srowen commented on a change in pull request #30810:
URL: https://github.com/apache/spark/pull/30810#discussion_r609144837
##########
File path: project/SparkBuild.scala
##########
@@ -291,6 +291,12 @@ object SparkBuild extends PomBuild {
publishLocal in SbtCompile := publishTask(publishLocalConfiguration in
SbtCompile).value,
publishLocal := Seq(publishLocal in MavenCompile, publishLocal in
SbtCompile).dependOn.value,
+ javaOptions ++= {
+ val versionParts = System.getProperty("java.version").split("[+.\\-]+",
3)
+ var major = versionParts(0).toInt
+ if (major >= 16) Seq("--add-modules=jdk.incubator.vector") else Seq.empty
Review comment:
Would a Spark cluster have to launch its JVMs with this flag to make it
work? that's OK if so, we might just add a note in the docs about it.
##########
File path: mllib-local/pom.xml
##########
@@ -81,6 +88,34 @@
</dependency>
</dependencies>
</profile>
+ <profile>
+ <id>vectorized</id>
+ <properties>
+ <extra.source.dir>src/vectorized/java</extra.source.dir>
Review comment:
Just some nits now: I wonder if we should call this profile something
more like `jvm-vectorized` to be more specific? but it doesn't matter much
--
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.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
