[AMD Official Use Only - AMD Internal Distribution Only]

Hi Mikhail,

Merging the patches were delayed because of the Thanksgiving holiday in Canada. 
It'll land on amd-staging-drm-next soon.

--

Regards,
Jay
________________________________
From: Mikhail Gavrilov <[email protected]>
Sent: Monday, October 13, 2025 7:03 AM
To: Pillai, Aurabindo <[email protected]>
Cc: Kazlauskas, Nicholas <[email protected]>; Wu, Ray 
<[email protected]>; Wheeler, Daniel <[email protected]>; Li, Roman 
<[email protected]>; Chung, ChiaHsuan (Tom) <[email protected]>; Deucher, 
Alexander <[email protected]>; amd-gfx list 
<[email protected]>; Linux List Kernel Mailing 
<[email protected]>; Linux regressions mailing list 
<[email protected]>
Subject: Re: 6.18/regression/bisected – BUG: sleeping function called from 
invalid context at ./include/linux/sched/mm.h:321 after 6d31602a9f57

On Tue, Oct 7, 2025 at 10:55 PM Pillai, Aurabindo
<[email protected]> wrote:
>
> [AMD Official Use Only - AMD Internal Distribution Only]
>
> Hi Mikhail,
>
> schedule_dc_vmin_vmax() has an allocation which is incorrectly using 
> GFP_KERNEL, which is likely the reason for the "sleeping function called from 
> invalid context". We have a fix queued for this week's update (switching it 
> to GFP_NOWAIT).
>

Hi Aurabindo,

6.18.0-rc1 still requires reverting commits 9d6939179491 and 6d31602a9f57
to avoid the “sleeping function called from invalid context” warnings.

The good news is that the separate WARN previously triggered at
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c:138 amdgpu_vm_set_pasid()
no longer appears in recent builds.
This WARN was a merge-only side effect (interaction between drm-next
and amd-drm-next),
though it’s unclear which exact change resolved it.

Logs for both cases are attached.

--
Best Regards,
Mike Gavrilov.

Reply via email to