[all-commits] [llvm/llvm-project] dad07b: [clang][AArch64][SVE] Avoid going through memory f...

[Joe Ellis via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: dad07baf123e672b1d5d5e7c21e73b92399d5a0c
      https://github.com/llvm/llvm-project/commit/dad07baf123e672b1d5d5e7c21e73b92399d5a0c
  Author: Joe Ellis <joe.ellis at arm.com>
  Date:   2020-12-16 (Wed, 16 Dec 2020)

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

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

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

Reviewed By: c-rhodes

Differential Revision: https://reviews.llvm.org/D92761