fix handling of base address for TypeId allocations #144187

RalfJung · 2025-07-19T18:06:09Z

This fixes the problems discovered by @theemathas in #142789:

const-eval would sometimes consider TypeId pointers to be null
the type ID is different in Miri than in regular executions

Both boil down to the same issue: the TypeId "allocation" has a guaranteed 0 base address, but const-eval assumes it was non-zero (like normal allocations) and Miri randomized it (like normal allocations).

r? @oli-obk

rustbot · 2025-07-19T18:06:15Z

The Miri subtree was changed

cc @rust-lang/miri

Some changes occurred to the CTFE / Miri interpreter

cc @rust-lang/miri

Some changes occurred to the CTFE machinery

cc @RalfJung, @oli-obk, @lcnr

RalfJung · 2025-07-19T18:06:58Z

compiler/rustc_const_eval/src/const_eval/machine.rs

Nothing really changes here, but I looked at the code to ensure it is correct for TypeId and realized some ways in which it, and the comments, could be improved.

theemathas · 2025-07-20T12:07:38Z

Is it intentional that this PR does not fix the offset_from behavior, which can assert whether two TypeIds are the same?

oli-obk · 2025-07-20T12:16:04Z

Is it intentional that this PR does not fix the offset_from behavior, which can assert whether two TypeIds are the same?

Yes, catching UB is not a requirement for CTFE as it has significant impact on compile times. A program that does UB if two type ids do not match may suddenly stop erroring with the next compiler release, or error even if the type ids match.

RalfJung · 2025-07-20T20:13:44Z

Also, you can't use it to get a bool in the code that tells you the answer, you can only use it to abort compilation, so it's less of a hazard IMO.

RalfJung · 2025-07-20T20:14:52Z

#144169 landed so this PR is no longer blocked.

oli-obk · 2025-07-21T08:17:01Z

@bors r+ rollup

bors · 2025-07-21T08:17:04Z

📌 Commit 3f9be40 has been approved by oli-obk

It is now in the queue for this repository.

fix handling of base address for TypeId allocations This fixes the problems discovered by `@theemathas` in rust-lang#142789: - const-eval would sometimes consider TypeId pointers to be null - the type ID is different in Miri than in regular executions Both boil down to the same issue: the TypeId "allocation" has a guaranteed 0 base address, but const-eval assumes it was non-zero (like normal allocations) and Miri randomized it (like normal allocations). r? `@oli-obk`

fix handling of base address for TypeId allocations This fixes the problems discovered by ``@theemathas`` in rust-lang#142789: - const-eval would sometimes consider TypeId pointers to be null - the type ID is different in Miri than in regular executions Both boil down to the same issue: the TypeId "allocation" has a guaranteed 0 base address, but const-eval assumes it was non-zero (like normal allocations) and Miri randomized it (like normal allocations). r? ``@oli-obk``

fix handling of base address for TypeId allocations This fixes the problems discovered by ```@theemathas``` in rust-lang#142789: - const-eval would sometimes consider TypeId pointers to be null - the type ID is different in Miri than in regular executions Both boil down to the same issue: the TypeId "allocation" has a guaranteed 0 base address, but const-eval assumes it was non-zero (like normal allocations) and Miri randomized it (like normal allocations). r? ```@oli-obk```

Rollup of 15 pull requests Successful merges: - #142097 (gpu offload host code generation) - #143430 (Lower extra lifetimes before normal generic params.) - #143672 (Fix Box allocator drop elaboration) - #143768 (Constify Try, From, TryFrom and relevant traits) - #143816 (Implement `check` for compiletest and RA using tool macro) - #143985 (rustc_public: de-StableMIR-ize) - #144027 (clippy: make tests work in stage 1) - #144080 (Mitigate `#[align]` name resolution ambiguity regression with a rename) - #144176 (Add approval blocking labels for new bors) - #144187 (fix handling of base address for TypeId allocations) - #144212 (Remove the ptr_unique lang item) - #144243 (Subtree update of `rust-analyzer`) - #144246 (Don't use another main test file as auxiliary) - #144251 (rustc-dev-guide subtree update) - #144254 (opt-dist: make `artifact-dir` an absolute path for `opt-dist local`) r? `@ghost` `@rustbot` modify labels: rollup

Rollup of 14 pull requests Successful merges: - #142097 (gpu offload host code generation) - #143430 (Lower extra lifetimes before normal generic params.) - #143768 (Constify Try, From, TryFrom and relevant traits) - #143816 (Implement `check` for compiletest and RA using tool macro) - #143985 (rustc_public: de-StableMIR-ize) - #144027 (clippy: make tests work in stage 1) - #144080 (Mitigate `#[align]` name resolution ambiguity regression with a rename) - #144176 (Add approval blocking labels for new bors) - #144187 (fix handling of base address for TypeId allocations) - #144212 (Remove the ptr_unique lang item) - #144243 (Subtree update of `rust-analyzer`) - #144246 (Don't use another main test file as auxiliary) - #144251 (rustc-dev-guide subtree update) - #144254 (opt-dist: make `artifact-dir` an absolute path for `opt-dist local`) r? `@ghost` `@rustbot` modify labels: rollup

