Aggregating bounded min/max values has to not ignore null values

[robert3005]

Propagating bounded min/max is not the same as propagating nulls. In the case of
bounded min/max we have to conclude that null value could have come from trying
to produce the bounded value which means that there's no valid bounded value while there's still valid values in the array

handle #8166

[gatesn]

This seems like a very odd behaviour for truncated min/max to me, just leaving a request changes so I can read thoroughly on Monday.

[codspeed-hq]

Merging this PR will degrade performance by 15.76%

⚠️ 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.

❌ 2 regressed benchmarks
✅ 1273 untouched benchmarks

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	chunked_varbinview_opt_canonical_into[(1000, 10)]	187.7 µs	225 µs	-16.56%
❌	WallTime	cuda/bitpacked_u8/unpack/3bw[100M]	299.6 µs	352.3 µs	-14.94%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing rk/fuzzer (944755b) with develop (57e1784)}

[robert3005]

This comes from the fact that bounded min/max is not transitive if you exclude nulls. The problem we run into is when producing a bounded max returns null because the bound exceeds the limits we have defined which then is not included in the min/max of the file which then can lead to skipped values.

[robert3005]

talked to @joseph-isaacs. There's a simpler fix, when we overflow on max we store truncated_len + 1 field which would be bigger than any of the existing values this doesn't work, we need to increment

[joseph-isaacs]

what is the counterexample?

[robert3005]

take upper_bound([FF, FF, FF, 0], 2) then [FF, FF, FF] is not >= [FF, FF, FF, 0]

[joseph-isaacs]

But 0xFF > 0xF?

[robert3005]

yes but I don't follow

[joseph-isaacs]

Okay we need to store a flag to specify MAX_VALUE, such that for x. x <= MAX_VALUE