Replace OffsetOf by an actual sum of calls to intrinsic. #148151

cjgillot · 2025-10-26T20:55:55Z

This PR changes the way we compute the value of the offset_of! macro in MIR. The current implementation uses a dedicated MIR rvalue.

This PR proposes to replace it by an inline constant which sums calls to a new intrinsic offset_of(variant index, field index). The desugaring is done at THIR building time, easier that doing it on MIR.

The new intrinsic is only meant to be used by const-eval. LLVM codegen will refuse to generate code for it.

We replace:

a = offset_of!(T, Variant1.Field1.Variant2.Field2);

By:

a = const {constant#n};

{constant#n}: usize = {
    _1 = offset_of::<T>(index of Variant1, index of Field1);
    _2 = offset_of::<U>(index of Variant2, index of Field2); // Where T::Variant1::Field1 has type U
    _0 = _1 + _2
}

The second commit modifies intrinsic const checking to take allow_internal_unstable into account. The new intrinsic should only be called from stable offset_of! macro. The intrinsic itself is unstable, const-unstable, but rustc_intrinsic_const_stable_indirect.

Fixes #123959
Fixes #125680
Fixes #129425
Fixes #136175

r? @ghost

library/core/src/intrinsics/mod.rs

cjgillot · 2025-10-27T00:53:27Z

@bors try @rust-timer queue

Replace OffsetOf by an actual sum of calls to intrinsic.

rust-bors · 2025-10-27T03:07:33Z

☀️ Try build successful (CI)
Build commit: 7721497 (772149783c4c312fd88b7dfaa68a045b14db3280, parent: f37aa9955f03bb1bc6fe08670cb1ecae534b5815)

rust-timer · 2025-10-27T05:51:54Z

Finished benchmarking commit (7721497): comparison URL.

Overall result: ❌✅ regressions and 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)	0.1%	[0.0%, 0.1%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.1%	[-0.1%, -0.1%]	3
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (secondary -1.9%)

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)	2.6%	[2.1%, 3.2%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-3.7%	[-4.9%, -1.5%]	5
All ❌✅ (primary)	-	-	0

Cycles

Results (secondary -0.8%)

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)	2.7%	[1.6%, 5.0%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-6.1%	[-7.9%, -4.2%]	2
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 473.876s -> 473.729s (-0.03%)
Artifact size: 390.50 MiB -> 390.46 MiB (-0.01%)

bors · 2025-10-31T16:23:55Z

☔ The latest upstream changes (presumably #148324) made this pull request unmergeable. Please resolve the merge conflicts.

rustbot · 2025-11-01T23:47:25Z

Some changes occurred in match checking

cc @Nadrieril

Some changes occurred to the CTFE machinery

cc @RalfJung, @oli-obk, @lcnr

⚠️ #[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

Some changes occurred to constck

cc @fee1-dead

Some changes occurred in src/tools/clippy

cc @rust-lang/clippy

This PR changes rustc_public

cc @oli-obk, @celinval, @ouz-a

Some changes occurred in rustc_ty_utils::consts.rs

cc @BoxyUwU

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

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

This PR changes MIR

cc @oli-obk, @RalfJung, @JakobDegen, @vakaras

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

Some changes occurred in compiler/rustc_codegen_cranelift

cc @bjorn3

Some changes occurred to the CTFE / Miri interpreter

cc @rust-lang/miri

Some changes occurred in compiler/rustc_codegen_ssa

cc @WaffleLapkin

This PR changes a file inside tests/crashes. If a crash was fixed, please move into the corresponding ui subdir and add 'Fixes #' to the PR description to autoclose the issue upon merge.

BoxyUwU · 2025-11-02T20:43:45Z

What's the justification/benefit for doing this, and what user facing impact does it have (i.e. why does it need a lang nomination)?

cjgillot · 2025-11-02T20:58:55Z

My justification is simplifying MIR. This feature was implemented as a specific MIR statement, but does not need to, an intrinsic is sufficient.

There should be no user-facing change because of the offset_of manipulation. Maybe some diagnostics, but not more.

However, there are 2 user-facing changes in this PR that t-lang may want to know:

const-stability of intrinsics now takes allow_internal_unstable into account;
we get an additional intrinsic with rustc_intrinsic_const_stable_indirect.

rust-rfcbot · 2025-11-05T16:49:50Z

🔔 This is now entering its final comment period, as per the review above. 🔔

RalfJung

rustc_const_eval changes LGTM.

