msokolov commented on PR #14874: URL: https://github.com/apache/lucene/pull/14874#issuecomment-3062970247
I did dump the generated machine instructions and while it's hard to really understand what is going on in there, there is clearly a major difference between the code being emitted for `dotProductBody256` and `dotProductBody256 @ 10` when the hack is in place vs not. So I think the theory that hot spot is optimizing differently in these two cases seems correct -- 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