Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Thu, 28 Jun 2018 16:33:23 +0200
From:   Dominique Martinet <asmadeus@codewreck.org>
To:     Matthew Wilcox <willy@infradead.org>
Cc:     v9fs-developer@lists.sourceforge.net,
        Latchesar Ionkov <lucho@ionkov.net>,
        Eric Van Hensbergen <ericvh@gmail.com>,
        linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org,
        Ron Minnich <rminnich@sandia.gov>
Subject: Re: [V9fs-developer] [PATCH 4/6] 9p: Remove an unnecessary memory
 barrier
Message-ID: <20180628143323.GA32547@nautica>
References: <20180628132629.3148-1-willy@infradead.org>
 <20180628132629.3148-5-willy@infradead.org>
 <20180628134029.GA24673@nautica>
 <20180628140358.GG7646@bombadil.infradead.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <20180628140358.GG7646@bombadil.infradead.org>
User-Agent: Mutt/1.5.21 (2010-09-15)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Matthew Wilcox wrote on Thu, Jun 28, 2018:
> How about this?
> 
> 	/*
> 	 * This barrier is needed to make sure any change made to req before
> -	 * the other thread wakes up will indeed be seen by the waiting side.
> +	 * the status change is visible to another thread
> 	 */

Yes, that sounds better.

This code is fairly old and I was wondering if the new WRITE_ONCE and
READ_ONCE macros would help but it looks like compile-time barriers only
so I do not think they would be enough...
Documentation/memory-barriers.txt still suggests something similar in
the "SLEEP AND WAKE-UP FUNCTIONS" section so I guess this is fine.


Thanks,
-- 
Dominique Martinet | Asmadeus