JVM Specification 第二天(Frame)

A  frameis used to store data and partial results, as well as to perform dynamic linking, return values for methods, and dispatch exceptions. 

A new frame is created each time a method is invoked. A frame is destroyed when its method invocation completes, whether that completion is normal or abrupt (it throws an uncaught exception). 

Frames are allocated from the Java virtual machine stack  of the thread creating the frame. Each frame has its own array of local variables , its own operand stack (§2.6.2), and a reference to the runtime constant pool  of the class of the current method

1. Local variables

Each frame  contains an array of variables known as its local variables.

A single local variable can hold a value of type boolean, byte, char, short, int, float,  reference, or  returnAddress. A pair of local variables can hold a value of type longor double.

2.Operand Stacks

Each frame (§2.6) contains a last-in-first-out (LIFO) stack known as its  operand stack.  The  maximum  depth  of  the  operand  stack  of  a  frame  is  determined  at compile-time and is supplied along with the code for the method associated with the frame

The  operand  stack  is  empty  when  the  frame  that  contains  it  is  created.  The Java virtual machine supplies instructions to load constants or values from local variables or fields onto the operand stack. Other Java virtual machine instructions take operands from the operand stack, operate on them, and push the result back onto the operand stack. The operand stack is also used to prepare parameters to be passed to methods and to receive method results.

For example, the iaddinstruction (§iadd) adds two intvalues together. It requires that the intvalues to be added be the top two values of the operand stack, pushed there by previous instructions. Both of the  intvalues are popped from the operand stack.  They  are  added,  and  their  sum  is  pushed  back  onto  the  operand  stack. Subcomputations may be nested on the operand stack, resulting in values that can be used by the encompassing computation.

3.Dynamic Linking

Each frame (§2.6) contains a reference to the runtime constant pool (§2.5.5) for the type of the current method to support  dynamic linkingof the method code.
The  classfile code for a method refers to methods to be invoked and variables to be accessed via symbolic references. Dynamic linking translates these symbolic method references into concrete method references, loading classes as necessary to resolve as-yet-undefined symbols, and translates variable accesses into appropriate offsets in storage structures associated with the runtime location of these variables.This late binding of the methods and variables makes changes in other classes that a method uses less likely to break this code.

Reference:http://blog.csdn.net/vernonzheng/article/details/8458483