From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([140.186.70.92]:46452)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <rusty@ozlabs.org>) id 1Qxu8F-0001yB-DP
	for qemu-devel@nongnu.org; Mon, 29 Aug 2011 01:11:36 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <rusty@ozlabs.org>) id 1Qxu8C-00085V-Mp
	for qemu-devel@nongnu.org; Mon, 29 Aug 2011 01:11:35 -0400
Received: from ozlabs.org ([203.10.76.45]:60202)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <rusty@ozlabs.org>) id 1Qxu8C-00085L-De
	for qemu-devel@nongnu.org; Mon, 29 Aug 2011 01:11:32 -0400
From: Rusty Russell <rusty@rustcorp.com.au>
In-Reply-To: <20110826054008.GL2308@yookeroo.fritz.box>
References: <20110826054008.GL2308@yookeroo.fritz.box>
Date: Mon, 29 Aug 2011 12:13:47 +0930
Message-ID: <87ipph2azg.fsf@rustcorp.com.au>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Subject: Re: [Qemu-devel] Missing barriers in hw/virtio.c
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: David Gibson <david@gibson.dropbear.id.au>, qemu-devel@nongnu.org
Cc: aik@ozlabs.ru, agraf@suse.de, Benjamin Herrenschmidt <benh@au.ibm.com>

On Fri, 26 Aug 2011 15:40:08 +1000, David Gibson <david@gibson.dropbear.id.au> wrote:
> Near the top of hw/virtio.c we have this:
> 
> /* QEMU doesn't strictly need write barriers since everything runs in
>  * lock-step.  We'll leave the calls to wmb() in though to make it
>  obvious for
>  * KVM or if kqemu gets SMP support.
>  * In any case, we must prevent the compiler from reordering the code.
>  * TODO: we likely need some rmb()/mb() as well.
>  */
> 
> #define wmb() __asm__ __volatile__("": : :"memory")
> 
> 
> However, as far as I can tell when using both kvm and io-thread, the
> assertion that barriers aren't necessary just isn't true.  Although it
> probably works on x86 with its strongly ordered model.

That looks very much like it's cribbed from the original lguest code.

Very wrong, please fix!

Thanks,
Rusty.