nfsd: don't abort copies early

The typical result of the backwards comparison here is that the source
server in a server-to-server copy will return BAD_STATEID within a few
seconds of the copy starting, instead of giving the copy a full lease
period, so the copy_file_range() call will end up unnecessarily
returning a short read.

Fixes: 624322f "NFSD add COPY_NOTIFY operation"
Signed-off-by: J. Bruce Fields <[email protected]>
Signed-off-by: Chuck Lever <[email protected]>

fs/nfsd/nfs4state.c

@@ -5389,7 +5389,7 @@ nfs4_laundromat(struct nfsd_net *nn)
 	idr_for_each_entry(&nn->s2s_cp_stateids, cps_t, i) {
 		cps = container_of(cps_t, struct nfs4_cpntf_state, cp_stateid);
 		if (cps->cp_stateid.sc_type == NFS4_COPYNOTIFY_STID &&
-				cps->cpntf_time > cutoff)
+				cps->cpntf_time < cutoff)
 			_free_cpntf_state_locked(nn, cps);
 	}
 	spin_unlock(&nn->s2s_cp_lock);