Expand CUDA module APIs, fix locality graph out-of-bounds

habanero-rice / hclib

A C/C++ task-based programming model for shared memory and distributed parallel computing.

http://habanero-rice.github.io/hclib/

BSD 3-Clause "New" or "Revised" License

71 stars 35 forks source link

Closed agrippa closed 4 years ago

agrippa commented 4 years ago

Some small extensions to the CUDA APIs to support waiting on multiple futures.

agrippa commented 4 years ago

@srirajpaul could you take a look at this small PR before I merge?

srirajpaul commented 4 years ago

@agrippa Sure. I never used HClib with GPU. Will use this pull request as an excuse to start using HClib with CUDA.