Ephemerons

Gil Tene gil at azul.com
Sat Jan 23 07:47:57 UTC 2016


> On Jan 22, 2016, at 2:49 PM, Peter Levart <peter.levart at gmail.com> wrote:
> 
> Hi Gil,
> 
> Thanks for taking a look at the Ephemerons for Java. It's great to have a big mind joining the discussion.
> 
> On 01/22/2016 07:12 PM, Gil Tene wrote:
>> Peter,
>> 
>> I've been following Ephemerons in other GC'ed environments, and wondering when someone will bring it up for Java. Happy to see attentions given to it. The "conditional weak reference hashmap/table/dictionary" use case seems to be a common primary motivator, and it's a very valid one. Given that we currently do completely concurrent ref processing (in Zing at least) for all reference types, adding Ephemerons to the spec'ed behavior will add some interesting challenges for keeping things concurrent, but I don't think that it is fundamentally undoable (just "hard" and interesting to fully work out).
> 
> The algorithm for processing ephemerons is not much different from that for processing normal references. It's just that it is iterative until it converges to a stable state. If you allow mutator threads to at least in some phases execute concurrently with normal reference processing then I suspect it should be possible to do that for ephemerons too, but I don't have a clue what tricks you perform to be able to allow mutator threads to be concurrent with reference processing.

The subject of Concurrent Reference Processing has not been widely written about. While we've been doing it in our production collectors (PGC, C4 in our Vega and Zing products) since roughly 2007, we had not published papers about the mechanisms involved. Partially because it seemed to be an "obscure" area of GC. I believe there are versions of Metronome that also dealt with non-STW ref processing, but without publishing the mechanisms either. In fact, the first publication I am aware of that deals with the subject of Concurrent (aka "on the fly") collection and reference types is "Reference Object Processing in On-The-Fly Garbage Collection" by Ugawa, Jones, and Ritson (published in 2014 at ISMM'14, and can be found here <https://kar.kent.ac.uk/40820/>, with a presentation here <https://www.cs.kent.ac.uk/projects/gc/mirrorgc/ismm-2014-reftypes.pdf>).

I believe that I can say (with some authority) that there are sound and reliable ways to achieve fully concurrent (as in Zero STW) reference processing for the non-strong reference types in Java today (soft, weak, phantom, final). But that doing so requires some interesting state machine modeling and careful "throwing of gates" to achieve atomic transitions in behavior without requiring STW processing of references (which is probably a must), and without even requiring an STW transition point (which is a nice bonus for the perfectionists among us, but probably acceptable to do without if the transition included no processing of references). The mechanisms we use rely on an ability to carefully "throw a gate" at a point in the tracing where strong ref processing (including all soft refs that have been policy-chosen to be strengthened in this cycle) is known to be complete, and after which references to non-strongly-marked objects can safely return null (and before which they return a strong reference). Ephemerons will almost certainly mess with how that gate currently works...

So while I believe that dealing with Ephemerons in a STW ref processing world is fairly straight forward, adding the Ephemeron behavior into a concurrent ref processing world will "mess" with the delicate transitions and invariants that make the current mechanisms work. Still, I'm fairly confident that we could still find a way to deal with that. But I don't have one formulated fully at the moment, and don't know of anyone else who has...

