On 1/8/26 13:54, Liang, Prike wrote:
> [Public]
> 
> Hi Alex
> 
> This patch should fix the TLB flush support issue on the older GFX.
> Do you think we should revert the following change to clean up the VM layer 
> flush quirk?

Yeah, agree. The patch below is much cleaner than this quirk.

Regards,
Christian.

> 
> https://lists.freedesktop.org/archives/amd-gfx/2025-December/134959.html
> 
> Regards,
>       Prike
> 
>> -----Original Message-----
>> From: Koenig, Christian <[email protected]>
>> Sent: Thursday, January 8, 2026 8:44 PM
>> To: Liang, Prike <[email protected]>; [email protected]
>> Cc: Deucher, Alexander <[email protected]>
>> Subject: Re: [PATCH] drm/amdgpu: validate the flush_gpu_tlb_pasid()
>>
>> On 1/6/26 14:19, Prike Liang wrote:
>>> Validate flush_gpu_tlb_pasid() availability before flushing tlb.
>>>
>>> Signed-off-by: Prike Liang <[email protected]>
>>
>> Reviewed-by: Christian König <[email protected]>
>>
>>> ---
>>>  drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c | 4 ++++
>>>  1 file changed, 4 insertions(+)
>>>
>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
>>> index cd4acc6adc9e..f871f1693483 100644
>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
>>> @@ -780,6 +780,10 @@ int amdgpu_gmc_flush_gpu_tlb_pasid(struct
>> amdgpu_device *adev, uint16_t pasid,
>>>             return 0;
>>>
>>>     if (!adev->gmc.flush_pasid_uses_kiq || !ring->sched.ready) {
>>> +
>>> +           if (!adev->gmc.gmc_funcs->flush_gpu_tlb_pasid)
>>> +                   return 0;
>>> +
>>>             if (adev->gmc.flush_tlb_needs_extra_type_2)
>>>                     adev->gmc.gmc_funcs->flush_gpu_tlb_pasid(adev, pasid,
>>>                                                              2, all_hub,
> 

Reply via email to