jpountz commented on PR #13658: URL: https://github.com/apache/lucene/pull/13658#issuecomment-2294992594
Sadly, this seems to have caused a minor regression on nightly benchmarks, e.g. [`AndHighMed`](https://people.apache.org/~mikemccand/lucenebench/AndHighMed.html) lost 2% with a p-value of 0. I cannot reproduce the slowdown locally. However, one thing I am observing is that `splitLongs` is never inlined into the `decodeX` methods, which is annoying as I would expect it to be important performance-wise to unroll various loops. I am not sure if this is due to insufficient warming or if we could organize the code a bit differently to help the JVM make better inlining decisions. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
