[kixelated]

Problem

The nightly Tests (rs windows-test) job has failed every night since it was added: noq::tests::pinned_ipv6_connection (and usually ..._with_host_override) panic with CreateEndpoint(Os { code: 10022 }) (WSAEINVAL). Any real server or client binding a specific IPv6 address on Windows hits the same error.

moq_sock::bind::udp cleared IPV6_V6ONLY on every IPv6 socket, including [::1]. Linux forces v6-only when a socket binds a specific IPv6 address, so this was harmless there. Windows keeps the cleared flag, so noq-udp's UdpSocketState::new reads the socket as dual-stack and sets IPv4-level options on it, which Windows rejects with WSAEINVAL. Upstream quinn-udp has the same logic. The bug is ours: the socket claims an IPv4 side it can't have.

Approach

Clear IPV6_V6ONLY only for [::] and IPv4-mapped binds. Set it for every other IPv6 address, which matches what Linux already does. This also covers bind::tcp, which shares the helper. A new moq-sock test checks that a [::1] bind comes back v6-only. Only Windows can regress it, and it runs there in the nightly windows-test.

The other Windows failure (a_refused_upgrade_keeps_websocket, one night) was "could not bind a matching TCP and UDP port after 20 attempts". That is a separate port-collision flake.

Verification

Windows was verified through a temporary commit, now dropped. It added a step to the windows job in platform.yml that ran just rs test -p moq-sock -p moq-tokio, plus a diagnostic that ran noq's socket setup on a [::1] socket with IPV6_V6ONLY both off and on.

Impact

• No public API or wire change. Behavior change: a socket bound to a specific IPv6 address is now v6-only on every platform, as it already was on Linux.

Alternatives

• Have noq-udp tolerate WSAEINVAL on the IPv4 options. Rejected because it hides the mismatch rather than fixing the socket we hand it.

Follow-ups

• connection::tests::quic_winning_opens_one_session fails locally under heavy load, on main too (QUIC loses the race to WebSocket). It is unrelated to this change.

(Written by Opus 5.5)

🤖 Generated with Claude Code

---

^{Need help on this PR? Tag @codesmith-bot with what you need. Autofix is disabled.}

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-12T01:05:38.330961Z	a0e8577	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[coderabbitai]

Warning

Review limit reached

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Next included review available in 58 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used all 4 included reviews currently available.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 162150ab-014e-43ce-813f-94ac5d4ff9a1

---

📥 Commits

Reviewing files that changed from the base of the PR and between 5897d91 and a0e8577.

---

📒 Files selected for processing (1)

• rs/moq-sock/src/bind.rs

---

• 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

---

_{Comment @coderabbitai help to get the list of available commands.}

[kixelated]

Merge summary: CI green and Codex reviewed the head (a0e8577) with no findings. No changes made at merge time; enabling auto-merge on the current head.

(Written by Claude Opus 5.5)

[kixelated]

Dequeued by the merge queue: cargo fmt --all --check fails on test/interop/clients/rust/tail.rs, which is broken on main since #5377 and unrelated to this change. Open fixes are #5381 (adds a rustfmt config under test/interop) and #5383 (reformats the file). Re-enable auto-merge once one of them lands.

(Written by Claude Opus 5.5)