From 85b1d7b3f475c6425707b2e064f9c0f54058019d Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Mon, 4 Nov 2013 11:52:45 -0800 Subject: drm/i915: move VLV DDR freq fetch into init_clock_gating We don't want it delayed with the RPS work. Signed-off-by: Jesse Barnes Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 31 ++++++++++++++++++------------- 1 file changed, 18 insertions(+), 13 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 09ac9e79830f..13afd54ca6e9 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4064,19 +4064,6 @@ static void valleyview_enable_rps(struct drm_device *dev) I915_WRITE(GEN6_RC_CONTROL, rc6_mode); val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS); - switch ((val >> 6) & 3) { - case 0: - case 1: - dev_priv->mem_freq = 800; - break; - case 2: - dev_priv->mem_freq = 1066; - break; - case 3: - dev_priv->mem_freq = 1333; - break; - } - DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq); DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & 0x10 ? "yes" : "no"); DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val); @@ -5325,6 +5312,24 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) static void valleyview_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; + u32 val; + + mutex_lock(&dev_priv->rps.hw_lock); + val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS); + mutex_unlock(&dev_priv->rps.hw_lock); + switch ((val >> 6) & 3) { + case 0: + case 1: + dev_priv->mem_freq = 800; + break; + case 2: + dev_priv->mem_freq = 1066; + break; + case 3: + dev_priv->mem_freq = 1333; + break; + } + DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq); I915_WRITE(DSPCLK_GATE_D, VRHUNIT_CLOCK_GATE_DISABLE); -- cgit v1.2.3 From f64a28a7c5ab2fc342326de9e126acf3cc0f91d6 Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Mon, 4 Nov 2013 16:07:00 -0800 Subject: drm/i915/vlv: fixup DDR freq detection per Punit spec MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Either the docs were wrong or the values have changed since the old days before we had wheels. Reported-by: Ville Syrjälä Signed-off-by: Jesse Barnes Reviewed-by: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 13afd54ca6e9..a5778e59cc15 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5319,15 +5319,22 @@ static void valleyview_init_clock_gating(struct drm_device *dev) mutex_unlock(&dev_priv->rps.hw_lock); switch ((val >> 6) & 3) { case 0: - case 1: dev_priv->mem_freq = 800; break; - case 2: + case 1: dev_priv->mem_freq = 1066; break; - case 3: + case 2: dev_priv->mem_freq = 1333; break; + case 3: + /* + * Probably a BIOS/Punit bug, or a new platform we don't + * support yet. + */ + WARN(1, "invalid DDR freq detected, assuming 800MHz\n"); + dev_priv->mem_freq = 800; + break; } DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq); -- cgit v1.2.3 From 07ab118b393410c65146f44c17b0ae5373eb972e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 5 Nov 2013 22:42:28 +0200 Subject: drm/i915: Improve vlv_gpu_freq() and vlv_freq_opcode() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We're currently miscalculating the VLV graphics clock a little bit. This is caused by rounding the step to integer MHz, which does not match reality. Change the formula to match the GUnit HAS to give more accurate answers. Signed-off-by: Ville Syrjälä Reviewed-by: Jesse Barnes Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 35 ++++++++++++----------------------- 1 file changed, 12 insertions(+), 23 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index a5778e59cc15..865035beccc7 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5947,57 +5947,46 @@ int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val) int vlv_gpu_freq(int ddr_freq, int val) { - int mult, base; + int div; + /* 4 x czclk */ switch (ddr_freq) { case 800: - mult = 20; - base = 120; + div = 10; break; case 1066: - mult = 22; - base = 133; + div = 12; break; case 1333: - mult = 21; - base = 125; + div = 16; break; default: return -1; } - return ((val - 0xbd) * mult) + base; + return DIV_ROUND_CLOSEST(ddr_freq * (val + 6 - 0xbd), 4 * div); } int vlv_freq_opcode(int ddr_freq, int val) { - int mult, base; + int mul; + /* 4 x czclk */ switch (ddr_freq) { case 800: - mult = 20; - base = 120; + mul = 10; break; case 1066: - mult = 22; - base = 133; + mul = 12; break; case 1333: - mult = 21; - base = 125; + mul = 16; break; default: return -1; } - val /= mult; - val -= base / mult; - val += 0xbd; - - if (val > 0xea) - val = 0xea; - - return val; + return DIV_ROUND_CLOSEST(4 * mul * val, ddr_freq) + 0xbd - 6; } void intel_pm_init(struct drm_device *dev) -- cgit v1.2.3 From 2ec3815f29d1b7659ecf3f1791e7e394efdd6969 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 5 Nov 2013 22:42:29 +0200 Subject: drm/i915: Pass dev_priv to vlv_gpu_freq() and vlv_freq_opcode() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We'll be looking at more than just mem_freq from dev_priv, so just pass the whole thing. Signed-off-by: Ville Syrjälä Reviewed-by: Jesse Barnes Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 38 ++++++++++++++++---------------------- 1 file changed, 16 insertions(+), 22 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 865035beccc7..f5bb9b349487 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -3609,9 +3609,9 @@ static void vlv_update_rps_cur_delay(struct drm_i915_private *dev_priv) if (pval != dev_priv->rps.cur_delay) DRM_DEBUG_DRIVER("Punit overrode GPU freq: %d MHz (%u) requested, but got %d Mhz (%u)\n", - vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.cur_delay), + vlv_gpu_freq(dev_priv, dev_priv->rps.cur_delay), dev_priv->rps.cur_delay, - vlv_gpu_freq(dev_priv->mem_freq, pval), pval); + vlv_gpu_freq(dev_priv, pval), pval); dev_priv->rps.cur_delay = pval; } @@ -3629,10 +3629,9 @@ void valleyview_set_rps(struct drm_device *dev, u8 val) vlv_update_rps_cur_delay(dev_priv); DRM_DEBUG_DRIVER("GPU freq request from %d MHz (%u) to %d MHz (%u)\n", - vlv_gpu_freq(dev_priv->mem_freq, - dev_priv->rps.cur_delay), + vlv_gpu_freq(dev_priv, dev_priv->rps.cur_delay), dev_priv->rps.cur_delay, - vlv_gpu_freq(dev_priv->mem_freq, val), val); + vlv_gpu_freq(dev_priv, val), val); if (val == dev_priv->rps.cur_delay) return; @@ -3641,7 +3640,7 @@ void valleyview_set_rps(struct drm_device *dev, u8 val) dev_priv->rps.cur_delay = val; - trace_intel_gpu_freq_change(vlv_gpu_freq(dev_priv->mem_freq, val)); + trace_intel_gpu_freq_change(vlv_gpu_freq(dev_priv, val)); } static void gen6_disable_rps_interrupts(struct drm_device *dev) @@ -4070,32 +4069,27 @@ static void valleyview_enable_rps(struct drm_device *dev) dev_priv->rps.cur_delay = (val >> 8) & 0xff; DRM_DEBUG_DRIVER("current GPU freq: %d MHz (%u)\n", - vlv_gpu_freq(dev_priv->mem_freq, - dev_priv->rps.cur_delay), + vlv_gpu_freq(dev_priv, dev_priv->rps.cur_delay), dev_priv->rps.cur_delay); dev_priv->rps.max_delay = valleyview_rps_max_freq(dev_priv); dev_priv->rps.hw_max = dev_priv->rps.max_delay; DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n", - vlv_gpu_freq(dev_priv->mem_freq, - dev_priv->rps.max_delay), + vlv_gpu_freq(dev_priv, dev_priv->rps.max_delay), dev_priv->rps.max_delay); dev_priv->rps.rpe_delay = valleyview_rps_rpe_freq(dev_priv); DRM_DEBUG_DRIVER("RPe GPU freq: %d MHz (%u)\n", - vlv_gpu_freq(dev_priv->mem_freq, - dev_priv->rps.rpe_delay), + vlv_gpu_freq(dev_priv, dev_priv->rps.rpe_delay), dev_priv->rps.rpe_delay); dev_priv->rps.min_delay = valleyview_rps_min_freq(dev_priv); DRM_DEBUG_DRIVER("min GPU freq: %d MHz (%u)\n", - vlv_gpu_freq(dev_priv->mem_freq, - dev_priv->rps.min_delay), + vlv_gpu_freq(dev_priv, dev_priv->rps.min_delay), dev_priv->rps.min_delay); DRM_DEBUG_DRIVER("setting GPU freq to %d MHz (%u)\n", - vlv_gpu_freq(dev_priv->mem_freq, - dev_priv->rps.rpe_delay), + vlv_gpu_freq(dev_priv, dev_priv->rps.rpe_delay), dev_priv->rps.rpe_delay); valleyview_set_rps(dev_priv->dev, dev_priv->rps.rpe_delay); @@ -5945,12 +5939,12 @@ int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val) return 0; } -int vlv_gpu_freq(int ddr_freq, int val) +int vlv_gpu_freq(struct drm_i915_private *dev_priv, int val) { int div; /* 4 x czclk */ - switch (ddr_freq) { + switch (dev_priv->mem_freq) { case 800: div = 10; break; @@ -5964,15 +5958,15 @@ int vlv_gpu_freq(int ddr_freq, int val) return -1; } - return DIV_ROUND_CLOSEST(ddr_freq * (val + 6 - 0xbd), 4 * div); + return DIV_ROUND_CLOSEST(dev_priv->mem_freq * (val + 6 - 0xbd), 4 * div); } -int vlv_freq_opcode(int ddr_freq, int val) +int vlv_freq_opcode(struct drm_i915_private *dev_priv, int val) { int mul; /* 4 x czclk */ - switch (ddr_freq) { + switch (dev_priv->mem_freq) { case 800: mul = 10; break; @@ -5986,7 +5980,7 @@ int vlv_freq_opcode(int ddr_freq, int val) return -1; } - return DIV_ROUND_CLOSEST(4 * mul * val, ddr_freq) + 0xbd - 6; + return DIV_ROUND_CLOSEST(4 * mul * val, dev_priv->mem_freq) + 0xbd - 6; } void intel_pm_init(struct drm_device *dev) -- cgit v1.2.3 From 2325991e021fb0da21d8d08009d1eea78133745a Mon Sep 17 00:00:00 2001 From: Chon Ming Lee Date: Thu, 7 Nov 2013 15:23:26 +0800 Subject: drm/i915/vlv: Workaround a punit issue in DDR data rate for 1333. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For DDR data rate reporting by Punit in PUNIT_GPU_FREQ_STS, the actual data encoding is 00b=800, 01b=1066, 10b=1333, 11b=1333. Some premium VLV sku will get the DDR_DATA_RATE set as 11. As a result, the turbo frequency reporting will be incorrect without this workaround. Signed-off-by: Chon Ming Lee Reviewed-by: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index f5bb9b349487..38943f8ef260 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5322,12 +5322,7 @@ static void valleyview_init_clock_gating(struct drm_device *dev) dev_priv->mem_freq = 1333; break; case 3: - /* - * Probably a BIOS/Punit bug, or a new platform we don't - * support yet. - */ - WARN(1, "invalid DDR freq detected, assuming 800MHz\n"); - dev_priv->mem_freq = 800; + dev_priv->mem_freq = 1333; break; } DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq); -- cgit v1.2.3 From 6917c7b9d9083272ddf7e64f5482e8820a31fb3c Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 6 Nov 2013 13:56:26 -0200 Subject: drm/i915: Initialise min/max frequencies before updating RPS registers MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The RPS register writing routines use the current value of min/max to set certain limits and interrupt gating. If we set those afterwards, we risk setting up the hw incorrectly and losing power management events, and worse, trigger some internal assertions. Reorder the calling sequences to be correct, and remove the then unrequired clamping from inside set_rps(). And for a bonus, fix the bug of calling gen6_set_rps() from Valleyview. Signed-off-by: Chris Wilson Signed-off-by: Rodrigo Vivi Reviewed-by: Ville Syrjälä CC: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 19 +++++-------------- 1 file changed, 5 insertions(+), 14 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 38943f8ef260..e37860377285 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -3414,26 +3414,19 @@ static void ironlake_disable_drps(struct drm_device *dev) * ourselves, instead of doing a rmw cycle (which might result in us clearing * all limits and the gpu stuck at whatever frequency it is at atm). */ -static u32 gen6_rps_limits(struct drm_i915_private *dev_priv, u8 *val) +static u32 gen6_rps_limits(struct drm_i915_private *dev_priv, u8 val) { u32 limits; - limits = 0; - - if (*val >= dev_priv->rps.max_delay) - *val = dev_priv->rps.max_delay; - limits |= dev_priv->rps.max_delay << 24; - /* Only set the down limit when we've reached the lowest level to avoid * getting more interrupts, otherwise leave this clear. This prevents a * race in the hw when coming out of rc6: There's a tiny window where * the hw runs at the minimal clock before selecting the desired * frequency, if the down threshold expires in that window we will not * receive a down interrupt. */ - if (*val <= dev_priv->rps.min_delay) { - *val = dev_priv->rps.min_delay; + limits = dev_priv->rps.max_delay << 24; + if (val <= dev_priv->rps.min_delay) limits |= dev_priv->rps.min_delay << 16; - } return limits; } @@ -3533,7 +3526,6 @@ static void gen6_set_rps_thresholds(struct drm_i915_private *dev_priv, u8 val) void gen6_set_rps(struct drm_device *dev, u8 val) { struct drm_i915_private *dev_priv = dev->dev_private; - u32 limits = gen6_rps_limits(dev_priv, &val); WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); WARN_ON(val > dev_priv->rps.max_delay); @@ -3556,7 +3548,8 @@ void gen6_set_rps(struct drm_device *dev, u8 val) /* 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)); POSTING_READ(GEN6_RPNSWREQ); @@ -3620,8 +3613,6 @@ void valleyview_set_rps(struct drm_device *dev, u8 val) { struct drm_i915_private *dev_priv = dev->dev_private; - gen6_rps_limits(dev_priv, &val); - WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); WARN_ON(val > dev_priv->rps.max_delay); WARN_ON(val < dev_priv->rps.min_delay); -- cgit v1.2.3 From 4c7915616a7a09be0c1e5b76c26381df15fe671b Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 7 Nov 2013 19:57:48 +0200 Subject: drm/i915: Kill vlv_update_rps_cur_delay() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Polling to make sure the current GPU frequency matches the last requested frequency should not be necessay, and if there's some throttling involved, the two might not match anyway. Since we're still seeing this trigger occasionally, and it just introduces a rather pointless 10 ms delay, it seems like better to kill it off. Signed-off-by: Ville Syrjälä Reviewed-by: Jesse Barnes Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 27 --------------------------- 1 file changed, 27 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index e37860377285..542b4448ccb7 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -3584,31 +3584,6 @@ void gen6_rps_boost(struct drm_i915_private *dev_priv) mutex_unlock(&dev_priv->rps.hw_lock); } -/* - * Wait until the previous freq change has completed, - * or the timeout elapsed, and then update our notion - * of the current GPU frequency. - */ -static void vlv_update_rps_cur_delay(struct drm_i915_private *dev_priv) -{ - u32 pval; - - WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); - - if (wait_for(((pval = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS)) & GENFREQSTATUS) == 0, 10)) - DRM_DEBUG_DRIVER("timed out waiting for Punit\n"); - - pval >>= 8; - - if (pval != dev_priv->rps.cur_delay) - DRM_DEBUG_DRIVER("Punit overrode GPU freq: %d MHz (%u) requested, but got %d Mhz (%u)\n", - vlv_gpu_freq(dev_priv, dev_priv->rps.cur_delay), - dev_priv->rps.cur_delay, - vlv_gpu_freq(dev_priv, pval), pval); - - dev_priv->rps.cur_delay = pval; -} - void valleyview_set_rps(struct drm_device *dev, u8 val) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -3617,8 +3592,6 @@ void valleyview_set_rps(struct drm_device *dev, u8 val) WARN_ON(val > dev_priv->rps.max_delay); WARN_ON(val < dev_priv->rps.min_delay); - vlv_update_rps_cur_delay(dev_priv); - DRM_DEBUG_DRIVER("GPU freq request from %d MHz (%u) to %d MHz (%u)\n", vlv_gpu_freq(dev_priv, dev_priv->rps.cur_delay), dev_priv->rps.cur_delay, -- cgit v1.2.3 From b19870ee67ad8b552db0e7456a8d1d33f96421c0 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Wed, 6 Nov 2013 23:02:25 +0200 Subject: drm/i915: Use plane_name() in gen7_enable_fbc() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit All the other .enable_fbc() funcs use plane_name(). Make gen7_enable_fbc() do the same. Signed-off-by: Ville Syrjälä Reviewed-by: Rodrigo Vivi Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 172efa0bfb86..0fcd591b0a4b 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -295,7 +295,7 @@ static void gen7_enable_fbc(struct drm_crtc *crtc, unsigned long interval) sandybridge_blit_fbc_update(dev); - DRM_DEBUG_KMS("enabled fbc on plane %d\n", intel_crtc->plane); + DRM_DEBUG_KMS("enabled fbc on plane %c\n", plane_name(intel_crtc->plane)); } bool intel_fbc_enabled(struct drm_device *dev) -- cgit v1.2.3 From d629336b6af9ff214e9d1e7224946a000fc8f70f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 21 Nov 2013 21:29:45 +0200 Subject: drm/i915: Don't set the fence number in DPFC_CTL on SNB MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit SNB has another register where the actual FBC CPU fence number is stored. The documenation explicitly states that the fence number in DPFC_CTL must be 0 on SNB. And in fact when it's not zero, the GTT tracking simply doesn't work. Signed-off-by: Ville Syrjälä Reviewed-by: Chris Wilson Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 0fcd591b0a4b..de4cf565ec4c 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -222,7 +222,9 @@ static void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval) 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); + dpfc_ctl |= DPFC_CTL_FENCE_EN; + if (IS_GEN5(dev)) + dpfc_ctl |= obj->fence_reg; I915_WRITE(ILK_DPFC_CHICKEN, DPFC_HT_MODIFY); I915_WRITE(ILK_DPFC_RECOMP_CTL, DPFC_RECOMP_STALL_EN | -- cgit v1.2.3 From fbeeaa2306fcff7614c8b41b420a400503b6d28e Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 25 Nov 2013 17:15:28 +0200 Subject: drm/i915: add audio power domain This way the code is simpler and can also be used for other platforms where the audio power domain->power well mapping is different. Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index de4cf565ec4c..8f472daa2793 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5775,10 +5775,7 @@ void i915_request_power_well(void) dev_priv = container_of(hsw_pwr, struct drm_i915_private, power_domains); - - mutex_lock(&hsw_pwr->lock); - __intel_power_well_get(dev_priv->dev, &hsw_pwr->power_wells[0]); - mutex_unlock(&hsw_pwr->lock); + intel_display_power_get(dev_priv->dev, POWER_DOMAIN_AUDIO); } EXPORT_SYMBOL_GPL(i915_request_power_well); @@ -5792,10 +5789,7 @@ void i915_release_power_well(void) dev_priv = container_of(hsw_pwr, struct drm_i915_private, power_domains); - - mutex_lock(&hsw_pwr->lock); - __intel_power_well_put(dev_priv->dev, &hsw_pwr->power_wells[0]); - mutex_unlock(&hsw_pwr->lock); + intel_display_power_put(dev_priv->dev, POWER_DOMAIN_AUDIO); } EXPORT_SYMBOL_GPL(i915_release_power_well); -- cgit v1.2.3 From c1ca727f89450cbc560af93045d57a186b83b0dc Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 25 Nov 2013 17:15:29 +0200 Subject: drm/i915: support for multiple power wells HW generations so far had only one always-on power well and optionally one dynamic power well. Upcoming HW gens may have multiple dynamic power wells, so add some infrastructure to support them. The idea is to keep the existing power domain API used by the rest of the driver and create a mapping between these power domains and the underlying power wells. This mapping can differ from one HW to another but high level driver code doesn't need to know about this. Through the existing get/put API it would just ask for a given power domain and the power domain framework would make sure the relevant power wells get enabled in the right order. Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 116 ++++++++++++++++++++++++++++++++++------ 1 file changed, 99 insertions(+), 17 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 8f472daa2793..2b3173415212 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5627,15 +5627,41 @@ static bool is_always_on_power_domain(struct drm_device *dev, return BIT(domain) & always_on_domains; } +#define for_each_power_well(i, power_well, domain_mask, power_domains) \ + for (i = 0; \ + i < (power_domains)->power_well_count && \ + ((power_well) = &(power_domains)->power_wells[i]); \ + i++) \ + if ((power_well)->domains & (domain_mask)) + +#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_device *dev, + struct i915_power_well *power_well) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + return I915_READ(HSW_PWR_WELL_DRIVER) == + (HSW_PWR_WELL_ENABLE_REQUEST | HSW_PWR_WELL_STATE_ENABLED); +} + bool intel_display_power_enabled(struct drm_device *dev, enum intel_display_power_domain domain) { struct drm_i915_private *dev_priv = dev->dev_private; + struct i915_power_domains *power_domains; + struct i915_power_well *power_well; + bool is_enabled; + int i; if (!HAS_POWER_WELL(dev)) return true; @@ -5643,11 +5669,24 @@ bool intel_display_power_enabled(struct drm_device *dev, if (is_always_on_power_domain(dev, domain)) return true; - return I915_READ(HSW_PWR_WELL_DRIVER) == - (HSW_PWR_WELL_ENABLE_REQUEST | HSW_PWR_WELL_STATE_ENABLED); + power_domains = &dev_priv->power_domains; + + is_enabled = true; + + mutex_lock(&power_domains->lock); + for_each_power_well_rev(i, power_well, BIT(domain), power_domains) { + if (!power_well->is_enabled(dev, power_well)) { + is_enabled = false; + break; + } + } + mutex_unlock(&power_domains->lock); + + return is_enabled; } -static void __intel_set_power_well(struct drm_device *dev, bool enable) +static void hsw_set_power_well(struct drm_device *dev, + struct i915_power_well *power_well, bool enable) { struct drm_i915_private *dev_priv = dev->dev_private; bool is_enabled, enable_requested; @@ -5713,16 +5752,17 @@ static void __intel_set_power_well(struct drm_device *dev, bool enable) static void __intel_power_well_get(struct drm_device *dev, struct i915_power_well *power_well) { - if (!power_well->count++) - __intel_set_power_well(dev, true); + if (!power_well->count++ && power_well->set) + power_well->set(dev, power_well, true); } static void __intel_power_well_put(struct drm_device *dev, struct i915_power_well *power_well) { WARN_ON(!power_well->count); - if (!--power_well->count && i915_disable_power_well) - __intel_set_power_well(dev, false); + + if (!--power_well->count && power_well->set && i915_disable_power_well) + power_well->set(dev, power_well, false); } void intel_display_power_get(struct drm_device *dev, @@ -5730,6 +5770,8 @@ void intel_display_power_get(struct drm_device *dev, { struct drm_i915_private *dev_priv = dev->dev_private; struct i915_power_domains *power_domains; + struct i915_power_well *power_well; + int i; if (!HAS_POWER_WELL(dev)) return; @@ -5740,7 +5782,8 @@ void intel_display_power_get(struct drm_device *dev, power_domains = &dev_priv->power_domains; mutex_lock(&power_domains->lock); - __intel_power_well_get(dev, &power_domains->power_wells[0]); + for_each_power_well(i, power_well, BIT(domain), power_domains) + __intel_power_well_get(dev, power_well); mutex_unlock(&power_domains->lock); } @@ -5749,6 +5792,8 @@ void intel_display_power_put(struct drm_device *dev, { struct drm_i915_private *dev_priv = dev->dev_private; struct i915_power_domains *power_domains; + struct i915_power_well *power_well; + int i; if (!HAS_POWER_WELL(dev)) return; @@ -5759,7 +5804,8 @@ void intel_display_power_put(struct drm_device *dev, power_domains = &dev_priv->power_domains; mutex_lock(&power_domains->lock); - __intel_power_well_put(dev, &power_domains->power_wells[0]); + for_each_power_well_rev(i, power_well, BIT(domain), power_domains) + __intel_power_well_put(dev, power_well); mutex_unlock(&power_domains->lock); } @@ -5793,17 +5839,52 @@ void i915_release_power_well(void) } EXPORT_SYMBOL_GPL(i915_release_power_well); +static struct i915_power_well hsw_power_wells[] = { + { + .name = "display", + .domains = POWER_DOMAIN_MASK & ~HSW_ALWAYS_ON_POWER_DOMAINS, + .is_enabled = hsw_power_well_enabled, + .set = hsw_set_power_well, + }, +}; + +static struct i915_power_well bdw_power_wells[] = { + { + .name = "display", + .domains = POWER_DOMAIN_MASK & ~BDW_ALWAYS_ON_POWER_DOMAINS, + .is_enabled = hsw_power_well_enabled, + .set = hsw_set_power_well, + }, +}; + +#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_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; struct i915_power_domains *power_domains = &dev_priv->power_domains; - struct i915_power_well *power_well; + + if (!HAS_POWER_WELL(dev)) + return 0; mutex_init(&power_domains->lock); - hsw_pwr = power_domains; - power_well = &power_domains->power_wells[0]; - power_well->count = 0; + /* + * The enabling order will be from lower to higher indexed wells, + * the disabling order is reversed. + */ + if (IS_HASWELL(dev)) { + set_power_wells(power_domains, hsw_power_wells); + hsw_pwr = power_domains; + } else if (IS_BROADWELL(dev)) { + set_power_wells(power_domains, bdw_power_wells); + hsw_pwr = power_domains; + } else { + WARN_ON(1); + } return 0; } @@ -5818,15 +5899,16 @@ static void intel_power_domains_resume(struct drm_device *dev) struct drm_i915_private *dev_priv = dev->dev_private; struct i915_power_domains *power_domains = &dev_priv->power_domains; struct i915_power_well *power_well; + int i; if (!HAS_POWER_WELL(dev)) return; mutex_lock(&power_domains->lock); - - power_well = &power_domains->power_wells[0]; - __intel_set_power_well(dev, power_well->count > 0); - + for_each_power_well(i, power_well, POWER_DOMAIN_MASK, power_domains) { + if (power_well->set) + power_well->set(dev, power_well, power_well->count > 0); + } mutex_unlock(&power_domains->lock); } -- cgit v1.2.3 From 6f3ef5ddabc0ad321678ee091c75b1f082a42707 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 25 Nov 2013 17:15:30 +0200 Subject: drm/i915: add always-on power wells instead of special casing them Instead of using a separate function to check whether a power domain is is always on, add an always-on power well covering all these power domains and do the usual get/put on these unconditionally. Since we don't assign a .set handler for these the get/put won't have any effect besides the adjusted refcount. This makes the code more readable and provides debug info also on the use of always-on power wells (once the relevant debugfs entry is added.) v3: make is_always_on to be bool instead of a bit field (Paulo) Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 41 +++++++++++++---------------------------- 1 file changed, 13 insertions(+), 28 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 2b3173415212..9230781824e5 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5608,25 +5608,6 @@ void intel_suspend_hw(struct drm_device *dev) lpt_suspend_hw(dev); } -static bool is_always_on_power_domain(struct drm_device *dev, - enum intel_display_power_domain domain) -{ - unsigned long always_on_domains; - - BUG_ON(BIT(domain) & ~POWER_DOMAIN_MASK); - - if (IS_BROADWELL(dev)) { - always_on_domains = BDW_ALWAYS_ON_POWER_DOMAINS; - } else if (IS_HASWELL(dev)) { - always_on_domains = HSW_ALWAYS_ON_POWER_DOMAINS; - } else { - WARN_ON(1); - return true; - } - - return BIT(domain) & always_on_domains; -} - #define for_each_power_well(i, power_well, domain_mask, power_domains) \ for (i = 0; \ i < (power_domains)->power_well_count && \ @@ -5666,15 +5647,15 @@ bool intel_display_power_enabled(struct drm_device *dev, if (!HAS_POWER_WELL(dev)) return true; - if (is_always_on_power_domain(dev, domain)) - return true; - power_domains = &dev_priv->power_domains; is_enabled = true; mutex_lock(&power_domains->lock); for_each_power_well_rev(i, power_well, BIT(domain), power_domains) { + if (power_well->always_on) + continue; + if (!power_well->is_enabled(dev, power_well)) { is_enabled = false; break; @@ -5776,9 +5757,6 @@ void intel_display_power_get(struct drm_device *dev, if (!HAS_POWER_WELL(dev)) return; - if (is_always_on_power_domain(dev, domain)) - return; - power_domains = &dev_priv->power_domains; mutex_lock(&power_domains->lock); @@ -5798,9 +5776,6 @@ void intel_display_power_put(struct drm_device *dev, if (!HAS_POWER_WELL(dev)) return; - if (is_always_on_power_domain(dev, domain)) - return; - power_domains = &dev_priv->power_domains; mutex_lock(&power_domains->lock); @@ -5840,6 +5815,11 @@ void i915_release_power_well(void) EXPORT_SYMBOL_GPL(i915_release_power_well); static struct i915_power_well hsw_power_wells[] = { + { + .name = "always-on", + .always_on = 1, + .domains = HSW_ALWAYS_ON_POWER_DOMAINS, + }, { .name = "display", .domains = POWER_DOMAIN_MASK & ~HSW_ALWAYS_ON_POWER_DOMAINS, @@ -5849,6 +5829,11 @@ static struct i915_power_well hsw_power_wells[] = { }; static struct i915_power_well bdw_power_wells[] = { + { + .name = "always-on", + .always_on = 1, + .domains = BDW_ALWAYS_ON_POWER_DOMAINS, + }, { .name = "display", .domains = POWER_DOMAIN_MASK & ~BDW_ALWAYS_ON_POWER_DOMAINS, -- cgit v1.2.3 From f7243ac9a25810140f5910c69821d5fa1e3e2387 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 25 Nov 2013 17:15:33 +0200 Subject: drm/i915: don't do BDW/HSW specific powerdomains init on other platforms Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 9230781824e5..5331925e37fa 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5914,6 +5914,9 @@ void intel_power_domains_init_hw(struct drm_device *dev) intel_display_set_init_power(dev, true); intel_power_domains_resume(dev); + if (!(IS_HASWELL(dev) || IS_BROADWELL(dev))) + return; + /* We're taking over the BIOS, so clear any requests made by it since * the driver is in charge now. */ if (I915_READ(HSW_PWR_WELL_BIOS) & HSW_PWR_WELL_ENABLE_REQUEST) -- cgit v1.2.3 From 1c2256df26f697bcd20016c271ddc8fd44653919 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 25 Nov 2013 17:15:34 +0200 Subject: drm/i915: add a default always-on power well So far we distinguished platforms without a dynamic power well with the HAS_POWER_WELL macro and for such platforms we didn't call any power domain functions. Instead of doing this check we can add an always-on power well for these platforms and call the power domain functions unconditionally. For always-on power wells we only increase/decrease their refcounts, otherwise they are nop. This makes high level driver code more readable and as a bonus provides some idea of the current power domains state for all platforms (once the relevant debugfs entry is added). v3: rename intel_power_wells to i9xx_always_on_power_well (Paulo) Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 28 +++++++++------------------- 1 file changed, 9 insertions(+), 19 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 5331925e37fa..6b4f91ee486c 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5644,9 +5644,6 @@ bool intel_display_power_enabled(struct drm_device *dev, bool is_enabled; int i; - if (!HAS_POWER_WELL(dev)) - return true; - power_domains = &dev_priv->power_domains; is_enabled = true; @@ -5754,9 +5751,6 @@ void intel_display_power_get(struct drm_device *dev, struct i915_power_well *power_well; int i; - if (!HAS_POWER_WELL(dev)) - return; - power_domains = &dev_priv->power_domains; mutex_lock(&power_domains->lock); @@ -5773,9 +5767,6 @@ void intel_display_power_put(struct drm_device *dev, struct i915_power_well *power_well; int i; - if (!HAS_POWER_WELL(dev)) - return; - power_domains = &dev_priv->power_domains; mutex_lock(&power_domains->lock); @@ -5814,6 +5805,14 @@ void i915_release_power_well(void) } EXPORT_SYMBOL_GPL(i915_release_power_well); +static struct i915_power_well i9xx_always_on_power_well[] = { + { + .name = "always-on", + .always_on = 1, + .domains = POWER_DOMAIN_MASK, + }, +}; + static struct i915_power_well hsw_power_wells[] = { { .name = "always-on", @@ -5852,9 +5851,6 @@ int intel_power_domains_init(struct drm_device *dev) struct drm_i915_private *dev_priv = dev->dev_private; struct i915_power_domains *power_domains = &dev_priv->power_domains; - if (!HAS_POWER_WELL(dev)) - return 0; - mutex_init(&power_domains->lock); /* @@ -5868,7 +5864,7 @@ int intel_power_domains_init(struct drm_device *dev) set_power_wells(power_domains, bdw_power_wells); hsw_pwr = power_domains; } else { - WARN_ON(1); + set_power_wells(power_domains, i9xx_always_on_power_well); } return 0; @@ -5886,9 +5882,6 @@ static void intel_power_domains_resume(struct drm_device *dev) struct i915_power_well *power_well; int i; - if (!HAS_POWER_WELL(dev)) - return; - mutex_lock(&power_domains->lock); for_each_power_well(i, power_well, POWER_DOMAIN_MASK, power_domains) { if (power_well->set) @@ -5907,9 +5900,6 @@ void intel_power_domains_init_hw(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - if (!HAS_POWER_WELL(dev)) - return; - /* For now, we need the power well to be always enabled. */ intel_display_set_init_power(dev, true); intel_power_domains_resume(dev); -- cgit v1.2.3 From 1da51581b00b3fc3ac72156e2a69c6bab71f7794 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 25 Nov 2013 17:15:35 +0200 Subject: drm/i915: add a debugfs entry for power domain info Add a debugfs entry showing the use-count for all power domains of each power well. v3: address comments from Paulo: - simplify power_domain_str() by using a switch table - move power_well::domain_count to power_domains - WARN_ON decrementing a 0 refcount Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 6b4f91ee486c..2f8399c347ee 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5754,8 +5754,13 @@ void intel_display_power_get(struct drm_device *dev, power_domains = &dev_priv->power_domains; mutex_lock(&power_domains->lock); + +#if IS_ENABLED(CONFIG_DEBUG_FS) + power_domains->domain_use_count[domain]++; +#endif for_each_power_well(i, power_well, BIT(domain), power_domains) __intel_power_well_get(dev, power_well); + mutex_unlock(&power_domains->lock); } @@ -5770,8 +5775,15 @@ void intel_display_power_put(struct drm_device *dev, power_domains = &dev_priv->power_domains; mutex_lock(&power_domains->lock); + for_each_power_well_rev(i, power_well, BIT(domain), power_domains) __intel_power_well_put(dev, power_well); + +#if IS_ENABLED(CONFIG_DEBUG_FS) + WARN_ON(!power_domains->domain_use_count[domain]); + power_domains->domain_use_count[domain]--; +#endif + mutex_unlock(&power_domains->lock); } -- cgit v1.2.3 From c8d9a5905e45d856fb21cce2e20f186ce6719560 Mon Sep 17 00:00:00 2001 From: Deepak S Date: Sat, 23 Nov 2013 14:55:42 +0530 Subject: drm/i915: Add power well arguments to force wake routines. Added power well arguments to all the force wake routines to help us individually control power well based on the scenario. Signed-off-by: Deepak S Reviewed-by: Jesse Barnes [danvet: Resolve conflict with the removed forcewake hack and drop one spurious hunk Jesse noticed.