Merge tag 'fbdev-main-3.15' of git://git.kernel.org/pub/scm/linux/kernel/git/tomba/linux

Pull fbdev changes from Tomi Valkeinen:
 "Various fbdev fixes and improvements, but nothing big"

* tag 'fbdev-main-3.15' of git://git.kernel.org/pub/scm/linux/kernel/git/tomba/linux: (38 commits)
  fbdev: Make the switch from generic to native driver less alarming
  Video: atmel: avoid the id of fix screen info is overwritten
  video: imxfb: Add DT default contrast control register property.
  video: atmel_lcdfb: ensure the hardware is initialized with the correct mode
  fbdev: vesafb: add dev->remove() callback
  fbdev: efifb: add dev->remove() callback
  video: pxa3xx-gcu: switch to devres functions
  video: pxa3xx-gcu: provide an empty .open call
  video: pxa3xx-gcu: pass around struct device *
  video: pxa3xx-gcu: rename some symbols
  sisfb: fix 1280x720 resolution support
  video: fbdev: uvesafb: Remove impossible code path in uvesafb_init_info
  video: fbdev: uvesafb: Remove redundant NULL check in uvesafb_remove
  fbdev: FB_OPENCORES should depend on HAS_DMA
  OMAPDSS: convert pixel clock to common videomode style
  OMAPDSS: Remove unused get_context_loss_count support
  OMAPDSS: use DISPC register to detect context loss
  video: da8xx-fb: Use "SIMPLE_DEV_PM_OPS" macro
  video: imxfb: Convert to SIMPLE_DEV_PM_OPS
  video: imxfb: Resolve mismatch between backlight/contrast
  ...

52 files changed, 695 insertions, 872 deletions

diff --git a/Documentation/devicetree/bindings/video/fsl,imx-fb.txt b/Documentation/devicetree/bindings/video/fsl,imx-fb.txt
index 46da08db186..0329f60d431 100644
--- a/Documentation/devicetree/bindings/video/fsl,imx-fb.txt
+++ b/Documentation/devicetree/bindings/video/fsl,imx-fb.txt
@@ -15,8 +15,12 @@ Required nodes:
 	- fsl,pcr: LCDC PCR value
 
 Optional properties:
+- lcd-supply: Regulator for LCD supply voltage.
 - fsl,dmacr: DMA Control Register value. This is optional. By default, the
 	register is not modified as recommended by the datasheet.
+- fsl,lpccr: Contrast Control Register value. This property provides the
+	default value for the contrast control register.
+	If that property is ommited, the register is zeroed.
 - fsl,lscr1: LCDC Sharp Configuration Register value.
 
 Example:
diff --git a/arch/arm/mach-imx/mach-mx27ads.c b/arch/arm/mach-imx/mach-mx27ads.c
index 9821b824dca..a7a4a9c6761 100644
--- a/arch/arm/mach-imx/mach-mx27ads.c
+++ b/arch/arm/mach-imx/mach-mx27ads.c
@@ -21,6 +21,10 @@
 #include <linux/mtd/physmap.h>
 #include <linux/i2c.h>
 #include <linux/irq.h>
+
+#include <linux/regulator/fixed.h>
+#include <linux/regulator/machine.h>
+
 #include <asm/mach-types.h>
 #include <asm/mach/arch.h>
 #include <asm/mach/time.h>
@@ -195,14 +199,58 @@ static const struct imxi2c_platform_data mx27ads_i2c1_data __initconst = {
 static struct i2c_board_info mx27ads_i2c_devices[] = {
 };
 
-void lcd_power(int on)
+static void vgpio_set(struct gpio_chip *chip, unsigned offset, int value)
 {
-	if (on)
+	if (value)
 		__raw_writew(PBC_BCTRL1_LCDON, PBC_BCTRL1_SET_REG);
 	else
 		__raw_writew(PBC_BCTRL1_LCDON, PBC_BCTRL1_CLEAR_REG);
 }
 
