[all-commits] [llvm/llvm-project] c87408: InstSimplify: Add baseline tests for frexp handling

[Matt Arsenault via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: c874082a0aca323865b42f2b01f83d22920eabe5
      https://github.com/llvm/llvm-project/commit/c874082a0aca323865b42f2b01f83d22920eabe5
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-17 (Mon, 17 Jul 2023)

  Changed paths:
    A llvm/test/Transforms/InstSimplify/frexp.ll

  Log Message:
  -----------
  InstSimplify: Add baseline tests for frexp handling


  Commit: 29d2a9bf9dbf030b43bb46d5611a751c91defc02
      https://github.com/llvm/llvm-project/commit/29d2a9bf9dbf030b43bb46d5611a751c91defc02
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-17 (Mon, 17 Jul 2023)

  Changed paths:
    M llvm/lib/Analysis/ConstantFolding.cpp
    M llvm/lib/Analysis/InstructionSimplify.cpp
    M llvm/test/Transforms/InstSimplify/frexp.ll

  Log Message:
  -----------
  InstSimplify: Handle basic folds for frexp

Handle constant folding and idempotent folding. Not sure
this is an appropriate use of undef for the inf/nan case. The
C version says the second result is "unspecified". The AMDGPU
instruction returns 0.


  Commit: 296e24cd2ee9ee65fb4036b673ff98dd1079d60f
      https://github.com/llvm/llvm-project/commit/296e24cd2ee9ee65fb4036b673ff98dd1079d60f
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-17 (Mon, 17 Jul 2023)

  Changed paths:
    M llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
    M llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp
    A llvm/test/CodeGen/AMDGPU/frexp-constant-fold.ll

  Log Message:
  -----------
  DAG: Constant fold frexp nodes

Special casing the nonfinite exponent value everywhere is kind of
annoying.


  Commit: 467df9c591dfdffef9af94b2d7e7a9df79afa551
      https://github.com/llvm/llvm-project/commit/467df9c591dfdffef9af94b2d7e7a9df79afa551
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-17 (Mon, 17 Jul 2023)

  Changed paths:
    M llvm/test/CodeGen/AMDGPU/fdiv32-to-rcp-folding.ll
    M llvm/test/CodeGen/AMDGPU/rcp-pattern.ll
    A llvm/test/CodeGen/AMDGPU/rsq.f32.ll
    A llvm/test/CodeGen/AMDGPU/rsq.f64.ll
    R llvm/test/CodeGen/AMDGPU/rsq.ll

  Log Message:
  -----------
  AMDGPU: Split and convert some rcp and rsq tests to generated checks


Compare: https://github.com/llvm/llvm-project/compare/82c65cc8f856...467df9c591df