[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: realloc question
From:       Szymon Stefanek <stefanek () tin ! it>
Date:       1999-01-02 2:37:09
[Download RAW message or body]

On Wed, 30 Dec 1998, you wrote:
>On Wed, 30 Dec 1998, Szymon Stefanek wrote:
>
>> Hi all!
>> 
>> I think that this is a good place to ask :)
>> 
>> char * buffer_ptr = malloc(10);
>> memmove(buffer_ptr,"1234567890",10);
>
>You are aware of the fact that buffer_ptr doesn't point to a real string
>(not /0 terminated), are you?

:)

>
>> char * pointer_to_five = buffer_ptr+4;
>> buffer_ptr = realloc(20);
>> 
>> now buffer_ptr points to the beginning of a 20 bytes buffer like this:
>> "1234567890**********"
>> 
>> but where points 'pointer_to_five' ? :)
>
>pointer_to_five _might_ point to "567890**********". It might.

It might ?
So it is NOT sure that it will point there ? :)

>
>If you increase a buffer by using realloc() two different scenarios are
>possible (besides 'out of memory' of course):
>1.
>There is sufficient memory available directly behind the old buffer. In
>this case realloc() indeed extends the old buffer and pointer_to_five
>is valid.
>2.
>There isn't sufficient memory available directly behind the old buffer.
>In this case realloc() allocates a completely new buffer of the new size
>somewhere else, copies the content of the old buffer to the new one, and
>deallocates the old buffer. Your pointer_to_five points to a portion of
>memory which is not in use.  :-(
>

Thanks :)

Szymon Stefanek

[prev in list] [next in list] [prev in thread] [next in thread]