+static int vgpio_dir_out(struct gpio_chip *chip, unsigned offset, int value)
+{
+	return 0;
+}
+
+#define MX27ADS_LCD_GPIO	(6 * 32)
+
+static struct regulator_consumer_supply mx27ads_lcd_regulator_consumer =
+	REGULATOR_SUPPLY("lcd", "imx-fb.0");
+
+static struct regulator_init_data mx27ads_lcd_regulator_init_data = {
+	.constraints	= {
+		.valid_ops_mask	= REGULATOR_CHANGE_STATUS,
+},
+	.consumer_supplies	= &mx27ads_lcd_regulator_consumer,
+	.num_consumer_supplies	= 1,
+};
+
+static struct fixed_voltage_config mx27ads_lcd_regulator_pdata = {
+	.supply_name	= "LCD",
+	.microvolts	= 3300000,
+	.gpio		= MX27ADS_LCD_GPIO,
+	.init_data	= &mx27ads_lcd_regulator_init_data,
+};
+
+static void __init mx27ads_regulator_init(void)
+{
+	struct gpio_chip *vchip;
+
+	vchip = kzalloc(sizeof(*vchip), GFP_KERNEL);
+	vchip->owner		= THIS_MODULE;
+	vchip->label		= "LCD";
+	vchip->base		= MX27ADS_LCD_GPIO;
+	vchip->ngpio		= 1;
+	vchip->direction_output	= vgpio_dir_out;
+	vchip->set		= vgpio_set;
+	gpiochip_add(vchip);
+
+	platform_device_register_data(&platform_bus, "reg-fixed-voltage",
+				      PLATFORM_DEVID_AUTO,
+				      &mx27ads_lcd_regulator_pdata,
+				      sizeof(mx27ads_lcd_regulator_pdata));
+}
+
 static struct imx_fb_videomode mx27ads_modes[] = {
 	{
 		.mode = {
@@ -239,8 +287,6 @@ static const struct imx_fb_platform_data mx27ads_fb_data __initconst = {
 	.pwmr		= 0x00A903FF,
 	.lscr1		= 0x00120300,
 	.dmacr		= 0x00020010,
-
-	.lcd_power	= lcd_power,
 };
 
 static int mx27ads_sdhc1_init(struct device *dev, irq_handler_t detect_irq,
@@ -304,6 +350,7 @@ static void __init mx27ads_board_init(void)
 	i2c_register_board_info(1, mx27ads_i2c_devices,
 				ARRAY_SIZE(mx27ads_i2c_devices));
 	imx27_add_imx_i2c(1, &mx27ads_i2c1_data);
+	mx27ads_regulator_init();
 	imx27_add_imx_fb(&mx27ads_fb_data);
 	imx27_add_mxc_mmc(0, &sdhc1_pdata);
 	imx27_add_mxc_mmc(1, &sdhc2_pdata);
diff --git a/arch/arm/mach-omap2/display.c b/arch/arm/mach-omap2/display.c
index 4cf165502b3..93ebb4007ea 100644
--- a/arch/arm/mach-omap2/display.c
+++ b/arch/arm/mach-omap2/display.c
@@ -301,7 +301,6 @@ int __init omap_display_init(struct omap_dss_board_info *board_data)
 	board_data->version = ver;
 	board_data->dsi_enable_pads = omap_dsi_enable_pads;
 	board_data->dsi_disable_pads = omap_dsi_disable_pads;
-	board_data->get_context_loss_count = omap_pm_get_dev_context_loss_count;
 	board_data->set_min_bus_tput = omap_dss_set_min_bus_tput;
 
 	omap_display_device.dev.platform_data = board_data;
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/video/Kconfig b/drivers/video/Kconfig
index b4b209ce802..6c793bc683d 100644
--- a/drivers/video/Kconfig
+++ b/drivers/video/Kconfig
@@ -967,7 +967,7 @@ config FB_PVR2
 
