Subject: Re: problem with fsck_ffs
To: Thorsten Frueauf <s_frueau@ira.uka.de>
From: Frederick Bruckman <fredb@fb.sa.enteract.com>
List: port-pmax
Date: 04/23/1998 13:40:06
On Thu, 23 Apr 1998, Thorsten Frueauf wrote:

> Hello!
> 
> > I seem to remember a warning somewhere, that fsck will choke and die on
> > trying to fix certain problems if it doesn't have any swap space.  That
> > would normally be the case when it's started by "rc." Perhaps you could
> > try `swapctl -A ; fsck -p' from single user mode.
> 
> I had time to try this out. The swap partition was already added at that
> point, swapctl -l listed it properly, so this did not help me :-(
> 
> My problem with fsck still remains, I was not able to get the fs clean
> yet. Maybe I should send-pr it.

So much for that idea. I guess that change was made in rc some time ago. 
Interestingly enough, I just came across a similar problem on my system. 
fsck was dumping core after the very last question. I forget, but it was
something like "...fix summary info." ``fsck_ffs -f -b 32 /dev/rsdxx''
worked for me.  You might want to check out the manpage for "fsck_ffs"; it
lists more options than generic "fsck", even though fsck invokes fsck_ffs
anyway.

I doubt that your problem would be considered a bug in fsck. There's only
so much you can expect it to do. The ultimate solution would be to
dump|tar|pax that whole filesystem, then newfs and restore it. Good luck.
:-)