diff options
Diffstat (limited to 'drivers/gpu/drm/i915/intel_pm.c')
| -rw-r--r-- | drivers/gpu/drm/i915/intel_pm.c | 5020 |
1 files changed, 3725 insertions, 1295 deletions
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 72f41aaa71f..ee72807069e 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -30,8 +30,30 @@ #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> -#define FORCEWAKE_ACK_TIMEOUT_MS 2 +/** + * RC6 is a special power stage which allows the GPU to enter an very + * low-voltage mode when idle, using down to 0V while at this stage. This + * stage is entered automatically when the GPU is idle when RC6 support is + * enabled, and as soon as new workload arises GPU wakes up automatically as well. + * + * There are different RC6 modes available in Intel GPU, which differentiate + * among each other with the latency required to enter and leave RC6 and + * voltage consumed by the GPU in different states. + * + * The combination of the following flags define which states GPU is allowed + * to enter, while RC6 is the normal RC6 state, RC6p is the deep RC6, and + * RC6pp is deepest RC6. Their support by hardware varies according to the + * GPU, BIOS, chipset and platform. RC6 is usually the safest one and the one + * which brings the most power savings; deeper states save more power, but + * require higher latency to switch to and wake up. + */ +#define INTEL_RC6_ENABLE (1<<0) +#define INTEL_RC6p_ENABLE (1<<1) +#define INTEL_RC6pp_ENABLE (1<<2) /* FBC, or Frame Buffer Compression, is a technique employed to compress the * framebuffer contents in-memory, aiming at reducing the required bandwidth @@ -66,47 +88,54 @@ 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->cfb_size / FBC_LL_SIZE; + 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 %d, ", - cfb_pitch, crtc->y, intel_crtc->plane); + DRM_DEBUG_KMS("enabled FBC, pitch %d, yoff %d, plane %c\n", + cfb_pitch, crtc->y, plane_name(intel_crtc->plane)); } static bool i8xx_fbc_enabled(struct drm_device *dev) @@ -116,31 +145,29 @@ 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 %d\n", intel_crtc->plane); + DRM_DEBUG_KMS("enabled fbc on plane %c\n", plane_name(intel_crtc->plane)); } static void g4x_disable_fbc(struct drm_device *dev) @@ -171,7 +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; @@ -182,34 +213,31 @@ 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, obj->gtt_offset | ILK_FBC_RT_VALID); + I915_WRITE(ILK_FBC_RT_BASE, i915_gem_obj_ggtt_offset(obj) | ILK_FBC_RT_VALID); /* enable it... */ I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN); @@ -220,7 +248,7 @@ static void ironlake_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)); } static void ironlake_disable_fbc(struct drm_device *dev) @@ -245,6 +273,46 @@ 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) +{ + struct drm_device *dev = crtc->dev; + struct drm_i915_private *dev_priv = dev->dev_private; + 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; + + 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 | DPFC_CTL_EN); + + if (IS_IVYBRIDGE(dev)) { + /* WaFbcAsynchFlipDisableFbcQueue:ivb */ + I915_WRITE(ILK_DISPLAY_CHICKEN1, + I915_READ(ILK_DISPLAY_CHICKEN1) | + ILK_FBCQ_DIS); + } else { + /* 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, + SNB_CPU_FENCE_ENABLE | obj->fence_reg); + I915_WRITE(DPFC_CPU_FENCE_OFFSET, crtc->y); + + sandybridge_blit_fbc_update(dev); + + DRM_DEBUG_KMS("enabled fbc on plane %c\n", plane_name(intel_crtc->plane)); +} + bool intel_fbc_enabled(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -264,20 +332,19 @@ static void intel_fbc_work_fn(struct work_struct *__work) struct drm_i915_private *dev_priv = dev->dev_private; mutex_lock(&dev->struct_mutex); - if (work == dev_priv->fbc_work) { + if (work == dev_priv->fbc.fbc_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->cfb_plane = to_intel_crtc(work->crtc)->plane; - dev_priv->cfb_fb = work->crtc->fb->base.id; - dev_priv->cfb_y = work->crtc->y; + dev_priv->fbc.plane = to_intel_crtc(work->crtc)->plane; + dev_priv->fbc.fb_id = work->crtc->primary->fb->base.id; + dev_priv->fbc.y = work->crtc->y; } - dev_priv->fbc_work = NULL; + dev_priv->fbc.fbc_work = NULL; } mutex_unlock(&dev->struct_mutex); @@ -286,28 +353,28 @@ static void intel_fbc_work_fn(struct work_struct *__work) static void intel_cancel_fbc_work(struct drm_i915_private *dev_priv) { - if (dev_priv->fbc_work == NULL) + if (dev_priv->fbc.fbc_work == NULL) return; DRM_DEBUG_KMS("cancelling pending FBC enable\n"); /* Synchronisation is provided by struct_mutex and checking of - * dev_priv->fbc_work, so we can perform the cancellation + * dev_priv->fbc.fbc_work, so we can perform the cancellation * entirely asynchronously. */ - if (cancel_delayed_work(&dev_priv->fbc_work->work)) + if (cancel_delayed_work(&dev_priv->fbc.fbc_work->work)) /* tasklet was killed before being run, clean up */ - kfree(dev_priv->fbc_work); + kfree(dev_priv->fbc.fbc_work); /* Mark the work as no longer wanted so that if it does * wake-up (because the work was already running and waiting * for our mutex), it will discover that is no longer * necessary to run. */ - dev_priv->fbc_work = NULL; + dev_priv->fbc.fbc_work = NULL; } -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; @@ -318,20 +385,18 @@ void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval) intel_cancel_fbc_work(dev_priv); - work = kzalloc(sizeof *work, GFP_KERNEL); + work = kzalloc(sizeof(*work), GFP_KERNEL); if (work == NULL) { - dev_priv->display.enable_fbc(crtc, interval); + DRM_ERROR("Failed to allocate FBC work structure\n"); + 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_work = work; - - DRM_DEBUG_KMS("scheduling delayed FBC enable\n"); + dev_priv->fbc.fbc_work = work; /* Delay the actual enabling to let pageflipping cease and the * display to settle before starting the compression. Note that @@ -343,6 +408,8 @@ void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval) * following the termination of the page-flipping sequence * and indeed performing the enable as a co-routine and not * waiting synchronously upon the vblank. + * + * WaFbcWaitForVBlankBeforeEnable:ilk,snb */ schedule_delayed_work(&work->work, msecs_to_jiffies(50)); } @@ -357,7 +424,17 @@ void intel_disable_fbc(struct drm_device *dev) return; dev_priv->display.disable_fbc(dev); - dev_priv->cfb_plane = -1; + dev_priv->fbc.plane = -1; +} + +static bool set_no_fbc_reason(struct drm_i915_private *dev_priv, + enum no_fbc_reason reason) +{ + if (dev_priv->fbc.no_fbc_reason == reason) + return false; + + dev_priv->fbc.no_fbc_reason = reason; + return true; } /** @@ -370,7 +447,7 @@ void intel_disable_fbc(struct drm_device *dev) * - no pixel mulitply/line duplication * - no alpha buffer discard * - no dual wide - * - framebuffer <= 2048 in width, 1536 in height + * - framebuffer <= max_hdisplay in width, max_vdisplay in height * * We can't assume that any compression will take place (worst case), * so the compressed buffer has to be the same size as the uncompressed @@ -387,13 +464,19 @@ void intel_update_fbc(struct drm_device *dev) struct drm_framebuffer *fb; struct intel_framebuffer *intel_fb; struct drm_i915_gem_object *obj; - int enable_fbc; + const struct drm_display_mode *adjusted_mode; + unsigned int max_width, max_height; - if (!i915_powersave) + if (!HAS_FBC(dev)) { + set_no_fbc_reason(dev_priv, FBC_UNSUPPORTED); return; + } - if (!I915_HAS_FBC(dev)) + if (!i915.powersave) { + if (set_no_fbc_reason(dev_priv, FBC_MODULE_PARAM)) + DRM_DEBUG_KMS("fbc disabled per module param\n"); return; + } /* * If FBC is already on, we just have to verify that we can @@ -404,64 +487,65 @@ 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) { - if (tmp_crtc->enabled && - !to_intel_crtc(tmp_crtc)->primary_disabled && - tmp_crtc->fb) { + for_each_crtc(dev, tmp_crtc) { + if (intel_crtc_active(tmp_crtc) && + to_intel_crtc(tmp_crtc)->primary_enabled) { if (crtc) { - DRM_DEBUG_KMS("more than one pipe active, disabling compression\n"); - dev_priv->no_fbc_reason = FBC_MULTIPLE_PIPES; + if (set_no_fbc_reason(dev_priv, FBC_MULTIPLE_PIPES)) + DRM_DEBUG_KMS("more than one pipe active, disabling compression\n"); goto out_disable; } crtc = tmp_crtc; } } - if (!crtc || crtc->fb == NULL) { - DRM_DEBUG_KMS("no output, disabling\n"); - dev_priv->no_fbc_reason = FBC_NO_OUTPUT; + 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; - enable_fbc = i915_enable_fbc; - if (enable_fbc < 0) { - DRM_DEBUG_KMS("fbc set to per-chip default\n"); - enable_fbc = 1; - if (INTEL_INFO(dev)->gen <= 6) - enable_fbc = 0; - } - if (!enable_fbc) { - DRM_DEBUG_KMS("fbc disabled per module param\n"); - dev_priv->no_fbc_reason = FBC_MODULE_PARAM; + 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 (intel_fb->obj->base.size > dev_priv->cfb_size) { - DRM_DEBUG_KMS("framebuffer too large, disabling " - "compression\n"); - dev_priv->no_fbc_reason = FBC_STOLEN_TOO_SMALL; + 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; } - if ((crtc->mode.flags & DRM_MODE_FLAG_INTERLACE) || - (crtc->mode.flags & DRM_MODE_FLAG_DBLSCAN)) { - DRM_DEBUG_KMS("mode incompatible with compression, " - "disabling\n"); - dev_priv->no_fbc_reason = FBC_UNSUPPORTED_MODE; + if ((adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE) || + (adjusted_mode->flags & DRM_MODE_FLAG_DBLSCAN)) { + if (set_no_fbc_reason(dev_priv, FBC_UNSUPPORTED_MODE)) + DRM_DEBUG_KMS("mode incompatible with compression, " + "disabling\n"); goto out_disable; } - if ((crtc->mode.hdisplay > 2048) || - (crtc->mode.vdisplay > 1536)) { - DRM_DEBUG_KMS("mode too large for compression, disabling\n"); - dev_priv->no_fbc_reason = FBC_MODE_TOO_LARGE; + + if (IS_G4X(dev) || INTEL_INFO(dev)->gen >= 5) { + max_width = 4096; + max_height = 2048; + } else { + max_width = 2048; + max_height = 1536; + } + if (intel_crtc->config.pipe_src_w > max_width || + intel_crtc->config.pipe_src_h > max_height) { + if (set_no_fbc_reason(dev_priv, FBC_MODE_TOO_LARGE)) + DRM_DEBUG_KMS("mode too large for compression, disabling\n"); goto out_disable; } - if ((IS_I915GM(dev) || IS_I945GM(dev)) && intel_crtc->plane != 0) { - DRM_DEBUG_KMS("plane not 0, disabling compression\n"); - dev_priv->no_fbc_reason = FBC_BAD_PLANE; + 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 A, disabling compression\n"); goto out_disable; } @@ -470,8 +554,8 @@ void intel_update_fbc(struct drm_device *dev) */ if (obj->tiling_mode != I915_TILING_X || obj->fence_reg == I915_FENCE_REG_NONE) { - DRM_DEBUG_KMS("framebuffer not tiled or fenced, disabling compression\n"); - dev_priv->no_fbc_reason = FBC_NOT_TILED; + if (set_no_fbc_reason(dev_priv, FBC_NOT_TILED)) + DRM_DEBUG_KMS("framebuffer not tiled or fenced, disabling compression\n"); goto out_disable; } @@ -479,14 +563,20 @@ void intel_update_fbc(struct drm_device *dev) if (in_dbg_master()) goto out_disable; + if (i915_gem_stolen_setup_compression(dev, intel_fb->obj->base.size)) { + if (set_no_fbc_reason(dev_priv, FBC_STOLEN_TOO_SMALL)) + DRM_DEBUG_KMS("framebuffer too large, disabling compression\n"); + goto out_disable; + } + /* If the scanout has not changed, don't modify the FBC settings. * Note that we make the fundamental assumption that the fb->obj * cannot be unpinned (and have its GTT offset and fence revoked) * without first being decoupled from the scanout and FBC disabled. */ - if (dev_priv->cfb_plane == intel_crtc->plane && - dev_priv->cfb_fb == fb->base.id && - dev_priv->cfb_y == crtc->y) + if (dev_priv->fbc.plane == intel_crtc->plane && + dev_priv->fbc.fb_id == fb->base.id && + dev_priv->fbc.y == crtc->y) return; if (intel_fbc_enabled(dev)) { @@ -517,7 +607,8 @@ 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; out_disable: @@ -526,11 +617,12 @@ out_disable: DRM_DEBUG_KMS("unsupported config, disabling FBC\n"); intel_disable_fbc(dev); } + i915_gem_stolen_cleanup_compression(dev); } 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); @@ -569,7 +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); @@ -737,7 +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); @@ -770,21 +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, @@ -863,14 +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, @@ -878,65 +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 @@ -991,8 +1009,8 @@ 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) { - if (crtc->enabled && crtc->fb) { + for_each_crtc(dev, crtc) { + if (intel_crtc_active(crtc)) { if (enabled) return NULL; enabled = crtc; @@ -1002,8 +1020,9 @@ static struct drm_crtc *single_enabled_crtc(struct drm_device *dev) return enabled; } -static void pineview_update_wm(struct drm_device *dev) +static void pineview_update_wm(struct drm_crtc *unused_crtc) { + struct drm_device *dev = unused_crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct drm_crtc *crtc; const struct cxsr_latency *latency; @@ -1020,8 +1039,12 @@ static void pineview_update_wm(struct drm_device *dev) crtc = single_enabled_crtc(dev); if (crtc) { - int clock = crtc->mode.clock; - int pixel_size = crtc->fb->bits_per_pixel / 8; + const struct drm_display_mode *adjusted_mode; + int pixel_size = crtc->primary->fb->bits_per_pixel / 8; + int clock; + + adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode; + clock = adjusted_mode->crtc_clock; /* Display SR */ wm = intel_calculate_wm(clock, &pineview_display_wm, @@ -1081,21 +1104,23 @@ static bool g4x_compute_wm0(struct drm_device *dev, int *cursor_wm) { struct drm_crtc *crtc; + const struct drm_display_mode *adjusted_mode; int htotal, hdisplay, clock, pixel_size; int line_time_us, line_count; int entries, tlb_miss; crtc = intel_get_crtc_for_plane(dev, plane); - if (crtc->fb == NULL || !crtc->enabled) { + if (!intel_crtc_active(crtc)) { *cursor_wm = cursor->guard_size; *plane_wm = display->guard_size; return false; } - htotal = crtc->mode.htotal; - hdisplay = crtc->mode.hdisplay; - clock = crtc->mode.clock; - pixel_size = crtc->fb->bits_per_pixel / 8; + 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->primary->fb->bits_per_pixel / 8; /* Use the small buffer method to calculate plane watermark */ entries = ((clock * pixel_size / 1000) * display_latency_ns) / 