Page History

To avoid data corruption in multithreaded Java programs, shared data must be protected from concurrent modifications and accesses. This Locking can be performed at the object level by using synchronized methods or , synchronized blocks, or by using the java.util.concurrent dynamic lock objects. However, excessive use of locking may can result in deadlocks (See CON08-J. Do not call alien methods that synchronize on the same objects as any callers in the execution chain). For instance, to avoid deadlocks deadlocks.

Java neither prevents deadlocks nor requires their detection [JLS 2015]. Deadlock can occur when two or more threads request and release locks in different orders. Consequently, programs are required to avoid deadlock by acquiring and releasing locks in the same order.

Additionally, synchronization should be limited to cases where it is absolutely necessary. For example, the paint(), dispose(), stop(), and destroy() methods should never be synchronized in an applet should not be synchronized because they are always called and used from dedicated threads.

"The Java programming language neither prevents nor requires detection of deadlock conditions." \[[JLS 05|AA. Java References#JLS 05]\]. Deadlocks can arise when two or more threads request and release locks in different orders. 

Noncompliant Code Example

. Furthermore, the Thread.stop() and Thread.destroy() methods are deprecated (see THI05-J. Do not use Thread.stop() to terminate threads for more information).

This rule also applies to programs that need to work with a limited set of resources. For example, liveness issues can arise when two or more threads are waiting for each other to release resources such as database connections. These issues can be resolved by letting each waiting thread retry the operation at random intervals until they successfully acquire the resource.

Noncompliant Code Example (Different Lock Orders)

This noncompliant code example can deadlock because of excessive synchronization. The balanceAmount field represents the total balance available for a particular BankAccount object. Users are allowed to initiate an operation that atomically transfers a specified amount from one account to anotherThis noncompliant code example can deadlock because of excessive synchronization. Assume that an attacker has two bank accounts and is capable of requesting two depositAllAmount() operations in succession, one each from the two threads started in main().

final class BankAccount {
  private intdouble balanceAmount;  // Total amount in bank account
	 
  private BankAccount(intdouble balance) {
    this.balanceAmount = balance;
  }

  // Deposits the amount from this object instance
  // to BankAccount instance argument ba 
  private void depositAllAmountdepositAmount(BankAccount ba, double amount) {
    synchronized (this) {
      synchronized (ba) {
        if ba.balanceAmount += this.balanceAmount;(amount > balanceAmount) {
        this.balanceAmount = 0;throw // withdraw all amount from this instance
new IllegalArgumentException(
             ba.displayAllAmount();  //"Transfer Displaycannot thebe newcompleted"
 balanceAmount in ba (may cause deadlock)
    );
        }
     } 
  }
     ba.balanceAmount += amount;
        this.balanceAmount -= amount;
  private synchronized void displayAllAmount() {}
    System.out.println(balanceAmount);}
  }

  public static void initiateTransfer(final BankAccount first,
    final BankAccount second, final double amount) {

    Thread ttransfer = new Thread(new Runnable() {
        public void run() {
          first.depositAllAmountdepositAmount(second, amount);
        }
    });
    ttransfer.start();
  }
}

Objects of class BankAccount represent bank accounts. The balanceAmount field represents the total balance amount available for a particular object (bank account). A user is allowed to initiate an operation deposit all amount that transfers the balance amount from one account to another. This is equivalent to closing a bank account and transferring the balance to a different (existing or new) account.

Objects of this class are deadlock-prone. An attacker may cause the program to construct two threads that initiate balance transfers from two different BankAccount object instances, a and b. Consider the following code that does this:

BankAccount a = new BankAccount(5000);
BankAccount b = new BankAccount(6000);
initiateTransfer(a, b); // starts thread 1
initiateTransfer(b, a); // starts thread 2

this class are prone to deadlock. An attacker who has two bank accounts can construct two threads that initiate balance transfers from two different BankAccount object instances a and b. For example, consider the following code:

