[v3,4/5] drm/i915: Do not call drm_crtc_arm_vblank_event in async flips

Commit Message

Karthik B S May 28, 2020, 5:39 a.m. UTC

Since the flip done event will be sent in the flip_done_handler,
no need to add the event to the list and delay it for later.

v2: -Moved the async check above vblank_get as it
     was causing issues for PSR.

v3: -No need to wait for vblank to pass, as this wait was causing a
     16ms delay once every few flips.

Signed-off-by: Karthik B S <[e-mail]>
---
 drivers/gpu/drm/i915/display/intel_sprite.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

Patch

diff --git a/drivers/gpu/drm/i915/display/intel_sprite.c b/drivers/gpu/drm/i915/display/intel_sprite.c
index 571c36f929bd..a67621887c42 100644
--- a/drivers/gpu/drm/i915/display/intel_sprite.c
+++ b/drivers/gpu/drm/i915/display/intel_sprite.c
@@ -92,6 +92,9 @@  void intel_pipe_update_start(const struct intel_crtc_state *new_crtc_state)
 	DEFINE_WAIT(wait);
 	u32 psr_status;
 
+	if (new_crtc_state->uapi.async_flip)
+		goto irq_disable;
+
 	vblank_start = adjusted_mode->crtc_vblank_start;
 	if (adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE)
 		vblank_start = DIV_ROUND_UP(vblank_start, 2);
@@ -205,7 +208,7 @@  void intel_pipe_update_end(struct intel_crtc_state *new_crtc_state)
 	 * Would be slightly nice to just grab the vblank count and arm the
 	 * event outside of the critical section - the spinlock might spin for a
 	 * while ... */
-	if (new_crtc_state->uapi.event) {
+	if (new_crtc_state->uapi.event && !new_crtc_state->uapi.async_flip) {
 		drm_WARN_ON(&dev_priv->drm,
 			    drm_crtc_vblank_get(&crtc->base) != 0);
 
@@ -219,6 +222,9 @@  void intel_pipe_update_end(struct intel_crtc_state *new_crtc_state)
 
 	local_irq_enable();
 
+	if (new_crtc_state->uapi.async_flip)
+		return;
+
 	if (intel_vgpu_active(dev_priv))
 		return;