From: Ville Syrjälä ville.syrjala@linux.intel.com
We have a few more corner cases with the drm_vblank_on/off stuff. This series aims to fix those.
I couldn't think of a good way to write a test case for the drm_vblank_get() issue since it's a transient state that occurs briefly during modeset. So I added asserts for it instead.
So far I didn't write a test for the intel_sanitize_crtc() case either. But I can try to think of something for that...
Ville Syrjälä (3): drm: Always reject drm_vblank_get() after drm_vblank_off() drm/i915: Warn if drm_vblank_get() still works after drm_vblank_off() drm/i915: Re-enable vblank irqs for already active pipes
drivers/gpu/drm/drm_irq.c | 14 ++++++++++++++ drivers/gpu/drm/i915/intel_display.c | 21 +++++++++++++++++++++ 2 files changed, 35 insertions(+)
From: Ville Syrjälä ville.syrjala@linux.intel.com
Make sure drm_vblank_get() never succeeds when called between drm_vblank_off() and drm_vblank_on(). Borrow a trick from the old drm_vblank_{pre,post}_modeset() functions and just bump the refcount in drm_vblank_off() and drop it in drm_vblank_on().
Hopefully the use of inmodeset won't conflict badly with drm_vblank_{pre,post}_modeset().
For i915 there's a window between drm_vblank_off() and marking the crtc as inactive where the current code still allows drm_vblank_get().
Signed-off-by: Ville Syrjälä ville.syrjala@linux.intel.com --- drivers/gpu/drm/drm_irq.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+)
diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c index dd786d8..69a2b2a 100644 --- a/drivers/gpu/drm/drm_irq.c +++ b/drivers/gpu/drm/drm_irq.c @@ -958,6 +958,15 @@ void drm_vblank_off(struct drm_device *dev, int crtc) } spin_unlock(&dev->event_lock);
+ /* + * Prevent subsequent drm_vblank_get() from re-enabling + * the vblank interrupt by bumping the refcount. + */ + if (!dev->vblank[crtc].inmodeset) { + atomic_inc(&dev->vblank[crtc].refcount); + dev->vblank[crtc].inmodeset = 1; + } + spin_unlock_irqrestore(&dev->vbl_lock, irqflags); } EXPORT_SYMBOL(drm_vblank_off); @@ -972,6 +981,11 @@ void drm_vblank_on(struct drm_device *dev, int crtc) unsigned long irqflags;
spin_lock_irqsave(&dev->vbl_lock, irqflags); + /* Drop our private "prevent drm_vblank_get" refcount */ + if (dev->vblank[crtc].inmodeset) { + atomic_dec(&dev->vblank[crtc].refcount); + dev->vblank[crtc].inmodeset = 0; + } /* re-enable interrupts if there's are users left */ if (atomic_read(&dev->vblank[crtc].refcount) != 0) WARN_ON(drm_vblank_enable(dev, crtc));
From: Ville Syrjälä ville.syrjala@linux.intel.com
Signed-off-by: Ville Syrjälä ville.syrjala@linux.intel.com --- drivers/gpu/drm/i915/intel_display.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index df58afc..9420f4f 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -1298,6 +1298,17 @@ static void assert_sprites_disabled(struct drm_i915_private *dev_priv, } }
+static void assert_vblank_disabled(struct intel_crtc *crtc) +{ + struct drm_device *dev = crtc->base.dev; + enum pipe pipe = crtc->pipe; + + if (WARN_ON(drm_vblank_get(dev, pipe) == 0)) { + drm_vblank_put(dev, pipe); + drm_vblank_off(dev, pipe); + } +} + static void ibx_assert_pch_refclk_enabled(struct drm_i915_private *dev_priv) { u32 val; @@ -3863,6 +3874,8 @@ static void intel_crtc_enable_planes(struct drm_crtc *crtc) int pipe = intel_crtc->pipe; int plane = intel_crtc->plane;
+ assert_vblank_disabled(intel_crtc); + intel_enable_primary_hw_plane(dev_priv, plane, pipe); intel_enable_planes(crtc); /* The fixup needs to happen before cursor is enabled */ @@ -3899,6 +3912,8 @@ static void intel_crtc_disable_planes(struct drm_crtc *crtc) intel_crtc_update_cursor(crtc, false); intel_disable_planes(crtc); intel_disable_primary_hw_plane(dev_priv, plane, pipe); + + assert_vblank_disabled(intel_crtc); }
static void ironlake_crtc_enable(struct drm_crtc *crtc)
From: Ville Syrjälä ville.syrjala@linux.intel.com
If a pipe is already active when we init/resume there might not be a full modeset afterwards so drm_vblank_on() may not get called. In such a case if someone is holding a vblank reference across a suspend/resume cycle drm_vblank_get() called after resuming won't re-enable the vblank interrupts.
