hi @theonlychant can you please give some background on what you are contributing? thank you!

To be honest I just added this in to try and understand how you guys(AMD) work with python and c++ before I apply for a job there(I have also made other PRs to AMD prior) I was testing to see if the samples would work and right now all it does is exposes a small CPU kernels from the repo to Python via Cython, building a single extension module (compute). Sorry if this wasted your time

@theonlychant no waste of time :) I'll look at this asap, thanks

I added nput validation + contiguity/dtype coercion to the PR changes to the wrappers to prevent crashes and ensure correct dtypes. You can review and tell me if it's garbage or not I don't mind I'll like working on my code for you guys(AMD) more right now I feel like I'm starting to understand how AMD writes code now and it's pretty easy to read for the most part thanks for reviewing when you do(if you're free)

Hi @theonlychant thanks again for submitting this PR. We appreciate you for wanting to contribute and we look forward to more contributions from you in the future. Regarding this specific PR, I would like to point you to the existing Python/hip-python/cython_hip_example/ and Python/hip-python/cython_basic/, and to the OpenMP-offload variants under Pragma_Examples/OpenMP/C/1_saxpy/..... Note that a CPU-only Cython benchmark doesn't quite belong in this repo given that its main focus is the AMD GPU software stack, so I am closing without merging.