On Sat, 21 Apr 2012 21:54:13 +0800, zfyuan <[email protected]> wrote: > Hi, everyone. > > I want to use pycuda to realize some algorithm in my field. It's a > good choice for NVidia's GPU computing. However when I install it on > Fedora 16, I met a problem, and need help. > > I have installed CUDA Toolkit on Fedora 16, successfully compiled the > CUDA SDK. However when I compile pycuda with GL extension supported, > there exists en fatal error: > > " > In file included from src/wrapper/wrap_cudagl.cpp:5:0: > src/cpp/cuda_gl.hpp:10:19: fatal error：GL/gl.h： No such file or > directory > compilation terminated。 > error: command 'gcc' failed with exit status 1 > make: *** [all] error 1 > " > > But I can find the GL/gl.h file in the below directory, and > "/usr/local/cuda" has been in the PATH variable. > > "/usr/local/cuda/extras/CUPTI/include/GL/gl.h" > > Could anybody help me how to solve this problem? Thanks a lot: )
Can you please post your siteconf.py and the entire output of your PyCUDA compilation process? Andreas
pgpQ2a4u84uZo.pgp
Description: PGP signature
_______________________________________________ PyCUDA mailing list [email protected] http://lists.tiker.net/listinfo/pycuda
