diff --git a/net/rds/connection.c b/net/rds/connection.c
index 1fa75ab7b7332..6a5ebdea7d2e9 100644
--- a/net/rds/connection.c
+++ b/net/rds/connection.c
@@ -333,11 +333,19 @@ void rds_conn_shutdown(struct rds_conn_path *cp)
 		rds_conn_path_reset(cp);
 
 		if (!rds_conn_path_transition(cp, RDS_CONN_DISCONNECTING,
+					      RDS_CONN_DOWN) &&
+		    !rds_conn_path_transition(cp, RDS_CONN_ERROR,
 					      RDS_CONN_DOWN)) {
 			/* This can happen - eg when we're in the middle of tearing
 			 * down the connection, and someone unloads the rds module.
-			 * Quite reproduceable with loopback connections.
+			 * Quite reproducible with loopback connections.
 			 * Mostly harmless.
+			 *
+			 * Note that this also happens with rds-tcp because
+			 * we could have triggered rds_conn_path_drop in irq
+			 * mode from rds_tcp_state change on the receipt of
+			 * a FIN, thus we need to recheck for RDS_CONN_ERROR
+			 * here.
 			 */
 			rds_conn_path_error(cp, "%s: failed to transition "
 					    "to state DOWN, current state "
diff --git a/net/rds/threads.c b/net/rds/threads.c
index e36e333a0aa0d..3e447d056d092 100644
--- a/net/rds/threads.c
+++ b/net/rds/threads.c
@@ -156,7 +156,7 @@ void rds_connect_worker(struct work_struct *work)
 	struct rds_connection *conn = cp->cp_conn;
 	int ret;
 
-	if (cp->cp_index > 1 && cp->cp_conn->c_laddr > cp->cp_conn->c_faddr)
+	if (cp->cp_index > 0 && cp->cp_conn->c_laddr > cp->cp_conn->c_faddr)
 		return;
 	clear_bit(RDS_RECONNECT_PENDING, &cp->cp_flags);
 	ret = rds_conn_path_transition(cp, RDS_CONN_DOWN, RDS_CONN_CONNECTING);