Re: [PATCH 1/2] NFS: Fix a number of bugs in the idmapper

["J. Bruce Fields" <bfields@fieldses.org>]

On Tue, Jul 31, 2012 at 03:07:42PM +0000, Myklebust, Trond wrote:
> On Tue, 2012-07-31 at 15:56 +0100, David Howells wrote:
> > Myklebust, Trond <Trond.Myklebust@netapp.com> wrote:
> > 
> > > This patch should already be present in nfs-for-next as of yesterday.
> > > (I've been on a 2 week vacation).
> > 
> > How about the second patch?  I posted an updated version of it just now.
> > 
> > Note that the second patch or something like it is likely to be necessary to
> > reduce the load on the keyring (you end up having duplicate keys of different
> > key types that can't be shared).
> 
> We don't expect people to be running both types of idmapper at the same
> time. Normally, the distribution will set up one or the other. However
> if both idmappers do happen to be configured, then we want to always
> choose the non-legacy type...

On a quick check....  It looks like F17 has /usr/sbin/nfsidmap, and also
runs rpc.idmapd for the server.  While rpc.idmap does have a "-S"
option, it's not using that.

But that should be easy to fix--I'll go look....

--b.