From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner+w=401wt.eu-S1762461AbZEOFRg@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1762461AbZEOFRg (ORCPT <rfc822;w@1wt.eu>);
	Fri, 15 May 2009 01:17:36 -0400
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1761421AbZEOFN0
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Fri, 15 May 2009 01:13:26 -0400
Received: from bombadil.infradead.org ([18.85.46.34]:33885 "EHLO
	bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1761503AbZEOFNZ (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Fri, 15 May 2009 01:13:25 -0400
Subject: Re: [patch 1/2] sched, timers: move calc_load() to scheduler
From: Peter Zijlstra <peterz@infradead.org>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: LKML <linux-kernel@vger.kernel.org>,
       Andrew Morton <akpm@linux-foundation.org>,
       Dimitri Sivanich <sivanich@sgi.com>, Ingo Molnar <mingo@elte.hu>
In-Reply-To: <alpine.LFD.2.00.0905142224230.3561@localhost.localdomain>
References: <20090514105915.973318588@linutronix.de>
	 <20090514110712.521114248@linutronix.de>
	 <1242331227.6642.1637.camel@laptop>
	 <alpine.LFD.2.00.0905142224230.3561@localhost.localdomain>
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
Date: Fri, 15 May 2009 07:13:11 +0200
Message-Id: <1242364391.26820.49.camel@twins>
Mime-Version: 1.0
X-Mailer: Evolution 2.26.1 
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Thu, 2009-05-14 at 22:40 +0200, Thomas Gleixner wrote:
> On Thu, 14 May 2009, Peter Zijlstra wrote:
> > On Thu, 2009-05-14 at 11:21 +0000, Thomas Gleixner wrote:
> > > plain text document attachment (move-calc-load-to-scheduler-v1.patch)
> > 
> > > +/*
> > > + * calc_load - update the avenrun load estimates 10 ticks after the
> > > + * CPUs have updated calc_load_tasks.
> > > + */
> > > +void calc_global_load(void)
> > > +{
> > > +	unsigned long upd = calc_load_update + 10;
> > > +	long active;
> > > +
> > > +	if (time_before(jiffies, upd))
> > > +		return;
> > >  
> > > +	active = atomic_long_read(&calc_load_tasks);
> > > +	active = active > 0 ? active * FIXED_1 : 0;
> > >  
> > > +	avenrun[0] = calc_load(avenrun[0], EXP_1, active);
> > > +	avenrun[1] = calc_load(avenrun[1], EXP_5, active);
> > > +	avenrun[2] = calc_load(avenrun[2], EXP_15, active);
> > > +
> > > +	calc_load_update += LOAD_FREQ;
> > > +}
> > 
> > > @@ -1211,7 +1160,8 @@ static inline void update_times(unsigned
> > >  void do_timer(unsigned long ticks)
> > >  {
> > >  	jiffies_64 += ticks;
> > > -	update_times(ticks);
> > > +	update_wall_time();
> > > +	calc_global_load();
> > >  }
> > 
> > I can see multiple cpus fall into calc_global_load() concurrently, which
> > would 'age' the load faster than expected.
> > 
> > Should we plug that hole?
> 
> They can't. do_timer() is called by exactly one CPU under xtime
> lock. What we removed is the loop over all online CPUs to retrieve the
> number of active tasks.

Ah, D'0h, yes.

Patches look good then.

Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>