From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:36957)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <chenliang88@huawei.com>) id 1WiajM-0004LP-1H
	for qemu-devel@nongnu.org; Thu, 08 May 2014 22:40:18 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <chenliang88@huawei.com>) id 1WiajF-0001vC-Ta
	for qemu-devel@nongnu.org; Thu, 08 May 2014 22:40:11 -0400
Received: from szxga01-in.huawei.com ([119.145.14.64]:10949)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <chenliang88@huawei.com>) id 1WiajF-0001jC-6j
	for qemu-devel@nongnu.org; Thu, 08 May 2014 22:40:05 -0400
Message-ID: <536C3FF0.7010907@huawei.com>
Date: Fri, 9 May 2014 10:39:44 +0800
From: ChenLiang <chenliang88@huawei.com>
MIME-Version: 1.0
References: <536C0E5A.2030703@kamp.de>
In-Reply-To: <536C0E5A.2030703@kamp.de>
Content-Type: text/plain; charset="ISO-8859-15"
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] Question about RAM migration flags
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: Peter Lieven <pl@kamp.de>
Cc: weidong.huang@huawei.com, quintela@redhat.com, "qemu-devel@nongnu.org" <qemu-devel@nongnu.org>, Orit Wasserman <owasserm@redhat.com>, Stefan Hajnoczi <stefanha@redhat.com>, Paolo Bonzini <pbonzini@redhat.com>

On 2014/5/9 7:08, Peter Lieven wrote:

> Hi,
> 
> while working on ram migration and reading through the code I realized that
> qemu does not stop loading a saved VM or rejecting an incoming migration
> if there is a flag in the stream that it does not understand. An unknown flag
> is simply ignored.
> 
> In the block migration code there is a catch at the end complaining about
> unknown flags, but in RAM migration there isn't.
> 
> Is this on purpose or an error?
> 
> Peter
> 
> 
> 


IMO, the flag in ram_load always is correct if the code of ram_save is correct.
Of course, it will be much rubost to do it as block migration.