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 2f8399c347ee..fd2537d429f2 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -191,7 +191,8 @@ 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); + gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL); + blt_ecoskpd = I915_READ(GEN6_BLITTER_ECOSKPD); blt_ecoskpd |= GEN6_BLITTER_FBC_NOTIFY << GEN6_BLITTER_LOCK_SHIFT; @@ -202,7 +203,8 @@ 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_ALL); } static void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval) @@ -3739,7 +3741,7 @@ static void gen8_enable_rps(struct drm_device *dev) /* 1c & 1d: Get forcewake during program sequence. Although the driver * hasn't enabled a state yet where we need forcewake, BIOS may have.*/ - gen6_gt_force_wake_get(dev_priv); + gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL); /* 2a: Disable RC states. */ I915_WRITE(GEN6_RC_CONTROL, 0); @@ -3796,7 +3798,7 @@ static void gen8_enable_rps(struct drm_device *dev) gen6_enable_rps_interrupts(dev); - gen6_gt_force_wake_put(dev_priv); + gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL); } static void gen6_enable_rps(struct drm_device *dev) @@ -3826,7 +3828,7 @@ 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); @@ -3918,7 +3920,7 @@ static void gen6_enable_rps(struct drm_device *dev) DRM_ERROR("Couldn't fix incorrect rc6 voltage\n"); } - gen6_gt_force_wake_put(dev_priv); + gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL); } void gen6_update_ring_freq(struct drm_device *dev) @@ -4080,7 +4082,8 @@ static void valleyview_enable_rps(struct drm_device *dev) valleyview_setup_pctx(dev); - gen6_gt_force_wake_get(dev_priv); + /* If VLV, Forcewake all wells, else re-direct to regular path */ + gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL); I915_WRITE(GEN6_RP_UP_THRESHOLD, 59400); I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 245000); @@ -4152,7 +4155,7 @@ static void valleyview_enable_rps(struct drm_device *dev) gen6_enable_rps_interrupts(dev); - gen6_gt_force_wake_put(dev_priv); + gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL); } void ironlake_teardown_rc6(struct drm_device *dev) -- cgit v1.2.3 From 940aece471bd6656b86f5d77132b6670a3b88dc8 Mon Sep 17 00:00:00 2001 From: Deepak S Date: Sat, 23 Nov 2013 14:55:43 +0530 Subject: drm/i915/vlv: Valleyview support for forcewake Individual power wells. Split vlv force wake routines to help individually control Media/Render well based on the register access. We've seen power savings in the lower sub-1W range on workloads that only need on of the power wells, e.g. glbenchmark, media playback Note: The same split isn't there for the forcewake queue, only the forcwake domains are split. Signed-off-by: Deepak S Reviewed-by: Jesse Barnes [danvet: Rebase on top of the removed forcewake hack in the ring irq get/put code and add a note to add Deepak's answer to Chris question.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index fd2537d429f2..1659265a7f7a 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -191,7 +191,10 @@ 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, FORCEWAKE_ALL); + + /* 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 << @@ -204,7 +207,7 @@ static void sandybridge_blit_fbc_update(struct drm_device *dev) I915_WRITE(GEN6_BLITTER_ECOSKPD, blt_ecoskpd); POSTING_READ(GEN6_BLITTER_ECOSKPD); - gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL); + gen6_gt_force_wake_put(dev_priv, FORCEWAKE_MEDIA); } static void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval) -- cgit v1.2.3 From 2f0aa30425414f3105bc29dd39bfdb40cb684393 Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Fri, 15 Nov 2013 09:32:11 -0800 Subject: drm/i915/vlv: use a lower RC6 timeout on VLV We use timeout mode, and we need to lower the timeout to get good RC6 residency when loads are running. This gets me from 0% residency during glxgears to 77%, which is a pretty good improvement. This value also matches the current BWG recommentations. Tested-by: "Meng, Mengmeng" Signed-off-by: Jesse Barnes Reviewed-by: Deepak S Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 1659265a7f7a..2d74ae83b175 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4110,7 +4110,7 @@ static void valleyview_enable_rps(struct drm_device *dev) for_each_ring(ring, dev_priv, i) I915_WRITE(RING_MAX_IDLE(ring->mmio_base), 10); - I915_WRITE(GEN6_RC6_THRESHOLD, 0xc350); + I915_WRITE(GEN6_RC6_THRESHOLD, 0x557); /* allows RC6 residency counter to work */ I915_WRITE(VLV_COUNTER_CONTROL, -- cgit v1.2.3 From 6b88f295690d3340fcbd01c1ff46991ea4f83c9f Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Fri, 15 Nov 2013 09:32:12 -0800 Subject: drm/i915/vlv: use parallel context restore when coming out of RC6 Setting this bit restores all ring contexts in parallel rather than serially. Matches current BWG recommendations. Tested-by: "Meng, Mengmeng" Signed-off-by: Jesse Barnes Reviewed-by: Deepak S Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 2d74ae83b175..eac7c4eb345c 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4118,7 +4118,7 @@ static void valleyview_enable_rps(struct drm_device *dev) VLV_MEDIA_RC6_COUNT_EN | VLV_RENDER_RC6_COUNT_EN)); if (intel_enable_rc6(dev) & INTEL_RC6_ENABLE) - rc6_mode = GEN7_RC_CTL_TO_MODE; + rc6_mode = GEN7_RC_CTL_TO_MODE | VLV_RC_CTL_CTX_RST_PARALLEL; intel_print_rc6_info(dev, rc6_mode); -- cgit v1.2.3 From ddf9c536295b9d7fcfd0bfc377593b41f2a4dc02 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Wed, 27 Nov 2013 22:02:02 +0200 Subject: drm/i915: add intel_display_power_enabled_sw() for use in atomic ctx MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Atm we call intel_display_power_enabled() from i915_capture_error_state() in IRQ context and then take a mutex. To fix this add a new intel_display_power_enabled_sw() which returns the domain state based on software tracking as opposed to reading the actual HW state. Since we use domain_use_count for this without locking on the reader side make sure we increase the counter only after enabling all required power wells and decrease it before disabling any of these power wells. Regression introduced in commit 1b02383464b4a915627ef3b8fd0ad7f07168c54c Author: Imre Deak Date:   Tue Sep 24 16:17:09 2013 +0300     drm/i915: support for multiple power wells Note that atm we depend on the value returned by intel_display_power_enabled_sw() in i915_capture_error_state() to avoid unclaimed register access reports. This was never guaranteed though, since another thread can disable the power concurrently. If this is a problem we need another explicit way to disable the reporting during error captures. v2: - remove barriers as the caller can't depend on the value returned from i915_capture_error_state_sw() anyway (Ville) - dump the state of pipe/transcoder power domain state (Daniel) Reported-by: Chris Wilson Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index eac7c4eb345c..ff47520f8d40 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5641,6 +5641,17 @@ static bool hsw_power_well_enabled(struct drm_device *dev, (HSW_PWR_WELL_ENABLE_REQUEST | HSW_PWR_WELL_STATE_ENABLED); } +bool intel_display_power_enabled_sw(struct drm_device *dev, + enum intel_display_power_domain domain) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + struct i915_power_domains *power_domains; + + power_domains = &dev_priv->power_domains; + + return power_domains->domain_use_count[domain]; +} + bool intel_display_power_enabled(struct drm_device *dev, enum intel_display_power_domain domain) { @@ -5761,12 +5772,11 @@ void intel_display_power_get(struct drm_device *dev, mutex_lock(&power_domains->lock); -#if IS_ENABLED(CONFIG_DEBUG_FS) - power_domains->domain_use_count[domain]++; -#endif for_each_power_well(i, power_well, BIT(domain), power_domains) __intel_power_well_get(dev, power_well); + power_domains->domain_use_count[domain]++; + mutex_unlock(&power_domains->lock); } @@ -5782,13 +5792,11 @@ void intel_display_power_put(struct drm_device *dev, mutex_lock(&power_domains->lock); - for_each_power_well_rev(i, power_well, BIT(domain), power_domains) - __intel_power_well_put(dev, power_well); - -#if IS_ENABLED(CONFIG_DEBUG_FS) WARN_ON(!power_domains->domain_use_count[domain]); power_domains->domain_use_count[domain]--; -#endif + + for_each_power_well_rev(i, power_well, BIT(domain), power_domains) + __intel_power_well_put(dev, power_well); mutex_unlock(&power_domains->lock); } -- cgit v1.2.3 From c5a44aa012ee86b3dfd0c6050ba34cd6eb412875 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 28 Nov 2013 17:29:58 +0200 Subject: drm/i915: Fix FBC1 plane checks for gen2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On gen2 and gen3 chipsets FBC is supported only on plane A. Fix (and simplify) the plane checks in intel_update_fbc() accordingly. Signed-off-by: Ville Syrjälä Reviewed-by: Chris Wilson Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index ff47520f8d40..d389078f0fe1 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -537,10 +537,10 @@ void intel_update_fbc(struct drm_device *dev) DRM_DEBUG_KMS("mode too large for compression, disabling\n"); goto out_disable; } - if ((IS_I915GM(dev) || IS_I945GM(dev) || IS_HASWELL(dev)) && - intel_crtc->plane != 0) { + if ((INTEL_INFO(dev)->gen < 4 || IS_HASWELL(dev)) && + intel_crtc->plane != PLANE_A) { if (set_no_fbc_reason(dev_priv, FBC_BAD_PLANE)) - DRM_DEBUG_KMS("plane not 0, disabling compression\n"); + DRM_DEBUG_KMS("plane not A, disabling compression\n"); goto out_disable; } -- cgit v1.2.3 From 40045465a91cad4e4bcd2691f0bece5f8b2910e3 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 28 Nov 2013 17:29:59 +0200 Subject: drm/i915: Reorganize FBC function pointer initializaition MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Initialize the FBC vfuncs on gen2 and gen3 chipsets. Also make a clean split for gen7+ vs. gen5+ vfunc initialization. Signed-off-by: Ville Syrjälä Reviewed-by: Chris Wilson Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 16 +++++++--------- 1 file changed, 7 insertions(+), 9 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index d389078f0fe1..ac4a74a41344 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5956,25 +5956,23 @@ 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 (INTEL_INFO(dev)->gen >= 7) { dev_priv->display.fbc_enabled = ironlake_fbc_enabled; - if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev)) - dev_priv->display.enable_fbc = - gen7_enable_fbc; - else - dev_priv->display.enable_fbc = - ironlake_enable_fbc; + dev_priv->display.enable_fbc = gen7_enable_fbc; + dev_priv->display.disable_fbc = ironlake_disable_fbc; + } else if (INTEL_INFO(dev)->gen >= 5) { + dev_priv->display.fbc_enabled = ironlake_fbc_enabled; + dev_priv->display.enable_fbc = ironlake_enable_fbc; dev_priv->display.disable_fbc = ironlake_disable_fbc; } else if (IS_GM45(dev)) { dev_priv->display.fbc_enabled = g4x_fbc_enabled; dev_priv->display.enable_fbc = g4x_enable_fbc; dev_priv->display.disable_fbc = g4x_disable_fbc; - } else if (IS_CRESTLINE(dev)) { + } else { dev_priv->display.fbc_enabled = i8xx_fbc_enabled; dev_priv->display.enable_fbc = i8xx_enable_fbc; dev_priv->display.disable_fbc = i8xx_disable_fbc; } - /* 855GM needs testing */ } /* For cxsr */ -- cgit v1.2.3 From d62292c8f778772d1b6ec125d461c8c16fdc0417 Mon Sep 17 00:00:00 2001 From: Paulo Zanoni Date: Wed, 27 Nov 2013 17:59:22 -0200 Subject: drm/i915: get a PC8 reference when enabling the power well In the current code, at haswell_modeset_global_resources, first we decide if we want to enable/disable the power well, then we decide if we want to enable/disable PC8. On the case where we're enabling PC8 this works fine, but on the case where we disable PC8 due to a non-eDP monitor being enabled, we first enable the power well and then disable PC8. Although wrong, this doesn't seem to be causing any problems now, and we don't even see anything in dmesg. But the patches for runtime D3 turn this problem into a real bug, so we need to fix it. This fixes the "modeset-non-lpsp" subtest from the "pm_pc8" test from intel-gpu-tools. v2: - Rebase (i915_disable_power_well). v3: - More reabase. Signed-off-by: Paulo Zanoni Reviewed-by: Rodrigo Vivi Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 41b6e080e362..cd3f511847ec 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5688,6 +5688,8 @@ static void hsw_set_power_well(struct drm_device *dev, unsigned long irqflags; uint32_t tmp; + WARN_ON(dev_priv->pc8.enabled); + tmp = I915_READ(HSW_PWR_WELL_DRIVER); is_enabled = tmp & HSW_PWR_WELL_STATE_ENABLED; enable_requested = tmp & HSW_PWR_WELL_ENABLE_REQUEST; @@ -5747,17 +5749,26 @@ static void hsw_set_power_well(struct drm_device *dev, static void __intel_power_well_get(struct drm_device *dev, struct i915_power_well *power_well) { - if (!power_well->count++ && power_well->set) + struct drm_i915_private *dev_priv = dev->dev_private; + + if (!power_well->count++ && power_well->set) { + hsw_disable_package_c8(dev_priv); power_well->set(dev, power_well, true); + } } static void __intel_power_well_put(struct drm_device *dev, struct i915_power_well *power_well) { + struct drm_i915_private *dev_priv = dev->dev_private; + WARN_ON(!power_well->count); - if (!--power_well->count && power_well->set && i915_disable_power_well) + if (!--power_well->count && power_well->set && + i915_disable_power_well) { power_well->set(dev, power_well, false); + hsw_enable_package_c8(dev_priv); + } } void intel_display_power_get(struct drm_device *dev, -- cgit v1.2.3 From 8a1874559f222efcae0c0c41b180f6e1af6b9d2e Mon Sep 17 00:00:00 2001 From: Paulo Zanoni Date: Fri, 6 Dec 2013 20:32:13 -0200 Subject: drm/i915: add initial Runtime PM functions This patch adds the initial infrastructure to allow a Runtime PM implementation that sets the device to its D3 state. The patch just adds the necessary callbacks and the initial infrastructure. We still don't have any platform that actually uses this infrastructure, we still don't call get/put in all the places we need to, and we don't have any function to save/restore the state of the registers. This is not a problem since no platform uses the code added by this patch. We have a few people simultaneously working on runtime PM, so this initial code could help everybody make their plans. V2: - Move some functions to intel_pm.c - Remove useless pm_runtime_allow() call at init - Remove useless pm_runtime_mark_last_busy() call at get - Use pm_runtime_get_sync() instead of 2 calls - Add a WARN to check if we're really awake V3: - Rebase. V4: - Don't need to call pci_{save,restore}_state and pci_set_power_sate, since they're already called by the PCI layer - Remove wrong pm_runtime_enable() call at init_runtime_pm Signed-off-by: Paulo Zanoni Reviewed-by: Rodrigo Vivi Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 55 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 55 insertions(+) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index cd3f511847ec..2590a5c90725 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -31,6 +31,7 @@ #include "../../../platform/x86/intel_ips.h" #include #include +#include /** * RC6 is a special power stage which allows the GPU to enter an very @@ -5961,6 +5962,60 @@ void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv) hsw_enable_package_c8(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; + + pm_runtime_get_sync(device); + WARN(dev_priv->pm.suspended, "Device still suspended.