From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:60035)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kwolf@redhat.com>) id 1Vr2gW-0002h1-UF
	for qemu-devel@nongnu.org; Thu, 12 Dec 2013 04:35:58 -0500
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <kwolf@redhat.com>) id 1Vr2gV-0007IE-PI
	for qemu-devel@nongnu.org; Thu, 12 Dec 2013 04:35:56 -0500
Received: from mx1.redhat.com ([209.132.183.28]:51702)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kwolf@redhat.com>) id 1Vr2gV-0007I6-Fj
	for qemu-devel@nongnu.org; Thu, 12 Dec 2013 04:35:55 -0500
Date: Thu, 12 Dec 2013 10:35:45 +0100
From: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20131212093545.GE10007@dhcp-200-207.str.redhat.com>
References: <1386796109-15264-1-git-send-email-kwolf@redhat.com>
	<1386796109-15264-2-git-send-email-kwolf@redhat.com>
	<52A954FB.7090603@kamp.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <52A954FB.7090603@kamp.de>
Subject: Re: [Qemu-devel] [PATCH 01/22] block: Move initialisation of
 BlockLimits to bdrv_refresh_limits()
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: pbonzini@redhat.com, qemu-devel@nongnu.org, stefanha@redhat.com

Am 12.12.2013 um 07:17 hat Peter Lieven geschrieben:
> Am 11.12.2013 22:08, schrieb Kevin Wolf:
> > This function separates filling the BlockLimits from bdrv_open(), which
> > allows it to call it from other operations which may change the limits
> > (e.g. modifications to the backing file chain or bdrv_reopen)
> >
> > Signed-off-by: Kevin Wolf <kwolf@redhat.com>

> > +static int iscsi_refresh_limits(BlockDriverState *bs)
> > +{
> > +    IscsiLun *iscsilun = bs->opaque;
> > +    struct scsi_task *task = NULL;
> > +    int ret;
> > +
> > +    memset(&bs->bl, 0, sizeof(bs->bl));
> > +
> you do that memset in bdrv_refresh_limits already.

Right, I'll drop it here.

> >      if (iscsilun->lbp.lbpu || iscsilun->lbp.lbpws) {
> >          struct scsi_inquiry_block_limits *inq_bl;
> >          task = iscsi_do_inquiry(iscsilun->iscsi, iscsilun->lun, 1,
> > @@ -1462,48 +1512,14 @@ static int iscsi_open(BlockDriverState *bs, QDict *options, int flags,
> 
> iscsi_do_inquiry does a sync call to request the pages from the target.
> you can only do that here if you can guarantee that iscsi_refresh_limits will
> be called with no pending requests in-flight otherwise this will end in a mess.
> 
> otherwise i would propose to leave the inquiry call ins iscsi_open and just fill
> the bs->bl struc tin iscsi_refresh_limits. the problem is if we make this async
> we have to handle all possible I/O callbacks whiile in iscsi_refresh_limits.
> i do not see how an iscsi target can change its limits while a target is mounted.

Yes, I'll just split the block in two, leaving the actual request in
iscsi_open().

Thanks for your review.

Kevin