PolytreeDE/picom
1
0
Fork 0
mirror of https://github.com/yshui/picom.git synced 2024-11-25 14:06:08 -05:00
Code Releases Activity

vblank: testing

Browse source 
Signed-off-by: Yuxuan Shui <yshuiv7@gmail.com>
This commit is contained in:
Yuxuan Shui 2024-08-07 21:35:49 +01:00
parent 81de8762c5
commit 7280ee3446
No known key found for this signature in database
GPG key ID: D3A4405BE6CC17F4
1 changed files with 34 additions and 17 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

51
src/vblank.c
View file

@ -87,6 +87,21 @@ struct sgi_video_sync_vblank_scheduler {
pthread_t sync_thread; pthread_t sync_thread;
bool running, error, vblank_requested; bool running, error, vblank_requested;
unsigned int last_msc; unsigned int last_msc;
/// Number of synthesized vblank event. Currently these are being inserted when
/// the driver reports duplicate msc to us.
///
/// This is because the NVIDIA driver has been observed to recover from a
/// duplicated msc loop by us rendering a new frame. So we insert a fake vblank
/// event so the render loop can progress, but doing so makes our msc desync from
/// the driver's msc. An hypothetical sequence of events go like this:
///
/// - driver -> msc 1, we report msc 1
/// - driver -> msc 1 (duplicate msc! if we don't render a new frame, we'll
/// be stuck here forever)
/// - we report msc 2
/// - new frame rendered, driver recover
/// - driver -> msc 2, but we already reported msc 2, so we have to report msc 3
unsigned int vblank_inserted;
/// Protects `running`, and `vblank_requested` /// Protects `running`, and `vblank_requested`
pthread_mutex_t vblank_requested_mtx; pthread_mutex_t vblank_requested_mtx;
@ -323,28 +338,30 @@ static void sgi_video_sync_scheduler_deinit(struct vblank_scheduler *base) {
static void static void
sgi_video_sync_scheduler_callback(EV_P attr_unused, ev_async *w, int attr_unused revents) { sgi_video_sync_scheduler_callback(EV_P attr_unused, ev_async *w, int attr_unused revents) {
auto sched = container_of(w, struct sgi_video_sync_vblank_scheduler, notify); auto sched = container_of(w, struct sgi_video_sync_vblank_scheduler, notify);
auto msc = atomic_load(&sched->current_msc); auto msc = atomic_load(&sched->current_msc) + sched->vblank_inserted;
if (sched->last_msc == msc) { auto ust = atomic_load(&sched->current_ust);
if (sched->last_msc >= msc) {
// NVIDIA spams us with duplicate vblank events after a suspend/resume // NVIDIA spams us with duplicate vblank events after a suspend/resume
// cycle. Recreating the X connection and GLX context seems to fix this. // cycle, or when the monitor turns off.
// Oh NVIDIA. // Fake a vblank event in this case. See comments on `vblank_inserted`
log_warn("Duplicate vblank event found with msc %d. Possible NVIDIA bug?", msc); // for more details.
log_warn("Resetting the vblank scheduler"); log_warn("Duplicate vblank event found with msc %d. Possible NVIDIA bug? "
sgi_video_sync_scheduler_deinit(&sched->base); "Number of duplicates so far: %d",
sched->base.vblank_event_requested = false; msc, sched->vblank_inserted);
if (!sgi_video_sync_scheduler_init(&sched->base)) { sched->last_msc++;
log_error("Failed to reset the vblank scheduler"); sched->vblank_inserted++;
} else {
sgi_video_sync_scheduler_schedule(&sched->base); struct timespec now;
} clock_gettime(CLOCK_MONOTONIC, &now);
return; ust = (uint64_t)(now.tv_sec * 1000000 + now.tv_nsec / 1000);
} else {
sched->last_msc = msc;
} }
auto event = (struct vblank_event){ auto event = (struct vblank_event){
.msc = msc, .msc = sched->last_msc,
.ust = atomic_load(&sched->current_ust), .ust = ust,
}; };
sched->base.vblank_event_requested = false; sched->base.vblank_event_requested = false;
sched->last_msc = msc;
log_verbose("Received vblank event for msc %" PRIu64, event.msc); log_verbose("Received vblank event for msc %" PRIu64, event.msc);
vblank_scheduler_invoke_callbacks(&sched->base, &event); vblank_scheduler_invoke_callbacks(&sched->base, &event);
} }