SUMMARY Since updating KDE Frameworks from 5.99.0 to 5.102.0 and after rebooting, my panel is gone. I re-created the panel, but after the next boot, it is gone again. This is quite bothersome. I cannot keep on recreating the panel every time, certainly because I do modify it quite a bit from the default. STEPS TO REPRODUCE 1. KDE Frameworks from 5.99.0 to 5.102.0 2. Reboot OBSERVED RESULT No panel EXPECTED RESULT Panel SOFTWARE/OS VERSIONS KDE Plasma Version: 5.26.5 KDE Frameworks Version: 5.102.0 Qt Version: 5.15.8 ADDITIONAL INFORMATION It may be that starting a new session is enough to trigger the issue. But I have not tried that yet. Plasma-related entries in my logs: --- plasmashell[1162]: Aborting shell load: The activity manager daemon (kactivitymanagerd) is not running. plasmashell[1162]: If this Plasma has been installed into a custom prefix, verify that its D-Bus services dir is known to the system for the daemon to be activatable. […] plasmashell[1162]: Aborting shell load: The activity manager daemon (kactivitymanagerd) is not running. plasmashell[1162]: If this Plasma has been installed into a custom prefix, verify that its D-Bus services dir is known to the system for the daemon to be activatable. […] plasmashell[1162]: kf.plasma.quick: Applet preload policy set to 1 […] plasmashell[1162]: org.kde.plasma.containmentlayoutmanager: Error: cannot change the containment to AppletsLayout --- Here, all but the 5th line are at the error level. The fifth line is at the info level.
> plasmashell[1162]: Aborting shell load: The activity manager daemon (kactivitymanagerd) is not running. This indicates that your system is broken, so Plasma doesn't load properly. kactivitymanagerd needs to be running. Any idea why it might not be?
(In reply to Nate Graham from comment #1) > > plasmashell[1162]: Aborting shell load: The activity manager daemon (kactivitymanagerd) is not running. > This indicates that your system is broken, so Plasma doesn't load properly. > kactivitymanagerd needs to be running. Any idea why it might not be? I looked a bit broader and kactivitymanagerd is running: --- dbus-daemon[1035]: [session uid=1000 pid=1035] Activating via systemd: service name='org.kde.ActivityManager' unit='plasma-kactivitymanagerd.service' requested by ':1.14' (uid=1000 pid… … systemd[1013]: Starting plasma-kactivitymanagerd.service... … plasmashell[1131]: Aborting shell load: The activity manager daemon (kactivitymanagerd) is not running. plasmashell[1131]: If this Plasma has been installed into a custom prefix, verify that its D-Bus services dir is known to the system for the daemon to be activatable. … dbus-daemon[1035]: [session uid=1000 pid=1035] Successfully activated service 'org.kde.ActivityManager' systemd[1013]: Started plasma-kactivitymanagerd.service. plasmashell[1131]: Aborting shell load: The activity manager daemon (kactivitymanagerd) is not running. plasmashell[1131]: If this Plasma has been installed into a custom prefix, verify that its D-Bus services dir is known to the system for the daemon to be activatable. --- So the first seems to be a race condition: plasmashell expects kactivitymanagerd to be ready while it is still loading. I guess the second time the warning is logged may also be a race condition, but then with output arriving in the log once it has already been resolved (this is all happening within the same second). So it seems that plasmashell should explicitly be told to wait for kactivitymanagerd to be loaded, which is likely another bug.
Hmm, kactivitymanagerd's service file has this in it: Before=graphical-session.target
(In reply to Nate Graham from comment #3) > Hmm, kactivitymanagerd's service file has this in it: > > Before=graphical-session.target But is graphical-session.target the one that matters here? I have the feeling it is not. In the log, I see --- systemd[1128]: Starting plasma-kactivitymanagerd.service... systemd[1128]: Started plasma-plasmashell.service. --- So plasmashell is started essentially concurrently with kactivitymanagerd. This is not strange, as we can see from their unit sections in their service files: ---plasma-plasmashell.service--- [Unit] Description=KDE Plasma Workspace After=plasma-ksmserver.service plasma-kcminit.service PartOf=graphical-session.target --- ---plasma-kactivitymanagerd.service--- [Unit] Description=KActivityManager Activity manager Service PartOf=graphical-session.target Before=graphical-session.target --- So AFAIU, in plasma-plasmashell.service, we should have the following as the ‘After’ line: --- After=plasma-ksmserver.service plasma-kcminit.service plasma-kactivitymanagerd.service ---
Cool, feel free to submit a patch!
Don't make a patch doing that. Blocking plasma from starting until KAMD is fully up would make everything slower and is missing some key parts of the story. KActivityManager daemon is DBus activated. Plasmashell will make a call to kamd track progress that way. It does it internally, that's why we don't have a strict systemd dependency. we can see someone doing that here: >dbus-daemon[1035]: [session uid=1000 pid=1035] Activating via systemd: service name='org.kde.ActivityManager' unit='plasma-kactivitymanagerd.service' requested by ':1.14' (uid=1000 pid… dbus-daemon should keep track of that all things and let plasmashell know which has it's own tracking. If we get in to the error path listed it means we've been told that starting If you get the error listed it means plasmashell explicitly thinks kamd failed to start rather than a race. Can you attach the non-abridged logs.
