[PATCH v2 i-g-t 2/3] tests/kms_hdr: Set HDR meta-data while testing for panel capability

[Borah, Chaitanya Kumar]

On 8/26/2025 5:23 PM, Sharma, Swati2 wrote:
>> @@ -652,6 +652,7 @@ static void test_hdr(data_t *data, uint32_t flags)
>>   {
>>       igt_display_t *display = &data->display;
>>       igt_output_t *output;
>> +    struct hdr_output_metadata hdr;
>>       igt_display_reset(display);
>> @@ -701,6 +702,12 @@ static void test_hdr(data_t *data, uint32_t flags)
>>               prepare_test(data, output, pipe);
>> +            /* Signal HDR requirement via metadata */
>> +            fill_hdr_output_metadata_st2048(&hdr);
>> +            set_hdr_output_metadata(data, &hdr);
>> +            igt_display_commit2(display, display->is_atomic ?
>> +                        COMMIT_ATOMIC : COMMIT_LEGACY);
> If this commit fails, test will fail. Ideally, we should skip.
> Can we add try_commit() instead?
> if (igt_display_try_commit_atomic(&data->display,
>  >------->------->------->------- DRM_MODE_ATOMIC_ALLOW_MODESET,
>  >------->------->------->-------  NULL) != 0) {
>  >-------igt_info("Commit failed, couldn't set HDR metadata in 
> infoframe\n");
>  >-------break;
> }


Thank you Swati for the review.

I am planning to use

igt_display_try_commit2(&data->display, COMMIT_ATOMIC);

as we don't want to allow modeset while setting HDR metadata.
Nor do I have any other flags to set. Let me know if you have any 
objections.

==
Chaitanya