THIR patterns: Use `ty::Value` in more places throughout `const_to_pat` #151295

Zalathar · 2026-01-18T11:13:08Z

This PR changes ConstToPat::valtree_to_pat to take a ty::Value instead of separate type and valtree arguments, and propagates that change throughout other parts of the overall const_to_pat implementation.

The biggest under-the-hood change is that instead of combining valtrees from the constant with child types from the parent type, we now just rely on the types already embedded in children of the valtree.

I'm not entirely sure that this is correct, but it doesn't seem to cause any failures in the test suite, and from what I can tell the old way was a holdover from a time when the types weren't readily available from the value side.

My ulterior motive for this cleanup is that I'm hoping to eventually add some extra data to thir::PatKind::Const, which will be a little easier if creation of that PatKind::Const doesn't have to assemble a ty::Value from component parts.

rustbot · 2026-01-18T11:13:13Z

Some changes occurred in match checking

cc @Nadrieril

rustbot · 2026-01-18T11:13:15Z

r? @davidtwco

rustbot has assigned @davidtwco.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Zalathar · 2026-01-18T11:32:27Z

compiler/rustc_mir_build/src/thir/pattern/const_to_pat.rs

-            .map(|(idx, (val, ty))| {
-                let field = FieldIdx::new(idx);
-                // Patterns can only use monomorphic types.
-                let ty = self.tcx.normalize_erasing_regions(self.typing_env, ty);


If I simply delete this normalize_erasing_regions and run the ui test suite with no other changes, I get a bunch of failures.

But if I remove the normalization in conjunction with using types from the valtree, everything appears to work. That suggests to me that using the types from the valtree makes the normalization step redundant.

Yeah the types stored in the valtree will come directly from CTFE and are fully normalized (though contain erased regions, but this is fine for THIR building)

Zalathar · 2026-01-18T11:51:07Z

Hmm, from looking at #150624 it seems like there might be reasons to not rely on the types in a valtree just yet.

cc @BoxyUwU Is it OK to rely to use the types embedded in a valtree for the purposes of lowering constants/literals to patterns, or is that going to interfere with other planned changes?

BoxyUwU · 2026-01-18T12:22:59Z

I think its OK. I'm not sure I'll do any of the follow up work in that issue for a while. It turns out to be really annoying to implement a fix for the 'erased regions problem while the old solver is still around, and while generic_const_exprs exists.

The solution I'd like to implement won't have any bad interactions here I believe. We've already fed ty::Value into other parts of the pattern matching infra so this is consistent:tm:. I think I may have even done something like this PR locally while experimenting with solutions to that issue.

r? BoxyUwU

BoxyUwU · 2026-01-19T09:09:59Z

@bors r+

rust-bors · 2026-01-19T09:10:02Z

📌 Commit 97603c0 has been approved by BoxyUwU

It is now in the queue for this repository.

RalfJung · 2026-01-19T09:28:16Z

@bors rollup=never
Some of the code here is quite perf-critical... I think I tired something similar in the previous PR in this area and then changed my mind based on perf numbers.

BoxyUwU · 2026-01-19T09:35:19Z

@bors r-

ah maybe we should just do a perf run then to avoid a potential land-revert churn

rust-bors · 2026-01-19T09:35:24Z

Commit 97603c0 has been unapproved.

BoxyUwU · 2026-01-19T09:35:31Z

@bors try @rust-timer queue

THIR patterns: Use `ty::Value` in more places throughout `const_to_pat`

rust-bors · 2026-01-19T11:53:06Z

☀️ Try build successful (CI)
Build commit: 04732d4 (04732d4c95e8c3e5c7063767637b578ebf91e919, parent: 3d087e6044bddc65723bf42c76fe4cc33a0076b0)

rust-timer · 2026-01-19T12:33:57Z

Finished benchmarking commit (04732d4): comparison URL.

Overall result: no relevant changes - 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

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

Max RSS (memory usage)

Results (secondary -0.4%)

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)	4.3%	[4.3%, 4.3%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.8%	[-3.4%, -2.2%]	2
All ❌✅ (primary)	-	-	0

Cycles

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

Binary size

Results (primary 0.0%, secondary 0.0%)

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

	mean	range	count
Regressions ❌ (primary)	0.0%	[0.0%, 0.0%]	13
Regressions ❌ (secondary)	0.0%	[0.0%, 0.1%]	6
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.0%	[0.0%, 0.0%]	13

Bootstrap: 473.646s -> 473.434s (-0.04%)
Artifact size: 383.28 MiB -> 383.31 MiB (0.01%)

RalfJung · 2026-01-19T12:35:07Z

Seems perf-neutral. :)

@bors r=BoxyUwU rollup=maybe

rust-bors · 2026-01-19T12:35:10Z

📌 Commit 97603c0 has been approved by BoxyUwU

It is now in the queue for this repository.

