Subject: Re: syslogd restart triggers a kernel trap?
To: Tom Haapanen <tomh@metrics.com>
From: Charles M. Hannum <root@ihack.net>
List: port-alpha
Date: 03/03/2001 12:24:22
On Sat, Mar 03, 2001 at 02:29:43PM -0500, Tom Haapanen wrote:
> Charles,
> 
> >> Mar  3 01:44:18 omega syslogd: restart
> >> Mar  3 01:44:18 omega /netbsd:
> >> Mar  3 01:44:18 omega /netbsd: fatal kernel trap:
> > ...
> >> Mar  3 01:44:19 omega /netbsd: dumping to dev 4,1 offset 263023
> >> Mar  3 01:44:19 omega /netbsd: dump 128 127 126 125 124 123 122 121 120
> 119
> >> 118
> 
> > Since it dumped, you should be able to inspect the core dump and find
> > out where it crashed.  A stack trace would be most helpful.
> 
> Never having partaken in kernel debugging before, how do I get gdb to read
> the core dump (so that I can get a stack trace)?  It looks like it was
> written to /dev/wd0b, which is a 256 MB swap partition (the system has 128
> MB of physical memory).  I trid specifying -c /dev/wd0b for gdb, but it
> claimed it could not recognize the core file.

Assuming you have enough space in /var/crash, savecore will normally
save the dump there at boot time.  You can look at it with:

cd /var/crash
gdb netbsd.12
(gdb) target kcore netbsd.12.core
(gdb) bt

Etc.