Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757079AbZCDMT4 (ORCPT ); Wed, 4 Mar 2009 07:19:56 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751618AbZCDMTs (ORCPT ); Wed, 4 Mar 2009 07:19:48 -0500 Received: from e28smtp08.in.ibm.com ([59.145.155.8]:49595 "EHLO e28smtp08.in.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751206AbZCDMTr (ORCPT ); Wed, 4 Mar 2009 07:19:47 -0500 Date: Wed, 4 Mar 2009 17:49:37 +0530 From: Arun R Bharadwaj To: linux-kernel@vger.kernel.org, linux-pm@lists.linux-foundation.org Cc: a.p.zijlstra@chello.nl, ego@in.ibm.com, tglx@linutronix.de, mingo@elte.hu, andi@firstfloor.org, venkatesh.pallipadi@intel.com, vatsa@linux.vnet.ibm.com, arjan@infradead.org, svaidy@linux.vnet.ibm.com, Arun Bharadwaj Subject: [v2 PATCH 4/4] timers: logic to enable timer migration. Message-ID: <20090304121937.GE9855@linux.vnet.ibm.com> Reply-To: arun@linux.vnet.ibm.com References: <20090304121249.GA9855@linux.vnet.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: <20090304121249.GA9855@linux.vnet.ibm.com> User-Agent: Mutt/1.5.18 (2008-05-17) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3685 Lines: 117 * Arun R Bharadwaj [2009-03-04 17:42:49]: This patch migrates all non pinned timers and hrtimers to the current idle load balancer, from all the idle CPUs. Timers firing on busy CPUs are not migrated. Signed-off-by: Arun R Bharadwaj --- include/linux/sched.h | 1 + kernel/hrtimer.c | 12 +++++++++++- kernel/sched.c | 5 +++++ kernel/timer.c | 13 ++++++++++++- 4 files changed, 29 insertions(+), 2 deletions(-) Index: linux.trees.git/kernel/timer.c =================================================================== --- linux.trees.git.orig/kernel/timer.c +++ linux.trees.git/kernel/timer.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #include @@ -628,7 +629,7 @@ __mod_timer(struct timer_list *timer, un { struct tvec_base *base, *new_base; unsigned long flags; - int ret; + int ret, current_cpu, preferred_cpu; ret = 0; @@ -649,6 +650,16 @@ __mod_timer(struct timer_list *timer, un new_base = __get_cpu_var(tvec_bases); + current_cpu = smp_processor_id(); + preferred_cpu = get_nohz_load_balancer(); + if (enable_timer_migration && !tbase_get_pinned(timer->base) && + idle_cpu(current_cpu) && preferred_cpu != -1) { + new_base = per_cpu(tvec_bases, preferred_cpu); + timer_set_base(timer, new_base); + timer->expires = expires; + internal_add_timer(new_base, timer); + goto out_unlock; + } if (base != new_base) { /* * We are trying to schedule the timer on the local CPU. Index: linux.trees.git/kernel/hrtimer.c =================================================================== --- linux.trees.git.orig/kernel/hrtimer.c +++ linux.trees.git/kernel/hrtimer.c @@ -43,6 +43,8 @@ #include #include #include +#include +#include #include @@ -198,8 +200,16 @@ int pinned) { struct hrtimer_clock_base *new_base; struct hrtimer_cpu_base *new_cpu_base; + int current_cpu, preferred_cpu; + + current_cpu = smp_processor_id(); + preferred_cpu = get_nohz_load_balancer(); + if (enable_timer_migration && !pinned && preferred_cpu != -1 && + idle_cpu(current_cpu)) + new_cpu_base = &per_cpu(hrtimer_bases, preferred_cpu); + else + new_cpu_base = &__get_cpu_var(hrtimer_bases); - new_cpu_base = &__get_cpu_var(hrtimer_bases); new_base = &new_cpu_base->clock_base[base->index]; if (base != new_base) { Index: linux.trees.git/include/linux/sched.h =================================================================== --- linux.trees.git.orig/include/linux/sched.h +++ linux.trees.git/include/linux/sched.h @@ -265,6 +265,7 @@ static inline int select_nohz_load_balan } #endif +extern int get_nohz_load_balancer(void); /* * Only dump TASK_* tasks. (0 for all tasks) */ Index: linux.trees.git/kernel/sched.c =================================================================== --- linux.trees.git.orig/kernel/sched.c +++ linux.trees.git/kernel/sched.c @@ -4009,6 +4009,11 @@ static struct { .load_balancer = ATOMIC_INIT(-1), }; +inline int get_nohz_load_balancer(void) +{ + return atomic_read(&nohz.load_balancer); +} + /* * This routine will try to nominate the ilb (idle load balancing) * owner among the cpus whose ticks are stopped. ilb owner will do the idle -- 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/