labuladong commented on code in PR #18340:
URL: https://github.com/apache/pulsar/pull/18340#discussion_r1014954124
##########
managed-ledger/src/main/java/org/apache/bookkeeper/mledger/impl/ManagedCursorImpl.java:
##########
@@ -2707,6 +2707,25 @@ public void operationFailed(ManagedLedgerException
exception) {
});
}
+ /**
+ * Try set {@link #state} to {@link State#Closing}.
+ * @return false if the {@link #state} already is {@link State#Closing} or
{@link State#Closed}.
+ */
+ private boolean trySetStateToClosing() {
+ final AtomicBoolean notClosing = new AtomicBoolean(false);
+ STATE_UPDATER.updateAndGet(this, state -> {
+ switch (state){
+ case Closing:
+ case Closed: return state;
+ default: {
+ notClosing.set(true);
+ return State.Closing;
+ }
Review Comment:
[`updateAndGet`
doc](https://docs.oracle.com/javase/8/docs/api/java/util/concurrent/atomic/AtomicReferenceFieldUpdater.html#updateAndGet-T-java.util.function.UnaryOperator-)
says:
> The function should be side-effect-free, since it may be re-applied when
attempted updates fail due to contention among threads.
So I guess `notClosing.set(true)` may cause tricky bugs. How about just
comparing the state that before and after `updateAndGet`?
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
