[Bug fortran/127873] New: With -std=... permit using DoubleComplex as derived-type name

[burnus at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=127873

            Bug ID: 127873
           Summary: With -std=... permit using DoubleComplex as
                    derived-type name
           Product: gcc
           Version: 17.0
            Status: UNCONFIRMED
          Keywords: rejects-valid
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: burnus at gcc dot gnu.org
  Target Milestone: ---

Code like
https://github.com/elwasif/HeCBench-Fortran-OMP/blob/a411f31324832d6a5001d658ab8fb3bab4c4990c/src/complex-omp-fortran/main.f90

  type :: DoubleComplex
    real(real64) :: x, y
  end type DoubleComplex

gets rejected in GCC:

src/complex-omp-fortran/main.f90:11:23:

   11 |   type :: DoubleComplex
      |                       1
Error: Type name ‘doublecomplex’ at (1) cannot be the same as an intrinsic type


ISO Fortran restriction is (only):
"C734 (R727) A derived type type-name shall not be DOUBLEPRECISION or the same
as the name of any intrinsic type defined in this document."


NOTE: doublecomplex is a common compiler extension, dating from the time
kind arguments were not really used.

Still, it should be usable with doublecomplex:

Possible solution:

* Accept it with -std=... and only error with --std=gnu...

* Always accept it - implies: 'type(doublecomplex) ::' shall not match that
  intrinsic type - while 'doublecomplex ::

Possible better is the second option?!? And we may want to update the
documentation.

Cf. also
https://flang.llvm.org/docs/Extensions.html#extensions-deletions-and-legacy-features-supported-by-default