Add very basic "comptime" fn implementation #148820

oli-obk · 2025-11-11T12:25:32Z

Implements functions that cannot be called at runtime (and thus also not from normal const functions, as those could be called from runtime).

This is done via the internal attribute rustc_comptime that can be added to normal functions, turning them into compile-time-only functions.

Because @fee1-dead and @compiler-errors did amazing work, we even get trait bounds that work inside comptime fns: via unconditionally-const const Trait bounds.

Use cases are

Reflection MVP #146923
const heap intrinsics
and the other various intrinsics (e.g. size_of 😆) that will just ICE in codegen or panic at runtime if they actually end up in runtime code

project goal issue: rust-lang/rust-project-goals#406

no tracking issue until we have a feature gate and some sort of syntax

cc @scottmcm as the T-lang goal champion

rustbot · 2025-11-11T12:25:35Z

Some changes occurred in compiler/rustc_attr_parsing

cc @jdonszelmann

Some changes occurred in compiler/rustc_passes/src/check_attr.rs

cc @jdonszelmann

Some changes occurred in compiler/rustc_hir/src/attrs

cc @jdonszelmann

Some changes occurred to the CTFE machinery

cc @RalfJung, @oli-obk, @lcnr

Some changes occurred in src/tools/rustfmt

cc @rust-lang/rustfmt

rustbot · 2025-11-11T12:25:38Z

r? @JonathanBrouwer

rustbot has assigned @JonathanBrouwer.
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

rustbot · 2025-11-11T12:37:13Z

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

RalfJung · 2025-11-11T12:42:42Z

compiler/rustc_hir/src/hir.rs

+    Comptime,
    Const,


Seems worth doc comments distinguishing these two.

fmease · 2025-11-11T12:46:49Z

compiler/rustc_parse/src/parser/item.rs

        let fn_span = self.token.span;
-        let header = self.parse_fn_front_matter(vis, case, FrontMatterParsingMode::Function)?; // `const ... fn`
+        let mut header = self.parse_fn_front_matter(vis, case, FrontMatterParsingMode::Function)?; // `const ... fn`
+        if let Some(attr) = attrs.iter().find(|attr| attr.has_name(sym::rustc_comptime)) {


This doesn't account for #[cfg_attr(…, rustc_comptime)]. I would just leave the parser and ast::Const alone and do the lowering to hir::Constness (which would still have Comptime) in AST lowering. This should simplify things a lot.

Even if you anticipate it being turned into a keyword / qualifier eventually, you can't assume that and it would be super easy anyway to update the parser & AST once it's truly necessary.

fmease · 2025-11-11T13:05:39Z

compiler/rustc_hir/src/hir.rs

 }

 #[derive(Copy, Clone, PartialEq, Eq, Debug, Encodable, Decodable, HashStable_Generic)]
 pub enum Constness {


Personally speaking I would rename the variants like so:

pub enum Constness { Always, // `#[rustc_comptime]` Maybe, // `const` Never, }

to mirror BoundConstness which is Always (for const), Maybe (for [const]) and Never. This makes it crystal clear what they represent and leaves no room for ambiguity.

If it were up to me, I would even rename Constness to ItemConstness if traits will be allowed to be compiletime, too. Otherwise, I would split it into FnConstness and TraitConstness.

I prefer this naming over the suggestion I made in https://github.com/rust-lang/rust/pull/148820/files#r2517464261 actually

fee1-dead · 2025-11-11T18:36:25Z

compiler/rustc_feature/src/builtin_attrs.rs

    ),
+    rustc_attr!(
+        rustc_comptime, Normal, template!(Word), ErrorFollowing,
+        EncodeCrossCrate::Yes,


It this necessary? The other queries would store this information

find_attr!(tcx.get_all_attrs(), AttributeKind::Comptime(..))

would do the same (slightly slower?) but avoids a new query which might be worth it. You can set CrossCrateEncoding for attributes in compiler/rustc_hir/src/attrs/encode_cross_crate.rs

