I have got problem when compile the cuda_conv(Google Code Archive - Long-term storage for Google Code Project Hosting.), my platform is ubuntu11.10 64bit, with cudatool 4.0.17. When compile the nvmatrix/nvmatrix.cu, it failed and displayed the error message “killed!! error 0x89”. Then I turn on the verbose option, the compile is die at ptaxs one file.
Can anybody help me? All the dependent package have already install, that’s no problem.
How can I solve it?
I did a test on my system (Ubuntu 12.10 x64, gcc (Ubuntu/Linaro 4.4.7-2ubuntu1) 4.4.7, CUDA 4.2, and the required python modules as in [url]http://code.google.com/p/cuda-convnet/wiki/Compiling[/url]. I was able to compile it and generate the _ConvNet.so library.
Did you edit the paths in the build.sh file to correspond to the correct ones on your system? Make sure the includes and libraries are on the respective directories.
The ones I had to change from his defaults were CUDA_SDK_PATH, which I installed elsewhere, and NUMPY_INCLUDE_PATH which I had to look in [url]Ubuntu – Error to see where arrayobject.h lived – /usr/lib/python2.7/dist-packages/numpy/core/include/numpy
You might want to do step 7 from the link below to install and use gcc 4.4, instead of the newer one. CUDA has issues sometimes with the newer gcc compilers:
[url]http://b-feng.blogspot.com/2012/10/install-cuda-5-on-ubuntu-1210.html[/url]
Thanks for your reply. I’m afraid my gcc have chosed the right version and the numpy path won’t be wrong, if nvidia company can give such error code message would be better. I’ll try other sdk toolkid to compile it.