Bug 1215695 - amdgpu ring gfx_low timeout
Summary: amdgpu ring gfx_low timeout
Status: RESOLVED DUPLICATE of bug 1215470
Alias: None
Product: openSUSE Tumbleweed
Classification: openSUSE
Component: Kernel (show other bugs)
Version: Current
Hardware: x86-64 Other
: P5 - None : Normal (vote)
Target Milestone: ---
Assignee: openSUSE Kernel Bugs
QA Contact: E-mail List
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2023-09-25 19:43 UTC by Eyad Issa
Modified: 2023-09-29 17:00 UTC (History)
3 users (show)

See Also:
Found By: ---
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---


Attachments
journalctl logs (9.87 KB, text/plain)
2023-09-25 19:43 UTC, Eyad Issa
Details
journalctl logs 2 (8.46 KB, text/plain)
2023-09-28 21:08 UTC, Eyad Issa
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Eyad Issa 2023-09-25 19:43:38 UTC
Created attachment 869735 [details]
journalctl logs

Sometimes the display crashes and seems to restart, only to crash again and again.

The only way out is forcing a reboot of the system.

Upon inspection of journalctl I found that it could be the amdgpu driver, but I can't seem to understand what is happening.

I'm using Gnome 45 with Wayland.


~> cat /etc/os-release
NAME="openSUSE Tumbleweed"
# VERSION="20230922"
Comment 1 Eyad Issa 2023-09-25 19:49:09 UTC
I forgot to add that in X11 the bug is not present
Comment 2 Daniel Garcia 2023-09-27 06:14:46 UTC
I've done some investigation about this bug and it looks like it's related to the kernel real-time schedule [3].

There's a patch sent to the kernel to try to solve this issue [1].
There's also some work done in mutter project to avoid the problem [2].

Current upstream proposed patches:
 * [1] kernel: https://patchwork.kernel.org/project/dri-devel/patch/20230926170549.2589045-1-halfline@gmail.com/
 * [2] mutter: https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3293

Related reported bugs for other distributions:
 * [3] freedesktop: https://gitlab.freedesktop.org/drm/amd/-/issues/2861
 * [4] ubuntu: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/2034619
 * [5] gnome: https://gitlab.gnome.org/GNOME/mutter/-/issues/3037
Comment 3 Atri Bhattacharya 2023-09-28 15:57:03 UTC
(In reply to Daniel Garcia from comment #2)
> I've done some investigation about this bug and it looks like it's related
> to the kernel real-time schedule [3].
> 
> There's a patch sent to the kernel to try to solve this issue [1].
> There's also some work done in mutter project to avoid the problem [2].
> 
> Current upstream proposed patches:
>  * [1] kernel:
> https://patchwork.kernel.org/project/dri-devel/patch/20230926170549.2589045-
> 1-halfline@gmail.com/
>  * [2] mutter: https://gitlab.gnome.org/GNOME/mutter/-/merge_requests/3293
> 
> Related reported bugs for other distributions:
>  * [3] freedesktop: https://gitlab.freedesktop.org/drm/amd/-/issues/2861
>  * [4] ubuntu: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/2034619
>  * [5] gnome: https://gitlab.gnome.org/GNOME/mutter/-/issues/3037

Could you please try the mutter package from https://build.opensuse.org/project/show/home:badshah400:boo1215695 and see if it helps. Please remember to try in a session without the env `MUTTER_DEBUG_KMS_THREAD_TYPE=user` set and where
you have not already executed
`chrt -a -o -p 0 $(pgrep -x gnome-shell)` or similar.

Thanks in advance.
Comment 4 Eyad Issa 2023-09-28 21:08:36 UTC
Created attachment 869827 [details]
journalctl logs 2

I don't know if it's the same bug, but I've managed to reproduce it even on X11 (still don't know how though xD)
Comment 5 Eyad Issa 2023-09-28 21:23:10 UTC
> Could you please try the mutter package from https://build.opensuse.org/project/show/home:badshah400:boo1215695 and see if it helps.

I'm trying it right now. Reporting tomorrow.
Comment 6 Eyad Issa 2023-09-28 21:48:51 UTC
(In reply to Eyad Issa from comment #5)
> > Could you please try the mutter package from https://build.opensuse.org/project/show/home:badshah400:boo1215695 and see if it helps.
> 
> I'm trying it right now. Reporting tomorrow.

Still crashing unfortunately...
Comment 7 Eyad Issa 2023-09-29 16:49:20 UTC
I forgot to add that it doesn't happen when NOT using an external monitor as a second screen (in "extend" mode).

Also, I've seen that the latest kernel (6.5.5) has multiple commits about amdgpu driver, so maybe it could be fixed by it (I hope!).
Comment 8 Eyad Issa 2023-09-29 17:00:01 UTC
Should be a duplicate of the one linked. 

Found it while looking for updates to the kernel in build.o.o.

*** This bug has been marked as a duplicate of bug 1215470 ***