Reduce allocation rate in HNSW concurrent merge #14011
Merged
Conversation
This file contains 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
The PR removes the allocation of a new `LockedRow` for each locking operation in `HnswLock`. Even though the object was very quickly released, and JIT supports on-stack allocation, it didn't happen in my experiments on OpenJDK 21 - it's easy to avoid the allocation, rather than rely on the JIT.
|
Thanks @villam-durina! This looks good. I guess we were trying to enforce access to the rows, requiring that callers acquire a lock to obtain them, but it was really just a fig leaf anyway, since you could still just get the row through the normal unlocked API anyway, so this doesn't really eliminate any safety. I'll merge |
|
@msokolov we should have |
viliam-durina
added a commit
to viliam-durina/apache-lucene
that referenced
this pull request
Dec 2, 2024
The PR removes the allocation of a new `LockedRow` for each locking operation in `HnswLock`. Even though the object was very quickly released, and JIT supports on-stack allocation, it didn't happen in my experiments on OpenJDK 21 - it's easy to avoid the allocation, rather than rely on the JIT. (cherry picked from commit 532d267)
viliam-durina
added a commit
to viliam-durina/apache-lucene
that referenced
this pull request
Dec 4, 2024
The PR removes the allocation of a new `LockedRow` for each locking operation in `HnswLock`. Even though the object was very quickly released, and JIT supports on-stack allocation, it didn't happen in my experiments on OpenJDK 21 - it's easy to avoid the allocation, rather than rely on the JIT. (cherry picked from commit 532d267)
msokolov
pushed a commit
that referenced
this pull request
Dec 4, 2024
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.
The PR removes the allocation of a new
LockedRowfor each locking operation inHnswLock. Even though the object was very quickly released, and JIT supports on-stack allocation, it didn't happen in my experiments on OpenJDK 21 - it's easy to avoid the allocation, rather than rely on the JIT.Comparison of the change processing the same workload. The effect on overall execution time was negligible.
Before:
After:
