[SPARK-6966][SQL] Use correct ClassLoader for JDBC Driver

Otherwise we cannot add jars with drivers after the fact.

Author: Michael Armbrust <michael@databricks.com>

Closes #5543 from marmbrus/jdbcClassloader and squashes the following commits:

d9930f3 [Michael Armbrust] fix imports
73d0614 [Michael Armbrust] [SPARK-6966][SQL] Use correct ClassLoader for JDBC Driver

sql/core/src/main/scala/org/apache/spark/sql/jdbc/JDBCRelation.scala

@@ -28,6 +28,7 @@ import org.apache.spark.sql.SQLContext
 import org.apache.spark.sql.catalyst.expressions.Row
 import org.apache.spark.sql.sources._
 import org.apache.spark.sql.types.StructType
+import org.apache.spark.util.Utils
 
 /**
  * Data corresponding to one partition of a JDBCRDD.
@@ -99,7 +100,7 @@ private[sql] class DefaultSource extends RelationProvider {
     val upperBound = parameters.getOrElse("upperBound", null)
     val numPartitions = parameters.getOrElse("numPartitions", null)
 
-    if (driver != null) Class.forName(driver)
+    if (driver != null) Utils.getContextOrSparkClassLoader.loadClass(driver)
 
     if (partitionColumn != null
         && (lowerBound == null || upperBound == null || numPartitions == null)) {