module: cache nearest parent package.json per directory by codebytere · Pull Request #65326 · nodejs/node

codebytere · 2026-08-16T15:51:12Z

Cold require() of a 1000-module CommonJS tree gets ~6 % faster (100 runs, ***) and module/module-require.js type='.js' +13.7 %, by looking up the nearest parent package.json once per directory instead of once per file.

module/module-require.js n=10000 type='.js' *** 13.69 % ±2.36%
module/module-require.js n=10000 type='.json' | 'dir' (don't use this path) -2.1 % (relink noise is ±1 %)
cold require() of a generated 1000-module tree, flat layout (100 runs) *** 6.37 % ±1.76%
same, pnpm-style symlinked node_modules *** 6.56 % ±0.90%
same, with NODE_COMPILE_CACHE warm *** 5.10 % ±1.43%

getNearestParentPackageJSON(checkPath) - called for every CommonJS module to find "type", and by findPackageJSON - memoized per file, so N modules in a directory meant N calls into the native package.json reader. The result only depends on the directory the traversal starts from, so the JS-side cache is now keyed by that directory, derived with the same rule the native side uses (trailing separator → that directory, else dirname). With the permission model enabled the per-path behavior is kept, because the native call performs a per-path read check. Same return values, same objects, same lifetime (both caches were already permanent). 1000 native calls → 236 for the tree above.

Tests: behavior is unchanged; test-module-*, test-require-*, test-esm-*, es-module and test-permission-* (which exercises the per-path branch) pass. Happy to add a call-count test if wanted.

Disclosure: the code, test, measurements and this description were written by Claude Code, directed and reviewed by @codebytere.

nodejs-github-bot · 2026-08-16T15:51:17Z

Review requested:

@nodejs/loaders

codecov · 2026-08-16T17:41:36Z

Codecov Report

❌ Patch coverage is 97.95918% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 90.13%. Comparing base (30bff4a) to head (feee201).
⚠️ Report is 43 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/internal/modules/package_json_reader.js	97.95%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #65326 +/- ##
=======================================
Coverage 90.13% 90.13% =======================================
Files 752 752 Lines 251568 251639 +71 Branches 47270 47285 +15 =======================================
+ Hits 226759 226827 +68 - Misses 16168 16174 +6 + Partials 8641 8638 -3

Files with missing lines	Coverage Δ
lib/internal/modules/package_json_reader.js	`99.23% <97.95%> (-0.22%)`	⬇️

... and 38 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

nodejs-github-bot · 2026-08-16T18:43:54Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75894/

ljharb · 2026-08-17T06:22:23Z

This seems good, but surely there should be some tests that assert the caching behavior, to avoid regressions?

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com>

codebytere · 2026-08-17T17:36:20Z

@ljharb done

nodejs-github-bot · 2026-08-18T16:00:19Z

Commit Queue failed

- Loading data for nodejs/node/pull/65326
✔ Done loading data for nodejs/node/pull/65326
----------------------------------- PR info ------------------------------------
Title module: cache nearest parent package.json per directory (#65326)
Author Shelley Vohr <shelley.vohr@gmail.com> (@codebytere)
Branch codebytere:perf/module-parent-package-json-dir-cache -> nodejs:main
Labels module, commit-queue
Commits 1
- module: cache nearest parent package.json per directory
Committers 1
- Shelley Vohr <shelley.vohr@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
------------------------------ Generated metadata ------------------------------
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
--------------------------------------------------------------------------------
ℹ This PR was created on Sun, 16 Aug 2026 15:51:12 GMT
✔ Approvals: 3
✔ - Yagiz Nizipli (@anonrig) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4947040173
✔ - James M Snell (@jasnell) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4948264808
✔ - Jordan Harband (@ljharb): https://github.com/nodejs/node/pull/65326#pullrequestreview-4953177652
✔ Last GitHub CI successful
ℹ Last Full PR CI on 2026-08-16T18:43:54Z: https://ci.nodejs.org/job/node-test-pull-request/75894/
⚠ Commits were pushed after the last Full PR CI run:
⚠ - module: cache nearest parent package.json per directory
- Querying data for job/node-test-pull-request/75894/
✔ Build data downloaded
- Querying failures of job/node-test-commit/90623/
✔ Data downloaded
✘ 2 failure(s) on the last Jenkins CI run
--------------------------------------------------------------------------------
✔ Aborted `git node land` session in /home/runner/work/node/node/.ncu

https://github.com/nodejs/node/actions/runs/32157693157

nodejs-github-bot · 2026-08-18T19:32:28Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75973/

nodejs-github-bot · 2026-08-19T08:52:08Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75982/

nodejs-github-bot · 2026-08-19T13:18:13Z

Landed in 9c641c5

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com> PR-URL: #65326 Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Jordan Harband <ljharb@gmail.com>

codebytere · 2026-08-16T15:51:12Z

Cold require() of a 1000-module CommonJS tree gets ~6 % faster (100 runs, ***) and module/module-require.js type='.js' +13.7 %, by looking up the nearest parent package.json once per directory instead of once per file.

