[joseph-isaacs]

Summary

Aggregate accumulators look up AggregateFnSession on every batch and on every pass of their execute loop. The grouped accumulator does two lookups per pass. Each lookup loads the session map, does two TypeId lookups and then loads the registry. This PR snapshots the aggregate kernel registries into the ExecutionCtx on first use, the same way execute-parent kernels are already snapshotted, so those lookups only read a plain HashMap.

This follows the idea in #10541, which avoids an unused session lookup.

[codspeed]

Merging this PR will degrade performance by 13.1%

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚠️ 1 benchmark measured no execution time

Nothing ran under measurement, usually because the compiler removed the code under test. This result is not comparable, so it counts as unchanged.

Preventing compiler optimizations

• bench_compare_primitive[(10000, 2)]

⚠️ 19 benchmarks spent significant time in system calls

System calls cannot be consistently instrumented, so they are not included in the measure, which understates the real cost. Please switch to the Walltime instrument to accurately measure system calls.

Measurement and system calls

• random_i128[0.01]
• test_scan_to_arrow_array[int]
• test_scan_filter[int_float]
• test_repeated_scan_to_arrow_array[int_float]
• test_repeated_scan[int_float]
• test_repeated_scan_to_arrow_array[int_str]
• decompress[u32, (1000, 256)]
• decompress[u16, (4000, 256)]
• test_scan_filter[int_str]
• multipolygon_mixed_validity
• …and 9 more

ℹ️ 40 benchmarks have no flame graph

The flamegraph could not be generated because we failed to resolve the symbols in the profiling trace. This happens when we fail to harvest information about language frames.

• test_scan_filter[int]
• test_scan_to_arrow_array[int]
• test_scan_filter[int_float]
• test_repeated_scan_to_arrow_array[int_float]
• test_repeated_scan_scalar_at[int]
• test_repeated_scan[int_float]
• test_scan_scalar_at[int]
• test_scan_scalar_at[int_str]
• test_repeated_scan_to_arrow_array[int_str]
• test_repeated_scan_scalar_at[int_str_float]
• …and 30 more

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

❌ 6 regressed benchmarks
✅ 2402 untouched benchmarks
⏩ 12 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌ ⚠️	Simulation	random_i128[0.01]	4.8 µs	6 µs	-19.42%
❌	Simulation	random_i8[0.01]	101.3 µs	117 µs	-13.46%
❌	WallTime	starts_with_avx2[64, VarBin]	53.3 µs	60.4 µs	-11.75%
❌	WallTime	bitpack_blocked_compress_avx2	6.7 µs	7.6 µs	-11.5%
❌ ℹ️	Simulation	test_scan_filter[int]	3.1 ms	3.5 ms	-11.15%
❌	Simulation	isolated_runs[u32, (16, 0)]	78.5 µs	88.2 µs	-10.97%

Tip

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

---

_{Comparing ji/vibrant-ritchie-0ct678 (2a2cde3) with develop (5d8fa18)}

Footnotes

1. 12 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. ↩