[libc-commits] [libc] [libc][math][c23] Add hypotf16 function (PR #131991)

[Tejas Vipin via libc-commits]

meltq wrote:

Added negative exhaustive tests, removed `hypotf16_hard_to_round` since exhaustive tests would finish executing much faster if `hypotf16` used an optimized version of `fputil::cast` that didn't emulate (14 minutes -> 1 min 45 s on my system). Added a comment suggesting `fputil::hypot<float16>` for cases where cast needs to be emulated, created and linked an issue regarding the optimization of `fputil::cast` https://github.com/llvm/llvm-project/issues/133517

https://github.com/llvm/llvm-project/pull/131991