Repository navigation
fix(pty): keep PtyManager spawn off the async runtime worker (KI-16) - #141
Conversation
Test-First: src-tauri/src/pty.rs::blocking_spawn_work_leaves_the_runtime_worker_free Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> (cherry picked from commit fc8470c)
PtyManager::spawn_with_id ran openpty and process creation inline on the Tokio worker. It now hands the worker over with block_in_place on a multi-thread runtime; order, errors and signatures are unchanged. Regression-For: f58728f Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> (cherry picked from commit dd322cb)
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: b99574e7f5
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".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| use super::*; | ||
|
|
||
| #[test] | ||
| fn blocking_spawn_work_leaves_the_runtime_worker_free() { |
There was a problem hiding this comment.
When this exact candidate is checked against its sole parent with BASE_SHA=1947110 HEAD_SHA=95eb4f3 bash scripts/ci/red-first.sh --plan, the gate exits 1 because the source-changing commit has no recognized trailer. The regression test appears in the same commit as the fix, so this history contains no compiling, failing revision for KI-16 despite the report claiming one; place the test in a preceding red commit with Test-First: and reference it from the fix with Regression-For:.
AGENTS.md reference: AGENTS.md:L16-L18
Useful? React with 👍 / 👎.
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
…n-blocking-v2 # Conflicts: # KNOWN_ISSUES.md
|
@Mergifyio queue |
Merge Queue Status
This pull request spent 47 minutes 5 seconds in the queue, including 10 minutes 12 seconds running CI. Required conditions to merge
|
Der Merge-Konflikt mit
mainist gelöst und der PR steht wieder auf einem konfliktfreien Stand.Die Auflösung bewahrt die KI-16-Nachführung und übernimmt zugleich die aktuellen Einträge aus
main.Die vollständige lokale Prüfbahn ist grün; der PR kann wieder in die Mergify-Queue.
Report
What changed
origin/mainintoclaude/ki-16-pty-spawn-blocking-v2as7ce90da.KNOWN_ISSUES.mdconflict by retaining the KI-16 partial-fix entry and KI-19 entry from this branch while preserving the currentmaincontent.mainchanges insrc-tauri/src/api.rsandsrc-tauri/src/providers.rs.Evidence
git fetch origin main: exit 0.git merge origin/main: initially reported the expectedKNOWN_ISSUES.mdconflict; resolved and committed as7ce90da.gitleaks detect --no-banner --redact --source .: exit 0; no leaks found.git commit --no-edit: exit 0; precommit secrets, format, cargo-check and typecheck gates passed.bash scripts/ci/gates.sh lane prepush: exit 0; 392 frontend tests, 502 HQ tests, 1660 Rust tests passed, 7 skipped, and clippy passed.bash scripts/ci/gates.sh lane prepush: exit 0; same lane passed again before push.git push -u origin claude/ki-16-pty-spawn-blocking-v2: exit 0; remote updated to7ce90da.git ls-remote origin claude/ki-16-pty-spawn-blocking-v2: exit 0; remote head is7ce90da0e50fe29da96aa1278df218b17f9af432.NICHT ABGEDECKT
#[cfg(windows)]tests (run_in_pty, npm_shim, ConPTY, DPAPI) or the Windows-only clippy arms; the Windows lane runs in the merge queue.Review
Review: ausstehend (Stufe A, PTY).
Open points
spawn_ptyinmain.rsremains a separate follow-up as described in the original PR scope.