Re: kern/15421 (vnode VOP()erations handle SAVESTART inconsistently)

[Chris Jepeway <jepeway%blasted-heath.com@localhost>]

> State-Changed-From-To: open->feedback
> State-Changed-By: pooka%narn.netbsd.org@localhost
> State-Changed-When: Sat, 19 Jan 2008 22:17:32 +0200
> State-Changed-Why:
> While the PR is in principle correct, SAVESTART is unlikely to get
> any divine interventions with the current lookup labyrinth and I'd
> like to close this PR.  Do you want to use SAVESTART for something
> besides rename?
As I recall (mostly from what I wrote in the PR),
I was trying to get a layered f/s working that
relied on SAVESTART.  However, since the UFS VOPs
don't use it consistently, and other f/s's don't
follow UFS's inconsistencies, I gave up.

Now, *why* that layered f/s cared about SAVESTART,
I just don't remember.  I mentioned ufs_rename only
to point out some gnarly use of SAVESTART that I
couldn't follow.

As for closing the PR because you want to,
well, I'm not doing kernel work anymore, so I
no longer have a practical use for a fix to this bug.

So...however you like, so far as I'm concerned.

Thanks, though, for following up!

Chris.