From: Ville Syrjälä <[email protected]> We currently hold a vblank referenced while trying to evade the vblank, but we drop it as soon as we've done that. After all the planes have been committed we are quite likely to grab a new vblank reference for delivering the flip event. This causes the vblank interrupt to do a enable->enable->disable ping-pong during many commits. If we instead hang on to the original vblank reference across the entire commit we can eliminate that ping-pong.
Signed-off-by: Ville Syrjälä <[email protected]> --- drivers/gpu/drm/i915/intel_sprite.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 26c8bdcd7e72..337e884252e5 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -137,8 +137,6 @@ void intel_pipe_update_start(struct intel_crtc *crtc) finish_wait(wq, &wait); - drm_crtc_vblank_put(&crtc->base); - crtc->debug.scanline_start = scanline; crtc->debug.start_vbl_time = ktime_get(); crtc->debug.start_vbl_count = intel_crtc_get_vblank_counter(crtc); @@ -185,6 +183,15 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work crtc->base.state->event = NULL; } + /* + * The reference was taken in intel_pipe_update_start(). It could + * have been dropped as soon as the vblank was evaded, but we hold + * on to it until this time to avoid the extra vblank interrupt + * enable->disable->enable ping-pong whenever we have to deliver + * an event. + */ + drm_crtc_vblank_put(&crtc->base); + local_irq_enable(); if (intel_vgpu_active(dev_priv)) -- 2.10.2 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