BankAccount a = new BankAccount(5000);
BankAccount b = new BankAccount(6000);
BankAccount.initiateTransfer(a, b, 1000); // starts thread 1
BankAccount.initiateTransfer(b, a, 1000); // starts thread 2

Each transfer is performed in its own threadThe two transfers are performed in their own threads, from instance a to b and b to a. The first thread atomically transfers the amount from a to b by depositing the balance from a to it in account b and then withdrawing the entire balance same amount from a. The second thread performs the reverse operation, ; that is, it transfers the balance amount from b to a and withdraws the balance from b. When executing depositAllAmountdepositAmount(), the first thread might acquire acquires a lock on object a while the . The second thread may could acquire a lock on object b before the first thread can. Subsequently, the first thread requests would request a lock on b, which is already held by the second thread and the . The second thread requests would request a lock on a, which is already held by the first thread. This constitutes a deadlock condition , as because neither thread can proceed.The threads in this program request monitors in varying order

This noncompliant code example may or may not deadlock, depending on the interleaving of method calls. If Thread T1 finishes executing before Thread T2, or T2 before T1, there are no issues because in these cases, locks are acquired and released in the same order. Sequences where the threads alternate, such as, T1, T2, T1, T2 may deadlock.

Compliant Solution (single private lock)

scheduling details of the platform. Deadlock occurs when (1) two threads request the same two locks in different orders, and (2) each thread obtains a lock that prevents the other thread from completing its transfer. Deadlock is avoided when two threads request the same two locks but one thread completes its transfer before the other thread begins. Similarly, deadlock is avoided if the two threads request the same two locks in the same order (which would happen if they both transfer money from one account to a second account) or if two transfers involving distinct accounts occur concurrently.

Compliant Solution (Private Static Final Lock Object)

This compliant solution avoids deadlock by synchronizing on a private static final lock object before performing any account transfers:The deadlock can be avoided by using a single lock to acquire before doing any account transfers.

final class BankAccount {
  private static intdouble balanceAmount;  // Total amount in bank account
	 
 private privatestatic final Object lock = new Object();

  private BankAccount(intdouble balance) {
    this.balanceAmount = balance;
    this.lock = new Object();
  }

  // Deposits the amount from this object instance
  // to BankAccount instance argument ba 
  private void depositAllAmountdepositAmount(BankAccount ba, double amount) {
    synchronized (lock) {
      ba.balanceAmount += this.balanceAmount;if (amount > balanceAmount) {
      this.balanceAmount = 0;throw // withdraw all amount from this instance
new IllegalArgumentException(
          ba.displayAllAmount();  //"Transfer Displaycannot the new balanceAmount in ba (may cause deadlock)be completed");
      }
    } 
  }
  ba.balanceAmount += amount;
  private void displayAllAmount() {
 this.balanceAmount   synchronized (lock) {
      System.out.println(balanceAmount)-= amount;
    }
  }

  public static void initiateTransfer(final BankAccount first,
    final BankAccount second, final double amount) {

    Thread ttransfer = new Thread(new Runnable() {
        @Override public void run() {
          first.depositAllAmountdepositAmount(second, amount);
        }
    });
    ttransfer.start();
  }
}

In this scenario, if deadlock cannot occur when two threads with two different BankAccount objects try to tranfer transfer to each othersother's accounts simultaneously, deadlock cannot occur. One thread will acquire the private lock, complete its transfer, and release the lock , before the other thread may can proceed.

This solution comes with imposes a performance penalty , as because a private static lock restricts the system to only performing one transfer at a timeperforming transfers sequentially. Two transfers involving four distinct accounts (and with distinct target accounts) may not happen cannot be performed concurrently. The impact of this This penalty increases considerably as the number of BankAccount objects increase. Consequently, this solution does not fails to scale very well.

Compliant Solution (

...

