dont assemble shadowed impl candidates

[lcnr]

Fixes rust-lang/trait-system-refactor-initiative#109.

I've originally intended to fix this by supporting lazy reevaluation when rerunning cycles. This ended up being really difficult, see https://github.com/lcnr/search_graph for my notes used while working on this. It is also insufficient for the rayon-hang-2.rs test as we end up with goals which we need to rerun for all combinations of provisional results. While landing such an optimization in the future may still be desirable, it is very difficult and insufficient to fix these hangs. Also see the relevant zulip thread.

I was previously opposed to avoiding assembling shadowed impls as it may prevent future improvements in this area, cc #141226. Going to track this and the reasoning behind it in rust-lang/trait-system-refactor-initiative#226.

r? @BoxyUwU @compiler-errors

[rustbot]

BoxyUwU is currently at their maximum review capacity.
They may take a while to respond.

[lcnr]

@bors rollup=never

We've already benchmarked this before, it's a small perf win, no reason to do it again before merging this PR

[compiler-errors]

r=me after squashing

[lcnr]

@bors r=compiler-errors

[bors]

📌 Commit 9931a92 has been approved by compiler-errors

It is now in the queue for this repository.

[lcnr]

@bors r=compiler-errors

[bors]

📌 Commit a78f92b has been approved by compiler-errors

It is now in the queue for this repository.

[bors]

⌛ Testing commit a78f92b with merge f34ba77...

[bors]

☀️ Test successful - checks-actions
Approved by: compiler-errors
Pushing f34ba77 to master...

[github-actions]

What is this?

This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.

Comparing 7cd9505 (parent) -> f34ba77 (this PR)

Test differences

Show 12 test diffs

Stage 1

• [ui] tests/ui/traits/next-solver/cycles/rayon-hang-1.rs: [missing] -> pass (J1)
• [ui] tests/ui/traits/next-solver/cycles/rayon-hang-2.rs: [missing] -> pass (J1)

Stage 2

• [ui] tests/ui/traits/next-solver/cycles/rayon-hang-1.rs: [missing] -> pass (J0)
• [ui] tests/ui/traits/next-solver/cycles/rayon-hang-2.rs: [missing] -> pass (J0)

Additionally, 8 doctest diffs were found. These are ignored, as they are noisy.

Job group index

• J0: aarch64-apple, aarch64-gnu, aarch64-gnu-llvm-19-1, aarch64-msvc-1, arm-android, armhf-gnu, dist-i586-gnu-i586-i686-musl, i686-gnu-1, i686-gnu-nopt-1, i686-msvc-1, test-various, x86_64-apple-2, x86_64-gnu, x86_64-gnu-llvm-19, x86_64-gnu-llvm-19-2, x86_64-gnu-llvm-20-2, x86_64-gnu-nopt, x86_64-gnu-stable, x86_64-mingw-1, x86_64-msvc-1
• J1: x86_64-gnu-llvm-19-3, x86_64-gnu-llvm-20-3

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard f34ba774c78ea32b7c40598b8ad23e75cdac42a6 --output-dir test-dashboard

And then open test-dashboard/index.html in your browser to see an overview of all executed tests.

Job duration changes

1. dist-aarch64-linux: 5634.6s -> 8809.5s (56.3%)
2. dist-aarch64-apple: 5691.0s -> 6727.0s (18.2%)
3. pr-check-1: 1446.4s -> 1651.2s (14.2%)
4. aarch64-gnu-llvm-19-2: 2154.8s -> 2441.8s (13.3%)
5. aarch64-gnu-debug: 3823.9s -> 4293.3s (12.3%)
6. x86_64-gnu-llvm-19: 2528.0s -> 2774.2s (9.7%)
7. pr-check-2: 2456.3s -> 2691.9s (9.6%)
8. x86_64-rust-for-linux: 2624.7s -> 2875.2s (9.5%)
9. x86_64-gnu-miri: 4588.6s -> 5013.5s (9.3%)
10. x86_64-apple-2: 4072.8s -> 4417.6s (8.5%)

How to interpret the job duration changes?

Job durations can vary a lot, based on the actual runner instance
that executed the job, system noise, invalidated caches, etc. The table above is provided
mostly for t-infra members, for simpler debugging of potential CI slow-downs.

[rust-timer]

Finished benchmarking commit (f34ba77): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.3%	[-2.9%, -0.1%]	7
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary -0.4%, secondary -4.0%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	5.0%	[5.0%, 5.0%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-3.0%	[-5.2%, -0.9%]	2
Improvements ✅ (secondary)	-4.0%	[-10.5%, -0.9%]	8
All ❌✅ (primary)	-0.4%	[-5.2%, 5.0%]	3

Cycles

Results (primary 2.7%, secondary -2.7%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	2.7%	[2.7%, 2.7%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.7%	[-3.0%, -2.4%]	3
All ❌✅ (primary)	2.7%	[2.7%, 2.7%]	2

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 468.551s -> 467.629s (-0.20%)
Artifact size: 376.95 MiB -> 376.97 MiB (0.01%)