Commit ea1dc6fc authored by Peter Zijlstra's avatar Peter Zijlstra Committed by Ingo Molnar
Browse files

sched/fair: Fix calc_cfs_shares() fixed point arithmetics width confusion 



Commit:

  fde7d22e ("sched/fair: Fix overly small weight for interactive group entities")

did something non-obvious but also did it buggy yet latent.

The problem was exposed for real by a later commit in the v4.7 merge window:

  2159197d ("sched/core: Enable increased load resolution on 64-bit kernels")

... after which tg->load_avg and cfs_rq->load.weight had different
units (10 bit fixed point and 20 bit fixed point resp.).

Add a comment to explain the use of cfs_rq->load.weight over the
'natural' cfs_rq->avg.load_avg and add scale_load_down() to correct
for the difference in unit.

Since this is (now, as per a previous commit) the only user of
calc_tg_weight(), collapse it.

The effects of this bug should be randomly inconsistent SMP-balancing
of cgroups workloads.

Reported-by: default avatarJirka Hladky <jhladky@redhat.com>
Signed-off-by: default avatarPeter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: 2159197d ("sched/core: Enable increased load resolution on 64-bit kernels")
Fixes: fde7d22e ("sched/fair: Fix overly small weight for interactive group entities")
Signed-off-by: default avatarIngo Molnar <mingo@kernel.org>
parent 7dd49125

kernel/sched/fair.c

+11 −16
Original line number Original line Diff line number Diff line
@@ -2497,28 +2497,22 @@ account_entity_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se) 




#ifdef CONFIG_FAIR_GROUP_SCHED

#ifdef CONFIG_FAIR_GROUP_SCHED

# ifdef CONFIG_SMP

# ifdef CONFIG_SMP

static inline long calc_tg_weight(struct task_group *tg, struct cfs_rq *cfs_rq)

static long calc_cfs_shares(struct cfs_rq *cfs_rq, struct task_group *tg)

{

{

	long tg_weight;

	long tg_weight, load, shares;





	/*

	/*

	 * Use this CPU's real-time load instead of the last load contribution

	 * This really should be: cfs_rq->avg.load_avg, but instead we use

	 * as the updating of the contribution is delayed, and we will use the

	 * cfs_rq->load.weight, which is its upper bound. This helps ramp up

	 * the real-time load to calc the share. See update_tg_load_avg().

	 * the shares for small weight interactive tasks.

	 */

	 */

	tg_weight = atomic_long_read(&tg->load_avg);

	load = scale_load_down(cfs_rq->load.weight);

	tg_weight -= cfs_rq->tg_load_avg_contrib;

	tg_weight += cfs_rq->load.weight;



	return tg_weight;

}





static long calc_cfs_shares(struct cfs_rq *cfs_rq, struct task_group *tg)

	tg_weight = atomic_long_read(&tg->load_avg);

{

	long tg_weight, load, shares;





	tg_weight = calc_tg_weight(tg, cfs_rq);

	/* Ensure tg_weight >= load */

	load = cfs_rq->load.weight;

	tg_weight -= cfs_rq->tg_load_avg_contrib;

	tg_weight += load;





	shares = (tg->shares * load);

	shares = (tg->shares * load);

	if (tg_weight)

	if (tg_weight)
@@ -2537,6 +2531,7 @@ static inline long calc_cfs_shares(struct cfs_rq *cfs_rq, struct task_group *tg) 
	return tg->shares;

	return tg->shares;

}

}

# endif /* CONFIG_SMP */

# endif /* CONFIG_SMP */



static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,

static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,

			    unsigned long weight)

			    unsigned long weight)

{

{

CRA Git | Maintained and supported by SUSTech CRA and CCSE