From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugs-return-440128-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 30358 invoked by alias); 12 Jan 2014 15:06:44 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Received: (qmail 30325 invoked by uid 48); 12 Jan 2014 15:06:39 -0000
From: "dominiq at lps dot ens.fr" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug libfortran/59774] [Regression] Inconsistent rounding between -m32 and -m64
Date: Sun, 12 Jan 2014 15:06:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: libfortran
X-Bugzilla-Version: 4.9.0
X-Bugzilla-Keywords:
X-Bugzilla-Severity: normal
X-Bugzilla-Who: dominiq at lps dot ens.fr
X-Bugzilla-Status: NEW
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: jvdelisle at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields:
Message-ID: <bug-59774-4-auGjwCY4M8@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-59774-4@http.gcc.gnu.org/bugzilla/>
References: <bug-59774-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2014-01/txt/msg01270.txt.bz2

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59774

--- Comment #5 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
For the tests in comment 1, I think PR323 strikes back again. I the same
results for both -m32 and -m64 with the following patch

--- ../_clean/libgfortran/io/write_float.def    2014-01-04 15:51:53.000000000
+0100
+++ libgfortran/io/write_float.def    2014-01-12 15:55:24.000000000 +0100
@@ -1043,10 +1043,11 @@ output_float_FMT_G_ ## x (st_parameter_d
     break;\
     }\
 \
-  rexp_d = calculate_exp_ ## x (-d);\
-  if ((m > 0.0 && ((m < 0.1 - 0.1 * r * rexp_d) || (rexp_d * (m + r) >=
1.0)))\
+  rexp_d = calculate_exp_ ## x (d);\
+  if ((m > 0.0 && ((10.0 * rexp_d * m < rexp_d - r ) || ((m + r) >= rexp_d)))\
       || ((m == 0.0) && !(compile_options.allow_std\
-              & (GFC_STD_F2003 | GFC_STD_F2008))))\
+              & (GFC_STD_F2003 | GFC_STD_F2008)))\
+      ||  d == 0)\
     { \
       newf.format = FMT_E;\
       newf.u.real.w = w;\
@@ -1069,7 +1070,7 @@ output_float_FMT_G_ ## x (st_parameter_d
       volatile GFC_REAL_ ## x temp;\
       mid = (low + high) / 2;\
 \
-      temp = (calculate_exp_ ## x (mid - 1) * (1 - r * rexp_d));\
+      temp = (calculate_exp_ ## x (mid - 1) * (rexp_d - r) / rexp_d);\
 \
       if (m < temp)\
         { \

I think in order to avoid overflows (10.0 * rexp_d * m < rexp_d - r ) and ((m +
r) >= rexp_d) should be exchanged.