[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[patch] GCC 3.0 configury hack
From: |
Mumit Khan |
Subject: |
[patch] GCC 3.0 configury hack |
Date: |
Tue, 19 Jun 2001 17:08:32 -0500 (CDT) |
[ I'm not subscribe to this list, so please copy me if appropriate ]
This is a sad hack, but one that is rather necessary. Even as libstdc++
matures, I don't really foresee this hack going away anytime soon given
the mess of a task that libstdc++ faces. The other alternative is to
"dumb down" the autoconf macro OCTAVE_CXX_ISO_COMPLIANT_LIBRARY (see
the definition in aclocal.m4), but I personally prefer this workaround,
at least for now.
2001-06-19 Mumit Khan <address@hidden>
* configure.in: Assume that GCC versions >= 3.0 is ISO C++
compliant to work around platform-specific problems.
Index: configure.in
===================================================================
RCS file: /cvs/octave/configure.in,v
retrieving revision 1.348
diff -u -3 -p -r1.348 configure.in
--- configure.in 2001/05/23 06:41:58 1.348
+++ configure.in 2001/06/19 21:26:00
@@ -185,6 +185,22 @@ changequote([,])dnl
;;
esac
+#
+# The test for ISO C++ compliance with GCC versions >= 3.0 (and possibly
+# with RedHat's GCC 2.96, which is based on pre-3.0 snapshot) fails on a
+# few platforms due to lack of integration with the host environment
+# (eg., missing iswblank on Solaris causes error when including cwtype),
+# so we force it here. This is a complete hack, but it's better than
+# having configure guess incorrectly on various platforms, only to have
+# the compilation fail mysteriously later on. For now, we assume that
+# GCC versions >= 3.0 is ISO C++ compliant.
+#
+
+case "$gxx_version" in
+ 1.* | 2.[[0-8]].* | 2.9[[0-6]].*) ;;
+ *) octave_cv_cxx_iso_compliant_library='yes'; ;;
+esac
+
CXX_VERSION=
if test -n "$gxx_version"; then
CXX_VERSION="$gxx_version"
Regards,
Mumit
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [patch] GCC 3.0 configury hack,
Mumit Khan <=