jack2 vs jack1 setup and performance tuning

[Jörn Nettingsmeier-5]

hi *!


i've just moved my studio machine to jack2 temporarily in order to test
the netone port. after two days, i'm pretty sure that it xruns more
often than jack1. i'm re-using my jack1 commandline, with the exception
of the -Z nozombify option which jack2 doesn't seem to support any more
(then again, it apparently doesn't zombify clients, so that's not a
problem).

here's my configuration and the startup messages, pasted from the
qjackctl log:

10:13:21.585 /usr/local/bin/jackd -R -p1024 -t9999 -u -dalsa -dhw:1
-r48000 -p64 -n2 -zs -M
10:13:21.595 JACK was started with PID=4222.
no message buffer overruns
no message buffer overruns
jackdmp 1.9.4
Copyright 2001-2005 Paul Davis and others.
Copyright 2004-2009 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
JACK server starting in realtime mode with priority 10
creating alsa driver ... hw:1|hw:1|64|2|48000|0|0|nomon|hwmeter|-|32bit
Using ALSA driver RME9652 running on card 1 - RME Digi9652 (Rev 1.5) at
0xe8000000, irq 17
configuring for 48000Hz, period = 64 frames (1.3 ms), buffer = 2 periods
ALSA: final selected sample format for capture: 32bit integer little-endian
ALSA: use 2 periods for capture
ALSA: final selected sample format for playback: 32bit integer little-endian
ALSA: use 2 periods for playback


i see error messages like this very frequently:
JackEngine::XRun: client = edcast was not run: state = 2
JackAudioDriver::ProcessAsync Process error
12:08:15.242 XRUN callback (1).

also the "process error" msgs often come in groups of 20-40.

kernel is 2.6.31.4-rt14, hardware is single-core amd64 and an rme digi9652.

one other thing i changed is that i'm using verbose messages. might that
be triggering cascade problems (i.e. one info message gets sent, timing
issues occur, more messages, xrun...)?


best,

jörn






--
Jörn Nettingsmeier

Meister für Veranstaltungstechnik

Audio and event engineer
Ambisonic surround recordings

http://stackingdwarves.net
[phone]

_______________________________________________
Jack-Devel mailing list
Jack-Devel@...
http://lists.jackaudio.org/listinfo.cgi/jack-devel-jackaudio.org

[Stéphane Letz]

Le 30 oct. 2009 à 13:34, Jörn Nettingsmeier a écrit :

> hi *!
>
>
> i've just moved my studio machine to jack2 temporarily in order to  
> test
> the netone port. after two days, i'm pretty sure that it xruns more
> often than jack1. i'm re-using my jack1 commandline, with the  
> exception
> of the -Z nozombify option which jack2 doesn't seem to support any  
> more
> (then again, it apparently doesn't zombify clients, so that's not a
> problem).
>
> here's my configuration and the startup messages, pasted from the
> qjackctl log:
>
> 10:13:21.585 /usr/local/bin/jackd -R -p1024 -t9999 -u -dalsa -dhw:1
> -r48000 -p64 -n2 -zs -M
> 10:13:21.595 JACK was started with PID=4222.
> no message buffer overruns
> no message buffer overruns
> jackdmp 1.9.4
> Copyright 2001-2005 Paul Davis and others.
> Copyright 2004-2009 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
> JACK server starting in realtime mode with priority 10
> creating alsa driver ... hw:1|hw:1|64|2|48000|0|0|nomon|hwmeter|-|
> 32bit
> Using ALSA driver RME9652 running on card 1 - RME Digi9652 (Rev 1.5)  
> at
> 0xe8000000, irq 17
> configuring for 48000Hz, period = 64 frames (1.3 ms), buffer = 2  
> periods
> ALSA: final selected sample format for capture: 32bit integer little-
> endian
> ALSA: use 2 periods for capture
> ALSA: final selected sample format for playback: 32bit integer  
> little-endian
> ALSA: use 2 periods for playback
>
>
> i see error messages like this very frequently:
> JackEngine::XRun: client = edcast was not run: state = 2
> JackAudioDriver::ProcessAsync Process error
> 12:08:15.242 XRUN callback (1).
>
> also the "process error" msgs often come in groups of 20-40.
>
> kernel is 2.6.31.4-rt14, hardware is single-core amd64 and an rme  
> digi9652.

I've just switch scheduling policy to SCHED_FIFO (what SCHED_RR)

Can you test latest SVN rev 3683?

>
> one other thing i changed is that i'm using verbose messages. might  
> that
> be triggering cascade problems (i.e. one info message gets sent,  
> timing
> issues occur, more messages, xrun...)?
>
>
> best,
>
> jörn

Should not since printing verbose messages is done from another thread  
with a lock-free way of sending messages to print.

Stéphane
_______________________________________________
Jack-Devel mailing list
Jack-Devel@...
http://lists.jackaudio.org/listinfo.cgi/jack-devel-jackaudio.org

[Jörn Nettingsmeier-5]

Stéphane Letz wrote:

>
> Le 30 oct. 2009 à 13:34, Jörn Nettingsmeier a écrit :
>
>> hi *!
>>
>>
>> i've just moved my studio machine to jack2 temporarily in order to test
>> the netone port. after two days, i'm pretty sure that it xruns more
>> often than jack1.
>
> I've just switch scheduling policy to SCHED_FIFO (what SCHED_RR)
>
> Can you test latest SVN rev 3683?

yup. i updated but didn't see any effect. however, geoff pointed out i
should be using -P 60, and that did it. nedko mentioned that jack2's
default rtprio was 10, which would explain what i was seeing.
stable now, thanks.



--
Jörn Nettingsmeier

Meister für Veranstaltungstechnik

Audio and event engineer
Ambisonic surround recordings

http://stackingdwarves.net
[phone]

_______________________________________________
Jack-Devel mailing list
Jack-Devel@...
http://lists.jackaudio.org/listinfo.cgi/jack-devel-jackaudio.org