>For instance, it's unclear to me why fsck would claim the symlinks
>were two blocks long; that's a pretty long symlink!

Because it`s an 8k/1k filesystem, and the filesystem cannot allocate
less than a fragment??


>If you fsck with that patch in, and a problem is corrected,
>and the superblock is written out, the wrong data (i.e. 0, rather than
> -1) will be written out for the fs_maxsymlinklen field.

Yes, but I thought NetBSD wasn't ever meant to mount 4.2 filesystems
read-write, so this shouldn't be a problem. That patch I called
"GROSS" only changes fs_maxsymlinklen in the superblock when fsck'ing
a filesystem with fs_postblformat equal to FS_42POSTBLFMT.  This
should only happen when converting ffs filesystems with fsck -c 1 (or
greater.)  In that case, fsck conveniently sets fs_maxsymlinklen to an
appropriate value.






------------------------------------------------------------------------------