On Fri, 19 Aug 2005 15:43:21 +0200 (CEST)
"Timo Schoeler" <timo.schoeler@riscworks.net> wrote:
> > | that's why i wrote 'wrong configured RAIDframe sets' -- the machine is
> > | colocated, i'm neither able nor willing to risk reconfiguration stuff
> > | (except for e.g. building a tweaked kernel and reboot)... ;)
> >
> > Ah, okay. Then comment out the dumpsys() code on i386/i386/machdep.c.

Eek!  Perhaps there is a smiley missing here.  Given his situation and
the conservatism that he expresses, I would say that messing with
source and rebuilding the kernel has to be at the bottom of the list of
things to try.  Building a new kernel with the config changes described
in other posts would probably be #2 on my list.  #1 would be adding the
following lines to /etc/sysctl.conf and rebooting.  In fact, a reboot
is not strictly necessary as you can run "/etc/rc.d/sysctl restart"
instead.  I generally like to reboot if it is an option just to make
sure that something can't go wrong if it reboots for any reason while
no one is around to fix it.

kern.dump_on_panic=0
ddb.onpanic=0

That would be #1 IMO simply because he has a working kernel and this
does not risk messing up the compile and winding up with a new kernel
that does not behave exactly like the existing one.

-- 
D'Arcy J.M. Cain <darcy@NetBSD.org>
http://www.NetBSD.org/