Rollup merge of #144187 - RalfJung:type-id-base-addr, r=oli-obk fix handling of base address for TypeId allocations This fixes the problems discovered by ````@theemathas```` in #142789: - const-eval would sometimes consider TypeId pointers to be null - the type ID is different in Miri than in regular executions Both boil down to the same issue: the TypeId "allocation" has a guaranteed 0 base address, but const-eval assumes it was non-zero (like normal allocations) and Miri randomized it (like normal allocations). r? ````@oli-obk````

Rollup of 14 pull requests Successful merges: - rust-lang/rust#142097 (gpu offload host code generation) - rust-lang/rust#143430 (Lower extra lifetimes before normal generic params.) - rust-lang/rust#143768 (Constify Try, From, TryFrom and relevant traits) - rust-lang/rust#143816 (Implement `check` for compiletest and RA using tool macro) - rust-lang/rust#143985 (rustc_public: de-StableMIR-ize) - rust-lang/rust#144027 (clippy: make tests work in stage 1) - rust-lang/rust#144080 (Mitigate `#[align]` name resolution ambiguity regression with a rename) - rust-lang/rust#144176 (Add approval blocking labels for new bors) - rust-lang/rust#144187 (fix handling of base address for TypeId allocations) - rust-lang/rust#144212 (Remove the ptr_unique lang item) - rust-lang/rust#144243 (Subtree update of `rust-analyzer`) - rust-lang/rust#144246 (Don't use another main test file as auxiliary) - rust-lang/rust#144251 (rustc-dev-guide subtree update) - rust-lang/rust#144254 (opt-dist: make `artifact-dir` an absolute path for `opt-dist local`) r? `@ghost` `@rustbot` modify labels: rollup

rustbot assigned oli-obk Jul 19, 2025

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Jul 19, 2025

RalfJung mentioned this pull request Jul 19, 2025

Stabilize const TypeId::of #144133

Merged

RalfJung commented Jul 19, 2025

View reviewed changes

This comment has been minimized.

Sign in to view

RalfJung force-pushed the type-id-base-addr branch from 31b35e0 to 2ad5678 Compare July 19, 2025 20:17

workingjubilee closed this Jul 20, 2025

rustbot removed the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jul 20, 2025

workingjubilee reopened this Jul 20, 2025

rustbot added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jul 20, 2025

theemathas mentioned this pull request Jul 20, 2025

Tracking issue for comparing raw pointers in constants #53020

Open

fix handling of base address for TypeId allocations

3f9be40

RalfJung force-pushed the type-id-base-addr branch from 2ad5678 to 3f9be40 Compare July 20, 2025 20:14

oli-obk approved these changes Jul 21, 2025

View reviewed changes

bors 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 Jul 21, 2025

jieyouxu mentioned this pull request Jul 21, 2025

Rollup of 10 pull requests #144245

Closed

jieyouxu mentioned this pull request Jul 21, 2025

Rollup of 14 pull requests #144261

Closed

jieyouxu mentioned this pull request Jul 21, 2025

Rollup of 15 pull requests #144262

Closed

jieyouxu mentioned this pull request Jul 21, 2025

Rollup of 14 pull requests #144269

Merged

bors merged commit 9e9399f into rust-lang:master Jul 21, 2025
11 checks passed

rustbot added this to the 1.90.0 milestone Jul 21, 2025

RalfJung deleted the type-id-base-addr branch July 23, 2025 06:43

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

fix handling of base address for TypeId allocations #144187

fix handling of base address for TypeId allocations #144187

Uh oh!

RalfJung commented Jul 19, 2025 •

edited

Loading

Uh oh!

rustbot commented Jul 19, 2025

Uh oh!

RalfJung Jul 19, 2025

Uh oh!

This comment has been minimized.

theemathas commented Jul 20, 2025

Uh oh!

oli-obk commented Jul 20, 2025

Uh oh!

RalfJung commented Jul 20, 2025

Uh oh!

RalfJung commented Jul 20, 2025

Uh oh!

oli-obk commented Jul 21, 2025

Uh oh!

bors commented Jul 21, 2025

Uh oh!

Uh oh!

Uh oh!

fix handling of base address for TypeId allocations #144187

fix handling of base address for TypeId allocations #144187

Uh oh!

Conversation

RalfJung commented Jul 19, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Jul 19, 2025

Uh oh!

RalfJung Jul 19, 2025

Choose a reason for hiding this comment

Uh oh!

This comment has been minimized.

theemathas commented Jul 20, 2025

Uh oh!

oli-obk commented Jul 20, 2025

Uh oh!

RalfJung commented Jul 20, 2025

Uh oh!

RalfJung commented Jul 20, 2025

Uh oh!

oli-obk commented Jul 21, 2025

Uh oh!

bors commented Jul 21, 2025

Uh oh!

Uh oh!

Uh oh!

RalfJung commented Jul 19, 2025 •

edited

Loading