I recently replaced one of my monitors with a new high refresh rate monitor and now i have a issue with all the windows on the new monitor (primary) freezing randomly. The desktop on the new monitor (the high refresh rate monitor) will randomly freeze. No icons on the desktop can be selected and any open windows become unresponsive. The taskbar however remains alive and if i select one of the frozen windows from the taskbar, that window will then become temporarily responsive until i click out of it.
I have tried all combinations of refresh rate and Adaptive Sync on both monitors with no improvement. I never had any issues with the old monitor. The new monitor is an Alienware AW2725DM. Currently the settings on both are 2561 x 1440 resolution, 59.95 Hz refresh and adaptive sync set to never. I’m using the display port and have tried 2 different cables also without success.
This is beyond my ability to diagnose other than trying different combinations of display settings which I’ve done with not success.
You can try to see if it is by firing up a terminal immediately after a freeze happens (or SSH in from another device if the freeze is total) and run journalctl --user-unit plasma-kwin_wayland --boot 0 | tail -100 or sudo dmesg | grep -i pageflip; if you see something like “Pageflip timed out!” that’s a bug in the nvidia-drm kernel driver and would confirm you’re affected by it.
Bryan, “journalctl --user-unit plasma-kwin_wayland --boot 0 | tail -100” output below. I’m at a complete loss of what else to try. It only happens on the new monitor. I hope it’s not a hardware issue as my window to return is past by a few days.
> journalctl --user-unit plasma-kwin_wayland --boot 0 | tail -100
> Aug 13 01:32:55 fedora44 systemd\[4503\]: Starting plasma-kwin_wayland.service - KDE Wayland Compositor...
> Aug 13 01:32:55 fedora44 systemd\[4503\]: Started plasma-kwin_wayland.service - KDE Wayland Compositor.
> Aug 13 01:32:55 fedora44 kwin_wayland\[4791\]: No backend specified, automatically choosing drm
> Aug 13 01:32:56 fedora44 plasma-keyboard\[4849\]: qtvirtualkeyboard currently is not supported at client-side, use QT_IM_MODULES=qtvirtualkeyboard at compositor
> -side.
> Aug 13 01:32:56 fedora44 plasma-keyboard\[4849\]: Member contentWidth of the object PopupList_QMLTYPE_69 overrides a member of the base object. Consider renamin
> g it or adding final or override specifier
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4865\]: The XKEYBOARD keymap compiler (xkbcomp) reports:
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4865\]: > Warning: Multiple symbols for level 1/group 1 on key <FK23>
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4865\]: > Using F23, ignoring XF86TouchpadOff
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4865\]: > Warning: Symbol map for key <FK23> redefined
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4865\]: > Using last definition for conflicting fields
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4865\]: > Warning: Symbol map for key <FK24> redefined
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4865\]: > Using last definition for conflicting fields
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4865\]: Errors from xkbcomp are not fatal to the X server
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4874\]: The XKEYBOARD keymap compiler (xkbcomp) reports:
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4874\]: > Warning: Unsupported maximum keycode 709, clipping.
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4874\]: > X11 cannot support keycodes above 255.
