FYI: new LLVM IR for parallelism


#1

Hi,
I found this article yesterday. I’m just posting the link here because I thought it might be interesting for julia threading in a future version as a compiler optimization.