Compare primitive values with RowFn

[connortsui20]

Summary

• Tracking Issue: Epic: Row-oriented scalar functions #9128
• Replaces Compare primitive values with RowFn #9587

Moves primitive comparisons onto RowFn and removes the handwritten primitive lane-kernel fallback.

Changes

PrimitiveCompare dispatches every primitive type, comparison operator, and constant orientation through multiversioned RowVisitor::visit_bool. Cross-generated optimized x86-64 IR confirms vectorized row comparisons, but does not establish end-to-end benchmark parity.

CodSpeed compares this layer directly against develop. #9694 adds canonical validity attachment above it.

[codspeed-hq]

Merging this PR will degrade performance by 8.07%

⚠️ 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.

⚠️ 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

⚡ 9 improved benchmarks
❌ 24 regressed benchmarks
✅ 2057 untouched benchmarks
⏩ 206 skipped benchmarks¹
🗄️ 4 archived benchmarks run²

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	WallTime	lt_i64_nullable_avx2	2.9 µs	4.4 µs	-32.84%
❌	WallTime	lt_i64_nullable_avx512	2.9 µs	4.4 µs	-32.56%
❌	WallTime	lt_i64_nullable_neon	3.5 µs	5.2 µs	-32.44%
❌	WallTime	compare_int_nullable_avx512	3.7 µs	5 µs	-27.22%
❌	Simulation	compact_sliced[(2048, 10)]	183.5 µs	251.4 µs	-27.01%
❌	WallTime	compare_int_nullable_avx2	3.9 µs	5.1 µs	-24.02%
❌	Simulation	compact_sliced[(1024, 10)]	115.3 µs	149.9 µs	-23.08%
❌	WallTime	compare_int_nullable_neon	5 µs	6.2 µs	-18.62%
❌	Simulation	int_gt[16]	96.5 µs	117.6 µs	-18%
❌	Simulation	or_chain[16]	1.1 ms	1.4 ms	-16.32%
❌	Simulation	int_gt[1024]	103.5 µs	122.5 µs	-15.56%
❌	Simulation	float_gt[1024]	259.9 µs	306.3 µs	-15.16%
❌	Simulation	or_chain[1024]	1.2 ms	1.4 ms	-15.14%
❌	Simulation	float_gt[16]	237.9 µs	280.2 µs	-15.09%
❌	Simulation	baseline_lt[4, 1024]	81.3 µs	93.7 µs	-13.19%
❌	Simulation	is_not_null_pred[1024]	205.6 µs	236.3 µs	-13%
❌	Simulation	is_not_null_pred[16]	196.9 µs	223.8 µs	-12%
❌	Simulation	baseline_eq[4, 1024]	80.8 µs	91.3 µs	-11.48%
❌	Simulation	int_gt[8192]	144.9 µs	163.4 µs	-11.31%
❌	Simulation	baseline_eq[16, 1024]	82.2 µs	92.5 µs	-11.16%
...	...	...	...	...	...

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 ct/primitive-comparison-simd (d7f1683) with develop (97953be)}

Footnotes

1. 206 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. ↩

2. 4 benchmarks were run, but are now archived. If they were deleted in another branch, consider rebasing to remove them from the report. Instead if they were added back, click here to restore them. ↩