BUG/MAJOR: leastconn: do not loop forever when facing saturated servers

Since commit 9fe72bba3 ("MAJOR: leastconn; Revamp the way servers are
ordered."), there's no way to escape the loop visiting the mt_list heads
in fwlc_get_next_server if all servers in the list are saturated,
resulting in a watchdog panic. It can be reproduced with this config
and injecting with more than 2 concurrent conns:

    balance leastconn
    server s1 127.0.0.1:8000 maxconn 1
    server s2 127.0.0.1:8000 maxconn 1

Here we count the number of saturated servers that were encountered, and
escape the loop once the number of remaining servers exceeds the number
of saturated ones. No backport is needed since this arrived in 3.2.

src/lb_fwlc.c

@@ -762,13 +762,14 @@ redo:
 	while (node) {
 		struct fwlc_tree_elt *tree_elt;
 		struct server *s;
+		int unusable = 0;
 		int orig_nb;
 		int i = 0;
 
 		tree_elt = eb32_entry(node, struct fwlc_tree_elt, lb_node);
 		orig_nb = statistical_prng_range(FWLC_LISTS_NB);
 
-		while (_HA_ATOMIC_LOAD(&tree_elt->elements) > 0) {
+		while (_HA_ATOMIC_LOAD(&tree_elt->elements) > unusable) {
 			struct mt_list mt_list;
 			mt_list.next = _HA_ATOMIC_LOAD(&tree_elt->srv_list[(i + orig_nb) % FWLC_LISTS_NB].next);
 
@@ -802,6 +803,8 @@ redo:
                                         }
 					avoided = s;
 				}
+				else
+					unusable++;
 				i++;
 			} else if (mt_list.next == &tree_elt->srv_list[(i + orig_nb) % FWLC_LISTS_NB]) {
 				i++;