Yuxin Tan created FLINK-28942:
---------------------------------
Summary: Deadlock may occurs when releasing readers for
SortMergeResultPartition
Key: FLINK-28942
URL: https://issues.apache.org/jira/browse/FLINK-28942
Project: Flink
Issue Type: Bug
Affects Versions: 1.16.0
Reporter: Yuxin Tan
After adding the logic of recycling buffers in CompositeBuffer in
https://issues.apache.org/jira/browse/FLINK-28373, when reading data and
recycling buffers simultaneously, the deadlock between the lock of
SortMergeResultPartition and the lock of SingleInputGate may occur.
In short, the deadlock may occur as follows.
1. SingleInputGate.getNextBufferOrEvent (SingleInputGate lock)
CompositeBuffer.getFullBufferData -> CompositeBuffer.recycleBuffer -> waiting
for
SortMergeResultPartition lock;
2. ResultPartitionManager.releasePartition (SortMergeResultPartition lock) ->
SortMergeSubpartitionReader.notifyDataAvailable ->
SingleInputGate.notifyChannelNonEmpty -> waiting for SingleInputGate lock.
The possibility of this deadlock is very small, but we should fix the bug as
soon as possible.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
