Question

I am a bit confused about the behaviour of a C program from another programmer I am working now with. What I can not understand is the following:

1) a variable is defined this way

typedef float (array3d_i)[3];
array3d_i d_i[NMAX];

2) once some values are assgined to all the d_i’s, a function is called which is like this:

void calc(elem3d_i d_element);

which is called from main using:

calc(d_i[i]);

in a loop.

When the d_i’s are initialized in main, each element gets an address in memory, I guess in the stack or somewhere else. When we call the function “calc”, I would expect that inside the function, a copy of the variable is created, in anoother address. But I debugged the program, and I can see that inside the function “calc”, the variable “d_elemt” gets the same address than d_i in main.

Is it normal or not?

I am even more confused because later there is call to another function, very similar situation except that now the variables are float type and also an array of them is initialized, and inside the function, the variables are given a different address than the one in main.

How can this be? Why the difference? Is the code or the debugger doing something weird?

Thanks

Answer 1

Arrays are passed by reference in C, while simple values will be passed by value. Or, rather, arrays are also passed by value, but the “value” of an array in this context is a reference to its first element. This is the “decay” Charles refers to in his comment.

By “reference”, I mean pointer of course since C doesn’t have references like C++ does.

C does not have a higher-level array concept, which is also why you can’t compute the length of the array in the called function.