View changes since this review

RalfJung · 2025-11-06T22:24:31Z

compiler/rustc_const_eval/src/interpret/intrinsics.rs

+                let tp_ty = instance.args.type_at(0);
+
+                let u32_layout = self.layout_of(self.tcx.types.u32)?;
+                let variant = self.read_scalar(&args[0])?.to_bits(u32_layout.size)? as u32;


Suggested change

let variant = self.read_scalar(&args[0])?.to_bits(u32_layout.size)? as u32;

let variant = self.read_scalar(&args[0])?.to_u32()?;

RalfJung · 2025-11-06T22:24:54Z

compiler/rustc_const_eval/src/interpret/intrinsics.rs

+
+                let u32_layout = self.layout_of(self.tcx.types.u32)?;
+                let variant = self.read_scalar(&args[0])?.to_bits(u32_layout.size)? as u32;
+                let field = self.read_scalar(&args[1])?.to_bits(u32_layout.size)? as usize;


Suggested change

let field = self.read_scalar(&args[1])?.to_bits(u32_layout.size)? as usize;

let field = self.read_scalar(&args[1])?.to_u32()? as usize;

RalfJung · 2025-11-06T22:27:55Z

library/core/src/intrinsics/mod.rs

+/// Therefore, implementations must not require the user to uphold
+/// any safety invariants.
+///
+/// The stabilized version of this intrinsic is [`core::mem::offset_of`].


This can also only be called at compile time, right? Please add that note to the comment.

Also, would be good to explain why this is a lang item, since that is very unusual for intrinsics.

bors · 2025-11-09T13:06:58Z

☔ The latest upstream changes (presumably #148721) made this pull request unmergeable. Please resolve the merge conflicts.

bors · 2025-11-11T20:48:14Z

☔ The latest upstream changes (presumably #148658) made this pull request unmergeable. Please resolve the merge conflicts.

rust-rfcbot · 2025-11-15T16:54:30Z

The final comment period, with a disposition to merge, as per the review above, is now complete.

As the automated representative of the governance process, I would like to thank the author for their work and everyone else who contributed.

This will be merged soon.

library/core/src/intrinsics/mod.rs

scottmcm · 2025-11-15T20:29:24Z

tests/mir-opt/building/offset_of.rs

+// CHECK: concrete::[[const_z0]]: usize
+// CHECK: [[z:_.*]] = offset_of::<Alpha>(const 0_u32, const 2_u32)
+// CHECK: [[z0:_.*]] = offset_of::<Beta>(const 0_u32, const 0_u32)
+// CHECK: [[sum:_.*]] = AddWithOverflow(copy [[z]], copy [[z0]]);


pondering: Can this ever overflow? Wouldn't the type have to be malformed for it to happen? Since we're only ever running it in CTFE, could it be AddUnchecked instead, and if it ever does overflow it'll get trapped by the UB-in-CTFE checks?

(Aside: appreciate the test here -- I came looking for it as soon as I was reading the code in mir_build.)

Actually, I'm kinda confused how it's AddWithOverflow here -- at https://github.com/rust-lang/rust/pull/148151/files#diff-ad5988b23007e07f23d5459907b7aa75c9dc6ae82623be8cd4bc6acc58cb3397R863 it's BinOp::Add, not BinOp::AddWithOverflow.

Is there a TerminatorKind::Assert in here too somehow?

It cannot possibly overflow, but as I made the transform happen on THIR, removing the overflow checks would only make stuff more complex.

I added a THIR-print test to show what the generated THIR looks like.

compiler/rustc_mir_build/src/thir/cx/expr.rs

scottmcm · 2025-11-15T20:41:55Z

library/core/src/intrinsics/mod.rs

+#[rustc_intrinsic_const_stable_indirect]
+#[rustc_intrinsic]
+#[lang = "offset_of"]
+pub const fn offset_of<T: PointeeSized>(variant: u32, field: u32) -> usize;


Musing: Is there ever a need to pass values into this? Why is it phrased this way as opposed to, say

Suggested change

pub const fn offset_of<T: PointeeSized>(variant: u32, field: u32) -> usize;

pub const fn offset_of<T: PointeeSized, const Variant: u32, const Field: u32>() -> usize;

(Not that the way it is is bad -- I think it's fine -- just came to mind as I was reading through stuff.)

scottmcm

r=me with the trivial comment fix. I left a bunch of other things on the way by that you can think about as well, but I don't think any of them are critical.

View changes since this review

bors · 2025-11-16T21:09:06Z

☔ The latest upstream changes (presumably #148957) made this pull request unmergeable. Please resolve the merge conflicts.

rustbot · 2025-11-16T22:00:39Z

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.

cjgillot force-pushed the no-offset-of branch from 0dc894f to e46804b Compare October 26, 2025 20:56

This comment has been minimized.

Sign in to view

cjgillot commented Oct 26, 2025

View reviewed changes

library/core/src/intrinsics/mod.rs Outdated Show resolved Hide resolved

cjgillot force-pushed the no-offset-of branch from e46804b to 8f24a32 Compare October 27, 2025 00:50

This comment has been minimized.

Sign in to view

rust-bors bot added a commit that referenced this pull request Oct 27, 2025

Auto merge of #148151 - cjgillot:no-offset-of, r=<try>

7721497

Replace OffsetOf by an actual sum of calls to intrinsic.

