Hi,

On Tue, Jan 11, 2005 at 07:43:15PM +0100, Manuel Bouyer wrote:
> > Every night between 2 and 5 a.m. - usually at a time with *NO* major
> > activity going on at all (!) - the machine dies.  Sometimes, there is
> > no console message at all, sometimes it prints this:
> > 
> > -------------- snip ---------------
> > panic: lockmgr: no context
[..]
> > Any ideas what might be triggering it?  And (even better) how to avoid
> > triggering it?
> 
> This could be the pmap bug which was fixed recently. Update from CVS using
> the netbsd-2-0 tag, and build a new kernel.

I missed that last change to pmap.c by one day :-) - I did have the
two-line patch that was posted in here (moving blast_dcache()), but the
now-netbsd-2-0-pmap.c has some more fixes now.

I've just built and booted a new kernel, and will report tomorrow whether
it crashed (and if yes, whether I could type "bt" in ddb and get a back
trace, as has been recommended to me today).

Also I'm running a "sysstat vmstat 1" in a ssh window right now, so
maybe that will tell me what the machine did before it died.  "top" didn't...

gert

-- 
USENET is *not* the non-clickable part of WWW!
                                                           //www.muc.de/~gert/
Gert Doering - Munich, Germany                             gert@greenie.muc.de
fax: +49-89-35655025                        gert@net.informatik.tu-muenchen.de