On Tue, Sep 27, 2022 at 05:44:12PM -0400, Jason Merrill wrote:
> On 9/27/22 16:26, Marek Polacek wrote:
> > --- a/gcc/cp/typeck.cc
> > +++ b/gcc/cp/typeck.cc
> > @@ -11042,8 +11042,13 @@ check_return_expr (tree retval, bool *no_warning)
> > the conditions for the named return value optimization. */
> > bool converted = false;
> > tree moved;
> > - /* This is only interesting for class type. */
> > - if (CLASS_TYPE_P (functype)
> > + /* Until C++23, this was only interesting for class type... */
> > + if ((CLASS_TYPE_P (functype)
> > + /* ...but in C++23, we should do the below when we're converting
> > + from/to a class/reference (a non-scalar type). */
> > + || (cxx_dialect >= cxx23
> > + && (!SCALAR_TYPE_P (functype)
> > + || !SCALAR_TYPE_P (TREE_TYPE (retval)))))
>
> You might reformat this as
> (cxx_dialect < cxx23
> ? CLASS...
> : (!SCALAR...
Done, I like that better.
> > --- a/gcc/testsuite/g++.dg/cpp0x/move-return3.C
> > +++ b/gcc/testsuite/g++.dg/cpp0x/move-return3.C
> > @@ -1,6 +1,7 @@
> > // PR c++/91212
> > // Test that C++11 implicit move semantics don't call the const copy.
> > -// { dg-do link }
> > +// In C++23, we call #2.
>
> I guess that behavior is tested by elision2.C:twelve()?
Yeah, I think that's exactly the same case.
> > --- a/gcc/testsuite/g++.old-deja/g++.mike/p2846b.C
> > +++ b/gcc/testsuite/g++.old-deja/g++.mike/p2846b.C
> > @@ -1,4 +1,4 @@
> > -// { dg-do run }
> > +// { dg-do run { target c++20_down } }
> > // Shows that problem of initializing one object's secondary base from
> > // another object via a user defined copy constructor for that base,
> > // the pointer for the secondary vtable is not set after implicit
> > @@ -11,6 +11,8 @@
> > // prms-id: 2846
> > +// This test fails in C++23 due to P2266.
>
> Instead of disabling this test for C++23, let's add a cast to B& in the
> return statement.
Fixed.
> OK with that change and optionally the ?: reformatting above.
Thanks a lot; patch pushed.
Marek
