[all-commits] [llvm/llvm-project] a26288: [X86][Atom] Fix vector fadd/fcmp/fmul resource/thr...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: a26288e8030a44a5696fe170de3fea79800d7bcd
      https://github.com/llvm/llvm-project/commit/a26288e8030a44a5696fe170de3fea79800d7bcd
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2021-05-20 (Thu, 20 May 2021)

  Changed paths:
    M llvm/lib/Target/X86/X86ScheduleAtom.td
    M llvm/test/tools/llvm-mca/X86/Atom/resources-sse1.s
    M llvm/test/tools/llvm-mca/X86/Atom/resources-sse2.s
    M llvm/test/tools/llvm-mca/X86/Atom/resources-sse3.s
    M llvm/test/tools/llvm-mca/X86/Atom/resources-x87.s

  Log Message:
  -----------
  [X86][Atom] Fix vector fadd/fcmp/fmul resource/throughputs

Match whats documented in the Intel AOM - these are all fadd/fcmp use Port1 and fmul uses Port1, but in many cases BOTH ports are required - this was being incorrectly modelled as EITHER port.

Discovered while investigating the correct fptoui costs to fix the regressions in D101555.

Now that we can use in-order models in llvm-mca, the atom model is a good "worst case scenario" analysis for x86.