Question about synchronized singleton classes

Hi,
I have a singleton class, with only 1 method (which is static), I want access to that method to be synchronized, but someone suggested that I make my getInstance() method synchronized as well.
Is this neccessary, or is it overkill?
thanks!

Example:
static Instance getInstance() {
if (instance == null) {
instance = new Instance();
return instance;
}Two threads call it simultaneously:
1. if (instance == null) { // evaluates true
2. if (instance == null) { // evaluates true
1. instance = new Instance(); // first instance
2. instance = new Instance(); // second instance,
deleting the reference to the first instance
1. return instance; // which is not the originally
created one but the second instance
2. return instance;There's actually a worse consequence.
1) T1 sees null
2) T1 sets instance to point to where the object will live
4) T2 sees non-null
3) T2 returns, with a pointer to what's supposed to be an object, but isn't yet inited.
4) T1 eventually inits the instance
This can happen even if you use the "double check locking" pattern of synchronization.
At least under the old memory model. I think the new JMM in effect in 5.0 (and possibly later 1.4) versions fixes it. I don't know if the fact that it can happen in absence of sync/DCL is an error or not, so I'm not sure if that's addressed in the new JMM, or if it's only the synced case that was a bug. (That is, it might be okay that it happens without syncing.)

Similar Messages

Question about required workshop classes for OCP

Im working on finishing up my OCP in a few months. Quick question about the workshop class. Is there a test at the end of the week?
Also how do most people pay for these classes? They are pretty steep and my company doesnt pay for stuff like this for me. Im paying totally out of my own pocket. Also why is the online class the same amount as the instructor lead class? Its not like im in a room at some building that Oracle has to lease for use or something.

RedDeuce wrote:
Im working on finishing up my OCP in a few months.What Certificate exactly are you working towards?
Also how do most people pay for these classes? They are pretty steep and my company doesnt pay for stuff like this for me. Im paying totally out of my own pocket.Then talk to your employer again or maybe look for other jobs that support your plan better.
Also why is the online class the same amount as the instructor lead class? Its not like im in a room at some building that Oracle has to lease for use or something.They know that you might save on expenses for travel and hotel etc.

Less dumb follow-up question about super/sub classes in WDJ?

This is a follow-up question to the question which Maksim answered in this thread:
Dumb question about super/sub classes in WDJ
Question:
Is there any kind of weird C++-like statement that you can put at the top of a WDJ module to force the module to interpret any reference to superclass A as a reference to some specific subclass B of A ???

David,
1. Java has no preprocessor, so C++ tricks are not available. Also I would not recommend such tricks even in C++ if you don't want to turn your colleagues working with same code into personal enemies.
2. The phrase "easier to create a WDJ custom class loader " makes me smile. First, it's not that simple to interfere WDJ class loading scheme. Plus custom class loaders is not trivial Java topic per se.
3. The problem "replace all A-s with B-s" is typically solved using one or another GoF creation patterns, like <a href="http://en.wikipedia.org/wiki/Abstract_factory_pattern">Abstract Factory</a> or <a href="http://en.wikipedia.org/wiki/Factory_method_pattern">Factory Method</a>. You may use them with custom class loader, if you really want to
By the way, all UI controls in WD are created using Abstract Factory (role played by view). So you may use this as good example.
Valery Silaev
SaM Solutions
http://www.sam-solutions.net

Question about "synchronized" and "wait"

Hello, everyone!
I have seen a piece of code like this,
synchronized (lock)
//do operation 1
lock.wait();
//do operation 2
I think in the above piece of code, when a thead finished operation 1, it will release the lock and let other threads waiting for the lock have chance to run the same block of code. Am I correct?
If I am correct, I have a further question, a "synchronized" block means a set of operations that can not be interrupted, but when invoke "wait" method, the thread running in the "synchronized" block will be terminated (interrupted) by release the lock and other threads will have chance to enter the "synchronized" block. In short, the execution inside the "synchronized" block will be interrupted. So, I wonder whether the above piece of code is correct and conforms to the principle of "synchronized". And what is its real use?
Thanks in advance,
George

