05.09.2020 13:34, Mikko Perttunen пишет:
> With job recovery becoming optional, syncpoints may have a mismatch
> between their value and max value when freed. As such, when freeing,
> set the max value to the current value of the syncpoint so that it
> is in a sane state for the next user.
>
> Signed-off-by: Mikko Perttunen <mperttunen@xxxxxxxxxx>
> ---
> drivers/gpu/host1x/syncpt.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/drivers/gpu/host1x/syncpt.c b/drivers/gpu/host1x/syncpt.c
> index 2fad8b2a55cc..82ecb4ac387e 100644
> --- a/drivers/gpu/host1x/syncpt.c
> +++ b/drivers/gpu/host1x/syncpt.c
> @@ -385,6 +385,7 @@ static void syncpt_release(struct kref *ref)
> {
> struct host1x_syncpt *sp = container_of(ref, struct host1x_syncpt, ref);
>
> + atomic_set(&sp->max_val, host1x_syncpt_read_min(sp));
> sp->locked = false;
>
> mutex_lock(&sp->host->syncpt_mutex);
>
Please note that the sync point state actually needs to be completely
reset at the sync point request-time because both downstream fastboot
and upstream u-boot [1] are needlessly enabling display VBLANK interrupt
that continuously increments sync point #26 during of kernel boot until
display controller is reset.
[1] https://github.com/u-boot/u-boot/blob/master/drivers/video/tegra.c#L155
Hence once sync point #26 is requested, it will have a dirty state. So
far this doesn't have any visible effect because sync points aren't used
much.
