BSI: fix ClearValues being slower than a serial loop#515
Merged
lemire merged 2 commits intoRoaringBitmap:masterfrom Mar 9, 2026
Merged
BSI: fix ClearValues being slower than a serial loop#515lemire merged 2 commits intoRoaringBitmap:masterfrom
lemire merged 2 commits intoRoaringBitmap:masterfrom
Conversation
The goroutine-per-bit-plane approach it replaced was slower in practice. Goroutine creation overhead dominates for typical BSI sizes, and the cost compounds when ClearValues is called once per term in a deletion pass over an entire index (the pattern used in caterwaul's DeleteGroups).
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
ClearValuespreviously spawned one goroutine per bit plane plus one forthe existence bitmap on every call. The goroutine creation overhead
dominated for typical BSI sizes, making the function slower than a simple
serial loop — in practice, an order of magnitude or more.
The fix replaces the parallel implementation with a straightforward serial
loop. The performance case for parallelism doesn't hold here: the work per
bit plane is a single
AndNoton an already-vectorized bitmap, whichcompletes faster than the goroutine scheduler can dispatch. The cost also
compounds when
ClearValuesis called in a tight loop, such as once perterm when purging deleted documents from an entire index.