perf: remove unneeded scalar_at and valid_at and use arrays instead #8307
+85
−41
CodSpeed HQ / CodSpeed Performance Analysis
failed
Jun 9, 2026 in 0s
2 benchmarks regressed
⚠️ 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.
⚡ 13 improved benchmarks
❌ 2 regressed benchmarks
✅ 1508 untouched benchmarks
Warning
Please fix the performance issues or acknowledge them on CodSpeed.
Performance Changes
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ❌ | Simulation | decompress_rd[f64, (100000, 0.0)] |
845.9 µs | 1,024.1 µs | -17.4% |
| ❌ | Simulation | decompress_rd[f32, (100000, 0.0)] |
499.5 µs | 586.5 µs | -14.85% |
| ⚡ | Simulation | extend_from_array_non_zctl_overlapping[(10000, 8)] |
4 ms | 2.3 ms | +70.67% |
| ⚡ | Simulation | extend_from_array_non_zctl_overlapping[(1000, 8)] |
452.9 µs | 287 µs | +57.79% |
| ⚡ | Simulation | chunked_bool_canonical_into[(1000, 10)] |
46.7 µs | 31.7 µs | +47.23% |
| ⚡ | Simulation | varbinview_zip_block_mask |
3.7 ms | 2.9 ms | +27.49% |
| ⚡ | Simulation | bitwise_not_vortex_buffer_mut[128] |
275.3 ns | 216.9 ns | +26.89% |
| ⚡ | Simulation | chunked_varbinview_canonical_into[(1000, 10)] |
197.9 µs | 161.9 µs | +22.26% |
| ⚡ | Simulation | extend_from_array_non_zctl_overlapping[(1000, 32)] |
948.6 µs | 782.4 µs | +21.24% |
| ⚡ | Simulation | bitwise_not_vortex_buffer_mut[1024] |
336.9 ns | 278.6 ns | +20.94% |
| ⚡ | Simulation | chunked_varbinview_into_canonical[(1000, 10)] |
213.6 µs | 177.3 µs | +20.48% |
| ⚡ | Simulation | bitwise_not_vortex_buffer_mut[2048] |
400.6 ns | 342.2 ns | +17.05% |
| ⚡ | Simulation | decompress_rd[f64, (100000, 0.01)] |
981.6 µs | 845.5 µs | +16.11% |
| ⚡ | Simulation | decompress_rd[f64, (100000, 0.1)] |
981.6 µs | 845.4 µs | +16.1% |
| ⚡ | Simulation | varbinview_zip_fragmented_mask |
6.9 ms | 6.2 ms | +12.4% |
Tip
Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.
Comparing ji/row-accessor-remove (6626533) with develop (f2148d4)
Loading