[pulseaudio-tickets] [Bug 48608] "Assertion 'b' failed at pulsecore/memblock.c:454, function pa_memblock_acquire()" when running anything that deals with sound levels

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=48608

--- Comment #2 from Tanu Kaskinen <tanuk at iki.fi> 2012-05-02 09:54:35 UTC ---
It seems that pa_stream_peek() doesn't take into account the possibility that
there's a gap in s->record_memblockq. I'm not sure what would cause that gap,
but the stack trace shows that there is a gap - there's no other explanation
for s->peek_memchunk.memblock being NULL after a successful pa_memblockq_peek()
call. The fix might be as simple as giving a silence memchunk to
pa_memblockq_new() so that the gaps will be filled by silence, but maybe first
it should be understood in which conditions record streams can have gaps - if
there shouldn't be gaps at all, then the silence memchunk solution is only a
workaround.


(In reply to comment #1)
> This seems to correlate with the 'pid' and 'dbus-socket' files going missing
> from the pulseaudio directory in /tmp.

That's peculiar. I don't know what that means. Especially the pid file
disappearing sounds odd.

> $ cd $TMP
> 
> $ find pulse-*
> pulse-fxXeyjndr5n9
> pulse-fxXeyjndr5n9/native
> 
> $ pavucontrol 
> Assertion 'b' failed at pulsecore/memblock.c:454, function
> pa_memblock_acquire(). Aborting.
> Aborted
> 
> $ pulseaudio -k
> E: [pulseaudio] main.c: Failed to kill daemon: No such process
> 
> $ kill 93621

What's 93621?

> $ pavucontrol
> 
> (at this point pulseaudio is re-launched and operates normally)
> 
> $ find pulse-*
> pulse-fxXeyjndr5n9
> pulse-fxXeyjndr5n9/native
> pulse-fxXeyjndr5n9/dbus-socket
> pulse-fxXeyjndr5n9/pid

-- 
Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA Contact for the bug.
You are the assignee for the bug.