cli(install): fix node-gyp bootstrap walkup causing bats parallel hang by jdx · Pull Request #220 · endevco/aube

jdx · 2026-04-22T23:16:23Z

Summary

Found the actual root cause of the post-#217 bats parallel deadlock: #217's workspace-root walk-up in install::run interacted catastrophically with the recursive aube install that node-gyp bootstrap spawns.

The deadlock

node_gyp_bootstrap::bootstrap_blocking puts its scratch project at $XDG_CACHE_HOME/aube/tools/node-gyp/v12/ and spawns a recursive aube install there.
Under bats, HOME=$TEST_TEMP_DIR and XDG_CACHE_HOME=$TEST_TEMP_DIR/.cache. If the test has a pnpm-workspace.yaml at $TEST_TEMP_DIR/ (very common — workspace tests, allowBuilds, lifecycle tests, etc.), then:
The recursive aube install starts at tool_dir, walks up via find_workspace_root (added in cli: anchor aube install at workspace root from member subdir #217), and finds the outer test's pnpm-workspace.yaml.
The recursive process tries to install the outer workspace and blocks on the outer project lock at $TEST_TEMP_DIR/node_modules/ — the lock the parent aube install is already holding.
Two-process deadlock. Both wedge forever.

Why it only reproduced in GHA

GHA runners hit the lifecycle-script path more densely under --jobs >= 2 because multiple parallel tests concurrently trigger postinstall hooks.
Dev machines usually have node-gyp already on PATH, so ensure() short-circuits and the recursive install never runs.

Fix

Write an empty aube-workspace.yaml in tool_dir alongside the synthetic package.json. find_workspace_root's first-ancestor check finds the marker at tool_dir and returns immediately — the walk no longer escapes. workspace_packages stays empty so has_workspace is false and the install runs as a single-package install.

Test plan

cargo clippy --all-targets -- -D warnings — clean
Manual reasoning: the marker kills the walk at its first ancestor, so the recursive install can no longer target the outer workspace, breaking the lock-cycle
CI green on shards that previously hung (this PR restores --jobs parallelism in ci(bats): restore parallel jobs after #220 fixed the bootstrap deadlock #219)

Follow-up

Close #219 (was a CI-only workaround); the parallel-hang it papered over is fixed here at the root.

🤖 Generated with Claude Code

Note

Low Risk
Low risk: changes only the node-gyp bootstrap scratch directory setup by adding a local workspace marker to avoid workspace root walk-up and related lock deadlocks during recursive installs.

Overview
Prevents the recursive aube install used for node-gyp bootstrapping from walking up into an outer workspace (e.g., test fixtures with pnpm-workspace.yaml) and deadlocking on the parent install’s project lock.

This does so by writing an empty aube-workspace.yaml into the node-gyp tool cache directory before spawning the recursive install, ensuring workspace-root discovery stops at tool_dir and runs as a single-package install.

^{Reviewed by Cursor Bugbot for commit 93b9e1f. Bugbot is set up for automated code reviews on this repo. Configure here.}

#217's workspace-root walk-up in install::run interacted catastrophically with the recursive `aube install` that node-gyp bootstrap spawns. The bootstrap puts its scratch project under `$XDG_CACHE_HOME/aube/tools/node-gyp/v12/`. Inside any HOME with a `pnpm-workspace.yaml` at the top — which is exactly the shape every bats test creates (HOME=$TEST_TEMP_DIR with the test workspace yaml at the root) — the recursive `aube install` walked out of `tool_dir`, discovered the outer workspace yaml, and tried to install the *outer* workspace while the outer process held its project lock. Two-process deadlock; both wedged forever. This was the actual cause of the post-#217 GHA bats hang on shards 1+2 ubuntu and both macOS shards (any test that triggered a node-gyp bootstrap inside a workspace got stuck). It only reproduced under `--jobs >= 2` because GHA runners hit the lifecycle-script path more densely under parallelism, and never reproduced locally because most dev machines have node-gyp on PATH (bootstrap short-circuits). Fix: write an empty `aube-workspace.yaml` in `tool_dir` alongside the synthetic `package.json`. `find_workspace_root`'s first-marker check now returns `tool_dir` immediately and the recursive install runs as a single-package install (`workspace_packages` is empty so `has_workspace` stays false). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

greptile-apps · 2026-04-22T23:18:28Z

Greptile Summary

This PR writes an empty aube-workspace.yaml sentinel into the node-gyp bootstrap tool_dir before spawning the recursive aube install. Because find_workspace_root checks start itself first via Path::ancestors(), the walk stops immediately at tool_dir instead of climbing into the outer test's temp workspace and deadlocking on its project lock. WorkspaceConfig::load already handles an empty yaml file by returning Self::default() (empty packages list), so the recursive install correctly runs as a single-package install with has_workspace = false.

