Package: aspectc++
Version: 1:2.2+git20170823-7
Followup-For: Bug #907632
User: ubuntu-de...@lists.ubuntu.com
Usertags: origin-ubuntu disco ubuntu-patch

Hello,

Based on the bug history, I don't think this is a bug in gcc-8 but in
aspectc++ (or in llvm), so I have reassigned.

I have also refined the patch provided by Frédéric, to avoid repetition in
the code; please find it attached.

I have uploaded this patch to Ubuntu, where I confirm it fixes the build
failure on ppc64el.

Cheers,
-- 
Steve Langasek                   Give me a lever long enough and a Free OS
Debian Developer                   to set it on, and I can move the world.
Ubuntu Developer                                   https://www.debian.org/
slanga...@ubuntu.com                                     vor...@debian.org
diff -Nru aspectc++-2.2+git20170823/debian/control 
aspectc++-2.2+git20170823/debian/control
--- aspectc++-2.2+git20170823/debian/control    2018-08-24 23:29:52.000000000 
-0500
+++ aspectc++-2.2+git20170823/debian/control    2018-12-25 10:15:09.000000000 
-0600
@@ -1,8 +1,7 @@
 Source: aspectc++
 Section: devel
 Priority: optional
-Maintainer: Ubuntu Developers <ubuntu-devel-disc...@lists.ubuntu.com>
-XSBC-Original-Maintainer: Reinhard Tartler <siret...@tauware.de>
+Maintainer: Reinhard Tartler <siret...@tauware.de>
 Build-Depends: debhelper (>= 9), libxml2-dev, docbook-to-man, zlib1g-dev, 
libedit-dev, llvm-4.0-dev, libclang-4.0-dev
 Build-Depends-Indep: doxygen, graphviz, gsfonts
 Standards-Version: 4.0.0
diff -Nru aspectc++-2.2+git20170823/debian/patches/ppc64el-no-float128.patch 
aspectc++-2.2+git20170823/debian/patches/ppc64el-no-float128.patch
--- aspectc++-2.2+git20170823/debian/patches/ppc64el-no-float128.patch  
1969-12-31 18:00:00.000000000 -0600
+++ aspectc++-2.2+git20170823/debian/patches/ppc64el-no-float128.patch  
2018-12-25 10:15:01.000000000 -0600
@@ -0,0 +1,23 @@
+Description: adjust ppc64el compiler options for compatibility with gcc-8
+ gcc-8 now uses -mfloat128 by default on ppc64el, which results in
+ incompatible type definitions being passed in to llvm.  Override this by
+ explicitly passing -mno-float128 on ppc64el.
+Author: Frédéric Bonnard <fre...@debian.org>
+Bug-Debian: https://bugs.debian.org/907632
+Last-Modified: 2018-12-25
+
+Index: aspectc++-2.2+git20170823/Ag++/PumaConfigFile.cc
+===================================================================
+--- aspectc++-2.2+git20170823.orig/Ag++/PumaConfigFile.cc
++++ aspectc++-2.2+git20170823/Ag++/PumaConfigFile.cc
+@@ -118,7 +118,10 @@
+     config_command_str = "\"" + _config.cc_bin() + "\" "
+         + _config.optvec().getString(
+             (OptionItem::OPT_GCC | OptionItem::OPT_CONFIG))
++#ifdef __powerpc__ && __powerpc64__ && __LITTLE_ENDIAN__
++        + " -mno-float128 "
++#endif
+         + " -E -dM -v -x c++ \"" + empty_file_name + "\"";
+   }
+ 
+   // get c compiler output
diff -Nru aspectc++-2.2+git20170823/debian/patches/series 
aspectc++-2.2+git20170823/debian/patches/series
--- aspectc++-2.2+git20170823/debian/patches/series     2018-06-25 
09:59:10.000000000 -0500
+++ aspectc++-2.2+git20170823/debian/patches/series     2018-12-25 
10:12:04.000000000 -0600
@@ -1,2 +1,3 @@
 0001-lexertl-Fix-FTBFS-with-gcc-8.patch
 auto-gitignore
+ppc64el-no-float128.patch

Reply via email to