Thanks, pkwooster buddy!You're welcome
I just wondered whether "wait inside a synchronized
block" technology is thread safe. Please look at the
following example,wait and synchronized are thread safe.
public class Foo {
int mVal= 0;
public final Object mLock = ...;
public void doIt() {
synchronized(mLock) {
mVal = ...;
mLock.wait();
if (mVal == ...) {
// do something
} else {
// do something else
}If we have two threads, T1 and T2, enter the block in
their respective order, and T1 is woken up first, T2
may have tampered with T1's execution path because T2
changed mVal while T1 was asleep. So T2 manipulate
instance field mVal is a side-effect.when you do the wait() you give up the lock and the other threads get a chance to run. When you exit the wait() you get the new value of the myVal variable which may have been changed. This is exactly what you want. To make that not thread save you could do
int temp = myVal;
wait();
if(temp == ...)in this case the variable temp contains the old vale of myVal.
>
I think the most safest way is never wait inside a
synchronized block, but it is less efficient. What do
you think about the trick of wait inside a
synchronized block? I think you are very experienced
in thread field from your kind reply.
Thanks in advance,
Georgewait(), notify() and notifyAll() are very useful. You need them when you want threads to cooperate in an predictable manner. I recommend that you review the section on consumer and producer classes and wait and notify in the Threads Tutorial. It gives good examples. For a practical example of this you could also take a look at my Multithreaded Server Example. It implements a simple chat server that switches String messages. Look especially at the send(), doSend() and popMessage() methods in the StreamConnection class. These allow the receive thread of one user to send messages out using the send thread of another user. If you have questions about that example, please post the questions on that topic.
Hope this helps.

Few questions about Calendar / SimpleDateFormat classes

Hi.
I have few questions about Calendar class:
1. How can I get only date representation (without the time)?
after doing:
Calendar cal = Calendar.getInstance();
I tried to clear unecessary fields:
cal.clear(Calendar.SECOND);
cal.clear(Calendar.MINUTE);
cal.clear(Calendar.HOUR_OF_DAY);
But after printing the time, it seems that the HOUR was not cleared:
SimpleDateFormat sdf1 = new SimpleDateFormat("dd/MM/yyyy HH:mm:ss");
System.out.println(sdf1.format(cal.getTime()));
---> 03/11/2004 17:00:00
Am I missing somthing?
2. I want to make sure that two different formats couldn't be compared. i.e., if I'll try to parse one String according to different format -- ParseException will be thrown.
String date = "19/04/2004 13:06:10";
SimpleDateFormat sdf = new SimpleDateFormat("dd/MM/yyyy");
Date dateObj = sdf.parse(date);
However, even though that formats are different, no exception is thrown and the return Date is 19/04/2004 00:00:00.
How can I cause to exception to be thrown if formats are not identical?
Thanks in advanced

The Calendar class has a few of what Microsoft would call 'features'. :^)
To clear the time, call:
calendar.set(calendar.get(Calendar.YEAR), calendar.get(Calendar.MONTH), calendar.get(Calendar.DAY), 0, 0, 0);
If you want 'pessimistic' rather than 'optimistic' date parsing, you have two options:
1) Call calendar.setLenient(false);
See if that is strict enough for your needs, otherwise:
2) Write code to ensure a stricter parsing of the string passed in. For example, you could very simply check the length of the string to determine if time data is also present.
When parsing, a string like '02/29/02' would parse to the date '03/01/02'. Setting lenient to false may fix this, but the surest way is to do some testing and then make the parsing more pessimistic where it suits your needs.
- Saish
"My karma ran over your dogma." - Anon

Question about using container class - Map

hi all
i have a question about the following :
i have two sets of data, both use a common filed as key for the map:
A = {a,b,d,f,g,h} and B = {a,d,e,g,i,k}, each key has a value associated with.
i need to find out the commonality, and the difference - i.e fields are in map A but not in Map B, and some fields that are in B but not in A. is there any quick method that comes with the collection class that can calculate this? thanks

yes, the keyset will be used to access values.
so, SetA.retainAll(SetB) will modify the SetA so that it contains keys that are in both SetA and SetB - the intersection.
and SetA.removeAll(SetB) will modify the SetA so that it only contains keys that are in SetA, but not in SetB?
correct?

