Repository navigation
deps: bump moq-noq to 1.3.3 for RUSTSEC-2026-0185 - #4702
Conversation
noq 1.3.3 bounds the stream/CRYPTO reassembly chunk count (RUSTSEC-2026-0185) and fixes compaction dropping unread data on unordered streams. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
MERGE — reviewed Security dep bump of the moq-noq stack on Advisory / version check
Findings: none. Public API unchanged; wire change is only closing peers that exceed the reassembly cap (intended). Finite receive-window default remains with #4605 as noted. CI: Check / Test / platform jobs still pending at review time — wait for green before merge. This is an automated review, not the maintainer's decision |
kixelated
left a comment
There was a problem hiding this comment.
Automated review by review (OpenAI)
Reviewed a09e1ae.
No actionable introduced bugs found. Cargo.toml:154–155,239 raises the three minimum versions coherently, and Cargo.lock:4538,4560,4591,9850 updates exactly the four fork packages. Independently confirming the existing review: all four 1.3.3 checksums match the published, non-yanked registry entries. Features are unchanged; their Rust 1.88 minimum fits the workspace's 1.91 requirement.
Direction is right: ship the released backport through both moq-tokio and moq-uring, without unrelated dependency churn. The upstream source includes the STREAM/CRYPTO cap, unordered-compaction fix, header-reset handling, and close-capsule lifetime fix. The raw-QUIC reset-code change from 2.0.1 is excluded from 1.3.3. This fulfills the fork pin step; the separate upstream noq-proto 1.3.0 used by iroh and the finite receive-window default remain outside this fix.
Verification: static manifest/lock, registry, consumer wiring, and upstream patch review. No local builds or tests run; the author's just check result was not independently reproduced. Exact-head CI was still running.
(Written by review (OpenAI))
|
Navigate logical layers of code changes, visualize relationships, and explore their blast radius. No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Advanced Run ID: ⛔ Files ignored due to path filters (1)
📒 Files selected for processing (1)
Included review availability: This review used your included allowance. Your plan provides up to 4 included reviews per hour; 1 remain after this review. WalkthroughThe workspace dependency versions for Priority: ⬆️ High Merge Risk: ⚪ Minimal · up to The dependency update includes the reported fixes, and no concrete compatibility issue was found in inspected callers. No actionable merge-blocking risk is established. Architecture SummaryArchitecture risk: 🔵 Low · up to The change affects 1 system. Changed systems: Architecture concerns Review detailsSystems and components
Before / after behavior
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches✨ Simplify code
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. Comment |
|
Merge summary: bumps the moq-noq stack (moq-noq, moq-noq-proto, moq-noq-udp, web-transport-moq) from 1.3.2 to 1.3.3 for RUSTSEC-2026-0185. Cargo.toml pins and Cargo.lock only; no public API change. All CI green on a09e1ae. Three automated reviews (Grok, OpenAI, CodeRabbit) found nothing actionable; maintainer approved. Out of scope follow-ups already tracked: the upstream noq-proto used by iroh, and the finite receive-window default (#4605). (Written by Claude Opus 5.5) |
Bumps the noq stack on
mainfrom 1.3.2 to 1.3.3 (moq-noq-proto,moq-noq-udp,web-transport-moqpins; only those four crates plusmoq-noqmove inCargo.lock).noq 1.3.3 (moq-dev/noq#26, #21) carries:
UnknownSession.This covers the
mainpin step ofquest/m0/noq-reassembly-cap.md; the finite receive-window default stays with #4605.Public API: none. Wire: none.
just checkpasses.🤖 Generated with Claude Code