(By always marking data in use before it's used and ensuring that
data is not used before it's marked free, a crash can leave some
blocks or inodes missing from the free map, and inodes with
reference counts too high, so fsck is a good idea for garbage
collection in case of a crash, but it's very optional - you can
bring up the system read/write and only suffer a little bit of
"leaked" disk space.)
The code is heavily fuzzed where SYSV ufs internals are concerned
to avoid trade secret problems, but it's just replaced by stubs
with english descriptions of what they do.
I'm not saying that "somebody oughta", because if so, then *I* oughta,
but in case this is of interest to anybody, the URL is
http://www.pdos.lcs.mit.edu/~ganger/papers/CSE-TR-254-95/
-- -Colin