Introduce `Unnormalized` wrapper by adwinwhite · Pull Request #155083 · rust-lang/rust

adwinwhite · 2026-04-10T07:23:20Z

This is the first step of the eager normalization series.

This PR introduce an Unnormalized wrapper and make most normalization routines consume it. The purpose is to make normalization explicit.
This PR contains no behavior change.

API changes are in the first two commit.
There're some normalization routines left untouched:

normalize in the type checker of borrowck: better do it together with field.ty() returning Unnormalized.
normalize_with_depth: only used inside the old solver. Can be done later.
query_normalize: rarely used.
misc local normalization helpers.

The compiler errors are mostly fixed via ast-grep, with exceptions handled manually.

adwinwhite · 2026-04-10T07:24:17Z

@bors try @rust-timer queue

Introduce `Unnormalized` wrapper

adwinwhite · 2026-04-10T07:38:24Z

@bors try cancel

rust-bors · 2026-04-10T07:38:29Z

Try build cancelled. Cancelled workflows:

https://github.com/rust-lang/rust/actions/runs/24231666610

adwinwhite · 2026-04-10T09:31:13Z

@bors try @rust-timer queue

Introduce `Unnormalized` wrapper

rust-bors · 2026-04-10T11:43:50Z

☀️ Try build successful (CI)
Build commit: 10d0ef4 (10d0ef4e7995ca1d162f402be9f0b39b15d7a78a, parent: 7659cec4ed1457bc0d1f636127e66e8fe5008928)

rust-timer · 2026-04-10T12:24:19Z

Finished benchmarking commit (10d0ef4): comparison URL.

Overall result: no relevant changes - no action needed

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.

@bors rollup=never
@rustbot label: -S-waiting-on-perf -perf-regression

Instruction count

This perf run didn't have relevant results for this metric.

Max RSS (memory usage)

Results (primary -2.4%, secondary -6.4%)

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)	-2.4%	[-2.4%, -2.4%]	1
Improvements ✅ (secondary)	-6.4%	[-6.4%, -6.4%]	1
All ❌✅ (primary)	-2.4%	[-2.4%, -2.4%]	1

Cycles

Results (secondary -0.7%)

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)	2.4%	[2.1%, 2.7%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.7%	[-3.8%, -3.6%]	2
All ❌✅ (primary)	-	-	0

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 492.281s -> 489.321s (-0.60%)
Artifact size: 395.57 MiB -> 395.56 MiB (-0.00%)

adwinwhite · 2026-04-10T13:41:30Z

r? @lcnr

lcnr

Things I'd like to see in this PR maybe?

TypeErrCtxt::normalize_fn_sig should take Unnormalized. that should simplify the diff
explicit_item_self_bounds is an iterator over Unnormalized<(Whatever, Span)>, it should also have the span outside of it

Things for future PRs

I dislike .map(|(c, s)| (c.skip_norm_wip(), s)) as a pattern, we should skip_norm_wip once we iterate over the collection
all of the t == cx.tcx.normalize_erasing_regions(cx.typing_env(), t) asserts should be some assert_fully_normalized and don't use Unnormalized::new_wip
a bunch of these normalize calls in codegen are unnecessary. we should change them to use assert_fully_normalized instead
a big one is struct_tail_raw whose normalize should take an Unnormalized
all calls to skip_norm_wip().def_id should be replaced with a def_id() function on Unnormalized<Whateverhasadefid>
we should add fn Unnormalized::<T>::map<U>(self, f: impl FnOnce(T) -> U) -> Unnormalized<U>
unsure what to do in type_of and predicates_of as these queries very intentionally ferry unnormalized types around. I guess we just have a .skip_normalization() for places where it's intentional 🤔
also slightly unsure about elaborate, I think it should take Unnormalized as it also just returns unnormalized things?
field.ty should return Unnormalized
normalize_with_depth_to should take Unnormalized 🤔
coherence::ImplHeader should its things be Unnormalized I think?
DropckConstraint should have Unnormalized fields? 🤔
the DeeplyNormalize QueryTypeOp should contain an Unnormalized<T> as input