@ghost

…uwer Rollup of 15 pull requests Successful merges: - #148623 (Ignore `#[doc(hidden)]` items when computing trimmed paths for printing) - #150550 (Miscellaneous cleanups to borrowck related code) - #150879 (Remove the diagnostic lints) - #150895 (rustc_errors: Add (heuristic) Syntax Highlighting for `rustc --explain`) - #150987 (remote-test-server: Fix header in batch mode) - #151004 (std: implement `sleep_until` on Apple platforms) - #151045 (Simplify some literal-value negations with `u128::wrapping_neg`) - #151119 (Support pointers in type reflection) - #151171 (Do not recover from `Trait()` if generic list is unterminated) - #151231 (HIR typeck cleanup: clarify and re-style `check_expr_unop`) - #151249 (Parse ident with allowing recovery when trying to diagnose) - #151295 (THIR patterns: Use `ty::Value` in more places throughout `const_to_pat`) - #151326 (Remove `DiagMessage::Translated` in favour of `DiagMessage::Str`) - #151361 (add test for issue 61463) - #151371 (Add `S-blocked` to `labels_blocking_approval`) r? @ghost

Rollup merge of #151295 - const-to-pat, r=BoxyUwU THIR patterns: Use `ty::Value` in more places throughout `const_to_pat` This PR changes `ConstToPat::valtree_to_pat` to take a `ty::Value` instead of separate type and valtree arguments, and propagates that change throughout other parts of the overall `const_to_pat` implementation. The biggest under-the-hood change is that instead of combining valtrees from the constant with child types from the parent type, we now just rely on the types already embedded in children of the valtree. I'm not *entirely* sure that this is correct, but it doesn't seem to cause any failures in the test suite, and from what I can tell the old way was a holdover from a time when the types weren't readily available from the value side. --- My ulterior motive for this cleanup is that I'm hoping to eventually add some extra data to `thir::PatKind::Const`, which will be a little easier if creation of that `PatKind::Const` doesn't have to assemble a `ty::Value` from component parts.

@ghost

…uwer Rollup of 15 pull requests Successful merges: - rust-lang/rust#148623 (Ignore `#[doc(hidden)]` items when computing trimmed paths for printing) - rust-lang/rust#150550 (Miscellaneous cleanups to borrowck related code) - rust-lang/rust#150879 (Remove the diagnostic lints) - rust-lang/rust#150895 (rustc_errors: Add (heuristic) Syntax Highlighting for `rustc --explain`) - rust-lang/rust#150987 (remote-test-server: Fix header in batch mode) - rust-lang/rust#151004 (std: implement `sleep_until` on Apple platforms) - rust-lang/rust#151045 (Simplify some literal-value negations with `u128::wrapping_neg`) - rust-lang/rust#151119 (Support pointers in type reflection) - rust-lang/rust#151171 (Do not recover from `Trait()` if generic list is unterminated) - rust-lang/rust#151231 (HIR typeck cleanup: clarify and re-style `check_expr_unop`) - rust-lang/rust#151249 (Parse ident with allowing recovery when trying to diagnose) - rust-lang/rust#151295 (THIR patterns: Use `ty::Value` in more places throughout `const_to_pat`) - rust-lang/rust#151326 (Remove `DiagMessage::Translated` in favour of `DiagMessage::Str`) - rust-lang/rust#151361 (add test for issue 61463) - rust-lang/rust#151371 (Add `S-blocked` to `labels_blocking_approval`) r? @ghost

Use ty::Value in more places throughout const_to_pat

97603c0

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 Jan 18, 2026

rustbot assigned davidtwco Jan 18, 2026

Zalathar commented Jan 18, 2026

View reviewed changes

rustbot assigned BoxyUwU and unassigned davidtwco Jan 18, 2026

rust-bors bot 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 Jan 19, 2026

rust-bors bot 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-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Jan 19, 2026

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Jan 19, 2026

This comment has been minimized.

Sign in to view

rust-bors bot pushed a commit that referenced this pull request Jan 19, 2026

Auto merge of #151295 - Zalathar:const-to-pat, r=<try>

04732d4

THIR patterns: Use `ty::Value` in more places throughout `const_to_pat`

This comment has been minimized.

Sign in to view

rustbot removed the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Jan 19, 2026

rust-bors bot 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-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Jan 19, 2026

JonathanBrouwer mentioned this pull request Jan 19, 2026

Rollup of 15 pull requests #151381

Merged

Zalathar mentioned this pull request Jan 19, 2026

THIR patterns: Explicitly distinguish &pin from plain &/&mut #151282

Open

rust-bors bot merged commit b41e8f0 into rust-lang:main Jan 20, 2026
12 checks passed

rustbot added this to the 1.95.0 milestone Jan 20, 2026

Zalathar deleted the const-to-pat branch January 20, 2026 02:29

Uh oh!

THIR patterns: Use ty::Value in more places throughout const_to_pat #151295

THIR patterns: Use ty::Value in more places throughout const_to_pat #151295

Conversation

Zalathar commented Jan 18, 2026

Uh oh!

rustbot commented Jan 18, 2026

Uh oh!

rustbot commented Jan 18, 2026

Uh oh!

Zalathar Jan 18, 2026

Choose a reason for hiding this comment

Uh oh!

BoxyUwU Jan 18, 2026

Choose a reason for hiding this comment

Uh oh!

Zalathar commented Jan 18, 2026

Uh oh!

BoxyUwU commented Jan 18, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

BoxyUwU commented Jan 19, 2026

Uh oh!

rust-bors bot commented Jan 19, 2026

Uh oh!

RalfJung commented Jan 19, 2026

Uh oh!

BoxyUwU commented Jan 19, 2026

Uh oh!

rust-bors bot commented Jan 19, 2026

Uh oh!

BoxyUwU commented Jan 19, 2026

Uh oh!

This comment has been minimized.

This comment has been minimized.

rust-bors bot commented Jan 19, 2026

Uh oh!

This comment has been minimized.

rust-timer commented Jan 19, 2026

Overall result: no relevant changes - no action needed

Instruction count

Max RSS (memory usage)

Cycles

Binary size

Uh oh!

RalfJung commented Jan 19, 2026

Uh oh!

rust-bors bot commented Jan 19, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

THIR patterns: Use `ty::Value` in more places throughout `const_to_pat` #151295

THIR patterns: Use `ty::Value` in more places throughout `const_to_pat` #151295

BoxyUwU commented Jan 18, 2026 •

edited

Loading