_priv *priv;
-	int ret;
-
-	ret = nouveau_i2c_create(parent, engine, oclass, nv94_i2c_sclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	return 0;
-}
-
-struct nouveau_oclass
-nv94_i2c_oclass = {
-	.handle = NV_SUBDEV(I2C, 0x94),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv94_i2c_ctor,
+struct nouveau_oclass *
+nv94_i2c_oclass = &(struct nouveau_i2c_impl) {
+	.base.handle = NV_SUBDEV(I2C, 0x94),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_i2c_ctor,
 		.dtor = _nouveau_i2c_dtor,
 		.init = _nouveau_i2c_init,
 		.fini = _nouveau_i2c_fini,
 	},
-};
+	.sclass = nv94_i2c_sclass,
+	.pad_x = &nv04_i2c_pad_oclass,
+	.pad_s = &nv94_i2c_pad_oclass,
+	.aux = 4,
+	.aux_stat = nv94_aux_stat,
+	.aux_mask = nv94_aux_mask,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nvd0.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/nvd0.c
index 29967d30f97..364ddb1c5f0 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/i2c/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nvd0.c
@@ -42,8 +42,6 @@ nvd0_i2c_sense_sda(struct nouveau_i2c_port *base)
 
 static const struct nouveau_i2c_func
 nvd0_i2c_func = {
-	.acquire   = nv94_i2c_acquire,
-	.release   = nv94_i2c_release,
 	.drive_scl = nv50_i2c_drive_scl,
 	.drive_sda = nv50_i2c_drive_sda,
 	.sense_scl = nvd0_i2c_sense_scl,
@@ -75,7 +73,7 @@ nvd0_i2c_port_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	return 0;
 }
 
-static struct nouveau_oclass
+struct nouveau_oclass
 nvd0_i2c_sclass[] = {
 	{ .handle = NV_I2C_TYPE_DCBI2C(DCB_I2C_NVIO_BIT),
 	  .ofuncs = &(struct nouveau_ofuncs) {
@@ -96,29 +94,19 @@ nvd0_i2c_sclass[] = {
 	{}
 };
 
-static int
-nvd0_i2c_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	      struct nouveau_oclass *oclass, void *data, u32 size,
-	      struct nouveau_object **pobject)
-{
-	struct nv50_i2c_priv *priv;
-	int ret;
-
-	ret = nouveau_i2c_create(parent, engine, oclass, nvd0_i2c_sclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	return 0;
-}
-
-struct nouveau_oclass
-nvd0_i2c_oclass = {
-	.handle = NV_SUBDEV(I2C, 0xd0),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nvd0_i2c_ctor,
+struct nouveau_oclass *
+nvd0_i2c_oclass = &(struct nouveau_i2c_impl) {
+	.base.handle = NV_SUBDEV(I2C, 0xd0),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_i2c_ctor,
 		.dtor = _nouveau_i2c_dtor,
 		.init = _nouveau_i2c_init,
 		.fini = _nouveau_i2c_fini,
 	},
-};
+	.sclass = nvd0_i2c_sclass,
+	.pad_x = &nv04_i2c_pad_oclass,
+	.pad_s = &nv94_i2c_pad_oclass,
+	.aux = 4,
+	.aux_stat = nv94_aux_stat,
+	.aux_mask = nv94_aux_mask,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/nve0.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/nve0.c
new file mode 100644
index 00000000000..cae77e1ad8d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/nve0.c
@@ -0,0 +1,72 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nv50.h"
+
+static void
+nve0_aux_stat(struct nouveau_i2c *i2c, u32 *hi, u32 *lo, u32 *rq, u32 *tx)
+{
+	u32 intr = nv_rd32(i2c, 0x00dc60);
+	u32 stat = nv_rd32(i2c, 0x00dc68) & intr, i;
+	for (i = 0, *hi = *lo = *rq = *tx = 0; i < 8; i++) {
+		if ((stat & (1 << (i * 4)))) *hi |= 1 << i;
+		if ((stat & (2 << (i * 4)))) *lo |= 1 << i;
+		if ((stat & (4 << (i * 4)))) *rq |= 1 << i;
+		if ((stat & (8 << (i * 4)))) *tx |= 1 << i;
+	}
+	nv_wr32(i2c, 0x00dc60, intr);
+}
+
+static void
+nve0_aux_mask(struct nouveau_i2c *i2c, u32 type, u32 mask, u32 data)
+{
+	u32 temp = nv_rd32(i2c, 0x00dc68), i;
+	for (i = 0; i < 8; i++) {
+		if (mask & (1 << i)) {
+			if (!(data & (1 << i))) {
+				temp &= ~(type << (i * 4));
+				continue;
+			}
+			temp |= type << (i * 4);
+		}
+	}
+	nv_wr32(i2c, 0x00dc68, temp);
+}
+
+struct nouveau_oclass *
+nve0_i2c_oclass = &(struct nouveau_i2c_impl) {
+	.base.handle = NV_SUBDEV(I2C, 0xe0),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nouveau_i2c_ctor,
+		.dtor = _nouveau_i2c_dtor,
+		.init = _nouveau_i2c_init,
+		.fini = _nouveau_i2c_fini,
+	},
+	.sclass = nvd0_i2c_sclass,
+	.pad_x = &nv04_i2c_pad_oclass,
+	.pad_s = &nv94_i2c_pad_oclass,
+	.aux = 4,
+	.aux_stat = nve0_aux_stat,
+	.aux_mask = nve0_aux_mask,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/pad.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/pad.c
new file mode 100644
index 00000000000..e9e412477c1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/pad.c
@@ -0,0 +1,84 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "pad.h"
+
+int
+_nvkm_i2c_pad_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nvkm_i2c_pad *pad = (void *)object;
+	DBG("-> NULL\n");
+	pad->port = NULL;
+	return nouveau_object_fini(&pad->base, suspend);
+}
+
+int
+_nvkm_i2c_pad_init(struct nouveau_object *object)
+{
+	struct nvkm_i2c_pad *pad = (void *)object;
+	DBG("-> PORT:%02x\n", pad->next->index);
+	pad->port = pad->next;
+	return nouveau_object_init(&pad->base);
+}
+
+int
+nvkm_i2c_pad_create_(struct nouveau_object *parent,
+		     struct nouveau_object *engine,
+		     struct nouveau_oclass *oclass, int index,
+		     int size, void **pobject)
+{
+	struct nouveau_i2c *i2c = (void *)engine;
+	struct nouveau_i2c_port *port;
+	struct nvkm_i2c_pad *pad;
+	int ret;
+
+	list_for_each_entry(port, &i2c->ports, head) {
+		pad = nvkm_i2c_pad(port);
+		if (pad->index == index) {
+			atomic_inc(&nv_object(pad)->refcount);
+			*pobject = pad;
+			return 1;
+		}
+	}
+
+	ret = nouveau_object_create_(parent, engine, oclass, 0, size, pobject);
+	pad = *pobject;
+	if (ret)
+		return ret;
+
+	pad->index = index;
+	return 0;
+}
+
+int
+_nvkm_i2c_pad_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+		   struct nouveau_oclass *oclass, void *data, u32 index,
+		   struct nouveau_object **pobject)
+{
+	struct nvkm_i2c_pad *pad;
+	int ret;
+	ret = nvkm_i2c_pad_create(parent, engine, oclass, index, &pad);
+	*pobject = nv_object(pad);
+	return ret;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/pad.h b/drivers/gpu/drm/nouveau/core/subdev/i2c/pad.h
new file mode 100644
index 00000000000..452ac10c300
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/pad.h
@@ -0,0 +1,58 @@
+#ifndef __NVKM_I2C_PAD_H__
+#define __NVKM_I2C_PAD_H__
+
+#include "priv.h"
+
+struct nvkm_i2c_pad {
+	struct nouveau_object base;
+	int index;
+	struct nouveau_i2c_port *port;
+	struct nouveau_i2c_port *next;
+};
+
+static inline struct nvkm_i2c_pad *
+nvkm_i2c_pad(struct nouveau_i2c_port *port)
+{
+	struct nouveau_object *pad = nv_object(port);
+	while (pad->parent)
+		pad = pad->parent;
+	return (void *)pad;
+}
+
+#define nvkm_i2c_pad_create(p,e,o,i,d)                                         \
+	nvkm_i2c_pad_create_((p), (e), (o), (i), sizeof(**d), (void **)d)
+#define nvkm_i2c_pad_destroy(p) ({                                             \
+	struct nvkm_i2c_pad *_p = (p);                                         \
+	_nvkm_i2c_pad_dtor(nv_object(_p));                                     \
+})
+#define nvkm_i2c_pad_init(p) ({                                                \
+	struct nvkm_i2c_pad *_p = (p);                                         \
+	_nvkm_i2c_pad_init(nv_object(_p));                                     \
+})
+#define nvkm_i2c_pad_fini(p,s) ({                                              \
+	struct nvkm_i2c_pad *_p = (p);                                         \
+	_nvkm_i2c_pad_fini(nv_object(_p), (s));                                \
+})
+
+int nvkm_i2c_pad_create_(struct nouveau_object *, struct nouveau_object *,
+			 struct nouveau_oclass *, int index, int, void **);
+
+int _nvkm_i2c_pad_ctor(struct nouveau_object *, struct nouveau_object *,
+		       struct nouveau_oclass *, void *, u32,
+		       struct nouveau_object **);
+#define _nvkm_i2c_pad_dtor nouveau_object_destroy
+int _nvkm_i2c_pad_init(struct nouveau_object *);
+int _nvkm_i2c_pad_fini(struct nouveau_object *, bool);
+
+#ifndef MSG
+#define MSG(l,f,a...) do {                                                     \
+	struct nvkm_i2c_pad *_pad = (void *)pad;                               \
+	nv_##l(nv_object(_pad)->engine, "PAD:%c:%02x: "f,                      \
+	       _pad->index >= 0x100 ? 'X' : 'S',                               \
+	       _pad->index >= 0x100 ? _pad->index - 0x100 : _pad->index, ##a); \
+} while(0)
+#define DBG(f,a...) MSG(debug, f, ##a)
+#define ERR(f,a...) MSG(error, f, ##a)
+#endif
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/padnv04.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/padnv04.c
new file mode 100644
index 00000000000..2c4b61296dd
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/padnv04.c
@@ -0,0 +1,35 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "pad.h"
+
+struct nouveau_oclass
+nv04_i2c_pad_oclass = {
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = _nvkm_i2c_pad_ctor,
+		.dtor = _nvkm_i2c_pad_dtor,
+		.init = _nvkm_i2c_pad_init,
+		.fini = _nvkm_i2c_pad_fini,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/padnv94.c b/drivers/gpu/drm/nouveau/core/subdev/i2c/padnv94.c
new file mode 100644
index 00000000000..0dc6753014f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/padnv94.c
@@ -0,0 +1,86 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "pad.h"
+
+struct nv94_i2c_pad {
+	struct nvkm_i2c_pad base;
+	int addr;
+};
+
+static int
+nv94_i2c_pad_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nouveau_i2c *i2c = (void *)object->engine;
+	struct nv94_i2c_pad *pad = (void *)object;
+	nv_mask(i2c, 0x00e50c + pad->addr, 0x00000001, 0x00000001);
+	return nvkm_i2c_pad_fini(&pad->base, suspend);
+}
+
+static int
+nv94_i2c_pad_init(struct nouveau_object *object)
+{
+	struct nouveau_i2c *i2c = (void *)object->engine;
+	struct nv94_i2c_pad *pad = (void *)object;
+
+	switch (nv_oclass(pad->base.next)->handle) {
+	case NV_I2C_TYPE_DCBI2C(DCB_I2C_NVIO_AUX):
+		nv_mask(i2c, 0x00e500 + pad->addr, 0x0000c003, 0x00000002);
+		break;
+	case NV_I2C_TYPE_DCBI2C(DCB_I2C_NVIO_BIT):
+	default:
+		nv_mask(i2c, 0x00e500 + pad->addr, 0x0000c003, 0x0000c001);
+		break;
+	}
+
+	nv_mask(i2c, 0x00e50c + pad->addr, 0x00000001, 0x00000000);
+	return nvkm_i2c_pad_init(&pad->base);
+}
+
+static int
+nv94_i2c_pad_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+		  struct nouveau_oclass *oclass, void *data, u32 index,
+		  struct nouveau_object **pobject)
+{
+	struct nv94_i2c_pad *pad;
+	int ret;
+
+	ret = nvkm_i2c_pad_create(parent, engine, oclass, index, &pad);
+	*pobject = nv_object(pad);
+	if (ret)
+		return ret;
+
+	pad->addr = index * 0x50;;
+	return 0;
+}
+
+struct nouveau_oclass
+nv94_i2c_pad_oclass = {
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv94_i2c_pad_ctor,
+		.dtor = _nvkm_i2c_pad_dtor,
+		.init = nv94_i2c_pad_init,
+		.fini = nv94_i2c_pad_fini,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/port.h b/drivers/gpu/drm/nouveau/core/subdev/i2c/port.h
new file mode 100644
index 00000000000..a8ff6e077af
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/port.h
@@ -0,0 +1,15 @@
+#ifndef __NVKM_I2C_PORT_H__
+#define __NVKM_I2C_PORT_H__
+
+#include "priv.h"
+
+#ifndef MSG
+#define MSG(l,f,a...) do {                                                     \
+	struct nouveau_i2c_port *_port = (void *)port;                         \
+	nv_##l(nv_object(_port)->engine, "PORT:%02x: "f, _port->index, ##a);   \
+} while(0)
+#define DBG(f,a...) MSG(debug, f, ##a)
+#define ERR(f,a...) MSG(error, f, ##a)
+#endif
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/i2c/priv.h b/drivers/gpu/drm/nouveau/core/subdev/i2c/priv.h
new file mode 100644
index 00000000000..780090b6425
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/i2c/priv.h
@@ -0,0 +1,85 @@
+#ifndef __NVKM_I2C_H__
+#define __NVKM_I2C_H__
+
+#include <subdev/i2c.h>
+
+extern struct nouveau_oclass nv04_i2c_pad_oclass;
+extern struct nouveau_oclass nv94_i2c_pad_oclass;
+
+#define nouveau_i2c_port_create(p,e,o,i,a,f,d)                                 \
+	nouveau_i2c_port_create_((p), (e), (o), (i), (a), (f),                 \
+				 sizeof(**d), (void **)d)
+#define nouveau_i2c_port_destroy(p) ({                                         \
+	struct nouveau_i2c_port *port = (p);                                   \
+	_nouveau_i2c_port_dtor(nv_object(i2c));                                \
+})
+#define nouveau_i2c_port_init(p)                                               \
+	nouveau_object_init(&(p)->base)
+#define nouveau_i2c_port_fini(p,s)                                             \
+	nouveau_object_fini(&(p)->base, (s))
+
+int nouveau_i2c_port_create_(struct nouveau_object *, struct nouveau_object *,
+			     struct nouveau_oclass *, u8,
+			     const struct i2c_algorithm *,
+			     const struct nouveau_i2c_func *,
+			     int, void **);
+void _nouveau_i2c_port_dtor(struct nouveau_object *);
+#define _nouveau_i2c_port_init nouveau_object_init
+int  _nouveau_i2c_port_fini(struct nouveau_object *, bool);
+
+#define nouveau_i2c_create(p,e,o,d)                                            \
+	nouveau_i2c_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_i2c_destroy(p) ({                                              \
+	struct nouveau_i2c *i2c = (p);                                         \
+	_nouveau_i2c_dtor(nv_object(i2c));                                     \
+})
+#define nouveau_i2c_init(p) ({                                                 \
+	struct nouveau_i2c *i2c = (p);                                         \
+	_nouveau_i2c_init(nv_object(i2c));                                     \
+})
+#define nouveau_i2c_fini(p,s) ({                                               \
+	struct nouveau_i2c *i2c = (p);                                         \
+	_nouveau_i2c_fini(nv_object(i2c), (s));                                \
+})
+
+int nouveau_i2c_create_(struct nouveau_object *, struct nouveau_object *,
+			struct nouveau_oclass *, int, void **);
+int  _nouveau_i2c_ctor(struct nouveau_object *, struct nouveau_object *,
+		       struct nouveau_oclass *, void *, u32,
+		       struct nouveau_object **);
+void _nouveau_i2c_dtor(struct nouveau_object *);
+int  _nouveau_i2c_init(struct nouveau_object *);
+int  _nouveau_i2c_fini(struct nouveau_object *, bool);
+
+extern struct nouveau_oclass nouveau_anx9805_sclass[];
+extern struct nouveau_oclass nvd0_i2c_sclass[];
+
+extern const struct i2c_algorithm nouveau_i2c_bit_algo;
+extern const struct i2c_algorithm nouveau_i2c_aux_algo;
+
+struct nouveau_i2c_impl {
+	struct nouveau_oclass base;
+
+	/* supported i2c port classes */
+	struct nouveau_oclass *sclass;
+	struct nouveau_oclass *pad_x;
+	struct nouveau_oclass *pad_s;
+
+	/* number of native dp aux channels present */
+	int aux;
+
+	/* read and ack pending interrupts, returning only data
+	 * for ports that have not been masked off, while still
+	 * performing the ack for anything that was pending.
+	 */
+	void (*aux_stat)(struct nouveau_i2c *, u32 *, u32 *, u32 *, u32 *);
+
+	/* mask on/off interrupt types for a given set of auxch
+	 */
+	void (*aux_mask)(struct nouveau_i2c *, u32, u32, u32);
+};
+
+void nv94_aux_stat(struct nouveau_i2c *, u32 *, u32 *, u32 *, u32 *);
+void nv94_aux_mask(struct nouveau_i2c *, u32, u32, u32);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ibus/gk20a.c b/drivers/gpu/drm/nouveau/core/subdev/ibus/gk20a.c
new file mode 100644
index 00000000000..245f0ebaa6a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/ibus/gk20a.c
@@ -0,0 +1,103 @@
+/*
+ * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include <subdev/ibus.h>
+#include <subdev/timer.h>
+
+struct gk20a_ibus_priv {
+	struct nouveau_ibus base;
+};
+
+static void
+gk20a_ibus_init_priv_ring(struct gk20a_ibus_priv *priv)
+{
+	nv_mask(priv, 0x137250, 0x3f, 0);
+
+	nv_mask(priv, 0x000200, 0x20, 0);
+	usleep_range(20, 30);
+	nv_mask(priv, 0x000200, 0x20, 0x20);
+
+	nv_wr32(priv, 0x12004c, 0x4);
+	nv_wr32(priv, 0x122204, 0x2);
+	nv_rd32(priv, 0x122204);
+}
+
+static void
+gk20a_ibus_intr(struct nouveau_subdev *subdev)
+{
+	struct gk20a_ibus_priv *priv = (void *)subdev;
+	u32 status0 = nv_rd32(priv, 0x120058);
+
+	if (status0 & 0x7) {
+		nv_debug(priv, "resetting priv ring\n");
+		gk20a_ibus_init_priv_ring(priv);
+	}
+
+	/* Acknowledge interrupt */
+	nv_mask(priv, 0x12004c, 0x2, 0x2);
+
+	if (!nv_wait(subdev, 0x12004c, 0x3f, 0x00))
+		nv_warn(priv, "timeout waiting for ringmaster ack\n");
+}
+
+static int
+gk20a_ibus_init(struct nouveau_object *object)
+{
+	struct gk20a_ibus_priv *priv = (void *)object;
+	int ret;
+
+	ret = _nouveau_ibus_init(object);
+	if (ret)
+		return ret;
+
+	gk20a_ibus_init_priv_ring(priv);
+
+	return 0;
+}
+
+static int
+gk20a_ibus_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct gk20a_ibus_priv *priv;
+	int ret;
+
+	ret = nouveau_ibus_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	nv_subdev(priv)->intr = gk20a_ibus_intr;
+	return 0;
+}
+
+struct nouveau_oclass
+gk20a_ibus_oclass = {
+	.handle = NV_SUBDEV(IBUS, 0xea),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gk20a_ibus_ctor,
+		.dtor = _nouveau_ibus_dtor,
+		.init = gk20a_ibus_init,
+		.fini = _nouveau_ibus_fini,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ibus/nve0.c b/drivers/gpu/drm/nouveau/core/subdev/ibus/nve0.c
index 7120124dcea..ebef970a064 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ibus/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/ibus/nve0.c
@@ -95,6 +95,23 @@ nve0_ibus_intr(struct nouveau_subdev *subdev)
 }
 
 static int
+nve0_ibus_init(struct nouveau_object *object)
+{
+	struct nve0_ibus_priv *priv = (void *)object;
+	int ret = nouveau_ibus_init(&priv->base);
+	if (ret == 0) {
+		nv_mask(priv, 0x122318, 0x0003ffff, 0x00001000);
+		nv_mask(priv, 0x12231c, 0x0003ffff, 0x00000200);
+		nv_mask(priv, 0x122310, 0x0003ffff, 0x00000800);
+		nv_mask(priv, 0x122348, 0x0003ffff, 0x00000100);
+		nv_mask(priv, 0x1223b0, 0x0003ffff, 0x00000fff);
+		nv_mask(priv, 0x122348, 0x0003ffff, 0x00000200);
+		nv_mask(priv, 0x122358, 0x0003ffff, 0x00002880);
+	}
+	return ret;
+}
+
+static int
 nve0_ibus_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	       struct nouveau_oclass *oclass, void *data, u32 size,
 	       struct nouveau_object **pobject)
@@ -117,7 +134,7 @@ nve0_ibus_oclass = {
 	.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nve0_ibus_ctor,
 		.dtor = _nouveau_ibus_dtor,
-		.init = _nouveau_ibus_init,
+		.init = nve0_ibus_init,
 		.fini = _nouveau_ibus_fini,
 	},
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/base.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/base.c
index 6565f3dbbe0..14706d9842c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/base.c
@@ -22,7 +22,24 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/instmem.h>
+#include "priv.h"
+
+/******************************************************************************
+ * instmem object base implementation
+ *****************************************************************************/
+
+void
+_nouveau_instobj_dtor(struct nouveau_object *object)
+{
+	struct nouveau_instmem *imem = (void *)object->engine;
+	struct nouveau_instobj *iobj = (void *)object;
+
+	mutex_lock(&nv_subdev(imem)->mutex);
+	list_del(&iobj->head);
+	mutex_unlock(&nv_subdev(imem)->mutex);
+
+	return nouveau_object_destroy(&iobj->base);
+}
 
 int
 nouveau_instobj_create_(struct nouveau_object *parent,
@@ -46,73 +63,26 @@ nouveau_instobj_create_(struct nouveau_object *parent,
 	return 0;
 }
 
-void
-nouveau_instobj_destroy(struct nouveau_instobj *iobj)
-{
-	struct nouveau_subdev *subdev = nv_subdev(iobj->base.engine);
+/******************************************************************************
+ * instmem subdev base implementation
+ *****************************************************************************/
 
-	mutex_lock(&subdev->mutex);
-	list_del(&iobj->head);
-	mutex_unlock(&subdev->mutex);
-
-	return nouveau_object_destroy(&iobj->base);
-}
-
-void
-_nouveau_instobj_dtor(struct nouveau_object *object)
+static int
+nouveau_instmem_alloc(struct nouveau_instmem *imem,
+		      struct nouveau_object *parent, u32 size, u32 align,
+		      struct nouveau_object **pobject)
 {
-	struct nouveau_instobj *iobj = (void *)object;
-	return nouveau_instobj_destroy(iobj);
+	struct nouveau_object *engine = nv_object(imem);
+	struct nouveau_instmem_impl *impl = (void *)engine->oclass;
+	struct nouveau_instobj_args args = { .size = size, .align = align };
+	return nouveau_object_ctor(parent, engine, impl->instobj, &args,
+				   sizeof(args), pobject);
 }
 
 int
-nouveau_instmem_create_(struct nouveau_object *parent,
-			struct nouveau_object *engine,
-			struct nouveau_oclass *oclass,
-			int length, void **pobject)
-{
-	struct nouveau_instmem *imem;
-	int ret;
-
-	ret = nouveau_subdev_create_(parent, engine, oclass, 0,
-				     "INSTMEM", "instmem", length, pobject);
-	imem = *pobject;
-	if (ret)
-		return ret;
-
-	INIT_LIST_HEAD(&imem->list);
-	return 0;
-}
-
-int
-nouveau_instmem_init(struct nouveau_instmem *imem)
-{
-	struct nouveau_instobj *iobj;
-	int ret, i;
-
-	ret = nouveau_subdev_init(&imem->base);
-	if (ret)
-		return ret;
-
-	mutex_lock(&imem->base.mutex);
-
-	list_for_each_entry(iobj, &imem->list, head) {
-		if (iobj->suspend) {
-			for (i = 0; i < iobj->size; i += 4)
-				nv_wo32(iobj, i, iobj->suspend[i / 4]);
-			vfree(iobj->suspend);
-			iobj->suspend = NULL;
-		}
-	}
-
-	mutex_unlock(&imem->base.mutex);
-
-	return 0;
-}
-
-int
-nouveau_instmem_fini(struct nouveau_instmem *imem, bool suspend)
+_nouveau_instmem_fini(struct nouveau_object *object, bool suspend)
 {
+	struct nouveau_instmem *imem = (void *)object;
 	struct nouveau_instobj *iobj;
 	int i, ret = 0;
 
@@ -143,12 +113,45 @@ int
 _nouveau_instmem_init(struct nouveau_object *object)
 {
 	struct nouveau_instmem *imem = (void *)object;
-	return nouveau_instmem_init(imem);
+	struct nouveau_instobj *iobj;
+	int ret, i;
+
+	ret = nouveau_subdev_init(&imem->base);
+	if (ret)
+		return ret;
+
+	mutex_lock(&imem->base.mutex);
+
+	list_for_each_entry(iobj, &imem->list, head) {
+		if (iobj->suspend) {
+			for (i = 0; i < iobj->size; i += 4)
+				nv_wo32(iobj, i, iobj->suspend[i / 4]);
+			vfree(iobj->suspend);
+			iobj->suspend = NULL;
+		}
+	}
+
+	mutex_unlock(&imem->base.mutex);
+
+	return 0;
 }
 
 int
-_nouveau_instmem_fini(struct nouveau_object *object, bool suspend)
+nouveau_instmem_create_(struct nouveau_object *parent,
+			struct nouveau_object *engine,
+			struct nouveau_oclass *oclass,
+			int length, void **pobject)
 {
-	struct nouveau_instmem *imem = (void *)object;
-	return nouveau_instmem_fini(imem, suspend);
+	struct nouveau_instmem *imem;
+	int ret;
+
+	ret = nouveau_subdev_create_(parent, engine, oclass, 0,
+				     "INSTMEM", "instmem", length, pobject);
+	imem = *pobject;
+	if (ret)
+		return ret;
+
+	INIT_LIST_HEAD(&imem->list);
+	imem->alloc = nouveau_instmem_alloc;
+	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
index 795393d7b2f..7b64befee48 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
@@ -22,10 +22,35 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/fb.h>
-
 #include "nv04.h"
 
+/******************************************************************************
+ * instmem object implementation
+ *****************************************************************************/
+
+static u32
+nv04_instobj_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct nv04_instobj_priv *node = (void *)object;
+	return nv_ro32(object->engine, node->mem->offset + addr);
+}
+
+static void
+nv04_instobj_wr32(struct nouveau_object *object, u64 addr, u32 data)
+{
+	struct nv04_instobj_priv *node = (void *)object;
+	nv_wo32(object->engine, node->mem->offset + addr, data);
+}
+
+static void
+nv04_instobj_dtor(struct nouveau_object *object)
+{
+	struct nv04_instmem_priv *priv = (void *)object->engine;
+	struct nv04_instobj_priv *node = (void *)object;
+	nouveau_mm_free(&priv->heap, &node->mem);
+	nouveau_instobj_destroy(&node->base);
+}
+
 static int
 nv04_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
@@ -33,18 +58,19 @@ nv04_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 {
 	struct nv04_instmem_priv *priv = (void *)engine;
 	struct nv04_instobj_priv *node;
-	int ret, align;
+	struct nouveau_instobj_args *args = data;
+	int ret;
 
-	align = (unsigned long)data;
-	if (!align)
-		align = 1;
+	if (!args->align)
+		args->align = 1;
 
 	ret = nouveau_instobj_create(parent, engine, oclass, &node);
 	*pobject = nv_object(node);
 	if (ret)
 		return ret;
 
-	ret = nouveau_mm_head(&priv->heap, 1, size, size, align, &node->mem);
+	ret = nouveau_mm_head(&priv->heap, 1, args->size, args->size,
+			      args->align, &node->mem);
 	if (ret)
 		return ret;
 
@@ -53,32 +79,9 @@ nv04_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	return 0;
 }
 
-static void
-nv04_instobj_dtor(struct nouveau_object *object)
-{
-	struct nv04_instmem_priv *priv = (void *)object->engine;
-	struct nv04_instobj_priv *node = (void *)object;
-	nouveau_mm_free(&priv->heap, &node->mem);
-	nouveau_instobj_destroy(&node->base);
-}
-
-static u32
-nv04_instobj_rd32(struct nouveau_object *object, u64 addr)
-{
-	struct nv04_instobj_priv *node = (void *)object;
-	return nv_ro32(object->engine, node->mem->offset + addr);
-}
-
-static void
-nv04_instobj_wr32(struct nouveau_object *object, u64 addr, u32 data)
-{
-	struct nv04_instobj_priv *node = (void *)object;
-	nv_wo32(object->engine, node->mem->offset + addr, data);
-}
-
-static struct nouveau_oclass
+struct nouveau_instobj_impl
 nv04_instobj_oclass = {
-	.ofuncs = &(struct nouveau_ofuncs) {
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv04_instobj_ctor,
 		.dtor = nv04_instobj_dtor,
 		.init = _nouveau_instobj_init,
@@ -88,19 +91,34 @@ nv04_instobj_oclass = {
 	},
 };
 
-int
-nv04_instmem_alloc(struct nouveau_instmem *imem, struct nouveau_object *parent,
-		   u32 size, u32 align, struct nouveau_object **pobject)
+/******************************************************************************
+ * instmem subdev implementation
+ *****************************************************************************/
+
+static u32
+nv04_instmem_rd32(struct nouveau_object *object, u64 addr)
 {
-	struct nouveau_object *engine = nv_object(imem);
-	int ret;
+	return nv_rd32(object, 0x700000 + addr);
+}
 
-	ret = nouveau_object_ctor(parent, engine, &nv04_instobj_oclass,
-				  (void *)(unsigned long)align, size, pobject);
-	if (ret)
-		return ret;
+static void
+nv04_instmem_wr32(struct nouveau_object *object, u64 addr, u32 data)
+{
+	return nv_wr32(object, 0x700000 + addr, data);
+}
 
-	return 0;
+void
+nv04_instmem_dtor(struct nouveau_object *object)
+{
+	struct nv04_instmem_priv *priv = (void *)object;
+	nouveau_gpuobj_ref(NULL, &priv->ramfc);
+	nouveau_gpuobj_ref(NULL, &priv->ramro);
+	nouveau_ramht_ref(NULL, &priv->ramht);
+	nouveau_gpuobj_ref(NULL, &priv->vbios);
+	nouveau_mm_fini(&priv->heap);
+	if (priv->iomem)
+		iounmap(priv->iomem);
+	nouveau_instmem_destroy(&priv->base);
 }
 
 static int
@@ -118,7 +136,6 @@ nv04_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 	/* PRAMIN aperture maps over the end of VRAM, reserve it */
 	priv->base.reserved = 512 * 1024;
-	priv->base.alloc    = nv04_instmem_alloc;
 
 	ret = nouveau_mm_init(&priv->heap, 0, priv->base.reserved, 1);
 	if (ret)
@@ -150,36 +167,10 @@ nv04_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	return 0;
 }
 
-void
-nv04_instmem_dtor(struct nouveau_object *object)
-{
-	struct nv04_instmem_priv *priv = (void *)object;
-	nouveau_gpuobj_ref(NULL, &priv->ramfc);
-	nouveau_gpuobj_ref(NULL, &priv->ramro);
-	nouveau_ramht_ref(NULL, &priv->ramht);
-	nouveau_gpuobj_ref(NULL, &priv->vbios);
-	nouveau_mm_fini(&priv->heap);
-	if (priv->iomem)
-		iounmap(priv->iomem);
-	nouveau_instmem_destroy(&priv->base);
-}
-
-static u32
-nv04_instmem_rd32(struct nouveau_object *object, u64 addr)
-{
-	return nv_rd32(object, 0x700000 + addr);
-}
-
-static void
-nv04_instmem_wr32(struct nouveau_object *object, u64 addr, u32 data)
-{
-	return nv_wr32(object, 0x700000 + addr, data);
-}
-
-struct nouveau_oclass
-nv04_instmem_oclass = {
-	.handle = NV_SUBDEV(INSTMEM, 0x04),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv04_instmem_oclass = &(struct nouveau_instmem_impl) {
+	.base.handle = NV_SUBDEV(INSTMEM, 0x04),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv04_instmem_ctor,
 		.dtor = nv04_instmem_dtor,
 		.init = _nouveau_instmem_init,
@@ -187,4 +178,5 @@ nv04_instmem_oclass = {
 		.rd32 = nv04_instmem_rd32,
 		.wr32 = nv04_instmem_wr32,
 	},
-};
+	.instobj = &nv04_instobj_oclass.base,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.h b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.h
index b15b6131023..095fbc6fc09 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.h
@@ -5,7 +5,9 @@
 #include <core/ramht.h>
 #include <core/mm.h>
 
-#include <subdev/instmem.h>
+#include "priv.h"
+
+extern struct nouveau_instobj_impl nv04_instobj_oclass;
 
 struct nv04_instmem_priv {
 	struct nouveau_instmem base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv40.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv40.c
index b10a143787a..8803809f9fc 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv40.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv40.c
@@ -26,13 +26,30 @@
 
 #include "nv04.h"
 
+/******************************************************************************
+ * instmem subdev implementation
+ *****************************************************************************/
+
+static u32
+nv40_instmem_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct nv04_instmem_priv *priv = (void *)object;
+	return ioread32_native(priv->iomem + addr);
+}
+
+static void
+nv40_instmem_wr32(struct nouveau_object *object, u64 addr, u32 data)
+{
+	struct nv04_instmem_priv *priv = (void *)object;
+	iowrite32_native(data, priv->iomem + addr);
+}
+
 static int
 nv40_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
 		  struct nouveau_object **pobject)
 {
 	struct nouveau_device *device = nv_device(parent);
-	struct pci_dev *pdev = device->pdev;
 	struct nv04_instmem_priv *priv;
 	int ret, bar, vs;
 
@@ -42,13 +59,13 @@ nv40_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		return ret;
 
 	/* map bar */
-	if (pci_resource_len(pdev, 2))
+	if (nv_device_resource_len(device, 2))
 		bar = 2;
 	else
 		bar = 3;
 
-	priv->iomem = ioremap(pci_resource_start(pdev, bar),
-			      pci_resource_len(pdev, bar));
+	priv->iomem = ioremap(nv_device_resource_start(device, bar),
+			      nv_device_resource_len(device, bar));
 	if (!priv->iomem) {
 		nv_error(priv, "unable to map PRAMIN BAR\n");
 		return -EFAULT;
@@ -69,7 +86,6 @@ nv40_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->base.reserved += 512 * 1024;	/* object storage */
 
 	priv->base.reserved = round_up(priv->base.reserved, 4096);
-	priv->base.alloc    = nv04_instmem_alloc;
 
 	ret = nouveau_mm_init(&priv->heap, 0, priv->base.reserved, 1);
 	if (ret)
@@ -106,24 +122,10 @@ nv40_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	return 0;
 }
 
-static u32
-nv40_instmem_rd32(struct nouveau_object *object, u64 addr)
-{
-	struct nv04_instmem_priv *priv = (void *)object;
-	return ioread32_native(priv->iomem + addr);
-}
-
-static void
-nv40_instmem_wr32(struct nouveau_object *object, u64 addr, u32 data)
-{
-	struct nv04_instmem_priv *priv = (void *)object;
-	iowrite32_native(data, priv->iomem + addr);
-}
-
-struct nouveau_oclass
-nv40_instmem_oclass = {
-	.handle = NV_SUBDEV(INSTMEM, 0x40),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv40_instmem_oclass = &(struct nouveau_instmem_impl) {
+	.base.handle = NV_SUBDEV(INSTMEM, 0x40),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv40_instmem_ctor,
 		.dtor = nv04_instmem_dtor,
 		.init = _nouveau_instmem_init,
@@ -131,4 +133,5 @@ nv40_instmem_oclass = {
 		.rd32 = nv40_instmem_rd32,
 		.wr32 = nv40_instmem_wr32,
 	},
-};
+	.instobj = &nv04_instobj_oclass.base,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv50.c
index 97bc5dff93e..7cb3b098a08 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv50.c
@@ -22,11 +22,11 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/instmem.h>
 #include <subdev/fb.h>
-
 #include <core/mm.h>
 
+#include "priv.h"
+
 struct nv50_instmem_priv {
 	struct nouveau_instmem base;
 	spinlock_t lock;
@@ -38,42 +38,9 @@ struct nv50_instobj_priv {
 	struct nouveau_mem *mem;
 };
 
-static int
-nv50_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nouveau_fb *pfb = nouveau_fb(parent);
-	struct nv50_instobj_priv *node;
-	u32 align = (unsigned long)data;
-	int ret;
-
-	size  = max((size  + 4095) & ~4095, (u32)4096);
-	align = max((align + 4095) & ~4095, (u32)4096);
-
-	ret = nouveau_instobj_create(parent, engine, oclass, &node);
-	*pobject = nv_object(node);
-	if (ret)
-		return ret;
-
-	ret = pfb->ram->get(pfb, size, align, 0, 0x800, &node->mem);
-	if (ret)
-		return ret;
-
-	node->base.addr = node->mem->offset;
-	node->base.size = node->mem->size << 12;
-	node->mem->page_shift = 12;
-	return 0;
-}
-
-static void
-nv50_instobj_dtor(struct nouveau_object *object)
-{
-	struct nv50_instobj_priv *node = (void *)object;
-	struct nouveau_fb *pfb = nouveau_fb(object);
-	pfb->ram->put(pfb, &node->mem);
-	nouveau_instobj_destroy(&node->base);
-}
+/******************************************************************************
+ * instmem object implementation
+ *****************************************************************************/
 
 static u32
 nv50_instobj_rd32(struct nouveau_object *object, u64 offset)
@@ -113,9 +80,46 @@ nv50_instobj_wr32(struct nouveau_object *object, u64 offset, u32 data)
 	spin_unlock_irqrestore(&priv->lock, flags);
 }
 
-static struct nouveau_oclass
+static void
+nv50_instobj_dtor(struct nouveau_object *object)
+{
+	struct nv50_instobj_priv *node = (void *)object;
+	struct nouveau_fb *pfb = nouveau_fb(object);
+	pfb->ram->put(pfb, &node->mem);
+	nouveau_instobj_destroy(&node->base);
+}
+
+static int
+nv50_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+		  struct nouveau_oclass *oclass, void *data, u32 size,
+		  struct nouveau_object **pobject)
+{
+	struct nouveau_fb *pfb = nouveau_fb(parent);
+	struct nouveau_instobj_args *args = data;
+	struct nv50_instobj_priv *node;
+	int ret;
+
+	args->size  = max((args->size  + 4095) & ~4095, (u32)4096);
+	args->align = max((args->align + 4095) & ~4095, (u32)4096);
+
+	ret = nouveau_instobj_create(parent, engine, oclass, &node);
+	*pobject = nv_object(node);
+	if (ret)
+		return ret;
+
+	ret = pfb->ram->get(pfb, args->size, args->align, 0, 0x800, &node->mem);
+	if (ret)
+		return ret;
+
+	node->base.addr = node->mem->offset;
+	node->base.size = node->mem->size << 12;
+	node->mem->page_shift = 12;
+	return 0;
+}
+
+static struct nouveau_instobj_impl
 nv50_instobj_oclass = {
-	.ofuncs = &(struct nouveau_ofuncs) {
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv50_instobj_ctor,
 		.dtor = nv50_instobj_dtor,
 		.init = _nouveau_instobj_init,
@@ -125,13 +129,16 @@ nv50_instobj_oclass = {
 	},
 };
 
+/******************************************************************************
+ * instmem subdev implementation
+ *****************************************************************************/
+
 static int
-nv50_instmem_alloc(struct nouveau_instmem *imem, struct nouveau_object *parent,
-		   u32 size, u32 align, struct nouveau_object **pobject)
+nv50_instmem_fini(struct nouveau_object *object, bool suspend)
 {
-	struct nouveau_object *engine = nv_object(imem);
-	return nouveau_object_ctor(parent, engine, &nv50_instobj_oclass,
-				   (void *)(unsigned long)align, size, pobject);
+	struct nv50_instmem_priv *priv = (void *)object;
+	priv->addr = ~0ULL;
+	return nouveau_instmem_fini(&priv->base, suspend);
 }
 
 static int
@@ -148,25 +155,17 @@ nv50_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		return ret;
 
 	spin_lock_init(&priv->lock);
-	priv->base.alloc = nv50_instmem_alloc;
 	return 0;
 }
 
-static int
-nv50_instmem_fini(struct nouveau_object *object, bool suspend)
-{
-	struct nv50_instmem_priv *priv = (void *)object;
-	priv->addr = ~0ULL;
-	return nouveau_instmem_fini(&priv->base, suspend);
-}
-
-struct nouveau_oclass
-nv50_instmem_oclass = {
-	.handle = NV_SUBDEV(INSTMEM, 0x50),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv50_instmem_oclass = &(struct nouveau_instmem_impl) {
+	.base.handle = NV_SUBDEV(INSTMEM, 0x50),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv50_instmem_ctor,
 		.dtor = _nouveau_instmem_dtor,
 		.init = _nouveau_instmem_init,
 		.fini = nv50_instmem_fini,
 	},
-};
+	.instobj = &nv50_instobj_oclass.base,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/priv.h b/drivers/gpu/drm/nouveau/core/subdev/instmem/priv.h
new file mode 100644
index 00000000000..8d67dedc5bb
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/priv.h
@@ -0,0 +1,56 @@
+#ifndef __NVKM_INSTMEM_PRIV_H__
+#define __NVKM_INSTMEM_PRIV_H__
+
+#include <subdev/instmem.h>
+
+struct nouveau_instobj_impl {
+	struct nouveau_oclass base;
+};
+
+struct nouveau_instobj_args {
+	u32 size;
+	u32 align;
+};
+
+#define nouveau_instobj_create(p,e,o,d)                                        \
+	nouveau_instobj_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_instobj_destroy(p) ({                                          \
+	struct nouveau_instobj *iobj = (p);                                    \
+	_nouveau_instobj_dtor(nv_object(iobj));                                \
+})
+#define nouveau_instobj_init(p)                                                \
+	nouveau_object_init(&(p)->base)
+#define nouveau_instobj_fini(p,s)                                              \
+	nouveau_object_fini(&(p)->base, (s))
+
+int  nouveau_instobj_create_(struct nouveau_object *, struct nouveau_object *,
+			     struct nouveau_oclass *, int, void **);
+void _nouveau_instobj_dtor(struct nouveau_object *);
+#define _nouveau_instobj_init nouveau_object_init
+#define _nouveau_instobj_fini nouveau_object_fini
+
+struct nouveau_instmem_impl {
+	struct nouveau_oclass base;
+	struct nouveau_oclass *instobj;
+};
+
+#define nouveau_instmem_create(p,e,o,d)                                        \
+	nouveau_instmem_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_instmem_destroy(p)                                             \
+	nouveau_subdev_destroy(&(p)->base)
+#define nouveau_instmem_init(p) ({                                             \
+	struct nouveau_instmem *imem = (p);                                    \
+	_nouveau_instmem_init(nv_object(imem));                                \
+})
+#define nouveau_instmem_fini(p,s) ({                                           \
+	struct nouveau_instmem *imem = (p);                                    \
+	_nouveau_instmem_fini(nv_object(imem), (s));                           \
+})
+
+int nouveau_instmem_create_(struct nouveau_object *, struct nouveau_object *,
+			    struct nouveau_oclass *, int, void **);
+#define _nouveau_instmem_dtor _nouveau_subdev_dtor
+int _nouveau_instmem_init(struct nouveau_object *);
+int _nouveau_instmem_fini(struct nouveau_object *, bool);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltcg/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/ltcg/gf100.c
index cce65cc5651..f2f3338a967 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/ltcg/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltcg/gf100.c
@@ -22,44 +22,35 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/ltcg.h>
 #include <subdev/fb.h>
 #include <subdev/timer.h>
 
-struct nvc0_ltcg_priv {
-	struct nouveau_ltcg base;
-	u32 part_nr;
-	u32 subp_nr;
-	u32 num_tags;
-	u32 tag_base;
-	struct nouveau_mm tags;
-	struct nouveau_mm_node *tag_ram;
-};
+#include "gf100.h"
 
 static void
-nvc0_ltcg_subp_isr(struct nvc0_ltcg_priv *priv, int unit, int subp)
+gf100_ltcg_lts_isr(struct gf100_ltcg_priv *priv, int ltc, int lts)
 {
-	u32 subp_base = 0x141000 + (unit * 0x2000) + (subp * 0x400);
-	u32 stat = nv_rd32(priv, subp_base + 0x020);
+	u32 base = 0x141000 + (ltc * 0x2000) + (lts * 0x400);
+	u32 stat = nv_rd32(priv, base + 0x020);
 
 	if (stat) {
-		nv_info(priv, "LTC%d_LTS%d: 0x%08x\n", unit, subp, stat);
-		nv_wr32(priv, subp_base + 0x020, stat);
+		nv_info(priv, "LTC%d_LTS%d: 0x%08x\n", ltc, lts, stat);
+		nv_wr32(priv, base + 0x020, stat);
 	}
 }
 
 static void
-nvc0_ltcg_intr(struct nouveau_subdev *subdev)
+gf100_ltcg_intr(struct nouveau_subdev *subdev)
 {
-	struct nvc0_ltcg_priv *priv = (void *)subdev;
-	u32 units;
-
-	units = nv_rd32(priv, 0x00017c);
-	while (units) {
-		u32 subp, unit = ffs(units) - 1;
-		for (subp = 0; subp < priv->subp_nr; subp++)
-			nvc0_ltcg_subp_isr(priv, unit, subp);
-		units &= ~(1 << unit);
+	struct gf100_ltcg_priv *priv = (void *)subdev;
+	u32 mask;
+
+	mask = nv_rd32(priv, 0x00017c);
+	while (mask) {
+		u32 lts, ltc = __ffs(mask);
+		for (lts = 0; lts < priv->lts_nr; lts++)
+			gf100_ltcg_lts_isr(priv, ltc, lts);
+		mask &= ~(1 << ltc);
 	}
 
 	/* we do something horribly wrong and upset PMFB a lot, so mask off
@@ -68,11 +59,11 @@ nvc0_ltcg_intr(struct nouveau_subdev *subdev)
 	nv_mask(priv, 0x000640, 0x02000000, 0x00000000);
 }
 
-static int
-nvc0_ltcg_tags_alloc(struct nouveau_ltcg *ltcg, u32 n,
+int
+gf100_ltcg_tags_alloc(struct nouveau_ltcg *ltcg, u32 n,
 		     struct nouveau_mm_node **pnode)
 {
-	struct nvc0_ltcg_priv *priv = (struct nvc0_ltcg_priv *)ltcg;
+	struct gf100_ltcg_priv *priv = (struct gf100_ltcg_priv *)ltcg;
 	int ret;
 
 	ret = nouveau_mm_head(&priv->tags, 1, n, n, 1, pnode);
@@ -82,18 +73,18 @@ nvc0_ltcg_tags_alloc(struct nouveau_ltcg *ltcg, u32 n,
 	return ret;
 }
 
-static void
-nvc0_ltcg_tags_free(struct nouveau_ltcg *ltcg, struct nouveau_mm_node **pnode)
+void
+gf100_ltcg_tags_free(struct nouveau_ltcg *ltcg, struct nouveau_mm_node **pnode)
 {
-	struct nvc0_ltcg_priv *priv = (struct nvc0_ltcg_priv *)ltcg;
+	struct gf100_ltcg_priv *priv = (struct gf100_ltcg_priv *)ltcg;
 
 	nouveau_mm_free(&priv->tags, pnode);
 }
 
 static void
-nvc0_ltcg_tags_clear(struct nouveau_ltcg *ltcg, u32 first, u32 count)
+gf100_ltcg_tags_clear(struct nouveau_ltcg *ltcg, u32 first, u32 count)
 {
-	struct nvc0_ltcg_priv *priv = (struct nvc0_ltcg_priv *)ltcg;
+	struct gf100_ltcg_priv *priv = (struct gf100_ltcg_priv *)ltcg;
 	u32 last = first + count - 1;
 	int p, i;
 
@@ -104,16 +95,16 @@ nvc0_ltcg_tags_clear(struct nouveau_ltcg *ltcg, u32 first, u32 count)
 	nv_wr32(priv, 0x17e8c8, 0x4); /* trigger clear */
 
 	/* wait until it's finished with clearing */
-	for (p = 0; p < priv->part_nr; ++p) {
-		for (i = 0; i < priv->subp_nr; ++i)
+	for (p = 0; p < priv->ltc_nr; ++p) {
+		for (i = 0; i < priv->lts_nr; ++i)
 			nv_wait(priv, 0x1410c8 + p * 0x2000 + i * 0x400, ~0, 0);
 	}
 }
 
 /* TODO: Figure out tag memory details and drop the over-cautious allocation.
  */
-static int
-nvc0_ltcg_init_tag_ram(struct nouveau_fb *pfb, struct nvc0_ltcg_priv *priv)
+int
+gf100_ltcg_init_tag_ram(struct nouveau_fb *pfb, struct gf100_ltcg_priv *priv)
 {
 	u32 tag_size, tag_margin, tag_align;
 	int ret;
@@ -124,7 +115,7 @@ nvc0_ltcg_init_tag_ram(struct nouveau_fb *pfb, struct nvc0_ltcg_priv *priv)
 		priv->num_tags = 1 << 17; /* we have 17 bits in PTE */
 	priv->num_tags = (priv->num_tags + 63) & ~63; /* round up to 64 */
 
-	tag_align = priv->part_nr * 0x800;
+	tag_align = priv->ltc_nr * 0x800;
 	tag_margin = (tag_align < 0x6000) ? 0x6000 : tag_align;
 
 	/* 4 part 4 sub: 0x2000 bytes for 56 tags */
@@ -157,11 +148,11 @@ nvc0_ltcg_init_tag_ram(struct nouveau_fb *pfb, struct nvc0_ltcg_priv *priv)
 }
 
 static int
-nvc0_ltcg_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+gf100_ltcg_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	       struct nouveau_oclass *oclass, void *data, u32 size,
 	       struct nouveau_object **pobject)
 {
-	struct nvc0_ltcg_priv *priv;
+	struct gf100_ltcg_priv *priv;
 	struct nouveau_fb *pfb = nouveau_fb(parent);
 	u32 parts, mask;
 	int ret, i;
@@ -175,27 +166,27 @@ nvc0_ltcg_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	mask = nv_rd32(priv, 0x022554);
 	for (i = 0; i < parts; i++) {
 		if (!(mask & (1 << i)))
-			priv->part_nr++;
+			priv->ltc_nr++;
 	}
-	priv->subp_nr = nv_rd32(priv, 0x17e8dc) >> 28;
+	priv->lts_nr = nv_rd32(priv, 0x17e8dc) >> 28;
 
-	ret = nvc0_ltcg_init_tag_ram(pfb, priv);
+	ret = gf100_ltcg_init_tag_ram(pfb, priv);
 	if (ret)
 		return ret;
 
-	priv->base.tags_alloc = nvc0_ltcg_tags_alloc;
-	priv->base.tags_free  = nvc0_ltcg_tags_free;
-	priv->base.tags_clear = nvc0_ltcg_tags_clear;
+	priv->base.tags_alloc = gf100_ltcg_tags_alloc;
+	priv->base.tags_free  = gf100_ltcg_tags_free;
+	priv->base.tags_clear = gf100_ltcg_tags_clear;
 
-	nv_subdev(priv)->intr = nvc0_ltcg_intr;
+	nv_subdev(priv)->intr = gf100_ltcg_intr;
 	return 0;
 }
 
-static void
-nvc0_ltcg_dtor(struct nouveau_object *object)
+void
+gf100_ltcg_dtor(struct nouveau_object *object)
 {
 	struct nouveau_ltcg *ltcg = (struct nouveau_ltcg *)object;
-	struct nvc0_ltcg_priv *priv = (struct nvc0_ltcg_priv *)ltcg;
+	struct gf100_ltcg_priv *priv = (struct gf100_ltcg_priv *)ltcg;
 	struct nouveau_fb *pfb = nouveau_fb(ltcg->base.base.parent);
 
 	nouveau_mm_fini(&priv->tags);
@@ -205,10 +196,10 @@ nvc0_ltcg_dtor(struct nouveau_object *object)
 }
 
 static int
-nvc0_ltcg_init(struct nouveau_object *object)
+gf100_ltcg_init(struct nouveau_object *object)
 {
 	struct nouveau_ltcg *ltcg = (struct nouveau_ltcg *)object;
-	struct nvc0_ltcg_priv *priv = (struct nvc0_ltcg_priv *)ltcg;
+	struct gf100_ltcg_priv *priv = (struct gf100_ltcg_priv *)ltcg;
 	int ret;
 
 	ret = nouveau_ltcg_init(ltcg);
@@ -216,20 +207,20 @@ nvc0_ltcg_init(struct nouveau_object *object)
 		return ret;
 
 	nv_mask(priv, 0x17e820, 0x00100000, 0x00000000); /* INTR_EN &= ~0x10 */
-	nv_wr32(priv, 0x17e8d8, priv->part_nr);
+	nv_wr32(priv, 0x17e8d8, priv->ltc_nr);
 	if (nv_device(ltcg)->card_type >= NV_E0)
-		nv_wr32(priv, 0x17e000, priv->part_nr);
+		nv_wr32(priv, 0x17e000, priv->ltc_nr);
 	nv_wr32(priv, 0x17e8d4, priv->tag_base);
 	return 0;
 }
 
-struct nouveau_oclass
-nvc0_ltcg_oclass = {
+struct nouveau_oclass *
+gf100_ltcg_oclass = &(struct nouveau_oclass) {
 	.handle = NV_SUBDEV(LTCG, 0xc0),
 	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nvc0_ltcg_ctor,
-		.dtor = nvc0_ltcg_dtor,
-		.init = nvc0_ltcg_init,
+		.ctor = gf100_ltcg_ctor,
+		.dtor = gf100_ltcg_dtor,
+		.init = gf100_ltcg_init,
 		.fini = _nouveau_ltcg_fini,
 	},
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltcg/gf100.h b/drivers/gpu/drm/nouveau/core/subdev/ltcg/gf100.h
new file mode 100644
index 00000000000..87b10b8412e
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltcg/gf100.h
@@ -0,0 +1,21 @@
+#ifndef __NVKM_LTCG_PRIV_GF100_H__
+#define __NVKM_LTCG_PRIV_GF100_H__
+
+#include <subdev/ltcg.h>
+
+struct gf100_ltcg_priv {
+	struct nouveau_ltcg base;
+	u32 ltc_nr;
+	u32 lts_nr;
+	u32 num_tags;
+	u32 tag_base;
+	struct nouveau_mm tags;
+	struct nouveau_mm_node *tag_ram;
+};
+
+void gf100_ltcg_dtor(struct nouveau_object *);
+int  gf100_ltcg_init_tag_ram(struct nouveau_fb *, struct gf100_ltcg_priv *);
+int  gf100_ltcg_tags_alloc(struct nouveau_ltcg *, u32, struct nouveau_mm_node **);
+void gf100_ltcg_tags_free(struct nouveau_ltcg *, struct nouveau_mm_node **);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/ltcg/gm107.c b/drivers/gpu/drm/nouveau/core/subdev/ltcg/gm107.c
new file mode 100644
index 00000000000..e79d0e81de4
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/ltcg/gm107.c
@@ -0,0 +1,142 @@
+/*
+ * Copyright 2014 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include <subdev/fb.h>
+#include <subdev/timer.h>
+
+#include "gf100.h"
+
+static void
+gm107_ltcg_lts_isr(struct gf100_ltcg_priv *priv, int ltc, int lts)
+{
+	u32 base = 0x140000 + (ltc * 0x2000) + (lts * 0x400);
+	u32 stat = nv_rd32(priv, base + 0x00c);
+
+	if (stat) {
+		nv_info(priv, "LTC%d_LTS%d: 0x%08x\n", ltc, lts, stat);
+		nv_wr32(priv, base + 0x00c, stat);
+	}
+}
+
+static void
+gm107_ltcg_intr(struct nouveau_subdev *subdev)
+{
+	struct gf100_ltcg_priv *priv = (void *)subdev;
+	u32 mask;
+
+	mask = nv_rd32(priv, 0x00017c);
+	while (mask) {
+		u32 lts, ltc = __ffs(mask);
+		for (lts = 0; lts < priv->lts_nr; lts++)
+			gm107_ltcg_lts_isr(priv, ltc, lts);
+		mask &= ~(1 << ltc);
+	}
+
+	/* we do something horribly wrong and upset PMFB a lot, so mask off
+	 * interrupts from it after the first one until it's fixed
+	 */
+	nv_mask(priv, 0x000640, 0x02000000, 0x00000000);
+}
+
+static void
+gm107_ltcg_tags_clear(struct nouveau_ltcg *ltcg, u32 first, u32 count)
+{
+	struct gf100_ltcg_priv *priv = (struct gf100_ltcg_priv *)ltcg;
+	u32 last = first + count - 1;
+	int p, i;
+
+	BUG_ON((first > last) || (last >= priv->num_tags));
+
+	nv_wr32(priv, 0x17e270, first);
+	nv_wr32(priv, 0x17e274, last);
+	nv_wr32(priv, 0x17e26c, 0x4); /* trigger clear */
+
+	/* wait until it's finished with clearing */
+	for (p = 0; p < priv->ltc_nr; ++p) {
+		for (i = 0; i < priv->lts_nr; ++i)
+			nv_wait(priv, 0x14046c + p * 0x2000 + i * 0x200, ~0, 0);
+	}
+}
+
+static int
+gm107_ltcg_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
+{
+	struct gf100_ltcg_priv *priv;
+	struct nouveau_fb *pfb = nouveau_fb(parent);
+	u32 parts, mask;
+	int ret, i;
+
+	ret = nouveau_ltcg_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	parts = nv_rd32(priv, 0x022438);
+	mask = nv_rd32(priv, 0x021c14);
+	for (i = 0; i < parts; i++) {
+		if (!(mask & (1 << i)))
+			priv->ltc_nr++;
+	}
+	priv->lts_nr = nv_rd32(priv, 0x17e280) >> 28;
+
+	ret = gf100_ltcg_init_tag_ram(pfb, priv);
+	if (ret)
+		return ret;
+
+	priv->base.tags_alloc = gf100_ltcg_tags_alloc;
+	priv->base.tags_free  = gf100_ltcg_tags_free;
+	priv->base.tags_clear = gm107_ltcg_tags_clear;
+
+	nv_subdev(priv)->intr = gm107_ltcg_intr;
+	return 0;
+}
+
+static int
+gm107_ltcg_init(struct nouveau_object *object)
+{
+	struct nouveau_ltcg *ltcg = (struct nouveau_ltcg *)object;
+	struct gf100_ltcg_priv *priv = (struct gf100_ltcg_priv *)ltcg;
+	int ret;
+
+	ret = nouveau_ltcg_init(ltcg);
+	if (ret)
+		return ret;
+
+	nv_wr32(priv, 0x17e27c, priv->ltc_nr);
+	nv_wr32(priv, 0x17e278, priv->tag_base);
+	return 0;
+}
+
+struct nouveau_oclass *
+gm107_ltcg_oclass = &(struct nouveau_oclass) {
+	.handle = NV_SUBDEV(LTCG, 0xff),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = gm107_ltcg_ctor,
+		.dtor = gf100_ltcg_dtor,
+		.init = gm107_ltcg_init,
+		.fini = _nouveau_ltcg_fini,
+	},
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/base.c b/drivers/gpu/drm/nouveau/core/subdev/mc/base.c
index b4b9943773b..8a5555192fa 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mc/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/mc/base.c
@@ -93,7 +93,7 @@ _nouveau_mc_dtor(struct nouveau_object *object)
 {
 	struct nouveau_device *device = nv_device(object);
 	struct nouveau_mc *pmc = (void *)object;
-	free_irq(device->pdev->irq, pmc);
+	free_irq(pmc->irq, pmc);
 	if (pmc->use_msi)
 		pci_disable_msi(device->pdev);
 	nouveau_subdev_destroy(&pmc->base);
@@ -114,33 +114,44 @@ nouveau_mc_create_(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	switch (device->pdev->device & 0x0ff0) {
-	case 0x00f0:
-	case 0x02e0:
-		/* BR02? NFI how these would be handled yet exactly */
-		break;
-	default:
-		switch (device->chipset) {
-		case 0xaa: break; /* reported broken, nv also disable it */
-		default:
-			pmc->use_msi = true;
+	if (nv_device_is_pci(device))
+		switch (device->pdev->device & 0x0ff0) {
+		case 0x00f0:
+		case 0x02e0:
+			/* BR02? NFI how these would be handled yet exactly */
 			break;
+		default:
+			switch (device->chipset) {
+			case 0xaa:
+				/* reported broken, nv also disable it */
+				break;
+			default:
+				pmc->use_msi = true;
+				break;
 		}
-	}
 
-	pmc->use_msi = nouveau_boolopt(device->cfgopt, "NvMSI", pmc->use_msi);
-	if (pmc->use_msi && oclass->msi_rearm) {
-		pmc->use_msi = pci_enable_msi(device->pdev) == 0;
-		if (pmc->use_msi) {
-			nv_info(pmc, "MSI interrupts enabled\n");
-			oclass->msi_rearm(pmc);
+		pmc->use_msi = nouveau_boolopt(device->cfgopt, "NvMSI",
+					       pmc->use_msi);
+
+		if (pmc->use_msi && oclass->msi_rearm) {
+			pmc->use_msi = pci_enable_msi(device->pdev) == 0;
+			if (pmc->use_msi) {
+				nv_info(pmc, "MSI interrupts enabled\n");
+				oclass->msi_rearm(pmc);
+			}
+		} else {
+			pmc->use_msi = false;
 		}
-	} else {
-		pmc->use_msi = false;
 	}
 
-	ret = request_irq(device->pdev->irq, nouveau_mc_intr,
-			  IRQF_SHARED, "nouveau", pmc);
+	ret = nv_device_get_irq(device, true);
+	if (ret < 0)
+		return ret;
+	pmc->irq = ret;
+
+	ret = request_irq(pmc->irq, nouveau_mc_intr, IRQF_SHARED, "nouveau",
+			  pmc);
+
 	if (ret < 0)
 		return ret;
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nv04.h b/drivers/gpu/drm/nouveau/core/subdev/mc/nv04.h
index b0d5c31606c..81a408e7d03 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mc/nv04.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nv04.h
@@ -14,6 +14,7 @@ int  nv04_mc_ctor(struct nouveau_object *, struct nouveau_object *,
 extern const struct nouveau_mc_intr nv04_mc_intr[];
 int  nv04_mc_init(struct nouveau_object *);
 void nv40_mc_msi_rearm(struct nouveau_mc *);
+int  nv44_mc_init(struct nouveau_object *object);
 int  nv50_mc_init(struct nouveau_object *);
 extern const struct nouveau_mc_intr nv50_mc_intr[];
 extern const struct nouveau_mc_intr nvc0_mc_intr[];
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nv44.c b/drivers/gpu/drm/nouveau/core/subdev/mc/nv44.c
index 3bfee5c6c4f..cc4d0d2d886 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mc/nv44.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nv44.c
@@ -24,7 +24,7 @@
 
 #include "nv04.h"
 
-static int
+int
 nv44_mc_init(struct nouveau_object *object)
 {
 	struct nv04_mc_priv *priv = (void *)object;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c b/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c
new file mode 100644
index 00000000000..a75c35ccf25
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nv4c.c
@@ -0,0 +1,45 @@
+/*
+ * Copyright 2014 Ilia Mirkin
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ilia Mirkin
+ */
+
+#include "nv04.h"
+
+static void
+nv4c_mc_msi_rearm(struct nouveau_mc *pmc)
+{
+	struct nv04_mc_priv *priv = (void *)pmc;
+	nv_wr08(priv, 0x088050, 0xff);
+}
+
+struct nouveau_oclass *
+nv4c_mc_oclass = &(struct nouveau_mc_oclass) {
+	.base.handle = NV_SUBDEV(MC, 0x4c),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_mc_ctor,
+		.dtor = _nouveau_mc_dtor,
+		.init = nv44_mc_init,
+		.fini = _nouveau_mc_fini,
+	},
+	.intr = nv04_mc_intr,
+	.msi_rearm = nv4c_mc_msi_rearm,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/mc/nv50.c
index e8822a934c4..9ca93e2718f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mc/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nv50.c
@@ -26,6 +26,7 @@
 
 const struct nouveau_mc_intr
 nv50_mc_intr[] = {
+	{ 0x04000000, NVDEV_ENGINE_DISP },  /* DISP before FIFO, so pageflip-timestamping works! */
 	{ 0x00000001, NVDEV_ENGINE_MPEG },
 	{ 0x00000100, NVDEV_ENGINE_FIFO },
 	{ 0x00001000, NVDEV_ENGINE_GR },
@@ -33,8 +34,8 @@ nv50_mc_intr[] = {
 	{ 0x00008000, NVDEV_ENGINE_BSP },	/* NV84- */
 	{ 0x00020000, NVDEV_ENGINE_VP },	/* NV84- */
 	{ 0x00100000, NVDEV_SUBDEV_TIMER },
-	{ 0x00200000, NVDEV_SUBDEV_GPIO },
-	{ 0x04000000, NVDEV_ENGINE_DISP },
+	{ 0x00200000, NVDEV_SUBDEV_GPIO },	/* PMGR->GPIO */
+	{ 0x00200000, NVDEV_SUBDEV_I2C }, 	/* PMGR->I2C/AUX */
 	{ 0x10000000, NVDEV_SUBDEV_BUS },
 	{ 0x80000000, NVDEV_ENGINE_SW },
 	{ 0x0002d101, NVDEV_SUBDEV_FB },
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nv98.c b/drivers/gpu/drm/nouveau/core/subdev/mc/nv98.c
index f8a6f18e2d3..3c76d9038f3 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mc/nv98.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nv98.c
@@ -26,6 +26,7 @@
 
 static const struct nouveau_mc_intr
 nv98_mc_intr[] = {
+	{ 0x04000000, NVDEV_ENGINE_DISP },  /* DISP first, so pageflip timestamps work */
 	{ 0x00000001, NVDEV_ENGINE_PPP },
 	{ 0x00000100, NVDEV_ENGINE_FIFO },
 	{ 0x00001000, NVDEV_ENGINE_GR },
@@ -35,9 +36,9 @@ nv98_mc_intr[] = {
 	{ 0x00040000, NVDEV_SUBDEV_PWR },	/* NVA3:NVC0 */
 	{ 0x00080000, NVDEV_SUBDEV_THERM },	/* NVA3:NVC0 */
 	{ 0x00100000, NVDEV_SUBDEV_TIMER },
-	{ 0x00200000, NVDEV_SUBDEV_GPIO },
+	{ 0x00200000, NVDEV_SUBDEV_GPIO },	/* PMGR->GPIO */
+	{ 0x00200000, NVDEV_SUBDEV_I2C }, 	/* PMGR->I2C/AUX */
 	{ 0x00400000, NVDEV_ENGINE_COPY0 },	/* NVA3-     */
-	{ 0x04000000, NVDEV_ENGINE_DISP },
 	{ 0x10000000, NVDEV_SUBDEV_BUS },
 	{ 0x80000000, NVDEV_ENGINE_SW },
 	{ 0x0042d101, NVDEV_SUBDEV_FB },
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/mc/nvc0.c
index c02b4763a2d..f9c6a678b47 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mc/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nvc0.c
@@ -26,20 +26,23 @@
 
 const struct nouveau_mc_intr
 nvc0_mc_intr[] = {
+	{ 0x04000000, NVDEV_ENGINE_DISP },  /* DISP first, so pageflip timestamps work. */
 	{ 0x00000001, NVDEV_ENGINE_PPP },
 	{ 0x00000020, NVDEV_ENGINE_COPY0 },
 	{ 0x00000040, NVDEV_ENGINE_COPY1 },
 	{ 0x00000080, NVDEV_ENGINE_COPY2 },
 	{ 0x00000100, NVDEV_ENGINE_FIFO },
 	{ 0x00001000, NVDEV_ENGINE_GR },
+	{ 0x00002000, NVDEV_SUBDEV_FB },
 	{ 0x00008000, NVDEV_ENGINE_BSP },
 	{ 0x00040000, NVDEV_SUBDEV_THERM },
 	{ 0x00020000, NVDEV_ENGINE_VP },
 	{ 0x00100000, NVDEV_SUBDEV_TIMER },
-	{ 0x00200000, NVDEV_SUBDEV_GPIO },
+	{ 0x00200000, NVDEV_SUBDEV_GPIO },	/* PMGR->GPIO */
+	{ 0x00200000, NVDEV_SUBDEV_I2C },	/* PMGR->I2C/AUX */
 	{ 0x01000000, NVDEV_SUBDEV_PWR },
 	{ 0x02000000, NVDEV_SUBDEV_LTCG },
-	{ 0x04000000, NVDEV_ENGINE_DISP },
+	{ 0x08000000, NVDEV_SUBDEV_FB },
 	{ 0x10000000, NVDEV_SUBDEV_BUS },
 	{ 0x40000000, NVDEV_SUBDEV_IBUS },
 	{ 0x80000000, NVDEV_ENGINE_SW },
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mxm/base.c b/drivers/gpu/drm/nouveau/core/subdev/mxm/base.c
index 129120473f6..51fcf796041 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mxm/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/mxm/base.c
@@ -87,55 +87,39 @@ mxm_shadow_dsm(struct nouveau_mxm *mxm, u8 version)
 		0xB8, 0x9C, 0x79, 0xB6, 0x2F, 0xD5, 0x56, 0x65
 	};
 	u32 mxms_args[] = { 0x00000000 };
-	union acpi_object args[4] = {
-		/* _DSM MUID */
-		{ .buffer.type = 3,
-		  .buffer.length = sizeof(muid),
-		  .buffer.pointer = muid,
-		},
-		/* spec says this can be zero to mean "highest revision", but
-		 * of course there's at least one bios out there which fails
-		 * unless you pass in exactly the version it supports..
-		 */
-		{ .integer.type = ACPI_TYPE_INTEGER,
-		  .integer.value = (version & 0xf0) << 4 | (version & 0x0f),
-		},
-		/* MXMS function */
-		{ .integer.type = ACPI_TYPE_INTEGER,
-		  .integer.value = 0x00000010,
-		},
-		/* Pointer to MXMS arguments */
-		{ .buffer.type = ACPI_TYPE_BUFFER,
-		  .buffer.length = sizeof(mxms_args),
-		  .buffer.pointer = (char *)mxms_args,
-		},
+	union acpi_object argv4 = {
+		.buffer.type = ACPI_TYPE_BUFFER,
+		.buffer.length = sizeof(mxms_args),
+		.buffer.pointer = (char *)mxms_args,
 	};
-	struct acpi_object_list list = { ARRAY_SIZE(args), args };
-	struct acpi_buffer retn = { ACPI_ALLOCATE_BUFFER, NULL };
 	union acpi_object *obj;
 	acpi_handle handle;
-	int ret;
+	int rev;
 
-	handle = ACPI_HANDLE(&device->pdev->dev);
+	handle = ACPI_HANDLE(nv_device_base(device));
 	if (!handle)
 		return false;
 
-	ret = acpi_evaluate_object(handle, "_DSM", &list, &retn);
-	if (ret) {
-		nv_debug(mxm, "DSM MXMS failed: %d\n", ret);
+	/*
+	 * spec says this can be zero to mean "highest revision", but
+	 * of course there's at least one bios out there which fails
+	 * unless you pass in exactly the version it supports..
+	 */
+	rev = (version & 0xf0) << 4 | (version & 0x0f);
+	obj = acpi_evaluate_dsm(handle, muid, rev, 0x00000010, &argv4);
+	if (!obj) {
+		nv_debug(mxm, "DSM MXMS failed\n");
 		return false;
 	}
 
-	obj = retn.pointer;
 	if (obj->type == ACPI_TYPE_BUFFER) {
 		mxm->mxms = kmemdup(obj->buffer.pointer,
 					 obj->buffer.length, GFP_KERNEL);
-	} else
-	if (obj->type == ACPI_TYPE_INTEGER) {
+	} else if (obj->type == ACPI_TYPE_INTEGER) {
 		nv_debug(mxm, "DSM MXMS returned 0x%llx\n", obj->integer.value);
 	}
 
-	kfree(obj);
+	ACPI_FREE(obj);
 	return mxm->mxms != NULL;
 }
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mxm/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/mxm/nv50.c
index 64f8b4702bf..fcaabe8456e 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mxm/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/mxm/nv50.c
@@ -150,7 +150,7 @@ mxm_dcb_sanitise_entry(struct nouveau_bios *bios, void *data, int idx, u16 pdcb)
 	 * common example is DP->eDP.
 	 */
 	conn  = bios->data;
-	conn += dcb_conn(bios, (ctx.outp[0] & 0x0000f000) >> 12, &ver, &len);
+	conn += nvbios_connEe(bios, (ctx.outp[0] & 0x0000f000) >> 12, &ver, &len);
 	type  = conn[0];
 	switch (ctx.desc.conn_type) {
 	case 0x01: /* LVDS */
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/host.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/host.fuc
index 2284ecb1c9b..c2bb616a8da 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/host.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/host.fuc
@@ -83,7 +83,7 @@ host_send:
 		// increment GET
 		add b32 $r1 0x1
 		and $r14 $r1 #fifo_qmaskf
-		nv_iowr(NV_PPWR_FIFO_GET(0), $r1)
+		nv_iowr(NV_PPWR_FIFO_GET(0), $r14)
 		bra #host_send
 	host_send_done:
 	ret
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/i2c_.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/i2c_.fuc
new file mode 100644
index 00000000000..757dda70002
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/i2c_.fuc
@@ -0,0 +1,393 @@
+/*
+ * Copyright 2013 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#define T_TIMEOUT  2200000
+#define T_RISEFALL 1000
+#define T_HOLD     5000
+
+#ifdef INCLUDE_PROC
+process(PROC_I2C_, #i2c_init, #i2c_recv)
+#endif
+
+/******************************************************************************
+ * I2C_ data segment
+ *****************************************************************************/
+#ifdef INCLUDE_DATA
+i2c_scl_map:
+.b32 NV_PPWR_OUTPUT_I2C_0_SCL
+.b32 NV_PPWR_OUTPUT_I2C_1_SCL
+.b32 NV_PPWR_OUTPUT_I2C_2_SCL
+.b32 NV_PPWR_OUTPUT_I2C_3_SCL
+.b32 NV_PPWR_OUTPUT_I2C_4_SCL
+.b32 NV_PPWR_OUTPUT_I2C_5_SCL
+.b32 NV_PPWR_OUTPUT_I2C_6_SCL
+.b32 NV_PPWR_OUTPUT_I2C_7_SCL
+.b32 NV_PPWR_OUTPUT_I2C_8_SCL
+.b32 NV_PPWR_OUTPUT_I2C_9_SCL
+i2c_sda_map:
+.b32 NV_PPWR_OUTPUT_I2C_0_SDA
+.b32 NV_PPWR_OUTPUT_I2C_1_SDA
+.b32 NV_PPWR_OUTPUT_I2C_2_SDA
+.b32 NV_PPWR_OUTPUT_I2C_3_SDA
+.b32 NV_PPWR_OUTPUT_I2C_4_SDA
+.b32 NV_PPWR_OUTPUT_I2C_5_SDA
+.b32 NV_PPWR_OUTPUT_I2C_6_SDA
+.b32 NV_PPWR_OUTPUT_I2C_7_SDA
+.b32 NV_PPWR_OUTPUT_I2C_8_SDA
+.b32 NV_PPWR_OUTPUT_I2C_9_SDA
+#if NVKM_PPWR_CHIPSET < GF119
+i2c_ctrl:
+.b32 0x00e138
+.b32 0x00e150
+.b32 0x00e168
+.b32 0x00e180
+.b32 0x00e254
+.b32 0x00e274
+.b32 0x00e764
+.b32 0x00e780
+.b32 0x00e79c
+.b32 0x00e7b8
+#endif
+#endif
+
+/******************************************************************************
+ * I2C_ code segment
+ *****************************************************************************/
+#ifdef INCLUDE_CODE
+
+// $r3  - value
+// $r2  - sda line
+// $r1  - scl line
+// $r0  - zero
+i2c_drive_scl:
+	cmp b32 $r3 0
+	bra e #i2c_drive_scl_lo
+	nv_iowr(NV_PPWR_OUTPUT_SET, $r1)
+	ret
+	i2c_drive_scl_lo:
+	nv_iowr(NV_PPWR_OUTPUT_CLR, $r1)
+	ret
+
+i2c_drive_sda:
+	cmp b32 $r3 0
+	bra e #i2c_drive_sda_lo
+	nv_iowr(NV_PPWR_OUTPUT_SET, $r2)
+	ret
+	i2c_drive_sda_lo:
+	nv_iowr(NV_PPWR_OUTPUT_CLR, $r2)
+	ret
+
+i2c_sense_scl:
+	bclr $flags $p1
+	nv_iord($r3, NV_PPWR_INPUT)
+	and $r3 $r1
+	bra z #i2c_sense_scl_done
+		bset $flags $p1
+	i2c_sense_scl_done:
+	ret
+
+i2c_sense_sda:
+	bclr $flags $p1
+	nv_iord($r3, NV_PPWR_INPUT)
+	and $r3 $r2
+	bra z #i2c_sense_sda_done
+		bset $flags $p1
+	i2c_sense_sda_done:
+	ret
+
+#define i2c_drive_scl(v) /*
+*/	mov $r3 (v) /*
+*/	call(i2c_drive_scl)
+#define i2c_drive_sda(v) /*
+*/	mov $r3 (v) /*
+*/	call(i2c_drive_sda)
+#define i2c_sense_scl() /*
+*/	call(i2c_sense_scl)
+#define i2c_sense_sda() /*
+*/	call(i2c_sense_sda)
+#define i2c_delay(v) /*
+*/	mov $r14 (v) /*
+*/	call(nsec)
+
+#define i2c_trace_init() /*
+*/	imm32($r6, 0x10000000) /*
+*/	sub b32 $r7 $r6 1 /*
+*/
+#define i2c_trace_down() /*
+*/	shr b32 $r6 4 /*
+*/	push $r5 /*
+*/	shl b32 $r5 $r6 4 /*
+*/	sub b32 $r5 $r6 /*
+*/	not b32 $r5 /*
+*/	and $r7 $r5 /*
+*/	pop $r5 /*
+*/
+#define i2c_trace_exit() /*
+*/	shl b32 $r6 4 /*
+*/
+#define i2c_trace_next() /*
+*/	add b32 $r7 $r6 /*
+*/
+#define i2c_trace_call(func) /*
+*/	i2c_trace_next() /*
+*/	i2c_trace_down() /*
+*/	call(func) /*
+*/	i2c_trace_exit() /*
+*/
+
+i2c_raise_scl:
+	push $r4
+	mov $r4 (T_TIMEOUT / T_RISEFALL)
+	i2c_drive_scl(1)
+	i2c_raise_scl_wait:
+		i2c_delay(T_RISEFALL)
+		i2c_sense_scl()
+		bra $p1 #i2c_raise_scl_done
+		sub b32 $r4 1
+		bra nz #i2c_raise_scl_wait
+	i2c_raise_scl_done:
+	pop $r4
+	ret
+
+i2c_start:
+	i2c_sense_scl()
+	bra not $p1 #i2c_start_rep
+	i2c_sense_sda()
+	bra not $p1 #i2c_start_rep
+	bra #i2c_start_send
+	i2c_start_rep:
+		i2c_drive_scl(0)
+		i2c_drive_sda(1)
+		i2c_trace_call(i2c_raise_scl)
+		bra not $p1 #i2c_start_out
+	i2c_start_send:
+	i2c_drive_sda(0)
+	i2c_delay(T_HOLD)
+	i2c_drive_scl(0)
+	i2c_delay(T_HOLD)
+	i2c_start_out:
+	ret
+
+i2c_stop:
+	i2c_drive_scl(0)
+	i2c_drive_sda(0)
+	i2c_delay(T_RISEFALL)
+	i2c_drive_scl(1)
+	i2c_delay(T_HOLD)
+	i2c_drive_sda(1)
+	i2c_delay(T_HOLD)
+	ret
+
+// $r3  - value
+// $r2  - sda line
+// $r1  - scl line
+// $r0  - zero
+i2c_bitw:
+	call(i2c_drive_sda)
+	i2c_delay(T_RISEFALL)
+	i2c_trace_call(i2c_raise_scl)
+	bra not $p1 #i2c_bitw_out
+	i2c_delay(T_HOLD)
+	i2c_drive_scl(0)
+	i2c_delay(T_HOLD)
+	i2c_bitw_out:
+	ret
+
+// $r3  - value (out)
+// $r2  - sda line
+// $r1  - scl line
+// $r0  - zero
+i2c_bitr:
+	i2c_drive_sda(1)
+	i2c_delay(T_RISEFALL)
+	i2c_trace_call(i2c_raise_scl)
+	bra not $p1 #i2c_bitr_done
+	i2c_sense_sda()
+	i2c_drive_scl(0)
+	i2c_delay(T_HOLD)
+	xbit $r3 $flags $p1
+	bset $flags $p1
+	i2c_bitr_done:
+	ret
+
+i2c_get_byte:
+	mov $r5 0
+	mov $r4 8
+	i2c_get_byte_next:
+		shl b32 $r5 1
+		i2c_trace_call(i2c_bitr)
+		bra not $p1 #i2c_get_byte_done
+		or $r5 $r3
+		sub b32 $r4 1
+		bra nz #i2c_get_byte_next
+	mov $r3 1
+	i2c_trace_call(i2c_bitw)
+	i2c_get_byte_done:
+	ret
+
+i2c_put_byte:
+	mov $r4 8
+	i2c_put_byte_next:
+		sub b32 $r4 1
+		xbit $r3 $r5 $r4
+		i2c_trace_call(i2c_bitw)
+		bra not $p1 #i2c_put_byte_done
+		cmp b32 $r4 0
+		bra ne #i2c_put_byte_next
+	i2c_trace_call(i2c_bitr)
+	bra not $p1 #i2c_put_byte_done
+	i2c_trace_next()
+	cmp b32 $r3 1
+	bra ne #i2c_put_byte_done
+	bclr $flags $p1	// nack
+	i2c_put_byte_done:
+	ret
+
+i2c_addr:
+	i2c_trace_call(i2c_start)
+	bra not $p1 #i2c_addr_done
+	extr $r3 $r12 I2C__MSG_DATA0_ADDR
+	shl b32 $r3 1
+	or $r5 $r3
+	i2c_trace_call(i2c_put_byte)
+	i2c_addr_done:
+	ret
+
+i2c_acquire_addr:
+	extr $r14 $r12 I2C__MSG_DATA0_PORT
+#if NVKM_PPWR_CHIPSET < GF119
+	shl b32 $r14 2
+	add b32 $r14 #i2c_ctrl
+	ld b32 $r14 D[$r14]
+#else
+	shl b32 $r14 5
+	add b32 $r14 0x00d014
+#endif
+	ret
+
+i2c_acquire:
+	call(i2c_acquire_addr)
+	call(rd32)
+	bset $r13 3
+	call(wr32)
+	ret
+
+i2c_release:
+	call(i2c_acquire_addr)
+	call(rd32)
+	bclr $r13 3
+	call(wr32)
+	ret
+
+// description
+//
+// $r15 - current (i2c)
+// $r14 - sender process name
+// $r13 - message
+// $r12 - data0
+// $r11 - data1
+// $r0  - zero
+i2c_recv:
+	bclr $flags $p1
+	extr $r1 $r12 I2C__MSG_DATA0_PORT
+	shl b32 $r1 2
+	cmp b32 $r1 (#i2c_sda_map - #i2c_scl_map)
+	bra ge #i2c_recv_done
+	add b32 $r3 $r1 #i2c_sda_map
+	ld b32 $r2 D[$r3]
+	add b32 $r3 $r1 #i2c_scl_map
+	ld b32 $r1 D[$r3]
+
+	bset $flags $p2
+	push $r13
+	push $r14
+
+	push $r13
+	i2c_trace_init()
+	i2c_trace_call(i2c_acquire)
+	pop $r13
+
+	cmp b32 $r13 I2C__MSG_RD08
+	bra ne #i2c_recv_not_rd08
+		mov $r5 0
+		i2c_trace_call(i2c_addr)
+		bra not $p1 #i2c_recv_done
+		extr $r5 $r12 I2C__MSG_DATA0_RD08_REG
+		i2c_trace_call(i2c_put_byte)
+		bra not $p1 #i2c_recv_done
+		mov $r5 1
+		i2c_trace_call(i2c_addr)
+		bra not $p1 #i2c_recv_done
+		i2c_trace_call(i2c_get_byte)
+		bra not $p1 #i2c_recv_done
+		ins $r11 $r5 I2C__MSG_DATA1_RD08_VAL
+		i2c_trace_call(i2c_stop)
+		mov b32 $r11 $r5
+		clear b32 $r7
+		bra #i2c_recv_done
+
+	i2c_recv_not_rd08:
+	cmp b32 $r13 I2C__MSG_WR08
+	bra ne #i2c_recv_not_wr08
+		mov $r5 0
+		call(i2c_addr)
+		bra not $p1 #i2c_recv_done
+		extr $r5 $r12 I2C__MSG_DATA0_WR08_REG
+		call(i2c_put_byte)
+		bra not $p1 #i2c_recv_done
+		mov $r5 0
+		call(i2c_addr)
+		bra not $p1 #i2c_recv_done
+		extr $r5 $r11 I2C__MSG_DATA1_WR08_VAL
+		call(i2c_put_byte)
+		bra not $p1 #i2c_recv_done
+		call(i2c_stop)
+		clear b32 $r7
+		extr $r5 $r12 I2C__MSG_DATA0_WR08_SYNC
+		bra nz #i2c_recv_done
+		bclr $flags $p2
+		bra #i2c_recv_done
+
+	i2c_recv_not_wr08:
+
+	i2c_recv_done:
+	extr $r14 $r12 I2C__MSG_DATA0_PORT
+	call(i2c_release)
+
+	pop $r14
+	pop $r13
+	bra not $p2 #i2c_recv_exit
+	mov b32 $r12 $r7
+	call(send)
+
+	i2c_recv_exit:
+	ret
+
+// description
+//
+// $r15 - current (i2c)
+// $r0  - zero
+i2c_init:
+	ret
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
index 0a7b05fa5c1..8f29badd785 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
@@ -51,12 +51,12 @@ time_next: .b32 0
 // $r0  - zero
 rd32:
 	nv_iowr(NV_PPWR_MMIO_ADDR, $r14)
-	mov $r14 NV_PPWR_MMIO_CTRL_OP_RD
-	sethi $r14 NV_PPWR_MMIO_CTRL_TRIGGER
-	nv_iowr(NV_PPWR_MMIO_CTRL, $r14)
+	mov $r13 NV_PPWR_MMIO_CTRL_OP_RD
+	sethi $r13 NV_PPWR_MMIO_CTRL_TRIGGER
+	nv_iowr(NV_PPWR_MMIO_CTRL, $r13)
 	rd32_wait:
-		nv_iord($r14, NV_PPWR_MMIO_CTRL)
-		and $r14 NV_PPWR_MMIO_CTRL_STATUS
+		nv_iord($r13, NV_PPWR_MMIO_CTRL)
+		and $r13 NV_PPWR_MMIO_CTRL_STATUS
 		bra nz #rd32_wait
 	nv_iord($r13, NV_PPWR_MMIO_DATA)
 	ret
@@ -70,23 +70,25 @@ rd32:
 wr32:
 	nv_iowr(NV_PPWR_MMIO_ADDR, $r14)
 	nv_iowr(NV_PPWR_MMIO_DATA, $r13)
-	mov $r14 NV_PPWR_MMIO_CTRL_OP_WR
-	or $r14 NV_PPWR_MMIO_CTRL_MASK_B32_0
-	sethi $r14 NV_PPWR_MMIO_CTRL_TRIGGER
+	mov $r13 NV_PPWR_MMIO_CTRL_OP_WR
+	or $r13 NV_PPWR_MMIO_CTRL_MASK_B32_0
+	sethi $r13 NV_PPWR_MMIO_CTRL_TRIGGER
 
 #ifdef NVKM_FALCON_MMIO_TRAP
-	mov $r8 NV_PPWR_INTR_TRIGGER_USER1
-	nv_iowr(NV_PPWR_INTR_TRIGGER, $r8)
+	push $r13
+	mov $r13 NV_PPWR_INTR_TRIGGER_USER1
+	nv_iowr(NV_PPWR_INTR_TRIGGER, $r13)
 	wr32_host:
-		nv_iord($r8, NV_PPWR_INTR)
-		and $r8 NV_PPWR_INTR_USER1
+		nv_iord($r13, NV_PPWR_INTR)
+		and $r13 NV_PPWR_INTR_USER1
 		bra nz #wr32_host
+	pop $r13
 #endif
 
-	nv_iowr(NV_PPWR_MMIO_CTRL, $r14)
+	nv_iowr(NV_PPWR_MMIO_CTRL, $r13)
 	wr32_wait:
-		nv_iord($r14, NV_PPWR_MMIO_CTRL)
-		and $r14 NV_PPWR_MMIO_CTRL_STATUS
+		nv_iord($r13, NV_PPWR_MMIO_CTRL)
+		and $r13 NV_PPWR_MMIO_CTRL_STATUS
 		bra nz #wr32_wait
 	ret
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
index 2a74ea90760..e2a63ac5422 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
@@ -83,6 +83,50 @@
 #define NV_PPWR_MMIO_CTRL_OP_WR                                      0x00000002
 #define NV_PPWR_OUTPUT                                                   0x07c0
 #define NV_PPWR_OUTPUT_FB_PAUSE                                      0x00000004
+#if NVKM_PPWR_CHIPSET < GF119
+#define NV_PPWR_OUTPUT_I2C_3_SCL                                     0x00000100
+#define NV_PPWR_OUTPUT_I2C_3_SDA                                     0x00000200
+#define NV_PPWR_OUTPUT_I2C_0_SCL                                     0x00001000
+#define NV_PPWR_OUTPUT_I2C_0_SDA                                     0x00002000
+#define NV_PPWR_OUTPUT_I2C_1_SCL                                     0x00004000
+#define NV_PPWR_OUTPUT_I2C_1_SDA                                     0x00008000
+#define NV_PPWR_OUTPUT_I2C_2_SCL                                     0x00010000
+#define NV_PPWR_OUTPUT_I2C_2_SDA                                     0x00020000
+#define NV_PPWR_OUTPUT_I2C_4_SCL                                     0x00040000
+#define NV_PPWR_OUTPUT_I2C_4_SDA                                     0x00080000
+#define NV_PPWR_OUTPUT_I2C_5_SCL                                     0x00100000
+#define NV_PPWR_OUTPUT_I2C_5_SDA                                     0x00200000
+#define NV_PPWR_OUTPUT_I2C_6_SCL                                     0x00400000
+#define NV_PPWR_OUTPUT_I2C_6_SDA                                     0x00800000
+#define NV_PPWR_OUTPUT_I2C_7_SCL                                     0x01000000
+#define NV_PPWR_OUTPUT_I2C_7_SDA                                     0x02000000
+#define NV_PPWR_OUTPUT_I2C_8_SCL                                     0x04000000
+#define NV_PPWR_OUTPUT_I2C_8_SDA                                     0x08000000
+#define NV_PPWR_OUTPUT_I2C_9_SCL                                     0x10000000
+#define NV_PPWR_OUTPUT_I2C_9_SDA                                     0x20000000
+#else
+#define NV_PPWR_OUTPUT_I2C_0_SCL                                     0x00000400
+#define NV_PPWR_OUTPUT_I2C_1_SCL                                     0x00000800
+#define NV_PPWR_OUTPUT_I2C_2_SCL                                     0x00001000
+#define NV_PPWR_OUTPUT_I2C_3_SCL                                     0x00002000
+#define NV_PPWR_OUTPUT_I2C_4_SCL                                     0x00004000
+#define NV_PPWR_OUTPUT_I2C_5_SCL                                     0x00008000
+#define NV_PPWR_OUTPUT_I2C_6_SCL                                     0x00010000
+#define NV_PPWR_OUTPUT_I2C_7_SCL                                     0x00020000
+#define NV_PPWR_OUTPUT_I2C_8_SCL                                     0x00040000
+#define NV_PPWR_OUTPUT_I2C_9_SCL                                     0x00080000
+#define NV_PPWR_OUTPUT_I2C_0_SDA                                     0x00100000
+#define NV_PPWR_OUTPUT_I2C_1_SDA                                     0x00200000
+#define NV_PPWR_OUTPUT_I2C_2_SDA                                     0x00400000
+#define NV_PPWR_OUTPUT_I2C_3_SDA                                     0x00800000
+#define NV_PPWR_OUTPUT_I2C_4_SDA                                     0x01000000
+#define NV_PPWR_OUTPUT_I2C_5_SDA                                     0x02000000
+#define NV_PPWR_OUTPUT_I2C_6_SDA                                     0x04000000
+#define NV_PPWR_OUTPUT_I2C_7_SDA                                     0x08000000
+#define NV_PPWR_OUTPUT_I2C_8_SDA                                     0x10000000
+#define NV_PPWR_OUTPUT_I2C_9_SDA                                     0x20000000
+#endif
+#define NV_PPWR_INPUT                                                    0x07c4
 #define NV_PPWR_OUTPUT_SET                                               0x07e0
 #define NV_PPWR_OUTPUT_SET_FB_PAUSE                                  0x00000004
 #define NV_PPWR_OUTPUT_CLR                                               0x07e4
@@ -125,6 +169,15 @@
 */	.b32 0 /*
 */	.skip 64
 
+#if NV_PPWR_CHIPSET < GK208
+#define imm32(reg,val) /*
+*/	movw reg  ((val) & 0x0000ffff) /*
+*/	sethi reg ((val) & 0xffff0000)
+#else
+#define imm32(reg,val) /*
+*/	mov reg (val)
+#endif
+
 #ifndef NVKM_FALCON_UNSHIFTED_IO
 #define nv_iord(reg,ior) /*
 */	mov reg ior /*
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
index 947be536dae..17a8a383d91 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
@@ -37,6 +37,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_PROC
@@ -46,6 +47,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_DATA
@@ -57,6 +59,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h
index 9342e2d7d3b..39a5dc150a0 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc.h
@@ -89,33 +89,13 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x54534554,
-	0x00000494,
-	0x00000475,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
+	0x5f433249,
+	0x00000877,
+	0x0000071e,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x454c4449,
-	0x0000049f,
-	0x0000049d,
-	0x00000000,
-	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -131,14 +111,13 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x54534554,
+	0x00000898,
+	0x00000879,
 	0x00000000,
 	0x00000000,
-/* 0x0210: proc_list_tail */
-/* 0x0210: time_prev */
 	0x00000000,
-/* 0x0214: time_next */
 	0x00000000,
-/* 0x0218: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -154,6 +133,9 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x454c4449,
+	0x000008a3,
+	0x000008a1,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -171,11 +153,14 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0298: rfifo_queue */
 	0x00000000,
 	0x00000000,
+/* 0x0268: proc_list_tail */
+/* 0x0268: time_prev */
 	0x00000000,
+/* 0x026c: time_next */
 	0x00000000,
+/* 0x0270: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -204,31 +189,11 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0318: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x000003a9,
-/* 0x0324: memx_func_next */
-	0x00000001,
-	0x00000000,
-	0x000003c7,
-	0x00000002,
-	0x00000002,
-	0x000003df,
-	0x00040003,
-	0x00000000,
-	0x00000407,
-	0x00010004,
-	0x00000000,
-	0x00000421,
-/* 0x0354: memx_func_tail */
-/* 0x0354: memx_data_head */
-	0x00000000,
-	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x02f0: rfifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -261,10 +226,25 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0370: memx_func_head */
+	0x00010000,
 	0x00000000,
+	0x000003a9,
+/* 0x037c: memx_func_next */
+	0x00000001,
 	0x00000000,
+	0x000003c7,
+	0x00000002,
+	0x00000002,
+	0x000003df,
+	0x00040003,
 	0x00000000,
+	0x00000407,
+	0x00010004,
 	0x00000000,
+	0x00000421,
+/* 0x03ac: memx_func_tail */
+/* 0x03ac: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -735,7 +715,6 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0b54: memx_data_tail */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -778,6 +757,29 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bac: memx_data_tail */
+/* 0x0bac: i2c_scl_map */
+	0x00000400,
+	0x00000800,
+	0x00001000,
+	0x00002000,
+	0x00004000,
+	0x00008000,
+	0x00010000,
+	0x00020000,
+	0x00040000,
+	0x00080000,
+/* 0x0bd4: i2c_sda_map */
+	0x00100000,
+	0x00200000,
+	0x00400000,
+	0x00800000,
+	0x01000000,
+	0x02000000,
+	0x04000000,
+	0x08000000,
+	0x10000000,
+	0x20000000,
 	0x00000000,
 };
 
@@ -786,13 +788,13 @@ uint32_t nv108_pwr_code[] = {
 /* 0x0004: rd32 */
 	0xf607a040,
 	0x04bd000e,
-	0xe3f0010e,
+	0xd3f0010d,
 	0x07ac4001,
-	0xbd000ef6,
+	0xbd000df6,
 /* 0x0019: rd32_wait */
-	0x07ac4e04,
-	0xf100eecf,
-	0xf47000e4,
+	0x07ac4d04,
+	0xf100ddcf,
+	0xf47000d4,
 	0xa44df61b,
 	0x00ddcf07,
 /* 0x002e: wr32 */
@@ -800,14 +802,14 @@ uint32_t nv108_pwr_code[] = {
 	0x000ef607,
 	0xa44004bd,
 	0x000df607,
-	0x020e04bd,
-	0xf0f0e5f0,
-	0xac4001e3,
-	0x000ef607,
+	0x020d04bd,
+	0xf0f0d5f0,
+	0xac4001d3,
+	0x000df607,
 /* 0x004e: wr32_wait */
-	0xac4e04bd,
-	0x00eecf07,
-	0x7000e4f1,
+	0xac4d04bd,
+	0x00ddcf07,
+	0x7000d4f1,
 	0xf8f61bf4,
 /* 0x005d: nsec */
 	0xcf2c0800,
@@ -832,20 +834,20 @@ uint32_t nv108_pwr_code[] = {
 	0x03e99800,
 	0xf40096b0,
 	0x0a98280b,
-	0x029abb84,
+	0x029abb9a,
 	0x0d0e1cf4,
 	0x01de7e01,
 	0xf494bd00,
 /* 0x00b2: intr_watchdog_next_time */
 	0x0a98140e,
-	0x00a6b085,
+	0x00a6b09b,
 	0xa6080bf4,
 	0x061cf49a,
 /* 0x00c0: intr_watchdog_next_time_set */
 /* 0x00c3: intr_watchdog_next_proc */
-	0xb58509b5,
+	0xb59b09b5,
 	0xe0b603e9,
-	0x10e6b158,
+	0x68e6b158,
 	0xc81bf402,
 /* 0x00d2: intr */
 	0x00f900f8,
@@ -862,15 +864,15 @@ uint32_t nv108_pwr_code[] = {
 	0x080804bd,
 	0xc40088cf,
 	0x0bf40289,
-	0x8500b51f,
+	0x9b00b51f,
 	0x957e580e,
 	0x09980000,
-	0x0096b085,
+	0x0096b09b,
 	0x000d0bf4,
 	0x0009f634,
 	0x09b504bd,
 /* 0x0125: intr_skip_watchdog */
-	0x0089e484,
+	0x0089e49a,
 	0x360bf408,
 	0xcf068849,
 	0x9ac40099,
@@ -918,7 +920,7 @@ uint32_t nv108_pwr_code[] = {
 /* 0x01c6: timer_reset */
 	0x3400161e,
 	0xbd000ef6,
-	0x840eb504,
+	0x9a0eb504,
 /* 0x01d0: timer_enable */
 	0x38000108,
 	0xbd0008f6,
@@ -949,7 +951,7 @@ uint32_t nv108_pwr_code[] = {
 	0xa6008a98,
 	0x100bf4ae,
 	0xb15880b6,
-	0xf4021086,
+	0xf4026886,
 	0x32f4f11b,
 /* 0x0239: find_done */
 	0xfc8eb201,
@@ -1009,14 +1011,14 @@ uint32_t nv108_pwr_code[] = {
 	0x0bf412a6,
 	0x071ec42e,
 	0xb704ee94,
-	0x980218e0,
+	0x980270e0,
 	0xec9803eb,
 	0x01ed9802,
 	0x7e00ee98,
 	0xb600023f,
 	0x1ec40110,
 	0x04b0400f,
-	0xbd0001f6,
+	0xbd000ef6,
 	0xc70ef404,
 /* 0x0328: host_send_done */
 /* 0x032a: host_recv */
@@ -1031,7 +1033,7 @@ uint32_t nv108_pwr_code[] = {
 	0xf412a608,
 	0x23c4ef0b,
 	0x0434b607,
-	0x029830b7,
+	0x02f030b7,
 	0xb5033bb5,
 	0x3db5023c,
 	0x003eb501,
@@ -1044,11 +1046,11 @@ uint32_t nv108_pwr_code[] = {
 /* 0x0379: host_init */
 	0x00804100,
 	0xf11014b6,
-	0x40021815,
+	0x40027015,
 	0x01f604d0,
 	0x4104bd00,
 	0x14b60080,
-	0x9815f110,
+	0xf015f110,
 	0x04dc4002,
 	0xbd0001f6,
 	0x40010104,
@@ -1101,13 +1103,13 @@ uint32_t nv108_pwr_code[] = {
 	0x001398b2,
 	0x950410b6,
 	0x30f01034,
-	0xc835980c,
+	0xde35980c,
 	0x12a655f9,
 	0xfced1ef4,
 	0x7ee0fcd0,
 	0xf800023f,
 /* 0x0455: memx_info */
-	0x03544c00,
+	0x03ac4c00,
 	0x7e08004b,
 	0xf800023f,
 /* 0x0461: memx_recv */
@@ -1119,7 +1121,301 @@ uint32_t nv108_pwr_code[] = {
 /* 0x0471: perf_recv */
 /* 0x0473: perf_init */
 	0xf800f800,
-/* 0x0475: test_recv */
+/* 0x0475: i2c_drive_scl */
+	0x0036b000,
+	0x400d0bf4,
+	0x01f607e0,
+	0xf804bd00,
+/* 0x0485: i2c_drive_scl_lo */
+	0x07e44000,
+	0xbd0001f6,
+/* 0x048f: i2c_drive_sda */
+	0xb000f804,
+	0x0bf40036,
+	0x07e0400d,
+	0xbd0002f6,
+/* 0x049f: i2c_drive_sda_lo */
+	0x4000f804,
+	0x02f607e4,
+	0xf804bd00,
+/* 0x04a9: i2c_sense_scl */
+	0x0132f400,
+	0xcf07c443,
+	0x31fd0033,
+	0x060bf404,
+/* 0x04bb: i2c_sense_scl_done */
+	0xf80131f4,
+/* 0x04bd: i2c_sense_sda */
+	0x0132f400,
+	0xcf07c443,
+	0x32fd0033,
+	0x060bf404,
+/* 0x04cf: i2c_sense_sda_done */
+	0xf80131f4,
+/* 0x04d1: i2c_raise_scl */
+	0x4440f900,
+	0x01030898,
+	0x0004757e,
+/* 0x04dc: i2c_raise_scl_wait */
+	0x7e03e84e,
+	0x7e00005d,
+	0xf40004a9,
+	0x42b60901,
+	0xef1bf401,
+/* 0x04f0: i2c_raise_scl_done */
+	0x00f840fc,
+/* 0x04f4: i2c_start */
+	0x0004a97e,
+	0x7e0d11f4,
+	0xf40004bd,
+	0x0ef40611,
+/* 0x0505: i2c_start_rep */
+	0x7e00032e,
+	0x03000475,
+	0x048f7e01,
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0xd17e50fc,
+	0x64b60004,
+	0x1d11f404,
+/* 0x0530: i2c_start_send */
+	0x8f7e0003,
+	0x884e0004,
+	0x005d7e13,
+	0x7e000300,
+	0x4e000475,
+	0x5d7e1388,
+/* 0x054a: i2c_start_out */
+	0x00f80000,
+/* 0x054c: i2c_stop */
+	0x757e0003,
+	0x00030004,
+	0x00048f7e,
+	0x7e03e84e,
+	0x0300005d,
+	0x04757e01,
+	0x13884e00,
+	0x00005d7e,
+	0x8f7e0103,
+	0x884e0004,
+	0x005d7e13,
+/* 0x057b: i2c_bitw */
+	0x7e00f800,
+	0x4e00048f,
+	0x5d7e03e8,
+	0x76bb0000,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb60004d1,
+	0x11f40464,
+	0x13884e17,
+	0x00005d7e,
+	0x757e0003,
+	0x884e0004,
+	0x005d7e13,
+/* 0x05b9: i2c_bitw_out */
+/* 0x05bb: i2c_bitr */
+	0x0300f800,
+	0x048f7e01,
+	0x03e84e00,
+	0x00005d7e,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x04d17e50,
+	0x0464b600,
+	0x7e1a11f4,
+	0x030004bd,
+	0x04757e00,
+	0x13884e00,
+	0x00005d7e,
+	0xf4013cf0,
+/* 0x05fe: i2c_bitr_done */
+	0x00f80131,
+/* 0x0600: i2c_get_byte */
+	0x08040005,
+/* 0x0604: i2c_get_byte_next */
+	0xbb0154b6,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x0005bb7e,
+	0xf40464b6,
+	0x53fd2a11,
+	0x0142b605,
+	0x03d81bf4,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x7b7e50fc,
+	0x64b60005,
+/* 0x064d: i2c_get_byte_done */
+/* 0x064f: i2c_put_byte */
+	0x0400f804,
+/* 0x0651: i2c_put_byte_next */
+	0x0142b608,
+	0xbb3854ff,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x00057b7e,
+	0xf40464b6,
+	0x46b03411,
+	0xd81bf400,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x05bb7e50,
+	0x0464b600,
+	0xbb0f11f4,
+	0x36b00076,
+	0x061bf401,
+/* 0x06a7: i2c_put_byte_done */
+	0xf80132f4,
+/* 0x06a9: i2c_addr */
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0xf47e50fc,
+	0x64b60004,
+	0x2911f404,
+	0x012ec3e7,
+	0xfd0134b6,
+	0x76bb0553,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb600064f,
+/* 0x06ee: i2c_addr_done */
+	0x00f80464,
+/* 0x06f0: i2c_acquire_addr */
+	0xb6f8cec7,
+	0xe0b705e4,
+	0x00f8d014,
+/* 0x06fc: i2c_acquire */
+	0x0006f07e,
+	0x0000047e,
+	0x7e03d9f0,
+	0xf800002e,
+/* 0x070d: i2c_release */
+	0x06f07e00,
+	0x00047e00,
+	0x03daf000,
+	0x00002e7e,
+/* 0x071e: i2c_recv */
+	0x32f400f8,
+	0xf8c1c701,
+	0xb00214b6,
+	0x1ff52816,
+	0x13b80137,
+	0x98000bd4,
+	0x13b80032,
+	0x98000bac,
+	0x31f40031,
+	0xf9d0f902,
+	0xf1d0f9e0,
+	0xf1000067,
+	0x92100063,
+	0x76bb0167,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb60006fc,
+	0xd0fc0464,
+	0xf500d6b0,
+	0x0500b01b,
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0xa97e50fc,
+	0x64b60006,
+	0xcc11f504,
+	0xe0c5c700,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x064f7e50,
+	0x0464b600,
+	0x00a911f5,
+	0x76bb0105,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb60006a9,
+	0x11f50464,
+	0x76bb0087,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb6000600,
+	0x11f40464,
+	0xe05bcb67,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x054c7e50,
+	0x0464b600,
+	0x74bd5bb2,
+/* 0x0823: i2c_recv_not_rd08 */
+	0xb0410ef4,
+	0x1bf401d6,
+	0x7e00053b,
+	0xf40006a9,
+	0xc5c73211,
+	0x064f7ee0,
+	0x2811f400,
+	0xa97e0005,
+	0x11f40006,
+	0xe0b5c71f,
+	0x00064f7e,
+	0x7e1511f4,
+	0xbd00054c,
+	0x08c5c774,
+	0xf4091bf4,
+	0x0ef40232,
+/* 0x0861: i2c_recv_not_wr08 */
+/* 0x0861: i2c_recv_done */
+	0xf8cec703,
+	0x00070d7e,
+	0xd0fce0fc,
+	0xb20912f4,
+	0x023f7e7c,
+/* 0x0875: i2c_recv_exit */
+/* 0x0877: i2c_init */
+	0xf800f800,
+/* 0x0879: test_recv */
 	0x04584100,
 	0xb60011cf,
 	0x58400110,
@@ -1128,26 +1424,26 @@ uint32_t nv108_pwr_code[] = {
 	0xe3f1d900,
 	0x967e134f,
 	0x00f80001,
-/* 0x0494: test_init */
+/* 0x0898: test_init */
 	0x7e08004e,
 	0xf8000196,
-/* 0x049d: idle_recv */
-/* 0x049f: idle */
+/* 0x08a1: idle_recv */
+/* 0x08a3: idle */
 	0xf400f800,
 	0x54410031,
 	0x0011cf04,
 	0x400110b6,
 	0x01f60454,
-/* 0x04b3: idle_loop */
+/* 0x08b7: idle_loop */
 	0x0104bd00,
 	0x0232f458,
-/* 0x04b8: idle_proc */
-/* 0x04b8: idle_proc_exec */
+/* 0x08bc: idle_proc */
+/* 0x08bc: idle_proc_exec */
 	0x1eb210f9,
 	0x0002487e,
 	0x11f410fc,
 	0x0231f409,
-/* 0x04cb: idle_proc_next */
+/* 0x08cf: idle_proc_next */
 	0xb6f00ef4,
 	0x1fa65810,
 	0xf4e81bf4,
@@ -1161,5 +1457,4 @@ uint32_t nv108_pwr_code[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x00000000,
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
index 6fde0b89e5a..6744fcc0615 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
@@ -37,6 +37,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_PROC
@@ -46,6 +47,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_DATA
@@ -57,6 +59,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h
index 0fa4d7dcd40..254205cd516 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc.h
@@ -89,9 +89,31 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x5f433249,
+	0x00000982,
+	0x00000825,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x54534554,
-	0x0000057b,
-	0x00000554,
+	0x000009ab,
+	0x00000984,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +134,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x00000587,
-	0x00000585,
+	0x000009b7,
+	0x000009b5,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -133,12 +155,12 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0210: proc_list_tail */
-/* 0x0210: time_prev */
+/* 0x0268: proc_list_tail */
+/* 0x0268: time_prev */
 	0x00000000,
-/* 0x0214: time_next */
+/* 0x026c: time_next */
 	0x00000000,
-/* 0x0218: fifo_queue */
+/* 0x0270: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -171,7 +193,7 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0298: rfifo_queue */
+/* 0x02f0: rfifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -204,11 +226,11 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0318: memx_func_head */
+/* 0x0370: memx_func_head */
 	0x00010000,
 	0x00000000,
 	0x0000046f,
-/* 0x0324: memx_func_next */
+/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
 	0x00000496,
@@ -221,8 +243,19 @@ uint32_t nva3_pwr_data[] = {
 	0x00010004,
 	0x00000000,
 	0x000004fc,
-/* 0x0354: memx_func_tail */
-/* 0x0354: memx_data_head */
+/* 0x03ac: memx_func_tail */
+/* 0x03ac: memx_data_head */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -724,6 +757,41 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bac: memx_data_tail */
+/* 0x0bac: i2c_scl_map */
+	0x00001000,
+	0x00004000,
+	0x00010000,
+	0x00000100,
+	0x00040000,
+	0x00100000,
+	0x00400000,
+	0x01000000,
+	0x04000000,
+	0x10000000,
+/* 0x0bd4: i2c_sda_map */
+	0x00002000,
+	0x00008000,
+	0x00020000,
+	0x00000200,
+	0x00080000,
+	0x00200000,
+	0x00800000,
+	0x02000000,
+	0x08000000,
+	0x20000000,
+/* 0x0bfc: i2c_ctrl */
+	0x0000e138,
+	0x0000e150,
+	0x0000e168,
+	0x0000e180,
+	0x0000e254,
+	0x0000e274,
+	0x0000e764,
+	0x0000e780,
+	0x0000e79c,
+	0x0000e7b8,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -735,7 +803,6 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0b54: memx_data_tail */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -787,15 +854,15 @@ uint32_t nva3_pwr_code[] = {
 	0x07a007f1,
 	0xd00604b6,
 	0x04bd000e,
-	0xf001e7f0,
-	0x07f101e3,
+	0xf001d7f0,
+	0x07f101d3,
 	0x04b607ac,
-	0x000ed006,
+	0x000dd006,
 /* 0x0022: rd32_wait */
-	0xe7f104bd,
-	0xe4b607ac,
-	0x00eecf06,
-	0x7000e4f1,
+	0xd7f104bd,
+	0xd4b607ac,
+	0x00ddcf06,
+	0x7000d4f1,
 	0xf1f21bf4,
 	0xb607a4d7,
 	0xddcf06d4,
@@ -807,15 +874,15 @@ uint32_t nva3_pwr_code[] = {
 	0xb607a407,
 	0x0dd00604,
 	0xf004bd00,
-	0xe5f002e7,
-	0x01e3f0f0,
+	0xd5f002d7,
+	0x01d3f0f0,
 	0x07ac07f1,
 	0xd00604b6,
-	0x04bd000e,
+	0x04bd000d,
 /* 0x006c: wr32_wait */
-	0x07ace7f1,
-	0xcf06e4b6,
-	0xe4f100ee,
+	0x07acd7f1,
+	0xcf06d4b6,
+	0xd4f100dd,
 	0x1bf47000,
 /* 0x007f: nsec */
 	0xf000f8f2,
@@ -845,21 +912,21 @@ uint32_t nva3_pwr_code[] = {
 	0x9800f8df,
 	0x96b003e9,
 	0x2a0bf400,
-	0xbb840a98,
+	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
 	0x025421f5,
 	0x0ef494bd,
 /* 0x00e9: intr_watchdog_next_time */
-	0x850a9815,
+	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
 /* 0x00f8: intr_watchdog_next_time_set */
 /* 0x00fb: intr_watchdog_next_proc */
-	0x80850980,
+	0x809b0980,
 	0xe0b603e9,
-	0x10e6b158,
+	0x68e6b158,
 	0xc61bf402,
 /* 0x010a: intr */
 	0x00f900f8,
@@ -880,15 +947,15 @@ uint32_t nva3_pwr_code[] = {
 	0x0088cf06,
 	0xf40289c4,
 	0x0080230b,
-	0x58e7f085,
+	0x58e7f09b,
 	0x98cb21f4,
-	0x96b08509,
+	0x96b09b09,
 	0x110bf400,
 	0xb63407f0,
 	0x09d00604,
 	0x8004bd00,
 /* 0x016e: intr_skip_watchdog */
-	0x89e48409,
+	0x89e49a09,
 	0x0bf40800,
 	0x8897f148,
 	0x0694b606,
@@ -948,7 +1015,7 @@ uint32_t nva3_pwr_code[] = {
 	0x000ed006,
 	0x0e8004bd,
 /* 0x0241: timer_enable */
-	0x0187f084,
+	0x0187f09a,
 	0xb63807f0,
 	0x08d00604,
 /* 0x024f: timer_done */
@@ -979,7 +1046,7 @@ uint32_t nva3_pwr_code[] = {
 	0xb8008a98,
 	0x0bf406ae,
 	0x5880b610,
-	0x021086b1,
+	0x026886b1,
 	0xf4f01bf4,
 /* 0x02b2: find_done */
 	0x8eb90132,
@@ -1049,7 +1116,7 @@ uint32_t nva3_pwr_code[] = {
 	0x320bf406,
 	0x94071ec4,
 	0xe0b704ee,
-	0xeb980218,
+	0xeb980270,
 	0x02ec9803,
 	0x9801ed98,
 	0x21f500ee,
@@ -1057,7 +1124,7 @@ uint32_t nva3_pwr_code[] = {
 	0x0f1ec401,
 	0x04b007f1,
 	0xd00604b6,
-	0x04bd0001,
+	0x04bd000e,
 /* 0x03cb: host_send_done */
 	0xf8ba0ef4,
 /* 0x03cd: host_recv */
@@ -1075,7 +1142,7 @@ uint32_t nva3_pwr_code[] = {
 	0xe60bf406,
 	0xb60723c4,
 	0x30b70434,
-	0x3b800298,
+	0x3b8002f0,
 	0x023c8003,
 	0x80013d80,
 	0x20b6003e,
@@ -1090,13 +1157,13 @@ uint32_t nva3_pwr_code[] = {
 /* 0x0430: host_init */
 	0x008017f1,
 	0xf11014b6,
-	0xf1021815,
+	0xf1027015,
 	0xb604d007,
 	0x01d00604,
 	0xf104bd00,
 	0xb6008017,
 	0x15f11014,
-	0x07f10298,
+	0x07f102f0,
 	0x04b604dc,
 	0x0001d006,
 	0x17f004bd,
@@ -1156,14 +1223,14 @@ uint32_t nva3_pwr_code[] = {
 	0x00139802,
 	0x950410b6,
 	0x30f01034,
-	0xc835980c,
+	0xde35980c,
 	0x12b855f9,
 	0xec1ef406,
 	0xe0fcd0fc,
 	0x02b921f5,
 /* 0x0532: memx_info */
 	0xc7f100f8,
-	0xb7f10354,
+	0xb7f103ac,
 	0x21f50800,
 	0x00f802b9,
 /* 0x0540: memx_recv */
@@ -1175,7 +1242,312 @@ uint32_t nva3_pwr_code[] = {
 /* 0x0550: perf_recv */
 /* 0x0552: perf_init */
 	0x00f800f8,
-/* 0x0554: test_recv */
+/* 0x0554: i2c_drive_scl */
+	0xf40036b0,
+	0x07f1110b,
+	0x04b607e0,
+	0x0001d006,
+	0x00f804bd,
+/* 0x0568: i2c_drive_scl_lo */
+	0x07e407f1,
+	0xd00604b6,
+	0x04bd0001,
+/* 0x0576: i2c_drive_sda */
+	0x36b000f8,
+	0x110bf400,
+	0x07e007f1,
+	0xd00604b6,
+	0x04bd0002,
+/* 0x058a: i2c_drive_sda_lo */
+	0x07f100f8,
+	0x04b607e4,
+	0x0002d006,
+	0x00f804bd,
+/* 0x0598: i2c_sense_scl */
+	0xf10132f4,
+	0xb607c437,
+	0x33cf0634,
+	0x0431fd00,
+	0xf4060bf4,
+/* 0x05ae: i2c_sense_scl_done */
+	0x00f80131,
+/* 0x05b0: i2c_sense_sda */
+	0xf10132f4,
+	0xb607c437,
+	0x33cf0634,
+	0x0432fd00,
+	0xf4060bf4,
+/* 0x05c6: i2c_sense_sda_done */
+	0x00f80131,
+/* 0x05c8: i2c_raise_scl */
+	0x47f140f9,
+	0x37f00898,
+	0x5421f501,
+/* 0x05d5: i2c_raise_scl_wait */
+	0xe8e7f105,
+	0x7f21f403,
+	0x059821f5,
+	0xb60901f4,
+	0x1bf40142,
+/* 0x05e9: i2c_raise_scl_done */
+	0xf840fcef,
+/* 0x05ed: i2c_start */
+	0x9821f500,
+	0x0d11f405,
+	0x05b021f5,
+	0xf40611f4,
+/* 0x05fe: i2c_start_rep */
+	0x37f0300e,
+	0x5421f500,
+	0x0137f005,
+	0x057621f5,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xc821f550,
+	0x0464b605,
+/* 0x062b: i2c_start_send */
+	0xf01f11f4,
+	0x21f50037,
+	0xe7f10576,
+	0x21f41388,
+	0x0037f07f,
+	0x055421f5,
+	0x1388e7f1,
+/* 0x0647: i2c_start_out */
+	0xf87f21f4,
+/* 0x0649: i2c_stop */
+	0x0037f000,
+	0x055421f5,
+	0xf50037f0,
+	0xf1057621,
+	0xf403e8e7,
+	0x37f07f21,
+	0x5421f501,
+	0x88e7f105,
+	0x7f21f413,
+	0xf50137f0,
+	0xf1057621,
+	0xf41388e7,
+	0x00f87f21,
+/* 0x067c: i2c_bitw */
+	0x057621f5,
+	0x03e8e7f1,
+	0xbb7f21f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05c821f5,
+	0xf40464b6,
+	0xe7f11811,
+	0x21f41388,
+	0x0037f07f,
+	0x055421f5,
+	0x1388e7f1,
+/* 0x06bb: i2c_bitw_out */
+	0xf87f21f4,
+/* 0x06bd: i2c_bitr */
+	0x0137f000,
+	0x057621f5,
+	0x03e8e7f1,
+	0xbb7f21f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05c821f5,
+	0xf40464b6,
+	0x21f51b11,
+	0x37f005b0,
+	0x5421f500,
+	0x88e7f105,
+	0x7f21f413,
+	0xf4013cf0,
+/* 0x0702: i2c_bitr_done */
+	0x00f80131,
+/* 0x0704: i2c_get_byte */
+	0xf00057f0,
+/* 0x070a: i2c_get_byte_next */
+	0x54b60847,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b606bd,
+	0x2b11f404,
+	0xb60553fd,
+	0x1bf40142,
+	0x0137f0d8,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x7c21f550,
+	0x0464b606,
+/* 0x0754: i2c_get_byte_done */
+/* 0x0756: i2c_put_byte */
+	0x47f000f8,
+/* 0x0759: i2c_put_byte_next */
+	0x0142b608,
+	0xbb3854ff,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x067c21f5,
+	0xf40464b6,
+	0x46b03411,
+	0xd81bf400,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xbd21f550,
+	0x0464b606,
+	0xbb0f11f4,
+	0x36b00076,
+	0x061bf401,
+/* 0x07af: i2c_put_byte_done */
+	0xf80132f4,
+/* 0x07b1: i2c_addr */
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b605ed,
+	0x2911f404,
+	0x012ec3e7,
+	0xfd0134b6,
+	0x76bb0553,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6075621,
+/* 0x07f6: i2c_addr_done */
+	0x00f80464,
+/* 0x07f8: i2c_acquire_addr */
+	0xb6f8cec7,
+	0xe0b702e4,
+	0xee980bfc,
+/* 0x0807: i2c_acquire */
+	0xf500f800,
+	0xf407f821,
+	0xd9f00421,
+	0x3f21f403,
+/* 0x0816: i2c_release */
+	0x21f500f8,
+	0x21f407f8,
+	0x03daf004,
+	0xf83f21f4,
+/* 0x0825: i2c_recv */
+	0x0132f400,
+	0xb6f8c1c7,
+	0x16b00214,
+	0x3a1ff528,
+	0xd413a001,
+	0x0032980b,
+	0x0bac13a0,
+	0xf4003198,
+	0xd0f90231,
+	0xd0f9e0f9,
+	0x000067f1,
+	0x100063f1,
+	0xbb016792,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x080721f5,
+	0xfc0464b6,
+	0x00d6b0d0,
+	0x00b31bf5,
+	0xbb0057f0,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x07b121f5,
+	0xf50464b6,
+	0xc700d011,
+	0x76bbe0c5,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6075621,
+	0x11f50464,
+	0x57f000ad,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b607b1,
+	0x8a11f504,
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b60704,
+	0x6a11f404,
+	0xbbe05bcb,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x064921f5,
+	0xb90464b6,
+	0x74bd025b,
+/* 0x092b: i2c_recv_not_rd08 */
+	0xb0430ef4,
+	0x1bf401d6,
+	0x0057f03d,
+	0x07b121f5,
+	0xc73311f4,
+	0x21f5e0c5,
+	0x11f40756,
+	0x0057f029,
+	0x07b121f5,
+	0xc71f11f4,
+	0x21f5e0b5,
+	0x11f40756,
+	0x4921f515,
+	0xc774bd06,
+	0x1bf408c5,
+	0x0232f409,
+/* 0x096b: i2c_recv_not_wr08 */
+/* 0x096b: i2c_recv_done */
+	0xc7030ef4,
+	0x21f5f8ce,
+	0xe0fc0816,
+	0x12f4d0fc,
+	0x027cb90a,
+	0x02b921f5,
+/* 0x0980: i2c_recv_exit */
+/* 0x0982: i2c_init */
+	0x00f800f8,
+/* 0x0984: test_recv */
 	0x05d817f1,
 	0xcf0614b6,
 	0x10b60011,
@@ -1185,12 +1557,12 @@ uint32_t nva3_pwr_code[] = {
 	0x00e7f104,
 	0x4fe3f1d9,
 	0xf521f513,
-/* 0x057b: test_init */
+/* 0x09ab: test_init */
 	0xf100f801,
 	0xf50800e7,
 	0xf801f521,
-/* 0x0585: idle_recv */
-/* 0x0587: idle */
+/* 0x09b5: idle_recv */
+/* 0x09b7: idle */
 	0xf400f800,
 	0x17f10031,
 	0x14b605d4,
@@ -1198,32 +1570,20 @@ uint32_t nva3_pwr_code[] = {
 	0xf10110b6,
 	0xb605d407,
 	0x01d00604,
-/* 0x05a3: idle_loop */
+/* 0x09d3: idle_loop */
 	0xf004bd00,
 	0x32f45817,
-/* 0x05a9: idle_proc */
-/* 0x05a9: idle_proc_exec */
+/* 0x09d9: idle_proc */
+/* 0x09d9: idle_proc_exec */
 	0xb910f902,
 	0x21f5021e,
 	0x10fc02c2,
 	0xf40911f4,
 	0x0ef40231,
-/* 0x05bd: idle_proc_next */
+/* 0x09ed: idle_proc_next */
 	0x5810b6ef,
 	0xf4061fb8,
 	0x02f4e61b,
 	0x0028f4dd,
 	0x00bb0ef4,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
index eaa64da68e3..48f79434a44 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
@@ -37,6 +37,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_PROC
@@ -46,6 +47,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_DATA
@@ -57,6 +59,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h
index 82c8e8b8891..7ac87405d01 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc.h
@@ -89,9 +89,31 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x5f433249,
+	0x00000982,
+	0x00000825,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x54534554,
-	0x0000057b,
-	0x00000554,
+	0x000009ab,
+	0x00000984,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +134,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x00000587,
-	0x00000585,
+	0x000009b7,
+	0x000009b5,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -133,12 +155,12 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0210: proc_list_tail */
-/* 0x0210: time_prev */
+/* 0x0268: proc_list_tail */
+/* 0x0268: time_prev */
 	0x00000000,
-/* 0x0214: time_next */
+/* 0x026c: time_next */
 	0x00000000,
-/* 0x0218: fifo_queue */
+/* 0x0270: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -171,7 +193,7 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0298: rfifo_queue */
+/* 0x02f0: rfifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -204,11 +226,11 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0318: memx_func_head */
+/* 0x0370: memx_func_head */
 	0x00010000,
 	0x00000000,
 	0x0000046f,
-/* 0x0324: memx_func_next */
+/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
 	0x00000496,
@@ -221,8 +243,19 @@ uint32_t nvc0_pwr_data[] = {
 	0x00010004,
 	0x00000000,
 	0x000004fc,
-/* 0x0354: memx_func_tail */
-/* 0x0354: memx_data_head */
+/* 0x03ac: memx_func_tail */
+/* 0x03ac: memx_data_head */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -724,6 +757,41 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bac: memx_data_tail */
+/* 0x0bac: i2c_scl_map */
+	0x00001000,
+	0x00004000,
+	0x00010000,
+	0x00000100,
+	0x00040000,
+	0x00100000,
+	0x00400000,
+	0x01000000,
+	0x04000000,
+	0x10000000,
+/* 0x0bd4: i2c_sda_map */
+	0x00002000,
+	0x00008000,
+	0x00020000,
+	0x00000200,
+	0x00080000,
+	0x00200000,
+	0x00800000,
+	0x02000000,
+	0x08000000,
+	0x20000000,
+/* 0x0bfc: i2c_ctrl */
+	0x0000e138,
+	0x0000e150,
+	0x0000e168,
+	0x0000e180,
+	0x0000e254,
+	0x0000e274,
+	0x0000e764,
+	0x0000e780,
+	0x0000e79c,
+	0x0000e7b8,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -735,7 +803,6 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0b54: memx_data_tail */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -787,15 +854,15 @@ uint32_t nvc0_pwr_code[] = {
 	0x07a007f1,
 	0xd00604b6,
 	0x04bd000e,
-	0xf001e7f0,
-	0x07f101e3,
+	0xf001d7f0,
+	0x07f101d3,
 	0x04b607ac,
-	0x000ed006,
+	0x000dd006,
 /* 0x0022: rd32_wait */
-	0xe7f104bd,
-	0xe4b607ac,
-	0x00eecf06,
-	0x7000e4f1,
+	0xd7f104bd,
+	0xd4b607ac,
+	0x00ddcf06,
+	0x7000d4f1,
 	0xf1f21bf4,
 	0xb607a4d7,
 	0xddcf06d4,
@@ -807,15 +874,15 @@ uint32_t nvc0_pwr_code[] = {
 	0xb607a407,
 	0x0dd00604,
 	0xf004bd00,
-	0xe5f002e7,
-	0x01e3f0f0,
+	0xd5f002d7,
+	0x01d3f0f0,
 	0x07ac07f1,
 	0xd00604b6,
-	0x04bd000e,
+	0x04bd000d,
 /* 0x006c: wr32_wait */
-	0x07ace7f1,
-	0xcf06e4b6,
-	0xe4f100ee,
+	0x07acd7f1,
+	0xcf06d4b6,
+	0xd4f100dd,
 	0x1bf47000,
 /* 0x007f: nsec */
 	0xf000f8f2,
@@ -845,21 +912,21 @@ uint32_t nvc0_pwr_code[] = {
 	0x9800f8df,
 	0x96b003e9,
 	0x2a0bf400,
-	0xbb840a98,
+	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
 	0x025421f5,
 	0x0ef494bd,
 /* 0x00e9: intr_watchdog_next_time */
-	0x850a9815,
+	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
 /* 0x00f8: intr_watchdog_next_time_set */
 /* 0x00fb: intr_watchdog_next_proc */
-	0x80850980,
+	0x809b0980,
 	0xe0b603e9,
-	0x10e6b158,
+	0x68e6b158,
 	0xc61bf402,
 /* 0x010a: intr */
 	0x00f900f8,
@@ -880,15 +947,15 @@ uint32_t nvc0_pwr_code[] = {
 	0x0088cf06,
 	0xf40289c4,
 	0x0080230b,
-	0x58e7f085,
+	0x58e7f09b,
 	0x98cb21f4,
-	0x96b08509,
+	0x96b09b09,
 	0x110bf400,
 	0xb63407f0,
 	0x09d00604,
 	0x8004bd00,
 /* 0x016e: intr_skip_watchdog */
-	0x89e48409,
+	0x89e49a09,
 	0x0bf40800,
 	0x8897f148,
 	0x0694b606,
@@ -948,7 +1015,7 @@ uint32_t nvc0_pwr_code[] = {
 	0x000ed006,
 	0x0e8004bd,
 /* 0x0241: timer_enable */
-	0x0187f084,
+	0x0187f09a,
 	0xb63807f0,
 	0x08d00604,
 /* 0x024f: timer_done */
@@ -979,7 +1046,7 @@ uint32_t nvc0_pwr_code[] = {
 	0xb8008a98,
 	0x0bf406ae,
 	0x5880b610,
-	0x021086b1,
+	0x026886b1,
 	0xf4f01bf4,
 /* 0x02b2: find_done */
 	0x8eb90132,
@@ -1049,7 +1116,7 @@ uint32_t nvc0_pwr_code[] = {
 	0x320bf406,
 	0x94071ec4,
 	0xe0b704ee,
-	0xeb980218,
+	0xeb980270,
 	0x02ec9803,
 	0x9801ed98,
 	0x21f500ee,
@@ -1057,7 +1124,7 @@ uint32_t nvc0_pwr_code[] = {
 	0x0f1ec401,
 	0x04b007f1,
 	0xd00604b6,
-	0x04bd0001,
+	0x04bd000e,
 /* 0x03cb: host_send_done */
 	0xf8ba0ef4,
 /* 0x03cd: host_recv */
@@ -1075,7 +1142,7 @@ uint32_t nvc0_pwr_code[] = {
 	0xe60bf406,
 	0xb60723c4,
 	0x30b70434,
-	0x3b800298,
+	0x3b8002f0,
 	0x023c8003,
 	0x80013d80,
 	0x20b6003e,
@@ -1090,13 +1157,13 @@ uint32_t nvc0_pwr_code[] = {
 /* 0x0430: host_init */
 	0x008017f1,
 	0xf11014b6,
-	0xf1021815,
+	0xf1027015,
 	0xb604d007,
 	0x01d00604,
 	0xf104bd00,
 	0xb6008017,
 	0x15f11014,
-	0x07f10298,
+	0x07f102f0,
 	0x04b604dc,
 	0x0001d006,
 	0x17f004bd,
@@ -1156,14 +1223,14 @@ uint32_t nvc0_pwr_code[] = {
 	0x00139802,
 	0x950410b6,
 	0x30f01034,
-	0xc835980c,
+	0xde35980c,
 	0x12b855f9,
 	0xec1ef406,
 	0xe0fcd0fc,
 	0x02b921f5,
 /* 0x0532: memx_info */
 	0xc7f100f8,
-	0xb7f10354,
+	0xb7f103ac,
 	0x21f50800,
 	0x00f802b9,
 /* 0x0540: memx_recv */
@@ -1175,7 +1242,312 @@ uint32_t nvc0_pwr_code[] = {
 /* 0x0550: perf_recv */
 /* 0x0552: perf_init */
 	0x00f800f8,
-/* 0x0554: test_recv */
+/* 0x0554: i2c_drive_scl */
+	0xf40036b0,
+	0x07f1110b,
+	0x04b607e0,
+	0x0001d006,
+	0x00f804bd,
+/* 0x0568: i2c_drive_scl_lo */
+	0x07e407f1,
+	0xd00604b6,
+	0x04bd0001,
+/* 0x0576: i2c_drive_sda */
+	0x36b000f8,
+	0x110bf400,
+	0x07e007f1,
+	0xd00604b6,
+	0x04bd0002,
+/* 0x058a: i2c_drive_sda_lo */
+	0x07f100f8,
+	0x04b607e4,
+	0x0002d006,
+	0x00f804bd,
+/* 0x0598: i2c_sense_scl */
+	0xf10132f4,
+	0xb607c437,
+	0x33cf0634,
+	0x0431fd00,
+	0xf4060bf4,
+/* 0x05ae: i2c_sense_scl_done */
+	0x00f80131,
+/* 0x05b0: i2c_sense_sda */
+	0xf10132f4,
+	0xb607c437,
+	0x33cf0634,
+	0x0432fd00,
+	0xf4060bf4,
+/* 0x05c6: i2c_sense_sda_done */
+	0x00f80131,
+/* 0x05c8: i2c_raise_scl */
+	0x47f140f9,
+	0x37f00898,
+	0x5421f501,
+/* 0x05d5: i2c_raise_scl_wait */
+	0xe8e7f105,
+	0x7f21f403,
+	0x059821f5,
+	0xb60901f4,
+	0x1bf40142,
+/* 0x05e9: i2c_raise_scl_done */
+	0xf840fcef,
+/* 0x05ed: i2c_start */
+	0x9821f500,
+	0x0d11f405,
+	0x05b021f5,
+	0xf40611f4,
+/* 0x05fe: i2c_start_rep */
+	0x37f0300e,
+	0x5421f500,
+	0x0137f005,
+	0x057621f5,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xc821f550,
+	0x0464b605,
+/* 0x062b: i2c_start_send */
+	0xf01f11f4,
+	0x21f50037,
+	0xe7f10576,
+	0x21f41388,
+	0x0037f07f,
+	0x055421f5,
+	0x1388e7f1,
+/* 0x0647: i2c_start_out */
+	0xf87f21f4,
+/* 0x0649: i2c_stop */
+	0x0037f000,
+	0x055421f5,
+	0xf50037f0,
+	0xf1057621,
+	0xf403e8e7,
+	0x37f07f21,
+	0x5421f501,
+	0x88e7f105,
+	0x7f21f413,
+	0xf50137f0,
+	0xf1057621,
+	0xf41388e7,
+	0x00f87f21,
+/* 0x067c: i2c_bitw */
+	0x057621f5,
+	0x03e8e7f1,
+	0xbb7f21f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05c821f5,
+	0xf40464b6,
+	0xe7f11811,
+	0x21f41388,
+	0x0037f07f,
+	0x055421f5,
+	0x1388e7f1,
+/* 0x06bb: i2c_bitw_out */
+	0xf87f21f4,
+/* 0x06bd: i2c_bitr */
+	0x0137f000,
+	0x057621f5,
+	0x03e8e7f1,
+	0xbb7f21f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05c821f5,
+	0xf40464b6,
+	0x21f51b11,
+	0x37f005b0,
+	0x5421f500,
+	0x88e7f105,
+	0x7f21f413,
+	0xf4013cf0,
+/* 0x0702: i2c_bitr_done */
+	0x00f80131,
+/* 0x0704: i2c_get_byte */
+	0xf00057f0,
+/* 0x070a: i2c_get_byte_next */
+	0x54b60847,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b606bd,
+	0x2b11f404,
+	0xb60553fd,
+	0x1bf40142,
+	0x0137f0d8,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x7c21f550,
+	0x0464b606,
+/* 0x0754: i2c_get_byte_done */
+/* 0x0756: i2c_put_byte */
+	0x47f000f8,
+/* 0x0759: i2c_put_byte_next */
+	0x0142b608,
+	0xbb3854ff,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x067c21f5,
+	0xf40464b6,
+	0x46b03411,
+	0xd81bf400,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xbd21f550,
+	0x0464b606,
+	0xbb0f11f4,
+	0x36b00076,
+	0x061bf401,
+/* 0x07af: i2c_put_byte_done */
+	0xf80132f4,
+/* 0x07b1: i2c_addr */
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b605ed,
+	0x2911f404,
+	0x012ec3e7,
+	0xfd0134b6,
+	0x76bb0553,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6075621,
+/* 0x07f6: i2c_addr_done */
+	0x00f80464,
+/* 0x07f8: i2c_acquire_addr */
+	0xb6f8cec7,
+	0xe0b702e4,
+	0xee980bfc,
+/* 0x0807: i2c_acquire */
+	0xf500f800,
+	0xf407f821,
+	0xd9f00421,
+	0x3f21f403,
+/* 0x0816: i2c_release */
+	0x21f500f8,
+	0x21f407f8,
+	0x03daf004,
+	0xf83f21f4,
+/* 0x0825: i2c_recv */
+	0x0132f400,
+	0xb6f8c1c7,
+	0x16b00214,
+	0x3a1ff528,
+	0xd413a001,
+	0x0032980b,
+	0x0bac13a0,
+	0xf4003198,
+	0xd0f90231,
+	0xd0f9e0f9,
+	0x000067f1,
+	0x100063f1,
+	0xbb016792,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x080721f5,
+	0xfc0464b6,
+	0x00d6b0d0,
+	0x00b31bf5,
+	0xbb0057f0,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x07b121f5,
+	0xf50464b6,
+	0xc700d011,
+	0x76bbe0c5,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6075621,
+	0x11f50464,
+	0x57f000ad,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b607b1,
+	0x8a11f504,
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b60704,
+	0x6a11f404,
+	0xbbe05bcb,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x064921f5,
+	0xb90464b6,
+	0x74bd025b,
+/* 0x092b: i2c_recv_not_rd08 */
+	0xb0430ef4,
+	0x1bf401d6,
+	0x0057f03d,
+	0x07b121f5,
+	0xc73311f4,
+	0x21f5e0c5,
+	0x11f40756,
+	0x0057f029,
+	0x07b121f5,
+	0xc71f11f4,
+	0x21f5e0b5,
+	0x11f40756,
+	0x4921f515,
+	0xc774bd06,
+	0x1bf408c5,
+	0x0232f409,
+/* 0x096b: i2c_recv_not_wr08 */
+/* 0x096b: i2c_recv_done */
+	0xc7030ef4,
+	0x21f5f8ce,
+	0xe0fc0816,
+	0x12f4d0fc,
+	0x027cb90a,
+	0x02b921f5,
+/* 0x0980: i2c_recv_exit */
+/* 0x0982: i2c_init */
+	0x00f800f8,
+/* 0x0984: test_recv */
 	0x05d817f1,
 	0xcf0614b6,
 	0x10b60011,
@@ -1185,12 +1557,12 @@ uint32_t nvc0_pwr_code[] = {
 	0x00e7f104,
 	0x4fe3f1d9,
 	0xf521f513,
-/* 0x057b: test_init */
+/* 0x09ab: test_init */
 	0xf100f801,
 	0xf50800e7,
 	0xf801f521,
-/* 0x0585: idle_recv */
-/* 0x0587: idle */
+/* 0x09b5: idle_recv */
+/* 0x09b7: idle */
 	0xf400f800,
 	0x17f10031,
 	0x14b605d4,
@@ -1198,32 +1570,20 @@ uint32_t nvc0_pwr_code[] = {
 	0xf10110b6,
 	0xb605d407,
 	0x01d00604,
-/* 0x05a3: idle_loop */
+/* 0x09d3: idle_loop */
 	0xf004bd00,
 	0x32f45817,
-/* 0x05a9: idle_proc */
-/* 0x05a9: idle_proc_exec */
+/* 0x09d9: idle_proc */
+/* 0x09d9: idle_proc_exec */
 	0xb910f902,
 	0x21f5021e,
 	0x10fc02c2,
 	0xf40911f4,
 	0x0ef40231,
-/* 0x05bd: idle_proc_next */
+/* 0x09ed: idle_proc_next */
 	0x5810b6ef,
 	0xf4061fb8,
 	0x02f4e61b,
 	0x0028f4dd,
 	0x00bb0ef4,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
 };
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
index 32d65ea254d..8a89dfe41ce 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
@@ -37,6 +37,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_PROC
@@ -46,6 +47,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_DATA
@@ -57,6 +59,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h
index ce65e2a4b78..cd9ff1a7328 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc.h
@@ -89,34 +89,9 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x54534554,
-	0x000004eb,
-	0x000004ca,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x454c4449,
-	0x000004f7,
-	0x000004f5,
-	0x00000000,
-	0x00000000,
-	0x00000000,
+	0x5f433249,
+	0x000008e3,
+	0x00000786,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -133,14 +108,12 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0210: proc_list_tail */
-/* 0x0210: time_prev */
 	0x00000000,
-/* 0x0214: time_next */
-	0x00000000,
-/* 0x0218: fifo_queue */
 	0x00000000,
 	0x00000000,
+	0x54534554,
+	0x00000906,
+	0x000008e5,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -160,6 +133,9 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x454c4449,
+	0x00000912,
+	0x00000910,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -171,7 +147,6 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0298: rfifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -180,8 +155,12 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0268: proc_list_tail */
+/* 0x0268: time_prev */
 	0x00000000,
+/* 0x026c: time_next */
 	0x00000000,
+/* 0x0270: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -204,31 +183,17 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0318: memx_func_head */
-	0x00010000,
 	0x00000000,
-	0x000003f4,
-/* 0x0324: memx_func_next */
-	0x00000001,
 	0x00000000,
-	0x00000415,
-	0x00000002,
-	0x00000002,
-	0x00000430,
-	0x00040003,
 	0x00000000,
-	0x00000458,
-	0x00010004,
 	0x00000000,
-	0x00000472,
-/* 0x0354: memx_func_tail */
-/* 0x0354: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x02f0: rfifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -261,10 +226,25 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0370: memx_func_head */
+	0x00010000,
 	0x00000000,
+	0x000003f4,
+/* 0x037c: memx_func_next */
+	0x00000001,
 	0x00000000,
+	0x00000415,
+	0x00000002,
+	0x00000002,
+	0x00000430,
+	0x00040003,
 	0x00000000,
+	0x00000458,
+	0x00010004,
 	0x00000000,
+	0x00000472,
+/* 0x03ac: memx_func_tail */
+/* 0x03ac: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -735,7 +715,6 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0b54: memx_data_tail */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -778,6 +757,29 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bac: memx_data_tail */
+/* 0x0bac: i2c_scl_map */
+	0x00000400,
+	0x00000800,
+	0x00001000,
+	0x00002000,
+	0x00004000,
+	0x00008000,
+	0x00010000,
+	0x00020000,
+	0x00040000,
+	0x00080000,
+/* 0x0bd4: i2c_sda_map */
+	0x00100000,
+	0x00200000,
+	0x00400000,
+	0x00800000,
+	0x01000000,
+	0x02000000,
+	0x04000000,
+	0x08000000,
+	0x10000000,
+	0x20000000,
 	0x00000000,
 };
 
@@ -786,14 +788,14 @@ uint32_t nvd0_pwr_code[] = {
 /* 0x0004: rd32 */
 	0x07a007f1,
 	0xbd000ed0,
-	0x01e7f004,
-	0xf101e3f0,
+	0x01d7f004,
+	0xf101d3f0,
 	0xd007ac07,
-	0x04bd000e,
+	0x04bd000d,
 /* 0x001c: rd32_wait */
-	0x07ace7f1,
-	0xf100eecf,
-	0xf47000e4,
+	0x07acd7f1,
+	0xf100ddcf,
+	0xf47000d4,
 	0xd7f1f51b,
 	0xddcf07a4,
 /* 0x0033: wr32 */
@@ -802,14 +804,14 @@ uint32_t nvd0_pwr_code[] = {
 	0x04bd000e,
 	0x07a407f1,
 	0xbd000dd0,
-	0x02e7f004,
-	0xf0f0e5f0,
-	0x07f101e3,
-	0x0ed007ac,
+	0x02d7f004,
+	0xf0f0d5f0,
+	0x07f101d3,
+	0x0dd007ac,
 /* 0x0057: wr32_wait */
 	0xf104bd00,
-	0xcf07ace7,
-	0xe4f100ee,
+	0xcf07acd7,
+	0xd4f100dd,
 	0x1bf47000,
 /* 0x0067: nsec */
 	0xf000f8f5,
@@ -836,21 +838,21 @@ uint32_t nvd0_pwr_code[] = {
 	0x9800f8e2,
 	0x96b003e9,
 	0x2a0bf400,
-	0xbb840a98,
+	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
 	0x020621f5,
 	0x0ef494bd,
 /* 0x00c5: intr_watchdog_next_time */
-	0x850a9815,
+	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
 /* 0x00d4: intr_watchdog_next_time_set */
 /* 0x00d7: intr_watchdog_next_proc */
-	0x80850980,
+	0x809b0980,
 	0xe0b603e9,
-	0x10e6b158,
+	0x68e6b158,
 	0xc61bf402,
 /* 0x00e6: intr */
 	0x00f900f8,
@@ -868,15 +870,15 @@ uint32_t nvd0_pwr_code[] = {
 	0x0887f004,
 	0xc40088cf,
 	0x0bf40289,
-	0x85008020,
+	0x9b008020,
 	0xf458e7f0,
 	0x0998a721,
-	0x0096b085,
+	0x0096b09b,
 	0xf00e0bf4,
 	0x09d03407,
 	0x8004bd00,
 /* 0x013e: intr_skip_watchdog */
-	0x89e48409,
+	0x89e49a09,
 	0x0bf40800,
 	0x8897f13c,
 	0x0099cf06,
@@ -929,7 +931,7 @@ uint32_t nvd0_pwr_code[] = {
 	0x0ed03407,
 	0x8004bd00,
 /* 0x01f6: timer_enable */
-	0x87f0840e,
+	0x87f09a0e,
 	0x3807f001,
 	0xbd0008d0,
 /* 0x0201: timer_done */
@@ -960,7 +962,7 @@ uint32_t nvd0_pwr_code[] = {
 	0x06aeb800,
 	0xb6100bf4,
 	0x86b15880,
-	0x1bf40210,
+	0x1bf40268,
 	0x0132f4f0,
 /* 0x0264: find_done */
 	0xfc028eb9,
@@ -1024,14 +1026,14 @@ uint32_t nvd0_pwr_code[] = {
 	0x0bf40612,
 	0x071ec42f,
 	0xb704ee94,
-	0x980218e0,
+	0x980270e0,
 	0xec9803eb,
 	0x01ed9802,
 	0xf500ee98,
 	0xb6026b21,
 	0x1ec40110,
 	0xb007f10f,
-	0x0001d004,
+	0x000ed004,
 	0x0ef404bd,
 /* 0x0365: host_send_done */
 /* 0x0367: host_recv */
@@ -1048,7 +1050,7 @@ uint32_t nvd0_pwr_code[] = {
 	0xec0bf406,
 	0xb60723c4,
 	0x30b70434,
-	0x3b800298,
+	0x3b8002f0,
 	0x023c8003,
 	0x80013d80,
 	0x20b6003e,
@@ -1061,12 +1063,12 @@ uint32_t nvd0_pwr_code[] = {
 /* 0x03be: host_init */
 	0x17f100f8,
 	0x14b60080,
-	0x1815f110,
+	0x7015f110,
 	0xd007f102,
 	0x0001d004,
 	0x17f104bd,
 	0x14b60080,
-	0x9815f110,
+	0xf015f110,
 	0xdc07f102,
 	0x0001d004,
 	0x17f004bd,
@@ -1122,13 +1124,13 @@ uint32_t nvd0_pwr_code[] = {
 	0x10b60013,
 	0x10349504,
 	0x980c30f0,
-	0x55f9c835,
+	0x55f9de35,
 	0xf40612b8,
 	0xd0fcec1e,
 	0x21f5e0fc,
 	0x00f8026b,
 /* 0x04a8: memx_info */
-	0x0354c7f1,
+	0x03acc7f1,
 	0x0800b7f1,
 	0x026b21f5,
 /* 0x04b6: memx_recv */
@@ -1140,49 +1142,342 @@ uint32_t nvd0_pwr_code[] = {
 /* 0x04c6: perf_recv */
 	0x00f800f8,
 /* 0x04c8: perf_init */
-/* 0x04ca: test_recv */
-	0x17f100f8,
-	0x11cf05d8,
-	0x0110b600,
-	0x05d807f1,
-	0xbd0001d0,
-	0x00e7f104,
-	0x4fe3f1d9,
-	0xb621f513,
-/* 0x04eb: test_init */
-	0xf100f801,
-	0xf50800e7,
-	0xf801b621,
-/* 0x04f5: idle_recv */
-/* 0x04f7: idle */
-	0xf400f800,
-	0x17f10031,
-	0x11cf05d4,
-	0x0110b600,
-	0x05d407f1,
+/* 0x04ca: i2c_drive_scl */
+	0x36b000f8,
+	0x0e0bf400,
+	0x07e007f1,
 	0xbd0001d0,
-/* 0x050d: idle_loop */
-	0x5817f004,
-/* 0x0513: idle_proc */
-/* 0x0513: idle_proc_exec */
-	0xf90232f4,
-	0x021eb910,
-	0x027421f5,
-	0x11f410fc,
-	0x0231f409,
-/* 0x0527: idle_proc_next */
-	0xb6ef0ef4,
-	0x1fb85810,
-	0xe61bf406,
-	0xf4dd02f4,
-	0x0ef40028,
-	0x000000c1,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
+/* 0x04db: i2c_drive_scl_lo */
+	0xf100f804,
+	0xd007e407,
+	0x04bd0001,
+/* 0x04e6: i2c_drive_sda */
+	0x36b000f8,
+	0x0e0bf400,
+	0x07e007f1,
+	0xbd0002d0,
+/* 0x04f7: i2c_drive_sda_lo */
+	0xf100f804,
+	0xd007e407,
+	0x04bd0002,
+/* 0x0502: i2c_sense_scl */
+	0x32f400f8,
+	0xc437f101,
+	0x0033cf07,
+	0xf40431fd,
+	0x31f4060b,
+/* 0x0515: i2c_sense_scl_done */
+/* 0x0517: i2c_sense_sda */
+	0xf400f801,
+	0x37f10132,
+	0x33cf07c4,
+	0x0432fd00,
+	0xf4060bf4,
+/* 0x052a: i2c_sense_sda_done */
+	0x00f80131,
+/* 0x052c: i2c_raise_scl */
+	0x47f140f9,
+	0x37f00898,
+	0xca21f501,
+/* 0x0539: i2c_raise_scl_wait */
+	0xe8e7f104,
+	0x6721f403,
+	0x050221f5,
+	0xb60901f4,
+	0x1bf40142,
+/* 0x054d: i2c_raise_scl_done */
+	0xf840fcef,
+/* 0x0551: i2c_start */
+	0x0221f500,
+	0x0d11f405,
+	0x051721f5,
+	0xf40611f4,
+/* 0x0562: i2c_start_rep */
+	0x37f0300e,
+	0xca21f500,
+	0x0137f004,
+	0x04e621f5,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x2c21f550,
+	0x0464b605,
+/* 0x058f: i2c_start_send */
+	0xf01f11f4,
+	0x21f50037,
+	0xe7f104e6,
+	0x21f41388,
+	0x0037f067,
+	0x04ca21f5,
+	0x1388e7f1,
+/* 0x05ab: i2c_start_out */
+	0xf86721f4,
+/* 0x05ad: i2c_stop */
+	0x0037f000,
+	0x04ca21f5,
+	0xf50037f0,
+	0xf104e621,
+	0xf403e8e7,
+	0x37f06721,
+	0xca21f501,
+	0x88e7f104,
+	0x6721f413,
+	0xf50137f0,
+	0xf104e621,
+	0xf41388e7,
+	0x00f86721,
+/* 0x05e0: i2c_bitw */
+	0x04e621f5,
+	0x03e8e7f1,
+	0xbb6721f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x052c21f5,
+	0xf40464b6,
+	0xe7f11811,
+	0x21f41388,
+	0x0037f067,
+	0x04ca21f5,
+	0x1388e7f1,
+/* 0x061f: i2c_bitw_out */
+	0xf86721f4,
+/* 0x0621: i2c_bitr */
+	0x0137f000,
+	0x04e621f5,
+	0x03e8e7f1,
+	0xbb6721f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x052c21f5,
+	0xf40464b6,
+	0x21f51b11,
+	0x37f00517,
+	0xca21f500,
+	0x88e7f104,
+	0x6721f413,
+	0xf4013cf0,
+/* 0x0666: i2c_bitr_done */
+	0x00f80131,
+/* 0x0668: i2c_get_byte */
+	0xf00057f0,
+/* 0x066e: i2c_get_byte_next */
+	0x54b60847,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b60621,
+	0x2b11f404,
+	0xb60553fd,
+	0x1bf40142,
+	0x0137f0d8,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xe021f550,
+	0x0464b605,
+/* 0x06b8: i2c_get_byte_done */
+/* 0x06ba: i2c_put_byte */
+	0x47f000f8,
+/* 0x06bd: i2c_put_byte_next */
+	0x0142b608,
+	0xbb3854ff,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05e021f5,
+	0xf40464b6,
+	0x46b03411,
+	0xd81bf400,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x2121f550,
+	0x0464b606,
+	0xbb0f11f4,
+	0x36b00076,
+	0x061bf401,
+/* 0x0713: i2c_put_byte_done */
+	0xf80132f4,
+/* 0x0715: i2c_addr */
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b60551,
+	0x2911f404,
+	0x012ec3e7,
+	0xfd0134b6,
+	0x76bb0553,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb606ba21,
+/* 0x075a: i2c_addr_done */
+	0x00f80464,
+/* 0x075c: i2c_acquire_addr */
+	0xb6f8cec7,
+	0xe0b705e4,
+	0x00f8d014,
+/* 0x0768: i2c_acquire */
+	0x075c21f5,
+	0xf00421f4,
+	0x21f403d9,
+/* 0x0777: i2c_release */
+	0xf500f833,
+	0xf4075c21,
+	0xdaf00421,
+	0x3321f403,
+/* 0x0786: i2c_recv */
+	0x32f400f8,
+	0xf8c1c701,
+	0xb00214b6,
+	0x1ff52816,
+	0x13a0013a,
+	0x32980bd4,
+	0xac13a000,
+	0x0031980b,
+	0xf90231f4,
+	0xf9e0f9d0,
+	0x0067f1d0,
+	0x0063f100,
+	0x01679210,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x6821f550,
+	0x0464b607,
+	0xd6b0d0fc,
+	0xb31bf500,
+	0x0057f000,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x1521f550,
+	0x0464b607,
+	0x00d011f5,
+	0xbbe0c5c7,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x06ba21f5,
+	0xf50464b6,
+	0xf000ad11,
+	0x76bb0157,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6071521,
+	0x11f50464,
+	0x76bb008a,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6066821,
+	0x11f40464,
+	0xe05bcb6a,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xad21f550,
+	0x0464b605,
+	0xbd025bb9,
+	0x430ef474,
+/* 0x088c: i2c_recv_not_rd08 */
+	0xf401d6b0,
+	0x57f03d1b,
+	0x1521f500,
+	0x3311f407,
+	0xf5e0c5c7,
+	0xf406ba21,
+	0x57f02911,
+	0x1521f500,
+	0x1f11f407,
+	0xf5e0b5c7,
+	0xf406ba21,
+	0x21f51511,
+	0x74bd05ad,
+	0xf408c5c7,
+	0x32f4091b,
+	0x030ef402,
+/* 0x08cc: i2c_recv_not_wr08 */
+/* 0x08cc: i2c_recv_done */
+	0xf5f8cec7,
+	0xfc077721,
+	0xf4d0fce0,
+	0x7cb90a12,
+	0x6b21f502,
+/* 0x08e1: i2c_recv_exit */
+/* 0x08e3: i2c_init */
+	0xf800f802,
+/* 0x08e5: test_recv */
+	0xd817f100,
+	0x0011cf05,
+	0xf10110b6,
+	0xd005d807,
+	0x04bd0001,
+	0xd900e7f1,
+	0x134fe3f1,
+	0x01b621f5,
+/* 0x0906: test_init */
+	0xe7f100f8,
+	0x21f50800,
+	0x00f801b6,
+/* 0x0910: idle_recv */
+/* 0x0912: idle */
+	0x31f400f8,
+	0xd417f100,
+	0x0011cf05,
+	0xf10110b6,
+	0xd005d407,
+	0x04bd0001,
+/* 0x0928: idle_loop */
+	0xf45817f0,
+/* 0x092e: idle_proc */
+/* 0x092e: idle_proc_exec */
+	0x10f90232,
+	0xf5021eb9,
+	0xfc027421,
+	0x0911f410,
+	0xf40231f4,
+/* 0x0942: idle_proc_next */
+	0x10b6ef0e,
+	0x061fb858,
+	0xf4e61bf4,
+	0x28f4dd02,
+	0xc10ef400,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
index 5fb0cccc6c6..574acfa44c8 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
@@ -7,6 +7,7 @@
 #define PROC_HOST 0x54534f48
 #define PROC_MEMX 0x584d454d
 #define PROC_PERF 0x46524550
+#define PROC_I2C_ 0x5f433249
 #define PROC_TEST 0x54534554
 
 /* KERN: message identifiers */
@@ -24,4 +25,22 @@
 #define MEMX_WAIT   3
 #define MEMX_DELAY  4
 
+/* I2C_: message identifiers */
+#define I2C__MSG_RD08 0
+#define I2C__MSG_WR08 1
+
+#define I2C__MSG_DATA0_PORT 24:31
+#define I2C__MSG_DATA0_ADDR 14:23
+
+#define I2C__MSG_DATA0_RD08_PORT I2C__MSG_DATA0_PORT
+#define I2C__MSG_DATA0_RD08_ADDR I2C__MSG_DATA0_ADDR
+#define I2C__MSG_DATA0_RD08_REG 0:7
+#define I2C__MSG_DATA1_RD08_VAL 0:7
+
+#define I2C__MSG_DATA0_WR08_PORT I2C__MSG_DATA0_PORT
+#define I2C__MSG_DATA0_WR08_ADDR I2C__MSG_DATA0_ADDR
+#define I2C__MSG_DATA0_WR08_SYNC 8:8
+#define I2C__MSG_DATA0_WR08_REG 0:7
+#define I2C__MSG_DATA1_WR08_VAL 0:7
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/base.c b/drivers/gpu/drm/nouveau/core/subdev/therm/base.c
index 80e584a1bd1..9ad01da6eac 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/base.c
@@ -110,16 +110,18 @@ nouveau_therm_update(struct nouveau_therm *therm, int mode)
 		poll = false;
 		break;
 	case NOUVEAU_THERM_CTRL_AUTO:
-		if (priv->fan->bios.nr_fan_trip) {
+		switch(priv->fan->bios.fan_mode) {
+		case NVBIOS_THERM_FAN_TRIP:
 			duty = nouveau_therm_update_trip(therm);
-		} else
-		if (priv->fan->bios.linear_min_temp ||
-		    priv->fan->bios.linear_max_temp) {
+			break;
+		case NVBIOS_THERM_FAN_LINEAR:
 			duty = nouveau_therm_update_linear(therm);
-		} else {
+			break;
+		case NVBIOS_THERM_FAN_OTHER:
 			if (priv->cstate)
 				duty = priv->cstate;
 			poll = false;
+			break;
 		}
 		immd = false;
 		break;
@@ -179,7 +181,7 @@ nouveau_therm_fan_mode(struct nouveau_therm *therm, int mode)
 
 	/* do not allow automatic fan management if the thermal sensor is
 	 * not available */
-	if (priv->mode == NOUVEAU_THERM_CTRL_AUTO && therm->temp_get(therm) < 0)
+	if (mode == NOUVEAU_THERM_CTRL_AUTO && therm->temp_get(therm) < 0)
 		return -EINVAL;
 
 	if (priv->mode == mode)
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
index 95f6129eeed..016990a8252 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
@@ -54,8 +54,10 @@ nouveau_fan_update(struct nouveau_fan *fan, bool immediate, int target)
 
 	/* check that we're not already at the target duty cycle */
 	duty = fan->get(therm);
-	if (duty == target)
-		goto done;
+	if (duty == target) {
+		spin_unlock_irqrestore(&fan->lock, flags);
+		return 0;
+	}
 
 	/* smooth out the fanspeed increase/decrease */
 	if (!immediate && duty >= 0) {
@@ -73,8 +75,15 @@ nouveau_fan_update(struct nouveau_fan *fan, bool immediate, int target)
 
 	nv_debug(therm, "FAN update: %d\n", duty);
 	ret = fan->set(therm, duty);
-	if (ret)
-		goto done;
+	if (ret) {
+		spin_unlock_irqrestore(&fan->lock, flags);
+		return ret;
+	}
+
+	/* fan speed updated, drop the fan lock before grabbing the
+	 * alarm-scheduling lock and risking a deadlock
+	 */
+	spin_unlock_irqrestore(&fan->lock, flags);
 
 	/* schedule next fan update, if not at target speed already */
 	if (list_empty(&fan->alarm.head) && target != duty) {
@@ -92,8 +101,6 @@ nouveau_fan_update(struct nouveau_fan *fan, bool immediate, int target)
 		ptimer->alarm(ptimer, delay * 1000 * 1000, &fan->alarm);
 	}
 
-done:
-	spin_unlock_irqrestore(&fan->lock, flags);
 	return ret;
 }
 
@@ -185,11 +192,8 @@ nouveau_therm_fan_set_defaults(struct nouveau_therm *therm)
 	priv->fan->bios.max_duty = 100;
 	priv->fan->bios.bump_period = 500;
 	priv->fan->bios.slow_down_period = 2000;
-/*XXX: talk to mupuf */
-#if 0
 	priv->fan->bios.linear_min_temp = 40;
 	priv->fan->bios.linear_max_temp = 85;
-#endif
 }
 
 static void
@@ -235,7 +239,8 @@ nouveau_therm_fan_ctor(struct nouveau_therm *therm)
 	/* attempt to locate a drivable fan, and determine control method */
 	ret = gpio->find(gpio, 0, DCB_GPIO_FAN, 0xff, &func);
 	if (ret == 0) {
-		if (func.log[0] & DCB_GPIO_LOG_DIR_IN) {
+		/* FIXME: is this really the place to perform such checks ? */
+		if (func.line != 16 && func.log[0] & DCB_GPIO_LOG_DIR_IN) {
 			nv_debug(therm, "GPIO_FAN is in input mode\n");
 			ret = -EINVAL;
 		} else {
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c b/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c
index 5f71db8e899..9a5c0734026 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/fanpwm.c
@@ -67,7 +67,7 @@ nouveau_fanpwm_set(struct nouveau_therm *therm, int percent)
 	if (priv->base.bios.pwm_freq) {
 		divs = 1;
 		if (therm->pwm_clock)
-			divs = therm->pwm_clock(therm);
+			divs = therm->pwm_clock(therm, priv->func.line);
 		divs /= priv->base.bios.pwm_freq;
 	}
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/ic.c b/drivers/gpu/drm/nouveau/core/subdev/therm/ic.c
index 7610fc5f8fa..ca9ad9fd47b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/ic.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/ic.c
@@ -60,9 +60,9 @@ static struct nouveau_i2c_board_info
 nv_board_infos[] = {
 	{ { I2C_BOARD_INFO("w83l785ts", 0x2d) }, 0 },
 	{ { I2C_BOARD_INFO("w83781d", 0x2d) }, 0  },
-	{ { I2C_BOARD_INFO("adt7473", 0x2e) }, 20  },
-	{ { I2C_BOARD_INFO("adt7473", 0x2d) }, 20  },
-	{ { I2C_BOARD_INFO("adt7473", 0x2c) }, 20  },
+	{ { I2C_BOARD_INFO("adt7473", 0x2e) }, 40  },
+	{ { I2C_BOARD_INFO("adt7473", 0x2d) }, 40  },
+	{ { I2C_BOARD_INFO("adt7473", 0x2c) }, 40  },
 	{ { I2C_BOARD_INFO("f75375", 0x2e) }, 0  },
 	{ { I2C_BOARD_INFO("lm99", 0x4c) }, 0  },
 	{ { I2C_BOARD_INFO("lm90", 0x4c) }, 0  },
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nv50.c
index 8cf7597a218..321db927d63 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nv50.c
@@ -93,7 +93,7 @@ nv50_fan_pwm_set(struct nouveau_therm *therm, int line, u32 divs, u32 duty)
 }
 
 int
-nv50_fan_pwm_clock(struct nouveau_therm *therm)
+nv50_fan_pwm_clock(struct nouveau_therm *therm, int line)
 {
 	int chipset = nv_device(therm)->chipset;
 	int crystal = nv_device(therm)->crystal;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c
index 3b2c4580098..0478b2e3fb1 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nva3.c
@@ -36,7 +36,7 @@ nva3_therm_fan_sense(struct nouveau_therm *therm)
 	u32 tach = nv_rd32(therm, 0x00e728) & 0x0000ffff;
 	u32 ctrl = nv_rd32(therm, 0x00e720);
 	if (ctrl & 0x00000001)
-		return tach * 60;
+		return tach * 60 / 2;
 	return -ENODEV;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
index 4dd4f81ae87..bbf117be572 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
@@ -32,12 +32,15 @@ static int
 pwm_info(struct nouveau_therm *therm, int line)
 {
 	u32 gpio = nv_rd32(therm, 0x00d610 + (line * 0x04));
+
 	switch (gpio & 0x000000c0) {
 	case 0x00000000: /* normal mode, possibly pwm forced off by us */
 	case 0x00000040: /* nvio special */
 		switch (gpio & 0x0000001f) {
+		case 0x00: return 2;
 		case 0x19: return 1;
 		case 0x1c: return 0;
+		case 0x1e: return 2;
 		default:
 			break;
 		}
@@ -56,8 +59,9 @@ nvd0_fan_pwm_ctrl(struct nouveau_therm *therm, int line, bool enable)
 	int indx = pwm_info(therm, line);
 	if (indx < 0)
 		return indx;
-
-	nv_mask(therm, 0x00d610 + (line * 0x04), 0x000000c0, data);
+	else if (indx < 2)
+		nv_mask(therm, 0x00d610 + (line * 0x04), 0x000000c0, data);
+	/* nothing to do for indx == 2, it seems hardwired to PTHERM */
 	return 0;
 }
 
@@ -67,10 +71,15 @@ nvd0_fan_pwm_get(struct nouveau_therm *therm, int line, u32 *divs, u32 *duty)
 	int indx = pwm_info(therm, line);
 	if (indx < 0)
 		return indx;
-
-	if (nv_rd32(therm, 0x00d610 + (line * 0x04)) & 0x00000040) {
-		*divs = nv_rd32(therm, 0x00e114 + (indx * 8));
-		*duty = nv_rd32(therm, 0x00e118 + (indx * 8));
+	else if (indx < 2) {
+		if (nv_rd32(therm, 0x00d610 + (line * 0x04)) & 0x00000040) {
+			*divs = nv_rd32(therm, 0x00e114 + (indx * 8));
+			*duty = nv_rd32(therm, 0x00e118 + (indx * 8));
+			return 0;
+		}
+	} else if (indx == 2) {
+		*divs = nv_rd32(therm, 0x0200d8) & 0x1fff;
+		*duty = nv_rd32(therm, 0x0200dc) & 0x1fff;
 		return 0;
 	}
 
@@ -83,16 +92,26 @@ nvd0_fan_pwm_set(struct nouveau_therm *therm, int line, u32 divs, u32 duty)
 	int indx = pwm_info(therm, line);
 	if (indx < 0)
 		return indx;
-
-	nv_wr32(therm, 0x00e114 + (indx * 8), divs);
-	nv_wr32(therm, 0x00e118 + (indx * 8), duty | 0x80000000);
+	else if (indx < 2) {
+		nv_wr32(therm, 0x00e114 + (indx * 8), divs);
+		nv_wr32(therm, 0x00e118 + (indx * 8), duty | 0x80000000);
+	} else if (indx == 2) {
+		nv_mask(therm, 0x0200d8, 0x1fff, divs); /* keep the high bits */
+		nv_wr32(therm, 0x0200dc, duty | 0x40000000);
+	}
 	return 0;
 }
 
 static int
-nvd0_fan_pwm_clock(struct nouveau_therm *therm)
+nvd0_fan_pwm_clock(struct nouveau_therm *therm, int line)
 {
-	return (nv_device(therm)->crystal * 1000) / 20;
+	int indx = pwm_info(therm, line);
+	if (indx < 0)
+		return 0;
+	else if (indx < 2)
+		return (nv_device(therm)->crystal * 1000) / 20;
+	else
+		return nv_device(therm)->crystal * 1000 / 10;
 }
 
 static int
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h b/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h
index 96f8f95693c..916fca5c781 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/priv.h
@@ -143,7 +143,7 @@ void nv40_therm_intr(struct nouveau_subdev *);
 int nv50_fan_pwm_ctrl(struct nouveau_therm *, int, bool);
 int nv50_fan_pwm_get(struct nouveau_therm *, int, u32 *, u32 *);
 int nv50_fan_pwm_set(struct nouveau_therm *, int, u32, u32);
-int nv50_fan_pwm_clock(struct nouveau_therm *);
+int nv50_fan_pwm_clock(struct nouveau_therm *, int);
 int nv84_temp_get(struct nouveau_therm *therm);
 int nv84_therm_fini(struct nouveau_object *object, bool suspend);
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/temp.c b/drivers/gpu/drm/nouveau/core/subdev/therm/temp.c
index cfde9eb44ad..6212537b90c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/temp.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/temp.c
@@ -192,11 +192,11 @@ alarm_timer_callback(struct nouveau_alarm *alarm)
 	nouveau_therm_threshold_hyst_polling(therm, &sensor->thrs_shutdown,
 					     NOUVEAU_THERM_THRS_SHUTDOWN);
 
+	spin_unlock_irqrestore(&priv->sensor.alarm_program_lock, flags);
+
 	/* schedule the next poll in one second */
 	if (therm->temp_get(therm) >= 0 && list_empty(&alarm->head))
-		ptimer->alarm(ptimer, 1000 * 1000 * 1000, alarm);
-
-	spin_unlock_irqrestore(&priv->sensor.alarm_program_lock, flags);
+		ptimer->alarm(ptimer, 1000000000ULL, alarm);
 }
 
 void
diff --git a/drivers/gpu/drm/nouveau/core/subdev/timer/gk20a.c b/drivers/gpu/drm/nouveau/core/subdev/timer/gk20a.c
new file mode 100644
index 00000000000..37484db1f7f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/timer/gk20a.c
@@ -0,0 +1,57 @@
+/*
+ * Copyright 2012 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+
+#include "nv04.h"
+
+static int
+gk20a_timer_init(struct nouveau_object *object)
+{
+	struct nv04_timer_priv *priv = (void *)object;
+	u32 hi = upper_32_bits(priv->suspend_time);
+	u32 lo = lower_32_bits(priv->suspend_time);
+	int ret;
+
+	ret = nouveau_timer_init(&priv->base);
+	if (ret)
+		return ret;
+
+	nv_debug(priv, "time low        : 0x%08x\n", lo);
+	nv_debug(priv, "time high       : 0x%08x\n", hi);
+
+	/* restore the time before suspend */
+	nv_wr32(priv, NV04_PTIMER_TIME_1, hi);
+	nv_wr32(priv, NV04_PTIMER_TIME_0, lo);
+	return 0;
+}
+
+struct nouveau_oclass
+gk20a_timer_oclass = {
+	.handle = NV_SUBDEV(TIMER, 0xff),
+	.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_timer_ctor,
+		.dtor = nv04_timer_dtor,
+		.init = gk20a_timer_init,
+		.fini = nv04_timer_fini,
+	}
+};
diff --git a/drivers/gpu/drm/nouveau/core/subdev/timer/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/timer/nv04.c
index c0bdd10358d..240ed0b983a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/timer/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/timer/nv04.c
@@ -22,22 +22,7 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/timer.h>
-
-#define NV04_PTIMER_INTR_0      0x009100
-#define NV04_PTIMER_INTR_EN_0   0x009140
-#define NV04_PTIMER_NUMERATOR   0x009200
-#define NV04_PTIMER_DENOMINATOR 0x009210
-#define NV04_PTIMER_TIME_0      0x009400
-#define NV04_PTIMER_TIME_1      0x009410
-#define NV04_PTIMER_ALARM_0     0x009420
-
-struct nv04_timer_priv {
-	struct nouveau_timer base;
-	struct list_head alarms;
-	spinlock_t lock;
-	u64 suspend_time;
-};
+#include "nv04.h"
 
 static u64
 nv04_timer_read(struct nouveau_timer *ptimer)
@@ -142,35 +127,14 @@ nv04_timer_intr(struct nouveau_subdev *subdev)
 	}
 }
 
-static int
-nv04_timer_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		struct nouveau_oclass *oclass, void *data, u32 size,
-		struct nouveau_object **pobject)
-{
-	struct nv04_timer_priv *priv;
-	int ret;
-
-	ret = nouveau_timer_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.base.intr = nv04_timer_intr;
-	priv->base.read = nv04_timer_read;
-	priv->base.alarm = nv04_timer_alarm;
-	priv->base.alarm_cancel = nv04_timer_alarm_cancel;
-	priv->suspend_time = 0;
-
-	INIT_LIST_HEAD(&priv->alarms);
-	spin_lock_init(&priv->lock);
-	return 0;
-}
-
-static void
-nv04_timer_dtor(struct nouveau_object *object)
+int
+nv04_timer_fini(struct nouveau_object *object, bool suspend)
 {
 	struct nv04_timer_priv *priv = (void *)object;
-	return nouveau_timer_destroy(&priv->base);
+	if (suspend)
+		priv->suspend_time = nv04_timer_read(&priv->base);
+	nv_wr32(priv, NV04_PTIMER_INTR_EN_0, 0x00000000);
+	return nouveau_timer_fini(&priv->base, suspend);
 }
 
 static int
@@ -257,14 +221,35 @@ nv04_timer_init(struct nouveau_object *object)
 	return 0;
 }
 
-static int
-nv04_timer_fini(struct nouveau_object *object, bool suspend)
+void
+nv04_timer_dtor(struct nouveau_object *object)
 {
 	struct nv04_timer_priv *priv = (void *)object;
-	if (suspend)
-		priv->suspend_time = nv04_timer_read(&priv->base);
-	nv_wr32(priv, NV04_PTIMER_INTR_EN_0, 0x00000000);
-	return nouveau_timer_fini(&priv->base, suspend);
+	return nouveau_timer_destroy(&priv->base);
+}
+
+int
+nv04_timer_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+		struct nouveau_oclass *oclass, void *data, u32 size,
+		struct nouveau_object **pobject)
+{
+	struct nv04_timer_priv *priv;
+	int ret;
+
+	ret = nouveau_timer_create(parent, engine, oclass, &priv);
+	*pobject = nv_object(priv);
+	if (ret)
+		return ret;
+
+	priv->base.base.intr = nv04_timer_intr;
+	priv->base.read = nv04_timer_read;
+	priv->base.alarm = nv04_timer_alarm;
+	priv->base.alarm_cancel = nv04_timer_alarm_cancel;
+	priv->suspend_time = 0;
+
+	INIT_LIST_HEAD(&priv->alarms);
+	spin_lock_init(&priv->lock);
+	return 0;
 }
 
 struct nouveau_oclass
diff --git a/drivers/gpu/drm/nouveau/core/subdev/timer/nv04.h b/drivers/gpu/drm/nouveau/core/subdev/timer/nv04.h
new file mode 100644
index 00000000000..4bc152697c3
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/timer/nv04.h
@@ -0,0 +1,27 @@
+#ifndef __NVKM_TIMER_NV04_H__
+#define __NVKM_TIMER_NV04_H__
+
+#include "priv.h"
+
+#define NV04_PTIMER_INTR_0      0x009100
+#define NV04_PTIMER_INTR_EN_0   0x009140
+#define NV04_PTIMER_NUMERATOR   0x009200
+#define NV04_PTIMER_DENOMINATOR 0x009210
+#define NV04_PTIMER_TIME_0      0x009400
+#define NV04_PTIMER_TIME_1      0x009410
+#define NV04_PTIMER_ALARM_0     0x009420
+
+struct nv04_timer_priv {
+	struct nouveau_timer base;
+	struct list_head alarms;
+	spinlock_t lock;
+	u64 suspend_time;
+};
+
+int  nv04_timer_ctor(struct nouveau_object *, struct nouveau_object *,
+		     struct nouveau_oclass *, void *, u32,
+		     struct nouveau_object **);
+void nv04_timer_dtor(struct nouveau_object *);
+int  nv04_timer_fini(struct nouveau_object *, bool);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/timer/priv.h b/drivers/gpu/drm/nouveau/core/subdev/timer/priv.h
new file mode 100644
index 00000000000..799dae3f230
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/timer/priv.h
@@ -0,0 +1,6 @@
+#ifndef __NVKM_TIMER_PRIV_H__
+#define __NVKM_TIMER_PRIV_H__
+
+#include <subdev/timer.h>
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/vm/base.c b/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
index ef3133e7575..7dd680ff2f6 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
@@ -72,13 +72,7 @@ nouveau_vm_map_at(struct nouveau_vma *vma, u64 delta, struct nouveau_mem *node)
 	vmm->flush(vm);
 }
 
-void
-nouveau_vm_map(struct nouveau_vma *vma, struct nouveau_mem *node)
-{
-	nouveau_vm_map_at(vma, 0, node);
-}
-
-void
+static void
 nouveau_vm_map_sg_table(struct nouveau_vma *vma, u64 delta, u64 length,
 			struct nouveau_mem *mem)
 {
@@ -136,7 +130,7 @@ finish:
 	vmm->flush(vm);
 }
 
-void
+static void
 nouveau_vm_map_sg(struct nouveau_vma *vma, u64 delta, u64 length,
 		  struct nouveau_mem *mem)
 {
@@ -175,6 +169,18 @@ nouveau_vm_map_sg(struct nouveau_vma *vma, u64 delta, u64 length,
 }
 
 void
+nouveau_vm_map(struct nouveau_vma *vma, struct nouveau_mem *node)
+{
+	if (node->sg)
+		nouveau_vm_map_sg_table(vma, 0, node->size << 12, node);
+	else
+	if (node->pages)
+		nouveau_vm_map_sg(vma, 0, node->size << 12, node);
+	else
+		nouveau_vm_map_at(vma, 0, node);
+}
+
+void
 nouveau_vm_unmap_at(struct nouveau_vma *vma, u64 delta, u64 length)
 {
 	struct nouveau_vm *vm = vma->vm;
diff --git a/drivers/gpu/drm/nouveau/dispnv04/crtc.c b/drivers/gpu/drm/nouveau/dispnv04/crtc.c
index 0e3270c3ffd..41be3424c90 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/crtc.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/crtc.c
@@ -239,7 +239,7 @@ nv_crtc_mode_set_vga(struct drm_crtc *crtc, struct drm_display_mode *mode)
 	struct drm_device *dev = crtc->dev;
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
 	struct nv04_crtc_reg *regp = &nv04_display(dev)->mode_reg.crtc_reg[nv_crtc->index];
-	struct drm_framebuffer *fb = crtc->fb;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 
 	/* Calculate our timings */
 	int horizDisplay	= (mode->crtc_hdisplay >> 3)		- 1;
@@ -574,7 +574,7 @@ nv_crtc_mode_set_regs(struct drm_crtc *crtc, struct drm_display_mode * mode)
 		regp->CRTC[NV_CIO_CRE_86] = 0x1;
 	}
 
-	regp->CRTC[NV_CIO_CRE_PIXEL_INDEX] = (crtc->fb->depth + 1) / 8;
+	regp->CRTC[NV_CIO_CRE_PIXEL_INDEX] = (crtc->primary->fb->depth + 1) / 8;
 	/* Enable slaved mode (called MODE_TV in nv4ref.h) */
 	if (lvds_output || tmds_output || tv_output)
 		regp->CRTC[NV_CIO_CRE_PIXEL_INDEX] |= (1 << 7);
@@ -588,7 +588,7 @@ nv_crtc_mode_set_regs(struct drm_crtc *crtc, struct drm_display_mode * mode)
 	regp->ramdac_gen_ctrl = NV_PRAMDAC_GENERAL_CONTROL_BPC_8BITS |
 				NV_PRAMDAC_GENERAL_CONTROL_VGA_STATE_SEL |
 				NV_PRAMDAC_GENERAL_CONTROL_PIXMIX_ON;
-	if (crtc->fb->depth == 16)
+	if (crtc->primary->fb->depth == 16)
 		regp->ramdac_gen_ctrl |= NV_PRAMDAC_GENERAL_CONTROL_ALT_MODE_SEL;
 	if (nv_device(drm->device)->chipset >= 0x11)
 		regp->ramdac_gen_ctrl |= NV_PRAMDAC_GENERAL_CONTROL_PIPE_LONG;
@@ -609,7 +609,7 @@ static int
 nv_crtc_swap_fbs(struct drm_crtc *crtc, struct drm_framebuffer *old_fb)
 {
 	struct nv04_display *disp = nv04_display(crtc->dev);
-	struct nouveau_framebuffer *nvfb = nouveau_framebuffer(crtc->fb);
+	struct nouveau_framebuffer *nvfb = nouveau_framebuffer(crtc->primary->fb);
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
 	int ret;
 
@@ -808,7 +808,7 @@ nv_crtc_gamma_set(struct drm_crtc *crtc, u16 *r, u16 *g, u16 *b, uint32_t start,
 	 * mark the lut values as dirty by setting depth==0, and it'll be
 	 * uploaded on the first mode_set_base()
 	 */
-	if (!nv_crtc->base.fb) {
+	if (!nv_crtc->base.primary->fb) {
 		nv_crtc->lut.depth = 0;
 		return;
 	}
@@ -832,7 +832,7 @@ nv04_crtc_do_mode_set_base(struct drm_crtc *crtc,
 	NV_DEBUG(drm, "index %d\n", nv_crtc->index);
 
 	/* no fb bound */
-	if (!atomic && !crtc->fb) {
+	if (!atomic && !crtc->primary->fb) {
 		NV_DEBUG(drm, "No FB bound\n");
 		return 0;
 	}
@@ -844,8 +844,8 @@ nv04_crtc_do_mode_set_base(struct drm_crtc *crtc,
 		drm_fb = passed_fb;
 		fb = nouveau_framebuffer(passed_fb);
 	} else {
-		drm_fb = crtc->fb;
-		fb = nouveau_framebuffer(crtc->fb);
+		drm_fb = crtc->primary->fb;
+		fb = nouveau_framebuffer(crtc->primary->fb);
 	}
 
 	nv_crtc->fb.offset = fb->nvbo->bo.offset;
@@ -857,9 +857,9 @@ nv04_crtc_do_mode_set_base(struct drm_crtc *crtc,
 
 	/* Update the framebuffer format. */
 	regp->CRTC[NV_CIO_CRE_PIXEL_INDEX] &= ~3;
-	regp->CRTC[NV_CIO_CRE_PIXEL_INDEX] |= (crtc->fb->depth + 1) / 8;
+	regp->CRTC[NV_CIO_CRE_PIXEL_INDEX] |= (crtc->primary->fb->depth + 1) / 8;
 	regp->ramdac_gen_ctrl &= ~NV_PRAMDAC_GENERAL_CONTROL_ALT_MODE_SEL;
-	if (crtc->fb->depth == 16)
+	if (crtc->primary->fb->depth == 16)
 		regp->ramdac_gen_ctrl |= NV_PRAMDAC_GENERAL_CONTROL_ALT_MODE_SEL;
 	crtc_wr_cio_state(crtc, regp, NV_CIO_CRE_PIXEL_INDEX);
 	NVWriteRAMDAC(dev, nv_crtc->index, NV_PRAMDAC_GENERAL_CONTROL,
@@ -1048,7 +1048,7 @@ nouveau_crtc_set_config(struct drm_mode_set *set)
 
 	/* get a pm reference here */
 	ret = pm_runtime_get_sync(dev->dev);
-	if (ret < 0)
+	if (ret < 0 && ret != -EACCES)
 		return ret;
 
 	ret = drm_crtc_helper_set_config(set);
diff --git a/drivers/gpu/drm/nouveau/dispnv04/dac.c b/drivers/gpu/drm/nouveau/dispnv04/dac.c
index 434b920f6bd..a96dda48718 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/dac.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/dac.c
@@ -414,7 +414,7 @@ static void nv04_dac_commit(struct drm_encoder *encoder)
 	helper->dpms(encoder, DRM_MODE_DPMS_ON);
 
 	NV_DEBUG(drm, "Output %s is running on CRTC %d using output %c\n",
-		 drm_get_connector_name(&nouveau_encoder_connector_get(nv_encoder)->base),
+		 nouveau_encoder_connector_get(nv_encoder)->base.name,
 		 nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
 }
 
diff --git a/drivers/gpu/drm/nouveau/dispnv04/dfp.c b/drivers/gpu/drm/nouveau/dispnv04/dfp.c
index 7fdc51e2a57..e57babb206d 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/dfp.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/dfp.c
@@ -415,7 +415,7 @@ static void nv04_dfp_mode_set(struct drm_encoder *encoder,
 	/* Output property. */
 	if ((nv_connector->dithering_mode == DITHERING_MODE_ON) ||
 	    (nv_connector->dithering_mode == DITHERING_MODE_AUTO &&
-	     encoder->crtc->fb->depth > connector->display_info.bpc * 3)) {
+	     encoder->crtc->primary->fb->depth > connector->display_info.bpc * 3)) {
 		if (nv_device(drm->device)->chipset == 0x11)
 			regp->dither = savep->dither | 0x00010000;
 		else {
@@ -477,7 +477,7 @@ static void nv04_dfp_commit(struct drm_encoder *encoder)
 	helper->dpms(encoder, DRM_MODE_DPMS_ON);
 
 	NV_DEBUG(drm, "Output %s is running on CRTC %d using output %c\n",
-		 drm_get_connector_name(&nouveau_encoder_connector_get(nv_encoder)->base),
+		 nouveau_encoder_connector_get(nv_encoder)->base.name,
 		 nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
 }
 
diff --git a/drivers/gpu/drm/nouveau/dispnv04/disp.c b/drivers/gpu/drm/nouveau/dispnv04/disp.c
index b13ff0fc42d..4342fdaee70 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/disp.c
@@ -77,11 +77,6 @@ nv04_display_create(struct drm_device *dev)
 
 	nouveau_hw_save_vga_fonts(dev, 1);
 
-	ret = nouveau_object_new(nv_object(drm), NVDRM_DEVICE, 0xd1500000,
-				 NV04_DISP_CLASS, NULL, 0, &disp->core);
-	if (ret)
-		return ret;
-
 	nv04_crtc_create(dev, 0);
 	if (nv_two_heads(dev))
 		nv04_crtc_create(dev, 1);
@@ -120,7 +115,7 @@ nv04_display_create(struct drm_device *dev)
 				 &dev->mode_config.connector_list, head) {
 		if (!connector->encoder_ids[0]) {
 			NV_WARN(drm, "%s has no encoders, removing\n",
-				drm_get_connector_name(connector));
+				connector->name);
 			connector->funcs->destroy(connector);
 		}
 	}
diff --git a/drivers/gpu/drm/nouveau/dispnv04/disp.h b/drivers/gpu/drm/nouveau/dispnv04/disp.h
index 56a28db0400..4245fc3dab7 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/disp.h
+++ b/drivers/gpu/drm/nouveau/dispnv04/disp.h
@@ -80,7 +80,6 @@ struct nv04_display {
 	struct nv04_mode_state saved_reg;
 	uint32_t saved_vga_font[4][16384];
 	uint32_t dac_users[4];
-	struct nouveau_object *core;
 	struct nouveau_bo *image[2];
 };
 
diff --git a/drivers/gpu/drm/nouveau/dispnv04/overlay.c b/drivers/gpu/drm/nouveau/dispnv04/overlay.c
index 32e7064b819..ab03f7719d2 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/overlay.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/overlay.c
@@ -55,9 +55,12 @@ struct nouveau_plane {
 	int hue;
 	int saturation;
 	int iturbt_709;
+
+	void (*set_params)(struct nouveau_plane *);
 };
 
 static uint32_t formats[] = {
+	DRM_FORMAT_YUYV,
 	DRM_FORMAT_UYVY,
 	DRM_FORMAT_NV12,
 };
@@ -140,10 +143,10 @@ nv10_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	nv_wr32(dev, NV_PVIDEO_POINT_OUT(flip), crtc_y << 16 | crtc_x);
 	nv_wr32(dev, NV_PVIDEO_SIZE_OUT(flip), crtc_h << 16 | crtc_w);
 
-	if (fb->pixel_format == DRM_FORMAT_NV12) {
+	if (fb->pixel_format != DRM_FORMAT_UYVY)
 		format |= NV_PVIDEO_FORMAT_COLOR_LE_CR8YB8CB8YA8;
+	if (fb->pixel_format == DRM_FORMAT_NV12)
 		format |= NV_PVIDEO_FORMAT_PLANAR;
-	}
 	if (nv_plane->iturbt_709)
 		format |= NV_PVIDEO_FORMAT_MATRIX_ITURBT709;
 	if (nv_plane->colorkey & (1 << 24))
@@ -182,9 +185,9 @@ nv10_disable_plane(struct drm_plane *plane)
 }
 
 static void
-nv10_destroy_plane(struct drm_plane *plane)
+nv_destroy_plane(struct drm_plane *plane)
 {
-	nv10_disable_plane(plane);
+	plane->funcs->disable_plane(plane);
 	drm_plane_cleanup(plane);
 	kfree(plane);
 }
@@ -217,9 +220,9 @@ nv10_set_params(struct nouveau_plane *plane)
 }
 
 static int
-nv10_set_property(struct drm_plane *plane,
-		  struct drm_property *property,
-		  uint64_t value)
+nv_set_property(struct drm_plane *plane,
+		struct drm_property *property,
+		uint64_t value)
 {
 	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
 
@@ -238,15 +241,16 @@ nv10_set_property(struct drm_plane *plane,
 	else
 		return -EINVAL;
 
-	nv10_set_params(nv_plane);
+	if (nv_plane->set_params)
+		nv_plane->set_params(nv_plane);
 	return 0;
 }
 
 static const struct drm_plane_funcs nv10_plane_funcs = {
 	.update_plane = nv10_update_plane,
 	.disable_plane = nv10_disable_plane,
-	.set_property = nv10_set_property,
-	.destroy = nv10_destroy_plane,
+	.set_property = nv_set_property,
+	.destroy = nv_destroy_plane,
 };
 
 static void
@@ -266,7 +270,7 @@ nv10_overlay_init(struct drm_device *device)
 	case 0x15:
 	case 0x1a:
 	case 0x20:
-		num_formats = 1;
+		num_formats = 2;
 		break;
 	}
 
@@ -321,8 +325,159 @@ nv10_overlay_init(struct drm_device *device)
 	drm_object_attach_property(&plane->base.base,
 				   plane->props.iturbt_709, plane->iturbt_709);
 
+	plane->set_params = nv10_set_params;
 	nv10_set_params(plane);
-	nv_wr32(dev, NV_PVIDEO_STOP, 1);
+	nv10_disable_plane(&plane->base);
+	return;
+cleanup:
+	drm_plane_cleanup(&plane->base);
+err:
+	kfree(plane);
+	nv_error(dev, "Failed to create plane\n");
+}
+
+static int
+nv04_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
+		  struct drm_framebuffer *fb, int crtc_x, int crtc_y,
+		  unsigned int crtc_w, unsigned int crtc_h,
+		  uint32_t src_x, uint32_t src_y,
+		  uint32_t src_w, uint32_t src_h)
+{
+	struct nouveau_device *dev = nouveau_dev(plane->dev);
+	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_framebuffer *nv_fb = nouveau_framebuffer(fb);
+	struct nouveau_bo *cur = nv_plane->cur;
+	uint32_t overlay = 1;
+	int brightness = (nv_plane->brightness - 512) * 62 / 512;
+	int pitch, ret, i;
+
+	/* Source parameters given in 16.16 fixed point, ignore fractional. */
+	src_x >>= 16;
+	src_y >>= 16;
+	src_w >>= 16;
+	src_h >>= 16;
+
+	pitch = ALIGN(src_w * 4, 0x100);
+
+	if (pitch > 0xffff)
+		return -ERANGE;
+
+	/* TODO: Compute an offset? Not sure how to do this for YUYV. */
+	if (src_x != 0 || src_y != 0)
+		return -ERANGE;
+
+	if (crtc_w < src_w || crtc_h < src_h)
+		return -ERANGE;
+
+	ret = nouveau_bo_pin(nv_fb->nvbo, TTM_PL_FLAG_VRAM);
+	if (ret)
+		return ret;
+
+	nv_plane->cur = nv_fb->nvbo;
+
+	nv_wr32(dev, NV_PVIDEO_OE_STATE, 0);
+	nv_wr32(dev, NV_PVIDEO_SU_STATE, 0);
+	nv_wr32(dev, NV_PVIDEO_RM_STATE, 0);
+
+	for (i = 0; i < 2; i++) {
+		nv_wr32(dev, NV_PVIDEO_BUFF0_START_ADDRESS + 4 * i,
+			nv_fb->nvbo->bo.offset);
+		nv_wr32(dev, NV_PVIDEO_BUFF0_PITCH_LENGTH + 4 * i, pitch);
+		nv_wr32(dev, NV_PVIDEO_BUFF0_OFFSET + 4 * i, 0);
+	}
+	nv_wr32(dev, NV_PVIDEO_WINDOW_START, crtc_y << 16 | crtc_x);
+	nv_wr32(dev, NV_PVIDEO_WINDOW_SIZE, crtc_h << 16 | crtc_w);
+	nv_wr32(dev, NV_PVIDEO_STEP_SIZE,
+		(uint32_t)(((src_h - 1) << 11) / (crtc_h - 1)) << 16 | (uint32_t)(((src_w - 1) << 11) / (crtc_w - 1)));
+
+	/* It should be possible to convert hue/contrast to this */
+	nv_wr32(dev, NV_PVIDEO_RED_CSC_OFFSET, 0x69 - brightness);
+	nv_wr32(dev, NV_PVIDEO_GREEN_CSC_OFFSET, 0x3e + brightness);
+	nv_wr32(dev, NV_PVIDEO_BLUE_CSC_OFFSET, 0x89 - brightness);
+	nv_wr32(dev, NV_PVIDEO_CSC_ADJUST, 0);
+
+	nv_wr32(dev, NV_PVIDEO_CONTROL_Y, 0x001); /* (BLUR_ON, LINE_HALF) */
+	nv_wr32(dev, NV_PVIDEO_CONTROL_X, 0x111); /* (WEIGHT_HEAVY, SHARPENING_ON, SMOOTHING_ON) */
+
+	nv_wr32(dev, NV_PVIDEO_FIFO_BURST_LENGTH, 0x03);
+	nv_wr32(dev, NV_PVIDEO_FIFO_THRES_SIZE, 0x38);
+
+	nv_wr32(dev, NV_PVIDEO_KEY, nv_plane->colorkey);
+
+	if (nv_plane->colorkey & (1 << 24))
+		overlay |= 0x10;
+	if (fb->pixel_format == DRM_FORMAT_YUYV)
+		overlay |= 0x100;
+
+	nv_wr32(dev, NV_PVIDEO_OVERLAY, overlay);
+
+	nv_wr32(dev, NV_PVIDEO_SU_STATE, nv_rd32(dev, NV_PVIDEO_SU_STATE) ^ (1 << 16));
+
+	if (cur)
+		nouveau_bo_unpin(cur);
+
+	return 0;
+}
+
+static int
+nv04_disable_plane(struct drm_plane *plane)
+{
+	struct nouveau_device *dev = nouveau_dev(plane->dev);
+	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+
+	nv_mask(dev, NV_PVIDEO_OVERLAY, 1, 0);
+	nv_wr32(dev, NV_PVIDEO_OE_STATE, 0);
+	nv_wr32(dev, NV_PVIDEO_SU_STATE, 0);
+	nv_wr32(dev, NV_PVIDEO_RM_STATE, 0);
+	if (nv_plane->cur) {
+		nouveau_bo_unpin(nv_plane->cur);
+		nv_plane->cur = NULL;
+	}
+
+	return 0;
+}
+
+static const struct drm_plane_funcs nv04_plane_funcs = {
+	.update_plane = nv04_update_plane,
+	.disable_plane = nv04_disable_plane,
+	.set_property = nv_set_property,
+	.destroy = nv_destroy_plane,
+};
+
+static void
+nv04_overlay_init(struct drm_device *device)
+{
+	struct nouveau_device *dev = nouveau_dev(device);
+	struct nouveau_plane *plane = kzalloc(sizeof(struct nouveau_plane), GFP_KERNEL);
+	int ret;
+
+	if (!plane)
+		return;
+
+	ret = drm_plane_init(device, &plane->base, 1 /* single crtc */,
+			     &nv04_plane_funcs,
+			     formats, 2, false);
+	if (ret)
+		goto err;
+
+	/* Set up the plane properties */
+	plane->props.colorkey = drm_property_create_range(
+			device, 0, "colorkey", 0, 0x01ffffff);
+	plane->props.brightness = drm_property_create_range(
+			device, 0, "brightness", 0, 1024);
+	if (!plane->props.colorkey ||
+	    !plane->props.brightness)
+		goto cleanup;
+
+	plane->colorkey = 0;
+	drm_object_attach_property(&plane->base.base,
+				   plane->props.colorkey, plane->colorkey);
+
+	plane->brightness = 512;
+	drm_object_attach_property(&plane->base.base,
+				   plane->props.brightness, plane->brightness);
+
+	nv04_disable_plane(&plane->base);
 	return;
 cleanup:
 	drm_plane_cleanup(&plane->base);
@@ -335,6 +490,8 @@ void
 nouveau_overlay_init(struct drm_device *device)
 {
 	struct nouveau_device *dev = nouveau_dev(device);
-	if (dev->chipset >= 0x10 && dev->chipset <= 0x40)
+	if (dev->chipset < 0x10)
+		nv04_overlay_init(device);
+	else if (dev->chipset <= 0x40)
 		nv10_overlay_init(device);
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c b/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c
index 244822df8ff..8667620b703 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c
@@ -171,7 +171,8 @@ static void nv04_tv_commit(struct drm_encoder *encoder)
 	helper->dpms(encoder, DRM_MODE_DPMS_ON);
 
 	NV_DEBUG(drm, "Output %s is running on CRTC %d using output %c\n",
-		 drm_get_connector_name(&nouveau_encoder_connector_get(nv_encoder)->base), nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
+		 nouveau_encoder_connector_get(nv_encoder)->base.name,
+		 nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
 }
 
 static void nv04_tv_destroy(struct drm_encoder *encoder)
diff --git a/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c b/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c
index acef48f4a4e..195bd8e86c6 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c
@@ -612,8 +612,7 @@ static void nv17_tv_commit(struct drm_encoder *encoder)
 	helper->dpms(encoder, DRM_MODE_DPMS_ON);
 
 	NV_INFO(drm, "Output %s is running on CRTC %d using output %c\n",
-		drm_get_connector_name(
-			&nouveau_encoder_connector_get(nv_encoder)->base),
+		nouveau_encoder_connector_get(nv_encoder)->base.name,
 		nv_crtc->index, '@' + ffs(nv_encoder->dcb->or));
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_abi16.c b/drivers/gpu/drm/nouveau/nouveau_abi16.c
index 900fae01793..b13f441c643 100644
--- a/drivers/gpu/drm/nouveau/nouveau_abi16.c
+++ b/drivers/gpu/drm/nouveau/nouveau_abi16.c
@@ -97,6 +97,7 @@ nouveau_abi16_swclass(struct nouveau_drm *drm)
 	case NV_C0:
 	case NV_D0:
 	case NV_E0:
+	case GM100:
 		return 0x906e;
 	}
 
@@ -139,7 +140,7 @@ nouveau_abi16_chan_fini(struct nouveau_abi16 *abi16,
 
 	/* destroy channel object, all children will be killed too */
 	if (chan->chan) {
-		abi16->handles &= ~(1 << (chan->chan->handle & 0xffff));
+		abi16->handles &= ~(1ULL << (chan->chan->handle & 0xffff));
 		nouveau_channel_del(&chan->chan);
 	}
 
@@ -179,12 +180,21 @@ nouveau_abi16_ioctl_getparam(ABI16_IOCTL_ARGS)
 		getparam->value = device->chipset;
 		break;
 	case NOUVEAU_GETPARAM_PCI_VENDOR:
-		getparam->value = dev->pdev->vendor;
+		if (nv_device_is_pci(device))
+			getparam->value = dev->pdev->vendor;
+		else
+			getparam->value = 0;
 		break;
 	case NOUVEAU_GETPARAM_PCI_DEVICE:
-		getparam->value = dev->pdev->device;
+		if (nv_device_is_pci(device))
+			getparam->value = dev->pdev->device;
+		else
+			getparam->value = 0;
 		break;
 	case NOUVEAU_GETPARAM_BUS_TYPE:
+		if (!nv_device_is_pci(device))
+			getparam->value = 3;
+		else
 		if (drm_pci_device_is_agp(dev))
 			getparam->value = 0;
 		else
@@ -270,8 +280,8 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS)
 		return nouveau_abi16_put(abi16, -EINVAL);
 
 	/* allocate "abi16 channel" data and make up a handle for it */
-	init->channel = ffsll(~abi16->handles);
-	if (!init->channel--)
+	init->channel = __ffs64(~abi16->handles);
+	if (~abi16->handles == 0)
 		return nouveau_abi16_put(abi16, -ENOSPC);
 
 	chan = kzalloc(sizeof(*chan), GFP_KERNEL);
@@ -280,7 +290,7 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS)
 
 	INIT_LIST_HEAD(&chan->notifiers);
 	list_add(&chan->head, &abi16->channels);
-	abi16->handles |= (1 << init->channel);
+	abi16->handles |= (1ULL << init->channel);
 
 	/* create channel object and initialise dma and fence management */
 	ret = nouveau_channel_new(drm, cli, NVDRM_DEVICE, NVDRM_CHAN |
diff --git a/drivers/gpu/drm/nouveau/nouveau_acpi.c b/drivers/gpu/drm/nouveau/nouveau_acpi.c
index ba0183fb84f..279206997e5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_acpi.c
+++ b/drivers/gpu/drm/nouveau/nouveau_acpi.c
@@ -1,15 +1,10 @@
 #include <linux/pci.h>
 #include <linux/acpi.h>
 #include <linux/slab.h>
-#include <acpi/acpi_drivers.h>
-#include <acpi/acpi_bus.h>
-#include <acpi/video.h>
-#include <acpi/acpi.h>
 #include <linux/mxm-wmi.h>
-
 #include <linux/vga_switcheroo.h>
-
 #include <drm/drm_edid.h>
+#include <acpi/video.h>
 
 #include "nouveau_drm.h"
 #include "nouveau_acpi.h"
@@ -66,6 +61,7 @@ bool nouveau_is_v1_dsm(void) {
 #define NOUVEAU_DSM_HAS_MUX 0x1
 #define NOUVEAU_DSM_HAS_OPT 0x2
 
+#ifdef CONFIG_VGA_SWITCHEROO
 static const char nouveau_dsm_muid[] = {
 	0xA0, 0xA0, 0x95, 0x9D, 0x60, 0x00, 0x48, 0x4D,
 	0xB3, 0x4D, 0x7E, 0x5F, 0xEA, 0x12, 0x9F, 0xD4,
@@ -78,124 +74,89 @@ static const char nouveau_op_dsm_muid[] = {
 
 static int nouveau_optimus_dsm(acpi_handle handle, int func, int arg, uint32_t *result)
 {
-	struct acpi_buffer output = { ACPI_ALLOCATE_BUFFER, NULL };
-	struct acpi_object_list input;
-	union acpi_object params[4];
+	int i;
 	union acpi_object *obj;
-	int i, err;
 	char args_buff[4];
+	union acpi_object argv4 = {
+		.buffer.type = ACPI_TYPE_BUFFER,
+		.buffer.length = 4,
+		.buffer.pointer = args_buff
+	};
 
-	input.count = 4;
-	input.pointer = params;
-	params[0].type = ACPI_TYPE_BUFFER;
-	params[0].buffer.length = sizeof(nouveau_op_dsm_muid);
-	params[0].buffer.pointer = (char *)nouveau_op_dsm_muid;
-	params[1].type = ACPI_TYPE_INTEGER;
-	params[1].integer.value = 0x00000100;
-	params[2].type = ACPI_TYPE_INTEGER;
-	params[2].integer.value = func;
-	params[3].type = ACPI_TYPE_BUFFER;
-	params[3].buffer.length = 4;
 	/* ACPI is little endian, AABBCCDD becomes {DD,CC,BB,AA} */
 	for (i = 0; i < 4; i++)
 		args_buff[i] = (arg >> i * 8) & 0xFF;
-	params[3].buffer.pointer = args_buff;
-
-	err = acpi_evaluate_object(handle, "_DSM", &input, &output);
-	if (err) {
-		printk(KERN_INFO "failed to evaluate _DSM: %d\n", err);
-		return err;
-	}
-
-	obj = (union acpi_object *)output.pointer;
-
-	if (obj->type == ACPI_TYPE_INTEGER)
-		if (obj->integer.value == 0x80000002) {
-			return -ENODEV;
-		}
 
-	if (obj->type == ACPI_TYPE_BUFFER) {
-		if (obj->buffer.length == 4 && result) {
-			*result = 0;
+	*result = 0;
+	obj = acpi_evaluate_dsm_typed(handle, nouveau_op_dsm_muid, 0x00000100,
+				      func, &argv4, ACPI_TYPE_BUFFER);
+	if (!obj) {
+		acpi_handle_info(handle, "failed to evaluate _DSM\n");
+		return AE_ERROR;
+	} else {
+		if (obj->buffer.length == 4) {
 			*result |= obj->buffer.pointer[0];
 			*result |= (obj->buffer.pointer[1] << 8);
 			*result |= (obj->buffer.pointer[2] << 16);
 			*result |= (obj->buffer.pointer[3] << 24);
 		}
+		ACPI_FREE(obj);
 	}
 
-	kfree(output.pointer);
 	return 0;
 }
 
-static int nouveau_dsm(acpi_handle handle, int func, int arg, uint32_t *result)
+/*
+ * On some platforms, _DSM(nouveau_op_dsm_muid, func0) has special
+ * requirements on the fourth parameter, so a private implementation
+ * instead of using acpi_check_dsm().
+ */
+static int nouveau_check_optimus_dsm(acpi_handle handle)
 {
-	struct acpi_buffer output = { ACPI_ALLOCATE_BUFFER, NULL };
-	struct acpi_object_list input;
-	union acpi_object params[4];
-	union acpi_object *obj;
-	int err;
-
-	input.count = 4;
-	input.pointer = params;
-	params[0].type = ACPI_TYPE_BUFFER;
-	params[0].buffer.length = sizeof(nouveau_dsm_muid);
-	params[0].buffer.pointer = (char *)nouveau_dsm_muid;
-	params[1].type = ACPI_TYPE_INTEGER;
-	params[1].integer.value = 0x00000102;
-	params[2].type = ACPI_TYPE_INTEGER;
-	params[2].integer.value = func;
-	params[3].type = ACPI_TYPE_INTEGER;
-	params[3].integer.value = arg;
-
-	err = acpi_evaluate_object(handle, "_DSM", &input, &output);
-	if (err) {
-		printk(KERN_INFO "failed to evaluate _DSM: %d\n", err);
-		return err;
-	}
-
-	obj = (union acpi_object *)output.pointer;
+	int result;
 
-	if (obj->type == ACPI_TYPE_INTEGER)
-		if (obj->integer.value == 0x80000002)
-			return -ENODEV;
-
-	if (obj->type == ACPI_TYPE_BUFFER) {
-		if (obj->buffer.length == 4 && result) {
-			*result = 0;
-			*result |= obj->buffer.pointer[0];
-			*result |= (obj->buffer.pointer[1] << 8);
-			*result |= (obj->buffer.pointer[2] << 16);
-			*result |= (obj->buffer.pointer[3] << 24);
-		}
-	}
+	/*
+	 * Function 0 returns a Buffer containing available functions.
+	 * The args parameter is ignored for function 0, so just put 0 in it
+	 */
+	if (nouveau_optimus_dsm(handle, 0, 0, &result))
+		return 0;
 
-	kfree(output.pointer);
-	return 0;
+	/*
+	 * ACPI Spec v4 9.14.1: if bit 0 is zero, no function is supported.
+	 * If the n-th bit is enabled, function n is supported
+	 */
+	return result & 1 && result & (1 << NOUVEAU_DSM_OPTIMUS_CAPS);
 }
 
-/* Returns 1 if a DSM function is usable and 0 otherwise */
-static int nouveau_test_dsm(acpi_handle test_handle,
-	int (*dsm_func)(acpi_handle, int, int, uint32_t *),
-	int sfnc)
+static int nouveau_dsm(acpi_handle handle, int func, int arg)
 {
-	u32 result = 0;
-
-	/* Function 0 returns a Buffer containing available functions. The args
-	 * parameter is ignored for function 0, so just put 0 in it */
-	if (dsm_func(test_handle, 0, 0, &result))
-		return 0;
+	int ret = 0;
+	union acpi_object *obj;
+	union acpi_object argv4 = {
+		.integer.type = ACPI_TYPE_INTEGER,
+		.integer.value = arg,
+	};
+
+	obj = acpi_evaluate_dsm_typed(handle, nouveau_dsm_muid, 0x00000102,
+				      func, &argv4, ACPI_TYPE_INTEGER);
+	if (!obj) {
+		acpi_handle_info(handle, "failed to evaluate _DSM\n");
+		return AE_ERROR;
+	} else {
+		if (obj->integer.value == 0x80000002)
+			ret = -ENODEV;
+		ACPI_FREE(obj);
+	}
 
-	/* ACPI Spec v4 9.14.1: if bit 0 is zero, no function is supported. If
-	 * the n-th bit is enabled, function n is supported */
-	return result & 1 && result & (1 << sfnc);
+	return ret;
 }
 
 static int nouveau_dsm_switch_mux(acpi_handle handle, int mux_id)
 {
 	mxm_wmi_call_mxmx(mux_id == NOUVEAU_DSM_LED_STAMINA ? MXM_MXDS_ADAPTER_IGD : MXM_MXDS_ADAPTER_0);
 	mxm_wmi_call_mxds(mux_id == NOUVEAU_DSM_LED_STAMINA ? MXM_MXDS_ADAPTER_IGD : MXM_MXDS_ADAPTER_0);
-	return nouveau_dsm(handle, NOUVEAU_DSM_LED, mux_id, NULL);
+	return nouveau_dsm(handle, NOUVEAU_DSM_LED, mux_id);
 }
 
 static int nouveau_dsm_set_discrete_state(acpi_handle handle, enum vga_switcheroo_state state)
@@ -205,7 +166,7 @@ static int nouveau_dsm_set_discrete_state(acpi_handle handle, enum vga_switchero
 		arg = NOUVEAU_DSM_POWER_SPEED;
 	else
 		arg = NOUVEAU_DSM_POWER_STAMINA;
-	nouveau_dsm(handle, NOUVEAU_DSM_POWER, arg, NULL);
+	nouveau_dsm(handle, NOUVEAU_DSM_POWER, arg);
 	return 0;
 }
 
@@ -265,11 +226,11 @@ static int nouveau_dsm_pci_probe(struct pci_dev *pdev)
 		nouveau_dsm_priv.other_handle = dhandle;
 		return false;
 	}
-	if (nouveau_test_dsm(dhandle, nouveau_dsm, NOUVEAU_DSM_POWER))
+	if (acpi_check_dsm(dhandle, nouveau_dsm_muid, 0x00000102,
+			   1 << NOUVEAU_DSM_POWER))
 		retval |= NOUVEAU_DSM_HAS_MUX;
 
-	if (nouveau_test_dsm(dhandle, nouveau_optimus_dsm,
-		NOUVEAU_DSM_OPTIMUS_CAPS))
+	if (nouveau_check_optimus_dsm(dhandle))
 		retval |= NOUVEAU_DSM_HAS_OPT;
 
 	if (retval & NOUVEAU_DSM_HAS_OPT) {
@@ -388,6 +349,11 @@ void nouveau_unregister_dsm_handler(void)
 	if (nouveau_dsm_priv.optimus_detected || nouveau_dsm_priv.dsm_detected)
 		vga_switcheroo_unregister_handler();
 }
+#else
+void nouveau_register_dsm_handler(void) {}
+void nouveau_unregister_dsm_handler(void) {}
+void nouveau_switcheroo_optimus_dsm(void) {}
+#endif
 
 /* retrieve the ROM in 4k blocks */
 static int nouveau_rom_call(acpi_handle rom_handle, uint8_t *bios,
@@ -423,9 +389,6 @@ bool nouveau_acpi_rom_supported(struct pci_dev *pdev)
 	acpi_status status;
 	acpi_handle dhandle, rom_handle;
 
-	if (!nouveau_dsm_priv.dsm_detected && !nouveau_dsm_priv.optimus_detected)
-		return false;
-
 	dhandle = ACPI_HANDLE(&pdev->dev);
 	if (!dhandle)
 		return false;
diff --git a/drivers/gpu/drm/nouveau/nouveau_agp.c b/drivers/gpu/drm/nouveau/nouveau_agp.c
index 2953c4e91e1..51666daddb9 100644
--- a/drivers/gpu/drm/nouveau/nouveau_agp.c
+++ b/drivers/gpu/drm/nouveau/nouveau_agp.c
@@ -75,7 +75,7 @@ nouveau_agp_enabled(struct nouveau_drm *drm)
 {
 	struct drm_device *dev = drm->dev;
 
-	if (!drm_pci_device_is_agp(dev) || !dev->agp)
+	if (!dev->pdev || !drm_pci_device_is_agp(dev) || !dev->agp)
 		return false;
 
 	if (drm->agp.stat == UNKNOWN) {
diff --git a/drivers/gpu/drm/nouveau/nouveau_backlight.c b/drivers/gpu/drm/nouveau/nouveau_backlight.c
index 630f6e84fc0..2c1e4aad7da 100644
--- a/drivers/gpu/drm/nouveau/nouveau_backlight.c
+++ b/drivers/gpu/drm/nouveau/nouveau_backlight.c
@@ -31,7 +31,6 @@
  */
 
 #include <linux/backlight.h>
-#include <linux/acpi.h>
 
 #include "nouveau_drm.h"
 #include "nouveau_reg.h"
@@ -222,14 +221,6 @@ nouveau_backlight_init(struct drm_device *dev)
 	struct nouveau_device *device = nv_device(drm->device);
 	struct drm_connector *connector;
 
-#ifdef CONFIG_ACPI
-	if (acpi_video_backlight_support()) {
-		NV_INFO(drm, "ACPI backlight interface available, "
-			     "not registering our own\n");
-		return 0;
-	}
-#endif
-
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
 		if (connector->connector_type != DRM_MODE_CONNECTOR_LVDS &&
 		    connector->connector_type != DRM_MODE_CONNECTOR_eDP)
diff --git a/drivers/gpu/drm/nouveau/nouveau_bios.c b/drivers/gpu/drm/nouveau/nouveau_bios.c
index 4c3feaaa103..8268a4ccac1 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bios.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bios.c
@@ -1474,9 +1474,12 @@ parse_dcb20_entry(struct drm_device *dev, struct dcb_table *dcb,
 		case 0:
 			entry->dpconf.link_bw = 162000;
 			break;
-		default:
+		case 1:
 			entry->dpconf.link_bw = 270000;
 			break;
+		default:
+			entry->dpconf.link_bw = 540000;
+			break;
 		}
 		switch ((conf & 0x0f000000) >> 24) {
 		case 0xf:
@@ -2069,6 +2072,10 @@ nouveau_bios_init(struct drm_device *dev)
 	struct nvbios *bios = &drm->vbios;
 	int ret;
 
+	/* only relevant for PCI devices */
+	if (!dev->pdev)
+		return 0;
+
 	if (!NVInitVBIOS(dev))
 		return -ENODEV;
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index c0fde6b9393..b6dc85c614b 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -560,28 +560,6 @@ nouveau_bo_evict_flags(struct ttm_buffer_object *bo, struct ttm_placement *pl)
 }
 
 
-/* GPU-assisted copy using NV_MEMORY_TO_MEMORY_FORMAT, can access
- * TTM_PL_{VRAM,TT} directly.
- */
-
-static int
-nouveau_bo_move_accel_cleanup(struct nouveau_channel *chan,
-			      struct nouveau_bo *nvbo, bool evict,
-			      bool no_wait_gpu, struct ttm_mem_reg *new_mem)
-{
-	struct nouveau_fence *fence = NULL;
-	int ret;
-
-	ret = nouveau_fence_new(chan, false, &fence);
-	if (ret)
-		return ret;
-
-	ret = ttm_bo_move_accel_cleanup(&nvbo->bo, fence, evict,
-					no_wait_gpu, new_mem);
-	nouveau_fence_unref(&fence);
-	return ret;
-}
-
 static int
 nve0_bo_move_init(struct nouveau_channel *chan, u32 handle)
 {
@@ -798,25 +776,25 @@ nv50_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
 		  struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem)
 {
 	struct nouveau_mem *node = old_mem->mm_node;
-	struct nouveau_bo *nvbo = nouveau_bo(bo);
 	u64 length = (new_mem->num_pages << PAGE_SHIFT);
 	u64 src_offset = node->vma[0].offset;
 	u64 dst_offset = node->vma[1].offset;
+	int src_tiled = !!node->memtype;
+	int dst_tiled = !!((struct nouveau_mem *)new_mem->mm_node)->memtype;
 	int ret;
 
 	while (length) {
 		u32 amount, stride, height;
 
+		ret = RING_SPACE(chan, 18 + 6 * (src_tiled + dst_tiled));
+		if (ret)
+			return ret;
+
 		amount  = min(length, (u64)(4 * 1024 * 1024));
 		stride  = 16 * 4;
 		height  = amount / stride;
 
-		if (old_mem->mem_type == TTM_PL_VRAM &&
-		    nouveau_bo_tile_layout(nvbo)) {
-			ret = RING_SPACE(chan, 8);
-			if (ret)
-				return ret;
-
+		if (src_tiled) {
 			BEGIN_NV04(chan, NvSubCopy, 0x0200, 7);
 			OUT_RING  (chan, 0);
 			OUT_RING  (chan, 0);
@@ -826,19 +804,10 @@ nv50_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
 			OUT_RING  (chan, 0);
 			OUT_RING  (chan, 0);
 		} else {
-			ret = RING_SPACE(chan, 2);
-			if (ret)
-				return ret;
-
 			BEGIN_NV04(chan, NvSubCopy, 0x0200, 1);
 			OUT_RING  (chan, 1);
 		}
-		if (new_mem->mem_type == TTM_PL_VRAM &&
-		    nouveau_bo_tile_layout(nvbo)) {
-			ret = RING_SPACE(chan, 8);
-			if (ret)
-				return ret;
-
+		if (dst_tiled) {
 			BEGIN_NV04(chan, NvSubCopy, 0x021c, 7);
 			OUT_RING  (chan, 0);
 			OUT_RING  (chan, 0);
@@ -848,18 +817,10 @@ nv50_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
 			OUT_RING  (chan, 0);
 			OUT_RING  (chan, 0);
 		} else {
-			ret = RING_SPACE(chan, 2);
-			if (ret)
-				return ret;
-
 			BEGIN_NV04(chan, NvSubCopy, 0x021c, 1);
 			OUT_RING  (chan, 1);
 		}
 
-		ret = RING_SPACE(chan, 14);
-		if (ret)
-			return ret;
-
 		BEGIN_NV04(chan, NvSubCopy, 0x0238, 2);
 		OUT_RING  (chan, upper_32_bits(src_offset));
 		OUT_RING  (chan, upper_32_bits(dst_offset));
@@ -953,23 +914,28 @@ nv04_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
 }
 
 static int
-nouveau_vma_getmap(struct nouveau_channel *chan, struct nouveau_bo *nvbo,
-		   struct ttm_mem_reg *mem, struct nouveau_vma *vma)
+nouveau_bo_move_prep(struct nouveau_drm *drm, struct ttm_buffer_object *bo,
+		     struct ttm_mem_reg *mem)
 {
-	struct nouveau_mem *node = mem->mm_node;
+	struct nouveau_mem *old_node = bo->mem.mm_node;
+	struct nouveau_mem *new_node = mem->mm_node;
+	u64 size = (u64)mem->num_pages << PAGE_SHIFT;
 	int ret;
 
-	ret = nouveau_vm_get(nv_client(chan->cli)->vm, mem->num_pages <<
-			     PAGE_SHIFT, node->page_shift,
-			     NV_MEM_ACCESS_RW, vma);
+	ret = nouveau_vm_get(nv_client(drm)->vm, size, old_node->page_shift,
+			     NV_MEM_ACCESS_RW, &old_node->vma[0]);
 	if (ret)
 		return ret;
 
-	if (mem->mem_type == TTM_PL_VRAM)
-		nouveau_vm_map(vma, node);
-	else
-		nouveau_vm_map_sg(vma, 0, mem->num_pages << PAGE_SHIFT, node);
+	ret = nouveau_vm_get(nv_client(drm)->vm, size, new_node->page_shift,
+			     NV_MEM_ACCESS_RW, &old_node->vma[1]);
+	if (ret) {
+		nouveau_vm_put(&old_node->vma[0]);
+		return ret;
+	}
 
+	nouveau_vm_map(&old_node->vma[0], old_node);
+	nouveau_vm_map(&old_node->vma[1], new_node);
 	return 0;
 }
 
@@ -979,35 +945,34 @@ nouveau_bo_move_m2mf(struct ttm_buffer_object *bo, int evict, bool intr,
 {
 	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct nouveau_channel *chan = drm->ttm.chan;
-	struct nouveau_bo *nvbo = nouveau_bo(bo);
-	struct ttm_mem_reg *old_mem = &bo->mem;
+	struct nouveau_fence *fence;
 	int ret;
 
-	mutex_lock_nested(&chan->cli->mutex, SINGLE_DEPTH_NESTING);
-
 	/* create temporary vmas for the transfer and attach them to the
 	 * old nouveau_mem node, these will get cleaned up after ttm has
 	 * destroyed the ttm_mem_reg
 	 */
 	if (nv_device(drm->device)->card_type >= NV_50) {
-		struct nouveau_mem *node = old_mem->mm_node;
-
-		ret = nouveau_vma_getmap(chan, nvbo, old_mem, &node->vma[0]);
-		if (ret)
-			goto out;
-
-		ret = nouveau_vma_getmap(chan, nvbo, new_mem, &node->vma[1]);
+		ret = nouveau_bo_move_prep(drm, bo, new_mem);
 		if (ret)
-			goto out;
+			return ret;
 	}
 
-	ret = drm->ttm.move(chan, bo, &bo->mem, new_mem);
+	mutex_lock_nested(&chan->cli->mutex, SINGLE_DEPTH_NESTING);
+	ret = nouveau_fence_sync(bo->sync_obj, chan);
 	if (ret == 0) {
-		ret = nouveau_bo_move_accel_cleanup(chan, nvbo, evict,
-						    no_wait_gpu, new_mem);
+		ret = drm->ttm.move(chan, bo, &bo->mem, new_mem);
+		if (ret == 0) {
+			ret = nouveau_fence_new(chan, false, &fence);
+			if (ret == 0) {
+				ret = ttm_bo_move_accel_cleanup(bo, fence,
+								evict,
+								no_wait_gpu,
+								new_mem);
+				nouveau_fence_unref(&fence);
+			}
+		}
 	}
-
-out:
 	mutex_unlock(&chan->cli->mutex);
 	return ret;
 }
@@ -1147,19 +1112,10 @@ nouveau_bo_move_ntfy(struct ttm_buffer_object *bo, struct ttm_mem_reg *new_mem)
 		return;
 
 	list_for_each_entry(vma, &nvbo->vma_list, head) {
-		if (new_mem && new_mem->mem_type == TTM_PL_VRAM) {
+		if (new_mem && new_mem->mem_type != TTM_PL_SYSTEM &&
+			      (new_mem->mem_type == TTM_PL_VRAM ||
+			       nvbo->page_shift != vma->vm->vmm->lpg_shift)) {
 			nouveau_vm_map(vma, new_mem->mm_node);
-		} else
-		if (new_mem && new_mem->mem_type == TTM_PL_TT &&
-		    nvbo->page_shift == vma->vm->vmm->spg_shift) {
-			if (((struct nouveau_mem *)new_mem->mm_node)->sg)
-				nouveau_vm_map_sg_table(vma, 0, new_mem->
-						  num_pages << PAGE_SHIFT,
-						  new_mem->mm_node);
-			else
-				nouveau_vm_map_sg(vma, 0, new_mem->
-						  num_pages << PAGE_SHIFT,
-						  new_mem->mm_node);
 		} else {
 			nouveau_vm_unmap(vma);
 		}
@@ -1224,28 +1180,27 @@ nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr,
 		goto out;
 	}
 
-	/* CPU copy if we have no accelerated method available */
-	if (!drm->ttm.move) {
-		ret = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
-		goto out;
-	}
-
 	/* Hardware assisted copy. */
-	if (new_mem->mem_type == TTM_PL_SYSTEM)
-		ret = nouveau_bo_move_flipd(bo, evict, intr,
-					    no_wait_gpu, new_mem);
-	else if (old_mem->mem_type == TTM_PL_SYSTEM)
-		ret = nouveau_bo_move_flips(bo, evict, intr,
-					    no_wait_gpu, new_mem);
-	else
-		ret = nouveau_bo_move_m2mf(bo, evict, intr,
-					   no_wait_gpu, new_mem);
-
-	if (!ret)
-		goto out;
+	if (drm->ttm.move) {
+		if (new_mem->mem_type == TTM_PL_SYSTEM)
+			ret = nouveau_bo_move_flipd(bo, evict, intr,
+						    no_wait_gpu, new_mem);
+		else if (old_mem->mem_type == TTM_PL_SYSTEM)
+			ret = nouveau_bo_move_flips(bo, evict, intr,
+						    no_wait_gpu, new_mem);
+		else
+			ret = nouveau_bo_move_m2mf(bo, evict, intr,
+						   no_wait_gpu, new_mem);
+		if (!ret)
+			goto out;
+	}
 
 	/* Fallback to software copy. */
-	ret = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
+	spin_lock(&bo->bdev->fence_lock);
+	ret = ttm_bo_wait(bo, true, intr, no_wait_gpu);
+	spin_unlock(&bo->bdev->fence_lock);
+	if (ret == 0)
+		ret = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
 
 out:
 	if (nv_device(drm->device)->card_type < NV_50) {
@@ -1271,6 +1226,7 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem)
 {
 	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
 	struct nouveau_drm *drm = nouveau_bdev(bdev);
+	struct nouveau_mem *node = mem->mm_node;
 	struct drm_device *dev = drm->dev;
 	int ret;
 
@@ -1293,14 +1249,16 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem)
 			mem->bus.is_iomem = !dev->agp->cant_use_aperture;
 		}
 #endif
-		break;
+		if (nv_device(drm->device)->card_type < NV_50 || !node->memtype)
+			/* untiled */
+			break;
+		/* fallthrough, tiled memory */
 	case TTM_PL_VRAM:
 		mem->bus.offset = mem->start << PAGE_SHIFT;
-		mem->bus.base = pci_resource_start(dev->pdev, 1);
+		mem->bus.base = nv_device_resource_start(nouveau_dev(dev), 1);
 		mem->bus.is_iomem = true;
 		if (nv_device(drm->device)->card_type >= NV_50) {
 			struct nouveau_bar *bar = nouveau_bar(drm->device);
-			struct nouveau_mem *node = mem->mm_node;
 
 			ret = bar->umap(bar, node, NV_MEM_ACCESS_RW,
 					&node->bar_vma);
@@ -1335,7 +1293,8 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
 	struct nouveau_device *device = nv_device(drm->device);
-	u32 mappable = pci_resource_len(device->pdev, 1) >> PAGE_SHIFT;
+	u32 mappable = nv_device_resource_len(device, 1) >> PAGE_SHIFT;
+	int ret;
 
 	/* as long as the bo isn't in vram, and isn't tiled, we've got
 	 * nothing to do here.
@@ -1344,10 +1303,20 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 		if (nv_device(drm->device)->card_type < NV_50 ||
 		    !nouveau_bo_tile_layout(nvbo))
 			return 0;
+
+		if (bo->mem.mem_type == TTM_PL_SYSTEM) {
+			nouveau_bo_placement_set(nvbo, TTM_PL_TT, 0);
+
+			ret = nouveau_bo_validate(nvbo, false, false);
+			if (ret)
+				return ret;
+		}
+		return 0;
 	}
 
 	/* make sure bo is in mappable vram */
-	if (bo->mem.start + bo->mem.num_pages < mappable)
+	if (nv_device(drm->device)->card_type >= NV_50 ||
+	    bo->mem.start + bo->mem.num_pages < mappable)
 		return 0;
 
 
@@ -1362,6 +1331,7 @@ nouveau_ttm_tt_populate(struct ttm_tt *ttm)
 {
 	struct ttm_dma_tt *ttm_dma = (void *)ttm;
 	struct nouveau_drm *drm;
+	struct nouveau_device *device;
 	struct drm_device *dev;
 	unsigned i;
 	int r;
@@ -1379,6 +1349,7 @@ nouveau_ttm_tt_populate(struct ttm_tt *ttm)
 	}
 
 	drm = nouveau_bdev(ttm->bdev);
+	device = nv_device(drm->device);
 	dev = drm->dev;
 
 #if __OS_HAS_AGP
@@ -1399,13 +1370,12 @@ nouveau_ttm_tt_populate(struct ttm_tt *ttm)
 	}
 
 	for (i = 0; i < ttm->num_pages; i++) {
-		ttm_dma->dma_address[i] = pci_map_page(dev->pdev, ttm->pages[i],
-						   0, PAGE_SIZE,
-						   PCI_DMA_BIDIRECTIONAL);
-		if (pci_dma_mapping_error(dev->pdev, ttm_dma->dma_address[i])) {
+		ttm_dma->dma_address[i] = nv_device_map_page(device,
+							     ttm->pages[i]);
+		if (!ttm_dma->dma_address[i]) {
 			while (--i) {
-				pci_unmap_page(dev->pdev, ttm_dma->dma_address[i],
-					       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
+				nv_device_unmap_page(device,
+						     ttm_dma->dma_address[i]);
 				ttm_dma->dma_address[i] = 0;
 			}
 			ttm_pool_unpopulate(ttm);
@@ -1420,6 +1390,7 @@ nouveau_ttm_tt_unpopulate(struct ttm_tt *ttm)
 {
 	struct ttm_dma_tt *ttm_dma = (void *)ttm;
 	struct nouveau_drm *drm;
+	struct nouveau_device *device;
 	struct drm_device *dev;
 	unsigned i;
 	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
@@ -1428,6 +1399,7 @@ nouveau_ttm_tt_unpopulate(struct ttm_tt *ttm)
 		return;
 
 	drm = nouveau_bdev(ttm->bdev);
+	device = nv_device(drm->device);
 	dev = drm->dev;
 
 #if __OS_HAS_AGP
@@ -1446,8 +1418,7 @@ nouveau_ttm_tt_unpopulate(struct ttm_tt *ttm)
 
 	for (i = 0; i < ttm->num_pages; i++) {
 		if (ttm_dma->dma_address[i]) {
-			pci_unmap_page(dev->pdev, ttm_dma->dma_address[i],
-				       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
+			nv_device_unmap_page(device, ttm_dma->dma_address[i]);
 		}
 	}
 
@@ -1535,7 +1506,6 @@ nouveau_bo_vma_add(struct nouveau_bo *nvbo, struct nouveau_vm *vm,
 		   struct nouveau_vma *vma)
 {
 	const u32 size = nvbo->bo.mem.num_pages << PAGE_SHIFT;
-	struct nouveau_mem *node = nvbo->bo.mem.mm_node;
 	int ret;
 
 	ret = nouveau_vm_get(vm, size, nvbo->page_shift,
@@ -1543,15 +1513,10 @@ nouveau_bo_vma_add(struct nouveau_bo *nvbo, struct nouveau_vm *vm,
 	if (ret)
 		return ret;
 
-	if (nvbo->bo.mem.mem_type == TTM_PL_VRAM)
+	if ( nvbo->bo.mem.mem_type != TTM_PL_SYSTEM &&
+	    (nvbo->bo.mem.mem_type == TTM_PL_VRAM ||
+	     nvbo->page_shift != vma->vm->vmm->lpg_shift))
 		nouveau_vm_map(vma, nvbo->bo.mem.mm_node);
-	else if (nvbo->bo.mem.mem_type == TTM_PL_TT &&
-		 nvbo->page_shift == vma->vm->vmm->spg_shift) {
-		if (node->sg)
-			nouveau_vm_map_sg_table(vma, 0, size, node);
-		else
-			nouveau_vm_map_sg(vma, 0, size, node);
-	}
 
 	list_add_tail(&vma->head, &nvbo->vma_list);
 	vma->refcount = 1;
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c
index cc5152be2cf..ccb6b452d6d 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.c
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.c
@@ -154,7 +154,7 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nouveau_cli *cli,
 			 * nfi why this exists, it came from the -nv ddx.
 			 */
 			args.flags = NV_DMA_TARGET_PCI | NV_DMA_ACCESS_RDWR;
-			args.start = pci_resource_start(device->pdev, 1);
+			args.start = nv_device_resource_start(device, 1);
 			args.limit = args.start + limit;
 		} else {
 			args.flags = NV_DMA_TARGET_VRAM | NV_DMA_ACCESS_RDWR;
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c
index 1674882d60d..1fa222e8f00 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.c
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.c
@@ -44,6 +44,7 @@
 
 #include <subdev/i2c.h>
 #include <subdev/gpio.h>
+#include <engine/disp.h>
 
 MODULE_PARM_DESC(tv_disable, "Disable TV-out detection");
 static int nouveau_tv_disable = 0;
@@ -75,7 +76,8 @@ find_encoder(struct drm_connector *connector, int type)
 			continue;
 		nv_encoder = nouveau_encoder(obj_to_encoder(obj));
 
-		if (type == DCB_OUTPUT_ANY || nv_encoder->dcb->type == type)
+		if (type == DCB_OUTPUT_ANY ||
+		    (nv_encoder->dcb && nv_encoder->dcb->type == type))
 			return nv_encoder;
 	}
 
@@ -100,22 +102,24 @@ static void
 nouveau_connector_destroy(struct drm_connector *connector)
 {
 	struct nouveau_connector *nv_connector = nouveau_connector(connector);
-	nouveau_event_ref(NULL, &nv_connector->hpd_func);
+	nouveau_event_ref(NULL, &nv_connector->hpd);
 	kfree(nv_connector->edid);
 	drm_sysfs_connector_remove(connector);
 	drm_connector_cleanup(connector);
+	if (nv_connector->aux.transfer)
+		drm_dp_aux_unregister(&nv_connector->aux);
 	kfree(connector);
 }
 
-static struct nouveau_i2c_port *
-nouveau_connector_ddc_detect(struct drm_connector *connector,
-			     struct nouveau_encoder **pnv_encoder)
+static struct nouveau_encoder *
+nouveau_connector_ddc_detect(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
 	struct nouveau_connector *nv_connector = nouveau_connector(connector);
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nouveau_gpio *gpio = nouveau_gpio(drm->device);
-	struct nouveau_i2c_port *port = NULL;
+	struct nouveau_encoder *nv_encoder;
+	struct drm_mode_object *obj;
 	int i, panel = -ENODEV;
 
 	/* eDP panels need powering on by us (if the VBIOS doesn't default it
@@ -130,13 +134,9 @@ nouveau_connector_ddc_detect(struct drm_connector *connector,
 		}
 	}
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		struct nouveau_encoder *nv_encoder;
-		struct drm_mode_object *obj;
-		int id;
-
-		id = connector->encoder_ids[i];
-		if (!id)
+	for (i = 0; nv_encoder = NULL, i < DRM_CONNECTOR_MAX_ENCODER; i++) {
+		int id = connector->encoder_ids[i];
+		if (id == 0)
 			break;
 
 		obj = drm_mode_object_find(dev, id, DRM_MODE_OBJECT_ENCODER);
@@ -144,22 +144,24 @@ nouveau_connector_ddc_detect(struct drm_connector *connector,
 			continue;
 		nv_encoder = nouveau_encoder(obj_to_encoder(obj));
 
-		port = nv_encoder->i2c;
-		if (port && nv_probe_i2c(port, 0x50)) {
-			*pnv_encoder = nv_encoder;
-			break;
+		if (nv_encoder->dcb->type == DCB_OUTPUT_DP) {
+			int ret = nouveau_dp_detect(nv_encoder);
+			if (ret == 0)
+				break;
+		} else
+		if (nv_encoder->i2c) {
+			if (nv_probe_i2c(nv_encoder->i2c, 0x50))
+				break;
 		}
-
-		port = NULL;
 	}
 
 	/* eDP panel not detected, restore panel power GPIO to previous
 	 * state to avoid confusing the SOR for other output types.
 	 */
-	if (!port && panel == 0)
+	if (!nv_encoder && panel == 0)
 		gpio->set(gpio, 0, DCB_GPIO_PANEL_POWER, 0xff, panel);
 
-	return port;
+	return nv_encoder;
 }
 
 static struct nouveau_encoder *
@@ -255,28 +257,20 @@ nouveau_connector_detect(struct drm_connector *connector, bool force)
 	}
 
 	ret = pm_runtime_get_sync(connector->dev->dev);
-	if (ret < 0)
+	if (ret < 0 && ret != -EACCES)
 		return conn_status;
 
-	i2c = nouveau_connector_ddc_detect(connector, &nv_encoder);
-	if (i2c) {
+	nv_encoder = nouveau_connector_ddc_detect(connector);
+	if (nv_encoder && (i2c = nv_encoder->i2c) != NULL) {
 		nv_connector->edid = drm_get_edid(connector, &i2c->adapter);
 		drm_mode_connector_update_edid_property(connector,
 							nv_connector->edid);
 		if (!nv_connector->edid) {
 			NV_ERROR(drm, "DDC responded, but no EDID for %s\n",
-				 drm_get_connector_name(connector));
+				 connector->name);
 			goto detect_analog;
 		}
 
-		if (nv_encoder->dcb->type == DCB_OUTPUT_DP &&
-		    !nouveau_dp_detect(to_drm_encoder(nv_encoder))) {
-			NV_ERROR(drm, "Detected %s, but failed init\n",
-				 drm_get_connector_name(connector));
-			conn_status = connector_status_disconnected;
-			goto out;
-		}
-
 		/* Override encoder type for DVI-I based on whether EDID
 		 * says the display is digital or analog, both use the
 		 * same i2c channel so the value returned from ddc_detect
@@ -437,7 +431,7 @@ nouveau_connector_force(struct drm_connector *connector)
 	nv_encoder = find_encoder(connector, type);
 	if (!nv_encoder) {
 		NV_ERROR(drm, "can't find encoder to force %s on!\n",
-			 drm_get_connector_name(connector));
+			 connector->name);
 		connector->status = connector_status_disconnected;
 		return;
 	}
@@ -912,33 +906,103 @@ nouveau_connector_funcs_lvds = {
 };
 
 static void
+nouveau_connector_dp_dpms(struct drm_connector *connector, int mode)
+{
+	struct nouveau_encoder *nv_encoder = NULL;
+
+	if (connector->encoder)
+		nv_encoder = nouveau_encoder(connector->encoder);
+	if (nv_encoder && nv_encoder->dcb &&
+	    nv_encoder->dcb->type == DCB_OUTPUT_DP) {
+		if (mode == DRM_MODE_DPMS_ON) {
+			u8 data = DP_SET_POWER_D0;
+			nv_wraux(nv_encoder->i2c, DP_SET_POWER, &data, 1);
+			usleep_range(1000, 2000);
+		} else {
+			u8 data = DP_SET_POWER_D3;
+			nv_wraux(nv_encoder->i2c, DP_SET_POWER, &data, 1);
+		}
+	}
+
+	drm_helper_connector_dpms(connector, mode);
+}
+
+static const struct drm_connector_funcs
+nouveau_connector_funcs_dp = {
+	.dpms = nouveau_connector_dp_dpms,
+	.save = NULL,
+	.restore = NULL,
+	.detect = nouveau_connector_detect,
+	.destroy = nouveau_connector_destroy,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.set_property = nouveau_connector_set_property,
+	.force = nouveau_connector_force
+};
+
+static void
 nouveau_connector_hotplug_work(struct work_struct *work)
 {
 	struct nouveau_connector *nv_connector =
-		container_of(work, struct nouveau_connector, hpd_work);
+		container_of(work, typeof(*nv_connector), work);
 	struct drm_connector *connector = &nv_connector->base;
-	struct drm_device *dev = connector->dev;
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nouveau_gpio *gpio = nouveau_gpio(drm->device);
-	bool plugged = gpio->get(gpio, 0, nv_connector->hpd.func, 0xff);
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	const char *name = connector->name;
 
-	NV_DEBUG(drm, "%splugged %s\n", plugged ? "" : "un",
-		 drm_get_connector_name(connector));
+	if (nv_connector->status & NVKM_HPD_IRQ) {
+	} else {
+		bool plugged = (nv_connector->status != NVKM_HPD_UNPLUG);
 
-	if (plugged)
-		drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON);
-	else
-		drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF);
+		NV_DEBUG(drm, "%splugged %s\n", plugged ? "" : "un", name);
 
-	drm_helper_hpd_irq_event(dev);
+		if (plugged)
+			drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON);
+		else
+			drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF);
+		drm_helper_hpd_irq_event(connector->dev);
+	}
+
+	nouveau_event_get(nv_connector->hpd);
 }
 
 static int
-nouveau_connector_hotplug(void *data, int index)
+nouveau_connector_hotplug(void *data, u32 type, int index)
 {
 	struct nouveau_connector *nv_connector = data;
-	schedule_work(&nv_connector->hpd_work);
-	return NVKM_EVENT_KEEP;
+	nv_connector->status = type;
+	schedule_work(&nv_connector->work);
+	return NVKM_EVENT_DROP;
+}
+
+static ssize_t
+nouveau_connector_aux_xfer(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg)
+{
+	struct nouveau_connector *nv_connector =
+		container_of(aux, typeof(*nv_connector), aux);
+	struct nouveau_encoder *nv_encoder;
+	struct nouveau_i2c_port *port;
+	int ret;
+
+	nv_encoder = find_encoder(&nv_connector->base, DCB_OUTPUT_DP);
+	if (!nv_encoder || !(port = nv_encoder->i2c))
+		return -ENODEV;
+	if (WARN_ON(msg->size > 16))
+		return -E2BIG;
+	if (msg->size == 0)
+		return msg->size;
+
+	ret = nouveau_i2c(port)->acquire(port, 0);
+	if (ret)
+		return ret;
+
+	ret = port->func->aux(port, false, msg->request, msg->address,
+			      msg->buffer, msg->size);
+	nouveau_i2c(port)->release(port);
+	if (ret >= 0) {
+		msg->reply = ret;
+		return msg->size;
+	}
+
+	return ret;
 }
 
 static int
@@ -960,7 +1024,8 @@ drm_conntype_from_dcb(enum dcb_connector_type dcb)
 	case DCB_CONNECTOR_DP       : return DRM_MODE_CONNECTOR_DisplayPort;
 	case DCB_CONNECTOR_eDP      : return DRM_MODE_CONNECTOR_eDP;
 	case DCB_CONNECTOR_HDMI_0   :
-	case DCB_CONNECTOR_HDMI_1   : return DRM_MODE_CONNECTOR_HDMIA;
+	case DCB_CONNECTOR_HDMI_1   :
+	case DCB_CONNECTOR_HDMI_C   : return DRM_MODE_CONNECTOR_HDMIA;
 	default:
 		break;
 	}
@@ -973,9 +1038,9 @@ nouveau_connector_create(struct drm_device *dev, int index)
 {
 	const struct drm_connector_funcs *funcs = &nouveau_connector_funcs;
 	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nouveau_gpio *gpio = nouveau_gpio(drm->device);
 	struct nouveau_display *disp = nouveau_display(dev);
 	struct nouveau_connector *nv_connector = NULL;
+	struct nouveau_disp *pdisp = nouveau_disp(drm->device);
 	struct drm_connector *connector;
 	int type, ret = 0;
 	bool dummy;
@@ -991,33 +1056,15 @@ nouveau_connector_create(struct drm_device *dev, int index)
 		return ERR_PTR(-ENOMEM);
 
 	connector = &nv_connector->base;
-	INIT_WORK(&nv_connector->hpd_work, nouveau_connector_hotplug_work);
 	nv_connector->index = index;
 