 config FB_OPENCORES
 	tristate "OpenCores VGA/LCD core 2.0 framebuffer support"
-	depends on FB
+	depends on FB && HAS_DMA
 	select FB_CFB_FILLRECT
 	select FB_CFB_COPYAREA
 	select FB_CFB_IMAGEBLIT
diff --git a/drivers/video/atmel_lcdfb.c b/drivers/video/atmel_lcdfb.c
index cd961622f9c..e683b6ef959 100644
--- a/drivers/video/atmel_lcdfb.c
+++ b/drivers/video/atmel_lcdfb.c
@@ -1190,12 +1190,12 @@ static int __init atmel_lcdfb_probe(struct platform_device *pdev)
 	if (!sinfo->config)
 		goto free_info;
 
-	strcpy(info->fix.id, sinfo->pdev->name);
 	info->flags = ATMEL_LCDFB_FBINFO_DEFAULT;
 	info->pseudo_palette = sinfo->pseudo_palette;
 	info->fbops = &atmel_lcdfb_ops;
 
 	info->fix = atmel_lcdfb_fix;
+	strcpy(info->fix.id, sinfo->pdev->name);
 
 	/* Enable LCDC Clocks */
 	sinfo->bus_clk = clk_get(dev, "hclk");
@@ -1298,6 +1298,12 @@ static int __init atmel_lcdfb_probe(struct platform_device *pdev)
 		goto unregister_irqs;
 	}
 
+	ret = atmel_lcdfb_set_par(info);
+	if (ret < 0) {
+		dev_err(dev, "set par failed: %d\n", ret);
+		goto unregister_irqs;
+	}
+
 	dev_set_drvdata(dev, info);
 
 	/*
diff --git a/drivers/video/aty/atyfb_base.c b/drivers/video/aty/atyfb_base.c
index 28fafbf864a..c3d0074a32d 100644
--- a/drivers/video/aty/atyfb_base.c
+++ b/drivers/video/aty/atyfb_base.c
@@ -862,8 +862,8 @@ static int aty_var_to_crtc(const struct fb_info *info,
 	h_sync_pol = sync & FB_SYNC_HOR_HIGH_ACT ? 0 : 1;
 	v_sync_pol = sync & FB_SYNC_VERT_HIGH_ACT ? 0 : 1;
 
-	if ((xres > 1600) || (yres > 1200)) {
-		FAIL("MACH64 chips are designed for max 1600x1200\n"
+	if ((xres > 1920) || (yres > 1200)) {
+		FAIL("MACH64 chips are designed for max 1920x1200\n"
 		     "select another resolution.");
 	}
 	h_sync_strt = h_disp + var->right_margin;
@@ -2653,7 +2653,8 @@ static int aty_init(struct fb_info *info)
 		      FBINFO_HWACCEL_IMAGEBLIT |
 		      FBINFO_HWACCEL_FILLRECT  |
 		      FBINFO_HWACCEL_COPYAREA  |
-		      FBINFO_HWACCEL_YPAN;
+		      FBINFO_HWACCEL_YPAN      |
+		      FBINFO_READS_FAST;
 
 #ifdef CONFIG_PMAC_BACKLIGHT
 	if (M64_HAS(G3_PB_1_1) && of_machine_is_compatible("PowerBook1,1")) {
diff --git a/drivers/video/aty/mach64_accel.c b/drivers/video/aty/mach64_accel.c
index e45833ce975..182bd680141 100644
--- a/drivers/video/aty/mach64_accel.c
+++ b/drivers/video/aty/mach64_accel.c
@@ -4,6 +4,7 @@
  */
 
