[all-commits] [llvm/llvm-project] 8db31e: [NVPTX] Do not addrspacecast AS-specific kernel ar...

[Artem Belevich via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 8db31e932d33b63449d3727c0496fce29883ebb1
      https://github.com/llvm/llvm-project/commit/8db31e932d33b63449d3727c0496fce29883ebb1
  Author: Artem Belevich <tra at google.com>
  Date:   2023-01-26 (Thu, 26 Jan 2023)

  Changed paths:
    M llvm/lib/Target/NVPTX/NVPTX.h
    M llvm/lib/Target/NVPTX/NVPTXLowerArgs.cpp
    M llvm/lib/Target/NVPTX/NVPTXTargetMachine.cpp
    M llvm/test/CodeGen/NVPTX/lower-args.ll

  Log Message:
  -----------
  [NVPTX] Do not addrspacecast AS-specific kernel arguments.

Fixes https://github.com/llvm/llvm-project/issues/46954

The assumption that generic pointers passed to a CUDA kernel is CUDA-specific
and should not be applied to non-CUDA compilations.  Addrspacecasts to global AS
and back should never be applied to AS-specific pointers.

In order to make tests actually do the testing for non-CUDA compilation, we need
to get TargetMachine from the TargetPassConfig, instead of passing it explicitly
as a pass constructor argument.

Differential Revision: https://reviews.llvm.org/D142581