Date: Fri, 12 Aug 2016 04:54:51 +0000 From: bugzilla-noreply@freebsd.org To: freebsd-ports-bugs@FreeBSD.org Subject: [Bug 211760] graphics/Coin builds fine with clang 3.8 Message-ID: <bug-211760-13-pXJNcB5sb8@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-211760-13@https.bugs.freebsd.org/bugzilla/> References: <bug-211760-13@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D211760 Shane <FreeBSD@ShaneWare.Biz> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |FreeBSD@ShaneWare.Biz --- Comment #6 from Shane <FreeBSD@ShaneWare.Biz> --- The USE_GCC option is historic, it was used to make the port build with gcc= 48 instead of the base gcc4.2 which is still the base compiler on 9.3 that is supported until December 2016. The correct replacement for USE_GCC would now be USE=3Dcompiler:c++11-lang = - see /usr/ports/Mk/Uses/compiler.mk for available options. As for the build failure it starts with "error: use of undeclared identifier 'NO_SINGLEPREC'" these can be found in work/Coin-3.1.3/include/Inventor/C/base/math-undefs.h - reading the comments there shows this as their approach to preventing the use of math functions = that take a float parameter, these functions are not used within the Coin source code and the deliberate build breakage can be removed from the build by commenting out the #define lines in that file. I have not dug into this so = am unsure if some math functions are called indirectly and lead to float/double conversions that they don't want. Removing the #defines in math-undefs.h should allow any clang or gcc>4.2 to= be used. It was about 2 years ago I looked at this and think this got it to compile on 8.x and 9.x --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-211760-13-pXJNcB5sb8>