Replace OffsetOf by an actual sum of calls to intrinsic.

[cjgillot]

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

[cjgillot]

@bors try @rust-timer queue

[rust-bors]

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

[rust-timer]

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]

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

[rustbot]

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]

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]

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-timer]

Finished benchmarking commit (f9e7961): comparison URL.

Overall result: ❌✅ regressions and improvements - please read the text below

Our benchmarks found a performance regression caused by this PR.
This might be an actual regression, but it can also be just noise.

Next Steps:

• If the regression was expected or you think it can be justified,
  please write a comment with sufficient written justification, and add
  @rustbot label: +perf-regression-triaged to it, to mark the regression as triaged.
• If you think that you know of a way to resolve the regression, try to create
  a new PR with a fix for the regression.
• If you do not understand the regression or you think that it is just noise,
  you can ask the @rust-lang/wg-compiler-performance working group for help (members of this group
  were already notified of this PR).

@rustbot label: +perf-regression
cc @rust-lang/wg-compiler-performance

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)	3.1%	[3.1%, 3.1%]	1
Regressions ❌ (secondary)	0.3%	[0.1%, 0.6%]	6
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.4%	[-0.9%, -0.0%]	7
All ❌✅ (primary)	3.1%	[3.1%, 3.1%]	1

Max RSS (memory usage)

Results (primary 2.1%, secondary -0.6%)

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

	mean	range	count
Regressions ❌ (primary)	2.1%	[2.0%, 2.2%]	2
Regressions ❌ (secondary)	3.4%	[2.4%, 4.4%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.6%	[-5.5%, -0.7%]	6
All ❌✅ (primary)	2.1%	[2.0%, 2.2%]	2

Cycles

Results (secondary -0.3%)

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)	4.0%	[2.3%, 6.6%]	5
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.5%	[-7.6%, -1.3%]	5
All ❌✅ (primary)	-	-	0

Binary size

Results (primary 1.1%)

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

	mean	range	count
Regressions ❌ (primary)	1.1%	[1.1%, 1.1%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.1%	[1.1%, 1.1%]	1

Bootstrap: 474.876s -> 471.624s (-0.68%)
Artifact size: 388.76 MiB -> 388.73 MiB (-0.01%)

[lqd]

clap_derive is noise, include-blob looks like noise.

[panstromek]

perf triage:

As indicated by the comment above, clap_derive is noise, include-blob is also noise (at least the opt variants, other ones didn't return to previous state, but they are tiny in absolute numbers).

@rustbot label: +perf-regression-triaged