CON11-J. Do not assume that declaring an object volatile guarantees visibility of its members

According to the Java Language Specification [[JLS 05]], section, 8.3.1.4 volatile Fields:

A field may be declared volatile, in which case the Java memory model (§17) ensures that all threads see a consistent value for the variable.

Notably, this applies only to fields and not to the contents of objects that are declared volatile. A thread may not observe a recent write to the object's field from another thread.

Declaring an object volatile in order to ensure visibility of the most up-to-date object state does not work without the use of explicit synchronization, unless the object is [immutable].

In the absence of synchronization, the effect of declaring a field volatile is that when one thread sets the field to a new value, other threads can see the new object reference immediately. If the referenced object is immutable, this has the effect that other threads also see a consistent view of the state of the object. However, if the object is mutable, other threads may see a partially-constructed object, or an object in a (temporarily) inconsistent state [[Goetz 07]]. Declaring the object volatile does not prevent this issue.

Technically the object does not have to be strictly immutable. If it can be proved that the object is thread-safe by design, then the field that will hold its reference may be declared as volatile.

Noncompliant Code Example (Arrays)

This noncompliant code example shows an array object (arrays are objects in Java) that is declared volatile. It appears that when, a value is written by a thread to one of the array elements, it will be visible to other threads immediately. This is misleading because the volatile keyword just makes the array reference visible to all threads and does not affect the actual data contained within the array.

volatile int[] arr = new int[20];
// ...
arr[1] = 10;

It is possible for one thread to set a new value to arr[1] while another thread attempts to read the value of arr[1], with the result that the reading thread receives an inconsistent value for arr[1].

Compliant Solution (AtomicIntegerArray)

This compliant solution suggests using the java.util.concurrent.atomic.AtomicIntegerArray concurrency utility. Using its set(index, value) method ensures that the write is atomic and the resulting value is immediately visible to other threads. The other threads can retrieve a value from a specific index by using the get(index) method.

AtomicIntegerArray aia = new AtomicIntegerArray(5);
// ...
aia.set(1, 10);

Compliant Solution (synchronization)

To ensure visibility, accessor methods may synchronize access while performing operations on nonvolatile elements of an array which is declared as volatile.

volatile int[] arr = new int[20];
// ...
synchronized(lock) {
  arr[1] = 10; 
}

Noncompliant Code Example (Mutable object)

This noncompliant code example declares an instance field of type Properties as volatile. The field can be mutated using the put() method. This makes objects of class Foo mutable.

class Foo {
  private volatile Properties properties;

  public Foo() {
    properties = new Properties();
    // Load some useful values into properties
  }

  public String get(String s) {
    return properties.getProperty(s);
  }

  public void put(String key, String value) {
    // Perform validation of value before inserting
    properties.setProperty(key, value);
  }
}

This class permits a race condition. If one thread calls get() while another calls put(), the first thread may receive a stale value, or an internally inconsistent value from the properties object because the operations within put() are nonatomic. Declaring the object volatile does not prevent this data race.

Even if the client thread sees the new reference to properties, the object state that it observes may change in the meantime.

Because the object is not immutable, it is unsafe for use in a multi-threaded environment.

Compliant Solution (immutable)

This compliant solution renders the Foo class immutable. Consequently, once it is properly constructed, no thread can modify properties and cause a race condition.

class Foo {
  private final Properties properties;

  public Foo() {
    properties = new Properties();
    // Load some useful values into properties
  }

  public String get(String s) {
    return properties.getProperty(s);
  }
}

The drawback of this solution is that the put() method cannot be accommodated if the goal is to ensure immutability. The Foo class is [immutable] because all its fields are final and the properties field is being safely published.

