[all-commits] [llvm/llvm-project] a0c2c6: [X86][AVX] Fold CONCAT(HOP(X, Y), HOP(Z, W)) -> HOP(C...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: a0c2c6aa420e572520337b782410c0c04ba87b69
      https://github.com/llvm/llvm-project/commit/a0c2c6aa420e572520337b782410c0c04ba87b69
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2020-08-12 (Wed, 12 Aug 2020)

  Changed paths:
    M llvm/lib/Target/X86/X86ISelLowering.cpp
    M llvm/test/CodeGen/X86/haddsub-2.ll
    M llvm/test/CodeGen/X86/haddsub-undef.ll

  Log Message:
  -----------
  [X86][AVX] Fold CONCAT(HOP(X,Y),HOP(Z,W)) -> HOP(CONCAT(X,Z),CONCAT(Y,W)) for float types

Only do this for AVX2+ targets as we still get some regressions on AVX1 without PERMPD/PERMQ