Cleanup: Use rustc's same_type for our same_tys

[phansch]

This delegates our same_tys to ty::TyS::same_type to
remove some duplication.

Our same_tys was introduced 4 years ago in #730. Before, it was
building an inference context to be able to call
can_eq to compare the types. The rustc-dev-guide has some more details
about can_eq in Type Inference -> Trying equality.

Now, using the rustc function, we are essentially comparing the DefIds
of the given types, which also makes more sense, IMO.

I also confirmed that the FIXME is resolved via a bit of dbg!, however
no UI tests seem to have been affected.

---

changelog: none

[phansch]

Ah, it might make sense to also check whether or not the FIXME was resolved already without this PR. Will do that tomorrow.

[phansch]

Confirmed that the FIXME was already resolved without these changes:

I removed our same_tys and updated all call sites to use the rustc function instead

[flip1995]

Hmm, generally TyS should never be used outside of the ty module. This should be enforced by an internal rustc lint. But this lint is one of the lints, I wanted to fix for quite some time. Also same_type would have to be re-exported to prevent the usage of TyS in Clippy.

[phansch]

Hmm, generally TyS should never be used outside of the ty module.

A right, I'm going to give re-exporting a try later today 👍

[bors]

☔ The latest upstream changes (presumably #5583) made this pull request unmergeable. Please resolve the merge conflicts.

[flip1995]

@phansch Did you have any success with reexporting?

[phansch]

@phansch Did you have any success with reexporting?

Not really. same_type is the only associated function of TyS. Everything else is a method and accessible via instances of Ty. As far as I can tell the function would either have to be turned into a method or moved out of TyS?

[flip1995]

As far as I can tell the function would either have to be turned into a method or moved out of TyS?

We can just keep it as it is right now, since the internal rustc lint doesn't trigger here. So I would be fine with this right now. (This again reminds me, that I wanted to revisit this internal rustc lint...)

[flip1995]

so rebase + CI pass = r=me

[phansch]

@bors r=flip1995

[bors]

📌 Commit 623faac has been approved by flip1995

[bors]

⌛ Testing commit 623faac with merge e6f05ad...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: flip1995
Pushing e6f05ad to master...