1000; @@ -1108,9 +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; @@ -1165,6 +1190,7 @@ static bool g4x_compute_srwm(struct drm_device *dev, int *display_wm, int *cursor_wm) { struct drm_crtc *crtc; + const struct drm_display_mode *adjusted_mode; int hdisplay, htotal, pixel_size, clock; unsigned long line_time_us; int line_count, line_size; @@ -1177,12 +1203,13 @@ static bool g4x_compute_srwm(struct drm_device *dev, } crtc = intel_get_crtc_for_plane(dev, plane); - hdisplay = crtc->mode.hdisplay; - htotal = crtc->mode.htotal; - clock = crtc->mode.clock; - pixel_size = crtc->fb->bits_per_pixel / 8; + 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->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; @@ -1194,7 +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; @@ -1215,11 +1242,11 @@ static bool vlv_compute_drain_latency(struct drm_device *dev, int entries; crtc = intel_get_crtc_for_plane(dev, plane); - if (crtc->fb == NULL || !crtc->enabled) + if (!intel_crtc_active(crtc)) return false; - clock = crtc->mode.clock; /* VESA DOT Clock */ - pixel_size = crtc->fb->bits_per_pixel / 8; /* BPP */ + clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock; + pixel_size = crtc->primary->fb->bits_per_pixel / 8; /* BPP */ entries = (clock / 1000) * pixel_size; *plane_prec_mult = (entries > 256) ? @@ -1280,39 +1307,47 @@ static void vlv_update_drain_latency(struct drm_device *dev) #define single_plane_enabled(mask) is_power_of_2(mask) -static void valleyview_update_wm(struct drm_device *dev) +static void valleyview_update_wm(struct drm_crtc *crtc) { + struct drm_device *dev = crtc->dev; static const int sr_latency_ns = 12000; struct drm_i915_private *dev_priv = dev->dev_private; int planea_wm, planeb_wm, cursora_wm, cursorb_wm; int plane_sr, cursor_sr; + int ignore_plane_sr, ignore_cursor_sr; unsigned int enabled = 0; vlv_update_drain_latency(dev); - if (g4x_compute_wm0(dev, 0, + if (g4x_compute_wm0(dev, PIPE_A, &valleyview_wm_info, latency_ns, &valleyview_cursor_wm_info, latency_ns, &planea_wm, &cursora_wm)) - enabled |= 1; + enabled |= 1 << PIPE_A; - if (g4x_compute_wm0(dev, 1, + if (g4x_compute_wm0(dev, PIPE_B, &valleyview_wm_info, latency_ns, &valleyview_cursor_wm_info, latency_ns, &planeb_wm, &cursorb_wm)) - enabled |= 2; + enabled |= 1 << PIPE_B; - plane_sr = cursor_sr = 0; if (single_plane_enabled(enabled) && g4x_compute_srwm(dev, ffs(enabled) - 1, sr_latency_ns, &valleyview_wm_info, &valleyview_cursor_wm_info, - &plane_sr, &cursor_sr)) + &plane_sr, &ignore_cursor_sr) && + g4x_compute_srwm(dev, ffs(enabled) - 1, + 2*sr_latency_ns, + &valleyview_wm_info, + &valleyview_cursor_wm_info, + &ignore_plane_sr, &cursor_sr)) { I915_WRITE(FW_BLC_SELF_VLV, FW_CSPWRDWNEN); - else + } else { I915_WRITE(FW_BLC_SELF_VLV, I915_READ(FW_BLC_SELF_VLV) & ~FW_CSPWRDWNEN); + plane_sr = cursor_sr = 0; + } DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n", planea_wm, cursora_wm, @@ -1325,43 +1360,46 @@ static void valleyview_update_wm(struct drm_device *dev) (planeb_wm << DSPFW_PLANEB_SHIFT) | planea_wm); I915_WRITE(DSPFW2, - (I915_READ(DSPFW2) & DSPFW_CURSORA_MASK) | + (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) | (cursora_wm << DSPFW_CURSORA_SHIFT)); I915_WRITE(DSPFW3, - (I915_READ(DSPFW3) | (cursor_sr << DSPFW_CURSOR_SR_SHIFT))); + (I915_READ(DSPFW3) & ~DSPFW_CURSOR_SR_MASK) | + (cursor_sr << DSPFW_CURSOR_SR_SHIFT)); } -static void g4x_update_wm(struct drm_device *dev) +static void g4x_update_wm(struct drm_crtc *crtc) { + struct drm_device *dev = crtc->dev; static const int sr_latency_ns = 12000; struct drm_i915_private *dev_priv = dev->dev_private; int planea_wm, planeb_wm, cursora_wm, cursorb_wm; int plane_sr, cursor_sr; unsigned int enabled = 0; - if (g4x_compute_wm0(dev, 0, + if (g4x_compute_wm0(dev, PIPE_A, &g4x_wm_info, latency_ns, &g4x_cursor_wm_info, latency_ns, &planea_wm, &cursora_wm)) - enabled |= 1; + enabled |= 1 << PIPE_A; - if (g4x_compute_wm0(dev, 1, + if (g4x_compute_wm0(dev, PIPE_B, &g4x_wm_info, latency_ns, &g4x_cursor_wm_info, latency_ns, &planeb_wm, &cursorb_wm)) - enabled |= 2; + enabled |= 1 << PIPE_B; - plane_sr = cursor_sr = 0; if (single_plane_enabled(enabled) && g4x_compute_srwm(dev, ffs(enabled) - 1, sr_latency_ns, &g4x_wm_info, &g4x_cursor_wm_info, - &plane_sr, &cursor_sr)) + &plane_sr, &cursor_sr)) { I915_WRITE(FW_BLC_SELF, FW_BLC_SELF_EN); - else + } else { I915_WRITE(FW_BLC_SELF, I915_READ(FW_BLC_SELF) & ~FW_BLC_SELF_EN); + plane_sr = cursor_sr = 0; + } DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n", planea_wm, cursora_wm, @@ -1374,16 +1412,17 @@ static void g4x_update_wm(struct drm_device *dev) (planeb_wm << DSPFW_PLANEB_SHIFT) | planea_wm); I915_WRITE(DSPFW2, - (I915_READ(DSPFW2) & DSPFW_CURSORA_MASK) | + (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) | (cursora_wm << DSPFW_CURSORA_SHIFT)); /* HPLL off in SR has some issues on G4x... disable it */ I915_WRITE(DSPFW3, - (I915_READ(DSPFW3) & ~DSPFW_HPLL_SR_EN) | + (I915_READ(DSPFW3) & ~(DSPFW_HPLL_SR_EN | DSPFW_CURSOR_SR_MASK)) | (cursor_sr << DSPFW_CURSOR_SR_SHIFT)); } -static void i965_update_wm(struct drm_device *dev) +static void i965_update_wm(struct drm_crtc *unused_crtc) { + struct drm_device *dev = unused_crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct drm_crtc *crtc; int srwm = 1; @@ -1394,14 +1433,16 @@ static void i965_update_wm(struct drm_device *dev) if (crtc) { /* self-refresh has much higher latency */ static const int sr_latency_ns = 12000; - int clock = crtc->mode.clock; - int htotal = crtc->mode.htotal; - int hdisplay = crtc->mode.hdisplay; - int pixel_size = crtc->fb->bits_per_pixel / 8; + const struct drm_display_mode *adjusted_mode = + &to_intel_crtc(crtc)->config.adjusted_mode; + 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->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) * @@ -1415,7 +1456,7 @@ static void i965_update_wm(struct drm_device *dev) 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 - @@ -1447,8 +1488,9 @@ static void i965_update_wm(struct drm_device *dev) I915_WRITE(DSPFW3, (cursor_sr << DSPFW_CURSOR_SR_SHIFT)); } -static void i9xx_update_wm(struct drm_device *dev) +static void i9xx_update_wm(struct drm_crtc *unused_crtc) { + struct drm_device *dev = unused_crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; const struct intel_watermark_params *wm_info; uint32_t fwater_lo; @@ -1463,14 +1505,19 @@ static void i9xx_update_wm(struct drm_device *dev) 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 (crtc->enabled && crtc->fb) { - planea_wm = intel_calculate_wm(crtc->mode.clock, - wm_info, fifo_size, - crtc->fb->bits_per_pixel / 8, + if (intel_crtc_active(crtc)) { + const struct drm_display_mode *adjusted_mode; + int cpp = crtc->primary->fb->bits_per_pixel / 8; + if (IS_GEN2(dev)) + cpp = 4; + + adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode; + planea_wm = intel_calculate_wm(adjusted_mode->crtc_clock, + wm_info, fifo_size, cpp, latency_ns); enabled = crtc; } else @@ -1478,10 +1525,15 @@ static void i9xx_update_wm(struct drm_device *dev) fifo_size = dev_priv->display.get_fifo_size(dev, 1); crtc = intel_get_crtc_for_plane(dev, 1); - if (crtc->enabled && crtc->fb) { - planeb_wm = intel_calculate_wm(crtc->mode.clock, - wm_info, fifo_size, - crtc->fb->bits_per_pixel / 8, + if (intel_crtc_active(crtc)) { + const struct drm_display_mode *adjusted_mode; + int cpp = crtc->primary->fb->bits_per_pixel / 8; + if (IS_GEN2(dev)) + cpp = 4; + + adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode; + planeb_wm = intel_calculate_wm(adjusted_mode->crtc_clock, + wm_info, fifo_size, cpp, latency_ns); if (enabled == NULL) enabled = crtc; @@ -1492,6 +1544,16 @@ static void i9xx_update_wm(struct drm_device *dev) 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. */ @@ -1501,20 +1563,22 @@ static void i9xx_update_wm(struct drm_device *dev) 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) { /* self-refresh has much higher latency */ static const int sr_latency_ns = 6000; - int clock = enabled->mode.clock; - int htotal = enabled->mode.htotal; - int hdisplay = enabled->mode.hdisplay; - int pixel_size = enabled->fb->bits_per_pixel / 8; + const struct drm_display_mode *adjusted_mode = + &to_intel_crtc(enabled)->config.adjusted_mode; + 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->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) * @@ -1551,17 +1615,19 @@ static void i9xx_update_wm(struct drm_device *dev) 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_device *dev) +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; struct drm_crtc *crtc; + const struct drm_display_mode *adjusted_mode; uint32_t fwater_lo; int planea_wm; @@ -1569,10 +1635,11 @@ static void i830_update_wm(struct drm_device *dev) if (crtc == NULL) return; - planea_wm = intel_calculate_wm(crtc->mode.clock, &i830_wm_info, + adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode; + planea_wm = intel_calculate_wm(adjusted_mode->crtc_clock, + &i845_wm_info, dev_priv->display.get_fifo_size(dev, 0), - crtc->fb->bits_per_pixel / 8, - latency_ns); + 4, latency_ns); fwater_lo = I915_READ(FW_BLC) & ~0xfff; fwater_lo |= (3<<8) | planea_wm; @@ -1581,490 +1648,1166 @@ static void i830_update_wm(struct drm_device *dev) I915_WRITE(FW_BLC, fwater_lo); } -#define ILK_LP0_PLANE_LATENCY 700 -#define ILK_LP0_CURSOR_LATENCY 1300 +static uint32_t ilk_pipe_pixel_rate(struct drm_device *dev, + struct drm_crtc *crtc) +{ + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + uint32_t pixel_rate; + + pixel_rate = intel_crtc->config.adjusted_mode.crtc_clock; + + /* We only use IF-ID interlacing. If we ever use PF-ID we'll need to + * adjust the pixel_rate here. */ + + if (intel_crtc->config.pch_pfit.enabled) { + uint64_t pipe_w, pipe_h, pfit_w, pfit_h; + uint32_t pfit_size = intel_crtc->config.pch_pfit.size; + + pipe_w = intel_crtc->config.pipe_src_w; + pipe_h = intel_crtc->config.pipe_src_h; + pfit_w = (pfit_size >> 16) & 0xFFFF; + pfit_h = pfit_size & 0xFFFF; + if (pipe_w < pfit_w) + pipe_w = pfit_w; + if (pipe_h < pfit_h) + pipe_h = pfit_h; + + pixel_rate = div_u64((uint64_t) pixel_rate * pipe_w * pipe_h, + pfit_w * pfit_h); + } + + return pixel_rate; +} + +/* latency must be in 0.1us units. */ +static uint32_t ilk_wm_method1(uint32_t pixel_rate, uint8_t bytes_per_pixel, + uint32_t latency) +{ + uint64_t ret; + + if (WARN(latency == 0, "Latency value missing\n")) + return UINT_MAX; + + ret = (uint64_t) pixel_rate * bytes_per_pixel * latency; + ret = DIV_ROUND_UP_ULL(ret, 64 * 10000) + 2; + + return ret; +} + +/* latency must be in 0.1us units. */ +static uint32_t ilk_wm_method2(uint32_t pixel_rate, uint32_t pipe_htotal, + uint32_t horiz_pixels, uint8_t bytes_per_pixel, + uint32_t latency) +{ + uint32_t ret; + + if (WARN(latency == 0, "Latency value missing\n")) + return UINT_MAX; + + ret = (latency * pixel_rate) / (pipe_htotal * 10000); + ret = (ret + 1) * horiz_pixels * bytes_per_pixel; + ret = DIV_ROUND_UP(ret, 64) + 2; + return ret; +} + +static uint32_t ilk_wm_fbc(uint32_t pri_val, uint32_t horiz_pixels, + uint8_t bytes_per_pixel) +{ + return DIV_ROUND_UP(pri_val * 64, horiz_pixels * bytes_per_pixel) + 2; +} + +struct ilk_pipe_wm_parameters { + bool active; + uint32_t pipe_htotal; + uint32_t pixel_rate; + struct intel_plane_wm_parameters pri; + struct intel_plane_wm_parameters spr; + struct intel_plane_wm_parameters cur; +}; + +struct ilk_wm_maximums { + uint16_t pri; + uint16_t spr; + uint16_t cur; + uint16_t fbc; +}; + +/* used in computing the new watermarks state */ +struct intel_wm_config { + unsigned int num_pipes_active; + bool sprites_enabled; + bool sprites_scaled; +}; /* - * 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. + * For both WM_PIPE and WM_LP. + * mem_value must be in 0.1us units. */ -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) +static uint32_t ilk_compute_pri_wm(const struct ilk_pipe_wm_parameters *params, + uint32_t mem_value, + bool is_lp) { - struct drm_i915_private *dev_priv = dev->dev_private; + uint32_t method1, method2; - DRM_DEBUG_KMS("watermark %d: display plane %d, fbc lines %d," - " cursor %d\n", level, display_wm, fbc_wm, cursor_wm); + if (!params->active || !params->pri.enabled) + return 0; - 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); + method1 = ilk_wm_method1(params->pixel_rate, + params->pri.bytes_per_pixel, + mem_value); - /* 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; - } + if (!is_lp) + return method1; - 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; - } + method2 = ilk_wm_method2(params->pixel_rate, + params->pipe_htotal, + params->pri.horiz_pixels, + params->pri.bytes_per_pixel, + mem_value); - 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; - } + return min(method1, method2); +} - if (!