From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43)
	id 1MInq3-0008LH-3U
	for qemu-devel@nongnu.org; Mon, 22 Jun 2009 14:01:51 -0400
Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43)
	id 1MInpy-0008FX-DB
	for qemu-devel@nongnu.org; Mon, 22 Jun 2009 14:01:50 -0400
Received: from [199.232.76.173] (port=37549 helo=monty-python.gnu.org)
	by lists.gnu.org with esmtp (Exim 4.43) id 1MInpy-0008FN-9Y
	for qemu-devel@nongnu.org; Mon, 22 Jun 2009 14:01:46 -0400
Received: from mx2.redhat.com ([66.187.237.31]:59853)
	by monty-python.gnu.org with esmtp (Exim 4.60)
	(envelope-from <avi@redhat.com>) id 1MInpx-00042e-KY
	for qemu-devel@nongnu.org; Mon, 22 Jun 2009 14:01:45 -0400
Message-ID: <4A3FC711.9000200@redhat.com>
Date: Mon, 22 Jun 2009 21:01:53 +0300
From: Avi Kivity <avi@redhat.com>
MIME-Version: 1.0
Subject: Re: [Qemu-devel] Re: [Qemu-commits] [COMMIT 3086844] Instead of
	writing a zero page, madvise it away
References: <200906221549.n5MFn3Qd015389@d03av02.boulder.ibm.com>
	<4A3FAD69.60507@redhat.com> <4A3FB077.4040607@codemonkey.ws>
	<4A3FB390.4060809@redhat.com> <4A3FB95D.3060404@us.ibm.com>
	<4A3FBD61.8030109@redhat.com> <4A3FC15D.1020705@us.ibm.com>
In-Reply-To: <4A3FC15D.1020705@us.ibm.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
List-Id: qemu-devel.nongnu.org
List-Unsubscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.gnu.org/pipermail/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Anthony Liguori <aliguori@us.ibm.com>
Cc: qemu-devel <qemu-devel@nongnu.org>

On 06/22/2009 08:37 PM, Anthony Liguori wrote:
> I'm saying, always enable dirty tracking from start-of-day with KVM.  
> Then the QEMU dirty bitmap is always accurate.  The trick is to never 
> start resetting it until you need to do live migration.

That disables large pages and incurs a performance penalty (needlessly 
setting bits).

> The idea being that once the dirty bits have been set, the overhead 
> (hopefully) should be zero.
>

It's not zero.  You also need to clear the bits somehow when you balloon.

-- 
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.