From: ahoward <ahoward@...>
Date: 2002-10-22T06:55:00+09:00
Subject: Re: rb_gc_register_address problem

On Tue, 22 Oct 2002, Paul Brannan wrote:

>   std::vector<char *> foo;
>   void func(int argc, VALUE * argv, VALUE self) {
>     foo.push_back(STR2CSTR(argv[0]));
>     rb_global_variable(&argv[0]);
>   }

can i see more of this?

> So my question is: why does rb_gc_register_address() store VALUE*'s
> instead of VALUE's?  I suppose that if it stored a VALUE, then it would
> be possible to register an object as a global but not keep a reference
> to that object; this may not be desirable.  In our case, though, we just
> wanted the string to remain valid.

i'm not _positive_, but if you did not store pointers to pointers to objects,
you would not be able to change, in the gc, what that pointerer was pointing
to!  you would only be able to change the thing pointed to itself! eg.  the gc
would never be able to set an object pointer to nil, or anywhere else...

the difference is illustrated here :

  #include <stdlib.h>
  #include <stdio.h>

  void
  i_can_change_what_memory_you_point_to (ptr)
    char **ptr;
  {
    *ptr = "barfoo";
  }

  void
  i_can_change_the_memory_you_point_to (ptr)
    char *ptr;
  {
    *ptr = 'b';
  }


  int
  main (int argc, char **argv, char **env)
  {
    char *buf = malloc(42);
    char *ptr;

    ptr = buf;
    strcpy (ptr, "foobar");
    printf("%s\n", ptr);                  /* -->> "foobar" */
    i_can_change_what_memory_you_point_to (&ptr);
    printf("%s\n", ptr);                  /* -->> "barfoo" */

    ptr = buf;
    strcpy (ptr, "foo");
    printf("%s\n", ptr);                  /* -->> "foo" */
    i_can_change_the_memory_you_point_to (ptr);
    printf("%s\n", ptr);                  /* -->> "boo"    */

    return 0;
  }


the gc must have a prototype like 'i_can_change_what_memory_you_point_to'


> If I do have a bug like this in the future, what techniques can I use to
> track it down?

gcc has built in memory leak debugging - but it only works with c.  a good
reason not to use c++ imo. ;-)

i actually have a modification, somewhere, which works with c++, want it?

-a

-- 

 ====================================
 | Ara Howard
 | NOAA Forecast Systems Laboratory
 | Information and Technology Services
 | Data Systems Group
 | R/FST 325 Broadway
 | Boulder, CO 80305-3328
 | Email: ahoward@fsl.noaa.gov
 | Phone:  303-497-7238
 | Fax:    303-497-7259
 ====================================