src: improve utf8 string generation performance #54873

anonrig · 2024-09-10T14:29:07Z

This speeds up Utf8Value across all Node.js codebase. We add a fast path for one byte strings and use the new V8 API (String::ValueView) with simdutf for non-byte strings. Once simdutf supports null termination and invalid character replacement, we can improve the performance more.

cc @nodejs/performance @lemire

Benchmark CI: https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/1637/

github-actions · 2024-09-10T14:52:39Z

Failed to start CI

- Validating Jenkins credentials
✔  Jenkins credentials valid
- Starting PR CI job
✘  Failed to start PR CI: 400 Bad Request

https://github.com/nodejs/node/actions/runs/10795045779

jasnell · 2024-09-10T16:01:00Z

CI failures look relevant to this change.

src/util.cc

RafaelGSS · 2024-09-10T17:05:02Z

Since the PR is claiming for performance, we should have benchmarks.

codecov · 2024-09-10T17:55:39Z

Codecov Report

Attention: Patch coverage is 84.61538% with 4 lines in your changes missing coverage. Please review.

Project coverage is 88.08%. Comparing base (9db6327) to head (4b28cf1).
Report is 271 commits behind head on main.

Files with missing lines	Patch %	Lines
src/string_bytes.cc	66.66%	2 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #54873      +/-   ##
==========================================
+ Coverage   88.05%   88.08%   +0.02%     
==========================================
  Files         651      651              
  Lines      183405   183551     +146     
  Branches    35822    35867      +45     
==========================================
+ Hits       161499   161676     +177     
+ Misses      15159    15133      -26     
+ Partials     6747     6742       -5

Files with missing lines	Coverage Δ
src/util.cc	`87.34% <100.00%> (+0.32%)`	⬆️
src/string_bytes.cc	`68.43% <66.66%> (+4.48%)`	⬆️

... and 50 files with indirect coverage changes

anonrig · 2024-09-10T18:10:21Z

Since the PR is claiming for performance, we should have benchmarks.

I don't know how to have an isolated benchmark for this. Any recommendations @RafaelGSS?

RafaelGSS · 2024-09-10T18:13:26Z

I don't know how to have an isolated benchmark for this.

So, how do you know this PR is improving performance then?

Any recommendations?

I suspect if you measure any public API that makes use of this new branch behind the scenes you should get a feasible result

mcollina

RSLGTM

src/util.cc

Co-authored-by: Robert Nagy <[email protected]>

src/util.cc

Co-authored-by: Robert Nagy <[email protected]>

nodejs-github-bot · 2024-09-15T01:49:16Z

CI: https://ci.nodejs.org/job/node-test-pull-request/62446/

nodejs-github-bot · 2024-09-15T14:16:54Z

CI: https://ci.nodejs.org/job/node-test-pull-request/62453/

nodejs-github-bot · 2024-09-15T19:36:32Z

CI: https://ci.nodejs.org/job/node-test-pull-request/62460/

anonrig · 2024-09-15T19:37:28Z

@ronag Can you review?

nodejs-github-bot · 2024-09-16T22:14:33Z

CI: https://ci.nodejs.org/job/node-test-pull-request/62487/

Addressed the recommendation. Dismissing.

nodejs-github-bot · 2024-09-17T00:19:48Z

Landed in 8191e1f

PR-URL: nodejs#54873 Reviewed-By: Daniel Lemire <[email protected]> Reviewed-By: Matteo Collina <[email protected]> Reviewed-By: James M Snell <[email protected]> Reviewed-By: Stephen Belanger <[email protected]>

