From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([140.186.70.92]:33028)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1R5Xpw-0007Ww-In
	for qemu-devel@nongnu.org; Mon, 19 Sep 2011 03:00:17 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1R5Xpv-0007qG-Ge
	for qemu-devel@nongnu.org; Mon, 19 Sep 2011 03:00:16 -0400
Message-ID: <4E76E86C.3010707@redhat.com>
Date: Mon, 19 Sep 2011 08:59:56 +0200
From: Paolo Bonzini <pbonzini@redhat.com>
MIME-Version: 1.0
References: <1315989802-18753-1-git-send-email-agraf@suse.de>
	<1315989802-18753-2-git-send-email-agraf@suse.de>
	<20110915031437.GR9025@yookeroo.fritz.box>
	<4E71A2D2.2080008@redhat.com>
	<20110916030616.GZ9025@yookeroo.fritz.box>
	<4E7327E4.7040708@redhat.com>
	<20110916152731.27f76f7d@BR8GGW75.de.ibm.com>
	<1316188291.2777.25.camel@pasglop>
	<20110919085555.67e7e4b8@BR8GGW75.de.ibm.com>
In-Reply-To: <20110919085555.67e7e4b8@BR8GGW75.de.ibm.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [Qemu-ppc] [PATCH 01/58] spapr: proper
	qdevification
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: Thomas Huth <thuth@linux.vnet.ibm.com>
Cc: David Gibson <david@gibson.dropbear.id.au>, qemu-ppc@nongnu.org, Alexander Graf <agraf@suse.de>, qemu-devel Developers <qemu-devel@nongnu.org>

On 09/19/2011 08:55 AM, Thomas Huth wrote:
>> >  Note that in addition to that, the PAPR spec specifies only one
>> >  "device" (whatever that means) per vscsi instance.
> Really? In that case, I wonder why Linux is using the "Logical unit
> addressing format" with target IDs and bus numbers instead of the
> "Flat space addressing method" for vscsi ... according to
> drivers/scsi/ibmvscsi/ibmvscsi.c :
>
> static inline u16 lun_from_dev(struct scsi_device *dev)
> {
> 	return (0x2<<  14) | (dev->id<<  8) | (dev->channel<<  5) | dev->lun;
> }
>
> In case there's really only one device per vscsi instance, shouldn't
> that code use addressing method 0x1 instead of 0x2 here?

As long as dev->id == 0, dev->channel == 0, dev->lun < 31, the three 
addressing methods are all equivalent.

Some comments in ibmvscsi.c say that iOS needs non-zero channels, so 
there does seem to be someone else who doesn't follow the PAPR spec too 
well. :)

Paolo