Don't mark "safe" intrinsics as unsafe

[GuillaumeGomez]

A good example of this is intrinsics::abort.

Before:

After:

cc @jyn514
r? @lqd

[jyn514]

Can you add a test for this?

[GuillaumeGomez]

Well, I have no idea how to do that... Only a few specific intrinsics listed in the typeck function are safe to call, and we don't run DOM tests on std/core. If you have an idea by any chance?

[jyn514]

I thought intrinsics were part of the compiler, not the standard library?

Anyway, you can replicate however the standard library does it in a standalone file with #![no_core] and then test that.

[GuillaumeGomez]

Oh good idea!

[GuillaumeGomez]

Done, thanks again for the tip!

[lqd]

Nicely done. The test is exactly what I was hoping for (thanks @jyn514).

Thanks for taking care of it so quickly !
@bors r+

[bors]

📌 Commit 58617eb492a72efb581884f48565a2e7e8c2dcfe has been approved by lqd

[lqd]

@bors r- until we check whether this is applied to all hir::ForeignItem and not just the intrinsics ?

[GuillaumeGomez]

I added the abi check.

[lqd]

Could we add something like this to the test ? core::intrinsics::needs_drop is safe as well

extern "C" {
    // @has 'foo/fn.needs_drop.html'
    // @has - '//pre[@class="rust fn"]' 'pub unsafe extern "C" fn needs_drop() -> !'
    pub fn needs_drop() -> !;
}

[GuillaumeGomez]

@lqd Added your test as well!

[lqd]

@bors r+

[bors]

📌 Commit f683040 has been approved by lqd