Hi

When I log out of a KDE session, instead of taking me to the logon screen it drops out to a command prompt. I can log in there but cannot start any desktop. the only option is to type reboot and start again.

This happens maybe once in 10 logouts. It never fails if I switch user.

This is my setup

Operating System: Kubuntu 21.10
KDE Plasma Version: 5.22.5
KDE Frameworks Version: 5.86.0
Qt Version: 5.15.2
Kernel Version: 5.13.0-35-generic (64-bit)
Graphics Platform: Wayland
Processors: 12 × Intel® Core™ i5-10400 CPU @ 2.90GHz
Memory: 7.5 GiB of RAM
Graphics Processor: Mesa Intel® UHD Graphics 630

Has anyone else seen this or has anyone got any suggestions on diagnosing/fixing?

thanks

Yeah, I'm also getting this issue

I'm currently running:
OS: Fedora Linux 35 (KDE Plasma) x86_64
Host: Laptop AB
Kernel: 5.16.15-201.fc35.x86_64
Uptime: 35 mins
Packages: 2669 (rpm), 16 (flatpak)
Shell: bash 5.1.8
Resolution: 2256x1504
DE: Plasma 5.24.3
WM: KWin
Theme: Breeze Dark [Plasma], Adwaita [GTK2]
Icons: [Plasma], breeze-dark [GTK2/3]
Terminal: konsole
CPU: 11th Gen Intel i7-1165G7 (8) @ 4.700GHz
GPU: Intel TigerLake-LP GT2 [Iris Xe Graphics]
Memory: 5671MiB / 15775MiB

I also notice this issue only happens to the account I created when I first set up my computer, since another account that I created on my computer doesn't have this issue

What do the logfiles contain after the crash?

Code: Select all

journalctl -xep err

(Shows all Errors, beginning at the bottom. Scroll up to read more)

koffeinfriedhof wrote:What do the logfiles contain after the crash?

Code: Select all

journalctl -xep err

(Shows all Errors, beginning at the bottom. Scroll up to read more)

I'l l look next time it does it...

Both accounts on my computer are ones that I created in the LTS version of Kubuntu 2 years ago. Haven't tried creating new accounts

koffeinfriedhof wrote:What do the logfiles contain after the crash?

Code: Select all

journalctl -xep err

(Shows all Errors, beginning at the bottom. Scroll up to read more)

I'll try this out too later today

So I let it crash again and ran the command

Code: Select all

journalctl -xep err

and got the following:

Code: Select all

-- Boot b3570ece3944408e931e763b4ee33d34 --
Mar 22 12:38:58 dell-chromebook-3100 kded5[9216]: org.kde.plasma.dataengine.geolocation: error:  "Unknown host location.services.mozilla.com: Host not found"
Mar 22 12:38:58 dell-chromebook-3100 kded5[9216]: org.kde.plasma.dataengine.geolocation: error:  "Unknown host location.services.mozilla.com: Host not found"
Mar 22 12:39:05 dell-chromebook-3100 pipewire-pulse[10371]: pw.conf: execvp error 'pactl': No such file or directory
Mar 22 12:39:59 dell-chromebook-3100 systemd[9008]: Failed to start Kite Updater.
░░ Subject: A start job for unit UNIT has failed
░░ Defined-By: systemd
░░ Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
░░
░░ A start job for unit UNIT has finished with a failure.
░░
░░ The job identifier is 531 and the job result is failed.
-- Boot 3480f507b81c4d679a08fcb003c6906c --
Mar 22 13:15:23 dell-chromebook-3100 kded5[1958]: org.kde.plasma.dataengine.geolocation: error:  "Unknown host location.services.mozilla.com: Host not found"
Mar 22 13:15:23 dell-chromebook-3100 kded5[1958]: org.kde.plasma.dataengine.geolocation: error:  "Unknown host location.services.mozilla.com: Host not found"
Mar 22 13:15:31 dell-chromebook-3100 pipewire-pulse[3097]: pw.conf: execvp error 'pactl': No such file or directory

This is everything between what I assume is for the last two boots

Did you enter the command directly after the crash or did you reboot first? There is no hint for a crash. If you reboot you must add -b -1 to this command to get the log from previous boot. See man journalctl for details.

koffeinfriedhof wrote:Did you enter the command directly after the crash or did you reboot first? There is no hint for a crash. If you reboot you must add -b -1 to this command to get the log from previous boot. See man journalctl for details.

