Re: Locating where first param of an invoke gets put on the stack

Hi Jon,

You will need to implement a  simple symbolic interpreter to know this data
flow information. I.e, that the definition of  `i` reaches the invoke of
the method.

I have implemented a simple tool that may help you with that. The tool is
available here https://github.com/saeg/asm-defuse. I have no documentation
(shame on me). I implemented it for data flow testing coverage, but I think
that it may be useful for you needs.

Best regards

Roberto

On Fri, Dec 1, 2017, 2:02 PM Jon Sawyer <[email protected]> wrote:

> I want to thank everyone who worked on asm, such a wonderful framework. I
> wish I had found it earlier, it would have saved me so much time. I have a
> bad habit of reinventing the wheel when I need something, trying to avoid
> that here if possible. Really hoping ASM contains some feature that could
> make this easier.
>
> I'd like to inject some code prior to the first param of an invoke* is put
> on the stack. My problem here is locating when the first param (or "sub
> param" enters the stack);
>
> If it is something like
>
> Log.print("hello","world");
> ---
>  ldc           #29                 // String TAG
>  ldc           #31                 // String MSG
> 15: invokestatic  #37                 // Method
> logg/loggl/Log.print:(Ljava/lang/String;Ljava/lang/String;)I
>
> it is fairly simple, I can just move back two positions and be in front of
> the ldc "TAG", and then inject my code.
>
>
> But when something like this happens it gets complicated , dosomething
> takes 7 ints. I need to jump to right before the bipush 100, so I can't
> simply move back 7 positions.
>
>  int i = 100;
>  dosomething(x(32,i/2),1,20 * 2,1,1,81,1);
> ---
> bipush        100
> istore_2
> bipush        32
> iload_2
> iconst_2
> idiv
> invokestatic  #6                  // Method x:(II)I
> iconst_1
> bipush        40
> iconst_1
> iconst_1
> bipush        81
> iconst_1
> invokestatic  #7                  // Method
> dosomething:(IIIIIII)Ljava/lang/String;
>
> So, does ASM hold any magic for me here? Something easier than just
> keeping track of the effects of each instruction on the stack, and walking
> backwards?
>
> Thank You,
> Jon
>
> --
> You receive this message as a subscriber of the [email protected] mailing list.
> To unsubscribe: mailto:[email protected]
> For general help: mailto:[email protected]?subject=help
> OW2 mailing lists service home page: http://www.ow2.org/wws
>


-- 
You receive this message as a subscriber of the [email protected] mailing list.
To unsubscribe: mailto:[email protected]
For general help: mailto:[email protected]?subject=help
OW2 mailing lists service home page: http://www.ow2.org/wws