[Bf-committers] Buildbot cuda-hack

[Francesc Juhe]

But the cuda-hack commit was for CUDA 7.5 initially, I just ask about it now because I am creating a new build environment with el capitan and wanted to know more of what this “hack” did to better understand the build process and be prepared if I had the same issue.

About the openmp, I have been using clang 3.5 from blender libs  to get openmp support without finding any issues but since the change from scons to cmake on osx buildbot is no longer using it (scons buildbot config pointed to that clang from blender libs but cmake doesn't) and so it is missing openmp support but as the release build has openmp I wanted to know if release builders were using this clang from blender libs or a different one. In my build I added an osx config file to point to clang 3.5 and I think it could be better if official buildbot had openmp back like before when it used scons so that buildbot builds are more similar to releases, otherwise there could be issues not present on buildbot but affecting release. Don’t know, just my opinion.

Also, there is another thing missing since scons was removed in favor of cmake, osx buildbot does not include the blenderplayer binary inside blenderplayer.app, I fix it in my build with this diff:

diff --git a/source/blenderplayer/CMakeLists.txt b/source/blenderplayer/CMakeLists.txt
index ca84195..0da6e70 100644 (file)
--- a/source/blenderplayer/CMakeLists.txt
+++ b/source/blenderplayer/CMakeLists.txt
@@ -76,6 +76,11 @@ elseif(APPLE)
                MACOSX_BUNDLE_SHORT_VERSION_STRING ${BLENDER_VERSION}
                MACOSX_BUNDLE_LONG_VERSION_STRING "${BLENDER_VERSION} ${BLENDER_DATE}")
 
+       install(
+               TARGETS blenderplayer
+               DESTINATION "."
+       )
+
 else()
        add_executable(blenderplayer bad_level_call_stubs/stubs.c)

Thanks,
Francesc

On 26 Aug 2016, at 19:08, Jens Verwiebe <info at jensverwiebe.de> wrote:

> I'd say cuda8 is pointless for you right now, cause pascal gpu is not 
> yet supported by the NV/OSX (Quadro-) drivers.
> 
> You would definitely loose openmp capability unless you compile and use 
> a vanilla clang 3.7 which supports it.
> 
> 
> Jens
> 
> 
> Am 26.08.2016 um 18:11 schrieb Francesc Juhe:
>> Hi,
>> 
>> Some time ago noticed this commit https://developer.blender.org/rB3a24e6320f31a790ea9d3c912a35db2186140b71 changing the nvcc compiler on builbot but I did not give it much though as I was having no issues compiling with cuda 7.5 and clang 3.5 from libs folder on OSX 10.9. But now with CUDA 8 it seems I have to update to OSX 10.11 and Xcode 7.3.
>> 
>> So, how does this cuda-hack work?
>> Also, OSX buildbot builds seem to be compiled without OpenMP but release builds are compiled with it. What are the appropriate steps for building a release build?
>> 
>> Thank you,
>> Francesc
>> 
>> _______________________________________________
>> Bf-committers mailing list
>> Bf-committers at blender.org
>> https://lists.blender.org/mailman/listinfo/bf-committers
>> 
> 
> -- 
> 
> Jens Verwiebe
> Allerskehre 44 - 22309 Hamburg
> 
> Tel.: +49 40 68 78 50
> mobile: +49 172 400 49 07
> mailto: info at jensverwiebe.de
> web: http://www.jensverwiebe.de
> 
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> https://lists.blender.org/mailman/listinfo/bf-committers