[PATCH] D146212: Use NEON' tbl1 for 16xi8 build vector with mask.

[Lawrence Benson via Phabricator via llvm-commits]

lawben created this revision.
Herald added subscribers: hiraditya, kristof.beyls.
Herald added a project: All.
lawben requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

When using Clang's __builtin_shufflevector with a 16xi8 source and runtime mask on an AArch64 target, LLVM currently generates 16 extract+and+insert operations. This patch replaces 
these 16 inserts with a vector AND + NEON's tbl1 intruction.

Issue: https://github.com/llvm/llvm-project/issues/60515


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D146212

Files:
  llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
  llvm/test/CodeGen/AArch64/neon-shuffle-vector-tbl.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D146212.505756.patch
Type: text/x-patch
Size: 7123 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230316/4bdbb461/attachment.bin>