GeezerVison

Copy link

Copy Markdown

Member

merceyz, bpasero, ShenHongFei, exoego, robpalme, dubzzz, Bnaya, lin72h, btea, optionsx, and 5 more reacted with thumbs up emoji ShenHongFei, Mike-Dax, exoego, morgante, lin72h, optionsx, XinChou16, and RareScrap reacted with rocket emoji

avivkeller reviewed Aug 22, 2024

View reviewed changes

benjamingr reviewed Aug 22, 2024

View reviewed changes

benjamingr reviewed Aug 22, 2024

View reviewed changes

benjamingr reviewed Aug 22, 2024

View reviewed changes

benjamingr approved these changes Aug 22, 2024

View reviewed changes

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

Benchmark 1: ~/work/node/out/Release/node ./built/local/_tsc.js --version
  Time (mean ± σ):     120.5 ms ±   1.7 ms    [User: 108.6 ms, System: 11.2 ms]
  Range (min … max):   117.5 ms … 129.1 ms    100 runs
 
Benchmark 2: ~/work/node/out/Release/node ./built/local/tsc.js --version
  Time (mean ± σ):      51.8 ms ±   1.0 ms    [User: 42.3 ms, System: 9.0 ms]
  Range (min … max):    50.4 ms …  56.6 ms    100 runs
 
Summary
  '~/work/node/out/Release/node ./built/local/tsc.js --version' ran
    2.33 ± 0.06 times faster than '~/work/node/out/Release/node ./built/local/_tsc.js --version'

Benchmark 1: echo '{"seq": 1, "command": "status"}' | ~/work/node/out/Release/node ./built/local/_tsserver.js --disableAutomaticTypingAcquisition
  Time (mean ± σ):     192.3 ms ±   2.5 ms    [User: 178.3 ms, System: 19.8 ms]
  Range (min … max):   188.1 ms … 201.1 ms    100 runs
 
Benchmark 2: echo '{"seq": 1, "command": "status"}' | ~/work/node/out/Release/node ./built/local/tsserver.js --disableAutomaticTypingAcquisition
  Time (mean ± σ):      84.6 ms ±   1.1 ms    [User: 76.5 ms, System: 14.5 ms]
  Range (min … max):    82.6 ms …  88.6 ms    100 runs
 
Summary
  'echo '{"seq": 1, "command": "status"}' | ~/work/node/out/Release/node ./built/local/tsserver.js --disableAutomaticTypingAcquisition' ran
    2.27 ± 0.04 times faster than 'echo '{"seq": 1, "command": "status"}' | ~/work/node/out/Release/node ./built/local/_tsserver.js --disableAutomaticTypingAcquisition'

@@           Coverage Diff            @@
##             main   #54501    +/-   ##
========================================
  Coverage   87.34%   87.34%            
========================================
  Files         649      649            
  Lines      182544   182666   +122     
  Branches    35030    35042    +12     
========================================
+ Hits       159445   159555   +110     
- Misses      16372    16387    +15     
+ Partials     6727     6724     -3     

Benchmark 1: NODE_DISABLE_COMPILE_CACHE=true ~/work/node/out/Release/node ./node_modules/eslint/bin/eslint.js --version
  Time (mean ± σ):     129.9 ms ±   7.8 ms    [User: 128.6 ms, System: 17.4 ms]
  Range (min … max):   126.1 ms … 204.6 ms    100 runs
 
  Warning: Statistical outliers were detected. Consider re-running this benchmark on a quiet PC without any interferences from other programs. It might help to use the '--warmup' or '--prepare' options.
 
Benchmark 2: ~/work/node/out/Release/node ./node_modules/eslint/bin/eslint.js --version
  Time (mean ± σ):     105.5 ms ±   1.8 ms    [User: 103.3 ms, System: 18.6 ms]
  Range (min … max):   101.7 ms … 114.5 ms    100 runs
 
Summary
  '~/work/node/out/Release/node ./node_modules/eslint/bin/eslint.js --version' ran
    1.23 ± 0.08 times faster than 'NODE_DISABLE_COMPILE_CACHE=true ~/work/node/out/Release/node ./node_modules/eslint/bin/eslint.js --version'

Benchmark 1: NODE_DISABLE_COMPILE_CACHE=true ~/work/node/out/Release/node ./node_modules/eslint/bin/eslint.js --print-config ./src/compiler/debug.ts
  Time (mean ± σ):     791.6 ms ±   6.2 ms    [User: 987.2 ms, System: 86.3 ms]
  Range (min … max):   779.1 ms … 811.3 ms    100 runs
 
