From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-116832-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 14194 invoked by alias); 5 Jul 2005 20:36:18 -0000
Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
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
Received: (qmail 14180 invoked by uid 22791); 5 Jul 2005 20:36:14 -0000
Received: from mtagate3.de.ibm.com (HELO mtagate3.de.ibm.com) (195.212.29.152)
    by sourceware.org (qpsmtpd/0.30-dev) with ESMTP; Tue, 05 Jul 2005 20:36:14 +0000
Received: from d12nrmr1607.megacenter.de.ibm.com (d12nrmr1607.megacenter.de.ibm.com [9.149.167.49])
	by mtagate3.de.ibm.com (8.12.10/8.12.10) with ESMTP id j65KaCmf208414
	for <gcc@gcc.gnu.org>; Tue, 5 Jul 2005 20:36:12 GMT
Received: from d12av04.megacenter.de.ibm.com (d12av04.megacenter.de.ibm.com [9.149.165.229])
	by d12nrmr1607.megacenter.de.ibm.com (8.12.10/NCO/VER6.6) with ESMTP id j65KaBvc132066
	for <gcc@gcc.gnu.org>; Tue, 5 Jul 2005 22:36:11 +0200
Received: from d12av04.megacenter.de.ibm.com (loopback [127.0.0.1])
	by d12av04.megacenter.de.ibm.com (8.12.11/8.13.3) with ESMTP id j65KaB0W023381
	for <gcc@gcc.gnu.org>; Tue, 5 Jul 2005 22:36:11 +0200
Received: from d12ml102.megacenter.de.ibm.com (d12ml102.megacenter.de.ibm.com [9.149.166.138])
	by d12av04.megacenter.de.ibm.com (8.12.11/8.12.11) with ESMTP id j65KaBAx023373;
	Tue, 5 Jul 2005 22:36:11 +0200
In-Reply-To: <OF45C5314C.15FDC5D6-ON43257035.004B3EF6-43257035.00565586@il.ibm.com>
Subject: Re: tr1::unordered_set<double> bizarre rounding behavior (x86)
To: Michael Veksler <VEKSLER@il.ibm.com>
Cc: gcc@gcc.gnu.org, Paolo Carlini <pcarlini@suse.de>,
        Gabriel Dos Reis <gdr@integrable-solutions.net>
Message-ID: <OF3D8F4C02.ABB9D5E5-ONC2257035.006FEEE2-C2257035.00712F30@il.ibm.com>
From: Michael Veksler <VEKSLER@il.ibm.com>
Date: Tue, 05 Jul 2005 20:36:00 -0000
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
X-SW-Source: 2005-07/txt/msg00196.txt.bz2





There is one more thing to consider: the ABI.
By changing the code in the header file will break the ABI
of tr1::unordered_set. Code compiled with older gcc and
newer and fixed-gcc will not interoperate.

I don't see an easy path to avoid ABI breakage this time,
however how about preparing for future breakage when
std::tr1::hash<T> changes again?
Maybe move all the specialized hash<T> code from
tr/functional into libstdc++.so? Maybe move specialized
hashtable<double>, unordered_XXX<float> and friends into
libstdc++.so? This way, when hash<T> and
std::equal_to<double> change again, the ABI will be not
as affected.


  Michael.