module/module-require.js n=10000 type='.js' *** 13.69 % ±2.36%
module/module-require.js n=10000 type='.json' | 'dir' (don't use this path) -2.1 % (relink noise is ±1 %)
cold require() of a generated 1000-module tree, flat layout (100 runs) *** 6.37 % ±1.76%
same, pnpm-style symlinked node_modules *** 6.56 % ±0.90%
same, with NODE_COMPILE_CACHE warm *** 5.10 % ±1.43%

getNearestParentPackageJSON(checkPath) - called for every CommonJS module to find "type", and by findPackageJSON - memoized per file, so N modules in a directory meant N calls into the native package.json reader. The result only depends on the directory the traversal starts from, so the JS-side cache is now keyed by that directory, derived with the same rule the native side uses (trailing separator → that directory, else dirname). With the permission model enabled the per-path behavior is kept, because the native call performs a per-path read check. Same return values, same objects, same lifetime (both caches were already permanent). 1000 native calls → 236 for the tree above.

Tests: behavior is unchanged; test-module-*, test-require-*, test-esm-*, es-module and test-permission-* (which exercises the per-path branch) pass. Happy to add a call-count test if wanted.

Disclosure: the code, test, measurements and this description were written by Claude Code, directed and reviewed by @codebytere.

nodejs-github-bot · 2026-08-16T15:51:17Z

Review requested:

@nodejs/loaders

codecov · 2026-08-16T17:41:36Z

Codecov Report

❌ Patch coverage is 97.95918% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 90.13%. Comparing base (30bff4a) to head (feee201).
⚠️ Report is 43 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/internal/modules/package_json_reader.js	97.95%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #65326 +/- ##
=======================================
Coverage 90.13% 90.13% =======================================
Files 752 752 Lines 251568 251639 +71 Branches 47270 47285 +15 =======================================
+ Hits 226759 226827 +68 - Misses 16168 16174 +6 + Partials 8641 8638 -3

Files with missing lines	Coverage Δ
lib/internal/modules/package_json_reader.js	`99.23% <97.95%> (-0.22%)`	⬇️

... and 38 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

nodejs-github-bot · 2026-08-16T18:43:54Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75894/

ljharb · 2026-08-17T06:22:23Z

This seems good, but surely there should be some tests that assert the caching behavior, to avoid regressions?

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com>

codebytere · 2026-08-17T17:36:20Z

@ljharb done

nodejs-github-bot · 2026-08-18T16:00:19Z

Commit Queue failed

- Loading data for nodejs/node/pull/65326
✔ Done loading data for nodejs/node/pull/65326
----------------------------------- PR info ------------------------------------
Title module: cache nearest parent package.json per directory (#65326)
Author Shelley Vohr <shelley.vohr@gmail.com> (@codebytere)
Branch codebytere:perf/module-parent-package-json-dir-cache -> nodejs:main
Labels module, commit-queue
Commits 1
- module: cache nearest parent package.json per directory
Committers 1
- Shelley Vohr <shelley.vohr@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
------------------------------ Generated metadata ------------------------------
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
--------------------------------------------------------------------------------
ℹ This PR was created on Sun, 16 Aug 2026 15:51:12 GMT
✔ Approvals: 3
✔ - Yagiz Nizipli (@anonrig) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4947040173
✔ - James M Snell (@jasnell) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4948264808
✔ - Jordan Harband (@ljharb): https://github.com/nodejs/node/pull/65326#pullrequestreview-4953177652
✔ Last GitHub CI successful
ℹ Last Full PR CI on 2026-08-16T18:43:54Z: https://ci.nodejs.org/job/node-test-pull-request/75894/
⚠ Commits were pushed after the last Full PR CI run:
⚠ - module: cache nearest parent package.json per directory
- Querying data for job/node-test-pull-request/75894/
✔ Build data downloaded
- Querying failures of job/node-test-commit/90623/
✔ Data downloaded
✘ 2 failure(s) on the last Jenkins CI run
--------------------------------------------------------------------------------
✔ Aborted `git node land` session in /home/runner/work/node/node/.ncu

https://github.com/nodejs/node/actions/runs/32157693157

nodejs-github-bot · 2026-08-18T19:32:28Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75973/

nodejs-github-bot · 2026-08-19T08:52:08Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75982/

nodejs-github-bot · 2026-08-19T13:18:13Z

Landed in 9c641c5

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com> PR-URL: #65326 Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Jordan Harband <ljharb@gmail.com>

codebytere · 2026-08-16T15:51:12Z

Cold require() of a 1000-module CommonJS tree gets ~6 % faster (100 runs, ***) and module/module-require.js type='.js' +13.7 %, by looking up the nearest parent package.json once per directory instead of once per file.

module/module-require.js n=10000 type='.js' *** 13.69 % ±2.36%
module/module-require.js n=10000 type='.json' | 'dir' (don't use this path) -2.1 % (relink noise is ±1 %)
cold require() of a generated 1000-module tree, flat layout (100 runs) *** 6.37 % ±1.76%
same, pnpm-style symlinked node_modules *** 6.56 % ±0.90%
same, with NODE_COMPILE_CACHE warm *** 5.10 % ±1.43%

getNearestParentPackageJSON(checkPath) - called for every CommonJS module to find "type", and by findPackageJSON - memoized per file, so N modules in a directory meant N calls into the native package.json reader. The result only depends on the directory the traversal starts from, so the JS-side cache is now keyed by that directory, derived with the same rule the native side uses (trailing separator → that directory, else dirname). With the permission model enabled the per-path behavior is kept, because the native call performs a per-path read check. Same return values, same objects, same lifetime (both caches were already permanent). 1000 native calls → 236 for the tree above.

