GUI freezes randomly, kernel: [drm:amdgpu_dm_commit_planes [amdgpu]] *ERROR* Waiting for fences timed out!

[Alexandru Bolboaca]

When working normally, random freezes happen. Everything freezes, except the mouse cursor.

I tried from keyboard to: switch to another window, start a terminal, run a command to restart gnome, enabled ctrl+alt+backspace and tried it to restart X windows, ctrl+alt+delete, switch to another session with ctrl+alt+F1-F9 - nothing works.

If I'm listening to something, the sound goes on for a while; sometimes it stops by itself, other times it keeps going.

The only way to recover that I've found is to stop the laptop from the power button.

For a while I thought it was due to firefox; I installed the deb version and then switched to chrome - same behavior.

It didn't happen when I was playing Hollow Knight in Steam, or when I've used the terminal and nothing else.

ProblemType: Bug
DistroRelease: Ubuntu 22.04
Package: gnome-shell 42.2-0ubuntu0.2
ProcVersionSignature: Ubuntu 5.15.0-40.43-generic 5.15.35
Uname: Linux 5.15.0-40-generic x86_64
NonfreeKernelModules: nvidia_modeset nvidia
ApportVersion: 2.20.11-0ubuntu82.1
Architecture: amd64
CasperMD5CheckResult: pass
CurrentDesktop: ubuntu:GNOME
Date: Wed Jul 6 14:27:36 2022
DisplayManager: gdm3
InstallationDate: Installed on 2022-06-16 (19 days ago)
InstallationMedia: Ubuntu 22.04 LTS "Jammy Jellyfish" - Release amd64 (20220419)
RelatedPackageVersions: mutter-common 42.2-0ubuntu1
SourcePackage: gnome-shell
UpgradeStatus: No upgrade log present (probably fresh install)

[Daniel van Vugt (vanvugt)]

Thanks for the bug report.

First we need to exclude the most common cause of shell bugs: unsupported extensions. Please uninstall:

  '<email address hidden>',
  '<email address hidden>',
  '<email address hidden>',
  '<email address hidden>',
  '<email address hidden>'

and then log in again.

If the bug still occurs then next please check for crashes by following these steps:

1. Look in /var/crash for crash files and if found run:

    ubuntu-bug YOURFILE.crash

   Then tell us the ID of the newly-created bug.

2. If step 1 failed then look at https://errors.ubuntu.com/user/ID where ID is the content of file /var/lib/whoopsie/whoopsie-id on the machine. Do you find any links to recent problems on that page? If so then please send the links to us.

[Igor Kobzev (whiteczar-deactivatedaccount)]

Have you tried just pressing the Super ('Windows') key to get into the Activities overview? I've had this bug ever since I started using Gnome: in Ubuntu 20.04, in PopOS, in Manjaro, in Fedora. The Super key trick works every time. It'd still be nice to fix the underlying issue, though.

[Alexandru Bolboaca (alex-boly)]

I wasn't able to do anything about this - had COVID, fortunately the light form, but I still had to sleep a lot.

However, after upgrading the system I haven't had this issue anymore.

I will still keep it under observation for a few weeks or so and keep this updated.

[Alexandru Bolboaca (alex-boly)]

Of course, I say that and it happens again 5 minutes later ...

There's nothing in /var/crash.

I still had my extensions active. I will disable them and let you know what happens.

[Alexandru Bolboaca (alex-boly)]

I experimented starting Firefox using the NVIDIA card and it didn't froze for a few days.

[Alexandru Bolboaca (alex-boly)]

It froze again today, two times. It happened while working in Firefox on the NVIDIA card.

I then disabled the GNOME extensions I had installed and it still froze once (until now).

Still nothing in /var/crash.

I'm at a loss now, I don't know what else to try.

[Daniel van Vugt (vanvugt)]

Please go to step 2 in comment #2.

Please also delete the aforementioned extensions because disabling them is not reliable;

  cd ~/.local/share/gnome-shell/
  rm -rf extensions

and then log in again.

[Alexandru Bolboaca (alex-boly)]

Removed all extensions, as instructed, and rebooted. Will monitor for more freezes.

Step 2 in comment #2:

Looked at recent reports, and the most recent is a crash in color-picker on 11 July (two weeks ago): https://errors.ubuntu.com/oops/259cc932-011b-11ed-a45f-fa163e55efd0.

I've had other crashes in: pysolfc, nautilus, terminator, tracker-mine-fs, and obs-studio before this.

