Re: [patch] Real-Time Preemption, -RT-2.6.12-rc6-V0.7.48-00

[William Weston <weston@sysex.net>]

On Sat, 25 Jun 2005, Ingo Molnar wrote:

> > > do you have the NMI watchdog enabled? Find below 
> > > serial-logging-earlyprintk-nmi.txt.
> > 
> > I had a serial console up, but not NMI watchdog until now.  Here's 
> > some NMI watchdog traces from both -50-17 and -50-18:
> 
> all of these traces seem to have lockupcli involved - is that correct?  
> lockupcli is just a userspace test-app to artificially trigger a 
> hard-lockup (it disables interrupts and goes into an infinite loop). So 
> the NMI watchdog triggering on lockupcli would be normal and expected.  
> So once this works, it would be nice to reproduce whatever hard lockup 
> you are seeing and see whether the NMI watchdog produces any output to 
> the serial log. (if such log is supposed to be included in your dmesg 
> then it somehow got intermixed with lockupcli logs)

I still haven't been able to get any NMI watchdog traces with the lockups
induced by VLC and burnP6.  Early printk is enabled on the serial console.
I have noticed, however, that scheduling performance slowly degrades 
during the ~1 minute before locking up.  Once I was able to get a delayed 
SysRq response (~30s) from the serial console after the X console became 
unresponsive.  Is this possibly a scheduler starvation issue that affects 
everything, including the NMI watchdog?  Any more suggestions for catching 
a trace?

--ww
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/