Null fn lints

[Niki4tap]

Adds lints to check for code, that assumes nullable fn().

Lint examples:

transmute_null_to_fn:

error: transmuting a known null pointer into a function pointer
  --> $DIR/transmute_null_to_fn.rs:9:23
   |
LL |         let _: fn() = std::mem::transmute(std::ptr::null::<()>());
   |                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ this transmute results in undefined behavior
   |
   = help: try wrapping your function pointer type in `Option<T>` instead, and using `None` as a null pointer value

fn_null_check:

error: function pointer assumed to be nullable, even though it isn't
  --> $DIR/fn_null_check.rs:13:8
   |
LL |     if (fn_ptr as *mut ()).is_null() {}
   |        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
   |
   = help: try wrapping your function pointer type in `Option<T>` instead, and using `is_none` to check for null pointer value

Closes #1644

---

changelog: Improvement: [transmuting_null]: Now detects const pointers to all types
#10099
changelog: New lint: [transmute_null_to_fn]
#10099
changelog: New lint: [fn_null_check]
#10099

[rustbot]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @llogiq (or someone else) soon.

Please see the contribution instructions for more information.

[Niki4tap]

@rustbot label: +A-lint, +C-enhancement, +L-correctness

[llogiq]

This looks good in general, I have a few suggestions regarding code style.

[Niki4tap]

@llogiq I believe all code style issues have been resolved, one thing I would still like to discuss though, is whether fn_null_check lint should be in a different category.

Everything I've said in the suggestion still stands, and I just mostly wanted to mention this concern here specifically, as this is not just a code style issue, and might need to be discussed outside of a suggestion. My opinion is that, while fn_null_check might not seem as impactful as transmute_null_to_fn, they both are checking for the same root issue, which is assuming that fn() can be null, so I believe that their severity should not be separated.

[Niki4tap]

Thanks for the review @llogiq! Is this now ready for merge?

[llogiq]

Thank you!

I'm a bit wary about the correctness group for the null check lint, but that may be overcautious. We can revisit this if there are any reports of false positives or something.

@bors r+

[bors]

📌 Commit 691df70 has been approved by llogiq

It is now in the queue for this repository.

[bors]

⌛ Testing commit 691df70 with merge b3145fe...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: llogiq
Pushing b3145fe to master...