From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-163469-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 31267 invoked by alias); 28 Jul 2010 15:53:33 -0000
Received: (qmail 31113 invoked by uid 22791); 28 Jul 2010 15:53:32 -0000
X-SWARE-Spam-Status: No, hits=-6.2 required=5.0	tests=AWL,BAYES_00,DKIM_SIGNED,DKIM_VALID,RCVD_IN_DNSWL_HI,SPF_HELO_PASS,T_RP_MATCHES_RCVD
X-Spam-Check-By: sourceware.org
Received: from mail1.commerzbank.com (HELO mail1.commerzbank.com) (212.149.48.100)    by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Wed, 28 Jul 2010 15:53:23 +0000
Received: by mail1.commerzbank.com (Commerzbank Mail-System, from userid 1002)	id 5FB061A83EE; Wed, 28 Jul 2010 17:53:20 +0200 (CEST)
X-Spam-Virus: No
Received: from mail1.commerzbank.com (localhost [127.0.0.1])	by mail1.commerzbank.com (Commerzbank Mail-System) with ESMTP id AB7281A83DA;	Wed, 28 Jul 2010 17:53:16 +0200 (CEST)
Received: from pfx1.commerzbank.com (intern.postfix.commerzbank.com [172.16.71.213])	by mail1.commerzbank.com (Commerzbank Mail-System) with ESMTPS id A55221A83D6;	Wed, 28 Jul 2010 17:53:16 +0200 (CEST)
Received: by pfx1.commerzbank.com (Commerzbank Internal Mail-System, from userid 1001)	id 838553B561; Wed, 28 Jul 2010 17:53:16 +0200 (CEST)
Received: from sv035511.ztb.icb.commerzbank.com (sv035511.ztb.icb.commerzbank.com [140.13.159.110]) 	by pfx1.commerzbank.com (Commerzbank Internal Mail-System) with ESMTP id 6BAE72B2C7; 	Wed, 28 Jul 2010 17:53:16 +0200 (CEST)
Received: from SE000319.ztb.icb.commerzbank.com ([140.26.56.30]) by sv035511.ztb.icb.commerzbank.com with Microsoft SMTPSVC(6.0.3790.4675); 	 Wed, 28 Jul 2010 17:53:16 +0200
MIME-Version: 1.0
Subject: RE: optimization question: mpl
Date: Wed, 28 Jul 2010 15:53:00 -0000
Message-ID: <824586E7387D1443B5DD024DAC75B91C139049@SE000319.ztb.icb.commerzbank.com>
In-Reply-To: <AANLkTinSN+dOi-oi1ZWBa+4P55+Wvn=DtPasMJ_usOUb@mail.gmail.com>
References: <824586E7387D1443B5DD024DAC75B91C139048@SE000319.ztb.icb.commerzbank.com> <AANLkTinSN+dOi-oi1ZWBa+4P55+Wvn=DtPasMJ_usOUb@mail.gmail.com>
From: "Hite, Christopher" <Christopher.Hite@partner.commerzbank.com>
To: "Richard Guenther" <richard.guenther@gmail.com>
Cc: <gcc@gcc.gnu.org>
Content-class: urn:content-classes:message
Content-Type: text/plain; 	charset="us-ascii"
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: 2010-07/txt/msg00404.txt.bz2

> Generally without knowing the compiler version you are using
> it is hard to tell.=20=20
I'll use whatever's best.  Right now I'm still on 4.4.3.  I'll probably
upgrade soon to 4.5.

> The same is true without a complete compilable testcase.
I didn't want to make a test case that depends on boost::mpl. How's
this:

struct DecodeContext;

struct M1{
	static const int id=3D1;
	static void decode(DecodeContext& ){}
};

struct M2{
	static const int id=3D2;
	static void decode(DecodeContext& ){}
};

struct M3{
	static const int id=3D3;
	static void decode(DecodeContext& ){}
};


template <int> struct ListMember;

template <> struct ListMember<1>{ typedef M1 type; };
template <> struct ListMember<2>{ typedef M2 type; };
template <> struct ListMember<3>{ typedef M3 type; };

template<int i>
void foo(int id, DecodeContext& dc) {
	typedef typename ListMember<i>::type M;
	if(M::id=3D=3Did)
		M::decode(dc);
	else
		foo<i-1>(id,dc);
}

template<>
void foo<0>(int id, DecodeContext& dc) {}



int main(){
	DecodeContext& dc=3D *(DecodeContext*)0;// junk for now
	int id=3D2; //sometime runtime dependent
	foo<3>(id,dc);
	return 0;
}


> You can use the flatten attribute to tell the compiler to inline all
> calls in a given function, like
>=20
> void __attribute__((flatten)) foo(void)
> {
> ...
> decode1();
> ...
> }

That would cause decode1() to be inlined, which might not be what you
want.

Hmm maybe I could rewrite things so the switch case returns a function
pointer.  I'm guessing that would make things slower though.