gov.nasa.jpf.jvm.bytecode

Class ARRAYLENGTH

java.lang.Object

gov.nasa.jpf.vm.Instruction

gov.nasa.jpf.jvm.bytecode.ARRAYLENGTH

All Implemented Interfaces:

JVMInstruction, Attributable, InstructionInterface, java.lang.Cloneable

public class ARRAYLENGTH
extends Instruction
implements JVMInstruction

Get length of array ..., arrayref => ..., length

Field Summary

Fields

Modifier and Type	Field and Description
protected int	arrayRef

Fields inherited from class gov.nasa.jpf.vm.Instruction

attr, insnIndex, mi, position

Constructor Summary

Constructors

Constructor and Description
ARRAYLENGTH()

Method Summary

Modifier and Type	Method and Description
void	accept(JVMInstructionVisitor insVisitor)
Instruction	execute(ThreadInfo ti) this is the real workhorse returns next instruction to enter in this thread <2do> it's unfortunate we roll every side effect into this method, because it diminishes the value of the 'executeInstruction' notification: all insns that require some sort of late binding (InvokeVirtual, GetField, ..) are not yet fully analyzable (e.g.
int	getArrayRef(ThreadInfo ti) only makes sense from an executeInstruction() or instructionExecuted() listener, it is undefined outside of insn exec notifications
int	getByteCode()
ElementInfo	peekArrayElementInfo(ThreadInfo ti)
protected int	peekArrayRef(ThreadInfo ti)

Methods inherited from class gov.nasa.jpf.vm.Instruction

addAttr, asInstruction, attrIterator, attrIterator, cleanupTransients, getAttr, getAttr, getFileLocation, getFilePos, getInstructionIndex, getLength, getLineNumber, getMethodInfo, getMnemonic, getNext, getNext, getNextAttr, getPosition, getPrev, getSourceLine, getSourceLocation, getSourceOrLocation, hasAttr, hasAttr, init, isBackJump, isCompleted, isExtendedInstruction, isFirstInstruction, isMonitorEnterPrologue, isSchedulingRelevant, removeAttr, replaceAttr, requiresClinitExecution, setAttr, setContext, setLocation, setMethodInfo, toPostExecString, toString, typeSafeClone

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Methods inherited from interface gov.nasa.jpf.vm.bytecode.InstructionInterface

asInstruction, getFileLocation, getFilePos, getInstructionIndex, getLength, getLineNumber, getMethodInfo, getMnemonic, getNext, getNext, getPosition, getPrev, getSourceLine, getSourceLocation, isBackJump, isCompleted, isExtendedInstruction, isFirstInstruction, toPostExecString

Methods inherited from interface gov.nasa.jpf.util.Attributable

addAttr, attrIterator, attrIterator, getAttr, getAttr, getNextAttr, hasAttr, hasAttr, removeAttr, replaceAttr, setAttr

Field Detail

arrayRef

protected int arrayRef

Constructor Detail

ARRAYLENGTH

public ARRAYLENGTH()

Method Detail

getArrayRef

public int getArrayRef(ThreadInfo ti)

only makes sense from an executeInstruction() or instructionExecuted() listener, it is undefined outside of insn exec notifications

peekArrayElementInfo

public ElementInfo peekArrayElementInfo(ThreadInfo ti)

execute

public Instruction execute(ThreadInfo ti)

Description copied from class: Instruction

this is the real workhorse returns next instruction to enter in this thread <2do> it's unfortunate we roll every side effect into this method, because it diminishes the value of the 'executeInstruction' notification: all insns that require some sort of late binding (InvokeVirtual, GetField, ..) are not yet fully analyzable (e.g. the callee of InvokeVirtuals is not known yet), putting the burden of duplicating the related code of enter() in the listener. It would be better if we factor this 'prepareExecution' out of enter()

Specified by:

execute in interface InstructionInterface

Specified by:

execute in class Instruction

getByteCode

public int getByteCode()

Specified by:

getByteCode in interface InstructionInterface

accept

public void accept(JVMInstructionVisitor insVisitor)

Specified by:

accept in interface JVMInstruction

peekArrayRef

protected int peekArrayRef(ThreadInfo ti)