Bug 66131

The backtrace shows no crash. It looks like the backtrace is taken during normal execution. How exactly did you get that gdb log?

In data lunedì 24 giugno 2013 18:56:10, hai scritto:

Comment # 1 on bug 66131 from Tanu Kaskinen 
The backtrace shows no crash. It looks like the backtrace is taken during
normal execution. How exactly did you get that gdb log? 
You are receiving this mail because: 
You reported the bug. 


I'm sorry! It is the first time that I try to report a bug and I confess my 
inexperience with gdb! I use pulseaudio-module-jack but that always happens 
with linux-rt pulseaudio crashes after a few minutes. If you use pulseaudio 
without jackd there is no crash. But with any other kernel-module-jack 
pulseaudio not crashed!

(In reply to comment #2)
> I'm sorry! It is the first time that I try to report a bug and I confess my 
> inexperience with gdb!

No worries. The only problem with the backtrace is that it is taken at a wrong time - it looks like pulseaudio hadn't yet crashed when you took the backtrace. You need to wait until pulseaudio crashes (while it's running inside gdb), then the execution will automatically stop and you can run the "thread apply all bt full" command. The output will then hopefully show where exactly the crash happened.

In data lunedì 24 giugno 2013 19:18:06, hai scritto:

No worries. 
Well! While I learn to use GDB ...... You could try to use pulseaduio-module-
jack on a kernel-rt and see if it crashes on you too? With me pulseaudio with 
jackd crashes only with linux-rt, whereas with a normal kernel works very 
well! I assume that you know how to run pulseaudio on jackd: if not I send you 
the scripts that I use via qjackctl! Ok? thanks!

I installed Ubuntu's lowlatency kernel, which has the preempt patches. I'm running pulseaudio on top of jack, and I'm playing some music. No crashes yet, but I'll keep using this setup for a while.

Do you have some particular problems with gdb? You already got a backtrace, so I would imagine that there are no problems. You only need to reproduce the crash while pulseaudio is running in gdb.

In data martedì 25 giugno 2013 07:41:00, hai scritto:

I installed Ubuntu's lowlatency kernel, which has the preempt patches. 
A kernel with this patch?: 
https://www.kernel.org/pub/linux/kernel/projects/rt/3.8/

I'm running pulseaudio on top of jack, and I'm playing some music. No crashes 
yet, but I'll keep using this setup for a while. 
Do you have some particular problems with gdb? 
Yes, I can not launch jackd with qjackctl after pulseaudio in GDB because it 
crashes! Out of GDB but it's okay so I did not backtrace.
You already got a backtrace, so I would imagine that there are no problems. 
You only need to reproduce the crash while pulseaudio is running in gdb.

(In reply to comment #5)
> I installed Ubuntu's lowlatency kernel, which has the preempt patches. 

Actually the only difference between the lowlatency kernel and the generic kernel are in the configuration. There are no rt patches, which is why it is called a "lowlatency" kernel instead of a "rt" kernel.

OK, my mistake.

Compiling a custom kernel is more effort than I'm willing to do at this point.

In data martedì 25 giugno 2013 08:21:07, hai scritto:

Actually the only difference between the lowlatency kernel and the generic 
kernel are in the configuration. There are no rt patches, which is why it is 
called a "lowlatency" kernel instead of a "rt" kernel.


In fact, everything works fine except always with a kernel-rt with the patches 
that I mentioned earlier. The performance of a kernel-rt are much higher and, 
with it, jackd works really well. I use ROSA-linux. Here the kernels that I 
use or would like to use:
http://mib.pianetalinux.org/forum/viewtopic.php?p=22391#p22391

In data martedì 25 giugno 2013 08:27:57, hai scritto:

OK, my mistake. Compiling a custom kernel is more effort than I'm willing to 
do at this point.


Great! Thank you!

Hi!
Here is the report of gdb on pulseaudio with kernel-rt!:
*_Starting program: /usr/bin/pulseaudio _*
*_warning: no loadable sections found in added symbol-file system-supplied DSO at 
0x7ffff7ffa000_*
*_[Thread debugging using libthread_db enabled]_*
*_Using host libthread_db library "/lib64/libthread_db.so.1"._*
*_process 23390 is executing new program: /usr/bin/pulseaudio_*
*_[Thread debugging using libthread_db enabled]_*
*_Using host libthread_db library "/lib64/libthread_db.so.1"._*
*_W: [pulseaudio] pid.c: Stale PID file, overwriting._*
*_[New Thread 0x7fffec2a7700 (LWP 23553)]_*
*_[New Thread 0x7fffebaa6700 (LWP 23554)]_*
*_[New Thread 0x7fffeb2a5700 (LWP 23555)]_*
*_Detaching after fork from child process 23556._*
*_E: [alsa-sink-ALC269VB Analog] alsa-sink.c: Attivazione da parte di ALSA per scrivere nuovi 
dati sul dispositivo, ma non c'era nulla da scrivere._*
*_E: [alsa-sink-ALC269VB Analog] alsa-sink.c: Molto probabilmente si tratta di un bug nei 
driver ALSA "snd_hda_intel". Segnalare questo problema agli sviluppatori di ALSA._*
*_E: [alsa-sink-ALC269VB Analog] alsa-sink.c: Attivazione avvenuta con POLLOUT impostato 
-- tuttavia, una successiva snd_pcm_avail() ha ritornato 0 o un altro valore < min_avail._*
*_[Thread 0x7fffeb2a5700 (LWP 23555) exited]_*
*_[Thread 0x7fffebaa6700 (LWP 23554) exited]_*
*_[Thread 0x7fffec2a7700 (LWP 23553) exited]_*

*_Program terminated with signal SIGKILL, Killed._*
*_The program no longer exists._*
*_(gdb) _*


oiram/bin/selom
Da ognuno secondo le proprie capacità ad ognuno secondo i propri bisogni.
MIB-kernellinux-tester
Linux[1] 
MIB[2] Lilypond[3] Frescobaldi[4] Rosegarden[5] 

--------
[1] https://www.kernel.org/
[2] http://mib.pianetalinux.org/blog/
[3] http://lilypond.org/
[4] http://www.frescobaldi.org/
[5] http://www.rosegardenmusic.com

Hi!
I solved:
https://launchpad.net/kxstudio/+announcement/6610
Thanks!-- 
oiram/bin/selom
Da ognuno secondo le proprie capacità ad ognuno secondo i propri bisogni.
MIB-kernellinux-tester
Linux[1] 
MIB[2] Lilypond[3] Frescobaldi[4] Rosegarden[5] 

--------
[1] https://www.kernel.org/
[2] http://mib.pianetalinux.org/blog/
[3] http://lilypond.org/
[4] http://www.frescobaldi.org/
[5] http://www.rosegardenmusic.com

-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/pulseaudio/pulseaudio/issues/276.