[BUG] Build failures against Spark 3.2.1 rc1 and make 3.2.1 non snapshot

[tgravescs]

Describe the bug
I was building against the spark 3.2.1 rc1 that is up for vote and got some build failures:

[ERROR] [Error] /home/tgraves/workspace/spark-rapids/sql-plugin/src/main/320until322/scala/com/nvidia/spark/rapids/shims/v2/Spark320until322Shims.scala:40: value datetimeRebaseMode is not a member of object org.apache.spark.sql.execution.datasources.DataSourceUtils
possible cause: maybe a semicolon is missing before `value datetimeRebaseMode'?
[ERROR] [Error] /home/tgraves/workspace/spark-rapids/sql-plugin/src/main/320until322/scala/org/apache/spark/sql/execution/datasources/parquet/rapids/shims/v2/ShimVectorizedColumnReader.scala:43: type mismatch;
 found   : org.apache.spark.sql.internal.SQLConf.LegacyBehaviorPolicy.Value
 required: org.apache.spark.sql.catalyst.util.RebaseDateTime.RebaseSpec
[INFO] [Info] : org.apache.spark.sql.internal.SQLConf.LegacyBehaviorPolicy.Value <: org.apache.spark.sql.catalyst.util.RebaseDateTime.RebaseSpec?
[INFO] [Info] : false
[ERROR] [Error] /home/tgraves/workspace/spark-rapids/sql-plugin/src/main/320until322/scala/org/apache/spark/sql/execution/datasources/parquet/rapids/shims/v2/ShimVectorizedColumnReader.scala:44: type mismatch;
 found   : org.apache.spark.sql.internal.SQLConf.LegacyBehaviorPolicy.Value
 required: org.apache.spark.sql.catalyst.util.RebaseDateTime.RebaseSpec
[INFO] [Info] : org.apache.spark.sql.internal.SQLConf.LegacyBehaviorPolicy.Value <: org.apache.spark.sql.catalyst.util.RebaseDateTime.RebaseSpec?
[INFO] [Info] : false
[ERROR] [Error] /home/tgraves/workspace/spark-rapids/sql-plugin/src/main/320until322/scala/org/apache/spark/sql/execution/datasources/parquet/rapids/shims/v2/ShimVectorizedColumnReader.scala:56: not enough arguments for constructor VectorizedColumnReader: (x$1: org.apache.parquet.column.ColumnDescriptor, x$2: org.apache.parquet.schema.LogicalTypeAnnotation, x$3: org.apache.parquet.column.page.PageReader, x$4: java.util.PrimitiveIterator.OfLong, x$5: java.time.ZoneId, x$6: String, x$7: String, x$8: String, x$9: String)org.apache.spark.sql.execution.datasources.parquet.VectorizedColumnReader.
Unspecified value parameters x$8, x$9.
[ERROR] four errors found

Looks like the datetimeRebase mode caused by https://issues.apache.org/jira/browse/SPARK-37705

[tgravescs]

Note this was initially handled for 3.3.0 and 3.2.2 in 9281caa#diff-91b290931d42094ddbdde567e494e940cfb9c53cbbbe1f679eaa3b513afe2723

and #4465

But it looks like our build wasn't fully up to date due tot the way the spark release candidate version was already incremented.

[Niharikadutta]

@tgravescs Do we have a working change ready for this? We are seeing these issues while building against Spark 3.2.1.

[jlowe]

@Niharikadutta yes, the linked PR #4508 should resolve the issue.

[tgravescs]

pr merged