 	/* attempt to parse vbios connector type and hotplug gpio */
 	nv_connector->dcb = olddcb_conn(dev, index);
 	if (nv_connector->dcb) {
-		static const u8 hpd[16] = {
-			0xff, 0x07, 0x08, 0xff, 0xff, 0x51, 0x52, 0xff,
-			0xff, 0xff, 0xff, 0xff, 0xff, 0x5e, 0x5f, 0x60,
-		};
-
 		u32 entry = ROM16(nv_connector->dcb[0]);
 		if (olddcb_conntab(dev)[3] >= 4)
 			entry |= (u32)ROM16(nv_connector->dcb[2]) << 16;
 
-		ret = gpio->find(gpio, 0, hpd[ffs((entry & 0x07033000) >> 12)],
-				 DCB_GPIO_UNUSED, &nv_connector->hpd);
-		if (ret)
-			nv_connector->hpd.func = DCB_GPIO_UNUSED;
-
-		if (nv_connector->hpd.func != DCB_GPIO_UNUSED) {
-			nouveau_event_new(gpio->events, nv_connector->hpd.line,
-					  nouveau_connector_hotplug,
-					  nv_connector,
-					 &nv_connector->hpd_func);
-		}
-
 		nv_connector->type = nv_connector->dcb[0];
 		if (drm_conntype_from_dcb(nv_connector->type) ==
 					  DRM_MODE_CONNECTOR_Unknown) {
@@ -1039,7 +1086,6 @@ nouveau_connector_create(struct drm_device *dev, int index)
 		}
 	} else {
 		nv_connector->type = DCB_CONNECTOR_NONE;
-		nv_connector->hpd.func = DCB_GPIO_UNUSED;
 	}
 
 	/* no vbios data, or an unknown dcb connector type - attempt to
@@ -1079,8 +1125,8 @@ nouveau_connector_create(struct drm_device *dev, int index)
 		}
 	}
 
-	type = drm_conntype_from_dcb(nv_connector->type);
-	if (type == DRM_MODE_CONNECTOR_LVDS) {
+	switch ((type = drm_conntype_from_dcb(nv_connector->type))) {
+	case DRM_MODE_CONNECTOR_LVDS:
 		ret = nouveau_bios_parse_lvds_table(dev, 0, &dummy, &dummy);
 		if (ret) {
 			NV_ERROR(drm, "Error parsing LVDS table, disabling\n");
@@ -1089,8 +1135,23 @@ nouveau_connector_create(struct drm_device *dev, int index)
 		}
 
 		funcs = &nouveau_connector_funcs_lvds;
-	} else {
+		break;
+	case DRM_MODE_CONNECTOR_DisplayPort:
+	case DRM_MODE_CONNECTOR_eDP:
+		nv_connector->aux.dev = dev->dev;
+		nv_connector->aux.transfer = nouveau_connector_aux_xfer;
+		ret = drm_dp_aux_register(&nv_connector->aux);
+		if (ret) {
+			NV_ERROR(drm, "failed to register aux channel\n");
+			kfree(nv_connector);
+			return ERR_PTR(ret);
+		}
+
+		funcs = &nouveau_connector_funcs_dp;
+		break;
+	default:
 		funcs = &nouveau_connector_funcs;
+		break;
 	}
 
 	/* defaults, will get overridden in detect() */
@@ -1165,10 +1226,16 @@ nouveau_connector_create(struct drm_device *dev, int index)
 		break;
 	}
 
-	connector->polled = DRM_CONNECTOR_POLL_CONNECT;
-	if (nv_connector->hpd.func != DCB_GPIO_UNUSED)
+	ret = nouveau_event_new(pdisp->hpd, NVKM_HPD, index,
+				nouveau_connector_hotplug,
+				nv_connector, &nv_connector->hpd);
+	if (ret)
+		connector->polled = DRM_CONNECTOR_POLL_CONNECT;
+	else
 		connector->polled = DRM_CONNECTOR_POLL_HPD;
 
+	INIT_WORK(&nv_connector->work, nouveau_connector_hotplug_work);
+
 	drm_sysfs_connector_add(connector);
 	return connector;
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.h b/drivers/gpu/drm/nouveau/nouveau_connector.h
index 264a778f473..8861b6c579a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.h
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.h
@@ -28,12 +28,12 @@
 #define __NOUVEAU_CONNECTOR_H__
 
 #include <drm/drm_edid.h>
+#include <drm/drm_dp_helper.h>
 #include "nouveau_crtc.h"
 
 #include <core/event.h>
 
 #include <subdev/bios.h>
-#include <subdev/bios/gpio.h>
 
 struct nouveau_i2c_port;
 
@@ -67,9 +67,11 @@ struct nouveau_connector {
 	u8 index;
 	u8 *dcb;
 
-	struct dcb_gpio_func hpd;
-	struct work_struct hpd_work;
-	struct nouveau_eventh *hpd_func;
+	struct nouveau_eventh *hpd;
+	u32 status;
+	struct work_struct work;
+
+	struct drm_dp_aux aux;
 
 	int dithering_mode;
 	int dithering_depth;
diff --git a/drivers/gpu/drm/nouveau/nouveau_crtc.h b/drivers/gpu/drm/nouveau/nouveau_crtc.h
index d1e5890784d..a0534489d23 100644
--- a/drivers/gpu/drm/nouveau/nouveau_crtc.h
+++ b/drivers/gpu/drm/nouveau/nouveau_crtc.h
@@ -74,7 +74,7 @@ struct nouveau_crtc {
 
 static inline struct nouveau_crtc *nouveau_crtc(struct drm_crtc *crtc)
 {
-	return container_of(crtc, struct nouveau_crtc, base);
+	return crtc ? container_of(crtc, struct nouveau_crtc, base) : NULL;
 }
 
 static inline struct drm_crtc *to_drm_crtc(struct nouveau_crtc *crtc)
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 25ea82f8def..47ad74255bf 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -42,7 +42,7 @@
 #include <core/class.h>
 
 static int
-nouveau_display_vblank_handler(void *data, int head)
+nouveau_display_vblank_handler(void *data, u32 type, int head)
 {
 	struct nouveau_drm *drm = data;
 	drm_handle_vblank(drm->dev, head);
@@ -68,20 +68,100 @@ nouveau_display_vblank_disable(struct drm_device *dev, int head)
 		nouveau_event_put(disp->vblank[head]);
 }
 
+static inline int
+calc(int blanks, int blanke, int total, int line)
+{
+	if (blanke >= blanks) {
+		if (line >= blanks)
+			line -= total;
+	} else {
+		if (line >= blanks)
+			line -= total;
+		line -= blanke + 1;
+	}
+	return line;
+}
+
+int
+nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos,
+				ktime_t *stime, ktime_t *etime)
+{
+	const u32 mthd = NV04_DISP_SCANOUTPOS + nouveau_crtc(crtc)->index;
+	struct nouveau_display *disp = nouveau_display(crtc->dev);
+	struct nv04_display_scanoutpos args;
+	int ret, retry = 1;
+
+	do {
+		ret = nv_exec(disp->core, mthd, &args, sizeof(args));
+		if (ret != 0)
+			return 0;
+
+		if (args.vline) {
+			ret |= DRM_SCANOUTPOS_ACCURATE;
+			ret |= DRM_SCANOUTPOS_VALID;
+			break;
+		}
+
+		if (retry) ndelay(crtc->linedur_ns);
+	} while (retry--);
+
+	*hpos = args.hline;
+	*vpos = calc(args.vblanks, args.vblanke, args.vtotal, args.vline);
+	if (stime) *stime = ns_to_ktime(args.time[0]);
+	if (etime) *etime = ns_to_ktime(args.time[1]);
+
+	if (*vpos < 0)
+		ret |= DRM_SCANOUTPOS_INVBL;
+	return ret;
+}
+
+int
+nouveau_display_scanoutpos(struct drm_device *dev, int head, unsigned int flags,
+			   int *vpos, int *hpos, ktime_t *stime, ktime_t *etime)
+{
+	struct drm_crtc *crtc;
+
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+		if (nouveau_crtc(crtc)->index == head) {
+			return nouveau_display_scanoutpos_head(crtc, vpos, hpos,
+							       stime, etime);
+		}
+	}
+
+	return 0;
+}
+
+int
+nouveau_display_vblstamp(struct drm_device *dev, int head, int *max_error,
+			 struct timeval *time, unsigned flags)
+{
+	struct drm_crtc *crtc;
+
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+		if (nouveau_crtc(crtc)->index == head) {
+			return drm_calc_vbltimestamp_from_scanoutpos(dev,
+					head, max_error, time, flags, crtc,
+					&crtc->hwmode);
+		}
+	}
+
+	return -EINVAL;
+}
+
 static void
 nouveau_display_vblank_fini(struct drm_device *dev)
 {
 	struct nouveau_display *disp = nouveau_display(dev);
 	int i;
 
+	drm_vblank_cleanup(dev);
+
 	if (disp->vblank) {
 		for (i = 0; i < dev->mode_config.num_crtc; i++)
 			nouveau_event_ref(NULL, &disp->vblank[i]);
 		kfree(disp->vblank);
 		disp->vblank = NULL;
 	}
-
-	drm_vblank_cleanup(dev);
 }
 
 static int
@@ -98,7 +178,7 @@ nouveau_display_vblank_init(struct drm_device *dev)
 		return -ENOMEM;
 
 	for (i = 0; i < dev->mode_config.num_crtc; i++) {
-		ret = nouveau_event_new(pdisp->vblank, i,
+		ret = nouveau_event_new(pdisp->vblank, 1, i,
 					nouveau_display_vblank_handler,
 					drm, &disp->vblank[i]);
 		if (ret) {
@@ -313,7 +393,7 @@ nouveau_display_init(struct drm_device *dev)
 	/* enable hotplug interrupts */
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
 		struct nouveau_connector *conn = nouveau_connector(connector);
-		if (conn->hpd_func) nouveau_event_get(conn->hpd_func);
+		if (conn->hpd) nouveau_event_get(conn->hpd);
 	}
 
 	return ret;
@@ -328,7 +408,7 @@ nouveau_display_fini(struct drm_device *dev)
 	/* disable hotplug interrupts */
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
 		struct nouveau_connector *conn = nouveau_connector(connector);
-		if (conn->hpd_func) nouveau_event_put(conn->hpd_func);
+		if (conn->hpd) nouveau_event_put(conn->hpd);
 	}
 
 	drm_kms_helper_poll_disable(dev);
@@ -339,6 +419,7 @@ int
 nouveau_display_create(struct drm_device *dev)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nouveau_device *device = nouveau_dev(dev);
 	struct nouveau_display *disp;
 	int ret, gen;
 
@@ -379,7 +460,7 @@ nouveau_display_create(struct drm_device *dev)
 	}
 
 	dev->mode_config.funcs = &nouveau_mode_config_funcs;
-	dev->mode_config.fb_base = pci_resource_start(dev->pdev, 1);
+	dev->mode_config.fb_base = nv_device_resource_start(device, 1);
 
 	dev->mode_config.min_width = 0;
 	dev->mode_config.min_height = 0;
@@ -407,10 +488,32 @@ nouveau_display_create(struct drm_device *dev)
 	drm_kms_helper_poll_disable(dev);
 
 	if (drm->vbios.dcb.entries) {
-		if (nv_device(drm->device)->card_type < NV_50)
-			ret = nv04_display_create(dev);
-		else
-			ret = nv50_display_create(dev);
+		static const u16 oclass[] = {
+			GM107_DISP_CLASS,
+			NVF0_DISP_CLASS,
+			NVE0_DISP_CLASS,
+			NVD0_DISP_CLASS,
+			NVA3_DISP_CLASS,
+			NV94_DISP_CLASS,
+			NVA0_DISP_CLASS,
+			NV84_DISP_CLASS,
+			NV50_DISP_CLASS,
+			NV04_DISP_CLASS,
+		};
+		int i;
+
+		for (i = 0, ret = -ENODEV; ret && i < ARRAY_SIZE(oclass); i++) {
+			ret = nouveau_object_new(nv_object(drm), NVDRM_DEVICE,
+						 NVDRM_DISPLAY, oclass[i],
+						 NULL, 0, &disp->core);
+		}
+
+		if (ret == 0) {
+			if (nv_mclass(disp->core) < NV50_DISP_CLASS)
+				ret = nv04_display_create(dev);
+			else
+				ret = nv50_display_create(dev);
+		}
 	} else {
 		ret = 0;
 	}
@@ -439,6 +542,7 @@ void
 nouveau_display_destroy(struct drm_device *dev)
 {
 	struct nouveau_display *disp = nouveau_display(dev);
+	struct nouveau_drm *drm = nouveau_drm(dev);
 
 	nouveau_backlight_exit(dev);
 	nouveau_display_vblank_fini(dev);
@@ -449,6 +553,8 @@ nouveau_display_destroy(struct drm_device *dev)
 	if (disp->dtor)
 		disp->dtor(dev);
 
+	nouveau_object_del(nv_object(drm), NVDRM_DEVICE, NVDRM_DISPLAY);
+
 	nouveau_drm(dev)->display = NULL;
 	kfree(disp);
 }
@@ -465,7 +571,7 @@ nouveau_display_suspend(struct drm_device *dev)
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct nouveau_framebuffer *nouveau_fb;
 
-		nouveau_fb = nouveau_framebuffer(crtc->fb);
+		nouveau_fb = nouveau_framebuffer(crtc->primary->fb);
 		if (!nouveau_fb || !nouveau_fb->nvbo)
 			continue;
 
@@ -492,7 +598,7 @@ nouveau_display_repin(struct drm_device *dev)
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct nouveau_framebuffer *nouveau_fb;
 
-		nouveau_fb = nouveau_framebuffer(crtc->fb);
+		nouveau_fb = nouveau_framebuffer(crtc->primary->fb);
 		if (!nouveau_fb || !nouveau_fb->nvbo)
 			continue;
 
@@ -589,7 +695,7 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	const int swap_interval = (flags & DRM_MODE_PAGE_FLIP_ASYNC) ? 0 : 1;
 	struct drm_device *dev = crtc->dev;
 	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nouveau_bo *old_bo = nouveau_framebuffer(crtc->fb)->nvbo;
+	struct nouveau_bo *old_bo = nouveau_framebuffer(crtc->primary->fb)->nvbo;
 	struct nouveau_bo *new_bo = nouveau_framebuffer(fb)->nvbo;
 	struct nouveau_page_flip_state *s;
 	struct nouveau_channel *chan = drm->channel;
@@ -603,6 +709,14 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	if (!s)
 		return -ENOMEM;
 
+	if (new_bo != old_bo) {
+		ret = nouveau_bo_pin(new_bo, TTM_PL_FLAG_VRAM);
+		if (ret)
+			goto fail_free;
+	}
+
+	mutex_lock(&chan->cli->mutex);
+
 	/* synchronise rendering channel with the kernel's channel */
 	spin_lock(&new_bo->bo.bdev->fence_lock);
 	fence = nouveau_fence_ref(new_bo->bo.sync_obj);
@@ -610,15 +724,8 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	ret = nouveau_fence_sync(fence, chan);
 	nouveau_fence_unref(&fence);
 	if (ret)
-		goto fail_free;
-
-	if (new_bo != old_bo) {
-		ret = nouveau_bo_pin(new_bo, TTM_PL_FLAG_VRAM);
-		if (ret)
-			goto fail_free;
-	}
+		goto fail_unpin;
 
-	mutex_lock(&chan->cli->mutex);
 	ret = ttm_bo_reserve(&old_bo->bo, true, false, false, NULL);
 	if (ret)
 		goto fail_unpin;
@@ -629,6 +736,9 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 		  fb->bits_per_pixel, fb->pitches[0], crtc->x, crtc->y,
 		  new_bo->bo.offset };
 
+	/* Keep vblanks on during flip, for the target crtc of this flip */
+	drm_vblank_get(dev, nouveau_crtc(crtc)->index);
+
 	/* Emit a page flip */
 	if (nv_device(drm->device)->card_type >= NV_50) {
 		ret = nv50_display_flip_next(crtc, fb, chan, swap_interval);
@@ -657,12 +767,12 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	}
 
 	ret = nouveau_page_flip_emit(chan, old_bo, new_bo, s, &fence);
-	mutex_unlock(&chan->cli->mutex);
 	if (ret)
 		goto fail_unreserve;
+	mutex_unlock(&chan->cli->mutex);
 
 	/* Update the crtc struct and cleanup */
-	crtc->fb = fb;
+	crtc->primary->fb = fb;
 
 	nouveau_bo_fence(old_bo, fence);
 	ttm_bo_unreserve(&old_bo->bo);
@@ -672,6 +782,7 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	return 0;
 
 fail_unreserve:
+	drm_vblank_put(dev, nouveau_crtc(crtc)->index);
 	ttm_bo_unreserve(&old_bo->bo);
 fail_unpin:
 	mutex_unlock(&chan->cli->mutex);
@@ -691,6 +802,7 @@ nouveau_finish_page_flip(struct nouveau_channel *chan,
 	struct drm_device *dev = drm->dev;
 	struct nouveau_page_flip_state *s;
 	unsigned long flags;
+	int crtcid = -1;
 
 	spin_lock_irqsave(&dev->event_lock, flags);
 
@@ -701,8 +813,16 @@ nouveau_finish_page_flip(struct nouveau_channel *chan,
 	}
 
 	s = list_first_entry(&fctx->flip, struct nouveau_page_flip_state, head);
-	if (s->event)
-		drm_send_vblank_event(dev, s->crtc, s->event);
+	if (s->event) {
+		/* Vblank timestamps/counts are only correct on >= NV-50 */
+		if (nv_device(drm->device)->card_type >= NV_50)
+			crtcid = s->crtc;
+
+		drm_send_vblank_event(dev, crtcid, s->event);
+	}
+
+	/* Give up ownership of vblank for page-flipped crtc */
+	drm_vblank_put(dev, s->crtc);
 
 	list_del(&s->head);
 	if (ps)
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.h b/drivers/gpu/drm/nouveau/nouveau_display.h
index 8bc8bab90e8..a71cf77e55b 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.h
+++ b/drivers/gpu/drm/nouveau/nouveau_display.h
@@ -36,6 +36,7 @@ struct nouveau_display {
 	int  (*init)(struct drm_device *);
 	void (*fini)(struct drm_device *);
 
+	struct nouveau_object *core;
 	struct nouveau_eventh **vblank;
 
 	struct drm_property *dithering_mode;
@@ -63,6 +64,10 @@ void nouveau_display_repin(struct drm_device *dev);
 void nouveau_display_resume(struct drm_device *dev);
 int  nouveau_display_vblank_enable(struct drm_device *, int);
 void nouveau_display_vblank_disable(struct drm_device *, int);
+int  nouveau_display_scanoutpos(struct drm_device *, int, unsigned int,
+				int *, int *, ktime_t *, ktime_t *);
+int  nouveau_display_vblstamp(struct drm_device *, int, int *,
+			      struct timeval *, unsigned);
 
 int  nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 			    struct drm_pending_vblank_event *event,
diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.c b/drivers/gpu/drm/nouveau/nouveau_dma.c
index 40f91e1e584..c177272152e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.c
@@ -100,7 +100,7 @@ nv50_dma_push(struct nouveau_channel *chan, struct nouveau_bo *bo,
 
 	chan->dma.ib_put = (chan->dma.ib_put + 1) & chan->dma.ib_max;
 
-	DRM_MEMORYBARRIER();
+	mb();
 	/* Flush writes. */
 	nouveau_bo_rd32(pb, 0);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.h b/drivers/gpu/drm/nouveau/nouveau_dma.h
index 984004d66a6..dc0e0c5cadb 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.h
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.h
@@ -155,7 +155,7 @@ BEGIN_IMC0(struct nouveau_channel *chan, int subc, int mthd, u16 data)
 }
 
 #define WRITE_PUT(val) do {                                                    \
-	DRM_MEMORYBARRIER();                                                   \
+	mb();                                                   \
 	nouveau_bo_rd32(chan->push.buffer, 0);                                 \
 	nv_wo32(chan->object, chan->user_put, ((val) << 2) + chan->push.vma.offset);  \
 } while (0)
diff --git a/drivers/gpu/drm/nouveau/nouveau_dp.c b/drivers/gpu/drm/nouveau/nouveau_dp.c
index 36fd2250056..5675ffc175a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dp.c
+++ b/drivers/gpu/drm/nouveau/nouveau_dp.c
@@ -55,11 +55,10 @@ nouveau_dp_probe_oui(struct drm_device *dev, struct nouveau_i2c_port *auxch,
 
 }
 
-bool
-nouveau_dp_detect(struct drm_encoder *encoder)
+int
+nouveau_dp_detect(struct nouveau_encoder *nv_encoder)
 {
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct drm_device *dev = encoder->dev;
+	struct drm_device *dev = nv_encoder->base.base.dev;
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nouveau_i2c_port *auxch;
 	u8 *dpcd = nv_encoder->dp.dpcd;
@@ -67,11 +66,11 @@ nouveau_dp_detect(struct drm_encoder *encoder)
 
 	auxch = nv_encoder->i2c;
 	if (!auxch)
-		return false;
+		return -ENODEV;
 
 	ret = nv_rdaux(auxch, DP_DPCD_REV, dpcd, 8);
 	if (ret)
-		return false;
+		return ret;
 
 	nv_encoder->dp.link_bw = 27000 * dpcd[1];
 	nv_encoder->dp.link_nr = dpcd[2] & DP_MAX_LANE_COUNT_MASK;
@@ -91,6 +90,5 @@ nouveau_dp_detect(struct drm_encoder *encoder)
 		     nv_encoder->dp.link_nr, nv_encoder->dp.link_bw);
 
 	nouveau_dp_probe_oui(dev, auxch, dpcd);
-
-	return true;
+	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 98a22e6e27a..5425ffe3931 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -33,6 +33,7 @@
 #include <core/client.h>
 #include <core/gpuobj.h>
 #include <core/class.h>
+#include <core/option.h>
 
 #include <engine/device.h>
 #include <engine/disp.h>
@@ -81,7 +82,7 @@ module_param_named(runpm, nouveau_runtime_pm, int, 0400);
 static struct drm_driver driver;
 
 static u64
-nouveau_name(struct pci_dev *pdev)
+nouveau_pci_name(struct pci_dev *pdev)
 {
 	u64 name = (u64)pci_domain_nr(pdev->bus) << 32;
 	name |= pdev->bus->number << 16;
@@ -89,15 +90,30 @@ nouveau_name(struct pci_dev *pdev)
 	return name | PCI_FUNC(pdev->devfn);
 }
 
+static u64
+nouveau_platform_name(struct platform_device *platformdev)
+{
+	return platformdev->id;
+}
+
+static u64
+nouveau_name(struct drm_device *dev)
+{
+	if (dev->pdev)
+		return nouveau_pci_name(dev->pdev);
+	else
+		return nouveau_platform_name(dev->platformdev);
+}
+
 static int
-nouveau_cli_create(struct pci_dev *pdev, const char *name,
+nouveau_cli_create(u64 name, const char *sname,
 		   int size, void **pcli)
 {
 	struct nouveau_cli *cli;
 	int ret;
 
 	*pcli = NULL;
-	ret = nouveau_client_create_(name, nouveau_name(pdev), nouveau_config,
+	ret = nouveau_client_create_(sname, name, nouveau_config,
 				     nouveau_debug, size, pcli);
 	cli = *pcli;
 	if (ret) {
@@ -281,7 +297,8 @@ static int nouveau_drm_probe(struct pci_dev *pdev,
 	remove_conflicting_framebuffers(aper, "nouveaufb", boot);
 	kfree(aper);
 
-	ret = nouveau_device_create(pdev, nouveau_name(pdev), pci_name(pdev),
+	ret = nouveau_device_create(pdev, NOUVEAU_BUS_PCI,
+				    nouveau_pci_name(pdev), pci_name(pdev),
 				    nouveau_config, nouveau_debug, &device);
 	if (ret)
 		return ret;
@@ -300,22 +317,27 @@ static int nouveau_drm_probe(struct pci_dev *pdev,
 #define PCI_CLASS_MULTIMEDIA_HD_AUDIO 0x0403
 
 static void
-nouveau_get_hdmi_dev(struct drm_device *dev)
+nouveau_get_hdmi_dev(struct nouveau_drm *drm)
 {
-	struct nouveau_drm *drm = dev->dev_private;
-	struct pci_dev *pdev = dev->pdev;
+	struct pci_dev *pdev = drm->dev->pdev;
+
+	if (!pdev) {
+		DRM_INFO("not a PCI device; no HDMI\n");
+		drm->hdmi_device = NULL;
+		return;
+	}
 
 	/* subfunction one is a hdmi audio device? */
 	drm->hdmi_device = pci_get_bus_and_slot((unsigned int)pdev->bus->number,
 						PCI_DEVFN(PCI_SLOT(pdev->devfn), 1));
 
 	if (!drm->hdmi_device) {
-		DRM_INFO("hdmi device  not found %d %d %d\n", pdev->bus->number, PCI_SLOT(pdev->devfn), 1);
+		NV_DEBUG(drm, "hdmi device not found %d %d %d\n", pdev->bus->number, PCI_SLOT(pdev->devfn), 1);
 		return;
 	}
 
 	if ((drm->hdmi_device->class >> 8) != PCI_CLASS_MULTIMEDIA_HD_AUDIO) {
-		DRM_INFO("possible hdmi device  not audio %d\n", drm->hdmi_device->class);
+		NV_DEBUG(drm, "possible hdmi device not audio %d\n", drm->hdmi_device->class);
 		pci_dev_put(drm->hdmi_device);
 		drm->hdmi_device = NULL;
 		return;
@@ -330,22 +352,24 @@ nouveau_drm_load(struct drm_device *dev, unsigned long flags)
 	struct nouveau_drm *drm;
 	int ret;
 
-	ret = nouveau_cli_create(pdev, "DRM", sizeof(*drm), (void**)&drm);
+	ret = nouveau_cli_create(nouveau_name(dev), "DRM", sizeof(*drm),
+				 (void **)&drm);
 	if (ret)
 		return ret;
 
 	dev->dev_private = drm;
 	drm->dev = dev;
+	nouveau_client(drm)->debug = nouveau_dbgopt(nouveau_debug, "DRM");
 
 	INIT_LIST_HEAD(&drm->clients);
 	spin_lock_init(&drm->tile.lock);
 
-	nouveau_get_hdmi_dev(dev);
+	nouveau_get_hdmi_dev(drm);
 
 	/* make sure AGP controller is in a consistent state before we
 	 * (possibly) execute vbios init tables (see nouveau_agp.h)
 	 */
-	if (drm_pci_device_is_agp(dev) && dev->agp) {
+	if (pdev && drm_pci_device_is_agp(dev) && dev->agp) {
 		/* dummy device object, doesn't init anything, but allows
 		 * agp code access to registers
 		 */
@@ -376,6 +400,8 @@ nouveau_drm_load(struct drm_device *dev, unsigned long flags)
 	if (ret)
 		goto fail_device;
 
+	dev->irq_enabled = true;
+
 	/* workaround an odd issue on nvc1 by disabling the device's
 	 * nosnoop capability.  hopefully won't cause issues until a
 	 * better fix is found - assuming there is one...
@@ -475,6 +501,7 @@ nouveau_drm_remove(struct pci_dev *pdev)
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nouveau_object *device;
 
+	dev->irq_enabled = false;
 	device = drm->client.base.device;
 	drm_put_dev(dev);
 
@@ -483,13 +510,13 @@ nouveau_drm_remove(struct pci_dev *pdev)
 }
 
 static int
-nouveau_do_suspend(struct drm_device *dev)
+nouveau_do_suspend(struct drm_device *dev, bool runtime)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nouveau_cli *cli;
 	int ret;
 
-	if (dev->mode_config.num_crtc) {
+	if (dev->mode_config.num_crtc && !runtime) {
 		NV_INFO(drm, "suspending display...\n");
 		ret = nouveau_display_suspend(dev);
 		if (ret)
@@ -503,19 +530,21 @@ nouveau_do_suspend(struct drm_device *dev)
 	if (drm->cechan) {
 		ret = nouveau_channel_idle(drm->cechan);
 		if (ret)
-			return ret;
+			goto fail_display;
 	}
 
 	if (drm->channel) {
 		ret = nouveau_channel_idle(drm->channel);
 		if (ret)
-			return ret;
+			goto fail_display;
 	}
 
 	NV_INFO(drm, "suspending client object trees...\n");
 	if (drm->fence && nouveau_fence(drm)->suspend) {
-		if (!nouveau_fence(drm)->suspend(drm))
-			return -ENOMEM;
+		if (!nouveau_fence(drm)->suspend(drm)) {
+			ret = -ENOMEM;
+			goto fail_display;
+		}
 	}
 
 	list_for_each_entry(cli, &drm->clients, head) {
@@ -537,6 +566,10 @@ fail_client:
 		nouveau_client_init(&cli->base);
 	}
 
+	if (drm->fence && nouveau_fence(drm)->resume)
+		nouveau_fence(drm)->resume(drm);
+
+fail_display:
 	if (dev->mode_config.num_crtc) {
 		NV_INFO(drm, "resuming display...\n");
 		nouveau_display_resume(dev);
@@ -557,7 +590,7 @@ int nouveau_pmops_suspend(struct device *dev)
 	if (drm_dev->mode_config.num_crtc)
 		nouveau_fbcon_set_suspend(drm_dev, 1);
 
-	ret = nouveau_do_suspend(drm_dev);
+	ret = nouveau_do_suspend(drm_dev, false);
 	if (ret)
 		return ret;
 
@@ -619,12 +652,12 @@ int nouveau_pmops_resume(struct device *dev)
 	ret = nouveau_do_resume(drm_dev);
 	if (ret)
 		return ret;
-	if (drm_dev->mode_config.num_crtc)
-		nouveau_fbcon_set_suspend(drm_dev, 0);
 
-	nouveau_fbcon_zfill_all(drm_dev);
-	if (drm_dev->mode_config.num_crtc)
+	if (drm_dev->mode_config.num_crtc) {
 		nouveau_display_resume(drm_dev);
+		nouveau_fbcon_set_suspend(drm_dev, 0);
+	}
+
 	return 0;
 }
 
@@ -637,7 +670,7 @@ static int nouveau_pmops_freeze(struct device *dev)
 	if (drm_dev->mode_config.num_crtc)
 		nouveau_fbcon_set_suspend(drm_dev, 1);
 
-	ret = nouveau_do_suspend(drm_dev);
+	ret = nouveau_do_suspend(drm_dev, false);
 	return ret;
 }
 
@@ -650,11 +683,12 @@ static int nouveau_pmops_thaw(struct device *dev)
 	ret = nouveau_do_resume(drm_dev);
 	if (ret)
 		return ret;
-	if (drm_dev->mode_config.num_crtc)
-		nouveau_fbcon_set_suspend(drm_dev, 0);
-	nouveau_fbcon_zfill_all(drm_dev);
-	if (drm_dev->mode_config.num_crtc)
+
+	if (drm_dev->mode_config.num_crtc) {
 		nouveau_display_resume(drm_dev);
+		nouveau_fbcon_set_suspend(drm_dev, 0);
+	}
+
 	return 0;
 }
 
@@ -662,7 +696,6 @@ static int nouveau_pmops_thaw(struct device *dev)
 static int
 nouveau_drm_open(struct drm_device *dev, struct drm_file *fpriv)
 {
-	struct pci_dev *pdev = dev->pdev;
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nouveau_cli *cli;
 	char name[32], tmpname[TASK_COMM_LEN];
@@ -670,13 +703,15 @@ nouveau_drm_open(struct drm_device *dev, struct drm_file *fpriv)
 
 	/* need to bring up power immediately if opening device */
 	ret = pm_runtime_get_sync(dev->dev);
-	if (ret < 0)
+	if (ret < 0 && ret != -EACCES)
 		return ret;
 
 	get_task_comm(tmpname, current);
 	snprintf(name, sizeof(name), "%s[%d]", tmpname, pid_nr(fpriv->pid));
 
-	ret = nouveau_cli_create(pdev, name, sizeof(*cli), (void **)&cli);
+	ret = nouveau_cli_create(nouveau_name(dev), name, sizeof(*cli),
+			(void **)&cli);
+
 	if (ret)
 		goto out_suspend;
 
@@ -753,7 +788,7 @@ long nouveau_drm_ioctl(struct file *filp,
 	dev = file_priv->minor->dev;
 
 	ret = pm_runtime_get_sync(dev->dev);
-	if (ret < 0)
+	if (ret < 0 && ret != -EACCES)
 		return ret;
 
 	ret = drm_ioctl(filp, cmd, arg);
@@ -798,6 +833,8 @@ driver = {
 	.get_vblank_counter = drm_vblank_count,
 	.enable_vblank = nouveau_display_vblank_enable,
 	.disable_vblank = nouveau_display_vblank_disable,
+	.get_scanout_position = nouveau_display_scanoutpos,
+	.get_vblank_timestamp = nouveau_display_vblstamp,
 
 	.ioctls = nouveau_ioctls,
 	.num_ioctls = ARRAY_SIZE(nouveau_ioctls),
@@ -855,20 +892,23 @@ static int nouveau_pmops_runtime_suspend(struct device *dev)
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
 	int ret;
 
-	if (nouveau_runtime_pm == 0)
-		return -EINVAL;
+	if (nouveau_runtime_pm == 0) {
+		pm_runtime_forbid(dev);
+		return -EBUSY;
+	}
 
 	/* are we optimus enabled? */
 	if (nouveau_runtime_pm == -1 && !nouveau_is_optimus() && !nouveau_is_v1_dsm()) {
 		DRM_DEBUG_DRIVER("failing to power off - not optimus\n");
-		return -EINVAL;
+		pm_runtime_forbid(dev);
+		return -EBUSY;
 	}
 
 	nv_debug_level(SILENT);
 	drm_kms_helper_poll_disable(drm_dev);
 	vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_OFF);
 	nouveau_switcheroo_optimus_dsm();
-	ret = nouveau_do_suspend(drm_dev);
+	ret = nouveau_do_suspend(drm_dev, true);
 	pci_save_state(pdev);
 	pci_disable_device(pdev);
 	pci_set_power_state(pdev, PCI_D3cold);
@@ -894,8 +934,6 @@ static int nouveau_pmops_runtime_resume(struct device *dev)
 	pci_set_master(pdev);
 
 	ret = nouveau_do_resume(drm_dev);
-	if (drm_dev->mode_config.num_crtc)
-		nouveau_display_resume(drm_dev);
 	drm_kms_helper_poll_enable(drm_dev);
 	/* do magic */
 	nv_mask(device, 0x88488, (1 << 25), (1 << 25));
@@ -912,12 +950,15 @@ static int nouveau_pmops_runtime_idle(struct device *dev)
 	struct nouveau_drm *drm = nouveau_drm(drm_dev);
 	struct drm_crtc *crtc;
 
-	if (nouveau_runtime_pm == 0)
+	if (nouveau_runtime_pm == 0) {
+		pm_runtime_forbid(dev);
 		return -EBUSY;
+	}
 
 	/* are we optimus enabled? */
 	if (nouveau_runtime_pm == -1 && !nouveau_is_optimus() && !nouveau_is_v1_dsm()) {
 		DRM_DEBUG_DRIVER("failing to power off - not optimus\n");
+		pm_runtime_forbid(dev);
 		return -EBUSY;
 	}
 
@@ -963,6 +1004,25 @@ nouveau_drm_pci_driver = {
 	.driver.pm = &nouveau_pm_ops,
 };
 
+int nouveau_drm_platform_probe(struct platform_device *pdev)
+{
+	struct nouveau_device *device;
+	int ret;
+
+	ret = nouveau_device_create(pdev, NOUVEAU_BUS_PLATFORM,
+				    nouveau_platform_name(pdev),
+				    dev_name(&pdev->dev), nouveau_config,
+				    nouveau_debug, &device);
+
+	ret = drm_platform_init(&driver, pdev);
+	if (ret) {
+		nouveau_object_ref(NULL, (struct nouveau_object **)&device);
+		return ret;
+	}
+
+	return ret;
+}
+
 static int __init
 nouveau_drm_init(void)
 {
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.h b/drivers/gpu/drm/nouveau/nouveau_drm.h
index 4b0fb6c66be..7efbafaf7c1 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.h
@@ -54,6 +54,7 @@ enum nouveau_drm_handle {
 	NVDRM_CLIENT  = 0xffffffff,
 	NVDRM_DEVICE  = 0xdddddddd,
 	NVDRM_CONTROL = 0xdddddddc,
+	NVDRM_DISPLAY = 0xd1500000,
 	NVDRM_PUSH    = 0xbbbb0000, /* |= client chid */
 	NVDRM_CHAN    = 0xcccc0000, /* |= client chid */
 	NVDRM_NVSW    = 0x55550000,
@@ -160,10 +161,7 @@ int nouveau_pmops_resume(struct device *);
 #define NV_ERROR(cli, fmt, args...) nv_error((cli), fmt, ##args)
 #define NV_WARN(cli, fmt, args...) nv_warn((cli), fmt, ##args)
 #define NV_INFO(cli, fmt, args...) nv_info((cli), fmt, ##args)
-#define NV_DEBUG(cli, fmt, args...) do {                                       \
-	if (drm_debug & DRM_UT_DRIVER)                                         \
-		nv_info((cli), fmt, ##args);                                   \
-} while (0)
+#define NV_DEBUG(cli, fmt, args...) nv_debug((cli), fmt, ##args)
 
 extern int nouveau_modeset;
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_encoder.h b/drivers/gpu/drm/nouveau/nouveau_encoder.h
index 24660c0f713..5f0e37fc284 100644
--- a/drivers/gpu/drm/nouveau/nouveau_encoder.h
+++ b/drivers/gpu/drm/nouveau/nouveau_encoder.h
@@ -46,6 +46,7 @@ struct nouveau_encoder {
 	/* different to drm_encoder.crtc, this reflects what's
 	 * actually programmed on the hw, not the proposed crtc */
 	struct drm_crtc *crtc;
+	u32 ctrl;
 
 	struct drm_display_mode mode;
 	int last_dpms;
@@ -84,9 +85,7 @@ get_slave_funcs(struct drm_encoder *enc)
 }
 
 /* nouveau_dp.c */
-bool nouveau_dp_detect(struct drm_encoder *);
-void nouveau_dp_dpms(struct drm_encoder *, int mode, u32 datarate,
-		     struct nouveau_object *);
+int nouveau_dp_detect(struct nouveau_encoder *);
 
 struct nouveau_connector *
 nouveau_encoder_connector_get(struct nouveau_encoder *encoder);
diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.c b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
index 7903e0ed3c7..191665ee7f5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fbcon.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
@@ -528,20 +528,13 @@ nouveau_fbcon_set_suspend(struct drm_device *dev, int state)
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	if (drm->fbcon) {
 		console_lock();
-		if (state == 0)
+		if (state == 1)
 			nouveau_fbcon_save_disable_accel(dev);
 		fb_set_suspend(drm->fbcon->helper.fbdev, state);
-		if (state == 1)
+		if (state == 0) {
 			nouveau_fbcon_restore_accel(dev);
+			nouveau_fbcon_zfill(dev, drm->fbcon);
+		}
 		console_unlock();
 	}
 }
-
-void
-nouveau_fbcon_zfill_all(struct drm_device *dev)
-{
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	if (drm->fbcon) {
-		nouveau_fbcon_zfill(dev, drm->fbcon);
-	}
-}
diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.h b/drivers/gpu/drm/nouveau/nouveau_fbcon.h
index fdfc0c94fbc..fcff797d208 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fbcon.h
+++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.h
@@ -61,7 +61,6 @@ void nouveau_fbcon_gpu_lockup(struct fb_info *info);
 int nouveau_fbcon_init(struct drm_device *dev);
 void nouveau_fbcon_fini(struct drm_device *dev);
 void nouveau_fbcon_set_suspend(struct drm_device *dev, int state);
-void nouveau_fbcon_zfill_all(struct drm_device *dev);
 void nouveau_fbcon_save_disable_accel(struct drm_device *dev);
 void nouveau_fbcon_restore_accel(struct drm_device *dev);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
index 40cf52e6d6d..ab5ea3b0d66 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
@@ -143,7 +143,7 @@ nouveau_fence_emit(struct nouveau_fence *fence, struct nouveau_channel *chan)
 	int ret;
 
 	fence->channel  = chan;
-	fence->timeout  = jiffies + (15 * DRM_HZ);
+	fence->timeout  = jiffies + (15 * HZ);
 	fence->sequence = ++fctx->sequence;
 
 	ret = fctx->emit(fence);
@@ -166,7 +166,7 @@ nouveau_fence_done(struct nouveau_fence *fence)
 }
 
 static int
-nouveau_fence_wait_uevent_handler(void *data, int index)
+nouveau_fence_wait_uevent_handler(void *data, u32 type, int index)
 {
 	struct nouveau_fence_priv *priv = data;
 	wake_up_all(&priv->waiting);
@@ -183,7 +183,7 @@ nouveau_fence_wait_uevent(struct nouveau_fence *fence, bool intr)
 	struct nouveau_eventh *handler;
 	int ret = 0;
 
-	ret = nouveau_event_new(pfifo->uevent, 0,
+	ret = nouveau_event_new(pfifo->uevent, 1, 0,
 				nouveau_fence_wait_uevent_handler,
 				priv, &handler);
 	if (ret)
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index 78a27f8ad7d..c90c0dc0afe 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -228,8 +228,6 @@ nouveau_gem_ioctl_new(struct drm_device *dev, void *data,
 	struct nouveau_bo *nvbo = NULL;
 	int ret = 0;
 
-	drm->ttm.bdev.dev_mapping = drm->dev->dev_mapping;
-
 	if (!pfb->memtype_valid(pfb, req->info.tile_flags)) {
 		NV_ERROR(cli, "bad page flags: 0x%08x\n", req->info.tile_flags);
 		return -EINVAL;
@@ -463,12 +461,6 @@ validate_list(struct nouveau_channel *chan, struct nouveau_cli *cli,
 	list_for_each_entry(nvbo, list, entry) {
 		struct drm_nouveau_gem_pushbuf_bo *b = &pbbo[nvbo->pbbo_index];
 
-		ret = validate_sync(chan, nvbo);
-		if (unlikely(ret)) {
-			NV_ERROR(cli, "fail pre-validate sync\n");
-			return ret;
-		}
-
 		ret = nouveau_gem_set_domain(&nvbo->gem, b->read_domains,
 					     b->write_domains,
 					     b->valid_domains);
@@ -506,7 +498,7 @@ validate_list(struct nouveau_channel *chan, struct nouveau_cli *cli,
 			b->presumed.valid = 0;
 			relocs++;
 
-			if (DRM_COPY_TO_USER(&upbbo[nvbo->pbbo_index].presumed,
+			if (copy_to_user(&upbbo[nvbo->pbbo_index].presumed,
 					     &b->presumed, sizeof(b->presumed)))
 				return -EFAULT;
 		}
@@ -593,7 +585,7 @@ u_memcpya(uint64_t user, unsigned nmemb, unsigned size)
 	if (!mem)
 		return ERR_PTR(-ENOMEM);
 
-	if (DRM_COPY_FROM_USER(mem, userptr, size)) {
+	if (copy_from_user(mem, userptr, size)) {
 		u_free(mem);
 		return ERR_PTR(-EFAULT);
 	}
diff --git a/drivers/gpu/drm/nouveau/nouveau_hwmon.c b/drivers/gpu/drm/nouveau/nouveau_hwmon.c
index 4aff04fa483..19fd767bab1 100644
--- a/drivers/gpu/drm/nouveau/nouveau_hwmon.c
+++ b/drivers/gpu/drm/nouveau/nouveau_hwmon.c
@@ -383,8 +383,9 @@ nouveau_hwmon_set_pwm1_enable(struct device *d, struct device_attribute *a,
 	long value;
 	int ret;
 
-	if (strict_strtol(buf, 10, &value) == -EINVAL)
-		return -EINVAL;
+	ret = kstrtol(buf, 10, &value);
+	if (ret)
+		return ret;
 
 	ret = therm->attr_set(therm, NOUVEAU_THERM_ATTR_FAN_MODE, value);
 	if (ret)
@@ -587,18 +588,14 @@ nouveau_hwmon_init(struct drm_device *dev)
 
 	/* set the default attributes */
 	ret = sysfs_create_group(&hwmon_dev->kobj, &hwmon_default_attrgroup);
-	if (ret) {
-		if (ret)
-			goto error;
-	}
+	if (ret)
+		goto error;
 
 	/* if the card has a working thermal sensor */
 	if (therm->temp_get(therm) >= 0) {
 		ret = sysfs_create_group(&hwmon_dev->kobj, &hwmon_temp_attrgroup);
-		if (ret) {
-			if (ret)
-				goto error;
-		}
+		if (ret)
+			goto error;
 	}
 
 	/* if the card has a pwm fan */
diff --git a/drivers/gpu/drm/nouveau/nouveau_ioc32.c b/drivers/gpu/drm/nouveau/nouveau_ioc32.c
index c1a7e5a73a2..462679a8fec 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ioc32.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ioc32.c
@@ -57,7 +57,7 @@ long nouveau_compat_ioctl(struct file *filp, unsigned int cmd,
 		return drm_compat_ioctl(filp, cmd, arg);
 
 #if 0
-	if (nr < DRM_COMMAND_BASE + DRM_ARRAY_SIZE(mga_compat_ioctls))
+	if (nr < DRM_COMMAND_BASE + ARRAY_SIZE(mga_compat_ioctls))
 		fn = nouveau_compat_ioctls[nr - DRM_COMMAND_BASE];
 #endif
 	if (fn != NULL)
diff --git a/drivers/gpu/drm/nouveau/nouveau_sgdma.c b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
index 0843ebc910d..a4d22e5eb17 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sgdma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
@@ -31,16 +31,17 @@ nv04_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *mem)
 {
 	struct nouveau_sgdma_be *nvbe = (struct nouveau_sgdma_be *)ttm;
 	struct nouveau_mem *node = mem->mm_node;
-	u64 size = mem->num_pages << 12;
 
 	if (ttm->sg) {
-		node->sg = ttm->sg;
-		nouveau_vm_map_sg_table(&node->vma[0], 0, size, node);
+		node->sg    = ttm->sg;
+		node->pages = NULL;
 	} else {
+		node->sg    = NULL;
 		node->pages = nvbe->ttm.dma_address;
-		nouveau_vm_map_sg(&node->vma[0], 0, size, node);
 	}
+	node->size = (mem->num_pages << PAGE_SHIFT) >> 12;
 
+	nouveau_vm_map(&node->vma[0], node);
 	nvbe->node = node;
 	return 0;
 }
@@ -67,9 +68,13 @@ nv50_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *mem)
 
 	/* noop: bound in move_notify() */
 	if (ttm->sg) {
-		node->sg = ttm->sg;
-	} else
+		node->sg    = ttm->sg;
+		node->pages = NULL;
+	} else {
+		node->sg    = NULL;
 		node->pages = nvbe->ttm.dma_address;
+	}
+	node->size = (mem->num_pages << PAGE_SHIFT) >> 12;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_sysfs.c b/drivers/gpu/drm/nouveau/nouveau_sysfs.c
index 89201a17ce7..75dda2b0717 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sysfs.c
+++ b/drivers/gpu/drm/nouveau/nouveau_sysfs.c
@@ -30,7 +30,7 @@
 static inline struct drm_device *
 drm_device(struct device *d)
 {
-	return pci_get_drvdata(to_pci_dev(d));
+	return dev_get_drvdata(d);
 }
 
 #define snappendf(p,r,f,a...) do {                                             \
@@ -132,9 +132,10 @@ nouveau_sysfs_fini(struct drm_device *dev)
 {
 	struct nouveau_sysfs *sysfs = nouveau_sysfs(dev);
 	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nouveau_device *device = nv_device(drm->device);
 
 	if (sysfs->ctrl) {
-		device_remove_file(&dev->pdev->dev, &dev_attr_pstate);
+		device_remove_file(nv_device_base(device), &dev_attr_pstate);
 		nouveau_object_del(nv_object(drm), NVDRM_DEVICE, NVDRM_CONTROL);
 	}
 
@@ -146,6 +147,7 @@ int
 nouveau_sysfs_init(struct drm_device *dev)
 {
 	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nouveau_device *device = nv_device(drm->device);
 	struct nouveau_sysfs *sysfs;
 	int ret;
 
@@ -156,7 +158,7 @@ nouveau_sysfs_init(struct drm_device *dev)
 	ret = nouveau_object_new(nv_object(drm), NVDRM_DEVICE, NVDRM_CONTROL,
 				 NV_CONTROL_CLASS, NULL, 0, &sysfs->ctrl);
 	if (ret == 0)
-		device_create_file(&dev->pdev->dev, &dev_attr_pstate);
+		device_create_file(nv_device_base(device), &dev_attr_pstate);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index 19e3757291f..ab0228f640a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -171,6 +171,7 @@ nouveau_gart_manager_new(struct ttm_mem_type_manager *man,
 	node = kzalloc(sizeof(*node), GFP_KERNEL);
 	if (!node)
 		return -ENOMEM;
+
 	node->page_shift = 12;
 
 	switch (nv_device(drm->device)->card_type) {
@@ -353,21 +354,26 @@ int
 nouveau_ttm_init(struct nouveau_drm *drm)
 {
 	struct drm_device *dev = drm->dev;
+	struct nouveau_device *device = nv_device(drm->device);
 	u32 bits;
 	int ret;
 
 	bits = nouveau_vmmgr(drm->device)->dma_bits;
-	if ( drm->agp.stat == ENABLED ||
-	    !pci_dma_supported(dev->pdev, DMA_BIT_MASK(bits)))
-		bits = 32;
-
-	ret = pci_set_dma_mask(dev->pdev, DMA_BIT_MASK(bits));
-	if (ret)
-		return ret;
-
-	ret = pci_set_consistent_dma_mask(dev->pdev, DMA_BIT_MASK(bits));
-	if (ret)
-		pci_set_consistent_dma_mask(dev->pdev, DMA_BIT_MASK(32));
+	if (nv_device_is_pci(device)) {
+		if (drm->agp.stat == ENABLED ||
+		     !pci_dma_supported(dev->pdev, DMA_BIT_MASK(bits)))
+			bits = 32;
+
+		ret = pci_set_dma_mask(dev->pdev, DMA_BIT_MASK(bits));
+		if (ret)
+			return ret;
+
+		ret = pci_set_consistent_dma_mask(dev->pdev,
+						  DMA_BIT_MASK(bits));
+		if (ret)
+			pci_set_consistent_dma_mask(dev->pdev,
+						    DMA_BIT_MASK(32));
+	}
 
 	ret = nouveau_ttm_global_init(drm);
 	if (ret)
@@ -375,7 +381,9 @@ nouveau_ttm_init(struct nouveau_drm *drm)
 
 	ret = ttm_bo_device_init(&drm->ttm.bdev,
 				  drm->ttm.bo_global_ref.ref.object,
-				  &nouveau_bo_driver, DRM_FILE_PAGE_OFFSET,
+				  &nouveau_bo_driver,
+				  dev->anon_inode->i_mapping,
+				  DRM_FILE_PAGE_OFFSET,
 				  bits <= 32 ? true : false);
 	if (ret) {
 		NV_ERROR(drm, "error initialising bo driver, %d\n", ret);
@@ -393,8 +401,8 @@ nouveau_ttm_init(struct nouveau_drm *drm)
 		return ret;
 	}
 
-	drm->ttm.mtrr = arch_phys_wc_add(pci_resource_start(dev->pdev, 1),
-					 pci_resource_len(dev->pdev, 1));
+	drm->ttm.mtrr = arch_phys_wc_add(nv_device_resource_start(device, 1),
+					 nv_device_resource_len(device, 1));
 
 	/* GART init */
 	if (drm->agp.stat != ENABLED) {
diff --git a/drivers/gpu/drm/nouveau/nouveau_vga.c b/drivers/gpu/drm/nouveau/nouveau_vga.c
index 81638d7f2ef..4f4c3fec691 100644
--- a/drivers/gpu/drm/nouveau/nouveau_vga.c
+++ b/drivers/gpu/drm/nouveau/nouveau_vga.c
@@ -14,7 +14,9 @@ nouveau_vga_set_decode(void *priv, bool state)
 {
 	struct nouveau_device *device = nouveau_dev(priv);
 
-	if (device->chipset >= 0x40)
+	if (device->card_type == NV_40 && device->chipset >= 0x4c)
+		nv_wr32(device, 0x088060, state);
+	else if (device->chipset >= 0x40)
 		nv_wr32(device, 0x088054, state);
 	else
 		nv_wr32(device, 0x001854, state);
@@ -62,12 +64,13 @@ static bool
 nouveau_switcheroo_can_switch(struct pci_dev *pdev)
 {
 	struct drm_device *dev = pci_get_drvdata(pdev);
-	bool can_switch;
 
-	spin_lock(&dev->count_lock);
-	can_switch = (dev->open_count == 0);
-	spin_unlock(&dev->count_lock);
-	return can_switch;
+	/*
+	 * FIXME: open_count is protected by drm_global_mutex but that would lead to
+	 * locking inversion with the driver load path. And the access here is
+	 * completely racy anyway. So don't bother with locking for now.
+	 */
+	return dev->open_count == 0;
 }
 
 static const struct vga_switcheroo_client_ops
@@ -82,6 +85,11 @@ nouveau_vga_init(struct nouveau_drm *drm)
 {
 	struct drm_device *dev = drm->dev;
 	bool runtime = false;
+
+	/* only relevant for PCI devices */
+	if (!dev->pdev)
+		return;
+
 	vga_client_register(dev->pdev, dev, NULL, nouveau_vga_set_decode);
 
 	if (nouveau_runtime_pm == 1)
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index 4e384a2f99c..4c534b7b04d 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -1,4 +1,4 @@
-	/*
+/*
  * Copyright 2011 Red Hat Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
@@ -26,6 +26,7 @@
 
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_dp_helper.h>
 
 #include "nouveau_drm.h"
 #include "nouveau_dma.h"
@@ -651,7 +652,7 @@ nv50_crtc_set_dither(struct nouveau_crtc *nv_crtc, bool update)
 	nv_connector = nouveau_crtc_connector_get(nv_crtc);
 	connector = &nv_connector->base;
 	if (nv_connector->dithering_mode == DITHERING_MODE_AUTO) {
-		if (nv_crtc->base.fb->depth > connector->display_info.bpc * 3)
+		if (nv_crtc->base.primary->fb->depth > connector->display_info.bpc * 3)
 			mode = DITHERING_MODE_DYNAMIC2X2;
 	} else {
 		mode = nv_connector->dithering_mode;
@@ -785,7 +786,8 @@ nv50_crtc_set_scale(struct nouveau_crtc *nv_crtc, bool update)
 
 		if (update) {
 			nv50_display_flip_stop(crtc);
-			nv50_display_flip_next(crtc, crtc->fb, NULL, 1);
+			nv50_display_flip_next(crtc, crtc->primary->fb,
+					       NULL, 1);
 		}
 	}
 
@@ -956,7 +958,7 @@ nv50_crtc_prepare(struct drm_crtc *crtc)
 
 	nv50_display_flip_stop(crtc);
 
-	push = evo_wait(mast, 2);
+	push = evo_wait(mast, 6);
 	if (push) {
 		if (nv50_vers(mast) < NV84_DISP_MAST_CLASS) {
 			evo_mthd(push, 0x0874 + (nv_crtc->index * 0x400), 1);
@@ -1028,20 +1030,21 @@ nv50_crtc_commit(struct drm_crtc *crtc)
 	}
 
 	nv50_crtc_cursor_show_hide(nv_crtc, nv_crtc->cursor.visible, true);
-	nv50_display_flip_next(crtc, crtc->fb, NULL, 1);
+	nv50_display_flip_next(crtc, crtc->primary->fb, NULL, 1);
 }
 
 static bool
 nv50_crtc_mode_fixup(struct drm_crtc *crtc, const struct drm_display_mode *mode,
 		     struct drm_display_mode *adjusted_mode)
 {
+	drm_mode_set_crtcinfo(adjusted_mode, CRTC_INTERLACE_HALVE_V);
 	return true;
 }
 
 static int
 nv50_crtc_swap_fbs(struct drm_crtc *crtc, struct drm_framebuffer *old_fb)
 {
-	struct nouveau_framebuffer *nvfb = nouveau_framebuffer(crtc->fb);
+	struct nouveau_framebuffer *nvfb = nouveau_framebuffer(crtc->primary->fb);
 	struct nv50_head *head = nv50_head(crtc);
 	int ret;
 
@@ -1138,7 +1141,7 @@ nv50_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *umode,
 	nv50_crtc_set_dither(nv_crtc, false);
 	nv50_crtc_set_scale(nv_crtc, false);
 	nv50_crtc_set_color_vibrance(nv_crtc, false);
-	nv50_crtc_set_image(nv_crtc, crtc->fb, x, y, false);
+	nv50_crtc_set_image(nv_crtc, crtc->primary->fb, x, y, false);
 	return 0;
 }
 
@@ -1150,7 +1153,7 @@ nv50_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
 	int ret;
 
-	if (!crtc->fb) {
+	if (!crtc->primary->fb) {
 		NV_DEBUG(drm, "No FB bound\n");
 		return 0;
 	}
@@ -1160,8 +1163,8 @@ nv50_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 		return ret;
 
 	nv50_display_flip_stop(crtc);
-	nv50_crtc_set_image(nv_crtc, crtc->fb, x, y, true);
-	nv50_display_flip_next(crtc, crtc->fb, NULL, 1);
+	nv50_crtc_set_image(nv_crtc, crtc->primary->fb, x, y, true);
+	nv50_display_flip_next(crtc, crtc->primary->fb, NULL, 1);
 	return 0;
 }
 
@@ -1205,6 +1208,7 @@ static void
 nv50_crtc_disable(struct drm_crtc *crtc)
 {
 	struct nv50_head *head = nv50_head(crtc);
+	evo_sync(crtc->dev);
 	if (head->image)
 		nouveau_bo_unpin(head->image);
 	nouveau_bo_ref(NULL, &head->image);
@@ -1698,10 +1702,9 @@ nv50_hdmi_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode)
 }
 
 static void
-nv50_hdmi_disconnect(struct drm_encoder *encoder)
+nv50_hdmi_disconnect(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
 {
 	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(nv_encoder->crtc);
 	struct nv50_disp *disp = nv50_disp(encoder->dev);
 	const u32 moff = (nv_crtc->index << 3) | nv_encoder->or;
 
@@ -1720,7 +1723,7 @@ nv50_sor_dpms(struct drm_encoder *encoder, int mode)
 	struct drm_device *dev = encoder->dev;
 	struct nv50_disp *disp = nv50_disp(dev);
 	struct drm_encoder *partner;
-	int or = nv_encoder->or;
+	u32 mthd;
 
 	nv_encoder->last_dpms = mode;
 
@@ -1738,7 +1741,18 @@ nv50_sor_dpms(struct drm_encoder *encoder, int mode)
 		}
 	}
 
-	nv_call(disp->core, NV50_DISP_SOR_PWR + or, (mode == DRM_MODE_DPMS_ON));
+	mthd  = (ffs(nv_encoder->dcb->heads) - 1) << 3;
+	mthd |= (ffs(nv_encoder->dcb->sorconf.link) - 1) << 2;
+	mthd |= nv_encoder->or;
+
+	if (nv_encoder->dcb->type == DCB_OUTPUT_DP) {
+		nv_call(disp->core, NV50_DISP_SOR_PWR | mthd, 1);
+		mthd |= NV94_DISP_SOR_DP_PWR;
+	} else {
+		mthd |= NV50_DISP_SOR_PWR;
+	}
+
+	nv_call(disp->core, mthd, (mode == DRM_MODE_DPMS_ON));
 }
 
 static bool
@@ -1762,33 +1776,36 @@ nv50_sor_mode_fixup(struct drm_encoder *encoder,
 }
 
 static void
-nv50_sor_disconnect(struct drm_encoder *encoder)
+nv50_sor_ctrl(struct nouveau_encoder *nv_encoder, u32 mask, u32 data)
 {
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
-	const int or = nv_encoder->or;
-	u32 *push;
-
-	if (nv_encoder->crtc) {
-		nv50_crtc_prepare(nv_encoder->crtc);
-
-		push = evo_wait(mast, 4);
-		if (push) {
-			if (nv50_vers(mast) < NVD0_DISP_MAST_CLASS) {
-				evo_mthd(push, 0x0600 + (or * 0x40), 1);
-				evo_data(push, 0x00000000);
-			} else {
-				evo_mthd(push, 0x0200 + (or * 0x20), 1);
-				evo_data(push, 0x00000000);
-			}
-			evo_kick(push, mast);
+	struct nv50_mast *mast = nv50_mast(nv_encoder->base.base.dev);
+	u32 temp = (nv_encoder->ctrl & ~mask) | (data & mask), *push;
+	if (temp != nv_encoder->ctrl && (push = evo_wait(mast, 2))) {
+		if (nv50_vers(mast) < NVD0_DISP_MAST_CLASS) {
+			evo_mthd(push, 0x0600 + (nv_encoder->or * 0x40), 1);
+			evo_data(push, (nv_encoder->ctrl = temp));
+		} else {
+			evo_mthd(push, 0x0200 + (nv_encoder->or * 0x20), 1);
+			evo_data(push, (nv_encoder->ctrl = temp));
 		}
-
-		nv50_hdmi_disconnect(encoder);
+		evo_kick(push, mast);
 	}
+}
+
+static void
+nv50_sor_disconnect(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(nv_encoder->crtc);
 
 	nv_encoder->last_dpms = DRM_MODE_DPMS_OFF;
 	nv_encoder->crtc = NULL;
+
+	if (nv_crtc) {
+		nv50_crtc_prepare(&nv_crtc->base);
+		nv50_sor_ctrl(nv_encoder, 1 << nv_crtc->index, 0);
+		nv50_hdmi_disconnect(&nv_encoder->base.base, nv_crtc);
+	}
 }
 
 static void
@@ -1808,12 +1825,14 @@ nv50_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *umode,
 	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
 	struct nouveau_connector *nv_connector;
 	struct nvbios *bios = &drm->vbios;
-	u32 *push, lvds = 0;
+	u32 lvds = 0, mask, ctrl;
 	u8 owner = 1 << nv_crtc->index;
 	u8 proto = 0xf;
 	u8 depth = 0x0;
 
 	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	nv_encoder->crtc = encoder->crtc;
+
 	switch (nv_encoder->dcb->type) {
 	case DCB_OUTPUT_TMDS:
 		if (nv_encoder->dcb->sorconf.link & 1) {
@@ -1825,7 +1844,7 @@ nv50_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *umode,
 			proto = 0x2;
 		}
 
-		nv50_hdmi_mode_set(encoder, mode);
+		nv50_hdmi_mode_set(&nv_encoder->base.base, mode);
 		break;
 	case DCB_OUTPUT_LVDS:
 		proto = 0x0;
@@ -1881,19 +1900,11 @@ nv50_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *umode,
 		break;
 	}
 
-	nv50_sor_dpms(encoder, DRM_MODE_DPMS_ON);
+	nv50_sor_dpms(&nv_encoder->base.base, DRM_MODE_DPMS_ON);
 
-	push = evo_wait(nv50_mast(dev), 8);
-	if (push) {
-		if (nv50_vers(mast) < NVD0_DISP_CLASS) {
-			u32 ctrl = (depth << 16) | (proto << 8) | owner;
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				ctrl |= 0x00001000;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				ctrl |= 0x00002000;
-			evo_mthd(push, 0x0600 + (nv_encoder->or * 0x040), 1);
-			evo_data(push, ctrl);
-		} else {
+	if (nv50_vers(mast) >= NVD0_DISP_CLASS) {
+		u32 *push = evo_wait(mast, 3);
+		if (push) {
 			u32 magic = 0x31ec6000 | (nv_crtc->index << 25);
 			u32 syncs = 0x00000001;
 
@@ -1908,14 +1919,21 @@ nv50_sor_mode_set(struct drm_encoder *encoder, struct drm_display_mode *umode,
 			evo_mthd(push, 0x0404 + (nv_crtc->index * 0x300), 2);
 			evo_data(push, syncs | (depth << 6));
 			evo_data(push, magic);
-			evo_mthd(push, 0x0200 + (nv_encoder->or * 0x020), 1);
-			evo_data(push, owner | (proto << 8));
+			evo_kick(push, mast);
 		}
 
-		evo_kick(push, mast);
+		ctrl = proto << 8;
+		mask = 0x00000f00;
+	} else {
+		ctrl = (depth << 16) | (proto << 8);
+		if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+			ctrl |= 0x00001000;
+		if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+			ctrl |= 0x00002000;
+		mask = 0x000f3f00;
 	}
 
-	nv_encoder->crtc = encoder->crtc;
+	nv50_sor_ctrl(nv_encoder, mask | owner, ctrl | owner);
 }
 
 static void
@@ -2199,16 +2217,6 @@ nv50_display_destroy(struct drm_device *dev)
 int
 nv50_display_create(struct drm_device *dev)
 {
-	static const u16 oclass[] = {
-		NVF0_DISP_CLASS,
-		NVE0_DISP_CLASS,
-		NVD0_DISP_CLASS,
-		NVA3_DISP_CLASS,
-		NV94_DISP_CLASS,
-		NVA0_DISP_CLASS,
-		NV84_DISP_CLASS,
-		NV50_DISP_CLASS,
-	};
 	struct nouveau_device *device = nouveau_dev(dev);
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct dcb_table *dcb = &drm->vbios.dcb;
@@ -2225,6 +2233,7 @@ nv50_display_create(struct drm_device *dev)
 	nouveau_display(dev)->dtor = nv50_display_destroy;
 	nouveau_display(dev)->init = nv50_display_init;
 	nouveau_display(dev)->fini = nv50_display_fini;
+	disp->core = nouveau_display(dev)->core;
 
 	/* small shared memory area we use for notifiers and semaphores */
 	ret = nouveau_bo_new(dev, 4096, 0x1000, TTM_PL_FLAG_VRAM,
@@ -2243,17 +2252,6 @@ nv50_display_create(struct drm_device *dev)
 	if (ret)
 		goto out;
 
-	/* attempt to allocate a supported evo display class */
-	ret = -ENODEV;
-	for (i = 0; ret && i < ARRAY_SIZE(oclass); i++) {
-		ret = nouveau_object_new(nv_object(drm), NVDRM_DEVICE,
-					 0xd1500000, oclass[i], NULL, 0,
-					 &disp->core);
-	}
-
-	if (ret)
-		goto out;
-
 	/* allocate master evo channel */
 	ret = nv50_dmac_create(disp->core, NV50_DISP_MAST_CLASS, 0,
 			      &(struct nv50_display_mast_class) {
@@ -2313,7 +2311,7 @@ nv50_display_create(struct drm_device *dev)
 			continue;
 
 		NV_WARN(drm, "%s has no encoders, removing\n",
-			drm_get_connector_name(connector));
+			connector->name);
 		connector->funcs->destroy(connector);
 	}
 
diff --git a/drivers/gpu/drm/omapdrm/omap_connector.c b/drivers/gpu/drm/omapdrm/omap_connector.c
index 912759daf56..86f4ead0441 100644
--- a/drivers/gpu/drm/omapdrm/omap_connector.c
+++ b/drivers/gpu/drm/omapdrm/omap_connector.c
@@ -37,7 +37,7 @@ struct omap_connector {
 void copy_timings_omap_to_drm(struct drm_display_mode *mode,
 		struct omap_video_timings *timings)
 {
-	mode->clock = timings->pixel_clock;
+	mode->clock = timings->pixelclock / 1000;
 
 	mode->hdisplay = timings->x_res;
 	mode->hsync_start = mode->hdisplay + timings->hfp;
@@ -68,7 +68,7 @@ void copy_timings_omap_to_drm(struct drm_display_mode *mode,
 void copy_timings_drm_to_omap(struct omap_video_timings *timings,
 		struct drm_display_mode *mode)
 {
-	timings->pixel_clock = mode->clock;
+	timings->pixelclock = mode->clock * 1000;
 
 	timings->x_res = mode->hdisplay;
 	timings->hfp = mode->hsync_start - mode->hdisplay;
@@ -220,7 +220,7 @@ static int omap_connector_mode_valid(struct drm_connector *connector,
 	if (!r) {
 		/* check if vrefresh is still valid */
 		new_mode = drm_mode_duplicate(dev, mode);
-		new_mode->clock = timings.pixel_clock;
+		new_mode->clock = timings.pixelclock / 1000;
 		new_mode->vrefresh = 0;
 		if (mode->vrefresh == drm_mode_vrefresh(new_mode))
 			ret = MODE_OK;
diff --git a/drivers/gpu/drm/omapdrm/omap_crtc.c b/drivers/gpu/drm/omapdrm/omap_crtc.c
index 0fd2eb139f6..2d28dc337cf 100644
--- a/drivers/gpu/drm/omapdrm/omap_crtc.c
+++ b/drivers/gpu/drm/omapdrm/omap_crtc.c
@@ -33,6 +33,7 @@ struct omap_crtc {
 	int pipe;
 	enum omap_channel channel;
 	struct omap_overlay_manager_info info;
+	struct drm_encoder *current_encoder;
 
 	/*
 	 * Temporary: eventually this will go away, but it is needed
@@ -120,13 +121,25 @@ static void omap_crtc_start_update(struct omap_overlay_manager *mgr)
 {
 }
 
+static void set_enabled(struct drm_crtc *crtc, bool enable);
+
 static int omap_crtc_enable(struct omap_overlay_manager *mgr)
 {
+	struct omap_crtc *omap_crtc = omap_crtcs[mgr->id];
+
+	dispc_mgr_setup(omap_crtc->channel, &omap_crtc->info);
+	dispc_mgr_set_timings(omap_crtc->channel,
+			&omap_crtc->timings);
+	set_enabled(&omap_crtc->base, true);
+
 	return 0;
 }
 
 static void omap_crtc_disable(struct omap_overlay_manager *mgr)
 {
+	struct omap_crtc *omap_crtc = omap_crtcs[mgr->id];
+
+	set_enabled(&omap_crtc->base, false);
 }
 
 static void omap_crtc_set_timings(struct omap_overlay_manager *mgr,
@@ -184,7 +197,6 @@ static void omap_crtc_destroy(struct drm_crtc *crtc)
 	WARN_ON(omap_crtc->apply_irq.registered);
 	omap_irq_unregister(crtc->dev, &omap_crtc->error_irq);
 
-	omap_crtc->plane->funcs->destroy(omap_crtc->plane);
 	drm_crtc_cleanup(crtc);
 
 	kfree(omap_crtc);
@@ -245,7 +257,7 @@ static int omap_crtc_mode_set(struct drm_crtc *crtc,
 	copy_timings_drm_to_omap(&omap_crtc->timings, mode);
 	omap_crtc->full_update = true;
 
-	return omap_plane_mode_set(omap_crtc->plane, crtc, crtc->fb,
+	return omap_plane_mode_set(omap_crtc->plane, crtc, crtc->primary->fb,
 			0, 0, mode->hdisplay, mode->vdisplay,
 			x << 16, y << 16,
 			mode->hdisplay << 16, mode->vdisplay << 16,
@@ -273,7 +285,7 @@ static int omap_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 	struct drm_plane *plane = omap_crtc->plane;
 	struct drm_display_mode *mode = &crtc->mode;
 
-	return omap_plane_mode_set(plane, crtc, crtc->fb,
+	return omap_plane_mode_set(plane, crtc, crtc->primary->fb,
 			0, 0, mode->hdisplay, mode->vdisplay,
 			x << 16, y << 16,
 			mode->hdisplay << 16, mode->vdisplay << 16,
@@ -307,15 +319,15 @@ static void page_flip_worker(struct work_struct *work)
 	struct drm_display_mode *mode = &crtc->mode;
 	struct drm_gem_object *bo;
 
-	mutex_lock(&crtc->mutex);
-	omap_plane_mode_set(omap_crtc->plane, crtc, crtc->fb,
+	drm_modeset_lock(&crtc->mutex, NULL);
+	omap_plane_mode_set(omap_crtc->plane, crtc, crtc->primary->fb,
 			0, 0, mode->hdisplay, mode->vdisplay,
 			crtc->x << 16, crtc->y << 16,
 			mode->hdisplay << 16, mode->vdisplay << 16,
 			vblank_cb, crtc);
-	mutex_unlock(&crtc->mutex);
+	drm_modeset_unlock(&crtc->mutex);
 
-	bo = omap_framebuffer_bo(crtc->fb, 0);
+	bo = omap_framebuffer_bo(crtc->primary->fb, 0);
 	drm_gem_object_unreference_unlocked(bo);
 }
 
@@ -336,18 +348,25 @@ static int omap_crtc_page_flip_locked(struct drm_crtc *crtc,
 {
 	struct drm_device *dev = crtc->dev;
 	struct omap_crtc *omap_crtc = to_omap_crtc(crtc);
+	struct drm_plane *primary = crtc->primary;
 	struct drm_gem_object *bo;
+	unsigned long flags;
 
-	DBG("%d -> %d (event=%p)", crtc->fb ? crtc->fb->base.id : -1,
+	DBG("%d -> %d (event=%p)", primary->fb ? primary->fb->base.id : -1,
 			fb->base.id, event);
 
+	spin_lock_irqsave(&dev->event_lock, flags);
+
 	if (omap_crtc->old_fb) {
+		spin_unlock_irqrestore(&dev->event_lock, flags);
 		dev_err(dev->dev, "already a pending flip\n");
 		return -EINVAL;
 	}
 
 	omap_crtc->event = event;
-	crtc->fb = fb;
+	omap_crtc->old_fb = primary->fb = fb;
+
+	spin_unlock_irqrestore(&dev->event_lock, flags);
 
 	/*
 	 * Hold a reference temporarily until the crtc is updated
@@ -411,7 +430,7 @@ static void omap_crtc_error_irq(struct omap_drm_irq *irq, uint32_t irqstatus)
 	struct drm_crtc *crtc = &omap_crtc->base;
 	DRM_ERROR("%s: errors: %08x\n", omap_crtc->name, irqstatus);
 	/* avoid getting in a flood, unregister the irq until next vblank */
-	omap_irq_unregister(crtc->dev, &omap_crtc->error_irq);
+	__omap_irq_unregister(crtc->dev, &omap_crtc->error_irq);
 }
 
 static void omap_crtc_apply_irq(struct omap_drm_irq *irq, uint32_t irqstatus)
@@ -421,13 +440,13 @@ static void omap_crtc_apply_irq(struct omap_drm_irq *irq, uint32_t irqstatus)
 	struct drm_crtc *crtc = &omap_crtc->base;
 
 	if (!omap_crtc->error_irq.registered)
-		omap_irq_register(crtc->dev, &omap_crtc->error_irq);
+		__omap_irq_register(crtc->dev, &omap_crtc->error_irq);
 
 	if (!dispc_mgr_go_busy(omap_crtc->channel)) {
 		struct omap_drm_private *priv =
 				crtc->dev->dev_private;
 		DBG("%s: apply done", omap_crtc->name);
-		omap_irq_unregister(crtc->dev, &omap_crtc->apply_irq);
+		__omap_irq_unregister(crtc->dev, &omap_crtc->apply_irq);
 		queue_work(priv->wq, &omap_crtc->apply_work);
 	}
 }
@@ -446,7 +465,7 @@ static void apply_worker(struct work_struct *work)
 	 * the callbacks and list modification all serialized
 	 * with respect to modesetting ioctls from userspace.
 	 */
-	mutex_lock(&crtc->mutex);
+	drm_modeset_lock(&crtc->mutex, NULL);
 	dispc_runtime_get();
 
 	/*
@@ -491,7 +510,7 @@ static void apply_worker(struct work_struct *work)
 
 out:
 	dispc_runtime_put();
-	mutex_unlock(&crtc->mutex);
+	drm_modeset_unlock(&crtc->mutex);
 }
 
 int omap_crtc_apply(struct drm_crtc *crtc,
@@ -499,7 +518,7 @@ int omap_crtc_apply(struct drm_crtc *crtc,
 {
 	struct omap_crtc *omap_crtc = to_omap_crtc(crtc);
 
-	WARN_ON(!mutex_is_locked(&crtc->mutex));
+	WARN_ON(!drm_modeset_is_locked(&crtc->mutex));
 
 	/* no need to queue it again if it is already queued: */
 	if (apply->queued)
@@ -527,38 +546,46 @@ static void set_enabled(struct drm_crtc *crtc, bool enable)
 	struct drm_device *dev = crtc->dev;
 	struct omap_crtc *omap_crtc = to_omap_crtc(crtc);
 	enum omap_channel channel = omap_crtc->channel;
-	struct omap_irq_wait *wait = NULL;
+	struct omap_irq_wait *wait;
+	u32 framedone_irq, vsync_irq;
+	int ret;
 
 	if (dispc_mgr_is_enabled(channel) == enable)
 		return;
 
-	/* ignore sync-lost irqs during enable/disable */
+	/*
+	 * Digit output produces some sync lost interrupts during the first
+	 * frame when enabling, so we need to ignore those.
+	 */
 	omap_irq_unregister(crtc->dev, &omap_crtc->error_irq);
 
-	if (dispc_mgr_get_framedone_irq(channel)) {
-		if (!enable) {
-			wait = omap_irq_wait_init(dev,
-					dispc_mgr_get_framedone_irq(channel), 1);
-		}
+	framedone_irq = dispc_mgr_get_framedone_irq(channel);
+	vsync_irq = dispc_mgr_get_vsync_irq(channel);
+
+	if (enable) {
+		wait = omap_irq_wait_init(dev, vsync_irq, 1);
 	} else {
 		/*
-		 * When we disable digit output, we need to wait until fields
-		 * are done.  Otherwise the DSS is still working, and turning
-		 * off the clocks prevents DSS from going to OFF mode. And when
-		 * enabling, we need to wait for the extra sync losts
+		 * When we disable the digit output, we need to wait for
+		 * FRAMEDONE to know that DISPC has finished with the output.
+		 *
+		 * OMAP2/3 does not have FRAMEDONE irq for digit output, and in
+		 * that case we need to use vsync interrupt, and wait for both
+		 * even and odd frames.
 		 */
-		wait = omap_irq_wait_init(dev,
-				dispc_mgr_get_vsync_irq(channel), 2);
+
+		if (framedone_irq)
+			wait = omap_irq_wait_init(dev, framedone_irq, 1);
+		else
+			wait = omap_irq_wait_init(dev, vsync_irq, 2);
 	}
 
 	dispc_mgr_enable(channel, enable);
 
-	if (wait) {
-		int ret = omap_irq_wait(dev, wait, msecs_to_jiffies(100));
-		if (ret) {
-			dev_err(dev->dev, "%s: timeout waiting for %s\n",
-					omap_crtc->name, enable ? "enable" : "disable");
-		}
+	ret = omap_irq_wait(dev, wait, msecs_to_jiffies(100));
+	if (ret) {
+		dev_err(dev->dev, "%s: timeout waiting for %s\n",
+				omap_crtc->name, enable ? "enable" : "disable");
 	}
 
 	omap_irq_register(crtc->dev, &omap_crtc->error_irq);
@@ -585,8 +612,12 @@ static void omap_crtc_pre_apply(struct omap_drm_apply *apply)
 		}
 	}
 
+	if (omap_crtc->current_encoder && encoder != omap_crtc->current_encoder)
+		omap_encoder_set_enabled(omap_crtc->current_encoder, false);
+
+	omap_crtc->current_encoder = encoder;
+
 	if (!omap_crtc->enabled) {
-		set_enabled(&omap_crtc->base, false);
 		if (encoder)
 			omap_encoder_set_enabled(encoder, false);
 	} else {
@@ -595,13 +626,7 @@ static void omap_crtc_pre_apply(struct omap_drm_apply *apply)
 			omap_encoder_update(encoder, omap_crtc->mgr,
 					&omap_crtc->timings);
 			omap_encoder_set_enabled(encoder, true);
-			omap_crtc->full_update = false;
 		}
-
-		dispc_mgr_setup(omap_crtc->channel, &omap_crtc->info);
-		dispc_mgr_set_timings(omap_crtc->channel,
-				&omap_crtc->timings);
-		set_enabled(&omap_crtc->base, true);
 	}
 
 	omap_crtc->full_update = false;
@@ -612,10 +637,30 @@ static void omap_crtc_post_apply(struct omap_drm_apply *apply)
 	/* nothing needed for post-apply */
 }
 
+void omap_crtc_flush(struct drm_crtc *crtc)
+{
+	struct omap_crtc *omap_crtc = to_omap_crtc(crtc);
+	int loops = 0;
+
+	while (!list_empty(&omap_crtc->pending_applies) ||
+		!list_empty(&omap_crtc->queued_applies) ||
+		omap_crtc->event || omap_crtc->old_fb) {
+
+		if (++loops > 10) {
+			dev_err(crtc->dev->dev,
+				"omap_crtc_flush() timeout\n");
+			break;
+		}
+
+		schedule_timeout_uninterruptible(msecs_to_jiffies(20));
+	}
+}
+
 static const char *channel_names[] = {
 		[OMAP_DSS_CHANNEL_LCD] = "lcd",
 		[OMAP_DSS_CHANNEL_DIGIT] = "tv",
 		[OMAP_DSS_CHANNEL_LCD2] = "lcd2",
+		[OMAP_DSS_CHANNEL_LCD3] = "lcd3",
 };
 
 void omap_crtc_pre_init(void)
@@ -623,6 +668,11 @@ void omap_crtc_pre_init(void)
 	dss_install_mgr_ops(&mgr_ops);
 }
 
+void omap_crtc_pre_uninit(void)
+{
+	dss_uninstall_mgr_ops();
+}
+
 /* initialize crtc */
 struct drm_crtc *omap_crtc_init(struct drm_device *dev,
 		struct drm_plane *plane, enum omap_channel channel, int id)
diff --git a/drivers/gpu/drm/omapdrm/omap_debugfs.c b/drivers/gpu/drm/omapdrm/omap_debugfs.c
index c27f59da7f2..d4c04d69fc4 100644
--- a/drivers/gpu/drm/omapdrm/omap_debugfs.c
+++ b/drivers/gpu/drm/omapdrm/omap_debugfs.c
@@ -48,7 +48,7 @@ static int mm_show(struct seq_file *m, void *arg)
 {
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
-	return drm_mm_dump_table(m, dev->mm_private);
+	return drm_mm_dump_table(m, &dev->vma_offset_manager->vm_addr_space_mm);
 }
 
 static int fb_show(struct seq_file *m, void *arg)
diff --git a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
index 701c4c10e08..f926b4caf44 100644
--- a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
+++ b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
@@ -969,12 +969,21 @@ static const struct dev_pm_ops omap_dmm_pm_ops = {
 };
 #endif
 
+#if defined(CONFIG_OF)
+static const struct of_device_id dmm_of_match[] = {
+	{ .compatible = "ti,omap4-dmm", },
+	{ .compatible = "ti,omap5-dmm", },
+	{},
+};
+#endif
+
 struct platform_driver omap_dmm_driver = {
 	.probe = omap_dmm_probe,
 	.remove = omap_dmm_remove,
 	.driver = {
 		.owner = THIS_MODULE,
 		.name = DMM_DRIVER_NAME,
+		.of_match_table = of_match_ptr(dmm_of_match),
 #ifdef CONFIG_PM
 		.pm = &omap_dmm_pm_ops,
 #endif
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.c b/drivers/gpu/drm/omapdrm/omap_drv.c
index e7fa3cd9674..002b9721e85 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.c
+++ b/drivers/gpu/drm/omapdrm/omap_drv.c
@@ -86,6 +86,47 @@ static bool channel_used(struct drm_device *dev, enum omap_channel channel)
 
 	return false;
 }
+static void omap_disconnect_dssdevs(void)
+{
+	struct omap_dss_device *dssdev = NULL;
+
+	for_each_dss_dev(dssdev)
+		dssdev->driver->disconnect(dssdev);
+}
+
+static int omap_connect_dssdevs(void)
+{
+	int r;
+	struct omap_dss_device *dssdev = NULL;
+	bool no_displays = true;
+
+	for_each_dss_dev(dssdev) {
+		r = dssdev->driver->connect(dssdev);
+		if (r == -EPROBE_DEFER) {
+			omap_dss_put_device(dssdev);
+			goto cleanup;
+		} else if (r) {
+			dev_warn(dssdev->dev, "could not connect display: %s\n",
+				dssdev->name);
+		} else {
+			no_displays = false;
+		}
+	}
+
+	if (no_displays)
+		return -EPROBE_DEFER;
+
+	return 0;
+
+cleanup:
+	/*
+	 * if we are deferring probe, we disconnect the devices we previously
+	 * connected
+	 */
+	omap_disconnect_dssdevs();
+
+	return r;
+}
 
 static int omap_modeset_init(struct drm_device *dev)
 {
@@ -95,9 +136,6 @@ static int omap_modeset_init(struct drm_device *dev)
 	int num_mgrs = dss_feat_get_num_mgrs();
 	int num_crtcs;
 	int i, id = 0;
-	int r;
-
-	omap_crtc_pre_init();
 
 	drm_mode_config_init(dev);
 
@@ -119,26 +157,8 @@ static int omap_modeset_init(struct drm_device *dev)
 		enum omap_channel channel;
 		struct omap_overlay_manager *mgr;
 
-		if (!dssdev->driver) {
-			dev_warn(dev->dev, "%s has no driver.. skipping it\n",
-					dssdev->name);
-			continue;
-		}
-
-		if (!(dssdev->driver->get_timings ||
-					dssdev->driver->read_edid)) {
-			dev_warn(dev->dev, "%s driver does not support "
-				"get_timings or read_edid.. skipping it!\n",
-				dssdev->name);
-			continue;
-		}
-
-		r = dssdev->driver->connect(dssdev);
-		if (r) {
-			dev_err(dev->dev, "could not connect display: %s\n",
-					dssdev->name);
+		if (!omapdss_device_is_connected(dssdev))
 			continue;
-		}
 
 		encoder = omap_encoder_init(dev, dssdev);
 
@@ -493,20 +513,26 @@ static int dev_load(struct drm_device *dev, unsigned long flags)
 static int dev_unload(struct drm_device *dev)
 {
 	struct omap_drm_private *priv = dev->dev_private;
+	int i;
 
 	DBG("unload: dev=%p", dev);
 
 	drm_kms_helper_poll_fini(dev);
-	drm_vblank_cleanup(dev);
-	omap_drm_irq_uninstall(dev);
 
 	omap_fbdev_free(dev);
+
+	/* flush crtcs so the fbs get released */
+	for (i = 0; i < priv->num_crtcs; i++)
+		omap_crtc_flush(priv->crtcs[i]);
+
 	omap_modeset_free(dev);
 	omap_gem_deinit(dev);
 
-	flush_workqueue(priv->wq);
 	destroy_workqueue(priv->wq);
 
+	drm_vblank_cleanup(dev);
+	omap_drm_irq_uninstall(dev);
+
 	kfree(dev->dev_private);
 	dev->dev_private = NULL;
 
@@ -562,9 +588,7 @@ static void dev_lastclose(struct drm_device *dev)
 		}
 	}
 
-	drm_modeset_lock_all(dev);
-	ret = drm_fb_helper_restore_fbdev_mode(priv->fbdev);
-	drm_modeset_unlock_all(dev);
+	ret = drm_fb_helper_restore_fbdev_mode_unlocked(priv->fbdev);
 	if (ret)
 		DBG("failed to restore crtc mode");
 }
@@ -655,9 +679,19 @@ static void pdev_shutdown(struct platform_device *device)
 
 static int pdev_probe(struct platform_device *device)
 {
+	int r;
+
 	if (omapdss_is_initialized() == false)
 		return -EPROBE_DEFER;
 
+	omap_crtc_pre_init();
+
+	r = omap_connect_dssdevs();
+	if (r) {
+		omap_crtc_pre_uninit();
+		return r;
+	}
+
 	DBG("%s", device->name);
 	return drm_platform_init(&omap_drm_driver, device);
 }
@@ -665,9 +699,12 @@ static int pdev_probe(struct platform_device *device)
 static int pdev_remove(struct platform_device *device)
 {
 	DBG("");
-	drm_platform_exit(&omap_drm_driver, device);
 
-	platform_driver_unregister(&omap_dmm_driver);
+	drm_put_dev(platform_get_drvdata(device));
+
+	omap_disconnect_dssdevs();
+	omap_crtc_pre_uninit();
+
 	return 0;
 }
 
@@ -694,18 +731,33 @@ static struct platform_driver pdev = {
 
 static int __init omap_drm_init(void)
 {
+	int r;
+
 	DBG("init");
-	if (platform_driver_register(&omap_dmm_driver)) {
-		/* we can continue on without DMM.. so not fatal */
-		dev_err(NULL, "DMM registration failed\n");
+
+	r = platform_driver_register(&omap_dmm_driver);
+	if (r) {
+		pr_err("DMM driver registration failed\n");
+		return r;
+	}
+
+	r = platform_driver_register(&pdev);
+	if (r) {
+		pr_err("omapdrm driver registration failed\n");
+		platform_driver_unregister(&omap_dmm_driver);
+		return r;
 	}
-	return platform_driver_register(&pdev);
+
+	return 0;
 }
 
 static void __exit omap_drm_fini(void)
 {
 	DBG("fini");
+
 	platform_driver_unregister(&pdev);
+
+	platform_driver_unregister(&omap_dmm_driver);
 }
 
 /* need late_initcall() so we load after dss_driver's are loaded */
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.h b/drivers/gpu/drm/omapdrm/omap_drv.h
index 07847693cf4..284b80fc3c5 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.h
+++ b/drivers/gpu/drm/omapdrm/omap_drv.h
@@ -141,10 +141,12 @@ int omap_gem_resume(struct device *dev);
 
 int omap_irq_enable_vblank(struct drm_device *dev, int crtc_id);
 void omap_irq_disable_vblank(struct drm_device *dev, int crtc_id);
-irqreturn_t omap_irq_handler(DRM_IRQ_ARGS);
+irqreturn_t omap_irq_handler(int irq, void *arg);
 void omap_irq_preinstall(struct drm_device *dev);
 int omap_irq_postinstall(struct drm_device *dev);
 void omap_irq_uninstall(struct drm_device *dev);
+void __omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq);
+void __omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq);
 void omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq);
 void omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq);
 int omap_drm_irq_uninstall(struct drm_device *dev);
@@ -158,8 +160,10 @@ enum omap_channel omap_crtc_channel(struct drm_crtc *crtc);
 int omap_crtc_apply(struct drm_crtc *crtc,
 		struct omap_drm_apply *apply);
 void omap_crtc_pre_init(void);
+void omap_crtc_pre_uninit(void);
 struct drm_crtc *omap_crtc_init(struct drm_device *dev,
 		struct drm_plane *plane, enum omap_channel channel, int id);
+void omap_crtc_flush(struct drm_crtc *crtc);
 
 struct drm_plane *omap_plane_init(struct drm_device *dev,
 		int plane_id, bool private_plane);
diff --git a/drivers/gpu/drm/omapdrm/omap_encoder.c b/drivers/gpu/drm/omapdrm/omap_encoder.c
index 6a12e899235..5290a88c681 100644
--- a/drivers/gpu/drm/omapdrm/omap_encoder.c
+++ b/drivers/gpu/drm/omapdrm/omap_encoder.c
@@ -51,6 +51,9 @@ struct omap_dss_device *omap_encoder_get_dssdev(struct drm_encoder *encoder)
 static void omap_encoder_destroy(struct drm_encoder *encoder)
 {
 	struct omap_encoder *omap_encoder = to_omap_encoder(encoder);
+
+	omap_encoder_set_enabled(encoder, false);
+
 	drm_encoder_cleanup(encoder);
 	kfree(omap_encoder);
 }
diff --git a/drivers/gpu/drm/omapdrm/omap_fb.c b/drivers/gpu/drm/omapdrm/omap_fb.c
index f2b8f0668c0..2a5cacdc344 100644
--- a/drivers/gpu/drm/omapdrm/omap_fb.c
+++ b/drivers/gpu/drm/omapdrm/omap_fb.c
@@ -123,12 +123,16 @@ static int omap_framebuffer_dirty(struct drm_framebuffer *fb,
 {
 	int i;
 
+	drm_modeset_lock_all(fb->dev);
+
 	for (i = 0; i < num_clips; i++) {
 		omap_framebuffer_flush(fb, clips[i].x1, clips[i].y1,
 					clips[i].x2 - clips[i].x1,
 					clips[i].y2 - clips[i].y1);
 	}
 
+	drm_modeset_unlock_all(fb->dev);
+
 	return 0;
 }
 
@@ -214,6 +218,20 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 		info->rotation_type = OMAP_DSS_ROT_TILER;
 		info->screen_width  = omap_gem_tiled_stride(plane->bo, orient);
 	} else {
+		switch (win->rotation & 0xf) {
+		case 0:
+		case BIT(DRM_ROTATE_0):
+			/* OK */
+			break;
+
+		default:
+			dev_warn(fb->dev->dev,
+				"rotation '%d' ignored for non-tiled fb\n",
+				win->rotation);
+			win->rotation = 0;
+			break;
+		}
+
 		info->paddr         = get_linear_addr(plane, format, 0, x, y);
 		info->rotation_type = OMAP_DSS_ROT_DMA;
 		info->screen_width  = plane->pitch;
@@ -302,13 +320,14 @@ struct drm_connector *omap_framebuffer_get_next_connector(
 	struct drm_connector *connector = from;
 
 	if (!from)
-		return list_first_entry(connector_list, typeof(*from), head);
+		return list_first_entry_or_null(connector_list, typeof(*from),
+						head);
 
 	list_for_each_entry_from(connector, connector_list, head) {
 		if (connector != from) {
 			struct drm_encoder *encoder = connector->encoder;
 			struct drm_crtc *crtc = encoder ? encoder->crtc : NULL;
-			if (crtc && crtc->fb == fb)
+			if (crtc && crtc->primary->fb == fb)
 				return connector;
 
 		}
@@ -327,6 +346,7 @@ void omap_framebuffer_flush(struct drm_framebuffer *fb,
 
 	VERB("flush: %d,%d %dx%d, fb=%p", x, y, w, h, fb);
 
+	/* FIXME: This is racy - no protection against modeset config changes. */
 	while ((connector = omap_framebuffer_get_next_connector(fb, connector))) {
 		/* only consider connectors that are part of a chain */
 		if (connector->encoder && connector->encoder->crtc) {
diff --git a/drivers/gpu/drm/omapdrm/omap_fbdev.c b/drivers/gpu/drm/omapdrm/omap_fbdev.c
index 002988d0902..1388ca7f87e 100644
--- a/drivers/gpu/drm/omapdrm/omap_fbdev.c
+++ b/drivers/gpu/drm/omapdrm/omap_fbdev.c
@@ -371,6 +371,9 @@ void omap_fbdev_free(struct drm_device *dev)
 
 	fbdev = to_omap_fbdev(priv->fbdev);
 
+	/* release the ref taken in omap_fbdev_create() */
+	omap_gem_put_paddr(fbdev->bo);
+
 	/* this will free the backing object */
 	if (fbdev->fb) {
 		drm_framebuffer_unregister_private(fbdev->fb);
diff --git a/drivers/gpu/drm/omapdrm/omap_gem.c b/drivers/gpu/drm/omapdrm/omap_gem.c
index 5aec3e81fe2..95dbce286a4 100644
--- a/drivers/gpu/drm/omapdrm/omap_gem.c
+++ b/drivers/gpu/drm/omapdrm/omap_gem.c
@@ -153,24 +153,24 @@ static struct {
 static void evict_entry(struct drm_gem_object *obj,
 		enum tiler_fmt fmt, struct usergart_entry *entry)
 {
-	if (obj->dev->dev_mapping) {
-		struct omap_gem_object *omap_obj = to_omap_bo(obj);
-		int n = usergart[fmt].height;
-		size_t size = PAGE_SIZE * n;
-		loff_t off = mmap_offset(obj) +
-				(entry->obj_pgoff << PAGE_SHIFT);
-		const int m = 1 + ((omap_obj->width << fmt) / PAGE_SIZE);
-		if (m > 1) {
-			int i;
-			/* if stride > than PAGE_SIZE then sparse mapping: */
-			for (i = n; i > 0; i--) {
-				unmap_mapping_range(obj->dev->dev_mapping,
-						off, PAGE_SIZE, 1);
-				off += PAGE_SIZE * m;
-			}
-		} else {
-			unmap_mapping_range(obj->dev->dev_mapping, off, size, 1);
+	struct omap_gem_object *omap_obj = to_omap_bo(obj);
+	int n = usergart[fmt].height;
+	size_t size = PAGE_SIZE * n;
+	loff_t off = mmap_offset(obj) +
+			(entry->obj_pgoff << PAGE_SHIFT);
+	const int m = 1 + ((omap_obj->width << fmt) / PAGE_SIZE);
+
+	if (m > 1) {
+		int i;
+		/* if stride > than PAGE_SIZE then sparse mapping: */
+		for (i = n; i > 0; i--) {
+			unmap_mapping_range(obj->dev->anon_inode->i_mapping,
+					    off, PAGE_SIZE, 1);
+			off += PAGE_SIZE * m;
 		}
+	} else {
+		unmap_mapping_range(obj->dev->anon_inode->i_mapping,
+				    off, size, 1);
 	}
 
 	entry->obj = NULL;
@@ -980,12 +980,9 @@ int omap_gem_resume(struct device *dev)
 #ifdef CONFIG_DEBUG_FS
 void omap_gem_describe(struct drm_gem_object *obj, struct seq_file *m)
 {
-	struct drm_device *dev = obj->dev;
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	uint64_t off;
 
-	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
-
 	off = drm_vma_node_start(&obj->vma_node);
 
 	seq_printf(m, "%08x: %2d (%2d) %08llx %08Zx (%2d) %p %4d",
@@ -1050,10 +1047,10 @@ static inline bool is_waiting(struct omap_gem_sync_waiter *waiter)
 {
 	struct omap_gem_object *omap_obj = waiter->omap_obj;
 	if ((waiter->op & OMAP_GEM_READ) &&
-			(omap_obj->sync->read_complete < waiter->read_target))
+			(omap_obj->sync->write_complete < waiter->write_target))
 		return true;
 	if ((waiter->op & OMAP_GEM_WRITE) &&
-			(omap_obj->sync->write_complete < waiter->write_target))
+			(omap_obj->sync->read_complete < waiter->read_target))
 		return true;
 	return false;
 }
@@ -1229,6 +1226,8 @@ int omap_gem_op_async(struct drm_gem_object *obj, enum omap_gem_op op,
 		}
 
 		spin_unlock(&sync_lock);
+
+		kfree(waiter);
 	}
 
 	/* no waiting.. */
diff --git a/drivers/gpu/drm/omapdrm/omap_irq.c b/drivers/gpu/drm/omapdrm/omap_irq.c
index cb858600185..f035d2bceae 100644
--- a/drivers/gpu/drm/omapdrm/omap_irq.c
+++ b/drivers/gpu/drm/omapdrm/omap_irq.c
@@ -45,12 +45,11 @@ static void omap_irq_update(struct drm_device *dev)
 	dispc_read_irqenable();        /* flush posted write */
 }
 
-void omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq)
+void __omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq)
 {
 	struct omap_drm_private *priv = dev->dev_private;
 	unsigned long flags;
 
-	dispc_runtime_get();
 	spin_lock_irqsave(&list_lock, flags);
 
 	if (!WARN_ON(irq->registered)) {
@@ -60,14 +59,21 @@ void omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq)
 	}
 
 	spin_unlock_irqrestore(&list_lock, flags);
+}
+
+void omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq)
+{
+	dispc_runtime_get();
+
+	__omap_irq_register(dev, irq);
+
 	dispc_runtime_put();
 }
 
-void omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq)
+void __omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq)
 {
 	unsigned long flags;
 
-	dispc_runtime_get();
 	spin_lock_irqsave(&list_lock, flags);
 
 	if (!WARN_ON(!irq->registered)) {
@@ -77,6 +83,14 @@ void omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq)
 	}
 
 	spin_unlock_irqrestore(&list_lock, flags);
+}
+
+void omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq)
+{
+	dispc_runtime_get();
+
+	__omap_irq_unregister(dev, irq);
+
 	dispc_runtime_put();
 }
 
@@ -173,7 +187,7 @@ void omap_irq_disable_vblank(struct drm_device *dev, int crtc_id)
 	dispc_runtime_put();
 }
 
-irqreturn_t omap_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t omap_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	struct omap_drm_private *priv = dev->dev_private;
@@ -308,7 +322,7 @@ int omap_drm_irq_uninstall(struct drm_device *dev)
 	if (dev->num_crtcs) {
 		spin_lock_irqsave(&dev->vbl_lock, irqflags);
 		for (i = 0; i < dev->num_crtcs; i++) {
-			DRM_WAKEUP(&dev->vblank[i].queue);
+			wake_up(&dev->vblank[i].queue);
 			dev->vblank[i].enabled = false;
 			dev->vblank[i].last =
 				dev->driver->get_vblank_counter(dev, i);
diff --git a/drivers/gpu/drm/omapdrm/omap_plane.c b/drivers/gpu/drm/omapdrm/omap_plane.c
index 046d5e660c0..3cf31ee59aa 100644
--- a/drivers/gpu/drm/omapdrm/omap_plane.c
+++ b/drivers/gpu/drm/omapdrm/omap_plane.c
@@ -225,6 +225,11 @@ int omap_plane_mode_set(struct drm_plane *plane,
 		omap_plane->apply_done_cb.arg = arg;
 	}
 
+	if (plane->fb)
+		drm_framebuffer_unreference(plane->fb);
+
+	drm_framebuffer_reference(fb);
+
 	plane->fb = fb;
 	plane->crtc = crtc;
 
@@ -241,10 +246,13 @@ static int omap_plane_update(struct drm_plane *plane,
 	struct omap_plane *omap_plane = to_omap_plane(plane);
 	omap_plane->enabled = true;
 
-	if (plane->fb)
-		drm_framebuffer_unreference(plane->fb);
-
-	drm_framebuffer_reference(fb);
+	/* omap_plane_mode_set() takes adjusted src */
+	switch (omap_plane->win.rotation & 0xf) {
+	case BIT(DRM_ROTATE_90):
+	case BIT(DRM_ROTATE_270):
+		swap(src_w, src_h);
+		break;
+	}
 
 	return omap_plane_mode_set(plane, crtc, fb,
 			crtc_x, crtc_y, crtc_w, crtc_h,
diff --git a/drivers/gpu/drm/panel/Kconfig b/drivers/gpu/drm/panel/Kconfig
new file mode 100644
index 00000000000..4ec874da566
--- /dev/null
+++ b/drivers/gpu/drm/panel/Kconfig
@@ -0,0 +1,33 @@
+config DRM_PANEL
+	bool
+	depends on DRM
+	help
+	  Panel registration and lookup framework.
+
+menu "Display Panels"
+	depends on DRM_PANEL
+
+config DRM_PANEL_SIMPLE
+	tristate "support for simple panels"
+	depends on OF
+	help
+	  DRM panel driver for dumb panels that need at most a regulator and
+	  a GPIO to be powered up. Optionally a backlight can be attached so
+	  that it can be automatically turned off when the panel goes into a
+	  low power state.
+
+config DRM_PANEL_LD9040
+	tristate "LD9040 RGB/SPI panel"
+	depends on DRM && DRM_PANEL
+	depends on OF
+	select SPI
+	select VIDEOMODE_HELPERS
+
+config DRM_PANEL_S6E8AA0
+	tristate "S6E8AA0 DSI video mode panel"
+	depends on DRM && DRM_PANEL
+	depends on OF
+	select DRM_MIPI_DSI
+	select VIDEOMODE_HELPERS
+
+endmenu
diff --git a/drivers/gpu/drm/panel/Makefile b/drivers/gpu/drm/panel/Makefile
new file mode 100644
index 00000000000..8b929212fad
--- /dev/null
+++ b/drivers/gpu/drm/panel/Makefile
@@ -0,0 +1,3 @@
+obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o
+obj-$(CONFIG_DRM_PANEL_LD9040) += panel-ld9040.o
+obj-$(CONFIG_DRM_PANEL_S6E8AA0) += panel-s6e8aa0.o
diff --git a/drivers/gpu/drm/panel/panel-ld9040.c b/drivers/gpu/drm/panel/panel-ld9040.c
new file mode 100644
index 00000000000..db1601fdbe2
--- /dev/null
+++ b/drivers/gpu/drm/panel/panel-ld9040.c
@@ -0,0 +1,379 @@
+/*
+ * ld9040 AMOLED LCD drm_panel driver.
+ *
+ * Copyright (c) 2014 Samsung Electronics Co., Ltd
+ * Derived from drivers/video/backlight/ld9040.c
+ *
+ * Andrzej Hajda <a.hajda@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+*/
+
+#include <drm/drmP.h>
+#include <drm/drm_panel.h>
+
+#include <linux/gpio/consumer.h>
+#include <linux/regulator/consumer.h>
+#include <linux/spi/spi.h>
+
+#include <video/mipi_display.h>
+#include <video/of_videomode.h>
+#include <video/videomode.h>
+
+/* Manufacturer Command Set */
+#define MCS_MANPWR		0xb0
+#define MCS_ELVSS_ON		0xb1
+#define MCS_USER_SETTING	0xf0
+#define MCS_DISPCTL		0xf2
+#define MCS_POWER_CTRL		0xf4
+#define MCS_GTCON		0xf7
+#define MCS_PANEL_CONDITION	0xf8
+#define MCS_GAMMA_SET1		0xf9
+#define MCS_GAMMA_CTRL		0xfb
+
+/* array of gamma tables for gamma value 2.2 */
+static u8 const ld9040_gammas[25][22] = {
+	{ 0xf9, 0x00, 0x13, 0xb2, 0xba, 0xd2, 0x00, 0x30, 0x00, 0xaf, 0xc0,
+	  0xb8, 0xcd, 0x00, 0x3d, 0x00, 0xa8, 0xb8, 0xb7, 0xcd, 0x00, 0x44 },
+	{ 0xf9, 0x00, 0x13, 0xb9, 0xb9, 0xd0, 0x00, 0x3c, 0x00, 0xaf, 0xbf,
+	  0xb6, 0xcb, 0x00, 0x4b, 0x00, 0xa8, 0xb9, 0xb5, 0xcc, 0x00, 0x52 },
+	{ 0xf9, 0x00, 0x13, 0xba, 0xb9, 0xcd, 0x00, 0x41, 0x00, 0xb0, 0xbe,
+	  0xb5, 0xc9, 0x00, 0x51, 0x00, 0xa9, 0xb9, 0xb5, 0xca, 0x00, 0x57 },
+	{ 0xf9, 0x00, 0x13, 0xb9, 0xb8, 0xcd, 0x00, 0x46, 0x00, 0xb1, 0xbc,
+	  0xb5, 0xc8, 0x00, 0x56, 0x00, 0xaa, 0xb8, 0xb4, 0xc9, 0x00, 0x5d },
+	{ 0xf9, 0x00, 0x13, 0xba, 0xb8, 0xcb, 0x00, 0x4b, 0x00, 0xb3, 0xbc,
+	  0xb4, 0xc7, 0x00, 0x5c, 0x00, 0xac, 0xb8, 0xb4, 0xc8, 0x00, 0x62 },
+	{ 0xf9, 0x00, 0x13, 0xbb, 0xb7, 0xca, 0x00, 0x4f, 0x00, 0xb4, 0xbb,
+	  0xb3, 0xc7, 0x00, 0x60, 0x00, 0xad, 0xb8, 0xb4, 0xc7, 0x00, 0x67 },
+	{ 0xf9, 0x00, 0x47, 0xba, 0xb6, 0xca, 0x00, 0x53, 0x00, 0xb5, 0xbb,
+	  0xb3, 0xc6, 0x00, 0x65, 0x00, 0xae, 0xb8, 0xb3, 0xc7, 0x00, 0x6c },
+	{ 0xf9, 0x00, 0x71, 0xbb, 0xb5, 0xc8, 0x00, 0x57, 0x00, 0xb5, 0xbb,
+	  0xb0, 0xc5, 0x00, 0x6a, 0x00, 0xae, 0xb9, 0xb1, 0xc6, 0x00, 0x70 },
+	{ 0xf9, 0x00, 0x7b, 0xbb, 0xb4, 0xc8, 0x00, 0x5b, 0x00, 0xb5, 0xba,
+	  0xb1, 0xc4, 0x00, 0x6e, 0x00, 0xae, 0xb9, 0xb0, 0xc5, 0x00, 0x75 },
+	{ 0xf9, 0x00, 0x82, 0xba, 0xb4, 0xc7, 0x00, 0x5f, 0x00, 0xb5, 0xba,
+	  0xb0, 0xc3, 0x00, 0x72, 0x00, 0xae, 0xb8, 0xb0, 0xc3, 0x00, 0x7a },
+	{ 0xf9, 0x00, 0x89, 0xba, 0xb3, 0xc8, 0x00, 0x62, 0x00, 0xb6, 0xba,
+	  0xaf, 0xc3, 0x00, 0x76, 0x00, 0xaf, 0xb7, 0xae, 0xc4, 0x00, 0x7e },
+	{ 0xf9, 0x00, 0x8b, 0xb9, 0xb3, 0xc7, 0x00, 0x65, 0x00, 0xb7, 0xb8,
+	  0xaf, 0xc3, 0x00, 0x7a, 0x00, 0x80, 0xb6, 0xae, 0xc4, 0x00, 0x81 },
+	{ 0xf9, 0x00, 0x93, 0xba, 0xb3, 0xc5, 0x00, 0x69, 0x00, 0xb8, 0xb9,
+	  0xae, 0xc1, 0x00, 0x7f, 0x00, 0xb0, 0xb6, 0xae, 0xc3, 0x00, 0x85 },
+	{ 0xf9, 0x00, 0x97, 0xba, 0xb2, 0xc5, 0x00, 0x6c, 0x00, 0xb8, 0xb8,
+	  0xae, 0xc1, 0x00, 0x82, 0x00, 0xb0, 0xb6, 0xae, 0xc2, 0x00, 0x89 },
+	{ 0xf9, 0x00, 0x9a, 0xba, 0xb1, 0xc4, 0x00, 0x6f, 0x00, 0xb8, 0xb8,
+	  0xad, 0xc0, 0x00, 0x86, 0x00, 0xb0, 0xb7, 0xad, 0xc0, 0x00, 0x8d },
+	{ 0xf9, 0x00, 0x9c, 0xb9, 0xb0, 0xc4, 0x00, 0x72, 0x00, 0xb8, 0xb8,
+	  0xac, 0xbf, 0x00, 0x8a, 0x00, 0xb0, 0xb6, 0xac, 0xc0, 0x00, 0x91 },
+	{ 0xf9, 0x00, 0x9e, 0xba, 0xb0, 0xc2, 0x00, 0x75, 0x00, 0xb9, 0xb8,
+	  0xab, 0xbe, 0x00, 0x8e, 0x00, 0xb0, 0xb6, 0xac, 0xbf, 0x00, 0x94 },
+	{ 0xf9, 0x00, 0xa0, 0xb9, 0xaf, 0xc3, 0x00, 0x77, 0x00, 0xb9, 0xb7,
+	  0xab, 0xbe, 0x00, 0x90, 0x00, 0xb0, 0xb6, 0xab, 0xbf, 0x00, 0x97 },
+	{ 0xf9, 0x00, 0xa2, 0xb9, 0xaf, 0xc2, 0x00, 0x7a, 0x00, 0xb9, 0xb7,
+	  0xaa, 0xbd, 0x00, 0x94, 0x00, 0xb0, 0xb5, 0xab, 0xbf, 0x00, 0x9a },
+	{ 0xf9, 0x00, 0xa4, 0xb9, 0xaf, 0xc1, 0x00, 0x7d, 0x00, 0xb9, 0xb6,
+	  0xaa, 0xbb, 0x00, 0x97, 0x00, 0xb1, 0xb5, 0xaa, 0xbf, 0x00, 0x9d },
+	{ 0xf9, 0x00, 0xa4, 0xb8, 0xb0, 0xbf, 0x00, 0x80, 0x00, 0xb8, 0xb6,
+	  0xaa, 0xbc, 0x00, 0x9a, 0x00, 0xb0, 0xb5, 0xab, 0xbd, 0x00, 0xa0 },
+	{ 0xf9, 0x00, 0xa8, 0xb8, 0xae, 0xbe, 0x00, 0x84, 0x00, 0xb9, 0xb7,
+	  0xa8, 0xbc, 0x00, 0x9d, 0x00, 0xb2, 0xb5, 0xaa, 0xbc, 0x00, 0xa4 },
+	{ 0xf9, 0x00, 0xa9, 0xb6, 0xad, 0xbf, 0x00, 0x86, 0x00, 0xb8, 0xb5,
+	  0xa8, 0xbc, 0x00, 0xa0, 0x00, 0xb3, 0xb3, 0xa9, 0xbc, 0x00, 0xa7 },
+	{ 0xf9, 0x00, 0xa9, 0xb7, 0xae, 0xbd, 0x00, 0x89, 0x00, 0xb7, 0xb6,
+	  0xa8, 0xba, 0x00, 0xa4, 0x00, 0xb1, 0xb4, 0xaa, 0xbb, 0x00, 0xaa },
+	{ 0xf9, 0x00, 0xa7, 0xb4, 0xae, 0xbf, 0x00, 0x91, 0x00, 0xb2, 0xb4,
+	  0xaa, 0xbb, 0x00, 0xac, 0x00, 0xb3, 0xb1, 0xaa, 0xbc, 0x00, 0xb3 },
+};
+
+struct ld9040 {
+	struct device *dev;
+	struct drm_panel panel;
+
+	struct regulator_bulk_data supplies[2];
+	struct gpio_desc *reset_gpio;
+	u32 power_on_delay;
+	u32 reset_delay;
+	struct videomode vm;
+	u32 width_mm;
+	u32 height_mm;
+
+	int brightness;
+
+	/* This field is tested by functions directly accessing bus before
+	 * transfer, transfer is skipped if it is set. In case of transfer
+	 * failure or unexpected response the field is set to error value.
+	 * Such construct allows to eliminate many checks in higher level
+	 * functions.
+	 */
+	int error;
+};
+
+#define panel_to_ld9040(p) container_of(p, struct ld9040, panel)
+
+static int ld9040_clear_error(struct ld9040 *ctx)
+{
+	int ret = ctx->error;
+
+	ctx->error = 0;
+	return ret;
+}
+
+static int ld9040_spi_write_word(struct ld9040 *ctx, u16 data)
+{
+	struct spi_device *spi = to_spi_device(ctx->dev);
+	struct spi_transfer xfer = {
+		.len		= 2,
+		.tx_buf		= &data,
+	};
+	struct spi_message msg;
+
+	spi_message_init(&msg);
+	spi_message_add_tail(&xfer, &msg);
+
+	return spi_sync(spi, &msg);
+}
+
+static void ld9040_dcs_write(struct ld9040 *ctx, const u8 *data, size_t len)
+{
+	int ret = 0;
+
+	if (ctx->error < 0 || len == 0)
+		return;
+
+	dev_dbg(ctx->dev, "writing dcs seq: %*ph\n", len, data);
+	ret = ld9040_spi_write_word(ctx, *data);
+
+	while (!ret && --len) {
+		++data;
+		ret = ld9040_spi_write_word(ctx, *data | 0x100);
+	}
+
+	if (ret) {
+		dev_err(ctx->dev, "error %d writing dcs seq: %*ph\n", ret, len,
+			data);
+		ctx->error = ret;
+	}
+
+	usleep_range(300, 310);
+}
+
+#define ld9040_dcs_write_seq_static(ctx, seq...) \
+({\
+	static const u8 d[] = { seq };\
+	ld9040_dcs_write(ctx, d, ARRAY_SIZE(d));\
+})
+
+static void ld9040_brightness_set(struct ld9040 *ctx)
+{
+	ld9040_dcs_write(ctx, ld9040_gammas[ctx->brightness],
+			 ARRAY_SIZE(ld9040_gammas[ctx->brightness]));
+
+	ld9040_dcs_write_seq_static(ctx, MCS_GAMMA_CTRL, 0x02, 0x5a);
+}
+
+static void ld9040_init(struct ld9040 *ctx)
+{
+	ld9040_dcs_write_seq_static(ctx, MCS_USER_SETTING, 0x5a, 0x5a);
+	ld9040_dcs_write_seq_static(ctx, MCS_PANEL_CONDITION,
+		0x05, 0x65, 0x96, 0x71, 0x7d, 0x19, 0x3b, 0x0d,
+		0x19, 0x7e, 0x0d, 0xe2, 0x00, 0x00, 0x7e, 0x7d,
+		0x07, 0x07, 0x20, 0x20, 0x20, 0x02, 0x02);
+	ld9040_dcs_write_seq_static(ctx, MCS_DISPCTL,
+		0x02, 0x08, 0x08, 0x10, 0x10);
+	ld9040_dcs_write_seq_static(ctx, MCS_MANPWR, 0x04);
+	ld9040_dcs_write_seq_static(ctx, MCS_POWER_CTRL,
+		0x0a, 0x87, 0x25, 0x6a, 0x44, 0x02, 0x88);
+	ld9040_dcs_write_seq_static(ctx, MCS_ELVSS_ON, 0x0d, 0x00, 0x16);
+	ld9040_dcs_write_seq_static(ctx, MCS_GTCON, 0x09, 0x00, 0x00);
+	ld9040_brightness_set(ctx);
+	ld9040_dcs_write_seq_static(ctx, MIPI_DCS_EXIT_SLEEP_MODE);
+	ld9040_dcs_write_seq_static(ctx, MIPI_DCS_SET_DISPLAY_ON);
+}
+
+static int ld9040_power_on(struct ld9040 *ctx)
+{
+	int ret;
+
+	ret = regulator_bulk_enable(ARRAY_SIZE(ctx->supplies), ctx->supplies);
+	if (ret < 0)
+		return ret;
+
+	msleep(ctx->power_on_delay);
+	gpiod_set_value(ctx->reset_gpio, 0);
+	msleep(ctx->reset_delay);
+	gpiod_set_value(ctx->reset_gpio, 1);
+	msleep(ctx->reset_delay);
+
+	return 0;
+}
+
+static int ld9040_power_off(struct ld9040 *ctx)
+{
+	return regulator_bulk_disable(ARRAY_SIZE(ctx->supplies), ctx->supplies);
+}
+
+static int ld9040_disable(struct drm_panel *panel)
+{
+	struct ld9040 *ctx = panel_to_ld9040(panel);
+
+	msleep(120);
+	ld9040_dcs_write_seq_static(ctx, MIPI_DCS_SET_DISPLAY_OFF);
+	ld9040_dcs_write_seq_static(ctx, MIPI_DCS_ENTER_SLEEP_MODE);
+	msleep(40);
+
+	ld9040_clear_error(ctx);
+
+	return ld9040_power_off(ctx);
+}
+
+static int ld9040_enable(struct drm_panel *panel)
+{
+	struct ld9040 *ctx = panel_to_ld9040(panel);
+	int ret;
+
+	ret = ld9040_power_on(ctx);
+	if (ret < 0)
+		return ret;
+
+	ld9040_init(ctx);
+
+	ret = ld9040_clear_error(ctx);
+
+	if (ret < 0)
+		ld9040_disable(panel);
+
+	return ret;
+}
+
+static int ld9040_get_modes(struct drm_panel *panel)
+{
+	struct drm_connector *connector = panel->connector;
+	struct ld9040 *ctx = panel_to_ld9040(panel);
+	struct drm_display_mode *mode;
+
+	mode = drm_mode_create(connector->dev);
+	if (!mode) {
+		DRM_ERROR("failed to create a new display mode\n");
+		return 0;
+	}
+
+	drm_display_mode_from_videomode(&ctx->vm, mode);
+	mode->width_mm = ctx->width_mm;
+	mode->height_mm = ctx->height_mm;
+	connector->display_info.width_mm = mode->width_mm;
+	connector->display_info.height_mm = mode->height_mm;
+
+	mode->type = DRM_MODE_TYPE_DRIVER | DRM_MODE_TYPE_PREFERRED;
+	drm_mode_probed_add(connector, mode);
+
+	return 1;
+}
+
+static const struct drm_panel_funcs ld9040_drm_funcs = {
+	.disable = ld9040_disable,
+	.enable = ld9040_enable,
+	.get_modes = ld9040_get_modes,
+};
+
+static int ld9040_parse_dt(struct ld9040 *ctx)
+{
+	struct device *dev = ctx->dev;
+	struct device_node *np = dev->of_node;
+	int ret;
+
+	ret = of_get_videomode(np, &ctx->vm, 0);
+	if (ret < 0)
+		return ret;
+
+	of_property_read_u32(np, "power-on-delay", &ctx->power_on_delay);
+	of_property_read_u32(np, "reset-delay", &ctx->reset_delay);
+	of_property_read_u32(np, "panel-width-mm", &ctx->width_mm);
+	of_property_read_u32(np, "panel-height-mm", &ctx->height_mm);
+
+	return 0;
+}
+
+static int ld9040_probe(struct spi_device *spi)
+{
+	struct device *dev = &spi->dev;
+	struct ld9040 *ctx;
+	int ret;
+
+	ctx = devm_kzalloc(dev, sizeof(struct ld9040), GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+
+	spi_set_drvdata(spi, ctx);
+
+	ctx->dev = dev;
+	ctx->brightness = ARRAY_SIZE(ld9040_gammas) - 1;
+
+	ret = ld9040_parse_dt(ctx);
+	if (ret < 0)
+		return ret;
+
+	ctx->supplies[0].supply = "vdd3";
+	ctx->supplies[1].supply = "vci";
+	ret = devm_regulator_bulk_get(dev, ARRAY_SIZE(ctx->supplies),
+				      ctx->supplies);
+	if (ret < 0)
+		return ret;
+
+	ctx->reset_gpio = devm_gpiod_get(dev, "reset");
+	if (IS_ERR(ctx->reset_gpio)) {
+		dev_err(dev, "cannot get reset-gpios %ld\n",
+			PTR_ERR(ctx->reset_gpio));
+		return PTR_ERR(ctx->reset_gpio);
+	}
+	ret = gpiod_direction_output(ctx->reset_gpio, 1);
+	if (ret < 0) {
+		dev_err(dev, "cannot configure reset-gpios %d\n", ret);
+		return ret;
+	}
+
+	spi->bits_per_word = 9;
+	ret = spi_setup(spi);
+	if (ret < 0) {
+		dev_err(dev, "spi setup failed.\n");
+		return ret;
+	}
+
+	drm_panel_init(&ctx->panel);
+	ctx->panel.dev = dev;
+	ctx->panel.funcs = &ld9040_drm_funcs;
+
+	return drm_panel_add(&ctx->panel);
+}
+
+static int ld9040_remove(struct spi_device *spi)
+{
+	struct ld9040 *ctx = spi_get_drvdata(spi);
+
+	ld9040_power_off(ctx);
+	drm_panel_remove(&ctx->panel);
+
+	return 0;
+}
+
+static struct of_device_id ld9040_of_match[] = {
+	{ .compatible = "samsung,ld9040" },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, ld9040_of_match);
+
+static struct spi_driver ld9040_driver = {
+	.probe		= ld9040_probe,
+	.remove		= ld9040_remove,
+	.driver = {
+		.name	= "ld9040",
+		.owner	= THIS_MODULE,
+		.of_match_table = ld9040_of_match,
+	},
+};
+module_spi_driver(ld9040_driver);
+
+MODULE_AUTHOR("Andrzej Hajda <a.hajda@samsung.com>");
+MODULE_DESCRIPTION("ld9040 LCD Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/panel/panel-s6e8aa0.c b/drivers/gpu/drm/panel/panel-s6e8aa0.c
new file mode 100644
index 00000000000..06e57a26db7
--- /dev/null
+++ b/drivers/gpu/drm/panel/panel-s6e8aa0.c
@@ -0,0 +1,1070 @@
+/*
+ * MIPI-DSI based s6e8aa0 AMOLED LCD 5.3 inch panel driver.
+ *
+ * Copyright (c) 2013 Samsung Electronics Co., Ltd
+ *
+ * Inki Dae, <inki.dae@samsung.com>
+ * Donghwa Lee, <dh09.lee@samsung.com>
+ * Joongmock Shin <jmock.shin@samsung.com>
+ * Eunchul Kim <chulspro.kim@samsung.com>
+ * Tomasz Figa <t.figa@samsung.com>
+ * Andrzej Hajda <a.hajda@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+*/
+
+#include <drm/drmP.h>
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_panel.h>
+
+#include <linux/gpio/consumer.h>
+#include <linux/regulator/consumer.h>
+
+#include <video/mipi_display.h>
+#include <video/of_videomode.h>
+#include <video/videomode.h>
+
+#define LDI_MTP_LENGTH			24
+#define GAMMA_LEVEL_NUM			25
+#define GAMMA_TABLE_LEN			26
+
+#define PANELCTL_SS_MASK		(1 << 5)
+#define PANELCTL_SS_1_800		(0 << 5)
+#define PANELCTL_SS_800_1		(1 << 5)
+#define PANELCTL_GTCON_MASK		(7 << 2)
+#define PANELCTL_GTCON_110		(6 << 2)
+#define PANELCTL_GTCON_111		(7 << 2)
+
+#define PANELCTL_CLK1_CON_MASK		(7 << 3)
+#define PANELCTL_CLK1_000		(0 << 3)
+#define PANELCTL_CLK1_001		(1 << 3)
+#define PANELCTL_CLK2_CON_MASK		(7 << 0)
+#define PANELCTL_CLK2_000		(0 << 0)
+#define PANELCTL_CLK2_001		(1 << 0)
+
+#define PANELCTL_INT1_CON_MASK		(7 << 3)
+#define PANELCTL_INT1_000		(0 << 3)
+#define PANELCTL_INT1_001		(1 << 3)
+#define PANELCTL_INT2_CON_MASK		(7 << 0)
+#define PANELCTL_INT2_000		(0 << 0)
+#define PANELCTL_INT2_001		(1 << 0)
+
+#define PANELCTL_BICTL_CON_MASK		(7 << 3)
+#define PANELCTL_BICTL_000		(0 << 3)
+#define PANELCTL_BICTL_001		(1 << 3)
+#define PANELCTL_BICTLB_CON_MASK	(7 << 0)
+#define PANELCTL_BICTLB_000		(0 << 0)
+#define PANELCTL_BICTLB_001		(1 << 0)
+
+#define PANELCTL_EM_CLK1_CON_MASK	(7 << 3)
+#define PANELCTL_EM_CLK1_110		(6 << 3)
+#define PANELCTL_EM_CLK1_111		(7 << 3)
+#define PANELCTL_EM_CLK1B_CON_MASK	(7 << 0)
+#define PANELCTL_EM_CLK1B_110		(6 << 0)
+#define PANELCTL_EM_CLK1B_111		(7 << 0)
+
+#define PANELCTL_EM_CLK2_CON_MASK	(7 << 3)
+#define PANELCTL_EM_CLK2_110		(6 << 3)
+#define PANELCTL_EM_CLK2_111		(7 << 3)
+#define PANELCTL_EM_CLK2B_CON_MASK	(7 << 0)
+#define PANELCTL_EM_CLK2B_110		(6 << 0)
+#define PANELCTL_EM_CLK2B_111		(7 << 0)
+
+#define PANELCTL_EM_INT1_CON_MASK	(7 << 3)
+#define PANELCTL_EM_INT1_000		(0 << 3)
+#define PANELCTL_EM_INT1_001		(1 << 3)
+#define PANELCTL_EM_INT2_CON_MASK	(7 << 0)
+#define PANELCTL_EM_INT2_000		(0 << 0)
+#define PANELCTL_EM_INT2_001		(1 << 0)
+
+#define AID_DISABLE			(0x4)
+#define AID_1				(0x5)
+#define AID_2				(0x6)
+#define AID_3				(0x7)
+
+typedef u8 s6e8aa0_gamma_table[GAMMA_TABLE_LEN];
+
+struct s6e8aa0_variant {
+	u8 version;
+	const s6e8aa0_gamma_table *gamma_tables;
+};
+
+struct s6e8aa0 {
+	struct device *dev;
+	struct drm_panel panel;
+
+	struct regulator_bulk_data supplies[2];
+	struct gpio_desc *reset_gpio;
+	u32 power_on_delay;
+	u32 reset_delay;
+	u32 init_delay;
+	bool flip_horizontal;
+	bool flip_vertical;
+	struct videomode vm;
+	u32 width_mm;
+	u32 height_mm;
+
+	u8 version;
+	u8 id;
+	const struct s6e8aa0_variant *variant;
+	int brightness;
+
+	/* This field is tested by functions directly accessing DSI bus before
+	 * transfer, transfer is skipped if it is set. In case of transfer
+	 * failure or unexpected response the field is set to error value.
+	 * Such construct allows to eliminate many checks in higher level
+	 * functions.
+	 */
+	int error;
+};
+
+#define panel_to_s6e8aa0(p) container_of(p, struct s6e8aa0, panel)
+
+static int s6e8aa0_clear_error(struct s6e8aa0 *ctx)
+{
+	int ret = ctx->error;
+
+	ctx->error = 0;
+	return ret;
+}
+
+static void s6e8aa0_dcs_write(struct s6e8aa0 *ctx, const void *data, size_t len)
+{
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(ctx->dev);
+	int ret;
+
+	if (ctx->error < 0)
+		return;
+
+	ret = mipi_dsi_dcs_write(dsi, dsi->channel, data, len);
+	if (ret < 0) {
+		dev_err(ctx->dev, "error %d writing dcs seq: %*ph\n", ret, len,
+			data);
+		ctx->error = ret;
+	}
+}
+
+static int s6e8aa0_dcs_read(struct s6e8aa0 *ctx, u8 cmd, void *data, size_t len)
+{
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(ctx->dev);
+	int ret;
+
+	if (ctx->error < 0)
+		return ctx->error;
+
+	ret = mipi_dsi_dcs_read(dsi, dsi->channel, cmd, data, len);
+	if (ret < 0) {
+		dev_err(ctx->dev, "error %d reading dcs seq(%#x)\n", ret, cmd);
+		ctx->error = ret;
+	}
+
+	return ret;
+}
+
+#define s6e8aa0_dcs_write_seq(ctx, seq...) \
+({\
+	const u8 d[] = { seq };\
+	BUILD_BUG_ON_MSG(ARRAY_SIZE(d) > 64, "DCS sequence too big for stack");\
+	s6e8aa0_dcs_write(ctx, d, ARRAY_SIZE(d));\
+})
+
+#define s6e8aa0_dcs_write_seq_static(ctx, seq...) \
+({\
+	static const u8 d[] = { seq };\
+	s6e8aa0_dcs_write(ctx, d, ARRAY_SIZE(d));\
+})
+
+static void s6e8aa0_apply_level_1_key(struct s6e8aa0 *ctx)
+{
+	s6e8aa0_dcs_write_seq_static(ctx, 0xf0, 0x5a, 0x5a);
+}
+
+static void s6e8aa0_panel_cond_set_v142(struct s6e8aa0 *ctx)
+{
+	static const u8 aids[] = {
+		0x04, 0x04, 0x04, 0x04, 0x04, 0x60, 0x80, 0xA0
+	};
+	u8 aid = aids[ctx->id >> 5];
+	u8 cfg = 0x3d;
+	u8 clk_con = 0xc8;
+	u8 int_con = 0x08;
+	u8 bictl_con = 0x48;
+	u8 em_clk1_con = 0xff;
+	u8 em_clk2_con = 0xff;
+	u8 em_int_con = 0xc8;
+
+	if (ctx->flip_vertical) {
+		/* GTCON */
+		cfg &= ~(PANELCTL_GTCON_MASK);
+		cfg |= (PANELCTL_GTCON_110);
+	}
+
+	if (ctx->flip_horizontal) {
+		/* SS */
+		cfg &= ~(PANELCTL_SS_MASK);
+		cfg |= (PANELCTL_SS_1_800);
+	}
+
+	if (ctx->flip_horizontal || ctx->flip_vertical) {
+		/* CLK1,2_CON */
+		clk_con &= ~(PANELCTL_CLK1_CON_MASK |
+			PANELCTL_CLK2_CON_MASK);
+		clk_con |= (PANELCTL_CLK1_000 | PANELCTL_CLK2_001);
+
+		/* INT1,2_CON */
+		int_con &= ~(PANELCTL_INT1_CON_MASK |
+			PANELCTL_INT2_CON_MASK);
+		int_con |= (PANELCTL_INT1_000 | PANELCTL_INT2_001);
+
+		/* BICTL,B_CON */
+		bictl_con &= ~(PANELCTL_BICTL_CON_MASK |
+			PANELCTL_BICTLB_CON_MASK);
+		bictl_con |= (PANELCTL_BICTL_000 |
+			PANELCTL_BICTLB_001);
+
+		/* EM_CLK1,1B_CON */
+		em_clk1_con &= ~(PANELCTL_EM_CLK1_CON_MASK |
+			PANELCTL_EM_CLK1B_CON_MASK);
+		em_clk1_con |= (PANELCTL_EM_CLK1_110 |
+			PANELCTL_EM_CLK1B_110);
+
+		/* EM_CLK2,2B_CON */
+		em_clk2_con &= ~(PANELCTL_EM_CLK2_CON_MASK |
+			PANELCTL_EM_CLK2B_CON_MASK);
+		em_clk2_con |= (PANELCTL_EM_CLK2_110 |
+			PANELCTL_EM_CLK2B_110);
+
+		/* EM_INT1,2_CON */
+		em_int_con &= ~(PANELCTL_EM_INT1_CON_MASK |
+			PANELCTL_EM_INT2_CON_MASK);
+		em_int_con |= (PANELCTL_EM_INT1_000 |
+			PANELCTL_EM_INT2_001);
+	}
+
+	s6e8aa0_dcs_write_seq(ctx,
+		0xf8, cfg, 0x35, 0x00, 0x00, 0x00, 0x93, 0x00,
+		0x3c, 0x78, 0x08, 0x27, 0x7d, 0x3f, 0x00, 0x00,
+		0x00, 0x20, aid, 0x08, 0x6e, 0x00, 0x00, 0x00,
+		0x02, 0x07, 0x07, 0x23, 0x23, 0xc0, clk_con, int_con,
+		bictl_con, 0xc1, 0x00, 0xc1, em_clk1_con, em_clk2_con,
+		em_int_con);
+}
+
+static void s6e8aa0_panel_cond_set(struct s6e8aa0 *ctx)
+{
+	if (ctx->version < 142)
+		s6e8aa0_dcs_write_seq_static(ctx,
+			0xf8, 0x19, 0x35, 0x00, 0x00, 0x00, 0x94, 0x00,
+			0x3c, 0x78, 0x10, 0x27, 0x08, 0x6e, 0x00, 0x00,
+			0x00, 0x00, 0x04, 0x08, 0x6e, 0x00, 0x00, 0x00,
+			0x00, 0x07, 0x07, 0x23, 0x6e, 0xc0, 0xc1, 0x01,
+			0x81, 0xc1, 0x00, 0xc3, 0xf6, 0xf6, 0xc1
+		);
+	else
+		s6e8aa0_panel_cond_set_v142(ctx);
+}
+
+static void s6e8aa0_display_condition_set(struct s6e8aa0 *ctx)
+{
+	s6e8aa0_dcs_write_seq_static(ctx, 0xf2, 0x80, 0x03, 0x0d);
+}
+
+static void s6e8aa0_etc_source_control(struct s6e8aa0 *ctx)
+{
+	s6e8aa0_dcs_write_seq_static(ctx, 0xf6, 0x00, 0x02, 0x00);
+}
+
+static void s6e8aa0_etc_pentile_control(struct s6e8aa0 *ctx)
+{
+	static const u8 pent32[] = {
+		0xb6, 0x0c, 0x02, 0x03, 0x32, 0xc0, 0x44, 0x44, 0xc0, 0x00
+	};
+
+	static const u8 pent142[] = {
+		0xb6, 0x0c, 0x02, 0x03, 0x32, 0xff, 0x44, 0x44, 0xc0, 0x00
+	};
+
+	if (ctx->version < 142)
+		s6e8aa0_dcs_write(ctx, pent32, ARRAY_SIZE(pent32));
+	else
+		s6e8aa0_dcs_write(ctx, pent142, ARRAY_SIZE(pent142));
+}
+
+static void s6e8aa0_etc_power_control(struct s6e8aa0 *ctx)
+{
+	static const u8 pwr142[] = {
+		0xf4, 0xcf, 0x0a, 0x12, 0x10, 0x1e, 0x33, 0x02
+	};
+
+	static const u8 pwr32[] = {
+		0xf4, 0xcf, 0x0a, 0x15, 0x10, 0x19, 0x33, 0x02
+	};
+
+	if (ctx->version < 142)
+		s6e8aa0_dcs_write(ctx, pwr32, ARRAY_SIZE(pwr32));
+	else
+		s6e8aa0_dcs_write(ctx, pwr142, ARRAY_SIZE(pwr142));
+}
+
+static void s6e8aa0_etc_elvss_control(struct s6e8aa0 *ctx)
+{
+	u8 id = ctx->id ? 0 : 0x95;
+
+	s6e8aa0_dcs_write_seq(ctx, 0xb1, 0x04, id);
+}
+
+static void s6e8aa0_elvss_nvm_set_v142(struct s6e8aa0 *ctx)
+{
+	u8 br;
+
+	switch (ctx->brightness) {
+	case 0 ... 6: /* 30cd ~ 100cd */
+		br = 0xdf;
+		break;
+	case 7 ... 11: /* 120cd ~ 150cd */
+		br = 0xdd;
+		break;
+	case 12 ... 15: /* 180cd ~ 210cd */
+	default:
+		br = 0xd9;
+		break;
+	case 16 ... 24: /* 240cd ~ 300cd */
+		br = 0xd0;
+		break;
+	}
+
+	s6e8aa0_dcs_write_seq(ctx, 0xd9, 0x14, 0x40, 0x0c, 0xcb, 0xce, 0x6e,
+		0xc4, 0x0f, 0x40, 0x41, br, 0x00, 0x60, 0x19);
+}
+
+static void s6e8aa0_elvss_nvm_set(struct s6e8aa0 *ctx)
+{
+	if (ctx->version < 142)
+		s6e8aa0_dcs_write_seq_static(ctx,
+			0xd9, 0x14, 0x40, 0x0c, 0xcb, 0xce, 0x6e, 0xc4, 0x07,
+			0x40, 0x41, 0xc1, 0x00, 0x60, 0x19);
+	else
+		s6e8aa0_elvss_nvm_set_v142(ctx);
+};
+
+static void s6e8aa0_apply_level_2_key(struct s6e8aa0 *ctx)
+{
+	s6e8aa0_dcs_write_seq_static(ctx, 0xfc, 0x5a, 0x5a);
+}
+
+static const s6e8aa0_gamma_table s6e8aa0_gamma_tables_v142[GAMMA_LEVEL_NUM] = {
+	{
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0x62, 0x55, 0x55,
+		0xaf, 0xb1, 0xb1, 0xbd, 0xce, 0xb7, 0x9a, 0xb1,
+		0x90, 0xb2, 0xc4, 0xae, 0x00, 0x60, 0x00, 0x40,
+		0x00, 0x70,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0x74, 0x68, 0x69,
+		0xb8, 0xc1, 0xb7, 0xbd, 0xcd, 0xb8, 0x93, 0xab,
+		0x88, 0xb4, 0xc4, 0xb1, 0x00, 0x6b, 0x00, 0x4d,
+		0x00, 0x7d,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0x95, 0x8a, 0x89,
+		0xb4, 0xc6, 0xb2, 0xc5, 0xd2, 0xbf, 0x90, 0xa8,
+		0x85, 0xb5, 0xc4, 0xb3, 0x00, 0x7b, 0x00, 0x5d,
+		0x00, 0x8f,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0x9f, 0x98, 0x92,
+		0xb3, 0xc4, 0xb0, 0xbc, 0xcc, 0xb4, 0x91, 0xa6,
+		0x87, 0xb5, 0xc5, 0xb4, 0x00, 0x87, 0x00, 0x6a,
+		0x00, 0x9e,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0x99, 0x93, 0x8b,
+		0xb2, 0xc2, 0xb0, 0xbd, 0xce, 0xb4, 0x90, 0xa6,
+		0x87, 0xb3, 0xc3, 0xb2, 0x00, 0x8d, 0x00, 0x70,
+		0x00, 0xa4,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa7, 0xa5, 0x99,
+		0xb2, 0xc2, 0xb0, 0xbb, 0xcd, 0xb1, 0x93, 0xa7,
+		0x8a, 0xb2, 0xc1, 0xb0, 0x00, 0x92, 0x00, 0x75,
+		0x00, 0xaa,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa0, 0xa0, 0x93,
+		0xb6, 0xc4, 0xb4, 0xb5, 0xc8, 0xaa, 0x94, 0xa9,
+		0x8c, 0xb2, 0xc0, 0xb0, 0x00, 0x97, 0x00, 0x7a,
+		0x00, 0xaf,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa3, 0xa7, 0x96,
+		0xb3, 0xc2, 0xb0, 0xba, 0xcb, 0xb0, 0x94, 0xa8,
+		0x8c, 0xb0, 0xbf, 0xaf, 0x00, 0x9f, 0x00, 0x83,
+		0x00, 0xb9,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0x9d, 0xa2, 0x90,
+		0xb6, 0xc5, 0xb3, 0xb8, 0xc9, 0xae, 0x94, 0xa8,
+		0x8d, 0xaf, 0xbd, 0xad, 0x00, 0xa4, 0x00, 0x88,
+		0x00, 0xbf,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa6, 0xac, 0x97,
+		0xb4, 0xc4, 0xb1, 0xbb, 0xcb, 0xb2, 0x93, 0xa7,
+		0x8d, 0xae, 0xbc, 0xad, 0x00, 0xa7, 0x00, 0x8c,
+		0x00, 0xc3,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa2, 0xa9, 0x93,
+		0xb6, 0xc5, 0xb2, 0xba, 0xc9, 0xb0, 0x93, 0xa7,
+		0x8d, 0xae, 0xbb, 0xac, 0x00, 0xab, 0x00, 0x90,
+		0x00, 0xc8,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0x9e, 0xa6, 0x8f,
+		0xb7, 0xc6, 0xb3, 0xb8, 0xc8, 0xb0, 0x93, 0xa6,
+		0x8c, 0xae, 0xbb, 0xad, 0x00, 0xae, 0x00, 0x93,
+		0x00, 0xcc,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xab, 0xb4, 0x9c,
+		0xb3, 0xc3, 0xaf, 0xb7, 0xc7, 0xaf, 0x93, 0xa6,
+		0x8c, 0xaf, 0xbc, 0xad, 0x00, 0xb1, 0x00, 0x97,
+		0x00, 0xcf,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa6, 0xb1, 0x98,
+		0xb1, 0xc2, 0xab, 0xba, 0xc9, 0xb2, 0x93, 0xa6,
+		0x8d, 0xae, 0xba, 0xab, 0x00, 0xb5, 0x00, 0x9b,
+		0x00, 0xd4,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa3, 0xae, 0x94,
+		0xb2, 0xc3, 0xac, 0xbb, 0xca, 0xb4, 0x91, 0xa4,
+		0x8a, 0xae, 0xba, 0xac, 0x00, 0xb8, 0x00, 0x9e,
+		0x00, 0xd8,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xab, 0xb7, 0x9c,
+		0xae, 0xc0, 0xa9, 0xba, 0xc9, 0xb3, 0x92, 0xa5,
+		0x8b, 0xad, 0xb9, 0xab, 0x00, 0xbb, 0x00, 0xa1,
+		0x00, 0xdc,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa7, 0xb4, 0x97,
+		0xb0, 0xc1, 0xaa, 0xb9, 0xc8, 0xb2, 0x92, 0xa5,
+		0x8c, 0xae, 0xb9, 0xab, 0x00, 0xbe, 0x00, 0xa4,
+		0x00, 0xdf,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa3, 0xb0, 0x94,
+		0xb0, 0xc2, 0xab, 0xbb, 0xc9, 0xb3, 0x91, 0xa4,
+		0x8b, 0xad, 0xb8, 0xaa, 0x00, 0xc1, 0x00, 0xa8,
+		0x00, 0xe2,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa3, 0xb0, 0x94,
+		0xae, 0xbf, 0xa8, 0xb9, 0xc8, 0xb3, 0x92, 0xa4,
+		0x8b, 0xad, 0xb7, 0xa9, 0x00, 0xc4, 0x00, 0xab,
+		0x00, 0xe6,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa7, 0xb6, 0x98,
+		0xaf, 0xc0, 0xa8, 0xb8, 0xc7, 0xb2, 0x93, 0xa5,
+		0x8d, 0xad, 0xb7, 0xa9, 0x00, 0xc7, 0x00, 0xae,
+		0x00, 0xe9,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa4, 0xb3, 0x95,
+		0xaf, 0xc1, 0xa9, 0xb9, 0xc8, 0xb3, 0x92, 0xa4,
+		0x8b, 0xad, 0xb7, 0xaa, 0x00, 0xc9, 0x00, 0xb0,
+		0x00, 0xec,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa4, 0xb3, 0x95,
+		0xac, 0xbe, 0xa6, 0xbb, 0xc9, 0xb4, 0x90, 0xa3,
+		0x8a, 0xad, 0xb7, 0xa9, 0x00, 0xcc, 0x00, 0xb4,
+		0x00, 0xf0,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa0, 0xb0, 0x91,
+		0xae, 0xc0, 0xa6, 0xba, 0xc8, 0xb4, 0x91, 0xa4,
+		0x8b, 0xad, 0xb7, 0xa9, 0x00, 0xcf, 0x00, 0xb7,
+		0x00, 0xf3,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa7, 0xb8, 0x98,
+		0xab, 0xbd, 0xa4, 0xbb, 0xc9, 0xb5, 0x91, 0xa3,
+		0x8b, 0xac, 0xb6, 0xa8, 0x00, 0xd1, 0x00, 0xb9,
+		0x00, 0xf6,
+	}, {
+		0xfa, 0x01, 0x71, 0x31, 0x7b, 0xa4, 0xb5, 0x95,
+		0xa9, 0xbc, 0xa1, 0xbb, 0xc9, 0xb5, 0x91, 0xa3,
+		0x8a, 0xad, 0xb6, 0xa8, 0x00, 0xd6, 0x00, 0xbf,
+		0x00, 0xfc,
+	},
+};
+
+static const s6e8aa0_gamma_table s6e8aa0_gamma_tables_v96[GAMMA_LEVEL_NUM] = {
+	{
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x00, 0xff,
+		0xdf, 0x1f, 0xd7, 0xdc, 0xb7, 0xe1, 0xc0, 0xaf,
+		0xc4, 0xd2, 0xd0, 0xcf, 0x00, 0x4d, 0x00, 0x40,
+		0x00, 0x5f,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x00, 0xff,
+		0xd5, 0x35, 0xcf, 0xdc, 0xc1, 0xe1, 0xbf, 0xb3,
+		0xc1, 0xd2, 0xd1, 0xce,	0x00, 0x53, 0x00, 0x46,
+		0x00, 0x67,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x00, 0xff,
+		0xd2, 0x64, 0xcf, 0xdb, 0xc6, 0xe1, 0xbd, 0xb3,
+		0xbd, 0xd2, 0xd2, 0xce,	0x00, 0x59, 0x00, 0x4b,
+		0x00, 0x6e,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x00, 0xff,
+		0xd0, 0x7c, 0xcf, 0xdb, 0xc9, 0xe0, 0xbc, 0xb4,
+		0xbb, 0xcf, 0xd1, 0xcc, 0x00, 0x5f, 0x00, 0x50,
+		0x00, 0x75,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x00, 0xff,
+		0xd0, 0x8e, 0xd1, 0xdb, 0xcc, 0xdf, 0xbb, 0xb6,
+		0xb9, 0xd0, 0xd1, 0xcd,	0x00, 0x63, 0x00, 0x54,
+		0x00, 0x7a,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x00, 0xff,
+		0xd1, 0x9e, 0xd5, 0xda, 0xcd, 0xdd, 0xbb, 0xb7,
+		0xb9, 0xce, 0xce, 0xc9,	0x00, 0x68, 0x00, 0x59,
+		0x00, 0x81,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x00, 0xff,
+		0xd0, 0xa5, 0xd6, 0xda, 0xcf, 0xdd, 0xbb, 0xb7,
+		0xb8, 0xcc, 0xcd, 0xc7,	0x00, 0x6c, 0x00, 0x5c,
+		0x00, 0x86,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x1f, 0xfe,
+		0xd0, 0xae, 0xd7, 0xd9, 0xd0, 0xdb, 0xb9, 0xb6,
+		0xb5, 0xca, 0xcc, 0xc5,	0x00, 0x74, 0x00, 0x63,
+		0x00, 0x90,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x1f, 0xf9,
+		0xcf, 0xb0, 0xd6, 0xd9, 0xd1, 0xdb, 0xb9, 0xb6,
+		0xb4, 0xca, 0xcb, 0xc5,	0x00, 0x77, 0x00, 0x66,
+		0x00, 0x94,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xff, 0x1f, 0xf7,
+		0xcf, 0xb3, 0xd7, 0xd8, 0xd1, 0xd9, 0xb7, 0xb6,
+		0xb3, 0xc9, 0xca, 0xc3,	0x00, 0x7b, 0x00, 0x69,
+		0x00, 0x99,
+
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xfd, 0x2f, 0xf7,
+		0xdf, 0xb5, 0xd6, 0xd8, 0xd1, 0xd8, 0xb6, 0xb5,
+		0xb2, 0xca, 0xcb, 0xc4,	0x00, 0x7e, 0x00, 0x6c,
+		0x00, 0x9d,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xfa, 0x2f, 0xf5,
+		0xce, 0xb6, 0xd5, 0xd7, 0xd2, 0xd8, 0xb6, 0xb4,
+		0xb0, 0xc7, 0xc9, 0xc1,	0x00, 0x84, 0x00, 0x71,
+		0x00, 0xa5,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xf7, 0x2f, 0xf2,
+		0xce, 0xb9, 0xd5, 0xd8, 0xd2, 0xd8, 0xb4, 0xb4,
+		0xaf, 0xc7, 0xc9, 0xc1,	0x00, 0x87, 0x00, 0x73,
+		0x00, 0xa8,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xf5, 0x2f, 0xf0,
+		0xdf, 0xba, 0xd5, 0xd7, 0xd2, 0xd7, 0xb4, 0xb4,
+		0xaf, 0xc5, 0xc7, 0xbf,	0x00, 0x8a, 0x00, 0x76,
+		0x00, 0xac,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xf2, 0x2f, 0xed,
+		0xcE, 0xbb, 0xd4, 0xd6, 0xd2, 0xd6, 0xb5, 0xb4,
+		0xaF, 0xc5, 0xc7, 0xbf,	0x00, 0x8c, 0x00, 0x78,
+		0x00, 0xaf,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xef, 0x2f, 0xeb,
+		0xcd, 0xbb, 0xd2, 0xd7, 0xd3, 0xd6, 0xb3, 0xb4,
+		0xae, 0xc5, 0xc6, 0xbe,	0x00, 0x91, 0x00, 0x7d,
+		0x00, 0xb6,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xee, 0x2f, 0xea,
+		0xce, 0xbd, 0xd4, 0xd6, 0xd2, 0xd5, 0xb2, 0xb3,
+		0xad, 0xc3, 0xc4, 0xbb,	0x00, 0x94, 0x00, 0x7f,
+		0x00, 0xba,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xec, 0x2f, 0xe8,
+		0xce, 0xbe, 0xd3, 0xd6, 0xd3, 0xd5, 0xb2, 0xb2,
+		0xac, 0xc3, 0xc5, 0xbc,	0x00, 0x96, 0x00, 0x81,
+		0x00, 0xbd,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xeb, 0x2f, 0xe7,
+		0xce, 0xbf, 0xd3, 0xd6, 0xd2, 0xd5, 0xb1, 0xb2,
+		0xab, 0xc2, 0xc4, 0xbb,	0x00, 0x99, 0x00, 0x83,
+		0x00, 0xc0,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xef, 0x5f, 0xe9,
+		0xca, 0xbf, 0xd3, 0xd5, 0xd2, 0xd4, 0xb2, 0xb2,
+		0xab, 0xc1, 0xc4, 0xba,	0x00, 0x9b, 0x00, 0x85,
+		0x00, 0xc3,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xea, 0x5f, 0xe8,
+		0xee, 0xbf, 0xd2, 0xd5, 0xd2, 0xd4, 0xb1, 0xb2,
+		0xab, 0xc1, 0xc2, 0xb9,	0x00, 0x9D, 0x00, 0x87,
+		0x00, 0xc6,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xe9, 0x5f, 0xe7,
+		0xcd, 0xbf, 0xd2, 0xd6, 0xd2, 0xd4, 0xb1, 0xb2,
+		0xab, 0xbe, 0xc0, 0xb7,	0x00, 0xa1, 0x00, 0x8a,
+		0x00, 0xca,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xe8, 0x61, 0xe6,
+		0xcd, 0xbf, 0xd1, 0xd6, 0xd3, 0xd4, 0xaf, 0xb0,
+		0xa9, 0xbe, 0xc1, 0xb7,	0x00, 0xa3, 0x00, 0x8b,
+		0x00, 0xce,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xe8, 0x62, 0xe5,
+		0xcc, 0xc0, 0xd0, 0xd6, 0xd2, 0xd4, 0xaf, 0xb1,
+		0xa9, 0xbd, 0xc0, 0xb6,	0x00, 0xa5, 0x00, 0x8d,
+		0x00, 0xd0,
+	}, {
+		0xfa, 0x01, 0x1f, 0x1f, 0x1f, 0xe7, 0x7f, 0xe3,
+		0xcc, 0xc1, 0xd0, 0xd5, 0xd3, 0xd3, 0xae, 0xaf,
+		0xa8, 0xbe, 0xc0, 0xb7,	0x00, 0xa8, 0x00, 0x90,
+		0x00, 0xd3,
+	}
+};
+
+static const s6e8aa0_gamma_table s6e8aa0_gamma_tables_v32[GAMMA_LEVEL_NUM] = {
+	{
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0x72, 0x5e, 0x6b,
+		0xa1, 0xa7, 0x9a, 0xb4, 0xcb, 0xb8, 0x92, 0xac,
+		0x97, 0xb4, 0xc3, 0xb5, 0x00, 0x4e, 0x00, 0x37,
+		0x00, 0x58,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0x85, 0x71, 0x7d,
+		0xa6, 0xb6, 0xa1, 0xb5, 0xca, 0xba, 0x93, 0xac,
+		0x98, 0xb2, 0xc0, 0xaf, 0x00, 0x59, 0x00, 0x43,
+		0x00, 0x64,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xa4, 0x94, 0x9e,
+		0xa0, 0xbb, 0x9c, 0xc3, 0xd2, 0xc6, 0x93, 0xaa,
+		0x95, 0xb7, 0xc2, 0xb4, 0x00, 0x65, 0x00, 0x50,
+		0x00, 0x74,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xad, 0xa1, 0xa6,
+		0xa0, 0xb9, 0x9b, 0xc3, 0xd1, 0xc8, 0x90, 0xa6,
+		0x90, 0xbb, 0xc3, 0xb7, 0x00, 0x6f, 0x00, 0x5b,
+		0x00, 0x80,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xa6, 0x9d, 0x9f,
+		0x9f, 0xb8, 0x9a, 0xc7, 0xd5, 0xcc, 0x90, 0xa5,
+		0x8f, 0xb8, 0xc1, 0xb6, 0x00, 0x74, 0x00, 0x60,
+		0x00, 0x85,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xb3, 0xae, 0xae,
+		0x9e, 0xb7, 0x9a, 0xc8, 0xd6, 0xce, 0x91, 0xa6,
+		0x90, 0xb6, 0xc0, 0xb3, 0x00, 0x78, 0x00, 0x65,
+		0x00, 0x8a,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xad, 0xa9, 0xa8,
+		0xa3, 0xb9, 0x9e, 0xc4, 0xd3, 0xcb, 0x94, 0xa6,
+		0x90, 0xb6, 0xbf, 0xb3, 0x00, 0x7c, 0x00, 0x69,
+		0x00, 0x8e,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xaf, 0xaf, 0xa9,
+		0xa5, 0xbc, 0xa2, 0xc7, 0xd5, 0xcd, 0x93, 0xa5,
+		0x8f, 0xb4, 0xbd, 0xb1, 0x00, 0x83, 0x00, 0x70,
+		0x00, 0x96,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xa9, 0xab, 0xa3,
+		0xaa, 0xbf, 0xa7, 0xc5, 0xd3, 0xcb, 0x93, 0xa5,
+		0x8f, 0xb2, 0xbb, 0xb0, 0x00, 0x86, 0x00, 0x74,
+		0x00, 0x9b,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xb1, 0xb5, 0xab,
+		0xab, 0xc0, 0xa9, 0xc7, 0xd4, 0xcc, 0x94, 0xa4,
+		0x8f, 0xb1, 0xbb, 0xaf, 0x00, 0x8a, 0x00, 0x77,
+		0x00, 0x9e,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xad, 0xb2, 0xa7,
+		0xae, 0xc2, 0xab, 0xc5, 0xd3, 0xca, 0x93, 0xa4,
+		0x8f, 0xb1, 0xba, 0xae, 0x00, 0x8d, 0x00, 0x7b,
+		0x00, 0xa2,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xa9, 0xaf, 0xa3,
+		0xb0, 0xc3, 0xae, 0xc4, 0xd1, 0xc8, 0x93, 0xa4,
+		0x8f, 0xb1, 0xba, 0xaf, 0x00, 0x8f, 0x00, 0x7d,
+		0x00, 0xa5,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xb4, 0xbd, 0xaf,
+		0xae, 0xc1, 0xab, 0xc2, 0xd0, 0xc6, 0x94, 0xa4,
+		0x8f, 0xb1, 0xba, 0xaf, 0x00, 0x92, 0x00, 0x80,
+		0x00, 0xa8,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xb0, 0xb9, 0xac,
+		0xad, 0xc1, 0xab, 0xc4, 0xd1, 0xc7, 0x95, 0xa4,
+		0x90, 0xb0, 0xb9, 0xad, 0x00, 0x95, 0x00, 0x84,
+		0x00, 0xac,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xad, 0xb6, 0xa7,
+		0xaf, 0xc2, 0xae, 0xc5, 0xd1, 0xc7, 0x93, 0xa3,
+		0x8e, 0xb0, 0xb9, 0xad, 0x00, 0x98, 0x00, 0x86,
+		0x00, 0xaf,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xb4, 0xbf, 0xaf,
+		0xad, 0xc1, 0xab, 0xc3, 0xd0, 0xc6, 0x94, 0xa3,
+		0x8f, 0xaf, 0xb8, 0xac, 0x00, 0x9a, 0x00, 0x89,
+		0x00, 0xb2,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xb0, 0xbc, 0xac,
+		0xaf, 0xc2, 0xad, 0xc2, 0xcf, 0xc4, 0x94, 0xa3,
+		0x90, 0xaf, 0xb8, 0xad, 0x00, 0x9c, 0x00, 0x8b,
+		0x00, 0xb5,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xad, 0xb9, 0xa7,
+		0xb1, 0xc4, 0xaf, 0xc3, 0xcf, 0xc5, 0x94, 0xa3,
+		0x8f, 0xae, 0xb7, 0xac, 0x00, 0x9f, 0x00, 0x8e,
+		0x00, 0xb8,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xad, 0xb9, 0xa7,
+		0xaf, 0xc2, 0xad, 0xc1, 0xce, 0xc3, 0x95, 0xa3,
+		0x90, 0xad, 0xb6, 0xab, 0x00, 0xa2, 0x00, 0x91,
+		0x00, 0xbb,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xb1, 0xbe, 0xac,
+		0xb1, 0xc4, 0xaf, 0xc1, 0xcd, 0xc1, 0x95, 0xa4,
+		0x91, 0xad, 0xb6, 0xab, 0x00, 0xa4, 0x00, 0x93,
+		0x00, 0xbd,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xad, 0xbb, 0xa8,
+		0xb3, 0xc5, 0xb2, 0xc1, 0xcd, 0xc2, 0x95, 0xa3,
+		0x90, 0xad, 0xb6, 0xab, 0x00, 0xa6, 0x00, 0x95,
+		0x00, 0xc0,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xad, 0xbb, 0xa8,
+		0xb0, 0xc3, 0xaf, 0xc2, 0xce, 0xc2, 0x94, 0xa2,
+		0x90, 0xac, 0xb6, 0xab, 0x00, 0xa8, 0x00, 0x98,
+		0x00, 0xc3,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xa9, 0xb8, 0xa5,
+		0xb3, 0xc5, 0xb2, 0xc1, 0xcc, 0xc0, 0x95, 0xa2,
+		0x90, 0xad, 0xb6, 0xab, 0x00, 0xaa, 0x00, 0x9a,
+		0x00, 0xc5,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xb0, 0xc0, 0xac,
+		0xb0, 0xc3, 0xaf, 0xc1, 0xcd, 0xc1, 0x95, 0xa2,
+		0x90, 0xac, 0xb5, 0xa9, 0x00, 0xac, 0x00, 0x9c,
+		0x00, 0xc8,
+	}, {
+		0xfa, 0x01, 0x43, 0x14, 0x45, 0xad, 0xbd, 0xa8,
+		0xaf, 0xc2, 0xaf, 0xc1, 0xcc, 0xc0, 0x95, 0xa2,
+		0x90, 0xac, 0xb5, 0xaa, 0x00, 0xb1, 0x00, 0xa1,
+		0x00, 0xcc,
+	},
+};
+
+static const struct s6e8aa0_variant s6e8aa0_variants[] = {
+	{
+		.version = 32,
+		.gamma_tables = s6e8aa0_gamma_tables_v32,
+	}, {
+		.version = 96,
+		.gamma_tables = s6e8aa0_gamma_tables_v96,
+	}, {
+		.version = 142,
+		.gamma_tables = s6e8aa0_gamma_tables_v142,
+	}, {
+		.version = 210,
+		.gamma_tables = s6e8aa0_gamma_tables_v142,
+	}
+};
+
+static void s6e8aa0_brightness_set(struct s6e8aa0 *ctx)
+{
+	const u8 *gamma;
+
+	if (ctx->error)
+		return;
+
+	gamma = ctx->variant->gamma_tables[ctx->brightness];
+
+	if (ctx->version >= 142)
+		s6e8aa0_elvss_nvm_set(ctx);
+
+	s6e8aa0_dcs_write(ctx, gamma, GAMMA_TABLE_LEN);
+
+	/* update gamma table. */
+	s6e8aa0_dcs_write_seq_static(ctx, 0xf7, 0x03);
+}
+
+static void s6e8aa0_panel_init(struct s6e8aa0 *ctx)
+{
+	s6e8aa0_apply_level_1_key(ctx);
+	s6e8aa0_apply_level_2_key(ctx);
+	msleep(20);
+
+	s6e8aa0_dcs_write_seq_static(ctx, MIPI_DCS_EXIT_SLEEP_MODE);
+	msleep(40);
+
+	s6e8aa0_panel_cond_set(ctx);
+	s6e8aa0_display_condition_set(ctx);
+	s6e8aa0_brightness_set(ctx);
+	s6e8aa0_etc_source_control(ctx);
+	s6e8aa0_etc_pentile_control(ctx);
+	s6e8aa0_elvss_nvm_set(ctx);
+	s6e8aa0_etc_power_control(ctx);
+	s6e8aa0_etc_elvss_control(ctx);
+	msleep(ctx->init_delay);
+}
+
+static void s6e8aa0_set_maximum_return_packet_size(struct s6e8aa0 *ctx,
+						   int size)
+{
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(ctx->dev);
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+	u8 buf[] = {size, 0};
+	struct mipi_dsi_msg msg = {
+		.channel = dsi->channel,
+		.type = MIPI_DSI_SET_MAXIMUM_RETURN_PACKET_SIZE,
+		.tx_len = sizeof(buf),
+		.tx_buf = buf
+	};
+	int ret;
+
+	if (ctx->error < 0)
+		return;
+
+	if (!ops || !ops->transfer)
+		ret = -EIO;
+	else
+		ret = ops->transfer(dsi->host, &msg);
+
+	if (ret < 0) {
+		dev_err(ctx->dev,
+			"error %d setting maximum return packet size to %d\n",
+			ret, size);
+		ctx->error = ret;
+	}
+}
+
+static void s6e8aa0_read_mtp_id(struct s6e8aa0 *ctx)
+{
+	u8 id[3];
+	int ret, i;
+
+	ret = s6e8aa0_dcs_read(ctx, 0xd1, id, ARRAY_SIZE(id));
+	if (ret < ARRAY_SIZE(id) || id[0] == 0x00) {
+		dev_err(ctx->dev, "read id failed\n");
+		ctx->error = -EIO;
+		return;
+	}
+
+	dev_info(ctx->dev, "ID: 0x%2x, 0x%2x, 0x%2x\n", id[0], id[1], id[2]);
+
+	for (i = 0; i < ARRAY_SIZE(s6e8aa0_variants); ++i) {
+		if (id[1] == s6e8aa0_variants[i].version)
+			break;
+	}
+	if (i >= ARRAY_SIZE(s6e8aa0_variants)) {
+		dev_err(ctx->dev, "unsupported display version %d\n", id[1]);
+		ctx->error = -EINVAL;
+		return;
+	}
+
+	ctx->variant = &s6e8aa0_variants[i];
+	ctx->version = id[1];
+	ctx->id = id[2];
+}
+
+static void s6e8aa0_set_sequence(struct s6e8aa0 *ctx)
+{
+	s6e8aa0_set_maximum_return_packet_size(ctx, 3);
+	s6e8aa0_read_mtp_id(ctx);
+	s6e8aa0_panel_init(ctx);
+	s6e8aa0_dcs_write_seq_static(ctx, MIPI_DCS_SET_DISPLAY_ON);
+}
+
+static int s6e8aa0_power_on(struct s6e8aa0 *ctx)
+{
+	int ret;
+
+	ret = regulator_bulk_enable(ARRAY_SIZE(ctx->supplies), ctx->supplies);
+	if (ret < 0)
+		return ret;
+
+	msleep(ctx->power_on_delay);
+
+	gpiod_set_value(ctx->reset_gpio, 0);
+	usleep_range(10000, 11000);
+	gpiod_set_value(ctx->reset_gpio, 1);
+
+	msleep(ctx->reset_delay);
+
+	return 0;
+}
+
+static int s6e8aa0_power_off(struct s6e8aa0 *ctx)
+{
+	return regulator_bulk_disable(ARRAY_SIZE(ctx->supplies), ctx->supplies);
+}
+
+static int s6e8aa0_disable(struct drm_panel *panel)
+{
+	struct s6e8aa0 *ctx = panel_to_s6e8aa0(panel);
+
+	s6e8aa0_dcs_write_seq_static(ctx, MIPI_DCS_ENTER_SLEEP_MODE);
+	s6e8aa0_dcs_write_seq_static(ctx, MIPI_DCS_SET_DISPLAY_OFF);
+	msleep(40);
+
+	s6e8aa0_clear_error(ctx);
+
+	return s6e8aa0_power_off(ctx);
+}
+
+static int s6e8aa0_enable(struct drm_panel *panel)
+{
+	struct s6e8aa0 *ctx = panel_to_s6e8aa0(panel);
+	int ret;
+
+	ret = s6e8aa0_power_on(ctx);
+	if (ret < 0)
+		return ret;
+
+	s6e8aa0_set_sequence(ctx);
+	ret = ctx->error;
+
+	if (ret < 0)
+		s6e8aa0_disable(panel);
+
+	return ret;
+}
+
+static int s6e8aa0_get_modes(struct drm_panel *panel)
+{
+	struct drm_connector *connector = panel->connector;
+	struct s6e8aa0 *ctx = panel_to_s6e8aa0(panel);
+	struct drm_display_mode *mode;
+
+	mode = drm_mode_create(connector->dev);
+	if (!mode) {
+		DRM_ERROR("failed to create a new display mode\n");
+		return 0;
+	}
+
+	drm_display_mode_from_videomode(&ctx->vm, mode);
+	mode->width_mm = ctx->width_mm;
+	mode->height_mm = ctx->height_mm;
+	connector->display_info.width_mm = mode->width_mm;
+	connector->display_info.height_mm = mode->height_mm;
+
+	mode->type = DRM_MODE_TYPE_DRIVER | DRM_MODE_TYPE_PREFERRED;
+	drm_mode_probed_add(connector, mode);
+
+	return 1;
+}
+
+static const struct drm_panel_funcs s6e8aa0_drm_funcs = {
+	.disable = s6e8aa0_disable,
+	.enable = s6e8aa0_enable,
+	.get_modes = s6e8aa0_get_modes,
+};
+
+static int s6e8aa0_parse_dt(struct s6e8aa0 *ctx)
+{
+	struct device *dev = ctx->dev;
+	struct device_node *np = dev->of_node;
+	int ret;
+
+	ret = of_get_videomode(np, &ctx->vm, 0);
+	if (ret < 0)
+		return ret;
+
+	of_property_read_u32(np, "power-on-delay", &ctx->power_on_delay);
+	of_property_read_u32(np, "reset-delay", &ctx->reset_delay);
+	of_property_read_u32(np, "init-delay", &ctx->init_delay);
+	of_property_read_u32(np, "panel-width-mm", &ctx->width_mm);
+	of_property_read_u32(np, "panel-height-mm", &ctx->height_mm);
+
+	ctx->flip_horizontal = of_property_read_bool(np, "flip-horizontal");
+	ctx->flip_vertical = of_property_read_bool(np, "flip-vertical");
+
+	return 0;
+}
+
+static int s6e8aa0_probe(struct mipi_dsi_device *dsi)
+{
+	struct device *dev = &dsi->dev;
+	struct s6e8aa0 *ctx;
+	int ret;
+
+	ctx = devm_kzalloc(dev, sizeof(struct s6e8aa0), GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+
+	mipi_dsi_set_drvdata(dsi, ctx);
+
+	ctx->dev = dev;
+
+	dsi->lanes = 4;
+	dsi->format = MIPI_DSI_FMT_RGB888;
+	dsi->mode_flags = MIPI_DSI_MODE_VIDEO | MIPI_DSI_MODE_VIDEO_BURST
+		| MIPI_DSI_MODE_VIDEO_HFP | MIPI_DSI_MODE_VIDEO_HBP
+		| MIPI_DSI_MODE_VIDEO_HSA | MIPI_DSI_MODE_EOT_PACKET
+		| MIPI_DSI_MODE_VSYNC_FLUSH | MIPI_DSI_MODE_VIDEO_AUTO_VERT;
+
+	ret = s6e8aa0_parse_dt(ctx);
+	if (ret < 0)
+		return ret;
+
+	ctx->supplies[0].supply = "vdd3";
+	ctx->supplies[1].supply = "vci";
+	ret = devm_regulator_bulk_get(dev, ARRAY_SIZE(ctx->supplies),
+				      ctx->supplies);
+	if (ret < 0) {
+		dev_err(dev, "failed to get regulators: %d\n", ret);
+		return ret;
+	}
+
+	ctx->reset_gpio = devm_gpiod_get(dev, "reset");
+	if (IS_ERR(ctx->reset_gpio)) {
+		dev_err(dev, "cannot get reset-gpios %ld\n",
+			PTR_ERR(ctx->reset_gpio));
+		return PTR_ERR(ctx->reset_gpio);
+	}
+	ret = gpiod_direction_output(ctx->reset_gpio, 1);
+	if (ret < 0) {
+		dev_err(dev, "cannot configure reset-gpios %d\n", ret);
+		return ret;
+	}
+
+	ctx->brightness = GAMMA_LEVEL_NUM - 1;
+
+	drm_panel_init(&ctx->panel);
+	ctx->panel.dev = dev;
+	ctx->panel.funcs = &s6e8aa0_drm_funcs;
+
+	ret = drm_panel_add(&ctx->panel);
+	if (ret < 0)
+		return ret;
+
+	ret = mipi_dsi_attach(dsi);
+	if (ret < 0)
+		drm_panel_remove(&ctx->panel);
+
+	return ret;
+}
+
+static int s6e8aa0_remove(struct mipi_dsi_device *dsi)
+{
+	struct s6e8aa0 *ctx = mipi_dsi_get_drvdata(dsi);
+
+	mipi_dsi_detach(dsi);
+	drm_panel_remove(&ctx->panel);
+
+	return 0;
+}
+
+static struct of_device_id s6e8aa0_of_match[] = {
+	{ .compatible = "samsung,s6e8aa0" },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, s6e8aa0_of_match);
+
+static struct mipi_dsi_driver s6e8aa0_driver = {
+	.probe = s6e8aa0_probe,
+	.remove = s6e8aa0_remove,
+	.driver = {
+		.name = "panel_s6e8aa0",
+		.owner = THIS_MODULE,
+		.of_match_table = s6e8aa0_of_match,
+	},
+};
+module_mipi_dsi_driver(s6e8aa0_driver);
+
+MODULE_AUTHOR("Donghwa Lee <dh09.lee@samsung.com>");
+MODULE_AUTHOR("Inki Dae <inki.dae@samsung.com>");
+MODULE_AUTHOR("Joongmock Shin <jmock.shin@samsung.com>");
+MODULE_AUTHOR("Eunchul Kim <chulspro.kim@samsung.com>");
+MODULE_AUTHOR("Tomasz Figa <t.figa@samsung.com>");
+MODULE_AUTHOR("Andrzej Hajda <a.hajda@samsung.com>");
+MODULE_DESCRIPTION("MIPI-DSI based s6e8aa0 AMOLED LCD Panel Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/panel/panel-simple.c b/drivers/gpu/drm/panel/panel-simple.c
new file mode 100644
index 00000000000..a25136132c3
--- /dev/null
+++ b/drivers/gpu/drm/panel/panel-simple.c
@@ -0,0 +1,702 @@
+/*
+ * Copyright (C) 2013, NVIDIA Corporation.  All rights reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sub license,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/backlight.h>
+#include <linux/gpio/consumer.h>
+#include <linux/module.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/regulator/consumer.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_panel.h>
+
+struct panel_desc {
+	const struct drm_display_mode *modes;
+	unsigned int num_modes;
+
+	struct {
+		unsigned int width;
+		unsigned int height;
+	} size;
+};
+
+struct panel_simple {
+	struct drm_panel base;
+	bool enabled;
+
+	const struct panel_desc *desc;
+
+	struct backlight_device *backlight;
+	struct regulator *supply;
+	struct i2c_adapter *ddc;
+
+	struct gpio_desc *enable_gpio;
+};
+
+static inline struct panel_simple *to_panel_simple(struct drm_panel *panel)
+{
+	return container_of(panel, struct panel_simple, base);
+}
+
+static int panel_simple_get_fixed_modes(struct panel_simple *panel)
+{
+	struct drm_connector *connector = panel->base.connector;
+	struct drm_device *drm = panel->base.drm;
+	struct drm_display_mode *mode;
+	unsigned int i, num = 0;
+
+	if (!panel->desc)
+		return 0;
+
+	for (i = 0; i < panel->desc->num_modes; i++) {
+		const struct drm_display_mode *m = &panel->desc->modes[i];
+
+		mode = drm_mode_duplicate(drm, m);
+		if (!mode) {
+			dev_err(drm->dev, "failed to add mode %ux%u@%u\n",
+				m->hdisplay, m->vdisplay, m->vrefresh);
+			continue;
+		}
+
+		drm_mode_set_name(mode);
+
+		drm_mode_probed_add(connector, mode);
+		num++;
+	}
+
+	connector->display_info.width_mm = panel->desc->size.width;
+	connector->display_info.height_mm = panel->desc->size.height;
+
+	return num;
+}
+
+static int panel_simple_disable(struct drm_panel *panel)
+{
+	struct panel_simple *p = to_panel_simple(panel);
+
+	if (!p->enabled)
+		return 0;
+
+	if (p->backlight) {
+		p->backlight->props.power = FB_BLANK_POWERDOWN;
+		backlight_update_status(p->backlight);
+	}
+
+	if (p->enable_gpio)
+		gpiod_set_value_cansleep(p->enable_gpio, 0);
+
+	regulator_disable(p->supply);
+	p->enabled = false;
+
+	return 0;
+}
+
+static int panel_simple_enable(struct drm_panel *panel)
+{
+	struct panel_simple *p = to_panel_simple(panel);
+	int err;
+
+	if (p->enabled)
+		return 0;
+
+	err = regulator_enable(p->supply);
+	if (err < 0) {
+		dev_err(panel->dev, "failed to enable supply: %d\n", err);
+		return err;
+	}
+
+	if (p->enable_gpio)
+		gpiod_set_value_cansleep(p->enable_gpio, 1);
+
+	if (p->backlight) {
+		p->backlight->props.power = FB_BLANK_UNBLANK;
+		backlight_update_status(p->backlight);
+	}
+
+	p->enabled = true;
+
+	return 0;
+}
+
+static int panel_simple_get_modes(struct drm_panel *panel)
+{
+	struct panel_simple *p = to_panel_simple(panel);
+	int num = 0;
+
+	/* probe EDID if a DDC bus is available */
+	if (p->ddc) {
+		struct edid *edid = drm_get_edid(panel->connector, p->ddc);
+		drm_mode_connector_update_edid_property(panel->connector, edid);
+		if (edid) {
+			num += drm_add_edid_modes(panel->connector, edid);
+			kfree(edid);
+		}
+	}
+
+	/* add hard-coded panel modes */
+	num += panel_simple_get_fixed_modes(p);
+
+	return num;
+}
+
+static const struct drm_panel_funcs panel_simple_funcs = {
+	.disable = panel_simple_disable,
+	.enable = panel_simple_enable,
+	.get_modes = panel_simple_get_modes,
+};
+
+static int panel_simple_probe(struct device *dev, const struct panel_desc *desc)
+{
+	struct device_node *backlight, *ddc;
+	struct panel_simple *panel;
+	int err;
+
+	panel = devm_kzalloc(dev, sizeof(*panel), GFP_KERNEL);
+	if (!panel)
+		return -ENOMEM;
+
+	panel->enabled = false;
+	panel->desc = desc;
+
+	panel->supply = devm_regulator_get(dev, "power");
+	if (IS_ERR(panel->supply))
+		return PTR_ERR(panel->supply);
+
+	panel->enable_gpio = devm_gpiod_get(dev, "enable");
+	if (IS_ERR(panel->enable_gpio)) {
+		err = PTR_ERR(panel->enable_gpio);
+		if (err != -ENOENT) {
+			dev_err(dev, "failed to request GPIO: %d\n", err);
+			return err;
+		}
+
+		panel->enable_gpio = NULL;
+	} else {
+		err = gpiod_direction_output(panel->enable_gpio, 0);
+		if (err < 0) {
+			dev_err(dev, "failed to setup GPIO: %d\n", err);
+			return err;
+		}
+	}
+
+	backlight = of_parse_phandle(dev->of_node, "backlight", 0);
+	if (backlight) {
+		panel->backlight = of_find_backlight_by_node(backlight);
+		of_node_put(backlight);
+
+		if (!panel->backlight)
+			return -EPROBE_DEFER;
+	}
+
+	ddc = of_parse_phandle(dev->of_node, "ddc-i2c-bus", 0);
+	if (ddc) {
+		panel->ddc = of_find_i2c_adapter_by_node(ddc);
+		of_node_put(ddc);
+
+		if (!panel->ddc) {
+			err = -EPROBE_DEFER;
+			goto free_backlight;
+		}
+	}
+
+	drm_panel_init(&panel->base);
+	panel->base.dev = dev;
+	panel->base.funcs = &panel_simple_funcs;
+
+	err = drm_panel_add(&panel->base);
+	if (err < 0)
+		goto free_ddc;
+
+	dev_set_drvdata(dev, panel);
+
+	return 0;
+
+free_ddc:
+	if (panel->ddc)
+		put_device(&panel->ddc->dev);
+free_backlight:
+	if (panel->backlight)
+		put_device(&panel->backlight->dev);
+
+	return err;
+}
+
+static int panel_simple_remove(struct device *dev)
+{
+	struct panel_simple *panel = dev_get_drvdata(dev);
+
+	drm_panel_detach(&panel->base);
+	drm_panel_remove(&panel->base);
+
+	panel_simple_disable(&panel->base);
+
+	if (panel->ddc)
+		put_device(&panel->ddc->dev);
+
+	if (panel->backlight)
+		put_device(&panel->backlight->dev);
+
+	return 0;
+}
+
+static void panel_simple_shutdown(struct device *dev)
+{
+	struct panel_simple *panel = dev_get_drvdata(dev);
+
+	panel_simple_disable(&panel->base);
+}
+
+static const struct drm_display_mode auo_b101aw03_mode = {
+	.clock = 51450,
+	.hdisplay = 1024,
+	.hsync_start = 1024 + 156,
+	.hsync_end = 1024 + 156 + 8,
+	.htotal = 1024 + 156 + 8 + 156,
+	.vdisplay = 600,
+	.vsync_start = 600 + 16,
+	.vsync_end = 600 + 16 + 6,
+	.vtotal = 600 + 16 + 6 + 16,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc auo_b101aw03 = {
+	.modes = &auo_b101aw03_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 223,
+		.height = 125,
+	},
+};
+
+static const struct drm_display_mode auo_b133xtn01_mode = {
+	.clock = 69500,
+	.hdisplay = 1366,
+	.hsync_start = 1366 + 48,
+	.hsync_end = 1366 + 48 + 32,
+	.htotal = 1366 + 48 + 32 + 20,
+	.vdisplay = 768,
+	.vsync_start = 768 + 3,
+	.vsync_end = 768 + 3 + 6,
+	.vtotal = 768 + 3 + 6 + 13,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc auo_b133xtn01 = {
+	.modes = &auo_b133xtn01_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 293,
+		.height = 165,
+	},
+};
+
+static const struct drm_display_mode chunghwa_claa101wa01a_mode = {
+	.clock = 72070,
+	.hdisplay = 1366,
+	.hsync_start = 1366 + 58,
+	.hsync_end = 1366 + 58 + 58,
+	.htotal = 1366 + 58 + 58 + 58,
+	.vdisplay = 768,
+	.vsync_start = 768 + 4,
+	.vsync_end = 768 + 4 + 4,
+	.vtotal = 768 + 4 + 4 + 4,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc chunghwa_claa101wa01a = {
+	.modes = &chunghwa_claa101wa01a_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 220,
+		.height = 120,
+	},
+};
+
+static const struct drm_display_mode chunghwa_claa101wb01_mode = {
+	.clock = 69300,
+	.hdisplay = 1366,
+	.hsync_start = 1366 + 48,
+	.hsync_end = 1366 + 48 + 32,
+	.htotal = 1366 + 48 + 32 + 20,
+	.vdisplay = 768,
+	.vsync_start = 768 + 16,
+	.vsync_end = 768 + 16 + 8,
+	.vtotal = 768 + 16 + 8 + 16,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc chunghwa_claa101wb01 = {
+	.modes = &chunghwa_claa101wb01_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 223,
+		.height = 125,
+	},
+};
+
+static const struct drm_display_mode edt_et057090dhu_mode = {
+	.clock = 25175,
+	.hdisplay = 640,
+	.hsync_start = 640 + 16,
+	.hsync_end = 640 + 16 + 30,
+	.htotal = 640 + 16 + 30 + 114,
+	.vdisplay = 480,
+	.vsync_start = 480 + 10,
+	.vsync_end = 480 + 10 + 3,
+	.vtotal = 480 + 10 + 3 + 32,
+	.vrefresh = 60,
+	.flags = DRM_MODE_FLAG_NVSYNC | DRM_MODE_FLAG_NHSYNC,
+};
+
+static const struct panel_desc edt_et057090dhu = {
+	.modes = &edt_et057090dhu_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 115,
+		.height = 86,
+	},
+};
+
+static const struct drm_display_mode edt_etm0700g0dh6_mode = {
+	.clock = 33260,
+	.hdisplay = 800,
+	.hsync_start = 800 + 40,
+	.hsync_end = 800 + 40 + 128,
+	.htotal = 800 + 40 + 128 + 88,
+	.vdisplay = 480,
+	.vsync_start = 480 + 10,
+	.vsync_end = 480 + 10 + 2,
+	.vtotal = 480 + 10 + 2 + 33,
+	.vrefresh = 60,
+	.flags = DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC,
+};
+
+static const struct panel_desc edt_etm0700g0dh6 = {
+	.modes = &edt_etm0700g0dh6_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 152,
+		.height = 91,
+	},
+};
+
+static const struct drm_display_mode lg_lp129qe_mode = {
+	.clock = 285250,
+	.hdisplay = 2560,
+	.hsync_start = 2560 + 48,
+	.hsync_end = 2560 + 48 + 32,
+	.htotal = 2560 + 48 + 32 + 80,
+	.vdisplay = 1700,
+	.vsync_start = 1700 + 3,
+	.vsync_end = 1700 + 3 + 10,
+	.vtotal = 1700 + 3 + 10 + 36,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc lg_lp129qe = {
+	.modes = &lg_lp129qe_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 272,
+		.height = 181,
+	},
+};
+
+static const struct drm_display_mode samsung_ltn101nt05_mode = {
+	.clock = 54030,
+	.hdisplay = 1024,
+	.hsync_start = 1024 + 24,
+	.hsync_end = 1024 + 24 + 136,
+	.htotal = 1024 + 24 + 136 + 160,
+	.vdisplay = 600,
+	.vsync_start = 600 + 3,
+	.vsync_end = 600 + 3 + 6,
+	.vtotal = 600 + 3 + 6 + 61,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc samsung_ltn101nt05 = {
+	.modes = &samsung_ltn101nt05_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 1024,
+		.height = 600,
+	},
+};
+
+static const struct of_device_id platform_of_match[] = {
+	{
+		.compatible = "auo,b101aw03",
+		.data = &auo_b101aw03,
+	}, {
+		.compatible = "auo,b133xtn01",
+		.data = &auo_b133xtn01,
+	}, {
+		.compatible = "chunghwa,claa101wa01a",
+		.data = &chunghwa_claa101wa01a
+	}, {
+		.compatible = "chunghwa,claa101wb01",
+		.data = &chunghwa_claa101wb01
+	}, {
+		.compatible = "edt,et057090dhu",
+		.data = &edt_et057090dhu,
+	}, {
+		.compatible = "edt,et070080dh6",
+		.data = &edt_etm0700g0dh6,
+	}, {
+		.compatible = "edt,etm0700g0dh6",
+		.data = &edt_etm0700g0dh6,
+	}, {
+		.compatible = "lg,lp129qe",
+		.data = &lg_lp129qe,
+	}, {
+		.compatible = "samsung,ltn101nt05",
+		.data = &samsung_ltn101nt05,
+	}, {
+		.compatible = "simple-panel",
+	}, {
+		/* sentinel */
+	}
+};
+MODULE_DEVICE_TABLE(of, platform_of_match);
+
+static int panel_simple_platform_probe(struct platform_device *pdev)
+{
+	const struct of_device_id *id;
+
+	id = of_match_node(platform_of_match, pdev->dev.of_node);
+	if (!id)
+		return -ENODEV;
+
+	return panel_simple_probe(&pdev->dev, id->data);
+}
+
+static int panel_simple_platform_remove(struct platform_device *pdev)
+{
+	return panel_simple_remove(&pdev->dev);
+}
+
+static void panel_simple_platform_shutdown(struct platform_device *pdev)
+{
+	panel_simple_shutdown(&pdev->dev);
+}
+
+static struct platform_driver panel_simple_platform_driver = {
+	.driver = {
+		.name = "panel-simple",
+		.owner = THIS_MODULE,
+		.of_match_table = platform_of_match,
+	},
+	.probe = panel_simple_platform_probe,
+	.remove = panel_simple_platform_remove,
+	.shutdown = panel_simple_platform_shutdown,
+};
+
+struct panel_desc_dsi {
+	struct panel_desc desc;
+
+	unsigned long flags;
+	enum mipi_dsi_pixel_format format;
+	unsigned int lanes;
+};
+
+static const struct drm_display_mode lg_ld070wx3_sl01_mode = {
+	.clock = 71000,
+	.hdisplay = 800,
+	.hsync_start = 800 + 32,
+	.hsync_end = 800 + 32 + 1,
+	.htotal = 800 + 32 + 1 + 57,
+	.vdisplay = 1280,
+	.vsync_start = 1280 + 28,
+	.vsync_end = 1280 + 28 + 1,
+	.vtotal = 1280 + 28 + 1 + 14,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc_dsi lg_ld070wx3_sl01 = {
+	.desc = {
+		.modes = &lg_ld070wx3_sl01_mode,
+		.num_modes = 1,
+		.size = {
+			.width = 94,
+			.height = 151,
+		},
+	},
+	.flags = MIPI_DSI_MODE_VIDEO,
+	.format = MIPI_DSI_FMT_RGB888,
+	.lanes = 4,
+};
+
+static const struct drm_display_mode lg_lh500wx1_sd03_mode = {
+	.clock = 67000,
+	.hdisplay = 720,
+	.hsync_start = 720 + 12,
+	.hsync_end = 720 + 12 + 4,
+	.htotal = 720 + 12 + 4 + 112,
+	.vdisplay = 1280,
+	.vsync_start = 1280 + 8,
+	.vsync_end = 1280 + 8 + 4,
+	.vtotal = 1280 + 8 + 4 + 12,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc_dsi lg_lh500wx1_sd03 = {
+	.desc = {
+		.modes = &lg_lh500wx1_sd03_mode,
+		.num_modes = 1,
+		.size = {
+			.width = 62,
+			.height = 110,
+		},
+	},
+	.flags = MIPI_DSI_MODE_VIDEO,
+	.format = MIPI_DSI_FMT_RGB888,
+	.lanes = 4,
+};
+
+static const struct drm_display_mode panasonic_vvx10f004b00_mode = {
+	.clock = 157200,
+	.hdisplay = 1920,
+	.hsync_start = 1920 + 154,
+	.hsync_end = 1920 + 154 + 16,
+	.htotal = 1920 + 154 + 16 + 32,
+	.vdisplay = 1200,
+	.vsync_start = 1200 + 17,
+	.vsync_end = 1200 + 17 + 2,
+	.vtotal = 1200 + 17 + 2 + 16,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc_dsi panasonic_vvx10f004b00 = {
+	.desc = {
+		.modes = &panasonic_vvx10f004b00_mode,
+		.num_modes = 1,
+		.size = {
+			.width = 217,
+			.height = 136,
+		},
+	},
+	.flags = MIPI_DSI_MODE_VIDEO | MIPI_DSI_MODE_VIDEO_SYNC_PULSE,
+	.format = MIPI_DSI_FMT_RGB888,
+	.lanes = 4,
+};
+
+static const struct of_device_id dsi_of_match[] = {
+	{
+		.compatible = "lg,ld070wx3-sl01",
+		.data = &lg_ld070wx3_sl01
+	}, {
+		.compatible = "lg,lh500wx1-sd03",
+		.data = &lg_lh500wx1_sd03
+	}, {
+		.compatible = "panasonic,vvx10f004b00",
+		.data = &panasonic_vvx10f004b00
+	}, {
+		/* sentinel */
+	}
+};
+MODULE_DEVICE_TABLE(of, dsi_of_match);
+
+static int panel_simple_dsi_probe(struct mipi_dsi_device *dsi)
+{
+	const struct panel_desc_dsi *desc;
+	const struct of_device_id *id;
+	int err;
+
+	id = of_match_node(dsi_of_match, dsi->dev.of_node);
+	if (!id)
+		return -ENODEV;
+
+	desc = id->data;
+
+	err = panel_simple_probe(&dsi->dev, &desc->desc);
+	if (err < 0)
+		return err;
+
+	dsi->mode_flags = desc->flags;
+	dsi->format = desc->format;
+	dsi->lanes = desc->lanes;
+
+	return mipi_dsi_attach(dsi);
+}
+
+static int panel_simple_dsi_remove(struct mipi_dsi_device *dsi)
+{
+	int err;
+
+	err = mipi_dsi_detach(dsi);
+	if (err < 0)
+		dev_err(&dsi->dev, "failed to detach from DSI host: %d\n", err);
+
+	return panel_simple_remove(&dsi->dev);
+}
+
+static void panel_simple_dsi_shutdown(struct mipi_dsi_device *dsi)
+{
+	panel_simple_shutdown(&dsi->dev);
+}
+
+static struct mipi_dsi_driver panel_simple_dsi_driver = {
+	.driver = {
+		.name = "panel-simple-dsi",
+		.owner = THIS_MODULE,
+		.of_match_table = dsi_of_match,
+	},
+	.probe = panel_simple_dsi_probe,
+	.remove = panel_simple_dsi_remove,
+	.shutdown = panel_simple_dsi_shutdown,
+};
+
+static int __init panel_simple_init(void)
+{
+	int err;
+
+	err = platform_driver_register(&panel_simple_platform_driver);
+	if (err < 0)
+		return err;
+
+	if (IS_ENABLED(CONFIG_DRM_MIPI_DSI)) {
+		err = mipi_dsi_driver_register(&panel_simple_dsi_driver);
+		if (err < 0)
+			return err;
+	}
+
+	return 0;
+}
+module_init(panel_simple_init);
+
+static void __exit panel_simple_exit(void)
+{
+	if (IS_ENABLED(CONFIG_DRM_MIPI_DSI))
+		mipi_dsi_driver_unregister(&panel_simple_dsi_driver);
+
+	platform_driver_unregister(&panel_simple_platform_driver);
+}
+module_exit(panel_simple_exit);
+
+MODULE_AUTHOR("Thierry Reding <treding@nvidia.com>");
+MODULE_DESCRIPTION("DRM Driver for Simple Panels");
+MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c
index d70aafb8330..5d7ea246185 100644
--- a/drivers/gpu/drm/qxl/qxl_display.c
+++ b/drivers/gpu/drm/qxl/qxl_display.c
@@ -399,10 +399,14 @@ static int qxl_framebuffer_surface_dirty(struct drm_framebuffer *fb,
 	struct qxl_bo *qobj;
 	int inc = 1;
 
+	drm_modeset_lock_all(fb->dev);
+
 	qobj = gem_to_qxl_bo(qxl_fb->obj);
 	/* if we aren't primary surface ignore this */
-	if (!qobj->is_primary)
+	if (!qobj->is_primary) {
+		drm_modeset_unlock_all(fb->dev);
 		return 0;
+	}
 
 	if (!num_clips) {
 		num_clips = 1;
@@ -417,6 +421,9 @@ static int qxl_framebuffer_surface_dirty(struct drm_framebuffer *fb,
 
 	qxl_draw_dirty_fb(qdev, qxl_fb, qobj, flags, color,
 			  clips, num_clips, inc);
+
+	drm_modeset_unlock_all(fb->dev);
+
 	return 0;
 }
 
@@ -520,7 +527,7 @@ static int qxl_crtc_mode_set(struct drm_crtc *crtc,
 	bool recreate_primary = false;
 	int ret;
 	int surf_id;
-	if (!crtc->fb) {
+	if (!crtc->primary->fb) {
 		DRM_DEBUG_KMS("No FB bound\n");
 		return 0;
 	}
@@ -529,7 +536,7 @@ static int qxl_crtc_mode_set(struct drm_crtc *crtc,
 		qfb = to_qxl_framebuffer(old_fb);
 		old_bo = gem_to_qxl_bo(qfb->obj);
 	}
-	qfb = to_qxl_framebuffer(crtc->fb);
+	qfb = to_qxl_framebuffer(crtc->primary->fb);
 	bo = gem_to_qxl_bo(qfb->obj);
 	if (!m)
 		/* and do we care? */
@@ -567,6 +574,10 @@ static int qxl_crtc_mode_set(struct drm_crtc *crtc,
 			   bo->surf.height, bo->surf.stride, bo->surf.format);
 		qxl_io_create_primary(qdev, base_offset, bo);
 		bo->is_primary = true;
+	}
+
+	if (bo->is_primary) {
+		DRM_DEBUG_KMS("setting surface_id to 0 for primary surface %d on crtc %d\n", bo->surface_id, qcrtc->index);
 		surf_id = 0;
 	} else {
 		surf_id = bo->surface_id;
@@ -602,14 +613,14 @@ static void qxl_crtc_disable(struct drm_crtc *crtc)
 	struct qxl_crtc *qcrtc = to_qxl_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct qxl_device *qdev = dev->dev_private;
-	if (crtc->fb) {
-		struct qxl_framebuffer *qfb = to_qxl_framebuffer(crtc->fb);
+	if (crtc->primary->fb) {
+		struct qxl_framebuffer *qfb = to_qxl_framebuffer(crtc->primary->fb);
 		struct qxl_bo *bo = gem_to_qxl_bo(qfb->obj);
 		int ret;
 		ret = qxl_bo_reserve(bo, false);
 		qxl_bo_unpin(bo);
 		qxl_bo_unreserve(bo);
-		crtc->fb = NULL;
+		crtc->primary->fb = NULL;
 	}
 
 	qxl_monitors_config_set(qdev, qcrtc->index, 0, 0, 0, 0, 0);
@@ -834,7 +845,7 @@ static const struct drm_connector_funcs qxl_connector_funcs = {
 	.save = qxl_conn_save,
 	.restore = qxl_conn_restore,
 	.detect = qxl_conn_detect,
-	.fill_modes = drm_helper_probe_single_connector_modes,
+	.fill_modes = drm_helper_probe_single_connector_modes_nomerge,
 	.set_property = qxl_conn_set_property,
 	.destroy = qxl_conn_destroy,
 };
diff --git a/drivers/gpu/drm/qxl/qxl_drv.c b/drivers/gpu/drm/qxl/qxl_drv.c
index fee8748bdca..6e936634d65 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.c
+++ b/drivers/gpu/drm/qxl/qxl_drv.c
@@ -214,7 +214,6 @@ static struct pci_driver qxl_pci_driver = {
 static struct drm_driver qxl_driver = {
 	.driver_features = DRIVER_GEM | DRIVER_MODESET |
 			   DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED,
-	.dev_priv_size = 0,
 	.load = qxl_driver_load,
 	.unload = qxl_driver_unload,
 
diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h
index 7bda32f68d3..36ed40ba773 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.h
+++ b/drivers/gpu/drm/qxl/qxl_drv.h
@@ -534,7 +534,7 @@ void qxl_debugfs_takedown(struct drm_minor *minor);
 
 /* qxl_irq.c */
 int qxl_irq_init(struct qxl_device *qdev);
-irqreturn_t qxl_irq_handler(DRM_IRQ_ARGS);
+irqreturn_t qxl_irq_handler(int irq, void *arg);
 
 /* qxl_fb.c */
 int qxl_fb_init(struct qxl_device *qdev);
diff --git a/drivers/gpu/drm/qxl/qxl_ioctl.c b/drivers/gpu/drm/qxl/qxl_ioctl.c
index 7b95c75e962..b110883f825 100644
--- a/drivers/gpu/drm/qxl/qxl_ioctl.c
+++ b/drivers/gpu/drm/qxl/qxl_ioctl.c
@@ -200,7 +200,7 @@ static int qxl_process_single_command(struct qxl_device *qdev,
 	for (i = 0; i < cmd->relocs_num; ++i) {
 		struct drm_qxl_reloc reloc;
 
-		if (DRM_COPY_FROM_USER(&reloc,
+		if (copy_from_user(&reloc,
 				       &((struct drm_qxl_reloc *)(uintptr_t)cmd->relocs)[i],
 				       sizeof(reloc))) {
 			ret = -EFAULT;
@@ -297,7 +297,7 @@ static int qxl_execbuffer_ioctl(struct drm_device *dev, void *data,
 		struct drm_qxl_command *commands =
 			(struct drm_qxl_command *)(uintptr_t)execbuffer->commands;
 
-		if (DRM_COPY_FROM_USER(&user_cmd, &commands[cmd_num],
+		if (copy_from_user(&user_cmd, &commands[cmd_num],
 				       sizeof(user_cmd)))
 			return -EFAULT;
 
@@ -451,4 +451,4 @@ const struct drm_ioctl_desc qxl_ioctls[] = {
 			  DRM_AUTH|DRM_UNLOCKED),
 };
 
-int qxl_max_ioctls = DRM_ARRAY_SIZE(qxl_ioctls);
+int qxl_max_ioctls = ARRAY_SIZE(qxl_ioctls);
diff --git a/drivers/gpu/drm/qxl/qxl_irq.c b/drivers/gpu/drm/qxl/qxl_irq.c
index 21393dc4700..0bf1e20c6e4 100644
--- a/drivers/gpu/drm/qxl/qxl_irq.c
+++ b/drivers/gpu/drm/qxl/qxl_irq.c
@@ -25,7 +25,7 @@
 
 #include "qxl_drv.h"
 
-irqreturn_t qxl_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t qxl_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	struct qxl_device *qdev = (struct qxl_device *)dev->dev_private;
@@ -33,6 +33,9 @@ irqreturn_t qxl_irq_handler(DRM_IRQ_ARGS)
 
 	pending = xchg(&qdev->ram_header->int_pending, 0);
 
+	if (!pending)
+		return IRQ_NONE;
+
 	atomic_inc(&qdev->irq_received);
 
 	if (pending & QXL_INTERRUPT_DISPLAY) {
@@ -87,7 +90,7 @@ int qxl_irq_init(struct qxl_device *qdev)
 	atomic_set(&qdev->irq_received_cursor, 0);
 	atomic_set(&qdev->irq_received_io_cmd, 0);
 	qdev->irq_received_error = 0;
-	ret = drm_irq_install(qdev->ddev);
+	ret = drm_irq_install(qdev->ddev, qdev->ddev->pdev->irq);
 	qdev->ram_header->int_mask = QXL_INTERRUPT_MASK;
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("Failed installing irq: %d\n", ret);
diff --git a/drivers/gpu/drm/qxl/qxl_kms.c b/drivers/gpu/drm/qxl/qxl_kms.c
index e5ca498be92..fd88eb4a3f7 100644
--- a/drivers/gpu/drm/qxl/qxl_kms.c
+++ b/drivers/gpu/drm/qxl/qxl_kms.c
@@ -115,7 +115,7 @@ static void qxl_gc_work(struct work_struct *work)
 	qxl_garbage_collect(qdev);
 }
 
-int qxl_device_init(struct qxl_device *qdev,
+static int qxl_device_init(struct qxl_device *qdev,
 		    struct drm_device *ddev,
 		    struct pci_dev *pdev,
 		    unsigned long flags)
diff --git a/drivers/gpu/drm/qxl/qxl_object.c b/drivers/gpu/drm/qxl/qxl_object.c
index 8691c76c5ef..b95f144f0b4 100644
--- a/drivers/gpu/drm/qxl/qxl_object.c
+++ b/drivers/gpu/drm/qxl/qxl_object.c
@@ -82,8 +82,6 @@ int qxl_bo_create(struct qxl_device *qdev,
 	enum ttm_bo_type type;
 	int r;
 
-	if (unlikely(qdev->mman.bdev.dev_mapping == NULL))
-		qdev->mman.bdev.dev_mapping = qdev->ddev->dev_mapping;
 	if (kernel)
 		type = ttm_bo_type_kernel;
 	else
diff --git a/drivers/gpu/drm/qxl/qxl_release.c b/drivers/gpu/drm/qxl/qxl_release.c
index 821ab7b9409..14e776f1d14 100644
--- a/drivers/gpu/drm/qxl/qxl_release.c
+++ b/drivers/gpu/drm/qxl/qxl_release.c
@@ -349,7 +349,7 @@ void qxl_release_fence_buffer_objects(struct qxl_release *release)
 		qxl_fence_add_release_locked(&qbo->fence, release->id);
 
 		ttm_bo_add_to_lru(bo);
-		ww_mutex_unlock(&bo->resv->lock);
+		__ttm_bo_unreserve(bo);
 		entry->reserved = false;
 	}
 	spin_unlock(&bdev->fence_lock);
diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c
index c7e7e6590c2..71a1baeac14 100644
--- a/drivers/gpu/drm/qxl/qxl_ttm.c
+++ b/drivers/gpu/drm/qxl/qxl_ttm.c
@@ -109,13 +109,11 @@ static const struct vm_operations_struct *ttm_vm_ops;
 static int qxl_ttm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 {
 	struct ttm_buffer_object *bo;
-	struct qxl_device *qdev;
 	int r;
 
 	bo = (struct ttm_buffer_object *)vma->vm_private_data;
 	if (bo == NULL)
 		return VM_FAULT_NOPAGE;
-	qdev = qxl_get_qdev(bo->bdev);
 	r = ttm_vm_ops->fault(vma, vmf);
 	return r;
 }
@@ -162,10 +160,6 @@ static int qxl_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags)
 static int qxl_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 			     struct ttm_mem_type_manager *man)
 {
-	struct qxl_device *qdev;
-
-	qdev = qxl_get_qdev(bdev);
-
 	switch (type) {
 	case TTM_PL_SYSTEM:
 		/* System memory */
@@ -433,6 +427,7 @@ static int qxl_sync_obj_flush(void *sync_obj)
 
 static void qxl_sync_obj_unref(void **sync_obj)
 {
+	*sync_obj = NULL;
 }
 
 static void *qxl_sync_obj_ref(void *sync_obj)
@@ -493,7 +488,9 @@ int qxl_ttm_init(struct qxl_device *qdev)
 	/* No others user of address space so set it to 0 */
 	r = ttm_bo_device_init(&qdev->mman.bdev,
 			       qdev->mman.bo_global_ref.ref.object,
-			       &qxl_bo_driver, DRM_FILE_PAGE_OFFSET, 0);
+			       &qxl_bo_driver,
+			       qdev->ddev->anon_inode->i_mapping,
+			       DRM_FILE_PAGE_OFFSET, 0);
 	if (r) {
 		DRM_ERROR("failed initializing buffer object driver(%d).\n", r);
 		return r;
@@ -518,8 +515,6 @@ int qxl_ttm_init(struct qxl_device *qdev)
 		 ((unsigned)num_io_pages * PAGE_SIZE) / (1024 * 1024));
 	DRM_INFO("qxl: %uM of Surface memory size\n",
 		 (unsigned)qdev->surfaceram_size / (1024 * 1024));
-	if (unlikely(qdev->mman.bdev.dev_mapping == NULL))
-		qdev->mman.bdev.dev_mapping = qdev->ddev->dev_mapping;
 	r = qxl_ttm_debugfs_init(qdev);
 	if (r) {
 		DRM_ERROR("Failed to init debugfs\n");
diff --git a/drivers/gpu/drm/r128/r128_cce.c b/drivers/gpu/drm/r128/r128_cce.c
index c451257f08f..59459fe4e8c 100644
--- a/drivers/gpu/drm/r128/r128_cce.c
+++ b/drivers/gpu/drm/r128/r128_cce.c
@@ -892,10 +892,10 @@ static int r128_cce_get_buffers(struct drm_device *dev,
 
 		buf->file_priv = file_priv;
 
-		if (DRM_COPY_TO_USER(&d->request_indices[i], &buf->idx,
+		if (copy_to_user(&d->request_indices[i], &buf->idx,
 				     sizeof(buf->idx)))
 			return -EFAULT;
-		if (DRM_COPY_TO_USER(&d->request_sizes[i], &buf->total,
+		if (copy_to_user(&d->request_sizes[i], &buf->total,
 				     sizeof(buf->total)))
 			return -EFAULT;
 
diff --git a/drivers/gpu/drm/r128/r128_drv.h b/drivers/gpu/drm/r128/r128_drv.h
index 56eb5e3f543..5bf3f5ff805 100644
--- a/drivers/gpu/drm/r128/r128_drv.h
+++ b/drivers/gpu/drm/r128/r128_drv.h
@@ -154,7 +154,7 @@ extern int r128_do_cleanup_cce(struct drm_device *dev);
 extern int r128_enable_vblank(struct drm_device *dev, int crtc);
 extern void r128_disable_vblank(struct drm_device *dev, int crtc);
 extern u32 r128_get_vblank_counter(struct drm_device *dev, int crtc);
-extern irqreturn_t r128_driver_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t r128_driver_irq_handler(int irq, void *arg);
 extern void r128_driver_irq_preinstall(struct drm_device *dev);
 extern int r128_driver_irq_postinstall(struct drm_device *dev);
 extern void r128_driver_irq_uninstall(struct drm_device *dev);
@@ -514,7 +514,7 @@ do {									\
 	if (R128_VERBOSE)						\
 		DRM_INFO("COMMIT_RING() tail=0x%06x\n",			\
 			 dev_priv->ring.tail);				\
-	DRM_MEMORYBARRIER();						\
+	mb();						\
 	R128_WRITE(R128_PM4_BUFFER_DL_WPTR, dev_priv->ring.tail);	\
 	R128_READ(R128_PM4_BUFFER_DL_WPTR);				\
 } while (0)
diff --git a/drivers/gpu/drm/r128/r128_ioc32.c b/drivers/gpu/drm/r128/r128_ioc32.c
index a954c548201..663f38c63ba 100644
--- a/drivers/gpu/drm/r128/r128_ioc32.c
+++ b/drivers/gpu/drm/r128/r128_ioc32.c
@@ -33,6 +33,7 @@
 
 #include <drm/drmP.h>
 #include <drm/r128_drm.h>
+#include "r128_drv.h"
 
 typedef struct drm_r128_init32 {
 	int func;
@@ -202,7 +203,7 @@ long r128_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	if (nr < DRM_COMMAND_BASE)
 		return drm_compat_ioctl(filp, cmd, arg);
 
-	if (nr < DRM_COMMAND_BASE + DRM_ARRAY_SIZE(r128_compat_ioctls))
+	if (nr < DRM_COMMAND_BASE + ARRAY_SIZE(r128_compat_ioctls))
 		fn = r128_compat_ioctls[nr - DRM_COMMAND_BASE];
 
 	if (fn != NULL)
diff --git a/drivers/gpu/drm/r128/r128_irq.c b/drivers/gpu/drm/r128/r128_irq.c
index 2ea4f09d269..c2ae496babb 100644
--- a/drivers/gpu/drm/r128/r128_irq.c
+++ b/drivers/gpu/drm/r128/r128_irq.c
@@ -44,7 +44,7 @@ u32 r128_get_vblank_counter(struct drm_device *dev, int crtc)
 	return atomic_read(&dev_priv->vbl_received);
 }
 
-irqreturn_t r128_driver_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t r128_driver_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	drm_r128_private_t *dev_priv = (drm_r128_private_t *) dev->dev_private;
diff --git a/drivers/gpu/drm/r128/r128_state.c b/drivers/gpu/drm/r128/r128_state.c
index 01dd9aef9f0..575e986f82a 100644
--- a/drivers/gpu/drm/r128/r128_state.c
+++ b/drivers/gpu/drm/r128/r128_state.c
@@ -895,31 +895,22 @@ static int r128_cce_dispatch_write_span(struct drm_device *dev,
 	if (count > 4096 || count <= 0)
 		return -EMSGSIZE;
 
-	if (DRM_COPY_FROM_USER(&x, depth->x, sizeof(x)))
+	if (copy_from_user(&x, depth->x, sizeof(x)))
 		return -EFAULT;
-	if (DRM_COPY_FROM_USER(&y, depth->y, sizeof(y)))
+	if (copy_from_user(&y, depth->y, sizeof(y)))
 		return -EFAULT;
 
 	buffer_size = depth->n * sizeof(u32);
-	buffer = kmalloc(buffer_size, GFP_KERNEL);
-	if (buffer == NULL)
-		return -ENOMEM;
-	if (DRM_COPY_FROM_USER(buffer, depth->buffer, buffer_size)) {
-		kfree(buffer);
-		return -EFAULT;
-	}
+	buffer = memdup_user(depth->buffer, buffer_size);
+	if (IS_ERR(buffer))
+		return PTR_ERR(buffer);
 
 	mask_size = depth->n * sizeof(u8);
 	if (depth->mask) {
-		mask = kmalloc(mask_size, GFP_KERNEL);
-		if (mask == NULL) {
+		mask = memdup_user(depth->mask, mask_size);
+		if (IS_ERR(mask)) {
 			kfree(buffer);
-			return -ENOMEM;
-		}
-		if (DRM_COPY_FROM_USER(mask, depth->mask, mask_size)) {
-			kfree(buffer);
-			kfree(mask);
-			return -EFAULT;
+			return PTR_ERR(mask);
 		}
 
 		for (i = 0; i < count; i++, x++) {
@@ -999,46 +990,33 @@ static int r128_cce_dispatch_write_pixels(struct drm_device *dev,
 		kfree(x);
 		return -ENOMEM;
 	}
-	if (DRM_COPY_FROM_USER(x, depth->x, xbuf_size)) {
+	if (copy_from_user(x, depth->x, xbuf_size)) {
 		kfree(x);
 		kfree(y);
 		return -EFAULT;
 	}
-	if (DRM_COPY_FROM_USER(y, depth->y, xbuf_size)) {
+	if (copy_from_user(y, depth->y, xbuf_size)) {
 		kfree(x);
 		kfree(y);
 		return -EFAULT;
 	}
 
 	buffer_size = depth->n * sizeof(u32);
-	buffer = kmalloc(buffer_size, GFP_KERNEL);
-	if (buffer == NULL) {
-		kfree(x);
-		kfree(y);
-		return -ENOMEM;
-	}
-	if (DRM_COPY_FROM_USER(buffer, depth->buffer, buffer_size)) {
+	buffer = memdup_user(depth->buffer, buffer_size);
+	if (IS_ERR(buffer)) {
 		kfree(x);
 		kfree(y);
-		kfree(buffer);
-		return -EFAULT;
+		return PTR_ERR(buffer);
 	}
 
 	if (depth->mask) {
 		mask_size = depth->n * sizeof(u8);
-		mask = kmalloc(mask_size, GFP_KERNEL);
-		if (mask == NULL) {
-			kfree(x);
-			kfree(y);
-			kfree(buffer);
-			return -ENOMEM;
-		}
-		if (DRM_COPY_FROM_USER(mask, depth->mask, mask_size)) {
+		mask = memdup_user(depth->mask, mask_size);
+		if (IS_ERR(mask)) {
 			kfree(x);
 			kfree(y);
 			kfree(buffer);
-			kfree(mask);
-			return -EFAULT;
+			return PTR_ERR(mask);
 		}
 
 		for (i = 0; i < count; i++) {
@@ -1107,9 +1085,9 @@ static int r128_cce_dispatch_read_span(struct drm_device *dev,
 	if (count > 4096 || count <= 0)
 		return -EMSGSIZE;
 
-	if (DRM_COPY_FROM_USER(&x, depth->x, sizeof(x)))
+	if (copy_from_user(&x, depth->x, sizeof(x)))
 		return -EFAULT;
-	if (DRM_COPY_FROM_USER(&y, depth->y, sizeof(y)))
+	if (copy_from_user(&y, depth->y, sizeof(y)))
 		return -EFAULT;
 
 	BEGIN_RING(7);
@@ -1162,12 +1140,12 @@ static int r128_cce_dispatch_read_pixels(struct drm_device *dev,
 		kfree(x);
 		return -ENOMEM;
 	}
-	if (DRM_COPY_FROM_USER(x, depth->x, xbuf_size)) {
+	if (copy_from_user(x, depth->x, xbuf_size)) {
 		kfree(x);
 		kfree(y);
 		return -EFAULT;
 	}
-	if (DRM_COPY_FROM_USER(y, depth->y, ybuf_size)) {
+	if (copy_from_user(y, depth->y, ybuf_size)) {
 		kfree(x);
 		kfree(y);
 		return -EFAULT;
@@ -1524,7 +1502,7 @@ static int r128_cce_stipple(struct drm_device *dev, void *data, struct drm_file
 
 	DEV_INIT_TEST_WITH_RETURN(dev_priv);
 
-	if (DRM_COPY_FROM_USER(&mask, stipple->mask, 32 * sizeof(u32)))
+	if (copy_from_user(&mask, stipple->mask, 32 * sizeof(u32)))
 		return -EFAULT;
 
 	RING_SPACE_TEST_WITH_RETURN(dev_priv);
@@ -1616,13 +1594,13 @@ static int r128_getparam(struct drm_device *dev, void *data, struct drm_file *fi
 
 	switch (param->param) {
 	case R128_PARAM_IRQ_NR:
-		value = drm_dev_to_irq(dev);
+		value = dev->pdev->irq;
 		break;
 	default:
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_TO_USER(param->value, &value, sizeof(int))) {
+	if (copy_to_user(param->value, &value, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
@@ -1663,4 +1641,4 @@ const struct drm_ioctl_desc r128_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(R128_GETPARAM, r128_getparam, DRM_AUTH),
 };
 
-int r128_max_ioctl = DRM_ARRAY_SIZE(r128_ioctls);
+int r128_max_ioctl = ARRAY_SIZE(r128_ioctls);
diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile
index 306364a1ecd..dbcbfe80aac 100644
--- a/drivers/gpu/drm/radeon/Makefile
+++ b/drivers/gpu/drm/radeon/Makefile
@@ -72,7 +72,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \
 	radeon_cs.o radeon_bios.o radeon_benchmark.o r100.o r300.o r420.o \
 	rs400.o rs600.o rs690.o rv515.o r520.o r600.o rv770.o radeon_test.o \
 	r200.o radeon_legacy_tv.o r600_cs.o r600_blit_shaders.o \
-	radeon_pm.o atombios_dp.o r600_audio.o r600_hdmi.o \
+	radeon_pm.o atombios_dp.o r600_audio.o r600_hdmi.o dce3_1_afmt.o \
 	evergreen.o evergreen_cs.o evergreen_blit_shaders.o \
 	evergreen_hdmi.o radeon_trace_points.o ni.o cayman_blit_shaders.o \
 	atombios_encoders.o radeon_semaphore.o radeon_sa.o atombios_i2c.o si.o \
@@ -80,7 +80,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \
 	r600_dpm.o rs780_dpm.o rv6xx_dpm.o rv770_dpm.o rv730_dpm.o rv740_dpm.o \
 	rv770_smc.o cypress_dpm.o btc_dpm.o sumo_dpm.o sumo_smc.o trinity_dpm.o \
 	trinity_smc.o ni_dpm.o si_smc.o si_dpm.o kv_smc.o kv_dpm.o ci_smc.o \
-	ci_dpm.o dce6_afmt.o
+	ci_dpm.o dce6_afmt.o radeon_vm.o
 
 # add async DMA block
 radeon-y += \
@@ -99,6 +99,12 @@ radeon-y += \
 	uvd_v3_1.o \
 	uvd_v4_2.o
 
+# add VCE block
+radeon-y += \
+	radeon_vce.o \
+	vce_v1_0.o \
+	vce_v2_0.o \
+
 radeon-$(CONFIG_COMPAT) += radeon_ioc32.o
 radeon-$(CONFIG_VGA_SWITCHEROO) += radeon_atpx_handler.o
 radeon-$(CONFIG_ACPI) += radeon_acpi.o
diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c
index 0b9621c9aee..30d242b2507 100644
--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -209,6 +209,16 @@ static void atombios_enable_crtc_memreq(struct drm_crtc *crtc, int state)
 	atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
 }
 
+static const u32 vga_control_regs[6] =
+{
+	AVIVO_D1VGA_CONTROL,
+	AVIVO_D2VGA_CONTROL,
+	EVERGREEN_D3VGA_CONTROL,
+	EVERGREEN_D4VGA_CONTROL,
+	EVERGREEN_D5VGA_CONTROL,
+	EVERGREEN_D6VGA_CONTROL,
+};
+
 static void atombios_blank_crtc(struct drm_crtc *crtc, int state)
 {
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
@@ -216,13 +226,23 @@ static void atombios_blank_crtc(struct drm_crtc *crtc, int state)
 	struct radeon_device *rdev = dev->dev_private;
 	int index = GetIndexIntoMasterTable(COMMAND, BlankCRTC);
 	BLANK_CRTC_PS_ALLOCATION args;
+	u32 vga_control = 0;
 
 	memset(&args, 0, sizeof(args));
 
+	if (ASIC_IS_DCE8(rdev)) {
+		vga_control = RREG32(vga_control_regs[radeon_crtc->crtc_id]);
+		WREG32(vga_control_regs[radeon_crtc->crtc_id], vga_control | 1);
+	}
+
 	args.ucCRTC = radeon_crtc->crtc_id;
 	args.ucBlanking = state;
 
 	atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+
+	if (ASIC_IS_DCE8(rdev)) {
+		WREG32(vga_control_regs[radeon_crtc->crtc_id], vga_control);
+	}
 }
 
 static void atombios_powergate_crtc(struct drm_crtc *crtc, int state)
@@ -250,8 +270,6 @@ void atombios_crtc_dpms(struct drm_crtc *crtc, int mode)
 	switch (mode) {
 	case DRM_MODE_DPMS_ON:
 		radeon_crtc->enabled = true;
-		/* adjust pm to dpms changes BEFORE enabling crtcs */
-		radeon_pm_compute_clocks(rdev);
 		atombios_enable_crtc(crtc, ATOM_ENABLE);
 		if (ASIC_IS_DCE3(rdev) && !ASIC_IS_DCE6(rdev))
 			atombios_enable_crtc_memreq(crtc, ATOM_ENABLE);
@@ -269,10 +287,10 @@ void atombios_crtc_dpms(struct drm_crtc *crtc, int mode)
 			atombios_enable_crtc_memreq(crtc, ATOM_DISABLE);
 		atombios_enable_crtc(crtc, ATOM_DISABLE);
 		radeon_crtc->enabled = false;
-		/* adjust pm to dpms changes AFTER disabling crtcs */
-		radeon_pm_compute_clocks(rdev);
 		break;
 	}
+	/* adjust pm to dpms */
+	radeon_pm_compute_clocks(rdev);
 }
 
 static void
@@ -423,7 +441,17 @@ static void atombios_crtc_program_ss(struct radeon_device *rdev,
 	int index = GetIndexIntoMasterTable(COMMAND, EnableSpreadSpectrumOnPPLL);
 	union atom_enable_ss args;
 
-	if (!enable) {
+	if (enable) {
+		/* Don't mess with SS if percentage is 0 or external ss.
+		 * SS is already disabled previously, and disabling it
+		 * again can cause display problems if the pll is already
+		 * programmed.
+		 */
+		if (ss->percentage == 0)
+			return;
+		if (ss->type & ATOM_EXTERNAL_SS_MASK)
+			return;
+	} else {
 		for (i = 0; i < rdev->num_crtc; i++) {
 			if (rdev->mode_info.crtcs[i] &&
 			    rdev->mode_info.crtcs[i]->enabled &&
@@ -459,8 +487,6 @@ static void atombios_crtc_program_ss(struct radeon_device *rdev,
 		args.v3.usSpreadSpectrumAmount = cpu_to_le16(ss->amount);
 		args.v3.usSpreadSpectrumStep = cpu_to_le16(ss->step);
 		args.v3.ucEnable = enable;
-		if ((ss->percentage == 0) || (ss->type & ATOM_EXTERNAL_SS_MASK) || ASIC_IS_DCE61(rdev))
-			args.v3.ucEnable = ATOM_DISABLE;
 	} else if (ASIC_IS_DCE4(rdev)) {
 		args.v2.usSpreadSpectrumPercentage = cpu_to_le16(ss->percentage);
 		args.v2.ucSpreadSpectrumType = ss->type & ATOM_SS_CENTRE_SPREAD_MODE_MASK;
@@ -480,8 +506,6 @@ static void atombios_crtc_program_ss(struct radeon_device *rdev,
 		args.v2.usSpreadSpectrumAmount = cpu_to_le16(ss->amount);
 		args.v2.usSpreadSpectrumStep = cpu_to_le16(ss->step);
 		args.v2.ucEnable = enable;
-		if ((ss->percentage == 0) || (ss->type & ATOM_EXTERNAL_SS_MASK) || ASIC_IS_DCE41(rdev))
-			args.v2.ucEnable = ATOM_DISABLE;
 	} else if (ASIC_IS_DCE3(rdev)) {
 		args.v1.usSpreadSpectrumPercentage = cpu_to_le16(ss->percentage);
 		args.v1.ucSpreadSpectrumType = ss->type & ATOM_SS_CENTRE_SPREAD_MODE_MASK;
@@ -503,8 +527,7 @@ static void atombios_crtc_program_ss(struct radeon_device *rdev,
 		args.lvds_ss_2.ucSpreadSpectrumRange = ss->range;
 		args.lvds_ss_2.ucEnable = enable;
 	} else {
-		if ((enable == ATOM_DISABLE) || (ss->percentage == 0) ||
-		    (ss->type & ATOM_EXTERNAL_SS_MASK)) {
+		if (enable == ATOM_DISABLE) {
 			atombios_disable_ss(rdev, pll_id);
 			return;
 		}
@@ -534,7 +557,8 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc,
 	u32 adjusted_clock = mode->clock;
 	int encoder_mode = atombios_get_encoder_mode(encoder);
 	u32 dp_clock = mode->clock;
-	int bpc = radeon_get_monitor_bpc(connector);
+	u32 clock = mode->clock;
+	int bpc = radeon_crtc->bpc;
 	bool is_duallink = radeon_dig_monitor_is_duallink(encoder, mode->clock);
 
 	/* reset the pll flags */
@@ -609,6 +633,24 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc,
 			radeon_crtc->pll_flags |= RADEON_PLL_USE_REF_DIV;
 	}
 
+	/* adjust pll for deep color modes */
+	if (encoder_mode == ATOM_ENCODER_MODE_HDMI) {
+		switch (bpc) {
+		case 8:
+		default:
+			break;
+		case 10:
+			clock = (clock * 5) / 4;
+			break;
+		case 12:
+			clock = (clock * 3) / 2;
+			break;
+		case 16:
+			clock = clock * 2;
+			break;
+		}
+	}
+
 	/* DCE3+ has an AdjustDisplayPll that will adjust the pixel clock
 	 * accordingly based on the encoder/transmitter to work around
 	 * special hw requirements.
@@ -630,7 +672,7 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc,
 			switch (crev) {
 			case 1:
 			case 2:
-				args.v1.usPixelClock = cpu_to_le16(mode->clock / 10);
+				args.v1.usPixelClock = cpu_to_le16(clock / 10);
 				args.v1.ucTransmitterID = radeon_encoder->encoder_id;
 				args.v1.ucEncodeMode = encoder_mode;
 				if (radeon_crtc->ss_enabled && radeon_crtc->ss.percentage)
@@ -642,7 +684,7 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc,
 				adjusted_clock = le16_to_cpu(args.v1.usPixelClock) * 10;
 				break;
 			case 3:
-				args.v3.sInput.usPixelClock = cpu_to_le16(mode->clock / 10);
+				args.v3.sInput.usPixelClock = cpu_to_le16(clock / 10);
 				args.v3.sInput.ucTransmitterID = radeon_encoder->encoder_id;
 				args.v3.sInput.ucEncodeMode = encoder_mode;
 				args.v3.sInput.ucDispPllConfig = 0;
@@ -656,10 +698,6 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc,
 					args.v3.sInput.usPixelClock = cpu_to_le16(dp_clock / 10);
 				} else if (radeon_encoder->devices & (ATOM_DEVICE_DFP_SUPPORT)) {
 					struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv;
-					if (encoder_mode == ATOM_ENCODER_MODE_HDMI)
-						/* deep color support */
-						args.v3.sInput.usPixelClock =
-							cpu_to_le16((mode->clock * bpc / 8) / 10);
 					if (dig->coherent_mode)
 						args.v3.sInput.ucDispPllConfig |=
 							DISPPLL_CONFIG_COHERENT_MODE;
@@ -839,14 +877,21 @@ static void atombios_crtc_program_pll(struct drm_crtc *crtc,
 			args.v5.ucMiscInfo = 0; /* HDMI depth, etc. */
 			if (ss_enabled && (ss->type & ATOM_EXTERNAL_SS_MASK))
 				args.v5.ucMiscInfo |= PIXEL_CLOCK_V5_MISC_REF_DIV_SRC;
-			switch (bpc) {
-			case 8:
-			default:
-				args.v5.ucMiscInfo |= PIXEL_CLOCK_V5_MISC_HDMI_24BPP;
-				break;
-			case 10:
-				args.v5.ucMiscInfo |= PIXEL_CLOCK_V5_MISC_HDMI_30BPP;
-				break;
+			if (encoder_mode == ATOM_ENCODER_MODE_HDMI) {
+				switch (bpc) {
+				case 8:
+				default:
+					args.v5.ucMiscInfo |= PIXEL_CLOCK_V5_MISC_HDMI_24BPP;
+					break;
+				case 10:
+					/* yes this is correct, the atom define is wrong */
+					args.v5.ucMiscInfo |= PIXEL_CLOCK_V5_MISC_HDMI_32BPP;
+					break;
+				case 12:
+					/* yes this is correct, the atom define is wrong */
+					args.v5.ucMiscInfo |= PIXEL_CLOCK_V5_MISC_HDMI_30BPP;
+					break;
+				}
 			}
 			args.v5.ucTransmitterID = encoder_id;
 			args.v5.ucEncoderMode = encoder_mode;
@@ -861,20 +906,22 @@ static void atombios_crtc_program_pll(struct drm_crtc *crtc,
 			args.v6.ucMiscInfo = 0; /* HDMI depth, etc. */
 			if (ss_enabled && (ss->type & ATOM_EXTERNAL_SS_MASK))
 				args.v6.ucMiscInfo |= PIXEL_CLOCK_V6_MISC_REF_DIV_SRC;
-			switch (bpc) {
-			case 8:
-			default:
-				args.v6.ucMiscInfo |= PIXEL_CLOCK_V6_MISC_HDMI_24BPP;
-				break;
-			case 10:
-				args.v6.ucMiscInfo |= PIXEL_CLOCK_V6_MISC_HDMI_30BPP;
-				break;
-			case 12:
-				args.v6.ucMiscInfo |= PIXEL_CLOCK_V6_MISC_HDMI_36BPP;
-				break;
-			case 16:
-				args.v6.ucMiscInfo |= PIXEL_CLOCK_V6_MISC_HDMI_48BPP;
-				break;
+			if (encoder_mode == ATOM_ENCODER_MODE_HDMI) {
+				switch (bpc) {
+				case 8:
+				default:
+					args.v6.ucMiscInfo |= PIXEL_CLOCK_V6_MISC_HDMI_24BPP;
+					break;
+				case 10:
+					args.v6.ucMiscInfo |= PIXEL_CLOCK_V6_MISC_HDMI_30BPP_V6;
+					break;
+				case 12:
+					args.v6.ucMiscInfo |= PIXEL_CLOCK_V6_MISC_HDMI_36BPP_V6;
+					break;
+				case 16:
+					args.v6.ucMiscInfo |= PIXEL_CLOCK_V6_MISC_HDMI_48BPP;
+					break;
+				}
 			}
 			args.v6.ucTransmitterID = encoder_id;
 			args.v6.ucEncoderMode = encoder_mode;
@@ -915,6 +962,9 @@ static bool atombios_crtc_prepare_pll(struct drm_crtc *crtc, struct drm_display_
 		struct radeon_connector_atom_dig *dig_connector =
 			radeon_connector->con_priv;
 		int dp_clock;
+
+		/* Assign mode clock for hdmi deep color max clock limit check */
+		radeon_connector->pixelclock_for_modeset = mode->clock;
 		radeon_crtc->bpc = radeon_get_monitor_bpc(connector);
 
 		switch (encoder_mode) {
@@ -938,11 +988,14 @@ static bool atombios_crtc_prepare_pll(struct drm_crtc *crtc, struct drm_display_
 							radeon_atombios_get_ppll_ss_info(rdev,
 											 &radeon_crtc->ss,
 											 ATOM_DP_SS_ID1);
-				} else
+				} else {
 					radeon_crtc->ss_enabled =
 						radeon_atombios_get_ppll_ss_info(rdev,
 										 &radeon_crtc->ss,
 										 ATOM_DP_SS_ID1);
+				}
+				/* disable spread spectrum on DCE3 DP */
+				radeon_crtc->ss_enabled = false;
 			}
 			break;
 		case ATOM_ENCODER_MODE_LVDS:
@@ -993,10 +1046,17 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode
 	struct radeon_encoder *radeon_encoder =
 		to_radeon_encoder(radeon_crtc->encoder);
 	u32 pll_clock = mode->clock;
+	u32 clock = mode->clock;
 	u32 ref_div = 0, fb_div = 0, frac_fb_div = 0, post_div = 0;
 	struct radeon_pll *pll;
 	int encoder_mode = atombios_get_encoder_mode(radeon_crtc->encoder);
 
+	/* pass the actual clock to atombios_crtc_program_pll for DCE5,6 for HDMI */
+	if (ASIC_IS_DCE5(rdev) &&
+	    (encoder_mode == ATOM_ENCODER_MODE_HDMI) &&
+	    (radeon_crtc->bpc > 8))
+		clock = radeon_crtc->adjusted_clock;
+
 	switch (radeon_crtc->pll_id) {
 	case ATOM_PPLL1:
 		pll = &rdev->clock.p1pll;
@@ -1031,7 +1091,7 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode
 				 radeon_crtc->crtc_id, &radeon_crtc->ss);
 
 	atombios_crtc_program_pll(crtc, radeon_crtc->crtc_id, radeon_crtc->pll_id,
-				  encoder_mode, radeon_encoder->encoder_id, mode->clock,
+				  encoder_mode, radeon_encoder->encoder_id, clock,
 				  ref_div, fb_div, frac_fb_div, post_div,
 				  radeon_crtc->bpc, radeon_crtc->ss_enabled, &radeon_crtc->ss);
 
@@ -1039,15 +1099,17 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode
 		/* calculate ss amount and step size */
 		if (ASIC_IS_DCE4(rdev)) {
 			u32 step_size;
-			u32 amount = (((fb_div * 10) + frac_fb_div) * radeon_crtc->ss.percentage) / 10000;
+			u32 amount = (((fb_div * 10) + frac_fb_div) *
+				      (u32)radeon_crtc->ss.percentage) /
+				(100 * (u32)radeon_crtc->ss.percentage_divider);
 			radeon_crtc->ss.amount = (amount / 10) & ATOM_PPLL_SS_AMOUNT_V2_FBDIV_MASK;
 			radeon_crtc->ss.amount |= ((amount - (amount / 10)) << ATOM_PPLL_SS_AMOUNT_V2_NFRAC_SHIFT) &
 				ATOM_PPLL_SS_AMOUNT_V2_NFRAC_MASK;
 			if (radeon_crtc->ss.type & ATOM_PPLL_SS_TYPE_V2_CENTRE_SPREAD)
-				step_size = (4 * amount * ref_div * (radeon_crtc->ss.rate * 2048)) /
+				step_size = (4 * amount * ref_div * ((u32)radeon_crtc->ss.rate * 2048)) /
 					(125 * 25 * pll->reference_freq / 100);
 			else
-				step_size = (2 * amount * ref_div * (radeon_crtc->ss.rate * 2048)) /
+				step_size = (2 * amount * ref_div * ((u32)radeon_crtc->ss.rate * 2048)) /
 					(125 * 25 * pll->reference_freq / 100);
 			radeon_crtc->ss.step = step_size;
 		}
@@ -1074,9 +1136,10 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	u32 fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_NONE);
 	u32 tmp, viewport_w, viewport_h;
 	int r;
+	bool bypass_lut = false;
 
 	/* no fb bound */
-	if (!atomic && !crtc->fb) {
+	if (!atomic && !crtc->primary->fb) {
 		DRM_DEBUG_KMS("No FB bound\n");
 		return 0;
 	}
@@ -1086,8 +1149,8 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 		target_fb = fb;
 	}
 	else {
-		radeon_fb = to_radeon_framebuffer(crtc->fb);
-		target_fb = crtc->fb;
+		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+		target_fb = crtc->primary->fb;
 	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
@@ -1112,33 +1175,73 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	radeon_bo_get_tiling_flags(rbo, &tiling_flags, NULL);
 	radeon_bo_unreserve(rbo);
 
-	switch (target_fb->bits_per_pixel) {
-	case 8:
+	switch (target_fb->pixel_format) {
+	case DRM_FORMAT_C8:
 		fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_8BPP) |
 			     EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_INDEXED));
 		break;
-	case 15:
+	case DRM_FORMAT_XRGB4444:
+	case DRM_FORMAT_ARGB4444:
+		fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) |
+			     EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB4444));
+#ifdef __BIG_ENDIAN
+		fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16);
+#endif
+		break;
+	case DRM_FORMAT_XRGB1555:
+	case DRM_FORMAT_ARGB1555:
 		fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) |
 			     EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB1555));
+#ifdef __BIG_ENDIAN
+		fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16);
+#endif
+		break;
+	case DRM_FORMAT_BGRX5551:
+	case DRM_FORMAT_BGRA5551:
+		fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) |
+			     EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_BGRA5551));
+#ifdef __BIG_ENDIAN
+		fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16);
+#endif
 		break;
-	case 16:
+	case DRM_FORMAT_RGB565:
 		fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) |
 			     EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB565));
 #ifdef __BIG_ENDIAN
 		fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16);
 #endif
 		break;
-	case 24:
-	case 32:
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_ARGB8888:
 		fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) |
 			     EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB8888));
 #ifdef __BIG_ENDIAN
 		fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32);
 #endif
 		break;
+	case DRM_FORMAT_XRGB2101010:
+	case DRM_FORMAT_ARGB2101010:
+		fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) |
+			     EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB2101010));
+#ifdef __BIG_ENDIAN
+		fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32);
+#endif
+		/* Greater 8 bpc fb needs to bypass hw-lut to retain precision */
+		bypass_lut = true;
+		break;
+	case DRM_FORMAT_BGRX1010102:
+	case DRM_FORMAT_BGRA1010102:
+		fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) |
+			     EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_BGRA1010102));
+#ifdef __BIG_ENDIAN
+		fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32);
+#endif
+		/* Greater 8 bpc fb needs to bypass hw-lut to retain precision */
+		bypass_lut = true;
+		break;
 	default:
-		DRM_ERROR("Unsupported screen depth %d\n",
-			  target_fb->bits_per_pixel);
+		DRM_ERROR("Unsupported screen format %s\n",
+			  drm_get_format_name(target_fb->pixel_format));
 		return -EINVAL;
 	}
 
@@ -1146,31 +1249,48 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 		evergreen_tiling_fields(tiling_flags, &bankw, &bankh, &mtaspect, &tile_split);
 
 		/* Set NUM_BANKS. */
-		if (rdev->family >= CHIP_BONAIRE) {
-			unsigned tileb, index, num_banks, tile_split_bytes;
+		if (rdev->family >= CHIP_TAHITI) {
+			unsigned index, num_banks;
 
-			/* Calculate the macrotile mode index. */
-			tile_split_bytes = 64 << tile_split;
-			tileb = 8 * 8 * target_fb->bits_per_pixel / 8;
-			tileb = min(tile_split_bytes, tileb);
+			if (rdev->family >= CHIP_BONAIRE) {
+				unsigned tileb, tile_split_bytes;
 
-			for (index = 0; tileb > 64; index++) {
-				tileb >>= 1;
-			}
+				/* Calculate the macrotile mode index. */
+				tile_split_bytes = 64 << tile_split;
+				tileb = 8 * 8 * target_fb->bits_per_pixel / 8;
+				tileb = min(tile_split_bytes, tileb);
+
+				for (index = 0; tileb > 64; index++)
+					tileb >>= 1;
+
+				if (index >= 16) {
+					DRM_ERROR("Wrong screen bpp (%u) or tile split (%u)\n",
+						  target_fb->bits_per_pixel, tile_split);
+					return -EINVAL;
+				}
 
-			if (index >= 16) {
-				DRM_ERROR("Wrong screen bpp (%u) or tile split (%u)\n",
-					  target_fb->bits_per_pixel, tile_split);
-				return -EINVAL;
+				num_banks = (rdev->config.cik.macrotile_mode_array[index] >> 6) & 0x3;
+			} else {
+				switch (target_fb->bits_per_pixel) {
+				case 8:
+					index = 10;
+					break;
+				case 16:
+					index = SI_TILE_MODE_COLOR_2D_SCANOUT_16BPP;
+					break;
+				default:
+				case 32:
+					index = SI_TILE_MODE_COLOR_2D_SCANOUT_32BPP;
+					break;
+				}
+
+				num_banks = (rdev->config.si.tile_mode_array[index] >> 20) & 0x3;
 			}
 
-			num_banks = (rdev->config.cik.macrotile_mode_array[index] >> 6) & 0x3;
 			fb_format |= EVERGREEN_GRPH_NUM_BANKS(num_banks);
 		} else {
-			/* SI and older. */
-			if (rdev->family >= CHIP_TAHITI)
-				tmp = rdev->config.si.tile_config;
-			else if (rdev->family >= CHIP_CAYMAN)
+			/* NI and older. */
+			if (rdev->family >= CHIP_CAYMAN)
 				tmp = rdev->config.cayman.tile_config;
 			else
 				tmp = rdev->config.evergreen.tile_config;
@@ -1250,6 +1370,18 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(EVERGREEN_GRPH_CONTROL + radeon_crtc->crtc_offset, fb_format);
 	WREG32(EVERGREEN_GRPH_SWAP_CONTROL + radeon_crtc->crtc_offset, fb_swap);
 
+	/*
+	 * The LUT only has 256 slots for indexing by a 8 bpc fb. Bypass the LUT
+	 * for > 8 bpc scanout to avoid truncation of fb indices to 8 msb's, to
+	 * retain the full precision throughout the pipeline.
+	 */
+	WREG32_P(EVERGREEN_GRPH_LUT_10BIT_BYPASS_CONTROL + radeon_crtc->crtc_offset,
+		 (bypass_lut ? EVERGREEN_LUT_10BIT_BYPASS_EN : 0),
+		 ~EVERGREEN_LUT_10BIT_BYPASS_EN);
+
+	if (bypass_lut)
+		DRM_DEBUG_KMS("Bypassing hardware LUT due to 10 bit fb scanout.\n");
+
 	WREG32(EVERGREEN_GRPH_SURFACE_OFFSET_X + radeon_crtc->crtc_offset, 0);
 	WREG32(EVERGREEN_GRPH_SURFACE_OFFSET_Y + radeon_crtc->crtc_offset, 0);
 	WREG32(EVERGREEN_GRPH_X_START + radeon_crtc->crtc_offset, 0);
@@ -1282,10 +1414,10 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	tmp &= ~EVERGREEN_GRPH_SURFACE_UPDATE_H_RETRACE_EN;
 	WREG32(EVERGREEN_GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset, tmp);
 
-	/* set pageflip to happen anywhere in vblank interval */
-	WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 0);
+	/* set pageflip to happen only at start of vblank interval (front porch) */
+	WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3);
 
-	if (!atomic && fb && fb != crtc->fb) {
+	if (!atomic && fb && fb != crtc->primary->fb) {
 		radeon_fb = to_radeon_framebuffer(fb);
 		rbo = gem_to_radeon_bo(radeon_fb->obj);
 		r = radeon_bo_reserve(rbo, false);
@@ -1317,9 +1449,10 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	u32 fb_swap = R600_D1GRPH_SWAP_ENDIAN_NONE;
 	u32 tmp, viewport_w, viewport_h;
 	int r;
+	bool bypass_lut = false;
 
 	/* no fb bound */
-	if (!atomic && !crtc->fb) {
+	if (!atomic && !crtc->primary->fb) {
 		DRM_DEBUG_KMS("No FB bound\n");
 		return 0;
 	}
@@ -1329,8 +1462,8 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 		target_fb = fb;
 	}
 	else {
-		radeon_fb = to_radeon_framebuffer(crtc->fb);
-		target_fb = crtc->fb;
+		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+		target_fb = crtc->primary->fb;
 	}
 
 	obj = radeon_fb->obj;
@@ -1354,18 +1487,30 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	radeon_bo_get_tiling_flags(rbo, &tiling_flags, NULL);
 	radeon_bo_unreserve(rbo);
 
-	switch (target_fb->bits_per_pixel) {
-	case 8:
+	switch (target_fb->pixel_format) {
+	case DRM_FORMAT_C8:
 		fb_format =
 		    AVIVO_D1GRPH_CONTROL_DEPTH_8BPP |
 		    AVIVO_D1GRPH_CONTROL_8BPP_INDEXED;
 		break;
-	case 15:
+	case DRM_FORMAT_XRGB4444:
+	case DRM_FORMAT_ARGB4444:
+		fb_format =
+		    AVIVO_D1GRPH_CONTROL_DEPTH_16BPP |
+		    AVIVO_D1GRPH_CONTROL_16BPP_ARGB4444;
+#ifdef __BIG_ENDIAN
+		fb_swap = R600_D1GRPH_SWAP_ENDIAN_16BIT;
+#endif
+		break;
+	case DRM_FORMAT_XRGB1555:
 		fb_format =
 		    AVIVO_D1GRPH_CONTROL_DEPTH_16BPP |
 		    AVIVO_D1GRPH_CONTROL_16BPP_ARGB1555;
+#ifdef __BIG_ENDIAN
+		fb_swap = R600_D1GRPH_SWAP_ENDIAN_16BIT;
+#endif
 		break;
-	case 16:
+	case DRM_FORMAT_RGB565:
 		fb_format =
 		    AVIVO_D1GRPH_CONTROL_DEPTH_16BPP |
 		    AVIVO_D1GRPH_CONTROL_16BPP_RGB565;
@@ -1373,8 +1518,8 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 		fb_swap = R600_D1GRPH_SWAP_ENDIAN_16BIT;
 #endif
 		break;
-	case 24:
-	case 32:
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_ARGB8888:
 		fb_format =
 		    AVIVO_D1GRPH_CONTROL_DEPTH_32BPP |
 		    AVIVO_D1GRPH_CONTROL_32BPP_ARGB8888;
@@ -1382,9 +1527,20 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 		fb_swap = R600_D1GRPH_SWAP_ENDIAN_32BIT;
 #endif
 		break;
+	case DRM_FORMAT_XRGB2101010:
+	case DRM_FORMAT_ARGB2101010:
+		fb_format =
+		    AVIVO_D1GRPH_CONTROL_DEPTH_32BPP |
+		    AVIVO_D1GRPH_CONTROL_32BPP_ARGB2101010;
+#ifdef __BIG_ENDIAN
+		fb_swap = R600_D1GRPH_SWAP_ENDIAN_32BIT;
+#endif
+		/* Greater 8 bpc fb needs to bypass hw-lut to retain precision */
+		bypass_lut = true;
+		break;
 	default:
-		DRM_ERROR("Unsupported screen depth %d\n",
-			  target_fb->bits_per_pixel);
+		DRM_ERROR("Unsupported screen format %s\n",
+			  drm_get_format_name(target_fb->pixel_format));
 		return -EINVAL;
 	}
 
@@ -1423,6 +1579,13 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	if (rdev->family >= CHIP_R600)
 		WREG32(R600_D1GRPH_SWAP_CONTROL + radeon_crtc->crtc_offset, fb_swap);
 
