Re: [PATCH v2 2/4] drm/v3d: Refactor perfmon locking

[Maíra Canal <mcanal@igalia.com>]

Hi Iago,

Thanks for your review!

On 6/1/26 08:52, Iago Toral wrote:
> El dom, 31-05-2026 a las 17:18 -0300, Maíra Canal escribió:
> 
> (...)
>> diff --git a/drivers/gpu/drm/v3d/v3d_gem.c
>> b/drivers/gpu/drm/v3d/v3d_gem.c
>> index 1ee3c038d5f6..9487ab7acd03 100644
>> --- a/drivers/gpu/drm/v3d/v3d_gem.c
>> +++ b/drivers/gpu/drm/v3d/v3d_gem.c
>> @@ -137,7 +137,8 @@ v3d_reset(struct v3d_dev *v3d)
>>   	v3d_mmu_set_page_table(v3d);
>>   	v3d_irq_reset(v3d);
>>   
>> -	v3d_perfmon_stop(v3d, v3d->active_perfmon, false);
>> +	/* Re-arm the global perfmon HW counters that the reset
>> zeroed. */
>> +	v3d_perfmon_resume(v3d);
> 
> What would happen if the reset happens when a non-global perfmon is
> active?
> 

Nothing. In v3d_gpu_reset_for_timeout(), before calling v3d_reset(), we
stop job->perfmon and set v3d->perfmon_state.active = NULL.

@@ -711,6 +693,8 @@ v3d_gpu_reset_for_timeout(struct v3d_dev *v3d, 
struct drm_sched_job *sched_job,
  	if (sched_job)
  		drm_sched_increase_karma(sched_job);

+	v3d_perfmon_stop(v3d, job->perfmon, false);
+
  	/* get the GPU back into the init state */
  	v3d_reset(v3d);

Best regards,
- Maíra

> Iago
>