> 
>> Scanning through the proposal and Java class (mostly JavaDoc spec), I have the following question:
>> 
>> Do we really need a separate "ephemerally reachable" strength below week? The was you extended the definition to the weak definition to say "An object is weakly reachable if it is neither strongly nor softly reachable but can be reached by traversing a weak reference or by traversing an ephemeron through it's value while the ephemeron's key is at least weakly reachable." would [naively] seem to be sufficient to add and fully describe the desired Ephemeron behavior from a reference strength perspective.
> 
> Ephemeron always touches definitions of at least two consecutive strengths of reachabilities. The prototype says:
> 
>  * <li> An object is <em>weakly reachable</em> if it is neither
>  * strongly nor softly reachable but can be reached by traversing a
>  * weak reference or by traversing an ephemeron through it's value while
>  * the ephemeron's key is at least weakly reachable.
> 
>  * <li> An object is <em>ephemerally reachable</em> if it is neither
>  * strongly, softly nor weakly reachable but can be reached by traversing an
>  * ephemeron through it's key or by traversing an ephemeron through it's value
>  * while it's key is at most ephemerally reachable. When the ephemerons that
>  * refer to ephemerally reachable key object are cleared, the key object becomes
>  * eligible for finalization.
> 
> 
> But Ephemeron does not need a special reachability strength. It could be merged with WeakReference as far as the reachability of it's key is concerned. In that case it would touch at least the definitions of softly-reachable and weakly-reachable:
> 
>  * <li> An object is <em>softly reachable</em> if it is not strongly
>  * reachable but can be reached by traversing a soft reference or
>  * by traversing an ephemeron through it's value while
>  * the ephemeron's key is at least softly reachable.
> 
>  * <li> An object is <em>weakly reachable</em> if it is neither
>  * strongly nor softly reachable but can be reached by traversing a
>  * weak reference (including ephemeron through it's key) or by traversing
>  * an ephemeron through it's value while the ephemeron's key is at most
>  * weakly reachable. When the weak  references to a weakly-reachable
>  * object are cleared (which includes ephemerons to a weakly-reachable key),
>  * the object becomes eligible for finalization.
> 
> 
> Presently the hotspot handles different strengths of references by maintaining a distinct set of discovered references for each type and then processing these sets in sequence from the strongest to the weakest. If Ephemerons and WeakReferences were kept in the same discovered set, they would present the same reachability strength for their referent (called also the key in case of ephemeron).
> 
>> Specifically, would modifying your implementation such that Ephemeron<K, V> extended WeakReference<K> (instead of Reference<K>), and it's V value was tracked with "private WeakReference<V> valueRef" (instead of "private V value") [along with the semi-obvious internal changes that would result] provide what is needed to support proper Ephemeron behavior we just added the "…or by traversing an ephemeron through it's value while the ephemeron's key is at least weakly reachable." to the spec'ed definition of "weakly reachable" (and modified the associated JVM ref processing accordingly, which you already have to do to support your wider definitions anyway)?
> 
> In case the strength of ephemeron's key was the same as that of WeakReference's referent, it would make sense for Ephemeron to extend WeakReference, since it would then just be a special kind of WeakReference. But I don't think that changing "private V value" into "private WeakReference<V> valueRef" would buy anything in terms of processing algorithm simplicity. At least in hotspot it would increase the code complexity, since marking the ephemeron's value 'alive' as a consequence of finding it's key marked alive during iteration over the list of ephemerons and removing the current ephemeron from the pending list, would also require removing of the valuRef WeakReference from the pending list at the same time but that would require finding it in the list 1st, since it's a single-linked-list or removing it at next iteration which would increase the number of iterations through the pending set before the state converges...
> 
> There's no problem in making the VM to treat the "private V value" specially. In HotSpot it is only required to update the OopMap for the Ephemeron class to exclude the value field - analogue to how this is done for Reference fields. There's no need for indirection through another WeakReference to get special treatment for value (exclusion from traversals).
> 
> Regards, Peter
> 
> 
> 
>  — Gil.
>> 
>>> On Jan 17, 2016, at 10:18 AM, Peter Levart <peter.levart at gmail.com> <mailto:peter.levart at gmail.com> wrote:
>>> 
>>> Hi,
>>> 
>>> Ephemerons are special kind of references described by Barry Hayes [1]. In the simple variant, they contain two "referents". One is called the key, which can be viewed as a "weak" referent in the style of Java reference types (SoftReference, WeakReference, PhantomReference), the other is called the value, which is also treated specially by GC. It's reachability is dependent on the reachability of the key.
>>> 
>>> Ephemerons solve the problem seen for example in the java.util.WeakHashMap when a value in this map directly or indirectly refers to it's key. Such entry will never be purged as the value is always reachable from the WeakHashMap instance and through the value, it's key is reachable too. There are other places where Ephemerons could be used - for example in ClassValue API. Try googling for "java ephemeron" (without quotes) to find out other situations where Ephemerons would be beneficial.
>>> 
>>> If one was to implement Ephemerons in Java, the main question he would be asking is how Ephemeron(s) were supposed to interact with existing Java reference types. Hayes only defines their behavior in isolation, but Java already has 4 "weak" reference types with different "strengths": SoftReference, WeakReference, FinalReference (for implementing finalization) and PhantomReference. In the prototype I choose to define Ephemerons as a new reference type with it's own "strength" for the key referent positioned between WeakReference and FinalReference. It would be possible to merge it with an existing reference type like WeakReference or position it between SoftReference and WeakReference or even entirely "on the top" of the strength list, but I think it would not be wise to position it below FinalReference or even PhantomReference. What's best is an open question. What is also not so obvious is how to define the reachability of the Ephemeron's value and it's interaction with exis
>>>  ting refe
>>> rence types. I think I defined it soundly (see the reachability types in the javadoc of [4]). The reason for defining the reachability of the value to alternate between ephemeraly-reachable and weakly-reachable and not between ephemeraly-reachable and strongly-reachable, while theoretically possible, is the desire to have a separate phase of processing for each reachability strength, like it is done currently, and which doesn't affect the performance of processing of existing reference types.
>>> 
>>> Having skimmed through the reference processing code in the VM for a couple of times, I thought it should not be too complicated to implement another type of "weak" reference. Encouraged by how little changes were needed to remove the sun.misc.Cleaner special type of reference [2], I began experimenting and came up with a prototype that seems to work [3]. Luckily most of the logic to process Reference objects in VM is encapsulated in the ReferenceProcessor class and this logic is invoked from various GC implementations of the hotspot. Changes needed are therefore mostly localized to this class. The webrev also contains recent changes from JDK-8143847 [2] that have not yet propagated to the jdk9-dev repo. When they do, I will prepare a rebased webrev without those changes.
>>> 
>>> I'm publishing this prototype here to get the answer to the main question: Is there an interest to have Ephemeron reference type in Java?
>>> 
>>> Given that there was interest, I would also like to initiate a discussion about:
>>> - What would be the most appropriate "strength" of reachabillity for Epehmeron referents (key and value) and the desired interactions with existing reachabilities.
>>> - The prototype itself. Since I'm new to this part of code (that's my 1st not so shallow dive into the VM), I surely must have missed places that should be changed. Although the prototype seems to work, I have not created extensive tests for the functionality and have not exposed it to all the various GC algorithms and situations yet. I could use some advise on how to exercise all GC algorithm combinations possible in the VM (what flags to use, for example).
>>> 
>>> Regards, Peter
>>> 
>>> [1] https://static.aminer.org/pdf/PDF/000/522/273/ephemerons_a_new_finalization_mechanism.pdf <https://static.aminer.org/pdf/PDF/000/522/273/ephemerons_a_new_finalization_mechanism.pdf>
>>> [2] https://bugs.openjdk.java.net/browse/JDK-8143847 <https://bugs.openjdk.java.net/browse/JDK-8143847>
>>> [3] http://cr.openjdk.java.net/~plevart/misc/Ephemeron/ <http://cr.openjdk.java.net/~plevart/misc/Ephemeron/>
>>> [4] http://cr.openjdk.java.net/~plevart/misc/Ephemeron/webrev.jdk.01/src/java.base/share/classes/java/lang/ref/Ephemeron.java.html <http://cr.openjdk.java.net/~plevart/misc/Ephemeron/webrev.jdk.01/src/java.base/share/classes/java/lang/ref/Ephemeron.java.html>
>>> 
>>> P.S.
>>> 
>>> To wet some appetite, with the above prototype (applied to current jdk9-dev), it is possible to run the following example:
>>> 
>>> import java.lang.ref.Ephemeron;
>>> import java.util.ArrayList;
>>> import java.util.List;
>>> 
>>> public class EphemeronTest {
>>> 
>>>    static class Key {
>>>        final int i;
>>> 
>>>        Key(int i) {
>>>            this.i = i;
>>>        }
>>> 
>>>        @Override
>>>        public String toString() {
>>>            return "k" + i;
>>>        }
>>>    }
>>> 
>>>    static class Value {
>>>        final Key key;
>>> 
>>>        Value(Key key) {
>>>            this.key = key;
>>>        }
>>> 
>>>        @Override
>>>        public String toString() {
>>>            return "v(" + key + ")";
>>>        }
>>>    }
>>> 
>>>    static class Eph extends Ephemeron<Key, Value> {
>>>        public Eph(Key key, Value value) {
>>>            super(key, value);
>>>        }
>>> 
>>>        @Override
>>>        public String toString() {
>>>            return getKey() + "=" + getValue();
>>>        }
>>>    }
>>> 
>>>    static void test(int size, boolean forwardChaining) throws Exception {
>>>        System.out.println();
>>>        System.out.println((forwardChaining ? "Forward" : "Backward") +
>>>                           " chaining of value->key ...");
>>>        System.out.println();
>>>        List<Eph> ephs = new ArrayList<>(size);
>>> 
>>>        Key k1 = new Key(1);
>>>        {
>>>            Key kp = k1;
>>>            for (int i = 2; i <= size; i++) {
>>>                Key ki = new Key(i);
>>>                ephs.add(
>>>                    forwardChaining
>>>                    ? new Eph(kp, new Value(ki))
>>>                    : new Eph(ki, new Value(kp))
>>>                );
>>>                kp = ki;
>>>            }
>>>            ephs.add(
>>>                forwardChaining
>>>                ? new Eph(kp, new Value(k1))
>>>                : new Eph(k1, new Value(kp))
>>>            );
>>>            kp = null;
>>>        }
>>> 
>>>        System.gc();
>>>        System.out.println("1: " + ephs);
>>>        k1.getClass(); // reachabilityFence
>>> 
>>>        k1 = null;
>>>        System.gc();
>>>        System.out.println("2: " + ephs);
>>>    }
>>> 
>>>    public static void main(String[] args) throws Exception {
>>> 
>>>        int size = (args.length < 1)
>>>                   ? 5
>>>                   : Math.max(2, Integer.parseInt(args[0]));
>>> 
>>>        test(size, true);
>>>        test(size, false);
>>>    }
>>> }
>>> 
>>> 
>>> Which prints:
>>> 
>>> Forward chaining of value->key ...
>>> 
>>> 1: [k1=v(k2), k2=v(k3), k3=v(k4), k4=v(k5), k5=v(k1)]
>>> 2: [null=null, null=null, null=null, null=null, null=null]
>>> 
>>> Backward chaining of value->key ...
>>> 
>>> 1: [k2=v(k1), k3=v(k2), k4=v(k3), k5=v(k4), k1=v(k5)]
>>> 2: [null=null, null=null, null=null, null=null, null=null]
>>> 
>>> 
>>> If you compile the JDK with --enable-debug and use the following VM switches when running: -XX:+PrintGCDetails -XX:+TraceReferenceGC, you can also observe the inner workings of the reference processing, including Ephemerons.
>>> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.java.net/pipermail/hotspot-gc-dev/attachments/20160123/c3dfda9a/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 842 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <https://mail.openjdk.java.net/pipermail/hotspot-gc-dev/attachments/20160123/c3dfda9a/signature.asc>


More information about the hotspot-gc-dev mailing list