Re: CLOSE/OPEN race

[Jeff Layton <jlayton@redhat.com>]

On Sat, 2016-11-12 at 18:16 +0000, Trond Myklebust wrote:
> > On Nov 12, 2016, at 06:08, Benjamin Coddington <bcodding@redhat.com> wrote:
> > 
> > I've been seeing the following on a modified version of generic/089
> > that gets the client stuck sending LOCK with NFS4ERR_OLD_STATEID.
> > 
> > 1. Client has open stateid A, sends a CLOSE
> > 2. Client sends OPEN with same owner
> > 3. Client sends another OPEN with same owner
> > 4. Client gets a reply to OPEN in 3, stateid is B.2 (stateid B sequence 2)
> > 5. Client does LOCK,LOCKU,FREE_STATEID from B.2
> > 6. Client gets a reply to CLOSE in 1
> > 7. Client gets reply to OPEN in 2, stateid is B.1
> > 8. Client sends LOCK with B.1 - OLD_STATEID, now stuck in a loop
> > 
> > The CLOSE response in 6 causes us to clear NFS_OPEN_STATE, so that the OPEN
> > response in 7 is able to update the open_stateid even though it has a lower
> > sequence number.
> 
> Hmm… We probably should not do that if the stateid.other field of A (i.e. the one supplied as the argument to CLOSE) does not match the stateid.other of B.
> In fact, the reply in (4), where the stateid changes to B, should be the thing that resets the OPEN state.NrybXǧv^)޺{.n+{"^nrz\x1ah&\x1eGh\x03(階ݢj"\x1a^[mzޖfh~m

It looks like that's already the case in nfs_clear_open_stateid_locked,
though I don't think we ought to be doing anything with the stateid in
the CLOSE response. I sent a draft patch in another part of this
thread, but I don't quite see how that would cause this problem.
-- 
Jeff Layton <jlayton@redhat.com>