Re: [PATCH i-g-t v2] tests/kms_properties: Add skip on headless machines

["Kishore, Sumit" <sumit.kishore@intel.com>]

On 19-03-2026 00:36, Ville Syrjälä wrote:
> On Wed, Mar 18, 2026 at 07:31:20PM +0530, sumit kishore wrote:
>> Print debug messages for sub-tests which are expected to
>> skip on headless display systems.
>>
>> v2: Use existing library function, instead of
>>      defining new (karthik)
>>
>> Signed-off-by: sumit kishore <sumit.kishore@intel.com>
>> ---
>>   tests/kms_properties.c | 4 ++++
>>   1 file changed, 4 insertions(+)
>>
>> diff --git a/tests/kms_properties.c b/tests/kms_properties.c
>> index f735839da..9f5649c90 100644
>> --- a/tests/kms_properties.c
>> +++ b/tests/kms_properties.c
>> @@ -377,6 +377,8 @@ static void plane_properties(igt_display_t *display, bool atomic)
>>   	igt_output_t *output;
>>   	igt_crtc_t *crtc;
>>   
>> +	igt_display_require_output(display);
>> +
>>   	for_each_crtc_with_single_output(display, crtc, output) {
>>   		igt_display_reset(display);
>>   
>> @@ -400,6 +402,8 @@ static void crtc_properties(igt_display_t *display, bool atomic)
>>   	igt_crtc_t *crtc;
>>   	igt_output_t *output;
>>   
>> +	igt_display_require_output(display);
>> +
> I'm thinking these tests shouldn't even need an output. It should be
> possible to set (at least most) plane/crtc properties regardless of
> whether they're active or not.

this test is already being skipped on headless displays, when it hits 
the check "for_each_crtc_with_single_output".

my changes are printing the reason for the skip to add more clarity in 
the logs.

in case if we need to set the properties at crtc level, regardless of 
headless/active displays,

it can be taken as a separate iGT improvement task as it will require 
changes at multiple places, or re-design of the test case itself.

>>   	for_each_crtc_with_single_output(display, crtc, output) {
>>   		igt_display_reset(display);
>>   
>> -- 
>> 2.43.0