In fact, you did mark it as cross-crate encodable oli, so if we keep the query you can turn cross crate encoding of the attribute off (see my other comment)

fee1-dead · 2025-11-11T18:40:11Z

compiler/rustc_hir_typeck/src/callee.rs

+        if let hir::Constness::Comptime = self.tcx.constness(callee_did) {
+            match const_context {
+                Some(hir::ConstContext::Const { .. } | hir::ConstContext::Static(_)) => {}
+                Some(hir::ConstContext::ConstFn) | None => {
+                    self.dcx().span_err(span, "comptime fns can only be called at compile time");
+                }
+            }


This doesn't enforce cross crate if I use a comptime function in fn if my crate doesn't enable const_trait_impl.

Would it make sense to also add logic to constck? (check_const.rs)

Well I'm working on turning this check on unconditionally. I can just move the comptime check before the feature gate check

jdonszelmann · 2025-11-12T09:00:19Z

compiler/rustc_ast/src/ast.rs

 #[derive(HashStable_Generic, Walkable)]
 pub enum Const {
+    Comptime(Span),
    Yes(Span),


Would be nice to some day make all the variant names meaningful. Something like

enum EvaluationTime { Comptime, MaybeComptime, Runtime, }

Looking at your PR yday that's one of the first things I thought. First decided not to comment it because it might not be for this PR but coming back I thought I'd mention it anyway.

jdonszelmann · 2025-11-12T09:03:12Z

compiler/rustc_hir/src/attrs/encode_cross_crate.rs

            BodyStability { .. } => No,
            Coinductive(..) => No,
            Cold(..) => No,
+            Comptime(..) => Yes,


can be No if we do the query

jdonszelmann · 2025-11-12T09:06:00Z

compiler/rustc_hir_pretty/src/lib.rs

        match s {
            hir::Constness::NotConst => {}
            hir::Constness::Const => self.word_nbsp("const"),
+            hir::Constness::Comptime => self.word_nbsp("comptime"),


Comptime is an attribute for now right; #[rustc_comptime]? Pretty printing doesn't matter so much, but this makes it seem like it's already a keyword. You could make it an unstable keyword ofc. But otherwise, I think the attr is already printed itself in hir pretty printing, so you might not need to emit anything here

jdonszelmann · 2025-11-12T09:08:07Z

compiler/rustc_middle/src/ty/mod.rs

            self.def_kind(def_id),
            DefKind::Fn | DefKind::AssocFn | DefKind::Ctor(_, CtorKind::Fn) | DefKind::Closure
-        ) && self.constness(def_id) == hir::Constness::Const
+        ) && self.constness(def_id) != hir::Constness::NotConst


I know this enum is unlikely to change any time soon, but might be nicer to write as a match exhaustively naming the variants

…d in comptime fn

…rgetted at what it does

rustbot · 2025-11-21T11:27:05Z

Some changes occurred in src/tools/clippy

cc @rust-lang/clippy

Some changes occurred to constck

cc @fee1-dead

rustbot · 2025-11-21T11:27:08Z

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.

rust-log-analyzer · 2025-11-21T12:04:27Z

The job x86_64-gnu-gcc failed! Check out the build log: (web) (plain enhanced) (plain)

Click to see the possible cause of the failure (guessed by this bot)

test [crashes] tests/crashes/131787.rs ... ok
test [crashes] tests/crashes/131886.rs ... ok
test [crashes] tests/crashes/132126.rs ... ok
test [crashes] tests/crashes/132765.rs ... ok
2025-11-21T12:04:11.543685Z ERROR compiletest::runtest: fatal error, panic: "crashtest no longer crashes/triggers ICE, hooray! Please give it a meaningful name, add a doc-comment to the start of the test explaining why it exists and move it to tests/ui or wherever you see fit. Adding 'Fixes #<issueNr>' to your PR description ensures that the corresponding ticket is auto-closed upon merge. If you want to see verbose output, set `COMPILETEST_VERBOSE_CRASHES=1`."
test [crashes] tests/crashes/132960.rs ... FAILED
test [crashes] tests/crashes/133965.rs ... ignored, ignored if rustc wasn't built with debug assertions
test [crashes] tests/crashes/133613.rs ... ok
test [crashes] tests/crashes/134061.rs ... ignored, ignored if rustc wasn't built with debug assertions
test [crashes] tests/crashes/134479.rs ... ignored, gcc backend is marked as ignore
---

