From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([140.186.70.92]:35096)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1Ri0C4-0004I6-Ot
	for qemu-devel@nongnu.org; Tue, 03 Jan 2012 03:58:05 -0500
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1Ri0C3-0005L2-O0
	for qemu-devel@nongnu.org; Tue, 03 Jan 2012 03:58:04 -0500
Received: from mx1.redhat.com ([209.132.183.28]:62845)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <pbonzini@redhat.com>) id 1Ri0C3-0005Kt-AO
	for qemu-devel@nongnu.org; Tue, 03 Jan 2012 03:58:03 -0500
Message-ID: <4F02C316.2070606@redhat.com>
Date: Tue, 03 Jan 2012 09:57:58 +0100
From: Paolo Bonzini <pbonzini@redhat.com>
MIME-Version: 1.0
References: <1324399916-21315-1-git-send-email-aliguori@us.ibm.com>	<1324399916-21315-2-git-send-email-aliguori@us.ibm.com>
	<4F01F075.2070706@redhat.com> <4F025757.2060106@codemonkey.ws>
In-Reply-To: <4F025757.2060106@codemonkey.ws>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [PATCH 01/27] qom: add the base Object class
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: Anthony Liguori <anthony@codemonkey.ws>
Cc: Anthony Liguori <aliguori@us.ibm.com>, qemu-devel@nongnu.org, Markus Armbruster <armbru@redhat.com>

On 01/03/2012 02:18 AM, Anthony Liguori wrote:
>>
>> 2) Why do you need a list for the interface objects at all? You can make
>> obj->interfaces point to a single blob, and initialize the objects in
>> place. You
>> only need to store the number of interfaces into the TypeImpl or in
>> the object
>> header, or add a dummy Interface at the end whose contents are all-zero.
>
> Interfaces aren't really handled in a special way.  You could easily
> make it so that the infrastructure could be modified to allow true
> multiple inheritance.
>
> I'm not sure I see an obvious benefit to making interfaces handled more
> as a special case.  Right now, the only place that really even knows
> about interfaces is dynamic_cast and initialization.  That seems like a
> good thing to me.

That wouldn't really be a special case, not any more than initializing a 
device's child in-place within the parent struct.

Anyhow, nothing that can't be done as a followup.  Will review v2 now.

Paolo