Retry with smaller split on CudfColumnSizeOverflowException

Depends on rapidsai/cudf#13911.

When a CUDF operation causes a column's size to exceed the valid range
for CUDF columns (i.e. cudf::size_type), CUDF will throw an exception.

Prior to this commit, the `RmmRapidsRetryIterator` does not attempt retries
with smaller splits, in this case. Instead, the overflow is treated as
a generic exception.

This commit allows the RmmRapidsRetryIterator to recognize the exception
specific to the overflow case (i.e. `CudfColumnSizeOverflowException`),
and attempt a split-retry.

Note: This error condition is difficult to reproduce. The catch/retry is
a "best effort" attempt not to fail the entire task.

Signed-off-by: MithunR <[email protected]>

sql-plugin/src/main/scala/com/nvidia/spark/rapids/RmmRapidsRetryIterator.scala

@@ -18,6 +18,8 @@ package com.nvidia.spark.rapids
 
 import scala.collection.mutable
 
+import ai.rapids.cudf.CudfColumnSizeOverflowException
+
 import com.nvidia.spark.rapids.Arm.{closeOnExcept, withResource}
 import com.nvidia.spark.rapids.RapidsPluginImplicits._
 import com.nvidia.spark.rapids.ScalableTaskCompletion.onTaskCompletion
@@ -580,9 +582,14 @@ object RmmRapidsRetryIterator extends Logging {
             lastException = ex
 
             if (!topLevelIsRetry && !causedByRetry) {
-              // we want to throw early here, since we got an exception
-              // we were not prepared to handle
-              throw lastException
+              // If the exception is the result of a CUDF column size overflow, attempt split-retry.
+              ex match {
+                case _: CudfColumnSizeOverflowException => doSplit = true
+                case _ =>
+                  // we want to throw early here, since we got an exception
+                  // we were not prepared to handle
+                  throw lastException
+              }
             }
             // else another exception wrapped a retry. So we are going to try again
         }