[SOLVED] Graphical issues after resume from suspend under Plasma 6

brachiosnorus · 2024-03-09 13:04:05

I'm having issues with a few applications after resuming from suspend under Plasma 6.

Firefox: always crashes on resume. I always get the Firefox crash handler upon resuming. I tried disabling hardware acceleration but it had no effect.

Yakuake: after resume the window is not positioned correctly. Instead of the window being positioned at the top left-hand corner, there is a large space on the top and left. Otherwise it works fine.

Kmail: the list of emails shows correctly but clicking on an individual email results in a black panel where the email contents should be shown.

All three of these are resolved by restarting the applications.

System information:
- Lenovo ThinkBook 16p Gen 2
- Wayland
- AMD Radeon Graphics (radeonsi, renoir, LLVM 17.0.6, DRM 3.57, 6.7.9-arch1-1) but there is also a NVIDIA card onboard, which is not used unless I explicitly start applications with the right command-line parameters. I confirmed that both KDE and Plasma are running with the Radeon graphics and not NVIDIA.
- Arch is fully updated and rebooted

Last edited by brachiosnorus (2024-05-27 05:17:26)

M440 · 2024-03-09 15:32:28

if you have nvidia gpu in your system and load drivers for it i would try to set the correct suspend options for it anyway

https://bbs.archlinux.org/viewtopic.php?id=293500

Last edited by M440 (2024-03-09 15:32:52)

brachiosnorus · 2024-03-09 16:02:47

Thanks for the reply.

I should have mentioned that I already had nvidia_drm.modeset=1 set as a boot parameter. However, I did not have nvidia.NVreg_PreserveVideoMemoryAllocations=1 set, but adding that now prevents suspend-to-RAM altogether.

[  395.962620] NVRM: GPU 0000:01:00.0: PreserveVideoMemoryAllocations module parameter is set. System Power Management attempted without driver procfs suspend interface. Please refer to the 'Configuring Power Management Support' section in the driver README.
[  395.962628] nvidia 0000:01:00.0: PM: pci_pm_suspend(): nv_pmops_suspend+0x0/0x30 [nvidia] returns -5
[  395.962941] nvidia 0000:01:00.0: PM: dpm_run_callback(): pci_pm_suspend+0x0/0x170 returns -5
[  395.962947] nvidia 0000:01:00.0: PM: failed to suspend async: error -5
[  396.684789] PM: Some devices failed to suspend, or early wake event detected

Last edited by brachiosnorus (2024-03-09 17:15:41)

brachiosnorus · 2024-03-09 18:22:00

Looking at logs closer, it seems like Wayland is crashing on resume and that's the underlying cause of the issues. And I'm also seeing that this crash only happens when I am using my external monitor; if I am using the laptop screen, nothing goes wrong at all.

Relevant part of logs that shows up only when the monitor is attached:

Mar 09 19:52:36 myhostname kwin_wayland[3598]: kwin_core: Applying output config failed!
Mar 09 19:52:36 myhostname kernel: traps: kwin_wayland[3598] general protection fault ip:7f70acb90a5c sp:7ffd35e9baf0 error:0 in libQt6Core.so.6.6.2[7f70aca88000+35b000]
Mar 09 19:52:36 myhostname systemd[1]: Started Process Core Dump (PID 6097/UID 0).
Mar 09 19:52:37 myhostname systemd-coredump[6098]: [?] Process 3598 (kwin_wayland) of user 1000 dumped core.
                                                    
