Implement RowFn row execution

[connortsui20]

Rationale for this change

• Tracking issue: Tracking Issue: Execute RowFn over Vortex arrays #9130
• API tracking: Tracking Issue: Define the RowFn API #9129
• Epic: Epic: Row-oriented scalar functions #9128

Implements typed row execution for the contracts in #9386. Batch planning and orchestration are split into #9450.

What changes are included in this PR?

Adds owned-output and sink-writing row loops, prepared and deferred execution, the ViewLen contract, and decoding validation. This layer does not yet connect the row loops to execute_rows.

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

Extends the existing unstable_row_fns API. There are no stable API changes.

[codspeed-hq]

Merging this PR will degrade performance by 12.34%

⚠️ 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

❌ 7 regressed benchmarks
✅ 1993 untouched benchmarks
⏩ 89 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	take[small_m/shuffled/primitive/nonnull/chunks=1024/indices=1000]	5.7 ms	6.6 ms	-14.2%
❌	Simulation	take[small_m/shuffled/primitive/nonnull/chunks=1024/indices=256]	2.2 ms	2.5 ms	-13.6%
❌	Simulation	take[small_m/shuffled/primitive/nonnull/chunks=16384/indices=1000]	9.6 ms	11 ms	-12.95%
❌	Simulation	compress_fsst[(1000, 64, 8)]	1 ms	1.2 ms	-12.2%
❌	Simulation	take[core/shuffled/primitive/nonnull/chunks=1024/indices=10000]	14.3 ms	16.2 ms	-11.87%
❌	Simulation	take[small_m/shuffled/primitive/nonnull/chunks=1024/indices=16]	278.6 µs	312.8 µs	-10.94%
❌	Simulation	take[small_m/shuffled/primitive/nonnull/chunks=16384/indices=256]	3.2 ms	3.5 ms	-10.54%

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-framework (d65a9da) with develop (42a72d9)}

Footnotes

1. 89 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. ↩

[joseph-isaacs]

this is from a bad InputElement impl only right?

[connortsui20]

yes, but it also helps LLVM know that the loop bound is correct (see owned.rs). I guess I can try and see if removing it changes anything but obviously it is not going to improve perf in any way, and it makes more sense to be defensive here imo

[connortsui20]

actually I also want to check if we bound it by the new ViewLen trait I added if that changes the codegen, if not I will change it

[connortsui20]

ok I changed it to be an equality check. We definitely still need this check though

[joseph-isaacs]

LG