From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1751253AbcGMUvO (ORCPT <rfc822;w@1wt.eu>);
	Wed, 13 Jul 2016 16:51:14 -0400
Received: from bombadil.infradead.org ([198.137.202.9]:50641 "EHLO
	bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1750883AbcGMUvG (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Wed, 13 Jul 2016 16:51:06 -0400
Date: Wed, 13 Jul 2016 22:51:02 +0200
From: Peter Zijlstra <peterz@infradead.org>
To: Tejun Heo <tj@kernel.org>
Cc: John Stultz <john.stultz@linaro.org>, Ingo Molnar <mingo@redhat.com>,
        lkml <linux-kernel@vger.kernel.org>,
        Dmitry Shmidt <dimitrysh@google.com>,
        Rom Lemarchand <romlem@google.com>, Colin Cross <ccross@google.com>,
        Todd Kjos <tkjos@google.com>, Oleg Nesterov <oleg@redhat.com>,
        "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Subject: Re: Severe performance regression w/ 4.4+ on Android due to cgroup
 locking changes
Message-ID: <20160713205102.GZ30909@twins.programming.kicks-ass.net>
References: <CALAqxLV2H06M2AxP2kK7pGHjLgQAaGdHB9HJEGZjH_Rouy4WhQ@mail.gmail.com>
 <20160713182102.GJ4065@mtj.duckdns.org>
 <20160713183347.GK4065@mtj.duckdns.org>
 <CALAqxLU76kVm8b5GNnSk25-Xk6D1efiWtwNJbY1th_pcN9Ks=g@mail.gmail.com>
 <20160713201823.GB29670@mtj.duckdns.org>
 <20160713202657.GW30154@twins.programming.kicks-ass.net>
 <20160713203944.GC29670@mtj.duckdns.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20160713203944.GC29670@mtj.duckdns.org>
User-Agent: Mutt/1.5.23.1 (2014-03-12)
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Wed, Jul 13, 2016 at 04:39:44PM -0400, Tejun Heo wrote:

> > There is a synchronize_sched() in there, so sorta. That thing is heavily
> > geared towards readers, as is the only 'sane' choice for global locks.
> 
> It used to use the expedited variant until 001dac627ff3
> ("locking/percpu-rwsem: Make use of the rcu_sync infrastructure"), so
> it might have been okay before then.

Right, but expedited stuff sprays IPIs around the entire system. That's
stuff other people complain about.

> The options that I can see are
> 
> 1. Somehow make percpu_rwsem's write behavior more responsive in a way
>    which is acceptable all use cases.  This would be great but
>    probably impossible.
> 
> 2. Add a fast-writer option to percpu_rwsem so that users which care
>    about write latency can opt in for higher processing overhead for
>    lower latency.

So, IIRC, the trade-off is a full memory barrier in read_lock and
read_unlock() vs sync_sched() in write.

Full memory barriers are expensive and while the combined cost might
well exceed the cost of the sync_sched() it doesn't suffer the latency
issues.

Not sure if we can frob the two in a single codebase, but I can have a
poke if Oleg or Paul doesn't beat me to it.