(fbc_wm || display_wm || cursor_wm)) { - DRM_DEBUG_KMS("latency %d is 0, disabling wm%d+\n", level, level); - return false; - } +/* + * For both WM_PIPE and WM_LP. + * mem_value must be in 0.1us units. + */ +static uint32_t ilk_compute_spr_wm(const struct ilk_pipe_wm_parameters *params, + uint32_t mem_value) +{ + uint32_t method1, method2; - return true; + if (!params->active || !params->spr.enabled) + return 0; + + method1 = ilk_wm_method1(params->pixel_rate, + params->spr.bytes_per_pixel, + mem_value); + method2 = ilk_wm_method2(params->pixel_rate, + params->pipe_htotal, + params->spr.horiz_pixels, + params->spr.bytes_per_pixel, + mem_value); + return min(method1, method2); } /* - * Compute watermark values of WM[1-3], + * For both WM_PIPE and WM_LP. + * mem_value must be in 0.1us units. */ -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) +static uint32_t ilk_compute_cur_wm(const struct ilk_pipe_wm_parameters *params, + uint32_t mem_value) { - struct drm_crtc *crtc; - unsigned long line_time_us; - int hdisplay, htotal, pixel_size, clock; - int line_count, line_size; - int small, large; - int entries; + if (!params->active || !params->cur.enabled) + return 0; - if (!latency_ns) { - *fbc_wm = *display_wm = *cursor_wm = 0; - return false; + return ilk_wm_method2(params->pixel_rate, + params->pipe_htotal, + params->cur.horiz_pixels, + params->cur.bytes_per_pixel, + mem_value); +} + +/* Only for WM_LP. */ +static uint32_t ilk_compute_fbc_wm(const struct ilk_pipe_wm_parameters *params, + uint32_t pri_val) +{ + if (!params->active || !params->pri.enabled) + return 0; + + return ilk_wm_fbc(pri_val, + params->pri.horiz_pixels, + params->pri.bytes_per_pixel); +} + +static unsigned int ilk_display_fifo_size(const struct drm_device *dev) +{ + if (INTEL_INFO(dev)->gen >= 8) + return 3072; + else if (INTEL_INFO(dev)->gen >= 7) + return 768; + else + 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, + const struct intel_wm_config *config, + enum intel_ddb_partitioning ddb_partitioning, + bool is_sprite) +{ + unsigned int fifo_size = ilk_display_fifo_size(dev); + + /* if sprites aren't enabled, sprites get nothing */ + if (is_sprite && !config->sprites_enabled) + return 0; + + /* HSW allows LP1+ watermarks even with multiple pipes */ + if (level == 0 || config->num_pipes_active > 1) { + fifo_size /= INTEL_INFO(dev)->num_pipes; + + /* + * For some reason the non self refresh + * FIFO size is only half of the self + * refresh FIFO size on ILK/SNB. + */ + if (INTEL_INFO(dev)->gen <= 6) + fifo_size /= 2; } - crtc = intel_get_crtc_for_plane(dev, plane); - hdisplay = crtc->mode.hdisplay; - htotal = crtc->mode.htotal; - clock = crtc->mode.clock; - pixel_size = crtc->fb->bits_per_pixel / 8; + if (config->sprites_enabled) { + /* level 0 is always calculated with 1:1 split */ + if (level > 0 && ddb_partitioning == INTEL_DDB_PART_5_6) { + if (is_sprite) + fifo_size *= 5; + fifo_size /= 6; + } else { + fifo_size /= 2; + } + } - line_time_us = (htotal * 1000) / clock; - line_count = (latency_ns / line_time_us + 1000) / 1000; - line_size = hdisplay * pixel_size; + /* clamp to max that the registers can hold */ + return min(fifo_size, ilk_plane_wm_reg_max(dev, level, is_sprite)); +} - /* 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; +/* Calculate the maximum cursor plane watermark */ +static unsigned int ilk_cursor_wm_max(const struct drm_device *dev, + int level, + const struct intel_wm_config *config) +{ + /* HSW LP1+ watermarks w/ multiple pipes */ + if (level > 0 && config->num_pipes_active > 1) + return 64; - entries = DIV_ROUND_UP(min(small, large), display->cacheline_size); - *display_wm = entries + display->guard_size; + /* otherwise just report max that registers can hold */ + return ilk_cursor_wm_reg_max(dev, level); +} + +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 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_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 ilk_wm_maximums *max, + struct intel_wm_level *result) +{ + bool ret; + + /* already determined to be invalid? */ + if (!result->enable) + return false; + + result->enable = result->pri_val <= max->pri && + result->spr_val <= max->spr && + result->cur_val <= max->cur; + + ret = result->enable; /* - * Spec says: - * FBC WM = ((Final Primary WM * 64) / number of bytes per line) + 2 + * HACK until we can pre-compute everything, + * and thus fail gracefully if LP0 watermarks + * are exceeded... */ - *fbc_wm = DIV_ROUND_UP(*display_wm * 64, line_size) + 2; + if (level == 0 && !result->enable) { + if (result->pri_val > max->pri) + DRM_DEBUG_KMS("Primary WM%d too large %u (max %u)\n", + level, result->pri_val, max->pri); + if (result->spr_val > max->spr) + DRM_DEBUG_KMS("Sprite WM%d too large %u (max %u)\n", + level, result->spr_val, max->spr); + if (result->cur_val > max->cur) + DRM_DEBUG_KMS("Cursor WM%d too large %u (max %u)\n", + level, result->cur_val, max->cur); + + result->pri_val = min_t(uint32_t, result->pri_val, max->pri); + result->spr_val = min_t(uint32_t, result->spr_val, max->spr); + result->cur_val = min_t(uint32_t, result->cur_val, max->cur); + result->enable = true; + } - /* 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 ret; +} + +static void ilk_compute_wm_level(const struct drm_i915_private *dev_priv, + int level, + const struct ilk_pipe_wm_parameters *p, + struct intel_wm_level *result) +{ + uint16_t pri_latency = dev_priv->wm.pri_latency[level]; + uint16_t spr_latency = dev_priv->wm.spr_latency[level]; + uint16_t cur_latency = dev_priv->wm.cur_latency[level]; + + /* WM1+ latency values stored in 0.5us units */ + if (level > 0) { + pri_latency *= 5; + spr_latency *= 5; + cur_latency *= 5; + } + + result->pri_val = ilk_compute_pri_wm(p, pri_latency, level); + result->spr_val = ilk_compute_spr_wm(p, spr_latency); + result->cur_val = ilk_compute_cur_wm(p, cur_latency); + result->fbc_val = ilk_compute_fbc_wm(p, result->pri_val); + result->enable = true; +} + +static uint32_t +hsw_compute_linetime_wm(struct drm_device *dev, struct drm_crtc *crtc) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + struct drm_display_mode *mode = &intel_crtc->config.adjusted_mode; + u32 linetime, ips_linetime; + + if (!intel_crtc_active(crtc)) + return 0; - return ironlake_check_srwm(dev, level, - *fbc_wm, *display_wm, *cursor_wm, - display, cursor); + /* The WM are computed with base on how long it takes to fill a single + * row at the given clock rate, multiplied by 8. + * */ + linetime = DIV_ROUND_CLOSEST(mode->crtc_htotal * 1000 * 8, + mode->crtc_clock); + ips_linetime = DIV_ROUND_CLOSEST(mode->crtc_htotal * 1000 * 8, + intel_ddi_get_cdclk_freq(dev_priv)); + + return PIPE_WM_LINETIME_IPS_LINETIME(ips_linetime) | + PIPE_WM_LINETIME_TIME(linetime); } -static void ironlake_update_wm(struct drm_device *dev) +static void intel_read_wm_latency(struct drm_device *dev, uint16_t wm[5]) { 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, 0, - &ironlake_display_wm_info, - ILK_LP0_PLANE_LATENCY, - &ironlake_cursor_wm_info, - ILK_LP0_CURSOR_LATENCY, - &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; - } - - if (g4x_compute_wm0(dev, 1, - &ironlake_display_wm_info, - ILK_LP0_PLANE_LATENCY, - &ironlake_cursor_wm_info, - ILK_LP0_CURSOR_LATENCY, - &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 |= 2; + + if (IS_HASWELL(dev) || IS_BROADWELL(dev)) { + uint64_t sskpd = I915_READ64(MCH_SSKPD); + + wm[0] = (sskpd >> 56) & 0xFF; + if (wm[0] == 0) + wm[0] = sskpd & 0xF; + wm[1] = (sskpd >> 4) & 0xFF; + wm[2] = (sskpd >> 12) & 0xFF; + wm[3] = (sskpd >> 20) & 0x1FF; + wm[4] = (sskpd >> 32) & 0x1FF; + } else if (INTEL_INFO(dev)->gen >= 6) { + uint32_t sskpd = I915_READ(MCH_SSKPD); + + wm[0] = (sskpd >> SSKPD_WM0_SHIFT) & SSKPD_WM_MASK; + wm[1] = (sskpd >> SSKPD_WM1_SHIFT) & SSKPD_WM_MASK; + wm[2] = (sskpd >> SSKPD_WM2_SHIFT) & SSKPD_WM_MASK; + wm[3] = (sskpd >> SSKPD_WM3_SHIFT) & SSKPD_WM_MASK; + } else if (INTEL_INFO(dev)->gen >= 5) { + uint32_t mltr = I915_READ(MLTR_ILK); + + /* ILK primary LP0 latency is 700 ns */ + wm[0] = 7; + wm[1] = (mltr >> MLTR_WM1_SHIFT) & ILK_SRLT_MASK; + wm[2] = (mltr >> MLTR_WM2_SHIFT) & ILK_SRLT_MASK; } +} - /* - * 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); +static void intel_fixup_spr_wm_latency(struct drm_device *dev, uint16_t wm[5]) +{ + /* ILK sprite LP0 latency is 1300 ns */ + if (INTEL_INFO(dev)->gen == 5) + wm[0] = 13; +} - if (!single_plane_enabled(enabled)) - return; - enabled = ffs(enabled) - 1; - - /* WM1 */ - if (!ironlake_compute_srwm(dev, 1, enabled, - ILK_READ_WM1_LATENCY() * 500, - &ironlake_display_srwm_info, - &ironlake_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - return; +static void intel_fixup_cur_wm_latency(struct drm_device *dev, uint16_t wm[5]) +{ + /* ILK cursor LP0 latency is 1300 ns */ + if (INTEL_INFO(dev)->gen == 5) + wm[0] = 13; - I915_WRITE(WM1_LP_ILK, - WM1_LP_SR_EN | - (ILK_READ_WM1_LATENCY() << 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, - ILK_READ_WM2_LATENCY() * 500, - &ironlake_display_srwm_info, - &ironlake_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - return; + /* WaDoubleCursorLP3Latency:ivb */ + if (IS_IVYBRIDGE(dev)) + wm[3] *= 2; +} - I915_WRITE(WM2_LP_ILK, - WM2_LP_EN | - (ILK_READ_WM2_LATENCY() << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); +int ilk_wm_max_level(const struct drm_device *dev) +{ + /* how many WM levels are we expecting */ + if (IS_HASWELL(dev) || IS_BROADWELL(dev)) + return 4; + else if (INTEL_INFO(dev)->gen >= 6) + return 3; + else + return 2; +} - /* - * WM3 is unsupported on ILK, probably because we don't have latency - * data for that power state - */ +static void intel_print_wm_latency(struct drm_device *dev, + const char *name, + const uint16_t wm[5]) +{ + int level, max_level = ilk_wm_max_level(dev); + + for (level = 0; level <= max_level; level++) { + unsigned int latency = wm[level]; + + if (latency == 0) { + DRM_ERROR("%s WM%d latency not provided\n", + name, level); + continue; + } + + /* WM1+ latency values in 0.5us units */ + if (level > 0) + latency *= 5; + + DRM_DEBUG_KMS("%s WM%d latency %u (%u.%u usec)\n", + name, level, wm[level], + latency / 10, latency % 10); + } +} + +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 sandybridge_update_wm(struct drm_device *dev) +static void snb_wm_latency_quirk(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - int latency = SNB_READ_WM0_LATENCY() * 100; /* In unit 0.1us */ - u32 val; - int fbc_wm, plane_wm, cursor_wm; - unsigned int enabled; - - enabled = 0; - if (g4x_compute_wm0(dev, 0, - &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; - } - - if (g4x_compute_wm0(dev, 1, - &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 |= 2; - } - - if ((dev_priv->num_pipe == 3) && - g4x_compute_wm0(dev, 2, - &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 |= 3; - } + bool changed; /* - * 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 - * + * The BIOS provided WM memory latency values are often + * inadequate for high resolution displays. Adjust them. */ - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); + 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 (!single_plane_enabled(enabled) || - dev_priv->sprite_scaling_enabled) - return; - enabled = ffs(enabled) - 1; - - /* WM1 */ - if (!ironlake_compute_srwm(dev, 1, enabled, - SNB_READ_WM1_LATENCY() * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) + if (!changed) return; - I915_WRITE(WM1_LP_ILK, - WM1_LP_SR_EN | - (SNB_READ_WM1_LATENCY() << 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, - SNB_READ_WM2_LATENCY() * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - 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; + + intel_read_wm_latency(dev, dev_priv->wm.pri_latency); + + memcpy(dev_priv->wm.spr_latency, dev_priv->wm.pri_latency, + sizeof(dev_priv->wm.pri_latency)); + memcpy(dev_priv->wm.cur_latency, dev_priv->wm.pri_latency, + sizeof(dev_priv->wm.pri_latency)); + + intel_fixup_spr_wm_latency(dev, dev_priv->wm.spr_latency); + intel_fixup_cur_wm_latency(dev, dev_priv->wm.cur_latency); + + 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 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; - I915_WRITE(WM2_LP_ILK, - WM2_LP_EN | - (SNB_READ_WM2_LATENCY() << 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, - SNB_READ_WM3_LATENCY() * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) + if (!intel_crtc_active(crtc)) return; - I915_WRITE(WM3_LP_ILK, - WM3_LP_EN | - (SNB_READ_WM3_LATENCY() << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); + 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) { + p->spr = intel_plane->wm; + break; + } + } } -static void -haswell_update_linetime_wm(struct drm_device *dev, int pipe, - struct drm_display_mode *mode) +static void ilk_compute_wm_config(struct drm_device *dev, + struct intel_wm_config *config) { - struct drm_i915_private *dev_priv = dev->dev_private; - u32 temp; + struct intel_crtc *intel_crtc; - temp = I915_READ(PIPE_WM_LINETIME(pipe)); - temp &= ~PIPE_WM_LINETIME_MASK; + /* Compute the currently _active_ config */ + for_each_intel_crtc(dev, intel_crtc) { + const struct intel_pipe_wm *wm = &intel_crtc->wm.active; - /* The WM are computed with base on how long it takes to fill a single - * row at the given clock rate, multiplied by 8. - * */ - temp |= PIPE_WM_LINETIME_TIME( - ((mode->crtc_hdisplay * 1000) / mode->clock) * 8); + if (!wm->pipe_enabled) + continue; - /* IPS watermarks are only used by pipe A, and are ignored by - * pipes B and C. They are calculated similarly to the common - * linetime values, except that we are using CD clock frequency - * in MHz instead of pixel rate for the division. - * - * This is a placeholder for the IPS watermark calculation code. + 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 ilk_pipe_wm_parameters *params, + struct intel_pipe_wm *pipe_wm) +{ + struct drm_device *dev = crtc->dev; + 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 = { + .num_pipes_active = 1, + .sprites_enabled = params->spr.enabled, + .sprites_scaled = params->spr.scaled, + }; + 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); + + /* At least LP0 must be valid */ + if (!ilk_validate_wm_level(0, &max, &pipe_wm->wm[0])) + return false; + + ilk_compute_wm_reg_maximums(dev, 1, &max); + + 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; +} + +/* + * Merge the watermarks from all active pipes for a specific level. + */ +static void ilk_merge_wm_level(struct drm_device *dev, + int level, + struct intel_wm_level *ret_wm) +{ + const struct intel_crtc *intel_crtc; + + 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) + 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); + } +} + +/* + * Merge all low power watermarks for all active pipes. + */ +static void ilk_wm_merge(struct drm_device *dev, + 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; + + /* 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++) { + struct intel_wm_level *wm = &merged->wm[level]; + + ilk_merge_wm_level(dev, level, wm); + + 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) { + 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]; - I915_WRITE(PIPE_WM_LINETIME(pipe), temp); + wm->enable = false; + } + } } -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) +static int ilk_wm_lp_to_level(int wm_lp, const struct intel_pipe_wm *pipe_wm) { - struct drm_crtc *crtc; - int clock; - int entries, tlb_miss; + /* LP1,LP2,LP3 levels are