[all-commits] [llvm/llvm-project] f26ed6: llc: Change behavior of -mcpu with existing attribute

[Matt Arsenault via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: f26ed6e47cb8b080c236d11c4942a12265180084
      https://github.com/llvm/llvm-project/commit/f26ed6e47cb8b080c236d11c4942a12265180084
  Author: Matt Arsenault <Matthew.Arsenault at amd.com>
  Date:   2020-01-07 (Tue, 07 Jan 2020)

  Changed paths:
    M llvm/include/llvm/CodeGen/CommandFlags.inc
    M llvm/test/CodeGen/AMDGPU/memory-legalizer-atomic-insert-end.mir
    M llvm/test/CodeGen/MIR/AMDGPU/llc-target-cpu-attr-from-cmdline-ir.mir
    M llvm/test/CodeGen/X86/avoid-sfb-overlaps.ll
    M llvm/test/CodeGen/X86/avoid-sfb.ll
    M llvm/test/CodeGen/X86/llc-override-mcpu-mattr.ll
    M llvm/test/DebugInfo/COFF/inlining-files.ll
    M llvm/test/DebugInfo/COFF/inlining-header.ll
    M llvm/test/Other/opt-override-mcpu-mattr.ll
    M llvm/test/Transforms/LoopVectorize/X86/tail_folding_and_assume_safety.ll
    M llvm/test/tools/llvm-objdump/AMDGPU/source-lines.ll

  Log Message:
  -----------
  llc: Change behavior of -mcpu with existing attribute

Don't overwrite existing target-cpu attributes.

I've often found the replacement behavior annoying, and this is
inconsistent with how the fast math command line flags interact with
the function attributes.

Does not yet change target-features, since I think that should behave
as a concatenation.