This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
[libgfortran,committed] Remove use of HAVE_MATH_H in acinclude.m4
- From: FX Coudert <fxcoudert at gmail dot com>
- To: Fortran List <fortran at gcc dot gnu dot org>, gcc-patches list <gcc-patches at gcc dot gnu dot org>
- Date: Sun, 24 Feb 2008 11:59:15 +0000
- Subject: [libgfortran,committed] Remove use of HAVE_MATH_H in acinclude.m4
Checks for broken isfinite(), isnan() and fpclassify() in libgfortran/
acinclude.m4 include <math.h> conditionaly on HAVE_MATH_H being
defined (original patch by Stevel Ellcey, http://gcc.gnu.org/ml/
fortran/2005-09/msg00443.html, okayed by yours truly). Unfortunately,
HAVE_MATH_H is never defined, ever. Thus, the tests are done without
including <math.h>, and can sometimes give wrong results (one example
being powerpc-apple-darwin9.2.0, see comment #28 in PR 32841).
The patch below, committed as rev. 132587 after bootstrap on powerpc-
apple-darwin9.2.0, fixes that issue. (And that's my first 2008
library patch!)
FX
2008-02-24 Francois-Xavier Coudert <fxcoudert@gcc.gnu.org>
PR libfortran/32841
* acinclude.m4: Don't use HAVE_MATH_H.
* configure: Regenerate.
Index: acinclude.m4
===================================================================
--- acinclude.m4 (revision 132578)
+++ acinclude.m4 (working copy)
@@ -204,9 +204,7 @@
libgfor_check_for_broken_isfinite_save_LIBS=$LIBS
LIBS="$LIBS -lm"
AC_TRY_RUN([
-#ifdef HAVE_MATH_H
#include <math.h>
-#endif
#include <float.h>
int main ()
{
@@ -237,9 +235,7 @@
libgfor_check_for_broken_isnan_save_LIBS=$LIBS
LIBS="$LIBS -lm"
AC_TRY_RUN([
-#ifdef HAVE_MATH_H
#include <math.h>
-#endif
#include <float.h>
int main ()
{
@@ -288,9 +284,7 @@
libgfor_check_for_broken_fpclassify_save_LIBS=$LIBS
LIBS="$LIBS -lm"
AC_TRY_RUN([
-#ifdef HAVE_MATH_H
#include <math.h>
-#endif
#include <float.h>
int main ()
{
--
François-Xavier Coudert
http://www.homepages.ucl.ac.uk/~uccafco/