From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-patches-return-293068-listarch-gcc-patches=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 28659 invoked by alias); 28 May 2011 14:19:52 -0000
Received: (qmail 28648 invoked by uid 22791); 28 May 2011 14:19:51 -0000
X-SWARE-Spam-Status: No, hits=-2.3 required=5.0	tests=AWL,BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FROM,RCVD_IN_DNSWL_LOW,RFC_ABUSE_POST
X-Spam-Check-By: sourceware.org
Received: from mail-qy0-f175.google.com (HELO mail-qy0-f175.google.com) (209.85.216.175)    by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Sat, 28 May 2011 14:19:38 +0000
Received: by qyk35 with SMTP id 35so236058qyk.20        for <gcc-patches@gcc.gnu.org>; Sat, 28 May 2011 07:19:37 -0700 (PDT)
MIME-Version: 1.0
Received: by 10.229.118.72 with SMTP id u8mr2274579qcq.1.1306592377394; Sat, 28 May 2011 07:19:37 -0700 (PDT)
Received: by 10.229.11.143 with HTTP; Sat, 28 May 2011 07:19:37 -0700 (PDT)
In-Reply-To: <BANLkTimh+Dr2M7BR7xkf-TAysE=VHt94qA@mail.gmail.com>
References: <BANLkTimh+Dr2M7BR7xkf-TAysE=VHt94qA@mail.gmail.com>
Date: Sat, 28 May 2011 17:25:00 -0000
Message-ID: <BANLkTimFPUmuGtOo19f__6M0iWAjZn+VMQ@mail.gmail.com>
Subject: Re: [patch] [2/2] Support reduction in loop SLP
From: "H.J. Lu" <hjl.tools@gmail.com>
To: Ira Rosen <ira.rosen@linaro.org>
Cc: gcc-patches@gcc.gnu.org, Patch Tracking <patches@linaro.org>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
X-IsSubscribed: yes
Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-patches.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-patches/>
List-Post: <mailto:gcc-patches@gcc.gnu.org>
List-Help: <mailto:gcc-patches-help@gcc.gnu.org>
Sender: gcc-patches-owner@gcc.gnu.org
X-SW-Source: 2011-05/txt/msg02239.txt.bz2

On Tue, May 17, 2011 at 11:03 PM, Ira Rosen <ira.rosen@linaro.org> wrote:
> This part adds the actual code for reduction support.
>
> Bootstrapped and tested on powerpc64-suse-linux.
> I am planning to apply it later today.
>
> Ira
>
> ChangeLog:
>
> =A0 =A0 =A0 =A0PR tree-optimization/41881
> =A0 =A0 =A0 =A0* tree-vectorizer.h (struct _loop_vec_info): Add new field
> reduction_chains along with a macro for
> =A0 =A0 =A0 =A0its access.
> =A0 =A0 =A0 =A0* tree-vect-loop.c (new_loop_vec_info): Initialize reducti=
on chains.
> =A0 =A0 =A0 =A0(destroy_loop_vec_info): Free reduction chains.
> =A0 =A0 =A0 =A0(vect_analyze_loop_2): Return false if vect_analyze_slp() =
returns false.
> =A0 =A0 =A0 =A0(vect_is_slp_reduction): New function.
> =A0 =A0 =A0 =A0(vect_is_simple_reduction_1): Call vect_is_slp_reduction.
> =A0 =A0 =A0 =A0(vect_create_epilog_for_reduction): Support SLP reduction =
chains.
> =A0 =A0 =A0 =A0* tree-vect-slp.c (vect_get_and_check_slp_defs): Allow dif=
ferent
> definition types for reduction
> =A0 =A0 =A0 =A0chains.
> =A0 =A0 =A0 =A0(vect_supported_load_permutation_p): Don't allow permutati=
ons for
> reduction chains.
> =A0 =A0 =A0 =A0(vect_analyze_slp_instance): Support reduction chains.
> =A0 =A0 =A0 =A0(vect_analyze_slp): Try to build SLP instance from reducti=
on chains.
> =A0 =A0 =A0 =A0(vect_get_constant_vectors): =A0Handle reduction chains.
> =A0 =A0 =A0 =A0(vect_schedule_slp_instance): Mark the first statement of =
the
> reduction chain as reduction.
>
> testsuite/ChangeLog:
>
> =A0 =A0 =A0 =A0PR tree-optimization/41881
> =A0 =A0 =A0 =A0* gcc.dg/vect/O3-pr41881.c: New test.
> =A0 =A0 =A0 =A0* gcc.dg/vect/O3-slp-reduc-10.c: New test.
>

This caused:

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D49199

--=20
H.J.