Skip to content

Instantly share code, notes, and snippets.

@xlanor

xlanor/archlinux376.md

Last active July 2, 2019 13:21
Show Gist options
  • Save xlanor/fa6af632fd7345b2b1cbdfab633233c5 to your computer and use it in GitHub Desktop.
Save xlanor/fa6af632fd7345b2b1cbdfab633233c5 to your computer and use it in GitHub Desktop.
Download ZIP
CSCI376 - building on archlinux.
Raw
archlinux376.md

Setup for CSCI376 multicore & GPU processing for an Arch Linux system

To install amdapp-sdk:
Using yay,

yay -S --editmenu amdapp-sdk


Edit PKGBuild, remove:

  • !upx (depreceated pacman 5.10)\

Replace:

This will get you the amd app SDK that is being used for windows in this course.
Unfortunately I still can't get this to work. It is much easier to use the nvidia drivers.

I randomly picked a few tutorials with the nvidia drivers and they all seem to work.

Instructions to install nvidia drivers on arch:
(From https://wiki.archlinux.org/index.php/GPGPU)\
There's alot of information there, but primarily what you need is this:

yay -S opencl-nvidia
yay -S ocl-icd
yay -S opencl-headers

These 3 should allow you to compile any files

Generally, by installing the above libraries, /usr/bin/ld should have been setup nicely, so you can now do this

g++ *.cpp -lOpenCl

to compile the files.

For some reason you still need to execute the output as root to access the openCL api.

arch

Things to note:

  • ALL nvidia drivers must be on the same versions.
  • Intel works too, just that you need to install the following

intel-opencl
intel-opencl-runtime

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment