Github user cloud-fan commented on a diff in the pull request:
https://github.com/apache/spark/pull/20224#discussion_r163773021
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/internal/SQLConf.scala ---
@@ -629,6 +629,14 @@ object SQLConf {
.booleanConf
.createWithDefault(true)
+ val WHOLESTAGE_CODEGEN_USE_ID_IN_CLASS_NAME =
+ buildConf("spark.sql.codegen.useIdInClassName")
+ .internal()
+ .doc("When true, embed the (whole-stage) codegen stage ID into " +
+ "the class name of the generated class as a suffix")
+ .booleanConf
+ .createWithDefault(true)
--- End diff --
I think it's always good to have id in explain and generated classes. The
only concern is we may have codegen cache issues if putting id in the class
name, so we need a config to turn it off.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