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4874\]: > Warning: Virtual modifier Hyper multiply defined
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4874\]: > Using 0, ignoring 0
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4874\]: > Warning: Virtual modifier ScrollLock multiply defined
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4874\]: > Using 0, ignoring 0
> Aug 13 01:32:56 fedora44 kwin_wayland_wrapper\[4874\]: Errors from xkbcomp are not fatal to the X server
> Aug 13 01:33:01 fedora44 plasma-keyboard\[4849\]: Failed to register with host portal QDBusError("org.freedesktop.portal.Error.Failed", "Could not register app
> ID: Connection already associated with an application ID")
> Aug 13 11:17:58 fedora44 kwin_wayland\[4791\]: QDBusConnection: couldn't handle call to Teardown, no slot matched
> Aug 13 11:17:58 fedora44 kwin_wayland\[4791\]: QDBusConnection: couldn't handle call to Teardown, no slot matched
> Aug 13 11:17:58 fedora44 kwin_wayland\[4791\]: Could not find slot Krunner1Adaptor::Teardown
> Aug 13 11:45:54 fedora44 kwin_wayland\[4791\]: qrc:/qt/qml/org/kde/kwin/overview/qml/Main.qml:851:5: QML Connections: Implicitly defined onFoo properties in Con
> nections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
> Aug 13 11:45:54 fedora44 kwin_wayland\[4791\]: qrc:/qt/qml/org/kde/kwin/overview/qml/Main.qml:851:5: QML Connections: Implicitly defined onFoo properties in Con
> nections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }
> Aug 13 11:45:56 fedora44 kwin_wayland\[4791\]: Could not delete render time query because no context is current
> Aug 13 11:45:56 fedora44 kwin_wayland\[4791\]: Could not delete render time query because no context is current
> Aug 13 11:45:56 fedora44 kwin_wayland\[4791\]: Could not delete render time query because no context is current
> Aug 13 11:45:56 fedora44 kwin_wayland\[4791\]: Could not delete render time query because no context is current
> Aug 13 15:12:07 fedora44 kwin_wayland\[4791\]: Could not load window switcher package "org.kde.breeze.desktop" . Falling back to default
> Aug 13 15:12:07 fedora44 kwin_wayland\[4791\]: Member visible of the object PlasmaQuick::Dialog overrides a member of the base object. Consider renaming it or a
> dding final or override specifier
> Aug 13 15:12:07 fedora44 kwin_wayland\[4791\]: Failed to register with host portal QDBusError("org.freedesktop.portal.Error.Failed", "Could not register app ID:
> Unable to open /proc/4791/root")
Hmm… I don’t see anything about pageflip, so maybe that’s not what’s happening here. There are a bunch of invasive things you could try, like switching drivers or disabling GSP firmware offloading, but I’d suggest one last “low cost” one: in the monitor’s own on-screen menu, check for and disable anything like “Panel Self Refresh (PSR),” “Eco/power-saving display mode,” or “DP power management.” I’ve seen some posts that think those could be linked to pageflip stalls on some monitors independent of the OS-side settings.
I can’t say that the output was “immediately” from after the freeze. I was working and stepped away for a few minutes and when i came back it was frozen. I then ran the command. I haven’t had it happen since I started this thread while i was sitting at the computer. I’ll continue to monitor and try to catch it quicker next time.
As far as the monitor’s menu goes i don’t see anything that resembles what you describe.
Not really sure if our issue is really related, but for me, it freezes randomly, but happens more when doing heavy r/w on disks. For me nothing work after system freeze, had to force-restart the machine. I cannot reproduce this on kernel 6.19.10
$ uname -ar
Linux fedodesk 6.19.10-300.fc44.x86_64 #1 SMP PREEMPT_DYNAMIC Wed Mar 25 18:23:49 UTC 2026 x86_64 GNU/Linux
Your issue seems distinct from this one, @yuemiyuki, so it would be best for you to open your own thread; that way, it can catch the eyes of those looking into possible BTRFS hangs.
One potential fix would be to use NVIDIA’s open-source kernel modules, if you’re willing, @tired-curmudgeon
That would involve confirming you’re on RPM Fusion’s akmod-nvidia by running rpm -qa | grep -i nvidia, where you should see packages like akmod-nvidia, kmod-nvidia-<kernel-version>, xorg-x11-drv-nvidia, xorg-x11-drv-nvidia-cuda, etc. If you installed nvidia-driver or nvidia-open from a different repo, like negativo17, or used NVIDIA’s .run installer, that would change the way you’d move forward.
NVIDIA’s open kernel module lives under RPM Fusion’s nonfree-tainted repo, so you’d need to run sudo dnf install rpmfusion-nonfree-release-tainted. Then, just swap the driver package with sudo dnf swap akmod-nvidia akmod-nvidia-open. DNF will do the heavy lifting and remove akmod-nvidia and pull in akmod-nvidia-open along with its dependencies. It won’t touch xorg-x11-drv-nvidia or the CUDA packages.
