Consider NonNull as a pointer type

[qwaz]

PR 1/2 for issue #8045. Add NonNull as a pointer class to suppress false positives like UnsafeCell<NonNull<()>>. However, this change is not sufficient to handle the cases shared in gtk-rs and Rug in the issue.

changelog: none

r? @xFrednet

[qwaz]

I tried and failed to avoid using hard-coded paths for NonNull. I couldn't find NonNull in diagnostic items. Is there a better way to handle that comparison?

[xFrednet]

I tried and failed to avoid using hard-coded paths for NonNull. I couldn't find NonNull in diagnostic items. Is there a better way to handle that comparison?

It sadly doesn't have another identifier besides that. We can add them in the rust repo, but that will take at least a week to be available here again. It's okay to have it like this until we go through them again and add them in rust 🙃

[xFrednet]

The changes are looking good, thank you very much! 🙃

@bors r+

[bors]

📌 Commit 844996b has been approved by xFrednet

[bors]

⌛ Testing commit 844996b with merge 907f6d9...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: xFrednet
Pushing 907f6d9 to master...