Tests: behavior is unchanged; test-module-*, test-require-*, test-esm-*, es-module and test-permission-* (which exercises the per-path branch) pass. Happy to add a call-count test if wanted.

Disclosure: the code, test, measurements and this description were written by Claude Code, directed and reviewed by @codebytere.

nodejs-github-bot · 2026-08-16T15:51:17Z

Review requested:

@nodejs/loaders

codecov · 2026-08-16T17:41:36Z

Codecov Report

❌ Patch coverage is 97.95918% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 90.13%. Comparing base (30bff4a) to head (feee201).
⚠️ Report is 43 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/internal/modules/package_json_reader.js	97.95%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #65326 +/- ##
=======================================
Coverage 90.13% 90.13% =======================================
Files 752 752 Lines 251568 251639 +71 Branches 47270 47285 +15 =======================================
+ Hits 226759 226827 +68 - Misses 16168 16174 +6 + Partials 8641 8638 -3

Files with missing lines	Coverage Δ
lib/internal/modules/package_json_reader.js	`99.23% <97.95%> (-0.22%)`	⬇️

... and 38 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

nodejs-github-bot · 2026-08-16T18:43:54Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75894/

ljharb · 2026-08-17T06:22:23Z

This seems good, but surely there should be some tests that assert the caching behavior, to avoid regressions?

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com>

codebytere · 2026-08-17T17:36:20Z

@ljharb done

nodejs-github-bot · 2026-08-18T16:00:19Z

Commit Queue failed

- Loading data for nodejs/node/pull/65326
✔ Done loading data for nodejs/node/pull/65326
----------------------------------- PR info ------------------------------------
Title module: cache nearest parent package.json per directory (#65326)
Author Shelley Vohr <shelley.vohr@gmail.com> (@codebytere)
Branch codebytere:perf/module-parent-package-json-dir-cache -> nodejs:main
Labels module, commit-queue
Commits 1
- module: cache nearest parent package.json per directory
Committers 1
- Shelley Vohr <shelley.vohr@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
------------------------------ Generated metadata ------------------------------
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
--------------------------------------------------------------------------------
ℹ This PR was created on Sun, 16 Aug 2026 15:51:12 GMT
✔ Approvals: 3
✔ - Yagiz Nizipli (@anonrig) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4947040173
✔ - James M Snell (@jasnell) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4948264808
✔ - Jordan Harband (@ljharb): https://github.com/nodejs/node/pull/65326#pullrequestreview-4953177652
✔ Last GitHub CI successful
ℹ Last Full PR CI on 2026-08-16T18:43:54Z: https://ci.nodejs.org/job/node-test-pull-request/75894/
⚠ Commits were pushed after the last Full PR CI run:
⚠ - module: cache nearest parent package.json per directory
- Querying data for job/node-test-pull-request/75894/
✔ Build data downloaded
- Querying failures of job/node-test-commit/90623/
✔ Data downloaded
✘ 2 failure(s) on the last Jenkins CI run
--------------------------------------------------------------------------------
✔ Aborted `git node land` session in /home/runner/work/node/node/.ncu

https://github.com/nodejs/node/actions/runs/32157693157

nodejs-github-bot · 2026-08-18T19:32:28Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75973/

nodejs-github-bot · 2026-08-19T08:52:08Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75982/

nodejs-github-bot · 2026-08-19T13:18:13Z

Landed in 9c641c5

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com> PR-URL: #65326 Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Jordan Harband <ljharb@gmail.com>

codebytere · 2026-08-16T15:51:12Z

Cold require() of a 1000-module CommonJS tree gets ~6 % faster (100 runs, ***) and module/module-require.js type='.js' +13.7 %, by looking up the nearest parent package.json once per directory instead of once per file.

module/module-require.js n=10000 type='.js' *** 13.69 % ±2.36%
module/module-require.js n=10000 type='.json' | 'dir' (don't use this path) -2.1 % (relink noise is ±1 %)
cold require() of a generated 1000-module tree, flat layout (100 runs) *** 6.37 % ±1.76%
same, pnpm-style symlinked node_modules *** 6.56 % ±0.90%
same, with NODE_COMPILE_CACHE warm *** 5.10 % ±1.43%

getNearestParentPackageJSON(checkPath) - called for every CommonJS module to find "type", and by findPackageJSON - memoized per file, so N modules in a directory meant N calls into the native package.json reader. The result only depends on the directory the traversal starts from, so the JS-side cache is now keyed by that directory, derived with the same rule the native side uses (trailing separator → that directory, else dirname). With the permission model enabled the per-path behavior is kept, because the native call performs a per-path read check. Same return values, same objects, same lifetime (both caches were already permanent). 1000 native calls → 236 for the tree above.