\n"); +} + +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; + + dev_priv->pm.suspended = false; + + if (!HAS_RUNTIME_PM(dev)) + return; + + pm_runtime_set_active(device); + + pm_runtime_set_autosuspend_delay(device, 10000); /* 10s */ + pm_runtime_mark_last_busy(device); + pm_runtime_use_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; + + /* Make sure we're not suspended first. */ + pm_runtime_get_sync(device); + pm_runtime_disable(device); +} + /* Set up chip specific power management-related functions */ void intel_init_pm(struct drm_device *dev) { -- cgit v1.2.3 From d5e8fdc8c10bdff9b8af9cc8b25607ae71e26d3b Mon Sep 17 00:00:00 2001 From: Paulo Zanoni Date: Wed, 11 Dec 2013 18:50:09 -0200 Subject: drm/i915: extract hsw_power_well_post_{enable, disable} I want to add more code to the post_enable function. Signed-off-by: Paulo Zanoni Reviewed-by: Damien Lespiau Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 75 ++++++++++++++++++++++++----------------- 1 file changed, 44 insertions(+), 31 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 2590a5c90725..d8fb00a2b565 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5681,12 +5681,53 @@ bool intel_display_power_enabled(struct drm_device *dev, return is_enabled; } +static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv) +{ + struct drm_device *dev = dev_priv->dev; + unsigned long irqflags; + + if (IS_BROADWELL(dev)) { + spin_lock_irqsave(&dev_priv->irq_lock, irqflags); + I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_B), + dev_priv->de_irq_mask[PIPE_B]); + I915_WRITE(GEN8_DE_PIPE_IER(PIPE_B), + ~dev_priv->de_irq_mask[PIPE_B] | + GEN8_PIPE_VBLANK); + I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_C), + dev_priv->de_irq_mask[PIPE_C]); + I915_WRITE(GEN8_DE_PIPE_IER(PIPE_C), + ~dev_priv->de_irq_mask[PIPE_C] | + GEN8_PIPE_VBLANK); + POSTING_READ(GEN8_DE_PIPE_IER(PIPE_C)); + spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); + } +} + +static void hsw_power_well_post_disable(struct drm_i915_private *dev_priv) +{ + struct drm_device *dev = dev_priv->dev; + enum pipe p; + unsigned long irqflags; + + /* + * After this, the registers on the pipes that are part of the power + * well will become zero, so we have to adjust our counters according to + * that. + * + * FIXME: Should we do this in general in drm_vblank_post_modeset? + */ + spin_lock_irqsave(&dev->vbl_lock, irqflags); + for_each_pipe(p) + if (p != PIPE_A) + dev->vblank[p].last = 0; + spin_unlock_irqrestore(&dev->vbl_lock, irqflags); +} + static void hsw_set_power_well(struct drm_device *dev, struct i915_power_well *power_well, bool enable) { struct drm_i915_private *dev_priv = dev->dev_private; bool is_enabled, enable_requested; - unsigned long irqflags; uint32_t tmp; WARN_ON(dev_priv->pc8.enabled); @@ -5707,42 +5748,14 @@ static void hsw_set_power_well(struct drm_device *dev, DRM_ERROR("Timeout enabling power well\n"); } - if (IS_BROADWELL(dev)) { - spin_lock_irqsave(&dev_priv->irq_lock, irqflags); - I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_B), - dev_priv->de_irq_mask[PIPE_B]); - I915_WRITE(GEN8_DE_PIPE_IER(PIPE_B), - ~dev_priv->de_irq_mask[PIPE_B] | - GEN8_PIPE_VBLANK); - I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_C), - dev_priv->de_irq_mask[PIPE_C]); - I915_WRITE(GEN8_DE_PIPE_IER(PIPE_C), - ~dev_priv->de_irq_mask[PIPE_C] | - GEN8_PIPE_VBLANK); - POSTING_READ(GEN8_DE_PIPE_IER(PIPE_C)); - spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags); - } + hsw_power_well_post_enable(dev_priv); } else { if (enable_requested) { - enum pipe p; - I915_WRITE(HSW_PWR_WELL_DRIVER, 0); POSTING_READ(HSW_PWR_WELL_DRIVER); DRM_DEBUG_KMS("Requesting to disable the power well\n"); - /* - * After this, the registers on the pipes that are part - * of the power well will become zero, so we have to - * adjust our counters according to that. - * - * FIXME: Should we do this in general in - * drm_vblank_post_modeset? - */ - spin_lock_irqsave(&dev->vbl_lock, irqflags); - for_each_pipe(p) - if (p != PIPE_A) - dev->vblank[p].last = 0; - spin_unlock_irqrestore(&dev->vbl_lock, irqflags); + hsw_power_well_post_disable(dev_priv); } } } -- cgit v1.2.3 From f9dcb0dfee98406c9c64e1aad10af427d644b78f Mon Sep 17 00:00:00 2001 From: Paulo Zanoni Date: Wed, 11 Dec 2013 18:50:10 -0200 Subject: drm/i915: touch VGA MSR after we enable the power well Fixes regression introduced by: commit bf51d5e2cda5d36d98e4b46ac7fca9461e512c41 Author: Paulo Zanoni Date: Wed Jul 3 17:12:13 2013 -0300 drm/i915: switch disable_power_well default value to 1 The bug I'm seeing can be reproduced with: - Have vgacon configured/enabled - Make sure the power well gets disabled, then enabled. You can check this by seeing the messages print by hsw_set_power_well - Stop your display manager - echo 0 > /sys/class/vtconsole/vtcon1/bind I can easily reproduce this by blacklising snd_hda_intel and booting with eDP+HDMI. If you do this and then look at dmesg, you'll see we're printing infinite "Unclaimed register" messages. This is happening because we're stuck on an infinite loop inside console_unlock(), which is calling many functions from vgacon.c. And the code that's triggering the error messages is from vgacon_set_cursor_size(). After we re-enable the power well, every time we read/write the VGA address 0x3d5 we get an "unclaimed register" interrupt (ERR_INT) and print error messages. If we write anything to the VGA MSR register (it doesn't really matter which value you write to bit 0), any reads/writes to 0x3d5 _don't_ trigger the "unclaimed register" errors anymore (even if MSR bit 0 is zero). So what happens with the current code is that when we unbind i915 and bind vgacon, we call console_unlock(). Function console_unlock() is responsible for printing any messages that were supposed to be print when the console was locked, so it calls the TTY layer, which calls the console layer, which calls vgacon to print the messages. At this point, vgacon eventually calls vgacon_set_cursor_size(), which touches 0x3d5, which triggers unclaimed register interrupts. The problem is that when we get these interrupts, we print the error messages, so we add more work to console_unlock(), which will try to print it again, and then call vgacon again, trigger a new interrupt, which will put more stuff to the buffer, and then we'll be stuck at console_unlock() forever. If you patch intel_uncore.c to not print anything when we detect unclaimed registers, we won't get into the console_unlock() infinite loop and the driver unbind will work just fine. We will still be getting interrupts every time vgacon touches those registers, but we will survive. This is a valid experiment, but IMHO it's not the real fix: if we don't print any error messages we will still keep getting the interrupts, and if we disable ERR_INT we won't get the interrupt anymore, but we will also stop getting all the other error interrupts. I talked about this problem with the HW engineer and his recommendation is "So don't do any VGA I/O or memory access while the power well is disabled, and make to re-program MSR after enabling the power well and before using VGA I/O or memory accesses.". Notice that this is just a partial fix to fd.o #67813. This fixes the case where the power well is already enabled when we unbind, not when it's disabled when we unbind. V2: - Rebase (first version was sent in September). V3: - Complete rewrite of the same fix: smaller implementation, improved commit message. Testcase: igt/drv_module_reload Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=67813 Signed-off-by: Paulo Zanoni Reviewed-by: Damien Lespiau Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index d8fb00a2b565..3cd521f3823b 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -30,6 +30,7 @@ #include "intel_drv.h" #include "../../../platform/x86/intel_ips.h" #include +#include #include #include @@ -5686,6 +5687,20 @@ static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv) struct drm_device *dev = dev_priv->dev; unsigned long irqflags; + /* + * After we re-enable the power well, if we touch VGA register 0x3d5 + * we'll get unclaimed register interrupts. This stops after we write + * anything to the VGA MSR register. The vgacon module uses this + * register all the time, so if we unbind our driver and, as a + * consequence, bind vgacon, we'll get stuck in an infinite loop at + * console_unlock(). So make here we touch the VGA MSR register, making + * sure vgacon can keep working normally without triggering interrupts + * and error messages. + */ + vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO); + outb(inb(VGA_MSR_READ), VGA_MSR_WRITE); + vga_put(dev->pdev, VGA_RSRC_LEGACY_IO); + if (IS_BROADWELL(dev)) { spin_lock_irqsave(&dev_priv->irq_lock, irqflags); I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_B), -- cgit v1.2.3 From 42a430f51c7500be41ca4cbd5b3de930853bd5ee Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 28 Nov 2013 17:29:56 +0200 Subject: drm/i915: Gen2 FBC1 CFB pitch wants 32B units MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On gen2 the compressed frame buffer pitch is specified in 32B units rather than the 64B units used on gen3+. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 3cd521f3823b..599d445f7ce9 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -104,8 +104,11 @@ static void i8xx_enable_fbc(struct drm_crtc *crtc, unsigned long interval) if (fb->pitches[0] < cfb_pitch) cfb_pitch = fb->pitches[0]; - /* FBC_CTL wants 64B units */ - cfb_pitch = (cfb_pitch / 64) - 1; + /* FBC_CTL wants 32B or 64B units */ + if (IS_GEN2(dev)) + cfb_pitch = (cfb_pitch / 32) - 1; + else + cfb_pitch = (cfb_pitch / 64) - 1; plane = intel_crtc->plane == 0 ? FBC_CTL_PLANEA : FBC_CTL_PLANEB; /* Clear old tags */ -- cgit v1.2.3 From 159f98750e413f5d6e63d1131417c5310ea175fb Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 28 Nov 2013 17:29:57 +0200 Subject: drm/i915: FBC_CONTROL2 is gen4 only MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Gen2 and gen3 don't have the FBC_CONTROL2 register, so don't touch it. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 599d445f7ce9..d7cb6bf5712d 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -98,7 +98,7 @@ static void i8xx_enable_fbc(struct drm_crtc *crtc, unsigned long interval) struct intel_crtc *intel_crtc = to_intel_crtc(crtc); int cfb_pitch; int plane, i; - u32 fbc_ctl, fbc_ctl2; + u32 fbc_ctl; cfb_pitch = dev_priv->fbc.size / FBC_LL_SIZE; if (fb->pitches[0] < cfb_pitch) @@ -115,11 +115,15 @@ static void i8xx_enable_fbc(struct drm_crtc *crtc, unsigned long interval) 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 |= plane; + I915_WRITE(FBC_CONTROL2, fbc_ctl2); + I915_WRITE(FBC_FENCE_OFF, crtc->y); + } /* enable it... */ fbc_ctl = FBC_CTL_EN | FBC_CTL_PERIODIC; -- cgit v1.2.3 From 993495ae992c91e98ade8fbe977c57bfd81910c1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Dec 2013 17:27:40 +0200 Subject: drm/i915: Rework the FBC interval/stall stuff a bit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Don't touch DPFC_RECOMP_CTL on FBC2, use RMW to update the FBC_CONTROL on FBC1 to make it easier for people to experiment with different numbers. Also fix the interval mask for FBC1. v2: Rebased Reviewed-by: Imre Deak Signed-off-by: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 34 ++++++++++++++-------------------- 1 file changed, 14 insertions(+), 20 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index d7cb6bf5712d..c68abf718825 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -88,7 +88,7 @@ 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; @@ -126,11 +126,12 @@ static void i8xx_enable_fbc(struct drm_crtc *crtc, unsigned long interval) } /* 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); @@ -145,7 +146,7 @@ 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; @@ -154,16 +155,12 @@ static void g4x_enable_fbc(struct drm_crtc *crtc, unsigned long interval) 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_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... */ @@ -219,7 +216,7 @@ static void sandybridge_blit_fbc_update(struct drm_device *dev) 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; @@ -228,7 +225,6 @@ static void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval) 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); @@ -241,9 +237,6 @@ static void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval) dpfc_ctl |= 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)); I915_WRITE(ILK_DPFC_FENCE_YOFF, crtc->y); I915_WRITE(ILK_FBC_RT_BASE, i915_gem_obj_ggtt_offset(obj) | ILK_FBC_RT_VALID); /* enable it... */ @@ -281,7 +274,7 @@ static bool ironlake_fbc_enabled(struct drm_device *dev) return I915_READ(ILK_DPFC_CONTROL) & DPFC_CTL_EN; } -static void gen7_enable_fbc(struct drm_crtc *crtc, unsigned long interval) +static void gen7_enable_fbc(struct drm_crtc *crtc) { struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; @@ -338,8 +331,7 @@ static void intel_fbc_work_fn(struct work_struct *__work) * the prior work. */ if (work->crtc->fb == work->fb) { - dev_priv->display.enable_fbc(work->crtc, - work->interval); + dev_priv->display.enable_fbc(work->crtc); dev_priv->fbc.plane = to_intel_crtc(work->crtc)->plane; dev_priv->fbc.fb_id = work->crtc->fb->base.id; @@ -376,7 +368,7 @@ static void intel_cancel_fbc_work(struct drm_i915_private *dev_priv) dev_priv->fbc.fbc_work = NULL; } -static void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval) +static void intel_enable_fbc(struct drm_crtc *crtc) { struct intel_fbc_work *work; struct drm_device *dev = crtc->dev; @@ -390,13 +382,12 @@ static void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval) work = kzalloc(sizeof(*work), GFP_KERNEL); if (work == NULL) { DRM_ERROR("Failed to allocate FBC work structure\n"); - dev_priv->display.enable_fbc(crtc, interval); + dev_priv->display.enable_fbc(crtc); return; } work->crtc = crtc; work->fb = crtc->fb; - work->interval = interval; INIT_DELAYED_WORK(&work->work, intel_fbc_work_fn); dev_priv->fbc.fbc_work = work; @@ -611,7 +602,7 @@ void intel_update_fbc(struct drm_device *dev) intel_disable_fbc(dev); } - intel_enable_fbc(crtc, 500); + intel_enable_fbc(crtc); dev_priv->fbc.no_fbc_reason = FBC_OK; return; @@ -6073,6 +6064,9 @@ void intel_init_pm(struct drm_device *dev) 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); } } -- cgit v1.2.3 From 63801f211c6eeb6def635ceee39d165e00fd6e09 Mon Sep 17 00:00:00 2001 From: Ben Widawsky Date: Thu, 12 Dec 2013 17:26:03 -0800 Subject: drm/i915/bdw: Force all Data Cache Data Port access to be Non-Coherent I stumbled on to some unimplemented errata. To be honest, I am not really sure of the impact, just that the docs say to do. No w/a name for this one. v2: v1 was a stale thing which should have never seen the light of day. (Haihao) Cc: Kenneth Graunke Signed-off-by: Ben Widawsky Reviewed-by: Damien Lespiau Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index c68abf718825..791fbe386b7c 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5269,6 +5269,14 @@ static void gen8_init_clock_gating(struct drm_device *dev) I915_READ(CHICKEN_PIPESL_1(i) | 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)); } static void haswell_init_clock_gating(struct drm_device *dev) -- cgit v1.2.3 From ab57fff1302c485d74992d34df24ccb5efda244e Mon Sep 17 00:00:00 2001 From: Ben Widawsky Date: Thu, 12 Dec 2013 15:28:04 -0800 Subject: drm/i915/bdw: Implement ff workarounds WaVSRefCountFullforceMissDisable and WaDSRefCountFullforceMissDisable VS is a carry-over from HSW, and DS is likely not used by anyone yet. Cc: Kenneth Graunke Signed-off-by: Ben Widawsky Reviewed-by: Damien Lespiau [danvet: Line of 106 chars is too long. Really.