Refactor the code in the `convert_while_ascii` helper function to make
it more suitable for auto-vectorization and also process the full ascii
prefix of the string. The generic case conversion logic will only be
invoked starting from the first non-ascii character.

The runtime on microbenchmarks with ascii-only inputs improves between
1.5x for short and 4x for long inputs on x86_64 and aarch64.

The new implementation also encapsulates all unsafe inside the
`convert_while_ascii` function.

Fixes rust-lang#123712

… builds