either 1,2,3 or 1,3,4 */ + return wm_lp + (wm_lp >= 2 && pipe_wm->wm[4].enable); +} - crtc = intel_get_crtc_for_plane(dev, plane); - if (crtc->fb == NULL || !crtc->enabled) { - *sprite_wm = display->guard_size; - return false; +/* 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 ilk_wm_values *results) +{ + struct intel_crtc *intel_crtc; + int level, wm_lp; + + results->enable_fbc_wm = merged->fbc_wm_enabled; + results->partitioning = partitioning; + + /* LP1+ register values */ + for (wm_lp = 1; wm_lp <= 3; wm_lp++) { + const struct intel_wm_level *r; + + level = ilk_wm_lp_to_level(wm_lp, merged); + + r = &merged->wm[level]; + + /* + * 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; + else + results->wm_lp[wm_lp - 1] |= + r->fbc_val << WM1_LP_FBC_SHIFT; + + /* + * 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; } - clock = crtc->mode.clock; + /* LP0 register values */ + 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]; - /* 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; + if (WARN_ON(!r->enable)) + continue; - return true; + results->wm_linetime[pipe] = intel_crtc->wm.active.linetime; + + results->wm_pipe[pipe] = + (r->pri_val << WM0_PIPE_PLANE_SHIFT) | + (r->spr_val << WM0_PIPE_SPRITE_SHIFT) | + r->cur_val; + } } -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) +/* 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 *ilk_find_best_result(struct drm_device *dev, + struct intel_pipe_wm *r1, + struct intel_pipe_wm *r2) { - struct drm_crtc *crtc; - unsigned long line_time_us; - int clock; - int line_count, line_size; - int small, large; - int entries; + int level, max_level = ilk_wm_max_level(dev); + int level1 = 0, level2 = 0; + + for (level = 1; level <= max_level; level++) { + if (r1->wm[level].enable) + level1 = level; + if (r2->wm[level].enable) + level2 = level; + } - if (!latency_ns) { - *sprite_wm = 0; - return false; + if (level1 == level2) { + if (r2->fbc_wm_enabled && !r1->fbc_wm_enabled) + return r2; + else + return r1; + } else if (level1 > level2) { + return r1; + } else { + return r2; } +} - crtc = intel_get_crtc_for_plane(dev, plane); - clock = crtc->mode.clock; - if (!clock) { - *sprite_wm = 0; - return false; +/* dirty bits used to track which watermarks need changes */ +#define WM_DIRTY_PIPE(pipe) (1 << (pipe)) +#define WM_DIRTY_LINETIME(pipe) (1 << (8 + (pipe))) +#define WM_DIRTY_LP(wm_lp) (1 << (15 + (wm_lp))) +#define WM_DIRTY_LP_ALL (WM_DIRTY_LP(1) | WM_DIRTY_LP(2) | WM_DIRTY_LP(3)) +#define WM_DIRTY_FBC (1 << 24) +#define WM_DIRTY_DDB (1 << 25) + +static unsigned int ilk_compute_wm_dirty(struct drm_device *dev, + const struct ilk_wm_values *old, + const struct ilk_wm_values *new) +{ + unsigned int dirty = 0; + enum pipe pipe; + int wm_lp; + + for_each_pipe(pipe) { + if (old->wm_linetime[pipe] != new->wm_linetime[pipe]) { + dirty |= WM_DIRTY_LINETIME(pipe); + /* Must disable LP1+ watermarks too */ + dirty |= WM_DIRTY_LP_ALL; + } + + if (old->wm_pipe[pipe] != new->wm_pipe[pipe]) { + dirty |= WM_DIRTY_PIPE(pipe); + /* Must disable LP1+ watermarks too */ + dirty |= WM_DIRTY_LP_ALL; + } } - line_time_us = (sprite_width * 1000) / clock; - if (!line_time_us) { - *sprite_wm = 0; - return false; + if (old->enable_fbc_wm != new->enable_fbc_wm) { + dirty |= WM_DIRTY_FBC; + /* Must disable LP1+ watermarks too */ + dirty |= WM_DIRTY_LP_ALL; } - line_count = (latency_ns / line_time_us + 1000) / 1000; - line_size = sprite_width * pixel_size; + if (old->partitioning != new->partitioning) { + dirty |= WM_DIRTY_DDB; + /* Must disable LP1+ watermarks too */ + dirty |= WM_DIRTY_LP_ALL; + } - /* 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; + /* LP1+ watermarks already deemed dirty, no need to continue */ + if (dirty & WM_DIRTY_LP_ALL) + return dirty; - entries = DIV_ROUND_UP(min(small, large), display->cacheline_size); - *sprite_wm = entries + display->guard_size; + /* Find the lowest numbered LP1+ watermark in need of an update... */ + for (wm_lp = 1; wm_lp <= 3; wm_lp++) { + if (old->wm_lp[wm_lp - 1] != new->wm_lp[wm_lp - 1] || + old->wm_lp_spr[wm_lp - 1] != new->wm_lp_spr[wm_lp - 1]) + break; + } - return *sprite_wm > 0x3ff ? false : true; + /* ...and mark it and all higher numbered LP1+ watermarks as dirty */ + for (; wm_lp <= 3; wm_lp++) + dirty |= WM_DIRTY_LP(wm_lp); + + return dirty; } -static void sandybridge_update_sprite_wm(struct drm_device *dev, int pipe, - uint32_t sprite_width, int pixel_size) +static bool _ilk_disable_lp_wm(struct drm_i915_private *dev_priv, + unsigned int dirty) { - struct drm_i915_private *dev_priv = dev->dev_private; - int latency = SNB_READ_WM0_LATENCY() * 100; /* In unit 0.1us */ - u32 val; - int sprite_wm, reg; - int ret; + struct ilk_wm_values *previous = &dev_priv->wm.hw; + bool changed = false; - 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 */ + 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; +} - 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 %d\n", - pipe); +/* + * 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 ilk_write_wm_values(struct drm_i915_private *dev_priv, + struct ilk_wm_values *results) +{ + 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, previous, results); + if (!dirty) return; + + _ilk_disable_lp_wm(dev_priv, dirty); + + if (dirty & WM_DIRTY_PIPE(PIPE_A)) + I915_WRITE(WM0_PIPEA_ILK, results->wm_pipe[0]); + if (dirty & WM_DIRTY_PIPE(PIPE_B)) + I915_WRITE(WM0_PIPEB_ILK, results->wm_pipe[1]); + if (dirty & WM_DIRTY_PIPE(PIPE_C)) + I915_WRITE(WM0_PIPEC_IVB, results->wm_pipe[2]); + + if (dirty & WM_DIRTY_LINETIME(PIPE_A)) + I915_WRITE(PIPE_WM_LINETIME(PIPE_A), results->wm_linetime[0]); + if (dirty & WM_DIRTY_LINETIME(PIPE_B)) + I915_WRITE(PIPE_WM_LINETIME(PIPE_B), results->wm_linetime[1]); + if (dirty & WM_DIRTY_LINETIME(PIPE_C)) + I915_WRITE(PIPE_WM_LINETIME(PIPE_C), results->wm_linetime[2]); + + if (dirty & WM_DIRTY_DDB) { + 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); + } } - 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 %d - %d\n", pipe, sprite_wm); + if (dirty & WM_DIRTY_FBC) { + val = I915_READ(DISP_ARB_CTL); + if (results->enable_fbc_wm) + val &= ~DISP_FBC_WM_DIS; + else + val |= DISP_FBC_WM_DIS; + I915_WRITE(DISP_ARB_CTL, val); + } + 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]); - ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width, - pixel_size, - &sandybridge_display_srwm_info, - SNB_READ_WM1_LATENCY() * 500, - &sprite_wm); - if (!ret) { - DRM_DEBUG_KMS("failed to compute sprite lp1 wm on pipe %d\n", - pipe); - return; + 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]); } - I915_WRITE(WM1S_LP_ILK, sprite_wm); - /* Only IVB has two more LP watermarks for sprite */ - if (!IS_IVYBRIDGE(dev)) - return; + 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) && previous->wm_lp[1] != results->wm_lp[1]) + I915_WRITE(WM2_LP_ILK, results->wm_lp[1]); + 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; +} - ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width, - pixel_size, - &sandybridge_display_srwm_info, - SNB_READ_WM2_LATENCY() * 500, - &sprite_wm); - if (!ret) { - DRM_DEBUG_KMS("failed to compute sprite lp2 wm on pipe %d\n", - pipe); +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 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 = {}; + + ilk_compute_wm_parameters(crtc, ¶ms); + + intel_compute_pipe_wm(crtc, ¶ms, &pipe_wm); + + if (!memcmp(&intel_crtc->wm.active, &pipe_wm, sizeof(pipe_wm))) return; + + 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, &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, &config, &max, &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; } - I915_WRITE(WM2S_LP_IVB, sprite_wm); - ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width, - pixel_size, - &sandybridge_display_srwm_info, - SNB_READ_WM3_LATENCY() * 500, - &sprite_wm); - if (!ret) { - DRM_DEBUG_KMS("failed to compute sprite lp3 wm on pipe %d\n", - pipe); - return; + partitioning = (best_lp_wm == &lp_wm_1_2) ? + INTEL_DDB_PART_1_2 : INTEL_DDB_PART_5_6; + + ilk_compute_wm_results(dev, best_lp_wm, partitioning, &results); + + ilk_write_wm_values(dev_priv, &results); +} + +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; + intel_plane->wm.scaled = scaled; + intel_plane->wm.horiz_pixels = sprite_width; + intel_plane->wm.bytes_per_pixel = pixel_size; + + /* + * 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); + + 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 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; + static const unsigned int wm0_pipe_reg[] = { + [PIPE_A] = WM0_PIPEA_ILK, + [PIPE_B] = WM0_PIPEB_ILK, + [PIPE_C] = WM0_PIPEC_IVB, + }; + + hw->wm_pipe[pipe] = I915_READ(wm0_pipe_reg[pipe]); + if (IS_HASWELL(dev) || IS_BROADWELL(dev)) + hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe)); + + active->pipe_enabled = intel_crtc_active(crtc); + + if (active->pipe_enabled) { + u32 tmp = hw->wm_pipe[pipe]; + + /* + * For active pipes LP0 watermark is marked as + * enabled, and LP1+ watermaks as disabled since + * we can't really reverse compute them in case + * multiple pipes are active. + */ + active->wm[0].enable = true; + active->wm[0].pri_val = (tmp & WM0_PIPE_PLANE_MASK) >> WM0_PIPE_PLANE_SHIFT; + active->wm[0].spr_val = (tmp & WM0_PIPE_SPRITE_MASK) >> WM0_PIPE_SPRITE_SHIFT; + active->wm[0].cur_val = tmp & WM0_PIPE_CURSOR_MASK; + active->linetime = hw->wm_linetime[pipe]; + } else { + int level, max_level = ilk_wm_max_level(dev); + + /* + * For inactive pipes, all watermark levels + * should be marked as enabled but zeroed, + * which is what we'd compute them to. + */ + for (level = 0; level <= max_level; level++) + active->wm[level].enable = true; + } +} + +void ilk_wm_get_hw_state(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + struct ilk_wm_values *hw = &dev_priv->wm.hw; + struct drm_crtc *crtc; + + for_each_crtc(dev, crtc) + ilk_pipe_wm_get_hw_state(crtc); + + hw->wm_lp[0] = I915_READ(WM1_LP_ILK); + hw->wm_lp[1] = I915_READ(WM2_LP_ILK); + hw->wm_lp[2] = I915_READ(WM3_LP_ILK); + + hw->wm_lp_spr[0] = I915_READ(WM1S_LP_ILK); + 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); } - I915_WRITE(WM3S_LP_IVB, sprite_wm); + + 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); } /** @@ -2099,31 +2842,24 @@ static void sandybridge_update_sprite_wm(struct drm_device *dev, int pipe, * We don't use the sprite, so we can ignore that. And on Crestline we have * to set the non-SR watermarks to 8. */ -void intel_update_watermarks(struct drm_device *dev) +void intel_update_watermarks(struct drm_crtc *crtc) { - struct drm_i915_private *dev_priv = dev->dev_private; + struct drm_i915_private *dev_priv = crtc->dev->dev_private; if (dev_priv->display.update_wm) - dev_priv->display.update_wm(dev); -} - -void intel_update_linetime_watermarks(struct drm_device *dev, - int pipe, struct drm_display_mode *mode) -{ - struct drm_i915_private *dev_priv = dev->dev_private; - - if (dev_priv->display.update_linetime_wm) - dev_priv->display.update_linetime_wm(dev, pipe, mode); + dev_priv->display.update_wm(crtc); } -void intel_update_sprite_watermarks(struct drm_device *dev, int pipe, - uint32_t sprite_width, int pixel_size) +void intel_update_sprite_watermarks(struct drm_plane *plane, + struct drm_crtc *crtc, + uint32_t sprite_width, int pixel_size, + bool enabled, bool scaled) { - struct drm_i915_private *dev_priv = dev->dev_private; + struct drm_i915_private *dev_priv = plane->dev->dev_private; if (dev_priv->display.update_sprite_wm) - dev_priv->display.update_sprite_wm(dev, pipe, sprite_width, - pixel_size); + dev_priv->display.update_sprite_wm(plane, crtc, sprite_width, + pixel_size, enabled, scaled); } static struct drm_i915_gem_object * @@ -2140,7 +2876,7 @@ intel_alloc_context_page(struct drm_device *dev) return NULL; } - ret = i915_gem_object_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; @@ -2155,10 +2891,9 @@ 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); - mutex_unlock(&dev->struct_mutex); return NULL; } @@ -2294,122 +3029,520 @@ 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; } +static void gen6_set_rps_thresholds(struct drm_i915_private *dev_priv, u8 val) +{ + int new_power; + + new_power = dev_priv->rps.power; + switch (dev_priv->rps.power) { + case LOW_POWER: + 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.efficient_freq && val < dev_priv->rps.cur_freq) + new_power = LOW_POWER; + 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_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_freq_softlimit) + new_power = LOW_POWER; + if (val == dev_priv->rps.max_freq_softlimit) + new_power = HIGH_POWER; + if (new_power == dev_priv->rps.power) + return; + + /* Note the units here are not exactly 1us, but 1280ns. */ + switch (new_power) { + case LOW_POWER: + /* Upclock if more than 95% busy over 16ms */ + I915_WRITE(GEN6_RP_UP_EI, 12500); + I915_WRITE(GEN6_RP_UP_THRESHOLD, 11800); + + /* Downclock if less than 85% busy over 32ms */ + I915_WRITE(GEN6_RP_DOWN_EI, 25000); + I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 21250); + + I915_WRITE(GEN6_RP_CONTROL, + GEN6_RP_MEDIA_TURBO | + GEN6_RP_MEDIA_HW_NORMAL_MODE | + GEN6_RP_MEDIA_IS_GFX | + GEN6_RP_ENABLE | + GEN6_RP_UP_BUSY_AVG | + GEN6_RP_DOWN_IDLE_AVG); + break; + + case BETWEEN: + /* Upclock if more than 90% busy over 13ms */ + I915_WRITE(GEN6_RP_UP_EI, 10250); + I915_WRITE(GEN6_RP_UP_THRESHOLD, 9225); + + /* Downclock if less than 75% busy over 32ms */ + I915_WRITE(GEN6_RP_DOWN_EI, 25000); + I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 18750); + + I915_WRITE(GEN6_RP_CONTROL, + GEN6_RP_MEDIA_TURBO | + GEN6_RP_MEDIA_HW_NORMAL_MODE | + GEN6_RP_MEDIA_IS_GFX | + GEN6_RP_ENABLE | + GEN6_RP_UP_BUSY_AVG | + GEN6_RP_DOWN_IDLE_AVG); + break; + + case HIGH_POWER: + /* Upclock if more than 85% busy over 10ms */ + I915_WRITE(GEN6_RP_UP_EI, 8000); + I915_WRITE(GEN6_RP_UP_THRESHOLD, 6800); + + /* Downclock if less than 60% busy over 32ms */ + I915_WRITE(GEN6_RP_DOWN_EI, 25000); + I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 15000); + + I915_WRITE(GEN6_RP_CONTROL, + GEN6_RP_MEDIA_TURBO | + GEN6_RP_MEDIA_HW_NORMAL_MODE | + GEN6_RP_MEDIA_IS_GFX | + GEN6_RP_ENABLE | + GEN6_RP_UP_BUSY_AVG | + GEN6_RP_DOWN_IDLE_AVG); + break; + } + + dev_priv->rps.power = new_power; + 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->struct_mutex)); - WARN_ON(val > dev_priv->rps.max_delay); - WARN_ON(val < dev_priv->rps.min_delay); + 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 (val == dev_priv->rps.cur_delay) - return; + /* 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); - 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); } -static void gen6_disable_rps(struct drm_device *dev) +/* 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 (IS_VALLEYVIEW(dev)) + vlv_set_rps_idle(dev_priv); + else + 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); +} + +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 (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_freq_softlimit); + dev_priv->rps.last_adj = 0; + } + mutex_unlock(&dev_priv->rps.hw_lock); +} + +void valleyview_set_rps(struct drm_device *dev, u8 val) +{ + 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); + + 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); + + if (val != dev_priv->rps.cur_freq) + vlv_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ, val); + + I915_WRITE(GEN6_PMINTRMSK, gen6_rps_pm_mask(dev_priv, val)); + + dev_priv->rps.cur_freq = val; + trace_intel_gpu_freq_change(vlv_gpu_freq(dev_priv, val)); +} + +static void gen8_disable_rps_interrupts(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + 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. */ + + spin_lock_irq(&dev_priv->irq_lock); + dev_priv->rps.pm_iir = 0; + spin_unlock_irq(&dev_priv->irq_lock); + + I915_WRITE(GEN8_GT_IIR(2), dev_priv->pm_rps_events); +} + +static void gen6_disable_rps_interrupts(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - I915_WRITE(GEN6_RC_CONTROL, 0); - I915_WRITE(GEN6_RPNSWREQ, 1 << 31); I915_WRITE(GEN6_PMINTRMSK, 0xffffffff); - I915_WRITE(GEN6_PMIER, 0); + 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 * stale bits in PMIIR and PMIMR which gen6_enable_rps will clean up. */ - spin_lock_irq(&dev_priv->rps.lock); + spin_lock_irq(&dev_priv->irq_lock); dev_priv->rps.pm_iir = 0; - spin_unlock_irq(&dev_priv->rps.lock); + spin_unlock_irq(&dev_priv->irq_lock); - I915_WRITE(GEN6_PMIIR, I915_READ(GEN6_PMIIR)); + I915_WRITE(GEN6_PMIIR, dev_priv->pm_rps_events); } -int intel_enable_rc6(const struct drm_device *dev) +static void gen6_disable_rps(struct drm_device *dev) { - /* Respect the kernel parameter if it is set */ - if (i915_enable_rc6 >= 0) - return i915_enable_rc6; + struct drm_i915_private *dev_priv = dev->dev_private; - if (INTEL_INFO(dev)->gen == 5) { -#ifdef CONFIG_INTEL_IOMMU - /* Disable rc6 on ilk if VT-d is on. */ - if (intel_iommu_gfx_mapped) - return false; -#endif - DRM_DEBUG_DRIVER("Ironlake: only RC6 available\n"); - return INTEL_RC6_ENABLE; - } + I915_WRITE(GEN6_RC_CONTROL, 0); + I915_WRITE(GEN6_RPNSWREQ, 1 << 31); + + if (IS_BROADWELL(dev)) + gen8_disable_rps_interrupts(dev); + else + gen6_disable_rps_interrupts(dev); +} + +static void valleyview_disable_rps(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + I915_WRITE(GEN6_RC_CONTROL, 0); + + gen6_disable_rps_interrupts(dev); +} - if (IS_HASWELL(dev)) { - DRM_DEBUG_DRIVER("Haswell: only RC6 available\n"); - return INTEL_RC6_ENABLE; +static void intel_print_rc6_info(struct drm_device *dev, u32 mode) +{ + 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"); +} + +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 (enable_rc6 >= 0) { + int mask; - /* snb/ivb have more than one rc6 state. */ - if (INTEL_INFO(dev)->gen == 6) { - DRM_DEBUG_DRIVER("Sandybridge: deep RC6 disabled\n"); - return INTEL_RC6_ENABLE; + 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; } - DRM_DEBUG_DRIVER("RC6 and deep RC6 enabled\n"); - return (INTEL_RC6_ENABLE | INTEL_RC6p_ENABLE); + /* Disable RC6 on Ironlake */ + if (INTEL_INFO(dev)->gen == 5) + return 0; + + 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; +} + +static void gen8_enable_rps_interrupts(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + 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; + + spin_lock_irq(&dev_priv->irq_lock); + WARN_ON(dev_priv->rps.pm_iir); + 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); +} + +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; + + /* 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; +} + +static void gen8_enable_rps(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + struct intel_engine_cs *ring; + uint32_t rc6_mask = 0, rp_state_cap; + int unused; + + /* 1a: Software RC state - RC0 */ + I915_WRITE(GEN6_RC_STATE, 0); + + /* 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, 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); + I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */ + I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */ + for_each_ring(ring, dev_priv, unused) + I915_WRITE(RING_MAX_IDLE(ring->mmio_base), 10); + I915_WRITE(GEN6_RC_SLEEP, 0); + I915_WRITE(GEN6_RC6_THRESHOLD, 50000); /* 50/125ms per EI */ + + /* 3: Enable RC6 */ + if (intel_enable_rc6(dev) & INTEL_RC6_ENABLE) + rc6_mask = GEN6_RC_CTL_RC6_ENABLE; + 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); + + /* 4 Program defaults and thresholds for RPS*/ + 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_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%*/ + I915_WRITE(GEN6_RP_UP_EI, 66000); /* 84.48ms, XXX: random? */ + I915_WRITE(GEN6_RP_DOWN_EI, 350000); /* 448ms, XXX: random? */ + + I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10); + + /* 5: Enable RPS */ + I915_WRITE(GEN6_RP_CONTROL, + GEN6_RP_MEDIA_TURBO | + GEN6_RP_MEDIA_HW_NORMAL_MODE | + GEN6_RP_MEDIA_IS_GFX | + GEN6_RP_ENABLE | + GEN6_RP_UP_BUSY_AVG | + GEN6_RP_DOWN_IDLE_AVG); + + /* 6: Ring frequency + overclocking (our driver does this later */ + + gen6_set_rps(dev, (I915_READ(GEN6_GT_PERF_STATUS) & 0xff00) >> 8); + + gen8_enable_rps_interrupts(dev); + + 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 pcu_mbox, rc6_mask = 0; + u32 rc6vids, pcu_mbox = 0, rc6_mask = 0; u32 gtfifodbg; int rc6_mode; - int i; + int i, ret; - WARN_ON(!mutex_is_locked(&dev->struct_mutex)); + WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); /* Here begins a magic sequence of register writes to enable * auto-downclocking. @@ -2425,15 +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 100MHz */ - dev_priv->rps.max_delay = rp_state_cap & 0xff; - dev_priv->rps.min_delay = (rp_state_cap & 0xff0000) >> 16; - 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); @@ -2449,8 +3579,11 @@ static void gen6_enable_rps(struct drm_device *dev) I915_WRITE(GEN6_RC_SLEEP, 0); I915_WRITE(GEN6_RC1e_THRESHOLD, 1000); - I915_WRITE(GEN6_RC6_THRESHOLD, 50000); - I915_WRITE(GEN6_RC6p_THRESHOLD, 100000); + if (IS_IVYBRIDGE(dev)) + I915_WRITE(GEN6_RC6_THRESHOLD, 125000); + else + I915_WRITE(GEN6_RC6_THRESHOLD, 50000); + I915_WRITE(GEN6_RC6p_THRESHOLD, 150000); I915_WRITE(GEN6_RC6pp_THRESHOLD, 64000); /* unused */ /* Check if we are enabling RC6 */ @@ -2467,148 +3600,357 @@ static void gen6_enable_rps(struct drm_device *dev) rc6_mask |= GEN6_RC_CTL_RC6pp_ENABLE; } - DRM_INFO("Enabling RC6 states: RC6 %s, RC6p %s, RC6pp %s\n", - (rc6_mask & GEN6_RC_CTL_RC6_ENABLE) ? "on" : "off", - (rc6_mask & GEN6_RC_CTL_RC6p_ENABLE) ? "on" : "off", - (rc6_mask & GEN6_RC_CTL_RC6pp_ENABLE) ? "on" : "off"); + intel_print_rc6_info(dev, rc6_mask); I915_WRITE(GEN6_RC_CONTROL, rc6_mask | GEN6_RC_CTL_EI_MODE(1) | GEN6_RC_CTL_HW_ENABLE); - I915_WRITE(GEN6_RPNSWREQ, - GEN6_FREQUENCY(10) | - GEN6_OFFSET(0) | - GEN6_AGGRESSIVE_TURBO); - I915_WRITE(GEN6_RC_VIDEO_FREQ, - GEN6_FREQUENCY(12)); - - I915_WRITE(GEN6_RP_DOWN_TIMEOUT, 1000000); - I915_WRITE(GEN6_RP_INTERRUPT_LIMITS, - dev_priv->rps.max_delay << 24 | - dev_priv->rps.min_delay << 16); - - I915_WRITE(GEN6_RP_UP_THRESHOLD, 59400); - I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 245000); - I915_WRITE(GEN6_RP_UP_EI, 66000); - I915_WRITE(GEN6_RP_DOWN_EI, 350000); - + /* Power down if completely idle for over 50ms */ + I915_WRITE(GEN6_RP_DOWN_TIMEOUT, 50000); I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10); - I915_WRITE(GEN6_RP_CONTROL, - GEN6_RP_MEDIA_TURBO | - GEN6_RP_MEDIA_HW_NORMAL_MODE | - GEN6_RP_MEDIA_IS_GFX | - GEN6_RP_ENABLE | - GEN6_RP_UP_BUSY_AVG | - (IS_HASWELL(dev) ? GEN7_RP_DOWN_IDLE_AVG : GEN6_RP_DOWN_IDLE_CONT)); - if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0, - 500)) - DRM_ERROR("timeout waiting for pcode mailbox to become idle\n"); - - I915_WRITE(GEN6_PCODE_DATA, 0); - I915_WRITE(GEN6_PCODE_MAILBOX, - GEN6_PCODE_READY | - GEN6_PCODE_WRITE_MIN_FREQ_TABLE); - if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0, - 500)) - DRM_ERROR("timeout waiting for pcode mailbox to finish\n"); - - /* Check for overclock support */ - if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0, - 500)) - DRM_ERROR("timeout waiting for pcode mailbox to become idle\n"); - I915_WRITE(GEN6_PCODE_MAILBOX, GEN6_READ_OC_PARAMS); - pcu_mbox = I915_READ(GEN6_PCODE_DATA); - if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0, - 500)) - DRM_ERROR("timeout waiting for pcode mailbox to finish\n"); - if (pcu_mbox & (1<<31)) { /* OC supported */ - dev_priv->rps.max_delay = pcu_mbox & 0xff; - DRM_DEBUG_DRIVER("overclocking supported, adjusting frequency max to %dMHz\n", pcu_mbox * 50); + ret = sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_MIN_FREQ_TABLE, 0); + 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; } - gen6_set_rps(dev_priv->dev, (gt_perf_status & 0xff00) >> 8); - - /* requires MSI enabled */ - I915_WRITE(GEN6_PMIER, GEN6_PM_DEFERRED_EVENTS); - spin_lock_irq(&dev_priv->rps.lock); - WARN_ON(dev_priv->rps.pm_iir != 0); - I915_WRITE(GEN6_PMIMR, 0); - spin_unlock_irq(&dev_priv->rps.lock); - /* enable all PM interrupts */ - I915_WRITE(GEN6_PMINTRMSK, 0); + dev_priv->rps.power = HIGH_POWER; /* force a reset */ + gen6_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit); + + gen6_enable_rps_interrupts(dev); + + rc6vids = 0; + ret = sandybridge_pcode_read(dev_priv, GEN6_PCODE_READ_RC6VIDS, &rc6vids); + if (IS_GEN6(dev) && ret) { + DRM_DEBUG_DRIVER("Couldn't check for BIOS workaround\n"); + } else if (IS_GEN6(dev) && (GEN6_DECODE_RC6_VID(rc6vids & 0xff) < 450)) { + DRM_DEBUG_DRIVER("You should update your BIOS. Correcting minimum rc6 voltage (%dmV->%dmV)\n", + GEN6_DECODE_RC6_VID(rc6vids & 0xff), 450); + rc6vids &= 0xffff00; + rc6vids |= GEN6_ENCODE_RC6_VID(450); + ret = sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_RC6VIDS, rc6vids); + if (ret) + 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); } -static 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; - int gpu_freq, ia_freq, max_ia_freq; + unsigned int gpu_freq; + unsigned int max_ia_freq, min_ring_freq; int scaling_factor = 180; + struct cpufreq_policy *policy; - WARN_ON(!mutex_is_locked(&dev->struct_mutex)); + WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); - max_ia_freq = cpufreq_quick_get_max(0); - /* - * Default to measured freq if none found, PCU will ensure we don't go - * over - */ - if (!max_ia_freq) + policy = cpufreq_cpu_get(0); + if (policy) { + max_ia_freq = policy->cpuinfo.max_freq; + cpufreq_cpu_put(policy); + } else { + /* + * Default to measured freq if none found, PCU will ensure we + * don't go over + */ max_ia_freq = tsc_khz; + } /* Convert from kHz to MHz */ max_ia_freq /= 1000; + min_ring_freq = I915_READ(DCLK) & 0xf; + /* convert DDR frequency from units of 266.6MHz to bandwidth */ + min_ring_freq = mult_frac(min_ring_freq, 8, 3); + /* * For each potential GPU frequency, load a ring frequency we'd like * 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; - /* - * For GPU frequencies less than 750MHz, just use the lowest - * ring freq. - */ - if (gpu_freq < min_freq) - ia_freq = 800; - else - ia_freq = max_ia_freq - ((diff * scaling_factor) / 2); - ia_freq = DIV_ROUND_CLOSEST(ia_freq, 100); - - I915_WRITE(GEN6_PCODE_DATA, - (ia_freq << GEN6_PCODE_FREQ_IA_RATIO_SHIFT) | - gpu_freq); - I915_WRITE(GEN6_PCODE_MAILBOX, GEN6_PCODE_READY | - GEN6_PCODE_WRITE_MIN_FREQ_TABLE); - if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & - GEN6_PCODE_READY) == 0, 10)) { - DRM_ERROR("pcode write of freq table timed out\n"); - continue; + if (INTEL_INFO(dev)->gen >= 8) { + /* max(2 * GT, DDR). NB: GT is 50MHz units */ + ring_freq = max(min_ring_freq, gpu_freq); + } else if (IS_HASWELL(dev)) { + ring_freq = mult_frac(gpu_freq, 5, 4); + ring_freq = max(min_ring_freq, ring_freq); + /* leave ia_freq as the default, chosen by cpufreq */ + } else { + /* On older processors, there is no separate ring + * clock domain, so in order to boost the bandwidth + * of the ring, we need to upclock the CPU (ia_freq). + * + * For GPU frequencies less than 750MHz, + * just use the lowest ring freq. + */ + if (gpu_freq < min_freq) + ia_freq = 800; + else + ia_freq = max_ia_freq - ((diff * scaling_factor) / 2); + ia_freq = DIV_ROUND_CLOSEST(ia_freq, 100); } + + sandybridge_pcode_write(dev_priv, + GEN6_PCODE_WRITE_MIN_FREQ_TABLE, + ia_freq << GEN6_PCODE_FREQ_IA_RATIO_SHIFT | + ring_freq << GEN6_PCODE_FREQ_RING_RATIO_SHIFT | + gpu_freq); + } +} + +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; + + val = vlv_nc_read(dev_priv, IOSF_NC_FB_GFX_FREQ_FUSE); + + rp0 = (val & FB_GFX_MAX_FREQ_FUSE_MASK) >> FB_GFX_MAX_FREQ_FUSE_SHIFT; + /* Clamp to max */ + rp0 = min_t(u32, rp0, 0xea); + + return rp0; +} + +static int valleyview_rps_rpe_freq(struct drm_i915_private *dev_priv) +{ + u32 val, rpe; + + val = vlv_nc_read(dev_priv, IOSF_NC_FB_GFX_FMAX_FUSE_LO); + rpe = (val & FB_FMAX_VMIN_FREQ_LO_MASK) >> FB_FMAX_VMIN_FREQ_LO_SHIFT; + val = vlv_nc_read(dev_priv, IOSF_NC_FB_GFX_FMAX_FUSE_HI); + rpe |= (val & FB_FMAX_VMIN_FREQ_HI_MASK) << 5; + + return rpe; +} + +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; + struct drm_i915_gem_object *pctx; + unsigned long pctx_paddr; + 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 */ + int pcbr_offset; + + pcbr_offset = (pcbr & (~4095)) - dev_priv->mm.stolen_base; + pctx = i915_gem_object_create_stolen_for_preallocated(dev_priv->dev, + pcbr_offset, + I915_GTT_OFFSET_NONE, + pctx_size); + goto out; + } + + /* + * From the Gunit register HAS: + * The Gfx driver is expected to program this register and ensure + * proper allocation within Gfx stolen memory. For example, this + * register should be programmed such than the PCBR range does not + * overlap with other ranges, such as the frame buffer, protected + * memory, or any other relevant ranges. + */ + pctx = i915_gem_object_create_stolen(dev, pctx_size); + if (!pctx) { + DRM_DEBUG("not enough stolen space for PCTX, disabling\n"); + return; + } + + pctx_paddr = dev_priv->mm.stolen_base + pctx->stolen->start; + I915_WRITE(VLV_PCBR, pctx_paddr); + +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_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); } + + /* 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); + I915_WRITE(GEN6_RP_UP_EI, 66000); + I915_WRITE(GEN6_RP_DOWN_EI, 350000); + + I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10); + + I915_WRITE(GEN6_RP_CONTROL, + GEN6_RP_MEDIA_TURBO | + GEN6_RP_MEDIA_HW_NORMAL_MODE | + GEN6_RP_MEDIA_IS_GFX | + GEN6_RP_ENABLE | + GEN6_RP_UP_BUSY_AVG | + GEN6_RP_DOWN_IDLE_CONT); + + I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 0x00280000); + I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); + I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); + + for_each_ring(ring, dev_priv, i) + I915_WRITE(RING_MAX_IDLE(ring->mmio_base), 10); + + I915_WRITE(GEN6_RC6_THRESHOLD, 0x557); + + /* allows RC6 residency counter to work */ + I915_WRITE(VLV_COUNTER_CONTROL, + _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH | + 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 | 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); + + DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & 0x10 ? "yes" : "no"); + DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val); + + dev_priv->rps.cur_freq = (val >> 8) & 0xff; + DRM_DEBUG_DRIVER("current GPU freq: %d MHz (%u)\n", + 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, dev_priv->rps.efficient_freq), + dev_priv->rps.efficient_freq); + + valleyview_set_rps(dev_priv->dev, dev_priv->rps.efficient_freq); + + gen6_enable_rps_interrupts(dev); + + gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL); } void ironlake_teardown_rc6(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - if (dev_priv->renderctx) { - i915_gem_object_unpin(dev_priv->renderctx); - drm_gem_object_unreference(&dev_priv->renderctx->base); - dev_priv->renderctx = NULL; + if (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->pwrctx) { - i915_gem_object_unpin(dev_priv->pwrctx); - drm_gem_object_unreference(&dev_priv->pwrctx->base); - dev_priv->pwrctx = NULL; + if (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; } } @@ -2634,14 +3976,14 @@ static int ironlake_setup_rc6(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - if (dev_priv->renderctx == NULL) - dev_priv->renderctx = intel_alloc_context_page(dev); - if (!dev_priv->renderctx) + if (dev_priv->ips.renderctx == NULL) + dev_priv->ips.renderctx = intel_alloc_context_page(dev); + if (!dev_priv->ips.renderctx) return -ENOMEM; - if (dev_priv->pwrctx == NULL) - dev_priv->pwrctx = intel_alloc_context_page(dev); - if (!dev_priv->pwrctx) { + if (dev_priv->ips.pwrctx == NULL) + dev_priv->ips.pwrctx = intel_alloc_context_page(dev); + if (!dev_priv->ips.pwrctx) { ironlake_teardown_rc6(dev); return -ENOMEM; } @@ -2652,7 +3994,8 @@ 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; /* rc6 disabled by default due to repeated reports of hanging during @@ -2667,6 +4010,9 @@ static void ironlake_enable_rc6(struct drm_device *dev) if (ret) return; + was_interruptible = dev_priv->mm.interruptible; + dev_priv->mm.interruptible = false; + /* * GPU can automatically power down the render unit if given a page * to save state. @@ -2674,12 +4020,13 @@ static void ironlake_enable_rc6(struct drm_device *dev) ret = intel_ring_begin(ring, 6); if (ret) { ironlake_teardown_rc6(dev); + dev_priv->mm.interruptible = was_interruptible; return; } intel_ring_emit(ring, MI_SUSPEND_FLUSH | MI_SUSPEND_FLUSH_EN); intel_ring_emit(ring, MI_SET_CONTEXT); - intel_ring_emit(ring, dev_priv->renderctx->gtt_offset | + intel_ring_emit(ring, i915_gem_obj_ggtt_offset(dev_priv->ips.renderctx) | MI_MM_SPACE_GTT | MI_SAVE_EXT_STATE_EN | MI_RESTORE_EXT_STATE_EN | @@ -2694,15 +4041,18 @@ static void ironlake_enable_rc6(struct drm_device *dev) * does an implicit flush, combined with MI_FLUSH above, it should be * safe to assume that renderctx is valid */ - ret = intel_wait_ring_idle(ring); + ret = intel_ring_idle(ring); + dev_priv->mm.interruptible = was_interruptible; if (ret) { - DRM_ERROR("failed to enable ironlake power power savings\n"); + DRM_ERROR("failed to enable ironlake power savings\n"); ironlake_teardown_rc6(dev); return; } - I915_WRITE(PWRCTXA, dev_priv->pwrctx->gtt_offset | PWRCTX_EN); + 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, GEN6_RC_CTL_RC6_ENABLE); } static unsigned long intel_pxfreq(u32 vidfreq) @@ -2790,9 +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); @@ -2821,6 +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 */ @@ -2954,7 +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; @@ -2997,7 +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); @@ -3014,7 +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); @@ -3046,9 +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); @@ -3151,7 +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; @@ -3237,6 +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; @@ -3308,39 +4664,157 @@ 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; + + /* Interrupts should be disabled already to avoid re-arming. */ + WARN_ON(dev->irq_enabled); + if (IS_IRONLAKE_M(dev)) { ironlake_disable_drps(dev); ironlake_disable_rc6(dev); - } else if (INTEL_INFO(dev)->gen >= 6 && !IS_VALLEYVIEW(dev)) { - gen6_disable_rps(dev); + } 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)) + valleyview_disable_rps(dev); + else + gen6_disable_rps(dev); + dev_priv->rps.enabled = false; + mutex_unlock(&dev_priv->rps.hw_lock); + } +} + +static void intel_gen6_powersave_work(struct work_struct *work) +{ + struct drm_i915_private *dev_priv = + container_of(work, struct drm_i915_private, + rps.delayed_resume_work.work); + struct drm_device *dev = dev_priv->dev; + + mutex_lock(&dev_priv->rps.hw_lock); + + if (IS_VALLEYVIEW(dev)) { + valleyview_enable_rps(dev); + } else if (IS_BROADWELL(dev)) { + gen8_enable_rps(dev); + __gen6_update_ring_freq(dev); + } else { + gen6_enable_rps(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) { + 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)) && !IS_VALLEYVIEW(dev)) { - gen6_enable_rps(dev); - gen6_update_ring_freq(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). + */ + 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; + + /* + * On Ibex Peak and Cougar Point, we need to disable clock + * gating for the panel power sequencer or it will fail to + * start up when no ports are active. + */ + I915_WRITE(SOUTH_DSPCLK_GATE_D, PCH_DPLSUNIT_CLOCK_GATE_DISABLE); +} + +static void g4x_disable_trickle_feed(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + int pipe; + + for_each_pipe(pipe) { + I915_WRITE(DSPCNTR(pipe), + I915_READ(DSPCNTR(pipe)) | + DISPPLANE_TRICKLE_FEED_DISABLE); + intel_flush_primary_plane(dev_priv, pipe); } } +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; - uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE; + uint32_t dspclk_gate = ILK_VRHUNIT_CLOCK_GATE_DISABLE; - /* Required for FBC */ - dspclk_gate |= DPFCUNIT_CLOCK_GATE_DISABLE | - DPFCRUNIT_CLOCK_GATE_DISABLE | - DPFDUNIT_CLOCK_GATE_DISABLE; - /* Required for CxSR */ - dspclk_gate |= DPARBUNIT_CLOCK_GATE_DISABLE; + /* + * Required for FBC + * WaFbcDisableDpfcClockGating:ilk + */ + dspclk_gate |= ILK_DPFCRUNIT_CLOCK_GATE_DISABLE | + ILK_DPFCUNIT_CLOCK_GATE_DISABLE | + ILK_DPFDUNIT_CLOCK_GATE_ENABLE; I915_WRITE(PCH_3DCGDIS0, MARIUNIT_CLOCK_GATE_DISABLE | @@ -3348,8 +4822,6 @@ static void ironlake_init_clock_gating(struct drm_device *dev) I915_WRITE(PCH_3DCGDIS1, VFMUNIT_CLOCK_GATE_DISABLE); - I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate); - /* * According to the spec the following bits should be set in * order to enable memory self-refresh @@ -3360,15 +4832,12 @@ static void ironlake_init_clock_gating(struct drm_device *dev) I915_WRITE(ILK_DISPLAY_CHICKEN2, (I915_READ(ILK_DISPLAY_CHICKEN2) | ILK_DPARB_GATE | ILK_VSDPFD_FULL)); - I915_WRITE(ILK_DSPCLK_GATE, - (I915_READ(ILK_DSPCLK_GATE) | - ILK_DPARB_CLK_GATE)); + dspclk_gate |= ILK_DPARBUNIT_CLOCK_GATE_ENABLE; 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 @@ -3378,42 +4847,121 @@ static void ironlake_init_clock_gating(struct drm_device *dev) * The bit 7,8,9 of 0x42020. */ if (IS_IRONLAKE_M(dev)) { + /* WaFbcAsynchFlipDisableFbcQueue:ilk */ I915_WRITE(ILK_DISPLAY_CHICKEN1, I915_READ(ILK_DISPLAY_CHICKEN1) | ILK_FBCQ_DIS); I915_WRITE(ILK_DISPLAY_CHICKEN2, I915_READ(ILK_DISPLAY_CHICKEN2) | ILK_DPARB_GATE); - I915_WRITE(ILK_DSPCLK_GATE, - I915_READ(ILK_DSPCLK_GATE) | - ILK_DPFC_DIS1 | - ILK_DPFC_DIS2 | - ILK_CLK_FBC); } + I915_WRITE(ILK_DSPCLK_GATE_D, dspclk_gate); + I915_WRITE(ILK_DISPLAY_CHICKEN2, I915_READ(ILK_DISPLAY_CHICKEN2) | ILK_ELPIN_409_SELECT); I915_WRITE(_3D_CHICKEN2, _3D_CHICKEN2_WM_READ_PIPELINED << 16 | _3D_CHICKEN2_WM_READ_PIPELINED); + + /* WaDisableRenderCachePipelinedFlush:ilk */ + 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); } -static void gen6_init_clock_gating(struct drm_device *dev) +static void cpt_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; int pipe; - uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE; + uint32_t val; - I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate); + /* + * On Ibex Peak and Cougar Point, we need to disable clock + * gating for the panel power sequencer or it will fail to + * start up when no ports are active. + */ + I915_WRITE(SOUTH_DSPCLK_GATE_D, PCH_DPLSUNIT_CLOCK_GATE_DISABLE | + PCH_DPLUNIT_CLOCK_GATE_DISABLE | + PCH_CPUNIT_CLOCK_GATE_DISABLE); + I915_WRITE(SOUTH_CHICKEN2, I915_READ(SOUTH_CHICKEN2) | + DPLS_EDP_PPS_FIX_DIS); + /* The below fixes the weird display corruption, a few pixels shifted + * downward, on (only) LVDS of some HP laptops with IVY. + */ + for_each_pipe(pipe) { + val = I915_READ(TRANS_CHICKEN2(pipe)); + val |= TRANS_CHICKEN2_TIMING_OVERRIDE; + val &= ~TRANS_CHICKEN2_FDI_POLARITY_REVERSED; + if (dev_priv->vbt.fdi_rx_polarity_inverted) + val |= TRANS_CHICKEN2_FDI_POLARITY_REVERSED; + val &= ~TRANS_CHICKEN2_FRAME_START_DELAY_MASK; + val &= ~TRANS_CHICKEN2_DISABLE_DEEP_COLOR_COUNTER; + val &= ~TRANS_CHICKEN2_DISABLE_DEEP_COLOR_MODESWITCH; + I915_WRITE(TRANS_CHICKEN2(pipe), val); + } + /* WADP0ClockGatingDisable */ + for_each_pipe(pipe) { + I915_WRITE(TRANS_CHICKEN1(pipe), + TRANS_CHICKEN1_DP0UNIT_GC_DISABLE); + } +} + +static void gen6_check_mch_setup(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + uint32_t tmp; + + tmp = I915_READ(MCH_SSKPD); + if ((tmp & MCH_SSKPD_WM0_MASK) != MCH_SSKPD_WM0_VAL) { + DRM_INFO("Wrong MCH_SSKPD value: 0x%08x\n", tmp); + DRM_INFO("This can cause pipe underruns and display issues.\n"); + DRM_INFO("Please upgrade your BIOS to fix this.\n"); + } +} + +static void gen6_init_clock_gating(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + uint32_t dspclk_gate = ILK_VRHUNIT_CLOCK_GATE_DISABLE; + + I915_WRITE(ILK_DSPCLK_GATE_D, dspclk_gate); I915_WRITE(ILK_DISPLAY_CHICKEN2, I915_READ(ILK_DISPLAY_CHICKEN2) | ILK_ELPIN_409_SELECT); - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); + /* WaDisableHiZPlanesWhenMSAAEnabled:snb */ + I915_WRITE(_3D_CHICKEN, + _MASKED_BIT_ENABLE(_3D_CHICKEN_HIZ_PLANE_DISABLE_MSAA_4X_SNB)); + + /* WaSetupGtModeTdRowDispatch:snb */ + if (IS_SNB_GT1(dev)) + I915_WRITE(GEN6_GT_MODE, + _MASKED_BIT_ENABLE(GEN6_TD_FOUR_ROW_DISPATCH_DISABLE)); + + /* 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)); @@ -3433,17 +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 and - * WaDisableRCPBUnitClockGating. + * 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 @@ -3453,6 +5008,8 @@ static void gen6_init_clock_gating(struct drm_device *dev) * The bit5 and bit7 of 0x42020 * The bit14 of 0x70180 * The bit14 of 0x71180 + * + * WaFbcAsynchFlipDisableFbcQueue:snb */ I915_WRITE(ILK_DISPLAY_CHICKEN1, I915_READ(ILK_DISPLAY_CHICKEN1) | @@ -3460,31 +5017,28 @@ static void gen6_init_clock_gating(struct drm_device *dev) I915_WRITE(ILK_DISPLAY_CHICKEN2, I915_READ(ILK_DISPLAY_CHICKEN2) | ILK_DPARB_GATE | ILK_VSDPFD_FULL); - I915_WRITE(ILK_DSPCLK_GATE, - I915_READ(ILK_DSPCLK_GATE) | - ILK_DPARB_CLK_GATE | - ILK_DPFD_CLK_GATE); + I915_WRITE(ILK_DSPCLK_GATE_D, + I915_READ(ILK_DSPCLK_GATE_D) | + ILK_DPARBUNIT_CLOCK_GATE_ENABLE | + ILK_DPFDUNIT_CLOCK_GATE_ENABLE); - I915_WRITE(GEN6_MBCTL, I915_READ(GEN6_MBCTL) | - GEN6_MBCTL_ENABLE_BOOT_FETCH); + g4x_disable_trickle_feed(dev); - for_each_pipe(pipe) { - I915_WRITE(DSPCNTR(pipe), - I915_READ(DSPCNTR(pipe)) | - DISPPLANE_TRICKLE_FEED_DISABLE); - intel_flush_display_plane(dev_priv, pipe); - } + cpt_init_clock_gating(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)); + gen6_check_mch_setup(dev); } 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; @@ -3493,223 +5047,430 @@ static void gen7_setup_fixed_func_scheduler(struct drm_i915_private *dev_priv) I915_WRITE(GEN7_FF_THREAD_MODE, reg); } -static void haswell_init_clock_gating(struct drm_device *dev) +static void lpt_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - int pipe; - uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE; - I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate); + /* + * TODO: this bit should only be enabled when really needed, then + * disabled when not needed anymore in order to save power. + */ + if (dev_priv->pch_id == INTEL_PCH_LPT_LP_DEVICE_ID_TYPE) + I915_WRITE(SOUTH_DSPCLK_GATE_D, + I915_READ(SOUTH_DSPCLK_GATE_D) | + PCH_LP_PARTITION_LEVEL_DISABLE); + + /* WADPOClockGatingDisable:hsw */ + I915_WRITE(_TRANSA_CHICKEN1, + I915_READ(_TRANSA_CHICKEN1) | + TRANS_CHICKEN1_DP0UNIT_GC_DISABLE); +} + +static void lpt_suspend_hw(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + if (dev_priv->pch_id == INTEL_PCH_LPT_LP_DEVICE_ID_TYPE) { + uint32_t val = I915_READ(SOUTH_DSPCLK_GATE_D); + + val &= ~PCH_LP_PARTITION_LEVEL_DISABLE; + I915_WRITE(SOUTH_DSPCLK_GATE_D, val); + } +} + +static void gen8_init_clock_gating(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + enum pipe pipe; I915_WRITE(WM3_LP_ILK, 0); I915_WRITE(WM2_LP_ILK, 0); I915_WRITE(WM1_LP_ILK, 0); - /* According to the spec, bit 13 (RCZUNIT) must be set on IVB. - * This implements the WaDisableRCZUnitClockGating workaround. + /* FIXME(BDW): Check all the w/a, some might only apply to + * pre-production hw. */ + + /* 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(GEN6_UCGCTL2, GEN6_RCZUNIT_CLOCK_GATE_DISABLE); + I915_WRITE(HALF_SLICE_CHICKEN3, + _MASKED_BIT_ENABLE(GEN8_CENTROID_PIXEL_OPT_DIS)); + I915_WRITE(HALF_SLICE_CHICKEN3, + _MASKED_BIT_ENABLE(GEN8_SAMPLER_POWER_BYPASS_DIS)); + I915_WRITE(GAMTARBMODE, _MASKED_BIT_ENABLE(ARB_MODE_BWGTLB_DISABLE)); - I915_WRITE(ILK_DSPCLK_GATE, IVB_VRHUNIT_CLK_GATE); + I915_WRITE(_3D_CHICKEN3, + _3D_CHICKEN_SDE_LIMIT_FIFO_POLY_DEPTH(2)); - I915_WRITE(IVB_CHICKEN3, - CHICKEN3_DGMG_REQ_OUT_FIX_DISABLE | - CHICKEN3_DGMG_DONE_FIX_DISABLE); + I915_WRITE(COMMON_SLICE_CHICKEN2, + _MASKED_BIT_ENABLE(GEN8_CSC2_SBE_VUE_CACHE_CONSERVATIVE)); - /* Apply the WaDisableRHWOOptimizationForRenderHang workaround. */ - I915_WRITE(GEN7_COMMON_SLICE_CHICKEN1, - GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC); + I915_WRITE(GEN7_HALF_SLICE_CHICKEN1, + _MASKED_BIT_ENABLE(GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE)); - /* WaApplyL3ControlAndL3ChickenMode requires those two on Ivy Bridge */ - I915_WRITE(GEN7_L3CNTLREG1, - GEN7_WA_FOR_GEN7_L3_CONTROL); - I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER, - GEN7_WA_L3_CHICKEN_MODE); + /* 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:bdw */ + I915_WRITE(CHICKEN_PAR1_1, + I915_READ(CHICKEN_PAR1_1) | DPA_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); + } + + /* 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)); + + /* 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)); + + /* + * 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); + + I915_WRITE(GEN6_RC_SLEEP_PSMI_CONTROL, + _MASKED_BIT_ENABLE(GEN8_RC_SEMA_IDLE_MSG_DISABLE)); + + /* 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; - /* This is required by WaCatErrorRejectionIssue */ + 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); + I915_WRITE(HSW_ROW_CHICKEN3, + _MASKED_BIT_ENABLE(HSW_ROW_CHICKEN3_L3_GLOBAL_ATOMICS_DISABLE)); + + /* This is required by WaCatErrorRejectionIssue:hsw */ I915_WRITE(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG, I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) | GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB); - for_each_pipe(pipe) { - I915_WRITE(DSPCNTR(pipe), - I915_READ(DSPCNTR(pipe)) | - DISPPLANE_TRICKLE_FEED_DISABLE); - intel_flush_display_plane(dev_priv, pipe); - } + /* WaVSRefCountFullforceMissDisable:hsw */ + I915_WRITE(GEN7_FF_THREAD_MODE, + I915_READ(GEN7_FF_THREAD_MODE) & ~GEN7_FF_VS_REF_CNT_FFME); - gen7_setup_fixed_func_scheduler(dev_priv); + /* 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 */ + /* WaDisable4x2SubspanOptimization:hsw */ I915_WRITE(CACHE_MODE_1, _MASKED_BIT_ENABLE(PIXEL_SUBSPAN_COLLECT_OPT_DISABLE)); - /* XXX: This is a workaround for early silicon revisions and should be - * removed later. + /* + * 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(WM_DBG, - I915_READ(WM_DBG) | - WM_DBG_DISALLOW_MULTIPLE_LP | - WM_DBG_DISALLOW_SPRITE | - WM_DBG_DISALLOW_MAXFIFO); + 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); + /* WaRsPkgCStateDisplayPMReq:hsw */ + I915_WRITE(CHICKEN_PAR1_1, + I915_READ(CHICKEN_PAR1_1) | FORCE_ARB_IDLE_PLANES); + + lpt_init_clock_gating(dev); } static void ivybridge_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - int pipe; - uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE; uint32_t snpcr; - I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate); + ilk_init_lp_watermarks(dev); - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); + I915_WRITE(ILK_DSPCLK_GATE_D, ILK_VRHUNIT_CLOCK_GATE_DISABLE); - I915_WRITE(ILK_DSPCLK_GATE, IVB_VRHUNIT_CLK_GATE); + /* WaDisableEarlyCull:ivb */ + I915_WRITE(_3D_CHICKEN3, + _MASKED_BIT_ENABLE(_3D_CHICKEN_SF_DISABLE_OBJEND_CULL)); + /* WaDisableBackToBackFlipFix:ivb */ I915_WRITE(IVB_CHICKEN3, CHICKEN3_DGMG_REQ_OUT_FIX_DISABLE | CHICKEN3_DGMG_DONE_FIX_DISABLE); - /* Apply the WaDisableRHWOOptimizationForRenderHang workaround. */ + /* WaDisablePSDDualDispatchEnable:ivb */ + if (IS_IVB_GT1(dev)) + I915_WRITE(GEN7_HALF_SLICE_CHICKEN1, + _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, GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC); - /* WaApplyL3ControlAndL3ChickenMode requires those two on Ivy Bridge */ + /* WaApplyL3ControlAndL3ChickenMode:ivb */ I915_WRITE(GEN7_L3CNTLREG1, GEN7_WA_FOR_GEN7_L3_CONTROL); I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER, - GEN7_WA_L3_CHICKEN_MODE); + GEN7_WA_L3_CHICKEN_MODE); + if (IS_IVB_GT1(dev)) + I915_WRITE(GEN7_ROW_CHICKEN2, + _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE)); + 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)); + } - /* 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. - * + /* WaForceL3Serialization:ivb */ + I915_WRITE(GEN7_L3SQCREG4, I915_READ(GEN7_L3SQCREG4) & + ~L3SQ_URB_READ_CAM_MATCH_DISABLE); + + /* * According to the spec, bit 13 (RCZUNIT) must be set on IVB. - * This implements the WaDisableRCZUnitClockGating workaround. + * 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 */ + /* This is required by WaCatErrorRejectionIssue:ivb */ I915_WRITE(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG, I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) | GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB); - for_each_pipe(pipe) { - I915_WRITE(DSPCNTR(pipe), - I915_READ(DSPCNTR(pipe)) | - DISPPLANE_TRICKLE_FEED_DISABLE); - intel_flush_display_plane(dev_priv, pipe); - } - - I915_WRITE(GEN6_MBCTL, I915_READ(GEN6_MBCTL) | - GEN6_MBCTL_ENABLE_BOOT_FETCH); + g4x_disable_trickle_feed(dev); gen7_setup_fixed_func_scheduler(dev_priv); - /* WaDisable4x2SubspanOptimization */ + 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; I915_WRITE(GEN6_MBCUNIT_SNPCR, snpcr); + + if (!HAS_PCH_NOP(dev)) + cpt_init_clock_gating(dev); + + gen6_check_mch_setup(dev); } static void valleyview_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - int pipe; - uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE; + u32 val; - I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate); + 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); - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); + 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); - I915_WRITE(ILK_DSPCLK_GATE, IVB_VRHUNIT_CLK_GATE); + /* WaDisableEarlyCull:vlv */ + I915_WRITE(_3D_CHICKEN3, + _MASKED_BIT_ENABLE(_3D_CHICKEN_SF_DISABLE_OBJEND_CULL)); + /* WaDisableBackToBackFlipFix:vlv */ I915_WRITE(IVB_CHICKEN3, CHICKEN3_DGMG_REQ_OUT_FIX_DISABLE | CHICKEN3_DGMG_DONE_FIX_DISABLE); - /* Apply the WaDisableRHWOOptimizationForRenderHang workaround. */ - I915_WRITE(GEN7_COMMON_SLICE_CHICKEN1, - GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC); + /* WaPsdDispatchEnable:vlv */ + /* WaDisablePSDDualDispatchEnable:vlv */ + I915_WRITE(GEN7_HALF_SLICE_CHICKEN1, + _MASKED_BIT_ENABLE(GEN7_MAX_PS_THREAD_DEP | + GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE)); + + /* WaDisable_RenderCache_OperationalFlush:vlv */ + I915_WRITE(CACHE_MODE_0_GEN7, _MASKED_BIT_DISABLE(RC_OP_FLUSH_ENABLE)); - /* WaApplyL3ControlAndL3ChickenMode requires those two on Ivy Bridge */ - I915_WRITE(GEN7_L3CNTLREG1, GEN7_WA_FOR_GEN7_L3_CONTROL); - I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER, GEN7_WA_L3_CHICKEN_MODE); + /* WaForceL3Serialization:vlv */ + I915_WRITE(GEN7_L3SQCREG4, I915_READ(GEN7_L3SQCREG4) & + ~L3SQ_URB_READ_CAM_MATCH_DISABLE); - /* This is required by WaCatErrorRejectionIssue */ + /* WaDisableDopClockGating:vlv */ + I915_WRITE(GEN7_ROW_CHICKEN2, + _MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE)); + + /* This is required by WaCatErrorRejectionIssue:vlv */ I915_WRITE(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG, I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) | GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB); - I915_WRITE(GEN6_MBCTL, I915_READ(GEN6_MBCTL) | - GEN6_MBCTL_ENABLE_BOOT_FETCH); - + gen7_setup_fixed_func_scheduler(dev_priv); - /* 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 workaround. - * - * Also apply WaDisableVDSUnitClockGating and - * WaDisableRCPBUnitClockGating. + * This implements the WaDisableRCZUnitClockGating:vlv workaround. */ 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); - for_each_pipe(pipe) { - I915_WRITE(DSPCNTR(pipe), - I915_READ(DSPCNTR(pipe)) | - DISPPLANE_TRICKLE_FEED_DISABLE); - intel_flush_display_plane(dev_priv, pipe); - } + 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)); /* - * On ValleyView, the GUnit needs to signal the GT - * when flip and other events complete. So enable - * all the GUnit->GT interrupts here + * WaIncreaseL3CreditsForVLVB0:vlv + * This is the hardware default actually. */ - I915_WRITE(VLV_DPFLIPSTAT, PIPEB_LINE_COMPARE_INT_EN | - PIPEB_HLINE_INT_EN | PIPEB_VBLANK_INT_EN | - SPRITED_FLIPDONE_INT_EN | SPRITEC_FLIPDONE_INT_EN | - PLANEB_FLIPDONE_INT_EN | PIPEA_LINE_COMPARE_INT_EN | - PIPEA_HLINE_INT_EN | PIPEA_VBLANK_INT_EN | - SPRITEB_FLIPDONE_INT_EN | SPRITEA_FLIPDONE_INT_EN | - PLANEA_FLIPDONE_INT_EN); + 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, 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); + + /* 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) @@ -3728,6 +5489,15 @@ static void g4x_init_clock_gating(struct drm_device *dev) if (IS_GM45(dev)) dspclk_gate |= DSSUNIT_CLOCK_GATE_DISABLE; I915_WRITE(DSPCLK_GATE_D, dspclk_gate); + + /* WaDisableRenderCachePipelinedFlush */ + 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); } static void crestline_init_clock_gating(struct drm_device *dev) @@ -3739,6 +5509,11 @@ static void crestline_init_clock_gating(struct drm_device *dev) I915_WRITE(DSPCLK_GATE_D, 0); I915_WRITE(RAMCLK_GATE_D, 0); 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) @@ -3751,6 +5526,11 @@ static void broadwater_init_clock_gating(struct drm_device *dev) I965_ISC_CLOCK_GATE_DISABLE | I965_FBC_CLOCK_GATE_DISABLE); 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) @@ -3767,6 +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) @@ -3774,6 +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) @@ -3783,75 +5573,824 @@ static void i830_init_clock_gating(struct drm_device *dev) I915_WRITE(DSPCLK_GATE_D, OVRUNIT_CLOCK_GATE_DISABLE); } -static void ibx_init_clock_gating(struct drm_device *dev) +void intel_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; + dev_priv->display.init_clock_gating(dev); +} + +void intel_suspend_hw(struct drm_device *dev) +{ + if (HAS_PCH_LPT(dev)) + lpt_suspend_hw(dev); +} + +#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)) + +#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. + */ +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 i915_power_domains *power_domains; + bool ret; + + power_domains = &dev_priv->power_domains; + + mutex_lock(&power_domains->lock); + ret = intel_display_power_enabled_unlocked(dev_priv, domain); + mutex_unlock(&power_domains->lock); + + return ret; +} + +/* + * 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_device *dev = dev_priv->dev; + unsigned long irqflags; + /* - * On Ibex Peak and Cougar Point, we need to disable clock - * gating for the panel power sequencer or it will fail to - * start up when no ports are active. + * 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. */ - I915_WRITE(SOUTH_DSPCLK_GATE_D, PCH_DPLSUNIT_CLOCK_GATE_DISABLE); + 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 cpt_init_clock_gating(struct drm_device *dev) +static void hsw_set_power_well(struct drm_i915_private *dev_priv, + struct i915_power_well *power_well, bool enable) { - struct drm_i915_private *dev_priv = dev->dev_private; - int pipe; + bool is_enabled, enable_requested; + uint32_t tmp; + + tmp = I915_READ(HSW_PWR_WELL_DRIVER); + is_enabled = tmp & HSW_PWR_WELL_STATE_ENABLED; + enable_requested = tmp & HSW_PWR_WELL_ENABLE_REQUEST; + + if (enable) { + if (!enable_requested) + I915_WRITE(HSW_PWR_WELL_DRIVER, + HSW_PWR_WELL_ENABLE_REQUEST); + + if (!is_enabled) { + DRM_DEBUG_KMS("Enabling power well\n"); + if (wait_for((I915_READ(HSW_PWR_WELL_DRIVER) & + HSW_PWR_WELL_STATE_ENABLED), 20)) + DRM_ERROR("Timeout enabling power well\n"); + } + + hsw_power_well_post_enable(dev_priv); + } else { + if (enable_requested) { + 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); /* - * On Ibex Peak and Cougar Point, we need to disable clock - * gating for the panel power sequencer or it will fail to - * start up when no ports are active. + * We're taking over the BIOS, so clear any requests made by it since + * the driver is in charge now. */ - I915_WRITE(SOUTH_DSPCLK_GATE_D, PCH_DPLSUNIT_CLOCK_GATE_DISABLE); - I915_WRITE(SOUTH_CHICKEN2, I915_READ(SOUTH_CHICKEN2) | - DPLS_EDP_PPS_FIX_DIS); - /* Without this, mode sets may fail silently on FDI */ - for_each_pipe(pipe) - I915_WRITE(TRANS_CHICKEN2(pipe), TRANS_AUTOTRAIN_GEN_STALL_DIS); + if (I915_READ(HSW_PWR_WELL_BIOS) & HSW_PWR_WELL_ENABLE_REQUEST) + I915_WRITE(HSW_PWR_WELL_BIOS, 0); } -void intel_init_clock_gating(struct drm_device *dev) +static void hsw_power_well_enable(struct drm_i915_private *dev_priv, + struct i915_power_well *power_well) { - struct drm_i915_private *dev_priv = dev->dev_private; + hsw_set_power_well(dev_priv, power_well, true); +} - dev_priv->display.init_clock_gating(dev); +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) { + /* + * Enable the CRI clock source so we can get at the + * display and