+	/* LUT only has 256 slots for 8 bpc fb. Bypass for > 8 bpc scanout for precision */
+	WREG32_P(AVIVO_D1GRPH_LUT_SEL + radeon_crtc->crtc_offset,
+		 (bypass_lut ? AVIVO_LUT_10BIT_BYPASS_EN : 0), ~AVIVO_LUT_10BIT_BYPASS_EN);
+
+	if (bypass_lut)
+		DRM_DEBUG_KMS("Bypassing hardware LUT due to 10 bit fb scanout.\n");
+
 	WREG32(AVIVO_D1GRPH_SURFACE_OFFSET_X + radeon_crtc->crtc_offset, 0);
 	WREG32(AVIVO_D1GRPH_SURFACE_OFFSET_Y + radeon_crtc->crtc_offset, 0);
 	WREG32(AVIVO_D1GRPH_X_START + radeon_crtc->crtc_offset, 0);
@@ -1451,10 +1614,10 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	tmp &= ~AVIVO_D1GRPH_SURFACE_UPDATE_H_RETRACE_EN;
 	WREG32(AVIVO_D1GRPH_FLIP_CONTROL + radeon_crtc->crtc_offset, tmp);
 
-	/* set pageflip to happen anywhere in vblank interval */
-	WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 0);
+	/* set pageflip to happen only at start of vblank interval (front porch) */
+	WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3);
 
-	if (!atomic && fb && fb != crtc->fb) {
+	if (!atomic && fb && fb != crtc->primary->fb) {
 		radeon_fb = to_radeon_framebuffer(fb);
 		rbo = gem_to_radeon_bo(radeon_fb->obj);
 		r = radeon_bo_reserve(rbo, false);
@@ -1689,8 +1852,9 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc)
 		}
 		/* otherwise, pick one of the plls */
 		if ((rdev->family == CHIP_KAVERI) ||
-		    (rdev->family == CHIP_KABINI)) {
-			/* KB/KV has PPLL1 and PPLL2 */
+		    (rdev->family == CHIP_KABINI) ||
+		    (rdev->family == CHIP_MULLINS)) {
+			/* KB/KV/ML has PPLL1 and PPLL2 */
 			pll_in_use = radeon_get_pll_use_mask(crtc);
 			if (!(pll_in_use & (1 << ATOM_PPLL2)))
 				return ATOM_PPLL2;
@@ -1743,6 +1907,20 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc)
 			return ATOM_PPLL1;
 		DRM_ERROR("unable to allocate a PPLL\n");
 		return ATOM_PPLL_INVALID;
+	} else if (ASIC_IS_DCE41(rdev)) {
+		/* Don't share PLLs on DCE4.1 chips */
+		if (ENCODER_MODE_IS_DP(atombios_get_encoder_mode(radeon_crtc->encoder))) {
+			if (rdev->clock.dp_extclk)
+				/* skip PPLL programming if using ext clock */
+				return ATOM_PPLL_INVALID;
+		}
+		pll_in_use = radeon_get_pll_use_mask(crtc);
+		if (!(pll_in_use & (1 << ATOM_PPLL1)))
+			return ATOM_PPLL1;
+		if (!(pll_in_use & (1 << ATOM_PPLL2)))
+			return ATOM_PPLL2;
+		DRM_ERROR("unable to allocate a PPLL\n");
+		return ATOM_PPLL_INVALID;
 	} else if (ASIC_IS_DCE4(rdev)) {
 		/* in DP mode, the DP ref clock can come from PPLL, DCPLL, or ext clock,
 		 * depending on the asic:
@@ -1770,7 +1948,7 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc)
 				if (pll != ATOM_PPLL_INVALID)
 					return pll;
 			}
-		} else if (!ASIC_IS_DCE41(rdev)) { /* Don't share PLLs on DCE4.1 chips */
+		} else {
 			/* use the same PPLL for all monitors with the same clock */
 			pll = radeon_get_shared_nondp_ppll(crtc);
 			if (pll != ATOM_PPLL_INVALID)
@@ -1840,6 +2018,9 @@ int atombios_crtc_mode_set(struct drm_crtc *crtc,
 	    (ATOM_DEVICE_TV_SUPPORT | ATOM_DEVICE_CV_SUPPORT))
 		is_tvcv = true;
 
+	if (!radeon_crtc->adjusted_clock)
+		return -EINVAL;
+
 	atombios_crtc_set_pll(crtc, adjusted_mode);
 
 	if (ASIC_IS_DCE4(rdev))
@@ -1927,12 +2108,12 @@ static void atombios_crtc_disable(struct drm_crtc *crtc)
 	int i;
 
 	atombios_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
-	if (crtc->fb) {
+	if (crtc->primary->fb) {
 		int r;
 		struct radeon_framebuffer *radeon_fb;
 		struct radeon_bo *rbo;
 
-		radeon_fb = to_radeon_framebuffer(crtc->fb);
+		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
 		rbo = gem_to_radeon_bo(radeon_fb->obj);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r))
diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c
index fb3ae07a146..b1e11f8434e 100644
--- a/drivers/gpu/drm/radeon/atombios_dp.c
+++ b/drivers/gpu/drm/radeon/atombios_dp.c
@@ -95,9 +95,12 @@ static int radeon_process_aux_ch(struct radeon_i2c_chan *chan,
 	int index = GetIndexIntoMasterTable(COMMAND, ProcessAuxChannelTransaction);
 	unsigned char *base;
 	int recv_bytes;
+	int r = 0;
 
 	memset(&args, 0, sizeof(args));
 
+	mutex_lock(&chan->mutex);
+
 	base = (unsigned char *)(rdev->mode_info.atom_context->scratch + 1);
 
 	radeon_atom_copy_swap(base, send, send_bytes, true);
@@ -117,19 +120,22 @@ static int radeon_process_aux_ch(struct radeon_i2c_chan *chan,
 	/* timeout */
 	if (args.v1.ucReplyStatus == 1) {
 		DRM_DEBUG_KMS("dp_aux_ch timeout\n");
-		return -ETIMEDOUT;
+		r = -ETIMEDOUT;
+		goto done;
 	}
 
 	/* flags not zero */
 	if (args.v1.ucReplyStatus == 2) {
 		DRM_DEBUG_KMS("dp_aux_ch flags not zero\n");
-		return -EBUSY;
+		r = -EIO;
+		goto done;
 	}
 
 	/* error */
 	if (args.v1.ucReplyStatus == 3) {
 		DRM_DEBUG_KMS("dp_aux_ch error\n");
-		return -EIO;
+		r = -EIO;
+		goto done;
 	}
 
 	recv_bytes = args.v1.ucDataOutLen;
@@ -139,187 +145,89 @@ static int radeon_process_aux_ch(struct radeon_i2c_chan *chan,
 	if (recv && recv_size)
 		radeon_atom_copy_swap(recv, base + 16, recv_bytes, false);
 
-	return recv_bytes;
-}
-
-static int radeon_dp_aux_native_write(struct radeon_connector *radeon_connector,
-				      u16 address, u8 *send, u8 send_bytes, u8 delay)
-{
-	struct radeon_connector_atom_dig *dig_connector = radeon_connector->con_priv;
-	int ret;
-	u8 msg[20];
-	int msg_bytes = send_bytes + 4;
-	u8 ack;
-	unsigned retry;
-
-	if (send_bytes > 16)
-		return -1;
-
-	msg[0] = address;
-	msg[1] = address >> 8;
-	msg[2] = AUX_NATIVE_WRITE << 4;
-	msg[3] = (msg_bytes << 4) | (send_bytes - 1);
-	memcpy(&msg[4], send, send_bytes);
-
-	for (retry = 0; retry < 4; retry++) {
-		ret = radeon_process_aux_ch(dig_connector->dp_i2c_bus,
-					    msg, msg_bytes, NULL, 0, delay, &ack);
-		if (ret == -EBUSY)
-			continue;
-		else if (ret < 0)
-			return ret;
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK)
-			return send_bytes;
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
-			udelay(400);
-		else
-			return -EIO;
-	}
+	r = recv_bytes;
+done:
+	mutex_unlock(&chan->mutex);
 
-	return -EIO;
+	return r;
 }
 
-static int radeon_dp_aux_native_read(struct radeon_connector *radeon_connector,
-				     u16 address, u8 *recv, int recv_bytes, u8 delay)
+#define BARE_ADDRESS_SIZE 3
+#define HEADER_SIZE (BARE_ADDRESS_SIZE + 1)
+
+static ssize_t
+radeon_dp_aux_transfer(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg)
 {
-	struct radeon_connector_atom_dig *dig_connector = radeon_connector->con_priv;
-	u8 msg[4];
-	int msg_bytes = 4;
-	u8 ack;
+	struct radeon_i2c_chan *chan =
+		container_of(aux, struct radeon_i2c_chan, aux);
 	int ret;
-	unsigned retry;
-
-	msg[0] = address;
-	msg[1] = address >> 8;
-	msg[2] = AUX_NATIVE_READ << 4;
-	msg[3] = (msg_bytes << 4) | (recv_bytes - 1);
-
-	for (retry = 0; retry < 4; retry++) {
-		ret = radeon_process_aux_ch(dig_connector->dp_i2c_bus,
-					    msg, msg_bytes, recv, recv_bytes, delay, &ack);
-		if (ret == -EBUSY)
-			continue;
-		else if (ret < 0)
-			return ret;
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK)
-			return ret;
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
-			udelay(400);
-		else if (ret == 0)
-			return -EPROTO;
+	u8 tx_buf[20];
+	size_t tx_size;
+	u8 ack, delay = 0;
+
+	if (WARN_ON(msg->size > 16))
+		return -E2BIG;
+
+	tx_buf[0] = msg->address & 0xff;
+	tx_buf[1] = msg->address >> 8;
+	tx_buf[2] = msg->request << 4;
+	tx_buf[3] = msg->size ? (msg->size - 1) : 0;
+
+	switch (msg->request & ~DP_AUX_I2C_MOT) {
+	case DP_AUX_NATIVE_WRITE:
+	case DP_AUX_I2C_WRITE:
+		/* tx_size needs to be 4 even for bare address packets since the atom
+		 * table needs the info in tx_buf[3].
+		 */
+		tx_size = HEADER_SIZE + msg->size;
+		if (msg->size == 0)
+			tx_buf[3] |= BARE_ADDRESS_SIZE << 4;
 		else
-			return -EIO;
+			tx_buf[3] |= tx_size << 4;
+		memcpy(tx_buf + HEADER_SIZE, msg->buffer, msg->size);
+		ret = radeon_process_aux_ch(chan,
+					    tx_buf, tx_size, NULL, 0, delay, &ack);
+		if (ret >= 0)
+			/* Return payload size. */
+			ret = msg->size;
+		break;
+	case DP_AUX_NATIVE_READ:
+	case DP_AUX_I2C_READ:
+		/* tx_size needs to be 4 even for bare address packets since the atom
+		 * table needs the info in tx_buf[3].
+		 */
+		tx_size = HEADER_SIZE;
+		if (msg->size == 0)
+			tx_buf[3] |= BARE_ADDRESS_SIZE << 4;
+		else
+			tx_buf[3] |= tx_size << 4;
+		ret = radeon_process_aux_ch(chan,
+					    tx_buf, tx_size, msg->buffer, msg->size, delay, &ack);
+		break;
+	default:
+		ret = -EINVAL;
+		break;
 	}
 
-	return -EIO;
-}
-
-static void radeon_write_dpcd_reg(struct radeon_connector *radeon_connector,
-				 u16 reg, u8 val)
-{
-	radeon_dp_aux_native_write(radeon_connector, reg, &val, 1, 0);
-}
-
-static u8 radeon_read_dpcd_reg(struct radeon_connector *radeon_connector,
-			       u16 reg)
-{
-	u8 val = 0;
-
-	radeon_dp_aux_native_read(radeon_connector, reg, &val, 1, 0);
+	if (ret >= 0)
+		msg->reply = ack >> 4;
 
-	return val;
+	return ret;
 }
 
-int radeon_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
-			 u8 write_byte, u8 *read_byte)
+void radeon_dp_aux_init(struct radeon_connector *radeon_connector)
 {
-	struct i2c_algo_dp_aux_data *algo_data = adapter->algo_data;
-	struct radeon_i2c_chan *auxch = (struct radeon_i2c_chan *)adapter;
-	u16 address = algo_data->address;
-	u8 msg[5];
-	u8 reply[2];
-	unsigned retry;
-	int msg_bytes;
-	int reply_bytes = 1;
 	int ret;
-	u8 ack;
 
-	/* Set up the command byte */
-	if (mode & MODE_I2C_READ)
-		msg[2] = AUX_I2C_READ << 4;
-	else
-		msg[2] = AUX_I2C_WRITE << 4;
-
-	if (!(mode & MODE_I2C_STOP))
-		msg[2] |= AUX_I2C_MOT << 4;
-
-	msg[0] = address;
-	msg[1] = address >> 8;
+	radeon_connector->ddc_bus->rec.hpd = radeon_connector->hpd.hpd;
+	radeon_connector->ddc_bus->aux.dev = radeon_connector->base.kdev;
+	radeon_connector->ddc_bus->aux.transfer = radeon_dp_aux_transfer;
 
-	switch (mode) {
-	case MODE_I2C_WRITE:
-		msg_bytes = 5;
-		msg[3] = msg_bytes << 4;
-		msg[4] = write_byte;
-		break;
-	case MODE_I2C_READ:
-		msg_bytes = 4;
-		msg[3] = msg_bytes << 4;
-		break;
-	default:
-		msg_bytes = 4;
-		msg[3] = 3 << 4;
-		break;
-	}
-
-	for (retry = 0; retry < 4; retry++) {
-		ret = radeon_process_aux_ch(auxch,
-					    msg, msg_bytes, reply, reply_bytes, 0, &ack);
-		if (ret == -EBUSY)
-			continue;
-		else if (ret < 0) {
-			DRM_DEBUG_KMS("aux_ch failed %d\n", ret);
-			return ret;
-		}
-
-		switch (ack & AUX_NATIVE_REPLY_MASK) {
-		case AUX_NATIVE_REPLY_ACK:
-			/* I2C-over-AUX Reply field is only valid
-			 * when paired with AUX ACK.
-			 */
-			break;
-		case AUX_NATIVE_REPLY_NACK:
-			DRM_DEBUG_KMS("aux_ch native nack\n");
-			return -EREMOTEIO;
-		case AUX_NATIVE_REPLY_DEFER:
-			DRM_DEBUG_KMS("aux_ch native defer\n");
-			udelay(400);
-			continue;
-		default:
-			DRM_ERROR("aux_ch invalid native reply 0x%02x\n", ack);
-			return -EREMOTEIO;
-		}
-
-		switch (ack & AUX_I2C_REPLY_MASK) {
-		case AUX_I2C_REPLY_ACK:
-			if (mode == MODE_I2C_READ)
-				*read_byte = reply[0];
-			return ret;
-		case AUX_I2C_REPLY_NACK:
-			DRM_DEBUG_KMS("aux_i2c nack\n");
-			return -EREMOTEIO;
-		case AUX_I2C_REPLY_DEFER:
-			DRM_DEBUG_KMS("aux_i2c defer\n");
-			udelay(400);
-			break;
-		default:
-			DRM_ERROR("aux_i2c invalid reply 0x%02x\n", ack);
-			return -EREMOTEIO;
-		}
-	}
+	ret = drm_dp_aux_register(&radeon_connector->ddc_bus->aux);
+	if (!ret)
+		radeon_connector->ddc_bus->has_aux = true;
 
-	DRM_DEBUG_KMS("aux i2c too many retries, giving up\n");
-	return -EREMOTEIO;
+	WARN(ret, "drm_dp_aux_register() failed with error %d\n", ret);
 }
 
 /***** general DP utility functions *****/
@@ -384,6 +292,19 @@ static int dp_get_max_dp_pix_clock(int link_rate,
 
 /***** radeon specific DP functions *****/
 
+static int radeon_dp_get_max_link_rate(struct drm_connector *connector,
+				       u8 dpcd[DP_DPCD_SIZE])
+{
+	int max_link_rate;
+
+	if (radeon_connector_is_dp12_capable(connector))
+		max_link_rate = min(drm_dp_max_link_rate(dpcd), 540000);
+	else
+		max_link_rate = min(drm_dp_max_link_rate(dpcd), 270000);
+
+	return max_link_rate;
+}
+
 /* First get the min lane# when low rate is used according to pixel clock
  * (prefer low rate), second check max lane# supported by DP panel,
  * if the max lane# < low rate lane# then use max lane# instead.
@@ -393,7 +314,7 @@ static int radeon_dp_get_dp_lane_number(struct drm_connector *connector,
 					int pix_clock)
 {
 	int bpp = convert_bpc_to_bpp(radeon_get_monitor_bpc(connector));
-	int max_link_rate = drm_dp_max_link_rate(dpcd);
+	int max_link_rate = radeon_dp_get_max_link_rate(connector, dpcd);
 	int max_lane_num = drm_dp_max_lane_count(dpcd);
 	int lane_num;
 	int max_dp_pix_clock;
@@ -431,7 +352,7 @@ static int radeon_dp_get_dp_link_clock(struct drm_connector *connector,
 			return 540000;
 	}
 
-	return drm_dp_max_link_rate(dpcd);
+	return radeon_dp_get_max_link_rate(connector, dpcd);
 }
 
 static u8 radeon_dp_encoder_service(struct radeon_device *rdev,
@@ -454,12 +375,11 @@ static u8 radeon_dp_encoder_service(struct radeon_device *rdev,
 
 u8 radeon_dp_getsinktype(struct radeon_connector *radeon_connector)
 {
-	struct radeon_connector_atom_dig *dig_connector = radeon_connector->con_priv;
 	struct drm_device *dev = radeon_connector->base.dev;
 	struct radeon_device *rdev = dev->dev_private;
 
 	return radeon_dp_encoder_service(rdev, ATOM_DP_ACTION_GET_SINK_TYPE, 0,
-					 dig_connector->dp_i2c_bus->rec.i2c_id, 0);
+					 radeon_connector->ddc_bus->rec.i2c_id, 0);
 }
 
 static void radeon_dp_probe_oui(struct radeon_connector *radeon_connector)
@@ -470,11 +390,11 @@ static void radeon_dp_probe_oui(struct radeon_connector *radeon_connector)
 	if (!(dig_connector->dpcd[DP_DOWN_STREAM_PORT_COUNT] & DP_OUI_SUPPORT))
 		return;
 
-	if (radeon_dp_aux_native_read(radeon_connector, DP_SINK_OUI, buf, 3, 0))
+	if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_SINK_OUI, buf, 3) == 3)
 		DRM_DEBUG_KMS("Sink OUI: %02hx%02hx%02hx\n",
 			      buf[0], buf[1], buf[2]);
 
-	if (radeon_dp_aux_native_read(radeon_connector, DP_BRANCH_OUI, buf, 3, 0))
+	if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_BRANCH_OUI, buf, 3) == 3)
 		DRM_DEBUG_KMS("Branch OUI: %02hx%02hx%02hx\n",
 			      buf[0], buf[1], buf[2]);
 }
@@ -483,16 +403,18 @@ bool radeon_dp_getdpcd(struct radeon_connector *radeon_connector)
 {
 	struct radeon_connector_atom_dig *dig_connector = radeon_connector->con_priv;
 	u8 msg[DP_DPCD_SIZE];
-	int ret, i;
+	int ret;
+
+	char dpcd_hex_dump[DP_DPCD_SIZE * 3];
 
-	ret = radeon_dp_aux_native_read(radeon_connector, DP_DPCD_REV, msg,
-					DP_DPCD_SIZE, 0);
+	ret = drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_DPCD_REV, msg,
+			       DP_DPCD_SIZE);
 	if (ret > 0) {
 		memcpy(dig_connector->dpcd, msg, DP_DPCD_SIZE);
-		DRM_DEBUG_KMS("DPCD: ");
-		for (i = 0; i < DP_DPCD_SIZE; i++)
-			DRM_DEBUG_KMS("%02x ", msg[i]);
-		DRM_DEBUG_KMS("\n");
+
+		hex_dump_to_buffer(dig_connector->dpcd, sizeof(dig_connector->dpcd),
+				   32, 1, dpcd_hex_dump, sizeof(dpcd_hex_dump), false);
+		DRM_DEBUG_KMS("DPCD: %s\n", dpcd_hex_dump);
 
 		radeon_dp_probe_oui(radeon_connector);
 
@@ -508,6 +430,7 @@ int radeon_dp_get_panel_mode(struct drm_encoder *encoder,
 	struct drm_device *dev = encoder->dev;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
+	struct radeon_connector_atom_dig *dig_connector;
 	int panel_mode = DP_PANEL_MODE_EXTERNAL_DP_MODE;
 	u16 dp_bridge = radeon_connector_encoder_get_dp_bridge_encoder_id(connector);
 	u8 tmp;
@@ -515,21 +438,30 @@ int radeon_dp_get_panel_mode(struct drm_encoder *encoder,
 	if (!ASIC_IS_DCE4(rdev))
 		return panel_mode;
 
+	if (!radeon_connector->con_priv)
+		return panel_mode;
+
+	dig_connector = radeon_connector->con_priv;
+
 	if (dp_bridge != ENCODER_OBJECT_ID_NONE) {
 		/* DP bridge chips */
-		tmp = radeon_read_dpcd_reg(radeon_connector, DP_EDP_CONFIGURATION_CAP);
-		if (tmp & 1)
-			panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE;
-		else if ((dp_bridge == ENCODER_OBJECT_ID_NUTMEG) ||
-			 (dp_bridge == ENCODER_OBJECT_ID_TRAVIS))
-			panel_mode = DP_PANEL_MODE_INTERNAL_DP1_MODE;
-		else
-			panel_mode = DP_PANEL_MODE_EXTERNAL_DP_MODE;
+		if (drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux,
+				      DP_EDP_CONFIGURATION_CAP, &tmp) == 1) {
+			if (tmp & 1)
+				panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE;
+			else if ((dp_bridge == ENCODER_OBJECT_ID_NUTMEG) ||
+				 (dp_bridge == ENCODER_OBJECT_ID_TRAVIS))
+				panel_mode = DP_PANEL_MODE_INTERNAL_DP1_MODE;
+			else
+				panel_mode = DP_PANEL_MODE_EXTERNAL_DP_MODE;
+		}
 	} else if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) {
 		/* eDP */
-		tmp = radeon_read_dpcd_reg(radeon_connector, DP_EDP_CONFIGURATION_CAP);
-		if (tmp & 1)
-			panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE;
+		if (drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux,
+				      DP_EDP_CONFIGURATION_CAP, &tmp) == 1) {
+			if (tmp & 1)
+				panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE;
+		}
 	}
 
 	return panel_mode;
@@ -575,37 +507,43 @@ int radeon_dp_mode_valid_helper(struct drm_connector *connector,
 	return MODE_OK;
 }
 
-static bool radeon_dp_get_link_status(struct radeon_connector *radeon_connector,
-				      u8 link_status[DP_LINK_STATUS_SIZE])
-{
-	int ret;
-	ret = radeon_dp_aux_native_read(radeon_connector, DP_LANE0_1_STATUS,
-					link_status, DP_LINK_STATUS_SIZE, 100);
-	if (ret <= 0) {
-		return false;
-	}
-
-	DRM_DEBUG_KMS("link status %6ph\n", link_status);
-	return true;
-}
-
 bool radeon_dp_needs_link_train(struct radeon_connector *radeon_connector)
 {
 	u8 link_status[DP_LINK_STATUS_SIZE];
 	struct radeon_connector_atom_dig *dig = radeon_connector->con_priv;
 
-	if (!radeon_dp_get_link_status(radeon_connector, link_status))
+	if (drm_dp_dpcd_read_link_status(&radeon_connector->ddc_bus->aux, link_status)
+	    <= 0)
 		return false;
 	if (drm_dp_channel_eq_ok(link_status, dig->dp_lane_count))
 		return false;
 	return true;
 }
 
+void radeon_dp_set_rx_power_state(struct drm_connector *connector,
+				  u8 power_state)
+{
+	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
+	struct radeon_connector_atom_dig *dig_connector;
+
+	if (!radeon_connector->con_priv)
+		return;
+
+	dig_connector = radeon_connector->con_priv;
+
+	/* power up/down the sink */
+	if (dig_connector->dpcd[0] >= 0x11) {
+		drm_dp_dpcd_writeb(&radeon_connector->ddc_bus->aux,
+				   DP_SET_POWER, power_state);
+		usleep_range(1000, 2000);
+	}
+}
+
+
 struct radeon_dp_link_train_info {
 	struct radeon_device *rdev;
 	struct drm_encoder *encoder;
 	struct drm_connector *connector;
-	struct radeon_connector *radeon_connector;
 	int enc_id;
 	int dp_clock;
 	int dp_lane_count;
@@ -615,6 +553,7 @@ struct radeon_dp_link_train_info {
 	u8 link_status[DP_LINK_STATUS_SIZE];
 	u8 tries;
 	bool use_dpencoder;
+	struct drm_dp_aux *aux;
 };
 
 static void radeon_dp_update_vs_emph(struct radeon_dp_link_train_info *dp_info)
@@ -625,8 +564,8 @@ static void radeon_dp_update_vs_emph(struct radeon_dp_link_train_info *dp_info)
 				       0, dp_info->train_set[0]); /* sets all lanes at once */
 
 	/* set the vs/emph on the sink */
-	radeon_dp_aux_native_write(dp_info->radeon_connector, DP_TRAINING_LANE0_SET,
-				   dp_info->train_set, dp_info->dp_lane_count, 0);
+	drm_dp_dpcd_write(dp_info->aux, DP_TRAINING_LANE0_SET,
+			  dp_info->train_set, dp_info->dp_lane_count);
 }
 
 static void radeon_dp_set_tp(struct radeon_dp_link_train_info *dp_info, int tp)
@@ -661,7 +600,7 @@ static void radeon_dp_set_tp(struct radeon_dp_link_train_info *dp_info, int tp)
 	}
 
 	/* enable training pattern on the sink */
-	radeon_write_dpcd_reg(dp_info->radeon_connector, DP_TRAINING_PATTERN_SET, tp);
+	drm_dp_dpcd_writeb(dp_info->aux, DP_TRAINING_PATTERN_SET, tp);
 }
 
 static int radeon_dp_link_train_init(struct radeon_dp_link_train_info *dp_info)
@@ -671,32 +610,30 @@ static int radeon_dp_link_train_init(struct radeon_dp_link_train_info *dp_info)
 	u8 tmp;
 
 	/* power up the sink */
-	if (dp_info->dpcd[0] >= 0x11)
-		radeon_write_dpcd_reg(dp_info->radeon_connector,
-				      DP_SET_POWER, DP_SET_POWER_D0);
+	radeon_dp_set_rx_power_state(dp_info->connector, DP_SET_POWER_D0);
 
 	/* possibly enable downspread on the sink */
 	if (dp_info->dpcd[3] & 0x1)
-		radeon_write_dpcd_reg(dp_info->radeon_connector,
-				      DP_DOWNSPREAD_CTRL, DP_SPREAD_AMP_0_5);
+		drm_dp_dpcd_writeb(dp_info->aux,
+				   DP_DOWNSPREAD_CTRL, DP_SPREAD_AMP_0_5);
 	else
-		radeon_write_dpcd_reg(dp_info->radeon_connector,
-				      DP_DOWNSPREAD_CTRL, 0);
+		drm_dp_dpcd_writeb(dp_info->aux,
+				   DP_DOWNSPREAD_CTRL, 0);
 
 	if ((dp_info->connector->connector_type == DRM_MODE_CONNECTOR_eDP) &&
 	    (dig->panel_mode == DP_PANEL_MODE_INTERNAL_DP2_MODE)) {
-		radeon_write_dpcd_reg(dp_info->radeon_connector, DP_EDP_CONFIGURATION_SET, 1);
+		drm_dp_dpcd_writeb(dp_info->aux, DP_EDP_CONFIGURATION_SET, 1);
 	}
 
 	/* set the lane count on the sink */
 	tmp = dp_info->dp_lane_count;
 	if (drm_dp_enhanced_frame_cap(dp_info->dpcd))
 		tmp |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
-	radeon_write_dpcd_reg(dp_info->radeon_connector, DP_LANE_COUNT_SET, tmp);
+	drm_dp_dpcd_writeb(dp_info->aux, DP_LANE_COUNT_SET, tmp);
 
 	/* set the link rate on the sink */
 	tmp = drm_dp_link_rate_to_bw_code(dp_info->dp_clock);
-	radeon_write_dpcd_reg(dp_info->radeon_connector, DP_LINK_BW_SET, tmp);
+	drm_dp_dpcd_writeb(dp_info->aux, DP_LINK_BW_SET, tmp);
 
 	/* start training on the source */
 	if (ASIC_IS_DCE4(dp_info->rdev) || !dp_info->use_dpencoder)
@@ -707,9 +644,9 @@ static int radeon_dp_link_train_init(struct radeon_dp_link_train_info *dp_info)
 					  dp_info->dp_clock, dp_info->enc_id, 0);
 
 	/* disable the training pattern on the sink */
-	radeon_write_dpcd_reg(dp_info->radeon_connector,
-			      DP_TRAINING_PATTERN_SET,
-			      DP_TRAINING_PATTERN_DISABLE);
+	drm_dp_dpcd_writeb(dp_info->aux,
+			   DP_TRAINING_PATTERN_SET,
+			   DP_TRAINING_PATTERN_DISABLE);
 
 	return 0;
 }
@@ -719,9 +656,9 @@ static int radeon_dp_link_train_finish(struct radeon_dp_link_train_info *dp_info
 	udelay(400);
 
 	/* disable the training pattern on the sink */
-	radeon_write_dpcd_reg(dp_info->radeon_connector,
-			      DP_TRAINING_PATTERN_SET,
-			      DP_TRAINING_PATTERN_DISABLE);
+	drm_dp_dpcd_writeb(dp_info->aux,
+			   DP_TRAINING_PATTERN_SET,
+			   DP_TRAINING_PATTERN_DISABLE);
 
 	/* disable the training pattern on the source */
 	if (ASIC_IS_DCE4(dp_info->rdev) || !dp_info->use_dpencoder)
@@ -753,7 +690,8 @@ static int radeon_dp_link_train_cr(struct radeon_dp_link_train_info *dp_info)
 	while (1) {
 		drm_dp_link_train_clock_recovery_delay(dp_info->dpcd);
 
-		if (!radeon_dp_get_link_status(dp_info->radeon_connector, dp_info->link_status)) {
+		if (drm_dp_dpcd_read_link_status(dp_info->aux,
+						 dp_info->link_status) <= 0) {
 			DRM_ERROR("displayport link status failed\n");
 			break;
 		}
@@ -815,7 +753,8 @@ static int radeon_dp_link_train_ce(struct radeon_dp_link_train_info *dp_info)
 	while (1) {
 		drm_dp_link_train_channel_eq_delay(dp_info->dpcd);
 
-		if (!radeon_dp_get_link_status(dp_info->radeon_connector, dp_info->link_status)) {
+		if (drm_dp_dpcd_read_link_status(dp_info->aux,
+						 dp_info->link_status) <= 0) {
 			DRM_ERROR("displayport link status failed\n");
 			break;
 		}
@@ -898,19 +837,23 @@ void radeon_dp_link_train(struct drm_encoder *encoder,
 	else
 		dp_info.enc_id |= ATOM_DP_CONFIG_LINK_A;
 
-	tmp = radeon_read_dpcd_reg(radeon_connector, DP_MAX_LANE_COUNT);
-	if (ASIC_IS_DCE5(rdev) && (tmp & DP_TPS3_SUPPORTED))
-		dp_info.tp3_supported = true;
-	else
+	if (drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux, DP_MAX_LANE_COUNT, &tmp)
+	    == 1) {
+		if (ASIC_IS_DCE5(rdev) && (tmp & DP_TPS3_SUPPORTED))
+			dp_info.tp3_supported = true;
+		else
+			dp_info.tp3_supported = false;
+	} else {
 		dp_info.tp3_supported = false;
+	}
 
 	memcpy(dp_info.dpcd, dig_connector->dpcd, DP_RECEIVER_CAP_SIZE);
 	dp_info.rdev = rdev;
 	dp_info.encoder = encoder;
 	dp_info.connector = connector;
-	dp_info.radeon_connector = radeon_connector;
 	dp_info.dp_lane_count = dig_connector->dp_lane_count;
 	dp_info.dp_clock = dig_connector->dp_clock;
+	dp_info.aux = &radeon_connector->ddc_bus->aux;
 
 	if (radeon_dp_link_train_init(&dp_info))
 		goto done;
diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c
index a42d61571f4..7d68203a373 100644
--- a/drivers/gpu/drm/radeon/atombios_encoders.c
+++ b/drivers/gpu/drm/radeon/atombios_encoders.c
@@ -183,7 +183,6 @@ void radeon_atom_backlight_init(struct radeon_encoder *radeon_encoder,
 	struct backlight_properties props;
 	struct radeon_backlight_privdata *pdata;
 	struct radeon_encoder_atom_dig *dig;
-	u8 backlight_level;
 	char bl_name[16];
 
 	/* Mac laptops with multiple GPUs use the gmux driver for backlight
@@ -222,12 +221,17 @@ void radeon_atom_backlight_init(struct radeon_encoder *radeon_encoder,
 
 	pdata->encoder = radeon_encoder;
 
-	backlight_level = radeon_atom_get_backlight_level_from_reg(rdev);
-
 	dig = radeon_encoder->enc_priv;
 	dig->bl_dev = bd;
 
 	bd->props.brightness = radeon_atom_backlight_get_brightness(bd);
+	/* Set a reasonable default here if the level is 0 otherwise
+	 * fbdev will attempt to turn the backlight on after console
+	 * unblanking and it will try and restore 0 which turns the backlight
+	 * off again.
+	 */
+	if (bd->props.brightness == 0)
+		bd->props.brightness = RADEON_MAX_BL_LEVEL;
 	bd->props.power = FB_BLANK_UNBLANK;
 	backlight_update_status(bd);
 
@@ -464,11 +468,12 @@ atombios_tv_setup(struct drm_encoder *encoder, int action)
 
 static u8 radeon_atom_get_bpc(struct drm_encoder *encoder)
 {
-	struct drm_connector *connector = radeon_get_connector_for_encoder(encoder);
 	int bpc = 8;
 
-	if (connector)
-		bpc = radeon_get_monitor_bpc(connector);
+	if (encoder->crtc) {
+		struct radeon_crtc *radeon_crtc = to_radeon_crtc(encoder->crtc);
+		bpc = radeon_crtc->bpc;
+	}
 
 	switch (bpc) {
 	case 0:
@@ -1313,7 +1318,7 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t
 			}
 			if (is_dp)
 				args.v5.ucLaneNum = dp_lane_count;
-			else if (radeon_encoder->pixel_clock > 165000)
+			else if (radeon_dig_monitor_is_duallink(encoder, radeon_encoder->pixel_clock))
 				args.v5.ucLaneNum = 8;
 			else
 				args.v5.ucLaneNum = 4;
@@ -1632,10 +1637,16 @@ radeon_atom_encoder_dpms_dig(struct drm_encoder *encoder, int mode)
 	struct drm_connector *connector = radeon_get_connector_for_encoder(encoder);
 	struct radeon_connector *radeon_connector = NULL;
 	struct radeon_connector_atom_dig *radeon_dig_connector = NULL;
+	bool travis_quirk = false;
 
 	if (connector) {
 		radeon_connector = to_radeon_connector(connector);
 		radeon_dig_connector = radeon_connector->con_priv;
+		if ((radeon_connector_encoder_get_dp_bridge_encoder_id(connector) ==
+		     ENCODER_OBJECT_ID_TRAVIS) &&
+		    (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) &&
+		    !ASIC_IS_DCE5(rdev))
+			travis_quirk = true;
 	}
 
 	switch (mode) {
@@ -1656,17 +1667,13 @@ radeon_atom_encoder_dpms_dig(struct drm_encoder *encoder, int mode)
 					atombios_external_encoder_setup(encoder, ext_encoder,
 									EXTERNAL_ENCODER_ACTION_V3_ENCODER_SETUP);
 			}
-			atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE, 0, 0);
 		} else if (ASIC_IS_DCE4(rdev)) {
 			/* setup and enable the encoder */
 			atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_SETUP, 0);
-			/* enable the transmitter */
-			atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE, 0, 0);
 		} else {
 			/* setup and enable the encoder and transmitter */
 			atombios_dig_encoder_setup(encoder, ATOM_ENABLE, 0);
 			atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_SETUP, 0, 0);
-			atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE, 0, 0);
 		}
 		if (ENCODER_MODE_IS_DP(atombios_get_encoder_mode(encoder)) && connector) {
 			if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) {
@@ -1674,68 +1681,56 @@ radeon_atom_encoder_dpms_dig(struct drm_encoder *encoder, int mode)
 							     ATOM_TRANSMITTER_ACTION_POWER_ON);
 				radeon_dig_connector->edp_on = true;
 			}
+		}
+		/* enable the transmitter */
+		atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_ENABLE, 0, 0);
+		if (ENCODER_MODE_IS_DP(atombios_get_encoder_mode(encoder)) && connector) {
+			/* DP_SET_POWER_D0 is set in radeon_dp_link_train */
 			radeon_dp_link_train(encoder, connector);
 			if (ASIC_IS_DCE4(rdev))
 				atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_DP_VIDEO_ON, 0);
 		}
 		if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
-			atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_LCD_BLON, 0, 0);
+			atombios_dig_transmitter_setup(encoder,
+						       ATOM_TRANSMITTER_ACTION_LCD_BLON, 0, 0);
+		if (ext_encoder)
+			atombios_external_encoder_setup(encoder, ext_encoder, ATOM_ENABLE);
 		break;
 	case DRM_MODE_DPMS_STANDBY:
 	case DRM_MODE_DPMS_SUSPEND:
 	case DRM_MODE_DPMS_OFF:
 		if (ASIC_IS_DCE4(rdev)) {
+			if (ENCODER_MODE_IS_DP(atombios_get_encoder_mode(encoder)) && connector)
+				atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_DP_VIDEO_OFF, 0);
+		}
+		if (ext_encoder)
+			atombios_external_encoder_setup(encoder, ext_encoder, ATOM_DISABLE);
+		if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
+			atombios_dig_transmitter_setup(encoder,
+						       ATOM_TRANSMITTER_ACTION_LCD_BLOFF, 0, 0);
+
+		if (ENCODER_MODE_IS_DP(atombios_get_encoder_mode(encoder)) &&
+		    connector && !travis_quirk)
+			radeon_dp_set_rx_power_state(connector, DP_SET_POWER_D3);
+		if (ASIC_IS_DCE4(rdev)) {
 			/* disable the transmitter */
-			atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0);
+			atombios_dig_transmitter_setup(encoder,
+						       ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0);
 		} else {
 			/* disable the encoder and transmitter */
-			atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0);
+			atombios_dig_transmitter_setup(encoder,
+						       ATOM_TRANSMITTER_ACTION_DISABLE, 0, 0);
 			atombios_dig_encoder_setup(encoder, ATOM_DISABLE, 0);
 		}
 		if (ENCODER_MODE_IS_DP(atombios_get_encoder_mode(encoder)) && connector) {
-			if (ASIC_IS_DCE4(rdev))
-				atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_DP_VIDEO_OFF, 0);
+			if (travis_quirk)
+				radeon_dp_set_rx_power_state(connector, DP_SET_POWER_D3);
 			if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) {
 				atombios_set_edp_panel_power(connector,
 							     ATOM_TRANSMITTER_ACTION_POWER_OFF);
 				radeon_dig_connector->edp_on = false;
 			}
 		}
-		if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
-			atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_LCD_BLOFF, 0, 0);
-		break;
-	}
-}
-
-static void
-radeon_atom_encoder_dpms_ext(struct drm_encoder *encoder,
-			     struct drm_encoder *ext_encoder,
-			     int mode)
-{
-	struct drm_device *dev = encoder->dev;
-	struct radeon_device *rdev = dev->dev_private;
-
-	switch (mode) {
-	case DRM_MODE_DPMS_ON:
-	default:
-		if (ASIC_IS_DCE41(rdev) || ASIC_IS_DCE61(rdev)) {
-			atombios_external_encoder_setup(encoder, ext_encoder,
-							EXTERNAL_ENCODER_ACTION_V3_ENABLE_OUTPUT);
-			atombios_external_encoder_setup(encoder, ext_encoder,
-							EXTERNAL_ENCODER_ACTION_V3_ENCODER_BLANKING_OFF);
-		} else
-			atombios_external_encoder_setup(encoder, ext_encoder, ATOM_ENABLE);
-		break;
-	case DRM_MODE_DPMS_STANDBY:
-	case DRM_MODE_DPMS_SUSPEND:
-	case DRM_MODE_DPMS_OFF:
-		if (ASIC_IS_DCE41(rdev) || ASIC_IS_DCE61(rdev)) {
-			atombios_external_encoder_setup(encoder, ext_encoder,
-							EXTERNAL_ENCODER_ACTION_V3_ENCODER_BLANKING);
-			atombios_external_encoder_setup(encoder, ext_encoder,
-							EXTERNAL_ENCODER_ACTION_V3_DISABLE_OUTPUT);
-		} else
-			atombios_external_encoder_setup(encoder, ext_encoder, ATOM_DISABLE);
 		break;
 	}
 }
@@ -1746,7 +1741,6 @@ radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode)
 	struct drm_device *dev = encoder->dev;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-	struct drm_encoder *ext_encoder = radeon_get_external_encoder(encoder);
 
 	DRM_DEBUG_KMS("encoder dpms %d to mode %d, devices %08x, active_devices %08x\n",
 		  radeon_encoder->encoder_id, mode, radeon_encoder->devices,
@@ -1806,9 +1800,6 @@ radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode)
 		return;
 	}
 
-	if (ext_encoder)
-		radeon_atom_encoder_dpms_ext(encoder, ext_encoder, mode);
-
 	radeon_atombios_encoder_dpms_scratch_regs(encoder, (mode == DRM_MODE_DPMS_ON) ? true : false);
 
 }
@@ -1897,8 +1888,11 @@ atombios_set_encoder_crtc_source(struct drm_encoder *encoder)
 					args.v2.ucEncodeMode = ATOM_ENCODER_MODE_CRT;
 				else
 					args.v2.ucEncodeMode = atombios_get_encoder_mode(encoder);
-			} else
+			} else if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) {
+				args.v2.ucEncodeMode = ATOM_ENCODER_MODE_LVDS;
+			} else {
 				args.v2.ucEncodeMode = atombios_get_encoder_mode(encoder);
+			}
 			switch (radeon_encoder->encoder_id) {
 			case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
 			case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
diff --git a/drivers/gpu/drm/radeon/atombios_i2c.c b/drivers/gpu/drm/radeon/atombios_i2c.c
index f685035dbe3..9c570fb15b8 100644
--- a/drivers/gpu/drm/radeon/atombios_i2c.c
+++ b/drivers/gpu/drm/radeon/atombios_i2c.c
@@ -27,8 +27,6 @@
 #include "radeon.h"
 #include "atom.h"
 
-extern void radeon_atom_copy_swap(u8 *dst, u8 *src, u8 num_bytes, bool to_le);
-
 #define TARGET_HW_I2C_CLOCK 50
 
 /* these are a limitation of ProcessI2cChannelTransaction not the hw */
@@ -45,15 +43,19 @@ static int radeon_process_i2c_ch(struct radeon_i2c_chan *chan,
 	int index = GetIndexIntoMasterTable(COMMAND, ProcessI2cChannelTransaction);
 	unsigned char *base;
 	u16 out = cpu_to_le16(0);
+	int r = 0;
 
 	memset(&args, 0, sizeof(args));
 
+	mutex_lock(&chan->mutex);
+
 	base = (unsigned char *)rdev->mode_info.atom_context->scratch;
 
 	if (flags & HW_I2C_WRITE) {
 		if (num > ATOM_MAX_HW_I2C_WRITE) {
 			DRM_ERROR("hw i2c: tried to write too many bytes (%d vs 3)\n", num);
-			return -EINVAL;
+			r = -EINVAL;
+			goto done;
 		}
 		if (buf == NULL)
 			args.ucRegIndex = 0;
@@ -67,7 +69,8 @@ static int radeon_process_i2c_ch(struct radeon_i2c_chan *chan,
 	} else {
 		if (num > ATOM_MAX_HW_I2C_READ) {
 			DRM_ERROR("hw i2c: tried to read too many bytes (%d vs 255)\n", num);
-			return -EINVAL;
+			r = -EINVAL;
+			goto done;
 		}
 		args.ucRegIndex = 0;
 		args.lpI2CDataOut = 0;
@@ -84,13 +87,17 @@ static int radeon_process_i2c_ch(struct radeon_i2c_chan *chan,
 	/* error */
 	if (args.ucStatus != HW_ASSISTED_I2C_STATUS_SUCCESS) {
 		DRM_DEBUG_KMS("hw_i2c error\n");
-		return -EIO;
+		r = -EIO;
+		goto done;
 	}
 
 	if (!(flags & HW_I2C_WRITE))
 		radeon_atom_copy_swap(buf, base, num, false);
 
-	return 0;
+done:
+	mutex_unlock(&chan->mutex);
+
+	return r;
 }
 
 int radeon_atom_hw_i2c_xfer(struct i2c_adapter *i2c_adap,
diff --git a/drivers/gpu/drm/radeon/btc_dpm.c b/drivers/gpu/drm/radeon/btc_dpm.c
index 9b6950d9b3c..f81d7ca134d 100644
--- a/drivers/gpu/drm/radeon/btc_dpm.c
+++ b/drivers/gpu/drm/radeon/btc_dpm.c
@@ -29,6 +29,7 @@
 #include "cypress_dpm.h"
 #include "btc_dpm.h"
 #include "atom.h"
+#include <linux/seq_file.h>
 
 #define MC_CG_ARB_FREQ_F0           0x0a
 #define MC_CG_ARB_FREQ_F1           0x0b
@@ -49,6 +50,7 @@ struct rv7xx_ps *rv770_get_ps(struct radeon_ps *rps);
 struct rv7xx_power_info *rv770_get_pi(struct radeon_device *rdev);
 struct evergreen_power_info *evergreen_get_pi(struct radeon_device *rdev);
 
+extern int ni_mc_load_microcode(struct radeon_device *rdev);
 
 //********* BARTS **************//
 static const u32 barts_cgcg_cgls_default[] =
@@ -2510,21 +2512,6 @@ int btc_dpm_enable(struct radeon_device *rdev)
 	if (eg_pi->ls_clock_gating)
 		btc_ls_clock_gating_enable(rdev, true);
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-		PPSMC_Result result;
-
-		ret = rv770_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-		result = rv770_send_msg_to_smc(rdev, PPSMC_MSG_EnableThermalInterrupt);
-
-		if (result != PPSMC_Result_OK)
-			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
-	}
-
 	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
 	btc_init_stutter_mode(rdev);
@@ -2576,7 +2563,11 @@ void btc_dpm_disable(struct radeon_device *rdev)
 void btc_dpm_setup_asic(struct radeon_device *rdev)
 {
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	int r;
 
+	r = ni_mc_load_microcode(rdev);
+	if (r)
+		DRM_ERROR("Failed to load MC firmware!\n");
 	rv770_get_memory_type(rdev);
 	rv740_read_clock_registers(rdev);
 	btc_read_arb_registers(rdev);
@@ -2610,6 +2601,10 @@ int btc_dpm_init(struct radeon_device *rdev)
 	pi->min_vddc_in_table = 0;
 	pi->max_vddc_in_table = 0;
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = rv7xx_parse_power_table(rdev);
 	if (ret)
 		return ret;
@@ -2766,6 +2761,37 @@ void btc_dpm_fini(struct radeon_device *rdev)
 	r600_free_extended_power_table(rdev);
 }
 
+void btc_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
+						     struct seq_file *m)
+{
+	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	struct radeon_ps *rps = &eg_pi->current_rps;
+	struct rv7xx_ps *ps = rv770_get_ps(rps);
+	struct rv7xx_pl *pl;
+	u32 current_index =
+		(RREG32(TARGET_AND_CURRENT_PROFILE_INDEX) & CURRENT_PROFILE_INDEX_MASK) >>
+		CURRENT_PROFILE_INDEX_SHIFT;
+
+	if (current_index > 2) {
+		seq_printf(m, "invalid dpm profile %d\n", current_index);
+	} else {
+		if (current_index == 0)
+			pl = &ps->low;
+		else if (current_index == 1)
+			pl = &ps->medium;
+		else /* current_index == 2 */
+			pl = &ps->high;
+		seq_printf(m, "uvd    vclk: %d dclk: %d\n", rps->vclk, rps->dclk);
+		if (rdev->family >= CHIP_CEDAR) {
+			seq_printf(m, "power level %d    sclk: %u mclk: %u vddc: %u vddci: %u\n",
+				   current_index, pl->sclk, pl->mclk, pl->vddc, pl->vddci);
+		} else {
+			seq_printf(m, "power level %d    sclk: %u mclk: %u vddc: %u\n",
+				   current_index, pl->sclk, pl->mclk, pl->vddc);
+		}
+	}
+}
+
 u32 btc_dpm_get_sclk(struct radeon_device *rdev, bool low)
 {
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
diff --git a/drivers/gpu/drm/radeon/btcd.h b/drivers/gpu/drm/radeon/btcd.h
index 29e32de7e02..9c65be2d55a 100644
--- a/drivers/gpu/drm/radeon/btcd.h
+++ b/drivers/gpu/drm/radeon/btcd.h
@@ -44,6 +44,10 @@
 #       define DYN_SPREAD_SPECTRUM_EN                   (1 << 23)
 #       define AC_DC_SW                                 (1 << 24)
 
+#define TARGET_AND_CURRENT_PROFILE_INDEX                  0x66c
+#       define CURRENT_PROFILE_INDEX_MASK                 (0xf << 4)
+#       define CURRENT_PROFILE_INDEX_SHIFT                4
+
 #define	CG_BIF_REQ_AND_RSP				0x7f4
 #define		CG_CLIENT_REQ(x)			((x) << 0)
 #define		CG_CLIENT_REQ_MASK			(0xff << 0)
diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c
index 1ed47997635..584090ac3eb 100644
--- a/drivers/gpu/drm/radeon/ci_dpm.c
+++ b/drivers/gpu/drm/radeon/ci_dpm.c
@@ -21,8 +21,10 @@
  *
  */
 
+#include <linux/firmware.h>
 #include "drmP.h"
 #include "radeon.h"
+#include "radeon_ucode.h"
 #include "cikd.h"
 #include "r600_dpm.h"
 #include "ci_dpm.h"
@@ -171,6 +173,7 @@ extern void si_trim_voltage_table_to_fit_state_table(struct radeon_device *rdev,
 						     struct atom_voltage_table *voltage_table);
 extern void cik_enter_rlc_safe_mode(struct radeon_device *rdev);
 extern void cik_exit_rlc_safe_mode(struct radeon_device *rdev);
+extern int ci_mc_load_microcode(struct radeon_device *rdev);
 extern void cik_update_cg(struct radeon_device *rdev,
 			  u32 block, bool enable);
 
@@ -201,24 +204,29 @@ static void ci_initialize_powertune_defaults(struct radeon_device *rdev)
 	struct ci_power_info *pi = ci_get_pi(rdev);
 
 	switch (rdev->pdev->device) {
+	case 0x6649:
 	case 0x6650:
+	case 0x6651:
 	case 0x6658:
 	case 0x665C:
+	case 0x665D:
 	default:
 		pi->powertune_defaults = &defaults_bonaire_xt;
 		break;
-	case 0x6651:
-	case 0x665D:
-		pi->powertune_defaults = &defaults_bonaire_pro;
-		break;
 	case 0x6640:
-		pi->powertune_defaults = &defaults_saturn_xt;
-		break;
 	case 0x6641:
-		pi->powertune_defaults = &defaults_saturn_pro;
+	case 0x6646:
+	case 0x6647:
+		pi->powertune_defaults = &defaults_saturn_xt;
 		break;
 	case 0x67B8:
 	case 0x67B0:
+		pi->powertune_defaults = &defaults_hawaii_xt;
+		break;
+	case 0x67BA:
+	case 0x67B1:
+		pi->powertune_defaults = &defaults_hawaii_pro;
+		break;
 	case 0x67A0:
 	case 0x67A1:
 	case 0x67A2:
@@ -227,11 +235,7 @@ static void ci_initialize_powertune_defaults(struct radeon_device *rdev)
 	case 0x67AA:
 	case 0x67B9:
 	case 0x67BE:
-		pi->powertune_defaults = &defaults_hawaii_xt;
-		break;
-	case 0x67BA:
-	case 0x67B1:
-		pi->powertune_defaults = &defaults_hawaii_pro;
+		pi->powertune_defaults = &defaults_bonaire_xt;
 		break;
 	}
 
@@ -747,6 +751,14 @@ static void ci_apply_state_adjust_rules(struct radeon_device *rdev,
 	u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
 	int i;
 
+	if (rps->vce_active) {
+		rps->evclk = rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].evclk;
+		rps->ecclk = rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].ecclk;
+	} else {
+		rps->evclk = 0;
+		rps->ecclk = 0;
+	}
+
 	if ((rdev->pm.dpm.new_active_crtc_count > 1) ||
 	    ci_dpm_vblank_too_short(rdev))
 		disable_mclk_switching = true;
@@ -805,6 +817,13 @@ static void ci_apply_state_adjust_rules(struct radeon_device *rdev,
 		sclk = ps->performance_levels[0].sclk;
 	}
 
+	if (rps->vce_active) {
+		if (sclk < rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].sclk)
+			sclk = rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].sclk;
+		if (mclk < rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].mclk)
+			mclk = rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].mclk;
+	}
+
 	ps->performance_levels[0].sclk = sclk;
 	ps->performance_levels[0].mclk = mclk;
 
@@ -1160,7 +1179,7 @@ static int ci_stop_dpm(struct radeon_device *rdev)
 	tmp &= ~GLOBAL_PWRMGT_EN;
 	WREG32_SMC(GENERAL_PWRMGT, tmp);
 
-	tmp = RREG32(SCLK_PWRMGT_CNTL);
+	tmp = RREG32_SMC(SCLK_PWRMGT_CNTL);
 	tmp &= ~DYNAMIC_PM_EN;
 	WREG32_SMC(SCLK_PWRMGT_CNTL, tmp);
 
@@ -3469,7 +3488,6 @@ static int ci_enable_uvd_dpm(struct radeon_device *rdev, bool enable)
 		0 : -EINVAL;
 }
 
-#if 0
 static int ci_enable_vce_dpm(struct radeon_device *rdev, bool enable)
 {
 	struct ci_power_info *pi = ci_get_pi(rdev);
@@ -3502,6 +3520,7 @@ static int ci_enable_vce_dpm(struct radeon_device *rdev, bool enable)
 		0 : -EINVAL;
 }
 
+#if 0
 static int ci_enable_samu_dpm(struct radeon_device *rdev, bool enable)
 {
 	struct ci_power_info *pi = ci_get_pi(rdev);
@@ -3588,7 +3607,6 @@ static int ci_update_uvd_dpm(struct radeon_device *rdev, bool gate)
 	return ci_enable_uvd_dpm(rdev, !gate);
 }
 
-#if 0
 static u8 ci_get_vce_boot_level(struct radeon_device *rdev)
 {
 	u8 i;
@@ -3609,15 +3627,15 @@ static int ci_update_vce_dpm(struct radeon_device *rdev,
 			     struct radeon_ps *radeon_current_state)
 {
 	struct ci_power_info *pi = ci_get_pi(rdev);
-	bool new_vce_clock_non_zero = (radeon_new_state->evclk != 0);
-	bool old_vce_clock_non_zero = (radeon_current_state->evclk != 0);
 	int ret = 0;
 	u32 tmp;
 
-	if (new_vce_clock_non_zero != old_vce_clock_non_zero) {
-		if (new_vce_clock_non_zero) {
-			pi->smc_state_table.VceBootLevel = ci_get_vce_boot_level(rdev);
+	if (radeon_current_state->evclk != radeon_new_state->evclk) {
+		if (radeon_new_state->evclk) {
+			/* turn the clocks on when encoding */
+			cik_update_cg(rdev, RADEON_CG_BLOCK_VCE, false);
 
+			pi->smc_state_table.VceBootLevel = ci_get_vce_boot_level(rdev);
 			tmp = RREG32_SMC(DPM_TABLE_475);
 			tmp &= ~VceBootLevel_MASK;
 			tmp |= VceBootLevel(pi->smc_state_table.VceBootLevel);
@@ -3625,12 +3643,16 @@ static int ci_update_vce_dpm(struct radeon_device *rdev,
 
 			ret = ci_enable_vce_dpm(rdev, true);
 		} else {
+			/* turn the clocks off when not encoding */
+			cik_update_cg(rdev, RADEON_CG_BLOCK_VCE, true);
+
 			ret = ci_enable_vce_dpm(rdev, false);
 		}
 	}
 	return ret;
 }
 
+#if 0
 static int ci_update_samu_dpm(struct radeon_device *rdev, bool gate)
 {
 	return ci_enable_samu_dpm(rdev, gate);
@@ -4503,8 +4525,8 @@ static void ci_get_memory_type(struct radeon_device *rdev)
 
 }
 
-void ci_update_current_ps(struct radeon_device *rdev,
-			  struct radeon_ps *rps)
+static void ci_update_current_ps(struct radeon_device *rdev,
+				 struct radeon_ps *rps)
 {
 	struct ci_ps *new_ps = ci_get_ps(rps);
 	struct ci_power_info *pi = ci_get_pi(rdev);
@@ -4514,8 +4536,8 @@ void ci_update_current_ps(struct radeon_device *rdev,
 	pi->current_rps.ps_priv = &pi->current_ps;
 }
 
-void ci_update_requested_ps(struct radeon_device *rdev,
-			    struct radeon_ps *rps)
+static void ci_update_requested_ps(struct radeon_device *rdev,
+				   struct radeon_ps *rps)
 {
 	struct ci_ps *new_ps = ci_get_ps(rps);
 	struct ci_power_info *pi = ci_get_pi(rdev);
@@ -4549,6 +4571,11 @@ void ci_dpm_post_set_power_state(struct radeon_device *rdev)
 
 void ci_dpm_setup_asic(struct radeon_device *rdev)
 {
+	int r;
+
+	r = ci_mc_load_microcode(rdev);
+	if (r)
+		DRM_ERROR("Failed to load MC firmware!\n");
 	ci_read_clock_registers(rdev);
 	ci_get_memory_type(rdev);
 	ci_enable_acpi_power_management(rdev);
@@ -4561,13 +4588,6 @@ int ci_dpm_enable(struct radeon_device *rdev)
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
 	int ret;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	if (ci_is_smc_running(rdev))
 		return -EINVAL;
 	if (pi->voltage_control != CISLANDS_VOLTAGE_CONTROL_NONE) {
@@ -4665,6 +4685,18 @@ int ci_dpm_enable(struct radeon_device *rdev)
 		DRM_ERROR("ci_enable_power_containment failed\n");
 		return ret;
 	}
+
+	ci_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
+
+	ci_update_current_ps(rdev, boot_ps);
+
+	return 0;
+}
+
+int ci_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 #if 0
@@ -4685,19 +4717,8 @@ int ci_dpm_enable(struct radeon_device *rdev)
 #endif
 	}
 
-	ci_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
-
 	ci_dpm_powergate_uvd(rdev, true);
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), true);
-
-	ci_update_current_ps(rdev, boot_ps);
-
 	return 0;
 }
 
@@ -4706,12 +4727,6 @@ void ci_dpm_disable(struct radeon_device *rdev)
 	struct ci_power_info *pi = ci_get_pi(rdev);
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	ci_dpm_powergate_uvd(rdev, false);
 
 	if (!ci_is_smc_running(rdev))
@@ -4742,13 +4757,6 @@ int ci_dpm_set_power_state(struct radeon_device *rdev)
 	struct radeon_ps *old_ps = &pi->current_rps;
 	int ret;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	ci_find_dpm_states_clocks_in_dpm_table(rdev, new_ps);
 	if (pi->pcie_performance_request)
 		ci_request_link_speed_change_before_state_change(rdev, new_ps, old_ps);
@@ -4767,13 +4775,13 @@ int ci_dpm_set_power_state(struct radeon_device *rdev)
 		DRM_ERROR("ci_generate_dpm_level_enable_mask failed\n");
 		return ret;
 	}
-#if 0
+
 	ret = ci_update_vce_dpm(rdev, new_ps, old_ps);
 	if (ret) {
 		DRM_ERROR("ci_update_vce_dpm failed\n");
 		return ret;
 	}
-#endif
+
 	ret = ci_update_sclk_t(rdev);
 	if (ret) {
 		DRM_ERROR("ci_update_sclk_t failed\n");
@@ -4804,13 +4812,6 @@ int ci_dpm_set_power_state(struct radeon_device *rdev)
 	if (pi->pcie_performance_request)
 		ci_notify_link_speed_change_after_state_change(rdev, new_ps, old_ps);
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), true);
-
 	return 0;
 }
 
@@ -4981,9 +4982,6 @@ static int ci_parse_power_table(struct radeon_device *rdev)
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
-	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
-	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
-	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
 	for (i = 0; i < state_array->ucNumEntries; i++) {
 		u8 *idx;
 		power_state = (union pplib_power_state *)power_state_offset;
@@ -5020,11 +5018,26 @@ static int ci_parse_power_table(struct radeon_device *rdev)
 		power_state_offset += 2 + power_state->v2.ucNumDPMLevels;
 	}
 	rdev->pm.dpm.num_ps = state_array->ucNumEntries;
+
+	/* fill in the vce power states */
+	for (i = 0; i < RADEON_MAX_VCE_LEVELS; i++) {
+		u32 sclk, mclk;
+		clock_array_index = rdev->pm.dpm.vce_states[i].clk_idx;
+		clock_info = (union pplib_clock_info *)
+			&clock_info_array->clockInfo[clock_array_index * clock_info_array->ucEntrySize];
+		sclk = le16_to_cpu(clock_info->ci.usEngineClockLow);
+		sclk |= clock_info->ci.ucEngineClockHigh << 16;
+		mclk = le16_to_cpu(clock_info->ci.usMemoryClockLow);
+		mclk |= clock_info->ci.ucMemoryClockHigh << 16;
+		rdev->pm.dpm.vce_states[i].sclk = sclk;
+		rdev->pm.dpm.vce_states[i].mclk = mclk;
+	}
+
 	return 0;
 }
 
-int ci_get_vbios_boot_values(struct radeon_device *rdev,
-			     struct ci_vbios_boot_state *boot_state)
+static int ci_get_vbios_boot_values(struct radeon_device *rdev,
+				    struct ci_vbios_boot_state *boot_state)
 {
 	struct radeon_mode_info *mode_info = &rdev->mode_info;
 	int index = GetIndexIntoMasterTable(DATA, FirmwareInfo);
@@ -5099,17 +5112,25 @@ int ci_dpm_init(struct radeon_device *rdev)
 		ci_dpm_fini(rdev);
 		return ret;
 	}
-	ret = ci_parse_power_table(rdev);
+
+	ret = r600_get_platform_caps(rdev);
 	if (ret) {
 		ci_dpm_fini(rdev);
 		return ret;
 	}
+
 	ret = r600_parse_extended_power_table(rdev);
 	if (ret) {
 		ci_dpm_fini(rdev);
 		return ret;
 	}
 
+	ret = ci_parse_power_table(rdev);
+	if (ret) {
+		ci_dpm_fini(rdev);
+		return ret;
+	}
+
         pi->dll_default_on = false;
         pi->sram_end = SMC_RAM_END;
 
@@ -5128,6 +5149,12 @@ int ci_dpm_init(struct radeon_device *rdev)
 	pi->mclk_dpm_key_disabled = 0;
 	pi->pcie_dpm_key_disabled = 0;
 
+	/* mclk dpm is unstable on some R7 260X cards with the old mc ucode */
+	if ((rdev->pdev->device == 0x6658) &&
+	    (rdev->mc_fw->size == (BONAIRE_MC_UCODE_SIZE * 4))) {
+		pi->mclk_dpm_key_disabled = 1;
+	}
+
 	pi->caps_sclk_ds = true;
 
 	pi->mclk_strobe_mode_threshold = 40000;
@@ -5142,6 +5169,7 @@ int ci_dpm_init(struct radeon_device *rdev)
 	pi->caps_sclk_throttle_low_notification = false;
 
 	pi->caps_uvd_dpm = true;
+	pi->caps_vce_dpm = true;
 
         ci_get_leakage_voltages(rdev);
         ci_patch_dependency_tables_with_leakage(rdev);
diff --git a/drivers/gpu/drm/radeon/ci_smc.c b/drivers/gpu/drm/radeon/ci_smc.c
index 9c745dd2243..8debc9d4736 100644
--- a/drivers/gpu/drm/radeon/ci_smc.c
+++ b/drivers/gpu/drm/radeon/ci_smc.c
@@ -28,6 +28,7 @@
 #include "cikd.h"
 #include "ppsmc.h"
 #include "radeon_ucode.h"
+#include "ci_dpm.h"
 
 static int ci_set_smc_sram_address(struct radeon_device *rdev,
 				   u32 smc_address, u32 limit)
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index e950fabd7f5..c0ea66192fe 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -38,6 +38,7 @@ MODULE_FIRMWARE("radeon/BONAIRE_me.bin");
 MODULE_FIRMWARE("radeon/BONAIRE_ce.bin");
 MODULE_FIRMWARE("radeon/BONAIRE_mec.bin");
 MODULE_FIRMWARE("radeon/BONAIRE_mc.bin");
+MODULE_FIRMWARE("radeon/BONAIRE_mc2.bin");
 MODULE_FIRMWARE("radeon/BONAIRE_rlc.bin");
 MODULE_FIRMWARE("radeon/BONAIRE_sdma.bin");
 MODULE_FIRMWARE("radeon/BONAIRE_smc.bin");
@@ -46,6 +47,7 @@ MODULE_FIRMWARE("radeon/HAWAII_me.bin");
 MODULE_FIRMWARE("radeon/HAWAII_ce.bin");
 MODULE_FIRMWARE("radeon/HAWAII_mec.bin");
 MODULE_FIRMWARE("radeon/HAWAII_mc.bin");
+MODULE_FIRMWARE("radeon/HAWAII_mc2.bin");
 MODULE_FIRMWARE("radeon/HAWAII_rlc.bin");
 MODULE_FIRMWARE("radeon/HAWAII_sdma.bin");
 MODULE_FIRMWARE("radeon/HAWAII_smc.bin");
@@ -61,6 +63,12 @@ MODULE_FIRMWARE("radeon/KABINI_ce.bin");
 MODULE_FIRMWARE("radeon/KABINI_mec.bin");
 MODULE_FIRMWARE("radeon/KABINI_rlc.bin");
 MODULE_FIRMWARE("radeon/KABINI_sdma.bin");
+MODULE_FIRMWARE("radeon/MULLINS_pfp.bin");
+MODULE_FIRMWARE("radeon/MULLINS_me.bin");
+MODULE_FIRMWARE("radeon/MULLINS_ce.bin");
+MODULE_FIRMWARE("radeon/MULLINS_mec.bin");
+MODULE_FIRMWARE("radeon/MULLINS_rlc.bin");
+MODULE_FIRMWARE("radeon/MULLINS_sdma.bin");
 
 extern int r600_ih_ring_alloc(struct radeon_device *rdev);
 extern void r600_ih_ring_fini(struct radeon_device *rdev);
@@ -72,9 +80,11 @@ extern int sumo_rlc_init(struct radeon_device *rdev);
 extern void si_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc);
 extern void si_rlc_reset(struct radeon_device *rdev);
 extern void si_init_uvd_internal_cg(struct radeon_device *rdev);
+static u32 cik_get_cu_active_bitmap(struct radeon_device *rdev, u32 se, u32 sh);
 extern int cik_sdma_resume(struct radeon_device *rdev);
 extern void cik_sdma_enable(struct radeon_device *rdev, bool enable);
 extern void cik_sdma_fini(struct radeon_device *rdev);
+extern void vce_v2_0_enable_mgcg(struct radeon_device *rdev, bool enable);
 static void cik_rlc_stop(struct radeon_device *rdev);
 static void cik_pcie_gen3_enable(struct radeon_device *rdev);
 static void cik_program_aspm(struct radeon_device *rdev);
@@ -1095,7 +1105,7 @@ static const u32 spectre_golden_registers[] =
 	0x8a14, 0xf000003f, 0x00000007,
 	0x8b24, 0xffffffff, 0x00ffffff,
 	0x28350, 0x3f3f3fff, 0x00000082,
-	0x28355, 0x0000003f, 0x00000000,
+	0x28354, 0x0000003f, 0x00000000,
 	0x3e78, 0x00000001, 0x00000002,
 	0x913c, 0xffff03df, 0x00000004,
 	0xc768, 0x00000008, 0x00000008,
@@ -1470,6 +1480,43 @@ static const u32 hawaii_mgcg_cgcg_init[] =
 	0xd80c, 0xff000ff0, 0x00000100
 };
 
+static const u32 godavari_golden_registers[] =
+{
+	0x55e4, 0xff607fff, 0xfc000100,
+	0x6ed8, 0x00010101, 0x00010000,
+	0x9830, 0xffffffff, 0x00000000,
+	0x98302, 0xf00fffff, 0x00000400,
+	0x6130, 0xffffffff, 0x00010000,
+	0x5bb0, 0x000000f0, 0x00000070,
+	0x5bc0, 0xf0311fff, 0x80300000,
+	0x98f8, 0x73773777, 0x12010001,
+	0x98fc, 0xffffffff, 0x00000010,
+	0x8030, 0x00001f0f, 0x0000100a,
+	0x2f48, 0x73773777, 0x12010001,
+	0x2408, 0x000fffff, 0x000c007f,
+	0x8a14, 0xf000003f, 0x00000007,
+	0x8b24, 0xffffffff, 0x00ff0fff,
+	0x30a04, 0x0000ff0f, 0x00000000,
+	0x28a4c, 0x07ffffff, 0x06000000,
+	0x4d8, 0x00000fff, 0x00000100,
+	0xd014, 0x00010000, 0x00810001,
+	0xd814, 0x00010000, 0x00810001,
+	0x3e78, 0x00000001, 0x00000002,
+	0xc768, 0x00000008, 0x00000008,
+	0xc770, 0x00000f00, 0x00000800,
+	0xc774, 0x00000f00, 0x00000800,
+	0xc798, 0x00ffffff, 0x00ff7fbf,
+	0xc79c, 0x00ffffff, 0x00ff7faf,
+	0x8c00, 0x000000ff, 0x00000001,
+	0x214f8, 0x01ff01ff, 0x00000002,
+	0x21498, 0x007ff800, 0x00200000,
+	0x2015c, 0xffffffff, 0x00000f40,
+	0x88c4, 0x001f3ae3, 0x00000082,
+	0x88d4, 0x0000001f, 0x00000010,
+	0x30934, 0xffffffff, 0x00000000
+};
+
+
 static void cik_init_golden_registers(struct radeon_device *rdev)
 {
 	switch (rdev->family) {
@@ -1501,6 +1548,20 @@ static void cik_init_golden_registers(struct radeon_device *rdev)
 						 kalindi_golden_spm_registers,
 						 (const u32)ARRAY_SIZE(kalindi_golden_spm_registers));
 		break;
+	case CHIP_MULLINS:
+		radeon_program_register_sequence(rdev,
+						 kalindi_mgcg_cgcg_init,
+						 (const u32)ARRAY_SIZE(kalindi_mgcg_cgcg_init));
+		radeon_program_register_sequence(rdev,
+						 godavari_golden_registers,
+						 (const u32)ARRAY_SIZE(godavari_golden_registers));
+		radeon_program_register_sequence(rdev,
+						 kalindi_golden_common_registers,
+						 (const u32)ARRAY_SIZE(kalindi_golden_common_registers));
+		radeon_program_register_sequence(rdev,
+						 kalindi_golden_spm_registers,
+						 (const u32)ARRAY_SIZE(kalindi_golden_spm_registers));
+		break;
 	case CHIP_KAVERI:
 		radeon_program_register_sequence(rdev,
 						 spectre_mgcg_cgcg_init,
@@ -1697,25 +1758,25 @@ static void cik_srbm_select(struct radeon_device *rdev,
  * Load the GDDR MC ucode into the hw (CIK).
  * Returns 0 on success, error on failure.
  */
-static int ci_mc_load_microcode(struct radeon_device *rdev)
+int ci_mc_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
 	u32 running, blackout = 0;
 	u32 *io_mc_regs;
-	int i, ucode_size, regs_size;
+	int i, regs_size, ucode_size;
 
 	if (!rdev->mc_fw)
 		return -EINVAL;
 
+	ucode_size = rdev->mc_fw->size / 4;
+
 	switch (rdev->family) {
 	case CHIP_BONAIRE:
 		io_mc_regs = (u32 *)&bonaire_io_mc_regs;
-		ucode_size = CIK_MC_UCODE_SIZE;
 		regs_size = BONAIRE_IO_MC_REGS_SIZE;
 		break;
 	case CHIP_HAWAII:
 		io_mc_regs = (u32 *)&hawaii_io_mc_regs;
-		ucode_size = HAWAII_MC_UCODE_SIZE;
 		regs_size = HAWAII_IO_MC_REGS_SIZE;
 		break;
 	default:
@@ -1782,7 +1843,7 @@ static int cik_init_microcode(struct radeon_device *rdev)
 	const char *chip_name;
 	size_t pfp_req_size, me_req_size, ce_req_size,
 		mec_req_size, rlc_req_size, mc_req_size = 0,
-		sdma_req_size, smc_req_size = 0;
+		sdma_req_size, smc_req_size = 0, mc2_req_size = 0;
 	char fw_name[30];
 	int err;
 
@@ -1796,7 +1857,8 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		ce_req_size = CIK_CE_UCODE_SIZE * 4;
 		mec_req_size = CIK_MEC_UCODE_SIZE * 4;
 		rlc_req_size = BONAIRE_RLC_UCODE_SIZE * 4;
-		mc_req_size = CIK_MC_UCODE_SIZE * 4;
+		mc_req_size = BONAIRE_MC_UCODE_SIZE * 4;
+		mc2_req_size = BONAIRE_MC2_UCODE_SIZE * 4;
 		sdma_req_size = CIK_SDMA_UCODE_SIZE * 4;
 		smc_req_size = ALIGN(BONAIRE_SMC_UCODE_SIZE, 4);
 		break;
@@ -1808,6 +1870,7 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		mec_req_size = CIK_MEC_UCODE_SIZE * 4;
 		rlc_req_size = BONAIRE_RLC_UCODE_SIZE * 4;
 		mc_req_size = HAWAII_MC_UCODE_SIZE * 4;
+		mc2_req_size = HAWAII_MC2_UCODE_SIZE * 4;
 		sdma_req_size = CIK_SDMA_UCODE_SIZE * 4;
 		smc_req_size = ALIGN(HAWAII_SMC_UCODE_SIZE, 4);
 		break;
@@ -1829,6 +1892,15 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		rlc_req_size = KB_RLC_UCODE_SIZE * 4;
 		sdma_req_size = CIK_SDMA_UCODE_SIZE * 4;
 		break;
+	case CHIP_MULLINS:
+		chip_name = "MULLINS";
+		pfp_req_size = CIK_PFP_UCODE_SIZE * 4;
+		me_req_size = CIK_ME_UCODE_SIZE * 4;
+		ce_req_size = CIK_CE_UCODE_SIZE * 4;
+		mec_req_size = CIK_MEC_UCODE_SIZE * 4;
+		rlc_req_size = ML_RLC_UCODE_SIZE * 4;
+		sdma_req_size = CIK_SDMA_UCODE_SIZE * 4;
+		break;
 	default: BUG();
 	}
 
@@ -1903,16 +1975,22 @@ static int cik_init_microcode(struct radeon_device *rdev)
 
 	/* No SMC, MC ucode on APUs */
 	if (!(rdev->flags & RADEON_IS_IGP)) {
-		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name);
+		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc2.bin", chip_name);
 		err = request_firmware(&rdev->mc_fw, fw_name, rdev->dev);
-		if (err)
-			goto out;
-		if (rdev->mc_fw->size != mc_req_size) {
+		if (err) {
+			snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name);
+			err = request_firmware(&rdev->mc_fw, fw_name, rdev->dev);
+			if (err)
+				goto out;
+		}
+		if ((rdev->mc_fw->size != mc_req_size) &&
+		    (rdev->mc_fw->size != mc2_req_size)){
 			printk(KERN_ERR
 			       "cik_mc: Bogus length %zu in firmware \"%s\"\n",
 			       rdev->mc_fw->size, fw_name);
 			err = -EINVAL;
 		}
+		DRM_INFO("%s: %zu bytes\n", fw_name, rdev->mc_fw->size);
 
 		snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name);
 		err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev);
@@ -2028,6 +2106,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 5:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
 				break;
 			case 6:
@@ -2048,6 +2127,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 9:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
 				break;
 			case 10:
@@ -2070,6 +2150,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 13:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
 				break;
 			case 14:
@@ -2092,6 +2173,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 27:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
 				break;
 			case 28:
@@ -2209,6 +2291,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				gb_tile_moden = 0;
 				break;
 			}
+			rdev->config.cik.macrotile_mode_array[reg_offset] = gb_tile_moden;
 			WREG32(GB_MACROTILE_MODE0 + (reg_offset * 4), gb_tile_moden);
 		}
 	} else if (num_pipe_configs == 8) {
@@ -2246,6 +2329,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 5:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
 				break;
 			case 6:
@@ -2266,6 +2350,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 9:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
 				break;
 			case 10:
@@ -2288,6 +2373,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 13:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
 				break;
 			case 14:
@@ -2310,6 +2396,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 27:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P8_32x32_16x16) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
 				break;
 			case 28:
@@ -2466,6 +2553,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 					break;
 				case 5:
 					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
 							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
 					break;
 				case 6:
@@ -2486,6 +2574,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 					break;
 				case 9:
 					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
 							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
 					break;
 				case 10:
@@ -2508,6 +2597,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 					break;
 				case 13:
 					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
 							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
 					break;
 				case 14:
@@ -2530,6 +2620,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 					break;
 				case 27:
 					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+							 PIPE_CONFIG(ADDR_SURF_P4_16x16) |
 							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
 					break;
 				case 28:
@@ -2592,6 +2683,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 					break;
 				case 5:
 					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
 							 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
 					break;
 				case 6:
@@ -2612,6 +2704,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 					break;
 				case 9:
 					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
 							 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
 					break;
 				case 10:
@@ -2634,6 +2727,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 					break;
 				case 13:
 					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
 							 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
 					break;
 				case 14:
@@ -2656,6 +2750,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 					break;
 				case 27:
 					gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+							 PIPE_CONFIG(ADDR_SURF_P4_8x16) |
 							 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
 					break;
 				case 28:
@@ -2812,6 +2907,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 5:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P2) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
 				break;
 			case 6:
@@ -2827,11 +2923,13 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 						 TILE_SPLIT(split_equal_to_row_size));
 				break;
 			case 8:
-				gb_tile_moden = ARRAY_MODE(ARRAY_LINEAR_ALIGNED);
+				gb_tile_moden = ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+						PIPE_CONFIG(ADDR_SURF_P2);
 				break;
 			case 9:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING));
+						 MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
+						 PIPE_CONFIG(ADDR_SURF_P2));
 				break;
 			case 10:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
@@ -2853,6 +2951,7 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 13:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+						 PIPE_CONFIG(ADDR_SURF_P2) |
 						 MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING));
 				break;
 			case 14:
@@ -2875,7 +2974,8 @@ static void cik_tiling_mode_table_init(struct radeon_device *rdev)
 				break;
 			case 27:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
-						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING));
+						 MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
+						 PIPE_CONFIG(ADDR_SURF_P2));
 				break;
 			case 28:
 				gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_2D_TILED_THIN1) |
@@ -3046,7 +3146,7 @@ static u32 cik_create_bitmask(u32 bit_width)
 }
 
 /**
- * cik_select_se_sh - select which SE, SH to address
+ * cik_get_rb_disabled - computes the mask of disabled RBs
  *
  * @rdev: radeon_device pointer
  * @max_rb_num: max RBs (render backends) for the asic
@@ -3159,7 +3259,7 @@ static void cik_gpu_init(struct radeon_device *rdev)
 	u32 mc_shared_chmap, mc_arb_ramcfg;
 	u32 hdp_host_path_cntl;
 	u32 tmp;
-	int i, j;
+	int i, j, k;
 
 	switch (rdev->family) {
 	case CHIP_BONAIRE:
@@ -3240,6 +3340,7 @@ static void cik_gpu_init(struct radeon_device *rdev)
 		gb_addr_config = BONAIRE_GB_ADDR_CONFIG_GOLDEN;
 		break;
 	case CHIP_KABINI:
+	case CHIP_MULLINS:
 	default:
 		rdev->config.cik.max_shader_engines = 1;
 		rdev->config.cik.max_tile_pipes = 2;
@@ -3347,6 +3448,15 @@ static void cik_gpu_init(struct radeon_device *rdev)
 		     rdev->config.cik.max_sh_per_se,
 		     rdev->config.cik.max_backends_per_se);
 
+	for (i = 0; i < rdev->config.cik.max_shader_engines; i++) {
+		for (j = 0; j < rdev->config.cik.max_sh_per_se; j++) {
+			for (k = 0; k < rdev->config.cik.max_cu_per_sh; k++) {
+				rdev->config.cik.active_cus +=
+					hweight32(cik_get_cu_active_bitmap(rdev, i, j));
+			}
+		}
+	}
+
 	/* set HW defaults for 3D engine */
 	WREG32(CP_MEQ_THRESHOLDS, MEQ1_START(0x30) | MEQ2_START(0x60));
 
@@ -3487,6 +3597,51 @@ int cik_ring_test(struct radeon_device *rdev, struct radeon_ring *ring)
 }
 
 /**
+ * cik_hdp_flush_cp_ring_emit - emit an hdp flush on the cp
+ *
+ * @rdev: radeon_device pointer
+ * @ridx: radeon ring index
+ *
+ * Emits an hdp flush on the cp.
+ */
+static void cik_hdp_flush_cp_ring_emit(struct radeon_device *rdev,
+				       int ridx)
+{
+	struct radeon_ring *ring = &rdev->ring[ridx];
+	u32 ref_and_mask;
+
+	switch (ring->idx) {
+	case CAYMAN_RING_TYPE_CP1_INDEX:
+	case CAYMAN_RING_TYPE_CP2_INDEX:
+	default:
+		switch (ring->me) {
+		case 0:
+			ref_and_mask = CP2 << ring->pipe;
+			break;
+		case 1:
+			ref_and_mask = CP6 << ring->pipe;
+			break;
+		default:
+			return;
+		}
+		break;
+	case RADEON_RING_TYPE_GFX_INDEX:
+		ref_and_mask = CP0;
+		break;
+	}
+
+	radeon_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5));
+	radeon_ring_write(ring, (WAIT_REG_MEM_OPERATION(1) | /* write, wait, write */
+				 WAIT_REG_MEM_FUNCTION(3) |  /* == */
+				 WAIT_REG_MEM_ENGINE(1)));   /* pfp */
+	radeon_ring_write(ring, GPU_HDP_FLUSH_REQ >> 2);
+	radeon_ring_write(ring, GPU_HDP_FLUSH_DONE >> 2);
+	radeon_ring_write(ring, ref_and_mask);
+	radeon_ring_write(ring, ref_and_mask);
+	radeon_ring_write(ring, 0x20); /* poll interval */
+}
+
+/**
  * cik_fence_gfx_ring_emit - emit a fence on the gfx ring
  *
  * @rdev: radeon_device pointer
@@ -3512,15 +3667,7 @@ void cik_fence_gfx_ring_emit(struct radeon_device *rdev,
 	radeon_ring_write(ring, fence->seq);
 	radeon_ring_write(ring, 0);
 	/* HDP flush */
-	/* We should be using the new WAIT_REG_MEM special op packet here
-	 * but it causes the CP to hang
-	 */
-	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
-	radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) |
-				 WRITE_DATA_DST_SEL(0)));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 0);
+	cik_hdp_flush_cp_ring_emit(rdev, fence->ring);
 }
 
 /**
@@ -3550,15 +3697,7 @@ void cik_fence_compute_ring_emit(struct radeon_device *rdev,
 	radeon_ring_write(ring, fence->seq);
 	radeon_ring_write(ring, 0);
 	/* HDP flush */
-	/* We should be using the new WAIT_REG_MEM special op packet here
-	 * but it causes the CP to hang
-	 */
-	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
-	radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) |
-				 WRITE_DATA_DST_SEL(0)));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 0);
+	cik_hdp_flush_cp_ring_emit(rdev, fence->ring);
 }
 
 bool cik_semaphore_ring_emit(struct radeon_device *rdev,
@@ -3566,19 +3705,14 @@ bool cik_semaphore_ring_emit(struct radeon_device *rdev,
 			     struct radeon_semaphore *semaphore,
 			     bool emit_wait)
 {
-/* TODO: figure out why semaphore cause lockups */
-#if 0
 	uint64_t addr = semaphore->gpu_addr;
 	unsigned sel = emit_wait ? PACKET3_SEM_SEL_WAIT : PACKET3_SEM_SEL_SIGNAL;
 
 	radeon_ring_write(ring, PACKET3(PACKET3_MEM_SEMAPHORE, 1));
-	radeon_ring_write(ring, addr & 0xffffffff);
+	radeon_ring_write(ring, lower_32_bits(addr));
 	radeon_ring_write(ring, (upper_32_bits(addr) & 0xffff) | sel);
 
 	return true;
-#else
-	return false;
-#endif
 }
 
 /**
@@ -3646,6 +3780,7 @@ int cik_copy_cpdma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
@@ -3694,7 +3829,7 @@ void cik_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 			radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
 			radeon_ring_write(ring, WRITE_DATA_DST_SEL(1));
 			radeon_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc);
-			radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff);
+			radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr));
 			radeon_ring_write(ring, next_rptr);
 		}
 
@@ -3816,6 +3951,8 @@ static void cik_cp_gfx_enable(struct radeon_device *rdev, bool enable)
 	if (enable)
 		WREG32(CP_ME_CNTL, 0);
 	else {
+		if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+			radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 		WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT));
 		rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
 	}
@@ -4004,8 +4141,6 @@ static int cik_cp_gfx_resume(struct radeon_device *rdev)
 	WREG32(CP_RB0_BASE, rb_addr);
 	WREG32(CP_RB0_BASE_HI, upper_32_bits(rb_addr));
 
-	ring->rptr = RREG32(CP_RB0_RPTR);
-
 	/* start the ring */
 	cik_cp_gfx_start(rdev);
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = true;
@@ -4014,18 +4149,50 @@ static int cik_cp_gfx_resume(struct radeon_device *rdev)
 		rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
 		return r;
 	}
+
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+
 	return 0;
 }
 
-u32 cik_compute_ring_get_rptr(struct radeon_device *rdev,
-			      struct radeon_ring *ring)
+u32 cik_gfx_get_rptr(struct radeon_device *rdev,
+		     struct radeon_ring *ring)
 {
 	u32 rptr;
 
+	if (rdev->wb.enabled)
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	else
+		rptr = RREG32(CP_RB0_RPTR);
 
+	return rptr;
+}
+
+u32 cik_gfx_get_wptr(struct radeon_device *rdev,
+		     struct radeon_ring *ring)
+{
+	u32 wptr;
+
+	wptr = RREG32(CP_RB0_WPTR);
+
+	return wptr;
+}
+
+void cik_gfx_set_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	WREG32(CP_RB0_WPTR, ring->wptr);
+	(void)RREG32(CP_RB0_WPTR);
+}
+
+u32 cik_compute_get_rptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring)
+{
+	u32 rptr;
 
 	if (rdev->wb.enabled) {
-		rptr = le32_to_cpu(rdev->wb.wb[ring->rptr_offs/4]);
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
 	} else {
 		mutex_lock(&rdev->srbm_mutex);
 		cik_srbm_select(rdev, ring->me, ring->pipe, ring->queue, 0);
@@ -4037,13 +4204,14 @@ u32 cik_compute_ring_get_rptr(struct radeon_device *rdev,
 	return rptr;
 }
 
-u32 cik_compute_ring_get_wptr(struct radeon_device *rdev,
-			      struct radeon_ring *ring)
+u32 cik_compute_get_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring)
 {
 	u32 wptr;
 
 	if (rdev->wb.enabled) {
-		wptr = le32_to_cpu(rdev->wb.wb[ring->wptr_offs/4]);
+		/* XXX check if swapping is necessary on BE */
+		wptr = rdev->wb.wb[ring->wptr_offs/4];
 	} else {
 		mutex_lock(&rdev->srbm_mutex);
 		cik_srbm_select(rdev, ring->me, ring->pipe, ring->queue, 0);
@@ -4055,10 +4223,11 @@ u32 cik_compute_ring_get_wptr(struct radeon_device *rdev,
 	return wptr;
 }
 
-void cik_compute_ring_set_wptr(struct radeon_device *rdev,
-			       struct radeon_ring *ring)
+void cik_compute_set_wptr(struct radeon_device *rdev,
+			  struct radeon_ring *ring)
 {
-	rdev->wb.wb[ring->wptr_offs/4] = cpu_to_le32(ring->wptr);
+	/* XXX check if swapping is necessary on BE */
+	rdev->wb.wb[ring->wptr_offs/4] = ring->wptr;
 	WDOORBELL32(ring->doorbell_index, ring->wptr);
 }
 
@@ -4074,8 +4243,11 @@ static void cik_cp_compute_enable(struct radeon_device *rdev, bool enable)
 {
 	if (enable)
 		WREG32(CP_MEC_CNTL, 0);
-	else
+	else {
 		WREG32(CP_MEC_CNTL, (MEC_ME1_HALT | MEC_ME2_HALT));
+		rdev->ring[CAYMAN_RING_TYPE_CP1_INDEX].ready = false;
+		rdev->ring[CAYMAN_RING_TYPE_CP2_INDEX].ready = false;
+	}
 	udelay(50);
 }
 
@@ -4526,8 +4698,7 @@ static int cik_cp_compute_resume(struct radeon_device *rdev)
 		rdev->ring[idx].wptr = 0;
 		mqd->queue_state.cp_hqd_pq_wptr = rdev->ring[idx].wptr;
 		WREG32(CP_HQD_PQ_WPTR, mqd->queue_state.cp_hqd_pq_wptr);
-		rdev->ring[idx].rptr = RREG32(CP_HQD_PQ_RPTR);
-		mqd->queue_state.cp_hqd_pq_rptr = rdev->ring[idx].rptr;
+		mqd->queue_state.cp_hqd_pq_rptr = RREG32(CP_HQD_PQ_RPTR);
 
 		/* set the vmid for the queue */
 		mqd->queue_state.cp_hqd_vmid = 0;
@@ -4852,6 +5023,160 @@ static void cik_gpu_soft_reset(struct radeon_device *rdev, u32 reset_mask)
 	cik_print_gpu_status_regs(rdev);
 }
 
+struct kv_reset_save_regs {
+	u32 gmcon_reng_execute;
+	u32 gmcon_misc;
+	u32 gmcon_misc3;
+};
+
+static void kv_save_regs_for_reset(struct radeon_device *rdev,
+				   struct kv_reset_save_regs *save)
+{
+	save->gmcon_reng_execute = RREG32(GMCON_RENG_EXECUTE);
+	save->gmcon_misc = RREG32(GMCON_MISC);
+	save->gmcon_misc3 = RREG32(GMCON_MISC3);
+
+	WREG32(GMCON_RENG_EXECUTE, save->gmcon_reng_execute & ~RENG_EXECUTE_ON_PWR_UP);
+	WREG32(GMCON_MISC, save->gmcon_misc & ~(RENG_EXECUTE_ON_REG_UPDATE |
+						STCTRL_STUTTER_EN));
+}
+
+static void kv_restore_regs_for_reset(struct radeon_device *rdev,
+				      struct kv_reset_save_regs *save)
+{
+	int i;
+
+	WREG32(GMCON_PGFSM_WRITE, 0);
+	WREG32(GMCON_PGFSM_CONFIG, 0x200010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0);
+	WREG32(GMCON_PGFSM_CONFIG, 0x300010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x210000);
+	WREG32(GMCON_PGFSM_CONFIG, 0xa00010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x21003);
+	WREG32(GMCON_PGFSM_CONFIG, 0xb00010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x2b00);
+	WREG32(GMCON_PGFSM_CONFIG, 0xc00010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0);
+	WREG32(GMCON_PGFSM_CONFIG, 0xd00010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x420000);
+	WREG32(GMCON_PGFSM_CONFIG, 0x100010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x120202);
+	WREG32(GMCON_PGFSM_CONFIG, 0x500010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x3e3e36);
+	WREG32(GMCON_PGFSM_CONFIG, 0x600010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x373f3e);
+	WREG32(GMCON_PGFSM_CONFIG, 0x700010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x3e1332);
+	WREG32(GMCON_PGFSM_CONFIG, 0xe00010ff);
+
+	WREG32(GMCON_MISC3, save->gmcon_misc3);
+	WREG32(GMCON_MISC, save->gmcon_misc);
+	WREG32(GMCON_RENG_EXECUTE, save->gmcon_reng_execute);
+}
+
+static void cik_gpu_pci_config_reset(struct radeon_device *rdev)
+{
+	struct evergreen_mc_save save;
+	struct kv_reset_save_regs kv_save = { 0 };
+	u32 tmp, i;
+
+	dev_info(rdev->dev, "GPU pci config reset\n");
+
+	/* disable dpm? */
+
+	/* disable cg/pg */
+	cik_fini_pg(rdev);
+	cik_fini_cg(rdev);
+
+	/* Disable GFX parsing/prefetching */
+	WREG32(CP_ME_CNTL, CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT);
+
+	/* Disable MEC parsing/prefetching */
+	WREG32(CP_MEC_CNTL, MEC_ME1_HALT | MEC_ME2_HALT);
+
+	/* sdma0 */
+	tmp = RREG32(SDMA0_ME_CNTL + SDMA0_REGISTER_OFFSET);
+	tmp |= SDMA_HALT;
+	WREG32(SDMA0_ME_CNTL + SDMA0_REGISTER_OFFSET, tmp);
+	/* sdma1 */
+	tmp = RREG32(SDMA0_ME_CNTL + SDMA1_REGISTER_OFFSET);
+	tmp |= SDMA_HALT;
+	WREG32(SDMA0_ME_CNTL + SDMA1_REGISTER_OFFSET, tmp);
+	/* XXX other engines? */
+
+	/* halt the rlc, disable cp internal ints */
+	cik_rlc_stop(rdev);
+
+	udelay(50);
+
+	/* disable mem access */
+	evergreen_mc_stop(rdev, &save);
+	if (evergreen_mc_wait_for_idle(rdev)) {
+		dev_warn(rdev->dev, "Wait for MC idle timed out !\n");
+	}
+
+	if (rdev->flags & RADEON_IS_IGP)
+		kv_save_regs_for_reset(rdev, &kv_save);
+
+	/* disable BM */
+	pci_clear_master(rdev->pdev);
+	/* reset */
+	radeon_pci_config_reset(rdev);
+
+	udelay(100);
+
+	/* wait for asic to come out of reset */
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CONFIG_MEMSIZE) != 0xffffffff)
+			break;
+		udelay(1);
+	}
+
+	/* does asic init need to be run first??? */
+	if (rdev->flags & RADEON_IS_IGP)
+		kv_restore_regs_for_reset(rdev, &kv_save);
+}
+
 /**
  * cik_asic_reset - soft reset GPU
  *
@@ -4870,10 +5195,17 @@ int cik_asic_reset(struct radeon_device *rdev)
 	if (reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, true);
 
+	/* try soft reset */
 	cik_gpu_soft_reset(rdev, reset_mask);
 
 	reset_mask = cik_gpu_check_soft_reset(rdev);
 
+	/* try pci config reset */
+	if (reset_mask && radeon_hard_reset)
+		cik_gpu_pci_config_reset(rdev);
+
+	reset_mask = cik_gpu_check_soft_reset(rdev);
+
 	if (!reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, false);
 
@@ -4896,11 +5228,9 @@ bool cik_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 	if (!(reset_mask & (RADEON_RESET_GFX |
 			    RADEON_RESET_COMPUTE |
 			    RADEON_RESET_CP))) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force CP activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
@@ -5077,6 +5407,7 @@ static int cik_pcie_gart_enable(struct radeon_device *rdev)
 	WREG32(MC_VM_MX_L1_TLB_CNTL,
 	       (0xA << 7) |
 	       ENABLE_L1_TLB |
+	       ENABLE_L1_FRAGMENT_PROCESSING |
 	       SYSTEM_ACCESS_MODE_NOT_IN_SYS |
 	       ENABLE_ADVANCED_DRIVER_MODEL |
 	       SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU);
@@ -5089,7 +5420,8 @@ static int cik_pcie_gart_enable(struct radeon_device *rdev)
 	       CONTEXT1_IDENTITY_ACCESS_MODE(1));
 	WREG32(VM_L2_CNTL2, INVALIDATE_ALL_L1_TLBS | INVALIDATE_L2_CACHE);
 	WREG32(VM_L2_CNTL3, L2_CACHE_BIGK_ASSOCIATIVITY |
-	       L2_CACHE_BIGK_FRAGMENT_SIZE(6));
+	       BANK_SELECT(4) |
+	       L2_CACHE_BIGK_FRAGMENT_SIZE(4));
 	/* setup context0 */
 	WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR, rdev->mc.gtt_start >> 12);
 	WREG32(VM_CONTEXT0_PAGE_TABLE_END_ADDR, rdev->mc.gtt_end >> 12);
@@ -5125,6 +5457,7 @@ static int cik_pcie_gart_enable(struct radeon_device *rdev)
 	       (u32)(rdev->dummy_page.addr >> 12));
 	WREG32(VM_CONTEXT1_CNTL2, 4);
 	WREG32(VM_CONTEXT1_CNTL, ENABLE_CONTEXT | PAGE_TABLE_DEPTH(1) |
+				PAGE_TABLE_BLOCK_SIZE(radeon_vm_block_size - 9) |
 				RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT |
 				RANGE_PROTECTION_FAULT_ENABLE_DEFAULT |
 				DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT |
@@ -5138,20 +5471,6 @@ static int cik_pcie_gart_enable(struct radeon_device *rdev)
 				WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT |
 				WRITE_PROTECTION_FAULT_ENABLE_DEFAULT);
 
-	/* TC cache setup ??? */
-	WREG32(TC_CFG_L1_LOAD_POLICY0, 0);
-	WREG32(TC_CFG_L1_LOAD_POLICY1, 0);
-	WREG32(TC_CFG_L1_STORE_POLICY, 0);
-
-	WREG32(TC_CFG_L2_LOAD_POLICY0, 0);
-	WREG32(TC_CFG_L2_LOAD_POLICY1, 0);
-	WREG32(TC_CFG_L2_STORE_POLICY0, 0);
-	WREG32(TC_CFG_L2_STORE_POLICY1, 0);
-	WREG32(TC_CFG_L2_ATOMIC_POLICY, 0);
-
-	WREG32(TC_CFG_L1_VOLATILE, 0);
-	WREG32(TC_CFG_L2_VOLATILE, 0);
-
 	if (rdev->family == CHIP_KAVERI) {
 		u32 tmp = RREG32(CHUB_CONTROL);
 		tmp &= ~BYPASS_VM;
@@ -5367,16 +5686,7 @@ void cik_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
 	radeon_ring_write(ring, VMID(0));
 
 	/* HDP flush */
-	/* We should be using the WAIT_REG_MEM packet here like in
-	 * cik_fence_ring_emit(), but it causes the CP to hang in this
-	 * context...
-	 */
-	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
-	radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) |
-				 WRITE_DATA_DST_SEL(0)));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 0);
+	cik_hdp_flush_cp_ring_emit(rdev, ridx);
 
 	/* bits 0-15 are the VM contexts0-15 */
 	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
@@ -5572,6 +5882,9 @@ static int cik_rlc_resume(struct radeon_device *rdev)
 	case CHIP_KABINI:
 		size = KB_RLC_UCODE_SIZE;
 		break;
+	case CHIP_MULLINS:
+		size = ML_RLC_UCODE_SIZE;
+		break;
 	}
 
 	cik_rlc_stop(rdev);
@@ -5943,6 +6256,10 @@ void cik_update_cg(struct radeon_device *rdev,
 		cik_enable_hdp_mgcg(rdev, enable);
 		cik_enable_hdp_ls(rdev, enable);
 	}
+
+	if (block & RADEON_CG_BLOCK_VCE) {
+		vce_v2_0_enable_mgcg(rdev, enable);
+	}
 }
 
 static void cik_init_cg(struct radeon_device *rdev)
@@ -6316,12 +6633,13 @@ void cik_get_csb_buffer(struct radeon_device *rdev, volatile u32 *buffer)
 		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
 	case CHIP_KABINI:
+	case CHIP_MULLINS:
 		buffer[count++] = cpu_to_le32(0x00000000); /* XXX */
 		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
 	case CHIP_HAWAII:
-		buffer[count++] = 0x3a00161a;
-		buffer[count++] = 0x0000002e;
+		buffer[count++] = cpu_to_le32(0x3a00161a);
+		buffer[count++] = cpu_to_le32(0x0000002e);
 		break;
 	default:
 		buffer[count++] = cpu_to_le32(0x00000000);
@@ -6461,6 +6779,19 @@ static void cik_disable_interrupt_state(struct radeon_device *rdev)
 		WREG32(LB_INTERRUPT_MASK + EVERGREEN_CRTC4_REGISTER_OFFSET, 0);
 		WREG32(LB_INTERRUPT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
 	}
+	/* pflip */
+	if (rdev->num_crtc >= 2) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, 0);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, 0);
+	}
+	if (rdev->num_crtc >= 4) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, 0);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, 0);
+	}
+	if (rdev->num_crtc >= 6) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, 0);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
+	}
 
 	/* dac hotplug */
 	WREG32(DAC_AUTODETECT_INT_CONTROL, 0);
@@ -6817,6 +7148,25 @@ int cik_irq_set(struct radeon_device *rdev)
 		WREG32(LB_INTERRUPT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, crtc6);
 	}
 
+	if (rdev->num_crtc >= 2) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+	}
+	if (rdev->num_crtc >= 4) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+	}
+	if (rdev->num_crtc >= 6) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+	}
+
 	WREG32(DC_HPD1_INT_CONTROL, hpd1);
 	WREG32(DC_HPD2_INT_CONTROL, hpd2);
 	WREG32(DC_HPD3_INT_CONTROL, hpd3);
@@ -6853,6 +7203,29 @@ static inline void cik_irq_ack(struct radeon_device *rdev)
 	rdev->irq.stat_regs.cik.disp_int_cont5 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE5);
 	rdev->irq.stat_regs.cik.disp_int_cont6 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE6);
 
+	rdev->irq.stat_regs.cik.d1grph_int = RREG32(GRPH_INT_STATUS +
+		EVERGREEN_CRTC0_REGISTER_OFFSET);
+	rdev->irq.stat_regs.cik.d2grph_int = RREG32(GRPH_INT_STATUS +
+		EVERGREEN_CRTC1_REGISTER_OFFSET);
+	if (rdev->num_crtc >= 4) {
+		rdev->irq.stat_regs.cik.d3grph_int = RREG32(GRPH_INT_STATUS +
+			EVERGREEN_CRTC2_REGISTER_OFFSET);
+		rdev->irq.stat_regs.cik.d4grph_int = RREG32(GRPH_INT_STATUS +
+			EVERGREEN_CRTC3_REGISTER_OFFSET);
+	}
+	if (rdev->num_crtc >= 6) {
+		rdev->irq.stat_regs.cik.d5grph_int = RREG32(GRPH_INT_STATUS +
+			EVERGREEN_CRTC4_REGISTER_OFFSET);
+		rdev->irq.stat_regs.cik.d6grph_int = RREG32(GRPH_INT_STATUS +
+			EVERGREEN_CRTC5_REGISTER_OFFSET);
+	}
+
+	if (rdev->irq.stat_regs.cik.d1grph_int & GRPH_PFLIP_INT_OCCURRED)
+		WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_CLEAR);
+	if (rdev->irq.stat_regs.cik.d2grph_int & GRPH_PFLIP_INT_OCCURRED)
+		WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_CLEAR);
 	if (rdev->irq.stat_regs.cik.disp_int & LB_D1_VBLANK_INTERRUPT)
 		WREG32(LB_VBLANK_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, VBLANK_ACK);
 	if (rdev->irq.stat_regs.cik.disp_int & LB_D1_VLINE_INTERRUPT)
@@ -6863,6 +7236,12 @@ static inline void cik_irq_ack(struct radeon_device *rdev)
 		WREG32(LB_VLINE_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, VLINE_ACK);
 
 	if (rdev->num_crtc >= 4) {
+		if (rdev->irq.stat_regs.cik.d3grph_int & GRPH_PFLIP_INT_OCCURRED)
+			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET,
+			       GRPH_PFLIP_INT_CLEAR);
+		if (rdev->irq.stat_regs.cik.d4grph_int & GRPH_PFLIP_INT_OCCURRED)
+			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET,
+			       GRPH_PFLIP_INT_CLEAR);
 		if (rdev->irq.stat_regs.cik.disp_int_cont2 & LB_D3_VBLANK_INTERRUPT)
 			WREG32(LB_VBLANK_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, VBLANK_ACK);
 		if (rdev->irq.stat_regs.cik.disp_int_cont2 & LB_D3_VLINE_INTERRUPT)
@@ -6874,6 +7253,12 @@ static inline void cik_irq_ack(struct radeon_device *rdev)
 	}
 
 	if (rdev->num_crtc >= 6) {
+		if (rdev->irq.stat_regs.cik.d5grph_int & GRPH_PFLIP_INT_OCCURRED)
+			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET,
+			       GRPH_PFLIP_INT_CLEAR);
+		if (rdev->irq.stat_regs.cik.d6grph_int & GRPH_PFLIP_INT_OCCURRED)
+			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET,
+			       GRPH_PFLIP_INT_CLEAR);
 		if (rdev->irq.stat_regs.cik.disp_int_cont4 & LB_D5_VBLANK_INTERRUPT)
 			WREG32(LB_VBLANK_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, VBLANK_ACK);
 		if (rdev->irq.stat_regs.cik.disp_int_cont4 & LB_D5_VLINE_INTERRUPT)
@@ -6992,6 +7377,7 @@ static inline u32 cik_get_ih_wptr(struct radeon_device *rdev)
 		tmp = RREG32(IH_RB_CNTL);
 		tmp |= IH_WPTR_OVERFLOW_CLEAR;
 		WREG32(IH_RB_CNTL, tmp);
+		wptr &= ~RB_OVERFLOW;
 	}
 	return (wptr & rdev->ih.ptr_mask);
 }
@@ -7079,7 +7465,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[0]))
-						radeon_crtc_handle_flip(rdev, 0);
+						radeon_crtc_handle_vblank(rdev, 0);
 					rdev->irq.stat_regs.cik.disp_int &= ~LB_D1_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D1 vblank\n");
 				}
@@ -7105,7 +7491,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[1]))
-						radeon_crtc_handle_flip(rdev, 1);
+						radeon_crtc_handle_vblank(rdev, 1);
 					rdev->irq.stat_regs.cik.disp_int_cont &= ~LB_D2_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D2 vblank\n");
 				}
@@ -7131,7 +7517,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[2]))
-						radeon_crtc_handle_flip(rdev, 2);
+						radeon_crtc_handle_vblank(rdev, 2);
 					rdev->irq.stat_regs.cik.disp_int_cont2 &= ~LB_D3_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D3 vblank\n");
 				}
@@ -7157,7 +7543,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[3]))
-						radeon_crtc_handle_flip(rdev, 3);
+						radeon_crtc_handle_vblank(rdev, 3);
 					rdev->irq.stat_regs.cik.disp_int_cont3 &= ~LB_D4_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D4 vblank\n");
 				}
@@ -7183,7 +7569,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[4]))
-						radeon_crtc_handle_flip(rdev, 4);
+						radeon_crtc_handle_vblank(rdev, 4);
 					rdev->irq.stat_regs.cik.disp_int_cont4 &= ~LB_D5_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D5 vblank\n");
 				}
@@ -7209,7 +7595,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[5]))
-						radeon_crtc_handle_flip(rdev, 5);
+						radeon_crtc_handle_vblank(rdev, 5);
 					rdev->irq.stat_regs.cik.disp_int_cont5 &= ~LB_D6_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D6 vblank\n");
 				}
@@ -7225,6 +7611,15 @@ restart_ih:
 				break;
 			}
 			break;
+		case 8: /* D1 page flip */
+		case 10: /* D2 page flip */
+		case 12: /* D3 page flip */
+		case 14: /* D4 page flip */
+		case 16: /* D5 page flip */
+		case 18: /* D6 page flip */
+			DRM_DEBUG("IH: D%d flip\n", ((src_id - 8) >> 1) + 1);
+			radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1);
+			break;
 		case 42: /* HPD hotplug */
 			switch (src_data) {
 			case 0:
@@ -7283,14 +7678,30 @@ restart_ih:
 			addr = RREG32(VM_CONTEXT1_PROTECTION_FAULT_ADDR);
 			status = RREG32(VM_CONTEXT1_PROTECTION_FAULT_STATUS);
 			mc_client = RREG32(VM_CONTEXT1_PROTECTION_FAULT_MCCLIENT);
+			/* reset addr and status */
+			WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1);
+			if (addr == 0x0 && status == 0x0)
+				break;
 			dev_err(rdev->dev, "GPU fault detected: %d 0x%08x\n", src_id, src_data);
 			dev_err(rdev->dev, "  VM_CONTEXT1_PROTECTION_FAULT_ADDR   0x%08X\n",
 				addr);
 			dev_err(rdev->dev, "  VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n",
 				status);
 			cik_vm_decode_fault(rdev, status, addr, mc_client);
-			/* reset addr and status */
-			WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1);
+			break;
+		case 167: /* VCE */
+			DRM_DEBUG("IH: VCE int: 0x%08x\n", src_data);
+			switch (src_data) {
+			case 0:
+				radeon_fence_process(rdev, TN_RING_TYPE_VCE1_INDEX);
+				break;
+			case 1:
+				radeon_fence_process(rdev, TN_RING_TYPE_VCE2_INDEX);
+				break;
+			default:
+				DRM_ERROR("Unhandled interrupt: %d %d\n", src_id, src_data);
+				break;
+			}
 			break;
 		case 176: /* GFX RB CP_INT */
 		case 177: /* GFX IB CP_INT */
@@ -7503,26 +7914,7 @@ static int cik_startup(struct radeon_device *rdev)
 
 	cik_mc_program(rdev);
 
-	if (rdev->flags & RADEON_IS_IGP) {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
-		    !rdev->mec_fw || !rdev->sdma_fw || !rdev->rlc_fw) {
-			r = cik_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
-	} else {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
-		    !rdev->mec_fw || !rdev->sdma_fw || !rdev->rlc_fw ||
-		    !rdev->mc_fw) {
-			r = cik_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
-
+	if (!(rdev->flags & RADEON_IS_IGP) && !rdev->pm.dpm_enabled) {
 		r = ci_mc_load_microcode(rdev);
 		if (r) {
 			DRM_ERROR("Failed to load MC firmware!\n");
@@ -7610,6 +8002,22 @@ static int cik_startup(struct radeon_device *rdev)
 	if (r)
 		rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
 
+	r = radeon_vce_resume(rdev);
+	if (!r) {
+		r = vce_v2_0_resume(rdev);
+		if (!r)
+			r = radeon_fence_driver_start_ring(rdev,
+							   TN_RING_TYPE_VCE1_INDEX);
+		if (!r)
+			r = radeon_fence_driver_start_ring(rdev,
+							   TN_RING_TYPE_VCE2_INDEX);
+	}
+	if (r) {
+		dev_err(rdev->dev, "VCE init error (%d).\n", r);
+		rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size = 0;
+		rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_size = 0;
+	}
+
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
 		r = radeon_irq_kms_init(rdev);
@@ -7627,7 +8035,6 @@ static int cik_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     CP_RB0_RPTR, CP_RB0_WPTR,
 			     PACKET3(PACKET3_NOP, 0x3FFF));
 	if (r)
 		return r;
@@ -7636,7 +8043,6 @@ static int cik_startup(struct radeon_device *rdev)
 	/* type-2 packets are deprecated on MEC, use type-3 instead */
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP1_RPTR_OFFSET,
-			     CP_HQD_PQ_RPTR, CP_HQD_PQ_WPTR,
 			     PACKET3(PACKET3_NOP, 0x3FFF));
 	if (r)
 		return r;
@@ -7648,7 +8054,6 @@ static int cik_startup(struct radeon_device *rdev)
 	/* type-2 packets are deprecated on MEC, use type-3 instead */
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP2_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP2_RPTR_OFFSET,
-			     CP_HQD_PQ_RPTR, CP_HQD_PQ_WPTR,
 			     PACKET3(PACKET3_NOP, 0x3FFF));
 	if (r)
 		return r;
@@ -7660,16 +8065,12 @@ static int cik_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     SDMA0_GFX_RB_RPTR + SDMA0_REGISTER_OFFSET,
-			     SDMA0_GFX_RB_WPTR + SDMA0_REGISTER_OFFSET,
 			     SDMA_PACKET(SDMA_OPCODE_NOP, 0, 0));
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, CAYMAN_WB_DMA1_RPTR_OFFSET,
-			     SDMA0_GFX_RB_RPTR + SDMA1_REGISTER_OFFSET,
-			     SDMA0_GFX_RB_WPTR + SDMA1_REGISTER_OFFSET,
 			     SDMA_PACKET(SDMA_OPCODE_NOP, 0, 0));
 	if (r)
 		return r;
@@ -7685,7 +8086,6 @@ static int cik_startup(struct radeon_device *rdev)
 	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 	if (ring->ring_size) {
 		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 				     RADEON_CP_PACKET2);
 		if (!r)
 			r = uvd_v1_0_init(rdev);
@@ -7693,6 +8093,23 @@ static int cik_startup(struct radeon_device *rdev)
 			DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
 	}
 
+	r = -ENOENT;
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+	if (ring->ring_size)
+		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
+				     VCE_CMD_NO_OP);
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+	if (ring->ring_size)
+		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
+				     VCE_CMD_NO_OP);
+
+	if (!r)
+		r = vce_v1_0_init(rdev);
+	else if (r != -ENOENT)
+		DRM_ERROR("radeon: failed initializing VCE (%d).\n", r);
+
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
 		dev_err(rdev->dev, "IB initialization failed (%d).\n", r);
@@ -7731,6 +8148,9 @@ int cik_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	cik_init_golden_registers(rdev);
 
+	if (rdev->pm.pm_method == PM_METHOD_DPM)
+		radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = cik_startup(rdev);
 	if (r) {
@@ -7754,12 +8174,14 @@ int cik_resume(struct radeon_device *rdev)
  */
 int cik_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	dce6_audio_fini(rdev);
 	radeon_vm_manager_fini(rdev);
 	cik_cp_enable(rdev, false);
 	cik_sdma_enable(rdev, false);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_suspend(rdev);
+	radeon_vce_suspend(rdev);
 	cik_fini_pg(rdev);
 	cik_fini_cg(rdev);
 	cik_irq_suspend(rdev);
@@ -7835,6 +8257,30 @@ int cik_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (rdev->flags & RADEON_IS_IGP) {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
+		    !rdev->mec_fw || !rdev->sdma_fw || !rdev->rlc_fw) {
+			r = cik_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	} else {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
+		    !rdev->mec_fw || !rdev->sdma_fw || !rdev->rlc_fw ||
+		    !rdev->mc_fw) {
+			r = cik_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	ring->ring_obj = NULL;
 	r600_ring_init(rdev, ring, 1024 * 1024);
@@ -7868,6 +8314,17 @@ int cik_init(struct radeon_device *rdev)
 		r600_ring_init(rdev, ring, 4096);
 	}
 
+	r = radeon_vce_init(rdev);
+	if (!r) {
+		ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+		ring->ring_obj = NULL;
+		r600_ring_init(rdev, ring, 4096);
+
+		ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+		ring->ring_obj = NULL;
+		r600_ring_init(rdev, ring, 4096);
+	}
+
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
 
@@ -7915,6 +8372,7 @@ int cik_init(struct radeon_device *rdev)
  */
 void cik_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	cik_cp_fini(rdev);
 	cik_sdma_fini(rdev);
 	cik_fini_pg(rdev);
@@ -7928,6 +8386,7 @@ void cik_fini(struct radeon_device *rdev)
 	radeon_irq_kms_fini(rdev);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_fini(rdev);
+	radeon_vce_fini(rdev);
 	cik_pcie_gart_fini(rdev);
 	r600_vram_scratch_fini(rdev);
 	radeon_gem_fini(rdev);
@@ -8666,6 +9125,41 @@ int cik_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk)
 	return r;
 }
 
+int cik_set_vce_clocks(struct radeon_device *rdev, u32 evclk, u32 ecclk)
+{
+	int r, i;
+	struct atom_clock_dividers dividers;
+	u32 tmp;
+
+	r = radeon_atom_get_clock_dividers(rdev, COMPUTE_GPUCLK_INPUT_FLAG_DEFAULT_GPUCLK,
+					   ecclk, false, &dividers);
+	if (r)
+		return r;
+
+	for (i = 0; i < 100; i++) {
+		if (RREG32_SMC(CG_ECLK_STATUS) & ECLK_STATUS)
+			break;
+		mdelay(10);
+	}
+	if (i == 100)
+		return -ETIMEDOUT;
+
+	tmp = RREG32_SMC(CG_ECLK_CNTL);
+	tmp &= ~(ECLK_DIR_CNTL_EN|ECLK_DIVIDER_MASK);
+	tmp |= dividers.post_divider;
+	WREG32_SMC(CG_ECLK_CNTL, tmp);
+
+	for (i = 0; i < 100; i++) {
+		if (RREG32_SMC(CG_ECLK_STATUS) & ECLK_STATUS)
+			break;
+		mdelay(10);
+	}
+	if (i == 100)
+		return -ETIMEDOUT;
+
+	return 0;
+}
+
 static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 {
 	struct pci_dev *root = rdev->pdev->bus->self;
diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c
index d08b83c6267..8e9d0f1d858 100644
--- a/drivers/gpu/drm/radeon/cik_sdma.c
+++ b/drivers/gpu/drm/radeon/cik_sdma.c
@@ -52,6 +52,75 @@ u32 cik_gpu_check_soft_reset(struct radeon_device *rdev);
  */
 
 /**
+ * cik_sdma_get_rptr - get the current read pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Get the current rptr from the hardware (CIK+).
+ */
+uint32_t cik_sdma_get_rptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	u32 rptr, reg;
+
+	if (rdev->wb.enabled) {
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	} else {
+		if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+			reg = SDMA0_GFX_RB_RPTR + SDMA0_REGISTER_OFFSET;
+		else
+			reg = SDMA0_GFX_RB_RPTR + SDMA1_REGISTER_OFFSET;
+
+		rptr = RREG32(reg);
+	}
+
+	return (rptr & 0x3fffc) >> 2;
+}
+
+/**
+ * cik_sdma_get_wptr - get the current write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Get the current wptr from the hardware (CIK+).
+ */
+uint32_t cik_sdma_get_wptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	u32 reg;
+
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		reg = SDMA0_GFX_RB_WPTR + SDMA0_REGISTER_OFFSET;
+	else
+		reg = SDMA0_GFX_RB_WPTR + SDMA1_REGISTER_OFFSET;
+
+	return (RREG32(reg) & 0x3fffc) >> 2;
+}
+
+/**
+ * cik_sdma_set_wptr - commit the write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Write the wptr back to the hardware (CIK+).
+ */
+void cik_sdma_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring)
+{
+	u32 reg;
+
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		reg = SDMA0_GFX_RB_WPTR + SDMA0_REGISTER_OFFSET;
+	else
+		reg = SDMA0_GFX_RB_WPTR + SDMA1_REGISTER_OFFSET;
+
+	WREG32(reg, (ring->wptr << 2) & 0x3fffc);
+}
+
+/**
  * cik_sdma_ring_ib_execute - Schedule an IB on the DMA engine
  *
  * @rdev: radeon_device pointer
@@ -72,7 +141,7 @@ void cik_sdma_ring_ib_execute(struct radeon_device *rdev,
 		next_rptr += 4;
 		radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0));
 		radeon_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc);
-		radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff);
+		radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr));
 		radeon_ring_write(ring, 1); /* number of DWs to follow */
 		radeon_ring_write(ring, next_rptr);
 	}
@@ -82,12 +151,41 @@ void cik_sdma_ring_ib_execute(struct radeon_device *rdev,
 		radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_NOP, 0, 0));
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_INDIRECT_BUFFER, 0, extra_bits));
 	radeon_ring_write(ring, ib->gpu_addr & 0xffffffe0); /* base must be 32 byte aligned */
-	radeon_ring_write(ring, upper_32_bits(ib->gpu_addr) & 0xffffffff);
+	radeon_ring_write(ring, upper_32_bits(ib->gpu_addr));
 	radeon_ring_write(ring, ib->length_dw);
 
 }
 
 /**
+ * cik_sdma_hdp_flush_ring_emit - emit an hdp flush on the DMA ring
+ *
+ * @rdev: radeon_device pointer
+ * @ridx: radeon ring index
+ *
+ * Emit an hdp flush packet on the requested DMA ring.
+ */
+static void cik_sdma_hdp_flush_ring_emit(struct radeon_device *rdev,
+					 int ridx)
+{
+	struct radeon_ring *ring = &rdev->ring[ridx];
+	u32 extra_bits = (SDMA_POLL_REG_MEM_EXTRA_OP(1) |
+			  SDMA_POLL_REG_MEM_EXTRA_FUNC(3)); /* == */
+	u32 ref_and_mask;
+
+	if (ridx == R600_RING_TYPE_DMA_INDEX)
+		ref_and_mask = SDMA0;
+	else
+		ref_and_mask = SDMA1;
+
+	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_POLL_REG_MEM, 0, extra_bits));
+	radeon_ring_write(ring, GPU_HDP_FLUSH_DONE);
+	radeon_ring_write(ring, GPU_HDP_FLUSH_REQ);
+	radeon_ring_write(ring, ref_and_mask); /* reference */
+	radeon_ring_write(ring, ref_and_mask); /* mask */
+	radeon_ring_write(ring, (0xfff << 16) | 10); /* retry count, poll interval */
+}
+
+/**
  * cik_sdma_fence_ring_emit - emit a fence on the DMA ring
  *
  * @rdev: radeon_device pointer
@@ -105,18 +203,13 @@ void cik_sdma_fence_ring_emit(struct radeon_device *rdev,
 
 	/* write the fence */
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_FENCE, 0, 0));
-	radeon_ring_write(ring, addr & 0xffffffff);
-	radeon_ring_write(ring, upper_32_bits(addr) & 0xffffffff);
+	radeon_ring_write(ring, lower_32_bits(addr));
+	radeon_ring_write(ring, upper_32_bits(addr));
 	radeon_ring_write(ring, fence->seq);
 	/* generate an interrupt */
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_TRAP, 0, 0));
 	/* flush HDP */
-	/* We should be using the new POLL_REG_MEM special op packet here
-	 * but it causes sDMA to hang sometimes
-	 */
-	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
+	cik_sdma_hdp_flush_ring_emit(rdev, fence->ring);
 }
 
 /**
@@ -140,7 +233,7 @@ bool cik_sdma_semaphore_ring_emit(struct radeon_device *rdev,
 
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SEMAPHORE, 0, extra_bits));
 	radeon_ring_write(ring, addr & 0xfffffff8);
-	radeon_ring_write(ring, upper_32_bits(addr) & 0xffffffff);
+	radeon_ring_write(ring, upper_32_bits(addr));
 
 	return true;
 }
@@ -157,7 +250,9 @@ static void cik_sdma_gfx_stop(struct radeon_device *rdev)
 	u32 rb_cntl, reg_offset;
 	int i;
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if ((rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX) ||
+	    (rdev->asic->copy.copy_ring_index == CAYMAN_RING_TYPE_DMA1_INDEX))
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 
 	for (i = 0; i < 2; i++) {
 		if (i == 0)
@@ -169,6 +264,8 @@ static void cik_sdma_gfx_stop(struct radeon_device *rdev)
 		WREG32(SDMA0_GFX_RB_CNTL + reg_offset, rb_cntl);
 		WREG32(SDMA0_GFX_IB_CNTL + reg_offset, 0);
 	}
+	rdev->ring[R600_RING_TYPE_DMA_INDEX].ready = false;
+	rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX].ready = false;
 }
 
 /**
@@ -196,6 +293,11 @@ void cik_sdma_enable(struct radeon_device *rdev, bool enable)
 	u32 me_cntl, reg_offset;
 	int i;
 
+	if (enable == false) {
+		cik_sdma_gfx_stop(rdev);
+		cik_sdma_rlc_stop(rdev);
+	}
+
 	for (i = 0; i < 2; i++) {
 		if (i == 0)
 			reg_offset = SDMA0_REGISTER_OFFSET;
@@ -267,8 +369,6 @@ static int cik_sdma_gfx_resume(struct radeon_device *rdev)
 		ring->wptr = 0;
 		WREG32(SDMA0_GFX_RB_WPTR + reg_offset, ring->wptr << 2);
 
-		ring->rptr = RREG32(SDMA0_GFX_RB_RPTR + reg_offset) >> 2;
-
 		/* enable DMA RB */
 		WREG32(SDMA0_GFX_RB_CNTL + reg_offset, rb_cntl | SDMA_RB_ENABLE);
 
@@ -288,7 +388,9 @@ static int cik_sdma_gfx_resume(struct radeon_device *rdev)
 		}
 	}
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+	if ((rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX) ||
+	    (rdev->asic->copy.copy_ring_index == CAYMAN_RING_TYPE_DMA1_INDEX))
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
 
 	return 0;
 }
@@ -323,10 +425,6 @@ static int cik_sdma_load_microcode(struct radeon_device *rdev)
 	if (!rdev->sdma_fw)
 		return -EINVAL;
 
-	/* stop the gfx rings and rlc compute queues */
-	cik_sdma_gfx_stop(rdev);
-	cik_sdma_rlc_stop(rdev);
-
 	/* halt the MEs */
 	cik_sdma_enable(rdev, false);
 
@@ -395,9 +493,6 @@ int cik_sdma_resume(struct radeon_device *rdev)
  */
 void cik_sdma_fini(struct radeon_device *rdev)
 {
-	/* stop the gfx rings and rlc compute queues */
-	cik_sdma_gfx_stop(rdev);
-	cik_sdma_rlc_stop(rdev);
 	/* halt the MEs */
 	cik_sdma_enable(rdev, false);
 	radeon_ring_fini(rdev, &rdev->ring[R600_RING_TYPE_DMA_INDEX]);
@@ -456,10 +551,10 @@ int cik_copy_dma(struct radeon_device *rdev,
 		radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_COPY, SDMA_COPY_SUB_OPCODE_LINEAR, 0));
 		radeon_ring_write(ring, cur_size_in_bytes);
 		radeon_ring_write(ring, 0); /* src/dst endian swap */
-		radeon_ring_write(ring, src_offset & 0xffffffff);
-		radeon_ring_write(ring, upper_32_bits(src_offset) & 0xffffffff);
-		radeon_ring_write(ring, dst_offset & 0xffffffff);
-		radeon_ring_write(ring, upper_32_bits(dst_offset) & 0xffffffff);
+		radeon_ring_write(ring, lower_32_bits(src_offset));
+		radeon_ring_write(ring, upper_32_bits(src_offset));
+		radeon_ring_write(ring, lower_32_bits(dst_offset));
+		radeon_ring_write(ring, upper_32_bits(dst_offset));
 		src_offset += cur_size_in_bytes;
 		dst_offset += cur_size_in_bytes;
 	}
@@ -467,6 +562,7 @@ int cik_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
@@ -502,14 +598,14 @@ int cik_sdma_ring_test(struct radeon_device *rdev,
 	tmp = 0xCAFEDEAD;
 	writel(tmp, ptr);
 
-	r = radeon_ring_lock(rdev, ring, 4);
+	r = radeon_ring_lock(rdev, ring, 5);
 	if (r) {
 		DRM_ERROR("radeon: dma failed to lock ring %d (%d).\n", ring->idx, r);
 		return r;
 	}
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0));
 	radeon_ring_write(ring, rdev->vram_scratch.gpu_addr & 0xfffffffc);
-	radeon_ring_write(ring, upper_32_bits(rdev->vram_scratch.gpu_addr) & 0xffffffff);
+	radeon_ring_write(ring, upper_32_bits(rdev->vram_scratch.gpu_addr));
 	radeon_ring_write(ring, 1); /* number of DWs to follow */
 	radeon_ring_write(ring, 0xDEADBEEF);
 	radeon_ring_unlock_commit(rdev, ring);
@@ -564,7 +660,7 @@ int cik_sdma_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 
 	ib.ptr[0] = SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0);
 	ib.ptr[1] = rdev->vram_scratch.gpu_addr & 0xfffffffc;
-	ib.ptr[2] = upper_32_bits(rdev->vram_scratch.gpu_addr) & 0xffffffff;
+	ib.ptr[2] = upper_32_bits(rdev->vram_scratch.gpu_addr);
 	ib.ptr[3] = 1;
 	ib.ptr[4] = 0xDEADBEEF;
 	ib.length_dw = 5;
@@ -616,11 +712,9 @@ bool cik_sdma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 		mask = RADEON_RESET_DMA1;
 
 	if (!(reset_mask & mask)) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force ring activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
@@ -648,7 +742,26 @@ void cik_sdma_vm_set_page(struct radeon_device *rdev,
 
 	trace_radeon_vm_set_page(pe, addr, count, incr, flags);
 
-	if (flags & R600_PTE_SYSTEM) {
+	if (flags == R600_PTE_GART) {
+		uint64_t src = rdev->gart.table_addr + (addr >> 12) * 8;
+		while (count) {
+			unsigned bytes = count * 8;
+			if (bytes > 0x1FFFF8)
+				bytes = 0x1FFFF8;
+
+			ib->ptr[ib->length_dw++] = SDMA_PACKET(SDMA_OPCODE_COPY, SDMA_WRITE_SUB_OPCODE_LINEAR, 0);
+			ib->ptr[ib->length_dw++] = bytes;
+			ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */
+			ib->ptr[ib->length_dw++] = lower_32_bits(src);
+			ib->ptr[ib->length_dw++] = upper_32_bits(src);
+			ib->ptr[ib->length_dw++] = lower_32_bits(pe);
+			ib->ptr[ib->length_dw++] = upper_32_bits(pe);
+
+			pe += bytes;
+			src += bytes;
+			count -= bytes / 8;
+		}
+	} else if (flags & R600_PTE_SYSTEM) {
 		while (count) {
 			ndw = count * 2;
 			if (ndw > 0xFFFFE)
@@ -747,12 +860,7 @@ void cik_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm
 	radeon_ring_write(ring, VMID(0));
 
 	/* flush HDP */
-	/* We should be using the new POLL_REG_MEM special op packet here
-	 * but it causes sDMA to hang sometimes
-	 */
-	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
+	cik_sdma_hdp_flush_ring_emit(rdev, ridx);
 
 	/* flush TLB */
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
diff --git a/drivers/gpu/drm/radeon/cikd.h b/drivers/gpu/drm/radeon/cikd.h
index 5964af5e5b2..0c6e1b55d96 100644
--- a/drivers/gpu/drm/radeon/cikd.h
+++ b/drivers/gpu/drm/radeon/cikd.h
@@ -203,6 +203,12 @@
 #define		CTF_TEMP_MASK				0x0003fe00
 #define		CTF_TEMP_SHIFT				9
 
+#define CG_ECLK_CNTL                                    0xC05000AC
+#       define ECLK_DIVIDER_MASK                        0x7f
+#       define ECLK_DIR_CNTL_EN                         (1 << 8)
+#define CG_ECLK_STATUS                                  0xC05000B0
+#       define ECLK_STATUS                              (1 << 0)
+
 #define	CG_SPLL_FUNC_CNTL				0xC0500140
 #define		SPLL_RESET				(1 << 0)
 #define		SPLL_PWRON				(1 << 1)
@@ -476,6 +482,7 @@
 #define		READ_PROTECTION_FAULT_ENABLE_DEFAULT		(1 << 16)
 #define		WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT		(1 << 18)
 #define		WRITE_PROTECTION_FAULT_ENABLE_DEFAULT		(1 << 19)
+#define		PAGE_TABLE_BLOCK_SIZE(x)			(((x) & 0xF) << 24)
 #define VM_CONTEXT1_CNTL				0x1414
 #define VM_CONTEXT0_CNTL2				0x1430
 #define VM_CONTEXT1_CNTL2				0x1434
@@ -724,6 +731,17 @@
 
 #define ATC_MISC_CG           				0x3350
 
+#define GMCON_RENG_EXECUTE				0x3508
+#define 	RENG_EXECUTE_ON_PWR_UP			(1 << 0)
+#define GMCON_MISC					0x350c
+#define 	RENG_EXECUTE_ON_REG_UPDATE		(1 << 11)
+#define 	STCTRL_STUTTER_EN			(1 << 16)
+
+#define GMCON_PGFSM_CONFIG				0x3538
+#define GMCON_PGFSM_WRITE				0x353c
+#define GMCON_PGFSM_READ				0x3540
+#define GMCON_MISC3					0x3544
+
 #define MC_SEQ_CNTL_3                                     0x3600
 #       define CAC_EN                                     (1 << 31)
 #define MC_SEQ_G5PDX_CTRL                                 0x3604
@@ -871,6 +889,15 @@
 #       define DC_HPD6_RX_INTERRUPT                     (1 << 18)
 #define DISP_INTERRUPT_STATUS_CONTINUE6                 0x6780
 
+/* 0x6858, 0x7458, 0x10058, 0x10c58, 0x11858, 0x12458 */
+#define GRPH_INT_STATUS                                 0x6858
+#       define GRPH_PFLIP_INT_OCCURRED                  (1 << 0)
+#       define GRPH_PFLIP_INT_CLEAR                     (1 << 8)
+/* 0x685c, 0x745c, 0x1005c, 0x10c5c, 0x1185c, 0x1245c */
+#define GRPH_INT_CONTROL                                0x685c
+#       define GRPH_PFLIP_INT_MASK                      (1 << 0)
+#       define GRPH_PFLIP_INT_TYPE                      (1 << 8)
+
 #define	DAC_AUTODETECT_INT_CONTROL			0x67c8
 
 #define DC_HPD1_INT_STATUS                              0x601c
@@ -1725,12 +1752,12 @@
 #define		EOP_TC_WB_ACTION_EN                     (1 << 15) /* L2 */
 #define		EOP_TCL1_ACTION_EN                      (1 << 16)
 #define		EOP_TC_ACTION_EN                        (1 << 17) /* L2 */
+#define		EOP_TCL2_VOLATILE                       (1 << 24)
 #define		EOP_CACHE_POLICY(x)                     ((x) << 25)
                 /* 0 - LRU
 		 * 1 - Stream
 		 * 2 - Bypass
 		 */
-#define		EOP_TCL2_VOLATILE                       (1 << 27)
 #define		DATA_SEL(x)                             ((x) << 29)
                 /* 0 - discard
 		 * 1 - send low 32bit data
@@ -1999,4 +2026,47 @@
 /* UVD CTX indirect */
 #define	UVD_CGC_MEM_CTRL				0xC0
 
+/* VCE */
+
+#define VCE_VCPU_CACHE_OFFSET0		0x20024
+#define VCE_VCPU_CACHE_SIZE0		0x20028
+#define VCE_VCPU_CACHE_OFFSET1		0x2002c
+#define VCE_VCPU_CACHE_SIZE1		0x20030
+#define VCE_VCPU_CACHE_OFFSET2		0x20034
+#define VCE_VCPU_CACHE_SIZE2		0x20038
+#define VCE_RB_RPTR2			0x20178
+#define VCE_RB_WPTR2			0x2017c
+#define VCE_RB_RPTR			0x2018c
+#define VCE_RB_WPTR			0x20190
+#define VCE_CLOCK_GATING_A		0x202f8
+#	define CGC_CLK_GATE_DLY_TIMER_MASK	(0xf << 0)
+#	define CGC_CLK_GATE_DLY_TIMER(x)	((x) << 0)
+#	define CGC_CLK_GATER_OFF_DLY_TIMER_MASK	(0xff << 4)
+#	define CGC_CLK_GATER_OFF_DLY_TIMER(x)	((x) << 4)
+#	define CGC_UENC_WAIT_AWAKE	(1 << 18)
+#define VCE_CLOCK_GATING_B		0x202fc
+#define VCE_CGTT_CLK_OVERRIDE		0x207a0
+#define VCE_UENC_CLOCK_GATING		0x207bc
+#	define CLOCK_ON_DELAY_MASK	(0xf << 0)
+#	define CLOCK_ON_DELAY(x)	((x) << 0)
+#	define CLOCK_OFF_DELAY_MASK	(0xff << 4)
+#	define CLOCK_OFF_DELAY(x)	((x) << 4)
+#define VCE_UENC_REG_CLOCK_GATING	0x207c0
+#define VCE_SYS_INT_EN			0x21300
+#	define VCE_SYS_INT_TRAP_INTERRUPT_EN	(1 << 3)
+#define VCE_LMI_CTRL2			0x21474
+#define VCE_LMI_CTRL			0x21498
+#define VCE_LMI_VM_CTRL			0x214a0
+#define VCE_LMI_SWAP_CNTL		0x214b4
+#define VCE_LMI_SWAP_CNTL1		0x214b8
+#define VCE_LMI_CACHE_CTRL		0x214f4
+
+#define VCE_CMD_NO_OP		0x00000000
+#define VCE_CMD_END		0x00000001
+#define VCE_CMD_IB		0x00000002
+#define VCE_CMD_FENCE		0x00000003
+#define VCE_CMD_TRAP		0x00000004
+#define VCE_CMD_IB_AUTO		0x00000005
+#define VCE_CMD_SEMAPHORE	0x00000006
+
 #endif
diff --git a/drivers/gpu/drm/radeon/clearstate_cayman.h b/drivers/gpu/drm/radeon/clearstate_cayman.h
index aa908c55a51..e48a14037b7 100644
--- a/drivers/gpu/drm/radeon/clearstate_cayman.h
+++ b/drivers/gpu/drm/radeon/clearstate_cayman.h
@@ -1050,7 +1050,7 @@ static const struct cs_extent_def SECT_CONTEXT_defs[] =
     {SECT_CONTEXT_def_5, 0x0000a29e, 5 },
     {SECT_CONTEXT_def_6, 0x0000a2a5, 56 },
     {SECT_CONTEXT_def_7, 0x0000a2de, 290 },
-    { 0, 0, 0 }
+    { NULL, 0, 0 }
 };
 static const u32 SECT_CLEAR_def_1[] =
 {
@@ -1061,7 +1061,7 @@ static const u32 SECT_CLEAR_def_1[] =
 static const struct cs_extent_def SECT_CLEAR_defs[] =
 {
     {SECT_CLEAR_def_1, 0x0000ffc0, 3 },
-    { 0, 0, 0 }
+    { NULL, 0, 0 }
 };
 static const u32 SECT_CTRLCONST_def_1[] =
 {
@@ -1071,11 +1071,11 @@ static const u32 SECT_CTRLCONST_def_1[] =
 static const struct cs_extent_def SECT_CTRLCONST_defs[] =
 {
     {SECT_CTRLCONST_def_1, 0x0000f3fc, 2 },
-    { 0, 0, 0 }
+    { NULL, 0, 0 }
 };
 static const struct cs_section_def cayman_cs_data[] = {
     { SECT_CONTEXT_defs, SECT_CONTEXT },
     { SECT_CLEAR_defs, SECT_CLEAR },
     { SECT_CTRLCONST_defs, SECT_CTRLCONST },
-    { 0, SECT_NONE }
+    { NULL, SECT_NONE }
 };
diff --git a/drivers/gpu/drm/radeon/clearstate_ci.h b/drivers/gpu/drm/radeon/clearstate_ci.h
index c3982f9475f..f55d06664e3 100644
--- a/drivers/gpu/drm/radeon/clearstate_ci.h
+++ b/drivers/gpu/drm/radeon/clearstate_ci.h
@@ -936,9 +936,9 @@ static const struct cs_extent_def ci_SECT_CONTEXT_defs[] =
     {ci_SECT_CONTEXT_def_5, 0x0000a2a0, 2 },
     {ci_SECT_CONTEXT_def_6, 0x0000a2a3, 1 },
     {ci_SECT_CONTEXT_def_7, 0x0000a2a5, 233 },
-    { 0, 0, 0 }
+    { NULL, 0, 0 }
 };
 static const struct cs_section_def ci_cs_data[] = {
     { ci_SECT_CONTEXT_defs, SECT_CONTEXT },
-    { 0, SECT_NONE }
+    { NULL, SECT_NONE }
 };
diff --git a/drivers/gpu/drm/radeon/clearstate_si.h b/drivers/gpu/drm/radeon/clearstate_si.h
index b994cb2a35a..66e39cdb5cb 100644
--- a/drivers/gpu/drm/radeon/clearstate_si.h
+++ b/drivers/gpu/drm/radeon/clearstate_si.h
@@ -933,9 +933,9 @@ static const struct cs_extent_def si_SECT_CONTEXT_defs[] =
     {si_SECT_CONTEXT_def_5, 0x0000a2a1, 1 },
     {si_SECT_CONTEXT_def_6, 0x0000a2a3, 1 },
     {si_SECT_CONTEXT_def_7, 0x0000a2a5, 233 },
-    { 0, 0, 0 }
+    { NULL, 0, 0 }
 };
 static const struct cs_section_def si_cs_data[] = {
     { si_SECT_CONTEXT_defs, SECT_CONTEXT },
-    { 0, SECT_NONE }
+    { NULL, SECT_NONE }
 };
diff --git a/drivers/gpu/drm/radeon/cypress_dpm.c b/drivers/gpu/drm/radeon/cypress_dpm.c
index 920e1e4a52c..47d31e91575 100644
--- a/drivers/gpu/drm/radeon/cypress_dpm.c
+++ b/drivers/gpu/drm/radeon/cypress_dpm.c
@@ -1551,7 +1551,7 @@ int cypress_populate_smc_voltage_tables(struct radeon_device *rdev,
 
 		table->voltageMaskTable.highMask[RV770_SMC_VOLTAGEMASK_VDDCI] = 0;
 		table->voltageMaskTable.lowMask[RV770_SMC_VOLTAGEMASK_VDDCI] =
-			cpu_to_be32(eg_pi->vddc_voltage_table.mask_low);
+			cpu_to_be32(eg_pi->vddci_voltage_table.mask_low);
 	}
 
 	return 0;
@@ -1905,21 +1905,6 @@ int cypress_dpm_enable(struct radeon_device *rdev)
 	if (pi->mg_clock_gating)
 		cypress_mg_clock_gating_enable(rdev, true);
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-		PPSMC_Result result;
-
-		ret = rv770_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-		result = rv770_send_msg_to_smc(rdev, PPSMC_MSG_EnableThermalInterrupt);
-
-		if (result != PPSMC_Result_OK)
-			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
-	}
-
 	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
 	return 0;
@@ -2051,6 +2036,10 @@ int cypress_dpm_init(struct radeon_device *rdev)
 	pi->min_vddc_in_table = 0;
 	pi->max_vddc_in_table = 0;
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = rv7xx_parse_power_table(rdev);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/radeon/dce3_1_afmt.c b/drivers/gpu/drm/radeon/dce3_1_afmt.c
new file mode 100644
index 00000000000..51800e340a5
--- /dev/null
+++ b/drivers/gpu/drm/radeon/dce3_1_afmt.c
@@ -0,0 +1,244 @@
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ * Copyright 2014 Rafał Miłecki
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include <linux/hdmi.h>
+#include <drm/drmP.h>
+#include "radeon.h"
+#include "radeon_asic.h"
+#include "r600d.h"
+
+static void dce3_2_afmt_write_speaker_allocation(struct drm_encoder *encoder)
+{
+	struct radeon_device *rdev = encoder->dev->dev_private;
+	struct drm_connector *connector;
+	struct radeon_connector *radeon_connector = NULL;
+	u32 tmp;
+	u8 *sadb;
+	int sad_count;
+
+	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+		if (connector->encoder == encoder) {
+			radeon_connector = to_radeon_connector(connector);
+			break;
+		}
+	}
+
+	if (!radeon_connector) {
+		DRM_ERROR("Couldn't find encoder's connector\n");
+		return;
+	}
+
+	sad_count = drm_edid_to_speaker_allocation(radeon_connector->edid, &sadb);
+	if (sad_count < 0) {
+		DRM_ERROR("Couldn't read Speaker Allocation Data Block: %d\n", sad_count);
+		return;
+	}
+
+	/* program the speaker allocation */
+	tmp = RREG32(AZ_F0_CODEC_PIN0_CONTROL_CHANNEL_SPEAKER);
+	tmp &= ~(DP_CONNECTION | SPEAKER_ALLOCATION_MASK);
+	/* set HDMI mode */
+	tmp |= HDMI_CONNECTION;
+	if (sad_count)
+		tmp |= SPEAKER_ALLOCATION(sadb[0]);
+	else
+		tmp |= SPEAKER_ALLOCATION(5); /* stereo */
+	WREG32(AZ_F0_CODEC_PIN0_CONTROL_CHANNEL_SPEAKER, tmp);
+
+	kfree(sadb);
+}
+
+static void dce3_2_afmt_write_sad_regs(struct drm_encoder *encoder)
+{
+	struct radeon_device *rdev = encoder->dev->dev_private;
+	struct drm_connector *connector;
+	struct radeon_connector *radeon_connector = NULL;
+	struct cea_sad *sads;
+	int i, sad_count;
+
+	static const u16 eld_reg_to_type[][2] = {
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR0, HDMI_AUDIO_CODING_TYPE_PCM },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR1, HDMI_AUDIO_CODING_TYPE_AC3 },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR2, HDMI_AUDIO_CODING_TYPE_MPEG1 },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR3, HDMI_AUDIO_CODING_TYPE_MP3 },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR4, HDMI_AUDIO_CODING_TYPE_MPEG2 },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR5, HDMI_AUDIO_CODING_TYPE_AAC_LC },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR6, HDMI_AUDIO_CODING_TYPE_DTS },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR7, HDMI_AUDIO_CODING_TYPE_ATRAC },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR9, HDMI_AUDIO_CODING_TYPE_EAC3 },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR10, HDMI_AUDIO_CODING_TYPE_DTS_HD },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR11, HDMI_AUDIO_CODING_TYPE_MLP },
+		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR13, HDMI_AUDIO_CODING_TYPE_WMA_PRO },
+	};
+
+	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
+		if (connector->encoder == encoder) {
+			radeon_connector = to_radeon_connector(connector);
+			break;
+		}
+	}
+
+	if (!radeon_connector) {
+		DRM_ERROR("Couldn't find encoder's connector\n");
+		return;
+	}
+
+	sad_count = drm_edid_to_sad(radeon_connector->edid, &sads);
+	if (sad_count < 0) {
+		DRM_ERROR("Couldn't read SADs: %d\n", sad_count);
+		return;
+	}
+	BUG_ON(!sads);
+
+	for (i = 0; i < ARRAY_SIZE(eld_reg_to_type); i++) {
+		u32 value = 0;
+		u8 stereo_freqs = 0;
+		int max_channels = -1;
+		int j;
+
+		for (j = 0; j < sad_count; j++) {
+			struct cea_sad *sad = &sads[j];
+
+			if (sad->format == eld_reg_to_type[i][1]) {
+				if (sad->channels > max_channels) {
+					value = MAX_CHANNELS(sad->channels) |
+						DESCRIPTOR_BYTE_2(sad->byte2) |
+						SUPPORTED_FREQUENCIES(sad->freq);
+					max_channels = sad->channels;
+				}
+
+				if (sad->format == HDMI_AUDIO_CODING_TYPE_PCM)
+					stereo_freqs |= sad->freq;
+				else
+					break;
+			}
+		}
+
+		value |= SUPPORTED_FREQUENCIES_STEREO(stereo_freqs);
+
+		WREG32(eld_reg_to_type[i][0], value);
+	}
+
+	kfree(sads);
+}
+
+/*
+ * update the info frames with the data from the current display mode
+ */
+void dce3_1_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mode)
+{
+	struct drm_device *dev = encoder->dev;
+	struct radeon_device *rdev = dev->dev_private;
+	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
+	struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv;
+	u8 buffer[HDMI_INFOFRAME_HEADER_SIZE + HDMI_AVI_INFOFRAME_SIZE];
+	struct hdmi_avi_infoframe frame;
+	uint32_t offset;
+	ssize_t err;
+
+	if (!dig || !dig->afmt)
+		return;
+
+	/* Silent, r600_hdmi_enable will raise WARN for us */
+	if (!dig->afmt->enabled)
+		return;
+	offset = dig->afmt->offset;
+
+	/* disable audio prior to setting up hw */
+	dig->afmt->pin = r600_audio_get_pin(rdev);
+	r600_audio_enable(rdev, dig->afmt->pin, false);
+
+	r600_audio_set_dto(encoder, mode->clock);
+
+	WREG32(HDMI0_VBI_PACKET_CONTROL + offset,
+	       HDMI0_NULL_SEND); /* send null packets when required */
+
+	WREG32(HDMI0_AUDIO_CRC_CONTROL + offset, 0x1000);
+
+	if (ASIC_IS_DCE32(rdev)) {
+		WREG32(HDMI0_AUDIO_PACKET_CONTROL + offset,
+		       HDMI0_AUDIO_DELAY_EN(1) | /* default audio delay */
+		       HDMI0_AUDIO_PACKETS_PER_LINE(3)); /* should be suffient for all audio modes and small enough for all hblanks */
+		WREG32(AFMT_AUDIO_PACKET_CONTROL + offset,
+		       AFMT_AUDIO_SAMPLE_SEND | /* send audio packets */
+		       AFMT_60958_CS_UPDATE); /* allow 60958 channel status fields to be updated */
+	} else {
+		WREG32(HDMI0_AUDIO_PACKET_CONTROL + offset,
+		       HDMI0_AUDIO_SAMPLE_SEND | /* send audio packets */
+		       HDMI0_AUDIO_DELAY_EN(1) | /* default audio delay */
+		       HDMI0_AUDIO_PACKETS_PER_LINE(3) | /* should be suffient for all audio modes and small enough for all hblanks */
+		       HDMI0_60958_CS_UPDATE); /* allow 60958 channel status fields to be updated */
+	}
+
+	if (ASIC_IS_DCE32(rdev)) {
+		dce3_2_afmt_write_speaker_allocation(encoder);
+		dce3_2_afmt_write_sad_regs(encoder);
+	}
+
+	WREG32(HDMI0_ACR_PACKET_CONTROL + offset,
+	       HDMI0_ACR_SOURCE | /* select SW CTS value - XXX verify that hw CTS works on all families */
+	       HDMI0_ACR_AUTO_SEND); /* allow hw to sent ACR packets when required */
+
+	WREG32(HDMI0_VBI_PACKET_CONTROL + offset,
+	       HDMI0_NULL_SEND | /* send null packets when required */
+	       HDMI0_GC_SEND | /* send general control packets */
+	       HDMI0_GC_CONT); /* send general control packets every frame */
+
+	/* TODO: HDMI0_AUDIO_INFO_UPDATE */
+	WREG32(HDMI0_INFOFRAME_CONTROL0 + offset,
+	       HDMI0_AVI_INFO_SEND | /* enable AVI info frames */
+	       HDMI0_AVI_INFO_CONT | /* send AVI info frames every frame/field */
+	       HDMI0_AUDIO_INFO_SEND | /* enable audio info frames (frames won't be set until audio is enabled) */
+	       HDMI0_AUDIO_INFO_CONT); /* send audio info frames every frame/field */
+
+	WREG32(HDMI0_INFOFRAME_CONTROL1 + offset,
+	       HDMI0_AVI_INFO_LINE(2) | /* anything other than 0 */
+	       HDMI0_AUDIO_INFO_LINE(2)); /* anything other than 0 */
+
+	WREG32(HDMI0_GC + offset, 0); /* unset HDMI0_GC_AVMUTE */
+
+	err = drm_hdmi_avi_infoframe_from_display_mode(&frame, mode);
+	if (err < 0) {
+		DRM_ERROR("failed to setup AVI infoframe: %zd\n", err);
+		return;
+	}
+
+	err = hdmi_avi_infoframe_pack(&frame, buffer, sizeof(buffer));
+	if (err < 0) {
+		DRM_ERROR("failed to pack AVI infoframe: %zd\n", err);
+		return;
+	}
+
+	r600_hdmi_update_avi_infoframe(encoder, buffer, sizeof(buffer));
+	r600_hdmi_update_ACR(encoder, mode->clock);
+
+	/* it's unknown what these bits do excatly, but it's indeed quite useful for debugging */
+	WREG32(HDMI0_RAMP_CONTROL0 + offset, 0x00FFFFFF);
+	WREG32(HDMI0_RAMP_CONTROL1 + offset, 0x007FFFFF);
+	WREG32(HDMI0_RAMP_CONTROL2 + offset, 0x00000001);
+	WREG32(HDMI0_RAMP_CONTROL3 + offset, 0x00000001);
+
+	r600_hdmi_audio_workaround(encoder);
+
+	/* enable audio after to setting up hw */
+	r600_audio_enable(rdev, dig->afmt->pin, true);
+}
diff --git a/drivers/gpu/drm/radeon/dce6_afmt.c b/drivers/gpu/drm/radeon/dce6_afmt.c
index 713a5d35990..0a65dc7e93e 100644
--- a/drivers/gpu/drm/radeon/dce6_afmt.c
+++ b/drivers/gpu/drm/radeon/dce6_afmt.c
@@ -278,13 +278,15 @@ static int dce6_audio_chipset_supported(struct radeon_device *rdev)
 	return !ASIC_IS_NODCE(rdev);
 }
 
-static void dce6_audio_enable(struct radeon_device *rdev,
-			      struct r600_audio_pin *pin,
-			      bool enable)
+void dce6_audio_enable(struct radeon_device *rdev,
+		       struct r600_audio_pin *pin,
+		       bool enable)
 {
+	if (!pin)
+		return;
+
 	WREG32_ENDPOINT(pin->offset, AZ_F0_CODEC_PIN_CONTROL_HOTPLUG_CONTROL,
-			AUDIO_ENABLED);
-	DRM_INFO("%s audio %d support\n", enable ? "Enabling" : "Disabling", pin->id);
+			enable ? AUDIO_ENABLED : 0);
 }
 
 static const u32 pin_offsets[7] =
@@ -307,11 +309,17 @@ int dce6_audio_init(struct radeon_device *rdev)
 
 	rdev->audio.enabled = true;
 
-	if (ASIC_IS_DCE8(rdev))
+	if (ASIC_IS_DCE81(rdev)) /* KV: 4 streams, 7 endpoints */
+		rdev->audio.num_pins = 7;
+	else if (ASIC_IS_DCE83(rdev)) /* KB: 2 streams, 3 endpoints */
+		rdev->audio.num_pins = 3;
+	else if (ASIC_IS_DCE8(rdev)) /* BN/HW: 6 streams, 7 endpoints */
+		rdev->audio.num_pins = 7;
+	else if (ASIC_IS_DCE61(rdev)) /* TN: 4 streams, 6 endpoints */
 		rdev->audio.num_pins = 6;
-	else if (ASIC_IS_DCE61(rdev))
-		rdev->audio.num_pins = 4;
-	else
+	else if (ASIC_IS_DCE64(rdev)) /* OL: 2 streams, 2 endpoints */
+		rdev->audio.num_pins = 2;
+	else /* SI: 6 streams, 6 endpoints */
 		rdev->audio.num_pins = 6;
 
 	for (i = 0; i < rdev->audio.num_pins; i++) {
@@ -323,7 +331,8 @@ int dce6_audio_init(struct radeon_device *rdev)
 		rdev->audio.pin[i].connected = false;
 		rdev->audio.pin[i].offset = pin_offsets[i];
 		rdev->audio.pin[i].id = i;
-		dce6_audio_enable(rdev, &rdev->audio.pin[i], true);
+		/* disable audio.  it will be set up later */
+		dce6_audio_enable(rdev, &rdev->audio.pin[i], false);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index 9702e55e924..15e4f28015e 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -146,6 +146,7 @@ extern u32 si_get_csb_size(struct radeon_device *rdev);
 extern void si_get_csb_buffer(struct radeon_device *rdev, volatile u32 *buffer);
 extern u32 cik_get_csb_size(struct radeon_device *rdev);
 extern void cik_get_csb_buffer(struct radeon_device *rdev, volatile u32 *buffer);
+extern void rv770_set_clk_bypass_mode(struct radeon_device *rdev);
 
 static const u32 evergreen_golden_registers[] =
 {
@@ -188,7 +189,7 @@ static const u32 evergreen_golden_registers[] =
 	0x8c1c, 0xffffffff, 0x00001010,
 	0x28350, 0xffffffff, 0x00000000,
 	0xa008, 0xffffffff, 0x00010000,
-	0x5cc, 0xffffffff, 0x00000001,
+	0x5c4, 0xffffffff, 0x00000001,
 	0x9508, 0xffffffff, 0x00000002,
 	0x913c, 0x0000000f, 0x0000000a
 };
@@ -475,7 +476,7 @@ static const u32 cedar_golden_registers[] =
 	0x8c1c, 0xffffffff, 0x00001010,
 	0x28350, 0xffffffff, 0x00000000,
 	0xa008, 0xffffffff, 0x00010000,
-	0x5cc, 0xffffffff, 0x00000001,
+	0x5c4, 0xffffffff, 0x00000001,
 	0x9508, 0xffffffff, 0x00000002
 };
 
@@ -634,7 +635,7 @@ static const u32 juniper_mgcg_init[] =
 static const u32 supersumo_golden_registers[] =
 {
 	0x5eb4, 0xffffffff, 0x00000002,
-	0x5cc, 0xffffffff, 0x00000001,
+	0x5c4, 0xffffffff, 0x00000001,
 	0x7030, 0xffffffff, 0x00000011,
 	0x7c30, 0xffffffff, 0x00000011,
 	0x6104, 0x01000300, 0x00000000,
@@ -718,7 +719,7 @@ static const u32 sumo_golden_registers[] =
 static const u32 wrestler_golden_registers[] =
 {
 	0x5eb4, 0xffffffff, 0x00000002,
-	0x5cc, 0xffffffff, 0x00000001,
+	0x5c4, 0xffffffff, 0x00000001,
 	0x7030, 0xffffffff, 0x00000011,
 	0x7c30, 0xffffffff, 0x00000011,
 	0x6104, 0x01000300, 0x00000000,
@@ -1300,36 +1301,6 @@ void dce4_wait_for_vblank(struct radeon_device *rdev, int crtc)
 }
 
 /**
- * radeon_irq_kms_pflip_irq_get - pre-pageflip callback.
- *
- * @rdev: radeon_device pointer
- * @crtc: crtc to prepare for pageflip on
- *
- * Pre-pageflip callback (evergreen+).
- * Enables the pageflip irq (vblank irq).
- */
-void evergreen_pre_page_flip(struct radeon_device *rdev, int crtc)
-{
-	/* enable the pflip int */
-	radeon_irq_kms_pflip_irq_get(rdev, crtc);
-}
-
-/**
- * evergreen_post_page_flip - pos-pageflip callback.
- *
- * @rdev: radeon_device pointer
- * @crtc: crtc to cleanup pageflip on
- *
- * Post-pageflip callback (evergreen+).
- * Disables the pageflip irq (vblank irq).
- */
-void evergreen_post_page_flip(struct radeon_device *rdev, int crtc)
-{
-	/* disable the pflip int */
-	radeon_irq_kms_pflip_irq_put(rdev, crtc);
-}
-
-/**
  * evergreen_page_flip - pageflip callback.
  *
  * @rdev: radeon_device pointer
@@ -1342,7 +1313,7 @@ void evergreen_post_page_flip(struct radeon_device *rdev, int crtc)
  * double buffered update to take place.
  * Returns the current update pending status.
  */
-u32 evergreen_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
+void evergreen_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 {
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 	u32 tmp = RREG32(EVERGREEN_GRPH_UPDATE + radeon_crtc->crtc_offset);
@@ -1374,9 +1345,23 @@ u32 evergreen_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 	/* Unlock the lock, so double-buffering can take place inside vblank */
 	tmp &= ~EVERGREEN_GRPH_UPDATE_LOCK;
 	WREG32(EVERGREEN_GRPH_UPDATE + radeon_crtc->crtc_offset, tmp);
+}
+
+/**
+ * evergreen_page_flip_pending - check if page flip is still pending
+ *
+ * @rdev: radeon_device pointer
+ * @crtc_id: crtc to check
+ *
+ * Returns the current update pending status.
+ */
+bool evergreen_page_flip_pending(struct radeon_device *rdev, int crtc_id)
+{
+	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 
 	/* Return current update_pending status: */
-	return RREG32(EVERGREEN_GRPH_UPDATE + radeon_crtc->crtc_offset) & EVERGREEN_GRPH_SURFACE_UPDATE_PENDING;
+	return !!(RREG32(EVERGREEN_GRPH_UPDATE + radeon_crtc->crtc_offset) &
+		EVERGREEN_GRPH_SURFACE_UPDATE_PENDING);
 }
 
 /* get temperature in millidegrees */
@@ -1679,7 +1664,7 @@ bool evergreen_hpd_sense(struct radeon_device *rdev, enum radeon_hpd_id hpd)
 	case RADEON_HPD_6:
 		if (RREG32(DC_HPD6_INT_STATUS) & DC_HPDx_SENSE)
 			connected = true;
-			break;
+		break;
 	default:
 		break;
 	}
@@ -2657,8 +2642,9 @@ void evergreen_mc_resume(struct radeon_device *rdev, struct evergreen_mc_save *s
 	for (i = 0; i < rdev->num_crtc; i++) {
 		if (save->crtc_enabled[i]) {
 			tmp = RREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i]);
-			if ((tmp & 0x3) != 0) {
-				tmp &= ~0x3;
+			if ((tmp & 0x7) != 3) {
+				tmp &= ~0x7;
+				tmp |= 0x3;
 				WREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i], tmp);
 			}
 			tmp = RREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i]);
@@ -2989,8 +2975,6 @@ static int evergreen_cp_resume(struct radeon_device *rdev)
 	WREG32(CP_RB_BASE, ring->gpu_addr >> 8);
 	WREG32(CP_DEBUG, (1 << 27) | (1 << 28));
 
-	ring->rptr = RREG32(CP_RB_RPTR);
-
 	evergreen_cp_start(rdev);
 	ring->ready = true;
 	r = radeon_ring_test(rdev, RADEON_RING_TYPE_GFX_INDEX, ring);
@@ -3354,6 +3338,18 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
 			disabled_rb_mask &= ~(1 << i);
 	}
 
+	for (i = 0; i < rdev->config.evergreen.num_ses; i++) {
+		u32 simd_disable_bitmap;
+
+		WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i));
+		WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i));
+		simd_disable_bitmap = (RREG32(CC_GC_SHADER_PIPE_CONFIG) & 0xffff0000) >> 16;
+		simd_disable_bitmap |= 0xffffffff << rdev->config.evergreen.max_simds;
+		tmp <<= 16;
+		tmp |= simd_disable_bitmap;
+	}
+	rdev->config.evergreen.active_simds = hweight32(~tmp);
+
 	WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES);
 	WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES);
 
@@ -3867,6 +3863,48 @@ static void evergreen_gpu_soft_reset(struct radeon_device *rdev, u32 reset_mask)
 	evergreen_print_gpu_status_regs(rdev);
 }
 
+void evergreen_gpu_pci_config_reset(struct radeon_device *rdev)
+{
+	struct evergreen_mc_save save;
+	u32 tmp, i;
+
+	dev_info(rdev->dev, "GPU pci config reset\n");
+
+	/* disable dpm? */
+
+	/* Disable CP parsing/prefetching */
+	WREG32(CP_ME_CNTL, CP_ME_HALT | CP_PFP_HALT);
+	udelay(50);
+	/* Disable DMA */
+	tmp = RREG32(DMA_RB_CNTL);
+	tmp &= ~DMA_RB_ENABLE;
+	WREG32(DMA_RB_CNTL, tmp);
+	/* XXX other engines? */
+
+	/* halt the rlc */
+	r600_rlc_stop(rdev);
+
+	udelay(50);
+
+	/* set mclk/sclk to bypass */
+	rv770_set_clk_bypass_mode(rdev);
+	/* disable BM */
+	pci_clear_master(rdev->pdev);
+	/* disable mem access */
+	evergreen_mc_stop(rdev, &save);
+	if (evergreen_mc_wait_for_idle(rdev)) {
+		dev_warn(rdev->dev, "Wait for MC idle timed out !\n");
+	}
+	/* reset */
+	radeon_pci_config_reset(rdev);
+	/* wait for asic to come out of reset */
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CONFIG_MEMSIZE) != 0xffffffff)
+			break;
+		udelay(1);
+	}
+}
+
 int evergreen_asic_reset(struct radeon_device *rdev)
 {
 	u32 reset_mask;
@@ -3876,10 +3914,17 @@ int evergreen_asic_reset(struct radeon_device *rdev)
 	if (reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, true);
 
+	/* try soft reset */
 	evergreen_gpu_soft_reset(rdev, reset_mask);
 
 	reset_mask = evergreen_gpu_check_soft_reset(rdev);
 
+	/* try pci config reset */
+	if (reset_mask && radeon_hard_reset)
+		evergreen_gpu_pci_config_reset(rdev);
+
+	reset_mask = evergreen_gpu_check_soft_reset(rdev);
+
 	if (!reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, false);
 
@@ -3902,11 +3947,9 @@ bool evergreen_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *rin
 	if (!(reset_mask & (RADEON_RESET_GFX |
 			    RADEON_RESET_COMPUTE |
 			    RADEON_RESET_CP))) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force CP activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
@@ -4298,8 +4341,8 @@ void evergreen_disable_interrupt_state(struct radeon_device *rdev)
 		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
 	}
 
-	/* only one DAC on DCE6 */
-	if (!ASIC_IS_DCE6(rdev))
+	/* only one DAC on DCE5 */
+	if (!ASIC_IS_DCE5(rdev))
 		WREG32(DACA_AUTODETECT_INT_CONTROL, 0);
 	WREG32(DACB_AUTODETECT_INT_CONTROL, 0);
 
@@ -4325,7 +4368,6 @@ int evergreen_irq_set(struct radeon_device *rdev)
 	u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0;
 	u32 hpd1, hpd2, hpd3, hpd4, hpd5, hpd6;
 	u32 grbm_int_cntl = 0;
-	u32 grph1 = 0, grph2 = 0, grph3 = 0, grph4 = 0, grph5 = 0, grph6 = 0;
 	u32 afmt1 = 0, afmt2 = 0, afmt3 = 0, afmt4 = 0, afmt5 = 0, afmt6 = 0;
 	u32 dma_cntl, dma_cntl1 = 0;
 	u32 thermal_int = 0;
@@ -4508,15 +4550,21 @@ int evergreen_irq_set(struct radeon_device *rdev)
 		WREG32(INT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, crtc6);
 	}
 
-	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, grph1);
-	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, grph2);
+	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET,
+	       GRPH_PFLIP_INT_MASK);
+	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET,
+	       GRPH_PFLIP_INT_MASK);
 	if (rdev->num_crtc >= 4) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, grph3);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, grph4);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 	if (rdev->num_crtc >= 6) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, grph5);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, grph6);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 
 	WREG32(DC_HPD1_INT_CONTROL, hpd1);
@@ -4708,6 +4756,7 @@ static u32 evergreen_get_ih_wptr(struct radeon_device *rdev)
 		tmp = RREG32(IH_RB_CNTL);
 		tmp |= IH_WPTR_OVERFLOW_CLEAR;
 		WREG32(IH_RB_CNTL, tmp);
+		wptr &= ~RB_OVERFLOW;
 	}
 	return (wptr & rdev->ih.ptr_mask);
 }
@@ -4759,7 +4808,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[0]))
-						radeon_crtc_handle_flip(rdev, 0);
+						radeon_crtc_handle_vblank(rdev, 0);
 					rdev->irq.stat_regs.evergreen.disp_int &= ~LB_D1_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D1 vblank\n");
 				}
@@ -4785,7 +4834,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[1]))
-						radeon_crtc_handle_flip(rdev, 1);
+						radeon_crtc_handle_vblank(rdev, 1);
 					rdev->irq.stat_regs.evergreen.disp_int_cont &= ~LB_D2_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D2 vblank\n");
 				}
@@ -4811,7 +4860,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[2]))
-						radeon_crtc_handle_flip(rdev, 2);
+						radeon_crtc_handle_vblank(rdev, 2);
 					rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~LB_D3_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D3 vblank\n");
 				}
@@ -4837,7 +4886,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[3]))
-						radeon_crtc_handle_flip(rdev, 3);
+						radeon_crtc_handle_vblank(rdev, 3);
 					rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~LB_D4_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D4 vblank\n");
 				}
@@ -4863,7 +4912,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[4]))
-						radeon_crtc_handle_flip(rdev, 4);
+						radeon_crtc_handle_vblank(rdev, 4);
 					rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~LB_D5_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D5 vblank\n");
 				}
@@ -4889,7 +4938,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[5]))
-						radeon_crtc_handle_flip(rdev, 5);
+						radeon_crtc_handle_vblank(rdev, 5);
 					rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~LB_D6_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D6 vblank\n");
 				}
@@ -4905,6 +4954,15 @@ restart_ih:
 				break;
 			}
 			break;
+		case 8: /* D1 page flip */
+		case 10: /* D2 page flip */
+		case 12: /* D3 page flip */
+		case 14: /* D4 page flip */
+		case 16: /* D5 page flip */
+		case 18: /* D6 page flip */
+			DRM_DEBUG("IH: D%d flip\n", ((src_id - 8) >> 1) + 1);
+			radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1);
+			break;
 		case 42: /* HPD hotplug */
 			switch (src_data) {
 			case 0:
@@ -5010,14 +5068,16 @@ restart_ih:
 		case 147:
 			addr = RREG32(VM_CONTEXT1_PROTECTION_FAULT_ADDR);
 			status = RREG32(VM_CONTEXT1_PROTECTION_FAULT_STATUS);
+			/* reset addr and status */
+			WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1);
+			if (addr == 0x0 && status == 0x0)
+				break;
 			dev_err(rdev->dev, "GPU fault detected: %d 0x%08x\n", src_id, src_data);
 			dev_err(rdev->dev, "  VM_CONTEXT1_PROTECTION_FAULT_ADDR   0x%08X\n",
 				addr);
 			dev_err(rdev->dev, "  VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n",
 				status);
 			cayman_vm_decode_fault(rdev, status, addr);
-			/* reset addr and status */
-			WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1);
 			break;
 		case 176: /* CP_INT in ring buffer */
 		case 177: /* CP_INT in IB1 */
@@ -5109,27 +5169,12 @@ static int evergreen_startup(struct radeon_device *rdev)
 
 	evergreen_mc_program(rdev);
 
-	if (ASIC_IS_DCE5(rdev)) {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw || !rdev->mc_fw) {
-			r = ni_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
+	if (ASIC_IS_DCE5(rdev) && !rdev->pm.dpm_enabled) {
 		r = ni_mc_load_microcode(rdev);
 		if (r) {
 			DRM_ERROR("Failed to load MC firmware!\n");
 			return r;
 		}
-	} else {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
-			r = r600_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
 	}
 
 	if (rdev->flags & RADEON_IS_AGP) {
@@ -5199,14 +5244,12 @@ static int evergreen_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     R600_CP_RB_RPTR, R600_CP_RB_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR, DMA_RB_WPTR,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0));
 	if (r)
 		return r;
@@ -5224,7 +5267,6 @@ static int evergreen_startup(struct radeon_device *rdev)
 	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 	if (ring->ring_size) {
 		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 				     RADEON_CP_PACKET2);
 		if (!r)
 			r = uvd_v1_0_init(rdev);
@@ -5267,6 +5309,9 @@ int evergreen_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	evergreen_init_golden_registers(rdev);
 
+	if (rdev->pm.pm_method == PM_METHOD_DPM)
+		radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = evergreen_startup(rdev);
 	if (r) {
@@ -5281,6 +5326,7 @@ int evergreen_resume(struct radeon_device *rdev)
 
 int evergreen_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_suspend(rdev);
@@ -5357,6 +5403,27 @@ int evergreen_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (ASIC_IS_DCE5(rdev)) {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw || !rdev->mc_fw) {
+			r = ni_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	} else {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
+			r = r600_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024);
 
@@ -5409,6 +5476,7 @@ int evergreen_init(struct radeon_device *rdev)
 
 void evergreen_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r600_audio_fini(rdev);
 	r700_cp_fini(rdev);
 	r600_dma_fini(rdev);
@@ -5418,9 +5486,9 @@ void evergreen_fini(struct radeon_device *rdev)
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
 	radeon_irq_kms_fini(rdev);
-	evergreen_pcie_gart_fini(rdev);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_fini(rdev);
+	evergreen_pcie_gart_fini(rdev);
 	r600_vram_scratch_fini(rdev);
 	radeon_gem_fini(rdev);
 	radeon_fence_driver_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/evergreen_cs.c b/drivers/gpu/drm/radeon/evergreen_cs.c
index eb8ac315f92..5c8b358f9fb 100644
--- a/drivers/gpu/drm/radeon/evergreen_cs.c
+++ b/drivers/gpu/drm/radeon/evergreen_cs.c
@@ -967,7 +967,10 @@ static int evergreen_cs_track_check(struct radeon_cs_parser *p)
 	if (track->cb_dirty) {
 		tmp = track->cb_target_mask;
 		for (i = 0; i < 8; i++) {
-			if ((tmp >> (i * 4)) & 0xF) {
+			u32 format = G_028C70_FORMAT(track->cb_color_info[i]);
+
+			if (format != V_028C70_COLOR_INVALID &&
+			    (tmp >> (i * 4)) & 0xF) {
 				/* at least one component is enabled */
 				if (track->cb_color_bo[i] == NULL) {
 					dev_warn(p->dev, "%s:%d mask 0x%08X | 0x%08X no cb for %d\n",
@@ -1162,7 +1165,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 					"0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		break;
 	case DB_DEPTH_CONTROL:
 		track->db_depth_control = radeon_get_ib_value(p, idx);
@@ -1193,12 +1196,12 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			}
 			ib[idx] &= ~Z_ARRAY_MODE(0xf);
 			track->db_z_info &= ~Z_ARRAY_MODE(0xf);
-			ib[idx] |= Z_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->lobj.tiling_flags));
-			track->db_z_info |= Z_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->lobj.tiling_flags));
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) {
+			ib[idx] |= Z_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->tiling_flags));
+			track->db_z_info |= Z_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->tiling_flags));
+			if (reloc->tiling_flags & RADEON_TILING_MACRO) {
 				unsigned bankw, bankh, mtaspect, tile_split;
 
-				evergreen_tiling_fields(reloc->lobj.tiling_flags,
+				evergreen_tiling_fields(reloc->tiling_flags,
 							&bankw, &bankh, &mtaspect,
 							&tile_split);
 				ib[idx] |= DB_NUM_BANKS(evergreen_cs_get_num_banks(track->nbanks));
@@ -1234,7 +1237,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		track->db_z_read_offset = radeon_get_ib_value(p, idx);
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->db_z_read_bo = reloc->robj;
 		track->db_dirty = true;
 		break;
@@ -1246,7 +1249,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		track->db_z_write_offset = radeon_get_ib_value(p, idx);
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->db_z_write_bo = reloc->robj;
 		track->db_dirty = true;
 		break;
@@ -1258,7 +1261,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		track->db_s_read_offset = radeon_get_ib_value(p, idx);
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->db_s_read_bo = reloc->robj;
 		track->db_dirty = true;
 		break;
@@ -1270,7 +1273,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		track->db_s_write_offset = radeon_get_ib_value(p, idx);
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->db_s_write_bo = reloc->robj;
 		track->db_dirty = true;
 		break;
@@ -1294,7 +1297,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 		}
 		tmp = (reg - VGT_STRMOUT_BUFFER_BASE_0) / 16;
 		track->vgt_strmout_bo_offset[tmp] = radeon_get_ib_value(p, idx) << 8;
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->vgt_strmout_bo[tmp] = reloc->robj;
 		track->streamout_dirty = true;
 		break;
@@ -1314,7 +1317,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 					"0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 	case CB_TARGET_MASK:
 		track->cb_target_mask = radeon_get_ib_value(p, idx);
 		track->cb_dirty = true;
@@ -1378,8 +1381,8 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 						"0x%04X\n", reg);
 				return -EINVAL;
 			}
-			ib[idx] |= CB_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->lobj.tiling_flags));
-			track->cb_color_info[tmp] |= CB_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->lobj.tiling_flags));
+			ib[idx] |= CB_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->tiling_flags));
+			track->cb_color_info[tmp] |= CB_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->tiling_flags));
 		}
 		track->cb_dirty = true;
 		break;
@@ -1396,8 +1399,8 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 						"0x%04X\n", reg);
 				return -EINVAL;
 			}
-			ib[idx] |= CB_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->lobj.tiling_flags));
-			track->cb_color_info[tmp] |= CB_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->lobj.tiling_flags));
+			ib[idx] |= CB_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->tiling_flags));
+			track->cb_color_info[tmp] |= CB_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->tiling_flags));
 		}
 		track->cb_dirty = true;
 		break;
@@ -1458,10 +1461,10 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		if (!(p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS)) {
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) {
+			if (reloc->tiling_flags & RADEON_TILING_MACRO) {
 				unsigned bankw, bankh, mtaspect, tile_split;
 
-				evergreen_tiling_fields(reloc->lobj.tiling_flags,
+				evergreen_tiling_fields(reloc->tiling_flags,
 							&bankw, &bankh, &mtaspect,
 							&tile_split);
 				ib[idx] |= CB_NUM_BANKS(evergreen_cs_get_num_banks(track->nbanks));
@@ -1486,10 +1489,10 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		if (!(p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS)) {
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) {
+			if (reloc->tiling_flags & RADEON_TILING_MACRO) {
 				unsigned bankw, bankh, mtaspect, tile_split;
 
-				evergreen_tiling_fields(reloc->lobj.tiling_flags,
+				evergreen_tiling_fields(reloc->tiling_flags,
 							&bankw, &bankh, &mtaspect,
 							&tile_split);
 				ib[idx] |= CB_NUM_BANKS(evergreen_cs_get_num_banks(track->nbanks));
@@ -1517,7 +1520,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			dev_err(p->dev, "bad SET_CONTEXT_REG 0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->cb_color_fmask_bo[tmp] = reloc->robj;
 		break;
 	case CB_COLOR0_CMASK:
@@ -1534,7 +1537,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			dev_err(p->dev, "bad SET_CONTEXT_REG 0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->cb_color_cmask_bo[tmp] = reloc->robj;
 		break;
 	case CB_COLOR0_FMASK_SLICE:
@@ -1575,7 +1578,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 		}
 		tmp = (reg - CB_COLOR0_BASE) / 0x3c;
 		track->cb_color_bo_offset[tmp] = radeon_get_ib_value(p, idx);
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->cb_color_bo[tmp] = reloc->robj;
 		track->cb_dirty = true;
 		break;
@@ -1591,7 +1594,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 		}
 		tmp = ((reg - CB_COLOR8_BASE) / 0x1c) + 8;
 		track->cb_color_bo_offset[tmp] = radeon_get_ib_value(p, idx);
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->cb_color_bo[tmp] = reloc->robj;
 		track->cb_dirty = true;
 		break;
@@ -1603,7 +1606,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		track->htile_offset = radeon_get_ib_value(p, idx);
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->htile_bo = reloc->robj;
 		track->db_dirty = true;
 		break;
@@ -1720,7 +1723,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 					"0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		break;
 	case SX_MEMORY_EXPORT_BASE:
 		if (p->rdev->family >= CHIP_CAYMAN) {
@@ -1734,7 +1737,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 					"0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		break;
 	case CAYMAN_SX_SCATTER_EXPORT_BASE:
 		if (p->rdev->family < CHIP_CAYMAN) {
@@ -1748,7 +1751,7 @@ static int evergreen_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 					"0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		break;
 	case SX_MISC:
 		track->sx_misc_kill_all_prims = (radeon_get_ib_value(p, idx) & 0x1) != 0;
@@ -1833,7 +1836,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 			return -EINVAL;
 		}
 
-		offset = reloc->lobj.gpu_offset +
+		offset = reloc->gpu_offset +
 		         (idx_value & 0xfffffff0) +
 		         ((u64)(tmp & 0xff) << 32);
 
@@ -1879,7 +1882,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 			return -EINVAL;
 		}
 
-		offset = reloc->lobj.gpu_offset +
+		offset = reloc->gpu_offset +
 		         idx_value +
 		         ((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
 
@@ -1906,7 +1909,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 			return -EINVAL;
 		}
 
-		offset = reloc->lobj.gpu_offset +
+		offset = reloc->gpu_offset +
 		         idx_value +
 		         ((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
 
@@ -1934,7 +1937,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 			return -EINVAL;
 		}
 
-		offset = reloc->lobj.gpu_offset +
+		offset = reloc->gpu_offset +
 		         radeon_get_ib_value(p, idx+1) +
 		         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
@@ -2024,7 +2027,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 			DRM_ERROR("bad DISPATCH_INDIRECT\n");
 			return -EINVAL;
 		}
-		ib[idx+0] = idx_value + (u32)(reloc->lobj.gpu_offset & 0xffffffff);
+		ib[idx+0] = idx_value + (u32)(reloc->gpu_offset & 0xffffffff);
 		r = evergreen_cs_track_check(p);
 		if (r) {
 			dev_warn(p->dev, "%s:%d invalid cmd stream\n", __func__, __LINE__);
@@ -2046,7 +2049,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 				return -EINVAL;
 			}
 
-			offset = reloc->lobj.gpu_offset +
+			offset = reloc->gpu_offset +
 			         (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
 			         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
@@ -2103,7 +2106,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 				tmp = radeon_get_ib_value(p, idx) +
 					((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
 
-				offset = reloc->lobj.gpu_offset + tmp;
+				offset = reloc->gpu_offset + tmp;
 
 				if ((tmp + size) > radeon_bo_size(reloc->robj)) {
 					dev_warn(p->dev, "CP DMA src buffer too small (%llu %lu)\n",
@@ -2141,7 +2144,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 				tmp = radeon_get_ib_value(p, idx+2) +
 					((u64)(radeon_get_ib_value(p, idx+3) & 0xff) << 32);
 
-				offset = reloc->lobj.gpu_offset + tmp;
+				offset = reloc->gpu_offset + tmp;
 
 				if ((tmp + size) > radeon_bo_size(reloc->robj)) {
 					dev_warn(p->dev, "CP DMA dst buffer too small (%llu %lu)\n",
@@ -2171,7 +2174,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 				DRM_ERROR("bad SURFACE_SYNC\n");
 				return -EINVAL;
 			}
-			ib[idx+2] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+			ib[idx+2] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		}
 		break;
 	case PACKET3_EVENT_WRITE:
@@ -2187,7 +2190,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 				DRM_ERROR("bad EVENT_WRITE\n");
 				return -EINVAL;
 			}
-			offset = reloc->lobj.gpu_offset +
+			offset = reloc->gpu_offset +
 			         (radeon_get_ib_value(p, idx+1) & 0xfffffff8) +
 			         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
@@ -2209,7 +2212,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 			return -EINVAL;
 		}
 
-		offset = reloc->lobj.gpu_offset +
+		offset = reloc->gpu_offset +
 		         (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
 		         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
@@ -2231,7 +2234,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 			return -EINVAL;
 		}
 
-		offset = reloc->lobj.gpu_offset +
+		offset = reloc->gpu_offset +
 		         (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
 		         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
@@ -2299,11 +2302,11 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 				}
 				if (!(p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS)) {
 					ib[idx+1+(i*8)+1] |=
-						TEX_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->lobj.tiling_flags));
-					if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) {
+						TEX_ARRAY_MODE(evergreen_cs_get_aray_mode(reloc->tiling_flags));
+					if (reloc->tiling_flags & RADEON_TILING_MACRO) {
 						unsigned bankw, bankh, mtaspect, tile_split;
 
-						evergreen_tiling_fields(reloc->lobj.tiling_flags,
+						evergreen_tiling_fields(reloc->tiling_flags,
 									&bankw, &bankh, &mtaspect,
 									&tile_split);
 						ib[idx+1+(i*8)+6] |= TEX_TILE_SPLIT(tile_split);
@@ -2315,7 +2318,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 					}
 				}
 				texture = reloc->robj;
-				toffset = (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+				toffset = (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 
 				/* tex mip base */
 				tex_dim = ib[idx+1+(i*8)+0] & 0x7;
@@ -2334,7 +2337,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 						DRM_ERROR("bad SET_RESOURCE (tex)\n");
 						return -EINVAL;
 					}
-					moffset = (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+					moffset = (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 					mipmap = reloc->robj;
 				}
 
@@ -2361,7 +2364,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 					ib[idx+1+(i*8)+1] = radeon_bo_size(reloc->robj) - offset;
 				}
 
-				offset64 = reloc->lobj.gpu_offset + offset;
+				offset64 = reloc->gpu_offset + offset;
 				ib[idx+1+(i*8)+0] = offset64;
 				ib[idx+1+(i*8)+2] = (ib[idx+1+(i*8)+2] & 0xffffff00) |
 						    (upper_32_bits(offset64) & 0xff);
@@ -2442,7 +2445,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 					  offset + 4, radeon_bo_size(reloc->robj));
 				return -EINVAL;
 			}
-			offset += reloc->lobj.gpu_offset;
+			offset += reloc->gpu_offset;
 			ib[idx+1] = offset;
 			ib[idx+2] = upper_32_bits(offset) & 0xff;
 		}
@@ -2461,7 +2464,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 					  offset + 4, radeon_bo_size(reloc->robj));
 				return -EINVAL;
 			}
-			offset += reloc->lobj.gpu_offset;
+			offset += reloc->gpu_offset;
 			ib[idx+3] = offset;
 			ib[idx+4] = upper_32_bits(offset) & 0xff;
 		}
@@ -2490,7 +2493,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 				  offset + 8, radeon_bo_size(reloc->robj));
 			return -EINVAL;
 		}
-		offset += reloc->lobj.gpu_offset;
+		offset += reloc->gpu_offset;
 		ib[idx+0] = offset;
 		ib[idx+1] = upper_32_bits(offset) & 0xff;
 		break;
@@ -2515,7 +2518,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 					  offset + 4, radeon_bo_size(reloc->robj));
 				return -EINVAL;
 			}
-			offset += reloc->lobj.gpu_offset;
+			offset += reloc->gpu_offset;
 			ib[idx+1] = offset;
 			ib[idx+2] = upper_32_bits(offset) & 0xff;
 		} else {
@@ -2539,7 +2542,7 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p,
 					  offset + 4, radeon_bo_size(reloc->robj));
 				return -EINVAL;
 			}
-			offset += reloc->lobj.gpu_offset;
+			offset += reloc->gpu_offset;
 			ib[idx+3] = offset;
 			ib[idx+4] = upper_32_bits(offset) & 0xff;
 		} else {
@@ -2714,7 +2717,7 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 				dst_offset = radeon_get_ib_value(p, idx+1);
 				dst_offset <<= 8;
 
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset >> 8);
 				p->idx += count + 7;
 				break;
 			/* linear */
@@ -2722,8 +2725,8 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 				dst_offset = radeon_get_ib_value(p, idx+1);
 				dst_offset |= ((u64)(radeon_get_ib_value(p, idx+2) & 0xff)) << 32;
 
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+2] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+2] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
 				p->idx += count + 3;
 				break;
 			default:
@@ -2765,10 +2768,10 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 							dst_offset + (count * 4), radeon_bo_size(dst_reloc->robj));
 					return -EINVAL;
 				}
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+2] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+3] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
-				ib[idx+4] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+2] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+3] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
+				ib[idx+4] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 				p->idx += 5;
 				break;
 			/* Copy L2T/T2L */
@@ -2778,22 +2781,22 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 					/* tiled src, linear dst */
 					src_offset = radeon_get_ib_value(p, idx+1);
 					src_offset <<= 8;
-					ib[idx+1] += (u32)(src_reloc->lobj.gpu_offset >> 8);
+					ib[idx+1] += (u32)(src_reloc->gpu_offset >> 8);
 
 					dst_offset = radeon_get_ib_value(p, idx + 7);
 					dst_offset |= ((u64)(radeon_get_ib_value(p, idx+8) & 0xff)) << 32;
-					ib[idx+7] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+8] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
+					ib[idx+7] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+8] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
 				} else {
 					/* linear src, tiled dst */
 					src_offset = radeon_get_ib_value(p, idx+7);
 					src_offset |= ((u64)(radeon_get_ib_value(p, idx+8) & 0xff)) << 32;
-					ib[idx+7] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+8] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+					ib[idx+7] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+8] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 
 					dst_offset = radeon_get_ib_value(p, idx+1);
 					dst_offset <<= 8;
-					ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
+					ib[idx+1] += (u32)(dst_reloc->gpu_offset >> 8);
 				}
 				if ((src_offset + (count * 4)) > radeon_bo_size(src_reloc->robj)) {
 					dev_warn(p->dev, "DMA L2T, src buffer too small (%llu %lu)\n",
@@ -2824,10 +2827,10 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 							dst_offset + count, radeon_bo_size(dst_reloc->robj));
 					return -EINVAL;
 				}
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset & 0xffffffff);
-				ib[idx+2] += (u32)(src_reloc->lobj.gpu_offset & 0xffffffff);
-				ib[idx+3] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
-				ib[idx+4] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset & 0xffffffff);
+				ib[idx+2] += (u32)(src_reloc->gpu_offset & 0xffffffff);
+				ib[idx+3] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
+				ib[idx+4] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 				p->idx += 5;
 				break;
 			/* Copy L2L, partial */
@@ -2837,10 +2840,10 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 					DRM_ERROR("L2L Partial is cayman only !\n");
 					return -EINVAL;
 				}
-				ib[idx+1] += (u32)(src_reloc->lobj.gpu_offset & 0xffffffff);
-				ib[idx+2] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
-				ib[idx+4] += (u32)(dst_reloc->lobj.gpu_offset & 0xffffffff);
-				ib[idx+5] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
+				ib[idx+1] += (u32)(src_reloc->gpu_offset & 0xffffffff);
+				ib[idx+2] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
+				ib[idx+4] += (u32)(dst_reloc->gpu_offset & 0xffffffff);
+				ib[idx+5] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
 
 				p->idx += 9;
 				break;
@@ -2873,12 +2876,12 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 							dst2_offset + (count * 4), radeon_bo_size(dst2_reloc->robj));
 					return -EINVAL;
 				}
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+2] += (u32)(dst2_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+3] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+4] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
-				ib[idx+5] += upper_32_bits(dst2_reloc->lobj.gpu_offset) & 0xff;
-				ib[idx+6] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+2] += (u32)(dst2_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+3] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+4] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
+				ib[idx+5] += upper_32_bits(dst2_reloc->gpu_offset) & 0xff;
+				ib[idx+6] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 				p->idx += 7;
 				break;
 			/* Copy L2T Frame to Field */
@@ -2913,10 +2916,10 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 							dst2_offset + (count * 4), radeon_bo_size(dst2_reloc->robj));
 					return -EINVAL;
 				}
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
-				ib[idx+2] += (u32)(dst2_reloc->lobj.gpu_offset >> 8);
-				ib[idx+8] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+9] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset >> 8);
+				ib[idx+2] += (u32)(dst2_reloc->gpu_offset >> 8);
+				ib[idx+8] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+9] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 				p->idx += 10;
 				break;
 			/* Copy L2T/T2L, partial */
@@ -2929,16 +2932,16 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 				/* detile bit */
 				if (radeon_get_ib_value(p, idx + 2) & (1 << 31)) {
 					/* tiled src, linear dst */
-					ib[idx+1] += (u32)(src_reloc->lobj.gpu_offset >> 8);
+					ib[idx+1] += (u32)(src_reloc->gpu_offset >> 8);
 
-					ib[idx+7] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+8] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
+					ib[idx+7] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+8] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
 				} else {
 					/* linear src, tiled dst */
-					ib[idx+7] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+8] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+					ib[idx+7] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+8] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 
-					ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
+					ib[idx+1] += (u32)(dst_reloc->gpu_offset >> 8);
 				}
 				p->idx += 12;
 				break;
@@ -2975,10 +2978,10 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 							dst2_offset + (count * 4), radeon_bo_size(dst2_reloc->robj));
 					return -EINVAL;
 				}
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
-				ib[idx+2] += (u32)(dst2_reloc->lobj.gpu_offset >> 8);
-				ib[idx+8] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+9] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset >> 8);
+				ib[idx+2] += (u32)(dst2_reloc->gpu_offset >> 8);
+				ib[idx+8] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+9] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 				p->idx += 10;
 				break;
 			/* Copy L2T/T2L (tile units) */
@@ -2989,22 +2992,22 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 					/* tiled src, linear dst */
 					src_offset = radeon_get_ib_value(p, idx+1);
 					src_offset <<= 8;
-					ib[idx+1] += (u32)(src_reloc->lobj.gpu_offset >> 8);
+					ib[idx+1] += (u32)(src_reloc->gpu_offset >> 8);
 
 					dst_offset = radeon_get_ib_value(p, idx+7);
 					dst_offset |= ((u64)(radeon_get_ib_value(p, idx+8) & 0xff)) << 32;
-					ib[idx+7] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+8] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
+					ib[idx+7] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+8] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
 				} else {
 					/* linear src, tiled dst */
 					src_offset = radeon_get_ib_value(p, idx+7);
 					src_offset |= ((u64)(radeon_get_ib_value(p, idx+8) & 0xff)) << 32;
-					ib[idx+7] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+8] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+					ib[idx+7] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+8] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 
 					dst_offset = radeon_get_ib_value(p, idx+1);
 					dst_offset <<= 8;
-					ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
+					ib[idx+1] += (u32)(dst_reloc->gpu_offset >> 8);
 				}
 				if ((src_offset + (count * 4)) > radeon_bo_size(src_reloc->robj)) {
 					dev_warn(p->dev, "DMA L2T, T2L src buffer too small (%llu %lu)\n",
@@ -3025,8 +3028,8 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 					DRM_ERROR("L2T, T2L Partial is cayman only !\n");
 					return -EINVAL;
 				}
-				ib[idx+1] += (u32)(src_reloc->lobj.gpu_offset >> 8);
-				ib[idx+4] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
+				ib[idx+1] += (u32)(src_reloc->gpu_offset >> 8);
+				ib[idx+4] += (u32)(dst_reloc->gpu_offset >> 8);
 				p->idx += 13;
 				break;
 			/* Copy L2T broadcast (tile units) */
@@ -3062,10 +3065,10 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 							dst2_offset + (count * 4), radeon_bo_size(dst2_reloc->robj));
 					return -EINVAL;
 				}
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
-				ib[idx+2] += (u32)(dst2_reloc->lobj.gpu_offset >> 8);
-				ib[idx+8] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+9] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset >> 8);
+				ib[idx+2] += (u32)(dst2_reloc->gpu_offset >> 8);
+				ib[idx+8] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+9] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 				p->idx += 10;
 				break;
 			default:
@@ -3086,8 +3089,8 @@ int evergreen_dma_cs_parse(struct radeon_cs_parser *p)
 					 dst_offset, radeon_bo_size(dst_reloc->robj));
 				return -EINVAL;
 			}
-			ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-			ib[idx+3] += (upper_32_bits(dst_reloc->lobj.gpu_offset) << 16) & 0x00ff0000;
+			ib[idx+1] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+			ib[idx+3] += (upper_32_bits(dst_reloc->gpu_offset) << 16) & 0x00ff0000;
 			p->idx += 4;
 			break;
 		case DMA_PACKET_NOP:
diff --git a/drivers/gpu/drm/radeon/evergreen_dma.c b/drivers/gpu/drm/radeon/evergreen_dma.c
index a37b5443638..478caefe0fe 100644
--- a/drivers/gpu/drm/radeon/evergreen_dma.c
+++ b/drivers/gpu/drm/radeon/evergreen_dma.c
@@ -151,6 +151,7 @@ int evergreen_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
@@ -174,11 +175,9 @@ bool evergreen_dma_is_lockup(struct radeon_device *rdev, struct radeon_ring *rin
 	u32 reset_mask = evergreen_gpu_check_soft_reset(rdev);
 
 	if (!(reset_mask & RADEON_RESET_DMA)) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force ring activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
diff --git a/drivers/gpu/drm/radeon/evergreen_hdmi.c b/drivers/gpu/drm/radeon/evergreen_hdmi.c
index 0c6d5cef4cf..1ec0e6e83f9 100644
--- a/drivers/gpu/drm/radeon/evergreen_hdmi.c
+++ b/drivers/gpu/drm/radeon/evergreen_hdmi.c
@@ -293,10 +293,13 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
 	struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv;
+	struct drm_connector *connector = radeon_get_connector_for_encoder(encoder);
 	u8 buffer[HDMI_INFOFRAME_HEADER_SIZE + HDMI_AVI_INFOFRAME_SIZE];
 	struct hdmi_avi_infoframe frame;
 	uint32_t offset;
 	ssize_t err;
+	uint32_t val;
+	int bpc = 8;
 
 	if (!dig || !dig->afmt)
 		return;
@@ -306,6 +309,21 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode
 		return;
 	offset = dig->afmt->offset;
 
+	/* hdmi deep color mode general control packets setup, if bpc > 8 */
+	if (encoder->crtc) {
+		struct radeon_crtc *radeon_crtc = to_radeon_crtc(encoder->crtc);
+		bpc = radeon_crtc->bpc;
+	}
+
+	/* disable audio prior to setting up hw */
+	if (ASIC_IS_DCE6(rdev)) {
+		dig->afmt->pin = dce6_audio_get_pin(rdev);
+		dce6_audio_enable(rdev, dig->afmt->pin, false);
+	} else {
+		dig->afmt->pin = r600_audio_get_pin(rdev);
+		r600_audio_enable(rdev, dig->afmt->pin, false);
+	}
+
 	evergreen_audio_set_dto(encoder, mode->clock);
 
 	WREG32(HDMI_VBI_PACKET_CONTROL + offset,
@@ -313,6 +331,35 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode
 
 	WREG32(AFMT_AUDIO_CRC_CONTROL + offset, 0x1000);
 
+	val = RREG32(HDMI_CONTROL + offset);
+	val &= ~HDMI_DEEP_COLOR_ENABLE;
+	val &= ~HDMI_DEEP_COLOR_DEPTH_MASK;
+
+	switch (bpc) {
+		case 0:
+		case 6:
+		case 8:
+		case 16:
+		default:
+			DRM_DEBUG("%s: Disabling hdmi deep color for %d bpc.\n",
+					 connector->name, bpc);
+			break;
+		case 10:
+			val |= HDMI_DEEP_COLOR_ENABLE;
+			val |= HDMI_DEEP_COLOR_DEPTH(HDMI_30BIT_DEEP_COLOR);
+			DRM_DEBUG("%s: Enabling hdmi deep color 30 for 10 bpc.\n",
+					 connector->name);
+			break;
+		case 12:
+			val |= HDMI_DEEP_COLOR_ENABLE;
+			val |= HDMI_DEEP_COLOR_DEPTH(HDMI_36BIT_DEEP_COLOR);
+			DRM_DEBUG("%s: Enabling hdmi deep color 36 for 12 bpc.\n",
+					 connector->name);
+			break;
+	}
+
+	WREG32(HDMI_CONTROL + offset, val);
+
 	WREG32(HDMI_VBI_PACKET_CONTROL + offset,
 	       HDMI_NULL_SEND | /* send null packets when required */
 	       HDMI_GC_SEND | /* send general control packets */
@@ -339,9 +386,13 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode
 
 	/* fglrx clears sth in AFMT_AUDIO_PACKET_CONTROL2 here */
 
-	WREG32(HDMI_ACR_PACKET_CONTROL + offset,
-	       HDMI_ACR_SOURCE | /* select SW CTS value */
-	       HDMI_ACR_AUTO_SEND); /* allow hw to sent ACR packets when required */
+	if (bpc > 8)
+		WREG32(HDMI_ACR_PACKET_CONTROL + offset,
+		       HDMI_ACR_AUTO_SEND); /* allow hw to sent ACR packets when required */
+	else
+		WREG32(HDMI_ACR_PACKET_CONTROL + offset,
+		       HDMI_ACR_SOURCE | /* select SW CTS value */
+		       HDMI_ACR_AUTO_SEND); /* allow hw to sent ACR packets when required */
 
 	evergreen_hdmi_update_ACR(encoder, mode->clock);
 
@@ -409,12 +460,16 @@ void evergreen_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode
 	WREG32(AFMT_RAMP_CONTROL1 + offset, 0x007FFFFF);
 	WREG32(AFMT_RAMP_CONTROL2 + offset, 0x00000001);
 	WREG32(AFMT_RAMP_CONTROL3 + offset, 0x00000001);
+
+	/* enable audio after to setting up hw */
+	if (ASIC_IS_DCE6(rdev))
+		dce6_audio_enable(rdev, dig->afmt->pin, true);
+	else
+		r600_audio_enable(rdev, dig->afmt->pin, true);
 }
 
 void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable)
 {
-	struct drm_device *dev = encoder->dev;
-	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
 	struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv;
 
@@ -427,15 +482,6 @@ void evergreen_hdmi_enable(struct drm_encoder *encoder, bool enable)
 	if (!enable && !dig->afmt->enabled)
 		return;
 
-	if (enable) {
-		if (ASIC_IS_DCE6(rdev))
-			dig->afmt->pin = dce6_audio_get_pin(rdev);
-		else
-			dig->afmt->pin = r600_audio_get_pin(rdev);
-	} else {
-		dig->afmt->pin = NULL;
-	}
-
 	dig->afmt->enabled = enable;
 
 	DRM_DEBUG("%sabling HDMI interface @ 0x%04X for encoder 0x%x\n",
diff --git a/drivers/gpu/drm/radeon/evergreen_reg.h b/drivers/gpu/drm/radeon/evergreen_reg.h
index 8a4e641f0e3..23bff590fb6 100644
--- a/drivers/gpu/drm/radeon/evergreen_reg.h
+++ b/drivers/gpu/drm/radeon/evergreen_reg.h
@@ -33,6 +33,7 @@
 #define EVERGREEN_PIF_PHY0_DATA                         0xc
 #define EVERGREEN_PIF_PHY1_INDEX                        0x10
 #define EVERGREEN_PIF_PHY1_DATA                         0x14
+#define EVERGREEN_MM_INDEX_HI                           0x18
 
 #define EVERGREEN_VGA_MEMORY_BASE_ADDRESS               0x310
 #define EVERGREEN_VGA_MEMORY_BASE_ADDRESS_HIGH          0x324
@@ -115,6 +116,8 @@
 #       define EVERGREEN_GRPH_ARRAY_LINEAR_ALIGNED      1
 #       define EVERGREEN_GRPH_ARRAY_1D_TILED_THIN1      2
 #       define EVERGREEN_GRPH_ARRAY_2D_TILED_THIN1      4
+#define EVERGREEN_GRPH_LUT_10BIT_BYPASS_CONTROL         0x6808
+#       define EVERGREEN_LUT_10BIT_BYPASS_EN            (1 << 8)
 #define EVERGREEN_GRPH_SWAP_CONTROL                     0x680c
 #       define EVERGREEN_GRPH_ENDIAN_SWAP(x)            (((x) & 0x3) << 0)
 #       define EVERGREEN_GRPH_ENDIAN_NONE               0
@@ -236,7 +239,6 @@
 #       define EVERGREEN_CRTC_V_BLANK                   (1 << 0)
 #define EVERGREEN_CRTC_STATUS_POSITION                  0x6e90
 #define EVERGREEN_CRTC_STATUS_HV_COUNT                  0x6ea0
-#define EVERGREEN_MASTER_UPDATE_MODE                    0x6ef8
 #define EVERGREEN_CRTC_UPDATE_LOCK                      0x6ed4
 #define EVERGREEN_MASTER_UPDATE_LOCK                    0x6ef4
 #define EVERGREEN_MASTER_UPDATE_MODE                    0x6ef8
diff --git a/drivers/gpu/drm/radeon/evergreen_smc.h b/drivers/gpu/drm/radeon/evergreen_smc.h
index 76ada8cfe90..3a03ba37d04 100644
--- a/drivers/gpu/drm/radeon/evergreen_smc.h
+++ b/drivers/gpu/drm/radeon/evergreen_smc.h
@@ -57,7 +57,7 @@ typedef struct SMC_Evergreen_MCRegisters SMC_Evergreen_MCRegisters;
 
 #define EVERGREEN_SMC_FIRMWARE_HEADER_LOCATION 0x100
 
-#define EVERGREEN_SMC_FIRMWARE_HEADER_softRegisters   0x0
+#define EVERGREEN_SMC_FIRMWARE_HEADER_softRegisters   0x8
 #define EVERGREEN_SMC_FIRMWARE_HEADER_stateTable      0xC
 #define EVERGREEN_SMC_FIRMWARE_HEADER_mcRegisterTable 0x20
 
diff --git a/drivers/gpu/drm/radeon/evergreend.h b/drivers/gpu/drm/radeon/evergreend.h
index 17f99079899..b066d6711b8 100644
--- a/drivers/gpu/drm/radeon/evergreend.h
+++ b/drivers/gpu/drm/radeon/evergreend.h
@@ -82,12 +82,16 @@
 #define	CG_SPLL_FUNC_CNTL_2				0x604
 #define		SCLK_MUX_SEL(x)				((x) << 0)
 #define		SCLK_MUX_SEL_MASK			(0x1ff << 0)
+#define		SCLK_MUX_UPDATE				(1 << 26)
 #define	CG_SPLL_FUNC_CNTL_3				0x608
 #define		SPLL_FB_DIV(x)				((x) << 0)
 #define		SPLL_FB_DIV_MASK			(0x3ffffff << 0)
 #define		SPLL_DITHEN				(1 << 28)
+#define	CG_SPLL_STATUS					0x60c
+#define		SPLL_CHG_STATUS				(1 << 1)
 
 #define MPLL_CNTL_MODE                                  0x61c
+#       define MPLL_MCLK_SEL                            (1 << 11)
 #       define SS_SSEN                                  (1 << 24)
 #       define SS_DSMODE_EN                             (1 << 25)
 
@@ -513,10 +517,11 @@
 #       define HDMI_ERROR_ACK                (1 << 8)
 #       define HDMI_ERROR_MASK               (1 << 9)
 #       define HDMI_DEEP_COLOR_ENABLE        (1 << 24)
-#       define HDMI_DEEP_COLOR_DEPTH         (((x) & 3) << 28)
+#       define HDMI_DEEP_COLOR_DEPTH(x)      (((x) & 3) << 28)
 #       define HDMI_24BIT_DEEP_COLOR         0
 #       define HDMI_30BIT_DEEP_COLOR         1
 #       define HDMI_36BIT_DEEP_COLOR         2
+#       define HDMI_DEEP_COLOR_DEPTH_MASK    (3 << 28)
 #define HDMI_STATUS                          0x7034
 #       define HDMI_ACTIVE_AVMUTE            (1 << 0)
 #       define HDMI_AUDIO_PACKET_ERROR       (1 << 16)
diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c
index b41905573cd..9ef8c38f2d6 100644
--- a/drivers/gpu/drm/radeon/kv_dpm.c
+++ b/drivers/gpu/drm/radeon/kv_dpm.c
@@ -546,6 +546,52 @@ static int kv_set_divider_value(struct radeon_device *rdev,
 	return 0;
 }
 
+static u32 kv_convert_vid2_to_vid7(struct radeon_device *rdev,
+				   struct sumo_vid_mapping_table *vid_mapping_table,
+				   u32 vid_2bit)
+{
+	struct radeon_clock_voltage_dependency_table *vddc_sclk_table =
+		&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk;
+	u32 i;
+
+	if (vddc_sclk_table && vddc_sclk_table->count) {
+		if (vid_2bit < vddc_sclk_table->count)
+			return vddc_sclk_table->entries[vid_2bit].v;
+		else
+			return vddc_sclk_table->entries[vddc_sclk_table->count - 1].v;
+	} else {
+		for (i = 0; i < vid_mapping_table->num_entries; i++) {
+			if (vid_mapping_table->entries[i].vid_2bit == vid_2bit)
+				return vid_mapping_table->entries[i].vid_7bit;
+		}
+		return vid_mapping_table->entries[vid_mapping_table->num_entries - 1].vid_7bit;
+	}
+}
+
+static u32 kv_convert_vid7_to_vid2(struct radeon_device *rdev,
+				   struct sumo_vid_mapping_table *vid_mapping_table,
+				   u32 vid_7bit)
+{
+	struct radeon_clock_voltage_dependency_table *vddc_sclk_table =
+		&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk;
+	u32 i;
+
+	if (vddc_sclk_table && vddc_sclk_table->count) {
+		for (i = 0; i < vddc_sclk_table->count; i++) {
+			if (vddc_sclk_table->entries[i].v == vid_7bit)
+				return i;
+		}
+		return vddc_sclk_table->count - 1;
+	} else {
+		for (i = 0; i < vid_mapping_table->num_entries; i++) {
+			if (vid_mapping_table->entries[i].vid_7bit == vid_7bit)
+				return vid_mapping_table->entries[i].vid_2bit;
+		}
+
+		return vid_mapping_table->entries[vid_mapping_table->num_entries - 1].vid_2bit;
+	}
+}
+
 static u16 kv_convert_8bit_index_to_voltage(struct radeon_device *rdev,
 					    u16 voltage)
 {
@@ -556,9 +602,9 @@ static u16 kv_convert_2bit_index_to_voltage(struct radeon_device *rdev,
 					    u32 vid_2bit)
 {
 	struct kv_power_info *pi = kv_get_pi(rdev);
-	u32 vid_8bit = sumo_convert_vid2_to_vid7(rdev,
-						 &pi->sys_info.vid_mapping_table,
-						 vid_2bit);
+	u32 vid_8bit = kv_convert_vid2_to_vid7(rdev,
+					       &pi->sys_info.vid_mapping_table,
+					       vid_2bit);
 
 	return kv_convert_8bit_index_to_voltage(rdev, (u16)vid_8bit);
 }
@@ -639,7 +685,7 @@ static int kv_force_lowest_valid(struct radeon_device *rdev)
 
 static int kv_unforce_levels(struct radeon_device *rdev)
 {
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return kv_notify_message_to_smu(rdev, PPSMC_MSG_NoForcedLevel);
 	else
 		return kv_set_enabled_levels(rdev);
@@ -1126,11 +1172,6 @@ int kv_dpm_enable(struct radeon_device *rdev)
 	struct kv_power_info *pi = kv_get_pi(rdev);
 	int ret;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	ret = kv_process_firmware_header(rdev);
 	if (ret) {
 		DRM_ERROR("kv_process_firmware_header failed\n");
@@ -1215,6 +1256,21 @@ int kv_dpm_enable(struct radeon_device *rdev)
 
 	kv_reset_acp_boot_level(rdev);
 
+	ret = kv_smc_bapm_enable(rdev, false);
+	if (ret) {
+		DRM_ERROR("kv_smc_bapm_enable failed\n");
+		return ret;
+	}
+
+	kv_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
+
+	return ret;
+}
+
+int kv_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret = 0;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		ret = kv_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
@@ -1226,35 +1282,17 @@ int kv_dpm_enable(struct radeon_device *rdev)
 		radeon_irq_set(rdev);
 	}
 
-	ret = kv_smc_bapm_enable(rdev, false);
-	if (ret) {
-		DRM_ERROR("kv_smc_bapm_enable failed\n");
-		return ret;
-	}
-
 	/* powerdown unused blocks for now */
 	kv_dpm_powergate_acp(rdev, true);
 	kv_dpm_powergate_samu(rdev, true);
 	kv_dpm_powergate_vce(rdev, true);
 	kv_dpm_powergate_uvd(rdev, true);
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), true);
-
-	kv_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
-
 	return ret;
 }
 
 void kv_dpm_disable(struct radeon_device *rdev)
 {
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	kv_smc_bapm_enable(rdev, false);
 
 	/* powerup blocks */
@@ -1346,13 +1384,11 @@ static int kv_enable_uvd_dpm(struct radeon_device *rdev, bool enable)
 					PPSMC_MSG_UVDDPM_Enable : PPSMC_MSG_UVDDPM_Disable);
 }
 
-#if 0
 static int kv_enable_vce_dpm(struct radeon_device *rdev, bool enable)
 {
 	return kv_notify_message_to_smu(rdev, enable ?
 					PPSMC_MSG_VCEDPM_Enable : PPSMC_MSG_VCEDPM_Disable);
 }
-#endif
 
 static int kv_enable_samu_dpm(struct radeon_device *rdev, bool enable)
 {
@@ -1372,13 +1408,20 @@ static int kv_update_uvd_dpm(struct radeon_device *rdev, bool gate)
 	struct radeon_uvd_clock_voltage_dependency_table *table =
 		&rdev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table;
 	int ret;
+	u32 mask;
 
 	if (!gate) {
-		if (!pi->caps_uvd_dpm || table->count || pi->caps_stable_p_state)
+		if (table->count)
 			pi->uvd_boot_level = table->count - 1;
 		else
 			pi->uvd_boot_level = 0;
 
+		if (!pi->caps_uvd_dpm || pi->caps_stable_p_state) {
+			mask = 1 << pi->uvd_boot_level;
+		} else {
+			mask = 0x1f;
+		}
+
 		ret = kv_copy_bytes_to_smc(rdev,
 					   pi->dpm_table_start +
 					   offsetof(SMU7_Fusion_DpmTable, UvdBootLevel),
@@ -1387,17 +1430,14 @@ static int kv_update_uvd_dpm(struct radeon_device *rdev, bool gate)
 		if (ret)
 			return ret;
 
-		if (!pi->caps_uvd_dpm ||
-		    pi->caps_stable_p_state)
-			kv_send_msg_to_smc_with_parameter(rdev,
-							  PPSMC_MSG_UVDDPM_SetEnabledMask,
-							  (1 << pi->uvd_boot_level));
+		kv_send_msg_to_smc_with_parameter(rdev,
+						  PPSMC_MSG_UVDDPM_SetEnabledMask,
+						  mask);
 	}
 
 	return kv_enable_uvd_dpm(rdev, !gate);
 }
 
-#if 0
 static u8 kv_get_vce_boot_level(struct radeon_device *rdev)
 {
 	u8 i;
@@ -1422,6 +1462,9 @@ static int kv_update_vce_dpm(struct radeon_device *rdev,
 	int ret;
 
 	if (radeon_new_state->evclk > 0 && radeon_current_state->evclk == 0) {
+		kv_dpm_powergate_vce(rdev, false);
+		/* turn the clocks on when encoding */
+		cik_update_cg(rdev, RADEON_CG_BLOCK_VCE, false);
 		if (pi->caps_stable_p_state)
 			pi->vce_boot_level = table->count - 1;
 		else
@@ -1444,11 +1487,13 @@ static int kv_update_vce_dpm(struct radeon_device *rdev,
 		kv_enable_vce_dpm(rdev, true);
 	} else if (radeon_new_state->evclk == 0 && radeon_current_state->evclk > 0) {
 		kv_enable_vce_dpm(rdev, false);
+		/* turn the clocks off when not encoding */
+		cik_update_cg(rdev, RADEON_CG_BLOCK_VCE, true);
+		kv_dpm_powergate_vce(rdev, true);
 	}
 
 	return 0;
 }
-#endif
 
 static int kv_update_samu_dpm(struct radeon_device *rdev, bool gate)
 {
@@ -1583,11 +1628,16 @@ static void kv_dpm_powergate_vce(struct radeon_device *rdev, bool gate)
 	pi->vce_power_gated = gate;
 
 	if (gate) {
-		if (pi->caps_vce_pg)
+		if (pi->caps_vce_pg) {
+			/* XXX do we need a vce_v1_0_stop() ?  */
 			kv_notify_message_to_smu(rdev, PPSMC_MSG_VCEPowerOFF);
+		}
 	} else {
-		if (pi->caps_vce_pg)
+		if (pi->caps_vce_pg) {
 			kv_notify_message_to_smu(rdev, PPSMC_MSG_VCEPowerON);
+			vce_v2_0_resume(rdev);
+			vce_v1_0_start(rdev);
+		}
 	}
 }
 
@@ -1618,7 +1668,7 @@ static void kv_dpm_powergate_acp(struct radeon_device *rdev, bool gate)
 	if (pi->acp_power_gated == gate)
 		return;
 
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return;
 
 	pi->acp_power_gated = gate;
@@ -1776,14 +1826,9 @@ int kv_dpm_set_power_state(struct radeon_device *rdev)
 {
 	struct kv_power_info *pi = kv_get_pi(rdev);
 	struct radeon_ps *new_ps = &pi->requested_rps;
-	/*struct radeon_ps *old_ps = &pi->current_rps;*/
+	struct radeon_ps *old_ps = &pi->current_rps;
 	int ret;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	if (pi->bapm_enable) {
 		ret = kv_smc_bapm_enable(rdev, rdev->pm.dpm.ac_power);
 		if (ret) {
@@ -1792,7 +1837,7 @@ int kv_dpm_set_power_state(struct radeon_device *rdev)
 		}
 	}
 
-	if (rdev->family == CHIP_KABINI) {
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) {
 		if (pi->enable_dpm) {
 			kv_set_valid_clock_range(rdev, new_ps);
 			kv_update_dfs_bypass_settings(rdev, new_ps);
@@ -1811,14 +1856,15 @@ int kv_dpm_set_power_state(struct radeon_device *rdev)
 			kv_set_enabled_levels(rdev);
 			kv_force_lowest_valid(rdev);
 			kv_unforce_levels(rdev);
-#if 0
+
 			ret = kv_update_vce_dpm(rdev, new_ps, old_ps);
 			if (ret) {
 				DRM_ERROR("kv_update_vce_dpm failed\n");
 				return ret;
 			}
-#endif
 			kv_update_sclk_t(rdev);
+			if (rdev->family == CHIP_MULLINS)
+				kv_enable_nb_dpm(rdev);
 		}
 	} else {
 		if (pi->enable_dpm) {
@@ -1836,24 +1882,17 @@ int kv_dpm_set_power_state(struct radeon_device *rdev)
 			kv_program_nbps_index_settings(rdev, new_ps);
 			kv_freeze_sclk_dpm(rdev, false);
 			kv_set_enabled_levels(rdev);
-#if 0
 			ret = kv_update_vce_dpm(rdev, new_ps, old_ps);
 			if (ret) {
 				DRM_ERROR("kv_update_vce_dpm failed\n");
 				return ret;
 			}
-#endif
 			kv_update_acp_boot_level(rdev);
 			kv_update_sclk_t(rdev);
 			kv_enable_nb_dpm(rdev);
 		}
 	}
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), true);
-
 	return 0;
 }
 
@@ -1876,7 +1915,7 @@ void kv_dpm_reset_asic(struct radeon_device *rdev)
 {
 	struct kv_power_info *pi = kv_get_pi(rdev);
 
-	if (rdev->family == CHIP_KABINI) {
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) {
 		kv_force_lowest_valid(rdev);
 		kv_init_graphics_levels(rdev);
 		kv_program_bootup_state(rdev);
@@ -1915,14 +1954,41 @@ static void kv_construct_max_power_limits_table(struct radeon_device *rdev,
 static void kv_patch_voltage_values(struct radeon_device *rdev)
 {
 	int i;
-	struct radeon_uvd_clock_voltage_dependency_table *table =
+	struct radeon_uvd_clock_voltage_dependency_table *uvd_table =
 		&rdev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table;
+	struct radeon_vce_clock_voltage_dependency_table *vce_table =
+		&rdev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table;
+	struct radeon_clock_voltage_dependency_table *samu_table =
+		&rdev->pm.dpm.dyn_state.samu_clock_voltage_dependency_table;
+	struct radeon_clock_voltage_dependency_table *acp_table =
+		&rdev->pm.dpm.dyn_state.acp_clock_voltage_dependency_table;
+
+	if (uvd_table->count) {
+		for (i = 0; i < uvd_table->count; i++)
+			uvd_table->entries[i].v =
+				kv_convert_8bit_index_to_voltage(rdev,
+								 uvd_table->entries[i].v);
+	}
 
-	if (table->count) {
-		for (i = 0; i < table->count; i++)
-			table->entries[i].v =
+	if (vce_table->count) {
+		for (i = 0; i < vce_table->count; i++)
+			vce_table->entries[i].v =
 				kv_convert_8bit_index_to_voltage(rdev,
-								 table->entries[i].v);
+								 vce_table->entries[i].v);
+	}
+
+	if (samu_table->count) {
+		for (i = 0; i < samu_table->count; i++)
+			samu_table->entries[i].v =
+				kv_convert_8bit_index_to_voltage(rdev,
+								 samu_table->entries[i].v);
+	}
+
+	if (acp_table->count) {
+		for (i = 0; i < acp_table->count; i++)
+			acp_table->entries[i].v =
+				kv_convert_8bit_index_to_voltage(rdev,
+								 acp_table->entries[i].v);
 	}
 
 }
@@ -1955,7 +2021,7 @@ static int kv_force_dpm_highest(struct radeon_device *rdev)
 			break;
 	}
 
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return kv_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_DPM_ForceState, i);
 	else
 		return kv_set_enabled_level(rdev, i);
@@ -1975,7 +2041,7 @@ static int kv_force_dpm_lowest(struct radeon_device *rdev)
 			break;
 	}
 
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return kv_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_DPM_ForceState, i);
 	else
 		return kv_set_enabled_level(rdev, i);
@@ -2055,6 +2121,14 @@ static void kv_apply_state_adjust_rules(struct radeon_device *rdev,
 	struct radeon_clock_and_voltage_limits *max_limits =
 		&rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
 
+	if (new_rps->vce_active) {
+		new_rps->evclk = rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].evclk;
+		new_rps->ecclk = rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].ecclk;
+	} else {
+		new_rps->evclk = 0;
+		new_rps->ecclk = 0;
+	}
+
 	mclk = max_limits->mclk;
 	sclk = min_sclk;
 
@@ -2074,6 +2148,11 @@ static void kv_apply_state_adjust_rules(struct radeon_device *rdev,
 		sclk = stable_p_state_sclk;
 	}
 
+	if (new_rps->vce_active) {
+		if (sclk < rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].sclk)
+			sclk = rdev->pm.dpm.vce_states[rdev->pm.dpm.vce_level].sclk;
+	}
+
 	ps->need_dfs_bypass = true;
 
 	for (i = 0; i < ps->num_levels; i++) {
@@ -2110,7 +2189,8 @@ static void kv_apply_state_adjust_rules(struct radeon_device *rdev,
 		}
 	}
 
-	pi->video_start = new_rps->dclk || new_rps->vclk;
+	pi->video_start = new_rps->dclk || new_rps->vclk ||
+		new_rps->evclk || new_rps->ecclk;
 
 	if ((new_rps->class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) ==
 	    ATOM_PPLIB_CLASSIFICATION_UI_BATTERY)
@@ -2118,7 +2198,7 @@ static void kv_apply_state_adjust_rules(struct radeon_device *rdev,
 	else
 		pi->battery_state = false;
 
-	if (rdev->family == CHIP_KABINI) {
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) {
 		ps->dpm0_pg_nb_ps_lo = 0x1;
 		ps->dpm0_pg_nb_ps_hi = 0x0;
 		ps->dpmx_nb_ps_lo = 0x1;
@@ -2179,7 +2259,7 @@ static int kv_calculate_nbps_level_settings(struct radeon_device *rdev)
 	if (pi->lowest_valid > pi->highest_valid)
 		return -EINVAL;
 
-	if (rdev->family == CHIP_KABINI) {
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) {
 		for (i = pi->lowest_valid; i <= pi->highest_valid; i++) {
 			pi->graphics_level[i].GnbSlow = 1;
 			pi->graphics_level[i].ForceNbPs1 = 0;
@@ -2253,9 +2333,9 @@ static void kv_init_graphics_levels(struct radeon_device *rdev)
 				break;
 
 			kv_set_divider_value(rdev, i, table->entries[i].clk);
-			vid_2bit = sumo_convert_vid7_to_vid2(rdev,
-							     &pi->sys_info.vid_mapping_table,
-							     table->entries[i].v);
+			vid_2bit = kv_convert_vid7_to_vid2(rdev,
+							   &pi->sys_info.vid_mapping_table,
+							   table->entries[i].v);
 			kv_set_vid(rdev, i, vid_2bit);
 			kv_set_at(rdev, i, pi->at[i]);
 			kv_dpm_power_level_enabled_for_throttle(rdev, i, true);
@@ -2324,7 +2404,7 @@ static void kv_program_nbps_index_settings(struct radeon_device *rdev,
 	struct kv_power_info *pi = kv_get_pi(rdev);
 	u32 nbdpmconfig1;
 
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return;
 
 	if (pi->sys_info.nb_dpm_enable) {
@@ -2556,9 +2636,6 @@ static int kv_parse_power_table(struct radeon_device *rdev)
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
-	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
-	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
-	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
 	for (i = 0; i < state_array->ucNumEntries; i++) {
 		u8 *idx;
 		power_state = (union pplib_power_state *)power_state_offset;
@@ -2595,6 +2672,19 @@ static int kv_parse_power_table(struct radeon_device *rdev)
 		power_state_offset += 2 + power_state->v2.ucNumDPMLevels;
 	}
 	rdev->pm.dpm.num_ps = state_array->ucNumEntries;
+
+	/* fill in the vce power states */
+	for (i = 0; i < RADEON_MAX_VCE_LEVELS; i++) {
+		u32 sclk;
+		clock_array_index = rdev->pm.dpm.vce_states[i].clk_idx;
+		clock_info = (union pplib_clock_info *)
+			&clock_info_array->clockInfo[clock_array_index * clock_info_array->ucEntrySize];
+		sclk = le16_to_cpu(clock_info->sumo.usEngineClockLow);
+		sclk |= clock_info->sumo.ucEngineClockHigh << 16;
+		rdev->pm.dpm.vce_states[i].sclk = sclk;
+		rdev->pm.dpm.vce_states[i].mclk = 0;
+	}
+
 	return 0;
 }
 
@@ -2608,6 +2698,10 @@ int kv_dpm_init(struct radeon_device *rdev)
 		return -ENOMEM;
 	rdev->pm.dpm.priv = pi;
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = r600_parse_extended_power_table(rdev);
 	if (ret)
 		return ret;
@@ -2617,9 +2711,6 @@ int kv_dpm_init(struct radeon_device *rdev)
 
         pi->sram_end = SMC_RAM_END;
 
-	if (rdev->family == CHIP_KABINI)
-		pi->high_voltage_t = 4001;
-
 	pi->enable_nb_dpm = true;
 
 	pi->caps_power_containment = true;
@@ -2635,13 +2726,13 @@ int kv_dpm_init(struct radeon_device *rdev)
 	pi->caps_sclk_ds = true;
 	pi->enable_auto_thermal_throttling = true;
 	pi->disable_nb_ps3_in_battery = false;
-	pi->bapm_enable = false;
+	pi->bapm_enable = true;
 	pi->voltage_drop_t = 0;
 	pi->caps_sclk_throttle_low_notification = false;
 	pi->caps_fps = false; /* true? */
 	pi->caps_uvd_pg = true;
 	pi->caps_uvd_dpm = true;
-	pi->caps_vce_pg = false;
+	pi->caps_vce_pg = false; /* XXX true */
 	pi->caps_samu_pg = false;
 	pi->caps_acp_pg = false;
 	pi->caps_stable_p_state = false;
diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c
index f59a9e9fccf..5a33ca68186 100644
--- a/drivers/gpu/drm/radeon/ni.c
+++ b/drivers/gpu/drm/radeon/ni.c
@@ -174,6 +174,7 @@ extern void evergreen_pcie_gen2_enable(struct radeon_device *rdev);
 extern void evergreen_program_aspm(struct radeon_device *rdev);
 extern void sumo_rlc_fini(struct radeon_device *rdev);
 extern int sumo_rlc_init(struct radeon_device *rdev);
+extern void evergreen_gpu_pci_config_reset(struct radeon_device *rdev);
 
 /* Firmware Names */
 MODULE_FIRMWARE("radeon/BARTS_pfp.bin");
@@ -1056,6 +1057,18 @@ static void cayman_gpu_init(struct radeon_device *rdev)
 			disabled_rb_mask &= ~(1 << i);
 	}
 
+	for (i = 0; i < rdev->config.cayman.max_shader_engines; i++) {
+		u32 simd_disable_bitmap;
+
+		WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i));
+		WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i));
+		simd_disable_bitmap = (RREG32(CC_GC_SHADER_PIPE_CONFIG) & 0xffff0000) >> 16;
+		simd_disable_bitmap |= 0xffffffff << rdev->config.cayman.max_simds_per_se;
+		tmp <<= 16;
+		tmp |= simd_disable_bitmap;
+	}
+	rdev->config.cayman.active_simds = hweight32(~tmp);
+
 	WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES);
 	WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES);
 
@@ -1227,12 +1240,14 @@ static int cayman_pcie_gart_enable(struct radeon_device *rdev)
 	       SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU);
 	/* Setup L2 cache */
 	WREG32(VM_L2_CNTL, ENABLE_L2_CACHE |
+	       ENABLE_L2_FRAGMENT_PROCESSING |
 	       ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE |
 	       ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE |
 	       EFFECTIVE_L2_QUEUE_SIZE(7) |
 	       CONTEXT1_IDENTITY_ACCESS_MODE(1));
 	WREG32(VM_L2_CNTL2, INVALIDATE_ALL_L1_TLBS | INVALIDATE_L2_CACHE);
 	WREG32(VM_L2_CNTL3, L2_CACHE_BIGK_ASSOCIATIVITY |
+	       BANK_SELECT(6) |
 	       L2_CACHE_BIGK_FRAGMENT_SIZE(6));
 	/* setup context0 */
 	WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR, rdev->mc.gtt_start >> 12);
@@ -1265,6 +1280,7 @@ static int cayman_pcie_gart_enable(struct radeon_device *rdev)
 	       (u32)(rdev->dummy_page.addr >> 12));
 	WREG32(VM_CONTEXT1_CNTL2, 4);
 	WREG32(VM_CONTEXT1_CNTL, ENABLE_CONTEXT | PAGE_TABLE_DEPTH(1) |
+				PAGE_TABLE_BLOCK_SIZE(radeon_vm_block_size - 9) |
 				RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT |
 				RANGE_PROTECTION_FAULT_ENABLE_DEFAULT |
 				DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT |
@@ -1330,20 +1346,19 @@ void cayman_fence_ring_emit(struct radeon_device *rdev,
 {
 	struct radeon_ring *ring = &rdev->ring[fence->ring];
 	u64 addr = rdev->fence_drv[fence->ring].gpu_addr;
+	u32 cp_coher_cntl = PACKET3_FULL_CACHE_ENA | PACKET3_TC_ACTION_ENA |
+		PACKET3_SH_ACTION_ENA;
 
 	/* flush read cache over gart for this vmid */
-	radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
-	radeon_ring_write(ring, (CP_COHER_CNTL2 - PACKET3_SET_CONFIG_REG_START) >> 2);
-	radeon_ring_write(ring, 0);
 	radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
-	radeon_ring_write(ring, PACKET3_TC_ACTION_ENA | PACKET3_SH_ACTION_ENA);
+	radeon_ring_write(ring, PACKET3_ENGINE_ME | cp_coher_cntl);
 	radeon_ring_write(ring, 0xFFFFFFFF);
 	radeon_ring_write(ring, 0);
 	radeon_ring_write(ring, 10); /* poll interval */
 	/* EVENT_WRITE_EOP - flush caches, send int */
 	radeon_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4));
 	radeon_ring_write(ring, EVENT_TYPE(CACHE_FLUSH_AND_INV_EVENT_TS) | EVENT_INDEX(5));
-	radeon_ring_write(ring, addr & 0xffffffff);
+	radeon_ring_write(ring, lower_32_bits(addr));
 	radeon_ring_write(ring, (upper_32_bits(addr) & 0xff) | DATA_SEL(1) | INT_SEL(2));
 	radeon_ring_write(ring, fence->seq);
 	radeon_ring_write(ring, 0);
@@ -1352,6 +1367,8 @@ void cayman_fence_ring_emit(struct radeon_device *rdev,
 void cayman_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 {
 	struct radeon_ring *ring = &rdev->ring[ib->ring];
+	u32 cp_coher_cntl = PACKET3_FULL_CACHE_ENA | PACKET3_TC_ACTION_ENA |
+		PACKET3_SH_ACTION_ENA;
 
 	/* set to DX10/11 mode */
 	radeon_ring_write(ring, PACKET3(PACKET3_MODE_CONTROL, 0));
@@ -1376,14 +1393,11 @@ void cayman_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 			  (ib->vm ? (ib->vm->id << 24) : 0));
 
 	/* flush read cache over gart for this vmid */
-	radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
-	radeon_ring_write(ring, (CP_COHER_CNTL2 - PACKET3_SET_CONFIG_REG_START) >> 2);
-	radeon_ring_write(ring, ib->vm ? ib->vm->id : 0);
 	radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
-	radeon_ring_write(ring, PACKET3_TC_ACTION_ENA | PACKET3_SH_ACTION_ENA);
+	radeon_ring_write(ring, PACKET3_ENGINE_ME | cp_coher_cntl);
 	radeon_ring_write(ring, 0xFFFFFFFF);
 	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 10); /* poll interval */
+	radeon_ring_write(ring, ((ib->vm ? ib->vm->id : 0) << 24) | 10); /* poll interval */
 }
 
 static void cayman_cp_enable(struct radeon_device *rdev, bool enable)
@@ -1391,13 +1405,63 @@ static void cayman_cp_enable(struct radeon_device *rdev, bool enable)
 	if (enable)
 		WREG32(CP_ME_CNTL, 0);
 	else {
-		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+		if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+			radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 		WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT));
 		WREG32(SCRATCH_UMSK, 0);
 		rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
 	}
 }
 
+u32 cayman_gfx_get_rptr(struct radeon_device *rdev,
+			struct radeon_ring *ring)
+{
+	u32 rptr;
+
+	if (rdev->wb.enabled)
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	else {
+		if (ring->idx == RADEON_RING_TYPE_GFX_INDEX)
+			rptr = RREG32(CP_RB0_RPTR);
+		else if (ring->idx == CAYMAN_RING_TYPE_CP1_INDEX)
+			rptr = RREG32(CP_RB1_RPTR);
+		else
+			rptr = RREG32(CP_RB2_RPTR);
+	}
+
+	return rptr;
+}
+
+u32 cayman_gfx_get_wptr(struct radeon_device *rdev,
+			struct radeon_ring *ring)
+{
+	u32 wptr;
+
+	if (ring->idx == RADEON_RING_TYPE_GFX_INDEX)
+		wptr = RREG32(CP_RB0_WPTR);
+	else if (ring->idx == CAYMAN_RING_TYPE_CP1_INDEX)
+		wptr = RREG32(CP_RB1_WPTR);
+	else
+		wptr = RREG32(CP_RB2_WPTR);
+
+	return wptr;
+}
+
+void cayman_gfx_set_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring)
+{
+	if (ring->idx == RADEON_RING_TYPE_GFX_INDEX) {
+		WREG32(CP_RB0_WPTR, ring->wptr);
+		(void)RREG32(CP_RB0_WPTR);
+	} else if (ring->idx == CAYMAN_RING_TYPE_CP1_INDEX) {
+		WREG32(CP_RB1_WPTR, ring->wptr);
+		(void)RREG32(CP_RB1_WPTR);
+	} else {
+		WREG32(CP_RB2_WPTR, ring->wptr);
+		(void)RREG32(CP_RB2_WPTR);
+	}
+}
+
 static int cayman_cp_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
@@ -1526,6 +1590,16 @@ static int cayman_cp_resume(struct radeon_device *rdev)
 		CP_RB1_BASE,
 		CP_RB2_BASE
 	};
+	static const unsigned cp_rb_rptr[] = {
+		CP_RB0_RPTR,
+		CP_RB1_RPTR,
+		CP_RB2_RPTR
+	};
+	static const unsigned cp_rb_wptr[] = {
+		CP_RB0_WPTR,
+		CP_RB1_WPTR,
+		CP_RB2_WPTR
+	};
 	struct radeon_ring *ring;
 	int i, r;
 
@@ -1583,9 +1657,9 @@ static int cayman_cp_resume(struct radeon_device *rdev)
 		ring = &rdev->ring[ridx[i]];
 		WREG32_P(cp_rb_cntl[i], RB_RPTR_WR_ENA, ~RB_RPTR_WR_ENA);
 
-		ring->rptr = ring->wptr = 0;
-		WREG32(ring->rptr_reg, ring->rptr);
-		WREG32(ring->wptr_reg, ring->wptr);
+		ring->wptr = 0;
+		WREG32(cp_rb_rptr[i], 0);
+		WREG32(cp_rb_wptr[i], ring->wptr);
 
 		mdelay(1);
 		WREG32_P(cp_rb_cntl[i], 0, ~RB_RPTR_WR_ENA);
@@ -1605,6 +1679,9 @@ static int cayman_cp_resume(struct radeon_device *rdev)
 		return r;
 	}
 
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+
 	return 0;
 }
 
@@ -1831,8 +1908,10 @@ int cayman_asic_reset(struct radeon_device *rdev)
 
 	reset_mask = cayman_gpu_check_soft_reset(rdev);
 
-	if (!reset_mask)
-		r600_set_bios_scratch_engine_hung(rdev, false);
+	if (reset_mask)
+		evergreen_gpu_pci_config_reset(rdev);
+
+	r600_set_bios_scratch_engine_hung(rdev, false);
 
 	return 0;
 }
@@ -1853,11 +1932,9 @@ bool cayman_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 	if (!(reset_mask & (RADEON_RESET_GFX |
 			    RADEON_RESET_COMPUTE |
 			    RADEON_RESET_CP))) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force CP activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
@@ -1878,23 +1955,7 @@ static int cayman_startup(struct radeon_device *rdev)
 
 	evergreen_mc_program(rdev);
 
-	if (rdev->flags & RADEON_IS_IGP) {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
-			r = ni_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
-	} else {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw || !rdev->mc_fw) {
-			r = ni_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
-
+	if (!(rdev->flags & RADEON_IS_IGP) && !rdev->pm.dpm_enabled) {
 		r = ni_mc_load_microcode(rdev);
 		if (r) {
 			DRM_ERROR("Failed to load MC firmware!\n");
@@ -1981,23 +2042,18 @@ static int cayman_startup(struct radeon_device *rdev)
 	evergreen_irq_set(rdev);
 
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     CP_RB0_RPTR, CP_RB0_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR + DMA0_REGISTER_OFFSET,
-			     DMA_RB_WPTR + DMA0_REGISTER_OFFSET,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, CAYMAN_WB_DMA1_RPTR_OFFSET,
-			     DMA_RB_RPTR + DMA1_REGISTER_OFFSET,
-			     DMA_RB_WPTR + DMA1_REGISTER_OFFSET,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
 	if (r)
 		return r;
@@ -2016,7 +2072,6 @@ static int cayman_startup(struct radeon_device *rdev)
 	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 	if (ring->ring_size) {
 		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 				     RADEON_CP_PACKET2);
 		if (!r)
 			r = uvd_v1_0_init(rdev);
@@ -2063,6 +2118,9 @@ int cayman_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	ni_init_golden_registers(rdev);
 
+	if (rdev->pm.pm_method == PM_METHOD_DPM)
+		radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = cayman_startup(rdev);
 	if (r) {
@@ -2075,6 +2133,7 @@ int cayman_resume(struct radeon_device *rdev)
 
 int cayman_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	if (ASIC_IS_DCE6(rdev))
 		dce6_audio_fini(rdev);
 	else
@@ -2145,6 +2204,27 @@ int cayman_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (rdev->flags & RADEON_IS_IGP) {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
+			r = ni_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	} else {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw || !rdev->mc_fw) {
+			r = ni_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	ring->ring_obj = NULL;
 	r600_ring_init(rdev, ring, 1024 * 1024);
 
@@ -2204,6 +2284,7 @@ int cayman_init(struct radeon_device *rdev)
 
 void cayman_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	cayman_cp_fini(rdev);
 	cayman_dma_fini(rdev);
 	r600_irq_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/ni_dma.c b/drivers/gpu/drm/radeon/ni_dma.c
index bdeb65ed365..6378e027669 100644
--- a/drivers/gpu/drm/radeon/ni_dma.c
+++ b/drivers/gpu/drm/radeon/ni_dma.c
@@ -43,6 +43,75 @@ u32 cayman_gpu_check_soft_reset(struct radeon_device *rdev);
  */
 
 /**
+ * cayman_dma_get_rptr - get the current read pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Get the current rptr from the hardware (cayman+).
+ */
+uint32_t cayman_dma_get_rptr(struct radeon_device *rdev,
+			     struct radeon_ring *ring)
+{
+	u32 rptr, reg;
+
+	if (rdev->wb.enabled) {
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	} else {
+		if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+			reg = DMA_RB_RPTR + DMA0_REGISTER_OFFSET;
+		else
+			reg = DMA_RB_RPTR + DMA1_REGISTER_OFFSET;
+
+		rptr = RREG32(reg);
+	}
+
+	return (rptr & 0x3fffc) >> 2;
+}
+
+/**
+ * cayman_dma_get_wptr - get the current write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Get the current wptr from the hardware (cayman+).
+ */
+uint32_t cayman_dma_get_wptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	u32 reg;
+
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		reg = DMA_RB_WPTR + DMA0_REGISTER_OFFSET;
+	else
+		reg = DMA_RB_WPTR + DMA1_REGISTER_OFFSET;
+
+	return (RREG32(reg) & 0x3fffc) >> 2;
+}
+
+/**
+ * cayman_dma_set_wptr - commit the write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Write the wptr back to the hardware (cayman+).
+ */
+void cayman_dma_set_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring)
+{
+	u32 reg;
+
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		reg = DMA_RB_WPTR + DMA0_REGISTER_OFFSET;
+	else
+		reg = DMA_RB_WPTR + DMA1_REGISTER_OFFSET;
+
+	WREG32(reg, (ring->wptr << 2) & 0x3fffc);
+}
+
+/**
  * cayman_dma_ring_ib_execute - Schedule an IB on the DMA engine
  *
  * @rdev: radeon_device pointer
@@ -88,7 +157,9 @@ void cayman_dma_stop(struct radeon_device *rdev)
 {
 	u32 rb_cntl;
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if ((rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX) ||
+	    (rdev->asic->copy.copy_ring_index == CAYMAN_RING_TYPE_DMA1_INDEX))
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 
 	/* dma0 */
 	rb_cntl = RREG32(DMA_RB_CNTL + DMA0_REGISTER_OFFSET);
@@ -177,8 +248,6 @@ int cayman_dma_resume(struct radeon_device *rdev)
 		ring->wptr = 0;
 		WREG32(DMA_RB_WPTR + reg_offset, ring->wptr << 2);
 
-		ring->rptr = RREG32(DMA_RB_RPTR + reg_offset) >> 2;
-
 		WREG32(DMA_RB_CNTL + reg_offset, rb_cntl | DMA_RB_ENABLE);
 
 		ring->ready = true;
@@ -190,7 +259,9 @@ int cayman_dma_resume(struct radeon_device *rdev)
 		}
 	}
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+	if ((rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX) ||
+	    (rdev->asic->copy.copy_ring_index == CAYMAN_RING_TYPE_DMA1_INDEX))
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
 
 	return 0;
 }
@@ -229,11 +300,9 @@ bool cayman_dma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 		mask = RADEON_RESET_DMA1;
 
 	if (!(reset_mask & mask)) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force ring activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c
index 49c4d48f54d..01fc4888e6f 100644
--- a/drivers/gpu/drm/radeon/ni_dpm.c
+++ b/drivers/gpu/drm/radeon/ni_dpm.c
@@ -720,6 +720,8 @@ static const u32 cayman_sysls_enable[] =
 struct rv7xx_power_info *rv770_get_pi(struct radeon_device *rdev);
 struct evergreen_power_info *evergreen_get_pi(struct radeon_device *rdev);
 
+extern int ni_mc_load_microcode(struct radeon_device *rdev);
+
 struct ni_power_info *ni_get_pi(struct radeon_device *rdev)
 {
         struct ni_power_info *pi = rdev->pm.dpm.priv;
@@ -1313,7 +1315,7 @@ static void ni_populate_smc_voltage_tables(struct radeon_device *rdev,
 
 		table->voltageMaskTable.highMask[NISLANDS_SMC_VOLTAGEMASK_VDDCI] = 0;
 		table->voltageMaskTable.lowMask[NISLANDS_SMC_VOLTAGEMASK_VDDCI] =
-			cpu_to_be32(eg_pi->vddc_voltage_table.mask_low);
+			cpu_to_be32(eg_pi->vddci_voltage_table.mask_low);
 	}
 }
 
@@ -2586,7 +2588,7 @@ static int ni_populate_sq_ramping_values(struct radeon_device *rdev,
 	if (NISLANDS_DPM2_SQ_RAMP_STI_SIZE > (STI_SIZE_MASK >> STI_SIZE_SHIFT))
 		enable_sq_ramping = false;
 
-	if (NISLANDS_DPM2_SQ_RAMP_LTI_RATIO <= (LTI_RATIO_MASK >> LTI_RATIO_SHIFT))
+	if (NISLANDS_DPM2_SQ_RAMP_LTI_RATIO > (LTI_RATIO_MASK >> LTI_RATIO_SHIFT))
 		enable_sq_ramping = false;
 
 	for (i = 0; i < state->performance_level_count; i++) {
@@ -3565,7 +3567,11 @@ void ni_set_uvd_clock_after_set_eng_clock(struct radeon_device *rdev,
 void ni_dpm_setup_asic(struct radeon_device *rdev)
 {
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	int r;
 
+	r = ni_mc_load_microcode(rdev);
+	if (r)
+		DRM_ERROR("Failed to load MC firmware!\n");
 	ni_read_clock_registers(rdev);
 	btc_read_arb_registers(rdev);
 	rv770_get_memory_type(rdev);
@@ -3710,21 +3716,6 @@ int ni_dpm_enable(struct radeon_device *rdev)
 	if (eg_pi->ls_clock_gating)
 		ni_ls_clockgating_enable(rdev, true);
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-		PPSMC_Result result;
-
-		ret = rv770_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, 0xff * 1000);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-		result = rv770_send_msg_to_smc(rdev, PPSMC_MSG_EnableThermalInterrupt);
-
-		if (result != PPSMC_Result_OK)
-			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
-	}
-
 	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
 	ni_update_current_ps(rdev, boot_ps);
@@ -3954,7 +3945,6 @@ static void ni_parse_pplib_clock_info(struct radeon_device *rdev,
 	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
 	struct ni_ps *ps = ni_get_ps(rps);
-	u16 vddc;
 	struct rv7xx_pl *pl = &ps->performance_levels[index];
 
 	ps->performance_level_count = index + 1;
@@ -3970,8 +3960,8 @@ static void ni_parse_pplib_clock_info(struct radeon_device *rdev,
 