After the new module builds, you need to rebuild initramfs and reboot; i.e., sudo dracut -f --regenerate-all and then reboot normally. You can verify everything took after the reboot with modinfo -F license nvidia and that should output “Dual MIT/GPL” because the proprietary one reports “NVIDIA”. You can also run nvidia-smi to confirm the GPU is detected and driver version looks right.
The akmod-nvidia-open package compiles the modules 100% from source, while akmod-nvidia use precompiled binary blobs like nvidia’s run installer and compile only a subset of sources.
Also had another freeze last night (again I didn’t notice it immediately as i was working in the other monitor and then got up to get coffee) so didn’t know exactly when the freeze happened. So within a few minutes of getting back to my desk i checked with. The line at Aug 13 20:47:01 got me thinking of another thread i read where a user stated he would get frozen windows when he raised and lowered the screen brightness. So…
> journalctl --since "2026-08-13 20:45:00" --until "2026-08-13 20:50:00"
> Aug 13 20:46:47 fedora44 cupsd\[1236\]: REQUEST localhost - - "POST / HTTP/1.1" 200 197 Renew-Subscription successful-ok
> Aug 13 20:46:49 fedora44 cupsd\[1236\]: REQUEST localhost - - "POST / HTTP/1.1" 200 197 Renew-Subscription successful-ok
> Aug 13 20:46:56 fedora44 kwin_wayland\[3600\]: **qrc:/qt/qml/org/kde/kwin/overview/qml/Main.qml:851:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }**
> Aug 13 20:46:56 fedora44 kwin_wayland\[3600\]: **qrc:/qt/qml/org/kde/kwin/overview/qml/Main.qml:851:5: QML Connections: Implicitly defined onFoo properties in Connections are deprecated. Use this syntax instead: function onFoo(<arguments>) { ... }**
> Aug 13 20:47:01 fedora44 org_kde_powerdevil\[3754\]: **\[DDCutilDisplay\]: ddca_get_non_table_vcp_value -3007**
> Aug 13 20:47:02 fedora44 kwin_wayland\[3600\]: **Could not delete render time query because no context is current**
> Aug 13 20:47:02 fedora44 kwin_wayland\[3600\]: **Could not delete render time query because no context is current**
> Aug 13 20:47:02 fedora44 kwin_wayland\[3600\]: **Could not delete render time query because no context is current**
> Aug 13 20:47:02 fedora44 kwin_wayland\[3600\]: **Could not delete render time query because no context is current**
...
Update: I disabled DDC/CI control on both monitors and haven’t had a freeze since. Not definitive because i haven’t used the system enough to call a win here. But, I used the system pretty heavily for a couple hours last night and no freezes. So we’ll see what happens today. The problem was happening quite often so if i get through today i"m going to call it solved. Hope I’m not being too optimistic but…we’ll see .
Update II: Disabling DDC/CI in settings and on the monitor seems to have solved it for me. It’s been over 24 hours. No freezes. I don’t need to manage monitor brightness from the desktop so a minor inconvenience.
Update III: Open windows on my primary monitor just froze. I thought i was home free but apparently not. I give up i guess. Since this is a situation i can’t live with i guess it’s time to reevaluate.
No momentum. Froze again after hours of no freeze.
I hate to say it but no matter what distro I’ve tried over the years and no matter how stable the system starts out it eventually borks itself with these endless updates and inablility to handle new hardware changes.
I’m afraid linux is never going to be a viable OS for the general public who doesn’t want to spend half their time troubleshooting with no help from the designers and only community support which unfortunately resembles more of a fishing expedition rather than anything.
Thanks for your time Bryan but after almost a year on fedora i’m going backt to Windows. I’ve got work to do and linux feels like something some kid made in his basement.
I can understand those frustrations… When I started my Linux journey on Ubuntu’s Warty Warthog after attending a tech conference and discovering it, every update seemed to break, well, everything. My experience now is much different and I find Windows, which I have to use at work, significantly less stable than Fedora; however, everyone has a different experience.
