reintroduce vectorcall optimization with new PyCallArgs trait
#4768
This check has been archived and is scheduled for deletion.
Learn more about checks retention
CodSpeed HQ / CodSpeed Performance Analysis
succeeded
Dec 24, 2024
Performance Gate Passed
Summary
⚡ 5 improvements
✅ 79 untouched benchmarks
Benchmarks breakdown
| Benchmark | main |
Icxolu:pycallargs |
Change | |
|---|---|---|---|---|
| ⚡ | call |
756.2 µs | 620.5 µs | +21.87% |
| ⚡ | call_1 |
380.1 µs | 271 µs | +40.27% |
| ⚡ | call_method_1 |
1,008.9 µs | 641.4 µs | +57.29% |
| ⚡ | call_method_one_arg |
965.8 µs | 611.6 µs | +57.92% |
| ⚡ | call_one_arg |
347.7 µs | 242.1 µs | +43.64% |
Loading