Issue #51 new

Fails to compile

Anonymous created an issue

See link for output. http://pastebin.com/MVZis4tF.

With glm 9.4 installed from Arch repo.

Also getting:

==== Building glload (debug) ==== Creating Debug gl_load_cpp.cpp gl_load.c glx_load_cpp.cpp glx_load.c In file included from include/glload/glx_all.h:5:0, from source/glx_load.c:4: include/glload/_int_glx_exts.h:166:0: warning: "GLX_BACK_EXT" redefined [enabled by default] #define GLX_BACK_EXT 0x20E0 ^ In file included from include/glload/_int_glx_type.h:11:0, from include/glload/glx_all.h:4, from source/glx_load.c:4: /usr/include/GL/glx.h:453:0: note: this is the location of the previous definition #define GLX_BACK_EXT GLX_BACK_LEFT_EXT ^ In file included from include/glload/glx_all.h:5:0, from source/glx_load.c:4: include/glload/_int_glx_exts.h:173:0: warning: "GLX_FRONT_EXT" redefined [enabled by default] #define GLX_FRONT_EXT 0x20DE ^ In file included from include/glload/_int_glx_type.h:11:0, from include/glload/glx_all.h:4, from source/glx_load.c:4: /usr/include/GL/glx.h:452:0: note: this is the location of the previous definition

Feel free to email me @ bruce dot tanner at gmail dot com

Comments (4)

  1. tannerb32

    Another issue: glsdk/glmesh/include/glmesh/VertexFormat.h:168,32 - Error - default initialization of an object of const type 'const glmesh::SeparateAttribFormatTag' requires a user-provided default constructor

    this is from Clang, compiles fine in GCC

  2. Lucas Jandrew

    I recently managed to stumble my way through getting this to compile after encountering the same set of errors. Since these fixes all involve editing the Makefiles generated by premake and I don't know anything about tooling in the C++ space I can only note my findings here instead of making a patch :(

    The biggest issue was that I had a few of these libraries already on my machine and the compiler was prioritizing them as system level libraries. Since I really didn't want to remove them as other things were utilizing them I went into the Makefile for each project that was failing and changed the INCLUDES lines to use the -isystem flag rather than -I. I believe in newer versions of GCC/G++ the -I flag only works on libraries in "quotes" and not in <brackets>, and -isystem bumps the path to the front of the system includes.

    The other issue I had was due to some failed type casting in the framework project. I had to add -fpermissive to the CPPFLAGS environment variable for the framework Makefile.

  3. Log in to comment