Cut over sum aggregate function

[gatesn]

Move sum compute function over to the Sum impl of AggregatFnVTable

[edit] need to plumb through an ExecutionCtx first!

[codspeed-hq]

Merging this PR will degrade performance by 26.55%

❌ 14 regressed benchmarks
✅ 973 untouched benchmarks
⏩ 1512 skipped benchmarks¹

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	null_count_run_end[(10000, 1024, 0.1)]	39.4 µs	53.6 µs	-26.43%
❌	Simulation	null_count_run_end[(10000, 1024, 0.5)]	39.5 µs	53.8 µs	-26.55%
❌	Simulation	null_count_run_end[(10000, 256, 0.5)]	41.2 µs	55.3 µs	-25.42%
❌	Simulation	null_count_run_end[(10000, 4, 0.5)]	118.1 µs	132.2 µs	-10.64%
❌	Simulation	null_count_run_end[(10000, 4, 0.1)]	106.4 µs	120.6 µs	-11.74%
❌	Simulation	null_count_run_end[(10000, 16, 0.1)]	56.9 µs	70.9 µs	-19.81%
❌	Simulation	null_count_run_end[(10000, 16, 0.5)]	60.9 µs	75.2 µs	-19.03%
❌	Simulation	null_count_run_end[(10000, 256, 0.1)]	41.1 µs	55.4 µs	-25.75%
❌	Simulation	null_count_run_end[(100000, 1024, 0.1)]	64.7 µs	79 µs	-18.12%
❌	Simulation	null_count_run_end[(100000, 1024, 0.5)]	65.1 µs	79.4 µs	-18.04%
❌	Simulation	null_count_run_end[(100000, 256, 0.1)]	72.3 µs	86.7 µs	-16.57%
❌	Simulation	null_count_run_end[(100000, 256, 0.01)]	72 µs	86.3 µs	-16.62%
❌	Simulation	null_count_run_end[(100000, 256, 0.5)]	74.4 µs	88.7 µs	-16.12%
❌	Simulation	null_count_run_end[(10000, 16, 0.01)]	55.7 µs	69.9 µs	-20.24%

---

_{Comparing ngates/sum-agg (d284073) with develop (2a92893)}

Footnotes

1. 1512 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. ↩

[robert3005]

We should make sure we port any exhisting behaviour we have in the current implementations

[gatesn]

Like what?

[joseph-isaacs]

NaN handlings at least. Its work fuzzing this on the side is possible