View changes since this review

lcnr · 2026-04-15T16:21:40Z

reviewed all changes, and would like to merge this soon 😁 makes it easier to work off it

adwinwhite · 2026-04-16T08:22:22Z

Do we need to make the normalze method in borrowck's type checker consume Unnormalized in this PR? Ask because it's not in the final summary comment. 🤔
I leave it out because it would create a smaller diff if it's done with field.ty api change.

lcnr · 2026-04-16T08:30:42Z

fine to do it later 👍

adwinwhite · 2026-04-16T14:00:12Z

explicit_item_self_bounds is an iterator over Unnormalized<(Whatever, Span)>, it should also have the span outside of it

explicit_item_self_bounds and other predicates_of variants return EarlyBinder. Their iter_instantiated(_copied) call can't be changed to return (Unnormalized<...>, Span).

Other than that, I should've resolved all review comments.

lcnr · 2026-04-17T08:23:13Z

+    let mut hybrid_preds: Vec<_> =
+        impl_predicates.instantiate_identity(tcx).predicates.into_iter().collect();


? what is this doing

View changes since the review

Ah, sorry, I reordered the code a bit. This collect isn't needed anymore.

lcnr · 2026-04-17T08:29:18Z

            let impl_args = self.fresh_args_for_item(base_expr.span, impl_def_id);
-            let impl_trait_ref =
-                self.tcx.impl_trait_ref(impl_def_id).instantiate(self.tcx, impl_args);
+            let impl_trait_ref = self
+                .tcx
+                .impl_trait_ref(impl_def_id)
+                .instantiate(self.tcx, impl_args)
+                .skip_norm_wip();
            let cause = self.misc(base_expr.span);

            // Match the impl self type against the base ty. If this fails,
            // we just skip this impl, since it's not particularly useful.
-            let impl_trait_ref = ocx.normalize(&cause, self.param_env, impl_trait_ref);
+            let impl_trait_ref =
+                ocx.normalize(&cause, self.param_env, Unnormalized::new_wip(impl_trait_ref));


that one is just skip_Norm_wip into new_wip

View changes since the review

lcnr · 2026-04-17T08:29:40Z

            let impl_args = self.fresh_args_for_item(base_expr.span, impl_def_id);
-            let impl_trait_ref =
-                self.tcx.impl_trait_ref(impl_def_id).instantiate(self.tcx, impl_args);
+            let impl_trait_ref = self
+                .tcx
+                .impl_trait_ref(impl_def_id)
+                .instantiate(self.tcx, impl_args)
+                .skip_norm_wip();
            let cause = self.misc(base_expr.span);

            // Match the impl self type against the base ty. If this fails,
            // we just skip this impl, since it's not particularly useful.
-            let impl_trait_ref = ocx.normalize(&cause, self.param_env, impl_trait_ref);
+            let impl_trait_ref =
+                ocx.normalize(&cause, self.param_env, Unnormalized::new_wip(impl_trait_ref));


lcnr · 2026-04-17T08:36:56Z

+/// A wrapper for value that needs normalization.
+/// FIXME: This is very WIP. The plan is to replace the `skip_norm_wip` spread
+/// throughout the codebase with proper normalization. This is the first step toward
+/// switching to eager normalization with the next solver. See the normalization refactor plan
+/// [here](https://rust-lang.zulipchat.com/#narrow/channel/364551-t-types.2Ftrait-system-refactor/topic/Eager.20normalization.2C.20ahoy.21/with/582996293)
+/// We're in a weird intermediate state due to the change is big to land in a single PR.
+/// A lot of active work is going on which is tracked in #155345.
+/// While this is in progress, if you encounter this type, just use `Unnormalized::new_wip`
+/// and `Unnormalized::skip_norm_wip` as needed.
+///
+/// The interner type parameter exists to constraint generic for certain impl, e.g.,
+/// `Unnormalized<I, I::Clause>`.


Suggested change

/// A wrapper for value that needs normalization.

/// FIXME: This is very WIP. The plan is to replace the `skip_norm_wip` spread

/// throughout the codebase with proper normalization. This is the first step toward

/// switching to eager normalization with the next solver. See the normalization refactor plan

/// [here](https://rust-lang.zulipchat.com/#narrow/channel/364551-t-types.2Ftrait-system-refactor/topic/Eager.20normalization.2C.20ahoy.21/with/582996293)

/// We're in a weird intermediate state due to the change is big to land in a single PR.

