antonha opened a new pull request, #13149: URL: https://github.com/apache/lucene/pull/13149
Instead of calling `IntersectVisitor.visit` for each doc in the `readDelta16` and `readInts32` methods, create a `DocIdSetIterator` and call `IntersectVisitor.visit(DocIdSetIterator)` instead. This seems to make Lucene faster at some sorting and range querying tasks - I saw **35-45% reduction in execution time**. In learnt this through running this benchmark setup by Elastic: https://github.com/elastic/elasticsearch-opensearch-benchmark. The hypothesis is that this is due to fewer virtual calls being made - once per BKD leaf, instead of once per document. Note that this is only measurable if the readInts methods have been called with at least 3 implementation of the IntersectVisitor interface - otherwise the JIT inlining takes away the virtual call. In real life Lucene deployments, I would judge that it is very likely that at least 3 implementations are used. For more details on method etc, there are details in this blog post: https://blunders.io/posts/es-benchmark-4-inlining I tried benchmarking this with luceneutil, but did not see any significant change with the default benchmark - I suspect that I'm using the wrong luceneutil tasks to see any major difference. Which luceneutils benchmarks should I be using for these changes? -- 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
