fix(buffer): include the bit offset in BitBufferMut::from_buffer's bounds check

[jackylee-ch]

BitBuffer::new_with_offset (bit/buf.rs:123), BitBufferView::new (bit/view.rs:77) and
BitBufferView::slice (bit/view.rs:269) all assert
len.saturating_add(offset) <= buffer.len().saturating_mul(8). BitBufferMut::from_buffer
asserted only len <= buffer.len() * 8, so a caller could build a bit range extending past the
allocation and have value/set read or write outside it — freeze() would then panic pointing
at new_with_offset, a constructor the caller never called.

The offset parameter arrived in #4952, two days after #4937 added this assertion without one.

The field doc claimed the offset is "always less than 8"; that holds for BitBuffer, which
normalises via offset / 8, but not here — BitPackedArray passes its u16 chunk offset
(bitpacking/compute/compare_fused.rs:122), so the doc is corrected rather than enforced.

Tests

cargo test --release -p vortex-buffer --lib: 880 passed, 871 before. -p vortex-fastlanes
(319 + 4) and -p vortex-array --lib (3454, two pre-existing arrays::listview failures) are
unchanged, which is what covers the two in-tree callers. Restoring the old assertion fails 4 of
the 9 new cases; the accepting cases pass either way.

AI assistance

Written with agentic AI assistance; I read all four assertions and both callers before changing
one.

[codspeed]

Merging this PR will regress 3 benchmarks

⚠️ 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

⚡ 6 improved benchmarks
❌ 3 regressed benchmarks
✅ 2189 untouched benchmarks
⏩ 218 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	random_i8[0.5]	70.2 µs	93.8 µs	-25.14%
❌	Simulation	decompress[u64, (4000, 1024)]	70.5 µs	87.5 µs	-19.38%
❌	WallTime	words_gather_scalar_avx2[65536]	8.3 µs	9.4 µs	-11.96%
⚡	WallTime	dict_canonicalize_gt_u8_avx2[16000000]	10.8 ms	6.8 ms	+59.93%
⚡	WallTime	arrow_checked_add_u32_neon[16384]	20.4 µs	13.4 µs	+52.6%
⚡	WallTime	filtered_owned_i64_avx2[OneNullInEight]	27.1 µs	21.9 µs	+23.81%
⚡	Simulation	random_i16[0.8]	95.4 µs	77.2 µs	+23.52%
⚡	WallTime	filtered_owned_i64_avx512[OneNullInEight]	26.7 µs	22.6 µs	+17.87%
⚡	Simulation	allocate_drop_bytes[0]	575.7 ns	521.6 ns	+10.39%

Tip

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

---

_{Comparing jackylee-ch:fix/bitbuffermut-offset-bound (c711221) with develop (6465462)}

Footnotes

1. 218 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. ↩