[all-commits] [llvm/llvm-project] e14889: [X86] Preserve volatile ATOMIC_LOAD_OR nodes

[Nabeel Omer via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: e148899ad9f384e77e1440f6cf8b03f5fbbc11d2
      https://github.com/llvm/llvm-project/commit/e148899ad9f384e77e1440f6cf8b03f5fbbc11d2
  Author: Nabeel Omer <Nabeel.Omer at sony.com>
  Date:   2023-07-11 (Tue, 11 Jul 2023)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/test/CodeGen/X86/pr63692.ll

  Log Message:
  -----------
  [X86] Preserve volatile ATOMIC_LOAD_OR nodes

Fixes #63692.

In reference to volatile memory accesses, the langref says:
> the backend should never split or merge target-legal volatile load/store instructions.

Differential Revision: https://reviews.llvm.org/D154609