nfsd: handle NFS4ERR_BADSLOT on CB_SEQUENCE better

Currently it just restarts the call, without getting a new slot.

Signed-off-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>

Author: jtlayton

fs/nfsd/nfs4callback.c

@@ -1389,6 +1389,13 @@ static bool nfsd4_cb_sequence_done(struct rpc_task *task, struct nfsd4_callback
 		rpc_delay(task, 2 * HZ);
 		return false;
 	case -NFS4ERR_BADSLOT:
+		/*
+		 * BADSLOT means that the client and server are out of sync
+		 * as to the backchannel parameters. Mark the backchannel faulty
+		 * and restart the RPC, but leak the slot so no one uses it.
+		 */
+		nfsd4_mark_cb_fault(cb->cb_clp);
+		cb->cb_held_slot = -1;
 		goto retry_nowait;
 	case -NFS4ERR_SEQ_MISORDERED:
 		if (session->se_cb_seq_nr[cb->cb_held_slot] != 1) {