[Mesa-dev] [PATCH 3/9] i965: Only wait for the fence bo to be signaled

[Chris Wilson]

Quoting Chris Wilson (2017-06-09 14:01:34)
> The fence bo may be reused as an input fence to another batch, which
> will cause us to treat it as busy until that subsequent batch is idle.
> We only need to check if the fence has been signaled, which we can do by
> checking the write status of the fence bo -- when the write is
> completed, the fence is signaled.
> 
> Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> Cc: Kenneth Graunke <kenneth at whitecape.org>
> Cc: Matt Turner <mattst88 at gmail.com>
> ---
>  src/mesa/drivers/dri/i965/brw_sync.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/src/mesa/drivers/dri/i965/brw_sync.c b/src/mesa/drivers/dri/i965/brw_sync.c
> index a8356c304f..427108610d 100644
> --- a/src/mesa/drivers/dri/i965/brw_sync.c
> +++ b/src/mesa/drivers/dri/i965/brw_sync.c
> @@ -179,7 +179,7 @@ brw_fence_has_completed_locked(struct brw_fence *fence)
>           return false;
>        }
>  
> -      if (brw_bo_busy(fence->batch_bo))
> +      if (brw_bo_write_busy(fence->batch_bo))
>           return false;

This is wrong. I was thinking of vk fence bo which are separate
entities, signaled with a dummy write. GL is using the batch as its
fence bo and so never flags it with a dummy write. We must there always
wait/check for exclusive access here.
-Chris