[API] Thread binding API for CUDA/OpenCL targets

cornell-zhang / hcl-dialect

HeteroCL-MLIR dialect for accelerator design

https://cornell-zhang.github.io/heterocl/index.html

Other

40 stars 17 forks source link

[API] Thread binding API for CUDA/OpenCL targets #62

Closed hecmay closed 2 years ago

hecmay commented 2 years ago

@chhzh123 @zzzDavid. I added a ThreadBindOp to hcl-dialect. this customization only adds the NDRange dimension information to the iteration. Can you please kindly review the PR?

My C/C++ formatter automatically formatted some code so there is a lot of changes in this PR. If you think this is a problem, i can revert the changes.

hecmay commented 2 years ago

fixed.

chhzh123 commented 2 years ago

You have already updated the HeteroCL repository, right?

hecmay commented 2 years ago

You have already updated the HeteroCL repository, right?

Right. I have a PR ready for review in HCL repo. We still need to convert to gpu dialect. I will leave that part to Alex

chhzh123 commented 2 years ago

Okay, thanks. I'll merge it later.