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Oct 27, 2025

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 Oct 27, 2025

cjgillot force-pushed the no-offset-of branch from 875c6f6 to 9fd16d6 Compare November 1, 2025 23:45

cjgillot marked this pull request as ready for review November 1, 2025 23:47

rustbot added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Nov 1, 2025

rustbot removed the S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. label Nov 1, 2025

cjgillot added the I-lang-nominated Nominated for discussion during a lang team meeting. label Nov 2, 2025

oli-obk self-assigned this Nov 2, 2025

This comment was marked as resolved.

Sign in to view

traviscross added the T-lang Relevant to the language team label Nov 5, 2025

RalfJung reviewed Nov 6, 2025

View reviewed changes

cjgillot force-pushed the no-offset-of branch from 4699a81 to d65ccdd Compare November 9, 2025 19:59

This comment has been minimized.

Sign in to view

cjgillot mentioned this pull request Nov 9, 2025

Replace Rvalue::NullaryOp by a variant in mir::ConstValue. #148766

Open

rust-rfcbot added finished-final-comment-period The final comment period is finished for this PR / Issue. and removed final-comment-period In the final comment period and will be merged soon unless new substantive objections are raised. labels Nov 15, 2025

rust-rfcbot added the to-announce Announce this issue on triage meeting label Nov 15, 2025

rustbot mentioned this pull request Nov 15, 2025

Tracking issue for release notes of #148151: Replace OffsetOf by an actual sum of calls to intrinsic. #148978

Open

scottmcm reviewed Nov 15, 2025

View reviewed changes

library/core/src/intrinsics/mod.rs Outdated Show resolved Hide resolved

scottmcm reviewed Nov 15, 2025

View reviewed changes

compiler/rustc_mir_build/src/thir/cx/expr.rs Show resolved Hide resolved

scottmcm reviewed Nov 15, 2025

View reviewed changes

scottmcm approved these changes Nov 15, 2025

View reviewed changes

cjgillot force-pushed the no-offset-of branch from d65ccdd to d78c14e Compare November 16, 2025 02:09

This comment has been minimized.

Sign in to view

cjgillot force-pushed the no-offset-of branch from d78c14e to fc6c34f Compare November 16, 2025 17:27

cjgillot force-pushed the no-offset-of branch from fc6c34f to 5b48fd9 Compare November 16, 2025 22:00

cjgillot added 5 commits November 16, 2025 22:03

Replace OffsetOf by an actual sum.

4ad1003

Honor allow_internal_unstable for const intrinsics.

b664ef0

Simplify implementation.

92e74f2

Complete doc.

0c82a1b

Add THIR building test.

85eca56

cjgillot force-pushed the no-offset-of branch from 5b48fd9 to 85eca56 Compare November 16, 2025 22:03

	let variant = self.read_scalar(&args[0])?.to_bits(u32_layout.size)? as u32;
	let variant = self.read_scalar(&args[0])?.to_u32()?;

	let field = self.read_scalar(&args[1])?.to_bits(u32_layout.size)? as usize;
	let field = self.read_scalar(&args[1])?.to_u32()? as usize;

	pub const fn offset_of<T: PointeeSized>(variant: u32, field: u32) -> usize;
	pub const fn offset_of<T: PointeeSized, const Variant: u32, const Field: u32>() -> usize;

Replace OffsetOf by an actual sum of calls to intrinsic. #148151

Are you sure you want to change the base?

Replace OffsetOf by an actual sum of calls to intrinsic. #148151

Uh oh!

Conversation

cjgillot commented Oct 26, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

This comment has been minimized.

Uh oh!

cjgillot commented Oct 27, 2025

Uh oh!

This comment has been minimized.

This comment has been minimized.

rust-bors bot commented Oct 27, 2025

Uh oh!

This comment has been minimized.

rust-timer commented Oct 27, 2025

Overall result: ❌✅ regressions and improvements - no action needed

Uh oh!

bors commented Oct 31, 2025

Uh oh!

rustbot commented Nov 1, 2025

Uh oh!

BoxyUwU commented Nov 2, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cjgillot commented Nov 2, 2025

Uh oh!

This comment was marked as resolved.

This comment was marked as resolved.

rust-rfcbot commented Nov 5, 2025

Uh oh!

RalfJung 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!

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!

bors commented Nov 9, 2025

Uh oh!

This comment has been minimized.

bors commented Nov 11, 2025

Uh oh!

rust-rfcbot commented Nov 15, 2025

Uh oh!

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!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

scottmcm 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!

This comment has been minimized.

This comment has been minimized.

bors commented Nov 16, 2025

Uh oh!

rustbot commented Nov 16, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

13 participants

cjgillot commented Oct 26, 2025 •

edited

Loading

BoxyUwU commented Nov 2, 2025 •

edited

Loading

RalfJung left a comment •

edited by rustbot

Loading

scottmcm left a comment •

edited by rustbot

Loading