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 791fbe386b7c..b35f65ed6c5e 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5256,14 +5256,14 @@ static void gen8_init_clock_gating(struct drm_device *dev) I915_WRITE(GEN7_HALF_SLICE_CHICKEN1, _MASKED_BIT_ENABLE(GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE)); - /* WaSwitchSolVfFArbitrationPriority */ + /* WaSwitchSolVfFArbitrationPriority:bdw */ I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL); - /* WaPsrDPAMaskVBlankInSRD */ + /* WaPsrDPAMaskVBlankInSRD:bdw */ I915_WRITE(CHICKEN_PAR1_1, I915_READ(CHICKEN_PAR1_1) | DPA_MASK_VBLANK_SRD); - /* WaPsrDPRSUnmaskVBlankInSRD */ + /* WaPsrDPRSUnmaskVBlankInSRD:bdw */ for_each_pipe(i) { I915_WRITE(CHICKEN_PIPESL_1(i), I915_READ(CHICKEN_PIPESL_1(i) | @@ -5277,6 +5277,12 @@ static void gen8_init_clock_gating(struct drm_device *dev) 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)); } static void haswell_init_clock_gating(struct drm_device *dev) -- cgit v1.2.3 From 691bb717544f3c1241f5a4e5fb884a0215a2dc9e Mon Sep 17 00:00:00 2001 From: Damien Lespiau Date: Thu, 12 Dec 2013 14:36:36 +0000 Subject: drm/i915: Use IS_VALLEYVIEW() to test the is_valleyview flag Signed-off-by: Damien Lespiau Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index b35f65ed6c5e..7fd3e67244ec 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -3582,9 +3582,11 @@ void gen6_set_rps(struct drm_device *dev, u8 val) void gen6_rps_idle(struct drm_i915_private *dev_priv) { + struct drm_device *dev = dev_priv->dev; + mutex_lock(&dev_priv->rps.hw_lock); if (dev_priv->rps.enabled) { - if (dev_priv->info->is_valleyview) + if (IS_VALLEYVIEW(dev)) valleyview_set_rps(dev_priv->dev, dev_priv->rps.min_delay); else gen6_set_rps(dev_priv->dev, dev_priv->rps.min_delay); @@ -3595,9 +3597,11 @@ void gen6_rps_idle(struct drm_i915_private *dev_priv) void gen6_rps_boost(struct drm_i915_private *dev_priv) { + struct drm_device *dev = dev_priv->dev; + mutex_lock(&dev_priv->rps.hw_lock); if (dev_priv->rps.enabled) { - if (dev_priv->info->is_valleyview) + if (IS_VALLEYVIEW(dev)) valleyview_set_rps(dev_priv->dev, dev_priv->rps.max_delay); else gen6_set_rps(dev_priv->dev, dev_priv->rps.max_delay); -- cgit v1.2.3 From ac9545fda6ed13c3a1205de138fa25c076cf4473 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:28 +0200 Subject: drm/i915: Add IVB DDB partitioning control MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On IVB the display data buffer partitioning control lives in the DISP_ARB_CTL2 register. Add the relevant defines/code for it. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 30 ++++++++++++++++++++++-------- 1 file changed, 22 insertions(+), 8 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 7fd3e67244ec..be8244282742 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2864,6 +2864,7 @@ static unsigned int ilk_compute_wm_dirty(struct drm_device *dev, static void hsw_write_wm_values(struct drm_i915_private *dev_priv, struct hsw_wm_values *results) { + struct drm_device *dev = dev_priv->dev; struct hsw_wm_values *previous = &dev_priv->wm.hw; unsigned int dirty; uint32_t val; @@ -2894,12 +2895,21 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv, I915_WRITE(PIPE_WM_LINETIME(PIPE_C), results->wm_linetime[2]); if (dirty & WM_DIRTY_DDB) { - val = I915_READ(WM_MISC); - if (results->partitioning == INTEL_DDB_PART_1_2) - val &= ~WM_MISC_DATA_PARTITION_5_6; - else - val |= WM_MISC_DATA_PARTITION_5_6; - I915_WRITE(WM_MISC, val); + if (IS_HASWELL(dev)) { + val = I915_READ(WM_MISC); + if (results->partitioning == INTEL_DDB_PART_1_2) + val &= ~WM_MISC_DATA_PARTITION_5_6; + else + val |= WM_MISC_DATA_PARTITION_5_6; + I915_WRITE(WM_MISC, val); + } else { + val = I915_READ(DISP_ARB_CTL2); + if (results->partitioning == INTEL_DDB_PART_1_2) + val &= ~DISP_DATA_PARTITION_5_6; + else + val |= DISP_DATA_PARTITION_5_6; + I915_WRITE(DISP_ARB_CTL2, val); + } } if (dirty & WM_DIRTY_FBC) { @@ -3210,8 +3220,12 @@ void ilk_wm_get_hw_state(struct drm_device *dev) hw->wm_lp_spr[1] = I915_READ(WM2S_LP_IVB); hw->wm_lp_spr[2] = I915_READ(WM3S_LP_IVB); - hw->partitioning = (I915_READ(WM_MISC) & WM_MISC_DATA_PARTITION_5_6) ? - INTEL_DDB_PART_5_6 : INTEL_DDB_PART_1_2; + if (IS_HASWELL(dev)) + 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); -- cgit v1.2.3 From a68d68eebc9820212186b6d02f2dce09deef0e53 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:29 +0200 Subject: drm/i915: Add ILK/SNB/IVB WM latency field support MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add a new function ilk_wm_lp_latency() which will tell us what to write into the WM_LPx register latency field. HSW is different from erlier gens in this regard. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index be8244282742..dbd025a4f22f 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2715,6 +2715,17 @@ static int ilk_wm_lp_to_level(int wm_lp, const struct intel_pipe_wm *pipe_wm) return wm_lp + (wm_lp >= 2 && pipe_wm->wm[4].enable); } +/* 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)) + return 2 * level; + else + return dev_priv->wm.pri_latency[level]; +} + static void hsw_compute_wm_results(struct drm_device *dev, const struct intel_pipe_wm *merged, enum intel_ddb_partitioning partitioning, @@ -2737,7 +2748,7 @@ static void hsw_compute_wm_results(struct drm_device *dev, break; results->wm_lp[wm_lp - 1] = WM3_LP_EN | - ((level * 2) << WM1_LP_LATENCY_SHIFT) | + (ilk_wm_lp_latency(dev, level) << WM1_LP_LATENCY_SHIFT) | (r->pri_val << WM1_LP_SR_SHIFT) | r->cur_val; -- cgit v1.2.3 From 7b39a0b791a356d09dd2517ec04b1bf9cdced4ab Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:30 +0200 Subject: drm/i915: Avoid computing invalid WM levels when sprites/scaling is enabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On ILK/SNB only LP0/1 watermarks can be enabled when sprites are enabled, and on ILK/SNB/IVB sprite scaling is limited to LP0 only. So we can avoid computing the extra levels we're never going to use. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index dbd025a4f22f..b06076d8f386 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2643,6 +2643,14 @@ static bool intel_compute_pipe_wm(struct drm_crtc *crtc, /* LP0 watermarks always use 1/2 DDB partitioning */ ilk_compute_wm_maximums(dev, 0, &config, INTEL_DDB_PART_1_2, &max); + /* 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; + for (level = 0; level <= max_level; level++) ilk_compute_wm_level(dev_priv, level, params, &pipe_wm->wm[level]); -- cgit v1.2.3 From 6cef2b8a5671dc6abbc0414a94be843590f93ede Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:32 +0200 Subject: drm/i915: Fix LP1 sprite watermarks for ILK/SNB MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ILK/SNB don't have LP2+ watermarks for sprites. Also the LP1 sprite watermark register has its own enable bit. Take these differences into account when programming the LP1+ registers. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 27 ++++++++++++++++++++------- 1 file changed, 20 insertions(+), 7 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index b06076d8f386..232d8c3b1f10 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2767,7 +2767,11 @@ static void hsw_compute_wm_results(struct drm_device *dev, results->wm_lp[wm_lp - 1] |= r->fbc_val << WM1_LP_FBC_SHIFT; - results->wm_lp_spr[wm_lp - 1] = r->spr_val; + if (INTEL_INFO(dev)->gen <= 6 && r->spr_val) { + WARN_ON(wm_lp != 1); + results->wm_lp_spr[wm_lp - 1] = WM1S_LP_EN | r->spr_val; + } else + results->wm_lp_spr[wm_lp - 1] = r->spr_val; } /* LP0 register values */ @@ -2899,6 +2903,10 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv, if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] != 0) I915_WRITE(WM1_LP_ILK, 0); + if (INTEL_INFO(dev)->gen <= 6 && + dirty & WM_DIRTY_LP(1) && previous->wm_lp_spr[0] != 0) + I915_WRITE(WM1S_LP_ILK, 0); + if (dirty & WM_DIRTY_PIPE(PIPE_A)) I915_WRITE(WM0_PIPEA_ILK, results->wm_pipe[0]); if (dirty & WM_DIRTY_PIPE(PIPE_B)) @@ -2940,12 +2948,17 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv, I915_WRITE(DISP_ARB_CTL, val); } - if (dirty & WM_DIRTY_LP(1) && previous->wm_lp_spr[0] != results->wm_lp_spr[0]) - I915_WRITE(WM1S_LP_ILK, results->wm_lp_spr[0]); - if (dirty & WM_DIRTY_LP(2) && previous->wm_lp_spr[1] != results->wm_lp_spr[1]) - I915_WRITE(WM2S_LP_IVB, results->wm_lp_spr[1]); - if (dirty & WM_DIRTY_LP(3) && previous->wm_lp_spr[2] != results->wm_lp_spr[2]) - I915_WRITE(WM3S_LP_IVB, results->wm_lp_spr[2]); + if (INTEL_INFO(dev)->gen <= 6) { + if (dirty & WM_DIRTY_LP(1) && results->wm_lp_spr[0] != 0) + I915_WRITE(WM1S_LP_ILK, results->wm_lp_spr[0]); + } else { + if (dirty & WM_DIRTY_LP(1) && previous->wm_lp_spr[0] != results->wm_lp_spr[0]) + I915_WRITE(WM1S_LP_ILK, results->wm_lp_spr[0]); + if (dirty & WM_DIRTY_LP(2) && previous->wm_lp_spr[1] != results->wm_lp_spr[1]) + I915_WRITE(WM2S_LP_IVB, results->wm_lp_spr[1]); + if (dirty & WM_DIRTY_LP(3) && previous->wm_lp_spr[2] != results->wm_lp_spr[2]) + I915_WRITE(WM3S_LP_IVB, results->wm_lp_spr[2]); + } if (dirty & WM_DIRTY_LP(1) && results->wm_lp[0] != 0) I915_WRITE(WM1_LP_ILK, results->wm_lp[0]); -- cgit v1.2.3 From facd619b8869b308e02104a200abf6f9d7cddcab Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:33 +0200 Subject: drm/i915: Fix LP1+ watermark disabling ILK MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On ILK disabling LP1+ watermarks must be done carefully to avoid underruns. If we just write 0 to the register in the middle of the scan cycle we often get an underrun. So instead we have to leave the actual watermark levels in the register intact, and just toggle the enable bit. Presumably the hardware takes a while to get out of low power mode, and so the watermark level need to stay valid until that time. We also have to be careful with the WM1S_LP_EN bit. It seems the hardware more or less treats it like the actual watermarks numbers, and so we must not toggle it too soon. Just leave it alone when disabling the LP1+ watermarks. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 33 ++++++++++++++++++++------------- 1 file changed, 20 insertions(+), 13 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 232d8c3b1f10..c43cf138c871 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2896,16 +2896,23 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv, if (!dirty) return; - if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] != 0) - I915_WRITE(WM3_LP_ILK, 0); - if (dirty & WM_DIRTY_LP(2) && previous->wm_lp[1] != 0) - I915_WRITE(WM2_LP_ILK, 0); - if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] != 0) - I915_WRITE(WM1_LP_ILK, 0); + 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]); + } + 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]); + } + 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]); + } - if (INTEL_INFO(dev)->gen <= 6 && - dirty & WM_DIRTY_LP(1) && previous->wm_lp_spr[0] != 0) - I915_WRITE(WM1S_LP_ILK, 0); + /* + * Don't touch WM1S_LP_EN here. + * Doing so could cause underruns. + */ if (dirty & WM_DIRTY_PIPE(PIPE_A)) I915_WRITE(WM0_PIPEA_ILK, results->wm_pipe[0]); @@ -2949,7 +2956,7 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv, } if (INTEL_INFO(dev)->gen <= 6) { - if (dirty & WM_DIRTY_LP(1) && results->wm_lp_spr[0] != 0) + if (dirty & WM_DIRTY_LP(1) && previous->wm_lp_spr[0] != results->wm_lp_spr[0]) I915_WRITE(WM1S_LP_ILK, results->wm_lp_spr[0]); } else { if (dirty & WM_DIRTY_LP(1) && previous->wm_lp_spr[0] != results->wm_lp_spr[0]) @@ -2960,11 +2967,11 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv, I915_WRITE(WM3S_LP_IVB, results->wm_lp_spr[2]); } - if (dirty & WM_DIRTY_LP(1) && results->wm_lp[0] != 0) + if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] != results->wm_lp[0]) I915_WRITE(WM1_LP_ILK, results->wm_lp[0]); - if (dirty & WM_DIRTY_LP(2) && results->wm_lp[1] != 0) + if (dirty & WM_DIRTY_LP(2) && previous->wm_lp[1] != results->wm_lp[1]) I915_WRITE(WM2_LP_ILK, results->wm_lp[1]); - if (dirty & WM_DIRTY_LP(3) && results->wm_lp[2] != 0) + if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] != results->wm_lp[2]) I915_WRITE(WM3_LP_ILK, results->wm_lp[2]); dev_priv->wm.hw = *results; -- cgit v1.2.3 From 0ba22e26fe47b2a216e5438292aeeb8e015e9d64 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:34 +0200 Subject: drm/i915: Don't merge LP1+ watermarks on ILK/SNB/IVB when multiple pipes are enabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Multi-pipe LP1+ watermarks are a HSW+ feature, so let's not do it on earlier generations. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index c43cf138c871..a65d8816c1e6 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2690,11 +2690,17 @@ static void ilk_merge_wm_level(struct drm_device *dev, * 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 hsw_wm_maximums *max, struct intel_pipe_wm *merged) { int level, max_level = ilk_wm_max_level(dev); + /* ILK/SNB/IVB: LP1+ watermarks only w/ single pipe */ + if ((INTEL_INFO(dev)->gen <= 6 || IS_IVYBRIDGE(dev)) && + config->num_pipes_active > 1) + return; + merged->fbc_wm_enabled = true; /* merge each WM1+ level */ @@ -3000,13 +3006,13 @@ static void haswell_update_wm(struct drm_crtc *crtc) intel_crtc->wm.active = pipe_wm; ilk_compute_wm_maximums(dev, 1, &config, INTEL_DDB_PART_1_2, &max); - ilk_wm_merge(dev, &max, &lp_wm_1_2); + ilk_wm_merge(dev, &config, &max, &lp_wm_1_2); /* 5/6 split only in single pipe config on IVB+ */ if (INTEL_INFO(dev)->gen >= 7 && config.num_pipes_active == 1 && config.sprites_enabled) { ilk_compute_wm_maximums(dev, 1, &config, INTEL_DDB_PART_5_6, &max); - ilk_wm_merge(dev, &max, &lp_wm_5_6); + ilk_wm_merge(dev, &config, &max, &lp_wm_5_6); best_lp_wm = hsw_find_best_result(dev, &lp_wm_1_2, &lp_wm_5_6); } else { -- cgit v1.2.3 From 6c8b6c288783b05733de31fb61fc8ebfa8ae0229 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:35 +0200 Subject: drm/i915: Disable FBC WM on ILK, and disable LP2+ when FBC is enabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ILK has a bunch of issues with FBC. First of all, BSpec tells us that FBC WM should never be enabled. Secondly when FBC is enabled with FBC WM disabled, LP2+ watermarks must be disabled. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index a65d8816c1e6..3d4daa1fc3be 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2701,7 +2701,8 @@ static void ilk_wm_merge(struct drm_device *dev, config->num_pipes_active > 1) return; - merged->fbc_wm_enabled = true; + /* 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++) { @@ -2721,6 +2722,20 @@ static void ilk_wm_merge(struct drm_device *dev, wm->fbc_val = 0; } } + + /* ILK: LP2+ must be disabled when FBC WM is disabled but FBC enabled */ + /* + * FIXME this is racy. FBC might get enabled later. + * What we should check here is whether FBC can be + * enabled sometime later. + */ + if (IS_GEN5(dev) && !merged->fbc_wm_enabled && intel_fbc_enabled(dev)) { + for (level = 2; level <= max_level; level++) { + struct intel_wm_level *wm = &merged->wm[level]; + + wm->enable = false; + } + } } static int ilk_wm_lp_to_level(int wm_lp, const struct intel_pipe_wm *pipe_wm) -- cgit v1.2.3 From ce0e0713a6b21ec611d4a2eb5c60d18dbb4bf479 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:36 +0200 Subject: drm/i915: Linetime watermarks are a HSW feature MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Linetime watermarks don't exist on ILK/SNB/IVB, so don't compute them except on HSW. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 3d4daa1fc3be..276e98df1c0d 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2655,7 +2655,8 @@ static bool intel_compute_pipe_wm(struct drm_crtc *crtc, ilk_compute_wm_level(dev_priv, level, params, &pipe_wm->wm[level]); - pipe_wm->linetime = hsw_compute_linetime_wm(dev, crtc); + if (IS_HASWELL(dev)) + pipe_wm->linetime = hsw_compute_linetime_wm(dev, crtc); /* At least LP0 must be valid */ return ilk_validate_wm_level(0, &max, &pipe_wm->wm[0]); @@ -3234,7 +3235,8 @@ static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc) }; hw->wm_pipe[pipe] = I915_READ(wm0_pipe_reg[pipe]); - hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe)); + if (IS_HASWELL(dev)) + hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe)); if (intel_crtc_active(crtc)) { u32 tmp = hw->wm_pipe[pipe]; -- cgit v1.2.3 From 017636cc09e2c58bc493cc1fe74d858f50d173af Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:37 +0200 Subject: drm/i915: Disable LP1+ watermarks safely in init MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ILK doesn't like if we just write the LP1+ watermarks registers with 0. We need to just disable the watermarks by clearing the enable bit. Use that method also when disabling LP1+ watermarks in init_clock_gating. It looks like disabling the sprite LP1 watermarks can cause underruns even if we just toggle the WM1S_LP_EN bit. So treat that bit like the actual watermark numbers and avoid setting it to 0 immediately. