[BUG]: `stdlib_float16_bits_t` is conditionally undefined on compilers with native Float16 support

[officiallyanee]

Description

Encountered an error when attempting to use stdlib_float16_bits_t on a compiler with native _Float16 support. The typedef for stdlib_float16_bits_t in ctor.h is placed inside the #else branch of the native Float16 detection block, meaning it is only defined when no native _Float16 support is detected. On such compilers, attempting to use it as shown in the README results in an "undeclared identifier" error. It's documentation says to prefer explicit helpers but doesn't mention that it's not supported on devices with native float16 support.

Related Issues

none.

Questions

No.

Reproduction

Running the readme usage example with appropriate headers.

Expected Results

For it to compile without errors.

Actual Results

example.c:28:23: error: expected ';' after expression
   28 |         stdlib_float16_bits_t y;
      |                              ^
      |                              ;
example.c:28:2: error: use of undeclared identifier 'stdlib_float16_bits_t'; did you mean 'stdlib_float16_to_bits'?
   28 |         stdlib_float16_bits_t y;
      |         ^~~~~~~~~~~~~~~~~~~~~
      |         stdlib_float16_to_bits

I ran it in the example file using make command.

Checklist

• Read and understood the Code of Conduct.
• Searched for existing issues and pull requests.