                                                    Stack trace of thread 3598:
                                                    #0  0x00007f70acb90a5c n/a (libQt6Core.so.6 + 0x190a5c)
                                                    #1  0x00007f70af3e8807 n/a (libkwin.so.6 + 0x3e8807)
                                                    #2  0x00007f70acb90ca9 n/a (libQt6Core.so.6 + 0x190ca9)
                                                    #3  0x00007f70af17ede7 _ZN4KWin7Session13deviceResumedEm (libkwin.so.6 + 0x17ede7)
                                                    #4  0x00007f70af18b9cb n/a (libkwin.so.6 + 0x18b9cb)
                                                    #5  0x00007f70af18bc01 n/a (libkwin.so.6 + 0x18bc01)
                                                    #6  0x00007f70ae1130ce n/a (libQt6DBus.so.6 + 0x390ce)
                                                    #7  0x00007f70acb7c2c7 _ZN7QObject5eventEP6QEvent (libQt6Core.so.6 + 0x17c2c7)
                                                    #8  0x00007f70adaf438b _ZN19QApplicationPrivate13notify_helperEP7QObjectP6QEvent (libQt6Widgets.so.6 + 0xf438b)
                                                    #9  0x00007f70acb39818 _ZN16QCoreApplication15notifyInternal2EP7QObjectP6QEvent (libQt6Core.so.6 + 0x139818)
                                                    #10 0x00007f70acb39b9b _ZN23QCoreApplicationPrivate16sendPostedEventsEP7QObjectiP11QThreadData (libQt6Core.so.6 + 0x139b9b)
                                                    #11 0x00007f70acc9f18f _ZN20QEventDispatcherUNIX13processEventsE6QFlagsIN10QEventLoop17ProcessEventsFlagEE (libQt6Core.so.6 + 0x29f18f)
                                                    #12 0x00007f70ad5b26e2 _ZN23QUnixEventDispatcherQPA13processEventsE6QFlagsIN10QEventLoop17ProcessEventsFlagEE (libQt6Gui.so.6 + 0x5b26e2)
                                                    #13 0x00007f70acb43d6e _ZN10QEventLoop4execE6QFlagsINS_17ProcessEventsFlagEE (libQt6Core.so.6 + 0x143d6e)
                                                    #14 0x00007f70acb3c2b8 _ZN16QCoreApplication4execEv (libQt6Core.so.6 + 0x13c2b8)
                                                    #15 0x000063862c38cdee n/a (kwin_wayland + 0x3cdee)
                                                    #16 0x00007f70ac443cd0 n/a (libc.so.6 + 0x25cd0)
                                                    #17 0x00007f70ac443d8a __libc_start_main (libc.so.6 + 0x25d8a)
                                                    #18 0x000063862c393425 n/a (kwin_wayland + 0x43425)

Last edited by brachiosnorus (2024-03-09 19:12:01)

stueja · 2024-03-14 07:06:10

Did you

sudo systemctl enable nvidia-suspend.service
sudo systemctl enable nvidia-hibernate.service
sudo systemctl enable nvidia-resume.service

alongside with adding

nvidia.NVreg_PreserveVideoMemoryAllocations=1

according to https://download.nvidia.com/XFree86/Lin … ement.html ?

brachiosnorus · 2024-03-14 07:27:16

Thanks. Things seemed to have improved with moving to X from Wayland, recent package updates and enabling those services (unclear which contributed what).

My system has been extremely unstable since nvidia-550 was released so I'm tentatively going to mark this as solved but there have been so many problems I will have to keep an eye on things.

Edit: Well Wayland seems like it's off the table; I can use the desktop for 10 seconds before I get a lock up and have to hard reboot (not sysrq, power button).

Last edited by brachiosnorus (2024-03-14 07:34:29)

brachiosnorus · 2024-03-16 15:58:07

I'm still getting all the same crashes on resume-from-suspend.

I've even tried downgrading to nvidia-545 but the system doesn't seem to be any more stable.

seth · 2024-03-16 17:24:29

The only crash you posted is from kwin_wayland.
Given

Wayland seems like it's off the table; I can use the desktop for 10 seconds before I get a lock up and have to hard reboot

post an updated and complete system journal that covers the crashes after the S3 ended.

brachiosnorus · 2024-03-18 06:40:45

Yes, I marked as solved because the initial reported issue is solvable by reverting to X11.

However, I encountered ongoing (but different!) instability, so I resumed addressing the original wayland issue by downgrading to nvidia-545 in an effort to get a working system, but I am back in the original state so I no longer consider this solved for me.

Here are the logs showing suspend and then resume: http://pastie.org/p/244oLGC8RKyve6VR4JdB2o

Edit: maybe it's a Radeon issue since there are references to radeonsi_dri.so in the stacktrace?

