[PATCH v2] drm/amdkfd: Handle deallocated VPGRs in gfx11+ trap handler

[Jay Cornwall]

On 5/29/2024 16:07, Lancelot SIX wrote:

> On 29/05/2024 20:35, Jay Cornwall wrote:
>> A wavefront may deallocate its VGPRs at the end of a program while
>> waiting for memory transactions to complete. If it subsequently
>> receives a context save exception it will be unable to save,
>> since this requires VGPRs. In this case the trap handler should
>> terminate the wavefront.
>>
>> Fixes intermittent VM faults under context switching load.
>>
>> V2: Use S_ENDPGM instead of S_ENDPGM_SAVED for performance counters
> 
> Hi Jay,
> 
> Thanks for the V2.
> 
> FYI,as far as I can see, the .h part of the patch does not seem to apply 
> directly on current amd-staging-drm-next, but I guess we just have a 
> different bases.

Sorry, it's parented to the commit below. This has been submitted but is 
working its way through post-submit testing.

Thanks for the review.


commit d6449614e21cc166f888b3d5fc59cd1156ed7e7d
Author: Jay Cornwall <jay.cornwall at amd.com>
Date:   Thu May 23 09:00:28 2024 -0500

     drm/amdkfd: gfx12 context save/restore trap handler fixes