From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 5056 invoked by alias); 16 Oct 2002 22:54:40 -0000 Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-owner@gcc.gnu.org Received: (qmail 5047 invoked from network); 16 Oct 2002 22:54:39 -0000 Received: from unknown (HELO moshier.moshier.net) (24.61.24.32) by sources.redhat.com with SMTP; 16 Oct 2002 22:54:39 -0000 Received: from www.moshier.net (moshier.ne.client2.attbi.com [24.61.24.32]) by moshier.moshier.net (8.11.6/8.11.6/SuSE Linux 0.5) with ESMTP id g9GMsZm22440; Wed, 16 Oct 2002 18:54:35 -0400 Date: Wed, 16 Oct 2002 17:20:00 -0000 From: Stephen L Moshier To: Richard Henderson cc: gcc@gcc.gnu.org Subject: Re: real.c fails floating point tests In-Reply-To: <20021016220028.GB30823@redhat.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-SW-Source: 2002-10/txt/msg01023.txt.bz2 > The harness in contrib/paranoia.cc calls into real_arithmetic > and real_compare, which calls each of these functions. That does not test the 160-bit internal arithmetic, which is what the posted report was all about. We seem to be talking about different things there. Also, beware that contrib/paranoia.cc is currently a watered-down test because the log, exp, pow functions use the host computer's math library instead of the arithmetic supposedly under test. I am mainly concerned about the decimal <-> binary conversion, which does exercise the internal arithmetic. I'll continue to test and post decimal-binary cases as time permits.