[PATCH] D86429: [X86] Make lowerShuffleAsLanePermuteAndPermute use sublanes on AVX2

[Simon Pilgrim via Phabricator via llvm-commits]

RKSimon added inline comments.


================
Comment at: llvm/lib/Target/X86/X86ISelLowering.cpp:15629
+  SDValue CrossLane = DAG.getVectorShuffle(VT, DL, V1, V2, CrossLaneMask);
+  return DAG.getVectorShuffle(VT, DL, CrossLane, DAG.getUNDEF(VT), InLaneMask);
 }
----------------
Please can you try pulling this into the lambda as well? And refactor to try the 128-bit lanes first, then 64/32-bits on AVX2/fast-AVX2 cases? Sorry for asking you to experiment like this but I'm away  from a devmachine for the next week.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D86429/new/

https://reviews.llvm.org/D86429