Question about dynamically allocated arrays in C++

[Merc248]

One thing that I thought was interesting was when my CS professor was showing us how to use GDB, where he exploits the fact that there is always a "0" right outside the bounds of a dynamically allocated integer array in order to prove a point in debugging a program. For instance:

int *array = new int [5];
int a = array[5];

And now, a will be equal to 0. My question to you guys is, why is this so in every single execution of the program? I just thought it was extremely peculiar behavior that happens too often for it to be a coincidence...

[Xerampelinae]

bren@itu:~/tmp$ cat test2.cpp
#include <iostream>
using namespace std;

int main(void)
{
int *array = new int[5];
int a = array[5];
cout << a << endl;
}

bren@itu:~/tmp$ g++ test2.cpp
bren@itu:~/tmp$ ./a.out
135145
bren@itu:~/tmp$

random junk here so ubuntuforums doesn't reject my post

[Merc248]

Weird... I actually get the same number on my end as well. I also tried array[6] and it gives 0 every time the program is executed. Same with new int[6] and array[6], it will produce 0 every execution. Why is that? Is it because it allocates the same portion of memory from the heap in every execution?

[Xerampelinae]

Weird... I actually get the same number on my end as well. I also tried array[6] and it gives 0 every time the program is executed. Same with new int[6] and array[6], it will produce 0 every execution. Why is that? Is it because it allocates the same portion of memory from the heap in every execution?

I think you're right that it takes the same memory from the heap each time. I guess you could make a for loop and repeatedly create new memory without freeing it to test..