Closed mzk32 closed 5 years ago
In principle you don't need to compile with -fPIC. You just need to have cuda on your $PATH variable. Make sure you do ./autogen.sh ; ./configure; make
Also, you should probably use 1.2.0; 1.2.1 has a problem with the solo version it was supposed to fix. (I am working on it)
I got an error when compiling version 1.2.1:
i have libcudart in this location:
i couldn't figure out how to
recompile with -fPIC
, please helpthank you