[PATCH] D92761: [clang][AArch64][SVE] Avoid going through memory for VLAT <-> VLST casts

[Joe Ellis via Phabricator via llvm-commits]

joechrisellis created this revision.
joechrisellis added reviewers: peterwaller-arm, DavidTruby.
Herald added subscribers: psnobl, kristof.beyls, tschuett.
Herald added a reviewer: efriedma.
joechrisellis requested review of this revision.
Herald added projects: clang, LLVM.
Herald added subscribers: llvm-commits, cfe-commits.

This change makes use of the llvm.vector.extract intrinsic to avoid
going through memory when performing bitcasts between vector-length
agnostic types and vector-length specific types.

Depends on D91362 <https://reviews.llvm.org/D91362>


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D92761

Files:
  clang/lib/CodeGen/CGExprScalar.cpp
  clang/test/CodeGen/aarch64-sve-acle-__ARM_FEATURE_SVE_VECTOR_OPERATORS.c
  clang/test/CodeGen/attr-arm-sve-vector-bits-bitcast.c
  clang/test/CodeGen/attr-arm-sve-vector-bits-call.c
  clang/test/CodeGen/attr-arm-sve-vector-bits-cast.c
  clang/test/CodeGen/attr-arm-sve-vector-bits-codegen.c
  clang/test/CodeGen/attr-arm-sve-vector-bits-globals.c
  llvm/include/llvm/IR/IRBuilder.h

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D92761.309901.patch
Type: text/x-patch
Size: 75884 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20201207/e0ebb2f1/attachment-0001.bin>