From: Stephen Kellett <snail@...>
Date: 2004-11-30T23:02:52+09:00
Subject: Re: How to implement set_trace_func callback in C/C++?

In message <1101771510.124768.9696.nullmailer@x31.priv.netlab.jp>,
Yukihiro Matsumoto <matz@ruby-lang.org> writes
>Try:
>
>  VALUE my_trace_function(int argc, VALUE *argv);

Thanks that works just fine. I've managed to decode 5 of the 6
parameters that can potentially be passed in. I know that sometimes I
don't get all 6 parameters. I'm having a problem decoding the "binding"
parameter. How do I do this? I'd like to be able to determine which
thread and what the local variables are and what the parameters are.

I thought the binding value (argv[4]) was really a struct BLOCK *, where
BLOCK is taken from eval.c

struct BLOCK {
        NODE *var;
        NODE *body;
        VALUE self;
        struct FRAME frame;
        struct SCOPE *scope;
        VALUE klass;
        NODE *cref;
        int iter;
        int vmode;
        int flags;
        struct RVarmap *dyna_vars;
        VALUE orig_thread;
        VALUE wrapper;
        VALUE block_obj;
        struct BLOCK *outer;
        struct BLOCK *prev;
};

However if I process argv[4] as a struct BLOCK * I find that the
"orig_thread" value varies wildly, and other values don't seem to make
much sense.

Here is my code, where v is the VALUE from argv[4]

        if (v != Qnil)
        {
                struct BLOCK    *blk;

                blk = (struct BLOCK *)v;
                printf(" binding:%d\n", v);
                printf("\tthread:%d\n", blk->orig_thread);
                printf("\tself:%d\n", blk->self);
                printf("\tframe:%d\n", blk->frame.self);
                printf("\tscope:%d\n", blk->scope);
        }
        else
        {
                printf(" binding:NIL");
        }

How should I be interpreting the binding parameter? I want to avoid
calling eval() and deal with values directly.

Hoping you can help.

Stephen
-- 
Stephen Kellett
Object Media Limited    http://www.objmedia.demon.co.uk
RSI Information:        http://www.objmedia.demon.co.uk/rsi.html