 #include <linux/delay.h>
+#include <asm/unaligned.h>
 #include <linux/fb.h>
 #include <video/mach64.h>
 #include "atyfb.h"
@@ -419,7 +420,7 @@ void atyfb_imageblit(struct fb_info *info, const struct fb_image *image)
 		u32 *pbitmap, dwords = (src_bytes + 3) / 4;
 		for (pbitmap = (u32*)(image->data); dwords; dwords--, pbitmap++) {
 			wait_for_fifo(1, par);
-			aty_st_le32(HOST_DATA0, le32_to_cpup(pbitmap), par);
+			aty_st_le32(HOST_DATA0, get_unaligned_le32(pbitmap), par);
 		}
 	}
 
diff --git a/drivers/video/aty/mach64_cursor.c b/drivers/video/aty/mach64_cursor.c
index 95ec042ddbf..0fe02e22d9a 100644
--- a/drivers/video/aty/mach64_cursor.c
+++ b/drivers/video/aty/mach64_cursor.c
@@ -5,6 +5,7 @@
 #include <linux/fb.h>
 #include <linux/init.h>
 #include <linux/string.h>
+#include "../fb_draw.h"
 
 #include <asm/io.h>
 
@@ -157,24 +158,33 @@ static int atyfb_cursor(struct fb_info *info, struct fb_cursor *cursor)
 
 	    for (i = 0; i < height; i++) {
 		for (j = 0; j < width; j++) {
+			u16 l = 0xaaaa;
 			b = *src++;
 			m = *msk++;
 			switch (cursor->rop) {
 			case ROP_XOR:
 			    // Upper 4 bits of mask data
-			    fb_writeb(cursor_bits_lookup[(b ^ m) >> 4], dst++);
+			    l = cursor_bits_lookup[(b ^ m) >> 4] |
 			    // Lower 4 bits of mask
-			    fb_writeb(cursor_bits_lookup[(b ^ m) & 0x0f],
-				      dst++);
+				    (cursor_bits_lookup[(b ^ m) & 0x0f] << 8);
 			    break;
 			case ROP_COPY:
 			    // Upper 4 bits of mask data
-			    fb_writeb(cursor_bits_lookup[(b & m) >> 4], dst++);
+			    l = cursor_bits_lookup[(b & m) >> 4] |
 			    // Lower 4 bits of mask
-			    fb_writeb(cursor_bits_lookup[(b & m) & 0x0f],
-				      dst++);
+				    (cursor_bits_lookup[(b & m) & 0x0f] << 8);
 			    break;
 			}
+			/*
+			 * If cursor size is not a multiple of 8 characters
+			 * we must pad it with transparent pattern (0xaaaa).
+			 */
+			if ((j + 1) * 8 > cursor->image.width) {
+				l = comp(l, 0xaaaa,
+				    (1 << ((cursor->image.width & 7) * 2)) - 1);
+			}
+			fb_writeb(l & 0xff, dst++);
+			fb_writeb(l >> 8, dst++);
 		}
 		dst += offset;
 	    }
diff --git a/drivers/video/cfbcopyarea.c b/drivers/video/cfbcopyarea.c
index bb5a96b1645..bcb57235fcc 100644
--- a/drivers/video/cfbcopyarea.c
+++ b/drivers/video/cfbcopyarea.c
@@ -43,13 +43,22 @@
      */
 
 static void
-bitcpy(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
-		const unsigned long __iomem *src, int src_idx, int bits,
+bitcpy(struct fb_info *p, unsigned long __iomem *dst, unsigned dst_idx,
+		const unsigned long __iomem *src, unsigned src_idx, int bits,
 		unsigned n, u32 bswapmask)
 {
 	unsigned long first, last;
 	int const shift = dst_idx-src_idx;
-	int left, right;
+
+#if 0
+	/*
+	 * If you suspect bug in this function, compare it with this simple
+	 * memmove implementation.
+	 */
+	fb_memmove((char *)dst + ((dst_idx & (bits - 1))) / 8,
+		   (char *)src + ((src_idx & (bits - 1))) / 8, n / 8);
+	return;
+#endif
 
 	first = fb_shifted_pixels_mask_long(p, dst_idx, bswapmask);
 	last = ~fb_shifted_pixels_mask_long(p, (dst_idx+n) % bits, bswapmask);
@@ -98,9 +107,8 @@ bitcpy(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
 		unsigned long d0, d1;
 		int m;
 