Tests: behavior is unchanged; test-module-*, test-require-*, test-esm-*, es-module and test-permission-* (which exercises the per-path branch) pass. Happy to add a call-count test if wanted.

Disclosure: the code, test, measurements and this description were written by Claude Code, directed and reviewed by @codebytere.

nodejs-github-bot · 2026-08-16T15:51:17Z

Review requested:

@nodejs/loaders

codecov · 2026-08-16T17:41:36Z

Codecov Report

❌ Patch coverage is 97.95918% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 90.13%. Comparing base (30bff4a) to head (feee201).
⚠️ Report is 43 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/internal/modules/package_json_reader.js	97.95%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #65326 +/- ##
=======================================
Coverage 90.13% 90.13% =======================================
Files 752 752 Lines 251568 251639 +71 Branches 47270 47285 +15 =======================================
+ Hits 226759 226827 +68 - Misses 16168 16174 +6 + Partials 8641 8638 -3

Files with missing lines	Coverage Δ
lib/internal/modules/package_json_reader.js	`99.23% <97.95%> (-0.22%)`	⬇️

... and 38 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

nodejs-github-bot · 2026-08-16T18:43:54Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75894/

ljharb · 2026-08-17T06:22:23Z

This seems good, but surely there should be some tests that assert the caching behavior, to avoid regressions?

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com>

codebytere · 2026-08-17T17:36:20Z

@ljharb done

nodejs-github-bot · 2026-08-18T16:00:19Z

Commit Queue failed

- Loading data for nodejs/node/pull/65326
✔ Done loading data for nodejs/node/pull/65326
----------------------------------- PR info ------------------------------------
Title module: cache nearest parent package.json per directory (#65326)
Author Shelley Vohr <shelley.vohr@gmail.com> (@codebytere)
Branch codebytere:perf/module-parent-package-json-dir-cache -> nodejs:main
Labels module, commit-queue
Commits 1
- module: cache nearest parent package.json per directory
Committers 1
- Shelley Vohr <shelley.vohr@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
------------------------------ Generated metadata ------------------------------
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
--------------------------------------------------------------------------------
ℹ This PR was created on Sun, 16 Aug 2026 15:51:12 GMT
✔ Approvals: 3
✔ - Yagiz Nizipli (@anonrig) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4947040173
✔ - James M Snell (@jasnell) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4948264808
✔ - Jordan Harband (@ljharb): https://github.com/nodejs/node/pull/65326#pullrequestreview-4953177652
✔ Last GitHub CI successful
ℹ Last Full PR CI on 2026-08-16T18:43:54Z: https://ci.nodejs.org/job/node-test-pull-request/75894/
⚠ Commits were pushed after the last Full PR CI run:
⚠ - module: cache nearest parent package.json per directory
- Querying data for job/node-test-pull-request/75894/
✔ Build data downloaded
- Querying failures of job/node-test-commit/90623/
✔ Data downloaded
✘ 2 failure(s) on the last Jenkins CI run
--------------------------------------------------------------------------------
✔ Aborted `git node land` session in /home/runner/work/node/node/.ncu

https://github.com/nodejs/node/actions/runs/32157693157

nodejs-github-bot · 2026-08-18T19:32:28Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75973/

nodejs-github-bot · 2026-08-19T08:52:08Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75982/

nodejs-github-bot · 2026-08-19T13:18:13Z

Landed in 9c641c5

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com> PR-URL: #65326 Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Jordan Harband <ljharb@gmail.com>

codebytere · 2026-08-16T15:51:12Z

Cold require() of a 1000-module CommonJS tree gets ~6 % faster (100 runs, ***) and module/module-require.js type='.js' +13.7 %, by looking up the nearest parent package.json once per directory instead of once per file.

module/module-require.js n=10000 type='.js' *** 13.69 % ±2.36%
module/module-require.js n=10000 type='.json' | 'dir' (don't use this path) -2.1 % (relink noise is ±1 %)
cold require() of a generated 1000-module tree, flat layout (100 runs) *** 6.37 % ±1.76%
same, pnpm-style symlinked node_modules *** 6.56 % ±0.90%
same, with NODE_COMPILE_CACHE warm *** 5.10 % ±1.43%

getNearestParentPackageJSON(checkPath) - called for every CommonJS module to find "type", and by findPackageJSON - memoized per file, so N modules in a directory meant N calls into the native package.json reader. The result only depends on the directory the traversal starts from, so the JS-side cache is now keyed by that directory, derived with the same rule the native side uses (trailing separator → that directory, else dirname). With the permission model enabled the per-path behavior is kept, because the native call performs a per-path read check. Same return values, same objects, same lifetime (both caches were already permanent). 1000 native calls → 236 for the tree above.

Tests: behavior is unchanged; test-module-*, test-require-*, test-esm-*, es-module and test-permission-* (which exercises the per-path branch) pass. Happy to add a call-count test if wanted.

