[all-commits] [llvm/llvm-project] 3e0e55: [CUDA] Fixed sm version constrain for __bmma_m8n8k...

JackAKirk via All-commits all-commits at lists.llvm.org
Fri Aug 5 12:14:33 PDT 2022


  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 3e0e5568a6a8c744d26f79a1e55360fe2655867c
      https://github.com/llvm/llvm-project/commit/3e0e5568a6a8c744d26f79a1e55360fe2655867c
  Author: Jack Kirk <jack.kirk at codeplay.com>
  Date:   2022-08-05 (Fri, 05 Aug 2022)

  Changed paths:
    M clang/include/clang/Basic/BuiltinsNVPTX.def
    M clang/test/CodeGen/builtins-nvptx-mma.cu
    M clang/test/CodeGen/builtins-nvptx-mma.py

  Log Message:
  -----------
  [CUDA] Fixed sm version constrain for __bmma_m8n8k128_mma_and_popc_b1.

As stated in
https://docs.nvidia.com/cuda/parallel-thread-execution/index.html#warp-level-matrix-instructions-wmma-mma:
".and operation in single-bit wmma requires sm_80 or higher."

tra@: Fixed a bug in builtins-nvptx-mma.py test generator and regenerated the tests.

Differential Revision: https://reviews.llvm.org/D131265




More information about the All-commits mailing list