Question about synchronized

Hello everyone,
Suppose I have a method which is a synchronized method (for example, method Foo in the following code block), and in this synchronized method another method which is not synchronized is invoked (for example, method Goo in the following code block). I am wondering when executing in Goo from Foo, whether we still have the synchronized feature (i.e. obtain the lock of "this" object).
synchronized public Foo()
    Goo();
private Goo()
}Thanks in advance,
George

Kaj,
Hi, pls see my post above - I replied to the same question in the other forum and I was pretty sure about my answer until I found it different from ur reply here.
Here is my reply. Can u kindly guide me to what's wrong in it ?
if a lock on a sync-ed method extends to all methods being called from it, then the jvm (bcos it has no means of knowing until runtime the method calls from a sync method) has to lock all the methods in a class that has atleast one sync method.
What this would mean potentially is that the effect of sync-ing one method would be equal to sync-ing all the methods bcos any of the other methods could be potentially called from the sync-ed one.
The example below is co-erced, but serves to illustrate the point.
     public class Test
    synchronized public void Foo()
        Goo();
    synchronized public void Moo()
        Goo();
        notify();
    private void Goo()
        System.out.println(Thread.currentThread().getName()+" accessing goo");
        try
            if(Thread.currentThread().getName().equals("Th one"))
                 //force thread1 to wait , lets see if thread2 can access it while thread1's waiting on it
                wait();
        catch(Exception e){}
        System.out.println(Thread.currentThread().getName()+" returning from goo");
//first thread
public class MyT implements Runnable
    Test t;
    public MyT(Test t)
        this.t = t;
    /* (non-Javadoc)
     * @see java.lang.Runnable#run()
    public void run()
        t.Foo();
//second thread
public class MyT2 implements Runnable
    Test t;
    public MyT2(Test t)
        this.t = t;
    /* (non-Javadoc)
     * @see java.lang.Runnable#run()
    public void run()
        t.Moo();
//Main
public class Main
    public static void main(String[] args)
        Test t = new Test();
        Thread th = new Thread(new MyT(t),"Th one");
        Thread th1 = new Thread(new MyT2(t), "Th two");
        th.start();
        th1.start();
And the o/p I got was
Th one accessing goo/*thread one in goo, is in wait() coz of the funny if clause there*/
Th two accessing goo /*there's no lock on goo, else with thread one in wait(), this wouldve been impossible*/
Th two returning from goo/*Thread 2 is over with Goo(), will go back and notify Thread 1*/
Th one returning from goo/*Thread 1 is notified, wakes up and returns :-)*/
tx,
ram.

Question about Abstract,Final Class

when we are using final keyword along with class definition .we are making that class final, means we can’t extend that class. The same thing happens when we make our constructors private. From usability perspective both are same ? or is there any difference?
Secondly accounting to java syntax a class can be either final or abstract and not both .Then why language specification doesn't impose any restriction on making abstract classes constructor private. We can create an Abstract class with private Constructor (Basically utility class with all methods are static) ,to make this class Singleton .This situation is equal to abstract final class ?
Thanks,
Paul.

EJP wrote:
when we are using final keyword along with class definition .we are making that class final, means we can’t extend that class. The same thing happens when we make our constructors private.No it doesn't.
Secondly accounting to java syntax a class can be either final or abstract and not both.Correct.
Then why language specification doesn't impose any restriction on making abstract classes constructor private.Why should it? That proposition doesn't follow from your previous sentence.I think OP is asking about this case
public abstract class WTF {
private WTF() {}
private WTF(...) {}
}where the class is abstract and all the c'tors are final. It's an abstract class that cannot be extended, much like a final abstract class would be if it were allowed. So, since purpose of abstract classes is to be extended, the OP seems to be asking, "Why can we, in this fashion, create an abstract class that cannot be extended?"
I don't know the answer to that, but I would guess that, while final is an explicit syntactical element for the purpose of making a class non-extensible, in the all-private-c'tors case, the non-extensibility is simply a side effect, and probably enough of a corner case that it wasn't worth the effort to explicitly forbid it in the spec.
I also think it's something not worth thinking much about. It was certainly not a key point of some grand design.

