[libc-commits] [libc] [libc] Alternative algorithm for decimal FP printf (PR #123643)

[Simon Tatham via libc-commits]

statham-arm wrote:

This PR has a performance dependency on #123580. I haven't listed a formal dependency between the two, because in _correctness_ terms they're independent. But the claim in my commit message here that the new USE_FLOAT320 algorithm is faster than USE_DYADIC_FLOAT is only true once IntegerToString is optimized. So I don't intend to land this until after #123580.

https://github.com/llvm/llvm-project/pull/123643