Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753487AbZCTEfP (ORCPT ); Fri, 20 Mar 2009 00:35:15 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1750820AbZCTEe7 (ORCPT ); Fri, 20 Mar 2009 00:34:59 -0400 Received: from e28smtp05.in.ibm.com ([59.145.155.5]:49523 "EHLO e28smtp05.in.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750703AbZCTEe6 (ORCPT ); Fri, 20 Mar 2009 00:34:58 -0400 Date: Fri, 20 Mar 2009 10:04:37 +0530 From: Gautham R Shenoy To: Vaidyanathan Srinivasan Cc: Peter Zijlstra , Ingo Molnar , linux-kernel@vger.kernel.org, Suresh Siddha , Balbir Singh Subject: Re: [PATCH 3 2/6] sched: Record the current active power savings level Message-ID: <20090320043437.GA6603@in.ibm.com> Reply-To: ego@in.ibm.com References: <20090318092054.24787.18730.stgit@sofia.in.ibm.com> <20090318092227.24787.22917.stgit@sofia.in.ibm.com> <20090319163205.GK2990@dirshya.in.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20090319163205.GK2990@dirshya.in.ibm.com> User-Agent: Mutt/1.5.17 (2007-11-01) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4613 Lines: 116 On Thu, Mar 19, 2009 at 10:02:05PM +0530, Vaidyanathan Srinivasan wrote: > * Gautham R Shenoy [2009-03-18 14:52:28]: > > > The current active power savings level of a system is defined as the > > maximum of the sched_mc_power_savings and the sched_smt_power_savings. > > > > The decisions during power-aware loadbalancing, depend on this value. > > > > Record this value in a read mostly global variable instead of having to > > compute it everytime. > > > > Signed-off-by: Gautham R Shenoy > > Cc: Peter Zijlstra > > --- > > > > include/linux/sched.h | 1 + > > kernel/sched.c | 8 ++++++-- > > kernel/sched_fair.c | 2 +- > > 3 files changed, 8 insertions(+), 3 deletions(-) > > > > diff --git a/include/linux/sched.h b/include/linux/sched.h > > index 37fecf7..7dc8aea 100644 > > --- a/include/linux/sched.h > > +++ b/include/linux/sched.h > > @@ -793,6 +793,7 @@ enum powersavings_balance_level { > > }; > > > > extern int sched_mc_power_savings, sched_smt_power_savings; > > Now we will need sched_mc_power_savings and sched_smt_power_savings > only until we rebuild the sched domains. These can be static > variables and we can perhaps remove the extern for them? Better still > if we can capture this information elsewhere until sched domain is > built and SD_POWERSAVINGS_BALANCE flags are set so as to not > have a need for these global variables. Right now we need these variables outside sched.c only while rebuilding the sched_domains, as you rightly pointed out. So, yes these variables should be _read_only outside sched.c, but they are required nevertheless. However, like we had discussed in one of the earlier posts, when we can have a single tunable that can capture in essence what these two variables seek to achieve, we can get rid of these variables. Till then I think we'll have to retain them. > > > +extern enum powersavings_balance_level active_power_savings_level; > > enum sched_domain_level { > > SD_LV_NONE = 0, > > diff --git a/kernel/sched.c b/kernel/sched.c > > index 8e2558c..407ee03 100644 > > --- a/kernel/sched.c > > +++ b/kernel/sched.c > > @@ -3398,7 +3398,7 @@ out_balanced: > > > > if (this == group_leader && group_leader != group_min) { > > *imbalance = min_load_per_task; > > - if (sched_mc_power_savings >= POWERSAVINGS_BALANCE_WAKEUP) { > > + if (active_power_savings_level >= POWERSAVINGS_BALANCE_WAKEUP) { > > cpu_rq(this_cpu)->rd->sched_mc_preferred_wakeup_cpu = > > cpumask_first(sched_group_cpus(group_leader)); > > } > > @@ -3683,7 +3683,7 @@ redo: > > !test_sd_parent(sd, SD_POWERSAVINGS_BALANCE)) > > return -1; > > > > - if (sched_mc_power_savings < POWERSAVINGS_BALANCE_WAKEUP) > > + if (active_power_savings_level < POWERSAVINGS_BALANCE_WAKEUP) > > return -1; > > > > if (sd->nr_balance_failed++ < 2) > > @@ -7206,6 +7206,8 @@ static void sched_domain_node_span(int node, struct cpumask *span) > > #endif /* CONFIG_NUMA */ > > > > int sched_smt_power_savings = 0, sched_mc_power_savings = 0; > > +/* Records the currently active power savings level */ > > +enum powersavings_balance_level __read_mostly active_power_savings_level; > > > > /* > > * The cpus mask in sched_group and sched_domain hangs off the end. > > @@ -8040,6 +8042,8 @@ static ssize_t sched_power_savings_store(const char *buf, size_t count, int smt) > > sched_smt_power_savings = level; > > else > > sched_mc_power_savings = level; > > + active_power_savings_level = max(sched_smt_power_savings, > > + sched_mc_power_savings); > > > > arch_reinit_sched_domains(); > > > > diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c > > index 0566f2a..a3583c6 100644 > > --- a/kernel/sched_fair.c > > +++ b/kernel/sched_fair.c > > @@ -1054,7 +1054,7 @@ static int wake_idle(int cpu, struct task_struct *p) > > chosen_wakeup_cpu = > > cpu_rq(this_cpu)->rd->sched_mc_preferred_wakeup_cpu; > > > > - if (sched_mc_power_savings >= POWERSAVINGS_BALANCE_WAKEUP && > > + if (active_power_savings_level >= POWERSAVINGS_BALANCE_WAKEUP && > > idle_cpu(cpu) && idle_cpu(this_cpu) && > > p->mm && !(p->flags & PF_KTHREAD) && > > cpu_isset(chosen_wakeup_cpu, p->cpus_allowed)) > > > > > Acked-by: Vaidyanathan Srinivasan -- Thanks and Regards gautham -- 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/