/// A lot of active work is going on which is tracked in #155345.

/// While this is in progress, if you encounter this type, just use `Unnormalized::new_wip`

/// and `Unnormalized::skip_norm_wip` as needed.

///

/// The interner type parameter exists to constraint generic for certain impl, e.g.,

/// `Unnormalized<I, I::Clause>`.

/// A wrapper for values that need normalization.

///

/// FIXME(#155345): This is very WIP. The plan is to replace the `skip_norm_wip`

/// spread throughout the codebase with proper normalization. This is the first

/// step toward switching to eager normalization with the next solver. See the

/// normalization refactor plan [here].

///

/// We're in a weird intermediate state as the change is too big to land in a

/// single PR. While this work is in progress, just use `Unnormalized::new_wip`

/// and `Unnormalized::skip_norm_wip` as needed.

///

/// The interner type parameter exists to constraint generic for certain impl,

/// e.g., `Unnormalized<I, I::Clause>`.

///

/// [here]: https://rust-lang.zulipchat.com/#narrow/channel/364551-t-types.2Ftrait-system-refactor/topic/Eager.20normalization.2C.20ahoy.21/with/582996293

View changes since the review

.explicit_super_predicates_of(def_id) .iter_identity_copied() .map(Unnormalized::skip_norm_wip)

the skipping above is needed since iter_identitidy_copied returns iterator over Unnoramlized<(..., span)> while the next method expects pair tuple.

lcnr · 2026-04-17T08:38:15Z

                    let ct = lower_const_arg_for_rustdoc(cx.tcx, const_arg, cx.tcx.types.usize);
                    let typing_env = ty::TypingEnv::post_analysis(cx.tcx, *def_id);
-                    let ct = cx.tcx.normalize_erasing_regions(typing_env, ct);
+                    let ct = cx.tcx.normalize_erasing_regions(typing_env, Unnormalized::new(ct));


Suggested change

let ct = cx.tcx.normalize_erasing_regions(typing_env, Unnormalized::new(ct));

let ct = cx.tcx.normalize_erasing_regions(typing_env, Unnormalized::new_wip(ct));

so for the other uses in this file

View changes since the review

lcnr · 2026-04-17T08:39:53Z

@@ -381,7 +382,7 @@ impl<'tcx> NonCopyConst<'tcx> {
    ) -> bool {
        // Make sure to instantiate all types coming from `typeck` with `gen_args`.
        let ty = EarlyBinder::bind(typeck.expr_ty(e)).instantiate(tcx, gen_args);
-        let ty = tcx.try_normalize_erasing_regions(typing_env, ty).unwrap_or(ty);
+        let ty = tcx.try_normalize_erasing_regions(typing_env, ty).unwrap_or(ty.skip_normalization());


Suggested change

let ty = tcx.try_normalize_erasing_regions(typing_env, ty).unwrap_or(ty.skip_normalization());

let ty = tcx.try_normalize_erasing_regions(typing_env, ty).unwrap_or(ty.skip_norm_wip());

View changes since the review

lcnr

final nit pls cleanup commit history, then r=me

View changes since this review

lcnr · 2026-04-17T09:31:58Z

@bors rollup=never

adwinwhite · 2026-04-17T10:07:07Z

final nit pls cleanup commit history

Should I just squash or isolate the API change in one commit for this big PR?

Also unsure why I can't reply to single inline review. If you missed it,
skip_norm_wip after explicit_super_predicates_of is unavoidable.
See comment

lcnr · 2026-04-17T11:56:09Z

Should I just squash or isolate the API change in one commit for this big PR?

whatever u prefer, as long as the "address review" commits are gone :3

lcnr · 2026-04-17T11:56:12Z

r=me

rustbot · 2026-04-17T17:41:46Z

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.

adwinwhite · 2026-04-17T22:29:11Z

@bors r=lcnr

rust-bors · 2026-04-17T22:29:17Z

📌 Commit a73a9c7 has been approved by lcnr

It is now in the queue for this repository.

rust-bors · 2026-04-18T02:16:37Z

☔ The latest upstream changes (presumably #155404) made this pull request unmergeable. Please resolve the merge conflicts.

This pull request was unapproved.