From owner-freebsd-arch  Sat Mar 22 11:49:21 2003
Delivered-To: freebsd-arch@freebsd.org
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP id D16A637B401
	for <arch@freebsd.org>; Sat, 22 Mar 2003 11:49:20 -0800 (PST)
Received: from magic.adaptec.com (magic-mail.adaptec.com [208.236.45.100])
	by mx1.FreeBSD.org (Postfix) with ESMTP id 3D5CD43FBD
	for <arch@freebsd.org>; Sat, 22 Mar 2003 11:49:20 -0800 (PST)
	(envelope-from gibbs@scsiguy.com)
Received: from redfish.adaptec.com (redfish.adaptec.com [162.62.50.11])
	by magic.adaptec.com (8.11.6/8.11.6) with ESMTP id h2MJmSl09193;
	Sat, 22 Mar 2003 11:48:28 -0800
Received: from btc.btc.adaptec.com (btc.btc.adaptec.com [10.100.0.52])
	by redfish.adaptec.com (8.8.8+Sun/8.8.8) with ESMTP id LAA06908;
	Sat, 22 Mar 2003 11:48:44 -0800 (PST)
Received: from [10.100.253.70] (aslan [10.100.253.70])
	by btc.btc.adaptec.com (8.8.8+Sun/8.8.8) with ESMTP id MAA06657;
	Sat, 22 Mar 2003 12:48:38 -0700 (MST)
Date: Sat, 22 Mar 2003 12:48:39 -0700
From: "Justin T. Gibbs" <gibbs@scsiguy.com>
Reply-To: "Justin T. Gibbs" <gibbs@scsiguy.com>
To: Poul-Henning Kamp <phk@phk.freebsd.dk>,
	Bakul Shah <bakul@bitblocks.com>
Cc: arch@freebsd.org
Subject: Re: memcpy() with specified direction of copying ? 
Message-ID: <8980000.1048362518@aslan.btc.adaptec.com>
In-Reply-To: <9531.1047848405@critter.freebsd.dk>
References:  <9531.1047848405@critter.freebsd.dk>
X-Mailer: Mulberry/3.0.2 (Linux/x86)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Sender: owner-freebsd-arch@FreeBSD.ORG
Precedence: bulk
List-ID: <freebsd-arch.FreeBSD.ORG>
List-Archive: <http://docs.freebsd.org/mail/> (Web Archive)
List-Help: <mailto:majordomo@FreeBSD.ORG?subject=help> (List Instructions)
List-Subscribe: <mailto:majordomo@FreeBSD.ORG?subject=subscribe%20freebsd-arch>
List-Unsubscribe: <mailto:majordomo@FreeBSD.ORG?subject=unsubscribe%20freebsd-arch>
X-Loop: FreeBSD.ORG

> Well, in my case it is a matter of shared memory where the order in
> which I snapshot a data structure is critical for detecting integrity
> issues.

Without memory barriers, the order is undefined on many cpus irrespective
of the direction of copy.  So, for a data structure like this:

	start_tag
	some data
	...
	end_tag

A client knows that the data is stable if start_tag and end_tag are
equal only if the data is written as:

	start_tag = new_tag;
	mb();
	Fill the middle in whatever order is convenient;
	mb();
	end_tag = new_tag;

--
Justin


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-arch" in the body of the message