Hello leif,
I'm using gcc 5.2.0 under ArchLinux. Issues with ncurses and gf2x remain.
I've created and tested a patch (attached) to extend some of the existing
fixes for gcc 5.0 and 5.1.
A post at sage-support [1] might also be relevant.
Best,
Fidel
[1]: https://groups.google.com/d/topic/sage-support/9C-C5LEqf7A/discussion
On Wednesday, April 22, 2015 at 12:11:30 PM UTC-4, leif wrote:
>
> On 04/22/2015 05:30 AM, leif wrote:
> > Release of GCC 5.1 is scheduled for Wednesday, April 22nd.
>
> It's out now:
>
> https://gcc.gnu.org/ml/gcc/2015-04/msg00287.html
>
>
> The errors with ncurses and gf2x remain. 8-/
>
>
> -leif
>
>
> > I took the first release candidate and tried to build Sage 6.6 with it:
> >
> > * MPIR fails in 'configure' -- easy to fix, see #18247
> >
> > * Lcalc fails to build -- usual C++ issues, subset of those
> > with clang, easy to fix, cf. #12437
> > ('-fpermissive' also suffices)
> >
> > * Sage library -- modules using Lcalc headers won't build for the
> > same reason (build with '-fpermissive' in
> > CFLAGS; distutils ignores CXXFLAGS)
> >
> > * ncurses (a superfluous package anyway) fails to build with
> > a syntax error(!) -- apparently a bug in the preprocessor,
> > will hopefully vanish with the final;
> > setting CPP to use another one works
> >
> > * gf2x fails in tuning -- results don't agree; presumably (but
> > not necessarily) a compiler bug,
> > '-O1' fails in the same way, '-O0'
> > succeeds; will wait for and retry
> > with the final, otherwise probably
> > harder to debug
> >
> > After all, all tests (ptestlong) passed (modulo #17907).
>
>
>
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at http://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.
>From 67c76ca18bfba62610a9dc3dfe5215422d836a6c Mon Sep 17 00:00:00 2001
From: fidbc <fidbc.c...@gmail.com>
Date: Wed, 29 Jul 2015 23:56:43 -0400
Subject: [PATCH] Added fix for gcc 5.2.0
---
build/pkgs/gf2x/spkg-install | 2 +-
.../ncurses/patches/work_around_changed_output_of_GNU_cpp_5.x.patch | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/build/pkgs/gf2x/spkg-install b/build/pkgs/gf2x/spkg-install
index ee7e6a0..bf9d2e7 100755
--- a/build/pkgs/gf2x/spkg-install
+++ b/build/pkgs/gf2x/spkg-install
@@ -26,7 +26,7 @@ touch aclocal.m4 configure Makefile.in gf2x/gf2x-config.h.in
if [ "$SAGE_DEBUG" = "yes" ]; then
echo "Building a debug version of gf2x."
export CFLAGS="-O0 -g $CFLAGS"
-elif $CC --version 2>/dev/null |grep 'gcc.* 5[.]1' >/dev/null; then
+elif $CC --version 2>/dev/null |grep 'gcc.* 5[.][12]' >/dev/null; then
echo "Using compiler flags to work around problems with GCC 5.1 (Trac #18580)"
export CFLAGS="-O2 -fno-forward-propagate -g $CFLAGS"
else
diff --git a/build/pkgs/ncurses/patches/work_around_changed_output_of_GNU_cpp_5.x.patch b/build/pkgs/ncurses/patches/work_around_changed_output_of_GNU_cpp_5.x.patch
index af82739..df8cb8a 100644
--- a/build/pkgs/ncurses/patches/work_around_changed_output_of_GNU_cpp_5.x.patch
+++ b/build/pkgs/ncurses/patches/work_around_changed_output_of_GNU_cpp_5.x.patch
@@ -19,7 +19,7 @@ output of 'cpp' w.r.t. line directives [1]. Anyway, the patch fixes the issue.)
+# Work around "unexpected" output of GCC 5.1.0's cpp w.r.t. #line directives
+# by simply suppressing them:
+case `$1 -dumpversion 2>/dev/null` in
-+ 5.[01].*) # assume a "broken" one
++ 5.[012].*) # assume a "broken" one
+ preprocessor="$1 -P -DNCURSES_INTERNALS -I../include"
+ ;;
+ *)
--
2.5.0