Disclosure: the code, test, measurements and this description were written by Claude Code, directed and reviewed by @codebytere.

nodejs-github-bot · 2026-08-16T15:51:17Z

Review requested:

@nodejs/loaders

codecov · 2026-08-16T17:41:36Z

Codecov Report

❌ Patch coverage is 97.95918% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 90.13%. Comparing base (30bff4a) to head (feee201).
⚠️ Report is 43 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/internal/modules/package_json_reader.js	97.95%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #65326 +/- ##
=======================================
Coverage 90.13% 90.13% =======================================
Files 752 752 Lines 251568 251639 +71 Branches 47270 47285 +15 =======================================
+ Hits 226759 226827 +68 - Misses 16168 16174 +6 + Partials 8641 8638 -3

Files with missing lines	Coverage Δ
lib/internal/modules/package_json_reader.js	`99.23% <97.95%> (-0.22%)`	⬇️

... and 38 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

nodejs-github-bot · 2026-08-16T18:43:54Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75894/

ljharb · 2026-08-17T06:22:23Z

This seems good, but surely there should be some tests that assert the caching behavior, to avoid regressions?

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com>

codebytere · 2026-08-17T17:36:20Z

@ljharb done

nodejs-github-bot · 2026-08-18T16:00:19Z

Commit Queue failed

- Loading data for nodejs/node/pull/65326
✔ Done loading data for nodejs/node/pull/65326
----------------------------------- PR info ------------------------------------
Title module: cache nearest parent package.json per directory (#65326)
Author Shelley Vohr <shelley.vohr@gmail.com> (@codebytere)
Branch codebytere:perf/module-parent-package-json-dir-cache -> nodejs:main
Labels module, commit-queue
Commits 1
- module: cache nearest parent package.json per directory
Committers 1
- Shelley Vohr <shelley.vohr@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
------------------------------ Generated metadata ------------------------------
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
--------------------------------------------------------------------------------
ℹ This PR was created on Sun, 16 Aug 2026 15:51:12 GMT
✔ Approvals: 3
✔ - Yagiz Nizipli (@anonrig) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4947040173
✔ - James M Snell (@jasnell) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4948264808
✔ - Jordan Harband (@ljharb): https://github.com/nodejs/node/pull/65326#pullrequestreview-4953177652
✔ Last GitHub CI successful
ℹ Last Full PR CI on 2026-08-16T18:43:54Z: https://ci.nodejs.org/job/node-test-pull-request/75894/
⚠ Commits were pushed after the last Full PR CI run:
⚠ - module: cache nearest parent package.json per directory
- Querying data for job/node-test-pull-request/75894/
✔ Build data downloaded
- Querying failures of job/node-test-commit/90623/
✔ Data downloaded
✘ 2 failure(s) on the last Jenkins CI run
--------------------------------------------------------------------------------
✔ Aborted `git node land` session in /home/runner/work/node/node/.ncu

https://github.com/nodejs/node/actions/runs/32157693157

nodejs-github-bot · 2026-08-18T19:32:28Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75973/

nodejs-github-bot · 2026-08-19T08:52:08Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75982/

nodejs-github-bot · 2026-08-19T13:18:13Z

Landed in 9c641c5

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com> PR-URL: #65326 Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Jordan Harband <ljharb@gmail.com>

codebytere · 2026-08-16T15:51:12Z

Cold require() of a 1000-module CommonJS tree gets ~6 % faster (100 runs, ***) and module/module-require.js type='.js' +13.7 %, by looking up the nearest parent package.json once per directory instead of once per file.

module/module-require.js n=10000 type='.js' *** 13.69 % ±2.36%
module/module-require.js n=10000 type='.json' | 'dir' (don't use this path) -2.1 % (relink noise is ±1 %)
cold require() of a generated 1000-module tree, flat layout (100 runs) *** 6.37 % ±1.76%
same, pnpm-style symlinked node_modules *** 6.56 % ±0.90%
same, with NODE_COMPILE_CACHE warm *** 5.10 % ±1.43%

getNearestParentPackageJSON(checkPath) - called for every CommonJS module to find "type", and by findPackageJSON - memoized per file, so N modules in a directory meant N calls into the native package.json reader. The result only depends on the directory the traversal starts from, so the JS-side cache is now keyed by that directory, derived with the same rule the native side uses (trailing separator → that directory, else dirname). With the permission model enabled the per-path behavior is kept, because the native call performs a per-path read check. Same return values, same objects, same lifetime (both caches were already permanent). 1000 native calls → 236 for the tree above.

Tests: behavior is unchanged; test-module-*, test-require-*, test-esm-*, es-module and test-permission-* (which exercises the per-path branch) pass. Happy to add a call-count test if wanted.

