⚡ Bolt: Unroll AVX2 match length comparison to 256 bytes#380
Conversation
Co-authored-by: 404Setup <[email protected]>
|
👋 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. |
404Setup
deleted the
bolt-optimize-match-len-avx2-11708908719364509432
branch
1 participant
Optimized
match_len_avx2insrc/compress/matchfinder.rsby unrolling the comparison loop to process 256 bytes per iteration (previously 128 bytes).This reduces the overhead of
_mm256_testz_si256(which checks for any mismatch) by performing it half as often for long matches. The implementation accumulates XOR differences using_mm256_or_si256(which is cheap) before performing the expensive test.Correctness was verified with
cargo test compress::matchfinder, including explicit tests for AVX2 and tail handling. The performance improvement targets long match scenarios (e.g. repetitive data) which are common in compression workloads.Note: An earlier attempt to optimize
adler32AVX2 by breaking dependency chains resulted in a regression due to register spilling, so that change was reverted.PR created automatically by Jules for task 11708908719364509432 started by @404Setup