GH-131296: fix clang-cl warning on Windows in _winapi.c

[chris-eibl]

I think this is a skip news?

• Issue: clang-cl issues many warnings when building on Windows #131296

[chris-eibl]

fix warning : incompatible pointer types passing 'int *' to parameter of type 'PDWORD'
(aka 'unsigned long *') [-Wincompatible-pointer-types] in ..\Modules_winapi.c(648,50):

cpython/Modules/_winapi.c

Lines 647 to 648 in bc26f95

	if (!AdjustTokenPrivileges(token, FALSE, &tp.base, sizeof(previousTp),
	&previousTp.base, &previousTpSize)){

[chris-eibl]

@zooba Since this touches _winapi.c, most probably you are a suitable reviewer?

I hope it is not too impolite to directly ask, I am still unsure about the procedure here.

In case I should wait until this is assigned to a core dev - please let me know.

OOI: how do PRs get assigned (if not automatically via code ownership - this is clear to me)?

Do core devs go through awaiting review and pick up according to their liking / bandwith?

Do triagers assign reviews?

[zooba]

Tagging me for Windows-related stuff is fine, or the triage team will know to do it. Usually CODEOWNERS should cover most PRs, but if not then yeah, either the triagers or code devs will review the queue.

Probably most often someone is already involved in the issue, and so they'll look at the PR. Usually we aren't doing as many PRs for a single issue as you have been here, but since I said to do it that way, you can keep pinging me on them.

[chris-eibl]

Yeah, I also had them all as a big change, but honestly, I wouldn't have liked to review it. It was just super big and touched many files with many unrelated warnings.

In hindsight - doing them on a per file basis seems too fine grained, but I could not find a better way, except for some recent PRs, where it felt quite natural to group by warning (e.g. -Wunused) - and because that splattered not all over the whole code base, but just covered a few files.

Anyway, thanks for bearing with me 👍

[chris-eibl]

Tagging me for Windows-related stuff is fine

Almost all of that is Windows-related, mostly in #ifed code parts, sometimes in Windows specific files, sometimes in vcproj files.

Because gcc and clang on Linux have already done their warning job :)

There have been only very few other occurrencies, e.g. when sizeof(long) == 4 in 64bit builds, but this is kind of Windows specific, too.

So I am afraid and still will have to get on your nerves on some more ...

... just don't wanna seem pushing - all of this can sit and wait.