Chri writes:
>the problem is not with mounting the FS read-write.  if fsck
>makes this modification to the superblock, and the modification
>gets written out, then a future _fsck_ might fail...

Um. But if NetBSD isn't  meant to be mounting 4.2 ffs filesystems
with "fixed" read-write, then its fsck shouldn't be writing back
modified superblocks to unconverted to 4.2 ffs filesystems.
Should it?

If it should, then an obvious, ugly fix is to add a
global variable to fsck that holds the size fsck
thinks is the maximum length of a symlink...



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