On Mon, May 16, 2011 at 03:43:40PM -0400, bjschuma@xxxxxxxxxx wrote:
> +static __be32
> +nfsd4_free_file_stateid(stateid_t *stateid)
> +{
> + struct nfs4_stateid *stp = search_for_stateid(stateid);
> + if (!stp)
> + return nfserr_bad_stateid;
> + if (stateid->si_generation != 0) {
> + if (stateid->si_generation < stp->st_stateid.si_generation)
> + return nfserr_old_stateid;
> + if (stateid->si_generation > stp->st_stateid.si_generation)
> + return nfserr_bad_stateid;
> + }
> +
> + if (check_for_locks(stp->st_file, stp->st_stateowner))
> + return nfserr_locks_held;
I think this catches a lock stateid, but not an open stateid that has
associated lock stateid's that in turn hold locks.
Hm, also:
"The FREE_STATEID operation is used to free a stateid that no
longer has any associated locks (including opens, byte-range
locks, delegations, and layouts)"
So an open stateid also shouldn't be freeable as long as there are opens
associated with it.
Also I guess a client shouldn't be permitted to free a delegation that
it still holds. That means we'll always just return nfserr_locks for
delegation stateid's. I assume free_stateid is only useful in this case
for the case where a server forcibly revokes part of the client's state
and leaves some "stub" stateid's around in place of the revoked ones.
--b.
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