-		right = shift & (bits - 1);
-		left = -shift & (bits - 1);
-		bswapmask &= shift;
+		int const left = shift & (bits - 1);
+		int const right = -shift & (bits - 1);
 
 		if (dst_idx+n <= bits) {
 			// Single destination word
@@ -110,15 +118,15 @@ bitcpy(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
 			d0 = fb_rev_pixels_in_long(d0, bswapmask);
 			if (shift > 0) {
 				// Single source word
-				d0 >>= right;
+				d0 <<= left;
 			} else if (src_idx+n <= bits) {
 				// Single source word
-				d0 <<= left;
+				d0 >>= right;
 			} else {
 				// 2 source words
 				d1 = FB_READL(src + 1);
 				d1 = fb_rev_pixels_in_long(d1, bswapmask);
-				d0 = d0<<left | d1>>right;
+				d0 = d0 >> right | d1 << left;
 			}
 			d0 = fb_rev_pixels_in_long(d0, bswapmask);
 			FB_WRITEL(comp(d0, FB_READL(dst), first), dst);
@@ -135,60 +143,59 @@ bitcpy(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
 			if (shift > 0) {
 				// Single source word
 				d1 = d0;
-				d0 >>= right;
-				dst++;
+				d0 <<= left;
 				n -= bits - dst_idx;
 			} else {
 				// 2 source words
 				d1 = FB_READL(src++);
 				d1 = fb_rev_pixels_in_long(d1, bswapmask);
 
-				d0 = d0<<left | d1>>right;
-				dst++;
+				d0 = d0 >> right | d1 << left;
 				n -= bits - dst_idx;
 			}
 			d0 = fb_rev_pixels_in_long(d0, bswapmask);
 			FB_WRITEL(comp(d0, FB_READL(dst), first), dst);
 			d0 = d1;
+			dst++;
 
 			// Main chunk
 			m = n % bits;
 			n /= bits;
 			while ((n >= 4) && !bswapmask) {
 				d1 = FB_READL(src++);
-				FB_WRITEL(d0 << left | d1 >> right, dst++);
+				FB_WRITEL(d0 >> right | d1 << left, dst++);
 				d0 = d1;
 				d1 = FB_READL(src++);
-				FB_WRITEL(d0 << left | d1 >> right, dst++);
+				FB_WRITEL(d0 >> right | d1 << left, dst++);
 				d0 = d1;
 				d1 = FB_READL(src++);
-				FB_WRITEL(d0 << left | d1 >> right, dst++);
+				FB_WRITEL(d0 >> right | d1 << left, dst++);
 				d0 = d1;
 				d1 = FB_READL(src++);
-				FB_WRITEL(d0 << left | d1 >> right, dst++);
+				FB_WRITEL(d0 >> right | d1 << left, dst++);
 				d0 = d1;
 				n -= 4;
 			}
 			while (n--) {
 				d1 = FB_READL(src++);
 				d1 = fb_rev_pixels_in_long(d1, bswapmask);
-				d0 = d0 << left | d1 >> right;
+				d0 = d0 >> right | d1 << left;
 				d0 = fb_rev_pixels_in_long(d0, bswapmask);
 				FB_WRITEL(d0, dst++);
 				d0 = d1;
 			}
 
 			// Trailing bits
-			if (last) {
-				if (m <= right) {
+			if (m) {
+				if (m <= bits - right) {
 					// Single source word
-					d0 <<= left;
+					d0 >>= right;
 				} else {
 					// 2 source words
 					d1 = FB_READL(src);
 					d1 = fb_rev_pixels_in_long(d1,
 								bswapmask);
-					d0 = d0<<left | d1>>right;
+					d0 = d0 >> right | d1 << left;
 				}
 				d0 = fb_rev_pixels_in_long(d0, bswapmask);
 				FB_WRITEL(comp(d0, FB_READL(dst), last), dst);
@@ -202,43 +209,46 @@ bitcpy(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
      */
 
 static void
-bitcpy_rev(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
-		const unsigned long __iomem *src, int src_idx, int bits,
+bitcpy_rev(struct fb_info *p, unsigned long __iomem *dst, unsigned dst_idx,
+		const unsigned long __iomem *src, unsigned src_idx, int bits,
 		unsigned n, u32 bswapmask)
 {
 	unsigned long first, last;
 	int shift;
 
