From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([140.186.70.92]:55771)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kwolf@redhat.com>) id 1QmQWK-0006af-1F
	for qemu-devel@nongnu.org; Thu, 28 Jul 2011 09:21:06 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <kwolf@redhat.com>) id 1QmQWJ-0003V9-87
	for qemu-devel@nongnu.org; Thu, 28 Jul 2011 09:21:00 -0400
Received: from mx1.redhat.com ([209.132.183.28]:12603)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kwolf@redhat.com>) id 1QmQWJ-0003Uw-0H
	for qemu-devel@nongnu.org; Thu, 28 Jul 2011 09:20:59 -0400
Message-ID: <4E3162D8.9080702@redhat.com>
Date: Thu, 28 Jul 2011 15:23:36 +0200
From: Kevin Wolf <kwolf@redhat.com>
MIME-Version: 1.0
References: <20110727113000.25109.16204.sendpatchset@skannery>
	<20110727113045.25109.54866.sendpatchset@skannery>
	<4E3015EC.7070004@msgid.tls.msk.ru>
	<CAJSP0QXNNahA5U35QvjsthLTw6tQN5HZd_Odd=obd9Vj_qba7w@mail.gmail.com>
	<4E31389A.4080609@redhat.com>
	<CAJSP0QWeBpUDWcdaPW0_ARROWRaG31qCu-EqXMV1zO3ETnLD9Q@mail.gmail.com>
In-Reply-To: <CAJSP0QWeBpUDWcdaPW0_ARROWRaG31qCu-EqXMV1zO3ETnLD9Q@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [V5 Patch 3/4]Qemu: Command "block_set" for
 dynamic block params change
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: Stefan Hajnoczi <stefanha@gmail.com>
Cc: Supriya Kannery <supriyak@linux.vnet.ibm.com>, Michael Tokarev <mjt@tls.msk.ru>, qemu-devel@nongnu.org, Christoph Hellwig <hch@lst.de>

Am 28.07.2011 15:10, schrieb Stefan Hajnoczi:
> On Thu, Jul 28, 2011 at 11:23 AM, Kevin Wolf <kwolf@redhat.com> wrote:
>> Am 27.07.2011 16:51, schrieb Stefan Hajnoczi:
>>> I'll think about this some more, there are a couple of solutions like
>>> keeping only the file descriptor around, introducing a flush command
>>> that makes sure the file is in a clean state, or changing QED to not
>>> do this.
>>
>> Changing the format drivers doesn't really look like the right solution.
>>
>> Keeping the fd around looks okay, we can probably achieve this by
>> introducing a bdrv_reopen function. It means that we may need to reopen
>> the format layer, but it can't really fail.
> 
> My concern is that this assumes a single file descriptor.  For vmdk
> there may be multiple split files.
> 
> I'm almost starting to think bdrv_reopen() should be recursive down
> the BlockDriverState tree.

Yes, VMDK would have to call bdrv_reopen() for each file that it uses.

Kevin