Document that t-lang does not need involvement for unobservable intrinsics by RalfJung · Pull Request #161675 · rust-lang/rust

RalfJung · 2026-08-24T14:08:16Z

This turns the FCP (completed here) about intrinsics that could already be written on stable into documentation.

Also use this opportunity to document miri::intrinsic_fallback_is_spec.

r? @tgross35
Cc @rust-lang/lang

rustbot · 2026-08-24T14:08:21Z

⚠️ #[rustc_intrinsic_const_stable_indirect] controls whether intrinsics can be exposed to stable const
code; adding it needs t-lang approval.

cc @rust-lang/wg-const-eval

⚠️ #[miri::intrinsic_fallback_is_spec] must only be used if the function actively checks for all UB cases,
and explores the possible non-determinism of the intrinsic.

cc @rust-lang/miri

Any special-casing of Miri in the standard library requires review.

cc @rust-lang/miri

Some changes occurred to the intrinsics. Make sure the CTFE / Miri interpreter
gets adapted for the changes, if necessary.

cc @rust-lang/miri, @oli-obk, @lcnr

rustbot · 2026-08-24T14:08:23Z

tgross35 is currently at their maximum review capacity.
They may take a while to respond.

…nsics

Urgau · 2026-08-24T15:56:00Z

+//! Intrinsics don't need a body. However, they optionally can have a body, which we call the
+//! "fallback body". This will be used by codegen backends that do not have a dedicated
+//! implementation of the intrinsic, making it easier to add new intrinsics for specific operations
+//! without having to implement them in each codegen backend. The fallback body obviously has to be
+//! a valid implementation of the documented specification of the intrinsic. In some cases, the
+//! fallback body will be *equivalent* to the specification. Note that this is a strong requirement:
+//! if the spec says "UB if input `x` is even", then a valid implementation can just ignore this and
+//! do whatever it wants in that case; an *equivalent* implementation needs to actually check this
+//! condition and trigger UB in that case (e.g. by using `hint::assert_unchecked()`). Similar, if
+//! the spec says "returns `x` or `y` non-deterministically", then an *equivalent* implementation
+//! must actually do non-deterministic choice and return either value (e.g. by invoking some other
+//! language operation that has the same non-determinism). Intrinsics with such a fallback body that
+//! is equivalent to the spec may be marked with `#[miri::intrinsic_fallback_is_spec]`; the fallback
+//! body will then also be used by Miri for UB checking. When in doubt, do not use this attribute or
+//! ask the Miri maintainers for advice.
+//!
+//! Intrinsics are, in general, language extensions. Therefore, t-lang should be involved whenever a
+//! new intrinsic is exposed to stable code. However, if an intrinsic is marked
+//! `#[miri::intrinsic_fallback_is_spec]` with a fallback body that only uses stable features (or if
+//! such a fallback body could be written, but for one reason or another the actual fallback body is
+//! different), and if it also does not make other promises that go beyond observable program
+//! behavior (such as steering the optimizer in a particular direction), then an intrinsic may be
+//! used without t-lang involvement.


Just a remark on my side, no need to do anything about it. This is quite a lot a word to say that if a intrinsic does something new (language wise, ie not using existing language features) it should be validated by T-lang, and everything else just need T-compiler approval.

View changes since the review

The first paragraph explains a lot more than that. It explains the intrinsic_fallback_is_spec which is non-trivial, that's why I used so many words for it.

The second paragraph explains what you summarized. Happy to make it shorter if you have suggestions for how to concretely do that. It's non-trivial to say precisely what "something new" means.

traviscross

Looks good to me. Thanks @RalfJung. r=me (when FCP completes and with @tgross35' r+).

View changes since this review

RalfJung · 2026-09-12T14:41:21Z

The FCP has completed.
@tgross35 friendly reminder that this has been waiting for a while.
Or should I ask for a different reviewer? @Urgau left a comment, maybe they want to r+.

Urgau · 2026-09-14T06:28:42Z

@bors r=traviscross,Urgau rollup

rust-bors · 2026-09-14T06:28:46Z

📌 Commit 93c8b74 has been approved by traviscross,Urgau

It is now in the queue for this repository.

