GlutenPerfBot opened a new pull request, #13401: URL: https://github.com/apache/gluten/pull/13401
Upstream Velox's New Commits: ``` 76bf7e9e5 by Maria Basmanova, fix(exec): Stop a fixed point's loop when its task ends (19666) be84a3153 by GALI PREM SAGAR, feat(presto): Add opt-in inf and nan spellings to CAST from VARCHAR 911930187 by Ke, feat(nimble): Wire `nimbleScanFastPath` toggle to ChunkedDecoder (19630) 910348a12 by Maria Basmanova, test(exec): Run the output buffer serde tests with every serde (19667) 9e8bd22a9 by Maria Basmanova, fix(exec): Wait for a fixed point's peers before its first iteration (19664) 4ccc87a30 by GALI PREM SAGAR, feat(cudf): Support Hive partition keys in the cuDF Hive connector 0d9c3cd00 by Pedro Pedreira, perf(functions): Group per-row element_at copies on flat maps e089fcc6d by macduan, fix(dwrf): Handle filter cache resets without an active stripe ac96dd08c by Maria Basmanova, fix(exec): Serve reads of an output buffer that does not exist yet (19662) 9bf75aa86 by Anshul Baliga, fix(sparksql): Honor duplicate map key config in str_to_map 99d6ba8b3 by Daniel Bauer, feat(ucx-exchange): Register the UCX transport and enable its build 101286a45 by Maria Basmanova, fix(exec): Start a fixed point's loop after the task's drivers (19659) 62c048a34 by macduan, fix(paimon): Rebind runtime statistics after data source takeover 9ffaceeeb by Maria Basmanova, docs: Add rules on one check per condition and one detail level 581db8a2a by Xiaoxuan, perf(nimble): Add a cacheEncodingView option to encoding views ``` velox_branch: https://github.com/IBM/velox/commits/dft-2026_10_21 -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
