[PATCH] drm/xe: Process message queue upon submission start

[Niranjana Vishwanathapura]

On Thu, Jun 27, 2024 at 06:01:37PM +0000, Matthew Brost wrote:
>On Thu, Jun 27, 2024 at 12:20:55AM -0700, Niranjana Vishwanathapura wrote:
>> Stopping submission also stops the drm scheduler message
>> queue. Hence, when starting the submission, also process
>> message queue if there is any message ready.
>>
>> Signed-off-by: Niranjana Vishwanathapura <niranjana.vishwanathapura at intel.com>
>> ---
>>  drivers/gpu/drm/xe/xe_gpu_scheduler.c | 1 +
>>  1 file changed, 1 insertion(+)
>>
>> diff --git a/drivers/gpu/drm/xe/xe_gpu_scheduler.c b/drivers/gpu/drm/xe/xe_gpu_scheduler.c
>> index e4ad1d6ce1d5..a77a018fd8fe 100644
>> --- a/drivers/gpu/drm/xe/xe_gpu_scheduler.c
>> +++ b/drivers/gpu/drm/xe/xe_gpu_scheduler.c
>> @@ -82,6 +82,7 @@ void xe_sched_submission_start(struct xe_gpu_scheduler *sched)
>>  {
>>  	drm_sched_wqueue_start(&sched->base);
>>  	queue_work(sched->base.submit_wq, &sched->work_process_msg);
>> +	xe_sched_process_msg_queue_if_ready(sched);
>
>I missed this in the internal review, I don't think this is required.
>
>The above line 'queue_work(sched->base.submit_wq,
>&sched->work_process_msg);' does the same thing as
>xe_sched_process_msg_queue_if_ready unless I am missing something.
>
>Can you double check if you pull this patch out internally everything
>still works?
>

Yah, looks like it is not needed.
Let us drop this change.

Thanks,
Niranjana

>Matt
>
>>  }
>>
>>  void xe_sched_submission_stop(struct xe_gpu_scheduler *sched)
>> --
>> 2.43.0
>>