Inspired this quiestion .
Why does -
void display(int p[][SIZE]) // allowed
and
void display(int p[][]) // not a
That's because caculations in pointers requires that.
When p points to 2 dimentional array that's size is SIZE
this operation p+=3
is calculated this way:
if the value of p is a constant equal ADRESS p receive this value ADRESS + 3*SIZE.
That's because arithmitic in pointers is different from arithmitic in real numbers. This calculation can't be done without knowing the size of the array.
Because arrays decay to pointers when passed to a function. If you do not provide the cardinality of the second dimension of the array, the compiler would not know how to dereference this pointer.
Here is a longer explanation: when you write this
p[index]
the compiler performs some pointer arithmetic to find the address of the element that it needs to reference: it multiplies index
by the size of p
's element, and adds it to the base address of p
:
address = <base address of p> + index * <size of p's element>
When you try passing an array like this, p[][]
, the compiler knows only the base address of p
, but not the size of its element. It is in order for the compiler to know the size of p
's element that you need to provide the cardinality of the second dimension.
Because a 2D array is stored row wise and hence the function needs the number of columns so that it knows when the next row begins.