make find_similar_impl_candidates even fuzzier

[lcnr]

continues the good work of @BGR360 in #92223. I might have overshot a bit and we're now slightly too fuzzy 😅

with this we can now also simplify simplify_type, which is nice :3

[rust-highfive]

r? @cjgillot

(rust-highfive has picked a reviewer for you, use r? to override)

[cjgillot]

Marking as blocked on #92223.

[lcnr]

😅 the issue was that @BGR360 doesn't have the capacity to finish #92223, so this PR is the only one which should get merged.

should probably have used a clearer wording here

[bors]

☔ The latest upstream changes (presumably #93348) made this pull request unmergeable. Please resolve the merge conflicts.

[cjgillot]

I like the simplification of the logic. However, I feel the compiler now emits too many low-relevance suggestions, for instance integer types. Is there a way to filter them out?

[cjgillot]

Why merge struct/enum/union categories?

[lcnr]

because for adts we always compare their AdtDef anyways, so giving the different kinds different categories doesn't matter

[cjgillot]

Do we need to collect, or can we return an impl Iterator?

[lcnr]

would need rpitit for this, as find_similar_impl_candidates is defined on InferCtxt using an extension trait

[bors]

☔ The latest upstream changes (presumably #92007) made this pull request unmergeable. Please resolve the merge conflicts.

[lcnr]

not 100% happy with which impls are suggested here, but efficiently fixing that probably needs a more general rework of this code which i don't have the capacity for rn

[cjgillot]

not 100% happy with which impls are suggested here, but efficiently fixing that probably needs a more general rework of this code which i don't have the capacity for rn

So let's go with this first step.
r=me after a rebase.

[bors]

☔ The latest upstream changes (presumably #93893) made this pull request unmergeable. Please resolve the merge conflicts.

[lcnr]

@bors r=cjgillot rollup

[bors]

📌 Commit f2aea1e has been approved by cjgillot

[bors]

⌛ Testing commit f2aea1e with merge e9ec070cc26027bcae2069663baaa1b027407c1b...

[rust-log-analyzer]

The job arm-android failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

test fs::tests::read_large_dir ... ok

failures:

---- sys::unix::kernel_copy::tests::bench_file_to_socket_copy stdout ----
thread 'sys::unix::kernel_copy::tests::bench_file_to_socket_copy' panicked at 'called `Result::unwrap()` on an `Err` value: Custom { kind: Uncategorized, error: "failed to lookup address information: No address associated with hostname" }', library/std/src/sys/unix/kernel_copy/tests.rs:130:69

failures:
    sys::unix::kernel_copy::tests::bench_file_to_socket_copy


test result: FAILED. 931 passed; 1 failed; 4 ignored; 0 measured; 0 filtered out; finished in 952.69s

error: test failed, to rerun pass '-p std --lib'

[bors]

💔 Test failed - checks-actions

[lcnr]

seems spurious

@bors retry

[bors]

⌛ Testing commit f2aea1e with merge 52dd59e...

[bors]

☀️ Test successful - checks-actions
Approved by: cjgillot
Pushing 52dd59e to master...

[rust-timer]

Finished benchmarking commit (52dd59e): comparison url.

Summary: This benchmark run shows 2 relevant improvements 🎉 but 6 relevant regressions 😿 to instruction counts.

• Average relevant regression: 0.7%
• Average relevant improvement: -0.3%
• Largest improvement in instruction counts: -0.3% on full builds of inflate check
• Largest regression in instruction counts: 0.8% on full builds of keccak check

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

Next Steps: If you can justify the regressions found in this perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please open an issue or create a new PR that fixes the regressions, add a comment linking to the newly created issue or PR, and then add the perf-regression-triaged label to this PR.

@rustbot label: +perf-regression

[lcnr]

how does this change perf? :o bit confused by the results.

keccak check, full has a 0.78% regression but its detailed data shows that it is 2.0% faster? https://perf.rust-lang.org/detailed-query.html?commit=52dd59ed2154f4158ae37e6994b678a6249a7bb0&base_commit=b321742c6c27494897a88cd5ac17ac20aa3469a1&benchmark=keccak-check&scenario=full

Maybe simplify_type can now get inlined in some places? All other changes should have only impacted error reporting