[PATCH] drm/amdgpu: stop scheduler when calling hw_fini

[Chen, Guchun]

[Public]

Hi Andrey and Christian,

Thanks for your comment, I will send out a new patch set later on after I verify it.

Regards,
Guchun

-----Original Message-----
From: Grodzovsky, Andrey <Andrey.Grodzovsky at amd.com> 
Sent: Saturday, August 28, 2021 2:28 AM
To: Koenig, Christian <Christian.Koenig at amd.com>; Chen, Guchun <Guchun.Chen at amd.com>; amd-gfx at lists.freedesktop.org; Zhang, Hawking <Hawking.Zhang at amd.com>; mike at fireburn.co.uk; Deucher, Alexander <Alexander.Deucher at amd.com>
Subject: Re: [PATCH] drm/amdgpu: stop scheduler when calling hw_fini

I don't think it will start/stop twice because amdgpu_fence_driver_hw_fini/inint is not called during reset.

I am worried about calling drm_sched_start without calling drm_sched_resubmit_job first since that the place where the jobs are actually restarted. Also calling drm_sched_start with false flag  wrong here since it skips all the pending list handling.

Andrey

On 2021-08-27 7:34 a.m., Christian König wrote:
> In general that looks good to me, but what could be is that we now try 
> to stop/start the scheduler during reset twice.
>
> Andrey what do you think?
>
> Christian.
>
> Am 27.08.21 um 12:40 schrieb Guchun Chen:
>> This gurantees no more work on the ring can be submitted to hardware 
>> in suspend/resume case, otherwise the ring will not be empty before 
>> suspend.
>>
>> Suggested-by: Christian König <christian.koenig at amd.com>
>> Signed-off-by: Guchun Chen <guchun.chen at amd.com>
>> ---
>>   drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 6 ++++++
>>   1 file changed, 6 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
>> index b439eb7d4177..d6e429e63604 100644
>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
>> @@ -552,6 +552,9 @@ void amdgpu_fence_driver_hw_fini(struct
>> amdgpu_device *adev)
>>           if (!ring || !ring->fence_drv.initialized)
>>               continue;
>>   +        if (!ring->no_scheduler)
>> +            drm_sched_stop(&ring->sched, NULL);
>> +
>>           /* You can't wait for HW to signal if it's gone */
>>           if (!drm_dev_is_unplugged(&adev->ddev))
>>               r = amdgpu_fence_wait_empty(ring); @@ -611,6 +614,9 @@ 
>> void amdgpu_fence_driver_hw_init(struct
>> amdgpu_device *adev)
>>           if (!ring || !ring->fence_drv.initialized)
>>               continue;
>>   +        if (!ring->no_scheduler)
>> +            drm_sched_start(&ring->sched, false);
>> +
>>           /* enable the interrupt */
>>           if (ring->fence_drv.irq_src)
>>               amdgpu_irq_get(adev, ring->fence_drv.irq_src,
>