Re: [PATCH] nfsd: just keep single lockd reference for nfsd

[Chris Vine <chris-TF6qbakwsgc2epGFuHBODCp2UmYkHbXO@public.gmane.org>]

On Fri, 18 Jun 2010 11:11:32 -0400
Jeff Layton <jlayton@redhat.com> wrote:
> On Fri, 18 Jun 2010 16:04:11 +0100
> Chris Vine <chris-TF6qbakwsgc2epGFuHBODCp2UmYkHbXO@public.gmane.org> wrote:
> 
> > On Fri, 18 Jun 2010 10:30:42 -0400
> > Jeff Layton <jlayton@redhat.com> wrote:
> > [snip]
> > > Thanks for testing them. No oops == improvement!
> > > 
> > > ...but it would still be good to know what's wrong here. It sounds
> > > like something is really odd with loopback communications on this
> > > box. Is the ipv4 loopback interface up at this time? Do you have
> > > any iptables stuff set up that might be filtering out portmap
> > > registration requests from the kernel? What happens if you run
> > > "rpcinfo"? Does it also hang?
> > > 
> > > The kernel uses TCP for talking to portmap these days, so it might
> > > also be good to see whether you can use rpcinfo to talk to it with
> > > TCP too...
> > 
> > You are right. Although the lo interface is up, localhost is broken
> > in 2.6.35-rc3. External networking works OK which is why I hadn't
> > noticed.
> > 
> > It looks like some other unrelated bug in 2.6.35 is at work. Quite
> > why it doesn't strike my Pentium machine I can't say.  The netbook
> > takes its network address from DHCP (my Pentium does not) but that
> > shouldn't make any difference to localhost.
> > 
> > I shall do a bit more looking around later today or tomorrow.
> > 
> > Chris
> > 
> > 
> 
> Ok, good to know. I'll still plan to pursue these fixes. The error
> handling in nfsd obviously an area that needs improvement. I think
> Bruce's current plan is to take the earlier patches in that stack for
> 2.6.36. If this one is OK, it's probably reasonable to go in at the
> same time.

The problem is indeed a loopback bug and for anyone else suffering this
with nfsd, the fix is a one liner, here:
http://marc.info/?l=linux-kernel&m=127646140827821

At least this had the benefit of improving the error handling of nfsd!

Chris