When the system freezes I'm not running color-picker, pysolfc or obs-studio. Never had a freeze while running terminator either. It usually happens in the browser (firefox or chrome), but that may be just because I always have a browser open. Also happened in zoom once or twice.

[Daniel van Vugt (vanvugt)]

OK, next time a freeze happens please:

1. Wait 10 seconds.

2. Reboot.

3. Run:

   journalctl -b-1 > prevboot.txt

4. Attach the resulting text file here.

[Alexandru Bolboaca (alex-boly)]

I didn't have to wait too long...

[Daniel van Vugt (vanvugt)]

Thanks. That shows it's the amdgpu kernel driver getting stuck:

iul 26 20:24:46 alexb kernel: [drm:amdgpu_dm_commit_planes [amdgpu]] *ERROR* Waiting for fences timed out!
iul 26 20:24:46 alexb kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring sdma0 timeout, signaled seq=1334, emitted seq=1336
iul 26 20:24:46 alexb kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process pid 0 thread pid 0
iul 26 20:24:46 alexb kernel: amdgpu 0000:08:00.0: amdgpu: GPU recovery disabled.
iul 26 20:24:46 alexb kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx_0.0.0 timeout, signaled seq=12220, emitted seq=12221
iul 26 20:24:46 alexb kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process Xorg pid 2753 thread Xorg:cs0 pid 2782
iul 26 20:24:46 alexb kernel: amdgpu 0000:08:00.0: amdgpu: GPU recovery disabled.

[Daniel van Vugt (vanvugt)]

The next step is to check whether this apparent kernel bug is already fixed in newer kernels like:

https://kernel.ubuntu.com/~kernel-ppa/mainline/drm-tip/2022-07-17/amd64/

https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.18.14/amd64/

[Alexandru Bolboaca (alex-boly)]

OK, thank you! That's brilliant.

I'll try to install the kernels from the PPAs and check if the bug is still present.

[Alexandru Bolboaca (alex-boly)]

I tried to install the mainline kernel and to sign it according to https://www.digitalpeer.com/blog/install-custom-or-mainline-kernel-on-ubuntu, https://ubuntu.com/blog/how-to-sign-things-for-secure-boot, https://github.com/berglh/ubuntu-sb-kernel-signing.

It seemed to work, but when I boot it says the shim signature is not recognized, or something of the sort.

I'll try some more, but perhaps someone can point me to relevant and up-to-date documentation? Thank you.

[Daniel van Vugt (vanvugt)]

Sorry I forgot to mention that you probably need to disable Secure Boot in the BIOS in order to test those.

[Alexandru Bolboaca (alex-boly)]

Thank you for the info!

Finally managed to install the latest mainline kernel (5.19.5).

Unfortunately, issue still persists. I started only firefox, using discrete video card, but in about 10 minutes it got stuck again.

Today was a pretty bad laptop day - it got stuck about 6 times in 3 hours.

Anyway, hope this helps. Anything else I can do?

[Daniel van Vugt (vanvugt)]

Try the latest drm-tip kernel:

  https://kernel.ubuntu.com/~kernel-ppa/mainline/drm-tip/2022-08-27/amd64/

and if the problem still persists with that then we would usually ask you to report it upstream but it sounds like other people already have:

  https://gitlab.freedesktop.org/drm/amd/-/issues/2135

  https://gitlab.freedesktop.org/drm/amd/-/issues

[Alexandru Bolboaca (alex-boly)]

Thank you for your support.

I will take a look at drm-tip kernels. I wish they were available through the Mainline application, it would make it much easier to install.

I guess I'll need to wait for the fix in the kernel anyway.

Cheers!

[Alexandru Bolboaca (alex-boly)]

I tried a workaround until the issue is resolved: switch to Nvidia performance profile. I find it weird that freezes still happen - less than when running AMDGPU but still happen. I also tried to blacklist AMDGPU but the login screen was not loading anymore.

Is there a way to run Ubuntu completely on Nvidia GPU?

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[janual57 (janual57)]

Hi, any work around to this topic? I've also tried latest kernels, but no improvement.
I have ROG Strix with Nvidia RTX3060 and AMD Radeon Rembrandt and I have the same error with the exact same symptoms and that makes Linux unusable.

[Daniel van Vugt (vanvugt)]

Yes it is discussed in:
https://gitlab.freedesktop.org/drm/amd/-/issues/2135

[Alexandru Bolboaca (alex-boly)]

It doesn't happen for me anymore.

I've used the laptop less for a while, but kept doing upgrades and using it for 3D apps that weren't causing it to block. I kept doing updates in this period, and occasionally tested it. I also subscribed to Ubuntu Pro with a personal license.

