Message-ID: <1398357472.3509.2.camel@j-VirtualBox>
Subject: Re: [PATCH 3/3] sched, fair: Stop searching for tasks in newidle
 balance if there are runnable tasks
From: Jason Low <jason.low2@hp.com>
To: Mike Galbraith <umgwanakikbuti@gmail.com>
Cc: mingo@kernel.org, peterz@infradead.org, linux-kernel@vger.kernel.org,
        daniel.lezcano@linaro.org, alex.shi@linaro.org,
        preeti@linux.vnet.ibm.com, vincent.guittot@linaro.org,
        morten.rasmussen@arm.com, aswin@hp.com, chegu_vinod@hp.com
Date: Thu, 24 Apr 2014 09:37:52 -0700
In-Reply-To: <1398307888.26653.18.camel@marge.simpson.net>
References: <1398303035-18255-1-git-send-email-jason.low2@hp.com>
	 <1398303035-18255-4-git-send-email-jason.low2@hp.com>
	 <1398307888.26653.18.camel@marge.simpson.net>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Mime-Version: 1.0
Sender: linux-kernel-owner@vger.kernel.org

On Thu, 2014-04-24 at 04:51 +0200, Mike Galbraith wrote:
> On Wed, 2014-04-23 at 18:30 -0700, Jason Low wrote: 
> > It was found that when running some workloads (such as AIM7) on large systems
> > with many cores, CPUs do not remain idle for long. Thus, tasks can
> > wake/get enqueued while doing idle balancing.
> > 
> > In this patch, while traversing the domains in idle balance, in addition to
> > checking for pulled_task, we add an extra check for this_rq->nr_running for
> > determining if we should stop searching for tasks to pull. If there are
> > runnable tasks on this rq, then we will stop traversing the domains. This
> > reduces the chance that idle balance delays a task from running.
> > 
> > This patch resulted in approximately a 6% performance improvement when
> > running a Java Server workload on an 8 socket machine.
> 
> Checking rq->lock for contention before ever going to idle balancing as
> well should give you a bit more.  No need to run around looking for work
> that's trying to arrive.  By not going there, perhaps stacking tasks,
> you may head off a future bounce as well.

Are you thinking of something along the lines of this:

@@ -6658,7 +6658,8 @@ static int idle_balance(struct rq *this_rq)
         */
        this_rq->idle_stamp = rq_clock(this_rq);
 
-       if (this_rq->avg_idle < sysctl_sched_migration_cost)
+       if (this_rq->avg_idle < sysctl_sched_migration_cost ||
+           spin_is_contended(&this_rq->lock))
                goto out;

--
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/