Re: [Intel-gfx] [PATCH] drm/i915/guc: Remove racey GEM_BUG_ON

2021-12-09 Thread Daniele Ceraolo Spurio
On 12/9/2021 11:57 AM, Matthew Brost wrote: On Thu, Dec 09, 2021 at 11:26:09AM -0800, Daniele Ceraolo Spurio wrote: On 12/9/2021 10:51 AM, Matthew Brost wrote: A full GT can race with the last context put resulting in the context forgot to mention earlier but you're missing "reset" here

Re: [Intel-gfx] [PATCH] drm/i915/guc: Remove racey GEM_BUG_ON

2021-12-09 Thread Matthew Brost
On Thu, Dec 09, 2021 at 11:26:09AM -0800, Daniele Ceraolo Spurio wrote: > > > On 12/9/2021 10:51 AM, Matthew Brost wrote: > > A full GT can race with the last context put resulting in the context > > ref count being zero but the destroyed bit not yet being set. Remove > > GEM_BUG_ON in scrub_guc_

Re: [Intel-gfx] [PATCH] drm/i915/guc: Remove racey GEM_BUG_ON

2021-12-09 Thread Daniele Ceraolo Spurio
On 12/9/2021 10:51 AM, Matthew Brost wrote: A full GT can race with the last context put resulting in the context ref count being zero but the destroyed bit not yet being set. Remove GEM_BUG_ON in scrub_guc_desc_for_outstanding_g2h that asserts the destroyed bit must be set in ref count is zer

[Intel-gfx] [PATCH] drm/i915/guc: Remove racey GEM_BUG_ON

2021-12-09 Thread Matthew Brost
A full GT can race with the last context put resulting in the context ref count being zero but the destroyed bit not yet being set. Remove GEM_BUG_ON in scrub_guc_desc_for_outstanding_g2h that asserts the destroyed bit must be set in ref count is zero. Signed-off-by: Matthew Brost --- drivers/gp