Repeated panics...Suspect USB issues

[Doug White]

On Fri, 11 Mar 2005, Josh Tolbert wrote:

> > A backtrace would be nice. See the developer's handbook on kernel
> > debugging for details.
>
> I'm trying to get a backtrace. I've built a kernel with all the right
> bits, set up dumpdir and dumpdev in /etc/rc.conf, etc...Apparently
> savecore doesn't like writing to the swap partition (/dev/ar0s1b). I can
> get to the ddb promp, obtain a trace, panic the machine, etc...But I
> don't think I can obtain a good core without putting another hard drive
> in the machine, which isn't going to happen till sometime next week...Is
> there anything I can provide from the ddb prompt?

savecore *reads* from the swap partition, not *writes* to it. It will drop
the recovered core in /var/crash unless you tell it otherwise.

The developer's handbook has a command list for ddb. "trace" is generally
what we want.

-- 
Doug White                    |  FreeBSD: The Power to Serve
dwhite at gumbysoft.com          |  www.FreeBSD.org