From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([208.118.235.92]:58262)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1SwSe3-00028J-9x
	for qemu-devel@nongnu.org; Wed, 01 Aug 2012 02:43:00 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1SwSe2-0003F1-DS
	for qemu-devel@nongnu.org; Wed, 01 Aug 2012 02:42:59 -0400
Received: from mx1.redhat.com ([209.132.183.28]:57228)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1SwSe2-0003Eq-5a
	for qemu-devel@nongnu.org; Wed, 01 Aug 2012 02:42:58 -0400
Message-ID: <5018CFDC.4020703@redhat.com>
Date: Wed, 01 Aug 2012 08:42:36 +0200
From: Paolo Bonzini <pbonzini@redhat.com>
MIME-Version: 1.0
References: <1343346200-19850-1-git-send-email-nab@linux-iscsi.org>
	<CAG7+5M35QAK=ggAJRZtga0PWw5VZFi-mraKEAkWiq16U3ysxjg@mail.gmail.com>
In-Reply-To: <CAG7+5M35QAK=ggAJRZtga0PWw5VZFi-mraKEAkWiq16U3ysxjg@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [RFC-v5] tcm_vhost: Initial merge for vhost level
 target fabric driver
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: Eric Northup <digitaleric@google.com>
Cc: Jens Axboe <axboe@kernel.dk>, Anthony Liguori <aliguori@us.ibm.com>, linux-scsi <linux-scsi@vger.kernel.org>, kvm-devel <kvm@vger.kernel.org>, "Michael S. Tsirkin" <mst@redhat.com>, Greg Kroah-Hartman <gregkh@linuxfoundation.org>, qemu-devel <qemu-devel@nongnu.org>, "Nicholas A. Bellinger" <nab@linux-iscsi.org>, lf-virt <virtualization@lists.linux-foundation.org>, Anthony Liguori <aliguori@linux.vnet.ibm.com>, target-devel <target-devel@vger.kernel.org>, Hannes Reinecke <hare@suse.de>, Zhi Yong Wu <wuzhy@cn.ibm.com>, Christoph Hellwig <hch@lst.de>, Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>

Il 31/07/2012 22:52, Eric Northup ha scritto:
> It seems to me like this is not the way that virtio devices are supposed
> to behave - if a guest splits a virtio_scsi_cmd_req or _resp across a
> page boundary, then this code won't work.

Buffers can cover several pages.  Of course, data buffers have to be at
least sector aligned, but this restriction does not apply to
request/response descriptors.

> Quoting the 'Message Framing' part of the virtio spec:
> 
> "In particular, no implementation should use the descriptor boundaries
> to determine the size of any header in a request. "

True,  but this has never matched reality.

Paolo