Re: RFC: attribute "unpadded"

[Mark Mitchell <mark at codesourcery dot com>]

--On Tuesday, August 27, 2002 03:10:00 AM +0100 Jamie Lokier 
<egcs@tantalophile.demon.co.uk> wrote:

> Mark Mitchell wrote:
> > This comes up in C++ like so:
> >
> >   struct A { virtual void f();  char c; };
> >   struct B : public A { char c2 };
> Here in C++, `sizeof(struct A)' is 8, and `sizeof(struct B)' is also 8.
>
> > Here, the size of B is 8, not 12, because "c2" is packed into
> > the tail padding of A.  If A were "unpadded", you could express
> > this in C as:
> >
> >   struct A __attribute__((unpadded)) { void *vptr; char c; };
> >   struct B { struct A __base; char c2; };
> This does _not_ express the same thing in C, because this C version of
> `sizeof(struct A)' is 5.
You're correct -- but that's not how you would map the C++ ABI onto C.

In C, you would need two versions of each class type; the type when it
is used as a base class, and the type when it is a complete object.

So, you would do this:

 struct A { void *vptr; char c; };
 struct A_as_base __attribute__((unpadded)) { void *vptr; char c; };
 struct B { struct A_as_base __base; char c2; };

Then everything matches up.

You need to have two versions of each class because the base version
needs to omit virtual base classes.

--
Mark Mitchell                mark@codesourcery.com
CodeSourcery, LLC            http://www.codesourcery.com