From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43)
	id 1MdhkJ-0004ZZ-GW
	for qemu-devel@nongnu.org; Wed, 19 Aug 2009 05:46:19 -0400
Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43)
	id 1MdhkE-0004XG-Bb
	for qemu-devel@nongnu.org; Wed, 19 Aug 2009 05:46:18 -0400
Received: from [199.232.76.173] (port=33670 helo=monty-python.gnu.org)
	by lists.gnu.org with esmtp (Exim 4.43) id 1MdhkE-0004X8-1l
	for qemu-devel@nongnu.org; Wed, 19 Aug 2009 05:46:14 -0400
Received: from mail-bw0-f227.google.com ([209.85.218.227]:49735)
	by monty-python.gnu.org with esmtp (Exim 4.60)
	(envelope-from <filip.navara@gmail.com>) id 1MdhkD-0001Op-Jb
	for qemu-devel@nongnu.org; Wed, 19 Aug 2009 05:46:13 -0400
Received: by bwz27 with SMTP id 27so3421990bwz.34
	for <qemu-devel@nongnu.org>; Wed, 19 Aug 2009 02:46:12 -0700 (PDT)
MIME-Version: 1.0
In-Reply-To: <20090818161424.GA1435@shareable.org>
References: <4A8ACF77.3090801@bandsman.co.uk>
	<20090818161424.GA1435@shareable.org>
Date: Wed, 19 Aug 2009 11:46:11 +0200
Message-ID: <5b31733c0908190246h3385623fid8183b7e8e67500a@mail.gmail.com>
Subject: Re: [Qemu-devel] Arm emulation tweak
From: Filip Navara <filip.navara@gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
List-Id: qemu-devel.nongnu.org
List-Unsubscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.gnu.org/pipermail/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Jamie Lokier <jamie@shareable.org>
Cc: Nigel Horne <njh@bandsman.co.uk>, qemu-devel@nongnu.org

On Tue, Aug 18, 2009 at 6:14 PM, Jamie Lokier<jamie@shareable.org> wrote:
> Nigel Horne wrote:
>> This small patch reduces (on my machine) the call to new_tmp by 16
>> bytes, which may not sound much, but it's called so often that it makes
>> a nice speed up of Arm emulation:
>
>> =A0 =A0 =A0if (GET_TCGV_I32(temps[num_temps]))
>> =A0 =A0 =A0 =A0return temps[num_temps++];
>>
>> ! =A0 =A0 tmp =3D tcg_temp_new_i32();
>> ! =A0 =A0 temps[num_temps++] =3D tmp;
>> ! =A0 =A0 return tmp;
>> =A0}
>
> becomes
>
>> =A0 =A0 =A0if (GET_TCGV_I32(temps[num_temps]))
>> =A0 =A0 =A0 =A0return temps[num_temps++];
>>
>> ! =A0 =A0 return temps[num_temps++] =3D tcg_temp_new_i32();
>> =A0}
>
> That's pretty tragic if GCC fails to compile them to identical code,
> assuming you had optimisation turned on.
>
> Have you tried building with -Os, -O2 etc. to see if that makes a
> difference?
>
> Which GCC version are you using?
>
> -- Jamie
>

BTW, I have even better optimizations for this code at
git://repo.or.cz/qemu/navara.git. These functions are no longer needed
in the form as they appear in the source code today, since TCG has
evolved.

F.