Repository navigation
perf: reuse one file-hash read buffer per thread - #644
Merged
tinder-maxwellelliott merged 1 commit intoNov 9, 2026
Merged
Conversation
hash_file_into declared a fresh [0; 64 KiB] buffer on every call, so every hashed source file paid for zeroing 64 KiB before the first read, even though most source files are a few hundred bytes. On the large synthetic workspace (16k source files) that memset was a fifth of generate-hashes' instructions. Keep one zeroed buffer per thread and read into it. The reads, their size and the hashed bytes are unchanged, so hashes are identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01KJs3XyWhGePPEpsP7LQtfC
tinder-maxwellelliott
marked this pull request as ready for review
November 9, 2026 14:38
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
hash_file_into declared a fresh [0; 64 KiB] buffer on every call, so every hashed source file paid for zeroing 64 KiB before the first read, even though most source files are a few hundred bytes. On the large synthetic workspace (16k source files) that memset was a fifth of generate-hashes' instructions.
Keep one zeroed buffer per thread and read into it. The reads, their size and the hashed bytes are unchanged, so hashes are identical.