[libcxx-commits] [PATCH] D97705: [RFC][libc++] Improve std::to_chars for base != 10.

[Louis Dionne via Phabricator via libcxx-commits]

ldionne added a comment.

In D97705#3262177 <https://reviews.llvm.org/D97705#3262177>, @Mordante wrote:

> Thanks for the review! The patch is still RFC to see whether this is the direction we want to go. Hence the remaining `#if` and the lack of polish.

Just to make sure I understand (the review summary is difficult to read for me due to formatting issues), we have a base implementation using `__to_chars_itoa`, and then we also have implementations specific to other bases that use `__to_chars_integral<2>` and friends. When those base-specific implementations are enabled, we get roughly 3x performance improvement on those bases, at the cost of < 256 bytes of size for a lookup table. That's right?

If so, I'd say make the optimization non-customizeable and always use it.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D97705/new/

https://reviews.llvm.org/D97705