[not-matthias]

We do the same already in the other frameworks

Refs COD-4028

[greptile-apps]

Confidence Score: 4/5

[Medium impact] The change is limited to synchronous runs in the Benchmark.js plugin.

The PR appears safe to merge, with a non-blocking request for regression tests.

Findings

1.  GC ordering has no test ▶

Fix with agent prompt

### Issue 1
packages/benchmark.js-plugin/src/index.ts:210
The new collection has no regression test. Existing tests check that `startBenchmark` runs, but not whether `global.gc` runs before it. Removing this fix or moving collection into the measured window would still pass those tests.

Add a temporary `global.gc` mock and check that collection runs after warmup and before `startBenchmark`. Also cover missing GC and fallback execution, and restore the original global afterward.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

Summary

Adds forced garbage collection before synchronous Benchmark.js measurement, matching the asynchronous path.

• Synchronous Benchmark.js runs collect warmup garbage before instrumentation starts.

_{Reviews (1) · Last reviewed commit: "fix(benchmark.js): collect garbage befor..." · Reviewed by Greptile}

[greptile-apps]

GC ordering has no test

The new collection has no regression test. Existing tests check that startBenchmark runs, but not whether global.gc runs before it. Removing this fix or moving collection into the measured window would still pass those tests.

Add a temporary global.gc mock and check that collection runs after warmup and before startBenchmark. Also cover missing GC and fallback execution, and restore the original global afterward.

Prompt To Fix With AI

This is a comment left during a code review.
Path: packages/benchmark.js-plugin/src/index.ts
Line: 210

Comment:
**GC ordering has no test**

The new collection has no regression test. Existing tests check that `startBenchmark` runs, but not whether `global.gc` runs before it. Removing this fix or moving collection into the measured window would still pass those tests.

Add a temporary `global.gc` mock and check that collection runs after warmup and before `startBenchmark`. Also cover missing GC and fallback execution, and restore the original global afterward.

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

[codspeed]

Merging this PR will degrade performance by 3.68%

3 warnings, 1 note

⚠️ Walltime results may be unreliable

Some benchmarks used the Walltime instrument outside CodSpeed Macro Runners, which can make the results inconsistent.

Learn more about Macro Runners

See affected benchmarks

⚠️ 14 benchmarks are dominated by system calls

System calls cannot be simulated consistently and are left out of the measure. The reported value understates the real cost. Switch to the Walltime instrument to measure them.

Learn more about system calls

See affected benchmarks

⚠️ Base and head ran on different runner types

Some benchmarks with significant performance changes were compared across different runner types, so the changes might not come from the code. Run base and head on the same runner type, or use CodSpeed Macro Runners for more stable results.

Learn more about Macro Runners

Open the report in CodSpeed to investigate

ℹ️ 16 benchmarks have no flamegraph

The flamegraph wasn't generated: most of the time was spent in runtime internals, which would leave nothing useful to analyze.

See affected benchmarks

⚡ 9 improved benchmarks
❌ 8 regressed benchmarks
✅ 247 untouched benchmarks
⏩ 56 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	CPU Simulation	test_recursive_fibo_10	339.8 µs	691.8 µs	-50.88%
❌	CPU Simulation	wait 500ms	29.5 ms	52 ms	-43.22%
❌	CPU Simulation	wait 1sec	59 ms	93.8 ms	-37.06%
❌	CPU Simulation	test_recursive_cached_fibo_10	49.3 µs	65.4 µs	-24.71%
❌	CPU Simulation	test_recursive_cached_fibo_20	64.3 µs	80.1 µs	-19.72%
❌	CPU Simulation	test_recursive_cached_fibo_30	72 µs	87.8 µs	-17.94%
❌	Walltime	switch 2	84 ns	96 ns	-12.5%
❌	Walltime	switch 2	108 ns	120 ns	-10%
⚡	Memory	one	2 KB	1.3 KB	+48.24%
⚡	Memory	count is back to zero	2.1 KB	1.5 KB	+44.09%
⚡	Memory	wait 1sec	28.7 KB	22.1 KB	+29.49%
⚡	CPU Simulation	wait 500ms	38.7 ms	30.1 ms	+28.53%
⚡	Memory	wait 1sec	27.5 KB	22 KB	+24.96%
⚡	Walltime	test sync baz 10	144 ns	120 ns	+20%
⚡	CPU Simulation	wait 1sec	72.9 ms	61.3 ms	+19.02%
⚡	Walltime	test sync baz 100	252 ns	228 ns	+10.53%
⚡	Walltime	test_iterative_fibo_10	132 ns	120 ns	+10%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing cod-4028-use-predictable-gc-schedule-in-codspeed-node-for-memory (6925940) with main (727b808)}

Footnotes

1. 56 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