[all-commits] [llvm/llvm-project] 076bc3: AMDGPU: Add some new baseline tests for exp lowering

[Matt Arsenault via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 076bc374fce5165ea12e2f77120a1796739a5143
      https://github.com/llvm/llvm-project/commit/076bc374fce5165ea12e2f77120a1796739a5143
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-20 (Thu, 20 Jul 2023)

  Changed paths:
    M llvm/test/CodeGen/AMDGPU/llvm.exp.ll
    M llvm/test/CodeGen/AMDGPU/llvm.exp2.ll

  Log Message:
  -----------
  AMDGPU: Add some new baseline tests for exp lowering


  Commit: 02955132380e5e2b6f44b1277c15178e343b63b3
      https://github.com/llvm/llvm-project/commit/02955132380e5e2b6f44b1277c15178e343b63b3
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-20 (Thu, 20 Jul 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUISelLowering.cpp
    M llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
    M llvm/test/CodeGen/AMDGPU/llvm.exp.ll

  Log Message:
  -----------
  AMDGPU: Filter out contract flags when lowering exp

It is unsafe to contract the fsub into the fmul. It also increases
code size by duplicating a constant.


  Commit: 5f1d3834a2bc3b77e126a278a0e7f00bce5576fc
      https://github.com/llvm/llvm-project/commit/5f1d3834a2bc3b77e126a278a0e7f00bce5576fc
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-20 (Thu, 20 Jul 2023)

  Changed paths:
    M clang/lib/Headers/__clang_hip_math.h
    M clang/test/Headers/__clang_hip_math.hip

  Log Message:
  -----------
  clang/HIP: Directly use f32 exp and log builtins

These are now lowered correctly by the backend, and you get proper
fast math flags when directly handled.


Compare: https://github.com/llvm/llvm-project/compare/9ef82be96eef...5f1d3834a2bc