Fix rustdoc ICE when checking if a generic arg can be elided

[Colepng]

Fixes #133637

The param_at call expects a index into the full generics list, while index is generated only from the children.

I also noticed a call to skip_norm_wip so I replaced that with skip_normalization

I have one question about adding the regression test, should it live under rustdoc-ui or rustdoc-ui/issue. I will add it later today.

@fmease

[rustbot]

r? @GuillaumeGomez

rustbot has assigned @GuillaumeGomez.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: rustdoc
• rustdoc expanded to 8 candidates
• Random selection from GuillaumeGomez, lolbinarycat, notriddle

[GuillaumeGomez]

Please add a regression test.

[GuillaumeGomez]

This is not the regression test for #133637 as it works fine with current nightly.

View changes since the review

[Colepng]

When I run the test locally, I run into the ICE. It might pass on current nightly since I don't believe it's built with debug assertions enabled.

[rustbot]

Reminder, once the PR becomes ready for a review, use @rustbot ready.

[Colepng]

@rustbot ready

[fmease]

Could you use a smaller reproducer? One, some or all from https://github.com/rust-lang/rust/pull/160295/changes#r3696140723. Thanks!

View changes since the review

[fmease]

Smaller reproducer:

trait Trait<T = ()> {
    type Type<'a, 'b>;
}

struct Struct<T: Trait<()>>(T::Type<'static, 'static>);

View changes since the review

[Colepng]

When I test locally these do not ICE for me. Do you have the backtrace?

[Colepng]

I was able to minimize it to this. I haven't been successful on reproducing the ICE with a struct.

trait Trait<Default = ()> {
    type Type<'a, 'b>;
}

type Type<T> = <T as Trait>::Type<'static, 'static>;