Should I use T foo[X] instead of T *foo if x is know at compile-time?

Question

Assuming this:

struct A
{
    char *a;
        char *b;
}

If I know size of a and b at compile-time is worth effort if I do use a fixed-array-size instead of or will it change nothing? it will save a lot of malloc calls, much of them in a loop and I think it's fastest.

struct A
{
    char a[256];
    char b[32];
};

Answer 1

If you can do everything with the second option that you want to do, then it's a better option. You should avoid dynamic memory allocation unless it's necessary. There are some reasons to go with the first option. The obvious one is that you might have a variable size, but you've indicated that that isn't the case. Here are some other reasons:

1. You might want to reassign a and b, which can be done with pointers but not arrays.
2. You might want to be able to efficiently move an A object.