From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:43759)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <benh@kernel.crashing.org>) id 1VtUcm-00043b-VX
	for qemu-devel@nongnu.org; Wed, 18 Dec 2013 22:50:17 -0500
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <benh@kernel.crashing.org>) id 1VtUch-0005Wo-Iy
	for qemu-devel@nongnu.org; Wed, 18 Dec 2013 22:50:12 -0500
Message-ID: <1387424981.15680.21.camel@pasglop>
From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Date: Thu, 19 Dec 2013 14:49:41 +1100
In-Reply-To: <3A95A68A-B1B7-4DD1-8B0D-FDF5CBF1B172@suse.de>
References: <1381488828-22575-1-git-send-email-mark.cave-ayland@ilande.co.uk>
	<155E836F-E5EF-4F8F-AF5B-B8E0EC35553D@suse.de>
	<527D633A.5020801@ilande.co.uk>
	<1E2ED24F-E936-4195-9B3A-C119A9C0582C@suse.de>
	<1387400656.15680.13.camel@pasglop>
	<A5D01862-1290-44DC-8E10-49DB66D443F7@suse.de>
	<1387404245.15680.19.camel@pasglop>
	<3A95A68A-B1B7-4DD1-8B0D-FDF5CBF1B172@suse.de>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [PATCH] PPC: fix PCI configuration space
 MemoryRegions for grackle/uninorth
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: Alexander Graf <agraf@suse.de>
Cc: qemu-ppc <qemu-ppc@nongnu.org>, Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>, QEMU Developers <qemu-devel@nongnu.org>, Andreas =?ISO-8859-1?Q?F=E4rber?= <afaerber@suse.de>, =?ISO-8859-1?Q?Herv=E9?= Poussineau <hpoussin@reactos.org>

On Wed, 2013-12-18 at 23:07 +0100, Alexander Graf wrote:
> On 18.12.2013, at 23:04, Benjamin Herrenschmidt <benh@kernel.crashing.org> wrote:
> 
> > On Wed, 2013-12-18 at 22:24 +0100, Alexander Graf wrote:
> >> Then I don't understand why we break when we limit the data region to
> >> 4 bytes.
> > 
> > This is old uninorth, not U3 HT right ? The latter is memory mapped.
> 
> Depends, we use the same code to cover both. With 32bit guests we expose an old UniNorth.
> With 64bit guests we have to expose a U3 as the guest doesn't know what old UniNorth is anymore.
> 
> So yeah, maybe that's biting us.

Well, it's different.

Old uninorth uses some form of indirect address/data registers, as does
U3 AGP... but U3 HT uses memory mapped. So U3 has a bit of both :)

I think U4 PCIe is yet another beast as well.

Cheers,
Ben.