Pointer Arithmetic

Definition

Pointer arithmetic allows us to do literal mathematical operations with pointer addresses, which allows us to do very interesting things.

This is very important when working with Arrays, but is used outside of that as well.

How it works

When you have a pointer T *p where T is the type and int i

p + i has the effect:

• TYPE(p+i) = pointer-to-T
• VAL(p+i) = VAL(p) + SIZE(T) * VAL(i)

So basically, the language moves the pointer i elements of type T over. So for an int type, we’d be moving over 4 bytes every time we do + 1 on the pointer to the int.

In terms of Arrays, s* = 'h' is equivalent to s[0] = 'h'

So, *(s+2) = *(s+4) is the same as s[2] = s[4]

Arithmetic with Casted Pointers

Recall Casting Pointers changes purely the type, not value. We can use this to our advantage.

So, given char s[8] = "Hello!\n;

• The expression (int *)s + 7 has type pointer to int and so the value is ADDR(s) + SIZE(int) * 7 (in this case, that turns out to be ADDR(s)+28)
• The expression (int *)(s + 7) has type pointer to int and so the value is ADDR(s) + SIZE(char) * 7 (in this case, that turns out to be ADDR(s) + 7)
• The statement int x = *((int *) s); will interpret the four bytes starting at address s as an int and set x equal to this value
  • It will just copy these four bytes over
  • The resulting value of x depends on the endianness of the machine