[SPARK-17710][FOLLOW UP] Add comments to state why 'Utils.classForName' is not used

## What changes were proposed in this pull request?
Add comments.

## How was this patch tested?
Build passed.

Author: Weiqing Yang <yangweiqing001@gmail.com>

Closes #15776 from weiqingy/SPARK-17710.

(cherry picked from commit 8a9ca192)
Signed-off-by: Reynold Xin <rxin@databricks.com>

core/src/main/scala/org/apache/spark/util/Utils.scala

@@ -2539,6 +2539,8 @@ private[util] object CallerContext extends Logging {
   val callerContextSupported: Boolean = {
     SparkHadoopUtil.get.conf.getBoolean("hadoop.caller.context.enabled", false) && {
       try {
+        // `Utils.classForName` will make `ReplSuite` fail with `ClassCircularityError` in
+        // master Maven build, so do not use it before resolving SPARK-17714.
         // scalastyle:off classforname
         Class.forName("org.apache.hadoop.ipc.CallerContext")
         Class.forName("org.apache.hadoop.ipc.CallerContext$Builder")
@@ -2604,6 +2606,8 @@ private[spark] class CallerContext(
   def setCurrentContext(): Unit = {
     if (CallerContext.callerContextSupported) {
       try {
+        // `Utils.classForName` will make `ReplSuite` fail with `ClassCircularityError` in
+        // master Maven build, so do not use it before resolving SPARK-17714.
         // scalastyle:off classforname
         val callerContext = Class.forName("org.apache.hadoop.ipc.CallerContext")
         val builder = Class.forName("org.apache.hadoop.ipc.CallerContext$Builder")