Re: 2.6.14-rc4-rt1

[Fernando Lopez-Lezcano <nando@ccrma.Stanford.EDU>]

On Fri, 2005-10-14 at 08:15 +0200, Ingo Molnar wrote:
> * Ingo Molnar <mingo@elte.hu> wrote:
> 
> > could you try:
> > 
> > 	strace -o log sleep 10
> > 
> > and wait for a failure, and send us the log? Is it perhaps nanosleep 
> > unexpectedly returning with -EAGAIN or -512? There's a transient 
> > nanosleep failure that happens on really fast boxes, which we havent 
> > gotten to the bottom yet. That problem is very sporadic, but maybe 
> > your box is just too fast and triggers it more likely :-)
> 
> is nanosleep returning -ERESTART_RESTARTBLOCK (-516) perhaps?

Yes, that is probably correct, I saw a few of these:
  sleep: cannot read realtime clock: Unknown error 516

I left the kernel running overnight and this morning I've been using it
with no key repeats or screen blanks so far. Go figure. I imagine they
would come back if I reboot. I'm still seeing the Jack warnings but
that's probably a related but somewhat separate issue. 

I'm building rt5 and will report later. 

-- Fernando