-	dst += (n-1)/bits;
-	src += (n-1)/bits;
-	if ((n-1) % bits) {
-		dst_idx += (n-1) % bits;
-		dst += dst_idx >> (ffs(bits) - 1);
-		dst_idx &= bits - 1;
-		src_idx += (n-1) % bits;
-		src += src_idx >> (ffs(bits) - 1);
-		src_idx &= bits - 1;
-	}
+#if 0
+	/*
+	 * If you suspect bug in this function, compare it with this simple
+	 * memmove implementation.
+	 */
+	fb_memmove((char *)dst + ((dst_idx & (bits - 1))) / 8,
+		   (char *)src + ((src_idx & (bits - 1))) / 8, n / 8);
+	return;
+#endif
+
+	dst += (dst_idx + n - 1) / bits;
+	src += (src_idx + n - 1) / bits;
+	dst_idx = (dst_idx + n - 1) % bits;
+	src_idx = (src_idx + n - 1) % bits;
 
 	shift = dst_idx-src_idx;
 
-	first = fb_shifted_pixels_mask_long(p, bits - 1 - dst_idx, bswapmask);
-	last = ~fb_shifted_pixels_mask_long(p, bits - 1 - ((dst_idx-n) % bits),
-					    bswapmask);
+	first = ~fb_shifted_pixels_mask_long(p, (dst_idx + 1) % bits, bswapmask);
+	last = fb_shifted_pixels_mask_long(p, (bits + dst_idx + 1 - n) % bits, bswapmask);
 
 	if (!shift) {
 		// Same alignment for source and dest
 
 		if ((unsigned long)dst_idx+1 >= n) {
 			// Single word
-			if (last)
-				first &= last;
-			FB_WRITEL( comp( FB_READL(src), FB_READL(dst), first), dst);
+			if (first)
+				last &= first;
+			FB_WRITEL( comp( FB_READL(src), FB_READL(dst), last), dst);
 		} else {
 			// Multiple destination words
 
 			// Leading bits
-			if (first != ~0UL) {
+			if (first) {
 				FB_WRITEL( comp( FB_READL(src), FB_READL(dst), first), dst);
 				dst--;
 				src--;
@@ -262,7 +272,7 @@ bitcpy_rev(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
 				FB_WRITEL(FB_READL(src--), dst--);
 
 			// Trailing bits
-			if (last)
+			if (last != -1UL)
 				FB_WRITEL( comp( FB_READL(src), FB_READL(dst), last), dst);
 		}
 	} else {
@@ -270,29 +280,28 @@ bitcpy_rev(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
 		unsigned long d0, d1;
 		int m;
 
-		int const left = -shift & (bits-1);
-		int const right = shift & (bits-1);
-		bswapmask &= shift;
+		int const left = shift & (bits-1);
+		int const right = -shift & (bits-1);
 
