I recently upgraded my PC with an RTX 5080 and updated my drivers from nvidia-dkms 565 to nvidia-open-dks 570, and since then, I’ve been experiencing complete system freezes. The only solution is to force a reboot.
This mainly happens when playing a video in fullscreen, though I’ve also had it occur while gaming. In one instance, the screen froze completely, but the audio kept playing. Before the full freeze, I also notice a weird refresh issue, my mouse moves but updates only every 5-10 seconds before everything locks up entirely.
Checking the kernel logs, I found an error occurring exactly at the time of the freeze:
“Idling display engine timed out” and then “Flip event timeout on head”
I suspect this might be related to the latest NVIDIA drivers. If that’s the case, is there any workaround to fix this issue? Having to force reboot my PC constantly is really frustrating.
finally someone has the same thing…
I understand it goes to a black screen when switching to full-screen viewing? And you press ESC, it will come back…
It has already happened to me twice that it freezes the moment I put a youtube video in full screen, but it has also happened to me after X minutes while already in full screen on a video or even during a game.
For me it freezes leaving the image, not putting a black screen and I can’t do anything anymore, I tried ESC or anything else.
Well, I come back if I exit full screen mode. There is no complete freeze.
But I have been writing about this for a long time, there is essentially no reaction. No one writes anymore. This happened with an older card (1070) but with new drivers. Now I have the same thing with another Nvidia card with new drivers.
Did you have the new 570 drivers on both cards when you had the problem ?
For me it is impossible to exit full screen mode because it is completely frozen. Are you sure it is the same problem? Do you have the same kernel errors?
Have you both tried creating new user logins, tried to reproduce this on other DEs/WMs/WCs, other kernels, other drivers (like nvidia vs nvidia-lts vs nvidia-5xx versions) BIOS updates, or even more drastic changes like other distros?
You could actually have a different experience by trying any/all of the above.
04:50 kate[529272]: This plugin supports grabbing the mouse only for popup windows
04:51 kate[529272]: This plugin supports grabbing the mouse only for popup windows
05:09 rtkit-daemon[1127]: Supervising 9 threads of 5 processes of 1 users.
05:09 rtkit-daemon[1127]: Supervising 9 threads of 5 processes of 1 users.
05:11 rtkit-daemon[1127]: Supervising 9 threads of 5 processes of 1 users.
05:11 rtkit-daemon[1127]: Supervising 9 threads of 5 processes of 1 users.
05:12 rtkit-daemon[1127]: Successfully made thread 532392 of process 531958 owned by '1000' RT at priority 10.
05:12 rtkit-daemon[1127]: Supervising 10 threads of 6 processes of 1 users.
05:41 kwin_wayland[1094]: kwin_core: Cannot grant a token to KWin::ClientConnection(0x5fc3ba448ac0)
05:41 kwin_wayland[1094]: kwin_core: Cannot grant a token to KWin::ClientConnection(0x5fc3ba448ac0)
05:42 kwin_wayland[1094]: kwin_core: Cannot grant a token to KWin::ClientConnection(0x5fc3ba448ac0)
05:42 kwin_wayland[1094]: kwin_core: Cannot grant a token to KWin::ClientConnection(0x5fc3ba448ac0)
05:42 kwin_wayland[1094]: kwin_core: Cannot grant a token to KWin::ClientConnection(0x5fc3ba448ac0)
05:42 kwin_wayland[1094]: kwin_core: Cannot grant a token to KWin::ClientConnection(0x5fc3ba448ac0)
05:43 kwin_wayland[1094]: kwin_core: Cannot grant a token to KWin::ClientConnection(0x5fc3ba448ac0)
05:43 kwin_wayland[1094]: kwin_core: Cannot grant a token to KWin::ClientConnection(0x5fc3ba448ac0)
06:23 systemd[1044]: Starting Spectacle screenshot capture utility...
░░ Subject: UNIT starts to run
░░ Defined-By: systemd
░░ Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
░░
░░ The process of starting the unit UNIT has begun.
06:24 systemd[1044]: Started Spectacle screenshot capture utility.
░░ Subject: Starting the unit UNIT has completed
░░ Defined-By: systemd
░░ Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
░░
░░ The process of starting the unit UNIT has completed.
░░
░░ Result: done.
06:24 spectacle[532447]: kpipewire_vaapi_logging: VAAPI: Mesa Gallium driver 24.3.4-arch1.1 for AMD Radeon Graphics (radeonsi, raphael_mendocino, LLVM 19.1.7, DRM 3.59, 6.13.2->
06:51 systemd[1044]: app-org.kde.spectacle.service: Consumed 626ms CPU time, 101.5M memory peak.