Disclosure: the code, test, measurements and this description were written by Claude Code, directed and reviewed by @codebytere.

nodejs-github-bot · 2026-08-16T15:51:17Z

Review requested:

@nodejs/loaders

codecov · 2026-08-16T17:41:36Z

Codecov Report

❌ Patch coverage is 97.95918% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 90.13%. Comparing base (30bff4a) to head (feee201).
⚠️ Report is 43 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/internal/modules/package_json_reader.js	97.95%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #65326 +/- ##
=======================================
Coverage 90.13% 90.13% =======================================
Files 752 752 Lines 251568 251639 +71 Branches 47270 47285 +15 =======================================
+ Hits 226759 226827 +68 - Misses 16168 16174 +6 + Partials 8641 8638 -3

Files with missing lines	Coverage Δ
lib/internal/modules/package_json_reader.js	`99.23% <97.95%> (-0.22%)`	⬇️

... and 38 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

nodejs-github-bot · 2026-08-16T18:43:54Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75894/

ljharb · 2026-08-17T06:22:23Z

This seems good, but surely there should be some tests that assert the caching behavior, to avoid regressions?

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com>

codebytere · 2026-08-17T17:36:20Z

@ljharb done

nodejs-github-bot · 2026-08-18T16:00:19Z

Commit Queue failed

- Loading data for nodejs/node/pull/65326
✔ Done loading data for nodejs/node/pull/65326
----------------------------------- PR info ------------------------------------
Title module: cache nearest parent package.json per directory (#65326)
Author Shelley Vohr <shelley.vohr@gmail.com> (@codebytere)
Branch codebytere:perf/module-parent-package-json-dir-cache -> nodejs:main
Labels module, commit-queue
Commits 1
- module: cache nearest parent package.json per directory
Committers 1
- Shelley Vohr <shelley.vohr@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
------------------------------ Generated metadata ------------------------------
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
--------------------------------------------------------------------------------
ℹ This PR was created on Sun, 16 Aug 2026 15:51:12 GMT
✔ Approvals: 3
✔ - Yagiz Nizipli (@anonrig) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4947040173
✔ - James M Snell (@jasnell) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4948264808
✔ - Jordan Harband (@ljharb): https://github.com/nodejs/node/pull/65326#pullrequestreview-4953177652
✔ Last GitHub CI successful
ℹ Last Full PR CI on 2026-08-16T18:43:54Z: https://ci.nodejs.org/job/node-test-pull-request/75894/
⚠ Commits were pushed after the last Full PR CI run:
⚠ - module: cache nearest parent package.json per directory
- Querying data for job/node-test-pull-request/75894/
✔ Build data downloaded
- Querying failures of job/node-test-commit/90623/
✔ Data downloaded
✘ 2 failure(s) on the last Jenkins CI run
--------------------------------------------------------------------------------
✔ Aborted `git node land` session in /home/runner/work/node/node/.ncu

https://github.com/nodejs/node/actions/runs/32157693157

nodejs-github-bot · 2026-08-18T19:32:28Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75973/

nodejs-github-bot · 2026-08-19T08:52:08Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75982/

nodejs-github-bot · 2026-08-19T13:18:13Z

Landed in 9c641c5

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com> PR-URL: #65326 Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Jordan Harband <ljharb@gmail.com>

codebytere · 2026-08-16T15:51:12Z

Cold require() of a 1000-module CommonJS tree gets ~6 % faster (100 runs, ***) and module/module-require.js type='.js' +13.7 %, by looking up the nearest parent package.json once per directory instead of once per file.

module/module-require.js n=10000 type='.js' *** 13.69 % ±2.36%
module/module-require.js n=10000 type='.json' | 'dir' (don't use this path) -2.1 % (relink noise is ±1 %)
cold require() of a generated 1000-module tree, flat layout (100 runs) *** 6.37 % ±1.76%
same, pnpm-style symlinked node_modules *** 6.56 % ±0.90%
same, with NODE_COMPILE_CACHE warm *** 5.10 % ±1.43%

getNearestParentPackageJSON(checkPath) - called for every CommonJS module to find "type", and by findPackageJSON - memoized per file, so N modules in a directory meant N calls into the native package.json reader. The result only depends on the directory the traversal starts from, so the JS-side cache is now keyed by that directory, derived with the same rule the native side uses (trailing separator → that directory, else dirname). With the permission model enabled the per-path behavior is kept, because the native call performs a per-path read check. Same return values, same objects, same lifetime (both caches were already permanent). 1000 native calls → 236 for the tree above.

Tests: behavior is unchanged; test-module-*, test-require-*, test-esm-*, es-module and test-permission-* (which exercises the per-path branch) pass. Happy to add a call-count test if wanted.

Disclosure: the code, test, measurements and this description were written by Claude Code, directed and reviewed by @codebytere.

nodejs-github-bot · 2026-08-16T15:51:17Z

Review requested:

@nodejs/loaders

codecov · 2026-08-16T17:41:36Z

Codecov Report

❌ Patch coverage is 97.95918% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 90.13%. Comparing base (30bff4a) to head (feee201).
⚠️ Report is 43 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/internal/modules/package_json_reader.js	97.95%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #65326 +/- ##
=======================================
Coverage 90.13% 90.13% =======================================
Files 752 752 Lines 251568 251639 +71 Branches 47270 47285 +15 =======================================
+ Hits 226759 226827 +68 - Misses 16168 16174 +6 + Partials 8641 8638 -3

Files with missing lines	Coverage Δ
lib/internal/modules/package_json_reader.js	`99.23% <97.95%> (-0.22%)`	⬇️

... and 38 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

nodejs-github-bot · 2026-08-16T18:43:54Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75894/

ljharb · 2026-08-17T06:22:23Z

This seems good, but surely there should be some tests that assert the caching behavior, to avoid regressions?

