Fix/benchmark bugs

[rodrigobnogueira]

What do these changes do?

Fixed two logic bugs in benchmark.py that prevented accurate performance measurements:

1. Cache Miss Benchmark: test_cache_miss_benchmark was previously improperly configured with only 128 distinct objects (matching the cache size), resulting in cache hits instead of misses. Increased the sample size to 2048 objects to ensure evictions and actual misses are measured.
2. Concurrent Hit Benchmark: test_concurrent_cache_hit_benchmark contained a bug where the loop returned after the first iteration. Removed the premature return to ensure the benchmark runs for the intended 10 iterations.

Are there changes in behavior for the user?

No. These changes only affect the internal benchmark suite used for performance verification.

Related issue number

N/A

Checklist

• I think the code is well written
• Unit tests for the changes exist
• Documentation reflects the changes

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.37%. Comparing base (4232c3f) to head (ba2c2a2).
⚠️ Report is 1 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #733   +/-   ##
=======================================
  Coverage   97.37%   97.37%           
=======================================
  Files          15       15           
  Lines         991      991           
  Branches       56       56           
=======================================
  Hits          965      965           
  Misses         23       23           
  Partials        3        3           

Flag	Coverage Δ
unit	96.26% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[codspeed-hq]

Merging this PR will degrade performance by 99.22%

❌ 16 regressed benchmarks
✅ 47 untouched benchmarks
⏩ 4 skipped benchmarks¹

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

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
❌	test_concurrent_cache_hit_benchmark[meth-bounded-ttl]	9.1 ms	85.9 ms	-89.41%
❌	test_cache_miss_benchmark[func-unbounded-ttl]	11.2 ms	173.4 ms	-93.54%
❌	test_cache_miss_benchmark[func-unbounded]	1.1 ms	144.2 ms	-99.21%
❌	test_cache_miss_benchmark[func-bounded-ttl]	11 ms	173.3 ms	-93.65%
❌	test_concurrent_cache_hit_benchmark[meth-unbounded-ttl]	10.9 ms	98.2 ms	-88.87%
❌	test_concurrent_cache_hit_benchmark[meth-bounded]	3.3 ms	30.7 ms	-89.33%
❌	test_cache_miss_benchmark[meth-bounded-ttl]	11.4 ms	179.1 ms	-93.65%
❌	test_concurrent_cache_hit_benchmark[meth-unbounded]	3.3 ms	30.7 ms	-89.3%
❌	test_cache_miss_benchmark[meth-bounded]	1.5 ms	150.1 ms	-99.02%
❌	test_cache_miss_benchmark[func-bounded]	1.1 ms	144.3 ms	-99.22%
❌	test_cache_miss_benchmark[meth-unbounded-ttl]	11.6 ms	179.2 ms	-93.54%
❌	test_cache_miss_benchmark[meth-unbounded]	1.5 ms	150.2 ms	-99.02%
❌	test_concurrent_cache_hit_benchmark[func-bounded-ttl]	8.5 ms	79.8 ms	-89.41%
❌	test_concurrent_cache_hit_benchmark[func-bounded]	2.8 ms	25.5 ms	-89.2%
❌	test_concurrent_cache_hit_benchmark[func-unbounded-ttl]	10.3 ms	91.9 ms	-88.75%
❌	test_concurrent_cache_hit_benchmark[func-unbounded]	2.7 ms	25.5 ms	-89.25%

---

_{Comparing rodrigobnogueira:fix/benchmark-bugs (ba2c2a2) with master (4232c3f)}

Footnotes

1. 4 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. ↩

[rodrigobnogueira]

there are some opportunities of refactoring in the benchmark code. But I've kept the changes to the minimum necessary to fix the issues.