SPARK-19276 ensured that FetchFailures do not get swallowed by other
layers of exception handling, but it also meant that a killed task could
look like a fetch failure.  This is particularly a problem with
speculative execution, where we expect to kill tasks as they are reading
shuffle data.  The fix is to ensure that we always check for killed
tasks first.

Added a new unit test which fails before the fix, ran it 1k times to
check for flakiness.  Full suite of tests on jenkins.

Author: Imran Rashid <irashid@cloudera.com>

Closes apache#20987 from squito/SPARK-23816.

(cherry picked from commit 10f45bb)
Signed-off-by: Marcelo Vanzin <vanzin@cloudera.com>

…merge

…enerate a wrong result by codegen.

`EqualNullSafe` for `FloatType` and `DoubleType` might generate a wrong result by codegen.

```scala
scala> val df = Seq((Some(-1.0d), None), (None, Some(-1.0d))).toDF()
df: org.apache.spark.sql.DataFrame = [_1: double, _2: double]

scala> df.show()
+----+----+
|  _1|  _2|
+----+----+
|-1.0|null|
|null|-1.0|
+----+----+

scala> df.filter("_1 <=> _2").show()
+----+----+
|  _1|  _2|
+----+----+
|-1.0|null|
|null|-1.0|
+----+----+
```

The result should be empty but the result remains two rows.

Added a test.

Author: Takuya UESHIN <ueshin@databricks.com>

Closes apache#21094 from ueshin/issues/SPARK-24007/equalnullsafe.

(cherry picked from commit f09a9e9)
Signed-off-by: gatorsmile <gatorsmile@gmail.com>

…n text-based Hive table

## What changes were proposed in this pull request?

TableReader would get disproportionately slower as the number of columns in the query increased.

I fixed the way TableReader was looking up metadata for each column in the row. Previously, it had been looking up this data in linked lists, accessing each linked list by an index (column number). Now it looks up this data in arrays, where indexing by column number works better.

## How was this patch tested?

Manual testing
All sbt unit tests
python sql tests

Author: Bruce Robbins <bersprockets@gmail.com>

Closes apache#21043 from bersprockets/tabreadfix.

## What changes were proposed in this pull request?
Fix comment. Change `BroadcastHashJoin.broadcastFuture` to `BroadcastExchangeExec.relationFuture`: https://github.com/apache/spark/blob/d28d5732ae205771f1f443b15b10e64dcffb5ff0/sql/core/src/main/scala/org/apache/spark/sql/execution/exchange/BroadcastExchangeExec.scala#L66

## How was this patch tested?
N/A

Author: seancxmao <seancxmao@gmail.com>

Closes apache#21113 from seancxmao/SPARK-13136.

(cherry picked from commit c303b1b)
Signed-off-by: hyukjinkwon <gurwls223@apache.org>

…merge

…-2.2-merge