---- [crashes] tests/crashes/132960.rs stdout ----
------rustc stdout------------------------------

------rustc stderr------------------------------
warning: the feature `generic_const_exprs` is incomplete and may not be safe to use and/or cause compiler crashes
##[warning] --> /checkout/tests/crashes/132960.rs:3:46
  |
3 | #![feature(adt_const_params, const_ptr_read, generic_const_exprs, unsized_const_params)]
  |                                              ^^^^^^^^^^^^^^^^^^^
  |
  = note: see issue #76560 <https://github.com/rust-lang/rust/issues/76560> for more information
---

error[E0601]: `main` function not found in crate `132960`
##[error]  --> /checkout/tests/crashes/132960.rs:36:48
   |
36 | const FOOBAR: &str = concat_strs::<FOO, BAR>();
   |                                                ^ consider adding a `main` function to `/checkout/tests/crashes/132960.rs`

error[E0517]: attribute should be applied to a struct, enum, or union
##[error]  --> /checkout/tests/crashes/132960.rs:11:12
   |
11 |     #[repr(C)]
   |            ^
...
14 |     const fn concat_arr<const M: usize, const N: usize>(a: [u8; M], b: [u8; N]) -> [u8; M + N] {}
   |     --------------------------------------------------------------------------------------------- not a struct, enum, or union

error[E0517]: attribute should be applied to a struct, enum, or union
##[error]  --> /checkout/tests/crashes/132960.rs:12:12
   |
12 |     #[repr(C)]
   |            ^
13 |
14 |     const fn concat_arr<const M: usize, const N: usize>(a: [u8; M], b: [u8; N]) -> [u8; M + N] {}
   |     --------------------------------------------------------------------------------------------- not a struct, enum, or union

warning: the feature `const_ptr_read` has been stable since 1.71.0 and no longer requires an attribute to enable
##[warning] --> /checkout/tests/crashes/132960.rs:3:30
  |
3 | #![feature(adt_const_params, const_ptr_read, generic_const_exprs, unsized_const_params)]
  |                              ^^^^^^^^^^^^^^
  |
  = note: `#[warn(stable_features)]` on by default

error[E0308]: mismatched types
##[error]  --> /checkout/tests/crashes/132960.rs:14:84
   |
14 |     const fn concat_arr<const M: usize, const N: usize>(a: [u8; M], b: [u8; N]) -> [u8; M + N] {}
   |              ----------                                                            ^^^^^^^^^^^ expected `[u8; M + N]`, found `()`
   |              |
   |              implicitly returns `()` as its body has no tail or `return` expression
   |
note: consider returning one of these bindings
  --> /checkout/tests/crashes/132960.rs:14:57
   |
14 |     const fn concat_arr<const M: usize, const N: usize>(a: [u8; M], b: [u8; N]) -> [u8; M + N] {}
   |                                                         ^           ^

warning: type annotations needed
##[warning]  --> /checkout/tests/crashes/132960.rs:25:35
   |
25 |                 A.as_ptr().cast().read(),
   |                                   ^^^^
   |
   = warning: this is accepted in the current edition (Rust 2015) but is a hard error in Rust 2018!
   = note: for more information, see issue #46906 <https://github.com/rust-lang/rust/issues/46906>
   = note: `#[warn(tyvar_behind_raw_pointer)]` (part of `#[warn(rust_2018_compatibility)]`) on by default

warning: type annotations needed
##[warning]  --> /checkout/tests/crashes/132960.rs:26:35
   |
26 |                 B.as_ptr().cast().read(),
   |                                   ^^^^
   |
   = warning: this is accepted in the current edition (Rust 2015) but is a hard error in Rust 2018!
   = note: for more information, see issue #46906 <https://github.com/rust-lang/rust/issues/46906>