nodejs/node#54873

* chore: upgrade Node.js to v24.10.0 * chore: fixup crypto patch * chore: fixup crypto test patch * src: prepare for v8 sandboxing nodejs/node#58376 * esm: fix module.exports export on CJS modules nodejs/node#57366 * chore: fixup lazyload fs patch * esm: Source Phase Imports for WebAssembly nodejs/node#56919 * module: remove --experimental-default-type nodejs/node#56092 * lib,src: refactor assert to load error source from memory nodejs/node#59751 * src: add source location to v8::TaskRunner nodejs/node#54077 * src: remove dependency on wrapper-descriptor-based CppHeap nodejs/node#54077 * src: do not use soon-to-be-deprecated V8 API nodejs/node#53174 * src: stop using deprecated fields of v8::FastApiCallbackOptions nodejs/node#54077 * test: update v8-stats test for V8 12.6 nodejs/node#54077 * esm: unflag --experimental-wasm-modules nodejs/node#57038 * test: adapt assert tests to stack trace changes nodejs/node#58070 * src,test: unregister the isolate after disposal and before freeing nodejs/node#58070 * src: use cppgc to manage ContextifyContext nodejs/node#56522 * src: replace uses of FastApiTypedArray nodejs/node#58070 * module: integrate TypeScript into compile cache nodejs/node#56629 * deps: update ada to 3.2.7 nodejs/node#59336 * src: make minor cleanups in encoding_binding.cc nodejs/node#57448 * src: switch from `Get/SetPrototype` to `Get/SetPrototypeV2` nodejs/node#55453 * src: use non-deprecated Get/SetPrototype methods nodejs/node#59671 * src: simplify string_bytes with views nodejs/node#54876 * src: improve utf8 string generation performance nodejs/node#54873 * src: use non-deprecated Utf8LengthV2() method nodejs/node#58070 * src: use non-deprecated WriteUtf8V2() method nodejs/node#58070 * src: refactor WriteUCS2 and remove flags argument nodejs/node#58163 * src: use String::WriteV2() in TwoByteValue nodejs/node#58164 * node-api: use WriteV2 in napi_get_value_string_utf16 nodejs/node#58165 * node-api: use WriteOneByteV2 in napi_get_value_string_latin1 nodejs/node#58325 * src: migrate WriteOneByte to WriteOneByteV2 nodejs/node#59634 * fs: introduce dirent\.parentPath nodejs/node#50976 * src: avoid copy by using std::views::keys nodejs/node#56080 * chore: fixup patch indices * fix: errant use of context->GetIsolate() * fix: tweak BoringSSL compat patch for new changes * fix: add back missing isolate dtor declaration * fixup! esm: fix module.exports export on CJS modules * cli: remove --no-experimental-fetch flag https://github.com/nodejs/node/pull/52611/files * esm: Source Phase Imports for WebAssembly nodejs/node#56919 * fixup! src: prepare for v8 sandboxing * chore: bump @types/node to v24 * chore: fix const assignment in crypto test * fix: sandbox pointer patch issues * chore: rework source phase import patch * src: add percentage support to --max-old-space-size nodejs/node#59082 * chore: fixup crypto tests * chore: HostImportModuleWithPhaseDynamically todo * fix: cjs esm failures * fix: v8::Object::Wrappable issues - v8/node@b72a615 - v8/node@490bac2 - v8/node@4896a0d * chore: remove deleted specs * src: use v8::ExternalMemoryAccounter nodejs/node#58070 * fs: port SonicBoom module to fs module as FastUtf8Stream nodejs/node#58897 * chore: tweak sandboxed pr patch * test: disable parallel/test-os-checked-function * test: use WHATWG URL instead of url.parse * fix: OPENSSL_secure_zalloc doesn't work in BoringSSL * chore: fix accidental extra line * 7017517: [defer-import-eval] Parse import defer syntax https://chromium-review.googlesource.com/c/v8/v8/+/7017517

