On Mon, Oct 14, 2013 at 3:33 AM, Jeremy Jongepier <jeremy@email-addr-hiddenwrote:
>
> > I think there must be some automated way for the system to recognize
> > that jack is out of control and kill it. [1] notes:
> >
> > -t, --timeout int
> > Set client timeout limit in milliseconds. The default
> > is 500
> > msec. In realtime mode the client timeout must be smaller
> > than
> > the watchdog timeout (5000 msec).
> >
> > Interesting... "watchdog timeout"... but I have never seen any effect
> > from the watchdog mentioned here. Jack just keeps locking up the system
> > for way longer than 5 seconds.
>
>
>
> Afaik Jack2 doesn't have a watchdog thread.
>
Neither will the next release of Jack1.
I believe that the lockups here are actually caused by interactions with
the kernel.
I also wasn't aware that anyone had added
robustness-against-device-disappearance to any version of JACK. It was on
my list of things to do while I'm on a JACK-hacking excursion.
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@email-addr-hidden
http://lists.linuxaudio.org/listinfo/linux-audio-user
Received on Tue Oct 15 00:15:07 2013
This archive was generated by hypermail 2.1.8 : Tue Oct 15 2013 - 00:15:08 EEST