error[E0716]: temporary value dropped while borrowed
##[error]  --> /checkout/tests/crashes/132960.rs:24:44
   |
24 |                std::str::from_utf8_unchecked(&concat_arr(
   |   _____________-                              ^
   |  |____________________________________________|
25 | ||                 A.as_ptr().cast().read(),
26 | ||                 B.as_ptr().cast().read(),
27 | ||             ))
   | ||_____________^- using this value as a constant requires that borrow lasts for `'static`
   | |______________|
   |                creates a temporary value which is freed while still in use
28 |            };
   |            - temporary value is freed at the end of this statement

error: aborting due to 5 previous errors; 5 warnings emitted

Some errors have detailed explanations: E0308, E0517, E0601, E0716.
For more information about an error, try `rustc --explain E0308`.

------------------------------------------

error: crashtest no longer crashes/triggers ICE, hooray! Please give it a meaningful name, add a doc-comment to the start of the test explaining why it exists and move it to tests/ui or wherever you see fit. Adding 'Fixes #<issueNr>' to your PR description ensures that the corresponding ticket is auto-closed upon merge. If you want to see verbose output, set `COMPILETEST_VERBOSE_CRASHES=1`.

thread '[crashes] tests/crashes/132960.rs' panicked at src/tools/compiletest/src/runtest/crashes.rs:17:18:
fatal error
stack backtrace:
   5: __rustc::rust_begin_unwind

For more information how to resolve CI failures of this job, visit this link.

rustbot assigned JonathanBrouwer Nov 11, 2025

This comment has been minimized.

Sign in to view

oli-obk force-pushed the comptime branch from 9623203 to 953a740 Compare November 11, 2025 12:37

RalfJung reviewed Nov 11, 2025

View reviewed changes

fmease reviewed Nov 11, 2025

View reviewed changes

This comment has been minimized.

Sign in to view

fmease reviewed Nov 11, 2025

View reviewed changes

jdonszelmann self-assigned this Nov 11, 2025

fee1-dead self-assigned this Nov 11, 2025

fee1-dead reviewed Nov 11, 2025

View reviewed changes

oli-obk mentioned this pull request Nov 11, 2025

const_allocate intrinsic: what should it do at runtime? #93935

Open

jdonszelmann reviewed Nov 12, 2025

View reviewed changes

nikomatsakis mentioned this pull request Nov 12, 2025

reflection and comptime rust-lang/rust-project-goals#406

Open

7 tasks

JonathanBrouwer removed their assignment Nov 15, 2025

oli-obk and others added 4 commits November 20, 2025 12:05

Add Constness::Comptime

de709e3

Check comptime fns

938dee6

Prevent comptime fns from being called at runtime

5107f3e

Allow comptime fns to call other comptime fns

5255e50

oli-obk added 3 commits November 20, 2025 12:05

Demonstrate that const Trait bounds allow trait methods to be calle…

26518bc

…d in comptime fn

Make the inline field on the Const ConstContext more clearly ta…

5a09813

…rgetted at what it does

Rename Constness variants to match BoundConstness

82c5924

oli-obk force-pushed the comptime branch from 953a740 to 82c5924 Compare November 21, 2025 11:27

rustbot added the T-clippy Relevant to the Clippy team. label Nov 21, 2025

Add very basic "comptime" fn implementation #148820

Are you sure you want to change the base?

Add very basic "comptime" fn implementation #148820

Conversation

oli-obk commented Nov 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Nov 11, 2025

Uh oh!

rustbot commented Nov 11, 2025

Uh oh!

This comment has been minimized.

rustbot commented Nov 11, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

fmease Nov 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

This comment has been minimized.

fmease Nov 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

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!

jdonszelmann Nov 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

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!

rustbot commented Nov 21, 2025

Uh oh!

rustbot commented Nov 21, 2025

Uh oh!

rust-log-analyzer commented Nov 21, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants

oli-obk commented Nov 11, 2025 •

edited

Loading

fmease Nov 11, 2025 •

edited

Loading

fmease Nov 11, 2025 •

edited

Loading

jdonszelmann Nov 12, 2025 •

edited

Loading