░░ Subject: Resources consumed by unit
░░ Defined-By: systemd
░░ Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
░░
░░ Unit UNIT has been terminated. Statistics on the resources it has consumed are provided resources.
07:14 rtkit-daemon[1127]: Supervising 9 threads of 5 processes of 1 users.
07:14 rtkit-daemon[1127]: Supervising 9 threads of 5 processes of 1 users.
07:17 rtkit-daemon[1127]: Supervising 9 threads of 5 processes of 1 users.
07:17 rtkit-daemon[1127]: Supervising 9 threads of 5 processes of 1 users.
07:17 rtkit-daemon[1127]: Successfully made thread 532621 of process 532076 owned by '1000' RT at priority 10.
07:17 rtkit-daemon[1127]: Supervising 10 threads of 6 processes of 1 users.
try to make “framed screenshots” or play video full screen in Haruna - for example. And then after the black screen press ESC - it should die.
The problem is that the screen goes to sleep sometimes. And it seems to you that the computer is frozen?
I think I can’t use an old driver because my graphics card is not supported. Since it’s brand new I think only the 570 version supports it.
Same for my BIOS because it’s a new PC and I had to update it. It’s on the latest version.
I couldn’t test on another distro, I don’t have any other installed on my PC and the problem is too random for me to use a live CD to try to get the bug.
I don’t have the problem on Windows, I don’t think it helps, except that we know that it’s a software problem and not a hardware problem.
I haven’t tried making a new account, or using another DE. Since the problem is quite random, I’m afraid I’ll have trouble reproducing it on a new account. For the DE, I don’t have any others installed but I can see about doing it (same for switching to x11, but I switched to wayland to avoid other problems so it bothers me).
You are probably correct about the BIOS and driver options, but everything else is something you can try. Up to you.
Live CD/USB was not an option I mentioned because it’s never a full and proper test like a fully installed system with the right and most up-to-date drivers would be.
The easiest of all the options is to make a new user account. The distro option is last resort.
I would have to find a way to trigger the problem in any case.To avoid having to spend 3 hours using my PC normally on a new account where I don’t have my data.
If anyone else has a more precise idea of the problem, I’m interested.
You won’t believe it, but somehow “miraculously” creating a new user solved the problem in the old user too.
The question remains open - for how long?
But it helped, there is no limit to the surprise.
but the BIOS memory error is present in the logs, if anything., etc.
BIOS-provided physical RAM map:
x67 kernel: BIOS-e820: [mem 0x0000000000000000-0x000000000009ffff] usable
konsole[3984]: kf.windowsystem: static int KX11Extras::currentDesktop() may only be used on X11
x67 konsole[3984]: kf.windowsystem: static void KX11Extras::setOnDesktop(WId, int) may only be used on X11
x67 konsole[3984]: QThreadStorage: Thread 0x5b12458e16a0 exited after QThreadStorage 9 destroyed
x67 systemd[2279]: app-org.kde.konsole-3984.scope: Consumed 1.276s CPU time, 39.3M memory peak.
x67 systemd[2279]: Started Konsole
x67 systemd[2279]: Started app-org.kde.konsole-4554.scope
x67 kwin_wayland[2326]: kf.config.core: Watching absolute paths is not supported "/home/eg/.local/share/color-schemes/Otto.colors"
x67 systemd[2279]: Starting Spectacle screenshot capture utility...
x67 systemd[2279]: Starting Spectacle screenshot capture utility...
x67 systemd[2279]: Started Spectacle screenshot capture utility.
x67 spectacle[4573]: kpipewire_vaapi_logging: VAAPI: Mesa Gallium driver 24.3.4-arch1.1 for AMD Radeon Graphics (radeonsi, raphael_mendocino, LLVM 19.1.7, DRM 3.59, 6.13.2-arch1-1) in use for device "/dev/dri/renderD1>
x67 spectacle[4573]: kf.notifications: Audio notification requested, but no sound name provided in notifyrc file, aborting audio notification
x67 systemd[2279]: app-org.kde.spectacle.service: Consumed 711ms CPU time, 95.7M memory peak.
Did you do anything specific to make the bug appear?For me it appears randomly so it’s complicated to test if such a solution works or not.If you have a way to trigger the bug in 100% of cases I would like to know.
I just tried, I spammed the fact of taking a screenshot and putting a youtube video in full screen then removing it and it loops more than a hundred times and I did not have the crash. As I said, mine is very random.
Are you sure we have the same problem?
Not really because you tell me that you can get out of the bug by pressing Esc.When I have the bug, it’s over , I can’t do anything anymore.The only way for me is to force restart the PC.
What did you mean by that? I didn’t understand…