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com>

codebytere · 2026-08-17T17:36:20Z

@ljharb done

nodejs-github-bot · 2026-08-18T16:00:19Z

Commit Queue failed

- Loading data for nodejs/node/pull/65326
✔ Done loading data for nodejs/node/pull/65326
----------------------------------- PR info ------------------------------------
Title module: cache nearest parent package.json per directory (#65326)
Author Shelley Vohr <shelley.vohr@gmail.com> (@codebytere)
Branch codebytere:perf/module-parent-package-json-dir-cache -> nodejs:main
Labels module, commit-queue
Commits 1
- module: cache nearest parent package.json per directory
Committers 1
- Shelley Vohr <shelley.vohr@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
------------------------------ Generated metadata ------------------------------
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
--------------------------------------------------------------------------------
ℹ This PR was created on Sun, 16 Aug 2026 15:51:12 GMT
✔ Approvals: 3
✔ - Yagiz Nizipli (@anonrig) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4947040173
✔ - James M Snell (@jasnell) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4948264808
✔ - Jordan Harband (@ljharb): https://github.com/nodejs/node/pull/65326#pullrequestreview-4953177652
✔ Last GitHub CI successful
ℹ Last Full PR CI on 2026-08-16T18:43:54Z: https://ci.nodejs.org/job/node-test-pull-request/75894/
⚠ Commits were pushed after the last Full PR CI run:
⚠ - module: cache nearest parent package.json per directory
- Querying data for job/node-test-pull-request/75894/
✔ Build data downloaded
- Querying failures of job/node-test-commit/90623/
✔ Data downloaded
✘ 2 failure(s) on the last Jenkins CI run
--------------------------------------------------------------------------------
✔ Aborted `git node land` session in /home/runner/work/node/node/.ncu

https://github.com/nodejs/node/actions/runs/32157693157

nodejs-github-bot · 2026-08-18T19:32:28Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75973/

nodejs-github-bot · 2026-08-19T08:52:08Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75982/

nodejs-github-bot · 2026-08-19T13:18:13Z

Landed in 9c641c5

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com> PR-URL: #65326 Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Jordan Harband <ljharb@gmail.com>

codebytere · 2026-08-16T15:51:12Z

Cold require() of a 1000-module CommonJS tree gets ~6 % faster (100 runs, ***) and module/module-require.js type='.js' +13.7 %, by looking up the nearest parent package.json once per directory instead of once per file.

module/module-require.js n=10000 type='.js' *** 13.69 % ±2.36%
module/module-require.js n=10000 type='.json' | 'dir' (don't use this path) -2.1 % (relink noise is ±1 %)
cold require() of a generated 1000-module tree, flat layout (100 runs) *** 6.37 % ±1.76%
same, pnpm-style symlinked node_modules *** 6.56 % ±0.90%
same, with NODE_COMPILE_CACHE warm *** 5.10 % ±1.43%

getNearestParentPackageJSON(checkPath) - called for every CommonJS module to find "type", and by findPackageJSON - memoized per file, so N modules in a directory meant N calls into the native package.json reader. The result only depends on the directory the traversal starts from, so the JS-side cache is now keyed by that directory, derived with the same rule the native side uses (trailing separator → that directory, else dirname). With the permission model enabled the per-path behavior is kept, because the native call performs a per-path read check. Same return values, same objects, same lifetime (both caches were already permanent). 1000 native calls → 236 for the tree above.

Tests: behavior is unchanged; test-module-*, test-require-*, test-esm-*, es-module and test-permission-* (which exercises the per-path branch) pass. Happy to add a call-count test if wanted.

Disclosure: the code, test, measurements and this description were written by Claude Code, directed and reviewed by @codebytere.

nodejs-github-bot · 2026-08-16T15:51:17Z

Review requested:

@nodejs/loaders

codecov · 2026-08-16T17:41:36Z

Codecov Report

❌ Patch coverage is 97.95918% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 90.13%. Comparing base (30bff4a) to head (feee201).
⚠️ Report is 43 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/internal/modules/package_json_reader.js	97.95%	1 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #65326 +/- ##
=======================================
Coverage 90.13% 90.13% =======================================
Files 752 752 Lines 251568 251639 +71 Branches 47270 47285 +15 =======================================
+ Hits 226759 226827 +68 - Misses 16168 16174 +6 + Partials 8641 8638 -3

Files with missing lines	Coverage Δ
lib/internal/modules/package_json_reader.js	`99.23% <97.95%> (-0.22%)`	⬇️

... and 38 files with indirect coverage changes

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

nodejs-github-bot · 2026-08-16T18:43:54Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75894/

ljharb · 2026-08-17T06:22:23Z

This seems good, but surely there should be some tests that assert the caching behavior, to avoid regressions?

