Page History

Composite operations consisting Compound operations are operations that consist of more than one discrete operation are, by definition, non-atomic. For example, the Java expression x++ is non-atomic because it is a composite operation consisting of three discrete operations: reading the current value of x, adding one to it, and writing the new, incremented value back to x.

"Sequential consistency and/or freedom from data races still allows errors arising from groups of operations that need to be perceived atomically and are not." \[[JLS 05|AA. Java References#JLS 05]\]. In such cases, the {{java.util.concurrent}} utilities can be used to atomically manipulate a shared variable. If these utilities do not provide the required functionality in the form of atomic methods, operations that use the variable should be correctly synchronized. 

Note that, as with volatile, updated values are immediately visible to other threads when either of these two techniques is used. Synchronization provides a way to safely share object state across multiple threads without the need to reason about reorderings, compiler optimizations and hardware specific behavior.

This rule specifically deals with primitive operators such as ++. For atomicity of a grouping of calls to independently atomic methods of the existing Java API, see CON07-J. Do not assume that a grouping of calls to independently atomic methods is atomic.

The Java Language Specification also permits reads and writes of 64-bit values to be non-atomic though this is not an issue with most modern JVMs (see CON25-J. Ensure atomicity when reading and writing 64-bit values).

Noncompliant Code Example (post-decrement composite operation)

This noncompliant code example contains a data race that may result in the itemsInInventory field being incorrectly decremented.

private int itemsInInventory = 100;

public final int removeItem() {
  if (itemsInInventory > 0) {
    return itemsInInventory--;  // Returns new count of items in inventory
  } 
  return -1; // Error code  
}

For example, if the removeItem() method is concurrently invoked by two threads, the execution of these threads may be interleaved so that:

1. The first thread reads the current value of itemsInInventory (100).
2. The second thread reads the current value of itemsInInventory (100).
3. The first thread decrements the locally cached value (99).
4. The second thread decrements the locally cached value (99).
5. The first thread writes the locally cached value to itemsInInventory (99).
6. The second thread writes the locally cached value to itemsInInventory (99).

As a result, a decrement operation is "lost" and the itemsInInventory value is now incorrect.

Similarly, a returnItem() method that increments itemsInInventory is also nonatomic:

public final int returnItem() {
  itemsInInventory++;
}

Noncompliant Code Example (volatile)

This noncompliant code example attempts to resolve the problem by declaring itemsInInventory as volatile.

private volatile int itemsInInventory = 100;

public final int removeItem() {
  if (itemsInInventory > 0) {
    return itemsInInventory--;  // Returns new count of items in inventory
  } 
  return -1; // Error code  
}

Volatile variables are unsuitable when more than one read/write operation needs to be atomic. The use of a volatile variable in this noncompliant code example guarantees that once itemsInInventory has been updated, the new value is immediately visible to all threads that read the field. However, because the post decrement operator is nonatomic, even when volatile is used, the interleaving described in the previous noncompliant code example is still possible.

Similarly, the returnItem() method does not perform the increment operation atomically.

public final int returnItem() {
  itemsInInventory++;
}

Compliant Solution (java.util.concurrent.atomic classes)

This compliant solution uses a java.util.concurrent.atomic.AtomicInteger variable which allows certain composite operations to be performed atomically.

private final AtomicInteger itemsInInventory = new AtomicInteger(100);

private final int removeItem() {
  for (;;) {
    int old = itemsInInventory.get();
    if (old > 0) {
      int next = old - 1; // Decrement 
      if (itemsInInventory.compareAndSet(old, next)) {
        return next;  // Returns new count of items in inventory
      }
    } else {
      return -1; // Error code
    }
  }
}

According to the Java API \[[API 06|AA. Java References#API 06]\], class {{AtomicInteger}} documentation:

Wiki Markup
\[AtomicInteger is an\] {{int}} value that may be updated atomically. An {{AtomicInteger}} is used in applications such as atomically incremented counters, and cannot be used as a replacement for an {{Integer}}. However, this class does extend {{Number}} to allow uniform access by tools and utilities that deal with numerically-based classes.

The {{compareAndSet()}} method takes two arguments, the expected value of a variable when the method is invoked and the updated value. This compliant solution uses this method to atomically set the value of {{itemsInInventory}} to the updated value if and only if the current value equals the expected value. \[[API 06|AA. Java References#API 06]\] 

The returnItem() method can be fixed by using the java.util.concurrent.atomic.AtomicInteger.getAndIncrement() method.

public final int returnItem() {
  int temp;   
  if((temp = itemsInInventory.get()) == Integer.MAX_VALUE) { // Check for integer overflow
    return -1; // Error Code
  }

  if(temp == itemsInInventory.get()) {  
    return itemsInInventory.getAndIncrement();
  } else
    return -1;  // Error Code
  } 
}

It is necessary to use a local variable temp because there is a time-of-check to time-of-use (TOCTOU) condition between checking whether the inventory count is less than Integer.MAX_VALUE and using getAndIncrement() to increment it. Notably, this functionality could also be implemented by using the compareAndSet() method. The getAndIncrement() alternative is useful when control over setting the returned value must lie in the hands of the caller instead of the invoked method (returnItem()).

Compliant Solution (method synchronization)

This compliant solution uses method synchronization to synchronize access to itemsInInventory. Consequently, access to itemsInInventory is mutually exclusive and its state consistent across all threads.

private final int itemsInInventory = 100;

public final synchronized int removeItem() {
  if (itemsInInventory > 0) {
    return itemsInInventory--;  // Returns new count of items in inventory
  } 
  return -1; // Error Code 
}

similarly, the returnItem() method can be fixed by synchronizing it:

public final synchronized int returnItem() {
  if(itemsInInventory == Integer.MAX_VALUE) { // Check for integer overflow
    return -1; // Error Code
  } else {
    return itemsInInventory++;
  }
}

Synchronization is more expensive than using the optimized java.util.concurrent utilities and should only be used when the utilities do not contain the facilities (methods) required to carry out the atomic operation. When synchronizing, care must be taken to avoid deadlocks (see CON12-J. Avoid deadlock by requesting and releasing locks in the same order).

Compliant Solution (block synchronization)

Constructors and methods can use an alternative representation called block synchronization which synchronizes a block of code rather than a method, as highlighted in this compliant solution.

private final int itemsInInventory = 100;

public int removeItem() {
  synchronized(this) {
    if (itemsInInventory > 0) {
      return itemsInInventory--;  // Returns new count of items in inventory    
    } 
    return -1; // Error code
  }
}

similarly, the returnItem() method can be fixed by using block synchronization:

public final int returnItem() {
  synchronized(this) {
    if(itemsInInventory == Integer.MAX_VALUE) { // Check for integer overflow
      return -1; // Error Code
    } else {
      return itemsInInventory++;
    }
  }
}

Block synchronization is preferable over method synchronization because it reduces the duration for which the lock is held and also protects against denial of service attacks. Block synchronization requires synchronizing on an internal private lock object instead of the intrinsic lock of the class's object (see CON04-J. Use the private lock object idiom instead of the Class object's intrinsic locking mechanism).

When the number of items is 0 most of the time, the synchronized block may be moved inside the if condition to reduce the performance cost associated with synchronization. In that case, the variable itemsInInventory must be declared as volatile because the check to determine whether it is greater than 0 should rely on the latest value of itemsInInventory.

Compliant Solution (ReentrantLock)

This compliant solution uses a java.util.concurrent.locks.ReentrantLock to atomically perform the post-decrement operation.

private int itemsInInventory = 100;
private final Lock lock = new ReentrantLock();
    
public int removeItem() {
  Boolean myLock = false;
    
  try {
    myLock = lock.tryLock(); 
    	
    if (itemsInInventory > 0) {
      return itemsInInventory--;
    }  
  } finally {
    if (myLock) {
      lock.unlock();
    }
  }
  return -1; // Error code
}

