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 S1752273AbcEWVFH (ORCPT <rfc822;w@1wt.eu>);
	Mon, 23 May 2016 17:05:07 -0400
Received: from mx2.suse.de ([195.135.220.15]:37998 "EHLO mx2.suse.de"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1752059AbcEWVFG (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Mon, 23 May 2016 17:05:06 -0400
Date: Mon, 23 May 2016 14:04:55 -0700
From: Davidlohr Bueso <dave@stgolabs.net>
To: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Jason Low <jason.low2@hpe.com>, Peter Hurley <peter@hurleysoftware.com>,
        jason.low2@hp.com, Waiman Long <waiman.long@hpe.com>,
        Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@redhat.com>,
        linux-kernel@vger.kernel.org, Dave Chinner <david@fromorbit.com>,
        Scott J Norton <scott.norton@hpe.com>,
        Douglas Hatch <doug.hatch@hpe.com>
Subject: Re: [PATCH v4 2/5] locking/rwsem: Protect all writes to owner by
 WRITE_ONCE
Message-ID: <20160523210455.7rycskjo7n5rw73k@linux-uzut.site>
References: <1463534783-38814-1-git-send-email-Waiman.Long@hpe.com>
 <1463534783-38814-3-git-send-email-Waiman.Long@hpe.com>
 <20160518140436.GA6273@linux-uzut.site>
 <1463592095.3369.10.camel@j-VirtualBox>
 <573CB496.4010707@hpe.com>
 <1463601515.2587.24.camel@j-VirtualBox>
 <574086FE.6080807@hurleysoftware.com>
 <1464029181.2479.21.camel@j-VirtualBox>
 <20160523194416.up365sfbm7f55kbq@linux-uzut.site>
 <20160523201502.GN3825@linux.vnet.ibm.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Disposition: inline
In-Reply-To: <20160523201502.GN3825@linux.vnet.ibm.com>
User-Agent: Mutt/1.6.1-neo ()
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Mon, 23 May 2016, Paul E. McKenney wrote:

>But rcu_read_lock() does not exclude updates, which is one reason why
>pointer reads use rcu_dereference() rather than normal assignments.

Yes, I was referring to readers. With updates, otoh, are done holding a number of locks.

>
>So I do not believe that rcu_read_lock() is helping you in this case.
>
>That said, it is a bit hard to imagine the compiler tearing a load from
>an int...

Oh, right, good point.

Thanks,
Davidlohr