[libc][math] Do not use float16 basic operations in hypotf16. (#155177)

Compiler runtimes for float16 basic operations might not be correctly
rounded for all rounding modes.

GitOrigin-RevId: af0f85ce0be02f9af028ca8a99248aca11f1b456
3 files changed