Similarly, the returnItem() method can be made atomic:

public int returnItem() {
  Boolean myLock = false;
    
  try {
    myLock = lock.tryLock(); 
    	
    if (itemsInInventory == Integer.MAX_VALUE) { // Check for integer overflow
      return -1; 
    } else {
      return itemsInInventory++;
    }  
  } finally {
    if (myLock) {
      lock.unlock();
    }
  }
  return -1; // Error code
}

Code that uses this lock behaves similar to synchronized code that uses the traditional monitor lock. ReentrantLock provides several other capabilities, for instance, the tryLock() method does not block waiting if another thread is already holding the lock. The class java.util.concurrent.locks.ReentrantReadWriteLock can be used when some thread requires a lock to write information while other threads require the lock to simultaneously read the information.

Noncompliant Code Example (addition)

In this noncompliant code example, the two fields a and b may be set by multiple threads, using the setValues() method.

private volatile int a;
private volatile int b;

public int getSum() throws ArithmeticException {
  // Check for integer overflow
  if( b > 0 ? a > Integer.MAX_VALUE - b : a < Integer.MIN_VALUE - b ) { 
    throw new ArithmeticException("Not in range");
  }

  return a + b;
}

public void setValues(int a, int b) {
  this.a = a;
  this.b = b;
}

The getSum() method may return a different sum every time it is invoked from different threads. For instance, if a and b currently have the value 0, and one thread calls getSum() while another calls setValues(1, 1), then getSum() might return 0, 1, or 2. Of these, the value 1 is unacceptable; it is returned when the first thread reads a and b, after the second thread has set the value of a but before it has set the value of b.

Compliant Solution (addition)

This compliant solution synchronizes the setValues() method so that the entire operation is atomic.

private volatile int a;
private volatile int b;

public synchronized int getSum() throws ArithmeticException {
  // Check for integer overflow
  if( b > 0 ? a > Integer.MAX_VALUE - b : a < Integer.MIN_VALUE - b ) { 
    throw new ArithmeticException("Not in range");
  }

  return a + b;
}

public synchronized void setValues(int a, int b) {
  this.a = a;
  this.b = b;
}

Unlike the noncompliant code example, if a and b currently have the value 0, and one thread calls getSum() while another calls setValues(1, 1), getSum() may return return 0, or 2, depending on which thread obtains the intrinsic lock first. The locking guarantees that getSum() will never return the unacceptable value 1.

Risk Assessment

If operations on shared variables are not atomic, unexpected results may be produced. For example, there can be inadvertent information disclosure as one user may be able to receive information about other users.

Automated Detection

TODO

Related Vulnerabilities

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

References

\[[API 06|AA. Java References#API 06]\] Class AtomicInteger
\[[JLS 05|AA. Java References#JLS 05]\] [Chapter 17, Threads and Locks|http://java.sun.com/docs/books/jls/third_edition/html/memory.html], section 17.4.5 Happens-before Order, section 17.4.3 Programs and Program Order, section 17.4.8 Executions and Causality Requirements
\[[Tutorials 08|AA. Java References#Tutorials 08]\] [Java Concurrency Tutorial|http://java.sun.com/docs/books/tutorial/essential/concurrency/index.html]
\[[Lea 00|AA. Java References#Lea 00]\] Sections, 2.2.7 The Java Memory Model, 2.2.5 Deadlock, 2.1.1.1 Objects and locks
\[[Bloch 08|AA. Java References#Bloch 08]\] Item 66: Synchronize access to shared mutable data
\[[Daconta 03|AA. Java References#Daconta 03]\] Item 31: Instance Variables in Servlets
\[[JavaThreads 04|AA. Java References#JavaThreads 04]\] Section 5.2 Atomic Variables
\[[Goetz 06|AA. Java References#Goetz 06]\] 2.3. "Locking"
\[[MITRE 09|AA. Java References#MITRE 09]\] [CWE ID 667|http://cwe.mitre.org/data/definitions/667.html] "Insufficient Locking", [CWE ID 413|http://cwe.mitre.org/data/definitions/413.html] "Insufficient Resource Locking", [CWE ID 366|http://cwe.mitre.org/data/definitions/366.html]  "Race Condition within a Thread", [CWE ID 567|http://cwe.mitre.org/data/definitions/567.html]  "Unsynchronized Access to Shared Data"

. Expressions that include postfix or prefix increment (++), postfix or prefix decrement (--), or compound assignment operators always result in compound operations. Compound assignment expressions use operators such as *=, /=, %=, +=, -=, <<=, >>=, >>>=, ^= and |= [JLS 2015]. Compound operations on shared variables must be performed atomically to prevent data races and race conditions.

