The usual way of initializing an object is to use a constructor. Sometimes it is required to limit the number of instances of the sub-object to just one (this is similar to a singleton, however, the sub-object may or may not be static
). In addition, a technique called lazy initialization is used to defer the construction of the sub-object until it is actually required.
For these purposes, instead of a constructor, a class or an instance method should be used for initialization, depending on whether the sub-object is static
or not. The method checks whether the instance has already been created and if not, creates it. If the instance already exists, it simply returns it. This is shown below:
// Correct single threaded version using lazy initialization class Foo { private Helper helper = null; public Helper getHelper() { if (helper == null) { helper = new Helper(); } return helper; } // ... }
In a multi-threading scenario, the initialization must be synchronized so that two or more threads do not create multiple instances of the sub-object. The code shown below is correctly synchronized, albeit slower than the previous, single threaded code example.
// Correct multithreaded version using synchronization class Foo { private Helper helper = null; public synchronized Helper getHelper() { if (helper == null) { helper = new Helper(); } return helper; } // Other functions and members... }
The double checked locking (DCL) idiom is sometimes used to provide lazy initialization in multithreaded code. In a multi-threading scenario, lazy initialization is supplemented by reducing the cost of synchronization on each method access by limiting the synchronization to the case where the instance is to be created and forgoing it when retrieving an already created instance.
The double-checked locking pattern eliminates method synchronization and uses block synchronization. It strives to make the previous code example faster by installing a null
check before attempting to synchronize. This makes expensive synchronization dispensable for the common case of retrieving the value. The noncompliant code example shows the originally proposed DCL pattern.
According to the Java Memory Model (discussion reference) [[Pugh 04]]:
... writes that initialize the
Helper
object and the write to thehelper
field can be done or perceived out of order. As a result, a thread which invokesgetHelper()
could see a non-null reference to ahelper
object, but see the default values for fields of thehelper
object, rather than the values set in the constructor.Even if the compiler does not reorder those writes, on a multiprocessor the processor or the memory system may reorder those writes, as perceived by a thread running on another processor.
This makes the originally proposed double-checked locking pattern insecure.
The rule CON26-J. Do not publish partially-constructed objects discusses further the possibility of a non-null reference to a helper
object with default values for fields in the helper
object.
Noncompliant Code Example
This noncompliant code example uses the incorrect form of the double checked locking idiom.
// "Double-Checked Locking" idiom class Foo { private Helper helper = null; public Helper getHelper() { if (helper == null) { synchronized(this) { if (helper == null) { helper = new Helper(); } } } return helper; } // other functions and members... }
Compliant Solution (volatile
)
This compliant solution declares the Helper
object as volatile
.
// Works with acquire/release semantics for volatile // Broken under JDK 1.4 and earlier class Foo { private volatile Helper helper = null; public Helper getHelper() { if (helper == null) { synchronized(this) { if (helper == null) { helper = new Helper(); // If the helper is null, create a new instance } } } return helper; // If helper is non-null, return its instance } }
JDK 5.0 allows a write of a volatile
variable to be reordered with respect to a previous read or write. A read of a volatile
variable cannot be reordered with respect to any following read or write. Because of this, the double checked locking idiom can work when helper
is declared volatile
. If a thread initializes the Helper
object, a happens-before relationship is established between this thread and another that retrieves and returns the instance. [[Pugh 04]] and [[Manson 04]]
Compliant Solution (immutable
)
In this solution the Foo
class is unchanged, but the Helper
class is immutable. In this case, the Helper
class is guaranteed to be fully constructed before becoming visible. The object must be truly immutable; it is not sufficient for the program to refrain from modifying the object.
public class Helper { private final int n; public Helper(int n) { this.n = n; } // other fields & methods, all fields are final } class Foo { private Helper helper = null; public Helper getHelper() { if (helper == null) { synchronized(this) { if (helper == null) { helper = new Helper(); // If the helper is null, create a new instance } } } return helper; // If helper is non-null, return its instance } }
Compliant Solution (static initialization)
This compliant solution initializes the helper
in the declaration of the static
variable [[Manson 06]].
class Foo { private static final Helper helper = new Helper(42); public static Helper getHelper() {return helper;} }
Variables declared static
are guaranteed to be initialized and made visible to other threads immediately. Static initializers also exhibit these properties. This approach should not be confused with eager initialization because in this case, the Java Language Specification guarantees lazy initialization of the class when it will be first used.
Compliant Solution (explicit static lazy initialization)
This compliant solution explicitly incorporates lazy initialization. It also uses a static
variable as suggested in the previous compliant solution. The variable is declared within a static
inner, Holder
class.
class Foo { static Helper helper; // Lazy initialization private static class Holder { static Helper helper = new Helper(); } public static Helper getInstance() { return Holder.helper; } }
This idiom is called the initialize-on-demand holder class idiom. Initialization of the Holder
class is deferred until the getInstance()
method is called, following which the helper
is initialized. The only limitation of this method is that it works only for static
fields and not instance fields. [[Bloch 01]]
Exceptions
EX1: Explicitly synchronized code (that uses method synchronization or proper block synchronization, that is, enclosing all initialization statements) does not require the use of double-checked locking.
EX2: "Although the double-checked locking idiom cannot be used for references to objects, it can work for 32-bit primitive values (e.g., int's or float's). Note that it does not work for long's or double's, since unsynchronized reads/writes of 64-bit primitives are not guaranteed to be atomic." [[Pugh 04]]. See [CON25-J. Ensure atomicity when reading and writing 64-bit values] for more Information.
Risk Assessment
Using incorrect forms of the double checked locking idiom can lead to synchronization issues.
Rule |
Severity |
Likelihood |
Remediation Cost |
Priority |
Level |
---|---|---|---|---|---|
CON22- J |
low |
probable |
medium |
P4 |
L3 |
Automated Detection
TODO
Related Vulnerabilities
Search for vulnerabilities resulting from the violation of this rule on the CERT website.
References
[[API 06]]
[[Pugh 04]]
[[Bloch 01]] Item 48: "Synchronize access to shared mutable data"
[[MITRE 09]] CWE ID 609 "Double-Checked Locking"
CON21-J. Facilitate thread reuse by using Thread Pools 11. Concurrency (CON) CON23-J. Address the shortcomings of the Singleton design pattern