On Fri Apr 06 2007 at 18:00:47 -0700, Bill Stouder-Studenmund wrote:
> > so I would think that the vnop implementation should do VOP_UNLOCK on
> > dvp.  But the coda coda does vput, and puffs does too.
> > 
> > I expected that the caller of VOP_CREATE would hold a ref, and that
> > VOP_CREATE would ride that ref, so it seems odd to release it.
> > 
> > What's the right behavior, and should vnodeops(9) be fixed?
> 
> Look in vnode_if.src. Unfortunately vnodeops(9) has been wrong since day 
> 1.

Technically it is correct, the node *is* unlocked ;)

Yes, vnodeops(9) should be fixed.  But maybe it should be fixed by
writing a common paragraph for all the operations which create a node
(create, mkdir, mknod, symlink) and putting a "see also" reference to
that paragraph in the description of each respective op.  Looks like
they already have copypasted duplicates for the bit about PNBUF_PUT.

-- 
Antti Kantee <pooka@iki.fi>                     Of course he runs NetBSD
http://www.iki.fi/pooka/                          http://www.NetBSD.org/
    "la qualité la plus indispensable du cuisinier est l'exactitude"