Move execute-parent kernels into session registry

[gatesn]

Removes the ScalarFnVTable::execute_parent function in favor of purely session-registered kernels.

We already had some kernels on the session, and some on the scalar function. This unifies everything on the session.

I may well follow up by removing ScalarFnVTable::execute also in favor of these kernels.

[codspeed-hq]

Merging this PR will not alter performance

⚡ 12 improved benchmarks
❌ 24 regressed benchmarks
✅ 1507 untouched benchmarks
⏩ 39 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	chunked_bool_canonical_into[(1000, 10)]	20.3 µs	29.7 µs	-31.73%
❌	Simulation	chunked_dict_primitive_into_canonical[u32, (1000, 10, 10)]	180.2 µs	233.4 µs	-22.8%
❌	Simulation	encode_varbin[(1000, 2)]	234.6 µs	289 µs	-18.82%
❌	Simulation	canonicalize_sparse_fixed_size_list[(1024, 17, 8)]	439.2 µs	540.9 µs	-18.8%
❌	Simulation	search_index_above_max	255.9 µs	312.3 µs	-18.07%
❌	Simulation	search_index_below_min	255.9 µs	312.3 µs	-18.07%
❌	Simulation	search_index_mixed_out_of_range	256 µs	312.4 µs	-18.06%
❌	Simulation	search_index_full_range_random	256.1 µs	312.6 µs	-18.06%
❌	Simulation	search_index_in_range	256.3 µs	312.8 µs	-18.05%
❌	Simulation	canonicalize_sparse_list[(1024, 17, 8)]	480.4 µs	581.9 µs	-17.45%
❌	Simulation	chunked_varbinview_into_canonical[(1000, 10)]	177.7 µs	214.2 µs	-17.08%
❌	Simulation	search_index_below_min_chunked	1.3 ms	1.6 ms	-15.24%
❌	Simulation	search_index_mixed_out_of_range_chunked	1.3 ms	1.6 ms	-15.04%
❌	Simulation	search_index_full_range_random_chunked	1.4 ms	1.6 ms	-14.99%
❌	Simulation	canonicalize_sparse_fixed_size_list[(512, 7, 4)]	286 µs	334.8 µs	-14.59%
❌	Simulation	search_index_above_max_chunked	1.7 ms	2 ms	-14.07%
❌	Simulation	canonicalize_sparse_list[(512, 7, 4)]	307 µs	356.7 µs	-13.94%
❌	Simulation	search_index_in_range_chunked	2 ms	2.3 ms	-13.51%
❌	Simulation	chunked_bool_canonical_into[(10, 1000)]	751.5 µs	866.4 µs	-13.26%
❌	Simulation	chunked_varbinview_opt_canonical_into[(100, 100)]	359 µs	408.5 µs	-12.13%
...	...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

Tip

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

---

_{Comparing ngates/session-parent-kernels (afb594f) with develop (69ce1ed)}

Footnotes

1. 39 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. ↩