Re: [PATCH] c++: get_nsdmi in template context [PR108116]

[Patrick Palka <ppalka@redhat.com>]

On Wed, 21 Dec 2022, Jason Merrill wrote:

> On 12/21/22 09:52, Patrick Palka wrote:
> > Here during ahead of time checking of C{}, we indirectly call get_nsdmi
> > for C::m from finish_compound_literal, which in turn calls
> > break_out_target_exprs for C::m's (non-templated) initializer, during
> > which we end up building a call to A::~A and checking expr_noexcept_p
> > for it (from build_vec_delete_1).  But this is all done with
> > processing_template_decl set, so the built A::~A call is templated
> > (whose form r12-6897-gdec8d0e5fa00ceb2 recently changed) which
> > expr_noexcept_p doesn't expect and we crash.
> > 
> > In r10-6183-g20afdcd3698275 we fixed a similar issue by guarding a
> > expr_noexcept_p call with !processing_template_decl, which works here
> > too.  But it seems to me since the initializer we obtain in get_nsdmi is
> > always non-templated, it should be calling break_out_target_exprs with
> > processing_template_decl cleared since otherwise the function might end
> > up mixing templated and non-templated trees.
> > 
> > I'm not sure about this though, perhaps this is not the best fix here.
> > Alternatively, when processing_template_decl we could make get_nsdmi
> > avoid calling break_out_target_exprs at all or something.  Additionally,
> > perhaps break_out_target_exprs should be a no-op more generally when
> > processing_template_decl since we shouldn't see any TARGET_EXPRs inside
> > a template?
> 
> Hmm.
> 
> Any time we would call break_out_target_exprs we're dealing with non-dependent
> expressions; if we're in a template, we're building up an initializer or a
> call that we'll soon throw away, just for the purpose of checking or type
> computation.
> 
> Furthermore, as you say, the argument is always a non-template tree, whether
> in get_nsdmi or convert_default_arg.  So having processing_template_decl
> cleared would be correct.
> 
> I don't think we can get away with not calling break_out_target_exprs at all
> in a template; if nothing else, we would lose immediate invocation expansion.
> However, we could probably skip the bot_manip tree walk, which should avoid
> the problem.
> 
> Either way we end up returning non-template trees, as we do now, and callers
> have to deal with transient CONSTRUCTORs containing such (as we do in
> massage_init_elt).

Ah I see, makes sense.

> 
> Does convert_default_arg not run into the same problem, e.g. when calling
> 
>   void g(B = {0});

In practice it seems not, because we don't call convert_default_arg
when processing_template_decl is set (verified with an assert to
that effect).  In build_over_call for example we exit early when
processing_template_decl is set, and return a templated CALL_EXPR
that doesn't include default arguments at all.  A consequence of
this is that we don't reject ahead of time a call that would use
an ill-formed dependent default argument, e.g.

  template<class T>
  void g(B = T{0});

  template<class>
  void f() {
    g<void>();
  }

since the default argument instantiation would be the responsibility
of convert_default_arg.

Thinking hypothetically here, if we do in the future want to include default
arguments in the templated form of a CALL_EXPR, we'd probably have to
instantiate them with processing_template_decl set so that the result is
templated.  And we'd subsequently want to call break_out_target_exprs on
the result also with processing_template_decl set IIUC, to perform
immediate invocation expansion.  This seems to be a potential use case
for being able to call break_out_target_exprs on templated trees, and so
unconditionally clearing p_t_d from break_out_target_exprs might not be
future proof.

In light of this, shall we go with the original approach to clear
processing_template_decl directly from get_nsdmi?

> 
> ?
> 
> > Bootstrapped and regtested on x86_64-pc-linux-gnu.
> > 
> > 	PR c++/108116
> > 
> > gcc/cp/ChangeLog:
> > 
> > 	* init.cc (get_nsdmi): Clear processing_template_decl before
> > 	processing the non-templated initializer.
> > 
> > gcc/testsuite/ChangeLog:
> > 
> > 	* g++.dg/cpp0x/nsdmi-template24.C: New test.
> > ---
> >   gcc/cp/init.cc                                |  8 ++++++-
> >   gcc/testsuite/g++.dg/cpp0x/nsdmi-template24.C | 22 +++++++++++++++++++
> >   2 files changed, 29 insertions(+), 1 deletion(-)
> >   create mode 100644 gcc/testsuite/g++.dg/cpp0x/nsdmi-template24.C
> > 
> > diff --git a/gcc/cp/init.cc b/gcc/cp/init.cc
> > index 73e6547c076..c4345ebdaea 100644
> > --- a/gcc/cp/init.cc
> > +++ b/gcc/cp/init.cc
> > @@ -561,7 +561,8 @@ perform_target_ctor (tree init)
> >     return init;
> >   }
> >   -/* Return the non-static data initializer for FIELD_DECL MEMBER.  */
> > +/* Return the non-static data initializer for FIELD_DECL MEMBER.
> > +   The initializer returned is always non-templated.  */
> >     static GTY((cache)) decl_tree_cache_map *nsdmi_inst;
> >   @@ -670,6 +671,11 @@ get_nsdmi (tree member, bool in_ctor, tsubst_flags_t
> > complain)
> >         current_class_ptr = build_address (current_class_ref);
> >       }
> >   +  /* Since INIT is always non-templated clear processing_template_decl
> > +     before processing it so that we don't interleave templated and
> > +     non-templated trees.  */
> > +  processing_template_decl_sentinel ptds;
> > +
> >     /* Strip redundant TARGET_EXPR so we don't need to remap it, and
> >        so the aggregate init code below will see a CONSTRUCTOR.  */
> >     bool simple_target = (init && SIMPLE_TARGET_EXPR_P (init));
> > diff --git a/gcc/testsuite/g++.dg/cpp0x/nsdmi-template24.C
> > b/gcc/testsuite/g++.dg/cpp0x/nsdmi-template24.C
> > new file mode 100644
> > index 00000000000..202c67d7321
> > --- /dev/null
> > +++ b/gcc/testsuite/g++.dg/cpp0x/nsdmi-template24.C
> > @@ -0,0 +1,22 @@
> > +// PR c++/108116
> > +// { dg-do compile { target c++11 } }
> > +
> > +#include <initializer_list>
> > +
> > +struct A {
> > +  A(int);
> > +  ~A();
> > +};
> > +
> > +struct B {
> > +  B(std::initializer_list<A>);
> > +};
> > +
> > +struct C {
> > +  B m{0};
> > +};
> > +
> > +template<class>
> > +void f() {
> > +  C c = C{};
> > +};
> 
>