Re: [PATCH] drm/etnaviv: don't block scheduler when GPU is still active

> > Since 45ecaea73883 ("drm/sched: Partial revert of 'drm/sched: Keep > s_fence->parent pointer'") still active jobs aren't put back in the > pending list on drm_sched_start(), as they don't have a active > parent fence anymore, so if the GPU is still working and the timeout > is extended, all cur

Re: [PATCH] drm/etnaviv: don't block scheduler when GPU is still active

On Fr, 2024-06-21 at 21:59 +0200, Lucas Stach wrote: > Since 45ecaea73883 ("drm/sched: Partial revert of 'drm/sched: Keep > s_fence->parent pointer'") still active jobs aren't put back in the > pending list on drm_sched_start(), as they don't have a active > parent fence anymore, so if the GPU is s

[PATCH] drm/etnaviv: don't block scheduler when GPU is still active

Since 45ecaea73883 ("drm/sched: Partial revert of 'drm/sched: Keep s_fence->parent pointer'") still active jobs aren't put back in the pending list on drm_sched_start(), as they don't have a active parent fence anymore, so if the GPU is still working and the timeout is extended, all currently activ

Re: [PATCH] drm/etnaviv: don't block scheduler when GPU is still active

Hi, does anyone have some bandwidth to review this? Regards, Lucas Am Freitag, dem 31.03.2023 um 13:00 +0200 schrieb Lucas Stach: > Since 45ecaea73883 ("drm/sched: Partial revert of 'drm/sched: Keep > s_fence->parent pointer'") still active jobs aren't put back in the > pending list on drm_sched

[PATCH] drm/etnaviv: don't block scheduler when GPU is still active

Since 45ecaea73883 ("drm/sched: Partial revert of 'drm/sched: Keep s_fence->parent pointer'") still active jobs aren't put back in the pending list on drm_sched_start(), as they don't have a active parent fence anymore, so if the GPU is still working and the timeout is extended, all currently activ