[PATCH] D92762: [clang][AArch64][SVE] Avoid going through memory for coerced VLST arguments

[Joe Ellis via Phabricator via cfe-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 a project: clang.
Herald added a subscriber: cfe-commits.

VLST arguments are coerced to VLATs at the function boundary for
consistency with the VLAT ABI. They are then bitcast back to VLSTs in
the function prolog. Previously, this conversion is done through memory.
With the introduction of the llvm.vector.{insert,extract} intrinsic, we
can avoid going through memory here.

Depends on D92761 <https://reviews.llvm.org/D92761>


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D92762

Files:
  clang/lib/CodeGen/CGCall.cpp
  clang/test/CodeGen/aarch64-sve-acle-__ARM_FEATURE_SVE_VECTOR_OPERATORS.c
  clang/test/CodeGen/aarch64-sve-acle-__ARM_FEATURE_SVE_VECTOR_OPERATORS.cpp
  clang/test/CodeGen/attr-arm-sve-vector-bits-call.c
  clang/test/CodeGen/attr-arm-sve-vector-bits-cast.c

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D92762.309903.patch
Type: text/x-patch
Size: 35792 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20201207/1f3860aa/attachment-0001.bin>