Github user srinathshankar commented on a diff in the pull request:
https://github.com/apache/spark/pull/15043#discussion_r79042389
--- Diff:
core/src/main/scala/org/apache/spark/storage/memory/MemoryStore.scala ---
@@ -751,23 +757,43 @@ private[storage] class PartiallySerializedBlock[T](
taskContext.addTaskCompletionListener { _ =>
// When a task completes, its unroll memory will automatically be
freed. Thus we do not call
// releaseUnrollMemoryForThisTask() here because we want to avoid
double-freeing.
- unrolled.dispose()
+ unrolledBuffer.dispose()
+ }
+ }
+
+ // Exposed for testing
+ private[storage] def getUnrolledChunkedByteBuffer: ChunkedByteBuffer =
unrolledBuffer
+
+ private[this] var discarded = false
+ private[this] var consumed = false
+
+ private def verifyNotConsumedAndNotDiscarded(): Unit = {
+ if (consumed) {
+ throw new IllegalStateException(
+ "Can only call one of finishWritingToStream() or valuesIterator()
and can only call once.")
+ }
+ if (discarded) {
+ throw new IllegalStateException("Cannot call methods on a discarded
PartiallySerializedBlock")
}
+ consumed = true
--- End diff --
Hmm, why set consumed = true here ? What's the problem with calling
verifyNotConsumedAndNotDiscarded() twice ?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