the reference clock for VGA + * hotplug / manual detection. + */ + 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 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) +{ + vlv_set_power_well(dev_priv, power_well, power_well->count > 0); +} + +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 vlv_power_well_disable(struct drm_i915_private *dev_priv, + struct i915_power_well *power_well) +{ + vlv_set_power_well(dev_priv, power_well, false); +} + +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; +} + +static void vlv_display_power_well_enable(struct drm_i915_private *dev_priv, + struct i915_power_well *power_well) +{ + 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); + + /* + * 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; + + 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); + + 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_i915_private *dev_priv, + enum intel_display_power_domain domain) +{ + struct i915_power_domains *power_domains; + struct i915_power_well *power_well; + int i; + + power_domains = &dev_priv->power_domains; + + mutex_lock(&power_domains->lock); + + 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); - if (dev_priv->display.init_pch_clock_gating) - dev_priv->display.init_pch_clock_gating(dev); + intel_runtime_pm_put(dev_priv); } -/* Starting with Haswell, we have different power wells for - * different parts of the GPU. This attempts to enable them all. +static struct i915_power_domains *hsw_pwr; + +/* Display audio driver power well request */ +int i915_request_power_well(void) +{ + struct drm_i915_private *dev_priv; + + if (!hsw_pwr) + return -ENODEV; + + dev_priv = container_of(hsw_pwr, struct drm_i915_private, + power_domains); + intel_display_power_get(dev_priv, POWER_DOMAIN_AUDIO); + return 0; +} +EXPORT_SYMBOL_GPL(i915_request_power_well); + +/* Display audio driver power well release */ +int i915_release_power_well(void) +{ + struct drm_i915_private *dev_priv; + + if (!hsw_pwr) + return -ENODEV; + + dev_priv = container_of(hsw_pwr, struct drm_i915_private, + power_domains); + intel_display_power_put(dev_priv, POWER_DOMAIN_AUDIO); + return 0; +} +EXPORT_SYMBOL_GPL(i915_release_power_well); + +/* + * 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. */ -void intel_init_power_wells(struct drm_device *dev) +int i915_get_cdclk_freq(void) { - struct drm_i915_private *dev_priv = dev->dev_private; - unsigned long power_wells[] = { - HSW_PWR_WELL_CTL1, - HSW_PWR_WELL_CTL2, - HSW_PWR_WELL_CTL4 - }; + 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 i915_power_domains *power_domains = &dev_priv->power_domains; + + mutex_init(&power_domains->lock); + + /* + * 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_i915_private *dev_priv) +{ + hsw_pwr = NULL; +} + +static void intel_power_domains_resume(struct drm_i915_private *dev_priv) +{ + struct i915_power_domains *power_domains = &dev_priv->power_domains; + struct i915_power_well *power_well; int i; - if (!IS_HASWELL(dev)) + 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); +} + +void intel_power_domains_init_hw(struct drm_i915_private *dev_priv) +{ + struct i915_power_domains *power_domains = &dev_priv->power_domains; + + 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; +} + +void intel_aux_display_runtime_get(struct drm_i915_private *dev_priv) +{ + intel_runtime_pm_get(dev_priv); +} + +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; - mutex_lock(&dev->struct_mutex); + pm_runtime_get_sync(device); + WARN(dev_priv->pm.suspended, "Device still suspended.\n"); +} - for (i = 0; i < ARRAY_SIZE(power_wells); i++) { - int well = I915_READ(power_wells[i]); +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 ((well & HSW_PWR_WELL_STATE) == 0) { - I915_WRITE(power_wells[i], well & HSW_PWR_WELL_ENABLE); - if (wait_for(I915_READ(power_wells[i] & HSW_PWR_WELL_STATE), 20)) - DRM_ERROR("Error enabling power well %lx\n", power_wells[i]); - } + if (!HAS_RUNTIME_PM(dev)) + return; + + WARN(dev_priv->pm.suspended, "Getting nosync-ref while suspended.\n"); + pm_runtime_get_noresume(device); +} + +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) +{ + 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; } - mutex_unlock(&dev->struct_mutex); + 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_fini_runtime_pm(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; + + 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 */ @@ -3859,8 +6398,12 @@ 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; + 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; @@ -3868,12 +6411,14 @@ void intel_init_pm(struct drm_device *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 */ @@ -3884,54 +6429,33 @@ void intel_init_pm(struct drm_device *dev) /* For FIFO watermark updates */ if (HAS_PCH_SPLIT(dev)) { - if (HAS_PCH_IBX(dev)) - dev_priv->display.init_pch_clock_gating = ibx_init_clock_gating; - else if (HAS_PCH_CPT(dev)) - dev_priv->display.init_pch_clock_gating = cpt_init_clock_gating; - - if (IS_GEN5(dev)) { - if (I915_READ(MLTR_ILK) & ILK_SRLT_MASK) - 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 (SNB_READ_WM0_LATENCY()) { - 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)) { - /* FIXME: detect B0+ stepping and use auto training */ - if (SNB_READ_WM0_LATENCY()) { - 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_IVYBRIDGE(dev)) dev_priv->display.init_clock_gating = ivybridge_init_clock_gating; - } else if (IS_HASWELL(dev)) { - if (SNB_READ_WM0_LATENCY()) { - dev_priv->display.update_wm = sandybridge_update_wm; - dev_priv->display.update_sprite_wm = sandybridge_update_sprite_wm; - dev_priv->display.update_linetime_wm = haswell_update_linetime_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 - dev_priv->display.update_wm = NULL; + else if (INTEL_INFO(dev)->gen == 8) + dev_priv->display.init_clock_gating = gen8_init_clock_gating; + } 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 = @@ -3965,218 +6489,124 @@ 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; - } -} - -static void __gen6_gt_wait_for_thread_c0(struct drm_i915_private *dev_priv) -{ - u32 gt_thread_status_mask; - - if (IS_HASWELL(dev_priv->dev)) - gt_thread_status_mask = GEN6_GT_THREAD_STATUS_CORE_MASK_HSW; - else - gt_thread_status_mask = GEN6_GT_THREAD_STATUS_CORE_MASK; - - /* w/a for a sporadic read returning 0 by waiting for the GT - * thread to wake up. - */ - if (wait_for_atomic_us((I915_READ_NOTRACE(GEN6_GT_THREAD_STATUS_REG) & gt_thread_status_mask) == 0, 500)) - DRM_ERROR("GT thread status wait timed out\n"); -} - -static void __gen6_gt_force_wake_get(struct drm_i915_private *dev_priv) -{ - u32 forcewake_ack; - - if (IS_HASWELL(dev_priv->dev)) - forcewake_ack = FORCEWAKE_ACK_HSW; - else - forcewake_ack = FORCEWAKE_ACK; - - if (wait_for_atomic((I915_READ_NOTRACE(forcewake_ack) & 1) == 0, - FORCEWAKE_ACK_TIMEOUT_MS)) - DRM_ERROR("Timed out waiting for forcewake old ack to clear.\n"); - - I915_WRITE_NOTRACE(FORCEWAKE, 1); - POSTING_READ(ECOBUS); /* something from same cacheline, but !FORCEWAKE */ - - if (wait_for_atomic((I915_READ_NOTRACE(forcewake_ack) & 1), - FORCEWAKE_ACK_TIMEOUT_MS)) - DRM_ERROR("Timed out waiting for forcewake to ack request.\n"); + } - __gen6_gt_wait_for_thread_c0(dev_priv); + 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"); + } } -static void __gen6_gt_force_wake_mt_get(struct drm_i915_private *dev_priv) +int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u8 mbox, u32 *val) { - u32 forcewake_ack; + WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); - if (IS_HASWELL(dev_priv->dev)) - forcewake_ack = FORCEWAKE_ACK_HSW; - else - forcewake_ack = FORCEWAKE_MT_ACK; + if (I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) { + DRM_DEBUG_DRIVER("warning: pcode (read) mailbox access failed\n"); + return -EAGAIN; + } - if (wait_for_atomic((I915_READ_NOTRACE(forcewake_ack) & 1) == 0, - FORCEWAKE_ACK_TIMEOUT_MS)) - DRM_ERROR("Timed out waiting for forcewake old ack to clear.\n"); + I915_WRITE(GEN6_PCODE_DATA, *val); + I915_WRITE(GEN6_PCODE_MAILBOX, GEN6_PCODE_READY | mbox); - I915_WRITE_NOTRACE(FORCEWAKE_MT, _MASKED_BIT_ENABLE(1)); - POSTING_READ(ECOBUS); /* something from same cacheline, but !FORCEWAKE */ + if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0, + 500)) { + DRM_ERROR("timeout waiting for pcode read (%d) to finish\n", mbox); + return -ETIMEDOUT; + } - if (wait_for_atomic((I915_READ_NOTRACE(forcewake_ack) & 1), - FORCEWAKE_ACK_TIMEOUT_MS)) - DRM_ERROR("Timed out waiting for forcewake to ack request.\n"); + *val = I915_READ(GEN6_PCODE_DATA); + I915_WRITE(GEN6_PCODE_DATA, 0); - __gen6_gt_wait_for_thread_c0(dev_priv); + return 0; } -/* - * Generally this is called implicitly by the register read function. However, - * if some sequence requires the GT to not power down then this function should - * 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) +int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val) { - unsigned long irqflags; + WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); - spin_lock_irqsave(&dev_priv->gt_lock, irqflags); - if (dev_priv->forcewake_count++ == 0) - dev_priv->gt.force_wake_get(dev_priv); - spin_unlock_irqrestore(&dev_priv->gt_lock, irqflags); -} - -void gen6_gt_check_fifodbg(struct drm_i915_private *dev_priv) -{ - u32 gtfifodbg; - gtfifodbg = I915_READ_NOTRACE(GTFIFODBG); - if (WARN(gtfifodbg & GT_FIFO_CPU_ERROR_MASK, - "MMIO read or write has been dropped %x\n", gtfifodbg)) - I915_WRITE_NOTRACE(GTFIFODBG, GT_FIFO_CPU_ERROR_MASK); -} + if (I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) { + DRM_DEBUG_DRIVER("warning: pcode (write) mailbox access failed\n"); + return -EAGAIN; + } -static void __gen6_gt_force_wake_put(struct drm_i915_private *dev_priv) -{ - I915_WRITE_NOTRACE(FORCEWAKE, 0); - /* gen6_gt_check_fifodbg doubles as the POSTING_READ */ - gen6_gt_check_fifodbg(dev_priv); -} + I915_WRITE(GEN6_PCODE_DATA, val); + I915_WRITE(GEN6_PCODE_MAILBOX, GEN6_PCODE_READY | mbox); -static void __gen6_gt_force_wake_mt_put(struct drm_i915_private *dev_priv) -{ - I915_WRITE_NOTRACE(FORCEWAKE_MT, _MASKED_BIT_DISABLE(1)); - /* gen6_gt_check_fifodbg doubles as the POSTING_READ */ - gen6_gt_check_fifodbg(dev_priv); -} + if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0, + 500)) { + DRM_ERROR("timeout waiting for pcode write (%d) to finish\n", mbox); + return -ETIMEDOUT; + } -/* - * see gen6_gt_force_wake_get() - */ -void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv) -{ - unsigned long irqflags; + I915_WRITE(GEN6_PCODE_DATA, 0); - spin_lock_irqsave(&dev_priv->gt_lock, irqflags); - if (--dev_priv->forcewake_count == 0) - dev_priv->gt.force_wake_put(dev_priv); - spin_unlock_irqrestore(&dev_priv->gt_lock, irqflags); + return 0; } -int __gen6_gt_wait_for_fifo(struct drm_i915_private *dev_priv) +int vlv_gpu_freq(struct drm_i915_private *dev_priv, int val) { - int ret = 0; + int div; - if (dev_priv->gt_fifo_count < GT_FIFO_NUM_RESERVED_ENTRIES) { - int loop = 500; - u32 fifo = I915_READ_NOTRACE(GT_FIFO_FREE_ENTRIES); - while (fifo <= GT_FIFO_NUM_RESERVED_ENTRIES && loop--) { - udelay(10); - fifo = I915_READ_NOTRACE(GT_FIFO_FREE_ENTRIES); - } - if (WARN_ON(loop < 0 && fifo <= GT_FIFO_NUM_RESERVED_ENTRIES)) - ++ret; - dev_priv->gt_fifo_count = fifo; + /* 4 x czclk */ + switch (dev_priv->mem_freq) { + case 800: + div = 10; + break; + case 1066: + div = 12; + break; + case 1333: + div = 16; + break; + default: + return -1; } - dev_priv->gt_fifo_count--; - return ret; + return DIV_ROUND_CLOSEST(dev_priv->mem_freq * (val + 6 - 0xbd), 4 * div); } -static void vlv_force_wake_get(struct drm_i915_private *dev_priv) +int vlv_freq_opcode(struct drm_i915_private *dev_priv, int val) { - if (wait_for_atomic((I915_READ_NOTRACE(FORCEWAKE_ACK_VLV) & 1) == 0, - FORCEWAKE_ACK_TIMEOUT_MS)) - DRM_ERROR("Timed out waiting for forcewake old ack to clear.\n"); + int mul; - I915_WRITE_NOTRACE(FORCEWAKE_VLV, _MASKED_BIT_ENABLE(1)); - - if (wait_for_atomic((I915_READ_NOTRACE(FORCEWAKE_ACK_VLV) & 1), - FORCEWAKE_ACK_TIMEOUT_MS)) - DRM_ERROR("Timed out waiting for forcewake to ack request.\n"); - - __gen6_gt_wait_for_thread_c0(dev_priv); -} + /* 4 x czclk */ + switch (dev_priv->mem_freq) { + case 800: + mul = 10; + break; + case 1066: + mul = 12; + break; + case 1333: + mul = 16; + break; + default: + return -1; + } -static void vlv_force_wake_put(struct drm_i915_private *dev_priv) -{ - I915_WRITE_NOTRACE(FORCEWAKE_VLV, _MASKED_BIT_DISABLE(1)); - /* The below doubles as a POSTING_READ */ - gen6_gt_check_fifodbg(dev_priv); + return DIV_ROUND_CLOSEST(4 * mul * val, dev_priv->mem_freq) + 0xbd - 6; } -void intel_gt_init(struct drm_device *dev) +void intel_pm_setup(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - spin_lock_init(&dev_priv->gt_lock); + mutex_init(&dev_priv->rps.hw_lock); - if (IS_VALLEYVIEW(dev)) { - dev_priv->gt.force_wake_get = vlv_force_wake_get; - dev_priv->gt.force_wake_put = vlv_force_wake_put; - } else if (INTEL_INFO(dev)->gen >= 6) { - dev_priv->gt.force_wake_get = __gen6_gt_force_wake_get; - dev_priv->gt.force_wake_put = __gen6_gt_force_wake_put; - - /* IVB configs may use multi-threaded forcewake */ - if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev)) { - u32 ecobus; - - /* A small trick here - if the bios hasn't configured - * MT forcewake, and if the device is in RC6, then - * force_wake_mt_get will not wake the device and the - * ECOBUS read will return zero. Which will be - * (correctly) interpreted by the test below as MT - * forcewake being disabled. - */ - mutex_lock(&dev->struct_mutex); - __gen6_gt_force_wake_mt_get(dev_priv); - ecobus = I915_READ_NOTRACE(ECOBUS); - __gen6_gt_force_wake_mt_put(dev_priv); - mutex_unlock(&dev->struct_mutex); - - if (ecobus & FORCEWAKE_MT_ENABLE) { - DRM_DEBUG_KMS("Using MT version of forcewake\n"); - dev_priv->gt.force_wake_get = - __gen6_gt_force_wake_mt_get; - dev_priv->gt.force_wake_put = - __gen6_gt_force_wake_mt_put; - } - } - } -} + INIT_DELAYED_WORK(&dev_priv->rps.delayed_resume_work, + intel_gen6_powersave_work); + dev_priv->pm.suspended = false; + dev_priv->pm.irqs_disabled = false; +} |