Then I realized in the past week or so that I didn't have the problem anymore. I decided to test it for longer and it has worked perfectly for the last 3 days.

I'm currently on Ubuntu 22.04.2 LTS, X Windows, kernel 5.19.0-32-generic #33~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC 64bits, and the latest updates from ubuntu sources.

[Daniel van Vugt (vanvugt)]

Confirmed the fix is in linux-hwe-5.19 ... it looks like this:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=77c092e054262b594614bad5e5f47e57c5d29639

[Daniel van Vugt (vanvugt)]

Nope, not fixed in 5.19.0-32-generic

https://bugs.launchpad.net/ubuntu/+bug/2008540/comments/5

[Daniel van Vugt (vanvugt)]

Now tracking in:
https://gitlab.freedesktop.org/drm/amd/-/issues/2282

[Tony Rmz (tnxhawks)]

I have updated the kernel to the version in the following link:

https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.19-rc6/amd64/

And so far everything has worked correctly, I will keep testing over the next few days and report back.

[Daniel van Vugt (vanvugt)]

That kernel is actually older than the Ubuntu 5.19 kernels. But there's always a chance it will work around the bug by pre-dating it... Not likely in this case though since we have two duplicate bugs showing it was an issue as far back as kernel 5.11.

[Tony Rmz (tnxhawks)]

If it helps at all, I also have my laptop connected to a Samsung monitor via a USB Type C port which it receives the video and charging signal.

As this article also mentions:
https://gitlab.freedesktop.org/drm/amd/-/issues/2282

[Brett Holman (holmanb)]

Since upgrading to 6.1.0-16-generic on lunar, I hit nearly the exact same error again, minus the "fences timed out" log:

[728875.081136] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring sdma0 timeout, signaled seq=263052, emitted seq=263054
[728875.081681] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process spotify pid 626324 thread spotify:cs0 pid 626354
[728875.082223] amdgpu 0000:1f:00.0: amdgpu: GPU reset begin!
[728875.674379] amdgpu: [powerplay] Failed message: 0x24, input parameter: 0x0, error code: 0xffffffff
[728875.674390] amdgpu: [powerplay] Failed message: 0x9, input parameter: 0xf4, error code: 0xffffffff
[728875.674395] amdgpu: [powerplay] Failed message: 0xa, input parameter: 0x103000, error code: 0xffffffff
[728875.674399] amdgpu: [powerplay] Failed message: 0xe, input parameter: 0x0, error code: 0xffffffff
[728875.674403] amdgpu: [powerplay] Failed message: 0x42, input parameter: 0x1, error code: 0xffffffff
[728875.674407] amdgpu: [powerplay] Failed message: 0x24, input parameter: 0x0, error code: 0xffffffff
[728875.707103] [drm] REG_WAIT timeout 10us * 3000 tries - dce110_stream_encoder_dp_blank line:936
[728895.793037] [drm:atom_op_jump [amdgpu]] *ERROR* atombios stuck in loop for more than 20secs aborting
[728895.793200] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing E18E (len 824, WS 0, PS 0) @ 0xE30E
[728895.793351] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing E048 (len 326, WS 0, PS 0) @ 0xE138
[728895.793501] [drm:dce110_link_encoder_disable_output [amdgpu]] *ERROR* dce110_link_encoder_disable_output: Failed to execute VBIOS command table!
[728915.796976] [drm:atom_op_jump [amdgpu]] *ERROR* atombios stuck in loop for more than 20secs aborting
[728915.797146] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing C30A (len 62, WS 0, PS 0) @ 0xC326
[728935.800920] [drm:atom_op_jump [amdgpu]] *ERROR* atombios stuck in loop for more than 20secs aborting
[728935.801087] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing E18E (len 824, WS 0, PS 0) @ 0xE30E
[728935.801238] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing E048 (len 326, WS 0, PS 0) @ 0xE138
[728935.801388] [drm:dce110_link_encoder_disable_output [amdgpu]] *ERROR* dce110_link_encoder_disable_output: Failed to execute VBIOS command table!
[728955.804861] [drm:atom_op_jump [amdgpu]] *ERROR* atombios stuck in loop for more than 20secs aborting
[728955.805022] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing C30A (len 62, WS 0, PS 0) @ 0xC326
[728975.808802] [drm:atom_op_jump [amdgpu]] *ERROR* atombios stuck in loop for more than 20secs aborting
[728975.808971] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing B802 (len 1359, WS 12, PS 8) @ 0xBB17

