RES: [Dev-C++] why do we have to rename free() ?

[Daniel K. O. <dan...@ya...>]

 >  when i checked a lot of big project's source codes, I have found them
 >  replacing basic funtion names with new function names, and
 >  sometimes with a
 >  new way of functioning for example:
 >
 >  free() -> OPENSSL_free()
 >  malloc() -> gnew(type, number of blocks)
 >
 >  how can this be useful ?, does it have a thing to do with
 >  portability ?, and
 >  if not what's the difference in functionality ?



For libraries, it can have some conveniences, like:

----------
#define g_new(struct_type, n_structs)		\
    ((struct_type *) g_malloc (((gsize) sizeof (struct_type)) * ((gsize)
(n_structs))))
----------

So you'll use it like


----------
struct foo { ... };

struct foo* many_foos = g_new(foo, 10);
----------


instead of


----------
many_foos = malloc(sizeof(foo) * 10);
or
many_foos = calloc(sizeof(foo), 10);
----------


Using macros, you can also include debuggin info, like:

---------
#ifdef DEBUG_INFO
#define my_malloc(n) \
        (alloc_bytes++,malloc(n))
#else
#define my_malloc  malloc
#endif
---------


or


---------
#ifdef DEBUG_INFO
 #define my_free my_debugged_free
#else
 #define my_free free
#endif

void my_debugged_free(void* p)
{
    if (!p) {
        puts("Hey you stupid, your're trying to free(NULL)!");
        abort();
    } else
        free(p);
}
---------


So it's easy to track programming errors.





Daniel K. O.


It's OK to be ignorant; it's not OK to play stupid.
http://www.catb.org/~esr/faqs/smart-questions.html