[
https://issues.apache.org/jira/browse/HIVE-19037?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Matt McCline updated HIVE-19037:
--------------------------------
Attachment: HIVE-19037.06.patch
> Vectorization: Miscellaneous cleanup
> ------------------------------------
>
> Key: HIVE-19037
> URL: https://issues.apache.org/jira/browse/HIVE-19037
> Project: Hive
> Issue Type: Bug
> Components: Hive
> Affects Versions: 3.0.0
> Reporter: Matt McCline
> Assignee: Matt McCline
> Priority: Critical
> Attachments: HIVE-19037.01.patch, HIVE-19037.02.patch,
> HIVE-19037.03.patch, HIVE-19037.04.patch, HIVE-19037.05.patch,
> HIVE-19037.06.patch
>
>
> # Extraneous INFO logging in VectorReduceSinkCommonOperator
> # NPE in EXPLAIN for some SelectColumnIsTrue vector expressions
> # Don't throw instantiation exception when UDF has NULL SCALAR that cannot
> match vectorization expression description. Catch it early and don't try to
> create vectorization expression for UDF. (try VectorUDFAdaptor instead).
> # Support vectorizing SELECT NULL.
> # Support constant casting from \{STRING|CHAR|VARCHAR} to
> \{LONG|DOUBLE|DECIMAL} case.
> # Suppress fatal Vectorizer exceptions. Put single line exception stack
> trace innotVectorizedReason. Add test only variable
> hive.test.vectorizer.suppress.fatal.exceptions to turn that off so Vectorizer
> execution failure can be found more easily.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
