memcpy/memset not inlined when __builtin_assume gurantee the data is short enough

[mapleFU]

The code below is in apache arrow cpp[1]. The arrow-rs also has similiar phenomenon[2].

To be short, when size is gurantee to be less or equal to 12, gcc would inline the memcpy and memset but the clang don't optimize this. See godbolt link [3]. The problem is still exists when -ffreestanding is enabled.

c_type makeInline1(const char* data, int32_t size) {
  ARROW_COMPILER_ASSUME(size <= kInlineSize); // __builtin_assume
  c_type out;
  out.inlined = {size, {}};
  // Memcpy for 0 to 12
  memcpy(&out.inlined.data, data, size);
  return out;
}

Would this being a problem? If it can be fixed with some compiler flags, what flag should I use?

[1] https://github.com/apache/arrow/blob/63b34c97c5d3ca6d20dacb9e92b404986f1d7d62/cpp/src/arrow/util/binary_view_util.h#L28 [2] https://github.com/apache/arrow-rs/issues/6034 [3] https://godbolt.org/z/47T8s69xK

[the8472]

Yeah we often see this in Rust that short but variable-length memcpy's don't optimize well.

https://rust.godbolt.org/z/4P9xxeYsc

With AVX it should be possible to do masked moves for multiple-of-word-size types. On AVX512 it would even work for byte-sized ones.

This frequently comes up when attempting to vectorize code by chunking slices into arrays. The variable-length tail then ends up being a small but variable-length copy.