-
Notifications
You must be signed in to change notification settings - Fork 135
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
System freezes randomly, Dbus Error pointing to lxqt-panel #2045
Comments
It seems something in Arch has caused strange problems (also see #2044; do you use KWin too?). When did it start to happen? After an Arch upgrade?
Unlikely. I have it here, in Manjaro, and have encountered no problem, although (as I mentioned in #2044) I use git LXQt with Qt6 components. |
I have seen that issue, it seems to me like a big coincidence that something similar has been described just 4 hours ago. I have done 2 system updates in the last 2 weeks, i think this could be something related to the previous one(I updated today but this happened before today too). Since i update at least 1 time per week, this should be cause by a recent change. Update: Could it be related to some kind of update on KDE's packages in the last days? I have seen there was some activity on kde's packages lately. |
That's possible in the case of #2044. But since you said that you didn't use KWin, I wonder how KDE updates could have affected you. |
The stranger thing is that Arch's upgrades come to Manjaro Testing after a few days at most, but I've had a stable LXQt+kwin_x11 session. Moreover, I upgraded to the Qt6-based git lxqt-panel only recently, and I had a fully stable session with the Qt5-based panel too. IMO, an important piece of info is missing. |
I have been reading the
Since this is a screen thing, it may be related. Now, looking carefully, it seems that the Something i have find in I have 2 coredumps of PS: I also remember that this happened after a system update: The first time this happened, it was right after the system update(i remember that the update was ongoing and suddenly i was not able to change tabs and cpu gone wild). This time, was like 3 or 4 hours after the system update. I know i have to restart the pc after an update, but this time i did not made the reboot right after the update. |
That file belongs to
I'm not saying that the problem you see is related to |
Now, I'm really suspicious of the recent changes in KDE Frameworks. In spite of including "5" in their Arch package names, something may have not been backward compatible. That may explain why I don't see any issue here: I don't have those packages with "5" in their names. If my suspicion is right, some components of LXQt 1.4.0 should be broken after the recent Arch upgrades. Whether there's a way of fixing that before LXQt 2.0.0, I have no idea. |
@yan12125, any thought on this? |
That backtrace for the 'kscreen_backend' segfault looks awfully similar to the one I mentioned here: |
In spite of similarities, I doubt that they're related. lxqt/lxqt-qtplugin#83 was about an old problem that happened when settings were applied while kglobalacceld was running. |
An update on the topic: Today i had a The coredump details were exactly the same as the one i shared 6 comments above. This time i was opening During the program applying the changes, more or less, the coredump happened. My last system upgrade was in 2024-03-28, yet nothing related to lxqt or Kde was upgraded.
Between other not related programs which were upgraded (i.e electron). Later attempts to reproduce the issue by changing colors in lxqt-config-appearance dont trigger the coredump on kscreen_backend. Another thing i have to add is that i have noticed that after startup, One of the
|
|
@Chiitoo mentioned the similarities of backtraces of kscreen_backend earlier. Although lxqt/lxqt-qtplugin#83 was about kglobalacceld, the fix was quite general. So, I recommend you upgrade your Qt6-based git lxqt-qtplugin to the latest git source. If you're familiar with making packages from sources, you might also want to upgrade your Qt5-based lxqt-qtplugin 1.4.0 to this branch: https://github.com/lxqt/lxqt-qtplugin/tree/fix_fontapply. See lxqt/lxqt#2546 for the main reason, but the Qt6-based fix is also included in that Qt5 branch. |
I have updated the For this, i used After successfully installing the package (prior renaming the folder obtained via the .zip) i rebooted. After reboot, i still see the PS: before doing this, i did a system upgrade, |
Therefore, if you see a BTW, we'll change the name of that branch from
You mean its Qt5 version belonging to lxqt-config 1.4.0? It depends on A backtrace? |
Yes, the 1.4.0 version which is qt5-based (the one present in arch repositories). The
|
Working on and testing the Qt6 port of LXQt, I don't have the Qt5-based lxqt-config or |
Yes, i will run the corresponding
I guess that if the problem is not present at the moment in the Qt6, the issue may be on the In the meanwhile, i will check for Thank you. |
Right, everything seems normal here. |
Well, a few days have passed, and there were no more The I was able to run Here is the relevant part:
I think this issue can be closed.
