Subject: Re: panic: lfs_truncate: negative bytes
To: Konrad Schroder <firstname.lastname@example.org>
From: Sean Davis <email@example.com>
Date: 06/27/2002 14:22:25
On Thu, Jun 27, 2002 at 11:11:58AM -0700, Konrad Schroder wrote:
> On Thu, 27 Jun 2002, Sean Davis wrote:
> > Any ideas? Should I recreate my LFS volume?
> > I was doing a build.sh, with sources and obj on lfs.
> > I can make the core/kernel available if needed.
> If "fsck_lfs -n -d /dev/rfoo" says there is nothing wrong with your disk
> (in particular, no "block in 'clean' segment" or "segment X contains Y,
> low by Z" problems), it might be useful to look at the kernel core,
> otherwise probably not.
I just recreated the volume, so I didn't have a chance to try that.
> How old was this volume, and when was the last time you updated your
> kernel and lfs_cleanerd?
Not sure, at least a couple months old, and the kernel was last updated
yesterday. lfs_cleanerd updated yesterday as well, and it still paniced
doing some things. After I remade the lfs filesystem, everything seems fine.
I've also stopped seeing the lfs_markv messages.
> If you decide to rebuild, I'd suggest using frag size == block size (say,
> 4k/4k) until I fix lfs_markv to handle expanding fragments properly.
> With a new kernel you shouldn't really have a problem but ... I'm not
I just used newfs_lfs /dev/rwd1a, without any special block/frag size
options, so I'll see how it goes. So far so good.
> Konrad Schroder
/~\ The ASCII Sean Davis
\ / Ribbon Campaign aka dive
X Against HTML
/ \ Email! http://endersgame.net/~dive/