This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: g++ and aliasing bools
- From: Daniel Berlin <dan at dberlin dot org>
- To: Mark Mitchell <mark at codesourcery dot com>
- Cc: Joe Buck <jbuck at synopsys dot COM>, Neil Booth <neil at daikokuya dot demon dot co dot uk>, Paolo Carlini <pcarlini at unitus dot it>, "gcc at gcc dot gnu dot org" <gcc at gcc dot gnu dot org>
- Date: Sun, 27 Jan 2002 02:52:37 -0500 (EST)
- Subject: Re: g++ and aliasing bools
On Sat, 26 Jan 2002, Mark Mitchell wrote:
> > The argument can be extended to handle non-virtual single or multiple
> > inheritance: from the C point of view,
>
> This isn't obvious yet.
>
> You have to at least discuss zero-sized base classes and whether or not
> GNU C handles them in the same way -- including cases like this:
>
> struct A {};
> struct B : public A {};
> struct C : public A {};
> struct D : public B, C {};
>
> Here, D has size two to avoid having two A's at the same address.
> If C did not derive from A, D would have size one. In GNU C, does:
>
> struct A {};
> struct B { struct A __base1; };
> struct C { struct A __base1; };
> struct D {
> struct B __base1;
> struct C __base2;
> };
>
> have size two?
>
No, it has size 0.
> Does the variant where C is empty have size 1?
Nope.
However, this will just cause us to be more conservative than necessary.
We'll think they have the same address, and thus, alias, rather than say
they *don't* alias.
Right?
--Dan