I will let you close the issue, in case you would like to see any more information about it. |
The crash in the Qt6-based kscreen_backend should have been fixed after lxqt/lxqt-qtplugin@b73fc19 (related @Chiitoo's report). Also, the Qt5-based kscreen_backend shouldn't crash after https://github.com/lxqt/lxqt-qtplugin/tree/1.4 (based on which, lxqt-qtplugin 1.4.1 will be released soon) for the same reason. |
Well, it seems like the crash still is present. I obtained 2 today (i have been playing with WMs and using The error or issue seems to point to the same thing as the last
Fortunately this did not crash the system but the problem is somehow around. |
If I were you, I'd check if Openbox has anything to do with this, by changing my WM (as in lxqt/lxqt-config#987). I'm not saying that Openbox is the culprit, but there have been cases where Openbox has caused serious troubles. Personally, I don't trust an app whose development stopped years ago to play a role as sensitive as that of a WM. |
An update into this: I have been using xfwm4 for a while to test the keyboard issue i have had. In the meanwhile, i have been following what happens with this issue. After i have installed the A few times i did not restarted the system for different reasons, and in some of them i noticed the following. Continuing with the
I have ran:
and the result seems to point to the font issue again (and that's why i add this long comment with the details):
And the
Here is the output of the coredump, it may contain more relevant information:
I know that LXQt 2.0 will arrive soon in the archlinux repositories and i will have to see what happens in there, but i considered important to show this before the switch to the new version. Information:
LXQt 1.4 (not using qt6 as far as i understand) I just realized that i have had |
If a problem happened after a system upgrade here, then I'd know that it should be about the system. If it disappeared after a reboot, I'd conclude that some basic daemons/processes needed to be restarted before having a stable system. |
Hi, this issue could be closed. In the followed time to this report, i have not experienced this anymore (and nobody else has experienced it either). I guess this was fixed somehow, in teh qt6 libkscreen version, or something was changes (indirectly) in the lxqt-panel configuration that made this solved. |
Can still be reopened. |
Expected Behavior
Nothing, normal system behavior.
Current Behavior
System panel and actions freezes and crash on random times(at the moment, this happened to me just 2 or 3 times).
The mouse is not able to click things but it conserves the movement.
Hotkeys seem to work, somehow, (qps shorcut key seems to be activated because the tray icon is loaded but qps does not show in the screen)
Time clock works( HH:MM:SS change)
CPU goes to 100% and i am forced to physically shutdown the PC.
Possible Solution
No idea.
Steps to Reproduce (for bugs)
I have no steps to reproduce it, seems to happen on random time.
Context
Some of the
journalctl
messages i have seen are:lxqt-panel[750]: void DBusMenuImporter::slotGetLayoutFinished(QDBusPendingCallWatcher*): "Remote peer disconnected"
lxqt-panel[750]: Error on DBus request(:1.445,/StatusNotifierItem): QDBusError(org.freedesktop.DBus.Error.NoReply, Remote peer disconnected)
lxqt-panel[750]: Error on DBus request(:1.445,/StatusNotifierItem): QDBusError(org.freedesktop.DBus.Error.UnknownObject, No such object path '/StatusNotifierItem')
lxqt-panel[750]: Error on DBus request(:1.38,/StatusNotifierItem): QDBusError(org.freedesktop.DBus.Error.UnknownProperty, Property org.kde.StatusNotifierItem.IconName was not found in object /StatusNotifierItem)
System Information
I have the full journactl log from where i have extracted those lines, i case more information is needed. Maybe this is related to the
dbus-broker
change made in Archlinux in January 9th ?PS:
I have the following AUR packages which may be involved:
The text was updated successfully, but these errors were encountered: