US LLVM Dev Meeting 2019 - Round Table - Challenges using LLVM for GPU compilation

Dear all,

I would like announce a round table planned for the upcoming LLVM Dev meeting next week that will cover various topics related to the use of LLVM in the compiler stacks for the GPUs.

Here is the initial list of discussion topics:

  • Canonicalization vs. GPUs: Type mutation;

  • Control flow mutation (graphics shaders are more sensitive to this);

  • Divergence/reconvergence sensitivity;

  • Address-space awareness;

The round table is preliminary planned for Tue 3-4pm or Wed 2-3pm. The exact slot will be listed in the online schedule soon: https://llvmdevmtg2019.sched.com/

Feel free to let us know of any other topics you would like to cover.

Kind regards,
Anastasia

Dear all,

Could I suggest adding one more topic to the list related to GPUs? We have just published a prototype for a LLVM backend generating SPIR-V at https://github.com/KhronosGroup/LLVM-SPIRV-Backend. While I won’t be present at the round table, inputs from the LLVM community on how to upstream this backend to LLVM would be greatly appreciated.

Thank you.

Kind regards,
Marco

Thanks, Marco! If there is enough interest in this topic we can also organize a separate round table for this discussion.

Cheers,
Anastasia

+1
I’m very interested.

Hi all,

The round table slot is now confirmed for Tue 3.10pm: https://llvmdevmtg2019.sched.com/event/WFbV

Cheers,
Anastasia