[all-commits] [llvm/llvm-project] b5f04d: [libc] Use different alignment for memcpy between ...

[Guillaume Chatelet via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: b5f04d81a2c0b01db32ed7c1685be37e6bdce171
      https://github.com/llvm/llvm-project/commit/b5f04d81a2c0b01db32ed7c1685be37e6bdce171
  Author: Guillaume Chatelet <gchatelet at google.com>
  Date:   2021-04-26 (Mon, 26 Apr 2021)

  Changed paths:
    M libc/src/string/aarch64/memcpy.cpp
    M libc/src/string/memcpy.cpp
    M libc/src/string/memory_utils/memcpy_utils.h
    M libc/src/string/x86/memcpy.cpp
    M libc/test/src/string/memory_utils/memcpy_utils_test.cpp

  Log Message:
  -----------
  [libc] Use different alignment for memcpy between ARM and x86.

Aligned copy used to be 'destination aligned' for x86 but this decision was reverted in D93457 where we noticed that it was better for ARM to be 'source aligned'.
More benchmarking confirmed that it can be up to 30% faster to align copy to destination for x86. This Patch offers both implementations and switches x86 back to destination aligned.
It also fixes alignment to 32 byte on x86.

Differential Revision: https://reviews.llvm.org/D101296