Compliant Solution (the cheap read-write lock trick)

This compliant solution uses method synchronization to ensure thread safety. It declares the properties field as volatile to guard retrievals that use the getter method. A synchronized setter method is used to set the value of the Properties object.

public class Foo {
  private volatile Properties properties;

  public Foo() {
    properties = new Properties();
    // Load some useful values into properties
  }

  public String get(String s) {
    return properties.getProperty(s);
  }

  public synchronized void put(String key, String value) {
    // Perform validation of value
    properties.setProperty(key, value);
  }
}

This compliant solution has the advantage that it can accommodate the setter method. Declaring the object as volatile for safe publication using getter methods is cheaper in terms of performance, than synchronizing the getters. However, synchronizing the setter methods is mandatory because they typically consist of multiple operations.

Noncompliant Code Example (mutable sub-object)

This noncompliant code example declares the field FORMAT as volatile. However, the field stores a reference to a mutable object, DateFormat.

class DateHandler {
  private static volatile DateFormat FORMAT =
    DateFormat.getDateInstance(DateFormat.MEDIUM);

  public static Date parse(String str) throws ParseException {
    return FORMAT.parse(str);
  }

  public static void main(String arg[]) throws ParseException {
    DateHandler.parse("Jan 1, 2010");
  }
}

In the presence of multiple threads, this results in subtle thread safety issues. For instance, a thread may observe correctly formatted output for a completely different date. This is because threads are allowed to change the state of the object instance by invoking the parse() method. The parse() method does not defensively copy the instance field before returning it (a violation of [OBJ11-J. Defensively copy private mutable class members before returning their references]). Even if FORMAT was declared as final, a thread could see the object in an inconsistent state.

Compliant Solution (instance per call/defensive copying)

This compliant solution creates and returns a new DateFormat instance for every invocation of the parse() method.

class DateHandler {
  public static Date parse(String str) throws ParseException {
    DateFormat format = DateFormat.getDateInstance(DateFormat.MEDIUM);
    return format.parse(str);
  }

  public static void main(String arg[]) throws ParseException {
    DateHandler.parse("Jan 1, 2010");
  }
}	

This does not violate [OBJ11-J. Defensively copy private mutable class members before returning their references] because the class no longer contains internal mutable state, but a local field, format.

Compliant Solution (synchronization)

This compliant solution synchronizes the parse() method and consequently, the class DateHandler is made thread-safe. There is no requirement for declaring the FORMAT field as volatile.

class DateHandler {
  private static DateFormat FORMAT =
    DateFormat.getDateInstance(DateFormat.MEDIUM);

  public static synchronized Date parse(String str) throws ParseException {
    return FORMAT.parse(str);
  }

  public static void main(String arg[]) throws ParseException {
    DateHandler.parse("Jan 1, 2010");
  }
}

According to the Java API [API 06], class DateFormat documentation:

Date formats are not synchronized. It is recommended to create separate format instances for each thread. If multiple threads access a format concurrently, it must be synchronized externally.

However, performance wise, synchronization is usually a more expensive option as compared to other compliant solutions.

Compliant Solution (ThreadLocal storage)

This compliant solution uses a ThreadLocal object to store one DateFormat instance per thread.

class DateHandler {
  private ThreadLocal<DateFormat> df = new ThreadLocal<DateFormat>() {
    protected DateFormat initialValue() {
      return DateFormat.getDateInstance(DateFormat.MEDIUM);
    }
  };
  // ...
}

This technique may be ineffective if the same DateFormat instance is required to be shared between multiple threads, but there is little reason to do so.

Risk Assessment

Failing to synchronize access to shared mutable data can cause different threads to observe different states of the object.

Automated Detection

TODO

Related Vulnerabilities

Search for vulnerabilities resulting from the violation of this rule on the CERT website.

References

[[Goetz 07]] Pattern #2: "one-time safe publication"
[[Miller 09]] Mutable Statics
[[API 06]] Class java.text.DateFormat
[[JLS 05]]

[!The CERT Sun Microsystems Secure Coding Standard for Java^button_arrow_left.png!]      [!The CERT Sun Microsystems Secure Coding Standard for Java^button_arrow_up.png!]      [!The CERT Sun Microsystems Secure Coding Standard for Java^button_arrow_right.png!]