[Daniel van Vugt (vanvugt)]

Brett,

That's a different bug so please open a new one by running:

  ubuntu-bug linux

[Tony Rmz (tnxhawks)]

Hi guys, is there any update on this error? My computer is getting worse every day

antonio@antoniormz:~$ last -x | egrep "boot|crash"
reboot   system boot  5.19.0-051900rc6 Sat Mar 11 20:00   still running
antonio  :0           :0               Sat Mar 11 19:58 - crash  (00:02)
reboot   system boot  5.19.0-051900rc6 Sat Mar 11 19:57   still running
antonio  :0           :0               Sat Mar 11 18:37 - crash  (01:20)
reboot   system boot  5.19.0-051900rc6 Sat Mar 11 18:36   still running
antonio  :0           :0               Sat Mar 11 16:49 - crash  (01:46)
reboot   system boot  5.19.0-051900rc6 Sat Mar 11 16:48   still running
antonio  :0           :0               Sat Mar 11 09:09 - crash  (07:38)
reboot   system boot  5.19.0-051900rc6 Sat Mar 11 09:09   still running
antonio  :0           :0               Fri Mar 10 21:02 - crash  (12:06)
reboot   system boot  5.19.0-051900rc6 Fri Mar 10 20:57   still running
antonio  :0           :0               Fri Mar 10 19:53 - crash  (01:04)
reboot   system boot  5.19.0-051900rc6 Fri Mar 10 19:53   still running
antonio  :0           :0               Fri Mar 10 14:51 - crash  (05:01)
reboot   system boot  5.19.0-051900rc6 Fri Mar 10 14:50   still running
antonio  :0           :0               Fri Mar 10 14:41 - crash  (00:08)
reboot   system boot  5.19.0-051900rc6 Fri Mar 10 14:41   still running
antonio  :0           :0               Fri Mar 10 12:44 - crash  (01:56)
reboot   system boot  5.19.0-051900rc6 Fri Mar 10 12:44   still running
antonio  :0           :0               Fri Mar 10 12:24 - crash  (00:19)
reboot   system boot  5.19.0-051900rc6 Fri Mar 10 12:23   still running
antonio  :0           :0               Fri Mar 10 12:13 - crash  (00:10)
reboot   system boot  5.19.0-051900rc6 Fri Mar 10 12:12   still running
antonio  :0           :0               Thu Mar  9 21:03 - crash  (15:09)
reboot   system boot  5.19.0-051900rc6 Thu Mar  9 21:02   still running
antonio  :0           :0               Thu Mar  9 19:53 - crash  (01:08)
reboot   system boot  5.19.0-051900rc6 Thu Mar  9 19:53   still running
antonio  :0           :0               Wed Mar  8 20:11 - crash  (23:41)
reboot   system boot  5.19.0-051900rc6 Wed Mar  8 20:10   still running

[Tony Rmz (tnxhawks)]

I have updated the kernel following this article https://linuxhint.com/install-upgrade-latest-kernel-ubuntu-22-04/

now I'm running
antonio@antoniormz:~$ uname -r
6.2.9-060209-generic

and it has already been 6 days without a crash, is this the solution?

antonio@antoniormz:~$ last -x | egrep "boot|crash" | head
reboot system boot 6.2.9-060209-gen Sun Apr 2 22:29 still running
reboot system boot 5.19.0-051900rc6 Sun Apr 2 20:32 - 22:29 (01:56)

[- (nicoforce)]

OP from Gitlab here, just updated to Ubuntu 23.04 and am currently running kernel 6.2.0-20-generic.

I still had the issue until 5.19.0-41-generic.

So far the system has only crashed once, but it was after reboot, all the screens stopped working, but then the screen connected to the nvidia gpu came back, everything else connected to amd gpu stayed frozen, and then nvidia screen stopped working as well.

x system boot 6.2.0-20-generic Sat Apr 29 12:06 still running
x :0 :0 Sat Apr 29 11:55 - crash (00:11)
x seat0 login screen Sat Apr 29 11:55 - crash (00:11)
reboot system boot 6.2.0-20-generic Sat Apr 29 11:55 still running
reboot system boot 5.19.0-41-generi Sat Apr 29 10:11 - 11:53 (01:42)
reboot system boot 5.19.0-41-generi Sat Apr 29 10:05 - 10:06 (00:00)
reboot system boot 5.19.0-38-generi Sat Apr 8 22:11 - 10:03 (20+11:52)
reboot system boot 5.19.0-38-generi Sat Apr 8 22:01 - 22:02 (00:00)
reboot system boot 5.19.0-38-generi Sat Apr 8 21:59 - 22:01 (00:01)
reboot system boot 5.19.0-38-generi Wed Mar 22 20:56 - 21:56 (17+01:00)

I wasn't able to get any logs that might explain the crash, but I will keep you posted if this solved the initial issue I'm having.

