[all-commits] [llvm/llvm-project] 6565b5: [X86][llvm-mc] Make the suffix matcher more accurate.

[Pengfei Wang via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 6565b5858444ba7dcf799467f5be63d2c2370715
      https://github.com/llvm/llvm-project/commit/6565b5858444ba7dcf799467f5be63d2c2370715
  Author: Wang, Pengfei <pengfei.wang at intel.com>
  Date:   2020-05-27 (Wed, 27 May 2020)

  Changed paths:
    M llvm/lib/Target/X86/AsmParser/X86AsmParser.cpp
    M llvm/lib/Target/X86/AsmParser/X86Operand.h
    M llvm/test/MC/X86/avx512-err.s
    M llvm/test/tools/llvm-mca/X86/BdVer2/dependent-pmuld-paddd.s
    M llvm/test/tools/llvm-mca/X86/BtVer2/dependent-pmuld-paddd.s

  Log Message:
  -----------
  [X86][llvm-mc] Make the suffix matcher more accurate.

Summary:
Some instruction like VPMULDQ is NOT the variant of VPMULD but a new
one.
So we should make sure the suffix matcher only works for memory variant
that has the same size with the suffix.
Currently we only check for SSE/AVX* instructions, because many legacy
instructions didn't declare the alias instructions of their variants.

Differential Revision: https://reviews.llvm.org/D80608