Add comments on load balancing special cases.

Ingo explains:

The condition is 'impossible', but the whole balancing code is
(intentionally) a bit racy:

                cpus_and(tmp, group->cpumask, cpu_online_map);
                if (!cpus_weight(tmp))
                        goto next_group;

                for_each_cpu_mask(i, tmp) {
                        if (!idle_cpu(i))
                                goto next_group;
                        push_cpu = i;
                }

                rq = cpu_rq(push_cpu);
                double_lock_balance(busiest, rq);
                move_tasks(rq, push_cpu, busiest, 1, sd, IDLE);

in the for_each_cpu_mask() loop we specifically check for each CPU in
the target group to be idle - so push_cpu's runqueue == busiest [==
current runqueue] cannot be true because the current CPU is not idle, we
are running in the migration thread ... But this is not a real problem,
load-balancing we do in a racy way to reduce overhead [and it's all
statistics anyway so absolute accuracy is impossible], and active
balancing itself is somewhat racy due to the migration-thread wakeup
(and the active_balance flag) going outside the runqueue locks [for
similar reasons].

so it all looks quite plausible - the normal SMP boxes dont trigger it,
but Bjorn's 128-CPU setup with a non-trivial domain hiearachy triggers
it.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

kernel/sched.c

@@ -1681,6 +1681,11 @@ static int load_balance(int this_cpu, runqueue_t *this_rq,
 	busiest = find_busiest_queue(group);
 	if (!busiest)
 		goto out_balanced;
+	/*
+	 * This should be "impossible", but since load
+	 * balancing is inherently racy and statistical,
+	 * it could happen in theory.
+	 */
 	if (unlikely(busiest == this_rq)) {
 		WARN_ON(1);
 		goto out_balanced;
@@ -1844,7 +1849,14 @@ static void active_load_balance(runqueue_t *busiest, int busiest_cpu)
  		}
 
 		rq = cpu_rq(push_cpu);
-		if (busiest == rq)
+
+		/*
+		 * This condition is "impossible", but since load
+		 * balancing is inherently a bit racy and statistical,
+		 * it can trigger.. Reported by Bjorn Helgaas on a
+		 * 128-cpu setup.
+		 */
+		if (unlikely(busiest == rq))
 			goto next_group;
 		double_lock_balance(busiest, rq);
 		move_tasks(rq, push_cpu, busiest, 1, sd, IDLE);