Benchmark 2: ~/work/node/out/Release/node ./node_modules/eslint/bin/eslint.js --print-config ./src/compiler/debug.ts
  Time (mean ± σ):     595.0 ms ±   6.4 ms    [User: 786.4 ms, System: 89.5 ms]
  Range (min … max):   583.7 ms … 621.3 ms    100 runs
 
Summary
  '~/work/node/out/Release/node ./node_modules/eslint/bin/eslint.js --print-config ./src/compiler/debug.ts' ran
    1.33 ± 0.02 times faster than 'NODE_DISABLE_COMPILE_CACHE=true ~/work/node/out/Release/node ./node_modules/eslint/bin/eslint.js --print-config ./src/compiler/debug.ts'

aduh95 reviewed Aug 23, 2024

View reviewed changes

aduh95 reviewed Aug 23, 2024

View reviewed changes

aduh95 reviewed Aug 23, 2024

View reviewed changes

aduh95 reviewed Aug 23, 2024

View reviewed changes

aduh95 reviewed Aug 23, 2024

View reviewed changes

aduh95 reviewed Aug 23, 2024

View reviewed changes

jasnell approved these changes Aug 23, 2024

View reviewed changes

This patch adds the following API for tools to enable compile
cache dynamically and query its status.

- module.enableCompileCache(cacheDir)
- module.getCompileCacheDir()

In addition this adds a NODE_DISABLE_COMPILE_CACHE environment
variable to disable the code cache enabled by the APIs as
an escape hatch to avoid unexpected/undesired effects of
the compile cache (e.g. less precise test coverage).

When the module.enableCompileCache() method is invoked without
a specified directory, Node.js will use the value of
the NODE_COMPILE_CACHE environment variable if it's set, or
defaults to `path.join(os.tmpdir(), 'node-compile-cache')`
otherwise. Therefore it's recommended for tools to call this
method without specifying the directory to allow overrides.

PR-URL: #54501
Fixes: #53639
Reviewed-By: Benjamin Gruenbaum <[email protected]>
Reviewed-By: James M Snell <[email protected]>
Reviewed-By: Antoine du Hamel <[email protected]>

Notable changes:

net:
  * (SEMVER-MINOR) exclude ipv6 loopback addresses from server.listen (Giovanni Bucci) #54264
src:
  * (SEMVER-MINOR) add JS APIs for compile cache and NODE_DISABLE_COMPILE_CACHE (Joyee Cheung) #54501
src,lib:
  * (SEMVER-MINOR) add performance.uvMetricsInfo (Rafael Gonzaga) #54413
test_runner:
  * (SEMVER-MINOR) add support for coverage thresholds (Aviv Keller) #54429
  * (SEMVER-MINOR) support running tests in process (Colin Ihrig) #53927
  * (SEMVER-MINOR) defer inheriting hooks until run() (Colin Ihrig) #53927
vm:
  * (SEMVER-MINOR) introduce vanilla contexts via vm.constants.DONT_CONTEXTIFY (Joyee Cheung) #54394

PR-URL: #54560

Notable changes:

net:
  * (SEMVER-MINOR) exclude ipv6 loopback addresses from server.listen (Giovanni Bucci) #54264
src:
  * (SEMVER-MINOR) add JS APIs for compile cache and NODE_DISABLE_COMPILE_CACHE (Joyee Cheung) #54501
src,lib:
  * (SEMVER-MINOR) add performance.uvMetricsInfo (Rafael Gonzaga) #54413
test_runner:
  * (SEMVER-MINOR) add support for coverage thresholds (Aviv Keller) #54429
  * (SEMVER-MINOR) support running tests in process (Colin Ihrig) #53927
  * (SEMVER-MINOR) defer inheriting hooks until run() (Colin Ihrig) #53927
vm:
  * (SEMVER-MINOR) introduce vanilla contexts via vm.constants.DONT_CONTEXTIFY (Joyee Cheung) #54394

PR-URL: #54560

Notable changes:

net:
  * (SEMVER-MINOR) exclude ipv6 loopback addresses from server.listen (Giovanni Bucci) #54264
src:
  * (SEMVER-MINOR) add JS APIs for compile cache and NODE_DISABLE_COMPILE_CACHE (Joyee Cheung) #54501
src,lib:
  * (SEMVER-MINOR) add performance.uvMetricsInfo (Rafael Gonzaga) #54413
test_runner:
  * (SEMVER-MINOR) add support for coverage thresholds (Aviv Keller) #54429
  * (SEMVER-MINOR) support running tests in process (Colin Ihrig) #53927
  * (SEMVER-MINOR) defer inheriting hooks until run() (Colin Ihrig) #53927
vm:
  * (SEMVER-MINOR) introduce vanilla contexts via vm.constants.DONT_CONTEXTIFY (Joyee Cheung) #54394

