[Bug c++/102184] New: Explicit template instantiation is wrongly considered as specialization

[Bug c++/102184] New: Explicit template instantiation is wrongly considered as specialization

[fchelnokov at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102184

            Bug ID: 102184
           Summary: Explicit template instantiation is wrongly considered
                    as specialization
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: fchelnokov at gmail dot com
  Target Milestone: ---

In this code:
```
template<int N> requires(N==1) int fib() { return 1; }
template<int N> requires(N==2) int fib() { return 1; }
template<int N> int fib() { return fib<N-1>() + fib<N-2>(); }
template int fib<1>();
```
the last line with function explicit instantiation produces GCC error:
ambiguous template specialization 'fib<1>' for 'int fib()'

The other compilers are fine with the code: https://gcc.godbolt.org/z/fbMGs3hqY

[Bug c++/102184] Explicit template instantiation is wrongly considered as specialization

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102184

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |diagnostic

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
GCC is saying there is an ambigous specialization.

If I do this, GCC accepts the code:
template<int N> requires(N==1) int fib() { return 1; }
template<int N> requires(N==2) int fib() { return 1; }
template<int N> int fib() requires(N!=1 && N!=2)  { return fib<N-1>() +
fib<N-2>(); }
template int fib<1>();

[Bug c++/102184] Explicit template instantiation thinks there is an ambiguous template specialization when there is none for concepts

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102184

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2021-09-03
           Keywords|                            |rejects-valid
            Summary|Explicit template           |Explicit template
                   |instantiation is wrongly    |instantiation thinks there
                   |considered as               |is an ambiguous template
                   |specialization              |specialization when there
                   |                            |is none for concepts

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Andrew Pinski from comment #1)
> GCC is saying there is an ambigous specialization.
> 
> If I do this, GCC accepts the code:
> template<int N> requires(N==1) int fib() { return 1; }
> template<int N> requires(N==2) int fib() { return 1; }
> template<int N> int fib() requires(N!=1 && N!=2)  { return fib<N-1>() +
> fib<N-2>(); }
> template int fib<1>();

But I think GCC does not know which specialization to instantiate here.
Because the following works just fine:
int t = fib<1>();