Ordered Locks)

This compliant solution ensures that multiple locks are acquired and released in the same order. It requires that an a consistent ordering over BankAccount objects is available. The ordering is enforced by having the class BankAccount extend . Consequently, the BankAccount class implements the java.lang.Comparable interface and overriding overrides the compareTo() method.

final class BankAccount implements ComparableComparable<BankAccount> {
  private intdouble balanceAmount;  // Total amount in bank account
	 
  private final Object lock;

  private final long id; // Unique for each BankAccount
  private static final AtomicLong nextID = new AtomicLong(0); // Next unused ID

  BankAccount(intdouble balance) {
    this.balanceAmount = balance;
    this.lock = new Object();
  }

  // Deposits the amount from this object instance to BankAccount instance argument ba 
  private void depositAllAmountthis.id = nextID.getAndIncrement();
  }

  @Override public int compareTo(BankAccount ba) {
     BankAccount former, latter;
    if (compareTo(ba) < 0) {
      former = this;
      latter = ba;
    } else {
      former = ba;
      latter = thisreturn (this.id > ba.id) ? 1 : (this.id < ba.id) ? -1 : 0;
  }

  // Deposits the amount from this object instance
  // to BankAccount instance argument ba
  public void depositAmount(BankAccount ba, double amount) {
    BankAccount former, latter;
    }
    synchronized (formerif (compareTo(ba) < 0) {
      synchronizedformer (latter) {= this;
      latter = ba.balanceAmount += this.balanceAmount;
  ;
    } else {
      this.balanceAmountformer = 0ba;
  //  withdraw all amountlatter from= this instance;
    }
    synchronized ba.displayAllAmount(former); //{
 Display the new balanceAmount in basynchronized (maylatter) cause{
 deadlock)
      } 
if (amount > balanceAmount) }{
  }
 
  private synchronized void displayAllAmount() {
  throw new System.out.println(balanceAmount);IllegalArgumentException(
  }

  public static void initiateTransfer(final BankAccount first, final BankAccount second) {
 "Transfer cannot be Thread t = new Thread(new Runnable() {
completed");
        }
       public void run() {ba.balanceAmount += amount;
        first.depositAllAmount(second)this.balanceAmount -= amount;
      }
    });
    t.start();
  }

  public intstatic void compareToinitiateTransfer(final BankAccount ba) {
first,
    if(this.balanceAmount < ba.balanceAmountfinal BankAccount second, final double amount) {

    Thread return -1;
   } else if(this.balanceAmount > ba.balanceAmounttransfer = new Thread(new Runnable() {
     return  1;
 @Override public }void elserun() {
     return 0;
   }
   first.depositAmount(second, amount);
        }
    });
    transfer.start();
  }
}

In this compliant solution, whenever Whenever a transfer occurs, the two BankAccount objects are ordered , with so that the first object's lock being is acquired before the second object's lock. Consequently if When two threads attempt transfers between the same two accounts, they will both each try to acquire the first account's lock before acquiring the second account's lock first, with the result that . Consequently, one thread will acquire acquires both locks, complete completes the transfer, and release releases both locks before the other may thread can proceed.

Unlike the previous compliant solution, this solution incurs no performance penalty, as multiple transfers can occur concurrently permits multiple concurrent transfers as long as the transfers involve distinct target accounts.

Compliant Solution (ReentrantLock)

In this compliant solution, each BankAccount has a java.util.concurrent.locks.ReentrantLock associated with it. This design permits the depositAllAmountdepositAmount() method to try acquiring to attempt to acquire the locks of both accounts' locks, but releasing its , to release the locks if it fails, and trying to try again later if necessary.

final class BankAccount {
  private intdouble balanceAmount;  // Total amount in bank account
  private final Lock lock = new ReentrantLock();
  Lockprivate getLock() {return lock;}
  private static final int TIME = 1000; // ms
	 
  private BankAccount(intfinal Random number = new Random(123L);

