Re: Vnode API cleanup pass 2a

[David Holland <dholland-tech%netbsd.org@localhost>]

On Wed, Jan 15, 2014 at 04:31:07PM +0100, J. Hannken-Illjes wrote:
 > I put a diff to http://www.netbsd.org/~hannken/vnode-pass2a-3.diff that
 > changes the vnode creation operations to keep the dvp locked.
 > 
 > Any objections or OK to commit?

I don't understand where all of the zfs changes arise (without wading
into the zfs code) but I'm not going to worry too much about that.

Otherwise it looks ok. I'm kind of distressed by how far afield some
of the vput calls are from the vop entry points in some of the fses,
but I suppose that's the usual level of code quality. :-/

I suppose there might be a problem in one of those cases where
leaving the directory locked causes something else wrong to happen.
But if so I think I'd rather tackle it head on.

(Function calls should be symmetric with respect to locking, all the
more so in what's ostensibly a supported API.)

Also I'm not sure that if there were something missing I'd necessarily
spot it from reading a diff, but that can't be helped.


Now somebody just has to fix the rest of the vops that do this...
maybe this weekend I can finally deal with the fix for PR 47040 and
then I can stop being wedged on vfs work.

-- 
David A. Holland
dholland%netbsd.org@localhost