On Wed, Feb 02, 2005 at 11:07:14PM +0100, Paul Kranenburg wrote:
> > I started seeing this on a dual-CPU ss20 running a 2.0_BETA kernel:
> > cpu0: bogus interrupt ipl 0xa pc=0xf00c442c npc=0xf00c4430 psr=400006c3<S,PS>
> > cpu0: bogus interrupt ipl 0xa pc=0xf00c4420 npc=0xf00c4424 psr=404006c3<S,PS>
> > 
> > Any idea what this means ?
> > 
> 
> It means that the cpu took the interrupt (at level 10, the system clock)
> but the interrupt register showed no interrupt was pending.
> 
> I've no idea why this can happen as currently only one cpu is designated
> to field device interrupts. Maybe it's useful to look at what code is
> at the PCs mentioned above.

All these addresses are in acquire(). I don't have the source code
for this kernel any more, so I don't know exactly where is it.

I also noticed a:
xcall(cpu1,0xf01754b8): couldn't ping cpus
on the console, which didn't make it though syslog (so I don't know
if it happended just after the bogus interrupt messages, or some time later).

-- 
Manuel Bouyer <bouyer@antioche.eu.org>
     NetBSD: 26 ans d'experience feront toujours la difference
--