  BankAccount(double balance) {
    this.balanceAmount = balance;
  }

  // Deposits the amount from this object instance
  // to BankAccount instance argument ba 
  private void depositAllAmountdepositAmount(BankAccount ba, double amount)
                             throws InterruptedException {
    while (true) {
      if (this.lock.tryLock()) {
        try {
          if (ba.lock.tryLock()) {
            try {
              ba.balanceAmount += this.balanceAmount;
   if (amount > balanceAmount) {
           this.balanceAmount = 0; // withdraw all amountthrow from this instancenew IllegalArgumentException(
              ba.displayAllAmount();  // Display the new balanceAmount"Transfer incannot ba (may cause deadlock)be completed");
              break;}
            } finally { ba.balanceAmount += amount;
              ba.getLock().unlock()this.balanceAmount -= amount;
            }
  break;
        }
    } finally {
  } finally {
          thisba.getLock()lock.unlock();
        }
      }
      sleep( TIME);
    }
  }
   
  private void} displayAllAmount()finally {
    while (true) {
      if (this.lock.tryLockunlock()) {
;
        }
     try {}
      int n =  System.out.println(balanceAmountnumber.nextInt(1000);
      int TIME = 1000 + breakn;
 // 1 second + random delay to }prevent finally {livelock
          lock.unlock(Thread.sleep(TIME);
      }
  }

  public static   }
      sleep( TIME);
    }
  }

  public static void initiateTransfer(void initiateTransfer(final BankAccount first,
    final BankAccount firstsecond, final BankAccountdouble secondamount) {

    Thread ttransfer = new Thread(new Runnable() {
        public void run() {
        first.depositAllAmount(second);
    try {
    }
    });
    tfirst.start();
  }
}
depositAmount(second, amount);
          } catch (InterruptedException e) {
            Thread.currentThread().interrupt(); // Reset interrupted status
          }
        }
    });
    transfer.start();
  }
}

Deadlock is impossible in this compliant solution because locks are never held indefinitely. If the current object's lock is acquired but the second lock is unavailable, the first lock is released and the thread sleeps for some specified amount of time before attempting to reacquire the lockDeadlock is impossible in this code, becase no routine grabs a lock and holds it indefinitely. If a lock is acquired, but the system cannot proceed immediately, it releases the lock and sleeps before trying the lock again.

Code that uses this lock behaves locking strategy has behavior similar to that of synchronized code that uses the traditional monitor lock. ReentrantLock also provides several other capabilities, for instance. For example, the tryLock() method does not block waiting if immediately returns false when another thread is already holding holds the lock. The class Further, the java.util.concurrent.locks.ReentrantReadWriteLock can be used when some thread requires class has multiple-readers/single-writer semantics and is useful when some threads require a lock to write information while other threads require the lock to concurrently read the information.

Noncompliant Code Example

...

(Different Lock Orders, Recursive)

The following immutable WebRequest class encapsulates a web request received by a server:

// Immutable WebRequest
public final class WebRequest {
  private final long bandwidth;
  private final long responseTime;

  public WebRequest(long bandwidth, long responseTime) {
    this.bandwidth = bandwidth;
    this.responseTime = responseTime;
  }

  public long getBandwidth

This noncompliant code example consists of an application to monitor a sports race. It maintains a list of racers. Each racer has two statistics that can be reported about them: their current speed, and their current distance travelled. This example provides methods to report on the average current speed of the racers, and the average distance of the racers. To be thread-safe, it locks each racer before it reads her statistics until after it reports the average. Consequently, the statics reported by the methods are accurate at the time the methods actually return their results.

public class Racer {
  public double getCurrentSpeed() {
    // ...
  }

  public double getDistance() {
    // ...return bandwidth;
  }

  //...
}

public classlong RacegetResponseTime() {
  final static int MAX = 20return responseTime;
  static}
}

Each request has a response time associated with it, along with a measurement of the network bandwidth required to fulfill the request.