Beginner's question about naming helper classes

Hello,
I am building a pretty basic multi-tier application. I want to reduce a much as possible the dependencies between the web tier and the ejb tier. Ideally I would like for the web tier to construct a request object (of class MyRequest) and send it to the ejb tier that would then return a results object (of class MyResults).
These would therefore be the only two classes that BOTH the ejb and web tier would be aware of. All other classes would either belong to the web tier OR the ejb tier.
My question relates to the naming of these two classes. I would like to have different package names for the two classes as follows:
-web-tier request object will be of type com.mycompany.web.MyRequest
-ejb-tier request object will be of type com.mycompany.ejb.MyRequest
-web-tier results object will be of type com.mycompany.web.MyResults
-ejb-tier results object will be of type com.mycompany.ejb.MyResults
I am getting errors with that. Is there a way around it?
Thanks in advance,
Julien Martin.

Hello PinkyDead and thanks for your answer,
I got the errors on another project and can't remember which. Class Cast Exceptions perhaps. I had to change the package names so that they matched in the web tier and the ejb tier. (which is what I don't like).
As far as the new project is concerned I am trying to design the application as well as possible and your comments are therefore useful. Better getting the design right before starting coding.
Can you or anyone tell me what architecture would be better that the one I proposed?? Don't you agree that at least some classes have to be shared by the web tier and the ejb tier? How can it be otherwise?
Thanks in advance,
Julien.

Question about Inner/Outer classes

Hello!
I'm still preparing for java cert exam and have a question. If I have Outer and Inner classes like this:
Outer o = new Outer();
Outer.Inner i = o.new Inner();Does this mean that 'i' has a reference to both Outer and Inner classes. Moreover, does 'i' have a reference to the 'this' of 'o'?
Thanks,
Victor.

"i" doesn't have a reference. "i" is a reference. I think you are trying to express this:
import java.lang.reflect.*;
public class Outer {
    class Inner {
        private int foo;
    public static void main(String[] args) {
        Outer o = new Outer();
        Outer.Inner i = o.new Inner();
        for(Field field : i.getClass().getDeclaredFields()) {
            System.out.format("%s %s %s;%n", Modifier.toString(field.getModifiers()),
                ((Class)field.getGenericType()).getName(), field.getName());
}

Question about synchronized static function.

Hi, everyone!
The synchronized statement often deals with an object (called monitor).
But a static function can be invoked without an object.
So if a static function is declared as static, what is the function of
static and synchronized here? Which lock it is using?
Can you give me a simple expplanation?
regards,
George

Hence try and avoid synchronizing static methods.
Try and use non static method if you are doing
multithreaded programming. That way you can have lockWell, I avoid static methods, not just when doing multithreading, but not for the reason you mention.
objects which allow non interfering methods to be
accessed in parallelThats easy to do with static methods anyway:
class Foo
static Object lock1 = new Object();
static Object lock2 = new Object();
static void method1()
synchronized( lock1 )
static void method2()
synchronized( lock2 )
}Maybe I just missunderstod you...

Looking for tjacobs to answer a few questions about his StandardPrint Class

Hello,
Hopefully tjacobs will see this post and answer it as soon as he has time. I am trying to use your Standard Print class in a program I am developing to Print a JPanel in a JScrollPane and am running into trouble. I will have more info for you if you can find it in your heart to reply.
Thank you,
Ray Jaramillo

I'm in exactly the same boat as you... gonna be using it as my primary computer, graduating college soon, and my first mac. I'll be going with the 13 inch MacBook Air.
The Pro and the Air are nearly identical performance-wise and so that means the only advantage of the Pro is the storage space (unless you count the cd drive as an advantage). The Air is way more portable and has a better display, I don't know about you but my current cd drive rarely gets used so might as well get rid of it.
I have also compared them side-by-side in the store and that was really what won me over to the Air. The 11-inch Air was there too but the screen just seems too small, especially if you are doing work/writing essays etc.
Hope that helped

