How the Java virtual machine handles exceptions

A detailed study with examples of classes and methods

By Bill Venners, JavaWorld.com, 01/01/97

Exceptions

class OverflowException extends Exception {
}
class UnderflowException extends Exception {
}
class DivideByZeroException extends Exception {
}

A simple method that catches and throws exceptions is the remainder method of class NitPickyMath:

static int remainder(int dividend, int divisor)
    throws DivideByZeroException {
    try {
        return dividend % divisor;
    }
    catch (ArithmeticException e) {
        throw new DivideByZeroException();
    }
}

The remainder method simply performs the remainder operation upon the two ints passed as arguments. The remainder operation throws an ArithmeticException if the divisor of the remainder operation is a zero. This method catches this ArithmeticException and throws a DivideByZeroException.

The difference between a DivideByZero and an ArithmeticException exception is that the DivideByZeroException is a checked exception and the ArithmeticException is unchecked. Because the ArithmeticException is unchecked, a method need not declare this exception in a throws clause even though it might throw it. Any exceptions that are subclasses of either Error or RuntimeException are unchecked. (ArithmeticException is a subclass of RuntimeException.) By catching ArithmeticException and then throwing DivideByZeroException, the remainder method forces its clients to deal with the possibility of a divide-by-zero exception, either by catching it or declaring DivideByZeroException in their own throws clauses. This is because checked exceptions, such as DivideByZeroException, thrown within a method must be either caught by the method or declared in the method's throws clause. Unchecked exceptions, such as ArithmeticException, need not be caught or declared in the throws clause.

javac generates the following bytecode sequence for the remainder method:

The main bytecode sequence for remainder:
   0 iload_0               // Push local variable 0 (arg passed as divisor)
   1 iload_1               // Push local variable 1 (arg passed as dividend)
   2 irem                  // Pop divisor, pop dividend, push remainder
   3 ireturn               // Return int on top of stack (the remainder)
The bytecode sequence for the catch (ArithmeticException) clause:
   4 pop                   // Pop the reference to the ArithmeticException
                           // because it isn't used by this catch clause. 
   5 new #5 <Class DivideByZeroException>
                           // Create and push reference to new object of class
                           // DivideByZeroException.
DivideByZeroException
   8 dup           // Duplicate the reference to the new
                           // object on the top of the stack because it 
                           // must be both initialized 
                           // and thrown. The initialization will consume
                           // the copy of the reference created by the dup.
   9 invokenonvirtual #9 <Method DivideByZeroException.<init>()V>
                           // Call the constructor for the DivideByZeroException
                           // to initialize it. This instruction
                           // will pop the top reference to the object.
  12 athrow                // Pop the reference to a Throwable object, in this
                           // case the DivideByZeroException, 
                           // and throw the exception.

The bytecode sequence of the remainder method has two separate parts. The first part is the normal path of execution for the method. This part goes from pc offset zero through three. The second part is the catch clause, which goes from pc offset four through twelve.

The irem instruction in the main bytecode sequence may throw an ArithmeticException. If this occurs, the Java virtual machine knows to jump to the bytecode sequence that implements the catch clause by looking up and finding the exception in a table. Each method that catches exceptions is associated with an exception table that is delivered in the class file along with the bytecode sequence of the method. The exception table has one entry for each exception that is caught by each try block. Each entry has four pieces of information: the start and end points, the pc offset within the bytecode sequence to jump to, and a constant pool index of the exception class that is being caught. The exception table for the remainder method of class NitPickyMath is shown below:

Exception table:
   from   to  target type
     0     4     4   <Class java.lang.ArithmeticException>

The above exception table indicates that from pc offset zero through three, inclusive, ArithmeticException is caught. The try block's endpoint value, listed in the table under the label "to", is always one more than the last pc offset for which the exception is caught. In this case the endpoint value is listed as four, but the last pc offset for which the exception is caught is three. This range, zero to three inclusive, corresponds to the bytecode sequence that implements the code inside the try block of remainder. The target listed in the table is the pc offset to jump to if an ArithmeticException is thrown between the pc offsets zero and three, inclusive.

Free Download - 5 Minute Product Review. When slow equals Off: Manage the complexity of Web applications - Symphoniq

Free Download - 5 Minute Product Review. Realize the benefits of real user monitoring in less than an hour. - Symphoniq