For information about the atomicity of a grouping of calls to independently atomic methods that belong to thread-safe classes, see VNA03-J. Do not assume that a group of calls to independently atomic methods is atomic.

The Java Language Specification also permits reads and writes of 64-bit values to be non-atomic (see rule VNA05-J. Ensure atomicity when reading and writing 64-bit values).

Noncompliant Code Example (Logical Negation)

This noncompliant code example declares a shared boolean flag variable and provides a toggle() method that negates the current value of flag:

final class Flag {
  private boolean flag = true;

  public void toggle() {  // Unsafe
    flag = !flag;
  }

  public boolean getFlag() { // Unsafe
    return flag;
  }
}

Execution of this code may result in a data race because the value of flag is read, negated, and written back.

Consider, for example, two threads that call toggle(). The expected effect of toggling flag twice is that it is restored to its original value. However, the following scenario leaves flag in the incorrect state:

As a result, the effect of the call by t₂ is not reflected in flag; the program behaves as if toggle() was called only once, not twice.

Noncompliant Code Example (Bitwise Negation)

The toggle() method may also use the compound assignment operator ^= to negate the current value of flag:

final class Flag {
  private boolean flag = true;

  public void toggle() {  // Unsafe
    flag ^= true;  // Same as flag = !flag;
  }

  public boolean getFlag() { // Unsafe
    return flag;
  }
}

This code is also not thread-safe. A data race exists because ^= is a non-atomic compound operation.

Noncompliant Code Example (Volatile)

Declaring flag volatile also fails to solve the problem:

final class Flag {
  private volatile boolean flag = true;

  public void toggle() {  // Unsafe
    flag ^= true;
  }

  public boolean getFlag() { // Safe
    return flag;
  }
}

This code remains unsuitable for multithreaded use because declaring a variable volatile fails to guarantee the atomicity of compound operations on the variable.

Compliant Solution (Synchronization)

This compliant solution declares both the toggle() and getFlag() methods as synchronized:

final class Flag {
  private boolean flag = true;

  public synchronized void toggle() {
    flag ^= true; // Same as flag = !flag;
  }

  public synchronized boolean getFlag() {
    return flag;
  }
}

This solution guards reads and writes to the flag field with a lock on the instance, that is, this. Furthermore, synchronization ensures that changes are visible to all threads. Now, only two execution orders are possible, one of which is shown in the following scenario:

The second execution order involves the same operations, but t₂ starts and finishes before t₁.
Compliance with LCK00-J. Use private final lock objects to synchronize classes that may interact with untrusted code can reduce the likelihood of misuse by ensuring that untrusted callers cannot access the lock object.

Compliant Solution (Volatile-Read, Synchronized-Write)

In this compliant solution, the getFlag() method is not synchronized, and flag is declared as volatile. This solution is compliant because the read of flag in the getFlag() method is an atomic operation and the volatile qualification assures visibility. The toggle() method still requires synchronization because it performs a non-atomic operation.

final class Flag {
  private volatile boolean flag = true;

  public synchronized void toggle() {
    flag ^= true; // Same as flag = !flag;
  }

  public boolean getFlag() {
    return flag;
  }
}

This approach must not be used for getter methods that perform any additional operations other than returning the value of a volatile field without use of synchronization. Unless read performance is critical, this technique may lack significant advantages over synchronization [Goetz 2006].