Question about synchronized re-entry

Suppose I have the following code:
* method performs actions that require synchronization on state but
* the programmer ensures manually that it is only called in a context
* where we are already synchronized on state
private void veryPrivateMethod() {
    //synchronized(state) { (*) required?
    if (state == 999) {
        // perform actions that require synchronization on state
public void anotherMethod() {
    synchronized(state) {
        if (state == 42)
            state = 43;
        veryPrivateMethod();
}See (*): If I ensure manually that veryPrivateMethod will only be called in a context where we are already synchronized on state, do I still need to specifically synchronize on state in the method (reentrant synchronization), or will it be better performance-wise and still safely synchronized if I omit the synchronized statement in veryPrivateMethod()?
Thanks!

gimbal2 wrote:
932196 wrote:
I synchronize on it whenever I change / compare-change it. Is this a bad ideaThat's a maybe. Its not a bad idea, as long as you realize that doing this has no special meaning and will offer no special protection. You can synchronize on just about any object you want, as long as other threads synchronize on the exact same object when wanting to execute the same section(s) of volatile code.the comment in the example is:
// perform actions that require synchronization on statethis leads me to believe that the code is not correctly synchronized currently. like i said, changing the reference to the object you are synchronizing on pretty much always leads to broken code.

A question about synchronized Lists...

If I have a synchronized list, and I have a synchronized block in one thread with a lock on that list (synchronized(List)) am i supposed to be able to add an item to that list directly (using the List.add() method) from another thread while I'm still in the synchronized block previously described?
I have some code here where I tried to test that and the answer that I got is that it is possible while (I believe) it's not possible theoretically speaking.
import java.util.*;
public class Warehouse implements Runnable{
     List<String> items = Collections.synchronizedList(new ArrayList<String>());
//     List<String> items = new ArrayList<String>();
     public static void main(String[] args){
          Warehouse w = new Warehouse();
          Manager m = new Manager(w);
          Thread one = new Thread(w);
          Thread two = new Thread(m);
          one.start();
          two.start();
     public void add(String t){
          synchronized(items){
               for(int i=0; i<2; ++i){
                    System.out.println("Warehouse added item...");
                    try{
                    Thread.sleep(500);
                    }catch(InterruptedException ie){
                         ie.printStackTrace();
               items.add(t);
               items.notify();
     public void run(){
          for(int i=0; i<1; ++i){
               synchronized(items){
                    while(items.isEmpty()){
                         try{
                         System.out.println("Warehouse waiting...");
                         items.wait();
//                         System.out.println("Warehouse after waiting...");
                         for(int j=0; j<2; ++j){
                              System.out.println("Warehouse after waiting...");
                              try{
                              Thread.sleep(500);
                              }catch(InterruptedException ie){
                                   ie.printStackTrace();
                         }catch(InterruptedException ie){
                              ie.printStackTrace();
class Manager implements Runnable{
     Warehouse w;
     public Manager(Warehouse w){
          this.w = w;
     public void run(){
          for(int i=0; i<1; ++i){
               System.out.println("Manager Adding item...");
               w.add("articulo");
               for(int j=0; j<5; ++j){
                    w.items.add("jeje");
                    try{
                    Thread.sleep(500);
                    }catch(Exception ex){ex.printStackTrace();}
                    System.out.println("Manager adding to items");
               System.out.println("Manager finished with warehouse...");
}Output:
Warehouse waiting...
Manager Adding item...
Warehouse added item...
Warehouse added item...
*Warehouse after waiting...*
*Manager adding to items*
*Warehouse after waiting...*
*Manager adding to items*
Manager adding to items
Manager adding to items
Manager adding to items
Manager finished with warehouse...The code in "bold" (with the tokens) isn't supposed to happen concurrently as I understand it since the "warehouse after waiting" statement it's in the synchronized block with a lock on the items list and the "manager adding to items" statement is adding an item to that list through a direct reference to that list...

