Re: [PATCH i-g-t] tests/intel/kms_sharpness_filter: Restructure modifier validation into per-modifier subtests

["Naladala, Ramanaidu" <Ramanaidu.naladala@intel.com>]

Hi Pranay,

On 2/17/2026 3:00 PM, Pranay Samala wrote:
> Previously, all modifiers were exercised within a single
> subtest, making failures/skips hard to isolate and debug.
> This new flow generates explicit subtest names (filter-modifier-%s)
> for each modifier, improving clarity and CI reporting.
>
> Each subtest now validates only the selected modifier, and
> unsupported format/modifier combinations are cleanly skipped.
>
> Signed-off-by: Pranay Samala <pranay.samala@intel.com>
> ---
>   tests/intel/kms_sharpness_filter.c | 34 +++++++++++++++++++++---------
>   1 file changed, 24 insertions(+), 10 deletions(-)
>
> diff --git a/tests/intel/kms_sharpness_filter.c b/tests/intel/kms_sharpness_filter.c
> index eddfad6b9..0140363bc 100644
> --- a/tests/intel/kms_sharpness_filter.c
> +++ b/tests/intel/kms_sharpness_filter.c
> @@ -21,9 +21,6 @@
>    * SUBTEST: filter-strength
>    * Description: Verify that varying strength (0-255), affects the degree of sharpeness applied.
>    *
> - * SUBTEST: filter-modifiers
> - * Description: Verify content adaptive sharpness filter with varying modifiers.
> - *
>    * SUBTEST: filter-rotations
>    * Description: Verify content adaptive sharpness filter with varying rotations.
>    *
> @@ -70,6 +67,17 @@
>    *              sharpness filter.
>   */
>   
> +/**
> + * SUBTEST: filter-modifier-%s
> + * Description: Verify content adaptive sharpness filter with varying modifiers.
> + *
> + * arg[1]:
> + *
> + * @linear:	linear
> + * @x-tiled:	x-tiled
> + * @4-tiled:	4-tiled
> + */
> +
>   IGT_TEST_DESCRIPTION("Test to validate content adaptive sharpness filter");
>   
>   /*
> @@ -596,16 +604,22 @@ int igt_main_args("l", NULL, help_str, opt_handler, &data)
>   		}
>   	}
>   
> -	igt_describe("Verify content adaptive sharpness filter with "
> -		     "varying modifiers.");
> -	igt_subtest_with_dynamic("filter-modifiers") {
> +	for (int i = 0; i < ARRAY_SIZE(modifiers); i++) {
>   		data.rotation = IGT_ROTATION_0;
>   		data.format = DRM_FORMAT_XRGB8888;
>   		data.filter_strength = MID_FILTER_STRENGTH;
> -
> -		for (int i = 0; i < ARRAY_SIZE(modifiers); i++) {
> -			data.modifier = modifiers[i].modifier;
> -			data.modifier_name = modifiers[i].name;
> +		data.modifier = modifiers[i].modifier;
> +		data.modifier_name = modifiers[i].name;
> +
> +		igt_describe("Verify content adaptive sharpness filter with "
> +			     "varying modifiers.");
> +		igt_subtest_with_dynamic_f("filter-modifier-%s", data.modifier_name) {
> +			igt_skip_on_f((!igt_display_has_format_mod(&data.display,
> +								   DRM_FORMAT_XBGR8888,
> +								   data.modifier)),
> +				       "Skipping: Modifier " IGT_MODIFIER_FMT
> +				       " is not supported on this platform\n",
> +				       IGT_MODIFIER_ARGS(data.modifier));
>   
>   			run_sharpness_filter_test(&data, TEST_FILTER_MODIFIERS);
IMHO, I observed that kms_sharpness_filter.c contains two separate sets 
of subtests for pixel formats and modifiers. Due to this separation, 
coverage is missing for other pixel formats when combined with 
modifiers. Currently, the test checks modifiers only with 
DRM_FORMAT_XRGB8888.

Instead, we can create a combined subtest using 
igt_subtest_with_dynamic_f("filter-pixel-format-%s-modifier-%s").

This would allow the sharpness filter to be tested across all pixel 
formats with all modifiers.


>   		}