NFSv4.1: freeze the session table upon receiving NFS4ERR_BADSESSION

When the client received NFS4ERR_BADSESSION, it schedules recovery
and start the state manager thread which in turn freezes the
session table and does not allow for any new requests to use the
no-longer valid session. However, it is possible that before
the state manager thread runs, a new operation would use the
released slot that received BADSESSION and was therefore not
updated its sequence number. Such re-use of the slot can lead
the application errors.

Fixes: 5c44154 ("NFSv4.x: Handle bad/dead sessions correctly in nfs41_sequence_process()")
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>

fs/nfs/nfs4proc.c

@@ -921,6 +921,7 @@ static int nfs41_sequence_process(struct rpc_task *task,
 out_noaction:
 	return ret;
 session_recover:
+	set_bit(NFS4_SLOT_TBL_DRAINING, &session->fc_slot_table.slot_tbl_state);
 	nfs4_schedule_session_recovery(session, status);
 	dprintk("%s ERROR: %d Reset session\n", __func__, status);
 	nfs41_sequence_free_slot(res);