If pulseaudio is non-responsive, plasmashell hangs

Bug #1694863 reported by Paul
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
plasmashell
Unknown
High
plasma-desktop (Ubuntu)
New
Undecided
Unassigned

Bug Description

If pulseaudio is non-responsive or not running for any reason (eg; messing with jack or a device error), pulseaudio will hang, most particularly if you try to adjust the volume. This essentially bricks the system unless you can manage to kill and restart pulseaudio without the plasma/KDE gui's and hotkeys.

KSysGuard and other plasma applications will also hang and crash alongside the shell.

TLDR; plasmashell has a bad blocking condition in it's audio code.

Revision history for this message
In , Paul (paul17041993) wrote :

If pulseaudio becomes unresponsive or is in a bad state, plasmashell can hang, leaving the entire graphical system unresponsive unless you can kill and restart pulseaudio.

This condition can occur when messing with jackd or if a device encounters an error or similar condition, of which if the user attempts to adjust the volume or trigger a plasma audio event, plasma will immediately hang indefinitely.

Audio calls need to not be blocking, or should be handled in a sandboxed thread with a timeout, there is no guarantee that they will finish immediately or at all (hence why window's did away with push events).

Changed in plasmashell:
importance: Unknown → High
status: Unknown → New
Revision history for this message
In , U26 (u26) wrote :

If it is hanging, can you get us a backtrace of it blocking please.

Changed in plasmashell:
status: New → Incomplete
Revision history for this message
In , Cfeck (cfeck) wrote :

If you can provide the information requested in comment #1, please add it.

You can interrupt gdb while plasmashell hangs to get the backtrace. Without this information, we have to assume that it is some upstream component that blocks.

Revision history for this message
In , Cfeck (cfeck) wrote :

To further investigate this issue, KDE developers need the information requested in comment #1. If you can provide it, or need help with finding that information, please add a comment.

Revision history for this message
In , Paul (paul17041993) wrote :

Created attachment 106365
stacktrace of plasmashell

sorry, had been busy with other things to be bricking my system again...

Revision history for this message
In , Cfeck (cfeck) wrote :

Thanks for the update; changing status.

I cannot see a cause for a freeze in the trace, but maybe plasma-pa developers have an idea where Plasma does blocking calls to pulseaudio.

Changed in plasmashell:
status: Incomplete → New
Revision history for this message
In , Justin Zobel (justin-zobel) wrote :

Thank you for the report, Paul.

As it has been a while since this was reported, can you please test and confirm if this issue is still occurring or if this bug report can be marked as resolved.

I have set the bug status to "needsinfo" pending your response, please change back to "reported" or "resolved/worksforme" when you respond, thank you.

Changed in plasmashell:
status: New → Incomplete
Revision history for this message
In , Paul (paul17041993) wrote :

I no-longer have the system, or a system that can reproduce it, so I guess we'll just close this report...

Changed in plasmashell:
status: Incomplete → Unknown
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.