[Intel-gfx] [PATCH i-g-t v6 08/21] tests/sw_sync: Add subtest test_sync_merge_same

[Chris Wilson]

On Thu, Nov 17, 2016 at 10:01:56AM -0500, robert.foss at collabora.com wrote:
> From: Robert Foss <robert.foss at collabora.com>
> 
> This subtest verifies merging a fence with itself does not fail.
> 
> Signed-off-by: Robert Foss <robert.foss at collabora.com>
> Reviewed-by: Eric Engestrom <eric at engestrom.ch>
> ---
>  tests/sw_sync.c | 27 +++++++++++++++++++++++++++
>  1 file changed, 27 insertions(+)
> 
> diff --git a/tests/sw_sync.c b/tests/sw_sync.c
> index 528be4d..51aa025 100644
> --- a/tests/sw_sync.c
> +++ b/tests/sw_sync.c
> @@ -179,6 +179,30 @@ static void test_sync_merge(void)
>  	close(timeline);
>  }
>  
> +static void test_sync_merge_same(void)
> +{
> +	int in_fence[2];
> +	int timeline;
> +	int signaled;
> +
> +	timeline = sw_sync_timeline_create();
> +	in_fence[0] = sw_sync_fence_create(timeline, 1);
> +	in_fence[1] = sync_merge(in_fence[0], in_fence[0]);
> +
> +	signaled = sync_fence_count_status(in_fence[0],
> +					      SW_SYNC_FENCE_STATUS_SIGNALED);
> +	igt_assert_f(signaled == 0, "fence signaled too early\n");
> +
> +	sw_sync_timeline_inc(timeline, 1);
> +	signaled = sync_fence_count_status(in_fence[0],
> +					      SW_SYNC_FENCE_STATUS_SIGNALED);
> +	igt_assert_f(signaled == 1, "fence did not signal\n");

But what in_fence[1]? How does the lifetime of in_fence[1] relate to
in_fence[0]?
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre