GNOME Bugzilla – Bug 329015
hangs on startup if esd not running
Last modified: 2009-05-10 21:36:30 UTC
I'm using the ubuntu snaps from http://snapshots.gnomemeeting.net/ (for breezy, but I'm using on dapper- maybe that is relevant). Anyway, esd does not run by default for me, and when I start ekiga it hangs, looking for esd:
+ Trace 65705
if I start esd, ekiga runs apparently normally. (Though it does not detect my sound devices, but maybe that is a different problem?)
Having now been told that there are dapper snaps... I've upgraded, but still get the same problem and same trace.
Hmmm... I'm using dapper, I never have esd running either but I never had that crash. I don't see what bad can happen when I look at the relevant code in src/devices/audio.cpp: void gnomemeeting_sound_daemons_suspend (void) { #ifndef DISABLE_GNOME int esd_client = 0; /* Put ESD into standby mode */ esd_client = esd_open_sound (NULL); if (esd_client >= 0) { esd_standby (esd_client); esd_close (esd_client); } #endif }
It isn't a crash, 'just' a hang on startup. The trace is from when I attach gdb to the hung process.
Well, this is confusing : one could think the hang is due to the esd_open_sound function not finding the daemon. This is very tempting. Unfortunately, since it doesn't hang here, that explanation has to be ruled out. Could you try "thread apply all bt full", to see if the hang doesn't happen somewhere else ? (say, like for dbus...)
The fact that it doesn't detect your sound devices indicates another problem, which, I think, is related to your configuration. About ESD, I don't run ESD either, and the ESD code is 4 years old in Ekiga, so I have no explanation unfortunately. Perhaps thread apply all bt full will give another hint.
Full trace: Program received signal SIGINT, Interrupt. [Switching to Thread -1243403616 (LWP 23806)] 0xffffe410 in __kernel_vsyscall () (gdb) thread apply all bt full
+ Trace 65712
Eh... nothing is initialized yet ! It happens really early... strange and interesting.
Created attachment 58583 [details] Esound suspend/resume. Let's see if/how that little test program works on your configuration... that should settle if we're chasing the right fox.
What should I be looking for from it? currently it just starts and exits.
I was trying to see if this program crashed, freezed of whatever... if it doesn't, then probably the problem is not where we thought. Now where is it !?
Does anybody else have the problem ? I'm really stuck here...
Someone reported another crash where the evil happens in esound code: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=189940 The backtrace is not exactly the same though :-/ (and the user doesn't use ubuntu)
*** Bug 346403 has been marked as a duplicate of this bug. ***
Two people reporting esound issues and I have no idea what the problem can be :-/ 100% pure joy
Reassigning the bug to ESD. We correctly ask to suspend ESD, so the problem is outside of Ekiga.
I guess this is rather OBSOLETE nowadays and can be closed?
Probably yes.
Closing as per comment #16.