@tgross35

…viscross,Urgau document that t-lang does not need involvement for unobservable intrisics This turns the [FCP](rust-lang#161081 (comment)) (completed [here](rust-lang#161081 (comment))) about intrinsics that could already be written on stable into documentation. Also use this opportunity to document `miri::intrinsic_fallback_is_spec`. r? @tgross35 Cc @rust-lang/lang

@tgross35

…viscross,Urgau document that t-lang does not need involvement for unobservable intrisics This turns the [FCP](rust-lang#161081 (comment)) (completed [here](rust-lang#161081 (comment))) about intrinsics that could already be written on stable into documentation. Also use this opportunity to document `miri::intrinsic_fallback_is_spec`. r? @tgross35 Cc @rust-lang/lang

…uwer Rollup of 11 pull requests Successful merges: - #162752 (`rust-analyzer` subtree update) - #161868 (libtest: never iterate over all tests in `--exact` mode) - #161903 (Fix initialization cycle in `target_config`) - #162240 (Garbage-collect old incremental compilation sessions) - #161675 (document that t-lang does not need involvement for unobservable intrisics) - #162630 (Simplify the `G` in `Diag<'a, G>`) - #162647 (Add regression test for previous overflow evaluating the requirement) - #162703 (regression test for valtree leaf const) - #162723 (Add regression test for unexpected type for constructor) - #162735 (Remove pointless `A: Allocator` bounds in boxed.rs) - #162736 (clean up trivial region constraint filtering)

@tgross35

…viscross,Urgau document that t-lang does not need involvement for unobservable intrisics This turns the [FCP](rust-lang#161081 (comment)) (completed [here](rust-lang#161081 (comment))) about intrinsics that could already be written on stable into documentation. Also use this opportunity to document `miri::intrinsic_fallback_is_spec`. r? @tgross35 Cc @rust-lang/lang

…uwer Rollup of 15 pull requests Successful merges: - #162752 (`rust-analyzer` subtree update) - #161903 (Fix initialization cycle in `target_config`) - #162240 (Garbage-collect old incremental compilation sessions) - #162610 (fix tailcall indirect return) - #162634 (Implement semantic analysis for named `Fn` trait params) - #161675 (document that t-lang does not need involvement for unobservable intrisics) - #162160 (turn aligned-in-packed error into lint) - #162504 (Stabilize `unsafe_cell_access`) - #162516 (tidy: Sort multi-line types by treating `>` as a closing bracket) - #162630 (Simplify the `G` in `Diag<'a, G>`) - #162647 (Add regression test for previous overflow evaluating the requirement) - #162703 (regression test for valtree leaf const) - #162723 (Add regression test for unexpected type for constructor) - #162735 (Remove pointless `A: Allocator` bounds in boxed.rs) - #162736 (clean up trivial region constraint filtering)

…uwer Rollup of 16 pull requests Successful merges: - #162762 (Subtree sync for rustc_codegen_cranelift) - #162752 (`rust-analyzer` subtree update) - #161903 (Fix initialization cycle in `target_config`) - #162240 (Garbage-collect old incremental compilation sessions) - #162610 (fix tailcall indirect return) - #162634 (Implement semantic analysis for named `Fn` trait params) - #161675 (document that t-lang does not need involvement for unobservable intrisics) - #162160 (turn aligned-in-packed error into lint) - #162504 (Stabilize `unsafe_cell_access`) - #162516 (tidy: Sort multi-line types by treating `>` as a closing bracket) - #162630 (Simplify the `G` in `Diag<'a, G>`) - #162647 (Add regression test for previous overflow evaluating the requirement) - #162703 (regression test for valtree leaf const) - #162723 (Add regression test for unexpected type for constructor) - #162735 (Remove pointless `A: Allocator` bounds in boxed.rs) - #162736 (clean up trivial region constraint filtering)

@tgross35

Rollup merge of #161675 - RalfJung:intrinsic-approval, r=traviscross,Urgau document that t-lang does not need involvement for unobservable intrisics This turns the [FCP](#161081 (comment)) (completed [here](#161081 (comment))) about intrinsics that could already be written on stable into documentation. Also use this opportunity to document `miri::intrinsic_fallback_is_spec`. r? @tgross35 Cc @rust-lang/lang

