From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-143334-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 28166 invoked by alias); 6 Jan 2008 20:42:00 -0000
Received: (qmail 28156 invoked by uid 22791); 6 Jan 2008 20:41:59 -0000
X-Spam-Check-By: sourceware.org
Received: from caffeine.csclub.uwaterloo.ca (HELO caffeine.csclub.uwaterloo.ca) (129.97.134.17)     by sourceware.org (qpsmtpd/0.31) with ESMTP; Sun, 06 Jan 2008 20:41:42 +0000
Received: from caffeine.csclub.uwaterloo.ca (localhost [127.0.0.1]) 	by caffeine.csclub.uwaterloo.ca (Postfix) with ESMTP id 4CEBE73D41 	for <gcc@gcc.gnu.org>; Sun,  6 Jan 2008 15:41:40 -0500 (EST)
Received: by caffeine.csclub.uwaterloo.ca (Postfix, from userid 20006) 	id 3F89B73D49; Sun,  6 Jan 2008 15:41:40 -0500 (EST)
To: gcc@gcc.gnu.org
Subject: Re: __builtin_expect for indirect function calls
Message-Id: <20080106204140.3F89B73D49@caffeine.csclub.uwaterloo.ca>
Date: Sun, 06 Jan 2008 20:42:00 -0000
From: rridge@csclub.uwaterloo.ca (Ross Ridge)
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: 2008-01/txt/msg00050.txt.bz2

Mark Mitchell writes:
> What do people think?  Do we have the leeway to change this?

If it were just cases where using __builtin_expect is pointless that
would break, like function overloading and sizeof then I don't think
it would be a problem.  However, it would change behaviour when C++
conversion operators are used and I can see these being legitimately
used with __builtin_expect.  Something like:

	struct C {
		operator long();
	};

	int foo() {
		if (__builtin_expect(C(), 0))
			return 1;
		return 2;
	}

If cases like these are rare enough it's probably an acceptable change
if they give an error because the argument types don't match.

					Ross Ridge