So in order to make sure vblank interrupts get re-enabled post-resume, call drm_vblank_on() in intel_sanitize_crtc() if the crtc is already active.
v2: Also drm_vblank_off() if the pipe got disabled magically
Signed-off-by: Ville Syrjälä ville.syrjala@linux.intel.com --- drivers/gpu/drm/i915/intel_display.c | 6 ++++++ 1 file changed, 6 insertions(+)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 9420f4f..2e9f0b0 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -11708,6 +11708,12 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc) reg = PIPECONF(crtc->config.cpu_transcoder); I915_WRITE(reg, I915_READ(reg) & ~PIPECONF_FRAME_START_DELAY_MASK);
+ /* restore vblank interrupts to correct state */ + if (crtc->active) + drm_vblank_on(dev, crtc->pipe); + else + drm_vblank_off(dev, crtc->pipe); + /* We need to sanitize the plane -> pipe mapping first because this will * disable the crtc (and hence change the state) if it is wrong. Note * that gen4+ has a fixed plane -> pipe mapping. */
On Tue, May 20, 2014 at 05:20:05PM +0300, ville.syrjala@linux.intel.com wrote:
From: Ville Syrjälä ville.syrjala@linux.intel.com
If a pipe is already active when we init/resume there might not be a full modeset afterwards so drm_vblank_on() may not get called. In such a case if someone is holding a vblank reference across a suspend/resume cycle drm_vblank_get() called after resuming won't re-enable the vblank interrupts.
So in order to make sure vblank interrupts get re-enabled post-resume, call drm_vblank_on() in intel_sanitize_crtc() if the crtc is already active.
v2: Also drm_vblank_off() if the pipe got disabled magically
Signed-off-by: Ville Syrjälä ville.syrjala@linux.intel.com
This seems to duct-tape over the funny failure I'm seeing on my snb where kms_flip/vblank-vs-modeset|dpms-suspend work nicely, but kms_flip/vblank-vs-suspend has some totally hilarious random vblank frame counter after the modeset.
Testecase: igt/kms_flip/vblank-vs-suspend Tested-by: Daniel Vetter daniel.vetter@ffwll.ch
Now the problem I have: We already call this in the crtc_enable hook. Why does calling this here again add the necessary magic?
/me has no clue
Cheers, Daniel
drivers/gpu/drm/i915/intel_display.c | 6 ++++++ 1 file changed, 6 insertions(+)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 9420f4f..2e9f0b0 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -11708,6 +11708,12 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc) reg = PIPECONF(crtc->config.cpu_transcoder); I915_WRITE(reg, I915_READ(reg) & ~PIPECONF_FRAME_START_DELAY_MASK);
- /* restore vblank interrupts to correct state */
- if (crtc->active)
drm_vblank_on(dev, crtc->pipe);- else
drm_vblank_off(dev, crtc->pipe);- /* We need to sanitize the plane -> pipe mapping first because this will
- disable the crtc (and hence change the state) if it is wrong. Note
- that gen4+ has a fixed plane -> pipe mapping. */
-- 1.8.5.5
Intel-gfx mailing list Intel-gfx@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/intel-gfx
On Tue, May 20, 2014 at 07:31:33PM +0200, Daniel Vetter wrote:
On Tue, May 20, 2014 at 05:20:05PM +0300, ville.syrjala@linux.intel.com wrote:
From: Ville Syrjälä ville.syrjala@linux.intel.com
If a pipe is already active when we init/resume there might not be a full modeset afterwards so drm_vblank_on() may not get called. In such a case if someone is holding a vblank reference across a suspend/resume cycle drm_vblank_get() called after resuming won't re-enable the vblank interrupts.
So in order to make sure vblank interrupts get re-enabled post-resume, call drm_vblank_on() in intel_sanitize_crtc() if the crtc is already active.
v2: Also drm_vblank_off() if the pipe got disabled magically
Signed-off-by: Ville Syrjälä ville.syrjala@linux.intel.com
This seems to duct-tape over the funny failure I'm seeing on my snb where kms_flip/vblank-vs-modeset|dpms-suspend work nicely, but kms_flip/vblank-vs-suspend has some totally hilarious random vblank frame counter after the modeset.
Testecase: igt/kms_flip/vblank-vs-suspend Tested-by: Daniel Vetter daniel.vetter@ffwll.ch
Now the problem I have: We already call this in the crtc_enable hook. Why does calling this here again add the necessary magic?
/me has no clue
Well, we can figure this out later.
Queued for -next, thanks for the patch. -Daniel
Cheers, Daniel
drivers/gpu/drm/i915/intel_display.c | 6 ++++++ 1 file changed, 6 insertions(+)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 9420f4f..2e9f0b0 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -11708,6 +11708,12 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc) reg = PIPECONF(crtc->config.cpu_transcoder); I915_WRITE(reg, I915_READ(reg) & ~PIPECONF_FRAME_START_DELAY_MASK);
- /* restore vblank interrupts to correct state */
- if (crtc->active)
drm_vblank_on(dev, crtc->pipe);- else
drm_vblank_off(dev, crtc->pipe);- /* We need to sanitize the plane -> pipe mapping first because this will
- disable the crtc (and hence change the state) if it is wrong. Note
- that gen4+ has a fixed plane -> pipe mapping. */
-- 1.8.5.5
Intel-gfx mailing list Intel-gfx@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/intel-gfx
-- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch
dri-devel@lists.freedesktop.org
-
Daniel Vetter -
ville.syrjala＠linux.intel.com