Thanks Robin for your massively useful tips there on how to (stress) test JACK - I think what you just said, in an edited form, should be included on the JACK FAQ page please Paul!<br><br><br><div class="gmail_quote">On Sun, Dec 12, 2010 at 12:41 PM, Robin Gareus <span dir="ltr"><<a href="mailto:robin@gareus.org">robin@gareus.org</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div class="im">On 12/12/2010 01:03 PM, ailo wrote:<br>
> I've been looking around for any tests made comparing the different<br>
> kernels, -rt, generic, or any other type of realtime enchanced kernel.<br>
> I haven't found any test results yet, at least none audio related. I did<br>
> find some testing tools at <a href="http://rt.wiki.kernel.org" target="_blank">rt.wiki.kernel.org</a>, but don't know if and how<br>
> they could be made relevant to audio low latency testing.<br>
><br>
> I suppose the most interesting results would come from testing different<br>
> kernels with jack/alsa and jack/ffado.<br>
><br>
> Has anyone done such tests?<br>
<br>
</div>It is not trivial to perform such tests and AFAIK there's no benchmark<br>
suite to automate the process.<br>
<br>
There are a few tools to test JACK's realtime performance:<br>
<br>
- the ardour-source includes `tools/jacktest.c` checks for the max DSP<br>
load at which an x-run occurs.<br>
<br>
- <a href="http://wiki.linuxaudio.org/wiki/jack_latency_tests" target="_blank">http://wiki.linuxaudio.org/wiki/jack_latency_tests</a><br>
git://<a href="http://rg42.org/latentor" target="_blank">rg42.org/latentor</a> is a tool to automate measuring round-trip<br>
audio latency iterating all JACKd -n/-p/-S parameters<br>
However latentor is a pretty recent development and does not yet<br>
report x-runs. We watch qjackctl's icon for now.<br>
<br>
AFAICT there's no recipe. It's a matter of knowing some internals about<br>
RT-linux to come up with a proper kernel .config and doing real-life<br>
tests. I think it is impossible to assign a number "suitability for<br>
pro-audio" to a kernel.<br>
<br>
For testing performance of the 64studio RT kernel: I do run a couple of<br>
heavy-sessions (e.g. 16 jconvolvers in a 16 track ardour session + jamin<br>
which procudes quite some DSP, system and IO load). If there's no x-run<br>
at 32fpp*2p/48kHz after 24 h while I to surf the web and read email and<br>
compile another kernel in the meantime I bless the build OK :) There's a<br>
few additional things: wifi, suspend/resume, freq scaling, etc on the<br>
checklist, too.<br>
<br>
2c,<br>
<font color="#888888">robin<br>
</font><div><div></div><div class="h5">_______________________________________________<br>
Linux-audio-user mailing list<br>
<a href="mailto:Linux-audio-user@lists.linuxaudio.org">Linux-audio-user@lists.linuxaudio.org</a><br>
<a href="http://lists.linuxaudio.org/listinfo/linux-audio-user" target="_blank">http://lists.linuxaudio.org/listinfo/linux-audio-user</a><br>
</div></div></blockquote></div><br>