Re: [PATCH 1/2] sched: fix and clean up calculate_imbalance

[Peter Zijlstra]

On Tue, Jul 29, 2014 at 04:53:08PM +0200, Peter Zijlstra wrote:
> Ah, wait, I think I see why you want that gone. I was only expecting a
> correction fix wrt changing pick_busiest(), not also behaviour changes.

---
Subject: sched,fair: Make calculate_imbalance() independent
From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Date: Tue Jul 29 17:00:21 CEST 2014

Rik noticed that calculate_imbalance() relies on
update_sd_pick_busiest() to guarantee that busiest->sum_nr_running >
busiest->group_capacity_factor.

Break this implicit assumption (with the intent of not providing it
anymore) by having calculat_imbalance() verify it and not rely on
others.

Reported-by: Rik van Riel <riel@xxxxxxxxxx>
Signed-off-by: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
---
 kernel/sched/fair.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -6248,7 +6248,7 @@ static inline void calculate_imbalance(s
 		return fix_small_imbalance(env, sds);
 	}
 
-	if (!busiest->group_imb) {
+	if (busiest->sum_nr_running > busiest->group_capacity_factor) {
 		/*
 		 * Don't want to pull so many tasks that a group would go idle.
 		 * Except of course for the group_imb case, since then we might
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/