Let's double check this with crater.

@bors2 try

⌛ Trying commit 75ec279 with merge 1383df0...

To cancel the try build, run the command @bors try cancel.

☀️ Try build successful (CI)
Build commit: 1383df0 (1383df06117fd218068d1f39fd5326256dd4a348, parent: ca77504943887037504c7fc0b9bf06dab3910373)

@craterbot check

👌 Experiment pr-145181 created and queued.
🤖 Automatically detected try build 1383df0
🔍 You can check out the queue and this experiment's details.

i️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

🚧 Experiment pr-145181 is now running

i️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[...] as related queries can now handle type inference.

I think the more likely thing here is that we're now being more responsible and only calling has_significant_drop on types that don't have infer vars.

I did actually try this- the inference block I removed is certainly being hit in a few cases. I can rerun that experiment to find exactly which if you'd like.

Are you thinking the related queries still don't cope with infer vars? If so, do you know of a way to test this? (locally, not with craterbot ofc)

@compiler-errors https://github.com/rust-lang/rust/blob/master/tests/incremental/issue-86753.rs is an example for where has_significant_drop is called with an infer var, although admittedly a complex one. Works fine with the block removed.

I do see an argument for favoring canonicalization regardless- I just see this as an avenue for reducing complexity. Whatever seems to be the most appropriate.

🎉 Experiment pr-145181 is completed!
📊 12 regressed and 6 fixed (679566 total)
📰 Open the summary report.

⚠️ If you notice any spurious failure please add them to the denylist!
i️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

r? @compiler-errors

compiler-errors is not on the review rotation at the moment.
They may take a while to respond.