[igt-dev] [PATCH i-g-t] tests/kms_lease: Terminate lease on bad connector id

Nautiyal, Ankit K ankit.k.nautiyal at intel.com
Thu Jun 8 05:51:45 UTC 2023


On 6/8/2023 11:12 AM, Nautiyal, Ankit K wrote:
> Hi Thasleem,
>
> I think we should have terminate lease in igt_fixture where we have 
> igt_display_fini and all.
>
> There are many places where lease->fd remains open, during different 
> asserts, skip_ons throughout the test.
>
> Regards,
>
> Ankit
>
> On 4/30/2022 7:13 AM, Mohammed Thasleem wrote:
>> igt_skip_on make resource busy for sub sequence tests.
>> To avoid this terminate the lease on bad connector id.
>>
>> Signed-off-by: Mohammed Thasleem <mohammed.thasleem at intel.com>
>> ---
>>   tests/kms_lease.c | 4 ++++
>>   1 file changed, 4 insertions(+)
>>
>> diff --git a/tests/kms_lease.c b/tests/kms_lease.c
>> index 29a774ae8..03538cf8c 100644
>> --- a/tests/kms_lease.c
>> +++ b/tests/kms_lease.c
>> @@ -682,6 +682,10 @@ static void lease_unleased_connector(data_t *data)
>>               bad_connector_id = data->master.display.outputs[o].id;
>>       }
>>   +    /* Terminate lease on bad connector id before skip. */
>> +    if (bad_connector_id == 0)
>> +        terminate_lease(&lease);
>
> Perhaps we can add igt_skip("Couldnt find another connector"); in the 
> if block only.
>
> This scenario can happen in another subtest as well


Please ignore this. as mentioned above, I think its better to have 
terminate_lease in igt_fixture.

Regards,

Ankit

>
>> +
>>       /* Give up if there isn't another connector */
>>       igt_skip_on(bad_connector_id == 0);


More information about the igt-dev mailing list