[Bug c++/71694] store-data race with bitfields and tail-padding in C++

[jason at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71694

--- Comment #4 from Jason Merrill <jason at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #3)
> The C++ FE identifies such types with
> 
>   CLASSTYPE_NON_LAYOUT_POD_P (t) || CLASSTYPE_EMPTY_P (t)
> 
> where only CLASSTYPE_NON_LAYOUT_POD_P is interesting to us.  Only available
> in struct lang_type.  Possibly able to expose that via a new langhook.

Perhaps a langhook that returns the size that is safe to access when we don't
know that we're dealing with a complete object of that type, i.e.
CLASSTYPE_SIZE for C++ classes.