From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:60107)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <hpoussin@reactos.org>) id 1VGaix-0001qQ-Tl
	for qemu-devel@nongnu.org; Mon, 02 Sep 2013 16:27:56 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <hpoussin@reactos.org>) id 1VGaip-0001j9-GK
	for qemu-devel@nongnu.org; Mon, 02 Sep 2013 16:27:47 -0400
Received: from smtp1-g21.free.fr ([2a01:e0c:1:1599::10]:35313)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <hpoussin@reactos.org>) id 1VGaig-0001UJ-Rh
	for qemu-devel@nongnu.org; Mon, 02 Sep 2013 16:27:39 -0400
Message-ID: <5224F473.2060208@reactos.org>
Date: Mon, 02 Sep 2013 22:26:27 +0200
From: =?UTF-8?B?SGVydsOpIFBvdXNzaW5lYXU=?= <hpoussin@reactos.org>
MIME-Version: 1.0
References: <1377283973-9320-1-git-send-email-hpoussin@reactos.org>
	<1377283973-9320-4-git-send-email-hpoussin@reactos.org>
	<5217BE5F.5000708@twiddle.net>
In-Reply-To: <5217BE5F.5000708@twiddle.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: quoted-printable
Subject: Re: [Qemu-devel] [PATCH 3/3] prep: improve Raven PCI host emulation
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: Richard Henderson <rth@twiddle.net>
Cc: =?UTF-8?B?QW5kcmVhcyBGw6RyYmVy?= <andreas.faerber@web.de>, qemu-devel@nongnu.org

Richard Henderson a =C3=A9crit :
> On 08/23/2013 11:52 AM, Herv=C3=A9 Poussineau wrote:
>> +    uint8_t buf[4];
>> +    uint64_t val;
>> +
>> +    if (s->contiguous_map =3D=3D 0) {
>> +        /* 64 KB contiguous space for IOs */
>> +        addr &=3D 0xFFFF;
>> +    } else {
>> +        /* 8 MB non-contiguous space for IOs */
>> +        addr =3D (addr & 0x1F) | ((addr & 0x007FFF000) >> 7);
>> +    }
>> +
>> +    address_space_read(&s->pci_io_as, addr + 0x80000000, buf, size);
>> +    memcpy(&val, buf, size);
>> +    return val;
>=20
> This memcpy can't be right, especially for big-endian host.

pci_io_as is supposed to contain words/longs in little-endian mode (PCI=20
is always little-endian).
prep_io_ops endianness is DEVICE_LITTLE_ENDIAN, so, AFAIK, this means=20
that returned val should be LE.
Why should it break on big-endian hosts?

Regards,

Herv=C3=A9