Enable -Znext-solver on nightly by default

[Kivooeo]

View all comments

Implementation of rust-lang/compiler-team#1014 cc rust-lang/blog.rust-lang.org#1896 #160895

This enables -Znext-solver=globally by default in nightly, but keeps -Znext-solver=coherence for most tests.

[rust-bors]

💔 Test for 13afda5 failed: CI. Failed job:

• auto - x86_64-gnu-stable (web logs, enhanced plaintext logs)

[Kobzol]

@bors try @rust-timer queue

[Kobzol]

Heh, I noticed the failing test while reading the code and was thinking "huh, how does this work?". And now I see it doesn't 😆 It either has to be skipped on nightly, or perhaps I'd just replace it with a check that compares the parsed value to the Default impl of the next solver config.

[rust-bors]

☀️ Try build successful (CI)
Build commit: 97e0817 (97e0817e27d8579734e091d063da910298d6891d)
Base parent: 0cfa1c4 (0cfa1c40c6a6165adbcdb2bc5f973a9d7389186e)

[lcnr]

not part of this PR, but we allow coherence: bool = true even without derive(Default) that's surprising to me 🤔

View changes since the review

[Kobzol]

I was also surprised :) If you read the RFC, it can be used when constructing values of the struct manually (you can skip fields with a default value).

[lcnr]

enabling the tracked check if the next-solver is enabled in 72c9724

I guess let's wait for perf before reapproving :>

[rust-timer]

Finished benchmarking commit (97e0817): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

Benchmarking means the PR may be perf-sensitive. It's automatically marked not fit for rolling up. Overriding is possible but disadvised: it risks changing compiler perf.

Next, please: If you can, justify the regressions found in this try perf run in writing along with @rustbot label: +perf-regression-triaged. If not, fix the regressions and do another perf run. Neutral or positive results will clear the label automatically.

@bors rollup=never rustc-perf
@rustbot label: -S-waiting-on-perf +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)	0.6%	[0.1%, 1.6%]	119
Regressions ❌ (secondary)	0.5%	[0.0%, 1.7%]	151
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.5%	[-8.1%, -0.3%]	27
All ❌✅ (primary)	0.6%	[0.1%, 1.6%]	119

Max RSS (memory usage)

Results (primary 1.3%, secondary -0.5%)

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

	mean	range	count
Regressions ❌ (primary)	1.3%	[0.9%, 3.1%]	8
Regressions ❌ (secondary)	2.0%	[0.8%, 3.2%]	8
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.0%	[-3.1%, -0.6%]	13
All ❌✅ (primary)	1.3%	[0.9%, 3.1%]	8

Cycles

Results (primary 2.5%, secondary -0.9%)

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

	mean	range	count
Regressions ❌ (primary)	2.5%	[2.5%, 2.5%]	1
Regressions ❌ (secondary)	6.9%	[5.0%, 8.9%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.1%	[-4.1%, -2.3%]	7
All ❌✅ (primary)	2.5%	[2.5%, 2.5%]	1

Binary size

Results (primary -0.0%)

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

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.0%	[-0.0%, -0.0%]	3
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.0%	[-0.0%, -0.0%]	3

Bootstrap: 456.922s -> 480.397s (5.14%)
Artifact size: 398.89 MiB -> 400.06 MiB (0.29%)

[jdonszelmann]

@bors r=lcnr,jdonszelmann,kobzol

[rust-bors]

📌 Commit 72c9724 has been approved by lcnr,jdonszelmann,kobzol

It is now in the queue for this repository.

[rust-bors]

☀️ Test successful - CI
Approved by: lcnr,jdonszelmann,kobzol
Duration: 2h 57m 49s
Pushing 526c36b to main...

[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 16a623a (parent) -> 526c36b (this PR)

Test differences

Show 2 test diffs

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

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 526c36b4c477994b7905ec6a084d6150e49076f5 --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. x86_64-gnu-llvm-22-2: 59m 44s -> 1h 39m (+67.2%)
2. armhf-gnu: 56m 28s -> 1h 31m (+61.9%)
3. x86_64-msvc-ext3: 1h 14m -> 1h 57m (+56.6%)
4. x86_64-gnu-gcc-core-tests: 10m 51s -> 16m 48s (+54.8%)
5. dist-powerpc64le-linux-musl: 1h 4m -> 1h 38m (+51.8%)
6. x86_64-gnu-stable: 1h 44m -> 2h 35m (+49.2%)
7. optional-x86_64-gnu-autodiff: 37m 29s -> 55m 5s (+47.0%)
8. x86_64-gnu-tools: 47m 3s -> 1h 8m (+46.3%)
9. x86_64-gnu-llvm-21-2: 1h 9m -> 1h 40m (+45.4%)
10. x86_64-gnu-llvm-21: 36m 39s -> 52m 7s (+42.2%)

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 (526c36b): comparison URL.

Overall result: ❌✅ regressions and improvements - please read:

Our benchmarks found a performance regression caused by this PR.
This might be an actual regression, but it can also be just noise.

Next Steps:

• If the regression was expected or you think it can be justified,
  please write a comment with sufficient written justification, and add
  @rustbot label: +perf-regression-triaged to it, to mark the regression as triaged.
• If you think that you know of a way to resolve the regression, try to create
  a new PR with a fix for the regression.
• If you do not understand the regression or you think that it is just noise,
  you can ask the @rust-lang/wg-compiler-performance working group for help (members of this group
  were already notified of this PR).

@rustbot label: +perf-regression
cc @rust-lang/wg-compiler-performance

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)	0.6%	[0.2%, 1.5%]	124
Regressions ❌ (secondary)	0.5%	[0.0%, 1.6%]	151
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.5%	[-8.1%, -0.3%]	27
All ❌✅ (primary)	0.6%	[0.2%, 1.5%]	124

Max RSS (memory usage)

Results (primary 0.8%, secondary 0.1%)

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

	mean	range	count
Regressions ❌ (primary)	1.0%	[0.5%, 1.9%]	11
Regressions ❌ (secondary)	2.7%	[1.8%, 7.6%]	12
Improvements ✅ (primary)	-1.1%	[-1.1%, -1.1%]	1
Improvements ✅ (secondary)	-2.1%	[-2.8%, -0.6%]	14
All ❌✅ (primary)	0.8%	[-1.1%, 1.9%]	12

Cycles

Results (primary 2.3%, secondary 0.3%)

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

	mean	range	count
Regressions ❌ (primary)	2.3%	[2.3%, 2.3%]	1
Regressions ❌ (secondary)	4.4%	[2.0%, 9.8%]	5
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.1%	[-5.1%, -2.1%]	6
All ❌✅ (primary)	2.3%	[2.3%, 2.3%]	1

Binary size

Results (primary 0.1%, secondary 0.1%)

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

	mean	range	count
Regressions ❌ (primary)	0.1%	[0.0%, 0.2%]	118
Regressions ❌ (secondary)	0.1%	[0.0%, 0.2%]	123
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[0.0%, 0.2%]	118

Bootstrap: 457.604s -> 470.08s (2.73%)
Artifact size: 398.89 MiB -> 400.00 MiB (0.28%)

[Kobzol]

Bootstrap regression is caused by us compiling rustc in rustc-perf with the next trait solver. The rest of the results are caused by PGO perturbations, because we now compile dependencies of benchmark crates using the next trait solver when training PGO.

@rustbot label: +perf-regression-triaged