From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([208.118.235.92]:44851)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kraxel@redhat.com>) id 1T9Ym8-0003Ja-7X
	for qemu-devel@nongnu.org; Thu, 06 Sep 2012 05:53:31 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <kraxel@redhat.com>) id 1T9Ym2-0003Sp-FD
	for qemu-devel@nongnu.org; Thu, 06 Sep 2012 05:53:28 -0400
Received: from mx1.redhat.com ([209.132.183.28]:28117)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kraxel@redhat.com>) id 1T9Ym2-0003SO-7R
	for qemu-devel@nongnu.org; Thu, 06 Sep 2012 05:53:22 -0400
Message-ID: <5048728B.2010902@redhat.com>
Date: Thu, 06 Sep 2012 11:53:15 +0200
From: Gerd Hoffmann <kraxel@redhat.com>
MIME-Version: 1.0
References: <1345604699-24851-1-git-send-email-david@gibson.dropbear.id.au>
In-Reply-To: <1345604699-24851-1-git-send-email-david@gibson.dropbear.id.au>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Subject: Re: [Qemu-devel] [PATCH] usb: Fix host-side endian bugs in hcd-ehci
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: David Gibson <david@gibson.dropbear.id.au>
Cc: aliguori@us.ibm.com, qemu-devel@nongnu.org

On 08/22/12 05:04, David Gibson wrote:
> The EHCI device model is horribly broken for big-endian hosts.  It uses a
> union of 'mmio' a byte array which is as-is as the device's MMIO space
> with the various internal registers.  The IO routines assume that mmio is
> laid out in little-endian order, but everything else in the code accesses
> the register variables directly assuming (host) native endian.
> 
> This fix is fairly ugly - a nicer approach would involve converting
> hcd-ehci to use the new-style memory region .read and .write functions -
> but it's the most minimal fix I can see to apply for the qemu 1.2 release.

Oops.  This one slipped through.  Just found it -- to late for 1.2 --
while wading through my inbox looking for unprocessed stuff.

Doing the real thing instead (i.e. convert memory regions) doesn't look
that horrible though, I think we can put that into stable-1.2.  I'll
send out the patch in a few moments.

cheers,
  Gerd