[Bug libstdc++/108846] std::copy, std::copy_n and std::copy_backward on potentially overlapping subobjects

["redi at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108846

--- Comment #24 from Jonathan Wakely <redi at gcc dot gnu.org> ---
(In reply to Arthur O'Dwyer from comment #23)
> - There may be padding in the middle of an object, and I'm not confident
> that the Standard actually forbids it from being used. Of course your
> approach works fine on the Itanium ABI, and probably works everywhere that
> actually exists. If you've got chapter+verse proving that it must work
> *everywhere*, I'd appreciate seeing it, just for my own information.

I don't care about everywhere, only GCC-like compilers on the targets they
support.

> - If GCC were ever to add a builtin for this notion, IMO the proper name
> would be `__datasizeof(T)`. See
> https://danakj.github.io/2023/01/15/trivially-relocatable.html#data-size

I think it should have a __builtin_ prefix unless it directly implements a
standard trait like __is_object.

> - You can implement your library trait like this; see
> https://quuxplusone.github.io/blog/2023/03/04/trivial-swap-prize-update/
> #step-1.-std-swap-can-t-assume-it
> 
>     template <class _Tp>
>     struct __libcpp_datasizeof {
>         struct _Up {
>             [[__no_unique_address__]] _Tp __t_;
>             char __c_;
>         };
>         static const size_t value = __builtin_offsetof(_Up, __c_);
>     };
> 
> Unfortunately it looks like GCC doesn't support
> `__attribute__((__no_unique_address__))` in C++03 mode. (Neither does Clang.
> What is up with that!)

Which is OK, because you can't have final types in C++03 either, so testing it
using inheritance is good enough for C++03.

> Your suggested trait implementation is slightly wrong for `is_final` types:
> you return 0 but really a final type _can_ have usable tail padding. See
> https://godbolt.org/z/P6x459MEq

Ah, I didn't think that padding would actually be used by an adjacent member
subobject. But of course it can be for an empty class, where it consists of a
single byte of padding which can be reused by another member of a different
type.