 	/* patch up vddc if necessary */
 	if (pl->vddc == 0xff01) {
-		if (radeon_atom_get_max_vddc(rdev, 0, 0, &vddc) == 0)
-			pl->vddc = vddc;
+		if (pi->max_vddc)
+			pl->vddc = pi->max_vddc;
 	}
 
 	if (rps->class & ATOM_PPLIB_CLASSIFICATION_ACPI) {
@@ -4035,9 +4025,6 @@ static int ni_parse_power_table(struct radeon_device *rdev)
 				  power_info->pplib.ucNumStates, GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
-	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
-	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
-	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
 
 	for (i = 0; i < power_info->pplib.ucNumStates; i++) {
 		power_state = (union pplib_power_state *)
@@ -4099,6 +4086,10 @@ int ni_dpm_init(struct radeon_device *rdev)
 	pi->min_vddc_in_table = 0;
 	pi->max_vddc_in_table = 0;
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = ni_parse_power_table(rdev);
 	if (ret)
 		return ret;
@@ -4331,7 +4322,8 @@ void ni_dpm_print_power_state(struct radeon_device *rdev,
 void ni_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
 						    struct seq_file *m)
 {
-	struct radeon_ps *rps = rdev->pm.dpm.current_ps;
+	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	struct radeon_ps *rps = &eg_pi->current_rps;
 	struct ni_ps *ps = ni_get_ps(rps);
 	struct rv7xx_pl *pl;
 	u32 current_index =
diff --git a/drivers/gpu/drm/radeon/nid.h b/drivers/gpu/drm/radeon/nid.h
index 22421bc80c0..2e12e4d6925 100644
--- a/drivers/gpu/drm/radeon/nid.h
+++ b/drivers/gpu/drm/radeon/nid.h
@@ -128,6 +128,7 @@
 #define		READ_PROTECTION_FAULT_ENABLE_DEFAULT		(1 << 16)
 #define		WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT		(1 << 18)
 #define		WRITE_PROTECTION_FAULT_ENABLE_DEFAULT		(1 << 19)
+#define		PAGE_TABLE_BLOCK_SIZE(x)			(((x) & 0xF) << 24)
 #define VM_CONTEXT1_CNTL				0x1414
 #define VM_CONTEXT0_CNTL2				0x1430
 #define VM_CONTEXT1_CNTL2				0x1434
@@ -1154,6 +1155,7 @@
 #              define PACKET3_DB_ACTION_ENA        (1 << 26)
 #              define PACKET3_SH_ACTION_ENA        (1 << 27)
 #              define PACKET3_SX_ACTION_ENA        (1 << 28)
+#              define PACKET3_ENGINE_ME            (1 << 31)
 #define	PACKET3_ME_INITIALIZE				0x44
 #define		PACKET3_ME_INITIALIZE_DEVICE_ID(x) ((x) << 16)
 #define	PACKET3_COND_WRITE				0x45
diff --git a/drivers/gpu/drm/radeon/pptable.h b/drivers/gpu/drm/radeon/pptable.h
index da43ab32883..2d532996c69 100644
--- a/drivers/gpu/drm/radeon/pptable.h
+++ b/drivers/gpu/drm/radeon/pptable.h
@@ -23,7 +23,7 @@
 #ifndef _PPTABLE_H
 #define _PPTABLE_H
 
-#pragma pack(push, 1)
+#pragma pack(1)
 
 typedef struct _ATOM_PPLIB_THERMALCONTROLLER
 
@@ -677,6 +677,6 @@ typedef struct _ATOM_PPLIB_PPM_Table
       ULONG  ulTjmax;
 } ATOM_PPLIB_PPM_Table;
 
-#pragma pack(pop)
+#pragma pack()
 
 #endif
diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c
index 10abc4d5a6c..1544efcf1c3 100644
--- a/drivers/gpu/drm/radeon/r100.c
+++ b/drivers/gpu/drm/radeon/r100.c
@@ -142,36 +142,6 @@ void r100_wait_for_vblank(struct radeon_device *rdev, int crtc)
 }
 
 /**
- * r100_pre_page_flip - pre-pageflip callback.
- *
- * @rdev: radeon_device pointer
- * @crtc: crtc to prepare for pageflip on
- *
- * Pre-pageflip callback (r1xx-r4xx).
- * Enables the pageflip irq (vblank irq).
- */
-void r100_pre_page_flip(struct radeon_device *rdev, int crtc)
-{
-	/* enable the pflip int */
-	radeon_irq_kms_pflip_irq_get(rdev, crtc);
-}
-
-/**
- * r100_post_page_flip - pos-pageflip callback.
- *
- * @rdev: radeon_device pointer
- * @crtc: crtc to cleanup pageflip on
- *
- * Post-pageflip callback (r1xx-r4xx).
- * Disables the pageflip irq (vblank irq).
- */
-void r100_post_page_flip(struct radeon_device *rdev, int crtc)
-{
-	/* disable the pflip int */
-	radeon_irq_kms_pflip_irq_put(rdev, crtc);
-}
-
-/**
  * r100_page_flip - pageflip callback.
  *
  * @rdev: radeon_device pointer
@@ -182,9 +152,8 @@ void r100_post_page_flip(struct radeon_device *rdev, int crtc)
  * During vblank we take the crtc lock and wait for the update_pending
  * bit to go high, when it does, we release the lock, and allow the
  * double buffered update to take place.
- * Returns the current update pending status.
  */
-u32 r100_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
+void r100_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 {
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 	u32 tmp = ((u32)crtc_base) | RADEON_CRTC_OFFSET__OFFSET_LOCK;
@@ -206,8 +175,24 @@ u32 r100_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 	tmp &= ~RADEON_CRTC_OFFSET__OFFSET_LOCK;
 	WREG32(RADEON_CRTC_OFFSET + radeon_crtc->crtc_offset, tmp);
 
+}
+
+/**
+ * r100_page_flip_pending - check if page flip is still pending
+ *
+ * @rdev: radeon_device pointer
+ * @crtc_id: crtc to check
+ *
+ * Check if the last pagefilp is still pending (r1xx-r4xx).
+ * Returns the current update pending status.
+ */
+bool r100_page_flip_pending(struct radeon_device *rdev, int crtc_id)
+{
+	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
+
 	/* Return current update_pending status: */
-	return RREG32(RADEON_CRTC_OFFSET + radeon_crtc->crtc_offset) & RADEON_CRTC_OFFSET__GUI_TRIG_OFFSET;
+	return !!(RREG32(RADEON_CRTC_OFFSET + radeon_crtc->crtc_offset) &
+		RADEON_CRTC_OFFSET__GUI_TRIG_OFFSET);
 }
 
 /**
@@ -697,15 +682,11 @@ void r100_pci_gart_disable(struct radeon_device *rdev)
 	WREG32(RADEON_AIC_HI_ADDR, 0);
 }
 
-int r100_pci_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr)
+void r100_pci_gart_set_page(struct radeon_device *rdev, unsigned i,
+			    uint64_t addr)
 {
 	u32 *gtt = rdev->gart.ptr;
-
-	if (i < 0 || i > rdev->gart.num_gpu_pages) {
-		return -EINVAL;
-	}
 	gtt[i] = cpu_to_le32(lower_32_bits(addr));
-	return 0;
 }
 
 void r100_pci_gart_fini(struct radeon_device *rdev)
@@ -794,7 +775,7 @@ int r100_irq_process(struct radeon_device *rdev)
 				wake_up(&rdev->irq.vblank_queue);
 			}
 			if (atomic_read(&rdev->irq.pflip[0]))
-				radeon_crtc_handle_flip(rdev, 0);
+				radeon_crtc_handle_vblank(rdev, 0);
 		}
 		if (status & RADEON_CRTC2_VBLANK_STAT) {
 			if (rdev->irq.crtc_vblank_int[1]) {
@@ -803,7 +784,7 @@ int r100_irq_process(struct radeon_device *rdev)
 				wake_up(&rdev->irq.vblank_queue);
 			}
 			if (atomic_read(&rdev->irq.pflip[1]))
-				radeon_crtc_handle_flip(rdev, 1);
+				radeon_crtc_handle_vblank(rdev, 1);
 		}
 		if (status & RADEON_FP_DETECT_STAT) {
 			queue_hotplug = true;
@@ -1050,6 +1031,36 @@ static int r100_cp_init_microcode(struct radeon_device *rdev)
 	return err;
 }
 
+u32 r100_gfx_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	u32 rptr;
+
+	if (rdev->wb.enabled)
+		rptr = le32_to_cpu(rdev->wb.wb[ring->rptr_offs/4]);
+	else
+		rptr = RREG32(RADEON_CP_RB_RPTR);
+
+	return rptr;
+}
+
+u32 r100_gfx_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	u32 wptr;
+
+	wptr = RREG32(RADEON_CP_RB_WPTR);
+
+	return wptr;
+}
+
+void r100_gfx_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring)
+{
+	WREG32(RADEON_CP_RB_WPTR, ring->wptr);
+	(void)RREG32(RADEON_CP_RB_WPTR);
+}
+
 static void r100_cp_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
@@ -1102,7 +1113,6 @@ int r100_cp_init(struct radeon_device *rdev, unsigned ring_size)
 	ring_size = (1 << (rb_bufsz + 1)) * 4;
 	r100_cp_load_microcode(rdev);
 	r = radeon_ring_init(rdev, ring, ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     RADEON_CP_RB_RPTR, RADEON_CP_RB_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r) {
 		return r;
@@ -1164,7 +1174,6 @@ int r100_cp_init(struct radeon_device *rdev, unsigned ring_size)
 
 	WREG32(RADEON_CP_RB_CNTL, tmp);
 	udelay(10);
-	ring->rptr = RREG32(RADEON_CP_RB_RPTR);
 	/* Set cp mode to bus mastering & enable cp*/
 	WREG32(RADEON_CP_CSQ_MODE,
 	       REG_SET(RADEON_INDIRECT2_START, indirect2_start) |
@@ -1246,12 +1255,12 @@ int r100_reloc_pitch_offset(struct radeon_cs_parser *p,
 
 	value = radeon_get_ib_value(p, idx);
 	tmp = value & 0x003fffff;
-	tmp += (((u32)reloc->lobj.gpu_offset) >> 10);
+	tmp += (((u32)reloc->gpu_offset) >> 10);
 
 	if (!(p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS)) {
-		if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+		if (reloc->tiling_flags & RADEON_TILING_MACRO)
 			tile_flags |= RADEON_DST_TILE_MACRO;
-		if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) {
+		if (reloc->tiling_flags & RADEON_TILING_MICRO) {
 			if (reg == RADEON_SRC_PITCH_OFFSET) {
 				DRM_ERROR("Cannot src blit from microtiled surface\n");
 				radeon_cs_dump_packet(p, pkt);
@@ -1297,7 +1306,7 @@ int r100_packet3_load_vbpntr(struct radeon_cs_parser *p,
 			return r;
 		}
 		idx_value = radeon_get_ib_value(p, idx);
-		ib[idx+1] = radeon_get_ib_value(p, idx + 1) + ((u32)reloc->lobj.gpu_offset);
+		ib[idx+1] = radeon_get_ib_value(p, idx + 1) + ((u32)reloc->gpu_offset);
 
 		track->arrays[i + 0].esize = idx_value >> 8;
 		track->arrays[i + 0].robj = reloc->robj;
@@ -1309,7 +1318,7 @@ int r100_packet3_load_vbpntr(struct radeon_cs_parser *p,
 			radeon_cs_dump_packet(p, pkt);
 			return r;
 		}
-		ib[idx+2] = radeon_get_ib_value(p, idx + 2) + ((u32)reloc->lobj.gpu_offset);
+		ib[idx+2] = radeon_get_ib_value(p, idx + 2) + ((u32)reloc->gpu_offset);
 		track->arrays[i + 1].robj = reloc->robj;
 		track->arrays[i + 1].esize = idx_value >> 24;
 		track->arrays[i + 1].esize &= 0x7F;
@@ -1323,7 +1332,7 @@ int r100_packet3_load_vbpntr(struct radeon_cs_parser *p,
 			return r;
 		}
 		idx_value = radeon_get_ib_value(p, idx);
-		ib[idx+1] = radeon_get_ib_value(p, idx + 1) + ((u32)reloc->lobj.gpu_offset);
+		ib[idx+1] = radeon_get_ib_value(p, idx + 1) + ((u32)reloc->gpu_offset);
 		track->arrays[i + 0].robj = reloc->robj;
 		track->arrays[i + 0].esize = idx_value >> 8;
 		track->arrays[i + 0].esize &= 0x7F;
@@ -1566,7 +1575,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 		track->zb.robj = reloc->robj;
 		track->zb.offset = idx_value;
 		track->zb_dirty = true;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case RADEON_RB3D_COLOROFFSET:
 		r = radeon_cs_packet_next_reloc(p, &reloc, 0);
@@ -1579,7 +1588,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 		track->cb[0].robj = reloc->robj;
 		track->cb[0].offset = idx_value;
 		track->cb_dirty = true;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case RADEON_PP_TXOFFSET_0:
 	case RADEON_PP_TXOFFSET_1:
@@ -1593,16 +1602,16 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 			return r;
 		}
 		if (!(p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS)) {
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			if (reloc->tiling_flags & RADEON_TILING_MACRO)
 				tile_flags |= RADEON_TXO_MACRO_TILE;
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+			if (reloc->tiling_flags & RADEON_TILING_MICRO)
 				tile_flags |= RADEON_TXO_MICRO_TILE_X2;
 
 			tmp = idx_value & ~(0x7 << 2);
 			tmp |= tile_flags;
-			ib[idx] = tmp + ((u32)reloc->lobj.gpu_offset);
+			ib[idx] = tmp + ((u32)reloc->gpu_offset);
 		} else
-			ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+			ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		track->textures[i].robj = reloc->robj;
 		track->tex_dirty = true;
 		break;
@@ -1620,7 +1629,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 			return r;
 		}
 		track->textures[0].cube_info[i].offset = idx_value;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		track->textures[0].cube_info[i].robj = reloc->robj;
 		track->tex_dirty = true;
 		break;
@@ -1638,7 +1647,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 			return r;
 		}
 		track->textures[1].cube_info[i].offset = idx_value;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		track->textures[1].cube_info[i].robj = reloc->robj;
 		track->tex_dirty = true;
 		break;
@@ -1656,7 +1665,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 			return r;
 		}
 		track->textures[2].cube_info[i].offset = idx_value;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		track->textures[2].cube_info[i].robj = reloc->robj;
 		track->tex_dirty = true;
 		break;
@@ -1674,9 +1683,9 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 			return r;
 		}
 		if (!(p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS)) {
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			if (reloc->tiling_flags & RADEON_TILING_MACRO)
 				tile_flags |= RADEON_COLOR_TILE_ENABLE;
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+			if (reloc->tiling_flags & RADEON_TILING_MICRO)
 				tile_flags |= RADEON_COLOR_MICROTILE_ENABLE;
 
 			tmp = idx_value & ~(0x7 << 16);
@@ -1744,7 +1753,7 @@ static int r100_packet0_check(struct radeon_cs_parser *p,
 			radeon_cs_dump_packet(p, pkt);
 			return r;
 		}
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case RADEON_PP_CNTL:
 		{
@@ -1904,7 +1913,7 @@ static int r100_packet3_check(struct radeon_cs_parser *p,
 			radeon_cs_dump_packet(p, pkt);
 			return r;
 		}
-		ib[idx+1] = radeon_get_ib_value(p, idx+1) + ((u32)reloc->lobj.gpu_offset);
+		ib[idx+1] = radeon_get_ib_value(p, idx+1) + ((u32)reloc->gpu_offset);
 		r = r100_cs_track_check_pkt3_indx_buffer(p, pkt, reloc->robj);
 		if (r) {
 			return r;
@@ -1918,7 +1927,7 @@ static int r100_packet3_check(struct radeon_cs_parser *p,
 			radeon_cs_dump_packet(p, pkt);
 			return r;
 		}
-		ib[idx] = radeon_get_ib_value(p, idx) + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = radeon_get_ib_value(p, idx) + ((u32)reloc->gpu_offset);
 		track->num_arrays = 1;
 		track->vtx_size = r100_get_vtx_size(radeon_get_ib_value(p, idx + 2));
 
@@ -2494,11 +2503,9 @@ bool r100_gpu_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 
 	rbbm_status = RREG32(R_000E40_RBBM_STATUS);
 	if (!G_000E40_GUI_ACTIVE(rbbm_status)) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force CP activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
@@ -3194,12 +3201,12 @@ void r100_bandwidth_update(struct radeon_device *rdev)
 
 	if (rdev->mode_info.crtcs[0]->base.enabled) {
 		mode1 = &rdev->mode_info.crtcs[0]->base.mode;
-		pixel_bytes1 = rdev->mode_info.crtcs[0]->base.fb->bits_per_pixel / 8;
+		pixel_bytes1 = rdev->mode_info.crtcs[0]->base.primary->fb->bits_per_pixel / 8;
 	}
 	if (!(rdev->flags & RADEON_SINGLE_CRTC)) {
 		if (rdev->mode_info.crtcs[1]->base.enabled) {
 			mode2 = &rdev->mode_info.crtcs[1]->base.mode;
-			pixel_bytes2 = rdev->mode_info.crtcs[1]->base.fb->bits_per_pixel / 8;
+			pixel_bytes2 = rdev->mode_info.crtcs[1]->base.primary->fb->bits_per_pixel / 8;
 		}
 	}
 
@@ -3923,6 +3930,7 @@ int r100_resume(struct radeon_device *rdev)
 
 int r100_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
 	r100_irq_disable(rdev);
@@ -3933,6 +3941,7 @@ int r100_suspend(struct radeon_device *rdev)
 
 void r100_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -4039,6 +4048,9 @@ int r100_init(struct radeon_device *rdev)
 	}
 	r100_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = r100_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/r200.c b/drivers/gpu/drm/radeon/r200.c
index b3807edb193..58f0473aa73 100644
--- a/drivers/gpu/drm/radeon/r200.c
+++ b/drivers/gpu/drm/radeon/r200.c
@@ -185,7 +185,7 @@ int r200_packet0_check(struct radeon_cs_parser *p,
 		track->zb.robj = reloc->robj;
 		track->zb.offset = idx_value;
 		track->zb_dirty = true;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case RADEON_RB3D_COLOROFFSET:
 		r = radeon_cs_packet_next_reloc(p, &reloc, 0);
@@ -198,7 +198,7 @@ int r200_packet0_check(struct radeon_cs_parser *p,
 		track->cb[0].robj = reloc->robj;
 		track->cb[0].offset = idx_value;
 		track->cb_dirty = true;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case R200_PP_TXOFFSET_0:
 	case R200_PP_TXOFFSET_1:
@@ -215,16 +215,16 @@ int r200_packet0_check(struct radeon_cs_parser *p,
 			return r;
 		}
 		if (!(p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS)) {
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			if (reloc->tiling_flags & RADEON_TILING_MACRO)
 				tile_flags |= R200_TXO_MACRO_TILE;
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+			if (reloc->tiling_flags & RADEON_TILING_MICRO)
 				tile_flags |= R200_TXO_MICRO_TILE;
 
 			tmp = idx_value & ~(0x7 << 2);
 			tmp |= tile_flags;
-			ib[idx] = tmp + ((u32)reloc->lobj.gpu_offset);
+			ib[idx] = tmp + ((u32)reloc->gpu_offset);
 		} else
-			ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+			ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		track->textures[i].robj = reloc->robj;
 		track->tex_dirty = true;
 		break;
@@ -268,7 +268,7 @@ int r200_packet0_check(struct radeon_cs_parser *p,
 			return r;
 		}
 		track->textures[i].cube_info[face - 1].offset = idx_value;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		track->textures[i].cube_info[face - 1].robj = reloc->robj;
 		track->tex_dirty = true;
 		break;
@@ -287,9 +287,9 @@ int r200_packet0_check(struct radeon_cs_parser *p,
 		}
 
 		if (!(p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS)) {
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			if (reloc->tiling_flags & RADEON_TILING_MACRO)
 				tile_flags |= RADEON_COLOR_TILE_ENABLE;
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+			if (reloc->tiling_flags & RADEON_TILING_MICRO)
 				tile_flags |= RADEON_COLOR_MICROTILE_ENABLE;
 
 			tmp = idx_value & ~(0x7 << 16);
@@ -362,7 +362,7 @@ int r200_packet0_check(struct radeon_cs_parser *p,
 			radeon_cs_dump_packet(p, pkt);
 			return r;
 		}
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case RADEON_PP_CNTL:
 		{
diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c
index d8dd269b915..3c21d77a483 100644
--- a/drivers/gpu/drm/radeon/r300.c
+++ b/drivers/gpu/drm/radeon/r300.c
@@ -72,13 +72,11 @@ void rv370_pcie_gart_tlb_flush(struct radeon_device *rdev)
 #define R300_PTE_WRITEABLE (1 << 2)
 #define R300_PTE_READABLE  (1 << 3)
 
-int rv370_pcie_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr)
+void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i,
+			      uint64_t addr)
 {
 	void __iomem *ptr = rdev->gart.ptr;
 
-	if (i < 0 || i > rdev->gart.num_gpu_pages) {
-		return -EINVAL;
-	}
 	addr = (lower_32_bits(addr) >> 8) |
 	       ((upper_32_bits(addr) & 0xff) << 24) |
 	       R300_PTE_WRITEABLE | R300_PTE_READABLE;
@@ -86,7 +84,6 @@ int rv370_pcie_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr)
 	 * on powerpc without HW swappers, it'll get swapped on way
 	 * into VRAM - so no need for cpu_to_le32 on VRAM tables */
 	writel(addr, ((void __iomem *)ptr) + (i * 4));
-	return 0;
 }
 
 int rv370_pcie_gart_init(struct radeon_device *rdev)
@@ -640,7 +637,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 		track->cb[i].robj = reloc->robj;
 		track->cb[i].offset = idx_value;
 		track->cb_dirty = true;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case R300_ZB_DEPTHOFFSET:
 		r = radeon_cs_packet_next_reloc(p, &reloc, 0);
@@ -653,7 +650,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 		track->zb.robj = reloc->robj;
 		track->zb.offset = idx_value;
 		track->zb_dirty = true;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case R300_TX_OFFSET_0:
 	case R300_TX_OFFSET_0+4:
@@ -682,16 +679,16 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 
 		if (p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS) {
 			ib[idx] = (idx_value & 31) | /* keep the 1st 5 bits */
-				  ((idx_value & ~31) + (u32)reloc->lobj.gpu_offset);
+				  ((idx_value & ~31) + (u32)reloc->gpu_offset);
 		} else {
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			if (reloc->tiling_flags & RADEON_TILING_MACRO)
 				tile_flags |= R300_TXO_MACRO_TILE;
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+			if (reloc->tiling_flags & RADEON_TILING_MICRO)
 				tile_flags |= R300_TXO_MICRO_TILE;
-			else if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO_SQUARE)
+			else if (reloc->tiling_flags & RADEON_TILING_MICRO_SQUARE)
 				tile_flags |= R300_TXO_MICRO_TILE_SQUARE;
 
-			tmp = idx_value + ((u32)reloc->lobj.gpu_offset);
+			tmp = idx_value + ((u32)reloc->gpu_offset);
 			tmp |= tile_flags;
 			ib[idx] = tmp;
 		}
@@ -753,11 +750,11 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 				return r;
 			}
 
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			if (reloc->tiling_flags & RADEON_TILING_MACRO)
 				tile_flags |= R300_COLOR_TILE_ENABLE;
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+			if (reloc->tiling_flags & RADEON_TILING_MICRO)
 				tile_flags |= R300_COLOR_MICROTILE_ENABLE;
-			else if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO_SQUARE)
+			else if (reloc->tiling_flags & RADEON_TILING_MICRO_SQUARE)
 				tile_flags |= R300_COLOR_MICROTILE_SQUARE_ENABLE;
 
 			tmp = idx_value & ~(0x7 << 16);
@@ -838,11 +835,11 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 				return r;
 			}
 
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+			if (reloc->tiling_flags & RADEON_TILING_MACRO)
 				tile_flags |= R300_DEPTHMACROTILE_ENABLE;
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+			if (reloc->tiling_flags & RADEON_TILING_MICRO)
 				tile_flags |= R300_DEPTHMICROTILE_TILED;
-			else if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO_SQUARE)
+			else if (reloc->tiling_flags & RADEON_TILING_MICRO_SQUARE)
 				tile_flags |= R300_DEPTHMICROTILE_TILED_SQUARE;
 
 			tmp = idx_value & ~(0x7 << 16);
@@ -1052,7 +1049,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 			radeon_cs_dump_packet(p, pkt);
 			return r;
 		}
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case 0x4e0c:
 		/* RB3D_COLOR_CHANNEL_MASK */
@@ -1097,7 +1094,7 @@ static int r300_packet0_check(struct radeon_cs_parser *p,
 		track->aa.robj = reloc->robj;
 		track->aa.offset = idx_value;
 		track->aa_dirty = true;
-		ib[idx] = idx_value + ((u32)reloc->lobj.gpu_offset);
+		ib[idx] = idx_value + ((u32)reloc->gpu_offset);
 		break;
 	case R300_RB3D_AARESOLVE_PITCH:
 		track->aa.pitch = idx_value & 0x3FFE;
@@ -1162,7 +1159,7 @@ static int r300_packet3_check(struct radeon_cs_parser *p,
 			radeon_cs_dump_packet(p, pkt);
 			return r;
 		}
-		ib[idx+1] = radeon_get_ib_value(p, idx + 1) + ((u32)reloc->lobj.gpu_offset);
+		ib[idx+1] = radeon_get_ib_value(p, idx + 1) + ((u32)reloc->gpu_offset);
 		r = r100_cs_track_check_pkt3_indx_buffer(p, pkt, reloc->robj);
 		if (r) {
 			return r;
@@ -1440,6 +1437,7 @@ int r300_resume(struct radeon_device *rdev)
 
 int r300_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
 	r100_irq_disable(rdev);
@@ -1452,6 +1450,7 @@ int r300_suspend(struct radeon_device *rdev)
 
 void r300_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -1538,6 +1537,9 @@ int r300_init(struct radeon_device *rdev)
 	}
 	r300_set_reg_safe(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = r300_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/r300_cmdbuf.c b/drivers/gpu/drm/radeon/r300_cmdbuf.c
index 60170ea5e3a..84b1d5367a1 100644
--- a/drivers/gpu/drm/radeon/r300_cmdbuf.c
+++ b/drivers/gpu/drm/radeon/r300_cmdbuf.c
@@ -75,7 +75,7 @@ static int r300_emit_cliprects(drm_radeon_private_t *dev_priv,
 		OUT_RING(CP_PACKET0(R300_RE_CLIPRECT_TL_0, nr * 2 - 1));
 
 		for (i = 0; i < nr; ++i) {
-			if (DRM_COPY_FROM_USER
+			if (copy_from_user
 			    (&box, &cmdbuf->boxes[n + i], sizeof(box))) {
 				DRM_ERROR("copy cliprect faulted\n");
 				return -EFAULT;
@@ -928,12 +928,12 @@ static int r300_scratch(drm_radeon_private_t *dev_priv,
 		buf_idx = drm_buffer_pointer_to_dword(cmdbuf->buffer, 0);
 		*buf_idx *= 2; /* 8 bytes per buf */
 
-		if (DRM_COPY_TO_USER(ref_age_base + *buf_idx,
+		if (copy_to_user(ref_age_base + *buf_idx,
 				&dev_priv->scratch_ages[header.scratch.reg],
 				sizeof(u32)))
 			return -EINVAL;
 
-		if (DRM_COPY_FROM_USER(&h_pending,
+		if (copy_from_user(&h_pending,
 				ref_age_base + *buf_idx + 1,
 				sizeof(u32)))
 			return -EINVAL;
@@ -943,7 +943,7 @@ static int r300_scratch(drm_radeon_private_t *dev_priv,
 
 		h_pending--;
 
-		if (DRM_COPY_TO_USER(ref_age_base + *buf_idx + 1,
+		if (copy_to_user(ref_age_base + *buf_idx + 1,
 					&h_pending,
 					sizeof(u32)))
 			return -EINVAL;
diff --git a/drivers/gpu/drm/radeon/r420.c b/drivers/gpu/drm/radeon/r420.c
index 6edf2b3a52b..802b19220a2 100644
--- a/drivers/gpu/drm/radeon/r420.c
+++ b/drivers/gpu/drm/radeon/r420.c
@@ -335,6 +335,7 @@ int r420_resume(struct radeon_device *rdev)
 
 int r420_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r420_cp_errata_fini(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
@@ -348,6 +349,7 @@ int r420_suspend(struct radeon_device *rdev)
 
 void r420_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -444,6 +446,9 @@ int r420_init(struct radeon_device *rdev)
 	}
 	r420_set_reg_safe(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = r420_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/r500_reg.h b/drivers/gpu/drm/radeon/r500_reg.h
index 1dd0d32993d..136b7bc7cd2 100644
--- a/drivers/gpu/drm/radeon/r500_reg.h
+++ b/drivers/gpu/drm/radeon/r500_reg.h
@@ -402,6 +402,7 @@
  * block and vice versa.  This applies to GRPH, CUR, etc.
  */
 #define AVIVO_D1GRPH_LUT_SEL                                    0x6108
+#       define AVIVO_LUT_10BIT_BYPASS_EN                        (1 << 8)
 #define AVIVO_D1GRPH_PRIMARY_SURFACE_ADDRESS                    0x6110
 #define R700_D1GRPH_PRIMARY_SURFACE_ADDRESS_HIGH                0x6914
 #define R700_D2GRPH_PRIMARY_SURFACE_ADDRESS_HIGH                0x6114
diff --git a/drivers/gpu/drm/radeon/r520.c b/drivers/gpu/drm/radeon/r520.c
index e1aece73b37..98d6053c36c 100644
--- a/drivers/gpu/drm/radeon/r520.c
+++ b/drivers/gpu/drm/radeon/r520.c
@@ -312,6 +312,9 @@ int r520_init(struct radeon_device *rdev)
 		return r;
 	rv515_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = r520_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index 9ad06732a78..3c69f58e46e 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -105,6 +105,7 @@ void r600_fini(struct radeon_device *rdev);
 void r600_irq_disable(struct radeon_device *rdev);
 static void r600_pcie_gen2_enable(struct radeon_device *rdev);
 extern int evergreen_rlc_resume(struct radeon_device *rdev);
+extern void rv770_set_clk_bypass_mode(struct radeon_device *rdev);
 
 /**
  * r600_get_xclk - get the xclk
@@ -1644,6 +1645,67 @@ static void r600_gpu_soft_reset(struct radeon_device *rdev, u32 reset_mask)
 	r600_print_gpu_status_regs(rdev);
 }
 
+static void r600_gpu_pci_config_reset(struct radeon_device *rdev)
+{
+	struct rv515_mc_save save;
+	u32 tmp, i;
+
+	dev_info(rdev->dev, "GPU pci config reset\n");
+
+	/* disable dpm? */
+
+	/* Disable CP parsing/prefetching */
+	if (rdev->family >= CHIP_RV770)
+		WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1) | S_0086D8_CP_PFP_HALT(1));
+	else
+		WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1));
+
+	/* disable the RLC */
+	WREG32(RLC_CNTL, 0);
+
+	/* Disable DMA */
+	tmp = RREG32(DMA_RB_CNTL);
+	tmp &= ~DMA_RB_ENABLE;
+	WREG32(DMA_RB_CNTL, tmp);
+
+	mdelay(50);
+
+	/* set mclk/sclk to bypass */
+	if (rdev->family >= CHIP_RV770)
+		rv770_set_clk_bypass_mode(rdev);
+	/* disable BM */
+	pci_clear_master(rdev->pdev);
+	/* disable mem access */
+	rv515_mc_stop(rdev, &save);
+	if (r600_mc_wait_for_idle(rdev)) {
+		dev_warn(rdev->dev, "Wait for MC idle timedout !\n");
+	}
+
+	/* BIF reset workaround.  Not sure if this is needed on 6xx */
+	tmp = RREG32(BUS_CNTL);
+	tmp |= VGA_COHE_SPEC_TIMER_DIS;
+	WREG32(BUS_CNTL, tmp);
+
+	tmp = RREG32(BIF_SCRATCH0);
+
+	/* reset */
+	radeon_pci_config_reset(rdev);
+	mdelay(1);
+
+	/* BIF reset workaround.  Not sure if this is needed on 6xx */
+	tmp = SOFT_RESET_BIF;
+	WREG32(SRBM_SOFT_RESET, tmp);
+	mdelay(1);
+	WREG32(SRBM_SOFT_RESET, 0);
+
+	/* wait for asic to come out of reset */
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CONFIG_MEMSIZE) != 0xffffffff)
+			break;
+		udelay(1);
+	}
+}
+
 int r600_asic_reset(struct radeon_device *rdev)
 {
 	u32 reset_mask;
@@ -1653,10 +1715,17 @@ int r600_asic_reset(struct radeon_device *rdev)
 	if (reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, true);
 
+	/* try soft reset */
 	r600_gpu_soft_reset(rdev, reset_mask);
 
 	reset_mask = r600_gpu_check_soft_reset(rdev);
 
+	/* try pci config reset */
+	if (reset_mask && radeon_hard_reset)
+		r600_gpu_pci_config_reset(rdev);
+
+	reset_mask = r600_gpu_check_soft_reset(rdev);
+
 	if (!reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, false);
 
@@ -1679,11 +1748,9 @@ bool r600_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 	if (!(reset_mask & (RADEON_RESET_GFX |
 			    RADEON_RESET_COMPUTE |
 			    RADEON_RESET_CP))) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force CP activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
@@ -1891,6 +1958,9 @@ static void r600_gpu_init(struct radeon_device *rdev)
 	if (tmp < rdev->config.r600.max_simds) {
 		rdev->config.r600.max_simds = tmp;
 	}
+	tmp = rdev->config.r600.max_simds -
+		r600_count_pipe_bits((cc_gc_shader_pipe_config >> 16) & R6XX_MAX_SIMDS_MASK);
+	rdev->config.r600.active_simds = tmp;
 
 	disabled_rb_mask = (RREG32(CC_RB_BACKEND_DISABLE) >> 16) & R6XX_MAX_BACKENDS_MASK;
 	tmp = (tiling_config & PIPE_TILING__MASK) >> PIPE_TILING__SHIFT;
@@ -2185,7 +2255,8 @@ void r600_pciep_wreg(struct radeon_device *rdev, u32 reg, u32 v)
  */
 void r600_cp_stop(struct radeon_device *rdev)
 {
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 	WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1));
 	WREG32(SCRATCH_UMSK, 0);
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
@@ -2382,6 +2453,36 @@ out:
 	return err;
 }
 
+u32 r600_gfx_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	u32 rptr;
+
+	if (rdev->wb.enabled)
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	else
+		rptr = RREG32(R600_CP_RB_RPTR);
+
+	return rptr;
+}
+
+u32 r600_gfx_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	u32 wptr;
+
+	wptr = RREG32(R600_CP_RB_WPTR);
+
+	return wptr;
+}
+
+void r600_gfx_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring)
+{
+	WREG32(R600_CP_RB_WPTR, ring->wptr);
+	(void)RREG32(R600_CP_RB_WPTR);
+}
+
 static int r600_cp_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
@@ -2504,8 +2605,6 @@ int r600_cp_resume(struct radeon_device *rdev)
 	WREG32(CP_RB_BASE, ring->gpu_addr >> 8);
 	WREG32(CP_DEBUG, (1 << 27) | (1 << 28));
 
-	ring->rptr = RREG32(CP_RB_RPTR);
-
 	r600_cp_start(rdev);
 	ring->ready = true;
 	r = radeon_ring_test(rdev, RADEON_RING_TYPE_GFX_INDEX, ring);
@@ -2513,6 +2612,10 @@ int r600_cp_resume(struct radeon_device *rdev)
 		ring->ready = false;
 		return r;
 	}
+
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+
 	return 0;
 }
 
@@ -2607,30 +2710,31 @@ void r600_fence_ring_emit(struct radeon_device *rdev,
 			  struct radeon_fence *fence)
 {
 	struct radeon_ring *ring = &rdev->ring[fence->ring];
+	u32 cp_coher_cntl = PACKET3_TC_ACTION_ENA | PACKET3_VC_ACTION_ENA |
+		PACKET3_SH_ACTION_ENA;
+
+	if (rdev->family >= CHIP_RV770)
+		cp_coher_cntl |= PACKET3_FULL_CACHE_ENA;
 
 	if (rdev->wb.use_event) {
 		u64 addr = rdev->fence_drv[fence->ring].gpu_addr;
 		/* flush read cache over gart */
 		radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
-		radeon_ring_write(ring, PACKET3_TC_ACTION_ENA |
-					PACKET3_VC_ACTION_ENA |
-					PACKET3_SH_ACTION_ENA);
+		radeon_ring_write(ring, cp_coher_cntl);
 		radeon_ring_write(ring, 0xFFFFFFFF);
 		radeon_ring_write(ring, 0);
 		radeon_ring_write(ring, 10); /* poll interval */
 		/* EVENT_WRITE_EOP - flush caches, send int */
 		radeon_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4));
 		radeon_ring_write(ring, EVENT_TYPE(CACHE_FLUSH_AND_INV_EVENT_TS) | EVENT_INDEX(5));
-		radeon_ring_write(ring, addr & 0xffffffff);
+		radeon_ring_write(ring, lower_32_bits(addr));
 		radeon_ring_write(ring, (upper_32_bits(addr) & 0xff) | DATA_SEL(1) | INT_SEL(2));
 		radeon_ring_write(ring, fence->seq);
 		radeon_ring_write(ring, 0);
 	} else {
 		/* flush read cache over gart */
 		radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
-		radeon_ring_write(ring, PACKET3_TC_ACTION_ENA |
-					PACKET3_VC_ACTION_ENA |
-					PACKET3_SH_ACTION_ENA);
+		radeon_ring_write(ring, cp_coher_cntl);
 		radeon_ring_write(ring, 0xFFFFFFFF);
 		radeon_ring_write(ring, 0);
 		radeon_ring_write(ring, 10); /* poll interval */
@@ -2662,7 +2766,7 @@ bool r600_semaphore_ring_emit(struct radeon_device *rdev,
 		sel |= PACKET3_SEM_WAIT_ON_SIGNAL;
 
 	radeon_ring_write(ring, PACKET3(PACKET3_MEM_SEMAPHORE, 1));
-	radeon_ring_write(ring, addr & 0xffffffff);
+	radeon_ring_write(ring, lower_32_bits(addr));
 	radeon_ring_write(ring, (upper_32_bits(addr) & 0xff) | sel);
 
 	return true;
@@ -2723,9 +2827,9 @@ int r600_copy_cpdma(struct radeon_device *rdev,
 		if (size_in_bytes == 0)
 			tmp |= PACKET3_CP_DMA_CP_SYNC;
 		radeon_ring_write(ring, PACKET3(PACKET3_CP_DMA, 4));
-		radeon_ring_write(ring, src_offset & 0xffffffff);
+		radeon_ring_write(ring, lower_32_bits(src_offset));
 		radeon_ring_write(ring, tmp);
-		radeon_ring_write(ring, dst_offset & 0xffffffff);
+		radeon_ring_write(ring, lower_32_bits(dst_offset));
 		radeon_ring_write(ring, upper_32_bits(dst_offset) & 0xff);
 		radeon_ring_write(ring, cur_size_in_bytes);
 		src_offset += cur_size_in_bytes;
@@ -2738,6 +2842,7 @@ int r600_copy_cpdma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
@@ -2775,14 +2880,6 @@ static int r600_startup(struct radeon_device *rdev)
 
 	r600_mc_program(rdev);
 
-	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
-		r = r600_init_microcode(rdev);
-		if (r) {
-			DRM_ERROR("Failed to load firmware!\n");
-			return r;
-		}
-	}
-
 	if (rdev->flags & RADEON_IS_AGP) {
 		r600_agp_enable(rdev);
 	} else {
@@ -2803,12 +2900,6 @@ static int r600_startup(struct radeon_device *rdev)
 		return r;
 	}
 
-	r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_DMA_INDEX);
-	if (r) {
-		dev_err(rdev->dev, "failed initializing DMA fences (%d).\n", r);
-		return r;
-	}
-
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
 		r = radeon_irq_kms_init(rdev);
@@ -2826,18 +2917,10 @@ static int r600_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     R600_CP_RB_RPTR, R600_CP_RB_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
-	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
-	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR, DMA_RB_WPTR,
-			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
-	if (r)
-		return r;
-
 	r = r600_cp_load_microcode(rdev);
 	if (r)
 		return r;
@@ -2845,10 +2928,6 @@ static int r600_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
-	r = r600_dma_resume(rdev);
-	if (r)
-		return r;
-
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
 		dev_err(rdev->dev, "IB initialization failed (%d).\n", r);
@@ -2889,6 +2968,9 @@ int r600_resume(struct radeon_device *rdev)
 	/* post card */
 	atom_asic_init(rdev->mode_info.atom_context);
 
+	if (rdev->pm.pm_method == PM_METHOD_DPM)
+		radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = r600_startup(rdev);
 	if (r) {
@@ -2902,9 +2984,9 @@ int r600_resume(struct radeon_device *rdev)
 
 int r600_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	r600_cp_stop(rdev);
-	r600_dma_stop(rdev);
 	r600_irq_suspend(rdev);
 	radeon_wb_disable(rdev);
 	r600_pcie_gart_disable(rdev);
@@ -2970,12 +3052,20 @@ int r600_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
+		r = r600_init_microcode(rdev);
+		if (r) {
+			DRM_ERROR("Failed to load firmware!\n");
+			return r;
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024);
 
-	rdev->ring[R600_RING_TYPE_DMA_INDEX].ring_obj = NULL;
-	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_DMA_INDEX], 64 * 1024);
-
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
 
@@ -2988,7 +3078,6 @@ int r600_init(struct radeon_device *rdev)
 	if (r) {
 		dev_err(rdev->dev, "disabling GPU acceleration\n");
 		r600_cp_fini(rdev);
-		r600_dma_fini(rdev);
 		r600_irq_fini(rdev);
 		radeon_wb_fini(rdev);
 		radeon_ib_pool_fini(rdev);
@@ -3002,9 +3091,9 @@ int r600_init(struct radeon_device *rdev)
 
 void r600_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r600_audio_fini(rdev);
 	r600_cp_fini(rdev);
-	r600_dma_fini(rdev);
 	r600_irq_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -3420,7 +3509,6 @@ int r600_irq_set(struct radeon_device *rdev)
 	u32 hpd1, hpd2, hpd3, hpd4 = 0, hpd5 = 0, hpd6 = 0;
 	u32 grbm_int_cntl = 0;
 	u32 hdmi0, hdmi1;
-	u32 d1grph = 0, d2grph = 0;
 	u32 dma_cntl;
 	u32 thermal_int = 0;
 
@@ -3529,8 +3617,8 @@ int r600_irq_set(struct radeon_device *rdev)
 	WREG32(CP_INT_CNTL, cp_int_cntl);
 	WREG32(DMA_CNTL, dma_cntl);
 	WREG32(DxMODE_INT_MASK, mode_int);
-	WREG32(D1GRPH_INTERRUPT_CONTROL, d1grph);
-	WREG32(D2GRPH_INTERRUPT_CONTROL, d2grph);
+	WREG32(D1GRPH_INTERRUPT_CONTROL, DxGRPH_PFLIP_INT_MASK);
+	WREG32(D2GRPH_INTERRUPT_CONTROL, DxGRPH_PFLIP_INT_MASK);
 	WREG32(GRBM_INT_CNTL, grbm_int_cntl);
 	if (ASIC_IS_DCE3(rdev)) {
 		WREG32(DC_HPD1_INT_CONTROL, hpd1);
@@ -3707,6 +3795,7 @@ static u32 r600_get_ih_wptr(struct radeon_device *rdev)
 		tmp = RREG32(IH_RB_CNTL);
 		tmp |= IH_WPTR_OVERFLOW_CLEAR;
 		WREG32(IH_RB_CNTL, tmp);
+		wptr &= ~RB_OVERFLOW;
 	}
 	return (wptr & rdev->ih.ptr_mask);
 }
@@ -3791,7 +3880,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[0]))
-						radeon_crtc_handle_flip(rdev, 0);
+						radeon_crtc_handle_vblank(rdev, 0);
 					rdev->irq.stat_regs.r600.disp_int &= ~LB_D1_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D1 vblank\n");
 				}
@@ -3817,7 +3906,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[1]))
-						radeon_crtc_handle_flip(rdev, 1);
+						radeon_crtc_handle_vblank(rdev, 1);
 					rdev->irq.stat_regs.r600.disp_int &= ~LB_D2_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D2 vblank\n");
 				}
@@ -3833,6 +3922,14 @@ restart_ih:
 				break;
 			}
 			break;
+		case 9: /* D1 pflip */
+			DRM_DEBUG("IH: D1 flip\n");
+			radeon_crtc_handle_flip(rdev, 0);
+			break;
+		case 11: /* D2 pflip */
+			DRM_DEBUG("IH: D2 flip\n");
+			radeon_crtc_handle_flip(rdev, 1);
+			break;
 		case 19: /* HPD/DAC hotplug */
 			switch (src_data) {
 			case 0:
@@ -3903,6 +4000,10 @@ restart_ih:
 				break;
 			}
 			break;
+		case 124: /* UVD */
+			DRM_DEBUG("IH: UVD int: 0x%08x\n", src_data);
+			radeon_fence_process(rdev, R600_RING_TYPE_UVD_INDEX);
+			break;
 		case 176: /* CP_INT in ring buffer */
 		case 177: /* CP_INT in IB1 */
 		case 178: /* CP_INT in IB2 */
diff --git a/drivers/gpu/drm/radeon/r600_audio.c b/drivers/gpu/drm/radeon/r600_audio.c
index 47fc2b88697..bffac10c429 100644
--- a/drivers/gpu/drm/radeon/r600_audio.c
+++ b/drivers/gpu/drm/radeon/r600_audio.c
@@ -142,12 +142,15 @@ void r600_audio_update_hdmi(struct work_struct *work)
 }
 
 /* enable the audio stream */
-static void r600_audio_enable(struct radeon_device *rdev,
-			      struct r600_audio_pin *pin,
-			      bool enable)
+void r600_audio_enable(struct radeon_device *rdev,
+		       struct r600_audio_pin *pin,
+		       bool enable)
 {
 	u32 value = 0;
 
+	if (!pin)
+		return;
+
 	if (ASIC_IS_DCE4(rdev)) {
 		if (enable) {
 			value |= 0x81000000; /* Required to enable audio */
@@ -158,7 +161,6 @@ static void r600_audio_enable(struct radeon_device *rdev,
 		WREG32_P(R600_AUDIO_ENABLE,
 			 enable ? 0x81000000 : 0x0, ~0x81000000);
 	}
-	DRM_INFO("%s audio %d support\n", enable ? "Enabling" : "Disabling", pin->id);
 }
 
 /*
@@ -178,8 +180,8 @@ int r600_audio_init(struct radeon_device *rdev)
 	rdev->audio.pin[0].status_bits = 0;
 	rdev->audio.pin[0].category_code = 0;
 	rdev->audio.pin[0].id = 0;
-
-	r600_audio_enable(rdev, &rdev->audio.pin[0], true);
+	/* disable audio.  it will be set up later */
+	r600_audio_enable(rdev, &rdev->audio.pin[0], false);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/radeon/r600_cp.c b/drivers/gpu/drm/radeon/r600_cp.c
index d8eb48bff0e..8c9b7e26533 100644
--- a/drivers/gpu/drm/radeon/r600_cp.c
+++ b/drivers/gpu/drm/radeon/r600_cp.c
@@ -2515,7 +2515,7 @@ int r600_cp_dispatch_texture(struct drm_device *dev,
 		buf = radeon_freelist_get(dev);
 		if (!buf) {
 			DRM_DEBUG("EAGAIN\n");
-			if (DRM_COPY_TO_USER(tex->image, image, sizeof(*image)))
+			if (copy_to_user(tex->image, image, sizeof(*image)))
 				return -EFAULT;
 			return -EAGAIN;
 		}
@@ -2528,7 +2528,7 @@ int r600_cp_dispatch_texture(struct drm_device *dev,
 		buffer =
 		    (u32 *) ((char *)dev->agp_buffer_map->handle + buf->offset);
 
-		if (DRM_COPY_FROM_USER(buffer, data, pass_size)) {
+		if (copy_from_user(buffer, data, pass_size)) {
 			DRM_ERROR("EFAULT on pad, %d bytes\n", pass_size);
 			return -EFAULT;
 		}
diff --git a/drivers/gpu/drm/radeon/r600_cs.c b/drivers/gpu/drm/radeon/r600_cs.c
index 5dceea6f71a..12511bb5fd6 100644
--- a/drivers/gpu/drm/radeon/r600_cs.c
+++ b/drivers/gpu/drm/radeon/r600_cs.c
@@ -749,7 +749,10 @@ static int r600_cs_track_check(struct radeon_cs_parser *p)
 		}
 
 		for (i = 0; i < 8; i++) {
-			if ((tmp >> (i * 4)) & 0xF) {
+			u32 format = G_0280A0_FORMAT(track->cb_color_info[i]);
+
+			if (format != V_0280A0_COLOR_INVALID &&
+			    (tmp >> (i * 4)) & 0xF) {
 				/* at least one component is enabled */
 				if (track->cb_color_bo[i] == NULL) {
 					dev_warn(p->dev, "%s:%d mask 0x%08X | 0x%08X no cb for %d\n",
@@ -1004,8 +1007,22 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 	case R_008C64_SQ_VSTMP_RING_SIZE:
 	case R_0288C8_SQ_GS_VERT_ITEMSIZE:
 		/* get value to populate the IB don't remove */
-		tmp =radeon_get_ib_value(p, idx);
-		ib[idx] = 0;
+		/*tmp =radeon_get_ib_value(p, idx);
+		  ib[idx] = 0;*/
+		break;
+	case SQ_ESGS_RING_BASE:
+	case SQ_GSVS_RING_BASE:
+	case SQ_ESTMP_RING_BASE:
+	case SQ_GSTMP_RING_BASE:
+	case SQ_PSTMP_RING_BASE:
+	case SQ_VSTMP_RING_BASE:
+		r = radeon_cs_packet_next_reloc(p, &reloc, 0);
+		if (r) {
+			dev_warn(p->dev, "bad SET_CONTEXT_REG "
+					"0x%04X\n", reg);
+			return -EINVAL;
+		}
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		break;
 	case SQ_CONFIG:
 		track->sq_config = radeon_get_ib_value(p, idx);
@@ -1026,7 +1043,7 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			track->db_depth_info = radeon_get_ib_value(p, idx);
 			ib[idx] &= C_028010_ARRAY_MODE;
 			track->db_depth_info &= C_028010_ARRAY_MODE;
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) {
+			if (reloc->tiling_flags & RADEON_TILING_MACRO) {
 				ib[idx] |= S_028010_ARRAY_MODE(V_028010_ARRAY_2D_TILED_THIN1);
 				track->db_depth_info |= S_028010_ARRAY_MODE(V_028010_ARRAY_2D_TILED_THIN1);
 			} else {
@@ -1067,9 +1084,9 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 		}
 		tmp = (reg - VGT_STRMOUT_BUFFER_BASE_0) / 16;
 		track->vgt_strmout_bo_offset[tmp] = radeon_get_ib_value(p, idx) << 8;
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->vgt_strmout_bo[tmp] = reloc->robj;
-		track->vgt_strmout_bo_mc[tmp] = reloc->lobj.gpu_offset;
+		track->vgt_strmout_bo_mc[tmp] = reloc->gpu_offset;
 		track->streamout_dirty = true;
 		break;
 	case VGT_STRMOUT_BUFFER_SIZE_0:
@@ -1088,7 +1105,7 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 					"0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		break;
 	case R_028238_CB_TARGET_MASK:
 		track->cb_target_mask = radeon_get_ib_value(p, idx);
@@ -1125,10 +1142,10 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			}
 			tmp = (reg - R_0280A0_CB_COLOR0_INFO) / 4;
 			track->cb_color_info[tmp] = radeon_get_ib_value(p, idx);
-			if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO) {
+			if (reloc->tiling_flags & RADEON_TILING_MACRO) {
 				ib[idx] |= S_0280A0_ARRAY_MODE(V_0280A0_ARRAY_2D_TILED_THIN1);
 				track->cb_color_info[tmp] |= S_0280A0_ARRAY_MODE(V_0280A0_ARRAY_2D_TILED_THIN1);
-			} else if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO) {
+			} else if (reloc->tiling_flags & RADEON_TILING_MICRO) {
 				ib[idx] |= S_0280A0_ARRAY_MODE(V_0280A0_ARRAY_1D_TILED_THIN1);
 				track->cb_color_info[tmp] |= S_0280A0_ARRAY_MODE(V_0280A0_ARRAY_1D_TILED_THIN1);
 			}
@@ -1197,7 +1214,7 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			}
 			track->cb_color_frag_bo[tmp] = reloc->robj;
 			track->cb_color_frag_offset[tmp] = (u64)ib[idx] << 8;
-			ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+			ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		}
 		if (G_0280A0_TILE_MODE(track->cb_color_info[tmp])) {
 			track->cb_dirty = true;
@@ -1228,7 +1245,7 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			}
 			track->cb_color_tile_bo[tmp] = reloc->robj;
 			track->cb_color_tile_offset[tmp] = (u64)ib[idx] << 8;
-			ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+			ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		}
 		if (G_0280A0_TILE_MODE(track->cb_color_info[tmp])) {
 			track->cb_dirty = true;
@@ -1264,10 +1281,10 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 		}
 		tmp = (reg - CB_COLOR0_BASE) / 4;
 		track->cb_color_bo_offset[tmp] = radeon_get_ib_value(p, idx) << 8;
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->cb_color_base_last[tmp] = ib[idx];
 		track->cb_color_bo[tmp] = reloc->robj;
-		track->cb_color_bo_mc[tmp] = reloc->lobj.gpu_offset;
+		track->cb_color_bo_mc[tmp] = reloc->gpu_offset;
 		track->cb_dirty = true;
 		break;
 	case DB_DEPTH_BASE:
@@ -1278,9 +1295,9 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		track->db_offset = radeon_get_ib_value(p, idx) << 8;
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->db_bo = reloc->robj;
-		track->db_bo_mc = reloc->lobj.gpu_offset;
+		track->db_bo_mc = reloc->gpu_offset;
 		track->db_dirty = true;
 		break;
 	case DB_HTILE_DATA_BASE:
@@ -1291,7 +1308,7 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 			return -EINVAL;
 		}
 		track->htile_offset = radeon_get_ib_value(p, idx) << 8;
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		track->htile_bo = reloc->robj;
 		track->db_dirty = true;
 		break;
@@ -1360,7 +1377,7 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 					"0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		break;
 	case SX_MEMORY_EXPORT_BASE:
 		r = radeon_cs_packet_next_reloc(p, &reloc, r600_nomm);
@@ -1369,7 +1386,7 @@ static int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx)
 					"0x%04X\n", reg);
 			return -EINVAL;
 		}
-		ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+		ib[idx] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		break;
 	case SX_MISC:
 		track->sx_misc_kill_all_prims = (radeon_get_ib_value(p, idx) & 0x1) != 0;
@@ -1655,7 +1672,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 			return -EINVAL;
 		}
 
-		offset = reloc->lobj.gpu_offset +
+		offset = reloc->gpu_offset +
 		         (idx_value & 0xfffffff0) +
 		         ((u64)(tmp & 0xff) << 32);
 
@@ -1696,7 +1713,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 			return -EINVAL;
 		}
 
-		offset = reloc->lobj.gpu_offset +
+		offset = reloc->gpu_offset +
 		         idx_value +
 		         ((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
 
@@ -1748,7 +1765,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 				return -EINVAL;
 			}
 
-			offset = reloc->lobj.gpu_offset +
+			offset = reloc->gpu_offset +
 			         (radeon_get_ib_value(p, idx+1) & 0xfffffff0) +
 			         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
@@ -1788,7 +1805,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 			tmp = radeon_get_ib_value(p, idx) +
 				((u64)(radeon_get_ib_value(p, idx+1) & 0xff) << 32);
 
-			offset = reloc->lobj.gpu_offset + tmp;
+			offset = reloc->gpu_offset + tmp;
 
 			if ((tmp + size) > radeon_bo_size(reloc->robj)) {
 				dev_warn(p->dev, "CP DMA src buffer too small (%llu %lu)\n",
@@ -1818,7 +1835,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 			tmp = radeon_get_ib_value(p, idx+2) +
 				((u64)(radeon_get_ib_value(p, idx+3) & 0xff) << 32);
 
-			offset = reloc->lobj.gpu_offset + tmp;
+			offset = reloc->gpu_offset + tmp;
 
 			if ((tmp + size) > radeon_bo_size(reloc->robj)) {
 				dev_warn(p->dev, "CP DMA dst buffer too small (%llu %lu)\n",
@@ -1844,7 +1861,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 				DRM_ERROR("bad SURFACE_SYNC\n");
 				return -EINVAL;
 			}
-			ib[idx+2] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+			ib[idx+2] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		}
 		break;
 	case PACKET3_EVENT_WRITE:
@@ -1860,7 +1877,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 				DRM_ERROR("bad EVENT_WRITE\n");
 				return -EINVAL;
 			}
-			offset = reloc->lobj.gpu_offset +
+			offset = reloc->gpu_offset +
 			         (radeon_get_ib_value(p, idx+1) & 0xfffffff8) +
 			         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
@@ -1882,7 +1899,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 			return -EINVAL;
 		}
 
-		offset = reloc->lobj.gpu_offset +
+		offset = reloc->gpu_offset +
 		         (radeon_get_ib_value(p, idx+1) & 0xfffffffc) +
 		         ((u64)(radeon_get_ib_value(p, idx+2) & 0xff) << 32);
 
@@ -1947,11 +1964,11 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 					DRM_ERROR("bad SET_RESOURCE\n");
 					return -EINVAL;
 				}
-				base_offset = (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+				base_offset = (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 				if (!(p->cs_flags & RADEON_CS_KEEP_TILING_FLAGS)) {
-					if (reloc->lobj.tiling_flags & RADEON_TILING_MACRO)
+					if (reloc->tiling_flags & RADEON_TILING_MACRO)
 						ib[idx+1+(i*7)+0] |= S_038000_TILE_MODE(V_038000_ARRAY_2D_TILED_THIN1);
-					else if (reloc->lobj.tiling_flags & RADEON_TILING_MICRO)
+					else if (reloc->tiling_flags & RADEON_TILING_MICRO)
 						ib[idx+1+(i*7)+0] |= S_038000_TILE_MODE(V_038000_ARRAY_1D_TILED_THIN1);
 				}
 				texture = reloc->robj;
@@ -1961,13 +1978,13 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 					DRM_ERROR("bad SET_RESOURCE\n");
 					return -EINVAL;
 				}
-				mip_offset = (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+				mip_offset = (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 				mipmap = reloc->robj;
 				r = r600_check_texture_resource(p,  idx+(i*7)+1,
 								texture, mipmap,
 								base_offset + radeon_get_ib_value(p, idx+1+(i*7)+2),
 								mip_offset + radeon_get_ib_value(p, idx+1+(i*7)+3),
-								reloc->lobj.tiling_flags);
+								reloc->tiling_flags);
 				if (r)
 					return r;
 				ib[idx+1+(i*7)+2] += base_offset;
@@ -1991,7 +2008,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 					ib[idx+1+(i*7)+1] = radeon_bo_size(reloc->robj) - offset;
 				}
 
-				offset64 = reloc->lobj.gpu_offset + offset;
+				offset64 = reloc->gpu_offset + offset;
 				ib[idx+1+(i*8)+0] = offset64;
 				ib[idx+1+(i*8)+2] = (ib[idx+1+(i*8)+2] & 0xffffff00) |
 						    (upper_32_bits(offset64) & 0xff);
@@ -2101,7 +2118,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 					  offset + 4, radeon_bo_size(reloc->robj));
 				return -EINVAL;
 			}
-			ib[idx+1] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
+			ib[idx+1] += (u32)((reloc->gpu_offset >> 8) & 0xffffffff);
 		}
 		break;
 	case PACKET3_SURFACE_BASE_UPDATE:
@@ -2134,7 +2151,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 					  offset + 4, radeon_bo_size(reloc->robj));
 				return -EINVAL;
 			}
-			offset += reloc->lobj.gpu_offset;
+			offset += reloc->gpu_offset;
 			ib[idx+1] = offset;
 			ib[idx+2] = upper_32_bits(offset) & 0xff;
 		}
@@ -2153,7 +2170,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 					  offset + 4, radeon_bo_size(reloc->robj));
 				return -EINVAL;
 			}
-			offset += reloc->lobj.gpu_offset;
+			offset += reloc->gpu_offset;
 			ib[idx+3] = offset;
 			ib[idx+4] = upper_32_bits(offset) & 0xff;
 		}
@@ -2182,7 +2199,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 				  offset + 8, radeon_bo_size(reloc->robj));
 			return -EINVAL;
 		}
-		offset += reloc->lobj.gpu_offset;
+		offset += reloc->gpu_offset;
 		ib[idx+0] = offset;
 		ib[idx+1] = upper_32_bits(offset) & 0xff;
 		break;
@@ -2207,7 +2224,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 					  offset + 4, radeon_bo_size(reloc->robj));
 				return -EINVAL;
 			}
-			offset += reloc->lobj.gpu_offset;
+			offset += reloc->gpu_offset;
 			ib[idx+1] = offset;
 			ib[idx+2] = upper_32_bits(offset) & 0xff;
 		} else {
@@ -2231,7 +2248,7 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
 					  offset + 4, radeon_bo_size(reloc->robj));
 				return -EINVAL;
 			}
-			offset += reloc->lobj.gpu_offset;
+			offset += reloc->gpu_offset;
 			ib[idx+3] = offset;
 			ib[idx+4] = upper_32_bits(offset) & 0xff;
 		} else {
@@ -2386,7 +2403,7 @@ int r600_cs_legacy(struct drm_device *dev, void *data, struct drm_file *filp,
 	ib_chunk = &parser.chunks[parser.chunk_ib_idx];
 	parser.ib.length_dw = ib_chunk->length_dw;
 	*l = parser.ib.length_dw;
-	if (DRM_COPY_FROM_USER(ib, ib_chunk->user_ptr, ib_chunk->length_dw * 4)) {
+	if (copy_from_user(ib, ib_chunk->user_ptr, ib_chunk->length_dw * 4)) {
 		r = -EFAULT;
 		r600_cs_parser_fini(&parser, r);
 		return r;
@@ -2488,14 +2505,14 @@ int r600_dma_cs_parse(struct radeon_cs_parser *p)
 				dst_offset = radeon_get_ib_value(p, idx+1);
 				dst_offset <<= 8;
 
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset >> 8);
 				p->idx += count + 5;
 			} else {
 				dst_offset = radeon_get_ib_value(p, idx+1);
 				dst_offset |= ((u64)(radeon_get_ib_value(p, idx+2) & 0xff)) << 32;
 
-				ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-				ib[idx+2] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
+				ib[idx+1] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+				ib[idx+2] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
 				p->idx += count + 3;
 			}
 			if ((dst_offset + (count * 4)) > radeon_bo_size(dst_reloc->robj)) {
@@ -2522,22 +2539,22 @@ int r600_dma_cs_parse(struct radeon_cs_parser *p)
 					/* tiled src, linear dst */
 					src_offset = radeon_get_ib_value(p, idx+1);
 					src_offset <<= 8;
-					ib[idx+1] += (u32)(src_reloc->lobj.gpu_offset >> 8);
+					ib[idx+1] += (u32)(src_reloc->gpu_offset >> 8);
 
 					dst_offset = radeon_get_ib_value(p, idx+5);
 					dst_offset |= ((u64)(radeon_get_ib_value(p, idx+6) & 0xff)) << 32;
-					ib[idx+5] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+6] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
+					ib[idx+5] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+6] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
 				} else {
 					/* linear src, tiled dst */
 					src_offset = radeon_get_ib_value(p, idx+5);
 					src_offset |= ((u64)(radeon_get_ib_value(p, idx+6) & 0xff)) << 32;
-					ib[idx+5] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+6] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+					ib[idx+5] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+6] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 
 					dst_offset = radeon_get_ib_value(p, idx+1);
 					dst_offset <<= 8;
-					ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset >> 8);
+					ib[idx+1] += (u32)(dst_reloc->gpu_offset >> 8);
 				}
 				p->idx += 7;
 			} else {
@@ -2547,10 +2564,10 @@ int r600_dma_cs_parse(struct radeon_cs_parser *p)
 					dst_offset = radeon_get_ib_value(p, idx+1);
 					dst_offset |= ((u64)(radeon_get_ib_value(p, idx+3) & 0xff)) << 32;
 
-					ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+2] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+3] += upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff;
-					ib[idx+4] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
+					ib[idx+1] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+2] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+3] += upper_32_bits(dst_reloc->gpu_offset) & 0xff;
+					ib[idx+4] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
 					p->idx += 5;
 				} else {
 					src_offset = radeon_get_ib_value(p, idx+2);
@@ -2558,10 +2575,10 @@ int r600_dma_cs_parse(struct radeon_cs_parser *p)
 					dst_offset = radeon_get_ib_value(p, idx+1);
 					dst_offset |= ((u64)(radeon_get_ib_value(p, idx+3) & 0xff0000)) << 16;
 
-					ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+2] += (u32)(src_reloc->lobj.gpu_offset & 0xfffffffc);
-					ib[idx+3] += upper_32_bits(src_reloc->lobj.gpu_offset) & 0xff;
-					ib[idx+3] += (upper_32_bits(dst_reloc->lobj.gpu_offset) & 0xff) << 16;
+					ib[idx+1] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+2] += (u32)(src_reloc->gpu_offset & 0xfffffffc);
+					ib[idx+3] += upper_32_bits(src_reloc->gpu_offset) & 0xff;
+					ib[idx+3] += (upper_32_bits(dst_reloc->gpu_offset) & 0xff) << 16;
 					p->idx += 4;
 				}
 			}
@@ -2593,8 +2610,8 @@ int r600_dma_cs_parse(struct radeon_cs_parser *p)
 					 dst_offset + (count * 4), radeon_bo_size(dst_reloc->robj));
 				return -EINVAL;
 			}
-			ib[idx+1] += (u32)(dst_reloc->lobj.gpu_offset & 0xfffffffc);
-			ib[idx+3] += (upper_32_bits(dst_reloc->lobj.gpu_offset) << 16) & 0x00ff0000;
+			ib[idx+1] += (u32)(dst_reloc->gpu_offset & 0xfffffffc);
+			ib[idx+3] += (upper_32_bits(dst_reloc->gpu_offset) << 16) & 0x00ff0000;
 			p->idx += 4;
 			break;
 		case DMA_PACKET_NOP:
diff --git a/drivers/gpu/drm/radeon/r600_dma.c b/drivers/gpu/drm/radeon/r600_dma.c
index 7844d15c139..4969cef44a1 100644
--- a/drivers/gpu/drm/radeon/r600_dma.c
+++ b/drivers/gpu/drm/radeon/r600_dma.c
@@ -51,7 +51,14 @@ u32 r600_gpu_check_soft_reset(struct radeon_device *rdev);
 uint32_t r600_dma_get_rptr(struct radeon_device *rdev,
 			   struct radeon_ring *ring)
 {
-	return (radeon_ring_generic_get_rptr(rdev, ring) & 0x3fffc) >> 2;
+	u32 rptr;
+
+	if (rdev->wb.enabled)
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	else
+		rptr = RREG32(DMA_RB_RPTR);
+
+	return (rptr & 0x3fffc) >> 2;
 }
 
 /**
@@ -65,7 +72,7 @@ uint32_t r600_dma_get_rptr(struct radeon_device *rdev,
 uint32_t r600_dma_get_wptr(struct radeon_device *rdev,
 			   struct radeon_ring *ring)
 {
-	return (RREG32(ring->wptr_reg) & 0x3fffc) >> 2;
+	return (RREG32(DMA_RB_WPTR) & 0x3fffc) >> 2;
 }
 
 /**
@@ -79,7 +86,7 @@ uint32_t r600_dma_get_wptr(struct radeon_device *rdev,
 void r600_dma_set_wptr(struct radeon_device *rdev,
 		       struct radeon_ring *ring)
 {
-	WREG32(ring->wptr_reg, (ring->wptr << 2) & 0x3fffc);
+	WREG32(DMA_RB_WPTR, (ring->wptr << 2) & 0x3fffc);
 }
 
 /**
@@ -93,7 +100,8 @@ void r600_dma_stop(struct radeon_device *rdev)
 {
 	u32 rb_cntl = RREG32(DMA_RB_CNTL);
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if (rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 
 	rb_cntl &= ~DMA_RB_ENABLE;
 	WREG32(DMA_RB_CNTL, rb_cntl);
@@ -168,8 +176,6 @@ int r600_dma_resume(struct radeon_device *rdev)
 	ring->wptr = 0;
 	WREG32(DMA_RB_WPTR, ring->wptr << 2);
 
-	ring->rptr = RREG32(DMA_RB_RPTR) >> 2;
-
 	WREG32(DMA_RB_CNTL, rb_cntl | DMA_RB_ENABLE);
 
 	ring->ready = true;
@@ -180,7 +186,8 @@ int r600_dma_resume(struct radeon_device *rdev)
 		return r;
 	}
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+	if (rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
 
 	return 0;
 }
@@ -212,11 +219,9 @@ bool r600_dma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 	u32 reset_mask = r600_gpu_check_soft_reset(rdev);
 
 	if (!(reset_mask & RADEON_RESET_DMA)) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force ring activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
@@ -484,6 +489,7 @@ int r600_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
diff --git a/drivers/gpu/drm/radeon/r600_dpm.c b/drivers/gpu/drm/radeon/r600_dpm.c
index 5513d8f0625..9c61b74ef44 100644
--- a/drivers/gpu/drm/radeon/r600_dpm.c
+++ b/drivers/gpu/drm/radeon/r600_dpm.c
@@ -158,16 +158,18 @@ u32 r600_dpm_get_vblank_time(struct radeon_device *rdev)
 	u32 line_time_us, vblank_lines;
 	u32 vblank_time_us = 0xffffffff; /* if the displays are off, vblank time is max */
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
-		radeon_crtc = to_radeon_crtc(crtc);
-		if (crtc->enabled && radeon_crtc->enabled && radeon_crtc->hw_mode.clock) {
-			line_time_us = (radeon_crtc->hw_mode.crtc_htotal * 1000) /
-				radeon_crtc->hw_mode.clock;
-			vblank_lines = radeon_crtc->hw_mode.crtc_vblank_end -
-				radeon_crtc->hw_mode.crtc_vdisplay +
-				(radeon_crtc->v_border * 2);
-			vblank_time_us = vblank_lines * line_time_us;
-			break;
+	if (rdev->num_crtc && rdev->mode_info.mode_config_initialized) {
+		list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+			radeon_crtc = to_radeon_crtc(crtc);
+			if (crtc->enabled && radeon_crtc->enabled && radeon_crtc->hw_mode.clock) {
+				line_time_us = (radeon_crtc->hw_mode.crtc_htotal * 1000) /
+					radeon_crtc->hw_mode.clock;
+				vblank_lines = radeon_crtc->hw_mode.crtc_vblank_end -
+					radeon_crtc->hw_mode.crtc_vdisplay +
+					(radeon_crtc->v_border * 2);
+				vblank_time_us = vblank_lines * line_time_us;
+				break;
+			}
 		}
 	}
 
@@ -181,14 +183,15 @@ u32 r600_dpm_get_vrefresh(struct radeon_device *rdev)
 	struct radeon_crtc *radeon_crtc;
 	u32 vrefresh = 0;
 
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
-		radeon_crtc = to_radeon_crtc(crtc);
-		if (crtc->enabled && radeon_crtc->enabled && radeon_crtc->hw_mode.clock) {
-			vrefresh = radeon_crtc->hw_mode.vrefresh;
-			break;
+	if (rdev->num_crtc && rdev->mode_info.mode_config_initialized) {
+		list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+			radeon_crtc = to_radeon_crtc(crtc);
+			if (crtc->enabled && radeon_crtc->enabled && radeon_crtc->hw_mode.clock) {
+				vrefresh = radeon_crtc->hw_mode.vrefresh;
+				break;
+			}
 		}
 	}
-
 	return vrefresh;
 }
 
@@ -729,8 +732,8 @@ bool r600_is_uvd_state(u32 class, u32 class2)
 	return false;
 }
 
-int r600_set_thermal_temperature_range(struct radeon_device *rdev,
-				       int min_temp, int max_temp)
+static int r600_set_thermal_temperature_range(struct radeon_device *rdev,
+					      int min_temp, int max_temp)
 {
 	int low_temp = 0 * 1000;
 	int high_temp = 255 * 1000;
@@ -777,6 +780,22 @@ bool r600_is_internal_thermal_sensor(enum radeon_int_thermal_type sensor)
 	}
 }
 
+int r600_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
+	if (rdev->irq.installed &&
+	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
+		ret = r600_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
+		if (ret)
+			return ret;
+		rdev->irq.dpm_thermal = true;
+		radeon_irq_set(rdev);
+	}
+
+	return 0;
+}
+
 union power_info {
 	struct _ATOM_POWERPLAY_INFO info;
 	struct _ATOM_POWERPLAY_INFO_V2 info_2;
@@ -818,6 +837,26 @@ static int r600_parse_clk_voltage_dep_table(struct radeon_clock_voltage_dependen
 	return 0;
 }
 
+int r600_get_platform_caps(struct radeon_device *rdev)
+{
+	struct radeon_mode_info *mode_info = &rdev->mode_info;
+	union power_info *power_info;
+	int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
+        u16 data_offset;
+	u8 frev, crev;
+
+	if (!atom_parse_data_header(mode_info->atom_context, index, NULL,
+				   &frev, &crev, &data_offset))
+		return -EINVAL;
+	power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
+
+	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
+	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
+	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
+
+	return 0;
+}
+
 /* sizeof(ATOM_PPLIB_EXTENDEDHEADER) */
 #define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V2 12
 #define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V3 14
@@ -1027,7 +1066,15 @@ int r600_parse_extended_power_table(struct radeon_device *rdev)
 				(mode_info->atom_context->bios + data_offset +
 				 le16_to_cpu(ext_hdr->usVCETableOffset) + 1 +
 				 1 + array->ucNumEntries * sizeof(VCEClockInfo));
+			ATOM_PPLIB_VCE_State_Table *states =
+				(ATOM_PPLIB_VCE_State_Table *)
+				(mode_info->atom_context->bios + data_offset +
+				 le16_to_cpu(ext_hdr->usVCETableOffset) + 1 +
+				 1 + (array->ucNumEntries * sizeof (VCEClockInfo)) +
+				 1 + (limits->numEntries * sizeof(ATOM_PPLIB_VCE_Clock_Voltage_Limit_Record)));
 			ATOM_PPLIB_VCE_Clock_Voltage_Limit_Record *entry;
+			ATOM_PPLIB_VCE_State_Record *state_entry;
+			VCEClockInfo *vce_clk;
 			u32 size = limits->numEntries *
 				sizeof(struct radeon_vce_clock_voltage_dependency_entry);
 			rdev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table.entries =
@@ -1039,8 +1086,9 @@ int r600_parse_extended_power_table(struct radeon_device *rdev)
 			rdev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table.count =
 				limits->numEntries;
 			entry = &limits->entries[0];
+			state_entry = &states->entries[0];
 			for (i = 0; i < limits->numEntries; i++) {
-				VCEClockInfo *vce_clk = (VCEClockInfo *)
+				vce_clk = (VCEClockInfo *)
 					((u8 *)&array->entries[0] +
 					 (entry->ucVCEClockInfoIndex * sizeof(VCEClockInfo)));
 				rdev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table.entries[i].evclk =
@@ -1052,6 +1100,23 @@ int r600_parse_extended_power_table(struct radeon_device *rdev)
 				entry = (ATOM_PPLIB_VCE_Clock_Voltage_Limit_Record *)
 					((u8 *)entry + sizeof(ATOM_PPLIB_VCE_Clock_Voltage_Limit_Record));
 			}
+			for (i = 0; i < states->numEntries; i++) {
+				if (i >= RADEON_MAX_VCE_LEVELS)
+					break;
+				vce_clk = (VCEClockInfo *)
+					((u8 *)&array->entries[0] +
+					 (state_entry->ucVCEClockInfoIndex * sizeof(VCEClockInfo)));
+				rdev->pm.dpm.vce_states[i].evclk =
+					le16_to_cpu(vce_clk->usEVClkLow) | (vce_clk->ucEVClkHigh << 16);
+				rdev->pm.dpm.vce_states[i].ecclk =
+					le16_to_cpu(vce_clk->usECClkLow) | (vce_clk->ucECClkHigh << 16);
+				rdev->pm.dpm.vce_states[i].clk_idx =
+					state_entry->ucClockInfoIndex & 0x3f;
+				rdev->pm.dpm.vce_states[i].pstate =
+					(state_entry->ucClockInfoIndex & 0xc0) >> 6;
+				state_entry = (ATOM_PPLIB_VCE_State_Record *)
+					((u8 *)state_entry + sizeof(ATOM_PPLIB_VCE_State_Record));
+			}
 		}
 		if ((le16_to_cpu(ext_hdr->usSize) >= SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V3) &&
 			ext_hdr->usUVDTableOffset) {
diff --git a/drivers/gpu/drm/radeon/r600_dpm.h b/drivers/gpu/drm/radeon/r600_dpm.h
index 1000bf9719f..46b9d2a0301 100644
--- a/drivers/gpu/drm/radeon/r600_dpm.h
+++ b/drivers/gpu/drm/radeon/r600_dpm.h
@@ -213,10 +213,10 @@ void r600_wait_for_power_level(struct radeon_device *rdev,
 void r600_start_dpm(struct radeon_device *rdev);
 void r600_stop_dpm(struct radeon_device *rdev);
 
-int r600_set_thermal_temperature_range(struct radeon_device *rdev,
-				       int min_temp, int max_temp);
 bool r600_is_internal_thermal_sensor(enum radeon_int_thermal_type sensor);
 
+int r600_get_platform_caps(struct radeon_device *rdev);
+
 int r600_parse_extended_power_table(struct radeon_device *rdev);
 void r600_free_extended_power_table(struct radeon_device *rdev);
 
diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c
index b7d3ecba43e..26ef8ced6f8 100644
--- a/drivers/gpu/drm/radeon/r600_hdmi.c
+++ b/drivers/gpu/drm/radeon/r600_hdmi.c
@@ -133,7 +133,7 @@ struct radeon_hdmi_acr r600_hdmi_acr(uint32_t clock)
 /*
  * update the N and CTS parameters for a given pixel clock rate
  */
-static void r600_hdmi_update_ACR(struct drm_encoder *encoder, uint32_t clock)
+void r600_hdmi_update_ACR(struct drm_encoder *encoder, uint32_t clock)
 {
 	struct drm_device *dev = encoder->dev;
 	struct radeon_device *rdev = dev->dev_private;
@@ -142,21 +142,33 @@ static void r600_hdmi_update_ACR(struct drm_encoder *encoder, uint32_t clock)
 	struct radeon_encoder_atom_dig *dig = radeon_encoder->enc_priv;
 	uint32_t offset = dig->afmt->offset;
 
-	WREG32(HDMI0_ACR_32_0 + offset, HDMI0_ACR_CTS_32(acr.cts_32khz));
-	WREG32(HDMI0_ACR_32_1 + offset, acr.n_32khz);
-
-	WREG32(HDMI0_ACR_44_0 + offset, HDMI0_ACR_CTS_44(acr.cts_44_1khz));
-	WREG32(HDMI0_ACR_44_1 + offset, acr.n_44_1khz);
-
-	WREG32(HDMI0_ACR_48_0 + offset, HDMI0_ACR_CTS_48(acr.cts_48khz));
-	WREG32(HDMI0_ACR_48_1 + offset, acr.n_48khz);
+	WREG32_P(HDMI0_ACR_32_0 + offset,
+		 HDMI0_ACR_CTS_32(acr.cts_32khz),
+		 ~HDMI0_ACR_CTS_32_MASK);
+	WREG32_P(HDMI0_ACR_32_1 + offset,
+		 HDMI0_ACR_N_32(acr.n_32khz),
+		 ~HDMI0_ACR_N_32_MASK);
+
+	WREG32_P(HDMI0_ACR_44_0 + offset,
+		 HDMI0_ACR_CTS_44(acr.cts_44_1khz),
+		 ~HDMI0_ACR_CTS_44_MASK);
+	WREG32_P(HDMI0_ACR_44_1 + offset,
+		 HDMI0_ACR_N_44(acr.n_44_1khz),
+		 ~HDMI0_ACR_N_44_MASK);
+
+	WREG32_P(HDMI0_ACR_48_0 + offset,
+		 HDMI0_ACR_CTS_48(acr.cts_48khz),
+		 ~HDMI0_ACR_CTS_48_MASK);
+	WREG32_P(HDMI0_ACR_48_1 + offset,
+		 HDMI0_ACR_N_48(acr.n_48khz),
+		 ~HDMI0_ACR_N_48_MASK);
 }
 
 /*
  * build a HDMI Video Info Frame
  */
-static void r600_hdmi_update_avi_infoframe(struct drm_encoder *encoder,
-					   void *buffer, size_t size)
+void r600_hdmi_update_avi_infoframe(struct drm_encoder *encoder, void *buffer,
+				    size_t size)
 {
 	struct drm_device *dev = encoder->dev;
 	struct radeon_device *rdev = dev->dev_private;
@@ -231,7 +243,7 @@ int r600_hdmi_buffer_status_changed(struct drm_encoder *encoder)
 /*
  * write the audio workaround status to the hardware
  */
-static void r600_hdmi_audio_workaround(struct drm_encoder *encoder)
+void r600_hdmi_audio_workaround(struct drm_encoder *encoder)
 {
 	struct drm_device *dev = encoder->dev;
 	struct radeon_device *rdev = dev->dev_private;
@@ -320,124 +332,6 @@ void r600_audio_set_dto(struct drm_encoder *encoder, u32 clock)
 	}
 }
 
-static void dce3_2_afmt_write_speaker_allocation(struct drm_encoder *encoder)
-{
-	struct radeon_device *rdev = encoder->dev->dev_private;
-	struct drm_connector *connector;
-	struct radeon_connector *radeon_connector = NULL;
-	u32 tmp;
-	u8 *sadb;
-	int sad_count;
-
-	/* XXX: setting this register causes hangs on some asics */
-	return;
-
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
-		if (connector->encoder == encoder) {
-			radeon_connector = to_radeon_connector(connector);
-			break;
-		}
-	}
-
-	if (!radeon_connector) {
-		DRM_ERROR("Couldn't find encoder's connector\n");
-		return;
-	}
-
-	sad_count = drm_edid_to_speaker_allocation(radeon_connector->edid, &sadb);
-	if (sad_count < 0) {
-		DRM_ERROR("Couldn't read Speaker Allocation Data Block: %d\n", sad_count);
-		return;
-	}
-
-	/* program the speaker allocation */
-	tmp = RREG32(AZ_F0_CODEC_PIN0_CONTROL_CHANNEL_SPEAKER);
-	tmp &= ~(DP_CONNECTION | SPEAKER_ALLOCATION_MASK);
-	/* set HDMI mode */
-	tmp |= HDMI_CONNECTION;
-	if (sad_count)
-		tmp |= SPEAKER_ALLOCATION(sadb[0]);
-	else
-		tmp |= SPEAKER_ALLOCATION(5); /* stereo */
-	WREG32(AZ_F0_CODEC_PIN0_CONTROL_CHANNEL_SPEAKER, tmp);
-
-	kfree(sadb);
-}
-
-static void dce3_2_afmt_write_sad_regs(struct drm_encoder *encoder)
-{
-	struct radeon_device *rdev = encoder->dev->dev_private;
-	struct drm_connector *connector;
-	struct radeon_connector *radeon_connector = NULL;
-	struct cea_sad *sads;
-	int i, sad_count;
-
-	static const u16 eld_reg_to_type[][2] = {
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR0, HDMI_AUDIO_CODING_TYPE_PCM },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR1, HDMI_AUDIO_CODING_TYPE_AC3 },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR2, HDMI_AUDIO_CODING_TYPE_MPEG1 },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR3, HDMI_AUDIO_CODING_TYPE_MP3 },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR4, HDMI_AUDIO_CODING_TYPE_MPEG2 },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR5, HDMI_AUDIO_CODING_TYPE_AAC_LC },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR6, HDMI_AUDIO_CODING_TYPE_DTS },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR7, HDMI_AUDIO_CODING_TYPE_ATRAC },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR9, HDMI_AUDIO_CODING_TYPE_EAC3 },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR10, HDMI_AUDIO_CODING_TYPE_DTS_HD },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR11, HDMI_AUDIO_CODING_TYPE_MLP },
-		{ AZ_F0_CODEC_PIN0_CONTROL_AUDIO_DESCRIPTOR13, HDMI_AUDIO_CODING_TYPE_WMA_PRO },
-	};
-
-	list_for_each_entry(connector, &encoder->dev->mode_config.connector_list, head) {
-		if (connector->encoder == encoder) {
-			radeon_connector = to_radeon_connector(connector);
-			break;
-		}
-	}
-
-	if (!radeon_connector) {
-		DRM_ERROR("Couldn't find encoder's connector\n");
-		return;
-	}
-
-	sad_count = drm_edid_to_sad(radeon_connector->edid, &sads);
-	if (sad_count < 0) {
-		DRM_ERROR("Couldn't read SADs: %d\n", sad_count);
-		return;
-	}
-	BUG_ON(!sads);
-
-	for (i = 0; i < ARRAY_SIZE(eld_reg_to_type); i++) {
-		u32 value = 0;
-		u8 stereo_freqs = 0;
-		int max_channels = -1;
-		int j;
-
-		for (j = 0; j < sad_count; j++) {
-			struct cea_sad *sad = &sads[j];
-
-			if (sad->format == eld_reg_to_type[i][1]) {
-				if (sad->channels > max_channels) {
-					value = MAX_CHANNELS(sad->channels) |
-						DESCRIPTOR_BYTE_2(sad->byte2) |
-						SUPPORTED_FREQUENCIES(sad->freq);
-					max_channels = sad->channels;
-				}
-
-				if (sad->format == HDMI_AUDIO_CODING_TYPE_PCM)
-					stereo_freqs |= sad->freq;
-				else
-					break;
-			}
-		}
-
-		value |= SUPPORTED_FREQUENCIES_STEREO(stereo_freqs);
-
-		WREG32(eld_reg_to_type[i][0], value);
-	}
-
-	kfree(sads);
-}
-
 /*
  * update the info frames with the data from the current display mode
  */
@@ -450,6 +344,7 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod
 	u8 buffer[HDMI_INFOFRAME_HEADER_SIZE + HDMI_AVI_INFOFRAME_SIZE];
 	struct hdmi_avi_infoframe frame;
 	uint32_t offset;
+	uint32_t acr_ctl;
 	ssize_t err;
 
 	if (!dig || !dig->afmt)
@@ -460,54 +355,50 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod
 		return;
 	offset = dig->afmt->offset;
 
-	r600_audio_set_dto(encoder, mode->clock);
+	/* disable audio prior to setting up hw */
+	dig->afmt->pin = r600_audio_get_pin(rdev);
+	r600_audio_enable(rdev, dig->afmt->pin, false);
 
-	WREG32(HDMI0_VBI_PACKET_CONTROL + offset,
-	       HDMI0_NULL_SEND); /* send null packets when required */
-
-	WREG32(HDMI0_AUDIO_CRC_CONTROL + offset, 0x1000);
-
-	if (ASIC_IS_DCE32(rdev)) {
-		WREG32(HDMI0_AUDIO_PACKET_CONTROL + offset,
-		       HDMI0_AUDIO_DELAY_EN(1) | /* default audio delay */
-		       HDMI0_AUDIO_PACKETS_PER_LINE(3)); /* should be suffient for all audio modes and small enough for all hblanks */
-		WREG32(AFMT_AUDIO_PACKET_CONTROL + offset,
-		       AFMT_AUDIO_SAMPLE_SEND | /* send audio packets */
-		       AFMT_60958_CS_UPDATE); /* allow 60958 channel status fields to be updated */
-	} else {
-		WREG32(HDMI0_AUDIO_PACKET_CONTROL + offset,
-		       HDMI0_AUDIO_SAMPLE_SEND | /* send audio packets */
-		       HDMI0_AUDIO_DELAY_EN(1) | /* default audio delay */
-		       HDMI0_AUDIO_PACKETS_PER_LINE(3) | /* should be suffient for all audio modes and small enough for all hblanks */
-		       HDMI0_60958_CS_UPDATE); /* allow 60958 channel status fields to be updated */
-	}
-
-	if (ASIC_IS_DCE32(rdev)) {
-		dce3_2_afmt_write_speaker_allocation(encoder);
-		dce3_2_afmt_write_sad_regs(encoder);
-	}
-
-	WREG32(HDMI0_ACR_PACKET_CONTROL + offset,
-	       HDMI0_ACR_SOURCE | /* select SW CTS value - XXX verify that hw CTS works on all families */
-	       HDMI0_ACR_AUTO_SEND); /* allow hw to sent ACR packets when required */
-
-	WREG32(HDMI0_VBI_PACKET_CONTROL + offset,
-	       HDMI0_NULL_SEND | /* send null packets when required */
-	       HDMI0_GC_SEND | /* send general control packets */
-	       HDMI0_GC_CONT); /* send general control packets every frame */
-
-	/* TODO: HDMI0_AUDIO_INFO_UPDATE */
-	WREG32(HDMI0_INFOFRAME_CONTROL0 + offset,
-	       HDMI0_AVI_INFO_SEND | /* enable AVI info frames */
-	       HDMI0_AVI_INFO_CONT | /* send AVI info frames every frame/field */
-	       HDMI0_AUDIO_INFO_SEND | /* enable audio info frames (frames won't be set until audio is enabled) */
-	       HDMI0_AUDIO_INFO_CONT); /* send audio info frames every frame/field */
-
-	WREG32(HDMI0_INFOFRAME_CONTROL1 + offset,
-	       HDMI0_AVI_INFO_LINE(2) | /* anything other than 0 */
-	       HDMI0_AUDIO_INFO_LINE(2)); /* anything other than 0 */
+	r600_audio_set_dto(encoder, mode->clock);
 
-	WREG32(HDMI0_GC + offset, 0); /* unset HDMI0_GC_AVMUTE */
+	WREG32_P(HDMI0_AUDIO_PACKET_CONTROL + offset,
+		 HDMI0_AUDIO_SAMPLE_SEND | /* send audio packets */
+		 HDMI0_AUDIO_DELAY_EN(1) | /* default audio delay */
+		 HDMI0_AUDIO_PACKETS_PER_LINE(3) | /* should be suffient for all audio modes and small enough for all hblanks */
+		 HDMI0_60958_CS_UPDATE, /* allow 60958 channel status fields to be updated */
+		 ~(HDMI0_AUDIO_SAMPLE_SEND |
+		   HDMI0_AUDIO_DELAY_EN_MASK |
+		   HDMI0_AUDIO_PACKETS_PER_LINE_MASK |
+		   HDMI0_60958_CS_UPDATE));
+
+	/* DCE 3.0 uses register that's normally for CRC_CONTROL */
+	acr_ctl = ASIC_IS_DCE3(rdev) ? DCE3_HDMI0_ACR_PACKET_CONTROL :
+				       HDMI0_ACR_PACKET_CONTROL;
+	WREG32_P(acr_ctl + offset,
+		 HDMI0_ACR_SOURCE | /* select SW CTS value - XXX verify that hw CTS works on all families */
+		 HDMI0_ACR_AUTO_SEND, /* allow hw to sent ACR packets when required */
+		 ~(HDMI0_ACR_SOURCE |
+		   HDMI0_ACR_AUTO_SEND));
+
+	WREG32_OR(HDMI0_VBI_PACKET_CONTROL + offset,
+		  HDMI0_NULL_SEND | /* send null packets when required */
+		  HDMI0_GC_SEND | /* send general control packets */
+		  HDMI0_GC_CONT); /* send general control packets every frame */
+
+	WREG32_OR(HDMI0_INFOFRAME_CONTROL0 + offset,
+		  HDMI0_AVI_INFO_SEND | /* enable AVI info frames */
+		  HDMI0_AVI_INFO_CONT | /* send AVI info frames every frame/field */
+		  HDMI0_AUDIO_INFO_SEND | /* enable audio info frames (frames won't be set until audio is enabled) */
+		  HDMI0_AUDIO_INFO_UPDATE); /* required for audio info values to be updated */
+
+	WREG32_P(HDMI0_INFOFRAME_CONTROL1 + offset,
+		 HDMI0_AVI_INFO_LINE(2) | /* anything other than 0 */
+		 HDMI0_AUDIO_INFO_LINE(2), /* anything other than 0 */
+		 ~(HDMI0_AVI_INFO_LINE_MASK |
+		   HDMI0_AUDIO_INFO_LINE_MASK));
+
+	WREG32_AND(HDMI0_GC + offset,
+		   ~HDMI0_GC_AVMUTE); /* unset HDMI0_GC_AVMUTE */
 
 	err = drm_hdmi_avi_infoframe_from_display_mode(&frame, mode);
 	if (err < 0) {
@@ -522,19 +413,45 @@ void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mod
 	}
 
 	r600_hdmi_update_avi_infoframe(encoder, buffer, sizeof(buffer));
+
+	/* fglrx duplicates INFOFRAME_CONTROL0 & INFOFRAME_CONTROL1 ops here */
+
+	WREG32_AND(HDMI0_GENERIC_PACKET_CONTROL + offset,
+		   ~(HDMI0_GENERIC0_SEND |
+		     HDMI0_GENERIC0_CONT |
+		     HDMI0_GENERIC0_UPDATE |
+		     HDMI0_GENERIC1_SEND |
+		     HDMI0_GENERIC1_CONT |
+		     HDMI0_GENERIC0_LINE_MASK |
+		     HDMI0_GENERIC1_LINE_MASK));
+
 	r600_hdmi_update_ACR(encoder, mode->clock);
 
+	WREG32_P(HDMI0_60958_0 + offset,
+		 HDMI0_60958_CS_CHANNEL_NUMBER_L(1),
+		 ~(HDMI0_60958_CS_CHANNEL_NUMBER_L_MASK |
+		   HDMI0_60958_CS_CLOCK_ACCURACY_MASK));
+
+	WREG32_P(HDMI0_60958_1 + offset,
+		 HDMI0_60958_CS_CHANNEL_NUMBER_R(2),
+		 ~HDMI0_60958_CS_CHANNEL_NUMBER_R_MASK);
+
 	/* it's unknown what these bits do excatly, but it's indeed quite useful for debugging */
 	WREG32(HDMI0_RAMP_CONTROL0 + offset, 0x00FFFFFF);
 	WREG32(HDMI0_RAMP_CONTROL1 + offset, 0x007FFFFF);
 	WREG32(HDMI0_RAMP_CONTROL2 + offset, 0x00000001);
 	WREG32(HDMI0_RAMP_CONTROL3 + offset, 0x00000001);
 
-	r600_hdmi_audio_workaround(encoder);
+	/* enable audio after to setting up hw */
+	r600_audio_enable(rdev, dig->afmt->pin, true);
 }
 
-/*
- * update settings with current parameters from audio engine
+/**
+ * r600_hdmi_update_audio_settings - Update audio infoframe
+ *
+ * @encoder: drm encoder
+ *
+ * Gets info about current audio stream and updates audio infoframe.
  */
 void r600_hdmi_update_audio_settings(struct drm_encoder *encoder)
 {
@@ -546,7 +463,7 @@ void r600_hdmi_update_audio_settings(struct drm_encoder *encoder)
 	uint8_t buffer[HDMI_INFOFRAME_HEADER_SIZE + HDMI_AUDIO_INFOFRAME_SIZE];
 	struct hdmi_audio_infoframe frame;
 	uint32_t offset;
-	uint32_t iec;
+	uint32_t value;
 	ssize_t err;
 
 	if (!dig->afmt || !dig->afmt->enabled)
@@ -559,60 +476,6 @@ void r600_hdmi_update_audio_settings(struct drm_encoder *encoder)
 	DRM_DEBUG("0x%02X IEC60958 status bits and 0x%02X category code\n",
 		  (int)audio.status_bits, (int)audio.category_code);
 
-	iec = 0;
-	if (audio.status_bits & AUDIO_STATUS_PROFESSIONAL)
-		iec |= 1 << 0;
-	if (audio.status_bits & AUDIO_STATUS_NONAUDIO)
-		iec |= 1 << 1;
-	if (audio.status_bits & AUDIO_STATUS_COPYRIGHT)
-		iec |= 1 << 2;
-	if (audio.status_bits & AUDIO_STATUS_EMPHASIS)
-		iec |= 1 << 3;
-
-	iec |= HDMI0_60958_CS_CATEGORY_CODE(audio.category_code);
-
-	switch (audio.rate) {
-	case 32000:
-		iec |= HDMI0_60958_CS_SAMPLING_FREQUENCY(0x3);
-		break;
-	case 44100:
-		iec |= HDMI0_60958_CS_SAMPLING_FREQUENCY(0x0);
-		break;
-	case 48000:
-		iec |= HDMI0_60958_CS_SAMPLING_FREQUENCY(0x2);
-		break;
-	case 88200:
-		iec |= HDMI0_60958_CS_SAMPLING_FREQUENCY(0x8);
-		break;
-	case 96000:
-		iec |= HDMI0_60958_CS_SAMPLING_FREQUENCY(0xa);
-		break;
-	case 176400:
-		iec |= HDMI0_60958_CS_SAMPLING_FREQUENCY(0xc);
-		break;
-	case 192000:
-		iec |= HDMI0_60958_CS_SAMPLING_FREQUENCY(0xe);
-		break;
-	}
-
-	WREG32(HDMI0_60958_0 + offset, iec);
-
-	iec = 0;
-	switch (audio.bits_per_sample) {
-	case 16:
-		iec |= HDMI0_60958_CS_WORD_LENGTH(0x2);
-		break;
-	case 20:
-		iec |= HDMI0_60958_CS_WORD_LENGTH(0x3);
-		break;
-	case 24:
-		iec |= HDMI0_60958_CS_WORD_LENGTH(0xb);
-		break;
-	}
-	if (audio.status_bits & AUDIO_STATUS_V)
-		iec |= 0x5 << 16;
-	WREG32_P(HDMI0_60958_1 + offset, iec, ~0x5000f);
-
 	err = hdmi_audio_infoframe_init(&frame);
 	if (err < 0) {
 		DRM_ERROR("failed to setup audio infoframe\n");
@@ -627,8 +490,22 @@ void r600_hdmi_update_audio_settings(struct drm_encoder *encoder)
 		return;
 	}
 
+	value = RREG32(HDMI0_AUDIO_PACKET_CONTROL + offset);
+	if (value & HDMI0_AUDIO_TEST_EN)
+		WREG32(HDMI0_AUDIO_PACKET_CONTROL + offset,
+		       value & ~HDMI0_AUDIO_TEST_EN);
+
+	WREG32_OR(HDMI0_CONTROL + offset,
+		  HDMI0_ERROR_ACK);
+
+	WREG32_AND(HDMI0_INFOFRAME_CONTROL0 + offset,
+		   ~HDMI0_AUDIO_INFO_SOURCE);
+
 	r600_hdmi_update_audio_infoframe(encoder, buffer, sizeof(buffer));
-	r600_hdmi_audio_workaround(encoder);
+
+	WREG32_OR(HDMI0_INFOFRAME_CONTROL0 + offset,
+		  HDMI0_AUDIO_INFO_CONT |
+		  HDMI0_AUDIO_INFO_UPDATE);
 }
 
 /*
@@ -651,11 +528,6 @@ void r600_hdmi_enable(struct drm_encoder *encoder, bool enable)
 	if (!enable && !dig->afmt->enabled)
 		return;
 
-	if (enable)
-		dig->afmt->pin = r600_audio_get_pin(rdev);
-	else
-		dig->afmt->pin = NULL;
-
 	/* Older chipsets require setting HDMI and routing manually */
 	if (!ASIC_IS_DCE3(rdev)) {
 		if (enable)
diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h
index ebe38724a97..f94e7a9afe7 100644
--- a/drivers/gpu/drm/radeon/r600d.h
+++ b/drivers/gpu/drm/radeon/r600d.h
@@ -701,11 +701,18 @@
 #define RLC_UCODE_DATA                                    0x3f30
 
 #define SRBM_SOFT_RESET                                   0xe60
+#       define SOFT_RESET_BIF                             (1 << 1)
 #       define SOFT_RESET_DMA                             (1 << 12)
 #       define SOFT_RESET_RLC                             (1 << 13)
 #       define SOFT_RESET_UVD                             (1 << 18)
 #       define RV770_SOFT_RESET_DMA                       (1 << 20)
 
+#define BIF_SCRATCH0                                      0x5438
+
+#define BUS_CNTL                                          0x5420
+#       define BIOS_ROM_DIS                               (1 << 1)
+#       define VGA_COHE_SPEC_TIMER_DIS                    (1 << 9)
+
 #define CP_INT_CNTL                                       0xc124
 #       define CNTX_BUSY_INT_ENABLE                       (1 << 19)
 #       define CNTX_EMPTY_INT_ENABLE                      (1 << 20)
@@ -1022,15 +1029,18 @@
 #define HDMI0_AUDIO_PACKET_CONTROL   0x7408
 #       define HDMI0_AUDIO_SAMPLE_SEND  (1 << 0)
 #       define HDMI0_AUDIO_DELAY_EN(x)  (((x) & 3) << 4)
+#       define HDMI0_AUDIO_DELAY_EN_MASK	(3 << 4)
 #       define HDMI0_AUDIO_SEND_MAX_PACKETS  (1 << 8)
 #       define HDMI0_AUDIO_TEST_EN         (1 << 12)
 #       define HDMI0_AUDIO_PACKETS_PER_LINE(x)  (((x) & 0x1f) << 16)
+#       define HDMI0_AUDIO_PACKETS_PER_LINE_MASK	(0x1f << 16)
 #       define HDMI0_AUDIO_CHANNEL_SWAP    (1 << 24)
 #       define HDMI0_60958_CS_UPDATE       (1 << 26)
 #       define HDMI0_AZ_FORMAT_WTRIG_MASK  (1 << 28)
 #       define HDMI0_AZ_FORMAT_WTRIG_ACK   (1 << 29)
 #define HDMI0_AUDIO_CRC_CONTROL      0x740c
 #       define HDMI0_AUDIO_CRC_EN    (1 << 0)
+#define DCE3_HDMI0_ACR_PACKET_CONTROL	0x740c
 #define HDMI0_VBI_PACKET_CONTROL     0x7410
 #       define HDMI0_NULL_SEND       (1 << 0)
 #       define HDMI0_GC_SEND         (1 << 4)
@@ -1047,7 +1057,9 @@
 #       define HDMI0_MPEG_INFO_UPDATE  (1 << 10)
 #define HDMI0_INFOFRAME_CONTROL1     0x7418
 #       define HDMI0_AVI_INFO_LINE(x)  (((x) & 0x3f) << 0)
+#       define HDMI0_AVI_INFO_LINE_MASK		(0x3f << 0)
 #       define HDMI0_AUDIO_INFO_LINE(x)  (((x) & 0x3f) << 8)
+#       define HDMI0_AUDIO_INFO_LINE_MASK	(0x3f << 8)
 #       define HDMI0_MPEG_INFO_LINE(x)  (((x) & 0x3f) << 16)
 #define HDMI0_GENERIC_PACKET_CONTROL 0x741c
 #       define HDMI0_GENERIC0_SEND   (1 << 0)
@@ -1056,7 +1068,9 @@
 #       define HDMI0_GENERIC1_SEND   (1 << 4)
 #       define HDMI0_GENERIC1_CONT   (1 << 5)
 #       define HDMI0_GENERIC0_LINE(x)  (((x) & 0x3f) << 16)
+#       define HDMI0_GENERIC0_LINE_MASK		(0x3f << 16)
 #       define HDMI0_GENERIC1_LINE(x)  (((x) & 0x3f) << 24)
+#       define HDMI0_GENERIC1_LINE_MASK		(0x3f << 24)
 #define HDMI0_GC                     0x7428
 #       define HDMI0_GC_AVMUTE       (1 << 0)
 #define HDMI0_AVI_INFO0              0x7454
@@ -1112,16 +1126,22 @@
 #define HDMI0_GENERIC1_6             0x74a8
 #define HDMI0_ACR_32_0               0x74ac
 #       define HDMI0_ACR_CTS_32(x)   (((x) & 0xfffff) << 12)
+#       define HDMI0_ACR_CTS_32_MASK		(0xfffff << 12)
 #define HDMI0_ACR_32_1               0x74b0
 #       define HDMI0_ACR_N_32(x)   (((x) & 0xfffff) << 0)
+#       define HDMI0_ACR_N_32_MASK		(0xfffff << 0)
 #define HDMI0_ACR_44_0               0x74b4
 #       define HDMI0_ACR_CTS_44(x)   (((x) & 0xfffff) << 12)
+#       define HDMI0_ACR_CTS_44_MASK		(0xfffff << 12)
 #define HDMI0_ACR_44_1               0x74b8
 #       define HDMI0_ACR_N_44(x)   (((x) & 0xfffff) << 0)
+#       define HDMI0_ACR_N_44_MASK		(0xfffff << 0)
 #define HDMI0_ACR_48_0               0x74bc
 #       define HDMI0_ACR_CTS_48(x)   (((x) & 0xfffff) << 12)
+#       define HDMI0_ACR_CTS_48_MASK		(0xfffff << 12)
 #define HDMI0_ACR_48_1               0x74c0
 #       define HDMI0_ACR_N_48(x)   (((x) & 0xfffff) << 0)
+#       define HDMI0_ACR_N_48_MASK		(0xfffff << 0)
 #define HDMI0_ACR_STATUS_0           0x74c4
 #define HDMI0_ACR_STATUS_1           0x74c8
 #define HDMI0_AUDIO_INFO0            0x74cc
@@ -1141,14 +1161,17 @@
 #       define HDMI0_60958_CS_CATEGORY_CODE(x)      (((x) & 0xff) << 8)
 #       define HDMI0_60958_CS_SOURCE_NUMBER(x)      (((x) & 0xf) << 16)
 #       define HDMI0_60958_CS_CHANNEL_NUMBER_L(x)   (((x) & 0xf) << 20)
+#       define HDMI0_60958_CS_CHANNEL_NUMBER_L_MASK	(0xf << 20)
 #       define HDMI0_60958_CS_SAMPLING_FREQUENCY(x) (((x) & 0xf) << 24)
 #       define HDMI0_60958_CS_CLOCK_ACCURACY(x)     (((x) & 3) << 28)
+#       define HDMI0_60958_CS_CLOCK_ACCURACY_MASK	(3 << 28)
 #define HDMI0_60958_1                0x74d8
 #       define HDMI0_60958_CS_WORD_LENGTH(x)        (((x) & 0xf) << 0)
 #       define HDMI0_60958_CS_ORIGINAL_SAMPLING_FREQUENCY(x)   (((x) & 0xf) << 4)
 #       define HDMI0_60958_CS_VALID_L(x)   (((x) & 1) << 16)
 #       define HDMI0_60958_CS_VALID_R(x)   (((x) & 1) << 18)
 #       define HDMI0_60958_CS_CHANNEL_NUMBER_R(x)   (((x) & 0xf) << 20)
+#       define HDMI0_60958_CS_CHANNEL_NUMBER_R_MASK	(0xf << 20)
 #define HDMI0_ACR_PACKET_CONTROL     0x74dc
 #       define HDMI0_ACR_SEND        (1 << 0)
 #       define HDMI0_ACR_CONT        (1 << 1)
@@ -1159,6 +1182,7 @@
 #       define HDMI0_ACR_48          3
 #       define HDMI0_ACR_SOURCE      (1 << 8) /* 0 - hw; 1 - cts value */
 #       define HDMI0_ACR_AUTO_SEND   (1 << 12)
+#define DCE3_HDMI0_AUDIO_CRC_CONTROL	0x74dc
 #define HDMI0_RAMP_CONTROL0          0x74e0
 #       define HDMI0_RAMP_MAX_COUNT(x)   (((x) & 0xffffff) << 0)
 #define HDMI0_RAMP_CONTROL1          0x74e4
@@ -1575,6 +1599,7 @@
 #              define PACKET3_CP_DMA_CMD_DAIC      (1 << 29)
 #define	PACKET3_SURFACE_SYNC				0x43
 #              define PACKET3_CB0_DEST_BASE_ENA    (1 << 6)
+#              define PACKET3_FULL_CACHE_ENA       (1 << 20) /* r7xx+ only */
 #              define PACKET3_TC_ACTION_ENA        (1 << 23)
 #              define PACKET3_VC_ACTION_ENA        (1 << 24)
 #              define PACKET3_CB_ACTION_ENA        (1 << 25)
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 45e1f447bc7..60c47f82912 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -99,6 +99,10 @@ extern int radeon_fastfb;
 extern int radeon_dpm;
 extern int radeon_aspm;
 extern int radeon_runtime_pm;
+extern int radeon_hard_reset;
+extern int radeon_vm_size;
+extern int radeon_vm_block_size;
+extern int radeon_deep_color;
 
 /*
  * Copy from radeon_drv.h so we don't have to include both and have conflicting
@@ -112,19 +116,16 @@ extern int radeon_runtime_pm;
 #define RADEONFB_CONN_LIMIT			4
 #define RADEON_BIOS_NUM_SCRATCH			8
 
-/* max number of rings */
-#define RADEON_NUM_RINGS			6
-
 /* fence seq are set to this number when signaled */
 #define RADEON_FENCE_SIGNALED_SEQ		0LL
 
 /* internal ring indices */
 /* r1xx+ has gfx CP ring */
-#define RADEON_RING_TYPE_GFX_INDEX	0
+#define RADEON_RING_TYPE_GFX_INDEX		0
 
 /* cayman has 2 compute CP rings */
-#define CAYMAN_RING_TYPE_CP1_INDEX	1
-#define CAYMAN_RING_TYPE_CP2_INDEX	2
+#define CAYMAN_RING_TYPE_CP1_INDEX		1
+#define CAYMAN_RING_TYPE_CP2_INDEX		2
 
 /* R600+ has an async dma ring */
 #define R600_RING_TYPE_DMA_INDEX		3
@@ -132,13 +133,29 @@ extern int radeon_runtime_pm;
 #define CAYMAN_RING_TYPE_DMA1_INDEX		4
 
 /* R600+ */
-#define R600_RING_TYPE_UVD_INDEX	5
+#define R600_RING_TYPE_UVD_INDEX		5
+
+/* TN+ */
+#define TN_RING_TYPE_VCE1_INDEX			6
+#define TN_RING_TYPE_VCE2_INDEX			7
+
+/* max number of rings */
+#define RADEON_NUM_RINGS			8
+
+/* number of hw syncs before falling back on blocking */
+#define RADEON_NUM_SYNCS			4
+
+/* number of hw syncs before falling back on blocking */
+#define RADEON_NUM_SYNCS			4
 
 /* hardcode those limit for now */
 #define RADEON_VA_IB_OFFSET			(1 << 20)
 #define RADEON_VA_RESERVED_SIZE			(8 << 20)
 #define RADEON_IB_VM_MAX_SIZE			(64 << 10)
 
+/* hard reset data */
+#define RADEON_ASIC_RESET_DATA                  0x39d5e86b
+
 /* reset flags */
 #define RADEON_RESET_GFX			(1 << 0)
 #define RADEON_RESET_COMPUTE			(1 << 1)
@@ -252,6 +269,7 @@ struct radeon_clock {
  * Power management
  */
 int radeon_pm_init(struct radeon_device *rdev);
+int radeon_pm_late_init(struct radeon_device *rdev);
 void radeon_pm_fini(struct radeon_device *rdev);
 void radeon_pm_compute_clocks(struct radeon_device *rdev);
 void radeon_pm_suspend(struct radeon_device *rdev);
@@ -348,9 +366,8 @@ int radeon_fence_emit(struct radeon_device *rdev, struct radeon_fence **fence, i
 void radeon_fence_process(struct radeon_device *rdev, int ring);
 bool radeon_fence_signaled(struct radeon_fence *fence);
 int radeon_fence_wait(struct radeon_fence *fence, bool interruptible);
-int radeon_fence_wait_locked(struct radeon_fence *fence);
-int radeon_fence_wait_next_locked(struct radeon_device *rdev, int ring);
-int radeon_fence_wait_empty_locked(struct radeon_device *rdev, int ring);
+int radeon_fence_wait_next(struct radeon_device *rdev, int ring);
+int radeon_fence_wait_empty(struct radeon_device *rdev, int ring);
 int radeon_fence_wait_any(struct radeon_device *rdev,
 			  struct radeon_fence **fences,
 			  bool intr);
@@ -413,6 +430,11 @@ struct radeon_mman {
 	struct ttm_bo_device		bdev;
 	bool				mem_global_referenced;
 	bool				initialized;
+
+#if defined(CONFIG_DEBUG_FS)
+	struct dentry			*vram;
+	struct dentry			*gtt;
+#endif
 };
 
 /* bo virtual address in a specific vm */
@@ -427,6 +449,7 @@ struct radeon_bo_va {
 
 	/* protected by vm mutex */
 	struct list_head		vm_list;
+	struct list_head		vm_status;
 
 	/* constant after initialization */
 	struct radeon_vm		*vm;
@@ -437,6 +460,7 @@ struct radeon_bo {
 	/* Protected by gem.mutex */
 	struct list_head		list;
 	/* Protected by tbo.reserved */
+	u32				initial_domain;
 	u32				placements[3];
 	struct ttm_placement		placement;
 	struct ttm_buffer_object	tbo;
@@ -459,16 +483,6 @@ struct radeon_bo {
 };
 #define gem_to_radeon_bo(gobj) container_of((gobj), struct radeon_bo, gem_base)
 
-struct radeon_bo_list {
-	struct ttm_validate_buffer tv;
-	struct radeon_bo	*bo;
-	uint64_t		gpu_offset;
-	bool			written;
-	unsigned		domain;
-	unsigned		alt_domain;
-	u32			tiling_flags;
-};
-
 int radeon_gem_debugfs_init(struct radeon_device *rdev);
 
 /* sub-allocation manager, it has to be protected by another lock.
@@ -544,7 +558,6 @@ int radeon_mode_dumb_mmap(struct drm_file *filp,
 /*
  * Semaphores.
  */
-/* everything here is constant */
 struct radeon_semaphore {
 	struct radeon_sa_bo		*sa_bo;
 	signed				waiters;
@@ -667,14 +680,15 @@ void radeon_doorbell_free(struct radeon_device *rdev, u32 doorbell);
  * IRQS.
  */
 
-struct radeon_unpin_work {
-	struct work_struct work;
-	struct radeon_device *rdev;
-	int crtc_id;
-	struct radeon_fence *fence;
+struct radeon_flip_work {
+	struct work_struct		flip_work;
+	struct work_struct		unpin_work;
+	struct radeon_device		*rdev;
+	int				crtc_id;
+	uint64_t			base;
 	struct drm_pending_vblank_event *event;
-	struct radeon_bo *old_rbo;
-	u64 new_crtc_base;
+	struct radeon_bo		*old_rbo;
+	struct radeon_fence		*fence;
 };
 
 struct r500_irq_stat_regs {
@@ -721,6 +735,12 @@ struct cik_irq_stat_regs {
 	u32 disp_int_cont4;
 	u32 disp_int_cont5;
 	u32 disp_int_cont6;
+	u32 d1grph_int;
+	u32 d2grph_int;
+	u32 d3grph_int;
+	u32 d4grph_int;
+	u32 d5grph_int;
+	u32 d6grph_int;
 };
 
 union radeon_irq_stat_regs {
@@ -730,10 +750,6 @@ union radeon_irq_stat_regs {
 	struct cik_irq_stat_regs cik;
 };
 
-#define RADEON_MAX_HPD_PINS 6
-#define RADEON_MAX_CRTCS 6
-#define RADEON_MAX_AFMT_BLOCKS 7
-
 struct radeon_irq {
 	bool				installed;
 	spinlock_t			lock;
@@ -777,20 +793,17 @@ struct radeon_ib {
 struct radeon_ring {
 	struct radeon_bo	*ring_obj;
 	volatile uint32_t	*ring;
-	unsigned		rptr;
 	unsigned		rptr_offs;
-	unsigned		rptr_reg;
 	unsigned		rptr_save_reg;
 	u64			next_rptr_gpu_addr;
 	volatile u32		*next_rptr_cpu_addr;
 	unsigned		wptr;
 	unsigned		wptr_old;
-	unsigned		wptr_reg;
 	unsigned		ring_size;
 	unsigned		ring_free_dw;
 	int			count_dw;
-	unsigned long		last_activity;
-	unsigned		last_rptr;
+	atomic_t		last_rptr;
+	atomic64_t		last_activity;
 	uint64_t		gpu_addr;
 	uint32_t		align_mask;
 	uint32_t		ptr_mask;
@@ -823,13 +836,8 @@ struct radeon_mec {
 /* maximum number of VMIDs */
 #define RADEON_NUM_VM	16
 
-/* defines number of bits in page table versus page directory,
- * a page is 4KB so we have 12 bits offset, 9 bits in the page
- * table and the remaining 19 bits are in the page directory */
-#define RADEON_VM_BLOCK_SIZE   9
-
 /* number of entries in page table */
-#define RADEON_VM_PTE_COUNT (1 << RADEON_VM_BLOCK_SIZE)
+#define RADEON_VM_PTE_COUNT (1 << radeon_vm_block_size)
 
 /* PTBs (Page Table Blocks) need to be aligned to 32K */
 #define RADEON_VM_PTB_ALIGN_SIZE   32768
@@ -842,30 +850,48 @@ struct radeon_mec {
 #define R600_PTE_READABLE	(1 << 5)
 #define R600_PTE_WRITEABLE	(1 << 6)
 
+/* PTE (Page Table Entry) fragment field for different page sizes */
+#define R600_PTE_FRAG_4KB	(0 << 7)
+#define R600_PTE_FRAG_64KB	(4 << 7)
+#define R600_PTE_FRAG_256KB	(6 << 7)
+
+/* flags used for GART page table entries on R600+ */
+#define R600_PTE_GART	( R600_PTE_VALID | R600_PTE_SYSTEM | R600_PTE_SNOOPED \
+			| R600_PTE_READABLE | R600_PTE_WRITEABLE)
+
+struct radeon_vm_pt {
+	struct radeon_bo		*bo;
+	uint64_t			addr;
+};
+
 struct radeon_vm {
-	struct list_head		list;
 	struct list_head		va;
 	unsigned			id;
 
+	/* BOs freed, but not yet updated in the PT */
+	struct list_head		freed;
+
 	/* contains the page directory */
-	struct radeon_sa_bo		*page_directory;
+	struct radeon_bo		*page_directory;
 	uint64_t			pd_gpu_addr;
+	unsigned			max_pde_used;
 
 	/* array of page tables, one for each page directory entry */
-	struct radeon_sa_bo		**page_tables;
+	struct radeon_vm_pt		*page_tables;
+
+	struct radeon_bo_va		*ib_bo_va;
 
 	struct mutex			mutex;
 	/* last fence for cs using this vm */
 	struct radeon_fence		*fence;
 	/* last flush or NULL if we still need to flush */
 	struct radeon_fence		*last_flush;
+	/* last use of vmid */
+	struct radeon_fence		*last_id_use;
 };
 
 struct radeon_vm_manager {
-	struct mutex			lock;
-	struct list_head		lru_vm;
 	struct radeon_fence		*active[RADEON_NUM_VM];
-	struct radeon_sa_manager	sa_manager;
 	uint32_t			max_pfn;
 	/* number of VMIDs */
 	unsigned			nvm;
@@ -941,15 +967,15 @@ void radeon_ring_unlock_commit(struct radeon_device *rdev, struct radeon_ring *c
 void radeon_ring_undo(struct radeon_ring *ring);
 void radeon_ring_unlock_undo(struct radeon_device *rdev, struct radeon_ring *cp);
 int radeon_ring_test(struct radeon_device *rdev, struct radeon_ring *cp);
-void radeon_ring_force_activity(struct radeon_device *rdev, struct radeon_ring *ring);
-void radeon_ring_lockup_update(struct radeon_ring *ring);
+void radeon_ring_lockup_update(struct radeon_device *rdev,
+			       struct radeon_ring *ring);
 bool radeon_ring_test_lockup(struct radeon_device *rdev, struct radeon_ring *ring);
 unsigned radeon_ring_backup(struct radeon_device *rdev, struct radeon_ring *ring,
 			    uint32_t **data);
 int radeon_ring_restore(struct radeon_device *rdev, struct radeon_ring *ring,
 			unsigned size, uint32_t *data);
 int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *cp, unsigned ring_size,
-		     unsigned rptr_offs, unsigned rptr_reg, unsigned wptr_reg, u32 nop);
+		     unsigned rptr_offs, u32 nop);
 void radeon_ring_fini(struct radeon_device *rdev, struct radeon_ring *cp);
 
 
@@ -968,9 +994,12 @@ void cayman_dma_fini(struct radeon_device *rdev);
 struct radeon_cs_reloc {
 	struct drm_gem_object		*gobj;
 	struct radeon_bo		*robj;
-	struct radeon_bo_list		lobj;
+	struct ttm_validate_buffer	tv;
+	uint64_t			gpu_offset;
+	unsigned			prefered_domains;
+	unsigned			allowed_domains;
+	uint32_t			tiling_flags;
 	uint32_t			handle;
-	uint32_t			flags;
 };
 
 struct radeon_cs_chunk {
@@ -994,6 +1023,7 @@ struct radeon_cs_parser {
 	unsigned		nrelocs;
 	struct radeon_cs_reloc	*relocs;
 	struct radeon_cs_reloc	**relocs_ptr;
+	struct radeon_cs_reloc	*vm_bos;
 	struct list_head	validated;
 	unsigned		dma_reloc_idx;
 	/* indices of various chunks */
@@ -1243,6 +1273,17 @@ enum radeon_dpm_event_src {
 	RADEON_DPM_EVENT_SRC_DIGIAL_OR_EXTERNAL = 4
 };
 
+#define RADEON_MAX_VCE_LEVELS 6
+
+enum radeon_vce_level {
+	RADEON_VCE_LEVEL_AC_ALL = 0,     /* AC, All cases */
+	RADEON_VCE_LEVEL_DC_EE = 1,      /* DC, entropy encoding */
+	RADEON_VCE_LEVEL_DC_LL_LOW = 2,  /* DC, low latency queue, res <= 720 */
+	RADEON_VCE_LEVEL_DC_LL_HIGH = 3, /* DC, low latency queue, 1080 >= res > 720 */
+	RADEON_VCE_LEVEL_DC_GP_LOW = 4,  /* DC, general purpose queue, res <= 720 */
+	RADEON_VCE_LEVEL_DC_GP_HIGH = 5, /* DC, general purpose queue, 1080 >= res > 720 */
+};
+
 struct radeon_ps {
 	u32 caps; /* vbios flags */
 	u32 class; /* vbios flags */
@@ -1253,6 +1294,8 @@ struct radeon_ps {
 	/* VCE clocks */
 	u32 evclk;
 	u32 ecclk;
+	bool vce_active;
+	enum radeon_vce_level vce_level;
 	/* asic priv */
 	void *ps_priv;
 };
@@ -1427,6 +1470,17 @@ enum radeon_dpm_forced_level {
 	RADEON_DPM_FORCED_LEVEL_HIGH = 2,
 };
 
+struct radeon_vce_state {
+	/* vce clocks */
+	u32 evclk;
+	u32 ecclk;
+	/* gpu clocks */
+	u32 sclk;
+	u32 mclk;
+	u8 clk_idx;
+	u8 pstate;
+};
+
 struct radeon_dpm {
 	struct radeon_ps        *ps;
 	/* number of valid power states */
@@ -1439,6 +1493,9 @@ struct radeon_dpm {
 	struct radeon_ps        *boot_ps;
 	/* default uvd power state */
 	struct radeon_ps        *uvd_ps;
+	/* vce requirements */
+	struct radeon_vce_state vce_states[RADEON_MAX_VCE_LEVELS];
+	enum radeon_vce_level vce_level;
 	enum radeon_pm_state_type state;
 	enum radeon_pm_state_type user_state;
 	u32                     platform_caps;
@@ -1464,6 +1521,7 @@ struct radeon_dpm {
 	/* special states active */
 	bool                    thermal_active;
 	bool                    uvd_active;
+	bool                    vce_active;
 	/* thermal handling */
 	struct radeon_dpm_thermal thermal;
 	/* forced levels */
@@ -1474,6 +1532,7 @@ struct radeon_dpm {
 };
 
 void radeon_dpm_enable_uvd(struct radeon_device *rdev, bool enable);
+void radeon_dpm_enable_vce(struct radeon_device *rdev, bool enable);
 
 struct radeon_pm {
 	struct mutex		mutex;
@@ -1579,6 +1638,46 @@ int radeon_uvd_calc_upll_dividers(struct radeon_device *rdev,
 int radeon_uvd_send_upll_ctlreq(struct radeon_device *rdev,
                                 unsigned cg_upll_func_cntl);
 
+/*
+ * VCE
+ */
+#define RADEON_MAX_VCE_HANDLES	16
+#define RADEON_VCE_STACK_SIZE	(1024*1024)
+#define RADEON_VCE_HEAP_SIZE	(4*1024*1024)
+
+struct radeon_vce {
+	struct radeon_bo	*vcpu_bo;
+	uint64_t		gpu_addr;
+	unsigned		fw_version;
+	unsigned		fb_version;
+	atomic_t		handles[RADEON_MAX_VCE_HANDLES];
+	struct drm_file		*filp[RADEON_MAX_VCE_HANDLES];
+	unsigned		img_size[RADEON_MAX_VCE_HANDLES];
+	struct delayed_work	idle_work;
+};
+
+int radeon_vce_init(struct radeon_device *rdev);
+void radeon_vce_fini(struct radeon_device *rdev);
+int radeon_vce_suspend(struct radeon_device *rdev);
+int radeon_vce_resume(struct radeon_device *rdev);
+int radeon_vce_get_create_msg(struct radeon_device *rdev, int ring,
+			      uint32_t handle, struct radeon_fence **fence);
+int radeon_vce_get_destroy_msg(struct radeon_device *rdev, int ring,
+			       uint32_t handle, struct radeon_fence **fence);
+void radeon_vce_free_handles(struct radeon_device *rdev, struct drm_file *filp);
+void radeon_vce_note_usage(struct radeon_device *rdev);
+int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi, unsigned size);
+int radeon_vce_cs_parse(struct radeon_cs_parser *p);
+bool radeon_vce_semaphore_emit(struct radeon_device *rdev,
+			       struct radeon_ring *ring,
+			       struct radeon_semaphore *semaphore,
+			       bool emit_wait);
+void radeon_vce_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
+void radeon_vce_fence_emit(struct radeon_device *rdev,
+			   struct radeon_fence *fence);
+int radeon_vce_ring_test(struct radeon_device *rdev, struct radeon_ring *ring);
+int radeon_vce_ib_test(struct radeon_device *rdev, struct radeon_ring *ring);
+
 struct r600_audio_pin {
 	int			channels;
 	int			rate;
@@ -1682,7 +1781,8 @@ struct radeon_asic {
 	/* gart */
 	struct {
 		void (*tlb_flush)(struct radeon_device *rdev);
-		int (*set_page)(struct radeon_device *rdev, int i, uint64_t addr);
+		void (*set_page)(struct radeon_device *rdev, unsigned i,
+				 uint64_t addr);
 	} gart;
 	struct {
 		int (*init)(struct radeon_device *rdev);
@@ -1768,6 +1868,7 @@ struct radeon_asic {
 		void (*set_pcie_lanes)(struct radeon_device *rdev, int lanes);
 		void (*set_clock_gating)(struct radeon_device *rdev, int enable);
 		int (*set_uvd_clocks)(struct radeon_device *rdev, u32 vclk, u32 dclk);
+		int (*set_vce_clocks)(struct radeon_device *rdev, u32 evclk, u32 ecclk);
 		int (*get_temperature)(struct radeon_device *rdev);
 	} pm;
 	/* dynamic power management */
@@ -1775,6 +1876,7 @@ struct radeon_asic {
 		int (*init)(struct radeon_device *rdev);
 		void (*setup_asic)(struct radeon_device *rdev);
 		int (*enable)(struct radeon_device *rdev);
+		int (*late_enable)(struct radeon_device *rdev);
 		void (*disable)(struct radeon_device *rdev);
 		int (*pre_set_power_state)(struct radeon_device *rdev);
 		int (*set_power_state)(struct radeon_device *rdev);
@@ -1792,9 +1894,8 @@ struct radeon_asic {
 	} dpm;
 	/* pageflipping */
 	struct {
-		void (*pre_page_flip)(struct radeon_device *rdev, int crtc);
-		u32 (*page_flip)(struct radeon_device *rdev, int crtc, u64 crtc_base);
-		void (*post_page_flip)(struct radeon_device *rdev, int crtc);
+		void (*page_flip)(struct radeon_device *rdev, int crtc, u64 crtc_base);
+		bool (*page_flip_pending)(struct radeon_device *rdev, int crtc);
 	} pflip;
 };
 
@@ -1833,6 +1934,7 @@ struct r600_asic {
 	unsigned		tiling_group_size;
 	unsigned		tile_config;
 	unsigned		backend_map;
+	unsigned		active_simds;
 };
 
 struct rv770_asic {
@@ -1858,6 +1960,7 @@ struct rv770_asic {
 	unsigned		tiling_group_size;
 	unsigned		tile_config;
 	unsigned		backend_map;
+	unsigned		active_simds;
 };
 
 struct evergreen_asic {
@@ -1884,6 +1987,7 @@ struct evergreen_asic {
 	unsigned tiling_group_size;
 	unsigned tile_config;
 	unsigned backend_map;
+	unsigned active_simds;
 };
 
 struct cayman_asic {
@@ -1922,6 +2026,7 @@ struct cayman_asic {
 	unsigned multi_gpu_tile_size;
 
 	unsigned tile_config;
+	unsigned active_simds;
 };
 
 struct si_asic {
@@ -1952,6 +2057,7 @@ struct si_asic {
 
 	unsigned tile_config;
 	uint32_t tile_mode_array[32];
+	uint32_t active_cus;
 };
 
 struct cik_asic {
@@ -1983,6 +2089,7 @@ struct cik_asic {
 	unsigned tile_config;
 	uint32_t tile_mode_array[32];
 	uint32_t macrotile_mode_array[16];
+	uint32_t active_cus;
 };
 
 union radeon_asic_config {
@@ -2028,6 +2135,8 @@ int radeon_gem_wait_idle_ioctl(struct drm_device *dev, void *data,
 			      struct drm_file *filp);
 int radeon_gem_va_ioctl(struct drm_device *dev, void *data,
 			  struct drm_file *filp);
+int radeon_gem_op_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *filp);
 int radeon_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp);
 int radeon_gem_set_tiling_ioctl(struct drm_device *dev, void *data,
 				struct drm_file *filp);
@@ -2173,6 +2282,7 @@ struct radeon_device {
 	struct radeon_gem		gem;
 	struct radeon_pm		pm;
 	struct radeon_uvd		uvd;
+	struct radeon_vce		vce;
 	uint32_t			bios_scratch[RADEON_BIOS_NUM_SCRATCH];
 	struct radeon_wb		wb;
 	struct radeon_dummy_page	dummy_page;
@@ -2192,6 +2302,7 @@ struct radeon_device {
 	const struct firmware *sdma_fw;	/* CIK SDMA firmware */
 	const struct firmware *smc_fw;	/* SMC firmware */
 	const struct firmware *uvd_fw;	/* UVD firmware */
+	const struct firmware *vce_fw;	/* VCE firmware */
 	struct r600_vram_scratch vram_scratch;
 	int msi_enabled; /* msi enabled */
 	struct r600_ih ih; /* r6/700 interrupt ring */
@@ -2216,6 +2327,10 @@ struct radeon_device {
 	/* virtual memory */
 	struct radeon_vm_manager	vm_manager;
 	struct mutex			gpu_clock_mutex;
+	/* memory stats */
+	atomic64_t			vram_usage;
+	atomic64_t			gtt_usage;
+	atomic64_t			num_bytes_moved;
 	/* ACPI interface */
 	struct radeon_atif		atif;
 	struct radeon_atcs		atcs;
@@ -2229,6 +2344,7 @@ struct radeon_device {
 	bool have_disp_power_ref;
 };
 
+bool radeon_is_px(struct drm_device *dev);
 int radeon_device_init(struct radeon_device *rdev,
 		       struct drm_device *ddev,
 		       struct pci_dev *pdev,
@@ -2539,6 +2655,10 @@ void r100_pll_errata_after_index(struct radeon_device *rdev);
 #define ASIC_IS_DCE64(rdev) ((rdev->family == CHIP_OLAND))
 #define ASIC_IS_NODCE(rdev) ((rdev->family == CHIP_HAINAN))
 #define ASIC_IS_DCE8(rdev) ((rdev->family >= CHIP_BONAIRE))
+#define ASIC_IS_DCE81(rdev) ((rdev->family == CHIP_KAVERI))
+#define ASIC_IS_DCE82(rdev) ((rdev->family == CHIP_BONAIRE))
+#define ASIC_IS_DCE83(rdev) ((rdev->family == CHIP_KABINI) || \
+			     (rdev->family == CHIP_MULLINS))
 
 #define ASIC_IS_LOMBOK(rdev) ((rdev->ddev->pdev->device == 0x6849) || \
 			      (rdev->ddev->pdev->device == 0x6850) || \
@@ -2626,6 +2746,7 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 #define radeon_set_pcie_lanes(rdev, l) (rdev)->asic->pm.set_pcie_lanes((rdev), (l))
 #define radeon_set_clock_gating(rdev, e) (rdev)->asic->pm.set_clock_gating((rdev), (e))
 #define radeon_set_uvd_clocks(rdev, v, d) (rdev)->asic->pm.set_uvd_clocks((rdev), (v), (d))
+#define radeon_set_vce_clocks(rdev, ev, ec) (rdev)->asic->pm.set_vce_clocks((rdev), (ev), (ec))
 #define radeon_get_temperature(rdev) (rdev)->asic->pm.get_temperature((rdev))
 #define radeon_set_surface_reg(rdev, r, f, p, o, s) ((rdev)->asic->surface.set_reg((rdev), (r), (f), (p), (o), (s)))
 #define radeon_clear_surface_reg(rdev, r) ((rdev)->asic->surface.clear_reg((rdev), (r)))
@@ -2640,9 +2761,8 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 #define radeon_pm_finish(rdev) (rdev)->asic->pm.finish((rdev))
 #define radeon_pm_init_profile(rdev) (rdev)->asic->pm.init_profile((rdev))
 #define radeon_pm_get_dynpm_state(rdev) (rdev)->asic->pm.get_dynpm_state((rdev))
-#define radeon_pre_page_flip(rdev, crtc) (rdev)->asic->pflip.pre_page_flip((rdev), (crtc))
 #define radeon_page_flip(rdev, crtc, base) (rdev)->asic->pflip.page_flip((rdev), (crtc), (base))
-#define radeon_post_page_flip(rdev, crtc) (rdev)->asic->pflip.post_page_flip((rdev), (crtc))
+#define radeon_page_flip_pending(rdev, crtc) (rdev)->asic->pflip.page_flip_pending((rdev), (crtc))
 #define radeon_wait_for_vblank(rdev, crtc) (rdev)->asic->display.wait_for_vblank((rdev), (crtc))
 #define radeon_mc_wait_for_idle(rdev) (rdev)->asic->mc_wait_for_idle((rdev))
 #define radeon_get_xclk(rdev) (rdev)->asic->get_xclk((rdev))
@@ -2650,6 +2770,7 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 #define radeon_dpm_init(rdev) rdev->asic->dpm.init((rdev))
 #define radeon_dpm_setup_asic(rdev) rdev->asic->dpm.setup_asic((rdev))
 #define radeon_dpm_enable(rdev) rdev->asic->dpm.enable((rdev))
+#define radeon_dpm_late_enable(rdev) rdev->asic->dpm.late_enable((rdev))
 #define radeon_dpm_disable(rdev) rdev->asic->dpm.disable((rdev))
 #define radeon_dpm_pre_set_power_state(rdev) rdev->asic->dpm.pre_set_power_state((rdev))
 #define radeon_dpm_set_power_state(rdev) rdev->asic->dpm.set_power_state((rdev))
@@ -2668,6 +2789,7 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 /* Common functions */
 /* AGP */
 extern int radeon_gpu_reset(struct radeon_device *rdev);
+extern void radeon_pci_config_reset(struct radeon_device *rdev);
 extern void r600_set_bios_scratch_engine_hung(struct radeon_device *rdev, bool hung);
 extern void radeon_agp_disable(struct radeon_device *rdev);
 extern int radeon_modeset_init(struct radeon_device *rdev);
@@ -2700,19 +2822,26 @@ extern void radeon_program_register_sequence(struct radeon_device *rdev,
  */
 int radeon_vm_manager_init(struct radeon_device *rdev);
 void radeon_vm_manager_fini(struct radeon_device *rdev);
-void radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm);
+int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm);
 void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm);
-int radeon_vm_alloc_pt(struct radeon_device *rdev, struct radeon_vm *vm);
-void radeon_vm_add_to_lru(struct radeon_device *rdev, struct radeon_vm *vm);
+struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev,
+					  struct radeon_vm *vm,
+                                          struct list_head *head);
 struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
 				       struct radeon_vm *vm, int ring);
+void radeon_vm_flush(struct radeon_device *rdev,
+                     struct radeon_vm *vm,
+                     int ring);
 void radeon_vm_fence(struct radeon_device *rdev,
 		     struct radeon_vm *vm,
 		     struct radeon_fence *fence);
 uint64_t radeon_vm_map_gart(struct radeon_device *rdev, uint64_t addr);
+int radeon_vm_update_page_directory(struct radeon_device *rdev,
+				    struct radeon_vm *vm);
+int radeon_vm_clear_freed(struct radeon_device *rdev,
+			  struct radeon_vm *vm);
 int radeon_vm_bo_update(struct radeon_device *rdev,
-			struct radeon_vm *vm,
-			struct radeon_bo *bo,
+			struct radeon_bo_va *bo_va,
 			struct ttm_mem_reg *mem);
 void radeon_vm_bo_invalidate(struct radeon_device *rdev,
 			     struct radeon_bo *bo);
@@ -2725,13 +2854,19 @@ int radeon_vm_bo_set_addr(struct radeon_device *rdev,
 			  struct radeon_bo_va *bo_va,
 			  uint64_t offset,
 			  uint32_t flags);
-int radeon_vm_bo_rmv(struct radeon_device *rdev,
-		     struct radeon_bo_va *bo_va);
+void radeon_vm_bo_rmv(struct radeon_device *rdev,
+		      struct radeon_bo_va *bo_va);
 
 /* audio */
 void r600_audio_update_hdmi(struct work_struct *work);
 struct r600_audio_pin *r600_audio_get_pin(struct radeon_device *rdev);
 struct r600_audio_pin *dce6_audio_get_pin(struct radeon_device *rdev);
+void r600_audio_enable(struct radeon_device *rdev,
+		       struct r600_audio_pin *pin,
+		       bool enable);
+void dce6_audio_enable(struct radeon_device *rdev,
+		       struct r600_audio_pin *pin,
+		       bool enable);
 
 /*
  * R600 vram scratch functions
diff --git a/drivers/gpu/drm/radeon/radeon_acpi.c b/drivers/gpu/drm/radeon/radeon_acpi.c
index 98a9074b306..77e9d07c55b 100644
--- a/drivers/gpu/drm/radeon/radeon_acpi.c
+++ b/drivers/gpu/drm/radeon/radeon_acpi.c
@@ -25,18 +25,14 @@
 #include <linux/acpi.h>
 #include <linux/slab.h>
 #include <linux/power_supply.h>
-#include <acpi/acpi_drivers.h>
-#include <acpi/acpi_bus.h>
+#include <linux/vga_switcheroo.h>
 #include <acpi/video.h>
-
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 #include "radeon.h"
 #include "radeon_acpi.h"
 #include "atom.h"
 
-#include <linux/vga_switcheroo.h>
-
 #define ACPI_AC_CLASS           "ac_adapter"
 
 extern void radeon_pm_acpi_event_handler(struct radeon_device *rdev);
diff --git a/drivers/gpu/drm/radeon/radeon_agp.c b/drivers/gpu/drm/radeon/radeon_agp.c
index 42433344cb1..a9297b2c352 100644
--- a/drivers/gpu/drm/radeon/radeon_agp.c
+++ b/drivers/gpu/drm/radeon/radeon_agp.c
@@ -117,9 +117,6 @@ static struct radeon_agpmode_quirk radeon_agpmode_quirk_list[] = {
 	/* ATI Host Bridge / RV280 [M9+] Needs AGPMode 1 (phoronix forum) */
 	{ PCI_VENDOR_ID_ATI, 0xcbb2, PCI_VENDOR_ID_ATI, 0x5c61,
 		PCI_VENDOR_ID_SONY, 0x8175, 1},
-	/* HP Host Bridge / R300 [FireGL X1] Needs AGPMode 2 (fdo #7770) */
-	{ PCI_VENDOR_ID_HP, 0x122e, PCI_VENDOR_ID_ATI, 0x4e47,
-		PCI_VENDOR_ID_ATI, 0x0152, 2},
 	{ 0, 0, 0, 0, 0, 0, 0 },
 };
 #endif
diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c
index c0425bb6223..34b9aa9e3c0 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.c
+++ b/drivers/gpu/drm/radeon/radeon_asic.c
@@ -182,9 +182,9 @@ static struct radeon_asic_ring r100_gfx_ring = {
 	.ring_test = &r100_ring_test,
 	.ib_test = &r100_ib_test,
 	.is_lockup = &r100_gpu_is_lockup,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &r100_gfx_get_rptr,
+	.get_wptr = &r100_gfx_get_wptr,
+	.set_wptr = &r100_gfx_set_wptr,
 };
 
 static struct radeon_asic r100_asic = {
@@ -248,9 +248,8 @@ static struct radeon_asic r100_asic = {
 		.set_clock_gating = &radeon_legacy_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &r100_pre_page_flip,
 		.page_flip = &r100_page_flip,
-		.post_page_flip = &r100_post_page_flip,
+		.page_flip_pending = &r100_page_flip_pending,
 	},
 };
 
@@ -315,9 +314,8 @@ static struct radeon_asic r200_asic = {
 		.set_clock_gating = &radeon_legacy_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &r100_pre_page_flip,
 		.page_flip = &r100_page_flip,
-		.post_page_flip = &r100_post_page_flip,
+		.page_flip_pending = &r100_page_flip_pending,
 	},
 };
 
@@ -330,9 +328,9 @@ static struct radeon_asic_ring r300_gfx_ring = {
 	.ring_test = &r100_ring_test,
 	.ib_test = &r100_ib_test,
 	.is_lockup = &r100_gpu_is_lockup,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &r100_gfx_get_rptr,
+	.get_wptr = &r100_gfx_get_wptr,
+	.set_wptr = &r100_gfx_set_wptr,
 };
 
 static struct radeon_asic r300_asic = {
@@ -396,9 +394,8 @@ static struct radeon_asic r300_asic = {
 		.set_clock_gating = &radeon_legacy_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &r100_pre_page_flip,
 		.page_flip = &r100_page_flip,
-		.post_page_flip = &r100_post_page_flip,
+		.page_flip_pending = &r100_page_flip_pending,
 	},
 };
 
@@ -463,9 +460,8 @@ static struct radeon_asic r300_asic_pcie = {
 		.set_clock_gating = &radeon_legacy_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &r100_pre_page_flip,
 		.page_flip = &r100_page_flip,
-		.post_page_flip = &r100_post_page_flip,
+		.page_flip_pending = &r100_page_flip_pending,
 	},
 };
 
@@ -530,9 +526,8 @@ static struct radeon_asic r420_asic = {
 		.set_clock_gating = &radeon_atom_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &r100_pre_page_flip,
 		.page_flip = &r100_page_flip,
-		.post_page_flip = &r100_post_page_flip,
+		.page_flip_pending = &r100_page_flip_pending,
 	},
 };
 
@@ -597,9 +592,8 @@ static struct radeon_asic rs400_asic = {
 		.set_clock_gating = &radeon_legacy_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &r100_pre_page_flip,
 		.page_flip = &r100_page_flip,
-		.post_page_flip = &r100_post_page_flip,
+		.page_flip_pending = &r100_page_flip_pending,
 	},
 };
 
@@ -666,9 +660,8 @@ static struct radeon_asic rs600_asic = {
 		.set_clock_gating = &radeon_atom_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &rs600_pre_page_flip,
 		.page_flip = &rs600_page_flip,
-		.post_page_flip = &rs600_post_page_flip,
+		.page_flip_pending = &rs600_page_flip_pending,
 	},
 };
 
@@ -735,9 +728,8 @@ static struct radeon_asic rs690_asic = {
 		.set_clock_gating = &radeon_atom_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &rs600_pre_page_flip,
 		.page_flip = &rs600_page_flip,
-		.post_page_flip = &rs600_post_page_flip,
+		.page_flip_pending = &rs600_page_flip_pending,
 	},
 };
 
@@ -802,9 +794,8 @@ static struct radeon_asic rv515_asic = {
 		.set_clock_gating = &radeon_atom_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &rs600_pre_page_flip,
 		.page_flip = &rs600_page_flip,
-		.post_page_flip = &rs600_post_page_flip,
+		.page_flip_pending = &rs600_page_flip_pending,
 	},
 };
 
@@ -869,9 +860,8 @@ static struct radeon_asic r520_asic = {
 		.set_clock_gating = &radeon_atom_set_clock_gating,
 	},
 	.pflip = {
-		.pre_page_flip = &rs600_pre_page_flip,
 		.page_flip = &rs600_page_flip,
-		.post_page_flip = &rs600_post_page_flip,
+		.page_flip_pending = &rs600_page_flip_pending,
 	},
 };
 
@@ -883,9 +873,9 @@ static struct radeon_asic_ring r600_gfx_ring = {
 	.ring_test = &r600_ring_test,
 	.ib_test = &r600_ib_test,
 	.is_lockup = &r600_gfx_is_lockup,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &r600_gfx_get_rptr,
+	.get_wptr = &r600_gfx_get_wptr,
+	.set_wptr = &r600_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring r600_dma_ring = {
@@ -968,9 +958,8 @@ static struct radeon_asic r600_asic = {
 		.get_temperature = &rv6xx_get_temp,
 	},
 	.pflip = {
-		.pre_page_flip = &rs600_pre_page_flip,
 		.page_flip = &rs600_page_flip,
-		.post_page_flip = &rs600_post_page_flip,
+		.page_flip_pending = &rs600_page_flip_pending,
 	},
 };
 
@@ -1045,6 +1034,7 @@ static struct radeon_asic rv6xx_asic = {
 		.init = &rv6xx_dpm_init,
 		.setup_asic = &rv6xx_setup_asic,
 		.enable = &rv6xx_dpm_enable,
+		.late_enable = &r600_dpm_late_enable,
 		.disable = &rv6xx_dpm_disable,
 		.pre_set_power_state = &r600_dpm_pre_set_power_state,
 		.set_power_state = &rv6xx_dpm_set_power_state,
@@ -1058,9 +1048,8 @@ static struct radeon_asic rv6xx_asic = {
 		.force_performance_level = &rv6xx_dpm_force_performance_level,
 	},
 	.pflip = {
-		.pre_page_flip = &rs600_pre_page_flip,
 		.page_flip = &rs600_page_flip,
-		.post_page_flip = &rs600_post_page_flip,
+		.page_flip_pending = &rs600_page_flip_pending,
 	},
 };
 
@@ -1135,6 +1124,7 @@ static struct radeon_asic rs780_asic = {
 		.init = &rs780_dpm_init,
 		.setup_asic = &rs780_dpm_setup_asic,
 		.enable = &rs780_dpm_enable,
+		.late_enable = &r600_dpm_late_enable,
 		.disable = &rs780_dpm_disable,
 		.pre_set_power_state = &r600_dpm_pre_set_power_state,
 		.set_power_state = &rs780_dpm_set_power_state,
@@ -1148,9 +1138,8 @@ static struct radeon_asic rs780_asic = {
 		.force_performance_level = &rs780_dpm_force_performance_level,
 	},
 	.pflip = {
-		.pre_page_flip = &rs600_pre_page_flip,
 		.page_flip = &rs600_page_flip,
-		.post_page_flip = &rs600_post_page_flip,
+		.page_flip_pending = &rs600_page_flip_pending,
 	},
 };
 
@@ -1199,7 +1188,7 @@ static struct radeon_asic rv770_asic = {
 		.set_backlight_level = &atombios_set_backlight_level,
 		.get_backlight_level = &atombios_get_backlight_level,
 		.hdmi_enable = &r600_hdmi_enable,
-		.hdmi_setmode = &r600_hdmi_setmode,
+		.hdmi_setmode = &dce3_1_hdmi_setmode,
 	},
 	.copy = {
 		.blit = &r600_copy_cpdma,
@@ -1239,6 +1228,7 @@ static struct radeon_asic rv770_asic = {
 		.init = &rv770_dpm_init,
 		.setup_asic = &rv770_dpm_setup_asic,
 		.enable = &rv770_dpm_enable,
+		.late_enable = &rv770_dpm_late_enable,
 		.disable = &rv770_dpm_disable,
 		.pre_set_power_state = &r600_dpm_pre_set_power_state,
 		.set_power_state = &rv770_dpm_set_power_state,
@@ -1253,9 +1243,8 @@ static struct radeon_asic rv770_asic = {
 		.vblank_too_short = &rv770_dpm_vblank_too_short,
 	},
 	.pflip = {
-		.pre_page_flip = &rs600_pre_page_flip,
 		.page_flip = &rv770_page_flip,
-		.post_page_flip = &rs600_post_page_flip,
+		.page_flip_pending = &rv770_page_flip_pending,
 	},
 };
 
@@ -1267,9 +1256,9 @@ static struct radeon_asic_ring evergreen_gfx_ring = {
 	.ring_test = &r600_ring_test,
 	.ib_test = &r600_ib_test,
 	.is_lockup = &evergreen_gfx_is_lockup,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &r600_gfx_get_rptr,
+	.get_wptr = &r600_gfx_get_wptr,
+	.set_wptr = &r600_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring evergreen_dma_ring = {
@@ -1357,6 +1346,7 @@ static struct radeon_asic evergreen_asic = {
 		.init = &cypress_dpm_init,
 		.setup_asic = &cypress_dpm_setup_asic,
 		.enable = &cypress_dpm_enable,
+		.late_enable = &rv770_dpm_late_enable,
 		.disable = &cypress_dpm_disable,
 		.pre_set_power_state = &r600_dpm_pre_set_power_state,
 		.set_power_state = &cypress_dpm_set_power_state,
@@ -1371,9 +1361,8 @@ static struct radeon_asic evergreen_asic = {
 		.vblank_too_short = &cypress_dpm_vblank_too_short,
 	},
 	.pflip = {
-		.pre_page_flip = &evergreen_pre_page_flip,
 		.page_flip = &evergreen_page_flip,
-		.post_page_flip = &evergreen_post_page_flip,
+		.page_flip_pending = &evergreen_page_flip_pending,
 	},
 };
 
@@ -1449,6 +1438,7 @@ static struct radeon_asic sumo_asic = {
 		.init = &sumo_dpm_init,
 		.setup_asic = &sumo_dpm_setup_asic,
 		.enable = &sumo_dpm_enable,
+		.late_enable = &sumo_dpm_late_enable,
 		.disable = &sumo_dpm_disable,
 		.pre_set_power_state = &sumo_dpm_pre_set_power_state,
 		.set_power_state = &sumo_dpm_set_power_state,
@@ -1462,9 +1452,8 @@ static struct radeon_asic sumo_asic = {
 		.force_performance_level = &sumo_dpm_force_performance_level,
 	},
 	.pflip = {
-		.pre_page_flip = &evergreen_pre_page_flip,
 		.page_flip = &evergreen_page_flip,
-		.post_page_flip = &evergreen_post_page_flip,
+		.page_flip_pending = &evergreen_page_flip_pending,
 	},
 };
 
@@ -1540,6 +1529,7 @@ static struct radeon_asic btc_asic = {
 		.init = &btc_dpm_init,
 		.setup_asic = &btc_dpm_setup_asic,
 		.enable = &btc_dpm_enable,
+		.late_enable = &rv770_dpm_late_enable,
 		.disable = &btc_dpm_disable,
 		.pre_set_power_state = &btc_dpm_pre_set_power_state,
 		.set_power_state = &btc_dpm_set_power_state,
@@ -1549,14 +1539,13 @@ static struct radeon_asic btc_asic = {
 		.get_sclk = &btc_dpm_get_sclk,
 		.get_mclk = &btc_dpm_get_mclk,
 		.print_power_state = &rv770_dpm_print_power_state,
-		.debugfs_print_current_performance_level = &rv770_dpm_debugfs_print_current_performance_level,
+		.debugfs_print_current_performance_level = &btc_dpm_debugfs_print_current_performance_level,
 		.force_performance_level = &rv770_dpm_force_performance_level,
 		.vblank_too_short = &btc_dpm_vblank_too_short,
 	},
 	.pflip = {
-		.pre_page_flip = &evergreen_pre_page_flip,
 		.page_flip = &evergreen_page_flip,
-		.post_page_flip = &evergreen_post_page_flip,
+		.page_flip_pending = &evergreen_page_flip_pending,
 	},
 };
 
@@ -1570,9 +1559,9 @@ static struct radeon_asic_ring cayman_gfx_ring = {
 	.ib_test = &r600_ib_test,
 	.is_lockup = &cayman_gfx_is_lockup,
 	.vm_flush = &cayman_vm_flush,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &cayman_gfx_get_rptr,
+	.get_wptr = &cayman_gfx_get_wptr,
+	.set_wptr = &cayman_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring cayman_dma_ring = {
@@ -1585,9 +1574,9 @@ static struct radeon_asic_ring cayman_dma_ring = {
 	.ib_test = &r600_dma_ib_test,
 	.is_lockup = &cayman_dma_is_lockup,
 	.vm_flush = &cayman_dma_vm_flush,
-	.get_rptr = &r600_dma_get_rptr,
-	.get_wptr = &r600_dma_get_wptr,
-	.set_wptr = &r600_dma_set_wptr
+	.get_rptr = &cayman_dma_get_rptr,
+	.get_wptr = &cayman_dma_get_wptr,
+	.set_wptr = &cayman_dma_set_wptr
 };
 
 static struct radeon_asic_ring cayman_uvd_ring = {
@@ -1683,6 +1672,7 @@ static struct radeon_asic cayman_asic = {
 		.init = &ni_dpm_init,
 		.setup_asic = &ni_dpm_setup_asic,
 		.enable = &ni_dpm_enable,
+		.late_enable = &rv770_dpm_late_enable,
 		.disable = &ni_dpm_disable,
 		.pre_set_power_state = &ni_dpm_pre_set_power_state,
 		.set_power_state = &ni_dpm_set_power_state,
@@ -1697,9 +1687,8 @@ static struct radeon_asic cayman_asic = {
 		.vblank_too_short = &ni_dpm_vblank_too_short,
 	},
 	.pflip = {
-		.pre_page_flip = &evergreen_pre_page_flip,
 		.page_flip = &evergreen_page_flip,
-		.post_page_flip = &evergreen_post_page_flip,
+		.page_flip_pending = &evergreen_page_flip_pending,
 	},
 };
 
@@ -1783,6 +1772,7 @@ static struct radeon_asic trinity_asic = {
 		.init = &trinity_dpm_init,
 		.setup_asic = &trinity_dpm_setup_asic,
 		.enable = &trinity_dpm_enable,
+		.late_enable = &trinity_dpm_late_enable,
 		.disable = &trinity_dpm_disable,
 		.pre_set_power_state = &trinity_dpm_pre_set_power_state,
 		.set_power_state = &trinity_dpm_set_power_state,
@@ -1797,9 +1787,8 @@ static struct radeon_asic trinity_asic = {
 		.enable_bapm = &trinity_dpm_enable_bapm,
 	},
 	.pflip = {
-		.pre_page_flip = &evergreen_pre_page_flip,
 		.page_flip = &evergreen_page_flip,
-		.post_page_flip = &evergreen_post_page_flip,
+		.page_flip_pending = &evergreen_page_flip_pending,
 	},
 };
 
@@ -1813,9 +1802,9 @@ static struct radeon_asic_ring si_gfx_ring = {
 	.ib_test = &r600_ib_test,
 	.is_lockup = &si_gfx_is_lockup,
 	.vm_flush = &si_vm_flush,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &cayman_gfx_get_rptr,
+	.get_wptr = &cayman_gfx_get_wptr,
+	.set_wptr = &cayman_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring si_dma_ring = {
@@ -1828,9 +1817,9 @@ static struct radeon_asic_ring si_dma_ring = {
 	.ib_test = &r600_dma_ib_test,
 	.is_lockup = &si_dma_is_lockup,
 	.vm_flush = &si_dma_vm_flush,
-	.get_rptr = &r600_dma_get_rptr,
-	.get_wptr = &r600_dma_get_wptr,
-	.set_wptr = &r600_dma_set_wptr,
+	.get_rptr = &cayman_dma_get_rptr,
+	.get_wptr = &cayman_dma_get_wptr,
+	.set_wptr = &cayman_dma_set_wptr,
 };
 
 static struct radeon_asic si_asic = {
@@ -1913,6 +1902,7 @@ static struct radeon_asic si_asic = {
 		.init = &si_dpm_init,
 		.setup_asic = &si_dpm_setup_asic,
 		.enable = &si_dpm_enable,
+		.late_enable = &si_dpm_late_enable,
 		.disable = &si_dpm_disable,
 		.pre_set_power_state = &si_dpm_pre_set_power_state,
 		.set_power_state = &si_dpm_set_power_state,
@@ -1927,9 +1917,8 @@ static struct radeon_asic si_asic = {
 		.vblank_too_short = &ni_dpm_vblank_too_short,
 	},
 	.pflip = {
-		.pre_page_flip = &evergreen_pre_page_flip,
 		.page_flip = &evergreen_page_flip,
-		.post_page_flip = &evergreen_post_page_flip,
+		.page_flip_pending = &evergreen_page_flip_pending,
 	},
 };
 
@@ -1943,9 +1932,9 @@ static struct radeon_asic_ring ci_gfx_ring = {
 	.ib_test = &cik_ib_test,
 	.is_lockup = &cik_gfx_is_lockup,
 	.vm_flush = &cik_vm_flush,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &cik_gfx_get_rptr,
+	.get_wptr = &cik_gfx_get_wptr,
+	.set_wptr = &cik_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring ci_cp_ring = {
@@ -1958,9 +1947,9 @@ static struct radeon_asic_ring ci_cp_ring = {
 	.ib_test = &cik_ib_test,
 	.is_lockup = &cik_gfx_is_lockup,
 	.vm_flush = &cik_vm_flush,
-	.get_rptr = &cik_compute_ring_get_rptr,
-	.get_wptr = &cik_compute_ring_get_wptr,
-	.set_wptr = &cik_compute_ring_set_wptr,
+	.get_rptr = &cik_compute_get_rptr,
+	.get_wptr = &cik_compute_get_wptr,
+	.set_wptr = &cik_compute_set_wptr,
 };
 
 static struct radeon_asic_ring ci_dma_ring = {
@@ -1973,9 +1962,22 @@ static struct radeon_asic_ring ci_dma_ring = {
 	.ib_test = &cik_sdma_ib_test,
 	.is_lockup = &cik_sdma_is_lockup,
 	.vm_flush = &cik_dma_vm_flush,
-	.get_rptr = &r600_dma_get_rptr,
-	.get_wptr = &r600_dma_get_wptr,
-	.set_wptr = &r600_dma_set_wptr,
+	.get_rptr = &cik_sdma_get_rptr,
+	.get_wptr = &cik_sdma_get_wptr,
+	.set_wptr = &cik_sdma_set_wptr,
+};
+
+static struct radeon_asic_ring ci_vce_ring = {
+	.ib_execute = &radeon_vce_ib_execute,
+	.emit_fence = &radeon_vce_fence_emit,
+	.emit_semaphore = &radeon_vce_semaphore_emit,
+	.cs_parse = &radeon_vce_cs_parse,
+	.ring_test = &radeon_vce_ring_test,
+	.ib_test = &radeon_vce_ib_test,
+	.is_lockup = &radeon_ring_test_lockup,
+	.get_rptr = &vce_v1_0_get_rptr,
+	.get_wptr = &vce_v1_0_get_wptr,
+	.set_wptr = &vce_v1_0_set_wptr,
 };
 
 static struct radeon_asic ci_asic = {
@@ -2006,6 +2008,8 @@ static struct radeon_asic ci_asic = {
 		[R600_RING_TYPE_DMA_INDEX] = &ci_dma_ring,
 		[CAYMAN_RING_TYPE_DMA1_INDEX] = &ci_dma_ring,
 		[R600_RING_TYPE_UVD_INDEX] = &cayman_uvd_ring,
+		[TN_RING_TYPE_VCE1_INDEX] = &ci_vce_ring,
+		[TN_RING_TYPE_VCE2_INDEX] = &ci_vce_ring,
 	},
 	.irq = {
 		.set = &cik_irq_set,
@@ -2052,12 +2056,14 @@ static struct radeon_asic ci_asic = {
 		.set_pcie_lanes = NULL,
 		.set_clock_gating = NULL,
 		.set_uvd_clocks = &cik_set_uvd_clocks,
+		.set_vce_clocks = &cik_set_vce_clocks,
 		.get_temperature = &ci_get_temp,
 	},
 	.dpm = {
 		.init = &ci_dpm_init,
 		.setup_asic = &ci_dpm_setup_asic,
 		.enable = &ci_dpm_enable,
+		.late_enable = &ci_dpm_late_enable,
 		.disable = &ci_dpm_disable,
 		.pre_set_power_state = &ci_dpm_pre_set_power_state,
 		.set_power_state = &ci_dpm_set_power_state,
@@ -2073,9 +2079,8 @@ static struct radeon_asic ci_asic = {
 		.powergate_uvd = &ci_dpm_powergate_uvd,
 	},
 	.pflip = {
-		.pre_page_flip = &evergreen_pre_page_flip,
 		.page_flip = &evergreen_page_flip,
-		.post_page_flip = &evergreen_post_page_flip,
+		.page_flip_pending = &evergreen_page_flip_pending,
 	},
 };
 
@@ -2107,6 +2112,8 @@ static struct radeon_asic kv_asic = {
 		[R600_RING_TYPE_DMA_INDEX] = &ci_dma_ring,
 		[CAYMAN_RING_TYPE_DMA1_INDEX] = &ci_dma_ring,
 		[R600_RING_TYPE_UVD_INDEX] = &cayman_uvd_ring,
+		[TN_RING_TYPE_VCE1_INDEX] = &ci_vce_ring,
+		[TN_RING_TYPE_VCE2_INDEX] = &ci_vce_ring,
 	},
 	.irq = {
 		.set = &cik_irq_set,
@@ -2153,12 +2160,14 @@ static struct radeon_asic kv_asic = {
 		.set_pcie_lanes = NULL,
 		.set_clock_gating = NULL,
 		.set_uvd_clocks = &cik_set_uvd_clocks,
+		.set_vce_clocks = &cik_set_vce_clocks,
 		.get_temperature = &kv_get_temp,
 	},
 	.dpm = {
 		.init = &kv_dpm_init,
 		.setup_asic = &kv_dpm_setup_asic,
 		.enable = &kv_dpm_enable,
+		.late_enable = &kv_dpm_late_enable,
 		.disable = &kv_dpm_disable,
 		.pre_set_power_state = &kv_dpm_pre_set_power_state,
 		.set_power_state = &kv_dpm_set_power_state,
@@ -2174,9 +2183,8 @@ static struct radeon_asic kv_asic = {
 		.enable_bapm = &kv_dpm_enable_bapm,
 	},
 	.pflip = {
-		.pre_page_flip = &evergreen_pre_page_flip,
 		.page_flip = &evergreen_page_flip,
-		.post_page_flip = &evergreen_post_page_flip,
+		.page_flip_pending = &evergreen_page_flip_pending,
 	},
 };
 
@@ -2449,7 +2457,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 			rdev->cg_flags =
 				RADEON_CG_SUPPORT_GFX_MGCG |
 				RADEON_CG_SUPPORT_GFX_MGLS |
-				/*RADEON_CG_SUPPORT_GFX_CGCG |*/
+				RADEON_CG_SUPPORT_GFX_CGCG |
 				RADEON_CG_SUPPORT_GFX_CGLS |
 				RADEON_CG_SUPPORT_GFX_CGTS |
 				RADEON_CG_SUPPORT_GFX_CGTS_LS |
@@ -2468,7 +2476,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 			rdev->cg_flags =
 				RADEON_CG_SUPPORT_GFX_MGCG |
 				RADEON_CG_SUPPORT_GFX_MGLS |
-				/*RADEON_CG_SUPPORT_GFX_CGCG |*/
+				RADEON_CG_SUPPORT_GFX_CGCG |
 				RADEON_CG_SUPPORT_GFX_CGLS |
 				RADEON_CG_SUPPORT_GFX_CGTS |
 				RADEON_CG_SUPPORT_GFX_CP_LS |
@@ -2486,6 +2494,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 		break;
 	case CHIP_KAVERI:
 	case CHIP_KABINI:
+	case CHIP_MULLINS:
 		rdev->asic = &kv_asic;
 		/* set num crtcs */
 		if (rdev->family == CHIP_KAVERI) {
@@ -2493,7 +2502,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 			rdev->cg_flags =
 				RADEON_CG_SUPPORT_GFX_MGCG |
 				RADEON_CG_SUPPORT_GFX_MGLS |
-				/*RADEON_CG_SUPPORT_GFX_CGCG |*/
+				RADEON_CG_SUPPORT_GFX_CGCG |
 				RADEON_CG_SUPPORT_GFX_CGLS |
 				RADEON_CG_SUPPORT_GFX_CGTS |
 				RADEON_CG_SUPPORT_GFX_CGTS_LS |
@@ -2521,7 +2530,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 			rdev->cg_flags =
 				RADEON_CG_SUPPORT_GFX_MGCG |
 				RADEON_CG_SUPPORT_GFX_MGLS |
-				/*RADEON_CG_SUPPORT_GFX_CGCG |*/
+				RADEON_CG_SUPPORT_GFX_CGCG |
 				RADEON_CG_SUPPORT_GFX_CGLS |
 				RADEON_CG_SUPPORT_GFX_CGTS |
 				RADEON_CG_SUPPORT_GFX_CGTS_LS |
diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h
index c9fd97b5807..01e7c0ad8f0 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.h
+++ b/drivers/gpu/drm/radeon/radeon_asic.h
@@ -47,13 +47,6 @@ u8 atombios_get_backlight_level(struct radeon_encoder *radeon_encoder);
 void radeon_legacy_set_backlight_level(struct radeon_encoder *radeon_encoder, u8 level);
 u8 radeon_legacy_get_backlight_level(struct radeon_encoder *radeon_encoder);
 
-u32 radeon_ring_generic_get_rptr(struct radeon_device *rdev,
-				 struct radeon_ring *ring);
-u32 radeon_ring_generic_get_wptr(struct radeon_device *rdev,
-				 struct radeon_ring *ring);
-void radeon_ring_generic_set_wptr(struct radeon_device *rdev,
-				  struct radeon_ring *ring);
-
 /*
  * r100,rv100,rs100,rv200,rs200
  */
@@ -74,7 +67,8 @@ bool r100_gpu_is_lockup(struct radeon_device *rdev, struct radeon_ring *cp);
 int r100_asic_reset(struct radeon_device *rdev);
 u32 r100_get_vblank_counter(struct radeon_device *rdev, int crtc);
 void r100_pci_gart_tlb_flush(struct radeon_device *rdev);
-int r100_pci_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr);
+void r100_pci_gart_set_page(struct radeon_device *rdev, unsigned i,
+			    uint64_t addr);
 void r100_ring_start(struct radeon_device *rdev, struct radeon_ring *ring);
 int r100_irq_set(struct radeon_device *rdev);
 int r100_irq_process(struct radeon_device *rdev);
@@ -142,12 +136,19 @@ extern void r100_pm_prepare(struct radeon_device *rdev);
 extern void r100_pm_finish(struct radeon_device *rdev);
 extern void r100_pm_init_profile(struct radeon_device *rdev);
 extern void r100_pm_get_dynpm_state(struct radeon_device *rdev);
-extern void r100_pre_page_flip(struct radeon_device *rdev, int crtc);
-extern u32 r100_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base);
-extern void r100_post_page_flip(struct radeon_device *rdev, int crtc);
+extern void r100_page_flip(struct radeon_device *rdev, int crtc,
+			   u64 crtc_base);
+extern bool r100_page_flip_pending(struct radeon_device *rdev, int crtc);
 extern void r100_wait_for_vblank(struct radeon_device *rdev, int crtc);
 extern int r100_mc_wait_for_idle(struct radeon_device *rdev);
 
+u32 r100_gfx_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+u32 r100_gfx_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+void r100_gfx_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring);
+
 /*
  * r200,rv250,rs300,rv280
  */
@@ -171,7 +172,8 @@ extern void r300_fence_ring_emit(struct radeon_device *rdev,
 				struct radeon_fence *fence);
 extern int r300_cs_parse(struct radeon_cs_parser *p);
 extern void rv370_pcie_gart_tlb_flush(struct radeon_device *rdev);
-extern int rv370_pcie_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr);
+extern void rv370_pcie_gart_set_page(struct radeon_device *rdev, unsigned i,
+				     uint64_t addr);
 extern void rv370_set_pcie_lanes(struct radeon_device *rdev, int lanes);
 extern int rv370_get_pcie_lanes(struct radeon_device *rdev);
 extern void r300_set_reg_safe(struct radeon_device *rdev);
@@ -206,7 +208,8 @@ extern void rs400_fini(struct radeon_device *rdev);
 extern int rs400_suspend(struct radeon_device *rdev);
 extern int rs400_resume(struct radeon_device *rdev);
 void rs400_gart_tlb_flush(struct radeon_device *rdev);
-int rs400_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr);
+void rs400_gart_set_page(struct radeon_device *rdev, unsigned i,
+			 uint64_t addr);
 uint32_t rs400_mc_rreg(struct radeon_device *rdev, uint32_t reg);
 void rs400_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
 int rs400_gart_init(struct radeon_device *rdev);
@@ -229,7 +232,8 @@ int rs600_irq_process(struct radeon_device *rdev);
 void rs600_irq_disable(struct radeon_device *rdev);
 u32 rs600_get_vblank_counter(struct radeon_device *rdev, int crtc);
 void rs600_gart_tlb_flush(struct radeon_device *rdev);
-int rs600_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr);
+void rs600_gart_set_page(struct radeon_device *rdev, unsigned i,
+			 uint64_t addr);
 uint32_t rs600_mc_rreg(struct radeon_device *rdev, uint32_t reg);
 void rs600_mc_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
 void rs600_bandwidth_update(struct radeon_device *rdev);
@@ -241,9 +245,9 @@ void rs600_hpd_set_polarity(struct radeon_device *rdev,
 extern void rs600_pm_misc(struct radeon_device *rdev);
 extern void rs600_pm_prepare(struct radeon_device *rdev);
 extern void rs600_pm_finish(struct radeon_device *rdev);
-extern void rs600_pre_page_flip(struct radeon_device *rdev, int crtc);
-extern u32 rs600_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base);
-extern void rs600_post_page_flip(struct radeon_device *rdev, int crtc);
+extern void rs600_page_flip(struct radeon_device *rdev, int crtc,
+			    u64 crtc_base);
+extern bool rs600_page_flip_pending(struct radeon_device *rdev, int crtc);
 void rs600_set_safe_registers(struct radeon_device *rdev);
 extern void avivo_wait_for_vblank(struct radeon_device *rdev, int crtc);
 extern int rs600_mc_wait_for_idle(struct radeon_device *rdev);
@@ -368,6 +372,12 @@ int r600_mc_wait_for_idle(struct radeon_device *rdev);
 int r600_pcie_gart_init(struct radeon_device *rdev);
 void r600_scratch_init(struct radeon_device *rdev);
 int r600_init_microcode(struct radeon_device *rdev);
+u32 r600_gfx_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+u32 r600_gfx_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+void r600_gfx_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring);
 /* r600 irq */
 int r600_irq_process(struct radeon_device *rdev);
 int r600_irq_init(struct radeon_device *rdev);
@@ -381,6 +391,11 @@ void r600_rlc_stop(struct radeon_device *rdev);
 int r600_audio_init(struct radeon_device *rdev);
 struct r600_audio_pin r600_audio_status(struct radeon_device *rdev);
 void r600_audio_fini(struct radeon_device *rdev);
+void r600_audio_set_dto(struct drm_encoder *encoder, u32 clock);
+void r600_hdmi_update_avi_infoframe(struct drm_encoder *encoder, void *buffer,
+				    size_t size);
+void r600_hdmi_update_ACR(struct drm_encoder *encoder, uint32_t clock);
+void r600_hdmi_audio_workaround(struct drm_encoder *encoder);
 int r600_hdmi_buffer_status_changed(struct drm_encoder *encoder);
 void r600_hdmi_update_audio_settings(struct drm_encoder *encoder);
 void r600_hdmi_enable(struct drm_encoder *encoder, bool enable);
@@ -392,6 +407,7 @@ int rv6xx_get_temp(struct radeon_device *rdev);
 int r600_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
 int r600_dpm_pre_set_power_state(struct radeon_device *rdev);
 void r600_dpm_post_set_power_state(struct radeon_device *rdev);
+int r600_dpm_late_enable(struct radeon_device *rdev);
 /* r600 dma */
 uint32_t r600_dma_get_rptr(struct radeon_device *rdev,
 			   struct radeon_ring *ring);
@@ -440,7 +456,8 @@ void rv770_fini(struct radeon_device *rdev);
 int rv770_suspend(struct radeon_device *rdev);
 int rv770_resume(struct radeon_device *rdev);
 void rv770_pm_misc(struct radeon_device *rdev);
-u32 rv770_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base);
+void rv770_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base);
+bool rv770_page_flip_pending(struct radeon_device *rdev, int crtc);
 void r700_vram_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc);
 void r700_cp_stop(struct radeon_device *rdev);
 void r700_cp_fini(struct radeon_device *rdev);
@@ -451,9 +468,12 @@ int rv770_copy_dma(struct radeon_device *rdev,
 u32 rv770_get_xclk(struct radeon_device *rdev);
 int rv770_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
 int rv770_get_temp(struct radeon_device *rdev);
+/* hdmi */
+void dce3_1_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mode);
 /* rv7xx pm */
 int rv770_dpm_init(struct radeon_device *rdev);
 int rv770_dpm_enable(struct radeon_device *rdev);
+int rv770_dpm_late_enable(struct radeon_device *rdev);
 void rv770_dpm_disable(struct radeon_device *rdev);
 int rv770_dpm_set_power_state(struct radeon_device *rdev);
 void rv770_dpm_setup_asic(struct radeon_device *rdev);
@@ -505,9 +525,9 @@ extern void sumo_pm_init_profile(struct radeon_device *rdev);
 extern void btc_pm_init_profile(struct radeon_device *rdev);
 int sumo_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
 int evergreen_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
-extern void evergreen_pre_page_flip(struct radeon_device *rdev, int crtc);
-extern u32 evergreen_page_flip(struct radeon_device *rdev, int crtc, u64 crtc_base);
-extern void evergreen_post_page_flip(struct radeon_device *rdev, int crtc);
+extern void evergreen_page_flip(struct radeon_device *rdev, int crtc,
+				u64 crtc_base);
+extern bool evergreen_page_flip_pending(struct radeon_device *rdev, int crtc);
 extern void dce4_wait_for_vblank(struct radeon_device *rdev, int crtc);
 void evergreen_disable_interrupt_state(struct radeon_device *rdev);
 int evergreen_mc_wait_for_idle(struct radeon_device *rdev);
@@ -543,8 +563,11 @@ void btc_dpm_fini(struct radeon_device *rdev);
 u32 btc_dpm_get_sclk(struct radeon_device *rdev, bool low);
 u32 btc_dpm_get_mclk(struct radeon_device *rdev, bool low);
 bool btc_dpm_vblank_too_short(struct radeon_device *rdev);
+void btc_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
+						     struct seq_file *m);
 int sumo_dpm_init(struct radeon_device *rdev);
 int sumo_dpm_enable(struct radeon_device *rdev);
+int sumo_dpm_late_enable(struct radeon_device *rdev);
 void sumo_dpm_disable(struct radeon_device *rdev);
 int sumo_dpm_pre_set_power_state(struct radeon_device *rdev);
 int sumo_dpm_set_power_state(struct radeon_device *rdev);
@@ -591,6 +614,19 @@ void cayman_dma_vm_set_page(struct radeon_device *rdev,
 
 void cayman_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
 
+u32 cayman_gfx_get_rptr(struct radeon_device *rdev,
+			struct radeon_ring *ring);
+u32 cayman_gfx_get_wptr(struct radeon_device *rdev,
+			struct radeon_ring *ring);
+void cayman_gfx_set_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring);
+uint32_t cayman_dma_get_rptr(struct radeon_device *rdev,
+			     struct radeon_ring *ring);
+uint32_t cayman_dma_get_wptr(struct radeon_device *rdev,
+			     struct radeon_ring *ring);
+void cayman_dma_set_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring);
+
 int ni_dpm_init(struct radeon_device *rdev);
 void ni_dpm_setup_asic(struct radeon_device *rdev);
 int ni_dpm_enable(struct radeon_device *rdev);
@@ -610,6 +646,7 @@ int ni_dpm_force_performance_level(struct radeon_device *rdev,
 bool ni_dpm_vblank_too_short(struct radeon_device *rdev);
 int trinity_dpm_init(struct radeon_device *rdev);
 int trinity_dpm_enable(struct radeon_device *rdev);
+int trinity_dpm_late_enable(struct radeon_device *rdev);
 void trinity_dpm_disable(struct radeon_device *rdev);
 int trinity_dpm_pre_set_power_state(struct radeon_device *rdev);
 int trinity_dpm_set_power_state(struct radeon_device *rdev);
@@ -669,6 +706,7 @@ int si_get_temp(struct radeon_device *rdev);
 int si_dpm_init(struct radeon_device *rdev);
 void si_dpm_setup_asic(struct radeon_device *rdev);
 int si_dpm_enable(struct radeon_device *rdev);
+int si_dpm_late_enable(struct radeon_device *rdev);
 void si_dpm_disable(struct radeon_device *rdev);
 int si_dpm_pre_set_power_state(struct radeon_device *rdev);
 int si_dpm_set_power_state(struct radeon_device *rdev);
@@ -691,6 +729,7 @@ u32 cik_get_xclk(struct radeon_device *rdev);
 uint32_t cik_pciep_rreg(struct radeon_device *rdev, uint32_t reg);
 void cik_pciep_wreg(struct radeon_device *rdev, uint32_t reg, uint32_t v);
 int cik_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
+int cik_set_vce_clocks(struct radeon_device *rdev, u32 evclk, u32 ecclk);
 void cik_sdma_fence_ring_emit(struct radeon_device *rdev,
 			      struct radeon_fence *fence);
 bool cik_sdma_semaphore_ring_emit(struct radeon_device *rdev,
@@ -739,17 +778,30 @@ void cik_sdma_vm_set_page(struct radeon_device *rdev,
 			  uint32_t incr, uint32_t flags);
 void cik_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
 int cik_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib);
-u32 cik_compute_ring_get_rptr(struct radeon_device *rdev,
-			      struct radeon_ring *ring);
-u32 cik_compute_ring_get_wptr(struct radeon_device *rdev,
-			      struct radeon_ring *ring);
-void cik_compute_ring_set_wptr(struct radeon_device *rdev,
-			       struct radeon_ring *ring);
+u32 cik_gfx_get_rptr(struct radeon_device *rdev,
+		     struct radeon_ring *ring);
+u32 cik_gfx_get_wptr(struct radeon_device *rdev,
+		     struct radeon_ring *ring);
+void cik_gfx_set_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+u32 cik_compute_get_rptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring);
+u32 cik_compute_get_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring);
+void cik_compute_set_wptr(struct radeon_device *rdev,
+			  struct radeon_ring *ring);
+u32 cik_sdma_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+u32 cik_sdma_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+void cik_sdma_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring);
 int ci_get_temp(struct radeon_device *rdev);
 int kv_get_temp(struct radeon_device *rdev);
 
 int ci_dpm_init(struct radeon_device *rdev);
 int ci_dpm_enable(struct radeon_device *rdev);
+int ci_dpm_late_enable(struct radeon_device *rdev);
 void ci_dpm_disable(struct radeon_device *rdev);
 int ci_dpm_pre_set_power_state(struct radeon_device *rdev);
 int ci_dpm_set_power_state(struct radeon_device *rdev);
@@ -770,6 +822,7 @@ void ci_dpm_powergate_uvd(struct radeon_device *rdev, bool gate);
 
 int kv_dpm_init(struct radeon_device *rdev);
 int kv_dpm_enable(struct radeon_device *rdev);
+int kv_dpm_late_enable(struct radeon_device *rdev);
 void kv_dpm_disable(struct radeon_device *rdev);
 int kv_dpm_pre_set_power_state(struct radeon_device *rdev);
 int kv_dpm_set_power_state(struct radeon_device *rdev);
@@ -823,4 +876,17 @@ bool uvd_v3_1_semaphore_emit(struct radeon_device *rdev,
 /* uvd v4.2 */
 int uvd_v4_2_resume(struct radeon_device *rdev);
 
+/* vce v1.0 */
+uint32_t vce_v1_0_get_rptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring);
+uint32_t vce_v1_0_get_wptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring);
+void vce_v1_0_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring);
+int vce_v1_0_init(struct radeon_device *rdev);
+int vce_v1_0_start(struct radeon_device *rdev);
+
+/* vce v2.0 */
+int vce_v2_0_resume(struct radeon_device *rdev);
+
 #endif
diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c
index 5c39bf7c3d8..173f378428a 100644
--- a/drivers/gpu/drm/radeon/radeon_atombios.c
+++ b/drivers/gpu/drm/radeon/radeon_atombios.c
@@ -30,27 +30,10 @@
 #include "atom.h"
 #include "atom-bits.h"
 
-/* from radeon_encoder.c */
-extern uint32_t
-radeon_get_encoder_enum(struct drm_device *dev, uint32_t supported_device,
-			uint8_t dac);
-extern void radeon_link_encoder_connector(struct drm_device *dev);
 extern void
 radeon_add_atom_encoder(struct drm_device *dev, uint32_t encoder_enum,
 			uint32_t supported_device, u16 caps);
 
-/* from radeon_connector.c */
-extern void
-radeon_add_atom_connector(struct drm_device *dev,
-			  uint32_t connector_id,
-			  uint32_t supported_device,
-			  int connector_type,
-			  struct radeon_i2c_bus_rec *i2c_bus,
-			  uint32_t igp_lane_info,
-			  uint16_t connector_object_id,
-			  struct radeon_hpd *hpd,
-			  struct radeon_router *router);
-
 /* from radeon_legacy_encoder.c */
 extern void
 radeon_add_legacy_encoder(struct drm_device *dev, uint32_t encoder_enum,
@@ -1244,11 +1227,19 @@ bool radeon_atom_get_clock_info(struct drm_device *dev)
 			rdev->clock.default_dispclk =
 				le32_to_cpu(firmware_info->info_21.ulDefaultDispEngineClkFreq);
 			if (rdev->clock.default_dispclk == 0) {
-				if (ASIC_IS_DCE5(rdev))
+				if (ASIC_IS_DCE6(rdev))
+					rdev->clock.default_dispclk = 60000; /* 600 Mhz */
+				else if (ASIC_IS_DCE5(rdev))
 					rdev->clock.default_dispclk = 54000; /* 540 Mhz */
 				else
 					rdev->clock.default_dispclk = 60000; /* 600 Mhz */
 			}
+			/* set a reasonable default for DP */
+			if (ASIC_IS_DCE6(rdev) && (rdev->clock.default_dispclk < 53900)) {
+				DRM_INFO("Changing default dispclk from %dMhz to 600Mhz\n",
+					 rdev->clock.default_dispclk / 100);
+				rdev->clock.default_dispclk = 60000;
+			}
 			rdev->clock.dp_extclk =
 				le16_to_cpu(firmware_info->info_21.usUniphyDPModeExtClkFreq);
 			rdev->clock.current_dispclk = rdev->clock.default_dispclk;
@@ -1528,6 +1519,7 @@ bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev,
 						le16_to_cpu(ss_assign->v1.usSpreadSpectrumPercentage);
 					ss->type = ss_assign->v1.ucSpreadSpectrumMode;
 					ss->rate = le16_to_cpu(ss_assign->v1.usSpreadRateInKhz);
+					ss->percentage_divider = 100;
 					return true;
 				}
 				ss_assign = (union asic_ss_assignment *)
@@ -1545,6 +1537,7 @@ bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev,
 						le16_to_cpu(ss_assign->v2.usSpreadSpectrumPercentage);
 					ss->type = ss_assign->v2.ucSpreadSpectrumMode;
 					ss->rate = le16_to_cpu(ss_assign->v2.usSpreadRateIn10Hz);
+					ss->percentage_divider = 100;
 					if ((crev == 2) &&
 					    ((id == ASIC_INTERNAL_ENGINE_SS) ||
 					     (id == ASIC_INTERNAL_MEMORY_SS)))
@@ -1566,6 +1559,11 @@ bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev,
 						le16_to_cpu(ss_assign->v3.usSpreadSpectrumPercentage);
 					ss->type = ss_assign->v3.ucSpreadSpectrumMode;
 					ss->rate = le16_to_cpu(ss_assign->v3.usSpreadRateIn10Hz);
+					if (ss_assign->v3.ucSpreadSpectrumMode &
+					    SS_MODE_V3_PERCENTAGE_DIV_BY_1000_MASK)
+						ss->percentage_divider = 1000;
+					else
+						ss->percentage_divider = 100;
 					if ((id == ASIC_INTERNAL_ENGINE_SS) ||
 					    (id == ASIC_INTERNAL_MEMORY_SS))
 						ss->rate /= 100;
@@ -1809,7 +1807,8 @@ bool radeon_atom_get_tv_timings(struct radeon_device *rdev, int index,
 		if (misc & ATOM_DOUBLE_CLOCK_MODE)
 			mode->flags |= DRM_MODE_FLAG_DBLSCAN;
 
-		mode->clock = le16_to_cpu(tv_info->aModeTimings[index].usPixelClock) * 10;
+		mode->crtc_clock = mode->clock =
+			le16_to_cpu(tv_info->aModeTimings[index].usPixelClock) * 10;
 
 		if (index == 1) {
 			/* PAL timings appear to have wrong values for totals */
@@ -1852,7 +1851,8 @@ bool radeon_atom_get_tv_timings(struct radeon_device *rdev, int index,
 		if (misc & ATOM_DOUBLE_CLOCK_MODE)
 			mode->flags |= DRM_MODE_FLAG_DBLSCAN;
 
-		mode->clock = le16_to_cpu(dtd_timings->usPixClk) * 10;
+		mode->crtc_clock = mode->clock =
+			le16_to_cpu(dtd_timings->usPixClk) * 10;
 		break;
 	}
 	return true;
@@ -3884,16 +3884,18 @@ int radeon_atom_init_mc_reg_table(struct radeon_device *rdev,
 							((u8 *)format + sizeof(ATOM_INIT_REG_INDEX_FORMAT));
 					}
 					reg_table->last = i;
-					while ((*(u32 *)reg_data != END_OF_REG_DATA_BLOCK) &&
+					while ((le32_to_cpu(*(u32 *)reg_data) != END_OF_REG_DATA_BLOCK) &&
 					       (num_ranges < VBIOS_MAX_AC_TIMING_ENTRIES)) {
-						t_mem_id = (u8)((*(u32 *)reg_data & MEM_ID_MASK) >> MEM_ID_SHIFT);
+						t_mem_id = (u8)((le32_to_cpu(*(u32 *)reg_data) & MEM_ID_MASK)
+								>> MEM_ID_SHIFT);
 						if (module_index == t_mem_id) {
 							reg_table->mc_reg_table_entry[num_ranges].mclk_max =
-								(u32)((*(u32 *)reg_data & CLOCK_RANGE_MASK) >> CLOCK_RANGE_SHIFT);
+								(u32)((le32_to_cpu(*(u32 *)reg_data) & CLOCK_RANGE_MASK)
+								      >> CLOCK_RANGE_SHIFT);
 							for (i = 0, j = 1; i < reg_table->last; i++) {
 								if ((reg_table->mc_reg_address[i].pre_reg_data & LOW_NIBBLE_MASK) == DATA_FROM_TABLE) {
 									reg_table->mc_reg_table_entry[num_ranges].mc_data[i] =
-										(u32)*((u32 *)reg_data + j);
+										(u32)le32_to_cpu(*((u32 *)reg_data + j));
 									j++;
 								} else if ((reg_table->mc_reg_address[i].pre_reg_data & LOW_NIBBLE_MASK) == DATA_EQU_PREV) {
 									reg_table->mc_reg_table_entry[num_ranges].mc_data[i] =
@@ -3905,7 +3907,7 @@ int radeon_atom_init_mc_reg_table(struct radeon_device *rdev,
 						reg_data = (ATOM_MEMORY_SETTING_DATA_BLOCK *)
 							((u8 *)reg_data + le16_to_cpu(reg_block->usRegDataBlkSize));
 					}
-					if (*(u32 *)reg_data != END_OF_REG_DATA_BLOCK)
+					if (le32_to_cpu(*(u32 *)reg_data) != END_OF_REG_DATA_BLOCK)
 						return -EINVAL;
 					reg_table->num_entries = num_ranges;
 				} else
@@ -3944,6 +3946,10 @@ void radeon_atom_initialize_bios_scratch_regs(struct drm_device *dev)
 	/* tell the bios not to handle mode switching */
 	bios_6_scratch |= ATOM_S6_ACC_BLOCK_DISPLAY_SWITCH;
 
+	/* clear the vbios dpms state */
+	if (ASIC_IS_DCE4(rdev))
+		bios_2_scratch &= ~ATOM_S2_DEVICE_DPMS_STATE;
+
 	if (rdev->family >= CHIP_R600) {
 		WREG32(R600_BIOS_2_SCRATCH, bios_2_scratch);
 		WREG32(R600_BIOS_6_SCRATCH, bios_6_scratch);
diff --git a/drivers/gpu/drm/radeon/radeon_atpx_handler.c b/drivers/gpu/drm/radeon/radeon_atpx_handler.c
index 485848f889f..a9fb0d016d3 100644
--- a/drivers/gpu/drm/radeon/radeon_atpx_handler.c
+++ b/drivers/gpu/drm/radeon/radeon_atpx_handler.c
@@ -59,7 +59,7 @@ struct atpx_mux {
 	u16 mux;
 } __packed;
 
-bool radeon_is_px(void) {
+bool radeon_has_atpx(void) {
 	return radeon_atpx_priv.atpx_detected;
 }
 
@@ -219,7 +219,8 @@ static int radeon_atpx_verify_interface(struct radeon_atpx *atpx)
 	memcpy(&output, info->buffer.pointer, size);
 
 	/* TODO: check version? */
-	printk("ATPX version %u\n", output.version);
+	printk("ATPX version %u, functions 0x%08x\n",
+	       output.version, output.function_bits);
 
 	radeon_atpx_parse_functions(&atpx->functions, output.function_bits);
 
@@ -527,6 +528,13 @@ static bool radeon_atpx_detect(void)
 		has_atpx |= (radeon_atpx_pci_probe_handle(pdev) == true);
 	}
 
+	/* some newer PX laptops mark the dGPU as a non-VGA display device */
+	while ((pdev = pci_get_class(PCI_CLASS_DISPLAY_OTHER << 8, pdev)) != NULL) {
+		vga_count++;
+
+		has_atpx |= (radeon_atpx_pci_probe_handle(pdev) == true);
+	}
+
 	if (has_atpx && vga_count == 2) {
 		acpi_get_name(radeon_atpx_priv.atpx.handle, ACPI_FULL_PATHNAME, &buffer);
 		printk(KERN_INFO "VGA switcheroo: detected switching method %s handle\n",
diff --git a/drivers/gpu/drm/radeon/radeon_bios.c b/drivers/gpu/drm/radeon/radeon_bios.c
index b3633d9a531..6a03624fada 100644
--- a/drivers/gpu/drm/radeon/radeon_bios.c
+++ b/drivers/gpu/drm/radeon/radeon_bios.c
@@ -196,6 +196,20 @@ static bool radeon_atrm_get_bios(struct radeon_device *rdev)
 		}
 	}
 
+	if (!found) {
+		while ((pdev = pci_get_class(PCI_CLASS_DISPLAY_OTHER << 8, pdev)) != NULL) {
+			dhandle = ACPI_HANDLE(&pdev->dev);
+			if (!dhandle)
+				continue;
+
+			status = acpi_get_handle(dhandle, "ATRM", &atrm_handle);
+			if (!ACPI_FAILURE(status)) {
+				found = true;
+				break;
+			}
+		}
+	}
+
 	if (!found)
 		return false;
 
@@ -612,7 +626,7 @@ static bool radeon_acpi_vfct_bios(struct radeon_device *rdev)
 	    vhdr->DeviceID != rdev->pdev->device) {
 		DRM_INFO("ACPI VFCT table is not for this card\n");
 		goto out_unmap;
-	};
+	}
 
 	if (vfct->VBIOSImageOffset + sizeof(VFCT_IMAGE_HEADER) + vhdr->ImageLength > tbl_size) {
 		DRM_ERROR("ACPI VFCT image truncated\n");
diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c
index 68ce3605601..6651177110f 100644
--- a/drivers/gpu/drm/radeon/radeon_combios.c
+++ b/drivers/gpu/drm/radeon/radeon_combios.c
@@ -37,22 +37,6 @@
 #include <asm/pci-bridge.h>
 #endif /* CONFIG_PPC_PMAC */
 
-/* from radeon_encoder.c */
-extern uint32_t
-radeon_get_encoder_enum(struct drm_device *dev, uint32_t supported_device,
-			uint8_t dac);
-extern void radeon_link_encoder_connector(struct drm_device *dev);
-
-/* from radeon_connector.c */
-extern void
-radeon_add_legacy_connector(struct drm_device *dev,
-			    uint32_t connector_id,
-			    uint32_t supported_device,
-			    int connector_type,
-			    struct radeon_i2c_bus_rec *i2c_bus,
-			    uint16_t connector_object_id,
-			    struct radeon_hpd *hpd);
-
 /* from radeon_legacy_encoder.c */
 extern void
 radeon_add_legacy_encoder(struct drm_device *dev, uint32_t encoder_enum,
diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c
index 20a768ac89a..44831197e82 100644
--- a/drivers/gpu/drm/radeon/radeon_connectors.c
+++ b/drivers/gpu/drm/radeon/radeon_connectors.c
@@ -33,15 +33,6 @@
 
 #include <linux/pm_runtime.h>
 
-extern void
-radeon_combios_connected_scratch_regs(struct drm_connector *connector,
-				      struct drm_encoder *encoder,
-				      bool connected);
-extern void
-radeon_atombios_connected_scratch_regs(struct drm_connector *connector,
-				       struct drm_encoder *encoder,
-				       bool connected);
-
 void radeon_connector_hotplug(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
@@ -57,6 +48,7 @@ void radeon_connector_hotplug(struct drm_connector *connector)
 	radeon_hpd_set_polarity(rdev, radeon_connector->hpd.hpd);
 
 	/* if the connector is already off, don't turn it back on */
+	/* FIXME: This access isn't protected by any locks. */
 	if (connector->dpms != DRM_MODE_DPMS_ON)
 		return;
 
@@ -98,7 +90,7 @@ static void radeon_property_change_mode(struct drm_encoder *encoder)
 
 	if (crtc && crtc->enabled) {
 		drm_crtc_helper_set_mode(crtc, &crtc->mode,
-					 crtc->x, crtc->y, crtc->fb);
+					 crtc->x, crtc->y, crtc->primary->fb);
 	}
 }
 
@@ -109,6 +101,7 @@ int radeon_get_monitor_bpc(struct drm_connector *connector)
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
 	struct radeon_connector_atom_dig *dig_connector;
 	int bpc = 8;
+	int mode_clock, max_tmds_clock;
 
 	switch (connector->connector_type) {
 	case DRM_MODE_CONNECTOR_DVII:
@@ -154,6 +147,64 @@ int radeon_get_monitor_bpc(struct drm_connector *connector)
 		}
 		break;
 	}
+
+	if (drm_detect_hdmi_monitor(radeon_connector->edid)) {
+		/* hdmi deep color only implemented on DCE4+ */
+		if ((bpc > 8) && !ASIC_IS_DCE4(rdev)) {
+			DRM_DEBUG("%s: HDMI deep color %d bpc unsupported. Using 8 bpc.\n",
+					  connector->name, bpc);
+			bpc = 8;
+		}
+
+		/*
+		 * Pre DCE-8 hw can't handle > 12 bpc, and more than 12 bpc doesn't make
+		 * much sense without support for > 12 bpc framebuffers. RGB 4:4:4 at
+		 * 12 bpc is always supported on hdmi deep color sinks, as this is
+		 * required by the HDMI-1.3 spec. Clamp to a safe 12 bpc maximum.
+		 */
+		if (bpc > 12) {
+			DRM_DEBUG("%s: HDMI deep color %d bpc unsupported. Using 12 bpc.\n",
+					  connector->name, bpc);
+			bpc = 12;
+		}
+
+		/* Any defined maximum tmds clock limit we must not exceed? */
+		if (connector->max_tmds_clock > 0) {
+			/* mode_clock is clock in kHz for mode to be modeset on this connector */
+			mode_clock = radeon_connector->pixelclock_for_modeset;
+
+			/* Maximum allowable input clock in kHz */
+			max_tmds_clock = connector->max_tmds_clock * 1000;
+
+			DRM_DEBUG("%s: hdmi mode dotclock %d kHz, max tmds input clock %d kHz.\n",
+					  connector->name, mode_clock, max_tmds_clock);
+
+			/* Check if bpc is within clock limit. Try to degrade gracefully otherwise */
+			if ((bpc == 12) && (mode_clock * 3/2 > max_tmds_clock)) {
+				if ((connector->display_info.edid_hdmi_dc_modes & DRM_EDID_HDMI_DC_30) &&
+					(mode_clock * 5/4 <= max_tmds_clock))
+					bpc = 10;
+				else
+					bpc = 8;
+
+				DRM_DEBUG("%s: HDMI deep color 12 bpc exceeds max tmds clock. Using %d bpc.\n",
+						  connector->name, bpc);
+			}
+
+			if ((bpc == 10) && (mode_clock * 5/4 > max_tmds_clock)) {
+				bpc = 8;
+				DRM_DEBUG("%s: HDMI deep color 10 bpc exceeds max tmds clock. Using %d bpc.\n",
+						  connector->name, bpc);
+			}
+		}
+	}
+
+	if ((radeon_deep_color == 0) && (bpc > 8))
+		bpc = 8;
+
+	DRM_DEBUG("%s: Display bpc=%d, returned bpc=%d\n",
+			  connector->name, connector->display_info.bpc, bpc);
+
 	return bpc;
 }
 
@@ -269,13 +320,17 @@ radeon_connector_analog_encoder_conflict_solve(struct drm_connector *connector,
 					continue;
 
 				if (priority == true) {
-					DRM_DEBUG_KMS("1: conflicting encoders switching off %s\n", drm_get_connector_name(conflict));
-					DRM_DEBUG_KMS("in favor of %s\n", drm_get_connector_name(connector));
+					DRM_DEBUG_KMS("1: conflicting encoders switching off %s\n",
+						      conflict->name);
+					DRM_DEBUG_KMS("in favor of %s\n",
+						      connector->name);
 					conflict->status = connector_status_disconnected;
 					radeon_connector_update_scratch_regs(conflict, connector_status_disconnected);
 				} else {
-					DRM_DEBUG_KMS("2: conflicting encoders switching off %s\n", drm_get_connector_name(connector));
-					DRM_DEBUG_KMS("in favor of %s\n", drm_get_connector_name(conflict));
+					DRM_DEBUG_KMS("2: conflicting encoders switching off %s\n",
+						      connector->name);
+					DRM_DEBUG_KMS("in favor of %s\n",
+						      conflict->name);
 					current_status = connector_status_disconnected;
 				}
 				break;
@@ -796,7 +851,7 @@ radeon_vga_detect(struct drm_connector *connector, bool force)
 
 		if (!radeon_connector->edid) {
 			DRM_ERROR("%s: probed a monitor but no|invalid EDID\n",
-					drm_get_connector_name(connector));
+					connector->name);
 			ret = connector_status_connected;
 		} else {
 			radeon_connector->use_digital = !!(radeon_connector->edid->input & DRM_EDID_INPUT_DIGITAL);
@@ -1019,12 +1074,13 @@ radeon_dvi_detect(struct drm_connector *connector, bool force)
 
 		if (!radeon_connector->edid) {
 			DRM_ERROR("%s: probed a monitor but no|invalid EDID\n",
-					drm_get_connector_name(connector));
+					connector->name);
 			/* rs690 seems to have a problem with connectors not existing and always
 			 * return a block of 0's. If we see this just stop polling on this output */
 			if ((rdev->family == CHIP_RS690 || rdev->family == CHIP_RS740) && radeon_connector->base.null_edid_counter) {
 				ret = connector_status_disconnected;
-				DRM_ERROR("%s: detected RS690 floating bus bug, stopping ddc detect\n", drm_get_connector_name(connector));
+				DRM_ERROR("%s: detected RS690 floating bus bug, stopping ddc detect\n",
+					  connector->name);
 				radeon_connector->ddc_bus = NULL;
 			} else {
 				ret = connector_status_connected;
@@ -1235,17 +1291,15 @@ static int radeon_dvi_mode_valid(struct drm_connector *connector,
 		    (radeon_connector->connector_object_id == CONNECTOR_OBJECT_ID_DUAL_LINK_DVI_D) ||
 		    (radeon_connector->connector_object_id == CONNECTOR_OBJECT_ID_HDMI_TYPE_B))
 			return MODE_OK;
-		else if (radeon_connector->connector_object_id == CONNECTOR_OBJECT_ID_HDMI_TYPE_A) {
-			if (ASIC_IS_DCE6(rdev)) {
-				/* HDMI 1.3+ supports max clock of 340 Mhz */
-				if (mode->clock > 340000)
-					return MODE_CLOCK_HIGH;
-				else
-					return MODE_OK;
-			} else
+		else if (ASIC_IS_DCE6(rdev) && drm_detect_hdmi_monitor(radeon_connector->edid)) {
+			/* HDMI 1.3+ supports max clock of 340 Mhz */
+			if (mode->clock > 340000)
 				return MODE_CLOCK_HIGH;
-		} else
+			else
+				return MODE_OK;
+		} else {
 			return MODE_CLOCK_HIGH;
+		}
 	}
 
 	/* check against the max pixel clock */
@@ -1270,21 +1324,6 @@ static const struct drm_connector_funcs radeon_dvi_connector_funcs = {
 	.force = radeon_dvi_force,
 };
 
-static void radeon_dp_connector_destroy(struct drm_connector *connector)
-{
-	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
-	struct radeon_connector_atom_dig *radeon_dig_connector = radeon_connector->con_priv;
-
-	if (radeon_connector->edid)
-		kfree(radeon_connector->edid);
-	if (radeon_dig_connector->dp_i2c_bus)
-		radeon_i2c_destroy(radeon_dig_connector->dp_i2c_bus);
-	kfree(radeon_connector->con_priv);
-	drm_sysfs_connector_remove(connector);
-	drm_connector_cleanup(connector);
-	kfree(connector);
-}
-
 static int radeon_dp_get_modes(struct drm_connector *connector)
 {
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
@@ -1411,7 +1450,7 @@ bool radeon_connector_is_dp12_capable(struct drm_connector *connector)
 	struct radeon_device *rdev = dev->dev_private;
 
 	if (ASIC_IS_DCE5(rdev) &&
-	    (rdev->clock.dp_extclk >= 53900) &&
+	    (rdev->clock.default_dispclk >= 53900) &&
 	    radeon_connector_encoder_is_hbr2(connector)) {
 		return true;
 	}
@@ -1511,6 +1550,8 @@ out:
 static int radeon_dp_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
+	struct drm_device *dev = connector->dev;
+	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
 	struct radeon_connector_atom_dig *radeon_dig_connector = radeon_connector->con_priv;
 
@@ -1541,14 +1582,23 @@ static int radeon_dp_mode_valid(struct drm_connector *connector,
 					return MODE_PANEL;
 			}
 		}
-		return MODE_OK;
 	} else {
 		if ((radeon_dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_DISPLAYPORT) ||
-		    (radeon_dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_eDP))
+		    (radeon_dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_eDP)) {
 			return radeon_dp_mode_valid_helper(connector, mode);
-		else
-			return MODE_OK;
+		} else {
+			if (ASIC_IS_DCE6(rdev) && drm_detect_hdmi_monitor(radeon_connector->edid)) {
+				/* HDMI 1.3+ supports max clock of 340 Mhz */
+				if (mode->clock > 340000)
+					return MODE_CLOCK_HIGH;
+			} else {
+				if (mode->clock > 165000)
+					return MODE_CLOCK_HIGH;
+			}
+		}
 	}
+
+	return MODE_OK;
 }
 
 static const struct drm_connector_helper_funcs radeon_dp_connector_helper_funcs = {
@@ -1562,7 +1612,7 @@ static const struct drm_connector_funcs radeon_dp_connector_funcs = {
 	.detect = radeon_dp_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = radeon_connector_set_property,
-	.destroy = radeon_dp_connector_destroy,
+	.destroy = radeon_connector_destroy,
 	.force = radeon_dvi_force,
 };
 
@@ -1571,7 +1621,7 @@ static const struct drm_connector_funcs radeon_edp_connector_funcs = {
 	.detect = radeon_dp_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = radeon_lvds_set_property,
-	.destroy = radeon_dp_connector_destroy,
+	.destroy = radeon_connector_destroy,
 	.force = radeon_dvi_force,
 };
 
@@ -1580,7 +1630,7 @@ static const struct drm_connector_funcs radeon_lvds_bridge_connector_funcs = {
 	.detect = radeon_dp_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.set_property = radeon_lvds_set_property,
-	.destroy = radeon_dp_connector_destroy,
+	.destroy = radeon_connector_destroy,
 	.force = radeon_dvi_force,
 };
 
@@ -1604,6 +1654,7 @@ radeon_add_atom_connector(struct drm_device *dev,
 	uint32_t subpixel_order = SubPixelNone;
 	bool shared_ddc = false;
 	bool is_dp_bridge = false;
+	bool has_aux = false;
 
 	if (connector_type == DRM_MODE_CONNECTOR_Unknown)
 		return;
@@ -1676,15 +1727,10 @@ radeon_add_atom_connector(struct drm_device *dev,
 		radeon_dig_connector->igp_lane_info = igp_lane_info;
 		radeon_connector->con_priv = radeon_dig_connector;
 		if (i2c_bus->valid) {
-			/* add DP i2c bus */
-			if (connector_type == DRM_MODE_CONNECTOR_eDP)
-				radeon_dig_connector->dp_i2c_bus = radeon_i2c_create_dp(dev, i2c_bus, "eDP-auxch");
-			else
-				radeon_dig_connector->dp_i2c_bus = radeon_i2c_create_dp(dev, i2c_bus, "DP-auxch");
-			if (!radeon_dig_connector->dp_i2c_bus)
-				DRM_ERROR("DP: Failed to assign dp ddc bus! Check dmesg for i2c errors.\n");
 			radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus);
-			if (!radeon_connector->ddc_bus)
+			if (radeon_connector->ddc_bus)
+				has_aux = true;
+			else
 				DRM_ERROR("DP: Failed to assign ddc bus! Check dmesg for i2c errors.\n");
 		}
 		switch (connector_type) {
@@ -1899,12 +1945,10 @@ radeon_add_atom_connector(struct drm_device *dev,
 			drm_connector_init(dev, &radeon_connector->base, &radeon_dp_connector_funcs, connector_type);
 			drm_connector_helper_add(&radeon_connector->base, &radeon_dp_connector_helper_funcs);
 			if (i2c_bus->valid) {
-				/* add DP i2c bus */
-				radeon_dig_connector->dp_i2c_bus = radeon_i2c_create_dp(dev, i2c_bus, "DP-auxch");
-				if (!radeon_dig_connector->dp_i2c_bus)
-					DRM_ERROR("DP: Failed to assign dp ddc bus! Check dmesg for i2c errors.\n");
 				radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus);
-				if (!radeon_connector->ddc_bus)
+				if (radeon_connector->ddc_bus)
+					has_aux = true;
+				else
 					DRM_ERROR("DP: Failed to assign ddc bus! Check dmesg for i2c errors.\n");
 			}
 			subpixel_order = SubPixelHorizontalRGB;
@@ -1946,12 +1990,10 @@ radeon_add_atom_connector(struct drm_device *dev,
 			drm_connector_init(dev, &radeon_connector->base, &radeon_edp_connector_funcs, connector_type);
 			drm_connector_helper_add(&radeon_connector->base, &radeon_dp_connector_helper_funcs);
 			if (i2c_bus->valid) {
-				/* add DP i2c bus */
-				radeon_dig_connector->dp_i2c_bus = radeon_i2c_create_dp(dev, i2c_bus, "eDP-auxch");
-				if (!radeon_dig_connector->dp_i2c_bus)
-					DRM_ERROR("DP: Failed to assign dp ddc bus! Check dmesg for i2c errors.\n");
 				radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus);
-				if (!radeon_connector->ddc_bus)
+				if (radeon_connector->ddc_bus)
+					has_aux = true;
+				else
 					DRM_ERROR("DP: Failed to assign ddc bus! Check dmesg for i2c errors.\n");
 			}
 			drm_object_attach_property(&radeon_connector->base.base,
@@ -2009,6 +2051,10 @@ radeon_add_atom_connector(struct drm_device *dev,
 
 	connector->display_info.subpixel_order = subpixel_order;
 	drm_sysfs_connector_add(connector);
+
+	if (has_aux)
+		radeon_dp_aux_init(radeon_connector);
+
 	return;
 
 failed:
diff --git a/drivers/gpu/drm/radeon/radeon_cp.c b/drivers/gpu/drm/radeon/radeon_cp.c
index 3cae2bbc185..bb0d5c3a831 100644
--- a/drivers/gpu/drm/radeon/radeon_cp.c
+++ b/drivers/gpu/drm/radeon/radeon_cp.c
@@ -2020,10 +2020,10 @@ static int radeon_cp_get_buffers(struct drm_device *dev,
 
 		buf->file_priv = file_priv;
 
-		if (DRM_COPY_TO_USER(&d->request_indices[i], &buf->idx,
+		if (copy_to_user(&d->request_indices[i], &buf->idx,
 				     sizeof(buf->idx)))
 			return -EFAULT;
-		if (DRM_COPY_TO_USER(&d->request_sizes[i], &buf->total,
+		if (copy_to_user(&d->request_sizes[i], &buf->total,
 				     sizeof(buf->total)))
 			return -EFAULT;
 
@@ -2228,7 +2228,7 @@ void radeon_commit_ring(drm_radeon_private_t *dev_priv)
 
 	dev_priv->ring.tail &= dev_priv->ring.tail_mask;
 
-	DRM_MEMORYBARRIER();
+	mb();
 	GET_RING_HEAD( dev_priv );
 
 	if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_R600) {
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index 0b366169d64..ae763f60c8a 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -24,16 +24,59 @@
  * Authors:
  *    Jerome Glisse <glisse@freedesktop.org>
  */
+#include <linux/list_sort.h>
 #include <drm/drmP.h>
 #include <drm/radeon_drm.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 #include "radeon_trace.h"
 
+#define RADEON_CS_MAX_PRIORITY		32u
+#define RADEON_CS_NUM_BUCKETS		(RADEON_CS_MAX_PRIORITY + 1)
+
+/* This is based on the bucket sort with O(n) time complexity.
+ * An item with priority "i" is added to bucket[i]. The lists are then
+ * concatenated in descending order.
+ */
+struct radeon_cs_buckets {
+	struct list_head bucket[RADEON_CS_NUM_BUCKETS];
+};
+
+static void radeon_cs_buckets_init(struct radeon_cs_buckets *b)
+{
+	unsigned i;
+
+	for (i = 0; i < RADEON_CS_NUM_BUCKETS; i++)
+		INIT_LIST_HEAD(&b->bucket[i]);
+}
+
+static void radeon_cs_buckets_add(struct radeon_cs_buckets *b,
+				  struct list_head *item, unsigned priority)
+{
+	/* Since buffers which appear sooner in the relocation list are
+	 * likely to be used more often than buffers which appear later
+	 * in the list, the sort mustn't change the ordering of buffers
+	 * with the same priority, i.e. it must be stable.
+	 */
+	list_add_tail(item, &b->bucket[min(priority, RADEON_CS_MAX_PRIORITY)]);
+}
+
+static void radeon_cs_buckets_get_list(struct radeon_cs_buckets *b,
+				       struct list_head *out_list)
+{
+	unsigned i;
+
+	/* Connect the sorted buckets in the output list. */
+	for (i = 0; i < RADEON_CS_NUM_BUCKETS; i++) {
+		list_splice(&b->bucket[i], out_list);
+	}
+}
+
 static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 {
 	struct drm_device *ddev = p->rdev->ddev;
 	struct radeon_cs_chunk *chunk;
+	struct radeon_cs_buckets buckets;
 	unsigned i, j;
 	bool duplicate;
 
@@ -52,8 +95,12 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 	if (p->relocs == NULL) {
 		return -ENOMEM;
 	}
+
+	radeon_cs_buckets_init(&buckets);
+
 	for (i = 0; i < p->nrelocs; i++) {
 		struct drm_radeon_cs_reloc *r;
+		unsigned priority;
 
 		duplicate = false;
 		r = (struct drm_radeon_cs_reloc *)&chunk->kdata[i*4];
@@ -78,8 +125,14 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 		}
 		p->relocs_ptr[i] = &p->relocs[i];
 		p->relocs[i].robj = gem_to_radeon_bo(p->relocs[i].gobj);
-		p->relocs[i].lobj.bo = p->relocs[i].robj;
-		p->relocs[i].lobj.written = !!r->write_domain;
+
+		/* The userspace buffer priorities are from 0 to 15. A higher
+		 * number means the buffer is more important.
+		 * Also, the buffers used for write have a higher priority than
+		 * the buffers used for read only, which doubles the range
+		 * to 0 to 31. 32 is reserved for the kernel driver.
+		 */
+		priority = (r->flags & 0xf) * 2 + !!r->write_domain;
 
 		/* the first reloc of an UVD job is the msg and that must be in
 		   VRAM, also but everything into VRAM on AGP cards to avoid
@@ -87,29 +140,44 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 		if (p->ring == R600_RING_TYPE_UVD_INDEX &&
 		    (i == 0 || drm_pci_device_is_agp(p->rdev->ddev))) {
 			/* TODO: is this still needed for NI+ ? */
-			p->relocs[i].lobj.domain =
+			p->relocs[i].prefered_domains =
 				RADEON_GEM_DOMAIN_VRAM;
 
-			p->relocs[i].lobj.alt_domain =
+			p->relocs[i].allowed_domains =
 				RADEON_GEM_DOMAIN_VRAM;
 
+			/* prioritize this over any other relocation */
+			priority = RADEON_CS_MAX_PRIORITY;
 		} else {
 			uint32_t domain = r->write_domain ?
 				r->write_domain : r->read_domains;
 
-			p->relocs[i].lobj.domain = domain;
+			if (domain & RADEON_GEM_DOMAIN_CPU) {
+				DRM_ERROR("RADEON_GEM_DOMAIN_CPU is not valid "
+					  "for command submission\n");
+				return -EINVAL;
+			}
+
+			p->relocs[i].prefered_domains = domain;
 			if (domain == RADEON_GEM_DOMAIN_VRAM)
 				domain |= RADEON_GEM_DOMAIN_GTT;
-			p->relocs[i].lobj.alt_domain = domain;
+			p->relocs[i].allowed_domains = domain;
 		}
 
-		p->relocs[i].lobj.tv.bo = &p->relocs[i].robj->tbo;
+		p->relocs[i].tv.bo = &p->relocs[i].robj->tbo;
 		p->relocs[i].handle = r->handle;
 
-		radeon_bo_list_add_object(&p->relocs[i].lobj,
-					  &p->validated);
+		radeon_cs_buckets_add(&buckets, &p->relocs[i].tv.head,
+				      priority);
 	}
-	return radeon_bo_list_validate(&p->ticket, &p->validated, p->ring);
+
+	radeon_cs_buckets_get_list(&buckets, &p->validated);
+
+	if (p->cs_flags & RADEON_CS_USE_VM)
+		p->vm_bos = radeon_vm_get_bos(p->rdev, p->ib.vm,
+					      &p->validated);
+
+	return radeon_bo_list_validate(p->rdev, &p->ticket, &p->validated, p->ring);
 }
 
 static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority)
@@ -138,7 +206,7 @@ static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority
 				p->ring = R600_RING_TYPE_DMA_INDEX;
 			else
 				p->ring = CAYMAN_RING_TYPE_DMA1_INDEX;
-		} else if (p->rdev->family >= CHIP_R600) {
+		} else if (p->rdev->family >= CHIP_RV770) {
 			p->ring = R600_RING_TYPE_DMA_INDEX;
 		} else {
 			return -EINVAL;
@@ -147,6 +215,10 @@ static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority
 	case RADEON_CS_RING_UVD:
 		p->ring = R600_RING_TYPE_UVD_INDEX;
 		break;
+	case RADEON_CS_RING_VCE:
+		/* TODO: only use the low priority ring for now */
+		p->ring = TN_RING_TYPE_VCE1_INDEX;
+		break;
 	}
 	return 0;
 }
@@ -192,7 +264,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 		return -ENOMEM;
 	}
 	chunk_array_ptr = (uint64_t *)(unsigned long)(cs->chunks);
-	if (DRM_COPY_FROM_USER(p->chunks_array, chunk_array_ptr,
+	if (copy_from_user(p->chunks_array, chunk_array_ptr,
 			       sizeof(uint64_t)*cs->num_chunks)) {
 		return -EFAULT;
 	}
@@ -208,7 +280,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 		uint32_t __user *cdata;
 
 		chunk_ptr = (void __user*)(unsigned long)p->chunks_array[i];
-		if (DRM_COPY_FROM_USER(&user_chunk, chunk_ptr,
+		if (copy_from_user(&user_chunk, chunk_ptr,
 				       sizeof(struct drm_radeon_cs_chunk))) {
 			return -EFAULT;
 		}
@@ -252,7 +324,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 		if (p->chunks[i].kdata == NULL) {
 			return -ENOMEM;
 		}
-		if (DRM_COPY_FROM_USER(p->chunks[i].kdata, cdata, size)) {
+		if (copy_from_user(p->chunks[i].kdata, cdata, size)) {
 			return -EFAULT;
 		}
 		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_FLAGS) {
@@ -276,16 +348,33 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 			return -EINVAL;
 
 		/* we only support VM on some SI+ rings */
-		if ((p->rdev->asic->ring[p->ring]->cs_parse == NULL) &&
-		   ((p->cs_flags & RADEON_CS_USE_VM) == 0)) {
-			DRM_ERROR("Ring %d requires VM!\n", p->ring);
-			return -EINVAL;
+		if ((p->cs_flags & RADEON_CS_USE_VM) == 0) {
+			if (p->rdev->asic->ring[p->ring]->cs_parse == NULL) {
+				DRM_ERROR("Ring %d requires VM!\n", p->ring);
+				return -EINVAL;
+			}
+		} else {
+			if (p->rdev->asic->ring[p->ring]->ib_parse == NULL) {
+				DRM_ERROR("VM not supported on ring %d!\n",
+					  p->ring);
+				return -EINVAL;
+			}
 		}
 	}
 
 	return 0;
 }
 
+static int cmp_size_smaller_first(void *priv, struct list_head *a,
+				  struct list_head *b)
+{
+	struct radeon_cs_reloc *la = list_entry(a, struct radeon_cs_reloc, tv.head);
+	struct radeon_cs_reloc *lb = list_entry(b, struct radeon_cs_reloc, tv.head);
+
+	/* Sort A before B if A is smaller. */
+	return (int)la->robj->tbo.num_pages - (int)lb->robj->tbo.num_pages;
+}
+
 /**
  * cs_parser_fini() - clean parser states
  * @parser:	parser structure holding parsing context.
@@ -299,6 +388,18 @@ static void radeon_cs_parser_fini(struct radeon_cs_parser *parser, int error, bo
 	unsigned i;
 
 	if (!error) {
+		/* Sort the buffer list from the smallest to largest buffer,
+		 * which affects the order of buffers in the LRU list.
+		 * This assures that the smallest buffers are added first
+		 * to the LRU list, so they are likely to be later evicted
+		 * first, instead of large buffers whose eviction is more
+		 * expensive.
+		 *
+		 * This slightly lowers the number of bytes moved by TTM
+		 * per frame under memory pressure.
+		 */
+		list_sort(NULL, &parser->validated, cmp_size_smaller_first);
+
 		ttm_eu_fence_buffer_objects(&parser->ticket,
 					    &parser->validated,
 					    parser->ib.fence);
@@ -316,6 +417,7 @@ static void radeon_cs_parser_fini(struct radeon_cs_parser *parser, int error, bo
 	kfree(parser->track);
 	kfree(parser->relocs);
 	kfree(parser->relocs_ptr);
+	kfree(parser->vm_bos);
 	for (i = 0; i < parser->nchunks; i++)
 		drm_free_large(parser->chunks[i].kdata);
 	kfree(parser->chunks);
@@ -343,6 +445,9 @@ static int radeon_cs_ib_chunk(struct radeon_device *rdev,
 
 	if (parser->ring == R600_RING_TYPE_UVD_INDEX)
 		radeon_uvd_note_usage(rdev);
+	else if ((parser->ring == TN_RING_TYPE_VCE1_INDEX) ||
+		 (parser->ring == TN_RING_TYPE_VCE2_INDEX))
+		radeon_vce_note_usage(rdev);
 
 	radeon_cs_sync_rings(parser);
 	r = radeon_ib_schedule(rdev, &parser->ib, NULL);
@@ -352,24 +457,48 @@ static int radeon_cs_ib_chunk(struct radeon_device *rdev,
 	return r;
 }
 
-static int radeon_bo_vm_update_pte(struct radeon_cs_parser *parser,
+static int radeon_bo_vm_update_pte(struct radeon_cs_parser *p,
 				   struct radeon_vm *vm)
 {
-	struct radeon_device *rdev = parser->rdev;
-	struct radeon_bo_list *lobj;
-	struct radeon_bo *bo;
-	int r;
+	struct radeon_device *rdev = p->rdev;
+	struct radeon_bo_va *bo_va;
+	int i, r;
 
-	r = radeon_vm_bo_update(rdev, vm, rdev->ring_tmp_bo.bo, &rdev->ring_tmp_bo.bo->tbo.mem);
-	if (r) {
+	r = radeon_vm_update_page_directory(rdev, vm);
+	if (r)
 		return r;
+
+	r = radeon_vm_clear_freed(rdev, vm);
+	if (r)
+		return r;
+
+	if (vm->ib_bo_va == NULL) {
+		DRM_ERROR("Tmp BO not in VM!\n");
+		return -EINVAL;
 	}
-	list_for_each_entry(lobj, &parser->validated, tv.head) {
-		bo = lobj->bo;
-		r = radeon_vm_bo_update(parser->rdev, vm, bo, &bo->tbo.mem);
-		if (r) {
-			return r;
+
+	r = radeon_vm_bo_update(rdev, vm->ib_bo_va,
+				&rdev->ring_tmp_bo.bo->tbo.mem);
+	if (r)
+		return r;
+
+	for (i = 0; i < p->nrelocs; i++) {
+		struct radeon_bo *bo;
+
+		/* ignore duplicates */
+		if (p->relocs_ptr[i] != &p->relocs[i])
+			continue;
+
+		bo = p->relocs[i].robj;
+		bo_va = radeon_vm_bo_find(vm, bo);
+		if (bo_va == NULL) {
+			dev_err(rdev->dev, "bo %p not in vm %p\n", bo, vm);
+			return -EINVAL;
 		}
+
+		r = radeon_vm_bo_update(rdev, bo_va, &bo->tbo.mem);
+		if (r)
+			return r;
 	}
 	return 0;
 }
@@ -401,20 +530,13 @@ static int radeon_cs_ib_vm_chunk(struct radeon_device *rdev,
 	if (parser->ring == R600_RING_TYPE_UVD_INDEX)
 		radeon_uvd_note_usage(rdev);
 
-	mutex_lock(&rdev->vm_manager.lock);
 	mutex_lock(&vm->mutex);
-	r = radeon_vm_alloc_pt(rdev, vm);
-	if (r) {
-		goto out;
-	}
 	r = radeon_bo_vm_update_pte(parser, vm);
 	if (r) {
 		goto out;
 	}
 	radeon_cs_sync_rings(parser);
 	radeon_semaphore_sync_to(parser->ib.semaphore, vm->fence);
-	radeon_semaphore_sync_to(parser->ib.semaphore,
-				 radeon_vm_grab_id(rdev, vm, parser->ring));
 
 	if ((rdev->family >= CHIP_TAHITI) &&
 	    (parser->chunk_const_ib_idx != -1)) {
@@ -423,14 +545,8 @@ static int radeon_cs_ib_vm_chunk(struct radeon_device *rdev,
 		r = radeon_ib_schedule(rdev, &parser->ib, NULL);
 	}
 
-	if (!r) {
-		radeon_vm_fence(rdev, vm, parser->ib.fence);
-	}
-
 out:
-	radeon_vm_add_to_lru(rdev, vm);
 	mutex_unlock(&vm->mutex);
-	mutex_unlock(&rdev->vm_manager.lock);
 	return r;
 }
 
@@ -472,7 +588,7 @@ static int radeon_cs_ib_fill(struct radeon_device *rdev, struct radeon_cs_parser
 			}
 			parser->const_ib.is_const_ib = true;
 			parser->const_ib.length_dw = ib_chunk->length_dw;
-			if (DRM_COPY_FROM_USER(parser->const_ib.ptr,
+			if (copy_from_user(parser->const_ib.ptr,
 					       ib_chunk->user_ptr,
 					       ib_chunk->length_dw * 4))
 				return -EFAULT;
@@ -495,7 +611,7 @@ static int radeon_cs_ib_fill(struct radeon_device *rdev, struct radeon_cs_parser
 	parser->ib.length_dw = ib_chunk->length_dw;
 	if (ib_chunk->kdata)
 		memcpy(parser->ib.ptr, ib_chunk->kdata, ib_chunk->length_dw * 4);
-	else if (DRM_COPY_FROM_USER(parser->ib.ptr, ib_chunk->user_ptr, ib_chunk->length_dw * 4))
+	else if (copy_from_user(parser->ib.ptr, ib_chunk->user_ptr, ib_chunk->length_dw * 4))
 		return -EFAULT;
 	return 0;
 }
@@ -698,9 +814,9 @@ int radeon_cs_packet_next_reloc(struct radeon_cs_parser *p,
 	/* FIXME: we assume reloc size is 4 dwords */
 	if (nomm) {
 		*cs_reloc = p->relocs;
-		(*cs_reloc)->lobj.gpu_offset =
+		(*cs_reloc)->gpu_offset =
 			(u64)relocs_chunk->kdata[idx + 3] << 32;
-		(*cs_reloc)->lobj.gpu_offset |= relocs_chunk->kdata[idx + 0];
+		(*cs_reloc)->gpu_offset |= relocs_chunk->kdata[idx + 0];
 	} else
 		*cs_reloc = p->relocs_ptr[(idx / 4)];
 	return 0;
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 39b033b441d..697add2cd4e 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -99,14 +99,18 @@ static const char radeon_family_name[][16] = {
 	"KAVERI",
 	"KABINI",
 	"HAWAII",
+	"MULLINS",
 	"LAST",
 };
 
-#if defined(CONFIG_VGA_SWITCHEROO)
-bool radeon_is_px(void);
-#else
-static inline bool radeon_is_px(void) { return false; }
-#endif
+bool radeon_is_px(struct drm_device *dev)
+{
+	struct radeon_device *rdev = dev->dev_private;
+
+	if (rdev->flags & RADEON_IS_PX)
+		return true;
+	return false;
+}
 
 /**
  * radeon_program_register_sequence - program an array of registers.
@@ -144,6 +148,11 @@ void radeon_program_register_sequence(struct radeon_device *rdev,
 	}
 }
 
+void radeon_pci_config_reset(struct radeon_device *rdev)
+{
+	pci_write_config_dword(rdev->pdev, 0x7c, RADEON_ASIC_RESET_DATA);
+}
+
 /**
  * radeon_surface_init - Clear GPU surface registers.
  *
@@ -249,7 +258,7 @@ void radeon_scratch_free(struct radeon_device *rdev, uint32_t reg)
  * Init doorbell driver information (CIK)
  * Returns 0 on success, error on failure.
  */
-int radeon_doorbell_init(struct radeon_device *rdev)
+static int radeon_doorbell_init(struct radeon_device *rdev)
 {
 	/* doorbell bar mapping */
 	rdev->doorbell.base = pci_resource_start(rdev->pdev, 2);
@@ -278,7 +287,7 @@ int radeon_doorbell_init(struct radeon_device *rdev)
  *
  * Tear down doorbell driver information (CIK)
  */
-void radeon_doorbell_fini(struct radeon_device *rdev)
+static void radeon_doorbell_fini(struct radeon_device *rdev)
 {
 	iounmap(rdev->doorbell.ptr);
 	rdev->doorbell.ptr = NULL;
@@ -1043,6 +1052,43 @@ static void radeon_check_arguments(struct radeon_device *rdev)
 		radeon_agpmode = 0;
 		break;
 	}
+
+	if (!radeon_check_pot_argument(radeon_vm_size)) {
+		dev_warn(rdev->dev, "VM size (%d) must be a power of 2\n",
+			 radeon_vm_size);
+		radeon_vm_size = 4;
+	}
+
+	if (radeon_vm_size < 1) {
+		dev_warn(rdev->dev, "VM size (%d) to small, min is 1GB\n",
+			 radeon_vm_size);
+		radeon_vm_size = 4;
+	}
+
+       /*
+        * Max GPUVM size for Cayman, SI and CI are 40 bits.
+        */
+	if (radeon_vm_size > 1024) {
+		dev_warn(rdev->dev, "VM size (%d) too large, max is 1TB\n",
+			 radeon_vm_size);
+		radeon_vm_size = 4;
+	}
+
+	/* defines number of bits in page table versus page directory,
+	 * a page is 4KB so we have 12 bits offset, minimum 9 bits in the
+	 * page table and the remaining bits are in the page directory */
+	if (radeon_vm_block_size < 9) {
+		dev_warn(rdev->dev, "VM page table size (%d) too small\n",
+			 radeon_vm_block_size);
+		radeon_vm_block_size = 9;
+	}
+
+	if (radeon_vm_block_size > 24 ||
+	    (radeon_vm_size * 1024) < (1ull << radeon_vm_block_size)) {
+		dev_warn(rdev->dev, "VM page table size (%d) too large\n",
+			 radeon_vm_block_size);
+		radeon_vm_block_size = 9;
+	}
 }
 
 /**
@@ -1077,7 +1123,7 @@ static void radeon_switcheroo_set_state(struct pci_dev *pdev, enum vga_switchero
 {
 	struct drm_device *dev = pci_get_drvdata(pdev);
 
-	if (radeon_is_px() && state == VGA_SWITCHEROO_OFF)
+	if (radeon_is_px(dev) && state == VGA_SWITCHEROO_OFF)
 		return;
 
 	if (state == VGA_SWITCHEROO_ON) {
@@ -1117,12 +1163,13 @@ static void radeon_switcheroo_set_state(struct pci_dev *pdev, enum vga_switchero
 static bool radeon_switcheroo_can_switch(struct pci_dev *pdev)
 {
 	struct drm_device *dev = pci_get_drvdata(pdev);
-	bool can_switch;
 
-	spin_lock(&dev->count_lock);
-	can_switch = (dev->open_count == 0);
-	spin_unlock(&dev->count_lock);
-	return can_switch;
+	/*
+	 * FIXME: open_count is protected by drm_global_mutex but that would lead to
+	 * locking inversion with the driver load path. And the access here is
+	 * completely racy anyway. So don't bother with locking for now.
+	 */
+	return dev->open_count == 0;
 }
 
 static const struct vga_switcheroo_client_ops radeon_switcheroo_ops = {
@@ -1186,20 +1233,17 @@ int radeon_device_init(struct radeon_device *rdev,
 	r = radeon_gem_init(rdev);
 	if (r)
 		return r;
-	/* initialize vm here */
-	mutex_init(&rdev->vm_manager.lock);
+
+	radeon_check_arguments(rdev);
 	/* Adjust VM size here.
-	 * Currently set to 4GB ((1 << 20) 4k pages).
-	 * Max GPUVM size for cayman and SI is 40 bits.
+	 * Max GPUVM size for cayman+ is 40 bits.
 	 */
-	rdev->vm_manager.max_pfn = 1 << 20;
-	INIT_LIST_HEAD(&rdev->vm_manager.lru_vm);
+	rdev->vm_manager.max_pfn = radeon_vm_size << 18;
 
 	/* Set asic functions */
 	r = radeon_asic_init(rdev);
 	if (r)
 		return r;
-	radeon_check_arguments(rdev);
 
 	/* all of the newer IGP chips have an internal gart
 	 * However some rs4xx report as AGP, so remove that here.
@@ -1298,9 +1342,7 @@ int radeon_device_init(struct radeon_device *rdev,
 	 * ignore it */
 	vga_client_register(rdev->pdev, rdev, NULL, radeon_vga_set_decode);
 
-	if (radeon_runtime_pm == 1)
-		runtime = true;
-	if ((radeon_runtime_pm == -1) && radeon_is_px())
+	if (rdev->flags & RADEON_IS_PX)
 		runtime = true;
 	vga_switcheroo_register_client(rdev->pdev, &radeon_switcheroo_ops, runtime);
 	if (runtime)
@@ -1330,6 +1372,7 @@ int radeon_device_init(struct radeon_device *rdev,
 		if (r)
 			return r;
 	}
+
 	if ((radeon_testing & 1)) {
 		if (rdev->accel_working)
 			radeon_test_moves(rdev);
@@ -1420,7 +1463,7 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 
 	/* unpin the front buffers */
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
-		struct radeon_framebuffer *rfb = to_radeon_framebuffer(crtc->fb);
+		struct radeon_framebuffer *rfb = to_radeon_framebuffer(crtc->primary->fb);
 		struct radeon_bo *robj;
 
 		if (rfb == NULL || rfb->obj == NULL) {
@@ -1439,10 +1482,9 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 	/* evict vram memory */
 	radeon_bo_evict_vram(rdev);
 
-	mutex_lock(&rdev->ring_lock);
 	/* wait for gpu to finish processing current batch */
 	for (i = 0; i < RADEON_NUM_RINGS; i++) {
-		r = radeon_fence_wait_empty_locked(rdev, i);
+		r = radeon_fence_wait_empty(rdev, i);
 		if (r) {
 			/* delay GPU reset to resume */
 			force_completion = true;
@@ -1451,11 +1493,9 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 	if (force_completion) {
 		radeon_fence_driver_force_completion(rdev);
 	}
-	mutex_unlock(&rdev->ring_lock);
 
 	radeon_save_bios_scratch_regs(rdev);
 
-	radeon_pm_suspend(rdev);
 	radeon_suspend(rdev);
 	radeon_hpd_fini(rdev);
 	/* evict remaining vram memory */
@@ -1516,14 +1556,20 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon)
 	if (r)
 		DRM_ERROR("ib ring test failed (%d).\n", r);
 
-	radeon_pm_resume(rdev);
+	if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled) {
+		/* do dpm late init */
+		r = radeon_pm_late_init(rdev);
+		if (r) {
+			rdev->pm.dpm_enabled = false;
+			DRM_ERROR("radeon_pm_late_init failed, disabling dpm\n");
+		}
+	} else {
+		/* resume old pm late */
+		radeon_pm_resume(rdev);
+	}
+
 	radeon_restore_bios_scratch_regs(rdev);
 
-	if (fbcon) {
-		radeon_fbdev_set_suspend(rdev, 0);
-		console_unlock();
-	}
-       
 	/* init dig PHYs, disp eng pll */
 	if (rdev->is_atom_bios) {
 		radeon_atom_encoder_init(rdev);
@@ -1539,13 +1585,25 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon)
 	/* reset hpd state */
 	radeon_hpd_init(rdev);
 	/* blat the mode back in */
-	drm_helper_resume_force_mode(dev);
-	/* turn on display hw */
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
-		drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON);
+	if (fbcon) {
+		drm_helper_resume_force_mode(dev);
+		/* turn on display hw */
+		list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+			drm_helper_connector_dpms(connector, DRM_MODE_DPMS_ON);
+		}
 	}
 
 	drm_kms_helper_poll_enable(dev);
+
+	/* set the power state here in case we are a PX system or headless */
+	if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled)
+		radeon_pm_compute_clocks(rdev);
+
+	if (fbcon) {
+		radeon_fbdev_set_suspend(rdev, 0);
+		console_unlock();
+	}
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index 7b253815a32..bf25061c8ac 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -34,6 +34,8 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_edid.h>
 
+#include <linux/gcd.h>
+
 static void avivo_crtc_load_lut(struct drm_crtc *crtc)
 {
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
@@ -64,7 +66,8 @@ static void avivo_crtc_load_lut(struct drm_crtc *crtc)
 			     (radeon_crtc->lut_b[i] << 0));
 	}
 
-	WREG32(AVIVO_D1GRPH_LUT_SEL + radeon_crtc->crtc_offset, radeon_crtc->crtc_id);
+	/* Only change bit 0 of LUT_SEL, other bits are set elsewhere */
+	WREG32_P(AVIVO_D1GRPH_LUT_SEL + radeon_crtc->crtc_offset, radeon_crtc->crtc_id, ~1);
 }
 
 static void dce4_crtc_load_lut(struct drm_crtc *crtc)
@@ -247,16 +250,21 @@ static void radeon_crtc_destroy(struct drm_crtc *crtc)
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 
 	drm_crtc_cleanup(crtc);
+	destroy_workqueue(radeon_crtc->flip_queue);
 	kfree(radeon_crtc);
 }
 
-/*
- * Handle unpin events outside the interrupt handler proper.
+/**
+ * radeon_unpin_work_func - unpin old buffer object
+ *
+ * @__work - kernel work item
+ *
+ * Unpin the old frame buffer object outside of the interrupt handler
  */
 static void radeon_unpin_work_func(struct work_struct *__work)
 {
-	struct radeon_unpin_work *work =
-		container_of(__work, struct radeon_unpin_work, work);
+	struct radeon_flip_work *work =
+		container_of(__work, struct radeon_flip_work, unpin_work);
 	int r;
 
 	/* unpin of the old buffer */
@@ -274,39 +282,34 @@ static void radeon_unpin_work_func(struct work_struct *__work)
 	kfree(work);
 }
 
-void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id)
+void radeon_crtc_handle_vblank(struct radeon_device *rdev, int crtc_id)
 {
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
-	struct radeon_unpin_work *work;
 	unsigned long flags;
 	u32 update_pending;
 	int vpos, hpos;
 
+	/* can happen during initialization */
+	if (radeon_crtc == NULL)
+		return;
+
 	spin_lock_irqsave(&rdev->ddev->event_lock, flags);
-	work = radeon_crtc->unpin_work;
-	if (work == NULL ||
-	    (work->fence && !radeon_fence_signaled(work->fence))) {
+	if (radeon_crtc->flip_status != RADEON_FLIP_SUBMITTED) {
+		DRM_DEBUG_DRIVER("radeon_crtc->flip_status = %d != "
+				 "RADEON_FLIP_SUBMITTED(%d)\n",
+				 radeon_crtc->flip_status,
+				 RADEON_FLIP_SUBMITTED);
 		spin_unlock_irqrestore(&rdev->ddev->event_lock, flags);
 		return;
 	}
-	/* New pageflip, or just completion of a previous one? */
-	if (!radeon_crtc->deferred_flip_completion) {
-		/* do the flip (mmio) */
-		update_pending = radeon_page_flip(rdev, crtc_id, work->new_crtc_base);
-	} else {
-		/* This is just a completion of a flip queued in crtc
-		 * at last invocation. Make sure we go directly to
-		 * completion routine.
-		 */
-		update_pending = 0;
-		radeon_crtc->deferred_flip_completion = 0;
-	}
+
+	update_pending = radeon_page_flip_pending(rdev, crtc_id);
 
 	/* Has the pageflip already completed in crtc, or is it certain
 	 * to complete in this vblank?
 	 */
 	if (update_pending &&
-	    (DRM_SCANOUTPOS_VALID & radeon_get_crtc_scanoutpos(rdev->ddev, crtc_id,
+	    (DRM_SCANOUTPOS_VALID & radeon_get_crtc_scanoutpos(rdev->ddev, crtc_id, 0,
 							       &vpos, &hpos, NULL, NULL)) &&
 	    ((vpos >= (99 * rdev->mode_info.crtcs[crtc_id]->base.hwmode.crtc_vdisplay)/100) ||
 	     (vpos < 0 && !ASIC_IS_AVIVO(rdev)))) {
@@ -318,19 +321,43 @@ void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id)
 		 */
 		update_pending = 0;
 	}
-	if (update_pending) {
-		/* crtc didn't flip in this target vblank interval,
-		 * but flip is pending in crtc. It will complete it
-		 * in next vblank interval, so complete the flip at
-		 * next vblank irq.
-		 */
-		radeon_crtc->deferred_flip_completion = 1;
+	spin_unlock_irqrestore(&rdev->ddev->event_lock, flags);
+	if (!update_pending)
+		radeon_crtc_handle_flip(rdev, crtc_id);
+}
+
+/**
+ * radeon_crtc_handle_flip - page flip completed
+ *
+ * @rdev: radeon device pointer
+ * @crtc_id: crtc number this event is for
+ *
+ * Called when we are sure that a page flip for this crtc is completed.
+ */
+void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id)
+{
+	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
+	struct radeon_flip_work *work;
+	unsigned long flags;
+
+	/* this can happen at init */
+	if (radeon_crtc == NULL)
+		return;
+
+	spin_lock_irqsave(&rdev->ddev->event_lock, flags);
+	work = radeon_crtc->flip_work;
+	if (radeon_crtc->flip_status != RADEON_FLIP_SUBMITTED) {
+		DRM_DEBUG_DRIVER("radeon_crtc->flip_status = %d != "
+				 "RADEON_FLIP_SUBMITTED(%d)\n",
+				 radeon_crtc->flip_status,
+				 RADEON_FLIP_SUBMITTED);
 		spin_unlock_irqrestore(&rdev->ddev->event_lock, flags);
 		return;
 	}
 
-	/* Pageflip (will be) certainly completed in this vblank. Clean up. */
-	radeon_crtc->unpin_work = NULL;
+	/* Pageflip completed. Clean up. */
+	radeon_crtc->flip_status = RADEON_FLIP_NONE;
+	radeon_crtc->flip_work = NULL;
 
 	/* wakeup userspace */
 	if (work->event)
@@ -339,9 +366,59 @@ void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id)
 	spin_unlock_irqrestore(&rdev->ddev->event_lock, flags);
 
 	drm_vblank_put(rdev->ddev, radeon_crtc->crtc_id);
-	radeon_fence_unref(&work->fence);
-	radeon_post_page_flip(work->rdev, work->crtc_id);
-	schedule_work(&work->work);
+	radeon_irq_kms_pflip_irq_put(rdev, work->crtc_id);
+	queue_work(radeon_crtc->flip_queue, &work->unpin_work);
+}
+
+/**
+ * radeon_flip_work_func - page flip framebuffer
+ *
+ * @work - kernel work item
+ *
+ * Wait for the buffer object to become idle and do the actual page flip
+ */
+static void radeon_flip_work_func(struct work_struct *__work)
+{
+	struct radeon_flip_work *work =
+		container_of(__work, struct radeon_flip_work, flip_work);
+	struct radeon_device *rdev = work->rdev;
+	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[work->crtc_id];
+
+	struct drm_crtc *crtc = &radeon_crtc->base;
+	unsigned long flags;
+	int r;
+
+        down_read(&rdev->exclusive_lock);
+	if (work->fence) {
+		r = radeon_fence_wait(work->fence, false);
+		if (r == -EDEADLK) {
+			up_read(&rdev->exclusive_lock);
+			r = radeon_gpu_reset(rdev);
+			down_read(&rdev->exclusive_lock);
+		}
+		if (r)
+			DRM_ERROR("failed to wait on page flip fence (%d)!\n", r);
+
+		/* We continue with the page flip even if we failed to wait on
+		 * the fence, otherwise the DRM core and userspace will be
+		 * confused about which BO the CRTC is scanning out
+		 */
+
+		radeon_fence_unref(&work->fence);
+	}
+
+	/* We borrow the event spin lock for protecting flip_status */
+	spin_lock_irqsave(&crtc->dev->event_lock, flags);
+
+	/* set the proper interrupt */
+	radeon_irq_kms_pflip_irq_get(rdev, radeon_crtc->crtc_id);
+
+	/* do the flip (mmio) */
+	radeon_page_flip(rdev, radeon_crtc->crtc_id, work->base);
+
+	radeon_crtc->flip_status = RADEON_FLIP_SUBMITTED;
+	spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
+	up_read(&rdev->exclusive_lock);
 }
 
 static int radeon_crtc_page_flip(struct drm_crtc *crtc,
@@ -355,69 +432,61 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 	struct radeon_framebuffer *old_radeon_fb;
 	struct radeon_framebuffer *new_radeon_fb;
 	struct drm_gem_object *obj;
-	struct radeon_bo *rbo;
-	struct radeon_unpin_work *work;
+	struct radeon_flip_work *work;
+	struct radeon_bo *new_rbo;
+	uint32_t tiling_flags, pitch_pixels;
+	uint64_t base;
 	unsigned long flags;
-	u32 tiling_flags, pitch_pixels;
-	u64 base;
 	int r;
 
 	work = kzalloc(sizeof *work, GFP_KERNEL);
 	if (work == NULL)
 		return -ENOMEM;
 
-	work->event = event;
+	INIT_WORK(&work->flip_work, radeon_flip_work_func);
+	INIT_WORK(&work->unpin_work, radeon_unpin_work_func);
+
 	work->rdev = rdev;
 	work->crtc_id = radeon_crtc->crtc_id;
-	old_radeon_fb = to_radeon_framebuffer(crtc->fb);
-	new_radeon_fb = to_radeon_framebuffer(fb);
+	work->event = event;
+
 	/* schedule unpin of the old buffer */
+	old_radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
 	obj = old_radeon_fb->obj;
+
 	/* take a reference to the old object */
 	drm_gem_object_reference(obj);
-	rbo = gem_to_radeon_bo(obj);
-	work->old_rbo = rbo;
-	obj = new_radeon_fb->obj;
-	rbo = gem_to_radeon_bo(obj);
-
-	spin_lock(&rbo->tbo.bdev->fence_lock);
-	if (rbo->tbo.sync_obj)
-		work->fence = radeon_fence_ref(rbo->tbo.sync_obj);
-	spin_unlock(&rbo->tbo.bdev->fence_lock);
+	work->old_rbo = gem_to_radeon_bo(obj);
 
