[SPARK-13306][SQL] Addendum to uncorrelated scalar subquery

[rxin]

What changes were proposed in this pull request?

This pull request fixes some minor issues (documentation, test flakiness, test organization) with #11190, which was merged earlier tonight.

How was the this patch tested?

unit tests.

[rxin]

this was unused. just deleting it.

[rxin]

cc @davies
can you explain what the following test case is testing?

 assertResult(Array(Row(1))) {
sql("with t2 as (select 1 as b, 2 as c) " +
"select a from (select 1 as a union all select 2 as a) t " +
"where a = (select max(b) from t2) ").collect()
}

[rxin]

Also explain why wholestagecodegen needs to call waitforsubqueries explicitly.

[davies]

For those operators that is generated in WholeStageCodegen, the execute() of them are not called, but produce() will be called, we should make sure that all the subqueries all executed before generating the code in doProduce() and doConsume()

[davies]

That case is used to test CTE and aggregation inside subquery.

[SparkQA]

Test build #51614 has finished for PR 11285 at commit de512f6.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds no public classes.

[SparkQA]

Test build #51611 has finished for PR 11285 at commit c9bd987.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds the following public classes (experimental):
  • case class PlanSubqueries(sqlContext: SQLContext) extends Rule[SparkPlan]

[rxin]

Merging this in master.