 		if ((unsigned long)dst_idx+1 >= n) {
 			// Single destination word
-			if (last)
-				first &= last;
+			if (first)
+				last &= first;
 			d0 = FB_READL(src);
 			if (shift < 0) {
 				// Single source word
-				d0 <<= left;
+				d0 >>= right;
 			} else if (1+(unsigned long)src_idx >= n) {
 				// Single source word
-				d0 >>= right;
+				d0 <<= left;
 			} else {
 				// 2 source words
 				d1 = FB_READL(src - 1);
 				d1 = fb_rev_pixels_in_long(d1, bswapmask);
-				d0 = d0>>right | d1<<left;
+				d0 = d0 << left | d1 >> right;
 			}
 			d0 = fb_rev_pixels_in_long(d0, bswapmask);
-			FB_WRITEL(comp(d0, FB_READL(dst), first), dst);
+			FB_WRITEL(comp(d0, FB_READL(dst), last), dst);
 		} else {
 			// Multiple destination words
 			/** We must always remember the last value read, because in case
@@ -307,12 +316,12 @@ bitcpy_rev(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
 			if (shift < 0) {
 				// Single source word
 				d1 = d0;
-				d0 <<= left;
+				d0 >>= right;
 			} else {
 				// 2 source words
 				d1 = FB_READL(src--);
 				d1 = fb_rev_pixels_in_long(d1, bswapmask);
-				d0 = d0>>right | d1<<left;
+				d0 = d0 << left | d1 >> right;
 			}
 			d0 = fb_rev_pixels_in_long(d0, bswapmask);
 			FB_WRITEL(comp(d0, FB_READL(dst), first), dst);
@@ -325,39 +334,39 @@ bitcpy_rev(struct fb_info *p, unsigned long __iomem *dst, int dst_idx,
 			n /= bits;
 			while ((n >= 4) && !bswapmask) {
 				d1 = FB_READL(src--);
-				FB_WRITEL(d0 >> right | d1 << left, dst--);
+				FB_WRITEL(d0 << left | d1 >> right, dst--);
 				d0 = d1;
 				d1 = FB_READL(src--);
-				FB_WRITEL(d0 >> right | d1 << left, dst--);
+				FB_WRITEL(d0 << left | d1 >> right, dst--);
 				d0 = d1;
 				d1 = FB_READL(src--);
-				FB_WRITEL(d0 >> right | d1 << left, dst--);
+				FB_WRITEL(d0 << left | d1 >> right, dst--);
 				d0 = d1;
 				d1 = FB_READL(src--);
-				FB_WRITEL(d0 >> right | d1 << left, dst--);
+				FB_WRITEL(d0 << left | d1 >> right, dst--);
 				d0 = d1;
 				n -= 4;
 			}
 			while (n--) {
 				d1 = FB_READL(src--);
 				d1 = fb_rev_pixels_in_long(d1, bswapmask);
-				d0 = d0 >> right | d1 << left;
+				d0 = d0 << left | d1 >> right;
 				d0 = fb_rev_pixels_in_long(d0, bswapmask);
 				FB_WRITEL(d0, dst--);
 				d0 = d1;
 			}
 
 			// Trailing bits
-			if (last) {
-				if (m <= left) {
+			if (m) {
+				if (m <= bits - left) {
 					// Single source word
-					d0 >>= right;
+					d0 <<= left;
 				} else {
 					// 2 source words
 					d1 = FB_READL(src);
 					d1 = fb_rev_pixels_in_long(d1,
 								bswapmask);
-					d0 = d0>>right | d1<<left;
+					d0 = d0 << left | d1 >> right;
 				}
 				d0 = fb_rev_pixels_in_long(d0, bswapmask);
 				FB_WRITEL(comp(d0, FB_READL(dst), last), dst);
@@ -371,9 +380,9 @@ void cfb_copyarea(struct fb_info *p, const struct fb_copyarea *area)
 	u32 dx = area->dx, dy = area->dy, sx = area->sx, sy = area->sy;
 	u32 height = area->height, width = area->width;
 	unsigned long const bits_per_line = p->fix.line_length*8u;
-	unsigned long __iomem *dst = NULL, *src = NULL;
+	unsigned long __iomem *base = NULL;
 	int bits = BITS_PER_LONG, bytes = bits >> 3;
-	int dst_idx = 0, src_idx = 0, rev_copy = 0;
+	unsigned dst_idx = 0, src_idx = 0, rev_copy = 0;
 	u32 bswapmask = fb_compute_bswapmask(p);
 
 	if (p->state != FBINFO_STATE_RUNNING)
@@ -389,7 +398,7 @@ void cfb_copyarea(struct fb_info *p, const struct fb_copyarea *area)
 
 	// split the base of the framebuffer into a long-aligned address and the
 	// index of the first bit
-	dst = src = (unsigned long __iomem *)((unsigned long)p->screen_base & ~(