[Mesa-dev] [PATCH 09/25] gallium/u_threaded: implement asynchronous flushes

[Marek Olšák]

On Sun, Oct 22, 2017 at 9:07 PM, Nicolai Hähnle <nhaehnle at gmail.com> wrote:
> @@ -107,20 +138,46 @@ static boolean si_fence_finish(struct pipe_screen *screen,
>                                uint64_t timeout)
>  {
>         struct radeon_winsys *rws = ((struct r600_common_screen*)screen)->ws;
>         struct si_multi_fence *rfence = (struct si_multi_fence *)fence;
>         struct r600_common_context *rctx;
>         int64_t abs_timeout = os_time_get_absolute_timeout(timeout);
>
>         ctx = threaded_context_unwrap_sync(ctx);
>         rctx = ctx ? (struct r600_common_context*)ctx : NULL;
>
> +       if (!util_queue_fence_is_signalled(&rfence->ready)) {
> +               if (!timeout)
> +                       return false;
> +
> +               if (rfence->tc_token) {
> +                       /* Ensure that si_flush_from_st will be called for
> +                        * this fence, but only if we're in the API thread
> +                        * where the context is current.
> +                        *
> +                        * Note that the batch containing the flush may already
> +                        * be in flight in the driver thread, so the fence
> +                        * may not be ready yet when this call returns.
> +                        */
> +                       threaded_context_flush(ctx, rfence->tc_token);
> +               }
> +
> +               if (timeout == PIPE_TIMEOUT_INFINITE) {
> +                       util_queue_fence_wait(&rfence->ready);
> +               } else {
> +                       if (!util_queue_fence_wait_timeout(&rfence->ready, abs_timeout))
> +                               return false;
> +               }
> +
> +               assert(!rfence->tc_token);

tc_token might be non-NULL if this code is executed right after
si_flush_from_st signals the fence.

Marek