-	INIT_WORK(&work->work, radeon_unpin_work_func);
+	new_radeon_fb = to_radeon_framebuffer(fb);
+	obj = new_radeon_fb->obj;
+	new_rbo = gem_to_radeon_bo(obj);
 
-	/* We borrow the event spin lock for protecting unpin_work */
-	spin_lock_irqsave(&dev->event_lock, flags);
-	if (radeon_crtc->unpin_work) {
-		DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
-		r = -EBUSY;
-		goto unlock_free;
-	}
-	radeon_crtc->unpin_work = work;
-	radeon_crtc->deferred_flip_completion = 0;
-	spin_unlock_irqrestore(&dev->event_lock, flags);
+	spin_lock(&new_rbo->tbo.bdev->fence_lock);
+	if (new_rbo->tbo.sync_obj)
+		work->fence = radeon_fence_ref(new_rbo->tbo.sync_obj);
+	spin_unlock(&new_rbo->tbo.bdev->fence_lock);
 
 	/* pin the new buffer */
-	DRM_DEBUG_DRIVER("flip-ioctl() cur_fbo = %p, cur_bbo = %p\n",
-			 work->old_rbo, rbo);
+	DRM_DEBUG_DRIVER("flip-ioctl() cur_rbo = %p, new_rbo = %p\n",
+			 work->old_rbo, new_rbo);
 
-	r = radeon_bo_reserve(rbo, false);
+	r = radeon_bo_reserve(new_rbo, false);
 	if (unlikely(r != 0)) {
 		DRM_ERROR("failed to reserve new rbo buffer before flip\n");
-		goto pflip_cleanup;
+		goto cleanup;
 	}
 	/* Only 27 bit offset for legacy CRTC */
-	r = radeon_bo_pin_restricted(rbo, RADEON_GEM_DOMAIN_VRAM,
+	r = radeon_bo_pin_restricted(new_rbo, RADEON_GEM_DOMAIN_VRAM,
 				     ASIC_IS_AVIVO(rdev) ? 0 : 1 << 27, &base);
 	if (unlikely(r != 0)) {
-		radeon_bo_unreserve(rbo);
+		radeon_bo_unreserve(new_rbo);
 		r = -EINVAL;
 		DRM_ERROR("failed to pin new rbo buffer before flip\n");
-		goto pflip_cleanup;
+		goto cleanup;
 	}
-	radeon_bo_get_tiling_flags(rbo, &tiling_flags, NULL);
-	radeon_bo_unreserve(rbo);
+	radeon_bo_get_tiling_flags(new_rbo, &tiling_flags, NULL);
+	radeon_bo_unreserve(new_rbo);
 
 	if (!ASIC_IS_AVIVO(rdev)) {
 		/* crtc offset is from display base addr not FB location */
@@ -454,41 +523,49 @@ static int radeon_crtc_page_flip(struct drm_crtc *crtc,
 		}
 		base &= ~7;
 	}
+	work->base = base;
 
-	spin_lock_irqsave(&dev->event_lock, flags);
-	work->new_crtc_base = base;
-	spin_unlock_irqrestore(&dev->event_lock, flags);
-
-	/* update crtc fb */
-	crtc->fb = fb;
-
-	r = drm_vblank_get(dev, radeon_crtc->crtc_id);
+	r = drm_vblank_get(crtc->dev, radeon_crtc->crtc_id);
 	if (r) {
 		DRM_ERROR("failed to get vblank before flip\n");
-		goto pflip_cleanup1;
+		goto pflip_cleanup;
 	}
 
-	/* set the proper interrupt */
-	radeon_pre_page_flip(rdev, radeon_crtc->crtc_id);
+	/* We borrow the event spin lock for protecting flip_work */
+	spin_lock_irqsave(&crtc->dev->event_lock, flags);
+
+	if (radeon_crtc->flip_status != RADEON_FLIP_NONE) {
+		DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
+		spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
+		r = -EBUSY;
+		goto vblank_cleanup;
+	}
+	radeon_crtc->flip_status = RADEON_FLIP_PENDING;
+	radeon_crtc->flip_work = work;
+
+	/* update crtc fb */
+	crtc->primary->fb = fb;
+
+	spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
 
+	queue_work(radeon_crtc->flip_queue, &work->flip_work);
 	return 0;
 
-pflip_cleanup1:
-	if (unlikely(radeon_bo_reserve(rbo, false) != 0)) {
+vblank_cleanup:
+	drm_vblank_put(crtc->dev, radeon_crtc->crtc_id);
+
+pflip_cleanup:
+	if (unlikely(radeon_bo_reserve(new_rbo, false) != 0)) {
 		DRM_ERROR("failed to reserve new rbo in error path\n");
-		goto pflip_cleanup;
+		goto cleanup;
 	}
-	if (unlikely(radeon_bo_unpin(rbo) != 0)) {
+	if (unlikely(radeon_bo_unpin(new_rbo) != 0)) {
 		DRM_ERROR("failed to unpin new rbo in error path\n");
 	}
-	radeon_bo_unreserve(rbo);
+	radeon_bo_unreserve(new_rbo);
 
-pflip_cleanup:
-	spin_lock_irqsave(&dev->event_lock, flags);
-	radeon_crtc->unpin_work = NULL;
-unlock_free:
-	spin_unlock_irqrestore(&dev->event_lock, flags);
-	drm_gem_object_unreference_unlocked(old_radeon_fb->obj);
+cleanup:
+	drm_gem_object_unreference_unlocked(&work->old_rbo->gem_base);
 	radeon_fence_unref(&work->fence);
 	kfree(work);
 
@@ -562,6 +639,7 @@ static void radeon_crtc_init(struct drm_device *dev, int index)
 
 	drm_mode_crtc_set_gamma_size(&radeon_crtc->base, 256);
 	radeon_crtc->crtc_id = index;
+	radeon_crtc->flip_queue = create_singlethread_workqueue("radeon-crtc");
 	rdev->mode_info.crtcs[index] = radeon_crtc;
 
 	if (rdev->family >= CHIP_BONAIRE) {
@@ -571,6 +649,8 @@ static void radeon_crtc_init(struct drm_device *dev, int index)
 		radeon_crtc->max_cursor_width = CURSOR_WIDTH;
 		radeon_crtc->max_cursor_height = CURSOR_HEIGHT;
 	}
+	dev->mode_config.cursor_width = radeon_crtc->max_cursor_width;
+	dev->mode_config.cursor_height = radeon_crtc->max_cursor_height;
 
 #if 0
 	radeon_crtc->mode_set.crtc = &radeon_crtc->base;
@@ -653,7 +733,7 @@ static void radeon_print_display_setup(struct drm_device *dev)
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
 		radeon_connector = to_radeon_connector(connector);
 		DRM_INFO("Connector %d:\n", i);
-		DRM_INFO("  %s\n", drm_get_connector_name(connector));
+		DRM_INFO("  %s\n", connector->name);
 		if (radeon_connector->hpd.hpd != RADEON_HPD_NONE)
 			DRM_INFO("  %s\n", hpd_names[radeon_connector->hpd.hpd]);
 		if (radeon_connector->ddc_bus) {
@@ -749,25 +829,28 @@ int radeon_ddc_get_modes(struct radeon_connector *radeon_connector)
 	struct radeon_device *rdev = dev->dev_private;
 	int ret = 0;
 
+	/* don't leak the edid if we already fetched it in detect() */
+	if (radeon_connector->edid)
+		goto got_edid;
+
 	/* on hw with routers, select right port */
 	if (radeon_connector->router.ddc_valid)
 		radeon_router_select_ddc_port(radeon_connector);
 
 	if (radeon_connector_encoder_get_dp_bridge_encoder_id(&radeon_connector->base) !=
 	    ENCODER_OBJECT_ID_NONE) {
-		struct radeon_connector_atom_dig *dig = radeon_connector->con_priv;
-
-		if (dig->dp_i2c_bus)
+		if (radeon_connector->ddc_bus->has_aux)
 			radeon_connector->edid = drm_get_edid(&radeon_connector->base,
-							      &dig->dp_i2c_bus->adapter);
+							      &radeon_connector->ddc_bus->aux.ddc);
 	} else if ((radeon_connector->base.connector_type == DRM_MODE_CONNECTOR_DisplayPort) ||
 		   (radeon_connector->base.connector_type == DRM_MODE_CONNECTOR_eDP)) {
 		struct radeon_connector_atom_dig *dig = radeon_connector->con_priv;
 
 		if ((dig->dp_sink_type == CONNECTOR_OBJECT_ID_DISPLAYPORT ||
-		     dig->dp_sink_type == CONNECTOR_OBJECT_ID_eDP) && dig->dp_i2c_bus)
+		     dig->dp_sink_type == CONNECTOR_OBJECT_ID_eDP) &&
+		    radeon_connector->ddc_bus->has_aux)
 			radeon_connector->edid = drm_get_edid(&radeon_connector->base,
-							      &dig->dp_i2c_bus->adapter);
+							      &radeon_connector->ddc_bus->aux.ddc);
 		else if (radeon_connector->ddc_bus && !radeon_connector->edid)
 			radeon_connector->edid = drm_get_edid(&radeon_connector->base,
 							      &radeon_connector->ddc_bus->adapter);
@@ -788,8 +871,10 @@ int radeon_ddc_get_modes(struct radeon_connector *radeon_connector)
 			radeon_connector->edid = radeon_bios_get_hardcoded_edid(rdev);
 	}
 	if (radeon_connector->edid) {
+got_edid:
 		drm_mode_connector_update_edid_property(&radeon_connector->base, radeon_connector->edid);
 		ret = drm_add_edid_modes(&radeon_connector->base, radeon_connector->edid);
+		drm_edid_to_eld(&radeon_connector->base, radeon_connector->edid);
 		return ret;
 	}
 	drm_mode_connector_update_edid_property(&radeon_connector->base, NULL);
@@ -797,66 +882,89 @@ int radeon_ddc_get_modes(struct radeon_connector *radeon_connector)
 }
 
 /* avivo */
-static void avivo_get_fb_div(struct radeon_pll *pll,
-			     u32 target_clock,
-			     u32 post_div,
-			     u32 ref_div,
-			     u32 *fb_div,
-			     u32 *frac_fb_div)
-{
-	u32 tmp = post_div * ref_div;
 
-	tmp *= target_clock;
-	*fb_div = tmp / pll->reference_freq;
-	*frac_fb_div = tmp % pll->reference_freq;
+/**
+ * avivo_reduce_ratio - fractional number reduction
+ *
+ * @nom: nominator
+ * @den: denominator
+ * @nom_min: minimum value for nominator
+ * @den_min: minimum value for denominator
+ *
+ * Find the greatest common divisor and apply it on both nominator and
+ * denominator, but make nominator and denominator are at least as large
+ * as their minimum values.
+ */
+static void avivo_reduce_ratio(unsigned *nom, unsigned *den,
+			       unsigned nom_min, unsigned den_min)
+{
+	unsigned tmp;
+
+	/* reduce the numbers to a simpler ratio */
+	tmp = gcd(*nom, *den);
+	*nom /= tmp;
+	*den /= tmp;
+
+	/* make sure nominator is large enough */
+        if (*nom < nom_min) {
+		tmp = DIV_ROUND_UP(nom_min, *nom);
+		*nom *= tmp;
+		*den *= tmp;
+	}
 
-        if (*fb_div > pll->max_feedback_div)
-		*fb_div = pll->max_feedback_div;
-        else if (*fb_div < pll->min_feedback_div)
-                *fb_div = pll->min_feedback_div;
+	/* make sure the denominator is large enough */
+	if (*den < den_min) {
+		tmp = DIV_ROUND_UP(den_min, *den);
+		*nom *= tmp;
+		*den *= tmp;
+	}
 }
 
-static u32 avivo_get_post_div(struct radeon_pll *pll,
-			      u32 target_clock)
+/**
+ * avivo_get_fb_ref_div - feedback and ref divider calculation
+ *
+ * @nom: nominator
+ * @den: denominator
+ * @post_div: post divider
+ * @fb_div_max: feedback divider maximum
+ * @ref_div_max: reference divider maximum
+ * @fb_div: resulting feedback divider
+ * @ref_div: resulting reference divider
+ *
+ * Calculate feedback and reference divider for a given post divider. Makes
+ * sure we stay within the limits.
+ */
+static void avivo_get_fb_ref_div(unsigned nom, unsigned den, unsigned post_div,
+				 unsigned fb_div_max, unsigned ref_div_max,
+				 unsigned *fb_div, unsigned *ref_div)
 {
-	u32 vco, post_div, tmp;
+	/* limit reference * post divider to a maximum */
+	ref_div_max = max(min(100 / post_div, ref_div_max), 1u);
 
-	if (pll->flags & RADEON_PLL_USE_POST_DIV)
-		return pll->post_div;
+	/* get matching reference and feedback divider */
+	*ref_div = min(max(DIV_ROUND_CLOSEST(den, post_div), 1u), ref_div_max);
+	*fb_div = DIV_ROUND_CLOSEST(nom * *ref_div * post_div, den);
 
-	if (pll->flags & RADEON_PLL_PREFER_MINM_OVER_MAXP) {
-		if (pll->flags & RADEON_PLL_IS_LCD)
-			vco = pll->lcd_pll_out_min;
-		else
-			vco = pll->pll_out_min;
-	} else {
-		if (pll->flags & RADEON_PLL_IS_LCD)
-			vco = pll->lcd_pll_out_max;
-		else
-			vco = pll->pll_out_max;
-	}
-
-	post_div = vco / target_clock;
-	tmp = vco % target_clock;
-
-	if (pll->flags & RADEON_PLL_PREFER_MINM_OVER_MAXP) {
-		if (tmp)
-			post_div++;
-	} else {
-		if (!tmp)
-			post_div--;
+	/* limit fb divider to its maximum */
+        if (*fb_div > fb_div_max) {
+		*ref_div = DIV_ROUND_CLOSEST(*ref_div * fb_div_max, *fb_div);
+		*fb_div = fb_div_max;
 	}
-
-	if (post_div > pll->max_post_div)
-		post_div = pll->max_post_div;
-	else if (post_div < pll->min_post_div)
-		post_div = pll->min_post_div;
-
-	return post_div;
 }
 
-#define MAX_TOLERANCE 10
-
+/**
+ * radeon_compute_pll_avivo - compute PLL paramaters
+ *
+ * @pll: information about the PLL
+ * @dot_clock_p: resulting pixel clock
+ * fb_div_p: resulting feedback divider
+ * frac_fb_div_p: fractional part of the feedback divider
+ * ref_div_p: resulting reference divider
+ * post_div_p: resulting reference divider
+ *
+ * Try to calculate the PLL parameters to generate the given frequency:
+ * dot_clock = (ref_freq * feedback_div) / (ref_div * post_div)
+ */
 void radeon_compute_pll_avivo(struct radeon_pll *pll,
 			      u32 freq,
 			      u32 *dot_clock_p,
@@ -865,53 +973,135 @@ void radeon_compute_pll_avivo(struct radeon_pll *pll,
 			      u32 *ref_div_p,
 			      u32 *post_div_p)
 {
-	u32 target_clock = freq / 10;
-	u32 post_div = avivo_get_post_div(pll, target_clock);
-	u32 ref_div = pll->min_ref_div;
-	u32 fb_div = 0, frac_fb_div = 0, tmp;
+	unsigned target_clock = pll->flags & RADEON_PLL_USE_FRAC_FB_DIV ?
+		freq : freq / 10;
 
-	if (pll->flags & RADEON_PLL_USE_REF_DIV)
-		ref_div = pll->reference_div;
+	unsigned fb_div_min, fb_div_max, fb_div;
+	unsigned post_div_min, post_div_max, post_div;
+	unsigned ref_div_min, ref_div_max, ref_div;
+	unsigned post_div_best, diff_best;
+	unsigned nom, den;
+
+	/* determine allowed feedback divider range */
+	fb_div_min = pll->min_feedback_div;
+	fb_div_max = pll->max_feedback_div;
 
 	if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV) {
-		avivo_get_fb_div(pll, target_clock, post_div, ref_div, &fb_div, &frac_fb_div);
-		frac_fb_div = (100 * frac_fb_div) / pll->reference_freq;
-		if (frac_fb_div >= 5) {
-			frac_fb_div -= 5;
-			frac_fb_div = frac_fb_div / 10;
-			frac_fb_div++;
+		fb_div_min *= 10;
+		fb_div_max *= 10;
+	}
+
+	/* determine allowed ref divider range */
+	if (pll->flags & RADEON_PLL_USE_REF_DIV)
+		ref_div_min = pll->reference_div;
+	else
+		ref_div_min = pll->min_ref_div;
+
+	if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV &&
+	    pll->flags & RADEON_PLL_USE_REF_DIV)
+		ref_div_max = pll->reference_div;
+	else
+		ref_div_max = pll->max_ref_div;
+
+	/* determine allowed post divider range */
+	if (pll->flags & RADEON_PLL_USE_POST_DIV) {
+		post_div_min = pll->post_div;
+		post_div_max = pll->post_div;
+	} else {
+		unsigned vco_min, vco_max;
+
+		if (pll->flags & RADEON_PLL_IS_LCD) {
+			vco_min = pll->lcd_pll_out_min;
+			vco_max = pll->lcd_pll_out_max;
+		} else {
+			vco_min = pll->pll_out_min;
+			vco_max = pll->pll_out_max;
 		}
-		if (frac_fb_div >= 10) {
-			fb_div++;
-			frac_fb_div = 0;
+
+		if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV) {
+			vco_min *= 10;
+			vco_max *= 10;
 		}
-	} else {
-		while (ref_div <= pll->max_ref_div) {
-			avivo_get_fb_div(pll, target_clock, post_div, ref_div,
-					 &fb_div, &frac_fb_div);
-			if (frac_fb_div >= (pll->reference_freq / 2))
-				fb_div++;
-			frac_fb_div = 0;
-			tmp = (pll->reference_freq * fb_div) / (post_div * ref_div);
-			tmp = (tmp * 10000) / target_clock;
-
-			if (tmp > (10000 + MAX_TOLERANCE))
-				ref_div++;
-			else if (tmp >= (10000 - MAX_TOLERANCE))
-				break;
-			else
-				ref_div++;
+
+		post_div_min = vco_min / target_clock;
+		if ((target_clock * post_div_min) < vco_min)
+			++post_div_min;
+		if (post_div_min < pll->min_post_div)
+			post_div_min = pll->min_post_div;
+
+		post_div_max = vco_max / target_clock;
+		if ((target_clock * post_div_max) > vco_max)
+			--post_div_max;
+		if (post_div_max > pll->max_post_div)
+			post_div_max = pll->max_post_div;
+	}
+
+	/* represent the searched ratio as fractional number */
+	nom = target_clock;
+	den = pll->reference_freq;
+
+	/* reduce the numbers to a simpler ratio */
+	avivo_reduce_ratio(&nom, &den, fb_div_min, post_div_min);
+
+	/* now search for a post divider */
+	if (pll->flags & RADEON_PLL_PREFER_MINM_OVER_MAXP)
+		post_div_best = post_div_min;
+	else
+		post_div_best = post_div_max;
+	diff_best = ~0;
+
+	for (post_div = post_div_min; post_div <= post_div_max; ++post_div) {
+		unsigned diff;
+		avivo_get_fb_ref_div(nom, den, post_div, fb_div_max,
+				     ref_div_max, &fb_div, &ref_div);
+		diff = abs(target_clock - (pll->reference_freq * fb_div) /
+			(ref_div * post_div));
+
+		if (diff < diff_best || (diff == diff_best &&
+		    !(pll->flags & RADEON_PLL_PREFER_MINM_OVER_MAXP))) {
+
+			post_div_best = post_div;
+			diff_best = diff;
+		}
+	}
+	post_div = post_div_best;
+
+	/* get the feedback and reference divider for the optimal value */
+	avivo_get_fb_ref_div(nom, den, post_div, fb_div_max, ref_div_max,
+			     &fb_div, &ref_div);
+
+	/* reduce the numbers to a simpler ratio once more */
+	/* this also makes sure that the reference divider is large enough */
+	avivo_reduce_ratio(&fb_div, &ref_div, fb_div_min, ref_div_min);
+
+	/* avoid high jitter with small fractional dividers */
+	if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV && (fb_div % 10)) {
+		fb_div_min = max(fb_div_min, (9 - (fb_div % 10)) * 20 + 50);
+		if (fb_div < fb_div_min) {
+			unsigned tmp = DIV_ROUND_UP(fb_div_min, fb_div);
+			fb_div *= tmp;
+			ref_div *= tmp;
 		}
 	}
 
-	*dot_clock_p = ((pll->reference_freq * fb_div * 10) + (pll->reference_freq * frac_fb_div)) /
-		(ref_div * post_div * 10);
-	*fb_div_p = fb_div;
-	*frac_fb_div_p = frac_fb_div;
+	/* and finally save the result */
+	if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV) {
+		*fb_div_p = fb_div / 10;
+		*frac_fb_div_p = fb_div % 10;
+	} else {
+		*fb_div_p = fb_div;
+		*frac_fb_div_p = 0;
+	}
+
+	*dot_clock_p = ((pll->reference_freq * *fb_div_p * 10) +
+			(pll->reference_freq * *frac_fb_div_p)) /
+		       (ref_div * post_div * 10);
 	*ref_div_p = ref_div;
 	*post_div_p = post_div;
-	DRM_DEBUG_KMS("%d, pll dividers - fb: %d.%d ref: %d, post %d\n",
-		      *dot_clock_p, fb_div, frac_fb_div, ref_div, post_div);
+
+	DRM_DEBUG_KMS("%d - %d, pll dividers - fb: %d.%d ref: %d, post %d\n",
+		      freq, *dot_clock_p * 10, *fb_div_p, *frac_fb_div_p,
+		      ref_div, post_div);
 }
 
 /* pre-avivo */
@@ -1464,12 +1654,22 @@ int radeon_modeset_init(struct radeon_device *rdev)
 	/* setup afmt */
 	radeon_afmt_init(rdev);
 
-	/* Initialize power management */
-	radeon_pm_init(rdev);
-
 	radeon_fbdev_init(rdev);
 	drm_kms_helper_poll_init(rdev->ddev);
 
+	if (rdev->pm.dpm_enabled) {
+		/* do dpm late init */
+		ret = radeon_pm_late_init(rdev);
+		if (ret) {
+			rdev->pm.dpm_enabled = false;
+			DRM_ERROR("radeon_pm_late_init failed, disabling dpm\n");
+		}
+		/* set the dpm state for PX since there won't be
+		 * a modeset to call this.
+		 */
+		radeon_pm_compute_clocks(rdev);
+	}
+
 	return 0;
 }
 
@@ -1477,7 +1677,6 @@ void radeon_modeset_fini(struct radeon_device *rdev)
 {
 	radeon_fbdev_fini(rdev);
 	kfree(rdev->mode_info.bios_hardcoded_edid);
-	radeon_pm_fini(rdev);
 
 	if (rdev->mode_info.mode_config_initialized) {
 		radeon_afmt_fini(rdev);
@@ -1601,6 +1800,7 @@ bool radeon_crtc_scaling_mode_fixup(struct drm_crtc *crtc,
  *
  * \param dev Device to query.
  * \param crtc Crtc to query.
+ * \param flags Flags from caller (DRM_CALLED_FROM_VBLIRQ or 0).
  * \param *vpos Location where vertical scanout position should be stored.
  * \param *hpos Location where horizontal scanout position should go.
  * \param *stime Target location for timestamp taken immediately before
@@ -1622,8 +1822,8 @@ bool radeon_crtc_scaling_mode_fixup(struct drm_crtc *crtc,
  * unknown small number of scanlines wrt. real scanout position.
  *
  */
-int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, int *vpos, int *hpos,
-			       ktime_t *stime, ktime_t *etime)
+int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, unsigned int flags,
+			       int *vpos, int *hpos, ktime_t *stime, ktime_t *etime)
 {
 	u32 stat_crtc = 0, vbl = 0, position = 0;
 	int vbl_start, vbl_end, vtotal, ret = 0;
@@ -1765,5 +1965,27 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, int *vpos, int
 	if (in_vbl)
 		ret |= DRM_SCANOUTPOS_INVBL;
 
+	/* Is vpos outside nominal vblank area, but less than
+	 * 1/100 of a frame height away from start of vblank?
+	 * If so, assume this isn't a massively delayed vblank
+	 * interrupt, but a vblank interrupt that fired a few
+	 * microseconds before true start of vblank. Compensate
+	 * by adding a full frame duration to the final timestamp.
+	 * Happens, e.g., on ATI R500, R600.
+	 *
+	 * We only do this if DRM_CALLED_FROM_VBLIRQ.
+	 */
+	if ((flags & DRM_CALLED_FROM_VBLIRQ) && !in_vbl) {
+		vbl_start = rdev->mode_info.crtcs[crtc]->base.hwmode.crtc_vdisplay;
+		vtotal = rdev->mode_info.crtcs[crtc]->base.hwmode.crtc_vtotal;
+
+		if (vbl_start - *vpos < vtotal / 100) {
+			*vpos -= vtotal;
+
+			/* Signal this correction as "applied". */
+			ret |= 0x8;
+		}
+	}
+
 	return ret;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index db39ea36bf2..e9e36108424 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -78,9 +78,13 @@
  *   2.34.0 - Add CIK tiling mode array query
  *   2.35.0 - Add CIK macrotile mode array query
  *   2.36.0 - Fix CIK DCE tiling setup
+ *   2.37.0 - allow GS ring setup on r6xx/r7xx
+ *   2.38.0 - RADEON_GEM_OP (GET_INITIAL_DOMAIN, SET_INITIAL_DOMAIN),
+ *            CIK: 1D and linear tiling modes contain valid PIPE_CONFIG
+ *   2.39.0 - Add INFO query for number of active CUs
  */
 #define KMS_DRIVER_MAJOR	2
-#define KMS_DRIVER_MINOR	36
+#define KMS_DRIVER_MINOR	39
 #define KMS_DRIVER_PATCHLEVEL	0
 int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags);
 int radeon_driver_unload_kms(struct drm_device *dev);
@@ -102,15 +106,17 @@ int radeon_get_vblank_timestamp_kms(struct drm_device *dev, int crtc,
 void radeon_driver_irq_preinstall_kms(struct drm_device *dev);
 int radeon_driver_irq_postinstall_kms(struct drm_device *dev);
 void radeon_driver_irq_uninstall_kms(struct drm_device *dev);
-irqreturn_t radeon_driver_irq_handler_kms(DRM_IRQ_ARGS);
+irqreturn_t radeon_driver_irq_handler_kms(int irq, void *arg);
 void radeon_gem_object_free(struct drm_gem_object *obj);
 int radeon_gem_object_open(struct drm_gem_object *obj,
 				struct drm_file *file_priv);
 void radeon_gem_object_close(struct drm_gem_object *obj,
 				struct drm_file *file_priv);
 extern int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc,
+				      unsigned int flags,
 				      int *vpos, int *hpos, ktime_t *stime,
 				      ktime_t *etime);
+extern bool radeon_is_px(struct drm_device *dev);
 extern const struct drm_ioctl_desc radeon_ioctls_kms[];
 extern int radeon_max_kms_ioctl;
 int radeon_mmap(struct file *filp, struct vm_area_struct *vma);
@@ -140,11 +146,9 @@ void radeon_debugfs_cleanup(struct drm_minor *minor);
 #if defined(CONFIG_VGA_SWITCHEROO)
 void radeon_register_atpx_handler(void);
 void radeon_unregister_atpx_handler(void);
-bool radeon_is_px(void);
 #else
 static inline void radeon_register_atpx_handler(void) {}
 static inline void radeon_unregister_atpx_handler(void) {}
-static inline bool radeon_is_px(void) { return false; }
 #endif
 
 int radeon_no_wb;
@@ -168,6 +172,10 @@ int radeon_fastfb = 0;
 int radeon_dpm = -1;
 int radeon_aspm = -1;
 int radeon_runtime_pm = -1;
+int radeon_hard_reset = 0;
+int radeon_vm_size = 4;
+int radeon_vm_block_size = 9;
+int radeon_deep_color = 0;
 
 MODULE_PARM_DESC(no_wb, "Disable AGP writeback for scratch registers");
 module_param_named(no_wb, radeon_no_wb, int, 0444);
@@ -181,7 +189,7 @@ module_param_named(dynclks, radeon_dynclks, int, 0444);
 MODULE_PARM_DESC(r4xx_atom, "Enable ATOMBIOS modesetting for R4xx");
 module_param_named(r4xx_atom, radeon_r4xx_atom, int, 0444);
 
-MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing");
+MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes");
 module_param_named(vramlimit, radeon_vram_limit, int, 0600);
 
 MODULE_PARM_DESC(agpmode, "AGP Mode (-1 == PCI)");
@@ -232,6 +240,18 @@ module_param_named(aspm, radeon_aspm, int, 0444);
 MODULE_PARM_DESC(runpm, "PX runtime pm (1 = force enable, 0 = disable, -1 = PX only default)");
 module_param_named(runpm, radeon_runtime_pm, int, 0444);
 
+MODULE_PARM_DESC(hard_reset, "PCI config reset (1 = force enable, 0 = disable (default))");
+module_param_named(hard_reset, radeon_hard_reset, int, 0444);
+
+MODULE_PARM_DESC(vm_size, "VM address space size in gigabytes (default 4GB)");
+module_param_named(vm_size, radeon_vm_size, int, 0444);
+
+MODULE_PARM_DESC(vm_block_size, "VM page table size in bits (default 9)");
+module_param_named(vm_block_size, radeon_vm_block_size, int, 0444);
+
+MODULE_PARM_DESC(deep_color, "Deep Color support (1 = enable, 0 = disable (default))");
+module_param_named(deep_color, radeon_deep_color, int, 0444);
+
 static struct pci_device_id pciidlist[] = {
 	radeon_PCI_IDS
 };
@@ -397,8 +417,10 @@ static int radeon_pmops_runtime_suspend(struct device *dev)
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
 	int ret;
 
-	if (radeon_runtime_pm == 0)
-		return -EINVAL;
+	if (!radeon_is_px(drm_dev)) {
+		pm_runtime_forbid(dev);
+		return -EBUSY;
+	}
 
 	drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
 	drm_kms_helper_poll_disable(drm_dev);
@@ -419,7 +441,7 @@ static int radeon_pmops_runtime_resume(struct device *dev)
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
 	int ret;
 
-	if (radeon_runtime_pm == 0)
+	if (!radeon_is_px(drm_dev))
 		return -EINVAL;
 
 	drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
@@ -444,12 +466,8 @@ static int radeon_pmops_runtime_idle(struct device *dev)
 	struct drm_device *drm_dev = pci_get_drvdata(pdev);
 	struct drm_crtc *crtc;
 
-	if (radeon_runtime_pm == 0)
-		return -EBUSY;
-
-	/* are we PX enabled? */
-	if (radeon_runtime_pm == -1 && !radeon_is_px()) {
-		DRM_DEBUG_DRIVER("failing to power off - not px\n");
+	if (!radeon_is_px(drm_dev)) {
+		pm_runtime_forbid(dev);
 		return -EBUSY;
 	}
 
@@ -514,7 +532,6 @@ static struct drm_driver kms_driver = {
 	    DRIVER_USE_AGP |
 	    DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED | DRIVER_GEM |
 	    DRIVER_PRIME | DRIVER_RENDER,
-	.dev_priv_size = 0,
 	.load = radeon_driver_load_kms,
 	.open = radeon_driver_open_kms,
 	.preclose = radeon_driver_preclose_kms,
diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h
index 00e0d449021..dafd812e457 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.h
+++ b/drivers/gpu/drm/radeon/radeon_drv.h
@@ -405,7 +405,7 @@ extern void radeon_do_release(struct drm_device * dev);
 extern u32 radeon_get_vblank_counter(struct drm_device *dev, int crtc);
 extern int radeon_enable_vblank(struct drm_device *dev, int crtc);
 extern void radeon_disable_vblank(struct drm_device *dev, int crtc);
-extern irqreturn_t radeon_driver_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t radeon_driver_irq_handler(int irq, void *arg);
 extern void radeon_driver_irq_preinstall(struct drm_device * dev);
 extern int radeon_driver_irq_postinstall(struct drm_device *dev);
 extern void radeon_driver_irq_uninstall(struct drm_device * dev);
diff --git a/drivers/gpu/drm/radeon/radeon_family.h b/drivers/gpu/drm/radeon/radeon_family.h
index 614ad549297..4b7b87f71a6 100644
--- a/drivers/gpu/drm/radeon/radeon_family.h
+++ b/drivers/gpu/drm/radeon/radeon_family.h
@@ -97,6 +97,7 @@ enum radeon_family {
 	CHIP_KAVERI,
 	CHIP_KABINI,
 	CHIP_HAWAII,
+	CHIP_MULLINS,
 	CHIP_LAST,
 };
 
@@ -115,6 +116,7 @@ enum radeon_chip_flags {
 	RADEON_NEW_MEMMAP = 0x00400000UL,
 	RADEON_IS_PCI = 0x00800000UL,
 	RADEON_IS_IGPGART = 0x01000000UL,
+	RADEON_IS_PX = 0x02000000UL,
 };
 
 #endif
diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c
index d3a86e43c01..913787085df 100644
--- a/drivers/gpu/drm/radeon/radeon_fence.c
+++ b/drivers/gpu/drm/radeon/radeon_fence.c
@@ -121,7 +121,7 @@ int radeon_fence_emit(struct radeon_device *rdev,
 	(*fence)->seq = ++rdev->fence_drv[ring].sync_seq[ring];
 	(*fence)->ring = ring;
 	radeon_fence_ring_emit(rdev, ring, *fence);
-	trace_radeon_fence_emit(rdev->ddev, (*fence)->seq);
+	trace_radeon_fence_emit(rdev->ddev, ring, (*fence)->seq);
 	return 0;
 }
 
@@ -288,7 +288,6 @@ static bool radeon_fence_any_seq_signaled(struct radeon_device *rdev, u64 *seq)
  * @rdev: radeon device pointer
  * @target_seq: sequence number(s) we want to wait for
  * @intr: use interruptable sleep
- * @lock_ring: whether the ring should be locked or not
  *
  * Wait for the requested sequence number(s) to be written by any ring
  * (all asics).  Sequnce number array is indexed by ring id.
@@ -299,7 +298,7 @@ static bool radeon_fence_any_seq_signaled(struct radeon_device *rdev, u64 *seq)
  * -EDEADLK is returned when a GPU lockup has been detected.
  */
 static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
-				 bool intr, bool lock_ring)
+				 bool intr)
 {
 	uint64_t last_seq[RADEON_NUM_RINGS];
 	bool signaled;
@@ -313,7 +312,7 @@ static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
 				continue;
 
 			last_seq[i] = atomic64_read(&rdev->fence_drv[i].last_seq);
-			trace_radeon_fence_wait_begin(rdev->ddev, target_seq[i]);
+			trace_radeon_fence_wait_begin(rdev->ddev, i, target_seq[i]);
 			radeon_irq_kms_sw_irq_get(rdev, i);
 		}
 
@@ -332,7 +331,7 @@ static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
 				continue;
 
 			radeon_irq_kms_sw_irq_put(rdev, i);
-			trace_radeon_fence_wait_end(rdev->ddev, target_seq[i]);
+			trace_radeon_fence_wait_end(rdev->ddev, i, target_seq[i]);
 		}
 
 		if (unlikely(r < 0))
@@ -358,9 +357,6 @@ static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
 			if (i != RADEON_NUM_RINGS)
 				continue;
 
-			if (lock_ring)
-				mutex_lock(&rdev->ring_lock);
-
 			for (i = 0; i < RADEON_NUM_RINGS; ++i) {
 				if (!target_seq[i])
 					continue;
@@ -378,14 +374,9 @@ static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
 
 				/* remember that we need an reset */
 				rdev->needs_reset = true;
-				if (lock_ring)
-					mutex_unlock(&rdev->ring_lock);
 				wake_up_all(&rdev->fence_queue);
 				return -EDEADLK;
 			}
-
-			if (lock_ring)
-				mutex_unlock(&rdev->ring_lock);
 		}
 	}
 	return 0;
@@ -416,7 +407,7 @@ int radeon_fence_wait(struct radeon_fence *fence, bool intr)
 	if (seq[fence->ring] == RADEON_FENCE_SIGNALED_SEQ)
 		return 0;
 
-	r = radeon_fence_wait_seq(fence->rdev, seq, intr, true);
+	r = radeon_fence_wait_seq(fence->rdev, seq, intr);
 	if (r)
 		return r;
 
@@ -464,7 +455,7 @@ int radeon_fence_wait_any(struct radeon_device *rdev,
 	if (num_rings == 0)
 		return -ENOENT;
 
-	r = radeon_fence_wait_seq(rdev, seq, intr, true);
+	r = radeon_fence_wait_seq(rdev, seq, intr);
 	if (r) {
 		return r;
 	}
@@ -472,37 +463,7 @@ int radeon_fence_wait_any(struct radeon_device *rdev,
 }
 
 /**
- * radeon_fence_wait_locked - wait for a fence to signal
- *
- * @fence: radeon fence object
- *
- * Wait for the requested fence to signal (all asics).
- * Returns 0 if the fence has passed, error for all other cases.
- */
-int radeon_fence_wait_locked(struct radeon_fence *fence)
-{
-	uint64_t seq[RADEON_NUM_RINGS] = {};
-	int r;
-
-	if (fence == NULL) {
-		WARN(1, "Querying an invalid fence : %p !\n", fence);
-		return -EINVAL;
-	}
-
-	seq[fence->ring] = fence->seq;
-	if (seq[fence->ring] == RADEON_FENCE_SIGNALED_SEQ)
-		return 0;
-
-	r = radeon_fence_wait_seq(fence->rdev, seq, false, false);
-	if (r)
-		return r;
-
-	fence->seq = RADEON_FENCE_SIGNALED_SEQ;
-	return 0;
-}
-
-/**
- * radeon_fence_wait_next_locked - wait for the next fence to signal
+ * radeon_fence_wait_next - wait for the next fence to signal
  *
  * @rdev: radeon device pointer
  * @ring: ring index the fence is associated with
@@ -511,7 +472,7 @@ int radeon_fence_wait_locked(struct radeon_fence *fence)
  * Returns 0 if the next fence has passed, error for all other cases.
  * Caller must hold ring lock.
  */
-int radeon_fence_wait_next_locked(struct radeon_device *rdev, int ring)
+int radeon_fence_wait_next(struct radeon_device *rdev, int ring)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
 
@@ -521,11 +482,11 @@ int radeon_fence_wait_next_locked(struct radeon_device *rdev, int ring)
 		   already the last emited fence */
 		return -ENOENT;
 	}
-	return radeon_fence_wait_seq(rdev, seq, false, false);
+	return radeon_fence_wait_seq(rdev, seq, false);
 }
 
 /**
- * radeon_fence_wait_empty_locked - wait for all fences to signal
+ * radeon_fence_wait_empty - wait for all fences to signal
  *
  * @rdev: radeon device pointer
  * @ring: ring index the fence is associated with
@@ -534,7 +495,7 @@ int radeon_fence_wait_next_locked(struct radeon_device *rdev, int ring)
  * Returns 0 if the fences have passed, error for all other cases.
  * Caller must hold ring lock.
  */
-int radeon_fence_wait_empty_locked(struct radeon_device *rdev, int ring)
+int radeon_fence_wait_empty(struct radeon_device *rdev, int ring)
 {
 	uint64_t seq[RADEON_NUM_RINGS] = {};
 	int r;
@@ -543,7 +504,7 @@ int radeon_fence_wait_empty_locked(struct radeon_device *rdev, int ring)
 	if (!seq[ring])
 		return 0;
 
-	r = radeon_fence_wait_seq(rdev, seq, false, false);
+	r = radeon_fence_wait_seq(rdev, seq, false);
 	if (r) {
 		if (r == -EDEADLK)
 			return -EDEADLK;
@@ -794,7 +755,7 @@ void radeon_fence_driver_fini(struct radeon_device *rdev)
 	for (ring = 0; ring < RADEON_NUM_RINGS; ring++) {
 		if (!rdev->fence_drv[ring].initialized)
 			continue;
-		r = radeon_fence_wait_empty_locked(rdev, ring);
+		r = radeon_fence_wait_empty(rdev, ring);
 		if (r) {
 			/* no need to trigger GPU reset as we are unloading */
 			radeon_fence_driver_force_completion(rdev);
@@ -841,6 +802,8 @@ static int radeon_debugfs_fence_info(struct seq_file *m, void *data)
 		if (!rdev->fence_drv[i].initialized)
 			continue;
 
+		radeon_fence_process(rdev, i);
+
 		seq_printf(m, "--- ring %d ---\n", i);
 		seq_printf(m, "Last signaled fence 0x%016llx\n",
 			   (unsigned long long)atomic64_read(&rdev->fence_drv[i].last_seq));
@@ -856,15 +819,35 @@ static int radeon_debugfs_fence_info(struct seq_file *m, void *data)
 	return 0;
 }
 
+/**
+ * radeon_debugfs_gpu_reset - manually trigger a gpu reset
+ *
+ * Manually trigger a gpu reset at the next fence wait.
+ */
+static int radeon_debugfs_gpu_reset(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct radeon_device *rdev = dev->dev_private;
+
+	down_read(&rdev->exclusive_lock);
+	seq_printf(m, "%d\n", rdev->needs_reset);
+	rdev->needs_reset = true;
+	up_read(&rdev->exclusive_lock);
+
+	return 0;
+}
+
 static struct drm_info_list radeon_debugfs_fence_list[] = {
 	{"radeon_fence_info", &radeon_debugfs_fence_info, 0, NULL},
+	{"radeon_gpu_reset", &radeon_debugfs_gpu_reset, 0, NULL}
 };
 #endif
 
 int radeon_debugfs_fence_init(struct radeon_device *rdev)
 {
 #if defined(CONFIG_DEBUG_FS)
-	return radeon_debugfs_add_files(rdev, radeon_debugfs_fence_list, 1);
+	return radeon_debugfs_add_files(rdev, radeon_debugfs_fence_list, 2);
 #else
 	return 0;
 #endif
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index 96e440061bd..2e723651069 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -28,8 +28,6 @@
 #include <drm/drmP.h>
 #include <drm/radeon_drm.h>
 #include "radeon.h"
-#include "radeon_reg.h"
-#include "radeon_trace.h"
 
 /*
  * GART
@@ -394,952 +392,3 @@ void radeon_gart_fini(struct radeon_device *rdev)
 
 	radeon_dummy_page_fini(rdev);
 }
-
-/*
- * GPUVM
- * GPUVM is similar to the legacy gart on older asics, however
- * rather than there being a single global gart table
- * for the entire GPU, there are multiple VM page tables active
- * at any given time.  The VM page tables can contain a mix
- * vram pages and system memory pages and system memory pages
- * can be mapped as snooped (cached system pages) or unsnooped
- * (uncached system pages).
- * Each VM has an ID associated with it and there is a page table
- * associated with each VMID.  When execting a command buffer,
- * the kernel tells the the ring what VMID to use for that command
- * buffer.  VMIDs are allocated dynamically as commands are submitted.
- * The userspace drivers maintain their own address space and the kernel
- * sets up their pages tables accordingly when they submit their
- * command buffers and a VMID is assigned.
- * Cayman/Trinity support up to 8 active VMs at any given time;
- * SI supports 16.
- */
-
-/*
- * vm helpers
- *
- * TODO bind a default page at vm initialization for default address
- */
-
-/**
- * radeon_vm_num_pde - return the number of page directory entries
- *
- * @rdev: radeon_device pointer
- *
- * Calculate the number of page directory entries (cayman+).
- */
-static unsigned radeon_vm_num_pdes(struct radeon_device *rdev)
-{
-	return rdev->vm_manager.max_pfn >> RADEON_VM_BLOCK_SIZE;
-}
-
-/**
- * radeon_vm_directory_size - returns the size of the page directory in bytes
- *
- * @rdev: radeon_device pointer
- *
- * Calculate the size of the page directory in bytes (cayman+).
- */
-static unsigned radeon_vm_directory_size(struct radeon_device *rdev)
-{
-	return RADEON_GPU_PAGE_ALIGN(radeon_vm_num_pdes(rdev) * 8);
-}
-
-/**
- * radeon_vm_manager_init - init the vm manager
- *
- * @rdev: radeon_device pointer
- *
- * Init the vm manager (cayman+).
- * Returns 0 for success, error for failure.
- */
-int radeon_vm_manager_init(struct radeon_device *rdev)
-{
-	struct radeon_vm *vm;
-	struct radeon_bo_va *bo_va;
-	int r;
-	unsigned size;
-
-	if (!rdev->vm_manager.enabled) {
-		/* allocate enough for 2 full VM pts */
-		size = radeon_vm_directory_size(rdev);
-		size += rdev->vm_manager.max_pfn * 8;
-		size *= 2;
-		r = radeon_sa_bo_manager_init(rdev, &rdev->vm_manager.sa_manager,
-					      RADEON_GPU_PAGE_ALIGN(size),
-					      RADEON_VM_PTB_ALIGN_SIZE,
-					      RADEON_GEM_DOMAIN_VRAM);
-		if (r) {
-			dev_err(rdev->dev, "failed to allocate vm bo (%dKB)\n",
-				(rdev->vm_manager.max_pfn * 8) >> 10);
-			return r;
-		}
-
-		r = radeon_asic_vm_init(rdev);
-		if (r)
-			return r;
-
-		rdev->vm_manager.enabled = true;
-
-		r = radeon_sa_bo_manager_start(rdev, &rdev->vm_manager.sa_manager);
-		if (r)
-			return r;
-	}
-
-	/* restore page table */
-	list_for_each_entry(vm, &rdev->vm_manager.lru_vm, list) {
-		if (vm->page_directory == NULL)
-			continue;
-
-		list_for_each_entry(bo_va, &vm->va, vm_list) {
-			bo_va->valid = false;
-		}
-	}
-	return 0;
-}
-
-/**
- * radeon_vm_free_pt - free the page table for a specific vm
- *
- * @rdev: radeon_device pointer
- * @vm: vm to unbind
- *
- * Free the page table of a specific vm (cayman+).
- *
- * Global and local mutex must be lock!
- */
-static void radeon_vm_free_pt(struct radeon_device *rdev,
-				    struct radeon_vm *vm)
-{
-	struct radeon_bo_va *bo_va;
-	int i;
-
-	if (!vm->page_directory)
-		return;
-
-	list_del_init(&vm->list);
-	radeon_sa_bo_free(rdev, &vm->page_directory, vm->fence);
-
-	list_for_each_entry(bo_va, &vm->va, vm_list) {
-		bo_va->valid = false;
-	}
-
-	if (vm->page_tables == NULL)
-		return;
-
-	for (i = 0; i < radeon_vm_num_pdes(rdev); i++)
-		radeon_sa_bo_free(rdev, &vm->page_tables[i], vm->fence);
-
-	kfree(vm->page_tables);
-}
-
-/**
- * radeon_vm_manager_fini - tear down the vm manager
- *
- * @rdev: radeon_device pointer
- *
- * Tear down the VM manager (cayman+).
- */
-void radeon_vm_manager_fini(struct radeon_device *rdev)
-{
-	struct radeon_vm *vm, *tmp;
-	int i;
-
-	if (!rdev->vm_manager.enabled)
-		return;
-
-	mutex_lock(&rdev->vm_manager.lock);
-	/* free all allocated page tables */
-	list_for_each_entry_safe(vm, tmp, &rdev->vm_manager.lru_vm, list) {
-		mutex_lock(&vm->mutex);
-		radeon_vm_free_pt(rdev, vm);
-		mutex_unlock(&vm->mutex);
-	}
-	for (i = 0; i < RADEON_NUM_VM; ++i) {
-		radeon_fence_unref(&rdev->vm_manager.active[i]);
-	}
-	radeon_asic_vm_fini(rdev);
-	mutex_unlock(&rdev->vm_manager.lock);
-
-	radeon_sa_bo_manager_suspend(rdev, &rdev->vm_manager.sa_manager);
-	radeon_sa_bo_manager_fini(rdev, &rdev->vm_manager.sa_manager);
-	rdev->vm_manager.enabled = false;
-}
-
-/**
- * radeon_vm_evict - evict page table to make room for new one
- *
- * @rdev: radeon_device pointer
- * @vm: VM we want to allocate something for
- *
- * Evict a VM from the lru, making sure that it isn't @vm. (cayman+).
- * Returns 0 for success, -ENOMEM for failure.
- *
- * Global and local mutex must be locked!
- */
-static int radeon_vm_evict(struct radeon_device *rdev, struct radeon_vm *vm)
-{
-	struct radeon_vm *vm_evict;
-
-	if (list_empty(&rdev->vm_manager.lru_vm))
-		return -ENOMEM;
-
-	vm_evict = list_first_entry(&rdev->vm_manager.lru_vm,
-				    struct radeon_vm, list);
-	if (vm_evict == vm)
-		return -ENOMEM;
-
-	mutex_lock(&vm_evict->mutex);
-	radeon_vm_free_pt(rdev, vm_evict);
-	mutex_unlock(&vm_evict->mutex);
-	return 0;
-}
-
-/**
- * radeon_vm_alloc_pt - allocates a page table for a VM
- *
- * @rdev: radeon_device pointer
- * @vm: vm to bind
- *
- * Allocate a page table for the requested vm (cayman+).
- * Returns 0 for success, error for failure.
- *
- * Global and local mutex must be locked!
- */
-int radeon_vm_alloc_pt(struct radeon_device *rdev, struct radeon_vm *vm)
-{
-	unsigned pd_size, pd_entries, pts_size;
-	struct radeon_ib ib;
-	int r;
-
-	if (vm == NULL) {
-		return -EINVAL;
-	}
-
-	if (vm->page_directory != NULL) {
-		return 0;
-	}
-
-	pd_size = radeon_vm_directory_size(rdev);
-	pd_entries = radeon_vm_num_pdes(rdev);
-
-retry:
-	r = radeon_sa_bo_new(rdev, &rdev->vm_manager.sa_manager,
-			     &vm->page_directory, pd_size,
-			     RADEON_VM_PTB_ALIGN_SIZE, false);
-	if (r == -ENOMEM) {
-		r = radeon_vm_evict(rdev, vm);
-		if (r)
-			return r;
-		goto retry;
-
-	} else if (r) {
-		return r;
-	}
-
-	vm->pd_gpu_addr = radeon_sa_bo_gpu_addr(vm->page_directory);
-
-	/* Initially clear the page directory */
-	r = radeon_ib_get(rdev, R600_RING_TYPE_DMA_INDEX, &ib,
-			  NULL, pd_entries * 2 + 64);
-	if (r) {
-		radeon_sa_bo_free(rdev, &vm->page_directory, vm->fence);
-		return r;
-	}
-
-	ib.length_dw = 0;
-
-	radeon_asic_vm_set_page(rdev, &ib, vm->pd_gpu_addr,
-				0, pd_entries, 0, 0);
-
-	radeon_semaphore_sync_to(ib.semaphore, vm->fence);
-	r = radeon_ib_schedule(rdev, &ib, NULL);
-	if (r) {
-		radeon_ib_free(rdev, &ib);
-		radeon_sa_bo_free(rdev, &vm->page_directory, vm->fence);
-		return r;
-	}
-	radeon_fence_unref(&vm->fence);
-	vm->fence = radeon_fence_ref(ib.fence);
-	radeon_ib_free(rdev, &ib);
-	radeon_fence_unref(&vm->last_flush);
-
-	/* allocate page table array */
-	pts_size = radeon_vm_num_pdes(rdev) * sizeof(struct radeon_sa_bo *);
-	vm->page_tables = kzalloc(pts_size, GFP_KERNEL);
-
-	if (vm->page_tables == NULL) {
-		DRM_ERROR("Cannot allocate memory for page table array\n");
-		radeon_sa_bo_free(rdev, &vm->page_directory, vm->fence);
-		return -ENOMEM;
-	}
-
-	return 0;
-}
-
-/**
- * radeon_vm_add_to_lru - add VMs page table to LRU list
- *
- * @rdev: radeon_device pointer
- * @vm: vm to add to LRU
- *
- * Add the allocated page table to the LRU list (cayman+).
- *
- * Global mutex must be locked!
- */
-void radeon_vm_add_to_lru(struct radeon_device *rdev, struct radeon_vm *vm)
-{
-	list_del_init(&vm->list);
-	list_add_tail(&vm->list, &rdev->vm_manager.lru_vm);
-}
-
-/**
- * radeon_vm_grab_id - allocate the next free VMID
- *
- * @rdev: radeon_device pointer
- * @vm: vm to allocate id for
- * @ring: ring we want to submit job to
- *
- * Allocate an id for the vm (cayman+).
- * Returns the fence we need to sync to (if any).
- *
- * Global and local mutex must be locked!
- */
-struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
-				       struct radeon_vm *vm, int ring)
-{
-	struct radeon_fence *best[RADEON_NUM_RINGS] = {};
-	unsigned choices[2] = {};
-	unsigned i;
-
-	/* check if the id is still valid */
-	if (vm->fence && vm->fence == rdev->vm_manager.active[vm->id])
-		return NULL;
-
-	/* we definately need to flush */
-	radeon_fence_unref(&vm->last_flush);
-
-	/* skip over VMID 0, since it is the system VM */
-	for (i = 1; i < rdev->vm_manager.nvm; ++i) {
-		struct radeon_fence *fence = rdev->vm_manager.active[i];
-
-		if (fence == NULL) {
-			/* found a free one */
-			vm->id = i;
-			return NULL;
-		}
-
-		if (radeon_fence_is_earlier(fence, best[fence->ring])) {
-			best[fence->ring] = fence;
-			choices[fence->ring == ring ? 0 : 1] = i;
-		}
-	}
-
-	for (i = 0; i < 2; ++i) {
-		if (choices[i]) {
-			vm->id = choices[i];
-			trace_radeon_vm_grab_id(vm->id, ring);
-			return rdev->vm_manager.active[choices[i]];
-		}
-	}
-
-	/* should never happen */
-	BUG();
-	return NULL;
-}
-
-/**
- * radeon_vm_fence - remember fence for vm
- *
- * @rdev: radeon_device pointer
- * @vm: vm we want to fence
- * @fence: fence to remember
- *
- * Fence the vm (cayman+).
- * Set the fence used to protect page table and id.
- *
- * Global and local mutex must be locked!
- */
-void radeon_vm_fence(struct radeon_device *rdev,
-		     struct radeon_vm *vm,
-		     struct radeon_fence *fence)
-{
-	radeon_fence_unref(&rdev->vm_manager.active[vm->id]);
-	rdev->vm_manager.active[vm->id] = radeon_fence_ref(fence);
-
-	radeon_fence_unref(&vm->fence);
-	vm->fence = radeon_fence_ref(fence);
-}
-
-/**
- * radeon_vm_bo_find - find the bo_va for a specific vm & bo
- *
- * @vm: requested vm
- * @bo: requested buffer object
- *
- * Find @bo inside the requested vm (cayman+).
- * Search inside the @bos vm list for the requested vm
- * Returns the found bo_va or NULL if none is found
- *
- * Object has to be reserved!
- */
-struct radeon_bo_va *radeon_vm_bo_find(struct radeon_vm *vm,
-				       struct radeon_bo *bo)
-{
-	struct radeon_bo_va *bo_va;
-
-	list_for_each_entry(bo_va, &bo->va, bo_list) {
-		if (bo_va->vm == vm) {
-			return bo_va;
-		}
-	}
-	return NULL;
-}
-
-/**
- * radeon_vm_bo_add - add a bo to a specific vm
- *
- * @rdev: radeon_device pointer
- * @vm: requested vm
- * @bo: radeon buffer object
- *
- * Add @bo into the requested vm (cayman+).
- * Add @bo to the list of bos associated with the vm
- * Returns newly added bo_va or NULL for failure
- *
- * Object has to be reserved!
- */
-struct radeon_bo_va *radeon_vm_bo_add(struct radeon_device *rdev,
-				      struct radeon_vm *vm,
-				      struct radeon_bo *bo)
-{
-	struct radeon_bo_va *bo_va;
-
-	bo_va = kzalloc(sizeof(struct radeon_bo_va), GFP_KERNEL);
-	if (bo_va == NULL) {
-		return NULL;
-	}
-	bo_va->vm = vm;
-	bo_va->bo = bo;
-	bo_va->soffset = 0;
-	bo_va->eoffset = 0;
-	bo_va->flags = 0;
-	bo_va->valid = false;
-	bo_va->ref_count = 1;
-	INIT_LIST_HEAD(&bo_va->bo_list);
-	INIT_LIST_HEAD(&bo_va->vm_list);
-
-	mutex_lock(&vm->mutex);
-	list_add(&bo_va->vm_list, &vm->va);
-	list_add_tail(&bo_va->bo_list, &bo->va);
-	mutex_unlock(&vm->mutex);
-
-	return bo_va;
-}
-
-/**
- * radeon_vm_bo_set_addr - set bos virtual address inside a vm
- *
- * @rdev: radeon_device pointer
- * @bo_va: bo_va to store the address
- * @soffset: requested offset of the buffer in the VM address space
- * @flags: attributes of pages (read/write/valid/etc.)
- *
- * Set offset of @bo_va (cayman+).
- * Validate and set the offset requested within the vm address space.
- * Returns 0 for success, error for failure.
- *
- * Object has to be reserved!
- */
-int radeon_vm_bo_set_addr(struct radeon_device *rdev,
-			  struct radeon_bo_va *bo_va,
-			  uint64_t soffset,
-			  uint32_t flags)
-{
-	uint64_t size = radeon_bo_size(bo_va->bo);
-	uint64_t eoffset, last_offset = 0;
-	struct radeon_vm *vm = bo_va->vm;
-	struct radeon_bo_va *tmp;
-	struct list_head *head;
-	unsigned last_pfn;
-
-	if (soffset) {
-		/* make sure object fit at this offset */
-		eoffset = soffset + size;
-		if (soffset >= eoffset) {
-			return -EINVAL;
-		}
-
-		last_pfn = eoffset / RADEON_GPU_PAGE_SIZE;
-		if (last_pfn > rdev->vm_manager.max_pfn) {
-			dev_err(rdev->dev, "va above limit (0x%08X > 0x%08X)\n",
-				last_pfn, rdev->vm_manager.max_pfn);
-			return -EINVAL;
-		}
-
-	} else {
-		eoffset = last_pfn = 0;
-	}
-
-	mutex_lock(&vm->mutex);
-	head = &vm->va;
-	last_offset = 0;
-	list_for_each_entry(tmp, &vm->va, vm_list) {
-		if (bo_va == tmp) {
-			/* skip over currently modified bo */
-			continue;
-		}
-
-		if (soffset >= last_offset && eoffset <= tmp->soffset) {
-			/* bo can be added before this one */
-			break;
-		}
-		if (eoffset > tmp->soffset && soffset < tmp->eoffset) {
-			/* bo and tmp overlap, invalid offset */
-			dev_err(rdev->dev, "bo %p va 0x%08X conflict with (bo %p 0x%08X 0x%08X)\n",
-				bo_va->bo, (unsigned)bo_va->soffset, tmp->bo,
-				(unsigned)tmp->soffset, (unsigned)tmp->eoffset);
-			mutex_unlock(&vm->mutex);
-			return -EINVAL;
-		}
-		last_offset = tmp->eoffset;
-		head = &tmp->vm_list;
-	}
-
-	bo_va->soffset = soffset;
-	bo_va->eoffset = eoffset;
-	bo_va->flags = flags;
-	bo_va->valid = false;
-	list_move(&bo_va->vm_list, head);
-
-	mutex_unlock(&vm->mutex);
-	return 0;
-}
-
-/**
- * radeon_vm_map_gart - get the physical address of a gart page
- *
- * @rdev: radeon_device pointer
- * @addr: the unmapped addr
- *
- * Look up the physical address of the page that the pte resolves
- * to (cayman+).
- * Returns the physical address of the page.
- */
-uint64_t radeon_vm_map_gart(struct radeon_device *rdev, uint64_t addr)
-{
-	uint64_t result;
-
-	/* page table offset */
-	result = rdev->gart.pages_addr[addr >> PAGE_SHIFT];
-
-	/* in case cpu page size != gpu page size*/
-	result |= addr & (~PAGE_MASK);
-
-	return result;
-}
-
-/**
- * radeon_vm_page_flags - translate page flags to what the hw uses
- *
- * @flags: flags comming from userspace
- *
- * Translate the flags the userspace ABI uses to hw flags.
- */
-static uint32_t radeon_vm_page_flags(uint32_t flags)
-{
-        uint32_t hw_flags = 0;
-        hw_flags |= (flags & RADEON_VM_PAGE_VALID) ? R600_PTE_VALID : 0;
-        hw_flags |= (flags & RADEON_VM_PAGE_READABLE) ? R600_PTE_READABLE : 0;
-        hw_flags |= (flags & RADEON_VM_PAGE_WRITEABLE) ? R600_PTE_WRITEABLE : 0;
-        if (flags & RADEON_VM_PAGE_SYSTEM) {
-                hw_flags |= R600_PTE_SYSTEM;
-                hw_flags |= (flags & RADEON_VM_PAGE_SNOOPED) ? R600_PTE_SNOOPED : 0;
-        }
-        return hw_flags;
-}
-
-/**
- * radeon_vm_update_pdes - make sure that page directory is valid
- *
- * @rdev: radeon_device pointer
- * @vm: requested vm
- * @start: start of GPU address range
- * @end: end of GPU address range
- *
- * Allocates new page tables if necessary
- * and updates the page directory (cayman+).
- * Returns 0 for success, error for failure.
- *
- * Global and local mutex must be locked!
- */
-static int radeon_vm_update_pdes(struct radeon_device *rdev,
-				 struct radeon_vm *vm,
-				 struct radeon_ib *ib,
-				 uint64_t start, uint64_t end)
-{
-	static const uint32_t incr = RADEON_VM_PTE_COUNT * 8;
-
-	uint64_t last_pde = ~0, last_pt = ~0;
-	unsigned count = 0;
-	uint64_t pt_idx;
-	int r;
-
-	start = (start / RADEON_GPU_PAGE_SIZE) >> RADEON_VM_BLOCK_SIZE;
-	end = (end / RADEON_GPU_PAGE_SIZE) >> RADEON_VM_BLOCK_SIZE;
-
-	/* walk over the address space and update the page directory */
-	for (pt_idx = start; pt_idx <= end; ++pt_idx) {
-		uint64_t pde, pt;
-
-		if (vm->page_tables[pt_idx])
-			continue;
-
-retry:
-		r = radeon_sa_bo_new(rdev, &rdev->vm_manager.sa_manager,
-				     &vm->page_tables[pt_idx],
-				     RADEON_VM_PTE_COUNT * 8,
-				     RADEON_GPU_PAGE_SIZE, false);
-
-		if (r == -ENOMEM) {
-			r = radeon_vm_evict(rdev, vm);
-			if (r)
-				return r;
-			goto retry;
-		} else if (r) {
-			return r;
-		}
-
-		pde = vm->pd_gpu_addr + pt_idx * 8;
-
-		pt = radeon_sa_bo_gpu_addr(vm->page_tables[pt_idx]);
-
-		if (((last_pde + 8 * count) != pde) ||
-		    ((last_pt + incr * count) != pt)) {
-
-			if (count) {
-				radeon_asic_vm_set_page(rdev, ib, last_pde,
-							last_pt, count, incr,
-							R600_PTE_VALID);
-
-				count *= RADEON_VM_PTE_COUNT;
-				radeon_asic_vm_set_page(rdev, ib, last_pt, 0,
-							count, 0, 0);
-			}
-
-			count = 1;
-			last_pde = pde;
-			last_pt = pt;
-		} else {
-			++count;
-		}
-	}
-
-	if (count) {
-		radeon_asic_vm_set_page(rdev, ib, last_pde, last_pt, count,
-					incr, R600_PTE_VALID);
-
-		count *= RADEON_VM_PTE_COUNT;
-		radeon_asic_vm_set_page(rdev, ib, last_pt, 0,
-					count, 0, 0);
-	}
-
-	return 0;
-}
-
-/**
- * radeon_vm_update_ptes - make sure that page tables are valid
- *
- * @rdev: radeon_device pointer
- * @vm: requested vm
- * @start: start of GPU address range
- * @end: end of GPU address range
- * @dst: destination address to map to
- * @flags: mapping flags
- *
- * Update the page tables in the range @start - @end (cayman+).
- *
- * Global and local mutex must be locked!
- */
-static void radeon_vm_update_ptes(struct radeon_device *rdev,
-				  struct radeon_vm *vm,
-				  struct radeon_ib *ib,
-				  uint64_t start, uint64_t end,
-				  uint64_t dst, uint32_t flags)
-{
-	static const uint64_t mask = RADEON_VM_PTE_COUNT - 1;
-
-	uint64_t last_pte = ~0, last_dst = ~0;
-	unsigned count = 0;
-	uint64_t addr;
-
-	start = start / RADEON_GPU_PAGE_SIZE;
-	end = end / RADEON_GPU_PAGE_SIZE;
-
-	/* walk over the address space and update the page tables */
-	for (addr = start; addr < end; ) {
-		uint64_t pt_idx = addr >> RADEON_VM_BLOCK_SIZE;
-		unsigned nptes;
-		uint64_t pte;
-
-		if ((addr & ~mask) == (end & ~mask))
-			nptes = end - addr;
-		else
-			nptes = RADEON_VM_PTE_COUNT - (addr & mask);
-
-		pte = radeon_sa_bo_gpu_addr(vm->page_tables[pt_idx]);
-		pte += (addr & mask) * 8;
-
-		if ((last_pte + 8 * count) != pte) {
-
-			if (count) {
-				radeon_asic_vm_set_page(rdev, ib, last_pte,
-							last_dst, count,
-							RADEON_GPU_PAGE_SIZE,
-							flags);
-			}
-
-			count = nptes;
-			last_pte = pte;
-			last_dst = dst;
-		} else {
-			count += nptes;
-		}
-
-		addr += nptes;
-		dst += nptes * RADEON_GPU_PAGE_SIZE;
-	}
-
-	if (count) {
-		radeon_asic_vm_set_page(rdev, ib, last_pte,
-					last_dst, count,
-					RADEON_GPU_PAGE_SIZE, flags);
-	}
-}
-
-/**
- * radeon_vm_bo_update - map a bo into the vm page table
- *
- * @rdev: radeon_device pointer
- * @vm: requested vm
- * @bo: radeon buffer object
- * @mem: ttm mem
- *
- * Fill in the page table entries for @bo (cayman+).
- * Returns 0 for success, -EINVAL for failure.
- *
- * Object have to be reserved & global and local mutex must be locked!
- */
-int radeon_vm_bo_update(struct radeon_device *rdev,
-			struct radeon_vm *vm,
-			struct radeon_bo *bo,
-			struct ttm_mem_reg *mem)
-{
-	struct radeon_ib ib;
-	struct radeon_bo_va *bo_va;
-	unsigned nptes, npdes, ndw;
-	uint64_t addr;
-	int r;
-
-	/* nothing to do if vm isn't bound */
-	if (vm->page_directory == NULL)
-		return 0;
-
-	bo_va = radeon_vm_bo_find(vm, bo);
-	if (bo_va == NULL) {
-		dev_err(rdev->dev, "bo %p not in vm %p\n", bo, vm);
-		return -EINVAL;
-	}
-
-	if (!bo_va->soffset) {
-		dev_err(rdev->dev, "bo %p don't has a mapping in vm %p\n",
-			bo, vm);
-		return -EINVAL;
-	}
-
-	if ((bo_va->valid && mem) || (!bo_va->valid && mem == NULL))
-		return 0;
-
-	bo_va->flags &= ~RADEON_VM_PAGE_VALID;
-	bo_va->flags &= ~RADEON_VM_PAGE_SYSTEM;
-	if (mem) {
-		addr = mem->start << PAGE_SHIFT;
-		if (mem->mem_type != TTM_PL_SYSTEM) {
-			bo_va->flags |= RADEON_VM_PAGE_VALID;
-			bo_va->valid = true;
-		}
-		if (mem->mem_type == TTM_PL_TT) {
-			bo_va->flags |= RADEON_VM_PAGE_SYSTEM;
-		} else {
-			addr += rdev->vm_manager.vram_base_offset;
-		}
-	} else {
-		addr = 0;
-		bo_va->valid = false;
-	}
-
-	trace_radeon_vm_bo_update(bo_va);
-
-	nptes = radeon_bo_ngpu_pages(bo);
-
-	/* assume two extra pdes in case the mapping overlaps the borders */
-	npdes = (nptes >> RADEON_VM_BLOCK_SIZE) + 2;
-
-	/* padding, etc. */
-	ndw = 64;
-
-	if (RADEON_VM_BLOCK_SIZE > 11)
-		/* reserve space for one header for every 2k dwords */
-		ndw += (nptes >> 11) * 4;
-	else
-		/* reserve space for one header for
-		    every (1 << BLOCK_SIZE) entries */
-		ndw += (nptes >> RADEON_VM_BLOCK_SIZE) * 4;
-
-	/* reserve space for pte addresses */
-	ndw += nptes * 2;
-
-	/* reserve space for one header for every 2k dwords */
-	ndw += (npdes >> 11) * 4;
-
-	/* reserve space for pde addresses */
-	ndw += npdes * 2;
-
-	/* reserve space for clearing new page tables */
-	ndw += npdes * 2 * RADEON_VM_PTE_COUNT;
-
-	/* update too big for an IB */
-	if (ndw > 0xfffff)
-		return -ENOMEM;
-
-	r = radeon_ib_get(rdev, R600_RING_TYPE_DMA_INDEX, &ib, NULL, ndw * 4);
-	if (r)
-		return r;
-	ib.length_dw = 0;
-
-	r = radeon_vm_update_pdes(rdev, vm, &ib, bo_va->soffset, bo_va->eoffset);
-	if (r) {
-		radeon_ib_free(rdev, &ib);
-		return r;
-	}
-
-	radeon_vm_update_ptes(rdev, vm, &ib, bo_va->soffset, bo_va->eoffset,
-			      addr, radeon_vm_page_flags(bo_va->flags));
-
-	radeon_semaphore_sync_to(ib.semaphore, vm->fence);
-	r = radeon_ib_schedule(rdev, &ib, NULL);
-	if (r) {
-		radeon_ib_free(rdev, &ib);
-		return r;
-	}
-	radeon_fence_unref(&vm->fence);
-	vm->fence = radeon_fence_ref(ib.fence);
-	radeon_ib_free(rdev, &ib);
-	radeon_fence_unref(&vm->last_flush);
-
-	return 0;
-}
-
-/**
- * radeon_vm_bo_rmv - remove a bo to a specific vm
- *
- * @rdev: radeon_device pointer
- * @bo_va: requested bo_va
- *
- * Remove @bo_va->bo from the requested vm (cayman+).
- * Remove @bo_va->bo from the list of bos associated with the bo_va->vm and
- * remove the ptes for @bo_va in the page table.
- * Returns 0 for success.
- *
- * Object have to be reserved!
- */
-int radeon_vm_bo_rmv(struct radeon_device *rdev,
-		     struct radeon_bo_va *bo_va)
-{
-	int r = 0;
-
-	mutex_lock(&rdev->vm_manager.lock);
-	mutex_lock(&bo_va->vm->mutex);
-	if (bo_va->soffset) {
-		r = radeon_vm_bo_update(rdev, bo_va->vm, bo_va->bo, NULL);
-	}
-	mutex_unlock(&rdev->vm_manager.lock);
-	list_del(&bo_va->vm_list);
-	mutex_unlock(&bo_va->vm->mutex);
-	list_del(&bo_va->bo_list);
-
-	kfree(bo_va);
-	return r;
-}
-
-/**
- * radeon_vm_bo_invalidate - mark the bo as invalid
- *
- * @rdev: radeon_device pointer
- * @vm: requested vm
- * @bo: radeon buffer object
- *
- * Mark @bo as invalid (cayman+).
- */
-void radeon_vm_bo_invalidate(struct radeon_device *rdev,
-			     struct radeon_bo *bo)
-{
-	struct radeon_bo_va *bo_va;
-
-	list_for_each_entry(bo_va, &bo->va, bo_list) {
-		bo_va->valid = false;
-	}
-}
-
-/**
- * radeon_vm_init - initialize a vm instance
- *
- * @rdev: radeon_device pointer
- * @vm: requested vm
- *
- * Init @vm fields (cayman+).
- */
-void radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm)
-{
-	vm->id = 0;
-	vm->fence = NULL;
-	mutex_init(&vm->mutex);
-	INIT_LIST_HEAD(&vm->list);
-	INIT_LIST_HEAD(&vm->va);
-}
-
-/**
- * radeon_vm_fini - tear down a vm instance
- *
- * @rdev: radeon_device pointer
- * @vm: requested vm
- *
- * Tear down @vm (cayman+).
- * Unbind the VM and remove all bos from the vm bo list
- */
-void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm)
-{
-	struct radeon_bo_va *bo_va, *tmp;
-	int r;
-
-	mutex_lock(&rdev->vm_manager.lock);
-	mutex_lock(&vm->mutex);
-	radeon_vm_free_pt(rdev, vm);
-	mutex_unlock(&rdev->vm_manager.lock);
-
-	if (!list_empty(&vm->va)) {
-		dev_err(rdev->dev, "still active bo inside vm\n");
-	}
-	list_for_each_entry_safe(bo_va, tmp, &vm->va, vm_list) {
-		list_del_init(&bo_va->vm_list);
-		r = radeon_bo_reserve(bo_va->bo, false);
-		if (!r) {
-			list_del_init(&bo_va->bo_list);
-			radeon_bo_unreserve(bo_va->bo);
-			kfree(bo_va);
-		}
-	}
-	radeon_fence_unref(&vm->fence);
-	radeon_fence_unref(&vm->last_flush);
-	mutex_unlock(&vm->mutex);
-}
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index 805c5e566b9..d09650c1d72 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -86,7 +86,7 @@ retry:
 	return 0;
 }
 
-int radeon_gem_set_domain(struct drm_gem_object *gobj,
+static int radeon_gem_set_domain(struct drm_gem_object *gobj,
 			  uint32_t rdomain, uint32_t wdomain)
 {
 	struct radeon_bo *robj;
@@ -344,18 +344,7 @@ int radeon_gem_busy_ioctl(struct drm_device *dev, void *data,
 	}
 	robj = gem_to_radeon_bo(gobj);
 	r = radeon_bo_wait(robj, &cur_placement, true);
-	switch (cur_placement) {
-	case TTM_PL_VRAM:
-		args->domain = RADEON_GEM_DOMAIN_VRAM;
-		break;
-	case TTM_PL_TT:
-		args->domain = RADEON_GEM_DOMAIN_GTT;
-		break;
-	case TTM_PL_SYSTEM:
-		args->domain = RADEON_GEM_DOMAIN_CPU;
-	default:
-		break;
-	}
+	args->domain = radeon_mem_type_to_domain(cur_placement);
 	drm_gem_object_unreference_unlocked(gobj);
 	r = radeon_gem_handle_lockup(rdev, r);
 	return r;
@@ -533,6 +522,42 @@ out:
 	return r;
 }
 
+int radeon_gem_op_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *filp)
+{
+	struct drm_radeon_gem_op *args = data;
+	struct drm_gem_object *gobj;
+	struct radeon_bo *robj;
+	int r;
+
+	gobj = drm_gem_object_lookup(dev, filp, args->handle);
+	if (gobj == NULL) {
+		return -ENOENT;
+	}
+	robj = gem_to_radeon_bo(gobj);
+	r = radeon_bo_reserve(robj, false);
+	if (unlikely(r))
+		goto out;
+
+	switch (args->op) {
+	case RADEON_GEM_OP_GET_INITIAL_DOMAIN:
+		args->value = robj->initial_domain;
+		break;
+	case RADEON_GEM_OP_SET_INITIAL_DOMAIN:
+		robj->initial_domain = args->value & (RADEON_GEM_DOMAIN_VRAM |
+						      RADEON_GEM_DOMAIN_GTT |
+						      RADEON_GEM_DOMAIN_CPU);
+		break;
+	default:
+		r = -EINVAL;
+	}
+
+	radeon_bo_unreserve(robj);
+out:
+	drm_gem_object_unreference_unlocked(gobj);
+	return r;
+}
+
 int radeon_mode_dumb_create(struct drm_file *file_priv,
 			    struct drm_device *dev,
 			    struct drm_mode_create_dumb *args)
diff --git a/drivers/gpu/drm/radeon/radeon_i2c.c b/drivers/gpu/drm/radeon/radeon_i2c.c
index fc60b74ee30..add62200840 100644
--- a/drivers/gpu/drm/radeon/radeon_i2c.c
+++ b/drivers/gpu/drm/radeon/radeon_i2c.c
@@ -64,8 +64,7 @@ bool radeon_ddc_probe(struct radeon_connector *radeon_connector, bool use_aux)
 		radeon_router_select_ddc_port(radeon_connector);
 
 	if (use_aux) {
-		struct radeon_connector_atom_dig *dig = radeon_connector->con_priv;
-		ret = i2c_transfer(&dig->dp_i2c_bus->adapter, msgs, 2);
+		ret = i2c_transfer(&radeon_connector->ddc_bus->aux.ddc, msgs, 2);
 	} else {
 		ret = i2c_transfer(&radeon_connector->ddc_bus->adapter, msgs, 2);
 	}
@@ -95,6 +94,8 @@ static int pre_xfer(struct i2c_adapter *i2c_adap)
 	struct radeon_i2c_bus_rec *rec = &i2c->rec;
 	uint32_t temp;
 
+	mutex_lock(&i2c->mutex);
+
 	/* RV410 appears to have a bug where the hw i2c in reset
 	 * holds the i2c port in a bad state - switch hw i2c away before
 	 * doing DDC - do this for all r200s/r300s/r400s for safety sake
@@ -171,6 +172,8 @@ static void post_xfer(struct i2c_adapter *i2c_adap)
 	temp = RREG32(rec->mask_data_reg) & ~rec->mask_data_mask;
 	WREG32(rec->mask_data_reg, temp);
 	temp = RREG32(rec->mask_data_reg);
+
+	mutex_unlock(&i2c->mutex);
 }
 
 static int get_clock(void *i2c_priv)
@@ -814,6 +817,8 @@ static int radeon_hw_i2c_xfer(struct i2c_adapter *i2c_adap,
 	struct radeon_i2c_bus_rec *rec = &i2c->rec;
 	int ret = 0;
 
+	mutex_lock(&i2c->mutex);
+
 	switch (rdev->family) {
 	case CHIP_R100:
 	case CHIP_RV100:
@@ -880,6 +885,8 @@ static int radeon_hw_i2c_xfer(struct i2c_adapter *i2c_adap,
 		break;
 	}
 
+	mutex_unlock(&i2c->mutex);
+
 	return ret;
 }
 
@@ -920,6 +927,7 @@ struct radeon_i2c_chan *radeon_i2c_create(struct drm_device *dev,
 	i2c->adapter.dev.parent = &dev->pdev->dev;
 	i2c->dev = dev;
 	i2c_set_adapdata(&i2c->adapter, i2c);
+	mutex_init(&i2c->mutex);
 	if (rec->mm_i2c ||
 	    (rec->hw_capable &&
 	     radeon_hw_i2c &&
@@ -950,16 +958,16 @@ struct radeon_i2c_chan *radeon_i2c_create(struct drm_device *dev,
 		/* set the radeon bit adapter */
 		snprintf(i2c->adapter.name, sizeof(i2c->adapter.name),
 			 "Radeon i2c bit bus %s", name);
-		i2c->adapter.algo_data = &i2c->algo.bit;
-		i2c->algo.bit.pre_xfer = pre_xfer;
-		i2c->algo.bit.post_xfer = post_xfer;
-		i2c->algo.bit.setsda = set_data;
-		i2c->algo.bit.setscl = set_clock;
-		i2c->algo.bit.getsda = get_data;
-		i2c->algo.bit.getscl = get_clock;
-		i2c->algo.bit.udelay = 10;
-		i2c->algo.bit.timeout = usecs_to_jiffies(2200);	/* from VESA */
-		i2c->algo.bit.data = i2c;
+		i2c->adapter.algo_data = &i2c->bit;
+		i2c->bit.pre_xfer = pre_xfer;
+		i2c->bit.post_xfer = post_xfer;
+		i2c->bit.setsda = set_data;
+		i2c->bit.setscl = set_clock;
+		i2c->bit.getsda = get_data;
+		i2c->bit.getscl = get_clock;
+		i2c->bit.udelay = 10;
+		i2c->bit.timeout = usecs_to_jiffies(2200);	/* from VESA */
+		i2c->bit.data = i2c;
 		ret = i2c_bit_add_bus(&i2c->adapter);
 		if (ret) {
 			DRM_ERROR("Failed to register bit i2c %s\n", name);
@@ -974,52 +982,22 @@ out_free:
 
 }
 
-struct radeon_i2c_chan *radeon_i2c_create_dp(struct drm_device *dev,
-					     struct radeon_i2c_bus_rec *rec,
-					     const char *name)
-{
-	struct radeon_i2c_chan *i2c;
-	int ret;
-
-	i2c = kzalloc(sizeof(struct radeon_i2c_chan), GFP_KERNEL);
-	if (i2c == NULL)
-		return NULL;
-
-	i2c->rec = *rec;
-	i2c->adapter.owner = THIS_MODULE;
-	i2c->adapter.class = I2C_CLASS_DDC;
-	i2c->adapter.dev.parent = &dev->pdev->dev;
-	i2c->dev = dev;
-	snprintf(i2c->adapter.name, sizeof(i2c->adapter.name),
-		 "Radeon aux bus %s", name);
-	i2c_set_adapdata(&i2c->adapter, i2c);
-	i2c->adapter.algo_data = &i2c->algo.dp;
-	i2c->algo.dp.aux_ch = radeon_dp_i2c_aux_ch;
-	i2c->algo.dp.address = 0;
-	ret = i2c_dp_aux_add_bus(&i2c->adapter);
-	if (ret) {
-		DRM_INFO("Failed to register i2c %s\n", name);
-		goto out_free;
-	}
-
-	return i2c;
-out_free:
-	kfree(i2c);
-	return NULL;
-
-}
-
 void radeon_i2c_destroy(struct radeon_i2c_chan *i2c)
 {
 	if (!i2c)
 		return;
 	i2c_del_adapter(&i2c->adapter);
+	if (i2c->has_aux)
+		drm_dp_aux_unregister(&i2c->aux);
 	kfree(i2c);
 }
 
 /* Add the default buses */
 void radeon_i2c_init(struct radeon_device *rdev)
 {
+	if (radeon_hw_i2c)
+		DRM_INFO("hw_i2c forced on, you may experience display detection problems!\n");
+
 	if (rdev->is_atom_bios)
 		radeon_atombios_i2c_init(rdev);
 	else
diff --git a/drivers/gpu/drm/radeon/radeon_ioc32.c b/drivers/gpu/drm/radeon/radeon_ioc32.c
index bdb0f93e73b..0b98ea13457 100644
--- a/drivers/gpu/drm/radeon/radeon_ioc32.c
+++ b/drivers/gpu/drm/radeon/radeon_ioc32.c
@@ -399,7 +399,7 @@ long radeon_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 	if (nr < DRM_COMMAND_BASE)
 		return drm_compat_ioctl(filp, cmd, arg);
 
-	if (nr < DRM_COMMAND_BASE + DRM_ARRAY_SIZE(radeon_compat_ioctls))
+	if (nr < DRM_COMMAND_BASE + ARRAY_SIZE(radeon_compat_ioctls))
 		fn = radeon_compat_ioctls[nr - DRM_COMMAND_BASE];
 
 	if (fn != NULL)
diff --git a/drivers/gpu/drm/radeon/radeon_irq.c b/drivers/gpu/drm/radeon/radeon_irq.c
index 8d68e972789..244b19bab2e 100644
--- a/drivers/gpu/drm/radeon/radeon_irq.c
+++ b/drivers/gpu/drm/radeon/radeon_irq.c
@@ -181,7 +181,7 @@ static u32 radeon_acknowledge_irqs(drm_radeon_private_t *dev_priv, u32 *r500_dis
  * tied to dma at all, this is just a hangover from dri prehistory.
  */
 
-irqreturn_t radeon_driver_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t radeon_driver_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	drm_radeon_private_t *dev_priv =
@@ -203,7 +203,7 @@ irqreturn_t radeon_driver_irq_handler(DRM_IRQ_ARGS)
 
 	/* SW interrupt */
 	if (stat & RADEON_SW_INT_TEST)
-		DRM_WAKEUP(&dev_priv->swi_queue);
+		wake_up(&dev_priv->swi_queue);
 
 	/* VBLANK interrupt */
 	if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_RS600) {
@@ -249,7 +249,7 @@ static int radeon_wait_irq(struct drm_device * dev, int swi_nr)
 
 	dev_priv->stats.boxes |= RADEON_BOX_WAIT_IDLE;
 
-	DRM_WAIT_ON(ret, dev_priv->swi_queue, 3 * DRM_HZ,
+	DRM_WAIT_ON(ret, dev_priv->swi_queue, 3 * HZ,
 		    RADEON_READ(RADEON_LAST_SWI_REG) >= swi_nr);
 
 	return ret;
@@ -302,7 +302,7 @@ int radeon_irq_emit(struct drm_device *dev, void *data, struct drm_file *file_pr
 
 	result = radeon_emit_irq(dev);
 
-	if (DRM_COPY_TO_USER(emit->irq_seq, &result, sizeof(int))) {
+	if (copy_to_user(emit->irq_seq, &result, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
@@ -354,7 +354,7 @@ int radeon_driver_irq_postinstall(struct drm_device *dev)
 	    (drm_radeon_private_t *) dev->dev_private;
 
 	atomic_set(&dev_priv->swi_emitted, 0);
-	DRM_INIT_WAITQUEUE(&dev_priv->swi_queue);
+	init_waitqueue_head(&dev_priv->swi_queue);
 
 	dev->max_vblank_count = 0x001fffff;
 
diff --git a/drivers/gpu/drm/radeon/radeon_irq_kms.c b/drivers/gpu/drm/radeon/radeon_irq_kms.c
index ec6240b0046..16807afab36 100644
--- a/drivers/gpu/drm/radeon/radeon_irq_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_irq_kms.c
@@ -39,13 +39,13 @@
 /**
  * radeon_driver_irq_handler_kms - irq handler for KMS
  *
- * @DRM_IRQ_ARGS: args
+ * @int irq, void *arg: args
  *
  * This is the irq handler for the radeon KMS driver (all asics).
  * radeon_irq_process is a macro that points to the per-asic
  * irq handler callback.
  */
-irqreturn_t radeon_driver_irq_handler_kms(DRM_IRQ_ARGS)
+irqreturn_t radeon_driver_irq_handler_kms(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	struct radeon_device *rdev = dev->dev_private;
@@ -287,7 +287,7 @@ int radeon_irq_kms_init(struct radeon_device *rdev)
 	INIT_WORK(&rdev->reset_work, radeon_irq_reset_work_func);
 
 	rdev->irq.installed = true;
-	r = drm_irq_install(rdev->ddev);
+	r = drm_irq_install(rdev->ddev, rdev->ddev->pdev->irq);
 	if (r) {
 		rdev->irq.installed = false;
 		flush_work(&rdev->hotplug_work);
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index 21d593c0eca..d25ae6acfd5 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -33,6 +33,13 @@
 #include <linux/vga_switcheroo.h>
 #include <linux/slab.h>
 #include <linux/pm_runtime.h>
+
+#if defined(CONFIG_VGA_SWITCHEROO)
+bool radeon_has_atpx(void);
+#else
+static inline bool radeon_has_atpx(void) { return false; }
+#endif
+
 /**
  * radeon_driver_unload_kms - Main unload function for KMS.
  *
@@ -100,6 +107,11 @@ int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags)
 		flags |= RADEON_IS_PCI;
 	}
 
+	if ((radeon_runtime_pm != 0) &&
+	    radeon_has_atpx() &&
+	    ((flags & RADEON_IS_IGP) == 0))
+		flags |= RADEON_IS_PX;
+
 	/* radeon_device_init should report only fatal error
 	 * like memory allocation failure or iomapping failure,
 	 * or memory manager initialization failure, it must
@@ -130,7 +142,7 @@ int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags)
 				"Error during ACPI methods call\n");
 	}
 
-	if (radeon_runtime_pm != 0) {
+	if (radeon_is_px(dev)) {
 		pm_runtime_use_autosuspend(dev->dev);
 		pm_runtime_set_autosuspend_delay(dev->dev, 5000);
 		pm_runtime_set_active(dev->dev);
@@ -191,7 +203,7 @@ static void radeon_set_filp_rights(struct drm_device *dev,
  * etc. (all asics).
  * Returns 0 on success, -EINVAL on failure.
  */
-int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
+static int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 {
 	struct radeon_device *rdev = dev->dev_private;
 	struct drm_radeon_info *info = data;
@@ -223,7 +235,7 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 			*value = rdev->accel_working;
 		break;
 	case RADEON_INFO_CRTC_FROM_ID:
-		if (DRM_COPY_FROM_USER(value, value_ptr, sizeof(uint32_t))) {
+		if (copy_from_user(value, value_ptr, sizeof(uint32_t))) {
 			DRM_ERROR("copy_from_user %s:%u\n", __func__, __LINE__);
 			return -EFAULT;
 		}
@@ -269,7 +281,7 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		 *
 		 * When returning, the value is 1 if filp owns hyper-z access,
 		 * 0 otherwise. */
-		if (DRM_COPY_FROM_USER(value, value_ptr, sizeof(uint32_t))) {
+		if (copy_from_user(value, value_ptr, sizeof(uint32_t))) {
 			DRM_ERROR("copy_from_user %s:%u\n", __func__, __LINE__);
 			return -EFAULT;
 		}
@@ -281,7 +293,7 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		break;
 	case RADEON_INFO_WANT_CMASK:
 		/* The same logic as Hyper-Z. */
-		if (DRM_COPY_FROM_USER(value, value_ptr, sizeof(uint32_t))) {
+		if (copy_from_user(value, value_ptr, sizeof(uint32_t))) {
 			DRM_ERROR("copy_from_user %s:%u\n", __func__, __LINE__);
 			return -EFAULT;
 		}
@@ -417,7 +429,7 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		*value = rdev->fastfb_working;
 		break;
 	case RADEON_INFO_RING_WORKING:
-		if (DRM_COPY_FROM_USER(value, value_ptr, sizeof(uint32_t))) {
+		if (copy_from_user(value, value_ptr, sizeof(uint32_t))) {
 			DRM_ERROR("copy_from_user %s:%u\n", __func__, __LINE__);
 			return -EFAULT;
 		}
@@ -433,6 +445,9 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		case RADEON_CS_RING_UVD:
 			*value = rdev->ring[R600_RING_TYPE_UVD_INDEX].ready;
 			break;
+		case RADEON_CS_RING_VCE:
+			*value = rdev->ring[TN_RING_TYPE_VCE1_INDEX].ready;
+			break;
 		default:
 			return -EINVAL;
 		}
@@ -470,11 +485,55 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 			DRM_DEBUG_KMS("BACKEND_ENABLED_MASK is si+ only!\n");
 		}
 		break;
+	case RADEON_INFO_MAX_SCLK:
+		if ((rdev->pm.pm_method == PM_METHOD_DPM) &&
+		    rdev->pm.dpm_enabled)
+			*value = rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.sclk * 10;
+		else
+			*value = rdev->pm.default_sclk * 10;
+		break;
+	case RADEON_INFO_VCE_FW_VERSION:
+		*value = rdev->vce.fw_version;
+		break;
+	case RADEON_INFO_VCE_FB_VERSION:
+		*value = rdev->vce.fb_version;
+		break;
+	case RADEON_INFO_NUM_BYTES_MOVED:
+		value = (uint32_t*)&value64;
+		value_size = sizeof(uint64_t);
+		value64 = atomic64_read(&rdev->num_bytes_moved);
+		break;
+	case RADEON_INFO_VRAM_USAGE:
+		value = (uint32_t*)&value64;
+		value_size = sizeof(uint64_t);
+		value64 = atomic64_read(&rdev->vram_usage);
+		break;
+	case RADEON_INFO_GTT_USAGE:
+		value = (uint32_t*)&value64;
+		value_size = sizeof(uint64_t);
+		value64 = atomic64_read(&rdev->gtt_usage);
+		break;
+	case RADEON_INFO_ACTIVE_CU_COUNT:
+		if (rdev->family >= CHIP_BONAIRE)
+			*value = rdev->config.cik.active_cus;
+		else if (rdev->family >= CHIP_TAHITI)
+			*value = rdev->config.si.active_cus;
+		else if (rdev->family >= CHIP_CAYMAN)
+			*value = rdev->config.cayman.active_simds;
+		else if (rdev->family >= CHIP_CEDAR)
+			*value = rdev->config.evergreen.active_simds;
+		else if (rdev->family >= CHIP_RV770)
+			*value = rdev->config.rv770.active_simds;
+		else if (rdev->family >= CHIP_R600)
+			*value = rdev->config.r600.active_simds;
+		else
+			*value = 1;
+		break;
 	default:
 		DRM_DEBUG_KMS("Invalid request %d\n", info->request);
 		return -EINVAL;
 	}
-	if (DRM_COPY_TO_USER(value_ptr, (char*)value, value_size)) {
+	if (copy_to_user(value_ptr, (char*)value, value_size)) {
 		DRM_ERROR("copy_to_user %s:%u\n", __func__, __LINE__);
 		return -EFAULT;
 	}
@@ -520,7 +579,7 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 	/* new gpu have virtual address space support */
 	if (rdev->family >= CHIP_CAYMAN) {
 		struct radeon_fpriv *fpriv;
-		struct radeon_bo_va *bo_va;
+		struct radeon_vm *vm;
 		int r;
 
 		fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL);
@@ -528,21 +587,37 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 			return -ENOMEM;
 		}
 
-		radeon_vm_init(rdev, &fpriv->vm);
-
-		/* map the ib pool buffer read only into
-		 * virtual address space */
-		bo_va = radeon_vm_bo_add(rdev, &fpriv->vm,
-					 rdev->ring_tmp_bo.bo);
-		r = radeon_vm_bo_set_addr(rdev, bo_va, RADEON_VA_IB_OFFSET,
-					  RADEON_VM_PAGE_READABLE |
-					  RADEON_VM_PAGE_SNOOPED);
+		vm = &fpriv->vm;
+		r = radeon_vm_init(rdev, vm);
 		if (r) {
-			radeon_vm_fini(rdev, &fpriv->vm);
 			kfree(fpriv);
 			return r;
 		}
 
+		if (rdev->accel_working) {
+			r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
+			if (r) {
+				radeon_vm_fini(rdev, vm);
+				kfree(fpriv);
+				return r;
+			}
+
+			/* map the ib pool buffer read only into
+			 * virtual address space */
+			vm->ib_bo_va = radeon_vm_bo_add(rdev, vm,
+							rdev->ring_tmp_bo.bo);
+			r = radeon_vm_bo_set_addr(rdev, vm->ib_bo_va,
+						  RADEON_VA_IB_OFFSET,
+						  RADEON_VM_PAGE_READABLE |
+						  RADEON_VM_PAGE_SNOOPED);
+
+			radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
+			if (r) {
+				radeon_vm_fini(rdev, vm);
+				kfree(fpriv);
+				return r;
+			}
+		}
 		file_priv->driver_priv = fpriv;
 	}
 
@@ -567,19 +642,19 @@ void radeon_driver_postclose_kms(struct drm_device *dev,
 	/* new gpu have virtual address space support */
 	if (rdev->family >= CHIP_CAYMAN && file_priv->driver_priv) {
 		struct radeon_fpriv *fpriv = file_priv->driver_priv;
-		struct radeon_bo_va *bo_va;
+		struct radeon_vm *vm = &fpriv->vm;
 		int r;
 
-		r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
-		if (!r) {
-			bo_va = radeon_vm_bo_find(&fpriv->vm,
-						  rdev->ring_tmp_bo.bo);
-			if (bo_va)
-				radeon_vm_bo_rmv(rdev, bo_va);
-			radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
+		if (rdev->accel_working) {
+			r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
+			if (!r) {
+				if (vm->ib_bo_va)
+					radeon_vm_bo_rmv(rdev, vm->ib_bo_va);
+				radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
+			}
 		}
 
-		radeon_vm_fini(rdev, &fpriv->vm);
+		radeon_vm_fini(rdev, vm);
 		kfree(fpriv);
 		file_priv->driver_priv = NULL;
 	}
@@ -603,6 +678,7 @@ void radeon_driver_preclose_kms(struct drm_device *dev,
 	if (rdev->cmask_filp == file_priv)
 		rdev->cmask_filp = NULL;
 	radeon_uvd_free_handles(rdev, file_priv);
+	radeon_vce_free_handles(rdev, file_priv);
 }
 
 /*
@@ -712,11 +788,12 @@ int radeon_get_vblank_timestamp_kms(struct drm_device *dev, int crtc,
 	/* Helper routine in DRM core does all the work: */
 	return drm_calc_vbltimestamp_from_scanoutpos(dev, crtc, max_error,
 						     vblank_time, flags,
-						     drmcrtc);
+						     drmcrtc, &drmcrtc->hwmode);
 }
 
 #define KMS_INVALID_IOCTL(name)						\
-int name(struct drm_device *dev, void *data, struct drm_file *file_priv)\
+static int name(struct drm_device *dev, void *data, struct drm_file	\
+		*file_priv)						\
 {									\
 	DRM_ERROR("invalid ioctl with kms %s\n", __func__);		\
 	return -EINVAL;							\
@@ -796,5 +873,6 @@ const struct drm_ioctl_desc radeon_ioctls_kms[] = {
 	DRM_IOCTL_DEF_DRV(RADEON_GEM_GET_TILING, radeon_gem_get_tiling_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(RADEON_GEM_BUSY, radeon_gem_busy_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(RADEON_GEM_VA, radeon_gem_va_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(RADEON_GEM_OP, radeon_gem_op_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 };
-int radeon_max_kms_ioctl = DRM_ARRAY_SIZE(radeon_ioctls_kms);
+int radeon_max_kms_ioctl = ARRAY_SIZE(radeon_ioctls_kms);
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
index 0b158f98d28..cafb1ccf2ec 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
@@ -385,7 +385,7 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 
 	DRM_DEBUG_KMS("\n");
 	/* no fb bound */
-	if (!atomic && !crtc->fb) {
+	if (!atomic && !crtc->primary->fb) {
 		DRM_DEBUG_KMS("No FB bound\n");
 		return 0;
 	}
@@ -395,8 +395,8 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 		target_fb = fb;
 	}
 	else {
-		radeon_fb = to_radeon_framebuffer(crtc->fb);
-		target_fb = crtc->fb;
+		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+		target_fb = crtc->primary->fb;
 	}
 
 	switch (target_fb->bits_per_pixel) {
@@ -444,7 +444,7 @@ retry:
 		 * We don't shutdown the display controller because new buffer
 		 * will end up in same spot.
 		 */
-		if (!atomic && fb && fb != crtc->fb) {
+		if (!atomic && fb && fb != crtc->primary->fb) {
 			struct radeon_bo *old_rbo;
 			unsigned long nsize, osize;
 
@@ -555,7 +555,7 @@ retry:
 	WREG32(RADEON_CRTC_OFFSET + radeon_crtc->crtc_offset, crtc_offset);
 	WREG32(RADEON_CRTC_PITCH + radeon_crtc->crtc_offset, crtc_pitch);
 
-	if (!atomic && fb && fb != crtc->fb) {
+	if (!atomic && fb && fb != crtc->primary->fb) {
 		radeon_fb = to_radeon_framebuffer(fb);
 		rbo = gem_to_radeon_bo(radeon_fb->obj);
 		r = radeon_bo_reserve(rbo, false);
@@ -599,7 +599,7 @@ static bool radeon_set_crtc_timing(struct drm_crtc *crtc, struct drm_display_mod
 		}
 	}
 
-	switch (crtc->fb->bits_per_pixel) {
+	switch (crtc->primary->fb->bits_per_pixel) {
 	case 8:
 		format = 2;
 		break;
@@ -1087,12 +1087,12 @@ static void radeon_crtc_commit(struct drm_crtc *crtc)
 static void radeon_crtc_disable(struct drm_crtc *crtc)
 {
 	radeon_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
-	if (crtc->fb) {
+	if (crtc->primary->fb) {
 		int r;
 		struct radeon_framebuffer *radeon_fb;
 		struct radeon_bo *rbo;
 
-		radeon_fb = to_radeon_framebuffer(crtc->fb);
+		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
 		rbo = gem_to_radeon_bo(radeon_fb->obj);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r))
diff --git a/drivers/gpu/drm/radeon/radeon_mem.c b/drivers/gpu/drm/radeon/radeon_mem.c
index d54d2d7c903..146d253f113 100644
--- a/drivers/gpu/drm/radeon/radeon_mem.c
+++ b/drivers/gpu/drm/radeon/radeon_mem.c
@@ -243,7 +243,7 @@ int radeon_mem_alloc(struct drm_device *dev, void *data, struct drm_file *file_p
 	if (!block)
 		return -ENOMEM;
 
-	if (DRM_COPY_TO_USER(alloc->region_offset, &block->start,
+	if (copy_to_user(alloc->region_offset, &block->start,
 			     sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index 3f0dd664af9..0592ddb0904 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -46,6 +46,10 @@ struct radeon_device;
 #define to_radeon_encoder(x) container_of(x, struct radeon_encoder, base)
 #define to_radeon_framebuffer(x) container_of(x, struct radeon_framebuffer, base)
 
+#define RADEON_MAX_HPD_PINS 7
+#define RADEON_MAX_CRTCS 6
+#define RADEON_MAX_AFMT_BLOCKS 7
+
 enum radeon_rmx_type {
 	RMX_OFF,
 	RMX_FULL,
@@ -187,11 +191,11 @@ struct radeon_pll {
 struct radeon_i2c_chan {
 	struct i2c_adapter adapter;
 	struct drm_device *dev;
-	union {
-		struct i2c_algo_bit_data bit;
-		struct i2c_algo_dp_aux_data dp;
-	} algo;
+	struct i2c_algo_bit_data bit;
 	struct radeon_i2c_bus_rec rec;
+	struct drm_dp_aux aux;
+	bool has_aux;
+	struct mutex mutex;
 };
 
 /* mostly for macs, but really any system without connector tables */
@@ -233,8 +237,8 @@ struct radeon_mode_info {
 	struct card_info *atom_card_info;
 	enum radeon_connector_table connector_table;
 	bool mode_config_initialized;
-	struct radeon_crtc *crtcs[6];
-	struct radeon_afmt *afmt[7];
+	struct radeon_crtc *crtcs[RADEON_MAX_CRTCS];
+	struct radeon_afmt *afmt[RADEON_MAX_AFMT_BLOCKS];
 	/* DVI-I properties */
 	struct drm_property *coherent_mode_property;
 	/* DAC enable load detect */
@@ -291,6 +295,7 @@ struct radeon_tv_regs {
 
 struct radeon_atom_ss {
 	uint16_t percentage;
+	uint16_t percentage_divider;
 	uint8_t type;
 	uint16_t step;
 	uint8_t delay;
@@ -301,6 +306,12 @@ struct radeon_atom_ss {
 	uint16_t amount;
 };
 
+enum radeon_flip_status {
+	RADEON_FLIP_NONE,
+	RADEON_FLIP_PENDING,
+	RADEON_FLIP_SUBMITTED
+};
+
 struct radeon_crtc {
 	struct drm_crtc base;
 	int crtc_id;
@@ -324,8 +335,9 @@ struct radeon_crtc {
 	struct drm_display_mode native_mode;
 	int pll_id;
 	/* page flipping */
-	struct radeon_unpin_work *unpin_work;
-	int deferred_flip_completion;
+	struct workqueue_struct *flip_queue;
+	struct radeon_flip_work *flip_work;
+	enum radeon_flip_status flip_status;
 	/* pll sharing */
 	struct radeon_atom_ss ss;
 	bool ss_enabled;
@@ -438,7 +450,6 @@ struct radeon_encoder {
 struct radeon_connector_atom_dig {
 	uint32_t igp_lane_info;
 	/* displayport */
-	struct radeon_i2c_chan *dp_i2c_bus;
 	u8 dpcd[DP_RECEIVER_CAP_SIZE];
 	u8 dp_sink_type;
 	int dp_clock;
@@ -506,6 +517,7 @@ struct radeon_connector {
 	struct radeon_i2c_chan *router_bus;
 	enum radeon_connector_audio audio;
 	enum radeon_connector_dither dither;
+	int pixelclock_for_modeset;
 };
 
 struct radeon_framebuffer {
@@ -624,6 +636,30 @@ struct atom_voltage_table
 	struct atom_voltage_table_entry entries[MAX_VOLTAGE_ENTRIES];
 };
 
+
+extern void
+radeon_add_atom_connector(struct drm_device *dev,
+			  uint32_t connector_id,
+			  uint32_t supported_device,
+			  int connector_type,
+			  struct radeon_i2c_bus_rec *i2c_bus,
+			  uint32_t igp_lane_info,
+			  uint16_t connector_object_id,
+			  struct radeon_hpd *hpd,
+			  struct radeon_router *router);
+extern void
+radeon_add_legacy_connector(struct drm_device *dev,
+			    uint32_t connector_id,
+			    uint32_t supported_device,
+			    int connector_type,
+			    struct radeon_i2c_bus_rec *i2c_bus,
+			    uint16_t connector_object_id,
+			    struct radeon_hpd *hpd);
+extern uint32_t
+radeon_get_encoder_enum(struct drm_device *dev, uint32_t supported_device,
+			uint8_t dac);
+extern void radeon_link_encoder_connector(struct drm_device *dev);
+
 extern enum radeon_tv_std
 radeon_combios_get_tv_info(struct radeon_device *rdev);
 extern enum radeon_tv_std
@@ -631,6 +667,15 @@ radeon_atombios_get_tv_info(struct radeon_device *rdev);
 extern void radeon_atombios_get_default_voltages(struct radeon_device *rdev,
 						 u16 *vddc, u16 *vddci, u16 *mvdd);
 
+extern void
+radeon_combios_connected_scratch_regs(struct drm_connector *connector,
+				      struct drm_encoder *encoder,
+				      bool connected);
+extern void
+radeon_atombios_connected_scratch_regs(struct drm_connector *connector,
+				       struct drm_encoder *encoder,
+				       bool connected);
+
 extern struct drm_connector *
 radeon_get_connector_for_encoder(struct drm_encoder *encoder);
 extern struct drm_connector *
@@ -656,6 +701,9 @@ extern u8 radeon_dp_getsinktype(struct radeon_connector *radeon_connector);
 extern bool radeon_dp_getdpcd(struct radeon_connector *radeon_connector);
 extern int radeon_dp_get_panel_mode(struct drm_encoder *encoder,
 				    struct drm_connector *connector);
+extern void radeon_dp_set_rx_power_state(struct drm_connector *connector,
+					 u8 power_state);
+extern void radeon_dp_aux_init(struct radeon_connector *radeon_connector);
 extern void atombios_dig_encoder_setup(struct drm_encoder *encoder, int action, int panel_mode);
 extern void radeon_atom_encoder_init(struct radeon_device *rdev);
 extern void radeon_atom_disp_eng_pll_init(struct radeon_device *rdev);
@@ -664,8 +712,7 @@ extern void atombios_dig_transmitter_setup(struct drm_encoder *encoder,
 					   uint8_t lane_set);
 extern void radeon_atom_ext_encoder_setup_ddc(struct drm_encoder *encoder);
 extern struct drm_encoder *radeon_get_external_encoder(struct drm_encoder *encoder);
-extern int radeon_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
-				u8 write_byte, u8 *read_byte);
+void radeon_atom_copy_swap(u8 *dst, u8 *src, u8 num_bytes, bool to_le);
 
 extern void radeon_i2c_init(struct radeon_device *rdev);
 extern void radeon_i2c_fini(struct radeon_device *rdev);
@@ -676,9 +723,6 @@ extern void radeon_i2c_add(struct radeon_device *rdev,
 			   const char *name);
 extern struct radeon_i2c_chan *radeon_i2c_lookup(struct radeon_device *rdev,
 						 struct radeon_i2c_bus_rec *i2c_bus);
-extern struct radeon_i2c_chan *radeon_i2c_create_dp(struct drm_device *dev,
-						    struct radeon_i2c_bus_rec *rec,
-						    const char *name);
 extern struct radeon_i2c_chan *radeon_i2c_create(struct drm_device *dev,
 						 struct radeon_i2c_bus_rec *rec,
 						 const char *name);
@@ -766,6 +810,7 @@ extern int radeon_crtc_cursor_move(struct drm_crtc *crtc,
 				   int x, int y);
 
 extern int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc,
+				      unsigned int flags,
 				      int *vpos, int *hpos, ktime_t *stime,
 				      ktime_t *etime);
 
@@ -874,6 +919,7 @@ bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj)
 
 void radeon_fb_output_poll_changed(struct radeon_device *rdev);
 
+void radeon_crtc_handle_vblank(struct radeon_device *rdev, int crtc_id);
 void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id);
 
 int radeon_align_pitch(struct radeon_device *rdev, int width, int bpp, bool tiled);
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index c0fa4aa9cee..6c717b257d6 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -46,7 +46,7 @@ static void radeon_bo_clear_surface_reg(struct radeon_bo *bo);
  * function are calling it.
  */
 
-void radeon_bo_clear_va(struct radeon_bo *bo)
+static void radeon_bo_clear_va(struct radeon_bo *bo)
 {
 	struct radeon_bo_va *bo_va, *tmp;
 
@@ -56,11 +56,36 @@ void radeon_bo_clear_va(struct radeon_bo *bo)
 	}
 }
 
+static void radeon_update_memory_usage(struct radeon_bo *bo,
+				       unsigned mem_type, int sign)
+{
+	struct radeon_device *rdev = bo->rdev;
+	u64 size = (u64)bo->tbo.num_pages << PAGE_SHIFT;
+
+	switch (mem_type) {
+	case TTM_PL_TT:
+		if (sign > 0)
+			atomic64_add(size, &rdev->gtt_usage);
+		else
+			atomic64_sub(size, &rdev->gtt_usage);
+		break;
+	case TTM_PL_VRAM:
+		if (sign > 0)
+			atomic64_add(size, &rdev->vram_usage);
+		else
+			atomic64_sub(size, &rdev->vram_usage);
+		break;
+	}
+}
+
 static void radeon_ttm_bo_destroy(struct ttm_buffer_object *tbo)
 {
 	struct radeon_bo *bo;
 
 	bo = container_of(tbo, struct radeon_bo, tbo);
+
+	radeon_update_memory_usage(bo, bo->tbo.mem.mem_type, -1);
+
 	mutex_lock(&bo->rdev->gem.mutex);
 	list_del_init(&bo->list);
 	mutex_unlock(&bo->rdev->gem.mutex);
@@ -79,7 +104,7 @@ bool radeon_ttm_bo_is_radeon_bo(struct ttm_buffer_object *bo)
 
 void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 {
-	u32 c = 0;
+	u32 c = 0, i;
 
 	rbo->placement.fpfn = 0;
 	rbo->placement.lpfn = 0;
@@ -106,6 +131,17 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 		rbo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
 	rbo->placement.num_placement = c;
 	rbo->placement.num_busy_placement = c;
+
+	/*
+	 * Use two-ended allocation depending on the buffer size to
+	 * improve fragmentation quality.
+	 * 512kb was measured as the most optimal number.
+	 */
+	if (rbo->tbo.mem.size > 512 * 1024) {
+		for (i = 0; i < c; i++) {
+			rbo->placements[i] |= TTM_PL_FLAG_TOPDOWN;
+		}
+	}
 }
 
 int radeon_bo_create(struct radeon_device *rdev,
@@ -120,7 +156,6 @@ int radeon_bo_create(struct radeon_device *rdev,
 
 	size = ALIGN(size, PAGE_SIZE);
 
-	rdev->mman.bdev.dev_mapping = rdev->ddev->dev_mapping;
 	if (kernel) {
 		type = ttm_bo_type_kernel;
 	} else if (sg) {
@@ -145,6 +180,9 @@ int radeon_bo_create(struct radeon_device *rdev,
 	bo->surface_reg = -1;
 	INIT_LIST_HEAD(&bo->list);
 	INIT_LIST_HEAD(&bo->va);
+	bo->initial_domain = domain & (RADEON_GEM_DOMAIN_VRAM |
+	                               RADEON_GEM_DOMAIN_GTT |
+	                               RADEON_GEM_DOMAIN_CPU);
 	radeon_ttm_placement_from_domain(bo, domain);
 	/* Kernel allocation are uninterruptible */
 	down_read(&rdev->pm.mclk_lock);
@@ -338,42 +376,109 @@ void radeon_bo_fini(struct radeon_device *rdev)
 	arch_phys_wc_del(rdev->mc.vram_mtrr);
 }
 
-void radeon_bo_list_add_object(struct radeon_bo_list *lobj,
-				struct list_head *head)
+/* Returns how many bytes TTM can move per IB.
+ */
+static u64 radeon_bo_get_threshold_for_moves(struct radeon_device *rdev)
 {
-	if (lobj->written) {
-		list_add(&lobj->tv.head, head);
-	} else {
-		list_add_tail(&lobj->tv.head, head);
-	}
+	u64 real_vram_size = rdev->mc.real_vram_size;
+	u64 vram_usage = atomic64_read(&rdev->vram_usage);
+
+	/* This function is based on the current VRAM usage.
+	 *
+	 * - If all of VRAM is free, allow relocating the number of bytes that
+	 *   is equal to 1/4 of the size of VRAM for this IB.
+
+	 * - If more than one half of VRAM is occupied, only allow relocating
+	 *   1 MB of data for this IB.
+	 *
+	 * - From 0 to one half of used VRAM, the threshold decreases
+	 *   linearly.
+	 *         __________________
+	 * 1/4 of -|\               |
+	 * VRAM    | \              |
+	 *         |  \             |
+	 *         |   \            |
+	 *         |    \           |
+	 *         |     \          |
+	 *         |      \         |
+	 *         |       \________|1 MB
+	 *         |----------------|
+	 *    VRAM 0 %             100 %
+	 *         used            used
+	 *
+	 * Note: It's a threshold, not a limit. The threshold must be crossed
+	 * for buffer relocations to stop, so any buffer of an arbitrary size
+	 * can be moved as long as the threshold isn't crossed before
+	 * the relocation takes place. We don't want to disable buffer
+	 * relocations completely.
+	 *
+	 * The idea is that buffers should be placed in VRAM at creation time
+	 * and TTM should only do a minimum number of relocations during
+	 * command submission. In practice, you need to submit at least
+	 * a dozen IBs to move all buffers to VRAM if they are in GTT.
+	 *
+	 * Also, things can get pretty crazy under memory pressure and actual
+	 * VRAM usage can change a lot, so playing safe even at 50% does
+	 * consistently increase performance.
+	 */
+
+	u64 half_vram = real_vram_size >> 1;
+	u64 half_free_vram = vram_usage >= half_vram ? 0 : half_vram - vram_usage;
+	u64 bytes_moved_threshold = half_free_vram >> 1;
+	return max(bytes_moved_threshold, 1024*1024ull);
 }
 
-int radeon_bo_list_validate(struct ww_acquire_ctx *ticket,
+int radeon_bo_list_validate(struct radeon_device *rdev,
+			    struct ww_acquire_ctx *ticket,
 			    struct list_head *head, int ring)
 {
-	struct radeon_bo_list *lobj;
+	struct radeon_cs_reloc *lobj;
 	struct radeon_bo *bo;
-	u32 domain;
 	int r;
+	u64 bytes_moved = 0, initial_bytes_moved;
+	u64 bytes_moved_threshold = radeon_bo_get_threshold_for_moves(rdev);
 
 	r = ttm_eu_reserve_buffers(ticket, head);
 	if (unlikely(r != 0)) {
 		return r;
 	}
+
 	list_for_each_entry(lobj, head, tv.head) {
-		bo = lobj->bo;
+		bo = lobj->robj;
 		if (!bo->pin_count) {
-			domain = lobj->domain;
-			
+			u32 domain = lobj->prefered_domains;
+			u32 current_domain =
+				radeon_mem_type_to_domain(bo->tbo.mem.mem_type);
+
+			/* Check if this buffer will be moved and don't move it
+			 * if we have moved too many buffers for this IB already.
+			 *
+			 * Note that this allows moving at least one buffer of
+			 * any size, because it doesn't take the current "bo"
+			 * into account. We don't want to disallow buffer moves
+			 * completely.
+			 */
+			if ((lobj->allowed_domains & current_domain) != 0 &&
+			    (domain & current_domain) == 0 && /* will be moved */
+			    bytes_moved > bytes_moved_threshold) {
+				/* don't move it */
+				domain = current_domain;
+			}
+
 		retry:
 			radeon_ttm_placement_from_domain(bo, domain);
 			if (ring == R600_RING_TYPE_UVD_INDEX)
 				radeon_uvd_force_into_uvd_segment(bo);
-			r = ttm_bo_validate(&bo->tbo, &bo->placement,
-						true, false);
+
+			initial_bytes_moved = atomic64_read(&rdev->num_bytes_moved);
+			r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+			bytes_moved += atomic64_read(&rdev->num_bytes_moved) -
+				       initial_bytes_moved;
+
 			if (unlikely(r)) {
-				if (r != -ERESTARTSYS && domain != lobj->alt_domain) {
-					domain = lobj->alt_domain;
+				if (r != -ERESTARTSYS &&
+				    domain != lobj->allowed_domains) {
+					domain = lobj->allowed_domains;
 					goto retry;
 				}
 				ttm_eu_backoff_reservation(ticket, head);
@@ -564,14 +669,23 @@ int radeon_bo_check_tiling(struct radeon_bo *bo, bool has_moved,
 }
 
 void radeon_bo_move_notify(struct ttm_buffer_object *bo,
-			   struct ttm_mem_reg *mem)
+			   struct ttm_mem_reg *new_mem)
 {
 	struct radeon_bo *rbo;
+
 	if (!radeon_ttm_bo_is_radeon_bo(bo))
 		return;
+
 	rbo = container_of(bo, struct radeon_bo, tbo);
 	radeon_bo_check_tiling(rbo, 0, 1);
 	radeon_vm_bo_invalidate(rbo->rdev, rbo);
+
+	/* update statistics */
+	if (!new_mem)
+		return;
+
+	radeon_update_memory_usage(rbo, bo->mem.mem_type, -1);
+	radeon_update_memory_usage(rbo, new_mem->mem_type, 1);
 }
 
 int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
@@ -586,22 +700,30 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 	rbo = container_of(bo, struct radeon_bo, tbo);
 	radeon_bo_check_tiling(rbo, 0, 0);
 	rdev = rbo->rdev;
-	if (bo->mem.mem_type == TTM_PL_VRAM) {
-		size = bo->mem.num_pages << PAGE_SHIFT;
-		offset = bo->mem.start << PAGE_SHIFT;
-		if ((offset + size) > rdev->mc.visible_vram_size) {
-			/* hurrah the memory is not visible ! */
-			radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM);
-			rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT;
-			r = ttm_bo_validate(bo, &rbo->placement, false, false);
-			if (unlikely(r != 0))
-				return r;
-			offset = bo->mem.start << PAGE_SHIFT;
-			/* this should not happen */
-			if ((offset + size) > rdev->mc.visible_vram_size)
-				return -EINVAL;
-		}
+	if (bo->mem.mem_type != TTM_PL_VRAM)
+		return 0;
+
+	size = bo->mem.num_pages << PAGE_SHIFT;
+	offset = bo->mem.start << PAGE_SHIFT;
+	if ((offset + size) <= rdev->mc.visible_vram_size)
+		return 0;
+
+	/* hurrah the memory is not visible ! */
+	radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM);
+	rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT;
+	r = ttm_bo_validate(bo, &rbo->placement, false, false);
+	if (unlikely(r == -ENOMEM)) {
+		radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_GTT);
+		return ttm_bo_validate(bo, &rbo->placement, false, false);
+	} else if (unlikely(r != 0)) {
+		return r;
 	}
+
+	offset = bo->mem.start << PAGE_SHIFT;
+	/* this should never happen */
+	if ((offset + size) > rdev->mc.visible_vram_size)
+		return -EINVAL;
+
 	return 0;
 }
 
@@ -609,7 +731,7 @@ int radeon_bo_wait(struct radeon_bo *bo, u32 *mem_type, bool no_wait)
 {
 	int r;
 
-	r = ttm_bo_reserve(&bo->tbo, true, no_wait, false, 0);
+	r = ttm_bo_reserve(&bo->tbo, true, no_wait, false, NULL);
 	if (unlikely(r != 0))
 		return r;
 	spin_lock(&bo->tbo.bdev->fence_lock);
diff --git a/drivers/gpu/drm/radeon/radeon_object.h b/drivers/gpu/drm/radeon/radeon_object.h
index 209b1115026..5a873f31a17 100644
--- a/drivers/gpu/drm/radeon/radeon_object.h
+++ b/drivers/gpu/drm/radeon/radeon_object.h
@@ -65,7 +65,7 @@ static inline int radeon_bo_reserve(struct radeon_bo *bo, bool no_intr)
 {
 	int r;
 
-	r = ttm_bo_reserve(&bo->tbo, !no_intr, false, false, 0);
+	r = ttm_bo_reserve(&bo->tbo, !no_intr, false, false, NULL);
 	if (unlikely(r != 0)) {
 		if (r != -ERESTARTSYS)
 			dev_err(bo->rdev->dev, "%p reserve failed\n", bo);
@@ -138,9 +138,8 @@ extern int radeon_bo_evict_vram(struct radeon_device *rdev);
 extern void radeon_bo_force_delete(struct radeon_device *rdev);
 extern int radeon_bo_init(struct radeon_device *rdev);
 extern void radeon_bo_fini(struct radeon_device *rdev);
-extern void radeon_bo_list_add_object(struct radeon_bo_list *lobj,
-				struct list_head *head);
-extern int radeon_bo_list_validate(struct ww_acquire_ctx *ticket,
+extern int radeon_bo_list_validate(struct radeon_device *rdev,
+				   struct ww_acquire_ctx *ticket,
 				   struct list_head *head, int ring);
 extern int radeon_bo_fbdev_mmap(struct radeon_bo *bo,
 				struct vm_area_struct *vma);
@@ -151,7 +150,7 @@ extern void radeon_bo_get_tiling_flags(struct radeon_bo *bo,
 extern int radeon_bo_check_tiling(struct radeon_bo *bo, bool has_moved,
 				bool force_drop);
 extern void radeon_bo_move_notify(struct ttm_buffer_object *bo,
-					struct ttm_mem_reg *mem);
+				  struct ttm_mem_reg *new_mem);
 extern int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo);
 extern int radeon_bo_get_surface_reg(struct radeon_bo *bo);
 
@@ -181,7 +180,7 @@ extern int radeon_sa_bo_manager_suspend(struct radeon_device *rdev,
 extern int radeon_sa_bo_new(struct radeon_device *rdev,
 			    struct radeon_sa_manager *sa_manager,
 			    struct radeon_sa_bo **sa_bo,
-			    unsigned size, unsigned align, bool block);
+			    unsigned size, unsigned align);
 extern void radeon_sa_bo_free(struct radeon_device *rdev,
 			      struct radeon_sa_bo **sa_bo,
 			      struct radeon_fence *fence);
diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c
index 984097b907e..e447e390d09 100644
--- a/drivers/gpu/drm/radeon/radeon_pm.c
+++ b/drivers/gpu/drm/radeon/radeon_pm.c
@@ -73,8 +73,10 @@ void radeon_pm_acpi_event_handler(struct radeon_device *rdev)
 			rdev->pm.dpm.ac_power = true;
 		else
 			rdev->pm.dpm.ac_power = false;
-		if (rdev->asic->dpm.enable_bapm)
-			radeon_dpm_enable_bapm(rdev, rdev->pm.dpm.ac_power);
+		if (rdev->family == CHIP_ARUBA) {
+			if (rdev->asic->dpm.enable_bapm)
+				radeon_dpm_enable_bapm(rdev, rdev->pm.dpm.ac_power);
+		}
 		mutex_unlock(&rdev->pm.mutex);
         } else if (rdev->pm.pm_method == PM_METHOD_PROFILE) {
 		if (rdev->pm.profile == PM_PROFILE_AUTO) {
@@ -260,7 +262,7 @@ static void radeon_pm_set_clocks(struct radeon_device *rdev)
 		if (!ring->ready) {
 			continue;
 		}
-		r = radeon_fence_wait_empty_locked(rdev, i);
+		r = radeon_fence_wait_empty(rdev, i);
 		if (r) {
 			/* needs a GPU reset dont reset here */
 			mutex_unlock(&rdev->ring_lock);
@@ -361,6 +363,11 @@ static ssize_t radeon_set_pm_profile(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct radeon_device *rdev = ddev->dev_private;
 
+	/* Can't set profile when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return -EINVAL;
+
 	mutex_lock(&rdev->pm.mutex);
 	if (rdev->pm.pm_method == PM_METHOD_PROFILE) {
 		if (strncmp("default", buf, strlen("default")) == 0)
@@ -409,6 +416,13 @@ static ssize_t radeon_set_pm_method(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct radeon_device *rdev = ddev->dev_private;
 
+	/* Can't set method when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) {
+		count = -EINVAL;
+		goto fail;
+	}
+
 	/* we don't support the legacy modes with dpm */
 	if (rdev->pm.pm_method == PM_METHOD_DPM) {
 		count = -EINVAL;
@@ -446,6 +460,10 @@ static ssize_t radeon_get_dpm_state(struct device *dev,
 	struct radeon_device *rdev = ddev->dev_private;
 	enum radeon_pm_state_type pm = rdev->pm.dpm.user_state;
 
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return snprintf(buf, PAGE_SIZE, "off\n");
+
 	return snprintf(buf, PAGE_SIZE, "%s\n",
 			(pm == POWER_STATE_TYPE_BATTERY) ? "battery" :
 			(pm == POWER_STATE_TYPE_BALANCED) ? "balanced" : "performance");
@@ -459,6 +477,11 @@ static ssize_t radeon_set_dpm_state(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct radeon_device *rdev = ddev->dev_private;
 
+	/* Can't set dpm state when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return -EINVAL;
+
 	mutex_lock(&rdev->pm.mutex);
 	if (strncmp("battery", buf, strlen("battery")) == 0)
 		rdev->pm.dpm.user_state = POWER_STATE_TYPE_BATTERY;
@@ -485,6 +508,10 @@ static ssize_t radeon_get_dpm_forced_performance_level(struct device *dev,
 	struct radeon_device *rdev = ddev->dev_private;
 	enum radeon_dpm_forced_level level = rdev->pm.dpm.forced_level;
 
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return snprintf(buf, PAGE_SIZE, "off\n");
+
 	return snprintf(buf, PAGE_SIZE, "%s\n",
 			(level == RADEON_DPM_FORCED_LEVEL_AUTO) ? "auto" :
 			(level == RADEON_DPM_FORCED_LEVEL_LOW) ? "low" : "high");
@@ -500,6 +527,11 @@ static ssize_t radeon_set_dpm_forced_performance_level(struct device *dev,
 	enum radeon_dpm_forced_level level;
 	int ret = 0;
 
+	/* Can't force performance level when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return -EINVAL;
+
 	mutex_lock(&rdev->pm.mutex);
 	if (strncmp("low", buf, strlen("low")) == 0) {
 		level = RADEON_DPM_FORCED_LEVEL_LOW;
@@ -538,8 +570,14 @@ static ssize_t radeon_hwmon_show_temp(struct device *dev,
 				      char *buf)
 {
 	struct radeon_device *rdev = dev_get_drvdata(dev);
+	struct drm_device *ddev = rdev->ddev;
 	int temp;
 
+	/* Can't get temperature when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return -EINVAL;
+
 	if (rdev->asic->pm.get_temperature)
 		temp = radeon_get_temperature(rdev);
 	else
@@ -603,7 +641,6 @@ static const struct attribute_group *hwmon_groups[] = {
 static int radeon_hwmon_init(struct radeon_device *rdev)
 {
 	int err = 0;
-	struct device *hwmon_dev;
 
 	switch (rdev->pm.int_thermal_type) {
 	case THERMAL_TYPE_RV6XX:
@@ -616,11 +653,11 @@ static int radeon_hwmon_init(struct radeon_device *rdev)
 	case THERMAL_TYPE_KV:
 		if (rdev->asic->pm.get_temperature == NULL)
 			return err;
-		hwmon_dev = hwmon_device_register_with_groups(rdev->dev,
-							      "radeon", rdev,
-							      hwmon_groups);
-		if (IS_ERR(hwmon_dev)) {
-			err = PTR_ERR(hwmon_dev);
+		rdev->pm.int_hwmon_dev = hwmon_device_register_with_groups(rdev->dev,
+									   "radeon", rdev,
+									   hwmon_groups);
+		if (IS_ERR(rdev->pm.int_hwmon_dev)) {
+			err = PTR_ERR(rdev->pm.int_hwmon_dev);
 			dev_err(rdev->dev,
 				"Unable to register hwmon device: %d\n", err);
 		}
@@ -632,6 +669,12 @@ static int radeon_hwmon_init(struct radeon_device *rdev)
 	return err;
 }
 
+static void radeon_hwmon_fini(struct radeon_device *rdev)
+{
+	if (rdev->pm.int_hwmon_dev)
+		hwmon_device_unregister(rdev->pm.int_hwmon_dev);
+}
+
 static void radeon_dpm_thermal_work_handler(struct work_struct *work)
 {
 	struct radeon_device *rdev =
@@ -826,6 +869,9 @@ static void radeon_dpm_change_power_state_locked(struct radeon_device *rdev)
 
 	/* no need to reprogram if nothing changed unless we are on BTC+ */
 	if (rdev->pm.dpm.current_ps == rdev->pm.dpm.requested_ps) {
+		/* vce just modifies an existing state so force a change */
+		if (ps->vce_active != rdev->pm.dpm.vce_active)
+			goto force;
 		if ((rdev->family < CHIP_BARTS) || (rdev->flags & RADEON_IS_IGP)) {
 			/* for pre-BTC and APUs if the num crtcs changed but state is the same,
 			 * all we need to do is update the display configuration.
@@ -862,16 +908,21 @@ static void radeon_dpm_change_power_state_locked(struct radeon_device *rdev)
 		}
 	}
 
+force:
 	if (radeon_dpm == 1) {
 		printk("switching from power state:\n");
 		radeon_dpm_print_power_state(rdev, rdev->pm.dpm.current_ps);
 		printk("switching to power state:\n");
 		radeon_dpm_print_power_state(rdev, rdev->pm.dpm.requested_ps);
 	}
+
 	mutex_lock(&rdev->ddev->struct_mutex);
 	down_write(&rdev->pm.mclk_lock);
 	mutex_lock(&rdev->ring_lock);
 
+	/* update whether vce is active */
+	ps->vce_active = rdev->pm.dpm.vce_active;
+
 	ret = radeon_dpm_pre_set_power_state(rdev);
 	if (ret)
 		goto done;
@@ -888,7 +939,7 @@ static void radeon_dpm_change_power_state_locked(struct radeon_device *rdev)
 	for (i = 0; i < RADEON_NUM_RINGS; i++) {
 		struct radeon_ring *ring = &rdev->ring[i];
 		if (ring->ready)
-			radeon_fence_wait_empty_locked(rdev, i);
+			radeon_fence_wait_empty(rdev, i);
 	}
 
 	/* program the new power state */
@@ -924,6 +975,10 @@ void radeon_dpm_enable_uvd(struct radeon_device *rdev, bool enable)
 
 	if (rdev->asic->dpm.powergate_uvd) {
 		mutex_lock(&rdev->pm.mutex);
+		/* don't powergate anything if we
+		   have active but pause streams */
+		enable |= rdev->pm.dpm.sd > 0;
+		enable |= rdev->pm.dpm.hd > 0;
 		/* enable/disable UVD */
 		radeon_dpm_powergate_uvd(rdev, !enable);
 		mutex_unlock(&rdev->pm.mutex);
@@ -956,6 +1011,23 @@ void radeon_dpm_enable_uvd(struct radeon_device *rdev, bool enable)
 	}
 }
 
+void radeon_dpm_enable_vce(struct radeon_device *rdev, bool enable)
+{
+	if (enable) {
+		mutex_lock(&rdev->pm.mutex);
+		rdev->pm.dpm.vce_active = true;
+		/* XXX select vce level based on ring/task */
+		rdev->pm.dpm.vce_level = RADEON_VCE_LEVEL_AC_ALL;
+		mutex_unlock(&rdev->pm.mutex);
+	} else {
+		mutex_lock(&rdev->pm.mutex);
+		rdev->pm.dpm.vce_active = false;
+		mutex_unlock(&rdev->pm.mutex);
+	}
+
+	radeon_pm_compute_clocks(rdev);
+}
+
 static void radeon_pm_suspend_old(struct radeon_device *rdev)
 {
 	mutex_lock(&rdev->pm.mutex);
@@ -1010,8 +1082,10 @@ static void radeon_pm_resume_old(struct radeon_device *rdev)
 	rdev->pm.current_clock_mode_index = 0;
 	rdev->pm.current_sclk = rdev->pm.default_sclk;
 	rdev->pm.current_mclk = rdev->pm.default_mclk;
-	rdev->pm.current_vddc = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.voltage;
-	rdev->pm.current_vddci = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.vddci;
+	if (rdev->pm.power_state) {
+		rdev->pm.current_vddc = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.voltage;
+		rdev->pm.current_vddci = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.vddci;
+	}
 	if (rdev->pm.pm_method == PM_METHOD_DYNPM
 	    && rdev->pm.dynpm_state == DYNPM_STATE_SUSPENDED) {
 		rdev->pm.dynpm_state = DYNPM_STATE_ACTIVE;
@@ -1032,25 +1106,26 @@ static void radeon_pm_resume_dpm(struct radeon_device *rdev)
 	radeon_dpm_setup_asic(rdev);
 	ret = radeon_dpm_enable(rdev);
 	mutex_unlock(&rdev->pm.mutex);
-	if (ret) {
-		DRM_ERROR("radeon: dpm resume failed\n");
-		if ((rdev->family >= CHIP_BARTS) &&
-		    (rdev->family <= CHIP_CAYMAN) &&
-		    rdev->mc_fw) {
-			if (rdev->pm.default_vddc)
-				radeon_atom_set_voltage(rdev, rdev->pm.default_vddc,
-							SET_VOLTAGE_TYPE_ASIC_VDDC);
-			if (rdev->pm.default_vddci)
-				radeon_atom_set_voltage(rdev, rdev->pm.default_vddci,
-							SET_VOLTAGE_TYPE_ASIC_VDDCI);
-			if (rdev->pm.default_sclk)
-				radeon_set_engine_clock(rdev, rdev->pm.default_sclk);
-			if (rdev->pm.default_mclk)
-				radeon_set_memory_clock(rdev, rdev->pm.default_mclk);
-		}
-	} else {
-		rdev->pm.dpm_enabled = true;
-		radeon_pm_compute_clocks(rdev);
+	if (ret)
+		goto dpm_resume_fail;
+	rdev->pm.dpm_enabled = true;
+	return;
+
+dpm_resume_fail:
+	DRM_ERROR("radeon: dpm resume failed\n");
+	if ((rdev->family >= CHIP_BARTS) &&
+	    (rdev->family <= CHIP_CAYMAN) &&
+	    rdev->mc_fw) {
+		if (rdev->pm.default_vddc)
+			radeon_atom_set_voltage(rdev, rdev->pm.default_vddc,
+						SET_VOLTAGE_TYPE_ASIC_VDDC);
+		if (rdev->pm.default_vddci)
+			radeon_atom_set_voltage(rdev, rdev->pm.default_vddci,
+						SET_VOLTAGE_TYPE_ASIC_VDDCI);
+		if (rdev->pm.default_sclk)
+			radeon_set_engine_clock(rdev, rdev->pm.default_sclk);
+		if (rdev->pm.default_mclk)
+			radeon_set_memory_clock(rdev, rdev->pm.default_mclk);
 	}
 }
 
@@ -1170,51 +1245,50 @@ static int radeon_pm_init_dpm(struct radeon_device *rdev)
 	radeon_dpm_setup_asic(rdev);
 	ret = radeon_dpm_enable(rdev);
 	mutex_unlock(&rdev->pm.mutex);
-	if (ret) {
-		rdev->pm.dpm_enabled = false;
-		if ((rdev->family >= CHIP_BARTS) &&
-		    (rdev->family <= CHIP_CAYMAN) &&
-		    rdev->mc_fw) {
-			if (rdev->pm.default_vddc)
-				radeon_atom_set_voltage(rdev, rdev->pm.default_vddc,
-							SET_VOLTAGE_TYPE_ASIC_VDDC);
-			if (rdev->pm.default_vddci)
-				radeon_atom_set_voltage(rdev, rdev->pm.default_vddci,
-							SET_VOLTAGE_TYPE_ASIC_VDDCI);
-			if (rdev->pm.default_sclk)
-				radeon_set_engine_clock(rdev, rdev->pm.default_sclk);
-			if (rdev->pm.default_mclk)
-				radeon_set_memory_clock(rdev, rdev->pm.default_mclk);
-		}
-		DRM_ERROR("radeon: dpm initialization failed\n");
-		return ret;
-	}
+	if (ret)
+		goto dpm_failed;
 	rdev->pm.dpm_enabled = true;
-	radeon_pm_compute_clocks(rdev);
 
-	if (rdev->pm.num_power_states > 1) {
-		ret = device_create_file(rdev->dev, &dev_attr_power_dpm_state);
-		if (ret)
-			DRM_ERROR("failed to create device file for dpm state\n");
-		ret = device_create_file(rdev->dev, &dev_attr_power_dpm_force_performance_level);
-		if (ret)
-			DRM_ERROR("failed to create device file for dpm state\n");
-		/* XXX: these are noops for dpm but are here for backwards compat */
-		ret = device_create_file(rdev->dev, &dev_attr_power_profile);
-		if (ret)
-			DRM_ERROR("failed to create device file for power profile\n");
-		ret = device_create_file(rdev->dev, &dev_attr_power_method);
-		if (ret)
-			DRM_ERROR("failed to create device file for power method\n");
-
-		if (radeon_debugfs_pm_init(rdev)) {
-			DRM_ERROR("Failed to register debugfs file for dpm!\n");
-		}
+	ret = device_create_file(rdev->dev, &dev_attr_power_dpm_state);
+	if (ret)
+		DRM_ERROR("failed to create device file for dpm state\n");
+	ret = device_create_file(rdev->dev, &dev_attr_power_dpm_force_performance_level);
+	if (ret)
+		DRM_ERROR("failed to create device file for dpm state\n");
+	/* XXX: these are noops for dpm but are here for backwards compat */
+	ret = device_create_file(rdev->dev, &dev_attr_power_profile);
+	if (ret)
+		DRM_ERROR("failed to create device file for power profile\n");
+	ret = device_create_file(rdev->dev, &dev_attr_power_method);
+	if (ret)
+		DRM_ERROR("failed to create device file for power method\n");
 
-		DRM_INFO("radeon: dpm initialized\n");
+	if (radeon_debugfs_pm_init(rdev)) {
+		DRM_ERROR("Failed to register debugfs file for dpm!\n");
 	}
 
+	DRM_INFO("radeon: dpm initialized\n");
+
 	return 0;
+
+dpm_failed:
+	rdev->pm.dpm_enabled = false;
+	if ((rdev->family >= CHIP_BARTS) &&
+	    (rdev->family <= CHIP_CAYMAN) &&
+	    rdev->mc_fw) {
+		if (rdev->pm.default_vddc)
+			radeon_atom_set_voltage(rdev, rdev->pm.default_vddc,
+						SET_VOLTAGE_TYPE_ASIC_VDDC);
+		if (rdev->pm.default_vddci)
+			radeon_atom_set_voltage(rdev, rdev->pm.default_vddci,
+						SET_VOLTAGE_TYPE_ASIC_VDDCI);
+		if (rdev->pm.default_sclk)
+			radeon_set_engine_clock(rdev, rdev->pm.default_sclk);
+		if (rdev->pm.default_mclk)
+			radeon_set_memory_clock(rdev, rdev->pm.default_mclk);
+	}
+	DRM_ERROR("radeon: dpm initialization failed\n");
+	return ret;
 }
 
 int radeon_pm_init(struct radeon_device *rdev)
@@ -1228,11 +1302,11 @@ int radeon_pm_init(struct radeon_device *rdev)
 	case CHIP_RV670:
 	case CHIP_RS780:
 	case CHIP_RS880:
+	case CHIP_RV770:
+	case CHIP_BARTS:
+	case CHIP_TURKS:
+	case CHIP_CAICOS:
 	case CHIP_CAYMAN:
-	case CHIP_BONAIRE:
-	case CHIP_KABINI:
-	case CHIP_KAVERI:
-	case CHIP_HAWAII:
 		/* DPM requires the RLC, RV770+ dGPU requires SMC */
 		if (!rdev->rlc_fw)
 			rdev->pm.pm_method = PM_METHOD_PROFILE;
@@ -1245,7 +1319,6 @@ int radeon_pm_init(struct radeon_device *rdev)
 		else
 			rdev->pm.pm_method = PM_METHOD_PROFILE;
 		break;
-	case CHIP_RV770:
 	case CHIP_RV730:
 	case CHIP_RV710:
 	case CHIP_RV740:
@@ -1257,15 +1330,17 @@ int radeon_pm_init(struct radeon_device *rdev)
 	case CHIP_PALM:
 	case CHIP_SUMO:
 	case CHIP_SUMO2:
-	case CHIP_BARTS:
-	case CHIP_TURKS:
-	case CHIP_CAICOS:
 	case CHIP_ARUBA:
 	case CHIP_TAHITI:
 	case CHIP_PITCAIRN:
 	case CHIP_VERDE:
 	case CHIP_OLAND:
 	case CHIP_HAINAN:
+	case CHIP_BONAIRE:
+	case CHIP_KABINI:
+	case CHIP_KAVERI:
+	case CHIP_HAWAII:
+	case CHIP_MULLINS:
 		/* DPM requires the RLC, RV770+ dGPU requires SMC */
 		if (!rdev->rlc_fw)
 			rdev->pm.pm_method = PM_METHOD_PROFILE;
@@ -1290,6 +1365,18 @@ int radeon_pm_init(struct radeon_device *rdev)
 		return radeon_pm_init_old(rdev);
 }
 
+int radeon_pm_late_init(struct radeon_device *rdev)
+{
+	int ret = 0;
+
+	if (rdev->pm.pm_method == PM_METHOD_DPM) {
+		mutex_lock(&rdev->pm.mutex);
+		ret = radeon_dpm_late_enable(rdev);
+		mutex_unlock(&rdev->pm.mutex);
+	}
+	return ret;
+}
+
 static void radeon_pm_fini_old(struct radeon_device *rdev)
 {
 	if (rdev->pm.num_power_states > 1) {
@@ -1312,6 +1399,8 @@ static void radeon_pm_fini_old(struct radeon_device *rdev)
 		device_remove_file(rdev->dev, &dev_attr_power_method);
 	}
 
+	radeon_hwmon_fini(rdev);
+
 	if (rdev->pm.power_state)
 		kfree(rdev->pm.power_state);
 }
@@ -1331,6 +1420,8 @@ static void radeon_pm_fini_dpm(struct radeon_device *rdev)
 	}
 	radeon_dpm_fini(rdev);
 
+	radeon_hwmon_fini(rdev);
+
 	if (rdev->pm.power_state)
 		kfree(rdev->pm.power_state);
 }
@@ -1356,12 +1447,14 @@ static void radeon_pm_compute_clocks_old(struct radeon_device *rdev)
 
 	rdev->pm.active_crtcs = 0;
 	rdev->pm.active_crtc_count = 0;
-	list_for_each_entry(crtc,
-		&ddev->mode_config.crtc_list, head) {
-		radeon_crtc = to_radeon_crtc(crtc);
-		if (radeon_crtc->enabled) {
-			rdev->pm.active_crtcs |= (1 << radeon_crtc->crtc_id);
-			rdev->pm.active_crtc_count++;
+	if (rdev->num_crtc && rdev->mode_info.mode_config_initialized) {
+		list_for_each_entry(crtc,
+				    &ddev->mode_config.crtc_list, head) {
+			radeon_crtc = to_radeon_crtc(crtc);
+			if (radeon_crtc->enabled) {
+				rdev->pm.active_crtcs |= (1 << radeon_crtc->crtc_id);
+				rdev->pm.active_crtc_count++;
+			}
 		}
 	}
 
@@ -1420,17 +1513,22 @@ static void radeon_pm_compute_clocks_dpm(struct radeon_device *rdev)
 	struct drm_crtc *crtc;
 	struct radeon_crtc *radeon_crtc;
 
+	if (!rdev->pm.dpm_enabled)
+		return;
+
 	mutex_lock(&rdev->pm.mutex);
 
 	/* update active crtc counts */
 	rdev->pm.dpm.new_active_crtcs = 0;
 	rdev->pm.dpm.new_active_crtc_count = 0;
-	list_for_each_entry(crtc,
-		&ddev->mode_config.crtc_list, head) {
-		radeon_crtc = to_radeon_crtc(crtc);
-		if (crtc->enabled) {
-			rdev->pm.dpm.new_active_crtcs |= (1 << radeon_crtc->crtc_id);
-			rdev->pm.dpm.new_active_crtc_count++;
+	if (rdev->num_crtc && rdev->mode_info.mode_config_initialized) {
+		list_for_each_entry(crtc,
+				    &ddev->mode_config.crtc_list, head) {
+			radeon_crtc = to_radeon_crtc(crtc);
+			if (crtc->enabled) {
+				rdev->pm.dpm.new_active_crtcs |= (1 << radeon_crtc->crtc_id);
+				rdev->pm.dpm.new_active_crtc_count++;
+			}
 		}
 	}
 
@@ -1464,7 +1562,7 @@ static bool radeon_pm_in_vbl(struct radeon_device *rdev)
 	 */
 	for (crtc = 0; (crtc < rdev->num_crtc) && in_vbl; crtc++) {
 		if (rdev->pm.active_crtcs & (1 << crtc)) {
-			vbl_status = radeon_get_crtc_scanoutpos(rdev->ddev, crtc, &vpos, &hpos, NULL, NULL);
+			vbl_status = radeon_get_crtc_scanoutpos(rdev->ddev, crtc, 0, &vpos, &hpos, NULL, NULL);
 			if ((vbl_status & DRM_SCANOUTPOS_VALID) &&
 			    !(vbl_status & DRM_SCANOUTPOS_INVBL))
 				in_vbl = false;
@@ -1556,8 +1654,12 @@ static int radeon_debugfs_pm_info(struct seq_file *m, void *data)
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct radeon_device *rdev = dev->dev_private;
+	struct drm_device *ddev = rdev->ddev;
 
-	if (rdev->pm.dpm_enabled) {
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) {
+		seq_printf(m, "PX asic powered off\n");
+	} else if (rdev->pm.dpm_enabled) {
 		mutex_lock(&rdev->pm.mutex);
 		if (rdev->asic->dpm.debugfs_print_current_performance_level)
 			radeon_dpm_debugfs_print_current_performance_level(rdev, m);
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index 9214403ae17..f8050f5429e 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -63,7 +63,7 @@ int radeon_ib_get(struct radeon_device *rdev, int ring,
 {
 	int r;
 
-	r = radeon_sa_bo_new(rdev, &rdev->ring_tmp_bo, &ib->sa_bo, size, 256, true);
+	r = radeon_sa_bo_new(rdev, &rdev->ring_tmp_bo, &ib->sa_bo, size, 256);
 	if (r) {
 		dev_err(rdev->dev, "failed to get a new IB (%d)\n", r);
 		return r;
@@ -139,12 +139,19 @@ int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib,
 	}
 
 	/* 64 dwords should be enough for fence too */
-	r = radeon_ring_lock(rdev, ring, 64 + RADEON_NUM_RINGS * 8);
+	r = radeon_ring_lock(rdev, ring, 64 + RADEON_NUM_SYNCS * 8);
 	if (r) {
 		dev_err(rdev->dev, "scheduling IB failed (%d).\n", r);
 		return r;
 	}
 
+	/* grab a vm id if necessary */
+	if (ib->vm) {
+		struct radeon_fence *vm_id_fence;
+		vm_id_fence = radeon_vm_grab_id(rdev, ib->vm, ib->ring);
+        	radeon_semaphore_sync_to(ib->semaphore, vm_id_fence);
+	}
+
 	/* sync with other rings */
 	r = radeon_semaphore_sync_rings(rdev, ib->semaphore, ib->ring);
 	if (r) {
@@ -153,11 +160,9 @@ int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib,
 		return r;
 	}
 
-	/* if we can't remember our last VM flush then flush now! */
-	/* XXX figure out why we have to flush for every IB */
-	if (ib->vm /*&& !ib->vm->last_flush*/) {
-		radeon_ring_vm_flush(rdev, ib->ring, ib->vm);
-	}
+	if (ib->vm)
+		radeon_vm_flush(rdev, ib->vm, ib->ring);
+
 	if (const_ib) {
 		radeon_ring_ib_execute(rdev, const_ib->ring, const_ib);
 		radeon_semaphore_free(rdev, &const_ib->semaphore, NULL);
@@ -172,10 +177,10 @@ int radeon_ib_schedule(struct radeon_device *rdev, struct radeon_ib *ib,
 	if (const_ib) {
 		const_ib->fence = radeon_fence_ref(ib->fence);
 	}
-	/* we just flushed the VM, remember that */
-	if (ib->vm && !ib->vm->last_flush) {
-		ib->vm->last_flush = radeon_fence_ref(ib->fence);
-	}
+
+	if (ib->vm)
+		radeon_vm_fence(rdev, ib->vm, ib->fence);
+
 	radeon_ring_unlock_commit(rdev, ring);
 	return 0;
 }
@@ -257,6 +262,7 @@ int radeon_ib_ring_tests(struct radeon_device *rdev)
 		r = radeon_ib_test(rdev, i, ring);
 		if (r) {
 			ring->ready = false;
+			rdev->needs_reset = false;
 
 			if (i == RADEON_RING_TYPE_GFX_INDEX) {
 				/* oh, oh, that's really bad */
@@ -332,36 +338,6 @@ bool radeon_ring_supports_scratch_reg(struct radeon_device *rdev,
 	}
 }
 
-u32 radeon_ring_generic_get_rptr(struct radeon_device *rdev,
-				 struct radeon_ring *ring)
-{
-	u32 rptr;
-
-	if (rdev->wb.enabled)
-		rptr = le32_to_cpu(rdev->wb.wb[ring->rptr_offs/4]);
-	else
-		rptr = RREG32(ring->rptr_reg);
-
-	return rptr;
-}
-
-u32 radeon_ring_generic_get_wptr(struct radeon_device *rdev,
-				 struct radeon_ring *ring)
-{
-	u32 wptr;
-
-	wptr = RREG32(ring->wptr_reg);
-
-	return wptr;
-}
-
-void radeon_ring_generic_set_wptr(struct radeon_device *rdev,
-				  struct radeon_ring *ring)
-{
-	WREG32(ring->wptr_reg, ring->wptr);
-	(void)RREG32(ring->wptr_reg);
-}
-
 /**
  * radeon_ring_free_size - update the free size
  *
@@ -372,13 +348,17 @@ void radeon_ring_generic_set_wptr(struct radeon_device *rdev,
  */
 void radeon_ring_free_size(struct radeon_device *rdev, struct radeon_ring *ring)
 {
-	ring->rptr = radeon_ring_get_rptr(rdev, ring);
+	uint32_t rptr = radeon_ring_get_rptr(rdev, ring);
+
 	/* This works because ring_size is a power of 2 */
-	ring->ring_free_dw = (ring->rptr + (ring->ring_size / 4));
+	ring->ring_free_dw = rptr + (ring->ring_size / 4);
 	ring->ring_free_dw -= ring->wptr;
 	ring->ring_free_dw &= ring->ptr_mask;
 	if (!ring->ring_free_dw) {
+		/* this is an empty ring */
 		ring->ring_free_dw = ring->ring_size / 4;
+		/*  update lockup info to avoid false positive */
+		radeon_ring_lockup_update(rdev, ring);
 	}
 }
 
@@ -402,19 +382,13 @@ int radeon_ring_alloc(struct radeon_device *rdev, struct radeon_ring *ring, unsi
 	/* Align requested size with padding so unlock_commit can
 	 * pad safely */
 	radeon_ring_free_size(rdev, ring);
-	if (ring->ring_free_dw == (ring->ring_size / 4)) {
-		/* This is an empty ring update lockup info to avoid
-		 * false positive.
-		 */
-		radeon_ring_lockup_update(ring);
-	}
 	ndw = (ndw + ring->align_mask) & ~ring->align_mask;
 	while (ndw > (ring->ring_free_dw - 1)) {
 		radeon_ring_free_size(rdev, ring);
 		if (ndw < ring->ring_free_dw) {
 			break;
 		}
-		r = radeon_fence_wait_next_locked(rdev, ring->idx);
+		r = radeon_fence_wait_next(rdev, ring->idx);
 		if (r)
 			return r;
 	}
@@ -463,7 +437,7 @@ void radeon_ring_commit(struct radeon_device *rdev, struct radeon_ring *ring)
 	while (ring->wptr & ring->align_mask) {
 		radeon_ring_write(ring, ring->nop);
 	}
-	DRM_MEMORYBARRIER();
+	mb();
 	radeon_ring_set_wptr(rdev, ring);
 }
 
@@ -508,39 +482,17 @@ void radeon_ring_unlock_undo(struct radeon_device *rdev, struct radeon_ring *rin
 }
 
 /**
- * radeon_ring_force_activity - add some nop packets to the ring
- *
- * @rdev: radeon_device pointer
- * @ring: radeon_ring structure holding ring information
- *
- * Add some nop packets to the ring to force activity (all asics).
- * Used for lockup detection to see if the rptr is advancing.
- */
-void radeon_ring_force_activity(struct radeon_device *rdev, struct radeon_ring *ring)
-{
-	int r;
-
-	radeon_ring_free_size(rdev, ring);
-	if (ring->rptr == ring->wptr) {
-		r = radeon_ring_alloc(rdev, ring, 1);
-		if (!r) {
-			radeon_ring_write(ring, ring->nop);
-			radeon_ring_commit(rdev, ring);
-		}
-	}
-}
-
-/**
  * radeon_ring_lockup_update - update lockup variables
  *
  * @ring: radeon_ring structure holding ring information
  *
  * Update the last rptr value and timestamp (all asics).
  */
-void radeon_ring_lockup_update(struct radeon_ring *ring)
+void radeon_ring_lockup_update(struct radeon_device *rdev,
+			       struct radeon_ring *ring)
 {
-	ring->last_rptr = ring->rptr;
-	ring->last_activity = jiffies;
+	atomic_set(&ring->last_rptr, radeon_ring_get_rptr(rdev, ring));
+	atomic64_set(&ring->last_activity, jiffies_64);
 }
 
 /**
@@ -548,40 +500,23 @@ void radeon_ring_lockup_update(struct radeon_ring *ring)
  * @rdev:       radeon device structure
  * @ring:       radeon_ring structure holding ring information
  *
- * We don't need to initialize the lockup tracking information as we will either
- * have CP rptr to a different value of jiffies wrap around which will force
- * initialization of the lockup tracking informations.
- *
- * A possible false positivie is if we get call after while and last_cp_rptr ==
- * the current CP rptr, even if it's unlikely it might happen. To avoid this
- * if the elapsed time since last call is bigger than 2 second than we return
- * false and update the tracking information. Due to this the caller must call
- * radeon_ring_test_lockup several time in less than 2sec for lockup to be reported
- * the fencing code should be cautious about that.
- *
- * Caller should write to the ring to force CP to do something so we don't get
- * false positive when CP is just gived nothing to do.
- *
- **/
+ */
 bool radeon_ring_test_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 {
-	unsigned long cjiffies, elapsed;
+	uint32_t rptr = radeon_ring_get_rptr(rdev, ring);
+	uint64_t last = atomic64_read(&ring->last_activity);
+	uint64_t elapsed;
 
-	cjiffies = jiffies;
-	if (!time_after(cjiffies, ring->last_activity)) {
-		/* likely a wrap around */
-		radeon_ring_lockup_update(ring);
-		return false;
-	}
-	ring->rptr = radeon_ring_get_rptr(rdev, ring);
-	if (ring->rptr != ring->last_rptr) {
-		/* CP is still working no lockup */
-		radeon_ring_lockup_update(ring);
+	if (rptr != atomic_read(&ring->last_rptr)) {
+		/* ring is still working, no lockup */
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	elapsed = jiffies_to_msecs(cjiffies - ring->last_activity);
+
+	elapsed = jiffies_to_msecs(jiffies_64 - last);
 	if (radeon_lockup_timeout && elapsed >= radeon_lockup_timeout) {
-		dev_err(rdev->dev, "GPU lockup CP stall for more than %lumsec\n", elapsed);
+		dev_err(rdev->dev, "ring %d stalled for more than %llumsec\n",
+			ring->idx, elapsed);
 		return true;
 	}
 	/* give a chance to the GPU ... */
@@ -689,22 +624,18 @@ int radeon_ring_restore(struct radeon_device *rdev, struct radeon_ring *ring,
  * @ring: radeon_ring structure holding ring information
  * @ring_size: size of the ring
  * @rptr_offs: offset of the rptr writeback location in the WB buffer
- * @rptr_reg: MMIO offset of the rptr register
- * @wptr_reg: MMIO offset of the wptr register
  * @nop: nop packet for this ring
  *
  * Initialize the driver information for the selected ring (all asics).
  * Returns 0 on success, error on failure.
  */
 int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsigned ring_size,
-		     unsigned rptr_offs, unsigned rptr_reg, unsigned wptr_reg, u32 nop)
+		     unsigned rptr_offs, u32 nop)
 {
 	int r;
 
 	ring->ring_size = ring_size;
 	ring->rptr_offs = rptr_offs;
-	ring->rptr_reg = rptr_reg;
-	ring->wptr_reg = wptr_reg;
 	ring->nop = nop;
 	/* Allocate ring buffer */
 	if (ring->ring_obj == NULL) {
@@ -743,7 +674,7 @@ int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsig
 	if (radeon_debugfs_ring_init(rdev, ring)) {
 		DRM_ERROR("Failed to register debugfs file for rings !\n");
 	}
-	radeon_ring_lockup_update(ring);
+	radeon_ring_lockup_update(rdev, ring);
 	return 0;
 }
 
@@ -790,34 +721,52 @@ static int radeon_debugfs_ring_info(struct seq_file *m, void *data)
 	struct radeon_device *rdev = dev->dev_private;
 	int ridx = *(int*)node->info_ent->data;
 	struct radeon_ring *ring = &rdev->ring[ridx];
+
+	uint32_t rptr, wptr, rptr_next;
 	unsigned count, i, j;
-	u32 tmp;
 
 	radeon_ring_free_size(rdev, ring);
 	count = (ring->ring_size / 4) - ring->ring_free_dw;
-	tmp = radeon_ring_get_wptr(rdev, ring);
-	seq_printf(m, "wptr(0x%04x): 0x%08x [%5d]\n", ring->wptr_reg, tmp, tmp);
-	tmp = radeon_ring_get_rptr(rdev, ring);
-	seq_printf(m, "rptr(0x%04x): 0x%08x [%5d]\n", ring->rptr_reg, tmp, tmp);
+
+	wptr = radeon_ring_get_wptr(rdev, ring);
+	seq_printf(m, "wptr: 0x%08x [%5d]\n",
+		   wptr, wptr);
+
+	rptr = radeon_ring_get_rptr(rdev, ring);
+	seq_printf(m, "rptr: 0x%08x [%5d]\n",
+		   rptr, rptr);
+
 	if (ring->rptr_save_reg) {
-		seq_printf(m, "rptr next(0x%04x): 0x%08x\n", ring->rptr_save_reg,
-			   RREG32(ring->rptr_save_reg));
-	}
-	seq_printf(m, "driver's copy of the wptr: 0x%08x [%5d]\n", ring->wptr, ring->wptr);
-	seq_printf(m, "driver's copy of the rptr: 0x%08x [%5d]\n", ring->rptr, ring->rptr);
-	seq_printf(m, "last semaphore signal addr : 0x%016llx\n", ring->last_semaphore_signal_addr);
-	seq_printf(m, "last semaphore wait addr   : 0x%016llx\n", ring->last_semaphore_wait_addr);
+		rptr_next = RREG32(ring->rptr_save_reg);
+		seq_printf(m, "rptr next(0x%04x): 0x%08x [%5d]\n",
+			   ring->rptr_save_reg, rptr_next, rptr_next);
+	} else
+		rptr_next = ~0;
+
+	seq_printf(m, "driver's copy of the wptr: 0x%08x [%5d]\n",
+		   ring->wptr, ring->wptr);
+	seq_printf(m, "last semaphore signal addr : 0x%016llx\n",
+		   ring->last_semaphore_signal_addr);
+	seq_printf(m, "last semaphore wait addr   : 0x%016llx\n",
+		   ring->last_semaphore_wait_addr);
 	seq_printf(m, "%u free dwords in ring\n", ring->ring_free_dw);
 	seq_printf(m, "%u dwords in ring\n", count);
+
+	if (!ring->ready)
+		return 0;
+
 	/* print 8 dw before current rptr as often it's the last executed
 	 * packet that is the root issue
 	 */
-	i = (ring->rptr + ring->ptr_mask + 1 - 32) & ring->ptr_mask;
-	if (ring->ready) {
-		for (j = 0; j <= (count + 32); j++) {
-			seq_printf(m, "r[%5d]=0x%08x\n", i, ring->ring[i]);
-			i = (i + 1) & ring->ptr_mask;
-		}
+	i = (rptr + ring->ptr_mask + 1 - 32) & ring->ptr_mask;
+	for (j = 0; j <= (count + 32); j++) {
+		seq_printf(m, "r[%5d]=0x%08x", i, ring->ring[i]);
+		if (rptr == i)
+			seq_puts(m, " *");
+		if (rptr_next == i)
+			seq_puts(m, " #");
+		seq_puts(m, "\n");
+		i = (i + 1) & ring->ptr_mask;
 	}
 	return 0;
 }
@@ -828,6 +777,8 @@ static int cayman_cp2_index = CAYMAN_RING_TYPE_CP2_INDEX;
 static int radeon_dma1_index = R600_RING_TYPE_DMA_INDEX;
 static int radeon_dma2_index = CAYMAN_RING_TYPE_DMA1_INDEX;
 static int r600_uvd_index = R600_RING_TYPE_UVD_INDEX;
+static int si_vce1_index = TN_RING_TYPE_VCE1_INDEX;
+static int si_vce2_index = TN_RING_TYPE_VCE2_INDEX;
 
 static struct drm_info_list radeon_debugfs_ring_info_list[] = {
 	{"radeon_ring_gfx", radeon_debugfs_ring_info, 0, &radeon_gfx_index},
@@ -836,6 +787,8 @@ static struct drm_info_list radeon_debugfs_ring_info_list[] = {
 	{"radeon_ring_dma1", radeon_debugfs_ring_info, 0, &radeon_dma1_index},
 	{"radeon_ring_dma2", radeon_debugfs_ring_info, 0, &radeon_dma2_index},
 	{"radeon_ring_uvd", radeon_debugfs_ring_info, 0, &r600_uvd_index},
+	{"radeon_ring_vce1", radeon_debugfs_ring_info, 0, &si_vce1_index},
+	{"radeon_ring_vce2", radeon_debugfs_ring_info, 0, &si_vce2_index},
 };
 
 static int radeon_debugfs_sa_info(struct seq_file *m, void *data)
diff --git a/drivers/gpu/drm/radeon/radeon_sa.c b/drivers/gpu/drm/radeon/radeon_sa.c
index f0bac68254b..adcf3e2f07d 100644
--- a/drivers/gpu/drm/radeon/radeon_sa.c
+++ b/drivers/gpu/drm/radeon/radeon_sa.c
@@ -312,7 +312,7 @@ static bool radeon_sa_bo_next_hole(struct radeon_sa_manager *sa_manager,
 int radeon_sa_bo_new(struct radeon_device *rdev,
 		     struct radeon_sa_manager *sa_manager,
 		     struct radeon_sa_bo **sa_bo,
-		     unsigned size, unsigned align, bool block)
+		     unsigned size, unsigned align)
 {
 	struct radeon_fence *fences[RADEON_NUM_RINGS];
 	unsigned tries[RADEON_NUM_RINGS];
@@ -353,14 +353,11 @@ int radeon_sa_bo_new(struct radeon_device *rdev,
 		r = radeon_fence_wait_any(rdev, fences, false);
 		spin_lock(&sa_manager->wq.lock);
 		/* if we have nothing to wait for block */
-		if (r == -ENOENT && block) {
+		if (r == -ENOENT) {
 			r = wait_event_interruptible_locked(
 				sa_manager->wq, 
 				radeon_sa_event(sa_manager, size, align)
 			);
-
-		} else if (r == -ENOENT) {
-			r = -ENOMEM;
 		}
 
 	} while (!r);
@@ -402,13 +399,15 @@ void radeon_sa_bo_dump_debug_info(struct radeon_sa_manager *sa_manager,
 
 	spin_lock(&sa_manager->wq.lock);
 	list_for_each_entry(i, &sa_manager->olist, olist) {
+		uint64_t soffset = i->soffset + sa_manager->gpu_addr;
+		uint64_t eoffset = i->eoffset + sa_manager->gpu_addr;
 		if (&i->olist == sa_manager->hole) {
 			seq_printf(m, ">");
 		} else {
 			seq_printf(m, " ");
 		}
-		seq_printf(m, "[0x%08x 0x%08x] size %8d",
-			   i->soffset, i->eoffset, i->eoffset - i->soffset);
+		seq_printf(m, "[0x%010llx 0x%010llx] size %8lld",
+			   soffset, eoffset, eoffset - soffset);
 		if (i->fence) {
 			seq_printf(m, " protected by 0x%016llx on ring %d",
 				   i->fence->seq, i->fence->ring);
diff --git a/drivers/gpu/drm/radeon/radeon_semaphore.c b/drivers/gpu/drm/radeon/radeon_semaphore.c
index 2b42aa1914f..dbd6bcde92d 100644
--- a/drivers/gpu/drm/radeon/radeon_semaphore.c
+++ b/drivers/gpu/drm/radeon/radeon_semaphore.c
@@ -34,14 +34,15 @@
 int radeon_semaphore_create(struct radeon_device *rdev,
 			    struct radeon_semaphore **semaphore)
 {
+	uint32_t *cpu_addr;
 	int i, r;
 
 	*semaphore = kmalloc(sizeof(struct radeon_semaphore), GFP_KERNEL);
 	if (*semaphore == NULL) {
 		return -ENOMEM;
 	}
-	r = radeon_sa_bo_new(rdev, &rdev->ring_tmp_bo,
-			     &(*semaphore)->sa_bo, 8, 8, true);
+	r = radeon_sa_bo_new(rdev, &rdev->ring_tmp_bo, &(*semaphore)->sa_bo,
+			     8 * RADEON_NUM_SYNCS, 8);
 	if (r) {
 		kfree(*semaphore);
 		*semaphore = NULL;
@@ -49,7 +50,10 @@ int radeon_semaphore_create(struct radeon_device *rdev,
 	}
 	(*semaphore)->waiters = 0;
 	(*semaphore)->gpu_addr = radeon_sa_bo_gpu_addr((*semaphore)->sa_bo);
-	*((uint64_t*)radeon_sa_bo_cpu_addr((*semaphore)->sa_bo)) = 0;
+
+	cpu_addr = radeon_sa_bo_cpu_addr((*semaphore)->sa_bo);
+	for (i = 0; i < RADEON_NUM_SYNCS; ++i)
+		cpu_addr[i] = 0;
 
 	for (i = 0; i < RADEON_NUM_RINGS; ++i)
 		(*semaphore)->sync_to[i] = NULL;
@@ -125,6 +129,7 @@ int radeon_semaphore_sync_rings(struct radeon_device *rdev,
 				struct radeon_semaphore *semaphore,
 				int ring)
 {
+	unsigned count = 0;
 	int i, r;
 
         for (i = 0; i < RADEON_NUM_RINGS; ++i) {
@@ -140,6 +145,14 @@ int radeon_semaphore_sync_rings(struct radeon_device *rdev,
 			return -EINVAL;
 		}
 
+		if (++count > RADEON_NUM_SYNCS) {
+			/* not enough room, wait manually */
+			r = radeon_fence_wait(fence, false);
+			if (r)
+				return r;
+			continue;
+		}
+
 		/* allocate enough space for sync command */
 		r = radeon_ring_alloc(rdev, &rdev->ring[i], 16);
 		if (r) {
@@ -150,7 +163,9 @@ int radeon_semaphore_sync_rings(struct radeon_device *rdev,
 		if (!radeon_semaphore_emit_signal(rdev, i, semaphore)) {
 			/* signaling wasn't successful wait manually */
 			radeon_ring_undo(&rdev->ring[i]);
-			radeon_fence_wait_locked(fence);
+			r = radeon_fence_wait(fence, false);
+			if (r)
+				return r;
 			continue;
 		}
 
@@ -158,12 +173,16 @@ int radeon_semaphore_sync_rings(struct radeon_device *rdev,
 		if (!radeon_semaphore_emit_wait(rdev, ring, semaphore)) {
 			/* waiting wasn't successful wait manually */
 			radeon_ring_undo(&rdev->ring[i]);
-			radeon_fence_wait_locked(fence);
+			r = radeon_fence_wait(fence, false);
+			if (r)
+				return r;
 			continue;
 		}
 
 		radeon_ring_commit(rdev, &rdev->ring[i]);
 		radeon_fence_note_sync(fence, ring);
+
+		semaphore->gpu_addr += 8;
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c
index 4d20910899d..23bb64fd775 100644
--- a/drivers/gpu/drm/radeon/radeon_state.c
+++ b/drivers/gpu/drm/radeon/radeon_state.c
@@ -1810,7 +1810,7 @@ static int radeon_cp_dispatch_texture(struct drm_device * dev,
 		}
 		if (!buf) {
 			DRM_DEBUG("EAGAIN\n");
-			if (DRM_COPY_TO_USER(tex->image, image, sizeof(*image)))
+			if (copy_to_user(tex->image, image, sizeof(*image)))
 				return -EFAULT;
 			return -EAGAIN;
 		}
@@ -1823,7 +1823,7 @@ static int radeon_cp_dispatch_texture(struct drm_device * dev,
 
 #define RADEON_COPY_MT(_buf, _data, _width) \
 	do { \
-		if (DRM_COPY_FROM_USER(_buf, _data, (_width))) {\
+		if (copy_from_user(_buf, _data, (_width))) {\
 			DRM_ERROR("EFAULT on pad, %d bytes\n", (_width)); \
 			return -EFAULT; \
 		} \
@@ -2168,7 +2168,7 @@ static int radeon_cp_clear(struct drm_device *dev, void *data, struct drm_file *
 	if (sarea_priv->nbox > RADEON_NR_SAREA_CLIPRECTS)
 		sarea_priv->nbox = RADEON_NR_SAREA_CLIPRECTS;
 
-	if (DRM_COPY_FROM_USER(&depth_boxes, clear->depth_boxes,
+	if (copy_from_user(&depth_boxes, clear->depth_boxes,
 			       sarea_priv->nbox * sizeof(depth_boxes[0])))
 		return -EFAULT;
 
@@ -2436,7 +2436,7 @@ static int radeon_cp_texture(struct drm_device *dev, void *data, struct drm_file
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_FROM_USER(&image,
+	if (copy_from_user(&image,
 			       (drm_radeon_tex_image_t __user *) tex->image,
 			       sizeof(image)))
 		return -EFAULT;
@@ -2460,7 +2460,7 @@ static int radeon_cp_stipple(struct drm_device *dev, void *data, struct drm_file
 
 	LOCK_TEST_WITH_RETURN(dev, file_priv);
 
-	if (DRM_COPY_FROM_USER(&mask, stipple->mask, 32 * sizeof(u32)))
+	if (copy_from_user(&mask, stipple->mask, 32 * sizeof(u32)))
 		return -EFAULT;
 
 	RING_SPACE_TEST_WITH_RETURN(dev_priv);
@@ -2585,13 +2585,13 @@ static int radeon_cp_vertex2(struct drm_device *dev, void *data, struct drm_file
 		drm_radeon_prim_t prim;
 		drm_radeon_tcl_prim_t tclprim;
 
-		if (DRM_COPY_FROM_USER(&prim, &vertex->prim[i], sizeof(prim)))
+		if (copy_from_user(&prim, &vertex->prim[i], sizeof(prim)))
 			return -EFAULT;
 
 		if (prim.stateidx != laststate) {
 			drm_radeon_state_t state;
 
-			if (DRM_COPY_FROM_USER(&state,
+			if (copy_from_user(&state,
 					       &vertex->state[prim.stateidx],
 					       sizeof(state)))
 				return -EFAULT;
@@ -2799,7 +2799,7 @@ static int radeon_emit_packet3_cliprect(struct drm_device *dev,
 
 	do {
 		if (i < cmdbuf->nbox) {
-			if (DRM_COPY_FROM_USER(&box, &boxes[i], sizeof(box)))
+			if (copy_from_user(&box, &boxes[i], sizeof(box)))
 				return -EFAULT;
 			/* FIXME The second and subsequent times round
 			 * this loop, send a WAIT_UNTIL_3D_IDLE before
@@ -3054,7 +3054,7 @@ static int radeon_cp_getparam(struct drm_device *dev, void *data, struct drm_fil
 		if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_R600)
 			value = 0;
 		else
-			value = drm_dev_to_irq(dev);
+			value = dev->pdev->irq;
 		break;
 	case RADEON_PARAM_GART_BASE:
 		value = dev_priv->gart_vm_start;
@@ -3116,7 +3116,7 @@ static int radeon_cp_getparam(struct drm_device *dev, void *data, struct drm_fil
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_TO_USER(param->value, &value, sizeof(int))) {
+	if (copy_to_user(param->value, &value, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
@@ -3258,4 +3258,4 @@ struct drm_ioctl_desc radeon_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(RADEON_CS, r600_cs_legacy_ioctl, DRM_AUTH)
 };
 
-int radeon_max_ioctl = DRM_ARRAY_SIZE(radeon_ioctls);
+int radeon_max_ioctl = ARRAY_SIZE(radeon_ioctls);
diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c
index 12e8099a082..3a13e0d1055 100644
--- a/drivers/gpu/drm/radeon/radeon_test.c
+++ b/drivers/gpu/drm/radeon/radeon_test.c
@@ -257,20 +257,36 @@ static int radeon_test_create_and_emit_fence(struct radeon_device *rdev,
 					     struct radeon_ring *ring,
 					     struct radeon_fence **fence)
 {
+	uint32_t handle = ring->idx ^ 0xdeafbeef;
 	int r;
 
 	if (ring->idx == R600_RING_TYPE_UVD_INDEX) {
-		r = radeon_uvd_get_create_msg(rdev, ring->idx, 1, NULL);
+		r = radeon_uvd_get_create_msg(rdev, ring->idx, handle, NULL);
 		if (r) {
 			DRM_ERROR("Failed to get dummy create msg\n");
 			return r;
 		}
 
-		r = radeon_uvd_get_destroy_msg(rdev, ring->idx, 1, fence);
+		r = radeon_uvd_get_destroy_msg(rdev, ring->idx, handle, fence);
 		if (r) {
 			DRM_ERROR("Failed to get dummy destroy msg\n");
 			return r;
 		}
+
+	} else if (ring->idx == TN_RING_TYPE_VCE1_INDEX ||
+		   ring->idx == TN_RING_TYPE_VCE2_INDEX) {
+		r = radeon_vce_get_create_msg(rdev, ring->idx, handle, NULL);
+		if (r) {
+			DRM_ERROR("Failed to get dummy create msg\n");
+			return r;
+		}
+
+		r = radeon_vce_get_destroy_msg(rdev, ring->idx, handle, fence);
+		if (r) {
+			DRM_ERROR("Failed to get dummy destroy msg\n");
+			return r;
+		}
+
 	} else {
 		r = radeon_ring_lock(rdev, ring, 64);
 		if (r) {
@@ -486,6 +502,16 @@ out_cleanup:
 		printk(KERN_WARNING "Error while testing ring sync (%d).\n", r);
 }
 
+static bool radeon_test_sync_possible(struct radeon_ring *ringA,
+				      struct radeon_ring *ringB)
+{
+	if (ringA->idx == TN_RING_TYPE_VCE2_INDEX &&
+	    ringB->idx == TN_RING_TYPE_VCE1_INDEX)
+		return false;
+
+	return true;
+}
+
 void radeon_test_syncing(struct radeon_device *rdev)
 {
 	int i, j, k;
@@ -500,6 +526,9 @@ void radeon_test_syncing(struct radeon_device *rdev)
 			if (!ringB->ready)
 				continue;
 
+			if (!radeon_test_sync_possible(ringA, ringB))
+				continue;
+
 			DRM_INFO("Testing syncing between rings %d and %d...\n", i, j);
 			radeon_test_ring_sync(rdev, ringA, ringB);
 
@@ -511,6 +540,12 @@ void radeon_test_syncing(struct radeon_device *rdev)
 				if (!ringC->ready)
 					continue;
 
+				if (!radeon_test_sync_possible(ringA, ringC))
+					continue;
+
+				if (!radeon_test_sync_possible(ringB, ringC))
+					continue;
+
 				DRM_INFO("Testing syncing between rings %d, %d and %d...\n", i, j, k);
 				radeon_test_ring_sync2(rdev, ringA, ringB, ringC);
 
diff --git a/drivers/gpu/drm/radeon/radeon_trace.h b/drivers/gpu/drm/radeon/radeon_trace.h
index 0473257d407..f749f2c3bbd 100644
--- a/drivers/gpu/drm/radeon/radeon_trace.h
+++ b/drivers/gpu/drm/radeon/radeon_trace.h
@@ -106,42 +106,45 @@ TRACE_EVENT(radeon_vm_set_page,
 
 DECLARE_EVENT_CLASS(radeon_fence_request,
 
-	    TP_PROTO(struct drm_device *dev, u32 seqno),
+	    TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
 
-	    TP_ARGS(dev, seqno),
+	    TP_ARGS(dev, ring, seqno),
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
+			     __field(int, ring)
 			     __field(u32, seqno)
 			     ),
 
 	    TP_fast_assign(
 			   __entry->dev = dev->primary->index;
+			   __entry->ring = ring;
 			   __entry->seqno = seqno;
 			   ),
 
-	    TP_printk("dev=%u, seqno=%u", __entry->dev, __entry->seqno)
+	    TP_printk("dev=%u, ring=%d, seqno=%u",
+		      __entry->dev, __entry->ring, __entry->seqno)
 );
 
 DEFINE_EVENT(radeon_fence_request, radeon_fence_emit,
 
-	    TP_PROTO(struct drm_device *dev, u32 seqno),
+	    TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
 
-	    TP_ARGS(dev, seqno)
+	    TP_ARGS(dev, ring, seqno)
 );
 
 DEFINE_EVENT(radeon_fence_request, radeon_fence_wait_begin,
 
-	    TP_PROTO(struct drm_device *dev, u32 seqno),
+	    TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
 
-	    TP_ARGS(dev, seqno)
+	    TP_ARGS(dev, ring, seqno)
 );
 
 DEFINE_EVENT(radeon_fence_request, radeon_fence_wait_end,
 
-	    TP_PROTO(struct drm_device *dev, u32 seqno),
+	    TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
 
-	    TP_ARGS(dev, seqno)
+	    TP_ARGS(dev, ring, seqno)
 );
 
 DECLARE_EVENT_CLASS(radeon_semaphore_request,
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 71245d6f34a..c8a8a5144ec 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -39,12 +39,14 @@
 #include <linux/seq_file.h>
 #include <linux/slab.h>
 #include <linux/swiotlb.h>
+#include <linux/debugfs.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 
 #define DRM_FILE_PAGE_OFFSET (0x100000000ULL >> PAGE_SHIFT)
 
 static int radeon_ttm_debugfs_init(struct radeon_device *rdev);
+static void radeon_ttm_debugfs_fini(struct radeon_device *rdev);
 
 static struct radeon_device *radeon_get_rdev(struct ttm_bo_device *bdev)
 {
@@ -142,7 +144,7 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 		man->flags = TTM_MEMTYPE_FLAG_MAPPABLE | TTM_MEMTYPE_FLAG_CMA;
 #if __OS_HAS_AGP
 		if (rdev->flags & RADEON_IS_AGP) {
-			if (!(drm_core_has_AGP(rdev->ddev) && rdev->ddev->agp)) {
+			if (!rdev->ddev->agp) {
 				DRM_ERROR("AGP is not enabled for memory type %u\n",
 					  (unsigned)type);
 				return -EINVAL;
@@ -404,8 +406,14 @@ static int radeon_bo_move(struct ttm_buffer_object *bo,
 	if (r) {
 memcpy:
 		r = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
+		if (r) {
+			return r;
+		}
 	}
-	return r;
+
+	/* update statistics */
+	atomic64_add((u64)bo->num_pages << PAGE_SHIFT, &rdev->num_bytes_moved);
+	return 0;
 }
 
 static int radeon_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem)
@@ -699,7 +707,9 @@ int radeon_ttm_init(struct radeon_device *rdev)
 	/* No others user of address space so set it to 0 */
 	r = ttm_bo_device_init(&rdev->mman.bdev,
 			       rdev->mman.bo_global_ref.ref.object,
-			       &radeon_bo_driver, DRM_FILE_PAGE_OFFSET,
+			       &radeon_bo_driver,
+			       rdev->ddev->anon_inode->i_mapping,
+			       DRM_FILE_PAGE_OFFSET,
 			       rdev->need_dma32);
 	if (r) {
 		DRM_ERROR("failed initializing buffer object driver(%d).\n", r);
@@ -712,6 +722,9 @@ int radeon_ttm_init(struct radeon_device *rdev)
 		DRM_ERROR("Failed initializing VRAM heap.\n");
 		return r;
 	}
+	/* Change the size here instead of the init above so only lpfn is affected */
+	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+
 	r = radeon_bo_create(rdev, 256 * 1024, PAGE_SIZE, true,
 			     RADEON_GEM_DOMAIN_VRAM,
 			     NULL, &rdev->stollen_vga_memory);
@@ -737,7 +750,6 @@ int radeon_ttm_init(struct radeon_device *rdev)
 	}
 	DRM_INFO("radeon: %uM of GTT memory ready.\n",
 		 (unsigned)(rdev->mc.gtt_size / (1024 * 1024)));
-	rdev->mman.bdev.dev_mapping = rdev->ddev->dev_mapping;
 
 	r = radeon_ttm_debugfs_init(rdev);
 	if (r) {
@@ -753,6 +765,7 @@ void radeon_ttm_fini(struct radeon_device *rdev)
 
 	if (!rdev->mman.initialized)
 		return;
+	radeon_ttm_debugfs_fini(rdev);
 	if (rdev->stollen_vga_memory) {
 		r = radeon_bo_reserve(rdev->stollen_vga_memory, false);
 		if (r == 0) {
@@ -832,16 +845,15 @@ int radeon_mmap(struct file *filp, struct vm_area_struct *vma)
 	return 0;
 }
 
-
-#define RADEON_DEBUGFS_MEM_TYPES 2
-
 #if defined(CONFIG_DEBUG_FS)
+
 static int radeon_mm_dump_table(struct seq_file *m, void *data)
 {
 	struct drm_info_node *node = (struct drm_info_node *)m->private;
-	struct drm_mm *mm = (struct drm_mm *)node->info_ent->data;
+	unsigned ttm_pl = *(int *)node->info_ent->data;
 	struct drm_device *dev = node->minor->dev;
 	struct radeon_device *rdev = dev->dev_private;
+	struct drm_mm *mm = (struct drm_mm *)rdev->mman.bdev.man[ttm_pl].priv;
 	int ret;
 	struct ttm_bo_global *glob = rdev->mman.bdev.glob;
 
@@ -850,46 +862,169 @@ static int radeon_mm_dump_table(struct seq_file *m, void *data)
 	spin_unlock(&glob->lru_lock);
 	return ret;
 }
+
+static int ttm_pl_vram = TTM_PL_VRAM;
+static int ttm_pl_tt = TTM_PL_TT;
+
+static struct drm_info_list radeon_ttm_debugfs_list[] = {
+	{"radeon_vram_mm", radeon_mm_dump_table, 0, &ttm_pl_vram},
+	{"radeon_gtt_mm", radeon_mm_dump_table, 0, &ttm_pl_tt},
+	{"ttm_page_pool", ttm_page_alloc_debugfs, 0, NULL},
+#ifdef CONFIG_SWIOTLB
+	{"ttm_dma_page_pool", ttm_dma_page_alloc_debugfs, 0, NULL}
 #endif
+};
 
-static int radeon_ttm_debugfs_init(struct radeon_device *rdev)
+static int radeon_ttm_vram_open(struct inode *inode, struct file *filep)
 {
-#if defined(CONFIG_DEBUG_FS)
-	static struct drm_info_list radeon_mem_types_list[RADEON_DEBUGFS_MEM_TYPES+2];
-	static char radeon_mem_types_names[RADEON_DEBUGFS_MEM_TYPES+2][32];
-	unsigned i;
+	struct radeon_device *rdev = inode->i_private;
+	i_size_write(inode, rdev->mc.mc_vram_size);
+	filep->private_data = inode->i_private;
+	return 0;
+}
 
-	for (i = 0; i < RADEON_DEBUGFS_MEM_TYPES; i++) {
-		if (i == 0)
-			sprintf(radeon_mem_types_names[i], "radeon_vram_mm");
-		else
-			sprintf(radeon_mem_types_names[i], "radeon_gtt_mm");
-		radeon_mem_types_list[i].name = radeon_mem_types_names[i];
-		radeon_mem_types_list[i].show = &radeon_mm_dump_table;
-		radeon_mem_types_list[i].driver_features = 0;
-		if (i == 0)
-			radeon_mem_types_list[i].data = rdev->mman.bdev.man[TTM_PL_VRAM].priv;
-		else
-			radeon_mem_types_list[i].data = rdev->mman.bdev.man[TTM_PL_TT].priv;
+static ssize_t radeon_ttm_vram_read(struct file *f, char __user *buf,
+				    size_t size, loff_t *pos)
+{
+	struct radeon_device *rdev = f->private_data;
+	ssize_t result = 0;
+	int r;
+
+	if (size & 0x3 || *pos & 0x3)
+		return -EINVAL;
+
+	while (size) {
+		unsigned long flags;
+		uint32_t value;
+
+		if (*pos >= rdev->mc.mc_vram_size)
+			return result;
+
+		spin_lock_irqsave(&rdev->mmio_idx_lock, flags);
+		WREG32(RADEON_MM_INDEX, ((uint32_t)*pos) | 0x80000000);
+		if (rdev->family >= CHIP_CEDAR)
+			WREG32(EVERGREEN_MM_INDEX_HI, *pos >> 31);
+		value = RREG32(RADEON_MM_DATA);
+		spin_unlock_irqrestore(&rdev->mmio_idx_lock, flags);
+
+		r = put_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
 
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
 	}
-	/* Add ttm page pool to debugfs */
-	sprintf(radeon_mem_types_names[i], "ttm_page_pool");
-	radeon_mem_types_list[i].name = radeon_mem_types_names[i];
-	radeon_mem_types_list[i].show = &ttm_page_alloc_debugfs;
-	radeon_mem_types_list[i].driver_features = 0;
-	radeon_mem_types_list[i++].data = NULL;
-#ifdef CONFIG_SWIOTLB
-	if (swiotlb_nr_tbl()) {
-		sprintf(radeon_mem_types_names[i], "ttm_dma_page_pool");
-		radeon_mem_types_list[i].name = radeon_mem_types_names[i];
-		radeon_mem_types_list[i].show = &ttm_dma_page_alloc_debugfs;
-		radeon_mem_types_list[i].driver_features = 0;
-		radeon_mem_types_list[i++].data = NULL;
+
+	return result;
+}
+
+static const struct file_operations radeon_ttm_vram_fops = {
+	.owner = THIS_MODULE,
+	.open = radeon_ttm_vram_open,
+	.read = radeon_ttm_vram_read,
+	.llseek = default_llseek
+};
+
+static int radeon_ttm_gtt_open(struct inode *inode, struct file *filep)
+{
+	struct radeon_device *rdev = inode->i_private;
+	i_size_write(inode, rdev->mc.gtt_size);
+	filep->private_data = inode->i_private;
+	return 0;
+}
+
+static ssize_t radeon_ttm_gtt_read(struct file *f, char __user *buf,
+				   size_t size, loff_t *pos)
+{
+	struct radeon_device *rdev = f->private_data;
+	ssize_t result = 0;
+	int r;
+
+	while (size) {
+		loff_t p = *pos / PAGE_SIZE;
+		unsigned off = *pos & ~PAGE_MASK;
+		size_t cur_size = min_t(size_t, size, PAGE_SIZE - off);
+		struct page *page;
+		void *ptr;
+
+		if (p >= rdev->gart.num_cpu_pages)
+			return result;
+
+		page = rdev->gart.pages[p];
+		if (page) {
+			ptr = kmap(page);
+			ptr += off;
+
+			r = copy_to_user(buf, ptr, cur_size);
+			kunmap(rdev->gart.pages[p]);
+		} else
+			r = clear_user(buf, cur_size);
+
+		if (r)
+			return -EFAULT;
+
+		result += cur_size;
+		buf += cur_size;
+		*pos += cur_size;
+		size -= cur_size;
 	}
+
+	return result;
+}
+
+static const struct file_operations radeon_ttm_gtt_fops = {
+	.owner = THIS_MODULE,
+	.open = radeon_ttm_gtt_open,
+	.read = radeon_ttm_gtt_read,
+	.llseek = default_llseek
+};
+
 #endif
-	return radeon_debugfs_add_files(rdev, radeon_mem_types_list, i);
 
+static int radeon_ttm_debugfs_init(struct radeon_device *rdev)
+{
+#if defined(CONFIG_DEBUG_FS)
+	unsigned count;
+
+	struct drm_minor *minor = rdev->ddev->primary;
+	struct dentry *ent, *root = minor->debugfs_root;
+
+	ent = debugfs_create_file("radeon_vram", S_IFREG | S_IRUGO, root,
+				  rdev, &radeon_ttm_vram_fops);
+	if (IS_ERR(ent))
+		return PTR_ERR(ent);
+	rdev->mman.vram = ent;
+
+	ent = debugfs_create_file("radeon_gtt", S_IFREG | S_IRUGO, root,
+				  rdev, &radeon_ttm_gtt_fops);
+	if (IS_ERR(ent))
+		return PTR_ERR(ent);
+	rdev->mman.gtt = ent;
+
+	count = ARRAY_SIZE(radeon_ttm_debugfs_list);
+
+#ifdef CONFIG_SWIOTLB
+	if (!swiotlb_nr_tbl())
+		--count;
 #endif
+
+	return radeon_debugfs_add_files(rdev, radeon_ttm_debugfs_list, count);
+#else
+
 	return 0;
+#endif
+}
+
+static void radeon_ttm_debugfs_fini(struct radeon_device *rdev)
+{
+#if defined(CONFIG_DEBUG_FS)
+
+	debugfs_remove(rdev->mman.vram);
+	rdev->mman.vram = NULL;
+
+	debugfs_remove(rdev->mman.gtt);
+	rdev->mman.gtt = NULL;
+#endif
 }
diff --git a/drivers/gpu/drm/radeon/radeon_ucode.h b/drivers/gpu/drm/radeon/radeon_ucode.h
index a77cd274dfc..4e7c3269b18 100644
--- a/drivers/gpu/drm/radeon/radeon_ucode.h
+++ b/drivers/gpu/drm/radeon/radeon_ucode.h
@@ -52,14 +52,20 @@
 #define BONAIRE_RLC_UCODE_SIZE       2048
 #define KB_RLC_UCODE_SIZE            2560
 #define KV_RLC_UCODE_SIZE            2560
+#define ML_RLC_UCODE_SIZE            2560
 
 /* MC */
 #define BTC_MC_UCODE_SIZE            6024
 #define CAYMAN_MC_UCODE_SIZE         6037
 #define SI_MC_UCODE_SIZE             7769
+#define TAHITI_MC_UCODE_SIZE         7808
+#define PITCAIRN_MC_UCODE_SIZE       7775
+#define VERDE_MC_UCODE_SIZE          7875
 #define OLAND_MC_UCODE_SIZE          7863
-#define CIK_MC_UCODE_SIZE            7866
+#define BONAIRE_MC_UCODE_SIZE        7866
+#define BONAIRE_MC2_UCODE_SIZE       7948
 #define HAWAII_MC_UCODE_SIZE         7933
+#define HAWAII_MC2_UCODE_SIZE        8091
 
 /* SDMA */
 #define CIK_SDMA_UCODE_SIZE          1050
diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c
index b9c0529b4a2..a4ad270e826 100644
--- a/drivers/gpu/drm/radeon/radeon_uvd.c
+++ b/drivers/gpu/drm/radeon/radeon_uvd.c
@@ -91,6 +91,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	case CHIP_VERDE:
 	case CHIP_PITCAIRN:
 	case CHIP_ARUBA:
+	case CHIP_OLAND:
 		fw_name = FIRMWARE_TAHITI;
 		break;
 
@@ -98,6 +99,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	case CHIP_KABINI:
 	case CHIP_KAVERI:
 	case CHIP_HAWAII:
+	case CHIP_MULLINS:
 		fw_name = FIRMWARE_BONAIRE;
 		break;
 
@@ -170,6 +172,8 @@ void radeon_uvd_fini(struct radeon_device *rdev)
 
 	radeon_bo_unref(&rdev->uvd.vcpu_bo);
 
+	radeon_ring_fini(rdev, &rdev->ring[R600_RING_TYPE_UVD_INDEX]);
+
 	release_firmware(rdev->uvd_fw);
 }
 
@@ -452,7 +456,7 @@ static int radeon_uvd_cs_reloc(struct radeon_cs_parser *p,
 	}
 
 	reloc = p->relocs_ptr[(idx / 4)];
-	start = reloc->lobj.gpu_offset;
+	start = reloc->gpu_offset;
 	end = start + radeon_bo_size(reloc->robj);
 	start += offset;
 
@@ -462,6 +466,10 @@ static int radeon_uvd_cs_reloc(struct radeon_cs_parser *p,
 	cmd = radeon_get_ib_value(p, p->idx) >> 1;
 
 	if (cmd < 0x4) {
+		if (end <= start) {
+			DRM_ERROR("invalid reloc offset %X!\n", offset);
+			return -EINVAL;
+		}
 		if ((end - start) < buf_sizes[cmd]) {
 			DRM_ERROR("buffer (%d) to small (%d / %d)!\n", cmd,
 				  (unsigned)(end - start), buf_sizes[cmd]);
@@ -778,6 +786,8 @@ static void radeon_uvd_idle_work_handler(struct work_struct *work)
 
 	if (radeon_fence_count_emitted(rdev, R600_RING_TYPE_UVD_INDEX) == 0) {
 		if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled) {
+			radeon_uvd_count_handles(rdev, &rdev->pm.dpm.sd,
+						 &rdev->pm.dpm.hd);
 			radeon_dpm_enable_uvd(rdev, false);
 		} else {
 			radeon_set_uvd_clocks(rdev, 0, 0);
diff --git a/drivers/gpu/drm/radeon/radeon_vce.c b/drivers/gpu/drm/radeon/radeon_vce.c
new file mode 100644
index 00000000000..aa21c31a846
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_vce.c
@@ -0,0 +1,771 @@
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * Authors: Christian König <christian.koenig@amd.com>
+ */
+
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <drm/drmP.h>
+#include <drm/drm.h>
+
+#include "radeon.h"
+#include "radeon_asic.h"
+#include "sid.h"
+
+/* 1 second timeout */
+#define VCE_IDLE_TIMEOUT_MS	1000
+
+/* Firmware Names */
+#define FIRMWARE_BONAIRE	"radeon/BONAIRE_vce.bin"
+
+MODULE_FIRMWARE(FIRMWARE_BONAIRE);
+
+static void radeon_vce_idle_work_handler(struct work_struct *work);
+
+/**
+ * radeon_vce_init - allocate memory, load vce firmware
+ *
+ * @rdev: radeon_device pointer
+ *
+ * First step to get VCE online, allocate memory and load the firmware
+ */
+int radeon_vce_init(struct radeon_device *rdev)
+{
+	static const char *fw_version = "[ATI LIB=VCEFW,";
+	static const char *fb_version = "[ATI LIB=VCEFWSTATS,";
+	unsigned long size;
+	const char *fw_name, *c;
+	uint8_t start, mid, end;
+	int i, r;
+
+	INIT_DELAYED_WORK(&rdev->vce.idle_work, radeon_vce_idle_work_handler);
+
+	switch (rdev->family) {
+	case CHIP_BONAIRE:
+	case CHIP_KAVERI:
+	case CHIP_KABINI:
+	case CHIP_HAWAII:
+	case CHIP_MULLINS:
+		fw_name = FIRMWARE_BONAIRE;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	r = request_firmware(&rdev->vce_fw, fw_name, rdev->dev);
+	if (r) {
+		dev_err(rdev->dev, "radeon_vce: Can't load firmware \"%s\"\n",
+			fw_name);
+		return r;
+	}
+
+	/* search for firmware version */
+
+	size = rdev->vce_fw->size - strlen(fw_version) - 9;
+	c = rdev->vce_fw->data;
+	for (;size > 0; --size, ++c)
+		if (strncmp(c, fw_version, strlen(fw_version)) == 0)
+			break;
+
+	if (size == 0)
+		return -EINVAL;
+
+	c += strlen(fw_version);
+	if (sscanf(c, "%2hhd.%2hhd.%2hhd]", &start, &mid, &end) != 3)
+		return -EINVAL;
+
+	/* search for feedback version */
+
+	size = rdev->vce_fw->size - strlen(fb_version) - 3;
+	c = rdev->vce_fw->data;
+	for (;size > 0; --size, ++c)
+		if (strncmp(c, fb_version, strlen(fb_version)) == 0)
+			break;
+
+	if (size == 0)
+		return -EINVAL;
+
+	c += strlen(fb_version);
+	if (sscanf(c, "%2u]", &rdev->vce.fb_version) != 1)
+		return -EINVAL;
+
+	DRM_INFO("Found VCE firmware/feedback version %hhd.%hhd.%hhd / %d!\n",
+		 start, mid, end, rdev->vce.fb_version);
+
+	rdev->vce.fw_version = (start << 24) | (mid << 16) | (end << 8);
+
+	/* we can only work with this fw version for now */
+	if (rdev->vce.fw_version != ((40 << 24) | (2 << 16) | (2 << 8)))
+		return -EINVAL;
+
+	/* allocate firmware, stack and heap BO */
+
+	size = RADEON_GPU_PAGE_ALIGN(rdev->vce_fw->size) +
+	       RADEON_VCE_STACK_SIZE + RADEON_VCE_HEAP_SIZE;
+	r = radeon_bo_create(rdev, size, PAGE_SIZE, true,
+			     RADEON_GEM_DOMAIN_VRAM, NULL, &rdev->vce.vcpu_bo);
+	if (r) {
+		dev_err(rdev->dev, "(%d) failed to allocate VCE bo\n", r);
+		return r;
+	}
+
+	r = radeon_bo_reserve(rdev->vce.vcpu_bo, false);
+	if (r) {
+		radeon_bo_unref(&rdev->vce.vcpu_bo);
+		dev_err(rdev->dev, "(%d) failed to reserve VCE bo\n", r);
+		return r;
+	}
+
+	r = radeon_bo_pin(rdev->vce.vcpu_bo, RADEON_GEM_DOMAIN_VRAM,
+			  &rdev->vce.gpu_addr);
+	radeon_bo_unreserve(rdev->vce.vcpu_bo);
+	if (r) {
+		radeon_bo_unref(&rdev->vce.vcpu_bo);
+		dev_err(rdev->dev, "(%d) VCE bo pin failed\n", r);
+		return r;
+	}
+
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		atomic_set(&rdev->vce.handles[i], 0);
+		rdev->vce.filp[i] = NULL;
+        }
+
+	return 0;
+}
+
+/**
+ * radeon_vce_fini - free memory
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Last step on VCE teardown, free firmware memory
+ */
+void radeon_vce_fini(struct radeon_device *rdev)
+{
+	if (rdev->vce.vcpu_bo == NULL)
+		return;
+
+	radeon_bo_unref(&rdev->vce.vcpu_bo);
+
+	release_firmware(rdev->vce_fw);
+}
+
+/**
+ * radeon_vce_suspend - unpin VCE fw memory
+ *
+ * @rdev: radeon_device pointer
+ *
+ */
+int radeon_vce_suspend(struct radeon_device *rdev)
+{
+	int i;
+
+	if (rdev->vce.vcpu_bo == NULL)
+		return 0;
+
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i)
+		if (atomic_read(&rdev->vce.handles[i]))
+			break;
+
+	if (i == RADEON_MAX_VCE_HANDLES)
+		return 0;
+
+	/* TODO: suspending running encoding sessions isn't supported */
+	return -EINVAL;
+}
+
+/**
+ * radeon_vce_resume - pin VCE fw memory
+ *
+ * @rdev: radeon_device pointer
+ *
+ */
+int radeon_vce_resume(struct radeon_device *rdev)
+{
+	void *cpu_addr;
+	int r;
+
+	if (rdev->vce.vcpu_bo == NULL)
+		return -EINVAL;
+
+	r = radeon_bo_reserve(rdev->vce.vcpu_bo, false);
+	if (r) {
+		dev_err(rdev->dev, "(%d) failed to reserve VCE bo\n", r);
+		return r;
+	}
+
+	r = radeon_bo_kmap(rdev->vce.vcpu_bo, &cpu_addr);
+	if (r) {
+		radeon_bo_unreserve(rdev->vce.vcpu_bo);
+		dev_err(rdev->dev, "(%d) VCE map failed\n", r);
+		return r;
+	}
+
+	memcpy(cpu_addr, rdev->vce_fw->data, rdev->vce_fw->size);
+
+	radeon_bo_kunmap(rdev->vce.vcpu_bo);
+
+	radeon_bo_unreserve(rdev->vce.vcpu_bo);
+
+	return 0;
+}
+
+/**
+ * radeon_vce_idle_work_handler - power off VCE
+ *
+ * @work: pointer to work structure
+ *
+ * power of VCE when it's not used any more
+ */
+static void radeon_vce_idle_work_handler(struct work_struct *work)
+{
+	struct radeon_device *rdev =
+		container_of(work, struct radeon_device, vce.idle_work.work);
+
+	if ((radeon_fence_count_emitted(rdev, TN_RING_TYPE_VCE1_INDEX) == 0) &&
+	    (radeon_fence_count_emitted(rdev, TN_RING_TYPE_VCE2_INDEX) == 0)) {
+		if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled) {
+			radeon_dpm_enable_vce(rdev, false);
+		} else {
+			radeon_set_vce_clocks(rdev, 0, 0);
+		}
+	} else {
+		schedule_delayed_work(&rdev->vce.idle_work,
+				      msecs_to_jiffies(VCE_IDLE_TIMEOUT_MS));
+	}
+}
+
+/**
+ * radeon_vce_note_usage - power up VCE
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Make sure VCE is powerd up when we want to use it
+ */
+void radeon_vce_note_usage(struct radeon_device *rdev)
+{
+	bool streams_changed = false;
+	bool set_clocks = !cancel_delayed_work_sync(&rdev->vce.idle_work);
+	set_clocks &= schedule_delayed_work(&rdev->vce.idle_work,
+					    msecs_to_jiffies(VCE_IDLE_TIMEOUT_MS));
+
+	if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled) {
+		/* XXX figure out if the streams changed */
+		streams_changed = false;
+	}
+
+	if (set_clocks || streams_changed) {
+		if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled) {
+			radeon_dpm_enable_vce(rdev, true);
+		} else {
+			radeon_set_vce_clocks(rdev, 53300, 40000);
+		}
+	}
+}
+
+/**
+ * radeon_vce_free_handles - free still open VCE handles
+ *
+ * @rdev: radeon_device pointer
+ * @filp: drm file pointer
+ *
+ * Close all VCE handles still open by this file pointer
+ */
+void radeon_vce_free_handles(struct radeon_device *rdev, struct drm_file *filp)
+{
+	int i, r;
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		uint32_t handle = atomic_read(&rdev->vce.handles[i]);
+		if (!handle || rdev->vce.filp[i] != filp)
+			continue;
+
+		radeon_vce_note_usage(rdev);
+
+		r = radeon_vce_get_destroy_msg(rdev, TN_RING_TYPE_VCE1_INDEX,
+					       handle, NULL);
+		if (r)
+			DRM_ERROR("Error destroying VCE handle (%d)!\n", r);
+
+		rdev->vce.filp[i] = NULL;
+		atomic_set(&rdev->vce.handles[i], 0);
+	}
+}
+
+/**
+ * radeon_vce_get_create_msg - generate a VCE create msg
+ *
+ * @rdev: radeon_device pointer
+ * @ring: ring we should submit the msg to
+ * @handle: VCE session handle to use
+ * @fence: optional fence to return
+ *
+ * Open up a stream for HW test
+ */
+int radeon_vce_get_create_msg(struct radeon_device *rdev, int ring,
+			      uint32_t handle, struct radeon_fence **fence)
+{
+	const unsigned ib_size_dw = 1024;
+	struct radeon_ib ib;
+	uint64_t dummy;
+	int i, r;
+
+	r = radeon_ib_get(rdev, ring, &ib, NULL, ib_size_dw * 4);
+	if (r) {
+		DRM_ERROR("radeon: failed to get ib (%d).\n", r);
+		return r;
+	}
+
+	dummy = ib.gpu_addr + 1024;
+
+	/* stitch together an VCE create msg */
+	ib.length_dw = 0;
+	ib.ptr[ib.length_dw++] = 0x0000000c; /* len */
+	ib.ptr[ib.length_dw++] = 0x00000001; /* session cmd */
+	ib.ptr[ib.length_dw++] = handle;
+
+	ib.ptr[ib.length_dw++] = 0x00000030; /* len */
+	ib.ptr[ib.length_dw++] = 0x01000001; /* create cmd */
+	ib.ptr[ib.length_dw++] = 0x00000000;
+	ib.ptr[ib.length_dw++] = 0x00000042;
+	ib.ptr[ib.length_dw++] = 0x0000000a;
+	ib.ptr[ib.length_dw++] = 0x00000001;
+	ib.ptr[ib.length_dw++] = 0x00000080;
+	ib.ptr[ib.length_dw++] = 0x00000060;
+	ib.ptr[ib.length_dw++] = 0x00000100;
+	ib.ptr[ib.length_dw++] = 0x00000100;
+	ib.ptr[ib.length_dw++] = 0x0000000c;
+	ib.ptr[ib.length_dw++] = 0x00000000;
+
+	ib.ptr[ib.length_dw++] = 0x00000014; /* len */
+	ib.ptr[ib.length_dw++] = 0x05000005; /* feedback buffer */
+	ib.ptr[ib.length_dw++] = upper_32_bits(dummy);
+	ib.ptr[ib.length_dw++] = dummy;
+	ib.ptr[ib.length_dw++] = 0x00000001;
+
+	for (i = ib.length_dw; i < ib_size_dw; ++i)
+		ib.ptr[i] = 0x0;
+
+	r = radeon_ib_schedule(rdev, &ib, NULL);
+	if (r) {
+	        DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
+	}
+
+	if (fence)
+		*fence = radeon_fence_ref(ib.fence);
+
+	radeon_ib_free(rdev, &ib);
+
+	return r;
+}
+
+/**
+ * radeon_vce_get_destroy_msg - generate a VCE destroy msg
+ *
+ * @rdev: radeon_device pointer
+ * @ring: ring we should submit the msg to
+ * @handle: VCE session handle to use
+ * @fence: optional fence to return
+ *
+ * Close up a stream for HW test or if userspace failed to do so
+ */
+int radeon_vce_get_destroy_msg(struct radeon_device *rdev, int ring,
+			       uint32_t handle, struct radeon_fence **fence)
+{
+	const unsigned ib_size_dw = 1024;
+	struct radeon_ib ib;
+	uint64_t dummy;
+	int i, r;
+
+	r = radeon_ib_get(rdev, ring, &ib, NULL, ib_size_dw * 4);
+	if (r) {
+		DRM_ERROR("radeon: failed to get ib (%d).\n", r);
+		return r;
+	}
+
+	dummy = ib.gpu_addr + 1024;
+
+	/* stitch together an VCE destroy msg */
+	ib.length_dw = 0;
+	ib.ptr[ib.length_dw++] = 0x0000000c; /* len */
+	ib.ptr[ib.length_dw++] = 0x00000001; /* session cmd */
+	ib.ptr[ib.length_dw++] = handle;
+
+	ib.ptr[ib.length_dw++] = 0x00000014; /* len */
+	ib.ptr[ib.length_dw++] = 0x05000005; /* feedback buffer */
+	ib.ptr[ib.length_dw++] = upper_32_bits(dummy);
+	ib.ptr[ib.length_dw++] = dummy;
+	ib.ptr[ib.length_dw++] = 0x00000001;
+
+	ib.ptr[ib.length_dw++] = 0x00000008; /* len */
+	ib.ptr[ib.length_dw++] = 0x02000001; /* destroy cmd */
+
+	for (i = ib.length_dw; i < ib_size_dw; ++i)
+		ib.ptr[i] = 0x0;
+
+	r = radeon_ib_schedule(rdev, &ib, NULL);
+	if (r) {
+	        DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
+	}
+
+	if (fence)
+		*fence = radeon_fence_ref(ib.fence);
+
+	radeon_ib_free(rdev, &ib);
+
+	return r;
+}
+
+/**
+ * radeon_vce_cs_reloc - command submission relocation
+ *
+ * @p: parser context
+ * @lo: address of lower dword
+ * @hi: address of higher dword
+ * @size: size of checker for relocation buffer
+ *
+ * Patch relocation inside command stream with real buffer address
+ */
+int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi,
+			unsigned size)
+{
+	struct radeon_cs_chunk *relocs_chunk;
+	struct radeon_cs_reloc *reloc;
+	uint64_t start, end, offset;
+	unsigned idx;
+
+	relocs_chunk = &p->chunks[p->chunk_relocs_idx];
+	offset = radeon_get_ib_value(p, lo);
+	idx = radeon_get_ib_value(p, hi);
+
+	if (idx >= relocs_chunk->length_dw) {
+		DRM_ERROR("Relocs at %d after relocations chunk end %d !\n",
+			  idx, relocs_chunk->length_dw);
+		return -EINVAL;
+	}
+
+	reloc = p->relocs_ptr[(idx / 4)];
+	start = reloc->gpu_offset;
+	end = start + radeon_bo_size(reloc->robj);
+	start += offset;
+
+	p->ib.ptr[lo] = start & 0xFFFFFFFF;
+	p->ib.ptr[hi] = start >> 32;
+
+	if (end <= start) {
+		DRM_ERROR("invalid reloc offset %llX!\n", offset);
+		return -EINVAL;
+	}
+	if ((end - start) < size) {
+		DRM_ERROR("buffer to small (%d / %d)!\n",
+			(unsigned)(end - start), size);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+/**
+ * radeon_vce_validate_handle - validate stream handle
+ *
+ * @p: parser context
+ * @handle: handle to validate
+ *
+ * Validates the handle and return the found session index or -EINVAL
+ * we we don't have another free session index.
+ */
+int radeon_vce_validate_handle(struct radeon_cs_parser *p, uint32_t handle)
+{
+	unsigned i;
+
+	/* validate the handle */
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		if (atomic_read(&p->rdev->vce.handles[i]) == handle)
+			return i;
+	}
+
+	/* handle not found try to alloc a new one */
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		if (!atomic_cmpxchg(&p->rdev->vce.handles[i], 0, handle)) {
+			p->rdev->vce.filp[i] = p->filp;
+			p->rdev->vce.img_size[i] = 0;
+			return i;
+		}
+	}
+
+	DRM_ERROR("No more free VCE handles!\n");
+	return -EINVAL;
+}
+
+/**
+ * radeon_vce_cs_parse - parse and validate the command stream
+ *
+ * @p: parser context
+ *
+ */
+int radeon_vce_cs_parse(struct radeon_cs_parser *p)
+{
+	int session_idx = -1;
+	bool destroyed = false;
+	uint32_t tmp, handle = 0;
+	uint32_t *size = &tmp;
+	int i, r;
+
+	while (p->idx < p->chunks[p->chunk_ib_idx].length_dw) {
+		uint32_t len = radeon_get_ib_value(p, p->idx);
+		uint32_t cmd = radeon_get_ib_value(p, p->idx + 1);
+
+		if ((len < 8) || (len & 3)) {
+			DRM_ERROR("invalid VCE command length (%d)!\n", len);
+                	return -EINVAL;
+		}
+
+		if (destroyed) {
+			DRM_ERROR("No other command allowed after destroy!\n");
+			return -EINVAL;
+		}
+
+		switch (cmd) {
+		case 0x00000001: // session
+			handle = radeon_get_ib_value(p, p->idx + 2);
+			session_idx = radeon_vce_validate_handle(p, handle);
+			if (session_idx < 0)
+				return session_idx;
+			size = &p->rdev->vce.img_size[session_idx];
+			break;
+
+		case 0x00000002: // task info
+			break;
+
+		case 0x01000001: // create
+			*size = radeon_get_ib_value(p, p->idx + 8) *
+				radeon_get_ib_value(p, p->idx + 10) *
+				8 * 3 / 2;
+			break;
+
+		case 0x04000001: // config extension
+		case 0x04000002: // pic control
+		case 0x04000005: // rate control
+		case 0x04000007: // motion estimation
+		case 0x04000008: // rdo
+			break;
+
+		case 0x03000001: // encode
+			r = radeon_vce_cs_reloc(p, p->idx + 10, p->idx + 9,
+						*size);
+			if (r)
+				return r;
+
+			r = radeon_vce_cs_reloc(p, p->idx + 12, p->idx + 11,
+						*size / 3);
+			if (r)
+				return r;
+			break;
+
+		case 0x02000001: // destroy
+			destroyed = true;
+			break;
+
+		case 0x05000001: // context buffer
+			r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2,
+						*size * 2);
+			if (r)
+				return r;
+			break;
+
+		case 0x05000004: // video bitstream buffer
+			tmp = radeon_get_ib_value(p, p->idx + 4);
+			r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2,
+						tmp);
+			if (r)
+				return r;
+			break;
+
+		case 0x05000005: // feedback buffer
+			r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2,
+						4096);
+			if (r)
+				return r;
+			break;
+
+		default:
+			DRM_ERROR("invalid VCE command (0x%x)!\n", cmd);
+			return -EINVAL;
+		}
+
+		if (session_idx == -1) {
+			DRM_ERROR("no session command at start of IB\n");
+			return -EINVAL;
+		}
+
+		p->idx += len / 4;
+	}
+
+	if (destroyed) {
+		/* IB contains a destroy msg, free the handle */
+		for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i)
+			atomic_cmpxchg(&p->rdev->vce.handles[i], handle, 0);
+	}
+
+	return 0;
+}
+
+/**
+ * radeon_vce_semaphore_emit - emit a semaphore command
+ *
+ * @rdev: radeon_device pointer
+ * @ring: engine to use
+ * @semaphore: address of semaphore
+ * @emit_wait: true=emit wait, false=emit signal
+ *
+ */
+bool radeon_vce_semaphore_emit(struct radeon_device *rdev,
+			       struct radeon_ring *ring,
+			       struct radeon_semaphore *semaphore,
+			       bool emit_wait)
+{
+	uint64_t addr = semaphore->gpu_addr;
+
+	radeon_ring_write(ring, VCE_CMD_SEMAPHORE);
+	radeon_ring_write(ring, (addr >> 3) & 0x000FFFFF);
+	radeon_ring_write(ring, (addr >> 23) & 0x000FFFFF);
+	radeon_ring_write(ring, 0x01003000 | (emit_wait ? 1 : 0));
+	if (!emit_wait)
+		radeon_ring_write(ring, VCE_CMD_END);
+
+	return true;
+}
+
+/**
+ * radeon_vce_ib_execute - execute indirect buffer
+ *
+ * @rdev: radeon_device pointer
+ * @ib: the IB to execute
+ *
+ */
+void radeon_vce_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
+{
+	struct radeon_ring *ring = &rdev->ring[ib->ring];
+	radeon_ring_write(ring, VCE_CMD_IB);
+	radeon_ring_write(ring, ib->gpu_addr);
+	radeon_ring_write(ring, upper_32_bits(ib->gpu_addr));
+	radeon_ring_write(ring, ib->length_dw);
+}
+
+/**
+ * radeon_vce_fence_emit - add a fence command to the ring
+ *
+ * @rdev: radeon_device pointer
+ * @fence: the fence
+ *
+ */
+void radeon_vce_fence_emit(struct radeon_device *rdev,
+			   struct radeon_fence *fence)
+{
+	struct radeon_ring *ring = &rdev->ring[fence->ring];
+	uint64_t addr = rdev->fence_drv[fence->ring].gpu_addr;
+
+	radeon_ring_write(ring, VCE_CMD_FENCE);
+	radeon_ring_write(ring, addr);
+	radeon_ring_write(ring, upper_32_bits(addr));
+	radeon_ring_write(ring, fence->seq);
+	radeon_ring_write(ring, VCE_CMD_TRAP);
+	radeon_ring_write(ring, VCE_CMD_END);
+}
+
+/**
+ * radeon_vce_ring_test - test if VCE ring is working
+ *
+ * @rdev: radeon_device pointer
+ * @ring: the engine to test on
+ *
+ */
+int radeon_vce_ring_test(struct radeon_device *rdev, struct radeon_ring *ring)
+{
+	uint32_t rptr = vce_v1_0_get_rptr(rdev, ring);
+	unsigned i;
+	int r;
+
+	r = radeon_ring_lock(rdev, ring, 16);
+	if (r) {
+		DRM_ERROR("radeon: vce failed to lock ring %d (%d).\n",
+			  ring->idx, r);
+		return r;
+	}
+	radeon_ring_write(ring, VCE_CMD_END);
+	radeon_ring_unlock_commit(rdev, ring);
+
+	for (i = 0; i < rdev->usec_timeout; i++) {
+	        if (vce_v1_0_get_rptr(rdev, ring) != rptr)
+	                break;
+	        DRM_UDELAY(1);
+	}
+
+	if (i < rdev->usec_timeout) {
+	        DRM_INFO("ring test on %d succeeded in %d usecs\n",
+	                 ring->idx, i);
+	} else {
+	        DRM_ERROR("radeon: ring %d test failed\n",
+	                  ring->idx);
+	        r = -ETIMEDOUT;
+	}
+
+	return r;
+}
+
+/**
+ * radeon_vce_ib_test - test if VCE IBs are working
+ *
+ * @rdev: radeon_device pointer
+ * @ring: the engine to test on
+ *
+ */
+int radeon_vce_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
+{
+	struct radeon_fence *fence = NULL;
+	int r;
+
+	r = radeon_vce_get_create_msg(rdev, ring->idx, 1, NULL);
+	if (r) {
+		DRM_ERROR("radeon: failed to get create msg (%d).\n", r);
+		goto error;
+	}
+
+	r = radeon_vce_get_destroy_msg(rdev, ring->idx, 1, &fence);
+	if (r) {
+		DRM_ERROR("radeon: failed to get destroy ib (%d).\n", r);
+		goto error;
+	}
+
+	r = radeon_fence_wait(fence, false);
+	if (r) {
+		DRM_ERROR("radeon: fence wait failed (%d).\n", r);
+	} else {
+	        DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
+	}
+error:
+	radeon_fence_unref(&fence);
+	return r;
+}
diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c
new file mode 100644
index 00000000000..725d3669014
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_vm.c
@@ -0,0 +1,1089 @@
+/*
+ * Copyright 2008 Advanced Micro Devices, Inc.
+ * Copyright 2008 Red Hat Inc.
+ * Copyright 2009 Jerome Glisse.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Dave Airlie
+ *          Alex Deucher
+ *          Jerome Glisse
+ */
+#include <drm/drmP.h>
+#include <drm/radeon_drm.h>
+#include "radeon.h"
+#include "radeon_trace.h"
+
+/*
+ * GPUVM
+ * GPUVM is similar to the legacy gart on older asics, however
+ * rather than there being a single global gart table
+ * for the entire GPU, there are multiple VM page tables active
+ * at any given time.  The VM page tables can contain a mix
+ * vram pages and system memory pages and system memory pages
+ * can be mapped as snooped (cached system pages) or unsnooped
+ * (uncached system pages).
+ * Each VM has an ID associated with it and there is a page table
+ * associated with each VMID.  When execting a command buffer,
+ * the kernel tells the the ring what VMID to use for that command
+ * buffer.  VMIDs are allocated dynamically as commands are submitted.
+ * The userspace drivers maintain their own address space and the kernel
+ * sets up their pages tables accordingly when they submit their
+ * command buffers and a VMID is assigned.
+ * Cayman/Trinity support up to 8 active VMs at any given time;
+ * SI supports 16.
+ */
+
+/**
+ * radeon_vm_num_pde - return the number of page directory entries
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Calculate the number of page directory entries (cayman+).
+ */
+static unsigned radeon_vm_num_pdes(struct radeon_device *rdev)
+{
+	return rdev->vm_manager.max_pfn >> radeon_vm_block_size;
+}
+
+/**
+ * radeon_vm_directory_size - returns the size of the page directory in bytes
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Calculate the size of the page directory in bytes (cayman+).
+ */
+static unsigned radeon_vm_directory_size(struct radeon_device *rdev)
+{
+	return RADEON_GPU_PAGE_ALIGN(radeon_vm_num_pdes(rdev) * 8);
+}
+
+/**
+ * radeon_vm_manager_init - init the vm manager
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Init the vm manager (cayman+).
+ * Returns 0 for success, error for failure.
+ */
+int radeon_vm_manager_init(struct radeon_device *rdev)
+{
+	int r;
+
+	if (!rdev->vm_manager.enabled) {
+		r = radeon_asic_vm_init(rdev);
+		if (r)
+			return r;
+
+		rdev->vm_manager.enabled = true;
+	}
+	return 0;
+}
+
+/**
+ * radeon_vm_manager_fini - tear down the vm manager
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Tear down the VM manager (cayman+).
+ */
+void radeon_vm_manager_fini(struct radeon_device *rdev)
+{
+	int i;
+
+	if (!rdev->vm_manager.enabled)
+		return;
+
+	for (i = 0; i < RADEON_NUM_VM; ++i)
+		radeon_fence_unref(&rdev->vm_manager.active[i]);
+	radeon_asic_vm_fini(rdev);
+	rdev->vm_manager.enabled = false;
+}
+
+/**
+ * radeon_vm_get_bos - add the vm BOs to a validation list
+ *
+ * @vm: vm providing the BOs
+ * @head: head of validation list
+ *
+ * Add the page directory to the list of BOs to
+ * validate for command submission (cayman+).
+ */
+struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev,
+					  struct radeon_vm *vm,
+					  struct list_head *head)
+{
+	struct radeon_cs_reloc *list;
+	unsigned i, idx;
+
+	list = kmalloc_array(vm->max_pde_used + 2,
+			     sizeof(struct radeon_cs_reloc), GFP_KERNEL);
+	if (!list)
+		return NULL;
+
+	/* add the vm page table to the list */
+	list[0].gobj = NULL;
+	list[0].robj = vm->page_directory;
+	list[0].prefered_domains = RADEON_GEM_DOMAIN_VRAM;
+	list[0].allowed_domains = RADEON_GEM_DOMAIN_VRAM;
+	list[0].tv.bo = &vm->page_directory->tbo;
+	list[0].tiling_flags = 0;
+	list[0].handle = 0;
+	list_add(&list[0].tv.head, head);
+
+	for (i = 0, idx = 1; i <= vm->max_pde_used; i++) {
+		if (!vm->page_tables[i].bo)
+			continue;
+
+		list[idx].gobj = NULL;
+		list[idx].robj = vm->page_tables[i].bo;
+		list[idx].prefered_domains = RADEON_GEM_DOMAIN_VRAM;
+		list[idx].allowed_domains = RADEON_GEM_DOMAIN_VRAM;
+		list[idx].tv.bo = &list[idx].robj->tbo;
+		list[idx].tiling_flags = 0;
+		list[idx].handle = 0;
+		list_add(&list[idx++].tv.head, head);
+	}
+
+	return list;
+}
+
+/**
+ * radeon_vm_grab_id - allocate the next free VMID
+ *
+ * @rdev: radeon_device pointer
+ * @vm: vm to allocate id for
+ * @ring: ring we want to submit job to
+ *
+ * Allocate an id for the vm (cayman+).
+ * Returns the fence we need to sync to (if any).
+ *
+ * Global and local mutex must be locked!
+ */
+struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
+				       struct radeon_vm *vm, int ring)
+{
+	struct radeon_fence *best[RADEON_NUM_RINGS] = {};
+	unsigned choices[2] = {};
+	unsigned i;
+
+	/* check if the id is still valid */
+	if (vm->last_id_use && vm->last_id_use == rdev->vm_manager.active[vm->id])
+		return NULL;
+
+	/* we definately need to flush */
+	radeon_fence_unref(&vm->last_flush);
+
+	/* skip over VMID 0, since it is the system VM */
+	for (i = 1; i < rdev->vm_manager.nvm; ++i) {
+		struct radeon_fence *fence = rdev->vm_manager.active[i];
+
+		if (fence == NULL) {
+			/* found a free one */
+			vm->id = i;
+			trace_radeon_vm_grab_id(vm->id, ring);
+			return NULL;
+		}
+
+		if (radeon_fence_is_earlier(fence, best[fence->ring])) {
+			best[fence->ring] = fence;
+			choices[fence->ring == ring ? 0 : 1] = i;
+		}
+	}
+
+	for (i = 0; i < 2; ++i) {
+		if (choices[i]) {
+			vm->id = choices[i];
+			trace_radeon_vm_grab_id(vm->id, ring);
+			return rdev->vm_manager.active[choices[i]];
+		}
+	}
+
+	/* should never happen */
+	BUG();
+	return NULL;
+}
+
+/**
+ * radeon_vm_flush - hardware flush the vm
+ *
+ * @rdev: radeon_device pointer
+ * @vm: vm we want to flush
+ * @ring: ring to use for flush
+ *
+ * Flush the vm (cayman+).
+ *
+ * Global and local mutex must be locked!
+ */
+void radeon_vm_flush(struct radeon_device *rdev,
+		     struct radeon_vm *vm,
+		     int ring)
+{
+	uint64_t pd_addr = radeon_bo_gpu_offset(vm->page_directory);
+
+	/* if we can't remember our last VM flush then flush now! */
+	/* XXX figure out why we have to flush all the time */
+	if (!vm->last_flush || true || pd_addr != vm->pd_gpu_addr) {
+		vm->pd_gpu_addr = pd_addr;
+		radeon_ring_vm_flush(rdev, ring, vm);
+	}
+}
+
+/**
+ * radeon_vm_fence - remember fence for vm
+ *
+ * @rdev: radeon_device pointer
+ * @vm: vm we want to fence
+ * @fence: fence to remember
+ *
+ * Fence the vm (cayman+).
+ * Set the fence used to protect page table and id.
+ *
+ * Global and local mutex must be locked!
+ */
+void radeon_vm_fence(struct radeon_device *rdev,
+		     struct radeon_vm *vm,
+		     struct radeon_fence *fence)
+{
+	radeon_fence_unref(&vm->fence);
+	vm->fence = radeon_fence_ref(fence);
+
+	radeon_fence_unref(&rdev->vm_manager.active[vm->id]);
+	rdev->vm_manager.active[vm->id] = radeon_fence_ref(fence);
+
+	radeon_fence_unref(&vm->last_id_use);
+	vm->last_id_use = radeon_fence_ref(fence);
+
+        /* we just flushed the VM, remember that */
+        if (!vm->last_flush)
+                vm->last_flush = radeon_fence_ref(fence);
+}
+
+/**
+ * radeon_vm_bo_find - find the bo_va for a specific vm & bo
+ *
+ * @vm: requested vm
+ * @bo: requested buffer object
+ *
+ * Find @bo inside the requested vm (cayman+).
+ * Search inside the @bos vm list for the requested vm
+ * Returns the found bo_va or NULL if none is found
+ *
+ * Object has to be reserved!
+ */
+struct radeon_bo_va *radeon_vm_bo_find(struct radeon_vm *vm,
+				       struct radeon_bo *bo)
+{
+	struct radeon_bo_va *bo_va;
+
+	list_for_each_entry(bo_va, &bo->va, bo_list) {
+		if (bo_va->vm == vm) {
+			return bo_va;
+		}
+	}
+	return NULL;
+}
+
+/**
+ * radeon_vm_bo_add - add a bo to a specific vm
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ * @bo: radeon buffer object
+ *
+ * Add @bo into the requested vm (cayman+).
+ * Add @bo to the list of bos associated with the vm
+ * Returns newly added bo_va or NULL for failure
+ *
+ * Object has to be reserved!
+ */
+struct radeon_bo_va *radeon_vm_bo_add(struct radeon_device *rdev,
+				      struct radeon_vm *vm,
+				      struct radeon_bo *bo)
+{
+	struct radeon_bo_va *bo_va;
+
+	bo_va = kzalloc(sizeof(struct radeon_bo_va), GFP_KERNEL);
+	if (bo_va == NULL) {
+		return NULL;
+	}
+	bo_va->vm = vm;
+	bo_va->bo = bo;
+	bo_va->soffset = 0;
+	bo_va->eoffset = 0;
+	bo_va->flags = 0;
+	bo_va->valid = false;
+	bo_va->ref_count = 1;
+	INIT_LIST_HEAD(&bo_va->bo_list);
+	INIT_LIST_HEAD(&bo_va->vm_list);
+	INIT_LIST_HEAD(&bo_va->vm_status);
+
+	mutex_lock(&vm->mutex);
+	list_add(&bo_va->vm_list, &vm->va);
+	list_add_tail(&bo_va->bo_list, &bo->va);
+	mutex_unlock(&vm->mutex);
+
+	return bo_va;
+}
+
+/**
+ * radeon_vm_clear_bo - initially clear the page dir/table
+ *
+ * @rdev: radeon_device pointer
+ * @bo: bo to clear
+ */
+static int radeon_vm_clear_bo(struct radeon_device *rdev,
+			      struct radeon_bo *bo)
+{
+        struct ttm_validate_buffer tv;
+        struct ww_acquire_ctx ticket;
+        struct list_head head;
+	struct radeon_ib ib;
+	unsigned entries;
+	uint64_t addr;
+	int r;
+
+        memset(&tv, 0, sizeof(tv));
+        tv.bo = &bo->tbo;
+
+        INIT_LIST_HEAD(&head);
+        list_add(&tv.head, &head);
+
+        r = ttm_eu_reserve_buffers(&ticket, &head);
+        if (r)
+		return r;
+
+        r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+        if (r)
+                goto error;
+
+	addr = radeon_bo_gpu_offset(bo);
+	entries = radeon_bo_size(bo) / 8;
+
+	r = radeon_ib_get(rdev, R600_RING_TYPE_DMA_INDEX, &ib,
+			  NULL, entries * 2 + 64);
+	if (r)
+                goto error;
+
+	ib.length_dw = 0;
+
+	radeon_asic_vm_set_page(rdev, &ib, addr, 0, entries, 0, 0);
+
+	r = radeon_ib_schedule(rdev, &ib, NULL);
+	if (r)
+                goto error;
+
+	ttm_eu_fence_buffer_objects(&ticket, &head, ib.fence);
+	radeon_ib_free(rdev, &ib);
+
+	return 0;
+
+error:
+	ttm_eu_backoff_reservation(&ticket, &head);
+	return r;
+}
+
+/**
+ * radeon_vm_bo_set_addr - set bos virtual address inside a vm
+ *
+ * @rdev: radeon_device pointer
+ * @bo_va: bo_va to store the address
+ * @soffset: requested offset of the buffer in the VM address space
+ * @flags: attributes of pages (read/write/valid/etc.)
+ *
+ * Set offset of @bo_va (cayman+).
+ * Validate and set the offset requested within the vm address space.
+ * Returns 0 for success, error for failure.
+ *
+ * Object has to be reserved!
+ */
+int radeon_vm_bo_set_addr(struct radeon_device *rdev,
+			  struct radeon_bo_va *bo_va,
+			  uint64_t soffset,
+			  uint32_t flags)
+{
+	uint64_t size = radeon_bo_size(bo_va->bo);
+	uint64_t eoffset, last_offset = 0;
+	struct radeon_vm *vm = bo_va->vm;
+	struct radeon_bo_va *tmp;
+	struct list_head *head;
+	unsigned last_pfn, pt_idx;
+	int r;
+
+	if (soffset) {
+		/* make sure object fit at this offset */
+		eoffset = soffset + size;
+		if (soffset >= eoffset) {
+			return -EINVAL;
+		}
+
+		last_pfn = eoffset / RADEON_GPU_PAGE_SIZE;
+		if (last_pfn > rdev->vm_manager.max_pfn) {
+			dev_err(rdev->dev, "va above limit (0x%08X > 0x%08X)\n",
+				last_pfn, rdev->vm_manager.max_pfn);
+			return -EINVAL;
+		}
+
+	} else {
+		eoffset = last_pfn = 0;
+	}
+
+	mutex_lock(&vm->mutex);
+	head = &vm->va;
+	last_offset = 0;
+	list_for_each_entry(tmp, &vm->va, vm_list) {
+		if (bo_va == tmp) {
+			/* skip over currently modified bo */
+			continue;
+		}
+
+		if (soffset >= last_offset && eoffset <= tmp->soffset) {
+			/* bo can be added before this one */
+			break;
+		}
+		if (eoffset > tmp->soffset && soffset < tmp->eoffset) {
+			/* bo and tmp overlap, invalid offset */
+			dev_err(rdev->dev, "bo %p va 0x%08X conflict with (bo %p 0x%08X 0x%08X)\n",
+				bo_va->bo, (unsigned)bo_va->soffset, tmp->bo,
+				(unsigned)tmp->soffset, (unsigned)tmp->eoffset);
+			mutex_unlock(&vm->mutex);
+			return -EINVAL;
+		}
+		last_offset = tmp->eoffset;
+		head = &tmp->vm_list;
+	}
+
+	if (bo_va->soffset) {
+		/* add a clone of the bo_va to clear the old address */
+		tmp = kzalloc(sizeof(struct radeon_bo_va), GFP_KERNEL);
+		if (!tmp) {
+			mutex_unlock(&vm->mutex);
+			return -ENOMEM;
+		}
+		tmp->soffset = bo_va->soffset;
+		tmp->eoffset = bo_va->eoffset;
+		tmp->vm = vm;
+		list_add(&tmp->vm_status, &vm->freed);
+	}
+
+	bo_va->soffset = soffset;
+	bo_va->eoffset = eoffset;
+	bo_va->flags = flags;
+	bo_va->valid = false;
+	list_move(&bo_va->vm_list, head);
+
+	soffset = (soffset / RADEON_GPU_PAGE_SIZE) >> radeon_vm_block_size;
+	eoffset = (eoffset / RADEON_GPU_PAGE_SIZE) >> radeon_vm_block_size;
+
+	BUG_ON(eoffset >= radeon_vm_num_pdes(rdev));
+
+	if (eoffset > vm->max_pde_used)
+		vm->max_pde_used = eoffset;
+
+	radeon_bo_unreserve(bo_va->bo);
+
+	/* walk over the address space and allocate the page tables */
+	for (pt_idx = soffset; pt_idx <= eoffset; ++pt_idx) {
+		struct radeon_bo *pt;
+
+		if (vm->page_tables[pt_idx].bo)
+			continue;
+
+		/* drop mutex to allocate and clear page table */
+		mutex_unlock(&vm->mutex);
+
+		r = radeon_bo_create(rdev, RADEON_VM_PTE_COUNT * 8,
+				     RADEON_GPU_PAGE_SIZE, true,
+				     RADEON_GEM_DOMAIN_VRAM, NULL, &pt);
+		if (r)
+			return r;
+
+		r = radeon_vm_clear_bo(rdev, pt);
+		if (r) {
+			radeon_bo_unref(&pt);
+			radeon_bo_reserve(bo_va->bo, false);
+			return r;
+		}
+
+		/* aquire mutex again */
+		mutex_lock(&vm->mutex);
+		if (vm->page_tables[pt_idx].bo) {
+			/* someone else allocated the pt in the meantime */
+			mutex_unlock(&vm->mutex);
+			radeon_bo_unref(&pt);
+			mutex_lock(&vm->mutex);
+			continue;
+		}
+
+		vm->page_tables[pt_idx].addr = 0;
+		vm->page_tables[pt_idx].bo = pt;
+	}
+
+	mutex_unlock(&vm->mutex);
+	return radeon_bo_reserve(bo_va->bo, false);
+}
+
+/**
+ * radeon_vm_map_gart - get the physical address of a gart page
+ *
+ * @rdev: radeon_device pointer
+ * @addr: the unmapped addr
+ *
+ * Look up the physical address of the page that the pte resolves
+ * to (cayman+).
+ * Returns the physical address of the page.
+ */
+uint64_t radeon_vm_map_gart(struct radeon_device *rdev, uint64_t addr)
+{
+	uint64_t result;
+
+	/* page table offset */
+	result = rdev->gart.pages_addr[addr >> PAGE_SHIFT];
+
+	/* in case cpu page size != gpu page size*/
+	result |= addr & (~PAGE_MASK);
+
+	return result;
+}
+
+/**
+ * radeon_vm_page_flags - translate page flags to what the hw uses
+ *
+ * @flags: flags comming from userspace
+ *
+ * Translate the flags the userspace ABI uses to hw flags.
+ */
+static uint32_t radeon_vm_page_flags(uint32_t flags)
+{
+        uint32_t hw_flags = 0;
+        hw_flags |= (flags & RADEON_VM_PAGE_VALID) ? R600_PTE_VALID : 0;
+        hw_flags |= (flags & RADEON_VM_PAGE_READABLE) ? R600_PTE_READABLE : 0;
+        hw_flags |= (flags & RADEON_VM_PAGE_WRITEABLE) ? R600_PTE_WRITEABLE : 0;
+        if (flags & RADEON_VM_PAGE_SYSTEM) {
+                hw_flags |= R600_PTE_SYSTEM;
+                hw_flags |= (flags & RADEON_VM_PAGE_SNOOPED) ? R600_PTE_SNOOPED : 0;
+        }
+        return hw_flags;
+}
+
+/**
+ * radeon_vm_update_pdes - make sure that page directory is valid
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ * @start: start of GPU address range
+ * @end: end of GPU address range
+ *
+ * Allocates new page tables if necessary
+ * and updates the page directory (cayman+).
+ * Returns 0 for success, error for failure.
+ *
+ * Global and local mutex must be locked!
+ */
+int radeon_vm_update_page_directory(struct radeon_device *rdev,
+				    struct radeon_vm *vm)
+{
+	struct radeon_bo *pd = vm->page_directory;
+	uint64_t pd_addr = radeon_bo_gpu_offset(pd);
+	uint32_t incr = RADEON_VM_PTE_COUNT * 8;
+	uint64_t last_pde = ~0, last_pt = ~0;
+	unsigned count = 0, pt_idx, ndw;
+	struct radeon_ib ib;
+	int r;
+
+	/* padding, etc. */
+	ndw = 64;
+
+	/* assume the worst case */
+	ndw += vm->max_pde_used * 16;
+
+	/* update too big for an IB */
+	if (ndw > 0xfffff)
+		return -ENOMEM;
+
+	r = radeon_ib_get(rdev, R600_RING_TYPE_DMA_INDEX, &ib, NULL, ndw * 4);
+	if (r)
+		return r;
+	ib.length_dw = 0;
+
+	/* walk over the address space and update the page directory */
+	for (pt_idx = 0; pt_idx <= vm->max_pde_used; ++pt_idx) {
+		struct radeon_bo *bo = vm->page_tables[pt_idx].bo;
+		uint64_t pde, pt;
+
+		if (bo == NULL)
+			continue;
+
+		pt = radeon_bo_gpu_offset(bo);
+		if (vm->page_tables[pt_idx].addr == pt)
+			continue;
+		vm->page_tables[pt_idx].addr = pt;
+
+		pde = pd_addr + pt_idx * 8;
+		if (((last_pde + 8 * count) != pde) ||
+		    ((last_pt + incr * count) != pt)) {
+
+			if (count) {
+				radeon_asic_vm_set_page(rdev, &ib, last_pde,
+							last_pt, count, incr,
+							R600_PTE_VALID);
+			}
+
+			count = 1;
+			last_pde = pde;
+			last_pt = pt;
+		} else {
+			++count;
+		}
+	}
+
+	if (count)
+		radeon_asic_vm_set_page(rdev, &ib, last_pde, last_pt, count,
+					incr, R600_PTE_VALID);
+
+	if (ib.length_dw != 0) {
+		radeon_semaphore_sync_to(ib.semaphore, pd->tbo.sync_obj);
+		radeon_semaphore_sync_to(ib.semaphore, vm->last_id_use);
+		r = radeon_ib_schedule(rdev, &ib, NULL);
+		if (r) {
+			radeon_ib_free(rdev, &ib);
+			return r;
+		}
+		radeon_fence_unref(&vm->fence);
+		vm->fence = radeon_fence_ref(ib.fence);
+		radeon_fence_unref(&vm->last_flush);
+	}
+	radeon_ib_free(rdev, &ib);
+
+	return 0;
+}
+
+/**
+ * radeon_vm_frag_ptes - add fragment information to PTEs
+ *
+ * @rdev: radeon_device pointer
+ * @ib: IB for the update
+ * @pe_start: first PTE to handle
+ * @pe_end: last PTE to handle
+ * @addr: addr those PTEs should point to
+ * @flags: hw mapping flags
+ *
+ * Global and local mutex must be locked!
+ */
+static void radeon_vm_frag_ptes(struct radeon_device *rdev,
+				struct radeon_ib *ib,
+				uint64_t pe_start, uint64_t pe_end,
+				uint64_t addr, uint32_t flags)
+{
+	/**
+	 * The MC L1 TLB supports variable sized pages, based on a fragment
+	 * field in the PTE. When this field is set to a non-zero value, page
+	 * granularity is increased from 4KB to (1 << (12 + frag)). The PTE
+	 * flags are considered valid for all PTEs within the fragment range
+	 * and corresponding mappings are assumed to be physically contiguous.
+	 *
+	 * The L1 TLB can store a single PTE for the whole fragment,
+	 * significantly increasing the space available for translation
+	 * caching. This leads to large improvements in throughput when the
+	 * TLB is under pressure.
+	 *
+	 * The L2 TLB distributes small and large fragments into two
+	 * asymmetric partitions. The large fragment cache is significantly
+	 * larger. Thus, we try to use large fragments wherever possible.
+	 * Userspace can support this by aligning virtual base address and
+	 * allocation size to the fragment size.
+	 */
+
+	/* NI is optimized for 256KB fragments, SI and newer for 64KB */
+	uint64_t frag_flags = rdev->family == CHIP_CAYMAN ?
+			R600_PTE_FRAG_256KB : R600_PTE_FRAG_64KB;
+	uint64_t frag_align = rdev->family == CHIP_CAYMAN ? 0x200 : 0x80;
+
+	uint64_t frag_start = ALIGN(pe_start, frag_align);
+	uint64_t frag_end = pe_end & ~(frag_align - 1);
+
+	unsigned count;
+
+	/* system pages are non continuously */
+	if ((flags & R600_PTE_SYSTEM) || !(flags & R600_PTE_VALID) ||
+	    (frag_start >= frag_end)) {
+
+		count = (pe_end - pe_start) / 8;
+		radeon_asic_vm_set_page(rdev, ib, pe_start, addr, count,
+					RADEON_GPU_PAGE_SIZE, flags);
+		return;
+	}
+
+	/* handle the 4K area at the beginning */
+	if (pe_start != frag_start) {
+		count = (frag_start - pe_start) / 8;
+		radeon_asic_vm_set_page(rdev, ib, pe_start, addr, count,
+					RADEON_GPU_PAGE_SIZE, flags);
+		addr += RADEON_GPU_PAGE_SIZE * count;
+	}
+
+	/* handle the area in the middle */
+	count = (frag_end - frag_start) / 8;
+	radeon_asic_vm_set_page(rdev, ib, frag_start, addr, count,
+				RADEON_GPU_PAGE_SIZE, flags | frag_flags);
+
+	/* handle the 4K area at the end */
+	if (frag_end != pe_end) {
+		addr += RADEON_GPU_PAGE_SIZE * count;
+		count = (pe_end - frag_end) / 8;
+		radeon_asic_vm_set_page(rdev, ib, frag_end, addr, count,
+					RADEON_GPU_PAGE_SIZE, flags);
+	}
+}
+
+/**
+ * radeon_vm_update_ptes - make sure that page tables are valid
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ * @start: start of GPU address range
+ * @end: end of GPU address range
+ * @dst: destination address to map to
+ * @flags: mapping flags
+ *
+ * Update the page tables in the range @start - @end (cayman+).
+ *
+ * Global and local mutex must be locked!
+ */
+static void radeon_vm_update_ptes(struct radeon_device *rdev,
+				  struct radeon_vm *vm,
+				  struct radeon_ib *ib,
+				  uint64_t start, uint64_t end,
+				  uint64_t dst, uint32_t flags)
+{
+	uint64_t mask = RADEON_VM_PTE_COUNT - 1;
+	uint64_t last_pte = ~0, last_dst = ~0;
+	unsigned count = 0;
+	uint64_t addr;
+
+	start = start / RADEON_GPU_PAGE_SIZE;
+	end = end / RADEON_GPU_PAGE_SIZE;
+
+	/* walk over the address space and update the page tables */
+	for (addr = start; addr < end; ) {
+		uint64_t pt_idx = addr >> radeon_vm_block_size;
+		struct radeon_bo *pt = vm->page_tables[pt_idx].bo;
+		unsigned nptes;
+		uint64_t pte;
+
+		radeon_semaphore_sync_to(ib->semaphore, pt->tbo.sync_obj);
+
+		if ((addr & ~mask) == (end & ~mask))
+			nptes = end - addr;
+		else
+			nptes = RADEON_VM_PTE_COUNT - (addr & mask);
+
+		pte = radeon_bo_gpu_offset(pt);
+		pte += (addr & mask) * 8;
+
+		if ((last_pte + 8 * count) != pte) {
+
+			if (count) {
+				radeon_vm_frag_ptes(rdev, ib, last_pte,
+						    last_pte + 8 * count,
+						    last_dst, flags);
+			}
+
+			count = nptes;
+			last_pte = pte;
+			last_dst = dst;
+		} else {
+			count += nptes;
+		}
+
+		addr += nptes;
+		dst += nptes * RADEON_GPU_PAGE_SIZE;
+	}
+
+	if (count) {
+		radeon_vm_frag_ptes(rdev, ib, last_pte,
+				    last_pte + 8 * count,
+				    last_dst, flags);
+	}
+}
+
+/**
+ * radeon_vm_bo_update - map a bo into the vm page table
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ * @bo: radeon buffer object
+ * @mem: ttm mem
+ *
+ * Fill in the page table entries for @bo (cayman+).
+ * Returns 0 for success, -EINVAL for failure.
+ *
+ * Object have to be reserved and mutex must be locked!
+ */
+int radeon_vm_bo_update(struct radeon_device *rdev,
+			struct radeon_bo_va *bo_va,
+			struct ttm_mem_reg *mem)
+{
+	struct radeon_vm *vm = bo_va->vm;
+	struct radeon_ib ib;
+	unsigned nptes, ndw;
+	uint64_t addr;
+	int r;
+
+
+	if (!bo_va->soffset) {
+		dev_err(rdev->dev, "bo %p don't has a mapping in vm %p\n",
+			bo_va->bo, vm);
+		return -EINVAL;
+	}
+
+	if ((bo_va->valid && mem) || (!bo_va->valid && mem == NULL))
+		return 0;
+
+	bo_va->flags &= ~RADEON_VM_PAGE_VALID;
+	bo_va->flags &= ~RADEON_VM_PAGE_SYSTEM;
+	if (mem) {
+		addr = mem->start << PAGE_SHIFT;
+		if (mem->mem_type != TTM_PL_SYSTEM) {
+			bo_va->flags |= RADEON_VM_PAGE_VALID;
+			bo_va->valid = true;
+		}
+		if (mem->mem_type == TTM_PL_TT) {
+			bo_va->flags |= RADEON_VM_PAGE_SYSTEM;
+		} else {
+			addr += rdev->vm_manager.vram_base_offset;
+		}
+	} else {
+		addr = 0;
+		bo_va->valid = false;
+	}
+
+	trace_radeon_vm_bo_update(bo_va);
+
+	nptes = (bo_va->eoffset - bo_va->soffset) / RADEON_GPU_PAGE_SIZE;
+
+	/* padding, etc. */
+	ndw = 64;
+
+	if (radeon_vm_block_size > 11)
+		/* reserve space for one header for every 2k dwords */
+		ndw += (nptes >> 11) * 4;
+	else
+		/* reserve space for one header for
+		    every (1 << BLOCK_SIZE) entries */
+		ndw += (nptes >> radeon_vm_block_size) * 4;
+
+	/* reserve space for pte addresses */
+	ndw += nptes * 2;
+
+	/* update too big for an IB */
+	if (ndw > 0xfffff)
+		return -ENOMEM;
+
+	r = radeon_ib_get(rdev, R600_RING_TYPE_DMA_INDEX, &ib, NULL, ndw * 4);
+	if (r)
+		return r;
+	ib.length_dw = 0;
+
+	radeon_vm_update_ptes(rdev, vm, &ib, bo_va->soffset, bo_va->eoffset,
+			      addr, radeon_vm_page_flags(bo_va->flags));
+
+	radeon_semaphore_sync_to(ib.semaphore, vm->fence);
+	r = radeon_ib_schedule(rdev, &ib, NULL);
+	if (r) {
+		radeon_ib_free(rdev, &ib);
+		return r;
+	}
+	radeon_fence_unref(&vm->fence);
+	vm->fence = radeon_fence_ref(ib.fence);
+	radeon_ib_free(rdev, &ib);
+	radeon_fence_unref(&vm->last_flush);
+
+	return 0;
+}
+
+/**
+ * radeon_vm_clear_freed - clear freed BOs in the PT
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ *
+ * Make sure all freed BOs are cleared in the PT.
+ * Returns 0 for success.
+ *
+ * PTs have to be reserved and mutex must be locked!
+ */
+int radeon_vm_clear_freed(struct radeon_device *rdev,
+			  struct radeon_vm *vm)
+{
+	struct radeon_bo_va *bo_va, *tmp;
+	int r;
+
+	list_for_each_entry_safe(bo_va, tmp, &vm->freed, vm_status) {
+		list_del(&bo_va->vm_status);
+		r = radeon_vm_bo_update(rdev, bo_va, NULL);
+		kfree(bo_va);
+		if (r)
+			return r;
+	}
+	return 0;
+
+}
+
+/**
+ * radeon_vm_bo_rmv - remove a bo to a specific vm
+ *
+ * @rdev: radeon_device pointer
+ * @bo_va: requested bo_va
+ *
+ * Remove @bo_va->bo from the requested vm (cayman+).
+ *
+ * Object have to be reserved!
+ */
+void radeon_vm_bo_rmv(struct radeon_device *rdev,
+		      struct radeon_bo_va *bo_va)
+{
+	struct radeon_vm *vm = bo_va->vm;
+
+	list_del(&bo_va->bo_list);
+
+	mutex_lock(&vm->mutex);
+	list_del(&bo_va->vm_list);
+
+	if (bo_va->soffset) {
+		bo_va->bo = NULL;
+		list_add(&bo_va->vm_status, &vm->freed);
+	} else {
+		kfree(bo_va);
+	}
+
+	mutex_unlock(&vm->mutex);
+}
+
+/**
+ * radeon_vm_bo_invalidate - mark the bo as invalid
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ * @bo: radeon buffer object
+ *
+ * Mark @bo as invalid (cayman+).
+ */
+void radeon_vm_bo_invalidate(struct radeon_device *rdev,
+			     struct radeon_bo *bo)
+{
+	struct radeon_bo_va *bo_va;
+
+	list_for_each_entry(bo_va, &bo->va, bo_list) {
+		bo_va->valid = false;
+	}
+}
+
+/**
+ * radeon_vm_init - initialize a vm instance
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ *
+ * Init @vm fields (cayman+).
+ */
+int radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm)
+{
+	const unsigned align = min(RADEON_VM_PTB_ALIGN_SIZE,
+		RADEON_VM_PTE_COUNT * 8);
+	unsigned pd_size, pd_entries, pts_size;
+	int r;
+
+	vm->id = 0;
+	vm->ib_bo_va = NULL;
+	vm->fence = NULL;
+	vm->last_flush = NULL;
+	vm->last_id_use = NULL;
+	mutex_init(&vm->mutex);
+	INIT_LIST_HEAD(&vm->va);
+	INIT_LIST_HEAD(&vm->freed);
+
+	pd_size = radeon_vm_directory_size(rdev);
+	pd_entries = radeon_vm_num_pdes(rdev);
+
+	/* allocate page table array */
+	pts_size = pd_entries * sizeof(struct radeon_vm_pt);
+	vm->page_tables = kzalloc(pts_size, GFP_KERNEL);
+	if (vm->page_tables == NULL) {
+		DRM_ERROR("Cannot allocate memory for page table array\n");
+		return -ENOMEM;
+	}
+
+	r = radeon_bo_create(rdev, pd_size, align, true,
+			     RADEON_GEM_DOMAIN_VRAM, NULL,
+			     &vm->page_directory);
+	if (r)
+		return r;
+
+	r = radeon_vm_clear_bo(rdev, vm->page_directory);
+	if (r) {
+		radeon_bo_unref(&vm->page_directory);
+		vm->page_directory = NULL;
+		return r;
+	}
+
+	return 0;
+}
+
+/**
+ * radeon_vm_fini - tear down a vm instance
+ *
+ * @rdev: radeon_device pointer
+ * @vm: requested vm
+ *
+ * Tear down @vm (cayman+).
+ * Unbind the VM and remove all bos from the vm bo list
+ */
+void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm)
+{
+	struct radeon_bo_va *bo_va, *tmp;
+	int i, r;
+
+	if (!list_empty(&vm->va)) {
+		dev_err(rdev->dev, "still active bo inside vm\n");
+	}
+	list_for_each_entry_safe(bo_va, tmp, &vm->va, vm_list) {
+		list_del_init(&bo_va->vm_list);
+		r = radeon_bo_reserve(bo_va->bo, false);
+		if (!r) {
+			list_del_init(&bo_va->bo_list);
+			radeon_bo_unreserve(bo_va->bo);
+			kfree(bo_va);
+		}
+	}
+	list_for_each_entry_safe(bo_va, tmp, &vm->freed, vm_status)
+		kfree(bo_va);
+
+	for (i = 0; i < radeon_vm_num_pdes(rdev); i++)
+		radeon_bo_unref(&vm->page_tables[i].bo);
+	kfree(vm->page_tables);
+
+	radeon_bo_unref(&vm->page_directory);
+
+	radeon_fence_unref(&vm->fence);
+	radeon_fence_unref(&vm->last_flush);
+	radeon_fence_unref(&vm->last_id_use);
+
+	mutex_destroy(&vm->mutex);
+}
diff --git a/drivers/gpu/drm/radeon/reg_srcs/r600 b/drivers/gpu/drm/radeon/reg_srcs/r600
index 20bfbda7b3f..ec0c6829c1d 100644
--- a/drivers/gpu/drm/radeon/reg_srcs/r600
+++ b/drivers/gpu/drm/radeon/reg_srcs/r600
@@ -18,6 +18,7 @@ r600 0x9400
 0x00028A3C VGT_GROUP_VECT_1_FMT_CNTL
 0x00028A40 VGT_GS_MODE
 0x00028A6C VGT_GS_OUT_PRIM_TYPE
+0x00028B38 VGT_GS_MAX_VERT_OUT
 0x000088C8 VGT_GS_PER_ES
 0x000088E8 VGT_GS_PER_VS
 0x000088D4 VGT_GS_VERTEX_REUSE
diff --git a/drivers/gpu/drm/radeon/rs400.c b/drivers/gpu/drm/radeon/rs400.c
index 9566b5940a5..a0f96decece 100644
--- a/drivers/gpu/drm/radeon/rs400.c
+++ b/drivers/gpu/drm/radeon/rs400.c
@@ -212,21 +212,16 @@ void rs400_gart_fini(struct radeon_device *rdev)
 #define RS400_PTE_WRITEABLE (1 << 2)
 #define RS400_PTE_READABLE  (1 << 3)
 
