[all-commits] [llvm/llvm-project] d020fa: [AA] Skip the layer of indirection in returning co...

[David Goldblatt via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: d020fa2b29d84b877d1e43dbfb3418275857bf87
      https://github.com/llvm/llvm-project/commit/d020fa2b29d84b877d1e43dbfb3418275857bf87
  Author: David Goldblatt <davidgoldblatt at fb.com>
  Date:   2023-07-28 (Fri, 28 Jul 2023)

  Changed paths:
    M llvm/lib/Analysis/BasicAliasAnalysis.cpp
    M llvm/lib/Analysis/GlobalsModRef.cpp
    M llvm/lib/Analysis/ObjCARCAliasAnalysis.cpp
    M llvm/lib/Analysis/ScalarEvolutionAliasAnalysis.cpp
    M llvm/lib/Analysis/ScopedNoAliasAA.cpp
    M llvm/lib/Analysis/TypeBasedAliasAnalysis.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUAliasAnalysis.cpp
    M llvm/lib/Target/NVPTX/NVPTXAliasAnalysis.cpp

  Log Message:
  -----------
  [AA] Skip the layer of indirection in returning conservative results.

Historically, AA implementations chained to a following implementation
to answer recursive queries. This is no longer the case, but the legacy
lives on in a confusing phrasing of the return-a-conservative-value
paths. Let's just return "don't know" directly, where appropriate; the
current two-step way is confusing.

Differential Revision: https://reviews.llvm.org/D149100