Thread: I read about memory, stack, etc but I didn't understand something!

This is the memory model

Code:

----------------------
Code segment
"Execution code"

-----------------------
Data segment
"Global variables and constants"

-----------------------------------
Heap
"Dynamically allocated memory"
   .
   .
   .

---------------------------------
Stack
"Holds function arguments and local variables and constants"
 
                  ^
--------------------------------

Right?

Now my question is when heap is between data segment and stack, how there is no limitation in using new? We can allocate memory as much as we want (With an unlimited swap file :-P).

They're 4 separate areas of memory, there is no "between" as such. Not all systems lay out memory in the same order.

> how there is no limitation in using new?
The OS will not allocate overlapping memory, so at some point asking the OS for more memory will fail, and at that point calling new will throw an exception.

The OS will not allocate overlapping memory, so at some point asking the OS for more memory will fail, and at that point calling new will throw an exception.

But it seems it never overlaps. It makes swap file bigger and bigger until it gets to its defined limitation.

That is because memory is randomly accessable... the execution code is given a pointer where to find the variable when it wants to read/write to it. There is no defined order to things, something to keep in mind is that programs grow and shrink in the amount of memory they use... and they have to share the envirorment with other programs who the OS can decide that it wants to give a slice to, both just would write to the next avaiable slot of memory (afaik).

Then again I could have interpreted things wrong when I read up on this. (I think I read it in one of my CS books.)