Attach RowFn validity directly to canonical output

[connortsui20]

Summary

• Tracking Issue: Epic: Row-oriented scalar functions #9128
• Depends on Decode RowFn constants directly #9680

Attaches input-derived validity directly when a RowFn kernel already returns a canonical array, avoiding an extra lazy vortex.mask execution. Non-canonical output keeps the existing lazy mask path.

Changes

Reuses mask_validity_canonical without copying canonical value buffers and preserves kernel-output validation, output relabeling, and outer nullability. Structural Boolean and primitive tests verify that nullable dense execution no longer runs vortex.mask.

[codspeed-hq]

Merging this PR will regress 4 benchmarks

⚠️ 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

⚡ 97 improved benchmarks
❌ 4 regressed benchmarks
✅ 2068 untouched benchmarks
⏩ 106 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	mul_u64_nonnull_neon	15.3 µs	20.7 µs	-26.2%
❌	Simulation	take[duplicates/repeated/primitive/nonnull/chunks=16/indices=1000]	201.4 µs	238 µs	-15.4%
❌	WallTime	multiply_shapes_neon[(16384, PerRowPerRow)]	17.4 µs	20.2 µs	-13.8%
❌	WallTime	mul_i64_nonnull_neon	17.4 µs	20 µs	-12.99%
⚡	WallTime	subtract_shapes_neon[(128, PerRowConstant)]	3.2 µs	1.9 µs	+65.2%
⚡	WallTime	infallible_bool_avx512[i32, ConstantPerRow]	2.9 µs	1.8 µs	+63.52%
⚡	WallTime	subtract_shapes_neon[(128, ConstantPerRow)]	3.2 µs	1.9 µs	+62.66%
⚡	WallTime	infallible_bool_avx512[i32, PerRowConstant]	2.8 µs	1.8 µs	+61.82%
⚡	WallTime	add_shapes_neon[(128, PerRowConstant)]	3.1 µs	2 µs	+59.86%
⚡	WallTime	multiply_shapes_neon[(128, PerRowConstant)]	3.3 µs	2 µs	+59.57%
⚡	WallTime	add_shapes_neon[(128, ConstantPerRow)]	3.1 µs	1.9 µs	+59.57%
⚡	WallTime	multiply_shapes_neon[(128, ConstantPerRow)]	3.2 µs	2 µs	+59.07%
⚡	WallTime	deferred_bool_avx512[i32, ConstantPerRow]	3 µs	1.9 µs	+57.76%
⚡	WallTime	deferred_bool_avx512[i32, PerRowConstant]	3 µs	1.9 µs	+57.37%
⚡	WallTime	subtract_shapes_neon[(128, PerRowNullableConstant)]	3.8 µs	2.5 µs	+55.76%
⚡	WallTime	multiply_shapes_neon[(128, PerRowNullableConstant)]	3.9 µs	2.6 µs	+53.73%
⚡	WallTime	add_shapes_avx2[(128, PerRowConstant)]	2.9 µs	1.9 µs	+53.36%
⚡	WallTime	subtract_shapes_avx512[(128, PerRowConstant)]	3 µs	1.9 µs	+53.3%
⚡	WallTime	subtract_shapes_avx2[(128, PerRowConstant)]	2.9 µs	1.9 µs	+53.17%
⚡	WallTime	add_shapes_neon[(128, PerRowNullableConstant)]	3.8 µs	2.5 µs	+52.89%
...	...	...	...	...	...

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/row-fn-canonical-validity (f17be3d) with develop (3a1838c)³}

Footnotes

1. 106 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. ↩

3. No successful run was found on develop (a745604) during the generation of this report, so 3a1838c was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

[robert3005]

generally not a fan of optimisations like these...

[connortsui20]

I think it shows a gap in our abstraction, ideally this is not required for this optimization?