Re: [Jack-Devel] jack2/ubuntu issue: jack2 does not quit
On 04/08/2012 03:16 PM, Nedko Arnaudov wrote:
> Tim Blechmann<[hidden]> writes:
>
>> after upgrading to ubuntu 12.04, jackd2 does not quit correctly:
>>
>> hitting `quit' gives me the following message:
>>
>> 11:17:31.920 Client deactivated.
>> 11:17:31.936 D-BUS: JACK server could not be stopped. Sorry
here it says that qjackctl failed to stop the jack-server via d-bus
something's wrong with jackdbus...
>> 11:17:31.937 ALSA connection graph change.
>> Sun Apr 8 11:17:31 2012: Client 'qjackctl' with PID 3821 is out
>> Sun Apr 8 11:17:31 2012: Stopping jack server...
>> Sun Apr 8 11:17:31 2012: Client 'system' with PID 0 is out
>> Sun Apr 8 11:17:31 2012: Client 'PulseAudio JACK Sink' with PID 2727 is out
>> Sun Apr 8 11:17:31 2012: Client 'PulseAudio JACK Source' with PID 2727 is out
>
> So you use qjackctl for controlling jackdbus...
>
>> Sun Apr 8 11:17:31 2012: ERROR: Abort!
>> Sun Apr 8 11:17:31 2012: ERROR: info.si_signo = 6
>> Sun Apr 8 11:17:31 2012: ERROR: info.si_errno = 0
>> Sun Apr 8 11:17:31 2012: ERROR: Segmentation Fault!
>> Sun Apr 8 11:17:31 2012: ERROR: info.si_signo = 11
>> Sun Apr 8 11:17:31 2012: ERROR: info.si_errno = 0
>> Sun Apr 8 11:17:31 2012: ERROR: info.si_code = 1 (SEGV_MAPERR)
>> Sun Apr 8 11:17:31 2012: ERROR: info.si_addr = 0x525245
>> Sun Apr 8 11:17:31 2012: ERROR: reg[00] = 0x00007f808029a582
>> Sun Apr 8 11:17:31 2012: ERROR: reg[01] = 0x0000000000000000
>> Sun Apr 8 11:17:31 2012: ERROR: reg[02] = 0x00007f808029a2b8
>> Sun Apr 8 11:17:31 2012: ERROR: reg[03] = 0x0000000000000246
>> Sun Apr 8 11:17:31 2012: ERROR: reg[04] = 0x00007f808029a6f8
>> Sun Apr 8 11:17:31 2012: ERROR: reg[05] = 0x0000000000413c3b
>> Sun Apr 8 11:17:31 2012: ERROR: reg[06] = 0x0000000000000000
>> Sun Apr 8 11:17:31 2012: ERROR: reg[07] = 0x0000000000000014
>> Sun Apr 8 11:17:31 2012: ERROR: reg[08] = 0x0000000000525245
>> Sun Apr 8 11:17:31 2012: ERROR: reg[09] = 0x0000000000413c4d
>> Sun Apr 8 11:17:31 2012: ERROR: reg[10] = 0x00007f808029a3f0
>> Sun Apr 8 11:17:31 2012: ERROR: reg[11] = 0x00007f808029a400
>> Sun Apr 8 11:17:31 2012: ERROR: reg[12] = 0x0000000000000018
>> Sun Apr 8 11:17:31 2012: ERROR: reg[13] = 0x0000000000000000
>> Sun Apr 8 11:17:31 2012: ERROR: reg[14] = 0xffffffffffffffff
>> Sun Apr 8 11:17:31 2012: ERROR: reg[15] = 0x00007f8080299d90
>> Sun Apr 8 11:17:31 2012: ERROR: reg[16] = 0x00007f808402cdba
>> Sun Apr 8 11:17:31 2012: ERROR: reg[17] = 0x0000000000010246
>> Sun Apr 8 11:17:31 2012: ERROR: reg[18] = 0x0000000000000033
>> Sun Apr 8 11:17:31 2012: ERROR: reg[19] = 0x0000000000000004
>> Sun Apr 8 11:17:31 2012: ERROR: reg[20] = 0x000000000000000e
>> Sun Apr 8 11:17:31 2012: ERROR: reg[21] = 0x0000000080005a27
>> Sun Apr 8 11:17:31 2012: ERROR: reg[22] = 0x0000000000525245
>> Sun Apr 8 11:17:31 2012: ERROR: Stack trace:
>> Sun Apr 8 11:17:31 2012: ERROR: 1: 0x7f808402cdba
>> <_IO_vfprintf+10042> (/lib/x86_64-linux-gnu/libc.so.6)
>> Sun Apr 8 11:17:31 2012: ERROR: 2: 0x7f80840ea580
>> <__vsnprintf_chk+176> (/lib/x86_64-linux-gnu/libc.so.6)
>> Sun Apr 8 11:17:31 2012: ERROR: End of stack trace
>>
jackdbus crashed (segfault) and spitted out some kind of glibc stack
trace...
>>
>> sometimes i even have to killall -9 jackdbus ...
>
> Unfortunately, this stack trace is not very useful. You should be able
> to get backtrace of all threads by using one of the techniques described
> here:
>
> http://ladish.org/wiki/gdb_dbus_service
>
> Please provide the output of the "thread apply all bt". Enabling verbose
> log (it is a jackdbus "engine" parameter) could give additional info.
>
> I'm not sure what would be the correct approach to debug jackdbus when
> qjackctl is in the mix. In case neither Rui nor annybody else knows, you
> should be able to reproduce the crash without qjackctl, by using the
> commandline frontend to jackdbus, the jack_control script that is part
> of the distribution tarball and should be installed on your distro along
> with jackdbus itself.
>
i cannot tell you why that happens but the remaining error/abort/crash
messages do come from the jackdbus log file (~/.log/jack/jackdbus.log)
-- you can tell that by the funny escape character sequences that appear
on each "ERROR:..." line, allegedly meaningful on a "polychromatic" tty
terminal only
byee
--
rncbc aka Rui Nuno Capela
[hidden]
1333900262.15727_0.ltw:2,a <4F81B3F1.2040808 at rncbc dot org>