[libcxx-commits] [libcxx] [libc++][ranges] Optimize the performance of `ranges::starts_with` (PR #84570)

[Louis Dionne via libcxx-commits]

ldionne wrote:

@philnik777 What do you think the next step should be? Vectorize `std::equal` when we're on a platform where we know we can do better than `memcmp`?

Note that I think this patch makes sense in principle, but I also don't think we should check it in as-is since it regresses on some platforms. I'd even be fine with checking it in with some sort of `#ifndef _LIBCPP_MISMATCH_FASTER_THAN_MEMCMP` that we could fix in the near future. WDYT?

https://github.com/llvm/llvm-project/pull/84570