From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-157842-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 8363 invoked by alias); 9 Nov 2009 12:49:57 -0000
Received: (qmail 8096 invoked by uid 22791); 9 Nov 2009 12:49:57 -0000
X-SWARE-Spam-Status: No, hits=-1.9 required=5.0 	tests=AWL,BAYES_00,SARE_MSGID_LONG40,SPF_PASS
X-Spam-Check-By: sourceware.org
Received: from mail-pz0-f181.google.com (HELO mail-pz0-f181.google.com) (209.85.222.181)     by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Mon, 09 Nov 2009 12:49:50 +0000
Received: by pzk11 with SMTP id 11so1981554pzk.14         for <gcc@gcc.gnu.org>; Mon, 09 Nov 2009 04:49:48 -0800 (PST)
MIME-Version: 1.0
Received: by 10.142.151.22 with SMTP id y22mr752893wfd.100.1257770988549; Mon,  	09 Nov 2009 04:49:48 -0800 (PST)
In-Reply-To: <4AF44C0D.5090902@redhat.com>
References: <ba0bd44d0908280651r676484a5r6b776a2f6d1f12b7@mail.gmail.com>  	<4A97E98C.8040009@redhat.com> <ba0bd44d0911060529r5f6b9a3al83a8dea954a836a5@mail.gmail.com>  	<4AF44C0D.5090902@redhat.com>
From: Mohamed Shafi <shafitvm@gmail.com>
Date: Mon, 09 Nov 2009 12:49:00 -0000
Message-ID: <ba0bd44d0911090449j2d42ffdcuf561aa49e90ae47b@mail.gmail.com>
Subject: Re: How to write shift and add pattern?
To: Richard Henderson <rth@redhat.com>
Cc: GCC <gcc@gcc.gnu.org>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
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: 2009-11/txt/msg00242.txt.bz2

2009/11/6 Richard Henderson <rth@redhat.com>:
> On 11/06/2009 05:29 AM, Mohamed Shafi wrote:
>>
>> =A0 =A0 The target that i am working on has 1& =A02 bit shift-add patter=
ns.
>> GCC is not generating shift-add patterns when the shift count is 1. It
>> is currently generating add operations. What should be done to
>> generate shift-add pattern instead of add-add pattern?
>
> I'm not sure. =A0You may have to resort to matching
>
> =A0(set (match_operand 0 "register_operand" "")
> =A0 =A0 =A0 (plus (plus (match_operand 1 "register_operand" "")
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 (match_dup 1))
> =A0 =A0 =A0 =A0 =A0 =A0 (match_operand 2 "register_operand" ""))))
>
> But you should debug make_compound_operation first to
> figure out what's going on for your port, because it's
> working for x86_64:
>
> =A0 =A0 =A0 =A0long foo(long a, long b) { return a*2 + b; }
>
> =A0 =A0 =A0 =A0leaq =A0 =A0(%rsi,%rdi,2), %rax =A0 =A0 # 8 =A0 =A0 *lea_2=
_rex64
> =A0 =A0 =A0 =A0ret =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0 # 26 =A0 =A0return_internal
>
>
> r~
>

   I have fixed this. The culprit was the cost factor. I added the
case in targetm.rtx_costs and now it works properly. But i am having
issues with the reload.

Regards,
Shafi