[PATCH] [X86][SSE] Vectorized i8 and i16 shift operators

[Simon Pilgrim]

I've added AVX2 optimizations for v16i16 shifts - I investigated using extend/shift/truncation but the unpack/shift/pack approach I went with required fewer instructions.


REPOSITORY
  rL LLVM

http://reviews.llvm.org/D9474

Files:
  lib/Target/X86/X86ISelLowering.cpp
  lib/Target/X86/X86TargetTransformInfo.cpp
  test/Analysis/CostModel/X86/testshiftashr.ll
  test/Analysis/CostModel/X86/testshiftlshr.ll
  test/Analysis/CostModel/X86/testshiftshl.ll
  test/CodeGen/X86/2011-12-15-vec_shift.ll
  test/CodeGen/X86/avx2-vector-shifts.ll
  test/CodeGen/X86/vec_shift8.ll

EMAIL PREFERENCES
  http://reviews.llvm.org/settings/panel/emailpreferences/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: D9474.26325.patch
Type: text/x-patch
Size: 51978 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150522/f1042fb0/attachment.bin>