Created attachment 156989 [details] Journal output until and including issue (In reply to David Edmundson from comment #6) > > If you get the error listed it means plasmashell explicitly thinks kamd failed to start rather than a race. Can you attach the non-abridged logs. These are the non-abridged logs up until and including the issue.
Dear Bug Submitter, This bug has been in NEEDSINFO status with no change for at least 15 days. Please provide the requested information as soon as possible and set the bug status as REPORTED. Due to regular bug tracker maintenance, if the bug is still in NEEDSINFO status with no change in 30 days the bug will be closed as RESOLVED > WORKSFORME due to lack of needed information. For more information about our bug triaging procedures please read the wiki located here: https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging If you have already provided the requested information, please mark the bug as REPORTED so that the KDE team knows that the bug is ready to be confirmed. Thank you for helping us make KDE software even better for everyone!
(In reply to Bug Janitor Service from comment #8) > Dear Bug Submitter, > > This bug has been in NEEDSINFO status with no change for at least > 15 days. Please provide the requested information as soon as > possible and set the bug status as REPORTED. Due to regular bug > tracker maintenance, if the bug is still in NEEDSINFO status with > no change in 30 days the bug will be closed as RESOLVED > WORKSFORME > due to lack of needed information. > > For more information about our bug triaging procedures please read the > wiki located here: > https://community.kde.org/Guidelines_and_HOWTOs/Bug_triaging > > If you have already provided the requested information, please > mark the bug as REPORTED so that the KDE team knows that the bug is > ready to be confirmed. > > Thank you for helping us make KDE software even better for everyone! I think the requested info has already been provided, should we change back the status to "Reported"?
Relevant log bits: 09:46:36 PCNAME dbus-daemon[1040]: [session uid=1000 pid=1040] Activating via systemd: service name='org.kde.ActivityManager' unit='plasma-kactivitymanagerd.service' requested by ':1.14' (uid=1000 pid=1133 comm="/usr/bin/plasmashell --no-respawn") ... 09:46:36 PCNAME systemd[1018]: Started plasma-plasmashell.service. ... 09:46:36 PCNAME plasmashell[1133]: Aborting shell load: The activity manager daemon (kactivitymanagerd) is not running. ... 09:46:36 PCNAME dbus-daemon[1040]: [session uid=1000 pid=1040] Successfully activated service 'org.kde.ActivityManager' 09:46:36 PCNAME systemd[1018]: Started plasma-kactivitymanagerd.service. 09:46:36 PCNAME plasmashell[1133]: Aborting shell load: The activity manager daemon (kactivitymanagerd) is not running. This sequence of events suggests that plasmashell is trying to load before kamd has finished.
*** Bug 466918 has been marked as a duplicate of this bug. ***
*** Bug 468627 has been marked as a duplicate of this bug. ***
>This sequence of events suggests that plasmashell is trying to load before kamd has finished. That's not the full story. Plasmashell will try to start kamd internally (via DBus activation) and only proceed when that gets a reply/failure to our original call. We need to work out what that call fails.
I noticed BUG 466918 was marked as a duplicate of this, which is very similar to what's happening to me, however there they said it happens consistently, while in my case it happens intermittently. This week Plasma (5.27.6) failed to load twice after starting up the computer (I see the wallpaper and the mouse cursor, but the shell never loads). Restarting the computer fixes the issue and it happens seemingly randomly. I noticed my logs show the same message "Aborting shell load: The activity manager daemon (kactivitymanagerd) is not running.", however, I checked 3 other successful boots and they also show the same message twice, even though the shell loads succefully. I also don't have any problems with the panel disappearing. I also tested my Plasma 6 VM and it shows the same message as well after a successful boot. Can someone confirm the same message also appears after a successful boot? Maybe the issue is something else entirely.
(In reply to guimarcalsilva from comment #14) > I noticed BUG 466918 was marked as a duplicate of this, which is very > similar to what's happening to me, however there they said it happens > consistently, while in my case it happens intermittently. This week Plasma > (5.27.6) failed to load twice after starting up the computer (I see the > wallpaper and the mouse cursor, but the shell never loads). Restarting the > computer fixes the issue and it happens seemingly randomly. I noticed my > logs show the same message "Aborting shell load: The activity manager daemon > (kactivitymanagerd) is not running.", however, I checked 3 other successful > boots and they also show the same message twice, even though the shell loads > succefully. I also don't have any problems with the panel disappearing. > > I also tested my Plasma 6 VM and it shows the same message as well after a > successful boot. > > Can someone confirm the same message also appears after a successful boot? > Maybe the issue is something else entirely. I can confirm this, 100% identical behaviour on my system. Also appears after a successful boot, and only very occasionally the shell actually fails to load.
Created attachment 160293 [details] Log when shell failed to load PS: I deleted the last few lines where I switched TTYs and tried to kill plasmashell with killall since that's probably not useful.
Created attachment 160294 [details] Known good boot Known good boot - Shows the same message
*** Bug 473514 has been marked as a duplicate of this bug. ***