franz1981 commented on a change in pull request #2645: ARTEMIS-2321 Paging
scalability and GC improvement
URL: https://github.com/apache/activemq-artemis/pull/2645#discussion_r279271513
##########
File path:
artemis-core-client/src/main/java/org/apache/activemq/artemis/utils/SoftValueLongObjectHashMap.java
##########
@@ -21,26 +21,26 @@
import java.util.ArrayList;
import java.util.Collection;
import java.util.Comparator;
-import java.util.HashMap;
-import java.util.HashSet;
+import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.TreeSet;
-import java.util.concurrent.atomic.AtomicLong;
+import io.netty.util.collection.LongObjectHashMap;
+import io.netty.util.collection.LongObjectMap;
import org.jboss.logging.Logger;
-public class SoftValueHashMap<K, V extends SoftValueHashMap.ValueCache>
implements Map<K, V> {
+public class SoftValueLongObjectHashMap<V extends
SoftValueLongObjectHashMap.ValueCache> implements LongObjectMap<V> {
- private static final Logger logger =
Logger.getLogger(SoftValueHashMap.class);
+ private static final Logger logger =
Logger.getLogger(SoftValueLongObjectHashMap.class);
// The soft references that are already good.
// too bad there's no way to override the queue method on ReferenceQueue,
so I wouldn't need this
private final ReferenceQueue<V> refQueue = new ReferenceQueue<>();
- private final Map<K, AggregatedSoftReference> mapDelegate = new HashMap<>();
+ private final LongObjectMap<AggregatedSoftReference<V>> mapDelegate = new
LongObjectHashMap<>();
Review comment:
That's clear to me, but let me explain better what I mean...
If we have:
```
private final LongObjectMap<AggregatedSoftReference<V>> mapDelegate
```
and externally we provide:
```
//(1)
V put(long k, V value) {
return mapDelegate.put(k, value);
}
//(2)
V put(Long k, V value) {
return mapDelegate.put(k, value);
}
```
If caller will call (1), the javac will hide a boxing (long -> Long) to
allow calling `map.put(k, value)` given that
mapDelegate has been declared as `Map<Long,...`.
If you will call (2) the boxing will be forced on the caller level, but the
effect is the same: an unecessary `Long` instance.
That's why I've declared the mapDelegate as `LongObjectMap`: to avoid such
hidden boxing.
The point is that
> t can implement the same name with a tighter typing and that will be
invoked instead.
The tigher typing is allowed (right now) only between reference types: value
types like primitives are not considered related to reference types and need
boxing/unboxing.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
