[all-commits] [llvm/llvm-project] b51c29: [Inline] Add test for PR48209 (NFC)

[Nikita Popov via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: b51c290663cf50372cd681917cc63379abcf104a
      https://github.com/llvm/llvm-project/commit/b51c290663cf50372cd681917cc63379abcf104a
  Author: Nikita Popov <nikita.ppv at gmail.com>
  Date:   2020-11-18 (Wed, 18 Nov 2020)

  Changed paths:
    A llvm/test/Transforms/Inline/pr48209.ll

  Log Message:
  -----------
  [Inline] Add test for PR48209 (NFC)

The test shows noalias metadata being incorrectly applied to the
instruction producing the argument for the call.


  Commit: 23aeadb89df38406dc4d929d08286f7ce31040eb
      https://github.com/llvm/llvm-project/commit/23aeadb89df38406dc4d929d08286f7ce31040eb
  Author: Nikita Popov <nikita.ppv at gmail.com>
  Date:   2020-11-18 (Wed, 18 Nov 2020)

  Changed paths:
    M llvm/lib/Transforms/Utils/InlineFunction.cpp
    M llvm/test/Transforms/Inline/pr48209.ll

  Log Message:
  -----------
  [Inline] Fix incorrect noalias metadata application (PR48209)

The VMap also contains a mapping from Argument => Instruction,
where the instruction is part of the original function, not the
inlined one. The code was assuming that all the instructions in
the VMap were inlined.

This was a pre-existing problem for the loop access metadata, but
was extended to the more common noalias metadata by
27f647d117087ca11959e232e6443f4aee31e966, thus causing miscompiles.

There is a similar assumption inside CloneAliasScopeMetadata(), so
that one likely needs to be fixed as well.


Compare: https://github.com/llvm/llvm-project/compare/d1b921e3bf29...23aeadb89df3