Last edited by brachiosnorus (2024-03-18 07:11:55)

seth · 2024-03-18 09:13:26

Backtrace

Mar 18 07:25:21 systemd-coredump[75261]: Process 72091 (kwin_wayland) of user 1000 dumped core.
                                         
                                         Stack trace of thread 72091:
                                         #0  0x000070c294f7b8d6 _ZNK14QObjectPrivate20maybeSignalConnectedEj (libQt6Core.so.6 + 0x17b8d6)
                                         #1  0x000070c294f90b00 n/a (libQt6Core.so.6 + 0x190b00)
                                         #2  0x000070c2979e8c07 n/a (libkwin.so.6 + 0x3e8c07)
                                         #3  0x000070c294f90ca9 n/a (libQt6Core.so.6 + 0x190ca9)
                                         #4  0x000070c29777edd7 _ZN4KWin7Session13deviceResumedEm (libkwin.so.6 + 0x17edd7)
                                         #5  0x000070c29778b9bb n/a (libkwin.so.6 + 0x18b9bb)
                                         #6  0x000070c29778bbf1 n/a (libkwin.so.6 + 0x18bbf1)
                                         #7  0x000070c295ded0ce n/a (libQt6DBus.so.6 + 0x390ce)
                                         #8  0x000070c294f7c2c7 _ZN7QObject5eventEP6QEvent (libQt6Core.so.6 + 0x17c2c7)
                                         #9  0x000070c2960f438b _ZN19QApplicationPrivate13notify_helperEP7QObjectP6QEvent (libQt6Widgets.so.6 + 0xf438b)
                                         #10 0x000070c294f39818 _ZN16QCoreApplication15notifyInternal2EP7QObjectP6QEvent (libQt6Core.so.6 + 0x139818)
                                         #11 0x000070c294f39b9b _ZN23QCoreApplicationPrivate16sendPostedEventsEP7QObjectiP11QThreadData (libQt6Core.so.6 + 0x139b9b)
                                         #12 0x000070c29509f18f _ZN20QEventDispatcherUNIX13processEventsE6QFlagsIN10QEventLoop17ProcessEventsFlagEE (libQt6Core.so.6 + 0x29f18f)
                                         #13 0x000070c2959b26e2 _ZN23QUnixEventDispatcherQPA13processEventsE6QFlagsIN10QEventLoop17ProcessEventsFlagEE (libQt6Gui.so.6 + 0x5b26e2)
                                         #14 0x000070c294f43d6e _ZN10QEventLoop4execE6QFlagsINS_17ProcessEventsFlagEE (libQt6Core.so.6 + 0x143d6e)
                                         #15 0x000070c294f3c2b8 _ZN16QCoreApplication4execEv (libQt6Core.so.6 + 0x13c2b8)
                                         #16 0x00006147a0086dee n/a (kwin_wayland + 0x3cdee)
                                         #17 0x000070c294843cd0 n/a (libc.so.6 + 0x25cd0)
                                         #18 0x000070c294843d8a __libc_start_main (libc.so.6 + 0x25d8a)
                                         #19 0x00006147a008d425 n/a (kwin_wayland + 0x43425)

smells like https://bugs.kde.org/show_bug.cgi?id=483008 which would be fixed in https://invent.kde.org/plasma/kwin/-/me … uests/5433

brachiosnorus · 2024-03-18 09:51:15

Great find, thanks Seth!

brachiosnorus · 2024-05-27 05:15:49

I ended up rebuilding kwin with this patch, and it finally fixed the issue! The patch will be available in 6.1.0, but cherry-picking that one commit works on 6.0.5.

Last edited by brachiosnorus (2024-05-28 14:37:38)

Arch Linux

#1 2024-03-09 13:04:05

[SOLVED] Graphical issues after resume from suspend under Plasma 6

#2 2024-03-09 15:32:28

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#3 2024-03-09 16:02:47

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#4 2024-03-09 18:22:00

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#5 2024-03-14 07:06:10

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#6 2024-03-14 07:27:16

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#7 2024-03-16 15:58:07

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#8 2024-03-16 17:24:29

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#9 2024-03-18 06:40:45

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#10 2024-03-18 09:13:26

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#11 2024-03-18 09:51:15

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

#12 2024-05-27 05:15:49

Re: [SOLVED] Graphical issues after resume from suspend under Plasma 6

Board footer