Kulikov Vasiliy <segooon@gmail.com> writes:

quoted text
> set_current_state() is called only once before the first iteration.
> After return from schedule_timeout() current state is TASK_RUNNING. If
> we are going to wait again, set_current_state() must be called.
>
> Signed-off-by: Kulikov Vasiliy <segooon@gmail.com>
> ---
>  drivers/gpu/drm/nouveau/nouveau_fence.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
>
Thanks, I've pushed this patch to the Nouveau kernel tree.

quoted text
> diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
> index 813d853..b7777ae 100644
> --- a/drivers/gpu/drm/nouveau/nouveau_fence.c
> +++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
> @@ -186,7 +186,6 @@ nouveau_fence_wait(void *sync_obj, void *sync_arg, bool lazy, bool intr)
>  	unsigned long timeout = jiffies + (3 * DRM_HZ);
>  	int ret = 0;
>  
> -	__set_current_state(intr ? TASK_INTERRUPTIBLE : TASK_UNINTERRUPTIBLE);
>  
>  	while (1) {
>  		if (nouveau_fence_signalled(sync_obj, sync_arg))
> @@ -197,6 +196,8 @@ nouveau_fence_wait(void *sync_obj, void *sync_arg, bool lazy, bool intr)
>  			break;
>  		}
>  
> +		__set_current_state(intr ? TASK_INTERRUPTIBLE
> +			: TASK_UNINTERRUPTIBLE);
>  		if (lazy)
>  			schedule_timeout(1);