This noncompliant code example monitors web requests and provides routines for calculating the average bandwidth and response time required to serve incoming requests.

public final class WebRequestAnalyzer Racer[] racers = new Racer[MAX];
 
  public Race() {
  private final forVector<WebRequest> (intrequests i= =new 0Vector<WebRequest>();

 i public < MAX; i++boolean addWebRequest(WebRequest request) {
    return  racers[i] = new Racer(/* initialize racer */);
requests.add(new WebRequest(request.getBandwidth(),
          }
  }

  double getAverageCurrentSpeed() {
    return averageCurrentSpeedCalculator(0, 0.0   request.getResponseTime()));
  }

  public double averageCurrentSpeedCalculator(int i, double currentSpeedgetAverageBandwidth() { // Acquires locks in increasing order
    if (requests.size(i) > MAX - 1== 0) {
      return currentSpeed / MAXthrow new IllegalStateException("The vector is empty!");
    }

    return synchronized(racers[i]) {
   calculateAverageBandwidth(0, 0);
  }

  public double getAverageResponseTime() {
   currentSpeed += racers[i].getCurrentSpeed();
if (requests.size() == 0) {
      throw returnnew averageCurrentSpeedCalculator(++i, currentSpeedIllegalStateException("The vector is empty!");
    }	 
  }


  double getAverageDistance() {
    return averageDistanceCalculator(MAXreturn calculateAverageResponseTime(requests.size() - 1, 0.0);
  }

  private double averageDistanceCalculatorcalculateAverageBandwidth(int i, doublelong distancebandwidth) { // Acquires locks in decreasing order
    if (i <== -1requests.size()) {		 
      return distancebandwidth / MAXrequests.size();
    } 

    synchronized(racers[i] (requests.elementAt(i)) {
      distancebandwidth += racers[i].getDistance();requests.get(i).getBandwidth();
      // Acquires locks in increasing order
      return averageDistanceCalculatorcalculateAverageBandwidth(--++i, distancebandwidth);
    }
  }
}

This implementation is prone to deadlock because the recursive calls occur within the synchronized regions of these methods and acquire locks in opposite numerical orders. That is, averageCurrentSpeedCalculator() requests locks from index 0 to MAX - 1 (19) whereas averageDistanceCalculator() requests them from index MAX - 1 (19) to 0. Because of recursion, no previously acquired locks are released by either method. A deadlock occurs when two threads call these methods out of order in that, one thread calls averageSpeedCalculator() while the other calls averageTimeCalculator() before either method has finished executing.

For example, if one thread calls getCurrentSpeed(), it will first acquire the intrinsic lock for Racer 0, the first in the array. Meanwhile a second thread calls getCurrentDistance(), and it will first acquire the intrinsic lock for Racer 19, the last in the array. Consequently, deadlock results, as neither thread can acquire all of the locks and proceed with the calculation.

Compliant Solution

In this compliant solution, the two calculation methods use the same ordering for locking the racer objects.

  private double calculateAverageResponseTime(int i, long responseTime) {
    if (i <= -1) {
      return responseTime / requests.size();
    }
    synchronized (requests.elementAt(i)) {
      responseTime += requests.get(i).getResponseTime();
      // Acquires locks in decreasing order
      return calculateAverageResponseTime(--i, responseTime);
    }
  }
}

The monitoring application is built around the WebRequestAnalyzer class, which maintains a list of web requests using the requests vector and includes the addWebRequest() setter method. Any thread can request the average bandwidth or average response time of all web requests by invoking the getAverageBandwidth() and getAverageResponseTime() methods.

These methods use fine-grained locking by holding locks on individual elements (web requests) of the vector. These locks permit new requests to be added while the computations are still underway. Consequently, the statistics reported by the methods are accurate when they return the results.

Unfortunately, this noncompliant code example is prone to deadlock because the recursive calls within the synchronized regions of these methods acquire the intrinsic locks in opposite numerical orders. That is, calculateAverageBandwidth() requests locks from index 0 up to requests.size() − 1, whereas calculateAverageResponseTime() requests them from index requests.size() − 1 down to 0. Because of recursion, previously acquired locks are never released by either method. Deadlock occurs when two threads call these methods out of order, because one thread calls calculateAverageBandwidth(), while the other calls calculateAverageResponseTime() before either method has finished executing.

For example, when there are 20 requests in the vector, and one thread calls getAverageBandwidth(), the thread acquires the intrinsic lock of WebRequest 0, the first element in the vector. Meanwhile, if a second thread calls getAverageResponseTime(), it acquires the intrinsic lock of WebRequest 19, the last element in the vector. Consequently, deadlock results because neither thread can acquire all of the locks required to proceed with its calculations.

