[linux-audio-user] Re: Recent LKML discussion on preempt/latency in 2.6 kernels

Fernando Pablo Lopez-Lezcano nando at ccrma.Stanford.EDU
Thu Mar 25 16:31:45 EST 2004

On Thu, 2004-03-25 at 13:05, Steve Harris wrote:
> On Thu, Mar 25, 2004 at 11:35:26AM -0800, Fernando Pablo Lopez-Lezcano wrote:
> > I was adding:
> >   realtime security module (LSM) 0.0.4
> >     http://www.joq.us/realtime
> >   low latency scheduling additions by Takashi Iwai
> >     http://kerneltrap.org/node/view/2702
> >   drm scheduling additions (radeon/mga/r128) --> broken!
> What kind of latency were you getting with this, compared to patched 2.4?

a) preempt enabled, Takashi's patches included, x11 sucks...
b) preempt enabled, added tweaked radeon low latency patch, much better:
c) preempt disabled, with Takashi's and radeon patches:
d) same as c), with preempt enabled again:
e) same as d), without loading the xfree86 dri module:

This is from the output of latencytest 0.52 and looks very good. But,
and it is not a small but, when I tried to use real jack and jack
applications I could not go down to 2x128 at all, kept getting small
xruns. But an idle jack at 2x64 plus disk activity ("tar cf usr.tar
/usr") would very seldom generate xruns. 

So there is something else that needs tweaking[*]. 

Takashi's patches do improve the situation quite a lot. 
-- Fernando

[*] plus my setup was not really that stable. For some reason depending
on how I had the hwconf setup done by kudzu (or not) I would get into
situations where moving the mouse was creating all sorts of small xruns
in an otherwise idle jack. That drove me nuts for the better part of a
day, where I could not get back to the nice "almost no xrun" situation
although I thought I was "undoing" everything I had done. 

More information about the Linux-audio-user mailing list