6.7/regression/KASAN: null-ptr-deref in amdgpu_ras_reset_error_count+0x2d6

[Alex Deucher]

On Tue, Nov 7, 2023 at 1:18 PM Mikhail Gavrilov
<mikhail.v.gavrilov at gmail.com> wrote:
>
> On Mon, Nov 6, 2023 at 8:29 PM Alex Deucher <alexdeucher at gmail.com> wrote:
> >
> > Already fixed in this commit:
> > https://gitlab.freedesktop.org/agd5f/linux/-/commit/d1d4c0b7b65b7fab2bc6f97af9e823b1c42ccdb0
> > Which is in included in last weeks PR.
> >
>
> Thanks, it fixed the issue above.
> But, unfortunately this is not the only problem which I see on my laptop.
> Now I am observing 100% GPU loading all the time.
> And it looks as I show on this screenshot: https://postimg.cc/QHLQncMg
>
> And another bisect round says that this commit is blame:
> ❯ git bisect good
> de59b69932e64d77445d973a101d81d6e7e670c6 is the first bad commit
> commit de59b69932e64d77445d973a101d81d6e7e670c6
> Author: Alex Deucher <alexander.deucher at amd.com>
> Date:   Wed Sep 20 13:27:58 2023 -0400
>
>     drm/amdgpu/gmc: set a default disable value for AGP
>
>     To disable AGP, the start needs to be set to a higher
>     value than the end.  Set a default disable value for
>     the AGP aperture and allow the IP specific GMC code
>     to enable it selectively be calling amdgpu_gmc_agp_location().
>
>     Reviewed-by: Christian König <christian.koenig at amd.com>
>     Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
>
>  drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c           | 27 ++++++++++++++++-------
>  drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h           |  2 ++
>  drivers/gpu/drm/amd/amdgpu/amdgpu_object.c        |  3 +++
>  drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c            |  3 ++-
>  drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c            |  3 ++-
>  drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c             |  4 ++--
>  drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c             |  4 ++--
>  drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c             |  4 ++--
>  drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c             |  3 ++-
>  drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |  2 +-
>  10 files changed, 37 insertions(+), 18 deletions(-)
>
> I checked twice and ensure that it not happens on commit
> 29495d81457a483c2859ccde59cc063034bfe47d

The attached patch should fix it.  Not sure why your GPU shows up as
busy.  The AGP aperture was just disabled.

Alex
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-drm-amdgpu-fix-AGP-init-order.patch
Type: text/x-patch
Size: 4823 bytes
Desc: not available
URL: <https://lists.freedesktop.org/archives/amd-gfx/attachments/20231107/1a6e9751/attachment.bin>