On 3/16/26 19:16, Alex Deucher wrote:
> Return -ETIME rather than -ECANCELED for guilty contexts.
> Userspace only considers contexts to be guilty if they
> return -ETIME. Returning -ECANCELED means the context was
> innocent.
>
> Reviewed-by: Jesse Zhang <[email protected]>
> Signed-off-by: Alex Deucher <[email protected]>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> index 53e2eec734b1a..06dfe01492616 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> @@ -62,7 +62,7 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p,
>
> if (atomic_read(&p->ctx->guilty)) {
> amdgpu_ctx_put(p->ctx);
> - return -ECANCELED;
> + return -ETIME;
I think we should rather stop using p->ctx->guilty at all.
This is just completely unnecessary churn. The per entity handling already
correctly returns -ECANCELED or -ETIME depending on the fence error.
Regards,
Christian.
> }
>
> amdgpu_sync_create(&p->sync);
