Remove context dependant `!` fallback #148871

WaffleLapkin · 2025-11-12T15:51:35Z

... and minor cleanup.

r? lcnr

WaffleLapkin · 2025-11-12T16:34:36Z

I have really low confidence in the never type tests, IMO they require a major refactoring, as a lot of them are based on outdated ideas and features...

jackh726 · 2025-11-12T19:57:01Z

For the tests, I would expect us to have revisions for edition 2021 and edition 2024 (for the different fallbacks)

WaffleLapkin · 2025-11-13T12:28:31Z

@jackh726 I'm not sure 2021vs2015 is relevant. But either way, yes, I agree that we should have tests for both current fallback behaviors based on the edition. I'll make a followup which cleans up the tests, wanna make sure that we have relevant and comprehensive tests.

lcnr · 2025-11-13T15:56:18Z

compiler/rustc_hir_typeck/src/fallback.rs

+            // phase of fallback. This means that we only replace
+            // inference variables with their underlying opaque types as a
+            // last resort.


existing, but I do not understand that comment?

Do you understand what this wants to say. It feels like this is a general "if fallback occurred, previously stalled goals may make progress again".

I really don't see how there's anything specific to opaques here.

I don't get it either .-.

the example below is supposed to explain this, but I still don't get it

hmm, want to rip this out and replace it with

if fallback occurred, previously stalled goals may make progress again

even if this comment was somehow relevant, given that we both don't get what it means, I don't think it does us any good

lcnr · 2025-11-13T15:59:35Z

compiler/rustc_trait_selection/src/error_reporting/traits/fulfillment_errors.rs

                            });
                            let unit_obligation = obligation.with(tcx, predicate);
                            if self.predicate_may_hold(&unit_obligation) {
+                                // FIXME: make a new issue for this


Can you elaborate on that FIXME? it feels unactionable for anyone who isn't you right now... or maybe just do that if you can xd

Sure, I opened #148922 and linked it from here. I tried my best to write a good explanation, but it's just confusing :(

lcnr · 2025-11-13T16:00:12Z

tests/ui/coercion/coerce-issue-49593-box-never.rs

@@ -1,9 +1,8 @@
 //@ revisions: nofallback fallback


want to rename the revisions to "unit_fallback never_fallback"?

In a follow up

lcnr · 2025-11-13T16:02:19Z

compiler/rustc_hir_typeck/src/fallback.rs

                    fallback_to(self.tcx.types.unit);
                }
-                DivergingFallbackBehavior::ContextDependent => {
-                    if found_infer_var_info.self_in_trait && found_infer_var_info.output {


found_infer_var_info is dead code now?

Did you mean InferVarInfo? Yes, it is. Removed.

jackh726 · 2025-11-13T16:48:26Z

@jackh726 I'm not sure 2021vs2015 is relevant. But either way, yes, I agree that we should have tests for both current fallback behaviors based on the edition. I'll make a followup which cleans up the tests, wanna make sure that we have relevant and comprehensive tests.

I don't think we need to worry about 2015, just 2024 vs 2021 since that highlights the difference in fallback behavior.

WaffleLapkin · 2025-11-13T21:37:09Z

@rustbot review

compiler/rustc_hir_typeck/src/fallback.rs

lcnr · 2025-11-16T09:37:46Z

@bors try @rust-timer queue

Remove context dependant `!` fallback

rust-bors · 2025-11-16T11:56:16Z

☀️ Try build successful (CI)
Build commit: 2fb4a7a (2fb4a7a100a7d913c8233bdd63e268de1930a7fc, parent: 67c4cf395f243afcb973dacdd39b16895c9ad295)

rust-timer · 2025-11-16T13:15:56Z

Finished benchmarking commit (2fb4a7a): comparison URL.

Overall result: ✅ improvements - no action needed

Benchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf.

