Isolate Linux audio decode and output failures from video

[zortos293]

Stack

Depends on #959. This PR contains only audio failure isolation on top of the PLC change.

Changes

Keep malformed Opus and unrecoverable audio-output failures inside the audio worker rather than failing the shared video session. Attribute decode failures to Opus and output failures to the actual sink.

Eight consecutive decode rejections trigger a decoder rebuild, with at most two rebuilds per session. A continuously rejected stream disables audio after 24 rejected packets. An unrecoverable sink failure also disables audio without switching a fixed output route or stopping video.

Publish terminal audio state before queue closure and classify concurrent submissions as audio-disabled rather than fatal queue errors. Advance the concealment timeline after a rejected packet so the next packet cannot replay an already-concealed interval.

Verification

This republishes unchanged commit c339fac004a689e8e7f9f3dd8b4e6bb0a470ff42 from the malformed #957 after the user-authorized branch repair. The exact work is now preserved on a dedicated source branch; dev was restored to the original baseline.

The orchestrator independently reviewed and tested this combined implementation before repair:

• Real-libopus audio tests: 20 passed, 2 existing hardware-dependent tests ignored.
• Linux session tests, including real ALSA null and /dev/full failure fixtures: 15 passed.
• Core audio feedback tests passed.
• Full combined streamer workspace passed, including 110 Linux platform and 192 transport tests.
• Formatting and Clippy with warnings denied passed.

CI will run again on this replacement PR.

Limits

No live GFN session, GPU-backed embedded monitor, or physical PipeWire device-change test was available. Worker and core-feedback tests establish local isolation; live Qt session behavior remains unverified. Audio is unavailable for the rest of the session after the bounded recovery policy is exhausted.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Advanced

Run ID: c675c6e0-d625-4eab-8bad-cbfd5415dfa9

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

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.}

[capy-ai]

Capy found no issues.

View 2 other findings in Capy.

[capy-ai]

Updated the topic with a normal merge of dev and the held successful-audio-output recovery correction in 551972ae. Both monitor paths retain audio loss/recovery reporting and the already-landed color-format handling. Recovery is reported only after the replacement sink accepts output; terminal failure does not report recovery.

Independent verification on this source passed 707 bundled-FFmpeg native tests with 15 ignored, strict locked Clippy covering the bundled feature set, and formatting. The complete expected non-Sony integration tree also includes this correction and passed its covering suite. Required cross-platform PR CI is now running on the updated head. No force push or direct push to dev was used.