[PlanetCCRMA] Configuring 2.6.11-0.3.rdt.rhfc2.ccrma for 0 xruns...

[Fernando Lopez-Lezcano]

On Thu, 2005-04-14 at 15:54, Rui Nuno Capela wrote:
> >> OK. That's it. Here goes a new rtirq tarball (rtirq-20050414.tar.gz)
> >> with this new configuration option:
> >>
> >>   RTIRQ_NON_THREADED="rtc snd"
> >>
> >> where one tells which IRQ handler services are to be left NOT threaded
> >> when rtirq start is issued. Default goes for the realtime clock (rtc)
> >> and every ALSA sound driver (snd).
> >>
> >> NOTE: This has been NOT tested, as I have no PREEMPT_DESKTOP kernel
> >> near me ;)
> >
> > I did a simple test and it looks like it works but is very verbose. This
> > is on my laptop, on which almost everything is hooked into irq 9, so
> > this is pathological... but still, see what happens when I start
> > rtirq...
> >
> > rtirq: start [rtc] irq=8 /proc/irq/8/rtc/threaded: OK(0).
> > rtirq: start [rtc] irq=8 /proc/irq/8/rtc/threaded pid=257 prio=80: OK.
> > rtirq: start [snd] irq=9 /proc/irq/9/acpi/threaded: OK(0).
> > rtirq: start [snd] irq=9 /proc/irq/9/acpi/threaded
> > /proc/irq/9/eth0/threaded: OK(0).
> > rtirq: start [snd] irq=9 /proc/irq/9/acpi/threaded
> > /proc/irq/9/eth0/threaded /proc/irq/9/Intel 82801CA-ICH3 Modem/threaded:
> > OK(0).
[MUNCH]
>
> Damn right pathological! IRQ 9 is really hot in there!

Oh yes, that pin in the processor is really really hot!

> But it looks like rtirq is doing the right thing anyway. Only too much
> verbosity, I'm sure.
> 
> Now that I|we believe its doing it, what about suppressing those messages?
> Next micro-tarball coming up... (rtirq-20050415.tar.gz).

Uh, looks back to normal!
Thanks as usual for the fast response.... getting a release ready :-)
-- Fernando