From mboxrd@z Thu Jan  1 00:00:00 1970
From: Olaf Hering <olaf@aepfle.de>
Subject: Re: [PATCH v4 3/5] libxl: add support for vscsi
Date: Wed, 22 Apr 2015 10:08:52 +0200
Message-ID: <20150422080852.GA12884@aepfle.de>
References: <1429259460-30491-1-git-send-email-olaf@aepfle.de>
	<1429259460-30491-4-git-send-email-olaf@aepfle.de>
	<20150421140537.GC32209@l.oracle.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Return-path: <xen-devel-bounces@lists.xen.org>
Content-Disposition: inline
In-Reply-To: <20150421140537.GC32209@l.oracle.com>
List-Unsubscribe: <http://lists.xen.org/cgi-bin/mailman/options/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xen.org>
List-Help: <mailto:xen-devel-request@lists.xen.org?subject=help>
List-Subscribe: <http://lists.xen.org/cgi-bin/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=subscribe>
Sender: xen-devel-bounces@lists.xen.org
Errors-To: xen-devel-bounces@lists.xen.org
To: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Ian Jackson <ian.jackson@eu.citrix.com>, Stefano Stabellini <stefano.stabellini@eu.citrix.com>, Wei Liu <wei.liu2@citrix.com>, Ian Campbell <ian.campbell@citrix.com>, xen-devel@lists.xen.org
List-Id: xen-devel@lists.xenproject.org

On Tue, Apr 21, Konrad Rzeszutek Wilk wrote:

> On Fri, Apr 17, 2015 at 08:30:58AM +0000, Olaf Hering wrote:
> > +=item B<vscsi=[ "VSCSI_SPEC_STRING", "VSCSI_SPEC_STRING", ...]>
> > +
> > +Specifies the PVSCSI devices to be provided to the guest. PVSCSI passes
> > +dom0 SCSI devices as-is to the guest.
> s/dom0/backend/

I will make that change.

> As in theory you could run an HVM guest with an SCSI device passed in - and
> use said backed to pass the SCSI device to another guest.

Sounds interesting. I assume you meant "backend"? How will the domU talk
to the other domU? Not sure if my libxl change covers that use case, and
how I would be able to verify such functionality. Hopefully merging my
series will not depend on such use case.

> 'as-is' ? I thought there were some filtering done in the backend?

Perhaps 'passthrough with filtering'? I will remove it.

> > +Each VSCSI_SPEC_STRING consists of "pdev,vdev[,options]".
> > +'pdev' describes the physical device, preferable in a persistant format.
> What is 'persistent' format?

Anything that uses identifiers provided by the hardware, like serial
numbers. /dev/disk/by-*/* for example. The h:c:t:l notation is not
persistant because the "h" part will change across reboots.

> > +
> > +=item C<option>
> > +
> > +No options recognized.
> > +
> > +=back
> > +
> > +=item B<Linux xenlinux>
> xenlinux? Classic Xen?

Anything based on linux-2.6.18-xen.hg. I will adjust this to "Linux
based on classic xen kernel"

> > +static int libxl__device_vscsi_new_backend(libxl__egc *egc,
> > +                                           libxl__ao_device *aodev,
> > +                                           libxl_device_vscsi *vscsi,
> > +                                           libxl_domain_config *d_config)
> > +{

> > +        rc = libxl__device_vscsi_dev_backend_set(gc, v, back);
> > +        if (rc) goto out;
> Don't you want 'return rc' as the out will try to call libxl__xs_transaction_abort?

The current code works because t is initialized to XBT_NULL, and
libxl__xs_transaction_abort will do nothing in that case. I will adjust
this as you suggested.

Olaf