Der Mouse already has such a patch.

Please note the context of the word "trivial", i.e. "in comparison to
the code already there."  From what I can divine from my admittedly
limited knowledge, all it requires (from a pseudo-code point of view)
is changing from:

	if (i->refcount > 0) {
		reconnect(i);
		adjust_link_count(i);
	}

to

	if ((refcount > 0) || zflag) {
		reconnect(i);
		adjust_link_count(i);
	}

Unless I'm really missing the boat on this one; I think I'm close enough
to at least smack my face on its gunwales as I leap haplessly from the
pier...

In NO WAY am I trying to dismiss the complexity of the code; perhaps
"trivial" and "context" don't mix.  Sorry if I'm out of line on this.

On Mon, 10 Jun 2002, gabriel rosenkoetter wrote:

# Date: Mon, 10 Jun 2002 11:19:07 -0400
# From: gabriel rosenkoetter <gr@eclipsed.net>
# To: NetBSD Kernel Technical Discussion List <tech-kern@netbsd.org>
# Subject: Re: FFS reliability problems
#
# On Mon, Jun 10, 2002 at 08:02:38AM -0700, Greywolf wrote:
# > 2.  The fix for this is, in comparison to the code already there, trivial.
#
# So when will we be seeing your patch?
#
# --
# gabriel rosenkoetter
# gr@eclipsed.net
#


				--*greywolf;
--
NetBSD: safe ports in a storm.