[igt-dev] [PATCH i-g-t] tests/kms_plane_cursor: Pipe-D is invalid in big joiner

[Modem, Bhanuprakash]

Hi Nidhi,

On Wed-14-06-2023 05:22 pm, Nidhi Gupta wrote:
> Pipe D is invalid in big joiner and because of that currently
> test throws as skip for all the pipes and as it is within the
> dynamic subtest block.
> Fixed in IGT to skip the big joiner validity check for pipe-D
> and execute for the rest of the pipes as it is.
> 
> Signed-off-by: Nidhi Gupta <nidhi1.gupta at intel.com>
> ---
>   tests/kms_plane_cursor.c | 5 ++++-
>   1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/tests/kms_plane_cursor.c b/tests/kms_plane_cursor.c
> index 59b105c6..38918bb7 100644
> --- a/tests/kms_plane_cursor.c
> +++ b/tests/kms_plane_cursor.c
> @@ -86,7 +86,10 @@ static void test_init(data_t *data, enum pipe pipe_id, igt_output_t *output)
>   	data->cursor = igt_pipe_get_plane_type(data->pipe, DRM_PLANE_TYPE_CURSOR);
>   
>   	igt_output_set_pipe(data->output, data->pipe_id);
> -	igt_require(i915_pipe_output_combo_valid(display));
> +
> +	/*Pipe D is invalid in big joiner*/
> +	if (data->pipe_id != PIPE_D)
> +		igt_require(i915_pipe_output_combo_valid(display));

Why do we need to call this helper for pipe-A/B/C as we already know 
these pipes can support bigjoiner? If upcoming hardware enables pip-E/F, 
then how this logic can handle?

i915_pipe_output_combo_valid() itself is taken care of last pipe (pipe-D 
in case of adl/dg2). Maybe you need to think about a way to check before 
calling the igt_dynamic().

- Bhanu

>   
>   	igt_info("Using (pipe %s + %s) to run the subtest.\n",
>   		 kmstest_pipe_name(data->pipe_id), igt_output_name(data->output));