[SQL] [MINOR] [TEST] Set spark.unsafe.exceptionOnMemoryLeak to true

[gatorsmile]

What changes were proposed in this pull request?

When running IntelliJ, we are unable to capture the exception of memory leak detection.

org.apache.spark.executor.Executor: Managed memory leak detected

Explicitly setting spark.unsafe.exceptionOnMemoryLeak in SparkConf when building the SparkSession, instead of reading it from system properties.

How was this patch tested?

N/A

[gatorsmile]

cc @cloud-fan@kiszk

[SparkQA]

Test build #80751 has finished for PR 18967 at commit c6cca96.

• This patch fails PySpark unit tests.
• This patch merges cleanly.
• This patch adds no public classes.

[gatorsmile]

retest this please

[cloud-fan]

LGTM

[kiszk]

LGTM

[cloud-fan]

shall we also update pom.xml and SparkBuild.scala to remove the special setting for this config?

[SparkQA]

Test build #80753 has finished for PR 18967 at commit c6cca96.

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

[dongjoon-hyun]

+1 for updating pom and SparkBuild.scala.

[gatorsmile]

Since the changes made in this PR only affect the tests in sql/core and sql/hive package, I will not change pom and SparkBuild.scala, which also cover the testing in the other packages.

[gatorsmile]

Thanks! Merging to master.