nodejs/node#54873 Co-authored-by: Shelley Vohr <[email protected]>

* chore: upgrade Node.js to v24.10.0 * chore: fixup crypto patch * chore: fixup crypto test patch * src: prepare for v8 sandboxing nodejs/node#58376 * esm: fix module.exports export on CJS modules nodejs/node#57366 * chore: fixup lazyload fs patch * esm: Source Phase Imports for WebAssembly nodejs/node#56919 * module: remove --experimental-default-type nodejs/node#56092 * lib,src: refactor assert to load error source from memory nodejs/node#59751 * src: add source location to v8::TaskRunner nodejs/node#54077 * src: remove dependency on wrapper-descriptor-based CppHeap nodejs/node#54077 * src: do not use soon-to-be-deprecated V8 API nodejs/node#53174 * src: stop using deprecated fields of v8::FastApiCallbackOptions nodejs/node#54077 * test: update v8-stats test for V8 12.6 nodejs/node#54077 * esm: unflag --experimental-wasm-modules nodejs/node#57038 * test: adapt assert tests to stack trace changes nodejs/node#58070 * src,test: unregister the isolate after disposal and before freeing nodejs/node#58070 * src: use cppgc to manage ContextifyContext nodejs/node#56522 * src: replace uses of FastApiTypedArray nodejs/node#58070 * module: integrate TypeScript into compile cache nodejs/node#56629 * deps: update ada to 3.2.7 nodejs/node#59336 * src: make minor cleanups in encoding_binding.cc nodejs/node#57448 * src: switch from `Get/SetPrototype` to `Get/SetPrototypeV2` nodejs/node#55453 * src: use non-deprecated Get/SetPrototype methods nodejs/node#59671 * src: simplify string_bytes with views nodejs/node#54876 * src: improve utf8 string generation performance nodejs/node#54873 * src: use non-deprecated Utf8LengthV2() method nodejs/node#58070 * src: use non-deprecated WriteUtf8V2() method nodejs/node#58070 * src: refactor WriteUCS2 and remove flags argument nodejs/node#58163 * src: use String::WriteV2() in TwoByteValue nodejs/node#58164 * node-api: use WriteV2 in napi_get_value_string_utf16 nodejs/node#58165 * node-api: use WriteOneByteV2 in napi_get_value_string_latin1 nodejs/node#58325 * src: migrate WriteOneByte to WriteOneByteV2 nodejs/node#59634 * fs: introduce dirent\.parentPath nodejs/node#50976 * src: avoid copy by using std::views::keys nodejs/node#56080 * chore: fixup patch indices * fix: errant use of context->GetIsolate() * fix: tweak BoringSSL compat patch for new changes * fix: add back missing isolate dtor declaration * fixup! esm: fix module.exports export on CJS modules * cli: remove --no-experimental-fetch flag https://github.com/nodejs/node/pull/52611/files * esm: Source Phase Imports for WebAssembly nodejs/node#56919 * fixup! src: prepare for v8 sandboxing * chore: bump @types/node to v24 * chore: fix const assignment in crypto test * fix: sandbox pointer patch issues * chore: rework source phase import patch * src: add percentage support to --max-old-space-size nodejs/node#59082 * chore: fixup crypto tests * chore: HostImportModuleWithPhaseDynamically todo * fix: cjs esm failures * fix: v8::Object::Wrappable issues - v8/node@b72a615 - v8/node@490bac2 - v8/node@4896a0d * chore: remove deleted specs * src: use v8::ExternalMemoryAccounter nodejs/node#58070 * fs: port SonicBoom module to fs module as FastUtf8Stream nodejs/node#58897 * chore: tweak sandboxed pr patch * test: disable parallel/test-os-checked-function * test: use WHATWG URL instead of url.parse * fix: OPENSSL_secure_zalloc doesn't work in BoringSSL * chore: fix accidental extra line * 7017517: [defer-import-eval] Parse import defer syntax https://chromium-review.googlesource.com/c/v8/v8/+/7017517

