[PATCH] sched: filter affine wakeups

)

From: Nick Piggin <nickpiggin@yahoo.com.au>

Track the last waker CPU, and only consider wakeup-balancing if there's a
match between current waker CPU and the previous waker CPU.  This ensures
that there is some correlation between two subsequent wakeup events before
we move the task.  Should help random-wakeup workloads on large SMP
systems, by reducing the migration attempts by a factor of nr_cpus.
Signed-off-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NNick Piggin <npiggin@suse.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

include/linux/sched.h

@@ -696,8 +696,11 @@ struct task_struct {
 
 	int lock_depth;		/* BKL lock depth */
 
-#if defined(CONFIG_SMP) && defined(__ARCH_WANT_UNLOCKED_CTXSW)
+#if defined(CONFIG_SMP)
+	int last_waker_cpu;	/* CPU that last woke this task up */
+#if defined(__ARCH_WANT_UNLOCKED_CTXSW)
 	int oncpu;
+#endif
 #endif
 	int prio, static_prio;
 	struct list_head run_list;

kernel/sched.c

@@ -1290,6 +1290,9 @@ static int try_to_wake_up(task_t *p, unsigned int state, int sync)
 		}
 	}
 
+	if (p->last_waker_cpu != this_cpu)
+		goto out_set_cpu;
+
 	if (unlikely(!cpu_isset(this_cpu, p->cpus_allowed)))
 		goto out_set_cpu;
 
@@ -1360,6 +1363,8 @@ static int try_to_wake_up(task_t *p, unsigned int state, int sync)
 		cpu = task_cpu(p);
 	}
 
+	p->last_waker_cpu = this_cpu;
+
 out_activate:
 #endif /* CONFIG_SMP */
 	if (old_state == TASK_UNINTERRUPTIBLE) {
@@ -1441,9 +1446,12 @@ void fastcall sched_fork(task_t *p, int clone_flags)
 #ifdef CONFIG_SCHEDSTATS
 	memset(&p->sched_info, 0, sizeof(p->sched_info));
 #endif
-#if defined(CONFIG_SMP) && defined(__ARCH_WANT_UNLOCKED_CTXSW)
+#if defined(CONFIG_SMP)
+	p->last_waker_cpu = cpu;
+#if defined(__ARCH_WANT_UNLOCKED_CTXSW)
 	p->oncpu = 0;
 #endif
+#endif
 #ifdef CONFIG_PREEMPT
 	/* Want to start with kernel preemption disabled. */
 	task_thread_info(p)->preempt_count = 1;