On Fri, 12 Jan 2024, Jonathan Wakely wrote:
> On Fri, 12 Jan 2024 at 17:55, Patrick Palka <ppa...@redhat.com> wrote:
> >
> > On Thu, 11 Jan 2024, Jonathan Wakely wrote:
> >
> > > I'd like to commit this to trunk for GCC 14. Please take a look.
> > >
> > > -- >8 --
> > >
> > > This is the last part of PR libstdc++/108822 implementing P2255R2, which
> > > makes it ill-formed to create a std::tuple that would bind a reference
> > > to a temporary.
> > >
> > > The dangling checks are implemented as deleted constructors for C++20
> > > and higher, and as Debug Mode static assertions in the constructor body
> > > for older standards. This is similar to the r13-6084-g916ce577ad109b
> > > changes for std::pair.
> > >
> > > As part of this change, I've reimplemented most of std::tuple for C++20,
> > > making use of concepts to replace the enable_if constraints, and using
> > > conditional explicit to avoid duplicating most constructors. We could
> > > use conditional explicit for the C++11 implementation too (with pragmas
> > > to disables the -Wc++17-extensions warnings), but that should be done as
> > > a stage 1 change for GCC 15 rather than now.
> > >
> > > The partial specialization for std::tuple<T1, T2> is no longer used for
> > > C++20 (or more precisely, for a C++20 compiler that supports concepts
> > > and conditional explicit). The additional constructors and assignment
> > > operators that take std::pair arguments have been added to the C++20
> > > implementation of the primary template, with sizeof...(_Elements)==2
> > > constraints. This avoids reimplementing all the other constructors in
> > > the std::tuple<T1, T2> partial specialization to use concepts. This way
> > > we avoid four implementations of every constructor and only have three!
> > > (The primary template has an implementation of each constructor for
> > > C++11 and another for C++20, and the tuple<T1,T2> specialization has an
> > > implementation of each for C++11, so that's three for each constructor.)
> > >
> > > In order to make the constraints more efficient on the C++20 version of
> > > the default constructor I've also added a variable template for the
> > > __is_implicitly_default_constructible trait, implemented using concepts.
> > >
> > > libstdc++-v3/ChangeLog:
> > >
> > > PR libstdc++/108822
> > > * include/std/tuple (tuple): Add checks for dangling references.
> > > Reimplement constraints and constant expressions using C++20
> > > features.
> > > * include/std/type_traits [C++20]
> > > (__is_implicitly_default_constructible_v): Define.
> > > (__is_implicitly_default_constructible): Use variable template.
> > > * testsuite/20_util/tuple/dangling_ref.cc: New test.
> > > ---
> > > libstdc++-v3/include/std/tuple | 1021 ++++++++++++-----
> > > libstdc++-v3/include/std/type_traits | 11 +
> > > .../testsuite/20_util/tuple/dangling_ref.cc | 105 ++
> > > 3 files changed, 841 insertions(+), 296 deletions(-)
> > > create mode 100644 libstdc++-v3/testsuite/20_util/tuple/dangling_ref.cc
> > >
> > > diff --git a/libstdc++-v3/include/std/tuple
> > > b/libstdc++-v3/include/std/tuple
> > > index 50e11843757..cd05b638923 100644
> > > --- a/libstdc++-v3/include/std/tuple
> > > +++ b/libstdc++-v3/include/std/tuple
> > > @@ -752,11 +752,467 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
> > > template<typename... _Elements>
> > > class tuple : public _Tuple_impl<0, _Elements...>
> > > {
> > > - typedef _Tuple_impl<0, _Elements...> _Inherited;
> > > + using _Inherited = _Tuple_impl<0, _Elements...>;
> > >
> > > template<bool _Cond>
> > > using _TCC = _TupleConstraints<_Cond, _Elements...>;
> >
> > I guess this should be moved into the #else branch if it's not used in
> > the new impl.
>
> Ah yes, I left them there until I was sure I wouldn't need them ...
> then didn't move them when I didn't need them.
>
> >
> > >
> > > +#if __cpp_concepts && __cpp_conditional_explicit // >= C++20
> > > + template<typename... _UTypes>
> > > + static consteval bool
> > > + __constructible()
> > > + {
> > > + if constexpr (sizeof...(_UTypes) == sizeof...(_Elements))
> > > + return (is_constructible_v<_Elements, _UTypes> && ...);
> >
> > IIUC this (and all the other new constraints) won't short-circuit like
> > the old versions do :/ Not sure how much that matters?
>
> Yeah, I thought about that, but we have efficient built-ins for these
> traits now, so I think it's probably OK?
Performance wise agreed, though I suppose removing the short circuiting
could break existing (though not necessarily valid) code that relied
on it to prevent an ill-formed template instantiation. It seems
the standard https://eel.is/c++draft/tuple uses conjunction_v in some
constraints, and fold-expressions in others, implying short circuiting
in some cases but not others?
>
> If not we could go back to sharing the _TupleConstraints implementations.
IMHO I'd be more comfortable with that.