* chore: upgrade Node.js to v24.10.0 Co-authored-by: Shelley Vohr <[email protected]> * chore: fixup crypto patch Co-authored-by: Shelley Vohr <[email protected]> * chore: fixup crypto test patch Co-authored-by: Shelley Vohr <[email protected]> * src: prepare for v8 sandboxing nodejs/node#58376 Co-authored-by: Shelley Vohr <[email protected]> * esm: fix module.exports export on CJS modules nodejs/node#57366 Co-authored-by: Shelley Vohr <[email protected]> * chore: fixup lazyload fs patch Co-authored-by: Shelley Vohr <[email protected]> * esm: Source Phase Imports for WebAssembly nodejs/node#56919 Co-authored-by: Shelley Vohr <[email protected]> * module: remove --experimental-default-type nodejs/node#56092 Co-authored-by: Shelley Vohr <[email protected]> * lib,src: refactor assert to load error source from memory nodejs/node#59751 Co-authored-by: Shelley Vohr <[email protected]> * src: add source location to v8::TaskRunner nodejs/node#54077 Co-authored-by: Shelley Vohr <[email protected]> * src: remove dependency on wrapper-descriptor-based CppHeap nodejs/node#54077 Co-authored-by: Shelley Vohr <[email protected]> * src: do not use soon-to-be-deprecated V8 API nodejs/node#53174 Co-authored-by: Shelley Vohr <[email protected]> * src: stop using deprecated fields of v8::FastApiCallbackOptions nodejs/node#54077 Co-authored-by: Shelley Vohr <[email protected]> * test: update v8-stats test for V8 12.6 nodejs/node#54077 Co-authored-by: Shelley Vohr <[email protected]> * esm: unflag --experimental-wasm-modules nodejs/node#57038 Co-authored-by: Shelley Vohr <[email protected]> * test: adapt assert tests to stack trace changes nodejs/node#58070 Co-authored-by: Shelley Vohr <[email protected]> * src,test: unregister the isolate after disposal and before freeing nodejs/node#58070 Co-authored-by: Shelley Vohr <[email protected]> * src: use cppgc to manage ContextifyContext nodejs/node#56522 Co-authored-by: Shelley Vohr <[email protected]> * src: replace uses of FastApiTypedArray nodejs/node#58070 Co-authored-by: Shelley Vohr <[email protected]> * module: integrate TypeScript into compile cache nodejs/node#56629 Co-authored-by: Shelley Vohr <[email protected]> * deps: update ada to 3.2.7 nodejs/node#59336 Co-authored-by: Shelley Vohr <[email protected]> * src: make minor cleanups in encoding_binding.cc nodejs/node#57448 Co-authored-by: Shelley Vohr <[email protected]> * src: switch from `Get/SetPrototype` to `Get/SetPrototypeV2` nodejs/node#55453 Co-authored-by: Shelley Vohr <[email protected]> * src: use non-deprecated Get/SetPrototype methods nodejs/node#59671 Co-authored-by: Shelley Vohr <[email protected]> * src: simplify string_bytes with views nodejs/node#54876 Co-authored-by: Shelley Vohr <[email protected]> * src: improve utf8 string generation performance nodejs/node#54873 Co-authored-by: Shelley Vohr <[email protected]> * src: use non-deprecated Utf8LengthV2() method nodejs/node#58070 Co-authored-by: Shelley Vohr <[email protected]> * src: use non-deprecated WriteUtf8V2() method nodejs/node#58070 Co-authored-by: Shelley Vohr <[email protected]> * src: refactor WriteUCS2 and remove flags argument nodejs/node#58163 Co-authored-by: Shelley Vohr <[email protected]> * src: use String::WriteV2() in TwoByteValue nodejs/node#58164 Co-authored-by: Shelley Vohr <[email