I think you’re right to blame “designers,” if you mean those who work for the hardware manufacturers who don’t seem to care a lick about Linux most of the time. I have steered well-clear of NVIDIA products specifically because of the issues I’ve read about concerning their hardware and Linux. Trust me, plenty of people insist they’ve never had a single issue with NVIDIA stuff, but I’ve read posts and threads by enough people that I have just stayed away.
Mark’s point about updating the BIOS is a good one, even if you’re thinking of heading back to Windows.
That last bit of journal you shared after the one hang is definitely different than the pageflip issue that could have been affecting you; in fact, this line
You can still take some steps to try and fix this, though…
You should be able to disable DDC/CI brightness control in “Power Management” settings; that is, open System Settings > Power Management > General (or search “brightness” in System Settings) and look for a toggle named something like “Also control brightness of external screens via DDC/CI” and turn it off. If that toggle doesn’t exist or doesn’t help, you could just remove ddcutilIf with sudo dnf remove ddcutil, which would remove external-monitor brightness control from the KDE brightness slider in exchange for stability.
It might also be worth it to check for a PowerDevil update with a fix: rpm -q plasma-workspace to see your current version because there is a pending KDE merge request upstream that might have already landed.
yeah DDC/CI was disabled already two days ago. I turned it off in system settings as well as in the monitor’s OSD. That’s why i initially thought i had it fixed because after doing so, i went more than a day without a freeze. When i tried to login today from the lock screen the desktop on the new monitor was frozen and i had to reboot. thinking we had it solved and then getting the freeze again is what pretty much pushed me over the edge.
Then I spent a few minutes in Winblows. That and my OCD is what brought me back to trying to fix it. So i did a bios update. Then spent an hour or so getting secure boot keys fixed for akmods and here i am. All set and ready to be tortured again. I’ll let you know if i ever get it sorted out. I’m half ready to reinstall or try yet another distro. if you have any new ideas let me know. in the mean time i need a glass of bourbon.
why should DDC/CI trigger a freeze?
Maybe it’s some plasma-login-manager sheenanigan.
Have you applied current desktop settings to the login screen? It’s available – or hidden – somewhere in the convoluted mess of KDE settings
Have you tried switching back to SDDM?
I wonder if you can reproduce this problem in another DE like Gnome or Cinnamon.
The freezes traced back to DDCutilitydisplay thowing an error -3007. Which if i understood what i looked up meant that communication failed and the maximum number of retries was exceeded. It may be in my post above.
Also.
Another user claimed to be able cause the freeze by changing the brightness up or down quickly on his system. Which also points to DDC/CI so disabling a function i don’t need or care about was an easy try and still may be involved.
Thanks. That will be my next step. For now, i changed from Display Port to HDMI. It’s been 16 hours and no freezes. If this doesn’t work out I’ll switch DE. Sorry if some of my troubleshooting seems ill conceived…I’m just throwing darts hoping one will hit the target as i’m way out of my league here.
Mark is probably right but it didn’t take any effort to try disabling DDC and I’m not giving up any real functionality that i use or need.
My Solution: It’s been 2-1/2 days since switching form Display Port to HDMI. Had to give up some refresh rate, but haven’t had any freezes since the change. the only other odd thing is that since moving to HDMI, when i restart or fresh boot, the login screen on both montors flickers until i click on it with my mouse. This does not happen if i lock the screen or log out of my session, only after a boot. So while not solved in the sense i figured out what was causing the freezes, this work around has settled it for me.
Corrected Solution: It’s been 2-1/2 days since switching form Display Port to HDMI AND I also had to set “Turnoff screen” to Never in Power Management. Had to give up some refresh rate, but haven’t had any freezes since the change. the only other odd thing is that since moving to HDMI, when i restart or fresh boot, the login screen on both montors flickers until i click on it with my mouse. This does not happen if i lock the screen or log out of my session, only after a boot. So while not solved in the sense i figured out what was causing the freezes, this work around has settled it for me.
