I've recently upgraded my machine from 2.0.2 to 3.0 and have been having
the kernel spit the dummy when under load (actually when doing a bulk
package build or, strangely, when doing a pkgsrc CVS update once).

It used to do the same with 2.0.2, so I don't suspect the new version of
NetBSD. Given these messages I think it may be a problem with memory.
Here's a (hand transcribed) copy of the messages that appear on the
console:

|uvm_fault(0xcc191b6c, 0, 0, 1) -> 0xe
|kernel: page fault trap, code = 0
|Sopped in pid 6374, 1 (make) at netbsd:getcwd_common+0x5d:  mov1  0(%edx),  %eax
|db> sync
|syncing disks...ex0: uplistprt was 0
|16 23 done
|unmounting file systems...panic: lockmgr: locking against myself
|Stopped in pid 6374,1 (make) at netbsd:cpu_Debugger+0x4:  leave
|db> reboot

Looking at that fragment, should I suspect memory or hard-drive? The two
disks in the machine are mirrored using RAIDFrame which hasn't given any
errors that I've noticed - of course when I reboot it has to resync the
mirror because it crashed, but I would have thought it would have made
more noise beforehand if it was a problem with the disks.

Oh, and swap is also on the RAID device, if that makes a difference.

I also have recently had the machine lock up solid whenever I run lame,
even to the point where it doesn't produce any console messages at
all. I'd updated some of the packages on there and may have gotten it
out of whack, so I was hoping to complete this run of bulk package
builds before trying that again, but now I'm wondering whether it may be
related ...

Thanks in advance,
Malcolm

-- 
Malcolm Herbert                                This brain intentionally
mjch@mjch.net                                                left blank