You've probably figured this out already.  But,
since I just ran into the same thing, I thought
I'd comment and maybe save others potential trouble.

The problem(at least in my case) was running
a newer kernel with the UFS2 support and
an older /sbin/fsck_ffs.  To fix I had to replace
/sbin/fsck_ffs with a copy from the latest
"base.tgz" set.  After that, "fsck" ran perfectly
and the system booted no problem...

See Frank Van Der Linden's mail to current-users
list on April 15th(and numerous prior dates) for
further explanation.  There are various ways the
UFS code changes can trip one up, depending
on what vintage your binaries, kernel, and source
are and whether they match...

Steve Brown

Reinier Jonker wrote:

> After an unexplanatory freeze on my Quadra 840 running 1.6R, I had to 
> manually reset the machine. Since this has happened once or twice 
> before, I expected fsck to be run and be finished in half an hour. 
> Unfortunately, fsck immediately exited reporting:
>
> dev/rsd1a: BAD SUPER BLOCK: VALUES IN SUPER BLOCK DISAGREE WITH THOSE 
> IN FIRST
> ALTERNATE
>
> The filesystem appeared to function normally. I read the man page for 
> fsck (couldn't find anything) and then the one for fsck_ffs. I found 
> out fsck_ffs -d would give me diagnostic reports. I saw some values 
> were replaced by noughts in the alternate super block.
>
> I can't get fsck(_ffs) to repair this problem. Help! ;-)
> Thanks very much in advance.
>