PR-URL: #54560

Notable changes:

net:
  * (SEMVER-MINOR) exclude ipv6 loopback addresses from server.listen (Giovanni Bucci) #54264
src:
  * (SEMVER-MINOR) add JS APIs for compile cache and NODE_DISABLE_COMPILE_CACHE (Joyee Cheung) #54501
src,lib:
  * (SEMVER-MINOR) add performance.uvMetricsInfo (Rafael Gonzaga) #54413
test_runner:
  * (SEMVER-MINOR) add support for coverage thresholds (Aviv Keller) #54429
  * (SEMVER-MINOR) support running tests in process (Colin Ihrig) #53927
  * (SEMVER-MINOR) defer inheriting hooks until run() (Colin Ihrig) #53927
vm:
  * (SEMVER-MINOR) introduce vanilla contexts via vm.constants.DONT_CONTEXTIFY (Joyee Cheung) #54394

PR-URL: #54560

Notable changes:

net:
  * (SEMVER-MINOR) exclude ipv6 loopback addresses from server.listen (Giovanni Bucci) #54264
src:
  * (SEMVER-MINOR) add JS APIs for compile cache and NODE_DISABLE_COMPILE_CACHE (Joyee Cheung) #54501
src,lib:
  * (SEMVER-MINOR) add performance.uvMetricsInfo (Rafael Gonzaga) #54413
test_runner:
  * (SEMVER-MINOR) add support for coverage thresholds (Aviv Keller) #54429
  * (SEMVER-MINOR) support running tests in process (Colin Ihrig) #53927
  * (SEMVER-MINOR) defer inheriting hooks until run() (Colin Ihrig) #53927
vm:
  * (SEMVER-MINOR) introduce vanilla contexts via vm.constants.DONT_CONTEXTIFY (Joyee Cheung) #54394

PR-URL: #54560

Notable changes:

net:
  * (SEMVER-MINOR) exclude ipv6 loopback addresses from server.listen (Giovanni Bucci) #54264
src:
  * (SEMVER-MINOR) add JS APIs for compile cache and NODE_DISABLE_COMPILE_CACHE (Joyee Cheung) #54501
src,lib:
  * (SEMVER-MINOR) add performance.uvMetricsInfo (Rafael Gonzaga) #54413
test_runner:
  * (SEMVER-MINOR) add support for coverage thresholds (Aviv Keller) #54429
  * (SEMVER-MINOR) support running tests in process (Colin Ihrig) #53927
  * (SEMVER-MINOR) defer inheriting hooks until run() (Colin Ihrig) #53927
vm:
  * (SEMVER-MINOR) introduce vanilla contexts via vm.constants.DONT_CONTEXTIFY (Joyee Cheung) #54394

PR-URL: #54560

CharlesCowan reviewed Oct 10, 2024

View reviewed changes

Sorry, something went wrong.

Uh oh!

There was an error while loading. Please reload this page.

Adding support to
[module.enableCompileCache](https://nodejs.org/docs/v22.11.0/api/module.html#moduleenablecompilecachecachedir),
added by nodejs/node#54501, Typescript got [2.5x
perf on
tsc](https://devblogs.microsoft.com/typescript/announcing-typescript-5-7-beta/#support-for-v8-compile-caching-in-node.js).

On my machine, for the command `--version`, I got the following results:

```bash
Benchmark 1: node bin/npm-cli.js --version
  Time (mean ± σ):      58.0 ms ±   1.1 ms    [User: 51.8 ms, System: 16.8 ms]
  Range (min … max):    56.5 ms …  62.6 ms    51 runs
 
Benchmark 2: node bin/npm-cli-with-cache.js --version
  Time (mean ± σ):      52.9 ms ±   1.0 ms    [User: 46.5 ms, System: 16.8 ms]
  Range (min … max):    51.4 ms …  55.5 ms    56 runs
 
Summary
  'node bin/npm-cli-with-cache.js --version' ran
    1.10 ± 0.03 times faster than 'node bin/npm-cli.js --version'
```

I run the same test with `npm run empty` and I got the same 10%
improvement.