src: move import.meta initializer to native land

[joyeecheung]

When the source text module is compiled without custom callbacks, instead of calling into JS land from the per-isolate import.meta initializer and then back to C++ land to set up lazy data properties, just do the initialization all in C++ land. Only import.meta.resolve initialization will call back into JS land to generate a closure that call the cascaded loader for resolution.

In addition, simplify the loader structure by merging allowImportMetaResolve into isForAsyncLoaderHookWorker - the two are essentially equivalent, as import.meta.resolve is only allowed in a non-loader-hook worker thread's loader.

[nodejs-github-bot]

Review requested:

• @nodejs/loaders
• @nodejs/vm

[codecov]

Codecov Report

❌ Patch coverage is 75.58140% with 42 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.53%. Comparing base (e72761f) to head (5d486b7).
⚠️ Report is 10 commits behind head on main.

Files with missing lines	Patch %	Lines
src/module_wrap.cc	62.16%	21 Missing and 21 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #60603      +/-   ##
==========================================
- Coverage   88.57%   88.53%   -0.05%     
==========================================
  Files         704      703       -1     
  Lines      208087   208075      -12     
  Branches    40090    40088       -2     
==========================================
- Hits       184306   184212      -94     
- Misses      15849    15886      +37     
- Partials     7932     7977      +45     

Files with missing lines	Coverage Δ
lib/internal/modules/esm/hooks.js	86.05% <ø> (-0.12%)	⬇️
lib/internal/modules/esm/loader.js	96.96% <100.00%> (+0.10%)	⬆️
lib/internal/modules/esm/utils.js	100.00% <100.00%> (ø)
lib/internal/modules/package_json_reader.js	97.61% <100.00%> (ø)
src/module_wrap.h	52.94% <ø> (ø)
src/node_modules.cc	77.83% <ø> (+1.61%)	⬆️
src/module_wrap.cc	74.20% <62.16%> (-1.49%)	⬇️

... and 39 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.

[aduh95]

LGTM modulo the lint failure

[GeoffreyBooth]

I would expect that this should result in performance improvements, especially for a project with a lot of modules; should we run a benchmark?

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/70081/

[joyeecheung]

I would expect that this should result in performance improvements, especially for a project with a lot of modules

Yes...and no. If I modify the benchmarks like what I just pushed so that it entirely focuses on import.meta initialization, not the entire module loading, then there's an improvement:

                                                              confidence improvement accuracy (*)   (**)  (***)
esm/import-meta.js valuesToRead='dirname-and-filename' n=1000        ***      5.41 %       ±2.06% ±2.74% ±3.57%
esm/import-meta.js valuesToRead='dirname' n=1000                     ***      7.17 %       ±3.18% ±4.26% ±5.60%
esm/import-meta.js valuesToRead='filename' n=1000                    ***      9.60 %       ±2.31% ±3.08% ±4.01%
esm/import-meta.js valuesToRead='url' n=1000                         ***      9.78 %       ±1.76% ±2.35% ±3.07%

For the original version of the benchmark, which measures the entire performance of running import(...), there's no significant performance difference, since import.meta initialization isn't all that much work compared the entire process of resolving/loading/compiling/evaluating a module.

Nonetheless it seems useful to have a more focused benchmark, since we already have other benchmarks for measuring the entire loading process.

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/70082/

[joyeecheung]

Pushed the new benchmark. Can I get another review please? @GeoffreyBooth @aduh95