sgodbillon/bytecodeparser

A Java library to parse JVM bytecode, simulate the stack and extract as much information as possible

bytecodeparser's Issues

const string paramter return null

class.forname("test") with example code:

            DecodedMethodInvocationOp dmio = (DecodedMethodInvocationOp) frame.decodedOp;
            MethodParams methodParams = DecodedMethodInvocationOp.resolveParameters(frame);
            MethodParam[] params = methodParams.merge();
            System.out.println("method '" + dmio.getName() + "' has been called with the following arguments: " + java.util.Arrays.toString(params));

params return null;

No support for Java8 / invokedynamic

When parsing Java8 code with lambdas (that use invokedynamic opcode), NPE is thrown because invokedynamic is not listed in Opcodes.java

Problem about DecodedLocalVariableOp.localVariable

Hi,

I'm a new user of this package. I'm now trying to retrieve the information of local variables as the sample you show on the website http://stephane.godbillon.com/BytecodeParser/.

The problem I encountered when doing this is that I cannot get the local variable for LOAD instructions. Specifically, I use the following code to retrieve the information I want:

@Override
public void handle(Op op, int idx) {
  if ((op instanceof LocalVariableOpcode) == false)
    return ;

  LocalVariableOpcode lv = op.as(LocalVariableOpcode.class);
  LocalVariable lvInfo = lv.decode(context, idx).localVariable;
  System.out.println(idx + " " + lvInfo.toString() + ": " + lvInfo.name);
}

However, I got NullPointerException of lvInfo when parsing the instruction ILOAD 5 in my program. Is it due to some wrong manipulations or restrictions?

Thanks.

SWAP incorrectly performs swap on Stack

The current SWAP takes the top element and puts it one from bottom instead of one from top(which it is supposed to do).

The pull request is
#3

I have absolutely no clue on how to write a test case since I used this branch to reproduce the issue

https://github.com/deanhiller/timecardz/tree/bytecodeissue

In my case, I had a stack with [XXXXX, YYYYYY, TOP, ZZZZZ] and doing a swap ended up with [YYYYYY, TOP, XXXXX, ZZZZZ] instead of the correct [YYYYYY, XXXXX, TOP, ZZZZ] which then resulted in popped a TOP runtimeexception later on.

exception on init method not called but enter called causing nullpointer

I am not sure how to reproduce this and I worked around it by just doing if something != null but basically the bytecode stuff was wrapping enterMethod and exitMethod to be called around one of my methods(it seemed to always be the first method in my catch block) and initMethod was nevercalled.

Since initMethod was never called, I ended up with a nullpointer in enterMethod as it requires initMethod to be called.

The logic on whether to wire initMethod is strange in my opinion. The way it should be coded I would think is if(enterMethodAdded || exitMethodAdded) addInitMethod()

but instead it depends on different stuff.

popped a TOP! issue

I can only produce this with playframework so I am not sure if it is playframework issue or a bytecodeparser issue. I recorded full details here

http://play.lighthouseapp.com/projects/57987-play-framework/tickets/1761-bytecodeparser-or-playframework-issue