It is production.  This week these systems go to 2.4.21-ac4 with the
2.8.0 i2c.  I only think it's not a straight up kernel bug because there
are 35ish identicle systems running all the same software with 0
problems.  In addition for some odd reason this one machine was giving
some very odd temp readings.


Thus spake Zwane Mwaikambo (zwane@arm.linux.org.uk):

> On Tue, 5 Aug 2003, Robert L. Harris wrote:
> 
> > Code:  Bad EIP value.
> > 
> > >>EIP; 8011c560 Before first symbol   <=====
> > Trace; c011c47d <tasklet_hi_action+5d/90>
> > Trace; c011c1ff <do_softirq+6f/d0>
> > Trace; c0108bdb <do_IRQ+db/f0>
> > Trace; c0105400 <default_idle+0/40>
> > Trace; c0105400 <default_idle+0/40>
> > Trace; c010ae78 <call_do_IRQ+5/d>
> > Trace; c0105400 <default_idle+0/40>
> > Trace; c0105400 <default_idle+0/40>
> > Trace; c010542c <default_idle+2c/40>
> > Trace; c01054a2 <cpu_idle+42/60>
> > Trace; c0117e7f <release_console_sem+8f/a0>
> > Trace; c0117d8e <printk+11e/140>
> > 
> >  <0>Kernel panic: Aiee, killing interrupt handler!
> 
> It looks like someone freed a tasklet without removing it. But considering 
> your kernel cocktail (imported i2c code) it makes it harder for us to 
> debug, perhaps if you could try on a newer kernel (i know it'd be hard to 
> do if it's production)
> 
> -- 
> function.linuxpower.ca

:wq!
---------------------------------------------------------------------------
Robert L. Harris                     | GPG Key ID: E344DA3B
                                         @ x-hkp://pgp.mit.edu
DISCLAIMER:
      These are MY OPINIONS ALONE.  I speak for no-one else.

Life is not a destination, it's a journey.
  Microsoft produces 15 car pileups on the highway.
    Don't stop traffic to stand and gawk at the tragedy.