Thanks for the clarification; I ran

Code: Select all

journalctl -xep err -b -1

with admin privileges and got

Code: Select all

$ journalctl -xep err -b -1
░░
░░ The job identifier is 2983 and the job result is failed.
Mar 23 10:58:30 dell-chromebook-3100 systemd[1751]: Failed to start Kite Updater.
░░ Subject: A start job for unit UNIT has failed
░░ Defined-By: systemd
░░ Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
░░
░░ A start job for unit UNIT has finished with a failure.
░░
░░ The job identifier is 3008 and the job result is failed.
Mar 23 12:16:03 dell-chromebook-3100 systemd[1751]: Failed to start Kite Updater.
░░ Subject: A start job for unit UNIT has failed
░░ Defined-By: systemd
░░ Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
░░
░░ A start job for unit UNIT has finished with a failure.
░░
░░ The job identifier is 3090 and the job result is failed.
Mar 23 12:31:06 dell-chromebook-3100 systemd[1751]: Failed to start Kite Updater.
░░ Subject: A start job for unit UNIT has failed
░░ Defined-By: systemd
░░ Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
░░
░░ A start job for unit UNIT has finished with a failure.
░░
░░ The job identifier is 3250 and the job result is failed.
Mar 23 12:40:43 dell-chromebook-3100 cupsd[1595]: [Client 9] Returning IPP client-error-bad-request for Create-Printer-Subscriptions (/) from localhost.
Mar 23 13:05:01 dell-chromebook-3100 systemd[1751]: Failed to start Kite Updater.
░░ Subject: A start job for unit UNIT has failed
░░ Defined-By: systemd
░░ Support: https://lists.freedesktop.org/mailman/listinfo/systemd-devel
░░
░░ A start job for unit UNIT has finished with a failure.
░░
░░ The job identifier is 3356 and the job result is failed.
Mar 23 13:15:28 dell-chromebook-3100 systemd-cryptsetup[48364]: Device luks-f103b474-6760-46ad-b29d-050ffc0399f0 is still in use.
Mar 23 13:15:28 dell-chromebook-3100 systemd-cryptsetup[48364]: Failed to deactivate: Device or resource busy
Mar 23 13:15:28 dell-chromebook-3100 kernel: watchdog: watchdog0: watchdog did not stop!


Okay. I expected to see the plasmashell or sddm crashing. You could follow the logfile on a virtual terminal (e.g. ctrl-alt-f3) with `jjournalctl -f` and see what happens while logging out of Plasma.

Do you get any messages when dropped on the shell as user? Does sddm still run (ctrl-alt-f1/f7)?

The first error with cups is some printer-stuff, like printer not reachable or similar. I would at first ignore it.
The problems with luks only give a hint to an encrypted device not being closed/unmounted which could be a hint on problems during poweroff. Probably only a timing issue which solves itself later in the log without throwing further errors.
Kite doesn't mean anything to me.

You can either view the full log or adding the warnings to get an idea what happens:

Code: Select all

journalctl -xep warning

will show all warnings, errors, critical and so on.

You could also check for system updates/upgrades via `apt update && apt full-upgrade && apt autoremove` as root.

Hi
I raised this topic on the 10th of March and since posting this it hasn't crashed.

One thing that's different that may be implicated, is that for years I've been running Boinc distributing computing in the background and it does occasionally use lots of memory or processing (or both).

Around about the time I posted this question I suspended Boinc so it's not running.

Also, if a session is using a lot of swap space logoffs are very slow. Related?

As I said, this may or may not have a bearing on this issue so I'm making a point of logging in and out rather than switching users to see what happens

I'll keep you posted.

It looks like TheGamer82060 is still having a lot of issues with this.

Samo

samo wrote:...if a session is using a lot of swap space logoffs are very slow. Related?...

Perhaps this is related. New allocated memory on starting new processes needs to be freed in RAM which results in swapping again . Processes being swapped must return to RAM to process. That is slow. If the available memory tends to zero, the Kernel starts killing processes to avoid a system crash.