perf: byte-level path handling on Unix to bypass Components iterator
#245Comparing
perf/ascii-fast-path
(8cefe78
) withmain
(080188f
)+5%
- Improvement
- 5
- Untouched
- 5
Benchmarks
10 total
resolver[single-thread]
benches/resolver.rs::resolver::bench_resolver+9%
50.3 ms46.1 ms
resolver[resolve from symlinks]
benches/resolver.rs::resolver::bench_resolver+5%
146.4 ms139.1 ms
resolver[[multi-threaded]resolve from symlinks]
benches/resolver.rs::resolver::bench_resolver+4%
75.6 ms72.6 ms
resolver[[multi-threaded]resolve]
benches/resolver.rs::resolver::bench_resolver+4%
57.9 ms55.7 ms
resolver[pnp resolve]
benches/resolver.rs::resolver::bench_resolver+3%
8.4 KB8.2 KB
resolver[[single-threaded]resolve with many extensions]
benches/resolver.rs::resolver::bench_resolver+2%
134.2 ms131 ms
resolver[single-thread]
benches/resolver.rs::resolver::bench_resolver+2%
8.1 MB8 MB
resolver[pnp resolve]
benches/resolver.rs::resolver::bench_resolver+2%
255.3 µs251.4 µs
resolver[resolve from symlinks]
benches/resolver.rs::resolver::bench_resolver0%
12 MB12 MB
resolver[[single-threaded]resolve with many extensions]
benches/resolver.rs::resolver::bench_resolver0%
12.8 MB12.8 MB
Commits
Click on a commit to change the comparison rangeBase
main
080188f
+5.13%
perf: byte-level path handling on Unix to bypass Components iterator
8cefe78
2 days ago
by stormslowly