[kixelated]

Fixes three regressions in the Nightly job of https://github.com/moq-dev/moq/actions/runs/38143901976. PR CI misses all three: they only show up under --all-features or cargo bench -- --test.

• Features (compile): feat(video)!: accept external Vulkan encoder inputs #4975 changed DmaBuf::new to (fd, DmaBufLayout, owner), but the VA-API test fixture in moq-video/src/render/dmabuf.rs still used the old 7-argument call. The fixture had a private DmaBufFrame wrapper that duplicated what the public API now does (hold the FD, refuse downloads). I deleted it, and the fixture now calls DmaBuf::new(fd, layout, ()).
• Features (clippy): moq-cli/src/play/media.rs used let _ = sink.write(..)? on a Result<()>, which clippy::let_unit_value rejects under --all-features.
• moq-net benchmark smoke: feat(net)!: restart announce consumers on a replaced broadcast, keep subscriptions sticky #5087 made a winner switch between epochless routes a Restart, and keeps open subscriptions on the old instance.
  • origin/reprice_duplicate still expected an Update and panicked. It now expects Restart.
  • origin/handoff waited for an open subscription to switch over to a standby broadcast, so it hung forever. The nightly never got that far, because it panicked on reprice_duplicate first. The bench now measures what a subscriber actually does after a Restart: re-request the broadcast and read the standby's first group. It asserts that group is sequence 1, so replaying the incumbent's cached group 0 cannot pass.

Verification

• cargo clippy -p moq-video -p moq-cli --all-targets --all-features -- -D warnings passes.
• cargo bench -p moq-net --features fuzz --bench '*' -- --test: all 465 pass.
• just check: lint and compile passed. The test stage hit load-related timeouts in moq-cli export/import tests (load average 130). Rerun at normal load, cargo nextest run -p moq-cli --all-targets passes 150/150.

API / wire impact

None. Only a test fixture, a bench, and a clippy-only edit change.

The other nightly failures have their own PRs (#5335 macOS, #5338 drill-sensitivity). Audio quality is tracked in quest/m1/audio-jitter-target.

(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 2026-10-12T01:04:17.326677Z	ab4a49f	PR opened

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]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: fb9aa07b-0327-4ba1-9570-5f40d2e601f3

📥 Commits

Reviewing files that changed from the base of the PR and between b0be4d2 and a469bb9.

📒 Files selected for processing (1)

• rs/moq-net/benches/origin.rs

Included review availability: This review used your included allowance. Your plan provides up to 4 included reviews per hour; 0 remain after this review.

Walkthrough

Audio playback now returns errors from silence and audio-frame writes. Origin benchmarks update a route-change event expectation and measure handoff through the replacement broadcast’s first group. The DMA-BUF test fixture now constructs a DmaBuf directly from a layout and cloned descriptor.

Priority: ⬇️ Low

Merge Risk: ⚪ Minimal · up to a469b

The reviewed changes show no new failure mode in audio error handling, benchmark cleanup, or DMA-BUF fixture behavior. No actionable merge blocker is established.

Pre-merge checks · 5

Passed checks (5 passed)

Check name	Status	Explanation
Description check		The description clearly explains the three nightly regressions and the corresponding fixture, clippy, and benchmark fixes.
Title check		The title concisely and accurately summarizes the pull request's primary purpose: repairing nightly-only compile, clippy, and benchmark failures.
Docstring Coverage		Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 4 functions across 3 files.
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.

✨ Finishing Touches

✨ Simplify code

• Commit to this branch
• Create a new PR

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

Warning

Some tools did not complete. Review the errors below.

🔧 Clippy (1.98.1)

Clippy execution timed out

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]

Automated review by review (OpenAI)

Reviewed commit: b0be4d2.

No blocking correctness findings in the full three-file change. The direction is sound: the CLI edit preserves writes and error propagation; the origin benchmarks follow the current Restart/re-request contract; and the DMA-BUF fixture uses the existing FD-owning API instead of duplicating it. The main-branch merge preserves these fixes and supplies explicit timed tracks to the handoff benchmark.

Optional test hardening: rs/moq-net/benches/origin.rs:660–662 only checks that a group arrives after re-requesting. Assert sequence 1 (or payload "two") so accidentally rejoining the incumbent and replaying its cached group cannot satisfy "standby group". This is a coverage improvement, not evidence of a current routing failure.

Verification limits: GitHub-only static inspection of the complete patch and relevant current API/benchmark context; no builds, benchmark runs, or VA-API hardware tests executed. The description's test results are author-reported, and current-head PR workflows are still running. Ordinary PR CI does not establish the nightly all-features/benchmark smoke paths.

[kixelated]

Merge summary: CI was green and the OpenAI review of b0be4d2 had no blocking findings. I applied its optional suggestion in 50d3db1: the handoff bench now asserts the re-requested group is sequence 1, so rejoining the incumbent and replaying its cached group 0 cannot pass. The assert sits after the timer stops. cargo bench -p moq-net --bench origin -- --test handoff passes for 1p and 1000p. Enabling auto-merge on 50d3db1.

(Written by Claude Opus 5.5)

[kixelated]

Automated review by review (OpenAI)

Reviewed commit: 50d3db1.

No actionable bugs found in the delta from b0be4d2. The added assertion at rs/moq-net/benches/origin.rs:662–665 addresses the optional coverage gap in my previous review: the standby explicitly produces sequence 1, so replaying the incumbent's cached group 0 now fails. The assertion is outside the accumulated timing interval. This is an appropriately small fix; the overall nightly-repair direction remains sound.

Verification limits: GitHub-only static review of the delta, full PR patch, and surrounding benchmark/closure code. No builds, benchmarks, or hardware tests executed; this does not independently confirm the author-reported test results.