fix(security): bound publish/play AMF waits with wall-clock deadline

[cursor]

Security review fix

Fixes #144

Finding (Medium)

Location: src/client/mod.rs

Impact: A malicious RTMP server that completes connect + createStream but withholds onStatus during publish() or play() could block the embedder's calling thread for up to ~650 seconds (64 recv polls × 10s RECV_POLL_TIMEOUT_MS, plus up to 10s on send).

Root cause

do_amf_connect() threads a wall-clock deadline through its AMF exchanges, but publish() and play() passed None, so wait_for_command() fell back to unbounded per-poll 10s waits.

Fix

• Add command_io_deadline() that reuses the configured connect_timeout (default 10s).
• Pass Some(deadline) from publish() and play() into send_command_msg() and wait_for_command().
• Add regression test publish_and_play_honor_command_io_deadline.

Remediation leverage

Single helper + two call-site changes; no API break. set_connect_timeout() now also governs publish/play blocking I/O, matching the existing comment that recv budgets apply during those phases.

  

---

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

---

Note

Medium Risk
Touches blocking client I/O and timeout behavior on publish/play; low API surface but security-relevant thread-blocking fix.

Overview
Fixes a denial-of-service path where a server that stalls after createStream could keep publish() or play() blocked for many minutes because those paths waited for onStatus with no overall deadline.

publish() and play() now compute a wall-clock deadline via new command_io_deadline(), reusing the existing connect_timeout / default 10s budget, and pass it into send_command_msg() and wait_for_command()—matching what do_amf_connect() already did. Docs on connect_timeout and set_connect_timeout() are updated to state that this budget applies to blocking I/O in connect(), publish(), and play().

A regression test asserts both methods return ErrorCode::Timeout near the configured deadline when the peer sends nothing.

^{Reviewed by Cursor Bugbot for commit 86c1b11. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_702b6dcf-faf4-4ec1-b722-1745f9a2a205)