[all-commits] [llvm/llvm-project] bd2dca: AMDGPU: Use hex floats instead of ugly bitcasting

[Matt Arsenault via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: bd2dca0f74e4666bc806abdfae9de06fe37a697c
      https://github.com/llvm/llvm-project/commit/bd2dca0f74e4666bc806abdfae9de06fe37a697c
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-17 (Mon, 17 Jul 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
    M llvm/lib/Target/AMDGPU/SIISelLowering.cpp
    M llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-intrinsic-amdgcn-fdiv-fast.mir

  Log Message:
  -----------
  AMDGPU: Use hex floats instead of ugly bitcasting


  Commit: 715b12746f2e3b90e791d7aa9eb2b01555b3683a
      https://github.com/llvm/llvm-project/commit/715b12746f2e3b90e791d7aa9eb2b01555b3683a
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-17 (Mon, 17 Jul 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIISelLowering.cpp

  Log Message:
  -----------
  AMDGPU: Use available subtarget member


  Commit: 04185f0b0b52a906b67c96b654fd35f25f18328d
      https://github.com/llvm/llvm-project/commit/04185f0b0b52a906b67c96b654fd35f25f18328d
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2023-07-17 (Mon, 17 Jul 2023)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIISelLowering.cpp
    M llvm/test/CodeGen/AMDGPU/GlobalISel/combine-fcanonicalize.mir
    M llvm/test/CodeGen/AMDGPU/fcanonicalize.ll

  Log Message:
  -----------
  AMDGPU: Fix broken denormal constant folding of canonicalize

This needs to consider the dynamic denormal mode. It should be
possible to implement a runtime DAZ check with a canonicalize.


Compare: https://github.com/llvm/llvm-project/compare/8f90a5cc45db...04185f0b0b52