add regression test

[lcnr]

closes #128887

the errors in that issue are due to two separate issues:

• MIR inlining causing the trait solver to hit the recursion limit (partially fixed in Use a reduced recursion limit in the MIR inliner's cycle breaker #129714)
• using subtyping in method selection for paths (fixed in Relate receiver invariantly in method probe for Mode::Path #129073)

We moved any remaining issues due to MIR inlining into #131960, but keeping #128887 open as well seems unhelpful and confusing.

[rustbot]

r? @jdonszelmann

rustbot has assigned @jdonszelmann.
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

[lcnr]

cc @rust-lang/infra why did rustbot stable nominate this PR 😅

[samueltardieu]

Couldn't this be due to #131960 being a regression from stable to stable? However, it is not even mentioned as being closed by this PR.

[theemathas]

It's configured to stable-nominate all T-compiler PRs that close issues that are regressions from stable to stable. https://github.com/rust-lang/rust/blob/master/triagebot.toml#L64-L67

The PR description says "closes #128887"

[lcnr]

Couldn't this be due to #131960 being a regression from stable to stable? However, it is not even mentioned as being closed by this PR.

What exactly do you mean?

#131960 is a regression from stable due to MIR inlining which still exists. We've made it a harder to hit in #129714 but it still triggers for some crates, so we're keeping #131960 open. #128887 also tracked a separately issue which has since been fully fixed in #129073.

It's configured to stable-nominate all T-compiler PRs that close issues that are regressions from stable to stable. https://github.com/rust-lang/rust/blob/master/triagebot.toml#L64-L67

The PR description says "closes #128887"

Ah, that makes a lot of sense, thanks 😅 👍

[theemathas]

Note: #128887 tracked two different regressions. One of them (triangulate) has been fixed by #129073, and tested in this PR. The other one (one of the geoarrow-rs issues) has already been fixed in #129714 and tested in #129757.