[Agi-Asi]

Description

Segmented video reversal runs a multiprocessing.Pool while the parent process still holds an open PyAV input container (the with av.open(src) wrapping the whole body of reverse_video_file). On Linux the default fork start method duplicates the parent mid-flight: a forked child can inherit an FFmpeg/PyAV (or tqdm) lock in a locked state with no owner left to release it, and the pool then deadlocks on its own queue semaphore.

This matches the reported failures exactly:

• renders hanging in Reversing large file by cutting it in segments ([BUG] Rendering stalled in 'Reversing large file by cutting it in segment' #562)
• a worker stuck in multiprocessing/queues.py get() → semlock.__enter__, only surfaced after Ctrl-C as a KeyboardInterrupt stack trace ([BUG] Concurrency bug when reversing videos #569 — the trace in that issue is precisely a worker blocked on the inherited queue rlock)

Both are probabilistic — the fork has to race a held lock — which is why small projects never hit it and larger ones (more segments → more workers → more forks) hit it once every 10–20 renders, as reported.

Fix: create the pool from an explicit spawn context. Freshly spawned interpreters cannot inherit held locks, and this matches the default start method on macOS and Windows — Linux was the odd one out. reverse_video_file_in_one_chunk is a module-level function taking picklable Path arguments, so it is already spawn-compatible; a few interpreter startups are noise next to the video re-encoding the workers perform.

Testing

• New regression test drives the segmented path (segment duration forced below the test asset's length, so the pool actually runs) and asserts a decodable, non-empty output.
• pytest tests/test_utils.py — 3 passed.
• ruff check / ruff format --check clean on the touched files.

Fixes #562
Fixes #569

[codacy-production]

Not up to standards ⛔

🔴 Issues 2 high

Alerts:
⚠ 2 issues (≤ 0 issues of at least minor severity)

Results:
2 new issues

Category	Results
Security	2 high

View in Codacy

🟢 Metrics 0 duplication

Metric	Results
Duplication	0

View in Codacy

_{NEW Get contextual insights on your PRs based on Codacy's metrics, along with PR and Jira context, without leaving GitHub. Enable AI reviewer
TIP This summary will be updated as you push new changes.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 79.10%. Comparing base (1549ef4) to head (e2b5ea8).

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #672   +/-   ##
=======================================
  Coverage   79.10%   79.10%           
=======================================
  Files          23       23           
  Lines        2178     2178           
=======================================
  Hits         1723     1723           
  Misses        455      455           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[jeertmans]

Hi @Agi-Asi, thanks for your contribution! However, I believe this is a duplicated of #668, isn't it?

[Agi-Asi]

You're right — #668 covers the same spawn-context fix (I missed it when scanning open PRs; my search matched issue references but #668's body used 'Closes #562/#569' formatting my scan tokenized differently — my mistake). Closing this in favor of #668, which is more complete (also sorts segment paths). Sorry for the noise, and thanks for the quick triage!

[jeertmans]

No worries!