Add bit-packed cast benchmark

[joseph-isaacs]

Summary

• Add a cast_bitpacked divan benchmark for widening bit-packed u16 arrays to u32.
• Compare the public array.cast(u32).execute() path against an explicit canonicalize-then-cast baseline.
• Cover single and chunked inputs with and without patches.

Checks

• cargo +nightly fmt --all
• cargo bench -p vortex-fastlanes --features _test-harness --bench cast_bitpacked --no-run

[codspeed-hq]

Merging this PR will not alter performance

⚠️ 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 improved benchmarks
❌ 1 regressed benchmark
✅ 1234 untouched benchmarks
🆕 6 new benchmarks

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	baseline_eq[16, 65536]	287.6 µs	259.6 µs	+10.75%
⚡	Simulation	baseline_lt[16, 65536]	302.7 µs	274.8 µs	+10.16%
🆕	Simulation	cast_execute[(262144, 1, 0.0)]	N/A	1.9 ms	N/A
🆕	Simulation	cast_execute[(262144, 1, 0.01)]	N/A	1.9 ms	N/A
🆕	Simulation	cast_execute[(65536, 1, 0.0)]	N/A	517.6 µs	N/A
🆕	Simulation	cast_execute[(65536, 1, 0.01)]	N/A	525.8 µs	N/A
🆕	Simulation	cast_execute[(65536, 4, 0.0)]	N/A	2.8 ms	N/A
🆕	Simulation	cast_execute[(65536, 4, 0.01)]	N/A	2.8 ms	N/A
❌	Simulation	fast_lt_out_of_range[4, 65536]	204.3 µs	262.3 µs	-22.12%

Tip

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

---

_{Comparing codex/cast-bitpacked-benchmark (05a73b7) with develop (012d0ec)}