On Thu, Nov 22, 2018 at 10:04:26AM +0200, Ville Voutilainen wrote:
> On Tue, Nov 20, 2018 at 04:59:46PM -0500, Jason Merrill wrote:
> > On 11/19/18 5:12 PM, Marek Polacek wrote:
> >> > + /* Don't forget that the innermost namespace might have been
> >> > + marked as inline. */
> >> > + is_inline |= nested_inline_p;
> >> This looks wrong: an inline namespace does not make its nested namespaces
> >> inline as well.
>
> >A nested namespace definition cannot be inline. This is supposed to handle
> >cases such as
> >namespace A::B::inline C { ... }
> >because after 'C' we don't see :: so it breaks and we call push_namespace
> >outside the for loop. So I still don't see a bug; do you have a test that
> >I got wrong?
>
> The way I read the question is "what does
>
> namespace A::inline B::C::D {...}
>
> do?".
Thanks. This case is still handled correctly; I just checked the
push_namespace calls in gdb and just the one for B is with make_inline=true.
Arguably I should've added tests testing that some of the namespaces *aren't*
inline.
> C and D are not inline. For what it's worth, I had an earlier very incomplete
> stab at it, haven't looked how complete it really was; I know that it didn't
> handle diagnostics as well as yours, and I have no recollection
> of whether it handles the cases like the above. See attached.
I wasn't aware you had worked on this. Perhaps we should track the progress of
C++20 features in Bugzilla (to keep track of who's working on what).
Marek
