Re: [AD] memory leak in readbmp.c?

[Laurence Withers <lwithers@xxxxxxxxxx>]

In reply to Peter Wang <tjaden@xxxxxxxxxx>:
[snip - readbmp.c]
>> Nope; the destructor frees the memory allocated by 
>> _register_bitmap_file_type_init().
>
>Ok, I'll take your word for it.
>
>What about this?
>$ mtrace examples/exalpha mtrace.log
>- 0x080ebb10 Free 2474 was never alloc'd /home/tjaden/cvs/allegro/./src/readbmp.
>c:166
[snip some more]

Looks like `mtrace', whatever it is, doesn't record memory allocated in
constructor functions. The readbmp.c free() calls are all cleaning up
the linked list (check out the ctor to see that the memory is actually
allocated), and the allegro.c free() call is removing the exit function
(which was registered in the ctor).

Perhaps `mtrace' uses a ctor function to register itself? If that is the
case, the order of ctor calls is not guaranteed (the only thing we can
take for granted is that simple variables -- those without a ctor --
will be initialised first. This lets us build linked lists from global
pointers, etc.)

Bye for now,
-- 
Laurence Withers, lwithers@xxxxxxxxxx
                http://www.lwithers.demon.co.uk/

Attachment: signature.asc
Description: PGP signature