From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-170872-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 27604 invoked by alias); 24 Oct 2011 16:19:55 -0000
Received: (qmail 27593 invoked by uid 22791); 24 Oct 2011 16:19:53 -0000
X-SWARE-Spam-Status: No, hits=-2.2 required=5.0	tests=AWL,BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FROM,RCVD_IN_DNSWL_LOW,TW_FC
X-Spam-Check-By: sourceware.org
Received: from mail-iy0-f175.google.com (HELO mail-iy0-f175.google.com) (209.85.210.175)    by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Mon, 24 Oct 2011 16:19:37 +0000
Received: by iafi7 with SMTP id i7so8974450iaf.20        for <gcc@gcc.gnu.org>; Mon, 24 Oct 2011 09:19:36 -0700 (PDT)
MIME-Version: 1.0
Received: by 10.42.115.133 with SMTP id k5mr5232768icq.16.1319473176462; Mon, 24 Oct 2011 09:19:36 -0700 (PDT)
Received: by 10.42.223.65 with HTTP; Mon, 24 Oct 2011 09:19:36 -0700 (PDT)
In-Reply-To: <CAFiYyc0xGcf5iphRq2Z59-R4GG6wUvWHMuP_GfCUTSqyokZD+w@mail.gmail.com>
References: <7FB04A5C213E9943A72EE127DB74F0ADD15FEADD80@SJEXCHCCR02.corp.ad.broadcom.com>	<CAFiYyc0xGcf5iphRq2Z59-R4GG6wUvWHMuP_GfCUTSqyokZD+w@mail.gmail.com>
Date: Tue, 25 Oct 2011 03:12:00 -0000
Message-ID: <CA+=Sn1mQ=XHw+J_mtVN=g54T3FNjagGOxNyO6ZPc1tf6_4_DqQ@mail.gmail.com>
Subject: Re: Why doesn't GCC generate conditional move for COND_EXPR?
From: Andrew Pinski <pinskia@gmail.com>
To: Richard Guenther <richard.guenther@gmail.com>
Cc: Bingfeng Mei <bmei@broadcom.com>, "gcc@gcc.gnu.org" <gcc@gcc.gnu.org>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
X-IsSubscribed: yes
Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc/>
List-Post: <mailto:gcc@gcc.gnu.org>
List-Help: <http://gcc.gnu.org/ml/>
Sender: gcc-owner@gcc.gnu.org
X-SW-Source: 2011-10/txt/msg00430.txt.bz2

On Mon, Oct 24, 2011 at 7:00 AM, Richard Guenther
<richard.guenther@gmail.com> wrote:
> On Mon, Oct 24, 2011 at 2:55 PM, Bingfeng Mei <bmei@broadcom.com> wrote:
>> Hello,
>> I noticed that COND_EXPR is not expanded to conditional move
>> as MIN_EXPR/MAX_EXPR are (assuming movmodecc is available).
>> I wonder why not?
>>
>> I have some loop that fails tree vectorization, but still contains
>> COND_EXPR from tree ifcvt pass. In the end, the generated code
>> is worse than if I don't turned -ftree-vectorize on. =C2=A0This
>> is on our private port.
>
> Because nobody touched COND_EXPR expansion since ages.

I have a patch which I will be submitting next week or so that does
this expansion correctly.  In fact I have a few patches which improves
the generation of COND_EXPR in simple cases (in PHI-OPT).

Thanks,
Andrew Pinski