Date: Wed, 29 Jun 2005 19:12:29 -0400 (EDT) From: Matt Juszczak <matt@atopia.net> To: Kris Kennaway <kris@obsecurity.org> Cc: Gleb Smirnoff <glebius@freebsd.org>, freebsd-stable@freebsd.org Subject: Re: FreeBSD -STABLE servers repeatedly crashing. Message-ID: <20050629191215.I34468@neptune.atopia.net> In-Reply-To: <20050629100535.GC27557@xor.obsecurity.org> References: <42BF8815.6090909@atopia.net> <20050627081933.GA97832@cell.sick.ru> <42C16394.4040904@atopia.net> <1119971279.36316.45.camel@buffy.york.ac.uk> <42C16C0E.9090002@atopia.net> <20050629100535.GC27557@xor.obsecurity.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 29 Jun 2005, Kris Kennaway wrote: > On Tue, Jun 28, 2005 at 11:26:06AM -0400, Matt Juszczak wrote: > >>> OK, when it crashes next and is sat at the "db>" prompt, type "tr" and >>> press enter to get a trace. Copy this down (or have a serial console to >>> capture the output). Also, try typing "call doadump()" and see if that >>> succeeds in generating a crash dump. How were you trying to generate >>> one before? >>> >>> Gavin >>> >>> >> >> I can't type anything. The machine locks up. >> >> See: http://paste.atopia.net/126 >> >> After CPUID: 1, the machine locks cold and nothing else is printed to >> the screen. > > Try two things: > > 1) adding 'options KDB_STOP_NMI' to your kernel config. > > 2) If you still can't get it to break to DDB, then compile up a > debugging kernel, run kgdb on it (as described in the developers' > handbook), and list *(0xblah) where that address is the value of the > instruction pointer in the trap message (e.g. 0xc6644eff in your paste > above). That might at least be a start. > > Kris OK :) I'll try this next time it crashes. I actually disabled ipf a few nights ago and it hasn't crashed since... knock on wood.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050629191215.I34468>