feat(buffer): add allocator-backed storage

[gatesn]

Summary

Add allocator-backed storage to vortex-buffer.

Design: vortex-data/rfcs#65

Changes

• Use allocator_api2 for owned buffers.
• Keep external buffers zero-copy.
• Preserve runtime alignment and bytes/Arrow interop.
• Add allocator-aware constructors and tests.

API Changes

Owned buffers no longer use bytes::Bytes as storage. The bytes crate remains for interop.

[codspeed-hq]

Merging this PR will degrade performance by 6.5%

⚠️ 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

⚡ 60 improved benchmarks
❌ 124 regressed benchmarks
✅ 2027 untouched benchmarks
⏩ 64 skipped benchmarks¹
🗄️ 5 archived benchmarks run²

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	decompress[datetime_for_bp]	220.5 µs	405.3 µs	-45.6%
❌	Simulation	new_bp_prim_test_between[i64, 2048]	89.9 µs	153.8 µs	-41.56%
❌	Simulation	new_alp_prim_test_between[f64, 2048]	106.7 µs	171.2 µs	-37.68%
❌	Simulation	decompress[alp_for_bp_f64]	146 µs	225.2 µs	-35.15%
❌	Simulation	decompress_rd[f64, (2000, 0.01)]	121.6 µs	187.3 µs	-35.07%
❌	Simulation	decompress_rd[f64, (2000, 0.1)]	121.7 µs	187.3 µs	-35.03%
❌	Simulation	decompress_rd[f64, (2000, 0.0)]	121.7 µs	187.2 µs	-35.01%
❌	Simulation	execute_scalar_struct_simple	111.1 µs	168.8 µs	-34.18%
❌	Simulation	decompress[u16, (4000, 256)]	37.5 µs	55.7 µs	-32.66%
❌	Simulation	chunked_opt_bool_into_canonical[(1000, 10)]	85.7 µs	120.8 µs	-29.04%
❌	Simulation	new_bp_prim_test_between[i64, 16384]	170.5 µs	236.8 µs	-27.99%
❌	Simulation	alp_rd_decompress_f64	169.6 µs	235.5 µs	-27.98%
❌	Simulation	compare[1]	211.1 µs	287.3 µs	-26.51%
❌	Simulation	new_alp_prim_test_between[f64, 16384]	193.3 µs	258.5 µs	-25.22%
❌	Simulation	compare[1]	207.2 µs	272.3 µs	-23.89%
❌	Simulation	compare[2]	213 µs	278.6 µs	-23.57%
❌	Simulation	compare[2]	216.1 µs	281.8 µs	-23.33%
❌	Simulation	compare[3]	220.4 µs	285.8 µs	-22.89%
❌	Simulation	compare[3]	223.4 µs	289.2 µs	-22.74%
❌	Simulation	compare[4]	225.9 µs	291.4 µs	-22.47%
...	...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

Tip

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

---

_{Comparing ngates/buffer-allocator-core (235c4c8) with develop (a745604)}

Footnotes

1. 64 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. ↩

2. 5 benchmarks were run, but are now archived. If they were deleted in another branch, consider rebasing to remove them from the report. Instead if they were added back, click here to restore them. ↩

[robert3005]

since I ran through experiments with this in the past - the Arc overhead here for small length dominates your allocations. tokio::Bytes plays game with pointer masking and stores AtomicPtr<()> and another pointer to avoid synchronistaion for trivial cases

[gatesn]

Yep... that's causing the slow downs I believe