[all-commits] [llvm/llvm-project] db6bc2: [AMDGPU][GlobalISel] Fold G_FNEG above when users ...

Wed Nov 17 05:27:45 PST 2021

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: db6bc2ab51371f69569c25514c20c3786398a032
      https://github.com/llvm/llvm-project/commit/db6bc2ab51371f69569c25514c20c3786398a032
  Author: Mirko Brkusanin <Mirko.Brkusanin at amd.com>
  Date:   2021-11-17 (Wed, 17 Nov 2021)

  Changed paths:
    M llvm/include/llvm/CodeGen/GlobalISel/CombinerHelper.h
    M llvm/lib/CodeGen/GlobalISel/CombinerHelper.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUCombine.td
    A llvm/lib/Target/AMDGPU/AMDGPUCombinerHelper.cpp
    A llvm/lib/Target/AMDGPU/AMDGPUCombinerHelper.h
    M llvm/lib/Target/AMDGPU/AMDGPUPostLegalizerCombiner.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUPreLegalizerCombiner.cpp
    M llvm/lib/Target/AMDGPU/CMakeLists.txt
    A llvm/test/CodeGen/AMDGPU/GlobalISel/combine-foldable-fneg.mir
    M llvm/utils/gn/secondary/llvm/lib/Target/AMDGPU/BUILD.gn

  Log Message:
  -----------
  [AMDGPU][GlobalISel] Fold G_FNEG above when users cannot fold mods

If possible fold fneg into instruction above if users cannot fold mods and we
know it will decrease instruction count.
Follows same logic as SDAG combiner in choosing opportunities to combine.

Differential Revision: https://reviews.llvm.org/D112827

  Commit: f2d402e58b90eb2ee747ed96a2f7c5e7667023bd
      https://github.com/llvm/llvm-project/commit/f2d402e58b90eb2ee747ed96a2f7c5e7667023bd
  Author: Mirko Brkusanin <Mirko.Brkusanin at amd.com>
  Date:   2021-11-17 (Wed, 17 Nov 2021)

  Changed paths:
    M llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-fceil.mir
    M llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-frint.mir
    M llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-intrinsic-trunc.mir

  Log Message:
  -----------
  [NFC][AMDGPU][GlobalISel] Fix some legalizer tests

Instructions being tested were accidentally left dead.

Compare: https://github.com/llvm/llvm-project/compare/3874277f415d...f2d402e58b90