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 31 +++++++++++++++++++------------ 1 file changed, 19 insertions(+), 12 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 276e98df1c0d..e321fbc28a0b 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5051,6 +5051,20 @@ static void g4x_disable_trickle_feed(struct drm_device *dev) } } +static void ilk_init_lp_watermarks(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = dev->dev_private; + + I915_WRITE(WM3_LP_ILK, I915_READ(WM3_LP_ILK) & ~WM1_LP_SR_EN); + I915_WRITE(WM2_LP_ILK, I915_READ(WM2_LP_ILK) & ~WM1_LP_SR_EN); + I915_WRITE(WM1_LP_ILK, I915_READ(WM1_LP_ILK) & ~WM1_LP_SR_EN); + + /* + * Don't touch WM1S_LP_EN here. + * Doing so could cause underruns. + */ +} + static void ironlake_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -5084,9 +5098,8 @@ static void ironlake_init_clock_gating(struct drm_device *dev) I915_WRITE(DISP_ARB_CTL, (I915_READ(DISP_ARB_CTL) | DISP_FBC_WM_DIS)); - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); + + ilk_init_lp_watermarks(dev); /* * Based on the document from hardware guys the following bits @@ -5193,9 +5206,7 @@ static void gen6_init_clock_gating(struct drm_device *dev) I915_WRITE(GEN6_GT_MODE, _MASKED_BIT_ENABLE(GEN6_TD_FOUR_ROW_DISPATCH_DISABLE)); - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); + ilk_init_lp_watermarks(dev); I915_WRITE(CACHE_MODE_0, _MASKED_BIT_DISABLE(CM0_STC_EVICT_DISABLE_LRA_SNB)); @@ -5369,9 +5380,7 @@ static void haswell_init_clock_gating(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); + ilk_init_lp_watermarks(dev); /* According to the spec, bit 13 (RCZUNIT) must be set on IVB. * This implements the WaDisableRCZUnitClockGating:hsw workaround. @@ -5420,9 +5429,7 @@ static void ivybridge_init_clock_gating(struct drm_device *dev) struct drm_i915_private *dev_priv = dev->dev_private; uint32_t snpcr; - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); + ilk_init_lp_watermarks(dev); I915_WRITE(ILK_DSPCLK_GATE_D, ILK_VRHUNIT_CLOCK_GATE_DISABLE); -- cgit v1.2.3 From 96f90c5421aaaba35ffd931728fabeb8f6cd8471 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:38 +0200 Subject: drm/i915: Move ILK/SNB/IVB over to the HSW WM code MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The new HSW watermark code can now handle ILK/SNB/IVB as well, so switch them over. Kill the old code. Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 656 +--------------------------------------- 1 file changed, 11 insertions(+), 645 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index e321fbc28a0b..7e028341b2f5 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -965,65 +965,6 @@ 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 @@ -1707,423 +1648,6 @@ static void i830_update_wm(struct drm_crtc *unused_crtc) I915_WRITE(FW_BLC, fwater_lo); } -/* - * Check the wm result. - * - * If any calculated watermark values is larger than the maximum value that - * can be programmed into the associated watermark register, that watermark - * must be disabled. - */ -static bool ironlake_check_srwm(struct drm_device *dev, int level, - int fbc_wm, int display_wm, int cursor_wm, - const struct intel_watermark_params *display, - const struct intel_watermark_params *cursor) -{ - struct drm_i915_private *dev_priv = dev->dev_private; - - DRM_DEBUG_KMS("watermark %d: display plane %d, fbc lines %d," - " cursor %d\n", level, display_wm, fbc_wm, cursor_wm); - - if (fbc_wm > SNB_FBC_MAX_SRWM) { - DRM_DEBUG_KMS("fbc watermark(%d) is too large(%d), disabling wm%d+\n", - fbc_wm, SNB_FBC_MAX_SRWM, level); - - /* fbc has it's own way to disable FBC WM */ - I915_WRITE(DISP_ARB_CTL, - I915_READ(DISP_ARB_CTL) | DISP_FBC_WM_DIS); - return false; - } else if (INTEL_INFO(dev)->gen >= 6) { - /* enable FBC WM (except on ILK, where it must remain off) */ - I915_WRITE(DISP_ARB_CTL, - I915_READ(DISP_ARB_CTL) & ~DISP_FBC_WM_DIS); - } - - if (display_wm > display->max_wm) { - DRM_DEBUG_KMS("display watermark(%d) is too large(%d), disabling wm%d+\n", - display_wm, SNB_DISPLAY_MAX_SRWM, level); - return false; - } - - if (cursor_wm > cursor->max_wm) { - DRM_DEBUG_KMS("cursor watermark(%d) is too large(%d), disabling wm%d+\n", - cursor_wm, SNB_CURSOR_MAX_SRWM, level); - return false; - } - - if (!(fbc_wm || display_wm || cursor_wm)) { - DRM_DEBUG_KMS("latency %d is 0, disabling wm%d+\n", level, level); - return false; - } - - return true; -} - -/* - * Compute watermark values of WM[1-3], - */ -static bool ironlake_compute_srwm(struct drm_device *dev, int level, int plane, - int latency_ns, - const struct intel_watermark_params *display, - const struct intel_watermark_params *cursor, - int *fbc_wm, int *display_wm, int *cursor_wm) -{ - struct drm_crtc *crtc; - const struct drm_display_mode *adjusted_mode; - unsigned long line_time_us; - int hdisplay, htotal, pixel_size, clock; - int line_count, line_size; - int small, large; - int entries; - - if (!latency_ns) { - *fbc_wm = *display_wm = *cursor_wm = 0; - return false; - } - - crtc = intel_get_crtc_for_plane(dev, plane); - adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode; - clock = adjusted_mode->crtc_clock; - htotal = adjusted_mode->htotal; - hdisplay = to_intel_crtc(crtc)->config.pipe_src_w; - pixel_size = crtc->fb->bits_per_pixel / 8; - - line_time_us = (htotal * 1000) / clock; - line_count = (latency_ns / line_time_us + 1000) / 1000; - line_size = hdisplay * pixel_size; - - /* Use the minimum of the small and large buffer method for primary */ - small = ((clock * pixel_size / 1000) * latency_ns) / 1000; - large = line_count * line_size; - - entries = DIV_ROUND_UP(min(small, large), display->cacheline_size); - *display_wm = entries + display->guard_size; - - /* - * Spec says: - * FBC WM = ((Final Primary WM * 64) / number of bytes per line) + 2 - */ - *fbc_wm = DIV_ROUND_UP(*display_wm * 64, line_size) + 2; - - /* calculate the self-refresh watermark for display cursor */ - entries = line_count * pixel_size * 64; - entries = DIV_ROUND_UP(entries, cursor->cacheline_size); - *cursor_wm = entries + cursor->guard_size; - - return ironlake_check_srwm(dev, level, - *fbc_wm, *display_wm, *cursor_wm, - display, cursor); -} - -static void ironlake_update_wm(struct drm_crtc *crtc) -{ - struct drm_device *dev = crtc->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - int fbc_wm, plane_wm, cursor_wm; - unsigned int enabled; - - enabled = 0; - if (g4x_compute_wm0(dev, PIPE_A, - &ironlake_display_wm_info, - dev_priv->wm.pri_latency[0] * 100, - &ironlake_cursor_wm_info, - dev_priv->wm.cur_latency[0] * 100, - &plane_wm, &cursor_wm)) { - I915_WRITE(WM0_PIPEA_ILK, - (plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm); - DRM_DEBUG_KMS("FIFO watermarks For pipe A -" - " plane %d, " "cursor: %d\n", - plane_wm, cursor_wm); - enabled |= 1 << PIPE_A; - } - - if (g4x_compute_wm0(dev, PIPE_B, - &ironlake_display_wm_info, - dev_priv->wm.pri_latency[0] * 100, - &ironlake_cursor_wm_info, - dev_priv->wm.cur_latency[0] * 100, - &plane_wm, &cursor_wm)) { - I915_WRITE(WM0_PIPEB_ILK, - (plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm); - DRM_DEBUG_KMS("FIFO watermarks For pipe B -" - " plane %d, cursor: %d\n", - plane_wm, cursor_wm); - enabled |= 1 << PIPE_B; - } - - /* - * Calculate and update the self-refresh watermark only when one - * display plane is used. - */ - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); - - if (!single_plane_enabled(enabled)) - return; - enabled = ffs(enabled) - 1; - - /* WM1 */ - if (!ironlake_compute_srwm(dev, 1, enabled, - dev_priv->wm.pri_latency[1] * 500, - &ironlake_display_srwm_info, - &ironlake_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - return; - - I915_WRITE(WM1_LP_ILK, - WM1_LP_SR_EN | - (dev_priv->wm.pri_latency[1] << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); - - /* WM2 */ - if (!ironlake_compute_srwm(dev, 2, enabled, - dev_priv->wm.pri_latency[2] * 500, - &ironlake_display_srwm_info, - &ironlake_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - return; - - I915_WRITE(WM2_LP_ILK, - WM2_LP_EN | - (dev_priv->wm.pri_latency[2] << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); - - /* - * WM3 is unsupported on ILK, probably because we don't have latency - * data for that power state - */ -} - -static void sandybridge_update_wm(struct drm_crtc *crtc) -{ - struct drm_device *dev = crtc->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - int latency = dev_priv->wm.pri_latency[0] * 100; /* In unit 0.1us */ - u32 val; - int fbc_wm, plane_wm, cursor_wm; - unsigned int enabled; - - enabled = 0; - if (g4x_compute_wm0(dev, PIPE_A, - &sandybridge_display_wm_info, latency, - &sandybridge_cursor_wm_info, latency, - &plane_wm, &cursor_wm)) { - val = I915_READ(WM0_PIPEA_ILK); - val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK); - I915_WRITE(WM0_PIPEA_ILK, val | - ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm)); - DRM_DEBUG_KMS("FIFO watermarks For pipe A -" - " plane %d, " "cursor: %d\n", - plane_wm, cursor_wm); - enabled |= 1 << PIPE_A; - } - - if (g4x_compute_wm0(dev, PIPE_B, - &sandybridge_display_wm_info, latency, - &sandybridge_cursor_wm_info, latency, - &plane_wm, &cursor_wm)) { - val = I915_READ(WM0_PIPEB_ILK); - val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK); - I915_WRITE(WM0_PIPEB_ILK, val | - ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm)); - DRM_DEBUG_KMS("FIFO watermarks For pipe B -" - " plane %d, cursor: %d\n", - plane_wm, cursor_wm); - enabled |= 1 << PIPE_B; - } - - /* - * Calculate and update the self-refresh watermark only when one - * display plane is used. - * - * SNB support 3 levels of watermark. - * - * WM1/WM2/WM2 watermarks have to be enabled in the ascending order, - * and disabled in the descending order - * - */ - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); - - if (!single_plane_enabled(enabled) || - dev_priv->sprite_scaling_enabled) - return; - enabled = ffs(enabled) - 1; - - /* WM1 */ - if (!ironlake_compute_srwm(dev, 1, enabled, - dev_priv->wm.pri_latency[1] * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - return; - - I915_WRITE(WM1_LP_ILK, - WM1_LP_SR_EN | - (dev_priv->wm.pri_latency[1] << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); - - /* WM2 */ - if (!ironlake_compute_srwm(dev, 2, enabled, - dev_priv->wm.pri_latency[2] * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - return; - - I915_WRITE(WM2_LP_ILK, - WM2_LP_EN | - (dev_priv->wm.pri_latency[2] << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); - - /* WM3 */ - if (!ironlake_compute_srwm(dev, 3, enabled, - dev_priv->wm.pri_latency[3] * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - return; - - I915_WRITE(WM3_LP_ILK, - WM3_LP_EN | - (dev_priv->wm.pri_latency[3] << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); -} - -static void ivybridge_update_wm(struct drm_crtc *crtc) -{ - struct drm_device *dev = crtc->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - int latency = dev_priv->wm.pri_latency[0] * 100; /* In unit 0.1us */ - u32 val; - int fbc_wm, plane_wm, cursor_wm; - int ignore_fbc_wm, ignore_plane_wm, ignore_cursor_wm; - unsigned int enabled; - - enabled = 0; - if (g4x_compute_wm0(dev, PIPE_A, - &sandybridge_display_wm_info, latency, - &sandybridge_cursor_wm_info, latency, - &plane_wm, &cursor_wm)) { - val = I915_READ(WM0_PIPEA_ILK); - val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK); - I915_WRITE(WM0_PIPEA_ILK, val | - ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm)); - DRM_DEBUG_KMS("FIFO watermarks For pipe A -" - " plane %d, " "cursor: %d\n", - plane_wm, cursor_wm); - enabled |= 1 << PIPE_A; - } - - if (g4x_compute_wm0(dev, PIPE_B, - &sandybridge_display_wm_info, latency, - &sandybridge_cursor_wm_info, latency, - &plane_wm, &cursor_wm)) { - val = I915_READ(WM0_PIPEB_ILK); - val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK); - I915_WRITE(WM0_PIPEB_ILK, val | - ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm)); - DRM_DEBUG_KMS("FIFO watermarks For pipe B -" - " plane %d, cursor: %d\n", - plane_wm, cursor_wm); - enabled |= 1 << PIPE_B; - } - - if (g4x_compute_wm0(dev, PIPE_C, - &sandybridge_display_wm_info, latency, - &sandybridge_cursor_wm_info, latency, - &plane_wm, &cursor_wm)) { - val = I915_READ(WM0_PIPEC_IVB); - val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK); - I915_WRITE(WM0_PIPEC_IVB, val | - ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm)); - DRM_DEBUG_KMS("FIFO watermarks For pipe C -" - " plane %d, cursor: %d\n", - plane_wm, cursor_wm); - enabled |= 1 << PIPE_C; - } - - /* - * Calculate and update the self-refresh watermark only when one - * display plane is used. - * - * SNB support 3 levels of watermark. - * - * WM1/WM2/WM2 watermarks have to be enabled in the ascending order, - * and disabled in the descending order - * - */ - I915_WRITE(WM3_LP_ILK, 0); - I915_WRITE(WM2_LP_ILK, 0); - I915_WRITE(WM1_LP_ILK, 0); - - if (!single_plane_enabled(enabled) || - dev_priv->sprite_scaling_enabled) - return; - enabled = ffs(enabled) - 1; - - /* WM1 */ - if (!ironlake_compute_srwm(dev, 1, enabled, - dev_priv->wm.pri_latency[1] * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - return; - - I915_WRITE(WM1_LP_ILK, - WM1_LP_SR_EN | - (dev_priv->wm.pri_latency[1] << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); - - /* WM2 */ - if (!ironlake_compute_srwm(dev, 2, enabled, - dev_priv->wm.pri_latency[2] * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &fbc_wm, &plane_wm, &cursor_wm)) - return; - - I915_WRITE(WM2_LP_ILK, - WM2_LP_EN | - (dev_priv->wm.pri_latency[2] << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); - - /* WM3, note we have to correct the cursor latency */ - if (!ironlake_compute_srwm(dev, 3, enabled, - dev_priv->wm.pri_latency[3] * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &fbc_wm, &plane_wm, &ignore_cursor_wm) || - !ironlake_compute_srwm(dev, 3, enabled, - dev_priv->wm.cur_latency[3] * 500, - &sandybridge_display_srwm_info, - &sandybridge_cursor_srwm_info, - &ignore_fbc_wm, &ignore_plane_wm, &cursor_wm)) - return; - - I915_WRITE(WM3_LP_ILK, - WM3_LP_EN | - (dev_priv->wm.pri_latency[3] << WM1_LP_LATENCY_SHIFT) | - (fbc_wm << WM1_LP_FBC_SHIFT) | - (plane_wm << WM1_LP_SR_SHIFT) | - cursor_wm); -} - static uint32_t ilk_pipe_pixel_rate(struct drm_device *dev, struct drm_crtc *crtc) { @@ -3058,168 +2582,6 @@ static void haswell_update_sprite_wm(struct drm_plane *plane, haswell_update_wm(crtc); } -static bool -sandybridge_compute_sprite_wm(struct drm_device *dev, int plane, - uint32_t sprite_width, int pixel_size, - const struct intel_watermark_params *display, - int display_latency_ns, int *sprite_wm) -{ - struct drm_crtc *crtc; - int clock; - int entries, tlb_miss; - - crtc = intel_get_crtc_for_plane(dev, plane); - if (!intel_crtc_active(crtc)) { - *sprite_wm = display->guard_size; - return false; - } - - clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock; - - /* Use the small buffer method to calculate the sprite watermark */ - entries = ((clock * pixel_size / 1000) * display_latency_ns) / 1000; - tlb_miss = display->fifo_size*display->cacheline_size - - sprite_width * 8; - if (tlb_miss > 0) - entries += tlb_miss; - entries = DIV_ROUND_UP(entries, display->cacheline_size); - *sprite_wm = entries + display->guard_size; - if (*sprite_wm > (int)display->max_wm) - *sprite_wm = display->max_wm; - - return true; -} - -static bool -sandybridge_compute_sprite_srwm(struct drm_device *dev, int plane, - uint32_t sprite_width, int pixel_size, - const struct intel_watermark_params *display, - int latency_ns, int *sprite_wm) -{ - struct drm_crtc *crtc; - unsigned long line_time_us; - int clock; - int line_count, line_size; - int small, large; - int entries; - - if (!latency_ns) { - *sprite_wm = 0; - return false; - } - - crtc = intel_get_crtc_for_plane(dev, plane); - clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock; - if (!clock) { - *sprite_wm = 0; - return false; - } - - line_time_us = (sprite_width * 1000) / clock; - if (!line_time_us) { - *sprite_wm = 0; - return false; - } - - line_count = (latency_ns / line_time_us + 1000) / 1000; - line_size = sprite_width * pixel_size; - - /* Use the minimum of the small and large buffer method for primary */ - small = ((clock * pixel_size / 1000) * latency_ns) / 1000; - large = line_count * line_size; - - entries = DIV_ROUND_UP(min(small, large), display->cacheline_size); - *sprite_wm = entries + display->guard_size; - - return *sprite_wm > 0x3ff ? false : true; -} - -static void sandybridge_update_sprite_wm(struct drm_plane *plane, - struct drm_crtc *crtc, - uint32_t sprite_width, int pixel_size, - bool enabled, bool scaled) -{ - struct drm_device *dev = plane->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - int pipe = to_intel_plane(plane)->pipe; - int latency = dev_priv->wm.spr_latency[0] * 100; /* In unit 0.1us */ - u32 val; - int sprite_wm, reg; - int ret; - - if (!enabled) - return; - - switch (pipe) { - case 0: - reg = WM0_PIPEA_ILK; - break; - case 1: - reg = WM0_PIPEB_ILK; - break; - case 2: - reg = WM0_PIPEC_IVB; - break; - default: - return; /* bad pipe */ - } - - ret = sandybridge_compute_sprite_wm(dev, pipe, sprite_width, pixel_size, - &sandybridge_display_wm_info, - latency, &sprite_wm); - if (!ret) { - DRM_DEBUG_KMS("failed to compute sprite wm for pipe %c\n", - pipe_name(pipe)); - return; - } - - val = I915_READ(reg); - val &= ~WM0_PIPE_SPRITE_MASK; - I915_WRITE(reg, val | (sprite_wm << WM0_PIPE_SPRITE_SHIFT)); - DRM_DEBUG_KMS("sprite watermarks For pipe %c - %d\n", pipe_name(pipe), sprite_wm); - - - ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width, - pixel_size, - &sandybridge_display_srwm_info, - dev_priv->wm.spr_latency[1] * 500, - &sprite_wm); - if (!ret) { - DRM_DEBUG_KMS("failed to compute sprite lp1 wm on pipe %c\n", - pipe_name(pipe)); - return; - } - I915_WRITE(WM1S_LP_ILK, sprite_wm); - - /* Only IVB has two more LP watermarks for sprite */ - if (!IS_IVYBRIDGE(dev)) - return; - - ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width, - pixel_size, - &sandybridge_display_srwm_info, - dev_priv->wm.spr_latency[2] * 500, - &sprite_wm); - if (!ret) { - DRM_DEBUG_KMS("failed to compute sprite lp2 wm on pipe %c\n", - pipe_name(pipe)); - return; - } - I915_WRITE(WM2S_LP_IVB, sprite_wm); - - ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width, - pixel_size, - &sandybridge_display_srwm_info, - dev_priv->wm.spr_latency[3] * 500, - &sprite_wm); - if (!ret) { - DRM_DEBUG_KMS("failed to compute sprite lp3 wm on pipe %c\n", - pipe_name(pipe)); - return; - } - I915_WRITE(WM3S_LP_IVB, sprite_wm); -} - static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc) { struct drm_device *dev = crtc->dev; @@ -6184,9 +5546,11 @@ void intel_init_pm(struct drm_device *dev) if (IS_GEN5(dev)) { if (dev_priv->wm.pri_latency[1] && dev_priv->wm.spr_latency[1] && - dev_priv->wm.cur_latency[1]) - dev_priv->display.update_wm = ironlake_update_wm; - else { + dev_priv->wm.cur_latency[1]) { + dev_priv->display.update_wm = haswell_update_wm; + dev_priv->display.update_sprite_wm = + haswell_update_sprite_wm; + } else { DRM_DEBUG_KMS("Failed to get proper latency. " "Disable CxSR\n"); dev_priv->display.update_wm = NULL; @@ -6196,8 +5560,9 @@ void intel_init_pm(struct drm_device *dev) if (dev_priv->wm.pri_latency[0] && dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0]) { - dev_priv->display.update_wm = sandybridge_update_wm; - dev_priv->display.update_sprite_wm = sandybridge_update_sprite_wm; + dev_priv->display.update_wm = haswell_update_wm; + dev_priv->display.update_sprite_wm = + haswell_update_sprite_wm; } else { DRM_DEBUG_KMS("Failed to read display plane latency. " "Disable CxSR\n"); @@ -6208,8 +5573,9 @@ void intel_init_pm(struct drm_device *dev) if (dev_priv->wm.pri_latency[0] && dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0]) { - dev_priv->display.update_wm = ivybridge_update_wm; - dev_priv->display.update_sprite_wm = sandybridge_update_sprite_wm; + dev_priv->display.update_wm = haswell_update_wm; + dev_priv->display.update_sprite_wm = + haswell_update_sprite_wm; } else { DRM_DEBUG_KMS("Failed to read display plane latency. " "Disable CxSR\n"); -- cgit v1.2.3 From 8553c18ea696827f26d3ba5652d3a2fa9ce8e874 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 5 Dec 2013 15:51:39 +0200 Subject: drm/i915: Try to fix the messy IVB sprite scaling workaround MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We now have a very clear method of disabling LP1+ wartermarks, and we can actually detect if we actually did disable them, or if they were already disabled. Use that to clean up the WaCxSRDisabledForSpriteScaling:ivb handling. I was hoping to apply the workaround in a way that wouldn't require a blocking wait, but sadly IVB really does appear to require LP1+ watermarks to be off for an entire frame before enabling sprite scaling. Simply disabling LP1+ watermarks during the previous frame is not enough, no matter how early in the frame we do it :( Signed-off-by: Ville Syrjälä Reviewed-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 58 ++++++++++++++++++++++++++++++++--------- 1 file changed, 45 insertions(+), 13 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 7e028341b2f5..ecdb64e44e2a 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2426,33 +2426,26 @@ static unsigned int ilk_compute_wm_dirty(struct drm_device *dev, return dirty; } -/* - * The spec says we shouldn't write when we don't need, because every write - * causes WMs to be re-evaluated, expending some power. - */ -static void hsw_write_wm_values(struct drm_i915_private *dev_priv, - struct hsw_wm_values *results) +static bool _ilk_disable_lp_wm(struct drm_i915_private *dev_priv, + unsigned int dirty) { - struct drm_device *dev = dev_priv->dev; struct hsw_wm_values *previous = &dev_priv->wm.hw; - unsigned int dirty; - uint32_t val; - - dirty = ilk_compute_wm_dirty(dev_priv->dev, previous, results); - if (!dirty) - return; + bool changed = false; if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] & WM1_LP_SR_EN) { previous->wm_lp[2] &= ~WM1_LP_SR_EN; I915_WRITE(WM3_LP_ILK, previous->wm_lp[2]); + changed = true; } if (dirty & WM_DIRTY_LP(2) && previous->wm_lp[1] & WM1_LP_SR_EN) { previous->wm_lp[1] &= ~WM1_LP_SR_EN; I915_WRITE(WM2_LP_ILK, previous->wm_lp[1]); + changed = true; } if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] & WM1_LP_SR_EN) { previous->wm_lp[0] &= ~WM1_LP_SR_EN; I915_WRITE(WM1_LP_ILK, previous->wm_lp[0]); + changed = true; } /* @@ -2460,6 +2453,27 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv, * Doing so could cause underruns. */ + return changed; +} + +/* + * The spec says we shouldn't write when we don't need, because every write + * causes WMs to be re-evaluated, expending some power. + */ +static void hsw_write_wm_values(struct drm_i915_private *dev_priv, + struct hsw_wm_values *results) +{ + struct drm_device *dev = dev_priv->dev; + struct hsw_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)) @@ -2523,6 +2537,13 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv, dev_priv->wm.hw = *results; } +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 haswell_update_wm(struct drm_crtc *crtc) { struct intel_crtc *intel_crtc = to_intel_crtc(crtc); @@ -2572,6 +2593,7 @@ static void haswell_update_sprite_wm(struct drm_plane *plane, 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; @@ -2579,6 +2601,16 @@ static void haswell_update_sprite_wm(struct drm_plane *plane, intel_plane->wm.horiz_pixels = sprite_width; intel_plane->wm.bytes_per_pixel = pixel_size; + /* + * 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); + haswell_update_wm(crtc); } -- cgit v1.2.3 From 954911ebb1cb5c965a127ec9c0fc86c069c71cf9 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Tue, 17 Dec 2013 14:46:34 +0200 Subject: drm/i915: simplify platform specific code in hsw_write_wm_values No functional change. Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index ecdb64e44e2a..9c33835e96f9 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2515,12 +2515,11 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv, I915_WRITE(DISP_ARB_CTL, val); } - if (INTEL_INFO(dev)->gen <= 6) { - 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]); - } else { - if (dirty & WM_DIRTY_LP(1) && previous->wm_lp_spr[0] != results->wm_lp_spr[0]) - I915_WRITE(WM1S_LP_ILK, results->wm_lp_spr[0]); + if (dirty & WM_DIRTY_LP(1) && + previous->wm_lp_spr[0] != results->wm_lp_spr[0]) + I915_WRITE(WM1S_LP_ILK, results->wm_lp_spr[0]); + + if (INTEL_INFO(dev)->gen >= 7) { if (dirty & WM_DIRTY_LP(2) && previous->wm_lp_spr[1] != results->wm_lp_spr[1]) I915_WRITE(WM2S_LP_IVB, results->wm_lp_spr[1]); if (dirty & WM_DIRTY_LP(3) && previous->wm_lp_spr[2] != results->wm_lp_spr[2]) -- cgit v1.2.3 From 820c1980353808717f78a0c4ca2c734f0247dc0c Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Tue, 17 Dec 2013 14:46:36 +0200 Subject: drm/i915: s/haswell_update_wm/ilk_update_wm/ We use this hook starting from ILK onwards, so change the prefix accordingly. Also rename functions/struct names used from haswell_update_wm that are relevant to ILK already. No functional change. Signed-off-by: Imre Deak Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 86 ++++++++++++++++++++--------------------- 1 file changed, 43 insertions(+), 43 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 9c33835e96f9..bbd6503a4cf5 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -1716,7 +1716,7 @@ static uint32_t ilk_wm_fbc(uint32_t pri_val, uint32_t horiz_pixels, return DIV_ROUND_UP(pri_val * 64, horiz_pixels * bytes_per_pixel) + 2; } -struct hsw_pipe_wm_parameters { +struct ilk_pipe_wm_parameters { bool active; uint32_t pipe_htotal; uint32_t pixel_rate; @@ -1725,7 +1725,7 @@ struct hsw_pipe_wm_parameters { struct intel_plane_wm_parameters cur; }; -struct hsw_wm_maximums { +struct ilk_wm_maximums { uint16_t pri; uint16_t spr; uint16_t cur; @@ -1743,7 +1743,7 @@ struct intel_wm_config { * For both WM_PIPE and WM_LP. * mem_value must be in 0.1us units. */ -static uint32_t ilk_compute_pri_wm(const struct hsw_pipe_wm_parameters *params, +static uint32_t ilk_compute_pri_wm(const struct ilk_pipe_wm_parameters *params, uint32_t mem_value, bool is_lp) { @@ -1772,7 +1772,7 @@ static uint32_t ilk_compute_pri_wm(const struct hsw_pipe_wm_parameters *params, * For both WM_PIPE and WM_LP. * mem_value must be in 0.1us units. */ -static uint32_t ilk_compute_spr_wm(const struct hsw_pipe_wm_parameters *params, +static uint32_t ilk_compute_spr_wm(const struct ilk_pipe_wm_parameters *params, uint32_t mem_value) { uint32_t method1, method2; @@ -1795,7 +1795,7 @@ static uint32_t ilk_compute_spr_wm(const struct hsw_pipe_wm_parameters *params, * For both WM_PIPE and WM_LP. * mem_value must be in 0.1us units. */ -static uint32_t ilk_compute_cur_wm(const struct hsw_pipe_wm_parameters *params, +static uint32_t ilk_compute_cur_wm(const struct ilk_pipe_wm_parameters *params, uint32_t mem_value) { if (!params->active || !params->cur.enabled) @@ -1809,7 +1809,7 @@ static uint32_t ilk_compute_cur_wm(const struct hsw_pipe_wm_parameters *params, } /* Only for WM_LP. */ -static uint32_t ilk_compute_fbc_wm(const struct hsw_pipe_wm_parameters *params, +static uint32_t ilk_compute_fbc_wm(const struct ilk_pipe_wm_parameters *params, uint32_t pri_val) { if (!params->active || !params->pri.enabled) @@ -1914,7 +1914,7 @@ static void ilk_compute_wm_maximums(struct drm_device *dev, int level, const struct intel_wm_config *config, enum intel_ddb_partitioning ddb_partitioning, - struct hsw_wm_maximums *max) + struct ilk_wm_maximums *max) { max->pri = ilk_plane_wm_max(dev, level, config, ddb_partitioning, false); max->spr = ilk_plane_wm_max(dev, level, config, ddb_partitioning, true); @@ -1923,7 +1923,7 @@ static void ilk_compute_wm_maximums(struct drm_device *dev, } static bool ilk_validate_wm_level(int level, - const struct hsw_wm_maximums *max, + const struct ilk_wm_maximums *max, struct intel_wm_level *result) { bool ret; @@ -1965,7 +1965,7 @@ static bool ilk_validate_wm_level(int level, static void ilk_compute_wm_level(struct drm_i915_private *dev_priv, int level, - const struct hsw_pipe_wm_parameters *p, + const struct ilk_pipe_wm_parameters *p, struct intel_wm_level *result) { uint16_t pri_latency = dev_priv->wm.pri_latency[level]; @@ -2112,8 +2112,8 @@ static void intel_setup_wm_latency(struct drm_device *dev) intel_print_wm_latency(dev, "Cursor", dev_priv->wm.cur_latency); } -static void hsw_compute_wm_parameters(struct drm_crtc *crtc, - struct hsw_pipe_wm_parameters *p, +static void ilk_compute_wm_parameters(struct drm_crtc *crtc, + struct ilk_pipe_wm_parameters *p, struct intel_wm_config *config) { struct drm_device *dev = crtc->dev; @@ -2150,7 +2150,7 @@ static void hsw_compute_wm_parameters(struct drm_crtc *crtc, /* Compute new watermarks for the pipe */ static bool intel_compute_pipe_wm(struct drm_crtc *crtc, - const struct hsw_pipe_wm_parameters *params, + const struct ilk_pipe_wm_parameters *params, struct intel_pipe_wm *pipe_wm) { struct drm_device *dev = crtc->dev; @@ -2162,7 +2162,7 @@ static bool intel_compute_pipe_wm(struct drm_crtc *crtc, .sprites_enabled = params->spr.enabled, .sprites_scaled = params->spr.scaled, }; - struct hsw_wm_maximums max; + struct ilk_wm_maximums max; /* LP0 watermarks always use 1/2 DDB partitioning */ ilk_compute_wm_maximums(dev, 0, &config, INTEL_DDB_PART_1_2, &max); @@ -2216,7 +2216,7 @@ static void ilk_merge_wm_level(struct drm_device *dev, */ static void ilk_wm_merge(struct drm_device *dev, const struct intel_wm_config *config, - const struct hsw_wm_maximums *max, + const struct ilk_wm_maximums *max, struct intel_pipe_wm *merged) { int level, max_level = ilk_wm_max_level(dev); @@ -2280,10 +2280,10 @@ static unsigned int ilk_wm_lp_latency(struct drm_device *dev, int level) return dev_priv->wm.pri_latency[level]; } -static void hsw_compute_wm_results(struct drm_device *dev, +static void ilk_compute_wm_results(struct drm_device *dev, const struct intel_pipe_wm *merged, enum intel_ddb_partitioning partitioning, - struct hsw_wm_values *results) + struct ilk_wm_values *results) { struct intel_crtc *intel_crtc; int level, wm_lp; @@ -2340,7 +2340,7 @@ static void hsw_compute_wm_results(struct drm_device *dev, /* Find the result with the highest level enabled. Check for enable_fbc_wm in * case both are at the same level. Prefer r1 in case they're the same. */ -static struct intel_pipe_wm *hsw_find_best_result(struct drm_device *dev, +static struct intel_pipe_wm *ilk_find_best_result(struct drm_device *dev, struct intel_pipe_wm *r1, struct intel_pipe_wm *r2) { @@ -2375,8 +2375,8 @@ static struct intel_pipe_wm *hsw_find_best_result(struct drm_device *dev, #define WM_DIRTY_DDB (1 << 25) static unsigned int ilk_compute_wm_dirty(struct drm_device *dev, - const struct hsw_wm_values *old, - const struct hsw_wm_values *new) + const struct ilk_wm_values *old, + const struct ilk_wm_values *new) { unsigned int dirty = 0; enum pipe pipe; @@ -2429,7 +2429,7 @@ static unsigned int ilk_compute_wm_dirty(struct drm_device *dev, static bool _ilk_disable_lp_wm(struct drm_i915_private *dev_priv, unsigned int dirty) { - struct hsw_wm_values *previous = &dev_priv->wm.hw; + struct ilk_wm_values *previous = &dev_priv->wm.hw; bool changed = false; if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] & WM1_LP_SR_EN) { @@ -2460,11 +2460,11 @@ static bool _ilk_disable_lp_wm(struct drm_i915_private *dev_priv, * The spec says we shouldn't write when we don't need, because every write * causes WMs to be re-evaluated, expending some power. */ -static void hsw_write_wm_values(struct drm_i915_private *dev_priv, - struct hsw_wm_values *results) +static void ilk_write_wm_values(struct drm_i915_private *dev_priv, + struct ilk_wm_values *results) { struct drm_device *dev = dev_priv->dev; - struct hsw_wm_values *previous = &dev_priv->wm.hw; + struct ilk_wm_values *previous = &dev_priv->wm.hw; unsigned int dirty; uint32_t val; @@ -2543,20 +2543,20 @@ static bool ilk_disable_lp_wm(struct drm_device *dev) return _ilk_disable_lp_wm(dev_priv, WM_DIRTY_LP_ALL); } -static void haswell_update_wm(struct drm_crtc *crtc) +static void ilk_update_wm(struct drm_crtc *crtc) { struct intel_crtc *intel_crtc = to_intel_crtc(crtc); struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct hsw_wm_maximums max; - struct hsw_pipe_wm_parameters params = {}; - struct hsw_wm_values results = {}; + struct ilk_wm_maximums max; + struct ilk_pipe_wm_parameters params = {}; + struct ilk_wm_values results = {}; enum intel_ddb_partitioning partitioning; struct intel_pipe_wm pipe_wm = {}; struct intel_pipe_wm lp_wm_1_2 = {}, lp_wm_5_6 = {}, *best_lp_wm; struct intel_wm_config config = {}; - hsw_compute_wm_parameters(crtc, ¶ms, &config); + ilk_compute_wm_parameters(crtc, ¶ms, &config); intel_compute_pipe_wm(crtc, ¶ms, &pipe_wm); @@ -2574,7 +2574,7 @@ static void haswell_update_wm(struct drm_crtc *crtc) 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 = hsw_find_best_result(dev, &lp_wm_1_2, &lp_wm_5_6); + best_lp_wm = ilk_find_best_result(dev, &lp_wm_1_2, &lp_wm_5_6); } else { best_lp_wm = &lp_wm_1_2; } @@ -2582,12 +2582,12 @@ static void haswell_update_wm(struct drm_crtc *crtc) partitioning = (best_lp_wm == &lp_wm_1_2) ? INTEL_DDB_PART_1_2 : INTEL_DDB_PART_5_6; - hsw_compute_wm_results(dev, best_lp_wm, partitioning, &results); + ilk_compute_wm_results(dev, best_lp_wm, partitioning, &results); - hsw_write_wm_values(dev_priv, &results); + ilk_write_wm_values(dev_priv, &results); } -static void haswell_update_sprite_wm(struct drm_plane *plane, +static void ilk_update_sprite_wm(struct drm_plane *plane, struct drm_crtc *crtc, uint32_t sprite_width, int pixel_size, bool enabled, bool scaled) @@ -2610,14 +2610,14 @@ static void haswell_update_sprite_wm(struct