[PATCH 2/2] drm/amdkfd: Stop using GFP_NOIO explicitly for GFX10

[Felix Kuehling]

On 2019-11-12 4:35 p.m., Yong Zhao wrote:
> Hi Felix,
>
> See one thing inline I am not too sure.
>
> Yong
>
> On 2019-11-12 4:30 p.m., Felix Kuehling wrote:
>> On 2019-11-12 4:26 p.m., Yong Zhao wrote:
>>> Adapt the change from 1cd106ecfc1f04
>>>
>>> The change is:
>>>
>>>      drm/amdkfd: Stop using GFP_NOIO explicitly
>>>
>>>      This is no longer needed with the memalloc_nofs_save/restore in
>>>      dqm_lock/unlock
>>>
>>> Change-Id: I42450b2c149d2b1842be99a8f355c829a0079e7c
>>> Signed-off-by: Yong Zhao <Yong.Zhao at amd.com>
>>
>> The series is
>>
>> Reviewed-by: Felix Kuehling <Felix.Kuehling at amd.com>
>>
>>
>>> ---
>>>   drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c | 2 +-
>>>   drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c  | 2 +-
>>>   2 files changed, 2 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c 
>>> b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c
>>> index 46ddb33b624a..579c5ffcfa79 100644
>>> --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c
>>> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c
>>> @@ -393,7 +393,7 @@ struct mqd_manager *mqd_manager_init_v10(enum 
>>> KFD_MQD_TYPE type,
>>>       if (WARN_ON(type >= KFD_MQD_TYPE_MAX))
>>>           return NULL;
>>>   -    mqd = kzalloc(sizeof(*mqd), GFP_NOIO);
>>> +    mqd = kzalloc(sizeof(*mqd), GFP_KERNEL);
>>>       if (!mqd)
>>>           return NULL;
>>>   diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c 
>>> b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
>>> index be27ff01cdb8..22a819c888d8 100644
>>> --- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
>>> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
>>> @@ -92,7 +92,7 @@ static struct kfd_mem_obj *allocate_mqd(struct 
>>> kfd_dev *kfd,
>>>        * instead of sub-allocation function.
>>>        */
>>>       if (kfd->cwsr_enabled && (q->type == KFD_QUEUE_TYPE_COMPUTE)) {
>>> -        mqd_mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO);
>>> +        mqd_mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_KERNEL);
> [yz] This should be kept probably. With the latest code, 
> allocate_mqd() is called outside of the dqm. So now the situation is 
> different from the original one.

Memory allocations outside the DQM lock are not problematic. Only memory 
allocations under the DQM lock need to be careful to avoid FS reclaim, 
because that can lead to deadlocks with MMU notifiers. GFP_NOIO is no 
longer needed for that.

Regards,
   Felix

>>>           if (!mqd_mem_obj)
>>>               return NULL;
>>>           retval = amdgpu_amdkfd_alloc_gtt_mem(kfd->kgd,