search

system76 / cuda

Packaging for NVIDIA's CUDA Toolkit
21 stars 10 forks source link

Why /usr/lib instead of /usr/local ? #13

Closed lhk closed 4 years ago

lhk commented 4 years ago

This package installs cuda to/usr/lib. However, the default seems to be/usr/local. For example, the mnistCUDNN sample code from the code samples available here only compiles after editing the Makefile to your path.

I've tried torch and it works fine. But there are packages that rely on the /usr/local as the installation directory. For example jax, which only worked after I symlinked /usr/lib/cuda to /usr/local/cuda

Is there a reason for this installation directory?

mmstick commented 4 years ago

/usr/local is for locally-installed packages & applications, not system-installed packages from the distribution. Debian packaging does not install files to /usr/local. Ideally these samples would be corrected to automatically detect the CUDA path.

lhk commented 4 years ago

oh, I didn't know that. Thanks for the clarification.