Compliant Solution (Read-Write Lock)

This compliant solution uses a read-write lock to ensure atomicity and visibility:

final class Flag {
  private boolean flag = true;
  private final ReadWriteLock lock = new ReentrantReadWriteLock();
  private final Lock readLock = lock.readLock();
  private final Lock writeLock = lock.writeLock();

  public void toggle() {
    writeLock.lock();
    try {
      flag ^= true; // Same as flag = !flag;
    } finally {
      writeLock.unlock();
    }
  }

  public boolean getFlag() {
    readLock.lock();
    try {
      return flag;
    } finally {
      readLock.unlock();
    }
  }
}

Read-write locks allow shared state to be accessed by multiple readers or a single writer but never both. According to Goetz [Goetz 2006]:

In practice, read-write locks can improve performance for frequently accessed read-mostly data structures on multiprocessor systems; under other conditions they perform slightly worse than exclusive locks due to their greater complexity.

Profiling the application can determine the suitability of read-write locks.

Compliant Solution (AtomicBoolean)

This compliant solution declares flag to be of type AtomicBoolean:

import java.util.concurrent.atomic.AtomicBoolean;

final class Flag {
  private AtomicBoolean flag = new AtomicBoolean(true);

  public void toggle() {
    boolean temp;
    do {
      temp = flag.get();
    } while (!flag.compareAndSet(temp, !temp));
  }

  public AtomicBoolean getFlag() {
    return flag;
  }
}

The flag variable is updated using the compareAndSet() method of the AtomicBoolean class. All updates are visible to other threads.

Noncompliant Code Example (Addition of Primitives)

In this noncompliant code example, multiple threads can invoke the setValues() method to set the a and b fields. Because this class fails to test for integer overflow, users of the Adder class must ensure that the arguments to the setValues() method can be added without overflow (see NUM00-J. Detect or prevent integer overflow for more information).

final class Adder {
  private int a;
  private int b;

  public int getSum() {
    return a + b;
  }

  public void setValues(int a, int b) {
    this.a = a;
    this.b = b;
  }
}

The getSum() method contains a race condition. For example, when a and b currently have the values 0 and Integer.MAX_VALUE, respectively, and one thread calls getSum() while another calls setValues(Integer.MAX_VALUE, 0), the getSum() method might return either 0 or Integer.MAX_VALUE, or it might overflow. Overflow will occur when the first thread reads a and b after the second thread has set the value of a to Integer.MAX_VALUE but before it has set the value of b to 0.

Note that declaring the variables as volatile fails to resolve the issue because these compound operations involve reads and writes of multiple variables.

Noncompliant Code Example (Addition of Atomic Integers)

In this noncompliant code example, a and b are replaced with atomic integers:

final class Adder {
  private final AtomicInteger a = new AtomicInteger();
  private final AtomicInteger b = new AtomicInteger();

  public int getSum() {
    return a.get() + b.get();
  }

  public void setValues(int a, int b) {
    this.a.set(a);
    this.b.set(b);
  }
}

The simple replacement of the two int fields with atomic integers fails to eliminate the race condition because the compound operation a.get() + b.get() is still non-atomic.

Compliant Solution (Addition)

This compliant solution synchronizes the setValues() and getSum() methods to ensure atomicity:

final class Adder {
  private int a;
  private int b;

  public synchronized int getSum() {
    // Check for overflow 
    return a + b;
  }

  public synchronized void setValues(int a, int b) {
    this.a = a;
    this.b = b;
  }
}

The operations within the synchronized methods are now atomic with respect to other synchronized methods that lock on that object's monitor (that is, its intrinsic lock). It is now possible, for example, to add overflow checking to the synchronized getSum() method without introducing the possibility of a race condition.

Risk Assessment

When operations on shared variables are not atomic, unexpected results can be produced. For example, information can be disclosed inadvertently because one user can receive information about other users.

Automated Detection

Some available static analysis tools can detect the instances of non-atomic update of a concurrently shared value. The result of the update is determined by the interleaving of thread execution. These tools can detect the instances where thread-shared data is accessed without holding an appropriate lock, possibly causing a race condition.

Related Guidelines

Bibliography

...

Image Added Image Added Image Added11. Concurrency (CON)      11. Concurrency (CON)      CON02-J. Always synchronize on the appropriate object