From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:51187)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <dgilbert@redhat.com>) id 1YWVod-0003wQ-Lf
	for qemu-devel@nongnu.org; Fri, 13 Mar 2015 16:04:16 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <dgilbert@redhat.com>) id 1YWVoa-0000LD-9U
	for qemu-devel@nongnu.org; Fri, 13 Mar 2015 16:04:15 -0400
Received: from mx1.redhat.com ([209.132.183.28]:35195)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <dgilbert@redhat.com>) id 1YWVoa-0000L6-3x
	for qemu-devel@nongnu.org; Fri, 13 Mar 2015 16:04:12 -0400
Date: Fri, 13 Mar 2015 20:04:06 +0000
From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Message-ID: <20150313200405.GP2486@work-vm>
References: <CAFEAcA-bfQH0WmtT+msVDZV9Fk5RaZCBCR=E2tpYSCRCdFch5Q@mail.gmail.com>
	<54EF614B.1010607@greensocs.com>
	<CAFEAcA8g1HTLXdTqmq0eR-oP5=UgXp+1Yq_ze3gLO1+k0uJmDg@mail.gmail.com>
	<73E10212-3603-4B64-BA61-66607C12A16A@greensocs.com>
	<CAFEAcA8Lg_iLhd2Vs8Q1KtvWUrqRyHGjmU92RemqNmC2MDEf0w@mail.gmail.com>
	<711BBCAD-9514-4523-A456-AD4F397C41CD@greensocs.com>
	<54F5D3F4.40705@redhat.com>
	<19B4F8D2-B464-4B00-900C-A057DCCD54B3@greensocs.com>
	<20150303154742.GK2293@work-vm> <55033C9B.3010507@twiddle.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <55033C9B.3010507@twiddle.net>
Subject: Re: [Qemu-devel] [RFC 01/10] target-arm: protect cpu_exclusive_*.
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Richard Henderson <rth@twiddle.net>
Cc: mttcg@greensocs.com, Peter Maydell <peter.maydell@linaro.org>, "J. Kiszka" <jan.kiszka@siemens.com>, Mark Burton <mark.burton@greensocs.com>, QEMU Developers <qemu-devel@nongnu.org>, Alexander Graf <agraf@suse.de>, Paolo Bonzini <pbonzini@redhat.com>, KONRAD Fr?d?ric <fred.konrad@greensocs.com>

* Richard Henderson (rth@twiddle.net) wrote:
> On 03/03/2015 07:47 AM, Dr. David Alan Gilbert wrote:
> > That works for ARM where you have to terminate a ldrex with a strex or clrex,
> > but not all architectures have the equivalent of a clrex; most as I remember
> > just let you do an ldrex equivalent, decide you don't want to do the strex equivalent
> > and get on with life.
> 
> 
> I'm pretty sure that's not the case.  In fact, I can guarantee you that GCC
> never issues clrex, but does in fact simply do nothing like you describe for
> other architectures if we decide not to do the store.

Oh well, that means this technique won't work even for ARM, where I thought
it might stand a chance for ARM but nothing else.
I've still got a vague memory that some ARM docs at one point told you that
you should terminate an LDREX by either an STREX or a CLREX; but it's ~3.5 years
since I did any arm wrestling.

Dave

> r~
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK