[all-commits] [llvm/llvm-project] 0f8e4a: [InstCombine] Add test for PR47730

[Nikita Popov via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 0f8e4a5ed0e556af82592f7cae4a6082ab238da3
      https://github.com/llvm/llvm-project/commit/0f8e4a5ed0e556af82592f7cae4a6082ab238da3
  Author: Nikita Popov <nikita.ppv at gmail.com>
  Date:   2020-10-05 (Mon, 05 Oct 2020)

  Changed paths:
    M llvm/test/Transforms/InstCombine/select.ll

  Log Message:
  -----------
  [InstCombine] Add test for PR47730


  Commit: 3641d375f6747237b1b55a25a55e9028d8a67a02
      https://github.com/llvm/llvm-project/commit/3641d375f6747237b1b55a25a55e9028d8a67a02
  Author: Nikita Popov <nikita.ppv at gmail.com>
  Date:   2020-10-05 (Mon, 05 Oct 2020)

  Changed paths:
    M llvm/lib/Transforms/InstCombine/InstCombineSelect.cpp
    M llvm/test/Transforms/InstCombine/select.ll

  Log Message:
  -----------
  [InstCombine] Handle GEP inbounds in select op replacement (PR47730)

When retrying the "simplify with operand replaced" select
optimization without poison flags, also handle inbounds on GEPs.

Of course, this particular example would also be safe to transform
while keeping inbounds, but the underlying machinery does not
know this (yet).


Compare: https://github.com/llvm/llvm-project/compare/95262ee2be75...3641d375f674