Re: [PATCH] c++: ICE with noexcept and canonical types [PR101715]

[Marek Polacek <polacek@redhat.com>]

On Sat, Jan 15, 2022 at 09:24:05AM -0500, Patrick Palka wrote:
> On Fri, 14 Jan 2022, Marek Polacek via Gcc-patches wrote:
> 
> > This is a "canonical types differ for identical types" ICE, which started
> > with r11-4682.  It's a bit tricky to explain.  Consider:
> > 
> >   template <typename T> struct S {
> >     S<T> bar() noexcept(T::value);  // #1
> >     S<T> foo() noexcept(T::value);  // #2
> >   };
> > 
> >   template <typename T> S<T> S<T>::foo() noexcept(T::value) {}  // #3
> > 
> > We ICE because #3 and #2 have the same type, but their canonical types
> > differ: TYPE_CANONICAL (#3) == #2 but TYPE_CANONICAL (#2) == #1.
> > 
> > The member functions #1 and #2 have the same type.  However, since their
> > noexcept-specifier is deferred, when parsing them, we create a variant for
> > both of them, because DEFERRED_PARSE cannot be compared.  In other words,
> > build_cp_fntype_variant's
> > 
> >   tree v = TYPE_MAIN_VARIANT (type);
> >   for (; v; v = TYPE_NEXT_VARIANT (v))
> >     if (cp_check_qualified_type (v, type, type_quals, rqual, raises, late))
> >       return v;
> > 
> > will *not* find an existing variant when creating a method_type for #2, so we
> > have to create a new one.
> > 
> > But then we perform delayed parsing and call fixup_deferred_exception_variants
> > for #1 and #2.  f_d_e_v will replace TYPE_RAISES_EXCEPTIONS with the newly
> > parsed noexcept-specifier.  It also sets TYPE_CANONICAL (#2) to #1.  Both
> > noexcepts turned out to be the same, so now we have two equivalent variants in
> > the list!  I.e.,
> > 
> > +-----------------+      +-----------------+      +-----------------+
> > |      main       |      |      #2         |      |      #1         |
> > | S S::<T379>(S*) |----->| S S::<T37c>(S*) |----->| S S::<T37a>(S*) |----->NULL
> > |    -            |      |  noex(T::value) |      |  noex(T::value) |
> > +-----------------+      +-----------------+      +-----------------+
> > 
> > Then we get to #3.  As for #1 and #2, grokdeclarator calls build_memfn_type,
> > which ends up calling build_cp_fntype_variant, which will use the loop
> > above to look for an existing variant.  The first one that matches
> > cp_check_qualified_type will be used, so we use #2 rather than #1, and the
> > TYPE_CANONICAL mismatch follows.  Hopefully that makes sense.
> > 
> > As for the fix, I didn't think I could rewrite the method_type #2 with #1
> > because the type may have escaped via decltype.  So my approach is to
> > elide #2 from the list, so when looking for a matching variant, we always
> > find #1 (#2 remains live though, which admittedly sounds sort of dodgy).
> 
> I wonder about instead making build_cp_fntype_variant set the TYPE_CANONICAL for
> #3 to TYPE_CANONICAL(#2) (i.e. #1) instead of to #2?  Something like:
> 
> -- >8 --
> 
>  gcc/cp/tree.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/gcc/cp/tree.c b/gcc/cp/tree.c
> index 7f7de86b4e8..b89135fa121 100644
> --- a/gcc/cp/tree.c
> +++ b/gcc/cp/tree.c
> @@ -2779,8 +2779,9 @@ build_cp_fntype_variant (tree type, cp_ref_qualifier rqual,
>    else if (TYPE_CANONICAL (type) != type || cr != raises || late)
>      /* Build the underlying canonical type, since it is different
>         from TYPE. */
> -    TYPE_CANONICAL (v) = build_cp_fntype_variant (TYPE_CANONICAL (type),
> -						  rqual, cr, false);
> +    TYPE_CANONICAL (v)
> +      = TYPE_CANONICAL (build_cp_fntype_variant (TYPE_CANONICAL (type),
> +						 rqual, cr, false));
>    else
>      /* T is its own canonical type. */
>      TYPE_CANONICAL (v) = v;

Thanks for looking.  I can dig that (and verified it works), but it strikes
me more as a workaround for the duplicity problem.  I also don't see
TYPE_CANONICAL (...) = TYPE_CANONICAL (build_cp_fntype_variant (...))
anywhere in the codebase, if that means anything.

Marek