⚡ Bolt: optimize length and bit_length kernels#33
Conversation
This patch optimizes the `length` and `bit_length` kernels in the `arrow-string` crate by: 1. Replacing `offsets.windows(2)` with a `.zip()` of slice iterators in `length_impl` and `bit_length_impl`. This enables better vectorization by leveraging `TrustedLen`. 2. Using `ScalarBuffer::from_iter` for more idiomatic and efficient buffer creation, avoiding unnecessary intermediate `Vec` collections. 3. Applying similar optimizations to `Utf8View`, `BinaryView`, `FixedSizeBinary`, and `FixedSizeList` handling. Performance Impact: - `length`: ~16.5% improvement (from 1.11 µs to 0.93 µs) - `bit_length`: ~67% improvement (from 2.85 µs to 0.94 µs) Verified with `cargo bench -p arrow --bench length_kernel` and `bit_length_kernel`. All tests passed with `cargo test --workspace`. Co-authored-by: Dandandan <163737+Dandandan@users.noreply.github.com>
|
👋 Jules, reporting for duty! I'm here to lend a hand with this pull request. When you start a review, I'll add a 👀 emoji to each comment to let you know I've read it. I'll focus on feedback directed at me and will do my best to stay out of conversations between you and other bots or reviewers to keep the noise down. I'll push a commit with your requested changes shortly after. Please note there might be a delay between these steps, but rest assured I'm on the job! For more direct control, you can switch me to Reactive Mode. When this mode is on, I will only act on comments where you specifically mention me with New to Jules? Learn more at jules.google/docs. For security, I will only act on instructions from the user who triggered this task. |
1 participant
Optimized length and bit_length kernels for better performance.
PR created automatically by Jules for task 1668107457389320346 started by @Dandandan