[PATCH] D107306: [AArch64] Prefer ZIP1 over INS to lower concat_vectors.

[Eli Friedman via Phabricator via llvm-commits]

efriedma created this revision.
efriedma added reviewers: RKSimon, SjoerdMeijer, dmgreen, paulwalker-arm.
Herald added subscribers: danielkiss, hiraditya, kristof.beyls.
efriedma requested review of this revision.
Herald added a project: LLVM.

ZIP1 should have comparable performance, and gives the register allocator more flexibility.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D107306

Files:
  llvm/lib/Target/AArch64/AArch64InstrInfo.td
  llvm/test/CodeGen/AArch64/GlobalISel/select-concat-vectors.mir
  llvm/test/CodeGen/AArch64/arm64-neon-copy.ll
  llvm/test/CodeGen/AArch64/arm64-vcombine.ll
  llvm/test/CodeGen/AArch64/bf16-vector-shuffle.ll
  llvm/test/CodeGen/AArch64/concat-vector.ll
  llvm/test/CodeGen/AArch64/dag-combine-trunc-build-vec.ll
  llvm/test/CodeGen/AArch64/fp16-v16-instructions.ll
  llvm/test/CodeGen/AArch64/fp16-v8-instructions.ll
  llvm/test/CodeGen/AArch64/fp16-vector-shuffle.ll
  llvm/test/CodeGen/AArch64/neon-bitwise-instructions.ll
  llvm/test/CodeGen/AArch64/shuffle-mask-legal.ll
  llvm/test/CodeGen/AArch64/speculation-hardening-loads.ll
  llvm/test/CodeGen/AArch64/sve-fixed-length-concat.ll
  llvm/test/CodeGen/AArch64/sve-fixed-length-fp-extend-trunc.ll
  llvm/test/CodeGen/AArch64/sve-fixed-length-fp-to-int.ll
  llvm/test/CodeGen/AArch64/sve-fixed-length-int-to-fp.ll
  llvm/test/CodeGen/AArch64/sve-fixed-length-masked-gather.ll
  llvm/test/CodeGen/AArch64/sve-fixed-length-trunc-stores.ll
  llvm/test/CodeGen/AArch64/vector-fcopysign.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D107306.363559.patch
Type: text/x-patch
Size: 29918 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20210802/1c1fef48/attachment-0001.bin>