Arch Linux

backtrace:

#0 0x00007ffff60a9bf3 in Jack::JackNetOneDriver::FreePorts() ()
from /usr/lib/jack/jack_netone.so
#1 0x00007ffff60a9e66 in Jack::JackNetOneDriver::Initialize() ()
from /usr/lib/jack/jack_netone.so
#2 0x00007ffff7b99251 in Jack::JackWaitThreadedDriver::JackDriverStarter::Execute() () from /usr/lib/libjackserver.so.0
#3 0x00007ffff7b8b240 in Jack::JackPosixThread::ThreadHandler(void*) ()
from /usr/lib/libjackserver.so.0
#4 0x00007ffff6e1ddd2 in start_thread () from /usr/lib/libpthread.so.0
#5 0x00007ffff7117cdd in clone () from /usr/lib/libc.so.6

Looks almost exactly the same as the debian bug report, which is confirmed, and yet I am not able to reproduce this. Are you on i686 or x86_64? In the meantime I'll have to look at how to test netjack functionality locally.

edit: And I also notice that bug is closed. Could you try to reproduce with jack2-git from AUR? [1] Thanks.

[1] https://aur.archlinux.org/packages/jack2-git/

The crash happens on x86_64 only, on i686 it works. No luck with jack2-git, though, leads to the same crash.

jackdmp 1.9.10
Copyright 2001-2005 Paul Davis and others.
Copyright 2004-2013 Grame.
jackdmp comes with ABSOLUTELY NO WARRANTY
This is free software, and you are welcome to redistribute it
under certain conditions; see the file COPYING for details
no message buffer overruns
no message buffer overruns
no message buffer overruns
JACK server starting in realtime mode with priority 10
Speicherzugriffsfehler (Speicherabzug geschrieben)

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffff7ebd700 (LWP 7098)]
0x00007ffff60a7bf3 in Jack::JackNetOneDriver::FreePorts() ()
from /usr/lib/jack/jack_netone.so
(gdb) bt
#0 0x00007ffff60a7bf3 in Jack::JackNetOneDriver::FreePorts() ()
from /usr/lib/jack/jack_netone.so
#1 0x00007ffff60a7e66 in Jack::JackNetOneDriver::Initialize() ()
from /usr/lib/jack/jack_netone.so
#2 0x00007ffff7b99661 in Jack::JackWaitThreadedDriver::JackDriverStarter::Execute() () from /usr/lib/libjackserver.so.0
#3 0x00007ffff7b895b0 in Jack::JackPosixThread::ThreadHandler(void*) ()
from /usr/lib/libjackserver.so.0
#4 0x00007ffff6e1bdd2 in start_thread () from /usr/lib/libpthread.so.0
#5 0x00007ffff7115cdd

Please try this build and report if there is any improvement: http://pkgbuild.com/~schiv/

Assigning another packager to help with debugging/testing on x86_64.

The build at http://pkgbuild.com/~schiv/ crashes in the same way.

Before I mark this as upstream, have you used this functionality before? Or is this the first time you are using -d netone? If yes, when is the last time that you remember this to have worked? State the package version and whether it was an Arch Linux package, or a build on another unrelated system.

If not, get an old binary from http://arm.konnichi.com/search/index.php?a=64&q=jack2&core=1&extra=1&community=1 or rebuild an old version from ABS. Preferably, the last release prior to the current (1.9.8-4). See if they exhibit the same problem.

I've used the netone backend before on the same Arch machine with the official package, but it's a while back (couple of months).

jack2-1.9.7-3-x86_64.pkg.tar.xz works, all later versions do crash. Thanks a lot.

I think you might be seeing a corner-case bug here. There are no similar reports around and the Debian bug report above clearly implies that the problem (likely to be similar to yours) is solved. I suggest you file a bug report upstream at https://github.com/jackaudio/jack2/issues?state=closed and cite this last working version.

Yep, another user has confirmed this, so this is not really as unique as I suspected. You should definitely file a bug report since you can reproduce this, since unfortunately I am on i686 and cannot follow through any troubleshooting that may be required, unless of course someone wants to give me shell access.

I have exactly the same problem. I am using the x86_64 but have not tried the i686 package.

I've raised an issue on github 2 months ago as proposed by schivmeister.

https://github.com/jackaudio/jack2/issues/38

Maybe it was the wrong channel, I'll try the mailing list next, seems to be more responsive...

I have retested current git HEAD and the issue seems to have disappeared in the meantime.
I'll stick to git until 1.9.10 is out.

More specifically: Also version 1.9.9.5 works fine when compiled without firewire support here.

_wafconf() {
python2 waf configure --prefix=/usr \
--alsa $@
}

Thanks for the quick response. I recompiled the package on my notebook without firewire support. That seems to fix the problem.