feat(benchmark): support profiling

[avivkeller]

Hiya! @ovflowd mentioned to me that y'all are looking to speed up parts of Webpack. That's great, and I'm happy to help. Anyhow, I wanted to help, so I figured I'd start by looking at the benchmarks, and immediately noticed a few things I could improve: Benchmarks don't profile

What kind of change does this PR introduce?

Refactoring

Did you add tests for your changes?

The tests are the benchmarks

Does this PR introduce a breaking change?

I certainly hope not.

What needs to be documented once your changes are merged?

The documentation has already been updated

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: avivkeller / name: Aviv Keller (f7858a5, e91bbe4, 0e10311)

[alexander-akait]

Let's do changes step by step, otherwise diff are dirty and hard to read and understand

1. Let's keep benchmarks in test directory right (we can move them at the last step)
2. We keep generated code in ./test/js/<name>, so let's store generated code there (we want to rename this directory to test/outputs/<name> soon)
3. Let's add profiling and other feature to BenchmarkTestCases.benchmark.mjs (and not only) in other PR, because I can't see diff due to moving files to other directory and it is hard to understand what we changed

[avivkeller]

Sorry! Indeed, I'll simplify this to only add profiling support and such

[alexander-akait]

Also we are trying to using jest-worker to speedup our benchmarks - #19780 make sense to look at this later

[alexander-akait]

Looks good

[codspeed-hq]

CodSpeed Performance Report

Merging #19789 will degrade performances by 75.03%

_{Comparing avivkeller:benchmark-changes (f7858a5) with main (1887f2c)}

Summary

❌ 2 regressions
✅ 40 untouched benchmarks

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Benchmarks breakdown

	Benchmark	BASE	HEAD	Change
❌	benchmark "devtool-eval-source-map", scenario '{"name":"mode-development-rebuild","mode":"development","watch":true}'	70.3 ms	116.5 ms	-39.64%
❌	benchmark "many-chunks-esm", scenario '{"name":"mode-development-rebuild","mode":"development","watch":true}'	29.3 ms	117.4 ms	-75.03%

[alexander-akait]

@avivkeller As you can see, our watch tests are not stable (not related to your changes), if you have time to look at what's wrong with them will be great