protected]> * node-api: use WriteV2 in napi_get_value_string_utf16 nodejs/node#58165 Co-authored-by: Shelley Vohr <[email protected]> * node-api: use WriteOneByteV2 in napi_get_value_string_latin1 nodejs/node#58325 Co-authored-by: Shelley Vohr <[email protected]> * src: migrate WriteOneByte to WriteOneByteV2 nodejs/node#59634 Co-authored-by: Shelley Vohr <[email protected]> * fs: introduce dirent\.parentPath nodejs/node#50976 Co-authored-by: Shelley Vohr <[email protected]> * src: avoid copy by using std::views::keys nodejs/node#56080 Co-authored-by: Shelley Vohr <[email protected]> * chore: fixup patch indices Co-authored-by: Shelley Vohr <[email protected]> * fix: errant use of context->GetIsolate() Co-authored-by: Shelley Vohr <[email protected]> * fix: tweak BoringSSL compat patch for new changes Co-authored-by: Shelley Vohr <[email protected]> * fix: add back missing isolate dtor declaration Co-authored-by: Shelley Vohr <[email protected]> * fixup! esm: fix module.exports export on CJS modules Co-authored-by: Shelley Vohr <[email protected]> * cli: remove --no-experimental-fetch flag https://github.com/nodejs/node/pull/52611/files Co-authored-by: Shelley Vohr <[email protected]> * esm: Source Phase Imports for WebAssembly nodejs/node#56919 Co-authored-by: Shelley Vohr <[email protected]> * fixup! src: prepare for v8 sandboxing Co-authored-by: Shelley Vohr <[email protected]> * chore: bump @types/node to v24 Co-authored-by: Shelley Vohr <[email protected]> * chore: fix const assignment in crypto test Co-authored-by: Shelley Vohr <[email protected]> * fix: sandbox pointer patch issues Co-authored-by: Shelley Vohr <[email protected]> * chore: rework source phase import patch Co-authored-by: Shelley Vohr <[email protected]> * src: add percentage support to --max-old-space-size nodejs/node#59082 Co-authored-by: Shelley Vohr <[email protected]> * chore: fixup crypto tests Co-authored-by: Shelley Vohr <[email protected]> * chore: HostImportModuleWithPhaseDynamically todo Co-authored-by: Shelley Vohr <[email protected]> * fix: cjs esm failures Co-authored-by: Shelley Vohr <[email protected]> * fix: v8::Object::Wrappable issues - v8/node@b72a615 - v8/node@490bac2 - v8/node@4896a0d Co-authored-by: Shelley Vohr <[email protected]> * chore: remove deleted specs Co-authored-by: Shelley Vohr <[email protected]> * src: use v8::ExternalMemoryAccounter nodejs/node#58070 Co-authored-by: Shelley Vohr <[email protected]> * fs: port SonicBoom module to fs module as FastUtf8Stream nodejs/node#58897 Co-authored-by: Shelley Vohr <[email protected]> * chore: tweak sandboxed pr patch Co-authored-by: Shelley Vohr <[email protected]> * test: disable parallel/test-os-checked-function Co-authored-by: Shelley Vohr <[email protected]> * test: use WHATWG URL instead of url.parse Co-authored-by: Shelley Vohr <[email protected]> * fix: OPENSSL_secure_zalloc doesn't work in BoringSSL Co-authored-by: Shelley Vohr <[email protected]> * chore: fix accidental extra line Co-authored-by: Shelley Vohr <[email protected]> * 7017517: [defer-import-eval] Parse import defer syntax https://chromium-review.googlesource.com/c/v8/v8/+/7017517 Co-authored-by: Shelley Vohr <[email protected]> --------- Co-authored-by: trop[bot] <37223003+trop[bot]@users.noreply.github.com> Co-authored-by: Shelley Vohr <[email protected]>