@bors rollup=never
@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
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.4%	[-3.0%, -0.1%]	36
Improvements ✅ (secondary)	-0.2%	[-0.5%, -0.1%]	35
All ❌✅ (primary)	-0.4%	[-3.0%, -0.1%]	36

Max RSS (memory usage)

Results (primary -1.1%)

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)	-1.1%	[-1.3%, -1.0%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-1.1%	[-1.3%, -1.0%]	2

Cycles

Results (secondary 3.8%)

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)	3.8%	[3.7%, 4.0%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 473.917s -> 472.516s (-0.30%)
Artifact size: 388.65 MiB -> 388.63 MiB (-0.00%)

lcnr · 2025-11-17T15:31:35Z

@rustbot author

The code supporting it is extremely confusing. At the same time, we have no plans to use this scheme, so there is no value in supporting it.

the only diagnostic that was using this field specifically сares for the never type fallback, not the integer fallback.

I think in this case the early return is harder to read than just an if. "if fallback hasn't occured we don't do..."

One is about context dependant case which I removed a few commits back, the other neither me nor lcnr could understand, soooo :upsidedown_smile:

... now that it's always the same for a certain crate

rustbot · 2025-11-27T21:48:41Z

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

lcnr · 2025-11-28T09:06:34Z

@bors r+ rollup=never

bors · 2025-11-28T09:06:38Z

📌 Commit 37ecc8e has been approved by lcnr

It is now in the queue for this repository.

bors · 2025-11-28T09:16:18Z

⌛ Testing commit 37ecc8e with merge d645a4c...

bors · 2025-11-28T12:26:26Z

☀️ Test successful - checks-actions
Approved by: lcnr
Pushing d645a4c to main...

github-actions · 2025-11-28T12:29:25Z

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 88bd39b (parent) -> d645a4c (this PR)

Test differences

Show 29 test diffs

Stage 1

[ui] tests/ui/never_type/diverging-fallback-control-flow.rs#fallback: pass -> [missing] (J0)
[ui] tests/ui/never_type/diverging-fallback-control-flow.rs#nofallback: pass -> [missing] (J0)
[ui] tests/ui/never_type/feature-gate-never_type_fallback.rs: pass -> [missing] (J0)

Stage 2

[ui] tests/ui/never_type/diverging-fallback-control-flow.rs#fallback: pass -> [missing] (J1)
[ui] tests/ui/never_type/diverging-fallback-control-flow.rs#nofallback: pass -> [missing] (J1)
[ui] tests/ui/never_type/feature-gate-never_type_fallback.rs: pass -> [missing] (J1)

Additionally, 23 doctest diffs were found. These are ignored, as they are noisy.

Job group index

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard d645a4c9c563b80048ce5f32845e754a67f11efa --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

dist-aarch64-apple: 5953.0s -> 8503.9s (+42.9%)
dist-x86_64-apple: 6037.7s -> 7909.2s (+31.0%)
x86_64-gnu-llvm-20: 2435.8s -> 2862.0s (+17.5%)
dist-i586-gnu-i586-i686-musl: 4883.9s -> 5617.4s (+15.0%)
x86_64-gnu-gcc: 3038.8s -> 3454.0s (+13.7%)
x86_64-gnu-tools: 3286.7s -> 3734.0s (+13.6%)
pr-check-1: 1644.1s -> 1859.6s (+13.1%)
i686-gnu-1: 7396.0s -> 8157.6s (+10.3%)
x86_64-gnu-llvm-21-1: 3161.5s -> 3480.5s (+10.1%)
armhf-gnu: 4887.3s -> 5371.1s (+9.9%)

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 · 2025-11-28T13:46:39Z

Finished benchmarking commit (d645a4c): 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)	-0.3%	[-0.4%, -0.1%]	38
Improvements ✅ (secondary)	-0.3%	[-0.5%, -0.1%]	40
All ❌✅ (primary)	-0.3%	[-0.4%, -0.1%]	38

Max RSS (memory usage)

Results (primary 2.0%, secondary 1.2%)

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

	mean	range	count
Regressions ❌ (primary)	2.0%	[0.7%, 3.4%]	2
Regressions ❌ (secondary)	1.2%	[1.2%, 1.2%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.0%	[0.7%, 3.4%]	2

Cycles

Results (secondary -3.4%)

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
Improvements ✅ (secondary)	-3.4%	[-3.4%, -3.4%]	1
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 472.29s -> 473.697s (0.30%)
Artifact size: 386.94 MiB -> 386.90 MiB (-0.01%)

WaffleLapkin added the F-never_type `#![feature(never_type)]` label Nov 12, 2025

rustbot assigned lcnr Nov 12, 2025

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. labels Nov 12, 2025

This comment has been minimized.