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com>

codebytere · 2026-08-17T17:36:20Z

@ljharb done

nodejs-github-bot · 2026-08-18T16:00:19Z

Commit Queue failed

- Loading data for nodejs/node/pull/65326
✔ Done loading data for nodejs/node/pull/65326
----------------------------------- PR info ------------------------------------
Title module: cache nearest parent package.json per directory (#65326)
Author Shelley Vohr <shelley.vohr@gmail.com> (@codebytere)
Branch codebytere:perf/module-parent-package-json-dir-cache -> nodejs:main
Labels module, commit-queue
Commits 1
- module: cache nearest parent package.json per directory
Committers 1
- Shelley Vohr <shelley.vohr@gmail.com>
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
------------------------------ Generated metadata ------------------------------
PR-URL: https://github.com/nodejs/node/pull/65326
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Jordan Harband <ljharb@gmail.com>
--------------------------------------------------------------------------------
ℹ This PR was created on Sun, 16 Aug 2026 15:51:12 GMT
✔ Approvals: 3
✔ - Yagiz Nizipli (@anonrig) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4947040173
✔ - James M Snell (@jasnell) (TSC): https://github.com/nodejs/node/pull/65326#pullrequestreview-4948264808
✔ - Jordan Harband (@ljharb): https://github.com/nodejs/node/pull/65326#pullrequestreview-4953177652
✔ Last GitHub CI successful
ℹ Last Full PR CI on 2026-08-16T18:43:54Z: https://ci.nodejs.org/job/node-test-pull-request/75894/
⚠ Commits were pushed after the last Full PR CI run:
⚠ - module: cache nearest parent package.json per directory
- Querying data for job/node-test-pull-request/75894/
✔ Build data downloaded
- Querying failures of job/node-test-commit/90623/
✔ Data downloaded
✘ 2 failure(s) on the last Jenkins CI run
--------------------------------------------------------------------------------
✔ Aborted `git node land` session in /home/runner/work/node/node/.ncu

https://github.com/nodejs/node/actions/runs/32157693157

nodejs-github-bot · 2026-08-18T19:32:28Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75973/

nodejs-github-bot · 2026-08-19T08:52:08Z

CI: https://ci.nodejs.org/job/node-test-pull-request/75982/

nodejs-github-bot · 2026-08-19T13:18:13Z

Landed in 9c641c5

getNearestParentPackageJSON() memoized its answer per file, so every module loaded still made one native call, and TraverseParent() builds several std::filesystem::path temporaries per directory level and serializes the whole package.json, which the JS side then usually discarded because it already had that package.json deserialized. The native traversal starts at the directory of the given path, so the answer only depends on that directory: key the memo by it (following NormalizePath()'s trailing-separator rule), so that all modules in a directory share one native call. When the permission model is enabled the traversal also depends on the read permissions in effect at call time, so that configuration keeps the per-file cache. Loading a 1000-module tree spread over ~240 directories goes from 1000 to 236 native calls. Signed-off-by: Shelley Vohr <shelley.vohr@gmail.com> PR-URL: #65326 Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Jordan Harband <ljharb@gmail.com>

nodejs-github-bot added module Issues and PRs related to the module subsystem. needs-ci PRs that need a full CI run. labels Aug 16, 2026

codebytere added request-ci Add this label to start a Jenkins CI on a PR. and removed needs-ci PRs that need a full CI run. labels Aug 16, 2026

codebytere requested review from GeoffreyBooth, IlyasShabi and anonrig August 16, 2026 18:36

anonrig approved these changes Aug 16, 2026
View reviewed changes

github-actionsBot removed the request-ci Add this label to start a Jenkins CI on a PR. label Aug 16, 2026

jasnell approved these changes Aug 17, 2026
View reviewed changes

codebytere force-pushed the perf/module-parent-package-json-dir-cache branch from cbba4de to feee201Compare August 17, 2026 06:25

ljharb approved these changes Aug 17, 2026
View reviewed changes

jsumners-nr approved these changes Aug 17, 2026
View reviewed changes

codebytere added the commit-queue PRs queued for automated landing through the Commit Queue. label Aug 18, 2026

nodejs-github-bot added commit-queue-failed PRs whose Commit Queue landing failed and need manual intervention before retrying. and removed commit-queue PRs queued for automated landing through the Commit Queue. labels Aug 18, 2026

codebytere added the request-ci Add this label to start a Jenkins CI on a PR. label Aug 18, 2026

github-actionsBot removed the request-ci Add this label to start a Jenkins CI on a PR. label Aug 18, 2026

codebytere added commit-queue PRs queued for automated landing through the Commit Queue. and removed commit-queue-failed PRs whose Commit Queue landing failed and need manual intervention before retrying. labels Aug 19, 2026

nodejs-github-bot merged commit 9c641c5 into nodejs:mainAug 19, 2026
78 checks passed

nodejs-github-bot removed the commit-queue PRs queued for automated landing through the Commit Queue. label Aug 19, 2026

github-actionsBot mentioned this pull request Aug 25, 2026
2026-08-26, Version 26.8.0 (Current) #65551
Merged