I'm happy to include a news entry and issue if desired, we chatted about this problem briefly in Discord and I figured I'd just quickly post the patch.

CI has some compiler warnings

Run python3 Tools/build/check_warnings.py --compiler-output-file-path=compiler_output_macos.txt --warning-ignore-file-path=Tools/build/.warningignore_macos --compiler-output-type=clang --fail-on-regression --fail-on-improvement --path-prefix="./"
Unexpected warnings:
Python/remote_debug.h expected 0 warnings, found 2
{'file': 'Python/remote_debug.h', 'line': '[13](https://github.com/python/cpython/actions/runs/15448318268/job/43483840500?pr=135146#step:10:14)5', 'column': '1', 'message': "unused function '_Py_RemoteDebug_ClearCache'", 'option': '-Wunused-function'}
{'file': 'Python/remote_debug.h', 'line': '993', 'column': '1', 'message': "unused function '_Py_RemoteDebug_PagedReadRemoteMemory'", 'option': '-Wunused-function'}

Yeah this file is included in two places but they don't use all the functions. We must silence these warnings because they are false positives

I'm happy to include a news entry and issue if desired, we chatted about this problem briefly in Discord and I figured I'd just quickly post the patch.

No need. This is not a user visible change and it doesn't really qualify as a bug fix (depends how you look at it). I am happy either way basically

@pablogsal I've silenced the warnings.

Hummm, I still see the warnings in GitHub in the https://github.com/python/cpython/pull/135146/files page. Is this expected?

Yeah I think so? I'm not familiar with the warnings tooling, but I believe I've just followed the "standard" way to silence new warnings in a file. If there's another way, we'll need to hear from someone who's an owner of such things :)

Maybe set this:

#ifdef __clang__
    #define UNUSED __attribute__((unused))
#elif defined(__GNUC__)
    #define UNUSED __attribute__((unused))
#elif defined(_MSC_VER)
    #define UNUSED __pragma(warning(suppress: 4505))
#else
    #define UNUSED
#endif

UNUSED static void my_function(void) {
    // Function body
}

Yeah, as @pablogsal suggests, I'd use diagnostic pragmas to silence these warnings (because MSVC does not feature __attribute__((unused)) AFAIK). There is already infrastructure for that in

I'd furthermore suggest to enable the warnings again using _Py_COMP_DIAG_POP to limit the scope of the supressions.

On a second thought please scratch my suggestion above. If using

#define UNUSED __pragma(warning(suppress: 4505))

works like @pablogsal suggested above, then we're fine, since suppress

Pushes the current state of the pragma on the stack, disables the specified warning for the next line, and then pops the warning stack so that the pragma state is reset

Sorry for the noise.