[GitHub] [spark] srowen commented on a diff in pull request #36496: [SPARK-39104][SQL] Synchronize isCachedColumnBuffersLoaded to avoid concurrency issue

[GitBox <gi...@apache.org>]

srowen commented on code in PR #36496:
URL: https://github.com/apache/spark/pull/36496#discussion_r869427441


##########
sql/core/src/main/scala/org/apache/spark/sql/execution/columnar/InMemoryRelation.scala:
##########
@@ -238,7 +238,12 @@ case class CachedRDDBuilder(
   }
 
   def isCachedColumnBuffersLoaded: Boolean = {
-    _cachedColumnBuffers != null && isCachedRDDLoaded
+    if (_cachedColumnBuffers != null) {
+      _cachedColumnBuffers.synchronized {
+        return _cachedColumnBuffers != null && isCachedRDDLoaded
+      }
+    }
+    false
   }
 
   def isCachedRDDLoaded: Boolean = {

Review Comment:
   Wait a sec though, _cachedColumnBuffers is set to null in a synchronized block (not synchronized on the object itself). Don't we just need a similar synchronized block? I'm always worried about holding two different locks in this class, as it can lead to deadlock



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org