[kixelated]

Problem

just ts moq, which just ts bench runs first, panics in rustc while the copy analysis reads moq-net:
<dyn bytes::Buf as std::marker::Sized> has escaping bound vars. moq-net's ErasedSend::poll_write_buf(&mut self, ..., buf: &mut dyn Buf) is called through a trait object. BodyWrites::visit_operand took that callee's signature with skip_binder(), so the input &'^0 mut (dyn Buf + '^0) kept its late-bound regions. #519 then added an is_sized query to replace, and that query panics on escaping bound vars.

Bisected: b3c119556 (#519) panics, its parent 49b68c72b translates moq-net. #521 and #511 are not the cause.

Approach

• Instantiate the callee's bound regions as erased (instantiate_bound_regions_with_erased), as mir/call.rs and iter.rs already do, instead of skip_binder(). The types it records are region-erased already (storage), so only the trait query sees a difference.
• The other skip_binder() signatures in ts/src only walk types syntactically and never reach a trait query.
• The stock fixture's Leased trait gains skip(&mut self, other: &mut dyn Buf), called through the Box<dyn Leased> in buf::dyns. This is moq's shape, and it reproduces the panic on main. It runs differentially against native Rust.

Impact

Public API: none.

Alternatives

• Skip types with escaping bound vars in replace. Rejected: the analysis would then miss what such a callee replaces.

Follow-ups

• just ts moq runs only in nightly.yml, not in just ci or the merge queue. That is how feat(ts): translate the rest of a trait object in a struct's last field #519 landed. The stock fixture now covers this shape in just ci. Gate moq-net's translation (quest/m0/ts-moq-gate.md) records gating the whole translation, pending a CPU measurement.
• Finishes quest/m0/ts-moq-dyn-panic.md, which #529 recorded, and deletes it.

Checks

• just ts moq passes.
• just ts test --test fixtures stock fails with the panic before the fix and passes after it.
• just ci passes on the merged head.

Review

Codex (quest review), gpt-6.1-sol, medium effort, round 1 on head 5c90d3f: no findings.

🤖 Generated with Claude Code

[coderabbitai]

No actionable comments were generated in the recent review.

Walkthrough

Operand lowering now instantiates function signatures with erased bound regions before checking the receiver and processing arguments. The Buf fixture adds and exercises a skip operation, with corresponding expected TypeScript updates. The translation notes replace the panic document with a proposal to add moq-net translation to CI; this change does not add the proposed gate.

Merge Risk: Minimal · up to b9cfe

No merge-blocking issue is established. The signature fix and regression fixture appear consistent; moq-net CI coverage remains planned follow-up work.

Pre-merge checks: 5

Check name	Status	Explanation
Title check		The title clearly describes the primary fix: erasing late-bound regions before inspecting a called function's inputs.
Description check		The description directly explains the panic, root cause, implementation, fixture coverage, testing, and follow-up work.
Docstring Coverage		Docstring coverage is 80.00% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff with cyclomatic complexity of at least 2. Documentati…
Linked Issues check		Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check		Check skipped because no linked issues were found for this pull request.

Full details: Docstring Coverage

Explanation

Docstring coverage is 80.00% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff with cyclomatic complexity of at least 2. Documentation-only edits and deletions within surviving functions are included; unchanged diff context and fully deleted functions are excluded. Analyzed 5 functions across 6 files. (2 skipped: 2 unsupported.)

Finishing Touches

Simplify code

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

• X
• Mastodon
• Reddit
• LinkedIn

CodeRabbit Verify

• Run CodeRabbit Verify to test the changes in this pull request.

Recent review info

Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 2453b816-b061-422e-b441-a1c05f7cd510

Commits

Reviewing files that changed from the base of the PR and between aff50d0 and b9cfe51.

Files selected for processing: 9

• quest/m0/README.md
• quest/m0/ts-moq-dyn-panic.md
• quest/m0/ts-moq-gate.md
• ts/fixtures/stock/expected/buf.Leased.ts
• ts/fixtures/stock/expected/buf.Pieces.ts
• ts/fixtures/stock/expected/buf.ts
• ts/fixtures/stock/expected/lib.ts
• ts/fixtures/stock/src/buf.rs
• ts/src/lower/copy.rs

Files with no reviewable changes: 1

• quest/m0/ts-moq-dyn-panic.md

Included review availability: This review used your included allowance. Your plan provides up to 1 included review per hour; 0 remain after this review.

[pullfrog]

ℹ️ No critical issues — one doc nit inline.

Reviewed changes

• Region erasure at calls: BodyWrites::visit_operand (ts/src/lower/copy.rs:1159) instantiates the callee's bound regions with erased instead of skip_binder(), so a callee like moq-net's ErasedSend::poll_write_buf no longer hands &mut dyn Buf's late-bound regions to the is_sized query in replace. This matches the same shape in ts/src/lower/iter.rs:1658.
• Fixture coverage: the stock fixture's Leased trait gains skip(&mut self, other: &mut dyn Buf) called through the Box<dyn Leased> in dyns, with snapshot and differential regressions under ts/fixtures/stock/.
• Quest tree: ts-moq-dyn-panic.md is deleted as finished, and the new ts-moq-gate.md ([S]) records gating moq-net's translation in just ci behind a CPU measurement against the checks budget, linked from quest/m0/README.md.

I traced every consumer of the changed sig values: recorded types go through storage() (erase_and_anonymize_regions), replaces(), fills_fresh(), and pointees() match kinds syntactically, and keeps/crate_fn are untouched, so the erasure only removes the panic path and records the same types as before. The fixture's TS output mirrors the native semantics too: bytes impls Buf for &[u8], so &mut skipped coerces to &mut dyn Buf and the translation's rt.Place models the in-place advance.

(Written by GLM-5.3-Flash)

  ^{｜ Fix all ➔ ｜ Fix 👍s ➔ ｜ View workflow run ｜ Using glm-5.3-flash ｜ 𝕏}

[pullfrog]

it most naturally binds to other, which reads as "skip discards everything other still holds", but the body advances self by min(other.remaining(), self.remaining()). "then skips as many bytes as the two have left" would say that.

Suggested change

	/// Reads a byte of `other`, then skips as many bytes as it has left.
	/// Reads a byte of `other`, then skips as many bytes as the two have left.