From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:55555)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1ZuO4y-0007qo-Iu
	for qemu-devel@nongnu.org; Thu, 05 Nov 2015 12:12:07 -0500
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1ZuO4x-0002Vv-NF
	for qemu-devel@nongnu.org; Thu, 05 Nov 2015 12:12:04 -0500
References: <1446084866-12451-1-git-send-email-famz@redhat.com>
	<1446084866-12451-9-git-send-email-famz@redhat.com>
	<20151105163539.GB1702@stefanha-x1.localdomain>
From: Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <563B8DDB.1000304@redhat.com>
Date: Thu, 5 Nov 2015 18:11:55 +0100
MIME-Version: 1.0
In-Reply-To: <20151105163539.GB1702@stefanha-x1.localdomain>
Content-Type: text/plain; charset=iso-8859-15
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [PATCH v2 8/9] block: Introduce
	BlockDriver.bdrv_drain callback
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@redhat.com>, Fam Zheng <famz@redhat.com>
Cc: Kevin Wolf <kwolf@redhat.com>, qemu-block@nongnu.org, Peter Lieven <pl@kamp.de>, qemu-devel@nongnu.org, Ronnie Sahlberg <ronniesahlberg@gmail.com>



On 05/11/2015 17:35, Stefan Hajnoczi wrote:
>> >  void bdrv_drain(BlockDriverState *bs)
>> >  {
>> > +    BdrvChild *child;
>> >      bool busy = true;
>> >  
>> > +    if (bs->drv && bs->drv->bdrv_drain) {
>> > +        bs->drv->bdrv_drain(bs);
>> > +    }
>> > +    QLIST_FOREACH(child, &bs->children, next) {
>> > +        bdrv_drain(child->bs);
>> > +    }
>> >      while (busy) {
>> >          /* Keep iterating */
>> >           bdrv_flush_io_queue(bs);
> Recursing calls aio_poll() once for each BDS node and I think it's not
> necessary and could be a performance problem.

Indeed, only the call to bs->drv->bdrv_drain should be done recursively.

Paolo