When the manager thread notifies the items (warehouse thread), the warehouse thread does not have to wake up automatically, it has to compete for the items lock then wait for its Thread context to execute. Which means that the manage is able to continue to execute - adding to items manually, and eventually sleep()ing. When the manager thread sleeps that is a good place for a Thread context switch, and if the JVM takes that opportunity to do so, then the warehouse thread regains the lock and would prevent further access to items. But at some point the Thread context will switch again (probably when warehouse thread goes to sleep). Now the manager thread can wake up, print that it is working because that code is not protected by any synchronized lock. Then when it attempts to add to items again it blocks. Another thread context switch allows the warehouse to complete its work, finish up, and release the items lock. Manager can then get the lock on items and finish its work. The results would be exactly as you display them. Here is a possible walk-through of the threading paths taken:
Thread == Warehouse
i = 0
lock on items
    items.isEmpty == true
      print "Warehouse is waiting"
      items.wait (unlock items)
/* CONTEXT SWITCH */
Thread == Manager
i = 0
print ""Manager Adding item..."
locking items
    i = 0
      printing "Warehouse added item..."
      sleeping 500
    i = 1
      printing "Warehouse added item..."
      sleeping 500
    i ! < 2
    locking items
      adding to items
    unlocking items
    notifying items (wait will unlock eventually)
unlocking items
j = 0
    locking items
      adding to items
    unlocking items
    sleeping 500
/* CONTEXT SWITCH */
Thread == Warehouse
      items.wait (regain lock and get context)
      j = 0
        prinitng "Warehouse after waiting..."
        sleeping 500
/* CONTEXT SWITCH */
Thread == Manager
    sleeping 500 is done
    printing "Manager adding to items"
j = 1
    locking items (blocking until items is available)
/* CONTEXT SWITCH */
Thread == Warehouse
        sleeping 500 is done
      j = 1
        printing "Warehouse after waiting..."
      j ! < 2
    itemse ! isEmpty
unlocking items
i ! < 1
/* End of Warehouse Thread */
/* CONTEXT SWITCH */
Thread == Manager
    locking items (lock available)
      adding to itemes
    unlock items
    sleeping 500
    printing "Manager adding to items"
j = 2
    locking items
      adding to itemes
    unlock items
    sleeping 500
    printing "Manager adding to items"
j = 3
    locking items
      adding to itemes
    unlock items
    sleeping 500
    printing "Manager adding to items"
j = 4
    locking items
      adding to itemes
    unlock items
    sleeping 500
    printing "Manager adding to items"
j ! < 5
printing "Manager finished with warehouse..."
i ! < 1
/* End of Manager Thread */So the theory an practice are the same - you can not access the items object when it is locked by another thread. The problem is that your test does not demonstrate when locks are held and released - some of the output you expect to be protected is not and so you mis-interpret the code.

A question about the Thread class

the code below are two way of invoke a thread to run:
A:
public class Test {
public static void main(String args[]) {
     Thread myThread = new Thread(){
          public void run(){
               System.out.println("in myThread! \n");
               System.out.println(Thread.currentThread());
     Thread th = new Thread(myThread,"new Thread");
     th.start();
B:
public class Test {
public static void main(String args[]) {
     Thread myThread = new Thread(){
          public void run(){
               System.out.println("in myThread! \n");
               System.out.println(Thread.currentThread());
     myThread.start();
i have read many free software's source code, and find many people would like to implement it like A instead of B
so my question is why A is better than B.
thanks!!

A is actually using the constructor which accepts a Runnable. In your case, since you are overriding only run(), you could have simply implemented the Runnable and created the thread th. However, in this model you are using a Runnable you are actually separating the task from the Thread. In B, thread is actually the task which couples them tightly. Therefore, from design point of view, A is recommended with Runnable rather than a Thread unless you have specific reasons to do that.

Question about synchronized singleton classes

Similar Messages

Maybe you are looking for