Execute owned RowFn outputs over valid rows

[connortsui20]

Rationale for this change

• Stacked on: Execute spatial predicates with RowFn #9350
• Tracking issue: Tracking Issue: Execute RowFn over Vortex arrays #9130
• Epic: Epic: Row-oriented scalar functions #9128

Adds direct valid-row execution for owned RowFn outputs before the first nullable deferred owned consumer.

What changes are included in this PR?

Adds Default to OutputElement and uses it only to initialize skipped output positions. Batch execution masks those positions before returning the array.

Owned infallible and deferred visits now execute directly over null-tolerant decoded inputs. Unsupported input encodings can still use the filter-and-scatter fallback from the spatial-distance layer.

What APIs are changed? Are there any user-facing changes?

Extends the unstable OutputElement contract with Default. There are no stable API changes.

[codspeed-hq]

Merging this PR will degrade performance by 1.51%

⚠️ 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.

⚡ 7 improved benchmarks
❌ 6 regressed benchmarks
✅ 2001 untouched benchmarks
⏩ 517 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	compact_sliced[(2048, 10)]	207.7 µs	278.3 µs	-25.34%
❌	Simulation	compact_sliced[(1024, 10)]	127.9 µs	163.9 µs	-21.95%
❌	Simulation	take[small_m/shuffled/primitive/nonnull/chunks=256/indices=64]	517.4 µs	598.3 µs	-13.53%
❌	Simulation	take[small_m/shuffled/primitive/nonnull/chunks=2048/indices=64]	685.8 µs	781 µs	-12.19%
❌	Simulation	cold_misaligned[(16, 64)]	376.6 µs	425 µs	-11.39%
❌	Simulation	take[small_m/shuffled/primitive/nonnull/chunks=256/indices=16]	227.4 µs	254.3 µs	-10.58%
⚡	Simulation	decompress[u64, (4000, 1024)]	103.6 µs	84.4 µs	+22.73%
⚡	WallTime	words_gather_scalar[65536]	9.4 µs	8.3 µs	+13.38%
⚡	Simulation	null_count_run_end[(10000, 1024, 0.01)]	8 µs	7.2 µs	+11.01%
⚡	Simulation	null_count_run_end[(10000, 256, 0.01)]	8 µs	7.2 µs	+11.01%
⚡	Simulation	null_count_run_end[(32000, 1024, 0.01)]	8 µs	7.2 µs	+11.01%
⚡	Simulation	null_count_run_end[(32000, 256, 0.01)]	8 µs	7.2 µs	+11.01%
⚡	Simulation	compress_fsst[(500, 64, 8)]	580.2 µs	523.8 µs	+10.77%

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-owned-valid-rows (d90f471) with ct/row-fn-spatial-predicates (e0e3fe7)}

Footnotes

1. 517 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. ↩

[connortsui20]

GitHub marked this PR merged while I reordered the native stack because the rewritten head became an ancestor of its old base. The same layer continues in #9500, now directly above #9450.