…uwer Rollup of 16 pull requests Successful merges: - rust-lang/rust#162762 (Subtree sync for rustc_codegen_cranelift) - rust-lang/rust#162752 (`rust-analyzer` subtree update) - rust-lang/rust#161903 (Fix initialization cycle in `target_config`) - rust-lang/rust#162240 (Garbage-collect old incremental compilation sessions) - rust-lang/rust#162610 (fix tailcall indirect return) - rust-lang/rust#162634 (Implement semantic analysis for named `Fn` trait params) - rust-lang/rust#161675 (document that t-lang does not need involvement for unobservable intrisics) - rust-lang/rust#162160 (turn aligned-in-packed error into lint) - rust-lang/rust#162504 (Stabilize `unsafe_cell_access`) - rust-lang/rust#162516 (tidy: Sort multi-line types by treating `>` as a closing bracket) - rust-lang/rust#162630 (Simplify the `G` in `Diag<'a, G>`) - rust-lang/rust#162647 (Add regression test for previous overflow evaluating the requirement) - rust-lang/rust#162703 (regression test for valtree leaf const) - rust-lang/rust#162723 (Add regression test for unexpected type for constructor) - rust-lang/rust#162735 (Remove pointless `A: Allocator` bounds in boxed.rs) - rust-lang/rust#162736 (clean up trivial region constraint filtering)

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Aug 24, 2026

rustbot assigned tgross35 Aug 24, 2026

RalfJung force-pushed the intrinsic-approval branch from e04e762 to 453be15 Compare August 24, 2026 14:12

This comment has been minimized.

Sign in to view

document that t-lang does not need involvement for unobservable intri…

93c8b74

…nsics

RalfJung force-pushed the intrinsic-approval branch from 453be15 to 93c8b74 Compare August 24, 2026 14:23

Urgau reviewed Aug 24, 2026

View reviewed changes

traviscross added T-lang Relevant to the language team I-lang-radar Items that are on lang's radar and will need eventual work or consideration. labels Aug 25, 2026

traviscross approved these changes Aug 25, 2026

View reviewed changes

Urgau approved these changes Sep 14, 2026

View reviewed changes

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 Sep 14, 2026

Zalathar mentioned this pull request Sep 14, 2026

Rollup of 7 pull requests #162753

Closed

JonathanBrouwer mentioned this pull request Sep 14, 2026

Rollup of 11 pull requests #162755

Closed

JonathanBrouwer mentioned this pull request Sep 14, 2026

Rollup of 15 pull requests #162764

Closed

JonathanBrouwer mentioned this pull request Sep 14, 2026

Rollup of 16 pull requests #162767

Merged

rust-bors Bot merged commit 41a5222 into rust-lang:main Sep 14, 2026
13 checks passed

rustbot added this to the 1.100.0 milestone Sep 14, 2026

traviscross changed the title ~~document that t-lang does not need involvement for unobservable intrisics~~ document that t-lang does not need involvement for unobservable intrinsics Sep 15, 2026

traviscross changed the title ~~document that t-lang does not need involvement for unobservable intrinsics~~ Document that t-lang does not need involvement for unobservable intrinsics Sep 15, 2026

RalfJung deleted the intrinsic-approval branch September 16, 2026 06:03

Navigation Menu

Uh oh!

Document that t-lang does not need involvement for unobservable intrinsics - #161675

Document that t-lang does not need involvement for unobservable intrinsics#161675
rust-bors[bot] merged 1 commit into
rust-lang:mainfrom
RalfJung:intrinsic-approval

RalfJung commented Aug 24, 2026 •

edited

Loading

Uh oh!

rustbot commented Aug 24, 2026

Uh oh!

rustbot commented Aug 24, 2026

Uh oh!

This comment has been minimized.

Urgau Aug 24, 2026 •

edited by rustbot

Loading

Uh oh!

RalfJung Aug 24, 2026

Uh oh!

traviscross left a comment •

edited by rustbot

Loading

Uh oh!

RalfJung commented Sep 12, 2026

Uh oh!

Urgau commented Sep 14, 2026

Uh oh!

rust-bors Bot commented Sep 14, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

Uh oh!

Conversation

RalfJung commented Aug 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Aug 24, 2026

Uh oh!

rustbot commented Aug 24, 2026

Uh oh!

This comment has been minimized.

Urgau Aug 24, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

RalfJung Aug 24, 2026

Choose a reason for hiding this comment

Uh oh!

traviscross left a comment • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

RalfJung commented Sep 12, 2026

Uh oh!

Urgau commented Sep 14, 2026

Uh oh!

rust-bors Bot commented Sep 14, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

RalfJung commented Aug 24, 2026 •

edited

Loading

Urgau Aug 24, 2026 •

edited by rustbot

Loading

traviscross left a comment •

edited by rustbot

Loading