mgaido91 commented on issue #23731: [SPARK-26572][SQL] fix aggregate codegen result evaluation URL: https://github.com/apache/spark/pull/23731#issuecomment-460171427 The changes makes sense to me, but I think this problem was introduced in SPARK-13404, which claimed to have a significant perf gain (about 30% on TPCDS Q55), so it would be great if we can fix this without introducing perf regression. @peter-toth may you please run (and post the results) the benchmarks in order to ensure we are not introducing a perf regression with this PR? @davies you are the author of that PR, do you have time to check this?
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