-int rs400_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr)
+void rs400_gart_set_page(struct radeon_device *rdev, unsigned i, uint64_t addr)
 {
 	uint32_t entry;
 	u32 *gtt = rdev->gart.ptr;
 
-	if (i < 0 || i > rdev->gart.num_gpu_pages) {
-		return -EINVAL;
-	}
-
 	entry = (lower_32_bits(addr) & PAGE_MASK) |
 		((upper_32_bits(addr) & 0xff) << 4) |
 		RS400_PTE_WRITEABLE | RS400_PTE_READABLE;
 	entry = cpu_to_le32(entry);
 	gtt[i] = entry;
-	return 0;
 }
 
 int rs400_mc_wait_for_idle(struct radeon_device *rdev)
@@ -484,6 +479,7 @@ int rs400_resume(struct radeon_device *rdev)
 
 int rs400_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
 	r100_irq_disable(rdev);
@@ -493,6 +489,7 @@ int rs400_suspend(struct radeon_device *rdev)
 
 void rs400_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -560,6 +557,9 @@ int rs400_init(struct radeon_device *rdev)
 		return r;
 	r300_set_reg_safe(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = rs400_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c
index 76cc8d3aafe..d1a35cb1c91 100644
--- a/drivers/gpu/drm/radeon/rs600.c
+++ b/drivers/gpu/drm/radeon/rs600.c
@@ -109,19 +109,7 @@ void avivo_wait_for_vblank(struct radeon_device *rdev, int crtc)
 	}
 }
 
-void rs600_pre_page_flip(struct radeon_device *rdev, int crtc)
-{
-	/* enable the pflip int */
-	radeon_irq_kms_pflip_irq_get(rdev, crtc);
-}
-
-void rs600_post_page_flip(struct radeon_device *rdev, int crtc)
-{
-	/* disable the pflip int */
-	radeon_irq_kms_pflip_irq_put(rdev, crtc);
-}
-
-u32 rs600_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
+void rs600_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 {
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 	u32 tmp = RREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset);
@@ -148,9 +136,15 @@ u32 rs600_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 	/* Unlock the lock, so double-buffering can take place inside vblank */
 	tmp &= ~AVIVO_D1GRPH_UPDATE_LOCK;
 	WREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset, tmp);
+}
+
+bool rs600_page_flip_pending(struct radeon_device *rdev, int crtc_id)
+{
+	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 
 	/* Return current update_pending status: */
-	return RREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset) & AVIVO_D1GRPH_SURFACE_UPDATE_PENDING;
+	return !!(RREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset) &
+		AVIVO_D1GRPH_SURFACE_UPDATE_PENDING);
 }
 
 void avivo_program_fmt(struct drm_encoder *encoder)
@@ -632,24 +626,16 @@ static void rs600_gart_fini(struct radeon_device *rdev)
 	radeon_gart_table_vram_free(rdev);
 }
 
-#define R600_PTE_VALID     (1 << 0)
-#define R600_PTE_SYSTEM    (1 << 1)
-#define R600_PTE_SNOOPED   (1 << 2)
-#define R600_PTE_READABLE  (1 << 5)
-#define R600_PTE_WRITEABLE (1 << 6)
-
-int rs600_gart_set_page(struct radeon_device *rdev, int i, uint64_t addr)
+void rs600_gart_set_page(struct radeon_device *rdev, unsigned i, uint64_t addr)
 {
 	void __iomem *ptr = (void *)rdev->gart.ptr;
 
-	if (i < 0 || i > rdev->gart.num_gpu_pages) {
-		return -EINVAL;
-	}
 	addr = addr & 0xFFFFFFFFFFFFF000ULL;
-	addr |= R600_PTE_VALID | R600_PTE_SYSTEM | R600_PTE_SNOOPED;
-	addr |= R600_PTE_READABLE | R600_PTE_WRITEABLE;
+	if (addr == rdev->dummy_page.addr)
+		addr |= R600_PTE_SYSTEM | R600_PTE_SNOOPED;
+	else
+		addr |= R600_PTE_GART;
 	writeq(addr, ptr + (i * 8));
-	return 0;
 }
 
 int rs600_irq_set(struct radeon_device *rdev)
@@ -787,7 +773,7 @@ int rs600_irq_process(struct radeon_device *rdev)
 				wake_up(&rdev->irq.vblank_queue);
 			}
 			if (atomic_read(&rdev->irq.pflip[0]))
-				radeon_crtc_handle_flip(rdev, 0);
+				radeon_crtc_handle_vblank(rdev, 0);
 		}
 		if (G_007EDC_LB_D2_VBLANK_INTERRUPT(rdev->irq.stat_regs.r500.disp_int)) {
 			if (rdev->irq.crtc_vblank_int[1]) {
@@ -796,7 +782,7 @@ int rs600_irq_process(struct radeon_device *rdev)
 				wake_up(&rdev->irq.vblank_queue);
 			}
 			if (atomic_read(&rdev->irq.pflip[1]))
-				radeon_crtc_handle_flip(rdev, 1);
+				radeon_crtc_handle_vblank(rdev, 1);
 		}
 		if (G_007EDC_DC_HOT_PLUG_DETECT1_INTERRUPT(rdev->irq.stat_regs.r500.disp_int)) {
 			queue_hotplug = true;
@@ -1058,6 +1044,7 @@ int rs600_resume(struct radeon_device *rdev)
 
 int rs600_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
@@ -1068,6 +1055,7 @@ int rs600_suspend(struct radeon_device *rdev)
 
 void rs600_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r600_audio_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
@@ -1136,6 +1124,9 @@ int rs600_init(struct radeon_device *rdev)
 		return r;
 	rs600_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = rs600_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/rs690.c b/drivers/gpu/drm/radeon/rs690.c
index e7dab069ccc..3462b64369b 100644
--- a/drivers/gpu/drm/radeon/rs690.c
+++ b/drivers/gpu/drm/radeon/rs690.c
@@ -766,6 +766,7 @@ int rs690_resume(struct radeon_device *rdev)
 
 int rs690_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
@@ -776,6 +777,7 @@ int rs690_suspend(struct radeon_device *rdev)
 
 void rs690_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r600_audio_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
@@ -845,6 +847,9 @@ int rs690_init(struct radeon_device *rdev)
 		return r;
 	rs600_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = rs690_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/rs780_dpm.c b/drivers/gpu/drm/radeon/rs780_dpm.c
index 6af8505cf4d..02f7710de47 100644
--- a/drivers/gpu/drm/radeon/rs780_dpm.c
+++ b/drivers/gpu/drm/radeon/rs780_dpm.c
@@ -623,14 +623,6 @@ int rs780_dpm_enable(struct radeon_device *rdev)
 	if (pi->gfx_clock_gating)
 		r600_gfx_clockgating_enable(rdev, true);
 
-	if (rdev->irq.installed && (rdev->pm.int_thermal_type == THERMAL_TYPE_RV6XX)) {
-		ret = r600_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-	}
-
 	return 0;
 }
 
@@ -815,9 +807,6 @@ static int rs780_parse_power_table(struct radeon_device *rdev)
 				  power_info->pplib.ucNumStates, GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
-	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
-	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
-	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
 
 	for (i = 0; i < power_info->pplib.ucNumStates; i++) {
 		power_state = (union pplib_power_state *)
@@ -867,6 +856,10 @@ int rs780_dpm_init(struct radeon_device *rdev)
 		return -ENOMEM;
 	rdev->pm.dpm.priv = pi;
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = rs780_parse_power_table(rdev);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/radeon/rv515.c b/drivers/gpu/drm/radeon/rv515.c
index 5d1c316115e..3e21e869015 100644
--- a/drivers/gpu/drm/radeon/rv515.c
+++ b/drivers/gpu/drm/radeon/rv515.c
@@ -406,8 +406,9 @@ void rv515_mc_resume(struct radeon_device *rdev, struct rv515_mc_save *save)
 	for (i = 0; i < rdev->num_crtc; i++) {
 		if (save->crtc_enabled[i]) {
 			tmp = RREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + crtc_offsets[i]);
-			if ((tmp & 0x3) != 0) {
-				tmp &= ~0x3;
+			if ((tmp & 0x7) != 3) {
+				tmp &= ~0x7;
+				tmp |= 0x3;
 				WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + crtc_offsets[i], tmp);
 			}
 			tmp = RREG32(AVIVO_D1GRPH_UPDATE + crtc_offsets[i]);
@@ -596,6 +597,7 @@ int rv515_resume(struct radeon_device *rdev)
 
 int rv515_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
 	rs600_irq_disable(rdev);
@@ -612,6 +614,7 @@ void rv515_set_safe_registers(struct radeon_device *rdev)
 
 void rv515_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -685,6 +688,9 @@ int rv515_init(struct radeon_device *rdev)
 		return r;
 	rv515_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = rv515_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/rv6xx_dpm.c b/drivers/gpu/drm/radeon/rv6xx_dpm.c
index 26633a02525..e7045b08571 100644
--- a/drivers/gpu/drm/radeon/rv6xx_dpm.c
+++ b/drivers/gpu/drm/radeon/rv6xx_dpm.c
@@ -1546,7 +1546,6 @@ int rv6xx_dpm_enable(struct radeon_device *rdev)
 {
 	struct rv6xx_power_info *pi = rv6xx_get_pi(rdev);
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
-	int ret;
 
 	if (r600_dynamicpm_enabled(rdev))
 		return -EINVAL;
@@ -1594,15 +1593,6 @@ int rv6xx_dpm_enable(struct radeon_device *rdev)
 	r600_power_level_enable(rdev, R600_POWER_LEVEL_MEDIUM, true);
 	r600_power_level_enable(rdev, R600_POWER_LEVEL_HIGH, true);
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-		ret = r600_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-	}
-
 	rv6xx_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
 	r600_start_dpm(rdev);
@@ -1901,9 +1891,6 @@ static int rv6xx_parse_power_table(struct radeon_device *rdev)
 				  power_info->pplib.ucNumStates, GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
-	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
-	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
-	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
 
 	for (i = 0; i < power_info->pplib.ucNumStates; i++) {
 		power_state = (union pplib_power_state *)
@@ -1953,6 +1940,10 @@ int rv6xx_dpm_init(struct radeon_device *rdev)
 		return -ENOMEM;
 	rdev->pm.dpm.priv = pi;
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = rv6xx_parse_power_table(rdev);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c
index 9f5846743c9..da8703d8d45 100644
--- a/drivers/gpu/drm/radeon/rv770.c
+++ b/drivers/gpu/drm/radeon/rv770.c
@@ -801,7 +801,7 @@ u32 rv770_get_xclk(struct radeon_device *rdev)
 	return reference_clock;
 }
 
-u32 rv770_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
+void rv770_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 {
 	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 	u32 tmp = RREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset);
@@ -835,9 +835,15 @@ u32 rv770_page_flip(struct radeon_device *rdev, int crtc_id, u64 crtc_base)
 	/* Unlock the lock, so double-buffering can take place inside vblank */
 	tmp &= ~AVIVO_D1GRPH_UPDATE_LOCK;
 	WREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset, tmp);
+}
+
+bool rv770_page_flip_pending(struct radeon_device *rdev, int crtc_id)
+{
+	struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[crtc_id];
 
 	/* Return current update_pending status: */
-	return RREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset) & AVIVO_D1GRPH_SURFACE_UPDATE_PENDING;
+	return !!(RREG32(AVIVO_D1GRPH_UPDATE + radeon_crtc->crtc_offset) &
+		AVIVO_D1GRPH_SURFACE_UPDATE_PENDING);
 }
 
 /* get temperature in millidegrees */
@@ -1071,7 +1077,8 @@ static void rv770_mc_program(struct radeon_device *rdev)
  */
 void r700_cp_stop(struct radeon_device *rdev)
 {
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 	WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT));
 	WREG32(SCRATCH_UMSK, 0);
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
@@ -1123,6 +1130,35 @@ void r700_cp_fini(struct radeon_device *rdev)
 	radeon_scratch_free(rdev, ring->rptr_save_reg);
 }
 
+void rv770_set_clk_bypass_mode(struct radeon_device *rdev)
+{
+	u32 tmp, i;
+
+	if (rdev->flags & RADEON_IS_IGP)
+		return;
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL_2);
+	tmp &= SCLK_MUX_SEL_MASK;
+	tmp |= SCLK_MUX_SEL(1) | SCLK_MUX_UPDATE;
+	WREG32(CG_SPLL_FUNC_CNTL_2, tmp);
+
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CG_SPLL_STATUS) & SPLL_CHG_STATUS)
+			break;
+		udelay(1);
+	}
+
+	tmp &= ~SCLK_MUX_UPDATE;
+	WREG32(CG_SPLL_FUNC_CNTL_2, tmp);
+
+	tmp = RREG32(MPLL_CNTL_MODE);
+	if ((rdev->family == CHIP_RV710) || (rdev->family == CHIP_RV730))
+		tmp &= ~RV730_MPLL_MCLK_SEL;
+	else
+		tmp &= ~MPLL_MCLK_SEL;
+	WREG32(MPLL_CNTL_MODE, tmp);
+}
+
 /*
  * Core functions
  */
@@ -1291,6 +1327,9 @@ static void rv770_gpu_init(struct radeon_device *rdev)
 	if (tmp < rdev->config.rv770.max_simds) {
 		rdev->config.rv770.max_simds = tmp;
 	}
+	tmp = rdev->config.rv770.max_simds -
+		r600_count_pipe_bits((cc_gc_shader_pipe_config >> 16) & R7XX_MAX_SIMDS_MASK);
+	rdev->config.rv770.active_simds = tmp;
 
 	switch (rdev->config.rv770.max_tile_pipes) {
 	case 1:
@@ -1665,14 +1704,6 @@ static int rv770_startup(struct radeon_device *rdev)
 
 	rv770_mc_program(rdev);
 
-	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
-		r = r600_init_microcode(rdev);
-		if (r) {
-			DRM_ERROR("Failed to load firmware!\n");
-			return r;
-		}
-	}
-
 	if (rdev->flags & RADEON_IS_AGP) {
 		rv770_agp_enable(rdev);
 	} else {
@@ -1728,14 +1759,12 @@ static int rv770_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     R600_CP_RB_RPTR, R600_CP_RB_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR, DMA_RB_WPTR,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
 	if (r)
 		return r;
@@ -1754,7 +1783,6 @@ static int rv770_startup(struct radeon_device *rdev)
 	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 	if (ring->ring_size) {
 		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 				     RADEON_CP_PACKET2);
 		if (!r)
 			r = uvd_v1_0_init(rdev);
@@ -1792,6 +1820,9 @@ int rv770_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	rv770_init_golden_registers(rdev);
 
+	if (rdev->pm.pm_method == PM_METHOD_DPM)
+		radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = rv770_startup(rdev);
 	if (r) {
@@ -1806,6 +1837,7 @@ int rv770_resume(struct radeon_device *rdev)
 
 int rv770_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_suspend(rdev);
@@ -1876,6 +1908,17 @@ int rv770_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
+		r = r600_init_microcode(rdev);
+		if (r) {
+			DRM_ERROR("Failed to load firmware!\n");
+			return r;
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024);
 
@@ -1915,15 +1958,16 @@ int rv770_init(struct radeon_device *rdev)
 
 void rv770_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r700_cp_fini(rdev);
 	r600_dma_fini(rdev);
 	r600_irq_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
 	radeon_irq_kms_fini(rdev);
-	rv770_pcie_gart_fini(rdev);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_fini(rdev);
+	rv770_pcie_gart_fini(rdev);
 	r600_vram_scratch_fini(rdev);
 	radeon_gem_fini(rdev);
 	radeon_fence_driver_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/rv770_dma.c b/drivers/gpu/drm/radeon/rv770_dma.c
index aca8cbe8a33..bbf2e076ee4 100644
--- a/drivers/gpu/drm/radeon/rv770_dma.c
+++ b/drivers/gpu/drm/radeon/rv770_dma.c
@@ -86,6 +86,7 @@ int rv770_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
diff --git a/drivers/gpu/drm/radeon/rv770_dpm.c b/drivers/gpu/drm/radeon/rv770_dpm.c
index 374499db20c..3c76e1dcdf0 100644
--- a/drivers/gpu/drm/radeon/rv770_dpm.c
+++ b/drivers/gpu/drm/radeon/rv770_dpm.c
@@ -1863,8 +1863,8 @@ void rv770_enable_auto_throttle_source(struct radeon_device *rdev,
 	}
 }
 
-int rv770_set_thermal_temperature_range(struct radeon_device *rdev,
-					int min_temp, int max_temp)
+static int rv770_set_thermal_temperature_range(struct radeon_device *rdev,
+					       int min_temp, int max_temp)
 {
 	int low_temp = 0 * 1000;
 	int high_temp = 255 * 1000;
@@ -1966,6 +1966,15 @@ int rv770_dpm_enable(struct radeon_device *rdev)
 	if (pi->mg_clock_gating)
 		rv770_mg_clock_gating_enable(rdev, true);
 
+	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
+
+	return 0;
+}
+
+int rv770_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		PPSMC_Result result;
@@ -1981,8 +1990,6 @@ int rv770_dpm_enable(struct radeon_device *rdev)
 			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
 	}
 
-	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
-
 	return 0;
 }
 
@@ -2167,7 +2174,6 @@ static void rv7xx_parse_pplib_clock_info(struct radeon_device *rdev,
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
 	struct rv7xx_ps *ps = rv770_get_ps(rps);
 	u32 sclk, mclk;
-	u16 vddc;
 	struct rv7xx_pl *pl;
 
 	switch (index) {
@@ -2207,8 +2213,8 @@ static void rv7xx_parse_pplib_clock_info(struct radeon_device *rdev,
 
 	/* patch up vddc if necessary */
 	if (pl->vddc == 0xff01) {
-		if (radeon_atom_get_max_vddc(rdev, 0, 0, &vddc) == 0)
-			pl->vddc = vddc;
+		if (pi->max_vddc)
+			pl->vddc = pi->max_vddc;
 	}
 
 	if (rps->class & ATOM_PPLIB_CLASSIFICATION_ACPI) {
@@ -2244,14 +2250,12 @@ static void rv7xx_parse_pplib_clock_info(struct radeon_device *rdev,
 		pl->vddci = vddci;
 	}
 
-	if (rdev->family >= CHIP_BARTS) {
-		if ((rps->class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) ==
-		    ATOM_PPLIB_CLASSIFICATION_UI_PERFORMANCE) {
-			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.sclk = pl->sclk;
-			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.mclk = pl->mclk;
-			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddc = pl->vddc;
-			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddci = pl->vddci;
-		}
+	if ((rps->class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) ==
+	    ATOM_PPLIB_CLASSIFICATION_UI_PERFORMANCE) {
+		rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.sclk = pl->sclk;
+		rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.mclk = pl->mclk;
+		rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddc = pl->vddc;
+		rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddci = pl->vddci;
 	}
 }
 
@@ -2277,9 +2281,6 @@ int rv7xx_parse_power_table(struct radeon_device *rdev)
 				  power_info->pplib.ucNumStates, GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
-	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
-	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
-	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
 
 	for (i = 0; i < power_info->pplib.ucNumStates; i++) {
 		power_state = (union pplib_power_state *)
@@ -2328,12 +2329,6 @@ void rv770_get_engine_memory_ss(struct radeon_device *rdev)
 	pi->mclk_ss = radeon_atombios_get_asic_ss_info(rdev, &ss,
 						       ASIC_INTERNAL_MEMORY_SS, 0);
 
-	/* disable ss, causes hangs on some cayman boards */
-	if (rdev->family == CHIP_CAYMAN) {
-		pi->sclk_ss = false;
-		pi->mclk_ss = false;
-	}
-
 	if (pi->sclk_ss || pi->mclk_ss)
 		pi->dynamic_ss = true;
 	else
@@ -2357,6 +2352,10 @@ int rv770_dpm_init(struct radeon_device *rdev)
 	pi->min_vddc_in_table = 0;
 	pi->max_vddc_in_table = 0;
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = rv7xx_parse_power_table(rdev);
 	if (ret)
 		return ret;
@@ -2522,14 +2521,13 @@ u32 rv770_dpm_get_mclk(struct radeon_device *rdev, bool low)
 bool rv770_dpm_vblank_too_short(struct radeon_device *rdev)
 {
 	u32 vblank_time = r600_dpm_get_vblank_time(rdev);
-	u32 switch_limit = 300;
-
-	/* quirks */
-	/* ASUS K70AF */
-	if ((rdev->pdev->device == 0x9553) &&
-	    (rdev->pdev->subsystem_vendor == 0x1043) &&
-	    (rdev->pdev->subsystem_device == 0x1c42))
-		switch_limit = 200;
+	u32 switch_limit = 200; /* 300 */
+
+	/* RV770 */
+	/* mclk switching doesn't seem to work reliably on desktop RV770s */
+	if ((rdev->family == CHIP_RV770) &&
+	    !(rdev->flags & RADEON_IS_MOBILITY))
+		switch_limit = 0xffffffff; /* disable mclk switching */
 
 	if (vblank_time < switch_limit)
 		return true;
diff --git a/drivers/gpu/drm/radeon/rv770_dpm.h b/drivers/gpu/drm/radeon/rv770_dpm.h
index 9244effc6b5..f776634840c 100644
--- a/drivers/gpu/drm/radeon/rv770_dpm.h
+++ b/drivers/gpu/drm/radeon/rv770_dpm.h
@@ -283,8 +283,4 @@ int rv770_read_smc_soft_register(struct radeon_device *rdev,
 int rv770_write_smc_soft_register(struct radeon_device *rdev,
 				  u16 reg_offset, u32 value);
 
-/* thermal */
-int rv770_set_thermal_temperature_range(struct radeon_device *rdev,
-					int min_temp, int max_temp);
-
 #endif
diff --git a/drivers/gpu/drm/radeon/rv770d.h b/drivers/gpu/drm/radeon/rv770d.h
index 1ae277152cc..3cf1e292154 100644
--- a/drivers/gpu/drm/radeon/rv770d.h
+++ b/drivers/gpu/drm/radeon/rv770d.h
@@ -100,14 +100,21 @@
 #define	CG_SPLL_FUNC_CNTL_2				0x604
 #define		SCLK_MUX_SEL(x)				((x) << 0)
 #define		SCLK_MUX_SEL_MASK			(0x1ff << 0)
+#define		SCLK_MUX_UPDATE				(1 << 26)
 #define	CG_SPLL_FUNC_CNTL_3				0x608
 #define		SPLL_FB_DIV(x)				((x) << 0)
 #define		SPLL_FB_DIV_MASK			(0x3ffffff << 0)
 #define		SPLL_DITHEN				(1 << 28)
+#define	CG_SPLL_STATUS					0x60c
+#define		SPLL_CHG_STATUS				(1 << 1)
 
 #define	SPLL_CNTL_MODE					0x610
 #define		SPLL_DIV_SYNC				(1 << 5)
 
+#define MPLL_CNTL_MODE                                  0x61c
+#       define MPLL_MCLK_SEL                            (1 << 11)
+#       define RV730_MPLL_MCLK_SEL                      (1 << 25)
+
 #define	MPLL_AD_FUNC_CNTL				0x624
 #define		CLKF(x)					((x) << 0)
 #define		CLKF_MASK				(0x7f << 0)
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index 85e1edfaa3b..9e854fd016d 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -39,33 +39,39 @@ MODULE_FIRMWARE("radeon/TAHITI_pfp.bin");
 MODULE_FIRMWARE("radeon/TAHITI_me.bin");
 MODULE_FIRMWARE("radeon/TAHITI_ce.bin");
 MODULE_FIRMWARE("radeon/TAHITI_mc.bin");
+MODULE_FIRMWARE("radeon/TAHITI_mc2.bin");
 MODULE_FIRMWARE("radeon/TAHITI_rlc.bin");
 MODULE_FIRMWARE("radeon/TAHITI_smc.bin");
 MODULE_FIRMWARE("radeon/PITCAIRN_pfp.bin");
 MODULE_FIRMWARE("radeon/PITCAIRN_me.bin");
 MODULE_FIRMWARE("radeon/PITCAIRN_ce.bin");
 MODULE_FIRMWARE("radeon/PITCAIRN_mc.bin");
+MODULE_FIRMWARE("radeon/PITCAIRN_mc2.bin");
 MODULE_FIRMWARE("radeon/PITCAIRN_rlc.bin");
 MODULE_FIRMWARE("radeon/PITCAIRN_smc.bin");
 MODULE_FIRMWARE("radeon/VERDE_pfp.bin");
 MODULE_FIRMWARE("radeon/VERDE_me.bin");
 MODULE_FIRMWARE("radeon/VERDE_ce.bin");
 MODULE_FIRMWARE("radeon/VERDE_mc.bin");
+MODULE_FIRMWARE("radeon/VERDE_mc2.bin");
 MODULE_FIRMWARE("radeon/VERDE_rlc.bin");
 MODULE_FIRMWARE("radeon/VERDE_smc.bin");
 MODULE_FIRMWARE("radeon/OLAND_pfp.bin");
 MODULE_FIRMWARE("radeon/OLAND_me.bin");
 MODULE_FIRMWARE("radeon/OLAND_ce.bin");
 MODULE_FIRMWARE("radeon/OLAND_mc.bin");
+MODULE_FIRMWARE("radeon/OLAND_mc2.bin");
 MODULE_FIRMWARE("radeon/OLAND_rlc.bin");
 MODULE_FIRMWARE("radeon/OLAND_smc.bin");
 MODULE_FIRMWARE("radeon/HAINAN_pfp.bin");
 MODULE_FIRMWARE("radeon/HAINAN_me.bin");
 MODULE_FIRMWARE("radeon/HAINAN_ce.bin");
 MODULE_FIRMWARE("radeon/HAINAN_mc.bin");
+MODULE_FIRMWARE("radeon/HAINAN_mc2.bin");
 MODULE_FIRMWARE("radeon/HAINAN_rlc.bin");
 MODULE_FIRMWARE("radeon/HAINAN_smc.bin");
 
+static u32 si_get_cu_active_bitmap(struct radeon_device *rdev, u32 se, u32 sh);
 static void si_pcie_gen3_enable(struct radeon_device *rdev);
 static void si_program_aspm(struct radeon_device *rdev);
 extern void sumo_rlc_fini(struct radeon_device *rdev);
@@ -80,6 +86,8 @@ extern void evergreen_print_gpu_status_regs(struct radeon_device *rdev);
 extern bool evergreen_is_display_hung(struct radeon_device *rdev);
 static void si_enable_gui_idle_interrupt(struct radeon_device *rdev,
 					 bool enable);
+static void si_init_pg(struct radeon_device *rdev);
+static void si_init_cg(struct radeon_device *rdev);
 static void si_fini_pg(struct radeon_device *rdev);
 static void si_fini_cg(struct radeon_device *rdev);
 static void si_rlc_stop(struct radeon_device *rdev);
@@ -1460,41 +1468,38 @@ static const u32 hainan_io_mc_regs[TAHITI_IO_MC_REGS_SIZE][2] = {
 };
 
 /* ucode loading */
-static int si_mc_load_microcode(struct radeon_device *rdev)
+int si_mc_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
 	u32 running, blackout = 0;
 	u32 *io_mc_regs;
-	int i, ucode_size, regs_size;
+	int i, regs_size, ucode_size;
 
 	if (!rdev->mc_fw)
 		return -EINVAL;
 
+	ucode_size = rdev->mc_fw->size / 4;
+
 	switch (rdev->family) {
 	case CHIP_TAHITI:
 		io_mc_regs = (u32 *)&tahiti_io_mc_regs;
-		ucode_size = SI_MC_UCODE_SIZE;
 		regs_size = TAHITI_IO_MC_REGS_SIZE;
 		break;
 	case CHIP_PITCAIRN:
 		io_mc_regs = (u32 *)&pitcairn_io_mc_regs;
-		ucode_size = SI_MC_UCODE_SIZE;
 		regs_size = TAHITI_IO_MC_REGS_SIZE;
 		break;
 	case CHIP_VERDE:
 	default:
 		io_mc_regs = (u32 *)&verde_io_mc_regs;
-		ucode_size = SI_MC_UCODE_SIZE;
 		regs_size = TAHITI_IO_MC_REGS_SIZE;
 		break;
 	case CHIP_OLAND:
 		io_mc_regs = (u32 *)&oland_io_mc_regs;
-		ucode_size = OLAND_MC_UCODE_SIZE;
 		regs_size = TAHITI_IO_MC_REGS_SIZE;
 		break;
 	case CHIP_HAINAN:
 		io_mc_regs = (u32 *)&hainan_io_mc_regs;
-		ucode_size = OLAND_MC_UCODE_SIZE;
 		regs_size = TAHITI_IO_MC_REGS_SIZE;
 		break;
 	}
@@ -1550,7 +1555,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 	const char *chip_name;
 	const char *rlc_chip_name;
 	size_t pfp_req_size, me_req_size, ce_req_size, rlc_req_size, mc_req_size;
-	size_t smc_req_size;
+	size_t smc_req_size, mc2_req_size;
 	char fw_name[30];
 	int err;
 
@@ -1565,6 +1570,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 		ce_req_size = SI_CE_UCODE_SIZE * 4;
 		rlc_req_size = SI_RLC_UCODE_SIZE * 4;
 		mc_req_size = SI_MC_UCODE_SIZE * 4;
+		mc2_req_size = TAHITI_MC_UCODE_SIZE * 4;
 		smc_req_size = ALIGN(TAHITI_SMC_UCODE_SIZE, 4);
 		break;
 	case CHIP_PITCAIRN:
@@ -1575,6 +1581,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 		ce_req_size = SI_CE_UCODE_SIZE * 4;
 		rlc_req_size = SI_RLC_UCODE_SIZE * 4;
 		mc_req_size = SI_MC_UCODE_SIZE * 4;
+		mc2_req_size = PITCAIRN_MC_UCODE_SIZE * 4;
 		smc_req_size = ALIGN(PITCAIRN_SMC_UCODE_SIZE, 4);
 		break;
 	case CHIP_VERDE:
@@ -1585,6 +1592,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 		ce_req_size = SI_CE_UCODE_SIZE * 4;
 		rlc_req_size = SI_RLC_UCODE_SIZE * 4;
 		mc_req_size = SI_MC_UCODE_SIZE * 4;
+		mc2_req_size = VERDE_MC_UCODE_SIZE * 4;
 		smc_req_size = ALIGN(VERDE_SMC_UCODE_SIZE, 4);
 		break;
 	case CHIP_OLAND:
@@ -1594,7 +1602,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 		me_req_size = SI_PM4_UCODE_SIZE * 4;
 		ce_req_size = SI_CE_UCODE_SIZE * 4;
 		rlc_req_size = SI_RLC_UCODE_SIZE * 4;
-		mc_req_size = OLAND_MC_UCODE_SIZE * 4;
+		mc_req_size = mc2_req_size = OLAND_MC_UCODE_SIZE * 4;
 		smc_req_size = ALIGN(OLAND_SMC_UCODE_SIZE, 4);
 		break;
 	case CHIP_HAINAN:
@@ -1604,7 +1612,7 @@ static int si_init_microcode(struct radeon_device *rdev)
 		me_req_size = SI_PM4_UCODE_SIZE * 4;
 		ce_req_size = SI_CE_UCODE_SIZE * 4;
 		rlc_req_size = SI_RLC_UCODE_SIZE * 4;
-		mc_req_size = OLAND_MC_UCODE_SIZE * 4;
+		mc_req_size = mc2_req_size = OLAND_MC_UCODE_SIZE * 4;
 		smc_req_size = ALIGN(HAINAN_SMC_UCODE_SIZE, 4);
 		break;
 	default: BUG();
@@ -1657,16 +1665,22 @@ static int si_init_microcode(struct radeon_device *rdev)
 		err = -EINVAL;
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc2.bin", chip_name);
 	err = request_firmware(&rdev->mc_fw, fw_name, rdev->dev);
-	if (err)
-		goto out;
-	if (rdev->mc_fw->size != mc_req_size) {
+	if (err) {
+		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name);
+		err = request_firmware(&rdev->mc_fw, fw_name, rdev->dev);
+		if (err)
+			goto out;
+	}
+	if ((rdev->mc_fw->size != mc_req_size) &&
+	    (rdev->mc_fw->size != mc2_req_size)) {
 		printk(KERN_ERR
 		       "si_mc: Bogus length %zu in firmware \"%s\"\n",
 		       rdev->mc_fw->size, fw_name);
 		err = -EINVAL;
 	}
+	DRM_INFO("%s: %zu bytes\n", fw_name, rdev->mc_fw->size);
 
 	snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name);
 	err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev);
@@ -2887,7 +2901,7 @@ static void si_gpu_init(struct radeon_device *rdev)
 	u32 sx_debug_1;
 	u32 hdp_host_path_cntl;
 	u32 tmp;
-	int i, j;
+	int i, j, k;
 
 	switch (rdev->family) {
 	case CHIP_TAHITI:
@@ -3085,6 +3099,14 @@ static void si_gpu_init(struct radeon_device *rdev)
 		     rdev->config.si.max_sh_per_se,
 		     rdev->config.si.max_cu_per_sh);
 
+	for (i = 0; i < rdev->config.si.max_shader_engines; i++) {
+		for (j = 0; j < rdev->config.si.max_sh_per_se; j++) {
+			for (k = 0; k < rdev->config.si.max_cu_per_sh; k++) {
+				rdev->config.si.active_cus +=
+					hweight32(si_get_cu_active_bitmap(rdev, i, j));
+			}
+		}
+	}
 
 	/* set HW defaults for 3D engine */
 	WREG32(CP_QUEUE_THRESHOLDS, (ROQ_IB1_START(0x16) |
@@ -3173,7 +3195,7 @@ void si_fence_ring_emit(struct radeon_device *rdev,
 	/* EVENT_WRITE_EOP - flush caches, send int */
 	radeon_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4));
 	radeon_ring_write(ring, EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) | EVENT_INDEX(5));
-	radeon_ring_write(ring, addr & 0xffffffff);
+	radeon_ring_write(ring, lower_32_bits(addr));
 	radeon_ring_write(ring, (upper_32_bits(addr) & 0xff) | DATA_SEL(1) | INT_SEL(2));
 	radeon_ring_write(ring, fence->seq);
 	radeon_ring_write(ring, 0);
@@ -3206,7 +3228,7 @@ void si_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 			radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
 			radeon_ring_write(ring, (1 << 8));
 			radeon_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc);
-			radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff);
+			radeon_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr));
 			radeon_ring_write(ring, next_rptr);
 		}
 
@@ -3247,7 +3269,8 @@ static void si_cp_enable(struct radeon_device *rdev, bool enable)
 	if (enable)
 		WREG32(CP_ME_CNTL, 0);
 	else {
-		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+		if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+			radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 		WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT));
 		WREG32(SCRATCH_UMSK, 0);
 		rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
@@ -3431,8 +3454,6 @@ static int si_cp_resume(struct radeon_device *rdev)
 
 	WREG32(CP_RB0_BASE, ring->gpu_addr >> 8);
 
-	ring->rptr = RREG32(CP_RB0_RPTR);
-
 	/* ring1  - compute only */
 	/* Set ring buffer size */
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP1_INDEX];
@@ -3457,8 +3478,6 @@ static int si_cp_resume(struct radeon_device *rdev)
 
 	WREG32(CP_RB1_BASE, ring->gpu_addr >> 8);
 
-	ring->rptr = RREG32(CP_RB1_RPTR);
-
 	/* ring2 - compute only */
 	/* Set ring buffer size */
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP2_INDEX];
@@ -3483,8 +3502,6 @@ static int si_cp_resume(struct radeon_device *rdev)
 
 	WREG32(CP_RB2_BASE, ring->gpu_addr >> 8);
 
-	ring->rptr = RREG32(CP_RB2_RPTR);
-
 	/* start the rings */
 	si_cp_start(rdev);
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = true;
@@ -3508,6 +3525,9 @@ static int si_cp_resume(struct radeon_device *rdev)
 
 	si_enable_gui_idle_interrupt(rdev, true);
 
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+
 	return 0;
 }
 
@@ -3724,6 +3744,106 @@ static void si_gpu_soft_reset(struct radeon_device *rdev, u32 reset_mask)
 	evergreen_print_gpu_status_regs(rdev);
 }
 
+static void si_set_clk_bypass_mode(struct radeon_device *rdev)
+{
+	u32 tmp, i;
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL);
+	tmp |= SPLL_BYPASS_EN;
+	WREG32(CG_SPLL_FUNC_CNTL, tmp);
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL_2);
+	tmp |= SPLL_CTLREQ_CHG;
+	WREG32(CG_SPLL_FUNC_CNTL_2, tmp);
+
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(SPLL_STATUS) & SPLL_CHG_STATUS)
+			break;
+		udelay(1);
+	}
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL_2);
+	tmp &= ~(SPLL_CTLREQ_CHG | SCLK_MUX_UPDATE);
+	WREG32(CG_SPLL_FUNC_CNTL_2, tmp);
+
+	tmp = RREG32(MPLL_CNTL_MODE);
+	tmp &= ~MPLL_MCLK_SEL;
+	WREG32(MPLL_CNTL_MODE, tmp);
+}
+
+static void si_spll_powerdown(struct radeon_device *rdev)
+{
+	u32 tmp;
+
+	tmp = RREG32(SPLL_CNTL_MODE);
+	tmp |= SPLL_SW_DIR_CONTROL;
+	WREG32(SPLL_CNTL_MODE, tmp);
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL);
+	tmp |= SPLL_RESET;
+	WREG32(CG_SPLL_FUNC_CNTL, tmp);
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL);
+	tmp |= SPLL_SLEEP;
+	WREG32(CG_SPLL_FUNC_CNTL, tmp);
+
+	tmp = RREG32(SPLL_CNTL_MODE);
+	tmp &= ~SPLL_SW_DIR_CONTROL;
+	WREG32(SPLL_CNTL_MODE, tmp);
+}
+
+static void si_gpu_pci_config_reset(struct radeon_device *rdev)
+{
+	struct evergreen_mc_save save;
+	u32 tmp, i;
+
+	dev_info(rdev->dev, "GPU pci config reset\n");
+
+	/* disable dpm? */
+
+	/* disable cg/pg */
+	si_fini_pg(rdev);
+	si_fini_cg(rdev);
+
+	/* Disable CP parsing/prefetching */
+	WREG32(CP_ME_CNTL, CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT);
+	/* dma0 */
+	tmp = RREG32(DMA_RB_CNTL + DMA0_REGISTER_OFFSET);
+	tmp &= ~DMA_RB_ENABLE;
+	WREG32(DMA_RB_CNTL + DMA0_REGISTER_OFFSET, tmp);
+	/* dma1 */
+	tmp = RREG32(DMA_RB_CNTL + DMA1_REGISTER_OFFSET);
+	tmp &= ~DMA_RB_ENABLE;
+	WREG32(DMA_RB_CNTL + DMA1_REGISTER_OFFSET, tmp);
+	/* XXX other engines? */
+
+	/* halt the rlc, disable cp internal ints */
+	si_rlc_stop(rdev);
+
+	udelay(50);
+
+	/* disable mem access */
+	evergreen_mc_stop(rdev, &save);
+	if (evergreen_mc_wait_for_idle(rdev)) {
+		dev_warn(rdev->dev, "Wait for MC idle timed out !\n");
+	}
+
+	/* set mclk/sclk to bypass */
+	si_set_clk_bypass_mode(rdev);
+	/* powerdown spll */
+	si_spll_powerdown(rdev);
+	/* disable BM */
+	pci_clear_master(rdev->pdev);
+	/* reset */
+	radeon_pci_config_reset(rdev);
+	/* wait for asic to come out of reset */
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CONFIG_MEMSIZE) != 0xffffffff)
+			break;
+		udelay(1);
+	}
+}
+
 int si_asic_reset(struct radeon_device *rdev)
 {
 	u32 reset_mask;
@@ -3733,10 +3853,17 @@ int si_asic_reset(struct radeon_device *rdev)
 	if (reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, true);
 
+	/* try soft reset */
 	si_gpu_soft_reset(rdev, reset_mask);
 
 	reset_mask = si_gpu_check_soft_reset(rdev);
 
+	/* try pci config reset */
+	if (reset_mask && radeon_hard_reset)
+		si_gpu_pci_config_reset(rdev);
+
+	reset_mask = si_gpu_check_soft_reset(rdev);
+
 	if (!reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, false);
 
@@ -3759,11 +3886,9 @@ bool si_gfx_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 	if (!(reset_mask & (RADEON_RESET_GFX |
 			    RADEON_RESET_COMPUTE |
 			    RADEON_RESET_CP))) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force CP activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
@@ -3928,18 +4053,21 @@ static int si_pcie_gart_enable(struct radeon_device *rdev)
 	WREG32(MC_VM_MX_L1_TLB_CNTL,
 	       (0xA << 7) |
 	       ENABLE_L1_TLB |
+	       ENABLE_L1_FRAGMENT_PROCESSING |
 	       SYSTEM_ACCESS_MODE_NOT_IN_SYS |
 	       ENABLE_ADVANCED_DRIVER_MODEL |
 	       SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU);
 	/* Setup L2 cache */
 	WREG32(VM_L2_CNTL, ENABLE_L2_CACHE |
+	       ENABLE_L2_FRAGMENT_PROCESSING |
 	       ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE |
 	       ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE |
 	       EFFECTIVE_L2_QUEUE_SIZE(7) |
 	       CONTEXT1_IDENTITY_ACCESS_MODE(1));
 	WREG32(VM_L2_CNTL2, INVALIDATE_ALL_L1_TLBS | INVALIDATE_L2_CACHE);
 	WREG32(VM_L2_CNTL3, L2_CACHE_BIGK_ASSOCIATIVITY |
-	       L2_CACHE_BIGK_FRAGMENT_SIZE(0));
+	       BANK_SELECT(4) |
+	       L2_CACHE_BIGK_FRAGMENT_SIZE(4));
 	/* setup context0 */
 	WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR, rdev->mc.gtt_start >> 12);
 	WREG32(VM_CONTEXT0_PAGE_TABLE_END_ADDR, rdev->mc.gtt_end >> 12);
@@ -3976,6 +4104,7 @@ static int si_pcie_gart_enable(struct radeon_device *rdev)
 	       (u32)(rdev->dummy_page.addr >> 12));
 	WREG32(VM_CONTEXT1_CNTL2, 4);
 	WREG32(VM_CONTEXT1_CNTL, ENABLE_CONTEXT | PAGE_TABLE_DEPTH(1) |
+				PAGE_TABLE_BLOCK_SIZE(radeon_vm_block_size - 9) |
 				RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT |
 				RANGE_PROTECTION_FAULT_ENABLE_DEFAULT |
 				DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT |
@@ -5212,8 +5341,8 @@ static void si_enable_hdp_ls(struct radeon_device *rdev,
 		WREG32(HDP_MEM_POWER_LS, data);
 }
 
-void si_update_cg(struct radeon_device *rdev,
-		  u32 block, bool enable)
+static void si_update_cg(struct radeon_device *rdev,
+			 u32 block, bool enable)
 {
 	if (block & RADEON_CG_BLOCK_GFX) {
 		si_enable_gui_idle_interrupt(rdev, false);
@@ -5379,6 +5508,9 @@ static void si_init_pg(struct radeon_device *rdev)
 		si_init_ao_cu_mask(rdev);
 		if (rdev->pg_flags & RADEON_PG_SUPPORT_GFX_PG) {
 			si_init_gfx_cgpg(rdev);
+		} else {
+			WREG32(RLC_SAVE_AND_RESTORE_BASE, rdev->rlc.save_restore_gpu_addr >> 8);
+			WREG32(RLC_CLEAR_STATE_RESTORE_BASE, rdev->rlc.clear_state_gpu_addr >> 8);
 		}
 		si_enable_dma_pg(rdev, true);
 		si_enable_gfx_cgpg(rdev, true);
@@ -5566,7 +5698,7 @@ static void si_disable_interrupt_state(struct radeon_device *rdev)
 	}
 
 	if (!ASIC_IS_NODCE(rdev)) {
-		WREG32(DACA_AUTODETECT_INT_CONTROL, 0);
+		WREG32(DAC_AUTODETECT_INT_CONTROL, 0);
 
 		tmp = RREG32(DC_HPD1_INT_CONTROL) & DC_HPDx_INT_POLARITY;
 		WREG32(DC_HPD1_INT_CONTROL, tmp);
@@ -5661,7 +5793,6 @@ int si_irq_set(struct radeon_device *rdev)
 	u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0;
 	u32 hpd1 = 0, hpd2 = 0, hpd3 = 0, hpd4 = 0, hpd5 = 0, hpd6 = 0;
 	u32 grbm_int_cntl = 0;
-	u32 grph1 = 0, grph2 = 0, grph3 = 0, grph4 = 0, grph5 = 0, grph6 = 0;
 	u32 dma_cntl, dma_cntl1;
 	u32 thermal_int = 0;
 
@@ -5800,16 +5931,22 @@ int si_irq_set(struct radeon_device *rdev)
 	}
 
 	if (rdev->num_crtc >= 2) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, grph1);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, grph2);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 	if (rdev->num_crtc >= 4) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, grph3);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, grph4);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 	if (rdev->num_crtc >= 6) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, grph5);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, grph6);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 
 	if (!ASIC_IS_NODCE(rdev)) {
@@ -5966,6 +6103,7 @@ static inline u32 si_get_ih_wptr(struct radeon_device *rdev)
 		tmp = RREG32(IH_RB_CNTL);
 		tmp |= IH_WPTR_OVERFLOW_CLEAR;
 		WREG32(IH_RB_CNTL, tmp);
+		wptr &= ~RB_OVERFLOW;
 	}
 	return (wptr & rdev->ih.ptr_mask);
 }
@@ -6027,7 +6165,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[0]))
-						radeon_crtc_handle_flip(rdev, 0);
+						radeon_crtc_handle_vblank(rdev, 0);
 					rdev->irq.stat_regs.evergreen.disp_int &= ~LB_D1_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D1 vblank\n");
 				}
@@ -6053,7 +6191,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[1]))
-						radeon_crtc_handle_flip(rdev, 1);
+						radeon_crtc_handle_vblank(rdev, 1);
 					rdev->irq.stat_regs.evergreen.disp_int_cont &= ~LB_D2_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D2 vblank\n");
 				}
@@ -6079,7 +6217,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[2]))
-						radeon_crtc_handle_flip(rdev, 2);
+						radeon_crtc_handle_vblank(rdev, 2);
 					rdev->irq.stat_regs.evergreen.disp_int_cont2 &= ~LB_D3_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D3 vblank\n");
 				}
@@ -6105,7 +6243,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[3]))
-						radeon_crtc_handle_flip(rdev, 3);
+						radeon_crtc_handle_vblank(rdev, 3);
 					rdev->irq.stat_regs.evergreen.disp_int_cont3 &= ~LB_D4_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D4 vblank\n");
 				}
@@ -6131,7 +6269,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[4]))
-						radeon_crtc_handle_flip(rdev, 4);
+						radeon_crtc_handle_vblank(rdev, 4);
 					rdev->irq.stat_regs.evergreen.disp_int_cont4 &= ~LB_D5_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D5 vblank\n");
 				}
@@ -6157,7 +6295,7 @@ restart_ih:
 						wake_up(&rdev->irq.vblank_queue);
 					}
 					if (atomic_read(&rdev->irq.pflip[5]))
-						radeon_crtc_handle_flip(rdev, 5);
+						radeon_crtc_handle_vblank(rdev, 5);
 					rdev->irq.stat_regs.evergreen.disp_int_cont5 &= ~LB_D6_VBLANK_INTERRUPT;
 					DRM_DEBUG("IH: D6 vblank\n");
 				}
@@ -6173,6 +6311,15 @@ restart_ih:
 				break;
 			}
 			break;
+		case 8: /* D1 page flip */
+		case 10: /* D2 page flip */
+		case 12: /* D3 page flip */
+		case 14: /* D4 page flip */
+		case 16: /* D5 page flip */
+		case 18: /* D6 page flip */
+			DRM_DEBUG("IH: D%d flip\n", ((src_id - 8) >> 1) + 1);
+			radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1);
+			break;
 		case 42: /* HPD hotplug */
 			switch (src_data) {
 			case 0:
@@ -6222,18 +6369,24 @@ restart_ih:
 				break;
 			}
 			break;
+		case 124: /* UVD */
+			DRM_DEBUG("IH: UVD int: 0x%08x\n", src_data);
+			radeon_fence_process(rdev, R600_RING_TYPE_UVD_INDEX);
+			break;
 		case 146:
 		case 147:
 			addr = RREG32(VM_CONTEXT1_PROTECTION_FAULT_ADDR);
 			status = RREG32(VM_CONTEXT1_PROTECTION_FAULT_STATUS);
+			/* reset addr and status */
+			WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1);
+			if (addr == 0x0 && status == 0x0)
+				break;
 			dev_err(rdev->dev, "GPU fault detected: %d 0x%08x\n", src_id, src_data);
 			dev_err(rdev->dev, "  VM_CONTEXT1_PROTECTION_FAULT_ADDR   0x%08X\n",
 				addr);
 			dev_err(rdev->dev, "  VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n",
 				status);
 			si_vm_decode_fault(rdev, status, addr);
-			/* reset addr and status */
-			WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1);
 			break;
 		case 176: /* RINGID0 CP_INT */
 			radeon_fence_process(rdev, RADEON_RING_TYPE_GFX_INDEX);
@@ -6324,21 +6477,14 @@ static int si_startup(struct radeon_device *rdev)
 
 	si_mc_program(rdev);
 
-	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
-	    !rdev->rlc_fw || !rdev->mc_fw) {
-		r = si_init_microcode(rdev);
+	if (!rdev->pm.dpm_enabled) {
+		r = si_mc_load_microcode(rdev);
 		if (r) {
-			DRM_ERROR("Failed to load firmware!\n");
+			DRM_ERROR("Failed to load MC firmware!\n");
 			return r;
 		}
 	}
 
-	r = si_mc_load_microcode(rdev);
-	if (r) {
-		DRM_ERROR("Failed to load MC firmware!\n");
-		return r;
-	}
-
 	r = si_pcie_gart_enable(rdev);
 	if (r)
 		return r;
@@ -6421,37 +6567,30 @@ static int si_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     CP_RB0_RPTR, CP_RB0_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP1_RPTR_OFFSET,
-			     CP_RB1_RPTR, CP_RB1_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP2_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP2_RPTR_OFFSET,
-			     CP_RB2_RPTR, CP_RB2_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR + DMA0_REGISTER_OFFSET,
-			     DMA_RB_WPTR + DMA0_REGISTER_OFFSET,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0));
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, CAYMAN_WB_DMA1_RPTR_OFFSET,
-			     DMA_RB_RPTR + DMA1_REGISTER_OFFSET,
-			     DMA_RB_WPTR + DMA1_REGISTER_OFFSET,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0));
 	if (r)
 		return r;
@@ -6471,7 +6610,6 @@ static int si_startup(struct radeon_device *rdev)
 		ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 		if (ring->ring_size) {
 			r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-					     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 					     RADEON_CP_PACKET2);
 			if (!r)
 				r = uvd_v1_0_init(rdev);
@@ -6513,6 +6651,9 @@ int si_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	si_init_golden_registers(rdev);
 
+	if (rdev->pm.pm_method == PM_METHOD_DPM)
+		radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = si_startup(rdev);
 	if (r) {
@@ -6527,6 +6668,7 @@ int si_resume(struct radeon_device *rdev)
 
 int si_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	dce6_audio_fini(rdev);
 	radeon_vm_manager_fini(rdev);
 	si_cp_enable(rdev, false);
@@ -6600,6 +6742,18 @@ int si_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
+	    !rdev->rlc_fw || !rdev->mc_fw) {
+		r = si_init_microcode(rdev);
+		if (r) {
+			DRM_ERROR("Failed to load firmware!\n");
+			return r;
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	ring->ring_obj = NULL;
 	r600_ring_init(rdev, ring, 1024 * 1024);
@@ -6666,6 +6820,7 @@ int si_init(struct radeon_device *rdev)
 
 void si_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	si_cp_fini(rdev);
 	cayman_dma_fini(rdev);
 	si_fini_pg(rdev);
diff --git a/drivers/gpu/drm/radeon/si_dma.c b/drivers/gpu/drm/radeon/si_dma.c
index 59be2cfcbb4..e24c94b6d14 100644
--- a/drivers/gpu/drm/radeon/si_dma.c
+++ b/drivers/gpu/drm/radeon/si_dma.c
@@ -49,11 +49,9 @@ bool si_dma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
 		mask = RADEON_RESET_DMA1;
 
 	if (!(reset_mask & mask)) {
-		radeon_ring_lockup_update(ring);
+		radeon_ring_lockup_update(rdev, ring);
 		return false;
 	}
-	/* force ring activities */
-	radeon_ring_force_activity(rdev, ring);
 	return radeon_ring_test_lockup(rdev, ring);
 }
 
@@ -81,7 +79,25 @@ void si_dma_vm_set_page(struct radeon_device *rdev,
 
 	trace_radeon_vm_set_page(pe, addr, count, incr, flags);
 
-	if (flags & R600_PTE_SYSTEM) {
+	if (flags == R600_PTE_GART) {
+		uint64_t src = rdev->gart.table_addr + (addr >> 12) * 8;
+		while (count) {
+			unsigned bytes = count * 8;
+			if (bytes > 0xFFFF8)
+				bytes = 0xFFFF8;
+
+			ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_COPY,
+							      1, 0, 0, bytes);
+			ib->ptr[ib->length_dw++] = lower_32_bits(pe);
+			ib->ptr[ib->length_dw++] = lower_32_bits(src);
+			ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff;
+			ib->ptr[ib->length_dw++] = upper_32_bits(src) & 0xff;
+
+			pe += bytes;
+			src += bytes;
+			count -= bytes / 8;
+		}
+	} else if (flags & R600_PTE_SYSTEM) {
 		while (count) {
 			ndw = count * 2;
 			if (ndw > 0xFFFFE)
@@ -204,8 +220,8 @@ int si_copy_dma(struct radeon_device *rdev,
 			cur_size_in_bytes = 0xFFFFF;
 		size_in_bytes -= cur_size_in_bytes;
 		radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_COPY, 1, 0, 0, cur_size_in_bytes));
-		radeon_ring_write(ring, dst_offset & 0xffffffff);
-		radeon_ring_write(ring, src_offset & 0xffffffff);
+		radeon_ring_write(ring, lower_32_bits(dst_offset));
+		radeon_ring_write(ring, lower_32_bits(src_offset));
 		radeon_ring_write(ring, upper_32_bits(dst_offset) & 0xff);
 		radeon_ring_write(ring, upper_32_bits(src_offset) & 0xff);
 		src_offset += cur_size_in_bytes;
@@ -215,6 +231,7 @@ int si_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index 0b00c790fb7..58918868f89 100644
--- a/drivers/gpu/drm/radeon/si_dpm.c
+++ b/drivers/gpu/drm/radeon/si_dpm.c
@@ -1738,6 +1738,8 @@ struct evergreen_power_info *evergreen_get_pi(struct radeon_device *rdev);
 struct ni_power_info *ni_get_pi(struct radeon_device *rdev);
 struct ni_ps *ni_get_ps(struct radeon_ps *rps);
 
+extern int si_mc_load_microcode(struct radeon_device *rdev);
+
 static int si_populate_voltage_value(struct radeon_device *rdev,
 				     const struct atom_voltage_table *table,
 				     u16 value, SISLANDS_SMC_VOLTAGE_VALUE *voltage);
@@ -1753,9 +1755,6 @@ static int si_calculate_sclk_params(struct radeon_device *rdev,
 				    u32 engine_clock,
 				    SISLANDS_SMC_SCLK_VALUE *sclk);
 
-extern void si_update_cg(struct radeon_device *rdev,
-			 u32 block, bool enable);
-
 static struct si_power_info *si_get_pi(struct radeon_device *rdev)
 {
         struct si_power_info *pi = rdev->pm.dpm.priv;
@@ -1949,6 +1948,10 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev)
 			si_pi->cac_weights = cac_weights_cape_verde_pro;
 			si_pi->dte_data = dte_data_cape_verde;
 			break;
+		case 0x682C:
+			si_pi->cac_weights = cac_weights_cape_verde_pro;
+			si_pi->dte_data = dte_data_sun_xt;
+			break;
 		case 0x6825:
 		case 0x6827:
 			si_pi->cac_weights = cac_weights_heathrow;
@@ -1972,10 +1975,9 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev)
 			si_pi->dte_data = dte_data_venus_xt;
 			break;
 		case 0x6823:
-			si_pi->cac_weights = cac_weights_chelsea_pro;
-			si_pi->dte_data = dte_data_venus_pro;
-			break;
 		case 0x682B:
+		case 0x6822:
+		case 0x682A:
 			si_pi->cac_weights = cac_weights_chelsea_pro;
 			si_pi->dte_data = dte_data_venus_pro;
 			break;
@@ -1989,6 +1991,7 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev)
 		case 0x6601:
 		case 0x6621:
 		case 0x6603:
+		case 0x6605:
 			si_pi->cac_weights = cac_weights_mars_pro;
 			si_pi->lcac_config = lcac_mars_pro;
 			si_pi->cac_override = cac_override_oland;
@@ -1999,6 +2002,7 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev)
 		case 0x6600:
 		case 0x6606:
 		case 0x6620:
+		case 0x6604:
 			si_pi->cac_weights = cac_weights_mars_xt;
 			si_pi->lcac_config = lcac_mars_pro;
 			si_pi->cac_override = cac_override_oland;
@@ -2007,6 +2011,8 @@ static void si_initialize_powertune_defaults(struct radeon_device *rdev)
 			update_dte_from_pl2 = true;
 			break;
 		case 0x6611:
+		case 0x6613:
+		case 0x6608:
 			si_pi->cac_weights = cac_weights_oland_pro;
 			si_pi->lcac_config = lcac_mars_pro;
 			si_pi->cac_override = cac_override_oland;
@@ -2396,7 +2402,7 @@ static int si_populate_sq_ramping_values(struct radeon_device *rdev,
 	if (SISLANDS_DPM2_SQ_RAMP_STI_SIZE > (STI_SIZE_MASK >> STI_SIZE_SHIFT))
 		enable_sq_ramping = false;
 
-	if (NISLANDS_DPM2_SQ_RAMP_LTI_RATIO <= (LTI_RATIO_MASK >> LTI_RATIO_SHIFT))
+	if (SISLANDS_DPM2_SQ_RAMP_LTI_RATIO > (LTI_RATIO_MASK >> LTI_RATIO_SHIFT))
 		enable_sq_ramping = false;
 
 	for (i = 0; i < state->performance_level_count; i++) {
@@ -3591,10 +3597,9 @@ static void si_program_display_gap(struct radeon_device *rdev)
 
 	/* Setting this to false forces the performance state to low if the crtcs are disabled.
 	 * This can be a problem on PowerXpress systems or if you want to use the card
-	 * for offscreen rendering or compute if there are no crtcs enabled.  Set it to
-	 * true for now so that performance scales even if the displays are off.
+	 * for offscreen rendering or compute if there are no crtcs enabled.
 	 */
-	si_notify_smc_display_change(rdev, true /*rdev->pm.dpm.new_active_crtc_count > 0*/);
+	si_notify_smc_display_change(rdev, rdev->pm.dpm.new_active_crtc_count > 0);
 }
 
 static void si_enable_spread_spectrum(struct radeon_device *rdev, bool enable)
@@ -5414,7 +5419,7 @@ static void si_populate_mc_reg_addresses(struct radeon_device *rdev,
 
 	for (i = 0, j = 0; j < si_pi->mc_reg_table.last; j++) {
 		if (si_pi->mc_reg_table.valid_flag & (1 << j)) {
-			if (i >= SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE)
+			if (i >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE)
 				break;
 			mc_reg_table->address[i].s0 =
 				cpu_to_be16(si_pi->mc_reg_table.mc_reg_address[j].s0);
@@ -5754,6 +5759,11 @@ static void si_set_pcie_lane_width_in_smc(struct radeon_device *rdev,
 
 void si_dpm_setup_asic(struct radeon_device *rdev)
 {
+	int r;
+
+	r = si_mc_load_microcode(rdev);
+	if (r)
+		DRM_ERROR("Failed to load MC firmware!\n");
 	rv770_get_memory_type(rdev);
 	si_read_clock_registers(rdev);
 	si_enable_acpi_power_management(rdev);
@@ -5791,13 +5801,6 @@ int si_dpm_enable(struct radeon_device *rdev)
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
 	int ret;
 
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), false);
-
 	if (si_is_smc_running(rdev))
 		return -EINVAL;
 	if (pi->voltage_control)
@@ -5900,6 +5903,17 @@ int si_dpm_enable(struct radeon_device *rdev)
 	si_enable_sclk_control(rdev, true);
 	si_start_dpm(rdev);
 
+	si_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
+
+	ni_update_current_ps(rdev, boot_ps);
+
+	return 0;
+}
+
+int si_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		PPSMC_Result result;
@@ -5915,17 +5929,6 @@ int si_dpm_enable(struct radeon_device *rdev)
 			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
 	}
 
-	si_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
-
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), true);
-
-	ni_update_current_ps(rdev, boot_ps);
-
 	return 0;
 }
 
@@ -5934,13 +5937,6 @@ void si_dpm_disable(struct radeon_device *rdev)
 	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
 
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), false);
-
 	if (!si_is_smc_running(rdev))
 		return;
 	si_disable_ulv(rdev);
@@ -6005,13 +6001,6 @@ int si_dpm_set_power_state(struct radeon_device *rdev)
 	struct radeon_ps *old_ps = &eg_pi->current_rps;
 	int ret;
 
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), false);
-
 	ret = si_disable_ulv(rdev);
 	if (ret) {
 		DRM_ERROR("si_disable_ulv failed\n");
@@ -6104,13 +6093,6 @@ int si_dpm_set_power_state(struct radeon_device *rdev)
 		return ret;
 	}
 
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), true);
-
 	return 0;
 }
 
@@ -6296,9 +6278,6 @@ static int si_parse_power_table(struct radeon_device *rdev)
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
-	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
-	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
-	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
 	for (i = 0; i < state_array->ucNumEntries; i++) {
 		u8 *idx;
 		power_state = (union pplib_power_state *)power_state_offset;
@@ -6375,6 +6354,10 @@ int si_dpm_init(struct radeon_device *rdev)
 	pi->min_vddc_in_table = 0;
 	pi->max_vddc_in_table = 0;
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = si_parse_power_table(rdev);
 	if (ret)
 		return ret;
@@ -6497,7 +6480,8 @@ void si_dpm_fini(struct radeon_device *rdev)
 void si_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
 						    struct seq_file *m)
 {
-	struct radeon_ps *rps = rdev->pm.dpm.current_ps;
+	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	struct radeon_ps *rps = &eg_pi->current_rps;
 	struct ni_ps *ps = ni_get_ps(rps);
 	struct rv7xx_pl *pl;
 	u32 current_index =
diff --git a/drivers/gpu/drm/radeon/si_smc.c b/drivers/gpu/drm/radeon/si_smc.c
index d422a1cbf72..e80efcf0c23 100644
--- a/drivers/gpu/drm/radeon/si_smc.c
+++ b/drivers/gpu/drm/radeon/si_smc.c
@@ -28,6 +28,7 @@
 #include "sid.h"
 #include "ppsmc.h"
 #include "radeon_ucode.h"
+#include "sislands_smc.h"
 
 static int si_set_smc_sram_address(struct radeon_device *rdev,
 				   u32 smc_address, u32 limit)
diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h
index b322acc4809..fd414d34d88 100644
--- a/drivers/gpu/drm/radeon/sid.h
+++ b/drivers/gpu/drm/radeon/sid.h
@@ -94,6 +94,8 @@
 #define	CG_SPLL_FUNC_CNTL_2				0x604
 #define		SCLK_MUX_SEL(x)				((x) << 0)
 #define		SCLK_MUX_SEL_MASK			(0x1ff << 0)
+#define		SPLL_CTLREQ_CHG				(1 << 23)
+#define		SCLK_MUX_UPDATE				(1 << 26)
 #define	CG_SPLL_FUNC_CNTL_3				0x608
 #define		SPLL_FB_DIV(x)				((x) << 0)
 #define		SPLL_FB_DIV_MASK			(0x3ffffff << 0)
@@ -101,9 +103,12 @@
 #define		SPLL_DITHEN				(1 << 28)
 #define	CG_SPLL_FUNC_CNTL_4				0x60c
 
+#define	SPLL_STATUS					0x614
+#define		SPLL_CHG_STATUS				(1 << 1)
 #define	SPLL_CNTL_MODE					0x618
-#	define SPLL_REFCLK_SEL(x)			((x) << 8)
-#	define SPLL_REFCLK_SEL_MASK			0xFF00
+#define		SPLL_SW_DIR_CONTROL			(1 << 0)
+#	define SPLL_REFCLK_SEL(x)			((x) << 26)
+#	define SPLL_REFCLK_SEL_MASK			(3 << 26)
 
 #define	CG_SPLL_SPREAD_SPECTRUM				0x620
 #define		SSEN					(1 << 0)
@@ -357,6 +362,7 @@
 #define		READ_PROTECTION_FAULT_ENABLE_DEFAULT		(1 << 16)
 #define		WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT		(1 << 18)
 #define		WRITE_PROTECTION_FAULT_ENABLE_DEFAULT		(1 << 19)
+#define		PAGE_TABLE_BLOCK_SIZE(x)			(((x) & 0xF) << 24)
 #define VM_CONTEXT1_CNTL				0x1414
 #define VM_CONTEXT0_CNTL2				0x1430
 #define VM_CONTEXT1_CNTL2				0x1434
@@ -559,6 +565,8 @@
 #       define MRDCK0_BYPASS                            (1 << 24)
 #       define MRDCK1_BYPASS                            (1 << 25)
 
+#define	MPLL_CNTL_MODE					0x2bb0
+#       define MPLL_MCLK_SEL                            (1 << 11)
 #define	MPLL_FUNC_CNTL					0x2bb4
 #define		BWCTRL(x)				((x) << 20)
 #define		BWCTRL_MASK				(0xff << 20)
@@ -815,7 +823,7 @@
 #       define GRPH_PFLIP_INT_MASK                      (1 << 0)
 #       define GRPH_PFLIP_INT_TYPE                      (1 << 8)
 
-#define	DACA_AUTODETECT_INT_CONTROL			0x66c8
+#define	DAC_AUTODETECT_INT_CONTROL			0x67c8
 
 #define DC_HPD1_INT_STATUS                              0x601c
 #define DC_HPD2_INT_STATUS                              0x6028
@@ -1791,4 +1799,51 @@
 #define	DMA_PACKET_CONSTANT_FILL			  0xd
 #define	DMA_PACKET_NOP					  0xf
 
+#define VCE_STATUS					0x20004
+#define VCE_VCPU_CNTL					0x20014
+#define		VCE_CLK_EN				(1 << 0)
+#define VCE_VCPU_CACHE_OFFSET0				0x20024
+#define VCE_VCPU_CACHE_SIZE0				0x20028
+#define VCE_VCPU_CACHE_OFFSET1				0x2002c
+#define VCE_VCPU_CACHE_SIZE1				0x20030
+#define VCE_VCPU_CACHE_OFFSET2				0x20034
+#define VCE_VCPU_CACHE_SIZE2				0x20038
+#define VCE_SOFT_RESET					0x20120
+#define 	VCE_ECPU_SOFT_RESET			(1 << 0)
+#define 	VCE_FME_SOFT_RESET			(1 << 2)
+#define VCE_RB_BASE_LO2					0x2016c
+#define VCE_RB_BASE_HI2					0x20170
+#define VCE_RB_SIZE2					0x20174
+#define VCE_RB_RPTR2					0x20178
+#define VCE_RB_WPTR2					0x2017c
+#define VCE_RB_BASE_LO					0x20180
+#define VCE_RB_BASE_HI					0x20184
+#define VCE_RB_SIZE					0x20188
+#define VCE_RB_RPTR					0x2018c
+#define VCE_RB_WPTR					0x20190
+#define VCE_CLOCK_GATING_A				0x202f8
+#define VCE_CLOCK_GATING_B				0x202fc
+#define VCE_UENC_CLOCK_GATING				0x205bc
+#define VCE_UENC_REG_CLOCK_GATING			0x205c0
+#define VCE_FW_REG_STATUS				0x20e10
+#	define VCE_FW_REG_STATUS_BUSY			(1 << 0)
+#	define VCE_FW_REG_STATUS_PASS			(1 << 3)
+#	define VCE_FW_REG_STATUS_DONE			(1 << 11)
+#define VCE_LMI_FW_START_KEYSEL				0x20e18
+#define VCE_LMI_FW_PERIODIC_CTRL			0x20e20
+#define VCE_LMI_CTRL2					0x20e74
+#define VCE_LMI_CTRL					0x20e98
+#define VCE_LMI_VM_CTRL					0x20ea0
+#define VCE_LMI_SWAP_CNTL				0x20eb4
+#define VCE_LMI_SWAP_CNTL1				0x20eb8
+#define VCE_LMI_CACHE_CTRL				0x20ef4
+
+#define VCE_CMD_NO_OP					0x00000000
+#define VCE_CMD_END					0x00000001
+#define VCE_CMD_IB					0x00000002
+#define VCE_CMD_FENCE					0x00000003
+#define VCE_CMD_TRAP					0x00000004
+#define VCE_CMD_IB_AUTO					0x00000005
+#define VCE_CMD_SEMAPHORE				0x00000006
+
 #endif
diff --git a/drivers/gpu/drm/radeon/sislands_smc.h b/drivers/gpu/drm/radeon/sislands_smc.h
index 5578e983702..10e945a4947 100644
--- a/drivers/gpu/drm/radeon/sislands_smc.h
+++ b/drivers/gpu/drm/radeon/sislands_smc.h
@@ -374,8 +374,6 @@ typedef struct Smc_SIslands_DTE_Configuration Smc_SIslands_DTE_Configuration;
 
 #pragma pack(pop)
 
-int si_set_smc_sram_address(struct radeon_device *rdev,
-			    u32 smc_address, u32 limit);
 int si_copy_bytes_to_smc(struct radeon_device *rdev,
 			 u32 smc_start_address,
 			 const u8 *src, u32 byte_count, u32 limit);
diff --git a/drivers/gpu/drm/radeon/sumo_dpm.c b/drivers/gpu/drm/radeon/sumo_dpm.c
index 96ea6db8bf5..3f0e8d7b8db 100644
--- a/drivers/gpu/drm/radeon/sumo_dpm.c
+++ b/drivers/gpu/drm/radeon/sumo_dpm.c
@@ -71,7 +71,7 @@ static const u32 sumo_dtc[SUMO_PM_NUMBER_OF_TC] =
 	SUMO_DTC_DFLT_14,
 };
 
-struct sumo_ps *sumo_get_ps(struct radeon_ps *rps)
+static struct sumo_ps *sumo_get_ps(struct radeon_ps *rps)
 {
 	struct sumo_ps *ps = rps->ps_priv;
 
@@ -1202,14 +1202,10 @@ static void sumo_update_requested_ps(struct radeon_device *rdev,
 int sumo_dpm_enable(struct radeon_device *rdev)
 {
 	struct sumo_power_info *pi = sumo_get_pi(rdev);
-	int ret;
 
 	if (sumo_dpm_enabled(rdev))
 		return -EINVAL;
 
-	ret = sumo_enable_clock_power_gating(rdev);
-	if (ret)
-		return ret;
 	sumo_program_bootup_state(rdev);
 	sumo_init_bsp(rdev);
 	sumo_reset_am(rdev);
@@ -1233,6 +1229,19 @@ int sumo_dpm_enable(struct radeon_device *rdev)
 	if (pi->enable_boost)
 		sumo_enable_boost_timer(rdev);
 
+	sumo_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
+
+	return 0;
+}
+
+int sumo_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
+	ret = sumo_enable_clock_power_gating(rdev);
+	if (ret)
+		return ret;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		ret = sumo_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
@@ -1242,8 +1251,6 @@ int sumo_dpm_enable(struct radeon_device *rdev)
 		radeon_irq_set(rdev);
 	}
 
-	sumo_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
-
 	return 0;
 }
 
@@ -1477,9 +1484,6 @@ static int sumo_parse_power_table(struct radeon_device *rdev)
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
-	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
-	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
-	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
 	for (i = 0; i < state_array->ucNumEntries; i++) {
 		u8 *idx;
 		power_state = (union pplib_power_state *)power_state_offset;
@@ -1765,6 +1769,10 @@ int sumo_dpm_init(struct radeon_device *rdev)
 
 	sumo_construct_boot_and_acpi_state(rdev);
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = sumo_parse_power_table(rdev);
 	if (ret)
 		return ret;
@@ -1800,7 +1808,7 @@ void sumo_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev
 						      struct seq_file *m)
 {
 	struct sumo_power_info *pi = sumo_get_pi(rdev);
-	struct radeon_ps *rps = rdev->pm.dpm.current_ps;
+	struct radeon_ps *rps = &pi->current_rps;
 	struct sumo_ps *ps = sumo_get_ps(rps);
 	struct sumo_pl *pl;
 	u32 current_index =
diff --git a/drivers/gpu/drm/radeon/sumo_smc.c b/drivers/gpu/drm/radeon/sumo_smc.c
index 18abba5b581..fb081d2ae37 100644
--- a/drivers/gpu/drm/radeon/sumo_smc.c
+++ b/drivers/gpu/drm/radeon/sumo_smc.c
@@ -31,7 +31,6 @@
 #define SUMO_SMU_SERVICE_ROUTINE_ALTVDDNB_NOTIFY  27
 #define SUMO_SMU_SERVICE_ROUTINE_GFX_SRV_ID_20  20
 
-struct sumo_ps *sumo_get_ps(struct radeon_ps *rps);
 struct sumo_power_info *sumo_get_pi(struct radeon_device *rdev);
 
 static void sumo_send_msg_to_smu(struct radeon_device *rdev, u32 id)
diff --git a/drivers/gpu/drm/radeon/trinity_dpm.c b/drivers/gpu/drm/radeon/trinity_dpm.c
index d700698a1f2..32e50be9c4a 100644
--- a/drivers/gpu/drm/radeon/trinity_dpm.c
+++ b/drivers/gpu/drm/radeon/trinity_dpm.c
@@ -342,14 +342,14 @@ static void trinity_apply_state_adjust_rules(struct radeon_device *rdev,
 					     struct radeon_ps *new_rps,
 					     struct radeon_ps *old_rps);
 
-struct trinity_ps *trinity_get_ps(struct radeon_ps *rps)
+static struct trinity_ps *trinity_get_ps(struct radeon_ps *rps)
 {
 	struct trinity_ps *ps = rps->ps_priv;
 
 	return ps;
 }
 
-struct trinity_power_info *trinity_get_pi(struct radeon_device *rdev)
+static struct trinity_power_info *trinity_get_pi(struct radeon_device *rdev)
 {
 	struct trinity_power_info *pi = rdev->pm.dpm.priv;
 
@@ -1082,7 +1082,6 @@ void trinity_dpm_enable_bapm(struct radeon_device *rdev, bool enable)
 int trinity_dpm_enable(struct radeon_device *rdev)
 {
 	struct trinity_power_info *pi = trinity_get_pi(rdev);
-	int ret;
 
 	trinity_acquire_mutex(rdev);
 
@@ -1091,7 +1090,6 @@ int trinity_dpm_enable(struct radeon_device *rdev)
 		return -EINVAL;
 	}
 
-	trinity_enable_clock_power_gating(rdev);
 	trinity_program_bootup_state(rdev);
 	sumo_program_vc(rdev, 0x00C00033);
 	trinity_start_am(rdev);
@@ -1105,6 +1103,18 @@ int trinity_dpm_enable(struct radeon_device *rdev)
 	trinity_dpm_bapm_enable(rdev, false);
 	trinity_release_mutex(rdev);
 
+	trinity_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
+
+	return 0;
+}
+
+int trinity_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
+	trinity_acquire_mutex(rdev);
+	trinity_enable_clock_power_gating(rdev);
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		ret = trinity_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
@@ -1115,8 +1125,7 @@ int trinity_dpm_enable(struct radeon_device *rdev)
 		rdev->irq.dpm_thermal = true;
 		radeon_irq_set(rdev);
 	}
-
-	trinity_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
+	trinity_release_mutex(rdev);
 
 	return 0;
 }
@@ -1685,9 +1694,6 @@ static int trinity_parse_power_table(struct radeon_device *rdev)
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
-	rdev->pm.dpm.platform_caps = le32_to_cpu(power_info->pplib.ulPlatformCaps);
-	rdev->pm.dpm.backbias_response_time = le16_to_cpu(power_info->pplib.usBackbiasTime);
-	rdev->pm.dpm.voltage_response_time = le16_to_cpu(power_info->pplib.usVoltageTime);
 	for (i = 0; i < state_array->ucNumEntries; i++) {
 		u8 *idx;
 		power_state = (union pplib_power_state *)power_state_offset;
@@ -1868,7 +1874,16 @@ int trinity_dpm_init(struct radeon_device *rdev)
 	for (i = 0; i < SUMO_MAX_HARDWARE_POWERLEVELS; i++)
 		pi->at[i] = TRINITY_AT_DFLT;
 
-	pi->enable_bapm = false;
+	/* There are stability issues reported on with
+	 * bapm enabled when switching between AC and battery
+	 * power.  At the same time, some MSI boards hang
+	 * if it's not enabled and dpm is enabled.  Just enable
+	 * it for MSI boards right now.
+	 */
+	if (rdev->pdev->subsystem_vendor == 0x1462)
+		pi->enable_bapm = true;
+	else
+		pi->enable_bapm = false;
 	pi->enable_nbps_policy = true;
 	pi->enable_sclk_ds = true;
 	pi->enable_gfx_power_gating = true;
@@ -1886,6 +1901,10 @@ int trinity_dpm_init(struct radeon_device *rdev)
 
 	trinity_construct_boot_state(rdev);
 
+	ret = r600_get_platform_caps(rdev);
+	if (ret)
+		return ret;
+
 	ret = trinity_parse_power_table(rdev);
 	if (ret)
 		return ret;
@@ -1917,7 +1936,8 @@ void trinity_dpm_print_power_state(struct radeon_device *rdev,
 void trinity_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
 							 struct seq_file *m)
 {
-	struct radeon_ps *rps = rdev->pm.dpm.current_ps;
+	struct trinity_power_info *pi = trinity_get_pi(rdev);
+	struct radeon_ps *rps = &pi->current_rps;
 	struct trinity_ps *ps = trinity_get_ps(rps);
 	struct trinity_pl *pl;
 	u32 current_index =
diff --git a/drivers/gpu/drm/radeon/trinity_smc.c b/drivers/gpu/drm/radeon/trinity_smc.c
index 9672bcbc731..99dd0455334 100644
--- a/drivers/gpu/drm/radeon/trinity_smc.c
+++ b/drivers/gpu/drm/radeon/trinity_smc.c
@@ -27,9 +27,6 @@
 #include "trinity_dpm.h"
 #include "ppsmc.h"
 
-struct trinity_ps *trinity_get_ps(struct radeon_ps *rps);
-struct trinity_power_info *trinity_get_pi(struct radeon_device *rdev);
-
 static int trinity_notify_message_to_smu(struct radeon_device *rdev, u32 id)
 {
 	int i;
diff --git a/drivers/gpu/drm/radeon/uvd_v1_0.c b/drivers/gpu/drm/radeon/uvd_v1_0.c
index d4a68af1a27..be42c812520 100644
--- a/drivers/gpu/drm/radeon/uvd_v1_0.c
+++ b/drivers/gpu/drm/radeon/uvd_v1_0.c
@@ -83,7 +83,10 @@ int uvd_v1_0_init(struct radeon_device *rdev)
 	int r;
 
 	/* raise clocks while booting up the VCPU */
-	radeon_set_uvd_clocks(rdev, 53300, 40000);
+	if (rdev->family < CHIP_RV740)
+		radeon_set_uvd_clocks(rdev, 10000, 10000);
+	else
+		radeon_set_uvd_clocks(rdev, 53300, 40000);
 
 	r = uvd_v1_0_start(rdev);
 	if (r)
@@ -262,7 +265,7 @@ int uvd_v1_0_start(struct radeon_device *rdev)
 	/* Initialize the ring buffer's read and write pointers */
 	WREG32(UVD_RBC_RB_RPTR, 0x0);
 
-	ring->wptr = ring->rptr = RREG32(UVD_RBC_RB_RPTR);
+	ring->wptr = RREG32(UVD_RBC_RB_RPTR);
 	WREG32(UVD_RBC_RB_WPTR, ring->wptr);
 
 	/* set the ring address */
@@ -407,7 +410,10 @@ int uvd_v1_0_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 	struct radeon_fence *fence = NULL;
 	int r;
 
-	r = radeon_set_uvd_clocks(rdev, 53300, 40000);
+	if (rdev->family < CHIP_RV740)
+		r = radeon_set_uvd_clocks(rdev, 10000, 10000);
+	else
+		r = radeon_set_uvd_clocks(rdev, 53300, 40000);
 	if (r) {
 		DRM_ERROR("radeon: failed to raise UVD clocks (%d).\n", r);
 		return r;
diff --git a/drivers/gpu/drm/radeon/uvd_v2_2.c b/drivers/gpu/drm/radeon/uvd_v2_2.c
index b19ef495108..8bfdadd5659 100644
--- a/drivers/gpu/drm/radeon/uvd_v2_2.c
+++ b/drivers/gpu/drm/radeon/uvd_v2_2.c
@@ -45,7 +45,7 @@ void uvd_v2_2_fence_emit(struct radeon_device *rdev,
 	radeon_ring_write(ring, PACKET0(UVD_CONTEXT_ID, 0));
 	radeon_ring_write(ring, fence->seq);
 	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA0, 0));
-	radeon_ring_write(ring, addr & 0xffffffff);
+	radeon_ring_write(ring, lower_32_bits(addr));
 	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_DATA1, 0));
 	radeon_ring_write(ring, upper_32_bits(addr) & 0xff);
 	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0));
@@ -57,7 +57,6 @@ void uvd_v2_2_fence_emit(struct radeon_device *rdev,
 	radeon_ring_write(ring, 0);
 	radeon_ring_write(ring, PACKET0(UVD_GPCOM_VCPU_CMD, 0));
 	radeon_ring_write(ring, 2);
-	return;
 }
 
 /**
@@ -153,6 +152,7 @@ int uvd_v2_2_resume(struct radeon_device *rdev)
 		chip_id = 0x01000015;
 		break;
 	case CHIP_PITCAIRN:
+	case CHIP_OLAND:
 		chip_id = 0x01000016;
 		break;
 	case CHIP_ARUBA:
diff --git a/drivers/gpu/drm/radeon/vce_v1_0.c b/drivers/gpu/drm/radeon/vce_v1_0.c
new file mode 100644
index 00000000000..b44d9c842f7
--- /dev/null
+++ b/drivers/gpu/drm/radeon/vce_v1_0.c
@@ -0,0 +1,187 @@
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * Authors: Christian König <christian.koenig@amd.com>
+ */
+
+#include <linux/firmware.h>
+#include <drm/drmP.h>
+#include "radeon.h"
+#include "radeon_asic.h"
+#include "sid.h"
+
+/**
+ * vce_v1_0_get_rptr - get read pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon_ring pointer
+ *
+ * Returns the current hardware read pointer
+ */
+uint32_t vce_v1_0_get_rptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	if (ring->idx == TN_RING_TYPE_VCE1_INDEX)
+		return RREG32(VCE_RB_RPTR);
+	else
+		return RREG32(VCE_RB_RPTR2);
+}
+
+/**
+ * vce_v1_0_get_wptr - get write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon_ring pointer
+ *
+ * Returns the current hardware write pointer
+ */
+uint32_t vce_v1_0_get_wptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	if (ring->idx == TN_RING_TYPE_VCE1_INDEX)
+		return RREG32(VCE_RB_WPTR);
+	else
+		return RREG32(VCE_RB_WPTR2);
+}
+
+/**
+ * vce_v1_0_set_wptr - set write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon_ring pointer
+ *
+ * Commits the write pointer to the hardware
+ */
+void vce_v1_0_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring)
+{
+	if (ring->idx == TN_RING_TYPE_VCE1_INDEX)
+		WREG32(VCE_RB_WPTR, ring->wptr);
+	else
+		WREG32(VCE_RB_WPTR2, ring->wptr);
+}
+
+/**
+ * vce_v1_0_start - start VCE block
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Setup and start the VCE block
+ */
+int vce_v1_0_start(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int i, j, r;
+
+	/* set BUSY flag */
+	WREG32_P(VCE_STATUS, 1, ~1);
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+	WREG32(VCE_RB_RPTR, ring->wptr);
+	WREG32(VCE_RB_WPTR, ring->wptr);
+	WREG32(VCE_RB_BASE_LO, ring->gpu_addr);
+	WREG32(VCE_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
+	WREG32(VCE_RB_SIZE, ring->ring_size / 4);
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+	WREG32(VCE_RB_RPTR2, ring->wptr);
+	WREG32(VCE_RB_WPTR2, ring->wptr);
+	WREG32(VCE_RB_BASE_LO2, ring->gpu_addr);
+	WREG32(VCE_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
+	WREG32(VCE_RB_SIZE2, ring->ring_size / 4);
+
+	WREG32_P(VCE_VCPU_CNTL, VCE_CLK_EN, ~VCE_CLK_EN);
+
+	WREG32_P(VCE_SOFT_RESET,
+		 VCE_ECPU_SOFT_RESET |
+		 VCE_FME_SOFT_RESET, ~(
+		 VCE_ECPU_SOFT_RESET |
+		 VCE_FME_SOFT_RESET));
+
+	mdelay(100);
+
+	WREG32_P(VCE_SOFT_RESET, 0, ~(
+		 VCE_ECPU_SOFT_RESET |
+		 VCE_FME_SOFT_RESET));
+
+	for (i = 0; i < 10; ++i) {
+		uint32_t status;
+		for (j = 0; j < 100; ++j) {
+			status = RREG32(VCE_STATUS);
+			if (status & 2)
+				break;
+			mdelay(10);
+		}
+		r = 0;
+		if (status & 2)
+			break;
+
+		DRM_ERROR("VCE not responding, trying to reset the ECPU!!!\n");
+		WREG32_P(VCE_SOFT_RESET, VCE_ECPU_SOFT_RESET, ~VCE_ECPU_SOFT_RESET);
+		mdelay(10);
+		WREG32_P(VCE_SOFT_RESET, 0, ~VCE_ECPU_SOFT_RESET);
+		mdelay(10);
+		r = -1;
+	}
+
+	/* clear BUSY flag */
+	WREG32_P(VCE_STATUS, 0, ~1);
+
+	if (r) {
+		DRM_ERROR("VCE not responding, giving up!!!\n");
+		return r;
+	}
+
+	return 0;
+}
+
+int vce_v1_0_init(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	r = vce_v1_0_start(rdev);
+	if (r)
+		return r;
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+	ring->ready = true;
+	r = radeon_ring_test(rdev, TN_RING_TYPE_VCE1_INDEX, ring);
+	if (r) {
+		ring->ready = false;
+		return r;
+	}
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+	ring->ready = true;
+	r = radeon_ring_test(rdev, TN_RING_TYPE_VCE2_INDEX, ring);
+	if (r) {
+		ring->ready = false;
+		return r;
+	}
+
+	DRM_INFO("VCE initialized successfully.\n");
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/radeon/vce_v2_0.c b/drivers/gpu/drm/radeon/vce_v2_0.c
new file mode 100644
index 00000000000..1ac7bb825a1
--- /dev/null
+++ b/drivers/gpu/drm/radeon/vce_v2_0.c
@@ -0,0 +1,181 @@
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * Authors: Christian König <christian.koenig@amd.com>
+ */
+
+#include <linux/firmware.h>
+#include <drm/drmP.h>
+#include "radeon.h"
+#include "radeon_asic.h"
+#include "cikd.h"
+
+static void vce_v2_0_set_sw_cg(struct radeon_device *rdev, bool gated)
+{
+	u32 tmp;
+
+	if (gated) {
+		tmp = RREG32(VCE_CLOCK_GATING_B);
+		tmp |= 0xe70000;
+		WREG32(VCE_CLOCK_GATING_B, tmp);
+
+		tmp = RREG32(VCE_UENC_CLOCK_GATING);
+		tmp |= 0xff000000;
+		WREG32(VCE_UENC_CLOCK_GATING, tmp);
+
+		tmp = RREG32(VCE_UENC_REG_CLOCK_GATING);
+		tmp &= ~0x3fc;
+		WREG32(VCE_UENC_REG_CLOCK_GATING, tmp);
+
+		WREG32(VCE_CGTT_CLK_OVERRIDE, 0);
+    } else {
+		tmp = RREG32(VCE_CLOCK_GATING_B);
+		tmp |= 0xe7;
+		tmp &= ~0xe70000;
+		WREG32(VCE_CLOCK_GATING_B, tmp);
+
+		tmp = RREG32(VCE_UENC_CLOCK_GATING);
+		tmp |= 0x1fe000;
+		tmp &= ~0xff000000;
+		WREG32(VCE_UENC_CLOCK_GATING, tmp);
+
+		tmp = RREG32(VCE_UENC_REG_CLOCK_GATING);
+		tmp |= 0x3fc;
+		WREG32(VCE_UENC_REG_CLOCK_GATING, tmp);
+	}
+}
+
+static void vce_v2_0_set_dyn_cg(struct radeon_device *rdev, bool gated)
+{
+	u32 orig, tmp;
+
+	tmp = RREG32(VCE_CLOCK_GATING_B);
+	tmp &= ~0x00060006;
+	if (gated) {
+		tmp |= 0xe10000;
+	} else {
+		tmp |= 0xe1;
+		tmp &= ~0xe10000;
+	}
+	WREG32(VCE_CLOCK_GATING_B, tmp);
+
+	orig = tmp = RREG32(VCE_UENC_CLOCK_GATING);
+	tmp &= ~0x1fe000;
+	tmp &= ~0xff000000;
+	if (tmp != orig)
+		WREG32(VCE_UENC_CLOCK_GATING, tmp);
+
+	orig = tmp = RREG32(VCE_UENC_REG_CLOCK_GATING);
+	tmp &= ~0x3fc;
+	if (tmp != orig)
+		WREG32(VCE_UENC_REG_CLOCK_GATING, tmp);
+
+	if (gated)
+		WREG32(VCE_CGTT_CLK_OVERRIDE, 0);
+}
+
+static void vce_v2_0_disable_cg(struct radeon_device *rdev)
+{
+	WREG32(VCE_CGTT_CLK_OVERRIDE, 7);
+}
+
+void vce_v2_0_enable_mgcg(struct radeon_device *rdev, bool enable)
+{
+	bool sw_cg = false;
+
+	if (enable && (rdev->cg_flags & RADEON_CG_SUPPORT_VCE_MGCG)) {
+		if (sw_cg)
+			vce_v2_0_set_sw_cg(rdev, true);
+		else
+			vce_v2_0_set_dyn_cg(rdev, true);
+	} else {
+		vce_v2_0_disable_cg(rdev);
+
+		if (sw_cg)
+			vce_v2_0_set_sw_cg(rdev, false);
+		else
+			vce_v2_0_set_dyn_cg(rdev, false);
+	}
+}
+
+static void vce_v2_0_init_cg(struct radeon_device *rdev)
+{
+	u32 tmp;
+
+	tmp = RREG32(VCE_CLOCK_GATING_A);
+	tmp &= ~(CGC_CLK_GATE_DLY_TIMER_MASK | CGC_CLK_GATER_OFF_DLY_TIMER_MASK);
+	tmp |= (CGC_CLK_GATE_DLY_TIMER(0) | CGC_CLK_GATER_OFF_DLY_TIMER(4));
+	tmp |= CGC_UENC_WAIT_AWAKE;
+	WREG32(VCE_CLOCK_GATING_A, tmp);
+
+	tmp = RREG32(VCE_UENC_CLOCK_GATING);
+	tmp &= ~(CLOCK_ON_DELAY_MASK | CLOCK_OFF_DELAY_MASK);
+	tmp |= (CLOCK_ON_DELAY(0) | CLOCK_OFF_DELAY(4));
+	WREG32(VCE_UENC_CLOCK_GATING, tmp);
+
+	tmp = RREG32(VCE_CLOCK_GATING_B);
+	tmp |= 0x10;
+	tmp &= ~0x100000;
+	WREG32(VCE_CLOCK_GATING_B, tmp);
+}
+
+int vce_v2_0_resume(struct radeon_device *rdev)
+{
+	uint64_t addr = rdev->vce.gpu_addr;
+	uint32_t size;
+
+	WREG32_P(VCE_CLOCK_GATING_A, 0, ~(1 << 16));
+	WREG32_P(VCE_UENC_CLOCK_GATING, 0x1FF000, ~0xFF9FF000);
+	WREG32_P(VCE_UENC_REG_CLOCK_GATING, 0x3F, ~0x3F);
+	WREG32(VCE_CLOCK_GATING_B, 0xf7);
+
+	WREG32(VCE_LMI_CTRL, 0x00398000);
+	WREG32_P(VCE_LMI_CACHE_CTRL, 0x0, ~0x1);
+	WREG32(VCE_LMI_SWAP_CNTL, 0);
+	WREG32(VCE_LMI_SWAP_CNTL1, 0);
+	WREG32(VCE_LMI_VM_CTRL, 0);
+
+	size = RADEON_GPU_PAGE_ALIGN(rdev->vce_fw->size);
+	WREG32(VCE_VCPU_CACHE_OFFSET0, addr & 0x7fffffff);
+	WREG32(VCE_VCPU_CACHE_SIZE0, size);
+
+	addr += size;
+	size = RADEON_VCE_STACK_SIZE;
+	WREG32(VCE_VCPU_CACHE_OFFSET1, addr & 0x7fffffff);
+	WREG32(VCE_VCPU_CACHE_SIZE1, size);
+
+	addr += size;
+	size = RADEON_VCE_HEAP_SIZE;
+	WREG32(VCE_VCPU_CACHE_OFFSET2, addr & 0x7fffffff);
+	WREG32(VCE_VCPU_CACHE_SIZE2, size);
+
+	WREG32_P(VCE_LMI_CTRL2, 0x0, ~0x100);
+
+	WREG32_P(VCE_SYS_INT_EN, VCE_SYS_INT_TRAP_INTERRUPT_EN,
+		 ~VCE_SYS_INT_TRAP_INTERRUPT_EN);
+
+	vce_v2_0_init_cg(rdev);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/rcar-du/Kconfig b/drivers/gpu/drm/rcar-du/Kconfig
index d8e835ac2c5..2e3d7b5b0ad 100644
--- a/drivers/gpu/drm/rcar-du/Kconfig
+++ b/drivers/gpu/drm/rcar-du/Kconfig
@@ -1,6 +1,7 @@
 config DRM_RCAR_DU
 	tristate "DRM Support for R-Car Display Unit"
 	depends on DRM && ARM
+	depends on ARCH_SHMOBILE || COMPILE_TEST
 	select DRM_KMS_HELPER
 	select DRM_KMS_CMA_HELPER
 	select DRM_GEM_CMA_HELPER
@@ -12,6 +13,7 @@ config DRM_RCAR_DU
 config DRM_RCAR_LVDS
 	bool "R-Car DU LVDS Encoder Support"
 	depends on DRM_RCAR_DU
+	depends on ARCH_R8A7790 || ARCH_R8A7791 || COMPILE_TEST
 	help
 	  Enable support the R-Car Display Unit embedded LVDS encoders
 	  (currently only on R8A7790).
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
index a9d24e4bf79..299267db289 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
@@ -299,7 +299,7 @@ static void rcar_du_crtc_update_base(struct rcar_du_crtc *rcrtc)
 {
 	struct drm_crtc *crtc = &rcrtc->crtc;
 
-	rcar_du_plane_compute_base(rcrtc->plane, crtc->fb);
+	rcar_du_plane_compute_base(rcrtc->plane, crtc->primary->fb);
 	rcar_du_plane_update_base(rcrtc->plane);
 }
 
@@ -358,10 +358,10 @@ static int rcar_du_crtc_mode_set(struct drm_crtc *crtc,
 	const struct rcar_du_format_info *format;
 	int ret;
 
-	format = rcar_du_format_info(crtc->fb->pixel_format);
+	format = rcar_du_format_info(crtc->primary->fb->pixel_format);
 	if (format == NULL) {
 		dev_dbg(rcdu->dev, "mode_set: unsupported format %08x\n",
-			crtc->fb->pixel_format);
+			crtc->primary->fb->pixel_format);
 		ret = -EINVAL;
 		goto error;
 	}
@@ -371,14 +371,13 @@ static int rcar_du_crtc_mode_set(struct drm_crtc *crtc,
 		goto error;
 
 	rcrtc->plane->format = format;
-	rcrtc->plane->pitch = crtc->fb->pitches[0];
 
 	rcrtc->plane->src_x = x;
 	rcrtc->plane->src_y = y;
 	rcrtc->plane->width = mode->hdisplay;
 	rcrtc->plane->height = mode->vdisplay;
 
-	rcar_du_plane_compute_base(rcrtc->plane, crtc->fb);
+	rcar_du_plane_compute_base(rcrtc->plane, crtc->primary->fb);
 
 	rcrtc->outputs = 0;
 
@@ -413,7 +412,7 @@ static int rcar_du_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 	rcrtc->plane->src_x = x;
 	rcrtc->plane->src_y = y;
 
-	rcar_du_crtc_update_base(to_rcar_crtc(crtc));
+	rcar_du_crtc_update_base(rcrtc);
 
 	return 0;
 }
@@ -511,7 +510,7 @@ static int rcar_du_crtc_page_flip(struct drm_crtc *crtc,
 	}
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	crtc->fb = fb;
+	crtc->primary->fb = fb;
 	rcar_du_crtc_update_base(rcrtc);
 
 	if (event) {
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
index 0023f9719cf..792fd1d20e8 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
@@ -224,7 +224,9 @@ static int rcar_du_probe(struct platform_device *pdev)
 
 static int rcar_du_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&rcar_du_driver, pdev);
+	struct rcar_du_device *rcdu = platform_get_drvdata(pdev);
+
+	drm_put_dev(rcdu->ddev);
 
 	return 0;
 }
@@ -249,8 +251,8 @@ static const struct rcar_du_device_info rcar_du_r8a7779_info = {
 };
 
 static const struct rcar_du_device_info rcar_du_r8a7790_info = {
-	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_ALIGN_128B
-		  | RCAR_DU_FEATURE_DEFR8,
+	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
+	.quirks = RCAR_DU_QUIRK_ALIGN_128B | RCAR_DU_QUIRK_LVDS_LANES,
 	.num_crtcs = 3,
 	.routes = {
 		/* R8A7790 has one RGB output, two LVDS outputs and one
@@ -272,9 +274,29 @@ static const struct rcar_du_device_info rcar_du_r8a7790_info = {
 	.num_lvds = 2,
 };
 
+static const struct rcar_du_device_info rcar_du_r8a7791_info = {
+	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_DEFR8,
+	.num_crtcs = 2,
+	.routes = {
+		/* R8A7791 has one RGB output, one LVDS output and one
+		 * (currently unsupported) TCON output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(1),
+			.encoder_type = DRM_MODE_ENCODER_NONE,
+		},
+		[RCAR_DU_OUTPUT_LVDS0] = {
+			.possible_crtcs = BIT(0),
+			.encoder_type = DRM_MODE_ENCODER_LVDS,
+		},
+	},
+	.num_lvds = 1,
+};
+
 static const struct platform_device_id rcar_du_id_table[] = {
 	{ "rcar-du-r8a7779", (kernel_ulong_t)&rcar_du_r8a7779_info },
 	{ "rcar-du-r8a7790", (kernel_ulong_t)&rcar_du_r8a7790_info },
+	{ "rcar-du-r8a7791", (kernel_ulong_t)&rcar_du_r8a7791_info },
 	{ }
 };
 
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.h b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
index 65d2d636b00..e31b735d3f2 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
@@ -28,8 +28,10 @@ struct rcar_du_device;
 struct rcar_du_lvdsenc;
 
 #define RCAR_DU_FEATURE_CRTC_IRQ_CLOCK	(1 << 0)	/* Per-CRTC IRQ and clock */
-#define RCAR_DU_FEATURE_ALIGN_128B	(1 << 1)	/* Align pitches to 128 bytes */
-#define RCAR_DU_FEATURE_DEFR8		(1 << 2)	/* Has DEFR8 register */
+#define RCAR_DU_FEATURE_DEFR8		(1 << 1)	/* Has DEFR8 register */
+
+#define RCAR_DU_QUIRK_ALIGN_128B	(1 << 0)	/* Align pitches to 128 bytes */
+#define RCAR_DU_QUIRK_LVDS_LANES	(1 << 1)	/* LVDS lanes 1 and 3 inverted */
 
 /*
  * struct rcar_du_output_routing - Output routing specification
@@ -48,12 +50,14 @@ struct rcar_du_output_routing {
 /*
  * struct rcar_du_device_info - DU model-specific information
  * @features: device features (RCAR_DU_FEATURE_*)
+ * @quirks: device quirks (RCAR_DU_QUIRK_*)
  * @num_crtcs: total number of CRTCs
  * @routes: array of CRTC to output routes, indexed by output (RCAR_DU_OUTPUT_*)
  * @num_lvds: number of internal LVDS encoders
  */
 struct rcar_du_device_info {
 	unsigned int features;
+	unsigned int quirks;
 	unsigned int num_crtcs;
 	struct rcar_du_output_routing routes[RCAR_DU_OUTPUT_MAX];
 	unsigned int num_lvds;
@@ -84,6 +88,12 @@ static inline bool rcar_du_has(struct rcar_du_device *rcdu,
 	return rcdu->info->features & feature;
 }
 
+static inline bool rcar_du_needs(struct rcar_du_device *rcdu,
+				 unsigned int quirk)
+{
+	return rcdu->info->quirks & quirk;
+}
+
 static inline u32 rcar_du_read(struct rcar_du_device *rcdu, u32 reg)
 {
 	return ioread32(rcdu->mmio + reg);
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index b31ac080c4a..a87edfac111 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -119,7 +119,7 @@ int rcar_du_dumb_create(struct drm_file *file, struct drm_device *dev,
 	/* The R8A7779 DU requires a 16 pixels pitch alignment as documented,
 	 * but the R8A7790 DU seems to require a 128 bytes pitch alignment.
 	 */
-	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_ALIGN_128B))
+	if (rcar_du_needs(rcdu, RCAR_DU_QUIRK_ALIGN_128B))
 		align = 128;
 	else
 		align = 16 * args->bpp / 8;
@@ -144,7 +144,7 @@ rcar_du_fb_create(struct drm_device *dev, struct drm_file *file_priv,
 		return ERR_PTR(-EINVAL);
 	}
 
-	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_ALIGN_128B))
+	if (rcar_du_needs(rcdu, RCAR_DU_QUIRK_ALIGN_128B))
 		align = 128;
 	else
 		align = 16 * format->bpp / 8;
@@ -248,7 +248,10 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 			continue;
 		}
 
-		rcar_du_encoder_init(rcdu, pdata->type, pdata->output, pdata);
+		ret = rcar_du_encoder_init(rcdu, pdata->type, pdata->output,
+					   pdata);
+		if (ret < 0)
+			return ret;
 	}
 
 	/* Set the possible CRTCs and possible clones. There's always at least
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c
index 4f3ba93cd91..289048d1c7b 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdscon.c
@@ -57,15 +57,8 @@ static int rcar_du_lvds_connector_get_modes(struct drm_connector *connector)
 	return 1;
 }
 
-static int rcar_du_lvds_connector_mode_valid(struct drm_connector *connector,
-					    struct drm_display_mode *mode)
-{
-	return MODE_OK;
-}
-
 static const struct drm_connector_helper_funcs connector_helper_funcs = {
 	.get_modes = rcar_du_lvds_connector_get_modes,
-	.mode_valid = rcar_du_lvds_connector_mode_valid,
 	.best_encoder = rcar_du_connector_best_encoder,
 };
 
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
index a0f6a178192..df30a075d79 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
@@ -44,6 +44,7 @@ static int rcar_du_lvdsenc_start(struct rcar_du_lvdsenc *lvds,
 	const struct drm_display_mode *mode = &rcrtc->crtc.mode;
 	unsigned int freq = mode->clock;
 	u32 lvdcr0;
+	u32 lvdhcr;
 	u32 pllcr;
 	int ret;
 
@@ -72,15 +73,19 @@ static int rcar_du_lvdsenc_start(struct rcar_du_lvdsenc *lvds,
 	 * VSYNC -> CTRL1
 	 * DISP  -> CTRL2
 	 * 0     -> CTRL3
-	 *
-	 * Channels 1 and 3 are switched on ES1.
 	 */
 	rcar_lvds_write(lvds, LVDCTRCR, LVDCTRCR_CTR3SEL_ZERO |
 			LVDCTRCR_CTR2SEL_DISP | LVDCTRCR_CTR1SEL_VSYNC |
 			LVDCTRCR_CTR0SEL_HSYNC);
-	rcar_lvds_write(lvds, LVDCHCR,
-			LVDCHCR_CHSEL_CH(0, 0) | LVDCHCR_CHSEL_CH(1, 3) |
-			LVDCHCR_CHSEL_CH(2, 2) | LVDCHCR_CHSEL_CH(3, 1));
+
+	if (rcar_du_needs(lvds->dev, RCAR_DU_QUIRK_LVDS_LANES))
+		lvdhcr = LVDCHCR_CHSEL_CH(0, 0) | LVDCHCR_CHSEL_CH(1, 3)
+		       | LVDCHCR_CHSEL_CH(2, 2) | LVDCHCR_CHSEL_CH(3, 1);
+	else
+		lvdhcr = LVDCHCR_CHSEL_CH(0, 0) | LVDCHCR_CHSEL_CH(1, 1)
+		       | LVDCHCR_CHSEL_CH(2, 2) | LVDCHCR_CHSEL_CH(3, 3);
+
+	rcar_lvds_write(lvds, LVDCHCR, lvdhcr);
 
 	/* Select the input, hardcode mode 0, enable LVDS operation and turn
 	 * bias circuitry on.
@@ -144,18 +149,9 @@ static int rcar_du_lvdsenc_get_resources(struct rcar_du_lvdsenc *lvds,
 	sprintf(name, "lvds.%u", lvds->index);
 
 	mem = platform_get_resource_byname(pdev, IORESOURCE_MEM, name);
-	if (mem == NULL) {
-		dev_err(&pdev->dev, "failed to get memory resource for %s\n",
-			name);
-		return -EINVAL;
-	}
-
 	lvds->mmio = devm_ioremap_resource(&pdev->dev, mem);
-	if (lvds->mmio == NULL) {
-		dev_err(&pdev->dev, "failed to remap memory resource for %s\n",
-			name);
-		return -ENOMEM;
-	}
+	if (IS_ERR(lvds->mmio))
+		return PTR_ERR(lvds->mmio);
 
 	lvds->clock = devm_clk_get(&pdev->dev, name);
 	if (IS_ERR(lvds->clock)) {
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_plane.c b/drivers/gpu/drm/rcar-du/rcar_du_plane.c
index 53000644733..3fb69d9ae61 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_plane.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_plane.c
@@ -104,6 +104,15 @@ void rcar_du_plane_update_base(struct rcar_du_plane *plane)
 {
 	struct rcar_du_group *rgrp = plane->group;
 	unsigned int index = plane->hwindex;
+	u32 mwr;
+
+	/* Memory pitch (expressed in pixels) */
+	if (plane->format->planes == 2)
+		mwr = plane->pitch;
+	else
+		mwr = plane->pitch * 8 / plane->format->bpp;
+
+	rcar_du_plane_write(rgrp, index, PnMWR, mwr);
 
 	/* The Y position is expressed in raster line units and must be doubled
 	 * for 32bpp formats, according to the R8A7790 datasheet. No mention of
@@ -133,6 +142,8 @@ void rcar_du_plane_compute_base(struct rcar_du_plane *plane,
 {
 	struct drm_gem_cma_object *gem;
 
+	plane->pitch = fb->pitches[0];
+
 	gem = drm_fb_cma_get_gem_obj(fb, 0);
 	plane->dma[0] = gem->paddr + fb->offsets[0];
 
@@ -209,7 +220,6 @@ static void __rcar_du_plane_setup(struct rcar_du_plane *plane,
 	struct rcar_du_group *rgrp = plane->group;
 	u32 ddcr2 = PnDDCR2_CODE;
 	u32 ddcr4;
-	u32 mwr;
 
 	/* Data format
 	 *
@@ -240,14 +250,6 @@ static void __rcar_du_plane_setup(struct rcar_du_plane *plane,
 	rcar_du_plane_write(rgrp, index, PnDDCR2, ddcr2);
 	rcar_du_plane_write(rgrp, index, PnDDCR4, ddcr4);
 
-	/* Memory pitch (expressed in pixels) */
-	if (plane->format->planes == 2)
-		mwr = plane->pitch;
-	else
-		mwr = plane->pitch * 8 / plane->format->bpp;
-
-	rcar_du_plane_write(rgrp, index, PnMWR, mwr);
-
 	/* Destination position and size */
 	rcar_du_plane_write(rgrp, index, PnDSXR, plane->width);
 	rcar_du_plane_write(rgrp, index, PnDSYR, plane->height);
@@ -309,7 +311,6 @@ rcar_du_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
 
 	rplane->crtc = crtc;
 	rplane->format = format;
-	rplane->pitch = fb->pitches[0];
 
 	rplane->src_x = src_x >> 16;
 	rplane->src_y = src_y >> 16;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c
index 41d563adfea..ccfe64c7188 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_vgacon.c
@@ -25,15 +25,8 @@ static int rcar_du_vga_connector_get_modes(struct drm_connector *connector)
 	return 0;
 }
 
-static int rcar_du_vga_connector_mode_valid(struct drm_connector *connector,
-					    struct drm_display_mode *mode)
-{
-	return MODE_OK;
-}
-
 static const struct drm_connector_helper_funcs connector_helper_funcs = {
 	.get_modes = rcar_du_vga_connector_get_modes,
-	.mode_valid = rcar_du_vga_connector_mode_valid,
 	.best_encoder = rcar_du_connector_best_encoder,
 };
 
diff --git a/drivers/gpu/drm/savage/savage_bci.c b/drivers/gpu/drm/savage/savage_bci.c
index b17d0710871..c97cdc9ab23 100644
--- a/drivers/gpu/drm/savage/savage_bci.c
+++ b/drivers/gpu/drm/savage/savage_bci.c
@@ -49,7 +49,7 @@ savage_bci_wait_fifo_shadow(drm_savage_private_t * dev_priv, unsigned int n)
 #endif
 
 	for (i = 0; i < SAVAGE_DEFAULT_USEC_TIMEOUT; i++) {
-		DRM_MEMORYBARRIER();
+		mb();
 		status = dev_priv->status_ptr[0];
 		if ((status & mask) < threshold)
 			return 0;
@@ -123,7 +123,7 @@ savage_bci_wait_event_shadow(drm_savage_private_t * dev_priv, uint16_t e)
 	int i;
 
 	for (i = 0; i < SAVAGE_EVENT_USEC_TIMEOUT; i++) {
-		DRM_MEMORYBARRIER();
+		mb();
 		status = dev_priv->status_ptr[1];
 		if ((((status & 0xffff) - e) & 0xffff) <= 0x7fff ||
 		    (status & 0xffff) == 0)
@@ -449,7 +449,7 @@ static void savage_dma_flush(drm_savage_private_t * dev_priv)
 		}
 	}
 
-	DRM_MEMORYBARRIER();
+	mb();
 
 	/* do flush ... */
 	phys_addr = dev_priv->cmd_dma->offset +
@@ -990,10 +990,10 @@ static int savage_bci_get_buffers(struct drm_device *dev,
 
 		buf->file_priv = file_priv;
 
-		if (DRM_COPY_TO_USER(&d->request_indices[i],
+		if (copy_to_user(&d->request_indices[i],
 				     &buf->idx, sizeof(buf->idx)))
 			return -EFAULT;
-		if (DRM_COPY_TO_USER(&d->request_sizes[i],
+		if (copy_to_user(&d->request_sizes[i],
 				     &buf->total, sizeof(buf->total)))
 			return -EFAULT;
 
@@ -1079,4 +1079,4 @@ const struct drm_ioctl_desc savage_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(SAVAGE_BCI_EVENT_WAIT, savage_bci_event_wait, DRM_AUTH),
 };
 
-int savage_max_ioctl = DRM_ARRAY_SIZE(savage_ioctls);
+int savage_max_ioctl = ARRAY_SIZE(savage_ioctls);
diff --git a/drivers/gpu/drm/savage/savage_state.c b/drivers/gpu/drm/savage/savage_state.c
index b35e75ed890..c01ad0aeaa5 100644
--- a/drivers/gpu/drm/savage/savage_state.c
+++ b/drivers/gpu/drm/savage/savage_state.c
@@ -992,7 +992,7 @@ int savage_bci_cmdbuf(struct drm_device *dev, void *data, struct drm_file *file_
 		if (kcmd_addr == NULL)
 			return -ENOMEM;
 
-		if (DRM_COPY_FROM_USER(kcmd_addr, cmdbuf->cmd_addr,
+		if (copy_from_user(kcmd_addr, cmdbuf->cmd_addr,
 				       cmdbuf->size * 8))
 		{
 			kfree(kcmd_addr);
@@ -1007,7 +1007,7 @@ int savage_bci_cmdbuf(struct drm_device *dev, void *data, struct drm_file *file_
 			goto done;
 		}
 
-		if (DRM_COPY_FROM_USER(kvb_addr, cmdbuf->vb_addr,
+		if (copy_from_user(kvb_addr, cmdbuf->vb_addr,
 				       cmdbuf->vb_size)) {
 			ret = -EFAULT;
 			goto done;
@@ -1022,7 +1022,7 @@ int savage_bci_cmdbuf(struct drm_device *dev, void *data, struct drm_file *file_
 			goto done;
 		}
 
-		if (DRM_COPY_FROM_USER(kbox_addr, cmdbuf->box_addr,
+		if (copy_from_user(kbox_addr, cmdbuf->box_addr,
 				       cmdbuf->nbox * sizeof(struct drm_clip_rect))) {
 			ret = -EFAULT;
 			goto done;
@@ -1032,7 +1032,7 @@ int savage_bci_cmdbuf(struct drm_device *dev, void *data, struct drm_file *file_
 
 	/* Make sure writes to DMA buffers are finished before sending
 	 * DMA commands to the graphics hardware. */
-	DRM_MEMORYBARRIER();
+	mb();
 
 	/* Coming from user space. Don't know if the Xserver has
 	 * emitted wait commands. Assuming the worst. */
diff --git a/drivers/gpu/drm/shmobile/Kconfig b/drivers/gpu/drm/shmobile/Kconfig
index 2ee44ca9d67..a50fe0eeaa0 100644
--- a/drivers/gpu/drm/shmobile/Kconfig
+++ b/drivers/gpu/drm/shmobile/Kconfig
@@ -1,6 +1,7 @@
 config DRM_SHMOBILE
 	tristate "DRM Support for SH Mobile"
-	depends on DRM && (ARM || SUPERH)
+	depends on DRM && ARM
+	depends on ARCH_SHMOBILE || COMPILE_TEST
 	select BACKLIGHT_CLASS_DEVICE
 	select DRM_KMS_HELPER
 	select DRM_KMS_FB_HELPER
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
index 562f9a401cf..faf176b2daf 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
@@ -37,14 +37,21 @@
  * Clock management
  */
 
-static void shmob_drm_clk_on(struct shmob_drm_device *sdev)
+static int shmob_drm_clk_on(struct shmob_drm_device *sdev)
 {
-	if (sdev->clock)
-		clk_prepare_enable(sdev->clock);
+	int ret;
+
+	if (sdev->clock) {
+		ret = clk_prepare_enable(sdev->clock);
+		if (ret < 0)
+			return ret;
+	}
 #if 0
 	if (sdev->meram_dev && sdev->meram_dev->pdev)
 		pm_runtime_get_sync(&sdev->meram_dev->pdev->dev);
 #endif
+
+	return 0;
 }
 
 static void shmob_drm_clk_off(struct shmob_drm_device *sdev)
@@ -161,16 +168,19 @@ static void shmob_drm_crtc_start(struct shmob_drm_crtc *scrtc)
 	struct drm_device *dev = sdev->ddev;
 	struct drm_plane *plane;
 	u32 value;
+	int ret;
 
 	if (scrtc->started)
 		return;
 
-	format = shmob_drm_format_info(crtc->fb->pixel_format);
+	format = shmob_drm_format_info(crtc->primary->fb->pixel_format);
 	if (WARN_ON(format == NULL))
 		return;
 
 	/* Enable clocks before accessing the hardware. */
-	shmob_drm_clk_on(sdev);
+	ret = shmob_drm_clk_on(sdev);
+	if (ret < 0)
+		return;
 
 	/* Reset and enable the LCDC. */
 	lcdc_write(sdev, LDCNT2R, lcdc_read(sdev, LDCNT2R) | LDCNT2R_BR);
@@ -237,7 +247,7 @@ static void shmob_drm_crtc_start(struct shmob_drm_crtc *scrtc)
 	lcdc_write(sdev, LDDDSR, value);
 
 	/* Setup planes. */
-	list_for_each_entry(plane, &dev->mode_config.plane_list, head) {
+	drm_for_each_legacy_plane(plane, &dev->mode_config.plane_list) {
 		if (plane->crtc == crtc)
 			shmob_drm_plane_setup(plane);
 	}
@@ -293,7 +303,7 @@ static void shmob_drm_crtc_compute_base(struct shmob_drm_crtc *scrtc,
 					int x, int y)
 {
 	struct drm_crtc *crtc = &scrtc->crtc;
-	struct drm_framebuffer *fb = crtc->fb;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	struct shmob_drm_device *sdev = crtc->dev->dev_private;
 	struct drm_gem_cma_object *gem;
 	unsigned int bpp;
@@ -372,15 +382,15 @@ static int shmob_drm_crtc_mode_set(struct drm_crtc *crtc,
 	const struct shmob_drm_format_info *format;
 	void *cache;
 
-	format = shmob_drm_format_info(crtc->fb->pixel_format);
+	format = shmob_drm_format_info(crtc->primary->fb->pixel_format);
 	if (format == NULL) {
 		dev_dbg(sdev->dev, "mode_set: unsupported format %08x\n",
-			crtc->fb->pixel_format);
+			crtc->primary->fb->pixel_format);
 		return -EINVAL;
 	}
 
 	scrtc->format = format;
-	scrtc->line_size = crtc->fb->pitches[0];
+	scrtc->line_size = crtc->primary->fb->pitches[0];
 
 	if (sdev->meram) {
 		/* Enable MERAM cache if configured. We need to de-init
@@ -392,7 +402,7 @@ static int shmob_drm_crtc_mode_set(struct drm_crtc *crtc,
 		}
 
 		cache = sh_mobile_meram_cache_alloc(sdev->meram, mdata,
-						    crtc->fb->pitches[0],
+						    crtc->primary->fb->pitches[0],
 						    adjusted_mode->vdisplay,
 						    format->meram,
 						    &scrtc->line_size);
@@ -479,7 +489,7 @@ static int shmob_drm_crtc_page_flip(struct drm_crtc *crtc,
 	}
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 
-	crtc->fb = fb;
+	crtc->primary->fb = fb;
 	shmob_drm_crtc_update_base(scrtc);
 
 	if (event) {
@@ -664,12 +674,6 @@ static int shmob_drm_connector_get_modes(struct drm_connector *connector)
 	return 1;
 }
 
-static int shmob_drm_connector_mode_valid(struct drm_connector *connector,
-					  struct drm_display_mode *mode)
-{
-	return MODE_OK;
-}
-
 static struct drm_encoder *
 shmob_drm_connector_best_encoder(struct drm_connector *connector)
 {
@@ -680,7 +684,6 @@ shmob_drm_connector_best_encoder(struct drm_connector *connector)
 
 static const struct drm_connector_helper_funcs connector_helper_funcs = {
 	.get_modes = shmob_drm_connector_get_modes,
-	.mode_valid = shmob_drm_connector_mode_valid,
 	.best_encoder = shmob_drm_connector_best_encoder,
 };
 
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_drv.c b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
index 015551866b4..82c84c7fd4f 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_drv.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
@@ -185,7 +185,7 @@ static int shmob_drm_load(struct drm_device *dev, unsigned long flags)
 		goto done;
 	}
 
-	ret = drm_irq_install(dev);
+	ret = drm_irq_install(dev, platform_get_irq(dev->platformdev, 0));
 	if (ret < 0) {
 		dev_err(&pdev->dev, "failed to install IRQ handler\n");
 		goto done;
@@ -336,7 +336,9 @@ static int shmob_drm_probe(struct platform_device *pdev)
 
 static int shmob_drm_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&shmob_drm_driver, pdev);
+	struct shmob_drm_device *sdev = platform_get_drvdata(pdev);
+
+	drm_put_dev(sdev->ddev);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/sis/sis_drv.c b/drivers/gpu/drm/sis/sis_drv.c
index 4383b74a3aa..756f787b714 100644
--- a/drivers/gpu/drm/sis/sis_drv.c
+++ b/drivers/gpu/drm/sis/sis_drv.c
@@ -94,7 +94,7 @@ static int sis_driver_open(struct drm_device *dev, struct drm_file *file)
 	return 0;
 }
 
-void sis_driver_postclose(struct drm_device *dev, struct drm_file *file)
+static void sis_driver_postclose(struct drm_device *dev, struct drm_file *file)
 {
 	struct sis_file_private *file_priv = file->driver_priv;
 
diff --git a/drivers/gpu/drm/sis/sis_mm.c b/drivers/gpu/drm/sis/sis_mm.c
index 01857d83635..77f288e4a0a 100644
--- a/drivers/gpu/drm/sis/sis_mm.c
+++ b/drivers/gpu/drm/sis/sis_mm.c
@@ -266,7 +266,7 @@ int sis_idle(struct drm_device *dev)
 	 * because its polling frequency is too low.
 	 */
 
-	end = jiffies + (DRM_HZ * 3);
+	end = jiffies + (HZ * 3);
 
 	for (i = 0; i < 4; ++i) {
 		do {
@@ -359,4 +359,4 @@ const struct drm_ioctl_desc sis_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(SIS_FB_INIT, sis_fb_init, DRM_AUTH | DRM_MASTER | DRM_ROOT_ONLY),
 };
 
-int sis_max_ioctl = DRM_ARRAY_SIZE(sis_ioctls);
+int sis_max_ioctl = ARRAY_SIZE(sis_ioctls);
diff --git a/drivers/gpu/drm/tegra/Kconfig b/drivers/gpu/drm/tegra/Kconfig
index 8961ba6a34b..354ddb29231 100644
--- a/drivers/gpu/drm/tegra/Kconfig
+++ b/drivers/gpu/drm/tegra/Kconfig
@@ -1,13 +1,12 @@
 config DRM_TEGRA
-	bool "NVIDIA Tegra DRM"
-	depends on ARCH_TEGRA || ARCH_MULTIPLATFORM
+	tristate "NVIDIA Tegra DRM"
+	depends on ARCH_TEGRA || (ARM && COMPILE_TEST)
 	depends on DRM
-	select TEGRA_HOST1X
+	depends on RESET_CONTROLLER
 	select DRM_KMS_HELPER
-	select DRM_KMS_FB_HELPER
-	select FB_SYS_FILLRECT
-	select FB_SYS_COPYAREA
-	select FB_SYS_IMAGEBLIT
+	select DRM_MIPI_DSI
+	select DRM_PANEL
+	select TEGRA_HOST1X
 	help
 	  Choose this option if you have an NVIDIA Tegra SoC.
 
@@ -16,6 +15,18 @@ config DRM_TEGRA
 
 if DRM_TEGRA
 
+config DRM_TEGRA_FBDEV
+	bool "Enable legacy fbdev support"
+	select DRM_KMS_FB_HELPER
+	select FB_SYS_FILLRECT
+	select FB_SYS_COPYAREA
+	select FB_SYS_IMAGEBLIT
+	default y
+	help
+	  Choose this option if you have a need for the legacy fbdev support.
+	  Note that this support also provides the Linux console on top of
+	  the Tegra modesetting driver.
+
 config DRM_TEGRA_DEBUG
 	bool "NVIDIA Tegra DRM debug support"
 	help
diff --git a/drivers/gpu/drm/tegra/Makefile b/drivers/gpu/drm/tegra/Makefile
index edc76abd58b..2c66a8db9da 100644
--- a/drivers/gpu/drm/tegra/Makefile
+++ b/drivers/gpu/drm/tegra/Makefile
@@ -1,7 +1,6 @@
 ccflags-$(CONFIG_DRM_TEGRA_DEBUG) += -DDEBUG
 
 tegra-drm-y := \
-	bus.o \
 	drm.o \
 	gem.o \
 	fb.o \
@@ -9,6 +8,10 @@ tegra-drm-y := \
 	output.o \
 	rgb.o \
 	hdmi.o \
+	mipi-phy.o \
+	dsi.o \
+	sor.o \
+	dpaux.o \
 	gr2d.o \
 	gr3d.o
 
diff --git a/drivers/gpu/drm/tegra/bus.c b/drivers/gpu/drm/tegra/bus.c
deleted file mode 100644
index 565f8f7b9a4..00000000000
--- a/drivers/gpu/drm/tegra/bus.c
+++ /dev/null
@@ -1,76 +0,0 @@
-/*
- * Copyright (C) 2013 NVIDIA Corporation
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- */
-
-#include "drm.h"
-
-static int drm_host1x_set_busid(struct drm_device *dev,
-				struct drm_master *master)
-{
-	const char *device = dev_name(dev->dev);
-	const char *driver = dev->driver->name;
-	const char *bus = dev->dev->bus->name;
-	int length;
-
-	master->unique_len = strlen(bus) + 1 + strlen(device);
-	master->unique_size = master->unique_len;
-
-	master->unique = kmalloc(master->unique_len + 1, GFP_KERNEL);
-	if (!master->unique)
-		return -ENOMEM;
-
-	snprintf(master->unique, master->unique_len + 1, "%s:%s", bus, device);
-
-	length = strlen(driver) + 1 + master->unique_len;
-
-	dev->devname = kmalloc(length + 1, GFP_KERNEL);
-	if (!dev->devname)
-		return -ENOMEM;
-
-	snprintf(dev->devname, length + 1, "%s@%s", driver, master->unique);
-
-	return 0;
-}
-
-static struct drm_bus drm_host1x_bus = {
-	.bus_type = DRIVER_BUS_HOST1X,
-	.set_busid = drm_host1x_set_busid,
-};
-
-int drm_host1x_init(struct drm_driver *driver, struct host1x_device *device)
-{
-	struct drm_device *drm;
-	int ret;
-
-	INIT_LIST_HEAD(&driver->device_list);
-	driver->bus = &drm_host1x_bus;
-
-	drm = drm_dev_alloc(driver, &device->dev);
-	if (!drm)
-		return -ENOMEM;
-
-	ret = drm_dev_register(drm, 0);
-	if (ret)
-		goto err_free;
-
-	DRM_INFO("Initialized %s %d.%d.%d %s on minor %d\n", driver->name,
-		 driver->major, driver->minor, driver->patchlevel,
-		 driver->date, drm->primary->index);
-
-	return 0;
-
-err_free:
-	drm_dev_free(drm);
-	return ret;
-}
-
-void drm_host1x_exit(struct drm_driver *driver, struct host1x_device *device)
-{
-	struct tegra_drm *tegra = dev_get_drvdata(&device->dev);
-
-	drm_put_dev(tegra->drm);
-}
diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index ae1cb31ead7..ef40381f390 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -8,13 +8,18 @@
  */
 
 #include <linux/clk.h>
-#include <linux/clk/tegra.h>
 #include <linux/debugfs.h>
+#include <linux/reset.h>
 
 #include "dc.h"
 #include "drm.h"
 #include "gem.h"
 
+struct tegra_dc_soc_info {
+	bool supports_interlacing;
+	bool supports_cursor;
+};
+
 struct tegra_plane {
 	struct drm_plane base;
 	unsigned int index;
@@ -25,6 +30,254 @@ static inline struct tegra_plane *to_tegra_plane(struct drm_plane *plane)
 	return container_of(plane, struct tegra_plane, base);
 }
 
+static unsigned int tegra_dc_format(uint32_t format, uint32_t *swap)
+{
+	/* assume no swapping of fetched data */
+	if (swap)
+		*swap = BYTE_SWAP_NOSWAP;
+
+	switch (format) {
+	case DRM_FORMAT_XBGR8888:
+		return WIN_COLOR_DEPTH_R8G8B8A8;
+
+	case DRM_FORMAT_XRGB8888:
+		return WIN_COLOR_DEPTH_B8G8R8A8;
+
+	case DRM_FORMAT_RGB565:
+		return WIN_COLOR_DEPTH_B5G6R5;
+
+	case DRM_FORMAT_UYVY:
+		return WIN_COLOR_DEPTH_YCbCr422;
+
+	case DRM_FORMAT_YUYV:
+		if (swap)
+			*swap = BYTE_SWAP_SWAP2;
+
+		return WIN_COLOR_DEPTH_YCbCr422;
+
+	case DRM_FORMAT_YUV420:
+		return WIN_COLOR_DEPTH_YCbCr420P;
+
+	case DRM_FORMAT_YUV422:
+		return WIN_COLOR_DEPTH_YCbCr422P;
+
+	default:
+		break;
+	}
+
+	WARN(1, "unsupported pixel format %u, using default\n", format);
+	return WIN_COLOR_DEPTH_B8G8R8A8;
+}
+
+static bool tegra_dc_format_is_yuv(unsigned int format, bool *planar)
+{
+	switch (format) {
+	case WIN_COLOR_DEPTH_YCbCr422:
+	case WIN_COLOR_DEPTH_YUV422:
+		if (planar)
+			*planar = false;
+
+		return true;
+
+	case WIN_COLOR_DEPTH_YCbCr420P:
+	case WIN_COLOR_DEPTH_YUV420P:
+	case WIN_COLOR_DEPTH_YCbCr422P:
+	case WIN_COLOR_DEPTH_YUV422P:
+	case WIN_COLOR_DEPTH_YCbCr422R:
+	case WIN_COLOR_DEPTH_YUV422R:
+	case WIN_COLOR_DEPTH_YCbCr422RA:
+	case WIN_COLOR_DEPTH_YUV422RA:
+		if (planar)
+			*planar = true;
+
+		return true;
+	}
+
+	return false;
+}
+
+static inline u32 compute_dda_inc(unsigned int in, unsigned int out, bool v,
+				  unsigned int bpp)
+{
+	fixed20_12 outf = dfixed_init(out);
+	fixed20_12 inf = dfixed_init(in);
+	u32 dda_inc;
+	int max;
+
+	if (v)
+		max = 15;
+	else {
+		switch (bpp) {
+		case 2:
+			max = 8;
+			break;
+
+		default:
+			WARN_ON_ONCE(1);
+			/* fallthrough */
+		case 4:
+			max = 4;
+			break;
+		}
+	}
+
+	outf.full = max_t(u32, outf.full - dfixed_const(1), dfixed_const(1));
+	inf.full -= dfixed_const(1);
+
+	dda_inc = dfixed_div(inf, outf);
+	dda_inc = min_t(u32, dda_inc, dfixed_const(max));
+
+	return dda_inc;
+}
+
+static inline u32 compute_initial_dda(unsigned int in)
+{
+	fixed20_12 inf = dfixed_init(in);
+	return dfixed_frac(inf);
+}
+
+static int tegra_dc_setup_window(struct tegra_dc *dc, unsigned int index,
+				 const struct tegra_dc_window *window)
+{
+	unsigned h_offset, v_offset, h_size, v_size, h_dda, v_dda, bpp;
+	unsigned long value;
+	bool yuv, planar;
+
+	/*
+	 * For YUV planar modes, the number of bytes per pixel takes into
+	 * account only the luma component and therefore is 1.
+	 */
+	yuv = tegra_dc_format_is_yuv(window->format, &planar);
+	if (!yuv)
+		bpp = window->bits_per_pixel / 8;
+	else
+		bpp = planar ? 1 : 2;
+
+	value = WINDOW_A_SELECT << index;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_WINDOW_HEADER);
+
+	tegra_dc_writel(dc, window->format, DC_WIN_COLOR_DEPTH);
+	tegra_dc_writel(dc, window->swap, DC_WIN_BYTE_SWAP);
+
+	value = V_POSITION(window->dst.y) | H_POSITION(window->dst.x);
+	tegra_dc_writel(dc, value, DC_WIN_POSITION);
+
+	value = V_SIZE(window->dst.h) | H_SIZE(window->dst.w);
+	tegra_dc_writel(dc, value, DC_WIN_SIZE);
+
+	h_offset = window->src.x * bpp;
+	v_offset = window->src.y;
+	h_size = window->src.w * bpp;
+	v_size = window->src.h;
+
+	value = V_PRESCALED_SIZE(v_size) | H_PRESCALED_SIZE(h_size);
+	tegra_dc_writel(dc, value, DC_WIN_PRESCALED_SIZE);
+
+	/*
+	 * For DDA computations the number of bytes per pixel for YUV planar
+	 * modes needs to take into account all Y, U and V components.
+	 */
+	if (yuv && planar)
+		bpp = 2;
+
+	h_dda = compute_dda_inc(window->src.w, window->dst.w, false, bpp);
+	v_dda = compute_dda_inc(window->src.h, window->dst.h, true, bpp);
+
+	value = V_DDA_INC(v_dda) | H_DDA_INC(h_dda);
+	tegra_dc_writel(dc, value, DC_WIN_DDA_INC);
+
+	h_dda = compute_initial_dda(window->src.x);
+	v_dda = compute_initial_dda(window->src.y);
+
+	tegra_dc_writel(dc, h_dda, DC_WIN_H_INITIAL_DDA);
+	tegra_dc_writel(dc, v_dda, DC_WIN_V_INITIAL_DDA);
+
+	tegra_dc_writel(dc, 0, DC_WIN_UV_BUF_STRIDE);
+	tegra_dc_writel(dc, 0, DC_WIN_BUF_STRIDE);
+
+	tegra_dc_writel(dc, window->base[0], DC_WINBUF_START_ADDR);
+
+	if (yuv && planar) {
+		tegra_dc_writel(dc, window->base[1], DC_WINBUF_START_ADDR_U);
+		tegra_dc_writel(dc, window->base[2], DC_WINBUF_START_ADDR_V);
+		value = window->stride[1] << 16 | window->stride[0];
+		tegra_dc_writel(dc, value, DC_WIN_LINE_STRIDE);
+	} else {
+		tegra_dc_writel(dc, window->stride[0], DC_WIN_LINE_STRIDE);
+	}
+
+	if (window->bottom_up)
+		v_offset += window->src.h - 1;
+
+	tegra_dc_writel(dc, h_offset, DC_WINBUF_ADDR_H_OFFSET);
+	tegra_dc_writel(dc, v_offset, DC_WINBUF_ADDR_V_OFFSET);
+
+	if (window->tiled) {
+		value = DC_WIN_BUFFER_ADDR_MODE_TILE_UV |
+			DC_WIN_BUFFER_ADDR_MODE_TILE;
+	} else {
+		value = DC_WIN_BUFFER_ADDR_MODE_LINEAR_UV |
+			DC_WIN_BUFFER_ADDR_MODE_LINEAR;
+	}
+
+	tegra_dc_writel(dc, value, DC_WIN_BUFFER_ADDR_MODE);
+
+	value = WIN_ENABLE;
+
+	if (yuv) {
+		/* setup default colorspace conversion coefficients */
+		tegra_dc_writel(dc, 0x00f0, DC_WIN_CSC_YOF);
+		tegra_dc_writel(dc, 0x012a, DC_WIN_CSC_KYRGB);
+		tegra_dc_writel(dc, 0x0000, DC_WIN_CSC_KUR);
+		tegra_dc_writel(dc, 0x0198, DC_WIN_CSC_KVR);
+		tegra_dc_writel(dc, 0x039b, DC_WIN_CSC_KUG);
+		tegra_dc_writel(dc, 0x032f, DC_WIN_CSC_KVG);
+		tegra_dc_writel(dc, 0x0204, DC_WIN_CSC_KUB);
+		tegra_dc_writel(dc, 0x0000, DC_WIN_CSC_KVB);
+
+		value |= CSC_ENABLE;
+	} else if (window->bits_per_pixel < 24) {
+		value |= COLOR_EXPAND;
+	}
+
+	if (window->bottom_up)
+		value |= V_DIRECTION;
+
+	tegra_dc_writel(dc, value, DC_WIN_WIN_OPTIONS);
+
+	/*
+	 * Disable blending and assume Window A is the bottom-most window,
+	 * Window C is the top-most window and Window B is in the middle.
+	 */
+	tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_NOKEY);
+	tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_1WIN);
+
+	switch (index) {
+	case 0:
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_2WIN_X);
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_2WIN_Y);
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_3WIN_XY);
+		break;
+
+	case 1:
+		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_2WIN_X);
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_2WIN_Y);
+		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_3WIN_XY);
+		break;
+
+	case 2:
+		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_2WIN_X);
+		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_2WIN_Y);
+		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_3WIN_XY);
+		break;
+	}
+
+	tegra_dc_writel(dc, WIN_A_UPDATE << index, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, WIN_A_ACT_REQ << index, DC_CMD_STATE_CONTROL);
+
+	return 0;
+}
+
 static int tegra_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
 			      struct drm_framebuffer *fb, int crtc_x,
 			      int crtc_y, unsigned int crtc_w,
@@ -45,7 +298,7 @@ static int tegra_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
 	window.dst.y = crtc_y;
 	window.dst.w = crtc_w;
 	window.dst.h = crtc_h;
-	window.format = tegra_dc_format(fb->pixel_format);
+	window.format = tegra_dc_format(fb->pixel_format, &window.swap);
 	window.bits_per_pixel = fb->bits_per_pixel;
 	window.bottom_up = tegra_fb_is_bottom_up(fb);
 	window.tiled = tegra_fb_is_tiled(fb);
@@ -113,6 +366,7 @@ static const uint32_t plane_formats[] = {
 	DRM_FORMAT_XRGB8888,
 	DRM_FORMAT_RGB565,
 	DRM_FORMAT_UYVY,
+	DRM_FORMAT_YUYV,
 	DRM_FORMAT_YUV420,
 	DRM_FORMAT_YUV422,
 };
@@ -146,9 +400,9 @@ static int tegra_dc_add_planes(struct drm_device *drm, struct tegra_dc *dc)
 static int tegra_dc_set_base(struct tegra_dc *dc, int x, int y,
 			     struct drm_framebuffer *fb)
 {
-	unsigned int format = tegra_dc_format(fb->pixel_format);
 	struct tegra_bo *bo = tegra_fb_get_plane(fb, 0);
 	unsigned int h_offset = 0, v_offset = 0;
+	unsigned int format, swap;
 	unsigned long value;
 
 	tegra_dc_writel(dc, WINDOW_A_SELECT, DC_CMD_DISPLAY_WINDOW_HEADER);
@@ -158,7 +412,10 @@ static int tegra_dc_set_base(struct tegra_dc *dc, int x, int y,
 
 	tegra_dc_writel(dc, bo->paddr + value, DC_WINBUF_START_ADDR);
 	tegra_dc_writel(dc, fb->pitches[0], DC_WIN_LINE_STRIDE);
+
+	format = tegra_dc_format(fb->pixel_format, &swap);
 	tegra_dc_writel(dc, format, DC_WIN_COLOR_DEPTH);
+	tegra_dc_writel(dc, swap, DC_WIN_BYTE_SWAP);
 
 	if (tegra_fb_is_tiled(fb)) {
 		value = DC_WIN_BUFFER_ADDR_MODE_TILE_UV |
@@ -173,13 +430,13 @@ static int tegra_dc_set_base(struct tegra_dc *dc, int x, int y,
 	/* make sure bottom-up buffers are properly displayed */
 	if (tegra_fb_is_bottom_up(fb)) {
 		value = tegra_dc_readl(dc, DC_WIN_WIN_OPTIONS);
-		value |= INVERT_V;
+		value |= V_DIRECTION;
 		tegra_dc_writel(dc, value, DC_WIN_WIN_OPTIONS);
 
 		v_offset += fb->height - 1;
 	} else {
 		value = tegra_dc_readl(dc, DC_WIN_WIN_OPTIONS);
-		value &= ~INVERT_V;
+		value &= ~V_DIRECTION;
 		tegra_dc_writel(dc, value, DC_WIN_WIN_OPTIONS);
 	}
 
@@ -221,6 +478,109 @@ void tegra_dc_disable_vblank(struct tegra_dc *dc)
 	spin_unlock_irqrestore(&dc->lock, flags);
 }
 
+static int tegra_dc_cursor_set2(struct drm_crtc *crtc, struct drm_file *file,
+				uint32_t handle, uint32_t width,
+				uint32_t height, int32_t hot_x, int32_t hot_y)
+{
+	unsigned long value = CURSOR_CLIP_DISPLAY;
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	struct drm_gem_object *gem;
+	struct tegra_bo *bo = NULL;
+
+	if (!dc->soc->supports_cursor)
+		return -ENXIO;
+
+	if (width != height)
+		return -EINVAL;
+
+	switch (width) {
+	case 32:
+		value |= CURSOR_SIZE_32x32;
+		break;
+
+	case 64:
+		value |= CURSOR_SIZE_64x64;
+		break;
+
+	case 128:
+		value |= CURSOR_SIZE_128x128;
+
+	case 256:
+		value |= CURSOR_SIZE_256x256;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	if (handle) {
+		gem = drm_gem_object_lookup(crtc->dev, file, handle);
+		if (!gem)
+			return -ENOENT;
+
+		bo = to_tegra_bo(gem);
+	}
+
+	if (bo) {
+		unsigned long addr = (bo->paddr & 0xfffffc00) >> 10;
+#ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
+		unsigned long high = (bo->paddr & 0xfffffffc) >> 32;
+#endif
+
+		tegra_dc_writel(dc, value | addr, DC_DISP_CURSOR_START_ADDR);
+
+#ifdef CONFIG_ARCH_DMA_ADDR_T_64BIT
+		tegra_dc_writel(dc, high, DC_DISP_CURSOR_START_ADDR_HI);
+#endif
+
+		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+		value |= CURSOR_ENABLE;
+		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+		value = tegra_dc_readl(dc, DC_DISP_BLEND_CURSOR_CONTROL);
+		value &= ~CURSOR_DST_BLEND_MASK;
+		value &= ~CURSOR_SRC_BLEND_MASK;
+		value |= CURSOR_MODE_NORMAL;
+		value |= CURSOR_DST_BLEND_NEG_K1_TIMES_SRC;
+		value |= CURSOR_SRC_BLEND_K1_TIMES_SRC;
+		value |= CURSOR_ALPHA;
+		tegra_dc_writel(dc, value, DC_DISP_BLEND_CURSOR_CONTROL);
+	} else {
+		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+		value &= ~CURSOR_ENABLE;
+		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+	}
+
+	tegra_dc_writel(dc, CURSOR_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, CURSOR_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	return 0;
+}
+
+static int tegra_dc_cursor_move(struct drm_crtc *crtc, int x, int y)
+{
+	struct tegra_dc *dc = to_tegra_dc(crtc);
+	unsigned long value;
+
+	if (!dc->soc->supports_cursor)
+		return -ENXIO;
+
+	value = ((y & 0x3fff) << 16) | (x & 0x3fff);
+	tegra_dc_writel(dc, value, DC_DISP_CURSOR_POSITION);
+
+	tegra_dc_writel(dc, CURSOR_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, CURSOR_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	/* XXX: only required on generations earlier than Tegra124? */
+	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	return 0;
+}
+
 static void tegra_dc_finish_page_flip(struct tegra_dc *dc)
 {
 	struct drm_device *drm = dc->base.dev;
@@ -231,14 +591,14 @@ static void tegra_dc_finish_page_flip(struct tegra_dc *dc)
 	if (!dc->event)
 		return;
 
-	bo = tegra_fb_get_plane(crtc->fb, 0);
+	bo = tegra_fb_get_plane(crtc->primary->fb, 0);
 
 	/* check if new start address has been latched */
 	tegra_dc_writel(dc, READ_MUX, DC_CMD_STATE_ACCESS);
 	base = tegra_dc_readl(dc, DC_WINBUF_START_ADDR);
 	tegra_dc_writel(dc, 0, DC_CMD_STATE_ACCESS);
 
-	if (base == bo->paddr + crtc->fb->offsets[0]) {
+	if (base == bo->paddr + crtc->primary->fb->offsets[0]) {
 		spin_lock_irqsave(&drm->event_lock, flags);
 		drm_send_vblank_event(drm, dc->pipe, dc->event);
 		drm_vblank_put(drm, dc->pipe);
@@ -280,7 +640,7 @@ static int tegra_dc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	}
 
 	tegra_dc_set_base(dc, 0, 0, fb);
-	crtc->fb = fb;
+	crtc->primary->fb = fb;
 
 	return 0;
 }
@@ -297,6 +657,8 @@ static void tegra_dc_destroy(struct drm_crtc *crtc)
 }
 
 static const struct drm_crtc_funcs tegra_crtc_funcs = {
+	.cursor_set2 = tegra_dc_cursor_set2,
+	.cursor_move = tegra_dc_cursor_move,
 	.page_flip = tegra_dc_page_flip,
 	.set_config = drm_crtc_helper_set_config,
 	.destroy = tegra_dc_destroy,
@@ -308,7 +670,7 @@ static void tegra_crtc_disable(struct drm_crtc *crtc)
 	struct drm_device *drm = crtc->dev;
 	struct drm_plane *plane;
 
-	list_for_each_entry(plane, &drm->mode_config.plane_list, head) {
+	drm_for_each_legacy_plane(plane, &drm->mode_config.plane_list) {
 		if (plane->crtc == crtc) {
 			tegra_plane_disable(plane);
 			plane->crtc = NULL;
@@ -330,52 +692,11 @@ static bool tegra_crtc_mode_fixup(struct drm_crtc *crtc,
 	return true;
 }
 
-static inline u32 compute_dda_inc(unsigned int in, unsigned int out, bool v,
-				  unsigned int bpp)
-{
-	fixed20_12 outf = dfixed_init(out);
-	fixed20_12 inf = dfixed_init(in);
-	u32 dda_inc;
-	int max;
-
-	if (v)
-		max = 15;
-	else {
-		switch (bpp) {
-		case 2:
-			max = 8;
-			break;
-
-		default:
-			WARN_ON_ONCE(1);
-			/* fallthrough */
-		case 4:
-			max = 4;
-			break;
-		}
-	}
-
-	outf.full = max_t(u32, outf.full - dfixed_const(1), dfixed_const(1));
-	inf.full -= dfixed_const(1);
-
-	dda_inc = dfixed_div(inf, outf);
-	dda_inc = min_t(u32, dda_inc, dfixed_const(max));
-
-	return dda_inc;
-}
-
-static inline u32 compute_initial_dda(unsigned int in)
-{
-	fixed20_12 inf = dfixed_init(in);
-	return dfixed_frac(inf);
-}
-
 static int tegra_dc_set_timings(struct tegra_dc *dc,
 				struct drm_display_mode *mode)
 {
-	/* TODO: For HDMI compliance, h & v ref_to_sync should be set to 1 */
-	unsigned int h_ref_to_sync = 0;
-	unsigned int v_ref_to_sync = 0;
+	unsigned int h_ref_to_sync = 1;
+	unsigned int v_ref_to_sync = 1;
 	unsigned long value;
 
 	tegra_dc_writel(dc, 0x0, DC_DISP_DISP_TIMING_OPTIONS);
@@ -402,13 +723,14 @@ static int tegra_dc_set_timings(struct tegra_dc *dc,
 }
 
 static int tegra_crtc_setup_clk(struct drm_crtc *crtc,
-				struct drm_display_mode *mode,
-				unsigned long *div)
+				struct drm_display_mode *mode)
 {
-	unsigned long pclk = mode->clock * 1000, rate;
+	unsigned long pclk = mode->clock * 1000;
 	struct tegra_dc *dc = to_tegra_dc(crtc);
 	struct tegra_output *output = NULL;
 	struct drm_encoder *encoder;
+	unsigned int div;
+	u32 value;
 	long err;
 
 	list_for_each_entry(encoder, &crtc->dev->mode_config.encoder_list, head)
@@ -421,235 +743,37 @@ static int tegra_crtc_setup_clk(struct drm_crtc *crtc,
 		return -ENODEV;
 
 	/*
-	 * This assumes that the display controller will divide its parent
-	 * clock by 2 to generate the pixel clock.
+	 * This assumes that the parent clock is pll_d_out0 or pll_d2_out
+	 * respectively, each of which divides the base pll_d by 2.
 	 */
-	err = tegra_output_setup_clock(output, dc->clk, pclk * 2);
+	err = tegra_output_setup_clock(output, dc->clk, pclk, &div);
 	if (err < 0) {
 		dev_err(dc->dev, "failed to setup clock: %ld\n", err);
 		return err;
 	}
 
-	rate = clk_get_rate(dc->clk);
-	*div = (rate * 2 / pclk) - 2;
-
-	DRM_DEBUG_KMS("rate: %lu, div: %lu\n", rate, *div);
-
-	return 0;
-}
-
-static bool tegra_dc_format_is_yuv(unsigned int format, bool *planar)
-{
-	switch (format) {
-	case WIN_COLOR_DEPTH_YCbCr422:
-	case WIN_COLOR_DEPTH_YUV422:
-		if (planar)
-			*planar = false;
-
-		return true;
-
-	case WIN_COLOR_DEPTH_YCbCr420P:
-	case WIN_COLOR_DEPTH_YUV420P:
-	case WIN_COLOR_DEPTH_YCbCr422P:
-	case WIN_COLOR_DEPTH_YUV422P:
-	case WIN_COLOR_DEPTH_YCbCr422R:
-	case WIN_COLOR_DEPTH_YUV422R:
-	case WIN_COLOR_DEPTH_YCbCr422RA:
-	case WIN_COLOR_DEPTH_YUV422RA:
-		if (planar)
-			*planar = true;
-
-		return true;
-	}
-
-	return false;
-}
-
-int tegra_dc_setup_window(struct tegra_dc *dc, unsigned int index,
-			  const struct tegra_dc_window *window)
-{
-	unsigned h_offset, v_offset, h_size, v_size, h_dda, v_dda, bpp;
-	unsigned long value;
-	bool yuv, planar;
-
-	/*
-	 * For YUV planar modes, the number of bytes per pixel takes into
-	 * account only the luma component and therefore is 1.
-	 */
-	yuv = tegra_dc_format_is_yuv(window->format, &planar);
-	if (!yuv)
-		bpp = window->bits_per_pixel / 8;
-	else
-		bpp = planar ? 1 : 2;
-
-	value = WINDOW_A_SELECT << index;
-	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_WINDOW_HEADER);
-
-	tegra_dc_writel(dc, window->format, DC_WIN_COLOR_DEPTH);
-	tegra_dc_writel(dc, 0, DC_WIN_BYTE_SWAP);
-
-	value = V_POSITION(window->dst.y) | H_POSITION(window->dst.x);
-	tegra_dc_writel(dc, value, DC_WIN_POSITION);
-
-	value = V_SIZE(window->dst.h) | H_SIZE(window->dst.w);
-	tegra_dc_writel(dc, value, DC_WIN_SIZE);
-
-	h_offset = window->src.x * bpp;
-	v_offset = window->src.y;
-	h_size = window->src.w * bpp;
-	v_size = window->src.h;
-
-	value = V_PRESCALED_SIZE(v_size) | H_PRESCALED_SIZE(h_size);
-	tegra_dc_writel(dc, value, DC_WIN_PRESCALED_SIZE);
-
-	/*
-	 * For DDA computations the number of bytes per pixel for YUV planar
-	 * modes needs to take into account all Y, U and V components.
-	 */
-	if (yuv && planar)
-		bpp = 2;
-
-	h_dda = compute_dda_inc(window->src.w, window->dst.w, false, bpp);
-	v_dda = compute_dda_inc(window->src.h, window->dst.h, true, bpp);
-
-	value = V_DDA_INC(v_dda) | H_DDA_INC(h_dda);
-	tegra_dc_writel(dc, value, DC_WIN_DDA_INC);
+	DRM_DEBUG_KMS("rate: %lu, div: %u\n", clk_get_rate(dc->clk), div);
 
-	h_dda = compute_initial_dda(window->src.x);
-	v_dda = compute_initial_dda(window->src.y);
-
-	tegra_dc_writel(dc, h_dda, DC_WIN_H_INITIAL_DDA);
-	tegra_dc_writel(dc, v_dda, DC_WIN_V_INITIAL_DDA);
-
-	tegra_dc_writel(dc, 0, DC_WIN_UV_BUF_STRIDE);
-	tegra_dc_writel(dc, 0, DC_WIN_BUF_STRIDE);
-
-	tegra_dc_writel(dc, window->base[0], DC_WINBUF_START_ADDR);
-
-	if (yuv && planar) {
-		tegra_dc_writel(dc, window->base[1], DC_WINBUF_START_ADDR_U);
-		tegra_dc_writel(dc, window->base[2], DC_WINBUF_START_ADDR_V);
-		value = window->stride[1] << 16 | window->stride[0];
-		tegra_dc_writel(dc, value, DC_WIN_LINE_STRIDE);
-	} else {
-		tegra_dc_writel(dc, window->stride[0], DC_WIN_LINE_STRIDE);
-	}
-
-	if (window->bottom_up)
-		v_offset += window->src.h - 1;
-
-	tegra_dc_writel(dc, h_offset, DC_WINBUF_ADDR_H_OFFSET);
-	tegra_dc_writel(dc, v_offset, DC_WINBUF_ADDR_V_OFFSET);
-
-	if (window->tiled) {
-		value = DC_WIN_BUFFER_ADDR_MODE_TILE_UV |
-			DC_WIN_BUFFER_ADDR_MODE_TILE;
-	} else {
-		value = DC_WIN_BUFFER_ADDR_MODE_LINEAR_UV |
-			DC_WIN_BUFFER_ADDR_MODE_LINEAR;
-	}
-
-	tegra_dc_writel(dc, value, DC_WIN_BUFFER_ADDR_MODE);
-
-	value = WIN_ENABLE;
-
-	if (yuv) {
-		/* setup default colorspace conversion coefficients */
-		tegra_dc_writel(dc, 0x00f0, DC_WIN_CSC_YOF);
-		tegra_dc_writel(dc, 0x012a, DC_WIN_CSC_KYRGB);
-		tegra_dc_writel(dc, 0x0000, DC_WIN_CSC_KUR);
-		tegra_dc_writel(dc, 0x0198, DC_WIN_CSC_KVR);
-		tegra_dc_writel(dc, 0x039b, DC_WIN_CSC_KUG);
-		tegra_dc_writel(dc, 0x032f, DC_WIN_CSC_KVG);
-		tegra_dc_writel(dc, 0x0204, DC_WIN_CSC_KUB);
-		tegra_dc_writel(dc, 0x0000, DC_WIN_CSC_KVB);
-
-		value |= CSC_ENABLE;
-	} else if (window->bits_per_pixel < 24) {
-		value |= COLOR_EXPAND;
-	}
-
-	if (window->bottom_up)
-		value |= INVERT_V;
-
-	tegra_dc_writel(dc, value, DC_WIN_WIN_OPTIONS);
-
-	/*
-	 * Disable blending and assume Window A is the bottom-most window,
-	 * Window C is the top-most window and Window B is in the middle.
-	 */
-	tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_NOKEY);
-	tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_1WIN);
-
-	switch (index) {
-	case 0:
-		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_2WIN_X);
-		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_2WIN_Y);
-		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_3WIN_XY);
-		break;
-
-	case 1:
-		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_2WIN_X);
-		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_2WIN_Y);
-		tegra_dc_writel(dc, 0x000000, DC_WIN_BLEND_3WIN_XY);
-		break;
-
-	case 2:
-		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_2WIN_X);
-		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_2WIN_Y);
-		tegra_dc_writel(dc, 0xffff00, DC_WIN_BLEND_3WIN_XY);
-		break;
-	}
-
-	tegra_dc_writel(dc, WIN_A_UPDATE << index, DC_CMD_STATE_CONTROL);
-	tegra_dc_writel(dc, WIN_A_ACT_REQ << index, DC_CMD_STATE_CONTROL);
+	value = SHIFT_CLK_DIVIDER(div) | PIXEL_CLK_DIVIDER_PCD1;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_CLOCK_CONTROL);
 
 	return 0;
 }
 
-unsigned int tegra_dc_format(uint32_t format)
-{
-	switch (format) {
-	case DRM_FORMAT_XBGR8888:
-		return WIN_COLOR_DEPTH_R8G8B8A8;
-
-	case DRM_FORMAT_XRGB8888:
-		return WIN_COLOR_DEPTH_B8G8R8A8;
-
-	case DRM_FORMAT_RGB565:
-		return WIN_COLOR_DEPTH_B5G6R5;
-
-	case DRM_FORMAT_UYVY:
-		return WIN_COLOR_DEPTH_YCbCr422;
-
-	case DRM_FORMAT_YUV420:
-		return WIN_COLOR_DEPTH_YCbCr420P;
-
-	case DRM_FORMAT_YUV422:
-		return WIN_COLOR_DEPTH_YCbCr422P;
-
-	default:
-		break;
-	}
-
-	WARN(1, "unsupported pixel format %u, using default\n", format);
-	return WIN_COLOR_DEPTH_B8G8R8A8;
-}
-
 static int tegra_crtc_mode_set(struct drm_crtc *crtc,
 			       struct drm_display_mode *mode,
 			       struct drm_display_mode *adjusted,
 			       int x, int y, struct drm_framebuffer *old_fb)
 {
-	struct tegra_bo *bo = tegra_fb_get_plane(crtc->fb, 0);
+	struct tegra_bo *bo = tegra_fb_get_plane(crtc->primary->fb, 0);
 	struct tegra_dc *dc = to_tegra_dc(crtc);
 	struct tegra_dc_window window;
-	unsigned long div, value;
+	u32 value;
 	int err;
 
 	drm_vblank_pre_modeset(crtc->dev, dc->pipe);
 
-	err = tegra_crtc_setup_clk(crtc, mode, &div);
+	err = tegra_crtc_setup_clk(crtc, mode);
 	if (err) {
 		dev_err(dc->dev, "failed to setup clock for CRTC: %d\n", err);
 		return err;
@@ -658,22 +782,12 @@ static int tegra_crtc_mode_set(struct drm_crtc *crtc,
 	/* program display mode */
 	tegra_dc_set_timings(dc, mode);
 
-	value = DE_SELECT_ACTIVE | DE_CONTROL_NORMAL;
-	tegra_dc_writel(dc, value, DC_DISP_DATA_ENABLE_OPTIONS);
-
-	value = tegra_dc_readl(dc, DC_COM_PIN_OUTPUT_POLARITY(1));
-	value &= ~LVS_OUTPUT_POLARITY_LOW;
-	value &= ~LHS_OUTPUT_POLARITY_LOW;
-	tegra_dc_writel(dc, value, DC_COM_PIN_OUTPUT_POLARITY(1));
-
-	value = DISP_DATA_FORMAT_DF1P1C | DISP_ALIGNMENT_MSB |
-		DISP_ORDER_RED_BLUE;
-	tegra_dc_writel(dc, value, DC_DISP_DISP_INTERFACE_CONTROL);
-
-	tegra_dc_writel(dc, 0x00010001, DC_DISP_SHIFT_CLOCK_OPTIONS);
-
-	value = SHIFT_CLK_DIVIDER(div) | PIXEL_CLK_DIVIDER_PCD1;
-	tegra_dc_writel(dc, value, DC_DISP_DISP_CLOCK_CONTROL);
+	/* interlacing isn't supported yet, so disable it */
+	if (dc->soc->supports_interlacing) {
+		value = tegra_dc_readl(dc, DC_DISP_INTERLACE_CONTROL);
+		value &= ~INTERLACE_ENABLE;
+		tegra_dc_writel(dc, value, DC_DISP_INTERLACE_CONTROL);
+	}
 
 	/* setup window parameters */
 	memset(&window, 0, sizeof(window));
@@ -685,9 +799,10 @@ static int tegra_crtc_mode_set(struct drm_crtc *crtc,
 	window.dst.y = 0;
 	window.dst.w = mode->hdisplay;
 	window.dst.h = mode->vdisplay;
-	window.format = tegra_dc_format(crtc->fb->pixel_format);
-	window.bits_per_pixel = crtc->fb->bits_per_pixel;
-	window.stride[0] = crtc->fb->pitches[0];
+	window.format = tegra_dc_format(crtc->primary->fb->pixel_format,
+					&window.swap);
+	window.bits_per_pixel = crtc->primary->fb->bits_per_pixel;
+	window.stride[0] = crtc->primary->fb->pitches[0];
 	window.base[0] = bo->paddr;
 
 	err = tegra_dc_setup_window(dc, 0, &window);
@@ -702,7 +817,7 @@ static int tegra_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 {
 	struct tegra_dc *dc = to_tegra_dc(crtc);
 
-	return tegra_dc_set_base(dc, x, y, crtc->fb);
+	return tegra_dc_set_base(dc, x, y, crtc->primary->fb);
 }
 
 static void tegra_crtc_prepare(struct drm_crtc *crtc)
@@ -712,7 +827,7 @@ static void tegra_crtc_prepare(struct drm_crtc *crtc)
 	unsigned long value;
 
 	/* hardware initialization */
-	tegra_periph_reset_deassert(dc->clk);
+	reset_control_deassert(dc->rst);
 	usleep_range(10000, 20000);
 
 	if (dc->pipe)
@@ -731,14 +846,6 @@ static void tegra_crtc_prepare(struct drm_crtc *crtc)
 		WIN_A_OF_INT | WIN_B_OF_INT | WIN_C_OF_INT;
 	tegra_dc_writel(dc, value, DC_CMD_INT_POLARITY);
 
-	value = PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
-		PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
-	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
-
-	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
-	value |= DISP_CTRL_MODE_C_DISPLAY;
-	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
-
 	/* initialize timer */
 	value = CURSOR_THRESHOLD(0) | WINDOW_A_THRESHOLD(0x20) |
 		WINDOW_B_THRESHOLD(0x20) | WINDOW_C_THRESHOLD(0x20);
@@ -998,6 +1105,8 @@ static int tegra_dc_show_regs(struct seq_file *s, void *data)
 	DUMP_REG(DC_DISP_SD_BL_CONTROL);
 	DUMP_REG(DC_DISP_SD_HW_K_VALUES);
 	DUMP_REG(DC_DISP_SD_MAN_K_VALUES);
+	DUMP_REG(DC_DISP_CURSOR_START_ADDR_HI);
+	DUMP_REG(DC_DISP_BLEND_CURSOR_CONTROL);
 	DUMP_REG(DC_WIN_WIN_OPTIONS);
 	DUMP_REG(DC_WIN_BYTE_SWAP);
 	DUMP_REG(DC_WIN_BUFFER_CONTROL);
@@ -1103,28 +1212,26 @@ static int tegra_dc_debugfs_exit(struct tegra_dc *dc)
 
 static int tegra_dc_init(struct host1x_client *client)
 {
-	struct tegra_drm *tegra = dev_get_drvdata(client->parent);
+	struct drm_device *drm = dev_get_drvdata(client->parent);
 	struct tegra_dc *dc = host1x_client_to_dc(client);
 	int err;
 
-	dc->pipe = tegra->drm->mode_config.num_crtc;
-
-	drm_crtc_init(tegra->drm, &dc->base, &tegra_crtc_funcs);
+	drm_crtc_init(drm, &dc->base, &tegra_crtc_funcs);
 	drm_mode_crtc_set_gamma_size(&dc->base, 256);
 	drm_crtc_helper_add(&dc->base, &tegra_crtc_helper_funcs);
 
-	err = tegra_dc_rgb_init(tegra->drm, dc);
+	err = tegra_dc_rgb_init(drm, dc);
 	if (err < 0 && err != -ENODEV) {
 		dev_err(dc->dev, "failed to initialize RGB output: %d\n", err);
 		return err;
 	}
 
-	err = tegra_dc_add_planes(tegra->drm, dc);
+	err = tegra_dc_add_planes(drm, dc);
 	if (err < 0)
 		return err;
 
 	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
-		err = tegra_dc_debugfs_init(dc, tegra->drm->primary);
+		err = tegra_dc_debugfs_init(dc, drm->primary);
 		if (err < 0)
 			dev_err(dc->dev, "debugfs setup failed: %d\n", err);
 	}
@@ -1167,8 +1274,74 @@ static const struct host1x_client_ops dc_client_ops = {
 	.exit = tegra_dc_exit,
 };
 
+static const struct tegra_dc_soc_info tegra20_dc_soc_info = {
+	.supports_interlacing = false,
+	.supports_cursor = false,
+};
+
+static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
+	.supports_interlacing = false,
+	.supports_cursor = false,
+};
+
+static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
+	.supports_interlacing = true,
+	.supports_cursor = true,
+};
+
+static const struct of_device_id tegra_dc_of_match[] = {
+	{
+		.compatible = "nvidia,tegra124-dc",
+		.data = &tegra124_dc_soc_info,
+	}, {
+		.compatible = "nvidia,tegra30-dc",
+		.data = &tegra30_dc_soc_info,
+	}, {
+		.compatible = "nvidia,tegra20-dc",
+		.data = &tegra20_dc_soc_info,
+	}, {
+		/* sentinel */
+	}
+};
+
+static int tegra_dc_parse_dt(struct tegra_dc *dc)
+{
+	struct device_node *np;
+	u32 value = 0;
+	int err;
+
+	err = of_property_read_u32(dc->dev->of_node, "nvidia,head", &value);
+	if (err < 0) {
+		dev_err(dc->dev, "missing \"nvidia,head\" property\n");
+
+		/*
+		 * If the nvidia,head property isn't present, try to find the
+		 * correct head number by looking up the position of this
+		 * display controller's node within the device tree. Assuming
+		 * that the nodes are ordered properly in the DTS file and
+		 * that the translation into a flattened device tree blob
+		 * preserves that ordering this will actually yield the right
+		 * head number.
+		 *
+		 * If those assumptions don't hold, this will still work for
+		 * cases where only a single display controller is used.
+		 */
+		for_each_matching_node(np, tegra_dc_of_match) {
+			if (np == dc->dev->of_node)
+				break;
+
+			value++;
+		}
+	}
+
+	dc->pipe = value;
+
+	return 0;
+}
+
 static int tegra_dc_probe(struct platform_device *pdev)
 {
+	const struct of_device_id *id;
 	struct resource *regs;
 	struct tegra_dc *dc;
 	int err;
@@ -1177,9 +1350,18 @@ static int tegra_dc_probe(struct platform_device *pdev)
 	if (!dc)
 		return -ENOMEM;
 
+	id = of_match_node(tegra_dc_of_match, pdev->dev.of_node);
+	if (!id)
+		return -ENODEV;
+
 	spin_lock_init(&dc->lock);
 	INIT_LIST_HEAD(&dc->list);
 	dc->dev = &pdev->dev;
+	dc->soc = id->data;
+
+	err = tegra_dc_parse_dt(dc);
+	if (err < 0)
+		return err;
 
 	dc->clk = devm_clk_get(&pdev->dev, NULL);
 	if (IS_ERR(dc->clk)) {
@@ -1187,6 +1369,12 @@ static int tegra_dc_probe(struct platform_device *pdev)
 		return PTR_ERR(dc->clk);
 	}
 
+	dc->rst = devm_reset_control_get(&pdev->dev, "dc");
+	if (IS_ERR(dc->rst)) {
+		dev_err(&pdev->dev, "failed to get reset\n");
+		return PTR_ERR(dc->rst);
+	}
+
 	err = clk_prepare_enable(dc->clk);
 	if (err < 0)
 		return err;
@@ -1247,12 +1435,6 @@ static int tegra_dc_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static struct of_device_id tegra_dc_of_match[] = {
-	{ .compatible = "nvidia,tegra30-dc", },
-	{ .compatible = "nvidia,tegra20-dc", },
-	{ },
-};
-
 struct platform_driver tegra_dc_driver = {
 	.driver = {
 		.name = "tegra-dc",
diff --git a/drivers/gpu/drm/tegra/dc.h b/drivers/gpu/drm/tegra/dc.h
index 91bbda29147..78c5feff95d 100644
--- a/drivers/gpu/drm/tegra/dc.h
+++ b/drivers/gpu/drm/tegra/dc.h
@@ -28,6 +28,7 @@
 #define DISP_CTRL_MODE_STOP (0 << 5)
 #define DISP_CTRL_MODE_C_DISPLAY (1 << 5)
 #define DISP_CTRL_MODE_NC_DISPLAY (2 << 5)
+#define DISP_CTRL_MODE_MASK (3 << 5)
 #define DC_CMD_SIGNAL_RAISE			0x033
 #define DC_CMD_DISPLAY_POWER_CONTROL		0x036
 #define PW0_ENABLE (1 <<  0)
@@ -66,10 +67,12 @@
 #define WIN_A_ACT_REQ   (1 <<  1)
 #define WIN_B_ACT_REQ   (1 <<  2)
 #define WIN_C_ACT_REQ   (1 <<  3)
+#define CURSOR_ACT_REQ  (1 <<  7)
 #define GENERAL_UPDATE  (1 <<  8)
 #define WIN_A_UPDATE    (1 <<  9)
 #define WIN_B_UPDATE    (1 << 10)
 #define WIN_C_UPDATE    (1 << 11)
+#define CURSOR_UPDATE   (1 << 15)
 #define NC_HOST_TRIG    (1 << 24)
 
 #define DC_CMD_DISPLAY_WINDOW_HEADER		0x042
@@ -115,7 +118,10 @@
 #define DC_DISP_DISP_SIGNAL_OPTIONS1		0x401
 
 #define DC_DISP_DISP_WIN_OPTIONS		0x402
-#define HDMI_ENABLE (1 << 30)
+#define HDMI_ENABLE	(1 << 30)
+#define DSI_ENABLE	(1 << 29)
+#define SOR_ENABLE	(1 << 25)
+#define CURSOR_ENABLE	(1 << 16)
 
 #define DC_DISP_DISP_MEM_HIGH_PRIORITY		0x403
 #define CURSOR_THRESHOLD(x)   (((x) & 0x03) << 24)
@@ -238,6 +244,8 @@
 #define DITHER_CONTROL_ERRDIFF (3 << 8)
 
 #define DC_DISP_SHIFT_CLOCK_OPTIONS		0x431
+#define  SC1_H_QUALIFIER_NONE	(1 << 16)
+#define  SC0_H_QUALIFIER_NONE	(1 <<  0)
 
 #define DC_DISP_DATA_ENABLE_OPTIONS		0x432
 #define DE_SELECT_ACTIVE_BLANK  (0 << 0)
@@ -261,6 +269,14 @@
 #define DC_DISP_CURSOR_BACKGROUND		0x43d
 
 #define DC_DISP_CURSOR_START_ADDR		0x43e
+#define CURSOR_CLIP_DISPLAY	(0 << 28)
+#define CURSOR_CLIP_WIN_A	(1 << 28)
+#define CURSOR_CLIP_WIN_B	(2 << 28)
+#define CURSOR_CLIP_WIN_C	(3 << 28)
+#define CURSOR_SIZE_32x32	(0 << 24)
+#define CURSOR_SIZE_64x64	(1 << 24)
+#define CURSOR_SIZE_128x128	(2 << 24)
+#define CURSOR_SIZE_256x256	(3 << 24)
 #define DC_DISP_CURSOR_START_ADDR_NS		0x43f
 
 #define DC_DISP_CURSOR_POSITION			0x440
@@ -292,6 +308,24 @@
 #define DC_DISP_SD_HW_K_VALUES			0x4dd
 #define DC_DISP_SD_MAN_K_VALUES			0x4de
 
+#define DC_DISP_INTERLACE_CONTROL		0x4e5
+#define  INTERLACE_STATUS (1 << 2)
+#define  INTERLACE_START  (1 << 1)
+#define  INTERLACE_ENABLE (1 << 0)
+
+#define DC_DISP_CURSOR_START_ADDR_HI		0x4ec
+#define DC_DISP_BLEND_CURSOR_CONTROL		0x4f1
+#define CURSOR_MODE_LEGACY			(0 << 24)
+#define CURSOR_MODE_NORMAL			(1 << 24)
+#define CURSOR_DST_BLEND_ZERO			(0 << 16)
+#define CURSOR_DST_BLEND_K1			(1 << 16)
+#define CURSOR_DST_BLEND_NEG_K1_TIMES_SRC	(2 << 16)
+#define CURSOR_DST_BLEND_MASK			(3 << 16)
+#define CURSOR_SRC_BLEND_K1			(0 << 8)
+#define CURSOR_SRC_BLEND_K1_TIMES_SRC		(1 << 8)
+#define CURSOR_SRC_BLEND_MASK			(3 << 8)
+#define CURSOR_ALPHA				0xff
+
 #define DC_WIN_CSC_YOF				0x611
 #define DC_WIN_CSC_KYRGB			0x612
 #define DC_WIN_CSC_KUR				0x613
@@ -302,7 +336,8 @@
 #define DC_WIN_CSC_KVB				0x618
 
 #define DC_WIN_WIN_OPTIONS			0x700
-#define INVERT_V     (1 <<  2)
+#define H_DIRECTION  (1 <<  0)
+#define V_DIRECTION  (1 <<  2)
 #define COLOR_EXPAND (1 <<  6)
 #define CSC_ENABLE   (1 << 18)
 #define WIN_ENABLE   (1 << 30)
diff --git a/drivers/gpu/drm/tegra/dpaux.c b/drivers/gpu/drm/tegra/dpaux.c
new file mode 100644
index 00000000000..3f132e356e9
--- /dev/null
+++ b/drivers/gpu/drm/tegra/dpaux.c
@@ -0,0 +1,573 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/gpio.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/of_gpio.h>
+#include <linux/platform_device.h>
+#include <linux/reset.h>
+#include <linux/regulator/consumer.h>
+#include <linux/workqueue.h>
+
+#include <drm/drm_dp_helper.h>
+#include <drm/drm_panel.h>
+
+#include "dpaux.h"
+#include "drm.h"
+
+static DEFINE_MUTEX(dpaux_lock);
+static LIST_HEAD(dpaux_list);
+
+struct tegra_dpaux {
+	struct drm_dp_aux aux;
+	struct device *dev;
+
+	void __iomem *regs;
+	int irq;
+
+	struct tegra_output *output;
+
+	struct reset_control *rst;
+	struct clk *clk_parent;
+	struct clk *clk;
+
+	struct regulator *vdd;
+
+	struct completion complete;
+	struct work_struct work;
+	struct list_head list;
+};
+
+static inline struct tegra_dpaux *to_dpaux(struct drm_dp_aux *aux)
+{
+	return container_of(aux, struct tegra_dpaux, aux);
+}
+
+static inline struct tegra_dpaux *work_to_dpaux(struct work_struct *work)
+{
+	return container_of(work, struct tegra_dpaux, work);
+}
+
+static inline unsigned long tegra_dpaux_readl(struct tegra_dpaux *dpaux,
+					      unsigned long offset)
+{
+	return readl(dpaux->regs + (offset << 2));
+}
+
+static inline void tegra_dpaux_writel(struct tegra_dpaux *dpaux,
+				      unsigned long value,
+				      unsigned long offset)
+{
+	writel(value, dpaux->regs + (offset << 2));
+}
+
+static void tegra_dpaux_write_fifo(struct tegra_dpaux *dpaux, const u8 *buffer,
+				   size_t size)
+{
+	unsigned long offset = DPAUX_DP_AUXDATA_WRITE(0);
+	size_t i, j;
+
+	for (i = 0; i < size; i += 4) {
+		size_t num = min_t(size_t, size - i, 4);
+		unsigned long value = 0;
+
+		for (j = 0; j < num; j++)
+			value |= buffer[i + j] << (j * 8);
+
+		tegra_dpaux_writel(dpaux, value, offset++);
+	}
+}
+
+static void tegra_dpaux_read_fifo(struct tegra_dpaux *dpaux, u8 *buffer,
+				  size_t size)
+{
+	unsigned long offset = DPAUX_DP_AUXDATA_READ(0);
+	size_t i, j;
+
+	for (i = 0; i < size; i += 4) {
+		size_t num = min_t(size_t, size - i, 4);
+		unsigned long value;
+
+		value = tegra_dpaux_readl(dpaux, offset++);
+
+		for (j = 0; j < num; j++)
+			buffer[i + j] = value >> (j * 8);
+	}
+}
+
+static ssize_t tegra_dpaux_transfer(struct drm_dp_aux *aux,
+				    struct drm_dp_aux_msg *msg)
+{
+	unsigned long timeout = msecs_to_jiffies(250);
+	struct tegra_dpaux *dpaux = to_dpaux(aux);
+	unsigned long status;
+	ssize_t ret = 0;
+	u32 value;
+
+	/* Tegra has 4x4 byte DP AUX transmit and receive FIFOs. */
+	if (msg->size > 16)
+		return -EINVAL;
+
+	/*
+	 * Allow zero-sized messages only for I2C, in which case they specify
+	 * address-only transactions.
+	 */
+	if (msg->size < 1) {
+		switch (msg->request & ~DP_AUX_I2C_MOT) {
+		case DP_AUX_I2C_WRITE:
+		case DP_AUX_I2C_READ:
+			value = DPAUX_DP_AUXCTL_CMD_ADDRESS_ONLY;
+			break;
+
+		default:
+			return -EINVAL;
+		}
+	} else {
+		/* For non-zero-sized messages, set the CMDLEN field. */
+		value = DPAUX_DP_AUXCTL_CMDLEN(msg->size - 1);
+	}
+
+	switch (msg->request & ~DP_AUX_I2C_MOT) {
+	case DP_AUX_I2C_WRITE:
+		if (msg->request & DP_AUX_I2C_MOT)
+			value |= DPAUX_DP_AUXCTL_CMD_MOT_WR;
+		else
+			value |= DPAUX_DP_AUXCTL_CMD_I2C_WR;
+
+		break;
+
+	case DP_AUX_I2C_READ:
+		if (msg->request & DP_AUX_I2C_MOT)
+			value |= DPAUX_DP_AUXCTL_CMD_MOT_RD;
+		else
+			value |= DPAUX_DP_AUXCTL_CMD_I2C_RD;
+
+		break;
+
+	case DP_AUX_I2C_STATUS:
+		if (msg->request & DP_AUX_I2C_MOT)
+			value |= DPAUX_DP_AUXCTL_CMD_MOT_RQ;
+		else
+			value |= DPAUX_DP_AUXCTL_CMD_I2C_RQ;
+
+		break;
+
+	case DP_AUX_NATIVE_WRITE:
+		value |= DPAUX_DP_AUXCTL_CMD_AUX_WR;
+		break;
+
+	case DP_AUX_NATIVE_READ:
+		value |= DPAUX_DP_AUXCTL_CMD_AUX_RD;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	tegra_dpaux_writel(dpaux, msg->address, DPAUX_DP_AUXADDR);
+	tegra_dpaux_writel(dpaux, value, DPAUX_DP_AUXCTL);
+
+	if ((msg->request & DP_AUX_I2C_READ) == 0) {
+		tegra_dpaux_write_fifo(dpaux, msg->buffer, msg->size);
+		ret = msg->size;
+	}
+
+	/* start transaction */
+	value = tegra_dpaux_readl(dpaux, DPAUX_DP_AUXCTL);
+	value |= DPAUX_DP_AUXCTL_TRANSACTREQ;
+	tegra_dpaux_writel(dpaux, value, DPAUX_DP_AUXCTL);
+
+	status = wait_for_completion_timeout(&dpaux->complete, timeout);
+	if (!status)
+		return -ETIMEDOUT;
+
+	/* read status and clear errors */
+	value = tegra_dpaux_readl(dpaux, DPAUX_DP_AUXSTAT);
+	tegra_dpaux_writel(dpaux, 0xf00, DPAUX_DP_AUXSTAT);
+
+	if (value & DPAUX_DP_AUXSTAT_TIMEOUT_ERROR)
+		return -ETIMEDOUT;
+
+	if ((value & DPAUX_DP_AUXSTAT_RX_ERROR) ||
+	    (value & DPAUX_DP_AUXSTAT_SINKSTAT_ERROR) ||
+	    (value & DPAUX_DP_AUXSTAT_NO_STOP_ERROR))
+		return -EIO;
+
+	switch ((value & DPAUX_DP_AUXSTAT_REPLY_TYPE_MASK) >> 16) {
+	case 0x00:
+		msg->reply = DP_AUX_NATIVE_REPLY_ACK;
+		break;
+
+	case 0x01:
+		msg->reply = DP_AUX_NATIVE_REPLY_NACK;
+		break;
+
+	case 0x02:
+		msg->reply = DP_AUX_NATIVE_REPLY_DEFER;
+		break;
+
+	case 0x04:
+		msg->reply = DP_AUX_I2C_REPLY_NACK;
+		break;
+
+	case 0x08:
+		msg->reply = DP_AUX_I2C_REPLY_DEFER;
+		break;
+	}
+
+	if ((msg->size > 0) && (msg->reply == DP_AUX_NATIVE_REPLY_ACK)) {
+		if (msg->request & DP_AUX_I2C_READ) {
+			size_t count = value & DPAUX_DP_AUXSTAT_REPLY_MASK;
+
+			if (WARN_ON(count != msg->size))
+				count = min_t(size_t, count, msg->size);
+
+			tegra_dpaux_read_fifo(dpaux, msg->buffer, count);
+			ret = count;
+		}
+	}
+
+	return ret;
+}
+
+static void tegra_dpaux_hotplug(struct work_struct *work)
+{
+	struct tegra_dpaux *dpaux = work_to_dpaux(work);
+
+	if (dpaux->output)
+		drm_helper_hpd_irq_event(dpaux->output->connector.dev);
+}
+
+static irqreturn_t tegra_dpaux_irq(int irq, void *data)
+{
+	struct tegra_dpaux *dpaux = data;
+	irqreturn_t ret = IRQ_HANDLED;
+	unsigned long value;
+
+	/* clear interrupts */
+	value = tegra_dpaux_readl(dpaux, DPAUX_INTR_AUX);
+	tegra_dpaux_writel(dpaux, value, DPAUX_INTR_AUX);
+
+	if (value & (DPAUX_INTR_PLUG_EVENT | DPAUX_INTR_UNPLUG_EVENT))
+		schedule_work(&dpaux->work);
+
+	if (value & DPAUX_INTR_IRQ_EVENT) {
+		/* TODO: handle this */
+	}
+
+	if (value & DPAUX_INTR_AUX_DONE)
+		complete(&dpaux->complete);
+
+	return ret;
+}
+
+static int tegra_dpaux_probe(struct platform_device *pdev)
+{
+	struct tegra_dpaux *dpaux;
+	struct resource *regs;
+	unsigned long value;
+	int err;
+
+	dpaux = devm_kzalloc(&pdev->dev, sizeof(*dpaux), GFP_KERNEL);
+	if (!dpaux)
+		return -ENOMEM;
+
+	INIT_WORK(&dpaux->work, tegra_dpaux_hotplug);
+	init_completion(&dpaux->complete);
+	INIT_LIST_HEAD(&dpaux->list);
+	dpaux->dev = &pdev->dev;
+
+	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dpaux->regs = devm_ioremap_resource(&pdev->dev, regs);
+	if (IS_ERR(dpaux->regs))
+		return PTR_ERR(dpaux->regs);
+
+	dpaux->irq = platform_get_irq(pdev, 0);
+	if (dpaux->irq < 0) {
+		dev_err(&pdev->dev, "failed to get IRQ\n");
+		return -ENXIO;
+	}
+
+	dpaux->rst = devm_reset_control_get(&pdev->dev, "dpaux");
+	if (IS_ERR(dpaux->rst))
+		return PTR_ERR(dpaux->rst);
+
+	dpaux->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(dpaux->clk))
+		return PTR_ERR(dpaux->clk);
+
+	err = clk_prepare_enable(dpaux->clk);
+	if (err < 0)
+		return err;
+
+	reset_control_deassert(dpaux->rst);
+
+	dpaux->clk_parent = devm_clk_get(&pdev->dev, "parent");
+	if (IS_ERR(dpaux->clk_parent))
+		return PTR_ERR(dpaux->clk_parent);
+
+	err = clk_prepare_enable(dpaux->clk_parent);
+	if (err < 0)
+		return err;
+
+	err = clk_set_rate(dpaux->clk_parent, 270000000);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to set clock to 270 MHz: %d\n",
+			err);
+		return err;
+	}
+
+	dpaux->vdd = devm_regulator_get(&pdev->dev, "vdd");
+	if (IS_ERR(dpaux->vdd))
+		return PTR_ERR(dpaux->vdd);
+
+	err = devm_request_irq(dpaux->dev, dpaux->irq, tegra_dpaux_irq, 0,
+			       dev_name(dpaux->dev), dpaux);
+	if (err < 0) {
+		dev_err(dpaux->dev, "failed to request IRQ#%u: %d\n",
+			dpaux->irq, err);
+		return err;
+	}
+
+	dpaux->aux.transfer = tegra_dpaux_transfer;
+	dpaux->aux.dev = &pdev->dev;
+
+	err = drm_dp_aux_register(&dpaux->aux);
+	if (err < 0)
+		return err;
+
+	/* enable and clear all interrupts */
+	value = DPAUX_INTR_AUX_DONE | DPAUX_INTR_IRQ_EVENT |
+		DPAUX_INTR_UNPLUG_EVENT | DPAUX_INTR_PLUG_EVENT;
+	tegra_dpaux_writel(dpaux, value, DPAUX_INTR_EN_AUX);
+	tegra_dpaux_writel(dpaux, value, DPAUX_INTR_AUX);
+
+	mutex_lock(&dpaux_lock);
+	list_add_tail(&dpaux->list, &dpaux_list);
+	mutex_unlock(&dpaux_lock);
+
+	platform_set_drvdata(pdev, dpaux);
+
+	return 0;
+}
+
+static int tegra_dpaux_remove(struct platform_device *pdev)
+{
+	struct tegra_dpaux *dpaux = platform_get_drvdata(pdev);
+
+	drm_dp_aux_unregister(&dpaux->aux);
+
+	mutex_lock(&dpaux_lock);
+	list_del(&dpaux->list);
+	mutex_unlock(&dpaux_lock);
+
+	cancel_work_sync(&dpaux->work);
+
+	clk_disable_unprepare(dpaux->clk_parent);
+	reset_control_assert(dpaux->rst);
+	clk_disable_unprepare(dpaux->clk);
+
+	return 0;
+}
+
+static const struct of_device_id tegra_dpaux_of_match[] = {
+	{ .compatible = "nvidia,tegra124-dpaux", },
+	{ },
+};
+
+struct platform_driver tegra_dpaux_driver = {
+	.driver = {
+		.name = "tegra-dpaux",
+		.of_match_table = tegra_dpaux_of_match,
+	},
+	.probe = tegra_dpaux_probe,
+	.remove = tegra_dpaux_remove,
+};
+
+struct tegra_dpaux *tegra_dpaux_find_by_of_node(struct device_node *np)
+{
+	struct tegra_dpaux *dpaux;
+
+	mutex_lock(&dpaux_lock);
+
+	list_for_each_entry(dpaux, &dpaux_list, list)
+		if (np == dpaux->dev->of_node) {
+			mutex_unlock(&dpaux_lock);
+			return dpaux;
+		}
+
+	mutex_unlock(&dpaux_lock);
+
+	return NULL;
+}
+
+int tegra_dpaux_attach(struct tegra_dpaux *dpaux, struct tegra_output *output)
+{
+	unsigned long timeout;
+	int err;
+
+	output->connector.polled = DRM_CONNECTOR_POLL_HPD;
+	dpaux->output = output;
+
+	err = regulator_enable(dpaux->vdd);
+	if (err < 0)
+		return err;
+
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	while (time_before(jiffies, timeout)) {
+		enum drm_connector_status status;
+
+		status = tegra_dpaux_detect(dpaux);
+		if (status == connector_status_connected)
+			return 0;
+
+		usleep_range(1000, 2000);
+	}
+
+	return -ETIMEDOUT;
+}
+
+int tegra_dpaux_detach(struct tegra_dpaux *dpaux)
+{
+	unsigned long timeout;
+	int err;
+
+	err = regulator_disable(dpaux->vdd);
+	if (err < 0)
+		return err;
+
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	while (time_before(jiffies, timeout)) {
+		enum drm_connector_status status;
+
+		status = tegra_dpaux_detect(dpaux);
+		if (status == connector_status_disconnected) {
+			dpaux->output = NULL;
+			return 0;
+		}
+
+		usleep_range(1000, 2000);
+	}
+
+	return -ETIMEDOUT;
+}
+
+enum drm_connector_status tegra_dpaux_detect(struct tegra_dpaux *dpaux)
+{
+	unsigned long value;
+
+	value = tegra_dpaux_readl(dpaux, DPAUX_DP_AUXSTAT);
+
+	if (value & DPAUX_DP_AUXSTAT_HPD_STATUS)
+		return connector_status_connected;
+
+	return connector_status_disconnected;
+}
+
+int tegra_dpaux_enable(struct tegra_dpaux *dpaux)
+{
+	unsigned long value;
+
+	value = DPAUX_HYBRID_PADCTL_AUX_CMH(2) |
+		DPAUX_HYBRID_PADCTL_AUX_DRVZ(4) |
+		DPAUX_HYBRID_PADCTL_AUX_DRVI(0x18) |
+		DPAUX_HYBRID_PADCTL_AUX_INPUT_RCV |
+		DPAUX_HYBRID_PADCTL_MODE_AUX;
+	tegra_dpaux_writel(dpaux, value, DPAUX_HYBRID_PADCTL);
+
+	value = tegra_dpaux_readl(dpaux, DPAUX_HYBRID_SPARE);
+	value &= ~DPAUX_HYBRID_SPARE_PAD_POWER_DOWN;
+	tegra_dpaux_writel(dpaux, value, DPAUX_HYBRID_SPARE);
+
+	return 0;
+}
+
+int tegra_dpaux_disable(struct tegra_dpaux *dpaux)
+{
+	unsigned long value;
+
+	value = tegra_dpaux_readl(dpaux, DPAUX_HYBRID_SPARE);
+	value |= DPAUX_HYBRID_SPARE_PAD_POWER_DOWN;
+	tegra_dpaux_writel(dpaux, value, DPAUX_HYBRID_SPARE);
+
+	return 0;
+}
+
+int tegra_dpaux_prepare(struct tegra_dpaux *dpaux, u8 encoding)
+{
+	int err;
+
+	err = drm_dp_dpcd_writeb(&dpaux->aux, DP_MAIN_LINK_CHANNEL_CODING_SET,
+				 encoding);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+
+int tegra_dpaux_train(struct tegra_dpaux *dpaux, struct drm_dp_link *link,
+		      u8 pattern)
+{
+	u8 tp = pattern & DP_TRAINING_PATTERN_MASK;
+	u8 status[DP_LINK_STATUS_SIZE], values[4];
+	unsigned int i;
+	int err;
+
+	err = drm_dp_dpcd_writeb(&dpaux->aux, DP_TRAINING_PATTERN_SET, pattern);
+	if (err < 0)
+		return err;
+
+	if (tp == DP_TRAINING_PATTERN_DISABLE)
+		return 0;
+
+	for (i = 0; i < link->num_lanes; i++)
+		values[i] = DP_TRAIN_MAX_PRE_EMPHASIS_REACHED |
+			    DP_TRAIN_PRE_EMPHASIS_0 |
+			    DP_TRAIN_MAX_SWING_REACHED |
+			    DP_TRAIN_VOLTAGE_SWING_400;
+
+	err = drm_dp_dpcd_write(&dpaux->aux, DP_TRAINING_LANE0_SET, values,
+				link->num_lanes);
+	if (err < 0)
+		return err;
+
+	usleep_range(500, 1000);
+
+	err = drm_dp_dpcd_read_link_status(&dpaux->aux, status);
+	if (err < 0)
+		return err;
+
+	switch (tp) {
+	case DP_TRAINING_PATTERN_1:
+		if (!drm_dp_clock_recovery_ok(status, link->num_lanes))
+			return -EAGAIN;
+
+		break;
+
+	case DP_TRAINING_PATTERN_2:
+		if (!drm_dp_channel_eq_ok(status, link->num_lanes))
+			return -EAGAIN;
+
+		break;
+
+	default:
+		dev_err(dpaux->dev, "unsupported training pattern %u\n", tp);
+		return -EINVAL;
+	}
+
+	err = drm_dp_dpcd_writeb(&dpaux->aux, DP_EDP_CONFIGURATION_SET, 0);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/tegra/dpaux.h b/drivers/gpu/drm/tegra/dpaux.h
new file mode 100644
index 00000000000..806e245ca78
--- /dev/null
+++ b/drivers/gpu/drm/tegra/dpaux.h
@@ -0,0 +1,74 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef DRM_TEGRA_DPAUX_H
+#define DRM_TEGRA_DPAUX_H
+
+#define DPAUX_CTXSW 0x00
+
+#define DPAUX_INTR_EN_AUX 0x01
+#define DPAUX_INTR_AUX 0x05
+#define DPAUX_INTR_AUX_DONE (1 << 3)
+#define DPAUX_INTR_IRQ_EVENT (1 << 2)
+#define DPAUX_INTR_UNPLUG_EVENT (1 << 1)
+#define DPAUX_INTR_PLUG_EVENT (1 << 0)
+
+#define DPAUX_DP_AUXDATA_WRITE(x) (0x09 + ((x) << 2))
+#define DPAUX_DP_AUXDATA_READ(x) (0x19 + ((x) << 2))
+#define DPAUX_DP_AUXADDR 0x29
+
+#define DPAUX_DP_AUXCTL 0x2d
+#define DPAUX_DP_AUXCTL_TRANSACTREQ (1 << 16)
+#define DPAUX_DP_AUXCTL_CMD_AUX_RD (9 << 12)
+#define DPAUX_DP_AUXCTL_CMD_AUX_WR (8 << 12)
+#define DPAUX_DP_AUXCTL_CMD_MOT_RQ (6 << 12)
+#define DPAUX_DP_AUXCTL_CMD_MOT_RD (5 << 12)
+#define DPAUX_DP_AUXCTL_CMD_MOT_WR (4 << 12)
+#define DPAUX_DP_AUXCTL_CMD_I2C_RQ (2 << 12)
+#define DPAUX_DP_AUXCTL_CMD_I2C_RD (1 << 12)
+#define DPAUX_DP_AUXCTL_CMD_I2C_WR (0 << 12)
+#define DPAUX_DP_AUXCTL_CMD_ADDRESS_ONLY (1 << 8)
+#define DPAUX_DP_AUXCTL_CMDLEN(x) ((x) & 0xff)
+
+#define DPAUX_DP_AUXSTAT 0x31
+#define DPAUX_DP_AUXSTAT_HPD_STATUS (1 << 28)
+#define DPAUX_DP_AUXSTAT_REPLY_TYPE_MASK (0xf0000)
+#define DPAUX_DP_AUXSTAT_NO_STOP_ERROR (1 << 11)
+#define DPAUX_DP_AUXSTAT_SINKSTAT_ERROR (1 << 10)
+#define DPAUX_DP_AUXSTAT_RX_ERROR (1 << 9)
+#define DPAUX_DP_AUXSTAT_TIMEOUT_ERROR (1 << 8)
+#define DPAUX_DP_AUXSTAT_REPLY_MASK (0xff)
+
+#define DPAUX_DP_AUX_SINKSTAT_LO 0x35
+#define DPAUX_DP_AUX_SINKSTAT_HI 0x39
+
+#define DPAUX_HPD_CONFIG 0x3d
+#define DPAUX_HPD_CONFIG_UNPLUG_MIN_TIME(x) (((x) & 0xffff) << 16)
+#define DPAUX_HPD_CONFIG_PLUG_MIN_TIME(x) ((x) & 0xffff)
+
+#define DPAUX_HPD_IRQ_CONFIG 0x41
+#define DPAUX_HPD_IRQ_CONFIG_MIN_LOW_TIME(x) ((x) & 0xffff)
+
+#define DPAUX_DP_AUX_CONFIG 0x45
+
+#define DPAUX_HYBRID_PADCTL 0x49
+#define DPAUX_HYBRID_PADCTL_AUX_CMH(x) (((x) & 0x3) << 12)
+#define DPAUX_HYBRID_PADCTL_AUX_DRVZ(x) (((x) & 0x7) << 8)
+#define DPAUX_HYBRID_PADCTL_AUX_DRVI(x) (((x) & 0x3f) << 2)
+#define DPAUX_HYBRID_PADCTL_AUX_INPUT_RCV (1 << 1)
+#define DPAUX_HYBRID_PADCTL_MODE_I2C (1 << 0)
+#define DPAUX_HYBRID_PADCTL_MODE_AUX (0 << 0)
+
+#define DPAUX_HYBRID_SPARE 0x4d
+#define DPAUX_HYBRID_SPARE_PAD_POWER_DOWN (1 << 0)
+
+#define DPAUX_SCRATCH_REG0 0x51
+#define DPAUX_SCRATCH_REG1 0x55
+#define DPAUX_SCRATCH_REG2 0x59
+
+#endif
diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 07eba596d45..3396f9f6a9f 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -33,7 +33,6 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 	if (!tegra)
 		return -ENOMEM;
 
-	dev_set_drvdata(drm->dev, tegra);
 	mutex_init(&tegra->clients_lock);
 	INIT_LIST_HEAD(&tegra->clients);
 	drm->dev_private = tegra;
@@ -104,9 +103,11 @@ static void tegra_drm_context_free(struct tegra_drm_context *context)
 
 static void tegra_drm_lastclose(struct drm_device *drm)
 {
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 	struct tegra_drm *tegra = drm->dev_private;
 
 	tegra_fbdev_restore_mode(tegra->fbdev);
+#endif
 }
 
 static struct host1x_bo *
@@ -578,7 +579,7 @@ static void tegra_debugfs_cleanup(struct drm_minor *minor)
 #endif
 
 static struct drm_driver tegra_drm_driver = {
-	.driver_features = DRIVER_MODESET | DRIVER_GEM,
+	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME,
 	.load = tegra_drm_load,
 	.unload = tegra_drm_unload,
 	.open = tegra_drm_open,
@@ -596,6 +597,12 @@ static struct drm_driver tegra_drm_driver = {
 
 	.gem_free_object = tegra_bo_free_object,
 	.gem_vm_ops = &tegra_bo_vm_ops,
+
+	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
+	.gem_prime_export = tegra_gem_prime_export,
+	.gem_prime_import = tegra_gem_prime_import,
+
 	.dumb_create = tegra_bo_dumb_create,
 	.dumb_map_offset = tegra_bo_dumb_map_offset,
 	.dumb_destroy = drm_gem_dumb_destroy,
@@ -632,14 +639,40 @@ int tegra_drm_unregister_client(struct tegra_drm *tegra,
 	return 0;
 }
 
-static int host1x_drm_probe(struct host1x_device *device)
+static int host1x_drm_probe(struct host1x_device *dev)
 {
-	return drm_host1x_init(&tegra_drm_driver, device);
+	struct drm_driver *driver = &tegra_drm_driver;
+	struct drm_device *drm;
+	int err;
+
+	drm = drm_dev_alloc(driver, &dev->dev);
+	if (!drm)
+		return -ENOMEM;
+
+	drm_dev_set_unique(drm, dev_name(&dev->dev));
+	dev_set_drvdata(&dev->dev, drm);
+
+	err = drm_dev_register(drm, 0);
+	if (err < 0)
+		goto unref;
+
+	DRM_INFO("Initialized %s %d.%d.%d %s on minor %d\n", driver->name,
+		 driver->major, driver->minor, driver->patchlevel,
+		 driver->date, drm->primary->index);
+
+	return 0;
+
+unref:
+	drm_dev_unref(drm);
+	return err;
 }
 
-static int host1x_drm_remove(struct host1x_device *device)
+static int host1x_drm_remove(struct host1x_device *dev)
 {
-	drm_host1x_exit(&tegra_drm_driver, device);
+	struct drm_device *drm = dev_get_drvdata(&dev->dev);
+
+	drm_dev_unregister(drm);
+	drm_dev_unref(drm);
 
 	return 0;
 }
@@ -653,8 +686,12 @@ static const struct of_device_id host1x_drm_subdevs[] = {
 	{ .compatible = "nvidia,tegra30-hdmi", },
 	{ .compatible = "nvidia,tegra30-gr2d", },
 	{ .compatible = "nvidia,tegra30-gr3d", },
+	{ .compatible = "nvidia,tegra114-dsi", },
 	{ .compatible = "nvidia,tegra114-hdmi", },
 	{ .compatible = "nvidia,tegra114-gr3d", },
+	{ .compatible = "nvidia,tegra124-dc", },
+	{ .compatible = "nvidia,tegra124-sor", },
+	{ .compatible = "nvidia,tegra124-hdmi", },
 	{ /* sentinel */ }
 };
 
@@ -677,14 +714,26 @@ static int __init host1x_drm_init(void)
 	if (err < 0)
 		goto unregister_host1x;
 
-	err = platform_driver_register(&tegra_hdmi_driver);
+	err = platform_driver_register(&tegra_dsi_driver);
 	if (err < 0)
 		goto unregister_dc;
 
-	err = platform_driver_register(&tegra_gr2d_driver);
+	err = platform_driver_register(&tegra_sor_driver);
+	if (err < 0)
+		goto unregister_dsi;
+
+	err = platform_driver_register(&tegra_hdmi_driver);
+	if (err < 0)
+		goto unregister_sor;
+
+	err = platform_driver_register(&tegra_dpaux_driver);
 	if (err < 0)
 		goto unregister_hdmi;
 
+	err = platform_driver_register(&tegra_gr2d_driver);
+	if (err < 0)
+		goto unregister_dpaux;
+
 	err = platform_driver_register(&tegra_gr3d_driver);
 	if (err < 0)
 		goto unregister_gr2d;
@@ -693,8 +742,14 @@ static int __init host1x_drm_init(void)
 
 unregister_gr2d:
 	platform_driver_unregister(&tegra_gr2d_driver);
+unregister_dpaux:
+	platform_driver_unregister(&tegra_dpaux_driver);
 unregister_hdmi:
 	platform_driver_unregister(&tegra_hdmi_driver);
+unregister_sor:
+	platform_driver_unregister(&tegra_sor_driver);
+unregister_dsi:
+	platform_driver_unregister(&tegra_dsi_driver);
 unregister_dc:
 	platform_driver_unregister(&tegra_dc_driver);
 unregister_host1x:
@@ -707,7 +762,10 @@ static void __exit host1x_drm_exit(void)
 {
 	platform_driver_unregister(&tegra_gr3d_driver);
 	platform_driver_unregister(&tegra_gr2d_driver);
+	platform_driver_unregister(&tegra_dpaux_driver);
 	platform_driver_unregister(&tegra_hdmi_driver);
+	platform_driver_unregister(&tegra_sor_driver);
+	platform_driver_unregister(&tegra_dsi_driver);
 	platform_driver_unregister(&tegra_dc_driver);
 	host1x_driver_unregister(&host1x_drm_driver);
 }
diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index 7da0b923131..6b8fe9d86ed 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -19,16 +19,20 @@
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_fixed.h>
 
+struct reset_control;
+
 struct tegra_fb {
 	struct drm_framebuffer base;
 	struct tegra_bo **planes;
 	unsigned int num_planes;
 };
 
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 struct tegra_fbdev {
 	struct drm_fb_helper base;
 	struct tegra_fb *fb;
 };
+#endif
 
 struct tegra_drm {
 	struct drm_device *drm;
@@ -36,7 +40,9 @@ struct tegra_drm {
 	struct mutex clients_lock;
 	struct list_head clients;
 
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 	struct tegra_fbdev *fbdev;
+#endif
 };
 
 struct tegra_drm_client;
@@ -74,14 +80,15 @@ host1x_to_drm_client(struct host1x_client *client)
 	return container_of(client, struct tegra_drm_client, base);
 }
 
-extern int tegra_drm_register_client(struct tegra_drm *tegra,
-				     struct tegra_drm_client *client);
-extern int tegra_drm_unregister_client(struct tegra_drm *tegra,
-				       struct tegra_drm_client *client);
+int tegra_drm_register_client(struct tegra_drm *tegra,
+			      struct tegra_drm_client *client);
+int tegra_drm_unregister_client(struct tegra_drm *tegra,
+				struct tegra_drm_client *client);
 
-extern int tegra_drm_init(struct tegra_drm *tegra, struct drm_device *drm);
-extern int tegra_drm_exit(struct tegra_drm *tegra);
+int tegra_drm_init(struct tegra_drm *tegra, struct drm_device *drm);
+int tegra_drm_exit(struct tegra_drm *tegra);
 
+struct tegra_dc_soc_info;
 struct tegra_output;
 
 struct tegra_dc {
@@ -93,6 +100,7 @@ struct tegra_dc {
 	int pipe;
 
 	struct clk *clk;
+	struct reset_control *rst;
 	void __iomem *regs;
 	int irq;
 
@@ -106,6 +114,8 @@ struct tegra_dc {
 
 	/* page-flip handling */
 	struct drm_pending_vblank_event *event;
+
+	const struct tegra_dc_soc_info *soc;
 };
 
 static inline struct tegra_dc *
@@ -146,6 +156,7 @@ struct tegra_dc_window {
 	} dst;
 	unsigned int bits_per_pixel;
 	unsigned int format;
+	unsigned int swap;
 	unsigned int stride[2];
 	unsigned long base[3];
 	bool bottom_up;
@@ -153,27 +164,26 @@ struct tegra_dc_window {
 };
 
 /* from dc.c */
-extern unsigned int tegra_dc_format(uint32_t format);
-extern int tegra_dc_setup_window(struct tegra_dc *dc, unsigned int index,
-				 const struct tegra_dc_window *window);
-extern void tegra_dc_enable_vblank(struct tegra_dc *dc);
-extern void tegra_dc_disable_vblank(struct tegra_dc *dc);
-extern void tegra_dc_cancel_page_flip(struct drm_crtc *crtc,
-				      struct drm_file *file);
+void tegra_dc_enable_vblank(struct tegra_dc *dc);
+void tegra_dc_disable_vblank(struct tegra_dc *dc);
+void tegra_dc_cancel_page_flip(struct drm_crtc *crtc, struct drm_file *file);
 
 struct tegra_output_ops {
 	int (*enable)(struct tegra_output *output);
 	int (*disable)(struct tegra_output *output);
 	int (*setup_clock)(struct tegra_output *output, struct clk *clk,
-			   unsigned long pclk);
+			   unsigned long pclk, unsigned int *div);
 	int (*check_mode)(struct tegra_output *output,
 			  struct drm_display_mode *mode,
 			  enum drm_mode_status *status);
+	enum drm_connector_status (*detect)(struct tegra_output *output);
 };
 
 enum tegra_output_type {
 	TEGRA_OUTPUT_RGB,
 	TEGRA_OUTPUT_HDMI,
+	TEGRA_OUTPUT_DSI,
+	TEGRA_OUTPUT_EDP,
 };
 
 struct tegra_output {
@@ -183,6 +193,7 @@ struct tegra_output {
 	const struct tegra_output_ops *ops;
 	enum tegra_output_type type;
 
+	struct drm_panel *panel;
 	struct i2c_adapter *ddc;
 	const struct edid *edid;
 	unsigned int hpd_irq;
@@ -219,10 +230,11 @@ static inline int tegra_output_disable(struct tegra_output *output)
 }
 
 static inline int tegra_output_setup_clock(struct tegra_output *output,
-					   struct clk *clk, unsigned long pclk)
+					   struct clk *clk, unsigned long pclk,
+					   unsigned int *div)
 {
 	if (output && output->ops && output->ops->setup_clock)
-		return output->ops->setup_clock(output, clk, pclk);
+		return output->ops->setup_clock(output, clk, pclk, div);
 
 	return output ? -ENOSYS : -EINVAL;
 }
@@ -237,33 +249,48 @@ static inline int tegra_output_check_mode(struct tegra_output *output,
 	return output ? -ENOSYS : -EINVAL;
 }
 
-/* from bus.c */
-int drm_host1x_init(struct drm_driver *driver, struct host1x_device *device);
-void drm_host1x_exit(struct drm_driver *driver, struct host1x_device *device);
-
 /* from rgb.c */
-extern int tegra_dc_rgb_probe(struct tegra_dc *dc);
-extern int tegra_dc_rgb_remove(struct tegra_dc *dc);
-extern int tegra_dc_rgb_init(struct drm_device *drm, struct tegra_dc *dc);
-extern int tegra_dc_rgb_exit(struct tegra_dc *dc);
+int tegra_dc_rgb_probe(struct tegra_dc *dc);
+int tegra_dc_rgb_remove(struct tegra_dc *dc);
+int tegra_dc_rgb_init(struct drm_device *drm, struct tegra_dc *dc);
+int tegra_dc_rgb_exit(struct tegra_dc *dc);
 
 /* from output.c */
-extern int tegra_output_probe(struct tegra_output *output);
-extern int tegra_output_remove(struct tegra_output *output);
-extern int tegra_output_init(struct drm_device *drm, struct tegra_output *output);
-extern int tegra_output_exit(struct tegra_output *output);
+int tegra_output_probe(struct tegra_output *output);
+int tegra_output_remove(struct tegra_output *output);
+int tegra_output_init(struct drm_device *drm, struct tegra_output *output);
+int tegra_output_exit(struct tegra_output *output);
+
+/* from dpaux.c */
+struct tegra_dpaux;
+struct drm_dp_link;
+
+struct tegra_dpaux *tegra_dpaux_find_by_of_node(struct device_node *np);
+enum drm_connector_status tegra_dpaux_detect(struct tegra_dpaux *dpaux);
+int tegra_dpaux_attach(struct tegra_dpaux *dpaux, struct tegra_output *output);
+int tegra_dpaux_detach(struct tegra_dpaux *dpaux);
+int tegra_dpaux_enable(struct tegra_dpaux *dpaux);
+int tegra_dpaux_disable(struct tegra_dpaux *dpaux);
+int tegra_dpaux_prepare(struct tegra_dpaux *dpaux, u8 encoding);
+int tegra_dpaux_train(struct tegra_dpaux *dpaux, struct drm_dp_link *link,
+		      u8 pattern);
 
 /* from fb.c */
 struct tegra_bo *tegra_fb_get_plane(struct drm_framebuffer *framebuffer,
 				    unsigned int index);
 bool tegra_fb_is_bottom_up(struct drm_framebuffer *framebuffer);
 bool tegra_fb_is_tiled(struct drm_framebuffer *framebuffer);
-extern int tegra_drm_fb_init(struct drm_device *drm);
-extern void tegra_drm_fb_exit(struct drm_device *drm);
-extern void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev);
+int tegra_drm_fb_init(struct drm_device *drm);
+void tegra_drm_fb_exit(struct drm_device *drm);
+#ifdef CONFIG_DRM_TEGRA_FBDEV
+void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev);
+#endif
 
 extern struct platform_driver tegra_dc_driver;
+extern struct platform_driver tegra_dsi_driver;
+extern struct platform_driver tegra_sor_driver;
 extern struct platform_driver tegra_hdmi_driver;
+extern struct platform_driver tegra_dpaux_driver;
 extern struct platform_driver tegra_gr2d_driver;
 extern struct platform_driver tegra_gr3d_driver;
 
diff --git a/drivers/gpu/drm/tegra/dsi.c b/drivers/gpu/drm/tegra/dsi.c
new file mode 100644
index 00000000000..bd56f2affa7
--- /dev/null
+++ b/drivers/gpu/drm/tegra/dsi.c
@@ -0,0 +1,993 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/debugfs.h>
+#include <linux/host1x.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/reset.h>
+
+#include <linux/regulator/consumer.h>
+
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_panel.h>
+
+#include <video/mipi_display.h>
+
+#include "dc.h"
+#include "drm.h"
+#include "dsi.h"
+#include "mipi-phy.h"
+
+#define DSI_VIDEO_FIFO_DEPTH (1920 / 4)
+#define DSI_HOST_FIFO_DEPTH 64
+
+struct tegra_dsi {
+	struct host1x_client client;
+	struct tegra_output output;
+	struct device *dev;
+
+	void __iomem *regs;
+
+	struct reset_control *rst;
+	struct clk *clk_parent;
+	struct clk *clk_lp;
+	struct clk *clk;
+
+	struct drm_info_list *debugfs_files;
+	struct drm_minor *minor;
+	struct dentry *debugfs;
+
+	unsigned long flags;
+	enum mipi_dsi_pixel_format format;
+	unsigned int lanes;
+
+	struct tegra_mipi_device *mipi;
+	struct mipi_dsi_host host;
+
+	struct regulator *vdd;
+	bool enabled;
+};
+
+static inline struct tegra_dsi *
+host1x_client_to_dsi(struct host1x_client *client)
+{
+	return container_of(client, struct tegra_dsi, client);
+}
+
+static inline struct tegra_dsi *host_to_tegra(struct mipi_dsi_host *host)
+{
+	return container_of(host, struct tegra_dsi, host);
+}
+
+static inline struct tegra_dsi *to_dsi(struct tegra_output *output)
+{
+	return container_of(output, struct tegra_dsi, output);
+}
+
+static inline unsigned long tegra_dsi_readl(struct tegra_dsi *dsi,
+					    unsigned long reg)
+{
+	return readl(dsi->regs + (reg << 2));
+}
+
+static inline void tegra_dsi_writel(struct tegra_dsi *dsi, unsigned long value,
+				    unsigned long reg)
+{
+	writel(value, dsi->regs + (reg << 2));
+}
+
+static int tegra_dsi_show_regs(struct seq_file *s, void *data)
+{
+	struct drm_info_node *node = s->private;
+	struct tegra_dsi *dsi = node->info_ent->data;
+
+#define DUMP_REG(name)						\
+	seq_printf(s, "%-32s %#05x %08lx\n", #name, name,	\
+		   tegra_dsi_readl(dsi, name))
+
+	DUMP_REG(DSI_INCR_SYNCPT);
+	DUMP_REG(DSI_INCR_SYNCPT_CONTROL);
+	DUMP_REG(DSI_INCR_SYNCPT_ERROR);
+	DUMP_REG(DSI_CTXSW);
+	DUMP_REG(DSI_RD_DATA);
+	DUMP_REG(DSI_WR_DATA);
+	DUMP_REG(DSI_POWER_CONTROL);
+	DUMP_REG(DSI_INT_ENABLE);
+	DUMP_REG(DSI_INT_STATUS);
+	DUMP_REG(DSI_INT_MASK);
+	DUMP_REG(DSI_HOST_CONTROL);
+	DUMP_REG(DSI_CONTROL);
+	DUMP_REG(DSI_SOL_DELAY);
+	DUMP_REG(DSI_MAX_THRESHOLD);
+	DUMP_REG(DSI_TRIGGER);
+	DUMP_REG(DSI_TX_CRC);
+	DUMP_REG(DSI_STATUS);
+
+	DUMP_REG(DSI_INIT_SEQ_CONTROL);
+	DUMP_REG(DSI_INIT_SEQ_DATA_0);
+	DUMP_REG(DSI_INIT_SEQ_DATA_1);
+	DUMP_REG(DSI_INIT_SEQ_DATA_2);
+	DUMP_REG(DSI_INIT_SEQ_DATA_3);
+	DUMP_REG(DSI_INIT_SEQ_DATA_4);
+	DUMP_REG(DSI_INIT_SEQ_DATA_5);
+	DUMP_REG(DSI_INIT_SEQ_DATA_6);
+	DUMP_REG(DSI_INIT_SEQ_DATA_7);
+
+	DUMP_REG(DSI_PKT_SEQ_0_LO);
+	DUMP_REG(DSI_PKT_SEQ_0_HI);
+	DUMP_REG(DSI_PKT_SEQ_1_LO);
+	DUMP_REG(DSI_PKT_SEQ_1_HI);
+	DUMP_REG(DSI_PKT_SEQ_2_LO);
+	DUMP_REG(DSI_PKT_SEQ_2_HI);
+	DUMP_REG(DSI_PKT_SEQ_3_LO);
+	DUMP_REG(DSI_PKT_SEQ_3_HI);
+	DUMP_REG(DSI_PKT_SEQ_4_LO);
+	DUMP_REG(DSI_PKT_SEQ_4_HI);
+	DUMP_REG(DSI_PKT_SEQ_5_LO);
+	DUMP_REG(DSI_PKT_SEQ_5_HI);
+
+	DUMP_REG(DSI_DCS_CMDS);
+
+	DUMP_REG(DSI_PKT_LEN_0_1);
+	DUMP_REG(DSI_PKT_LEN_2_3);
+	DUMP_REG(DSI_PKT_LEN_4_5);
+	DUMP_REG(DSI_PKT_LEN_6_7);
+
+	DUMP_REG(DSI_PHY_TIMING_0);
+	DUMP_REG(DSI_PHY_TIMING_1);
+	DUMP_REG(DSI_PHY_TIMING_2);
+	DUMP_REG(DSI_BTA_TIMING);
+
+	DUMP_REG(DSI_TIMEOUT_0);
+	DUMP_REG(DSI_TIMEOUT_1);
+	DUMP_REG(DSI_TO_TALLY);
+
+	DUMP_REG(DSI_PAD_CONTROL_0);
+	DUMP_REG(DSI_PAD_CONTROL_CD);
+	DUMP_REG(DSI_PAD_CD_STATUS);
+	DUMP_REG(DSI_VIDEO_MODE_CONTROL);
+	DUMP_REG(DSI_PAD_CONTROL_1);
+	DUMP_REG(DSI_PAD_CONTROL_2);
+	DUMP_REG(DSI_PAD_CONTROL_3);
+	DUMP_REG(DSI_PAD_CONTROL_4);
+
+	DUMP_REG(DSI_GANGED_MODE_CONTROL);
+	DUMP_REG(DSI_GANGED_MODE_START);
+	DUMP_REG(DSI_GANGED_MODE_SIZE);
+
+	DUMP_REG(DSI_RAW_DATA_BYTE_COUNT);
+	DUMP_REG(DSI_ULTRA_LOW_POWER_CONTROL);
+
+	DUMP_REG(DSI_INIT_SEQ_DATA_8);
+	DUMP_REG(DSI_INIT_SEQ_DATA_9);
+	DUMP_REG(DSI_INIT_SEQ_DATA_10);
+	DUMP_REG(DSI_INIT_SEQ_DATA_11);
+	DUMP_REG(DSI_INIT_SEQ_DATA_12);
+	DUMP_REG(DSI_INIT_SEQ_DATA_13);
+	DUMP_REG(DSI_INIT_SEQ_DATA_14);
+	DUMP_REG(DSI_INIT_SEQ_DATA_15);
+
+#undef DUMP_REG
+
+	return 0;
+}
+
+static struct drm_info_list debugfs_files[] = {
+	{ "regs", tegra_dsi_show_regs, 0, NULL },
+};
+
+static int tegra_dsi_debugfs_init(struct tegra_dsi *dsi,
+				  struct drm_minor *minor)
+{
+	const char *name = dev_name(dsi->dev);
+	unsigned int i;
+	int err;
+
+	dsi->debugfs = debugfs_create_dir(name, minor->debugfs_root);
+	if (!dsi->debugfs)
+		return -ENOMEM;
+
+	dsi->debugfs_files = kmemdup(debugfs_files, sizeof(debugfs_files),
+				     GFP_KERNEL);
+	if (!dsi->debugfs_files) {
+		err = -ENOMEM;
+		goto remove;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(debugfs_files); i++)
+		dsi->debugfs_files[i].data = dsi;
+
+	err = drm_debugfs_create_files(dsi->debugfs_files,
+				       ARRAY_SIZE(debugfs_files),
+				       dsi->debugfs, minor);
+	if (err < 0)
+		goto free;
+
+	dsi->minor = minor;
+
+	return 0;
+
+free:
+	kfree(dsi->debugfs_files);
+	dsi->debugfs_files = NULL;
+remove:
+	debugfs_remove(dsi->debugfs);
+	dsi->debugfs = NULL;
+
+	return err;
+}
+
+static int tegra_dsi_debugfs_exit(struct tegra_dsi *dsi)
+{
+	drm_debugfs_remove_files(dsi->debugfs_files, ARRAY_SIZE(debugfs_files),
+				 dsi->minor);
+	dsi->minor = NULL;
+
+	kfree(dsi->debugfs_files);
+	dsi->debugfs_files = NULL;
+
+	debugfs_remove(dsi->debugfs);
+	dsi->debugfs = NULL;
+
+	return 0;
+}
+
+#define PKT_ID0(id)	((((id) & 0x3f) <<  3) | (1 <<  9))
+#define PKT_LEN0(len)	(((len) & 0x07) <<  0)
+#define PKT_ID1(id)	((((id) & 0x3f) << 13) | (1 << 19))
+#define PKT_LEN1(len)	(((len) & 0x07) << 10)
+#define PKT_ID2(id)	((((id) & 0x3f) << 23) | (1 << 29))
+#define PKT_LEN2(len)	(((len) & 0x07) << 20)
+
+#define PKT_LP		(1 << 30)
+#define NUM_PKT_SEQ	12
+
+/*
+ * non-burst mode with sync pulses
+ */
+static const u32 pkt_seq_video_non_burst_sync_pulses[NUM_PKT_SEQ] = {
+	[ 0] = PKT_ID0(MIPI_DSI_V_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0) |
+	       PKT_LP,
+	[ 1] = 0,
+	[ 2] = PKT_ID0(MIPI_DSI_V_SYNC_END) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0) |
+	       PKT_LP,
+	[ 3] = 0,
+	[ 4] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0) |
+	       PKT_LP,
+	[ 5] = 0,
+	[ 6] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0),
+	[ 7] = PKT_ID0(MIPI_DSI_BLANKING_PACKET) | PKT_LEN0(2) |
+	       PKT_ID1(MIPI_DSI_PACKED_PIXEL_STREAM_24) | PKT_LEN1(3) |
+	       PKT_ID2(MIPI_DSI_BLANKING_PACKET) | PKT_LEN2(4),
+	[ 8] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0) |
+	       PKT_LP,
+	[ 9] = 0,
+	[10] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0),
+	[11] = PKT_ID0(MIPI_DSI_BLANKING_PACKET) | PKT_LEN0(2) |
+	       PKT_ID1(MIPI_DSI_PACKED_PIXEL_STREAM_24) | PKT_LEN1(3) |
+	       PKT_ID2(MIPI_DSI_BLANKING_PACKET) | PKT_LEN2(4),
+};
+
+/*
+ * non-burst mode with sync events
+ */
+static const u32 pkt_seq_video_non_burst_sync_events[NUM_PKT_SEQ] = {
+	[ 0] = PKT_ID0(MIPI_DSI_V_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_END_OF_TRANSMISSION) | PKT_LEN1(7) |
+	       PKT_LP,
+	[ 1] = 0,
+	[ 2] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_END_OF_TRANSMISSION) | PKT_LEN1(7) |
+	       PKT_LP,
+	[ 3] = 0,
+	[ 4] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_END_OF_TRANSMISSION) | PKT_LEN1(7) |
+	       PKT_LP,
+	[ 5] = 0,
+	[ 6] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(2) |
+	       PKT_ID2(MIPI_DSI_PACKED_PIXEL_STREAM_24) | PKT_LEN2(3),
+	[ 7] = PKT_ID0(MIPI_DSI_BLANKING_PACKET) | PKT_LEN0(4),
+	[ 8] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_END_OF_TRANSMISSION) | PKT_LEN1(7) |
+	       PKT_LP,
+	[ 9] = 0,
+	[10] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(2) |
+	       PKT_ID2(MIPI_DSI_PACKED_PIXEL_STREAM_24) | PKT_LEN2(3),
+	[11] = PKT_ID0(MIPI_DSI_BLANKING_PACKET) | PKT_LEN0(4),
+};
+
+static int tegra_dsi_set_phy_timing(struct tegra_dsi *dsi)
+{
+	struct mipi_dphy_timing timing;
+	unsigned long value, period;
+	long rate;
+	int err;
+
+	rate = clk_get_rate(dsi->clk);
+	if (rate < 0)
+		return rate;
+
+	period = DIV_ROUND_CLOSEST(1000000000UL, rate * 2);
+
+	err = mipi_dphy_timing_get_default(&timing, period);
+	if (err < 0)
+		return err;
+
+	err = mipi_dphy_timing_validate(&timing, period);
+	if (err < 0) {
+		dev_err(dsi->dev, "failed to validate D-PHY timing: %d\n", err);
+		return err;
+	}
+
+	/*
+	 * The D-PHY timing fields below are expressed in byte-clock cycles,
+	 * so multiply the period by 8.
+	 */
+	period *= 8;
+
+	value = DSI_TIMING_FIELD(timing.hsexit, period, 1) << 24 |
+		DSI_TIMING_FIELD(timing.hstrail, period, 0) << 16 |
+		DSI_TIMING_FIELD(timing.hszero, period, 3) << 8 |
+		DSI_TIMING_FIELD(timing.hsprepare, period, 1);
+	tegra_dsi_writel(dsi, value, DSI_PHY_TIMING_0);
+
+	value = DSI_TIMING_FIELD(timing.clktrail, period, 1) << 24 |
+		DSI_TIMING_FIELD(timing.clkpost, period, 1) << 16 |
+		DSI_TIMING_FIELD(timing.clkzero, period, 1) << 8 |
+		DSI_TIMING_FIELD(timing.lpx, period, 1);
+	tegra_dsi_writel(dsi, value, DSI_PHY_TIMING_1);
+
+	value = DSI_TIMING_FIELD(timing.clkprepare, period, 1) << 16 |
+		DSI_TIMING_FIELD(timing.clkpre, period, 1) << 8 |
+		DSI_TIMING_FIELD(0xff * period, period, 0) << 0;
+	tegra_dsi_writel(dsi, value, DSI_PHY_TIMING_2);
+
+	value = DSI_TIMING_FIELD(timing.taget, period, 1) << 16 |
+		DSI_TIMING_FIELD(timing.tasure, period, 1) << 8 |
+		DSI_TIMING_FIELD(timing.tago, period, 1);
+	tegra_dsi_writel(dsi, value, DSI_BTA_TIMING);
+
+	return 0;
+}
+
+static int tegra_dsi_get_muldiv(enum mipi_dsi_pixel_format format,
+				unsigned int *mulp, unsigned int *divp)
+{
+	switch (format) {
+	case MIPI_DSI_FMT_RGB666_PACKED:
+	case MIPI_DSI_FMT_RGB888:
+		*mulp = 3;
+		*divp = 1;
+		break;
+
+	case MIPI_DSI_FMT_RGB565:
+		*mulp = 2;
+		*divp = 1;
+		break;
+
+	case MIPI_DSI_FMT_RGB666:
+		*mulp = 9;
+		*divp = 4;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int tegra_dsi_get_format(enum mipi_dsi_pixel_format format,
+				enum tegra_dsi_format *fmt)
+{
+	switch (format) {
+	case MIPI_DSI_FMT_RGB888:
+		*fmt = TEGRA_DSI_FORMAT_24P;
+		break;
+
+	case MIPI_DSI_FMT_RGB666:
+		*fmt = TEGRA_DSI_FORMAT_18NP;
+		break;
+
+	case MIPI_DSI_FMT_RGB666_PACKED:
+		*fmt = TEGRA_DSI_FORMAT_18P;
+		break;
+
+	case MIPI_DSI_FMT_RGB565:
+		*fmt = TEGRA_DSI_FORMAT_16P;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int tegra_output_dsi_enable(struct tegra_output *output)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	struct drm_display_mode *mode = &dc->base.mode;
+	unsigned int hact, hsw, hbp, hfp, i, mul, div;
+	struct tegra_dsi *dsi = to_dsi(output);
+	enum tegra_dsi_format format;
+	unsigned long value;
+	const u32 *pkt_seq;
+	int err;
+
+	if (dsi->enabled)
+		return 0;
+
+	if (dsi->flags & MIPI_DSI_MODE_VIDEO_SYNC_PULSE) {
+		DRM_DEBUG_KMS("Non-burst video mode with sync pulses\n");
+		pkt_seq = pkt_seq_video_non_burst_sync_pulses;
+	} else {
+		DRM_DEBUG_KMS("Non-burst video mode with sync events\n");
+		pkt_seq = pkt_seq_video_non_burst_sync_events;
+	}
+
+	err = tegra_dsi_get_muldiv(dsi->format, &mul, &div);
+	if (err < 0)
+		return err;
+
+	err = tegra_dsi_get_format(dsi->format, &format);
+	if (err < 0)
+		return err;
+
+	err = clk_enable(dsi->clk);
+	if (err < 0)
+		return err;
+
+	reset_control_deassert(dsi->rst);
+
+	value = DSI_CONTROL_CHANNEL(0) | DSI_CONTROL_FORMAT(format) |
+		DSI_CONTROL_LANES(dsi->lanes - 1) |
+		DSI_CONTROL_SOURCE(dc->pipe);
+	tegra_dsi_writel(dsi, value, DSI_CONTROL);
+
+	tegra_dsi_writel(dsi, DSI_VIDEO_FIFO_DEPTH, DSI_MAX_THRESHOLD);
+
+	value = DSI_HOST_CONTROL_HS | DSI_HOST_CONTROL_CS |
+		DSI_HOST_CONTROL_ECC;
+	tegra_dsi_writel(dsi, value, DSI_HOST_CONTROL);
+
+	value = tegra_dsi_readl(dsi, DSI_CONTROL);
+	value |= DSI_CONTROL_HS_CLK_CTRL;
+	value &= ~DSI_CONTROL_TX_TRIG(3);
+	value &= ~DSI_CONTROL_DCS_ENABLE;
+	value |= DSI_CONTROL_VIDEO_ENABLE;
+	value &= ~DSI_CONTROL_HOST_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_CONTROL);
+
+	err = tegra_dsi_set_phy_timing(dsi);
+	if (err < 0)
+		return err;
+
+	for (i = 0; i < NUM_PKT_SEQ; i++)
+		tegra_dsi_writel(dsi, pkt_seq[i], DSI_PKT_SEQ_0_LO + i);
+
+	/* horizontal active pixels */
+	hact = mode->hdisplay * mul / div;
+
+	/* horizontal sync width */
+	hsw = (mode->hsync_end - mode->hsync_start) * mul / div;
+	hsw -= 10;
+
+	/* horizontal back porch */
+	hbp = (mode->htotal - mode->hsync_end) * mul / div;
+	hbp -= 14;
+
+	/* horizontal front porch */
+	hfp = (mode->hsync_start  - mode->hdisplay) * mul / div;
+	hfp -= 8;
+
+	tegra_dsi_writel(dsi, hsw << 16 | 0, DSI_PKT_LEN_0_1);
+	tegra_dsi_writel(dsi, hact << 16 | hbp, DSI_PKT_LEN_2_3);
+	tegra_dsi_writel(dsi, hfp, DSI_PKT_LEN_4_5);
+	tegra_dsi_writel(dsi, 0x0f0f << 16, DSI_PKT_LEN_6_7);
+
+	/* set SOL delay */
+	tegra_dsi_writel(dsi, 8 * mul / div, DSI_SOL_DELAY);
+
+	/* enable display controller */
+	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+	value |= DSI_ENABLE;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+	value &= ~DISP_CTRL_MODE_MASK;
+	value |= DISP_CTRL_MODE_C_DISPLAY;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
+
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+	value |= PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		 PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	/* enable DSI controller */
+	value = tegra_dsi_readl(dsi, DSI_POWER_CONTROL);
+	value |= DSI_POWER_CONTROL_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_POWER_CONTROL);
+
+	dsi->enabled = true;
+
+	return 0;
+}
+
+static int tegra_output_dsi_disable(struct tegra_output *output)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	struct tegra_dsi *dsi = to_dsi(output);
+	unsigned long value;
+
+	if (!dsi->enabled)
+		return 0;
+
+	/* disable DSI controller */
+	value = tegra_dsi_readl(dsi, DSI_POWER_CONTROL);
+	value &= ~DSI_POWER_CONTROL_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_POWER_CONTROL);
+
+	/*
+	 * The following accesses registers of the display controller, so make
+	 * sure it's only executed when the output is attached to one.
+	 */
+	if (dc) {
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+		value &= ~(PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+			   PW4_ENABLE | PM0_ENABLE | PM1_ENABLE);
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+		value &= ~DISP_CTRL_MODE_MASK;
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
+
+		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+		value &= ~DSI_ENABLE;
+		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+		tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+		tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+	}
+
+	clk_disable(dsi->clk);
+
+	dsi->enabled = false;
+
+	return 0;
+}
+
+static int tegra_output_dsi_setup_clock(struct tegra_output *output,
+					struct clk *clk, unsigned long pclk,
+					unsigned int *divp)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	struct drm_display_mode *mode = &dc->base.mode;
+	unsigned int timeout, mul, div, vrefresh;
+	struct tegra_dsi *dsi = to_dsi(output);
+	unsigned long bclk, plld, value;
+	int err;
+
+	err = tegra_dsi_get_muldiv(dsi->format, &mul, &div);
+	if (err < 0)
+		return err;
+
+	DRM_DEBUG_KMS("mul: %u, div: %u, lanes: %u\n", mul, div, dsi->lanes);
+	vrefresh = drm_mode_vrefresh(mode);
+	DRM_DEBUG_KMS("vrefresh: %u\n", vrefresh);
+
+	/* compute byte clock */
+	bclk = (pclk * mul) / (div * dsi->lanes);
+
+	/*
+	 * Compute bit clock and round up to the next MHz.
+	 */
+	plld = DIV_ROUND_UP(bclk * 8, 1000000) * 1000000;
+
+	/*
+	 * We divide the frequency by two here, but we make up for that by
+	 * setting the shift clock divider (further below) to half of the
+	 * correct value.
+	 */
+	plld /= 2;
+
+	err = clk_set_parent(clk, dsi->clk_parent);
+	if (err < 0) {
+		dev_err(dsi->dev, "failed to set parent clock: %d\n", err);
+		return err;
+	}
+
+	err = clk_set_rate(dsi->clk_parent, plld);
+	if (err < 0) {
+		dev_err(dsi->dev, "failed to set base clock rate to %lu Hz\n",
+			plld);
+		return err;
+	}
+
+	/*
+	 * Derive pixel clock from bit clock using the shift clock divider.
+	 * Note that this is only half of what we would expect, but we need
+	 * that to make up for the fact that we divided the bit clock by a
+	 * factor of two above.
+	 *
+	 * It's not clear exactly why this is necessary, but the display is
+	 * not working properly otherwise. Perhaps the PLLs cannot generate
+	 * frequencies sufficiently high.
+	 */
+	*divp = ((8 * mul) / (div * dsi->lanes)) - 2;
+
+	/*
+	 * XXX: Move the below somewhere else so that we don't need to have
+	 * access to the vrefresh in this function?
+	 */
+
+	/* one frame high-speed transmission timeout */
+	timeout = (bclk / vrefresh) / 512;
+	value = DSI_TIMEOUT_LRX(0x2000) | DSI_TIMEOUT_HTX(timeout);
+	tegra_dsi_writel(dsi, value, DSI_TIMEOUT_0);
+
+	/* 2 ms peripheral timeout for panel */
+	timeout = 2 * bclk / 512 * 1000;
+	value = DSI_TIMEOUT_PR(timeout) | DSI_TIMEOUT_TA(0x2000);
+	tegra_dsi_writel(dsi, value, DSI_TIMEOUT_1);
+
+	value = DSI_TALLY_TA(0) | DSI_TALLY_LRX(0) | DSI_TALLY_HTX(0);
+	tegra_dsi_writel(dsi, value, DSI_TO_TALLY);
+
+	return 0;
+}
+
+static int tegra_output_dsi_check_mode(struct tegra_output *output,
+				       struct drm_display_mode *mode,
+				       enum drm_mode_status *status)
+{
+	/*
+	 * FIXME: For now, always assume that the mode is okay.
+	 */
+
+	*status = MODE_OK;
+
+	return 0;
+}
+
+static const struct tegra_output_ops dsi_ops = {
+	.enable = tegra_output_dsi_enable,
+	.disable = tegra_output_dsi_disable,
+	.setup_clock = tegra_output_dsi_setup_clock,
+	.check_mode = tegra_output_dsi_check_mode,
+};
+
+static int tegra_dsi_pad_enable(struct tegra_dsi *dsi)
+{
+	unsigned long value;
+
+	value = DSI_PAD_CONTROL_VS1_PULLDN(0) | DSI_PAD_CONTROL_VS1_PDIO(0);
+	tegra_dsi_writel(dsi, value, DSI_PAD_CONTROL_0);
+
+	return 0;
+}
+
+static int tegra_dsi_pad_calibrate(struct tegra_dsi *dsi)
+{
+	unsigned long value;
+
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_0);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_1);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_2);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_3);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_4);
+
+	/* start calibration */
+	tegra_dsi_pad_enable(dsi);
+
+	value = DSI_PAD_SLEW_UP(0x7) | DSI_PAD_SLEW_DN(0x7) |
+		DSI_PAD_LP_UP(0x1) | DSI_PAD_LP_DN(0x1) |
+		DSI_PAD_OUT_CLK(0x0);
+	tegra_dsi_writel(dsi, value, DSI_PAD_CONTROL_2);
+
+	return tegra_mipi_calibrate(dsi->mipi);
+}
+
+static int tegra_dsi_init(struct host1x_client *client)
+{
+	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct tegra_dsi *dsi = host1x_client_to_dsi(client);
+	int err;
+
+	dsi->output.type = TEGRA_OUTPUT_DSI;
+	dsi->output.dev = client->dev;
+	dsi->output.ops = &dsi_ops;
+
+	err = tegra_output_init(drm, &dsi->output);
+	if (err < 0) {
+		dev_err(client->dev, "output setup failed: %d\n", err);
+		return err;
+	}
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_dsi_debugfs_init(dsi, drm->primary);
+		if (err < 0)
+			dev_err(dsi->dev, "debugfs setup failed: %d\n", err);
+	}
+
+	err = tegra_dsi_pad_calibrate(dsi);
+	if (err < 0) {
+		dev_err(dsi->dev, "MIPI calibration failed: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static int tegra_dsi_exit(struct host1x_client *client)
+{
+	struct tegra_dsi *dsi = host1x_client_to_dsi(client);
+	int err;
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_dsi_debugfs_exit(dsi);
+		if (err < 0)
+			dev_err(dsi->dev, "debugfs cleanup failed: %d\n", err);
+	}
+
+	err = tegra_output_disable(&dsi->output);
+	if (err < 0) {
+		dev_err(client->dev, "output failed to disable: %d\n", err);
+		return err;
+	}
+
+	err = tegra_output_exit(&dsi->output);
+	if (err < 0) {
+		dev_err(client->dev, "output cleanup failed: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static const struct host1x_client_ops dsi_client_ops = {
+	.init = tegra_dsi_init,
+	.exit = tegra_dsi_exit,
+};
+
+static int tegra_dsi_setup_clocks(struct tegra_dsi *dsi)
+{
+	struct clk *parent;
+	int err;
+
+	parent = clk_get_parent(dsi->clk);
+	if (!parent)
+		return -EINVAL;
+
+	err = clk_set_parent(parent, dsi->clk_parent);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+
+static int tegra_dsi_host_attach(struct mipi_dsi_host *host,
+				 struct mipi_dsi_device *device)
+{
+	struct tegra_dsi *dsi = host_to_tegra(host);
+	struct tegra_output *output = &dsi->output;
+
+	dsi->flags = device->mode_flags;
+	dsi->format = device->format;
+	dsi->lanes = device->lanes;
+
+	output->panel = of_drm_find_panel(device->dev.of_node);
+	if (output->panel) {
+		if (output->connector.dev)
+			drm_helper_hpd_irq_event(output->connector.dev);
+	}
+
+	return 0;
+}
+
+static int tegra_dsi_host_detach(struct mipi_dsi_host *host,
+				 struct mipi_dsi_device *device)
+{
+	struct tegra_dsi *dsi = host_to_tegra(host);
+	struct tegra_output *output = &dsi->output;
+
+	if (output->panel && &device->dev == output->panel->dev) {
+		if (output->connector.dev)
+			drm_helper_hpd_irq_event(output->connector.dev);
+
+		output->panel = NULL;
+	}
+
+	return 0;
+}
+
+static const struct mipi_dsi_host_ops tegra_dsi_host_ops = {
+	.attach = tegra_dsi_host_attach,
+	.detach = tegra_dsi_host_detach,
+};
+
+static int tegra_dsi_probe(struct platform_device *pdev)
+{
+	struct tegra_dsi *dsi;
+	struct resource *regs;
+	int err;
+
+	dsi = devm_kzalloc(&pdev->dev, sizeof(*dsi), GFP_KERNEL);
+	if (!dsi)
+		return -ENOMEM;
+
+	dsi->output.dev = dsi->dev = &pdev->dev;
+
+	err = tegra_output_probe(&dsi->output);
+	if (err < 0)
+		return err;
+
+	/*
+	 * Assume these values by default. When a DSI peripheral driver
+	 * attaches to the DSI host, the parameters will be taken from
+	 * the attached device.
+	 */
+	dsi->flags = MIPI_DSI_MODE_VIDEO;
+	dsi->format = MIPI_DSI_FMT_RGB888;
+	dsi->lanes = 4;
+
+	dsi->rst = devm_reset_control_get(&pdev->dev, "dsi");
+	if (IS_ERR(dsi->rst))
+		return PTR_ERR(dsi->rst);
+
+	dsi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(dsi->clk)) {
+		dev_err(&pdev->dev, "cannot get DSI clock\n");
+		return PTR_ERR(dsi->clk);
+	}
+
+	err = clk_prepare_enable(dsi->clk);
+	if (err < 0) {
+		dev_err(&pdev->dev, "cannot enable DSI clock\n");
+		return err;
+	}
+
+	dsi->clk_lp = devm_clk_get(&pdev->dev, "lp");
+	if (IS_ERR(dsi->clk_lp)) {
+		dev_err(&pdev->dev, "cannot get low-power clock\n");
+		return PTR_ERR(dsi->clk_lp);
+	}
+
+	err = clk_prepare_enable(dsi->clk_lp);
+	if (err < 0) {
+		dev_err(&pdev->dev, "cannot enable low-power clock\n");
+		return err;
+	}
+
+	dsi->clk_parent = devm_clk_get(&pdev->dev, "parent");
+	if (IS_ERR(dsi->clk_parent)) {
+		dev_err(&pdev->dev, "cannot get parent clock\n");
+		return PTR_ERR(dsi->clk_parent);
+	}
+
+	err = clk_prepare_enable(dsi->clk_parent);
+	if (err < 0) {
+		dev_err(&pdev->dev, "cannot enable parent clock\n");
+		return err;
+	}
+
+	dsi->vdd = devm_regulator_get(&pdev->dev, "avdd-dsi-csi");
+	if (IS_ERR(dsi->vdd)) {
+		dev_err(&pdev->dev, "cannot get VDD supply\n");
+		return PTR_ERR(dsi->vdd);
+	}
+
+	err = regulator_enable(dsi->vdd);
+	if (err < 0) {
+		dev_err(&pdev->dev, "cannot enable VDD supply\n");
+		return err;
+	}
+
+	err = tegra_dsi_setup_clocks(dsi);
+	if (err < 0) {
+		dev_err(&pdev->dev, "cannot setup clocks\n");
+		return err;
+	}
+
+	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dsi->regs = devm_ioremap_resource(&pdev->dev, regs);
+	if (IS_ERR(dsi->regs))
+		return PTR_ERR(dsi->regs);
+
+	dsi->mipi = tegra_mipi_request(&pdev->dev);
+	if (IS_ERR(dsi->mipi))
+		return PTR_ERR(dsi->mipi);
+
+	dsi->host.ops = &tegra_dsi_host_ops;
+	dsi->host.dev = &pdev->dev;
+
+	err = mipi_dsi_host_register(&dsi->host);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to register DSI host: %d\n", err);
+		return err;
+	}
+
+	INIT_LIST_HEAD(&dsi->client.list);
+	dsi->client.ops = &dsi_client_ops;
+	dsi->client.dev = &pdev->dev;
+
+	err = host1x_client_register(&dsi->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to register host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	platform_set_drvdata(pdev, dsi);
+
+	return 0;
+}
+
+static int tegra_dsi_remove(struct platform_device *pdev)
+{
+	struct tegra_dsi *dsi = platform_get_drvdata(pdev);
+	int err;
+
+	err = host1x_client_unregister(&dsi->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to unregister host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	mipi_dsi_host_unregister(&dsi->host);
+	tegra_mipi_free(dsi->mipi);
+
+	regulator_disable(dsi->vdd);
+	clk_disable_unprepare(dsi->clk_parent);
+	clk_disable_unprepare(dsi->clk_lp);
+	clk_disable_unprepare(dsi->clk);
+	reset_control_assert(dsi->rst);
+
+	err = tegra_output_remove(&dsi->output);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to remove output: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static const struct of_device_id tegra_dsi_of_match[] = {
+	{ .compatible = "nvidia,tegra114-dsi", },
+	{ },
+};
+
+struct platform_driver tegra_dsi_driver = {
+	.driver = {
+		.name = "tegra-dsi",
+		.of_match_table = tegra_dsi_of_match,
+	},
+	.probe = tegra_dsi_probe,
+	.remove = tegra_dsi_remove,
+};
diff --git a/drivers/gpu/drm/tegra/dsi.h b/drivers/gpu/drm/tegra/dsi.h
new file mode 100644
index 00000000000..5ce610d08d7
--- /dev/null
+++ b/drivers/gpu/drm/tegra/dsi.h
@@ -0,0 +1,130 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef DRM_TEGRA_DSI_H
+#define DRM_TEGRA_DSI_H
+
+#define DSI_INCR_SYNCPT			0x00
+#define DSI_INCR_SYNCPT_CONTROL		0x01
+#define DSI_INCR_SYNCPT_ERROR		0x02
+#define DSI_CTXSW			0x08
+#define DSI_RD_DATA			0x09
+#define DSI_WR_DATA			0x0a
+#define DSI_POWER_CONTROL		0x0b
+#define DSI_POWER_CONTROL_ENABLE	(1 << 0)
+#define DSI_INT_ENABLE			0x0c
+#define DSI_INT_STATUS			0x0d
+#define DSI_INT_MASK			0x0e
+#define DSI_HOST_CONTROL		0x0f
+#define DSI_HOST_CONTROL_RAW		(1 << 6)
+#define DSI_HOST_CONTROL_HS		(1 << 5)
+#define DSI_HOST_CONTROL_BTA		(1 << 2)
+#define DSI_HOST_CONTROL_CS		(1 << 1)
+#define DSI_HOST_CONTROL_ECC		(1 << 0)
+#define DSI_CONTROL			0x10
+#define DSI_CONTROL_HS_CLK_CTRL		(1 << 20)
+#define DSI_CONTROL_CHANNEL(c)		(((c) & 0x3) << 16)
+#define DSI_CONTROL_FORMAT(f)		(((f) & 0x3) << 12)
+#define DSI_CONTROL_TX_TRIG(x)		(((x) & 0x3) <<  8)
+#define DSI_CONTROL_LANES(n)		(((n) & 0x3) <<  4)
+#define DSI_CONTROL_DCS_ENABLE		(1 << 3)
+#define DSI_CONTROL_SOURCE(s)		(((s) & 0x1) <<  2)
+#define DSI_CONTROL_VIDEO_ENABLE	(1 << 1)
+#define DSI_CONTROL_HOST_ENABLE		(1 << 0)
+#define DSI_SOL_DELAY			0x11
+#define DSI_MAX_THRESHOLD		0x12
+#define DSI_TRIGGER			0x13
+#define DSI_TX_CRC			0x14
+#define DSI_STATUS			0x15
+#define DSI_STATUS_IDLE			(1 << 10)
+#define DSI_INIT_SEQ_CONTROL		0x1a
+#define DSI_INIT_SEQ_DATA_0		0x1b
+#define DSI_INIT_SEQ_DATA_1		0x1c
+#define DSI_INIT_SEQ_DATA_2		0x1d
+#define DSI_INIT_SEQ_DATA_3		0x1e
+#define DSI_INIT_SEQ_DATA_4		0x1f
+#define DSI_INIT_SEQ_DATA_5		0x20
+#define DSI_INIT_SEQ_DATA_6		0x21
+#define DSI_INIT_SEQ_DATA_7		0x22
+#define DSI_PKT_SEQ_0_LO		0x23
+#define DSI_PKT_SEQ_0_HI		0x24
+#define DSI_PKT_SEQ_1_LO		0x25
+#define DSI_PKT_SEQ_1_HI		0x26
+#define DSI_PKT_SEQ_2_LO		0x27
+#define DSI_PKT_SEQ_2_HI		0x28
+#define DSI_PKT_SEQ_3_LO		0x29
+#define DSI_PKT_SEQ_3_HI		0x2a
+#define DSI_PKT_SEQ_4_LO		0x2b
+#define DSI_PKT_SEQ_4_HI		0x2c
+#define DSI_PKT_SEQ_5_LO		0x2d
+#define DSI_PKT_SEQ_5_HI		0x2e
+#define DSI_DCS_CMDS			0x33
+#define DSI_PKT_LEN_0_1			0x34
+#define DSI_PKT_LEN_2_3			0x35
+#define DSI_PKT_LEN_4_5			0x36
+#define DSI_PKT_LEN_6_7			0x37
+#define DSI_PHY_TIMING_0		0x3c
+#define DSI_PHY_TIMING_1		0x3d
+#define DSI_PHY_TIMING_2		0x3e
+#define DSI_BTA_TIMING			0x3f
+
+#define DSI_TIMING_FIELD(value, period, hwinc) \
+	((DIV_ROUND_CLOSEST(value, period) - (hwinc)) & 0xff)
+
+#define DSI_TIMEOUT_0			0x44
+#define DSI_TIMEOUT_LRX(x)		(((x) & 0xffff) << 16)
+#define DSI_TIMEOUT_HTX(x)		(((x) & 0xffff) <<  0)
+#define DSI_TIMEOUT_1			0x45
+#define DSI_TIMEOUT_PR(x)		(((x) & 0xffff) << 16)
+#define DSI_TIMEOUT_TA(x)		(((x) & 0xffff) <<  0)
+#define DSI_TO_TALLY			0x46
+#define DSI_TALLY_TA(x)			(((x) & 0xff) << 16)
+#define DSI_TALLY_LRX(x)		(((x) & 0xff) <<  8)
+#define DSI_TALLY_HTX(x)		(((x) & 0xff) <<  0)
+#define DSI_PAD_CONTROL_0		0x4b
+#define DSI_PAD_CONTROL_VS1_PDIO(x)	(((x) & 0xf) <<  0)
+#define DSI_PAD_CONTROL_VS1_PDIO_CLK	(1 <<  8)
+#define DSI_PAD_CONTROL_VS1_PULLDN(x)	(((x) & 0xf) << 16)
+#define DSI_PAD_CONTROL_VS1_PULLDN_CLK	(1 << 24)
+#define DSI_PAD_CONTROL_CD		0x4c
+#define DSI_PAD_CD_STATUS		0x4d
+#define DSI_VIDEO_MODE_CONTROL		0x4e
+#define DSI_PAD_CONTROL_1		0x4f
+#define DSI_PAD_CONTROL_2		0x50
+#define DSI_PAD_OUT_CLK(x)		(((x) & 0x7) <<  0)
+#define DSI_PAD_LP_DN(x)		(((x) & 0x7) <<  4)
+#define DSI_PAD_LP_UP(x)		(((x) & 0x7) <<  8)
+#define DSI_PAD_SLEW_DN(x)		(((x) & 0x7) << 12)
+#define DSI_PAD_SLEW_UP(x)		(((x) & 0x7) << 16)
+#define DSI_PAD_CONTROL_3		0x51
+#define DSI_PAD_CONTROL_4		0x52
+#define DSI_GANGED_MODE_CONTROL		0x53
+#define DSI_GANGED_MODE_START		0x54
+#define DSI_GANGED_MODE_SIZE		0x55
+#define DSI_RAW_DATA_BYTE_COUNT		0x56
+#define DSI_ULTRA_LOW_POWER_CONTROL	0x57
+#define DSI_INIT_SEQ_DATA_8		0x58
+#define DSI_INIT_SEQ_DATA_9		0x59
+#define DSI_INIT_SEQ_DATA_10		0x5a
+#define DSI_INIT_SEQ_DATA_11		0x5b
+#define DSI_INIT_SEQ_DATA_12		0x5c
+#define DSI_INIT_SEQ_DATA_13		0x5d
+#define DSI_INIT_SEQ_DATA_14		0x5e
+#define DSI_INIT_SEQ_DATA_15		0x5f
+
+/*
+ * pixel format as used in the DSI_CONTROL_FORMAT field
+ */
+enum tegra_dsi_format {
+	TEGRA_DSI_FORMAT_16P,
+	TEGRA_DSI_FORMAT_18NP,
+	TEGRA_DSI_FORMAT_18P,
+	TEGRA_DSI_FORMAT_24P,
+};
+
+#endif
diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index a3835e7de18..9798a708032 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -18,10 +18,12 @@ static inline struct tegra_fb *to_tegra_fb(struct drm_framebuffer *fb)
 	return container_of(fb, struct tegra_fb, base);
 }
 
