Hello,
I have a program that compiles and runs just fine on linux but breaks with the following error on my Mac
ld: library not found for -lcuda
collect2: ld returned 1 exit status
make: *** [read_lib] Error 1
The last step in my Makefile is here:
gcc -o read_lib read_lib.cu.o read_and_store_library.o n_queries_mgf.o read_and_store_query.o bin_queries.o bin_library.o -L /usr/local/cuda/lib/ -lcuda -lcudart
I have added /usr/local/cuda/lib to my PATH and I have verified that the libraries are located at that path. Any tips on how to fix this problem would be greatly appreciated.
In another post someone said that you should make sure the permission on the /usr/local/cuda/lib folder are correct, I see here that libcuda is different but I don’t understand what I need to change…
-rwxrwxr-x 1 lash admin 15913168 Sep 3 01:00 libcublas.dylib
-rwxrwxr-x 1 lash admin 5444172 Sep 3 01:00 libcublasemu.dylib
-rwxr-xr-x 1 root wheel 3893996 Sep 3 00:16 libcuda.dylib
-rwxrwxr-x 1 lash admin 276652 Sep 3 01:00 libcudart.dylib
-rwxrwxr-x 1 lash admin 8481764 Sep 3 01:00 libcufft.dylib
-rwxrwxr-x 1 lash admin 1501720 Sep 3 01:00 libcufftemu.dylib
-rwxrwxr-x 1 lash admin 13100 Sep 3 01:00 libtlshook.dylib