From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-141599-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 659 invoked by alias); 29 Oct 2007 16:32:42 -0000
Received: (qmail 646 invoked by uid 22791); 29 Oct 2007 16:32:40 -0000
X-Spam-Check-By: sourceware.org
Received: from us02smtp1.synopsys.com (HELO vaxjo.synopsys.com) (198.182.60.75)     by sourceware.org (qpsmtpd/0.31) with ESMTP; Mon, 29 Oct 2007 16:32:33 +0000
Received: from crone.synopsys.com (crone.synopsys.com [146.225.7.23]) 	by vaxjo.synopsys.com (Postfix) with ESMTP id A0234DDDD; 	Mon, 29 Oct 2007 09:32:31 -0700 (PDT)
Received: from piper.synopsys.com (localhost [127.0.0.1]) 	by crone.synopsys.com (8.9.1/8.9.1) with ESMTP id JAA00657; 	Mon, 29 Oct 2007 09:30:50 -0700 (PDT)
Received: from piper.synopsys.com (localhost [127.0.0.1]) 	by piper.synopsys.com (8.12.11/8.12.3) with ESMTP id l9TGWUZB010758; 	Mon, 29 Oct 2007 09:32:30 -0700
Received: (from jbuck@localhost) 	by piper.synopsys.com (8.12.11/8.12.11/Submit) id l9TGWUW5010756; 	Mon, 29 Oct 2007 09:32:30 -0700
Date: Mon, 29 Oct 2007 16:47:00 -0000
From: Joe Buck <Joe.Buck@synopsys.COM>
To: David Miller <davem@davemloft.net>
Cc: darryl-mailinglists@netbauds.net, dave.korn@artimi.com,         tomash.brechko@gmail.com, matz@suse.de, gcc@gcc.gnu.org
Subject: Re: Optimization of conditional access to globals: thread-unsafe?
Message-ID: <20071029163230.GB10611@synopsys.com>
References: <02e801c819c9$48396120$2e08a8c0@CAM.ARTIMI.COM> <20071028.183730.37179404.davem@davemloft.net> <4725675D.6000404@netbauds.net> <20071028.221738.18826511.davem@davemloft.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20071028.221738.18826511.davem@davemloft.net>
User-Agent: Mutt/1.4.1i
Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc/>
List-Post: <mailto:gcc@gcc.gnu.org>
List-Help: <http://gcc.gnu.org/ml/>
Sender: gcc-owner@gcc.gnu.org
X-SW-Source: 2007-10/txt/msg00711.txt.bz2

On Sun, Oct 28, 2007 at 10:17:38PM -0700, David Miller wrote:
> From: Darryl Miles <darryl-mailinglists@netbauds.net>
> Date: Mon, 29 Oct 2007 04:53:49 +0000
> 
> > What are the issues with "speculative loads" ?
> 
> The conditional might be protecting whether the pointer is valid and
> can be dereferenced at all.
> 
> int *counter;
> 
> void foo(int counter_is_valid)
> {
> 	if (counter_is_valid)
> 		(*counter)++;
> }

GCC will never do a speculative access in this case.  That's because
"counter_is_valid" might really mean "counter_is_non_null".

It seems that the original issue can only occur if there is a direct
write to a global, not a write through a pointer.