Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751312AbaLPO2H (ORCPT ); Tue, 16 Dec 2014 09:28:07 -0500 Received: from relay.parallels.com ([195.214.232.42]:56551 "EHLO relay.parallels.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750863AbaLPO2G (ORCPT ); Tue, 16 Dec 2014 09:28:06 -0500 Message-ID: <1418740080.6074.16.camel@tkhai> Subject: Re: [PATCH] sched/fair: Make cfs_rq::decay_counter non-atomic From: Kirill Tkhai To: CC: Peter Zijlstra , Ingo Molnar , Ben Segall , Kirill Tkhai Date: Tue, 16 Dec 2014 17:28:00 +0300 In-Reply-To: <1418739933.6074.15.camel@tkhai> References: <1418739933.6074.15.camel@tkhai> Organization: Parallels Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.8.5-2+b3 MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Originating-IP: [10.30.26.172] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org This should go on top of yesterday's "sched/fair: Fix sched_entity::avg::decay_count initialization" В Вт, 16/12/2014 в 17:25 +0300, Kirill Tkhai пишет: > We update decay_counter in update_cfs_rq_blocked_load() > only. This function is always called with rq lock locked, > so we can kill atomic actions. > > Signed-off-by: Kirill Tkhai > --- > kernel/sched/fair.c | 11 ++++++----- > kernel/sched/sched.h | 2 +- > 2 files changed, 7 insertions(+), 6 deletions(-) > > diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c > index 5f3b5a7..af990c4 100644 > --- a/kernel/sched/fair.c > +++ b/kernel/sched/fair.c > @@ -2570,7 +2570,7 @@ static __always_inline int __update_entity_runnable_avg(u64 now, > static inline u64 __synchronize_entity_decay(struct sched_entity *se) > { > struct cfs_rq *cfs_rq = cfs_rq_of(se); > - u64 decays = atomic64_read(&cfs_rq->decay_counter); > + u64 decays = ACCESS_ONCE(cfs_rq->decay_counter); > > decays -= se->avg.decay_count; > if (!decays) > @@ -2767,7 +2767,8 @@ static void update_cfs_rq_blocked_load(struct cfs_rq *cfs_rq, int force_update) > if (decays) { > cfs_rq->blocked_load_avg = decay_load(cfs_rq->blocked_load_avg, > decays); > - atomic64_add(decays, &cfs_rq->decay_counter); > + lockdep_assert_held(&rq_of(cfs_rq)->lock); > + cfs_rq->decay_counter += decays; > cfs_rq->last_decay = now; > } > > @@ -2837,7 +2838,7 @@ static inline void dequeue_entity_load_avg(struct cfs_rq *cfs_rq, > cfs_rq->runnable_load_avg -= se->avg.load_avg_contrib; > if (sleep) { > cfs_rq->blocked_load_avg += se->avg.load_avg_contrib; > - se->avg.decay_count = atomic64_read(&cfs_rq->decay_counter); > + se->avg.decay_count = ACCESS_ONCE(cfs_rq->decay_counter); > } /* migrations, e.g. sleep=0 leave decay_count == 0 */ > } > > @@ -7876,7 +7877,7 @@ void init_cfs_rq(struct cfs_rq *cfs_rq) > cfs_rq->min_vruntime_copy = cfs_rq->min_vruntime; > #endif > #ifdef CONFIG_SMP > - atomic64_set(&cfs_rq->decay_counter, 1); > + cfs_rq->decay_counter = 1; > atomic_long_set(&cfs_rq->removed_load, 0); > #endif > } > @@ -7928,7 +7929,7 @@ static void task_move_group_fair(struct task_struct *p, int queued) > * contribution, but we must synchronize for ongoing future > * decay. > */ > - se->avg.decay_count = atomic64_read(&cfs_rq->decay_counter); > + se->avg.decay_count = ACCESS_ONCE(cfs_rq->decay_counter); > cfs_rq->blocked_load_avg += se->avg.load_avg_contrib; > #endif > } > diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h > index 9a2a45c..672f120 100644 > --- a/kernel/sched/sched.h > +++ b/kernel/sched/sched.h > @@ -364,7 +364,7 @@ struct cfs_rq { > * the FAIR_GROUP_SCHED case). > */ > unsigned long runnable_load_avg, blocked_load_avg; > - atomic64_t decay_counter; > + u64 decay_counter; > u64 last_decay; > atomic_long_t removed_load; > > > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/