Memory Allocation for Public and Private Fields - GCC Method

Question

This is not a duplicate of this question , I have read the answers, and I still have some questions on this.

I tested several classes like this:

class A { private: int b; public: char c; int a; private: char e; }; 

And I saw that the fields are stored as if there was no access specifier, this is not so, because:

N3376 (first C ++ 11 article) 9.2 [class.mem] / 13:

Non-stationary data members of a (non-unitary) class with the same access control (section 11) are allocated in such a way that later members have higher addresses in the class object. The distribution order of non-static data elements with different access control is not defined. Implementation alignment requirements may result in two adjacent elements not being immediately distributed one after another; therefore, space may be required to manage virtual functions (10.3) and virtual base classes (10.1).

What I still do not understand is:

The distribution order of non-static data elements with different access control is not defined.

What do they mean by unspecified, GCC, of ​​course, has a way to do this, they don’t just do it randomly, I think ... They don’t want the user to know this? Or are there many ways depending on the options?

Since all the examples I tried, I had the same order as in my file (+ addition)

I am using gcc 4.9.2. Does anyone know if the GCC has indicated its way to do this?

I need to know this because I am creating a program that calculates the filling between all fields, the program works with structures that currently do not have access specifiers. I will need to find a way to do this when there are various accessibility blocks

Answer 1

Unspecified means that the compiler:

• freely make any decision he likes
• and not required for documentation.

And the means defined by the implementation, the compiler can make any decision and must document it.

---

If you are considering this class (slightly modified your version):

 class X { private: int b; int z; public: char c; int a; private: char e; public: int d; }; 

then the text from spec means:

• It is guaranteed that
  • &c < &a - both refer to the same access control.
  • &b < &z - both refer to the same access control.
• It is also guaranteed that
  • &z < &e - both refer to the same interleaved access control.
  • &a < &d - both refer to the same interleaved access control.
• it is not guaranteed that:
  • &z < &c - both refer to different access controls.
  • &a < &e - both refer to different access controls.

I saw code that uses an access specifier for each variable, so the compiler can reorder them to make the size as small as possible.