refactor: pack and gather patches with explicit chunk widths

[mhk197]

Share packing and exception gathering between uniform and caller-supplied chunk layouts. The explicit encoder accepts independently chosen widths, including full-width and zero-width chunks, and persists only their byte boundaries in the offsets child. Global-width encoding entry points remain available.

Tests cover full-width chunks, exception gathering, zero-copy slicing with nonzero offset origins, and decoding the resulting layouts. This makes it possible for a later encoder to choose widths independently for each chunk.

Validation: 394 FastLanes/BtrBlocks tests passed (1 skipped).

[codspeed]

Merging this PR will improve performance by 12.2%

⚠️ 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.

⚡ 1 improved benchmark
✅ 221 untouched benchmarks
⏩ 2339 skipped benchmarks¹

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚡	filtered_owned_i64_avx512[OneNullInEight]	25.9 µs	23.1 µs	+12.2%

Tip

Curious why performance improved? Comment @codspeedbot explain why performance improved on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing mk/bitpacked-stack-05-explicit-packing (82c7619) with mk/bitpacked-stack-02-cpu-layout (797717f)}

Footnotes

1. 2339 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. ↩