anonrig added dont-land-on-v18.x dont-land-on-v20.x PRs that should not land on the v20.x-staging branch and should not be released in v20.x. labels Sep 10, 2024

anonrig requested review from RafaelGSS, addaleax and jasnell September 10, 2024 14:29

nodejs-github-bot added c++ Issues and PRs that require attention from people who are familiar with C++. needs-ci PRs that need a full CI run. labels Sep 10, 2024

anonrig force-pushed the speed-up-utf8value branch from 36e1983 to def9c45 Compare September 10, 2024 14:31

anonrig added the performance Issues and PRs related to the performance of Node.js. label Sep 10, 2024

anonrig force-pushed the speed-up-utf8value branch from def9c45 to 34b530b Compare September 10, 2024 14:35

anonrig requested a review from lemire September 10, 2024 14:35

lemire approved these changes Sep 10, 2024

View reviewed changes

anonrig added the request-ci Add this label to start a Jenkins CI on a PR. label Sep 10, 2024

github-actions bot added request-ci-failed An error occurred while starting CI via request-ci label, and manual interventon is needed. and removed request-ci Add this label to start a Jenkins CI on a PR. labels Sep 10, 2024

lemire reviewed Sep 10, 2024

View reviewed changes

src/util.cc Show resolved Hide resolved

addaleax reviewed Sep 10, 2024

View reviewed changes

src/util.cc Outdated Show resolved Hide resolved

RafaelGSS reviewed Sep 10, 2024

View reviewed changes

src/util.cc Outdated Show resolved Hide resolved

src/util.cc Outdated Show resolved Hide resolved

anonrig force-pushed the speed-up-utf8value branch 4 times, most recently from 7556dc8 to 70263e8 Compare September 11, 2024 15:22

mcollina approved these changes Sep 11, 2024

View reviewed changes

mcollina added the request-ci Add this label to start a Jenkins CI on a PR. label Sep 11, 2024

anonrig requested review from RafaelGSS and addaleax September 12, 2024 15:31

ronag previously requested changes Sep 12, 2024

View reviewed changes

src/util.cc Outdated Show resolved Hide resolved

Update src/util.cc

cb36b86

Co-authored-by: Robert Nagy <[email protected]>

anonrig requested a review from ronag September 12, 2024 15:57

ronag reviewed Sep 12, 2024

View reviewed changes

src/util.cc Outdated Show resolved Hide resolved

Update src/util.cc

4b28cf1

Co-authored-by: Robert Nagy <[email protected]>

Qard approved these changes Sep 16, 2024

View reviewed changes

anonrig added commit-queue Add this label to land a pull request using GitHub Actions. commit-queue-squash Add this label to instruct the Commit Queue to squash all the PR commits into the first one. labels Sep 17, 2024

nodejs-github-bot removed the commit-queue Add this label to land a pull request using GitHub Actions. label Sep 17, 2024

nodejs-github-bot merged commit 8191e1f into nodejs:main Sep 17, 2024

targos added the dont-land-on-v22.x PRs that should not land on the v22.x-staging branch and should not be released in v22.x. label Oct 4, 2024

aduh95 mentioned this pull request Oct 14, 2024

2024-10-16, Version 23.0.0 (Current) #55338

Merged

codebytere added a commit to electron/electron that referenced this pull request Oct 21, 2025

src: improve utf8 string generation performance

18eaa20

nodejs/node#54873

codebytere added a commit to electron/electron that referenced this pull request Oct 28, 2025

src: improve utf8 string generation performance

6255ad9

nodejs/node#54873

codebytere added a commit to electron/electron that referenced this pull request Oct 30, 2025

src: improve utf8 string generation performance

1cf77b4

nodejs/node#54873

trop bot added a commit to electron/electron that referenced this pull request Oct 30, 2025

src: improve utf8 string generation performance

a0e8a3c

nodejs/node#54873 Co-authored-by: Shelley Vohr <[email protected]>

Uh oh!

src: improve utf8 string generation performance #54873

src: improve utf8 string generation performance #54873

Conversation

anonrig commented Sep 10, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions bot commented Sep 10, 2024

Uh oh!

jasnell commented Sep 10, 2024

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

RafaelGSS commented Sep 10, 2024

Uh oh!

codecov bot commented Sep 10, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

anonrig commented Sep 10, 2024

Uh oh!

RafaelGSS commented Sep 10, 2024

Uh oh!

mcollina left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

nodejs-github-bot commented Sep 15, 2024

Uh oh!

nodejs-github-bot commented Sep 15, 2024

Uh oh!

nodejs-github-bot commented Sep 15, 2024

Uh oh!

anonrig commented Sep 15, 2024

Uh oh!

nodejs-github-bot commented Sep 16, 2024

Uh oh!

nodejs-github-bot commented Sep 17, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

10 participants

anonrig commented Sep 10, 2024 •

edited

Loading

codecov bot commented Sep 10, 2024 •

edited

Loading