[PATCH] D40865: X86 AVX2: Prefer one VPERMV over ShuffleAsRepeatedMaskAndLanePermute

[Zvi Rackover via Phabricator via llvm-commits]

zvi created this revision.

Reverse the order of precedence in lowering v8i32 shuffles by
prefering one permute by vector of indices over two shuffleis with
immediate masks. The benefit is reduction of shuffle port pressure.


https://reviews.llvm.org/D40865

Files:
  lib/Target/X86/X86ISelLowering.cpp
  test/CodeGen/X86/avx2-conversions.ll
  test/CodeGen/X86/avx2-vector-shifts.ll
  test/CodeGen/X86/avx512-shuffles/broadcast-vector-int.ll
  test/CodeGen/X86/combine-shl.ll
  test/CodeGen/X86/combine-sra.ll
  test/CodeGen/X86/combine-srl.ll
  test/CodeGen/X86/oddshuffles.ll
  test/CodeGen/X86/psubus.ll
  test/CodeGen/X86/reduce-trunc-shl.ll
  test/CodeGen/X86/shuffle-vs-trunc-256.ll
  test/CodeGen/X86/vector-shuffle-256-v8.ll
  test/CodeGen/X86/vector-shuffle-combining.ll
  test/CodeGen/X86/vector-trunc-math.ll
  test/CodeGen/X86/vector-trunc.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D40865.125631.patch
Type: text/x-patch
Size: 70093 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20171205/07465ecc/attachment-0001.bin>