drm_plane *plane, if (IS_IVYBRIDGE(dev) && scaled && ilk_disable_lp_wm(dev)) intel_wait_for_vblank(dev, intel_plane->pipe); - haswell_update_wm(crtc); + ilk_update_wm(crtc); } static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc) { struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct hsw_wm_values *hw = &dev_priv->wm.hw; + struct ilk_wm_values *hw = &dev_priv->wm.hw; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); struct intel_pipe_wm *active = &intel_crtc->wm.active; enum pipe pipe = intel_crtc->pipe; @@ -2661,7 +2661,7 @@ static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc) void ilk_wm_get_hw_state(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - struct hsw_wm_values *hw = &dev_priv->wm.hw; + struct ilk_wm_values *hw = &dev_priv->wm.hw; struct drm_crtc *crtc; list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) @@ -5578,9 +5578,9 @@ void intel_init_pm(struct drm_device *dev) if (dev_priv->wm.pri_latency[1] && dev_priv->wm.spr_latency[1] && dev_priv->wm.cur_latency[1]) { - dev_priv->display.update_wm = haswell_update_wm; + dev_priv->display.update_wm = ilk_update_wm; dev_priv->display.update_sprite_wm = - haswell_update_sprite_wm; + ilk_update_sprite_wm; } else { DRM_DEBUG_KMS("Failed to get proper latency. " "Disable CxSR\n"); @@ -5591,9 +5591,9 @@ void intel_init_pm(struct drm_device *dev) if (dev_priv->wm.pri_latency[0] && dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0]) { - dev_priv->display.update_wm = haswell_update_wm; + dev_priv->display.update_wm = ilk_update_wm; dev_priv->display.update_sprite_wm = - haswell_update_sprite_wm; + ilk_update_sprite_wm; } else { DRM_DEBUG_KMS("Failed to read display plane latency. " "Disable CxSR\n"); @@ -5604,9 +5604,9 @@ void intel_init_pm(struct drm_device *dev) if (dev_priv->wm.pri_latency[0] && dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0]) { - dev_priv->display.update_wm = haswell_update_wm; + dev_priv->display.update_wm = ilk_update_wm; dev_priv->display.update_sprite_wm = - haswell_update_sprite_wm; + ilk_update_sprite_wm; } else { DRM_DEBUG_KMS("Failed to read display plane latency. " "Disable CxSR\n"); @@ -5617,9 +5617,9 @@ void intel_init_pm(struct drm_device *dev) if (dev_priv->wm.pri_latency[0] && dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0]) { - dev_priv->display.update_wm = haswell_update_wm; + dev_priv->display.update_wm = ilk_update_wm; dev_priv->display.update_sprite_wm = - haswell_update_sprite_wm; + ilk_update_sprite_wm; } else { DRM_DEBUG_KMS("Failed to read display plane latency. " "Disable CxSR\n"); -- cgit v1.2.3 From a42a57196ad897e7686944488b23f5d2b6496372 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 7 Jan 2014 16:14:08 +0200 Subject: drm/i915: Fix watermark code for BDW MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Looks like I forgot to update the ILK/SNB/IVB watermark patches to deal with BDW. Add the relevant BDW checks to make sure we take the HSW codepaths on BDW as well. Signed-off-by: Ville Syrjälä Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index bbd6503a4cf5..78503157daf2 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2012,7 +2012,7 @@ static void intel_read_wm_latency(struct drm_device *dev, uint16_t wm[5]) { struct drm_i915_private *dev_priv = dev->dev_private; - if (IS_HASWELL(dev)) { + if (IS_HASWELL(dev) || IS_BROADWELL(dev)) { uint64_t sskpd = I915_READ64(MCH_SSKPD); wm[0] = (sskpd >> 56) & 0xFF; @@ -2060,7 +2060,7 @@ static void intel_fixup_cur_wm_latency(struct drm_device *dev, uint16_t wm[5]) static int ilk_wm_max_level(const struct drm_device *dev) { /* how many WM levels are we expecting */ - if (IS_HASWELL(dev)) + if (IS_HASWELL(dev) || IS_BROADWELL(dev)) return 4; else if (INTEL_INFO(dev)->gen >= 6) return 3; @@ -2179,7 +2179,7 @@ static bool intel_compute_pipe_wm(struct drm_crtc *crtc, ilk_compute_wm_level(dev_priv, level, params, &pipe_wm->wm[level]); - if (IS_HASWELL(dev)) + if (IS_HASWELL(dev) || IS_BROADWELL(dev)) pipe_wm->linetime = hsw_compute_linetime_wm(dev, crtc); /* At least LP0 must be valid */ @@ -2274,7 +2274,7 @@ 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)) + if (IS_HASWELL(dev) || IS_BROADWELL(dev)) return 2 * level; else return dev_priv->wm.pri_latency[level]; @@ -2489,7 +2489,7 @@ static void ilk_write_wm_values(struct drm_i915_private *dev_priv, I915_WRITE(PIPE_WM_LINETIME(PIPE_C), results->wm_linetime[2]); if (dirty & WM_DIRTY_DDB) { - if (IS_HASWELL(dev)) { + 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; @@ -2628,7 +2628,7 @@ static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc) }; hw->wm_pipe[pipe] = I915_READ(wm0_pipe_reg[pipe]); - if (IS_HASWELL(dev)) + if (IS_HASWELL(dev) || IS_BROADWELL(dev)) hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe)); if (intel_crtc_active(crtc)) { @@ -2675,7 +2675,7 @@ void ilk_wm_get_hw_state(struct drm_device *dev) hw->wm_lp_spr[1] = I915_READ(WM2S_LP_IVB); hw->wm_lp_spr[2] = I915_READ(WM3S_LP_IVB); - if (IS_HASWELL(dev)) + 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)) -- cgit v1.2.3 From 03dce881292f43b269f1f8acca69df2c4568ee5c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 7 Jan 2014 16:14:09 +0200 Subject: drm/i915: Enable watermarks for BDW MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We forgot to intialize the watermark vfuncs for BDW, and hence the watermarks were never updated. Signed-off-by: Ville Syrjälä Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 78503157daf2..835f86386b95 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5627,6 +5627,17 @@ void intel_init_pm(struct drm_device *dev) } dev_priv->display.init_clock_gating = haswell_init_clock_gating; } else if (INTEL_INFO(dev)->gen == 8) { + if (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"); + dev_priv->display.update_wm = NULL; + } dev_priv->display.init_clock_gating = gen8_init_clock_gating; } else dev_priv->display.update_wm = NULL; -- cgit v1.2.3 From bd602544712edf72a128bc81d02f0a8051cb1372 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Tue, 7 Jan 2014 16:14:10 +0200 Subject: drm/i915: Simplify watermark/init_clock_gating setup MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Avoid duplicating the same piece of code several times by separating the watemark vfunc setup from the init_clock_gating vfunc setup on PCH platforms. Signed-off-by: Ville Syrjälä Reviewed-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 78 +++++++++-------------------------------- 1 file changed, 16 insertions(+), 62 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 835f86386b95..743449a51be8 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -5574,73 +5574,27 @@ void intel_init_pm(struct drm_device *dev) if (HAS_PCH_SPLIT(dev)) { intel_setup_wm_latency(dev); - if (IS_GEN5(dev)) { - if (dev_priv->wm.pri_latency[1] && - dev_priv->wm.spr_latency[1] && - dev_priv->wm.cur_latency[1]) { - dev_priv->display.update_wm = ilk_update_wm; - dev_priv->display.update_sprite_wm = - ilk_update_sprite_wm; - } else { - DRM_DEBUG_KMS("Failed to get proper latency. " - "Disable CxSR\n"); - dev_priv->display.update_wm = NULL; - } + if ((IS_GEN5(dev) && dev_priv->wm.pri_latency[1] && + dev_priv->wm.spr_latency[1] && dev_priv->wm.cur_latency[1]) || + (!IS_GEN5(dev) && dev_priv->wm.pri_latency[0] && + dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0])) { + dev_priv->display.update_wm = ilk_update_wm; + dev_priv->display.update_sprite_wm = ilk_update_sprite_wm; + } else { + DRM_DEBUG_KMS("Failed to read display plane latency. " + "Disable CxSR\n"); + } + + if (IS_GEN5(dev)) dev_priv->display.init_clock_gating = ironlake_init_clock_gating; - } else if (IS_GEN6(dev)) { - if (dev_priv->wm.pri_latency[0] && - dev_priv->wm.spr_latency[0] && - dev_priv->wm.cur_latency[0]) { - dev_priv->display.update_wm = 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"); - dev_priv->display.update_wm = NULL; - } + else if (IS_GEN6(dev)) dev_priv->display.init_clock_gating = gen6_init_clock_gating; - } else if (IS_IVYBRIDGE(dev)) { - if (dev_priv->wm.pri_latency[0] && - dev_priv->wm.spr_latency[0] && - dev_priv->wm.cur_latency[0]) { - dev_priv->display.update_wm = 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"); - dev_priv->display.update_wm = NULL; - } + else if (IS_IVYBRIDGE(dev)) dev_priv->display.init_clock_gating = ivybridge_init_clock_gating; - } else if (IS_HASWELL(dev)) { - if (dev_priv->wm.pri_latency[0] && - dev_priv->wm.spr_latency[0] && - dev_priv->wm.cur_latency[0]) { - dev_priv->display.update_wm = 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"); - dev_priv->display.update_wm = NULL; - } + else if (IS_HASWELL(dev)) dev_priv->display.init_clock_gating = haswell_init_clock_gating; - } else if (INTEL_INFO(dev)->gen == 8) { - if (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"); - dev_priv->display.update_wm = NULL; - } + else if (INTEL_INFO(dev)->gen == 8) dev_priv->display.init_clock_gating = gen8_init_clock_gating; - } else - dev_priv->display.update_wm = NULL; } else if (IS_VALLEYVIEW(dev)) { dev_priv->display.update_wm = valleyview_update_wm; dev_priv->display.init_clock_gating = -- cgit v1.2.3 From 576b259e65baf15f691010353f6ab06f7190c33f Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Fri, 20 Dec 2013 13:08:00 -0800 Subject: drm/i915: use crtc_htotal when calculating ilk watermarks MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This was introduced in: commit 7c4a395ff8f441acb7876281c6777624e6410349 Author: Ville Syrjälä Date: Wed Oct 9 19:17:56 2013 +0300 drm/i915: Don't re-compute pipe watermarks except for the affected pipe and I missed fixing it in: commit fec8cba306f974f3a4491176994de5d821273643 Author: Jesse Barnes Date: Wed Nov 27 11:10:26 2013 -0800 drm/i915: use crtc_htotal in watermark calculations to match fastboot v2 It's needed for ILK+ platforms to fastboot without crashing on a divide by 0 after a DPMS on action. Note: Ville mentioned in his review that this confusion seems to go down to the original introduction of this code in commit 801bcfffbb0721d7131e930f9a46103e539c43a4 Author: Paulo Zanoni Date: Fri May 31 10:08:35 2013 -0300 drm/i915: properly set HSW WM_PIPE registers So it seems to have been missed both in the fastboot patch and in the 3d mode suppport (where only crtc_htotal reflects the real pipe width). Signed-off-by: Jesse Barnes Reviewed-by: Ville Syrjälä [danvet: Add note based on Ville's review.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 743449a51be8..534459944d70 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2123,7 +2123,7 @@ static void ilk_compute_wm_parameters(struct drm_crtc *crtc, p->active = intel_crtc_active(crtc); if (p->active) { - p->pipe_htotal = intel_crtc->config.adjusted_mode.htotal; + 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->fb->bits_per_pixel / 8; p->cur.bytes_per_pixel = 4; -- cgit v1.2.3 From 3a77c4c441d2bcc2b1e27d8aabee0c57aed66ed3 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Fri, 10 Jan 2014 08:50:12 +0100 Subject: drm/i915: Drop I915_ prefix from HAS_FBC MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit My OCD just couldn't let this slide. Spotted while reviewing Ville's patch to only flip planes when we have FBC. Cc: Ville Syrjälä Reviewed-by: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 534459944d70..26e6d1b5363f 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -461,7 +461,7 @@ void intel_update_fbc(struct drm_device *dev) const struct drm_display_mode *adjusted_mode; unsigned int max_width, max_height; - if (!I915_HAS_FBC(dev)) { + if (!HAS_FBC(dev)) { set_no_fbc_reason(dev_priv, FBC_UNSUPPORTED); return; } @@ -5541,7 +5541,7 @@ void intel_init_pm(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - if (I915_HAS_FBC(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; -- cgit v1.2.3 From feb56b934463a7339ebc3c3cf2497c7958fe5a60 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Sat, 14 Dec 2013 20:38:30 -0200 Subject: drm/i915: i830M has watermarks like i855 So shuffle the checks around a bit. Also give all the structs and functions proper prefixes: i830_ for the dual-pipe mobile platforms and i845_ for the two single-pipe desktop platforms. Note that the max fifo value isn't actually correct for the i830M, but since we don't frob the fifo split we don't actually need it. This is different for some gen3 devices where we need the full fifo for self refresh mode. Cc: Thomas Richter Signed-off-by: Rodrigo Vivi Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 53 +++++++++++++++-------------------------- 1 file changed, 19 insertions(+), 34 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 26e6d1b5363f..fca20e4992e1 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -824,7 +824,7 @@ 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); @@ -857,21 +857,6 @@ 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, @@ -950,14 +935,14 @@ 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, @@ -1515,7 +1500,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc) else if (!IS_GEN2(dev)) wm_info = &i915_wm_info; else - wm_info = &i855_wm_info; + wm_info = &i830_wm_info; fifo_size = dev_priv->display.get_fifo_size(dev, 0); crtc = intel_get_crtc_for_plane(dev, 0); @@ -1622,7 +1607,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc) } } -static void i830_update_wm(struct drm_crtc *unused_crtc) +static void i845_update_wm(struct drm_crtc *unused_crtc) { struct drm_device *dev = unused_crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; @@ -1637,7 +1622,7 @@ static void i830_update_wm(struct drm_crtc *unused_crtc) adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode; planea_wm = intel_calculate_wm(adjusted_mode->crtc_clock, - &i830_wm_info, + &i845_wm_info, dev_priv->display.get_fifo_size(dev, 0), 4, latency_ns); fwater_lo = I915_READ(FW_BLC) & ~0xfff; @@ -5628,21 +5613,21 @@ void intel_init_pm(struct drm_device *dev) dev_priv->display.update_wm = i9xx_update_wm; dev_priv->display.get_fifo_size = i9xx_get_fifo_size; dev_priv->display.init_clock_gating = gen3_init_clock_gating; - } else if (IS_I865G(dev)) { - dev_priv->display.update_wm = i830_update_wm; - dev_priv->display.init_clock_gating = i85x_init_clock_gating; - dev_priv->display.get_fifo_size = i830_get_fifo_size; - } else if (IS_I85X(dev)) { - dev_priv->display.update_wm = i9xx_update_wm; - dev_priv->display.get_fifo_size = i85x_get_fifo_size; - dev_priv->display.init_clock_gating = i85x_init_clock_gating; - } else { - dev_priv->display.update_wm = i830_update_wm; - dev_priv->display.init_clock_gating = i830_init_clock_gating; - if (IS_845G(dev)) + } else if (IS_GEN2(dev)) { + if (INTEL_INFO(dev)->num_pipes == 1) { + dev_priv->display.update_wm = i845_update_wm; dev_priv->display.get_fifo_size = i845_get_fifo_size; - else + } else { + dev_priv->display.update_wm = i9xx_update_wm; dev_priv->display.get_fifo_size = i830_get_fifo_size; + } + + if (IS_I85X(dev) || IS_I865G(dev)) + dev_priv->display.init_clock_gating = i85x_init_clock_gating; + else + dev_priv->display.init_clock_gating = i830_init_clock_gating; + } else { + DRM_ERROR("unexpected fall-through in intel_init_pm\n"); } } -- cgit v1.2.3 From 3f2dc5ac05714711fc14f2bf0ee5e42d5c08c581 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Fri, 10 Jan 2014 14:06:47 +0200 Subject: drm/i915: Fix 915GM self-refresh enable/disable MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit INSTPM is a masked register so use the _MASKED_BIT_{ENABLE,DISABLE} macros when enabling/disabling self-refresh on 915GM. Signed-off-by: Ville Syrjälä Reviewed-by: Chris Wilson Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/i915/intel_pm.c') diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index fca20e4992e1..9998185fdb22 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -1548,7 +1548,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc) if (IS_I945G(dev) || IS_I945GM(dev)) I915_WRITE(FW_BLC_SELF, FW_BLC_SELF_EN_MASK | 0); else if (IS_I915GM(dev)) - I915_WRITE(INSTPM, I915_READ(INSTPM) & ~INSTPM_SELF_EN); + I915_WRITE(INSTPM, _MASKED_BIT_DISABLE(INSTPM_SELF_EN)); /* Calc sr entries for one plane configs */ if (HAS_FW_BLC(dev) && enabled) { @@ -1600,7 +1600,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc) I915_WRITE(FW_BLC_SELF, FW_BLC_SELF_EN_MASK | FW_BLC_SELF_EN); else if (IS_I915GM(dev)) - I915_WRITE(INSTPM, I915_READ(INSTPM) | INSTPM_SELF_EN); + I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_SELF_EN)); DRM_DEBUG_KMS("memory self refresh enabled\n"); } else DRM_DEBUG_KMS("memory self refresh disabled\n"); -- cgit v1.2.3