Confidence Score: 5/5

Safe to merge — the single-line fix is minimal, well-reasoned, and confirmed correct against the actual WorkspaceConfig::load and find_workspace_root implementations.

The empty sentinel file is correctly handled by WorkspaceConfig::load (returns Self::default() for empty content), find_workspace_root stops at tool_dir on its first iteration since Path::ancestors() includes start, and has_workspace stays false because workspace_packages is empty. No P0/P1 findings.

No files require special attention.

Important Files Changed

Filename	Overview
crates/aube/src/commands/install/node_gyp_bootstrap.rs	Writes an empty `aube-workspace.yaml` sentinel in `tool_dir` before the recursive `aube install`, stopping `find_workspace_root`'s walk at the tool directory and preventing the two-process deadlock in parallel bats tests.

_{Reviews (1): Last reviewed commit: "cli(install): break workspace walk-up ou..." | Re-trigger Greptile}

#217 was squash-merged with a debug version of `mise-tasks/ci/bats/nonserial` (per-file `for` loop with 90s timeout, breaks on first failure, no junit reporting). The actual deadlock that the debug loop was trying to surface — node-gyp bootstrap recursing into the outer test workspace — was fixed in #220. Restore the original `bats --jobs N` parallel invocation with junit reporting now that the underlying deadlock is gone. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

…ck (#219) ## Summary #217 was squash-merged with a debug version of `mise-tasks/ci/bats/nonserial` — a per-file `for` loop with a 90s timeout, breaks-on-first-failure, no junit reporting. Main has been running on that debug loop ever since. #220 fixed the actual root cause of the bats parallel deadlock (node-gyp bootstrap recursing back into the outer test workspace via #217's walk-up). With that fix on main, the debug script is no longer needed. This PR restores the original `bats --jobs N` parallel invocation with junit reporting. ## Test plan - [x] CI green proves it: shards 1+2 ubuntu and both macOS shards previously hung indefinitely with `--jobs >= 2`. With #220 merged, the deadlock is gone, so this PR's parallel invocation should land bats wall-time back at ~2 min/shard. 🤖 Generated with [Claude Code](https://claude.com/claude-code)  --- > [!NOTE] > **Medium Risk** > Changes CI test execution from a per-file, timed loop back to a single parallel `bats` run, which may reintroduce hangs or alter failure/reporting behavior in CI. > > **Overview** > Replaces the temporary debug per-file/timeout loop in `mise-tasks/ci/bats/nonserial` with a single `bats` invocation over the shard’s file list. > > The new run restores JUnit output (`--report-formatter junit`, `--output`, `BATS_REPORT_FILENAME`) and uses `--jobs`/`--parallel-binary-name` for parallel execution while still filtering out `serial` tests. > > <sup>Reviewed by [Cursor Bugbot](https://cursor.com/bugbot) for commit cf4f390. Bugbot is set up for automated code reviews on this repo. Configure [here](https://www.cursor.com/dashboard/bugbot).</sup>  Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

jdx merged commit 55fc657 into main Apr 22, 2026
16 checks passed

jdx deleted the fix/node-gyp-bootstrap-walkup branch April 22, 2026 23:24

jdx mentioned this pull request Apr 22, 2026

ci(bats): restore parallel jobs after #220 fixed the bootstrap deadlock #219

Merged

1 task

greptile-apps Bot mentioned this pull request Apr 22, 2026

chore: release v1.0.0-beta.13 #221

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

cli(install): fix node-gyp bootstrap walkup causing bats parallel hang#220

cli(install): fix node-gyp bootstrap walkup causing bats parallel hang#220
jdx merged 1 commit intomainfrom
fix/node-gyp-bootstrap-walkup

jdx commented Apr 22, 2026 •

edited by cursor Bot

Loading

Uh oh!

greptile-apps Bot commented Apr 22, 2026

Important Files Changed

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Uh oh!

Conversation

jdx commented Apr 22, 2026 • edited by cursor Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

The deadlock

Why it only reproduced in GHA

Fix

Test plan

Follow-up

Uh oh!

greptile-apps Bot commented Apr 22, 2026

Greptile Summary

Confidence Score: 5/5

Important Files Changed

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

jdx commented Apr 22, 2026 •

edited by cursor Bot

Loading