Jesse Off <joff@gci-net.com> writes:

> Well, I've been using it for awhile now, without a hitch.  I haven't
> actually tested whether or not its working though. 

I have become a convert since playing around with this,
after the discussion was raised here.

See Keith Smith & Margo Seltzer's paper on the topic at
http://www.eecs.harvard.edu/~keith/papers/realloc.ps.gz

> What were the known bugs that were present in Lite so as
> to determine whether this mysterious sysctl is indeed
> safe to be set?  If not, it probably should be
> documented somewhere or removed altogether.

I found some complaints from 1994 by doing google on
"doreallocblks" from NetBSD and FreeBSD types, however I
have not seen any bad symptoms, and I have seen a
performance increase.

(It would be nice to be able to expose the fs_maxcluster
information though, I think...)

        Sean.