Sign in to view

WaffleLapkin force-pushed the never-simplifications branch from 3f4ec16 to b128eae Compare November 13, 2025 12:25

This comment has been minimized.

Sign in to view

WaffleLapkin force-pushed the never-simplifications branch from b128eae to 6b36e86 Compare November 13, 2025 13:26

lcnr reviewed Nov 13, 2025

View reviewed changes

lcnr added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Nov 13, 2025

WaffleLapkin force-pushed the never-simplifications branch from 0f425a6 to 39ac890 Compare November 13, 2025 21:33

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Nov 13, 2025

lcnr reviewed Nov 16, 2025

View reviewed changes

compiler/rustc_hir_typeck/src/fallback.rs Outdated Show resolved Hide resolved

This comment has been minimized.

Sign in to view

rust-bors bot added a commit that referenced this pull request Nov 16, 2025

Auto merge of #148871 - WaffleLapkin:never-simplifications, r=<try>

2fb4a7a

Remove context dependant `!` fallback

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Nov 16, 2025

This comment has been minimized.

Sign in to view

rustbot added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Nov 17, 2025

WaffleLapkin added 10 commits November 27, 2025 16:12

remove support for context dependant ! fallback

e44a6c3

The code supporting it is extremely confusing. At the same time, we have no plans to use this scheme, so there is no value in supporting it.

doc fixes

e492343

track never type fallback specifically

965feb1

the only diagnostic that was using this field specifically сares for the never type fallback, not the integer fallback.

remove an unnecessary negation

a2bed4c

I think in this case the early return is harder to read than just an if. "if fallback hasn't occured we don't do..."

remove outdated comments

6107753

One is about context dependant case which I removed a few commits back, the other neither me nor lcnr could understand, soooo :upsidedown_smile:

don't pass self.<...> as an argument to a method on the same type...

91b058c

don't recompute diverging fallback type

c8a060b

... now that it's always the same for a certain crate

remove feature(never_type_fallback)

b48d701

remove dead code

62c3736

use a new tracking issue for never type changes

37ecc8e

WaffleLapkin force-pushed the never-simplifications branch from 39ac890 to 37ecc8e Compare November 27, 2025 21:48

WaffleLapkin requested a review from lcnr November 27, 2025 21:50

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Nov 27, 2025

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Nov 28, 2025

bors added the merged-by-bors This PR was explicitly merged by bors. label Nov 28, 2025

bors merged commit d645a4c into rust-lang:main Nov 28, 2025
12 checks passed

rustbot added this to the 1.93.0 milestone Nov 28, 2025

WaffleLapkin deleted the never-simplifications branch November 28, 2025 13:33

Remove context dependant ! fallback #148871

Remove context dependant ! fallback #148871

Uh oh!

Conversation

WaffleLapkin commented Nov 12, 2025

Uh oh!

This comment has been minimized.

WaffleLapkin commented Nov 12, 2025

Uh oh!

This comment has been minimized.

jackh726 commented Nov 12, 2025

Uh oh!

WaffleLapkin commented Nov 13, 2025

Uh oh!

This comment has been minimized.

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jackh726 commented Nov 13, 2025

Uh oh!

WaffleLapkin commented Nov 13, 2025

Uh oh!

Uh oh!

lcnr commented Nov 16, 2025

Uh oh!

This comment has been minimized.

This comment has been minimized.

rust-bors bot commented Nov 16, 2025

Uh oh!

This comment has been minimized.

rust-timer commented Nov 16, 2025

Overall result: ✅ improvements - no action needed

Uh oh!

lcnr commented Nov 17, 2025

Uh oh!

rustbot commented Nov 27, 2025

Uh oh!

lcnr commented Nov 28, 2025

Uh oh!

bors commented Nov 28, 2025

Uh oh!

bors commented Nov 28, 2025

Uh oh!

bors commented Nov 28, 2025

Uh oh!

Uh oh!

github-actions bot commented Nov 28, 2025

Test differences

Stage 1

Stage 2

Job duration changes

Uh oh!

rust-timer commented Nov 28, 2025

Overall result: ✅ improvements - no action needed

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Remove context dependant `!` fallback #148871

Remove context dependant `!` fallback #148871