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 static inline struct tegra_fbdev *to_tegra_fbdev(struct drm_fb_helper *helper)
 {
 	return container_of(helper, struct tegra_fbdev, base);
 }
+#endif
 
 struct tegra_bo *tegra_fb_get_plane(struct drm_framebuffer *framebuffer,
 				    unsigned int index)
@@ -98,8 +100,10 @@ static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
 		return ERR_PTR(-ENOMEM);
 
 	fb->planes = kzalloc(num_planes * sizeof(*planes), GFP_KERNEL);
-	if (!fb->planes)
+	if (!fb->planes) {
+		kfree(fb);
 		return ERR_PTR(-ENOMEM);
+	}
 
 	fb->num_planes = num_planes;
 
@@ -172,6 +176,7 @@ unreference:
 	return ERR_PTR(err);
 }
 
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 static struct fb_ops tegra_fb_ops = {
 	.owner = THIS_MODULE,
 	.fb_fillrect = sys_fillrect,
@@ -339,6 +344,12 @@ static void tegra_fbdev_free(struct tegra_fbdev *fbdev)
 	kfree(fbdev);
 }
 
+void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev)
+{
+	if (fbdev)
+		drm_fb_helper_restore_fbdev_mode_unlocked(&fbdev->base);
+}
+
 static void tegra_fb_output_poll_changed(struct drm_device *drm)
 {
 	struct tegra_drm *tegra = drm->dev_private;
@@ -346,16 +357,20 @@ static void tegra_fb_output_poll_changed(struct drm_device *drm)
 	if (tegra->fbdev)
 		drm_fb_helper_hotplug_event(&tegra->fbdev->base);
 }
+#endif
 
 static const struct drm_mode_config_funcs tegra_drm_mode_funcs = {
 	.fb_create = tegra_fb_create,
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 	.output_poll_changed = tegra_fb_output_poll_changed,
+#endif
 };
 
 int tegra_drm_fb_init(struct drm_device *drm)
 {
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 	struct tegra_drm *tegra = drm->dev_private;
-	struct tegra_fbdev *fbdev;
+#endif
 
 	drm->mode_config.min_width = 0;
 	drm->mode_config.min_height = 0;
@@ -365,28 +380,21 @@ int tegra_drm_fb_init(struct drm_device *drm)
 
 	drm->mode_config.funcs = &tegra_drm_mode_funcs;
 
-	fbdev = tegra_fbdev_create(drm, 32, drm->mode_config.num_crtc,
-				   drm->mode_config.num_connector);
-	if (IS_ERR(fbdev))
-		return PTR_ERR(fbdev);
-
-	tegra->fbdev = fbdev;
+#ifdef CONFIG_DRM_TEGRA_FBDEV
+	tegra->fbdev = tegra_fbdev_create(drm, 32, drm->mode_config.num_crtc,
+					  drm->mode_config.num_connector);
+	if (IS_ERR(tegra->fbdev))
+		return PTR_ERR(tegra->fbdev);
+#endif
 
 	return 0;
 }
 
 void tegra_drm_fb_exit(struct drm_device *drm)
 {
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 	struct tegra_drm *tegra = drm->dev_private;
 
 	tegra_fbdev_free(tegra->fbdev);
-}
-
-void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev)
-{
-	if (fbdev) {
-		drm_modeset_lock_all(fbdev->base.dev);
-		drm_fb_helper_restore_fbdev_mode(&fbdev->base);
-		drm_modeset_unlock_all(fbdev->base.dev);
-	}
+#endif
 }
diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index 28a9cbc07ab..aa85b7b26f1 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -8,16 +8,12 @@
  *
  * Copyright (c) 2011 Samsung Electronics Co., Ltd.
  *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version 2
- * of the License, or (at your option) any later version.
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
  */
 
+#include <linux/dma-buf.h>
 #include <drm/tegra_drm.h>
 
 #include "gem.h"
@@ -83,7 +79,7 @@ static struct host1x_bo *tegra_bo_get(struct host1x_bo *bo)
 	return bo;
 }
 
-const struct host1x_bo_ops tegra_bo_ops = {
+static const struct host1x_bo_ops tegra_bo_ops = {
 	.get = tegra_bo_get,
 	.put = tegra_bo_put,
 	.pin = tegra_bo_pin,
@@ -145,7 +141,6 @@ err_dma:
 	kfree(bo);
 
 	return ERR_PTR(err);
-
 }
 
 struct tegra_bo *tegra_bo_create_with_handle(struct drm_file *file,
@@ -174,13 +169,88 @@ err:
 	return ERR_PTR(ret);
 }
 
+static struct tegra_bo *tegra_bo_import(struct drm_device *drm,
+					struct dma_buf *buf)
+{
+	struct dma_buf_attachment *attach;
+	struct tegra_bo *bo;
+	ssize_t size;
+	int err;
+
+	bo = kzalloc(sizeof(*bo), GFP_KERNEL);
+	if (!bo)
+		return ERR_PTR(-ENOMEM);
+
+	host1x_bo_init(&bo->base, &tegra_bo_ops);
+	size = round_up(buf->size, PAGE_SIZE);
+
+	err = drm_gem_object_init(drm, &bo->gem, size);
+	if (err < 0)
+		goto free;
+
+	err = drm_gem_create_mmap_offset(&bo->gem);
+	if (err < 0)
+		goto release;
+
+	attach = dma_buf_attach(buf, drm->dev);
+	if (IS_ERR(attach)) {
+		err = PTR_ERR(attach);
+		goto free_mmap;
+	}
+
+	get_dma_buf(buf);
+
+	bo->sgt = dma_buf_map_attachment(attach, DMA_TO_DEVICE);
+	if (!bo->sgt) {
+		err = -ENOMEM;
+		goto detach;
+	}
+
+	if (IS_ERR(bo->sgt)) {
+		err = PTR_ERR(bo->sgt);
+		goto detach;
+	}
+
+	if (bo->sgt->nents > 1) {
+		err = -EINVAL;
+		goto detach;
+	}
+
+	bo->paddr = sg_dma_address(bo->sgt->sgl);
+	bo->gem.import_attach = attach;
+
+	return bo;
+
+detach:
+	if (!IS_ERR_OR_NULL(bo->sgt))
+		dma_buf_unmap_attachment(attach, bo->sgt, DMA_TO_DEVICE);
+
+	dma_buf_detach(buf, attach);
+	dma_buf_put(buf);
+free_mmap:
+	drm_gem_free_mmap_offset(&bo->gem);
+release:
+	drm_gem_object_release(&bo->gem);
+free:
+	kfree(bo);
+
+	return ERR_PTR(err);
+}
+
 void tegra_bo_free_object(struct drm_gem_object *gem)
 {
 	struct tegra_bo *bo = to_tegra_bo(gem);
 
+	if (gem->import_attach) {
+		dma_buf_unmap_attachment(gem->import_attach, bo->sgt,
+					 DMA_TO_DEVICE);
+		drm_prime_gem_destroy(gem, NULL);
+	} else {
+		tegra_bo_destroy(gem->dev, bo);
+	}
+
 	drm_gem_free_mmap_offset(gem);
 	drm_gem_object_release(gem);
-	tegra_bo_destroy(gem->dev, bo);
 
 	kfree(bo);
 }
@@ -256,3 +326,120 @@ int tegra_drm_mmap(struct file *file, struct vm_area_struct *vma)
 
 	return ret;
 }
+
+static struct sg_table *
+tegra_gem_prime_map_dma_buf(struct dma_buf_attachment *attach,
+			    enum dma_data_direction dir)
+{
+	struct drm_gem_object *gem = attach->dmabuf->priv;
+	struct tegra_bo *bo = to_tegra_bo(gem);
+	struct sg_table *sgt;
+
+	sgt = kmalloc(sizeof(*sgt), GFP_KERNEL);
+	if (!sgt)
+		return NULL;
+
+	if (sg_alloc_table(sgt, 1, GFP_KERNEL)) {
+		kfree(sgt);
+		return NULL;
+	}
+
+	sg_dma_address(sgt->sgl) = bo->paddr;
+	sg_dma_len(sgt->sgl) = gem->size;
+
+	return sgt;
+}
+
+static void tegra_gem_prime_unmap_dma_buf(struct dma_buf_attachment *attach,
+					  struct sg_table *sgt,
+					  enum dma_data_direction dir)
+{
+	sg_free_table(sgt);
+	kfree(sgt);
+}
+
+static void tegra_gem_prime_release(struct dma_buf *buf)
+{
+	drm_gem_dmabuf_release(buf);
+}
+
+static void *tegra_gem_prime_kmap_atomic(struct dma_buf *buf,
+					 unsigned long page)
+{
+	return NULL;
+}
+
+static void tegra_gem_prime_kunmap_atomic(struct dma_buf *buf,
+					  unsigned long page,
+					  void *addr)
+{
+}
+
+static void *tegra_gem_prime_kmap(struct dma_buf *buf, unsigned long page)
+{
+	return NULL;
+}
+
+static void tegra_gem_prime_kunmap(struct dma_buf *buf, unsigned long page,
+				   void *addr)
+{
+}
+
+static int tegra_gem_prime_mmap(struct dma_buf *buf, struct vm_area_struct *vma)
+{
+	return -EINVAL;
+}
+
+static void *tegra_gem_prime_vmap(struct dma_buf *buf)
+{
+	struct drm_gem_object *gem = buf->priv;
+	struct tegra_bo *bo = to_tegra_bo(gem);
+
+	return bo->vaddr;
+}
+
+static void tegra_gem_prime_vunmap(struct dma_buf *buf, void *vaddr)
+{
+}
+
+static const struct dma_buf_ops tegra_gem_prime_dmabuf_ops = {
+	.map_dma_buf = tegra_gem_prime_map_dma_buf,
+	.unmap_dma_buf = tegra_gem_prime_unmap_dma_buf,
+	.release = tegra_gem_prime_release,
+	.kmap_atomic = tegra_gem_prime_kmap_atomic,
+	.kunmap_atomic = tegra_gem_prime_kunmap_atomic,
+	.kmap = tegra_gem_prime_kmap,
+	.kunmap = tegra_gem_prime_kunmap,
+	.mmap = tegra_gem_prime_mmap,
+	.vmap = tegra_gem_prime_vmap,
+	.vunmap = tegra_gem_prime_vunmap,
+};
+
+struct dma_buf *tegra_gem_prime_export(struct drm_device *drm,
+				       struct drm_gem_object *gem,
+				       int flags)
+{
+	return dma_buf_export(gem, &tegra_gem_prime_dmabuf_ops, gem->size,
+			      flags);
+}
+
+struct drm_gem_object *tegra_gem_prime_import(struct drm_device *drm,
+					      struct dma_buf *buf)
+{
+	struct tegra_bo *bo;
+
+	if (buf->ops == &tegra_gem_prime_dmabuf_ops) {
+		struct drm_gem_object *gem = buf->priv;
+
+		if (gem->dev == drm) {
+			drm_gem_object_reference(gem);
+			return gem;
+		}
+	}
+
+	bo = tegra_bo_import(drm, buf);
+	if (IS_ERR(bo))
+		return ERR_CAST(bo);
+
+	return &bo->gem;
+}
diff --git a/drivers/gpu/drm/tegra/gem.h b/drivers/gpu/drm/tegra/gem.h
index 7674000bf47..2f3fe96c515 100644
--- a/drivers/gpu/drm/tegra/gem.h
+++ b/drivers/gpu/drm/tegra/gem.h
@@ -3,17 +3,9 @@
  *
  * Copyright (c) 2012-2013, NVIDIA Corporation.
  *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
  */
 
 #ifndef __HOST1X_GEM_H
@@ -31,6 +23,7 @@ struct tegra_bo {
 	struct drm_gem_object gem;
 	struct host1x_bo base;
 	unsigned long flags;
+	struct sg_table *sgt;
 	dma_addr_t paddr;
 	void *vaddr;
 };
@@ -40,8 +33,6 @@ static inline struct tegra_bo *to_tegra_bo(struct drm_gem_object *gem)
 	return container_of(gem, struct tegra_bo, gem);
 }
 
-extern const struct host1x_bo_ops tegra_bo_ops;
-
 struct tegra_bo *tegra_bo_create(struct drm_device *drm, unsigned int size,
 				 unsigned long flags);
 struct tegra_bo *tegra_bo_create_with_handle(struct drm_file *file,
@@ -59,4 +50,10 @@ int tegra_drm_mmap(struct file *file, struct vm_area_struct *vma);
 
 extern const struct vm_operations_struct tegra_bo_vm_ops;
 
+struct dma_buf *tegra_gem_prime_export(struct drm_device *drm,
+				       struct drm_gem_object *gem,
+				       int flags);
+struct drm_gem_object *tegra_gem_prime_import(struct drm_device *drm,
+					      struct dma_buf *buf);
+
 #endif
diff --git a/drivers/gpu/drm/tegra/gr2d.c b/drivers/gpu/drm/tegra/gr2d.c
index 7ec4259ffde..7c53941f2a9 100644
--- a/drivers/gpu/drm/tegra/gr2d.c
+++ b/drivers/gpu/drm/tegra/gr2d.c
@@ -1,17 +1,9 @@
 /*
  * Copyright (c) 2012-2013, NVIDIA Corporation.
  *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
  */
 
 #include <linux/clk.h>
@@ -36,7 +28,7 @@ static inline struct gr2d *to_gr2d(struct tegra_drm_client *client)
 static int gr2d_init(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
-	struct tegra_drm *tegra = dev_get_drvdata(client->parent);
+	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
 	struct gr2d *gr2d = to_gr2d(drm);
 
@@ -50,17 +42,17 @@ static int gr2d_init(struct host1x_client *client)
 		return -ENOMEM;
 	}
 
-	return tegra_drm_register_client(tegra, drm);
+	return tegra_drm_register_client(dev->dev_private, drm);
 }
 
 static int gr2d_exit(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
-	struct tegra_drm *tegra = dev_get_drvdata(client->parent);
+	struct drm_device *dev = dev_get_drvdata(client->parent);
 	struct gr2d *gr2d = to_gr2d(drm);
 	int err;
 
-	err = tegra_drm_unregister_client(tegra, drm);
+	err = tegra_drm_unregister_client(dev->dev_private, drm);
 	if (err < 0)
 		return err;
 
diff --git a/drivers/gpu/drm/tegra/gr3d.c b/drivers/gpu/drm/tegra/gr3d.c
index 4cec8f526af..30f5ba9bd6d 100644
--- a/drivers/gpu/drm/tegra/gr3d.c
+++ b/drivers/gpu/drm/tegra/gr3d.c
@@ -11,6 +11,7 @@
 #include <linux/host1x.h>
 #include <linux/module.h>
 #include <linux/platform_device.h>
+#include <linux/reset.h>
 #include <linux/tegra-powergate.h>
 
 #include "drm.h"
@@ -22,6 +23,8 @@ struct gr3d {
 	struct host1x_channel *channel;
 	struct clk *clk_secondary;
 	struct clk *clk;
+	struct reset_control *rst_secondary;
+	struct reset_control *rst;
 
 	DECLARE_BITMAP(addr_regs, GR3D_NUM_REGS);
 };
@@ -34,7 +37,7 @@ static inline struct gr3d *to_gr3d(struct tegra_drm_client *client)
 static int gr3d_init(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
-	struct tegra_drm *tegra = dev_get_drvdata(client->parent);
+	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
 	struct gr3d *gr3d = to_gr3d(drm);
 
@@ -48,17 +51,17 @@ static int gr3d_init(struct host1x_client *client)
 		return -ENOMEM;
 	}
 
-	return tegra_drm_register_client(tegra, drm);
+	return tegra_drm_register_client(dev->dev_private, drm);
 }
 
 static int gr3d_exit(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
-	struct tegra_drm *tegra = dev_get_drvdata(client->parent);
+	struct drm_device *dev = dev_get_drvdata(client->parent);
 	struct gr3d *gr3d = to_gr3d(drm);
 	int err;
 
-	err = tegra_drm_unregister_client(tegra, drm);
+	err = tegra_drm_unregister_client(dev->dev_private, drm);
 	if (err < 0)
 		return err;
 
@@ -255,15 +258,29 @@ static int gr3d_probe(struct platform_device *pdev)
 		return PTR_ERR(gr3d->clk);
 	}
 
+	gr3d->rst = devm_reset_control_get(&pdev->dev, "3d");
+	if (IS_ERR(gr3d->rst)) {
+		dev_err(&pdev->dev, "cannot get reset\n");
+		return PTR_ERR(gr3d->rst);
+	}
+
 	if (of_device_is_compatible(np, "nvidia,tegra30-gr3d")) {
 		gr3d->clk_secondary = devm_clk_get(&pdev->dev, "3d2");
 		if (IS_ERR(gr3d->clk)) {
 			dev_err(&pdev->dev, "cannot get secondary clock\n");
 			return PTR_ERR(gr3d->clk);
 		}
+
+		gr3d->rst_secondary = devm_reset_control_get(&pdev->dev,
+								"3d2");
+		if (IS_ERR(gr3d->rst_secondary)) {
+			dev_err(&pdev->dev, "cannot get secondary reset\n");
+			return PTR_ERR(gr3d->rst_secondary);
+		}
 	}
 
-	err = tegra_powergate_sequence_power_up(TEGRA_POWERGATE_3D, gr3d->clk);
+	err = tegra_powergate_sequence_power_up(TEGRA_POWERGATE_3D, gr3d->clk,
+						gr3d->rst);
 	if (err < 0) {
 		dev_err(&pdev->dev, "failed to power up 3D unit\n");
 		return err;
@@ -271,7 +288,8 @@ static int gr3d_probe(struct platform_device *pdev)
 
 	if (gr3d->clk_secondary) {
 		err = tegra_powergate_sequence_power_up(TEGRA_POWERGATE_3D1,
-							gr3d->clk_secondary);
+							gr3d->clk_secondary,
+							gr3d->rst_secondary);
 		if (err < 0) {
 			dev_err(&pdev->dev,
 				"failed to power up secondary 3D unit\n");
diff --git a/drivers/gpu/drm/tegra/hdmi.c b/drivers/gpu/drm/tegra/hdmi.c
index 0cd9bc2056e..ba067bb767e 100644
--- a/drivers/gpu/drm/tegra/hdmi.c
+++ b/drivers/gpu/drm/tegra/hdmi.c
@@ -8,10 +8,10 @@
  */
 
 #include <linux/clk.h>
-#include <linux/clk/tegra.h>
 #include <linux/debugfs.h>
 #include <linux/hdmi.h>
 #include <linux/regulator/consumer.h>
+#include <linux/reset.h>
 
 #include "hdmi.h"
 #include "drm.h"
@@ -40,15 +40,18 @@ struct tegra_hdmi {
 	struct host1x_client client;
 	struct tegra_output output;
 	struct device *dev;
+	bool enabled;
 
-	struct regulator *vdd;
+	struct regulator *hdmi;
 	struct regulator *pll;
+	struct regulator *vdd;
 
 	void __iomem *regs;
 	unsigned int irq;
 
 	struct clk *clk_parent;
 	struct clk *clk;
+	struct reset_control *rst;
 
 	const struct tegra_hdmi_config *config;
 
@@ -315,6 +318,85 @@ static const struct tmds_config tegra114_tmds_config[] = {
 	},
 };
 
+static const struct tmds_config tegra124_tmds_config[] = {
+	{ /* 480p/576p / 25.2MHz/27MHz modes */
+		.pclk = 27000000,
+		.pll0 = SOR_PLL_ICHPMP(1) | SOR_PLL_BG_V17_S(3) |
+			SOR_PLL_VCOCAP(0) | SOR_PLL_RESISTORSEL,
+		.pll1 = SOR_PLL_LOADADJ(3) | SOR_PLL_TMDS_TERMADJ(0),
+		.pe_current = PE_CURRENT0(PE_CURRENT_0_mA_T114) |
+			PE_CURRENT1(PE_CURRENT_0_mA_T114) |
+			PE_CURRENT2(PE_CURRENT_0_mA_T114) |
+			PE_CURRENT3(PE_CURRENT_0_mA_T114),
+		.drive_current =
+			DRIVE_CURRENT_LANE0_T114(DRIVE_CURRENT_10_400_mA_T114) |
+			DRIVE_CURRENT_LANE1_T114(DRIVE_CURRENT_10_400_mA_T114) |
+			DRIVE_CURRENT_LANE2_T114(DRIVE_CURRENT_10_400_mA_T114) |
+			DRIVE_CURRENT_LANE3_T114(DRIVE_CURRENT_10_400_mA_T114),
+		.peak_current = PEAK_CURRENT_LANE0(PEAK_CURRENT_0_000_mA) |
+			PEAK_CURRENT_LANE1(PEAK_CURRENT_0_000_mA) |
+			PEAK_CURRENT_LANE2(PEAK_CURRENT_0_000_mA) |
+			PEAK_CURRENT_LANE3(PEAK_CURRENT_0_000_mA),
+	}, { /* 720p / 74.25MHz modes */
+		.pclk = 74250000,
+		.pll0 = SOR_PLL_ICHPMP(1) | SOR_PLL_BG_V17_S(3) |
+			SOR_PLL_VCOCAP(1) | SOR_PLL_RESISTORSEL,
+		.pll1 = SOR_PLL_PE_EN | SOR_PLL_LOADADJ(3) |
+			SOR_PLL_TMDS_TERMADJ(0),
+		.pe_current = PE_CURRENT0(PE_CURRENT_15_mA_T114) |
+			PE_CURRENT1(PE_CURRENT_15_mA_T114) |
+			PE_CURRENT2(PE_CURRENT_15_mA_T114) |
+			PE_CURRENT3(PE_CURRENT_15_mA_T114),
+		.drive_current =
+			DRIVE_CURRENT_LANE0_T114(DRIVE_CURRENT_10_400_mA_T114) |
+			DRIVE_CURRENT_LANE1_T114(DRIVE_CURRENT_10_400_mA_T114) |
+			DRIVE_CURRENT_LANE2_T114(DRIVE_CURRENT_10_400_mA_T114) |
+			DRIVE_CURRENT_LANE3_T114(DRIVE_CURRENT_10_400_mA_T114),
+		.peak_current = PEAK_CURRENT_LANE0(PEAK_CURRENT_0_000_mA) |
+			PEAK_CURRENT_LANE1(PEAK_CURRENT_0_000_mA) |
+			PEAK_CURRENT_LANE2(PEAK_CURRENT_0_000_mA) |
+			PEAK_CURRENT_LANE3(PEAK_CURRENT_0_000_mA),
+	}, { /* 1080p / 148.5MHz modes */
+		.pclk = 148500000,
+		.pll0 = SOR_PLL_ICHPMP(1) | SOR_PLL_BG_V17_S(3) |
+			SOR_PLL_VCOCAP(3) | SOR_PLL_RESISTORSEL,
+		.pll1 = SOR_PLL_PE_EN | SOR_PLL_LOADADJ(3) |
+			SOR_PLL_TMDS_TERMADJ(0),
+		.pe_current = PE_CURRENT0(PE_CURRENT_10_mA_T114) |
+			PE_CURRENT1(PE_CURRENT_10_mA_T114) |
+			PE_CURRENT2(PE_CURRENT_10_mA_T114) |
+			PE_CURRENT3(PE_CURRENT_10_mA_T114),
+		.drive_current =
+			DRIVE_CURRENT_LANE0_T114(DRIVE_CURRENT_12_400_mA_T114) |
+			DRIVE_CURRENT_LANE1_T114(DRIVE_CURRENT_12_400_mA_T114) |
+			DRIVE_CURRENT_LANE2_T114(DRIVE_CURRENT_12_400_mA_T114) |
+			DRIVE_CURRENT_LANE3_T114(DRIVE_CURRENT_12_400_mA_T114),
+		.peak_current = PEAK_CURRENT_LANE0(PEAK_CURRENT_0_000_mA) |
+			PEAK_CURRENT_LANE1(PEAK_CURRENT_0_000_mA) |
+			PEAK_CURRENT_LANE2(PEAK_CURRENT_0_000_mA) |
+			PEAK_CURRENT_LANE3(PEAK_CURRENT_0_000_mA),
+	}, { /* 225/297MHz modes */
+		.pclk = UINT_MAX,
+		.pll0 = SOR_PLL_ICHPMP(1) | SOR_PLL_BG_V17_S(3) |
+			SOR_PLL_VCOCAP(0xf) | SOR_PLL_RESISTORSEL,
+		.pll1 = SOR_PLL_LOADADJ(3) | SOR_PLL_TMDS_TERMADJ(7)
+			| SOR_PLL_TMDS_TERM_ENABLE,
+		.pe_current = PE_CURRENT0(PE_CURRENT_0_mA_T114) |
+			PE_CURRENT1(PE_CURRENT_0_mA_T114) |
+			PE_CURRENT2(PE_CURRENT_0_mA_T114) |
+			PE_CURRENT3(PE_CURRENT_0_mA_T114),
+		.drive_current =
+			DRIVE_CURRENT_LANE0_T114(DRIVE_CURRENT_25_200_mA_T114) |
+			DRIVE_CURRENT_LANE1_T114(DRIVE_CURRENT_25_200_mA_T114) |
+			DRIVE_CURRENT_LANE2_T114(DRIVE_CURRENT_25_200_mA_T114) |
+			DRIVE_CURRENT_LANE3_T114(DRIVE_CURRENT_19_200_mA_T114),
+		.peak_current = PEAK_CURRENT_LANE0(PEAK_CURRENT_3_000_mA) |
+			PEAK_CURRENT_LANE1(PEAK_CURRENT_3_000_mA) |
+			PEAK_CURRENT_LANE2(PEAK_CURRENT_3_000_mA) |
+			PEAK_CURRENT_LANE3(PEAK_CURRENT_0_800_mA),
+	},
+};
+
 static const struct tegra_hdmi_audio_config *
 tegra_hdmi_get_audio_config(unsigned int audio_freq, unsigned int pclk)
 {
@@ -378,7 +460,7 @@ static void tegra_hdmi_setup_audio_fs_tables(struct tegra_hdmi *hdmi)
 
 		if (f > 96000)
 			delta = 2;
-		else if (f > 480000)
+		else if (f > 48000)
 			delta = 6;
 		else
 			delta = 9;
@@ -698,6 +780,9 @@ static int tegra_output_hdmi_enable(struct tegra_output *output)
 	int retries = 1000;
 	int err;
 
+	if (hdmi->enabled)
+		return 0;
+
 	hdmi->dvi = !tegra_output_is_hdmi(output);
 
 	pclk = mode->clock * 1000;
@@ -711,13 +796,9 @@ static int tegra_output_hdmi_enable(struct tegra_output *output)
 		return err;
 	}
 
-	/*
-	 * This assumes that the display controller will divide its parent
-	 * clock by 2 to generate the pixel clock.
-	 */
-	err = tegra_output_setup_clock(output, hdmi->clk, pclk * 2);
+	err = regulator_enable(hdmi->vdd);
 	if (err < 0) {
-		dev_err(hdmi->dev, "failed to setup clock: %d\n", err);
+		dev_err(hdmi->dev, "failed to enable VDD regulator: %d\n", err);
 		return err;
 	}
 
@@ -725,15 +806,26 @@ static int tegra_output_hdmi_enable(struct tegra_output *output)
 	if (err < 0)
 		return err;
 
-	err = clk_enable(hdmi->clk);
+	err = clk_prepare_enable(hdmi->clk);
 	if (err < 0) {
 		dev_err(hdmi->dev, "failed to enable clock: %d\n", err);
 		return err;
 	}
 
-	tegra_periph_reset_assert(hdmi->clk);
+	reset_control_assert(hdmi->rst);
 	usleep_range(1000, 2000);
-	tegra_periph_reset_deassert(hdmi->clk);
+	reset_control_deassert(hdmi->rst);
+
+	/* power up sequence */
+	value = tegra_hdmi_readl(hdmi, HDMI_NV_PDISP_SOR_PLL0);
+	value &= ~SOR_PLL_PDBG;
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_PLL0);
+
+	usleep_range(10, 20);
+
+	value = tegra_hdmi_readl(hdmi, HDMI_NV_PDISP_SOR_PLL0);
+	value &= ~SOR_PLL_PWR;
+	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_PLL0);
 
 	tegra_dc_writel(dc, VSYNC_H_POSITION(1),
 			DC_DISP_DISP_TIMING_OPTIONS);
@@ -833,15 +925,15 @@ static int tegra_output_hdmi_enable(struct tegra_output *output)
 	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_SEQ_INST(0));
 	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_SEQ_INST(8));
 
-	value = 0x1c800;
+	value = tegra_hdmi_readl(hdmi, HDMI_NV_PDISP_SOR_CSTM);
 	value &= ~SOR_CSTM_ROTCLK(~0);
 	value |= SOR_CSTM_ROTCLK(2);
+	value |= SOR_CSTM_PLLDIV;
+	value &= ~SOR_CSTM_LVDS_ENABLE;
+	value &= ~SOR_CSTM_MODE_MASK;
+	value |= SOR_CSTM_MODE_TMDS;
 	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_CSTM);
 
-	tegra_dc_writel(dc, DISP_CTRL_MODE_STOP, DC_CMD_DISPLAY_COMMAND);
-	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
-	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
-
 	/* start SOR */
 	tegra_hdmi_writel(hdmi,
 			  SOR_PWR_NORMAL_STATE_PU |
@@ -891,39 +983,84 @@ static int tegra_output_hdmi_enable(struct tegra_output *output)
 			  HDMI_NV_PDISP_SOR_STATE1);
 	tegra_hdmi_writel(hdmi, 0, HDMI_NV_PDISP_SOR_STATE0);
 
-	tegra_dc_writel(dc, HDMI_ENABLE, DC_DISP_DISP_WIN_OPTIONS);
-
-	value = PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
-		PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
-	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+	value |= HDMI_ENABLE;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
 
-	value = DISP_CTRL_MODE_C_DISPLAY;
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+	value &= ~DISP_CTRL_MODE_MASK;
+	value |= DISP_CTRL_MODE_C_DISPLAY;
 	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
 
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+	value |= PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		 PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
 	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
 	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
 
 	/* TODO: add HDCP support */
 
+	hdmi->enabled = true;
+
 	return 0;
 }
 
 static int tegra_output_hdmi_disable(struct tegra_output *output)
 {
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
 	struct tegra_hdmi *hdmi = to_hdmi(output);
+	unsigned long value;
+
+	if (!hdmi->enabled)
+		return 0;
+
+	/*
+	 * The following accesses registers of the display controller, so make
+	 * sure it's only executed when the output is attached to one.
+	 */
+	if (dc) {
+		/*
+		 * XXX: We can't do this here because it causes HDMI to go
+		 * into an erroneous state with the result that HDMI won't
+		 * properly work once disabled. See also a similar symptom
+		 * for the SOR output.
+		 */
+		/*
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+		value &= ~(PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+			   PW4_ENABLE | PM0_ENABLE | PM1_ENABLE);
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+		*/
+
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+		value &= ~DISP_CTRL_MODE_MASK;
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
+
+		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+		value &= ~HDMI_ENABLE;
+		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+		tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+		tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+	}
 
-	tegra_periph_reset_assert(hdmi->clk);
-	clk_disable(hdmi->clk);
+	clk_disable_unprepare(hdmi->clk);
+	reset_control_assert(hdmi->rst);
+	regulator_disable(hdmi->vdd);
 	regulator_disable(hdmi->pll);
 
+	hdmi->enabled = false;
+
 	return 0;
 }
 
 static int tegra_output_hdmi_setup_clock(struct tegra_output *output,
-					 struct clk *clk, unsigned long pclk)
+					 struct clk *clk, unsigned long pclk,
+					 unsigned int *div)
 {
 	struct tegra_hdmi *hdmi = to_hdmi(output);
-	struct clk *base;
 	int err;
 
 	err = clk_set_parent(clk, hdmi->clk_parent);
@@ -932,17 +1069,12 @@ static int tegra_output_hdmi_setup_clock(struct tegra_output *output,
 		return err;
 	}
 
-	base = clk_get_parent(hdmi->clk_parent);
-
-	/*
-	 * This assumes that the parent clock is pll_d_out0 or pll_d2_out
-	 * respectively, each of which divides the base pll_d by 2.
-	 */
-	err = clk_set_rate(base, pclk * 2);
+	err = clk_set_rate(hdmi->clk_parent, pclk);
 	if (err < 0)
-		dev_err(output->dev,
-			"failed to set base clock rate to %lu Hz\n",
-			pclk * 2);
+		dev_err(output->dev, "failed to set clock rate to %lu Hz\n",
+			pclk);
+
+	*div = 0;
 
 	return 0;
 }
@@ -959,7 +1091,7 @@ static int tegra_output_hdmi_check_mode(struct tegra_output *output,
 	parent = clk_get_parent(hdmi->clk_parent);
 
 	err = clk_round_rate(parent, pclk * 4);
-	if (err < 0)
+	if (err <= 0)
 		*status = MODE_NOCLOCK;
 	else
 		*status = MODE_OK;
@@ -980,7 +1112,7 @@ static int tegra_hdmi_show_regs(struct seq_file *s, void *data)
 	struct tegra_hdmi *hdmi = node->info_ent->data;
 	int err;
 
-	err = clk_enable(hdmi->clk);
+	err = clk_prepare_enable(hdmi->clk);
 	if (err)
 		return err;
 
@@ -1149,7 +1281,7 @@ static int tegra_hdmi_show_regs(struct seq_file *s, void *data)
 
 #undef DUMP_REG
 
-	clk_disable(hdmi->clk);
+	clk_disable_unprepare(hdmi->clk);
 
 	return 0;
 }
@@ -1215,33 +1347,33 @@ static int tegra_hdmi_debugfs_exit(struct tegra_hdmi *hdmi)
 
 static int tegra_hdmi_init(struct host1x_client *client)
 {
-	struct tegra_drm *tegra = dev_get_drvdata(client->parent);
+	struct drm_device *drm = dev_get_drvdata(client->parent);
 	struct tegra_hdmi *hdmi = host1x_client_to_hdmi(client);
 	int err;
 
-	err = regulator_enable(hdmi->vdd);
-	if (err < 0) {
-		dev_err(client->dev, "failed to enable VDD regulator: %d\n",
-			err);
-		return err;
-	}
-
 	hdmi->output.type = TEGRA_OUTPUT_HDMI;
 	hdmi->output.dev = client->dev;
 	hdmi->output.ops = &hdmi_ops;
 
-	err = tegra_output_init(tegra->drm, &hdmi->output);
+	err = tegra_output_init(drm, &hdmi->output);
 	if (err < 0) {
 		dev_err(client->dev, "output setup failed: %d\n", err);
 		return err;
 	}
 
 	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
-		err = tegra_hdmi_debugfs_init(hdmi, tegra->drm->primary);
+		err = tegra_hdmi_debugfs_init(hdmi, drm->primary);
 		if (err < 0)
 			dev_err(client->dev, "debugfs setup failed: %d\n", err);
 	}
 
+	err = regulator_enable(hdmi->hdmi);
+	if (err < 0) {
+		dev_err(client->dev, "failed to enable HDMI regulator: %d\n",
+			err);
+		return err;
+	}
+
 	return 0;
 }
 
@@ -1250,6 +1382,8 @@ static int tegra_hdmi_exit(struct host1x_client *client)
 	struct tegra_hdmi *hdmi = host1x_client_to_hdmi(client);
 	int err;
 
+	regulator_disable(hdmi->hdmi);
+
 	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
 		err = tegra_hdmi_debugfs_exit(hdmi);
 		if (err < 0)
@@ -1269,8 +1403,6 @@ static int tegra_hdmi_exit(struct host1x_client *client)
 		return err;
 	}
 
-	regulator_disable(hdmi->vdd);
-
 	return 0;
 }
 
@@ -1303,7 +1435,16 @@ static const struct tegra_hdmi_config tegra114_hdmi_config = {
 	.has_sor_io_peak_current = true,
 };
 
+static const struct tegra_hdmi_config tegra124_hdmi_config = {
+	.tmds = tegra124_tmds_config,
+	.num_tmds = ARRAY_SIZE(tegra124_tmds_config),
+	.fuse_override_offset = HDMI_NV_PDISP_SOR_PAD_CTLS0,
+	.fuse_override_value = 1 << 31,
+	.has_sor_io_peak_current = true,
+};
+
 static const struct of_device_id tegra_hdmi_of_match[] = {
+	{ .compatible = "nvidia,tegra124-hdmi", .data = &tegra124_hdmi_config },
 	{ .compatible = "nvidia,tegra114-hdmi", .data = &tegra114_hdmi_config },
 	{ .compatible = "nvidia,tegra30-hdmi", .data = &tegra30_hdmi_config },
 	{ .compatible = "nvidia,tegra20-hdmi", .data = &tegra20_hdmi_config },
@@ -1338,28 +1479,26 @@ static int tegra_hdmi_probe(struct platform_device *pdev)
 		return PTR_ERR(hdmi->clk);
 	}
 
-	err = clk_prepare(hdmi->clk);
-	if (err < 0)
-		return err;
+	hdmi->rst = devm_reset_control_get(&pdev->dev, "hdmi");
+	if (IS_ERR(hdmi->rst)) {
+		dev_err(&pdev->dev, "failed to get reset\n");
+		return PTR_ERR(hdmi->rst);
+	}
 
 	hdmi->clk_parent = devm_clk_get(&pdev->dev, "parent");
 	if (IS_ERR(hdmi->clk_parent))
 		return PTR_ERR(hdmi->clk_parent);
 
-	err = clk_prepare(hdmi->clk_parent);
-	if (err < 0)
-		return err;
-
 	err = clk_set_parent(hdmi->clk, hdmi->clk_parent);
 	if (err < 0) {
 		dev_err(&pdev->dev, "failed to setup clocks: %d\n", err);
 		return err;
 	}
 
-	hdmi->vdd = devm_regulator_get(&pdev->dev, "vdd");
-	if (IS_ERR(hdmi->vdd)) {
-		dev_err(&pdev->dev, "failed to get VDD regulator\n");
-		return PTR_ERR(hdmi->vdd);
+	hdmi->hdmi = devm_regulator_get(&pdev->dev, "hdmi");
+	if (IS_ERR(hdmi->hdmi)) {
+		dev_err(&pdev->dev, "failed to get HDMI regulator\n");
+		return PTR_ERR(hdmi->hdmi);
 	}
 
 	hdmi->pll = devm_regulator_get(&pdev->dev, "pll");
@@ -1368,6 +1507,12 @@ static int tegra_hdmi_probe(struct platform_device *pdev)
 		return PTR_ERR(hdmi->pll);
 	}
 
+	hdmi->vdd = devm_regulator_get(&pdev->dev, "vdd");
+	if (IS_ERR(hdmi->vdd)) {
+		dev_err(&pdev->dev, "failed to get VDD regulator\n");
+		return PTR_ERR(hdmi->vdd);
+	}
+
 	hdmi->output.dev = &pdev->dev;
 
 	err = tegra_output_probe(&hdmi->output);
@@ -1375,9 +1520,6 @@ static int tegra_hdmi_probe(struct platform_device *pdev)
 		return err;
 
 	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (!regs)
-		return -ENXIO;
-
 	hdmi->regs = devm_ioremap_resource(&pdev->dev, regs);
 	if (IS_ERR(hdmi->regs))
 		return PTR_ERR(hdmi->regs);
@@ -1422,8 +1564,8 @@ static int tegra_hdmi_remove(struct platform_device *pdev)
 		return err;
 	}
 
-	clk_unprepare(hdmi->clk_parent);
-	clk_unprepare(hdmi->clk);
+	clk_disable_unprepare(hdmi->clk_parent);
+	clk_disable_unprepare(hdmi->clk);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/tegra/hdmi.h b/drivers/gpu/drm/tegra/hdmi.h
index 0aebc485f7f..919a19df4e1 100644
--- a/drivers/gpu/drm/tegra/hdmi.h
+++ b/drivers/gpu/drm/tegra/hdmi.h
@@ -190,6 +190,11 @@
 
 #define HDMI_NV_PDISP_SOR_CSTM					0x5a
 #define SOR_CSTM_ROTCLK(x) (((x) & 0xf) << 24)
+#define SOR_CSTM_PLLDIV (1 << 21)
+#define SOR_CSTM_LVDS_ENABLE (1 << 16)
+#define SOR_CSTM_MODE_LVDS (0 << 12)
+#define SOR_CSTM_MODE_TMDS (1 << 12)
+#define SOR_CSTM_MODE_MASK (3 << 12)
 
 #define HDMI_NV_PDISP_SOR_LVDS					0x5b
 #define HDMI_NV_PDISP_SOR_CRCA					0x5c
diff --git a/drivers/gpu/drm/tegra/mipi-phy.c b/drivers/gpu/drm/tegra/mipi-phy.c
new file mode 100644
index 00000000000..486d19d589c
--- /dev/null
+++ b/drivers/gpu/drm/tegra/mipi-phy.c
@@ -0,0 +1,124 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/errno.h>
+#include <linux/kernel.h>
+
+#include "mipi-phy.h"
+
+/*
+ * Default D-PHY timings based on MIPI D-PHY specification. Derived from
+ * the valid ranges specified in Section 5.9 of the D-PHY specification
+ * with minor adjustments.
+ */
+int mipi_dphy_timing_get_default(struct mipi_dphy_timing *timing,
+				 unsigned long period)
+{
+	timing->clkmiss = 0;
+	timing->clkpost = 70 + 52 * period;
+	timing->clkpre = 8;
+	timing->clkprepare = 65;
+	timing->clksettle = 95;
+	timing->clktermen = 0;
+	timing->clktrail = 80;
+	timing->clkzero = 260;
+	timing->dtermen = 0;
+	timing->eot = 0;
+	timing->hsexit = 120;
+	timing->hsprepare = 65 + 5 * period;
+	timing->hszero = 145 + 5 * period;
+	timing->hssettle = 85 + 6 * period;
+	timing->hsskip = 40;
+	timing->hstrail = max(8 * period, 60 + 4 * period);
+	timing->init = 100000;
+	timing->lpx = 60;
+	timing->taget = 5 * timing->lpx;
+	timing->tago = 4 * timing->lpx;
+	timing->tasure = 2 * timing->lpx;
+	timing->wakeup = 1000000;
+
+	return 0;
+}
+
+/*
+ * Validate D-PHY timing according to MIPI Alliance Specification for D-PHY,
+ * Section 5.9 "Global Operation Timing Parameters".
+ */
+int mipi_dphy_timing_validate(struct mipi_dphy_timing *timing,
+			      unsigned long period)
+{
+	if (timing->clkmiss > 60)
+		return -EINVAL;
+
+	if (timing->clkpost < (60 + 52 * period))
+		return -EINVAL;
+
+	if (timing->clkpre < 8)
+		return -EINVAL;
+
+	if (timing->clkprepare < 38 || timing->clkprepare > 95)
+		return -EINVAL;
+
+	if (timing->clksettle < 95 || timing->clksettle > 300)
+		return -EINVAL;
+
+	if (timing->clktermen > 38)
+		return -EINVAL;
+
+	if (timing->clktrail < 60)
+		return -EINVAL;
+
+	if (timing->clkprepare + timing->clkzero < 300)
+		return -EINVAL;
+
+	if (timing->dtermen > 35 + 4 * period)
+		return -EINVAL;
+
+	if (timing->eot > 105 + 12 * period)
+		return -EINVAL;
+
+	if (timing->hsexit < 100)
+		return -EINVAL;
+
+	if (timing->hsprepare < 40 + 4 * period ||
+	    timing->hsprepare > 85 + 6 * period)
+		return -EINVAL;
+
+	if (timing->hsprepare + timing->hszero < 145 + 10 * period)
+		return -EINVAL;
+
+	if ((timing->hssettle < 85 + 6 * period) ||
+	    (timing->hssettle > 145 + 10 * period))
+		return -EINVAL;
+
+	if (timing->hsskip < 40 || timing->hsskip > 55 + 4 * period)
+		return -EINVAL;
+
+	if (timing->hstrail < max(8 * period, 60 + 4 * period))
+		return -EINVAL;
+
+	if (timing->init < 100000)
+		return -EINVAL;
+
+	if (timing->lpx < 50)
+		return -EINVAL;
+
+	if (timing->taget != 5 * timing->lpx)
+		return -EINVAL;
+
+	if (timing->tago != 4 * timing->lpx)
+		return -EINVAL;
+
+	if (timing->tasure < timing->lpx || timing->tasure > 2 * timing->lpx)
+		return -EINVAL;
+
+	if (timing->wakeup < 1000000)
+		return -EINVAL;
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/tegra/mipi-phy.h b/drivers/gpu/drm/tegra/mipi-phy.h
new file mode 100644
index 00000000000..012ea8ac36d
--- /dev/null
+++ b/drivers/gpu/drm/tegra/mipi-phy.h
@@ -0,0 +1,51 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef DRM_TEGRA_MIPI_PHY_H
+#define DRM_TEGRA_MIPI_PHY_H
+
+/*
+ * D-PHY timing parameters
+ *
+ * A detailed description of these parameters can be found in the  MIPI
+ * Alliance Specification for D-PHY, Section 5.9 "Global Operation Timing
+ * Parameters".
+ *
+ * All parameters are specified in nanoseconds.
+ */
+struct mipi_dphy_timing {
+	unsigned int clkmiss;
+	unsigned int clkpost;
+	unsigned int clkpre;
+	unsigned int clkprepare;
+	unsigned int clksettle;
+	unsigned int clktermen;
+	unsigned int clktrail;
+	unsigned int clkzero;
+	unsigned int dtermen;
+	unsigned int eot;
+	unsigned int hsexit;
+	unsigned int hsprepare;
+	unsigned int hszero;
+	unsigned int hssettle;
+	unsigned int hsskip;
+	unsigned int hstrail;
+	unsigned int init;
+	unsigned int lpx;
+	unsigned int taget;
+	unsigned int tago;
+	unsigned int tasure;
+	unsigned int wakeup;
+};
+
+int mipi_dphy_timing_get_default(struct mipi_dphy_timing *timing,
+				 unsigned long period);
+int mipi_dphy_timing_validate(struct mipi_dphy_timing *timing,
+			      unsigned long period);
+
+#endif
diff --git a/drivers/gpu/drm/tegra/output.c b/drivers/gpu/drm/tegra/output.c
index 2cb0065e057..a3e4f1eca6f 100644
--- a/drivers/gpu/drm/tegra/output.c
+++ b/drivers/gpu/drm/tegra/output.c
@@ -9,6 +9,7 @@
 
 #include <linux/of_gpio.h>
 
+#include <drm/drm_panel.h>
 #include "drm.h"
 
 static int tegra_connector_get_modes(struct drm_connector *connector)
@@ -17,6 +18,16 @@ static int tegra_connector_get_modes(struct drm_connector *connector)
 	struct edid *edid = NULL;
 	int err = 0;
 
+	/*
+	 * If the panel provides one or more modes, use them exclusively and
+	 * ignore any other means of obtaining a mode.
+	 */
+	if (output->panel) {
+		err = output->panel->funcs->get_modes(output->panel);
+		if (err > 0)
+			return err;
+	}
+
 	if (output->edid)
 		edid = kmemdup(output->edid, sizeof(*edid), GFP_KERNEL);
 	else if (output->ddc)
@@ -66,12 +77,20 @@ tegra_connector_detect(struct drm_connector *connector, bool force)
 	struct tegra_output *output = connector_to_output(connector);
 	enum drm_connector_status status = connector_status_unknown;
 
+	if (output->ops->detect)
+		return output->ops->detect(output);
+
 	if (gpio_is_valid(output->hpd_gpio)) {
 		if (gpio_get_value(output->hpd_gpio) == 0)
 			status = connector_status_disconnected;
 		else
 			status = connector_status_connected;
 	} else {
+		if (!output->panel)
+			status = connector_status_disconnected;
+		else
+			status = connector_status_connected;
+
 		if (connector->connector_type == DRM_MODE_CONNECTOR_LVDS)
 			status = connector_status_connected;
 	}
@@ -115,6 +134,16 @@ static const struct drm_encoder_funcs encoder_funcs = {
 
 static void tegra_encoder_dpms(struct drm_encoder *encoder, int mode)
 {
+	struct tegra_output *output = encoder_to_output(encoder);
+	struct drm_panel *panel = output->panel;
+
+	if (mode != DRM_MODE_DPMS_ON) {
+		drm_panel_disable(panel);
+		tegra_output_disable(output);
+	} else {
+		tegra_output_enable(output);
+		drm_panel_enable(panel);
+	}
 }
 
 static bool tegra_encoder_mode_fixup(struct drm_encoder *encoder,
@@ -163,14 +192,22 @@ static irqreturn_t hpd_irq(int irq, void *data)
 
 int tegra_output_probe(struct tegra_output *output)
 {
+	struct device_node *ddc, *panel;
 	enum of_gpio_flags flags;
-	struct device_node *ddc;
-	size_t size;
-	int err;
+	int err, size;
 
 	if (!output->of_node)
 		output->of_node = output->dev->of_node;
 
+	panel = of_parse_phandle(output->of_node, "nvidia,panel", 0);
+	if (panel) {
+		output->panel = of_drm_find_panel(panel);
+		if (!output->panel)
+			return -EPROBE_DEFER;
+
+		of_node_put(panel);
+	}
+
 	output->edid = of_get_property(output->of_node, "nvidia,edid", &size);
 
 	ddc = of_parse_phandle(output->of_node, "nvidia,ddc-i2c-bus", 0);
@@ -185,9 +222,6 @@ int tegra_output_probe(struct tegra_output *output)
 		of_node_put(ddc);
 	}
 
-	if (!output->edid && !output->ddc)
-		return -ENODEV;
-
 	output->hpd_gpio = of_get_named_gpio_flags(output->of_node,
 						   "nvidia,hpd-gpio", 0,
 						   &flags);
@@ -256,6 +290,16 @@ int tegra_output_init(struct drm_device *drm, struct tegra_output *output)
 		encoder = DRM_MODE_ENCODER_TMDS;
 		break;
 
+	case TEGRA_OUTPUT_DSI:
+		connector = DRM_MODE_CONNECTOR_DSI;
+		encoder = DRM_MODE_ENCODER_DSI;
+		break;
+
+	case TEGRA_OUTPUT_EDP:
+		connector = DRM_MODE_CONNECTOR_eDP;
+		encoder = DRM_MODE_ENCODER_TMDS;
+		break;
+
 	default:
 		connector = DRM_MODE_CONNECTOR_Unknown;
 		encoder = DRM_MODE_ENCODER_NONE;
@@ -267,6 +311,9 @@ int tegra_output_init(struct drm_device *drm, struct tegra_output *output)
 	drm_connector_helper_add(&output->connector, &connector_helper_funcs);
 	output->connector.dpms = DRM_MODE_DPMS_OFF;
 
+	if (output->panel)
+		drm_panel_attach(output->panel, &output->connector);
+
 	drm_encoder_init(drm, &output->encoder, &encoder_funcs, encoder);
 	drm_encoder_helper_add(&output->encoder, &encoder_helper_funcs);
 
diff --git a/drivers/gpu/drm/tegra/rgb.c b/drivers/gpu/drm/tegra/rgb.c
index 3b29018913a..d6af9be48f4 100644
--- a/drivers/gpu/drm/tegra/rgb.c
+++ b/drivers/gpu/drm/tegra/rgb.c
@@ -15,6 +15,7 @@
 struct tegra_rgb {
 	struct tegra_output output;
 	struct tegra_dc *dc;
+	bool enabled;
 
 	struct clk *clk_parent;
 	struct clk *clk;
@@ -87,27 +88,109 @@ static void tegra_dc_write_regs(struct tegra_dc *dc,
 static int tegra_output_rgb_enable(struct tegra_output *output)
 {
 	struct tegra_rgb *rgb = to_rgb(output);
+	unsigned long value;
+
+	if (rgb->enabled)
+		return 0;
 
 	tegra_dc_write_regs(rgb->dc, rgb_enable, ARRAY_SIZE(rgb_enable));
 
+	value = DE_SELECT_ACTIVE | DE_CONTROL_NORMAL;
+	tegra_dc_writel(rgb->dc, value, DC_DISP_DATA_ENABLE_OPTIONS);
+
+	/* XXX: parameterize? */
+	value = tegra_dc_readl(rgb->dc, DC_COM_PIN_OUTPUT_POLARITY(1));
+	value &= ~LVS_OUTPUT_POLARITY_LOW;
+	value &= ~LHS_OUTPUT_POLARITY_LOW;
+	tegra_dc_writel(rgb->dc, value, DC_COM_PIN_OUTPUT_POLARITY(1));
+
+	/* XXX: parameterize? */
+	value = DISP_DATA_FORMAT_DF1P1C | DISP_ALIGNMENT_MSB |
+		DISP_ORDER_RED_BLUE;
+	tegra_dc_writel(rgb->dc, value, DC_DISP_DISP_INTERFACE_CONTROL);
+
+	/* XXX: parameterize? */
+	value = SC0_H_QUALIFIER_NONE | SC1_H_QUALIFIER_NONE;
+	tegra_dc_writel(rgb->dc, value, DC_DISP_SHIFT_CLOCK_OPTIONS);
+
+	value = tegra_dc_readl(rgb->dc, DC_CMD_DISPLAY_COMMAND);
+	value &= ~DISP_CTRL_MODE_MASK;
+	value |= DISP_CTRL_MODE_C_DISPLAY;
+	tegra_dc_writel(rgb->dc, value, DC_CMD_DISPLAY_COMMAND);
+
+	value = tegra_dc_readl(rgb->dc, DC_CMD_DISPLAY_POWER_CONTROL);
+	value |= PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		 PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
+	tegra_dc_writel(rgb->dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+	tegra_dc_writel(rgb->dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(rgb->dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	rgb->enabled = true;
+
 	return 0;
 }
 
 static int tegra_output_rgb_disable(struct tegra_output *output)
 {
 	struct tegra_rgb *rgb = to_rgb(output);
+	unsigned long value;
+
+	if (!rgb->enabled)
+		return 0;
+
+	value = tegra_dc_readl(rgb->dc, DC_CMD_DISPLAY_POWER_CONTROL);
+	value &= ~(PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		   PW4_ENABLE | PM0_ENABLE | PM1_ENABLE);
+	tegra_dc_writel(rgb->dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+	value = tegra_dc_readl(rgb->dc, DC_CMD_DISPLAY_COMMAND);
+	value &= ~DISP_CTRL_MODE_MASK;
+	tegra_dc_writel(rgb->dc, value, DC_CMD_DISPLAY_COMMAND);
+
+	tegra_dc_writel(rgb->dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(rgb->dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
 
 	tegra_dc_write_regs(rgb->dc, rgb_disable, ARRAY_SIZE(rgb_disable));
 
+	rgb->enabled = false;
+
 	return 0;
 }
 
 static int tegra_output_rgb_setup_clock(struct tegra_output *output,
-					struct clk *clk, unsigned long pclk)
+					struct clk *clk, unsigned long pclk,
+					unsigned int *div)
 {
 	struct tegra_rgb *rgb = to_rgb(output);
+	int err;
+
+	err = clk_set_parent(clk, rgb->clk_parent);
+	if (err < 0) {
+		dev_err(output->dev, "failed to set parent: %d\n", err);
+		return err;
+	}
 
-	return clk_set_parent(clk, rgb->clk_parent);
+	/*
+	 * We may not want to change the frequency of the parent clock, since
+	 * it may be a parent for other peripherals. This is due to the fact
+	 * that on Tegra20 there's only a single clock dedicated to display
+	 * (pll_d_out0), whereas later generations have a second one that can
+	 * be used to independently drive a second output (pll_d2_out0).
+	 *
+	 * As a way to support multiple outputs on Tegra20 as well, pll_p is
+	 * typically used as the parent clock for the display controllers.
+	 * But this comes at a cost: pll_p is the parent of several other
+	 * peripherals, so its frequency shouldn't change out of the blue.
+	 *
+	 * The best we can do at this point is to use the shift clock divider
+	 * and hope that the desired frequency can be matched (or at least
+	 * matched sufficiently close that the panel will still work).
+	 */
+
+	*div = ((clk_get_rate(clk) * 2) / pclk) - 2;
+
+	return 0;
 }
 
 static int tegra_output_rgb_check_mode(struct tegra_output *output,
@@ -213,7 +296,7 @@ int tegra_dc_rgb_init(struct drm_device *drm, struct tegra_dc *dc)
 	 * RGB outputs are an exception, so we make sure they can be attached
 	 * to only their parent display controller.
 	 */
-	rgb->output.encoder.possible_crtcs = 1 << dc->pipe;
+	rgb->output.encoder.possible_crtcs = drm_crtc_mask(&dc->base);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/tegra/sor.c b/drivers/gpu/drm/tegra/sor.c
new file mode 100644
index 00000000000..27c979b5011
--- /dev/null
+++ b/drivers/gpu/drm/tegra/sor.c
@@ -0,0 +1,1466 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/debugfs.h>
+#include <linux/io.h>
+#include <linux/platform_device.h>
+#include <linux/reset.h>
+#include <linux/tegra-powergate.h>
+
+#include <drm/drm_dp_helper.h>
+
+#include "dc.h"
+#include "drm.h"
+#include "sor.h"
+
+struct tegra_sor {
+	struct host1x_client client;
+	struct tegra_output output;
+	struct device *dev;
+
+	void __iomem *regs;
+
+	struct reset_control *rst;
+	struct clk *clk_parent;
+	struct clk *clk_safe;
+	struct clk *clk_dp;
+	struct clk *clk;
+
+	struct tegra_dpaux *dpaux;
+
+	struct mutex lock;
+	bool enabled;
+
+	struct dentry *debugfs;
+};
+
+struct tegra_sor_config {
+	u32 bits_per_pixel;
+
+	u32 active_polarity;
+	u32 active_count;
+	u32 tu_size;
+	u32 active_frac;
+	u32 watermark;
+
+	u32 hblank_symbols;
+	u32 vblank_symbols;
+};
+
+static inline struct tegra_sor *
+host1x_client_to_sor(struct host1x_client *client)
+{
+	return container_of(client, struct tegra_sor, client);
+}
+
+static inline struct tegra_sor *to_sor(struct tegra_output *output)
+{
+	return container_of(output, struct tegra_sor, output);
+}
+
+static inline unsigned long tegra_sor_readl(struct tegra_sor *sor,
+					    unsigned long offset)
+{
+	return readl(sor->regs + (offset << 2));
+}
+
+static inline void tegra_sor_writel(struct tegra_sor *sor, unsigned long value,
+				    unsigned long offset)
+{
+	writel(value, sor->regs + (offset << 2));
+}
+
+static int tegra_sor_dp_train_fast(struct tegra_sor *sor,
+				   struct drm_dp_link *link)
+{
+	unsigned long value;
+	unsigned int i;
+	u8 pattern;
+	int err;
+
+	/* setup lane parameters */
+	value = SOR_LANE_DRIVE_CURRENT_LANE3(0x40) |
+		SOR_LANE_DRIVE_CURRENT_LANE2(0x40) |
+		SOR_LANE_DRIVE_CURRENT_LANE1(0x40) |
+		SOR_LANE_DRIVE_CURRENT_LANE0(0x40);
+	tegra_sor_writel(sor, value, SOR_LANE_DRIVE_CURRENT_0);
+
+	value = SOR_LANE_PREEMPHASIS_LANE3(0x0f) |
+		SOR_LANE_PREEMPHASIS_LANE2(0x0f) |
+		SOR_LANE_PREEMPHASIS_LANE1(0x0f) |
+		SOR_LANE_PREEMPHASIS_LANE0(0x0f);
+	tegra_sor_writel(sor, value, SOR_LANE_PREEMPHASIS_0);
+
+	value = SOR_LANE_POST_CURSOR_LANE3(0x00) |
+		SOR_LANE_POST_CURSOR_LANE2(0x00) |
+		SOR_LANE_POST_CURSOR_LANE1(0x00) |
+		SOR_LANE_POST_CURSOR_LANE0(0x00);
+	tegra_sor_writel(sor, value, SOR_LANE_POST_CURSOR_0);
+
+	/* disable LVDS mode */
+	tegra_sor_writel(sor, 0, SOR_LVDS);
+
+	value = tegra_sor_readl(sor, SOR_DP_PADCTL_0);
+	value |= SOR_DP_PADCTL_TX_PU_ENABLE;
+	value &= ~SOR_DP_PADCTL_TX_PU_MASK;
+	value |= SOR_DP_PADCTL_TX_PU(2); /* XXX: don't hardcode? */
+	tegra_sor_writel(sor, value, SOR_DP_PADCTL_0);
+
+	value = tegra_sor_readl(sor, SOR_DP_PADCTL_0);
+	value |= SOR_DP_PADCTL_CM_TXD_3 | SOR_DP_PADCTL_CM_TXD_2 |
+		 SOR_DP_PADCTL_CM_TXD_1 | SOR_DP_PADCTL_CM_TXD_0;
+	tegra_sor_writel(sor, value, SOR_DP_PADCTL_0);
+
+	usleep_range(10, 100);
+
+	value = tegra_sor_readl(sor, SOR_DP_PADCTL_0);
+	value &= ~(SOR_DP_PADCTL_CM_TXD_3 | SOR_DP_PADCTL_CM_TXD_2 |
+		   SOR_DP_PADCTL_CM_TXD_1 | SOR_DP_PADCTL_CM_TXD_0);
+	tegra_sor_writel(sor, value, SOR_DP_PADCTL_0);
+
+	err = tegra_dpaux_prepare(sor->dpaux, DP_SET_ANSI_8B10B);
+	if (err < 0)
+		return err;
+
+	for (i = 0, value = 0; i < link->num_lanes; i++) {
+		unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
+				     SOR_DP_TPG_SCRAMBLER_NONE |
+				     SOR_DP_TPG_PATTERN_TRAIN1;
+		value = (value << 8) | lane;
+	}
+
+	tegra_sor_writel(sor, value, SOR_DP_TPG);
+
+	pattern = DP_TRAINING_PATTERN_1;
+
+	err = tegra_dpaux_train(sor->dpaux, link, pattern);
+	if (err < 0)
+		return err;
+
+	value = tegra_sor_readl(sor, SOR_DP_SPARE_0);
+	value |= SOR_DP_SPARE_SEQ_ENABLE;
+	value &= ~SOR_DP_SPARE_PANEL_INTERNAL;
+	value |= SOR_DP_SPARE_MACRO_SOR_CLK;
+	tegra_sor_writel(sor, value, SOR_DP_SPARE_0);
+
+	for (i = 0, value = 0; i < link->num_lanes; i++) {
+		unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
+				     SOR_DP_TPG_SCRAMBLER_NONE |
+				     SOR_DP_TPG_PATTERN_TRAIN2;
+		value = (value << 8) | lane;
+	}
+
+	tegra_sor_writel(sor, value, SOR_DP_TPG);
+
+	pattern = DP_LINK_SCRAMBLING_DISABLE | DP_TRAINING_PATTERN_2;
+
+	err = tegra_dpaux_train(sor->dpaux, link, pattern);
+	if (err < 0)
+		return err;
+
+	for (i = 0, value = 0; i < link->num_lanes; i++) {
+		unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
+				     SOR_DP_TPG_SCRAMBLER_GALIOS |
+				     SOR_DP_TPG_PATTERN_NONE;
+		value = (value << 8) | lane;
+	}
+
+	tegra_sor_writel(sor, value, SOR_DP_TPG);
+
+	pattern = DP_TRAINING_PATTERN_DISABLE;
+
+	err = tegra_dpaux_train(sor->dpaux, link, pattern);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+
+static void tegra_sor_super_update(struct tegra_sor *sor)
+{
+	tegra_sor_writel(sor, 0, SOR_SUPER_STATE_0);
+	tegra_sor_writel(sor, 1, SOR_SUPER_STATE_0);
+	tegra_sor_writel(sor, 0, SOR_SUPER_STATE_0);
+}
+
+static void tegra_sor_update(struct tegra_sor *sor)
+{
+	tegra_sor_writel(sor, 0, SOR_STATE_0);
+	tegra_sor_writel(sor, 1, SOR_STATE_0);
+	tegra_sor_writel(sor, 0, SOR_STATE_0);
+}
+
+static int tegra_sor_setup_pwm(struct tegra_sor *sor, unsigned long timeout)
+{
+	unsigned long value;
+
+	value = tegra_sor_readl(sor, SOR_PWM_DIV);
+	value &= ~SOR_PWM_DIV_MASK;
+	value |= 0x400; /* period */
+	tegra_sor_writel(sor, value, SOR_PWM_DIV);
+
+	value = tegra_sor_readl(sor, SOR_PWM_CTL);
+	value &= ~SOR_PWM_CTL_DUTY_CYCLE_MASK;
+	value |= 0x400; /* duty cycle */
+	value &= ~SOR_PWM_CTL_CLK_SEL; /* clock source: PCLK */
+	value |= SOR_PWM_CTL_TRIGGER;
+	tegra_sor_writel(sor, value, SOR_PWM_CTL);
+
+	timeout = jiffies + msecs_to_jiffies(timeout);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_PWM_CTL);
+		if ((value & SOR_PWM_CTL_TRIGGER) == 0)
+			return 0;
+
+		usleep_range(25, 100);
+	}
+
+	return -ETIMEDOUT;
+}
+
+static int tegra_sor_attach(struct tegra_sor *sor)
+{
+	unsigned long value, timeout;
+
+	/* wake up in normal mode */
+	value = tegra_sor_readl(sor, SOR_SUPER_STATE_1);
+	value |= SOR_SUPER_STATE_HEAD_MODE_AWAKE;
+	value |= SOR_SUPER_STATE_MODE_NORMAL;
+	tegra_sor_writel(sor, value, SOR_SUPER_STATE_1);
+	tegra_sor_super_update(sor);
+
+	/* attach */
+	value = tegra_sor_readl(sor, SOR_SUPER_STATE_1);
+	value |= SOR_SUPER_STATE_ATTACHED;
+	tegra_sor_writel(sor, value, SOR_SUPER_STATE_1);
+	tegra_sor_super_update(sor);
+
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_TEST);
+		if ((value & SOR_TEST_ATTACHED) != 0)
+			return 0;
+
+		usleep_range(25, 100);
+	}
+
+	return -ETIMEDOUT;
+}
+
+static int tegra_sor_wakeup(struct tegra_sor *sor)
+{
+	struct tegra_dc *dc = to_tegra_dc(sor->output.encoder.crtc);
+	unsigned long value, timeout;
+
+	/* enable display controller outputs */
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+	value |= PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		 PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	/* wait for head to wake up */
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_TEST);
+		value &= SOR_TEST_HEAD_MODE_MASK;
+
+		if (value == SOR_TEST_HEAD_MODE_AWAKE)
+			return 0;
+
+		usleep_range(25, 100);
+	}
+
+	return -ETIMEDOUT;
+}
+
+static int tegra_sor_power_up(struct tegra_sor *sor, unsigned long timeout)
+{
+	unsigned long value;
+
+	value = tegra_sor_readl(sor, SOR_PWR);
+	value |= SOR_PWR_TRIGGER | SOR_PWR_NORMAL_STATE_PU;
+	tegra_sor_writel(sor, value, SOR_PWR);
+
+	timeout = jiffies + msecs_to_jiffies(timeout);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_PWR);
+		if ((value & SOR_PWR_TRIGGER) == 0)
+			return 0;
+
+		usleep_range(25, 100);
+	}
+
+	return -ETIMEDOUT;
+}
+
+struct tegra_sor_params {
+	/* number of link clocks per line */
+	unsigned int num_clocks;
+	/* ratio between input and output */
+	u64 ratio;
+	/* precision factor */
+	u64 precision;
+
+	unsigned int active_polarity;
+	unsigned int active_count;
+	unsigned int active_frac;
+	unsigned int tu_size;
+	unsigned int error;
+};
+
+static int tegra_sor_compute_params(struct tegra_sor *sor,
+				    struct tegra_sor_params *params,
+				    unsigned int tu_size)
+{
+	u64 active_sym, active_count, frac, approx;
+	u32 active_polarity, active_frac = 0;
+	const u64 f = params->precision;
+	s64 error;
+
+	active_sym = params->ratio * tu_size;
+	active_count = div_u64(active_sym, f) * f;
+	frac = active_sym - active_count;
+
+	/* fraction < 0.5 */
+	if (frac >= (f / 2)) {
+		active_polarity = 1;
+		frac = f - frac;
+	} else {
+		active_polarity = 0;
+	}
+
+	if (frac != 0) {
+		frac = div_u64(f * f,  frac); /* 1/fraction */
+		if (frac <= (15 * f)) {
+			active_frac = div_u64(frac, f);
+
+			/* round up */
+			if (active_polarity)
+				active_frac++;
+		} else {
+			active_frac = active_polarity ? 1 : 15;
+		}
+	}
+
+	if (active_frac == 1)
+		active_polarity = 0;
+
+	if (active_polarity == 1) {
+		if (active_frac) {
+			approx = active_count + (active_frac * (f - 1)) * f;
+			approx = div_u64(approx, active_frac * f);
+		} else {
+			approx = active_count + f;
+		}
+	} else {
+		if (active_frac)
+			approx = active_count + div_u64(f, active_frac);
+		else
+			approx = active_count;
+	}
+
+	error = div_s64(active_sym - approx, tu_size);
+	error *= params->num_clocks;
+
+	if (error <= 0 && abs64(error) < params->error) {
+		params->active_count = div_u64(active_count, f);
+		params->active_polarity = active_polarity;
+		params->active_frac = active_frac;
+		params->error = abs64(error);
+		params->tu_size = tu_size;
+
+		if (error == 0)
+			return true;
+	}
+
+	return false;
+}
+
+static int tegra_sor_calc_config(struct tegra_sor *sor,
+				 struct drm_display_mode *mode,
+				 struct tegra_sor_config *config,
+				 struct drm_dp_link *link)
+{
+	const u64 f = 100000, link_rate = link->rate * 1000;
+	const u64 pclk = mode->clock * 1000;
+	u64 input, output, watermark, num;
+	struct tegra_sor_params params;
+	u32 num_syms_per_line;
+	unsigned int i;
+
+	if (!link_rate || !link->num_lanes || !pclk || !config->bits_per_pixel)
+		return -EINVAL;
+
+	output = link_rate * 8 * link->num_lanes;
+	input = pclk * config->bits_per_pixel;
+
+	if (input >= output)
+		return -ERANGE;
+
+	memset(&params, 0, sizeof(params));
+	params.ratio = div64_u64(input * f, output);
+	params.num_clocks = div_u64(link_rate * mode->hdisplay, pclk);
+	params.precision = f;
+	params.error = 64 * f;
+	params.tu_size = 64;
+
+	for (i = params.tu_size; i >= 32; i--)
+		if (tegra_sor_compute_params(sor, &params, i))
+			break;
+
+	if (params.active_frac == 0) {
+		config->active_polarity = 0;
+		config->active_count = params.active_count;
+
+		if (!params.active_polarity)
+			config->active_count--;
+
+		config->tu_size = params.tu_size;
+		config->active_frac = 1;
+	} else {
+		config->active_polarity = params.active_polarity;
+		config->active_count = params.active_count;
+		config->active_frac = params.active_frac;
+		config->tu_size = params.tu_size;
+	}
+
+	dev_dbg(sor->dev,
+		"polarity: %d active count: %d tu size: %d active frac: %d\n",
+		config->active_polarity, config->active_count,
+		config->tu_size, config->active_frac);
+
+	watermark = params.ratio * config->tu_size * (f - params.ratio);
+	watermark = div_u64(watermark, f);
+
+	watermark = div_u64(watermark + params.error, f);
+	config->watermark = watermark + (config->bits_per_pixel / 8) + 2;
+	num_syms_per_line = (mode->hdisplay * config->bits_per_pixel) *
+			    (link->num_lanes * 8);
+
+	if (config->watermark > 30) {
+		config->watermark = 30;
+		dev_err(sor->dev,
+			"unable to compute TU size, forcing watermark to %u\n",
+			config->watermark);
+	} else if (config->watermark > num_syms_per_line) {
+		config->watermark = num_syms_per_line;
+		dev_err(sor->dev, "watermark too high, forcing to %u\n",
+			config->watermark);
+	}
+
+	/* compute the number of symbols per horizontal blanking interval */
+	num = ((mode->htotal - mode->hdisplay) - 7) * link_rate;
+	config->hblank_symbols = div_u64(num, pclk);
+
+	if (link->capabilities & DP_LINK_CAP_ENHANCED_FRAMING)
+		config->hblank_symbols -= 3;
+
+	config->hblank_symbols -= 12 / link->num_lanes;
+
+	/* compute the number of symbols per vertical blanking interval */
+	num = (mode->hdisplay - 25) * link_rate;
+	config->vblank_symbols = div_u64(num, pclk);
+	config->vblank_symbols -= 36 / link->num_lanes + 4;
+
+	dev_dbg(sor->dev, "blank symbols: H:%u V:%u\n", config->hblank_symbols,
+		config->vblank_symbols);
+
+	return 0;
+}
+
+static int tegra_output_sor_enable(struct tegra_output *output)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	struct drm_display_mode *mode = &dc->base.mode;
+	unsigned int vbe, vse, hbe, hse, vbs, hbs, i;
+	struct tegra_sor *sor = to_sor(output);
+	struct tegra_sor_config config;
+	struct drm_dp_link link;
+	struct drm_dp_aux *aux;
+	unsigned long value;
+	int err = 0;
+
+	mutex_lock(&sor->lock);
+
+	if (sor->enabled)
+		goto unlock;
+
+	err = clk_prepare_enable(sor->clk);
+	if (err < 0)
+		goto unlock;
+
+	reset_control_deassert(sor->rst);
+
+	/* FIXME: properly convert to struct drm_dp_aux */
+	aux = (struct drm_dp_aux *)sor->dpaux;
+
+	if (sor->dpaux) {
+		err = tegra_dpaux_enable(sor->dpaux);
+		if (err < 0)
+			dev_err(sor->dev, "failed to enable DP: %d\n", err);
+
+		err = drm_dp_link_probe(aux, &link);
+		if (err < 0) {
+			dev_err(sor->dev, "failed to probe eDP link: %d\n",
+				err);
+			return err;
+		}
+	}
+
+	err = clk_set_parent(sor->clk, sor->clk_safe);
+	if (err < 0)
+		dev_err(sor->dev, "failed to set safe parent clock: %d\n", err);
+
+	memset(&config, 0, sizeof(config));
+	config.bits_per_pixel = 24; /* XXX: don't hardcode? */
+
+	err = tegra_sor_calc_config(sor, mode, &config, &link);
+	if (err < 0)
+		dev_err(sor->dev, "failed to compute link configuration: %d\n",
+			err);
+
+	value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
+	value &= ~SOR_CLK_CNTRL_DP_CLK_SEL_MASK;
+	value |= SOR_CLK_CNTRL_DP_CLK_SEL_SINGLE_DPCLK;
+	tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
+
+	value = tegra_sor_readl(sor, SOR_PLL_2);
+	value &= ~SOR_PLL_2_BANDGAP_POWERDOWN;
+	tegra_sor_writel(sor, value, SOR_PLL_2);
+	usleep_range(20, 100);
+
+	value = tegra_sor_readl(sor, SOR_PLL_3);
+	value |= SOR_PLL_3_PLL_VDD_MODE_V3_3;
+	tegra_sor_writel(sor, value, SOR_PLL_3);
+
+	value = SOR_PLL_0_ICHPMP(0xf) | SOR_PLL_0_VCOCAP_RST |
+		SOR_PLL_0_PLLREG_LEVEL_V45 | SOR_PLL_0_RESISTOR_EXT;
+	tegra_sor_writel(sor, value, SOR_PLL_0);
+
+	value = tegra_sor_readl(sor, SOR_PLL_2);
+	value |= SOR_PLL_2_SEQ_PLLCAPPD;
+	value &= ~SOR_PLL_2_SEQ_PLLCAPPD_ENFORCE;
+	value |= SOR_PLL_2_LVDS_ENABLE;
+	tegra_sor_writel(sor, value, SOR_PLL_2);
+
+	value = SOR_PLL_1_TERM_COMPOUT | SOR_PLL_1_TMDS_TERM;
+	tegra_sor_writel(sor, value, SOR_PLL_1);
+
+	while (true) {
+		value = tegra_sor_readl(sor, SOR_PLL_2);
+		if ((value & SOR_PLL_2_SEQ_PLLCAPPD_ENFORCE) == 0)
+			break;
+
+		usleep_range(250, 1000);
+	}
+
+	value = tegra_sor_readl(sor, SOR_PLL_2);
+	value &= ~SOR_PLL_2_POWERDOWN_OVERRIDE;
+	value &= ~SOR_PLL_2_PORT_POWERDOWN;
+	tegra_sor_writel(sor, value, SOR_PLL_2);
+
+	/*
+	 * power up
+	 */
+
+	/* set safe link bandwidth (1.62 Gbps) */
+	value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
+	value &= ~SOR_CLK_CNTRL_DP_LINK_SPEED_MASK;
+	value |= SOR_CLK_CNTRL_DP_LINK_SPEED_G1_62;
+	tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
+
+	/* step 1 */
+	value = tegra_sor_readl(sor, SOR_PLL_2);
+	value |= SOR_PLL_2_SEQ_PLLCAPPD_ENFORCE | SOR_PLL_2_PORT_POWERDOWN |
+		 SOR_PLL_2_BANDGAP_POWERDOWN;
+	tegra_sor_writel(sor, value, SOR_PLL_2);
+
+	value = tegra_sor_readl(sor, SOR_PLL_0);
+	value |= SOR_PLL_0_VCOPD | SOR_PLL_0_POWER_OFF;
+	tegra_sor_writel(sor, value, SOR_PLL_0);
+
+	value = tegra_sor_readl(sor, SOR_DP_PADCTL_0);
+	value &= ~SOR_DP_PADCTL_PAD_CAL_PD;
+	tegra_sor_writel(sor, value, SOR_DP_PADCTL_0);
+
+	/* step 2 */
+	err = tegra_io_rail_power_on(TEGRA_IO_RAIL_LVDS);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to power on I/O rail: %d\n", err);
+		goto unlock;
+	}
+
+	usleep_range(5, 100);
+
+	/* step 3 */
+	value = tegra_sor_readl(sor, SOR_PLL_2);
+	value &= ~SOR_PLL_2_BANDGAP_POWERDOWN;
+	tegra_sor_writel(sor, value, SOR_PLL_2);
+
+	usleep_range(20, 100);
+
+	/* step 4 */
+	value = tegra_sor_readl(sor, SOR_PLL_0);
+	value &= ~SOR_PLL_0_POWER_OFF;
+	value &= ~SOR_PLL_0_VCOPD;
+	tegra_sor_writel(sor, value, SOR_PLL_0);
+
+	value = tegra_sor_readl(sor, SOR_PLL_2);
+	value &= ~SOR_PLL_2_SEQ_PLLCAPPD_ENFORCE;
+	tegra_sor_writel(sor, value, SOR_PLL_2);
+
+	usleep_range(200, 1000);
+
+	/* step 5 */
+	value = tegra_sor_readl(sor, SOR_PLL_2);
+	value &= ~SOR_PLL_2_PORT_POWERDOWN;
+	tegra_sor_writel(sor, value, SOR_PLL_2);
+
+	/* switch to DP clock */
+	err = clk_set_parent(sor->clk, sor->clk_dp);
+	if (err < 0)
+		dev_err(sor->dev, "failed to set DP parent clock: %d\n", err);
+
+	/* power DP lanes */
+	value = tegra_sor_readl(sor, SOR_DP_PADCTL_0);
+
+	if (link.num_lanes <= 2)
+		value &= ~(SOR_DP_PADCTL_PD_TXD_3 | SOR_DP_PADCTL_PD_TXD_2);
+	else
+		value |= SOR_DP_PADCTL_PD_TXD_3 | SOR_DP_PADCTL_PD_TXD_2;
+
+	if (link.num_lanes <= 1)
+		value &= ~SOR_DP_PADCTL_PD_TXD_1;
+	else
+		value |= SOR_DP_PADCTL_PD_TXD_1;
+
+	if (link.num_lanes == 0)
+		value &= ~SOR_DP_PADCTL_PD_TXD_0;
+	else
+		value |= SOR_DP_PADCTL_PD_TXD_0;
+
+	tegra_sor_writel(sor, value, SOR_DP_PADCTL_0);
+
+	value = tegra_sor_readl(sor, SOR_DP_LINKCTL_0);
+	value &= ~SOR_DP_LINKCTL_LANE_COUNT_MASK;
+	value |= SOR_DP_LINKCTL_LANE_COUNT(link.num_lanes);
+	tegra_sor_writel(sor, value, SOR_DP_LINKCTL_0);
+
+	/* start lane sequencer */
+	value = SOR_LANE_SEQ_CTL_TRIGGER | SOR_LANE_SEQ_CTL_SEQUENCE_DOWN |
+		SOR_LANE_SEQ_CTL_POWER_STATE_UP;
+	tegra_sor_writel(sor, value, SOR_LANE_SEQ_CTL);
+
+	while (true) {
+		value = tegra_sor_readl(sor, SOR_LANE_SEQ_CTL);
+		if ((value & SOR_LANE_SEQ_CTL_TRIGGER) == 0)
+			break;
+
+		usleep_range(250, 1000);
+	}
+
+	/* set link bandwidth */
+	value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
+	value &= ~SOR_CLK_CNTRL_DP_LINK_SPEED_MASK;
+	value |= drm_dp_link_rate_to_bw_code(link.rate) << 2;
+	tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
+
+	/* set linkctl */
+	value = tegra_sor_readl(sor, SOR_DP_LINKCTL_0);
+	value |= SOR_DP_LINKCTL_ENABLE;
+
+	value &= ~SOR_DP_LINKCTL_TU_SIZE_MASK;
+	value |= SOR_DP_LINKCTL_TU_SIZE(config.tu_size);
+
+	value |= SOR_DP_LINKCTL_ENHANCED_FRAME;
+	tegra_sor_writel(sor, value, SOR_DP_LINKCTL_0);
+
+	for (i = 0, value = 0; i < 4; i++) {
+		unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
+				     SOR_DP_TPG_SCRAMBLER_GALIOS |
+				     SOR_DP_TPG_PATTERN_NONE;
+		value = (value << 8) | lane;
+	}
+
+	tegra_sor_writel(sor, value, SOR_DP_TPG);
+
+	value = tegra_sor_readl(sor, SOR_DP_CONFIG_0);
+	value &= ~SOR_DP_CONFIG_WATERMARK_MASK;
+	value |= SOR_DP_CONFIG_WATERMARK(config.watermark);
+
+	value &= ~SOR_DP_CONFIG_ACTIVE_SYM_COUNT_MASK;
+	value |= SOR_DP_CONFIG_ACTIVE_SYM_COUNT(config.active_count);
+
+	value &= ~SOR_DP_CONFIG_ACTIVE_SYM_FRAC_MASK;
+	value |= SOR_DP_CONFIG_ACTIVE_SYM_FRAC(config.active_frac);
+
+	if (config.active_polarity)
+		value |= SOR_DP_CONFIG_ACTIVE_SYM_POLARITY;
+	else
+		value &= ~SOR_DP_CONFIG_ACTIVE_SYM_POLARITY;
+
+	value |= SOR_DP_CONFIG_ACTIVE_SYM_ENABLE;
+	value |= SOR_DP_CONFIG_DISPARITY_NEGATIVE;
+	tegra_sor_writel(sor, value, SOR_DP_CONFIG_0);
+
+	value = tegra_sor_readl(sor, SOR_DP_AUDIO_HBLANK_SYMBOLS);
+	value &= ~SOR_DP_AUDIO_HBLANK_SYMBOLS_MASK;
+	value |= config.hblank_symbols & 0xffff;
+	tegra_sor_writel(sor, value, SOR_DP_AUDIO_HBLANK_SYMBOLS);
+
+	value = tegra_sor_readl(sor, SOR_DP_AUDIO_VBLANK_SYMBOLS);
+	value &= ~SOR_DP_AUDIO_VBLANK_SYMBOLS_MASK;
+	value |= config.vblank_symbols & 0xffff;
+	tegra_sor_writel(sor, value, SOR_DP_AUDIO_VBLANK_SYMBOLS);
+
+	/* enable pad calibration logic */
+	value = tegra_sor_readl(sor, SOR_DP_PADCTL_0);
+	value |= SOR_DP_PADCTL_PAD_CAL_PD;
+	tegra_sor_writel(sor, value, SOR_DP_PADCTL_0);
+
+	if (sor->dpaux) {
+		u8 rate, lanes;
+
+		err = drm_dp_link_probe(aux, &link);
+		if (err < 0) {
+			dev_err(sor->dev, "failed to probe eDP link: %d\n",
+				err);
+			goto unlock;
+		}
+
+		err = drm_dp_link_power_up(aux, &link);
+		if (err < 0) {
+			dev_err(sor->dev, "failed to power up eDP link: %d\n",
+				err);
+			goto unlock;
+		}
+
+		err = drm_dp_link_configure(aux, &link);
+		if (err < 0) {
+			dev_err(sor->dev, "failed to configure eDP link: %d\n",
+				err);
+			goto unlock;
+		}
+
+		rate = drm_dp_link_rate_to_bw_code(link.rate);
+		lanes = link.num_lanes;
+
+		value = tegra_sor_readl(sor, SOR_CLK_CNTRL);
+		value &= ~SOR_CLK_CNTRL_DP_LINK_SPEED_MASK;
+		value |= SOR_CLK_CNTRL_DP_LINK_SPEED(rate);
+		tegra_sor_writel(sor, value, SOR_CLK_CNTRL);
+
+		value = tegra_sor_readl(sor, SOR_DP_LINKCTL_0);
+		value &= ~SOR_DP_LINKCTL_LANE_COUNT_MASK;
+		value |= SOR_DP_LINKCTL_LANE_COUNT(lanes);
+
+		if (link.capabilities & DP_LINK_CAP_ENHANCED_FRAMING)
+			value |= SOR_DP_LINKCTL_ENHANCED_FRAME;
+
+		tegra_sor_writel(sor, value, SOR_DP_LINKCTL_0);
+
+		/* disable training pattern generator */
+
+		for (i = 0; i < link.num_lanes; i++) {
+			unsigned long lane = SOR_DP_TPG_CHANNEL_CODING |
+					     SOR_DP_TPG_SCRAMBLER_GALIOS |
+					     SOR_DP_TPG_PATTERN_NONE;
+			value = (value << 8) | lane;
+		}
+
+		tegra_sor_writel(sor, value, SOR_DP_TPG);
+
+		err = tegra_sor_dp_train_fast(sor, &link);
+		if (err < 0) {
+			dev_err(sor->dev, "DP fast link training failed: %d\n",
+				err);
+			goto unlock;
+		}
+
+		dev_dbg(sor->dev, "fast link training succeeded\n");
+	}
+
+	err = tegra_sor_power_up(sor, 250);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to power up SOR: %d\n", err);
+		goto unlock;
+	}
+
+	/* start display controller in continuous mode */
+	value = tegra_dc_readl(dc, DC_CMD_STATE_ACCESS);
+	value |= WRITE_MUX;
+	tegra_dc_writel(dc, value, DC_CMD_STATE_ACCESS);
+
+	tegra_dc_writel(dc, VSYNC_H_POSITION(1), DC_DISP_DISP_TIMING_OPTIONS);
+	tegra_dc_writel(dc, DISP_CTRL_MODE_C_DISPLAY, DC_CMD_DISPLAY_COMMAND);
+
+	value = tegra_dc_readl(dc, DC_CMD_STATE_ACCESS);
+	value &= ~WRITE_MUX;
+	tegra_dc_writel(dc, value, DC_CMD_STATE_ACCESS);
+
+	/*
+	 * configure panel (24bpp, vsync-, hsync-, DP-A protocol, complete
+	 * raster, associate with display controller)
+	 */
+	value = SOR_STATE_ASY_VSYNCPOL |
+		SOR_STATE_ASY_HSYNCPOL |
+		SOR_STATE_ASY_PROTOCOL_DP_A |
+		SOR_STATE_ASY_CRC_MODE_COMPLETE |
+		SOR_STATE_ASY_OWNER(dc->pipe + 1);
+
+	switch (config.bits_per_pixel) {
+	case 24:
+		value |= SOR_STATE_ASY_PIXELDEPTH_BPP_24_444;
+		break;
+
+	case 18:
+		value |= SOR_STATE_ASY_PIXELDEPTH_BPP_18_444;
+		break;
+
+	default:
+		BUG();
+		break;
+	}
+
+	tegra_sor_writel(sor, value, SOR_STATE_1);
+
+	/*
+	 * TODO: The video timing programming below doesn't seem to match the
+	 * register definitions.
+	 */
+
+	value = ((mode->vtotal & 0x7fff) << 16) | (mode->htotal & 0x7fff);
+	tegra_sor_writel(sor, value, SOR_HEAD_STATE_1(0));
+
+	vse = mode->vsync_end - mode->vsync_start - 1;
+	hse = mode->hsync_end - mode->hsync_start - 1;
+
+	value = ((vse & 0x7fff) << 16) | (hse & 0x7fff);
+	tegra_sor_writel(sor, value, SOR_HEAD_STATE_2(0));
+
+	vbe = vse + (mode->vsync_start - mode->vdisplay);
+	hbe = hse + (mode->hsync_start - mode->hdisplay);
+
+	value = ((vbe & 0x7fff) << 16) | (hbe & 0x7fff);
+	tegra_sor_writel(sor, value, SOR_HEAD_STATE_3(0));
+
+	vbs = vbe + mode->vdisplay;
+	hbs = hbe + mode->hdisplay;
+
+	value = ((vbs & 0x7fff) << 16) | (hbs & 0x7fff);
+	tegra_sor_writel(sor, value, SOR_HEAD_STATE_4(0));
+
+	/* CSTM (LVDS, link A/B, upper) */
+	value = SOR_CSTM_LVDS | SOR_CSTM_LINK_ACT_A | SOR_CSTM_LINK_ACT_B |
+		SOR_CSTM_UPPER;
+	tegra_sor_writel(sor, value, SOR_CSTM);
+
+	/* PWM setup */
+	err = tegra_sor_setup_pwm(sor, 250);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to setup PWM: %d\n", err);
+		goto unlock;
+	}
+
+	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+	value |= SOR_ENABLE;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+	tegra_sor_update(sor);
+
+	err = tegra_sor_attach(sor);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to attach SOR: %d\n", err);
+		goto unlock;
+	}
+
+	err = tegra_sor_wakeup(sor);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to enable DC: %d\n", err);
+		goto unlock;
+	}
+
+	sor->enabled = true;
+
+unlock:
+	mutex_unlock(&sor->lock);
+	return err;
+}
+
+static int tegra_sor_detach(struct tegra_sor *sor)
+{
+	unsigned long value, timeout;
+
+	/* switch to safe mode */
+	value = tegra_sor_readl(sor, SOR_SUPER_STATE_1);
+	value &= ~SOR_SUPER_STATE_MODE_NORMAL;
+	tegra_sor_writel(sor, value, SOR_SUPER_STATE_1);
+	tegra_sor_super_update(sor);
+
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_PWR);
+		if (value & SOR_PWR_MODE_SAFE)
+			break;
+	}
+
+	if ((value & SOR_PWR_MODE_SAFE) == 0)
+		return -ETIMEDOUT;
+
+	/* go to sleep */
+	value = tegra_sor_readl(sor, SOR_SUPER_STATE_1);
+	value &= ~SOR_SUPER_STATE_HEAD_MODE_MASK;
+	tegra_sor_writel(sor, value, SOR_SUPER_STATE_1);
+	tegra_sor_super_update(sor);
+
+	/* detach */
+	value = tegra_sor_readl(sor, SOR_SUPER_STATE_1);
+	value &= ~SOR_SUPER_STATE_ATTACHED;
+	tegra_sor_writel(sor, value, SOR_SUPER_STATE_1);
+	tegra_sor_super_update(sor);
+
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_TEST);
+		if ((value & SOR_TEST_ATTACHED) == 0)
+			break;
+
+		usleep_range(25, 100);
+	}
+
+	if ((value & SOR_TEST_ATTACHED) != 0)
+		return -ETIMEDOUT;
+
+	return 0;
+}
+
+static int tegra_sor_power_down(struct tegra_sor *sor)
+{
+	unsigned long value, timeout;
+	int err;
+
+	value = tegra_sor_readl(sor, SOR_PWR);
+	value &= ~SOR_PWR_NORMAL_STATE_PU;
+	value |= SOR_PWR_TRIGGER;
+	tegra_sor_writel(sor, value, SOR_PWR);
+
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_PWR);
+		if ((value & SOR_PWR_TRIGGER) == 0)
+			return 0;
+
+		usleep_range(25, 100);
+	}
+
+	if ((value & SOR_PWR_TRIGGER) != 0)
+		return -ETIMEDOUT;
+
+	err = clk_set_parent(sor->clk, sor->clk_safe);
+	if (err < 0)
+		dev_err(sor->dev, "failed to set safe parent clock: %d\n", err);
+
+	value = tegra_sor_readl(sor, SOR_DP_PADCTL_0);
+	value &= ~(SOR_DP_PADCTL_PD_TXD_3 | SOR_DP_PADCTL_PD_TXD_0 |
+		   SOR_DP_PADCTL_PD_TXD_1 | SOR_DP_PADCTL_PD_TXD_2);
+	tegra_sor_writel(sor, value, SOR_DP_PADCTL_0);
+
+	/* stop lane sequencer */
+	value = SOR_LANE_SEQ_CTL_TRIGGER | SOR_LANE_SEQ_CTL_SEQUENCE_UP |
+		SOR_LANE_SEQ_CTL_POWER_STATE_DOWN;
+	tegra_sor_writel(sor, value, SOR_LANE_SEQ_CTL);
+
+	timeout = jiffies + msecs_to_jiffies(250);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_LANE_SEQ_CTL);
+		if ((value & SOR_LANE_SEQ_CTL_TRIGGER) == 0)
+			break;
+
+		usleep_range(25, 100);
+	}
+
+	if ((value & SOR_LANE_SEQ_CTL_TRIGGER) != 0)
+		return -ETIMEDOUT;
+
+	value = tegra_sor_readl(sor, SOR_PLL_2);
+	value |= SOR_PLL_2_PORT_POWERDOWN;
+	tegra_sor_writel(sor, value, SOR_PLL_2);
+
+	usleep_range(20, 100);
+
+	value = tegra_sor_readl(sor, SOR_PLL_0);
+	value |= SOR_PLL_0_POWER_OFF;
+	value |= SOR_PLL_0_VCOPD;
+	tegra_sor_writel(sor, value, SOR_PLL_0);
+
+	value = tegra_sor_readl(sor, SOR_PLL_2);
+	value |= SOR_PLL_2_SEQ_PLLCAPPD;
+	value |= SOR_PLL_2_SEQ_PLLCAPPD_ENFORCE;
+	tegra_sor_writel(sor, value, SOR_PLL_2);
+
+	usleep_range(20, 100);
+
+	return 0;
+}
+
+static int tegra_output_sor_disable(struct tegra_output *output)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	struct tegra_sor *sor = to_sor(output);
+	unsigned long value;
+	int err = 0;
+
+	mutex_lock(&sor->lock);
+
+	if (!sor->enabled)
+		goto unlock;
+
+	err = tegra_sor_detach(sor);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to detach SOR: %d\n", err);
+		goto unlock;
+	}
+
+	tegra_sor_writel(sor, 0, SOR_STATE_1);
+	tegra_sor_update(sor);
+
+	/*
+	 * The following accesses registers of the display controller, so make
+	 * sure it's only executed when the output is attached to one.
+	 */
+	if (dc) {
+		/*
+		 * XXX: We can't do this here because it causes the SOR to go
+		 * into an erroneous state and the output will look scrambled
+		 * the next time it is enabled. Presumably this is because we
+		 * should be doing this only on the next VBLANK. A possible
+		 * solution would be to queue a "power-off" event to trigger
+		 * this code to be run during the next VBLANK.
+		 */
+		/*
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+		value &= ~(PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+			   PW4_ENABLE | PM0_ENABLE | PM1_ENABLE);
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+		*/
+
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+		value &= ~DISP_CTRL_MODE_MASK;
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
+
+		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+		value &= ~SOR_ENABLE;
+		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+		tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+		tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+	}
+
+	err = tegra_sor_power_down(sor);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to power down SOR: %d\n", err);
+		goto unlock;
+	}
+
+	if (sor->dpaux) {
+		err = tegra_dpaux_disable(sor->dpaux);
+		if (err < 0) {
+			dev_err(sor->dev, "failed to disable DP: %d\n", err);
+			goto unlock;
+		}
+	}
+
+	err = tegra_io_rail_power_off(TEGRA_IO_RAIL_LVDS);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to power off I/O rail: %d\n", err);
+		goto unlock;
+	}
+
+	reset_control_assert(sor->rst);
+	clk_disable_unprepare(sor->clk);
+
+	sor->enabled = false;
+
+unlock:
+	mutex_unlock(&sor->lock);
+	return err;
+}
+
+static int tegra_output_sor_setup_clock(struct tegra_output *output,
+					struct clk *clk, unsigned long pclk,
+					unsigned int *div)
+{
+	struct tegra_sor *sor = to_sor(output);
+	int err;
+
+	err = clk_set_parent(clk, sor->clk_parent);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to set parent clock: %d\n", err);
+		return err;
+	}
+
+	err = clk_set_rate(sor->clk_parent, pclk);
+	if (err < 0) {
+		dev_err(sor->dev, "failed to set clock rate to %lu Hz\n", pclk);
+		return err;
+	}
+
+	*div = 0;
+
+	return 0;
+}
+
+static int tegra_output_sor_check_mode(struct tegra_output *output,
+				       struct drm_display_mode *mode,
+				       enum drm_mode_status *status)
+{
+	/*
+	 * FIXME: For now, always assume that the mode is okay.
+	 */
+
+	*status = MODE_OK;
+
+	return 0;
+}
+
+static enum drm_connector_status
+tegra_output_sor_detect(struct tegra_output *output)
+{
+	struct tegra_sor *sor = to_sor(output);
+
+	if (sor->dpaux)
+		return tegra_dpaux_detect(sor->dpaux);
+
+	return connector_status_unknown;
+}
+
+static const struct tegra_output_ops sor_ops = {
+	.enable = tegra_output_sor_enable,
+	.disable = tegra_output_sor_disable,
+	.setup_clock = tegra_output_sor_setup_clock,
+	.check_mode = tegra_output_sor_check_mode,
+	.detect = tegra_output_sor_detect,
+};
+
+static int tegra_sor_crc_open(struct inode *inode, struct file *file)
+{
+	file->private_data = inode->i_private;
+
+	return 0;
+}
+
+static int tegra_sor_crc_release(struct inode *inode, struct file *file)
+{
+	return 0;
+}
+
+static int tegra_sor_crc_wait(struct tegra_sor *sor, unsigned long timeout)
+{
+	u32 value;
+
+	timeout = jiffies + msecs_to_jiffies(timeout);
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_sor_readl(sor, SOR_CRC_A);
+		if (value & SOR_CRC_A_VALID)
+			return 0;
+
+		usleep_range(100, 200);
+	}
+
+	return -ETIMEDOUT;
+}
+
+static ssize_t tegra_sor_crc_read(struct file *file, char __user *buffer,
+				  size_t size, loff_t *ppos)
+{
+	struct tegra_sor *sor = file->private_data;
+	ssize_t num, err;
+	char buf[10];
+	u32 value;
+
+	mutex_lock(&sor->lock);
+
+	if (!sor->enabled) {
+		err = -EAGAIN;
+		goto unlock;
+	}
+
+	value = tegra_sor_readl(sor, SOR_STATE_1);
+	value &= ~SOR_STATE_ASY_CRC_MODE_MASK;
+	tegra_sor_writel(sor, value, SOR_STATE_1);
+
+	value = tegra_sor_readl(sor, SOR_CRC_CNTRL);
+	value |= SOR_CRC_CNTRL_ENABLE;
+	tegra_sor_writel(sor, value, SOR_CRC_CNTRL);
+
+	value = tegra_sor_readl(sor, SOR_TEST);
+	value &= ~SOR_TEST_CRC_POST_SERIALIZE;
+	tegra_sor_writel(sor, value, SOR_TEST);
+
+	err = tegra_sor_crc_wait(sor, 100);
+	if (err < 0)
+		goto unlock;
+
+	tegra_sor_writel(sor, SOR_CRC_A_RESET, SOR_CRC_A);
+	value = tegra_sor_readl(sor, SOR_CRC_B);
+
+	num = scnprintf(buf, sizeof(buf), "%08x\n", value);
+
+	err = simple_read_from_buffer(buffer, size, ppos, buf, num);
+
+unlock:
+	mutex_unlock(&sor->lock);
+	return err;
+}
+
+static const struct file_operations tegra_sor_crc_fops = {
+	.owner = THIS_MODULE,
+	.open = tegra_sor_crc_open,
+	.read = tegra_sor_crc_read,
+	.release = tegra_sor_crc_release,
+};
+
+static int tegra_sor_debugfs_init(struct tegra_sor *sor,
+				  struct drm_minor *minor)
+{
+	struct dentry *entry;
+	int err = 0;
+
+	sor->debugfs = debugfs_create_dir("sor", minor->debugfs_root);
+	if (!sor->debugfs)
+		return -ENOMEM;
+
+	entry = debugfs_create_file("crc", 0644, sor->debugfs, sor,
+				    &tegra_sor_crc_fops);
+	if (!entry) {
+		dev_err(sor->dev,
+			"cannot create /sys/kernel/debug/dri/%s/sor/crc\n",
+			minor->debugfs_root->d_name.name);
+		err = -ENOMEM;
+		goto remove;
+	}
+
+	return err;
+
+remove:
+	debugfs_remove(sor->debugfs);
+	sor->debugfs = NULL;
+	return err;
+}
+
+static int tegra_sor_debugfs_exit(struct tegra_sor *sor)
+{
+	debugfs_remove_recursive(sor->debugfs);
+	sor->debugfs = NULL;
+
+	return 0;
+}
+
+static int tegra_sor_init(struct host1x_client *client)
+{
+	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct tegra_sor *sor = host1x_client_to_sor(client);
+	int err;
+
+	if (!sor->dpaux)
+		return -ENODEV;
+
+	sor->output.type = TEGRA_OUTPUT_EDP;
+
+	sor->output.dev = sor->dev;
+	sor->output.ops = &sor_ops;
+
+	err = tegra_output_init(drm, &sor->output);
+	if (err < 0) {
+		dev_err(sor->dev, "output setup failed: %d\n", err);
+		return err;
+	}
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_sor_debugfs_init(sor, drm->primary);
+		if (err < 0)
+			dev_err(sor->dev, "debugfs setup failed: %d\n", err);
+	}
+
+	if (sor->dpaux) {
+		err = tegra_dpaux_attach(sor->dpaux, &sor->output);
+		if (err < 0) {
+			dev_err(sor->dev, "failed to attach DP: %d\n", err);
+			return err;
+		}
+	}
+
+	return 0;
+}
+
+static int tegra_sor_exit(struct host1x_client *client)
+{
+	struct tegra_sor *sor = host1x_client_to_sor(client);
+	int err;
+
+	err = tegra_output_disable(&sor->output);
+	if (err < 0) {
+		dev_err(sor->dev, "output failed to disable: %d\n", err);
+		return err;
+	}
+
+	if (sor->dpaux) {
+		err = tegra_dpaux_detach(sor->dpaux);
+		if (err < 0) {
+			dev_err(sor->dev, "failed to detach DP: %d\n", err);
+			return err;
+		}
+	}
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_sor_debugfs_exit(sor);
+		if (err < 0)
+			dev_err(sor->dev, "debugfs cleanup failed: %d\n", err);
+	}
+
+	err = tegra_output_exit(&sor->output);
+	if (err < 0) {
+		dev_err(sor->dev, "output cleanup failed: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static const struct host1x_client_ops sor_client_ops = {
+	.init = tegra_sor_init,
+	.exit = tegra_sor_exit,
+};
+
+static int tegra_sor_probe(struct platform_device *pdev)
+{
+	struct device_node *np;
+	struct tegra_sor *sor;
+	struct resource *regs;
+	int err;
+
+	sor = devm_kzalloc(&pdev->dev, sizeof(*sor), GFP_KERNEL);
+	if (!sor)
+		return -ENOMEM;
+
+	sor->output.dev = sor->dev = &pdev->dev;
+
+	np = of_parse_phandle(pdev->dev.of_node, "nvidia,dpaux", 0);
+	if (np) {
+		sor->dpaux = tegra_dpaux_find_by_of_node(np);
+		of_node_put(np);
+
+		if (!sor->dpaux)
+			return -EPROBE_DEFER;
+	}
+
+	err = tegra_output_probe(&sor->output);
+	if (err < 0)
+		return err;
+
+	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	sor->regs = devm_ioremap_resource(&pdev->dev, regs);
+	if (IS_ERR(sor->regs))
+		return PTR_ERR(sor->regs);
+
+	sor->rst = devm_reset_control_get(&pdev->dev, "sor");
+	if (IS_ERR(sor->rst))
+		return PTR_ERR(sor->rst);
+
+	sor->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(sor->clk))
+		return PTR_ERR(sor->clk);
+
+	sor->clk_parent = devm_clk_get(&pdev->dev, "parent");
+	if (IS_ERR(sor->clk_parent))
+		return PTR_ERR(sor->clk_parent);
+
+	err = clk_prepare_enable(sor->clk_parent);
+	if (err < 0)
+		return err;
+
+	sor->clk_safe = devm_clk_get(&pdev->dev, "safe");
+	if (IS_ERR(sor->clk_safe))
+		return PTR_ERR(sor->clk_safe);
+
+	err = clk_prepare_enable(sor->clk_safe);
+	if (err < 0)
+		return err;
+
+	sor->clk_dp = devm_clk_get(&pdev->dev, "dp");
+	if (IS_ERR(sor->clk_dp))
+		return PTR_ERR(sor->clk_dp);
+
+	err = clk_prepare_enable(sor->clk_dp);
+	if (err < 0)
+		return err;
+
+	INIT_LIST_HEAD(&sor->client.list);
+	sor->client.ops = &sor_client_ops;
+	sor->client.dev = &pdev->dev;
+
+	mutex_init(&sor->lock);
+
+	err = host1x_client_register(&sor->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to register host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	platform_set_drvdata(pdev, sor);
+
+	return 0;
+}
+
+static int tegra_sor_remove(struct platform_device *pdev)
+{
+	struct tegra_sor *sor = platform_get_drvdata(pdev);
+	int err;
+
+	err = host1x_client_unregister(&sor->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to unregister host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	clk_disable_unprepare(sor->clk_parent);
+	clk_disable_unprepare(sor->clk_safe);
+	clk_disable_unprepare(sor->clk_dp);
+	clk_disable_unprepare(sor->clk);
+
+	return 0;
+}
+
+static const struct of_device_id tegra_sor_of_match[] = {
+	{ .compatible = "nvidia,tegra124-sor", },
+	{ },
+};
+
+struct platform_driver tegra_sor_driver = {
+	.driver = {
+		.name = "tegra-sor",
+		.of_match_table = tegra_sor_of_match,
+	},
+	.probe = tegra_sor_probe,
+	.remove = tegra_sor_remove,
+};
diff --git a/drivers/gpu/drm/tegra/sor.h b/drivers/gpu/drm/tegra/sor.h
new file mode 100644
index 00000000000..a5f8853fedb
--- /dev/null
+++ b/drivers/gpu/drm/tegra/sor.h
@@ -0,0 +1,282 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef DRM_TEGRA_SOR_H
+#define DRM_TEGRA_SOR_H
+
+#define SOR_CTXSW 0x00
+
+#define SOR_SUPER_STATE_0 0x01
+
+#define SOR_SUPER_STATE_1 0x02
+#define  SOR_SUPER_STATE_ATTACHED		(1 << 3)
+#define  SOR_SUPER_STATE_MODE_NORMAL		(1 << 2)
+#define  SOR_SUPER_STATE_HEAD_MODE_MASK		(3 << 0)
+#define  SOR_SUPER_STATE_HEAD_MODE_AWAKE	(2 << 0)
+#define  SOR_SUPER_STATE_HEAD_MODE_SNOOZE	(1 << 0)
+#define  SOR_SUPER_STATE_HEAD_MODE_SLEEP	(0 << 0)
+
+#define SOR_STATE_0 0x03
+
+#define SOR_STATE_1 0x04
+#define  SOR_STATE_ASY_PIXELDEPTH_MASK		(0xf << 17)
+#define  SOR_STATE_ASY_PIXELDEPTH_BPP_18_444	(0x2 << 17)
+#define  SOR_STATE_ASY_PIXELDEPTH_BPP_24_444	(0x5 << 17)
+#define  SOR_STATE_ASY_VSYNCPOL			(1 << 13)
+#define  SOR_STATE_ASY_HSYNCPOL			(1 << 12)
+#define  SOR_STATE_ASY_PROTOCOL_MASK		(0xf << 8)
+#define  SOR_STATE_ASY_PROTOCOL_CUSTOM		(0xf << 8)
+#define  SOR_STATE_ASY_PROTOCOL_DP_A		(0x8 << 8)
+#define  SOR_STATE_ASY_PROTOCOL_DP_B		(0x9 << 8)
+#define  SOR_STATE_ASY_PROTOCOL_LVDS		(0x0 << 8)
+#define  SOR_STATE_ASY_CRC_MODE_MASK		(0x3 << 6)
+#define  SOR_STATE_ASY_CRC_MODE_NON_ACTIVE	(0x2 << 6)
+#define  SOR_STATE_ASY_CRC_MODE_COMPLETE	(0x1 << 6)
+#define  SOR_STATE_ASY_CRC_MODE_ACTIVE		(0x0 << 6)
+#define  SOR_STATE_ASY_OWNER(x)			(((x) & 0xf) << 0)
+
+#define SOR_HEAD_STATE_0(x) (0x05 + (x))
+#define SOR_HEAD_STATE_1(x) (0x07 + (x))
+#define SOR_HEAD_STATE_2(x) (0x09 + (x))
+#define SOR_HEAD_STATE_3(x) (0x0b + (x))
+#define SOR_HEAD_STATE_4(x) (0x0d + (x))
+#define SOR_HEAD_STATE_5(x) (0x0f + (x))
+#define SOR_CRC_CNTRL 0x11
+#define  SOR_CRC_CNTRL_ENABLE			(1 << 0)
+#define SOR_DP_DEBUG_MVID 0x12
+
+#define SOR_CLK_CNTRL 0x13
+#define  SOR_CLK_CNTRL_DP_LINK_SPEED_MASK	(0x1f << 2)
+#define  SOR_CLK_CNTRL_DP_LINK_SPEED(x)		(((x) & 0x1f) << 2)
+#define  SOR_CLK_CNTRL_DP_LINK_SPEED_G1_62	(0x06 << 2)
+#define  SOR_CLK_CNTRL_DP_LINK_SPEED_G2_70	(0x0a << 2)
+#define  SOR_CLK_CNTRL_DP_LINK_SPEED_G5_40	(0x14 << 2)
+#define  SOR_CLK_CNTRL_DP_CLK_SEL_MASK		(3 << 0)
+#define  SOR_CLK_CNTRL_DP_CLK_SEL_SINGLE_PCLK	(0 << 0)
+#define  SOR_CLK_CNTRL_DP_CLK_SEL_DIFF_PCLK	(1 << 0)
+#define  SOR_CLK_CNTRL_DP_CLK_SEL_SINGLE_DPCLK	(2 << 0)
+#define  SOR_CLK_CNTRL_DP_CLK_SEL_DIFF_DPCLK	(3 << 0)
+
+#define SOR_CAP 0x14
+
+#define SOR_PWR 0x15
+#define  SOR_PWR_TRIGGER			(1 << 31)
+#define  SOR_PWR_MODE_SAFE			(1 << 28)
+#define  SOR_PWR_NORMAL_STATE_PU		(1 << 0)
+
+#define SOR_TEST 0x16
+#define  SOR_TEST_CRC_POST_SERIALIZE		(1 << 23)
+#define  SOR_TEST_ATTACHED			(1 << 10)
+#define  SOR_TEST_HEAD_MODE_MASK		(3 << 8)
+#define  SOR_TEST_HEAD_MODE_AWAKE		(2 << 8)
+
+#define SOR_PLL_0 0x17
+#define  SOR_PLL_0_ICHPMP_MASK			(0xf << 24)
+#define  SOR_PLL_0_ICHPMP(x)			(((x) & 0xf) << 24)
+#define  SOR_PLL_0_VCOCAP_MASK			(0xf << 8)
+#define  SOR_PLL_0_VCOCAP(x)			(((x) & 0xf) << 8)
+#define  SOR_PLL_0_VCOCAP_RST			SOR_PLL_0_VCOCAP(3)
+#define  SOR_PLL_0_PLLREG_MASK			(0x3 << 6)
+#define  SOR_PLL_0_PLLREG_LEVEL(x)		(((x) & 0x3) << 6)
+#define  SOR_PLL_0_PLLREG_LEVEL_V25		SOR_PLL_0_PLLREG_LEVEL(0)
+#define  SOR_PLL_0_PLLREG_LEVEL_V15		SOR_PLL_0_PLLREG_LEVEL(1)
+#define  SOR_PLL_0_PLLREG_LEVEL_V35		SOR_PLL_0_PLLREG_LEVEL(2)
+#define  SOR_PLL_0_PLLREG_LEVEL_V45		SOR_PLL_0_PLLREG_LEVEL(3)
+#define  SOR_PLL_0_PULLDOWN			(1 << 5)
+#define  SOR_PLL_0_RESISTOR_EXT			(1 << 4)
+#define  SOR_PLL_0_VCOPD			(1 << 2)
+#define  SOR_PLL_0_POWER_OFF			(1 << 0)
+
+#define SOR_PLL_1 0x18
+/* XXX: read-only bit? */
+#define  SOR_PLL_1_TERM_COMPOUT			(1 << 15)
+#define  SOR_PLL_1_TMDS_TERM			(1 << 8)
+
+#define SOR_PLL_2 0x19
+#define  SOR_PLL_2_LVDS_ENABLE			(1 << 25)
+#define  SOR_PLL_2_SEQ_PLLCAPPD_ENFORCE		(1 << 24)
+#define  SOR_PLL_2_PORT_POWERDOWN		(1 << 23)
+#define  SOR_PLL_2_BANDGAP_POWERDOWN		(1 << 22)
+#define  SOR_PLL_2_POWERDOWN_OVERRIDE		(1 << 18)
+#define  SOR_PLL_2_SEQ_PLLCAPPD			(1 << 17)
+
+#define SOR_PLL_3 0x1a
+#define  SOR_PLL_3_PLL_VDD_MODE_V1_8 (0 << 13)
+#define  SOR_PLL_3_PLL_VDD_MODE_V3_3 (1 << 13)
+
+#define SOR_CSTM 0x1b
+#define  SOR_CSTM_LVDS				(1 << 16)
+#define  SOR_CSTM_LINK_ACT_B			(1 << 15)
+#define  SOR_CSTM_LINK_ACT_A			(1 << 14)
+#define  SOR_CSTM_UPPER				(1 << 11)
+
+#define SOR_LVDS 0x1c
+#define SOR_CRC_A 0x1d
+#define  SOR_CRC_A_VALID			(1 << 0)
+#define  SOR_CRC_A_RESET			(1 << 0)
+#define SOR_CRC_B 0x1e
+#define SOR_BLANK 0x1f
+#define SOR_SEQ_CTL 0x20
+
+#define SOR_LANE_SEQ_CTL 0x21
+#define  SOR_LANE_SEQ_CTL_TRIGGER		(1 << 31)
+#define  SOR_LANE_SEQ_CTL_SEQUENCE_UP		(0 << 20)
+#define  SOR_LANE_SEQ_CTL_SEQUENCE_DOWN		(1 << 20)
+#define  SOR_LANE_SEQ_CTL_POWER_STATE_UP	(0 << 16)
+#define  SOR_LANE_SEQ_CTL_POWER_STATE_DOWN	(1 << 16)
+
+#define SOR_SEQ_INST(x) (0x22 + (x))
+
+#define SOR_PWM_DIV 0x32
+#define  SOR_PWM_DIV_MASK			0xffffff
+
+#define SOR_PWM_CTL 0x33
+#define  SOR_PWM_CTL_TRIGGER			(1 << 31)
+#define  SOR_PWM_CTL_CLK_SEL			(1 << 30)
+#define  SOR_PWM_CTL_DUTY_CYCLE_MASK		0xffffff
+
+#define SOR_VCRC_A_0 0x34
+#define SOR_VCRC_A_1 0x35
+#define SOR_VCRC_B_0 0x36
+#define SOR_VCRC_B_1 0x37
+#define SOR_CCRC_A_0 0x38
+#define SOR_CCRC_A_1 0x39
+#define SOR_CCRC_B_0 0x3a
+#define SOR_CCRC_B_1 0x3b
+#define SOR_EDATA_A_0 0x3c
+#define SOR_EDATA_A_1 0x3d
+#define SOR_EDATA_B_0 0x3e
+#define SOR_EDATA_B_1 0x3f
+#define SOR_COUNT_A_0 0x40
+#define SOR_COUNT_A_1 0x41
+#define SOR_COUNT_B_0 0x42
+#define SOR_COUNT_B_1 0x43
+#define SOR_DEBUG_A_0 0x44
+#define SOR_DEBUG_A_1 0x45
+#define SOR_DEBUG_B_0 0x46
+#define SOR_DEBUG_B_1 0x47
+#define SOR_TRIG 0x48
+#define SOR_MSCHECK 0x49
+#define SOR_XBAR_CTRL 0x4a
+#define SOR_XBAR_POL 0x4b
+
+#define SOR_DP_LINKCTL_0 0x4c
+#define  SOR_DP_LINKCTL_LANE_COUNT_MASK		(0x1f << 16)
+#define  SOR_DP_LINKCTL_LANE_COUNT(x)		(((1 << (x)) - 1) << 16)
+#define  SOR_DP_LINKCTL_ENHANCED_FRAME		(1 << 14)
+#define  SOR_DP_LINKCTL_TU_SIZE_MASK		(0x7f << 2)
+#define  SOR_DP_LINKCTL_TU_SIZE(x)		(((x) & 0x7f) << 2)
+#define  SOR_DP_LINKCTL_ENABLE			(1 << 0)
+
+#define SOR_DP_LINKCTL_1 0x4d
+
+#define SOR_LANE_DRIVE_CURRENT_0 0x4e
+#define SOR_LANE_DRIVE_CURRENT_1 0x4f
+#define SOR_LANE4_DRIVE_CURRENT_0 0x50
+#define SOR_LANE4_DRIVE_CURRENT_1 0x51
+#define  SOR_LANE_DRIVE_CURRENT_LANE3(x) (((x) & 0xff) << 24)
+#define  SOR_LANE_DRIVE_CURRENT_LANE2(x) (((x) & 0xff) << 16)
+#define  SOR_LANE_DRIVE_CURRENT_LANE1(x) (((x) & 0xff) << 8)
+#define  SOR_LANE_DRIVE_CURRENT_LANE0(x) (((x) & 0xff) << 0)
+
+#define SOR_LANE_PREEMPHASIS_0 0x52
+#define SOR_LANE_PREEMPHASIS_1 0x53
+#define SOR_LANE4_PREEMPHASIS_0 0x54
+#define SOR_LANE4_PREEMPHASIS_1 0x55
+#define  SOR_LANE_PREEMPHASIS_LANE3(x) (((x) & 0xff) << 24)
+#define  SOR_LANE_PREEMPHASIS_LANE2(x) (((x) & 0xff) << 16)
+#define  SOR_LANE_PREEMPHASIS_LANE1(x) (((x) & 0xff) << 8)
+#define  SOR_LANE_PREEMPHASIS_LANE0(x) (((x) & 0xff) << 0)
+
+#define SOR_LANE_POST_CURSOR_0 0x56
+#define SOR_LANE_POST_CURSOR_1 0x57
+#define  SOR_LANE_POST_CURSOR_LANE3(x) (((x) & 0xff) << 24)
+#define  SOR_LANE_POST_CURSOR_LANE2(x) (((x) & 0xff) << 16)
+#define  SOR_LANE_POST_CURSOR_LANE1(x) (((x) & 0xff) << 8)
+#define  SOR_LANE_POST_CURSOR_LANE0(x) (((x) & 0xff) << 0)
+
+#define SOR_DP_CONFIG_0 0x58
+#define SOR_DP_CONFIG_DISPARITY_NEGATIVE	(1 << 31)
+#define SOR_DP_CONFIG_ACTIVE_SYM_ENABLE		(1 << 26)
+#define SOR_DP_CONFIG_ACTIVE_SYM_POLARITY	(1 << 24)
+#define SOR_DP_CONFIG_ACTIVE_SYM_FRAC_MASK	(0xf << 16)
+#define SOR_DP_CONFIG_ACTIVE_SYM_FRAC(x)	(((x) & 0xf) << 16)
+#define SOR_DP_CONFIG_ACTIVE_SYM_COUNT_MASK	(0x7f << 8)
+#define SOR_DP_CONFIG_ACTIVE_SYM_COUNT(x)	(((x) & 0x7f) << 8)
+#define SOR_DP_CONFIG_WATERMARK_MASK	(0x3f << 0)
+#define SOR_DP_CONFIG_WATERMARK(x)	(((x) & 0x3f) << 0)
+
+#define SOR_DP_CONFIG_1 0x59
+#define SOR_DP_MN_0 0x5a
+#define SOR_DP_MN_1 0x5b
+
+#define SOR_DP_PADCTL_0 0x5c
+#define  SOR_DP_PADCTL_PAD_CAL_PD	(1 << 23)
+#define  SOR_DP_PADCTL_TX_PU_ENABLE	(1 << 22)
+#define  SOR_DP_PADCTL_TX_PU_MASK	(0xff << 8)
+#define  SOR_DP_PADCTL_TX_PU(x)		(((x) & 0xff) << 8)
+#define  SOR_DP_PADCTL_CM_TXD_3		(1 << 7)
+#define  SOR_DP_PADCTL_CM_TXD_2		(1 << 6)
+#define  SOR_DP_PADCTL_CM_TXD_1		(1 << 5)
+#define  SOR_DP_PADCTL_CM_TXD_0		(1 << 4)
+#define  SOR_DP_PADCTL_PD_TXD_3		(1 << 3)
+#define  SOR_DP_PADCTL_PD_TXD_0		(1 << 2)
+#define  SOR_DP_PADCTL_PD_TXD_1		(1 << 1)
+#define  SOR_DP_PADCTL_PD_TXD_2		(1 << 0)
+
+#define SOR_DP_PADCTL_1 0x5d
+
+#define SOR_DP_DEBUG_0 0x5e
+#define SOR_DP_DEBUG_1 0x5f
+
+#define SOR_DP_SPARE_0 0x60
+#define  SOR_DP_SPARE_MACRO_SOR_CLK	(1 << 2)
+#define  SOR_DP_SPARE_PANEL_INTERNAL	(1 << 1)
+#define  SOR_DP_SPARE_SEQ_ENABLE	(1 << 0)
+
+#define SOR_DP_SPARE_1 0x61
+#define SOR_DP_AUDIO_CTRL 0x62
+
+#define SOR_DP_AUDIO_HBLANK_SYMBOLS 0x63
+#define SOR_DP_AUDIO_HBLANK_SYMBOLS_MASK (0x01ffff << 0)
+
+#define SOR_DP_AUDIO_VBLANK_SYMBOLS 0x64
+#define SOR_DP_AUDIO_VBLANK_SYMBOLS_MASK (0x1fffff << 0)
+
+#define SOR_DP_GENERIC_INFOFRAME_HEADER 0x65
+#define SOR_DP_GENERIC_INFOFRAME_SUBPACK_0 0x66
+#define SOR_DP_GENERIC_INFOFRAME_SUBPACK_1 0x67
+#define SOR_DP_GENERIC_INFOFRAME_SUBPACK_2 0x68
+#define SOR_DP_GENERIC_INFOFRAME_SUBPACK_3 0x69
+#define SOR_DP_GENERIC_INFOFRAME_SUBPACK_4 0x6a
+#define SOR_DP_GENERIC_INFOFRAME_SUBPACK_5 0x6b
+#define SOR_DP_GENERIC_INFOFRAME_SUBPACK_6 0x6c
+
+#define SOR_DP_TPG 0x6d
+#define  SOR_DP_TPG_CHANNEL_CODING	(1 << 6)
+#define  SOR_DP_TPG_SCRAMBLER_MASK	(3 << 4)
+#define  SOR_DP_TPG_SCRAMBLER_FIBONACCI	(2 << 4)
+#define  SOR_DP_TPG_SCRAMBLER_GALIOS	(1 << 4)
+#define  SOR_DP_TPG_SCRAMBLER_NONE	(0 << 4)
+#define  SOR_DP_TPG_PATTERN_MASK	(0xf << 0)
+#define  SOR_DP_TPG_PATTERN_HBR2	(0x8 << 0)
+#define  SOR_DP_TPG_PATTERN_CSTM	(0x7 << 0)
+#define  SOR_DP_TPG_PATTERN_PRBS7	(0x6 << 0)
+#define  SOR_DP_TPG_PATTERN_SBLERRRATE	(0x5 << 0)
+#define  SOR_DP_TPG_PATTERN_D102	(0x4 << 0)
+#define  SOR_DP_TPG_PATTERN_TRAIN3	(0x3 << 0)
+#define  SOR_DP_TPG_PATTERN_TRAIN2	(0x2 << 0)
+#define  SOR_DP_TPG_PATTERN_TRAIN1	(0x1 << 0)
+#define  SOR_DP_TPG_PATTERN_NONE	(0x0 << 0)
+
+#define SOR_DP_TPG_CONFIG 0x6e
+#define SOR_DP_LQ_CSTM_0 0x6f
+#define SOR_DP_LQ_CSTM_1 0x70
+#define SOR_DP_LQ_CSTM_2 0x71
+
+#endif
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
index d36efc13b16..d642d4a0213 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
@@ -74,7 +74,7 @@ static void set_scanout(struct drm_crtc *crtc, int n)
 		drm_flip_work_queue(&tilcdc_crtc->unref_work, tilcdc_crtc->scanout[n]);
 		drm_flip_work_commit(&tilcdc_crtc->unref_work, priv->wq);
 	}
-	tilcdc_crtc->scanout[n] = crtc->fb;
+	tilcdc_crtc->scanout[n] = crtc->primary->fb;
 	drm_framebuffer_reference(tilcdc_crtc->scanout[n]);
 	tilcdc_crtc->dirty &= ~stat[n];
 	pm_runtime_put_sync(dev->dev);
@@ -84,7 +84,7 @@ static void update_scanout(struct drm_crtc *crtc)
 {
 	struct tilcdc_crtc *tilcdc_crtc = to_tilcdc_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
-	struct drm_framebuffer *fb = crtc->fb;
+	struct drm_framebuffer *fb = crtc->primary->fb;
 	struct drm_gem_cma_object *gem;
 	unsigned int depth, bpp;
 
@@ -159,7 +159,7 @@ static int tilcdc_crtc_page_flip(struct drm_crtc *crtc,
 		return -EBUSY;
 	}
 
-	crtc->fb = fb;
+	crtc->primary->fb = fb;
 	tilcdc_crtc->event = event;
 	update_scanout(crtc);
 
@@ -339,7 +339,7 @@ static int tilcdc_crtc_mode_set(struct drm_crtc *crtc,
 	if (priv->rev == 2) {
 		unsigned int depth, bpp;
 
-		drm_fb_get_bpp_depth(crtc->fb->pixel_format, &depth, &bpp);
+		drm_fb_get_bpp_depth(crtc->primary->fb->pixel_format, &depth, &bpp);
 		switch (bpp) {
 		case 16:
 			break;
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_drv.c b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
index 116da199b94..b20b69488dc 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_drv.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
@@ -268,7 +268,7 @@ static int tilcdc_load(struct drm_device *dev, unsigned long flags)
 	}
 
 	pm_runtime_get_sync(dev->dev);
-	ret = drm_irq_install(dev);
+	ret = drm_irq_install(dev, platform_get_irq(dev->platformdev, 0));
 	pm_runtime_put_sync(dev->dev);
 	if (ret < 0) {
 		dev_err(dev->dev, "failed to install IRQ handler\n");
@@ -311,7 +311,7 @@ static void tilcdc_lastclose(struct drm_device *dev)
 	drm_fbdev_cma_restore_mode(priv->fbdev);
 }
 
-static irqreturn_t tilcdc_irq(DRM_IRQ_ARGS)
+static irqreturn_t tilcdc_irq(int irq, void *arg)
 {
 	struct drm_device *dev = arg;
 	struct tilcdc_drm_private *priv = dev->dev_private;
@@ -444,7 +444,7 @@ static int tilcdc_mm_show(struct seq_file *m, void *arg)
 {
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
-	return drm_mm_dump_table(m, dev->mm_private);
+	return drm_mm_dump_table(m, &dev->vma_offset_manager->vm_addr_space_mm);
 }
 
 static struct drm_info_list tilcdc_debugfs_list[] = {
@@ -594,7 +594,7 @@ static int tilcdc_pdev_probe(struct platform_device *pdev)
 
 static int tilcdc_pdev_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&tilcdc_driver, pdev);
+	drm_put_dev(platform_get_drvdata(pdev));
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/ttm/ttm_agp_backend.c b/drivers/gpu/drm/ttm/ttm_agp_backend.c
index 3302f99e749..764be36397f 100644
--- a/drivers/gpu/drm/ttm/ttm_agp_backend.c
+++ b/drivers/gpu/drm/ttm/ttm_agp_backend.c
@@ -126,6 +126,7 @@ struct ttm_tt *ttm_agp_tt_create(struct ttm_bo_device *bdev,
 	agp_be->ttm.func = &ttm_agp_func;
 
 	if (ttm_tt_init(&agp_be->ttm, bdev, size, page_flags, dummy_read_page)) {
+		kfree(agp_be);
 		return NULL;
 	}
 
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 07e02c4bf5a..4ab9f7171c4 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -351,9 +351,11 @@ static int ttm_bo_handle_move_mem(struct ttm_buffer_object *bo,
 
 moved:
 	if (bo->evicted) {
-		ret = bdev->driver->invalidate_caches(bdev, bo->mem.placement);
-		if (ret)
-			pr_err("Can not flush read caches\n");
+		if (bdev->driver->invalidate_caches) {
+			ret = bdev->driver->invalidate_caches(bdev, bo->mem.placement);
+			if (ret)
+				pr_err("Can not flush read caches\n");
+		}
 		bo->evicted = false;
 	}
 
@@ -410,7 +412,7 @@ static void ttm_bo_cleanup_refs_or_queue(struct ttm_buffer_object *bo)
 	int ret;
 
 	spin_lock(&glob->lru_lock);
-	ret = ttm_bo_reserve_nolru(bo, false, true, false, 0);
+	ret = __ttm_bo_reserve(bo, false, true, false, 0);
 
 	spin_lock(&bdev->fence_lock);
 	(void) ttm_bo_wait(bo, false, false, true);
@@ -441,7 +443,7 @@ static void ttm_bo_cleanup_refs_or_queue(struct ttm_buffer_object *bo)
 			ttm_bo_add_to_lru(bo);
 		}
 
-		ww_mutex_unlock(&bo->resv->lock);
+		__ttm_bo_unreserve(bo);
 	}
 
 	kref_get(&bo->list_kref);
@@ -492,7 +494,7 @@ static int ttm_bo_cleanup_refs_and_unlock(struct ttm_buffer_object *bo,
 		sync_obj = driver->sync_obj_ref(bo->sync_obj);
 		spin_unlock(&bdev->fence_lock);
 
-		ww_mutex_unlock(&bo->resv->lock);
+		__ttm_bo_unreserve(bo);
 		spin_unlock(&glob->lru_lock);
 
 		ret = driver->sync_obj_wait(sync_obj, false, interruptible);
@@ -512,7 +514,7 @@ static int ttm_bo_cleanup_refs_and_unlock(struct ttm_buffer_object *bo,
 			return ret;
 
 		spin_lock(&glob->lru_lock);
-		ret = ttm_bo_reserve_nolru(bo, false, true, false, 0);
+		ret = __ttm_bo_reserve(bo, false, true, false, 0);
 
 		/*
 		 * We raced, and lost, someone else holds the reservation now,
@@ -530,7 +532,7 @@ static int ttm_bo_cleanup_refs_and_unlock(struct ttm_buffer_object *bo,
 		spin_unlock(&bdev->fence_lock);
 
 	if (ret || unlikely(list_empty(&bo->ddestroy))) {
-		ww_mutex_unlock(&bo->resv->lock);
+		__ttm_bo_unreserve(bo);
 		spin_unlock(&glob->lru_lock);
 		return ret;
 	}
@@ -575,11 +577,11 @@ static int ttm_bo_delayed_delete(struct ttm_bo_device *bdev, bool remove_all)
 			kref_get(&nentry->list_kref);
 		}
 
-		ret = ttm_bo_reserve_nolru(entry, false, true, false, 0);
+		ret = __ttm_bo_reserve(entry, false, true, false, 0);
 		if (remove_all && ret) {
 			spin_unlock(&glob->lru_lock);
-			ret = ttm_bo_reserve_nolru(entry, false, false,
-						   false, 0);
+			ret = __ttm_bo_reserve(entry, false, false,
+					       false, 0);
 			spin_lock(&glob->lru_lock);
 		}
 
@@ -724,7 +726,7 @@ static int ttm_mem_evict_first(struct ttm_bo_device *bdev,
 
 	spin_lock(&glob->lru_lock);
 	list_for_each_entry(bo, &man->lru, lru) {
-		ret = ttm_bo_reserve_nolru(bo, false, true, false, 0);
+		ret = __ttm_bo_reserve(bo, false, true, false, 0);
 		if (!ret)
 			break;
 	}
@@ -957,7 +959,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 }
 EXPORT_SYMBOL(ttm_bo_mem_space);
 
-int ttm_bo_move_buffer(struct ttm_buffer_object *bo,
+static int ttm_bo_move_buffer(struct ttm_buffer_object *bo,
 			struct ttm_placement *placement,
 			bool interruptible,
 			bool no_wait_gpu)
@@ -1449,6 +1451,7 @@ EXPORT_SYMBOL(ttm_bo_device_release);
 int ttm_bo_device_init(struct ttm_bo_device *bdev,
 		       struct ttm_bo_global *glob,
 		       struct ttm_bo_driver *driver,
+		       struct address_space *mapping,
 		       uint64_t file_page_offset,
 		       bool need_dma32)
 {
@@ -1470,7 +1473,7 @@ int ttm_bo_device_init(struct ttm_bo_device *bdev,
 				    0x10000000);
 	INIT_DELAYED_WORK(&bdev->wq, ttm_bo_delayed_workqueue);
 	INIT_LIST_HEAD(&bdev->ddestroy);
-	bdev->dev_mapping = NULL;
+	bdev->dev_mapping = mapping;
 	bdev->glob = glob;
 	bdev->need_dma32 = need_dma32;
 	bdev->val_seq = 0;
@@ -1627,7 +1630,7 @@ static int ttm_bo_swapout(struct ttm_mem_shrink *shrink)
 
 	spin_lock(&glob->lru_lock);
 	list_for_each_entry(bo, &glob->swap_lru, swap) {
-		ret = ttm_bo_reserve_nolru(bo, false, true, false, 0);
+		ret = __ttm_bo_reserve(bo, false, true, false, 0);
 		if (!ret)
 			break;
 	}
@@ -1694,7 +1697,7 @@ out:
 	 * already swapped buffer.
 	 */
 
-	ww_mutex_unlock(&bo->resv->lock);
+	__ttm_bo_unreserve(bo);
 	kref_put(&bo->list_kref, ttm_bo_release_list);
 	return ret;
 }
@@ -1728,10 +1731,10 @@ int ttm_bo_wait_unreserved(struct ttm_buffer_object *bo)
 		return -ERESTARTSYS;
 	if (!ww_mutex_is_locked(&bo->resv->lock))
 		goto out_unlock;
-	ret = ttm_bo_reserve_nolru(bo, true, false, false, NULL);
+	ret = __ttm_bo_reserve(bo, true, false, false, NULL);
 	if (unlikely(ret != 0))
 		goto out_unlock;
-	ww_mutex_unlock(&bo->resv->lock);
+	__ttm_bo_unreserve(bo);
 
 out_unlock:
 	mutex_unlock(&bo->wu_mutex);
diff --git a/drivers/gpu/drm/ttm/ttm_bo_manager.c b/drivers/gpu/drm/ttm/ttm_bo_manager.c
index c58eba33bd5..bd850c9f4bc 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_manager.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_manager.c
@@ -55,6 +55,7 @@ static int ttm_bo_man_get_node(struct ttm_mem_type_manager *man,
 	struct ttm_range_manager *rman = (struct ttm_range_manager *) man->priv;
 	struct drm_mm *mm = &rman->mm;
 	struct drm_mm_node *node = NULL;
+	enum drm_mm_allocator_flags aflags = DRM_MM_CREATE_DEFAULT;
 	unsigned long lpfn;
 	int ret;
 
@@ -66,11 +67,15 @@ static int ttm_bo_man_get_node(struct ttm_mem_type_manager *man,
 	if (!node)
 		return -ENOMEM;
 
+	if (bo->mem.placement & TTM_PL_FLAG_TOPDOWN)
+		aflags = DRM_MM_CREATE_TOP;
+
 	spin_lock(&rman->lock);
-	ret = drm_mm_insert_node_in_range(mm, node, mem->num_pages,
-					  mem->page_alignment,
+	ret = drm_mm_insert_node_in_range_generic(mm, node, mem->num_pages,
+					  mem->page_alignment, 0,
 					  placement->fpfn, lpfn,
-					  DRM_MM_SEARCH_BEST);
+					  DRM_MM_SEARCH_BEST,
+					  aflags);
 	spin_unlock(&rman->lock);
 
 	if (unlikely(ret)) {
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index 40615215231..1df856f7856 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -187,7 +187,7 @@ void ttm_mem_io_free_vm(struct ttm_buffer_object *bo)
 	}
 }
 
-int ttm_mem_reg_ioremap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
+static int ttm_mem_reg_ioremap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
 			void **virtual)
 {
 	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
@@ -219,7 +219,7 @@ int ttm_mem_reg_ioremap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
 	return 0;
 }
 
-void ttm_mem_reg_iounmap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
+static void ttm_mem_reg_iounmap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
 			 void *virtual)
 {
 	struct ttm_mem_type_manager *man;
@@ -594,7 +594,7 @@ int ttm_bo_kmap(struct ttm_buffer_object *bo,
 	if (start_page > bo->num_pages)
 		return -EINVAL;
 #if 0
-	if (num_pages > 1 && !DRM_SUSER(DRM_CURPROC))
+	if (num_pages > 1 && !capable(CAP_SYS_ADMIN))
 		return -EPERM;
 #endif
 	(void) ttm_mem_io_lock(man, false);
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 6440eeac22d..0ce48e5a9cb 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -132,6 +132,15 @@ static int ttm_bo_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 		return VM_FAULT_NOPAGE;
 	}
 
+	/*
+	 * Refuse to fault imported pages. This should be handled
+	 * (if at all) by redirecting mmap to the exporter.
+	 */
+	if (bo->ttm && (bo->ttm->page_flags & TTM_PAGE_FLAG_SG)) {
+		retval = VM_FAULT_SIGBUS;
+		goto out_unlock;
+	}
+
 	if (bdev->driver->fault_reserve_notify) {
 		ret = bdev->driver->fault_reserve_notify(bo);
 		switch (ret) {
@@ -217,10 +226,17 @@ static int ttm_bo_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 			} else if (unlikely(!page)) {
 				break;
 			}
+			page->mapping = vma->vm_file->f_mapping;
+			page->index = drm_vma_node_start(&bo->vma_node) +
+				page_offset;
 			pfn = page_to_pfn(page);
 		}
 
-		ret = vm_insert_mixed(&cvma, address, pfn);
+		if (vma->vm_flags & VM_MIXEDMAP)
+			ret = vm_insert_mixed(&cvma, address, pfn);
+		else
+			ret = vm_insert_pfn(&cvma, address, pfn);
+
 		/*
 		 * Somebody beat us to this PTE or prefaulting to
 		 * an already populated PTE, or prefaulting error.
@@ -250,6 +266,8 @@ static void ttm_bo_vm_open(struct vm_area_struct *vma)
 	struct ttm_buffer_object *bo =
 	    (struct ttm_buffer_object *)vma->vm_private_data;
 
+	WARN_ON(bo->bdev->dev_mapping != vma->vm_file->f_mapping);
+
 	(void)ttm_bo_reference(bo);
 }
 
@@ -319,7 +337,16 @@ int ttm_bo_mmap(struct file *filp, struct vm_area_struct *vma,
 	 */
 
 	vma->vm_private_data = bo;
-	vma->vm_flags |= VM_IO | VM_MIXEDMAP | VM_DONTEXPAND | VM_DONTDUMP;
+
+	/*
+	 * We'd like to use VM_PFNMAP on shared mappings, where
+	 * (vma->vm_flags & VM_SHARED) != 0, for performance reasons,
+	 * but for some reason VM_PFNMAP + x86 PAT + write-combine is very
+	 * bad for performance. Until that has been sorted out, use
+	 * VM_MIXEDMAP on all mappings. See freedesktop.org bug #75719
+	 */
+	vma->vm_flags |= VM_MIXEDMAP;
+	vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP;
 	return 0;
 out_unref:
 	ttm_bo_unref(&bo);
@@ -334,7 +361,8 @@ int ttm_fbdev_mmap(struct vm_area_struct *vma, struct ttm_buffer_object *bo)
 
 	vma->vm_ops = &ttm_bo_vm_ops;
 	vma->vm_private_data = ttm_bo_reference(bo);
-	vma->vm_flags |= VM_IO | VM_MIXEDMAP | VM_DONTEXPAND;
+	vma->vm_flags |= VM_MIXEDMAP;
+	vma->vm_flags |= VM_IO | VM_DONTEXPAND;
 	return 0;
 }
 EXPORT_SYMBOL(ttm_fbdev_mmap);
diff --git a/drivers/gpu/drm/ttm/ttm_execbuf_util.c b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
index 479e9418e3d..e8dac875852 100644
--- a/drivers/gpu/drm/ttm/ttm_execbuf_util.c
+++ b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
@@ -46,7 +46,7 @@ static void ttm_eu_backoff_reservation_locked(struct list_head *list)
 			ttm_bo_add_to_lru(bo);
 			entry->removed = false;
 		}
-		ww_mutex_unlock(&bo->resv->lock);
+		__ttm_bo_unreserve(bo);
 	}
 }
 
@@ -140,8 +140,8 @@ retry:
 		if (entry->reserved)
 			continue;
 
-		ret = ttm_bo_reserve_nolru(bo, true, (ticket == NULL), true,
-					   ticket);
+		ret = __ttm_bo_reserve(bo, true, (ticket == NULL), true,
+				       ticket);
 
 		if (ret == -EDEADLK) {
 			/* uh oh, we lost out, drop every reservation and try
@@ -224,7 +224,7 @@ void ttm_eu_fence_buffer_objects(struct ww_acquire_ctx *ticket,
 		entry->old_sync_obj = bo->sync_obj;
 		bo->sync_obj = driver->sync_obj_ref(sync_obj);
 		ttm_bo_add_to_lru(bo);
-		ww_mutex_unlock(&bo->resv->lock);
+		__ttm_bo_unreserve(bo);
 		entry->reserved = false;
 	}
 	spin_unlock(&bdev->fence_lock);
diff --git a/drivers/gpu/drm/ttm/ttm_lock.c b/drivers/gpu/drm/ttm/ttm_lock.c
index 3daa9a3930b..6a954544727 100644
--- a/drivers/gpu/drm/ttm/ttm_lock.c
+++ b/drivers/gpu/drm/ttm/ttm_lock.c
@@ -186,14 +186,6 @@ int ttm_write_lock(struct ttm_lock *lock, bool interruptible)
 }
 EXPORT_SYMBOL(ttm_write_lock);
 
-void ttm_write_lock_downgrade(struct ttm_lock *lock)
-{
-	spin_lock(&lock->lock);
-	lock->rw = 1;
-	wake_up_all(&lock->queue);
-	spin_unlock(&lock->lock);
-}
-
 static int __ttm_vt_unlock(struct ttm_lock *lock)
 {
 	int ret = 0;
diff --git a/drivers/gpu/drm/ttm/ttm_object.c b/drivers/gpu/drm/ttm/ttm_object.c
index 6fe7b92a82d..d2a05335278 100644
--- a/drivers/gpu/drm/ttm/ttm_object.c
+++ b/drivers/gpu/drm/ttm/ttm_object.c
@@ -68,7 +68,7 @@
 
 struct ttm_object_file {
 	struct ttm_object_device *tdev;
-	rwlock_t lock;
+	spinlock_t lock;
 	struct list_head ref_list;
 	struct drm_open_hash ref_hash[TTM_REF_NUM];
 	struct kref refcount;
@@ -118,6 +118,7 @@ struct ttm_object_device {
  */
 
 struct ttm_ref_object {
+	struct rcu_head rcu_head;
 	struct drm_hash_item hash;
 	struct list_head head;
 	struct kref kref;
@@ -210,10 +211,9 @@ static void ttm_release_base(struct kref *kref)
 	 * call_rcu() or ttm_base_object_kfree().
 	 */
 
-	if (base->refcount_release) {
-		ttm_object_file_unref(&base->tfile);
+	ttm_object_file_unref(&base->tfile);
+	if (base->refcount_release)
 		base->refcount_release(&base);
-	}
 }
 
 void ttm_base_object_unref(struct ttm_base_object **p_base)
@@ -229,32 +229,92 @@ EXPORT_SYMBOL(ttm_base_object_unref);
 struct ttm_base_object *ttm_base_object_lookup(struct ttm_object_file *tfile,
 					       uint32_t key)
 {
-	struct ttm_object_device *tdev = tfile->tdev;
-	struct ttm_base_object *uninitialized_var(base);
+	struct ttm_base_object *base = NULL;
 	struct drm_hash_item *hash;
+	struct drm_open_hash *ht = &tfile->ref_hash[TTM_REF_USAGE];
 	int ret;
 
 	rcu_read_lock();
-	ret = drm_ht_find_item_rcu(&tdev->object_hash, key, &hash);
+	ret = drm_ht_find_item_rcu(ht, key, &hash);
 
 	if (likely(ret == 0)) {
-		base = drm_hash_entry(hash, struct ttm_base_object, hash);
-		ret = kref_get_unless_zero(&base->refcount) ? 0 : -EINVAL;
+		base = drm_hash_entry(hash, struct ttm_ref_object, hash)->obj;
+		if (!kref_get_unless_zero(&base->refcount))
+			base = NULL;
 	}
 	rcu_read_unlock();
 
-	if (unlikely(ret != 0))
-		return NULL;
+	return base;
+}
+EXPORT_SYMBOL(ttm_base_object_lookup);
 
-	if (tfile != base->tfile && !base->shareable) {
-		pr_err("Attempted access of non-shareable object\n");
-		ttm_base_object_unref(&base);
-		return NULL;
+struct ttm_base_object *
+ttm_base_object_lookup_for_ref(struct ttm_object_device *tdev, uint32_t key)
+{
+	struct ttm_base_object *base = NULL;
+	struct drm_hash_item *hash;
+	struct drm_open_hash *ht = &tdev->object_hash;
+	int ret;
+
+	rcu_read_lock();
+	ret = drm_ht_find_item_rcu(ht, key, &hash);
+
+	if (likely(ret == 0)) {
+		base = drm_hash_entry(hash, struct ttm_base_object, hash);
+		if (!kref_get_unless_zero(&base->refcount))
+			base = NULL;
 	}
+	rcu_read_unlock();
 
 	return base;
 }
-EXPORT_SYMBOL(ttm_base_object_lookup);
+EXPORT_SYMBOL(ttm_base_object_lookup_for_ref);
+
+/**
+ * ttm_ref_object_exists - Check whether a caller has a valid ref object
+ * (has opened) a base object.
+ *
+ * @tfile: Pointer to a struct ttm_object_file identifying the caller.
+ * @base: Pointer to a struct base object.
+ *
+ * Checks wether the caller identified by @tfile has put a valid USAGE
+ * reference object on the base object identified by @base.
+ */
+bool ttm_ref_object_exists(struct ttm_object_file *tfile,
+			   struct ttm_base_object *base)
+{
+	struct drm_open_hash *ht = &tfile->ref_hash[TTM_REF_USAGE];
+	struct drm_hash_item *hash;
+	struct ttm_ref_object *ref;
+
+	rcu_read_lock();
+	if (unlikely(drm_ht_find_item_rcu(ht, base->hash.key, &hash) != 0))
+		goto out_false;
+
+	/*
+	 * Verify that the ref object is really pointing to our base object.
+	 * Our base object could actually be dead, and the ref object pointing
+	 * to another base object with the same handle.
+	 */
+	ref = drm_hash_entry(hash, struct ttm_ref_object, hash);
+	if (unlikely(base != ref->obj))
+		goto out_false;
+
+	/*
+	 * Verify that the ref->obj pointer was actually valid!
+	 */
+	rmb();
+	if (unlikely(atomic_read(&ref->kref.refcount) == 0))
+		goto out_false;
+
+	rcu_read_unlock();
+	return true;
+
+ out_false:
+	rcu_read_unlock();
+	return false;
+}
+EXPORT_SYMBOL(ttm_ref_object_exists);
 
 int ttm_ref_object_add(struct ttm_object_file *tfile,
 		       struct ttm_base_object *base,
@@ -266,21 +326,25 @@ int ttm_ref_object_add(struct ttm_object_file *tfile,
 	struct ttm_mem_global *mem_glob = tfile->tdev->mem_glob;
 	int ret = -EINVAL;
 
+	if (base->tfile != tfile && !base->shareable)
+		return -EPERM;
+
 	if (existed != NULL)
 		*existed = true;
 
 	while (ret == -EINVAL) {
-		read_lock(&tfile->lock);
-		ret = drm_ht_find_item(ht, base->hash.key, &hash);
+		rcu_read_lock();
+		ret = drm_ht_find_item_rcu(ht, base->hash.key, &hash);
 
 		if (ret == 0) {
 			ref = drm_hash_entry(hash, struct ttm_ref_object, hash);
-			kref_get(&ref->kref);
-			read_unlock(&tfile->lock);
-			break;
+			if (kref_get_unless_zero(&ref->kref)) {
+				rcu_read_unlock();
+				break;
+			}
 		}
 
-		read_unlock(&tfile->lock);
+		rcu_read_unlock();
 		ret = ttm_mem_global_alloc(mem_glob, sizeof(*ref),
 					   false, false);
 		if (unlikely(ret != 0))
@@ -297,19 +361,19 @@ int ttm_ref_object_add(struct ttm_object_file *tfile,
 		ref->ref_type = ref_type;
 		kref_init(&ref->kref);
 
-		write_lock(&tfile->lock);
-		ret = drm_ht_insert_item(ht, &ref->hash);
+		spin_lock(&tfile->lock);
+		ret = drm_ht_insert_item_rcu(ht, &ref->hash);
 
 		if (likely(ret == 0)) {
 			list_add_tail(&ref->head, &tfile->ref_list);
 			kref_get(&base->refcount);
-			write_unlock(&tfile->lock);
+			spin_unlock(&tfile->lock);
 			if (existed != NULL)
 				*existed = false;
 			break;
 		}
 
-		write_unlock(&tfile->lock);
+		spin_unlock(&tfile->lock);
 		BUG_ON(ret != -EINVAL);
 
 		ttm_mem_global_free(mem_glob, sizeof(*ref));
@@ -330,17 +394,17 @@ static void ttm_ref_object_release(struct kref *kref)
 	struct ttm_mem_global *mem_glob = tfile->tdev->mem_glob;
 
 	ht = &tfile->ref_hash[ref->ref_type];
-	(void)drm_ht_remove_item(ht, &ref->hash);
+	(void)drm_ht_remove_item_rcu(ht, &ref->hash);
 	list_del(&ref->head);
-	write_unlock(&tfile->lock);
+	spin_unlock(&tfile->lock);
 
 	if (ref->ref_type != TTM_REF_USAGE && base->ref_obj_release)
 		base->ref_obj_release(base, ref->ref_type);
 
 	ttm_base_object_unref(&ref->obj);
 	ttm_mem_global_free(mem_glob, sizeof(*ref));
-	kfree(ref);
-	write_lock(&tfile->lock);
+	kfree_rcu(ref, rcu_head);
+	spin_lock(&tfile->lock);
 }
 
 int ttm_ref_object_base_unref(struct ttm_object_file *tfile,
@@ -351,15 +415,15 @@ int ttm_ref_object_base_unref(struct ttm_object_file *tfile,
 	struct drm_hash_item *hash;
 	int ret;
 
-	write_lock(&tfile->lock);
+	spin_lock(&tfile->lock);
 	ret = drm_ht_find_item(ht, key, &hash);
 	if (unlikely(ret != 0)) {
-		write_unlock(&tfile->lock);
+		spin_unlock(&tfile->lock);
 		return -EINVAL;
 	}
 	ref = drm_hash_entry(hash, struct ttm_ref_object, hash);
 	kref_put(&ref->kref, ttm_ref_object_release);
-	write_unlock(&tfile->lock);
+	spin_unlock(&tfile->lock);
 	return 0;
 }
 EXPORT_SYMBOL(ttm_ref_object_base_unref);
@@ -372,7 +436,7 @@ void ttm_object_file_release(struct ttm_object_file **p_tfile)
 	struct ttm_object_file *tfile = *p_tfile;
 
 	*p_tfile = NULL;
-	write_lock(&tfile->lock);
+	spin_lock(&tfile->lock);
 
 	/*
 	 * Since we release the lock within the loop, we have to
@@ -388,7 +452,7 @@ void ttm_object_file_release(struct ttm_object_file **p_tfile)
 	for (i = 0; i < TTM_REF_NUM; ++i)
 		drm_ht_remove(&tfile->ref_hash[i]);
 
-	write_unlock(&tfile->lock);
+	spin_unlock(&tfile->lock);
 	ttm_object_file_unref(&tfile);
 }
 EXPORT_SYMBOL(ttm_object_file_release);
@@ -404,7 +468,7 @@ struct ttm_object_file *ttm_object_file_init(struct ttm_object_device *tdev,
 	if (unlikely(tfile == NULL))
 		return NULL;
 
-	rwlock_init(&tfile->lock);
+	spin_lock_init(&tfile->lock);
 	tfile->tdev = tdev;
 	kref_init(&tfile->refcount);
 	INIT_LIST_HEAD(&tfile->ref_list);
diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c
index 210d5036516..75f31909004 100644
--- a/drivers/gpu/drm/ttm/ttm_tt.c
+++ b/drivers/gpu/drm/ttm/ttm_tt.c
@@ -170,9 +170,8 @@ void ttm_tt_destroy(struct ttm_tt *ttm)
 		ttm_tt_unbind(ttm);
 	}
 
-	if (ttm->state == tt_unbound) {
-		ttm->bdev->driver->ttm_tt_unpopulate(ttm);
-	}
+	if (ttm->state == tt_unbound)
+		ttm_tt_unpopulate(ttm);
 
 	if (!(ttm->page_flags & TTM_PAGE_FLAG_PERSISTENT_SWAP) &&
 	    ttm->swap_storage)
@@ -362,7 +361,7 @@ int ttm_tt_swapout(struct ttm_tt *ttm, struct file *persistent_swap_storage)
 		page_cache_release(to_page);
 	}
 
-	ttm->bdev->driver->ttm_tt_unpopulate(ttm);
+	ttm_tt_unpopulate(ttm);
 	ttm->swap_storage = swap_storage;
 	ttm->page_flags |= TTM_PAGE_FLAG_SWAPPED;
 	if (persistent_swap_storage)
@@ -375,3 +374,26 @@ out_err:
 
 	return ret;
 }
+
+static void ttm_tt_clear_mapping(struct ttm_tt *ttm)
+{
+	pgoff_t i;
+	struct page **page = ttm->pages;
+
+	if (ttm->page_flags & TTM_PAGE_FLAG_SG)
+		return;
+
+	for (i = 0; i < ttm->num_pages; ++i) {
+		(*page)->mapping = NULL;
+		(*page++)->index = 0;
+	}
+}
+
+void ttm_tt_unpopulate(struct ttm_tt *ttm)
+{
+	if (ttm->state == tt_unpopulated)
+		return;
+
+	ttm_tt_clear_mapping(ttm);
+	ttm->bdev->driver->ttm_tt_unpopulate(ttm);
+}
diff --git a/drivers/gpu/drm/udl/udl_fb.c b/drivers/gpu/drm/udl/udl_fb.c
index 97e9d614700..377176372da 100644
--- a/drivers/gpu/drm/udl/udl_fb.c
+++ b/drivers/gpu/drm/udl/udl_fb.c
@@ -403,15 +403,17 @@ static int udl_user_framebuffer_dirty(struct drm_framebuffer *fb,
 	int i;
 	int ret = 0;
 
+	drm_modeset_lock_all(fb->dev);
+
 	if (!ufb->active_16)
-		return 0;
+		goto unlock;
 
 	if (ufb->obj->base.import_attach) {
 		ret = dma_buf_begin_cpu_access(ufb->obj->base.import_attach->dmabuf,
 					       0, ufb->obj->base.size,
 					       DMA_FROM_DEVICE);
 		if (ret)
-			return ret;
+			goto unlock;
 	}
 
 	for (i = 0; i < num_clips; i++) {
@@ -427,6 +429,10 @@ static int udl_user_framebuffer_dirty(struct drm_framebuffer *fb,
 				       0, ufb->obj->base.size,
 				       DMA_FROM_DEVICE);
 	}
+
+ unlock:
+	drm_modeset_unlock_all(fb->dev);
+
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/udl/udl_gem.c b/drivers/gpu/drm/udl/udl_gem.c
index 8d67b943ac0..c041cd73f39 100644
--- a/drivers/gpu/drm/udl/udl_gem.c
+++ b/drivers/gpu/drm/udl/udl_gem.c
@@ -60,7 +60,7 @@ int udl_dumb_create(struct drm_file *file,
 		    struct drm_device *dev,
 		    struct drm_mode_create_dumb *args)
 {
-	args->pitch = args->width * ((args->bpp + 1) / 8);
+	args->pitch = args->width * DIV_ROUND_UP(args->bpp, 8);
 	args->size = args->pitch * args->height;
 	return udl_gem_create(file, dev,
 			      args->size, &args->handle);
@@ -177,8 +177,10 @@ void udl_gem_free_object(struct drm_gem_object *gem_obj)
 	if (obj->vmapping)
 		udl_gem_vunmap(obj);
 
-	if (gem_obj->import_attach)
+	if (gem_obj->import_attach) {
 		drm_prime_gem_destroy(gem_obj, obj->sg);
+		put_device(gem_obj->dev->dev);
+	}
 
 	if (obj->pages)
 		udl_gem_put_pages(obj);
@@ -256,9 +258,12 @@ struct drm_gem_object *udl_gem_prime_import(struct drm_device *dev,
 	int ret;
 
 	/* need to attach */
+	get_device(dev->dev);
 	attach = dma_buf_attach(dma_buf, dev->dev);
-	if (IS_ERR(attach))
+	if (IS_ERR(attach)) {
+		put_device(dev->dev);
 		return ERR_CAST(attach);
+	}
 
 	get_dma_buf(dma_buf);
 
@@ -282,6 +287,6 @@ fail_unmap:
 fail_detach:
 	dma_buf_detach(dma_buf, attach);
 	dma_buf_put(dma_buf);
-
+	put_device(dev->dev);
 	return ERR_PTR(ret);
 }
diff --git a/drivers/gpu/drm/udl/udl_main.c b/drivers/gpu/drm/udl/udl_main.c
index f5ae57406f3..7094b92d1ec 100644
--- a/drivers/gpu/drm/udl/udl_main.c
+++ b/drivers/gpu/drm/udl/udl_main.c
@@ -283,7 +283,7 @@ int udl_submit_urb(struct drm_device *dev, struct urb *urb, size_t len)
 int udl_driver_load(struct drm_device *dev, unsigned long flags)
 {
 	struct udl_device *udl;
-	int ret;
+	int ret = -ENOMEM;
 
 	DRM_DEBUG("\n");
 	udl = kzalloc(sizeof(struct udl_device), GFP_KERNEL);
@@ -294,12 +294,12 @@ int udl_driver_load(struct drm_device *dev, unsigned long flags)
 	dev->dev_private = udl;
 
 	if (!udl_parse_vendor_descriptor(dev, dev->usbdev)) {
+		ret = -ENODEV;
 		DRM_ERROR("firmware not recognized. Assume incompatible device\n");
 		goto err;
 	}
 
 	if (!udl_alloc_urb_list(dev, WRITES_IN_FLIGHT, MAX_TRANSFER)) {
-		ret = -ENOMEM;
 		DRM_ERROR("udl_alloc_urb_list failed\n");
 		goto err;
 	}
diff --git a/drivers/gpu/drm/udl/udl_modeset.c b/drivers/gpu/drm/udl/udl_modeset.c
index 2ae1eb7d163..cddc4fcf35c 100644
--- a/drivers/gpu/drm/udl/udl_modeset.c
+++ b/drivers/gpu/drm/udl/udl_modeset.c
@@ -310,7 +310,7 @@ static int udl_crtc_mode_set(struct drm_crtc *crtc,
 
 {
 	struct drm_device *dev = crtc->dev;
-	struct udl_framebuffer *ufb = to_udl_fb(crtc->fb);
+	struct udl_framebuffer *ufb = to_udl_fb(crtc->primary->fb);
 	struct udl_device *udl = dev->dev_private;
 	char *buf;
 	char *wrptr;
diff --git a/drivers/gpu/drm/via/via_dma.c b/drivers/gpu/drm/via/via_dma.c
index 652f9b43ec9..6fc0648dd37 100644
--- a/drivers/gpu/drm/via/via_dma.c
+++ b/drivers/gpu/drm/via/via_dma.c
@@ -60,7 +60,7 @@
 	dev_priv->dma_low += 8;					\
 }
 
-#define via_flush_write_combine() DRM_MEMORYBARRIER()
+#define via_flush_write_combine() mb()
 
 #define VIA_OUT_RING_QW(w1, w2)	do {		\
 	*vb++ = (w1);				\
@@ -234,13 +234,13 @@ static int via_dma_init(struct drm_device *dev, void *data, struct drm_file *fil
 
 	switch (init->func) {
 	case VIA_INIT_DMA:
-		if (!DRM_SUSER(DRM_CURPROC))
+		if (!capable(CAP_SYS_ADMIN))
 			retcode = -EPERM;
 		else
 			retcode = via_initialize(dev, dev_priv, init);
 		break;
 	case VIA_CLEANUP_DMA:
-		if (!DRM_SUSER(DRM_CURPROC))
+		if (!capable(CAP_SYS_ADMIN))
 			retcode = -EPERM;
 		else
 			retcode = via_dma_cleanup(dev);
@@ -273,7 +273,7 @@ static int via_dispatch_cmdbuffer(struct drm_device *dev, drm_via_cmdbuffer_t *c
 	if (cmd->size > VIA_PCI_BUF_SIZE)
 		return -ENOMEM;
 
-	if (DRM_COPY_FROM_USER(dev_priv->pci_buf, cmd->buf, cmd->size))
+	if (copy_from_user(dev_priv->pci_buf, cmd->buf, cmd->size))
 		return -EFAULT;
 
 	/*
@@ -346,7 +346,7 @@ static int via_dispatch_pci_cmdbuffer(struct drm_device *dev,
 
 	if (cmd->size > VIA_PCI_BUF_SIZE)
 		return -ENOMEM;
-	if (DRM_COPY_FROM_USER(dev_priv->pci_buf, cmd->buf, cmd->size))
+	if (copy_from_user(dev_priv->pci_buf, cmd->buf, cmd->size))
 		return -EFAULT;
 
 	if ((ret =
@@ -543,7 +543,7 @@ static void via_cmdbuf_start(drm_via_private_t *dev_priv)
 
 	VIA_WRITE(VIA_REG_TRANSPACE, pause_addr_hi);
 	VIA_WRITE(VIA_REG_TRANSPACE, pause_addr_lo);
-	DRM_WRITEMEMORYBARRIER();
+	wmb();
 	VIA_WRITE(VIA_REG_TRANSPACE, command | HC_HAGPCMNT_MASK);
 	VIA_READ(VIA_REG_TRANSPACE);
 
@@ -737,4 +737,4 @@ const struct drm_ioctl_desc via_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(VIA_BLIT_SYNC, via_dma_blit_sync, DRM_AUTH)
 };
 
-int via_max_ioctl = DRM_ARRAY_SIZE(via_ioctls);
+int via_max_ioctl = ARRAY_SIZE(via_ioctls);
diff --git a/drivers/gpu/drm/via/via_dmablit.c b/drivers/gpu/drm/via/via_dmablit.c
index 8b0f25904e6..ba33cf67918 100644
--- a/drivers/gpu/drm/via/via_dmablit.c
+++ b/drivers/gpu/drm/via/via_dmablit.c
@@ -217,7 +217,7 @@ via_fire_dmablit(struct drm_device *dev, drm_via_sg_info_t *vsg, int engine)
 	VIA_WRITE(VIA_PCI_DMA_MR0  + engine*0x04, VIA_DMA_MR_CM | VIA_DMA_MR_TDIE);
 	VIA_WRITE(VIA_PCI_DMA_BCR0 + engine*0x10, 0);
 	VIA_WRITE(VIA_PCI_DMA_DPR0 + engine*0x10, vsg->chain_start);
-	DRM_WRITEMEMORYBARRIER();
+	wmb();
 	VIA_WRITE(VIA_PCI_DMA_CSR0 + engine*0x04, VIA_DMA_CSR_DE | VIA_DMA_CSR_TS);
 	VIA_READ(VIA_PCI_DMA_CSR0 + engine*0x04);
 }
@@ -338,7 +338,7 @@ via_dmablit_handler(struct drm_device *dev, int engine, int from_irq)
 
 		blitq->blits[cur]->aborted = blitq->aborting;
 		blitq->done_blit_handle++;
-		DRM_WAKEUP(blitq->blit_queue + cur);
+		wake_up(blitq->blit_queue + cur);
 
 		cur++;
 		if (cur >= VIA_NUM_BLIT_SLOTS)
@@ -363,7 +363,7 @@ via_dmablit_handler(struct drm_device *dev, int engine, int from_irq)
 
 		via_abort_dmablit(dev, engine);
 		blitq->aborting = 1;
-		blitq->end = jiffies + DRM_HZ;
+		blitq->end = jiffies + HZ;
 	}
 
 	if (!blitq->is_active) {
@@ -372,7 +372,7 @@ via_dmablit_handler(struct drm_device *dev, int engine, int from_irq)
 			blitq->is_active = 1;
 			blitq->cur = cur;
 			blitq->num_outstanding--;
-			blitq->end = jiffies + DRM_HZ;
+			blitq->end = jiffies + HZ;
 			if (!timer_pending(&blitq->poll_timer))
 				mod_timer(&blitq->poll_timer, jiffies + 1);
 		} else {
@@ -436,7 +436,7 @@ via_dmablit_sync(struct drm_device *dev, uint32_t handle, int engine)
 	int ret = 0;
 
 	if (via_dmablit_active(blitq, engine, handle, &queue)) {
-		DRM_WAIT_ON(ret, *queue, 3 * DRM_HZ,
+		DRM_WAIT_ON(ret, *queue, 3 * HZ,
 			    !via_dmablit_active(blitq, engine, handle, NULL));
 	}
 	DRM_DEBUG("DMA blit sync handle 0x%x engine %d returned %d\n",
@@ -521,7 +521,7 @@ via_dmablit_workqueue(struct work_struct *work)
 
 		spin_unlock_irqrestore(&blitq->blit_lock, irqsave);
 
-		DRM_WAKEUP(&blitq->busy_queue);
+		wake_up(&blitq->busy_queue);
 
 		via_free_sg_info(dev->pdev, cur_sg);
 		kfree(cur_sg);
@@ -561,8 +561,8 @@ via_init_dmablit(struct drm_device *dev)
 		blitq->aborting = 0;
 		spin_lock_init(&blitq->blit_lock);
 		for (j = 0; j < VIA_NUM_BLIT_SLOTS; ++j)
-			DRM_INIT_WAITQUEUE(blitq->blit_queue + j);
-		DRM_INIT_WAITQUEUE(&blitq->busy_queue);
+			init_waitqueue_head(blitq->blit_queue + j);
+		init_waitqueue_head(&blitq->busy_queue);
 		INIT_WORK(&blitq->wq, via_dmablit_workqueue);
 		setup_timer(&blitq->poll_timer, via_dmablit_timer,
 				(unsigned long)blitq);
@@ -688,7 +688,7 @@ via_dmablit_grab_slot(drm_via_blitq_t *blitq, int engine)
 	while (blitq->num_free == 0) {
 		spin_unlock_irqrestore(&blitq->blit_lock, irqsave);
 
-		DRM_WAIT_ON(ret, blitq->busy_queue, DRM_HZ, blitq->num_free > 0);
+		DRM_WAIT_ON(ret, blitq->busy_queue, HZ, blitq->num_free > 0);
 		if (ret)
 			return (-EINTR == ret) ? -EAGAIN : ret;
 
@@ -713,7 +713,7 @@ via_dmablit_release_slot(drm_via_blitq_t *blitq)
 	spin_lock_irqsave(&blitq->blit_lock, irqsave);
 	blitq->num_free++;
 	spin_unlock_irqrestore(&blitq->blit_lock, irqsave);
-	DRM_WAKEUP(&blitq->busy_queue);
+	wake_up(&blitq->busy_queue);
 }
 
 /*
diff --git a/drivers/gpu/drm/via/via_drv.c b/drivers/gpu/drm/via/via_drv.c
index 92684a9b7e3..50abc2adfae 100644
--- a/drivers/gpu/drm/via/via_drv.c
+++ b/drivers/gpu/drm/via/via_drv.c
@@ -46,7 +46,7 @@ static int via_driver_open(struct drm_device *dev, struct drm_file *file)
 	return 0;
 }
 
-void via_driver_postclose(struct drm_device *dev, struct drm_file *file)
+static void via_driver_postclose(struct drm_device *dev, struct drm_file *file)
 {
 	struct via_file_private *file_priv = file->driver_priv;
 
diff --git a/drivers/gpu/drm/via/via_drv.h b/drivers/gpu/drm/via/via_drv.h
index a811ef2b505..ad0273256be 100644
--- a/drivers/gpu/drm/via/via_drv.h
+++ b/drivers/gpu/drm/via/via_drv.h
@@ -138,7 +138,7 @@ extern u32 via_get_vblank_counter(struct drm_device *dev, int crtc);
 extern int via_enable_vblank(struct drm_device *dev, int crtc);
 extern void via_disable_vblank(struct drm_device *dev, int crtc);
 
-extern irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t via_driver_irq_handler(int irq, void *arg);
 extern void via_driver_irq_preinstall(struct drm_device *dev);
 extern int via_driver_irq_postinstall(struct drm_device *dev);
 extern void via_driver_irq_uninstall(struct drm_device *dev);
diff --git a/drivers/gpu/drm/via/via_irq.c b/drivers/gpu/drm/via/via_irq.c
index ac98964297c..1319433816d 100644
--- a/drivers/gpu/drm/via/via_irq.c
+++ b/drivers/gpu/drm/via/via_irq.c
@@ -104,7 +104,7 @@ u32 via_get_vblank_counter(struct drm_device *dev, int crtc)
 	return atomic_read(&dev_priv->vbl_received);
 }
 
-irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t via_driver_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	drm_via_private_t *dev_priv = (drm_via_private_t *) dev->dev_private;
@@ -138,7 +138,7 @@ irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS)
 	for (i = 0; i < dev_priv->num_irqs; ++i) {
 		if (status & cur_irq->pending_mask) {
 			atomic_inc(&cur_irq->irq_received);
-			DRM_WAKEUP(&cur_irq->irq_queue);
+			wake_up(&cur_irq->irq_queue);
 			handled = 1;
 			if (dev_priv->irq_map[drm_via_irq_dma0_td] == i)
 				via_dmablit_handler(dev, 0, 1);
@@ -239,12 +239,12 @@ via_driver_irq_wait(struct drm_device *dev, unsigned int irq, int force_sequence
 	cur_irq = dev_priv->via_irqs + real_irq;
 
 	if (masks[real_irq][2] && !force_sequence) {
-		DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * DRM_HZ,
+		DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * HZ,
 			    ((VIA_READ(masks[irq][2]) & masks[irq][3]) ==
 			     masks[irq][4]));
 		cur_irq_sequence = atomic_read(&cur_irq->irq_received);
 	} else {
-		DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * DRM_HZ,
+		DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * HZ,
 			    (((cur_irq_sequence =
 			       atomic_read(&cur_irq->irq_received)) -
 			      *sequence) <= (1 << 23)));
@@ -287,7 +287,7 @@ void via_driver_irq_preinstall(struct drm_device *dev)
 			atomic_set(&cur_irq->irq_received, 0);
 			cur_irq->enable_mask = dev_priv->irq_masks[i][0];
 			cur_irq->pending_mask = dev_priv->irq_masks[i][1];
-			DRM_INIT_WAITQUEUE(&cur_irq->irq_queue);
+			init_waitqueue_head(&cur_irq->irq_queue);
 			dev_priv->irq_enable_mask |= cur_irq->enable_mask;
 			dev_priv->irq_pending_mask |= cur_irq->pending_mask;
 			cur_irq++;
diff --git a/drivers/gpu/drm/via/via_mm.c b/drivers/gpu/drm/via/via_mm.c
index 92788910548..d70b1e1544b 100644
--- a/drivers/gpu/drm/via/via_mm.c
+++ b/drivers/gpu/drm/via/via_mm.c
@@ -79,7 +79,7 @@ int via_final_context(struct drm_device *dev, int context)
 
 	/* Linux specific until context tracking code gets ported to BSD */
 	/* Last context, perform cleanup */
-	if (list_is_singular(&dev->ctxlist) && dev->dev_private) {
+	if (list_is_singular(&dev->ctxlist)) {
 		DRM_DEBUG("Last Context\n");
 		drm_irq_uninstall(dev);
 		via_cleanup_futex(dev_priv);
diff --git a/drivers/gpu/drm/via/via_video.c b/drivers/gpu/drm/via/via_video.c
index 6569efa2ff6..a9ffbad1cfd 100644
--- a/drivers/gpu/drm/via/via_video.c
+++ b/drivers/gpu/drm/via/via_video.c
@@ -36,7 +36,7 @@ void via_init_futex(drm_via_private_t *dev_priv)
 	DRM_DEBUG("\n");
 
 	for (i = 0; i < VIA_NR_XVMC_LOCKS; ++i) {
-		DRM_INIT_WAITQUEUE(&(dev_priv->decoder_queue[i]));
+		init_waitqueue_head(&(dev_priv->decoder_queue[i]));
 		XVMCLOCKPTR(dev_priv->sarea_priv, i)->lock = 0;
 	}
 }
@@ -58,7 +58,7 @@ void via_release_futex(drm_via_private_t *dev_priv, int context)
 		if ((_DRM_LOCKING_CONTEXT(*lock) == context)) {
 			if (_DRM_LOCK_IS_HELD(*lock)
 			    && (*lock & _DRM_LOCK_CONT)) {
-				DRM_WAKEUP(&(dev_priv->decoder_queue[i]));
+				wake_up(&(dev_priv->decoder_queue[i]));
 			}
 			*lock = 0;
 		}
@@ -83,10 +83,10 @@ int via_decoder_futex(struct drm_device *dev, void *data, struct drm_file *file_
 	switch (fx->func) {
 	case VIA_FUTEX_WAIT:
 		DRM_WAIT_ON(ret, dev_priv->decoder_queue[fx->lock],
-			    (fx->ms / 10) * (DRM_HZ / 100), *lock != fx->val);
+			    (fx->ms / 10) * (HZ / 100), *lock != fx->val);
 		return ret;
 	case VIA_FUTEX_WAKE:
-		DRM_WAKEUP(&(dev_priv->decoder_queue[fx->lock]));
+		wake_up(&(dev_priv->decoder_queue[fx->lock]));
 		return 0;
 	}
 	return 0;
diff --git a/drivers/gpu/drm/vmwgfx/Kconfig b/drivers/gpu/drm/vmwgfx/Kconfig
index b71bcd0bfbb..67720f70fe2 100644
--- a/drivers/gpu/drm/vmwgfx/Kconfig
+++ b/drivers/gpu/drm/vmwgfx/Kconfig
@@ -1,11 +1,14 @@
 config DRM_VMWGFX
 	tristate "DRM driver for VMware Virtual GPU"
-	depends on DRM && PCI && FB
+	depends on DRM && PCI
 	select FB_DEFERRED_IO
 	select FB_CFB_FILLRECT
 	select FB_CFB_COPYAREA
 	select FB_CFB_IMAGEBLIT
 	select DRM_TTM
+	# Only needed for the transitional use of drm_crtc_init - can be removed
+	# again once vmwgfx sets up the primary plane itself.
+	select DRM_KMS_HELPER
 	help
 	  Choose this option if you would like to run 3D acceleration
 	  in a VMware virtual machine.
@@ -14,7 +17,7 @@ config DRM_VMWGFX
 	  The compiled module will be called "vmwgfx.ko".
 
 config DRM_VMWGFX_FBCON
-	depends on DRM_VMWGFX
+	depends on DRM_VMWGFX && FB
 	bool "Enable framebuffer console under vmwgfx by default"
 	help
 	   Choose this option if you are shipping a new vmwgfx
diff --git a/drivers/gpu/drm/vmwgfx/Makefile b/drivers/gpu/drm/vmwgfx/Makefile
index 9f8b690bcf5..458cdf6d81e 100644
--- a/drivers/gpu/drm/vmwgfx/Makefile
+++ b/drivers/gpu/drm/vmwgfx/Makefile
@@ -6,6 +6,6 @@ vmwgfx-y := vmwgfx_execbuf.o vmwgfx_gmr.o vmwgfx_kms.o vmwgfx_drv.o \
 	    vmwgfx_fifo.o vmwgfx_irq.o vmwgfx_ldu.o vmwgfx_ttm_glue.o \
 	    vmwgfx_overlay.o vmwgfx_marker.o vmwgfx_gmrid_manager.o \
 	    vmwgfx_fence.o vmwgfx_dmabuf.o vmwgfx_scrn.o vmwgfx_context.o \
-	    vmwgfx_surface.o vmwgfx_prime.o
+	    vmwgfx_surface.o vmwgfx_prime.o vmwgfx_mob.o vmwgfx_shader.o
 
 obj-$(CONFIG_DRM_VMWGFX) := vmwgfx.o
diff --git a/drivers/gpu/drm/vmwgfx/svga3d_reg.h b/drivers/gpu/drm/vmwgfx/svga3d_reg.h
index d0e085ee824..f58dc7dd15c 100644
--- a/drivers/gpu/drm/vmwgfx/svga3d_reg.h
+++ b/drivers/gpu/drm/vmwgfx/svga3d_reg.h
@@ -34,6 +34,8 @@
 
 #include "svga_reg.h"
 
+typedef uint32 PPN;
+typedef __le64 PPN64;
 
 /*
  * 3D Hardware Version
@@ -71,6 +73,9 @@ typedef uint32 SVGA3dBool; /* 32-bit Bool definition */
 #define SVGA3D_MAX_CONTEXT_IDS                  256
 #define SVGA3D_MAX_SURFACE_IDS                  (32 * 1024)
 
+#define SVGA3D_NUM_TEXTURE_UNITS                32
+#define SVGA3D_NUM_LIGHTS                       8
+
 /*
  * Surface formats.
  *
@@ -81,6 +86,7 @@ typedef uint32 SVGA3dBool; /* 32-bit Bool definition */
  */
 
 typedef enum SVGA3dSurfaceFormat {
+   SVGA3D_FORMAT_MIN                   = 0,
    SVGA3D_FORMAT_INVALID               = 0,
 
    SVGA3D_X8R8G8B8                     = 1,
@@ -134,12 +140,6 @@ typedef enum SVGA3dSurfaceFormat {
    SVGA3D_RG_S10E5                     = 35,
    SVGA3D_RG_S23E8                     = 36,
 
-   /*
-    * Any surface can be used as a buffer object, but SVGA3D_BUFFER is
-    * the most efficient format to use when creating new surfaces
-    * expressly for index or vertex data.
-    */
-
    SVGA3D_BUFFER                       = 37,
 
    SVGA3D_Z_D24X8                      = 38,
@@ -159,15 +159,109 @@ typedef enum SVGA3dSurfaceFormat {
    /* Video format with alpha */
    SVGA3D_AYUV                         = 45,
 
+   SVGA3D_R32G32B32A32_TYPELESS        = 46,
+   SVGA3D_R32G32B32A32_FLOAT           = 25,
+   SVGA3D_R32G32B32A32_UINT            = 47,
+   SVGA3D_R32G32B32A32_SINT            = 48,
+   SVGA3D_R32G32B32_TYPELESS           = 49,
+   SVGA3D_R32G32B32_FLOAT              = 50,
+   SVGA3D_R32G32B32_UINT               = 51,
+   SVGA3D_R32G32B32_SINT               = 52,
+   SVGA3D_R16G16B16A16_TYPELESS        = 53,
+   SVGA3D_R16G16B16A16_FLOAT           = 24,
+   SVGA3D_R16G16B16A16_UNORM           = 41,
+   SVGA3D_R16G16B16A16_UINT            = 54,
+   SVGA3D_R16G16B16A16_SNORM           = 55,
+   SVGA3D_R16G16B16A16_SINT            = 56,
+   SVGA3D_R32G32_TYPELESS              = 57,
+   SVGA3D_R32G32_FLOAT                 = 36,
+   SVGA3D_R32G32_UINT                  = 58,
+   SVGA3D_R32G32_SINT                  = 59,
+   SVGA3D_R32G8X24_TYPELESS            = 60,
+   SVGA3D_D32_FLOAT_S8X24_UINT         = 61,
+   SVGA3D_R32_FLOAT_X8X24_TYPELESS     = 62,
+   SVGA3D_X32_TYPELESS_G8X24_UINT      = 63,
+   SVGA3D_R10G10B10A2_TYPELESS         = 64,
+   SVGA3D_R10G10B10A2_UNORM            = 26,
+   SVGA3D_R10G10B10A2_UINT             = 65,
+   SVGA3D_R11G11B10_FLOAT              = 66,
+   SVGA3D_R8G8B8A8_TYPELESS            = 67,
+   SVGA3D_R8G8B8A8_UNORM               = 68,
+   SVGA3D_R8G8B8A8_UNORM_SRGB          = 69,
+   SVGA3D_R8G8B8A8_UINT                = 70,
+   SVGA3D_R8G8B8A8_SNORM               = 28,
+   SVGA3D_R8G8B8A8_SINT                = 71,
+   SVGA3D_R16G16_TYPELESS              = 72,
+   SVGA3D_R16G16_FLOAT                 = 35,
+   SVGA3D_R16G16_UNORM                 = 40,
+   SVGA3D_R16G16_UINT                  = 73,
+   SVGA3D_R16G16_SNORM                 = 39,
+   SVGA3D_R16G16_SINT                  = 74,
+   SVGA3D_R32_TYPELESS                 = 75,
+   SVGA3D_D32_FLOAT                    = 76,
+   SVGA3D_R32_FLOAT                    = 34,
+   SVGA3D_R32_UINT                     = 77,
+   SVGA3D_R32_SINT                     = 78,
+   SVGA3D_R24G8_TYPELESS               = 79,
+   SVGA3D_D24_UNORM_S8_UINT            = 80,
+   SVGA3D_R24_UNORM_X8_TYPELESS        = 81,
+   SVGA3D_X24_TYPELESS_G8_UINT         = 82,
+   SVGA3D_R8G8_TYPELESS                = 83,
+   SVGA3D_R8G8_UNORM                   = 84,
+   SVGA3D_R8G8_UINT                    = 85,
+   SVGA3D_R8G8_SNORM                   = 27,
+   SVGA3D_R8G8_SINT                    = 86,
+   SVGA3D_R16_TYPELESS                 = 87,
+   SVGA3D_R16_FLOAT                    = 33,
+   SVGA3D_D16_UNORM                    = 8,
+   SVGA3D_R16_UNORM                    = 88,
+   SVGA3D_R16_UINT                     = 89,
+   SVGA3D_R16_SNORM                    = 90,
+   SVGA3D_R16_SINT                     = 91,
+   SVGA3D_R8_TYPELESS                  = 92,
+   SVGA3D_R8_UNORM                     = 93,
+   SVGA3D_R8_UINT                      = 94,
+   SVGA3D_R8_SNORM                     = 95,
+   SVGA3D_R8_SINT                      = 96,
+   SVGA3D_A8_UNORM                     = 32,
+   SVGA3D_R1_UNORM                     = 97,
+   SVGA3D_R9G9B9E5_SHAREDEXP           = 98,
+   SVGA3D_R8G8_B8G8_UNORM              = 99,
+   SVGA3D_G8R8_G8B8_UNORM              = 100,
+   SVGA3D_BC1_TYPELESS                 = 101,
+   SVGA3D_BC1_UNORM                    = 15,
+   SVGA3D_BC1_UNORM_SRGB               = 102,
+   SVGA3D_BC2_TYPELESS                 = 103,
+   SVGA3D_BC2_UNORM                    = 17,
+   SVGA3D_BC2_UNORM_SRGB               = 104,
+   SVGA3D_BC3_TYPELESS                 = 105,
+   SVGA3D_BC3_UNORM                    = 19,
+   SVGA3D_BC3_UNORM_SRGB               = 106,
+   SVGA3D_BC4_TYPELESS                 = 107,
    SVGA3D_BC4_UNORM                    = 108,
+   SVGA3D_BC4_SNORM                    = 109,
+   SVGA3D_BC5_TYPELESS                 = 110,
    SVGA3D_BC5_UNORM                    = 111,
+   SVGA3D_BC5_SNORM                    = 112,
+   SVGA3D_B5G6R5_UNORM                 = 3,
+   SVGA3D_B5G5R5A1_UNORM               = 5,
+   SVGA3D_B8G8R8A8_UNORM               = 2,
+   SVGA3D_B8G8R8X8_UNORM               = 1,
+   SVGA3D_R10G10B10_XR_BIAS_A2_UNORM   = 113,
+   SVGA3D_B8G8R8A8_TYPELESS            = 114,
+   SVGA3D_B8G8R8A8_UNORM_SRGB          = 115,
+   SVGA3D_B8G8R8X8_TYPELESS            = 116,
+   SVGA3D_B8G8R8X8_UNORM_SRGB          = 117,
 
    /* Advanced D3D9 depth formats. */
    SVGA3D_Z_DF16                       = 118,
    SVGA3D_Z_DF24                       = 119,
    SVGA3D_Z_D24S8_INT                  = 120,
 
-   SVGA3D_FORMAT_MAX
+   /* Planar video formats. */
+   SVGA3D_YV12                         = 121,
+
+   SVGA3D_FORMAT_MAX                   = 122,
 } SVGA3dSurfaceFormat;
 
 typedef uint32 SVGA3dColor; /* a, r, g, b */
@@ -957,15 +1051,21 @@ typedef enum {
 } SVGA3dCubeFace;
 
 typedef enum {
+   SVGA3D_SHADERTYPE_INVALID                    = 0,
+   SVGA3D_SHADERTYPE_MIN                        = 1,
    SVGA3D_SHADERTYPE_VS                         = 1,
    SVGA3D_SHADERTYPE_PS                         = 2,
-   SVGA3D_SHADERTYPE_MAX
+   SVGA3D_SHADERTYPE_MAX                        = 3,
+   SVGA3D_SHADERTYPE_GS                         = 3,
 } SVGA3dShaderType;
 
+#define SVGA3D_NUM_SHADERTYPE (SVGA3D_SHADERTYPE_MAX - SVGA3D_SHADERTYPE_MIN)
+
 typedef enum {
    SVGA3D_CONST_TYPE_FLOAT                      = 0,
    SVGA3D_CONST_TYPE_INT                        = 1,
    SVGA3D_CONST_TYPE_BOOL                       = 2,
+   SVGA3D_CONST_TYPE_MAX
 } SVGA3dShaderConstType;
 
 #define SVGA3D_MAX_SURFACE_FACES                6
@@ -1056,9 +1156,84 @@ typedef enum {
 #define SVGA_3D_CMD_GENERATE_MIPMAPS       SVGA_3D_CMD_BASE + 31
 #define SVGA_3D_CMD_ACTIVATE_SURFACE       SVGA_3D_CMD_BASE + 40
 #define SVGA_3D_CMD_DEACTIVATE_SURFACE     SVGA_3D_CMD_BASE + 41
-#define SVGA_3D_CMD_MAX                    SVGA_3D_CMD_BASE + 42
-
-#define SVGA_3D_CMD_FUTURE_MAX             2000
+#define SVGA_3D_CMD_SCREEN_DMA               1082
+#define SVGA_3D_CMD_SET_UNITY_SURFACE_COOKIE 1083
+#define SVGA_3D_CMD_OPEN_CONTEXT_SURFACE     1084
+
+#define SVGA_3D_CMD_LOGICOPS_BITBLT          1085
+#define SVGA_3D_CMD_LOGICOPS_TRANSBLT        1086
+#define SVGA_3D_CMD_LOGICOPS_STRETCHBLT      1087
+#define SVGA_3D_CMD_LOGICOPS_COLORFILL       1088
+#define SVGA_3D_CMD_LOGICOPS_ALPHABLEND      1089
+#define SVGA_3D_CMD_LOGICOPS_CLEARTYPEBLEND  1090
+
+#define SVGA_3D_CMD_SET_OTABLE_BASE          1091
+#define SVGA_3D_CMD_READBACK_OTABLE          1092
+
+#define SVGA_3D_CMD_DEFINE_GB_MOB            1093
+#define SVGA_3D_CMD_DESTROY_GB_MOB           1094
+#define SVGA_3D_CMD_REDEFINE_GB_MOB          1095
+#define SVGA_3D_CMD_UPDATE_GB_MOB_MAPPING    1096
+
+#define SVGA_3D_CMD_DEFINE_GB_SURFACE        1097
+#define SVGA_3D_CMD_DESTROY_GB_SURFACE       1098
+#define SVGA_3D_CMD_BIND_GB_SURFACE          1099
+#define SVGA_3D_CMD_COND_BIND_GB_SURFACE     1100
+#define SVGA_3D_CMD_UPDATE_GB_IMAGE          1101
+#define SVGA_3D_CMD_UPDATE_GB_SURFACE        1102
+#define SVGA_3D_CMD_READBACK_GB_IMAGE        1103
+#define SVGA_3D_CMD_READBACK_GB_SURFACE      1104
+#define SVGA_3D_CMD_INVALIDATE_GB_IMAGE      1105
+#define SVGA_3D_CMD_INVALIDATE_GB_SURFACE    1106
+
+#define SVGA_3D_CMD_DEFINE_GB_CONTEXT        1107
+#define SVGA_3D_CMD_DESTROY_GB_CONTEXT       1108
+#define SVGA_3D_CMD_BIND_GB_CONTEXT          1109
+#define SVGA_3D_CMD_READBACK_GB_CONTEXT      1110
+#define SVGA_3D_CMD_INVALIDATE_GB_CONTEXT    1111
+
+#define SVGA_3D_CMD_DEFINE_GB_SHADER         1112
+#define SVGA_3D_CMD_DESTROY_GB_SHADER        1113
+#define SVGA_3D_CMD_BIND_GB_SHADER           1114
+
+#define SVGA_3D_CMD_SET_OTABLE_BASE64        1115
+
+#define SVGA_3D_CMD_BEGIN_GB_QUERY           1116
+#define SVGA_3D_CMD_END_GB_QUERY             1117
+#define SVGA_3D_CMD_WAIT_FOR_GB_QUERY        1118
+
+#define SVGA_3D_CMD_NOP                      1119
+
+#define SVGA_3D_CMD_ENABLE_GART              1120
+#define SVGA_3D_CMD_DISABLE_GART             1121
+#define SVGA_3D_CMD_MAP_MOB_INTO_GART        1122
+#define SVGA_3D_CMD_UNMAP_GART_RANGE         1123
+
+#define SVGA_3D_CMD_DEFINE_GB_SCREENTARGET   1124
+#define SVGA_3D_CMD_DESTROY_GB_SCREENTARGET  1125
+#define SVGA_3D_CMD_BIND_GB_SCREENTARGET     1126
+#define SVGA_3D_CMD_UPDATE_GB_SCREENTARGET   1127
+
+#define SVGA_3D_CMD_READBACK_GB_IMAGE_PARTIAL   1128
+#define SVGA_3D_CMD_INVALIDATE_GB_IMAGE_PARTIAL 1129
+
+#define SVGA_3D_CMD_SET_GB_SHADERCONSTS_INLINE  1130
+#define SVGA_3D_CMD_GB_SCREEN_DMA               1131
+#define SVGA_3D_CMD_BIND_GB_SURFACE_WITH_PITCH  1132
+#define SVGA_3D_CMD_GB_MOB_FENCE                1133
+#define SVGA_3D_CMD_DEFINE_GB_SURFACE_V2        1134
+#define SVGA_3D_CMD_DEFINE_GB_MOB64          1135
+#define SVGA_3D_CMD_REDEFINE_GB_MOB64        1136
+#define SVGA_3D_CMD_NOP_ERROR                1137
+
+#define SVGA_3D_CMD_RESERVED1                1138
+#define SVGA_3D_CMD_RESERVED2                1139
+#define SVGA_3D_CMD_RESERVED3                1140
+#define SVGA_3D_CMD_RESERVED4                1141
+#define SVGA_3D_CMD_RESERVED5                1142
+
+#define SVGA_3D_CMD_MAX                      1142
+#define SVGA_3D_CMD_FUTURE_MAX               3000
 
 /*
  * Common substructures used in multiple FIFO commands:
@@ -1750,6 +1925,507 @@ struct {
 
 
 /*
+ * Guest-backed surface definitions.
+ */
+
+typedef uint32 SVGAMobId;
+
+typedef enum SVGAMobFormat {
+   SVGA3D_MOBFMT_INVALID = SVGA3D_INVALID_ID,
+   SVGA3D_MOBFMT_PTDEPTH_0 = 0,
+   SVGA3D_MOBFMT_PTDEPTH_1 = 1,
+   SVGA3D_MOBFMT_PTDEPTH_2 = 2,
+   SVGA3D_MOBFMT_RANGE     = 3,
+   SVGA3D_MOBFMT_PTDEPTH64_0 = 4,
+   SVGA3D_MOBFMT_PTDEPTH64_1 = 5,
+   SVGA3D_MOBFMT_PTDEPTH64_2 = 6,
+   SVGA3D_MOBFMT_MAX,
+} SVGAMobFormat;
+
+/*
+ * Sizes of opaque types.
+ */
+
+#define SVGA3D_OTABLE_MOB_ENTRY_SIZE 16
+#define SVGA3D_OTABLE_CONTEXT_ENTRY_SIZE 8
+#define SVGA3D_OTABLE_SURFACE_ENTRY_SIZE 64
+#define SVGA3D_OTABLE_SHADER_ENTRY_SIZE 16
+#define SVGA3D_OTABLE_SCREEN_TARGET_ENTRY_SIZE 64
+#define SVGA3D_CONTEXT_DATA_SIZE 16384
+
+/*
+ * SVGA3dCmdSetOTableBase --
+ *
+ * This command allows the guest to specify the base PPN of the
+ * specified object table.
+ */
+
+typedef enum {
+   SVGA_OTABLE_MOB           = 0,
+   SVGA_OTABLE_MIN           = 0,
+   SVGA_OTABLE_SURFACE       = 1,
+   SVGA_OTABLE_CONTEXT       = 2,
+   SVGA_OTABLE_SHADER        = 3,
+   SVGA_OTABLE_SCREEN_TARGET = 4,
+   SVGA_OTABLE_DX9_MAX       = 5,
+   SVGA_OTABLE_MAX           = 8
+} SVGAOTableType;
+
+typedef
+struct {
+   SVGAOTableType type;
+   PPN baseAddress;
+   uint32 sizeInBytes;
+   uint32 validSizeInBytes;
+   SVGAMobFormat ptDepth;
+} __packed
+SVGA3dCmdSetOTableBase;  /* SVGA_3D_CMD_SET_OTABLE_BASE */
+
+typedef
+struct {
+   SVGAOTableType type;
+   PPN64 baseAddress;
+   uint32 sizeInBytes;
+   uint32 validSizeInBytes;
+   SVGAMobFormat ptDepth;
+} __packed
+SVGA3dCmdSetOTableBase64;  /* SVGA_3D_CMD_SET_OTABLE_BASE64 */
+
+typedef
+struct {
+   SVGAOTableType type;
+} __packed
+SVGA3dCmdReadbackOTable;  /* SVGA_3D_CMD_READBACK_OTABLE */
+
+/*
+ * Define a memory object (Mob) in the OTable.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBMob {
+   SVGAMobId mobid;
+   SVGAMobFormat ptDepth;
+   PPN base;
+   uint32 sizeInBytes;
+} __packed
+SVGA3dCmdDefineGBMob;   /* SVGA_3D_CMD_DEFINE_GB_MOB */
+
+
+/*
+ * Destroys an object in the OTable.
+ */
+
+typedef
+struct SVGA3dCmdDestroyGBMob {
+   SVGAMobId mobid;
+} __packed
+SVGA3dCmdDestroyGBMob;   /* SVGA_3D_CMD_DESTROY_GB_MOB */
+
+/*
+ * Redefine an object in the OTable.
+ */
+
+typedef
+struct SVGA3dCmdRedefineGBMob {
+   SVGAMobId mobid;
+   SVGAMobFormat ptDepth;
+   PPN base;
+   uint32 sizeInBytes;
+} __packed
+SVGA3dCmdRedefineGBMob;   /* SVGA_3D_CMD_REDEFINE_GB_MOB */
+
+/*
+ * Define a memory object (Mob) in the OTable with a PPN64 base.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBMob64 {
+   SVGAMobId mobid;
+   SVGAMobFormat ptDepth;
+   PPN64 base;
+   uint32 sizeInBytes;
+} __packed
+SVGA3dCmdDefineGBMob64;   /* SVGA_3D_CMD_DEFINE_GB_MOB64 */
+
+/*
+ * Redefine an object in the OTable with PPN64 base.
+ */
+
+typedef
+struct SVGA3dCmdRedefineGBMob64 {
+   SVGAMobId mobid;
+   SVGAMobFormat ptDepth;
+   PPN64 base;
+   uint32 sizeInBytes;
+} __packed
+SVGA3dCmdRedefineGBMob64;   /* SVGA_3D_CMD_REDEFINE_GB_MOB64 */
+
+/*
+ * Notification that the page tables have been modified.
+ */
+
+typedef
+struct SVGA3dCmdUpdateGBMobMapping {
+   SVGAMobId mobid;
+} __packed
+SVGA3dCmdUpdateGBMobMapping;   /* SVGA_3D_CMD_UPDATE_GB_MOB_MAPPING */
+
+/*
+ * Define a guest-backed surface.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBSurface {
+   uint32 sid;
+   SVGA3dSurfaceFlags surfaceFlags;
+   SVGA3dSurfaceFormat format;
+   uint32 numMipLevels;
+   uint32 multisampleCount;
+   SVGA3dTextureFilter autogenFilter;
+   SVGA3dSize size;
+} __packed
+SVGA3dCmdDefineGBSurface;   /* SVGA_3D_CMD_DEFINE_GB_SURFACE */
+
+/*
+ * Destroy a guest-backed surface.
+ */
+
+typedef
+struct SVGA3dCmdDestroyGBSurface {
+   uint32 sid;
+} __packed
+SVGA3dCmdDestroyGBSurface;   /* SVGA_3D_CMD_DESTROY_GB_SURFACE */
+
+/*
+ * Bind a guest-backed surface to an object.
+ */
+
+typedef
+struct SVGA3dCmdBindGBSurface {
+   uint32 sid;
+   SVGAMobId mobid;
+} __packed
+SVGA3dCmdBindGBSurface;   /* SVGA_3D_CMD_BIND_GB_SURFACE */
+
+/*
+ * Conditionally bind a mob to a guest backed surface if testMobid
+ * matches the currently bound mob.  Optionally issue a readback on
+ * the surface while it is still bound to the old mobid if the mobid
+ * is changed by this command.
+ */
+
+#define SVGA3D_COND_BIND_GB_SURFACE_FLAG_READBACK (1 << 0)
+
+typedef
+struct{
+   uint32 sid;
+   SVGAMobId testMobid;
+   SVGAMobId mobid;
+   uint32 flags;
+} __packed
+SVGA3dCmdCondBindGBSurface;          /* SVGA_3D_CMD_COND_BIND_GB_SURFACE */
+
+/*
+ * Update an image in a guest-backed surface.
+ * (Inform the device that the guest-contents have been updated.)
+ */
+
+typedef
+struct SVGA3dCmdUpdateGBImage {
+   SVGA3dSurfaceImageId image;
+   SVGA3dBox box;
+} __packed
+SVGA3dCmdUpdateGBImage;   /* SVGA_3D_CMD_UPDATE_GB_IMAGE */
+
+/*
+ * Update an entire guest-backed surface.
+ * (Inform the device that the guest-contents have been updated.)
+ */
+
+typedef
+struct SVGA3dCmdUpdateGBSurface {
+   uint32 sid;
+} __packed
+SVGA3dCmdUpdateGBSurface;   /* SVGA_3D_CMD_UPDATE_GB_SURFACE */
+
+/*
+ * Readback an image in a guest-backed surface.
+ * (Request the device to flush the dirty contents into the guest.)
+ */
+
+typedef
+struct SVGA3dCmdReadbackGBImage {
+   SVGA3dSurfaceImageId image;
+} __packed
+SVGA3dCmdReadbackGBImage;   /* SVGA_3D_CMD_READBACK_GB_IMAGE*/
+
+/*
+ * Readback an entire guest-backed surface.
+ * (Request the device to flush the dirty contents into the guest.)
+ */
+
+typedef
+struct SVGA3dCmdReadbackGBSurface {
+   uint32 sid;
+} __packed
+SVGA3dCmdReadbackGBSurface;   /* SVGA_3D_CMD_READBACK_GB_SURFACE */
+
+/*
+ * Readback a sub rect of an image in a guest-backed surface.  After
+ * issuing this command the driver is required to issue an update call
+ * of the same region before issuing any other commands that reference
+ * this surface or rendering is not guaranteed.
+ */
+
+typedef
+struct SVGA3dCmdReadbackGBImagePartial {
+   SVGA3dSurfaceImageId image;
+   SVGA3dBox box;
+   uint32 invertBox;
+} __packed
+SVGA3dCmdReadbackGBImagePartial; /* SVGA_3D_CMD_READBACK_GB_IMAGE_PARTIAL */
+
+/*
+ * Invalidate an image in a guest-backed surface.
+ * (Notify the device that the contents can be lost.)
+ */
+
+typedef
+struct SVGA3dCmdInvalidateGBImage {
+   SVGA3dSurfaceImageId image;
+} __packed
+SVGA3dCmdInvalidateGBImage;   /* SVGA_3D_CMD_INVALIDATE_GB_IMAGE */
+
+/*
+ * Invalidate an entire guest-backed surface.
+ * (Notify the device that the contents if all images can be lost.)
+ */
+
+typedef
+struct SVGA3dCmdInvalidateGBSurface {
+   uint32 sid;
+} __packed
+SVGA3dCmdInvalidateGBSurface; /* SVGA_3D_CMD_INVALIDATE_GB_SURFACE */
+
+/*
+ * Invalidate a sub rect of an image in a guest-backed surface.  After
+ * issuing this command the driver is required to issue an update call
+ * of the same region before issuing any other commands that reference
+ * this surface or rendering is not guaranteed.
+ */
+
+typedef
+struct SVGA3dCmdInvalidateGBImagePartial {
+   SVGA3dSurfaceImageId image;
+   SVGA3dBox box;
+   uint32 invertBox;
+} __packed
+SVGA3dCmdInvalidateGBImagePartial; /* SVGA_3D_CMD_INVALIDATE_GB_IMAGE_PARTIAL */
+
+/*
+ * Define a guest-backed context.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBContext {
+   uint32 cid;
+} __packed
+SVGA3dCmdDefineGBContext;   /* SVGA_3D_CMD_DEFINE_GB_CONTEXT */
+
+/*
+ * Destroy a guest-backed context.
+ */
+
+typedef
+struct SVGA3dCmdDestroyGBContext {
+   uint32 cid;
+} __packed
+SVGA3dCmdDestroyGBContext;   /* SVGA_3D_CMD_DESTROY_GB_CONTEXT */
+
+/*
+ * Bind a guest-backed context.
+ *
+ * validContents should be set to 0 for new contexts,
+ * and 1 if this is an old context which is getting paged
+ * back on to the device.
+ *
+ * For new contexts, it is recommended that the driver
+ * issue commands to initialize all interesting state
+ * prior to rendering.
+ */
+
+typedef
+struct SVGA3dCmdBindGBContext {
+   uint32 cid;
+   SVGAMobId mobid;
+   uint32 validContents;
+} __packed
+SVGA3dCmdBindGBContext;   /* SVGA_3D_CMD_BIND_GB_CONTEXT */
+
+/*
+ * Readback a guest-backed context.
+ * (Request that the device flush the contents back into guest memory.)
+ */
+
+typedef
+struct SVGA3dCmdReadbackGBContext {
+   uint32 cid;
+} __packed
+SVGA3dCmdReadbackGBContext;   /* SVGA_3D_CMD_READBACK_GB_CONTEXT */
+
+/*
+ * Invalidate a guest-backed context.
+ */
+typedef
+struct SVGA3dCmdInvalidateGBContext {
+   uint32 cid;
+} __packed
+SVGA3dCmdInvalidateGBContext;   /* SVGA_3D_CMD_INVALIDATE_GB_CONTEXT */
+
+/*
+ * Define a guest-backed shader.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBShader {
+   uint32 shid;
+   SVGA3dShaderType type;
+   uint32 sizeInBytes;
+} __packed
+SVGA3dCmdDefineGBShader;   /* SVGA_3D_CMD_DEFINE_GB_SHADER */
+
+/*
+ * Bind a guest-backed shader.
+ */
+
+typedef struct SVGA3dCmdBindGBShader {
+   uint32 shid;
+   SVGAMobId mobid;
+   uint32 offsetInBytes;
+} __packed
+SVGA3dCmdBindGBShader;   /* SVGA_3D_CMD_BIND_GB_SHADER */
+
+/*
+ * Destroy a guest-backed shader.
+ */
+
+typedef struct SVGA3dCmdDestroyGBShader {
+   uint32 shid;
+} __packed
+SVGA3dCmdDestroyGBShader;   /* SVGA_3D_CMD_DESTROY_GB_SHADER */
+
+typedef
+struct {
+   uint32                  cid;
+   uint32                  regStart;
+   SVGA3dShaderType        shaderType;
+   SVGA3dShaderConstType   constType;
+
+   /*
+    * Followed by a variable number of shader constants.
+    *
+    * Note that FLOAT and INT constants are 4-dwords in length, while
+    * BOOL constants are 1-dword in length.
+    */
+} __packed
+SVGA3dCmdSetGBShaderConstInline;
+/* SVGA_3D_CMD_SET_GB_SHADERCONSTS_INLINE */
+
+typedef
+struct {
+   uint32               cid;
+   SVGA3dQueryType      type;
+} __packed
+SVGA3dCmdBeginGBQuery;           /* SVGA_3D_CMD_BEGIN_GB_QUERY */
+
+typedef
+struct {
+   uint32               cid;
+   SVGA3dQueryType      type;
+   SVGAMobId mobid;
+   uint32 offset;
+} __packed
+SVGA3dCmdEndGBQuery;                  /* SVGA_3D_CMD_END_GB_QUERY */
+
+
+/*
+ * SVGA_3D_CMD_WAIT_FOR_GB_QUERY --
+ *
+ *    The semantics of this command are identical to the
+ *    SVGA_3D_CMD_WAIT_FOR_QUERY except that the results are written
+ *    to a Mob instead of a GMR.
+ */
+
+typedef
+struct {
+   uint32               cid;
+   SVGA3dQueryType      type;
+   SVGAMobId mobid;
+   uint32 offset;
+} __packed
+SVGA3dCmdWaitForGBQuery;          /* SVGA_3D_CMD_WAIT_FOR_GB_QUERY */
+
+typedef
+struct {
+   SVGAMobId mobid;
+   uint32 fbOffset;
+   uint32 initalized;
+} __packed
+SVGA3dCmdEnableGart;              /* SVGA_3D_CMD_ENABLE_GART */
+
+typedef
+struct {
+   SVGAMobId mobid;
+   uint32 gartOffset;
+} __packed
+SVGA3dCmdMapMobIntoGart;          /* SVGA_3D_CMD_MAP_MOB_INTO_GART */
+
+
+typedef
+struct {
+   uint32 gartOffset;
+   uint32 numPages;
+} __packed
+SVGA3dCmdUnmapGartRange;          /* SVGA_3D_CMD_UNMAP_GART_RANGE */
+
+
+/*
+ * Screen Targets
+ */
+#define SVGA_STFLAG_PRIMARY (1 << 0)
+
+typedef
+struct {
+   uint32 stid;
+   uint32 width;
+   uint32 height;
+   int32 xRoot;
+   int32 yRoot;
+   uint32 flags;
+} __packed
+SVGA3dCmdDefineGBScreenTarget;    /* SVGA_3D_CMD_DEFINE_GB_SCREENTARGET */
+
+typedef
+struct {
+   uint32 stid;
+} __packed
+SVGA3dCmdDestroyGBScreenTarget;  /* SVGA_3D_CMD_DESTROY_GB_SCREENTARGET */
+
+typedef
+struct {
+   uint32 stid;
+   SVGA3dSurfaceImageId image;
+} __packed
+SVGA3dCmdBindGBScreenTarget;  /* SVGA_3D_CMD_BIND_GB_SCREENTARGET */
+
+typedef
+struct {
+   uint32 stid;
+   SVGA3dBox box;
+} __packed
+SVGA3dCmdUpdateGBScreenTarget;  /* SVGA_3D_CMD_UPDATE_GB_SCREENTARGET */
+
+/*
  * Capability query index.
  *
  * Notes:
@@ -1879,10 +2555,41 @@ typedef enum {
    SVGA3D_DEVCAP_SURFACEFMT_BC5_UNORM              = 83,
 
    /*
-    * Don't add new caps into the previous section; the values in this
-    * enumeration must not change. You can put new values right before
-    * SVGA3D_DEVCAP_MAX.
+    * Deprecated.
+    */
+   SVGA3D_DEVCAP_VGPU10                            = 84,
+
+   /*
+    * This contains several SVGA_3D_CAPS_VIDEO_DECODE elements
+    * ored together, one for every type of video decoding supported.
+    */
+   SVGA3D_DEVCAP_VIDEO_DECODE                      = 85,
+
+   /*
+    * This contains several SVGA_3D_CAPS_VIDEO_PROCESS elements
+    * ored together, one for every type of video processing supported.
+    */
+   SVGA3D_DEVCAP_VIDEO_PROCESS                     = 86,
+
+   SVGA3D_DEVCAP_LINE_AA                           = 87,  /* boolean */
+   SVGA3D_DEVCAP_LINE_STIPPLE                      = 88,  /* boolean */
+   SVGA3D_DEVCAP_MAX_LINE_WIDTH                    = 89,  /* float */
+   SVGA3D_DEVCAP_MAX_AA_LINE_WIDTH                 = 90,  /* float */
+
+   SVGA3D_DEVCAP_SURFACEFMT_YV12                   = 91,
+
+   /*
+    * Does the host support the SVGA logic ops commands?
+    */
+   SVGA3D_DEVCAP_LOGICOPS                          = 92,
+
+   /*
+    * What support does the host have for screen targets?
+    *
+    * See the SVGA3D_SCREENTARGET_CAP bits below.
     */
+   SVGA3D_DEVCAP_SCREENTARGETS                     = 93,
+
    SVGA3D_DEVCAP_MAX                                  /* This must be the last index. */
 } SVGA3dDevCapIndex;
 
@@ -1893,4 +2600,28 @@ typedef union {
    float  f;
 } SVGA3dDevCapResult;
 
+typedef enum {
+   SVGA3DCAPS_RECORD_UNKNOWN        = 0,
+   SVGA3DCAPS_RECORD_DEVCAPS_MIN    = 0x100,
+   SVGA3DCAPS_RECORD_DEVCAPS        = 0x100,
+   SVGA3DCAPS_RECORD_DEVCAPS_MAX    = 0x1ff,
+} SVGA3dCapsRecordType;
+
+typedef
+struct SVGA3dCapsRecordHeader {
+   uint32 length;
+   SVGA3dCapsRecordType type;
+}
+SVGA3dCapsRecordHeader;
+
+typedef
+struct SVGA3dCapsRecord {
+   SVGA3dCapsRecordHeader header;
+   uint32 data[1];
+}
+SVGA3dCapsRecord;
+
+
+typedef uint32 SVGA3dCapPair[2];
+
 #endif /* _SVGA3D_REG_H_ */
diff --git a/drivers/gpu/drm/vmwgfx/svga3d_surfacedefs.h b/drivers/gpu/drm/vmwgfx/svga3d_surfacedefs.h
index 8369c3ba10f..ef338509614 100644
--- a/drivers/gpu/drm/vmwgfx/svga3d_surfacedefs.h
+++ b/drivers/gpu/drm/vmwgfx/svga3d_surfacedefs.h
@@ -38,8 +38,11 @@
 
 #define DIV_ROUND_UP(x, y)  (((x) + (y) - 1) / (y))
 #define max_t(type, x, y)  ((x) > (y) ? (x) : (y))
+#define min_t(type, x, y)  ((x) < (y) ? (x) : (y))
 #define surf_size_struct SVGA3dSize
 #define u32 uint32
+#define u64 uint64_t
+#define U32_MAX ((u32)~0U)
 
 #endif /* __KERNEL__ */
 
@@ -704,8 +707,8 @@ static const struct svga3d_surface_desc svga3d_surface_descs[] = {
 
 static inline u32 clamped_umul32(u32 a, u32 b)
 {
-	uint64_t tmp = (uint64_t) a*b;
-	return (tmp > (uint64_t) ((u32) -1)) ? (u32) -1 : tmp;
+	u64 tmp = (u64) a*b;
+	return (tmp > (u64) U32_MAX) ? U32_MAX : tmp;
 }
 
 static inline const struct svga3d_surface_desc *
@@ -834,7 +837,7 @@ svga3dsurface_get_serialized_size(SVGA3dSurfaceFormat format,
 				  bool cubemap)
 {
 	const struct svga3d_surface_desc *desc = svga3dsurface_get_desc(format);
-	u32 total_size = 0;
+	u64 total_size = 0;
 	u32 mip;
 
 	for (mip = 0; mip < num_mip_levels; mip++) {
@@ -847,7 +850,7 @@ svga3dsurface_get_serialized_size(SVGA3dSurfaceFormat format,
 	if (cubemap)
 		total_size *= SVGA3D_MAX_SURFACE_FACES;
 
-	return total_size;
+	return (u32) min_t(u64, total_size, (u64) U32_MAX);
 }
 
 
diff --git a/drivers/gpu/drm/vmwgfx/svga_reg.h b/drivers/gpu/drm/vmwgfx/svga_reg.h
index 01f63cb4967..11323dd5196 100644
--- a/drivers/gpu/drm/vmwgfx/svga_reg.h
+++ b/drivers/gpu/drm/vmwgfx/svga_reg.h
@@ -169,7 +169,17 @@ enum {
    SVGA_REG_TRACES = 45,            /* Enable trace-based updates even when FIFO is on */
    SVGA_REG_GMRS_MAX_PAGES = 46,    /* Maximum number of 4KB pages for all GMRs */
    SVGA_REG_MEMORY_SIZE = 47,       /* Total dedicated device memory excluding FIFO */
-   SVGA_REG_TOP = 48,               /* Must be 1 more than the last register */
+   SVGA_REG_COMMAND_LOW = 48,       /* Lower 32 bits and submits commands */
+   SVGA_REG_COMMAND_HIGH = 49,      /* Upper 32 bits of command buffer PA */
+   SVGA_REG_MAX_PRIMARY_BOUNDING_BOX_MEM = 50,   /* Max primary memory */
+   SVGA_REG_SUGGESTED_GBOBJECT_MEM_SIZE_KB = 51, /* Suggested limit on mob mem */
+   SVGA_REG_DEV_CAP = 52,           /* Write dev cap index, read value */
+   SVGA_REG_CMD_PREPEND_LOW = 53,
+   SVGA_REG_CMD_PREPEND_HIGH = 54,
+   SVGA_REG_SCREENTARGET_MAX_WIDTH = 55,
+   SVGA_REG_SCREENTARGET_MAX_HEIGHT = 56,
+   SVGA_REG_MOB_MAX_SIZE = 57,
+   SVGA_REG_TOP = 58,               /* Must be 1 more than the last register */
 
    SVGA_PALETTE_BASE = 1024,        /* Base of SVGA color map */
    /* Next 768 (== 256*3) registers exist for colormap */
@@ -431,7 +441,10 @@ struct SVGASignedPoint {
 #define SVGA_CAP_TRACES             0x00200000
 #define SVGA_CAP_GMR2               0x00400000
 #define SVGA_CAP_SCREEN_OBJECT_2    0x00800000
-
+#define SVGA_CAP_COMMAND_BUFFERS    0x01000000
+#define SVGA_CAP_DEAD1              0x02000000
+#define SVGA_CAP_CMD_BUFFERS_2      0x04000000
+#define SVGA_CAP_GBOBJECTS          0x08000000
 
 /*
  * FIFO register indices.
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
index 0489c615248..6327cfc3680 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
@@ -40,6 +40,10 @@ static uint32_t vram_ne_placement_flags = TTM_PL_FLAG_VRAM |
 static uint32_t sys_placement_flags = TTM_PL_FLAG_SYSTEM |
 	TTM_PL_FLAG_CACHED;
 
+static uint32_t sys_ne_placement_flags = TTM_PL_FLAG_SYSTEM |
+	TTM_PL_FLAG_CACHED |
+	TTM_PL_FLAG_NO_EVICT;
+
 static uint32_t gmr_placement_flags = VMW_PL_FLAG_GMR |
 	TTM_PL_FLAG_CACHED;
 
@@ -47,6 +51,9 @@ static uint32_t gmr_ne_placement_flags = VMW_PL_FLAG_GMR |
 	TTM_PL_FLAG_CACHED |
 	TTM_PL_FLAG_NO_EVICT;
 
+static uint32_t mob_placement_flags = VMW_PL_FLAG_MOB |
+	TTM_PL_FLAG_CACHED;
+
 struct ttm_placement vmw_vram_placement = {
 	.fpfn = 0,
 	.lpfn = 0,
@@ -116,16 +123,26 @@ struct ttm_placement vmw_sys_placement = {
 	.busy_placement = &sys_placement_flags
 };
 
+struct ttm_placement vmw_sys_ne_placement = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.num_placement = 1,
+	.placement = &sys_ne_placement_flags,
+	.num_busy_placement = 1,
+	.busy_placement = &sys_ne_placement_flags
+};
+
 static uint32_t evictable_placement_flags[] = {
 	TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED,
 	TTM_PL_FLAG_VRAM | TTM_PL_FLAG_CACHED,
-	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED
+	VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED,
+	VMW_PL_FLAG_MOB | TTM_PL_FLAG_CACHED
 };
 
 struct ttm_placement vmw_evictable_placement = {
 	.fpfn = 0,
 	.lpfn = 0,
-	.num_placement = 3,
+	.num_placement = 4,
 	.placement = evictable_placement_flags,
 	.num_busy_placement = 1,
 	.busy_placement = &sys_placement_flags
@@ -140,10 +157,21 @@ struct ttm_placement vmw_srf_placement = {
 	.busy_placement = gmr_vram_placement_flags
 };
 
+struct ttm_placement vmw_mob_placement = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.num_placement = 1,
+	.num_busy_placement = 1,
+	.placement = &mob_placement_flags,
+	.busy_placement = &mob_placement_flags
+};
+
 struct vmw_ttm_tt {
 	struct ttm_dma_tt dma_ttm;
 	struct vmw_private *dev_priv;
 	int gmr_id;
+	struct vmw_mob *mob;
+	int mem_type;
 	struct sg_table sgt;
 	struct vmw_sg_table vsgt;
 	uint64_t sg_alloc_size;
@@ -244,6 +272,7 @@ void vmw_piter_start(struct vmw_piter *viter, const struct vmw_sg_table *vsgt,
 		viter->dma_address = &__vmw_piter_dma_addr;
 		viter->page = &__vmw_piter_non_sg_page;
 		viter->addrs = vsgt->addrs;
+		viter->pages = vsgt->pages;
 		break;
 	case vmw_dma_map_populate:
 	case vmw_dma_map_bind:
@@ -424,6 +453,63 @@ static void vmw_ttm_unmap_dma(struct vmw_ttm_tt *vmw_tt)
 	vmw_tt->mapped = false;
 }
 
+
+/**
+ * vmw_bo_map_dma - Make sure buffer object pages are visible to the device
+ *
+ * @bo: Pointer to a struct ttm_buffer_object
+ *
+ * Wrapper around vmw_ttm_map_dma, that takes a TTM buffer object pointer
+ * instead of a pointer to a struct vmw_ttm_backend as argument.
+ * Note that the buffer object must be either pinned or reserved before
+ * calling this function.
+ */
+int vmw_bo_map_dma(struct ttm_buffer_object *bo)
+{
+	struct vmw_ttm_tt *vmw_tt =
+		container_of(bo->ttm, struct vmw_ttm_tt, dma_ttm.ttm);
+
+	return vmw_ttm_map_dma(vmw_tt);
+}
+
+
+/**
+ * vmw_bo_unmap_dma - Make sure buffer object pages are visible to the device
+ *
+ * @bo: Pointer to a struct ttm_buffer_object
+ *
+ * Wrapper around vmw_ttm_unmap_dma, that takes a TTM buffer object pointer
+ * instead of a pointer to a struct vmw_ttm_backend as argument.
+ */
+void vmw_bo_unmap_dma(struct ttm_buffer_object *bo)
+{
+	struct vmw_ttm_tt *vmw_tt =
+		container_of(bo->ttm, struct vmw_ttm_tt, dma_ttm.ttm);
+
+	vmw_ttm_unmap_dma(vmw_tt);
+}
+
+
+/**
+ * vmw_bo_sg_table - Return a struct vmw_sg_table object for a
+ * TTM buffer object
+ *
+ * @bo: Pointer to a struct ttm_buffer_object
+ *
+ * Returns a pointer to a struct vmw_sg_table object. The object should
+ * not be freed after use.
+ * Note that for the device addresses to be valid, the buffer object must
+ * either be reserved or pinned.
+ */
+const struct vmw_sg_table *vmw_bo_sg_table(struct ttm_buffer_object *bo)
+{
+	struct vmw_ttm_tt *vmw_tt =
+		container_of(bo->ttm, struct vmw_ttm_tt, dma_ttm.ttm);
+
+	return &vmw_tt->vsgt;
+}
+
+
 static int vmw_ttm_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem)
 {
 	struct vmw_ttm_tt *vmw_be =
@@ -435,9 +521,27 @@ static int vmw_ttm_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem)
 		return ret;
 
 	vmw_be->gmr_id = bo_mem->start;
+	vmw_be->mem_type = bo_mem->mem_type;
+
+	switch (bo_mem->mem_type) {
+	case VMW_PL_GMR:
+		return vmw_gmr_bind(vmw_be->dev_priv, &vmw_be->vsgt,
+				    ttm->num_pages, vmw_be->gmr_id);
+	case VMW_PL_MOB:
+		if (unlikely(vmw_be->mob == NULL)) {
+			vmw_be->mob =
+				vmw_mob_create(ttm->num_pages);
+			if (unlikely(vmw_be->mob == NULL))
+				return -ENOMEM;
+		}
 
-	return vmw_gmr_bind(vmw_be->dev_priv, &vmw_be->vsgt,
-			    ttm->num_pages, vmw_be->gmr_id);
+		return vmw_mob_bind(vmw_be->dev_priv, vmw_be->mob,
+				    &vmw_be->vsgt, ttm->num_pages,
+				    vmw_be->gmr_id);
+	default:
+		BUG();
+	}
+	return 0;
 }
 
 static int vmw_ttm_unbind(struct ttm_tt *ttm)
@@ -445,7 +549,16 @@ static int vmw_ttm_unbind(struct ttm_tt *ttm)
 	struct vmw_ttm_tt *vmw_be =
 		container_of(ttm, struct vmw_ttm_tt, dma_ttm.ttm);
 
-	vmw_gmr_unbind(vmw_be->dev_priv, vmw_be->gmr_id);
+	switch (vmw_be->mem_type) {
+	case VMW_PL_GMR:
+		vmw_gmr_unbind(vmw_be->dev_priv, vmw_be->gmr_id);
+		break;
+	case VMW_PL_MOB:
+		vmw_mob_unbind(vmw_be->dev_priv, vmw_be->mob);
+		break;
+	default:
+		BUG();
+	}
 
 	if (vmw_be->dev_priv->map_mode == vmw_dma_map_bind)
 		vmw_ttm_unmap_dma(vmw_be);
@@ -453,6 +566,7 @@ static int vmw_ttm_unbind(struct ttm_tt *ttm)
 	return 0;
 }
 
+
 static void vmw_ttm_destroy(struct ttm_tt *ttm)
 {
 	struct vmw_ttm_tt *vmw_be =
@@ -463,9 +577,14 @@ static void vmw_ttm_destroy(struct ttm_tt *ttm)
 		ttm_dma_tt_fini(&vmw_be->dma_ttm);
 	else
 		ttm_tt_fini(ttm);
+
+	if (vmw_be->mob)
+		vmw_mob_destroy(vmw_be->mob);
+
 	kfree(vmw_be);
 }
 
+
 static int vmw_ttm_populate(struct ttm_tt *ttm)
 {
 	struct vmw_ttm_tt *vmw_tt =
@@ -500,6 +619,12 @@ static void vmw_ttm_unpopulate(struct ttm_tt *ttm)
 	struct vmw_private *dev_priv = vmw_tt->dev_priv;
 	struct ttm_mem_global *glob = vmw_mem_glob(dev_priv);
 
+
+	if (vmw_tt->mob) {
+		vmw_mob_destroy(vmw_tt->mob);
+		vmw_tt->mob = NULL;
+	}
+
 	vmw_ttm_unmap_dma(vmw_tt);
 	if (dev_priv->map_mode == vmw_dma_alloc_coherent) {
 		size_t size =
@@ -517,7 +642,7 @@ static struct ttm_backend_func vmw_ttm_func = {
 	.destroy = vmw_ttm_destroy,
 };
 
-struct ttm_tt *vmw_ttm_tt_create(struct ttm_bo_device *bdev,
+static struct ttm_tt *vmw_ttm_tt_create(struct ttm_bo_device *bdev,
 				 unsigned long size, uint32_t page_flags,
 				 struct page *dummy_read_page)
 {
@@ -530,6 +655,7 @@ struct ttm_tt *vmw_ttm_tt_create(struct ttm_bo_device *bdev,
 
 	vmw_be->dma_ttm.ttm.func = &vmw_ttm_func;
 	vmw_be->dev_priv = container_of(bdev, struct vmw_private, bdev);
+	vmw_be->mob = NULL;
 
 	if (vmw_be->dev_priv->map_mode == vmw_dma_alloc_coherent)
 		ret = ttm_dma_tt_init(&vmw_be->dma_ttm, bdev, size, page_flags,
@@ -546,12 +672,12 @@ out_no_init:
 	return NULL;
 }
 
-int vmw_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags)
+static int vmw_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags)
 {
 	return 0;
 }
 
-int vmw_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
+static int vmw_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 		      struct ttm_mem_type_manager *man)
 {
 	switch (type) {
@@ -571,6 +697,7 @@ int vmw_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 		man->default_caching = TTM_PL_FLAG_CACHED;
 		break;
 	case VMW_PL_GMR:
+	case VMW_PL_MOB:
 		/*
 		 * "Guest Memory Regions" is an aperture like feature with
 		 *  one slot per bo. There is an upper limit of the number of
@@ -589,7 +716,7 @@ int vmw_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 	return 0;
 }
 
-void vmw_evict_flags(struct ttm_buffer_object *bo,
+static void vmw_evict_flags(struct ttm_buffer_object *bo,
 		     struct ttm_placement *placement)
 {
 	*placement = vmw_sys_placement;
@@ -618,6 +745,7 @@ static int vmw_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg
 	switch (mem->mem_type) {
 	case TTM_PL_SYSTEM:
 	case VMW_PL_GMR:
+	case VMW_PL_MOB:
 		return 0;
 	case TTM_PL_VRAM:
 		mem->bus.offset = mem->start << PAGE_SHIFT;
@@ -677,6 +805,38 @@ static int vmw_sync_obj_wait(void *sync_obj, bool lazy, bool interruptible)
 				  VMW_FENCE_WAIT_TIMEOUT);
 }
 
+/**
+ * vmw_move_notify - TTM move_notify_callback
+ *
+ * @bo:             The TTM buffer object about to move.
+ * @mem:            The truct ttm_mem_reg indicating to what memory
+ *                  region the move is taking place.
+ *
+ * Calls move_notify for all subsystems needing it.
+ * (currently only resources).
+ */
+static void vmw_move_notify(struct ttm_buffer_object *bo,
+			    struct ttm_mem_reg *mem)
+{
+	vmw_resource_move_notify(bo, mem);
+}
+
+
+/**
+ * vmw_swap_notify - TTM move_notify_callback
+ *
+ * @bo:             The TTM buffer object about to be swapped out.
+ */
+static void vmw_swap_notify(struct ttm_buffer_object *bo)
+{
+	struct ttm_bo_device *bdev = bo->bdev;
+
+	spin_lock(&bdev->fence_lock);
+	ttm_bo_wait(bo, false, false, false);
+	spin_unlock(&bdev->fence_lock);
+}
+
+
 struct ttm_bo_driver vmw_bo_driver = {
 	.ttm_tt_create = &vmw_ttm_tt_create,
 	.ttm_tt_populate = &vmw_ttm_populate,
@@ -691,8 +851,8 @@ struct ttm_bo_driver vmw_bo_driver = {
 	.sync_obj_flush = vmw_sync_obj_flush,
 	.sync_obj_unref = vmw_sync_obj_unref,
 	.sync_obj_ref = vmw_sync_obj_ref,
-	.move_notify = NULL,
-	.swap_notify = NULL,
+	.move_notify = vmw_move_notify,
+	.swap_notify = vmw_swap_notify,
 	.fault_reserve_notify = &vmw_ttm_fault_reserve_notify,
 	.io_mem_reserve = &vmw_ttm_io_mem_reserve,
 	.io_mem_free = &vmw_ttm_io_mem_free,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
index 00ae0925aca..8bb26dcd9ea 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
@@ -32,12 +32,30 @@
 struct vmw_user_context {
 	struct ttm_base_object base;
 	struct vmw_resource res;
+	struct vmw_ctx_binding_state cbs;
 };
 
