Subject: Re: writing crashdump
To: Mipam <mipam@ux11.ltcm.net>
From: Pavel Cahyna <pavel@netbsd.org>
List: netbsd-users
Date: 05/23/2006 19:32:43
On Tue, May 23, 2006 at 07:27:52PM +0200, Mipam wrote:
> On Tue, 23 May 2006, Pavel Cahyna wrote:
> 
> > On Tue, May 23, 2006 at 04:25:42PM +0200, Mipam wrote:
> > > Hi All,
> > > 
> > > I wished to generate a crashdump.
> > > After rebooting with the kernel that crashes and ending up in db>
> > > I type: reboot 0x100
> > > However, i then see that it's unable to write the crashdump and reboots, 
> > > so i cannot get a crashdump, to make a backtrace later.
> > > Do i need to specify somewhere to write the crashdump to swap or?
> > > And how to do this?
> > 
> > Please specify more precisely what happens after you typed "reboot 0x100".
> > "it's unable to write" does not tell much.
> 
> Unable to write to dev 1,0 is what the message displays.
> This is all it mentiones. :-(

when the kernel starts, it should print something like
root on sd0a dumps on sd0b

the "dumps on" part tells what partition it will use to write crash dumps.
Usually this is the swap partition.

Please verify that this partition exists and is large enough for the dump
(larger than the amount of RAM).

Pavel