If issues start happening again, I'll go back to testing bleeding edge amd drivers on kisak repo.

EDIT: now that I read tnxhawks comments, I might be experiencing similar crashes, will update on that regard as well, but my initial issue was just related to the AMD driver freezing GUI with the following message: kernel: [drm:amdgpu_dm_commit_planes [amdgpu]] *ERROR* Waiting for fences timed out!

[- (nicoforce)]

After testing for this last few days, I haven't seen a similar stutter/freeze or this message on the logs at all "kernel: [drm:amdgpu_dm_commit_planes [amdgpu]] *ERROR* Waiting for fences timed out!"

However, my experience is much much worse now, instead of the stutters I used to have, which would last around 10 seconds and would happen many times a day, now the system just crashes, can only be saved by REISUB, and I can't get any logs related to it.

Tested Kernels:
- 6.1.26
- 6.2.0-20-generic (Provided by ubuntu)
- 6.2.9
- 6.2.13
- 6.3.0

All of them crash, while using the monitors connected to the usb C hub.

I might have to open a new issue with this new behavior, will do after testing 6.3.1, 6.2.14 and 6.1.27.

[Vladimir Argaard (kommado)]

Hello,

From the comments I gather that this has not been reproduced yet and no valid workaround is found,
Let me give you some things on that direction and a work around that worked for me.

First I use Ubuntu 22.04, I haven't touched the kernel it's current version is: 6.2.0-36-generic

My GPU is this: 03:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Navi 21 [Radeon RX 6800/6800 XT / 6900 XT] (rev c3) (prog-if 00 [VGA controller])
 Subsystem: Micro-Star International Co., Ltd. [MSI] Navi 21 [Radeon RX 6800/6800 XT / 6900 XT]
 Flags: bus master, fast devsel, latency 0, IRQ 229
 Memory at 6000000000 (64-bit, prefetchable) [size=16G]
 Memory at 6400000000 (64-bit, prefetchable) [size=256M]
 I/O ports at 4000 [size=256]
 Memory at 50200000 (32-bit, non-prefetchable) [size=1M]
 Expansion ROM at 50300000 [disabled] [size=128K]
 Capabilities: [48] Vendor Specific Information: Len=08 <?>
 Capabilities: [50] Power Management version 3
 Capabilities: [64] Express Legacy Endpoint, MSI 00
 Capabilities: [a0] MSI: Enable+ Count=1/1 Maskable- 64bit+
 Capabilities: [100] Vendor Specific Information: ID=0001 Rev=1 Len=010 <?>
 Capabilities: [150] Advanced Error Reporting
 Capabilities: [200] Physical Resizable BAR
 Capabilities: [240] Power Budgeting <?>
 Capabilities: [270] Secondary PCI Express
 Capabilities: [2a0] Access Control Services
 Capabilities: [2d0] Process Address Space ID (PASID)
 Capabilities: [320] Latency Tolerance Reporting
 Capabilities: [410] Physical Layer 16.0 GT/s <?>
 Capabilities: [440] Lane Margining at the Receiver <?>
 Kernel driver in use: amdgpu
 Kernel modules: amdgpu

Also I currently use gnome with X.Org (due to other issues with Wayland), but the issue existed in Wayland as well.

Just in the case that some application is responsible for this, in my day to day work I use the following software:
- MS teams preview
- Jetbrains IDEs
- Google chrome & Firefox (most of the day both browsers are always open)
- Discord
- Steam

I usually do not turn off my computer, I never get it to suspend mode, I only lock it when I'm away from the screen.

I have found a pattern for this issue it starts happening 2-3 days after my last reboot (if needed I can try to check the uptime when this happens next time), for a month in summer I was shutting down the pc during the night due to power cuts for this month the bug disappeared for me (so a not elegant workaround for the issue that worked for me is to reboot daily)

I hope this info helps on hunting the bug or as a workaround.

[Vladimir Argaard (kommado)]

Forgot to mention that I switch between applications using Activities or just the Super key, If i force myself to use Alt+Tab I never get a freeze

[Vladimir Argaard (kommado)]

It seems that I found a way to work around the issue without regularly rebooting my pc

I was looking at the accessibility settings and found an option to disable animations on Gnome, after disabling the animations this issue disappeared for me.