[igt-dev] [PATCH i-g-t v2] tests/kms_hdr: Fix failure to read output_bpc

[Rodrigo Siqueira Jordao]

Hi Lakshminarayana,

It looks like that we have another false-positive result from the CI in 
this patch.

Thanks
Siqueira

On 2021-10-08 4:59 a.m., Stylon Wang wrote:
> The first commit with TEST_ONLY flag in bpc-switch subtests
> could fail the test, because previous state can be with any bpc
> or even wth no crtc assigned to the connector.
> 
> v2:
> - Keep the test commit for its original purpose of checking
>    if 512x512 plane size works for the hardware.
> - Add new atomic commit so the setting of 8 bpc really takes effect.
> 
> Signed-off-by: Stylon Wang <stylon.wang at amd.com>
> ---
>   tests/kms_hdr.c | 22 +++++++++++++---------
>   1 file changed, 13 insertions(+), 9 deletions(-)
> 
> diff --git a/tests/kms_hdr.c b/tests/kms_hdr.c
> index 5b8529c8..273263db 100644
> --- a/tests/kms_hdr.c
> +++ b/tests/kms_hdr.c
> @@ -213,16 +213,27 @@ static void test_bpc_switch_on_output(data_t *data, igt_output_t *output,
>   
>   		draw_hdr_pattern(&afb);
>   
> -		/* Start in 8bpc. */
> +		/* Plane may be required to fit fullscreen. Check it here and allow
> +		 * smaller plane size in following tests.
> +		 */
>   		igt_plane_set_fb(data->primary, &afb);
>   		igt_plane_set_size(data->primary, data->w, data->h);
> -		igt_output_set_prop_value(data->output, IGT_CONNECTOR_MAX_BPC, 8);
>   		ret = igt_display_try_commit_atomic(display, DRM_MODE_ATOMIC_TEST_ONLY, NULL);
>   		if (!ret) {
>   			data->w = afb.width;
>   			data->h = afb.height;
>   		}
>   
> +		/* Start in 8bpc. */
> +		igt_output_set_prop_value(data->output, IGT_CONNECTOR_MAX_BPC, 8);
> +		igt_display_commit_atomic(display, DRM_MODE_ATOMIC_ALLOW_MODESET, NULL);
> +		/*
> +		 * i915 driver doesn't expose max bpc as debugfs entry,
> +		 * so limiting assert only for amd driver.
> +		 */
> +		if (is_amdgpu_device(data->fd))
> +			assert_output_bpc(data, 8);
> +
>   		/*
>   		 * amdgpu requires a primary plane when the CRTC is enabled.
>   		 * However, some older Intel hardware (hsw) have scaling
> @@ -232,13 +243,6 @@ static void test_bpc_switch_on_output(data_t *data, igt_output_t *output,
>   		if (!is_amdgpu_device(data->fd))
>   			igt_plane_set_fb(data->primary, NULL);
>   
> -		/*
> -		 * i915 driver doesn't expose max bpc as debugfs entry,
> -		 * so limiting assert only for amd driver.
> -		 */
> -		if (is_amdgpu_device(data->fd))
> -			assert_output_bpc(data, 8);
> -
>   		/* Switch to 10bpc. */
>   		igt_output_set_prop_value(data->output, IGT_CONNECTOR_MAX_BPC, 10);
>   		igt_display_commit_atomic(display, DRM_MODE_ATOMIC_ALLOW_MODESET, NULL);
>