drm/i915: Close race between processing unpin task and queueing the flip

Before queuing the flip but crucially after attaching the unpin-work to
the crtc, we continue to setup the unpin-work. However, should the
hardware fire early, we see the connected unpin-work and queue the task.
The task then promptly runs and unpins the fb before we finish taking
the required references or even pinning it... Havoc.

To close the race, we use the flip-pending atomic to indicate when the
flip is finally setup and enqueued. So during the flip-done processing,
we can check more accurately whether the flip was expected.

v2: Add the appropriate mb() to ensure that the writes to the page-flip
worker are complete prior to marking it active and emitting the MI_FLIP.
On the read side, the mb should be enforced by the spinlocks.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@vger.kernel.org
[danvet: Review the barriers a bit, we need a write barrier both
before and after updating ->pending. Similarly we need a read barrier
in the interrupt handler both before and after reading ->pending. With
well-ordered irqs only one barrier in each place should be required,
but since this patch explicitly sets out to combat spurious interrupts
with is staged activation of the unpin work we need to go full-bore on
the barriers, too. Discussed with Chris Wilson on irc and changes
acked by him.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

1 files changed, 32 insertions, 7 deletions

diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 78d12c471472..f8ee3d16109c 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -6929,11 +6929,18 @@ static void do_intel_finish_page_flip(struct drm_device *dev,
 
         spin_lock_irqsave(&dev->event_lock, flags);
         work = intel_crtc->unpin_work;
-        if (work == NULL || !work->pending) {
+
+        /* Ensure we don't miss a work->pending update ... */
+        smp_rmb();
+
+        if (work == NULL || atomic_read(&work->pending) < INTEL_FLIP_COMPLETE) {
                 spin_unlock_irqrestore(&dev->event_lock, flags);
                 return;
         }
 
+        /* and that the unpin work is consistent wrt ->pending. */
+        smp_rmb();
+
         intel_crtc->unpin_work = NULL;
 
         if (work->event)
@@ -6977,16 +6984,25 @@ void intel_prepare_page_flip(struct drm_device *dev, int plane)
                 to_intel_crtc(dev_priv->plane_to_crtc_mapping[plane]);
         unsigned long flags;
 
+        /* NB: An MMIO update of the plane base pointer will also
+         * generate a page-flip completion irq, i.e. every modeset
+         * is also accompanied by a spurious intel_prepare_page_flip().
+         */
         spin_lock_irqsave(&dev->event_lock, flags);
-        if (intel_crtc->unpin_work) {
-                if ((++intel_crtc->unpin_work->pending) > 1)
-                        DRM_ERROR("Prepared flip multiple times\n");
-        } else {
-                DRM_DEBUG_DRIVER("preparing flip with no unpin work?\n");
-        }
+        if (intel_crtc->unpin_work)
+                atomic_inc_not_zero(&intel_crtc->unpin_work->pending);
         spin_unlock_irqrestore(&dev->event_lock, flags);
 }
 
+inline static void intel_mark_page_flip_active(struct intel_crtc *intel_crtc)
+{
+        /* Ensure that the work item is consistent when activating it ... */
+        smp_wmb();
+        atomic_set(&intel_crtc->unpin_work->pending, INTEL_FLIP_PENDING);
+        /* and that it is marked active as soon as the irq could fire. */
+        smp_wmb();
+}
+
 static int intel_gen2_queue_flip(struct drm_device *dev,
                                  struct drm_crtc *crtc,
                                  struct drm_framebuffer *fb,
@@ -7020,6 +7036,8 @@ static int intel_gen2_queue_flip(struct drm_device *dev,
         intel_ring_emit(ring, fb->pitches[0]);
         intel_ring_emit(ring, obj->gtt_offset + intel_crtc->dspaddr_offset);
         intel_ring_emit(ring, 0); /* aux display base address, unused */
+
+        intel_mark_page_flip_active(intel_crtc);
         intel_ring_advance(ring);
         return 0;
 
@@ -7060,6 +7078,7 @@ static int intel_gen3_queue_flip(struct drm_device *dev,
         intel_ring_emit(ring, obj->gtt_offset + intel_crtc->dspaddr_offset);
         intel_ring_emit(ring, MI_NOOP);
 
+        intel_mark_page_flip_active(intel_crtc);
         intel_ring_advance(ring);
         return 0;
 
@@ -7106,6 +7125,8 @@ static int intel_gen4_queue_flip(struct drm_device *dev,
         pf = 0;
         pipesrc = I915_READ(PIPESRC(intel_crtc->pipe)) & 0x0fff0fff;
         intel_ring_emit(ring, pf | pipesrc);
+
+        intel_mark_page_flip_active(intel_crtc);
         intel_ring_advance(ring);
         return 0;
 
@@ -7148,6 +7169,8 @@ static int intel_gen6_queue_flip(struct drm_device *dev,
         pf = 0;
         pipesrc = I915_READ(PIPESRC(intel_crtc->pipe)) & 0x0fff0fff;
         intel_ring_emit(ring, pf | pipesrc);
+
+        intel_mark_page_flip_active(intel_crtc);
         intel_ring_advance(ring);
         return 0;
 
@@ -7202,6 +7225,8 @@ static int intel_gen7_queue_flip(struct drm_device *dev,
         intel_ring_emit(ring, (fb->pitches[0] | obj->tiling_mode));
         intel_ring_emit(ring, obj->gtt_offset + intel_crtc->dspaddr_offset);
         intel_ring_emit(ring, (MI_NOOP));
+
+        intel_mark_page_flip_active(intel_crtc);
         intel_ring_advance(ring);
         return 0;