Github user rdblue commented on a diff in the pull request:
https://github.com/apache/spark/pull/19568#discussion_r146975643
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/joins/SortMergeJoinExec.scala
---
@@ -585,21 +585,26 @@ case class SortMergeJoinExec(
val iterator = ctx.freshName("iterator")
val numOutput = metricTerm(ctx, "numOutputRows")
+ val joinedRow = ctx.freshName("joined")
--- End diff --
The second problem was fixed in this commit:
https://github.com/apache/spark/commit/6b6dd682e84d3b03d0b15fbd81a0d16729e521d2
I still think that the codegen problem should be fixed. Detecting
`CodgenFallback` is imperfect, but will still generate code and run it. I think
we should either remove codegen from `CodegenFallback` or add this fix to
ensure that code works, even if we don't expect to run it.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