+
+
+typedef int (*vmw_scrub_func)(struct vmw_ctx_bindinfo *, bool);
+
 static void vmw_user_context_free(struct vmw_resource *res);
 static struct vmw_resource *
 vmw_user_context_base_to_res(struct ttm_base_object *base);
 
+static int vmw_gb_context_create(struct vmw_resource *res);
+static int vmw_gb_context_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf);
+static int vmw_gb_context_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf);
+static int vmw_gb_context_destroy(struct vmw_resource *res);
+static int vmw_context_scrub_shader(struct vmw_ctx_bindinfo *bi, bool rebind);
+static int vmw_context_scrub_render_target(struct vmw_ctx_bindinfo *bi,
+					   bool rebind);
+static int vmw_context_scrub_texture(struct vmw_ctx_bindinfo *bi, bool rebind);
+static void vmw_context_binding_state_scrub(struct vmw_ctx_binding_state *cbs);
+static void vmw_context_binding_state_kill(struct vmw_ctx_binding_state *cbs);
 static uint64_t vmw_user_context_size;
 
 static const struct vmw_user_resource_conv user_context_conv = {
@@ -62,6 +80,23 @@ static const struct vmw_res_func vmw_legacy_context_func = {
 	.unbind = NULL
 };
 
+static const struct vmw_res_func vmw_gb_context_func = {
+	.res_type = vmw_res_context,
+	.needs_backup = true,
+	.may_evict = true,
+	.type_name = "guest backed contexts",
+	.backup_placement = &vmw_mob_placement,
+	.create = vmw_gb_context_create,
+	.destroy = vmw_gb_context_destroy,
+	.bind = vmw_gb_context_bind,
+	.unbind = vmw_gb_context_unbind
+};
+
+static const vmw_scrub_func vmw_scrub_funcs[vmw_ctx_binding_max] = {
+	[vmw_ctx_binding_shader] = vmw_context_scrub_shader,
+	[vmw_ctx_binding_rt] = vmw_context_scrub_render_target,
+	[vmw_ctx_binding_tex] = vmw_context_scrub_texture };
+
 /**
  * Context management:
  */
@@ -76,6 +111,20 @@ static void vmw_hw_context_destroy(struct vmw_resource *res)
 	} *cmd;
 
 
+	if (res->func->destroy == vmw_gb_context_destroy) {
+		mutex_lock(&dev_priv->cmdbuf_mutex);
+		mutex_lock(&dev_priv->binding_mutex);
+		(void) vmw_context_binding_state_kill
+			(&container_of(res, struct vmw_user_context, res)->cbs);
+		(void) vmw_gb_context_destroy(res);
+		mutex_unlock(&dev_priv->binding_mutex);
+		if (dev_priv->pinned_bo != NULL &&
+		    !dev_priv->query_cid_valid)
+			__vmw_execbuf_release_pinned_bo(dev_priv, NULL);
+		mutex_unlock(&dev_priv->cmdbuf_mutex);
+		return;
+	}
+
 	vmw_execbuf_release_pinned_bo(dev_priv);
 	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
 	if (unlikely(cmd == NULL)) {
@@ -92,6 +141,33 @@ static void vmw_hw_context_destroy(struct vmw_resource *res)
 	vmw_3d_resource_dec(dev_priv, false);
 }
 
+static int vmw_gb_context_init(struct vmw_private *dev_priv,
+			       struct vmw_resource *res,
+			       void (*res_free) (struct vmw_resource *res))
+{
+	int ret;
+	struct vmw_user_context *uctx =
+		container_of(res, struct vmw_user_context, res);
+
+	ret = vmw_resource_init(dev_priv, res, true,
+				res_free, &vmw_gb_context_func);
+	res->backup_size = SVGA3D_CONTEXT_DATA_SIZE;
+
+	if (unlikely(ret != 0)) {
+		if (res_free)
+			res_free(res);
+		else
+			kfree(res);
+		return ret;
+	}
+
+	memset(&uctx->cbs, 0, sizeof(uctx->cbs));
+	INIT_LIST_HEAD(&uctx->cbs.list);
+
+	vmw_resource_activate(res, vmw_hw_context_destroy);
+	return 0;
+}
+
 static int vmw_context_init(struct vmw_private *dev_priv,
 			    struct vmw_resource *res,
 			    void (*res_free) (struct vmw_resource *res))
@@ -103,6 +179,9 @@ static int vmw_context_init(struct vmw_private *dev_priv,
 		SVGA3dCmdDefineContext body;
 	} *cmd;
 
+	if (dev_priv->has_mob)
+		return vmw_gb_context_init(dev_priv, res, res_free);
+
 	ret = vmw_resource_init(dev_priv, res, false,
 				res_free, &vmw_legacy_context_func);
 
@@ -154,6 +233,180 @@ struct vmw_resource *vmw_context_alloc(struct vmw_private *dev_priv)
 	return (ret == 0) ? res : NULL;
 }
 
+
+static int vmw_gb_context_create(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	int ret;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineGBContext body;
+	} *cmd;
+
+	if (likely(res->id != -1))
+		return 0;
+
+	ret = vmw_resource_alloc_id(res);
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Failed to allocate a context id.\n");
+		goto out_no_id;
+	}
+
+	if (unlikely(res->id >= VMWGFX_NUM_GB_CONTEXT)) {
+		ret = -EBUSY;
+		goto out_no_fifo;
+	}
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for context "
+			  "creation.\n");
+		ret = -ENOMEM;
+		goto out_no_fifo;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DEFINE_GB_CONTEXT;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = res->id;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	(void) vmw_3d_resource_inc(dev_priv, false);
+
+	return 0;
+
+out_no_fifo:
+	vmw_resource_release_id(res);
+out_no_id:
+	return ret;
+}
+
+static int vmw_gb_context_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBContext body;
+	} *cmd;
+	struct ttm_buffer_object *bo = val_buf->bo;
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for context "
+			  "binding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_BIND_GB_CONTEXT;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = res->id;
+	cmd->body.mobid = bo->mem.start;
+	cmd->body.validContents = res->backup_dirty;
+	res->backup_dirty = false;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+static int vmw_gb_context_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct ttm_buffer_object *bo = val_buf->bo;
+	struct vmw_fence_obj *fence;
+	struct vmw_user_context *uctx =
+		container_of(res, struct vmw_user_context, res);
+
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdReadbackGBContext body;
+	} *cmd1;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBContext body;
+	} *cmd2;
+	uint32_t submit_size;
+	uint8_t *cmd;
+
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	mutex_lock(&dev_priv->binding_mutex);
+	vmw_context_binding_state_scrub(&uctx->cbs);
+
+	submit_size = sizeof(*cmd2) + (readback ? sizeof(*cmd1) : 0);
+
+	cmd = vmw_fifo_reserve(dev_priv, submit_size);
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for context "
+			  "unbinding.\n");
+		mutex_unlock(&dev_priv->binding_mutex);
+		return -ENOMEM;
+	}
+
+	cmd2 = (void *) cmd;
+	if (readback) {
+		cmd1 = (void *) cmd;
+		cmd1->header.id = SVGA_3D_CMD_READBACK_GB_CONTEXT;
+		cmd1->header.size = sizeof(cmd1->body);
+		cmd1->body.cid = res->id;
+		cmd2 = (void *) (&cmd1[1]);
+	}
+	cmd2->header.id = SVGA_3D_CMD_BIND_GB_CONTEXT;
+	cmd2->header.size = sizeof(cmd2->body);
+	cmd2->body.cid = res->id;
+	cmd2->body.mobid = SVGA3D_INVALID_ID;
+
+	vmw_fifo_commit(dev_priv, submit_size);
+	mutex_unlock(&dev_priv->binding_mutex);
+
+	/*
+	 * Create a fence object and fence the backup buffer.
+	 */
+
+	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
+					  &fence, NULL);
+
+	vmw_fence_single_bo(bo, fence);
+
+	if (likely(fence != NULL))
+		vmw_fence_obj_unreference(&fence);
+
+	return 0;
+}
+
+static int vmw_gb_context_destroy(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDestroyGBContext body;
+	} *cmd;
+
+	if (likely(res->id == -1))
+		return 0;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for context "
+			  "destruction.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DESTROY_GB_CONTEXT;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = res->id;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	if (dev_priv->query_cid == res->id)
+		dev_priv->query_cid_valid = false;
+	vmw_resource_release_id(res);
+	vmw_3d_resource_dec(dev_priv, false);
+
+	return 0;
+}
+
 /**
  * User-space context management:
  */
@@ -209,7 +462,6 @@ int vmw_context_define_ioctl(struct drm_device *dev, void *data,
 	struct vmw_resource *tmp;
 	struct drm_vmw_context_arg *arg = (struct drm_vmw_context_arg *)data;
 	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	int ret;
 
 
@@ -221,7 +473,7 @@ int vmw_context_define_ioctl(struct drm_device *dev, void *data,
 	if (unlikely(vmw_user_context_size == 0))
 		vmw_user_context_size = ttm_round_pot(sizeof(*ctx)) + 128;
 
-	ret = ttm_read_lock(&vmaster->lock, true);
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -268,7 +520,384 @@ int vmw_context_define_ioctl(struct drm_device *dev, void *data,
 out_err:
 	vmw_resource_unreference(&res);
 out_unlock:
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
 	return ret;
 
 }
+
+/**
+ * vmw_context_scrub_shader - scrub a shader binding from a context.
+ *
+ * @bi: single binding information.
+ * @rebind: Whether to issue a bind instead of scrub command.
+ */
+static int vmw_context_scrub_shader(struct vmw_ctx_bindinfo *bi, bool rebind)
+{
+	struct vmw_private *dev_priv = bi->ctx->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdSetShader body;
+	} *cmd;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_SET_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = bi->ctx->id;
+	cmd->body.type = bi->i1.shader_type;
+	cmd->body.shid = ((rebind) ? bi->res->id : SVGA3D_INVALID_ID);
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+/**
+ * vmw_context_scrub_render_target - scrub a render target binding
+ * from a context.
+ *
+ * @bi: single binding information.
+ * @rebind: Whether to issue a bind instead of scrub command.
+ */
+static int vmw_context_scrub_render_target(struct vmw_ctx_bindinfo *bi,
+					   bool rebind)
+{
+	struct vmw_private *dev_priv = bi->ctx->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdSetRenderTarget body;
+	} *cmd;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for render target "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_SETRENDERTARGET;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = bi->ctx->id;
+	cmd->body.type = bi->i1.rt_type;
+	cmd->body.target.sid = ((rebind) ? bi->res->id : SVGA3D_INVALID_ID);
+	cmd->body.target.face = 0;
+	cmd->body.target.mipmap = 0;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+/**
+ * vmw_context_scrub_texture - scrub a texture binding from a context.
+ *
+ * @bi: single binding information.
+ * @rebind: Whether to issue a bind instead of scrub command.
+ *
+ * TODO: Possibly complement this function with a function that takes
+ * a list of texture bindings and combines them to a single command.
+ */
+static int vmw_context_scrub_texture(struct vmw_ctx_bindinfo *bi,
+				     bool rebind)
+{
+	struct vmw_private *dev_priv = bi->ctx->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		struct {
+			SVGA3dCmdSetTextureState c;
+			SVGA3dTextureState s1;
+		} body;
+	} *cmd;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for texture "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+
+	cmd->header.id = SVGA_3D_CMD_SETTEXTURESTATE;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.c.cid = bi->ctx->id;
+	cmd->body.s1.stage = bi->i1.texture_stage;
+	cmd->body.s1.name = SVGA3D_TS_BIND_TEXTURE;
+	cmd->body.s1.value = ((rebind) ? bi->res->id : SVGA3D_INVALID_ID);
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+/**
+ * vmw_context_binding_drop: Stop tracking a context binding
+ *
+ * @cb: Pointer to binding tracker storage.
+ *
+ * Stops tracking a context binding, and re-initializes its storage.
+ * Typically used when the context binding is replaced with a binding to
+ * another (or the same, for that matter) resource.
+ */
+static void vmw_context_binding_drop(struct vmw_ctx_binding *cb)
+{
+	list_del(&cb->ctx_list);
+	if (!list_empty(&cb->res_list))
+		list_del(&cb->res_list);
+	cb->bi.ctx = NULL;
+}
+
+/**
+ * vmw_context_binding_add: Start tracking a context binding
+ *
+ * @cbs: Pointer to the context binding state tracker.
+ * @bi: Information about the binding to track.
+ *
+ * Performs basic checks on the binding to make sure arguments are within
+ * bounds and then starts tracking the binding in the context binding
+ * state structure @cbs.
+ */
+int vmw_context_binding_add(struct vmw_ctx_binding_state *cbs,
+			    const struct vmw_ctx_bindinfo *bi)
+{
+	struct vmw_ctx_binding *loc;
+
+	switch (bi->bt) {
+	case vmw_ctx_binding_rt:
+		if (unlikely((unsigned)bi->i1.rt_type >= SVGA3D_RT_MAX)) {
+			DRM_ERROR("Illegal render target type %u.\n",
+				  (unsigned) bi->i1.rt_type);
+			return -EINVAL;
+		}
+		loc = &cbs->render_targets[bi->i1.rt_type];
+		break;
+	case vmw_ctx_binding_tex:
+		if (unlikely((unsigned)bi->i1.texture_stage >=
+			     SVGA3D_NUM_TEXTURE_UNITS)) {
+			DRM_ERROR("Illegal texture/sampler unit %u.\n",
+				  (unsigned) bi->i1.texture_stage);
+			return -EINVAL;
+		}
+		loc = &cbs->texture_units[bi->i1.texture_stage];
+		break;
+	case vmw_ctx_binding_shader:
+		if (unlikely((unsigned)bi->i1.shader_type >=
+			     SVGA3D_SHADERTYPE_MAX)) {
+			DRM_ERROR("Illegal shader type %u.\n",
+				  (unsigned) bi->i1.shader_type);
+			return -EINVAL;
+		}
+		loc = &cbs->shaders[bi->i1.shader_type];
+		break;
+	default:
+		BUG();
+	}
+
+	if (loc->bi.ctx != NULL)
+		vmw_context_binding_drop(loc);
+
+	loc->bi = *bi;
+	loc->bi.scrubbed = false;
+	list_add_tail(&loc->ctx_list, &cbs->list);
+	INIT_LIST_HEAD(&loc->res_list);
+
+	return 0;
+}
+
+/**
+ * vmw_context_binding_transfer: Transfer a context binding tracking entry.
+ *
+ * @cbs: Pointer to the persistent context binding state tracker.
+ * @bi: Information about the binding to track.
+ *
+ */
+static void vmw_context_binding_transfer(struct vmw_ctx_binding_state *cbs,
+					 const struct vmw_ctx_bindinfo *bi)
+{
+	struct vmw_ctx_binding *loc;
+
+	switch (bi->bt) {
+	case vmw_ctx_binding_rt:
+		loc = &cbs->render_targets[bi->i1.rt_type];
+		break;
+	case vmw_ctx_binding_tex:
+		loc = &cbs->texture_units[bi->i1.texture_stage];
+		break;
+	case vmw_ctx_binding_shader:
+		loc = &cbs->shaders[bi->i1.shader_type];
+		break;
+	default:
+		BUG();
+	}
+
+	if (loc->bi.ctx != NULL)
+		vmw_context_binding_drop(loc);
+
+	if (bi->res != NULL) {
+		loc->bi = *bi;
+		list_add_tail(&loc->ctx_list, &cbs->list);
+		list_add_tail(&loc->res_list, &bi->res->binding_head);
+	}
+}
+
+/**
+ * vmw_context_binding_kill - Kill a binding on the device
+ * and stop tracking it.
+ *
+ * @cb: Pointer to binding tracker storage.
+ *
+ * Emits FIFO commands to scrub a binding represented by @cb.
+ * Then stops tracking the binding and re-initializes its storage.
+ */
+static void vmw_context_binding_kill(struct vmw_ctx_binding *cb)
+{
+	if (!cb->bi.scrubbed) {
+		(void) vmw_scrub_funcs[cb->bi.bt](&cb->bi, false);
+		cb->bi.scrubbed = true;
+	}
+	vmw_context_binding_drop(cb);
+}
+
+/**
+ * vmw_context_binding_state_kill - Kill all bindings associated with a
+ * struct vmw_ctx_binding state structure, and re-initialize the structure.
+ *
+ * @cbs: Pointer to the context binding state tracker.
+ *
+ * Emits commands to scrub all bindings associated with the
+ * context binding state tracker. Then re-initializes the whole structure.
+ */
+static void vmw_context_binding_state_kill(struct vmw_ctx_binding_state *cbs)
+{
+	struct vmw_ctx_binding *entry, *next;
+
+	list_for_each_entry_safe(entry, next, &cbs->list, ctx_list)
+		vmw_context_binding_kill(entry);
+}
+
+/**
+ * vmw_context_binding_state_scrub - Scrub all bindings associated with a
+ * struct vmw_ctx_binding state structure.
+ *
+ * @cbs: Pointer to the context binding state tracker.
+ *
+ * Emits commands to scrub all bindings associated with the
+ * context binding state tracker.
+ */
+static void vmw_context_binding_state_scrub(struct vmw_ctx_binding_state *cbs)
+{
+	struct vmw_ctx_binding *entry;
+
+	list_for_each_entry(entry, &cbs->list, ctx_list) {
+		if (!entry->bi.scrubbed) {
+			(void) vmw_scrub_funcs[entry->bi.bt](&entry->bi, false);
+			entry->bi.scrubbed = true;
+		}
+	}
+}
+
+/**
+ * vmw_context_binding_res_list_kill - Kill all bindings on a
+ * resource binding list
+ *
+ * @head: list head of resource binding list
+ *
+ * Kills all bindings associated with a specific resource. Typically
+ * called before the resource is destroyed.
+ */
+void vmw_context_binding_res_list_kill(struct list_head *head)
+{
+	struct vmw_ctx_binding *entry, *next;
+
+	list_for_each_entry_safe(entry, next, head, res_list)
+		vmw_context_binding_kill(entry);
+}
+
+/**
+ * vmw_context_binding_res_list_scrub - Scrub all bindings on a
+ * resource binding list
+ *
+ * @head: list head of resource binding list
+ *
+ * Scrub all bindings associated with a specific resource. Typically
+ * called before the resource is evicted.
+ */
+void vmw_context_binding_res_list_scrub(struct list_head *head)
+{
+	struct vmw_ctx_binding *entry;
+
+	list_for_each_entry(entry, head, res_list) {
+		if (!entry->bi.scrubbed) {
+			(void) vmw_scrub_funcs[entry->bi.bt](&entry->bi, false);
+			entry->bi.scrubbed = true;
+		}
+	}
+}
+
+/**
+ * vmw_context_binding_state_transfer - Commit staged binding info
+ *
+ * @ctx: Pointer to context to commit the staged binding info to.
+ * @from: Staged binding info built during execbuf.
+ *
+ * Transfers binding info from a temporary structure to the persistent
+ * structure in the context. This can be done once commands
+ */
+void vmw_context_binding_state_transfer(struct vmw_resource *ctx,
+					struct vmw_ctx_binding_state *from)
+{
+	struct vmw_user_context *uctx =
+		container_of(ctx, struct vmw_user_context, res);
+	struct vmw_ctx_binding *entry, *next;
+
+	list_for_each_entry_safe(entry, next, &from->list, ctx_list)
+		vmw_context_binding_transfer(&uctx->cbs, &entry->bi);
+}
+
+/**
+ * vmw_context_rebind_all - Rebind all scrubbed bindings of a context
+ *
+ * @ctx: The context resource
+ *
+ * Walks through the context binding list and rebinds all scrubbed
+ * resources.
+ */
+int vmw_context_rebind_all(struct vmw_resource *ctx)
+{
+	struct vmw_ctx_binding *entry;
+	struct vmw_user_context *uctx =
+		container_of(ctx, struct vmw_user_context, res);
+	struct vmw_ctx_binding_state *cbs = &uctx->cbs;
+	int ret;
+
+	list_for_each_entry(entry, &cbs->list, ctx_list) {
+		if (likely(!entry->bi.scrubbed))
+			continue;
+
+		if (WARN_ON(entry->bi.res == NULL || entry->bi.res->id ==
+			    SVGA3D_INVALID_ID))
+			continue;
+
+		ret = vmw_scrub_funcs[entry->bi.bt](&entry->bi, true);
+		if (unlikely(ret != 0))
+			return ret;
+
+		entry->bi.scrubbed = false;
+	}
+
+	return 0;
+}
+
+/**
+ * vmw_context_binding_list - Return a list of context bindings
+ *
+ * @ctx: The context resource
+ *
+ * Returns the current list of bindings of the given context. Note that
+ * this list becomes stale as soon as the dev_priv::binding_mutex is unlocked.
+ */
+struct list_head *vmw_context_binding_list(struct vmw_resource *ctx)
+{
+	return &(container_of(ctx, struct vmw_user_context, res)->cbs.list);
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
index d4e54fcc0ac..70ddce8358b 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
@@ -52,11 +52,10 @@ int vmw_dmabuf_to_placement(struct vmw_private *dev_priv,
 			    struct ttm_placement *placement,
 			    bool interruptible)
 {
-	struct vmw_master *vmaster = dev_priv->active_master;
 	struct ttm_buffer_object *bo = &buf->base;
 	int ret;
 
-	ret = ttm_write_lock(&vmaster->lock, interruptible);
+	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -71,7 +70,7 @@ int vmw_dmabuf_to_placement(struct vmw_private *dev_priv,
 	ttm_bo_unreserve(bo);
 
 err:
-	ttm_write_unlock(&vmaster->lock);
+	ttm_write_unlock(&dev_priv->reservation_sem);
 	return ret;
 }
 
@@ -95,12 +94,11 @@ int vmw_dmabuf_to_vram_or_gmr(struct vmw_private *dev_priv,
 			      struct vmw_dma_buffer *buf,
 			      bool pin, bool interruptible)
 {
-	struct vmw_master *vmaster = dev_priv->active_master;
 	struct ttm_buffer_object *bo = &buf->base;
 	struct ttm_placement *placement;
 	int ret;
 
-	ret = ttm_write_lock(&vmaster->lock, interruptible);
+	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -143,7 +141,7 @@ int vmw_dmabuf_to_vram_or_gmr(struct vmw_private *dev_priv,
 err_unreserve:
 	ttm_bo_unreserve(bo);
 err:
-	ttm_write_unlock(&vmaster->lock);
+	ttm_write_unlock(&dev_priv->reservation_sem);
 	return ret;
 }
 
@@ -198,7 +196,6 @@ int vmw_dmabuf_to_start_of_vram(struct vmw_private *dev_priv,
 				struct vmw_dma_buffer *buf,
 				bool pin, bool interruptible)
 {
-	struct vmw_master *vmaster = dev_priv->active_master;
 	struct ttm_buffer_object *bo = &buf->base;
 	struct ttm_placement placement;
 	int ret = 0;
@@ -209,7 +206,7 @@ int vmw_dmabuf_to_start_of_vram(struct vmw_private *dev_priv,
 		placement = vmw_vram_placement;
 	placement.lpfn = bo->num_pages;
 
-	ret = ttm_write_lock(&vmaster->lock, interruptible);
+	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -232,7 +229,7 @@ int vmw_dmabuf_to_start_of_vram(struct vmw_private *dev_priv,
 
 	ttm_bo_unreserve(bo);
 err_unlock:
-	ttm_write_unlock(&vmaster->lock);
+	ttm_write_unlock(&dev_priv->reservation_sem);
 
 	return ret;
 }
@@ -290,8 +287,7 @@ void vmw_bo_get_guest_ptr(const struct ttm_buffer_object *bo,
 /**
  * vmw_bo_pin - Pin or unpin a buffer object without moving it.
  *
- * @bo: The buffer object. Must be reserved, and present either in VRAM
- * or GMR memory.
+ * @bo: The buffer object. Must be reserved.
  * @pin: Whether to pin or unpin.
  *
  */
@@ -303,10 +299,9 @@ void vmw_bo_pin(struct ttm_buffer_object *bo, bool pin)
 	int ret;
 
 	lockdep_assert_held(&bo->resv->lock.base);
-	BUG_ON(old_mem_type != TTM_PL_VRAM &&
-	       old_mem_type != VMW_PL_GMR);
 
-	pl_flags = TTM_PL_FLAG_VRAM | VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED;
+	pl_flags = TTM_PL_FLAG_VRAM | VMW_PL_FLAG_GMR | VMW_PL_FLAG_MOB
+		| TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED;
 	if (pin)
 		pl_flags |= TTM_PL_FLAG_NO_EVICT;
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
index c7a549694e5..246a62bab37 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
@@ -112,6 +112,21 @@
 #define DRM_IOCTL_VMW_UPDATE_LAYOUT				\
 	DRM_IOW(DRM_COMMAND_BASE + DRM_VMW_UPDATE_LAYOUT,	\
 		 struct drm_vmw_update_layout_arg)
+#define DRM_IOCTL_VMW_CREATE_SHADER				\
+	DRM_IOWR(DRM_COMMAND_BASE + DRM_VMW_CREATE_SHADER,	\
+		 struct drm_vmw_shader_create_arg)
+#define DRM_IOCTL_VMW_UNREF_SHADER				\
+	DRM_IOW(DRM_COMMAND_BASE + DRM_VMW_UNREF_SHADER,	\
+		 struct drm_vmw_shader_arg)
+#define DRM_IOCTL_VMW_GB_SURFACE_CREATE				\
+	DRM_IOWR(DRM_COMMAND_BASE + DRM_VMW_GB_SURFACE_CREATE,	\
+		 union drm_vmw_gb_surface_create_arg)
+#define DRM_IOCTL_VMW_GB_SURFACE_REF				\
+	DRM_IOWR(DRM_COMMAND_BASE + DRM_VMW_GB_SURFACE_REF,	\
+		 union drm_vmw_gb_surface_reference_arg)
+#define DRM_IOCTL_VMW_SYNCCPU					\
+	DRM_IOW(DRM_COMMAND_BASE + DRM_VMW_SYNCCPU,		\
+		 struct drm_vmw_synccpu_arg)
 
 /**
  * The core DRM version of this macro doesn't account for
@@ -127,11 +142,11 @@
 
 static const struct drm_ioctl_desc vmw_ioctls[] = {
 	VMW_IOCTL_DEF(VMW_GET_PARAM, vmw_getparam_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_ALLOC_DMABUF, vmw_dmabuf_alloc_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_UNREF_DMABUF, vmw_dmabuf_unref_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_CURSOR_BYPASS,
 		      vmw_kms_cursor_bypass_ioctl,
 		      DRM_MASTER | DRM_CONTROL_ALLOW | DRM_UNLOCKED),
@@ -144,29 +159,28 @@ static const struct drm_ioctl_desc vmw_ioctls[] = {
 		      DRM_MASTER | DRM_CONTROL_ALLOW | DRM_UNLOCKED),
 
 	VMW_IOCTL_DEF(VMW_CREATE_CONTEXT, vmw_context_define_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_UNREF_CONTEXT, vmw_context_destroy_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_CREATE_SURFACE, vmw_surface_define_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_UNREF_SURFACE, vmw_surface_destroy_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_REF_SURFACE, vmw_surface_reference_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_EXECBUF, vmw_execbuf_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_FENCE_WAIT, vmw_fence_obj_wait_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_FENCE_SIGNALED,
 		      vmw_fence_obj_signaled_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_FENCE_UNREF, vmw_fence_obj_unref_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
-	VMW_IOCTL_DEF(VMW_FENCE_EVENT,
-		      vmw_fence_event_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_UNLOCKED | DRM_RENDER_ALLOW),
+	VMW_IOCTL_DEF(VMW_FENCE_EVENT, vmw_fence_event_ioctl,
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_GET_3D_CAP, vmw_get_cap_3d_ioctl,
-		      DRM_AUTH | DRM_UNLOCKED),
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
 
 	/* these allow direct access to the framebuffers mark as master only */
 	VMW_IOCTL_DEF(VMW_PRESENT, vmw_present_ioctl,
@@ -177,6 +191,21 @@ static const struct drm_ioctl_desc vmw_ioctls[] = {
 	VMW_IOCTL_DEF(VMW_UPDATE_LAYOUT,
 		      vmw_kms_update_layout_ioctl,
 		      DRM_MASTER | DRM_UNLOCKED),
+	VMW_IOCTL_DEF(VMW_CREATE_SHADER,
+		      vmw_shader_define_ioctl,
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
+	VMW_IOCTL_DEF(VMW_UNREF_SHADER,
+		      vmw_shader_destroy_ioctl,
+		      DRM_UNLOCKED | DRM_RENDER_ALLOW),
+	VMW_IOCTL_DEF(VMW_GB_SURFACE_CREATE,
+		      vmw_gb_surface_define_ioctl,
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
+	VMW_IOCTL_DEF(VMW_GB_SURFACE_REF,
+		      vmw_gb_surface_reference_ioctl,
+		      DRM_AUTH | DRM_UNLOCKED | DRM_RENDER_ALLOW),
+	VMW_IOCTL_DEF(VMW_SYNCCPU,
+		      vmw_user_dmabuf_synccpu_ioctl,
+		      DRM_UNLOCKED | DRM_RENDER_ALLOW),
 };
 
 static struct pci_device_id vmw_pci_id_list[] = {
@@ -189,6 +218,7 @@ static int enable_fbdev = IS_ENABLED(CONFIG_DRM_VMWGFX_FBCON);
 static int vmw_force_iommu;
 static int vmw_restrict_iommu;
 static int vmw_force_coherent;
+static int vmw_restrict_dma_mask;
 
 static int vmw_probe(struct pci_dev *, const struct pci_device_id *);
 static void vmw_master_init(struct vmw_master *);
@@ -203,6 +233,8 @@ MODULE_PARM_DESC(restrict_iommu, "Try to limit IOMMU usage for TTM pages");
 module_param_named(restrict_iommu, vmw_restrict_iommu, int, 0600);
 MODULE_PARM_DESC(force_coherent, "Force coherent TTM pages");
 module_param_named(force_coherent, vmw_force_coherent, int, 0600);
+MODULE_PARM_DESC(restrict_dma_mask, "Restrict DMA mask to 44 bits with IOMMU");
+module_param_named(restrict_dma_mask, vmw_restrict_dma_mask, int, 0600);
 
 
 static void vmw_print_capabilities(uint32_t capabilities)
@@ -240,38 +272,52 @@ static void vmw_print_capabilities(uint32_t capabilities)
 		DRM_INFO("  GMR2.\n");
 	if (capabilities & SVGA_CAP_SCREEN_OBJECT_2)
 		DRM_INFO("  Screen Object 2.\n");
+	if (capabilities & SVGA_CAP_COMMAND_BUFFERS)
+		DRM_INFO("  Command Buffers.\n");
+	if (capabilities & SVGA_CAP_CMD_BUFFERS_2)
+		DRM_INFO("  Command Buffers 2.\n");
+	if (capabilities & SVGA_CAP_GBOBJECTS)
+		DRM_INFO("  Guest Backed Resources.\n");
 }
 
-
 /**
- * vmw_execbuf_prepare_dummy_query - Initialize a query result structure at
- * the start of a buffer object.
+ * vmw_dummy_query_bo_create - create a bo to hold a dummy query result
  *
- * @dev_priv: The device private structure.
+ * @dev_priv: A device private structure.
  *
- * This function will idle the buffer using an uninterruptible wait, then
- * map the first page and initialize a pending occlusion query result structure,
- * Finally it will unmap the buffer.
+ * This function creates a small buffer object that holds the query
+ * result for dummy queries emitted as query barriers.
+ * The function will then map the first page and initialize a pending
+ * occlusion query result structure, Finally it will unmap the buffer.
+ * No interruptible waits are done within this function.
  *
- * TODO: Since we're only mapping a single page, we should optimize the map
- * to use kmap_atomic / iomap_atomic.
+ * Returns an error if bo creation or initialization fails.
  */
-static void vmw_dummy_query_bo_prepare(struct vmw_private *dev_priv)
+static int vmw_dummy_query_bo_create(struct vmw_private *dev_priv)
 {
+	int ret;
+	struct ttm_buffer_object *bo;
 	struct ttm_bo_kmap_obj map;
 	volatile SVGA3dQueryResult *result;
 	bool dummy;
-	int ret;
-	struct ttm_bo_device *bdev = &dev_priv->bdev;
-	struct ttm_buffer_object *bo = dev_priv->dummy_query_bo;
 
-	ttm_bo_reserve(bo, false, false, false, 0);
-	spin_lock(&bdev->fence_lock);
-	ret = ttm_bo_wait(bo, false, false, false);
-	spin_unlock(&bdev->fence_lock);
+	/*
+	 * Create the bo as pinned, so that a tryreserve will
+	 * immediately succeed. This is because we're the only
+	 * user of the bo currently.
+	 */
+	ret = ttm_bo_create(&dev_priv->bdev,
+			    PAGE_SIZE,
+			    ttm_bo_type_device,
+			    &vmw_sys_ne_placement,
+			    0, false, NULL,
+			    &bo);
+
 	if (unlikely(ret != 0))
-		(void) vmw_fallback_wait(dev_priv, false, true, 0, false,
-					 10*HZ);
+		return ret;
+
+	ret = ttm_bo_reserve(bo, false, true, false, 0);
+	BUG_ON(ret != 0);
 
 	ret = ttm_bo_kmap(bo, 0, 1, &map);
 	if (likely(ret == 0)) {
@@ -280,34 +326,19 @@ static void vmw_dummy_query_bo_prepare(struct vmw_private *dev_priv)
 		result->state = SVGA3D_QUERYSTATE_PENDING;
 		result->result32 = 0xff;
 		ttm_bo_kunmap(&map);
-	} else
-		DRM_ERROR("Dummy query buffer map failed.\n");
+	}
+	vmw_bo_pin(bo, false);
 	ttm_bo_unreserve(bo);
-}
 
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Dummy query buffer map failed.\n");
+		ttm_bo_unref(&bo);
+	} else
+		dev_priv->dummy_query_bo = bo;
 
-/**
- * vmw_dummy_query_bo_create - create a bo to hold a dummy query result
- *
- * @dev_priv: A device private structure.
- *
- * This function creates a small buffer object that holds the query
- * result for dummy queries emitted as query barriers.
- * No interruptible waits are done within this function.
- *
- * Returns an error if bo creation fails.
- */
-static int vmw_dummy_query_bo_create(struct vmw_private *dev_priv)
-{
-	return ttm_bo_create(&dev_priv->bdev,
-			     PAGE_SIZE,
-			     ttm_bo_type_device,
-			     &vmw_vram_sys_placement,
-			     0, false, NULL,
-			     &dev_priv->dummy_query_bo);
+	return ret;
 }
 
-
 static int vmw_request_device(struct vmw_private *dev_priv)
 {
 	int ret;
@@ -318,14 +349,24 @@ static int vmw_request_device(struct vmw_private *dev_priv)
 		return ret;
 	}
 	vmw_fence_fifo_up(dev_priv->fman);
+	if (dev_priv->has_mob) {
+		ret = vmw_otables_setup(dev_priv);
+		if (unlikely(ret != 0)) {
+			DRM_ERROR("Unable to initialize "
+				  "guest Memory OBjects.\n");
+			goto out_no_mob;
+		}
+	}
 	ret = vmw_dummy_query_bo_create(dev_priv);
 	if (unlikely(ret != 0))
 		goto out_no_query_bo;
-	vmw_dummy_query_bo_prepare(dev_priv);
 
 	return 0;
 
 out_no_query_bo:
+	if (dev_priv->has_mob)
+		vmw_otables_takedown(dev_priv);
+out_no_mob:
 	vmw_fence_fifo_down(dev_priv->fman);
 	vmw_fifo_release(dev_priv, &dev_priv->fifo);
 	return ret;
@@ -341,10 +382,13 @@ static void vmw_release_device(struct vmw_private *dev_priv)
 	BUG_ON(dev_priv->pinned_bo != NULL);
 
 	ttm_bo_unref(&dev_priv->dummy_query_bo);
+	if (dev_priv->has_mob)
+		vmw_otables_takedown(dev_priv);
 	vmw_fence_fifo_down(dev_priv->fman);
 	vmw_fifo_release(dev_priv, &dev_priv->fifo);
 }
 
+
 /**
  * Increase the 3d resource refcount.
  * If the count was prevously zero, initialize the fifo, switching to svga
@@ -510,6 +554,33 @@ out_fixup:
 	return 0;
 }
 
+/**
+ * vmw_dma_masks - set required page- and dma masks
+ *
+ * @dev: Pointer to struct drm-device
+ *
+ * With 32-bit we can only handle 32 bit PFNs. Optionally set that
+ * restriction also for 64-bit systems.
+ */
+#ifdef CONFIG_INTEL_IOMMU
+static int vmw_dma_masks(struct vmw_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+
+	if (intel_iommu_enabled &&
+	    (sizeof(unsigned long) == 4 || vmw_restrict_dma_mask)) {
+		DRM_INFO("Restricting DMA addresses to 44 bits.\n");
+		return dma_set_mask(dev->dev, DMA_BIT_MASK(44));
+	}
+	return 0;
+}
+#else
+static int vmw_dma_masks(struct vmw_private *dev_priv)
+{
+	return 0;
+}
+#endif
+
 static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 {
 	struct vmw_private *dev_priv;
@@ -532,7 +603,9 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	mutex_init(&dev_priv->hw_mutex);
 	mutex_init(&dev_priv->cmdbuf_mutex);
 	mutex_init(&dev_priv->release_mutex);
+	mutex_init(&dev_priv->binding_mutex);
 	rwlock_init(&dev_priv->resource_lock);
+	ttm_lock_init(&dev_priv->reservation_sem);
 
 	for (i = vmw_res_context; i < vmw_res_max; ++i) {
 		idr_init(&dev_priv->res_idr[i]);
@@ -578,14 +651,9 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 
 	vmw_get_initial_size(dev_priv);
 
-	if (dev_priv->capabilities & SVGA_CAP_GMR) {
-		dev_priv->max_gmr_descriptors =
-			vmw_read(dev_priv,
-				 SVGA_REG_GMR_MAX_DESCRIPTOR_LENGTH);
+	if (dev_priv->capabilities & SVGA_CAP_GMR2) {
 		dev_priv->max_gmr_ids =
 			vmw_read(dev_priv, SVGA_REG_GMR_MAX_IDS);
-	}
-	if (dev_priv->capabilities & SVGA_CAP_GMR2) {
 		dev_priv->max_gmr_pages =
 			vmw_read(dev_priv, SVGA_REG_GMRS_MAX_PAGES);
 		dev_priv->memory_size =
@@ -598,23 +666,45 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 		 */
 		dev_priv->memory_size = 512*1024*1024;
 	}
+	dev_priv->max_mob_pages = 0;
+	dev_priv->max_mob_size = 0;
+	if (dev_priv->capabilities & SVGA_CAP_GBOBJECTS) {
+		uint64_t mem_size =
+			vmw_read(dev_priv,
+				 SVGA_REG_SUGGESTED_GBOBJECT_MEM_SIZE_KB);
+
+		dev_priv->max_mob_pages = mem_size * 1024 / PAGE_SIZE;
+		dev_priv->prim_bb_mem =
+			vmw_read(dev_priv,
+				 SVGA_REG_MAX_PRIMARY_BOUNDING_BOX_MEM);
+		dev_priv->max_mob_size =
+			vmw_read(dev_priv, SVGA_REG_MOB_MAX_SIZE);
+	} else
+		dev_priv->prim_bb_mem = dev_priv->vram_size;
+
+	ret = vmw_dma_masks(dev_priv);
+	if (unlikely(ret != 0)) {
+		mutex_unlock(&dev_priv->hw_mutex);
+		goto out_err0;
+	}
+
+	if (unlikely(dev_priv->prim_bb_mem < dev_priv->vram_size))
+		dev_priv->prim_bb_mem = dev_priv->vram_size;
 
 	mutex_unlock(&dev_priv->hw_mutex);
 
 	vmw_print_capabilities(dev_priv->capabilities);
 
-	if (dev_priv->capabilities & SVGA_CAP_GMR) {
+	if (dev_priv->capabilities & SVGA_CAP_GMR2) {
 		DRM_INFO("Max GMR ids is %u\n",
 			 (unsigned)dev_priv->max_gmr_ids);
-		DRM_INFO("Max GMR descriptors is %u\n",
-			 (unsigned)dev_priv->max_gmr_descriptors);
-	}
-	if (dev_priv->capabilities & SVGA_CAP_GMR2) {
 		DRM_INFO("Max number of GMR pages is %u\n",
 			 (unsigned)dev_priv->max_gmr_pages);
 		DRM_INFO("Max dedicated hypervisor surface memory is %u kiB\n",
 			 (unsigned)dev_priv->memory_size / 1024);
 	}
+	DRM_INFO("Maximum display memory size is %u kiB\n",
+		 dev_priv->prim_bb_mem / 1024);
 	DRM_INFO("VRAM at 0x%08x size is %u kiB\n",
 		 dev_priv->vram_start, dev_priv->vram_size / 1024);
 	DRM_INFO("MMIO at 0x%08x size is %u kiB\n",
@@ -632,7 +722,9 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 
 	ret = ttm_bo_device_init(&dev_priv->bdev,
 				 dev_priv->bo_global_ref.ref.object,
-				 &vmw_bo_driver, VMWGFX_FILE_PAGE_OFFSET,
+				 &vmw_bo_driver,
+				 dev->anon_inode->i_mapping,
+				 VMWGFX_FILE_PAGE_OFFSET,
 				 false);
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("Failed initializing TTM buffer object driver.\n");
@@ -649,12 +741,22 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	dev_priv->has_gmr = true;
 	if (((dev_priv->capabilities & (SVGA_CAP_GMR | SVGA_CAP_GMR2)) == 0) ||
 	    refuse_dma || ttm_bo_init_mm(&dev_priv->bdev, VMW_PL_GMR,
-					 dev_priv->max_gmr_ids) != 0) {
+					 VMW_PL_GMR) != 0) {
 		DRM_INFO("No GMR memory available. "
 			 "Graphics memory resources are very limited.\n");
 		dev_priv->has_gmr = false;
 	}
 
+	if (dev_priv->capabilities & SVGA_CAP_GBOBJECTS) {
+		dev_priv->has_mob = true;
+		if (ttm_bo_init_mm(&dev_priv->bdev, VMW_PL_MOB,
+				   VMW_PL_MOB) != 0) {
+			DRM_INFO("No MOB memory available. "
+				 "3D will be disabled.\n");
+			dev_priv->has_mob = false;
+		}
+	}
+
 	dev_priv->mmio_mtrr = arch_phys_wc_add(dev_priv->mmio_start,
 					       dev_priv->mmio_size);
 
@@ -704,7 +806,7 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	}
 
 	if (dev_priv->capabilities & SVGA_CAP_IRQMASK) {
-		ret = drm_irq_install(dev);
+		ret = drm_irq_install(dev, dev->pdev->irq);
 		if (ret != 0) {
 			DRM_ERROR("Failed installing irq: %d\n", ret);
 			goto out_no_irq;
@@ -757,6 +859,8 @@ out_err4:
 	iounmap(dev_priv->mmio_virt);
 out_err3:
 	arch_phys_wc_del(dev_priv->mmio_mtrr);
+	if (dev_priv->has_mob)
+		(void) ttm_bo_clean_mm(&dev_priv->bdev, VMW_PL_MOB);
 	if (dev_priv->has_gmr)
 		(void) ttm_bo_clean_mm(&dev_priv->bdev, VMW_PL_GMR);
 	(void)ttm_bo_clean_mm(&dev_priv->bdev, TTM_PL_VRAM);
@@ -801,6 +905,8 @@ static int vmw_driver_unload(struct drm_device *dev)
 	ttm_object_device_release(&dev_priv->tdev);
 	iounmap(dev_priv->mmio_virt);
 	arch_phys_wc_del(dev_priv->mmio_mtrr);
+	if (dev_priv->has_mob)
+		(void) ttm_bo_clean_mm(&dev_priv->bdev, VMW_PL_MOB);
 	if (dev_priv->has_gmr)
 		(void)ttm_bo_clean_mm(&dev_priv->bdev, VMW_PL_GMR);
 	(void)ttm_bo_clean_mm(&dev_priv->bdev, TTM_PL_VRAM);
@@ -840,6 +946,7 @@ static void vmw_postclose(struct drm_device *dev,
 		drm_master_put(&vmw_fp->locked_master);
 	}
 
+	vmw_compat_shader_man_destroy(vmw_fp->shman);
 	ttm_object_file_release(&vmw_fp->tfile);
 	kfree(vmw_fp);
 }
@@ -859,22 +966,85 @@ static int vmw_driver_open(struct drm_device *dev, struct drm_file *file_priv)
 	if (unlikely(vmw_fp->tfile == NULL))
 		goto out_no_tfile;
 
+	vmw_fp->shman = vmw_compat_shader_man_create(dev_priv);
+	if (IS_ERR(vmw_fp->shman))
+		goto out_no_shman;
+
 	file_priv->driver_priv = vmw_fp;
-	dev_priv->bdev.dev_mapping = dev->dev_mapping;
 
 	return 0;
 
+out_no_shman:
+	ttm_object_file_release(&vmw_fp->tfile);
 out_no_tfile:
 	kfree(vmw_fp);
 	return ret;
 }
 
-static long vmw_unlocked_ioctl(struct file *filp, unsigned int cmd,
-			       unsigned long arg)
+static struct vmw_master *vmw_master_check(struct drm_device *dev,
+					   struct drm_file *file_priv,
+					   unsigned int flags)
+{
+	int ret;
+	struct vmw_fpriv *vmw_fp = vmw_fpriv(file_priv);
+	struct vmw_master *vmaster;
+
+	if (file_priv->minor->type != DRM_MINOR_LEGACY ||
+	    !(flags & DRM_AUTH))
+		return NULL;
+
+	ret = mutex_lock_interruptible(&dev->master_mutex);
+	if (unlikely(ret != 0))
+		return ERR_PTR(-ERESTARTSYS);
+
+	if (file_priv->is_master) {
+		mutex_unlock(&dev->master_mutex);
+		return NULL;
+	}
+
+	/*
+	 * Check if we were previously master, but now dropped.
+	 */
+	if (vmw_fp->locked_master) {
+		mutex_unlock(&dev->master_mutex);
+		DRM_ERROR("Dropped master trying to access ioctl that "
+			  "requires authentication.\n");
+		return ERR_PTR(-EACCES);
+	}
+	mutex_unlock(&dev->master_mutex);
+
+	/*
+	 * Taking the drm_global_mutex after the TTM lock might deadlock
+	 */
+	if (!(flags & DRM_UNLOCKED)) {
+		DRM_ERROR("Refusing locked ioctl access.\n");
+		return ERR_PTR(-EDEADLK);
+	}
+
+	/*
+	 * Take the TTM lock. Possibly sleep waiting for the authenticating
+	 * master to become master again, or for a SIGTERM if the
+	 * authenticating master exits.
+	 */
+	vmaster = vmw_master(file_priv->master);
+	ret = ttm_read_lock(&vmaster->lock, true);
+	if (unlikely(ret != 0))
+		vmaster = ERR_PTR(ret);
+
+	return vmaster;
+}
+
+static long vmw_generic_ioctl(struct file *filp, unsigned int cmd,
+			      unsigned long arg,
+			      long (*ioctl_func)(struct file *, unsigned int,
+						 unsigned long))
 {
 	struct drm_file *file_priv = filp->private_data;
 	struct drm_device *dev = file_priv->minor->dev;
 	unsigned int nr = DRM_IOCTL_NR(cmd);
+	struct vmw_master *vmaster;
+	unsigned int flags;
+	long ret;
 
 	/*
 	 * Do extra checking on driver private ioctls.
@@ -883,18 +1053,44 @@ static long vmw_unlocked_ioctl(struct file *filp, unsigned int cmd,
 	if ((nr >= DRM_COMMAND_BASE) && (nr < DRM_COMMAND_END)
 	    && (nr < DRM_COMMAND_BASE + dev->driver->num_ioctls)) {
 		const struct drm_ioctl_desc *ioctl =
-		    &vmw_ioctls[nr - DRM_COMMAND_BASE];
+			&vmw_ioctls[nr - DRM_COMMAND_BASE];
 
 		if (unlikely(ioctl->cmd_drv != cmd)) {
 			DRM_ERROR("Invalid command format, ioctl %d\n",
 				  nr - DRM_COMMAND_BASE);
 			return -EINVAL;
 		}
+		flags = ioctl->flags;
+	} else if (!drm_ioctl_flags(nr, &flags))
+		return -EINVAL;
+
+	vmaster = vmw_master_check(dev, file_priv, flags);
+	if (unlikely(IS_ERR(vmaster))) {
+		DRM_INFO("IOCTL ERROR %d\n", nr);
+		return PTR_ERR(vmaster);
 	}
 
-	return drm_ioctl(filp, cmd, arg);
+	ret = ioctl_func(filp, cmd, arg);
+	if (vmaster)
+		ttm_read_unlock(&vmaster->lock);
+
+	return ret;
+}
+
+static long vmw_unlocked_ioctl(struct file *filp, unsigned int cmd,
+			       unsigned long arg)
+{
+	return vmw_generic_ioctl(filp, cmd, arg, &drm_ioctl);
 }
 
+#ifdef CONFIG_COMPAT
+static long vmw_compat_ioctl(struct file *filp, unsigned int cmd,
+			     unsigned long arg)
+{
+	return vmw_generic_ioctl(filp, cmd, arg, &drm_compat_ioctl);
+}
+#endif
+
 static void vmw_lastclose(struct drm_device *dev)
 {
 	struct drm_crtc *crtc;
@@ -1063,12 +1259,11 @@ static int vmwgfx_pm_notifier(struct notifier_block *nb, unsigned long val,
 {
 	struct vmw_private *dev_priv =
 		container_of(nb, struct vmw_private, pm_nb);
-	struct vmw_master *vmaster = dev_priv->active_master;
 
 	switch (val) {
 	case PM_HIBERNATION_PREPARE:
 	case PM_SUSPEND_PREPARE:
-		ttm_suspend_lock(&vmaster->lock);
+		ttm_suspend_lock(&dev_priv->reservation_sem);
 
 		/**
 		 * This empties VRAM and unbinds all GMR bindings.
@@ -1082,7 +1277,7 @@ static int vmwgfx_pm_notifier(struct notifier_block *nb, unsigned long val,
 	case PM_POST_HIBERNATION:
 	case PM_POST_SUSPEND:
 	case PM_POST_RESTORE:
-		ttm_suspend_unlock(&vmaster->lock);
+		ttm_suspend_unlock(&dev_priv->reservation_sem);
 
 		break;
 	case PM_RESTORE_PREPARE:
@@ -1203,14 +1398,14 @@ static const struct file_operations vmwgfx_driver_fops = {
 	.poll = vmw_fops_poll,
 	.read = vmw_fops_read,
 #if defined(CONFIG_COMPAT)
-	.compat_ioctl = drm_compat_ioctl,
+	.compat_ioctl = vmw_compat_ioctl,
 #endif
 	.llseek = noop_llseek,
 };
 
 static struct drm_driver driver = {
 	.driver_features = DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED |
-	DRIVER_MODESET | DRIVER_PRIME,
+	DRIVER_MODESET | DRIVER_PRIME | DRIVER_RENDER,
 	.load = vmw_driver_load,
 	.unload = vmw_driver_unload,
 	.lastclose = vmw_lastclose,
@@ -1222,7 +1417,7 @@ static struct drm_driver driver = {
 	.enable_vblank = vmw_enable_vblank,
 	.disable_vblank = vmw_disable_vblank,
 	.ioctls = vmw_ioctls,
-	.num_ioctls = DRM_ARRAY_SIZE(vmw_ioctls),
+	.num_ioctls = ARRAY_SIZE(vmw_ioctls),
 	.master_create = vmw_master_create,
 	.master_destroy = vmw_master_destroy,
 	.master_set = vmw_master_set,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index 20890ad8408..6b252a887ae 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -40,9 +40,9 @@
 #include <drm/ttm/ttm_module.h>
 #include "vmwgfx_fence.h"
 
-#define VMWGFX_DRIVER_DATE "20120209"
+#define VMWGFX_DRIVER_DATE "20140325"
 #define VMWGFX_DRIVER_MAJOR 2
-#define VMWGFX_DRIVER_MINOR 4
+#define VMWGFX_DRIVER_MINOR 6
 #define VMWGFX_DRIVER_PATCHLEVEL 0
 #define VMWGFX_FILE_PAGE_OFFSET 0x00100000
 #define VMWGFX_FIFO_STATIC_SIZE (1024*1024)
@@ -50,19 +50,39 @@
 #define VMWGFX_MAX_VALIDATIONS 2048
 #define VMWGFX_MAX_DISPLAYS 16
 #define VMWGFX_CMD_BOUNCE_INIT_SIZE 32768
+#define VMWGFX_ENABLE_SCREEN_TARGET_OTABLE 0
+
+/*
+ * Perhaps we should have sysfs entries for these.
+ */
+#define VMWGFX_NUM_GB_CONTEXT 256
+#define VMWGFX_NUM_GB_SHADER 20000
+#define VMWGFX_NUM_GB_SURFACE 32768
+#define VMWGFX_NUM_GB_SCREEN_TARGET VMWGFX_MAX_DISPLAYS
+#define VMWGFX_NUM_MOB (VMWGFX_NUM_GB_CONTEXT +\
+			VMWGFX_NUM_GB_SHADER +\
+			VMWGFX_NUM_GB_SURFACE +\
+			VMWGFX_NUM_GB_SCREEN_TARGET)
 
 #define VMW_PL_GMR TTM_PL_PRIV0
 #define VMW_PL_FLAG_GMR TTM_PL_FLAG_PRIV0
+#define VMW_PL_MOB TTM_PL_PRIV1
+#define VMW_PL_FLAG_MOB TTM_PL_FLAG_PRIV1
 
 #define VMW_RES_CONTEXT ttm_driver_type0
 #define VMW_RES_SURFACE ttm_driver_type1
 #define VMW_RES_STREAM ttm_driver_type2
 #define VMW_RES_FENCE ttm_driver_type3
+#define VMW_RES_SHADER ttm_driver_type4
+
+struct vmw_compat_shader_manager;
 
 struct vmw_fpriv {
 	struct drm_master *locked_master;
 	struct ttm_object_file *tfile;
 	struct list_head fence_events;
+	bool gb_aware;
+	struct vmw_compat_shader_manager *shman;
 };
 
 struct vmw_dma_buffer {
@@ -82,6 +102,7 @@ struct vmw_dma_buffer {
 struct vmw_validate_buffer {
 	struct ttm_validate_buffer base;
 	struct drm_hash_item hash;
+	bool validate_as_mob;
 };
 
 struct vmw_res_func;
@@ -98,6 +119,7 @@ struct vmw_resource {
 	const struct vmw_res_func *func;
 	struct list_head lru_head; /* Protected by the resource lock */
 	struct list_head mob_head; /* Protected by @backup reserved */
+	struct list_head binding_head; /* Protected by binding_mutex */
 	void (*res_free) (struct vmw_resource *res);
 	void (*hw_destroy) (struct vmw_resource *res);
 };
@@ -106,6 +128,7 @@ enum vmw_res_type {
 	vmw_res_context,
 	vmw_res_surface,
 	vmw_res_stream,
+	vmw_res_shader,
 	vmw_res_max
 };
 
@@ -154,6 +177,7 @@ struct vmw_fifo_state {
 };
 
 struct vmw_relocation {
+	SVGAMobId *mob_loc;
 	SVGAGuestPtr *location;
 	uint32_t index;
 };
@@ -229,11 +253,77 @@ struct vmw_piter {
 	struct page *(*page)(struct vmw_piter *);
 };
 
+/*
+ * enum vmw_ctx_binding_type - abstract resource to context binding types
+ */
+enum vmw_ctx_binding_type {
+	vmw_ctx_binding_shader,
+	vmw_ctx_binding_rt,
+	vmw_ctx_binding_tex,
+	vmw_ctx_binding_max
+};
+
+/**
+ * struct vmw_ctx_bindinfo - structure representing a single context binding
+ *
+ * @ctx: Pointer to the context structure. NULL means the binding is not
+ * active.
+ * @res: Non ref-counted pointer to the bound resource.
+ * @bt: The binding type.
+ * @i1: Union of information needed to unbind.
+ */
+struct vmw_ctx_bindinfo {
+	struct vmw_resource *ctx;
+	struct vmw_resource *res;
+	enum vmw_ctx_binding_type bt;
+	bool scrubbed;
+	union {
+		SVGA3dShaderType shader_type;
+		SVGA3dRenderTargetType rt_type;
+		uint32 texture_stage;
+	} i1;
+};
+
+/**
+ * struct vmw_ctx_binding - structure representing a single context binding
+ *                        - suitable for tracking in a context
+ *
+ * @ctx_list: List head for context.
+ * @res_list: List head for bound resource.
+ * @bi: Binding info
+ */
+struct vmw_ctx_binding {
+	struct list_head ctx_list;
+	struct list_head res_list;
+	struct vmw_ctx_bindinfo bi;
+};
+
+
+/**
+ * struct vmw_ctx_binding_state - context binding state
+ *
+ * @list: linked list of individual bindings.
+ * @render_targets: Render target bindings.
+ * @texture_units: Texture units/samplers bindings.
+ * @shaders: Shader bindings.
+ *
+ * Note that this structure also provides storage space for the individual
+ * struct vmw_ctx_binding objects, so that no dynamic allocation is needed
+ * for individual bindings.
+ *
+ */
+struct vmw_ctx_binding_state {
+	struct list_head list;
+	struct vmw_ctx_binding render_targets[SVGA3D_RT_MAX];
+	struct vmw_ctx_binding texture_units[SVGA3D_NUM_TEXTURE_UNITS];
+	struct vmw_ctx_binding shaders[SVGA3D_SHADERTYPE_MAX];
+};
+
 struct vmw_sw_context{
 	struct drm_open_hash res_ht;
 	bool res_ht_initialized;
 	bool kernel; /**< is the called made from the kernel */
-	struct ttm_object_file *tfile;
+	struct vmw_fpriv *fp;
 	struct list_head validate_nodes;
 	struct vmw_relocation relocs[VMWGFX_MAX_RELOCATIONS];
 	uint32_t cur_reloc;
@@ -250,6 +340,8 @@ struct vmw_sw_context{
 	struct vmw_resource *last_query_ctx;
 	bool needs_post_query_barrier;
 	struct vmw_resource *error_resource;
+	struct vmw_ctx_binding_state staged_bindings;
+	struct list_head staged_shaders;
 };
 
 struct vmw_legacy_display;
@@ -281,6 +373,7 @@ struct vmw_private {
 	unsigned int io_start;
 	uint32_t vram_start;
 	uint32_t vram_size;
+	uint32_t prim_bb_mem;
 	uint32_t mmio_start;
 	uint32_t mmio_size;
 	uint32_t fb_max_width;
@@ -290,11 +383,13 @@ struct vmw_private {
 	__le32 __iomem *mmio_virt;
 	int mmio_mtrr;
 	uint32_t capabilities;
-	uint32_t max_gmr_descriptors;
 	uint32_t max_gmr_ids;
 	uint32_t max_gmr_pages;
+	uint32_t max_mob_pages;
+	uint32_t max_mob_size;
 	uint32_t memory_size;
 	bool has_gmr;
+	bool has_mob;
 	struct mutex hw_mutex;
 
 	/*
@@ -370,6 +465,7 @@ struct vmw_private {
 
 	struct vmw_sw_context ctx;
 	struct mutex cmdbuf_mutex;
+	struct mutex binding_mutex;
 
 	/**
 	 * Operating mode.
@@ -391,6 +487,11 @@ struct vmw_private {
 	uint32_t num_3d_resources;
 
 	/*
+	 * Replace this with an rwsem as soon as we have down_xx_interruptible()
+	 */
+	struct ttm_lock reservation_sem;
+
+	/*
 	 * Query processing. These members
 	 * are protected by the cmdbuf mutex.
 	 */
@@ -415,6 +516,12 @@ struct vmw_private {
 	 * DMA mapping stuff.
 	 */
 	enum vmw_dma_map_mode map_mode;
+
+	/*
+	 * Guest Backed stuff
+	 */
+	struct ttm_buffer_object *otable_bo;
+	struct vmw_otable *otables;
 };
 
 static inline struct vmw_surface *vmw_res_to_srf(struct vmw_resource *res)
@@ -471,23 +578,14 @@ extern void vmw_gmr_unbind(struct vmw_private *dev_priv, int gmr_id);
  * Resource utilities - vmwgfx_resource.c
  */
 struct vmw_user_resource_conv;
-extern const struct vmw_user_resource_conv *user_surface_converter;
-extern const struct vmw_user_resource_conv *user_context_converter;
 
-extern struct vmw_resource *vmw_context_alloc(struct vmw_private *dev_priv);
 extern void vmw_resource_unreference(struct vmw_resource **p_res);
 extern struct vmw_resource *vmw_resource_reference(struct vmw_resource *res);
+extern struct vmw_resource *
+vmw_resource_reference_unless_doomed(struct vmw_resource *res);
 extern int vmw_resource_validate(struct vmw_resource *res);
 extern int vmw_resource_reserve(struct vmw_resource *res, bool no_backup);
 extern bool vmw_resource_needs_backup(const struct vmw_resource *res);
-extern int vmw_context_destroy_ioctl(struct drm_device *dev, void *data,
-				     struct drm_file *file_priv);
-extern int vmw_context_define_ioctl(struct drm_device *dev, void *data,
-				    struct drm_file *file_priv);
-extern int vmw_context_check(struct vmw_private *dev_priv,
-			     struct ttm_object_file *tfile,
-			     int id,
-			     struct vmw_resource **p_res);
 extern int vmw_user_lookup_handle(struct vmw_private *dev_priv,
 				  struct ttm_object_file *tfile,
 				  uint32_t handle,
@@ -499,18 +597,6 @@ extern int vmw_user_resource_lookup_handle(
 	uint32_t handle,
 	const struct vmw_user_resource_conv *converter,
 	struct vmw_resource **p_res);
-extern void vmw_surface_res_free(struct vmw_resource *res);
-extern int vmw_surface_destroy_ioctl(struct drm_device *dev, void *data,
-				     struct drm_file *file_priv);
-extern int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
-				    struct drm_file *file_priv);
-extern int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
-				       struct drm_file *file_priv);
-extern int vmw_surface_check(struct vmw_private *dev_priv,
-			     struct ttm_object_file *tfile,
-			     uint32_t handle, int *id);
-extern int vmw_surface_validate(struct vmw_private *dev_priv,
-				struct vmw_surface *srf);
 extern void vmw_dmabuf_bo_free(struct ttm_buffer_object *bo);
 extern int vmw_dmabuf_init(struct vmw_private *dev_priv,
 			   struct vmw_dma_buffer *vmw_bo,
@@ -519,10 +605,21 @@ extern int vmw_dmabuf_init(struct vmw_private *dev_priv,
 			   void (*bo_free) (struct ttm_buffer_object *bo));
 extern int vmw_user_dmabuf_verify_access(struct ttm_buffer_object *bo,
 				  struct ttm_object_file *tfile);
+extern int vmw_user_dmabuf_alloc(struct vmw_private *dev_priv,
+				 struct ttm_object_file *tfile,
+				 uint32_t size,
+				 bool shareable,
+				 uint32_t *handle,
+				 struct vmw_dma_buffer **p_dma_buf);
+extern int vmw_user_dmabuf_reference(struct ttm_object_file *tfile,
+				     struct vmw_dma_buffer *dma_buf,
+				     uint32_t *handle);
 extern int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file_priv);
 extern int vmw_dmabuf_unref_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file_priv);
+extern int vmw_user_dmabuf_synccpu_ioctl(struct drm_device *dev, void *data,
+					 struct drm_file *file_priv);
 extern uint32_t vmw_dmabuf_validate_node(struct ttm_buffer_object *bo,
 					 uint32_t cur_validate_node);
 extern void vmw_dmabuf_validate_clear(struct ttm_buffer_object *bo);
@@ -622,10 +719,16 @@ extern struct ttm_placement vmw_vram_sys_placement;
 extern struct ttm_placement vmw_vram_gmr_placement;
 extern struct ttm_placement vmw_vram_gmr_ne_placement;
 extern struct ttm_placement vmw_sys_placement;
+extern struct ttm_placement vmw_sys_ne_placement;
 extern struct ttm_placement vmw_evictable_placement;
 extern struct ttm_placement vmw_srf_placement;
+extern struct ttm_placement vmw_mob_placement;
 extern struct ttm_bo_driver vmw_bo_driver;
 extern int vmw_dma_quiescent(struct drm_device *dev);
+extern int vmw_bo_map_dma(struct ttm_buffer_object *bo);
+extern void vmw_bo_unmap_dma(struct ttm_buffer_object *bo);
+extern const struct vmw_sg_table *
+vmw_bo_sg_table(struct ttm_buffer_object *bo);
 extern void vmw_piter_start(struct vmw_piter *viter,
 			    const struct vmw_sg_table *vsgt,
 			    unsigned long p_offs);
@@ -701,7 +804,7 @@ extern void vmw_execbuf_copy_fence_user(struct vmw_private *dev_priv,
  * IRQs and wating - vmwgfx_irq.c
  */
 
-extern irqreturn_t vmw_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t vmw_irq_handler(int irq, void *arg);
 extern int vmw_wait_seqno(struct vmw_private *dev_priv, bool lazy,
 			     uint32_t seqno, bool interruptible,
 			     unsigned long timeout);
@@ -832,6 +935,101 @@ extern int vmw_prime_handle_to_fd(struct drm_device *dev,
 				  uint32_t handle, uint32_t flags,
 				  int *prime_fd);
 
+/*
+ * MemoryOBject management -  vmwgfx_mob.c
+ */
+struct vmw_mob;
+extern int vmw_mob_bind(struct vmw_private *dev_priv, struct vmw_mob *mob,
+			const struct vmw_sg_table *vsgt,
+			unsigned long num_data_pages, int32_t mob_id);
+extern void vmw_mob_unbind(struct vmw_private *dev_priv,
+			   struct vmw_mob *mob);
+extern void vmw_mob_destroy(struct vmw_mob *mob);
+extern struct vmw_mob *vmw_mob_create(unsigned long data_pages);
+extern int vmw_otables_setup(struct vmw_private *dev_priv);
+extern void vmw_otables_takedown(struct vmw_private *dev_priv);
+
+/*
+ * Context management - vmwgfx_context.c
+ */
+
+extern const struct vmw_user_resource_conv *user_context_converter;
+
+extern struct vmw_resource *vmw_context_alloc(struct vmw_private *dev_priv);
+
+extern int vmw_context_check(struct vmw_private *dev_priv,
+			     struct ttm_object_file *tfile,
+			     int id,
+			     struct vmw_resource **p_res);
+extern int vmw_context_define_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv);
+extern int vmw_context_destroy_ioctl(struct drm_device *dev, void *data,
+				     struct drm_file *file_priv);
+extern int vmw_context_binding_add(struct vmw_ctx_binding_state *cbs,
+				   const struct vmw_ctx_bindinfo *ci);
+extern void
+vmw_context_binding_state_transfer(struct vmw_resource *res,
+				   struct vmw_ctx_binding_state *cbs);
+extern void vmw_context_binding_res_list_kill(struct list_head *head);
+extern void vmw_context_binding_res_list_scrub(struct list_head *head);
+extern int vmw_context_rebind_all(struct vmw_resource *ctx);
+extern struct list_head *vmw_context_binding_list(struct vmw_resource *ctx);
+
+/*
+ * Surface management - vmwgfx_surface.c
+ */
+
+extern const struct vmw_user_resource_conv *user_surface_converter;
+
+extern void vmw_surface_res_free(struct vmw_resource *res);
+extern int vmw_surface_destroy_ioctl(struct drm_device *dev, void *data,
+				     struct drm_file *file_priv);
+extern int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv);
+extern int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
+				       struct drm_file *file_priv);
+extern int vmw_gb_surface_define_ioctl(struct drm_device *dev, void *data,
+				       struct drm_file *file_priv);
+extern int vmw_gb_surface_reference_ioctl(struct drm_device *dev, void *data,
+					  struct drm_file *file_priv);
+extern int vmw_surface_check(struct vmw_private *dev_priv,
+			     struct ttm_object_file *tfile,
+			     uint32_t handle, int *id);
+extern int vmw_surface_validate(struct vmw_private *dev_priv,
+				struct vmw_surface *srf);
+
+/*
+ * Shader management - vmwgfx_shader.c
+ */
+
+extern const struct vmw_user_resource_conv *user_shader_converter;
+
+extern int vmw_shader_define_ioctl(struct drm_device *dev, void *data,
+				   struct drm_file *file_priv);
+extern int vmw_shader_destroy_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv);
+extern int vmw_compat_shader_lookup(struct vmw_compat_shader_manager *man,
+				    SVGA3dShaderType shader_type,
+				    u32 *user_key);
+extern void vmw_compat_shaders_commit(struct vmw_compat_shader_manager *man,
+				      struct list_head *list);
+extern void vmw_compat_shaders_revert(struct vmw_compat_shader_manager *man,
+				      struct list_head *list);
+extern int vmw_compat_shader_remove(struct vmw_compat_shader_manager *man,
+				    u32 user_key,
+				    SVGA3dShaderType shader_type,
+				    struct list_head *list);
+extern int vmw_compat_shader_add(struct vmw_compat_shader_manager *man,
+				 u32 user_key, const void *bytecode,
+				 SVGA3dShaderType shader_type,
+				 size_t size,
+				 struct ttm_object_file *tfile,
+				 struct list_head *list);
+extern struct vmw_compat_shader_manager *
+vmw_compat_shader_man_create(struct vmw_private *dev_priv);
+extern void
+vmw_compat_shader_man_destroy(struct vmw_compat_shader_manager *man);
+
 
 /**
  * Inline helper functions
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
index 599f6469a1e..87df0b3674f 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
@@ -54,6 +54,8 @@ struct vmw_resource_relocation {
  * @res: Ref-counted pointer to the resource.
  * @switch_backup: Boolean whether to switch backup buffer on unreserve.
  * @new_backup: Refcounted pointer to the new backup buffer.
+ * @staged_bindings: If @res is a context, tracks bindings set up during
+ * the command batch. Otherwise NULL.
  * @new_backup_offset: New backup buffer offset if @new_backup is non-NUll.
  * @first_usage: Set to true the first time the resource is referenced in
  * the command stream.
@@ -65,12 +67,32 @@ struct vmw_resource_val_node {
 	struct drm_hash_item hash;
 	struct vmw_resource *res;
 	struct vmw_dma_buffer *new_backup;
+	struct vmw_ctx_binding_state *staged_bindings;
 	unsigned long new_backup_offset;
 	bool first_usage;
 	bool no_buffer_needed;
 };
 
 /**
+ * struct vmw_cmd_entry - Describe a command for the verifier
+ *
+ * @user_allow: Whether allowed from the execbuf ioctl.
+ * @gb_disable: Whether disabled if guest-backed objects are available.
+ * @gb_enable: Whether enabled iff guest-backed objects are available.
+ */
+struct vmw_cmd_entry {
+	int (*func) (struct vmw_private *, struct vmw_sw_context *,
+		     SVGA3dCmdHeader *);
+	bool user_allow;
+	bool gb_disable;
+	bool gb_enable;
+};
+
+#define VMW_CMD_DEF(_cmd, _func, _user_allow, _gb_disable, _gb_enable)	\
+	[(_cmd) - SVGA_3D_CMD_BASE] = {(_func), (_user_allow),\
+				       (_gb_disable), (_gb_enable)}
+
+/**
  * vmw_resource_unreserve - unreserve resources previously reserved for
  * command submission.
  *
@@ -87,6 +109,18 @@ static void vmw_resource_list_unreserve(struct list_head *list,
 		struct vmw_dma_buffer *new_backup =
 			backoff ? NULL : val->new_backup;
 
+		/*
+		 * Transfer staged context bindings to the
+		 * persistent context binding tracker.
+		 */
+		if (unlikely(val->staged_bindings)) {
+			if (!backoff) {
+				vmw_context_binding_state_transfer
+					(val->res, val->staged_bindings);
+			}
+			kfree(val->staged_bindings);
+			val->staged_bindings = NULL;
+		}
 		vmw_resource_unreserve(res, new_backup,
 			val->new_backup_offset);
 		vmw_dmabuf_unreference(&val->new_backup);
@@ -146,6 +180,44 @@ static int vmw_resource_val_add(struct vmw_sw_context *sw_context,
 }
 
 /**
+ * vmw_resource_context_res_add - Put resources previously bound to a context on
+ * the validation list
+ *
+ * @dev_priv: Pointer to a device private structure
+ * @sw_context: Pointer to a software context used for this command submission
+ * @ctx: Pointer to the context resource
+ *
+ * This function puts all resources that were previously bound to @ctx on
+ * the resource validation list. This is part of the context state reemission
+ */
+static int vmw_resource_context_res_add(struct vmw_private *dev_priv,
+					struct vmw_sw_context *sw_context,
+					struct vmw_resource *ctx)
+{
+	struct list_head *binding_list;
+	struct vmw_ctx_binding *entry;
+	int ret = 0;
+	struct vmw_resource *res;
+
+	mutex_lock(&dev_priv->binding_mutex);
+	binding_list = vmw_context_binding_list(ctx);
+
+	list_for_each_entry(entry, binding_list, ctx_list) {
+		res = vmw_resource_reference_unless_doomed(entry->bi.res);
+		if (unlikely(res == NULL))
+			continue;
+
+		ret = vmw_resource_val_add(sw_context, entry->bi.res, NULL);
+		vmw_resource_unreference(&res);
+		if (unlikely(ret != 0))
+			break;
+	}
+
+	mutex_unlock(&dev_priv->binding_mutex);
+	return ret;
+}
+
+/**
  * vmw_resource_relocation_add - Add a relocation to the relocation list
  *
  * @list: Pointer to head of relocation list.
@@ -201,8 +273,12 @@ static void vmw_resource_relocations_apply(uint32_t *cb,
 {
 	struct vmw_resource_relocation *rel;
 
-	list_for_each_entry(rel, list, head)
-		cb[rel->offset] = rel->res->id;
+	list_for_each_entry(rel, list, head) {
+		if (likely(rel->res != NULL))
+			cb[rel->offset] = rel->res->id;
+		else
+			cb[rel->offset] = SVGA_3D_CMD_NOP;
+	}
 }
 
 static int vmw_cmd_invalid(struct vmw_private *dev_priv,
@@ -224,6 +300,7 @@ static int vmw_cmd_ok(struct vmw_private *dev_priv,
  *
  * @sw_context: The software context used for this command submission batch.
  * @bo: The buffer object to add.
+ * @validate_as_mob: Validate this buffer as a MOB.
  * @p_val_node: If non-NULL Will be updated with the validate node number
  * on return.
  *
@@ -232,6 +309,7 @@ static int vmw_cmd_ok(struct vmw_private *dev_priv,
  */
 static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
 				   struct ttm_buffer_object *bo,
+				   bool validate_as_mob,
 				   uint32_t *p_val_node)
 {
 	uint32_t val_node;
@@ -244,6 +322,10 @@ static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
 				    &hash) == 0)) {
 		vval_buf = container_of(hash, struct vmw_validate_buffer,
 					hash);
+		if (unlikely(vval_buf->validate_as_mob != validate_as_mob)) {
+			DRM_ERROR("Inconsistent buffer usage.\n");
+			return -EINVAL;
+		}
 		val_buf = &vval_buf->base;
 		val_node = vval_buf - sw_context->val_bufs;
 	} else {
@@ -266,6 +348,7 @@ static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
 		val_buf->bo = ttm_bo_reference(bo);
 		val_buf->reserved = false;
 		list_add_tail(&val_buf->head, &sw_context->validate_nodes);
+		vval_buf->validate_as_mob = validate_as_mob;
 	}
 
 	sw_context->fence_flags |= DRM_VMW_FENCE_FLAG_EXEC;
@@ -302,7 +385,8 @@ static int vmw_resources_reserve(struct vmw_sw_context *sw_context)
 			struct ttm_buffer_object *bo = &res->backup->base;
 
 			ret = vmw_bo_to_validate_list
-				(sw_context, bo, NULL);
+				(sw_context, bo,
+				 vmw_resource_needs_backup(res), NULL);
 
 			if (unlikely(ret != 0))
 				return ret;
@@ -339,22 +423,27 @@ static int vmw_resources_validate(struct vmw_sw_context *sw_context)
 }
 
 /**
- * vmw_cmd_res_check - Check that a resource is present and if so, put it
+ * vmw_cmd_compat_res_check - Check that a resource is present and if so, put it
  * on the resource validate list unless it's already there.
  *
  * @dev_priv: Pointer to a device private structure.
  * @sw_context: Pointer to the software context.
  * @res_type: Resource type.
  * @converter: User-space visisble type specific information.
- * @id: Pointer to the location in the command buffer currently being
+ * @id: user-space resource id handle.
+ * @id_loc: Pointer to the location in the command buffer currently being
  * parsed from where the user-space resource id handle is located.
+ * @p_val: Pointer to pointer to resource validalidation node. Populated
+ * on exit.
  */
-static int vmw_cmd_res_check(struct vmw_private *dev_priv,
-			     struct vmw_sw_context *sw_context,
-			     enum vmw_res_type res_type,
-			     const struct vmw_user_resource_conv *converter,
-			     uint32_t *id,
-			     struct vmw_resource_val_node **p_val)
+static int
+vmw_cmd_compat_res_check(struct vmw_private *dev_priv,
+			 struct vmw_sw_context *sw_context,
+			 enum vmw_res_type res_type,
+			 const struct vmw_user_resource_conv *converter,
+			 uint32_t id,
+			 uint32_t *id_loc,
+			 struct vmw_resource_val_node **p_val)
 {
 	struct vmw_res_cache_entry *rcache =
 		&sw_context->res_cache[res_type];
@@ -362,15 +451,22 @@ static int vmw_cmd_res_check(struct vmw_private *dev_priv,
 	struct vmw_resource_val_node *node;
 	int ret;
 
-	if (*id == SVGA3D_INVALID_ID)
+	if (id == SVGA3D_INVALID_ID) {
+		if (p_val)
+			*p_val = NULL;
+		if (res_type == vmw_res_context) {
+			DRM_ERROR("Illegal context invalid id.\n");
+			return -EINVAL;
+		}
 		return 0;
+	}
 
 	/*
 	 * Fastpath in case of repeated commands referencing the same
 	 * resource
 	 */
 
-	if (likely(rcache->valid && *id == rcache->handle)) {
+	if (likely(rcache->valid && id == rcache->handle)) {
 		const struct vmw_resource *res = rcache->res;
 
 		rcache->node->first_usage = false;
@@ -379,28 +475,28 @@ static int vmw_cmd_res_check(struct vmw_private *dev_priv,
 
 		return vmw_resource_relocation_add
 			(&sw_context->res_relocations, res,
-			 id - sw_context->buf_start);
+			 id_loc - sw_context->buf_start);
 	}
 
 	ret = vmw_user_resource_lookup_handle(dev_priv,
-					      sw_context->tfile,
-					      *id,
+					      sw_context->fp->tfile,
+					      id,
 					      converter,
 					      &res);
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("Could not find or use resource 0x%08x.\n",
-			  (unsigned) *id);
+			  (unsigned) id);
 		dump_stack();
 		return ret;
 	}
 
 	rcache->valid = true;
 	rcache->res = res;
-	rcache->handle = *id;
+	rcache->handle = id;
 
 	ret = vmw_resource_relocation_add(&sw_context->res_relocations,
 					  res,
-					  id - sw_context->buf_start);
+					  id_loc - sw_context->buf_start);
 	if (unlikely(ret != 0))
 		goto out_no_reloc;
 
@@ -411,6 +507,22 @@ static int vmw_cmd_res_check(struct vmw_private *dev_priv,
 	rcache->node = node;
 	if (p_val)
 		*p_val = node;
+
+	if (dev_priv->has_mob && node->first_usage &&
+	    res_type == vmw_res_context) {
+		ret = vmw_resource_context_res_add(dev_priv, sw_context, res);
+		if (unlikely(ret != 0))
+			goto out_no_reloc;
+		node->staged_bindings =
+			kzalloc(sizeof(*node->staged_bindings), GFP_KERNEL);
+		if (node->staged_bindings == NULL) {
+			DRM_ERROR("Failed to allocate context binding "
+				  "information.\n");
+			goto out_no_reloc;
+		}
+		INIT_LIST_HEAD(&node->staged_bindings->list);
+	}
+
 	vmw_resource_unreference(&res);
 	return 0;
 
@@ -422,6 +534,59 @@ out_no_reloc:
 }
 
 /**
+ * vmw_cmd_res_check - Check that a resource is present and if so, put it
+ * on the resource validate list unless it's already there.
+ *
+ * @dev_priv: Pointer to a device private structure.
+ * @sw_context: Pointer to the software context.
+ * @res_type: Resource type.
+ * @converter: User-space visisble type specific information.
+ * @id_loc: Pointer to the location in the command buffer currently being
+ * parsed from where the user-space resource id handle is located.
+ * @p_val: Pointer to pointer to resource validalidation node. Populated
+ * on exit.
+ */
+static int
+vmw_cmd_res_check(struct vmw_private *dev_priv,
+		  struct vmw_sw_context *sw_context,
+		  enum vmw_res_type res_type,
+		  const struct vmw_user_resource_conv *converter,
+		  uint32_t *id_loc,
+		  struct vmw_resource_val_node **p_val)
+{
+	return vmw_cmd_compat_res_check(dev_priv, sw_context, res_type,
+					converter, *id_loc, id_loc, p_val);
+}
+
+/**
+ * vmw_rebind_contexts - Rebind all resources previously bound to
+ * referenced contexts.
+ *
+ * @sw_context: Pointer to the software context.
+ *
+ * Rebind context binding points that have been scrubbed because of eviction.
+ */
+static int vmw_rebind_contexts(struct vmw_sw_context *sw_context)
+{
+	struct vmw_resource_val_node *val;
+	int ret;
+
+	list_for_each_entry(val, &sw_context->resource_list, head) {
+		if (likely(!val->staged_bindings))
+			continue;
+
+		ret = vmw_context_rebind_all(val->res);
+		if (unlikely(ret != 0)) {
+			if (ret != -ERESTARTSYS)
+				DRM_ERROR("Failed to rebind context.\n");
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+/**
  * vmw_cmd_cid_check - Check a command header for valid context information.
  *
  * @dev_priv: Pointer to a device private structure.
@@ -437,7 +602,7 @@ static int vmw_cmd_cid_check(struct vmw_private *dev_priv,
 {
 	struct vmw_cid_cmd {
 		SVGA3dCmdHeader header;
-		__le32 cid;
+		uint32_t cid;
 	} *cmd;
 
 	cmd = container_of(header, struct vmw_cid_cmd, header);
@@ -453,17 +618,35 @@ static int vmw_cmd_set_render_target_check(struct vmw_private *dev_priv,
 		SVGA3dCmdHeader header;
 		SVGA3dCmdSetRenderTarget body;
 	} *cmd;
+	struct vmw_resource_val_node *ctx_node;
+	struct vmw_resource_val_node *res_node;
 	int ret;
 
-	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	cmd = container_of(header, struct vmw_sid_cmd, header);
+
+	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				user_context_converter, &cmd->body.cid,
+				&ctx_node);
 	if (unlikely(ret != 0))
 		return ret;
 
-	cmd = container_of(header, struct vmw_sid_cmd, header);
 	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 				user_surface_converter,
-				&cmd->body.target.sid, NULL);
-	return ret;
+				&cmd->body.target.sid, &res_node);
+	if (unlikely(ret != 0))
+		return ret;
+
+	if (dev_priv->has_mob) {
+		struct vmw_ctx_bindinfo bi;
+
+		bi.ctx = ctx_node->res;
+		bi.res = res_node ? res_node->res : NULL;
+		bi.bt = vmw_ctx_binding_rt;
+		bi.i1.rt_type = cmd->body.type;
+		return vmw_context_binding_add(ctx_node->staged_bindings, &bi);
+	}
+
+	return 0;
 }
 
 static int vmw_cmd_surface_copy_check(struct vmw_private *dev_priv,
@@ -519,11 +702,6 @@ static int vmw_cmd_blt_surf_screen_check(struct vmw_private *dev_priv,
 
 	cmd = container_of(header, struct vmw_sid_cmd, header);
 
-	if (unlikely(!sw_context->kernel)) {
-		DRM_ERROR("Kernel only SVGA3d command: %u.\n", cmd->header.id);
-		return -EPERM;
-	}
-
 	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 				 user_surface_converter,
 				 &cmd->body.srcImage.sid, NULL);
@@ -541,11 +719,6 @@ static int vmw_cmd_present_check(struct vmw_private *dev_priv,
 
 	cmd = container_of(header, struct vmw_sid_cmd, header);
 
-	if (unlikely(!sw_context->kernel)) {
-		DRM_ERROR("Kernel only SVGA3d command: %u.\n", cmd->header.id);
-		return -EPERM;
-	}
-
 	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 				 user_surface_converter, &cmd->body.sid,
 				 NULL);
@@ -586,7 +759,7 @@ static int vmw_query_bo_switch_prepare(struct vmw_private *dev_priv,
 			sw_context->needs_post_query_barrier = true;
 			ret = vmw_bo_to_validate_list(sw_context,
 						      sw_context->cur_query_bo,
-						      NULL);
+						      dev_priv->has_mob, NULL);
 			if (unlikely(ret != 0))
 				return ret;
 		}
@@ -594,7 +767,7 @@ static int vmw_query_bo_switch_prepare(struct vmw_private *dev_priv,
 
 		ret = vmw_bo_to_validate_list(sw_context,
 					      dev_priv->dummy_query_bo,
-					      NULL);
+					      dev_priv->has_mob, NULL);
 		if (unlikely(ret != 0))
 			return ret;
 
@@ -672,6 +845,66 @@ static void vmw_query_bo_switch_commit(struct vmw_private *dev_priv,
 }
 
 /**
+ * vmw_translate_mob_pointer - Prepare to translate a user-space buffer
+ * handle to a MOB id.
+ *
+ * @dev_priv: Pointer to a device private structure.
+ * @sw_context: The software context used for this command batch validation.
+ * @id: Pointer to the user-space handle to be translated.
+ * @vmw_bo_p: Points to a location that, on successful return will carry
+ * a reference-counted pointer to the DMA buffer identified by the
+ * user-space handle in @id.
+ *
+ * This function saves information needed to translate a user-space buffer
+ * handle to a MOB id. The translation does not take place immediately, but
+ * during a call to vmw_apply_relocations(). This function builds a relocation
+ * list and a list of buffers to validate. The former needs to be freed using
+ * either vmw_apply_relocations() or vmw_free_relocations(). The latter
+ * needs to be freed using vmw_clear_validations.
+ */
+static int vmw_translate_mob_ptr(struct vmw_private *dev_priv,
+				 struct vmw_sw_context *sw_context,
+				 SVGAMobId *id,
+				 struct vmw_dma_buffer **vmw_bo_p)
+{
+	struct vmw_dma_buffer *vmw_bo = NULL;
+	struct ttm_buffer_object *bo;
+	uint32_t handle = *id;
+	struct vmw_relocation *reloc;
+	int ret;
+
+	ret = vmw_user_dmabuf_lookup(sw_context->fp->tfile, handle, &vmw_bo);
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Could not find or use MOB buffer.\n");
+		return -EINVAL;
+	}
+	bo = &vmw_bo->base;
+
+	if (unlikely(sw_context->cur_reloc >= VMWGFX_MAX_RELOCATIONS)) {
+		DRM_ERROR("Max number relocations per submission"
+			  " exceeded\n");
+		ret = -EINVAL;
+		goto out_no_reloc;
+	}
+
+	reloc = &sw_context->relocs[sw_context->cur_reloc++];
+	reloc->mob_loc = id;
+	reloc->location = NULL;
+
+	ret = vmw_bo_to_validate_list(sw_context, bo, true, &reloc->index);
+	if (unlikely(ret != 0))
+		goto out_no_reloc;
+
+	*vmw_bo_p = vmw_bo;
+	return 0;
+
+out_no_reloc:
+	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_p = NULL;
+	return ret;
+}
+
+/**
  * vmw_translate_guest_pointer - Prepare to translate a user-space buffer
  * handle to a valid SVGAGuestPtr
  *
@@ -701,7 +934,7 @@ static int vmw_translate_guest_ptr(struct vmw_private *dev_priv,
 	struct vmw_relocation *reloc;
 	int ret;
 
-	ret = vmw_user_dmabuf_lookup(sw_context->tfile, handle, &vmw_bo);
+	ret = vmw_user_dmabuf_lookup(sw_context->fp->tfile, handle, &vmw_bo);
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("Could not find or use GMR region.\n");
 		return -EINVAL;
@@ -718,7 +951,7 @@ static int vmw_translate_guest_ptr(struct vmw_private *dev_priv,
 	reloc = &sw_context->relocs[sw_context->cur_reloc++];
 	reloc->location = ptr;
 
-	ret = vmw_bo_to_validate_list(sw_context, bo, &reloc->index);
+	ret = vmw_bo_to_validate_list(sw_context, bo, false, &reloc->index);
 	if (unlikely(ret != 0))
 		goto out_no_reloc;
 
@@ -732,6 +965,30 @@ out_no_reloc:
 }
 
 /**
+ * vmw_cmd_begin_gb_query - validate a  SVGA_3D_CMD_BEGIN_GB_QUERY command.
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context used for this command submission.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_begin_gb_query(struct vmw_private *dev_priv,
+				  struct vmw_sw_context *sw_context,
+				  SVGA3dCmdHeader *header)
+{
+	struct vmw_begin_gb_query_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBeginGBQuery q;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_begin_gb_query_cmd,
+			   header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				 user_context_converter, &cmd->q.cid,
+				 NULL);
+}
+
+/**
  * vmw_cmd_begin_query - validate a  SVGA_3D_CMD_BEGIN_QUERY command.
  *
  * @dev_priv: Pointer to a device private struct.
@@ -750,12 +1007,64 @@ static int vmw_cmd_begin_query(struct vmw_private *dev_priv,
 	cmd = container_of(header, struct vmw_begin_query_cmd,
 			   header);
 
+	if (unlikely(dev_priv->has_mob)) {
+		struct {
+			SVGA3dCmdHeader header;
+			SVGA3dCmdBeginGBQuery q;
+		} gb_cmd;
+
+		BUG_ON(sizeof(gb_cmd) != sizeof(*cmd));
+
+		gb_cmd.header.id = SVGA_3D_CMD_BEGIN_GB_QUERY;
+		gb_cmd.header.size = cmd->header.size;
+		gb_cmd.q.cid = cmd->q.cid;
+		gb_cmd.q.type = cmd->q.type;
+
+		memcpy(cmd, &gb_cmd, sizeof(*cmd));
+		return vmw_cmd_begin_gb_query(dev_priv, sw_context, header);
+	}
+
 	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
 				 user_context_converter, &cmd->q.cid,
 				 NULL);
 }
 
 /**
+ * vmw_cmd_end_gb_query - validate a  SVGA_3D_CMD_END_GB_QUERY command.
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context used for this command submission.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_end_gb_query(struct vmw_private *dev_priv,
+				struct vmw_sw_context *sw_context,
+				SVGA3dCmdHeader *header)
+{
+	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_query_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdEndGBQuery q;
+	} *cmd;
+	int ret;
+
+	cmd = container_of(header, struct vmw_query_cmd, header);
+	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_translate_mob_ptr(dev_priv, sw_context,
+				    &cmd->q.mobid,
+				    &vmw_bo);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_query_bo_switch_prepare(dev_priv, &vmw_bo->base, sw_context);
+
+	vmw_dmabuf_unreference(&vmw_bo);
+	return ret;
+}
+
+/**
  * vmw_cmd_end_query - validate a  SVGA_3D_CMD_END_QUERY command.
  *
  * @dev_priv: Pointer to a device private struct.
@@ -774,6 +1083,25 @@ static int vmw_cmd_end_query(struct vmw_private *dev_priv,
 	int ret;
 
 	cmd = container_of(header, struct vmw_query_cmd, header);
+	if (dev_priv->has_mob) {
+		struct {
+			SVGA3dCmdHeader header;
+			SVGA3dCmdEndGBQuery q;
+		} gb_cmd;
+
+		BUG_ON(sizeof(gb_cmd) != sizeof(*cmd));
+
+		gb_cmd.header.id = SVGA_3D_CMD_END_GB_QUERY;
+		gb_cmd.header.size = cmd->header.size;
+		gb_cmd.q.cid = cmd->q.cid;
+		gb_cmd.q.type = cmd->q.type;
+		gb_cmd.q.mobid = cmd->q.guestResult.gmrId;
+		gb_cmd.q.offset = cmd->q.guestResult.offset;
+
+		memcpy(cmd, &gb_cmd, sizeof(*cmd));
+		return vmw_cmd_end_gb_query(dev_priv, sw_context, header);
+	}
+
 	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
 	if (unlikely(ret != 0))
 		return ret;
@@ -790,7 +1118,40 @@ static int vmw_cmd_end_query(struct vmw_private *dev_priv,
 	return ret;
 }
 
-/*
+/**
+ * vmw_cmd_wait_gb_query - validate a  SVGA_3D_CMD_WAIT_GB_QUERY command.
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context used for this command submission.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_wait_gb_query(struct vmw_private *dev_priv,
+				 struct vmw_sw_context *sw_context,
+				 SVGA3dCmdHeader *header)
+{
+	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_query_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdWaitForGBQuery q;
+	} *cmd;
+	int ret;
+
+	cmd = container_of(header, struct vmw_query_cmd, header);
+	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_translate_mob_ptr(dev_priv, sw_context,
+				    &cmd->q.mobid,
+				    &vmw_bo);
+	if (unlikely(ret != 0))
+		return ret;
+
+	vmw_dmabuf_unreference(&vmw_bo);
+	return 0;
+}
+
+/**
  * vmw_cmd_wait_query - validate a  SVGA_3D_CMD_WAIT_QUERY command.
  *
  * @dev_priv: Pointer to a device private struct.
@@ -809,6 +1170,25 @@ static int vmw_cmd_wait_query(struct vmw_private *dev_priv,
 	int ret;
 
 	cmd = container_of(header, struct vmw_query_cmd, header);
+	if (dev_priv->has_mob) {
+		struct {
+			SVGA3dCmdHeader header;
+			SVGA3dCmdWaitForGBQuery q;
+		} gb_cmd;
+
+		BUG_ON(sizeof(gb_cmd) != sizeof(*cmd));
+
+		gb_cmd.header.id = SVGA_3D_CMD_WAIT_FOR_GB_QUERY;
+		gb_cmd.header.size = cmd->header.size;
+		gb_cmd.q.cid = cmd->q.cid;
+		gb_cmd.q.type = cmd->q.type;
+		gb_cmd.q.mobid = cmd->q.guestResult.gmrId;
+		gb_cmd.q.offset = cmd->q.guestResult.offset;
+
+		memcpy(cmd, &gb_cmd, sizeof(*cmd));
+		return vmw_cmd_wait_gb_query(dev_priv, sw_context, header);
+	}
+
 	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
 	if (unlikely(ret != 0))
 		return ret;
@@ -834,14 +1214,36 @@ static int vmw_cmd_dma(struct vmw_private *dev_priv,
 		SVGA3dCmdSurfaceDMA dma;
 	} *cmd;
 	int ret;
+	SVGA3dCmdSurfaceDMASuffix *suffix;
+	uint32_t bo_size;
 
 	cmd = container_of(header, struct vmw_dma_cmd, header);
+	suffix = (SVGA3dCmdSurfaceDMASuffix *)((unsigned long) &cmd->dma +
+					       header->size - sizeof(*suffix));
+
+	/* Make sure device and verifier stays in sync. */
+	if (unlikely(suffix->suffixSize != sizeof(*suffix))) {
+		DRM_ERROR("Invalid DMA suffix size.\n");
+		return -EINVAL;
+	}
+
 	ret = vmw_translate_guest_ptr(dev_priv, sw_context,
 				      &cmd->dma.guest.ptr,
 				      &vmw_bo);
 	if (unlikely(ret != 0))
 		return ret;
 
+	/* Make sure DMA doesn't cross BO boundaries. */
+	bo_size = vmw_bo->base.num_pages * PAGE_SIZE;
+	if (unlikely(cmd->dma.guest.ptr.offset > bo_size)) {
+		DRM_ERROR("Invalid DMA offset.\n");
+		return -EINVAL;
+	}
+
+	bo_size -= cmd->dma.guest.ptr.offset;
+	if (unlikely(suffix->maximumOffset > bo_size))
+		suffix->maximumOffset = bo_size;
+
 	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 				user_surface_converter, &cmd->dma.host.sid,
 				NULL);
@@ -853,7 +1255,8 @@ static int vmw_cmd_dma(struct vmw_private *dev_priv,
 
 	srf = vmw_res_to_srf(sw_context->res_cache[vmw_res_surface].res);
 
-	vmw_kms_cursor_snoop(srf, sw_context->tfile, &vmw_bo->base, header);
+	vmw_kms_cursor_snoop(srf, sw_context->fp->tfile, &vmw_bo->base,
+			     header);
 
 out_no_surface:
 	vmw_dmabuf_unreference(&vmw_bo);
@@ -921,15 +1324,22 @@ static int vmw_cmd_tex_state(struct vmw_private *dev_priv,
 	struct vmw_tex_state_cmd {
 		SVGA3dCmdHeader header;
 		SVGA3dCmdSetTextureState state;
-	};
+	} *cmd;
 
 	SVGA3dTextureState *last_state = (SVGA3dTextureState *)
 	  ((unsigned long) header + header->size + sizeof(header));
 	SVGA3dTextureState *cur_state = (SVGA3dTextureState *)
 		((unsigned long) header + sizeof(struct vmw_tex_state_cmd));
+	struct vmw_resource_val_node *ctx_node;
+	struct vmw_resource_val_node *res_node;
 	int ret;
 
-	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	cmd = container_of(header, struct vmw_tex_state_cmd,
+			   header);
+
+	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				user_context_converter, &cmd->state.cid,
+				&ctx_node);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -939,9 +1349,20 @@ static int vmw_cmd_tex_state(struct vmw_private *dev_priv,
 
 		ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 					user_surface_converter,
-					&cur_state->value, NULL);
+					&cur_state->value, &res_node);
 		if (unlikely(ret != 0))
 			return ret;
+
+		if (dev_priv->has_mob) {
+			struct vmw_ctx_bindinfo bi;
+
+			bi.ctx = ctx_node->res;
+			bi.res = res_node ? res_node->res : NULL;
+			bi.bt = vmw_ctx_binding_tex;
+			bi.i1.texture_stage = cur_state->stage;
+			vmw_context_binding_add(ctx_node->staged_bindings,
+						&bi);
+		}
 	}
 
 	return 0;
@@ -971,6 +1392,314 @@ static int vmw_cmd_check_define_gmrfb(struct vmw_private *dev_priv,
 }
 
 /**
+ * vmw_cmd_switch_backup - Utility function to handle backup buffer switching
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @res_type: The resource type.
+ * @converter: Information about user-space binding for this resource type.
+ * @res_id: Pointer to the user-space resource handle in the command stream.
+ * @buf_id: Pointer to the user-space backup buffer handle in the command
+ * stream.
+ * @backup_offset: Offset of backup into MOB.
+ *
+ * This function prepares for registering a switch of backup buffers
+ * in the resource metadata just prior to unreserving.
+ */
+static int vmw_cmd_switch_backup(struct vmw_private *dev_priv,
+				 struct vmw_sw_context *sw_context,
+				 enum vmw_res_type res_type,
+				 const struct vmw_user_resource_conv
+				 *converter,
+				 uint32_t *res_id,
+				 uint32_t *buf_id,
+				 unsigned long backup_offset)
+{
+	int ret;
+	struct vmw_dma_buffer *dma_buf;
+	struct vmw_resource_val_node *val_node;
+
+	ret = vmw_cmd_res_check(dev_priv, sw_context, res_type,
+				converter, res_id, &val_node);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_translate_mob_ptr(dev_priv, sw_context, buf_id, &dma_buf);
+	if (unlikely(ret != 0))
+		return ret;
+
+	if (val_node->first_usage)
+		val_node->no_buffer_needed = true;
+
+	vmw_dmabuf_unreference(&val_node->new_backup);
+	val_node->new_backup = dma_buf;
+	val_node->new_backup_offset = backup_offset;
+
+	return 0;
+}
+
+/**
+ * vmw_cmd_bind_gb_surface - Validate an SVGA_3D_CMD_BIND_GB_SURFACE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_bind_gb_surface(struct vmw_private *dev_priv,
+				   struct vmw_sw_context *sw_context,
+				   SVGA3dCmdHeader *header)
+{
+	struct vmw_bind_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBSurface body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_bind_gb_surface_cmd, header);
+
+	return vmw_cmd_switch_backup(dev_priv, sw_context, vmw_res_surface,
+				     user_surface_converter,
+				     &cmd->body.sid, &cmd->body.mobid,
+				     0);
+}
+
+/**
+ * vmw_cmd_update_gb_image - Validate an SVGA_3D_CMD_UPDATE_GB_IMAGE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_update_gb_image(struct vmw_private *dev_priv,
+				   struct vmw_sw_context *sw_context,
+				   SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdUpdateGBImage body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.image.sid, NULL);
+}
+
+/**
+ * vmw_cmd_update_gb_surface - Validate an SVGA_3D_CMD_UPDATE_GB_SURFACE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_update_gb_surface(struct vmw_private *dev_priv,
+				     struct vmw_sw_context *sw_context,
+				     SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdUpdateGBSurface body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.sid, NULL);
+}
+
+/**
+ * vmw_cmd_readback_gb_image - Validate an SVGA_3D_CMD_READBACK_GB_IMAGE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_readback_gb_image(struct vmw_private *dev_priv,
+				     struct vmw_sw_context *sw_context,
+				     SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdReadbackGBImage body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.image.sid, NULL);
+}
+
+/**
+ * vmw_cmd_readback_gb_surface - Validate an SVGA_3D_CMD_READBACK_GB_SURFACE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_readback_gb_surface(struct vmw_private *dev_priv,
+				       struct vmw_sw_context *sw_context,
+				       SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdReadbackGBSurface body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.sid, NULL);
+}
+
+/**
+ * vmw_cmd_invalidate_gb_image - Validate an SVGA_3D_CMD_INVALIDATE_GB_IMAGE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_invalidate_gb_image(struct vmw_private *dev_priv,
+				       struct vmw_sw_context *sw_context,
+				       SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdInvalidateGBImage body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.image.sid, NULL);
+}
+
+/**
+ * vmw_cmd_invalidate_gb_surface - Validate an
+ * SVGA_3D_CMD_INVALIDATE_GB_SURFACE command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_invalidate_gb_surface(struct vmw_private *dev_priv,
+					 struct vmw_sw_context *sw_context,
+					 SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdInvalidateGBSurface body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.sid, NULL);
+}
+
+
+/**
+ * vmw_cmd_shader_define - Validate an SVGA_3D_CMD_SHADER_DEFINE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_shader_define(struct vmw_private *dev_priv,
+				 struct vmw_sw_context *sw_context,
+				 SVGA3dCmdHeader *header)
+{
+	struct vmw_shader_define_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineShader body;
+	} *cmd;
+	int ret;
+	size_t size;
+
+	cmd = container_of(header, struct vmw_shader_define_cmd,
+			   header);
+
+	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				user_context_converter, &cmd->body.cid,
+				NULL);
+	if (unlikely(ret != 0))
+		return ret;
+
+	if (unlikely(!dev_priv->has_mob))
+		return 0;
+
+	size = cmd->header.size - sizeof(cmd->body);
+	ret = vmw_compat_shader_add(sw_context->fp->shman,
+				    cmd->body.shid, cmd + 1,
+				    cmd->body.type, size,
+				    sw_context->fp->tfile,
+				    &sw_context->staged_shaders);
+	if (unlikely(ret != 0))
+		return ret;
+
+	return vmw_resource_relocation_add(&sw_context->res_relocations,
+					   NULL, &cmd->header.id -
+					   sw_context->buf_start);
+
+	return 0;
+}
+
+/**
+ * vmw_cmd_shader_destroy - Validate an SVGA_3D_CMD_SHADER_DESTROY
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_shader_destroy(struct vmw_private *dev_priv,
+				  struct vmw_sw_context *sw_context,
+				  SVGA3dCmdHeader *header)
+{
+	struct vmw_shader_destroy_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDestroyShader body;
+	} *cmd;
+	int ret;
+
+	cmd = container_of(header, struct vmw_shader_destroy_cmd,
+			   header);
+
+	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				user_context_converter, &cmd->body.cid,
+				NULL);
+	if (unlikely(ret != 0))
+		return ret;
+
+	if (unlikely(!dev_priv->has_mob))
+		return 0;
+
+	ret = vmw_compat_shader_remove(sw_context->fp->shman,
+				       cmd->body.shid,
+				       cmd->body.type,
+				       &sw_context->staged_shaders);
+	if (unlikely(ret != 0))
+		return ret;
+
+	return vmw_resource_relocation_add(&sw_context->res_relocations,
+					   NULL, &cmd->header.id -
+					   sw_context->buf_start);
+
+	return 0;
+}
+
+/**
  * vmw_cmd_set_shader - Validate an SVGA_3D_CMD_SET_SHADER
  * command
  *
@@ -986,18 +1715,105 @@ static int vmw_cmd_set_shader(struct vmw_private *dev_priv,
 		SVGA3dCmdHeader header;
 		SVGA3dCmdSetShader body;
 	} *cmd;
+	struct vmw_resource_val_node *ctx_node;
 	int ret;
 
 	cmd = container_of(header, struct vmw_set_shader_cmd,
 			   header);
 
-	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				user_context_converter, &cmd->body.cid,
+				&ctx_node);
 	if (unlikely(ret != 0))
 		return ret;
 
+	if (dev_priv->has_mob) {
+		struct vmw_ctx_bindinfo bi;
+		struct vmw_resource_val_node *res_node;
+		u32 shid = cmd->body.shid;
+
+		if (shid != SVGA3D_INVALID_ID)
+			(void) vmw_compat_shader_lookup(sw_context->fp->shman,
+							cmd->body.type,
+							&shid);
+
+		ret = vmw_cmd_compat_res_check(dev_priv, sw_context,
+					       vmw_res_shader,
+					       user_shader_converter,
+					       shid,
+					       &cmd->body.shid, &res_node);
+		if (unlikely(ret != 0))
+			return ret;
+
+		bi.ctx = ctx_node->res;
+		bi.res = res_node ? res_node->res : NULL;
+		bi.bt = vmw_ctx_binding_shader;
+		bi.i1.shader_type = cmd->body.type;
+		return vmw_context_binding_add(ctx_node->staged_bindings, &bi);
+	}
+
 	return 0;
 }
 
+/**
+ * vmw_cmd_set_shader_const - Validate an SVGA_3D_CMD_SET_SHADER_CONST
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_set_shader_const(struct vmw_private *dev_priv,
+				    struct vmw_sw_context *sw_context,
+				    SVGA3dCmdHeader *header)
+{
+	struct vmw_set_shader_const_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdSetShaderConst body;
+	} *cmd;
+	int ret;
+
+	cmd = container_of(header, struct vmw_set_shader_const_cmd,
+			   header);
+
+	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				user_context_converter, &cmd->body.cid,
+				NULL);
+	if (unlikely(ret != 0))
+		return ret;
+
+	if (dev_priv->has_mob)
+		header->id = SVGA_3D_CMD_SET_GB_SHADERCONSTS_INLINE;
+
+	return 0;
+}
+
+/**
+ * vmw_cmd_bind_gb_shader - Validate an SVGA_3D_CMD_BIND_GB_SHADER
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_bind_gb_shader(struct vmw_private *dev_priv,
+				  struct vmw_sw_context *sw_context,
+				  SVGA3dCmdHeader *header)
+{
+	struct vmw_bind_gb_shader_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBShader body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_bind_gb_shader_cmd,
+			   header);
+
+	return vmw_cmd_switch_backup(dev_priv, sw_context, vmw_res_shader,
+				     user_shader_converter,
+				     &cmd->body.shid, &cmd->body.mobid,
+				     cmd->body.offsetInBytes);
+}
+
 static int vmw_cmd_check_not_3d(struct vmw_private *dev_priv,
 				struct vmw_sw_context *sw_context,
 				void *buf, uint32_t *size)
@@ -1041,50 +1857,173 @@ static int vmw_cmd_check_not_3d(struct vmw_private *dev_priv,
 	return 0;
 }
 
-typedef int (*vmw_cmd_func) (struct vmw_private *,
-			     struct vmw_sw_context *,
-			     SVGA3dCmdHeader *);
-
-#define VMW_CMD_DEF(cmd, func) \
-	[cmd - SVGA_3D_CMD_BASE] = func
-
-static vmw_cmd_func vmw_cmd_funcs[SVGA_3D_CMD_MAX] = {
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DEFINE, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DESTROY, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_COPY, &vmw_cmd_surface_copy_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_STRETCHBLT, &vmw_cmd_stretch_blt_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DMA, &vmw_cmd_dma),
-	VMW_CMD_DEF(SVGA_3D_CMD_CONTEXT_DEFINE, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_CONTEXT_DESTROY, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETTRANSFORM, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETZRANGE, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETRENDERSTATE, &vmw_cmd_cid_check),
+static const struct vmw_cmd_entry vmw_cmd_entries[SVGA_3D_CMD_MAX] = {
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DEFINE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DESTROY, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_COPY, &vmw_cmd_surface_copy_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_STRETCHBLT, &vmw_cmd_stretch_blt_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DMA, &vmw_cmd_dma,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_CONTEXT_DEFINE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_CONTEXT_DESTROY, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETTRANSFORM, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETZRANGE, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETRENDERSTATE, &vmw_cmd_cid_check,
+		    true, false, false),
 	VMW_CMD_DEF(SVGA_3D_CMD_SETRENDERTARGET,
-		    &vmw_cmd_set_render_target_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETTEXTURESTATE, &vmw_cmd_tex_state),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETMATERIAL, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETLIGHTDATA, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETLIGHTENABLED, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETVIEWPORT, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETCLIPPLANE, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_CLEAR, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_PRESENT, &vmw_cmd_present_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SHADER_DEFINE, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SHADER_DESTROY, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SET_SHADER, &vmw_cmd_set_shader),
-	VMW_CMD_DEF(SVGA_3D_CMD_SET_SHADER_CONST, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_DRAW_PRIMITIVES, &vmw_cmd_draw),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETSCISSORRECT, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_BEGIN_QUERY, &vmw_cmd_begin_query),
-	VMW_CMD_DEF(SVGA_3D_CMD_END_QUERY, &vmw_cmd_end_query),
-	VMW_CMD_DEF(SVGA_3D_CMD_WAIT_FOR_QUERY, &vmw_cmd_wait_query),
-	VMW_CMD_DEF(SVGA_3D_CMD_PRESENT_READBACK, &vmw_cmd_ok),
+		    &vmw_cmd_set_render_target_check, true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETTEXTURESTATE, &vmw_cmd_tex_state,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETMATERIAL, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETLIGHTDATA, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETLIGHTENABLED, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETVIEWPORT, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETCLIPPLANE, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_CLEAR, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_PRESENT, &vmw_cmd_present_check,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SHADER_DEFINE, &vmw_cmd_shader_define,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SHADER_DESTROY, &vmw_cmd_shader_destroy,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_SHADER, &vmw_cmd_set_shader,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_SHADER_CONST, &vmw_cmd_set_shader_const,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_DRAW_PRIMITIVES, &vmw_cmd_draw,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETSCISSORRECT, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_BEGIN_QUERY, &vmw_cmd_begin_query,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_END_QUERY, &vmw_cmd_end_query,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_WAIT_FOR_QUERY, &vmw_cmd_wait_query,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_PRESENT_READBACK, &vmw_cmd_ok,
+		    true, false, false),
 	VMW_CMD_DEF(SVGA_3D_CMD_BLIT_SURFACE_TO_SCREEN,
-		    &vmw_cmd_blt_surf_screen_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DEFINE_V2, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_GENERATE_MIPMAPS, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_ACTIVATE_SURFACE, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_DEACTIVATE_SURFACE, &vmw_cmd_invalid),
+		    &vmw_cmd_blt_surf_screen_check, false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DEFINE_V2, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_GENERATE_MIPMAPS, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_ACTIVATE_SURFACE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEACTIVATE_SURFACE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SCREEN_DMA, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_UNITY_SURFACE_COOKIE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_OPEN_CONTEXT_SURFACE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_BITBLT, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_TRANSBLT, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_STRETCHBLT, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_COLORFILL, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_ALPHABLEND, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_CLEARTYPEBLEND, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_OTABLE_BASE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_OTABLE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_MOB, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_MOB, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_REDEFINE_GB_MOB, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UPDATE_GB_MOB_MAPPING, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_SURFACE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_SURFACE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_BIND_GB_SURFACE, &vmw_cmd_bind_gb_surface,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_COND_BIND_GB_SURFACE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UPDATE_GB_IMAGE, &vmw_cmd_update_gb_image,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UPDATE_GB_SURFACE,
+		    &vmw_cmd_update_gb_surface, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_GB_IMAGE,
+		    &vmw_cmd_readback_gb_image, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_GB_SURFACE,
+		    &vmw_cmd_readback_gb_surface, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_INVALIDATE_GB_IMAGE,
+		    &vmw_cmd_invalidate_gb_image, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_INVALIDATE_GB_SURFACE,
+		    &vmw_cmd_invalidate_gb_surface, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_BIND_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_INVALIDATE_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_SHADER, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_BIND_GB_SHADER, &vmw_cmd_bind_gb_shader,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_SHADER, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_OTABLE_BASE64, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_BEGIN_GB_QUERY, &vmw_cmd_begin_gb_query,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_END_GB_QUERY, &vmw_cmd_end_gb_query,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_WAIT_FOR_GB_QUERY, &vmw_cmd_wait_gb_query,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_NOP, &vmw_cmd_ok,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_ENABLE_GART, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DISABLE_GART, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_MAP_MOB_INTO_GART, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UNMAP_GART_RANGE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_SCREENTARGET, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_SCREENTARGET, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_BIND_GB_SCREENTARGET, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UPDATE_GB_SCREENTARGET, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_GB_IMAGE_PARTIAL, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_INVALIDATE_GB_IMAGE_PARTIAL, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_GB_SHADERCONSTS_INLINE, &vmw_cmd_cid_check,
+		    true, false, true)
 };
 
 static int vmw_cmd_check(struct vmw_private *dev_priv,
@@ -1095,6 +2034,8 @@ static int vmw_cmd_check(struct vmw_private *dev_priv,
 	uint32_t size_remaining = *size;
 	SVGA3dCmdHeader *header = (SVGA3dCmdHeader *) buf;
 	int ret;
+	const struct vmw_cmd_entry *entry;
+	bool gb = dev_priv->capabilities & SVGA_CAP_GBOBJECTS;
 
 	cmd_id = le32_to_cpu(((uint32_t *)buf)[0]);
 	/* Handle any none 3D commands */
@@ -1107,18 +2048,43 @@ static int vmw_cmd_check(struct vmw_private *dev_priv,
 
 	cmd_id -= SVGA_3D_CMD_BASE;
 	if (unlikely(*size > size_remaining))
-		goto out_err;
+		goto out_invalid;
 
 	if (unlikely(cmd_id >= SVGA_3D_CMD_MAX - SVGA_3D_CMD_BASE))
-		goto out_err;
+		goto out_invalid;
+
+	entry = &vmw_cmd_entries[cmd_id];
+	if (unlikely(!entry->func))
+		goto out_invalid;
+
+	if (unlikely(!entry->user_allow && !sw_context->kernel))
+		goto out_privileged;
 
-	ret = vmw_cmd_funcs[cmd_id](dev_priv, sw_context, header);
+	if (unlikely(entry->gb_disable && gb))
+		goto out_old;
+
+	if (unlikely(entry->gb_enable && !gb))
+		goto out_new;
+
+	ret = entry->func(dev_priv, sw_context, header);
 	if (unlikely(ret != 0))
-		goto out_err;
+		goto out_invalid;
 
 	return 0;
-out_err:
-	DRM_ERROR("Illegal / Invalid SVGA3D command: %d\n",
+out_invalid:
+	DRM_ERROR("Invalid SVGA3D command: %d\n",
+		  cmd_id + SVGA_3D_CMD_BASE);
+	return -EINVAL;
+out_privileged:
+	DRM_ERROR("Privileged SVGA3D command: %d\n",
+		  cmd_id + SVGA_3D_CMD_BASE);
+	return -EPERM;
+out_old:
+	DRM_ERROR("Deprecated (disallowed) SVGA3D command: %d\n",
+		  cmd_id + SVGA_3D_CMD_BASE);
+	return -EINVAL;
+out_new:
+	DRM_ERROR("SVGA3D command: %d not supported by virtual hardware.\n",
 		  cmd_id + SVGA_3D_CMD_BASE);
 	return -EINVAL;
 }
@@ -1174,6 +2140,9 @@ static void vmw_apply_relocations(struct vmw_sw_context *sw_context)
 		case VMW_PL_GMR:
 			reloc->location->gmrId = bo->mem.start;
 			break;
+		case VMW_PL_MOB:
+			*reloc->mob_loc = bo->mem.start;
+			break;
 		default:
 			BUG();
 		}
@@ -1198,6 +2167,8 @@ static void vmw_resource_list_unreference(struct list_head *list)
 	list_for_each_entry_safe(val, val_next, list, head) {
 		list_del_init(&val->head);
 		vmw_resource_unreference(&val->res);
+		if (unlikely(val->staged_bindings))
+			kfree(val->staged_bindings);
 		kfree(val);
 	}
 }
@@ -1224,7 +2195,8 @@ static void vmw_clear_validations(struct vmw_sw_context *sw_context)
 }
 
 static int vmw_validate_single_buffer(struct vmw_private *dev_priv,
-				      struct ttm_buffer_object *bo)
+				      struct ttm_buffer_object *bo,
+				      bool validate_as_mob)
 {
 	int ret;
 
@@ -1238,6 +2210,9 @@ static int vmw_validate_single_buffer(struct vmw_private *dev_priv,
 	     dev_priv->dummy_query_bo_pinned))
 		return 0;
 
+	if (validate_as_mob)
+		return ttm_bo_validate(bo, &vmw_mob_placement, true, false);
+
 	/**
 	 * Put BO in VRAM if there is space, otherwise as a GMR.
 	 * If there is no space in VRAM and GMR ids are all used up,
@@ -1259,7 +2234,6 @@ static int vmw_validate_single_buffer(struct vmw_private *dev_priv,
 	return ret;
 }
 
-
 static int vmw_validate_buffers(struct vmw_private *dev_priv,
 				struct vmw_sw_context *sw_context)
 {
@@ -1267,7 +2241,8 @@ static int vmw_validate_buffers(struct vmw_private *dev_priv,
 	int ret;
 
 	list_for_each_entry(entry, &sw_context->validate_nodes, base.head) {
-		ret = vmw_validate_single_buffer(dev_priv, entry->base.bo);
+		ret = vmw_validate_single_buffer(dev_priv, entry->base.bo,
+						 entry->validate_as_mob);
 		if (unlikely(ret != 0))
 			return ret;
 	}
@@ -1461,7 +2436,7 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 	} else
 		sw_context->kernel = true;
 
-	sw_context->tfile = vmw_fpriv(file_priv)->tfile;
+	sw_context->fp = vmw_fpriv(file_priv);
 	sw_context->cur_reloc = 0;
 	sw_context->cur_val_buf = 0;
 	sw_context->fence_flags = 0;
@@ -1478,16 +2453,17 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 			goto out_unlock;
 		sw_context->res_ht_initialized = true;
 	}
+	INIT_LIST_HEAD(&sw_context->staged_shaders);
 
 	INIT_LIST_HEAD(&resource_list);
 	ret = vmw_cmd_check_all(dev_priv, sw_context, kernel_commands,
 				command_size);
 	if (unlikely(ret != 0))
-		goto out_err;
+		goto out_err_nores;
 
 	ret = vmw_resources_reserve(sw_context);
 	if (unlikely(ret != 0))
-		goto out_err;
+		goto out_err_nores;
 
 	ret = ttm_eu_reserve_buffers(&ticket, &sw_context->validate_nodes);
 	if (unlikely(ret != 0))
@@ -1509,11 +2485,23 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 			goto out_err;
 	}
 
+	ret = mutex_lock_interruptible(&dev_priv->binding_mutex);
+	if (unlikely(ret != 0)) {
+		ret = -ERESTARTSYS;
+		goto out_err;
+	}
+
+	if (dev_priv->has_mob) {
+		ret = vmw_rebind_contexts(sw_context);
+		if (unlikely(ret != 0))
+			goto out_unlock_binding;
+	}
+
 	cmd = vmw_fifo_reserve(dev_priv, command_size);
 	if (unlikely(cmd == NULL)) {
 		DRM_ERROR("Failed reserving fifo space for commands.\n");
 		ret = -ENOMEM;
-		goto out_err;
+		goto out_unlock_binding;
 	}
 
 	vmw_apply_relocations(sw_context);
@@ -1538,6 +2526,8 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 		DRM_ERROR("Fence submission error. Syncing.\n");
 
 	vmw_resource_list_unreserve(&sw_context->resource_list, false);
+	mutex_unlock(&dev_priv->binding_mutex);
+
 	ttm_eu_fence_buffer_objects(&ticket, &sw_context->validate_nodes,
 				    (void *) fence);
 
@@ -1558,6 +2548,8 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 	}
 
 	list_splice_init(&sw_context->resource_list, &resource_list);
+	vmw_compat_shaders_commit(sw_context->fp->shman,
+				  &sw_context->staged_shaders);
 	mutex_unlock(&dev_priv->cmdbuf_mutex);
 
 	/*
@@ -1568,11 +2560,14 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 
 	return 0;
 
+out_unlock_binding:
+	mutex_unlock(&dev_priv->binding_mutex);
 out_err:
-	vmw_resource_relocations_free(&sw_context->res_relocations);
-	vmw_free_relocations(sw_context);
 	ttm_eu_backoff_reservation(&ticket, &sw_context->validate_nodes);
+out_err_nores:
 	vmw_resource_list_unreserve(&sw_context->resource_list, true);
+	vmw_resource_relocations_free(&sw_context->res_relocations);
+	vmw_free_relocations(sw_context);
 	vmw_clear_validations(sw_context);
 	if (unlikely(dev_priv->pinned_bo != NULL &&
 		     !dev_priv->query_cid_valid))
@@ -1581,6 +2576,8 @@ out_unlock:
 	list_splice_init(&sw_context->resource_list, &resource_list);
 	error_resource = sw_context->error_resource;
 	sw_context->error_resource = NULL;
+	vmw_compat_shaders_revert(sw_context->fp->shman,
+				  &sw_context->staged_shaders);
 	mutex_unlock(&dev_priv->cmdbuf_mutex);
 
 	/*
@@ -1737,7 +2734,6 @@ int vmw_execbuf_ioctl(struct drm_device *dev, void *data,
 {
 	struct vmw_private *dev_priv = vmw_priv(dev);
 	struct drm_vmw_execbuf_arg *arg = (struct drm_vmw_execbuf_arg *)data;
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	int ret;
 
 	/*
@@ -1754,7 +2750,7 @@ int vmw_execbuf_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
-	ret = ttm_read_lock(&vmaster->lock, true);
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -1770,6 +2766,6 @@ int vmw_execbuf_ioctl(struct drm_device *dev, void *data,
 	vmw_kms_cursor_post_execbuf(dev_priv);
 
 out_unlock:
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
 	return ret;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
index ed5ce2a41bb..b031b48dbb3 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
@@ -147,7 +147,7 @@ static int vmw_fb_check_var(struct fb_var_screeninfo *var,
 	}
 
 	if (!vmw_kms_validate_mode_vram(vmw_priv,
-					info->fix.line_length,
+					var->xres * var->bits_per_pixel/8,
 					var->yoffset + var->yres)) {
 		DRM_ERROR("Requested geom can not fit in framebuffer\n");
 		return -EINVAL;
@@ -162,6 +162,8 @@ static int vmw_fb_set_par(struct fb_info *info)
 	struct vmw_private *vmw_priv = par->vmw_priv;
 	int ret;
 
+	info->fix.line_length = info->var.xres * info->var.bits_per_pixel/8;
+
 	ret = vmw_kms_write_svga(vmw_priv, info->var.xres, info->var.yres,
 				 info->fix.line_length,
 				 par->bpp, par->depth);
@@ -377,14 +379,13 @@ static int vmw_fb_create_bo(struct vmw_private *vmw_priv,
 
 	ne_placement.lpfn = (size + PAGE_SIZE - 1) >> PAGE_SHIFT;
 
-	/* interuptable? */
-	ret = ttm_write_lock(&vmw_priv->fbdev_master.lock, false);
-	if (unlikely(ret != 0))
-		return ret;
+	(void) ttm_write_lock(&vmw_priv->reservation_sem, false);
 
 	vmw_bo = kmalloc(sizeof(*vmw_bo), GFP_KERNEL);
-	if (!vmw_bo)
+	if (!vmw_bo) {
+		ret = -ENOMEM;
 		goto err_unlock;
+	}
 
 	ret = vmw_dmabuf_init(vmw_priv, vmw_bo, size,
 			      &ne_placement,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
index c62d20e8a6f..436b013b423 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
@@ -271,7 +271,7 @@ void vmw_fence_obj_unreference(struct vmw_fence_obj **fence_p)
 	spin_unlock_irq(&fman->lock);
 }
 
-void vmw_fences_perform_actions(struct vmw_fence_manager *fman,
+static void vmw_fences_perform_actions(struct vmw_fence_manager *fman,
 				struct list_head *list)
 {
 	struct vmw_fence_action *action, *next_action;
@@ -897,7 +897,7 @@ static void vmw_event_fence_action_cleanup(struct vmw_fence_action *action)
  * Note that the action callbacks may be executed before this function
  * returns.
  */
-void vmw_fence_obj_add_action(struct vmw_fence_obj *fence,
+static void vmw_fence_obj_add_action(struct vmw_fence_obj *fence,
 			      struct vmw_fence_action *action)
 {
 	struct vmw_fence_manager *fman = fence->fman;
@@ -993,7 +993,7 @@ struct vmw_event_fence_pending {
 	struct drm_vmw_event_fence event;
 };
 
-int vmw_event_fence_action_create(struct drm_file *file_priv,
+static int vmw_event_fence_action_create(struct drm_file *file_priv,
 				  struct vmw_fence_obj *fence,
 				  uint32_t flags,
 				  uint64_t user_data,
@@ -1080,7 +1080,8 @@ int vmw_fence_event_ioctl(struct drm_device *dev, void *data,
 	 */
 	if (arg->handle) {
 		struct ttm_base_object *base =
-			ttm_base_object_lookup(vmw_fp->tfile, arg->handle);
+			ttm_base_object_lookup_for_ref(dev_priv->tdev,
+						       arg->handle);
 
 		if (unlikely(base == NULL)) {
 			DRM_ERROR("Fence event invalid fence object handle "
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
index 3eb148667d6..6ccd993e26b 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
@@ -35,6 +35,23 @@ bool vmw_fifo_have_3d(struct vmw_private *dev_priv)
 	uint32_t fifo_min, hwversion;
 	const struct vmw_fifo_state *fifo = &dev_priv->fifo;
 
+	if (!(dev_priv->capabilities & SVGA_CAP_3D))
+		return false;
+
+	if (dev_priv->capabilities & SVGA_CAP_GBOBJECTS) {
+		uint32_t result;
+
+		if (!dev_priv->has_mob)
+			return false;
+
+		mutex_lock(&dev_priv->hw_mutex);
+		vmw_write(dev_priv, SVGA_REG_DEV_CAP, SVGA3D_DEVCAP_3D);
+		result = vmw_read(dev_priv, SVGA_REG_DEV_CAP);
+		mutex_unlock(&dev_priv->hw_mutex);
+
+		return (result != 0);
+	}
+
 	if (!(dev_priv->capabilities & SVGA_CAP_EXTENDED_FIFO))
 		return false;
 
@@ -511,24 +528,16 @@ out_err:
 }
 
 /**
- * vmw_fifo_emit_dummy_query - emits a dummy query to the fifo.
+ * vmw_fifo_emit_dummy_legacy_query - emits a dummy query to the fifo using
+ * legacy query commands.
  *
  * @dev_priv: The device private structure.
  * @cid: The hardware context id used for the query.
  *
- * This function is used to emit a dummy occlusion query with
- * no primitives rendered between query begin and query end.
- * It's used to provide a query barrier, in order to know that when
- * this query is finished, all preceding queries are also finished.
- *
- * A Query results structure should have been initialized at the start
- * of the dev_priv->dummy_query_bo buffer object. And that buffer object
- * must also be either reserved or pinned when this function is called.
- *
- * Returns -ENOMEM on failure to reserve fifo space.
+ * See the vmw_fifo_emit_dummy_query documentation.
  */
-int vmw_fifo_emit_dummy_query(struct vmw_private *dev_priv,
-			      uint32_t cid)
+static int vmw_fifo_emit_dummy_legacy_query(struct vmw_private *dev_priv,
+					    uint32_t cid)
 {
 	/*
 	 * A query wait without a preceding query end will
@@ -566,3 +575,75 @@ int vmw_fifo_emit_dummy_query(struct vmw_private *dev_priv,
 
 	return 0;
 }
+
+/**
+ * vmw_fifo_emit_dummy_gb_query - emits a dummy query to the fifo using
+ * guest-backed resource query commands.
+ *
+ * @dev_priv: The device private structure.
+ * @cid: The hardware context id used for the query.
+ *
+ * See the vmw_fifo_emit_dummy_query documentation.
+ */
+static int vmw_fifo_emit_dummy_gb_query(struct vmw_private *dev_priv,
+					uint32_t cid)
+{
+	/*
+	 * A query wait without a preceding query end will
+	 * actually finish all queries for this cid
+	 * without writing to the query result structure.
+	 */
+
+	struct ttm_buffer_object *bo = dev_priv->dummy_query_bo;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdWaitForGBQuery body;
+	} *cmd;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Out of fifo space for dummy query.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_WAIT_FOR_GB_QUERY;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = cid;
+	cmd->body.type = SVGA3D_QUERYTYPE_OCCLUSION;
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+	cmd->body.mobid = bo->mem.start;
+	cmd->body.offset = 0;
+
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+
+/**
+ * vmw_fifo_emit_dummy_gb_query - emits a dummy query to the fifo using
+ * appropriate resource query commands.
+ *
+ * @dev_priv: The device private structure.
+ * @cid: The hardware context id used for the query.
+ *
+ * This function is used to emit a dummy occlusion query with
+ * no primitives rendered between query begin and query end.
+ * It's used to provide a query barrier, in order to know that when
+ * this query is finished, all preceding queries are also finished.
+ *
+ * A Query results structure should have been initialized at the start
+ * of the dev_priv->dummy_query_bo buffer object. And that buffer object
+ * must also be either reserved or pinned when this function is called.
+ *
+ * Returns -ENOMEM on failure to reserve fifo space.
+ */
+int vmw_fifo_emit_dummy_query(struct vmw_private *dev_priv,
+			      uint32_t cid)
+{
+	if (dev_priv->has_mob)
+		return vmw_fifo_emit_dummy_gb_query(dev_priv, cid);
+
+	return vmw_fifo_emit_dummy_legacy_query(dev_priv, cid);
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c b/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
index 6ef0b035bec..61d8d803199 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
@@ -125,181 +125,27 @@ static void vmw_gmr2_unbind(struct vmw_private *dev_priv,
 }
 
 
-static void vmw_gmr_free_descriptors(struct device *dev, dma_addr_t desc_dma,
-				     struct list_head *desc_pages)
-{
-	struct page *page, *next;
-	struct svga_guest_mem_descriptor *page_virtual;
-	unsigned int desc_per_page = PAGE_SIZE /
-		sizeof(struct svga_guest_mem_descriptor) - 1;
-
-	if (list_empty(desc_pages))
-		return;
-
-	list_for_each_entry_safe(page, next, desc_pages, lru) {
-		list_del_init(&page->lru);
-
-		if (likely(desc_dma != DMA_ADDR_INVALID)) {
-			dma_unmap_page(dev, desc_dma, PAGE_SIZE,
-				       DMA_TO_DEVICE);
-		}
-
-		page_virtual = kmap_atomic(page);
-		desc_dma = (dma_addr_t)
-			le32_to_cpu(page_virtual[desc_per_page].ppn) <<
-			PAGE_SHIFT;
-		kunmap_atomic(page_virtual);
-
-		__free_page(page);
-	}
-}
-
-/**
- * FIXME: Adjust to the ttm lowmem / highmem storage to minimize
- * the number of used descriptors.
- *
- */
-
-static int vmw_gmr_build_descriptors(struct device *dev,
-				     struct list_head *desc_pages,
-				     struct vmw_piter *iter,
-				     unsigned long num_pages,
-				     dma_addr_t *first_dma)
-{
-	struct page *page;
-	struct svga_guest_mem_descriptor *page_virtual = NULL;
-	struct svga_guest_mem_descriptor *desc_virtual = NULL;
-	unsigned int desc_per_page;
-	unsigned long prev_pfn;
-	unsigned long pfn;
-	int ret;
-	dma_addr_t desc_dma;
-
-	desc_per_page = PAGE_SIZE /
-	    sizeof(struct svga_guest_mem_descriptor) - 1;
-
-	while (likely(num_pages != 0)) {
-		page = alloc_page(__GFP_HIGHMEM);
-		if (unlikely(page == NULL)) {
-			ret = -ENOMEM;
-			goto out_err;
-		}
-
-		list_add_tail(&page->lru, desc_pages);
-		page_virtual = kmap_atomic(page);
-		desc_virtual = page_virtual - 1;
-		prev_pfn = ~(0UL);
-
-		while (likely(num_pages != 0)) {
-			pfn = vmw_piter_dma_addr(iter) >> PAGE_SHIFT;
-
-			if (pfn != prev_pfn + 1) {
-
-				if (desc_virtual - page_virtual ==
-				    desc_per_page - 1)
-					break;
-
-				(++desc_virtual)->ppn = cpu_to_le32(pfn);
-				desc_virtual->num_pages = cpu_to_le32(1);
-			} else {
-				uint32_t tmp =
-				    le32_to_cpu(desc_virtual->num_pages);
-				desc_virtual->num_pages = cpu_to_le32(tmp + 1);
-			}
-			prev_pfn = pfn;
-			--num_pages;
-			vmw_piter_next(iter);
-		}
-
-		(++desc_virtual)->ppn = DMA_PAGE_INVALID;
-		desc_virtual->num_pages = cpu_to_le32(0);
-		kunmap_atomic(page_virtual);
-	}
-
-	desc_dma = 0;
-	list_for_each_entry_reverse(page, desc_pages, lru) {
-		page_virtual = kmap_atomic(page);
-		page_virtual[desc_per_page].ppn = cpu_to_le32
-			(desc_dma >> PAGE_SHIFT);
-		kunmap_atomic(page_virtual);
-		desc_dma = dma_map_page(dev, page, 0, PAGE_SIZE,
-					DMA_TO_DEVICE);
-
-		if (unlikely(dma_mapping_error(dev, desc_dma)))
-			goto out_err;
-	}
-	*first_dma = desc_dma;
-
-	return 0;
-out_err:
-	vmw_gmr_free_descriptors(dev, DMA_ADDR_INVALID, desc_pages);
-	return ret;
-}
-
-static void vmw_gmr_fire_descriptors(struct vmw_private *dev_priv,
-				     int gmr_id, dma_addr_t desc_dma)
-{
-	mutex_lock(&dev_priv->hw_mutex);
-
-	vmw_write(dev_priv, SVGA_REG_GMR_ID, gmr_id);
-	wmb();
-	vmw_write(dev_priv, SVGA_REG_GMR_DESCRIPTOR, desc_dma >> PAGE_SHIFT);
-	mb();
-
-	mutex_unlock(&dev_priv->hw_mutex);
-
-}
-
 int vmw_gmr_bind(struct vmw_private *dev_priv,
 		 const struct vmw_sg_table *vsgt,
 		 unsigned long num_pages,
 		 int gmr_id)
 {
-	struct list_head desc_pages;
-	dma_addr_t desc_dma = 0;
-	struct device *dev = dev_priv->dev->dev;
 	struct vmw_piter data_iter;
-	int ret;
 
 	vmw_piter_start(&data_iter, vsgt, 0);
 
 	if (unlikely(!vmw_piter_next(&data_iter)))
 		return 0;
 
-	if (likely(dev_priv->capabilities & SVGA_CAP_GMR2))
-		return vmw_gmr2_bind(dev_priv, &data_iter, num_pages, gmr_id);
-
-	if (unlikely(!(dev_priv->capabilities & SVGA_CAP_GMR)))
-		return -EINVAL;
-
-	if (vsgt->num_regions > dev_priv->max_gmr_descriptors)
+	if (unlikely(!(dev_priv->capabilities & SVGA_CAP_GMR2)))
 		return -EINVAL;
 
-	INIT_LIST_HEAD(&desc_pages);
-
-	ret = vmw_gmr_build_descriptors(dev, &desc_pages, &data_iter,
-					num_pages, &desc_dma);
-	if (unlikely(ret != 0))
-		return ret;
-
-	vmw_gmr_fire_descriptors(dev_priv, gmr_id, desc_dma);
-	vmw_gmr_free_descriptors(dev, desc_dma, &desc_pages);
-
-	return 0;
+	return vmw_gmr2_bind(dev_priv, &data_iter, num_pages, gmr_id);
 }
 
 
 void vmw_gmr_unbind(struct vmw_private *dev_priv, int gmr_id)
 {
-	if (likely(dev_priv->capabilities & SVGA_CAP_GMR2)) {
+	if (likely(dev_priv->capabilities & SVGA_CAP_GMR2))
 		vmw_gmr2_unbind(dev_priv, gmr_id);
-		return;
-	}
-
-	mutex_lock(&dev_priv->hw_mutex);
-	vmw_write(dev_priv, SVGA_REG_GMR_ID, gmr_id);
-	wmb();
-	vmw_write(dev_priv, SVGA_REG_GMR_DESCRIPTOR, 0);
-	mb();
-	mutex_unlock(&dev_priv->hw_mutex);
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
index c5c054ae905..b1273e8e9a6 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
@@ -125,10 +125,21 @@ static int vmw_gmrid_man_init(struct ttm_mem_type_manager *man,
 		return -ENOMEM;
 
 	spin_lock_init(&gman->lock);
-	gman->max_gmr_pages = dev_priv->max_gmr_pages;
 	gman->used_gmr_pages = 0;
 	ida_init(&gman->gmr_ida);
-	gman->max_gmr_ids = p_size;
+
+	switch (p_size) {
+	case VMW_PL_GMR:
+		gman->max_gmr_ids = dev_priv->max_gmr_ids;
+		gman->max_gmr_pages = dev_priv->max_gmr_pages;
+		break;
+	case VMW_PL_MOB:
+		gman->max_gmr_ids = VMWGFX_NUM_MOB;
+		gman->max_gmr_pages = dev_priv->max_mob_pages;
+		break;
+	default:
+		BUG();
+	}
 	man->priv = (void *) gman;
 	return 0;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
index 45d5b5ab6ca..37881ecf5d7 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
@@ -29,12 +29,18 @@
 #include <drm/vmwgfx_drm.h>
 #include "vmwgfx_kms.h"
 
+struct svga_3d_compat_cap {
+	SVGA3dCapsRecordHeader header;
+	SVGA3dCapPair pairs[SVGA3D_DEVCAP_MAX];
+};
+
 int vmw_getparam_ioctl(struct drm_device *dev, void *data,
 		       struct drm_file *file_priv)
 {
 	struct vmw_private *dev_priv = vmw_priv(dev);
 	struct drm_vmw_getparam_arg *param =
 	    (struct drm_vmw_getparam_arg *)data;
+	struct vmw_fpriv *vmw_fp = vmw_fpriv(file_priv);
 
 	switch (param->param) {
 	case DRM_VMW_PARAM_NUM_STREAMS:
@@ -53,13 +59,18 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data,
 		param->value = dev_priv->fifo.capabilities;
 		break;
 	case DRM_VMW_PARAM_MAX_FB_SIZE:
-		param->value = dev_priv->vram_size;
+		param->value = dev_priv->prim_bb_mem;
 		break;
 	case DRM_VMW_PARAM_FIFO_HW_VERSION:
 	{
 		__le32 __iomem *fifo_mem = dev_priv->mmio_virt;
 		const struct vmw_fifo_state *fifo = &dev_priv->fifo;
 
+		if ((dev_priv->capabilities & SVGA_CAP_GBOBJECTS)) {
+			param->value = SVGA3D_HWVERSION_WS8_B1;
+			break;
+		}
+
 		param->value =
 			ioread32(fifo_mem +
 				 ((fifo->capabilities &
@@ -69,7 +80,30 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data,
 		break;
 	}
 	case DRM_VMW_PARAM_MAX_SURF_MEMORY:
-		param->value = dev_priv->memory_size;
+		if ((dev_priv->capabilities & SVGA_CAP_GBOBJECTS) &&
+		    !vmw_fp->gb_aware)
+			param->value = dev_priv->max_mob_pages * PAGE_SIZE / 2;
+		else
+			param->value = dev_priv->memory_size;
+		break;
+	case DRM_VMW_PARAM_3D_CAPS_SIZE:
+		if ((dev_priv->capabilities & SVGA_CAP_GBOBJECTS) &&
+		    vmw_fp->gb_aware)
+			param->value = SVGA3D_DEVCAP_MAX * sizeof(uint32_t);
+		else if (dev_priv->capabilities & SVGA_CAP_GBOBJECTS)
+			param->value = sizeof(struct svga_3d_compat_cap) +
+				sizeof(uint32_t);
+		else
+			param->value = (SVGA_FIFO_3D_CAPS_LAST -
+					SVGA_FIFO_3D_CAPS + 1) *
+				sizeof(uint32_t);
+		break;
+	case DRM_VMW_PARAM_MAX_MOB_MEMORY:
+		vmw_fp->gb_aware = true;
+		param->value = dev_priv->max_mob_pages * PAGE_SIZE;
+		break;
+	case DRM_VMW_PARAM_MAX_MOB_SIZE:
+		param->value = dev_priv->max_mob_size;
 		break;
 	default:
 		DRM_ERROR("Illegal vmwgfx get param request: %d\n",
@@ -80,6 +114,38 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data,
 	return 0;
 }
 
+static int vmw_fill_compat_cap(struct vmw_private *dev_priv, void *bounce,
+			       size_t size)
+{
+	struct svga_3d_compat_cap *compat_cap =
+		(struct svga_3d_compat_cap *) bounce;
+	unsigned int i;
+	size_t pair_offset = offsetof(struct svga_3d_compat_cap, pairs);
+	unsigned int max_size;
+
+	if (size < pair_offset)
+		return -EINVAL;
+
+	max_size = (size - pair_offset) / sizeof(SVGA3dCapPair);
+
+	if (max_size > SVGA3D_DEVCAP_MAX)
+		max_size = SVGA3D_DEVCAP_MAX;
+
+	compat_cap->header.length =
+		(pair_offset + max_size * sizeof(SVGA3dCapPair)) / sizeof(u32);
+	compat_cap->header.type = SVGA3DCAPS_RECORD_DEVCAPS;
+
+	mutex_lock(&dev_priv->hw_mutex);
+	for (i = 0; i < max_size; ++i) {
+		vmw_write(dev_priv, SVGA_REG_DEV_CAP, i);
+		compat_cap->pairs[i][0] = i;
+		compat_cap->pairs[i][1] = vmw_read(dev_priv, SVGA_REG_DEV_CAP);
+	}
+	mutex_unlock(&dev_priv->hw_mutex);
+
+	return 0;
+}
+
 
 int vmw_get_cap_3d_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file_priv)
@@ -92,29 +158,58 @@ int vmw_get_cap_3d_ioctl(struct drm_device *dev, void *data,
 	void __user *buffer = (void __user *)((unsigned long)(arg->buffer));
 	void *bounce;
 	int ret;
+	bool gb_objects = !!(dev_priv->capabilities & SVGA_CAP_GBOBJECTS);
+	struct vmw_fpriv *vmw_fp = vmw_fpriv(file_priv);
 
 	if (unlikely(arg->pad64 != 0)) {
 		DRM_ERROR("Illegal GET_3D_CAP argument.\n");
 		return -EINVAL;
 	}
 
-	size = (SVGA_FIFO_3D_CAPS_LAST - SVGA_FIFO_3D_CAPS + 1) << 2;
+	if (gb_objects && vmw_fp->gb_aware)
+		size = SVGA3D_DEVCAP_MAX * sizeof(uint32_t);
+	else if (gb_objects)
+		size = sizeof(struct svga_3d_compat_cap) + sizeof(uint32_t);
+	else
+		size = (SVGA_FIFO_3D_CAPS_LAST - SVGA_FIFO_3D_CAPS + 1) *
+			sizeof(uint32_t);
 
 	if (arg->max_size < size)
 		size = arg->max_size;
 
-	bounce = vmalloc(size);
+	bounce = vzalloc(size);
 	if (unlikely(bounce == NULL)) {
 		DRM_ERROR("Failed to allocate bounce buffer for 3D caps.\n");
 		return -ENOMEM;
 	}
 
-	fifo_mem = dev_priv->mmio_virt;
-	memcpy_fromio(bounce, &fifo_mem[SVGA_FIFO_3D_CAPS], size);
+	if (gb_objects && vmw_fp->gb_aware) {
+		int i, num;
+		uint32_t *bounce32 = (uint32_t *) bounce;
+
+		num = size / sizeof(uint32_t);
+		if (num > SVGA3D_DEVCAP_MAX)
+			num = SVGA3D_DEVCAP_MAX;
+
+		mutex_lock(&dev_priv->hw_mutex);
+		for (i = 0; i < num; ++i) {
+			vmw_write(dev_priv, SVGA_REG_DEV_CAP, i);
+			*bounce32++ = vmw_read(dev_priv, SVGA_REG_DEV_CAP);
+		}
+		mutex_unlock(&dev_priv->hw_mutex);
+	} else if (gb_objects) {
+		ret = vmw_fill_compat_cap(dev_priv, bounce, size);
+		if (unlikely(ret != 0))
+			goto out_err;
+	} else {
+		fifo_mem = dev_priv->mmio_virt;
+		memcpy_fromio(bounce, &fifo_mem[SVGA_FIFO_3D_CAPS], size);
+	}
 
 	ret = copy_to_user(buffer, bounce, size);
 	if (ret)
 		ret = -EFAULT;
+out_err:
 	vfree(bounce);
 
 	if (unlikely(ret != 0))
@@ -131,7 +226,6 @@ int vmw_present_ioctl(struct drm_device *dev, void *data,
 	struct drm_vmw_present_arg *arg =
 		(struct drm_vmw_present_arg *)data;
 	struct vmw_surface *surface;
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	struct drm_vmw_rect __user *clips_ptr;
 	struct drm_vmw_rect *clips = NULL;
 	struct drm_framebuffer *fb;
@@ -176,7 +270,7 @@ int vmw_present_ioctl(struct drm_device *dev, void *data,
 	}
 	vfb = vmw_framebuffer_to_vfb(fb);
 
-	ret = ttm_read_lock(&vmaster->lock, true);
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
 	if (unlikely(ret != 0))
 		goto out_no_ttm_lock;
 
@@ -196,7 +290,7 @@ int vmw_present_ioctl(struct drm_device *dev, void *data,
 	vmw_surface_unreference(&surface);
 
 out_no_surface:
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
 out_no_ttm_lock:
 	drm_framebuffer_unreference(fb);
 out_no_fb:
@@ -216,7 +310,6 @@ int vmw_present_readback_ioctl(struct drm_device *dev, void *data,
 	struct drm_vmw_fence_rep __user *user_fence_rep =
 		(struct drm_vmw_fence_rep __user *)
 		(unsigned long)arg->fence_rep;
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	struct drm_vmw_rect __user *clips_ptr;
 	struct drm_vmw_rect *clips = NULL;
 	struct drm_framebuffer *fb;
@@ -266,7 +359,7 @@ int vmw_present_readback_ioctl(struct drm_device *dev, void *data,
 		goto out_no_ttm_lock;
 	}
 
-	ret = ttm_read_lock(&vmaster->lock, true);
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
 	if (unlikely(ret != 0))
 		goto out_no_ttm_lock;
 
@@ -274,7 +367,7 @@ int vmw_present_readback_ioctl(struct drm_device *dev, void *data,
 			       vfb, user_fence_rep,
 			       clips, num_clips);
 
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
 out_no_ttm_lock:
 	drm_framebuffer_unreference(fb);
 out_no_fb:
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
index 4640adbcaf9..0c423766c44 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
@@ -30,7 +30,7 @@
 
 #define VMW_FENCE_WRAP (1 << 24)
 
-irqreturn_t vmw_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t vmw_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *)arg;
 	struct vmw_private *dev_priv = vmw_priv(dev);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
index 03f1c203863..8f3edc4710f 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
@@ -40,7 +40,7 @@ struct vmw_clip_rect {
  * Clip @num_rects number of @rects against @clip storing the
  * results in @out_rects and the number of passed rects in @out_num.
  */
-void vmw_clip_cliprects(struct drm_clip_rect *rects,
+static void vmw_clip_cliprects(struct drm_clip_rect *rects,
 			int num_rects,
 			struct vmw_clip_rect clip,
 			SVGASignedRect *out_rects,
@@ -187,7 +187,7 @@ int vmw_du_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
 	 * can do this since the caller in the drm core doesn't check anything
 	 * which is protected by any looks.
 	 */
-	mutex_unlock(&crtc->mutex);
+	drm_modeset_unlock(&crtc->mutex);
 	drm_modeset_lock_all(dev_priv->dev);
 
 	/* A lot of the code assumes this */
@@ -252,7 +252,7 @@ int vmw_du_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
 	ret = 0;
 out:
 	drm_modeset_unlock_all(dev_priv->dev);
-	mutex_lock(&crtc->mutex);
+	drm_modeset_lock(&crtc->mutex, NULL);
 
 	return ret;
 }
@@ -273,7 +273,7 @@ int vmw_du_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
 	 * can do this since the caller in the drm core doesn't check anything
 	 * which is protected by any looks.
 	 */
-	mutex_unlock(&crtc->mutex);
+	drm_modeset_unlock(&crtc->mutex);
 	drm_modeset_lock_all(dev_priv->dev);
 
 	vmw_cursor_update_position(dev_priv, shown,
@@ -281,7 +281,7 @@ int vmw_du_crtc_cursor_move(struct drm_crtc *crtc, int x, int y)
 				   du->cursor_y + du->hotspot_y);
 
 	drm_modeset_unlock_all(dev_priv->dev);
-	mutex_lock(&crtc->mutex);
+	drm_modeset_lock(&crtc->mutex, NULL);
 
 	return 0;
 }
@@ -423,7 +423,7 @@ struct vmw_framebuffer_surface {
 	struct drm_master *master;
 };
 
-void vmw_framebuffer_surface_destroy(struct drm_framebuffer *framebuffer)
+static void vmw_framebuffer_surface_destroy(struct drm_framebuffer *framebuffer)
 {
 	struct vmw_framebuffer_surface *vfbs =
 		vmw_framebuffer_to_vfbs(framebuffer);
@@ -468,7 +468,7 @@ static int do_surface_dirty_sou(struct vmw_private *dev_priv,
 	num_units = 0;
 	list_for_each_entry(crtc, &dev_priv->dev->mode_config.crtc_list,
 			    head) {
-		if (crtc->fb != &framebuffer->base)
+		if (crtc->primary->fb != &framebuffer->base)
 			continue;
 		units[num_units++] = vmw_crtc_to_du(crtc);
 	}
@@ -589,14 +589,13 @@ out_free_tmp:
 	return ret;
 }
 
-int vmw_framebuffer_surface_dirty(struct drm_framebuffer *framebuffer,
+static int vmw_framebuffer_surface_dirty(struct drm_framebuffer *framebuffer,
 				  struct drm_file *file_priv,
 				  unsigned flags, unsigned color,
 				  struct drm_clip_rect *clips,
 				  unsigned num_clips)
 {
 	struct vmw_private *dev_priv = vmw_priv(framebuffer->dev);
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	struct vmw_framebuffer_surface *vfbs =
 		vmw_framebuffer_to_vfbs(framebuffer);
 	struct drm_clip_rect norect;
@@ -609,9 +608,13 @@ int vmw_framebuffer_surface_dirty(struct drm_framebuffer *framebuffer,
 	if (!dev_priv->sou_priv)
 		return -EINVAL;
 
-	ret = ttm_read_lock(&vmaster->lock, true);
-	if (unlikely(ret != 0))
+	drm_modeset_lock_all(dev_priv->dev);
+
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
+	if (unlikely(ret != 0)) {
+		drm_modeset_unlock_all(dev_priv->dev);
 		return ret;
+	}
 
 	if (!num_clips) {
 		num_clips = 1;
@@ -628,7 +631,10 @@ int vmw_framebuffer_surface_dirty(struct drm_framebuffer *framebuffer,
 				   flags, color,
 				   clips, num_clips, inc, NULL);
 
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
+
+	drm_modeset_unlock_all(dev_priv->dev);
+
 	return 0;
 }
 
@@ -665,9 +671,9 @@ static int vmw_kms_new_framebuffer_surface(struct vmw_private *dev_priv,
 
 	if (unlikely(surface->mip_levels[0] != 1 ||
 		     surface->num_sizes != 1 ||
-		     surface->sizes[0].width < mode_cmd->width ||
-		     surface->sizes[0].height < mode_cmd->height ||
-		     surface->sizes[0].depth != 1)) {
+		     surface->base_size.width < mode_cmd->width ||
+		     surface->base_size.height < mode_cmd->height ||
+		     surface->base_size.depth != 1)) {
 		DRM_ERROR("Incompatible surface dimensions "
 			  "for requested mode.\n");
 		return -EINVAL;
@@ -754,7 +760,7 @@ struct vmw_framebuffer_dmabuf {
 	struct vmw_dma_buffer *buffer;
 };
 
-void vmw_framebuffer_dmabuf_destroy(struct drm_framebuffer *framebuffer)
+static void vmw_framebuffer_dmabuf_destroy(struct drm_framebuffer *framebuffer)
 {
 	struct vmw_framebuffer_dmabuf *vfbd =
 		vmw_framebuffer_to_vfbd(framebuffer);
@@ -876,7 +882,7 @@ static int do_dmabuf_dirty_sou(struct drm_file *file_priv,
 
 	num_units = 0;
 	list_for_each_entry(crtc, &dev_priv->dev->mode_config.crtc_list, head) {
-		if (crtc->fb != &framebuffer->base)
+		if (crtc->primary->fb != &framebuffer->base)
 			continue;
 		units[num_units++] = vmw_crtc_to_du(crtc);
 	}
@@ -940,22 +946,25 @@ static int do_dmabuf_dirty_sou(struct drm_file *file_priv,
 	return ret;
 }
 
-int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
+static int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
 				 struct drm_file *file_priv,
 				 unsigned flags, unsigned color,
 				 struct drm_clip_rect *clips,
 				 unsigned num_clips)
 {
 	struct vmw_private *dev_priv = vmw_priv(framebuffer->dev);
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	struct vmw_framebuffer_dmabuf *vfbd =
 		vmw_framebuffer_to_vfbd(framebuffer);
 	struct drm_clip_rect norect;
 	int ret, increment = 1;
 
-	ret = ttm_read_lock(&vmaster->lock, true);
-	if (unlikely(ret != 0))
+	drm_modeset_lock_all(dev_priv->dev);
+
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
+	if (unlikely(ret != 0)) {
+		drm_modeset_unlock_all(dev_priv->dev);
 		return ret;
+	}
 
 	if (!num_clips) {
 		num_clips = 1;
@@ -978,7 +987,10 @@ int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
 					  clips, num_clips, increment, NULL);
 	}
 
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
+
+	drm_modeset_unlock_all(dev_priv->dev);
+
 	return ret;
 }
 
@@ -1231,7 +1243,7 @@ int vmw_kms_present(struct vmw_private *dev_priv,
 
 	num_units = 0;
 	list_for_each_entry(crtc, &dev_priv->dev->mode_config.crtc_list, head) {
-		if (crtc->fb != &vfb->base)
+		if (crtc->primary->fb != &vfb->base)
 			continue;
 		units[num_units++] = vmw_crtc_to_du(crtc);
 	}
@@ -1368,7 +1380,7 @@ int vmw_kms_readback(struct vmw_private *dev_priv,
 
 	num_units = 0;
 	list_for_each_entry(crtc, &dev_priv->dev->mode_config.crtc_list, head) {
-		if (crtc->fb != &vfb->base)
+		if (crtc->primary->fb != &vfb->base)
 			continue;
 		units[num_units++] = vmw_crtc_to_du(crtc);
 	}
@@ -1631,7 +1643,7 @@ bool vmw_kms_validate_mode_vram(struct vmw_private *dev_priv,
 				uint32_t pitch,
 				uint32_t height)
 {
-	return ((u64) pitch * (u64) height) < (u64) dev_priv->vram_size;
+	return ((u64) pitch * (u64) height) < (u64) dev_priv->prim_bb_mem;
 }
 
 
@@ -1663,7 +1675,7 @@ void vmw_disable_vblank(struct drm_device *dev, int crtc)
  * Small shared kms functions.
  */
 
-int vmw_du_update_layout(struct vmw_private *dev_priv, unsigned num,
+static int vmw_du_update_layout(struct vmw_private *dev_priv, unsigned num,
 			 struct drm_vmw_rect *rects)
 {
 	struct drm_device *dev = dev_priv->dev;
@@ -1711,7 +1723,7 @@ int vmw_du_page_flip(struct drm_crtc *crtc,
 		     uint32_t page_flip_flags)
 {
 	struct vmw_private *dev_priv = vmw_priv(crtc->dev);
-	struct drm_framebuffer *old_fb = crtc->fb;
+	struct drm_framebuffer *old_fb = crtc->primary->fb;
 	struct vmw_framebuffer *vfb = vmw_framebuffer_to_vfb(fb);
 	struct drm_file *file_priv ;
 	struct vmw_fence_obj *fence = NULL;
@@ -1729,7 +1741,7 @@ int vmw_du_page_flip(struct drm_crtc *crtc,
 	if (!vmw_kms_screen_object_flippable(dev_priv, crtc))
 		return -EINVAL;
 
-	crtc->fb = fb;
+	crtc->primary->fb = fb;
 
 	/* do a full screen dirty update */
 	clips.x1 = clips.y1 = 0;
@@ -1769,7 +1781,7 @@ int vmw_du_page_flip(struct drm_crtc *crtc,
 	return ret;
 
 out_no_fence:
-	crtc->fb = old_fb;
+	crtc->primary->fb = old_fb;
 	return ret;
 }
 
@@ -1989,7 +2001,7 @@ int vmw_du_connector_fill_modes(struct drm_connector *connector,
 	if (du->pref_mode)
 		list_move(&du->pref_mode->head, &connector->probed_modes);
 
-	drm_mode_connector_list_update(connector);
+	drm_mode_connector_list_update(connector, true);
 
 	return 1;
 }
@@ -2008,7 +2020,6 @@ int vmw_kms_update_layout_ioctl(struct drm_device *dev, void *data,
 	struct vmw_private *dev_priv = vmw_priv(dev);
 	struct drm_vmw_update_layout_arg *arg =
 		(struct drm_vmw_update_layout_arg *)data;
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	void __user *user_rects;
 	struct drm_vmw_rect *rects;
 	unsigned rects_size;
@@ -2016,7 +2027,7 @@ int vmw_kms_update_layout_ioctl(struct drm_device *dev, void *data,
 	int i;
 	struct drm_mode_config *mode_config = &dev->mode_config;
 
-	ret = ttm_read_lock(&vmaster->lock, true);
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -2058,6 +2069,6 @@ int vmw_kms_update_layout_ioctl(struct drm_device *dev, void *data,
 out_free:
 	kfree(rects);
 out_unlock:
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
 	return ret;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
index a055a26819c..b2b9bd23aee 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
@@ -93,7 +93,7 @@ static int vmw_ldu_commit_list(struct vmw_private *dev_priv)
 
 		if (crtc == NULL)
 			return 0;
-		fb = entry->base.crtc.fb;
+		fb = entry->base.crtc.primary->fb;
 
 		return vmw_kms_write_svga(dev_priv, w, h, fb->pitches[0],
 					  fb->bits_per_pixel, fb->depth);
@@ -101,7 +101,7 @@ static int vmw_ldu_commit_list(struct vmw_private *dev_priv)
 
 	if (!list_empty(&lds->active)) {
 		entry = list_entry(lds->active.next, typeof(*entry), active);
-		fb = entry->base.crtc.fb;
+		fb = entry->base.crtc.primary->fb;
 
 		vmw_kms_write_svga(dev_priv, fb->width, fb->height, fb->pitches[0],
 				   fb->bits_per_pixel, fb->depth);
@@ -259,7 +259,7 @@ static int vmw_ldu_crtc_set_config(struct drm_mode_set *set)
 
 		connector->encoder = NULL;
 		encoder->crtc = NULL;
-		crtc->fb = NULL;
+		crtc->primary->fb = NULL;
 		crtc->enabled = false;
 
 		vmw_ldu_del_active(dev_priv, ldu);
@@ -280,7 +280,7 @@ static int vmw_ldu_crtc_set_config(struct drm_mode_set *set)
 
 	vmw_fb_off(dev_priv);
 
-	crtc->fb = fb;
+	crtc->primary->fb = fb;
 	encoder->crtc = crtc;
 	connector->encoder = encoder;
 	crtc->x = set->x;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
new file mode 100644
index 00000000000..04a64b8cd3c
--- /dev/null
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
@@ -0,0 +1,656 @@
+/**************************************************************************
+ *
+ * Copyright © 2012 VMware, Inc., Palo Alto, CA., USA
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ **************************************************************************/
+
+#include "vmwgfx_drv.h"
+
+/*
+ * If we set up the screen target otable, screen objects stop working.
+ */
+
+#define VMW_OTABLE_SETUP_SUB ((VMWGFX_ENABLE_SCREEN_TARGET_OTABLE) ? 0 : 1)
+
+#ifdef CONFIG_64BIT
+#define VMW_PPN_SIZE 8
+#define VMW_MOBFMT_PTDEPTH_0 SVGA3D_MOBFMT_PTDEPTH64_0
+#define VMW_MOBFMT_PTDEPTH_1 SVGA3D_MOBFMT_PTDEPTH64_1
+#define VMW_MOBFMT_PTDEPTH_2 SVGA3D_MOBFMT_PTDEPTH64_2
+#else
+#define VMW_PPN_SIZE 4
+#define VMW_MOBFMT_PTDEPTH_0 SVGA3D_MOBFMT_PTDEPTH_0
+#define VMW_MOBFMT_PTDEPTH_1 SVGA3D_MOBFMT_PTDEPTH_1
+#define VMW_MOBFMT_PTDEPTH_2 SVGA3D_MOBFMT_PTDEPTH_2
+#endif
+
+/*
+ * struct vmw_mob - Structure containing page table and metadata for a
+ * Guest Memory OBject.
+ *
+ * @num_pages       Number of pages that make up the page table.
+ * @pt_level        The indirection level of the page table. 0-2.
+ * @pt_root_page    DMA address of the level 0 page of the page table.
+ */
+struct vmw_mob {
+	struct ttm_buffer_object *pt_bo;
+	unsigned long num_pages;
+	unsigned pt_level;
+	dma_addr_t pt_root_page;
+	uint32_t id;
+};
+
+/*
+ * struct vmw_otable - Guest Memory OBject table metadata
+ *
+ * @size:           Size of the table (page-aligned).
+ * @page_table:     Pointer to a struct vmw_mob holding the page table.
+ */
+struct vmw_otable {
+	unsigned long size;
+	struct vmw_mob *page_table;
+};
+
+static int vmw_mob_pt_populate(struct vmw_private *dev_priv,
+			       struct vmw_mob *mob);
+static void vmw_mob_pt_setup(struct vmw_mob *mob,
+			     struct vmw_piter data_iter,
+			     unsigned long num_data_pages);
+
+/*
+ * vmw_setup_otable_base - Issue an object table base setup command to
+ * the device
+ *
+ * @dev_priv:       Pointer to a device private structure
+ * @type:           Type of object table base
+ * @offset          Start of table offset into dev_priv::otable_bo
+ * @otable          Pointer to otable metadata;
+ *
+ * This function returns -ENOMEM if it fails to reserve fifo space,
+ * and may block waiting for fifo space.
+ */
+static int vmw_setup_otable_base(struct vmw_private *dev_priv,
+				 SVGAOTableType type,
+				 unsigned long offset,
+				 struct vmw_otable *otable)
+{
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdSetOTableBase64 body;
+	} *cmd;
+	struct vmw_mob *mob;
+	const struct vmw_sg_table *vsgt;
+	struct vmw_piter iter;
+	int ret;
+
+	BUG_ON(otable->page_table != NULL);
+
+	vsgt = vmw_bo_sg_table(dev_priv->otable_bo);
+	vmw_piter_start(&iter, vsgt, offset >> PAGE_SHIFT);
+	WARN_ON(!vmw_piter_next(&iter));
+
+	mob = vmw_mob_create(otable->size >> PAGE_SHIFT);
+	if (unlikely(mob == NULL)) {
+		DRM_ERROR("Failed creating OTable page table.\n");
+		return -ENOMEM;
+	}
+
+	if (otable->size <= PAGE_SIZE) {
+		mob->pt_level = VMW_MOBFMT_PTDEPTH_0;
+		mob->pt_root_page = vmw_piter_dma_addr(&iter);
+	} else if (vsgt->num_regions == 1) {
+		mob->pt_level = SVGA3D_MOBFMT_RANGE;
+		mob->pt_root_page = vmw_piter_dma_addr(&iter);
+	} else {
+		ret = vmw_mob_pt_populate(dev_priv, mob);
+		if (unlikely(ret != 0))
+			goto out_no_populate;
+
+		vmw_mob_pt_setup(mob, iter, otable->size >> PAGE_SHIFT);
+		mob->pt_level += VMW_MOBFMT_PTDEPTH_1 - SVGA3D_MOBFMT_PTDEPTH_1;
+	}
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for OTable setup.\n");
+		ret = -ENOMEM;
+		goto out_no_fifo;
+	}
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->header.id = SVGA_3D_CMD_SET_OTABLE_BASE64;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.type = type;
+	cmd->body.baseAddress = cpu_to_le64(mob->pt_root_page >> PAGE_SHIFT);
+	cmd->body.sizeInBytes = otable->size;
+	cmd->body.validSizeInBytes = 0;
+	cmd->body.ptDepth = mob->pt_level;
+
+	/*
+	 * The device doesn't support this, But the otable size is
+	 * determined at compile-time, so this BUG shouldn't trigger
+	 * randomly.
+	 */
+	BUG_ON(mob->pt_level == VMW_MOBFMT_PTDEPTH_2);
+
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	otable->page_table = mob;
+
+	return 0;
+
+out_no_fifo:
+out_no_populate:
+	vmw_mob_destroy(mob);
+	return ret;
+}
+
+/*
+ * vmw_takedown_otable_base - Issue an object table base takedown command
+ * to the device
+ *
+ * @dev_priv:       Pointer to a device private structure
+ * @type:           Type of object table base
+ *
+ */
+static void vmw_takedown_otable_base(struct vmw_private *dev_priv,
+				     SVGAOTableType type,
+				     struct vmw_otable *otable)
+{
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdSetOTableBase body;
+	} *cmd;
+	struct ttm_buffer_object *bo;
+
+	if (otable->page_table == NULL)
+		return;
+
+	bo = otable->page_table->pt_bo;
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for OTable "
+			  "takedown.\n");
+	} else {
+		memset(cmd, 0, sizeof(*cmd));
+		cmd->header.id = SVGA_3D_CMD_SET_OTABLE_BASE;
+		cmd->header.size = sizeof(cmd->body);
+		cmd->body.type = type;
+		cmd->body.baseAddress = 0;
+		cmd->body.sizeInBytes = 0;
+		cmd->body.validSizeInBytes = 0;
+		cmd->body.ptDepth = SVGA3D_MOBFMT_INVALID;
+		vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	}
+
+	if (bo) {
+		int ret;
+
+		ret = ttm_bo_reserve(bo, false, true, false, NULL);
+		BUG_ON(ret != 0);
+
+		vmw_fence_single_bo(bo, NULL);
+		ttm_bo_unreserve(bo);
+	}
+
+	vmw_mob_destroy(otable->page_table);
+	otable->page_table = NULL;
+}
+
+/*
+ * vmw_otables_setup - Set up guest backed memory object tables
+ *
+ * @dev_priv:       Pointer to a device private structure
+ *
+ * Takes care of the device guest backed surface
+ * initialization, by setting up the guest backed memory object tables.
+ * Returns 0 on success and various error codes on failure. A succesful return
+ * means the object tables can be taken down using the vmw_otables_takedown
+ * function.
+ */
+int vmw_otables_setup(struct vmw_private *dev_priv)
+{
+	unsigned long offset;
+	unsigned long bo_size;
+	struct vmw_otable *otables;
+	SVGAOTableType i;
+	int ret;
+
+	otables = kzalloc(SVGA_OTABLE_DX9_MAX * sizeof(*otables),
+			  GFP_KERNEL);
+	if (unlikely(otables == NULL)) {
+		DRM_ERROR("Failed to allocate space for otable "
+			  "metadata.\n");
+		return -ENOMEM;
+	}
+
+	otables[SVGA_OTABLE_MOB].size =
+		VMWGFX_NUM_MOB * SVGA3D_OTABLE_MOB_ENTRY_SIZE;
+	otables[SVGA_OTABLE_SURFACE].size =
+		VMWGFX_NUM_GB_SURFACE * SVGA3D_OTABLE_SURFACE_ENTRY_SIZE;
+	otables[SVGA_OTABLE_CONTEXT].size =
+		VMWGFX_NUM_GB_CONTEXT * SVGA3D_OTABLE_CONTEXT_ENTRY_SIZE;
+	otables[SVGA_OTABLE_SHADER].size =
+		VMWGFX_NUM_GB_SHADER * SVGA3D_OTABLE_SHADER_ENTRY_SIZE;
+	otables[SVGA_OTABLE_SCREEN_TARGET].size =
+		VMWGFX_NUM_GB_SCREEN_TARGET *
+		SVGA3D_OTABLE_SCREEN_TARGET_ENTRY_SIZE;
+
+	bo_size = 0;
+	for (i = 0; i < SVGA_OTABLE_DX9_MAX; ++i) {
+		otables[i].size =
+			(otables[i].size + PAGE_SIZE - 1) & PAGE_MASK;
+		bo_size += otables[i].size;
+	}
+
+	ret = ttm_bo_create(&dev_priv->bdev, bo_size,
+			    ttm_bo_type_device,
+			    &vmw_sys_ne_placement,
+			    0, false, NULL,
+			    &dev_priv->otable_bo);
+
+	if (unlikely(ret != 0))
+		goto out_no_bo;
+
+	ret = ttm_bo_reserve(dev_priv->otable_bo, false, true, false, NULL);
+	BUG_ON(ret != 0);
+	ret = vmw_bo_driver.ttm_tt_populate(dev_priv->otable_bo->ttm);
+	if (unlikely(ret != 0))
+		goto out_unreserve;
+	ret = vmw_bo_map_dma(dev_priv->otable_bo);
+	if (unlikely(ret != 0))
+		goto out_unreserve;
+
+	ttm_bo_unreserve(dev_priv->otable_bo);
+
+	offset = 0;
+	for (i = 0; i < SVGA_OTABLE_DX9_MAX - VMW_OTABLE_SETUP_SUB; ++i) {
+		ret = vmw_setup_otable_base(dev_priv, i, offset,
+					    &otables[i]);
+		if (unlikely(ret != 0))
+			goto out_no_setup;
+		offset += otables[i].size;
+	}
+
+	dev_priv->otables = otables;
+	return 0;
+
+out_unreserve:
+	ttm_bo_unreserve(dev_priv->otable_bo);
+out_no_setup:
+	for (i = 0; i < SVGA_OTABLE_DX9_MAX - VMW_OTABLE_SETUP_SUB; ++i)
+		vmw_takedown_otable_base(dev_priv, i, &otables[i]);
+
+	ttm_bo_unref(&dev_priv->otable_bo);
+out_no_bo:
+	kfree(otables);
+	return ret;
+}
+
+
+/*
+ * vmw_otables_takedown - Take down guest backed memory object tables
+ *
+ * @dev_priv:       Pointer to a device private structure
+ *
+ * Take down the Guest Memory Object tables.
+ */
+void vmw_otables_takedown(struct vmw_private *dev_priv)
+{
+	SVGAOTableType i;
+	struct ttm_buffer_object *bo = dev_priv->otable_bo;
+	int ret;
+
+	for (i = 0; i < SVGA_OTABLE_DX9_MAX - VMW_OTABLE_SETUP_SUB; ++i)
+		vmw_takedown_otable_base(dev_priv, i,
+					 &dev_priv->otables[i]);
+
+	ret = ttm_bo_reserve(bo, false, true, false, NULL);
+	BUG_ON(ret != 0);
+
+	vmw_fence_single_bo(bo, NULL);
+	ttm_bo_unreserve(bo);
+
+	ttm_bo_unref(&dev_priv->otable_bo);
+	kfree(dev_priv->otables);
+	dev_priv->otables = NULL;
+}
+
+
+/*
+ * vmw_mob_calculate_pt_pages - Calculate the number of page table pages
+ * needed for a guest backed memory object.
+ *
+ * @data_pages:  Number of data pages in the memory object buffer.
+ */
+static unsigned long vmw_mob_calculate_pt_pages(unsigned long data_pages)
+{
+	unsigned long data_size = data_pages * PAGE_SIZE;
+	unsigned long tot_size = 0;
+
+	while (likely(data_size > PAGE_SIZE)) {
+		data_size = DIV_ROUND_UP(data_size, PAGE_SIZE);
+		data_size *= VMW_PPN_SIZE;
+		tot_size += (data_size + PAGE_SIZE - 1) & PAGE_MASK;
+	}
+
+	return tot_size >> PAGE_SHIFT;
+}
+
+/*
+ * vmw_mob_create - Create a mob, but don't populate it.
+ *
+ * @data_pages:  Number of data pages of the underlying buffer object.
+ */
+struct vmw_mob *vmw_mob_create(unsigned long data_pages)
+{
+	struct vmw_mob *mob = kzalloc(sizeof(*mob), GFP_KERNEL);
+
+	if (unlikely(mob == NULL))
+		return NULL;
+
+	mob->num_pages = vmw_mob_calculate_pt_pages(data_pages);
+
+	return mob;
+}
+
+/*
+ * vmw_mob_pt_populate - Populate the mob pagetable
+ *
+ * @mob:         Pointer to the mob the pagetable of which we want to
+ *               populate.
+ *
+ * This function allocates memory to be used for the pagetable, and
+ * adjusts TTM memory accounting accordingly. Returns ENOMEM if
+ * memory resources aren't sufficient and may cause TTM buffer objects
+ * to be swapped out by using the TTM memory accounting function.
+ */
+static int vmw_mob_pt_populate(struct vmw_private *dev_priv,
+			       struct vmw_mob *mob)
+{
+	int ret;
+	BUG_ON(mob->pt_bo != NULL);
+
+	ret = ttm_bo_create(&dev_priv->bdev, mob->num_pages * PAGE_SIZE,
+			    ttm_bo_type_device,
+			    &vmw_sys_ne_placement,
+			    0, false, NULL, &mob->pt_bo);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_bo_reserve(mob->pt_bo, false, true, false, NULL);
+
+	BUG_ON(ret != 0);
+	ret = vmw_bo_driver.ttm_tt_populate(mob->pt_bo->ttm);
+	if (unlikely(ret != 0))
+		goto out_unreserve;
+	ret = vmw_bo_map_dma(mob->pt_bo);
+	if (unlikely(ret != 0))
+		goto out_unreserve;
+
+	ttm_bo_unreserve(mob->pt_bo);
+	
+	return 0;
+
+out_unreserve:
+	ttm_bo_unreserve(mob->pt_bo);
+	ttm_bo_unref(&mob->pt_bo);
+
+	return ret;
+}
+
+/**
+ * vmw_mob_assign_ppn - Assign a value to a page table entry
+ *
+ * @addr: Pointer to pointer to page table entry.
+ * @val: The page table entry
+ *
+ * Assigns a value to a page table entry pointed to by *@addr and increments
+ * *@addr according to the page table entry size.
+ */
+#if (VMW_PPN_SIZE == 8)
+static void vmw_mob_assign_ppn(__le32 **addr, dma_addr_t val)
+{
+	*((__le64 *) *addr) = cpu_to_le64(val >> PAGE_SHIFT);
+	*addr += 2;
+}
+#else
+static void vmw_mob_assign_ppn(__le32 **addr, dma_addr_t val)
+{
+	*(*addr)++ = cpu_to_le32(val >> PAGE_SHIFT);
+}
+#endif
+
+/*
+ * vmw_mob_build_pt - Build a pagetable
+ *
+ * @data_addr:      Array of DMA addresses to the underlying buffer
+ *                  object's data pages.
+ * @num_data_pages: Number of buffer object data pages.
+ * @pt_pages:       Array of page pointers to the page table pages.
+ *
+ * Returns the number of page table pages actually used.
+ * Uses atomic kmaps of highmem pages to avoid TLB thrashing.
+ */
+static unsigned long vmw_mob_build_pt(struct vmw_piter *data_iter,
+				      unsigned long num_data_pages,
+				      struct vmw_piter *pt_iter)
+{
+	unsigned long pt_size = num_data_pages * VMW_PPN_SIZE;
+	unsigned long num_pt_pages = DIV_ROUND_UP(pt_size, PAGE_SIZE);
+	unsigned long pt_page;
+	__le32 *addr, *save_addr;
+	unsigned long i;
+	struct page *page;
+
+	for (pt_page = 0; pt_page < num_pt_pages; ++pt_page) {
+		page = vmw_piter_page(pt_iter);
+
+		save_addr = addr = kmap_atomic(page);
+
+		for (i = 0; i < PAGE_SIZE / VMW_PPN_SIZE; ++i) {
+			vmw_mob_assign_ppn(&addr,
+					   vmw_piter_dma_addr(data_iter));
+			if (unlikely(--num_data_pages == 0))
+				break;
+			WARN_ON(!vmw_piter_next(data_iter));
+		}
+		kunmap_atomic(save_addr);
+		vmw_piter_next(pt_iter);
+	}
+
+	return num_pt_pages;
+}
+
+/*
+ * vmw_mob_build_pt - Set up a multilevel mob pagetable
+ *
+ * @mob:            Pointer to a mob whose page table needs setting up.
+ * @data_addr       Array of DMA addresses to the buffer object's data
+ *                  pages.
+ * @num_data_pages: Number of buffer object data pages.
+ *
+ * Uses tail recursion to set up a multilevel mob page table.
+ */
+static void vmw_mob_pt_setup(struct vmw_mob *mob,
+			     struct vmw_piter data_iter,
+			     unsigned long num_data_pages)
+{
+	unsigned long num_pt_pages = 0;
+	struct ttm_buffer_object *bo = mob->pt_bo;
+	struct vmw_piter save_pt_iter;
+	struct vmw_piter pt_iter;
+	const struct vmw_sg_table *vsgt;
+	int ret;
+
+	ret = ttm_bo_reserve(bo, false, true, false, NULL);
+	BUG_ON(ret != 0);
+
+	vsgt = vmw_bo_sg_table(bo);
+	vmw_piter_start(&pt_iter, vsgt, 0);
+	BUG_ON(!vmw_piter_next(&pt_iter));
+	mob->pt_level = 0;
+	while (likely(num_data_pages > 1)) {
+		++mob->pt_level;
+		BUG_ON(mob->pt_level > 2);
+		save_pt_iter = pt_iter;
+		num_pt_pages = vmw_mob_build_pt(&data_iter, num_data_pages,
+						&pt_iter);
+		data_iter = save_pt_iter;
+		num_data_pages = num_pt_pages;
+	}
+
+	mob->pt_root_page = vmw_piter_dma_addr(&save_pt_iter);
+	ttm_bo_unreserve(bo);
+}
+
+/*
+ * vmw_mob_destroy - Destroy a mob, unpopulating first if necessary.
+ *
+ * @mob:            Pointer to a mob to destroy.
+ */
+void vmw_mob_destroy(struct vmw_mob *mob)
+{
+	if (mob->pt_bo)
+		ttm_bo_unref(&mob->pt_bo);
+	kfree(mob);
+}
+
+/*
+ * vmw_mob_unbind - Hide a mob from the device.
+ *
+ * @dev_priv:       Pointer to a device private.
+ * @mob_id:         Device id of the mob to unbind.
+ */
+void vmw_mob_unbind(struct vmw_private *dev_priv,
+		    struct vmw_mob *mob)
+{
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDestroyGBMob body;
+	} *cmd;
+	int ret;
+	struct ttm_buffer_object *bo = mob->pt_bo;
+
+	if (bo) {
+		ret = ttm_bo_reserve(bo, false, true, false, NULL);
+		/*
+		 * Noone else should be using this buffer.
+		 */
+		BUG_ON(ret != 0);
+	}
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for Memory "
+			  "Object unbinding.\n");
+	} else {
+		cmd->header.id = SVGA_3D_CMD_DESTROY_GB_MOB;
+		cmd->header.size = sizeof(cmd->body);
+		cmd->body.mobid = mob->id;
+		vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	}
+	if (bo) {
+		vmw_fence_single_bo(bo, NULL);
+		ttm_bo_unreserve(bo);
+	}
+	vmw_3d_resource_dec(dev_priv, false);
+}
+
+/*
+ * vmw_mob_bind - Make a mob visible to the device after first
+ *                populating it if necessary.
+ *
+ * @dev_priv:       Pointer to a device private.
+ * @mob:            Pointer to the mob we're making visible.
+ * @data_addr:      Array of DMA addresses to the data pages of the underlying
+ *                  buffer object.
+ * @num_data_pages: Number of data pages of the underlying buffer
+ *                  object.
+ * @mob_id:         Device id of the mob to bind
+ *
+ * This function is intended to be interfaced with the ttm_tt backend
+ * code.
+ */
+int vmw_mob_bind(struct vmw_private *dev_priv,
+		 struct vmw_mob *mob,
+		 const struct vmw_sg_table *vsgt,
+		 unsigned long num_data_pages,
+		 int32_t mob_id)
+{
+	int ret;
+	bool pt_set_up = false;
+	struct vmw_piter data_iter;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineGBMob64 body;
+	} *cmd;
+
+	mob->id = mob_id;
+	vmw_piter_start(&data_iter, vsgt, 0);
+	if (unlikely(!vmw_piter_next(&data_iter)))
+		return 0;
+
+	if (likely(num_data_pages == 1)) {
+		mob->pt_level = VMW_MOBFMT_PTDEPTH_0;
+		mob->pt_root_page = vmw_piter_dma_addr(&data_iter);
+	} else if (vsgt->num_regions == 1) {
+		mob->pt_level = SVGA3D_MOBFMT_RANGE;
+		mob->pt_root_page = vmw_piter_dma_addr(&data_iter);
+	} else if (unlikely(mob->pt_bo == NULL)) {
+		ret = vmw_mob_pt_populate(dev_priv, mob);
+		if (unlikely(ret != 0))
+			return ret;
+
+		vmw_mob_pt_setup(mob, data_iter, num_data_pages);
+		pt_set_up = true;
+		mob->pt_level += VMW_MOBFMT_PTDEPTH_1 - SVGA3D_MOBFMT_PTDEPTH_1;
+	}
+
+	(void) vmw_3d_resource_inc(dev_priv, false);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for Memory "
+			  "Object binding.\n");
+		goto out_no_cmd_space;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DEFINE_GB_MOB64;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.mobid = mob_id;
+	cmd->body.ptDepth = mob->pt_level;
+	cmd->body.base = cpu_to_le64(mob->pt_root_page >> PAGE_SHIFT);
+	cmd->body.sizeInBytes = num_data_pages * PAGE_SIZE;
+
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+
+out_no_cmd_space:
+	vmw_3d_resource_dec(dev_priv, false);
+	if (pt_set_up)
+		ttm_bo_unref(&mob->pt_bo);
+
+	return -ENOMEM;
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index 9b5ea2ac7dd..01d68f0a69d 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
@@ -88,6 +88,11 @@ struct vmw_resource *vmw_resource_reference(struct vmw_resource *res)
 	return res;
 }
 
+struct vmw_resource *
+vmw_resource_reference_unless_doomed(struct vmw_resource *res)
+{
+	return kref_get_unless_zero(&res->kref) ? res : NULL;
+}
 
 /**
  * vmw_resource_release_id - release a resource id to the id manager.
@@ -136,8 +141,12 @@ static void vmw_resource_release(struct kref *kref)
 		vmw_dmabuf_unreference(&res->backup);
 	}
 
-	if (likely(res->hw_destroy != NULL))
+	if (likely(res->hw_destroy != NULL)) {
 		res->hw_destroy(res);
+		mutex_lock(&dev_priv->binding_mutex);
+		vmw_context_binding_res_list_kill(&res->binding_head);
+		mutex_unlock(&dev_priv->binding_mutex);
+	}
 
 	id = res->id;
 	if (res->res_free != NULL)
@@ -215,6 +224,7 @@ int vmw_resource_init(struct vmw_private *dev_priv, struct vmw_resource *res,
 	res->func = func;
 	INIT_LIST_HEAD(&res->lru_head);
 	INIT_LIST_HEAD(&res->mob_head);
+	INIT_LIST_HEAD(&res->binding_head);
 	res->id = -1;
 	res->backup = NULL;
 	res->backup_offset = 0;
@@ -417,8 +427,7 @@ int vmw_dmabuf_init(struct vmw_private *dev_priv,
 	INIT_LIST_HEAD(&vmw_bo->res_list);
 
 	ret = ttm_bo_init(bdev, &vmw_bo->base, size,
-			  (user) ? ttm_bo_type_device :
-			  ttm_bo_type_kernel, placement,
+			  ttm_bo_type_device, placement,
 			  0, interruptible,
 			  NULL, acc_size, NULL, bo_free);
 	return ret;
@@ -441,6 +450,21 @@ static void vmw_user_dmabuf_release(struct ttm_base_object **p_base)
 	ttm_bo_unref(&bo);
 }
 
+static void vmw_user_dmabuf_ref_obj_release(struct ttm_base_object *base,
+					    enum ttm_ref_type ref_type)
+{
+	struct vmw_user_dma_buffer *user_bo;
+	user_bo = container_of(base, struct vmw_user_dma_buffer, prime.base);
+
+	switch (ref_type) {
+	case TTM_REF_SYNCCPU_WRITE:
+		ttm_bo_synccpu_write_release(&user_bo->dma.base);
+		break;
+	default:
+		BUG();
+	}
+}
+
 /**
  * vmw_user_dmabuf_alloc - Allocate a user dma buffer
  *
@@ -471,6 +495,8 @@ int vmw_user_dmabuf_alloc(struct vmw_private *dev_priv,
 	}
 
 	ret = vmw_dmabuf_init(dev_priv, &user_bo->dma, size,
+			      (dev_priv->has_mob) ?
+			      &vmw_sys_placement :
 			      &vmw_vram_sys_placement, true,
 			      &vmw_user_dmabuf_destroy);
 	if (unlikely(ret != 0))
@@ -482,7 +508,8 @@ int vmw_user_dmabuf_alloc(struct vmw_private *dev_priv,
 				    &user_bo->prime,
 				    shareable,
 				    ttm_buffer_type,
-				    &vmw_user_dmabuf_release, NULL);
+				    &vmw_user_dmabuf_release,
+				    &vmw_user_dmabuf_ref_obj_release);
 	if (unlikely(ret != 0)) {
 		ttm_bo_unref(&tmp);
 		goto out_no_base_object;
@@ -511,8 +538,137 @@ int vmw_user_dmabuf_verify_access(struct ttm_buffer_object *bo,
 		return -EPERM;
 
 	vmw_user_bo = vmw_user_dma_buffer(bo);
-	return (vmw_user_bo->prime.base.tfile == tfile ||
-		vmw_user_bo->prime.base.shareable) ? 0 : -EPERM;
+
+	/* Check that the caller has opened the object. */
+	if (likely(ttm_ref_object_exists(tfile, &vmw_user_bo->prime.base)))
+		return 0;
+
+	DRM_ERROR("Could not grant buffer access.\n");
+	return -EPERM;
+}
+
+/**
+ * vmw_user_dmabuf_synccpu_grab - Grab a struct vmw_user_dma_buffer for cpu
+ * access, idling previous GPU operations on the buffer and optionally
+ * blocking it for further command submissions.
+ *
+ * @user_bo: Pointer to the buffer object being grabbed for CPU access
+ * @tfile: Identifying the caller.
+ * @flags: Flags indicating how the grab should be performed.
+ *
+ * A blocking grab will be automatically released when @tfile is closed.
+ */
+static int vmw_user_dmabuf_synccpu_grab(struct vmw_user_dma_buffer *user_bo,
+					struct ttm_object_file *tfile,
+					uint32_t flags)
+{
+	struct ttm_buffer_object *bo = &user_bo->dma.base;
+	bool existed;
+	int ret;
+
+	if (flags & drm_vmw_synccpu_allow_cs) {
+		struct ttm_bo_device *bdev = bo->bdev;
+
+		spin_lock(&bdev->fence_lock);
+		ret = ttm_bo_wait(bo, false, true,
+				  !!(flags & drm_vmw_synccpu_dontblock));
+		spin_unlock(&bdev->fence_lock);
+		return ret;
+	}
+
+	ret = ttm_bo_synccpu_write_grab
+		(bo, !!(flags & drm_vmw_synccpu_dontblock));
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_ref_object_add(tfile, &user_bo->prime.base,
+				 TTM_REF_SYNCCPU_WRITE, &existed);
+	if (ret != 0 || existed)
+		ttm_bo_synccpu_write_release(&user_bo->dma.base);
+
+	return ret;
+}
+
+/**
+ * vmw_user_dmabuf_synccpu_release - Release a previous grab for CPU access,
+ * and unblock command submission on the buffer if blocked.
+ *
+ * @handle: Handle identifying the buffer object.
+ * @tfile: Identifying the caller.
+ * @flags: Flags indicating the type of release.
+ */
+static int vmw_user_dmabuf_synccpu_release(uint32_t handle,
+					   struct ttm_object_file *tfile,
+					   uint32_t flags)
+{
+	if (!(flags & drm_vmw_synccpu_allow_cs))
+		return ttm_ref_object_base_unref(tfile, handle,
+						 TTM_REF_SYNCCPU_WRITE);
+
+	return 0;
+}
+
+/**
+ * vmw_user_dmabuf_synccpu_release - ioctl function implementing the synccpu
+ * functionality.
+ *
+ * @dev: Identifies the drm device.
+ * @data: Pointer to the ioctl argument.
+ * @file_priv: Identifies the caller.
+ *
+ * This function checks the ioctl arguments for validity and calls the
+ * relevant synccpu functions.
+ */
+int vmw_user_dmabuf_synccpu_ioctl(struct drm_device *dev, void *data,
+				  struct drm_file *file_priv)
+{
+	struct drm_vmw_synccpu_arg *arg =
+		(struct drm_vmw_synccpu_arg *) data;
+	struct vmw_dma_buffer *dma_buf;
+	struct vmw_user_dma_buffer *user_bo;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	int ret;
+
+	if ((arg->flags & (drm_vmw_synccpu_read | drm_vmw_synccpu_write)) == 0
+	    || (arg->flags & ~(drm_vmw_synccpu_read | drm_vmw_synccpu_write |
+			       drm_vmw_synccpu_dontblock |
+			       drm_vmw_synccpu_allow_cs)) != 0) {
+		DRM_ERROR("Illegal synccpu flags.\n");
+		return -EINVAL;
+	}
+
+	switch (arg->op) {
+	case drm_vmw_synccpu_grab:
+		ret = vmw_user_dmabuf_lookup(tfile, arg->handle, &dma_buf);
+		if (unlikely(ret != 0))
+			return ret;
+
+		user_bo = container_of(dma_buf, struct vmw_user_dma_buffer,
+				       dma);
+		ret = vmw_user_dmabuf_synccpu_grab(user_bo, tfile, arg->flags);
+		vmw_dmabuf_unreference(&dma_buf);
+		if (unlikely(ret != 0 && ret != -ERESTARTSYS &&
+			     ret != -EBUSY)) {
+			DRM_ERROR("Failed synccpu grab on handle 0x%08x.\n",
+				  (unsigned int) arg->handle);
+			return ret;
+		}
+		break;
+	case drm_vmw_synccpu_release:
+		ret = vmw_user_dmabuf_synccpu_release(arg->handle, tfile,
+						      arg->flags);
+		if (unlikely(ret != 0)) {
+			DRM_ERROR("Failed synccpu release on handle 0x%08x.\n",
+				  (unsigned int) arg->handle);
+			return ret;
+		}
+		break;
+	default:
+		DRM_ERROR("Invalid synccpu operation.\n");
+		return -EINVAL;
+	}
+
+	return 0;
 }
 
 int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
@@ -525,10 +681,9 @@ int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
 	struct drm_vmw_dmabuf_rep *rep = &arg->rep;
 	struct vmw_dma_buffer *dma_buf;
 	uint32_t handle;
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	int ret;
 
-	ret = ttm_read_lock(&vmaster->lock, true);
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -545,7 +700,7 @@ int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
 	vmw_dmabuf_unreference(&dma_buf);
 
 out_no_dmabuf:
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
 
 	return ret;
 }
@@ -591,7 +746,8 @@ int vmw_user_dmabuf_lookup(struct ttm_object_file *tfile,
 }
 
 int vmw_user_dmabuf_reference(struct ttm_object_file *tfile,
-			      struct vmw_dma_buffer *dma_buf)
+			      struct vmw_dma_buffer *dma_buf,
+			      uint32_t *handle)
 {
 	struct vmw_user_dma_buffer *user_bo;
 
@@ -599,6 +755,8 @@ int vmw_user_dmabuf_reference(struct ttm_object_file *tfile,
 		return -EINVAL;
 
 	user_bo = container_of(dma_buf, struct vmw_user_dma_buffer, dma);
+
+	*handle = user_bo->prime.base.hash.key;
 	return ttm_ref_object_add(tfile, &user_bo->prime.base,
 				  TTM_REF_USAGE, NULL);
 }
@@ -719,7 +877,6 @@ int vmw_stream_claim_ioctl(struct drm_device *dev, void *data,
 	struct vmw_resource *tmp;
 	struct drm_vmw_stream_arg *arg = (struct drm_vmw_stream_arg *)data;
 	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	int ret;
 
 	/*
@@ -730,7 +887,7 @@ int vmw_stream_claim_ioctl(struct drm_device *dev, void *data,
 	if (unlikely(vmw_user_stream_size == 0))
 		vmw_user_stream_size = ttm_round_pot(sizeof(*stream)) + 128;
 
-	ret = ttm_read_lock(&vmaster->lock, true);
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -778,7 +935,7 @@ int vmw_stream_claim_ioctl(struct drm_device *dev, void *data,
 out_err:
 	vmw_resource_unreference(&res);
 out_unlock:
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
 	return ret;
 }
 
@@ -831,14 +988,13 @@ int vmw_dumb_create(struct drm_file *file_priv,
 		    struct drm_mode_create_dumb *args)
 {
 	struct vmw_private *dev_priv = vmw_priv(dev);
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	struct vmw_dma_buffer *dma_buf;
 	int ret;
 
 	args->pitch = args->width * ((args->bpp + 7) / 8);
 	args->size = args->pitch * args->height;
 
-	ret = ttm_read_lock(&vmaster->lock, true);
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -850,7 +1006,7 @@ int vmw_dumb_create(struct drm_file *file_priv,
 
 	vmw_dmabuf_unreference(&dma_buf);
 out_no_dmabuf:
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
 	return ret;
 }
 
@@ -1291,11 +1447,54 @@ void vmw_fence_single_bo(struct ttm_buffer_object *bo,
  * @mem:            The truct ttm_mem_reg indicating to what memory
  *                  region the move is taking place.
  *
- * For now does nothing.
+ * Evicts the Guest Backed hardware resource if the backup
+ * buffer is being moved out of MOB memory.
+ * Note that this function should not race with the resource
+ * validation code as long as it accesses only members of struct
+ * resource that remain static while bo::res is !NULL and
+ * while we have @bo reserved. struct resource::backup is *not* a
+ * static member. The resource validation code will take care
+ * to set @bo::res to NULL, while having @bo reserved when the
+ * buffer is no longer bound to the resource, so @bo:res can be
+ * used to determine whether there is a need to unbind and whether
+ * it is safe to unbind.
  */
 void vmw_resource_move_notify(struct ttm_buffer_object *bo,
 			      struct ttm_mem_reg *mem)
 {
+	struct vmw_dma_buffer *dma_buf;
+
+	if (mem == NULL)
+		return;
+
+	if (bo->destroy != vmw_dmabuf_bo_free &&
+	    bo->destroy != vmw_user_dmabuf_destroy)
+		return;
+
+	dma_buf = container_of(bo, struct vmw_dma_buffer, base);
+
+	if (mem->mem_type != VMW_PL_MOB) {
+		struct vmw_resource *res, *n;
+		struct ttm_bo_device *bdev = bo->bdev;
+		struct ttm_validate_buffer val_buf;
+
+		val_buf.bo = bo;
+
+		list_for_each_entry_safe(res, n, &dma_buf->res_list, mob_head) {
+
+			if (unlikely(res->func->unbind == NULL))
+				continue;
+
+			(void) res->func->unbind(res, true, &val_buf);
+			res->backup_dirty = true;
+			res->res_dirty = false;
+			list_del_init(&res->mob_head);
+		}
+
+		spin_lock(&bdev->fence_lock);
+		(void) ttm_bo_wait(bo, false, false, false);
+		spin_unlock(&bdev->fence_lock);
+	}
 }
 
 /**
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c b/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c
index 22406c8651e..a95d3a0cabe 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c
@@ -307,7 +307,7 @@ static int vmw_sou_crtc_set_config(struct drm_mode_set *set)
 
 		connector->encoder = NULL;
 		encoder->crtc = NULL;
-		crtc->fb = NULL;
+		crtc->primary->fb = NULL;
 		crtc->x = 0;
 		crtc->y = 0;
 		crtc->enabled = false;
@@ -368,7 +368,7 @@ static int vmw_sou_crtc_set_config(struct drm_mode_set *set)
 
 		connector->encoder = NULL;
 		encoder->crtc = NULL;
-		crtc->fb = NULL;
+		crtc->primary->fb = NULL;
 		crtc->x = 0;
 		crtc->y = 0;
 		crtc->enabled = false;
@@ -381,7 +381,7 @@ static int vmw_sou_crtc_set_config(struct drm_mode_set *set)
 	connector->encoder = encoder;
 	encoder->crtc = crtc;
 	crtc->mode = *mode;
-	crtc->fb = fb;
+	crtc->primary->fb = fb;
 	crtc->x = set->x;
 	crtc->y = set->y;
 	crtc->enabled = true;
@@ -572,5 +572,5 @@ void vmw_kms_screen_object_update_implicit_fb(struct vmw_private *dev_priv,
 	BUG_ON(!sou->base.is_implicit);
 
 	dev_priv->sou_priv->implicit_fb =
-		vmw_framebuffer_to_vfb(sou->base.crtc.fb);
+		vmw_framebuffer_to_vfb(sou->base.crtc.primary->fb);
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
new file mode 100644
index 00000000000..c1559eeaffe
--- /dev/null
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
@@ -0,0 +1,811 @@
+/**************************************************************************
+ *
+ * Copyright © 2009-2012 VMware, Inc., Palo Alto, CA., USA
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ **************************************************************************/
+
+#include "vmwgfx_drv.h"
+#include "vmwgfx_resource_priv.h"
+#include "ttm/ttm_placement.h"
+
+#define VMW_COMPAT_SHADER_HT_ORDER 12
+
+struct vmw_shader {
+	struct vmw_resource res;
+	SVGA3dShaderType type;
+	uint32_t size;
+};
+
+struct vmw_user_shader {
+	struct ttm_base_object base;
+	struct vmw_shader shader;
+};
+
+/**
+ * enum vmw_compat_shader_state - Staging state for compat shaders
+ */
+enum vmw_compat_shader_state {
+	VMW_COMPAT_COMMITED,
+	VMW_COMPAT_ADD,
+	VMW_COMPAT_DEL
+};
+
+/**
+ * struct vmw_compat_shader - Metadata for compat shaders.
+ *
+ * @handle: The TTM handle of the guest backed shader.
+ * @tfile: The struct ttm_object_file the guest backed shader is registered
+ * with.
+ * @hash: Hash item for lookup.
+ * @head: List head for staging lists or the compat shader manager list.
+ * @state: Staging state.
+ *
+ * The structure is protected by the cmdbuf lock.
+ */
+struct vmw_compat_shader {
+	u32 handle;
+	struct ttm_object_file *tfile;
+	struct drm_hash_item hash;
+	struct list_head head;
+	enum vmw_compat_shader_state state;
+};
+
+/**
+ * struct vmw_compat_shader_manager - Compat shader manager.
+ *
+ * @shaders: Hash table containing staged and commited compat shaders
+ * @list: List of commited shaders.
+ * @dev_priv: Pointer to a device private structure.
+ *
+ * @shaders and @list are protected by the cmdbuf mutex for now.
+ */
+struct vmw_compat_shader_manager {
+	struct drm_open_hash shaders;
+	struct list_head list;
+	struct vmw_private *dev_priv;
+};
+
+static void vmw_user_shader_free(struct vmw_resource *res);
+static struct vmw_resource *
+vmw_user_shader_base_to_res(struct ttm_base_object *base);
+
+static int vmw_gb_shader_create(struct vmw_resource *res);
+static int vmw_gb_shader_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf);
+static int vmw_gb_shader_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf);
+static int vmw_gb_shader_destroy(struct vmw_resource *res);
+
+static uint64_t vmw_user_shader_size;
+
+static const struct vmw_user_resource_conv user_shader_conv = {
+	.object_type = VMW_RES_SHADER,
+	.base_obj_to_res = vmw_user_shader_base_to_res,
+	.res_free = vmw_user_shader_free
+};
+
+const struct vmw_user_resource_conv *user_shader_converter =
+	&user_shader_conv;
+
+
+static const struct vmw_res_func vmw_gb_shader_func = {
+	.res_type = vmw_res_shader,
+	.needs_backup = true,
+	.may_evict = true,
+	.type_name = "guest backed shaders",
+	.backup_placement = &vmw_mob_placement,
+	.create = vmw_gb_shader_create,
+	.destroy = vmw_gb_shader_destroy,
+	.bind = vmw_gb_shader_bind,
+	.unbind = vmw_gb_shader_unbind
+};
+
+/**
+ * Shader management:
+ */
+
+static inline struct vmw_shader *
+vmw_res_to_shader(struct vmw_resource *res)
+{
+	return container_of(res, struct vmw_shader, res);
+}
+
+static void vmw_hw_shader_destroy(struct vmw_resource *res)
+{
+	(void) vmw_gb_shader_destroy(res);
+}
+
+static int vmw_gb_shader_init(struct vmw_private *dev_priv,
+			      struct vmw_resource *res,
+			      uint32_t size,
+			      uint64_t offset,
+			      SVGA3dShaderType type,
+			      struct vmw_dma_buffer *byte_code,
+			      void (*res_free) (struct vmw_resource *res))
+{
+	struct vmw_shader *shader = vmw_res_to_shader(res);
+	int ret;
+
+	ret = vmw_resource_init(dev_priv, res, true,
+				res_free, &vmw_gb_shader_func);
+
+
+	if (unlikely(ret != 0)) {
+		if (res_free)
+			res_free(res);
+		else
+			kfree(res);
+		return ret;
+	}
+
+	res->backup_size = size;
+	if (byte_code) {
+		res->backup = vmw_dmabuf_reference(byte_code);
+		res->backup_offset = offset;
+	}
+	shader->size = size;
+	shader->type = type;
+
+	vmw_resource_activate(res, vmw_hw_shader_destroy);
+	return 0;
+}
+
+static int vmw_gb_shader_create(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct vmw_shader *shader = vmw_res_to_shader(res);
+	int ret;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineGBShader body;
+	} *cmd;
+
+	if (likely(res->id != -1))
+		return 0;
+
+	ret = vmw_resource_alloc_id(res);
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Failed to allocate a shader id.\n");
+		goto out_no_id;
+	}
+
+	if (unlikely(res->id >= VMWGFX_NUM_GB_SHADER)) {
+		ret = -EBUSY;
+		goto out_no_fifo;
+	}
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "creation.\n");
+		ret = -ENOMEM;
+		goto out_no_fifo;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DEFINE_GB_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.shid = res->id;
+	cmd->body.type = shader->type;
+	cmd->body.sizeInBytes = shader->size;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	(void) vmw_3d_resource_inc(dev_priv, false);
+
+	return 0;
+
+out_no_fifo:
+	vmw_resource_release_id(res);
+out_no_id:
+	return ret;
+}
+
+static int vmw_gb_shader_bind(struct vmw_resource *res,
+			      struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBShader body;
+	} *cmd;
+	struct ttm_buffer_object *bo = val_buf->bo;
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "binding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_BIND_GB_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.shid = res->id;
+	cmd->body.mobid = bo->mem.start;
+	cmd->body.offsetInBytes = 0;
+	res->backup_dirty = false;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+static int vmw_gb_shader_unbind(struct vmw_resource *res,
+				bool readback,
+				struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBShader body;
+	} *cmd;
+	struct vmw_fence_obj *fence;
+
+	BUG_ON(res->backup->base.mem.mem_type != VMW_PL_MOB);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_BIND_GB_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.shid = res->id;
+	cmd->body.mobid = SVGA3D_INVALID_ID;
+	cmd->body.offsetInBytes = 0;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	/*
+	 * Create a fence object and fence the backup buffer.
+	 */
+
+	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
+					  &fence, NULL);
+
+	vmw_fence_single_bo(val_buf->bo, fence);
+
+	if (likely(fence != NULL))
+		vmw_fence_obj_unreference(&fence);
+
+	return 0;
+}
+
+static int vmw_gb_shader_destroy(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDestroyGBShader body;
+	} *cmd;
+
+	if (likely(res->id == -1))
+		return 0;
+
+	mutex_lock(&dev_priv->binding_mutex);
+	vmw_context_binding_res_list_scrub(&res->binding_head);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "destruction.\n");
+		mutex_unlock(&dev_priv->binding_mutex);
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DESTROY_GB_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.shid = res->id;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	mutex_unlock(&dev_priv->binding_mutex);
+	vmw_resource_release_id(res);
+	vmw_3d_resource_dec(dev_priv, false);
+
+	return 0;
+}
+
+/**
+ * User-space shader management:
+ */
+
+static struct vmw_resource *
+vmw_user_shader_base_to_res(struct ttm_base_object *base)
+{
+	return &(container_of(base, struct vmw_user_shader, base)->
+		 shader.res);
+}
+
+static void vmw_user_shader_free(struct vmw_resource *res)
+{
+	struct vmw_user_shader *ushader =
+		container_of(res, struct vmw_user_shader, shader.res);
+	struct vmw_private *dev_priv = res->dev_priv;
+
+	ttm_base_object_kfree(ushader, base);
+	ttm_mem_global_free(vmw_mem_glob(dev_priv),
+			    vmw_user_shader_size);
+}
+
+/**
+ * This function is called when user space has no more references on the
+ * base object. It releases the base-object's reference on the resource object.
+ */
+
+static void vmw_user_shader_base_release(struct ttm_base_object **p_base)
+{
+	struct ttm_base_object *base = *p_base;
+	struct vmw_resource *res = vmw_user_shader_base_to_res(base);
+
+	*p_base = NULL;
+	vmw_resource_unreference(&res);
+}
+
+int vmw_shader_destroy_ioctl(struct drm_device *dev, void *data,
+			      struct drm_file *file_priv)
+{
+	struct drm_vmw_shader_arg *arg = (struct drm_vmw_shader_arg *)data;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+
+	return ttm_ref_object_base_unref(tfile, arg->handle,
+					 TTM_REF_USAGE);
+}
+
+static int vmw_shader_alloc(struct vmw_private *dev_priv,
+			    struct vmw_dma_buffer *buffer,
+			    size_t shader_size,
+			    size_t offset,
+			    SVGA3dShaderType shader_type,
+			    struct ttm_object_file *tfile,
+			    u32 *handle)
+{
+	struct vmw_user_shader *ushader;
+	struct vmw_resource *res, *tmp;
+	int ret;
+
+	/*
+	 * Approximate idr memory usage with 128 bytes. It will be limited
+	 * by maximum number_of shaders anyway.
+	 */
+	if (unlikely(vmw_user_shader_size == 0))
+		vmw_user_shader_size =
+			ttm_round_pot(sizeof(struct vmw_user_shader)) + 128;
+
+	ret = ttm_mem_global_alloc(vmw_mem_glob(dev_priv),
+				   vmw_user_shader_size,
+				   false, true);
+	if (unlikely(ret != 0)) {
+		if (ret != -ERESTARTSYS)
+			DRM_ERROR("Out of graphics memory for shader "
+				  "creation.\n");
+		goto out;
+	}
+
+	ushader = kzalloc(sizeof(*ushader), GFP_KERNEL);
+	if (unlikely(ushader == NULL)) {
+		ttm_mem_global_free(vmw_mem_glob(dev_priv),
+				    vmw_user_shader_size);
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	res = &ushader->shader.res;
+	ushader->base.shareable = false;
+	ushader->base.tfile = NULL;
+
+	/*
+	 * From here on, the destructor takes over resource freeing.
+	 */
+
+	ret = vmw_gb_shader_init(dev_priv, res, shader_size,
+				 offset, shader_type, buffer,
+				 vmw_user_shader_free);
+	if (unlikely(ret != 0))
+		goto out;
+
+	tmp = vmw_resource_reference(res);
+	ret = ttm_base_object_init(tfile, &ushader->base, false,
+				   VMW_RES_SHADER,
+				   &vmw_user_shader_base_release, NULL);
+
+	if (unlikely(ret != 0)) {
+		vmw_resource_unreference(&tmp);
+		goto out_err;
+	}
+
+	if (handle)
+		*handle = ushader->base.hash.key;
+out_err:
+	vmw_resource_unreference(&res);
+out:
+	return ret;
+}
+
+
+int vmw_shader_define_ioctl(struct drm_device *dev, void *data,
+			     struct drm_file *file_priv)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	struct drm_vmw_shader_create_arg *arg =
+		(struct drm_vmw_shader_create_arg *)data;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	struct vmw_dma_buffer *buffer = NULL;
+	SVGA3dShaderType shader_type;
+	int ret;
+
+	if (arg->buffer_handle != SVGA3D_INVALID_ID) {
+		ret = vmw_user_dmabuf_lookup(tfile, arg->buffer_handle,
+					     &buffer);
+		if (unlikely(ret != 0)) {
+			DRM_ERROR("Could not find buffer for shader "
+				  "creation.\n");
+			return ret;
+		}
+
+		if ((u64)buffer->base.num_pages * PAGE_SIZE <
+		    (u64)arg->size + (u64)arg->offset) {
+			DRM_ERROR("Illegal buffer- or shader size.\n");
+			ret = -EINVAL;
+			goto out_bad_arg;
+		}
+	}
+
+	switch (arg->shader_type) {
+	case drm_vmw_shader_type_vs:
+		shader_type = SVGA3D_SHADERTYPE_VS;
+		break;
+	case drm_vmw_shader_type_ps:
+		shader_type = SVGA3D_SHADERTYPE_PS;
+		break;
+	case drm_vmw_shader_type_gs:
+		shader_type = SVGA3D_SHADERTYPE_GS;
+		break;
+	default:
+		DRM_ERROR("Illegal shader type.\n");
+		ret = -EINVAL;
+		goto out_bad_arg;
+	}
+
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
+	if (unlikely(ret != 0))
+		goto out_bad_arg;
+
+	ret = vmw_shader_alloc(dev_priv, buffer, arg->size, arg->offset,
+			       shader_type, tfile, &arg->shader_handle);
+
+	ttm_read_unlock(&dev_priv->reservation_sem);
+out_bad_arg:
+	vmw_dmabuf_unreference(&buffer);
+	return ret;
+}
+
+/**
+ * vmw_compat_shader_lookup - Look up a compat shader
+ *
+ * @man: Pointer to the compat shader manager.
+ * @shader_type: The shader type, that combined with the user_key identifies
+ * the shader.
+ * @user_key: On entry, this should be a pointer to the user_key.
+ * On successful exit, it will contain the guest-backed shader's TTM handle.
+ *
+ * Returns 0 on success. Non-zero on failure, in which case the value pointed
+ * to by @user_key is unmodified.
+ */
+int vmw_compat_shader_lookup(struct vmw_compat_shader_manager *man,
+			     SVGA3dShaderType shader_type,
+			     u32 *user_key)
+{
+	struct drm_hash_item *hash;
+	int ret;
+	unsigned long key = *user_key | (shader_type << 24);
+
+	ret = drm_ht_find_item(&man->shaders, key, &hash);
+	if (unlikely(ret != 0))
+		return ret;
+
+	*user_key = drm_hash_entry(hash, struct vmw_compat_shader,
+				   hash)->handle;
+
+	return 0;
+}
+
+/**
+ * vmw_compat_shader_free - Free a compat shader.
+ *
+ * @man: Pointer to the compat shader manager.
+ * @entry: Pointer to a struct vmw_compat_shader.
+ *
+ * Frees a struct vmw_compat_shder entry and drops its reference to the
+ * guest backed shader.
+ */
+static void vmw_compat_shader_free(struct vmw_compat_shader_manager *man,
+				   struct vmw_compat_shader *entry)
+{
+	list_del(&entry->head);
+	WARN_ON(drm_ht_remove_item(&man->shaders, &entry->hash));
+	WARN_ON(ttm_ref_object_base_unref(entry->tfile, entry->handle,
+					  TTM_REF_USAGE));
+	kfree(entry);
+}
+
+/**
+ * vmw_compat_shaders_commit - Commit a list of compat shader actions.
+ *
+ * @man: Pointer to the compat shader manager.
+ * @list: Caller's list of compat shader actions.
+ *
+ * This function commits a list of compat shader additions or removals.
+ * It is typically called when the execbuf ioctl call triggering these
+ * actions has commited the fifo contents to the device.
+ */
+void vmw_compat_shaders_commit(struct vmw_compat_shader_manager *man,
+			       struct list_head *list)
+{
+	struct vmw_compat_shader *entry, *next;
+
+	list_for_each_entry_safe(entry, next, list, head) {
+		list_del(&entry->head);
+		switch (entry->state) {
+		case VMW_COMPAT_ADD:
+			entry->state = VMW_COMPAT_COMMITED;
+			list_add_tail(&entry->head, &man->list);
+			break;
+		case VMW_COMPAT_DEL:
+			ttm_ref_object_base_unref(entry->tfile, entry->handle,
+						  TTM_REF_USAGE);
+			kfree(entry);
+			break;
+		default:
+			BUG();
+			break;
+		}
+	}
+}
+
+/**
+ * vmw_compat_shaders_revert - Revert a list of compat shader actions
+ *
+ * @man: Pointer to the compat shader manager.
+ * @list: Caller's list of compat shader actions.
+ *
+ * This function reverts a list of compat shader additions or removals.
+ * It is typically called when the execbuf ioctl call triggering these
+ * actions failed for some reason, and the command stream was never
+ * submitted.
+ */
+void vmw_compat_shaders_revert(struct vmw_compat_shader_manager *man,
+			       struct list_head *list)
+{
+	struct vmw_compat_shader *entry, *next;
+	int ret;
+
+	list_for_each_entry_safe(entry, next, list, head) {
+		switch (entry->state) {
+		case VMW_COMPAT_ADD:
+			vmw_compat_shader_free(man, entry);
+			break;
+		case VMW_COMPAT_DEL:
+			ret = drm_ht_insert_item(&man->shaders, &entry->hash);
+			list_del(&entry->head);
+			list_add_tail(&entry->head, &man->list);
+			entry->state = VMW_COMPAT_COMMITED;
+			break;
+		default:
+			BUG();
+			break;
+		}
+	}
+}
+
+/**
+ * vmw_compat_shader_remove - Stage a compat shader for removal.
+ *
+ * @man: Pointer to the compat shader manager
+ * @user_key: The key that is used to identify the shader. The key is
+ * unique to the shader type.
+ * @shader_type: Shader type.
+ * @list: Caller's list of staged shader actions.
+ *
+ * This function stages a compat shader for removal and removes the key from
+ * the shader manager's hash table. If the shader was previously only staged
+ * for addition it is completely removed (But the execbuf code may keep a
+ * reference if it was bound to a context between addition and removal). If
+ * it was previously commited to the manager, it is staged for removal.
+ */
+int vmw_compat_shader_remove(struct vmw_compat_shader_manager *man,
+			     u32 user_key, SVGA3dShaderType shader_type,
+			     struct list_head *list)
+{
+	struct vmw_compat_shader *entry;
+	struct drm_hash_item *hash;
+	int ret;
+
+	ret = drm_ht_find_item(&man->shaders, user_key | (shader_type << 24),
+			       &hash);
+	if (likely(ret != 0))
+		return -EINVAL;
+
+	entry = drm_hash_entry(hash, struct vmw_compat_shader, hash);
+
+	switch (entry->state) {
+	case VMW_COMPAT_ADD:
+		vmw_compat_shader_free(man, entry);
+		break;
+	case VMW_COMPAT_COMMITED:
+		(void) drm_ht_remove_item(&man->shaders, &entry->hash);
+		list_del(&entry->head);
+		entry->state = VMW_COMPAT_DEL;
+		list_add_tail(&entry->head, list);
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	return 0;
+}
+
+/**
+ * vmw_compat_shader_add - Create a compat shader and add the
+ * key to the manager
+ *
+ * @man: Pointer to the compat shader manager
+ * @user_key: The key that is used to identify the shader. The key is
+ * unique to the shader type.
+ * @bytecode: Pointer to the bytecode of the shader.
+ * @shader_type: Shader type.
+ * @tfile: Pointer to a struct ttm_object_file that the guest-backed shader is
+ * to be created with.
+ * @list: Caller's list of staged shader actions.
+ *
+ * Note that only the key is added to the shader manager's hash table.
+ * The shader is not yet added to the shader manager's list of shaders.
+ */
+int vmw_compat_shader_add(struct vmw_compat_shader_manager *man,
+			  u32 user_key, const void *bytecode,
+			  SVGA3dShaderType shader_type,
+			  size_t size,
+			  struct ttm_object_file *tfile,
+			  struct list_head *list)
+{
+	struct vmw_dma_buffer *buf;
+	struct ttm_bo_kmap_obj map;
+	bool is_iomem;
+	struct vmw_compat_shader *compat;
+	u32 handle;
+	int ret;
+
+	if (user_key > ((1 << 24) - 1) || (unsigned) shader_type > 16)
+		return -EINVAL;
+
+	/* Allocate and pin a DMA buffer */
+	buf = kzalloc(sizeof(*buf), GFP_KERNEL);
+	if (unlikely(buf == NULL))
+		return -ENOMEM;
+
+	ret = vmw_dmabuf_init(man->dev_priv, buf, size, &vmw_sys_ne_placement,
+			      true, vmw_dmabuf_bo_free);
+	if (unlikely(ret != 0))
+		goto out;
+
+	ret = ttm_bo_reserve(&buf->base, false, true, false, NULL);
+	if (unlikely(ret != 0))
+		goto no_reserve;
+
+	/* Map and copy shader bytecode. */
+	ret = ttm_bo_kmap(&buf->base, 0, PAGE_ALIGN(size) >> PAGE_SHIFT,
+			  &map);
+	if (unlikely(ret != 0)) {
+		ttm_bo_unreserve(&buf->base);
+		goto no_reserve;
+	}
+
+	memcpy(ttm_kmap_obj_virtual(&map, &is_iomem), bytecode, size);
+	WARN_ON(is_iomem);
+
+	ttm_bo_kunmap(&map);
+	ret = ttm_bo_validate(&buf->base, &vmw_sys_placement, false, true);
+	WARN_ON(ret != 0);
+	ttm_bo_unreserve(&buf->base);
+
+	/* Create a guest-backed shader container backed by the dma buffer */
+	ret = vmw_shader_alloc(man->dev_priv, buf, size, 0, shader_type,
+			       tfile, &handle);
+	vmw_dmabuf_unreference(&buf);
+	if (unlikely(ret != 0))
+		goto no_reserve;
+	/*
+	 * Create a compat shader structure and stage it for insertion
+	 * in the manager
+	 */
+	compat = kzalloc(sizeof(*compat), GFP_KERNEL);
+	if (compat == NULL)
+		goto no_compat;
+
+	compat->hash.key = user_key |  (shader_type << 24);
+	ret = drm_ht_insert_item(&man->shaders, &compat->hash);
+	if (unlikely(ret != 0))
+		goto out_invalid_key;
+
+	compat->state = VMW_COMPAT_ADD;
+	compat->handle = handle;
+	compat->tfile = tfile;
+	list_add_tail(&compat->head, list);
+
+	return 0;
+
+out_invalid_key:
+	kfree(compat);
+no_compat:
+	ttm_ref_object_base_unref(tfile, handle, TTM_REF_USAGE);
+no_reserve:
+out:
+	return ret;
+}
+
+/**
+ * vmw_compat_shader_man_create - Create a compat shader manager
+ *
+ * @dev_priv: Pointer to a device private structure.
+ *
+ * Typically done at file open time. If successful returns a pointer to a
+ * compat shader manager. Otherwise returns an error pointer.
+ */
+struct vmw_compat_shader_manager *
+vmw_compat_shader_man_create(struct vmw_private *dev_priv)
+{
+	struct vmw_compat_shader_manager *man;
+	int ret;
+
+	man = kzalloc(sizeof(*man), GFP_KERNEL);
+	if (man == NULL)
+		return ERR_PTR(-ENOMEM);
+
+	man->dev_priv = dev_priv;
+	INIT_LIST_HEAD(&man->list);
+	ret = drm_ht_create(&man->shaders, VMW_COMPAT_SHADER_HT_ORDER);
+	if (ret == 0)
+		return man;
+
+	kfree(man);
+	return ERR_PTR(ret);
+}
+
+/**
+ * vmw_compat_shader_man_destroy - Destroy a compat shader manager
+ *
+ * @man: Pointer to the shader manager to destroy.
+ *
+ * Typically done at file close time.
+ */
+void vmw_compat_shader_man_destroy(struct vmw_compat_shader_manager *man)
+{
+	struct vmw_compat_shader *entry, *next;
+
+	mutex_lock(&man->dev_priv->cmdbuf_mutex);
+	list_for_each_entry_safe(entry, next, &man->list, head)
+		vmw_compat_shader_free(man, entry);
+
+	mutex_unlock(&man->dev_priv->cmdbuf_mutex);
+	kfree(man);
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
index 7de2ea8bd55..4ecdbf3e59d 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
@@ -36,12 +36,13 @@
  * @base:           The TTM base object handling user-space visibility.
  * @srf:            The surface metadata.
  * @size:           TTM accounting size for the surface.
+ * @master:         master of the creating client. Used for security check.
  */
 struct vmw_user_surface {
 	struct ttm_prime_object prime;
 	struct vmw_surface srf;
 	uint32_t size;
-	uint32_t backup_handle;
+	struct drm_master *master;
 };
 
 /**
@@ -68,6 +69,14 @@ static int vmw_legacy_srf_unbind(struct vmw_resource *res,
 				 struct ttm_validate_buffer *val_buf);
 static int vmw_legacy_srf_create(struct vmw_resource *res);
 static int vmw_legacy_srf_destroy(struct vmw_resource *res);
+static int vmw_gb_surface_create(struct vmw_resource *res);
+static int vmw_gb_surface_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf);
+static int vmw_gb_surface_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf);
+static int vmw_gb_surface_destroy(struct vmw_resource *res);
+
 
 static const struct vmw_user_resource_conv user_surface_conv = {
 	.object_type = VMW_RES_SURFACE,
@@ -93,6 +102,18 @@ static const struct vmw_res_func vmw_legacy_surface_func = {
 	.unbind = &vmw_legacy_srf_unbind
 };
 
+static const struct vmw_res_func vmw_gb_surface_func = {
+	.res_type = vmw_res_surface,
+	.needs_backup = true,
+	.may_evict = true,
+	.type_name = "guest backed surfaces",
+	.backup_placement = &vmw_mob_placement,
+	.create = vmw_gb_surface_create,
+	.destroy = vmw_gb_surface_destroy,
+	.bind = vmw_gb_surface_bind,
+	.unbind = vmw_gb_surface_unbind
+};
+
 /**
  * struct vmw_surface_dma - SVGA3D DMA command
  */
@@ -291,6 +312,11 @@ static void vmw_hw_surface_destroy(struct vmw_resource *res)
 	struct vmw_surface *srf;
 	void *cmd;
 
+	if (res->func->destroy == vmw_gb_surface_destroy) {
+		(void) vmw_gb_surface_destroy(res);
+		return;
+	}
+
 	if (res->id != -1) {
 
 		cmd = vmw_fifo_reserve(dev_priv, vmw_surface_destroy_size());
@@ -549,12 +575,15 @@ static int vmw_surface_init(struct vmw_private *dev_priv,
 	struct vmw_resource *res = &srf->res;
 
 	BUG_ON(res_free == NULL);
-	(void) vmw_3d_resource_inc(dev_priv, false);
+	if (!dev_priv->has_mob)
+		(void) vmw_3d_resource_inc(dev_priv, false);
 	ret = vmw_resource_init(dev_priv, res, true, res_free,
+				(dev_priv->has_mob) ? &vmw_gb_surface_func :
 				&vmw_legacy_surface_func);
 
 	if (unlikely(ret != 0)) {
-		vmw_3d_resource_dec(dev_priv, false);
+		if (!dev_priv->has_mob)
+			vmw_3d_resource_dec(dev_priv, false);
 		res_free(res);
 		return ret;
 	}
@@ -597,6 +626,8 @@ static void vmw_user_surface_free(struct vmw_resource *res)
 	struct vmw_private *dev_priv = srf->res.dev_priv;
 	uint32_t size = user_srf->size;
 
+	if (user_srf->master)
+		drm_master_put(&user_srf->master);
 	kfree(srf->offsets);
 	kfree(srf->sizes);
 	kfree(srf->snooper.image);
@@ -670,7 +701,6 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 	struct vmw_surface_offset *cur_offset;
 	uint32_t num_sizes;
 	uint32_t size;
-	struct vmw_master *vmaster = vmw_master(file_priv->master);
 	const struct svga3d_surface_desc *desc;
 
 	if (unlikely(vmw_user_surface_size == 0))
@@ -696,7 +726,7 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
-	ret = ttm_read_lock(&vmaster->lock, true);
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -750,7 +780,7 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 
 	srf->base_size = *srf->sizes;
 	srf->autogen_filter = SVGA3D_TEX_FILTER_NONE;
-	srf->multisample_count = 1;
+	srf->multisample_count = 0;
 
 	cur_bo_offset = 0;
 	cur_offset = srf->offsets;
@@ -793,6 +823,8 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 
 	user_srf->prime.base.shareable = false;
 	user_srf->prime.base.tfile = NULL;
+	if (drm_is_primary_client(file_priv))
+		user_srf->master = drm_master_get(file_priv->master);
 
 	/**
 	 * From this point, the generic resource management functions
@@ -803,6 +835,24 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 	if (unlikely(ret != 0))
 		goto out_unlock;
 
+	/*
+	 * A gb-aware client referencing a shared surface will
+	 * expect a backup buffer to be present.
+	 */
+	if (dev_priv->has_mob && req->shareable) {
+		uint32_t backup_handle;
+
+		ret = vmw_user_dmabuf_alloc(dev_priv, tfile,
+					    res->backup_size,
+					    true,
+					    &backup_handle,
+					    &res->backup);
+		if (unlikely(ret != 0)) {
+			vmw_resource_unreference(&res);
+			goto out_unlock;
+		}
+	}
+
 	tmp = vmw_resource_reference(&srf->res);
 	ret = ttm_prime_object_init(tfile, res->backup_size, &user_srf->prime,
 				    req->shareable, VMW_RES_SURFACE,
@@ -817,7 +867,7 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 	rep->sid = user_srf->prime.base.hash.key;
 	vmw_resource_unreference(&res);
 
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
 	return 0;
 out_no_copy:
 	kfree(srf->offsets);
@@ -828,7 +878,81 @@ out_no_sizes:
 out_no_user_srf:
 	ttm_mem_global_free(vmw_mem_glob(dev_priv), size);
 out_unlock:
-	ttm_read_unlock(&vmaster->lock);
+	ttm_read_unlock(&dev_priv->reservation_sem);
+	return ret;
+}
+
+
+static int
+vmw_surface_handle_reference(struct vmw_private *dev_priv,
+			     struct drm_file *file_priv,
+			     uint32_t u_handle,
+			     enum drm_vmw_handle_type handle_type,
+			     struct ttm_base_object **base_p)
+{
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	struct vmw_user_surface *user_srf;
+	uint32_t handle;
+	struct ttm_base_object *base;
+	int ret;
+
+	if (handle_type == DRM_VMW_HANDLE_PRIME) {
+		ret = ttm_prime_fd_to_handle(tfile, u_handle, &handle);
+		if (unlikely(ret != 0))
+			return ret;
+	} else {
+		if (unlikely(drm_is_render_client(file_priv))) {
+			DRM_ERROR("Render client refused legacy "
+				  "surface reference.\n");
+			return -EACCES;
+		}
+		handle = u_handle;
+	}
+
+	ret = -EINVAL;
+	base = ttm_base_object_lookup_for_ref(dev_priv->tdev, handle);
+	if (unlikely(base == NULL)) {
+		DRM_ERROR("Could not find surface to reference.\n");
+		goto out_no_lookup;
+	}
+
+	if (unlikely(ttm_base_object_type(base) != VMW_RES_SURFACE)) {
+		DRM_ERROR("Referenced object is not a surface.\n");
+		goto out_bad_resource;
+	}
+
+	if (handle_type != DRM_VMW_HANDLE_PRIME) {
+		user_srf = container_of(base, struct vmw_user_surface,
+					prime.base);
+
+		/*
+		 * Make sure the surface creator has the same
+		 * authenticating master.
+		 */
+		if (drm_is_primary_client(file_priv) &&
+		    user_srf->master != file_priv->master) {
+			DRM_ERROR("Trying to reference surface outside of"
+				  " master domain.\n");
+			ret = -EACCES;
+			goto out_bad_resource;
+		}
+
+		ret = ttm_ref_object_add(tfile, base, TTM_REF_USAGE, NULL);
+		if (unlikely(ret != 0)) {
+			DRM_ERROR("Could not add a reference to a surface.\n");
+			goto out_bad_resource;
+		}
+	}
+
+	*base_p = base;
+	return 0;
+
+out_bad_resource:
+	ttm_base_object_unref(&base);
+out_no_lookup:
+	if (handle_type == DRM_VMW_HANDLE_PRIME)
+		(void) ttm_ref_object_base_unref(tfile, handle, TTM_REF_USAGE);
+
 	return ret;
 }
 
@@ -843,6 +967,7 @@ out_unlock:
 int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
 				struct drm_file *file_priv)
 {
+	struct vmw_private *dev_priv = vmw_priv(dev);
 	union drm_vmw_surface_reference_arg *arg =
 	    (union drm_vmw_surface_reference_arg *)data;
 	struct drm_vmw_surface_arg *req = &arg->req;
@@ -852,27 +977,16 @@ int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
 	struct vmw_user_surface *user_srf;
 	struct drm_vmw_size __user *user_sizes;
 	struct ttm_base_object *base;
-	int ret = -EINVAL;
-
-	base = ttm_base_object_lookup(tfile, req->sid);
-	if (unlikely(base == NULL)) {
-		DRM_ERROR("Could not find surface to reference.\n");
-		return -EINVAL;
-	}
+	int ret;
 
-	if (unlikely(ttm_base_object_type(base) != VMW_RES_SURFACE))
-		goto out_bad_resource;
+	ret = vmw_surface_handle_reference(dev_priv, file_priv, req->sid,
+					   req->handle_type, &base);
+	if (unlikely(ret != 0))
+		return ret;
 
 	user_srf = container_of(base, struct vmw_user_surface, prime.base);
 	srf = &user_srf->srf;
 
-	ret = ttm_ref_object_add(tfile, &user_srf->prime.base,
-				 TTM_REF_USAGE, NULL);
-	if (unlikely(ret != 0)) {
-		DRM_ERROR("Could not add a reference to a surface.\n");
-		goto out_no_reference;
-	}
-
 	rep->flags = srf->flags;
 	rep->format = srf->format;
 	memcpy(rep->mip_levels, srf->mip_levels, sizeof(srf->mip_levels));
@@ -880,15 +994,437 @@ int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
 	    rep->size_addr;
 
 	if (user_sizes)
-		ret = copy_to_user(user_sizes, srf->sizes,
-				   srf->num_sizes * sizeof(*srf->sizes));
+		ret = copy_to_user(user_sizes, &srf->base_size,
+				   sizeof(srf->base_size));
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("copy_to_user failed %p %u\n",
 			  user_sizes, srf->num_sizes);
+		ttm_ref_object_base_unref(tfile, base->hash.key, TTM_REF_USAGE);
 		ret = -EFAULT;
 	}
+
+	ttm_base_object_unref(&base);
+
+	return ret;
+}
+
+/**
+ * vmw_surface_define_encode - Encode a surface_define command.
+ *
+ * @srf: Pointer to a struct vmw_surface object.
+ * @cmd_space: Pointer to memory area in which the commands should be encoded.
+ */
+static int vmw_gb_surface_create(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct vmw_surface *srf = vmw_res_to_srf(res);
+	uint32_t cmd_len, submit_len;
+	int ret;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineGBSurface body;
+	} *cmd;
+
+	if (likely(res->id != -1))
+		return 0;
+
+	(void) vmw_3d_resource_inc(dev_priv, false);
+	ret = vmw_resource_alloc_id(res);
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Failed to allocate a surface id.\n");
+		goto out_no_id;
+	}
+
+	if (unlikely(res->id >= VMWGFX_NUM_GB_SURFACE)) {
+		ret = -EBUSY;
+		goto out_no_fifo;
+	}
+
+	cmd_len = sizeof(cmd->body);
+	submit_len = sizeof(*cmd);
+	cmd = vmw_fifo_reserve(dev_priv, submit_len);
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for surface "
+			  "creation.\n");
+		ret = -ENOMEM;
+		goto out_no_fifo;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DEFINE_GB_SURFACE;
+	cmd->header.size = cmd_len;
+	cmd->body.sid = srf->res.id;
+	cmd->body.surfaceFlags = srf->flags;
+	cmd->body.format = cpu_to_le32(srf->format);
+	cmd->body.numMipLevels = srf->mip_levels[0];
+	cmd->body.multisampleCount = srf->multisample_count;
+	cmd->body.autogenFilter = srf->autogen_filter;
+	cmd->body.size.width = srf->base_size.width;
+	cmd->body.size.height = srf->base_size.height;
+	cmd->body.size.depth = srf->base_size.depth;
+	vmw_fifo_commit(dev_priv, submit_len);
+
+	return 0;
+
+out_no_fifo:
+	vmw_resource_release_id(res);
+out_no_id:
+	vmw_3d_resource_dec(dev_priv, false);
+	return ret;
+}
+
+
+static int vmw_gb_surface_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBSurface body;
+	} *cmd1;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdUpdateGBSurface body;
+	} *cmd2;
+	uint32_t submit_size;
+	struct ttm_buffer_object *bo = val_buf->bo;
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	submit_size = sizeof(*cmd1) + (res->backup_dirty ? sizeof(*cmd2) : 0);
+
+	cmd1 = vmw_fifo_reserve(dev_priv, submit_size);
+	if (unlikely(cmd1 == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for surface "
+			  "binding.\n");
+		return -ENOMEM;
+	}
+
+	cmd1->header.id = SVGA_3D_CMD_BIND_GB_SURFACE;
+	cmd1->header.size = sizeof(cmd1->body);
+	cmd1->body.sid = res->id;
+	cmd1->body.mobid = bo->mem.start;
+	if (res->backup_dirty) {
+		cmd2 = (void *) &cmd1[1];
+		cmd2->header.id = SVGA_3D_CMD_UPDATE_GB_SURFACE;
+		cmd2->header.size = sizeof(cmd2->body);
+		cmd2->body.sid = res->id;
+		res->backup_dirty = false;
+	}
+	vmw_fifo_commit(dev_priv, submit_size);
+
+	return 0;
+}
+
+static int vmw_gb_surface_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct ttm_buffer_object *bo = val_buf->bo;
+	struct vmw_fence_obj *fence;
+
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdReadbackGBSurface body;
+	} *cmd1;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdInvalidateGBSurface body;
+	} *cmd2;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBSurface body;
+	} *cmd3;
+	uint32_t submit_size;
+	uint8_t *cmd;
+
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	submit_size = sizeof(*cmd3) + (readback ? sizeof(*cmd1) : sizeof(*cmd2));
+	cmd = vmw_fifo_reserve(dev_priv, submit_size);
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for surface "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+	if (readback) {
+		cmd1 = (void *) cmd;
+		cmd1->header.id = SVGA_3D_CMD_READBACK_GB_SURFACE;
+		cmd1->header.size = sizeof(cmd1->body);
+		cmd1->body.sid = res->id;
+		cmd3 = (void *) &cmd1[1];
+	} else {
+		cmd2 = (void *) cmd;
+		cmd2->header.id = SVGA_3D_CMD_INVALIDATE_GB_SURFACE;
+		cmd2->header.size = sizeof(cmd2->body);
+		cmd2->body.sid = res->id;
+		cmd3 = (void *) &cmd2[1];
+	}
+
+	cmd3->header.id = SVGA_3D_CMD_BIND_GB_SURFACE;
+	cmd3->header.size = sizeof(cmd3->body);
+	cmd3->body.sid = res->id;
+	cmd3->body.mobid = SVGA3D_INVALID_ID;
+
+	vmw_fifo_commit(dev_priv, submit_size);
+
+	/*
+	 * Create a fence object and fence the backup buffer.
+	 */
+
+	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
+					  &fence, NULL);
+
+	vmw_fence_single_bo(val_buf->bo, fence);
+
+	if (likely(fence != NULL))
+		vmw_fence_obj_unreference(&fence);
+
+	return 0;
+}
+
+static int vmw_gb_surface_destroy(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDestroyGBSurface body;
+	} *cmd;
+
+	if (likely(res->id == -1))
+		return 0;
+
+	mutex_lock(&dev_priv->binding_mutex);
+	vmw_context_binding_res_list_scrub(&res->binding_head);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for surface "
+			  "destruction.\n");
+		mutex_unlock(&dev_priv->binding_mutex);
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DESTROY_GB_SURFACE;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.sid = res->id;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	mutex_unlock(&dev_priv->binding_mutex);
+	vmw_resource_release_id(res);
+	vmw_3d_resource_dec(dev_priv, false);
+
+	return 0;
+}
+
+/**
+ * vmw_gb_surface_define_ioctl - Ioctl function implementing
+ *                               the user surface define functionality.
+ *
+ * @dev:            Pointer to a struct drm_device.
+ * @data:           Pointer to data copied from / to user-space.
+ * @file_priv:      Pointer to a drm file private structure.
+ */
+int vmw_gb_surface_define_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *file_priv)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	struct vmw_user_surface *user_srf;
+	struct vmw_surface *srf;
+	struct vmw_resource *res;
+	struct vmw_resource *tmp;
+	union drm_vmw_gb_surface_create_arg *arg =
+	    (union drm_vmw_gb_surface_create_arg *)data;
+	struct drm_vmw_gb_surface_create_req *req = &arg->req;
+	struct drm_vmw_gb_surface_create_rep *rep = &arg->rep;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	int ret;
+	uint32_t size;
+	const struct svga3d_surface_desc *desc;
+	uint32_t backup_handle;
+
+	if (unlikely(vmw_user_surface_size == 0))
+		vmw_user_surface_size = ttm_round_pot(sizeof(*user_srf)) +
+			128;
+
+	size = vmw_user_surface_size + 128;
+
+	desc = svga3dsurface_get_desc(req->format);
+	if (unlikely(desc->block_desc == SVGA3DBLOCKDESC_NONE)) {
+		DRM_ERROR("Invalid surface format for surface creation.\n");
+		return -EINVAL;
+	}
+
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_mem_global_alloc(vmw_mem_glob(dev_priv),
+				   size, false, true);
+	if (unlikely(ret != 0)) {
+		if (ret != -ERESTARTSYS)
+			DRM_ERROR("Out of graphics memory for surface"
+				  " creation.\n");
+		goto out_unlock;
+	}
+
+	user_srf = kzalloc(sizeof(*user_srf), GFP_KERNEL);
+	if (unlikely(user_srf == NULL)) {
+		ret = -ENOMEM;
+		goto out_no_user_srf;
+	}
+
+	srf = &user_srf->srf;
+	res = &srf->res;
+
+	srf->flags = req->svga3d_flags;
+	srf->format = req->format;
+	srf->scanout = req->drm_surface_flags & drm_vmw_surface_flag_scanout;
+	srf->mip_levels[0] = req->mip_levels;
+	srf->num_sizes = 1;
+	srf->sizes = NULL;
+	srf->offsets = NULL;
+	user_srf->size = size;
+	srf->base_size = req->base_size;
+	srf->autogen_filter = SVGA3D_TEX_FILTER_NONE;
+	srf->multisample_count = req->multisample_count;
+	res->backup_size = svga3dsurface_get_serialized_size
+	  (srf->format, srf->base_size, srf->mip_levels[0],
+	   srf->flags & SVGA3D_SURFACE_CUBEMAP);
+
+	user_srf->prime.base.shareable = false;
+	user_srf->prime.base.tfile = NULL;
+	if (drm_is_primary_client(file_priv))
+		user_srf->master = drm_master_get(file_priv->master);
+
+	/**
+	 * From this point, the generic resource management functions
+	 * destroy the object on failure.
+	 */
+
+	ret = vmw_surface_init(dev_priv, srf, vmw_user_surface_free);
+	if (unlikely(ret != 0))
+		goto out_unlock;
+
+	if (req->buffer_handle != SVGA3D_INVALID_ID) {
+		ret = vmw_user_dmabuf_lookup(tfile, req->buffer_handle,
+					     &res->backup);
+	} else if (req->drm_surface_flags &
+		   drm_vmw_surface_flag_create_buffer)
+		ret = vmw_user_dmabuf_alloc(dev_priv, tfile,
+					    res->backup_size,
+					    req->drm_surface_flags &
+					    drm_vmw_surface_flag_shareable,
+					    &backup_handle,
+					    &res->backup);
+
+	if (unlikely(ret != 0)) {
+		vmw_resource_unreference(&res);
+		goto out_unlock;
+	}
+
+	tmp = vmw_resource_reference(&srf->res);
+	ret = ttm_prime_object_init(tfile, res->backup_size, &user_srf->prime,
+				    req->drm_surface_flags &
+				    drm_vmw_surface_flag_shareable,
+				    VMW_RES_SURFACE,
+				    &vmw_user_surface_base_release, NULL);
+
+	if (unlikely(ret != 0)) {
+		vmw_resource_unreference(&tmp);
+		vmw_resource_unreference(&res);
+		goto out_unlock;
+	}
+
+	rep->handle = user_srf->prime.base.hash.key;
+	rep->backup_size = res->backup_size;
+	if (res->backup) {
+		rep->buffer_map_handle =
+			drm_vma_node_offset_addr(&res->backup->base.vma_node);
+		rep->buffer_size = res->backup->base.num_pages * PAGE_SIZE;
+		rep->buffer_handle = backup_handle;
+	} else {
+		rep->buffer_map_handle = 0;
+		rep->buffer_size = 0;
+		rep->buffer_handle = SVGA3D_INVALID_ID;
+	}
+
+	vmw_resource_unreference(&res);
+
+	ttm_read_unlock(&dev_priv->reservation_sem);
+	return 0;
+out_no_user_srf:
+	ttm_mem_global_free(vmw_mem_glob(dev_priv), size);
+out_unlock:
+	ttm_read_unlock(&dev_priv->reservation_sem);
+	return ret;
+}
+
+/**
+ * vmw_gb_surface_reference_ioctl - Ioctl function implementing
+ *                                  the user surface reference functionality.
+ *
+ * @dev:            Pointer to a struct drm_device.
+ * @data:           Pointer to data copied from / to user-space.
+ * @file_priv:      Pointer to a drm file private structure.
+ */
+int vmw_gb_surface_reference_ioctl(struct drm_device *dev, void *data,
+				   struct drm_file *file_priv)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	union drm_vmw_gb_surface_reference_arg *arg =
+	    (union drm_vmw_gb_surface_reference_arg *)data;
+	struct drm_vmw_surface_arg *req = &arg->req;
+	struct drm_vmw_gb_surface_ref_rep *rep = &arg->rep;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	struct vmw_surface *srf;
+	struct vmw_user_surface *user_srf;
+	struct ttm_base_object *base;
+	uint32_t backup_handle;
+	int ret = -EINVAL;
+
+	ret = vmw_surface_handle_reference(dev_priv, file_priv, req->sid,
+					   req->handle_type, &base);
+	if (unlikely(ret != 0))
+		return ret;
+
+	user_srf = container_of(base, struct vmw_user_surface, prime.base);
+	srf = &user_srf->srf;
+	if (srf->res.backup == NULL) {
+		DRM_ERROR("Shared GB surface is missing a backup buffer.\n");
+		goto out_bad_resource;
+	}
+
+	mutex_lock(&dev_priv->cmdbuf_mutex); /* Protect res->backup */
+	ret = vmw_user_dmabuf_reference(tfile, srf->res.backup,
+					&backup_handle);
+	mutex_unlock(&dev_priv->cmdbuf_mutex);
+
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Could not add a reference to a GB surface "
+			  "backup buffer.\n");
+		(void) ttm_ref_object_base_unref(tfile, base->hash.key,
+						 TTM_REF_USAGE);
+		goto out_bad_resource;
+	}
+
+	rep->creq.svga3d_flags = srf->flags;
+	rep->creq.format = srf->format;
+	rep->creq.mip_levels = srf->mip_levels[0];
+	rep->creq.drm_surface_flags = 0;
+	rep->creq.multisample_count = srf->multisample_count;
+	rep->creq.autogen_filter = srf->autogen_filter;
+	rep->creq.buffer_handle = backup_handle;
+	rep->creq.base_size = srf->base_size;
+	rep->crep.handle = user_srf->prime.base.hash.key;
+	rep->crep.backup_size = srf->res.backup_size;
+	rep->crep.buffer_handle = backup_handle;
+	rep->crep.buffer_map_handle =
+		drm_vma_node_offset_addr(&srf->res.backup->base.vma_node);
+	rep->crep.buffer_size = srf->res.backup->base.num_pages * PAGE_SIZE;
+
 out_bad_resource:
-out_no_reference:
 	ttm_base_object_unref(&base);
 
 	return ret;