Note that the addWebRequest() method also has a race condition with calculateAverageResponseTime(). While iterating over the vector, new elements can be added to the vector, invalidating the results of the previous computation. This race condition can be prevented by locking on the last element of the vector (when it contains at least one element) before inserting the element.

Compliant Solution

In this compliant solution, the two calculation methods acquire and release locks in the same order, beginning with the first web request in the vector.

public final class WebRequestAnalyzer {
  private final Vector<WebRequest> requests = new Vector<WebRequest>();

  public boolean addWebRequest(WebRequest request) {
    return requests.add(new WebRequest(request.getBandwidth(),
                        request.getResponseTime()));
  }

  public double getAverageBandwidth() {
    if (requests.size() == 0) {
      throw new IllegalStateException("The vector is empty!");
    }
    return calculateAverageBandwidth(0, 0);
  }

  public double getAverageResponseTime() {
    if (requests.size() == 0) {
      throw new IllegalStateException("The vector is empty!");
    }
    return calculateAverageResponseTime(0, 

public class Race {
  double getAverageCurrentSpeed() {
    return averageCurrentSpeedCalculator(0, 0.0);
  }

  private double averageCurrentSpeedCalculatorcalculateAverageBandwidth(int i, doublelong currentSpeedbandwidth) { // Acquires locks in increasing order
    if (i > MAX - 1== requests.size()) {
      return currentSpeedbandwidth / MAXrequests.size();
    }

    synchronized(racers[i]) { (requests.elementAt(i)) {
      // Acquires locks in increasing order
      currentSpeedbandwidth += racers[i].getCurrentSpeedrequests.get(i).getBandwidth();
      return averageCurrentSpeedCalculatorcalculateAverageBandwidth(++i, currentSpeedbandwidth);
    }	 
  }


  double getAverageDistance() {
    return averageDistanceCalculator(0, 0.0);
  }

  double averageDistanceCalculator(int i, double distance) { // Acquires locks in increasing order  private double calculateAverageResponseTime(int i, long responseTime) {
    if (i > MAX - 1== requests.size()) {		 
      return distanceresponseTime / MAXrequests.size();
    } 

    synchronized(racers[i] (requests.elementAt(i)) {
      // Acquires locks in increasing order
      distanceresponseTime += racers[i].getDistance requests.get(i).getResponseTime();
      return averageDistanceCalculatorcalculateAverageResponseTime(++i, distanceresponseTime);
    }
  }
}

Consequently, while one thread is producing a calculation, no other thread can produce a calculation. This is because the other thread would first have to synchronize on racers0, which is impossible until the first calculation is completecalculating the average bandwidth or response time, another thread cannot interfere or induce deadlock. Each thread must first synchronize on the first web request, which cannot happen until any prior calculation completes.

Locking on the last element of the vector in addWebRequest() is unnecessary for two reasons. First, the locks are acquired in increasing order in all the methods. Second, updates to the vector are reflected in the results of the computations.

Risk Assessment

Acquiring and releasing locks in the wrong order may can result in deadlocks deadlock.

Automated Detection

TODO

Related Vulnerabilities

Search for vulnerabilities resulting from the violation Some static analysis tools can detect violations of this rule on the CERT website.

References

\[[JLS 05|AA. Java References#JLS 05]\] [Chapter 17, Threads and Locks|http://java.sun.com/docs/books/jls/third_edition/html/memory.html]
\[[Halloway 00|AA. Java References#Halloway 00]\] 
\[[MITRE 09|AA. Java References#MITRE 09]\] [CWE ID 412|http://cwe.mitre.org/data/definitions/412.html] "Unrestricted Lock on Critical Resource"

Related Guidelines

Bibliography

...

Image Added Image Added Image AddedCON11-J. Do not assume that declaring an object volatile guarantees visibility of its members      11. Concurrency (CON)      CON13-J. Do not try to force thread shutdown